Sudden Death in Epilepsy: Forensic and Clinical Issues

SUDDEN DEATH in EPILEPSY FORENSIC AND CLINICAL ISSUES

SUDDEN DEATH in EPILEPSY FORENSIC AND CLINICAL ISSUES EDITED BY

CLAIRE M. LATHERS PAUL L. SCHR AEDER MICHAEL W. BUNGO JAN E. LEESTMA

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2011 by Taylor and Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-13: 978-1-4398-0223-6 (Ebook-PDF) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http:// www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

To (DS) a 4-year old boy with a history of nocturnal seizures who found hidden Christmas presents, including the one he wanted most of all, a remote control car two weeks before Christmas. The next morning he was found dead in bed. He and other victims of SUDEP challenge all of us to find preventive measures as quickly as possible. (CML) To Carol and Richard Lathers and to Marcel J. Lajoy for their love, support, and words of wisdom while completing this endeavor. (CML) To two university students (MB) and (AG) who were victims of SUDEP and to their courageous families who helped to found Epilepsy Bereaved in the United Kingdom. (PLS) To my patient wife Barbara and daughters Maria and Ellen who dealt with my many hours of editorial distraction with forbearance, encouragement and love. (PLS) To my grandson, Lucas, who missed quality time, so that this endeavor could be completed. (MWB)

Table of Contents

Foreword by Samuels Foreword by Schachter Preface Acknowledgments Editors Contributors

xv xvii xxi xxv xxvii xxxi

Section I Forensics of Sudden Death

1

Neurocardiologic Mechanistic Risk Factors in Sudden Unexpected Death in Epilepsy

3

Claire M. Lathers, Paul L. Schraeder, and Michael W. Bungo

2

Forensic Considerations and Sudden Unexpected Death in Epilepsy

37

Jan E. Leestma

3

Omega-3 Fatty Acids in Sudden Unexpected Death in€Epilepsy: Guardian of the Brain–Heart Connection

57

Fulvio A. Scorza, Esper A. Cavalheiro, Ricardo M. Arida, Vera C. Terra, Carla A. Scorza, Eliza Y. F. Sonoda, and Roberta M. Cysneiros

4

Unanswered Questions: SUDEP Studies Needed

67

Claire M. Lathers, Paul L. Schraeder, and Michael W. Bungo

5

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome: Implications in SUDEP David S. Paterson

vii

77

viii Table of Contents

6

Forensic Case Identification

95

Paul L. Schraeder, Elson L. So, and Claire M. Lathers

7

Sudden Unexpected Death in Epilepsy: Future Research Directions

109

Simona Parvulescu-Codrea

8

Forensic Postmortem Examination of Victims of Sudden Unexpected Death in Epilepsy

131

Claire M. Lathers, Paul L. Schraeder, Steven A. Koehler, and Cyril H. Wecht

9

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

145

Steven A. Koehler, Paul L. Schraeder, Claire M. Lathers, and Cyril H. Wecht

10

Drug Abuse and SUDEP

159

Steven B. Karch

11

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

169

Claire M. Lathers, Michelle M. Spino, Isha Agarwal, Laurie S. Y. Tyau, and Wallace B. Pickworth

12

Risk Factors for Sudden Death in Epilepsy

187

Thaddeus S. Walczak

13

EEG Findings in SUDEP

201

Maromi Nei and Nicole Simpkins

14

Severity of Seizures as a Forensic Risk and Case Reports

209

Edward H. Maa, Michael P. Earnest, Mark C. Spitz, and Jacquelyn Bainbridge

15

Intractable Epilepsy in the Setting of Malformations of Cortical Development as a Mechanism for SUDEP

221

Lara Jehi and Imad Najm

16

Neurogenic Cardiac Arrhythmias Howan Leung and Anne Y. Y. Chan

235

Table of Contents

17

Stress and SUDEP

ix

253

Claire M. Lathers and Paul L. Schraeder

18

Genetics of Sudden Death in Epilepsy

267

Neeti Ghali and Lina Nashef

19

Cardiac Channelopathies and Sudden Death

285

Benito Herreros

20

Sodium Channel Dysfunction: Common Physiopathologic Mechanism Associated with Sudden Death ECG Abnormalities in Brugada Syndrome and Some Types of€Epilepsy: Case Histories

303

Claire M. Lathers, Paul L. Schraeder, and Michael W. Bungo

21

Not Seizure but Syncope

311

Saumya Sharma, Trieu Ho, and Bharat K. Kantharia

22

Syncope, Seizures, and SUDEP: Case Histories

325

Claire M. Lathers, Paul L. Schraeder, and Michael W. Bungo

23

Sudden Death in Epilepsy: Relationship to the Sleep–Wake Circadian Cycle and Fractal Physiology

333

John D. Hughes and Susumu Sato

24

SUDEP: Medicolegal and Clinical Experiences

347

Braxton B. Wannamaker

Section II SUDEP Animal Models: MECHANISMS OF RISKS

25

Sudden Death: Animal Models to Study Nervous System Sites of Action for Disease and Pharmacological Intervention Claire M. Lathers

363

x Table of Contents

26

Synaptic Plasticity of Autonomic Ganglia: Role of Chronic Stress and Implication in Cardiovascular Diseases and Sudden Death

395

Karim A. Alkadhi and Karem H. Alzoubi

27

Animal Model for Sudden Cardiac Death: Autonomic Cardiac Sympathetic Nonuniform Neural Discharge

427

Claire M. Lathers

28

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

437

Claire M. Lathers and Paul L. Schraeder

29

A Characterization of the Lockstep Phenomenon in Phenobarbital-Pretreated Cats

465

Jeffrey M. Dodd-O and Claire M. Lathers

30

Relationship of the Lockstep Phenomenon and Precipitous Changes in Blood Pressure

481

Amy Z. Stauffer, Jeffrey M. Dodd-O, and Claire M. Lathers

31

Interspike Interval Histogram Characterization of€Synchronized Cardiac Sympathetic Neural Discharge and€Epileptogenic Activity in the Electrocorticogram of€the€Cat

495

Daniel K. O’Rourke and Claire M. Lathers

32

Power Spectral Analysis: A Procedure for Assessing Autonomic Activity Related to€Risk Factors for Sudden and Unexplained Death in€Epilepsy

513

Animal Model for Sudden Cardiac Death: Sympathetic Innervation and Myocardial Beta-Receptor Densities

539

Stephen R. Quint, John A. Messenheimer, and Michael B. Tennison

33

Claire M. Lathers and Robert M. Levin

34

Antiepileptic Activity of Beta-Blocking Agents

551

Claire M. Lathers, Kam F. Jim, William H. Spivey, Claire Kahn, Kathleen Dolce, and William D. Matthews

Table of Contents

35

Arrhythmias Associated with Epileptogenic Activity Elicited by Penicillin

xi

567

Claire M. Lathers and Paul L. Schraeder

36

Role of Neuropeptides in the Production of Epileptogenic€Activity and Arrhythmias

577

Claire M. Lathers

37

Sudden Epileptic Death in Experimental Animal Models

591

Ombretta Mameli and Marcello Alessandro Caria

38

Sympathetic Nervous System Dysregulation of Cardiac Function and Myocyte Potassium Channel Remodeling in Rodent Seizure Models: Candidate Mechanisms for SUDEP 615 Steven L. Bealer, Cameron S. Metcalf, Jason G. Little, Matteo Vatta, Amy Brewster, and Anne E. Anderson

39

The Urethane/Kainate Seizure Model as a Tool to Explore€Physiology and Death Associated with Seizures

627

Mark Stewart

40

Acute Cardiovascular Response during Kindled Seizures

645

Jeffrey H. Goodman, Richard W. Homan, and IsAac L. Crawford

41

DBA Mice as Models of Sudden Unexpected Death in€Epilepsy

659

Carl L. Faingold, Srinivasan Tupal, Yashanad Mhaskar, and Victor V. Uteshev

Section III Clinical Issues of Sudden Death

42

Cardiac and Pulmonary Risk Factors and Pathomechanisms€of Sudden Unexplained Death in Epilepsy€Patients Josef Finsterer and Claudia Stöllberger

679

xii Table of Contents

43

Neurocardiac Interactions in Sudden Unexpected Death in€Epilepsy: Can Ambulatory Electrocardiogram-Based Assessment of Autonomic Function and T-Wave Alternans Help to Evaluate Risk?

693

Richard L. Verrier and Steven C. Schachter

44

Arrhythmogenic, Respiratory, and Psychological Risk Factors for Sudden Unexpected Death and Epilepsy: Case Histories

711

Claire M. Lathers

45

Sudden Arrhythmic Death Syndrome: Underlying Cardiac Etiologies, Their Implications, and the Overlap with SUDEP

721

Paramdeep S. Dhillon and Elijah R. Behr

46

Odds Ratios Study of Antiepileptic Drugs: A Possible Approach to SUDEP Prevention?

743

Claire M. Lathers, Paul L. Schraeder, and H. Gregg Claycamp

47

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology: Possible Role in Cause and/or Prevention of SUDEP

755

Claire M. Lathers and Paul L. Schraeder

48

Clinical Pharmacology and SUDEP

789

Claire M. Lathers and Paul L. Schraeder

49

Experience-Based Teaching of Therapeutics and Clinical€Pharmacology of Antiepileptic Drugs: Sudden Unexplained Death in Epilepsy: Do Antiepileptic Drugs Have a Role?

801

Claire M. Lathers and Paul L. Schraeder

50

Clinical Pharmacology of Antiepileptic Drug Use: Clinical Pearls about the Perils of Patty

827

Paul L. Schraeder and Claire M. Lathers

51

Compliance with Antiepileptic Drug Treatment and the€Risk of Sudden Unexpected Death in Epilepsy Torbjörn Tomson

845

Table of Contents

52

SUDEP Clinical Case Histories: Typical and Atypical

xiii

853

Paul L. Schraeder

53

Cardiac Antiarrhythmic Agents: Pharmacological Basis for Their Antiarrhythmic and Proarrhythmic Effects

861

Saumya Sharma, Trieu Ho, and Bharat K. Kantharia

54

Could Beta–Blocker Antiarrhythmic and Antiseizure Activity Help Prevent SUDEP?

877

Claire M. Lathers

55

Decision Analysis and Risk Management

887

H. Gregg Claycamp

56

Epilepsy Surgery and the Prevention of SUDEP

905

Ryan S. Hays and Michael R. Sperling

57

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP

915

Jane Hanna and Rosemary Panelli

58

Bereavement and Sudden Unexpected Death in Epilepsy

937

Lina Nashef and Lene Sahlholdt

59

SUDEP: A Clinical and Communicative Conundrum

943

Paul L. Schraeder and Claire M. Lathers

60

Epilepsy and SUDEP: Lessons Learned: Scientific and Clinical Experience

953

Claire M. Lathers and Paul L. Schraeder

61

SUDEP: A Mystery Yet to Be Solved

967

Claire M. Lathers and Paul L. Schraeder

62

Forensic Evidence and Expert Witnesses: Scientific Evidence: Getting It in and Keeping It Out

973

Thomas L. Bohan

Index

983

Foreword by Samuels

Epilepsy is one of the most prevalent neurological diseases in the world. Known for ages and clearly described in some of the oldest medical treatises, it was greatly feared in the ancient world because of the belief that it could cause sudden death. As reasonable treatments arose—ἀrst bromides, then phenobarbital, and, in the mid-twentieth century, phenytoin, the ἀrst modern antiepileptic drug—the mainstream of the medical community came to believe that epileptic seizures themselves were relatively harmless. During my own training in internal medicine in the 1970s, it was widely taught that the best treatment for a convulsion was to simply place the patient in a safe environment and let the seizure take its course. With the development of newer, safer treatments that include drugs and nonpharmacological approaches such as surgery and, more recently, vagal and deep brain stimulation, it became progressively more apparent that the ancients were, in fact, correct. Epilepsy, a state of recurrent unprovoked seizures, itself indeed carried with it an increased risk for sudden unexpected death and the term SUDEP (sudden unexpected death in epilepsy) was born. SUDEP now joins a long list of sudden death syndromes, including sudden death in middle aged men, sudden unexpected nocturnal death syndrome (SUNDS), sudden death from fright, sudden infant death syndrome (SIDS), sudden death in young athletes, sudden death associated with drug use, sudden death from heart disease, sudden death during sedative drug (including alcohol) withdrawal, sudden death during delirium, sudden death from stroke (including subarachnoid hemorrhage), and sudden death from head injury. With the advent of long-term monitoring of various physiological parameters, including the electrocardiogram, arterial oxygen saturation, and the electroencephalogram, it became apparent that some of these sudden death syndromes had in common a capacity to produce malignant cardiac arrhythmias, respiratory arrest, or both. In his landmark paper of 1942, “Voodoo” Death, the eminent physiologist Walter B. Cannon recounted stories of sudden death from the anthropological literature and posited an autonomic storm as the unifying hypothesis. In the past half-century, much has been learned about the capacity of the brain to damage the visceral organs, but many features of SUDEP remain an enigma. Despite all of the advances in the diagnosis and treatment of seizure disorders, the threat of sudden death still hovers over the epileptic patient, much as it did in ancient times. In Sudden Death in Epilepsy: Forensic and Clinical Issues, four of the most eminent experts in SUDEP, Claire Lathers, Paul Schraeder, Michael Bungo, and Jan Leestma, have put together an impressive tome representing the state of this art and science. The book’s three sections, Forensics of Sudden Death, SUDEP Animal Models: Mechanisms of Risk, and Clinical Issues in Sudden Death, are written by a veritable who’s who in the ἀeld. The interested reader can ἀnd chapters on the history, diagnosis, phenomenology, mechanisms, genetics, animal models, pathology, physiology, and even the social repercussions of this devastating phenomenon. Books written by so many authors inevitably create a challenge with regard to organization and thematic coherence, but the four editors, themselves xv

xvi

Foreword by Samuels

each major contributors to the literature on SUDEP, have done an admirable job pulling together the disparate array of experts into a volume that holds together and is readable. Sudden Death in Epilepsy: Forensic and Clinical Issues should be of great interest to neurologists, psychiatrists, neurosurgeons, internists, cardiologists, neuroscientists, and cardiovascular specialists. Epileptologists and their trainees will ἀnd the content invaluable in their day-to-day lives of counseling and treating people with epilepsy. Taken as a whole, the content acts as a roadmap to those who hope to someday fully understand and prevent this dramatic and tragic event. Martin A. Samuels, MD, FAAN, MACP, DSc (hon) Chairman, Department of Neurology Brigham and Women’s Hospital and Professor of Neurology Harvard Medical School Boston, Massachusetts

Foreword by Schachter

If sudden death in epilepsy is the most feared and serious consequence of epilepsy, why is it seldom discussed and woefully under-researched? The editors point to many reasons for this in the Preface—inadequate animal models and basic understanding, lack of clinical recognition from treating physicians and medical examiners, inability to eliminate the possibility of sudden death in the nearly one in three patients with epilepsy whose seizures are drug-resistant, and continuing reluctance among physicians to discuss sudden death with patients and their families. Compounding these issues is the silo-style infrastructure of academic medicine, which creates intrinsic barriers to establishing clinical and research collaborations across relevant disciplines, such as epidemiology, neurology, cardiology and pulmonology, as well as between physicians and applied scientists, including electrical, mechanical, and computer engineers. Despite these and other problems, there are reasons to be optimistic that research will begin to solve the mysteries posed by Drs. Lathers and Schraeder at the end of this book. First, creative and passionate researchers are dedicated to eradicating sudden death in epilepsy, perhaps most of all the editors of this volume, who have put together the most comprehensive and current treatise on the topic. The like-minded authors span numerous disciplines and their chapters challenge current paradigms and suggest new ways of thinking about sudden death in epilepsy. Second, the National Institute of Neurological Disorders and Stroke (NINDS) is actively engaged. Curing Epilepsy 2007: Translating Discoveries into Therapies, organized by the NINDS, and attended by more than 400 researchers, health care professionals, patients, and family members, affirmed sudden death in epilepsy as a major target for research, with the development and validation of at least one prevention strategy to decrease its occurrence as a short-term goal, and identiἀcation of the responsible mechanisms, including effects of seizures on autonomic functioning, particularly cardiac and respiratory, as a longer-term goal. The NINDS also sponsored a workshop on sudden death in epilepsy, held in November, 2008, which brought together researchers, clinicians, and patient advocacy groups to discuss strategies and to make plans for research and outreach. Third, professional epilepsy societies have recently established committees and taskforces with patient advocacy organizations, and have committed resources to work together to educate epilepsy professionals, doctors in training, and patients and families about sudden unexpected death in epilepsy (SUDEP), and to chart research agendas. The recommendations of one such task force include “convening a multidisciplinary workshop to reἀne current lines of investigation and to identify additional areas of research for mechanisms underlying SUDEP; performing a survey of patients and their families and caregivers to identify effective means of education that will enhance participation in SUDEP research; conducting a campaign aimed at patients, families, caregivers, coroners, and medical examiners that emphasizes the need for complete autopsy examinations for patients with suspected SUDEP; and securing infrastructure grants to fund a consortium xvii

xviii

Foreword by Schachter

of centers that will conduct prospective clinical and basic research studies to identify preventable risk factors and mechanisms underlying SUDEP” [1]. Fourth, numerous patient advocacy groups and family survivors are talking openly and urgently about sudden death in epilepsy at conferences and on Web sites. In addition, these groups are aggressively funding research projects with the hope of uncovering answers that will move the ἀeld forward. Indeed, these and other developments will likely enable critically important progress to be made in the understanding of sudden death in epilepsy and its prevention. These answers cannot come soon enough. And yet answers alone are not sufficient. If lack of seizure control is a signiἀcant risk factor for sudden death in epilepsy, then we must also urgently address the treatment gap in epilepsy, both by making treatments available to the tens of millions of persons with epilepsy in developing countries who currently have no access to anticonvulsants, and by ensuring that all patients everywhere with epilepsy have access to state-of-the-art care, including the full range of pharmacological and nonpharmacological treatment options. Even if researchers identify the underlying mechanism(s) of sudden death in epilepsy and patients have full access to comprehensive care, there will still remain one major barrier—lack of effective communication between physicians and their patients. Here are the agonizing reflections of a physician about whether to inform patients about sudden death€[2]: As a doctor and an epileptologist, I experience the deepest frustration when I am notiἀed of the sudden and unexpected death of one of my patients. Often these patients have been found dead in bed, lying dead on the floor of their apartment, or drowned in the bathtub. They are invariably young and are often at the beginning of their lives. I never forget them, and I feel that I have truly and ultimately failed them. . . . one of my young patients, a 25-year-old man, died last year while in bed. His parents had encouraged him to live alone so that his life would be as normal as possible. They had visited him that evening and had eaten dinner with him. He had a girlfriend, but she had not been there that night. The next morning he was found lying face down on his bed, fully dressed. No one was sure whether or not he had experienced a seizure. When his parents called me, we talked about the problem of sudden death and epilepsy. They wondered why his other doctors and I had not told them or their son that this could occur. I asked them if they thought he would have lived any differently if he had known. If we told every patient with epilepsy about this possibility, some might not dare live independent lives or might be burdened by anxiety, knowing that they might not awake in the morning. To this day, I do not know if my answer was appropriate.

This challenge must be addressed and solved before the discoveries of research can be translated to patient care if we are ever to end the scourge of sudden death. As physicians caring for patients with epilepsy, we should not wait for all the answers. We must accept the imperfect state of knowledge and inform our patients and, where appropriate, their families in a meaningful and compassionate way about sudden death, and work with them as much as possible to reduce the risk factors, especially by completely controlling generalized convulsive seizures. The editors and authors of Sudden Death in Epilepsy: Forensic and Clinical Issues have produced a landmark book that holds extraordinary promise for meaningful progress, bringing us closer to the day when sudden death will be fully preventable. Until then, my

Foreword by Schachter

xix

hope is that this book will inspire researchers, and give some measure of comfort to the bereaved and hope to those living with epilepsy and to their loved ones. Steven C. Schachter, M.D. Departments of Neurology Beth Israel Deaconess Medical Center Harvard Medical School Chief Academic Officer Center for Integration of Medicine and Innovative Technology Boston, Massachusetts

References 1. So EL, Bainbridge J, Buchhalter JR, et al. Report of the American Epilepsy Society and the Epilepsy Foundation Joint Task Force on sudden unexplained death in epilepsy. Epilepsia 2009;50: 917–922. 2. Schachter SC, ed. Epilepsy in our experience: Accounts of health care professionals. Oxford: Oxford University Press; 2008, pp. 7–8.

Preface

This book should be considered a sequel to, rather than a substitute for, the 1990 book Epilepsy and Sudden Death, edited by Lathers and Schraeder. Much of the material in the 1990 book, especially the discussion of animal research data, is still timely, as little additional animal-based work is extant. In addition, the epidemiological, behavioral, and drug-abuse-related data is also currently relevant. This current volume is an expansion rather than an updated sequel to the previous book on sudden unexpected death in epilepsy (SUDEP). These statements are based on our extensive review of the literature for this current book, which indicates that most of the research questions regarding evaluation of mechanisms and prevention of SUDEP have, to date, not been adequately addressed. When the previous book was published, the phenomenon of sudden unexpected/unexplained death in persons with epilepsy was considered to be a controversial topic or a very rare phenomenon by clinicians and pathologists alike. The public had little or no knowledge of the phenomenon. However, one positive consequence of the work done in SUDEP over the ensuing years is that now the phenomenon is widely referred to as SUDEP and it is accepted as a complication of epilepsy worldwide. SUDEP is one of the most common causes of death in young adults with a history of epilepsy, and presents a spectrum of dilemmas to forensic experts, clinicians, and researchers. At a most basic level, as determined in a national survey of coroners and medical examiners and discussed in this book, even in self-evident cases, SUDEP is infrequently used on death certiἀcates as a ἀnal diagnosis or medical cause of death. Ironically, this survey found that the majority of those same coroners and medical examiners who did not routinely use SUDEP on death certiἀcates, nonetheless acknowledged its validity as a diagnosis in a theoretical case in which no cause of death in a person with epilepsy could be found on postmortem examination. This disconnect between intellectual acknowledgment of its existence, and actual use of SUDEP as a death certiἀcate diagnosis, in all likelihood, results in signiἀcant underreporting of the prevalence of this tragic phenomenon. It is hoped that this book will prove to be an important resource to improve the knowledge of coroners and medical examiners about the use of the term SUDEP in appropriate cases. Since a portion of this new book addresses forensic issues, it is intended to be a resource for forensic pathologists, attorneys, coroners, and medical examiners as they struggle to determine the cause of death in persons with epilepsy. Because most other clinical and research aspects of SUDEP are also addressed, neurologists, experts in epilepsy, cardiologists, clinical pharmacologists, pharmacists, nurses, students, and persons with epilepsy or with a family member so diagnosed should all ἀnd sections of interests. Over the past two decades, many more basic and clinical scientists doing research in the ἀeld of epilepsy are focusing on the problem of SUDEP, as it is one of the most common causes of death associated with having epilepsy. However, to date, most research has focused on clinical and epidemiological data, with relatively few investigators using xxi

xxii

Preface

animal models. In addition, there has been little effort to determine whether there are common links between the risk of SUDEP and sudden cardiac deaths. While cardiologists are focusing increasingly on genetic risk factors for potentially life-threatening arrhythmias in young persons without coronary artery disease, there has been little effort to establish collaboration between neuroscience investigators interested in SUDEP and cardiology researchers interested in unraveling the genetically determined risk factors for potentially fatal arrhythmias. This book will be of help in directing neurological and cardiologic investigators to common areas of interest, especially as relates to the unanswered question of why some persons with epilepsy seem to be at risk for neurogenically induced fatal cardiac events. The basic science of epilepsy as relates to SUDEP is updated and expanded in this book. The role of alcohol and other drugs as seizure enhancing or producing conditions are discussed. The role of medication, compliance with prescribed antiepileptic drugs, or lack thereof, is an even more important problem today and is addressed in several chapters. The relevance of epidemiological studies of SUDEP is also presented. Sudden death in the pediatric population is also reviewed. The neurocardiological aspects of SUDEP are addressed in some detail. While SUDEP remains a major unsolved problem, thanks to the efforts of those who have conducted animal and epidemiological studies, what was once denied and called a myth is an acknowledged reality that has to be dealt with as a multidisciplinary issue. To solve the mystery of SUDEP, a global focus is required. Persons at risk must be identiἀed and preventive treatment regimens developed to decrease the occurrence of SUDEP. Clinical chapters emphasize that sophisticated simultaneous ambulatory EKG and EEG telemetry and respiratory function monitoring of patients at risk for sudden death will help identify cardiac, respiratory, and epileptogenic interactions in order to decrease the risk of SUDEP. Basic scientiἀc research programs and clinical and epidemiology studies are needed. Multidisciplinary teams working in clinical settings and with laboratories must address the global issues of SUDEP. Forensic chapters in this book discuss the fact that if the correct term “SUDEP” is used on autopsy reports, and if postmortem verbal autopsies are conducted when needed, the true incidence of SUDEP well may be determined to be much higher than previously thought. Drug development can justify the need for new antiepileptics and drugs in other pharmacological classes that address the reduction of the risk of SUDEP. Animal model chapters discuss new data gleaned by building on the previously utilized models and the lessons learned during the last quarter century. The use of animal models continues to be one of the most useful approaches to better understanding SUDEP. Discussion in various chapters may be summarized by stating that it is important to “think out of the box” when evaluating an established animal model that has the potential, with modiἀcation(s), to investigate possible mechanisms of SUDEP. Various authors of the animal model chapters presented in this book emphasize that multiple models are needed to investigate the pathophysiology of SUDEP, to hypothesize about effective treatments, to develop pilot studies in persons with epilepsy, and to conduct conἀrmatory large-scale clinical trials. The ἀelds of pharmacology, clinical pharmacology, and cardiology have much to offer as we work to improve compliance, develop new antiepileptic drugs, and apply different categories of drugs that hopefully attenuate the chances of occurrence of SUDEP. Authors address the possible overlapping mechanisms that may apply to the risk of sudden

Preface

xxiii

unexpected death occurring in epilepsy and in cardiac disease. Several chapters explore the interaction between the central and peripheral autonomic nervous systems and the cardiopulmonary systems. Included is a discussion of the potential interactive role of genetically determined subtle cardiac risk factors for arrhythmias, with a predisposition for seizurerelated cardiac arrhythmias. The possible mechanisms that are operant in producing both epileptogenic and cardiogenic arrhythmias are addressed. Several chapters examine proposed mechanistic factors in SUDEP, listing risk categories of arrhythmogenic, respiratory, and hypoxia, and psychological factors and discussing mechanisms for risks associated with each category. Several chapters discuss patients with Brugada syndrome and an interesting, interpretive presentation of a hypothesis to explain a common pathophysiologic mechanism associated with sodium channel dysfunction that may be common to clinical electrophysiological abnormalities and some types of epilepsy. Clariἀcation of risk factors and establishment of the mechanism of SUDEP are important to establish preventative measures for SUDEP and emphasize the need to strive for full seizure control. Several chapters discuss the importance of encouraging patients with epilepsy to receive nonmedical, common sense, lifestyle-modifying interventions that have generally accepted public health beneἀts, even though there is as yet no consensus that they may or may not prevent sudden death. Cardiac patients, psychiatric patients, and certain ethnic groups experiencing acute stressful circumstances are at risk for unexpected sudden death. The impact of adverse emotional states on the autonomic control of cardiac rhythm is an established factor leading to cardiac dysrhythmias in humans and other species. Although stress is associated with changes in autonomic neural function, its role as a potential risk factor for SUDEP has not been investigated. The association of epilepsy with depression and anxiety indicates that emotional stress should be evaluated as a potential risk factor for SUDEP. The interactions between emotional factors and the arrythmogenic potential of epileptiform discharges, and the possibility of beneἀt from stress management intervention need investigation. Prospective studies of patients are needed to determine how we can identify which persons with epilepsy are at risk for SUDEP. In a number of chapters, the authors speculate about common potential preventive measures to minimize the risk of both sudden unexpected death in epilepsy and sudden cardiac death. Several chapters address the issue of clinicians who treat persons with epilepsy, manifesting reluctance to discuss the possibility of SUDEP with their patients. This reluctance seems to be the result of concern that even introducing the topic to the patient and family would be stressful for them. However, for the most part, families of SUDEP victims express disappointment that they had not known of this possibility and call for widespread acknowledgment of the potential for occurrence of sudden death in association with epilepsy. Epilepsy Bereaved in the United Kingdom, an organization founded by families of SUDEP victims, has been particularly successful in raising the level of awareness of SUDEP in persons with epilepsy and their families, and within the medical profession and the general public. This kind of advocacy and dissemination of information will serve to increase the availability of resources used to solve the tragic mystery of SUDEP. A primary purpose of this book is to provide clinicians with the knowledge necessary to improve their comfort level in discussing SUDEP with patients and families and thereby to allow for freer dissemination of information about minimizing the known risk factors for SUDEP (e.g., erratic compliance in taking prescribed antiepileptic drugs). Committed investigators in research must solve the mystery of SUDEP using a leadership philosophy foundation that provides innovative vision and approaches for SUDEP

xxiv

Preface

research and teaching programs. The interaction of teaching and research is essential. While a student is learning how to conduct research, that person must simultaneously learn to become a teacher. Academic fellowships and competitions and grant funding for medical students, postdoctoral fellows, residents, and faculty will attract medical and graduate trainees to work on SUDEP and move the SUDEP knowledge base forward. As self-learning exercises, we have incorporated a variety of case studies of sudden death within chapters and as standalone chapters as practical teaching exercises. Clinical and basic science investigators must provide vision and a fertile environment to teach students to become tomorrow’s leaders in the struggle to solve the mystery of SUDEP. Claire M. Lathers Paul L. Schraeder Michael W. Bungo Jan E. Leestma

Acknowledgments

I wish to acknowledge the dedication and hard work of co-editor Claire Lathers. (JEL) We would not have been able to complete this book in a timely manner without the diligent secretarial assistance of Marie Faiola. (CML, PLS, MWB, JEL).

xxv

Editors

Claire M. Lathers, PhD, FCP, has been credentialed as a Senior Biomedical Research Scientist by the U.S. Food and Drug Administration (FDA) for international recognition of her work in the two areas of cardiovascular autonomic dysfunction associated with sudden death in persons with epilepsy and with space flight. The primary focus of her international cardiovascular pharmacology research career has centered on autonomic peripheral and central mechanisms involved in the control and regulation of blood pressure, heart rate and rhythm, and the electroencephalogram. Dr. Lathers and Dr. Schraeder have collaborated and published numerous studies and two books focusing on epilepsy and sudden unexplained death. Dr. Lathers served the FDA for a total of 11 years, including four years as the Senior Advisor for Science to the director in the Center for Veterinary Medicine and director of the Office of New Animal Drug Evaluation and ἀve years in the Center for Drug Evaluation and Research as a pharmacology reviewer. Claire also served as a special government expert for 2 years. Dr. Lathers spent 14 years working as a visiting scientist at NASA/ USRA, collecting data from subjects in ground-based studies and from astronauts and cosmonauts before, during and after space flight. Dr. Lathers earned a BS in pharmacy from Albany College of Pharmacy, Union University and her PhD in pharmacology from the State University of New York at Buffalo School of Medicine. She completed an NIH funded two-year postdoctoral fellowship at the Medical College of Pennsylvania. Her academic faculty experience includes working at the Medical College of Pennsylvania (15 years), Albany College of Pharmacy (two years as president, dean, and tenured professor); Uniformed Services University of the Health Sciences (three years part time); and Gwynedd Mercy College (11 years part time). Dr. Lathers is currently collaborating with a number of academicians on scientiἀc issues of the nexus between human and veterinary medicine clinical pharmacology, antimicrobial resistance, food safety, and biodefense measures. In addition to her academic and government service, Dr. Lathers has worked in the pharmaceutical industry. She served as chief scientiἀc officer of Barr Pharmaceuticals for three years and worked part time with four other pharmaceutical companies during a 15-year period. Dr. Lathers has authored or co-authored over 300 publications and abstracts, has edited three books, and has presented data at over 140 international meetings. She is an emeritus fellow, an honorary member of the Board of Regents, and a past president of the American College of Clinical Pharmacology, having served as regent, treasurer, and president. Dr. Lathers also served as the section editor of the educational series entitled: “Innovative Teaching Methods in Clinical Pharmacology” for the Journal of Clinical Pharmacology for 17 years. Claire served as a member of the Board of the Annapolis Center, charged to evaluate risk assessments, and worked on the epidemiology, toxicology, and food safety workshops and accords. In recognition of her work, Dr. Lathers has been the recipient of numerous awards and honors. xxvii

xxviii Editors

Paul L. Schraeder, MD, FAAN, is professor emeritus of neurology at Drexel University College of Medicine, former chief of neurology at the Medical College of Pennsylvania Hospital, Philadelphia, Pennsylvania and former professor of medicine and neurology at the Robert Wood Johnson School of Medicine in Camden, New Jersey; head of the Division of Neurology at Cooper Hospital/University Medical Center, Camden, New Jersey; and former associate professor of neurology at the Medical College of Pennsylvania. He is a member of the Philadelphia Neurological Society, the American Epilepsy Society, and a fellow of the American Academy of Neurology. He has served on the professional advisory board of the Epilepsy Foundation of Southeastern Pennsylvania and the Epilepsy Foundation of America and as medical advisor to Epilepsy Bereaved, a support organization for surviving friends and family of victims of sudden unexplained death in epileptic persons (SUDEP) in the United Kingdom. Dr. Lathers and Dr. Schraeder co-edited the ἀrst book addressing the topic of Epilepsy and Sudden Death (Marcel Dekker, 1990). Dr. Lathers and he have collaborated for over three decades studying and investigating the mystery of SUDEP and developed the ἀrst experimental animal models of this fatal phenomenon. Dr. Schraeder organized a collaborative nationwide survey of how coroners and medical examiners evaluate the deaths of persons with a history of epilepsy. Dr. Schraeder received the AB€ degree from Bucknell University, Lewisburg, Pennsylvania and the MD degree from€Jefferson Medical€College, Philadelphia, Pennsylvania. He completed his residency in neurology and fellowship in electroencephalography and experimental epilepsy at the University of Wisconsin. Michael W. Bungo, MD, FACC, FACP, has agreed to serve as the third co-editor of this new book. After residency and fellowship training in cardiology at Harvard Medical School programs in Boston, Dr. Bungo worked full time for NASA as director of the Space Biomedical Research Institute at the Johnson Space Center. During his time at NASA, Dr. Bungo ἀrst cataloged arrhythmias occurring in spaceflight crews and hypothesized that the unique physiological and psychological environments of space flight may be arrhythmogenic. The Aerospace Medical Association awarded him the Louis H. Bauer Founders Award and NASA awarded him the NASA Medal for Exceptional Scientiἀc Achievement for his pioneering research work concerning the heart’s adaptation to zero gravity. While at NASA, Dr. Bungo served on the Joint U.S.–U.S.S.R. Working Group that developed the now combined space station science program for these prior competitors. He left NASA to assume the positions of director of the Heart Station, Division of Cardiology and vice chair for Inpatient Affairs, Department of Internal Medicine at the University of Texas Medical Branch in Galveston. Dr. Bungo subsequently moved to the University of Texas Medical School in Houston and served as chief of staff at the LBJ General Hospital, CEO of UT-Physicians, and vice dean for Clinical Affairs. He is currently a professor of medicine in the Division of Cardiology at that same institution. Dr. Bungo’s current research project funded by NASA is entitled “Cardiac Atrophy and Diastolic Dysfunction during and after Long Duration Spaceflight: Functional Consequences for Orthostatic Intolerance, Exercise Capacity, and Risk of Cardiac Arrhythmias.” In addition to numerous publications and presentations, Dr. Bungo and Dr. Lathers have collaborated since 1989 and have co-authored seven published papers. In 2008, Dr. Bungo has co-authored three papers and one book chapter with Drs. Lathers and Schraeder on the mystery of SUDEP. Dr.€Bungo has delivered scientiἀc presentations at three international symposiums organized by Dr.€Lathers.

Editors

xxix

Jan E. Leestma, MD, MM, is the lead author of the second edition of Forensic Neuropathology. He received the MD degree from the University of Michigan School of Medicine in 1964, and a Masters of Management (MM) degree from the J.L. Kellogg Graduate School of Management of Northwestern University, Evanston, Illinois in 1986. He completed residency training in anatomic pathology and neuropathology at the University of Colorado Medical Center, Denver, Colorado and a neuropathology fellowship at the€Albert Einstein College of Medicine, Bronx, New York. He is certiἀed in both anatomic pathology and neuropathology by the American Board of Pathology (1970). He served in the United States Air Force Medical Corps at the Armed Forces Institute of Pathology, Washington, D.C. (1968–1971) and was honorably discharged with the rank of major, USAF MC. He was an assistant and associate professor of pathology and neurology at Northwestern University School of Medicine (1971–1986) and served as chief of neuropathology at both Northwestern Memorial Hospital and the Children’s Memorial Hospitals, Chicago, Illinois. He was professor of pathology and neurology, and dean of students for the Division of the Biological Sciences and the Pritzker School of Medicine at the University of Chicago, Chicago, Illinois (1986–1987). He was an assistant medical examiner and neuropathology consultant to the Office of the Medical Examiner, Cook County, Illinois (1977–1987). He was a guest researcher at the Karolinska Institutet, Huddinge University Hospital, Pathology Institute, Stockholm, Sweden (1981–1982). He was associate medical director and chief of neuropathology at the Chicago Institute of Neurosurgery and Neuroresearch in Chicago (1987–2003). He has had a private consulting practice in forensic neuropathology since the early 1970s which continues to the present time. He has given expert testimony in more than 30 U.S. states, Canada, and the United Kingdom. He is the author of more than 100 professional publications including numerous book chapters in texts. He was the author of Forensic Neuropathology (ἀrst edition), Raven Press, New York, 1988. He is a member of the American Association of Neuropathologists, and of the American Academy of Forensic Sciences. Dr. Leestma wrote some chapters in Epilepsy and Sudden Death, edited by Drs. Lathers and Schraeder. Thus it is a natural extension of this collaboration for these three editors to work with co-editor Dr. Bungo and CRC Press is to expand the focus of forensics and clinical issues. The diverse scientiἀc expertise and endeavors of each of the four editors, working in the different ἀelds of forensics, neurology, cardiology, and clinical pharmacology, have united in this edition to produce a book with a special emphasis on the forensics and clinical issues associated with neurocardiology, epilepsy, arrhythmias, and sudden death.

Contributors

Isha Agarwal

Medical College of Pennsylvania Philadelphia, Pennsylvania

Karim A. Alkadhi, PhD

Department of Pharmacological and Pharmaceutical Sciences University of Houston Houston, Texas

Karem H. Alzoubi

Department of Clinical Pharmacy Jordan University of Science and Technology Irbid, Jordan

Anne E. Anderson, MD

Departments of Pediatrics, Neurology, Neuroscience Baylor College of Medicine Houston, Texas

Ricardo M. Arida

Departamento de Fisiologia Universidade Federal de São Paulo/Escola Paulista de Medicina (UNIFESP/EPM) São Paulo, Brasil

Jacquelyn Bainbridge, PharmD, FCCP School of Pharmacy University of Colorado Health Sciences Center Denver, Colorado

Steven L. Bealer, PhD

Department of Pharmacology and Toxicology University of Utah Salt Lake City, Utah

Elijah R. Behr, MD

Cardiac and Vascular Division St. George’s University of London London, United Kingdom

Amy Brewster

Department of Pediatrics Baylor College of Medicine Houston, Texas

Michael W. Bungo, MD, FACC, FACP,€CPE

Division of Cardiology University of Texas Medical School at Houston Houston, Texas

Esper A. Calvalheiro

Disciplina de Neurologia Experimental Universidade Federal de São Paulo/Escola Paulista de Medicina (UNIFESP/EPM) São Paulo, Brasil

Marcello Alessandro Caria

Department of Biomedical Sciences Human Physiology Division Sassari, Italy

Anne Y. Y. Chan

Department of Medicine and Therapeutics Chinese University of Hong Kong and Prince of Wales Hospital Hong Kong SAR, China

H. Gregg Claycamp, PhD

Center for Drug Evaluation and Research Office of Compliance U.S. Food and Drug Administration Silver Spring, Maryland

Isaac L. Crawford

Department of Neurology Southwestern Regional Epilepsy Center Veterans Administration Medical Center University of Texas Southwestern Medical Center Dallas, Texas

Thomas L. Bohan, PhD, JD, F-AAFS, D-IBFES MTC Forensics Peaks Island, Maine

xxxi

xxxii

Roberta M. Cysneiros

Programa de Pós-Graduação em Distúrbios do Desenvolvimento do Centro de Ciências Biológicas e da Saúde da Universidade Presbiteriana€Mackenzie São Paulo, Brasil

Paramdeep S. Dhillon

Cardiac and Vascular Sciences St. George’s University of London London, United Kingdom

Contributors

Benito Herreros, MD, PhD

Cardiology Department Hospital Universitario Rio Hortega Valladolid, Spain

Trieu Ho, MD

Division of Cardiology University of Texas Medical School at Houston Houston, Texas

Richard W. Homan, MD

Department of Physiology Johns Hopkins University Baltimore, Maryland

Department of Neurology University of Texas Southwestern Medical Center Southwestern Regional Epilepsy Center Veterans Administration Medical Center Dallas, Texas

Kathleen Dolce

John D. Hughes, MD

Jeffrey M. Dodd-O, MD

Department of Drug Metabolism Smith Kline & French Laboratories King of Prussia, Pennsylvania

Michael P. Earnest, MD Department of Neurology Denver Health and Hospitals Denver, Colorado

Carl L. Faingold, PhD

Department of Pharmacology Southern Illinois University School of Medicine Springἀeld, Illinois

Josef Finsterer

Krankenanstalt Rudolfstiftung Vienna, Austria

Neeti Ghali

Department of Clinical Genetics Guy’s Hospital London, United Kingdom

Jeffrey H. Goodman, PhD

Department of Physiology and Pharmacology State University of New York Brooklyn, New York

Jane Hanna, OBE, MA, BCL Epilepsy Bereaved Wantage, United Kingdom

Ryan S. Hays

Department of Neurology Thomas Jefferson University Philadelphia, Pennsylvania

National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland

Richard Hughes

Department of Neurology Denver Health and Hospitals Denver, Colorado

Lara Jehi, MD

Department of Neurology Cleveland Clinic Neurological Institute Cleveland, Ohio

Kam F. Jim, PhD

Clinical Documentation Sanoἀ-Aventis Great Valley, Pennsylvania

Claire Kahn, PhD

Department of Drug Metabolism Smith Kline & French Laboratories King of Prussia, Pennsylvania

Bharat K. Kantharia, MD

Division of Cardiology University of Texas Medical School at Houston Houston, Texas

Steven B. Karch, MD, FFFLM, FFDov Berkeley, California

Steven A. Koehler

Office of the Coroner of Allegheny County Pittsburgh, Pennsylvania

Contributors

Claire M. Lathers, PhD, Emeritus FCP Albany, New York

xxxiii

Lina Nashef, MBChB, MD, FRCP

Chicago, Illinois

Departments of Neurology and Clinical Neurosciences King’s College Hospital London, United Kingdom

Howan Leung

Maromi Nei, MD

Robert M. Levin

Daniel K. O’Rourke, MD

Albany College of Pharmacy Albany, New York

Medical College of Pennsylvania Philadelphia, Pennsylvania

Jason G. Little

Rosemary Panelli, PhD

Jan E. Leestma, MD

Department of Medicine and Therapeutics Chinese University of Hong Kong Hong Kong SAR, China

Department of Neurology Jefferson Comprehensive Epilepsy Center Philadelphia, Pennsylvania

Department of Pharmacology and Toxicology University of Utah Salt Lake City, Utah

Joint Epilepsy Council of Australia Seymour, Australia

Edward H. Maa, MD

Graduate School of Nursing University of Virginia Newport News, Virginia

Neurology Denver Health and Hospitals Denver, Colorado

Ombretta Mameli, MD

Department of Biomedical Sciences Human Physiology Division Sassari, Italy

William D. Matthews, PhD

Department of Drug Metabolism Smith Kline & French Laboratories King of Prussia, Pennsylvania

John A. Messenheimer, MD

Department of Neurology University of North Carolina School of Medicine Chapel Hill, North Carolina

Cameron S. Metcalf

Department of Pharmacology and Toxicology University of Utah Salt Lake City, Utah

Yashanad Mhaskar

Department of Pharmacology Southern Illinois University School of Medicine Springἀeld, Illinois

Imad Najm, MD

Department of Neurology Cleveland Clinic Neurological Institute Cleveland, Ohio

Simona Parvulescu-Codrea, MD, PhD

David S. Paterson, PhD Department of Pathology Children’s Hospital Boston Boston, Massachusetts

Wallace B. Pickworth, PhD Addiction Research Center National Institute on Drug Abuse Baltimore, Maryland

Stephen R. Quint, PhD

Department of Neurology University of North Carolina Chapel Hill, North Carolina

Lene Sahlholdt

Danish Epilepsy Centre Dianalund, Denmark

Martin Allen Samuels, MD Brigham and Women’s Hospital Harvard Medical School Boston, Massachusetts

Susumu Sato, MD

Office of Clinical Director National Institute of Neurological Disorders and Stroke Bethesda, Maryland

xxxiv

Steven C. Schachter, MD

Beth Israel Deaconess Medical Center Harvard Medical School Boston, Massachusetts

Paul L. Schraeder, MD, FAAN

Department of Neurology (emeritus) Drexel University College of Medicine Philadelphia, Pennsylvania

Carla A. Scorza

Disciplina de Neurologia Experimental Universidade Federal de São Paulo/Escola Paulista de Medicina (UNIFESP/EPM) São Paulo, Brasil

Fulvio Alexandre Scorza

Disciplina de Neurologia Experimental Universidade Federal de São Paulo/Escola Paulista de Medicina (UNIFESP/EPM) São Paulo, Brasil

Saumya Sharma, MD

Division of Cardiology University of Texas Medical School at Houston Houston, Texas

Nicole Simpkins

Jefferson Comprehensive Epilepsy Center Thomas Jefferson University Philadelphia, Pennsylvania

Elson L. So, MD

Section of Electroencephalography Mayo Clinic College of Medicine Rochester, Minnesota

Eliza Y. F. Sonoda

Disciplina de Neurologia Experimental Universidade Federal de São Paulo/Escola Paulista de Medicina (UNIFESP/EPM) São Paulo, Brasil

Michael R. Sperling, MD

Jefferson Comprehensive Epilepsy Center Thomas Jefferson University Philadelphia, Pennsylvania

Michele M. Spino

Medical College of Pennsylvania Philadelphia, Pennsylvania

Mark C. Spitz, MD Health Science Center University of Colorado Aurora, Colorado

Contributors

William H. Spivey, MD (deceased) Medical College of Pennsylvania Philadelphia, Pennsylvania

Amy Z. Stauffer, MD

Medical College of Pennsylvania Philadelphia, Pennsylvania

Mark Stewart, MD, PhD

Department of Physiology/Pharmacology and Neurology State University of New York Brooklyn, New York

Claudia Stöllberger

Second Medical Department Krankenanstalt Rudolfstiftung Vienna, Austria

Michael B. Tennison

School of Medicine University of North Carolina Chapel Hill, North Carolina

Vera C. Terra

Departamento de Neurologia, Psiquiatria e Psicologia Médica Universidade de São Paulo. Ribeirão Preto São Paulo, Brasil

Torbjörn Tomson

Department of Neurology Karolinska University Hospital Stockholm, Sweden

Srinivasan Tupal

Department of Pharmacology Southern Illinois University School of Medicine Springἀeld, Illinois

Laurie S. Y. Tyau, MD

Medical College of Pennsylvania Philadelphia, Pennsylvania

Victor V. Uteshev

Department of Pharmacology Southern Illinois University School of Medicine Springἀeld, Illinois

Matteo Vatta

Department of Pediatrics Baylor College of Medicine Houston, Texas

Contributors

Richard L. Verrier, PhD, FACC Beth Israel Deaconess Medical Center Harvard Medical School Boston, Massachusetts

ἀ addeus S. Walczak, MD MINCEP Epilepsy Care Minneapolis, Minnesota

xxxv

Braxton B. Wannamaker, MD Epilepsy Services and Research, Inc. Orangeburg, South Carolina

Cyril H. Wecht, MD, JD

Office of the Coroner of Allegheny County Pittsburgh, Pennsylvania

Forensics of Sudden Death

I

Neurocardiologic Mechanistic Risk Factors in Sudden Unexpected Death in Epilepsy

1

Claire M. Lathers Paul L. Schraeder Michael W. Bungo

Contents 1.1 Introduction 1.2 Risk Factors for SUDEP 1.3 SUDEP Animal Models 1.3.1 Therapies as Factors 1.4 Case of SUDEP with a Premorbid Diagnosis of Nonepileptic Seizures 1.5 Discussion 1.5.1 Future Steps References

3 3 15 24 24 26 26 26

1.1â•…Introduction There are at least three categories of factors that may be operative in the mechanisms for sudden unexpected death in epilepsy (SUDEP) (Lathers et al. 2008a): arrhythmogenic, including changes in autonomic neural and cardiac function, respiratory and hypoxia, and psychological. Each of these main risk factor categories, in all likelihood, includes many subcategories. For example, the arrhythmogenic category includes pharmacological drug effects, genetic ion channelopathies, and acquired heart disease. The psychological factors include stress, anxiety, and depression states. This list of three primary areas of mechanism for SUDEP was expanded to include eight topics (Lathers et al. 2008b, 2008c).

1.2â•…Risk Factors for SUDEP The risk factors for SUDEP are as follows:

1. Pathophysiological mechanisms: cardio/autonomic 2. Respiratory/autonomic factors and hypoxia 3. Syncope 4. Genetic/structural mechanistic factors 5. Risk factors for SUDEP 6. Therapies: increased or decreased risk 7. Psychological factors 8. Unusual factors possibly modifying risk 3

4 Sudden Death in Epilepsy: Forensic and Clinical Issues

The pathophysiological mechanisms of SUDEP are not extant (see Table 1.1). From a cardiac standpoint, there is no question that the sympathetic component of cardiac innervation is involved in the production of potentially fatal tachyarrhythmias. It has also been established that postganglionic cardiac sympathetic innervation is altered in association with temporal lobe epilepsy and may be a pathophysiological risk factor for SUDEP (Druschky et al. 2001). These clinical data correlate with animal studies that found epiÂ� leptiform discharge–related postganglionic cardiac sympathetic abnormalities that were associated with cardiac conduction disturbances and arrhythmias (Lathers and Schraeder 1982, 1987, 1990a; Schraeder and Lathers 1983, 1989; Lathers 2010a, 2010b; Lathers et€al. Table 1.1â•… Pathophysiological Mechanisms Model, Symptoms, and/or Changes Nonuniform autonomic cardiac postganglionic neural discharge associated with coronary occlusion of left anterior descending coronary artery and/or ouabain toxicity in cats.

Arrhythmia monitoring of pacemaker patients. Cardiac beta receptor distribution: Beta receptor density right atria signiἀcantly lower than left atria; right ventricle signiἀcantly lower than left ventricle; density in ventricles higher than atria; beta receptor density of distal distribution of left anterior descending coronary artery signiἀcantly higher than proximal distribution. These regional density differences are related to the cardiac contractile strength of different areas of heart. Regional differences in beta adrenoreceptor densities reflect differences in postganglionic cardiac sympathetic innervation of the myocardium (Randall et al. 1977, 1978, 1984). Regulation of cardiac neural discharge is a new paradigm in the management of sudden cardiac death.

Mechanisms of Sudden Death Nonuniform (increases, decreases, and/or no change) autonomic neural postganglionic cardiac sympathetic discharge traveling through the stellate ganglia causes cardiac€arrhythmias, ventricular ἀbrillation, and/or SUD in the manner described by Han and Moe (i.e., nonuniform recovery of excitability in ventricular muscle). Suggested neural discharge site for drugs to modify occurrence of arrhythmia and death (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978, 1981, 1988; Lathers 1980, 1981, 1982; Lathers and Roberts 1980, 1985; Spivey and Lathers 1985; Han and Moe 1964). Revealed tachycardias, most likely ventricular tachycardia, related to SUD, give insight into terminal event mechanisms (Wichter et al. 2005; Nagele et al. 2007). A correlation with the release of norepinephrine at sympathetic nerve terminals in the heart in a manner to produce arrhythmia. Innervation density is high in subepicardium and central conduction system. Nonuniform postganglionic cardiac sympathetic cardiac innervation is related to nonuniform beta sympathetic receptor locations in the heart and affects cardiac contractility and development of arrhythmias and/or death. In diseased hearts, cardiac innervation density varies and may lead to sudden cardiac death (Lathers et al. 1986, 1986b, 1988, 1990, 2010a, 2010b, 2010c). These data were conἀrmed by Druschky et al. (2007), Ieda et al. (2006, 2007, 2008), and Kimura et al. (2007). These site differences will vary release of norepinephrine in various sites of the heart. This modiἀes cardiac contractile function and may trigger development of arrhythmias and/or sudden death. Site difference, in part, is one component of the mechanism(s) involved in sudden cardiac death. The heart is extensively innervated and its performance is regulated by the balance of discharges within and between the autonomic nervous system divisions (Ieda et al. 2008; Lathers et al. 1977a, 1977b, 1978; Lathers 1980, 1981). (continued)

Neurocardiologic Mechanistic Risk Factors in SUDEP

5

Table 1.1â•… Pathophysiological Mechanisms (Continued) Model, Symptoms, and/or Changes Nonuniform autonomic cardiac postganglionic neural discharge associated with pentylenetetrazol-induced interictal epileptogenic activity in cats. Lockstep phenomenon (LSP): Cardiac postganglionic sympathetic and vagal discharges were synchronized one for one with both ictal and interictal discharges and premature ventricular contractions, ST/T changes, and conduction blocks, and precipitous changes in blood pressure occurring concurrent with interictal spikes.

Blockading GABAergic and glycinergic receptors in medulla slices of newborn rats evoked intermittent seizure-like ἀring of cardiac parasympathetic neurons, suggesting the seizure-like pattern of ἀring during an epileptic attack may cause neurogenic ictal bradyarrhythmias, cardiac asystole, or even sudden death in persons with epilepsy (Wang et al. 2006). Cardiac postganglionic denervation in patients with epilepsy examined to evaluate ictal asystole because tachyarrhythmias are common during epileptic seizures, while bradyarrhythmias or asystoles occur less frequently (Kerling et al. 2009).

Chaos science: Simple systems that manifest periodic activity are easily perturbed, and are less able to return to the preperturbed state.

Mechanisms of Sudden Death Nonuniform autonomic neural discharge, autonomic neural imbalance of postganglionic cardiac sympathetic, and vagal discharge causes cardiac arrhythmias, ventricular ἀbrillation, and/or asystole and/or SUDEP (Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1983, 1988; Lathers et al. 1984, 1990, 1993; Carnel et al. 1985; Tumer et al. 1985). Lockstep phenomenon is considered to be one potential mechanism for SUDEP (Lathers et al. 1987; Lathers and Schraeder 1990a, 1990b; Stauffer et al. 1989, 1990; Dodd-O and Lathers 1990; O’Rourke and Lathers 1990). 1. Spatial and temporal summation of neuronal discharges in a subcortical center producing a stimulus strong enough to overcome the cortical and ganglionic threshold (Dodd-O and Lathers 1990). 2. Increased synaptic recruitment, resulting in ampliἀcation of subcortical stimuli along their path, so when reaching the cortex and sympathetic ganglion, they are capable of causing susceptible neurons in these regions to discharge. 3. Increased irritability of all neurons so that subcortical impulses could stimulate cortical and ganglionic neurons (Dodd-O and Lathers 1990). Provides supporting data for the lockstep phenomenon ἀnding of Lathers et al. (1987) and colleagues. It is to be determined if the predisposition of central PNS ἀring correlates with electrical remodeling of myocardium, possibly secondary to epileptogenic activity related cardiac neural discharges (LSP).

Single-photon emission computed tomography examined I-(123)-meta-iodobenzylguanidine as a marker of postganglionic cardiac norepinephrine uptake. Pronounced reduction in cardiac single-photon emission computed tomography uptake in asystolic patients indicated postganglionic cardiac catecholamine disturbance. Impaired sympathetic cardiac innervation limits adjustment and modulation of heart rate and may increase the risk of asystolic events and, eventually, SUDEP (Kerling et al. 2009). Data of Kerling et al. support the ἀndings of Lathers et al. (1986a, 1986b, 1987, 1990), Lathers and Levin (2010), and those of Han and Moe (1964). A periodic rhythm in the brain, where normally rich complexity exists, implies a susceptibility to failure that may result in death (Gleick 1987). (continued)

6 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.1â•… Pathophysiological Mechanisms (Continued) Model, Symptoms, and/or Changes Mode locking: Epileptic focus and medullary cardiac center may be locked one to one during state of LSP in the manner suggested by Gleick (1987).

Power spectral analysis (Quint et al. 1990).

Sympathetic innervation is critical for effective cardiac function (Saffitz 2008). Developmental and regulatory mechanisms determining density and pattern of cardiac sympathetic innervation are unclear, as is role of innervation in arrhythmogenesis. Sema3a establishes cardiac sympathetic innervation patterning. Sema3a is abundantly expressed in the trabecular layer in early-stage embryos but is restricted to Purkinje ἀbers after birth, forming an epicardial-to-endocardial transmural sympathetic innervation pattern. Sema3a(−/−) mice lacked a cardiac sympathetic innervation gradient and exhibited stellate ganglia malformation, leading to marked sinus bradycardia due to sympathetic dysfunction. Sympathetic dysfunction: altered postganglionic cardiac sympathetic innervation found postmortem in patients with chronic temporal lobe epilepsy (Druschky et al. 2001).

Mechanisms of Sudden Death Occurrence of a very regular oscillator in the brain is theoretically dangerous, regardless of mechanism. Fractal processes are ubiquitous in biological systems, include brain electrical depolarizations, and are systems that convey different information. Complexity of a fractal process may at times depreciate into a simple periodic process representing decay of the system and dramatic change (Goldberger et al. 1985; Winfree 1987). Causes of death include failure of the brain and heart. If the brain sends a message to the heart triggering a fatal arrhythmia, or if the heart enters an arrhythmia via its own initiative, sudden death will occur. Perturbations of cardiac electrical depolarization may have several mechanisms, any or several of which may be operations (O’Rourke and Lathers 1990). Time-frequency domain analyses of heart rate variability in patients with epilepsy and time-frequency mapping of R–R intervals during partial seizure may provide procedures to assess autonomic activity related to risk factors for SUDEP (Novak et al. 1999; Everengul et al. 2005; Persson et al. 2007). Cardiac-speciἀc overexpression of Sema3a in transgenic mice (SemaTG) associated with reduced sympathetic innervation and attenuation of the epicardial-to-endocardial innervation gradient. SemaTG mice demonstrated sudden death and susceptibility to ventricular tachycardia, due to catecholamine supersensitivity and prolongation of the action potential duration. Concluded appropriate cardiac Sema3a expression is needed for sympathetic innervation patterning and is critical for heart rate control (Ieda et al. 2007).

Altered postganglionic cardiac sympathetic innervation may increase risk of cardiac abnormalities and/or SUDEP (Lathers et al. 1990, Chapter 22).

(continued)

Neurocardiologic Mechanistic Risk Factors in SUDEP

7

Table 1.1â•… Pathophysiological Mechanisms (Continued) Model, Symptoms, and/or Changes Amygdaloid kindled seizure effect on cardiovascular system was examined in rats. An abrupt 50% increase in mean arterial pressure (BP) lasting 20–30 s after initiation of seizure occurred with profound bradycardia, characterized by a rate about half of that recorded before stimulation. Changes in heart rate and BP observed during amygdaloid kindled seizures are similar to those observed during secondary spontaneous seizures. Effects apparently are independent of kindling stimulus because stimulusinduced cardiovascular changes were not present at the beginning of kindling. Use of electrical stimulation as a therapy for epilepsy is currently being studied in experimental animals and in patients with epilepsy. This study examined the effect of preemptive, low-frequency, 1-Hz sine wave stimulation (LFS) on incidence of amygdala kindled seizures in rats. Amygdaloid kindled seizures in unanesthetized rats induced abrupt elevation of blood pressure accompanied by a signiἀcant decrease in heart rate.

Mechanisms of Sudden Death Results suggest the kindling seizure model is useful to study underlying mechanisms of seizure-induced cardiac arrhythmias and possibly the clinical phenomenon of SUDEP (Vindrola et al. 1984; Goodman et al. 2005).

Dramatic decrease in incidence of stage 5 seizures in fully kindled animals after preemptive LFS suggests low-frequency stimulation may be an effective therapy for prevention of seizures in patients with epilepsy (Gary-Bobo and Bonvalette 1977). Muscarinic receptor blockade with atropine (1 mg/kg, i.v. abolished seizure-induced bradycardia. Seizure-induced hypertension was unaffected by beta-adrenergic blockade with timolol (1 mg/kg, i.v.), but reduced by phentolamine (5€mg/kg, s.c., an alpha-adrenergic receptor antagonist). Chemical sympathectomy was induced with 6-hydroxydopamine (100 mg/kg, i.v.), an agent that does not cross the blood–brain barrier, eliminated the pressor response but did not completely block seizure-induced bradycardia. Effectiveness of 6-hydroxydopamine was tested with tyramine (0.5 mg/kg, i.v.), an agent that releases endogenous catecholamines. Results indicate amygdaloid kindled seizures activate both branches of the autonomic nervous system. Bradycardia was mediated by the parasympathetic system; the pressor response was caused by an increase in peripheral resistance due to alpha-adrenergic receptor activation. Findings show kindling is a useful seizure model for future studies on the effect of seizures on cardiovascular function and possible mechanisms of seizure-related sudden unexplained death (Goodman et al. 1989, 1990).

Source: Lathers, C. M., P. L. Schraeder, M. W. Bungo, in Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Nova Science Publishers, Inc., Hauppauge, NY, 2008. With permission.

8 Sudden Death in Epilepsy: Forensic and Clinical Issues

1986a, 1986b, 1986c, 1987, 1993, 2008a, 2008b, 2008c). In addition to the risk of neurogenically induced arrhythmias, neurogenic apnea could be associated with SUDEP. Table 1.2 summarizes the SUDEP risk factors of respiratory/autonomic changes and hypoxia. Rare clinical case reports describe incidences of apnea associated with epileptiform activity (So et al. 2000). Additionally, a well-known sheep model of epilepsy (Simon et al. 1982; Table 1.2â•…Respiratory Factors and Hypoxia Model, Symptoms, and/or Change Postmortem found multiple areas of pulmonary punctuate hemorrhages and large areas of gross hemorrhage and edema in animals dying after induced epileptogenic activity, asystole, or ventricular ἀbrillation (Lathers and Schraeder 1982; Lathers et al. 1984; Carnel et al. 1985; Schraeder and Lathers 1983). In addition to the general and neurological risk factors, there is increasing evidence that cardiac (Aurlien et al. 2009; Kerling et al. 2009; Pezzella et al. 2009; Strzelczyk et al. 2008) and pulmonary (Scorza et al. 2007; Tavee and Morris 2008) changes additionally predispose a person to SUDEP. As with neurological risk factors, cardiac and pulmonary risk factors have neither been investigated by prospective observational follow-up studies nor by intervention studies. The pathomechanisms are still not established (Johnston and Smith 2007). Finsterer and Stöllberger (2010, this book) discuss recent ἀndings and practical implications concerning potential cardiac and pulmonary risk factors and pathomechanisms of SUDEP.

Changes in cardiac function alter cerebral blood flow, which, in turn, produces central hypoxia resulting in epileptogenic activity.

Mechanisms of Sudden Death Tissue hypoxia, hypercarbia, and alterations in acid–base balance may have contributed to the results in our model of experimental epilepsy. Acid–base balance was maintained only within physiological range before the initiation of epileptogenic activity.

So far, there is minimal evidence that any primary pulmonary disease could be a deἀnite risk factor for SUDEP (Finsterer and Stöllberger, 2010, this book). It is also unknown if patients with muscular respiratory insufficiency are at increased risk not to survive a tonic–clonic seizure. These patients appear particularly endangered because they often also have epilepsy and their epilepsy is often difficult to treat. Despite this uncertainty about pulmonary risk factors, there are frequent reports about patients who develop severe pulmonary problems during or after seizures, such as ictal hypoxemia or hypercapnia (Bateman et al. 2008), apnea (Bell and Sander 2006; Jehi and Najm 2008; Ryvlin et al. 2009), acute neurogenic pulmonary edema (Jehi and Najm 2008), or postictal laryngospasm (Tavee and Morris 2008). In a prospective autopsy series on 52 SUDEP patients, 80% had pulmonary congestion and edema (Leestma et al. 1989). It is uncertain if these abnormalities are due to primary pulmonary, cardiac, or laryngeal mechanisms. It is also unclear if only patients with previous lung disease, as opposed to previously healthy subjects, develop such problems. There is little information available about the effects of generalized tonic–clonic seizures on the respiratory system in general. Do generalized seizures induce bronchospasm or loss of tone of the muscles involved in respiration? Recent investigations have shown that at least the vital capacity, forced vital capacity, and forced expiratory volume within the ἀrst second (FEV1 etc.), are not signiἀcantly different between healthy subjects and epilepsy patients (Scorza et al. 2007) (Finsterer and Stöllberger 2010, this book). Some patients exhibit changes in cardiovascular status preceding the onset of convulsions (Schott et al. 1977; Schraeder et al. 1983). (continued)

Neurocardiologic Mechanistic Risk Factors in SUDEP

9

Table 1.2â•…Respiratory Factors and Hypoxia (Continued) Model, Symptoms, and/or Change The pulmonary edema model of status epilepticus in unanesthetized, chronically instrumented sheep in which sudden death and pulmonary edema occur.

Audiogenic seizures: respiratory arrest. Central alveolar hypoventilation syndrome (Ondine’s curse). Failure of automatic involuntary respiration with preservation of voluntary respiratory drive (Ondine’s curse) is rare, reported following a variety of morphologic lesions near respiratory centers in the lower brainstem. Risk factors include uncontrolled convulsive seizures, as well as respiratory and cardiac factors relating to treatment and supervision.

Epilepsy-related hypoxia. Ictal apnea. Postictal central apnea appears to be one potential mechanism for SUDEP. A 55-s convulsive seizure occurred in a 20-year-old female as she underwent video-EEG monitoring (So et al. 2000). Persistent apnea then developed. Electrocardiogram monitoring rhythm was not altered for the

Mechanisms of Sudden Death Catecholamine levels and seizure type and duration did not differ between animals dying suddenly or those surviving. Benign arrhythmias were generated in all animals; in no case did an arrhythmia account for death of an animal. Striking hypoventilation demonstrated in sudden death group but not in surviving animals. Differences in peak left atrial and pulmonary artery pressures, and in extravascular lung water; pulmonary edema did not account for the demise of sudden death animals. Thus, this model of epileptic sudden death supports a role of central hypoventilation in etiology of sudden unexpected death and shows an association, albeit not fatal, with pulmonary edema. The importance of arrhythmia in its pathogenesis is not conἀrmed (Simon et al. 1982; Johnston et al. 1995, 1997). Respiratory arrest mechanisms, modulated in part by serotonin, may cause SUDEP (Tupal and Faingold 2006a, 2006b). Central alveolar hypoventilation syndrome causes sudden death in a 39-year-old woman with heterotopia of the inferior olive (Matschke and Laas 2007). Neuropathologic examination disclosed preexisting malformation of the lower brain stem and acute local subarachnoid bleeding. Both respiratory and cardiac mechanisms are important. The apparent protective effect of lay supervision supports a role for respiratory factors, in part amenable to intervention by simple measures. Malignant tachyarrhythmias are rare during seizures and sinus bradycardia/arrest, although infrequent, occurs. Both types of cardiac arrhythmias can have a genetic basis as a contributory factor. Authors explore the potential of coexisting liability to cardiac arrhythmias as a contributory factor, but acknowledge that bridging evidence between inherited cardiac gene determinants and SUDEP is lacking (Terrence et al. 1981; Coulter 1984; Schraeder 1987; Nashef et al. 2007). Central apnea with seizures (Schraeder 1987) and neurogenic pulmonary edema and adult distress syndrome (Terrence et al. 1981). (Penἀeld and Jasper 1954; Bobo and Bonvallet 1975) So et al. (2000) note that although epileptic seizures may be associated with arrhythmogenic actions at the heart, in this patient the mechanism of marked central suppression of respiratory activity after seizures was clearly involved and almost resulted in sudden death. This case and Dr. Schraeder’s (1983) case highlight that both respiratory and cardiac changes do occur in persons with epilepsy. The (continued)

10 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.2â•…Respiratory Factors and Hypoxia (Continued) Model, Symptoms, and/or Change ἀrst 10 s, then it gradually and progressively slowed and stopped 57 s later. Cardiorespiratory resuscitation was successful. No evidence of airway obstruction or pulmonary edema was noted. One previous cardiorespiratory arrest after a complex partial seizure without secondary generalization had been reported for this patient. Postictal respiratory arrest was induced by serotonin receptor inhibition and prevented by selective serotonin reuptake inhibitor drugs (Tupal and Faingold 2006a, 2006b). Audiogenic seizure in mice. Obstructive sleep-apnea–induced cardiovascular complications.

Likely processes of the sudden infant death syndrome (SIDS) are identiἀed (apnea, failed arousal, failed autoresuscitation, etc.). The way in which epidemiological risk factors, genetics, neurotransmitter receptor defects, and neonatal cardiorespiratory reflex responses interact to lead to sudden death during sleep is unclear. 5-HT SIDS: Role of medullary 5-HT abnormalities in pathogenesis of SIDS, putative in utero origin of these abnormalities, and environmental and genetic factors interact to ultimately result in sudden death of the infant (Patterson 2010, this book).

Mechanisms of Sudden Death timing of events such as seizures, respiratory and/or laryngospasm, and cardiac EKG changes do vary in different patients.

The role of serotonin in SUDEP must be examined in future animal studies, including the DBA mice model of SUDEP (Faingold et al. 2010, this book). Suggested for use to study postictal respiratory arrest (So 2008). Implicated in pathogenesis of various cardiovascular diseases, including systemic hypertension, coronary artery disease, congestive heart failure, pulmonary hypertension, stroke, and cardiac arrhythmias. Mechanisms by which obstructive sleep apnea affects the cardiovascular system may involve mechanical effects on intrathoracic pressure, increased sympathetic activation, intermittent hypoxia, and endothelial dysfunction (Malow et al. 2000; Jain 2007). It is hypothesized that the neurophysiological basis of SIDS resides in a persistence of fetal reflex responses into the neonatal period, such as ampliἀcation of inhibitory cardiorespiratory reflex responses and reduced excitatory cardiorespiratory reflex responses. Explores ways in which multiple subtle abnormalities interact to lead to sudden death and emphasizes difficulty of ante-mortem identiἀcation of infants at risk for SIDS (Leiter and Böhm 2007). Underlying vulnerability involves a developmental defect in brainstem serotonergic (5-HT) systems that results in failure of protective cardiorespiratory responses to potentially life-threatening, but normally occurring events (e.g., hypoxia, hypercapnia), in the infant during sleep (Patterson 2010, Chapter 5, this book).

Source: Lathers C. M., P. L. Schraeder, M. W. Bungo, in Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Nova Science Publishers, Inc., Hauppauge, NY, 2008. With permission.

Johnston et al. 1995, 1997) found that although neurogenic pulmonary edema was commonly observed, the mechanism of death in the animals was central neurogenic hypoventilation. Likewise, syncope is a factor for sudden death (Table 1.3). It is evident that, in all likelihood, there are multiple contributing risk factors that, in unfortuitous combinations (including environmental circumstances) in individuals with epilepsy, may result in unexpected death (Nashef et al. 1998). This chapter consists of summary tables highlighting possible risk categories and mechanisms for risks. The reader is encouraged to use the table

Neurocardiologic Mechanistic Risk Factors in SUDEP

11

Table 1.3â•… Syncope Symptoms and/or Changes Syncope is a transient loss of consciousness and postural tone. Usually due to temporary, self-terminating global cerebral hypoperfusion. Important to differentiate from other nonsyncopal transient loss of consciousness attacks (Chen et al. 2008).

Although arising suddenly from prolonged recumbency or returning from weightlessness to earth’s gravity can result in syncope from orthostatic or vasovagal effects, there are many other possible causes. Cardiac causes are more likely to occur in the elderly; noncardiac causes are more common in the younger population. Cases described illustrate often unexpected mechanisms of syncope in otherwise healthy individuals. Cases of sudden collapse. At times, there are obvious cardiac etiologies or central nervous system etiologies; at other times, there are interactive cause and effects with cardiac disorders leading to central nervous system effects or visa versa. However, all too often there are subtle interplays with genetic predispositions, environmental interaction, therapeutic interventions, or unknown organic disease. As consequences are extreme, a thorough investigation and understanding of the mechanisms leading to syncope are of paramount importance. Psychogenic syncope and psychogenic seizures are common disorders but are difficult to identify. Headupright, tilt-table testing evaluates vasovagally mediated syncope and convulsive syncope. In eight patients with syncope and/or tonic–clonic motor activity, without changes in blood pressure and heart rate, transcranial Doppler cerebral blood flow velocity, and EEG monitoring revealed they were, in all instances, psychogenic or malingering. Combined long term ECG and EEG monitoring. Isolated cardiac rhythm abnormalities were noted in 21 patients, but none were symptomatic and no deἀnitive arrhythmias occurred. Isolated EEG abnormalities were noted in 11 patients, 5 of whom had EEG abnormalities consistent with seizure disorders. Simultaneous EEG and ECG abnormalities were seen in four patients. In two, a previously unsuspected etiology for syncope was found: seizures in one patient with heart disease and sinus pauses in another thought to have a seizure disorder.

Mechanisms of Sudden Death The most important screening tool in identifying mechanism(s) of syncope is a detailed history emphasizing a search for underlying disease, speciἀc associated circumstances, and pre- and post-event symptoms. The type of diagnostic studies (i.e., cardiac or neurologic) undertaken should be based on historical data. Seizures must be considered as a possible mechanism of otherwise unexplained loss of consciousness in nonelderly persons, including air crew members (Schraeder et al. 1994; Williams and Frenneaux 2007). Review recent research relevant to managing syncope in adults: syncope evaluation in the emergency department, effectiveness of a structured and standardized approach to syncope, role of the implantable loop recorder, and efficacy of nonpharmacological physical treatments (Chen et al. 2008; Bungo et al. 2010, this book). Clinical cases briefly explore the interactions of disordered electrical potentials in the brain and disordered electrical potentials in the heart (Lathers et al. 2010b, Chapter 22, this book).

It was concluded that patients who pass out or convulse during head-upright tilt without any change in physiologic parameters can be presumed psychogenic in origin and may be referred for psychiatric evaluation without further expensive diagnostic studies (Grubb et al. 1992). Combined ambulatory EEG/ECG monitoring may prove useful in evaluation of some patients with syncope (Beauregard et al. 1991).

(continued)

12 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.3â•… Syncope (Continued) Symptoms and/or Changes Vasovagal syncope. Much of the natural history is unknown. Study determined whether patients presenting for care had a recent increase in worsened syncope frequency. Insufficient cerebral perfusion is a common cause leading to a loss of consciousness with a critical reduction of blood flow to the reticular activating system. In neurally mediated syncope, a paradoxical reflex can occur that induces an increase in cerebrovascular resistance and contributes to the critical reduction of cerebral blood flow.

Mechanisms of Sudden Death Many syncope patients present for care after a recent worsening of their frequency of syncope (Franco 2007; Sheldon et al. 2007). Outlined anatomic structures involved in cerebral autoregulation, its mechanisms in normal and pathologic conditions, and noninvasive neuroimaging techniques used to study cerebral circulation and autoregulation. Emphasis placed on description of autoregulation pathophysiology in orthostatic and neurally mediated syncope (Franco 2007).

Source: Modiἀed and updated from Lathers C. M., P. L. Schraeder, M. W. Bungo, in Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Nova Science Publishers, Inc., Hauppauge, NY, 2008. With permission.

of contents of this book to identify which chapters are pertinent to a given topic within each table. The reader is referred especially to chapters in this book discussing risks for SUDEP by Walczak (2010), arrhythmias (Lathers and Schraeder 1982, 1987; Lathers et al. 1987), LSP (Schraeder and Lathers 1983, 1989), respiratory issues (Scorza et al. 2007; Finsterer and Stöllberger 2010; Lathers and Schraeder 1982; Carnel et al. 1985), syncope (Sharma et al. 2010, Chapter 21, this book; Lathers et al. 2010b, Chapter 22, this book), genetics (Herreros 2010), stress (Lathers and Schraeder 2006), therapies (Lathers and Schraeder 2002, 2010; Lathers et al. 2003a, 2003b, 2003c; Tompson 2008, 2010; Ryvlin et al. 2009), and unusual factors that may modify SUDEP risk (Scorza et al. 2008a, 2008b, 2010a, 2010b; Calderazzo et al. 2009; Terra-Bustamante et al. 2009). All are encouraged to read Brodie and Holmes (2008) and the chapter presenting how to educate persons with epilepsy about their risk factors and how to help families and survivors of SUDEP deal with the unexpected loss of a person with epilepsy (Hanna and Panelli, 2010, Chapter 57). There is general agreement that central nervous and autonomic nervous system/cardiorespiratory interactions include arrhythmias and apnea (Tables 1.1 and 1.2). Complicating the nervous system/cardiac relationship are relatively recent discoveries of genetically determined predispositions to arrhythmias that may result in seizure-like events at the time of the acute cardiac dysfunction (Table 1.4). How these genetic cardiac predispositions interact with the central and autonomic nervous systems is not understood. It is generally accepted that certain associated clinical circumstances [e.g., male gender (Table 1.5), poorly controlled generalized tonic clonic seizures, use of multiple antiepileptic drugs, changing doses or drugs, withdrawal of antiepileptic drugs, and poor compliance with antiepileptic drug use] are associated with an increased risk of SUDEP (Table 1.6). That psychogenic factors associated with life stresses are risk factors for cardiac arrhythmias and sudden death is well recognized by cardiologists (Lathers and Schraeder 2006). Also, certain ethnic groups have an increased incidence for stress-related sudden deaths [e.g., bangungut in healthy Filipino men (Table 1.7)]. There has been almost no organized effort to determine the role of stress as a risk factor for SUDEP. The stress response involves acute or chronic increases in sympathetic neural activity. Furthermore, Scorza and colleagues have identiἀed unusual factors that may modify the risk for SUDEP such as ambient temperature, the lunar phases of the moon,

Neurocardiologic Mechanistic Risk Factors in SUDEP

13

Table 1.4â•… Genetic/Structural Mechanistic Factors Associated Changes Four main inherited arrhythmia syndromes that are thought to be responsible for sudden death. Two clinical cases exhibit both Brugada syndrome and epilepsy produced by sodium channel dysfunction (Lathers et al. 2010b, this book). Brugada syndrome is produced by a mutation in gene SCN5A, which encodes the alpha subunit of the cardiac sodium channel (Antzelevitch et al. 2005a, 2005b; Aurlien et al. 2009. Herreros (2010, Chapter 19) note that some epileptic syndromes (Graves 2006) are due to different mutations in genes encoding alpha subunits of neuronal sodium channels (SCN1A, SCN2A) or in the beta subunit (SCN1B), common for both cardiac and neuronal isoforms (Lehmann-Horn and Jurkat-Rott 1999). The long QT syndrome: A genetically transmitted cardiac arrhythmia due to ion channel protein abnormalities, affecting the transport of potassium and sodium ions across the cell membrane. Patients may present with syncope, seizures, or aborted cardiac arrest. Long QT syndrome is an important cause of unexplained sudden cardiac death in the young.

Cardiac hypertrophy: An independent predictor of cardiovascular morbidity and mortality, predisposition to heart failure, QT interval prolongation, and ventricular arrhythmias.

Mechanisms of Sudden Death Four syndromes: long QT syndrome, Brugada syndrome, short QT syndrome, and catecholaminergic polymorphic ventricular tachycardia reviewed by Herreros (2010, Chapter 19, this book). Additional clinical, electrocardiographic, and genetic studies are needed to improve individual risk stratiἀcation and to determine any relationship among sodium channel dysfunction, Brugada ECG, and idiopathic epilepsy. Two patients presented in the cases appear to support the possibility that a common pathophysiologic mechanism associated with sodium channel dysfunction may be common to ECG abnormalities of Brugada syndrome and some types of epilepsy. Many patients must be screened to conἀrm. Risks factors for sudden death must be deἀned and linked with mechanisms for death (Kornick et al. 2003; Lathers et al. 2008a, 2008b, 2008c; Herreros, 2010, Chapter 19; Lathers et al. 2010a, Chapter 20).

Diagnosis of long QT syndrome depends on an ECG showing a prolonged QT interval (Kiehne and Kauferstein 2007). Establishment of a registry and discovery of genetic mutations causing the syndrome contribute greatly to understanding this condition and impetus to understanding other inherited cardiac arrhythmias. Genotype-phenotype correlation studies allow risk stratiἀcation of long QT syndrome patients. Lifestyle modiἀcation to avoid triggers for malignant cardiac arrhythmias, and use of beta blockers, pacemakers, and implantable deἀbrillators may reduce mortality in these patients (Vohra 2007). Mutations of cardiac ion channel genes affecting repolarization cause the majority of congenital cases. Despite detailed molecular characterizations of mutated ion channels, understanding how individual mutations may lead to arrhythmias and sudden death requires study of intact heart and modulation by autonomic nervous system. Studies of molecularly engineered mice with mutations in genes known to cause long QT syndrome in humans and speciἀc to cardiac repolarization in mice are reviewed (Salama and London 2007). Cardiac angiotensin II overproduction leads to long QT syndrome, resulting from IK1 potassium-dependent prolongation of action potential duration via modulation of channel subunit expression (Domenighetti et al. 2007). (continued)

14 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.4â•… Genetic/Structural Mechanistic Factors (Continued) Associated Changes Inherited arrhythmia syndromes, channelopathies.

Short QT syndrome: ECGs show a shortened QT interval and tall and narrow T waves. The cause is mutation of the potassium channel genes. Cardiac hypertrophy: associated with a dramatic change in gene expression proἀle of cardiac myocytes. Many genes important during development of the fetal heart but repressed in adult tissue are reexpressed, resulting in gross physiological changes and arrhythmias, cardiac failure, and sudden death. One transcription factor possibly important in repressing expression of fetal genes in the adult heart is REST (repressor element 1-silencing transcription factor).

Genetic factors in SUDEP: Likely to reflect underlying heterogeneous mechanisms common to the brain and heart, as has been shown to be the case with sudden cardiac death and sudden infant death syndrome (Ghali and Nashef 2010, this book).

Mechanisms of Sudden Death Long QT syndrome (Kiehne and Kauferstein 2007), short QT syndrome, catecholaminergic polymorphic ventricular tachycardia, Brugada syndrome, and overlapping phenotypes; established connections between these syndromes and idiopathic ventricular ἀbrillation (Sarkozy and Brugada 2005). Persons with short QT syndrome have increased familial risk of sudden cardiac death (Gaita et al. 2003, 2004). REST expression prevents increased BNP (Nppb) and ANP (Nppa) gene, encoding brain and atrial natriuretic peptides. Adult rat ventricular myocytes response to endothelin-1 and inhibition of REST results in increased expression of these genes in H9c2 cells. Increased expression of Nppb and Nppa correlates with increased histone H4 acetylation and histone H3 lysine 4 methylation of promoter-proximal regions of these genes. Deletions of individual REST repression domains combined activities of two domains of REST required to efficiently repress transcription of Nppb gene. A single repression domain is sufficient to repress Nppa gene. Data provide insight into molecular mechanisms for changes in gene expression proἀle cardiac hypertrophy (Bingham et al. 2006, 2007). Overlapping mechanisms. The increasing number of mongenically associated epilepsy syndromes raises the question of how epilepsy and arrythmogenic genetic disturbances may be concurrent. The association of uncontrolled epilepsy with risk of SUDEP leads to speculation that in some cases of idiopathic epilepsy, a genetic mutation may also cause a predisposition to fatal cardiac arrhythmias. The recent association of long QT syndrome with epilepsy suggests that ion channel gene mutations may be inherited susceptibility factors for neurogenic cardiac arrhythmia in some persons with epilepsy. Further investigation into overlap of epileptogenic and arrhythmogenic epidemiological and genetic factors is warranted.

and other factors (Table 1.8). As is evident from the listings in the above-mentioned tables, there is a likelihood that common risk factors for SUDEP and arrhythmogenic cardiac disease are related to centrally initiated peripheral autonomic dysfunction in association with epileptiform discharges and stress. Issues that need resolution include a better understanding of the individual risks, the mechanisms of cardiac arrhythmia and arrest in persons without a previously identiἀed structural heart disease, a deἀnition of abnormal interactions between the central nervous system (CNS) and the heart, the role of neurogenic pulmonary edema and central apnea in combination with cardiac autonomic neural and subtle anatomic and genetic factors as risk for SUDEP, and development of primary and secondary preventive measures along with educational programs to disseminate essential information to physicians, patients, and families.

Neurocardiologic Mechanistic Risk Factors in SUDEP

15

Table 1.5â•…Risk Factors for SUDEP SUDEP Risk Factors in a Swedish Population (Nilsson et al. 2001)

Retrospective Study of Risk Factors, United States Reviewed association between several clinical variables and SUDEP to elucidate risk factors. Characteristics of 67 cases correlated with published previous studies. Education of physicians about existence of SUDEP and risk factors is imperative to improving patient education and reduction in mortality (Lear-Kaul et al. 2005). Behavioral Risk Factor Surveillance System Data from South Carolina Compared health insurance coverage, health care visits, health-related behaviors among persons with epilepsy vs. general population (Ferguson et al. 2008).

Position of Patients at Night (Kinney et al. 2009; Monter et al. 2007; McGregor and Wheless 2006).

1. Higher number of seizures per year (relative risk of 10.16 in patients with more than 50 seizures/year compared to no more than 2 years) 2. Increased number of antiepileptic drugs (9.89 for three drugs vs. monotherapy) 3. Early vs. late onset epilepsy (7.72) 4. Frequent changes in antiepileptic drug dosage vs. unchanged dosage (6.08) 5. Risk and early onset and SUDEP risk and seizure frequency was weaker for females 6. Frequent dosage changes had a stronger association in females 7. Early age of onset and male sex 1. Long history of seizure disorder 2. Under medication or poorly controlled seizure activity 3. Male gender 4. Age younger than 40 years 5. Mental or physical stress

Persons with epilepsy: 1. Are more likely to smoke. 2. Have less physical activity. 3. Need better access to health care. 4. Need interventions focused on smoking cessation and increase in physical activity. 5. Lack money to visit a doctor. One-third of respondents with active epilepsy reported in past 12 months needed to see a doctor but could not because of cost. Note: Authors do not relate these behavior risk factors to increased likelihood of SUDEP per se, but one could also consider them to be risk factors for SUDEP. 1. Especially important in persons with uncontrolled epilepsy in institutional settings. 2. Avoid prone position and soft pillows when sleeping. 3. Use respiration monitors during sleep. 4. Train caregivers to be vigilant and to intervene to prevent respiratory compromise for institutionalized and at-home patients.

1.3â•… SUDEP Animal Models The importance of using many different animal models to study SUDEP in order to glean insights into the various mechanisms of risks and their contribution to the initiation of the death event is discussed by Lathers (2010a, Chapter 25) (Table 1.1). So (2008) emphasizes the signiἀcance of using audiogenic seizure mice to study postictal respiratory arrest. Postictal respiratory arrest was induced by serotonin receptor inhibition and prevented by

16 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.6â•… Therapies: Increased or Decreased Risk Symptoms and/or Changes

Mechanisms of Sudden Death

Increased risk SUDEP The putative advantage of new drugs is a smaller spectrum of Antiepileptic drugs—New antiepileptic possible adverse events, such as sedation and some may drugs, e.g., topiramate and lamotrigine, minimize noncompliance by reducing side effects of developed for chronic focal and secondarily lethargy and cognitive impairment. Many new antiepileptic generalized epileptic seizures. Therapeutic drugs do have less frequent interactions, leading to efficacy of these drugs does not seem to be improved tolerability with comedication (Lathers et al. superior to traditional anticonvulsants such 2003a, 2003b, 2003c; Walczak 2003). However, difficulty in as phenobarbital. achieving therapeutic dosage with some of the newer antiepileptic drugs because of side effects makes one question whether some newer agents are better than the older antiepileptic drugs. Suggests SUDEP rates reflect population rates and not a Risk of SUDEP rates in patients on speciἀc drug effect. FDA requires warning labels on the risk lamotrigine, gabapentin, topiramate, of SUDEP in association with use of each of the mentioned tigabine, and zonisamide are similar to drugs (Lathers and Schraeder 2002). those on standard antiepileptic drugs. Compliance with antiepileptic drugs In a coroner’s office review of forensic cases in Allegheny important in prevention of SUDEP. County for the year 2001, Lathers et al. (2003a, 2003b) found low, or no, levels of antiepileptic drugs postmortem in persons with epilepsy who died of SUDEP. Hughes (2009) deemed the most important SUDEP risk Noninterventional, single-arm study (Martin factor to be noncompliance with antiepileptic medication. et al. 2009). Explored the effectiveness and Ryvlin et al. (2009) found the risk of SUDEP is increased in behavioral outcomes in intellectually patients who have poor compliance and exhibit nocturnal disabled patients with topiramate for seizures and generalized tonic–clonic seizures. However, epilepsy. compliance is not the only therapeutic question to ask about victims of SUDEP. Others include: Is the correct dose of the antiepileptic drug being used? Was more than one drug prescribed and were there recent dose drug changes? All agree that maintenance of a stable therapeutic drug levels is crucial to avoid SUDEP. Some improvement in nearly all behavioral aspects was observed under concomitant topiramate therapy. In addition, seizure frequency decreased. Two unexpected deaths in patients taking the antiepileptic drug topiramate were attributed to SUDEP (Martin et al. 2009). A prolonged QT interval leads to increased risk for QT-prolonging drugs and risk of SUD. development of ventricular tachyarrhythmias, particularly Torsades de pointes is associated with early polymorphic ventricular tachycardia (torsades de pointes). cardiac depolarization. Drugs that prolong Polymorphic arrhythmia may rapidly develop into QT interval: class III antiarrhythmic agents, ventricular ἀbrillation and cause sudden death antimicrobial agents (fluoroquinolone and (Reingardiene and Vilcinskaite 2007). macrolide antibiotics), antipsychotic and antidepressant drugs, agents used in general anesthesia, and antimycotics (Reingardiene and Vilcinskaite 2007). Adverse cardiovascular effects of antipsychotic treatment: Cardiovascular Effects of Antipsychotics tachycardia, orthostatic hypotension, and rarely, SUD, Older antipsychotic literature was primarily muscarinic cholinergic antagonism, alpha(1)-adrenergic concerned with physiological consequences antagonism, or receptors associated with cardiac of muscarinic cholinergic antagonism, (continued)

Neurocardiologic Mechanistic Risk Factors in SUDEP

17

Table 1.6â•… Therapies: Increased or Decreased Risk (Continued) Symptoms and/or Changes

Mechanisms of Sudden Death

Alpha(1)-adrenergic antagonism, or receptors associated with cardiac conduction, but current literature recognizes that, for most antipsychoticexposed patients, the more signiἀcant cardiovascular burden of treatment is mediated by metabolic adverse effects such as weight gain, dyslipidemia, and diabetes mellitus.

conduction, metabolic adverse effects of weight gain, dyslipidemia, and diabetes mellitus (Michelsen and Meyer 2007). High-dose methadone can induce QT prolongation by hERG inhibition, resulting in QT prolongation, but new evidence shows that QT prolongation can occur at much lower doses, even when the drug is not given IV.

Methadone produces QT prolongation, arrhythmias, and sudden death (Karch 2010, this book, Chapter 10).

May be acting directly on myocardial conduction to produce arrhythmia and death (Lathers and Lipka 1986, 1987; Lathers et al. 1986a; Lipka and Lathers 1987; Lipka et al. 1988).

Chlorpromazine or thioridazine do not appear to produce arrhythmia or death via a central locus in an experimental cat model. Decreased risk SUDEP The question must be asked if persons thought to be at risk for SUDEP should be placed on a beta blocker, in addition to the prescribed anticonvulsant(s). Beta blockers are also used to reduce stress and persons with epilepsy generally are stressed by the disease and associated problems (Lathers and Schraeder 2006). Beta blockers produce anticonvulsant activity whether administered via the intraosseous route or intravenously (Jim et al. 1988, 1989; Spivey et al. 1987a, 1987b; Lathers et al. 1989, 1990, 2008). Note the intraosseous and endotracheal routes of administration of antiepileptic drugs may be used when IV access is not available in status epilepticus. These routes are appropriate for emergency room pediatric seizures or cardiac arrest when unable to establish a traditional IV line (Rusli et al. 1987; Jim et al. 1988, 1989; Spivey et al. 1987a, 1987b; Lathers et al. 1989, 1990, 2008). Postmortem found no serious pathology in the tibia bone after intraosseous administration, other than the expected needle track (Lathers et al. 1989). Selective serotonin reuptake inhibitor drugs. Postictal respiratory arrest was induced by serotonin receptor May be protective in some persons at risk inhibition and prevented by selective serotonin reuptake for SUDEP with a component of apnea. inhibitor drugs. The role of serotonin in SUDEP must be examined in future animal studies (Tupal and Faingold 2006). Vagal nerve stimulation may decrease risk of SUDEP by 50% Vagal nerve stimulation may decrease (Annegars et al. 2000). susceptibility to ventricular tachycardias (Verrier et al. 2009; Verrier and Schachter 2010). Beta blockers exhibited anticonvulsant activity.

Source: Lathers C. M., P. L. Schraeder, M. W. Bungo, in Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Nova Science Publishers, Inc., Hauppauge, NY, 2008. With permission.

18 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.7â•… Psychological Factors Symptoms and/or Changes Scared to death. A 21-year-old student had generalized tonic–clonic seizures induced by mental image of human pain. One ictal event occurred while listening to a description of suffering, as read from Fox’s Book of Martyrs. While again listening to the offending passage during EEG and ECG monitoring, had 25 s of asystole terminating in electrocerebral silence and a generalized tonic, tonic–clonic seizure. A 24-hour ambulatory monitor recorded episodes of progressive sinus bradycardia concomitant with P–R interval prolongation and Wenckebach atrioventricular block. Sinoatrial conduction times and sinus node recovery times were normal on atrial pacing (Schraeder et al. 1983). Emotional trauma and psychological stress. Stress response is associated with increased sympathetic activity and catecholamine levels that may be associated with increased risk of cardiac arrhythmias, especially in context of epileptiform cerebral discharges (Pickworth et al. 1990). Stress is associated with changes in autonomic neural function. Its role as a potential risk factor for SUDEP is not known. Association of epilepsy with cardiac abnormalities, such as neurogenic arrhythmias and microscopic perivascular and interstitial ἀbrosis, and with depression and anxiety indicates emotional stress with related increases in catecholamines should be evaluated as a potential risk factor for SUDEP. Stress and anger. Compare morphology and initiation pattern between ventricular arrhythmias that are triggered by anger and those that are not. At the time of shock, patients with implantable cardioverter-deἀbrillators recorded levels of deἀned mood states preceding the shock in a diary. Stress risk factors and triggers of sudden death. An autopsy proven case-control study (Owada et al. 1999).

Mechanisms of Sudden Death Since implantation of a permanent pacemaker, has been asymptomatic. Patient demonstrates advantages of reproducing the circumstances associated with an unexplained loss of consciousness while monitoring EEG and ECG (Schraeder et al. 1983).

Arrhythmogenic effects of efferent sympathetic drive precipitate cardiac arrhythmia and sudden death. Patients with preexisting heart disease are particularly at risk. Generation of proarrhythmic activity patterns within cerebral autonomic centers may be ampliἀed by afferent feedback from a dysfunctional myocardium (Gray et al. 2007). Impact of adverse emotional states on autonomic control of cardiac rhythm is a known important factor leading to cardiac dysrhythmias in humans and other species. Interaction between emotional factors and arrythmogenic potential of epileptiform discharges and possibility of beneἀt from stress management intervention need to be investigated (Lathers and Schraeder 2006). Ventricular arrhythmias occurring in setting of anger are more likely pause dependent and polymorphic. Suggests that in predisposed populations anger may create an arrhythmogenic substrate susceptible to more disorganized rhythms, a possible mechanism linking emotion and sudden death (Stopper et al. 2006). Identiἀed risk factors and triggers of sudden death in cases where causes of death were deἀnitely proven by autopsy in Japan. Legal and medical records for 4 years were investigated. Of 271 cases, 176 patients, 20 to 59 yrs-old were cases of sudden death in working generations. Of these, 91 cases, 52% could be analyzed by telephone interviews from close family members. One examiner undertook all phone interviews with the case subjects. As control subjects, 1167 employed persons who consulted for a health check. Of sudden death cases, ἀnal diagnosis in 29 cases was coronary artery disease (31.9%), 18 acute cardiac dysfunction (continued)

Neurocardiologic Mechanistic Risk Factors in SUDEP

19

Table 1.7â•… Psychological Factors (Continued) Symptoms and/or Changes

Retrospective study. Reviewed association between several clinical variables and SUDEP to elucidate risk factors. Characteristics of 67 cases correlated with published ἀndings in previous studies (Lear-Kaul et al. 2005).

Mechanisms of Sudden Death (19.8%), 6 other cardiac diseases (6.6%), 4 acute aortic dissection (4.4%), 4 cerebrovascular disease (4.4%), and 30 other diseases (32.9%). Through conditional logistic analysis, the following risk factors emerged as candidates: Long-term stress, history of heart disease, hypertension, chest symptoms, autonomic disturbance, short-term stress, and a smoking habit. Short-term stress, autonomic disturbance and a smoking habit increased the risk of sudden death due to coronary artery disease. Long-term stress was associated with an increased risk of sudden death due to acute cardiac dysfunction. Demonstrated that autonomic disturbance and stress were closely related to occurrence of sudden death (Owada et al. 1999). Attributes that deἀne an at-risk group of epileptics include: less than 40 years old, male, long history of seizure disorder, under medication or poorly controlled seizure activity, and mental or physical stress. Education of physicians about the existence of SUDEP and risk factors is imperative in improving patient education and reduction in mortality (Lear-Kaul et al. 2005).

Source: Lathers C. M., P. L. Schraeder, M. W. Bungo, in Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Nova Science Publishers, Inc., Hauppauge, NY, 2008. With permission.

selective serotonin reuptake inhibitor drugs. The possible role of serotonin in SUDEP must be examined in future animal studies (Faingold et al. 2010, Chapter 41). The ἀring pattern of cardiac parasympathetic neurons, as well as cardiac sympathetic postganglionic nerves, during an epileptic attack has been shown to change and has been termed the lockstep phenomenon (Lathers et al. 1987; Stauffer et al. 1989, 1990; Dodd-O and Lathers 1990; O’Rourke and Lathers 1990). The observation of autonomic neural discharges time-locked to cortical epileptiform is evidence that epileptogenic activation of the cardiac parasympathetic nerves, revealed by ictal bradyarrhythmias or cardiac asystole, may be one contributing cause of sudden death of patients with epilepsy. Likewise, epileptogenic activation of the cardiac sympathetic nerves may be another contributing cause of SUDEP. An imbalance between the two systems’ (i.e., the cardiac parasympathetic and the cardiac sympathetic) neural discharge patterns may contribute to SUDEP (Lathers and Schraeder 1982, 1986; Schraeder and Lathers 1983, 1989). Wang et al. (2006) provided supporting data for the LSP ἀnding. They examined blockade of inhibitory neurotransmission evoked seizure-like ἀring of cardiac parasympathetic neurons in brainstem slices of newborn rats. Speciἀcally, blockade of GABAergic and glycinergic receptors in medulla slices evoked intermittent seizure-like ἀring of cardiac parasympathetic neurons, suggesting the seizure-like pattern of ἀring during an epileptic attack may cause neurogenic ictal bradyarrhythmias, cardiac asystole, or even sudden death in persons with epilepsy. Nonuniform autonomic cardiac postganglionic neural discharges are associated with coronary occlusion of the left anterior descending coronary artery and/or ouabain toxicity in cats (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978; Lathers 1980, 1981). Nonuniform

20 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.8â•…Unusual Risk Factors Modify SUDEP Factor Unusual potential risk factors for SUDEP were examined by Scorza and colleagues (2008a, 2008b, 2010a, 2010b), Calderazzo et al. (2009), and Terra-Bustaman et al. (2008).

Physical activity (Scorza et al. 2008a).

SUDEP seems to occur more commonly during sleep (Asadi-Pooya and Sperling 2009). Chronobiology of acute pulmonary edema (Bilora et al. 1998).

Mechanism of Sudden Death Studies are needed to examine the role of unusual factors that are listed below to determine if they are applicable to SUDEP. These include life style modifying interventions with accepted public health beneἀts, but as yet with no consensus that they may or may not prevent sudden death. Animal studies and clinical studies are needed (Lathers et al. 2008b, reply; 2010c, Unanswered question). May modify the role of autonomic effects and cardiorespiratory disturbances. Beneἀts of physical activity may be related to reductions in sympathetic activity. Morbidity and mortality in cardiovascular disease are often associated with elevations in sympathetic activity (Scorza et al. 2008a; Schraeder et al. 1983). Increased physical action does have a beneἀcial cardiovascular effect. The question is, will increased physical activity be of beneἀt if one assumes cardiovascular sympathetic dysfunction or insufficiency in persons with epilepsy? In general, regular exercise does not have a downside, if the patient is cardiovascularly ἀt. The role of regular exercise in prevention of SUDEP is not established (Lathers et al. 2008b reply; 2010c, this book). Autonomic cardiorespiratory disturbances. Alterations in clinical parameters including ECG changes, blood pressure, respiration, and vasomotor tone (Hirsch and Martin 1971; Lathers and Schraeder 1982; Schraeder and Lathers 1983; Wannamaker 1985). ECG changes in experimental epilepsy include heart rate changes, arrhythmias, conduction blocks, altered ECG morphology, and QT interval). See discussion of Bilora et al. (1998) below and Chapter 23 by Hughes and Sato (2010, this book). Found pulmonary edema incidence higher during the night. No signiἀcant weekly or circannual rhythms detected. Examine possible relations with decreased heart output, increased sympathetic tone, or partial baroceptor desensitization occurring at night (Bilora et al. 1998). Evaluation of some cases of SUDEP and ners-SUDEP during long-term electroencephalography monitoring suggest that autonomic instability ending in cardiorespiratory arrest may be triggered by postictal suppression rather than by ictal activation of the autonomic nervous system. More epidemiologic studies on high-risk populations of persons with epilepsy are needed (Schuele 2009). The increase sympathetic tone or desensitization of baroceptors may contribute to occurrence of cardiac autonomic arrhythmias. See discussions of interictal activity and cardiac arrhythmias (continued)

Neurocardiologic Mechanistic Risk Factors in SUDEP

21

Table 1.8â•…Unusual Risk Factors Modify SUDEP (Continued) Factor

Seizure activity and neurogenesis (Scorza et al. 2008b). Omega-3 fatty acid nutritional deἀciency (Scorza et al. 2008a).

Low temperature (Scorza et al. 2008a).

Winter temperatures may lead to cardiac abnormalities and, hence, to sudden death. Are they a risk factor for SUDEP? (Colugnati et al. 2008).

Role of heart rate in rats with epilepsy during low temperature exposure (Sonoda et al. 2008). Minimum external temperature, mean external temperature (Bell et al. 2009). Climate fluctuations (Calderazzo et al. 2009). Sunlight may be a beneἀt in that vitamin D deἀciency may increase risk of SUDEP (Scorza et al. 2010a).

Mechanism of Sudden Death and baroreceptor changes by Lathers (2010a, 2010b) and cardiac and pulmonary risk factors and pathomechanisms of SUDEP by Finisterer and Stollberger (2010, Chapter€42). Aberrant dentate granule cell neurogenesis may influence negatively the cardiovascular system of a patient with epilepsy and lead to cardiac abnormalities and risk of SUDEP. Beneἀcial effect of nutritional aspects of omega-3 fatty acid may decrease cardiac arrhythmias and sudden death (SUD) in patients at risk for cardiac disease. No established answer at this time. See comments by Lathers et al. (2008b). Questions have been raised (Lathers et al. 2008a, 2008b, 2008c, 2010c) that need to be addressed: 1. Exactly how does one deἀne a low temperature? 2. What is the degree range used to deἀne a low temperature? 3. What is the duration at this low degree range? 4. Is there a dose response so that there is an increased risk of sudden death related to the lowest temperature? Mammals hibernate as a strategy to survive cold conditions. Hibernating mammals inherit a stable cardiovascular function. They show resistance to hypothermia at a cellular level, the membrane potential and excitability are more stable in their cardiac cells. Aortic smooth muscle cells maintain ionic gradients upon prolonged exposure to low temperature, cardiac myocytes from the mammals maintain constant levels of intracellular free calcium and forceful contractility at 10°C or lower. Postulate hibernating mammals have cardiovascular particularities that confer heart protection. The relevance to persons with epilepsy is to be determined. Low temperature increased the heart rate of patients with epilepsy. Authors suggest exposure to low temperatures could be a risk factors for cardiovascular abnormalities and, thus, for SUDEP. Found no evidence of association between either mean temperature group or minimum temperature group and SUDEP but there was a slight excess of SUDEP in the coldest (mean temperature) groups. May have an effect on SUDEP occurrence. Some clinical studies suggest low vitamin D levels may be associated with death from heart failure and sudden cardiac death. There also appears to be an association between low vitamin D and seizure occurrence. Scorza et al. suggest vitamin D may exhibit anticonvulsant action. More data are needed. (continued)

22 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 1.8â•…Unusual Risk Factors Modify SUDEP (Continued) Factor The lunar phase (Terra-Bustamante et al. 2009).

Lunar phase, month, season: no association found between these factors and SUDEP (Bell et al. 2009).

Mechanism of Sudden Death An epilepsy unit performed an 8-year analysis, examining possible associations between the phase of the moon and SUDEP. Incidence of SUDEP was highest at full moon (70%), followed by waxing moon (20%), and new moon (10%). No SUDEPs occurred during the waning cycle. Preliminary ἀndings suggest the full moon correlates with occurrence of SUDEP. Dataset of National Sentinel Clinical Audit of EpilepsyRelated Deaths examined all death certiἀcates of those who died in England and Wales between 9/1999 and 8/2000 to see if epilepsy was mentioned. Of 2412 deaths, 409 were identiἀed as probable SUDEP.

(increases, decreases, and/or no change) autonomic neural postganglionic cardiac sympathetic discharge traveling through the stellate ganglia, causes cardiac arrhythmias, ventricular ἀbrillation, and/or sudden cardiac death in the manner described by Han and Moe (1964), i.e., nonuniform recovery of excitability in ventricular muscle. It was suggested that the modiἀcation of neural discharge is a way for drugs to modify the occurrence of arrhythmia and death (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978; Lathers 1980, 1981; Han and Moe 1964). Nonuniform autonomic cardiac postganglionic neural discharge was also found to be associated with pentylenetetrazol-induced interictal epileptogenic activity in cats. Nonuniform autonomic neural discharge and autonomic neural imbalance of postganglionic cardiac sympathetic and vagal discharge cause cardiac arrhythmias, ventricular ἀbrillation, asystole, and/or SUDEP. Regional differences in the beta adrenoceptor densities reflects differences in postganglionic cardiac sympathetic innervation of the myocardium (Randall 1977, 1984; Randall et al. 1984). These site differences will vary the release of norepinephrine in various sympathetic nerve terminals and sites in the heart to modify cardiac contractile function and may trigger development of arrhythmias and/or sudden death. Innervation density of beta adrenoceptor densities is high in the subepicardium and central conduction system. Nonuniform postganglionic cardiac-sympathetic cardiac innervation is related to the nonuniform beta-sympathetic receptor locations in the heart and affects cardiac contractility and development of arrhythmias and/or death. In diseased hearts, cardiac innervation density varies and may lead to sudden cardiac death (Lathers et al. 1985, 1986a, 1986b, 1988, 1990, 2010; Lathers and Levin 2010, Chapter 33, this book). Seizures exert effects on cardiac function (Schuele, 2009). Since tachyarrhythmias are common during epileptic seizures, whereas bradyarrhythmias or asystoles occur less frequently, Kerling et al. (2009) evaluated cardiac postganglionic denervation in patients with epilepsy to determine if this was responsible for ictal asystole. I-(123)-metaiodobenzylguanidine was used as a marker of postganglionic cardiac norepinephrine uptake, using single-photon emission computed tomography. They concluded that a pronounced reduction in cardiac single-photon emission computed tomography uptake of asystole patients indicates postganglionic cardiac catecholamine disturbance. Impaired sympathetic cardiac innervation limits adjustment and heart-rate modulation and may increase the risk of asystole and, eventually, sudden unexpected death in persons with epilepsy. These data support the ἀnding of Lathers and colleagues described above. The

Neurocardiologic Mechanistic Risk Factors in SUDEP

23

reader is also referred to the chapter by Lathers and Levin (2010) that examines nonuniform sympathetic innervation and beta-sympathetic receptor locations and discusses how these factors affect cardiac contractility and the development of cardiac arrhythmias and/ or sudden death. Cerebral ischemia is associated with neuron degeneration. Accumulation of excess excitatory amino acids in the synaptic clef, activation of excitatory amino acid receptors, and influx of calcium into neurons are involved in the development of ischemia-induced neuronal death. Schwartz et al. (1995) hypothesized that neuroprotection may occur if inhibitory transmission via gamma-aminobutyric acid (GABA) is enhanced to offset excitation. Diazepam, a drug known to increase GABA-induced chloride channel opening, was studied in rats after hippocampal GABA levels had returned to post-transient global ischemia basal levels. The authors concluded that delayed enhancement of GABAergic neurotransmission directly at the site of vulnerability after an ischemic event does protect the vulnerable neurons from death. This model should be adapted to an animal model of seizure activity to study the effect of diazepam on GABA-mediated effects that may prevent ischemia-induced neuronal death, prevent the worsening of central neuronal communication due to epileptogenic activity, and eventually contribute a protective CNS effect that lessens individual at risk for SUDEP. Table 1.2 discusses postmortem multiple areas of punctuate hemorrhages and large areas of gross hemorrhage and edema in animals dying after inducing epileptogenic activity, asystole, or ventricular ἀbrillation (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers et al. 1984; Carnel et al. 1985). Tissue hypoxia, hypercarbia, and alterations in the acid–base balance may have contributed to the results in our model of experimental epilepsy. Acid–base balance was maintained only within physiological range before the initiation of epileptogenic activity (Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1983, 1989; Carnel et al. 1985; Stauffer et al. 1989, 1990). Changes in cardiac function alter cerebral blood flow, which, in turn, produces central hypoxia that results in epileptogenic activity. Some patients exhibit changes in cardiovascular status preceding the onset of convulsions (Schott et al. 1977; Schraeder et al. 1983). So (2008) has emphasized risk factors of postictal apnea and hypoxia, with or without pulmonary congestion, in combination with generalized tonic–clonic seizures and, to a lesser degree, with complex partial seizures and respiratory arrest as a mechanistic cause of SUDEP. So et al. (2000) also noted that, although epileptic seizures may be associated with arrhythmogenic actions of the heart, in their patient, the mechanism of marked central suppression of respiratory activity after seizures was clearly involved and almost always resulted in sudden death. This case, and the case of Schraeder et al. (1983), highlight that both respiratory and cardiac changes do occur in persons with epilepsy. The timing of events such as seizures, respiratory and/or laryngospasm, and cardiac EKG changes does vary in different patients. Ryvlin et al. (2009) note that the pathophysiology of SUDEP is not clear, but postictal central or obstructive apnea is one likely mechanism. However, as noted by Lathers (2010b, Chapter 44, this book): Caution must be exerted when concluding respiratory changes alone are the primary mechanism of death. At the ἀrst onset of the clinical problem cardiac arrhythmias may be felt by the patient but may not be visually detected by a witness. ‘Invisible’ cardiac arrhythmias may be initiated and then followed by ‘visible’ respiratory distress. Therefore, in addition to repositioning of the patient to ensure ease of respiration and/or stimulation of respiration, it is important, if possible, also to simultaneously monitor and medically support cardiac rate and rhythm.

24 Sudden Death in Epilepsy: Forensic and Clinical Issues

The reader is referred to additional discussion of actual case histories focused on arrhythmogenic, respiratory, and psychological risk factors in Chapter 44 of this book (Lathers 2010b). Most likely, different mechanisms and/or a different combination of mechanisms are responsible for death in different persons with epilepsy. 1.3.1â•… Therapies as Factors Data suggest that beta blockers exert a protective effect against seizure induction and/ or the development of cardiac arrhythmias with interictal and ictal activity (Spivey et al. 1987b; Jim et al. 1988, 1989; Lathers et al. 1989a, 1989b, 1990, 2008a) (Table 1.6). The question must be asked whether persons thought to be at risk for SUDEP should be placed on a beta blocker, in addition to the prescribed anticonvulsant(s). Celiker et al. (2008) reported clinical experiences of patients with catecholaminergic polymorphic ventricular tachycardia and concluded that medical treatment with propranolol and verapamil may decrease the incidence of arrhythmia. If patients are still refractory, implantation of intracardiac deἀbrillators should be considered. Obviously, a delay in diagnosis or inadequate treatment can result in sudden cardiac death. The same is true for persons at risk for SUDEP. See the case history of SUDEP by Schraeder (2010). Hughes (2009) addressed the issue of how to predict patients at risk for SUDEP by reviewing published reports of such deaths. With a mean incidence of SUDEP at 1.8/1000, a mean standardized mortality ratio of 6.8, and a mean percentage of SUDEP cases among deaths from epilepsy at 16.6%, the problem of SUDEP and risk factors is an important issue to be resolved. Hughes identiἀed 17 risk factors and concluded, just as Lathers et al. (2008a, 2008b, 2008c) stated, that a cardiac or pulmonary problem may be a primary risk factor in different patients. He deemed the most important risk factor to be noncompliance with antiepileptic medication. Maintenance of therapeutic drug levels is crucial to avoid SUDEP. Ryvlin et al. (2009) also found that the risk of SUDEP is increased in patients who have poor compliance, nocturnal seizures, and generalized tonic–clonic seizures. However, compliance is not the only risk factor to be addressed if a person is a victim of SUDEP. There are several postmortem questions to be asked (Lathers and Schraeder 2009) about a patient on an antiepileptic drug that still becomes a SUDEP victim. One must inquire whether the correct dose of the antiepileptic drug was being used to control seizures. Another clinical pharmacology question that must be asked is whether the patient was on the correct antiepileptic drug to control the particular type or mixture of seizures experienced. When evaluating the role of drugs as protectors of life, clinical pharmacologists (Lathers and Schraeder 2002) caution us to remember that the use of all drugs is a risk/beneἀt ratio evaluation (Lathers et al. 2003a, 2003b, 2003c). Thus, the use of antiepileptic drugs may not provide 100% protection to the patient against sudden death.

1.4â•…Case of SUDEP with a Premorbid Diagnosis of Nonepileptic Seizures Table 1.8 discusses unusual risk factors examined by the laboratory of Scorza to determine if they are risk factors for SUDEP and/or have been speculated to be risk factors. This includes asking whether the lunar phase has an effect on SUDEP (Terra-Bustamante et al. 2009). A retrospective examination of the incidence of SUDEP in an epilepsy unit over an

Neurocardiologic Mechanistic Risk Factors in SUDEP

25

A 35-year-old woman was found dead in bed by her parents. She had a longstanding psychiatric history with the occurrence of seizure-like events. On one occasion,€she had a nonepileptic seizure in her neurologist’s office consisting of sliding to the floor with bizarre asynchronous bilateral motor activity without loss of consciousness. This event occurred consequent to an emotionally tense situation at her home. Past observation of another type of event consisting of some automatisms and postÂ� ictal confusion complicated the history. Multiple routine EEGs over the years were unremarkable, save for one that showed unequivocal isolated left temporal interictal discharges. The patient was placed on carbamazepine and found to have consistent therapeutic levels. The patient recognized that the complex partial events were no longer occurring, but that others continued. After the retirement of her neurologist, the patient was seen at another center and subjected to several days of inpatient EEG monitoring, during which multiple clinical events were observed without any epileptiform activity being documented on any of the EEG recordings. As a result, the patient was informed that her seizures were nonepileptic and was advised to taper her antiepileptic medications. Several weeks later her parents notiἀed her former neurologist of her demise. Co mment s by L at hers, Schraeder, and Bung o Persons with only nonepileptic seizures would, by deἀnition, not be at risk for SUDEP. However, the unfortunate reality is that from 10% to 30% of persons who appear to have long-established nonepileptic seizures also have epilepsy (Betts 1997). Thus, although the majority of persons with nonepileptic seizures do not have concurrent epilepsy, in those who also have a bona-ἀde seizure disorder, as demonstrated in this case, there is a risk of SUDEP associated with withdrawal of antiepileptic drugs based on the observation of only nonepileptic events. The physician must consider all aspects of the history and all prior EEG data before having conἀdence that the antiepileptic medication can be safely withdrawn. One should also keep in mind that rapid discontinuation of medication that was at a therapeutic serum level could induce a withdrawal seizure even in a person without epilepsy, further clouding the issue of diagnosis. From Betts, T. 1997. Psychiatric aspects of nonepileptic seizures. In Epilepsy: A Comprehensive Textbook, ed. J. Engel and T. A. Pedley. Philadelphia, PA: LippincottRaven Publishers.

8-year period was reviewed to look for a possible association. The review found the number of SUDEP cases was highest at the time of a full moon (70%), followed by a waxing moon (20%), and a new moon (10%). There were no SUDEP cases during the waning cycle. The authors concluded that a full moon appears to correlate with SUDEP. This same laboratory has examined other factors that may be involved in SUDEP, from regular imbibing of sardines and salmon to the influence of climate fluctuations (Calderazzo et al. 2009). Scorza et al. (2008b) have raised the question of whether seizure activity also influences dentate granule cell neurogenesis, since neurogenesis persists throughout life in adult mammalian dentate gyrus and is regulated by environmental, physiological, and

26 Sudden Death in Epilepsy: Forensic and Clinical Issues

molecular factors. The presence of hilar ectopic dentate granule cells was studied after status-�epilepticus was induced experimentally and it was determined that these cells migrate aberrantly, are abnormally integrated, and that the resulting hyperexcitability may contribute to seizure generation and/or propagation. Since high seizure frequency is a potential risk factor for SUDEP, the authors hypothesize that cardiac arrhythmias during and between seizures or transmission of epileptic activity to the heart via the autonomic nervous system may play a role. Thus, the aberrant neurogenesis may negatively influence the cardiovascular system of the patient with epilepsy and lead to cardiac abnormalities and then to the unwanted event of SUDEP. Studies are needed to examine the role in SUDEP played by these unusual factors to determine if they are risk factors.

1.5â•…Discussion 1.5.1â•… Future Steps The American Epilepsy Society and the Epilepsy Foundation Joint Task Force on SUDEP (So et al. 2008) assessed knowledge about SUDEP and recommended the following steps: 1. Hold a multidisciplinary workshop to reἀne current lines of investigation and identify additional areas of research for mechanism underlying SUDEP. 2. Conduct a survey of patients, families, and caregivers to identify effective means of education to enhance participation in SUDEP research. 3. Campaign for emphasis of the need for complete autopsy examinations for patients with suspected SUDEP. 4. Secure infrastructure grants to fund a consortium of centers to conduct prospective clinical and basic research studies to identify preventable risk factors and mechanisms underlying SUDEP. During the interim before these steps have been achieved, it is most important to provide prompt and optimal control of seizures, especially generalized convulsive seizures, to prevent SUDEP. A global focus is needed to resolve the risk factors for and mechanisms of epilepsy and sudden death (Lathers 2009).

References Annegers, J. F., S. P. Coan, W. A. Hauser, and J. Leestma. J. 2000. Epilepsy, vagal nerve stimulation by the NCP system, all-cause mortality, and sudden, unexpected, unexplained death. Epilepsia 41: 549–553. Antzelevitch, C. 2005a. Cardiac repolarization. The long and short of it. Eurpace 7 (Suppl 2): 3–9. Antzelevitch, C. 2005b. Modulation of transmural repolarization. Ann N Y Acad Sci 1047: 314–323. Asadi-Pooya, A. A., and M. R. Sperling. 2009. Clinical features of sudden unexpected death in epilepsy. J Clin Neurophysiol Aug 24. [Epub ahead of print]. Aurelien, D., T. P. Leren, E. Taubøll, and L. Gjerstad. 2009. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18: 158–160. Bateman, L. M., C. S. Li, and M. Seyal. 2008. Ictal hypoxemia in localization-related epilepsy: Analysis of incidence, severity and risk factors. Brain 131: 3239–3245.

Neurocardiologic Mechanistic Risk Factors in SUDEP

27

Beauregard, L. A., R. Fabiszewski, C. H. Black et al. 1991. Combined ambulatory electroencephalographic and electrocardiographic recordings for evaluation of syncope. Am J Cardiol 68: 1067–1072. Bell, G. S., and J. W. Sander. 2006. Sudden unexpected death in epilepsy. Risk factors, possible mechanisms and prevention: a reappraisal. Acta Neurol Taiwan 15: 72–83. Bell, G. S., J. L. Peacock, and J. W. Sander. 2009. Seasonality as a risk factor for sudden unexpected death in epilepsy: A study in a large cohort. Epilepsia Oct 27. [Epub ahead of print]. Betts, T. 1997. Psychiatric aspects of nonepileptic seizures. In Epilepsy: A Comprehensive Textbook, ed. J. Engel and T. A. Pedley, Chapter 199. Philadelphia, PA: Lippincott-Raven. Bilora, F., G. Vettore, C. Maifredini, S. Rocco, M. Pastorello, and F. Petrobelli. 1998. Chronobiology of acute pulmonary edema in an emergency service. Cardiologia 43: 303–307. Bingham, A. J., L. Ooi, L. Kozera, E. White, and I. C. Wood. 2007. The repressor element 1-silencing transcription factor regulates heart-speciἀc gene expression using multiple chromatinÂ�modifying complexes. Mol Cell Biol 27: 4082–4092. Bingham, A. J., L. Ooi, and I. C. Wood. 2006. Multiple chromatin modiἀcations important for gene expression changes in cardiac hypertrophy. Biochem Soc Trans 34 (Pt 6): 1138–1140. Bobo, E. G., and M. Bonvallet. 1975. Amygdala and masseteric reflex: I. Facilitation, inhibition and diphasic modiἀcations of the reflex, induced by localized amygdaloid stimulation. Electroencephalogr Clin Neurophysiol 39: 329–339. Brodie, M. J., and G. L. Holmes. 2008. Should all patients be told about sudden unexpected death in epilepsy (SUDEP)? Pros and Cons. Epilepsia 49 (Suppl 9): 99–101. Calderazzo, L., R. M. Arida, R. M. Cysneiros, E. A. Cavalheiro, and F. A. Scorza. 2009. From sardines to salmon: Influence of climate fluctuations on sudden unexpected death in epilepsy. Epilepsy Behav Jan. 20. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazolâ•‚induced epileptogenic activity. Pharmacology 20: 225–240. Celiker, A., I. Erdogan, T. Karagoz, and S. Ozer. 2008. Clinical experiences of patients with catecholaminergic polymorphic ventricular tachycardia. Cardiol Young 23: 1–8. Chen, L. Y., D. G. Benditt, and S. Win-Kuang. 2008. Management of syncope in adults: An update. Mayo Clin Proc 83: 1280–1293. Colugnati D. B., R. M. Arida, S. L. Cravo, G. H. Schoorlemmer, A. C. de Almeida, E. A. Cavalheiro, and F. A. Scorza. 2008. Hibernating mammals in sudden cardiac death in epilepsy: What do they tell us? Med Hypotheses 70: 929–932. Coulter, D. L. 1984. Partial seizures with apnea and bradycardia. Arch Neurol 41: 173–174. Di Maio, T. G., and J. V. M. Di Maio, eds. 2006. Excited Delirium Syndrome. Cause of Death and Prevention. Boca Raton, FL: CRC Press. Dodd-O, J. M., and C. M. Lathers. 1990. A characterization of the lock step phenomenon in Phenobarbital treated cats. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 13, 199–220. New York, NY: Marcel Dekker. Domenighetti, A. A., C. Boixel, D. Cefai, H. Abriel, and T. Pedrazzini. 2007. Chronic angiotensin II stimulation in the heart produces an acquired long QT syndrome associated with IK1 potassium current downregulation. J Mol Cell Cardiol 42: 63–70. Druschky, A., M. J. Hiltz, P. Hopp et al. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [123I]metaiodobenzylguanidine-SPECT. Brain 124: 372–382. Everengul, H., H. Tanriverdi, D. Dursunoglu et al. 2005. Time frequency domain analyses of heart rate variability in patients with epilepsy. Epilepsy Res 63: 131–139. Faingold, C. L., S. Tupal, Y. Mhaskar, and V. V. Uteshev. 2010. DBA mice as models of sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 41. Boca Raton, FL: CRC Press.

28 Sudden Death in Epilepsy: Forensic and Clinical Issues Ferguson, P. L., J. Chiprich, G. Smith et al. 2008. Prevalence of self-reported epilepsy, health care access, and health behaviors among adults in South Carolina. Epilepsy Behav 13: 529–534. Finsterer, J., and C. Stöllberger. 2010. Cardiac and pulmonary risk factors and pathomechanisms of sudden unexplained death in epilepsy patients (SUDEP). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 42. Boca Raton, FL: CRC Press. Franco, F. A. 2007. Cerebral autoregulation and syncope. Prog Cardiovasc Dis 50: 49–80. Gaita, F., C. Giustetto, F. Bianchi et al. 2003. Short QT syndrome: A familial cause of sudden death. Circulation 108: 965–970. Gaita, F., C. Giustetto, F. Bianchi et al. 2004. Short QT syndrome: Pharmacological treatment. J Am Coll Crdiol 43: 1494–1499. Gary-Bobo, E. G., and M. Bonvalette. 1977. Commissural projection to the amygdala through the ἀmbria fornix system in the cat. Exp Brain Res 27: 61–70. Ghali and Nashef. 2010. Genetics of SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues,€ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 18. Boca Raton, FL: CRC Press. Gleick, J. 1987. Chaos: Making a New Science. New York, NY: Viking Press. Goldberger, A. L., V. Bhargava, and B. J. West. 1985. Nonlinear dynamics of the heart beat. Physica 17D: 207–214. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1989. Kindled seizures activate both branches of the autonomic nervous system. Epilepsy Res 34: 169–176. Goodman, H. J., R. W. Homan, and I. L. Crawford. 1990. Kindled seizures elevate blood pressure and induce cardiac arrhythmias. Epilepsia 31: 489–495. Goodman, J. H., R. E. Berger, and T. K. Tcheng. 2005. Preemptive low-frequency stimulation decreases the incidence of amygdala-kindled seizures. Epilepsia 46: 1–7. Graves, T. D. 2006. Ion channels and epilepsy. Q J Med 99: 201–217. Gray, M. A., P. Taggart, P. M. Sutton et al. 2007. A cortical potential reflecting cardiac function. Proc Natl Acad Sci USA 104: 6818–6823. Grubb, B. P., G. Gerard, D. A. Wolfe et al. 1992. Syncope and seizures of psychogenic origin: identiἀcation with head-upright tilt table testing. Clin Cardiol 15: 839–842. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hanna, N. J., and R. J. Panelli. 2010. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, J. E. Leestma, Chapter 57. Boca Raton, FL: CRC Press. Herreros, B. 2010. Cardiac channelopathies and sudden death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 19. Boca Raton, FL: CRC Press. Hirsch, C. S., and D. L. Martin. 1971. Unexpected death in young epileptics. Neurology 21: 682–690. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14: 280–287. Ieda, M., H. Kanazawa, Y. Ieda et al. 2006. Nerve growth factor is critical for cardiac sensory innervation and rescues neuropathy in diabetic hearts. Circulation 114: 2351–2363. Ieda, M., H. Kanazawa, K. Kimura et al. 2007. Sema3a maintains normal heart rhythm through sympathetic innervation patterning. Nat Med 13: 604–612. Ieda, M., K. Kimura, H. Kanazawa, and K. Fukuda. 2008. Regulation of cardiac nerves: a new paradigm in the management of sudden cardiac death? Curr Med Chem 15: 1731–1736. Jain, V. 2007. Clinical perspective of obstructive sleep apnea-induced cardiovascular complications. Antioxid Redox Signal 9: 701–710. Jehi, L., and I. M. Najm. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Clevel Clin J Med 75 (Suppl 2): S66–S70. Jim, K. F., C. M. Lathers, V. L. Farris, L. F. Pratt, and W. H. Spivey. 1989. Suppression of pentylenetetrazol elicited seizure activity by intraosseous lorazepam in pigs. Epilepsia 30: 480–486.

Neurocardiologic Mechanistic Risk Factors in SUDEP

29

Jim, K. F., C. M. Lathers, W. H. Spivey, W. D. Matthews, C. Kahn, and K. Dolce. 1988. Suppression of pentylenetetrazolâ•‚elicited seizure activity by intraosseous propranolol in pigs. J Clin Pharmacol 28: 1106–1111. Johnston, A., and P. Smith. 2007. Sudden unexpected death in epilepsy. Expert Rev Neurother 7: 1751–1761. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42: 588–594. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37: 531–537. Karch, S. 2010. Drug abuse and SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 10. Boca Raton, FL: CRC Press. Kerling, F., M. Dutsch, R. Linke, T. Kuwert, H. Stefan, and M. J. Hilz. 2009. Relation between ictal asystole and cardiac sympathetic dysfunctions shown by MIBG-SPECT. Acta Neurol Scand 120: 123–129. Kiehne, N., and S. Kauferstein. 2007. Mutations in the SCN5A gene: Evidence for a link between long TQ syndrome and sudden death? Forensic Sci Int Genet 1: 170–174. Kimura, K. M. Ieda, H. Kanazawa et al. 2007. Cardiac sympathetic rejuvenation: a link between nerve function and cardiac hypertrophy. Circ Res 100 (12): 1755–1764. Kinney, H., A. Chadwick, L. A. Crandall, M. Grafe, D. L. Armstrong, W. J. Kupsky, F. L. Trachtenberg, and H. F. Krous. 2009. Sudden death, febrile seizures, and hippocampal maldevelopment in toddlers: A new entity. Pediatr Dev Pathol Jul 16: 1. [Epub ahead of print]. Kornick, C. A., M. J. Kilborn, J. Santiago-Palmer et al. 2003. QTc interval prolongation associated with intravenous methadone. Pain 105 (3): 499–506. Lathers, C. M. 1980. Effect of timolol on postganglionic cardiac and preganglionic splanchnic sympathetic neural discharge associated with ouabainâ•‚induced arrhythmia. Eur J Pharmacol 64: 95–106. Lathers, C. M. 1981. Carnivores. Induced disease. Myocardial infarction. In Mammalian Models for Research on Aging 224–228. Washington, DC: National Academy Press. Lathers, C. M. 1982. Lack of effect of methylprednisolone on cardiac neural discharge associated with coronary occlusionâ•‚induced arrhythmia and death. Eur J Pharmacol 85: 233–238. Lathers, C. M. 2009. Epilepsy and sudden death: Personal reflections and call for global action. Epilepsy Behav 15 (3): 269–277. Lathers, C. M. 2010a. Sudden death animal models to study nervous system sites of action for disease and pharmacological intervention. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 25. Boca Raton, FL: CRC Press. Lathers, C. M. 2010b. Arrhythmogenic, respiratory, and psychological risk factors for sudden unexpected death and epilepsy. Case histories. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Chapter 44. Boca Raton, FL: CRC Press. Lathers, C. M., and R. M. Levin. 2010. Animal model for sudden cardiac death. Sympathetic innervation and myocardial beta receptor densities. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, J. E. Leestma. Chapter 33. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy. Characterization of autonomic cardiac neural discharge associated with pentylenetetrazolâ•‚induced epileptogenic activity. Epilepsia 23: 633–648. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., and P. L. Schraeder. 1990a. Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 485–505. New York, NY: Marcel Dekker.

30 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and P. L. Schraeder. 1990b. Synchronized cardiac neural discharge and epileptogenic activity, the Lock Step Phenomenon. Lack of correlation with cardiac arrhythmias. In Epilepsy and Sudden Death, ed. C. M Lathers and P. L. Schraeder, Chapter 12, 187–198. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42: 123–136. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies and SUDEP. Epilepsy Behav 14: 573–576. Lathers, C. M., and P. L. Schraeder. 2010. Antiepileptic drugs beneἀt/risk clinical pharmacology: Possible role in cause and/or prevention of SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, J. E. Leestma, Chapter 47. Boca Raton, FL: CRC Press. Lathers, C. M., R. F. Flax, and L. J. Lipka. 1986a. The effect of C1 spinal cord transection or bilateral adrenal vein ligation on thioridazineâ•‚induced arrhythmia and death in the cat. J Clin Pharmacol 26: 515–523. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986b. Regional distribution of myocardial receptors. Eur J Pharmacol 130: 111–117. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986c. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39: 2121–2141. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin, J. C. Krusz, G. J. Kelliher, and J. Roberts. 1981. The action of reserpine, 6-hydroxydopamine, and bretylium on digitalis-induced cardiotoxicity. Eur J Pharmacol 76: 371–379. Lathers, C. M., K. F. Jim, and W. H. Spivey. 1989a. A comparison of intraosseous and intravenous routes of administration for antiseizure agents. Epilepsia 30: 472–479. Lathers, C. M., K. F. Jim, W. High, W. H. Spivey, W. D. Matthews, and T. Ho. 1989b. An investigation of the pathological and physiological effects of intraosseous sodium bicarbonate in pigs. J Clin Pharmacol 29: 354–359. Lathers, C. M., K. F. Jim, W. H. Spivey, C. Kahn, K. Dolce, and W. D. Matthews. 1990. Antiepileptic activity of beta-blocking agents. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 24, 485–505. New York, NY: Marcel Dekker. Lathers, C. M., G. J. Kelliher, and J. Roberts. 1974a. Correlation of ouabainâ•‚induced arrhythmia and nonuniformity in cardiac accelerator nerves. Clin Res 22: 682A. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalisâ•‚induced arrhythmia. Circulation 57: 1058–1065. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1977a. Role of the adrenergic nervous system in arrhythmia produced by acute coronary artery occlusion. In Pathophysiology and Therapeutics of Myocardial Ischemia, ed. A. Lefer, G. Kelliher, and M. Rovetto, 123–147. Myocardial Ischemia Symposium, Philadelphia Physiological Society, May 6, 1976. New York, NY: Spectrum Publ. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977b. Correlation of ouabainâ•‚induced arrhythmia and nonuniformity in the histamineâ•‚evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., S. A. Koehler, C. H. Wecht, and P. L. Schraeder. 2003a. Forensic antiepileptic drug levels in 2001 autopsy cases of sudden, unexpected deaths in persons with epilepsy in Allegheny County Pennsylvania. FDA Science Forum, Washington, DC. Lathers, C. M., S. A. Koehler, C. H. Wecht, and P. L. Schraeder. 2003b. Forensic antiepileptic drug levels in 2001 autopsy cases of sudden, unexpected deaths in persons with epilepsy in Allegheny County, Pennsylvania. Annual Meeting of the American College Clinical Pharmacology, Orlando, FL.

Neurocardiologic Mechanistic Risk Factors in SUDEP

31

Lathers, C. M., P. L. Schraeder, and H. G. Claycamp. 2003c. Clinical pharmacology of topirimate vs. lamotrigine vs. Phenobarbital: Comparison of efficacy and side effects using odds ratios. J Clin Pharmacol 43: 491–503. Lathers, C. M., and L. J. Lipka. 1986. Chlorpromazine: Cardiac arrhythmogenicity in the cat. Life Sci 38: 521–538. Lathers, C. M., and L. J. Lipka. 1987. Cardiac arrhythmia, sudden death, and psychoactive agents. J€Clin Pharmacol 27: 1–14. Lathers, C. M., L. J. Lipka, and H. A. Klions. 1988. Digitalis glycosides: A discussion of the similarities and differences in actions and existing controversies. Rev Clin Basic Pharmacol 7: 1–108. Lathers, C. M., and J. Roberts. 1980. Digitalis cardiotoxicity revisited. Life Sci 27: 1713–1733. Lathers, C. M., and J. Roberts. 1985. Are the sympathetic neural effects of digoxin and quinidine involved in their action on cardiac rhythm? J Cardiovasc Pharmacol 7: 350–360. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1974b. Relationship between the effect of ouabain on€ arrhythmia and interspike intervals (ISI) of cardiac accelerator nerves. Pharmacologist 16:€201. Lathers, C. M., P. L. Schraeder, and M. Bungo. 2008a. Mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Reply. Epilepsy Behav 13: 265–269. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008c. Sudden death: Neurocardiologic mystery, Chapter 13. In: Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Chapter 13. New York, NY: Nova Science Publishers, Inc. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010a. Sodium channel dysfunction. Common physiopathologic mechanism associated with sudden death ECG abnormalities in Brugada syndrome and some types of epilepsy. Case histories. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, J. E. Leestma, Chapter 20, Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010b. Syncope, seizures and SUDEP. Case histories. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 22. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010c. Unanswered questions: SUDEP Clinical studies needed. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P.€L.€Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 4. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34: 1919–1936. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital on autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33: 837–844. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lock step phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28: 736–745. Lathers, C. M., W. H. Spivey, R. M. Levin, and N. Tumer. 1990. The effect of dilevalol on cardiac autonomic neural discharge, plasma catecholamines, and myocardial beta receptor density associated with coronary occlusion. J Clin Pharmacol 30: 241–253. Lear-Kaul, K. C., L. Coughlin, and M. J. Dobersen. 2005. Sudden unexpected death in epilepsy: a retrospective study. Am J Forensic Med Pathol 26: 11–17. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26: 195–203. Lehmann-Horn F., and K. Jurkat-Rott. 1999. Voltage-gated ion channels and hereditary disease. Physiol Rev 79: 1317–1372.

32 Sudden Death in Epilepsy: Forensic and Clinical Issues Leiter, J. C., and I. Böhm. 2007. Mechanisms of pathogenesis in the sudden infant death syndrome. Respir Physiol Neurobiol 159(2): 127–138. Lipka, L. J., and C. M. Lathers. 1987. Psychoactive agents, seizure production, and sudden death in epilepsy. J Clin Pharmacol 27: 169–183. Lipka, L. J., C. M. Lathers, and J. Roberts. 1988. Does chlorpromazine produce cardiac arrhythmia via the central nervous system? J Clin Pharmacol 28: 968–983. Malow, B. A., K. Levy, K. Maturen, and R. Bowes. 2000. Obstructive sleep apnea is common in medically refractory epilepsy patients. Neurology 55: 1002–1007. Martin, P., A. Schreiner, K. Rettig, and B. Schauble. 2009. Topiramate in patients with epilepsy and intellectual disability. Epilepsy Behav 14: 496–502. Matschke, J., and R. Laas. 2007. Central alveolar hypoventilation syndrome causes SUD. Am J Forensic Med Pathol 28: 141–144. Michelsen, J. W., and J. M. Meyer. 2007. Cardiovascular effects of antipsychotics. Expert Rev Neurother 7: 829–839. McGregor, A., and J. Wheless. 2006. Pediatric experience with sudden unexplained death in epilepsy at a tertiary epilepsy center. J Child Neurol 21: 782–787. Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16: 1–7. Nagele, H., S. Hashagen, M. Azizi, S. Behrens, and M. A. Castel. 2007. Analysis of terminal arrhythmias stored in the memory of pacemakers from patients dying suddenly. Europace 9: 380–384. Nashef, L., S. Garner, J. W. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64: 349–352. Nashef, L., N. Hindocha, and A. Makoff. 2007. Risk factors in sudden death in epilepsy (SUDEP): The quest for mechanisms. Epilepsia 48: 859–871. Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42: 667–673. Novak, V., A. L. Reeves, P. Novak, P. A. Low, and F. W. Sharbrough. 1999. Time-frequency mapping of R–R interval during complex partial seizures of temporal lobe origin. J Auton Nerv Syst 24: 195–202. O’Rourke, K., and C. Lathers. 1990. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 13, 239– 260. New York, NY: Marcel Dekker. Owada, M., Y. Aizawa, K. Kurihara, N. Tanabe, T. Aizaki, and T. Izumi. 1999. Risk factors and triggers of sudden death in the working generation: An autopsy proven case-control study. Tohoku J Exp Med 189: 245–258. Paterson, D. S. 2010. Medullary serotonergic abnormalities in sudden infant death syndrome. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 5. Boca Raton, FL: CRC Press. Penἀeld, W., and H. Jasper. 1954. Summary of clinical analysis and seizure patterns. In Epilepsy and the Functional Anatomy of Human Brain, 830–831. Boston, MA: Little, Brown. Persson, H., E. Kumlien, M. Ericson, and T. Topmson. 2007. Circadian variation in heart-rate variability in localization-related epilepsy. Epilepsia 48: 917–922. Pezzella, M., P. Striano, C. Ciampa, L Errichiello, P. Penza, and S. Striano. 2009. Severe pulmonary congestion in a near-miss at the ἀrst seizure: further evidence for respiratory dysfunction in sudden unexpected death in epilepsy. Epilepsy Behav (in press). Pickworth, W. B., J. Gerard-Ciminera, and C. M. Lathers. 1990. Stress, arrhythmias and seizures. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 22, 417–446. New York, NY: Marcel Dekker.

Neurocardiologic Mechanistic Risk Factors in SUDEP

33

Quint, S. R., J. A. Messenheimer, and M. B. Tennison. 1990. Power spectral analysis: a procedure for assessing autonomic activity related to risk factors for sudden and unexplained death in epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 16, 261–291. New York, NY: Marcel Dekker. Randall, W. C. 1977. Neural Regulation of the Heart, 1–440. New York, NY: Oxford University Press. Randall, W. C. 1984. Nervous Control of Cardiovascular Function, 1–476. New York, NY: Oxford University Press. Randall, W. C., J. X. Thomas, D. E. Euler, and G. J. Rozanski. 1978. Cardiac dysrhythmias associated with autonomic nervous system imbalance in the conscious dog. In Perspectives in Cardiovascular Research, Vol. 2, Neural Mechanisms in Cardiac Arrhythmias, ed. P. J. Schwartz, A. M. Brown, A. Malliani, and A. Zanchetti, 123–138. New York, NY: Raven Press. Reingardiene, D., and J. Vilcinskaite. 2007. QTc-prolonging drugs and the risk of sudden death. Medicina (Kaunas) 43: 347–353. Rusli, M., W. H. Spivey, H. Bonner, R. M. McNamara, C. K. Aaron, and C. M. Lathers. 1987. Pathological effects of endotracheal diazepam in cats. Ann Emerg Med 16: 314–318. Ryvlin, P., T. Tomson, and A. Montavont. 2009. Excess mortality and sudden unexpected death in epilepsy. Press Med 38 (6): 905–910. Saffitz, J. E. 2008. Sympathetic neural activity and the pathogenesis of sudden cardiac death. Heart Rhythm 5: 140–141. Salama, G., and B. London. 2007. Mouse models of long QT syndrome. J Physiol 578 (Pt 1): 43–53. Sarkozy, A., and P. Brugada. 2005. Sudden cardiac death: What is inside our genes? Can J Cardiol 21: 1099–1110. Schott, G. D., A. A. McLeod, and D. E. Jewitt. 1977. Cardiac arrhythmias that masquerade as epilepsy. Br Med J 1: 1454–1457. Schraeder, P. L. 1987. Adult respiratory distress syndrome (ARDS) associated with nonconvulsive status epilepticus. Epilepsia 28: 605. Schraeder, P. L. 2010. SUDEP case histories: Typical and atypical. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 52. Boca Raton, FL: CRC Press. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3: 55–62. Schraeder, P. L., C. M. Lathers, and J. B. Charles. 1994. The spectrum of syncope. J Clin Pharmacol 34: 454–459. Schraeder, P. L., R. Pontzer, and T. R. Engel. 1983. A case of being scared to death. Arch Intern Med 143: 1793–1794. Schuele, S. U. 2009. Effects of seizures on cardiac function. J Clin Neurophysiol. Sep 11. [Epub ahead of print]. Schwartz, R. D., X. Yu, M. R. Katzman, D. M. Hayden-Hixson, and J. M. Perry. 1995. Diazepam, given postischemia, protects selectively vulnerable neurons in the rat hippocampus and striatum. J Neurosci 15 (1 Pt 2): 529–539. Scorza, F. A., A. M. Abreu, M. Albuquerque et al. 2007. Quantiἀcation of respiratory parameters in patients with temporal lobe epilepsy. Arq Neuropsiquiatr 65: 450–453. Scorza, F. A., M. Albuquerque, R. M. Arida, V. C. Terra, H. R. Machado, and E. A. Cavalheiro. 2010a. Beneἀts of sunlight: Vitamin D deἀciency might increase the risk of sudden unexpected death in epilepsy. Med Hypotheses 74: 158–161. Scorza, F. A., E. A. Cavalheiro, R. M. Arida et al. 2010b. Omega-3 fatty acids in SUDEP: Guardian of the brain-heart connection. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Chapter 3. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 22. Boca Raton, FL: CRC Press.

34 Sudden Death in Epilepsy: Forensic and Clinical Issues Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008a. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13: 263–264. Scorza, F. A., R. M. Cysneiros, and R. M. Arida et al. 2008b. Adult hippocampal neurogenesis and sudden unexpected death in epilepsy: Reality or just an attractive history? Med Hypotheses 71: 914–922. Sharma, S., T. Ho, and B. K. Kantharia. 2010. Not seizure but syncope: Causes and investigations to establish the diagnosis. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, J. E. Leestma, Chapter 21. Boca Raton, FL: CRC Press. Simon, R. P., L. L. Bayne, R. F. Tranbaugh, and F. R. Lewis. 1982. Elevated pulmonary flow and protein content during status epilepticus in sheep. J Appl Physiol 52: 91–95. Sheldon, R. S., A. G. Sheldon, A. Serletis et al. 2007. Investigators of the prevention of syncope study. Worsening of symptoms before presentation with vasovagal syncope. J Cardiovasc Electrophysiol 18: 954–959. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (Suppl 9): 93–98. So, E. L., J. Brainbridge, and J. R. Buchhalter et al. 2008. Report of the American Epilepsy Society and the Epilepsy Foundation Joint Task Force on sudden unexplained death in epilepsy. Epilepsia 50: 917–922. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41: 1494–1497. Sonoda E. Y., D. B. Colugnati, C. A. Scorza, R. M. Arida, A. P. Pansani, A. C. de Almeida, E. A. Cavalheiro, and F. A. Scorza. 2008. Is cold the new hot in sudden unexpected death in epilepsy? Effect of low temperature on heart rate of rats with epilepsy. Arq Neuropsiquiatr 66: 848–852. Spivey, W. H., and C. M. Lathers. 1985. Effect of timolol on the sympathetic nervous system in coronary occlusion in cats. Ann Emerg Med 134: 939–944. Spivey, W. H., C. M. Lathers et al. 1985. A comparison of intraosseous, central and peripheral routes of sodium bicarbonate administration during CPR in pigs. Ann Emerg Med 14: 1135–1139. Spivey, W. H., H. D. Unger, R. M. McNamara, M. M. LaManna, T. Ho, and C. M. Lathers. 1987a. The effect of intraosseous sodium bicarbonate on bone in swine. Ann Emerg Med 16: 773–776. Spivey, W. H., H. D. Unger, C. M. Lathers, and R. M. McNamara. 1987b. Intraosseous diazepam suppression of pentylentetetrazol-induced epileptogenic activity in pigs. Ann Emerg Med 16: 56–59. Stauffer, A. Z., J. Doddâ•‚O, and C. M. Lathers. 1989. The relationship of the lock step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72: 340–345. Stauffer, A. Z., J. M. Dodd-O, and C. M. Lathers. 1990. Relationship of the lockstep phenomenon and precipitous changes in blood pressure. In Epilepsy and Sudden Death, ed. C. M. Lathers and P.€L. Schraeder, Chapter 14, 199–220. New York, NY: Marcel Dekker. Strzelczyk, A., S. Bauer, S. Knake, W. H. Oertel, H. M. Hamer, and F. Rosenow. 2008. Ictal asystole in temporal lobe epilepsy before and after pacemaker implantation. Epileptic Disord 10: 39–44. Stopper, M., T. Joska, and M. M. Burg et al. 2006. Electrophysiologic characteristics of anger-Â�triggered arrhythmias. Heart Rhythm 4: 268–273. Tavee, J., and H. Morris 3rd. 2008. Severe postictal laryngospasm as a potential mechanism for sudden unexpected death in epilepsy: a near-miss in an EMU. Epilepsia 49: 2113–2117. Terra-Bustamante, V. C., C. A. Scorza, and M. de Albuquerque et al. 2009. Does the lunar phase have an effect on sudden unexpected death in epilepsy? Epilepsy Behav 14: 404–406. Terrence, C. E., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death in epileptic patients. Ann Neurol 25: 594–595. Tomson, T. 2010. Compliance with antiepileptic drug treatment and the risk of SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapters 51 and 22. Boca Raton, FL: CRC Press.

Neurocardiologic Mechanistic Risk Factors in SUDEP

35

Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: current knowledge and future directions. Lancet Neurol 7: 1021–1031. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47: 21–26. Tupal, S., and C. L. Faingold. 2006. Respiratory arrest mechanisms, modulated in part by serotonin, may cause SUDEP. Epilepsia 47: 21–26. Tumer, N., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital upon autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. Epilepsia 26: 520. Verrier, R. L., K. Kumar, and B. D. Nearing. 2009. Basis for sudden cardiac death prediction by T-wave alternans from an integrative physiology perspective. Heart Rhythm 6 (3): 416–422. Verrier, R. L., and S. C. Schachter. 2010. Neurocardiac interactions in sudden unexpected death in epilepsy: Can ambulatory electrocardiogram-based assessment of autonomic function and T-wave alternans help to evaluate risk? In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo and J. E. Leestma, Chapters 43 and 22. Boca Raton, FL: CRC Press. Vindrola, O., M. Asai, M. Zubieta, E. Talavera, E. Rodriguez, and G. Linares. 1984. Pentylenetetrazol kindling produces a long-lasting elevation of IR-Met-enkephalin but not IR-Leu-enkephalin in rat brain. Brain Res 297: 121–125. Vohra, J. 2007. The Long QT Syndrome. Heart Lung Circ 16 (Suppl 3): S5–S12. Walczak, T. 2003. Do antiepileptic drugs play a role in sudden unexpected death in epilepsy? Drug Saf 26: 673–683. Walczak, T. 2010. Risk factors for SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 12. Boca Raton, FL: CRC Press. Wang, J., Y. Chen, K. Li, and L. Hou. 2006. Blockade of inhibitory neurotransmission evoked seizurelike ἀring of cardiac parasympathetic neurons in brainstem slices of newborn rats: Implications for sudden deaths in patients of epilepsy. Epilepsy Res 70: 172–183. Wannamaker, B. 1985. Autonomic nervous system and epilepsy. Epilepsia 26 (Suppl 1): S31–S39. Wichter, T., T. M. Paul, and L. Eckardt et al. 2005. Arrhythmogenic right ventricular cardiomyopathy. Antiarrhythmic drugs, catheter ablation, or ICD? Herz 30: 91–101. Williams, L., and M. Frenneaux. 2007. Syncope in hypertropic cardiomyopathy mechanism and consequences for treatment. Europace 9 (9): 817–822. Winfree, A. T. 1987. When Time Breaks Down: The Three-Dimensional Dynamics of Electrochemical Waves and Cardiac Arrhythmias. Princeton, NJ: Princeton University Press.

Forensic Considerations and Sudden Unexpected Death in Epilepsy

2

Jan E. Leestma

Contents 2.1 Background of Sudden or Unexpected Deaths 2.2 Forensic Issues Regarding Epilepsy and SUDEP Deaths 2.2.1 Classifying Deaths as SUDEP 2.3 Drowning 2.4 SUDEP Deaths in the Home or Workplace 2.5 SUDEP Deaths Outdoors 2.6 Role of SUDEP in Traffic Deaths 2.7 SUDEP and Deaths in Agitated Delirium or Restraint 2.8 SUDEP in Criminal Cases 2.9 SUDEP and Anticonvulsant Medication References

37 39 39 41 41 42 42 44 48 50 52

2.1â•… Background of Sudden or Unexpected Deaths Sudden and unexpected deaths generally occur outside hospitals, although victims may be brought to an emergency room. In most locales in the United States, and likely in other developed countries, such deaths are usually not attended by a physician and are without detailed historical or medical information. They will usually be brought to the attention of a medical examiner or coroner, who is responsible for determining the cause and manner of death and for generating a death certiἀcate before the remains may be interred or otherwise disposed of. These responsibilities are statutory in most jurisdictions, but how they are discharged varies widely (DiMaio and DiMaio 2001). When a coroner is the responsible party, this individual may be elected or appointed, and may or may not be a medical doctor. A local pathologist (forensically trained or not) will usually be employed to conduct an autopsy or other examination of the body to make the determination of cause and manner of death. When a medical examiner is involved, that individual is almost always a trained forensic pathologist or will employ a staff of forensic pathologists to perform the required examinations to determine the medical cause of the death and its manner, which by convention may be labeled as a homicide (death at the hands of another), suicide (death by one’s own hand), accident, natural (natural disease processes), or “undetermined.” The coroner or medical examiner may direct that an autopsy be performed and that other studies that may include toxicological examinations of tissues or body fluids be undertaken (DiMaio and DiMaio 2001; Knight 1996). 37

38 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 2.1â•… Manner of Death Statistics Manner of Death

Percentage of Death Certiἀcates (%)

Natural diseases Homicide Suicide Accidents Undetermined

64–66 10–11 4–5 20–22 4–5

Sources: Office of the Medical Examiner, Cook County, IL, Annual Report, 1977– 1979, Office of the Medical Examiner, Chicago, IL, 1979; Leestma, J. E., and E. W. Sharp, in Forensic Neuropathology, CRC Press, Boca Raton, FL, 2009. With permission.

So-called sudden deaths may or may not be truly sudden, with many deἀnitions that surround such deaths (Davis and Wright 1980; de la Grandmaison and Durigon 2002; Kuller 1966; Kuller et al. 1966; Luke and Helpern 1968; Moritz 1954). When one refers to sudden deaths, more often than not, one is really referring to the fact that the victim actually died unexpectedly and, perhaps, also suddenly. When such deaths are witnessed, death may be described as having occurred in minutes rather than hours or longer. When a death is not witnessed, but the victim is found in circumstances that indicate death occurred in the course of normal activities and does not suggest a protracted agonal period, the unexpected and sudden quality of the death can be presumed (Haerem 1978). Such circumstances often involve ἀnding the victim dead in bed, in the bathroom (not bathing), or in a den or living room (in a chair or on the floor). When the victim is found in the bath (submerged or not), swimming pool, sauna or Jacuzzi, near electrical equipment or machinery, or in a vehicle (not involved in a crash), interpretations regarding the exitus may become complicated. When death occurs in the context of an apparent accident, with or without trauma, many facts and factors will have to be considered in reaching a proper determination of cause and manner of death (Leestma 1990a) (see Table 2.1). Owing to the variability in coroners’ and medical examiners’ statistics, it is generally recognized that 10–15% of their cases occurred suddenly and/or unexpectedly (see Table 2.2). Not unexpectedly, the majority of these deaths are due to some form of heart attack. Often this cannot be proven anatomically, although it can be inferred (Haerem 1978; Fineschi et al. 2006; Greenberg and Dwyer 1982; Schwartz and Gerrity 1975). Most central nervous system diseases do not kill suddenly, even though they present suddenly

Table 2.2â•… Medical Causes of Death in Sudden and/or Unexpected Deaths Disease Process Heart and great vessels Respiratory system Brain or meninges Digestive or urogenital systems Miscellaneous

Percentage (%) 56.1 (±7.4%) 14.5 (±6.4%) 15.8 (±2.4%) 8 (±1.7%) 9.5 (±8.9%)

Sources: Kuller, L., J Chronic Dis, 19 (11), 1165–1192, 1966; Kuller, L., A. Lilienfeld, and R. Fisher, Circulation, 34 (6), 1056–1068, 1966; Leestma, J. E., in Forensic NeuroÂ� pathology, CRC Press, Boca Raton, FL, 2009. With permission.

Forensic Considerations and Sudden Unexpected Death in Epilepsy

39

and unexpectedly as in strokes (Leestma 2009a). The largest group of CNS-related sudden deaths appear to occur in the context of epilepsy [sudden death in epilepsy (SUDEP)] (Leestma 2009a, 1990b; Lhatoo and Sander 2002). A number of chapters in this volume deal with population statistics for SUDEP and these data will not be repeated here. Suffice it to say that deaths classiἀed as SUDEP—not a cause of death, but a category of death for which no anatomically evident cause is found, much like the label sudden infant death syndrome (SIDS)—forms an important and nonrare problem for forensic pathologists. As noted before in previous publications (Leestma 1990a, 1990b; Leestma et al. 1989, 1997), the apparent incidence of SUDEP cases at a large metropolitan medical examiner’s facility (Chicago-Cook County, IL) is 60–80 cases per year (Leestma et al. 1989). A casual survey of other large metropolitan offices by the author indicates similar numbers of cases in all of them (Los Angeles, Denver, and Miami). Collecting precise and accurate statistics on SUDEP incidence is problematic because there is wide variation in practices among the forensic community on how such cases are labeled on death certiἀcates as Schraeder et al. (2006) have reported. This is a function of the awareness or lack of awareness of the phenomenon and/or philosophical positions on how, presumably, epilepsy-related deaths should be classiἀed or labeled by forensic pathologists. As alluded to earlier, it is the task and responsibility of the forensic pathologist to determine the medical cause and the manner of death if possible. If this is not possible, then, after all the evidence is gathered, the medical cause of death and, sometimes, the manner of death, must remain undeclared (undetermined). This circumstance is certainly intellectually disappointing and frustrating and may lead to unwarranted speculations on the part of family, friends, attorneys, the media, and colleagues. Such cases, whether assigned the SUDEP designation or not, may ἀnd their way into the legal system as civil cases (torts) or even as criminal cases. The following discussion will explore some of these possibilities and, when possible, provide case examples of the issues involved.

2.2â•…Forensic Issues Regarding Epilepsy and SUDEP Deaths 2.2.1â•…Classifying Deaths as SUDEP Which cases are labeled with the designation SUDEP can be difficult and sometimes controversial. An attempt to provide some guidelines was arrived at in conjunction with data collected during the development of lamotrigine (Lamictal) at the behest of the then Burroughs-Wellcome pharmaceutical company, now Glaxo-SmithKline, Inc. In the clinical trial databases that drew upon studies in many countries with 4700 patients (5747 patient years of exposure) over several years, 45 sudden and/or unexpected deaths were discovered in all arms of the study. It was important to evaluate, if possible, if there was a different incidence of SUDEP deaths between the treated and control groups. To this end, a study group of professionals that included epidemiologists, epileptologists, and pathologists was impaneled to review the data and to determine what criteria might be applied to the cases in order to stratify them for analysis (Leestma et al. 1997; Tennis et al. 1995). As might be expected, the quality and quantity of individual case data varied widely. It was not possible to obtain further data beyond that reported. Cases were stratiἀed according to the following classiἀcation (Leestma et al. 1997):

40 Sudden Death in Epilepsy: Forensic and Clinical Issues

• SUDEP (definite or highly probable) The victim suffered from epilepsy as deἀned by Gastaut (1969) and Gastaut and Zifkin (1985) and had been treated with one or more anticonvulsive agents, usually for many years, and died unexpectedly in a reasonable state of health. The fatal attack occurred suddenly, but death might not have occurred for several hours, when it was usually associated with cardiorespiratory arrest, resuscitative efforts, and their complications. The attack occurred during normal activities in benign circumstances. An obvious medical cause of death was not found after autopsy. Cardiac arrhythmia may have been observed after the attack. If the victim was found in the bath but did not show evidence of drowning, the death may be assignable as SUDEP. If status epilepticus or acute neurotrauma occurred during seizure, the case was excluded. • Possible SUDEP These cases met most or all of the SUDEP criteria, but data suggested more than one possible cause of death associated with seizures, such as death while bathing, swimming, or due to aspiration, with or without an observed seizure. • Other non-SUDEP These cases could not be assigned to either of the ἀrst two categories and an obvious or likely cause of death had been established. • Insufficient data These cases could not be interpreted because of lack of information or ambiguous data concerning the circumstances of death or concurrent medical conditions in the victim. There are and were inevitable issues with the above classiἀcations, owing to all the informational problems that are inherent to any patient-based study. A particularly thorny one was the role of possible drowning in some of the victims, which is discussed later. Other confounding variables include the role of concurrent diseases in the victim, especially heart disease in the older victim, as well as conditions related to alcohol and drug abuse. There are a host of issues in infants and children that include possible SIDS-type deaths. Highly important to any death investigation of an individual who apparently dies suddenly and/or unexpectedly is the death scene investigation (DiMaio and DiMaio 2001; Knight 1996; Leestma 2009b). The physical death scene environment must be documented and carefully inspected. It is important to note if the body is in a position or location that would be consistent with normal activity that was interrupted by the fatal attack. The presence of bruises on the face, head, and extremities does not necessarily mean that the victim was assaulted; rather that if a seizure occurred, the victim may have fallen and been injured or become injured during the clonic phase of a seizure. Context and common sense in evaluating the scene is vital. Furthermore, the investigator should search for and collect any pill bottles that are present and any materials that suggest illicit drug use. The pill bottles may indicate the use of anticonvulsant medications and can provide the names of the pharmacies that dispensed the drug and the physician who prescribed it, who may be contacted later for additional information. If the pill bottles have pills in them and the date of prescription is on the label, or if the bottle is empty, one can make a judgment regarding medication compliance by the victim. If other medications are present, this can also give a

Forensic Considerations and Sudden Unexpected Death in Epilepsy

41

clue as to other medical problems the victim may have had and can provide some guidance as to which toxicological examinations may need to be made. If any witnesses to the attack are present, a detailed description of what was seen is vital. If the attack was not witnessed, it may be possible to determine the timeframe of the attack by inference (when the individual was last seen and found). If friends or relatives are available, they may have knowledge about the health status of the decedent and, speciἀcally, if the victim was epileptic. To aid investigators in collecting this vital information, the author has previously described a “check sheet” that can serve as a template for investigators at death scenes where the victim was epileptic and sudden or unexpected death has occurred (Leestma et al. 1989; Leestma 1990b, 2009b).

2.3â•…Drowning Epileptic individuals are frequently found dead in bathtubs, swimming pools, hot tubs, and natural bodies of water in conjunction with swimming. It is a common instruction to epileptic persons not to bathe upon arising or immediately before going to bed, rather to bathe at another time of the day to avoid the increased likelihood of seizures that occur proximate to sleep (Laidlaw and Richens 1982; Engel and Pedley 2008). The postmortem diagnosis of drowning is imprecise and subject to many evidentiary problems that are well-known to any forensic pathologist such as the so-called “wet” and “dry” forms of drowning (DiMaio and DiMaio 2001; Knight 1996); thus, it cannot be known with complete assurance that a victim who was found submerged in the bathtub or swimming pool, and whose lungs did not show evidence of water aspiration, did not drown (laryngospasm), but rather died from the supposed mechanism(s) typical of SUDEP and simply sank into the water after the attack, when the victim was lifeless. If the potential drowning occurred in a natural body of water, such as a stream, lake, or ocean, it may be possible to recover diatoms and other small particles from the water in the lower respiratory passages, which may indicate aspiration. The forensics of this process and its reliability is controversial (Piette and De Letter 2006; Modell et al. 1999; Pachar and Cameron 1993). If the victim is found in the bathtub but the head is not immersed, this does not necessarily mean that drowning did not occur. The onset of rigor mortis may have shifted the position of the body after death, causing the head to rise out of the bath water. A careful examination of the oral cavity at autopsy may reveals bites of the tongue, lips, or buccal mucosa, which is presumptive evidence that a recent seizure occurred but still does not necessarily mean that drowning did not occur (Ulrich and Maxeiner 2003).

2.4â•…SUDEP Deaths in the Home or Workplace The most common location for SUDEP deaths is in the home, in locations that bespeak an attack that took place during normal activity (Leestma et al. 1989; Leestma 1990b). The most common of these is in the bedroom. The victim may be found in bed in a normal position, or one that suggests some movement, as in a seizure with disordered bedding apparent. The victim may also be found on the floor in the bedroom dressed in normal

42 Sudden Death in Epilepsy: Forensic and Clinical Issues

clothing or in bed clothes beside the bed or in another position that suggests that the fatal attack took place proximate to going to bed. The association of SUDEP, and for that matter epileptic seizures in general (Laidlaw and Richens 1982; Engel and Pedley 2008), with sleep is well known and it is also not uncommon to ἀnd SUDEP victims dead in an easy chair in a den or living room, sometimes with the television on, which suggests that the victim may have been sleeping or dozing when the attack occurred. Death in the bathroom, discussed earlier with respect to drowning, may have no connection with bathing when the victim is found dead on the bathroom floor. In such circumstances, various injuries from the impact of the body surfaces with bathroom ἀxtures is common and may confuse the interpretation of the death scene, suggesting a homicidal attack or perhaps a suicide. A careful documentation of the death scene and blood spatter or blood flow (DiMaio and DiMaio 2001; James et al. 2005) as well as the autopsy may clear up suspicion of foul play. The body of a SUDEP victim may be found in other areas of the home, in the garage or workshop area, or in a workplace area that raises the issue of other means of death than a seizure-related cause. A careful scene inspection and investigation will go a long way in ruling in or out electrocution or intoxication, be they accidental or self-inflicted. Deaths may not be discovered immediately in SUDEP victims and the body may have begun to decompose, making interpretations complex and making a proper scene investigation even more germane. The process of decomposition has been the subject of careful study by forensic scientists and there is a large body of literature on the subject (Milroy 1999; Knight 1996). Occasionally, a body that is not immediately found in a home may be subject to predation by insects, rodents, or pets. Most experienced forensic pathologists are familiar with these phenomena and can differentiate marks upon the body from premortem injuries.

2.5╅SUDEP Deaths Outdoors While it is not common, SUDEP victims can experience a fatal attack outdoors with prompt or delayed discovery of the body. Obviously, the longer a body has lain outside, the more decomposition and environmental effects such as insects, vermin, and other animals will complicate examination of determinations on the body. Even in decomposed bodies, it may be possible to glean valuable information that may make a diagnosis of SUDEP practicable (Milroy 1999; Knight 1996). As with an indoor death scene investigation, the outdoor death scene can yield important information about what happened to the victim and rule in or€out foul play, suicide, or accident. The following case example (Case A) is illustrative.

2.6â•…Role of SUDEP in Traffic Deaths A not uncommon type of forensic case is that involving vehicular crashes in which the driver appears to have experienced some sort of attack that rendered that person unable to control the vehicle. These may be single-vehicle or multivehicle incidents. Such cases are often very difficult to evaluate from a pathological point of view, owing to the often overlying trauma to the body and brain caused by the accident, and also the coexistence of other disease processes, the most important and common being cardiovascular disease.

Forensic Considerations and Sudden Unexpected Death in Epilepsy

43

Case A A middle-aged man had been complaining of headaches and “strange” feelings and set out toward his automobile to go to his family physician for an examination. The man’s wife observed the man fall to the ground and apparently jerk beside his car. When she reached her husband, he was apparently dead. Autopsy revealed a 2.5-cm unruptured, encapsulated mass ἀlled with brownish greasy material indenting the right inferior frontal lobe anterior to the amygdala. Microscopic examination revealed the mass to be an epidermoid cyst. The remainder of the autopsy was unremarkable. It appeared that this man died from an attack that represented a generalized convulsion (GTC), apparently his ἀrst and last, from a lesion in a commonly epileptogenic region of the brain. From the standpoint of SUDEP criteria, this man was not known to be epileptic, although it appears that at the time of his death he experienced a GTC event and, thus, it could be argued the death was SUDEP.

A not uncommon cause of vehicular accidents is an epileptic attack that causes the driver-victim to lose control of the vehicle and incur injuries that may or may not prove to be fatal. In the case of fatal accidents, an examination of the vehicle, its contents, and condition may provide important information for the forensic pathologist. As is always the case, witness accounts and a medical history of the victim may point the way for further inquiry, especially if the victim was suffering from epilepsy and did not appear to have died from traumatic injuries. As in other cases of SUDEP, an effort should be made to determine what drugs the victim was taking, who prescribed them, and if the victim was compliant. A thorough autopsy examination will generally reveal if there are signiἀcant traumatic injuries and a similar thorough neuropathological examination will reveal any acute lesions, or chronic lesions in the brain that might have been epileptogenic. Examples of such lesions are old traumatic contusions, vascular anomalies, tumors, and congenital malformations (Leestma et al. 1989; Leestma 2009b). Autopsies of non-SUDEP deaths in epileptic persons generally show a relatively low incidence of structural lesions in the brain (about 10%), whereas SUDEP victims have a much higher incidence, between 50% and 70% (Leestma et al. 1985, 1989; Monte et al. 2007; Shields et al. 2002; Thom et al. 1999). Not all studies have shown signiἀcant neuropathology, however (Morentin and Alcaraz 2002). If there are acute traumatic lesions in the central nervous system, it may still be possible to determine if there are more chronic lesions present. It may, however, be unlikely to determine if the older lesions had anything to do with the vehicle crash, although sometimes collateral information may suggest a causal event. Examination of the oropharynx may reveal tongue, lip, or buccal bites that may indicate a recent seizure (Ulrich and Maxeiner 2003). Toxicological examination will reveal if anticonvulsants were present and if they were within therapeutic range. While many studies of SUDEP victims have shown that they have been or are taking multiple anticonvulsant medications, they are more likely than not to be noncompliant with respect to anticonvulsant medications (Bell and Sander 2006; George and Davis 1998; Langan 2000; Langan et al. 2005; Lear-Kaul et al. 2005; Lund and Gormsen 1985; McKee and Bodἀsh 2000; Monte et al. 2007; Nilsson et al. 1999; Tomson et al. 2005; Vlooswijk et al. 2007; Walczak 2003).

44 Sudden Death in Epilepsy: Forensic and Clinical Issues

Not all studies have shown this (Opeskin et al. 1999, 2000; Schwender and Troncoso 1986; Walczak et al. 2001), and some have suggested that toxicological studies may not truly reflect proper or improper drug levels (McGugan 1999), thus the role of medications on the risk of SUDEP is, at present, suggestive but has not been proven. Some have suggested that carbamazepine may play some role in SUDEP, either by some idiosyncratic reaction or in connection with changes in dosage of this drug (Hitiris et al. 2007; Nilsson et al. 1999, 2001; Timmings 1993; Walczak 2003). A thorough pathological examination of the heart must be performed to deἀne the extent of arteriosclerotic cardiovascular disease, previous areas of myocardial scarring or necrosis, and any other morphological abnormality of the heart or its valves. One inevitably encounters some degree of cardiovascular disease in accident victims and in sudden or unexpected deaths, and it becomes difficult to compare the importance of what is found to the death circumstances (Natelson et al. 1998; Scorza et al. 2007). There is a robust literature on the heart and sudden death (Okada and Kawai 1983; Rossi 1982; Schwartz and Gerrity 1975; Schwartz and Walsh 1971; Fineschi et al. 2006; Greenberg and Dwyer 1982; Bharati and Lev 1990) and on the so-called neurocardiology (Armour and Ardell 1984; Johnson et al. 1984), which cannot be reviewed here but the issue of cardiac dysfunction and pathology is discussed in detail elsewhere in this volume. Since SUDEP deaths are apparently physiological in the sense that the process that leads to SUDEP is electrophysiological and, thus, may or may not have a morphological counterpart such as a conduction system defect in the heart, an ion channelopathy, or some other deἀnable and demonstrable disease process, the science of forensic pathology can only go so far in determining the mechanism of death in some traffic-related deaths.

2.7â•…SUDEP and Deaths in Agitated Delirium or Restraint There is a category of sudden death in which the victim dies suddenly and unexpectedly while in a state of delirium that may involve an arrest or restraint circumstance, during which the victim dies. This problem has been the subject of a robust literature over many years and recent monographs by DiMaio and DiMaio (2006) and by Ross and Chan (2006) review much of it. It is germane to this discussion that some victims are epileptic. A typical restraint death circumstance is one where an individual is involved in an assault, robbery, disturbance, or some other event that may result in law enforcement personnel, emergency personnel, or rescue personnel being summoned to the scene. Once on the scene, there may be an apparent offender who is violent, disoriented, or delusional and may seem to need to be restrained or taken into custody to prevent flight, injury to the offender or to others. In the course of the activity of restraining the individual, many personnel may become involved in trying to subdue the violent person and in so doing may place handcuffs or restraint ties to the person and may place the individual in the prone position, “hog tie” them, and perhaps overlie the individual with several people. In this circumstance, the violent person may suddenly become quiet or complain of not being able to breathe and suffer a cardiorespiratory arrest from which the victim may or may not be able to be resuscitated. Such events almost inevitably result in some form of litigation in which the personnel involved in the restraint are held to account for what actions they may or may not have

Forensic Considerations and Sudden Unexpected Death in Epilepsy

Case B This 29-year-old man had apparently had a seizure disorder since childhood, for which he had been treated with a variety of anticonvulsant medications and a vagal nerve stimulator (about 3 years before this admission) in an effort to suppress his frequent seizures. On a recent occasion, he had been admitted to a hospital after having suffered a seizure. During admission the patient experienced a cardio/pulmonary arrest from which he was successfully resuscitated. The patient had a history of respiratory troubles including asthma. A brain CT on this admission was said to be negative. The shortness of breath the patient had complained of was thought possibly to be due to aspiration. As the patient began to emerge from his postictal state and resuscitation procedures, he was disoriented and agitated, for which he was medicated. He recovered from this episode. Seven months later in the presence of his mother, the patient experienced a seizure event and apparently, while postictal, he became delirious and combative and assaulted his mother, who called for assistance. Police and ἀre department personnel responded and subdued the man by placing him in handcuffs and hobbles in a prone position on a litter. Soon afterward, he became unconscious and pulseless. Immediately after emergency medical personnel arrived, the patient had a heart rate of 140 and a respiratory rate of 28, but then became pulseless. Resuscitative treatment was begun and the man was conveyed to an emergency facility where he was reported to be asystolic on admission. During resuscitation, the patient urinated and vomited. Resuscitation did not succeed and the patient was pronounced dead. The victim was known to be hypertensive and to have had temporal lobe epilepsy for many years. When he was postical, it was not uncommon for him to be delusional and agitated. Autopsy examination reported a body weight of 251 lb and a height of about 72 in. A number of bruises and abrasions to the face, arms, and hands were noted. There were petechial hemorrhages of the lip mucosa and right flank. A subcutaneous hemorrhage was noted near the left occiput. There was evidence of food aspiration. Heart–blood toxicological examination revealed 10 µg/ml barbiturates, <10 µg/ml Lamotrigine, 25 µg/ml Oxcarbazepine, and no Phenytoin. A vagus nerve stimulator was encountered and removed (no testing done). The internal autopsy examination reported no injuries or obstructions to the neck organs. No injury to the lungs was noted and no pneumothorax was noted. The heart weighed 450 g but no obvious atheromata were noted, although abundant epicardial fat that inἀltrated the myocardium and moderately severe myoἀber scarring and hypertrophy were found. The liver was enlarged (2840 g). Neuropathology examination revealed Purkinje cell loss and cortical atrophy of the cerebellum, mild early hypoxic/ischemic changes in neurons, no apparent abnormality of the hippocampal regions beyond extensive corpora amylacea deposition in the subpial region, and no obvious lesion to account for the epilepsy. Microscopically, the heart showed extensive myocardial ἀbrosis. The medical examiner ruled the manner of death a homicide. A succession of lawsuits ensued against the police and ἀre department personnel who had responded in the incident. It was alleged that the man died because of positional asphyxia and compressive forces on the man’s thorax during the restraint.

45

46 Sudden Death in Epilepsy: Forensic and Clinical Issues

Case C This was a 38-year-old white male who apparently suffered an epileptic seizure while at a university campus where he was attending classes. He allegedly exhibited bizarre behavior and began to bang his head and face against a wall and then experienced what was described as a generalized tonic–clonic (GTC) seizure by many observers. After the seizure he became agitated, delusional, and combative. Police were called and they attempted to restrain the man by handcuffing him in a prone position. Emergency medical personnel were called and transported him to hospital, arriving at 1909 hours. While being transported after glucagon was administered (blood sugar found to be 34), he became still and experienced cardiorespiratory arrest. He was cyanotic and pulseless, with ἀxed and dilated pupils. Cardiac monitoring revealed asystole. Resuscitative measures were instituted without success. He was pronounced dead at 1926 hours. The victim’s history was that he had been epileptic since childhood and two craniotomies had been performed on his right temporal region in an attempt to correct his epilepsy. He continued to have seizures that were described as GTC, for which he was treated with carbamazepine. In spite of medication, he would have periodic seizures with an aura, not described. He was somewhat impulsive and apparently either slightly aphasic or dyslexic, which caused him some frustration. He apparently did not use alcohol or drugs. His medical records indicated he had had diabetes insipidus, in conjunction with one of his craniotomies. A history of diabetes mellitus was unsubstantiated. Autopsy revealed evidence of medical treatment, and that he was somewhat plethoric and had a blotchy erythema of the face with several abrasions apparently due to his head-banging and also to efforts to subdue him. There was blood in the mouth and a large tongue bite was found. The autopsy pathologist did not describe petechial hemorrhages in eyes, viscera, or face and discounted the erythema of the skin as being petechial hemorrhages due to suffocation. He had a number of bruises and abrasions of the elbows, knees, etc., due to scuffling. His heart weighed 490 g and showed <30% atheroma of right coronary, left circumflex, but the left anterior descending artery showed 50% occlusion, and the aorta showed moderate atheroma with no obvious infarct. The lungs were 800 and 680 g, congested and wet. No airway obstruction nor injury to neck or hyoid was noted. The liver was 2600 g and congested, as was the spleen (300 g). The brain was 1320 g with evidence of prior surgery in the right temporal side. No subdural hemorrhage was found, and no traumatic injury noted. The case was ἀled as manner of death undetermined with seizure disorder with delirium, coronary artery disease, and positional restraint.

performed. Inevitably, their training and education regarding proper procedures for “take downs” are scrutinized, along with what sort of supervision occurred during the event (Ross and Chan 2006). Death during restraint victims quite frequently are obese, hypertensive, diabetic, or under the influence of substances such as cocaine, amphetamines, phencyclidine (“angel

Forensic Considerations and Sudden Unexpected Death in Epilepsy

47

dust” or PCP), alcohol, cannabis, barbiturates, or opiates, as well as tranquilizers, antipsychotic medication, anticonvulsants, and other drugs (DiMaio and DiMaio 2006). Often, the use of many drugs is demonstrated. Not infrequently, the victims are found to be hyperpyrexic, have cocaine in their blood, and may show myoglobinuria (Ruttenber et al. 1997, 1999; Ross and Chan 2006). Cardiovascular disease demonstrated at autopsy is common. Some victims have a history of epilepsy and may have been apparently postictal when they were in their delirious state. The following are two cases that illustrate the latter, and most of the problems associated with that. This case, like the previous one, was the subject of civil litigation against the responding personnel and their organizations. In both cases, ultimately, either settlements were reached or no judgments were returned against the defendants or their organizations. Both cases illustrate the complexities that may occur when an agitated or delirious epileptic patient needs to be restrained to prevent self-injury or injury to those nearby. It seems painfully obvious that some action on the part of responding personnel, be they police, ἀremen, emergency personnel, or bystanders may be called for in situations like those described above, and also that it seems inevitable when tragedy occurs that they will be blamed, appropriately or not (Ross and Chan 2006). The causal issue of prone restraint, so-called “hog tying,” and overlaying of the agitated and delirious victim is controversial and is still being debated in scientiἀc and forensic circles and in the courts. Other tactics used in restraint of violent persons may involve one or more variants of the chokehold, which may occlude the airway and/or compress the carotid artery and produce death by these means (Vilke 2006, Reay and Holloway 1982; Reay and Eisele 1982). With respect to “hog tying” and prone restraint, Reay et al. (1988, 1992) and O’Halloran et al. (O’Halloran and Lewman 1993; O’Halloran and Frank 2000; O’Halloran 2004) postulated that when in the prone position “hog tied” and when personnel overlay the agitated individual, respiratory function may become impaired and cause death. From a study of restraint deaths, it was common to ἀnd victims to be obese and it was inferred that when placed in the prone position their obesity may impede diaphragmatic movements and respiratory function. The term positional asphyxia has been applied to this and related phenomena. Other reports highlighted not only the position and manner of restraint but the frequent accompanying excited delirium and its physiological effects as causes of death in such cases (Stratton et al. 1995; Pollanen et al. 1998; DiMaio and DiMaio 2006). In experimental studies reported by Chan et al. (1997, 2004), volunteers were placed in the sitting, supine, and prone positions, restrained and loaded after a period of exercise, and their pulmonary function monitored. Chan et al. (1997, 2004) found no evidence that body position while hog-tied or hobbled could in itself cause asphyxiation and that other factors beyond body position were probably more important in restraint deaths. The authors did not rule out that very obese individuals, who were not tested, might have some special risk of asphyxia due to their obesity. Others have expressed concerns that restraint circumstances, including the body habitus of the victim, may produce respiratory embarrassment (Parkes 2008; DiMaio and DiMaio 2006; Pollanen et al. 1998). A factor in the excited and delirious victim deaths is the stressful general physiological state that exists when violent, struggling, and agitated conditions occur. Some have referred to this state as “catecholamine” storm (LaPosata 2006; DiMaio and DiMaio 2006). From the extensive reviews in other chapters of this volume, the importance of catecholamine excess and autonomic imbalance or over-activity in the pathomechanism of SUDEP cases,

48 Sudden Death in Epilepsy: Forensic and Clinical Issues

it is not surprising that it may come into play when some epileptic individuals become delirious and agitated, perhaps in the postictal state, as illustrated in the previous cases. These types of cases, like all the remainder without epilepsy, are difficult to analyze and develop an unambiguous cause and mechanism of death. One can only collect as much information as one can about a given case, and somehow rank the factors that are present in coming to a causal judgment, if one can. Like Case B, sometimes forensic pathologists feel compelled to render a manner of death that might seem provocative, such as homicide. In such cases, one must remember that “homicide” is not necessarily murder but simply death at the hands of another. In this case, the pathologist somehow felt the preponderance of the evidence indicated that someone was responsible. His pronouncement was challenged, of course, in the legal proceedings that followed. In general, it is probably most rational unless very special circumstances are present to make use of the “undetermined” manner of death designation.

2.8â•…SUDEP in Criminal Cases In Chapter 24 of this book, Dr. Wanamaker recounts the case of an unfortunate woman who was assaulted by a home intruder and, as a result, suffered permanent and debilitating head injuries that caused her to become epileptic. Many years after the beating, she was found dead at home in her bathroom with no obvious anatomic cause of death. Her death was judged to be a SUDEP, which, according to the judgment of the U.S. attorney (the

Case D This 14-year-old girl was born with a large ventriculoseptal heart defect and tricuspid valve abnormality, for which surgery had been performed. Apparently, the surgery was a so-called Fontan procedure, which essentially diverted venous blood from the right atrium to the pulmonary trunk, bypassing the right ventricle (Kerendi et al. 2009). This is a treatment for tricuspid atresia, which she had. Within a few years of€ birth, she also developed a seizure disorder that was characterized as being of the€GTC variety, for which she was medicated with phenytoin with variable success. Not long after her heart surgery, it was necessary for her to be anticoagulated and to be treated with digitalis. About 2 months before she died for unknown reasons, her anticonvulsant medication was discontinued. A few weeks before her death, she was given a number of immunizations that caused her to become quite ill and necessitated her being conἀned to bed. A week before death, she commenced her menarche and had difficulties with maintaining her International Normalized Ratio (INR) in acceptable ranges. It appears she also experienced one or more seizures, with one of them causing a bite on her tongue that continued to bleed. She complained of headaches. She was found dead beside her bed in the morning after having been ill for about 2 weeks. An autopsy was performed that showed evidence of a head impact in the left temporal region, an acute subdural hemorrhage, an apparent basal skull fracture, a large

Forensic Considerations and Sudden Unexpected Death in Epilepsy

laceration with hemorrhage of the tongue with aspiration of blood, a number of focal hemorrhagic sites in the gut, mesentery bladder, spleen, and kidney, and some bruising on the legs. The autopsy pathologist concluded that the girl had been assaulted and classiἀed her death as being due to blunt force injuries with manner of death as undetermined. The local prosecutor was pursuing an indictment against the mother of the child for homicide. Independent examination of the autopsy tissues with microscopic examinations indicated that the subdural hematoma was relatively thin and was 1 to 2 weeks old by histological criteria. A skull fracture was present. It was reasonable to attribute the numerous areas of hemorrhage to dosage problems with her anticoagulant and not to having been assaulted. As to the mechanism(s) of death, the victim clearly had serious cardiovascular disease, which in and of itself could account for a sudden death, but there was evidence of recent seizure activity as exempliἀed by the hemorrhagic tongue bite and aspiration of blood and the position of the body beside the bed. It might be said that SUDEP was a reasonable possibility, but could not be proven. The case has languished for many years and it is unlikely to be adjudicated.

Case E A 10-year-old boy who had experienced a birth-related injury that left him with a spastic form of diplegia, intractable epilepsy, and mental retardation had been cared for in various institutions and by the parents in the child’s home over his lifetime. The child was capable of ambulating but with difficulty, could not speak intelligibly but apparently could understand spoken language, and was not apparently visually impaired. After a number of years of gastrostomy feedings, the child was capable of being fed manually and even managed to partially feed himself. The child had a ventriculoperitoneal shunt for hydrocephalus with several shunt revisions, but apparently the hydrocephalus had stabilized. The child’s abiding problem was one of seizures that were cyclical in severity and frequency. A number of anticonvulsant medications had been tried, but never fully controlled the child’s epilepsy which was of the GTC type. One morning, the child was found dead in bed by the parents. An autopsy was performed by a medical examiner facility and the child was found to have some contractures of the lower extremity and many bruises on them and on the upper extremities. Several bruises were found on the child’s face and head. A number of healing abrasions were found on the arms and legs. There were no fractures. No pneumonia or other infections were found. The child’s brain showed typical rostral cortical atrophy with some cyst formation in the white matter, hydrocephalus that was not severe, and bilateral subdural hematoma membranes with a minimal amount of recent hemorrhage. The viscera were within normal limits. The autopsy pathologist concluded that the child had died from blunt force injuries that were inflicted and the manner of death was deemed homicide. The parents were charged with homicide.

49

50 Sudden Death in Epilepsy: Forensic and Clinical Issues

An independent examination of the case and the autopsy materials revealed the obvious cerebral palsy pathology of the child and revealed the subdural hematoma to be likely many years old and essentially undatable with accuracy. The recent hemorrhage was minimal. The skin bruises were of different ages by visual inspection, as were the healing abrasions. Put in the context of how the child functioned (impaired mobility, frequent falls with and without seizures), the injuries were not inconsistent with the circumstances of the child. No acute signiἀcant impact site was found on the head. The subdural hematoma bleeding was found to be typical for any chronic subdural hematoma and could not speciἀcally be related to any known traumatic episode. Given the location of the body (in bed), the chronic problem of seizures, and no anatomic cause of death, the child’s death was concluded to be SUDEP. Eventually, charges were dropped against the parents.

assault occurred on a federal reservation), was directly caused by the assailant who became known to him and was charged. Such a case is distinctly unusual, not only because of the basic facts of the case, but also because of the time that had elapsed between the inflicted injury and the woman’s death and when her assailant was brought to justice. Most criminal cases that might involve a SUDEP scenario are much less complex or unusual but nonetheless can be challenging if not controversial. A not uncommon circumstance that may involve criminal prosecution is the death of a neurologically compromised child, usually at home but occasionally in some form of care facility. In many of these cases, the child suffered from congenital brain malformations or birth-related injury (cerebral palsy) associated with chronic seizures. Sometimes, the child may have been the victim of an acute life-threatening event (ALTE) with persistent neurological impairment that may have been due to inflicted injury (abuse), accident, or to a circumstance that is never explained. When such a child dies, commonly due to aspiration and pneumonia, urinary tract infection, shunt malfunction, or other issue related to chronic illness, these conditions can usually be identiἀed and put in their proper context with a competent autopsy. Occasionally, the child will not have an obvious anatomic cause of death and may ἀt into the SUDEP category. The forensic pathologist must be mindful of this latter form of death and not leap to the conclusion that, absent an obvious cause of death, someone must have caused it. Cases such as this one point out the many challenges of evaluating and interpreting infant and childhood deaths, which almost invariably have a cloud of abusive injury hovering over them, deserved or not.

2.9â•…SUDEP and Anticonvulsant Medication Whether anticonvulsant medication by some aberrant pharmacological reaction, noncompliance, under-dosage, or cessation of administration can lead to SUDEP remains a controversial subject and is dealt with in detail in several chapters in the volume. This

Forensic Considerations and Sudden Unexpected Death in Epilepsy

51

discussion will not be repeated here. Suffice it to say that this issue can have forensic and litigative consequences. Dr. Wanamaker has highlighted some case examples of this problem that largely center about physician-patient interactions, namely the responsibility of the doctor to warn an epileptic patient about the possibility of SUDEP when prescribing, switching drugs, tapering, or discontinuing anticonvulsant drugs. Such cases come, not infrequently, to civil litigation. The above case is similar to many that have reached litigation in the United States and elsewhere in recent years. In many cases, the doctor has no record of patient conversations or instructions, relying on memory or usual practice. Recently such conversations are recorded, videotaped, or immortalized in some fashion, sometimes along with a signed statement by the patient that they were informed of risks of medication and its discontinuance, including the facts about SUDEP. Informational packets have been prepared by some physicians to aid in this process. To the author’s knowledge, the majority of SUDEP-related

Case F This 28-year-old woman had been epileptic since childhood, suffering ἀrst from absence seizures that then gave way to GTC seizures as an adult. Other than her epilepsy, she had enjoyed good health and had recently gotten married. Her seizure frequency was very low, and it had been several years since she had had a seizure. Since her marriage, she had wanted to become pregnant and had indeed done so, visiting her neurologist once she was aware of her pregnancy to seek advice about her anticonvulsant medication and its potential effects on her unborn baby. The doctor, who immortalized his conversation with his patient in a very brief note in his office records, indicated that probably the risk to the baby was quite low, but that if the patient was concerned about her medication, in the face of no seizures for several years, it would be appropriate to taper the medication and discontinue it for the duration of the pregnancy. He pointed out that if a seizure were to occur, they would simply begin medication again and go on. A few months later, the woman was found dead in bed. Autopsy revealed the pregnancy, but no anatomic cause of death. The husband reported that he had never observed a seizure in his wife nor had she said anything about having had a seizure. A lawsuit stipulating neglectful treatment and wrongful death was brought against the neurologist. At trial, the doctor’s records were offered in evidence to support the fact that he had a treatment plan and had discussed it with his patient, but apparently did not feel bound to warn her about the risk of SUDEP. At trial, various experts offered testimony regarding what SUDEP was, how it was diagnosed, and, probably, what caused it. The issue of anticonvulsant medication in relation to the possible cause and prevention of SUDEP was explored. The defense expert presented the results of several of his own studies and those of others and indicated that, while many studies showed a statistical correlation of subtherapeutic or no anticonvulsant medication with SUDEP, others did not and it could not be inferred that stopping medication was causal or could predict SUDEP. The jury in the case acquitted the doctor in the action.

52 Sudden Death in Epilepsy: Forensic and Clinical Issues

litigations have been adjudicated in favor of the physicians. Suffice it to say that it is vital that physicians share information with their patients and document that this occurred and what was communicated and when.

References Armour, J. A., and J. L. Ardell, eds. 1984. Neurocardiology. New York, NY: Oxford University Press. Bell, G. S., and J. W. Sander. 2006. Sudden unexpected death in epilepsy. Risk factors, possible mechanisms and prevention: A reappraisal. Acta Neurol Taiwan 15 (2): 72–83. Bharati, S., and M. Lev. 1990. The Cardiac Conduction System in Sudden Unexplained Death. Mt. Kisco, NY: Futura. Chan, T. C., G. M. Vilke, T. Neuman, and J. L. Clausen. 1997. Restraint position and positional asphyxia. Ann Emerg Med 30 (5): 578–586. Chan, T. C., T. Neuman, J. Clausen, J. Eisele, and G. M. Vilke. 2004. Weight force during prone restraint and respiratory function. Am J Forensic Med Pathol 25 (3): 185–189. Davis, J. H., and R. K. Wright. 1980. The very sudden cardiac death syndrome—a conceptual model for pathologists. Hum Pathol 11 (2): 117–121. de la Grandmaison, G. L., and M. Durigon. 2002. Sudden adult death: A medico-legal series of 77 cases between 1995 and 2000. Med Sci Law 42 (3): 225–232. DiMaio, T. G., and V. J. M. DiMaio, eds. 2006. Excited Delirium Syndrome. Cause of Death and Prevention. Boca Raton, FL: CRC Press. DiMaio, V. J., and D. DiMaio. 2001. Forensic pathology. In Practical Aspects of Criminal and Forensic Investigations, ed. V. Geberth. Boca Raton, FL: CRC Press. Engel, J., Jr., and T. A. Pedley, eds. 2008. Epilepsy: A Comprehensive Textbook. 2nd ed. Philadelphia, PA: Wolters Kluwer/Lippincott Williams & Wilkins. Fineschi, V., G. Baroldi, and M. D. Silver. 2006. Pathology of the Heart and Sudden Death in Forensic Medicine. Boca Raton, FL: CRC Press. Gastaut, H. 1969. Classiἀcation of the epilepsies. Proposal for an international classiἀcation. Epilepsia 10 (Suppl): 14–21. Gastaut, H., and B. G. Zifkin. 1985. Classiἀcation of the epilepsies. J Clin Neurophysiol 2 (4): 313–326. George, J. R., and G. G. Davis. 1998. Comparison of anti-epileptic drug levels in different cases of sudden death. J Forensic Sci 43 (3): 598–603. Greenberg, H. M., and E. M. Dwyer Jr., eds. 1982. Sudden Coronary Death, Vol. 382, Annals of the New York Academy of Sciences. New York, NY: New York Academy of Sciences. Haerem, J. W. 1978. Sudden, unexpected coronary death. The occurrence of platelet aggregates in the epicardial and myocardial vessels of man. Acta Pathol Microbiol Scand Suppl (265): 1–47. Hitiris, N., S. Suratman, K. Kelly, L. J. Stephen, G. J. Sills, and M. J. Brodie. 2007. Sudden unexpected death in epilepsy: A search for risk factors. Epilepsy Behav 10 (1): 138–141. James, S. H., P. E. Kish, and T. P. Sutton. 2005. Principles of bloodstain pattern analysis. Theory and practice. In Practical Aspects of Criminal and Forensic Investigations, ed. V. Geberth. Boca Raton, FL: CRC Press. Johnson, R. H., D. G. Lambie, and J. M. K. Spalding, eds. 1984. Neurocardiology, Major Problems in Neurology, Vol. 13, ed. J. Walton. Philadelphia, PA: W. B. Saunders. Kerendi, F., Z. B. Kramer, W. T. Mahle, B. E. Kogon, K. R. Kanter, and P. M. Kirshbom. 2009. Perioperative risks and outcomes of atrioventricular valve surgery in conjunction with Fontan procedure. Ann Thorac Surg 87 (5): 1484–1488; discussion 1488–1489. Knight, B. 1996. Forensic Pathology. 2nd ed. London: Arnold. Kuller, L. 1966. Sudden and unexpected non-traumatic deaths in adults: A review of epidemiological and clinical studies. J Chronic Dis 19 (11): 1165–1192.

Forensic Considerations and Sudden Unexpected Death in Epilepsy

53

Kuller, L., A. Lilienfeld, and R. Fisher. 1966. Epidemiological study of sudden and unexpected deaths due to arteriosclerotic heart disease. Circulation 34 (6): 1056–1068. Laidlaw, J., and A. Richens, eds. 1982. A Textbook of Epilepsy. Edinburgh: Churchill Livingstone. Langan, Y. 2000. Sudden unexpected death in epilepsy (SUDEP): Risk factors and case control studies. Seizure 9 (3): 179–183. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64 (7): 1131–1133. LaPosata, E. A. 2006. Restraint stress. In Sudden Deaths in Custody, ed. D. L. Ross and T. C. Chan. Totowa, NJ: Humana Press. Lear-Kaul, K. C., L. Coughlin, and M. J. Dobersen. 2005. Sudden unexpected death in epilepsy: A retrospective study. Am J Forensic Med Pathol 26 (1): 11–17. Leestma, J. E. 1990a. Natural history of epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E. 1990b. Sudden unexpected death associated with seizures: A pathological review. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E. 2009a. Forensic aspects of adult general neuropathology. In Forensic Neuropathology, ed. J. E. Leestma. Boca Raton, FL: CRC Press. Leestma, J. E. 2009b. Forensic aspects of complex neural function. In Forensic Neuropathology, ed. J.€E. Leestma. Boca Raton, FL: CRC Press. Leestma, J. E., and E. W. Sharp. 2009. Pathology and neuropathology in the forensic setting. In Forensic Neuropathology, ed. J. E. Leestma. Boca Raton, FL: CRC Press. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earl. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38 (1): 47–55. Leestma, J. E., J. R. Hughes, S. S. Teas, and M. B. Kalelkar. 1985. Sudden epilepsy deaths and the forensic pathologist. Am J Forensic Med Pathol 6 (3): 215–218. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26 (2): 195–203. Lhatoo, S. D., and J. W. Sander. 2002. Sudden unexpected death in epilepsy. Hong Kong Med J 8 (5): 354–358. Luke, J. L., and M. Helpern. 1968. Sudden unexpected death from natural causes in young adults. A review of 275 consecutive autopsied cases. Arch Pathol 85 (1): 10–17. Lund, A., and H. Gormsen. 1985. The role of antiepileptics in sudden death in epilepsy. Acta Neurol Scand 72 (4): 444–446. McGugan, E. A. 1999. Sudden unexpected deaths in epileptics—A literature review. Scott Med J 44 (5): 137–139. McKee, J. R., and J. W. Bodἀsh. 2000. Sudden unexpected death in epilepsy in adults with mental retardation. Am J Ment Retard 105 (4): 229–235. Milroy, C. M. 1999. Forensic taphonomy: The postmortem fate of human remains. BMJ 319 (7207): 458. Modell, J. H., M. Bellefleur, and J. H. Davis. 1999. Drowning without aspiration: Is this an appropriate diagnosis? J Forensic Sci 44 (6): 1119–1123. Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16 (1): 1–7. Morentin, B., and R. Alcaraz. 2002. Sudden unexpected death in epilepsy in children and adolescents. Rev Neurol 34 (5): 462–465. Moritz, A. R. 1954. Sudden and unexpected death due to disease. GP 10 (6): 35–42. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55 (6): 857–860.

54 Sudden Death in Epilepsy: Forensic and Clinical Issues Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 353 (9156): 888–893. Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42 (5): 667–673. Office of the Medical Examiner, Cook County, IL. 1979. Annual Report, 1977–1979. Chicago, IL: Office of the Medical Examiner. O’Halloran, R. L. 2004. Reenactment of circumstances in deaths related to restraint. Am J Forensic Med Pathol 25 (3): 190–193. O’Halloran, R. L., and J. G. Frank. 2000. Asphyxial death during prone restraint revisited: A report of 21 cases. Am J Forensic Med Pathol 21 (1): 39–52. O’Halloran, R. L., and L. V. Lewman. 1993. Restraint asphyxiation in excited delirium. Am J Forensic Med Pathol 14 (4): 289–295. Okada, R., and S. Kawai. 1983. Histopathology of the conduction system in sudden cardiac death. Jpn Circ J 47 (5): 573–580. Opeskin, K., A. S. Harvey, S. M. Cordner, and S. F. Berkovic. 2000. Sudden unexpected death in epilepsy in Victoria. J Clin Neurosci 7 (1): 34–37. Opeskin, K., M. P. Burke, S. M. Cordner, and S. F. Berkovic. 1999. Comparison of antiepileptic drug levels in sudden unexpected deaths in epilepsy with deaths from other causes. Epilepsia 40 (12): 1795–1798. Pachar, J. V., and J. M. Cameron. 1993. The diagnosis of drowning by quantitative and qualitative diatom analysis. Med Sci Law 33 (4): 291–299. Parkes, J. 2008. Sudden death during restraint: Do some positions affect lung function? Med Sci Law 48 (2): 137–141. Piette, M. H., and E. A. De Letter. 2006. Drowning: Still a difficult autopsy diagnosis. Forensic Sci Int 163 (1–2): 1–9. Pollanen, M. S., D. A. Chiasson, J. T. Cairns, and J. G. Young. 1998. Unexpected death related to restraint for excited delirium: A retrospective study of deaths in police custody and in the community. CMAJ 158 (12): 1603–1607. Reay, D. T., and J. W. Eisele. 1982. Death from law enforcement neck holds. Am J Forensic Med Pathol 3 (3): 253–258. Reay, D. T., and G. A. Holloway Jr. 1982. Changes in carotid blood flow produced by neck compression. Am J Forensic Med Pathol 3 (3): 199–202. Reay, D. T., C. L. Fligner, A. D. Stilwell, and J. Arnold. 1992. Positional asphyxia during law enforcement transport. Am J Forensic Med Pathol 13 (2): 90–97. Reay, D. T., J. D. Howard, C. L. Fligner, and R. J. Ward. 1988. Effects of positional restraint on oxygen saturation and heart rate following exercise. Am J Forensic Med Pathol 9 (1): 16–18. Ross, D. L., and T. C. Chan, eds. 2006. Sudden Deaths in Custody. Totowa, NJ: Humana Press. Rossi, L. 1982. Pathologic changes in the cardiac conduction and nervous system in sudden coronary death. Ann N Y Acad Sci 382: 50–68. Ruttenber, A. J., J. Lawler-Heavner, M. Yin, C. V. Wetli, W. L. Hearn, and D. C. Mash. 1997. Fatal excited delirium following cocaine use: Epidemiologic ἀndings provide new evidence for mechanisms of cocaine toxicity. J Forensic Sci 42 (1): 25–31. Ruttenber, A. J., H. B. McAnally, and C. V. Wetli. 1999. Cocaine-associated rhabdomyolysis and excited delirium: Different stages of the same syndrome. Am J Forensic Med Pathol 20 (2): 120–127. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schwartz, C. J., and R. G. Gerrity. 1975. Anatomical pathology of sudden unexpected cardiac death. Circulation 52 (6 Suppl): III18–III26. Schwartz, C. J., and W. J. Walsh. 1971. The pathologic basis of sudden death. Prog Cardiovasc Dis 13 (5): 465–481.

Forensic Considerations and Sudden Unexpected Death in Epilepsy

55

Schwender, L. A., and J. C. Troncoso. 1986. Evaluation of sudden death in epilepsy. Am J Forensic Med Pathol 7 (4): 283–287. Scorza, F. A., M. Albuquerque, R. M. Arida, and E. A. Cavalheiro. 2007. Cardiovascular alterations and sudden death in epilepsy. Arq Neuropsiquiatr 65 (2B): 461–466. Shields, L. B., D. M. Hunsaker, 3rd, J. C. Hunsaker, and J. C. Parker Jr. 2002. Sudden unexpected death in epilepsy: Neuropathologic ἀndings. Am J Forensic Med Pathol 23 (4): 307–314. Stratton, S. J., C. Rogers, and K. Green. 1995. Sudden death in individuals in hobble restraints during paramedic transport. Ann Emerg Med 25 (5): 710–712. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36 (1): 29–36. Thom, M., B. Griffin, J. W. Sander, and F. Scaravilli. 1999. Amygdala sclerosis in sudden and unexpected death in epilepsy. Epilepsy Res 37 (1): 53–62. Timmings, P. L. 1993. Sudden unexpected death in epilepsy: A local audit. Seizure 2 (4): 287–290. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61. Ulrich, J., and H. Maxeiner. 2003. Tongue bite injuries—a diagnostic criterium for death in epileptic seizure? Arch Kriminol 212 (1–2): 19–29. Vilke, G. M. 2006. Neck holds. In Sudden Deaths in Custody, ed. D. L. Ross and T. C. Chan. Totowa, NJ: Humana Press. Vlooswijk, M. C., H. J. Majoie, M. C. De Krom, I. Y. Tan, and A. P. Aldenkamp. 2007. SUDEP in the Netherlands: A retrospective study in a tertiary referral center. Seizure 16 (2): 153–159. Walczak, T. 2003. Do antiepileptic drugs play a role in sudden unexpected death in epilepsy? Drug Saf 26 (10): 673–683. Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. D. Cascino, J. F. Annegers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56 (4): 519–525.

Omega-3 Fatty Acids in Sudden Unexpected Death in Epilepsy Guardian of the Brain–€Heart Connection

3

Fulvio A. Scorza Esper A. Cavalheiro Ricardo M. Arida Vera C. Terra Carla A. Scorza Eliza Y. F. Sonoda Roberta M. Cysneiros

Contents 3.1 General Aspects of Sudden Unexpected Death in Epilepsy 3.2 Evidence for the Cardiovascular Beneἀts of Omega-3 3.3 The Potential Role of Omega-3 in the Management of Epilepsy 3.4 Final Considerations and Future Remarks Acknowledgment References

57 59 59 62 63 63

3.1â•…General Aspects of Sudden Unexpected Death in Epilepsy Epilepsy is the most common serious chronic neurological condition in the world. Approximately 50 million people worldwide have epilepsy (Annegers 1997; Sander 2003; Yuen and Sander 2004). Epidemiological studies suggest that, with proper treatment, between 70% and 80% of people developing epilepsy will go into remission, while the remaining patients continue to have seizures and are refractory to treatment with the currently available therapies (Sander 1993; Kwan and Sander 2004). The most common risk factors for epilepsy are cerebrovascular diseases, brain tumors, alcohol, traumatic head injuries, malformations of cortical development, genetic inheritance, and infections of the central nervous system (Halatchev 2000). In resource-poor countries, endemic infections, such as malaria and neurocysticercosis, seem to be major risk factors (Duncan et al. 2006). Temporal lobe epilepsy (TLE) is the most common form of partial epilepsy, affecting at least 20% of all patients with epilepsy (Babb 1999), and is the most common form of drugrefractory epilepsy (Engel 1993). Atrophy of mesial temporal structures is well known to be associated with TLE and hippocampal sclerosis, which is a very frequent histological abnormality in this form of epilepsy (Cendes et al. 1993). Sudden unexpected death in epilepsy (SUDEP) represents the leading cause of mortality€ in adults with refractory epilepsy (Tomson et al. 2005). SUDEP is deἀned as sudden, 57

58 Sudden Death in Epilepsy: Forensic and Clinical Issues

unexpected, witnessed or unwitnessed, nontraumatic, and nondrowning deaths in patients with epilepsy, with or without evidence of a seizure and excluding documented status epilepticus, in which postmortem examination does not reveal a toxicological or anatomical cause of death (Nashef 1997). Comparisons of incidence estimates for SUDEP are difficult since different deἀnitions of SUDEP are used, not all patients have a postmortem examination, and case ascertainment methods and source populations are varied (Tomson et al. 2005). The reported incidence of sudden unexpected deaths has been variably estimated as 3.5/1000 person-years* in a Lamotrigine clinical trial (Leestma et al. 1997), 0.5–1.4/1000 person-years in people with treated epilepsy (Tennis et al. 1995), 5.9/1000 person-years in outpatients with epilepsy at a tertiary referral center (Nashef et al. 1995), 9/1000 person-years in candidates for epilepsy surgery, and 0.35/1000 person-years in the general population (Ficker et al. 1988). Only one death attributable to SUDEP has been recorded in approximately 8000 personyears of follow-up in the National General Practice Study of Epilepsy in the United Kingdom, a community-based prospective observational study of incidence cases with epilepsy realized in a population of 792 patients with newly diagnosed epilepsy (Lhatoo et al. 1999). The results of the Medical Research Council Antiepileptic Drug Withdrawal Study showed that SUDEP among patients with epilepsy in remission is a rare event (Medical Research Council 1991). Information concerning risk factors for SUDEP is conflicting, but potential risk factors include age (Tennis et al. 1995), early onset of epilepsy (Nilsson et al. 1999), duration of epilepsy (Walczak et al. 2001), uncontrolled seizures, mainly TLE (Sperling et al. 1999; Langan et al. 2005), seizure frequency (Sperling et al. 1999; Walczak et al. 2001; Langan et al. 2005), seizure type (Kloster and Engelskjon 1999; Walczak et al. 2001; Langan et al. 2005), AED numbers (Nilsson et al. 1999; Walczak et al. 2001), and winter temperatures (Scorza et al. 2007). More recently, an elegant review article by Lathers and colleagues (2008) addressed other risk factors that could be operant in SUDEP and in cardiac disease, such as psychological factors (stress/anxiety/depression states and anger/frustration), patients’ lifestyle (e.g., cigarette smoking), and high prevalence of metabolic syndrome (obesity, dyslipidemia, glucose intolerance, and hypertension). Additionally, potential pathomechanisms for SUDEP are unknown, but a number of interictal, ictal, and postmortem cardiac abnormalities account for the possibility of seizure-induced cardiogenic SUDEP (Tomson et al. 2005; Scorza et al. 2007). Along these lines, early animal studies have also examined the role of the autonomic nervous system and cardiac arrhythmias, interictal and ictal epileptogenic activity, and SUDEP (Lathers and Schraeder 1982). Speciἀcally, Lathers and Schraeder demonstrated that intravenous pentylenetetrazol in anesthetized cats induced autonomic dysfunction associated with both interictal and ictal epileptogenic activity. The autonomic dysfunction was manifested by the fact that autonomic cardiac nerves did not always respond in a predictable manner to changes in blood pressure, the development of a marked increase in variability in mean autonomic cardiac nerve discharge, and the appearance of a very large increase in the variability of the discharge rate of parasympathetic nerves, ἀrst and then second, in sympathetic discharge. The authors concluded that altered autonomic cardiac nerve discharge was associated with interictal epileptogenic activity and arrhythmia, which may contribute to sudden unexplained death in patients with epilepsy (Lathers and Schraeder 1982).

* Person-years: the specific amount of time (years) members of the population experience over the risk period.

Omega-3 Fatty Acids in Sudden Unexpected Death in Epilepsy

59

3.2â•…Evidence for the Cardiovascular Benefits of Omega-3 The beneἀcial effects of omega-3 fatty acids in the cardiovascular system have been described since the 1970s. Dyerberg and Bang (1979) suggested the low prevalence of cardiac diseases in Eskimos might be due to high dietary ingestion of omega-3 fatty acids. In the same decade, a number of experimental studies showed an antiarrhythmic role of the omega-3 fatty acid (Gudbjarnason and Hallgrimsson 1976). McLennan et al. (1992) conducted several experiments to conἀrm the possible antiarrhythmic role of omega-3. Initially, it was demonstrated that a diet high in ἀsh oil (tuna) could prevent the ventricular ἀbrillation in rats induced by coronary artery occlusion following reperfusion (McLennan et al. 1992; McLennan 1993). Subsequent studies conἀrmed their previous data (McLennan et al. 1992; Yang et al. 1993; Jacobsen et al. 1997). Billman and colleagues (1997), using a sudden cardiac death model in dogs, observed a reduction of arrhythmias and sudden cardiac death following induction of ischemia after omega-3 administration (Billman et al. 1994, 1997). Moreover, several work studies demonstrated positive effects of omega-3 in arrhythmia reduction (Kang and Leaf 1994; Xiao et al. 1995; Kang and Leaf 1996; Xiao et al. 1997, 2000). The antiarrhythmic mechanism induced by omega-3 was modulated by Na+ and Ca2+ currents in cardiac cells (Leaf et al. 2003). In the same way, positive effects of omega-3 have also been observed in clinical studies. Accordingly, several studies in the 1990s demonstrated that ἀsh consumption one or two times per week was associated with a 50% reduction of sudden cardiac death (Albert et al. 1998; Siscovick et al. 2003). Furthermore, a clinical study for infarct prevention conducted by an Italian group (GISSIPrevenzione), using a dose of 850 mg/day of omega-3 in 11,324 subjects presenting a ἀrst myocardial infarction episode, observed a signiἀcant reduction of death caused by cardiovascular complications (30%) as well as in sudden death (45%) when compared to control subjects, suggesting an antiarrhythmic role of omega-3 (Marchioli 1999; Marchioli et al. 2002). In conclusion, human and animals studies have demonstrated a possible action of omega-3 in the prevention of cardiovascular abnormalities and the reduction of occurrence of sudden cardiac death.

3.3â•…The Potential Role of Omega-3 in the Management of Epilepsy As discussed above, SUDEP is the most important direct epilepsy-related situation of death (Tomson et al. 2005) and cardiovascular abnormalities during and between seizures is directly related to a high frequency of SUDEP (Stollberger and Finsterer 2004). In addition, substantial evidence from epidemiological and case-control studies indicates that omega-3 reduces the risk of cardiovascular mortality, with an especially potent effect on sudden cardiac death (Calder 2004). Thus, a possible relationship between omega-3, epilepsy, and SUDEP could be considered. Along these lines, some animal and clinical studies have shown that omega-3 could be useful in the prevention and treatment of epilepsy. In 1998, Voskuyl and colleagues, using the cortical stimulation seizure model in rats, demonstrated a modest anticonvulsant effect of long duration with the polyunsaturated fatty acids. Moreover, pharmacology studies show that polyunsaturated fatty acids applied extracellularly raise the stimulatory thresholds of CA1 neurons in hippocampal slices (Xiao and Li 1999). Recently, our group was the ἀrst to demonstrate that chronic treatment with omega-3 promotes neuroprotection

60 Sudden Death in Epilepsy: Forensic and Clinical Issues

and increases parvalbumin-positive neurons in the hippocampal formation of rats with epilepsy (Ferrari et al. 2008), suggesting that omega-3 leads to prominent positive plastic changes in the hippocampal formation of rats with epilepsy (Figure 3.1). From a clinical viewpoint, Schlanger et al. (2002) reported on an open clinical trial in which ἀve patients took omega-3 supplements. In this study, a special food spread containing 65% omega-3 fatty acids was added to the daily diet of ἀve patients with epilepsy. The patients consumed 5 g of this spread at every breakfast for 6 months and all of them showed a marked reduction in both frequency and severity of epileptic seizures, suggesting that a dietary supplement containing omega-3 may be beneἀcial in the suppression of some cases of epileptic seizures. Recently, in the ἀrst randomized, placebo-controlled, parallel group of omega-3 supplementation, seizure frequency was reduced over the ἀrst 6 weeks of treatment in the supplement group, but this effect was not sustained (Yuen and Sander 2004). The authors think the decrease in seizures following the initial 6 weeks is a result of omegaÂ�-3 preparations, doses, treatment duration, and sample sizes. Speciἀcally, the researchers used a total daily dose of approximately 1.7 g omega-3, approximately 1 g eicosapentaenoic acid (EPA)

CV

CW

CA1

EV

EW

HILUS CA3

CA1

CA3

Hilus

Figure 3.1╇ Photomicrographs of PV-stained coronal sections in the hippocampal formation (CA1, CA3 regions and hilus of the dentate gyrus) of animals treated daily with vehicle (cremophor 0.009%) (CV); animals treated daily with 85 mg/kg omega-3 (CW); animals with epilepsy treated with vehicle (EV) and animals with epilepsy treated with 85 mg/kg omega-3, on a daily basis (EW). PV-positive neurons are significantly higher in animals with epilepsy treated with omega-3 (EW) when compared with untreated animals with epilepsy (EV), and control animals (CV, CW). (Scale bar, 800 µm.)

Omega-3 Fatty Acids in Sudden Unexpected Death in Epilepsy

61

and 0.7 g docosahexaenoic acid (DHA), and suggested that additional doses and omega-3 preparations should be tested. Concerning treatment duration and sample sizes, a 12-week, double-blind, placebo-controlled, parallel group trial of omega-3 supplementation in 57 patients with refractory epilepsy was evaluated. Since a reduction in seizures within the ἀrst 6 weeks of starting supplementation with omega-3 was observed, the authors also purposed that studies using longer treatment durations with larger sample sizes should be undertaken. As omega-3 fatty acids, per se, have been shown to reduce cardiac arrhythmias and sudden cardiac deaths (Calder 2004), Yuen and Sander (2004) proposed the interesting hypothesis that omega-3 fatty acid supplementation in patients with refractory seizures may reduce seizures and seizure-associated cardiac arrhythmias and, hence, SUDEP. According to this reasoning, two experimental studies developed by our group could reinforce this hypothesis. First, we evaluated the heart rate, using in vivo (ECG) and isolated ex vivo preparations (Langendorf preparation) of rats with epilepsy (Colugnati et al. 2005). Our results showed signiἀcant differences in the mean heart rate in vivo between the groups. In contrast, we did not ἀnd differences during isolated ex vivo situations, suggesting a central nervous system modulation on the heart, which could explain a risk for SUDEP (Figure€3.2). Quite interestingly, Scorza and colleagues (unpublished data) evaluated the same set of experiments (heart rate in vivo and isolated ex vivo preparation) in rats

(a) 300 200 100 0

Heart rate in vitro (bpm)

200

Control

Rats with epilepsy Ventricular pressure (mm Hg) in vitro

Heart rate in vivo (bpm)

400

(b)

150 100 50 0

Control

Rats with epilepsy

80 70

(c)

60 50 40 30 20 10 0

Control

Rats with epilepsy

Figure 3.2╇ Results obtained by Colugnati, Gomes, and Arida (2005). (a) Heart rate in vivo; Heart rate of isolated ex vivo preparation (b); Ventricular pressure of isolated ex vivo situation (c) from control and rats with epilepsy. Note that heart rate in vivo is significantly higher in rats with epilepsy. *p < 0.05.

62 Sudden Death in Epilepsy: Forensic and Clinical Issues

with epilepsy before and after chronic omega-3 administration. The results showed differences in the mean heart rate in vivo but, surprisingly, no differences in heart rates could be observed in the isolated ex vivo condition. It has been suggested that the use of certain antiepileptic drugs, especially carbaÂ� mazepine (CBZ), may predispose patients to SUDEP (Stollberger and Finsterer 2004). Furthermore, carbamazepine has been shown to affect the autonomic nervous system and the conduction system of the heart, mainly when associated with elevated plasma levels above 40 µmol/l (Stollberger and Finsterer 2004). Following this reasoning, Yuen and coworkers (2008) reported very recently the fatty acids (FAs) proἀles in red blood cells and plasma obtained prior to and following omega-3 FA supplementation in 56 patients with epilepsy, providing an opportunity to examine potential effects of antiepileptic drugs on FA proἀles (Yuen et al. 2008). They showed that patients on CBZ exhibited a less favorable FAs proἀle, associated with a greater risk of coronary heart disease mortality. In summary, the authors concluded that, as arrhythmias are thought to be important mechanisms in coronary heart disease mortality and SUDEP, the effect of CBZ in reducing omega-3 FAs might potentially explain some cases of SUDEP among patients prescribed the antiepileptic drug. Taking all of the data together, the ἀrst randomized, placebo-controlled parallel group study of omega-3 supplementation in patients with chronic epilepsy showed only a transient effect on seizure frequency that was not conἀrmed by other research groups, but indicates that additional trials are required (DeGiorgio and Miller 2008; Scorza et al. 2008). These results did not totally conἀrm that omega-3 fatty acids reduce the frequency of epileptic seizures in patients with intractable epilepsy; however, they established the safety of omega-3 supplementation in people with epilepsy (DeGiorgio and Miller 2008). Quite interestingly, there is now great interest in n-3 fatty acids for the prevention of SUDEP (DeGiorgio and Miller 2008; Scorza et al. 2008) and we have to bear in mind that, as omega-3 fatty acids have been shown to reduce cardiac arrhythmias and sudden cardiac deaths, it could be suggested that omega-3 fatty acid supplementation in patients with refractory seizures may reduce seizures and seizure-associated cardiac arrhythmias and, hence, SUDEP. Nonetheless, it is very important to emphasize that nutritional therapy (e.g., omega-3 supplementation) is not a substitute for anticonvulsant medications.

3.4â•…Final Considerations and Future Remarks Polyunsaturated fatty acids are present at high levels in the brain (DeGiorgio and Miller 2008). The 04 polyunsaturated fatty acids are EPA and DHA, which are n-3 fatty acids (omega-3) and dihomogammalinolenic acid (DGLA) and arachidonic acid (AA), which are n-6 fatty acids (omega-6) (Calon and Cole 2007; Ohara 2007). Dietary consumption of the long-chain omega-3 fatty acids, commonly found in ἀsh and ἀsh oil, not only contribute to human central nervous system development, but also may modify the risk for certain adult nervous system diseases, including epilepsy (Voskuyl et al. 1998; Bourre 2004). Following this reasoning, eating ἀsh is certainly a good way to improve the development of the brain, ranging from the composition of cell membranes to cerebral function. All omega-3 fatty acids are important for treatment or prevention of cardiovascular and neurological diseases (Bourre and Paquotte 2008; Scorza et al. 2008). The only foods that provide large amounts of omega-3 are seafood such as certain types of ἀnἀsh and shellἀsh. The highest

Omega-3 Fatty Acids in Sudden Unexpected Death in Epilepsy

63

omega-3 concentrations with the lowest levels of mercury are anchovies, Atlantic herring, Atlantic mackerel, scallops, wild salmon, canned salmon, sardines, and trout (Smith and Sahyoun 2005; Bourre and Paquotte 2008). In sum, a number of different dietary modiἀcations, nutritional supplements, and hormones may help prevent seizures or improve other aspects of health in patients with epilepsy (Gaby 2007). Based on this, new diet-based supplementary therapeutic strategies should be developed with the aim of decreasing seizure frequency and severity in people with epilepsy. Concerning cardioprotective omega-3, it has long been believed that a daily intake of 3000 to 4000 mg of ἀsh oil supplements or two to three servings of fatty ἀsh per week are safe and effective for adults in general, including those with neurological, inflammatory, and neurodegenerative diseases, such as multiple sclerosis, Alzheimer’s disease, and Parkinson’s disease (Mazza et al. 2007). While it is very important to emphasize that nutritional therapy (including omega-3 supplementation) is not a substitute for anticonvulsant medications, experimental, epidemiological, and clinical studies should be evaluated to establish with precision the possible supplementary beneἀt of omega-3 in the treatment of epilepsy.

Acknowledgment The authors thank Danuza Ferrari for technical assistance, and FAPESP, CInAPCeFAPESP, and CNPq for supporting this study.

References Albert, C. M., C. H. Hennekens, C. J. O’Donnell, et al. 1998. Fish consumption and risk of sudden cardiac death. JAMA 279: 23–28. Annegers, J. F. 1997. Epidemiology of epilepsy. In The Treatment of Epilepsy: Principles and Practice, 2nd ed., ed. E. Wyllie, 165–172. Baltimore, MD: Williams & Wilkins. Babb, T. L. 1999. Synaptic reorganizations in human and rat hippocampal epilepsy. Adv Neurol 79: 763–779. Billman, G. E., H. Hallaq, and A. Leaf. 1994. Prevention of ischemia-induced ventricular ἀbrillation by omega 3 fatty acids. Proc Natl Acad Sci U S A (91): 4427–4430. Billman, G. E., J. X. Kang, and A. Leaf. 1997. Prevention of ischemia-induced cardiac sudden death by n-3 polyunsaturated fatty acids in dogs. Lipids 32: 1161–1168. Bourre, J. M. 2004. Roles of unsaturated fatty acids (especially omega-3 fatty acids) in the brain at various ages and during aging. J Nutr Health Aging 8: 163–174. Bourre, J. M., and P. Paquotte. 2008. Seafood (wild and farmed) for the elderly: Contribution to the dietary intakes of iodine, selenium, DHA and vitamins B12 and D. J Nutr Health Aging 12: 186–192. Calder, P. C. 2004. N-3 fatty acids and cardiovascular disease: Evidence explained and mechanisms explored. Clin Sci 107: 1–11. Calon, F., and G. Cole. 2007. Neuroprotective action of omega-3 polyunsaturated fatty acids against neurodegenerative diseases: Evidence from animal studies. Prostaglandins Leukot Essent Fatty Acids 77: 287–293. Cendes, F., F. Leproux, D. Melanson, et al. 1993. MRI of amygdala and hippocampus in temporal lobe epilepsy. J Comput Assist Tomogr 17: 206–210. Colugnati, D. B., P. A. Gomes, R. M. Arida, et al. 2005. Analysis of cardiac parameters in animals with epilepsy: Possible cause of sudden death? Arq Neuropsiquiatr 23: 1035–1041.

64 Sudden Death in Epilepsy: Forensic and Clinical Issues DeGiorgio, C. M., and P. Miller. 2008. N-3 fatty acids (eicosapentanoic and docosahexanoic acids) in epilepsy and for the prevention of sudden unexpected death in epilepsy. Epilepsy Behav 13: 712–713. Duncan, J. S., J. W. Sander, S. M. Sisodiya, and M. C. Walker. 2006. Adult epilepsy. Lancet 367: 1087–1100. Dyerberg, J., and H. O. Bang. 1979. Haemostatic function and platelet polyunsaturated fatty acids in Eskimos. Lancet 2: 433–435. Engel, Jr., J. 1993. Update on surgical treatment of the epilepsies. Summary of the Second International Palm Desert Conference on the Surgical Treatment of the Epilepsies. Neurology 43: 1612–1617. Ferrari, D., R. M. Cysneiros, C. A. Scorza, et al. 2008. Neuroprotective activity of omega-3 fatty acids against epilepsy-induced hippocampal damage: Quantiἀcation with immunohistochemical for calcium-binding proteins. Epilepsy Behav 13: 36–42. Ficker, D. M., E. L. So, W. K. Shen, et al. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 515: 1270–1274. Gaby, A. R. 2007. Natural approaches to epilepsy. Altern Med Rev 12: 9–24. Gudbjarnason, S., and J. Hallgrimsson. 1976. Prostaglandins and polyunsaturated fatty acids in heart muscle. Acta Biol Med Ger 35: 1069–1080. Halatchev, V. N. 2000. Epidemiology of epilepsy—recent achievements and future. Folia Med (Plovdiv) 42: 17–22. Jacobsen, A. N., X. J. Du, A. M. Dart, and E. A. Woodcock. 1997. Ins (1, 4, 5) P3 and arrhythmogenic responses during myocardial reperfusion: Evidence for receptor speciἀcity. Am J Physiol 273: H1119–H1125. Kang, J. X., and A. Leaf. 1994. Effects of long-chain polyunsaturated fatty acids on the contraction of neonatal rat cardiac myocytes. Proc Natl Acad Sci U S A 91: 9886–9890. Kang, J. X., and A. Leaf. 1996. Antiarrhythmic effects of polyunsaturated fatty acids. Recent studies. Circulation 94: 1774–1780. Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67: 439–444. Kwan, P., and J. W. Sander. 2004. The natural history of epilepsy: An epidemiological view. J Neurol Neurosurg Psychiatry (75): 1376–1381. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Leaf, A., J. X. Kang, Y. F. Xiao, and G. E. Billman. 2003. Clinical prevention of sudden cardiac death by n-3 polyunsaturated fatty acids and mechanism of prevention of arrhythmias by n-3 ἀsh oils. Circulation 107: 2646–2652. Leestma, J. E., J. F. Annegers, M. J. Brodie, et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. Lhatoo, S. D., Y. Langan, B. K. MacDonald, S. Zeidan, and J. W. Sander. 1999. Sudden unexpected death: A rare event in a large community based prospective cohort with newly diagnosed epilepsy and high remission rates. J Neurol Neurosurg Psychiatry 66: 692–693. Marchioli, R. 1999. Results of GISSI Prevenzione: Diet, drugs, and cardiovascular risk. Researchers of GISSI Prevenzione. Cardiologia 44: 745–746. Marchioli, R., F. Barzi, E. Bomba, et al. 2002. GISSI-Prevenzione Investigators. Early protection against sudden death by n-3 polyunsaturated fatty acids after myocardial infarction: Time-course analysis of the results of the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico (GISSI)-Prevenzione. Circulation 105: 1897–1903.

Omega-3 Fatty Acids in Sudden Unexpected Death in Epilepsy

65

Mazza, M., M. Pomponi, L. Janiri, P. Bria, and S. Mazza. 2007. Omega-3 fatty acids and antioxidants in neurological and psychiatric diseases: An overview. Prog Neuropsychopharmacol Biol Psychiatry 31: 12–26. McLennan, P. L. 1993. Relative effects of dietary saturated, monounsaturated, and polyunsaturated fatty acids on cardiac arrhythmias in rats. Am J Clin Nutr 57: 207–212. McLennan, P. S., T. M. Bridle, M. Y. Abeywardena, and J. S. Charnock. 1992. Dietary lipid modulation of ventricular ἀbrillation threshold in the marmoset monkey. Am Heart J 123: 1555–1561. Medical Research Council. 1991. Antiepileptic Drug Withdrawal Study Group. Randomised study of antiepileptic drug withdrawal in patients in remission. Lancet 337: 1175–1180. Nashef, L. 1997. Sudden unexpected death in epilepsy: Terminology and deἀnitions. Epilepsia 38: S6–S8. Nashef, L., D. R. Fish, J. W. Sander, and S. D. Shorvon. 1995. Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral centre. J Neurol Neurosurg Psychiatry 58: 462–464. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 353: 888–893. Ohara, K. 2007. The n-3 polyunsaturated fatty acid/dopamine hypothesis of schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry 31: 469–474. Sander, J. W. 1993. Some aspects of prognosis in the epilepsies: A review. Epilepsia 34: 1007–1016. Sander, J. W. 2003. The epidemiology of epilepsy revisited. Curr Opin Neurol 16: 165–170. Schlanger, S., M. Shinitzky, and D. Yam. 2002. Diet enriched with omega-3 fatty acids alleviates convulsion symptoms in epilepsy patients. Epilepsia 43: 103–104. Scorza, F. A., M. Albuquerque, R. M. Arida, and E. A. Cavalheiro. 2007. Sudden unexpected death in epilepsy: Are winter temperatures a new potential risk factor? Epilepsy Behav 10: 509–510. Scorza, F. A., R. M. Cysneiros, R. M. Arida, V. C. Terra-Bustamante, M. de Albuquerque, and E. A. Cavalheiro. 2008. The other side of the coin: Beneἀciary effect of omega-3 fatty acids in sudden unexpected death in epilepsy. Epilepsy Behav 13: 279–283. Siscovick, D. S., R. N. Lemaitre, and D. Mozaffarian. 2003. The ἀsh story: A diet-heart hypothesis with clinical implications: n-3 polyunsaturated fatty acids, myocardial vulnerability, and sudden death. Circulation 107: 2632–2634. Smith, K. M., and N. R. Sahyoun. 2005. Fish consumption: Recommendations versus advisories, can they be reconciled? Nutr Rev 63: 39–46. Sperling, M. R., H. Feldman, J. Kinman, J. D. Liporace, and M. J. O’Connor. 1999. Seizure control and mortality in epilepsy. Ann Neurol 46: 45–50. Stollberger, C., and J. Finsterer. 2004. Cardiorespiratory ἀndings in sudden unexplained/unexpected death in epilepsy (SUDEP). Epilepsy Res 59: 51–60. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36: 29–36. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46: 54–61. Voskuyl, R. A., M. Vreugdenhil, J. X. Kang, and A. Leaf. 1998. Anticonvulsant effect of polyunsaturated fatty acids in rats, using the cortical stimulation model. Eur J Pharmacol 341 (2–3): 145–52. Walczak, T. S., I. E. Leppik, M. D’Amelio, et al. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56: 519–525. Xiao, Y., and X. Li. 1999. Polyunsaturated fatty acids modify mouse hippocampal neuronal excitability during excitotoxic or convulsant stimulation. Brain Res 846: 112–121. Xiao, Y. F., A. M. Gomez, J. P. Morgan, W. J. Lederer, and A. Leaf. 1997. Suppression of voltage-gated L-type Ca2+ currents by polyunsaturated fatty acids in adult and neonatal rat ventricular myocytes. Proc Natl Acad Sci U S A 94: 4182–4187. Xiao, Y. F., J. X. Kang, J. P. Morgan, and A. Leaf. 1995. Blocking effects of polyunsaturated fatty acids on Na+ channels of neonatal rat ventricular myocytes. Proc Natl Acad Sci U S A 92: 11000–11004.

66 Sudden Death in Epilepsy: Forensic and Clinical Issues Xiao, Y. F., S. N. Wright, G. K. Wang, J. P. Morgan, and A. Leaf. 2000. Coexpression with beta(1)subunit modiἀes the kinetics and fatty acid block of hH1(alpha) Na(+) channels. Am J Physiol Heart Circ Physiol 279: H35–H46. Yang, B. C., T. G. Saldeen, J. L. Bryant, W. W. Nichols, and J. L. Mehta. 1993. Long-term dietary ἀsh oil supplementation protects against ischemia–reperfusion-induced myocardial dysfunction in isolated rat hearts. Am Heart J 126: 1287–1292. Yuen, A. W., and J. W. Sander. 2004. Is omega-3 fatty acid deἀciency a factor contributing to refractory seizures and SUDEP? A hypothesis. Seizure 13: 104–107. Yuen, A. W., J. W. Sander, D. Flugel, et al. 2008. Erythrocyte and plasma fatty acid proἀles in patients with epilepsy: Does carbamazepine affect omega-3 fatty acid concentrations? Epilepsy Behav 12: 317–323.

Unanswered Questions SUDEP Studies Needed Claire M. Lathers Paul L. Schraeder Michael W. Bungo

4

Contents 4.1 Omega-3 Fatty Acid Nutritional Deἀciency 4.2 Low Temperatures as a Risk Factor for Cardiovascular Abnormalities and Sudden Death 4.3 Mechanisms of SUDEP: Possible Role of Autonomic Effects Such as Cardiorespiratory Disturbances 4.3.1 Beneἀcial Effects of Physical Activity Related to Reductions in Sympathetic Activity 4.3.2 Autonomic Cardiorespiratory Disturbances as Factors for SUD or SUDEP 4.3.3 Langendorf Isolated Heart Preparation References

67 68 69 69 69 71 72

Our recent review article entitled “Mystery of Sudden Death: Mechanisms for Risks” (Lathers et al. 2008a) examined the overlapping mechanisms that apply to the risk of sudden unexpected death occurring in epilepsy (SUDEP) and in cardiac disease. The information in this article has been updated in Chapter 1 of this book (Lathers et al. 2010). The role of central and peripheral autonomic nervous systems and the cardiopulmonary systems and the interaction between them was explored. The potential interactive role of genetically determined subtle cardiac risk factors for arrhythmias with a predisposition for seizure-Â�related cardiac arrhythmias and the possible mechanisms that are operant in producing both epileptogenic and cardiogenic arrhythmias were addressed. Potential preventive measures to minimize risk of both SUDEP and sudden cardiac death were presented. This information has been updated in several chapters in this book (Lathers 2010, animal models; Lathers and Schraeder 2010b, AED; Lathers and Schraeder 2010c, SUDEP animal model). Scorza et al. (2008, 2010, Chapter 3), raised important issues above and beyond standard medical intervention for the prevention of sudden death in cardiac patients, or in patients with epilepsy at risk for sudden death, by focusing on the importance of lifestyle issues. We discuss these lifestyle issues and comment on several points of clariἀcation of the information presented by Scorza et al. (2008, 2010, Chapter 3).

4.1â•…Omega-3 Fatty Acid Nutritional Deficiency The beneἀcial effect of nutritional aspects of omega-3 fatty acids may decrease cardiac arrhythmias and sudden death (SUD) in patients with cardiac disease. Some questions must be raised for future studies; for example: 67

68 Sudden Death in Epilepsy: Forensic and Clinical Issues

1. Does a nutritional deἀciency in omega-3 fatty acids become a risk factor for SUDEP? Today, there is no answer. Whether Omega-3 fatty acids deἀciency is a risk factor for the relatively young population prone to SUDEP, mean age 32.5–43.5 years (Leestma 1990), is unknown. Studies designed to address this question are needed. Unless someone has familial hypercholesterolemia with very high levels of cholesterol at a young age, we do not see that cholesterol is a clinically signiἀcant risk or predisposition to cardiac disease for the young population at risk for SUDEP. The beneἀcial cardiovascular effect of omega-3 fatty acids to decrease sudden death in cardiac patients may be due to actions on cholesterol levels. 2. Will this cardioprotective effect, independent of the effect on cholesterol levels, be of beneἀt to patients with epilepsy? The answer does not exist at this time, but there is nothing bad to be said for including omega-3 fatty acids in your diet to gain long term nutritional beneἀts, since eating ἀsh containing omega-3 fatty acids may decrease the risk of sudden cardiac death. We do not see a downside to consuming omega-3 fatty acids to decrease the risk of SUDEP, even though at this time there is no proven possibility that omega-3 fatty acids may also have a speciἀc beneἀcial effect on SUDEP. While nutritional therapy that includes omega-3 fatty acids supplementation is not a substitute for anticonvulsant medications, there may be a role for these fatty acids as a supplement to antiepileptic drugs (Scorza et al. 2008).

4.2â•…Low Temperatures as a Risk Factor for Cardiovascular Abnormalities and Sudden Death Scorza et al. (2008) discussed low ambient temperatures as a potential risk factor for cardiovascular abnormalities and, possibly, sudden death. When considering the possible role of low temperatures as a risk factor, several questions must be raised: 1. Exactly how does one deἀne low temperatures? 2. What is the degree range used to deἀne low temperatures and the duration necessary in this low degree range? 3. Is there a dose response with an increased risk of sudden death related to the lower temperatures? 4. If low temperatures are a risk factor for sudden death, how does one counsel patients to avoid cold? Should patients be advised to not ski, snow shoe, ice skate, or walk in the cold and be advised to dress warmly, avoid hypothermia, and move to Houston, Florida, or Southern Spain? During the winter, increases in hemoconcentration (erythrocyte count, plasma cholesterol, and plasma ἀbrinogen levels) occur and may contribute to arterial thrombosis (Neild et al. 1994). It is important to differentiate between the pathological and normal response of the body for protection versus the response to hypothermia. If these physiological changes are normal defense mechanisms, they may be difficult to modify except for a physical move to a temperate climate. One would also not expect patients living in a colder climate to stop participation in outdoor winter activities. Additional studies are needed to

Unanswered Questions: SUDEP Studies Needed

69

conἀrm the basis of this cold theory since it has been challenged by several epidemiological studies, including the study of Kloner et al. (1999).

4.3â•…Mechanisms of SUDEP: Possible Role of Autonomic Effects Such as Cardiorespiratory Disturbances 4.3.1â•…Beneficial Effects of Physical Activity Related to Reductions in Sympathetic Activity Morbidity and mortality in cardiovascular disease are often associated with elevations in sympathetic nervous system activity (Scorza et al. 2008). Increased physical activity does have a beneἀcial cardiovascular interaction. This raises the question: “Will increased physical activity be of beneἀcial input if one assumes cardiovascular sympathetic dysfunction or insufficiency in persons with epilepsy?” In general, exercise does not have a downside, if the patient is sufficiently cardiovascularly ἀt to perform the exercise regimen selected. If exercise is beneἀcial in preventing SUDEP, it will be a great ἀnding. However, even if there is no protective or preventive effect for sudden death resulting from exercise, there is certainly no reason for a patient with epilepsy not to exercise. 4.3.2â•…Autonomic Cardiorespiratory Disturbances as Factors for SUD or SUDEP Autonomic involvement in SUDEP is associated with even minimal epileptogenic activity and may cause altered autonomic activity and balance (Scorza et al. 2008). This may be characterized by alterations in cardiorespiratory parameters including ECG changes, blood pressure, respiration, and vasomotor tone (Hirsh and Martin 1971; Lathers and Schraeder 1982; Schraeder and Lathers 1983; Wannamaker 1985). ECG changes observed in experimental epilepsy include heart rate changes, arrhythmias, conduction blocks, altered ECG morphology, and QT interval changes (Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1983; Lathers et al. 1987; Mameli et al. 1988). Paroxysmal autonomic dysfunction associated with epileptogenic discharges may be a factor in sudden death as they cause a disruption of cardiac and/or respiratory function (Schraeder and Lathers 1990; Lathers and Schraeder 2010c, Chapter 28). Stress itself must also be considered in sudden death, since the occurrence of a seizure may cause the release of catecholamines from the adrenal glands (Doba et al. 1975) and may predispose the heart to development of arrhythmias when confronted with a sudden sympathetic neural barrage induced by epileptogenic discharges (Lathers et al. 2008a, 2008b; Schraeder and Lathers 1990; Pickworth et al. 1990; Lathers and Schraeder 2006, 2010a, 2010b, 2010c). The central nervous system modiἀes the heart in response to environmental stress (Lathers and Schraeder 2006, 2010a). Adverse emotional states impact the autonomic control of cardiac rhythm and are known factors leading to cardiac dysrhythmias in humans. Interactions between emotional factors and the arrythmogenic potential of epileptiform discharges, as well as the possible beneἀt from stress management intervention have yet to be investigated (Lathers and Schraeder 2006). Central nervous system modulation of peripheral autonomic neural discharges to the heart is a risk factor for sudden death in patients with cardiac disease (Lathers et al.

70 Sudden Death in Epilepsy: Forensic and Clinical Issues

1977, 1978; Evans and Gillis 1978) and is also a risk factor for sudden death in persons with epilepsy. Lathers and Schraeder (1982) and Schraeder and Lathers (1983) developed a cat model with an intact central nervous system to explore epilepsy/autonomic neural/ cardiopulmonary system relationships as possible mechanisms of SUDEP. They reported cardiac sympathetic neural nonuniform discharges in association with seizure discharges congruent with increased temporal dispersion of the recovery of ventricular excitability, leading to an underlying electrical instability that predisposes the ventricular myocardium to arrhythmia (Han and Moe 1964). Druschky et al. (2001) reported sympathetic dysfunction in the form of altered postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy and suggested altered postganglionic cardiac sympathetic innervation may increase risk of cardiac abnormalities and/or SUDEP. Studies are needed to explore both the developmental and regulatory mechanisms responsible for determining the density and pattern of cardiac sympathetic innervation and the exact role of disturbances associated with the exact location of the innervation in arrhythmogenesis (see discussion by Lathers and Levin 2010, Chapter 33). When evaluating mechanisms of SUDEP, one must also consider that there may be an underlying, undeἀned genetic predisposition to arrhythmias. Thus, it is very likely that the mechanism of sudden death varies from patient to patient, depending on a predisposition to neurogenically induced autonomic dysfunction, be it sympathetic or parasympathetic, and on the status of symptoms and disease present at the time of death. Animal data (Lathers et al. 1987; Stauffer et al. 1989; Lathers et al. 1990) demonstrated the lockstep phenomenon (LSP), postganglionic cardiac sympathetic and vagal discharges that were time-locked to ictal and interictal cortical epileptiform activity. It was suggested that the LSP may explain propagation of electrical impulses to autonomic nervous system regulatory centers, thus initiating arrhythmogenic potentials and premature ventricular contractions, ST/T changes, and conduction blocks occurring concurrently with the cerebral discharges (see Chapter 28, Lathers and Schraeder 2010c, LSP). Numerous experimental preparations could be devised to study in vivo and in vitro methods to modify function of the central nervous system and peripheral autonomic neural activity—reserpine, 6-OH dopamine or bretylium denervation, surgical denervation, adrenalectomy, and various combinations (Lathers et al. 1981, 1982; Lathers 2010, animal models). Epileptogenic activity is associated with cardiac rate and rhythm changes (Howell and Blumhardt 1990). Quint et al. (1990) discussed power spectral analysis of the heart rate as a technique to assess autonomic activity related to risk factors for SUDEP. Goodman et al. (1990; 2010, Chapter 40) found a typical cardiovascular response during a generalized kindled seizure in rats consisted of a large increase in blood pressure accompanied by a profound bradycardia during the ἀrst 20–30 seconds of the seizure. Lathers and Schraeder (1982) reported bradycardia associated with interictal subconvulsant epileptogenic in cats that died in asystole; an increase in heart rate occurred in animals dying in ventricular ἀbrillation. GABA neurotransmission regulates, in part, the central control of cardiovascular function (Schwartz and Lathers 1990). Inhibition of central GABAergic tone using picrotoxin or bicuculline results in enhanced sympathetic outflow to the heart and increased coronary resistance and arrhythmias (Segal et al. 1984). GABAA, introduced into the cerebroventricles prior to the agonist muscimol, prevents these cardiac changes. It may be clearly seen that neuronal excitability/inhibitory state in the central nervous system is important for development of seizures and cardiovascular abnormalities, including arrhythmias. When patients with generalized tonic–clonic seizures on no medications were compared with normal controls, the patients exhibited higher standard deviation of all R–R

Unanswered Questions: SUDEP Studies Needed

71

intervals, standard deviation of mean R–R intervals in 5-min recordings, and heart rate variability compared to normal controls. Heart rate variability exhibited a reduction in high-frequency parasympathetic activity and an increase in low-frequency sympathetic values (Everengul et al. 2005). Heart rate variability in epilepsy patients is lower at night than during the daytime, indicating subtle autonomic dysfunction at night, when risk of SUDEP is greatest (Woodley et al. 1977). Control of the heart rate is perturbed by alterations in neuroautonomic function in clinical syndromes of sudden cardiac death, congestive failure, space sickness, and physiologic aging. Fractal mechanisms in cardiac electrophysiology suggest that loss of normal fractal complexity in the interbeat interval dynamics may not be detected using conventional statistics but are quantiἀed using methods derived from the “chaos theory” (Goldberger 1991). There is a loss of complex physiological variability in pathological conditions, including heart rate dynamics, before sudden death. Cardiac chaos is prevalent in healthy hearts and a decrease of cardiac chaos occurs in congestive heart failure (Poon and Merrill 1997). Use of detrended fluctuation analysis of heart beat time series may distinguish healthy from pathologic data sets (Peng et al. 1995). Studies on patients need to be done to determine how to identify which persons with epilepsy are at risk for SUDEP. Arrhythmogenesis and acute hypoxia have been shown to interact in a negative manner and the combination becomes a risk factor for SUDEP. Acute pulmonary edema may develop during a clinical seizure and lung weights are known to be increased in animal studies (Lathers and Schraeder 1982; Carnel et al. 1985) and in victims of SUDEP (Koehler et al. 2010, Chapter 9). Simon et al. (1982) and Johnston et al. (1995, 1997) demonstrated that seizure induction in a sheep model produced acute pulmonary edema and a neurogenically induced increase in pulmonary microvascular pressure with an accompanying prolonged change in endothelial conductance to protein. In some animals, the neurogenic pulmonary edema was linked to seizures associated with central nervous system–induced apnea. Thus, when an increased risk for arrhythmias is combined with acute hypoxemia associated with the occurrence of neurogenic pulmonary edema and/or associated central apnea, sudden death may be the outcome. For additional information about the role of arrhythmogenesis and acute hypoxia as risk factors for SUDEP, the reader is referred to Lathers et al. (2008), Mameli and Alessandro (2010, Chapter 37), and Lathers et al. (2010, Chapter 1). 4.3.3â•…Langendorf Isolated Heart Preparation Cardiac arrhythmias may trigger sudden death in cardiac patients, originating from three sources: directly from the heart; from the peripheral autonomic nervous system; from the central nervous system; or from combinations of some or all of these factors (Lathers et al. 1977, 1978). Central and peripheral autonomic neural input is not present in the Langendorf isolated beating heart preparation. This preparation is used to examine drug actions on the heart, independent of autonomic neural input and independent of circulating substances such as catecholamines (Lathers et al. 1981, 1982). In vivo animal data may be compared with data obtained in the Langendorf preparation to study the contributory role of the autonomic nervous system to cardiac arrhythmias and/or death. Langendorf hearts may be obtained from animals denervated chemically by pretreatment with reserpine, bretylium or 6-hydroxydopamine, or from rats whose hearts were surgically denervated. Comparison of data from these various animal models reveals data about the role

72 Sudden Death in Epilepsy: Forensic and Clinical Issues

of the actual heart rate in initiation of cardiac arrhythmias and/or sudden death. Thus, the various experimental designs allow clariἀcation of the role of autonomic innervation and influences of the central and peripheral nervous system in relation to epileptogenic activity, heart rate, arrhythmias, and sudden death. Comparison of mean heart rates in vivo and in isolated ex vivo preparation (Langendorf preparation) in rats with epilepsy has been done (Colugnati et al. 2005). Data showed differences in mean heart rates in vivo, but no differences in heart rate in ex vivo, suggesting a central nervous system modulation of the heart may lead to SUDEP. This conclusion seems logical since the Langendorf preparation is an isolated heart preparation devoid of input from the central nervous system. The Langendorf preparation has been used to study the ability of digitalis glycosides to induce arrhythmias and seizures (Lathers, 1990). In conclusion, mechanisms of SUDEP include risk categories of arrhythmogenic factors, respiratory factors and hypoxia, and psychological factors (Lathers et al. 2008a, 2008b). Clariἀcation of risk factors and establishment of the mechanisms of SUDEP will help to develop preventative measures for SUDEP. It appears to be most important to attempt full seizure control. Scorza et al. (2008) encourage patients with epilepsy to receive nonmedical, lifestyle-modifying interventions that have accepted public health beneἀts, even though there is as yet no consensus that they may or may not prevent sudden death. Both animal studies and clinical studies are needed to deἀnitely address the risk factors of omega-3 fatty acids, cold temperatures, exercise, and heart rate for development of cardiac arrhythmias and/or SUDEP. Prospective studies of persons with epilepsy are needed to determine how to identify which persons with epilepsy are at risk for SUDEP (Lathers et€al. 2008a, 2008b).

References Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity. Pharmacology 20: 225–240. Colugnati, D. B., P. A. Gomes, R. M. Arida et al. 2005. Analysis of cardiac parameters in animals with epilepsy: Possible cause of sudden death? Arq Neuropsiquiatr 46: 54–61. Doba, N., H. R. Beresford, and D. J. Reis. 1975. Changes in regional blood flow and cardiodynamics associated with electrically and chemically induced epilepsy in the cat. Brain Res 47: 487–491. Druschky, A., M. J. Hiltz, P. Hopp et al. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by 123I meta-iodobenzylguanidine-SPECT. Brain 124: 372–382. Evans, D. E., and R. A. Gillis. 1978. Reflex mechanisms involved in cardiac arrhythmias induced by hypothalamic stimulation. Am J Physiol 234: H199–H209. Everengul, H., H. Tanriverdi, D. Dursunoglu et al. 2005. Time frequency domain analyses of heart rate variability in patients with epilepsy. Epilepsy Res 63: 131–139. Fowler, R. S., L. Rathl, and J. D. Keith. 1964. Accidental digitalis intoxication in children. J Pediatr 64: 188–199. Goldberger, A. L. 1992. Fractal mechanisms in the electrophysiology of the heart. IEEE Eng Med Biol Mag 11: 47–52. Goldberger, A. L. 1991. Is the normal heartbeat chaotic or homeostatic? New Physiol Sci 6: 87–91. Goodman, J., R. W. Homan, and I. L. Crawford. 2010. Cardiovascular responses during kindled seizures. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 40. Boca Raton, FL: CRC Press.

Unanswered Questions: SUDEP Studies Needed

73

Goodman, J. H., R. W. Homan, and I. L. Crawford. 1990. Acute cardiovascular response during kindled seizures. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter€11, 169–186. New York, NY: Marcel Dekker. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hirsh, C. S., and L. M. Martin. 1971. Unexpected death in young epileptics. Neurology 21: 682–690. Howell, S. J., and L. D. Blumhardt. 1990. The role of EEG monitoring in the diagnosis of epilepsyrelated cardiac arrhythmias and of cardiac arrhythmias mimicking epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 7, 101–119. New York, NY: Marcel Dekker. Johnston, S. C., J. K. Horn, U. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37: 531–537. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42: 588–594. Kloner, R. A., W. K. Poole, and R. L. Perritt. 1999. When throughout the year is coronary death most likely to occur?: A 12-year population-based analysis of more than 200,000 cases. Circulation 100: 1630–1634. Koehler, S. A., P. L. Schraeder, C. M. Lathers, and C. Wecht. 2010. Forensic postmortem results of sudden epilepsy death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 9. Boca Raton, FL: CRC Press. Lathers, C. M. 2010. Sudden death animal models to study nervous system sites of action for disease and pharmacological intervention. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 25. Boca Raton, FL: CRC Press. Lathers, C. M. 1990. Glycoside-induced arrhythmias and seizures. 1990. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 20, 375–391. New York, NY: Marcel Dekker. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin, J. C. Krusz, G. J. Kelliher, and J. Roberts. 1981. The action of reserpine, 6-hydroxydopamine, and bretylium on digitalis-induced cardiotoxicity. Eur J Pharmacol 76: 371–379. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin et al. 1982. Role of the adrenergic nerve terminal in digitalis-induced cardiac toxicity: A study of the effects of pharmacological and surgical denervation. J Cardiovasc Pharmacol 4: 91–98. Lathers, C.M., K. F. Jim, W. H. Spivey, C. Kahn, K. Dolce, and W. D. Matthews. 1990. Antiepileptic activity of beta-blocking agents. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 24, 485–505. New York, NY: Marcel Dekker. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57: 1058–1065. Lathers, C. M., and R. M. Levin. 2010. Animal model for sudden cardiac death: Sympathetic innervation and myocardial beta-receptor densities. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 33. Boca Raton, FL: CRC Press. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias and epileptogenic activity. J Clin Pharmacol 27: 346–356.

74 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., and P. L. Schraeder. 2010a. Stress and sudden death in persons with epilepsy (SUDEP). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M.€W. Bungo, and J. E. Leestma, Chapter 17. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 2010b. Antiepileptic drugs. Beneἀt/risk clinical pharmacology. Possible role in cause or prevention of SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 47. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 2010c. Animal model for sudden unexpected death in persons with epilepsy (SUDEP). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 28. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and M. Bungo. 2008a. Mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Reply. Epilepsy Behav 13: 265–269. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Risk factors for sudden death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 1. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and F. L. Winer. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity. The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Leestma, J. E. 1990. Natural history of epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 1, 11–26. New York, NY: Marcel Dekker. Mameli, O., and C. M. Alessandro. 2010. Sudden epileptic death in experimental animal models. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 37. Boca Raton, FL: CRC Press. Mameli, P., O. Mameli, E. Tolu et al. 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29: 74–82. Neild, P. J., D. Syndercombe-Court, W. R. Keatinge, G. C. Donaldson, M. Mattock, and M. Caunce. 1994. Cold-induced increases in erythrocyte count, plasma cholesterol and plasma ἀbrinogen of elderly people without a comparable rise in protein C or factor X. Clin Sci 86: 43–48. Peng, C. K., S. Havlin, H. E. Stanley, and A. L. Goldberger. 1995. Quantiἀcation of scaling exponents and crossover phenomena in nonstationary heartbeat time series. Chaos 5: 82–87. Pickworth, W. B., J. Gerard-Ciminera, and C. M. Lathers. 1990. Stress, arrhythmias, and seizures. In Epilepsy and Sudden Death, ed C. M. Lathers and P. L. Schraeder, Chapter 22, 417–446. New York, NY: Marcel Dekker. Poon, C. S., and C. K. Merrill. 1997. Decrease of cardiac chaos in congestive heart failure. Nature 389: 492–495. Quint, S. R., J. A. Messenheimer, and M. B. Tennison. 1990. Power spectral analysis as a procedure for assessing autonomic activity related to risk factors for sudden and unexplained death in epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 16, 261–291. New York, NY: Marcel Dekker. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci. 32: 1371–1382. Schraeder, P. L., and C. M. Lathers. 1990. Paroxysmal autonomic dysfunction. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 8, 121–133. New York, NY: Marcel Dekker. Schwartz, R. D., and C. M. Lathers. 1990. CM GAMA transmission, epileptogenic activity, and cardiac arrhythmias. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. Chapter€17, 293–308. New York, NY: Marcel Dekker. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13: 263–244.

Unanswered Questions: SUDEP Studies Needed

75

Scorza, F. A., A. Esper, A. Cavalheiro et al. 2010. Omega-3 fatty acids in SUDEP: Guardian of the brain-heart connection. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 3. Boca Raton, FL: CRC Press. Segal, S. A., T. Jacob, and R. A. Gillis. 1984. Blockade of central nervous system GABAergic tone causes sympathetic-mediated increases in coronary vascular resistance in cats. Circ Res 55: 404–414. Simon, R. P., L. L. Bayne, R. F. Tranbaugh, and F. R. Lewis. 1982. Elevated pulmonary flow and protein content during status epilepticus in sheep. J Appl Physiol 52: 91–95. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72: 340–345. Wannamaker, B. B. 1985. Autonomic nervous system and epilepsy. Epilepsia 26 (S1): 31–39. Woodley, D., W. Chambers, H. Starke, B. Dzindzio, and A. D. Forker. 1977. Intermittent complete atrioventricular block masquerading as epilepsy in the mitral valve prolapse syndrome. Chest 72: 369–372.

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome Implications in SUDEP

5

David S. Paterson

Contents 5.1 5.2 5.3 5.4

Sudden Infant Death Syndrome Serotonin and the Medullary 5-HT System Abnormalities in the Medullary 5-HT System in SIDS Medullary 5-HT Abnormalities and SIDS Risk Factors 5.4.1 Sleep Environment and Infection 5.4.2 Male Gender 5.4.3 Maternal Smoking 5.4.4 Maternal Drinking 5.5 5-HT Gene Polymorphisms and SIDS 5.6 SIDS as a Developmental Disorder That Originates in Utero 5.7 SIDS Results from a Combination of€Environmental and Genetic Factors 5.8 Brain Stem 5-HT Dysfunction and SUDEP 5.9 Conclusions References

77 78 79 80 81 81 83 84 84 85 85 86 87 87

5.1â•…Sudden Infant Death Syndrome The sudden infant death syndrome (SIDS) is deἀned as the sudden death of an infant less than 1 year of age, with the onset of the fatal episode occurring during sleep, that remains unexplained after a thorough investigation, including performance of a complete autopsy and review of the circumstances of death and the clinical history (Krous et al. 2004). Despite signiἀcant reductions in SIDS rates in recent years due to successful risk-reduction campaigns, SIDS remains the leading type of death of infants between 1 month and 1 year of age in developed countries (Moon et al. 2007). The majority (90%) of SIDS deaths occur within the ἀrst 6 postnatal months, with the peak incidence observed at 2–4 months of age (Task Force on Sudden Infant Death Syndrome 2005). Multiple studies have identiἀed robust associations between SIDS and environmental risk factors, including prone or face-down sleeping, bed sharing, and overbundling (Task Force on Sudden Infant Death Syndrome 2005). Several risk factors for SIDS relate to the mother and pregnancy, including prematurity, low birth weight, maternal cigarette smoking, and alcohol, cocaine, and heroin use during pregnancy (Kinney and Paterson 2004; Task Force on Sudden Infant Death Syndrome 2005). Evidence of recent infection is a common ἀnding in SIDS cases 77

78 Sudden Death in Epilepsy: Forensic and Clinical Issues Critical period

2–4 Postnatal months

Underlying abnormality

e.g., 5-HT defect

Vulnerable infant

SIDS

Prone sleeping, overbundling, infection

Stressor

Figure 5.1╇ The Triple Risk Model of SIDS. SIDS occurs when three factors impinge on the

infant simultaneously—an underlying vulnerability in the infant, a critical period in developÂ� ment, and a homeostatic stressor at the time of infant death. (From Filiano, J. J., and H. C. Kinney, Biol Neonate, 65 (3–4), 194–197, 1994. With permission.)

at autopsy (Task Force on Sudden Infant Death Syndrome 2005). There is also a male bias in SIDS with twice as many boys dying of SIDS than girls (Moon et al. 2007). The triple risk hypothesis of SIDS has proven useful in thinking about SIDS causation (see Figure 5.1) (Filiano and Kinney 1994). It posits that SIDS occurs when three factors impinge on the infant simultaneously—an underlying vulnerability in the infant, a critical period in development (the ἀrst 6 postnatal months when 90% of SIDS occurs), and a homeostatic stressor at the time of infant death (e.g., hypercapnia from rebreathing exhaled air as a result of sleeping prone in the face-down position). Increasing evidence, including studies from our laboratory, suggests that the underlying vulnerability involves a developmental defect in brain stem serotonergic (5-HT) systems that results in the failure of protective cardiorespiratory responses to potentially life-threatening but normally occurring events (e.g., hypoxia, hypercapnia), in the infant during sleep. In this chapter, I review the evidence to support a role of medullary 5-HT abnormalities in the pathogenesis of SIDS, the putative in utero origin of these abnormalities, and how environmental and genetic factors interact to ultimately result in sudden death of the infant. I begin by describing the medullary 5-HT system and its role in maintenance of homeostasis.

5.2â•…Serotonin and the Medullary 5-HT System Serotonin (5-HT) is a monoamine neurotransmitter, the synthesis of which involves the rate-limiting enzyme tryptophan hydroxylase (TPH), which is in turn the marker of the 5-HT cell phenotype. Neurons synthesizing 5-HT (and expressing TPH) are exclusive to the brain stem in distinct cell groups classically deἀned as B1–B9 that differentially project to virtually all regions of the neuraxis (Hornung 2003; Tork and Hornung 1990). These cell groups are divided into a rostral domain consisting of groups B4–B9, and a caudal domain consisting of groups B1–B3. The two domains of the brain stem 5-HT neurons are distinct

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome Caudal

Mid

GC

HG

Arc

Rostral

NTS

NTS

PIO

79

GC

Rob

PGCL

IRZ

Rob

PGCL

PIO

PIO Arc

Arc

Figure 5.2╇ The medullary 5-HT system consists of 5-HT neurons (areas in red) in the midline

raphe [raphe obscurus (Rob)], extra-raphe [gigantocellularis (GC), paragigantocellularis lateralis (PGCL), and intermediate reticular nucleus (IRZ)], and the ventral medullary surface [arcuate nucleus (Arc)] and the sites that they project to (blue areas) that do not contain 5-HT neurons, but mediate homeostatic functions [e.g., the hypoglossal nucleus (HG) and the nucleus of the solitary tract (NTS)]. This figure shows coronal sections at the level of the caudal, mid, and rostral medulla. (From Kinney, H. C. et al., Auton Neurosci, 132 (1–2), 81–102, 2007. With permission.)

in their developmental origins, functions, and connectivity (Hornung 2003; Tork and Hornung 1990). The rostral domain, located in the upper brain stem, projects “rostrally” and diffusely to the cerebral cortex, thalamus, hypothalamus, basal ganglia, hippocampus, and amygdala. It participates in the mediation of arousal, cognition, mood, motor activity, and cerebral blood flow. The caudal domain in the lower brain stem or medulla projects “caudally” and diffusely to other brain stem sites and to the cerebellum and spinal cord (Hornung 2003; Tork and Hornung 1990). Extensive animal data indicates that 5-HT neurons in the caudal medullary domain influence multiple homeostatic functions mediated by the medulla, including chemosensitivity (Bernard et al. 1996; Dreshaj et al. 1998; Messier et al. 2004; Richerson 1995, 1997, 2004; Wang et al. 1998), respiratory rhythm generation (Pena and Ramirez 2002; Tryba et al. 2006), drive (Bou-Flores et al. 2000), plasticity (Baker-Herman et al. 2004; Ling et al. 2001), blood pressure regulation (Henderson et al. 2000), thermoregulation (Berner et al. 1999), upper airway reflexes (Hilaire et al. 1993; Holtman et al. 1987; Kubin et al. 1992), and arousal (Bartlett et al. 1990; Darnall et al. 2005; Krammer et al. 1979). The medullary 5-HT system, as deἀned by us, consists of the 5-HT neuronal cell bodies located in the midline raphe, lateral extra-raphe, and ventral surface of the medulla in the caudal 5-HT domain (see Figure 5.2) (Kinney and Paterson 2004). We view the medullary 5-HT system as the “central integrator” for the coordination of synchronous homeostatic function via projections to and modulation of effector neurons, or mediating circuits that produce speciἀc homeostatic responses (Kinney et al. 2007).

5.3â•…Abnormalities in the Medullary 5-HT System in SIDS Abnormalities in markers of 5-HT function have been observed in the medullary 5-HT system (i.e., raphe obscurus, gigantocellularis, paragigantocellularis lateralis, intermediate reticular zone, arcuate nucleus, and hypoglossal nucleus) in SIDS infants, including an increased number of 5-HT neurons, many of which are immature (Paterson et al. 2006),

80 Sudden Death in Epilepsy: Forensic and Clinical Issues 3

H-8-OH-DPAT

Control

5-HT1A Receptor Binding Medullary 5-HT Source Nuclei

3

SIDS

Rob

H-8-OH-DPAT Binding Density (fmol/mg)

70 60 50 40 30 20 10 0 (a)

(b)

Rob

GC

PGCL

IRN

ARC

Figure 5.3╇ 5-HT1A receptor binding density in SIDS and control cases. (a) Illustrative autoradiograms displaying 3H-8-OH-DPAT binding (fmol/mg ± SEM) to 5-HT1A receptors in a tissue section from a control and a SIDS case at the mid-medulla level. The density of 5-HT1A receptor binding sites, including in the raphe obscurus (Rob), is visually lower in the SIDS case compared to the control case. (b) Graph comparing the density of 3H-8-OH-DPAT binding to 5-HT1A receptors in medullary nuclei containing 5-HT cells in SIDS and control cases. 5-HT1A receptor binding density is significantly lower in SIDS cases compared to controls. (***p€< 0.001, **p€< 0.01, *p < 0.05, ANCOVA.) RO, raphe obscurus; GC, gigantocellularis nucleus; PGCL, paragigantocellularis lateralis nucleus; IRN, intermediate reticular nucleus; ARC, arcuate nucleus. (From Paterson, D. S. et al., JAMA, 296 (17), 2124–2132, 2006. With permission.)

reduced 5-HT1A (see Figure 5.3), and 5-HT2A receptor expression (Kinney et al. 2003; Ozawa and Okado 2002; Ozawa and Takashima 2002; Panigrahy et al. 2000; Paterson et al. 2006), reduced 5-HTT binding (Paterson et al. 2006), abnormal TPH expression (Machaalani et al. 2009; Sawaguchi et al. 2003), reduced brain 5-HT levels (Sparks and Hunsaker 1991), altered 5-HT turnover (Cann-Moisan et al. 1999), and altered 5-HT breakdown (Sparks and Hunsaker 1991). These observations inform the idea that multiple elements of respiratory and autonomic regulation, mediated by the 5-HT system, are defective in SIDS, including but not restricted to respiratory rhythmogenesis and respiratory responses to hypercapnic and/or hypoxic challenge. This idea is supported by a SIDS infant who was observed to have subtle respiratory and cardiac dysfunction at birth and 5-HT receptor binding abnormalities at autopsy 2 weeks later (Kinney et al. 2005). Taken together, the above observations provide evidence to support the idea that medullary 5-HT abnormalities cause homeostatic dysfunction that potentially contributes to the death of the infant in SIDS, i.e., in terms of the Triple Risk Model of SIDS, the medullary 5-HT defect is, or is part of, the underlying abnormality that predisposes the infant to sleep-related death, particularly when combined with an environmental stressor, such as prone sleeping, during the critical developmental period.

5.4â•…Medullary 5-HT Abnormalities and SIDS Risk Factors Multiple studies have identiἀed robust correlations between environmental factors such as sleep position, infection, and prenatal exposure to cigarette smoke and alcohol, and

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

81

increased risk of SIDS. The mechanisms by which these risk factors adversely affect and predispose an infant to SIDS, however, are unknown and their elucidation is a major goal of SIDS research. In our SIDS studies, we have explored the potential effects of recognized SIDS risk factors on markers of 5-HT function to determine their potential role in the pathogenesis of SIDS. We categorize SIDS risk factors as either vulnerability factors or stressors based on the nature of their contribution to the Triple Risk Model of SIDS. Vulnerability factors are related to the development of the underlying abnormality (i.e., medullary 5-HT abnormality) that predisposes infants to increased SIDS risk and include male gender, race, and prematurity. Stressors are exogenous environmental factors that impact upon the infant at the time of death including prone sleep position, face-down sleeping, bed sharing, and infection. Prenatal and neonatal exposure to alcohol and cigarette smoke can be considered as both vulnerability factors and stressors because they may contribute to the development of the underlying abnormality as well as have a direct adverse effect on the protective physiological responses to homeostatic challenges in the infant at the time of death. The relationship between SIDS risk factors and medullary 5-HT abnormalities in SIDS is discussed below. 5.4.1â•…Sleep Environment and Infection In our most recent study of SIDS (Paterson et al. 2006), despite public health campaigns promoting safe sleep practices for infants, we found that 77% (24/31) of SIDS deaths were associated with an unsafe sleep environment (i.e., prone/side sleeping, or bed-sharing), with 48% (15/31) found in the prone sleep position. Similarly, we found that 42% (13/31) of SIDS cases had a history of illness in the days prior to death. Unsafe sleep environment, prematurity, and infection remain, therefore, major risk factors for SIDS. However, we did not observe a signiἀcant association between any of these risk factors and markers of 5-HT function in SIDS. These observations are based on a relatively small sample size and, thus, need to be repeated in a larger dataset to conἀrm their validity. However, these observations support the idea that environmental stressors expose an underlying vulnerability (i.e., medullary 5-HT abnormalities) in SIDS infants that renders them unable to adequately respond to changing physiological conditions, ultimately leading to the death of the infant. 5.4.2â•…Male Gender Twice as many male infants die of SIDS as female infants (Task Force on Sudden Infant Death Syndrome 2005). The reason for this is unknown, but the identiἀcation of a signiἀcantly lower density of 5-HT1A receptor binding in male compared to female SIDS infants (see Figure 5.4) (Paterson et al. 2006) suggests that sexual dimorphism in 5-HT function may play a role in predisposing male infants to SIDS. Indeed, signiἀcant differences in TPH, 5-HT, 5-HT metabolites, and 5-HT receptor expression normally exist between males and females in several brain regions including a lower level of 5-HT1A receptors (Arango et al. 1995; Dillon et al. 1991; Ferrari et al. 1999; Parsey et al. 2002). Evidence from studies in animals with 5-HT lesions have reported male gender-speciἀc abnormalities in respiration, chemosensitivity, and thermoregulation (Hodges et al. 2008; Li and Nattie 2008; Penatti et al. 2006) and responses that are modulated in part by 5-HT1A receptors in the medullary raphe and extra-raphe (Brown et al. 2008; Darnall et al. 2005; Hoffman et al.

82 Sudden Death in Epilepsy: Forensic and Clinical Issues 70 5-HT1A binding (fmol/mg)

60 50

Controls n = 6 Female SIDS n = 6 Male SIDS n = 6

40 30 20 10 0

Figure 5.4╇ 5-HT1A receptor binding in the raphe obscurus by sex. 5-HT1A receptor binding is

lower in male SIDS cases. Graph comparing 5-HT1A receptor binding density measured with 3H 8-OH DPAT autoradiography in the raphe obscurus in male and female SIDS cases, compared to controls infants. 5-HT1A receptor binding density in male SIDS infants is significantly lower compared to female SIDS infants (*p = 0.04). 5-HT1A receptor binding in both male (p = 0.02) and female (p = 0.05) SIDS infants is significantly lower compared to controls. (From Paterson, D.S. et al., JAMA, 296 (17), 2124–2132, 2006. With permission.)

2007; Messier et al. 2004). These observations raise the possibility that male human infants may similarly have reduced sensitivity to CO2 and temperature and that loss of medullary 5-HT1A receptors, as observed in SIDS, may attenuate protective homeostatic responses to a greater extent in male compared to female infants, thus placing them at greater risk for SIDS. Evidence from animal studies also suggests reduced plasticity in 5-HT neuron function in males compared to females. Deἀcits in postnatal brain levels of 5-HT1A receptor expression following prenatal cocaine and cigarette smoke exposure persist for a greater length of time in male compared to female rats (Johns et al. 2002; Slotkin et al. 2007a, 2007b), suggesting that the neonatal male infant brain is less resilient to exposure to at least some pharmacologically active toxins affecting 5-HT function in the maternal circulation than the neonatal female brain. The reason for these differences is unclear, but may in part be due to intrinsic differences in the level of sex steroids in males and females as both testosterone and estrogen influence the 5-HT system and its control of respiration (Bayliss et al. 1990; Bayliss and Millhorn 1992; Emery et al. 1994; Fogel et al. 2001; Liu et al. 2003; Matsumoto et al. 1985; Pickett et al. 1989; Regensteiner et al. 1990; Zhou et al. 2003). Moreover, SIDS infants have elevated levels of serum testosterone compared to controls (Emery et al. 2005); preterm infants—a high risk population—have signiἀcantly higher adrenal-derived androgens in the ἀrst year of life compared to term infants (Tapanainen et al. 1981); and the peak SIDS incidence between 2 and 4 months of age coincides with the peak postnatal increase in gonadal steroids (Forest et al. 1980; Peterson and Kennedy 1979; Winter et al. 1976). Thus, the normal higher levels of testosterone in male infants compared to female infants may be responsible for blunted respiratory responses to homeostatic challenges such as hypercapnia, thereby contributing to their greater SIDS risk. Taking these observations together, intrinsic differences in baseline brain 5-HT function, 5-HT neuronal plasticity, and CO2 sensitivity between males and females provide evidence that may explain, at least in part, the greater risk of SIDS in male infants.

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

83

5.4.3â•…Maternal Smoking The Aberdeen Indian Health Service Infant Mortality Study (Randall et al. 2001) involved analysis of factors contributing to infant death in a high-risk population for SIDS, the American Indians in the Northern Plains (Kinney et al. 2003). At the time of the study (1992–1996), the SIDS rate in this population was 3.46/1000 live births, compared to an overall SIDS rate in the United States of 0.76/1000 live births, and the overall rate in the Indian Health Service of 1.6/1000 (Iyasu et al. 2002). As part of this study, we explored the effect of prenatal exposure to cigarette smoke on medullary 5-HT binding in a dataset of SIDS cases and controls. Multivariate analyses revealed that 5-HT receptor binding was signiἀcantly lower in infants whose mothers smoked during pregnancy (irrespective of the cause of infant death) (see Figure 5.5), and suggest that prenatal smoking exposures adversely affect postnatal 5-HT receptor binding in the medullary 5-HT system. Evidence from studies in experimental animals supports the idea that prenatal exposure to nicotine and cigarette smoke alters 5-HT development and function, including altered 5-HT neuron ἀring, 5-HT receptor expression (Kenny et al. 2001; Slotkin et al. 2006a, 2006b; Slotkin and Ryde 2007; Slotkin and Seidler 2006), 5-HTT expression (Muneoka et al. 2001) Slotkin et al. 2007a, 2007b; Slotkin and Ryde 2007; Xu et al. 2001), 5-HT turnover, and depletion of the brain 5-HT in the postnatal period (King et al. 1991; Muneoka et al. 1997). These adverse effects appear to result from the binding of nicotine to nicotinic and/or 5-HT receptors on 5-HT neurons (Aznar et al. 2005; Bitner and Nikkel 2002; Cucchiaro and Commons 2003). In the developing human medulla, we found that nicotinic receptors are expressed by 5-HT neurons throughout the medullary 5-HT system, including in the raphe obscurus and arcuate nucleus (Duncan et al. 2008a). These observations suggest, therefore, that the increased risk of SIDS associated with maternal smoking during pregnancy involves, at least in part, mechanisms that alter the development and function of the medullary 5-HT system in the infant.

15

No Yes

10

*

5

0

Maternal smoking

Maternal drinking

Figure 5.5╇ Effect of maternal smoking and drinking on infant 5-HT receptor binding. 5-HT

receptor binding is significantly lower in the arcuate nucleus of infants exposed to cigarette smoke prenatally. Graphs comparing the effects of prenatal smoking and prenatal drinking on the effects of 5-HT receptor binding measured with 3H LSD autoradiography in the infant postnatally. Maternal smoking during pregnancy is associated with a 40% reduction (*p = 0.011) and maternal drinking during pregnancy is associated with a 32% reduction (p = 0.075) in 5-HT receptor binding postnatally in the infant. (From Kinney, H. C. et al., J€Neuropathol Exp Neurol, 62 (11), 1178–1191, 2003. With permission.)

84 Sudden Death in Epilepsy: Forensic and Clinical Issues

5.4.4â•…Maternal Drinking In the Aberdeen Area Infant Mortality Study (AAIMS), we also explored the effect of prenatal exposure to alcohol on medullary 5-HT binding. A trend for reduced 5-HT receptor binding was observed in the arcuate nucleus of infants of mothers who drank alcohol compared to nondrinking mothers (see Figure 5.5) (Kinney et al. 2003). In animal models of prenatal alcohol exposure, various abnormalities of the 5-HT system have been observed, including reduced cortical 5-HT levels and 5-HT receptors (Druse and Paul 1988), retarded process outgrowth and migration of 5-HT neurons, reduced density of 5-HT ἀbers in the medial forebrain bundle, and reduced 5-HT neurons in the median and dorsal raphe (Sari et al. 2001; Zhou et al. 2002) and lower brain stem (Druse et al. 2004). Prenatal alcohol also adversely affects signaling molecules and transcription factors necessary for 5-HT development, e.g., sonic hedgehog, which is involved in the early speciἀcation of 5-HT precursors (Ahlgren and Bronner-Fraser 1999; Ahlgren et al. 2002). The delivery of the 5-HT1A agonist buspirone or ipsapirone in maternal rats prevents the alcohol-induced loss of brain stem 5-HT neurons in the pups, underscoring the potential value of targeting 5-HT receptors clinically as a successful therapeutic strategy (Druse et al. 2004; Kim and Druse 1996). Therefore, prenatal exposure to alcohol, particularly in combination with exposure to cigarette smoke, may adversely affect the development and function of the medullary 5-HT system, thereby potentially increasing the risk of the exposed infant to SIDS.

5.5â•…5 -HT Gene Polymorphisms and SIDS Several studies have identiἀed signiἀcant associations between SIDS and gene polymorphisms resulting in alterations in 5-HT neuronal function and development. These include two polymorphisms in the 5-HTT gene that result in differential gene expression. The ἀrst of these is an insertion/deletion polymorphism in the gene promoter region (5-HTTLPR), producing a short “S” and a long “L” allele that consist of 14 copies and 16 copies of the insertion, respectively (Heils et al. 1997; Lesch and Mossner 1998). The L allele and LL genotype are associated with increased gene expression and have been observed more frequently in SIDS cases compared to controls (Narita et al. 2001; Nonnis Marzano et al. 2008; Opdal et al. 2008; Weese-Mayer et al. 2003). The second polymorphism is a variable number tandem repeat (VNTR) polymorphism in the second intron of the 5-HTT gene (Intron 2) that consists of 9, 10, or 12 copies of the repeat element (Ogilvie et al. 1996). Similarly, the 12/12 genotype is associated with increased gene expression and is observed more frequently in SIDS populations (Fiskerstrand et al. 1999; MacKenzie and Quinn 1999; Weese-Mayer et al. 2003). Reuptake activity is thought to be higher in the brains of individuals with the L and/or 12 alleles because there is increased gene expression and efficiency of 5-HT uptake in vitro (Greenberg et al. 1999; Heils et al. 1997; Lesch and Mossner 1998; van Dyck 2004). Thus, 5-HT concentrations at the synapse are likely lower in SIDS infants with the LL and/or the 12/12 genotype, which may exacerbate an existing 5-HT defect. Another study identiἀed an increased frequency of a VNTR polymorphism in the promoter region of the monoamine oxidase A (MAOA) gene in SIDS cases, compared to controls (Nonnis Marzano et al. 2008). MAOA is the major catabolic enzyme for 5-HT degradation and, therefore, plays an important role in the regulation of brain 5-HT levels. The polymorphism consists of a 30-bp repeat sequence present in 3, 3.5, 4, or 5 copies that

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

85

affect gene transcriptional activity, with 3.5 or 4 copies of the repeat sequence associated with increased transcription (Deckert et al. 1999; Sabol et al. 1998). Filonzi et al. (2009) identiἀed an increased frequency of the 4/4 genotype in SIDS cases compared to controls, suggesting increased expression of MAOA and, by extrapolation, reduced levels of 5-HT in SIDS cases, compared to controls. Thus, each of the three polymorphisms described above putatively results in reduced availability of 5-HT in SIDS. A rare mutation (IVS2 191_190insA) upstream of the third exon of the human ἀfth Ewing variant (FEV) gene has also been associated with SIDS (Rand et al. 2007). FEV is the human homologue of the ETS domain transcription factor Pet-1 that is necessary for differentiation and development of 5-HT neurons (Hendricks et al. 1999), including regulation of TPH, 5-HTT, and 5-HT1A receptor gene expression (Hendricks et al. 1999, 2003; Iyo et al. 2005; Maurer et al. 2004; Pfaar et al. 2002). Loss of the Pet-1 gene in mice results in failure of approximately 70% of 5-HT neurons to differentiate (Hendricks et al. 1999) and deἀcient expression of genes required for 5-HT synthesis, uptake, and vesicular storage in the remaining 5-HT neurons (Hendricks et al. 2003). The FEV gene mutation may, therefore, result in or predispose an infant to medullary 5-HT dysfunction and, thus, SIDS.

5.6â•…SIDS as a Developmental Disorder That Originates in Utero Several observations indicate that SIDS is a developmental disorder that originates during fetal life. The incidence of SIDS is greater in preterm and growth-restricted infants; the peak incidence of SIDS is related to a critical and ἀnite early developmental period (2–4 postnatal months), and prenatal exposure to environmental toxins including cigarette smoke (Anderson et al. 2005; Blair et al. 1996; Fleming et al. 1996; Haglund 1993; MacDorman et al. 1997; Mitchell and Milerad 2006; Schoendorf and Kiely 1992; Wisborg et al. 2000) and alcohol (Alm et al. 1999; Duncan et al. 2008b; Filonzi et al. 2009; Iyasu et al. 2002; Kinney et al. 2003; Klug et al. 2003; Scragg et al. 1993) are major risk factors for SIDS. Evidence from postmortem human studies suggests that the development of the medullary 5-HT system is abnormal in SIDS, including an increased number of 5-HT neurons with immature morphology (Paterson et al. 2006), abnormal/immature synapse formation (Paterson et al. 2006), and differential age-related changes in 5-HT receptor binding (i.e., binding decreases signiἀcantly with postnatal age in SIDS cases but not in controls) in the medulla of SIDS cases compared to controls (Kinney et al. 2003; Panigrahy et al. 2000). These observations offer a possible explanation for the low incidence of SIDS during the ἀrst postnatal month, followed by the period of peak incidence at 2–4 months: At birth, 5-HT function is relatively normal, but becomes progressively defective during the ἀrst postnatal month as 5-HT receptor binding decreases; by 2–4 months, 5-HT dysfunction reaches the threshold, whereby the infant is unable to respond appropriately to an environmental stressor (e.g., hypoxia), ultimately leading to the sudden death of the infant.

5.7╅SIDS Results from a Combination of€Environmental and Genetic Factors Fetal programming refers to the concept that prenatal exposure to drugs or other toxic agents can permanently reorganize or reprogram neural circuitry, altering physiological

86 Sudden Death in Epilepsy: Forensic and Clinical Issues

and behavioral systems, and thereby increasing vulnerability to illness or disorders in later life (de Moura et al. 2008; Rinaudo and Lamb 2008; Tremblay and Hamet 2008). Despite the adverse effects of prenatal alcohol and cigarette smoke exposure on fetal medullary 5-HT function and development, it is unlikely that exposure to environmental toxins alone is sufficient to cause SIDS, as only some infants exposed to alcohol and cigarette smoke die while others do not. Similarly, not all infants who sleep prone die of SIDS. Thus, some infants must be inherently more susceptible to SIDS than others, e.g., those with 5-HT gene mutations that render the fetus less resilient to toxic exposures and homeostatic stressors. We propose, therefore, that SIDS is a multifactorial disease that results from the superimposition of environmental toxins upon a predisposing genetic background.

5.8â•…Brain Stem 5-HT Dysfunction and SUDEP Sudden unexplained death in epilepsy (SUDEP) is the sudden unexpected death of an individual with epilepsy, who was otherwise well, and in whom no other cause for death can be found, despite thorough postmortem examination and blood tests (Langan et al. 2005; Nei et al. 2004; Tomson et al. 2005). SUDEP is frequently associated with sleep and, while the cause is unknown, it is proposed to result from mechanisms involving cardiorespiratory and autonomic dysfunction (So et al. 2009). Indeed, patients with epilepsy commonly display acidosis, hypoxia (Swallow et al. 2002), apnea (Blum et al. 2000), bradycardia, asystole (Rugg-Gunn et al. 2004), and arrhythmias (Nei et al. 2004; Zijlmans et al. 2002) after seizure. SUDEP, therefore, shares a number of deἀning characteristics with SIDS, raising the possibility that the diseases share a common pathogenesis, i.e., SUDEP results from an underlying medullary 5-HT abnormality that results in failure of cardiorespiratory function, leading to the death of the individual following seizure. Indeed, 5-HT dysfunction is associated with epilepsy and seizure generation: Plasma levels and the brain uptake rate of the 5-HT precursor tryptophan are lower in epileptics compared to nonepileptics (Albano et al. 2006). Similarly, seizure activity in mice is increased by depletion of 5-HT following the ablation of midbrain 5-HT neurons (Merrill et al. 2007) and by administration of 5-HT receptor antagonists (Tupal and Faingold 2006). In contrast, agents that elevate the levels of available 5-HT, including administration of the 5-HT precursor 5-hydroxy tryptophan (Truscott 1975) and SSRIs such as fluoxetine, have consistently been observed to act as anticonvulsants in both rodents (Leander 1992; Pasini et al. 1992; Prendiville and Gale 1993; Wada et al. 1995) and humans (Albano et al. 2006). Recent observations indicate that 5-HT mechanisms may also play a role in the cardiorespiratory dysfunction proposed to underlie SUDEP. In a recent study, administration of fluoxetine 25 min before seizure stimulus, at a dose that did not reduce seizure severity, signiἀcantly reduced postseizure respiratory arrest in audiogenic seizing mice (Tupal and Faingold 2006). This observation supports the idea that a subtle abnormality in the medullary 5-HT system (as in SIDS) may predispose an individual to respiratory arrest and, thus, potentially SUDEP following seizure, regardless of the seizure-generating mechanism. This idea is also supported by the observation that in tilt stress tests, patients with epilepsy display abnormal heart rate variability (Ansakorpi et al. 2000) similar to that observed in an infant that subsequently died of SIDS and that was observed to have reduced medullary 5-HT receptor binding density at autopsy (Kinney et al. 2005). Taken together, these observations provide potential new avenues of research in SUDEP, including analysis of predisposing risk factors such as

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

87

prenatal exposures, and postmortem analysis of the brain stem to determine the presence of medullary 5-HT abnormalities.

5.9â•…Conclusions The observations discussed in the preceding sections support the idea that multiple abnormalities in the medullary 5-HT system, resulting in failure of homeostatic responses to normal challenges during sleep, play a role in the pathogenesis of SIDS. In addition, evidence suggests that the medullary 5-HT abnormality in SIDS is a progressive developmental disorder that originates in utero and involves exposure of the fetus to environmental toxins in combination with a predisposing genetic background. However, the speciἀc nature of the 5-HT dysfunction in SIDS is still unclear. Is there an excess or a deἀcit of available 5-HT in SIDS? The abnormalities in markers of 5-HT function described above may be interpreted as evidence that either an increased number of 5-HT neurons may lead to an excess of extracellular 5-HT and a compensatory downregulation of 5-HT receptors or, alternatively, 5-HT synthesis and/or release may be dysfunctional in the 5-HT neurons (which are overabundant in compensation), resulting in a deἀciency of extracellular 5-HT. Indeed, both an excess and a deἀcit in 5-HT levels during development and in the postnatal period produce respiratory dysfunction in animal models. Determination of the level of available 5-HT in SIDS medulla is, therefore, critical in determining the speciἀc nature and pathogenesis of 5-HT dysfunction in SIDS. Such studies are currently underway in our laboratory. Given the similarities in the deἀning characteristics of SIDS and SUDEP and the evidence implicating 5-HT dysfunction in seizure genesis and cardiorespiratory failure following seizure, it is reasonable to hypothesize that medullary 5-HT dysfunction also underlies the pathogenesis of SUDEP. The brain stem and other studies described above for SIDS, therefore, suggest a possible direction for SUDEP research and may provide novel insight into the pathogenesis of this disease.

References Ahlgren, S. C., and M. Bronner-Fraser. 1999. Inhibition of sonic hedgehog signaling in vivo results in craniofacial neural crest cell death. Curr Biol 9 (22): 1304–1314. Ahlgren, S. C., V. Thakur, and M. Bronner-Fraser. 2002. Sonic hedgehog rescues cranial neural crest from cell death induced by ethanol exposure. Proc Natl Acad Sci U S A 99 (16): 10476–10481. Albano, C., A. Cupello, P. Mainardi, S. Scarrone, and E. Favale. 2006. Successful treatment of epilepsy with serotonin reuptake inhibitors: Proposed mechanism. Neurochem Res 31 (4): 509–514. Alm, B., G. Wennergren, G. Norvenius, R. Skjaerven, N. Oyen, K. Helweg-Larsen, H. Lagercrantz, and L. M. Irgens. 1999. Caffeine and alcohol as risk factors for sudden infant death syndrome. Nordic Epidemiological SIDS Study. Arch Dis Child 81 (2): 107–111. Anderson, M. E., D. C. Johnson, and H. A. Batal. 2005. Sudden infant death syndrome and prenatal maternal smoking: Rising attributed risk in the back to sleep era. BMC Med 3: 4. Ansakorpi, H., J. T. Korpelainen, K. Suominen, U. Tolonen, V. V. Myllyla, and J. I. Isojarvi. 2000. Interictal cardiovascular autonomic responses in patients with temporal lobe epilepsy. Epilepsia 41 (1): 42–47. Arango, V., M. D. Underwood, A. V. Gubbi, and J. J. Mann. 1995. Localized alterations in pre- and postsynaptic serotonin binding sites in the ventrolateral prefrontal cortex of suicide victims. Brain Res 688 (1–2): 121–133.

88 Sudden Death in Epilepsy: Forensic and Clinical Issues Aznar, S, V. Kostova, S. H. Christiansen, and G. M. Knudsen. 2005. Alpha 7 nicotinic receptor subunit is present on serotonin neurons projecting to hippocampus and septum. Synapse 55 (3): 196–200. Baker-Herman, T. L., D. D. Fuller, R. W. Bavis, A. G. Zabka, F. J. Golder, N. J. Doperalski, R. A. Johnson, J. J. Watters, and G. S. Mitchell. 2004. BDNF is necessary and sufficient for spinal respiratory plasticity following intermittent hypoxia. Nat Neurosci 7 (1): 48–55. Bartlett, D., Jr., J. C. Leiter, and S. L. Knuth. 1990. Control and actions of the genioglossus muscle. Prog Clin Biol Res 345: 99–107; discussion 108. Bayliss, D. A., and D. E. Millhorn. 1992. Central neural mechanisms of progesterone action: Application to the respiratory system. J Appl Physiol 73 (2): 393–404. Bayliss, D. A., J. A. Cidlowski, and D. E. Millhorn. 1990. The stimulation of respiration by progesterone in ovariectomized cat is mediated by an estrogen-dependent hypothalamic mechanism requiring gene expression. Endocrinology 126 (1): 519–527. Bernard, D. G., A. Li, and E. E. Nattie. 1996. Evidence for central chemoreception in the midline raphe. J Appl Physiol 80 (1): 108–115. Berner, N. J., D. A. Grahn, and H. C. Heller. 1999. 8-OH-DPAT-sensitive neurons in the nucleus raphe magnus modulate thermoregulatory output in rats. Brain Res 831 (1–2): 155–164. Bitner, R. S., and A. L. Nikkel. 2002. Alpha-7 nicotinic receptor expression by two distinct cell types in the dorsal raphe nucleus and locus coeruleus of rat. Brain Res 938 (1–2): 45–54. Blair, P. S., P. J. Fleming, D. Bensley, I. Smith, C. Bacon, E. Taylor, J. Berry, J. Golding, and J. Tripp. 1996. Smoking and the sudden infant death syndrome: Results from 1993–5 case-control study for conἀdential inquiry into stillbirths and deaths in infancy. Conἀdential Enquiry into Stillbirths and Deaths Regional Coordinators and Researchers. BMJ 313 (7051): 195–198. Blum, A. S., J. R. Ives, A. L. Goldberger, I. C. Al-Aweel, K. B. Krishnamurthy, F. W. Drislane, and D.€L. Schomer. 2000. Oxygen desaturations triggered by partial seizures: Implications for cardiopulmonary instability in epilepsy. Epilepsia 41 (5): 536–541. Bou-Flores, C., A. M. Lajard, R. Monteau, E. De Maeyer, I. Seif, J. Lanoir, and G. Hilaire. 2000. Abnormal phrenic motoneuron activity and morphology in neonatal monoamine oxidase A–deἀcient transgenic mice: Possible role of a serotonin excess. J Neurosci 20 (12): 4646–4656. Brown, J. W., E. A. Sirlin, A. M. Benoit, J. M. Hoffman, and R. A. Darnall. 2008. Activation of 5-HT1A receptors in medullary raphe disrupts sleep and decreases shivering during cooling in the conscious piglet. Am J Physiol Regul Integr Comp Physiol 294 (3): R884–R894. Cann-Moisan, C., E. Girin, J. D. Giroux, P. Le Bras, and J. Caroff. 1999. Changes in cerebrospinal fluid monoamine metabolites, tryptophan, and gamma-aminobutyric acid during the 1st year of life in normal infants: Comparison with victims of sudden infant death syndrome. Biol Neonate 75 (3): 152–159. Cucchiaro, G., and K. G. Commons. 2003. Alpha 4 nicotinic acetylcholine receptor subunit links cholinergic to brainstem monoaminergic neurotransmission. Synapse 49 (3): 195–205. Darnall, R. A., M. B. Harris, W. H. Gill, J. M. Hoffman, J. W. Brown, and M. M. Niblock. 2005. Inhibition of serotonergic neurons in the nucleus paragigantocellularis lateralis fragments sleep and decreases rapid eye movement sleep in the piglet: Implications for sudden infant death syndrome. J Neurosci 25 (36): 8322–8332. de Moura, E. G., P. C. Lisboa, and M. C. Passos. 2008. Neonatal programming of neuroimmunomodulation—Role of adipocytokines and neuropeptides. Neuroimmunomodulation 15 (3): 176–188. Deckert, J., M. Catalano, Y. V. Syagailo, M. Bosi, O. Okladnova, D. Di Bella, M. M. Nothen et al. 1999. Excess of high activity monoamine oxidase A gene promoter alleles in female patients with panic disorder. Hum Mol Genet 8 (4): 621–624. Dillon, K. A., R. Gross-Isseroff, M. Israeli, and A. Biegon. 1991. Autoradiographic analysis of serotonin 5-HT1A receptor binding in the human brain postmortem: Effects of age and alcohol. Brain Res 554 (1–2): 56–64. Dreshaj, I. A., M. A. Haxhiu, and R. J. Martin. 1998. Role of the medullary raphe nuclei in the respiratory response to CO2. Respir Physiol 111 (1): 15–23.

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

89

Druse, M. J., and L. H. Paul. 1988. Effects of in utero ethanol exposure on serotonin uptake in cortical regions. Alcohol 5 (6): 455–459. Druse, M. J., N. F. Tajuddin, R. A. Gillespie, E. Dickson, M. Atieh, C. A. Pietrzak, and P. T. Le. 2004. The serotonin-1A agonist ipsapirone prevents ethanol-associated death of total rhombenceÂ� phalic neurons and prevents the reduction of fetal serotonin neurons. Brain Res Dev Brain Res 150 (2): 79–88. Duncan, J. R., D. S. Paterson, and H. C. Kinney. 2008a. The development of nicotinic receptors in the human medulla oblongata: Inter-relationship with the serotonergic system. Auton Neurosci 144 (1–2): 61–75. Duncan, J. R., L. L. Randall, R. A. Belliveau, F. L. Trachtenberg, B. Randall, D. Habbe, F. Mandell, T. K. Welty, S. Iyasu, and H. C. Kinney. 2008b. The effect of maternal smoking and drinking during pregnancy upon (3)H-nicotine receptor brainstem binding in infants dying of the sudden infant death syndrome: Initial observations in a high risk population. Brain Pathol 18 (1): 21–31. Emery, M. J., M. P. Hlastala, and A. M. Matsumoto. 1994. Depression of hypercapnic ventilatory drive by testosterone in the sleeping infant primate. J Appl Physiol 76 (4): 1786–1793. Emery, M. J., H. F. Krous, J. M. Nadeau-Manning, B. T. Marck, and A. M. Matsumoto. 2005. Serum testosterone and estradiol in sudden infant death. J Pediatr 147: 586–591. Ferrari, P. F., S. Lowther, H. Tidbury, P. Greengrass, C. A. Wilson, and R. W. Horton. 1999. The influence of gender and age on neonatal rat hypothalamic 5-HT1A and 5-HT2A receptors. Cell Mol Neurobiol 19 (6): 775–784. Filiano, J. J., and H. C. Kinney. 1994. A perspective on neuropathologic ἀndings in victims of the sudden infant death syndrome: The triple-risk model. Biol Neonate 65 (3–4): 194–197. Filonzi, L., C. Magnani, A. M. Lavezzi, G. Rindi, S. Parmigiani, G. Bevilacqua, L. Matturri, and F. N. Marzano. 2009. Association of dopamine transporter and monoamine oxidase molecular polymorphisms with sudden infant death syndrome and stillbirth: New insights into the serotonin hypothesis. Neurogenetics 10 (1): 65–72. Fiskerstrand, C. E., E. A. Lovejoy, and J. P. Quinn. 1999. An intronic polymorphic domain often associated with susceptibility to affective disorders has allele dependent differential enhancer activity in embryonic stem cells. FEBS Lett 458 (2): 171–174. Fleming, P. J., P. S. Blair, C. Bacon, D. Bensley, I. Smith, E. Taylor, J. Berry, J. Golding, and J. Tripp. 1996. Environment of infants during sleep and risk of the sudden infant death syndrome: Results of 1993–5 case-control study for conἀdential inquiry into stillbirths and deaths in infancy. Conἀdential Enquiry into Stillbirths and Deaths Regional Coordina. BMJ 313 (7051): 191–195. Fogel, R. B., A. Malhotra, G. Pillar, J. K. Edwards, J. Beauregard, S. A. Shea, and D. P. White. 2001. Genioglossal activation in patients with obstructive sleep apnea versus control subjects. Mechanisms of muscle control. Am J Respir Crit Care Med 164 (11): 2025–2030. Forest, M. G., E. de Peretti, and J. Bertrand. 1980. Testicular and adrenal androgens and their binding to plasma proteins in the perinatal period: Developmental patterns of plasma testosterone, 4-androstenedione, dehydroepiandrosterone and its sulfate in premature and small for date infants as compare. J Steroid Biochem 12: 25–36. Greenberg, B. D., T. J. Tolliver, S. J. Huang, Q. Li, D. Bengel, and D. L. Murphy. 1999. Genetic variation in the serotonin transporter promoter region affects serotonin uptake in human blood platelets. Am J Med Genet 88 (1): 83–87. Haglund, B. 1993. Cigarette smoking and sudden infant death syndrome: Some salient points in the debate. Acta Paediatr Suppl 82 (Suppl 389): 37–39. Heils, A., R. Mossner, and K. P. Lesch. 1997. The human serotonin transporter gene polymorphism— Basic research and clinical implications. J Neural Transm 104 (10): 1005–1014. Henderson, L. A., K. A. Keay, and R. Bandler. 2000. Caudal midline medulla mediates behaviourallycoupled but not baroreceptor-mediated vasodepression. Neuroscience 98 (4): 779–792. Hendricks, T., N. Francis, D. Fyodorov, and E. S. Deneris. 1999. The ETS domain factor Pet-1 is an early and precise marker of central serotonin neurons and interacts with a conserved element in serotonergic genes. J Neurosci 19 (23): 10348–10356.

90 Sudden Death in Epilepsy: Forensic and Clinical Issues Hendricks, T. J., D. V. Fyodorov, L. J. Wegman, N. B. Lelutiu, E. A. Pehek, B. Yamamoto, J. Silver, E. J. Weeber, J. D. Sweatt, and E. S. Deneris. 2003. Pet-1 ETS gene plays a critical role in 5-HT neuron development and is required for normal anxiety-like and aggressive behavior. Neuron 37 (2): 233–247. Hilaire, G., D. Morin, A. M. Lajard, and R. Monteau. 1993. Changes in serotonin metabolism may elicit obstructive apnoea in the newborn rat. J Physiol 466: 367–381. Hodges, M. R., G. J. Tattersall, M. B. Harris, S. D. McEvoy, D. N. Richerson, E. S. Deneris, R. L. Johnson, Z. F. Chen, and G. B. Richerson. 2008. Defects in breathing and thermoregulation in mice with near-complete absence of central serotonin neurons. J Neurosci 28 (10): 2495–2505. Hoffman, J. M., J. W. Brown, E. A. Sirlin, A. M. Benoit, W. H. Gill, M. B. Harris, and R. A. Darnall. 2007. Activation of 5-HT1A receptors in the paragigantocellularis lateralis decreases shivering during cooling in the conscious piglet. Am J Physiol Regul Integr Comp Physiol 293 (1): R518–R527. Holtman, J. R., Jr., T. E. Dick, and A. J. Berger. 1987. Serotonin-mediated excitation of recurrent laryngeal and phrenic motoneurons evoked by stimulation of the raphe obscurus. Brain Res 417 (1): 12–20. Hornung, J. P. 2003. The human raphe nuclei and the serotonergic system. J Chem Neuroanat 26 (4): 331–343. Iyasu, S., L. L. Randall, T. K. Welty, J. Hsia, H. C. Kinney, F. Mandell, M. McClain, B. Randall, D. Habbe, H. Wilson, and M. Willinger. 2002. Risk factors for sudden infant death syndrome among northern plains Indians. JAMA 288 (21): 2717–2723. Iyo, A. H., B. Porter, E. S. Deneris, and M. C. Austin. 2005. Regional distribution and cellular localization of the ETS-domain transcription factor, FEV, mRNA in the human postmortem brain. Synapse 57 (4): 223–228. Johns, J. M., D. A. Lubin, J. A. Lieberman, and J. M. Lauder. 2002. Developmental effects of prenatal cocaine exposure on 5-HT1A receptors in male and female rat offspring. Dev Neurosci 24 (6): 522–530. Kenny, P. J., S. E. File, and M. Rattray. 2001. Nicotine regulates 5-HT(1A) receptor gene expression in the cerebral cortex and dorsal hippocampus. Eur J Neurosci 13 (6): 1267–1271. Kim, J. A., and M. J. Druse. 1996. Protective effects of maternal buspirone treatment on serotonin reuptake sites in ethanol-exposed offspring. Brain Res Dev Brain Res 92 (2): 190–198. King, J. A., M. Davila-Garcia, E. C. Azmitia, and F. L. Strand. 1991. Differential effects of prenatal and postnatal ACTH or nicotine exposure on 5-HT high affinity uptake in the neonatal rat brain. Int J Dev Neurosci 9 (3): 281–286. Kinney, H. C., L. L. Randall, L. A. Sleeper, M. Willinger, R. A. Belliveau, N. Zec, L. A. Rava et al. 2003. Serotonergic brainstem abnormalities in Northern Plains Indians with the sudden infant death syndrome. J Neuropathol Exp Neurol 62 (11): 1178–1191. Kinney, H. C., M. M. Myers, R. A. Belliveau, L. L. Randall, F. L. Trachtenberg, S. T. Fingers, M. Youngman, D. Habbe, and W. P. Fifer. 2005. Subtle autonomic and respiratory dysfunction in sudden infant death syndrome associated with serotonergic brainstem abnormalities: A case report. J Neuropathol Exp Neurol 64 (8): 689–694. Kinney, H. C., R. A. Belliveau, F. L. Trachtenberg, L. A. Rava, and D. S. Paterson. 2007. The development of the medullary serotonergic system in early human life. Auton Neurosci 132 (1–2): 81–102. Kinney, H. C., and D. S. Paterson. 2004. Sudden infant death syndrome. In Pathololgy and Genetics: Developmental Neuropathology, ed. J. Golden and B. Harding. Basel: Neuropath Press. Klug, M. G., L. Burd, J. Kerbeshian, B. Benz, and J. T. Martsolf. 2003. A comparison of the effects of parental risk markers on pre- and perinatal variables in multiple patient cohorts with fetal alcohol syndrome, autism, Tourette syndrome, and sudden infant death syndrome: An enviromic analysis. Neurotoxicol Teratol 25 (6): 707–717. Krammer, E. B., T. Rath, and M. F. Lischka. 1979. Somatotopic organization of the hypoglossal nucleus: A HRP study in the rat. Brain Res 170 (3): 533–537.

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

91

Krous, H. F., J. B. Beckwith, R. W. Byard, T. O. Rognum, T. Bajanowski, T. Corey, E. Cutz, R. Hanzlick, T. G. Keens, and E. A. Mitchell. 2004. Sudden infant death syndrome and unclassiἀed sudden infant deaths: A deἀnitional and diagnostic approach. Pediatrics 114 (1): 234–238. Kubin, L., H. Tojima, R. O. Davies, and A. I. Pack. 1992. Serotonergic excitatory drive to hypoglossal motoneurons in the decerebrate cat. Neurosci Lett 139 (2): 243–248. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64 (7): 1131–1133. Leander, J. D. 1992. Fluoxetine, a selective serotonin-uptake inhibitor, enhances the anticonvulsant effects of phenytoin, carbamazepine, and ameltolide (LY201116). Epilepsia 33 (3): 573–576. Lesch, K. P., and R. Mossner. 1998. Genetically driven variation in serotonin uptake: Is there a link to affective spectrum, neurodevelopmental, and neurodegenerative disorders? Biol Psychiatry 44 (3): 179–192. Li, A., and E. Nattie. 2008. Serotonin transporter knockout mice have a reduced ventilatory response to hypercapnia (predominantly in males) but not to hypoxia. J Physiol 586 (9): 2321–2329. Ling, L., D. D. Fuller, K. B. Bach, R. Kinkead, E. B. Olson Jr., and G. S. Mitchell. 2001. Chronic intermittent hypoxia elicits serotonin-dependent plasticity in the central neural control of breathing. J Neurosci 21 (14): 5381–5388. Liu, P. Y., B. Yee, S. M. Wishart, M. Jimenez, D. G. Jung, R. R. Grunstein, and D. J. Handelsman. 2003. The short-term effects of high-dose testosterone on sleep, breathing, and function in older men. J Clin Endocrinol Metab 88 (8): 3605–3613. MacDorman, M. F., S. Cnattingius, H. J. Hoffman, M. S. Kramer, and B. Haglund. 1997. Sudden infant death syndrome and smoking in the United States and Sweden. Am J Epidemiol 146 (3): 249–257. Machaalani, R., M. Say, and K. A. Waters. 2009. Serotoninergic receptor 1A in the sudden infant death syndrome brainstem medulla and associations with clinical risk factors. Acta Neuropathol 117 (3): 257–265. MacKenzie, A., and J. Quinn. 1999. A serotonin transporter gene intron 2 polymorphic region, correlated with affective disorders, has allele-dependent differential enhancer-like properties in the mouse embryo. Proc Natl Acad Sci U S A 96 (26): 15251–15255. Matsumoto, A. M., R. E. Sandblom, R. B. Schoene, K. A. Lee, E. C. Giblin, D. J. Pierson, and W. J. Bremner. 1985. Testosterone replacement in hypogonadal men: Effects on obstructive sleep apnoea, respiratory drives, and sleep. Clin Endocrinol (Oxf) 22 (6): 713–721. Maurer, P., S. Rorive, d’Exaerde A. de Kerchove, S. N. Schiffmann, I. Salmon, and Y. de Launoit. 2004. The Ets transcription factor Fev is speciἀcally expressed in the human central serotonergic neurons. Neurosci Lett 357 (3): 215–218. Merrill, M. A., R. W. Clough, J. W. Dailey, P. C. Jobe, and R. A. Browning. 2007. Localization of the serotonergic terminal ἀelds modulating seizures in the genetically epilepsy-prone rat. Epilepsy Res 76 (2–3): 93–102. Messier, M. L., A. Li, and E. E. Nattie. 2004. Inhibition of medullary raphe serotonergic neurons has age-dependent effects on the CO2 response in newborn piglets. J Appl Physiol 96 (5): 1909–1919. Mitchell, E. A., and J. Milerad. 2006. Smoking and the sudden infant death syndrome. Rev Environ Health 21 (2): 81–103. Moon, R. Y., R. S. Horne, and F. R. Hauck. 2007. Sudden infant death syndrome. Lancet 370 (9598): 1578–1587. Muneoka, K., T. Ogawa, K. Kamei, S. Muraoka, R. Tomiyoshi, Y. Mimura, H. Kato, M. R. Suzuki, and M. Takigawa. 1997. Prenatal nicotine exposure affects the development of the central serotoÂ� nergic system as well as the dopaminergic system in rat offspring: Involvement of route of drug administrations. Brain Res Dev Brain Res 102 (1): 117–126. Muneoka, K., T. Ogawa, K. Kamei, Y. Mimura, H. Kato, and M. Takigawa. 2001. Nicotine exposure during pregnancy is a factor which influences serotonin transporter density in the rat brain. Eur J Pharmacol 411 (3): 279–282.

92 Sudden Death in Epilepsy: Forensic and Clinical Issues Narita, N., M. Narita, S. Takashima, M. Nakayama, T. Nagai, and N. Okado. 2001. Serotonin transporter gene variation is a risk factor for sudden infant death syndrome in the Japanese population. Pediatrics 107 (4): 690–692. Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. Nonnis Marzano, F., M. Maldini, L. Filonzi, A. M. Lavezzi, S. Parmigiani, C. Magnani, G. Bevilacqua, and L. Matturri. 2008. Genes regulating the serotonin metabolic pathway in the brain stem and their role in the etiopathogenesis of the sudden infant death syndrome. Genomics 91 (6): 485–491. Ogilvie, A. D., S. Battersby, V. J. Bubb, G. Fink, A. J. Harmar, G. M. Goodwim, and C. A. Smith. 1996. Polymorphism in serotonin transporter gene associated with susceptibility to major depression. Lancet 347 (9003): 731–733. Opdal, S. H., A. Vege, and T. O. Rognum. 2008. Serotonin transporter gene variation in sudden infant death syndrome. Acta Paediatr 97 (7): 861–865. Ozawa, Y., and N. Okado. 2002. Alteration of serotonergic receptors in the brain stems of human patients with respiratory disorders. Neuropediatrics 33 (3): 142–149. Ozawa, Y., and S. Takashima. 2002. Developmental neurotransmitter pathology in the brainstem of sudden infant death syndrome: A review and sleep position. Forensic Sci Int 130 (Suppl): S53–S59. Panigrahy, A., J. Filiano, L. A. Sleeper, F. Mandell, M. Valdes-Dapena, H. F. Krous, L. A. Rava, E. Foley, W. F. White, and H. C. Kinney. 2000. Decreased serotonergic receptor binding in rhombic lipderived regions of the medulla oblongata in the sudden infant death syndrome. J Neuropathol Exp Neurol 59 (5): 377–384. Parsey, R. V., M. A. Oquendo, N. R. Simpson, R. T. Ogden, R. Van Heertum, V. Arango, and J. J. Mann. 2002. Effects of sex, age, and aggressive traits in man on brain serotonin 5-HT1A receptor binding potential measured by PET using [C-11]WAY-100635. Brain Res 954 (2): 173–182. Pasini, A., A. Tortorella, and K. Gale. 1992. Anticonvulsant effect of intranigral fluoxetine. Brain Res 593 (2): 287–290. Paterson, D. S., F. L. Trachtenberg, E. G. Thompson, R. A. Belliveau, A. H. Beggs, R. Darnall, A. E. Chadwick, H. F. Krous, and H. C. Kinney. 2006. Multiple serotonergic brainstem abnormalities in sudden infant death syndrome. JAMA 296 (17): 2124–2132. Pena, F., and J. M. Ramirez. 2002. Endogenous activation of serotonin-2A receptors is required for respiratory rhythm generation in vitro. J Neurosci 22 (24): 11055–11064. Penatti, E. M., A. V. Berniker, B. Kereshi, C. Cafaro, M. L. Kelly, M. M. Niblock, H. G. Gao, H. C. Kinney, A. Li, and E. E. Nattie. 2006. Ventilatory response to hypercapnia and hypoxia after extensive lesion of medullary serotonergic neurons in newborn conscious piglets. J Appl Physiol 101 (4): 1177–1188. Peterson, B. A., and B. J. Kennedy. 1979. Aging and cancer management: Part I. Clinical observations. CA Cancer J Clin 29 (6): 322–332. Pfaar, H., A. von Holst, D. M. Vogt Weisenhorn, C. Brodski, J. Guimera, and W. Wurst. 2002. mPet-1, a mouse ETS-domain transcription factor, is expressed in central serotonergic neurons. Dev Genes Evol 212 (1): 43–46. Pickett, C. K., J. G. Regensteiner, W. D. Woodard, D. D. Hagerman, J. V. Weil, and L. G. Moore. 1989. Progestin and estrogen reduce sleep-disordered breathing in postmenopausal women. J Appl Physiol 66 (4): 1656–1661. Prendiville, S., and K. Gale. 1993. Anticonvulsant effect of fluoxetine on focally evoked limbic motor seizures in rats. Epilepsia 34 (2): 381–384. Rand, C. M., E. M. Berry-Kravis, L. Zhou, W. Fan, and D. E. Weese-Mayer. 2007. Sudden infant death syndrome: Rare mutation in the serotonin system FEV gene. Pediatr Res 62 (2): 180–182. Randall, L. L., C. Krogh, T. K. Welty, M. Willinger, and S. Iyasu. 2001. The Aberdeen Indian Health Service Infant Mortality Study: Design, methodology, and implementation. Am Ind Alaska Native Ment Hlth Res: J Natl Cent 10: 1–20.

Medullary Serotonergic Abnormalities in Sudden Infant Death Syndrome

93

Regensteiner, J. G., R. G. McCullough, R. E. McCullough, C. K. Pickett, and L. G. Moore. 1990. Combined effects of female hormones and exercise on hypoxic ventilatory response. Respir Physiol 82 (1): 107–114. Richerson, G. B. 1995. Response to CO2 of neurons in the rostral ventral medulla in vitro. J Neurophysiol 73 (3): 933–944. Richerson, G. B. 1997. Sudden infant death syndrome: The role of central chemosensitivity. Neuroscientist 3: 3–7. Richerson, G. B. 2004. Serotonergic neurons as carbon dioxide sensors that maintain pH homeostasis. Nat Rev Neurosci 5 (6): 449–461. Rinaudo, P. F., and J. Lamb. 2008. Fetal origins of perinatal morbidity and/or adult disease. Semin Reprod Med 26 (5): 436–445. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Sabol, S. Z., S. Hu, and D. Hamer. 1998. A functional polymorphism in the monoamine oxidase A gene promoter. Hum Genet 103 (3): 273–279. Sari, Y., T. Powrozek, and F. C. Zhou. 2001. Alcohol deters the outgrowth of serotonergic neurons at midgestation. J Biomed Sci 8 (1): 119–125. Sawaguchi, T., F. Patricia, H. Kadhim, J. Groswasser, M. Sottiaux, H. Nishida, and A. Kahn. 2003. The correlation between serotonergic neurons in the brainstem and sleep apnea in SIDS victims. Early Hum Dev 75 (Suppl): S31–S40. Schoendorf, K. C., and J. L. Kiely. 1992. Relationship of sudden infant death syndrome to maternal smoking during and after pregnancy. Pediatrics 90 (6): 905–908. Scragg, R., E. A. Mitchell, B. J. Taylor, A. W. Stewart, R. P. Ford, J. M. Thompson, E. M. Allen, and D. M. Becroft. 1993. Bed sharing, smoking, and alcohol in the sudden infant death syndrome. New Zealand Cot Death Study Group. BMJ 307 (6915): 1312–1318. Slotkin, T. A., K. E. Pinkerton, and F. J. Seidler. 2006a. Perinatal environmental tobacco smoke exposure in rhesus monkeys: Critical periods and regional selectivity for effects on brain cell development and lipid peroxidation. Environ Health Perspect 114 (1): 34–39. Slotkin, T. A., C. A. Tate, M. M. Cousins, and F. J. Seidler. 2006b. Prenatal nicotine exposure alters the responses to subsequent nicotine administration and withdrawal in adolescence: Serotonin receptors and cell signaling. Neuropsychopharmacology 31 (11): 2462–2475. Slotkin, T. A., E. A. MacKillop, C. L. Rudder, I. T. Ryde, C. A. Tate, and F. J. Seidler. 2007a. Permanent,€ sex-selective effects of prenatal or adolescent nicotine exposure, separately or sequentially, in rat€brain regions: Indices of cholinergic and serotonergic synaptic function, cell signaling, and neural cell number and size at 6 months of. Neuropsychopharmacology 32 (5): 1082–1097. Slotkin, T. A., I. T. Ryde, C. A. Tate, and F. J. Seidler. 2007b. Lasting effects of nicotine treatment and withdrawal on serotonergic systems and cell signaling in rat brain regions: Separate or sequential exposure during fetal development and adulthood. Brain Res Bull 73 (4–6): 259–272. Slotkin, T. A., and F. J. Seidler. 2006. Cholinergic receptor subtypes in the olfactory bulbectomy model of depression. Brain Res Bull 68 (5): 341–345. So, E. L., J. Bainbridge, J. R. Buchhalter, J. Donalty, E. J. Donner, A. Finucane, N. M. Graves et al. 2009. Report of the American Epilepsy Society and the Epilepsy Foundation joint task force on sudden unexplained death in epilepsy. Epilepsia 50 (4): 917–922. Sparks, D. L., and J. C. Hunsaker 3rd. 1991. Sudden infant death syndrome: Altered aminergic– cholinergic synaptic markers in hypothalamus. J Child Neurol 6 (4): 335–339. Swallow, R. A., C. E. Hillier, and P. E. Smith. 2002. Sudden unexplained death in epilepsy (SUDEP) following previous seizure-related pulmonary oedema: Case report and review of possible preventative treatment. Seizure 11 (7): 446–448. Task Force on Sudden Infant Death Syndrome. 2005. The changing concept of sudden infant death syndrome: Diagnostic coding shifts, controversies regarding the sleeping environment, and new variables to consider in reducing risk. Pediatrics 116 (5): 1245–1255.

94 Sudden Death in Epilepsy: Forensic and Clinical Issues Tapanainen, J., P. Kellokumpu-Lehtinen, L. Pelliniemi, and I. Huhtaniemi. 1981. Age-related changes in endogenous steroids of human fetal testis during early and midpregnancy. J Clin Endocrinol Metab 52 (1): 98–102. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61. Tork, I, and J. P. Hornung. 1990. Raphe nuclei and the serotonergic system. In The Human Nervous System, ed. G. Paxinos. San Diego, CA: Academic Press. Tremblay, J., and P. Hamet. 2008. Impact of genetic and epigenetic factors from early life to later disease. Metabolism 57 (Suppl 2): S27–S31. Truscott, T. C. 1975. Effects of phenylalanine and 5-hydroxytryptophan on seizure severity in mice. Pharmacol Biochem Behav 3 (5): 939–941. Tryba, A. K., F. Pena, and J. M. Ramirez. 2006. Gasping activity in vitro: A rhythm dependent on 5-HT2A receptors. J Neurosci 26 (10): 2623–2634. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. van Dyck, C. H. 2004. Understanding the latest advances in pharmacologic interventions for Alzheimer’s disease. CNS Spectr 9 (7 Suppl 5): 24–28. Wada, Y., J. Shiraishi, M. Nakamura, and H. Hasegawa. 1995. Prolonged but not acute fluoxetine adÂ�ministration produces its inhibitory effect on hippocampal seizures in rats. PsychopharmaÂ� cology (Berl) 118 (3): 305–309. Wang, W., J. H. Pizzonia, and G. B. Richerson. 1998. Chemosensitivity of rat medullary raphe neurones in primary tissue culture. J Physiol 511 (Pt 2): 433–450. Weese-Mayer, D. E., L. Zhou, E. M. Berry-Kravis, B. S. Maher, J. M. Silvestri, and M. L. Marazita. 2003. Association of the serotonin transporter gene with sudden infant death syndrome: A haplotype analysis. Am J Med Genet A 122A (3): 238–245. Winter, J. S., I. A. Hughes, F. I. Reyes, and C. Faiman. 1976. Pituitary–gonadal relations in infancy: 2.€Patterns of serum gonadal steroid concentrations in man from birth to two years of age. J Clin Endocrinol Metab 42 (4): 679–686. Wisborg, K., U. Kesmodel, T. B. Henriksen, S. F. Olsen, and N. J. Secher. 2000. A prospective study of smoking during pregnancy and SIDS. Arch Dis Child 83 (3): 203–206. Xu, Z., F. J. Seidler, S. F. Ali, W. Slikker Jr., and T. A. Slotkin. 2001. Fetal and adolescent nicotine administration: Effects on CNS serotonergic systems. Brain Res 914 (1–2): 166–178. Zhou, F. C., Y. Sari, T. K. Li, C. Goodlett, and E. C. Azmitia. 2002. Deviations in brain early serotoÂ� nergic development as a result of fetal alcohol exposure. Neurotox Res 4 (4): 337–342. Zhou, X. S., J. A. Rowley, F. Demirovic, M. P. Diamond, and M. S. Badr. 2003. Effect of testosterone on the apneic threshold in women during NREM sleep. J Appl Physiol 94 (1): 101–107. Zijlmans, M., D. Flanagan, and J. Gotman. 2002. Heart rate changes and ECG abnormalities during epileptic seizures: Prevalence and deἀnition of an objective clinical sign. Epilepsia 43 (8): 847–854.

Forensic Case Identification Paul L. Schraeder Elson L. So Claire M. Lathers

6

Contents 6.1 Postmortem Diagnosis in SUDEP 6.1.2 Strengths of Verbal Autopsy 6.1.3 Weaknesses of Verbal Autopsy References

95 103 103 105

6.1â•…Postmortem Diagnosis in SUDEP Sudden unexpected/unexplained death in epilepsy (SUDEP) is the death of a person with epilepsy that is sudden and unexpected, with no obvious circumstances other than a history of a seizure disorder, without the premorbid occurrence of status epilepticus, and with no cause of death being found on postmortem structural and toxicological examination (Nashef 1997; Leestma et al. 1997). Until recent years, the existence of SUDEP as a risk of having epilepsy and as a diagnosis of cause of death in persons with epilepsy has been a topic avoided at several levels, including neurologists, forensic pathologists, and lay support groups. What makes SUDEP a particularly egregious and tragic entity is that the average victim is a young adult in the prime of life, with a mean age in the early 30s. There is resistance by medical professionals to having it as a topic in information for patients and families. Several years ago, one of the authors (PLS) proposed having an informational session for a lay organization that would address the issue of SUDEP, only to be told by another epilepsy specialist that the topic was too controversial and potentially stressful for the patients and their families. It took several years before requests from families and patients resulted in impetus for this topic to be broached at an educational forum for the public. Since the sky did not fall consequent to open discussion of this important risk associated with having epilepsy, and since lay audience are receptive and even appreciative, there has been, in general, an increasing willingness to discuss this topic. At the other end of the spectrum, SUDEP as a ἀnal diagnosis of cause of death is usually not used on death certiἀcates in the circumstance of persons with epilepsy who die with no cause being found, other than having the diagnosis of a seizure disorder (Schraeder et al. 2006). Most epidemiological studies conἀrm that the overall mortality rate of persons with epilepsy is in general two to three times higher than that of the general population (Morgan and Kerr 2002), and the general spectrum of causes of deaths in this population is not much different from that of the general population. These causes include such entities as congenital neurological disorders, alcohol-related deaths, malignancies, cardiovascular disease, and cerebrovascular disorders. However, when the issue of nonaccidental 95

96 Sudden Death in Epilepsy: Forensic and Clinical Issues

unexpected death is examined, persons with epilepsy have up to a 24-fold greater risk of occurrence compared to the general population (Ficker et al. 1998). The issue of accuracy of the national data on deaths in persons with epilepsy seems to be a relatively widespread problem. For example, Tsai (2005) found in Taiwan that data derived from national vital statistics are inadequate to evaluate mortality in persons with epilepsy. This study compared the mortality ratio for epilepsy based on the national vital statistics for 1 year with the total population of Taiwan as the denominator. In Taiwan, almost every death is documented by a death certiἀcate. Using the official statistics, a crude death rate of 561 per 100,000 persons was found for the general population and 756 per 100,000 for the population with epilepsy. However, when the author applied the predicted death rate established for epilepsy from the Warsaw population study (Zielinski 1974), he speculated that persons with epilepsy were likely to have a 10-fold higher risk if the history of having epilepsy had been routinely reported on the death certiἀcates. The mortality data in a cohort of epilepsy cases from a university outpatient clinic was determined for 1224 persons with epilepsy who were followed for a total of 5704 patient years; 54 deaths occurred with a mortality rate of 0.9%. The diagnosis of SUDEP was listed as a cause of death in 2% of the vital statistics group and in 11% of the university clinic population. While this study seems to provide support for the idea that vital statistics data are inadequate to use in assessing the causes of mortality in persons with epilepsy, including SUDEP, the study population in a university epilepsy clinic is by nature biased in favor of patients with more severe epilepsy, and may be over-representative of a group at higher risk of dying. A review of studies conducted in developed countries (Forsgren et al. 2005) shows that the mortality rates for persons with epilepsy vary from country to country. There are only a few population-based studies of mortality in epilepsy. Standard mortality ratios range from 1.8 in Poland to 4.1 in France, with the United States, United Kingdom, Iceland, and Sweden in between. However, the standard mortality ratios vary within each study when comparing patients with different categories of etiology for their epilepsy. In all studies cited, persons with a severe neurological deἀcit had the highest standard mortality ratios (range, 7–50), whereas those with idiopathic epilepsy had the lowest standard mortality ratios (range, 1.1–1.8). Persons with remote symptomatic epilepsy had standard mortality ratios that were intermediate (range, 2.1–6.5). The selection of the population studied is a factor in explaining the standard mortality ratios’ variability. For example, selected epilepsy populations such as university epilepsy centers or epilepsy residential care facilities will have a higher standard mortality ratio than that of the general epilepsy population. The deaths in persons with epilepsy can be due to causes unrelated to epilepsy (e.g., infections, non-central-nervous-system neoplasm, stroke, or progressive degenerative disorders), related to the underlying cause of the epilepsy (e.g., brain tumor, cerebral vascular malformation), or related to a seizure episode (e.g., status epilepticus, drowning or other accidents caused by a seizure, and SUDEP). The rates of occurrence of SUDEP also differ based on the population studied, varying from 0.35 per 1000 person years in an unselectedÂ�population-based study, to 1.5–6 per 1000 person years in a less unselected population, to 1 per 100 person years in those referred for epilepsy surgery or in those with continued postsurgical seizures. Overall, the rate of SUDEP was 24 times greater than sudden death in the general population, making it the most common seizure-related cause of death in adolescents and young adults (Ficker et al. 1998). Data showing a marked increase in the occurrence of SUDEP so far are based on studies of localized populations (Leestma 1990; Earnest et al. 1992). Since SUDEP is

Forensic Case Identification

97

relatively infrequent compared to the more common causes of death, and since there is an acknowledged reluctance by coroners and medical examiners to utilize the term SUDEP on the death certiἀcate diagnosis in appropriate cases (Schraeder et al. 2006), the limited population-Â�based studies can give only an approximation of the extent of occurrence in persons with epilepsy. The use of death certiἀcate data to determine incidence or prevalence data on cause of death related to epilepsy is fraught with inaccuracy, since causes of such deaths are unreliable because both false positive and false negative information are common errors. Any national prevalence data based on death certiἀcate diagnoses, in all likelihood, are not reflective of the actual number of cases of SUDEP, even if adequately autopsied by coroners and medical examiners. Thus, if there is underutilization of the ἀnal diagnosis of SUDEP in thoroughly autopsied cases, one can only speculate about what the actual overall national prevalence of SUDEP is in cases of deaths in persons with epilepsy. Complicating the ability to determine even a reasonable approximation of the national prevalence of SUDEP is the fact that a diagnosis of deἀnite SUDEP can be made only in those victims who undergo a thorough autopsy and in which there was no other cause of death found, whereas a diagnosis of probable SUDEP is used in those that did not have an autopsy, but in which no other cause is extant from the evidence available. By deἀnition, then, a diagnosis of deἀnite SUDEP can only be made on the basis of an autopsy, and when an autopsy is not performed in the case of a sudden death of a person with epilepsy, in whom no other cause of death is evident (e.g., drowning, trauma, history of other underlying disorders such as cardiac disease, drug overdose, etc.), the diagnosis of probable SUDEP is appropriate. Ideally, the nonautopsy data should include a thorough gathering of information consisting of a description of the circumstances at time of death, review of the medical records, and interviews with family members in order to be conἀdent that a diagnosis of probable SUDEP is correct. Unless the coroners and medical examiners have sufficient resources to gather such premorbid information, conἀdence in a diagnosis of probable SUDEP is not adequate, leaving the diagnosis of possible SUDEP to be used in those cases without autopsy in which there may be another possible cause of death and in which historical information is incomplete (Leestma et al. 1997). Several methodological issues are inherent in the study of mortality in epilepsy (Logroscino and Hesdorffer 2005). Included in these issues is retrospective versus prospective design. The cohort can be undertaken either way with mortality ratios and case fatality rates as measures, along with standardized mortality ratios. Case control design can determine a proportionate mortality ratio. However, in any of these study designs, one must always be aware of the difference between incidence rate and prevalence rate. The former is the number of new cases of a disease in a population over a speciἀed period divided by the number of persons at risk for the disease during that period. The latter is the number of existing cases of the disease in a total population divided by the number of persons in that population. In entities with varying degrees of symptomatic expression and long survival, as in a chronic disorder such as epilepsy, prevalent cases would be predominantly survivors with the disorder. The differentiation between incidence and prevalence studies are important because they tend to measure different components of the same populations. For example, a prevalence cohort overestimate of mortality may occur if it includes persons with epilepsy who have not experienced recent seizures and are not included in the incidence cohort or may be due to unaccounted prevalent and incident cases. In any event, no study design is without flaws and in part explains the wide variability in results

98 Sudden Death in Epilepsy: Forensic and Clinical Issues

among studies of similar populations. Variables can only be minimized when studies are undertaken of population-based cohorts of incident cases that will allow complete collection of data and the observation of the disease, in this case epilepsy, from initial diagnosis to the time of death. Prevalence data for SUDEP are especially inaccurate because of its underutilization as a ἀnal diagnosis on death certiἀcates in appropriate cases. A better understanding of the prevalence, risk factors, and pathogenesis depends on the availability of complete and accurate postmortem data of persons with SUDEP. The lack of awareness of SUDEP as a common cause of death in persons with epilepsy is acknowledged in data from the United Kingdom (Lip and Brodie 1992; Coyle et al. 1994). In addition to being a problem in the United Kingdom, accurate determination of the national prevalence of SUDEP in the United States is hampered by the lack of mandatory requirements for postmortems, inaccurate death certiἀcate data, and the underuse of SUDEP as a ἀnal diagnosis, even in appropriate cases. Since prevalence data for SUDEP are hampered by the inaccuracy of death certiἀcate diagnoses, the questions of how and why these data are inaccurate need to be addressed. Why is there a bias toward underutilization of SUDEP as a death certiἀcate diagnosis in cases in which no other diagnosis is reasonable? A national survey on how coroners and medical examiners in the United States documented the deaths of persons with a history of epilepsy, and especially how they dealt with the question of SUDEP, found some surprising results (Schraeder et al. 2006). The professional background, demographics, numbers of cases with epilepsy seen annually, details of the seizure histories, percentage of postmortem examinations conducted, and causes of death, including use of SUDEP, were evaluated. Of the coroners and medical examiners who were trained as pathologists, the vast majority (83.5%) accepted SUDEP as a valid postmortem diagnosis, as opposed to 63% of physicians who were not trained as pathologists, and 58% of those who were not medical professionals. Other factors that influenced acceptance of SUDEP as a valid diagnosis included an urban setting, higher numbers of epilepsy-related deaths, and higher autopsy rates. Despite the relatively high level of acceptance of the validity of the diagnosis, the rate of use of SUDEP as an actual ἀnal diagnosis was low, regardless of the educational background of the coroners and medical examiners. For example, its use as a death certiἀcate diagnosis by pathologists as an actual ἀnal diagnosis in autopsied cases of persons with a history of epilepsy that had no other cause found was only 22.9%. The nature of the jurisdiction, i.e., urban versus nonurban, the number of annual cases with history of epilepsy, and the autopsy rate were not factors associated with the rate of utilization of SUDEP on the death certiἀcate in appropriate cases. The coroners and medical examiners more often used diagnoses such as status epilepticus, respiratory failure, fatal seizure, or cardiac arrhythmia in cases that should have been considered as SUDEP. The low utilization rate of this diagnosis, despite the acknowledgment of its being a valid diagnosis, underscores an unexplained paradox that supports two premises, namely, that public records underestimate the prevalence of SUDEP in the United States, and that having more accurate forensic data about the prevalence and cause of SUDEP is an important, but as of yet unachieved, need. The obvious dilemmas involved in establishing any semblance of an accurate rate of occurrence of SUDEP are related to the prevalent paucity of prospective cohort studies, the need for a higher rate of autopsy performance in cases of deaths in persons with a history of epilepsy, and overcoming the reluctance of coroners and medical examiners to recognize this diagnosis as a common cause of death in epilepsy in cases where no other explanation is

Forensic Case Identification

99

extant. The ἀrst dilemma can only be overcome by a well-funded, multiregional longitudinal study. The second can be overcome with a more zealous use of autopsies in persons with a history of epilepsy. The third dilemma requires an ongoing educational effort about the existence of SUDEP and its known risk factors directed to coroners and medical examiners. Any such educational program needs to be directed at reducing the reluctance of coroners and medical examiners, from all educational and professional backgrounds, to use SUDEP as a death certiἀcate diagnosis when appropriate, more than at educating them about its existence since, at least on an intellectual level, most seem to regard it as an entity. The diagnosis of SUDEP is fraught with a sense of uncertainty and is underutilized as a death certiἀcate diagnosis, even in cases where no other cause of death could be documented on autopsy. The reason for this resistance to use SUDEP as a ἀnal diagnosis by coroners and medical examiners is unknown. One could speculate that perhaps the detail of the postmortem examinations may be insufficient. The extent and quality of postmortem data collection in persons with a history of epilepsy was addressed in the United Kingdom (England, Scotland, Wales, and Northern Ireland) as part of a study of the quality of care for persons with epilepsy and published in the National Sentinel Clinical Audit of EpilepsyRelated Deaths Report (Hanna et al. 2002). During the 1-year study period of the audit, 2412 deaths with a history of epilepsy were examined, with 1023 having autopsies. The audit reviewed 439 (43%) of the autopsied cases and 156 (11%) of 1389 deaths that were certiἀed without autopsies. This national audit utilized investigation of 439 deaths with postmortem records and 156 doctor-certiἀed deaths without postmortems. The majority (87%) of deaths with postmortem examinations had an inadequate examination with a lack of or nonstandardized toxicological, histological, or neuropathology tests. Overall, 117 (27%) of the external exams and 29 (7%) of the internal examinations were unsatisfactory. The audit also found that information about epilepsy and the circumstances of death were variable, with pathologists having inconsistent access to such information. The documented inadequacies of the quality of postmortem performance in the United Kingdom were disappointing, especially in the context of there being detailed guidelines on autopsy practice for evaluation of deaths associated with epilepsy in existence (RCPath 1993). Certiἀcation of cause of death, with or without postmortem, was inconsistent and often inappropriate. Of the certiἀcations of death without postmortem as due to epilepsy, 38% were found to be sudden and/or unwitnessed and should have had autopsies. Of deaths having had a postmortem, certiἀcation was inadequate in 41%. This latter observation was the result of several behaviors: phrasing of cause of death was very variable, with the term SUDEP being used in only 10% of audited cases; causes such as asphyxia, aspiration, or status epilepticus were utilized despite a lack of pathological evidence; and, in some cases, every medical condition the person had was listed on the death certiἀcate, even though they were noncontributory to the death. If the deaths are not adequately investigated and categorized, the difficulty in establishing the true number of epilepsy-related deaths is self evident. Only 10% of deaths with no pathological evidence on autopsy and a history of epilepsy were categorized as SUDEP. One brief paragraph in this audit succinctly summarizes the problem: “Accurate death investigation and certiἀcation of epilepsy-related deaths is necessary if this is to be a reliable source of data for public health surveillance, research, and for effective policy-making in relation to the prevention of epilepsy-related deaths.” In what may be an indirect plea for supplementary verbal autopsy type of data, the audit panel also noted that it is important for the pathologists to have access to standardized information collected from all relevant sources, including family members, on the

100 Sudden Death in Epilepsy: Forensic and Clinical Issues

history of the deceased and the circumstances of death. As a result of the epilepsy audit, the Royal College of Pathology Working Party on the Autopsy (2005) developed guidelines on autopsy practice for deaths associated with epilepsy. The guidelines include the following categories: the role of the autopsy to establish that the epilepsy caused or contributed to the death; pathology encountered at autopsy, speciἀc health and safety aspects; clinical information relevant to the autopsy; the autopsy procedure; speciἀc signiἀcant organ systems (brain, heart, lung, bladder); organ retention and brain histopathology; recommended other blocks (organs) for histological examination; other samples required (including toxicology and frozen brain samples for future evaluation); clinicopathological summary; and cause of death opinions/statements. The leading subcategory of the last section deals with the deἀnition of SUDEP. The Royal College of Pathologists clearly is committed to expecting a high standard of evaluation of cause of death in persons with a history of epilepsy. Concern about the accuracy of prevalence data for the occurrence of SUDEP occurs on both sides of the Atlantic Ocean. In the United States, few data exist about how coroners and medical examiners utilized the diagnosis of SUDEP. The accurate and consistent identiἀcation of cases of SUDEP is, and for the foreseeable future will continue to be, problematic. While it is generally accepted that autopsy conἀrmation is the sine qua non for making a diagnosis of deἀnite SUDEP, the acknowledged lack of autopsies of persons with a history of epilepsy means that the diagnosis of probable SUDEP is, for practical purposes, as close as we can come to this standard. This unfortunate reality should provide impetus for the development of a system, a verbal autopsy if you will, that contains a sufficient spectrum and amount of information about the victim, providing the highest possible level of conἀdence in the accuracy of diagnosis of SUDEP, short of autopsy conἀrmation. In order to determine how to accumulate the most accurate non-autopsy-based diagnostic information, we suggest that a multicenter study should be designed that would fund autopsies for all persons who die with a history of epilepsy within each center’s jurisdiction. In parallel with this postmortem examination study, a well-designed set of investigative questions could be used to acquire all available relevant date about the victim. The goal of this comparative study would be to optimize the non-autopsy-based data with the expectation that it would approximate as close as possible the accuracy of the postmortem diagnoses. If it were to be determined that the so-called verbal autopsy data do not come close to approximating the autopsy-based data, the implication would be that the only feasible approach to improving the accuracy of diagnosis of SUDEP would be to create an ongoing educational effort for coroners and medical examiners, emphasizing the need to maximize the performance of autopsies in cases of epilepsy where the cause of death is unclear and in being comfortable with the diagnosis of SUDEP on death certiἀcates. When coroners and medical examiners in the United States were surveyed about the extent of autopsy examinations they actually performed on persons with epilepsy (Schraeder et al. 2009), it was determined that, regardless of the educational background of the officials, 82.2% stated that a forensic pathologist was engaged to perform the autopsy, and the rest used a general pathologist. The brain was more likely to be removed for examination when the coroner or medical examiner was a pathologist and was in an urban compared to a nonurban jurisdiction. Urban jurisdictions were more likely to collect postmortem blood samples for antiepileptic drug levels. Nonperformance of an autopsy was due to cost, lack of consent, lack of personnel, or lack of time. However, cost of autopsy was more of an issue in nonurban jurisdictions. It is clear from these data that there is a need for more routine brain removal and histopathology in jurisdictions not run by pathologists, and a need for

Forensic Case Identification

101

more diligent use of postmortem antiepileptic drug level determination. The overall quality of autopsy performance in the United States is adequate for the purpose of ἀnding a cause of death. The limitations in the cases of SUDEP are related to not doing autopsies in many of the appropriate cases, and a reluctance to use the diagnosis of SUDEP, even in autopsied cases where the diagnosis is indicated. To ἀnd solutions to a clinical problem, one must ἀrst understand how it is deἀned. To prevent SUDEP, one must understand the risk factors and mechanisms (Lathers et al. 2008). One obvious problem in deἀning the occurrence of SUDEP, as noted earlier, is that the diagnosis of SUDEP in the United States is underutilized as the actual diagnosis on death certiἀcates (Schraeder et al. 2006, 2008). Thus, the incidence of SUDEP, most likely, is not accurately reflected in any data utilizing death certiἀcate diagnoses and is, in reality, likely to be more prevalent than assumed. As early as 1994, Coyle et al. commented that the incidence of SUDEP might be difficult to ascertain due to variations in reporting the cause of death investigated by coroners. They examined this problem using information from the United Kingdom news press during the year 1992 to identify all likely cases of SUDEP. As required by law in the United Kingdom, these unexplained deaths were investigated by pathologists and coroners. All relevant information, including postmortem reports and witness statements, was considered in the 40 SUDEP cases. Coyle et al. (1994) noted inconsistencies in the investigations performed and the observations documented at the time of death, with widely differing degrees of detail concerning the type and history of epilepsy being found. In 70% of these cases, the type of epilepsy was not documented. Inconsistencies were also found concerning details of antiepileptic drug use, the position of the body, toxicology reporting, and the detailed exams of organs, including the brain. The causes of death attributed to the cases varied greatly, with epilepsy noted as a primary cause of death in fewer than half of the cases. Coroner verdicts varied, with no distinct pattern emerging in relation to attributed cause of death. Wide variation in practice of individual corners and pathologists in the investigation and registering of sudden deaths reflected inconsistent quality of performance. Since most deaths were unwitnessed, investigation by and statements from officials in the process of registering deaths are essential to providing accurate information. Finally, in order to obtain an accurate proἀle of individuals at risk for SUDEP, epilepsy, and especially SUDEP, must be written on the death certiἀcate. Until this happens, many more, if not the majority of these deaths, will continue to go unrecorded. Likewise, the data for third world countries, which is even more limited than that in Europe and North America, and that does not depend on the use of autopsies, indicates that SUDEP is an underreported cause of death in persons with epilepsy. In general, developing countries do not have a complete civil registration system in place to collect and provide data regarding public health issues (Hill et al. 2007). Collection of data regarding births, deaths and causes of death, population censuses, sample vital registry systems, descriptive epidemiology of the population, demographic surveillance sites, and internationally coordinated sample survey programs, combined with enhanced methods of assessment and analysis are not in place. For example, a SUDEP underreporting problem has been described in Nigeria where most deaths in individuals with epilepsy occur at home (Sanyo 2005), with autopsies usually not conducted to determine the cause of death, and thus are never reported. Several authors have discussed the burden of epilepsy in developing countries (Jallon 2002; Diop et al. 2005). Nearly 75% of 50 million patients with epilepsy live in countries where diagnostic and therapeutic options are either not affordable, inadequate, or

102 Sudden Death in Epilepsy: Forensic and Clinical Issues

nonexistent. Many patients with epilepsy do not receive treatment because drugs are not available or, even if available, may not be taken for cultural and/or socioeconomic reasons. Although one would not predict that the prevalence of epilepsy should differ in underdeveloped countries versus developed countries, epidemiological studies from developed countries do suggest differences, with a much higher occurrence in developed countries. The reason for this discrepancy is unknown, but stigmatization associated with having epilepsy, limited access to health care personnel and diagnostic and treatment options are likely factors. In underdeveloped countries in Africa, the incidence of epilepsy varies between 10 and 55 per 1000, with an estimated mean prevalence of 15 per 1000. This ἀgure has also been reported for Latin America. Higher rates are found in lower socioeconomic groups and in rural areas. Methodological limitations raise questions about the accuracy of the reported results from developing countries, and include case deἀnition, case diagnosis, classiἀcation, and selection bias. Jallon (2002) suggests that the World Health Organization questionnaire used in such studies does not provide adequate data on seizure type nor syndromic classiἀcation. Sanyo (2005) recommends, as one obvious approach to the problem of underreporting epilepsy as a cause of death, emphasizing the problem to clinicians through articles in the published literature, with a focus on case deἀnition and criteria for diagnosis, risk factors, pathophysiology discussion, and a discussion of treatment options. Since the incidence of SUDEP increases with seizure severity, early onset epilepsy, poor seizure control, generalized tonic–clonic seizures, use of multiple antiepileptic medications, and frequent adjustment of antiepileptic drugs, it is important to understand the interplay of these factors. The possibility of altered cardiorespiratory reflexes causing central apnea, hypoxia, and edema along with cardiac arrhythmia must be understood by all caregivers and family members. Patients, relatives, and caregivers should be educated about the clinical problem of SUDEP and the need to reduce the incidence of SUDEP. Optimal seizure management with effective monotherapy is a common target for the physician managing patients with epilepsy. If intractable epilepsy exists, vagal nerve stimulation and neurosurgery are options to be considered as early on in the treatment regimen as possible to prevent SUDEP. As deἀned earlier, SUDEP refers to the sudden death of an individual with a clinical history of epilepsy, in whom a postmortem examination fails to uncover a gross anatomic, toxicological, or environmental cause of death. Evidence of terminal seizure activity may not be present. In 2002, Shields et al. reported that 1–2% of natural deaths certiἀed by the medical legal death investigators in the United States are attributed to epilepsy and that increased microscopic examination of the brain postmortem has allowed identiἀcation of structural changes that are not grossly evident, but are, nonetheless, representative of the underlying substrates of epileptogenic foci. Over all, Shields et al. (2002) examined 70 death cases, all with known clinical history of seizures, and classiἀed them as:

1. Individuals who lacked a gross brain lesion 2. Individuals with a brain lesion demonstrable at autopsy 3. Individuals who lacked neuropathological evaluation because of decomposition 4. Individuals for whom only an external examination was done

One potential method to correct the problem of underreporting the occurrence of SUDEP is to use a version of verbal autopsies. To date, this technique that has been used primarily in underdeveloped countries. The methodology emphasizes the importance of

Forensic Case Identification

103

talking with members of the family and/or close friends of the patient who has died unexpectedly. Discussions must include determination of facts about the longitudinal history of epilepsy in the deceased, especially including information about the clinical events at the time of a witnessed death. Verbal autopsies are used to determine a likely cause of death for individuals with no direct clinical observation regarding the terminal event (Snow et al. 1993; Lathers and Schraeder 2009). In such situations, the interview conducted during the verbal autopsy seeks key symptoms and signs cited by the relatives and friends of the deceased. The ἀnal diagnosis of cause of death is assigned according to the symptom complexes. Hill et al. (2007) emphasize that these alternate methods do not replace the need for accurate cause of death registration and rather should be used as part of a full, comprehensive health information system. International organizations, national governments, and academics need to work on improving the collection of data and ensuring that methods to do so are continually updated. In many areas of Africa, with limited health personnel and facilities, data indicate 80–85% of those with epilepsy have not been diagnosed or treated, which is deἀned as the treatment gap (Diop et al. 2005). While worldwide data suggest that the mortality of those with epilepsy is two to three times greater than the general population, such information does not exist for Africa. Diop et al. (2005) also suggested that verbal autopsy studies are needed since they do not rely on mostly nonexistent physical autopsies and death certiἀcates. They examined existing studies and noted that one reported a 6-fold increase in mortality in Africa for those with epilepsy, a rate that is considerably higher than the 2- or 3-fold increase generally reported for developing countries. While the high prevalence of this problem means that the public health impact of epilepsy mortality is likely to be enormous, no mention is made of how many of these deaths in persons with epilepsy were SUDEP. As a consequence of the problems listed above, the use of verbal autopsy is increasing in low-income countries to ascertain the likely cause of death (Aspray 2005). Information is obtained on cause of death via interviews with relatives or other associates of the deceased. While it may be used in situations with inherently poor data on adult or pediatric mortality, one must be aware of the strengths and weaknesses of verbal autopsies. 6.1.2â•…Strengths of Verbal Autopsy

1. Verbal autopsy does not depend on access to clinic services. 2. Clinical experts are used for classiἀcation process. 3. Algorithms may be developed for automatic coding of the cause of death. 4. It offers insight into the cause of death in sudden, unexpected deaths in both developing and developed countries. 5. Alternatives such as death certiἀcates do not always provide complete information and may also be imprecise. 6.1.3â•…Weaknesses of Verbal Autopsy

1. Methodological issues of robustness when examining less-frequent causes of death 2. Inability to ascertain a deἀnite SUDEP mortality classiἀcation 3. Possible selection bias in assessing cause of death, e.g., social taboos 4. Reliability of the information provider

104 Sudden Death in Epilepsy: Forensic and Clinical Issues

Verbal autopsies are used as a means of determining probable causes of death for individuals with no reliable clinical information regarding the terminal illness (Snow et al. 1993). Chandramohan et al. (1998) examined maternal deaths using data obtained in a rural district hospital in Tanzania, a rural teaching hospital in Ethiopia, and a rural district hospital in Ghana. They found reasonable agreement between physician diagnoses and algorithms but reported individual misclassiἀcations of causes of death were higher in algorithm-based verbal autopsies. These authors also reported verbal autopsies by a panel of physicians performed better than those based on an algorithm. Validity of the verbal autopsy diagnosis was highest for acute febrile illnesses, TB/AIDS, tetanus, rabies, and injuries. In 1997, Reeves and Quigley reviewed data-derived methods for assigning causes of death from verbal autopsy data, citing three primary methods of empirical classiἀcation rules: linear and other discriminate techniques, probability density estimation and decision trees, and rule-based methods. They also discussed the factors that need to be taken into account when selecting a classiἀcation method to be used for assigning cause of death from verbal autopsy data. To compare the performance of classiἀers derived using different methods requires a large verbal autopsy data set. Quigley et al. (1999) compared diagnostic accuracy of physician review, expert algorithms, and data-derived algorithms in adult verbal autopsies conducted in hospitals in Tanzania, Ethiopia, and Ghana. They concluded that in settings where physician review is not feasible, expert and data-derived algorithms do provide a reasonable alternative approach for assigning many causes of death. Chandramohan et al. (2001) also discussed the fact that verbal autopsy is an indirect method of ascertaining cause of death from symptoms and signs obtained from bereaved relatives and friends and, thus, the mortality estimates are susceptible to bias due to misclassiἀcation of causes of death. These authors explored application of sensitivity and speciἀcity of verbal autopsy data obtained from a hospital-based validation study to adjust the effect of misclassiἀcation errors in data obtained during verbal autopsies from a demographic surveillance system. They concluded that it is not possible to use sensitivity and speciἀcity estimates derived from a location-speciἀc validation study to adjust for misclassiἀcation in verbal autopsy data from populations with substantially different patterns of cause-speciἀc mortality. Boulle et al. (2001) examined artiἀcial neural networks as a method of classifying the cause of death from verbal autopsy in a wide range of disciplines. Artiἀcial neural networks were applied to data from a verbal autopsy study as a means to classify cause of death. Large data sets are needed to improve performance of data-derived algorithms, especially in artiἀcial neural network models. Freeman et al. (2005) compared agreement between causes of death obtained from the verbal autopsy by physician review versus computer-based algorithms and recommended the analysis technique combine both methods. They concluded that further validation studies were needed to improve performance of the verbal autopsy for assigning causes of neonatal death. Verbal autopsy procedures, for use in sample vital registration, are important methods for deriving causespeciἀc mortality estimates where disease burdens are greatest and routine cause-speciἀc mortality data do not exist (Setel et al. 2006). Lee et al. (2008) used verbal autopsy methods to ascertain birth asphyxia deaths in a community-based setting in southern Nepal. The authors note that use of various verbal autopsy deἀnitions and hierarchical approaches to assign cause of death may substantially affect estimates of birth asphyxia–speciἀc mortality and analyses of risk factors. Verbal autopsy methods need to be standardized and validated to generate accurate global estimates to determine policy and resource allocation in developing countries.

Forensic Case Identification

105

To glean information about the circumstances of death in sudden death in epilepsy, Nashef et al. (1998) conducted interviews with referred bereaved relatives of patients with epilepsy who had died suddenly. These data were combined with information derived from coroners’ reports, postmortem reports, previous medical records, and EEG data. Thus, the verbal data obtained in this study was a useful supplement to the other data available and was helpful in addressing the question of the need of relatives of SUDEP victims to have had antemortem knowledge of the possible risk of SUDEP. Most of the relatives stated they would have preferred to have been told that epilepsy could be fatal. The interviews emphasized the needs of the bereaved relatives and their sense of isolation in the face of an unexpected death. Thus, in contrast to developing countries where verbal autopsy may be the only means of establishing a possible or probable cause of death, this technique may have a different use in more affluent countries, namely of helping to clarify questions not answered by the standard methods of coroner’s and postmortem reports and available medical records. In summary, the purpose of verbal autopsy can be multifaceted. When used in conjunction with postmortem autopsy data from persons with SUDEP, it can help in focusing on retrospective data that provide additional help in identifying more accurately the cause of death, and in conducting retrospective analyses of these postmortem examinations. The value of all of this cumulative information from all sources is that it provides information for future preventative policy. In circumstances where postmortem information was not or could not be collected, verbal autopsies do offer an alternative method to ἀnding information regarding the mechanism of death, whether conducted in developing countries or in developed countries. In either case, regardless of how it is acquired, the worldwide database on persons with epilepsy who die suddenly and unexpectedly will gain important information that will help in determining the prevalence of SUDEP and contribute to the quest for identiἀcation of preventive interventions. The technique of verbal autopsy makes the assumption that individual disease entities have discrete symptom complexes that can be easily and accurately recognized and remembered (Reeves and Quigley 1997). The accuracy of the verbal autopsy technique in cases of SUDEP needs to be established. One possible way of accomplishing this goal would be to compare the accuracy of this method of detailed information gathering in predicting a diagnosis of deἀnite SUDEP in the cases that end up with a postmortem examination and ἀnal autopsy diagnosis of SUDEP. If it were found that an adequate, standardized verbal autopsy compared favorably, in accuracy of diagnosis, with the ἀnal physical autopsy diagnosis, then in cases of epilepsy-related deaths in which, for whatever reason, an autopsy is not performed, the verbal autopsy could reasonably be used as a substitute. However, until such a study is forthcoming, the physical autopsy will continue to be essential to determine the diagnosis of deἀnite SUDEP.

References Aspray, T. J. 2005. The use of verbal autopsy in attributing cause of death from epilepsy. Epilepsia 46 (Suppl 11): 15–17. Boulle, A., D. Chandramohan, and P. Weller. 2001. A case study of using artiἀcial neural networks for classifying cause of death from verbal autopsy. Int J Epidemiol 30: 515–520. Chandramohan, D., G. H. Maude, L. C., and Rodriques, R. J. 1998a. Verbal autopsies for adult deaths: Their development and validation in a multicentre study. Trop Med Int Health 3: 436–446. Chandramohan, D., L. C. Rodriques, G. H. Maude, and R. J. Hayes. 1998b. The validity of verbal autopsies for assessing the causes of institutional maternal death. Stud Fam Plann 29: 414–422.

106 Sudden Death in Epilepsy: Forensic and Clinical Issues Chandramohan, D., P. Setel, and M. Quigley. 2001. Effect of misclassiἀcation of causes of death in verbal autopsy: Can it be adjusted? Int J Epidemiol 30: 509–514. Coyle, H. P., N. Baker-Brian, and S. W. Brown. 1994. Coroners’ autopsy reporting of sudden unexplained death in epilepsy (SUDEP) in the UK. Seizure 3: 247–254. Diop, A. G., D. C. Hesdorffer, G. Logroscino, and W. A. Hauser. 2005. Epilepsy and mortality in Africa: A review of the literature. Epilepsia 46: (Suppl 11): 33–35. Earnest, M. P., et al. 1992. The sudden unexplained death syndrome in epilepsy demographic, cÂ�linical, and postmortem features. Epilepsia 33: 310–316. Ficker, D. M., E. L. So, W. K. Shen et al. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51: 1270–1274. Forsgren, L., W. A. Hauser, E. Olafsson et al. 2005. Mortality of epilepsy in developed countries: A review. Epilepsia 46 (S11): 18–27. Freeman, J. V., P. Christian, S. K. Khatry et al. 2005. Evaluation of neonatal verbal autopsy using physician review versus algorithm-based cause-of-death assignment in rural Nepal. Pediatr Perinat Epidemiol 19: 323–331. Hanna, N. J., M. Black, J. W. S. Sander et al. 2002. The National Sentinel Clinical Audit of EpilepsyRelated Death; The Stationery Office. Hill, K., A. D. Lopez, K. Shibuya, and P. Jha. 2007. On behalf of the monitoring of vital events (MoVE) writing group. Interim measures for meeting needs for health sector data: Births, deaths, and causes of death. Lancet 368. Jallon, P. 2002. Epilepsy and epileptic disorders, an epidemiological marker? Contribution of descriptive epidemiology. Epileptic Disorders 4: 1–13. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies and SUDEP. Epilepsy Behav 14: 573–576. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Lee, A. C., L. C. Mullany, J. M. Tielsch et al. 2008. Verbal autopsy methods to ascertain birth asphyxia deaths in a community-based setting in southern Nepal. Pediatrics 121: E1372–E1380. Leestma, J. 1990. Sudden death associated with seizures: A pathological review. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 61–89. New York, NY: Marcel Dekker. Leestma, J. E., J. F. Annegers, M. J. Brodie et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. Lipp, G. Y., and M. J. Brodie. 1992. Sudden death in epilepsy: An avoidable outcome? J R Soc Med 85: 665–676. Logroscino, G., and D. C. Hesdorffer. 2005. Methodological issues in studies of mortality following epilepsy: Measures, types of studies, sources of cases, cohort effects, and competing risks. Epilepsia 46 (Suppl 11): 3–7. Morgan, C. L., and M. P. Kerr. 2002. Epilepsy and mortality: A record linkage study in a UK population. Epilepsia 43: 1251–1255. Nashef, L. 1997. Sudden unexpected death in epilepsy: Terminology and deἀnitions. Epilepsia 38 (sl11): S6–S8. Nashef, L., S. Garner, J. W. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64: 349–352. Quigley, M. A., D. Chandramohan, and L. C. Rodriques. 1999. Diagnostic accuracy of physician review, expert algorithms and data-derived algorithms in adult verbal autopsies. Int J Epidemiol 28: 1081–1087. RCPath. 1993. Guidelines for Post-Mortem Reports. Royal College of Pathologists, London. RCPath Working Party on Autopsy. 2005. Guidelines on Autopsy Practice. Scenario 6: Deaths Associated with Epilepsy. January 2005. Reeves, B. C., and M. Quigley. 1997. A review of data-derived methods for assigning causes of death from verbal autopsy data. Int J Epidemiol 26: 1080–1089.

Forensic Case Identification

107

Sanyo, E. O. 2005. Increasing awareness about sudden unexplained death in epilepsy—A review. Afr J Med Med Sci 34: 323–327. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68: 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy examinations in sudden unexplained death in epilepsy. Am J Forensic Medicine Pathol 30(2): 123–126. Setel, P. W., D. R. Whiting, Y. Hemed et al. 2006. Validity of verbal autopsy procedures for determining cause of death in Tanzania. Trop Med Int Health 11: 681–696. Shields, L. B., D. M. Hunsaker, J. C. Hunsaker 3rd, and J. C. Parker Jr. 2002. Sudden unexpected death in epilepsy: Neuropathological ἀndings. Am J Forensic Med Pathol 23: 307–314. Snow, R. W., I. Basto de Azevedo, D. Forster et al. 1993. Maternal recall of symptoms associated with childhood deaths in rural east Africa. Int J Epidemiol 22: 677–683. Tsai, J. J. Mortality of epilepsy from national vital statistics and university epilepsy clinic in Taiwan. Epilepsia 46 (S11): 8–10. Zielinski, J. J. 1974. Epilepsy and mortality rate and cause of death. Epilepsia 15: 191–201.

Sudden Unexpected Death in Epilepsy Future Research Directions

7

Simona Parvulescu-Codrea

Contents 7.1 7.2 7.3 7.4

Introduction Why Should We Focus on SUDEP? Epidemiologic Data Supportive of a Cardiac Mechanism of SUDEP Clinical Evidence in Support of a Cardiac Mechanism: The Importance of Increased Heart Rate Associated with Epileptic Seizures 7.5 In Vivo Evidence for an Arrhythmogenic Substrate in the Hearts of Epilepsy Patients at Risk for SUDEP 7.6 Evidence of Sinus Tachycardia Associated with Epileptic Seizures 7.7 Bradycardia and Asystole 7.8 Ictal Cardiac SPECT Imaging Supportive of a Cardiac Mechanism of SUDEP 7.9 Future Research Avenues for an Underlying Cardiac Mechanism of SUDEP: Epilepsy, Depression, and Cardiac Disease 7.10 Evidence of Cardiac Arrhythmogenic Substrate for SUDEP: The Influence of Antiepileptic Drugs on the Heart 7.11 Summary and Clinical Perspectives Acknowledgment References

109 110 110 111 112 115 119 119 121 123 123 124 124

7.1â•…Introduction The systems in which our healthcare workers function are fragmented, complex, and narrowly specialized, creating gaps in communication, numerous hands-off circumstances, and more discipline-centered rather than patient-centered care (Begun et al. 2006). Consequently, as emphasized by Lathers et al. (2008) in the article “The Mystery of Sudden Death—Mechanisms for Risks,” sudden unexpected death in epilepsy (SUDEP) is an excellent example of a medical enigma, which in order to be solved, challenges scientists to break the current work paradigm and mount their efforts in a multidisciplinary team collaboration.

109

110 Sudden Death in Epilepsy: Forensic and Clinical Issues

7.2â•…Why Should We Focus on SUDEP? At least 50 million people in the world have epilepsy (WHO 2009). The estimated portion of the general population with active epilepsy, i.e., continuing to have seizures or need for treatment, at any given time is from 4 to 10 per 1000. Spratling (1902) was one of the ἀrst neurologists to recognize epilepsy as a disease that can destroy life suddenly and without warning during a single brief attack. Subsequently, many epidemiologic studies conἀrmed that epilepsy increases the risk of premature death by at least twofold to threefold, compared to the general population (WHO 2009). SUDEP is considered to be the most important culprit for seizure-related deaths. Nevertheless, despite the fact that epilepsy is so common and that we have been aware of the associated risk with SUDEP, investigation into the mechanisms and risk factors has so far been frustratingly unproductive. As one bereaved mother of a victim of SUDEP wrote, “Nobody knows how my son died because, although epilepsy is the most common neurological brain disorder in the world, it is also the most neglected.” (Hanna 2007). Accordingly, it is of vital importance that this area receives the deserved attention (Willams and Richens 1998).

7.3â•…Epidemiologic Data Supportive of a Cardiac Mechanism of SUDEP The mechanism of SUDEP is the subject of an ongoing scientiἀc debate. Literature over the past decade provides mounting epidemiologic evidence for increased frequency of cardiac abnormalities in patients experiencing epileptic seizures. This association proves to be more frequent than previously thought, especially in relation to SUDEP (Tinuper et al. 2001; Tigaran et al. 2003; Nei et al. 2000a; Chin et al. 2004; Ho et al. 2006; So and Sperling 2007; DeGiorgio et al. 2008; Scorza et al. 2008; Strzelczyk et al. 2008). Consequently, when the high incidence of SUDEP is taken into account (Pedley and Hauser 2002), there is a need for prompt and correct diagnosis of epilepsy and early recognition of possible risk factors that may be important in the prevention of SUDEP. Furthermore, it should be of special interdisciplinary interest to diagnose cases of epileptic seizures where the only ictal manifestation is autonomic dysfunction that may produce changes in the heart rhythm and rate (Zijlmans et al. 2002). Postmortem abnormalities such as subendocardial scarring in the myocardium, along with pathologic changes consisting of perivascular and interstitial ἀbrosis, contraction band necrosis, and the myocyte vacuolization found in autopsied SUDEP victims are supportive of a cardiac mechanism for SUDEP (Manno et al. 2005; P-Codrea Tigaran et al. 2005; Natelson et al. 1998; Falconer and Rajs 1976). Because of the discrete microscopic nature of the myocardial ἀbrosis, we may not be able to detect these abnormalities in vivo (Tigaran et al. 2002). However, cardiac ischemia, leading to the discrete areas of ἀbrosis in the myocardium, can create the pathophysiological substrate for fatal re-entrant arrhythmias, which may result in sudden unexpected death. Because the patients can ultimately die due to fatal arrhythmias, there is no postmortem evidence of the proposed cardiac mechanisms, because arrhythmias do not give any pathologic residua in the myocardium. Thus, we face a postmortem pathologic puzzle. Accordingly, awareness of the fact that pathological cardiac changes such as myocardial ἀbrosis triggered by repetitive epileptic seizures (P-Codrea Tigaran et al. 2005) potentially can lead to SUDEP, may be increasingly important in understanding neurocardiac mechanisms for SUDEP (Beran 2006).

Sudden Unexpected Death in Epilepsy

111

7.4â•…Clinical Evidence in Support of a Cardiac Mechanism: The Importance of Increased Heart Rate Associated with Epileptic Seizures The subendocardium has long been considered the region most vulnerable to ischemia, which occurs whenever there is an imbalance between myocardial oxygen supply and oxygen demand (Hoffman and Buckberg 1975). In the experimental situation, the pressure overload consequent to aortic stenosis disrupts the adequacy of subendocardial perfusion, induces evidence of ischemia, such as lactate accumulation, in response to a heart rate increase from 120 to 180 beats/min or in response to an increase in the severity of aortic stenosis (Brazier and Buckberg 1975). Subendocardial ischemia and infarction occur in association with aortic stenosis despite the absence of anatomic obstruction to coronary blood flow. This ischemia occurs when myocardial oxygen demands are increased as a result of the tachycardia of exercise or catecholamine infusion in patients with aortic stenosis. Similar increments in heart rate or stress do not cause EKG abnormalities in patients without left ventricular outflow obstruction or coronary artery disease. These observations suggest that the ischemia occurs because subendocardial oxygen delivery does not rise sufficiently to meet myocardial oxygen demands. Tachycardia shortens diastole more than systole and would tend to jeopardize blood flow to the subendocardium, since this region receives its blood supply during this phase of the cardiac cycle. Normally, coronary flow is augmented by vasodilation during tachycardia to supply added oxygen to meet the rate-induced rise in metabolic needs. In aortic stenosis, the coronary arteries may be nearly maximally dilated under resting conditions so that coronary flow can meet the increased oxygen requirements. Increasing heart rate in aortic stenosis may therefore limit available subendocardial blood supply and cause ischemia. In these studies, the relationship between myocardial oxygen supply and demand was deἀned by the diastolic and sytolic tension time index. It was demonstrated that changes in these indices induced by tachycardia were related to redistribution of blood flow away from the subendocardium and a reduced subendocardial oxygen delivery per unit of demand (Brazier and Buckberg 1975). The use of these indices has been expanded to include situations where aortic stenosis is not present. Foster et al. (1981) demonstrated that humans showed slightly compromised left ventricle contractile function when subjected to sudden strenuous exercise, but normal responses when exposed to graded exercise. They then suggested that this mechanism might be applicable to a group reported to have ischemic changes with sudden strenÂ� uous exercise (Foster et al. 1981, 1982). Their hypothesis, that subendocardial ischemia could occur in response to sudden strenuous exercise due to supply and demand imbalÂ� ances, relies on these indices. This concept is further supported by data demonstrating that sudden strenuous exercise performed without warm-up (Duncan et al. 1987) results in a transitory decrease in coronary blood flow at the onset of exercise, at a time when myocardial oxygen requirements are rapidly increasing. This may be due to abrupt increases in alpha adrenergic tone that are known to occur with sudden exertion (Aung-Din et al. 1981; Murray and Vatner 1981). It was also documented (Katz and Feinberg 1958; Khouri et al. 1965) that there are marked increases in myocardial oxygen demand due to sudden stress, which then leads to widening of the systolic time tension index. Thus, there may be both an increase in myocardial oxygen demand and a reduction in supply in this situation, which could result in myocardial ischemia. Such a paradigm would closely ἀt the situation of seizure patients. Many of the seizures, especially those that caused

112 Sudden Death in Epilepsy: Forensic and Clinical Issues

ischemic changes as evidenced by the electrocardiographic ST segment changes, occurred while the patients were asleep (Tigaran et al. 2003), which during normal NREM sleep, is a time of decreased sympathetic tone and increased parasympathetic activity, normally creating a state of reduced activity, and which, during REM sleep, is characterized by increased parasympathetic activity and variable sympathetic activity associated with increased activation of certain brain functions (Harris 2005; Yasue 1983). Furthermore, epileptic patients, especially those with tonic–clonic seizures, are known to have apnea during seizures (Nashef et al. 1996; So et al. 2000; Johnston et al. 1995). Thus, in addition to sudden exertion and the abnormalities in supply and demand described above, there are addiÂ�tional provocative stimuli, such as hypoxia for cardiac ischemia, in these patients. Moreover, as shown by Nei et al. (2000a), epilepsy patients may have impaired autonomic function, which, in the context of seizure-related rapid high catecholamine turnover, represents an additional facilitator for cardiac ischemia and arrhythmia. These ἀndings are consonant with the pathology of formation of cardiac contraction band necrosis as a result of massive catecholamine release, as shown in patients who died as a result of status epilepticus (Manno et al. 2005), along with ἀndings of perivascular interstitial ἀbrosis and myocyte vacuolization often found in patients with SUDEP (P-Codrea Tigaran et al. 2005). It is well known that increased amounts of ἀbrotic tissue in the heart strongly correlates with an increased incidence of atrial and ventricular tachyarrhythmias and sudden cardiac death (Ten Tusscher and Panἀlov 2007). These islands of necrosis found in the hearts of some persons with epilepsy could provide the substrate for the stimulation of malignant arrhythmias, such as ventricular tachycardia (Espinosa et al. 2009), and may represent the structural damage required to initiate ventricular arrhythmias through a functional re-entrant mechanism (El Sherif 1994).

7.5â•…In Vivo Evidence for an Arrhythmogenic Substrate in the Hearts of Epilepsy Patients at Risk for SUDEP A special electrocardiographic technique, known as signal-averaged electrocardiography, or ventricular late potentials, can be used to examine the presence of an arrhythmoÂ� genic substrate such as the patchy subendocardial ἀbrosis detected when the otherwise healthy hearts of the SUDEP victims have been autopsied. It should be noted that, even in patients with known heart disease and known ventricular ectopy, late potentials have a good negative predictive value but have an extremely poor positive predictive value. Today in clinical cardiology, signal-averaged electrocardiography is being supplanted by other techniques such as the microvolt T-wave alternans (TWA) test (Cambridge Heart, Bedford, MA) (Heart 2009), which monitors for minute beat-to-beat variations in the electrocardiographic T-wave during exercise stress. Even though the data presented in this chapter are unique in their application to a group of persons with epilepsy, there is obviously a need for additional studies using the more sophisticated newer techniques and a larger number of patients. Late potentials are produced by delayed or slowed conduction of the ventricular myocardium. When applied in coronary heart disease, late potentials have been shown to represent diseased myocardium with delayed depolarization, which may serve as the substrate for the re-entrant mechanism causing ventricular tachycardia and sudden cardiac death (El Sherif 1994).

Sudden Unexpected Death in Epilepsy

113

The prevalence of electrocardiographic late potentials is generally increased by the amount of scarring of the myocardium, as noted previously. The prevalence of late potentials is highest in patients with the most severely damaged myocardium, such as in patients with reduced left ventricular function due to previous myocardial infarctions (Christiansen et al. 1995). In autopsy studies of SUDEP victims, the ἀndings were represented by subendocardial patchy ἀbrosis, which would very likely not be visible on echocardiography or in a 12-lead electrocardiogram, but only detectable by the signal averaging electrocarÂ�dioÂ�graphic technique, namely the late potentials. Accordingly, in 1999 we were the ἀrst to employ this technique in both unmedicated and medicated epilepsy patients (Tigaran et al. 1999). Twenty-three patients were enrolled in the study. All the patients had normal 12-lead electrocardiograms and echocardiograms, as well as normal myocardial scintigraphy stress tests. The patients were used as their own controls. In the ἀrst phase, the patients were medicated when they underwent 15 min of raw electrocardiographic late potentials recordÂ� ings study. Before the ἀrst electrocardiogram, a blood sample was drawn to determine the antiepileptic drug concentration. Three of those were on monotherapy with lamotrigine, one on monotherapy with vigabatrin, and one on topiramate. The remaining eight received polypharmacy, with two (n = 5 patients) and three (n = 3 patients) drugs. Only one patient receiving polypharmacy was on carbamazepine, the others received topiramate (n = 5 patients), clobazam (n = 5 patients), lamotrigine (n = 1 patient), vigabatrin (n = 1 patient), gabapentine (n = 1 patient), and oxcarbazepine (n = 4 patients). Two of these drugs do not have sodium channel blocker properties (vigabatrine, gabapentine). In this ἀrst phase, 59% (13 out of 22 patients) were late potentials positive (8 men and 5€women). In one patient, the data could not be analyzed for technical reasons, due to the presence of a continuous electrocardiographic P–Q interval. For the second phase, namely the off-drugs phase, the investigation was not performed in the ἀrst ἀve patients. For the remainder of the patients undergoing this late potential electrocardiographic study, before the completion of the study, a blood sample was drawn to determine if any antiepileptic drugs could be detected in plasma. Thus, we detected that in two patients the antiepileptic drugs were within the therapeutic range when the investigation was conducted, leaving 15 patients with pertinent results after undergoing both on- and off-drugs late potentials electrocardiography study. Out of these 15 patients, there were initially 7 (47%) late potentials positive. With the withdrawal of all drugs, ἀve became late potentials negative, leaving only two with persistently positive results [2 out of 15 patients (13%)]. One was on monotherapy with lamotrigine, and one received a combination of topiramate with clobazam. Both patients received a drug with an action mechanism related to the inactivation of the sodium channels. Our results showed a clear correlation between male gender and positive longer signaled averÂ�aged electrocardiographic QRS complex, which correlates with the data across most of the SUDEP literature showing higher SUDEP fatalities among men. Females tended to€normalÂ�ize when antiepileptic medications and all the other drugs were stopped (see€Figure 7.1). Likewise, a high diastolic pressure at study entry was a signiἀcant predictor for poor late potentials normalization when drugs were withdrawn. Statistical work-up was considered to be nonrelevant due to the small numbers of the different subgroups. Our observations raised an important question concerning the background of the high prevalence of the late potential results (see Figure 7.2), which lead us to the following hypotheses:

114 Sudden Death in Epilepsy: Forensic and Clinical Issues 25 20 15 10 5 0 –5 –10 –15 –20 –25

Males Females

RMS

QRS

LPD

Figure 7.1╇ Average increase in measurements induced by stopping medication. LP tends to normalize in women [late potential duration (LPD) and the root mean square (RMS) voltage of the terminal at 40 mS] when AEDs were stopped; LP were considered positive when two parameters were abnormal: QRS > 120 mm, RMS40 < 25 μV, LPD > 40 mS. (From Tigaran, S., Acta Neurol Scand Suppl, 177, 9–32, 2002. With permission.)

1. The changes are due to subendocardial scarring present in the epilepsy patients’ myocardium. However, because of the discrete nature of ἀbrosis found postmortem in the hearts of the epilepsy patients, we may not be able to detect these abnormalities with sufficient sensitivity by means of the present technique. The author must admit that the results following the investigation of the ἀrst ἀve patients enrolled in the study, all of which exhibited positive late potentials, were surprising. Consequently, it was decided to repeat the investigation after the antiepileptic drugs were withdrawn (the patients did not receive any other drugs but antiepileptics), and simultaneously control the antiepileptic drug plasma concentration when investigating the rest of the patients, both in the medicated and in the unmedicated Positive *Vector Butterworth 40–250 Hz

Negative RMS40 : 10.2 µV RMS50 : 19.2 µV QRS : 140 mS

*Vector Butterworth 40–250 Hz

RMS40 : 25.3 µV RMS50 : 40.1 µV QRS : 101 mS

LPD : 46 mS

LPD : 24 mS

Noise : 0.4 µV

Noise : 0.3 µV

200 mm/s 10 µV/10 mm

200 mm/s 10 µV/10 mm

Figure 7.2╇ Example positive versus negative LP image. (From Tigaran, S., Acta Neurol Scand Suppl, 177, 9–32, 2002. With permission.)

Sudden Unexpected Death in Epilepsy

115

phase. Our ἀrst conclusion, coupled with the delayed depolarization of the left ventricle observed in 47% (positive late potentials) of our patients on drugs when compared with the off-drugs phase, led us to the second working hypothesis. 2. The changes are due to the antiepileptic drugs, thus generating the hypothesis that antiepileptics might also influence the electrical properties of the myocyte due to their electrophysiological effects on the conduction system and the myocardium. As previously shown, many antiepileptic drugs exert their effects by blocking sodium channels, which generally slow conduction. Several cardiac antiarrhythmic drugs also exert their effects by blocking this channel (Anderson et al. 1994). This relationÂ� ship has also been observed in other clinical studies of sodium channel blocker drugs. Antiepileptic drugs may thus, through their as yet unrecognized influence on the depolarization process (Freedman and Steinberg 1991), potentially increase the propensity for malignant arrhythmias. In a larger perspective, this could be a potential so-called proarrhythmic effect, which could precipitate cardiac arrhythmias. To the best of the author’s knowledge, there are no available data describing the relationship between the occurrence of late potentials and antiepileptic drugs. While not claiming to have found the ultimate explanation, the author would like to propose that although the two hypotheses are apparently discordant, both variants may be true. Likewise a combination of both hypotheses would also be a possibility, an idea supported by the ἀndings of the electrocardiographic positive ST segment and the positive late potentials. However, the interpretation of these ἀndings requires much caution, warranting further studies to verify the reproductibility of the described ἀndings (Tigaran et al. 2002). Future studies employing the magnetic resonance imaging (MRI) technique will help in clarifying the questions rained by these two hypotheses.

7.6â•…Evidence of Sinus Tachycardia Associated with Epileptic Seizures More than 25 years ago, Marshall et al. (1983) wrote that, “a seldom-recognized accompaniment of temporal lobe seizures is tachycardia.” They were the very ἀrst authors to document by simultaneous “electroencephalographic, electrocardiographic, and videotape monitoring” the presence of this phenomenon in 12 consecutive patients with spontaneous seizures. Several subsequent articles have described this phenomenon, and as a result of the analysis of more than 1500 seizures from more than 1000 epilepsy patients, all of them established without any doubt that, despite the origin of the epileptic seizure, sinus tachycardia is by far the most common cardiac phenomenon temporally associated with epileptic seizures, as one would expect as part of a normal response to stress. Some of the previous articles in the literature are lacking in details related to the exact brain origin of the seizure or the exact type of seizure, i.e., secondary generalization versus complex versus simple partial seizures. In the older studies, this inconsistency is due to the technical limitations of the imaging methods employed at the time of the study. Unfortunately, at times, the electroencephalographic onset of the seizure can be deceiving due to either rapid propagation of the seizure or inadequacy of the scalp EEG recording(s), as shown by studies using intracranial electrodes or magnetoencephalogram studies (Shibasaki et al. 2007). Overall, the heart rate is thought to be signiἀcantly higher with seizures arising from the temporal than from the frontal lobe (Galimberti et al. 1996; Schernthaner et al.

116 Sudden Death in Epilepsy: Forensic and Clinical Issues

130

1999; Garcia et al. 2001; Opherk et al. 2002; Tigaran et al. 2002; Leutmezer et al. 2003) and to occur earlier in patients whose seizures arise during sleep (Opherk et al. 2002). Of note, Tigaran et al. (2003) showed that epileptic seizures induce a rapid increase in heart rate from resting levels to more than 180 beats/min shortly before the onset of electrocardiographic seizures. This early increase in heart rate indicates that seizures can induce a turning on/off of very high levels of efferent neural cardiac sympathetic activity (see Figure 7.3). Such high levels of sympathetic activity (Tigaran et al. 2001) become toxic principally to the cardiac myocyte and could, in the long run, be a contributing factor to the scarring of the myocardium observed in SUDEP victims (see Figure 7.4) (P-Codrea Tigaran et al. 2005; Natelson et al. 1998; Falconer and Rajs 1976), as well as the formation of cardiac contraction bands (Manno et al. 2005). Moreover, Tigaran et al. (2003) have demonstrated the occurrence of myocardial ischeÂ� mia triggered by high heart rate values, as shown by the presence of electrocardiographic ST-segment depression, which is associated with an epileptic seizure (see Figure 7.5), and could represent an additional contributing factor to the pathologic changes of the myocardium, such as ἀbrotic scarring. Strikingly, the failure of the recorded heart rate to recover to the baseline level for as long as 1 h after seizure cessation can be a possible result of the impaired autonomic function (Ushijima et al. 2009) such as Nei et al. (2000a) have demonstrated to be present in some of their epilepsy patients. It is important to consider that, in epilepsy patients, tachycardia triggering cardiac ischemia may have even more serious consequences, especially in patients who already have cardiac disease. This suggests that methods of addressing it might be important and worthy of investigation, such as, for example, blunting the ischemic and heart rate response with beta-blockers. The observations and present hypothesis represent an exciting challenge with regard to experimental clinical testing. Salutary in this context are studies looking into the association between tachycardia and diverse types of ethnic groups. A pioneer in this ἀeld of research, Wilder-Smith

110 100 90 70

80

Average heart rate

120

No ST-segment change ST-segment change

0

10

20

30

40

50

60

Minutes relative to seizure

Figure 7.3╇ Average heart rate (HR) over time by ST-segment changes. 0 = max HR during seizure. (From Tigaran, S., Acta Neurol Scand Suppl, 177, 9–32, 2002. With permission.)

Sudden Unexpected Death in Epilepsy 9

1

2

117 1 2 3 4 5 6 7 8 9 10 11 12 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

8

3

10

4

7

5

6

11

(a)

X

1 2 3 4 5 6 7 8 9 10 11 12 13

8

3

10

4 5

(b)

9

1

2

7 6

11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

X X X

Figure 7.4╇ Comparison between the distribution of the fibrotic changes of the deep and sub-

endocardial myocardium in SUDEP versus control patients. (a) Control patients for fibrotic myocardial areas. (b) Fibrotic areas in the hearts of SUDEP patients. 1–15 (vertical), subject’s identification number; 1–11 (horizontal), sections from the transmural tissue blocks containing the deep and endocardial myocardium; 12, longitudinal tissue block from intermedial septum; 13, longitudinal tissue block from the papillary muscle of the left ventricle; X, presence of fibrotic changes in singular areas which are not present in the stylized figure. Light gray: slight fibrosis: 1 patient with fibrotic changes confined to 1 area. Darker gray, moderate fibrosis: 2 patients with fibrotic changes observed in the same area. Darkest gray, severe fibrosis: 3 patients with fibrotic changes observed in the same area. (From P-Codrea, Tigaran, S. et al., Am J Forensic Med Pathol, 26 (2), 99–105, 2005. With permission.)

(Wilder-Smith and Lim 2001), published a study concerning the changes in heart rate amongst non-Caucasian Singaporean patients. Notably, this study concluded that sinus tachycardia was considerably less frequent in non-Caucasian epilepsy patients. Despite being a very comprehensive review of SUDEP and its mechanism, the Nigerian Kwara State article lacked regional ethnic data that could have addressed this particular topic (Sanya 2005). By the same token, it is quite possible that the American melting pot, more than anywhere else, will claim its own speciἀc studies looking at the issue of racial and ethnic differences in epilepsy-related cardiac arrhythmias.

118 Sudden Death in Epilepsy: Forensic and Clinical Issues

00:49:13 Pre start of Event

01:03:57 Pre start of Event

00:49:47 Max. depress.

01:08:28 Max. depress.

00:51:33 Post end of Event

01:12:11 Post end of Event

Figure 7.5╇ Example from Patient 1 (left) and Patient 9 displaying dynamic ST-segment depression in relation to seizure with secondary generalization. (From Tigaran, S., Acta Neurol Scand Suppl, 177, 9–32, 2002. With permission.)

A conceptually different idea, outside of the direct association between heart rate variations and SUDEP, involves the epidemiology of obesity, which is thought to be a speciἀc important risk factor for occult cardiovascular disease. Thus, complications of obesity in epilepsy patients might influence even more the ability of the heart to adjust to abrupt, sudden variations in its rate and represent a contributory mechanism in SUDEP. Obesity is a chronic metabolic disorder associated with cardiovascular disease and increased morbidity and mortality. It is a widely acknowledged risk factor for developing coronary artery disease (Franzosi 2006). The number of deaths per year attributable to obesity is about 30,000 in the United Kingdom, a country that has produced a signiἀcant amount of SUDEP data. This number is 10 times higher in the United States, roughly about 300,000 deaths per year (Allison et al. 1999), where more than half of the adults are overweight (Flegal et al. 1998) and where obesity is thought to have overtaken smoking in 2005 as the main preventable cause of illness and premature death (Franzosi 2006). If we do death rate extrapolations for the United States for obesity, there will be 300,000 deaths per year, 25,000 deaths per month, 5769 deaths per week, 821 deaths per day, and 34 deaths per hour (WrongDiagnosis.com 2009). Moreover, we also have to keep in mind the speciἀc association between obesity and obstructive sleep apnea, which represents an aggravating and important contributing factor to the genesis of cardiovascular diseases (Hiestand et€al. 2006). Since apnea has already been identiἀed to be associated with epileptic seizures and SUDEP (Nashef et al. 1996; So et al. 2000), this association should prompt even more attention and many more research studies. Another notable association, which might be very signiἀcant for epilepsy and SUDEP, is the signiἀcant increase in the older population and the speciἀc increase of epilepsy prevalence in this age group (Velez and Selwa 2003). Accordingly, the speciἀc concurrent morbidities, such as stroke and cardiac diseases, present in this age category will prompt an even higher awareness of the cardiac complications in epilepsy.

Sudden Unexpected Death in Epilepsy

119

7.7â•…Bradycardia and Asystole Asystole and atrioventricular block are recognized as potentially malignant conduction abnormalities that can lead to ventricular ἀbrillation (Tigaran et al. 2002). Thus, bradyarrhythmia could be particularly risky in patients with coexistent, underlying ischemic heart diseases; this emphasizes the importance of recognizing these cases. Ictal bradycardia, cardiac asystole, and total atrioventricular block associated with epileptic seizures is to date a more seldom but noteworthy complication of epilepsy. This association is probably still underestimated because of the lack of recognition of cardiac complications in epilepsy, as mentioned before. Because of its implication in the pathophysiology of SUDEP, there were two earlier literature reviews on bradyarrhythmias and asystole as cardiac consequences of epileptic seizures (Devinsky et al. 1997; Tinuper et al. 2001). Several other case reports have been published since 2001 (Rocamora et al. 2003; Venugopalan et al. 2001; Mondon et al. 2002; Leutmezer et al. 2003; Toth et al. 2008; Carinci et al. 2007; Strzelczyk et al. 2008; Schuele et al. 2008; Ghearing et al. 2007; Almansori et al. 2006; Mascia et al. 2005; So and Sperling 2007; Rugg-Gunn et al. 2004a; Carvalho et al. 2004), adding valuable information about the occurrence of bradycardia in epilepsy. Of note, in this context, are recent studies from the Mayo Clinic (Britton et al. 2006), which showed no consistent hemispheric lateralization of seizure activity at the onset of bradycardia. Another study (Ghearing et al. 2007) found over a 14-year period that only 29 seizures were associated with ictal bradycardia in 13 patients. Out of the 29 bradycardia episodes, 7 patients had a total of 11 complex partial seizures that were associated with asystole. Despite the apparent multitude of published case reports, this association still seems to be infrequent. A feasible explanation of the paucity of data regarding the association between epilepsy and bradycardia may be an underestimation of this association due to the lack of adequate monitoring of the epilepsy patients, with concurrent electrocardiographic and intracranial electroencephalographic analysis (Leung et al. 2006). Important in this context, Rugg-Gunn et al. (2004b) developed a new monitoring strategy. Epilepsy patients were implanted with subcutaneous electrocardiogram loop recorders for an average of 18 months. Of the 19 patients, 4 (21%) developed bradycardia or asystole, for which a permanent pacemaker was deemed appropriate. Three of these episodes occurred at the time of a clinical seizure and one was not associated with a known clinical event. This study clearly showed how inadequate it is to simply record electroencephalograms and electrocardiograms for a few days or a few seizures. Thus, the ἀndings of Rugg-Gunn and colleagues (2004b) should be applauded. Other investigators should try to replicate these ἀndings in order to acquire more evidence-based data before we apply these ἀndings to the routine care of people with epilepsy (Hirsch and Hauser 2004).

7.8â•…Ictal Cardiac SPECT Imaging Supportive of a Cardiac Mechanism of SUDEP To the best of our knowledge, there is only one study illustrating focal myocardial perfusion defects concurrent with electrocardiographic ST-segment depression at the cessation of an epileptic seizure as an indicator for myocardial ischemia (see Figure 7.6) (Tigaran 2001; Dam et al. 2001). Notably, the criteria and the methods available for the detection of the myocardial ischemia with myocardial perfusion scintigraphy in this study were not

120 Sudden Death in Epilepsy: Forensic and Clinical Issues

Seizure

Rest

Perfusion defect during seizure

Patient 8

Figure 7.6╇ Perfusion defect during epileptic seizure.

used to determine if ischemia could occur globally in the myocardium. For future studies, the limitation of employing this type of imaging technique could be potentially addressed by the use of newer techniques, such as hybrid positron emission tomography/computerized tomography (PET/CT) or PET/magnetic resonance imaging systems, which would be able to detect more accurately any perfusion defects (Slomka et al. 2008). In data from an unpublished study conducted by the author of this chapter, 14 men and 9 women (age range 20–59 years, mean 42) with intractable focal epilepsy for 4 to 55 years (mean age 26 years) participated in this prospective study. All patients had normal baseline results on electrocardiography, Holter monitoring, echocardiography, myocardial scintigraphy at rest, and coronary angiography. Technetium 99m Cardiolite (DuPont Pharma, UK) was injected into 18 unmedicated patients during ictal events in which nine patients experienced secondarily generalized tonic–clonic seizures with a mean of maximum ictal heart rate of 133 beats/min. The other seven patients were injected in relation to complex partial seizures with an average maximal ictal heart rate of 108 beats/min. One patient received the tracer during a simple partial status epilepticus episode, during which the Holter recording was not available. Of the 18 myocardial perfusion studies, there were several ictal studies positive for focal myocardial perfusion defects (n = 3 or 16%). Unfortunately, only one was a positive ictal study in which the tracer was injected in direct relationship to a seizure where a 1-mm ST-segment depression, which is reflective of myocardial ischemia, was detected on the Holter monitor subsequent to a complex partial seizure without any motor phenomenon, recorded at 03:05 p.m. The maximum ictal heart rate during this seizure was 134 beats/min. This patient, a 56-year-old female with MRI-veriἀed right-sided hippocampus gliosis, had both infero- and anteroseptal reversible defects. This particular ἀnding may advocate for SUDEP being more frequent in women than was previously estimated (Walczak et al. 2001). The patient was moderately obese and had a normal 12-lead electrocardiogram, rest and stress test, and myocardial scintigraphy, as well as normal coronary angiography. Her baseline plasma norepinephrine was higher than normal, namely 1033 pg/mL. Her urine norepinephrine was 55,200 pg/mL

Sudden Unexpected Death in Epilepsy

121

or 135 μg per 24 h. Her baseline plasma epinephrine was also elevated above the normal reference interval, up to 87 pg/mL, and her urine norepinephrine was 13,800 pg/mL or 34€μg per 24€h (diuresis was 2450 mL for 24 h when the values were collected). Normal reference ranges are laboratory-speciἀc, vary according to methodology of testing, and differ between blood and urine samples. Supine (lying down): epinephrine less than 50 pg/ mL, norepinephrine less than 410 pg/mL; reference ranges for urine catecholamines are: epinephrine 0–20 μg per 24 h and norepinephrine 15–80 μg per 24 h (Answers.com 2009). The ἀndings in the above case concur with those of Nei et al. (2004), who suggests “. . . that patients with evidence of a great degree of change in autonomic tone during seizures might be at increased risk for SUDEP.” During the epileptic seizure, an almost threefold increase was noted in both plasma and urine catecholamines (plasma norepinephrine, 2891 pg/mL; plasma epinephrine, 296 pg/mL; urine norepinephrine, 121,300 pg/mL; and urine epinephrine, 29,700 pg/mL) for a total diuresis of 2400 mL. During the seizure, a drop in O2 saturation from 100% to 93% was recorded by the pulse oximeter, which did not represent any signiἀcant drop in the O2 saturation. The oxygen saturation dropped, but not below the limit of 90%. Most physicians would not be concerned when the oxygen saturation is above 90% (93% in this study). The imaging results from a small cohort of drug refractory epilepsy patients, whose antiepileptic drugs were withdrawn during the course of the study, and without any evidence of cardiovascular disease, substantiate the hypothesis that cardiac ischemic abnormalities may exist and could potentially provide some of the pathophysiologic explanation for SUDEP. The data also suggest that epileptic seizures of temporal origin may induce myocardial ischemia in the absence of coronary pathology, presumably by autonomicÂ�mediated vasospasm. This ἀnding points again toward an association between cardiac abnormalities in epilepsy and the sudden, large imbalance in cardiac efferent autonomic activity (Druschky et al. 2001) that may result in autonomic-mediated ischemia and, thus, possible fatal arrhythmia. Unfortunately, this study (Tigaran 2001) has several limitations. The tracer was simultaneously injected during the electrocardiographic detected myocardial ischemia as reflected by the ST-segment changes in one patient out of the 23 who were enrolled in the study. Studies related to the anatomical brain localization of the site of the seizure employing the subtraction ictal SPECT coregistered to MRI (SISCOM) method documented the importance of the timing for the tracer injection (O’Brien et al. 1998). Late injection of the radiotracer (>45 s) after the onset of the epileptic seizure was associated with a falsely localizing or nonlocalizing SISCOM study, a ἀnding that could be extrapolated to our epilepsy study. This might represent the explanation for why 15 of the 18 scans were negative myocardial perfusion studies.

7.9â•…Future Research Avenues for an Underlying Cardiac Mechanism of SUDEP: Epilepsy, Depression, and Cardiac Disease For centuries poets and folklore have asserted that there is a relationship between the mind and body in general, and human moods and the heart in particular. Almost 400 years ago Shakespeare wrote, “. . . My life being made of four, with two alone sinks down to death, oppress’d with melancholy . . . .” (Shakespeare). However, only in the past few years has this

122 Sudden Death in Epilepsy: Forensic and Clinical Issues

conviction been scientiἀcally tested (Glassman and Shapiro 1998). Indeed, it has also been proved that depression is the most frequent psychiatric disorder in patients with epilepsy and that it is more common in patients with partial seizure disorders of temporal or frontal lobe origin and among patients with poorly controlled seizures. In three communitybased studies, prevalence rates of depression ranged between 21% and 33% among patients with persistent seizures and between 4% and 6% among seizure-free patients (LaFrance et al. 2008). Ettinger et al. (2005) reported the results of a population-based survey that investigated a lifetime prevalence of depression, epilepsy, diabetes, and asthma in 185,000 households. Among the 2900 patients with epilepsy, 32% reported having experienced at least one episode of depression. This contrasted with an 8.6% prevalence among healthy respondents, 13% among patients with diabetes, and 16% among people with asthma. For a long time, people speculated about whether depression increases the risk of mortality only in individuals with established coronary disease, or if those without a history of heart disease are at increased risk as well. Glassman and Shapiro (1998) concluded that both groups are at risk, but the risk is higher in groups with established coronary artery disease. This conclusion was validated in a study by Penninx et al. (2001), which concluded that, regardless of the preexistence of coronary artery disease, the effects of depression result from the same mechanisms in both groups. Because sudden cardiac death accounts for most of the excess mortality in association with depression in patients with established coronary artery disease, one should search for proarrhythmic mechanisms associated with depression (Carney et al. 2001). Multiple possibilities exist. Heart rate variability is lower in depressed than in nondepressed patients with established coronary artery disease. In addition, plasma catecholamines, known provokers of arrhythmias and sudden cardiac death, are elevated in depressed patients (Carney et al. 2001). This is also the case in epilepsy patients (Tigaran et al. 2001; Lathers et al. 1984, 2008). Especially supportive of the altered autonomic tone in depression is the study of Sheline et al. (1996) who demonstrated that patients with a history of major depression had signiἀcantly smaller hippocampal volumes bilaterally when compared with matched control subjects. The decrement in hippocampal volume correlates with the cumulative lifetime duration of major depression, possibly as result of a progressive process mediated by glucocorticoid neurotoxicity (Carney et al. 2001). This process may be responsible for an increase in corticotropin-releasing factor secretory drive and may thereby contribute to the elevated hypothalamic–pituitary–adrenal axis activity observed in depression. Corticotropin-releasing factor is also a potent stimulus for sympathetic nervous system activation, which may account for the sympathetic hyperactivity observed in major depression (Carney et al. 1999, 2001) and in epilepsy patients (Nei et al. 2000a). In concert with the hippocampal ἀndings in depression, recent imaging studies in epilepsy patients revealed the existence of similar pathological ἀndings, namely those with right temporal lobe epilepsy and depression have hippocampal atrophy that cannot be explained by epilepsy alone (Shamim et al. 2009). In this context and supportive of this theory is the recent study of Bernhardt et al. (2009), which provides a unique quantitative assessment in patients with temporal lobe epilepsy. The study indicates that regions of the brain remote from the lobe from which the seizure originates may be adversely affected with signiἀcant changes in cerebral cortical structures that may be important in the development of comorbidities (Cascino 2009). Unfortunately, the sample size was too small to allow any conclusions regarding the left side hippocampus. Until now, there have been no studies looking into the association between the size of the hippocampal sclerosis, epilepsy, and depression in SUDEP patients. The ἀndings of

Sudden Unexpected Death in Epilepsy

123

this recent imaging study (Shamim et al. 2009) may open new research avenues for a better understanding of the SUDEP phenomenon, and a possible explanation of why not all epilepsy patients have an equal risk of dying of SUDEP. Finally, depression is associated with platelet activation and with inflammatory processes that may increase the risk of developing coronary artery disease or, in patients with established coronary artery disease or myocardial infarction (Carney et al. 2001), which might also be the case in epilepsy.

7.10â•…Evidence of Cardiac Arrhythmogenic Substrate for SUDEP: The Influence of Antiepileptic Drugs on the Heart Blumhardt et al. (1986) suggested that treated patients with epilepsy had lower mean rates of the ictal cardiac acceleration than the untreated patients, whereas Opherk et al. (2002) stated that antiepileptic drug regimes were mostly similar in seizure patients with and without cardiac abnormalities. However, due to the small number of patients studied, further conclusions concerning the possible protective effect of antiepileptic drugs were precluded. Others studies evaluated the role of heart rate, but did not specify the part played by antiepileptic drugs upon the heart rate. In contradistinction to the antiepileptic drug protective role theory, Devinsky et al. (1994) showed that patients with epilepsy have greater blood pressure and heart rate variability and reactivity than control patients, with those ἀndings partly attributable to carbamazepine. Conversely, by employing the technique of the electrocardiographic late potentials (see Figure 7.2) (Tigaran et al. 2002), it has been shown, with no preference for any of the antiepileptic drugs, that all may also influence the electrical properties of the myocytes, mostly through an inactivation of the sodium channels, thus potentially precipitating cardiac arrhythmias. The electrocardiographic late potential technique was shown to reveal the presence of diseased myocardium with delayed depolarization, which may serve as the substrate for reentrant arrhythmias causing ventricular tachycardia and sudden cardiac death (Tigaran et al. 2002). Further work is needed to determine whether techniques such as signal-averaged electrocardiography (SAECG) or other, newer techniques will be useful screening tools to identify whether some persons with epilepsy are at risk for SUDEP.

7.11â•…Summary and Clinical Perspectives The concept that abnormal electrical discharges in the brain trigger cardiac arrhythmias does not seem to be widely recognized by clinicians, although it has long been proposed in the neurological literature (Blumhardt et al. 1986; Mameli et al. 2006; Jallon 1997). Unfortunately, even now, electrocardiographic recordings are not always conducted or reviewed during epilepsy monitoring. Thus, the deἀnite relationship between epileptic seizures and cardiac abnormalities is still the subject of an ongoing debate. The diversity of the methods employed in the assessment of this relationship, in contrast to the use of only a single electrocardiogram rhythm monitoring channel that is usually a supplement to the EEG recordings, along with the infrequent use of simultaneous long-term EEG-Holter recordings, presents another limitation of this type of investigation.

124 Sudden Death in Epilepsy: Forensic and Clinical Issues

Of note is the consistent lack of respiratory recordings, which need to be conducted during assessments related to electroencephalographic investigations before epilepsy surgery or admittance for differential diagnosis purposes, since apnea and hypoxia clearly represent a major contributory mechanism to the occurrence of cardiac events (Schulz et al. 2006). Nonetheless, the variable results from studies seeking a relationship between epilepsy and the risk of potentially fatal cardiac events, as discussed in this chapter, support the need for different animal models (Lathers et al. 2008, 2010, Chapter 28) that would provide comparison with clinical study data from patients with epilepsy. Also, studies addressing the association between epilepsy, depression, and cardiac abnormalities and SUDEP are deemed to be of future research importance. We need to encourage cardiologists to exercise caution when concluding from Holter monitoring alone, especially in the setting of ambulatory recordings that presumed cardiogenic arrhythmias, even when they coincide with symptoms, are the primary cause of any patient’s complaints. Only prolonged monitoring, deἀned as at least 24 h of recordings of simultaneous EEG and ECG records, will improve the diagnostic yield and, consequently, optimize the treatment of the patients with seizure-related arrhythmias and also of persons with cardiogenic arrhythmias that can result in seizure like activity. Moreover, besides the determination of the genetic etiology of epilepsy and association with depression, it would be useful to conduct collaborative studies aimed at evaluating both the epileptogenic and the cardiologic aspect of the underlying substrate that may lead to increasing the chance of potentially fatal events. Only by making both doctors and patients aware of the presence of the potential for adverse symbiosis between neurological and cardiac abnormalities will the quality of preventive services, especially as relates to SUDEP, be improved. Only time, education, and interdisciplinary joint efforts will determine whether well-conducted research will be feasible in large population samples.

Acknowledgment For valuable criticism when reviewing this book chapter, I have to thank Allan S. Jaffe, MD, Mayo Clinic, Rochester, MN.

References Allison, D. B., K. R. Fontaine, J. E. Manson, J. Stevens, and T. B. VanItallie. 1999. Annual deaths attributable to obesity in the United States. JAMA 282 (16): 1530–1538. Almansori, M., M. Ijaz, and S. N. Ahmed. 2006. Cerebral arrhythmia influencing cardiac rhythm: a case of ictal bradycardia. Seizure 15 (6): 459–461. Anderson, J. L., E. V. Platia, A. Hallstrom, R. W. Henthorn, T. A. Buckingham, M. D. Carlson, and P. E. Carson. 1994. Interaction of baseline characteristics with the hazard of encainide, flecainide, and moricizine therapy in patients with myocardial infarction. A possible explanation for increased mortality in the Cardiac Arrhythmia Suppression Trial (CAST). Circulation 90 (6): 2843–2852. Answers.com. 2009. Catecholamines Tests: Normal Results. Available from http://www.answers╉ .com/catecholamines-tests-normal-results (accessed February 2009).

Sudden Unexpected Death in Epilepsy

125

Aung-Din, R., J. H. Mitchell, and J. C. Longhurst. 1981. Reflex alpha-adrenergic coronary vasoconstriction during hindlimb static exercise in dogs. Circ Res 48 (4): 502–509. Begun, J. W., J. Tornabeni, and K. R. White. 2006. Opportunities for improving patient care through lateral integration: the clinical nurse leader. J Healthc Manag 51 (1): 19–25. Beran, R. G. 2006. SUDEP—To discuss or not discuss: That is the question. Lancet Neurol 5 (6): 464–465. Bernhardt, B. C., K. J. Worsley, H. Kim, A. C. Evans, A. Bernasconi, and N. Bernasconi. 2009. Longitudinal and cross-sectional analysis of atrophy in pharmacoresistant temporal lobe epilepsy. Neurology 72 (20): 1747–1754. Blumhardt, L. D., P. E. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1 (8489): 1051–1056. Brazier, J. R., and G. D. Buckberg. 1975. Effects of tachycardia on the adequacy of subendocardial oxygen delivery in experimental aortic stenosis. Am Heart J 90 (2): 222–230. Britton, J. W., G. R. Ghearing, E. E. Benarroch, and G. D. Cascino. 2006. The ictal bradycardia syndrome: localization and lateralization. Epilepsia 47 (4): 737–744. Carinci, V., G. Barbato, A. Baldrati, and G. Di Pasquale. 2007. Asystole induced by partial seizures: a rare cause of syncope. Pacing Clin Electrophysiol 30 (11): 1416–1419. Carney, R. M., K. E. Freedland, and A. S. Jaffe. 2001. Depression as a risk factor for coronary heart disease mortality. Arch Gen Psychiatry 58 (3): 229–230. Carney, R. M., K. E. Freedland, R. C. Veith, and A. S. Jaffe. 1999. Can treating depression reduce mortality after an acute myocardial infarction? Psychosom Med 61 (5): 666–675. Carvalho, K. S., V. Salanova, and O. N. Markand. 2004. Cardiac asystole during a temporal lobe seizure. Seizure 13 (8): 595–599. Cascino, G. D. 2009. Temporal lobe epilepsy is a progressive neurologic disorder: Time means neurons! Neurology 72 (20): 1718–1719. Chin, P. S., K. R. Branch, and K. J. Becker. 2004. Myocardial infarction following brief convulsive seizures. Neurology 63 (12): 2453–2454. Christiansen, E. H., L. Frost, H. Mlgaard, P. E. Thomsen, T. T. Nielsen, and A. K. Pedersen. 1995. The signal-averaged ECG becomes late potential-positive at low noise levels in healthy subjects. Eur Heart J 16 (11): 1731–1735. Dam, M., S. Tigaran, K. Rasmussen, M. Rehling, and H. Molgaard. 2001. Sudden unexpected death in epilepsy (SUDEP): Regional myocardial blood flow during seizure. Neurology 56 (S-3): A233. DeGiorgio, C. M., P. Miller, S. Meymandi, and J. A. Gornbein. 2008. n-3 fatty acids (ἀsh oil) for epilepsy, cardiac risk factors, and risk of SUDEP: Clues from a pilot, double-blind, exploratory study. Epilepsy Behav 13 (4): 681–684. Devinsky, O., K. Perrine, and W. H. Theodore. 1994. Interictal autonomic nervous system function in patients with epilepsy. Epilepsia 35 (1): 199–204. Devinsky, O., S. Pacia, and G. Tatambhotla. 1997. Bradycardia and asystole induced by partial seizures: a case report and literature review. Neurology 48 (6): 1712–1714. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Duncan, H. W., R. J. Barnard, G. K. Grimditch, J. Vinten-Johansen, and G. D. Buckberg. 1987. Cardiovascular response to sudden strenuous exercise. Basic Res Cardiol 82 (3): 226–232. El Sherif, N. 1994. Re-entrant mechanism in venticular arrhythmias. In Cardiac Electrophysiology: From Cell to Bedside, ed. D. Zipes and J. Jalife. Philadelphia, PA: W. B. Saunders. Espinosa, P. S., J. W. Lee, U. B. Tedrow, E. B. Bromἀeld, and B. A. Dworetzky. 2009. Sudden unexpected near death in epilepsy: Malignant arrhythmia from a partial seizure. Neurology 72 (19): 1702–1703. Ettinger, A. B., M. L. Reed, J. F. Goldberg, and R. M. Hirschfeld. 2005. Prevalence of bipolar symptoms in epilepsy vs other chronic health disorders. Neurology 65 (4): 535–540.

126 Sudden Death in Epilepsy: Forensic and Clinical Issues Falconer, B., and J. Rajs. 1976. Post-mortem ἀndings of cardiac lesions in epileptics: A preliminary report. Forensic Sci 8 (1): 63–71. Flegal, K. M., M. D. Carroll, R. J. Kuczmarski, and C. L. Johnson. 1998. Overweight and obesity in the United States: Prevalence and trends, 1960–1994. Int J Obes Relat Metab Disord 22 (1): 39–47. Foster, C., J. D. Anholm, C. K. Hellman, J. Carpenter, M. L. Pollock, and D. H. Schmidt. 1981. Left ventricular function during sudden strenuous exercise. Circulation 63 (3): 592–596. Foster, C., D. S. Dymond, J. Carpenter, and D. H. Schmidt. 1982. Effect of warm-up on left ventricular response to sudden strenuous exercise. J Appl Physiol 53 (2): 380–383. Franzosi, M. G. 2006. Should we continue to use BMI as a cardiovascular risk factor? Lancet 368 (9536): 624–625. Freedman, R. A., and J. S. Steinberg. 1991. Selective prolongation of QRS late potentials by sodium channel blocking antiarrhythmic drugs: Relation to slowing of ventricular tachycardia. Electrophysiologic Study Versus Electrocardiographic Monitoring Trial (ESVEM) Investigators. J Am Coll Cardiol 17 (5): 1017–1025. Galimberti, C. A., E. Marchioni, F. Barzizza, R. Manni, I. Sartori, and A. Tartara. 1996. Partial epileptic seizures of different origin variably affect cardiac rhythm. Epilepsia 37 (8): 742–747. Garcia, M., C. D’Giano, S. Estelles, R. Leiguarda, and A. Rabinowicz. 2001. Ictal tachycardia: Its discriminating potential between temporal and extratemporal seizure foci. Seizure 10 (6): 415–419. Ghearing, G. R., T. M. Munger, A. S. Jaffe, E. E. Benarroch, and J. W. Britton. 2007. Clinical cues for detecting ictal asystole. Clin Auton Res 17 (4): 221–226. Glassman, A. H., and P. A. Shapiro. 1998. Depression and the course of coronary artery disease. Am J Psychiatry 155 (1): 4–11. Hanna, J. 2007. Epilepsy and sudden death: A personal view. Epilepsia 38 (s11): S3–S5, doi: 10.1111/ j.1528-1157.1997.tb06121.x. Harris, C. D. 2005. Neurophysiology of sleep and wakefulness. Respir Care Clin N Am 11 (4): 567–586. Heart, Cambridge. 2009. T-Wave Alternans. Available from http://www.cambridgeheart.com. Hiestand, D. M., P. Britz, M. Goldman, and B. Phillips. 2006. Prevalence of symptoms and risk of sleep apnea in the US population: Results from the national sleep foundation sleep in America 2005 poll. Chest 130 (3): 780–786. Hirsch, L. J., and W. A. Hauser. 2004. Can sudden unexplained death in epilepsy be prevented? Lancet 364 (9452): 2157–2158. Ho, R. T., T. Wicks, D. Wyeth, and M. Nei. 2006. Generalized tonic–clonic seizures detected by implantable loop recorder devices: Diagnosing more than cardiac arrhythmias. Heart Rhythm 3 (7): 857–861. Hoffman, J. I., and G. D. Buckberg. 1975. Pathophysiology of subendocardial ischaemia. Br Med J 1€(5949): 76–79. Jallon, P. 1997. Epilepsy and the heart. Rev Neurol (Paris) 153 (3): 173–184. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37 (4): 531–537. Katz, L. N., and H. Feinberg. 1958. The relation of cardiac effort to myocardial oxygen consumption and coronary flow. Circ Res 6 (5): 656–669. Khouri, E. M., D. E. Gregg, and C. R. Rayford. 1965. Effect of exercise on cardiac output, left coronary flow and myocardial metabolism in the unanesthetized dog. Circ Res 17 (5): 427–437. LaFrance, W. C., Jr., A. M. Kanner, and B. Hermann. 2008. Psychiatric comorbidities in epilepsy. Int Rev Neurobiol 83: 347–383. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936.

Sudden Unexpected Death in Epilepsy

127

Lathers, C. M., and P. L. Schraeder. 2010. SUDEP animal models. Ch. 28 in Sudden Death in Epilepsy: Forensic and Clinical Issues. Boca Raton: CRC Press. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9 (1): 19–30. Leutmezer, F., C. Schernthaner, S. Lurger, K. Potzelberger, and C. Baumgartner. 2003. ElectroÂ� cardiographic changes at the onset of epileptic seizures. Epilepsia 44 (3): 348–354. Mameli, O., M. A. Caria, A. Pintus, G. Padua, and S. Mameli. 2006. Sudden death in epilepsy: an experimental animal model. Seizure 15 (5): 275–287. Manno, E. M., E. A. Pfeifer, G. D. Cascino, K. H. Noe, and E. F. Wijdicks. 2005. Cardiac pathology in status epilepticus. Ann Neurol 58 (6): 954–957. Marshall, D. W., B. F. Westmoreland, and F. W. Sharbrough. 1983. Ictal tachycardia during temporal lobe seizures. Mayo Clin Proc 58 (7): 443–446. Mascia, A., P. P. Quarato, A. Sparano, V. Esposito, F. Sebastiano, G. Occhiogrosso, and G. Di Gennaro. 2005. Cardiac asystole during right frontal lobe seizures: A case report. Neurol Sci 26 (5): 340–343. Mondon, K., B. Charbonnier, C. Hommet, P. Corcia, A. Autret, and B. de Toffol. 2002. Ictal bradycardia followed by cardiac asystole: A case report. Epileptic Disord 4 (4): 261–264. Murray, P. A., and S. F. Vatner. 1981. Carotid sinus baroreceptor control of right coronary circulation in normal, hypertrophied, and failing right ventricles of conscious dogs. Circ Res 49 (6): 1339–1349. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55 (6): 857–860. Nei, M., R. T. Ho, M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia May;41(5): 542–548. Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. O’Brien, T. J., E. L. So, B. P. Mullan, M. F. Hauser, B. H. Brinkmann, N. I. Bohnen, D. Hanson, G. D. Cascino, C. R. Jack Jr., and F. W. Sharbrough. 1998. Subtraction ictal SPECT co-registered to MRI improves clinical usefulness of SPECT in localizing the surgical seizure focus. Neurology 50 (2): 445–454. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52 (2): 117–127. P-Codrea Tigaran, S., S. Dalager-Pedersen, U. Baandrup, M. Dam, and A. Vesterby-Charles. 2005. Sudden unexpected death in epilepsy: Is death by seizures a cardiac disease? Am J Forensic Med Pathol 26 (2): 99–105. Pedley, T. A., and W. A. Hauser. 2002. Sudden death in epilepsy: A wake-up call for management. Lancet 359 (9320): 1790–1791. Penninx, B. W., A. T. Beekman, A. Honig, D. J. Deeg, R. A. Schoevers, J. T. van Eijk, and W. van Tilburg. 2001. Depression and cardiac mortality: Results from a community-based longitudinal study. Arch Gen Psychiatry 58 (3): 221–227. Rocamora, R., M. Kurthen, L. Lickfett, J. Von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44 (2): 179–185. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004a. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004b. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Sanya, E. O. 2005. Increasing awareness about sudden unexplained death in epilepsy—A review. Afr J Med Med Sci 34 (4): 323–327.

128 Sudden Death in Epilepsy: Forensic and Clinical Issues Schernthaner, C., G. Lindinger, K. Potzelberger, K. Zeiler, and C. Baumgartner. 1999. Autonomic epilepsy—The influence of epileptic discharges on heart rate and rhythm. Wien Klin Wochenschr 111 (10): 392–401. Schuele, S. U., A. C. Bermeo, E. Locatelli, R. C. Burgess, and H. O. Luders. 2008. Ictal asystole: A benign condition? Epilepsia 49 (1): 168–171. Schulz, R., M. Grebe, H. J. Eisele, K. Mayer, N. Weissmann, and W. Seeger. 2006. Obstructive sleep apnea-related cardiovascular disease. Med Klin (Munich) 101 (4): 321–327. Scorza, F. A., D. B. Colugnati, A. P. Pansani, E. Y. Sonoda, R. M. Arida, and E. A. Cavalheiro. 2008. Preventing tomorrow’s sudden cardiac death in epilepsy today: what should physicians know about this? Clinics (Sao Paulo) 63 (3): 389–94 LID-S1807-593220. Shakespeare, W. Sonnet 45. Available from http://www.shakespeare-online.com/sonnets/45.html. Shamim, S., G. Hasler, C. Liew, S. Sato, and W. H. Theodore. 2009. Temporal lobe epilepsy, depression, and hippocampal volume. Epilepsia 50 (5): 1067–1071. Sheline, Y. I., P. W. Wang, M. H. Gado, J. G. Csernansky, and M. W. Vannier. 1996. Hippocampal atrophy in recurrent major depression. Proc Natl Acad Sci U S A 93 (9): 3908–3913. Shibasaki, H., A. Ikeda, and T. Nagamine. 2007. Use of magnetoencephalography in the presurgical evaluation of epilepsy patients. Clin Neurophysiol 118 (7): 1438–1448. Slomka, P. J., D. S. Berman, and G. Germano. 2008. Applications and software techniques for integrated cardiac multimodality imaging. Expert Rev Cardiovasc Ther 6 (1): 27–41. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: evidence from near-SUDEP incident. Epilepsia 41 (11): 1494–1497. So, N. K., and M. R. Sperling. 2007. Ictal asystole and SUDEP. Neurology 69 (5): 423–424. Spratling, W. P. 1902. The cause and manner of death in epilepsy. Med News 80: 1225–1227. Strzelczyk, A., S. Bauer, S. Knake, W. H. Oertel, H. M. Hamer, and F. Rosenow. 2008. Ictal asystole in temporal lobe epilepsy before and after pacemaker implantation. Epileptic Disord 10 (1): 39–44. Ten Tusscher, K. H., and A. V. Panἀlov. 2007. Influence of diffuse ἀbrosis on wave propagation in human ventricular tissue. Europace 9 (Suppl 6): vi38–vi45. Tigaran, S. 2001. Regional myocardial blood flow during seizures. Paper read at the American Academy of Neurology Annual Meeting in Philadelphia, on May 9, 2001 during session no. 32. Tigaran, S. 2002. Cardiac abnormalities in patients with refractory epilepsy. Acta Neurol Scand Suppl 177: 9–32. Tigaran, S., F. Andreasen, and M. Dam. 2001. Changes in plasma catecholamines after epileptic seizures. Possible explanation of SUDEP. Epilepsia 42 (S-2): 158. Tigaran, S., H. Molgaard, and M. Dam. 1999. Detection of myocardial ἀbrosis by ECG technique in epilepsy patients—Further evidence for cardiac pathogenesis of SUDEP. Epilepsia 40 (S-7): 105. Tigaran, S., H. Molgaard, and M. Dam. 2002. Atrio-ventricular block: A possible explanation of sudden unexpected death in epilepsy. Acta Neurol Scand 106 (4): 229–233. Tigaran, S., H. Molgaard, R. McClelland, M. Dam, and A. S. Jaffe. 2003. Evidence of cardiac ischemia during seizures in drug refractory epilepsy patients. Neurology 60 (3): 492–495. Tinuper, P., F. Bisulli, A. Cerullo, R. Carcangiu, C. Marini, G. Pierangeli, and P. Cortelli. 2001. Ictal bradycardia in partial epileptic seizures: Autonomic investigation in three cases and literature review. Brain 124 (Pt 12): 2361–2371. Toth, V., L. Hejjel, Z. Kalmar, A. Fogarasi, T. Auer, C. Gyimesi, A. Szucs, and J. Janszky. 2008. Effect of epileptic seizures on the heart rate. Ideggyogy Sz 61 (5–6): 155–161. Ushijima, A., N. Fukuma, Y. Kato, N. Aisu, and K. Mizuno. 2009. Sympathetic excitation during exercise as a cause of attenuated heart rate recovery in patients with myocardial infarction. J€Nippon€Med Sch 76 (2): 76–83. Velez, L., and L. M. Selwa. 2003. Seizure disorders in the elderly. Am Fam Physician 67 (2): 325–332. Venugopalan, P., P. M. Nair, and R. L. Koul. 2001. An infant with seizure-related bradycardia and asystole. J Paediatr Child Health 37 (1): 96–97.

Sudden Unexpected Death in Epilepsy

129

Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. D. Cascino, J. F. Annegers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56 (4): 519–525. WHO. 2009. Epilepsy. 2009/01/01. Available from http://www.who.int/mediacentre/factsheets/ fs999/en/index.html (accessed February 2009). Wilder-Smith, E., and S. H. Lim. 2001. Heart rate changes during partial seizures: A study amongst Singaporean patients. BMC Neurol 1: 5. Willams, B., and A. Richens. 1998. Epilepsy and sudden death. UK winter 1998 Epilepsy Toronto: EA updates. London: Epilepsy Research Foundation. WrongDiagnosis.com. 2009. Deaths from obesity. Available from http://www.wrongdiagnosis╉ .com/o/obesity/deaths.htm. Yasue, H. 1983. Role of autonomic nervous system in the pathogenesis of angina pectoris. Arch Mal Coeur Vaiss 76 Spec No: 3–6. Zijlmans, M., D. Flanagan, and J. Gotman. 2002. Heart rate changes and ECG abnormalities during epileptic seizures: Prevalence and deἀnition of an objective clinical sign. Epilepsia 43 (8): 847–854.

Forensic Postmortem Examination of Victims of Sudden Unexpected Death in Epilepsy

8

Claire M. Lathers Paul L. Schraeder Steven A. Koehler Cyril H. Wecht

Contents 8.1 Introduction 8.2 Hypothesis and Future Retrospective Studies 8.3 Speciἀc Aims 8.3.1 Year One 8.3.2 Year Two 8.3.3 Year Three 8.4 Questions for Retroprospective Postmortem Studies References

131 135 135 136 136 136 137 139

8.1â•…Introduction Evaluation of all potential cardiac risks for sudden death needs to start with detailed cardiac postmortem examinations. Detailed postmortem examinations of hearts obtained from patients with epilepsy who experienced sudden and unexpected death (SUDEP) will be helpful in identifying associated cardiac risk factors in persons with epilepsy, and also family members that could be predisposed to sudden death. A proἀle of representative pathology associated with sudden death is found in Table 8.1. Research work in the areas of SUDEP suggests that changes in autonomic peripheral and central mechanisms involved in the control and regulation of blood pressure, heart rate, and rhythm may contribute to sudden death. More speciἀcally, changes in postganglionic cardiac sympathetic discharges, which result in nonuniform neural discharges associated with epileptiform discharges, may be part of the mechanism of risk. Another major risk factor seems to be noncompliance with antiepileptic drug use. A national survey of coroners and medical examiners throughout the United States was conducted to determine their depth of understanding of causes of death in persons with epilepsy and, in particular, the phenomenon of sudden death (Schraeder et al. 2006, 2009). The survey concluded that, while there is a valid classiἀcation of SUDEP, it is not routinely used in autopsy reports of patients with known seizures who die suddenly and unexpectedly with no cause of death found on postmortem examination. Instead, the autopsy reports emphasized the negative ἀndings, i.e., that the death might have been 131

132 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 8.1â•… Pathological Findings in Victims of Sudden Death Victims of Sudden Death Victims of stress-related death (Selye 1958) 51 SUDEP cases non-East Asian subjects including indigenous Saudis, 1/1995 to 6/1995; most subcontinent Indians (43%) (Elfawal 2000) Autopsies of sudden death cases in Japan, 5/1994 to 2/1998 Of 271 cases, 176 patients (20 to 59 years old) were sudden death (Owada et al. 1999) 200 cases of sudden death in persons younger than 35 years in Veneto region, Italy – Unexplained deaths occurred in 12 cases (6%) – More than 1 in 20 cases in sudden death not explained by structural risk factors (Basso et al. 1999)

Cardiac Pathology Microscopic myoἀbrillar degeneration or myocytolysis Identical in hearts of patients dying subarachnoid hemorÂ� rhage and other acute strokes Autopsies on 22 victims 7 mild to moderate cardiac hypertrophy and 2 mild to moderate coronary stenosis 4 similar degrees of coronary narrowing but no myocardial hypertorphy Of the sudden death cases, 29 (31.9%) were due to coronary artery disease 18 (19.8%) acute cardiac dysfunction 6 (6.6%) other cardiac diseases 4 (4.4%) acute aortic dissection 163 cases (81.5%) due to cardiovascular etiologies

Obstructive coronary atherosclerosis, 23%; arrhythmogenic right ventricular cardiomyopathy, 12.5%; mitral valve prolapse 10%; conduction system abnormalities, 10%; congenital coronary artery anomalies, 8.5%; myo�carditis, 7.5%; hyper�trophic cardiomyopathy 5.5%; aortic rupture, 5.5%; dilated cardiomyopathy, 5%; nonatherosclerotic-acquired coronary artery disease, 3.5%; postoperative congenial heart disease, 13%; aortic stenosis, 2%

Pulmonary Pathology

Cerebral Pathology

Other

18 of 22 with severe pulmonary congestion and alveolar hemorrhage

10 cases (5%) due to respiratory causes

4 (4.4%) due to cardiovascular disease

30 (32.9%) due to other diseases

15 cases (7.5%) due to cerebral causes

Other causes (2%)

Pulmonary embolism, 2%

caused by a seizure disorder when there was no pathological ἀnding. Two important questions remain: 1. How do we determine if the death of a person is a sudden unexpected death? 2. How do we establish a true prevalence of the phenomenon of SUDEP?

Forensic Postmortem Examination of Victims of SUDEP

133

In addition to the need for establishing the true prevalence of the phenomenon of SUDEP is the importance of looking for evidence of the roles of the brain and the heart in the cause of death when conducting an autopsy on a person with epilepsy who has died suddenly. SUDEP refers to sudden death of an individual with a clinical history of epilepsy, in whom a postmortem examination fails to uncover a gross anatomic, toxicological, or environmental cause of death. Evidence of terminal seizure activity may not be present. In 2002, Shields et al. reported that 1–2% of natural deaths certified by the medical–legal death investigators in the United States are attributed to epilepsy and that increased microscopic examination of the brain postmortem has allowed identification of structural changes representative of epileptogenic foci. They examined 70 death cases, all with known clinical history of seizures, and classified them as: 1. Individuals who lacked a gross brain lesion 2. Those with a brain lesion demonstrable at autopsy 3. Those who lacked neuropathological evaluation because of decomposition so that only an external examination was done. Microscopic ἀndings include neuronal clusters, increased perivascular oligodendrogÂ� lia, gliosis, cystic gliotic lesions, decreased myelin, cerebellar Bergmann’s gliosis, and folial atrophy were found to be present in a higher percentage of the brains of SUDEP victims, when compared to brains from age- and sex-matched control subjects. Additional autopsy results are needed to clarify the role of changes in the heart in SUDEP. Insight into the mechanism of death in persons with epilepsy who die unexpectedly is not much greater than what it was when the ἀrst book (Lathers and Schraeder 1990b) addressed the topic of epilepsy and sudden death in 1990. This chapter emphasizes the need for clinical studies to focus on the contribution of cardiac autonomic dysfunction and/or antiepileptic drug use, with or without other drugs, to the development of cardiac abnormalities to gain an understanding of the mechanism of death in these persons. While most investigators agree that disturbance in the function of the autonomic nervous system may be a contributory cause (Han and Moe 1964; Randall et al. 1968; Lathers 1975; Lathers 1980a, 1980b; Lathers and Roberts 1985; Lathers and Spivey 1987; Spivey and Lathers 1985; Lathers et al. 1977a, 1977b, 1978; Lathers and Schraeder 1982; Lathers and Roberts 1985; Lathers et al. 1986; Lipka et al. 1988), how disruption of autonomic function contributes to the risk of death is not known. There are some clues from clinical observations. For example, the occurrence of seizures is often associated with measurable changes in cardiac conduction and rhythm. Two recent papers are relevant. The ἀrst, by Chin et al. (2004), reports myocardial infarction following brief convulsive seizures. The second paper, also by Chin et al. (2005), describes the occurrence of “postictal neurogenic stunned myocardium,” resulting from the consequence of seizures. A reversible multifocal ventricular dysfunction, developed in a nonvascular pattern, is thought to be the result of a high sympathetic tone. The initiating factor(s) for a fatal outcome in some at-risk individuals has not been deἀned. Experimental evidence indicates that cortical epileptiform activity can affect cardiac autonomic regulatory function (Lathers and Schraeder 1982; Lathers et al. 1987; Lathers and Schraeder 1990a, 1995, 2002; Lathers et al. 2003a; Lathers and Schraeder 1983, 1989, 1995, 2006, 2009; Lathers et al. 1984, 1988,

134 Sudden Death in Epilepsy: Forensic and Clinical Issues

1989, 1993, 2008a; Scorza et al. 2008; Carnel et al. 1985; Kraras et al. 1987; Stauffer et al. 1989, 1990; Suter and Lathers 1984; Tumer et al. 1985) and the integrity of pulmonary vasculature function (Simon et al. 1982; Johnston et al. 1995, 1997). In young adults with epilepsy, there is an increased risk for sudden death, often, but not always, as a function of the severity of their seizure disorder. Why there is an increased risk of presumed cardiac arrhythmia in this group is not known. Many investigators assume that death in these patients is a function of neurogenically induced cardiac arrhythmias since, in most cases, no gross pathological explanations are found on postmortem examination. However, there is a small body of literature that suggests that microscopic changes in the subendocardial region of the heart may be a contributory factor to increased risk of death in this population. Natelson et al. (1998) found irreversible pathological changes in the form of subendocardial perivascular and interstitial ἀbrosis in four of seven hearts from persons with epilepsy who died suddenly and unexpectedly. These ἀndings support the premise that patients with epilepsy who die suddenly and unexpectedly have subtle microscopic cardiac pathological conditions that may be responsible for increased risk of neurogenically induced arrhythmias. A series of questions must be addressed to focus on the issue of whether drugs are a beneἀt or a risk for SUDEP (Lathers 2002, 2003; Lathers and Schraeder 1995, 2002; Lathers et al. 2003, 2008a, 2008b; Scorza et al. 2008; Leestma et al. 1997; Schraeder and Lathers 1995). The ἀrst question is: “Do cardiac changes induced by antiepileptic drugs contribute to the risk of sudden death in persons with epilepsy?” A second question is: “Are particular combinations of antiepileptic drugs more likely to be associated with an increased risk of sudden death?” A third question to be addressed is: “Does a combination of antiepileptic drugs with other drugs, i.e., psychotropics, increase the risk of dying in a sudden, unexpected manner?” The FDA recognizes the importance of addressing this series of questions related to the use of pharmacological agents in that most of the new antiepileptic drugs have to be evaluated for the risk of SUDEP. As such, the FDA requires a statement in the package insert addressing the relative risk of occurrence of SUDEP with use of the drug. It is important for postmortem protocols to address the details of types of drugs and dosage schedules and possible adverse interactions as contributing to SUDEP. Speculation as to risk factors for SUDEP and/or contributory causes of SUDEP include the following:

1. Autonomic neurohumoral dysfunction 2. Autonomic cardiac neural dysfunction 3. Cardiac changes of subendocardial perivascular and interstitial ἀbrosis 4. Other cardiac changes (e.g., coronary heart disease, cardiomyopathy, aortic valvular stenosis, right ventricular dysplasia, postictal neurogenic stunned myocardium, coronary artery thrombosis, and postmyocardial infarction ἀbrosis 5. Cardiac changes possibly induced by antiepileptic drugs 6. The likelihood of cardiac changes being induced by particular combinations of antiepileptic drugs and associated with an increased risk of SUDEP 7. Whether a combination of antiepileptic drugs and nonantiepileptic drugs are more likely to increase the risk of dying in SUDEP (nonantiepileptic drugs are deἀned as therapeutic drugs and do include drugs of abuse)

Forensic Postmortem Examination of Victims of SUDEP

135

8. Whether an inherited predisposition to increased risk of cardiac arrhythmias (e.g., long QT and Brugada syndromes) could also be an additional risk factor in persons with epilepsy

8.2â•…Hypothesis and Future Retrospective Studies Autonomic neurohumoral and autonomic cardiac neural dysfunction, in combination with susceptibility associated with cardiac pathological changes, are hypothesized to be risk factors for the development of circumstances that predispose a person with epilepsy to sudden death. The potential risk factors of antiepileptic drug use versus nonantiepileptic drug use also must be examined. Study design should evaluate and differentiate the contribution of intrinsic parameters of autonomic dysfunction and/or cardiac pathological changes from the extrinsic parameter of therapeutic pharmacological agents (antiepileptic drugs and/or nonantiepileptic drugs). Microscopic pathological cardiac markers manifest in the heart in association with autonomic catecholamine-mediated arrhythmias and cardiac damage and/or by the administration of drugs (antiepileptic drugs and nonantiepileptic drugs should be examined). It could be assumed that catecholamine toxicity at the local level in the myocardium is a contributory factor to the development of microἀbrotic changes thought to be a predisposing risk factor for cardiac arrhythmias. Another contributory factor to cardiac arrhythmias and/or sudden death is abnormal cardiac postganglionic discharge. It is theorized that the combination of nonuniform postganglionic cardiac discharge due to abnormal anatomy of the postganglionic nerves in combination with catecholamine toxicity at the local level in the myocardium, possibly in combination with a genetically determined cardiac predisposition, are risk factors that combine to contribute to arrhythmias culminating in a fatal sudden death event. It is unclear whether or not the use of certain antiepileptic drugs such as carbamazepine compound diminish the physiologic risk factors for SUDEP (Devinsky 2004). The role of genetics in epilepsy needs to be considered when evaluating risk factors for SUDEP and/or the role of sodium channel dysfunction. The reader is referred to the following references and chapters discussing the genetics of epilepsy in this book: Chioza et al. (2001, 2002a, 2002b, 2002c), Sisodiya et al. (2007), Hindocha et al. (2008), Helbig et al. (2009), Mullen and Scheffer (2009), Herreros (2010), Ghali and Nashef (2010), and Lathers et al. (2010 Chapter 1).

8.3â•…Specific Aims Study designs should examine the potential contributory role of the following parameters to the risk of SUDEP to determine if: 1. Consistent microscopic, cardiac changes of subendocardial, perivascular, and interstitial ἀbrosis are present and may have contributed to the SUDEP. 2. Changes in postganglionic cardiac sympathetic nerves may be present. 3. Cardiac changes are induced by speciἀc antiepileptic drugs that could contribute to the risk of SUDEP.

136 Sudden Death in Epilepsy: Forensic and Clinical Issues

4. Particular combinations of antiepileptic drugs are associated with an increased risk of SUDEP. 5. A combination of antiepileptic and nonantiepileptic drugs is likely to increase the risk of SUDEP. Nonantiepileptic drugs are deἀned as therapeutic agents and will not include drugs of abuse. 6. Controls for the retrospective and prospective arms of the studies will consist of matched persons without a history of epilepsy who have died. See the chapter by Ghali and Nashef (2010) for a discussion of what genetic issues should be considered. To determine the validity of the proposed hypothesis that autonomic neurohumoral and autonomic neural dysfunction, in combination with susceptibility to cardiac pathological changes and/or the use of antiepileptic or nonantiepileptic drugs, are risk factors for the development of circumstances that predispose a person with epilepsy to sudden death, the following timeline and methodology could be used by those designing retrospective postmortem studies. 8.3.1â•…Year One During the ἀrst year, study should be devoted to analyzing retrospectively the frozen samples of heart tissue obtained from victims of SUDEP in order to determine the optimal parameters of analysis of samples obtained from the prospective arm of the study. Collection of prospectively obtained samples of cardiac tissue and preparations for postmortem examination of cardiac postganglionic innervations, using the method of Druschky et al. (2001), in persons who have died suddenly and unexpectedly should be initiated during year one. Likewise, methods should be utilized to determine the presence or absence of genetic defects at the Na+ channel level (Mullen and Scheffer 2009; Herreros 2010; Lathers and Schraeder 2010). 8.3.2â•…Year Two During the second year, the retrospective analysis should be completed and the prospective acquisition of samples should be completed. 8.3.3â•…Year Three During the third year, analysis of the pathological materials should be completed. The following analyses should be done: • Microscopic histopathology of the heart, especially the conduction system and subendocardial tissue. • Transmission electron microscopy, scanning transmission electron microscopy, scanning electron microscopy, and laser scanning fluorescence confocal microscopy. • Use 123I-metaiodobenzylguanidine-SPECT to look for sympathetic changes in the postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy, since the study of Druschky et al. (2001) suggested that altered

Forensic Postmortem Examination of Victims of SUDEP

137

postganglionic cardiac sympathetic innervation may increase the risk of cardiac abnormalities and/or SUDEP.

8.4â•…Questions for Retroprospective Postmortem Studies

1. What was the gender of the victim? 2. What was the victim’s age? 3. Did the victim of sudden, unexpected death have a known seizure disorder? 4. Was the victim witnessed exhibiting seizure(s) at the time of death? If so, what type of seizure was observed? 5. What was the actual date of death? How does the date of death for this victim compare with the date of the last known seizure? 6. Was any other cause of death found upon autopsy for any of these patients, such as cardiomyopathy and glial tumor? Did the victim have a terminal illness or any known associated illnesses? 7. What was the extent of postmortems? Were microscopic examinations done of the heart, lungs, brain, and other organs? 8. What was the microscopic pathology of the lungs and the hearts? 9. How did you interpret ἀndings of “heavy lungs” and data from hearts obtained in the postmortem autopsy? What do these ἀndings mean in terms of mechanism and signiἀcance (see Terrence 1990)? 10. Did the autopsy conduct postmortem examination of postganglionic cardiac sympathetic innervation of the heart? 11. Were quantitative antiepileptic drug levels found in the toxicological screen conducted postmortem? What assay method was used? 12. What were the actual antiepileptic drug levels postmortem? Were they within the therapeutic range or were they subtherapeutic? Were any levels below the lower level of quantiἀcation of the assay method? 13. Did the postmortem examine tears and aqueous humor or just aqueous humor for antiepileptic drug levels? 14. Were hair samples obtained to examine antiepileptic drugs? 15. Was the patient considered to be a compliant patient prior to death? 16. Was any genetic testing for entities that predispose an individual to cardiac arrhythmias done on the victim and/or the victim’s family prior to death? 17. How can the diagnosis of a seizure-induced death be differentiated from a diagnosis of a SUDEP death? 18. Was there any prior history of cardiac rhythm irregularities? 19. Was there any family history of cardiac rhythm irregularities or any history of sudden deaths?

Sympathetic changes in the postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy suggest that altered postganglionic cardiac sympathetic innervation may increase the risk of cardiac abnormalities and/or SUDEP (Druschky et al. 2001). Developmental and regulatory mechanisms determining density and pattern of cardiac sympathetic innervation are still unclear. Likewise, the exact role of innervation

138 Sudden Death in Epilepsy: Forensic and Clinical Issues

in arrhythmogenesis is unclear. The clinical study of Druschky et al. (2001) conἀrms the animal studies conducted by Lathers and colleagues (Lathers 1975, 1980a, 1980b; Lathers and Roberts 1985; Lathers and Spivey 1987; Spivey and Lathers 1985; Lathers et al. 1977a, 1977b, 1978; Lathers et al. 1986; Lathers and Schraeder 1982, 1983, 1987, 1989, 1990a, 2002; Lathers et al. 1984, 1987, 2003a, 1989, 1993, 2008a; Scorza et al. 2008; Lathers et al 2008b) in which postganglionic cardiac sympathetic neural discharge was monitored before and as arrhythmias developed. When considering likely mechanisms of SUDEP, it is important to consider whether there is any underlying, undeἀned genetic predisposition to arrhythmias and, thus, the mechanism of sudden death. Obviously, genetic predisposition to arrhythmias varies from patient to patient. The status of any subtle symptomatic disease present at the time of death will also vary from patient to patient and emphasize the importance of a genetic component to the autopsy. Animal data (Lathers et al. 1987; Staufferet al. 1989, 1990; O’Rourke and Lathers 1990; Dodd-O and Lathers 1990) demonstrated the lockstep phenomenon, i.e., postganglionic cardiac sympathetic discharge time locked to cortical epileptiform activity. The lockstep phenomenon may explain propagation of electrical impulses to autonomic nervous system regulatory centers, thus initiating arrhythmogenic potentials. During the lockstep phenomenon, cardiac postganglionic sympathetic and vagal discharges were synchronized with both ictal and interictal discharges and premature ventricular contractions, ST/T changes, and conduction blocks occurring concurrent with interictal spikes at a time when nonuniform cardiac sympathetic and vagal discharges were also observed. These experimental observations suggest that altered cardiac sympathetic innervation of hearts in patients who die suddenly and unexpectedly may contribute to nonuniform neural discharge, arrhythmias, and/or death. Detailed microscopic pathologic examination of cardiac autonomic nerves should be part of an autopsy study of SUDEP. A regulatory response, resulting from increased awareness of SUDEP, occurred in 1993 when the FDA focused the attention of practitioners and pharmaceutical manufacturers on the question of whether the use of anticonvulsant drugs contributes to or prevents sudden unexpected death in epileptic persons (Lathers 2002, 2003; Lathers and Schraeder 1995, 2002; Lathers et al. 2008a, 2008b; Scorza et al. 2008; Leestma et al. 1997; Schraeder and Lathers 1995). The FDA-convened panel of scientists considered the prevalence of sudden unexpected death in patients involved in studies associated with developing new anticonvulsant drugs and reviewed data on the risk of sudden unexpected death in patients taking lamotrigine. The risk of SUDEP was no different from that found in the young epilepsy population in general. Estimated SUDEP rates in patients receiving the new anticonvulsant drugs lamotrigine, gabapentin, topiramate, tigabine, and zonisamide were found to be similar to those in patients receiving standard anticonvulsant drugs, suggesting that SUDEP rates reflect population rates and not a speciἀc drug effect. The FDA requires warning labels on the risk of SUDEP in association with the use of each of the above-mentioned drugs. However, little data is available on the relative risk associated with the older, commonly prescribed antiepileptic drugs. The reader is referred to the chapters by Lathers and Schraeder (2010) and by Lathers, Schraeder, and Claycamp (Lathers et al. 2010) in this book for an in-depth discussion of the possible role of antiepileptic drugs in the cause and/or prevention of SUDEP. Widdes-Walsh and Devinsky (2007) discuss drug-resistant epilepsy, stating that it is a prevalent problem in spite of the fact that multiple ADE drug options are available. Mimics of drug-resistant epilepsy exist and cause diagnostic confusion. Fortunately, advances in epilepsy research and pharmacogenomics

Forensic Postmortem Examination of Victims of SUDEP

139

prove that new understandings of the mechanisms of drug resistance and tolerance allow rational antiepileptic drug strategies to prevent drug resistance. A different question regarding the clinical pharmacology of SUDEP has been raised by Tigaran and coauthors (Tigaran et al. 1997; Tigaran 2002; Tigaran et al. 2002; Tigaran et al. 2003). The question to be addressed is whether some persons with severe drug-resistant epilepsy, and without any indication of previous cardiac disease, may experience a beneἀcial effect of prophylactic treatment with cardioactive drugs to reduce the risk of sudden death (Tigaran et al. 1997). These investigators found ECG changes and ST-segment depression—many of which were closely related to the occurrence of epileptic seizures— that were suggestive of myocardial ischemia in the patients with severe drug-resistant epilepsy and no previous indication of cardiac disease. Of twelve patients with medically intractable epilepsy in studies with both ECG and EEG recordings, the ECG recording found in one person with chest pain minor yet morphologically conspicuous changes in the ECG, suggestive of myocardial infarction. This person with epilepsy died in cardiac arrest. In two publications in 2002, these authors reported atrio-ventricular block occurring as a life threatening cardiac arrhythmia complicating epileptic seizures in one person with medically intractable epilepsy (Tigaran et al. 2002) and cardiac abnormalities in patients with refractory epilepsy (Tigaran 2002). In another study, Tigaran et al. (2003b) examined 23 patients with drug-refractory epilepsy and found ST-segment depression in 40% of the patients and this was associated with a higher maximum heart rate during seizures, indicative of myocardial ischemia occurring in these individuals. These studies emphasize the importance of conducting both EEG and ECG in persons with severe drug-resistant epilepsy and the necessity of evaluating whether these patients would beneἀt from prophylactic treatment with cardioactive drugs to reduce the risk of sudden death. This chapter focuses on the need for retrospective forensic postmortem examination of not only the brains but also the hearts, myocardial autonomic innervations, and antiepileptic drug levels obtained from patients with epilepsy who die suddenly. A recent series of publications has addressed the multiple risk factors for sudden death in cardiac patients and in persons with epilepsy who have died suddenly and unexpectedly (Lathers and Schraeder 2010; Lathers et al. 2008a, 2008b; Scorza et al. 2008; (Herreros 2010). An evaluation of all potential risks for sudden death requires that risk assessments be done (Lathers 2002). These risk assessment efforts will aid all working in clinical and research ἀelds as they work to unravel the mystery of SUDEP. We recommend that retrospective postmortem examinations, including genetic determinations, be done using hearts and, of course, brains, obtained from patients who have died suddenly and unexpectedly. In addition, an effort should be made to obtain data on preexisting diatheses for cardiac arrhythmias in SUDEP victims and family members. The ἀndings from these rÂ�etrospective studies will allow prospective identiἀcation of persons in families of patients with epilepsy who have died suddenly. Finally, in the event that a physical autopsy was not conducted, the value of verbal autopsies must be determined as a relative substitute (Aspray 2005; Nashef et al. 1998; Lathers and Schraeder 2009; Nashef and Sahloldt 2010; Schraeder et al. 2010).

References Aspray, T. J. 2005. The use of verbal autopsy in attributing cause of death from epilepsy. Epilepsia 46 (Suppl 11): 15–17.

140 Sudden Death in Epilepsy: Forensic and Clinical Issues Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30 (4): 225–240. Chin, P. S., K. R. Branch, and K. J. Becker. 2004. Myocardial infarction following brief convulsive seizures. Neurology 63 (12): 2453–2454. Chin, P. S., K. R. Branch, and K. J. Becker. 2005. Postictal neurogenic stunned myocardium. Neurology 64 (11): 1977–1978. Chioza, B., H. Wilkie, L. Nashef, J. Blower, D. McCormick, P. Sham, P. Asherson, and A. J. Makoff. 2001. Association between the alpha(1a) calcium channel gene CACNA1A and idiopathic generalized epilepsy. Neurology 56 (9): 1245–1246. Chioza, B., A. Osei-Lah, H. Wilkie, L. Nashef, D. McCormick, P. Asherson, and A. J. Makoff. 2002a. Suggestive evidence for association of two potassium channel genes with different idiopathic generalised epilepsy syndromes. Epilepsy Res 52 (2): 107–116. Chioza, B., A. Osei-Lah, L. Nashef, B. Suarez-Merino, H. Wilkie, P. Sham, J. Knight, P. Asherson, and A. J. Makoff. 2002b. Haplotype and linkage disequilibrium analysis to characterise a region in the calcium channel gene CACNA1A associated with idiopathic generalised epilepsy. Eur J Hum Genet 10 (12): 857–864. Chioza, B., L. Nashef, P. Asherson, and A. Makoff. 2002c. CACNA1A and P/Q-type calcium channels in epilepsy. Lancet 359 (9302): 258. Devinsky, O. 2004. Effects of seizures on autonomic and cardiovascular function. Epilepsy Curr 4 (2): 43–46. Dodd-O, J. H., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in pÂ�henobarbital-pretreated cats. In Epilepsy and Sudden Death, Chapter 13. New York, NY: Marcel Dekker. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Ghali, N., and L. Nashef. 2010. Genetics of sudden death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, Chapter 18. ed. C. M. Lathers, P. Schraeder, M. Bungo, and J. Leestma. Boca Raton: CRC Press. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Helbig, I., H. C. Mefford, A. J. Sharp, M. Guipponi, M. Fichera, A. Franke, H. Muhle et al. 2009. 15q13.3 microdeletions increase risk of idiopathic generalized epilepsy. Nat Genet 41 (2): 160–162. Herreros, B. 2010. Cardiac channelopathies and sudden death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, Chapter 19. ed. C. M. Lathers, P. Schraeder, M. Bungo, and J. Leestma. Boca Raton: CRC Press. Hindocha, N., L. Nashef, F. Elmslie, R. Birch, S. Zuberi, A. Al-Chalabi, L. Crotti, P. J. Schwartz, and A. Makoff. 2008. Two cases of sudden unexpected death in epilepsy in a GEFS+ family with an SCN1A mutation. Epilepsia 49 (2): 360–365. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37 (4): 531–537. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42 (4): 588–594. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. Lathers, C. M. 1975. Effect of practolol and stalol on ouabrain-induced nonuniform adrenergic nerve activity. Pharmacologist 34: 745. Lathers, C. M. 1980a. Effect of timolol on autonomic neural discharge associated with ouabaininduced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106.

Forensic Postmortem Examination of Victims of SUDEP

141

Lathers, C. M. 1980b. The effect of metoprolol on coronary occlusion-induced arrhythmia and autonomic neural discharge. Fed Proc 39: 771. Lathers, C. M. 2002. Risk assessment in regulatory policy making for human and veterinary public health. J Clin Pharmacol 42 (8): 846–866. Lathers, C. M. 2003. Challenges and opportunities in animal drug development: A regulatory perspective. Nat Rev Drug Discov 2 (11): 915–918. Lathers, C. M., and J. Roberts. 1985. Are the sympathetic neural effects of digoxin and quinidine involved in their action on cardiac rhythm? J Cardiovasc Pharmacol 7 (2): 350–360. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 1990a. Arrhythmias associated with epileptogenic activity elicited by penicillin. In Epilepsy and Sudden Death., ed. C. M. Lathers and P. Schraeder, Chapter€9. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder, eds. 1990b. Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 1995. Experience-based teaching of therapeutics and clinical pharmacology of antiepileptic drugs. Sudden unexplained death in epilepsy: Do antiepileptic drugs have a role? J Clin Pharmacol 35 (6): 573–586; quiz 586–587. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42 (2): 123–136. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies. Epilepsy Behav 14: 573–576. Lathers, C. M., and P. L. Schraeder. 2010. Antiepileptic drugs beneἀt/risk clinical pharmacology: Possible role in cause and/or prevention of SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, Chapter 47. ed. C. M. Lathers, S. P. L., M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Lathers, C. M., and W. H. Spivey. 1987. The effect of beta blockers on cardiac neural discharge associated with coronary occlusion in the cat. J Clin Pharmacol 27 (8): 582–592. Lathers, C. M., K. M. Keller, J. Roberts, and A. B. Beasley. 1977a. Role of the adrenergic nervous system in arrhythmia produced by acute coronary artery occlusion. In Pathophysiology and Therapeutics of Myocardial Ischemia, ed. L. A., G. J. Kelliher, and M. Rovetto, Chapter 5. New York, NY: Spectrum. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977b. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39 (22): 2121–2141. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259.

142 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., L. J. Lipka, and H. Klions. 1988. Digitalis glycosides: A discussion of the similarities and differences in actions and existing controversies. Rev Clin Basic Pharm 7 (1-4): 1–108. Lathers, C. M., A. Z. Stauffer, N. Tumer, C. M. Kraras, and B. D. Goldman. 1989. Anticonvulsant and antiarrhythmic actions of the beta blocking agent timolol. Epilepsy Res 4 (1): 42–54. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital on autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33 (9): 837–844. Lathers, C. M., P. L. Schraeder, and H. G. Claycamp. 2003a. Clinical pharmacology of topiramate versus lamotrigine versus phenobarbital: Comparison of efficacy and side effects using odds ratios. J Clin Pharmacol 43 (5): 491–503. Lathers, C. M., S. A. Koehler, C. H. Wecht, and P. L. Schraeder. 2003b. Forensic antiepileptic drug levels in 2001 autopsy cases of sudden, unexpected deaths in persons with epilepsy in Allegheny County Pennsylvania. Paper read at the Annual Meeting of American College of Clinical Pharmacology, September, Orlando, FL. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008a. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Chapter 13. Hauppauge, NY: Nova Science. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Sodium channel dysfunction: Common pathophysiologic mechanism associated with sudden death ECG abnormalities in Brugada Syndrome and some types of epilepsy. Case histories. In Sudden Death in Epilepsy: Forensic and Clinical Issues, Chapter 20. ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earl. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38 (1): 47–55. Lipka, L. J., C. M. Lathers, and J. Roberts. 1988. Does chlorpromazine produce cardiac arrhythmia via the central nervous system? J Clin Pharmacol 28 (11): 968–983. Mullen, S. A., and I. E. Scheffer. 2009. Translational research in epilepsy genetics: Sodium channels in man to interneuronopathy in mouse. Arch Neurol 66 (1): 21–26. Nashef, L., and L. Sahloldt. 2010. Bereavement and sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, Chapter 58. ed. C. M. Lathers, P.€Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Nashef, L., S. Garner, J. W. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64€(3): 349–352. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55 (6): 857–860. O’Rourke, D. K., and C. M. Lathers. 1990. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, Chapter 15. New York, NY: Marcel Dekker. Randall, W. C., M. Szentivanyi, J. B. Pace, J. S. Wechsler, and M. P. Kaye. 1968. Patterns of sympathetic nerve projections onto the canine heart. Circ Res 22 (3): 315–323. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3 (1): 55–62. Schraeder, P. L., and C. M. Lathers. 1995. Clinical pharmacology of antiepileptic drug use: Clinical pearls about the perils of patty. J Clin Pharmacol 35 (12): 1120–1135.

Forensic Postmortem Examination of Victims of SUDEP

143

Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Schraeder, P. L., E. L. So, and C. M. Lathers. 2010. Forensic case identiἀcation. In Sudden Death in Epilepsy: Forensic and Clinical Issues, Chapter 6. ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13 (1): 263–264; author reply 265. Shields, L. B., D. M. Hunsaker, 3rd, J. C. Hunsaker, and J. C. Parker Jr. 2002. Sudden unexpected death in epilepsy: Neuropathologic ἀndings. Am J Forensic Med Pathol 23 (4): 307–314. Simon, R. P., L. L. Bayne, R. F. Tranbaugh, and F. R. Lewis. 1982. Elevated pulmonary lymph flow and protein content during status epilepticus in sheep. J Appl Physiol 52 (1): 91–95. Sisodiya, S., J. H. Cross, I. Blumcke, D. Chadwick, J. Craig, P. B. Crino, P. Debenham et al. 2007. Genetics of epilepsy: Epilepsy research foundation workshop report. Epileptic Disord 9 (2): 194–236. Spivey, W. H., and C. M. Lathers. 1985. Effect of timolol on the sympathetic nervous system in coronary occlusion in cats. Ann Emerg Med 14 (10): 939–944. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1990. Relationship of the lockstep phenomenon and precipitous changes in blood pressure. In Epilepsy and Sudden Death, Chapter 14. New York, NY: Marcel Dekker. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15 (1): 15–30. Terrence, C. F. 1990. Unexpected, unexplained death of epileptic persons: clinical correlation including pulmonary changes. In Epilepsy and Sudden Death, ed. C. Lather and P. Schraeder, Chapter€6. New York, NY: Marcel Dekker. Tigaran, S. 2002. Cardiac abnormalities in patients with refractory epilepsy. Acta Neurol Scand Suppl 177: 9–32. Tigaran, S., V. Rasmussen, M. Dam, S. Pedersen, H. Hogenhaven, and B. Friberg. 1997. ECG changes in epilepsy patients. Acta Neurol Scand 96 (2): 72–75. Tigaran, S., H. Molgaard, and M. Dam. 2002. Atrio-ventricular block: A possible explanation of sudden unexpected death in epilepsy. Acta Neurol Scand 106 (4): 229–233. Tigaran, S., A. Cascino Fuglsang-Frederiksen, and G. D. Cascino. 2003a. Temporal distribution of partial seizures during the sleep–wake cycle: possible signiἀcance for sudden unexpected death. Epilepsia 44: 123. Tigaran, S., H. Molgaard, R. McClelland, M. Dam, and A. S. Jaffe. 2003b. Evidence of cardiac ischeÂ� mia during seizures in drug-refractory epilepsy patients. Neurology 60 (3): 492–495. Tumer, N., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital upon autonomic function and epileptogenic activity induced by hippocampal injection of penicillin in cats. Epilepsia 26: 520. Widdess-Walsh, P., and O. Devinsky. 2007. Antiepileptic drug resistance and tolerance in epilepsy. Rev Neurol Dis 4 (4): 194–202.

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

9

Steven A. Koehler Paul L. Schraeder Claire M. Lathers Cyril H. Wecht

Contents 9.1 Introduction 9.2 Methods 9.3 Results 9.4 Discussion Acknowledgments References

145 148 148 155 158 158

9.1â•…Introduction Individuals with epilepsy can die from the progression of the primary underlying brain process, status epileptics, trauma, drowning precipitated by a seizure, or causes totally unrelated to the epilepsy (Earnest et al. 1992). The medical literature reports sudden, unexpected and unexplained deaths among individuals with epilepsy (Earnest et al. 1992). This disorder has been termed sudden unexpected death in epilepsy (SUDEP). SUDEP has been deἀned as a sudden, unexpected, witnessed or unwitnessed death, nontraumatic and nondrowning, in patients with epilepsy, with or without evidence of a seizure, excluding documented status epileptics, in whom postmortem examination does not reveal an anatomical or toxicological cause of death (Nashef and Shorvon 1997). SUDEP occurs most frequently among young individuals with a history of generalized tonic-clonic seizures. SUDEP has been reported in the literature to be responsible for 1.7–17% of all deaths among individuals with epilepsy (Ficker et al. 1998; Ficker 2000). However, these prior studies to ascertain the incidence of SUDEP suffered from selection bias and methodological limitations resulting in a tenfold difference between the various studies. Populationbased and forensic (medical examiner or coroner’s office) studies are few. Postmortem examinations of victims of SUDEP fail to establish the cause of death and toxicological analysis reveals subtherapeutic antiepileptic drug blood levels (Earnest et al. 1992). Some of the commonly reported features of SUDEP are highlighted in Table 9.1. Previous postmortem-based studies have reported the following changes. The brains in these series showed increased weight. The cerebral hemispheres were anemic and congested, and there were signs of hypoxia in the hippocampal regions. The hearts showed 145

146 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 9.1â•… Characteristics of SUDEP Characteristics History of epilepsy Age Sex Race Seizure event Death Location found dead Postmortem ἀndings Toxicological ἀndings

Features of SUDEP Victims 8–17% of deaths Average age is 28–35 years, rare in children Number of male victims is twice the number of female victims More frequent among blacks Witnesses to the seizure event are rare Occurred within minutes Bed Autopsies do not reveal a cause of death Heavier than normal weights of the heart, lungs, and liver Postmortem antiepileptic drug (AED) levels were either subtherapeutic or absent

ἀbrosis localized in the conductive system around the atrioventriclular bundle, edema of the conductive tissue, perivascular and interstitial ἀbrosis, and reversible myocyte vacuolization. The lungs showed increased weight, mild to moderate pulmonary edema, and alveolar hemorrhage (Leestma et al. 1989; Terrence et al. 1981; Jallon 1999; So 2008). The livers showed increased weight and venous congestion (Leestma et al. 1989). Leestma stated that the degree of passive congestion in both the liver and lungs, as well as the edema in the lungs, suggests some element of acute backward cardiac failure in SUDEP cases (Leestma 1990). The true incidence of SUDEP is not known and the estimates vary greatly. Possible reasons for this may be linked to the fact that SUDEP is not a diagnosis as such, since it is assigned when all other possible diagnoses have been eliminated, making it a default label. Another may be the general lack of understanding by medical examiners and coroner’s office personnel of what deἀnes a SUDEP death and when they should list it on the death certiἀcate. One method of investigating the protocol used within a medical examiner’s or coroner’s office regarding SUDEP deaths is to examine the death scene investigation, the autopsy and toxicology results, and the resulting death certiἀcate. In addition, by examining all the deaths at the medical examiner’s or coroner’s office and calculating the number of deaths that conform to the SUDEP deἀnition, an approximate incidence rate of SUDEP can be ascertained. All deaths that are sudden, unexpected, and unexplained by past medical history are investigated by the medical examiner or coroner’s offices. They typically conduct a detailed death scene investigation, a forensic autopsy, and a toxicological analysis of the body fluids. The level of investigation and the anatomical and toxicological data collected by the medical examiner or coroner’s office offer investigators studying SUDEP a wealth of information. The Allegheny County Coroner’s Office was selected for several reasons. First, it was the site of a previous examination of SUDEP death and, second, the forensic examination of these types of deaths involves a thorough death scene investigation, complete postmortem examination, and a comprehensive toxicological analysis of body fluids. The ἀrst study to examine SUDEP deaths at the Allegheny County Coroner’s Office was a retrospective study conducted in 1981 (Terrence et al. 1981). The study examined SUDEP cases that occurred between January 1, 1978 and December 31, 1979 and located a total of 8 cases: 4 male, 4 female, 4 white, and 4 black, ranging in age from 9 to 31 years. The pathological ἀndings at autopsy are shown in Table 9.2 (Terrence et al. 1981). While, the

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

147

Table 9.2â•… Pathological Findings Heart Weight Normal Weight Range (250–350 g)

Combined Lung Weight Normal Weight Range (650–1140 g)

1 2 3 4 5

275 400 360 370 200

1090 850 800 1250 755

6 7 8

374 300 110

1325 930 350

Patient Number

Neuropathological Findings None Cerebral edema None None Old contusion of L temporal pole, olfactory bulb, and orbitofrontal surfaces None None Cerebral edema

Source: Terrence, C. F., Rao, G. R., and Perper, J. A., Ann Neurol, 9 (5), 458–464, 1981.

normal heart weight by sex could not be calculated, using the overall range of 250–350€g as the normal heart weight, 50% of the cases had an enlarged heart (Cotran et al. 1994). No signiἀcant abnormalities in the brain were noted. Using the range of 650–1140 g as a normal weight range of the combined lung weight, 25% of the cases had heavy lungs. The type, number of prescription medications, and the postmortem blood levels are shown in Table 9.3 (Terrence et al. 1981). Phenytoin and phenobarbital was prescribed to all 8€patients,€primidone prescribed to 4, and carbamazepine, valproic acid, and mephenytoin prescribed to one patient. Toxicological analysis revealed subtherapeutic or absent blood levels of phenytoin in 7 cases and phenobarbital in 4 out of the 8 cases. Levels for carbaÂ� mazepine, valproic acid, and mephenytoin were also measured postmortem. The purpose of the current study was to present an overview of all deaths examined at the Allegheny County Coroner’s Office among individuals with epilepsy in AlÂ�Â� legheny County during 2001, to describe epidemiological, anatomical, pathological, and Table 9.3â•… Prescribed Medications and the Postmortem Drug Levels Patient Number 1 2 3 4 5 6 7 8

Postmortem Drug Levels Drug Prescribed Phenytoin, phenobarbital Phenytoin, phenobarbital, primidone Phenytoin, phenobarbital, primidone Phenytoin, phenobarbital, primidone Phenytoin, phenobarbital, primidone Phenytoin, phenobarbital Phenytoin, phenobarbital Phenytoin, phenobarbital, carbamazepine, valproic acid, mephenytoin

Phenytoin

Phenobarbital

Other

None detected None detected

None detected None detected

None detected None detected

None detected

Therapeutic

None detected

Subtherapeutic

Therapeutic

None detected

Therapeutic

Therapeutic

None detected

None detected Subtherapeutic None detected

None detected Subtherapeutic Therapeutic

None detected None detected None detected

Source: Terrence, C. F., Rao, G. R., and Perper, J. A., Ann Neurol, 9 (5), 458–464, 1981.

148 Sudden Death in Epilepsy: Forensic and Clinical Issues

toxicological features of each SUDEP case, and to determine if SUDEP is present on the death certiἀcates issued by the coroner’s office.

9.2â•…Methods The Allegheny County Coroner’s Office has jurisdiction to investigate all deaths within Allegheny County, which is located in western Pennsylvania and encompasses a population of ~1.2 million individuals. All deaths investigated by the Allegheny County Coroner’s Office were reviewed from January 1, 2001 to December 31, 2001. This office investigates more than 6000 cases and conducts more than 1200 autopsies annually. All cases were identiἀed by conducting a computer and hand search of all the case reports by the Chief Forensic Epidemiologist (SAK) for all deaths with the words “seizure disorders,” “epilepsy,” or “SUDEP” appearing in Part I or Part II of the death certiἀcate. The following epidemiological information was collected: age, sex, race, time and date last seen alive, and the time and date of death. Seizure-related information collected included past medical€ history, prescribed medications, and who witnessed the event. Pathological information collected included the weights of the internal organs obtained from the forensic autopsy report. The normal weight parameters of the internal organs are based on published data and, where possible, separated by age and sex. Toxicological analysis was conducted on the blood, bile, urine, and eye fluid recovered during the autopsy. The blood used for the toxicological analysis was collected from the heart during the autopsy. The number and level of drugs detected in the body fluids were obtained from the toxicological report. The data was entered and analyzed by Statistical Package for the Social Sciences® (11.0 Chicago, IL). The anatomical and toxicological data for all deaths with a diagnosis of epilepsy were summarized. Based upon the summarized data, the cases that meet the classiἀcation criteria for SUDEP were determined.

9.3â•…Results A total of 12 deaths were identiἀed in which seizure disorder was listed as either the immediate cause of death or contributed to the death during the study period. Epidemiological characteristics are shown in Table 9.4. Among the 12 cases, 58.3% were male, 41.7% were female, and all cases were white. All cases were between the ages of 38 and 54 years old with the mean age of the males being slightly higher than that of the females. Overall, more than 90% of the seizures were unwitnessed events. The only witnessed seizure was that of a 42-year-old female. She was in her residence when she had a sudden seizure and became unresponsive. She was pronounced dead at the emergency room one hour after the seizure. Among the males, four were found in their bathroom and three in bed. Among the females, all were found in their residence (one on the kitchen floor, one on the bed, one in a chair, and one outside on the rear deck). Deaths were most frequent in January among the males and in September for the females. According to information contained within the death investigation report, all 12 cases had a past medical history for either a diagnosis of seizures or epilepsy. All cases would be considered as having a seizure disorder. A complete postmortem examination was conducted on all but one case, due to an advanced level of decomposition.

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

149

Table 9.4â•… Epidemiological Characteristics: Age, Sex, Race, Seizure Events, Month of Event, Medical History, and Type of Postmortem Examination Epidemiological Characteristics Total number of cases (all were Caucasian) Age range (mean) Seizure event witnessed Month of seizure

Medical history of seizures disorder or epilepsy Manner of death Type of postmortem examination a

Males

Females

7 38–50 (x = 45.4) Yes: 0 No: 7 January: 3 February: 1 March: 1 August: 1 October: 1 Yes: 7

5 39–54 (x = 44.8) Yes: 1 No: 4 June: 1 July: 1 August: 1 September: 2

Natural: 6 Accident: 1 Complete: 6 External only: 1a

Natural: 5

Yes: 5

Complete: 5 External only: 0

Decomposed.

An examination of the information contained on Part I or Part II of the death certiἀcate by sex is shown in Tables 9.5 and 9.6. The immediate cause of death is the event that directly and immediately resulted in the death. The contributory cause of death consists of conditions that play a role in causing the death, but do not cause immediate death. The section criteria for this study included any death where seizure disorder appeared on the death certiἀcate. Seizure disorder was listed as the immediate cause of death in 83.3% of the cases (6 males and 4 females), cardiomyopathy was listed in one male case, and a glial brain tumor was listed in one female case (Tables 9.5 and 9.6). Among males, seizure disorder was the immediate cause of death among 6 of the 7 deaths. In Male Case #7, the seizure disorder was considered by the forensic pathologist conducting the examination to have played a signiἀcant role in contributing to the death. The manner of death was ruled natural in six cases (85.7%) and accidental in one (14.3%). Among the females, seizure disorder Table 9.5â•…Immediate and Contributory Causes of Death for Males Death Certiἀcate Part I Immediate cause of death

Part II Contributory cause of death

Case No. 1

2

3

Seizure Seizure Seizure disorder disorder disorder (clinical)

4 Seizure disorder

5 Seizure disorder

6 Seizure disorder

Chronic Liver Liver Chronic cirrhosis cirrhosis obstructive obstructive disease pulmonary pulmonary disease disease

7 Dilated cardiomyopathy with arteriosclerotic cardiovascular disease, acute pneumonia, emphysema Seizure disorder

150 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 9.6â•…Immediate and Contributory Causes of Death for Females Case No.

Death Certiἀcate Part I Immediate cause of death

1

2

Seizure disorder

Seizure disorder (clinical)

3

4 Seizure disorder with hypertensive and arteriosclerotic cardiovascular disease

Seizure disorder (clinical) with arteriosclerotic cardiovascular disease

5 Convulsive seizure developed from grade II glial tumor

Part II Contributory cause of death

was the immediate cause of death among four of the ἀve deaths. In one case, the immediate cause of death was a convulsive seizure that developed in association with a grade II glial tumor. The manner of death was ruled natural in all ἀve cases. Overall, the incidence of death in persons with a history of seizures in which no cause of death was found on postmortem was 0.83% (10 in 1200 autopsies). The 2001 Allegheny County Coroner’s data found 0.833% (10 of 1200) of the autopsy cases met SUDEP criteria. Among the 12 cases, 85.7% (six of all males) and 80% (four of all females) represented SUDEP deaths. Autopsies were performed on all ἀve females and six of the seven males. The weights of the internal organs were available in 92% of the cases. The weights of the hearts, thicknesses of the left ventricles, lungs, and brains, and brain pathologies by sex are shown Table 9.7â•…Observed and Normal Organ Weights of the Heart, Lungs, and Brain by Sex

Sex

Weight (lb)

Heart Weight (g) (Normal Weight)a

Left Ventricle Thickness (cm)

Combined Lung Weight (g)b

Brain Weight (g)c

Male

118

310 (281)

1.2

1440

1070

121 131 145 240

445 (292) 360 (302) 400 (317) 455 (406)

1.4 1.2 1.2 1.5

1095 1200 1510 955

1240 1275 1760 1570

268 99.5 108 160 226

575 (432) 300 (230) 345 (243) 445 (284) 395 (329)

1.6 1.5 1.2 1.7 1.2

2065 915 550 1055 975

1760 1215 1205 1450 920

300

400 (371)

1.4

970

1365

Female

a b c

Normal heart range (mean expected heart weight based on body weight and sex). Normal combined weights of lungs: males, 720–1140 g; females: 650–960 g. Normal brain weights: males, 1100–1700 g; females: 1050–1550 g.

Brain Pathology Craniotomy, patchy subarachnoid hemorrhage Normal Slight swelling Normal Congestion, periventricular white matter Normal Chronic mild swelling Normal Normal Small crowded gyri, w/o atrophy Congested, mild swelling

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

151

700 600 500 400

Actual

300

Normal

200 100 0 1

2

3

4

5

6

Figure 9.1╇ Normal and actual heart weight (g) among males.

in Table 9.7. The normal ranges of the internal organs are also shown in Table 9.7. The observed weight of the heart was above the expected mean weight based on body weight and sex in all cases. Figures 9.1 and 9.2 show the normal and actual heart weights among males and females, respectively. The normal range was based on the weight and sex of the subjects. The weight of the heart was above the expected weight, after adjusting for body weight and sex in all cases. Left-ventricle hypertrophy was seen in two cases. The combined weight of the right and left lungs showed that more than 66% of the male and 60% of the female lungs exceeded normal parameters. The normal range of combined lung weights was based on sex. The mean weight for all the female lungs was normal (449.2 g), while that of the males was elevated (688.75 g). Figures 9.3 and 9.4 show the weight of the normal and actual combined lung weights by sex. Among males, the combined lung weight exceeds the upper limits of normal in four cases. Among females, the combined lung weight exceeds the upper limits of normal in three cases and dropped below the lower limits in one case. Examinations of the brains of the males showed that 33% were above normal limits, 50% were within normal limits, and 16% were below. The normal range of brain weights was based on sex. Examinations of the brains of the females showed that 80% were within

500 450 400 350 300

Actual

250

Normal

200 150 100 50 0 1

2

3

4

Figure 9.2╇ Normal and actual heart weight (g) among females.

5

152 Sudden Death in Epilepsy: Forensic and Clinical Issues 2500 2000 Actual

1500

Lower Limit Upper Limit

1000 500 0 1

2

3

4

5

6

Figure 9.3╇ Actual and normal upper and lower ranges of the combined weight of lungs (g) among males.

normal limits and 16% were below. Figures 9.5 and 9.6 show the normal and actual brain weight by sex. In the male group, the past medical history, prescribed medications, drugs identiἀed by the toxicology screen, level of postmortem compounds, and pathological factors are shown in Table 9.8. All seven had a history of seizures or epilepsy disorder. In ἀve cases (71%), antiepileptic drugs were prescribed (clonazepam, phenytoin). The death scene investigation failed to ascertain a list of prescription medications in two deaths. Toxicological analysis revealed that AED medications were detected in 57% of the cases, with phenytoin the most frequently detected. All the detected AED were at subtherapeutic levels. In only one case was the toxicological analysis negative. The bodies in three of the cases were in varying stages of decomposition. In the female group, the past medical history, prescribed medications, drugs identiἀed by the toxicology screen, level of postmortem compounds, and pathological factors are shown in Table 9.9. All cases had a past medical history of seizures. In two cases (40%), medications for seizures were prescribed (Dilantin, Lamotrigine, Tegretol). The death 1200 1000 800

Actual Lower Limit

600

Upper Limit

400 200 0 1

2

3

4

5

Figure 9.4╇ Actual and normal upper and lower ranges of the combined weight of lungs (g) among females.

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

153

2000 1800 1600 Male

1400

Lower Upper

1200 1000 800

118

121

131

145

240

268

Figure 9.5╇ Actual and normal upper and lower ranges of brain weight (g) among males by body weight (lb).

scene investigation failed to ascertain a list of prescription medications in one death. Toxicological analysis revealed that AED medications were detected in 60% of the cases, with phenytoin the most frequently detected. Toxicological analysis showed that the AED medications were at therapeutic levels in three, subtherapeutic in one, and above the therapeutic level in one case. Even though all the female cases had a diagnosis of seizure disorder, no information was collected as to why three of them were not taking any antiepileptic medications. This omission is one of the pitfalls of a retrospective study. This category of information (i.e., why a person with a diagnosis of epilepsy is not on medication) should be a speciἀc issue addressed in future studies.

1600 1500 1400 1300

Female

1200

Lower

1100

Upper

1000 900 800

99.5

108

160

226

300

Figure 9.6╇ Actual and normal upper and lower ranges of brain weight (g) among females by body weight (lb).

154 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 9.8â•… Medical History, Prescribed Medication, Toxicology Screen, and Pathological Features among Males

Medical History

Prescribed Medications

Drugs Identiἀed (with Levels) in Toxicology Screen (AED Drugs in€Bold)

Seizures

Dilantin

Acetone (6 mg%) Alcohol (5 mg%)

Seizures

Prinvil Clonazepam K-Dur Oramorph Depo Medrol Provential inhaler

Doxylamine (0.014 mg%) Benzodiazepines: ╇ Nordiazepam (0.01€mg%) ╇ Chlordiazepoxide (0.013 mg%) ╇ Demoxepam (0.023 mg%) ╇ Oxazepam (0.010 mg%) ╇ Clonazepam (too low to quantify) Morphine (0.087 μg/mL) Dextromethorphanpositive Ibuprofen-positive Diphenhydraminepositive Phenytoin (8.31 μg/mL)

Seizures

Dilantin

DVT, seizures, alcoholism

Dilantin Keflex

None detected

Seizures, alcoholism

Celexa Neurontin Phenytoin Indomethacin Unknown

Ethanol (0.04%) Phenytoin (1.05 μg/mL)

HIV, seizures, chronic lung disease Bipolar depression Epilepsy, depression

Unknown

Citalopram-positive Desmethylcitaloprampositive Ethanol (0.02%) Phenytoin (2.45 μg/mL) Olanzapine (0.041 mg%) Sertraline (0.021 mg%) Valproic acid (TDX) (8.99 μg/mL)

Levels of Postmortem Compounds — —

Above therapeutic levels Therapeutic level Subtherapeutic level — Subtherapeutic level — Subtherapeutic level — — — Subtherapeutic level —

— Subtherapeutic level — — — Subtherapeutic level Above therapeutic level Therapeutic level Subtherapeutic level

Pathological Features Alive: 1/7/01 Found dead: 1/11/01 12:30 p.m. Body moderately decomposed Alive: 1/17/01 4:30€p.m. Dead: 1/18/01 12:30 p.m.

Alive: 1/18/01 7:30€a.m. Found dead: 1/18/01 4:50 p.m. Alive: 2/11/01 1:00€a.m. Found dead: 8:00€p.m. Decomposed

Putriἀcation

Alive: 10/16/01 11:00 p.m. Found dead: 10/16/01 12:00€p.m.

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

155

Table 9.9â•… Medical History, Prescribed Medication, Toxicology Screen, and Pathological Features among Females

Medical History

Prescribed Medications

Drugs Identiἀed in Toxicology Screen (AED Drugs in Bold)

Level of Postmortem Compounds Subtherapeutic level Therapeutic level Subtherapeutic level Above therapeutic level

Hysterectomy, seizures

Dilantin Lamotrigine

Phenytoin (5.36 μg/mL) Lamotrigine (0.79 mg%) Ibuprofen (0.81 mg%)

Seizures

Neurontin Tegretol Dilantin

Clonazepram (8.37 μg/ mL)

Amyotrophic lateral sclerosis, depression, seizures

Bupropion Ditropanxl Zanaflex Baclofen Rilutek Unknown

Bupropion (0.006 mg%) Threoaminobupropionpositive

Therapeutic level —

Trazodone (0.296 mg%) Venlfaxine (0.175 mg%) O-Desmethylvenlafaxine (0.025 mg%) Carbamazepine (4.56 μg/mL) Phenytoin (14.68 μg/ mL) Butalbital (0.246 mg%) Acetaminophen (14.05 mg%)

Therapeutic level Therapeutic level —

Diabetes, ╇ seizures, ╇ hypertension, ╇ asthma Hypertension, ╇ seizures, TMJ, ╇ osteoporosis, ╇ cervical fusion ╇ psychiatric history

Prempro Gybutynin Butslbitalapac-caff-tap

Therapeutic level Therapeutic level Subtherapeutic level Nearly toxic

Pathological Features Alive: 6/7/01 evening Found dead: 6/8/01 3:30 p.m. Alive: 7/6/01 12:20€a.m. Found dead: 7/6/01 12:21 a.m. Witnesses to seizure Alive: Unknown Found dead: 8/7/01 5:37 p.m. Moderate putriἀcation Alive: 9/28/01 11:30€p.m. Found dead: 9/29/01 9:31 a.m. Alive: 9/4/01 2:30€p.m. Found dead: 9/5/01 11:32 p.m.

9.4â•…Discussion The sudden and unexpected deaths among individuals with a history of epilepsy are cases that fall under the jurisdiction of the medical examiner or coroner’s office for a forensic investigation. These investigations involve a detailed review of the victims’ past medical history, a list of current medications with dosage levels, and a determination of the events immediately surrounding the time of death. The body, in most cases, also undergoes a complete forensic autopsy and toxicological analysis. All these factors can provide a fairly accurate representation of the events at the time of death, including any anatomical or disease processes that might have contributed to the death and the circulating blood levels of any medications or other compounds at the time of death. Therefore, those studying SUDEP via forensic data will be provided with not only the circumstances surrounding the death, but the weights and state of the internal organs, and a detailed toxicological analysis of the body fluids. The difficulty is locating these types of cases within the medical examiners’ or coroners’ ἀles. When medical examiner’s or coroner’s offices are confronted with a death where the autopsy and toxicological analyses fail to identify a speciἀc cause of death, and that individual has a well-documented history of seizures, the cause of death on the death certiἀcate would be listed as seizure disorder and not as SUDEP. A lack of any other

156 Sudden Death in Epilepsy: Forensic and Clinical Issues

explanation for the cause of death indicates that these victims should be considered more appropriately as having a deἀnite classiἀcation of SUDEP. However, some authors have highlighted concerns with the death certiἀcate generated from the medical examiner and coroner data. If the diagnosis of epilepsy and/or seizure disorder is not stated on death certiἀcates with accuracy, many SUDEP deaths will go undetected. Coyle et al. (1994) examined 40 cases of SUDEP identiἀed in the United Kingdom in 1992. Postmortem reports and witness statements were examined to look at the accuracy of the coroners’ diagnoses. In 70% of those cases, the type of seizure either was not known or was not referred to. The review also found inconsistent reports of the position of the victim’s body, examination of the organs, especially the brain, details of the medication history, and the toxicology examinations conducted. Epilepsy or seizure disorder as an attributed cause of death was used in less than half of the cases even though the victims had a history consistent with this diagnosis. All of these ἀndings raise issues of the quality of data included in postmortem reports. Within the United States, the majority of medical examiners’ and coroners’ offices traditionally list such deaths as seizure disorders rather than epilepsy. This practice of not using the diagnosis of SUDEP on the death certiἀcate is not due to a lack of understanding or acknowledgment of SUDEP as a valid classiἀcation of the cause of death. The lack of utilization of SUDEP as a ἀnal diagnosis in appropriate cases is in agreement with data obtained in a recent nationwide survey (Schraeder et al. 2006) that found most medical examiners and coroners did not use the diagnosis of SUDEP when entering the cause of death on the death certiἀcate in those seizure disorder cases where no cause of death was identiἀed on postmortem. This lack of use of SUDEP on the death certiἀcates results in an underreporting of SUDEP and overreporting of deaths from seizure disorder. This study highlights the increased need to educate medical examiners’ and coroners’ offices regarding when the category of SUDEP should be used. Forensic pathologists face their greatest challenge when attempting to determine causes of death that rely on exclusion criteria, such as in SUDEP. In these types of cases it is important for the examination to attempt to rule out other possible causes before classifying the death as SUDEP on the death certiἀcate. In a case that is likely to be SUDEP, the following should be part of the standard protocol. First, conduct a detailed death scene investigation with an emphasis on collecting prescription medicines at the residence. Second, obtain a detailed past medical history with a thorough review of the medical history. Third, conduct a meticulous forensic autopsy with a comprehensive examination of the heart and its conductive system. Finally, review the results of the toxicological analysis of the body fluids with special attention to the postmortem levels of AED medications. When comparing the results of the study conducted in 1981 (Terrence et al. 1981) with our data, there were some similarities and some differences. Enlarged hearts were reported in 50% of the victims in the earlier study; our study reported that all victims had an enlarged heart. However, a retrospective case control study in a medical examiner’s population by Davis and McGwin (2004) found no difference in mean heart mass between the two groups. With increasing age, they did ἀnd competing causes of death in patients with epilepsy. Their ἀndings support the concept that SUDEP occurs in young adults. Opeskin et al. (2000) conducted cardiac pathology exams in 10 victims of SUDEP and in 10 controls. They found no abnormalities in the conduction system nor in the other cardiac pathological parameters studied. In contrast, the ἀndings in our study showed above-normal heart weights. Other studies found cardiac pathological changes obtained

One-Year Postmortem Forensic Analysis of Deaths in Persons with Epilepsy

157

Table 9.10â•… Questions to Be Asked by Death Investigators Relating to Possible Seizure-â•›Related Deaths Questions 1 2 3 4 5 6 7 8 9 10

Was the deceased diagnosed with a seizure disorder? At what age was the deceased diagnosed with a seizure disorder? Was the deceased taking any antiepileptic drugs? Do you have a list of the deceased’s antiepileptic drugs? Was the deceased compliant with taking his/her medication? What was the date of the deceased’s last seizure? Were the deceased’s seizures well-controlled? Is there a family history of seizures in the deceased’s family? When was the last time the deceased took his/her antiepileptic drugs? Do you have the name of the deceased’s treating physician?

from SUDEP victims. George and Davis (1998) looked at the pathology of arrhythmogenic right ventriclular cardiomyopathy/dysplasia. Inflammatory inἀltrates (i.e., lymphocytes) were observed in 60% of cases while myocyte necrosis was found in only one case. In the current study, 63% of the deaths had above-normal lungs weights while the earlier data reported that only 25% had heavy lungs. Both studies showed that the levels of AED medications detected during toxicological analysis were typically at the subtherapeutic level, or absent entirely. In this study, 7 of the 12 deaths were determined to have occurred in patients who had been prescribed anticonvulsant medications. Based on the postmortem analysis for antiepileptic drugs, ἀve were at subtheraputic levels, one was within the therapeutic range, one was€above therapeutic levels, and ἀve were totally absent of antiepileptic drugs. Based on these results and supported by the Terrence et al. (1981) study, a high percentage were either due to individuals being noncompliant with their medications or the detection methods of the antiepileptic drugs; levels in postmortem samples were below the lower limit of quantiἀcation of the assay method. The issue of using postmortem blood to estimate the circulating blood level in the living due to the phenomena of redistribution is a concern in forensic€toxicology. In this study, all the analyses were conducted on blood collected from the chambers of the heart. This blood typically shows a higher concentration of the drug than does peripheral blood. Therefore, the recoded levels of the antiepileptic drugs were lower in the circulating blood, meaning that the individuals with subtherapeutic heart blood levels have an even lower level of the drug in their circulating blood. The current observation of increased lung weights has also been reported to be present in SUDEP by other investigators. Noncompliance of anticonvulsants has been suggested as a risk factor for neurogenic pulmonary edema and the ἀnding of pulmonary edema is thought to be associated with an adrenergic component in the death event (Lathers and Schraeder 2002; Tomson et al. 2005; Hughes 2009). For future studies, there is a need to have more detailed data collection, for example, about seizure types, the premorbid antiepileptic drug levels, the general level of compliance, and any stressful circumstances preceding death. In essence, the need for detailed information could be obtained with oral autopsy data (Lathers and Schraeder 2009). Finally, there is a need to emphasize the importance of prospective data collection. Those investigating a possible death related to a seizure-related medical history should obtain a detailed description of the events leading up to the death from the next of kin and obtain all medical records and a complete list of prescription

158 Sudden Death in Epilepsy: Forensic and Clinical Issues

medications. To ensure that the important information is ascertained, Table 9.10 lists the key questions.

Acknowledgments The authors thank Shaum Ladham, MD, Leon Rozin, MD, Abdulrezak Shakir, MD, and Joseph Dominick, RN, LFD, for technical support in the data collection process.

References Cotran, R. S., V. Kumar, S. L. Robbins, and F. J. Schoen. 1994. Robbins Pathological Basis of Disease, 5th ed. Philadelphia, PA: W. B. Saunders. Coyle, H. P., N. Baker-Brian, and S. W. Brown. 1994. Coroners’ autopsy reporting of sudden unexplained death in epilepsy (SUDEP) in the UK. Seizure 3 (4): 247–254. Davis, G. G., and G. McGwin Jr. 2004. Comparison of heart mass in seizure patients dying of sudden unexplained death in epilepsy to sudden death due to some other cause. Am J Forensic Med Pathol 25 (1): 23–28. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33 (2): 310–316. Ficker, D. M. 2000. Sudden unexplained death and injury in epilepsy. Epilepsia 41 (Suppl 2): S7–S12. Ficker, D. M., E. L. So, W. K. Shen, J. F. Annegers, P. C. O’Brien, G. D. Cascino, and P. G. Belau. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51 (5): 1270–1274. George, J. R., and G. G. Davis. 1998. Comparison of anti-epileptic drug levels in different cases of sudden death. J Forensic Sci 43 (3): 598–603. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14 (2): 280–287. Jallon, P. 1999. Sudden death of epileptic patients. Presse Med 28 (11): 605–611. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42 (2): 123–136. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies. Epilepsy Behav 14: 573–576. Leestma, J. E. 1990. Sudden unexpected death associated with seizures: A pathological review. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26 (2): 195–203. Nashef, L., and S. D. Shorvon. 1997. Mortality in epilepsy. Epilepsia 38 (10):1059–1061. Opeskin, K., A. Thomas, and S. F. Berkovic. 2000. Does cardiac conduction pathology contribute to sudden unexpected death in epilepsy? Epilepsy Res 40 (1): 17–24. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (Suppl 9): 93–98. Terrence, C. F., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann Neurol 9 (5): 458–464. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61.

Drug Abuse and SUDEP Steven B. Karch

10

Contents 10.1 Introduction 10.2 Channelopathies and Abused Drugs 10.2.1 QT Interval Prolongation 10.2.2 QT Shortening 10.3 QT Dispersion 10.4 Abnormal Catecholamine Metabolism References

159 161 161 164 164 165 165

10.1â•…Introduction Nearly a quarter century ago, Leestma et al. (1985) described the ἀndings seen at autopsies of 66 epileptics who had died unexpectedly during or just after experiencing a seizure. The autopsy ἀndings were insufficient to explain the cause of death in any of the cases. The syndrome of sudden death during or immediately after a seizure has come to be called sudden unexplained death in epilepsy (SUDEP). Over a 10-year period, the mean age of epileptics with SUDEP in Leestma’s studies was 31.4 years. Of these, 37% were found dead in bed, 49% were black males, 25% were white males, 11% were black females, and 15% were white females ranging in age from 10 months to 60 years (mean age, 28 years); nearly half of the decedents were found “dead in bed,” a description that applies equally well to deaths from long QT syndrome (LQTS), an entity due to heritable channelopathies (Ackerman et al. 2001), or to diabetes with hypoglycemia (Rothenbuhler et al. 2008; Tu et al. 2008). By convention, the abbreviation sudden unexplained death syndrome (SUDS) is used to describe the death of anyone older than 2 years who dies in this fashion, whereas for children 2 years and younger, the diagnosis that is used is sudden infant death syndrome (SIDS). More recently, a similar syndrome has been recognized in young diabetics (Gill et al. 2009). As with SUDEP, the cause of death is not apparent in any of these disorders, although it seems probable that an explanation is to be found only at the molecular level. One popular hypothesis holds that SUDEP victims may have died because of QT interval prolongation (QTd) and that SUDEP is just another variety of LQTS. The presence of prolonged QT intervals reflects delayed cardiac repolarization. It can be caused by a variety of different abnormalities, including acute hypoglycemia, superimposed upon the presence of cardiac autonomic neuropathy (Tu et al. 2008). This process may account for the death of young diabetics as noted above, although why the QT interval in young diabetics should be prolonged has never really been explained. Other factors, such as channelopathies and cardiomyocyte membrane abnormalities, may also be involved. The notion that the same abnormality might account for both SUDEP and LQTS (Aurlien et al. 2009) has the potential to open up new areas of research. 159

160 Sudden Death in Epilepsy: Forensic and Clinical Issues

Numerous risk factors for SUDEP have been proposed and a few have been consistently identiἀed, including young age, early seizure onset, refractory seizures, generalized tonic clonic seizures, and male gender. Studies have found that many of these individuals are in bed, presumably asleep, at the time of death, that anticonvulsant blood levels are subtherapeutic, and that a structural brain lesion often can be identiἀed (Leestma et al. 1985). The current consensus seems to be that SUDEP is primarily a seizure-related cause of death, but the mechanisms underlying SUDEP are unknown (Jehi and Najm 2008; Hughes 2009). Most SUDEP victims are found alone in a room at home and only rarely have they been vigorously exercising. A recent meta-analysis, which included a review of the Cochrane database (Monte et al. 2007), found that patients were more likely to be discovered asleep or, at least, in their beds. Subtherapeutic concentrations of antiseizure medications were found in nearly 70% of Leestma’s cases, many of the decedents having no medications detectable at all. Signiἀcant but nonprogressive brain abnormalities were present in more €than half the decedents. Leestma et al. (1985) estimated the prevalence of seizureassociated SUDEP was between 1:525 and 1:2100 epileptics (Tellez-Zenteno et al. 2005). Other retrospective studies have reported both slightly higher and slightly lower incidence rates (Tellez-Zenteno et al. 2005; Hughes 2009), but the fact remains that SUDEP is the most common cause of death in persons with epilepsy. SUDEP is rare in those who have only recently been diagnosed with epilepsy, and it is equally uncommon in those who are in remission. Many theories accounting for the etiology of SUDEP have been proposed (Kloster and Engelskjon 1999; Lathers and Schraeder 2006; So and Sperling 2007; Aurlien et al. 2009). Most involve cardiac arrhythmias, mediated by sympathetic autonomic events, which are thought to occur during the seizures. The possibility of heritable channelopathy is receiving much closer scrutiny than in the past. Some candidate genes make the QT interval longer and some make it shorter, but both abnormalities are associated with a sudden arrhythmia-related death. These changes are important not only because they have been identiἀed in epileptics (Akalin et al. 2003), but also because they are associated with the use of some abused drugs, in particular cocaine and alcohol (Gamouras et al. 2000; Karle and Kiehn 2002; Uyarel et al. 2005; Yap et al. 2009). The possibility that genetic polymorphisms may be responsible for SIDS, SUDS, and even SUDEP cannot be ignored, but neither can the process known as myocardial remodeling, which causes the QT interval to be longer in some parts of the heart than in others. This phenomenon is referred to as QT interval dispersion; if the difference between the longest and shortest QT interval measured in a 12-lead electrocardiogram exceeds 80 ms, a state of QT dispersion (QTd) is said to exist (Anderson 2003). QTd has the same strong association with sudden cardiac death as interstitial ἀbrosis and channelopathy (Cuddy et al. 2009). The process of myocardial remodeling has received relatively little attention from neurologists, but is increasingly recognized as a cause of sudden death by cardiologists and electrophysiologists (Fischer et al. 2007). Stimulant drugs initiate the same remodeling process as seen in states of chronic catecholamine excess, and such a state can be said to exist during recurrent seizure activity (Meierkord et al. 1994; Henning and Cuevas 2006). Hypertension (Haider et al. 1998), catecholamine excess (Rona 1985), and stimulant drug abuse (Karch et al. 1998 1999) always lead to myocardial ἀbrosis and left ventricular hypertrophy (i.e., myocardial remodeling). Either or both of these processes can provide

Drug Abuse and SUDEP

161

the substrate for sudden arrhythmia death (John et al. 2004; Haider et al. 1998). Ventricular hypertrophy and ἀbrosis are both common ἀndings in SUDEP victims (P-Codrea Tigaran et al. 2005) and it may well be that QTd is responsible for some cases of SUDEP.

10.2â•…Channelopathies and Abused Drugs 10.2.1â•…QT Interval Prolongation Ackerman and others have proposed that perturbations of the hERG channels (“rapid delayed repolarizing channel”) can confer a susceptibility for seizures themselves or, alternatively, cause QT prolongation, resulting in syncopal episodes that trigger seizures (Johnson et al. 2009; Nemec et al. 2009). This is not an unreasonable theory given that many patients who have seizures and QT prolongation are often confused with patients suffering from primary seizure disorders, and are treated with antiepileptic drugs instead of the internal deἀbrillators they really need (Lathers et al. 2010). In a recently published controlled case study, seizure phenotype was recorded in 98 of 343 (29%) patients who were randomly checked for the most common channelopathies. A seizure phenotype was more common in individuals with LQT2 (36 of 77 or 47%) than those with LQT1 (16 of 72 or 22%, p < 0.002) and LQT3 (7 of 28 or 25%, p < 0.05, not signiἀcant). LQT1 and LQT3 combined cohorts did not differ signiἀcantly from the expected background rates of a seizure phenotype, but a personal history of seizures was much more commonly found in those individuals with LQT2 (30 of 77 or 39%) than all other subtypes of LQTS (11 of 106 or 10%, p < 0.001) (Johnson et al. 2009). If this connection can be proven in other studies, the observation is potentially very signiἀcant. The gene, KCNH2, is responsible for LQT2. The KCNH2 channel was cloned originally from the hippocampus; it encodes a potassium channel active in hippocampal astrocytes and in the heart (Zehelein et al. 2004). Not only can this channel interact with some antiepileptic medications, e.g., phentoyin, phenobarbital, toprimate, and flunariÂ� zine (Danielsson et al. 2003; Trepakova et al. 2006; Yang et al. 2008), but it also interacts with abused drugs such as cocaine, methamphetamine, methadone, anabolic steroids, and even alcohol. This observation suggests that, in some cases of SUDEP, drug abuse may be the cause of death. That this observation has not yet been recorded in the literature only reflects the fact that full toxicology screening for abused drugs is not part of the routine investigation of SUDEP. Indeed, the autopsy is often performed even before a history of epilepsy is established (Kloster and Engelskjon 1999). Lathers et al. (1990) suggest, “… since both cocaine and epilepsy alone are associated with sudden unexpected death and since both are capable of modifying cardiac sympathetic neural discharge to produce changes in heart rate and rhythm, the question of whether the use of cocaine in the epileptic person places this individual at risk for sudden death must be raised.” It is also known that alcohol use is associated with an increased risk of autonomic dysfunction, seizure, and sudden death (Chan et al. 1990). It is clear that no single neurochemical system can adequately explain the complex nature of epileptic seizures. The same can be said for the neurochemical mechanisms that underlie alcohol withdrawal reactions in that there are complex, dynamic interactions among the neurotransmitters and neuromodulator systems in the brain. Another complicating factor is that

162 Sudden Death in Epilepsy: Forensic and Clinical Issues

occurrence of withdrawal seizures is only one component of the withdrawal syndrome. The challenge is to separate the neurochemical changes that may trigger seizures from those that are actually the result of the seizures themselves, or some other complication of the withdrawal reaction. Some of these factors include the individual’s genetic makeup, seizure threshold, health status, malnutrition, polydrug abuse, a history of epilepsy, prior alcohol withdrawal seizures, and trauma. Despite the difficulties associated with interpreting the neurochemical changes associated with ethanol withdrawal, there do appear to be similarities between the abnormalities observed in alcoholics and those postulated to be involved in the mechanism of epileptic seizures (Lathers et al. 1990). Twenty years have lapsed since Lathers et al. (1990) and Chan et al. (1990) discussed the possible interactions of cocaine and alcohol and SUDEP. The time is long overdue for us to answer these questions. The results of very recent discoveries make the effort to answer these questions even more worthwhile. When genome-wide data from ἀve different population-based cohorts, composed of 15,842 individuals of European ancestry, were analyzed, a total of 10 loci associated with the occurrence of LQTS were identiἀed. Four of these loci map near the monogenic LQTS genes: KCNQ1, KCNH2, SCN5A, and KCNJ2. Two other loci, ATP1B1 and PLN, have already been shown to be genes with established electrophysiological functions, whereas three of the newly discovered genes map to RNF207, near LITAF and within the NDRG4–GINS3–SETD6–CNOT1 complex, respectively. Until this study was undertaken, not a single one of these genes was thought to have anything to do with cardiac function (Newton-Cheh et al. 2009; Pfeufer et al. 2009). Any one of these genes could be responsible for LQTS, torsades de pointes, and sudden cardiac death (Pfeufer et al. 2009). The most striking thing about this new discovery is that it was replicated within the same week by another group of scientists who came up with exactly the same results (Newton-Cheh et al. 2009). If this sequence of events were to occur in a person with epilepsy, it would be called SUDEP, since the genetic component would not be detected by the medical examiner.€Even though these new discoveries very strongly suggest a nexus between SUDEP, channelopathies, and drug abuse, further understanding of the pathophysiology is required. Associations between LQT2 and epilepsy that had not previously been suspected are now known to exist. This raises the possibility that LQT2 perturbations in the KCNH2-encoded potassium channel may confer susceptibility for recurrent seizure activity (Johnson et al. 2009). When a cardiomyocyte depolarizes, the rapid component of the delayed rectiἀer K+ current, abbreviated as IKr, plays a key role in cell repolarization. During the plateau phase of the depolarization and repolarization cycle, the current generated through the IKr channel is small. As repolarization proceeds, a transient increase in the IKr outward current occurs due to fast recovery from inactivation and slow deactivation, ultimately leading to repolarization of the cardiac cell. The LQTII or hERG gene controls the IKr channel (Thomas et al. 2006). In congenital forms of LQTII syndrome, flow through the channel is slowed because the structure of the potassium pore itself is abnormal. As a consequence, the action potential is prolonged. This leads to the occurrence of early after depolarization currents that, in turn, can lead to lethal arrhythmias (torsades de pointes, literally “twisting of the points,” a form of ventricular tachycardia). However, LQTS can occur even in someone with a perfectly normal hERG gene because so many drugs, both licit and illicit, such as ἀrst and second generation antidepressants, interact with the channel (Sala et al. 2006). Since the

Drug Abuse and SUDEP

163

majority of antiepileptic drugs block voltage-dependent sodium and calcium channels, enhance GABAergic transmission, and/or antagonize glutamate receptors, it only seems reasonable to assume that an epileptic who abuses drugs is at far greater risk for arrhythmia, even if the effects of their antiseizure medications predominate. Although there is nothing in the literature to suggest that any of the commonly used anti-seizure medications cause QT prolongation, there is ample evidence that some abused drugs do. Cocaine, alcohol, and cocaethylene (a cocaine metabolite that is formed only in the presence of signiἀcant amounts of alcohol) all block the hERG potassium channel, effectively producing an acquired form of LQTII (Karle and Kiehn 2002). Whether they interact with any of the other recently discovered channels is not known. When hERG currents are measured in vitro using the patch-clamp technique, cocaethylene (the only metabolite of cocaine that is psychoactive) increases the QT interval and causes torsades de pointes (Ferreira et al. 2001; O’Leary 2002). Other studies have shown cocaethylene accelerates the inactivation of hERG current without affecting recovery from inactivation. Depending on the molecular conἀguration of an ion pore, there are several different ways to disrupt normal IKr function. Cocaine and its metabolite, cocaethylene, produce what is called open channel block, which prevents the channel closing by putting a “foot in the door” (Wang et al. 2000). Much the same mechanism has been proposed for quaternary ammonium compounds. Study of cocaethylene is particularly important because the IKr current is found to be altered, even at realistic plasma drug concentrations (O’Leary 2002). hERG K+ channels have also been found in neurons, putatively contributing to the neuropsychological behavior of individuals who have been drinking alcohol and/or using cocaine. Methamphetamine abuse also causes QT prolongation, even though it does not interact with the hERG channel (or any other known cardiac channel) (Haning and Goebert 2007). This somewhat perplexing result may be a consequence of catecholamine toxicity, as it is well established that epinephrine induces LQTS (Urao et al. 2004), and methamphetamine usage causes catecholamine release from synaptic vesicles, thereby increasing circulating levels of catecholamines (Stuerenburg et al. 2002). Central sympathetic stimulation, such as occurs during seizures, also causes release of catecholamines and could provide a plausible explanation for this phenomena (Eckard et al. 1999). Clinical studies show that cerebral infarction causes various cardiovascular and electrocardiographic abnormalities, depending on the location and the size of the infarct. The two most frequently encountered abnormalities in patients with stroke are QT interval prolongation and widening of the QRS complex. Disease of the insular cortex seems to be very important for activation of the sympathetic nervous system. Patients with brain stem infarction have substantially higher mean plasma norepinephrine levels than patients with hemispheric infarction; on the other hand, hemispheric lesions are associated with a signiἀcantly higher incidence of cardiac arrhythmias when compared to patients with brain stem infarction (Klingelhofer and Sander 1997). Whether any of these changes can be related to the mechanism that causes arrhythmias in epileptics is not known. Methadone has recently been added to the list of drugs shown to produce QT prolongation, arrhythmias, and sudden death. It has been known for several years that highdose methadone can induce QT prolongation (Kornick et al. 2003) by hERG inhibition. However, new evidence shows that QT prolongation can occur at much lower doses of methadone, even when the drug is not given intravenously. Methadone is a chiral€drug, but only (R)-methadone provides any pain relief. Laboratory experiments have shown that

164 Sudden Death in Epilepsy: Forensic and Clinical Issues

(S)-methadone, which does not bind the mu receptor or provide pain relief, blocks the hERG current three-and-a-half times more potently than (R)-methadone. If an individual is a CYP2B6 slow metabolizer (SM), they can still metabolize (R)-methadone normally, but they cannot metabolize the (S) form. Concentrations of the (S) form of methadone will continue to rise, eventually leading to hERG blockade and, potentially, torsades de pointes (Eap et al. 2007). 10.2.2â•…QT Shortening Pathologic reduction in the QT interval is much less common than prolongation but it is also associated with sudden death. Brugada et al. (2004) linked SQTS to a KCNH2 gene mutation, the same gene responsible for QT prolongation. This disorder is characterized by a corrected QT (QTc) interval that is shorter than normal (QTc ≤€320 ms), and it is often associated with atrial ἀbrillation, syncopal episodes, and/or sudden cardiac death in patients who are said to have no anatomic evidence of heart disease (Zareba and Cygankiewicz 2008). While there is general agreement about what constitutes a prolonged QT interval, the deἀnition of short QT (SQTS) remains somewhat controversial, though most accept that the lower boundary of normal is on the order of 320 ms. Unlike the mutations in LQTS, which result in loss of function, mutations in SQTS cause an increase in function with rapid repolarization; cardiac arrest may be the ἀrst symptom (Giustetto et al. 2006). The recent discovery that anabolic steroid abuse is associated with a reduction in the QT interval (Bigi et al. 2009) is most provocative, given the repeated observation that abnormalities of reproductive endocrine hormones are more often found in men with epilepsy than in the general population (Roste et al. 2005). There is an ongoing debate whether this increased risk in males can be attributed to the use of antiepileptic drugs or the epilepsy itself. The corrected QT interval in proven anabolic steroid abusers is signiἀcantly shorter than the QT interval of drug-free bodybuilders and that of sedentary men. In this recently study, sedentary men were found to have a QTc interval of 418 ± 23.6 ms, drug-free bodybuilders had a QTc interval of 422 ± 24.5 ms, and steroid abusing bodybuilders had a QTc interval of 367 ± 17.1 ms (p < 0.01). In fact, the correlation between steroid abuse and QT interval is so strong that some have recommended EKG screening as a method of detecting steroid abusers (Bigi et al. 2009). Similar ἀndings have been produced by the administration of anabolic steroids to experimental animals (Fulop et al. 2006; Liu et al. 2003).

10.3╅QT Dispersion Animal studies suggest that hippocampal norepinephrine transporters are downregulated when chronically exposed to cocaine (Kitayama et al. 2006), and human studies have shown a decrease in dopamine and serotonin transporters in the same areas (Mash et al. 2000). It should not be forgotten that all local anesthetics are Na channel blockers, both in the brain and the heart, and cocaine is a local anesthetic. Thus, it might be reasonable to postulate that some drug abusers who also have epilepsy, of which there appear to be more than a few (Opeskin et al. 2000),€die as a consequence of the drug abuse acting in synergy with their primary disease to cause seizure-related death.

Drug Abuse and SUDEP

165

10.4â•…Abnormal Catecholamine Metabolism Disruption of postganglionic reuptake is thought to be the mechanism that leads to bradyarrhythmia or even asystole during epileptic seizures. In the heart, as opposed to the brain, the action of norepinephrine is terminated primarily by reuptake (Kloner et al. 1992; Schafers et al. 1998) by the actions of catechol-O-methyl transferase (COMT). Recent studies of epileptics who have experienced bradyarrhythmia or asystole have shown that these individuals have dramatically abnormal postganglionic cardiac norepinephrine uptake (Kerling et al. 2009), suggesting impaired sympathetic cardiac innervation, resulting in a limited ability to adjust and modulate heart rate, or even cause asystole. A large number of commonly abused drugs (cocaine, methamphetamine, and 3,4-methylenedioxymethamphetamine) profoundly disrupt catecholamine metabolism, and their use may well be responsible for the death of an occasional drug abuser who also has epilepsy. However, the most current studies suggest variations in COMT activity are more likely due to genetic polymorphisms and, no doubt, more work will be done in this area in the near future (Haile et al. 2009).

References Ackerman, M. J., D. J. Tester, and D. J. Driscoll. 2001. Molecular autopsy of sudden unexplained death in the young. Am J Forensic Med Pathol 22 (2): 105–111. Akalin, F., A. Tirtir, and Y. Yilmaz. 2003. Increased QT dispersion in epileptic children. Acta Paediatr 92 (8): 916–920. Anderson, K. P. 2003. Sympathetic nervous system activity and ventricular tachyarrhythmias: Recent advances. Ann Noninvasive Electrocardiol 8 (1): 75–89. Aurlien, D., T. P. Leren, E. Tauboll, and L. Gjerstad. 2009. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18 (2): 158–160. Bigi, M. A., A. Aslani, and A. Aslani. 2009. Short QT interval: A novel predictor of androgen abuse in strength trained athletes. Ann Noninvasive Electrocardiol 14 (1): 35–39. Brugada, R., K. Hong, R. Dumaine, J. Cordeiro, F. Gaita, M. Borggrefe, T. M. Menendez et al. 2004. Sudden death associated with short-QT syndrome linked to mutations in HERG. Circulation 109 (1): 30–35. Chan, A. W. K., C. M. Lathers, and J. E. Leestma. 1990. Chapter 19. Alcohol, arrhythmias, seizures, and sudden death. In Epilepsy and Sudden Death, ed. C. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Cuddy, T. E., P. S. Halli, and R. B. Tate. 2009. QT dispersion and heart rate predict the risk of sudden unexpected cardiac death in men: The Manitoba Follow-Up Study. Prev Cardiol 12 (1): 27–33. Danielsson, B. R., K. Lansdell, L. Patmore, and T. Tomson. 2003. Phenytoin and phenobarbital inhibit human HERG potassium channels. Epilepsy Res 55 (1–2): 147–157. Eap, C. B., S. Crettol, J. S. Rougier, J. Schlapfer, L. Sintra Grilo, J. J. Deglon, J. Besson, M. CroquetteKrokar, P. A. Carrupt, and H. Abriel. 2007. Stereoselective block of hERG channel by (S)-methadone and QT interval prolongation in CYP2B6 slow metabolizers. Clin Pharmacol Ther 81 (5): 719–728. Eckardt, M., L. Gerlach, and F. L. Welter. 1999. Prolongation of the frequency-corrected QT dispersion following cerebral strokes with involvement of the insula of Reil. Eur Neurol 42 (4): 190–193. Ferreira, S., W. J. Crumb Jr., C. G. Carlton, and C. W. Clarkson. 2001. Effects of cocaine and its major metabolites on the HERG-encoded potassium channel. J Pharmacol Exp Ther 299 (1): 220–226.

166 Sudden Death in Epilepsy: Forensic and Clinical Issues Fischer, R., R. Dechend, A. Gapelyuk, E. Shagdarsuren, K. Gruner, A. Gruner, P. Gratze et al. 2007. Angiotensin II-induced sudden arrhythmic death and electrical remodeling. Am J Physiol Heart Circ Physiol 293 (2): H1242–H1253. Fulop, L., T. Banyasz, G. Szabo, I. B. Toth, T. Biro, I. Lorincz, A. Balogh, K. Peto, I. Miko, and P. P. Nanasi. 2006. Effects of sex hormones on ECG parameters and expression of cardiac ion channels in dogs. Acta Physiol (Oxf) 188 (3–4): 163–171. Gamouras, G. A., G. Monir, K. Plunkitt, S. Gursoy, and L. S. Dreifus. 2000. Cocaine abuse: Repolarization abnormalities and ventricular arrhythmias. Am J Med Sci 320 (1): 9–12. Gill, G. V., A. Woodward, I. F. Casson, and P. J. Weston. 2009. Cardiac arrhythmia and nocturnal hypoglycaemia in type 1 diabetes—The ‘dead in bed’ syndrome revisited. Diabetologia 52 (1): 42–45. Giustetto, C., F. Di Monte, C. Wolpert, M. Borggrefe, R. Schimpf, P. Sbragia, G. Leone et al. 2006. Short QT syndrome: Clinical ἀndings and diagnostic–therapeutic implications. Eur Heart J 27€(20): 2440–2447. Haider, A. W., M. G. Larson, E. J. Benjamin, and D. Levy. 1998. Increased left ventricular mass and hypertrophy are associated with increased risk for sudden death. J Am Coll Cardiol 32 (5): 1454–1459. Haile, C. N., T. R. Kosten, and T. A. Kosten. 2009. Pharmacogenetic treatments for drug addiction: Cocaine, amphetamine and methamphetamine. Am J Drug Alcohol Abuse 35 (3): 161–177. Haning, W., and D. Goebert. 2007. Electrocardiographic abnormalities in methamphetamine abusers. Addiction 102 (Suppl 1): 70–75. Henning, R. J., and J. Cuevas. 2006. Cocaine activates calcium/calmodulin kinase II and causes cardiomyocyte hypertrophy. J Cardiovasc Pharmacol 48 (1): 802–813. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14 (2): 280–287. Jehi, L., and I. M. Najm. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Cleve Clin J Med 75 (Suppl 2): S66–S70. John, B. T., B. K. Tamarappoo, J. L. Titus, W. D. Edwards, W. K. Shen, and S. S. Chugh. 2004. Global remodeling of the ventricular interstitium in idiopathic myocardial ἀbrosis and sudden cardiac death. Heart Rhythm 1 (2): 141–149. Johnson, J. N., N. Hofman, C. M. Haglund, G. D. Cascino, A. A. Wilde, and M. J. Ackerman. 2009. Identiἀcation of a possible pathogenic link between congenital long QT syndrome and epilepsy. Neurology 72 (3): 224–231. Karch, S. B., B. G. Stephens, and C. H. Ho. 1999. Methamphetamine-related deaths in San Francisco: Demographic, pathologic, and toxicologic proἀles. J Forensic Sci 44 (2): 359–368. Karch, S. B., B. Stephens, and C. H. Ho. 1998. Relating cocaine blood concentrations to toxicity—An autopsy study of 99 cases. J Forensic Sci 43 (1): 41–45. Karle, C. A., and J. Kiehn. 2002. An ion channel ‘addicted’ to ether, alcohol and cocaine: The HERG potassium channel. Cardiovasc Res 53 (1): 6–8. Kerling, F., M. Dutsch, R. Linke, T. Kuwert, H. Stefan, and M. J. Hilz. 2009. Relation between ictal asystole and cardiac sympathetic dysfunction shown by MIBG-SPECT. Acta Neurol Scand 120€(2): 123–129. Kitayama, T., L. Song, K. Morita, N. Morioka, and T. Dohi. 2006. Down-regulation of norepinephrine transporter function induced by chronic administration of desipramine linking to the alteration of sensitivity of local-anesthetics–induced convulsions and the counteraction by coadministration with local anesthetic. Brain Res 1096 (1): 97–103. Klingelhofer, J., and D. Sander. 1997. Cardiovascular consequences of clinical stroke. Baillieres Clin Neurol 6 (2): 309–335. Kloner, R. A., S. Hale, K. Alker, and S. Rezkalla. 1992. The effects of acute and chronic cocaine use on the heart. Circulation 85 (2): 407–419. Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67 (4): 439–444.

Drug Abuse and SUDEP

167

Kornick, C. A., M. J. Kilborn, J. Santiago-Palma, G. Schulman, H. T. Thaler, D. L. Keefe, A. N. Katchman et al. 2003. QTc interval prolongation associated with intravenous methadone. Pain 105 (3): 499–506. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Neurocardiologic mechanistic risk factors in sudden unexpected death in epilepsy. Ch. 1 in Sudden Death in Epilepsy: Forensic and Clinical Issues, eds. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Lathers, C. M., M. M. Spino, I. Agarwal, L. S. Y. Tyau, and W. B. Pickworth. 1990. Chapter 27. Cocaine-induced seizures, arrhythmias, and sudden death. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E., J. R. Hughes, S. S. Teas, and M. B. Kalelkar. 1985. Sudden epilepsy deaths and the forensic pathologist. Am J Forensic Med Pathol 6 (3): 215–218. Liu, X. K., A. Katchman, B. H. Whitἀeld, G. Wan, E. M. Janowski, R. L. Woosley, and S. N. Ebert. 2003. In€vivo androgen treatment shortens the QT interval and increases the densities of inward and delayed rectiἀer potassium currents in orchiectomized male rabbits. Cardiovasc Res 57 (1): 28–36. Mash, D. C., J. K. Staley, S. Izenwasser, M. Basile, and A. J. Ruttenber. 2000. Serotonin transporters upregulate with chronic cocaine use. J Chem Neuroanat 20 (3–4): 271–280. Meierkord, H., S. Shorvon, and S. L. Lightman. 1994. Plasma concentrations of prolactin, noradrenaline, vasopressin and oxytocin during and after a prolonged epileptic seizure. Acta Neurol Scand 90 (2): 73–77. Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16 (1): 1–7. Nemec, J., M. Buncova, V. Shusterman, B. Winter, W. K. Shen, and M. J. Ackerman. 2009. QT interval variability and adaptation to heart rate changes in patients with long QT syndrome. Pacing Clin Electrophysiol 32 (1): 72–81. Newton-Cheh, C., M. Eijgelsheim, K. M. Rice, P. I. de Bakker, X. Yin, K. Estrada, J. C. Bis et al. 2009. Common variants at ten loci influence QT interval duration in the QTGEN Study. Nat Genet 41 (4): 399–406. O’Leary, M. E. 2002. Inhibition of HERG potassium channels by cocaethylene: A metabolite of cocaine and ethanol. Cardiovasc Res 53 (1): 59–67. Opeskin, K., A. S. Harvey, S. M. Cordner, and S. F. Berkovic. 2000. Sudden unexpected death in epilepsy in Victoria. J Clin Neurosci 7 (1): 34–37. P-Codrea Tigaran, S., S. Dalager-Pedersen, U. Baandrup, M. Dam, and A. Vesterby-Charles. 2005. Sudden unexpected death in epilepsy: Is death by seizures a cardiac disease? Am J Forensic Med Pathol 26 (2): 99–105. Pfeufer, A., S. Sanna, D. E. Arking, M. Muller, V. Gateva, C. Fuchsberger, G. B. Ehret et al. 2009. Common variants at ten loci modulate the QT interval duration in the QTSCD Study. Nat Genet 41 (4): 407–414. Rona, G. 1985. Catecholamine cardiotoxicity. J Mol Cell Cardiol 17 (4): 291–306. Roste, L. S., E. Tauboll, L. Morkrid, T. Bjornenak, E. R. Saetre, T. Morland, and L. Gjerstad. 2005. Antiepileptic drugs alter reproductive endocrine hormones in men with epilepsy. Eur J Neurol 12 (2): 118–124. Rothenbuhler, A., C. P. Bibal, S. Le Fur, and P. Bougneres. 2008. Effects of a controlled hypoglycaemia test on QTc in adolescents with Type 1 diabetes. Diabet Med 25 (12): 1483–1485. Sala, M., F. Coppa, C. Cappucciati, P. Brambilla, G. d’Allio, E. Caverzasi, F. Barale, and G. M. De Ferrari. 2006. Antidepressants: Their effects on cardiac channels, QT prolongation and Torsade de Pointes. Curr Opin Investig Drugs 7 (3): 256–263. Schafers, M., D. Dutka, C. G. Rhodes, A. A. Lammertsma, F. Hermansen, O. Schober, and P. G. Camici. 1998. Myocardial presynaptic and postsynaptic autonomic dysfunction in hypertrophic cardiomyopathy. Circ Res 82 (1): 57–62. So, N. K., and M. R. Sperling. 2007. Ictal asystole and SUDEP. Neurology 69 (5): 423–424.

168 Sudden Death in Epilepsy: Forensic and Clinical Issues Stuerenburg, H. J., K. Petersen, T. Baumer, M. Rosenkranz, C. Buhmann, and R. Thomasius. 2002. Plasma concentrations of 5-HT, 5-HIAA, norepinephrine, epinephrine and dopamine in ecstasy users. Neuro Endocrinol Lett 23 (3): 259–261. Tellez-Zenteno, J. F., L. H. Ronquillo, and S. Wiebe. 2005. Sudden unexpected death in epilepsy: Evidence-based analysis of incidence and risk factors. Epilepsy Res 65 (1–2): 101–115. Thomas, D., C. A. Karle, and J. Kiehn. 2006. The cardiac hERG/IKr potassium channel as pharmaÂ� cological target: Structure, function, regulation, and clinical applications. Curr Pharm Des 12€(18): 2271–2283. Trepakova, E. S., S. J. Dech, and J. J. Salata. 2006. Flunarizine is a highly potent inhibitor of cardiac hERG potassium current. J Cardiovasc Pharmacol 47 (2): 211–220. Tu, E., S. M. Twigg, and C. Semsarian. 2008. Sudden death in type 1 diabetes: The mystery of the ‘dead in bed’ syndrome. Int J Cardiol. Urao, N., H. Shiraishi, K. Ishibashi, M. Hyogo, M. Tsukamoto, N. Keira, S. Hirasaki, T. Shirayama, and M. Nakagawa. 2004. Idiopathic long QT syndrome with early after depolarization induced by epinephrine. A case report. Circ J 68 (6): 587–591. Uyarel, H., C. Ozdol, A. M. Gencer, E. Okmen, and N. Cam. 2005. Acute alcohol intake and QT dispersion in healthy subjects. J Stud Alcohol 66 (4): 555–558. Wang, J., C. D. Myers, and G. A. Robertson. 2000. Dynamic control of deactivation gating by a soluble amino-terminal domain in HERG K(+) channels. J Gen Physiol 115 (6): 749–758. Yang, Z. Q., J. C. Barrow, W. D. Shipe, K. A. Schlegel, Y. Shu, F. V. Yang, C. W. Lindsley et al. 2008. Discovery of 1,4-substituted piperidines as potent and selective inhibitors of T-type calcium channels. J Med Chem 51 (20): 6471–6477. Yap, Y. G., E. R. Behr, and A. J. Camm. 2009. Drug-induced Brugada syndrome. Europace 11 (8): 989–994. Zareba, W., and I. Cygankiewicz. 2008. Long QT syndrome and short QT syndrome. Prog Cardiovasc Dis 51 (3): 264–278. Zehelein, J., D. Thomas, M. Khalil, A. B. Wimmer, M. Koenen, M. Licka, K. Wu et al. 2004. Identiἀcation and characterisation of a novel KCNQ1 mutation in a family with Romano–Ward syndrome. Biochim Biophys Acta 1690 (3): 185–192.

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

11

Claire M. Lathers Michelle M. Spino Isha Agarwal Laurie S. Y. Tyau Wallace B. Pickworth

Contents 11.1 Introduction 11.2 Mechanisms of Action of Cocaine 11.3 Cocaine-Induced Sudden Death 11.3.1 Cocaine-Induced Changes in Mean Arterial Blood Pressure and Heart Rate 11.3.2 Cocaine-Induced Myocardial Ischemia, Infarction, Arrhythmia, and Cardiomyopathies 11.3.3 Cocaine-Induced Changes in Postganglionic Cardiac Sympathetic Neural Function 11.3.4 Central Actions of Cocaine 11.3.5 Cocaine-Induced Seizures 11.4 Treatment of Cocaine-Induced Arrhythmias and Seizures 11.5 Use of Cocaine in Persons with Epilepsy 11.6 Summary References

169 170 172 172 174 176 176 179 180 181 181 182

11.1â•…Introduction The presence of coca leaves in the tombs of South American Indian mummies suggests that cocaine was used as early as a .d. 600. The use of cocaine is prevalent in modern society. Cregler and Mark (1986a) reviewed the demographics of current cocaine users and found that approximately 1 out of every 10 Americans, have used cocaine at least once (Cregler and Mark 1987). The fallacy that cocaine is a benign, nonaddicting substance may be part of the reason for the alarming rise in abuse (Cregler and Mark 1986b). Although cocaine has been found to be a cardiotoxin, the pathogenesis of this toxicity is not well deἀned (Cregler and Mark 1987). Cocaine use has also been linked to the occurrence of subarachnoid hemorrhage, hypertension, ventricular arrhythmia, tachycardia, acute myocardial infarction, seizure, and sudden death (Lichtenfeld et al. 1984; Nahas et al. 1985; Tazelaar et al. 1987; Young and Glauber 1947). Persons with epilepsy have been shown to manifest autonomic dysfunctions similar to those manifested by cocaine users, 169

170 Sudden Death in Epilepsy: Forensic and Clinical Issues

including changes in blood pressure and heart rate and rhythm, phenomena that may be contributory to sudden unexpected death (Leestma et al. 1984; Penἀeld and Erickson 1941; Phizackerly et al. 1954; Walsh et al. 1968). Thus, one must ask whether the use of cocaine in individuals with epilepsy places these individuals at risk of dying in a sudden unexplained manner.

11.2â•…Mechanisms of Action of Cocaine Cocaine (Figure 11.1), extracted from the leaves of Erythroxylon coca, is a potent local anesthetic agent (Cregler and Mark 1986a; Gould et al. 1985) possessing membrane-stabilizing effects at low plasma levels (Tazelaar et al. 1987). It is also a sympathomimetic agent at higher plasma concentrations (Benchimol et al. 1978; Duke 1986). Cocaine ampliἀes the effect of catecholamines by blocking the reuptake at the synaptic junctions, causing a local excess of norepinephrine at the synaptic cleft. As a result of the excess of norepinephrine at the nerve terminal, there is a prolongation and potentiation of the activity of norepinephrine (Weiss 1986). Norepinephrine is the primary neurotransmitter of the sympathetic nervous system. Excitation of the sympathetic nervous system produces physiological characteristics, such as mobilization of adrenal catecholamines, causing an increase in blood pressure and the heart rate, dilatation of the pupils, a rise in blood sugar levels, vasoconstriction of vessels in the brain and muscles, tightening of the sphincters, and an elevation of body temperature. The intense peripheral vasoconstriction retards reabsorption. Drug effects include intense euphoria and elation, garrulousness, excitability, and irritability; with repeated administration, paranoid ideation, delirium, and assaultiveness occur. Table 11.1 summarizes the actions of cocaine on the cardiovascular, respiratory, and central nervous system (Gay 1982). Cocaine as a hydrochloride salt is brought into the United States with purity ranging up to 95% (Gay 1982). The purity is decreased to 25–90% of its original state through the addition of diluents and adulterants such as procaine, lidocaine (Cregler and Mark 1986b), caffeine, benzocaine, amphetamines, heroin, quinine, talc, and phencyclidine. All adulterants contribute to the toxicity of cocaine. Finally, it is combined with sugars such as mannitol, lactose, and glucose to attain a ἀnal volume and weight (Gay 1982). The resulting cocaine street product can be administered by various routes, including intravenous and subcutaneous injections, intranasal inhalation (snorting), and the current vogue of smoking a “freebase” form of cocaine (crack). Freebase smoking or intravenous injections of

CH3 N

COOCH3

OOCC6H5 H

Figure 11.1╇ Structure of cocaine (C17 H 21NO4).

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

171

Table 11.1â•… The Cocaine Reaction Phase I: Early stimulation

II: Advanced stimulation

III: Depressive

Central Nervous System

Cardiovascular System

Euphoria: stated feelings of “soaring,” well-being Elation, expansive good humor, laughing, mydriasis Talkative, garrulous Excited, flighty, emotionally unstable Restless, irritable, apprehensive, unable to sit still Stereotyped movements (such as “picking” or “stroking”), bruxism Nausea, vomiting, vertigo Sudden headache Cold sweats Tremor (nonintentional) Twitching of small muscles, especially of the face, ἀngers, and feet Tics, generalized Preconvulsive, tonic and clonic jerks Possible psychosis, hallucinations Core body temperature rises Verbalization of impending doom (precedes imminent total collapse) Unresponsive to voice; decreased responsiveness to all stimuli Increased deep tendon reflexes Generalized hyperreflexia Convulsions: tonic and clonic Status epilepticus Incontinence Malignant encephalopathy possible

Pulse vanes at ἀrst, may immediately slow because of reflex vagal effect; will increase 30% to 50% above normal with system absorption of 25 mg of cocaine Blood pressure usually elevates 15% to 20% above normal with similar dosages as noted above Skin pallor caused by vasoconstriction Premature ventricular contractions

Increased respiratory rate and depth Dyspnea

Increased pulse and blood pressure: high output failure possible Blood pressure falls as ventricular dysrhythmias supervene and inefficient cardiac output results Pulse becomes rapid, weak, and irregular Peripheral, then central cyanosis Ventricular ἀbrillation Circulation failure Ashen gray cyanosis No palpable pulse Cardiac arrest Paralysis of medullary brain center Exitus

Gasping, rapid, or irregular respiration (Chevne–Stokes)

Flaccid paralysis of muscles Coma Pupils ἀxed and dilated Loss of reflexes Loss of vital support functions Paralysis of medullary brain center Exitus

Respiratory System

Agonal gasps Respiratory failure Gross pulmonary edema Paralysis of medullary brain center Exitus

cocaine cause a euphoric or “rush” experience, which occurs within 45 s and is associated with a rapid increase in plasma cocaine concentrations. The effect lasts for approximately 20 min. In contrast, intranasal administration results in euphoria occurring within 3 to 5 min of administration and lasts for 1 to 1.5 h (Van Dyke and Byck 1983). Regardless of the route of administration, accounts of cocaine-induced sudden death have become common.

172 Sudden Death in Epilepsy: Forensic and Clinical Issues

11.3â•…Cocaine-Induced Sudden Death Sudden death has been shown to be induced by cocaine (Amon et al. 1986; Estroff and Gold 1986; Mittleman and Welti 1987; Welti and Fishbain 1985). Reports indicate 1.2 g to be a lethal dose; however, severe toxic effects have been reported with doses as low as 20 mg (Estroff and Gold 1986). Because it is so sudden, medical personnel do not ordinarily witness cocaineinduced death; victims usually collapse and die before resuscitation efforts can begin. Confusion or convulsions precede death induced by cocaine. Estroff and Gold (1986) reported seven cases of sudden death associated with the use of cocaine, in whom a state of excited delirium was the fatal symptom. The initial symptom was intense paranoia, followed by bizarre and violent behavior necessitating the use of force to restrain the patient. The unexpected outbursts of strength were associated with hyperthermia, which was thought to be due to a direct effect of cocaine on the central nervous system center for temperature regulation, and due to peripheral vasoconstriction, with resultant reduction in heat (Ritchie and Greene 1980). Status epilepticus, respiratory paralysis, or cardiac arrhythmias genrally precede sudden death induced by cocaine. Abramowicz (1986) suggested that most sudden deaths associated with cocaine use are caused by seizures leading to anoxia. Recent clinical data have been correlated with pathological ἀndings, generating several hypotheses that attempt to deἀne forensically the pathological mechanisms of cocaine-induced sudden death. 11.3.1â•…Cocaine-Induced Changes in Mean Arterial Blood Pressure and Heart Rate The circulatory effects of cocaine are believed to be of both central and peripherally induced vasoconstriction and cardioacceleration (Young and Glauber 1947). Change in heart rate is a sensitive measure of cocaine-induced cardiovascular effect (Fischman et al. 1976). Cocaine results in dose-related changes in heart rate (Javiad et al. 1978), with small doses decreasing heart rate via central vagal action and moderate doses increasing heart rate via atrial and peripheral sympathetic stimulation (Benchimol et al. 1978). Extremely high intravenous doses have direct toxic effects on the heart and cause immediate death (Nanji and Filipenko 1984; Young and Glauber 1947). The duration of the cardiovascular action is dependent on the dose of cocaine. Fischman et al. (1976) showed that an increase in heart rate was evident after intravenous injections of varying doses of cocaine; the increase began 2–5 min after infusion, peaked at 10 min, and rapidly returned to baseline (Figure 11.2). Cocaine also increased blood pressure in a dose-related manner, but more variability is seen in this measure. In one study (Pitts et al. 1987), cocaine was administered intravenously and evoked a rapid, transient, dose-dependent rise in mean arterial pressures (Figure 11.3). Jain et al. (1987) reported that the administration of cocaine (0.25 mg/kg, i.v.) to anesthetized cats increased systolic and diastolic blood pressure by 33 ± 11 and 31 ± 7 mm Hg, respectively. The dose also enhanced the pressor responses to intravenous norepinephrine and to bilateral carotid occlusion. Doses of 0.5 and 1.0 mg/kg (i.v.) also caused an increase in blood pressure and responses to intravenous norepinephrine but did not increase the blood pressure response to bilateral carotid occlusion. Higher doses had no additive effect on the blood pressure, but rather slowed the heart rate, attenuated blood pressure responses to norepinephrine, prolonged the QRS duration, and decreased tidal volumes. All effects

Mean heart rate (beats/min)

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50

173 4 mg 8 mg 16 mg 32 mg Saline

Pre-drug

0

8

16

24

32

40

48

56

Time since drug injection (min)

Figure 11.2╇ Mean heart rate as a function over time after cocaine is injected. 40 20 (a)

0 –20 –40 –60

Mean percent change

40 20 (b)

0 –20 –40 –60 40 20

(c)

0 –20 –40 –60

0

M 1

2

3

5 10 15 30

(min)

Figure 11.3╇ Time course for the cardiovascular and respiratory effects of three different doses of

cocaine: 0.312, 1.25, and 5 mg/kg (i.v.) depicted in (a) (N = 6), (b) (N = 5), and (c) (N = 4), respectively. The ordinate scale is percentage of change and the abscissa scale is time in minutes. M€= period during maximal pressor response. Squares represent mean values for arterial pressure. Vertical lines represent standard error of the mean. All animals were anesthetized with pentobarbital (6.5 mg/kg, i.p.). (From Pitts, D.K., Udom, C.E., and Marwah, J., Life Sci., 40, 1099–1111, 1987.)

174 Sudden Death in Epilepsy: Forensic and Clinical Issues

were increased after a dose of 4 mg/kg or greater of cocaine i.v., with arrhythmias occurring with 4 and 8 mg/kg. Doses as low as 0.25 mg/kg (i.v.) evoked substantial cardiovascular responses and lethal responses of apnea. In another study, the administration of cocaine intravenously to conscious rats increased arterial blood pressure (Rockhold et al. 1987). The heart rate was elevated initially but subsequently was decreased. With the onset of cocaine-induced seizures, a further elevation in heart rate and blood pressure occurred, ultimately progressing to cardiovascular collapse and death. Preliminary studies utilizing intravenous administration of cocaine to anesthetized dogs elicited a dose-dependent increase in blood pressure and heart rate and alterations in the ST segment (Tackett and Jones 1987). These changes were associated with elevated cerebrospinal fluid levels of norepinephrine and dopamine, ἀndings that suggest a role for central catecholaminergic mechanisms in the cardiovascular actions of cocaine. Therefore, as cocaine raises the blood pressure and heart rate to excessively high levels, there is an increased risk of aneurism, arteriovenous malformation, and stroke or hemorrhage from ruptures of cerebral arteries weakened by drug-related arteritis. 11.3.2â•…Cocaine-Induced Myocardial Ischemia, Infarction, Arrhythmia, and Cardiomyopathies There has been a recent and dramatic increase in cardiac abnormalities among cocaine users (Duke 1986; Wiener and Putnam 1987; Wiener et al. 1986) that has raised questions concerning the effect of cocaine on the cardiovascular system. Indeed, cocaine is clearly cardiotoxic, being temporally linked to myocardial ischemia, arrhythmias, and many cardiomyopathies. Cocaine use in the presence of preexisting coronary artery disease may predispose the individual to the development of angina, arrhythmias, or myocardial infarction (Coleman et al. 1982; Young and Glauber 1947). It is possible that a patient with hypercholesterolemia who is using cocaine may be further increasing the likelihood of coronary artery spasm (Rosendorff et al. 1981), leading to myocardial ischemia and necrosis. Numerous cases of suspected cocaine-induced myocardial ischemias and infarctions have been reported (Isner et al. 1985, 1986; Kassowsky and Lyon 1984; Mathias 1986; Rod and Zucker 1987; Rollingher et al. 1986; Schachne et al. 1984; Simpson and Edwards 1986). Simpson and Edwards (1986) reported a case of a 21-year-old man with a history of recreational intravenous cocaine abuse who developed chest pain within 1 min and cardiopulmonary collapse within 1 h after injection of cocaine. Postmortem ἀndings revealed severe coronary obstructive lesions and acute platelet thrombosis, with secondary chronic and acute myocardial ischemic lesions, focal endothelial injury, and platelet aggregations being observed. The author proposed that coronary artery spasm induced by cocaine caused the endothelial lesions and favored platelet adherence and aggregation. The chronic obstructions that were also found may have resulted from a similar mechanism. According to Weiss (1986), ἀxed coronary atherosclerotic lesions play a permissive role in the induction of coronary vasospasm. It has been proposed that the ability of both intrinsic atherosclerotic plaques and cocaine-induced norepinephrine uptake blockade increases local levels of catecholamine, producing coronary vasospasms. Furthermore, preexisting coronary artery disease sensitizes the vascular smooth muscle to norepinephrine-induced vasoconstriction, predisposing the cocaine user to life-threatening ischemia (Gould et al. 1985; Weiss 1986). Also, with chronic cocaine abuse, the excessive accumulation of norepinephrine may prime the myocardium for a fatal arrhythmia (Tazelaar et al. 1987).

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

175

Cocaine increases the local concentrations of catecholamines from blocked adrenergic nerve endings to other cell receptors by inhibiting the neuronal uptake of norepinephrine. Thus the adrenergic response in susceptible organs is increased, leading to the development of catecholamine supersensitivity (Benchimol et al. 1978; Trendelenburg 1968). Therefore, cocaine is capable of eliciting both an inhibitory and an excitatory response of sympathetically innervated structures to endogenous and exogenous catecholamines (Pitts et al. 1987). It has also been suggested that cocaine accentuates the action of norepinephrine on beta receptors in the heart (Nanji and Filipenko 1984) by increasing the concentration of norepinephrine at the synaptic cleft. Beta stimulation increases automaticity, heart rate, and the conduction velocity of the His–Purkinje system and decreases atrioventricular nodal refractoriness (Tazelaar et al. 1987). Tazelaar et al. (1987) characterized cocaine-induced pathophysiology in a postmortem study of 30 cocaine-related deaths. Morphologic characteristics of acute ischemia were observed in 93% of the cases. These involved the formation of myocardial contraction bands in association with polymorphonucleocytes in the initial 12–24 h; they were replaced with lymphocytes by 24–48 h. The formation of contraction bands and myocardial interstitial ἀbrosis may be one pathogenic mechanism for fatal arrhythmias. These contraction bands may also represent an anatomical route for reentrant mechanism, thus priming the heart for fatal arrhythmias. It is quite possible that arrhythmias may be generated by cocaine through a decrease in the refractoriness of the myocardial ἀbers, accumulation of excess norepinephrine, formation of contraction bands, and production of interstitial ἀbrosis. Cocaine causes myocardial ischemia by direct and indirect actions (Vitullo et al. 1987). Direct effects are the stimulation of the sinoatrial node, with an increase in heart rate, contractility, and wall tension. Indirectly, the effects can be due to the sympathetic vasoconstriction of the peripheral smooth muscle vasculature. Arterial vasoconstriction leads to an increase in afterload and blood pressure, which, in turn, increases the work that the heart must pump against. Both the direct and the indirect actions increase the oxygen consumption of the myocardium, with an ischemic event occurring when the demand for oxygen supersedes the supply (Tyau and Lathers 1988). Pasternack et al. (1985) reported three male patients in their middle to late thirties who were referred for coronary angiography after having angina pectoris and/or an acute myocardial infarction, coincident with an increase in the frequency of cocaine abuse. The onset of angina and acute myocardial infarctions may have been caused by a cocaineinduced potentiation of the activities of the central nervous system resulting in systemic hypertension and tachycardia. Isner et al. (1986) reported a temporal relationship between cocaine and cardiac sequelae in seven nonintravenous cocaine abusers. It was concluded that cocaine may precipitate fatal arrhythmias, myocarditis, acute infarctions, and possible sudden death in patients with either anatomically normal or abnormal coronary arteries. In these cases there was a temporal relationship between the administration of cocaine and the onset of a myocardial ischemia and/or infarction. Because of many other medical factors involved, it is difficult to discern the actual etiological mechanisms of the infarctions. However, in a case report by Howard et al. (1985), a young woman with normal coronary arteries, blood glucose, and lipid levels and no history of cardiovascular disease or smoking was admitted to the hospital for loss of consciousness and epigastric pain; 5€h earlier, she had inhaled 1.5 g of cocaine. On admission, an ECG showed precordial ST segment elevation and a loss of R waves. On the day following admission, an echocardiography revealed akinesis and dyskinesis of the left ventricle apex and septum. In this

176 Sudden Death in Epilepsy: Forensic and Clinical Issues

healthy individual with no coronary risk factors or demonstrable coronary artery disease, infarction occurred. This ἀnding suggests that cocaine-induced myocardial infarctions should be considered when examining individuals who may not appear to be vulnerable. The cardiovascular events produced through cocaine abuse can be seen to involve a range of pathological responses, including coronary vasospasm, arrhythmia, myocardial ischeÂ� mia, infarction, and cardiomyopathies. Therapeutic use of cocaine is not without risk. For example, Chiu et al. (1986) reported a patient who was anesthesized for a closed reduction of a nasal fracture by spraying 2 ml of 1% cocaine solution into the nasal airways. The patient complained of an acute onset of chest pain and shortness of breath. An electrocardiogram indicated ST-T wave changes in the precordial leads, suggestive of an acute coronary ischemic event. The rise in the MB creatine kinase fraction and reversed LDH isoenzyme fractional values were consistent with a small nontransmural myocardial infarction. The published accounts of cardiovascular events, myocardial infarction, and mortality related to cocaine use as described here and consist mostly of case reports (Loveys 1987). Experimental research looking into the cardiovascular effects of cocaine is warranted. 11.3.3â•…Cocaine-Induced Changes in Postganglionic Cardiac Sympathetic Neural Function Cocaine potentiates the ganglionic blocking action of norepinephrine (Christ et al. 1982). In the isolated hamster stellate ganglia preparation, cocaine exaggerated the inhibitory action of exogenously applied norepinephrine. In pithed rats, cocaine potentiated the pressor effect of norepinephrine more than it potentiated the pressor effect of sympathetic stimulation (Bayorh et al. 1983). Cocaine increased plasma norepinephrine levels and extended the inotropic and chronotropic responses to sympathetic neural stimulation in anesthetized dogs (Matsuda et al. 1980). These actions were attributed to the inhibition of the neuronal catecholamine uptake by cocaine (Matsuda et al. 1980). It is possible that cocaine-induced exaggeration of sympathetic discharge may enhance the arrhythmias experimentally caused by ouabain (Lathers et al. 1977), coronary occlusion (Lathers et al. 1978), and seizures (Lathers and Schraeder 1982; Schraeder and Lathers 1983). Experimental arrhythmias are hypothesized to be a useful model to study possible mechanisms of sudden death associated with myocardial infarctions (Lathers et al. 1986) and epilepsy (Lathers and Schraeder 1987; Schraeder and Lathers 1989). Consequently, any changes induced by cocaine in cardiac sympathetic neural discharge may well augment the development of arrhythmias and/or sudden death. On the other hand, Dart et al. (1983) demonstrated that stimulation of postganglionic cardiac sympathetic nerves in a Langendorff rat–isolated heart preparation produced a stimulation frequency-dependent overflow of endogenous norepinephrine into the venous effluent with an increase in the heart rate. Cocaine signiἀcantly reduced the norepinephrine outflow while the heart rate continued to increase. 11.3.4â•…Central Actions of Cocaine Many of the effects of cocaine result from actions in the central nervous system. Cocaineinduced euphoria, for example, was among the ἀrst effects described (Freud 1884) and is the most well-known central effect. Generalized convulsions, which often precede

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

177

cocaine-induced death, unfortunately are less well-known outside the medical community. It has been shown that seizures are a major determinant of cocaine-induced death (Catravas and Waters 1981; Catravas et al. 1978). The common concomitants of generalized seizures, including hyperthermia, acidosis, increased blood pressure, cardiac arrhythmia, and hypoventilation, may be responsible for the lethality. In animal studies, many of the effects of cocaine have been localized to limbic structures. For example, Castellani et al. (1983) found that cocaine initiated high-voltage spindles that began in the amygdala 5–25 s after the injection of cocaine (Figure 11.4) and spread within seconds to other olfactory sites in synchronous bursts. Cocaine induced an increase in spindle frequency in the olfactory bulb and amygdala that was inhibited by atropine administration. Yasuda et al. (1984) reported that low concentrations of cocaine potentiated a norepinephrine-induced increase in spike amplitude of hippocampal splices. The authors concluded that the action was due to the inhibition of catecholamine uptake based on the observation that other inhibitors had the same effect. Lesse and Collins (1979) found that cocaine increased the speed at which epileptiform discharges spread to the amygdala and hippocampus. They postulated that subconvulsive doses of cocaine have an excitatory effect on limbic structures, which increases their sensitivity to repetitive discharges from distant foci. Matsuzaki (1978) reported that chronic high doses of cocaine in the rhesus monkey engendered persistent behavioral depression, with cortical and limbic slowing of EEG. They concluded that it was the action of cocaine on limbic structures that played an important role in the persistence of these effects. Overall, the evidence indicates that cocaine enhances the propagation of limbic seizures. Since such activity has been associated with cardiac arrhythmia (Lathers and Schraeder 1982; Schraeder and Lathers 1983), it is quite possible that cocaine-induced seizures could be a factor in the deaths of persons using the drug. The cortical EEG effects of cocaine in humans were among the earliest documented effects of the drug (Berger 1937). Cocaine increases power in the fast frequency (beta bands) of the resting EEG after subcutaneous, intravenous, or oral administration (Berger 1937; Herning et al. 1985). Four-hour intravenous infusions of high doses of cocaine in humans sustains the increase in EEG beta power (Pickworth et al. 1986). The increase of power in

1 min after injection

7th week after cocaine initiation

µV/mm 15

L OLF BULB

7.5

R OLF BULB

7.5

L AMYG

15

R AMYG

10

L OLF TUB

7.5

L ACCUMB 1s

10 mm

Figure 11.4╇ Electrographic amygdala–olfactory spindling and spike response to 5 mg/kg (i.v.) cocaine during preseizure behaviors.

178 Sudden Death in Epilepsy: Forensic and Clinical Issues

the fast EEG frequency bands is ordinarily associated with increased attention, vigilance, or arousal. Pickworth et al. (1986) measured subjective, cardiovascular, and EEG effects of large (60 mg) intravenous doses of cocaine in human volunteers. Although the subjective report of “rush” lasted for only a few moments, the pressor effect and tachycardia persisted for up to 60 min (Figure 11.5). The rush, or intense cocaine-induced euphoria, is the effect for which the drug is self-administered. It is quite probable that inadvertent overdosage may occur when subjects readminister cocaine at a time when the cardiovascular and central nervous systems are at jeopardy. In reviewing the effect of cocaine on the electrophysiology of the central monoaminergic neurons, Pitts and Marwah (1986) found that intravenous cocaine activated cerebellar Purkinje neurons and inhibited serotonergic dorsal raphe and noradrenergic locus coeruÂ� leus neurons. The authors concluded that cocaine-induced increases in the mean arterial blood pressure were correlated with changes in the discharge of the central neurons. Pitts and Marwah (1987) also found that reserpine pretreatment diminished the inhibitory effects of intravenous cocaine on neuronal discharges in the locus coeruleus and dorsal raphe as well as the excitatory action of cocaine on the cerebellar Purkinje neurons. Thus, although stimulation of the inhibitory locus coeruleus afferent input to the cerebellar Purkinje neurons can reduce the activity of the Purkinje neurons via a betaadrenoceptor mechanism, intravenous cocaine (1 mg/kg) did not precipitate the inhibitory actions of locus coeruleus stimulation on cerebellar Purkinje neurons. This dose of cocaine also did not potentiate the inhibitory effects of iontophoretically applied norepinephrine or GABA on cerebellar Purkinje neurons. Locus coeruleous neurons were inhibited by intravenous cocaine (1 mg/kg) in conscious animals paralyzed with gallamine. It was proposed that intravenous cocaine (1 mg/kg) reduced impulse flow in locus coeruleus neurons, possibly through an alpha2-autoreceptor mechanism, without augmenting the effect

Systolic pressure (mm Hg)

4

120

160

105

150

90

140

2

75

130

1

60

120

0

Heart rate (beats/min) “Rush”

–30

–15

0 Cocaine

15

30

45

60

75

3

90

105

120

135

150

Time (min)

Figure 11.5╇ Effects of intravenous cocaine (60 mg) in seven drug-experienced volunteers. The high dose caused a transient “rush” (drug-induced euphoria) but prolonged increases in blood pressure and heart rate.

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

179

of norepinephrine at the level of the noradrenergic terminals impinging on postsynaptic cerebellar Purkinje neurons. The action of cocaine on norepinephrine-containing locus coeruleus neurons was also evaluated in freely moving, unanesthetized cats (Trulson and Trulson 1987). Cocaine eÂ�licited a dose-dependent reduction in the activity of the neurons, which was suppressed by a prior administration of an alpha2 antagonist, piperoxane. Also, the activity of the locus coeruleus remained unchanged by the administration of the structurally related local anesthetic agent procaine. It was concluded that the local anesthetic actions of cocaine were not the inhibitory factor in its effect on the activity of norepinephrine-containing nÂ�eurons. Nevertheless, as Yasuda et al. (1984) stated, “It is remarkable that its effects upon the electrophysiological activity of the brain remain virtually unknown.” Whether cocaineiÂ�nduced changes in the activity of central neurons contributes to sudden death remains to be determined. 11.3.5â•…Cocaine-Induced Seizures Seizures have been shown to play an essential role in the pathophysiology of cocaine toxicity. A major determinant of lethality in cocaine-treated dogs was the presence of seizures (Catravas and Waters 1981; Catravas et al. 1978). Cocaine infusions produced prolonged seizures that led to lactic acidosis and hyperthermia prior to death while cardiac output, systemic vascular resistance, respiration, and oxygenation were stable just prior to death. Seizures and death could have been prevented with pretreatment using diazepam, highdose chlorpromazine, or neuromuscular blockade with pancuronium (Antelman et al. 1981; Fekete and Borsy 1971). The use of diazepam is particularly important since it also counteracts the sympathomimetic action of cocaine on the heart. Treatment of acidosis alone did not prevent death, unless the animals were maintained in a hypothermic state. Jonsson et al. (1983) reported one patient who, as a result of cocaine intoxication, showed combined metabolic and respiratory acidosis consequent to seizures and hypoventilation. Improved ventilation and the administration of bicarbonate reversed the hypotension and accelerated idioventricular rhythm to sinus rhythm. Blood pH increased from 6.33 to normal and the pCO2 decreased from 70 to 46 mm Hg. Jonsson et al. concluded that respiratory arrest compounded the acidosis in patients intoxicated with cocaine and may have contributed signiἀcantly to their deaths. Acidosis has a particularly negative effect on myocardial contractility (Fabiato and Fabiato 1978; Spivey et al. 1985) and acidosis can heighten the effects of catecholamines on the heart (Ford et al. 1968; Lathers et al. 1988; Spivey et al. 1985), and thereby contribute to the initiation of arrhythmias by cocaine. Carbamazepine is an antiepileptic drug that seems to be particularly effective in treating limbic system seizures. In experimental studies, repeated high doses of cocaine produce a convulsive response classiἀed as pharmacologic kindling. Weiss et al. (1987) reported that chronic carbamazepine administration inhibited the development of lidocaine- or cocaine-kindled seizures and lethality. Chronic, but not acute, pretreatment with carbaÂ� mazepine inhibited the high-dose cocaine seizures. It was suggested that carbamazepine may interact with local anesthetic mechanisms mediating the progressive development of seizures and that the effects of this antiepileptic drug at the level of the sodium channels should be further explored since both carbamazepine and the local anesthetics are believed to interact at this site. Investigation of the mechanism responsible for this effect should be undertaken, as it may prove clinically useful in preventing cocaine toxicity.

180 Sudden Death in Epilepsy: Forensic and Clinical Issues

11.4â•…Treatment of Cocaine-Induced Arrhythmias and Seizures Treatment of cocaine toxicity must ultimately involve deconditioning therapy to reduce drug craving and drug-seeking behavior (Kumor et al. 1988). Tricyclic antidepressants, bromocriptine, amantadine, methylphenidate, and lithium may decrease cocaine selfmedication. The hypertension and tachycardia that follow administration of cocaine are mediated by both alpha-adrenergic and beta-adrenergic receptors to induce vasoconstriction and an increase in heart rate and cardiac output, respectively (Olsen et al. 1983). One clinical management procedure of the adrenergic cocaine crisis involves the judicious use of intravenous propranolol, given in doses of 1 mg at 1-min intervals to a total of up to 6€mg (Gay 1982). Although intervention calms the excitable patient and decreases tachyÂ� arrhythmias, the efficacy of propranolol is limited by its receptor sensitivity. It has been argued that although propranolol effectively blocks beta receptors to decrease heart rate, it leaves the alpha-adrenergic receptors unopposed (Olsen et al. 1983). Thus stimulation of the alpha1-adrenergic receptors in the smooth muscle vasculature results in a worsening of vasoconstriction with resultant dangerous hypertension. Olsen et al. (1983) propose the use of phentolamine or nitroprusside to effect rapid vasodilatation. However, Gay (1983) argues that the nonselective alpha-adrenergic blockade properties of phentolamine may, in fact, further aggravate matters. Indeed, phentolamine will block alpha1 postsynaptic receptors to decrease vasoconstriction, but it will also block alpha2 presynaptic receptors. This blocks the normal regulatory control of the catecholamines, resulting in an increase in synthesis and output of norepinephrine, and may even spur a reflex sympathetic response to the heart. One agent recently used to treat cocaine toxicity is labetalol, which possesses both alpha- and beta-blocking capabilities. Thus the establishment of the alpha blockade counters the cocaine-induced vasoconstriction and hypertension while the beta blockade decreases the tachyarrhythmias (Gay and Loper 1988). The use of chlorpromazine and haloperidol to calm the hyperkinetic state is contraindicated in the cocaine user as they can lower seizure threshold activity and cause cardiac arrhythmias and/or sudden death (Lathers and Lipka 1986, 1987; Lipka and Lathers 1987). Instead, an effective means of quieting the stimulatory phase of cocaine intoxication is the use of diazepam, 15–20 mg orally every 8 h. Antelman et al. (1981) serendipitously found that amytriptiline, a tricyclic antidepressant, protected against sudden cardiac death due to cocaine intoxication in animals. Amytriptiline, 10 mg/kg, administered experimentally in animals 1 h before intraperitoneal injection of cocaine (35 mg/kg) resulted in no protection against sudden death. However, 24-h pretreatment with a single injection of amytriptiline markedly increased survival, while 10-day pretreatment conferred complete protection against sudden death. The mechanism of action, to date, has not been deἀned. However, pretreatment is not a useful tool in the management of clinical toxicity associated with cocaine use. It has recently been suggested that Ca2+ channel blockers may be a useful antidote for cocaine toxicity (Duke 1986; Mittleman and Welti 1987; Trouve and Nahas 1986). Ca2+ channel blockers inhibit the vasoconstrictive effects of norepinephrine by blocking the release of Ca2+ into the smooth muscle of the vasculature. Trouve and Nahas (1986) studied the cocaine antagonistic effects of nitrendepine, a Ca2+ channel blocker, in animals. Nitrendipine was selected for its lack of myocardial depressant activity and its ability to cause coronary vasodilatation. Nitrendipine (1.46 × 10−3 mg/kg/min) was concomitantly administered with 2 mg/ kg/min of cocaine and caused an inhibition of cocaine-induced tachycardia, pressor, and

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

181

vasoconstriction. Nitrendipine also suppressed the cocaine-induced arrhythmias observed in control animals. In a comparative study, nitrendipine and propranolol were able to slow cocaine-induced tachycardia while nitrendepine alone increased coronary flow and pulse pressure. Nitrendipine, alone and in combination with propranolol, decreased coronary flow and performance (Trouve and Nahas 1986). In addition to its cardioprotective properties, nitrendipine appears to possess central activity. Nitrendipine prevented motor tremors, convulsions, and seizures (Trouve and Nahas 1986). It was concluded that the sympatho� mimetic properties of cocaine can be antagonized by Ca2+ channel blockers. Ca2+ channel blockers may become the drugs of choice for the treatment of cocaine intoxication.

11.5â•…Use of Cocaine in Persons with Epilepsy Since cocaine use has been reported to produce hypertension, ventricular arrhythmias, tachycardia, myocardial infarction, seizures, and sudden death (Nahas et al. 1985; Tazelaar et al. 1987), and since persons with epilepsy have been shown to manifest autonomic dysfunction, including changes in blood pressure, heart rate, and rhythm (Leestma et al. 1984; Penἀeld and Erickson 1941; Phizackerly et al. 1954; Walsh et al. 1968), one must raise the question of whether the use of cocaine in individuals with epilepsy places them at risk of dying in a sudden, unexplained manner. Furthermore, dysfunction in the activity of peripheral cardiac autonomic neural discharge contributes to the production of cardiac arrhythmias (Gillis 1969; Gillis et al. 1972; Lathers et al. 1974, 1977, 1978; Verrier and Lown 1978; Weaver et al. 1976) and to sudden death (Lown and Verrier 1978). Lathers et al. (1977, 1978) and Lathers (1980) reported that nonuniform discharge in the cardiac postganglionic nerves, i.e., simultaneous increases and decreases in the various sympathetic branches innervating the myocardium, contributes to the production of arrhythmias by altering ventricular automaticity and excitability in the manner reported by Han and Moe (1964). Similar autonomic cardiac neural dysfunction was reported in association with arrhythmias and interictal and ictal discharges (Carnel et al. 1985; Lathers and Schraeder 1982, 1987; Lathers et al. 1984, 1987; Schraeder and Lathers 1983, 1988). Cocaine also modiἀes postganglionic cardiac sympathetic neural function, increasing the inotropic and chronotropic responses to sympathetic neural stimulation (Matsuda et al. 1980). Thus it is possible that the actions of cocaine and the autonomic cardiac neural dysfunction associated with epileptogenic activity may combine to produce cardiac arrhythmias and, at worst, sudden unexpected death. The question of whether cocaine use in the individual with epilepsy places the individual at risk for sudden death should be examined.

11.6â•…Summary This chapter has reviewed the incidence, characteristics, risk factors, and clinical management of cocaine-induced toxicity. Cocaine causes death by actions on the cardiovascular system, including cardiomyopathy, arrhythmia production, accelerated heart rate, and increased blood pressure. Seizures often accompany cocaine toxicity, leading to death. Cocaine is known to activate the EEG in humans, cause seizures in animals, and lower the seizure threshold. Patients with preexisting risk factors for cardiovascular pathology (high cholesterol, high blood pressure, cardiac arrhythmia, etc.) and those with epilepsy may

182 Sudden Death in Epilepsy: Forensic and Clinical Issues

be especially sensitive to cocaine-induced toxicity. Most research in animals suggests that cocaine-induced cardiovascular responses are due to enhanced noradrenergic response on the heart and arteriolar smooth muscles. While there is controversy surrounding the management of cocaine-induced toxicity, a symptomatic approach involves controlling the seizures with diazepam, the cardiovascular response with beta-adrenergic blockers or labetolol, a combined alpha- and beta-blocking agent, while correcting the systemic acidosis and hyperthermia. Use of the Ca2+ channel blockers may represent a new, more effective treatment. Finally, since both cocaine and epilepsy alone are associated with sudden unexpected death and since both are capable of modifying cardiac sympathetic neural discharge to produce changes in heart rate and rhythm, the question of whether the use of cocaine in the epileptic person places this individual at risk for sudden death must be raised.

References Abramowicz, M. 1986. Crack. Med Lett Drug Ther 28: 69–70. Amon, C. A., L. G. Tate, R. K. Wright, and W. Matusiak. 1986. Sudden death due to ingestion of cocaine. J Ann Toxicol 10: 217–218. Antelman, M., D. Kocan, N. Rowland, L. De Giovanni, and L. A. Chiodo. 1981. Amitriptyline provides long-lasting immunization against sudden cardiac death from cocaine. Eur J Pharmacol 69: 119–120. Bayorh, M. A., Z. Zukowska-Grojec, and I. J. Kopin. 1983. Effect of desipramine and cocaine on plasma norepinephrine and pressor responses to adrenergic stimulation in pithed rats. J Clin Pharmacol 23: 24–31. Benchimol, A., H. Bartall, and K. Desser. 1978. Accelerated ventricular rhythm and cocaine abuse. Ann Intern Med 88: 519–520. Berger, H. 1937. Hans Berger on the electroencephalogram of man. Arch Psychiatr Nervenk 106: 577–584. Translated in P. Gloor (Ed.). EEG Clin Neurophysiol Suppl 28: 291–297. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30: 225–240. Castellani, S., E. H. Ellinwood Jr., M. M. Kilbey et al. 1983. Cholinergic effects on arousal and cocaineinduced olfactory-amygdala spindling and seizures in cats. Sociol Behav 31: 461–466. Catravas, J. D., and I. W. Waters. 1981. Acute cocaine intoxication in the conscious dog: Studies on the mechanism of lethality. J Pharmacol Exp Ther 217: 350–356. Catravas, J. D., I. W. Waters, M. A. Walz et al. 1978. Acute cocaine intoxication in the conscious dog: Pathophysiologic proἀle of acute lethality. Arch Int Pharmacodyn 235: 328–340. Chiu, Y. C., K. Brecht, D. S. DasGupta, and E. Mhoon. 1986. Myocardial infarction with topical cocaine anesthesia for nasal surgery. Arch Otolaryngol Head Neck Surg 112: 988–990. Christ, D., J. Curry, and T. Zitaglio. 1982. Potentiation of the ganglionic blocking action of norepinephrine by cocaine. J Pharmacol Exp Ther 220: 97–101. Coleman, D. L., T. F. Ross, and J. L. Naughton. 1982. Myocardial ischemia and infarction related to recreational cocaine use. West J Med 136: 444–446. Cregler, L. L., and H. Mark. 1986a. Medical complications of cocaine abuse. New Engl J Med 315: 1495–1500. Cregler, L. L., and H. Mark. 1986b. Cardiovascular dangers of cocaine abuse. Am J Cardiol 57: 1185–1186. Cregler, L. L., and H. Mark. 1987. Cocaine-induced myocardial ischemia. Am J Med 82: 388. Dart, A. M., R. Dietz, W. Kubler, A. Schömig, and R. Strasser. 1983. Effects of cocaine and desipramine on the neurally evoked overflow of endogenous noradrenaline from the rat heart. Br J Pharmacol 79: 71–74.

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

183

Duke, M. 1986. Cocaine, myocardial infarction and arrhythmias—A review. Conn Med 50: 440–442. Estroff, T. W., and M. S. Gold. 1986. Medical and psychiatric complications of cocaine abuse with possible points of pharmacological treatment. In Advances in Alcohol Substance Abuse, Vol. 5, 61–68. New York, NY: Haworth Press. Fabiato, A., and F. Fabiato. 1978. Effects of pH on the myoἀlaments and the sarcoplasmic reticulum of skinned cells from cardiac and skeletal muscle. J Physiol 276: 233–255. Fekete, M., and J. Borsy. 1971. Chlorpromazine–cocaine antagonism: Its relation to changes of dopaÂ� mine metabolism in the brain. Eur J Pharmacol 16: 171–175. Fischman, M. W., C. R. Schuster, L. Resnekov et al. 1976. Cardiovascular and subjective effects of intravenous cocaine administration in humans. Arch Gen Psych 33: 983–989. Freud, S. 1884. Oncoca. Centralblatt für die Gesammte Theupie (Wien). Vol. 2, 289–314. Reprinted in the Cocaine Papers (S. Edminster, trans.). Vienna: Dunquin Press, 1963. Ford, G. D., W. H. Cline Jr., and W. W. Flemin. 1968. Influence of lactic acidosis on cardiovascular response to sympathomimetic amines. Am J Physiol 215: 1123–1129. Gay, G. R. 1982. Clinical management of acute and chronic cocaine poisoning. Ann Emerg Med 11: 562–572. Gay, G. R. 1983. Letter to the editor: In response to management of cocaine poisoning. Ann Emerg Med 10: 656–657. Gay, G. R., and K. A. Loper. 1988. Control of cocaine-induced hypertension with labetalol. Anesth Analg 67: 92. Gillis, R. A. 1969. Cardiac sympathetic nerve activity: Changes induced by ouabain and propranolol. Science 166: 508–510. Gillis, R. A., A. Raines, Y. L. Sohn, B. Levitt, and F. G. Standaert. 1972. Neuroexcitatory effects of digitalis and their role in the development of cardiac arrhythmias. J Pharmacol Exp Ther 183: 154–168. Gould, L., C. Gopalaswamy, C. Patel, and R. Betzu. 1985. Cocaine-induced myocardial infarction. N€Y State J Med, November, 660–661. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Herning, R. I., R. T. Jones, Hooker et al. 1985. Cocaine increases EEG beta: A replication and extension of Hans Berger’s historic experiments. EEG Clin Neurophysiol 60: 470–477. Howard, R. E., D. C. Hueter, and G. J. Davis. 1985. Acute myocardial infarction following cocaine abuse in a young woman with normal coronary arteries. J Am Med Assoc 254: 95–96. Isner, J. M., N. A. M. Estes, P. D. Thompson, Mark et al. 1985. Cardiac consequences of cocaine: Premature myocardial infarction, ventricular tachyarrhythmia, myocarditis, and sudden death. Circulation 72 (Suppl 3): 414. Isner, J. M., N. A. M. Estes, P. D. Thompson et al. 1986. Acute cardial events temporally related to cocaine abuse. New Engl J Med 315: 1438–1443. Jain, R., P. J. Gatti, M. Visner, K. G. Albrecht, M. G. Moront, C. E. Rackley, and R. A. Gillis. 1987. Effects of cocaine on cardiorespiratory function and on cardiovascular responses produced by bilateral carotid occlusion and IV norepinephrine. Fed Proc 46: 402. Javiad, J. I., M. W. Fishman, C. R. Schuster et al. 1978. Cocaine plasma concentration: Relation to physiological and subjective effects in humans. Science 202: 227–228. Jonsson, S., M. O’Meara, and J. B. Young. 1983. Acute cocaine poisoning. Am J Med 75: 1061–1064. Kassowsky, W. A., and A. F. Lyon. 1984. Cocaine and acute myocardial infarction. A probable connection. Chest 5: 729–731. Lathers, C. M. 1980. Effect of timolol on postganglionic cardiac and preganglionic splanchnic sympathetic neural discharge associated with ouabain-induced arrhythmia. Eur J Pharmacol 64: 95–106. Lathers, C. M., and L. J. Lipka. 1986. Chlorpromazine: Cardiac arrhythmogenicity in the cat. Life Sci 38: 521–538.

184 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and L. J. Lipka. 1987. Cardiac arrhythmia, sudden death and psychoactive agents. J Clin Pharmacol 27: 1–14. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy. Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–648. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., N. Turner, and J. M. Schoffstall. 1989. The effect of different routes of sodium bicarbonate administration on plasma catecholamines, pH, and blood pressure during cardiac arrest in pigs. Resuscitation 18: 59–74. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1974. Relationship between the effect of ouabain on arrhythmia and interspike intervals (ISI) of cardiac accelerator nerves. Pharmacologist 16: 201. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57: 1058–1065. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Turner, and R. M. Levin. 1986. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmias, and beta receptor density associated with coronary occlusion in the cat. Life Sci 39: 2121–2141. Lathers, C. M., P. L. Schraeder, and S. B. Cornel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effects of phenobarbital. Life Sci 34: 1919–1936. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lathers, C. M., W. H. Spivey, and N. Turner. 1988. The effect of timolol given ἀve minutes post coronary occlusion on plasma catecholamines. J Clin Pharmacol 28: 289–299. Leestma, J. E., M. G. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Lesse, H., and J. P. Collins. 1979. Effects of cocaine on propagation of limbic seizure activity. Pharmacol Biochem Behav 11: 689–694. Lichtenfeld, P. J., D. B. Rubin, and R. S. Feldman. 1984. Subarachnoid hemorrhage precipitated by cocaine snorting. Arch Neurol 41: 223–224. Lipka, L. J., and C. M. Lathers. 1987. Psychoactive agents, seizure production, and sudden death in epilepsy. J Clin Pharmacol 27: 169–183. Loveys, B. J. 1987. Physiologic effects of cocaine with particular reference to the cardiovascular system. Heart Lung 16: 175–181. Lown, B., and R. L. Verrier. 1978. Neural factors and sudden death. In Perspectives in Cardiovascular Research. Vol. 2. Neural Mechanisms in Cardiac Arrhythmias, ed. P. J. Schwartz, A. M. Brown, A. Malliani, and A. Zanchetti, 87–98. New York, NY: Raven Press. Mathias, D. W. 1986. Cocaine associated myocardial ischemia. Am J Med 81: 675–678. Matsuda, Y., Y. Masuda, B. Blattberg, and M. N. Levy. 1980. The effects of cocaine, chlorpheniramine and tripelennamine on the cardiac responses to sympathetic nerve stimulation. Eur J Pharmacol 63: 25–33. Matsuzaki, M. 1978. Alteration in pattern of EEG activities and convulsant effect of cocaine following chronic administration in the rhesus monkey. EEG Clin Neurophysiol 45: 1–15. Mittleman, R. E., and C. V. Welti. 1987. Cocaine and sudden “natural” death. J Forensic Sci 32: 11–19. Nahas, G., R. Trouve, J. F. Demus, and M. von Sitbon. 1985. A calcium-channel blocker as antidote to the cardiac effects of cocaine intoxication. New Engl J Med 313: 519–520.

Cocaine-Induced Seizures, Arrhythmias, and Sudden Death

185

Nanji, A. A., and J. D. Filipenko. 1984. Asystole and ventricular ἀbrillation associated with cocaine intoxication. Chest 85: 132–133. Olsen, K., N. Benowitz, and P. Pentel. 1983. Management of cocaine poisoning. Ann Emerg Med 10: 655–656. Pasternack, P. F., S. B. Colvin, and F. G. Baumann. 1985. Cocaine-induced angina pectoris and acute myocardial infarction in patients younger than 40 years. Am J Cardiol 55: 847. Penἀeld, W., and T. C. Erickson. 1941. Epilepsy and Cerebral Localizations, 320–362. Springἀeld, IL: Charles C. Thomas. Phizackerly, P. J. R., E. W. Poole, and C. W. M. Whitty. 1954. Sinoauricular heart block as an epileptic manifestation: A case report. Epilepsia 3: 89–91. Pickworth, W. B., R. I. Herning, K. Kumor, and M. Sherer. 1986. Spontaneous EEG during chronic cocaine infusion. Pharmacologist 28: 236. Pitts, D. K., and J. Marwah. 1986. Electro physiological effects of cocaine on central monoaminergic neurons. Eur J Pharmacol 131: 95–98. Pitts, D. K., and J. Marwah. 1987. Cocaine inhibits central monoaminergic neurons and activates cerebellar Purkinje neurons. Fed Proc 46: 400. Pitts, D. K., C. E. Udom, and J. Marwah. 1987. Cardiovascular effects of cocaine in anesthetized and conscious rats. Life Sci 40: 1099–1111. Ritchie, J. M., and N. M. Greene. 1980. Local anesthetics. In The Pharmacological Basis of Therapeutics, ed. L. S. Goodman and A. Gilman, 302–321. New York, NY: Macmillan. Roberts, D. C. S., M. E. Corcoran, and H. C. Fibiger. 1977. On the role of ascending catecholaminergic systems in intravenous self-administration of cocaine. Pharmacol Biochem Behav 6: 615–620. Rockhold, R. W., B. Hoskins, and I. K. Ho. 1987. Spontaneously hypertensive rats are resistant to convulsive and lethal actions of cocaine. Fed Proc 46: 402. Rod, J. L., and R. P. Zucker. 1987. Acute myocardial infarction shortly after cocaine inhalation. Am J Cardiol 59: 161. Rollingher, I. M., A. S. Belzberg, and I. L. MacDonald. 1986. Cocaine-induced myocardial infarction. Can Med Assoc 135: 45–46. Rosendorff, C., J. T. E. Hoffman, E. D. Verrier, J. Rouleau, and L. E. Boerboom. 1981. Cholesterol potentiates the coronary artery response to norepinephrine in anesthetized and conscious dogs. Circ Res 45: 320–329. Schachne, J. S., B. H. Roberts, and P. D. Thompson. 1984. Coronary-artery spasm and myocardial infarction associated with cocaine use. New Engl J Med 310: 1665–1666. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained sudden death. Life Set 32: 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal cardiovascular dysfunction and epileptogenic activity. Epilepsy Res 3: 55–62. Simpson, R. W., and W. D. Edwards. 1986. Pathogenesis of cocaine-induced ischemic heart disease. Arch Pathol Lab Med 110: 479–484. Spivey, W. H., C. M. Lathers, D. R. Malone, H. D. Unger, S. Blat, R. M. McNamara, J. Schroffstall, and N. Turner. 1985. A comparison of intraosseous, central and peripheral routes of sodium bicarbonate administration during CPR in pigs. Arch Pathol Lab Med 14: 1135–1139. Tackett, R. L., and L. F. Jones. 1987. Central catecholaminergic changes and cardiovascular responses following acute administration of cocaine. Pharmacologist 29: 159. Tazelaar, H. D., S. B. Karch, B. G. Stephens, and M. E. Billingham. 1987. Cocaine and the heart. Hum Pathol 18: 195–199. Trendelenburg, U. 1968. The effect of cocaine on the pacemaker of isolated guinea-pig atria. J Pharmacol Exp Ther 161: 222–231. Trouve, R., and G. Nahas. 1986. Nitrendipine: An antidote to cardiac and lethal toxicity of cocaine. Proc Soc Exp Biol Med 183: 392–397. Trulson, T. J., and M. E. Trulson. 1987. Cocaine suppresses the activity of noradrenergic locus coeruÂ� leus neurons in freely moving cats. Pharmacologist 29: 159.

186 Sudden Death in Epilepsy: Forensic and Clinical Issues Tyau, L. S. Y., and C. M. Lathers. 1988. Cocaine-induced myocardial ischemia, infarction, and arrhythmias: A review of possible mechanisms causing sudden death. FASEB J 2: A1518. Van Dyke, C., and R. Byck. 1983. Cocaine use in man. Adv Subst Abuse 3: 1–24. Verrier, R. L., and B. Lown. 1978. Sympathetic–parasympathetic interactions and ventricular electrical stability. In Perspectives in Cardiovascular Research. Vol. 2. Neural Mechanisms in Cardiac Arrhythmias, ed. P. J. Schwartz, A. M. Brown, A. Malliani, and A. Zanchetti, 75–85. New York, NY: Raven Press. Vitullo, J. C., C. Lakios-Cherpas, and P. A. Khairallah. 1987. Effects of intraaortic injection of cocaine hydrochloride on rat electrocardiograms. Fed Proc 46: 1143. Walsh, G., W. Masland, and E. Goldenshon. 1968. Paroxysmal cerebral discharge associated with paroxysmal atrial tachycardia. Electroencephalogr Clin Neurophysiol 24: 187. Weaver, L. C., T. Akera, and T. M. Brody. 1976. Digoxin toxicity: Primary sites of drug action on the sympathetic nervous system. J Pharmacol Exp Ther 197: 1–9. Weiss, R. J. 1986. Recurrent myocardial infarction caused by cocaine abuse. Am Heart J 111: 793. Weiss, S. R. B., M. Costello, R. Woodward et al. 1987. Chronic carbamazepine inhibits the development of cocaine-kindled seizures. Abstr Soc Neurosci 13: 950. Welti, C. V., and D. A. Fishbain. 1985. Cocaine-induced psychosis and sudden death in recreational cocaine users. J Forensic Sci 30: 873–880. Wiener, M. D., and C. Putnam. 1987. Pain in the chest in a user of cocaine. J Am Med Assoc 258: 2087–2088. Wiener, R. S., J. T. Lockhart, and R. G. Schwartz. 1986. Dilated cardiomyopathy and cocaine abuse: Report of two cases. Am J Med 81: 699–701. Wilkerson, R. D. 1987. Yohimbine pretreatment enhances the cardiovascular actions of cocaine in anesthetized dogs. Fed Proc 46: 1143. Yasuda, R. P., N. R. Zahniser, and T. V. Dunwiddie. 1984. Electrophysiological effects of cocaine in the rat hippocampus in vitro. Neurosci Lett 45: 199–204. Young, D., and J. J. Glauber. 1947. Electrocardiographs changes resulting from acute cocaine intoxication. Am Heart J 34: 272–279.

Risk Factors for Sudden Death in Epilepsy Thaddeus s. Walczak

12

Contents 12.1 12.2 12.3 12.4 12.5

Introduction Epidemiologic Sources and Biases Traditional Risk Factors for SUDEP: Epilepsy Severity and Seizure Type SUDEP Is Not Conἀned to Severe Epilepsy Other Traditional Risk Factors 12.5.1 Age and Gender 12.5.2 Symptomatic Causes of Epilepsy 12.5.3 Sleep, Sleep Position, and Supervision during Sleep 12.5.4 Psychotropic Drugs 12.6 AED Use and SUDEP 12.6.1 AED Compliance and SUDEP 12.6.2 AED Polytherapy and SUDEP 12.6.3 Individual AEDs and SUDEP 12.7 Novel SUDEP Risk Factors: Recent Work 12.7.1 SCN1A Mutations 12.7.2 Heart Rate Variability and Cardiac Autonomic Instability 12.7.3 Anatomic and Electrophysiologic Substrates of SUDEP 12.7.3.1 Anatomic Substrates of Postictal Apnea 12.7.3.2 Anatomic Substrates of Cardiac Arrhythmia in SUDEP 12.7.3.3 Electrophysiologic Substrates of Cardiac Arrhythmia in€SUDEP 12.8 The Next Steps in the Study of SUDEP Risk Factors References

187 188 189 192 192 192 193 193 193 194 194 194 194 195 196 196 196 196 196 197 197 198

12.1â•…Introduction Understanding risk factors for sudden unexpected death in epilepsy (SUDEP) is important for both research-oriented and practical reasons. Current thinking regarding the pathophysiology and prevention of SUDEP remains largely speculative. Deἀning circumstances surrounding SUDEP and the patient and epilepsy characteristics associated with SUDEP could help direct investigations of pathophysiology. Risk factors can sometimes be modiἀed and could offer an approach to prevention, though we must remember that additional prospective studies with active intervention are necessary to demonstrate that modifying risk factors is effective. More practically, a consistent set of risk factors could deἀne a population that may be especially prone to SUDEP. This group can then be targeted for more detailed discussion of SUDEP and for any potential interventions. 187

188 Sudden Death in Epilepsy: Forensic and Clinical Issues

Growing interest in SUDEP over the past 25 years has prompted a series of investigations into the risk factors associated with this condition. This chapter attempts to review and synthesize this information. Such a synthesis is challenging because of differences in deἀnitions, study populations, and study design. However, the picture that emerges is remarkably consistent and corresponds well to what we are beginning to learn about pathophysiology.

12.2â•… Epidemiologic Sources and Biases Deἀning risk factors is initially a task for epidemiology. Selection and other biases associated with various epidemiologic approaches can signiἀcantly affect results. This is especially true when investigating an uncommon condition such as SUDEP. Different results in various studies may be due to the differing methodologies used. Before considering results, it is important to consider the biases associated with the various epidemiologic approaches used to determine SUDEP risk factors. Initial reports of SUDEP were uncontrolled case series collected at medical examiners’ offices (Leestma et al. 1989; Terrence et al. 1975; Freytag and Lindenberg 1964). These solidiἀed belief in SUDEP as an entity and provided an initial guess at risk factors. However, these populations are not representative of the great majority of people with epilepsy. Risk factor assessments based on this material are likely to be affected by selection bias, especially given the absence of control populations. Subsequent studies have evaluated SUDEP risk factors in more representative populations, including large prescription databases, cohorts of developmentally delayed persons with epilepsy, larger cohorts of persons with epilepsy at epilepsy centers, and drug and device development programs. These also do not represent the general population with epilepsy, though risk factors in SUDEP subjects are usually compared to nested controls in these studies, which reduces bias due to selection. A SUDEP population derived from a cohort of incident epilepsy would provide the best information on risk factors. However, such information is very difficult to obtain because SUDEP is rare in a population-based cohort with prevalent epilepsy (Ficker et al. 1998) and even rarer in a group with incident epilepsy (Lhatoo et al. 2001). Even if a reasonable number of cases were to be found in a large community-based cohort, it would be extremely difficult to assess circumstances of death and risk factors in retrospect. Less information is typically available about people with epilepsy drawn from a community; often the diagnosis of epilepsy is not secure. More information is usually available in groups followed in epilepsy clinics, where the diagnosis of epilepsy is probably more reliable. Most information regarding SUDEP risk factors is derived from retrospective SUDEP ascertainment. These studies often do not fully deἀne the cohort from which the deaths are drawn, so the population in which SUDEP is being described is often unclear. Further potential sources of bias include: 1. The possibility that not all deaths are ascertained 2. Difficulty in elucidating the circumstances of death several years later 3. Uncertainty about when risk factors for SUDEP and controls should be ascertained in the course of clinical follow-up 4. Difficulty in ascertaining risk factors in retrospect

Risk Factors for Sudden Death in Epilepsy

189

The general availability of mortality indices in developed countries allows reliable determination of whether a given individual is alive or not. However, determination of whether SUDEP occurred or not is very much dependent on a thorough understanding of the circumstances of death and this can be very difficult to reconstruct several years later. Deἀning a cohort and potential risk factors prospectively, assessing for death at regular intervals, and assessing for SUDEP promptly after death reduces these biases; however, such studies are more costly and challenging to perform as they need to utilize case investigators using a standardized set of questions to be asked of family and healthcare providers, as well as performance of autopsies for most, if not all, persons with a history of epilepsy who died. Choice of controls can also influence results. Some studies use people with epilepsy dying from causes other than SUDEP as controls. This is thought to provide information regarding circumstances of death (Tellez-Zenteno et al. 2005). However, mortality in the epilepsy population is mostly related to underlying causes of epilepsy. It is not clear that comparison of SUDEP deaths to epilepsy deaths (which are mostly related to the causes of epilepsy) adds much to understanding why SUDEP occurs. If the goal is determining which people with epilepsy are at risk for SUDEP, comparing risk factors in live people with epilepsy and SUDEP victims is more appropriate (Tomson et al. 2008). Finally, individual risk factors potentially associated with SUDEP may be related to other potential risk factors. An important example is seizure frequency, seizure severity, and anticonvulsant drug polytherapy, all of which may influence one another (Nilsson et al. 1999; Walczak et al. 2001; Langan et al. 2005). Establishing that a risk factor is independently responsible requires multivariate analysis, which, in turn, requires a reasonable number of cases and controls.

12.3â•…Traditional Risk Factors for SUDEP: Epilepsy Severity and Seizure Type Many potential risk factors have been examined (Table 12.1). Epilepsy severity and correlates such as epilepsy duration, seizure type, and seizure frequency have perhaps been most intensely studied. SUDEP risk has been consistently associated with more severe epilepsy, longer duration of epilepsy, and more frequent seizures. Table 12.2 summarizes SUDEP incidence in several populations with differing epilepsy severity. SUDEP incidence is very low in new onset (incident) epilepsy cohorts and somewhat increased in community-based prevalence cohorts. SUDEP incidence is higher in cohorts of persons with epilepsy at referral centers where more severe epilepsy cases may be expected to congregate. Incidence rates are still higher in drug and device development programs, which are usually limited to patients who have failed treatments with several antiepileptic drugs (AEDs). SUDEP incidence is highest in persons with epilepsy undergoing epilepsy surgery, where epilepsy is especially refractory. Furthermore, the percentage of all reported deaths due to SUDEP also increases with increasing epilepsy severity (Table 12.2). These two ἀndings establish a clear gradient of risk that strongly supports the idea that SUDEP risk increases with epilepsy severity. Most studies with live patient controls report that SUDEP is associated with higher seizure frequency (Table 12.1). The larger controlled studies (Nilsson et al. 1999; Langan et al. 2005; Walczak et al. 2001) all demonstrate progressively increased relative risk with

0

ns

20/80

154/616

62/124

+

0

18

0

0

0

ns

+

0

Male Sex

11/?

ns

0

14/1806

57/171

ns

11/20

Young Age

0

+

ns

+

0

ns

Epilepsy Duration

+

+

+

ns

+

+

ns

0

Frequent Seizures

ns

+

+

ns

+

+

0

ns

0

+

+

nsa

ns

ns

+

Mental Retardation

0

0

+

+

+

+

0

ns

AED Polytherapy at Time of Death

0

0

ns

ns

ns

ns

0

Lack of€super� vision at night, treatment with CBZ

Rx with€anti� psychotic drugs

Treatment with CBZ, lack of supervision at night Rx with€anti� psychotic drugs Nonambulatory status

Noncompliance with AED Treatment Other Risk Factors

Note: SUDEP, sudden unexpected death in epilepsy; AED, antiepileptic drug; CBZ, carbamazepine. 0, item was not a risk factor for SUDEP; +, item was a risk factor for SUDEP; ns, item was not studied. a All SUDEP cases had mental retardation in this study.

Hiltris et al. (2007)

Nilsson et al. (1999) McKee and Bodἀsh (2000) Tennis et al. (1995) Walczak et al. (2001) Langan et al. (2005)

Jick et al. (1992) Timmings (1993)

Cases/ Controls

Frequent Tonic– Clonic Seizures

Table 12.1â•…Risk Factors for SUDEP in Some Studies with Living Persons with Epilepsy as Controls

190 Sudden Death in Epilepsy: Forensic and Clinical Issues

Risk Factors for Sudden Death in Epilepsy

191

Table 12.2â•… SUDEP Incidence with Increasing Epilepsy Severity

Study Lhatoo et al. (2001) Ficker et al. (1998) Jick et al. (1992) Tennis et al. (1995) Timmings (1993) Lip and Brodie (1992) Leppik (1995) Leestma et al. (1997) Annegers et al. (2000) Sperling et al. (1999) Dashieff (1991)

Population

Incidence (per 1000 patient years)

Percentage of Deaths That Are SUDEP (%)a

Population-based cohort of incident epilepsy Population-based cohort of prevalent epilepsy Large prescription database Large prescription database Epilepsy clinic Epilepsy clinic

0.09

0.5

0.35

8.6

1.3 1.35 2.0 4.9

26 11 nr

Tiagabine clinical trial Lamotrigine clinical trial Vagus nerve stimulator clinical trial Patients who underwent epilepsy surgery Candidates for epilepsy surgery

3.9 3.5 4.1 4.0

29 40 52 54

10.0

Source: Leestma et al., Epilepsia, 38, 47–55, 1997. Note: nr, not reported. a SUDEP included deἀnite and probable SUDEP cases.

higher seizure frequencies. When multivariate models include generalized tonic–clonic seizures, other seizure types, and AEDs, SUDEP risk appears associated with generalized tonic–clonic seizures rather than partial seizures (Walczak et al. 2001; Langan et al. 2005; Tomson et al. 2005). As few as three tonic–clonic seizures per year signiἀcantly increase SUDEP risk, compared to no tonic–clonic seizures; risk increases further with more frequent tonic–clonic seizures. Other evidence strongly supports the position that tonic–clonic seizures are an important risk factor for epilepsy. All studies assessing seizure type report a history of tonic–clonic seizures in at least 90% of SUDEP cases. These ἀndings are consistent no matter what the study design or source population (Hirsch and Martin 1971; Terrence et al. 1975; Earnest et al. 1992; Leestma et al. 1989; Kloster and Torstein 1999; Walczak et al. 2001; Timmings 1993; Langan et al. 2005). Furthermore, studies addressing circumstances of death report tonic–clonic seizure prior to death in most cases (Terrence et al. 1975; Leestma et al. 1989; Earnest et al. 1992; Nillson 1999; Langan et al. 2000; Opeskin and Berkovic 2003). The consistency of these ἀndings indicates that tonic–clonic seizures are an important proximate cause of SUDEP. The relationship between epilepsy duration and SUDEP risk is less clear. SUDEP did not occur in studies assessing outcome in the several years after an initial seizure (Beghi et al. 2005). When risk is stratiἀed by duration of epilepsy, a clinically and statistically signiἀcant risk is noted after 10 to 30 years of epilepsy (Walczak et al. 2001; Leestma et al. 1989). One case control study found increased SUDEP risk with longer epilepsy duration (Walczak et al. 2001) while another did not (Langan et al. 2005). Age of seizure onset is generally lower in SUDEP cases than in controls (Nillson 1999; Jick et al. 1992; Kloster and Torstein 1999). This also supports the idea that duration of epilepsy is an important risk factor.

192 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 12.3â•… Expected Yearly Number of SUDEP Cases in a Hypothetical City of 6,000,000 Inhabitantsa People with Epilepsy

SUDEP Rates (patient years)

Number of SUDEP Cases

Percentage of All SUDEP Cases (%)

24,000

0.5/1000 patient years

12

50

6,000 30,000

2.0/1000 patient years

12 24

50

Well-controlled epilepsy Intractable epilepsy Total a

See Section 12.4 for details.

12.4╇SUDEP Is Not Confined to Severe Epilepsy This information relating higher SUDEP risk to increased seizure severity is leading to the impression that SUDEP is conἀned to people with severe epilepsy. Population-based medical examiner series and clinical experience clearly indicate that this is not the case. SUDEP also occurs in people with what are generally considered to be well-controlled seizures (Opeskin and Berkovic 2003). In fact, a SUDEP presenting to a random physician is probably equally likely to occur in a person with well-controlled seizures as in a person with poorly controlled seizures. This is because, on a population basis, a large majority of people with epilepsy have reasonable seizure control. A thought experiment illustrates why this is the case (Table 12.3). Consider a large city with 6 million inhabitants. Assuming an epilepsy prevalence of 0.5%, we would expect the city to contain 30,000 individuals with epilepsy. Let us further assume that 80% of the individuals with epilepsy have well-controlled seizures and a SUDEP incidence of 0.5/1000 patient years. Of the individuals with epilepsy, 20% have poorly controlled seizures and a SUDEP risk of 2.0/1000 patient years. With these assumptions (Hauser and Hesdorffer 1990, Table 2), we would expect 12 SUDEP cases per year in the 24,000 individuals with well-controlled seizures and 12 SUDEP cases in the 6000 individuals with poorly controlled seizures. Thus, a SUDEP case presenting to a random health care worker in this city is as likely to have had well-controlled seizures as poorly controlled seizures. General practitioners would be more likely to encounter the SUDEP cases whose seizures had been well-controlled because most people with well-controlled epilepsy are followed by community physicians.

12.5â•…Other Traditional Risk Factors 12.5.1╇Age and Gender Initial uncontrolled case series described SUDEP as a phenomenon found in young men with excessive alcohol use (Leestma et al. 1989; Terrence et al. 1975). This constellation of risk factors appears to have reflected the cases typically referred to medical examiners’ offices and has not been found in controlled studies. Mean age at death in most studies is between 25 and 39 years (see Tomson et al. 2005 for review). Somewhat higher ages were found in the large cohort-based studies (Walczak et al. 2001; Langan et al. 2005; Nillson 1999). Neither population-based studies (Ficker et al. 1998) nor the large case control studies found a male predominance. In fact, two controlled studies (Walczak et al. 2001; Opeskin

Risk Factors for Sudden Death in Epilepsy

193

and Berkovic 2003) found increased incidence of SUDEP among women. Controlled series have not found evidence for increased alcohol abuse among victims of SUDEP (Nilsson et al. 1999; Langan et al. 2005; Opeskin and Berkovic 2003; Kloster and Torstein 1999). However, SUDEP risk appears lower among children than adults. SUDEP incidence in four pediatric studies ranged between 0.11 and 0.43/1000 patient years (Harvey et al. 1993; Donner et al. 2001; Weber et al. 2005; Camἀeld et al. 2002). A prospective cohort study supports this: SUDEP incidence was 0/3625 patient years (0/1000) in those aged 0–19 years, 13/7792 (1.67/1000) in those aged 20–39 years, and 7/3958 patient years (1.77/1000) in those aged 40–59 years (Walczak et al. 2001). Children with the rare and severe myoclonic epilepsy of infancy are an exception; unusually high rates of SUDEP have been reported in this condition (Dravet et al. 2005). 12.5.2â•… Symptomatic Causes of Epilepsy Early series also found an apparent excess of epileptogenic lesions among SUDEP victims (Leestma et al. 1989; Terrence et al. 1975). This ἀnding also appears to be related to selection bias. Controlled clinical (Walczak et al. 2001) and medical examiner’s series (Kloster and Torstein 1999; Opeskin and Berkovic 2003) have not found this association. Controlled studies have not found either symptomatic epilepsy (Nillson 1999; Kloster and Torstein 1999) or any particular epilepsy syndrome (Walczak et al. 2001; Opeskin and Berkovic 2003) to be more common among SUDEP victims either. However, there is some evidence that mental retardation is more common among SUDEP victims (Jick et al. 1992; Walczak et al. 2001) and the inability to ambulate may confer further risk in mentally retarded people with epilepsy (McKee and Bodἀsh 2000). Nonetheless, most SUDEP cases are not afflicted with mental retardation or cerebral palsy. 12.5.3â•… Sleep, Sleep Position, and Supervision during Sleep SUDEP victims are found dead in bed in the majority of cases (Leestma et al. 1989; Terrence et al. 1975; Opeskin and Berkovic 2003; Kloster and Torstein 1999; Nashef et al. 1995), suggesting that sleep increases SUDEP risk. Most patients found dead in bed are found prone in both controlled and uncontrolled studies (Leestma et al. 1989; Kloster and Torstein 1999). Supervision at night (deἀned here as the presence in the bedroom of a responsible individual of normal intelligence), the use of a listening device, and regular checks throughout the night were associated with lower SUDEP rates in a large, retrospective, controlled study (Langan et al. 2005). These ἀndings are consistent with the idea that SUDEP related to postictal apnea is more likely to be lethal when appropriate positioning and ἀrst aid does not occur following a nocturnal seizure. Please see the chapter by Drs. Sato and Hughes discussing SUDEP and sleep in this book. 12.5.4â•… Psychotropic Drugs Antipsychotic drugs are associated with increased risk of sudden death (Ray et al. 2009) and treatment with selective serotonin reuptake inhibitors decreases SUDEP risk in an animal model of SUDEP (Tupal and Faingold 2006). Use of psychotropic medications is probably more common in persons with epilepsy than in the general population. This raises the question of whether use of psychotropic drugs affects SUDEP risk. Several controlled

194 Sudden Death in Epilepsy: Forensic and Clinical Issues

studies have found that the use of psychotropic drugs in general is not more common in SUDEP victims than in control persons with epilepsy (Tennis et al. 1995; Nilsson et al. 1999; Walczak et al. 2001; Langan et al. 2005; Opeskin and Berkovic 2003). Anxiolytic drugs speciἀcally were found to be more common in SUDEP in one study (Nilsson et al. 1999) and less common in SUDEP in another (Opeskin and Berkovic 2003). Antipsychotic use was speciἀcally examined in one study (Nillson 1999) and was not associated with SUDEP. The association between clinical use of selective serotonin reuptake inhibitors and SUDEP has not been examined.

12.6â•…A ED Use and SUDEP 12.6.1â•…A ED Compliance and SUDEP Initial medical examiner series found that most SUDEP victims had subtherapeutic AED levels (Leestma et al. 1989; Terrence et al. 1975; Kloster and Torstein 1999). This led to the conclusion that noncompliance was associated with SUDEP. Controlled medical examiner series have not consistently found subtherapeutic AED levels in SUDEP cases (George and Davis 1998; Opeskin et al. 1999). Furthermore, antiepileptic drug metabolism continues after death so postmortem AED levels may not accurately reflect compliance (see Walczak 2003 for review). Comparison of AED levels at antemortem visits in SUDEP and control patients has not found evidence of decreased noncompliance in SUDEP patients (Walczak et al. 2001; Nilsson et al. 2001; Langan et al. 2005). Overall, noncompliance with antepileptic drug treatment does not appear to be an important risk factor for SUDEP. 12.6.2â•…A ED Polytherapy and SUDEP AED polytherapy at death was noted to be common in initial studies (Leestma et al. 1989; Terrence et al. 1975; Tennis et al. 1995). This was initially thought to be a surrogate marker for severe epilepsy rather than an independent risk factor. However, two controlled studies found that treatment with more than two AEDs was associated with SUDEP, even after adjusting for the number of seizures and seizure type (Walczak et al. 2001; Nilsson et al. 1999). A third controlled study adjusting for seizure frequency and type did not ἀnd SUDEP to be associated with the number of AEDs at the time of death, but did ἀnd SUDEP to be associated with the number of AEDs ever used (Langan et al. 2005). Thus, it appears that AED polytherapy is associated with SUDEP, even after adjusting for seizure severity and frequency. The pathophysiologic implications of these ἀndings are unclear. AED polytherapy could have additive adverse effects on cardiac function (see Section 12.6.3). Alternatively, increased sedation associated with polytherapy could prolong recovery from the postictal state and cause greater susceptibility to the postictal central apnea and positional asphyxia that may play a role in SUDEP. 12.6.3â•…Individual AEDs and SUDEP This raises the question as to whether any individual AED or combination is associated with SUDEP. Carbamazepine use appears particularly likely to be associated with cardiac arrhythmia and alteration of cardiac autonomic function (see Walczak et al. 2003 for

Risk Factors for Sudden Death in Epilepsy

195

review). Consequently several reports have asked whether carbamazepine use is associated with SUDEP with mixed ἀndings. An uncontrolled series found what appeared to be a high rate of carbamazepine use in people succumbing to SUDEP (Timmings 1998). Four larger controlled series found that carbamazepine use was equally likely in SUDEP cases and in control subjects (Walczak et al. 2001; Nilsson et al. 1999; Kloster and Torstein 1999; Opeskin et al. 1999). A cohort based controlled study (Jick et al. 1992) found that carbaÂ� mazepine use was less likely in SUDEP cases. The largest SUDEP series published (Langan et al. 2005) found that SUDEP was associated with current carbamazepine use (odds ratio 2.0, 95% conἀdence interval 1.1–3.8), even after adjustment for potential confounders. This series of 154 SUDEP cases had the highest power to detect association between individual AED use and SUDEP. However, this was not a cohort-based study; cases and controls were drawn from differing sources, allowing an introduction of bias. Further analyses have asked whether carbamazepine toxicity is associated with SUDEP. Two studies have not found association between carbamazepine toxicity and SUDEP (Walczak et al. 2001; Opeskin et al. 1999). However, a more detailed analysis (Nilsson et al. 2001) found that SUDEP risk was increased more than nine-fold with toxic carbamazepine concentrations at the time of last visit, even after adjusting for confounders. Risk was further increased after adjusting for number of AED dose changes in the last year. SUDEP risk was also increased nine-fold with low carbamazepine concentrations at last visit, but only when more than one AED dose change had been made in the last year. In contrast, SUDEP risk was not increased with therapeutic carbamazepine concentrations, irrespective of how many AED dose changes had been made. The authors concluded that frequent changes of carbamazepine dose with concentrations outside therapeutic range were an independent risk factor for SUDEP after adjustment for seizure severity. This would go along with the idea that abrupt large fluctuations in carbamazepine levels exacerbate cardiac autonomic instability in people with epilepsy and increase SUDEP risk. The occurrence of SUDEP with other epilepsy treatments has not been thoroughly examined. SUDEP may be less frequent with phenytoin use than with carbamazepine use (Nilsson et al. 2001); it is not clear whether this reflects a deleterious effect of carbaÂ� mazepine or a protective effect of phenytoin. Single retrospective studies have reported that SUDEP is less common when lamotrigine (Leestma et al. 1997) or the vagal nerve stimulator (Annegers et al. 2000) is used.

12.7â•…Novel SUDEP Risk Factors: Recent Work Several risk factors have been proposed based on preliminary observations. We discuss them because they illustrate the study of SUDEP risk factors related to pathophysiologic theories, rather than standard demographic variables. In general, two major theories of SUDEP pathogenesis have been proposed (Tomson et al. 2008). One holds that severe postÂ� ictal cerebral inhibition leads to postictal central apnea, which, together with obstructive apnea, leads to arrhythmia and death. Another theory holds that the signiἀcant adrenergic stimulation associated with frequent tonic–clonic seizures results in microscopic cardiac lesions such as subendocardial ἀbrosis or contraction band necrosis. These then act as potential foci for arrhythmia, perhaps triggered by adrenergic stimulation associated with further tonic–clonic seizures. Arrhythmia risk may already be increased in this population because of the cardiac autonomic abnormalities thought to be more common in people

196 Sudden Death in Epilepsy: Forensic and Clinical Issues

with epilepsy, as previously discussed. Some literature has begun to address risk factors based on these pathophysiologic theories. 12.7.1â•…SCN1A Mutations Two cases of SUDEP have been reported in a family with generalized epilepsy with febrile seizures and with a mutation in the sodium channel gene SCN1A (Hindocha et al. 2008). Mortality and SUDEP rates are increased in severe myoclonic epilepsy of infancy—an epilepsy syndrome also caused by an SCN1A mutation (Dravet et al. 2005). SCN1A is expressed in the heart. This has led to the idea that SCN1A mutations may predispose people with epilepsy to arrhythmia and SUDEP (Nashef et al. 2007). In principle, comparing the prevalence of SCN1A mutations in SUDEP victims and persons with epilepsy controls dying of other causes could answer this question. This approach could be extended to other genes predisposing to cardiac arrhythmias (Nashef et al. 2007). 12.7.2â•…Heart Rate Variability and Cardiac Autonomic Instability Heart rate variability is an indicator of cardiac autonomic function and can be easily assessed by a standardized analysis of R–R intervals during electrocardiography. Heart rate variability changes are known to be associated with sudden death in the general population. Heart rate variability changes are more common in people with epilepsy, though it is not clear whether the epilepsy itself, use of AEDs, or conditions comorbid with epilepsy are responsible (Tomson et al. 1998; Walczak 2003). This has led to the idea that altered heart rate variability may be a marker of SUDEP risk (Yuen and Sander 2004; DeGiorgio et al. 2008). Observers have further noted that omega-3 fatty acids reduce sudden cardiac deaths in healthy subjects and may therefore help prevent SUDEP (Yuen and Sander 2004). Preliminary studies suggest that omega-3 fatty acids normalize HRV in people with epilepsy (DeGiorgio et al. 2008). A large study examining whether omega-3 fatty acids can prevent SUDEP is being planned. 12.7.3â•…A natomic and Electrophysiologic Substrates of SUDEP 12.7.3.1â•…A natomic Substrates of Postictal Apnea If postictal central and obstructive apnea are responsible for SUDEP, anatomic features associated with obstructive sleep apnea (increased body mass index, increased neck circumference, nasal obstruction, decreased pharyngeal diameter, etc.) should be more common in people with SUDEP than in persons with epilepsy controls. This information does not appear in the literature, though such data should be easy to obtain from material in medical examiner series. Similarly, a person with epilepsy will adjust spontaneously in the postictal state to maintain an open airway. One would expect such spontaneous adjustments to be less common in people with cerebral palsy or other conditions limiting movement. This idea is supported by reports that SUDEP is more common in people with developmental delay (Walczak et al. 2001) and inability to ambulate (McKee and Bodἀsh 2000). 12.7.3.2â•…A natomic Substrates of Cardiac Arrhythmia in SUDEP If cardiac arrhythmia due to microscopic cardiac lesions and the adrenergic surge associated with a tonic–clonic seizure is responsible for SUDEP, microscopic subendocardial

Risk Factors for Sudden Death in Epilepsy

197

abnormalities or conduction abnormalities should be more common in SUDEP victims than in control persons with epilepsy. Information from small, controlled series indicates that subendocardial abnormalities may be more common in persons with epilepsy than in control subjects without epilepsy (P-Codrea Tigaran et al. 2005; Natelson 1998). However, prevalence of subendocardial or conduction abnormalities does not differ between people with epilepsy dying from SUDEP and those dying of other causes (Opeskin et al. 2000). Nonetheless, these are small studies with inadequate power to exclude a potential contribution from subtle cardiac lesions, so further study is warranted. 12.7.3.3â•…Electrophysiologic Substrates of Cardiac Arrhythmia in SUDEP If the hearts of persons with epilepsy are, in fact, more susceptible to arrhythmia, one would expect nonfatal arrhythmia to be more common in persons with epilepsy than in control populations. A controlled study of 24 to 48 hours of cardiac monitoring has not found this to be the case (Blumhardt et al. 1986). Long-term electrocardiographic recordings of small numbers of people with severe epilepsy, usually lasting for many months, have found periods of asystole generally thought to require intervention in 15% (Rugg-Gunn et al. 2004). However, these studies are not controlled; the incidence of asystole during prolonged recordings in people with chronic disease (or, for that matter, in healthy young men) is not known. This amalgam of information raises the question whether people with severe epilepsy should undergo long term electrocardiographic monitoring and whether asymptomatic arrhythmias found by such monitoring should be treated. Long-term cardiographic monitoring of a larger group of persons with epilepsy is currently underway (Cooper 2008). However, this appears to be an uncontrolled study so it may be difficult to determine the clinical relevance of the information obtained.

12.8â•…The Next Steps in the Study of SUDEP Risk Factors The studies reviewed here have established a reasonably consistent risk proἀle for SUDEP. Persons with epilepsy succumbing to SUDEP suffer from generalized tonic–clonic seizures, have longer durations of epilepsy, and are often treated with multiple AEDs. Other potential risk factors such as treatment with speciἀc classes of psychotropic drugs, or treatment with speciἀc AEDs deserve further exploration. Examining these potential risk factors will be challenging for two reasons. First, SUDEP is uncommon in the general population and persistent surveillance of large groups is required for detection. Second, any risks associated with putative risk factors will require adjustment for risks known to be associated with the traditional risk factors described above. It is clear that future research should move beyond the old approach of retrospective case accumulation with convenience controls in convenience populations. Analysis of what we have called traditional risk factors together with animal studies have led to reasonably supported pathophysiologic theories (Lathers 2010, Chapter 25; Lathers and Schraeder 2010, Chapter 28; Lathers and Levin 2010, Chapter 33; Alkadhi and Alzoubi 2010, Chapter 26; Bealer et al. 2010, Chapter 38; Faingold et al. 2010, Chapter 41; Stewart 2010, Chapter 39; Goodman et al. 2010, Chapter 40). We are now in a position to study novel risk factors that support or detract from those theories. The state of knowledge is such that prospective studies validating traditional risk factors can now be undertaken. This would require periodic standardized longitudinal

198 Sudden Death in Epilepsy: Forensic and Clinical Issues

surveillance for SUDEP in multiple centers but should be feasible with a well-organized approach. With a little more information, hypothesis-based intervention studies in populations at high risk should soon be feasible as well. Much work will need to be done to increase recognition of SUDEP and set up the core infrastructure for such studies (So et al. 2009). Given what we have learned already, such an investment is quite likely to lead to more complete understanding and, ultimately, prevention of this tragic condition.

References Annegers, J. F., S. P. Coan, W. A. Hauser, and J. Leestma. 2000. Epilepsy, vagal nerve stimulation by the NCP system, all-cause mortality, and sudden, unexpected, unexplained death. Epilepsia 41: 549–553. Beghi, E., M. Leone, and A. Solari. 2005. Mortality in patients with a ἀrst unprovoked seizure. Epilepsia 46 (11): 40–42. Blumhardt, L. D., P. E. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1 (8489): 1051–1056. Camἀeld, C. S., P. R. Camἀeld, and P. J. Veugelers. 2002. Death in children with epilepsy: A populationbased study. Lancet 39: 1891–1895. Cooper, P. N. 2008. Cardiac arrhythmias and SUDEP—Commentary on Hindocha et al. Epilepsia 49: 366–367. Dashieff, R. M. 1991. Sudden unexpected death in epilepsy: A series from an epilepsy surgery program and speculation on the relationship to sudden cardiac death. J Clin Neurosci 8: 216–222. DeGiorgio, C. M., P. Miller, S. Meymandi, and J. A. Gombein. 2008. n-3 fatty acids (eicosapentanoic acid and docosahexanoic acids) in epilepsy and for the prevention of sudden death in epilepsy. Epilepsy Behav 13: 681–684. Donner, E. J., C. R. Smith, and O. C. Snead. 2001. Sudden unexplained death in children with epilepsy. Neurology 57: 430–434. Dravet, C., M. Bureau, H. Oguni, Y. Fukayama, and O. Cokar. 2005. Severe myoclonic epilepsy in infancy (Dravet syndrome). In Epileptic Syndromes in Infancy, Childhood and Adolescence, 4th ed, ed. J. Roger, M. Bureau, C. Dravet, P. Genton, C. A. Tassinari, and P. Wolf, 89–114. London: John Libbey. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossak. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical and postmortem features. Epilepsia 33: 310–316. Ficker, D. M., E. L. So, and W. K. Shen et al. 1998. Population based study of the incidence of sudden unexplained death in epilepsy. Neurology 51: 1270–1274. Freytag, E., and R. Lindenberg. 1964. 294 medicolegal autopsies in epileptics. Arch Pathol 78: 274–286. George, J. R., and G. G. Davis. 1998. Comparison of anti-epileptic drug levels in different cases of sudden death. J Forensic Sci 43: 598–603. Harvey, A. S., T. Nolan, and J. B. Carlin. 1993. Community based study of mortality in children with epilepsy. Epilepsia 34: 597–603. Hauser, W. A., and D. C. Hesdorffer. 1990. Incidence and prevalence. In Epilepsy: Frequency, Causes, and Consequences, 1–51. New York, NY: Demos Publications. Hiltris, N., S. Suratman, K. Kelly, L. J. Stephen, G. J. Sills, and M. J. Brodie. 2007. Sudden unexpected death in epilepsy: A search for risk factors. Epilepsy Behav 10: 138–141. Hindocha, N., L. Nashef, and F. Elmslie et al. 2008. Two cases of sudden unexpected death in epilepsy in a GEFS+ family with an SCN1A mutation. Epilepsia 49: 360–365. Hirsch, C. S., and D. L. Martin. 1971. Unexpected death in young epileptics. Neurology 21: 682–689. Jick, S. S., T. B. Cole, R. A. Mesher, P. Tennis, and H. S. Jick. 1992. Sudden unexpected death in young persons with primary epilepsy. Pharmacoepidemiol Drug Safety 1: 59–64.

Risk Factors for Sudden Death in Epilepsy

199

Kloster, R., and E. Torstein. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67: 439–444. Langan, Y., L. Nashef, and J. W. A. S. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Leestma, J. E., T. W. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26: 195–203. Leestma, J. E., J. F. Annegers, and M. J. Brodie et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. Leppik, I. E. 1995. Tiagabine: The safety landscape. Epilepsia 36: (Sl6): S10–S13. Lhatoo, S. D., A. L. Johnson, D. M. Goodrich, B. K. MacDonald, J. W. Snyder, and S. D. Shorvon. 2001. Mortality in epilepsy in the ἀrst 11 to 14 years after diagnosis: Multivariate analysis of a long-term prospective, population-based cohort. Ann Neurol 49: 336–344. Lip, G. W., and M. J. Brodie. 1992. Sudden death in epilepsy: An avoidable outcome? J R Soc Med 85: 609–611. McKee, J. R., and J. W. Bodἀsh. 2000. Sudden unexpected death in epilepsy in adults with mental retardation. Am J Ment Retard 105: 229–235. Nashef, L., D. R. Fish, J. W. A. S. Sander, and S. D. Shorvon. 1995. Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral center. J Neurol Neurosurg, Psychiatry 58: 462–464. Nashef, L., N. Hindocha, and A. Makoff. 2007. Risk factors in sudden death in epilepsy (SUDEP): The quest for mechanisms. Epilepsia 48: 859–871. Natelson, B. H., R. V. Suzrez, C. F. Terrence, and R. Turzio. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55: 857–60. Nilsson, L., U. Bergman, and V. Diwan. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42: 667–673. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case control study. Lancet 35: 888–893. Opeskin, K., and S. F. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12: 456–464. Opeskin, K., M. P. Burke, and S. M. Cordner. 1999. Comparison of antiepileptic drug levels in sudden unexpected deaths in epilepsy with deaths from other causes. Epilepsia 40: 1795–1798. Opeskin, K., A. Thomas, and S. F. Berkovic. 2000. Does cardiac conduction pathology contribute to sudden unexpected death in epilepsy? Epilepsy Research 40: 17–24. P-Codrea Tigaran, S., S. Dalanger-Pedersen, U. Baandrup, M. Dam, and A. Vesterby-Charles. 2005. Sudden unexpected death in epilepsy: Is death by seizures a cardiac disease? Am J Forensic Med Pathol 26: 99–105. Ray, W. A., C. P. Chung, K. Murray, K. Hall, and C. M. Stein. 2009. Atypical antipsychotic drugs and the risk of sudden cardiac death. N Engl J Med 360: 225–235. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364: 2212–2219. Sperling, M. R., H. Fedman, J. Kinman, J. D. Liporace, and M. J. O’Connor. 1999. Seizure control and mortality in epilepsy. Ann Neurol 46: 45–50. So, E. L., J. Bainbridge, J. R. Buchhalter, and J. Donalty et al. 2009. Report of the American Epilepsy Society and the Epilepsy Foundation Joint Task Force on Sudden Unexplained Death in Epilepsy. Epilepsia 50: 917–922. Tellez-Zenteno, J. F., L. H. Ronquillo, and S. Wiebe. 2005. Sudden unexpected death in epilepsy: Evidence-based analysis of incidence and risk factors. Epilepsy Res 65: 101–115. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36: 29–36.

200 Sudden Death in Epilepsy: Forensic and Clinical Issues Terrence, C. F., H. M. Wiszotzkey, and J. A. Perper. 1975. Unexpected, unexplained death in epileptic patients. Neurology 25: 594–598. Timmings, P. L. 1993. Sudden unexpected death in epilepsy: A local audit. Seizure 2: 287–290. Timmings, P. L. 1998. Sudden unexpected death in epilepsy: Is carbamazepine implicated? Seizure 7: 289–291. Tomson, T., M. Ericson, C. Ihrman, and L. E. Lindblad. 1998. Heart rate variability in patients with epilepsy. Epilepsy Res 30: 77–83. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7 (11): 1021–1031. Tomson, T., T. Walczak, M. Silanpaa, and J. W. A. S. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46: (S11): 54–61. Tupal, N. and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47: 21–26. Walczak, T. 2003. Do antiepileptic drugs play a role in sudden unexpected death in epilepsy? Drug Saf 26: 673–683. Walczak, T. S., I. E. Leppik, M. D’Amelio et al. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 6: 51–55. Weber, P., R. Bubl, U. Blauenstein, B. U. Tillman, and J. Lutschg. 2005. Sudden unexplained death in children with epilepsy: A cohort study with an eighteen-year follow up. Acta Paediatr 94: 564–567. Yuen, A. W., and J. W. Sander. 2004. Is omega-3 fatty acid deἀciency a factor contributing to refractory seizures and SUDEP? A hypothesis. Seizure 13: 104–107.

EEG Findings in SUDEP Maromi Nei Nicole Simpkins

13

Contents 13.1 Introduction 13.2 EEG Data in Patients Who Subsequently Died due to SUDEP 13.3 EEG in SUDEP and Near-SUDEP 13.4 Conclusions References

201 201 202 205 206

13.1â•…Introduction Most of the data regarding the EEG during SUDEP are obtained from case reports from patients undergoing ambulatory EEG or video-EEG monitoring. These data are limited, but the EEG during SUDEP has generally shown a generalized tonic–clonic seizure with subsequent suppression. Ictal and interictal EEG recordings in patients with SUDEP have revealed varied ἀndings.

13.2â•…EEG Data in Patients Who Subsequently Died due to SUDEP In the majority of cases of witnessed SUDEP, a seizure is reported to precede death. In 12 of 15 witnessed cases of SUDEP, a generalized tonic–clonic seizure preceded death (Langan et al. 2000). In the other three cases in this series, either an aura occurred or the patient was thought to be in a postictal state. Since seizures usually precede death in SUDEP, this suggests that seizures are often responsible for triggering the physiologic changes that ultimately lead to SUDEP. Thus, the evaluation of the EEG and other physiologic data during seizures and interictally could yield clues regarding the pathophysiology of SUDEP. The video-EEG monitoring data in individuals who subsequently died due to SUDEP have revealed seizures of various localization and lateralization (Nei et al. 2004). In this study, video-EEG data from 21 patients who subsequently died due to SUDEP were reviewed and compared with video-EEG data from a control refractory focal epilepsy population. The majority (86%) of patients had ictal EEG recorded, in addition to interictal EEG data. Localization of seizure onset was conἀdently identiἀed as arising from the left hemisphere in 29% and from the right hemisphere in 14%. The remainder had seizures that were either nonlateralized (24%), multifocal (5%), or generalized (14%) in onset. Of these patients, 43% were thought to have temporal lobe onset for their seizures and 19% had a probable frontal lobe onset; however, these data are likely biased since the majority of patients were admitted as part of their evaluations for potential epilepsy surgery. No speciἀc lobe of onset or lateralization for seizures was more common in SUDEP than in control patients. 201

202 Sudden Death in Epilepsy: Forensic and Clinical Issues

There are also data to suggest that prolonged seizures, generalized tonic–clonic seizures, as well as seizure clusters, might increase the risk for increased autonomic stimulation during seizures and might increase the risk for cardiac arrhythmias (Nei et al. 2000, 2004). Unfortunately, there were no speciἀc EEG ἀndings that were predictive of SUDEP. Most patients with SUDEP die in their sleep. EEG data also suggest that patients with SUDEP are more likely to have a history of and/or documentation of seizures arising from sleep, as compared with control patients (Nei et al. 2004). Opherk et al. also found that in patients with refractory epilepsy (in a non-SUDEP population), there was a trend toward increased ictal-related EKG abnormalities during seizures associated with sleep, suggesting a potential speciἀc sleep-related risk on autonomic status during seizures arising at this time, which might increase risk for SUDEP. The reader is referred to the chapter on sleep and SUDEP by Drs. Sato and Hughes in this book. While it is clear that seizures can cause ictal and postictal EKG rate and repolarization abnormalities, the potential interictal effects of epileptiform abnormalities on cardiac and pulmonary function are not as clear. One study evaluated the effect of interictal epileptiform EEG discharges on the QTc interval of the EKG in patients who subsequently died due to SUDEP (Tavernor et al. 1996). In this study, the EEGs influence on EKG data from eleven patients with SUDEP were compared with data from 11 age and sex matched control patients, also with uncontrolled tonic–clonic seizures who were alive at the time of the investigation. They found that only for those with SUDEP, the QTc interval was signiἀcantly prolonged during epileptiform EEG discharges. This led to the speculation that prolonged QTc intervals might increase the likelihood for potentially lethal ventricular arrhythmias and sudden cardiac death. However, additional data are needed to conἀrm these ἀndings. No speciἀc information regarding the type or localization of the EEG epileptiform abnormalities is available from this study. Most epidemiologic studies on SUDEP have focused on seizure type, rather than speciἀc EEG ἀndings. Generalized tonic–clonic seizures increase the risk for SUDEP and these may be either primarily or secondarily generalized seizures (Nei et al. 2004).

13.3â•…EEG in SUDEP and Near-SUDEP There are few case reports of SUDEP or near SUDEP captured during EEG recording available in the literature (see Table 13.1). One case of SUDEP captured during video-EEG monitoring includes a 41-year-old woman with refractory focal epilepsy since infancy (Lee 1998). Interictally, she had independent bitemporal sharp waves, with left greater than right. During sleep, she had an unwitnessed secondarily generalized tonic–clonic seizure lasting 70 seconds, which was followed by diffuse slowing on the EEG and left temporal sharp waves for 40 seconds, then marked suppression, with overlying EKG artifact seen that failed to recover. The EKG initially showed bradycardia to 30 beats per minute (bpm), which slowly increased to 70 bpm ἀve minutes after the seizure. However, the heart rate then began to slow and stopped 18 minutes after the seizure. There was no evidence of cardiovascular nor pulmonary abnormalities at autopsy, and no evidence of asphyxia. The cause of SUDEP in this case was postulated to be cessation of brain function. McLean and Wimalaratna (2007) reported a case of a woman in her ἀfties who died following a seizure while undergoing ambulatory EEG monitoring. Interictally, she had slow and sharp wave discharges that increased in frequency during sleep. During sleep,

EEG Findings in SUDEP

203

Table 13.1â•… EEG and Seizure Data: Cases of SUDEP and Near-SUDEP Interictal EEG

Report

Death

Dasheiff and Dickinson (1986)

Yes

N/A

Bird et al. (1997)

Yes

Lee et al. (1998)

Yes

L Temp, R temp, R occipital Bitemporal sharp waves, L€>€R

McLean and Wimalaratna (2007) So et al. (2000)

Yes

Tavee and Morris (2008) Espinosa et al. (2009)

No

Slow and sharp discharge Bifrontal sharp waves

No

L temp

No

Bitemporal theta/delta

Postictal EEG

Sz Type Prior to Event

N/A

N/A

CPS

R medial temporal

R suppressed, GTC then bilateral suppression GTC Diffusely Slow; L temp sharp waves × 40 secs Diffuse EEG c/w No suppression GTC

EKG: pulse artifact at 48 bpm EKG:€brady� cardia at 30 bpm€post� ictally

Bifrontal

Diffusely slow, then suppressed

GTC

Yes: 4 GTC in 6 h

L hem, maximal temp R temp

Continuous slowing

GTC

Yes: 1 aura, 4 partial and 2 GTC No

Apnea postictally, then€brady� cardia and asystole Stridor/ resp distress€af�� ter sz ended EKG: VT, then VF

Ictal EEG

N/A

Spike-wave discharge

Diffusely slow GTC

Sz Cluster Prior to Event Yes: cluster of 2 GTC and 1 CPS within 3 h Yes: cluster of 5 GTCs in 24 h No

Notes EKG: VF

—

Note: sz, seizure; L, left; R, right; temp, temporal; N/A, not available; hem, hemisphere; secs, seconds; GTC, generalized tonic–clonic seizures (either primary or secondarily generalized); c/w, consistent with; CPS, complex partial seizure; bpm, beats per minute; VF, ventricular ἀbrillation; resp, respiratory; VT, ventricular tachycardia.

a 52-second seizure beginning with spike and wave discharges was captured, followed abruptly by marked suppression that failed to recover. Further details regarding localization of ictal and interictal discharges are not available. Rhythmic movement artifact was seen at the T3 electrode, with associated muscle activity that became less frequent and disappeared completely 3 minutes after seizure termination, leaving a suppressed EEG. No EKG or respiratory data were available. One case of SUDEP during intracranial EEG monitoring was reported by Bird et al. (1997). A 47-year-old man with refractory seizures since age 19 was admitted for videoEEG monitoring. He had undergone ambulatory scalp monitoring and had three complex partial seizures with secondary generalization that began with right hemispheric slow waves, but were otherwise nonlocalized. Interictally, there were complex left temporal discharges, right temporal spikes, and occasional right occipital spikes. He then underwent implantation of bilateral temporal depth electrodes and subdural electrodes of the anterior

204 Sudden Death in Epilepsy: Forensic and Clinical Issues

and posterior temporal regions for further evaluation of his seizures. The operation and immediate postoperative period were uneventful. He subsequently had four secondarily generalized seizures and died following the ἀfth seizure at 3 a .m. Electrographic onset in all ἀve seizures was in the right medial temporal lobe. The ἀfth seizure occurred during sleep and began with head version to the left followed by turning of his body, then a generalized convulsion lasting for 2.5 minutes. At the end of the convulsion, he no longer moved. The EEG showed right mesial temporal onset, with the electrographic discharge spreading to the left hemisphere after 15 seconds, and subsequent generalized discharge lasting 2.5 minutes. The right-sided ictal discharge then briefly flattened, alternating with spindling spike discharges for 16 seconds, before stopping completely, leaving a suppressed background on the right. The left hemisphere continued to show spikes for 8 more seconds, then stopped. Pulse artifact, at 46 beats per minute, was seen for another 2 minutes and then gradually decreased in amplitude (continuing at the same rate) until the heart beat stopped. There were no respiratory or EKG recordings. Postmortem examination revealed mild congestion in the lungs and normal cardiac examination. The neuropathological examination showed an acute infarct in the right temporal lobe, attributed to insertion of the depth electrodes, acute hypoxic changes in the right hippocampus, evidence of old frontal contusions bilaterally, and bilateral occipital ulegyria. Dasheiff and Dickinson (1986) reported a case of sudden unexpected death in a patient undergoing video-EEG monitoring with intracranial electrodes. The patient was a 48-yearold man with a history of prior myocardial infarction and refractory focal epilepsy who had been implanted with depth electrodes for seizure focus localization. He had two secondarily generalized tonic–clonic seizures within 1 hour of each other. After the second seizure, the patient complained of chest and left arm pain. An EKG revealed ST segment elevation and inverted T waves. One hour later, the patient complained of chest pain, then was witnessed to have a complex partial seizure, becoming cyanotic and apneic. The ἀrst available EKG revealed coarse ventricular ἀbrillation, followed by asystole. Cardiopulmonary resuscitation was ineffective. Postmortem examination revealed no acute coronary artery thromboses or pulmonary emboli but did reveal an old myocardial infarction. The EEG ἀndings were not reported. It is likely that the acute sympathetic discharge due to the cluster of seizures precipitated both the angina and ventricular arrhythmia, to which he was already predisposed, due to his underlying previous myocardial infarction. A similar mechanism has been proposed as resulting in myocardial infarction in the setting of seizure (Chin et al. 2004). It is of interest to speculate that a vasospasm mechanism such as that observed in Prinzmetal Angina could account for the ST elevation and cause infarction in the absence of atherosclerosis. A case of near SUDEP due to postictal severe laryngospasm was reported by Tavee and Morris (2008). A 42-year-old man with refractory epilepsy since age 6 was admitted for video-EEG monitoring. Interictally, he had frequent left sphenoidal electrode sharp waves and less frequent posterior temporal sharp waves. He had one simple partial seizure and 5 complex partial seizures, two of which secondarily generalized. Ictal EEG showed left hemisphere, maximal in the temporal region, ictal fast activity. During seizure six, the patient awakened from sleep with right arm and face twitching, followed by right arm extension, left arm flexion, head version to the right and a generalized convulsion. The entire seizure lasted 82 seconds. Following the seizure, he developed inspiratory stridor and marked cyanosis, eventually requiring intubation for respiratory support. The anesthesiologist noted laryngospasm at the time of intubation. It was postulated that aspiraÂ� tion€may have triggered the laryngospasm. The postictal EEG showed diffuse slowing until

EEG Findings in SUDEP

205

6 hours after the seizure, when he developed nonconvulsive status epilepticus, which eventually resolved. He recovered to baseline and was ultimately discharged home. So et al. (2000) reported a case of near-SUDEP due to postictal central apnea. A 20-yearold woman with refractory epilepsy since age 1, was admitted for video EEG monitoring. She had frequent generalized convulsions and complex partial seizures and also had a history of convulsive status epilepticus at age 9 when she developed mononucleosis, which culminated in cardiorespiratory arrest. There were no permanent neurologic deἀcits after this event. She had no known cardiovascular or pulmonary disorders; however, following both clusters of seizures, as well as isolated, self-limited seizures, she had postictal respiratory arrest and had required cardiopulmonary resuscitation since age 10. It had been observed that following a seizure, her pulse was initially regular and strong, but this disappeared€as the apnea continued. Outpatient interictal EEG showed bisynchronous frontal sharp waves. During the monitoring, she had a cluster of four generalized convulsions within 6€hours, lasting 55–85 seconds each. Ictal EEG showed bifrontal slow waves at onset. She had fully recovered from the ἀrst three seizures when she had a fourth seizure, lasting 56 seconds. She was noted to have apnea immediately following the seizure. The EKG remained unchanged for 10 seconds, then gradually slowed until the heart stopped 57 seconds later. Postictally, the EEG showed diffuse slowing for 20 seconds, which was followed by marked suppression. She was successfully resuscitated, and follow-up cardiac evaluation showed no evidence of cardiac or pulmonary disease. A demand cardiac pacemaker was placed. Following implantation of the pacemaker, 10–15 second periods of apnea were noted after seizures; however, she no longer required cardiopulmonary resuscitation. Recently, a case of near-SUDEP revealed a right temporal complex partial seizure with secondary generalization associated with ventricular tachycardia (Espinosa et al. 2009). A 51-year-old woman underwent video-EEG monitoring for refractory focal epilepsy. Her interictal EEG revealed bitemporal independent theta and delta activity, and a baseline EKG revealed a ἀrst-degree atrioventricular block, with a normal QTc interval. She had a typical complex partial seizure with secondary generalization, which was associated with ventricular tachycardia, then ventricular ἀbrillation, toward the end of the seizure. The patient underwent successful cardiopulmonary resuscitation and subsequent deἀbrillator implantation.

13.4â•…Conclusions The EEG data from patients with SUDEP, near-SUDEP, or subsequent SUDEP reveal a variety of interictal ἀndings, with both focal and generalized interictal epileptiform abnormalities. While the case numbers are limited, the ictal EEG recordings from patients with SUDEP or near-SUDEP have uniformly recorded a terminal generalized tonic–clonic seizure, except in one case, which reported a cluster of two secondarily generalized tonic–clonic seizures and then a complex partial seizure just prior to death. This ἀnding is consistent with epidemiologic data that generalized tonic–clonic seizures increase risk for SUDEP. The data thus far do not implicate a speciἀc seizure focus lateralization or lobe of the brain being associated with a higher risk for SUDEP. These data, while limited, also suggest that seizure clusters might also increase risk for SUDEP. The EEGs from cases of SUDEP also reveal diffuse suppression after the seizure ends. Based on this ἀnding, the possibility of primary cerebral shutdown has been proposed

206 Sudden Death in Epilepsy: Forensic and Clinical Issues

(Bird et al. 1997). The sudden cessation of cerebral activity has been suggested to be due to primary irreversible brain failure, with cardiorespiratory failure occurring as a secondary consequence. However, it is important to note that there are limited cardiorespiratory data available in these cases. Alternatively, it is possible that seizures could result in concomitant cardiopulmonary abnormalities during the ictal or postictal phase of the seizure, resulting in either anoxia or decreased cardiac output. While there can be diffuse suppression of the EEG after an uncomplicated generalized tonic–clonic seizure, anoxia or decreased cardiac output, such as related to a seizure-related arrhythmia or pulseless electrical activity, may explain the persistence of this suppression and lack of recovery of the EEG in these SUDEP cases. While a primary cerebral etiology is possible, the data thus far suggest that the seizure itself is an important trigger for a cascade of respiratory and/or cardiac abnormalities that ultimately cause death. The near-SUDEP case of Espinosa et al. (2009) and the SUDEP case of Dasheiff and Dickinson (1986) were associated with ventricular tachyarrhythmias occurring during or toward the end of a seizure. In the two SUDEP cases with EKG or pulse artifact, bradycardia was recorded, suggesting that cardiac function was affected. The respiratory status in these cases is unknown, but it is possible that respiratory compromise could have occurred during the seizure, resulting in persistent postictal suppression of the EEG and reflex bradycardia. In the So et al. (2000) case, it appears that apnea triggered the bradycardia and treatment with a cardiac pacemaker insertion has been helpful. Even though the apnea was the initial event, the more concerning life-threatening cardiorespiratory response may have been the secondary cardiac effect of bradycardia, which had been eliminated by the pacemaker. More detailed analysis of EEG data and close correlation with cardiac and pulmonary function is needed to fully interpret the EEG ἀndings obtained thus far during SUDEP. Perhaps the use of more routine pulmonary and cardiac monitoring, along with EEG recordings, may yield further insights into the pathophysiology of SUDEP. Hopefully, specialized investigation, ideally evaluating the combined neurologic, cardiac, and pulmonary data in people with epilepsy during both the ictal as well as interictal states, will identify risk factors for SUDEP and provide targets for preventative therapy.

References Bird, J. M., K. A. T. Dembny, D. Sandeman, and S. Butler. 1997. Sudden unexplained death in epilepsy: An intracranially monitored case. Epilepsia 38 (S11): S52–S56. Chin, P. S., K. R. Branch, and K. J. Becker. 2004. Myocardial infarction following brief convulsive seizures. Neurology 63 (12): 2453. Dasheiff, R. M., and L. J. Dickinson. 1986. Sudden unexpected death of epileptic patient due to cardiac arrhythmia after seizure. Arch Neurol (43): 194–196. Espinosa, P. S., J. W. Lee, U. B. Tedrow, E. Bromἀeld, and B. A. Dworetzky. 2009. Sudden unexpected near death in epilepsy (SUNDEP): Malignant ventricular arrhythmia from a partial seizure. Neurology 72: 1702–1703. Langan, Y., L. Nashef, and J. W. A. S. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Lee, M. A. 1998. EEG video recording of sudden unexpected death in epilepsy. Epilepsia 39 (S6): 120–121. McClean, B. N., and S. Wimalaratna. 2007. Sudden death in epilepsy recorded in ambulatory EEG. J Neurol Neurosurg Psych 78: 1395–1397.

EEG Findings in SUDEP

207

Nei, M., R. T. Ho, and B. W. Abou-Khalil et al. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52: 117–127. So, E., M. Sam, and T. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41 (11): 1494–1497. Tavee, J., and H. Morris. 2008. Severe postictal laryngospasm as a potential mechanism for sudden unexpected death in epilepsy: A near-miss in an EMU. Epilepsia 49 (12): 2113–2117. Tavernor, S. J., S. W. Brown, R. M. E. Tavernor, and C. Gifford. 1996. Electrocardiograph QT lengthening associated with epileptiform EEG discharges—A role in sudden unexplained death in epilepsy? Seizure 5: 79–83.

Severity of Seizures as a Forensic Risk and Case Reports Edward H. Maa Michael P. Earnest Mark C. Spitz Jacquelyn Bainbridge

14

Contents 14.1 What Is a Severe Seizure? 14.2 Deἀnition 14.3 Epidemiology 14.4 SUDEP Risk Factors 14.5 Witnessed Cases 14.6 University of Colorado Epilepsy Monitoring Unit Case 14.7 Conclusion References

209 210 211 211 211 215 218 219

14.1â•…What Is a Severe Seizure? Seizures are a debilitating neurological condition characterized by the sudden evolution of synchronous electrical activity in the brain that can lead to loss of awareness, confusion, sudden falls, odd motor behaviors or sensations, and convulsions (Spitz 1998). Medically inexperienced, ἀrst-time witnesses of seizures are frightened or confused and often misinterpret the event as the impending death of the patient. What makes a seizure severe? The lay media implies the severity of seizures by the intense shaking, drooling, and eye rolling associated with a convulsion. The severity of a seizure in a typical medical encounter is judged by its duration and the presence of witnessed convulsion, tongue biting, and incontinence. To the patient stricken with seizures, however, each event is severe not only because of the clinical manifestation of the seizure, but also because of the psychosocial impact of the unpredictability of the seizure. The epileptologist’s approach to seizure severity is in alignment with the patient perspective but also emphasizes the likelihood of injury. This likelihood of injury is the foundation of seizure precautions and is intended as a safety recommendation because of the unpredictability of epileptic seizures. The sudden change or loss of awareness associated with seizures would not in and of itself be harmful if it only occurred in sleep. In the settings of operating heavy machinery, swimming or bathing alone, work or play at heights, tight spaces, or with open flames including cooking stoves, seizures can have devastating consequences. Even a single seizure in these settings increases the risk of death or serious injury. 209

210 Sudden Death in Epilepsy: Forensic and Clinical Issues

Unpredictability inevitably leads to anxiety. Any new neurology resident taking stroke calls can attest to many sleepless nights waiting for their ἀrst tissue plasminogen activator (tPA) candidate. New parents experience anxious and uneasy sleep, listening to the irregular breaths and sounds of a sleeping newborn infant. In much the same way, it is the overwhelming anticipation, rather than the seizure itself, that leads to chronic anxiety. In fact, of 1023 epilepsy patients who responded to an Epilepsy Foundation questionnaire, uncertainty and fear of the next seizure was rated as the worst thing about having epilepsy (Fisher et al. 2000). Fear eventually leads to restricted movements outside of the home, decreased productivity, and impaired livelihood. Psychiatric disturbances including depression and agoraphobia can be found in as many as 70%, and the risk of suicide in uncontrolled epilepsy patients is 13%, nine times that of the general population (Nowack 2006). Seizures themselves have graded severity; intuitively, the more violent the motor involvement, the more severe the seizure. This concept is manifest in the old terminology of “grand mal” and “petit mal” seizures. The imprecision of these terms from the perspective of therapeutics and modern epilepsy care ignores the fact that to the lay person, blinking and staring cannot possibly be as serious as convulsing and becoming cyanotic. The higher the frequency and longer the duration of an individual seizure, as well as the overall duration of epilepsy, add to the concept of severity, as do a number of surrogate markers including increased numbers and doses of medications, specialists, and surgeries. The ultimate marker of seizure and epilepsy severity is any seizure event that results in death. Recent sensationalized deaths in persons having a seizure emphasize the fact that uncontrolled epilepsy should not be thought of as a chronic condition that merely necessitates annual reἀlling of medications (Epilepsy Foundation 2009; Phillips 2008). Obviously, patients with persistent seizures are at increased risk of fractures, burns, and drowning, but they also exhibit increased rates of depression, anxiety, and suicide (Sperling 2004; Spitz et al. 1994). Less well known in the general medical community is that sudden unexplained death in epilepsy (SUDEP) is the most common cause of seizure-related death (Langan et al. 2000; Langan and Nashef 2003), accounting for as many as 50% of early deaths in refractory epilepsy patients (Sperling 2001).

14.2â•…Definition Sudden unexplained death in epilepsy can be established by applying the criteria developed by an expert panel (Leestma et al. 1997):

1. Diagnosis of epilepsy 2. Death occurring unexpectedly while in a reasonable state of health 3. Death occurring suddenly 4. Death occurring during normal activities and benign circumstances 5. No obvious medical cause of death determined during postmortem examination 6. Death is not the result of trauma, asphyxia from aspiration, or status epilepticus

Death from SUDEP is “deἀnite” if all conditions are satisἀed and “probable” if no postmortem data is available.

Severity of Seizures as a Forensic Risk and Case Reports

211

14.3â•…Epidemiology Fortunately, SUDEP remains relatively rare with annual incidence ἀgures ranging from 0 to 10 per 1000 patients, depending on the studied population (Tellez-Zenteno et al. 2005). Reflected in these numbers is the suggestion that seizure severity accounts for a higher incidence of SUDEP, with epilepsy surgical candidates representing the higher end of the spectrum and general population coroner’s cases representing the lower end of the spectrum. SUDEP appears to be mainly a problem in patients with refractory epilepsy. In the National General Practice Study of Epilepsy in the United Kingdom (NGPSE), there was only one conἀrmed case of SUDEP in 7147 person years. In this prospective cohort of 564 patients, 70% became seizure-free for at least 5 years of the mean 15 years of follow-up (Sander et al. 1990).

14.4â•…SUDEP Risk Factors Accumulated risk factors from descriptive cohort studies over the years include youth (Leestma et al. 1997; Opeskin and Berkovic 2003), male gender (Tennis et al. 1995), early onset of epilepsy (Kloster and Engelskjon 1999; Nilsson et al. 1999), duration of epilepsy and seizure frequency (Walczak et al. 2001; Leestma et al. 1997), poor control of seizures (Sperling et al. 1999), convulsive seizure type (Kloster and Engelskjon 1999; Birnbach et al. 1991), high antiepileptic drug number (Nilsson et al. 1999; Tennis et al. 1995; Walczak et al. 2001; Racoosin et al. 2001), frequency of antiepileptic drug changes (Nilsson et al. 1999), subtherapeutic antiepileptic drug levels (Kloster and Engelskjon 1999; Earnest et al. 1992), mental retardation (Walczaket al. 2001), concomitant use of psychotropic medications (Nilsson et al. 1999; Tennis et al. 1995), prone position (Kloster and Engelskjon 1999), and being found in bed or home (Kloster and Engelskjon 1999; Opeskin and Berkovic 2003). Despite inconsistent methodologies, patient populations, and autopsy availability, Stollberger and Finsterer (2004) and Tellez-Zenteno et al. (2005) summarized these ἀndings in their works, providing a picture of a mid to late 30s male living alone with poorly€controlled, long-standing symptomatic convulsive epilepsy, on multiple medications, found dead in bed in the prone position, often with evidence of a recent seizure.

14.5â•…Witnessed Cases The elusiveness of SUDEP’s etiology likely remains because of its predilection for unwitnessed sleep, relative rarity, and its multifactorial nature. Witnessed case reports over the years have shed some light on this devastating condition. Dasheiff and Dickinson (1986) described a 48-year-old male from the Wisconsin Epilepsy Center with a witnessed complex partial seizure who recovered, but then suffered what appeared to be a second seizure accompanied by ventricular ἀbrillation on EKG. Even though witnessed in the hospital, he was not successfully resuscitated and went on to receive an unrevealing autopsy. The authors suspected a cardiac origin of SUDEP. Later, Dasheiff (1991) described a much higher incidence of SUDEP, almost 1:100, than previously reported from their experience at the Pittsburgh Epilepsy Center. One of the seven patients he described was in the midst of transfer from the intensive care unit to

212 Sudden Death in Epilepsy: Forensic and Clinical Issues

the hospital floor after receiving a temporal lobectomy for refractory epilepsy when she suddenly expired while sitting unmonitored in bed. Efforts to resuscitate her were also fruitless and, despite another negative autopsy, the author suspected a cardiac cause as the leading etiologic hypothesis. Purves et al. (1992) reported a case of a 27-year-old woman from the British Columbia epilepsy program. She had complex partial seizures during her monitoring stay, but her last event was a secondarily generalized tonic–clonic seizure that resulted in her lying in the prone position. She was found 24 min later, cyanotic and unable to be resuscitated. Authors attributed this death to asphyxia following severe postictal depression. Bird et al. (1997) reported the ἀrst intracranially monitored case, from their experience in Bristol, England. The patient had bilateral, anterior and posterior temporal depth electrodes placed. Two days later, all medications were withdrawn and he had four typical seizures with right mesial temporal onset of electrical activity followed by blank staring, left head turn, then 2−4 min of convulsion. His ἀfth seizure occurred in sleep during which his left head turn was followed by his whole body turning to the prone position. He convulsed for 2.5 min then stopped moving and never recovered. Evidence of labored breathing or asphyxia was not seen by the authors and, unfortunately, there was no cardiac rhythm strip associated with this recording. The electroencephalogram (EEG) revealed an unusual right mesial spindling spike-discharge burst suppression activity for 16 s, followed by electrical silence. The left hemisphere showed rhythmic spike discharges for an

7 µV/mm

Fp1 - F7

7 µV/mm

F7 - T3 T3 - T5

7 µV/mm

T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm

F8 - T4

7 µV/mm

T4 - T6

7 µV/mm

T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm 7 µV/mm

F3 - C3

7 µV/mm

C3 - P3 P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm 7 µV/mm

F4 - C4

7 µV/mm

C4 - P4 P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm

Cz - Pz

7 µV/mm

ECG

50 µV/mm

00:17:34

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 50 µV/mm, LFF: 1 Hz, HFF: 70 Hz, Notch: Off

xltek

Figure 14.1╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

Severity of Seizures as a Forensic Risk and Case Reports

213

additional 8 s when they also disappeared permanently. Pulse artifact was seen for an additional 2 min, prompting the authors to downplay the cardiac arrhythmia hypothesis of SUDEP. The lack of slow wave changes on EEG was also not suggestive of cerebral anoxia. Based on the pattern of spindling spike discharges seen as the terminal event, the authors suggested that dysregulation of thalamocortical regulatory systems may not only shut off cortical activity, but may be responsible for cessation of all life functions. Lee (1998) reported a case from Calgary during which a woman’s convulsion was followed by bradycardia to a rate of 30 beats/min. A minute after seizure cessation, the EEG became silent and never recovered. Interestingly, the bradycardia resolved after 4 additional minutes, returning to a rate of 70 beats/min, until the patient ἀnally expired 18 min later. A cardiac cause was also not suspected as an explanation for the death. Finally, McLean and Wimalaratna (2007) described an ambulatory EEG case from the United Kingdom of a woman with poorly controlled epilepsy who underwent ambulatory monitoring. Her ambulatory recorder documented increasing spike frequency once she had fallen asleep, eventually coalescing into an ictal event at 08:27:18 the next morning. Reminiscent of the intracranial case, the seizure exhibited polyspike (up to 6 spikes) activity for 52 s, abruptly terminated at 08:28:14, leaving an isoelectric EEG. No cardiac rhythm information was recorded, but rhythmic movement artifact was seen in the left temporal leads associated with muscle activity. It slowed to complete cessation of activity over the next 3 min. The patient was found the next morning in her night clothes, prone on

7 µV/mm

Fp1 - F7

7 µV/mm

F7 - T3

7 µV/mm

T3 - T5 T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm 7 µV/mm

F8 - T4

7 µV/mm

T4 - T6 T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm 7 µV/mm

F3 - C3

7 µV/mm

C3 - P3

7 µV/mm

P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm

F4 - C4

7 µV/mm

C4 - P4 P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm 7 µV/mm

Cz - Pz

50 µV/mm

ECG

00:18:54

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 50 µV/mm, LFF: 1 Hz, HFF: 70 Hz, Notch: Off

xltek

Figure 14.2╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

214 Sudden Death in Epilepsy: Forensic and Clinical Issues

the floor with arm outstretched toward the telephone. The authors also could not attribute the death to hypotension or cerebral anoxia due to the lack of slow wave activity on the EEG (Neidemeyer and Da Silva 1987). They coined the phrase abrupt irreversible “cerebral electrical shutdown” to explain the primary mechanism of SUDEP. The largest case-control study of SUDEP is from the United Kingdom, where autopsies are routinely performed on suspected seizure deaths. Langan et al. (2005) published their results of 154 cases of conἀrmed SUDEP. Of the conἀrmed SUPDEP cases, 15% (23 cases) were witnessed, with the majority following a convulsive seizure and associated with breathing difficulties. In addition to previous accounts of cardiac arrhythmias, they suggest central and obstructive apneas are likely contributors to SUDEP mechanisms, based on the high frequency of labored breathing reported. In the controlled environment of epilepsy monitoring and telemetry units, cardiac cases may be enriched because nurse interventions, such as rolling patients into the recovery position, nasal cannula oxygen, as well as engaging the postictal patient in the neurological exam, may be sufficient to prevent deaths from apneic mechanisms. This argument is supported by their ἀnding that supervision was a protective factor in their study, and further supported by a study of SUDEP incidence at a residential school for children with epilepsy who were supervised at night and carefully monitored after a seizure. Of the 310 students enrolled between 1970 and 1993, there were no SUDEP deaths during term, but 14 sudden deaths while at home on vacation. Most were unwitnessed (Nashef et al. 1995).

Fp1 - F7

15 µV/mm

F7 - T3

15 µV/mm

T3 - T5

15 µV/mm

T5 - O1

15 µV/mm

Fp2 - F8

15 µV/mm

F8 - T4

15 µV/mm

T4 - T6

15 µV/mm

T6 - O2

15 µV/mm

Fp1 - F3

15 µV/mm

F3 - C3

15 µV/mm

C3 - P3

15 µV/mm

P3 - O1

15 µV/mm

Fp2 - F4

15 µV/mm

F4 - C4

15 µV/mm

C4 - P4

15 µV/mm

P4 - O2

15 µV/mm

Fz - Cz

15 µV/mm

Cz - Pz

15 µV/mm

ECG

50 µV/mm

00:19:54

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 15 µV/mm, LFF: 1 Hz, HFF: 15 Hz, Notch: Off

xltek

Figure 14.3╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

Severity of Seizures as a Forensic Risk and Case Reports

215

14.6â•…University of Colorado Epilepsy Monitoring Unit Case The patient was a 63-year-old, right-handed male with seizures beginning at age 6 or 7. His seizures began with a vague, “strange feeling,” followed quickly by a sensation of nervousness and feeling of “wanting to get away.” This lasted 30 s and was followed by lethargy and trouble with expressive speech for several minutes. These events occurred multiple times a day despite maximal medical management, and very rarely would be followed by a secondarily generalized tonic–clonic seizure lasting 1 to 2 min. His last convulsive seizure occurred more than 5 years ago, and he had no history of status epilepticus. He had failed multiple antiepileptic medications including phenobarbital, valproic acid, gabaÂ� pentin, lamotrigine, topiramate, zonisamide, and levetiracetam, and was currently being managed by phenytoin, 300 mg twice daily, and carbamazepine, 600 mg twice daily, by a community neurologist. His past medical history was signiἀcant for complex partial seizures, depression (fluoxetine, 20 mg nightly), hypertension (atenolol, 50 mg nightly), and a history of head trauma before 2 years of age during which he was unconscious for more than a week. He was the product of an uncomplicated pregnancy and birth and there was no family history of epilepsy. Physical examination was remarkable for blood pressure of 147/77 and pulse 62, II/VI systolic crescendo/decrescendo cardiac murmur at the right upper sternal border, and brisk but symmetric lower extremity reflexes with flexor plantar responses. Routine EEG revealed left temporal sharp waves. MRI scans were performed at an outside hospital and not available for review. Fp1 - F7

7 µV/mm

F7 - T3

7 µV/mm

T3 - T5

7 µV/mm

T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm

F8 - T4

7 µV/mm

T4 - T6

7 µV/mm

T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm

F3 - C3

7 µV/mm

C3 - P3

7 µV/mm

P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm

F4 - C4

7 µV/mm

C4 - P4

7 µV/mm

P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm

Cz - Pz

7 µV/mm

ECG

50 µV/mm

00:20:44

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 7 µV/mm, LFF: 1 Hz, HFF: 15 Hz, Notch: Off

xltek

Figure 14.4╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

216 Sudden Death in Epilepsy: Forensic and Clinical Issues

He was admitted for epilepsy monitoring for potential ablative epilepsy surgery. Phenytoin and carbamazepine doses were halved on day 2. Phenytoin was discontinued on day 3, and carbamazepine was discontinued on day 4. During the latter half of day 4, the patient had two of his typical complex partial events with the second one generalizing into a mild convulsion. These events evolved out of the left temporal lobe and, of note, the patient’s postictal pulse following his convulsion was 140 beats/min. In the early morning of day 5, he had a third and ἀnal event. The seizure began typically, evolving from the left temporal region, but clinically the patient was sleeping in a prone position. His seizure generalized after 45 s, but the patient remained prone, his face hidden in his pillow. Audio and visual conἀrmation of progressively labored postictal breathing with good chest expansion was evident after the gentle convulsion ended 1 min later. One-and-a-half minutes after this, his EEG attenuated to essentially a flat baseline with only pulse artifact appreciated. Less than 30 s later, audio and visual evidence of breathing stopped as the EKG rhythm strip also terminated in electrical silence (Figures 14.1 through 14.8). The cardiac rhythm strip is particularly insightful in this tragic case, and suggests that previous cases may beneἀt from reevaluation by a multidisciplinary team. The EKG following the cessation of the convulsive seizure begins to suggest a peaked T wave (Figure 14.4 compared with Figure 14.1). In Figure 14.5, junctional escape beats are appreciated as the T wave continues to exhibit a more peaked appearance. By Figure 14.7, signiἀcant ST elevation is appreciated as the cardiac rhythm begins to slow signiἀcantly and, by Figure 14.8, cardiac electrical activity is silent. In a review of these series of rhythm strips with Fp1 - F7

7 µV/mm

F7 - T3

7 µV/mm

T3 - T5

7 µV/mm

T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm

F8 - T4

7 µV/mm

T4 - T6

7 µV/mm

T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm

F3 - C3

7 µV/mm

C3 - P3

7 µV/mm

P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm

F4 - C4

7 µV/mm

C4 - P4

7 µV/mm

P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm

Cz - Pz

7 µV/mm

ECG

50 µV/mm 00:21:26

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 7 µV/mm, LFF: 1 Hz, HFF: 15 Hz, Notch: Off

xltek

Figure 14.5╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

Severity of Seizures as a Forensic Risk and Case Reports

217

a cardiologist, the peaking of T waves followed by junctional escape beats suggests acidemic hyperkalemia (C. Long, personal communication, April 21, 2009). Not having the beneἀt of the corresponding video that clearly demonstrates the patient as being postictal with labored, facedown breathing, Long suggested the EKG showed the patient becoming acidemic, possibly by a rebreathing mechanism. Acute potassium shifts associated with respiratory acidosis are predicted (Perez et al. 1981), but whether levels associated with rebreathing are sufficient to produce cardiac disturbance are less certain (S. Linus, personal communication, May 4, 2009). Linus reviewed the postictal EKG as well, but as a nephrologist, he was much less impressed with the EKG being explained by hyperkalemia, claiming that despite the minimal peaking of the T wave, the QRS complex remained narrow throughout. Montague et al. (2008) retrospectively reviewed the frequency of EKG changes in hyperkalemia. Despite subjective corroboration of peaking of the T wave with quantitative amplitude measurements, no diagnostic threshold could be established. In fact, the EKG had such poor sensitivity and speciἀcity for hyperkalemia, they recommended against EKGs in guiding treatment of hyperkalemia in stable patients. The cardiologic and nephrologic interpretation of this rhythm strip remains unresolved, but suggests a different and broader approach to the problem of SUDEP. Could a combination of convulsion-related lactic acidosis plus respiratory acidosis from prone rebreathing be sufficient to cause a potassium-related fatal arrhythmia? Much like the ambulatory and intracranial EEG cases, there was no evidence of slow waves to suggest that hypotension or cerebral anoxia were the cause of death. Additionally, the EEG Fp1 - F7

7 µV/mm

F7 - T3

7 µV/mm

T3 - T5

7 µV/mm

T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm

F8 - T4

7 µV/mm

T4 - T6

7 µV/mm

T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm

F3 - C3

7 µV/mm

C3 - P3

7 µV/mm

P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm

F4 - C4

7 µV/mm

C4 - P4

7 µV/mm

P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm

Cz - Pz

7 µV/mm

ECG

50 µV/mm 00:22:04

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 7 µV/mm, LFF: 1 Hz, HFF: 15 Hz, Notch: Off

xltek

Figure 14.6╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

218 Sudden Death in Epilepsy: Forensic and Clinical Issues

Fp1 - F7

7 µV/mm

F7 - T3

7 µV/mm

T3 - T5

7 µV/mm

T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm

F8 - T4

7 µV/mm

T4 - T6

7 µV/mm

T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm

F3 - C3

7 µV/mm

C3 - P3

7 µV/mm

P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm

F4 - C4

7 µV/mm

C4 - P4

7 µV/mm

P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm

Cz - Pz

7 µV/mm

ECG

50 µV/mm 00:24:49

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 7 µV/mm, LFF: 1 Hz, HFF: 15 Hz, Notch: Off

xltek

Figure 14.7╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

had become silent a full 4 min before the last EKG activity, possibly supporting McLean’s concept of “cerebral electrical shutdown.” Could hyperkalemia or respiratory acidosis contribute to this phenomenon? Or are these patient’s EKG changes wholly unrelated to the underlying mechanism of SUDEP? More questions than answers remain.

14.7â•…Conclusion Seizure severity appears to be a forensic risk for the development of SUDEP. Seizures that are convulsive, frequent, and of long duration appear to increase the risk of SUDEP to almost 1:100 per year (Tellez-Zenteno et al. 2005), but not all SUDEP deaths are preceded by convulsions. Instead complex partial seizures and recovery from complex partial seizures with death a short time later have been reported (Langan et al. 2005). The severity of a speciἀc convulsion is also not particularly helpful, as witnessed in the University of Colorado case. The patient began convulsing in a prone position and essentially never moved from the spot. The convulsion was not of an unusual duration or violent motor behavior, but appeared to be position dependent. The nexus of coroners’ (Leestma 1990) and witnessed cases of SUDEP seems to suggest that a signiἀcant number of deaths are associated with prone positioning, which really has nothing at all to do with seizure severity. As increased interest and research in SUDEP further clariἀes the pathophysiologic explanations of sudden death in epilepsy, our epidemiological concept of seizure severity

Severity of Seizures as a Forensic Risk and Case Reports

219

7 µV/mm

Fp1 - F7

7 µV/mm

F7 - T3

7 µV/mm

T3 - T5 T5 - O1

7 µV/mm

Fp2 - F8

7 µV/mm

F8 - T4

7 µV/mm

T4 - T6

7 µV/mm

T6 - O2

7 µV/mm

Fp1 - F3

7 µV/mm

F3 - C3

7 µV/mm

C3 - P3

7 µV/mm

P3 - O1

7 µV/mm

Fp2 - F4

7 µV/mm

F4 - C4

7 µV/mm

C4 - P4

7 µV/mm

P4 - O2

7 µV/mm

Fz - Cz

7 µV/mm

Cz - Pz

7 µV/mm

ECG

50 µV/mm

00:28:45

10.5 s/page, 1 s/div, 21.6 mm/s Gain: 7 µV/mm, LFF: 1 Hz, HFF: 15 Hz, Notch: Off

xltek

Figure 14.8╇ The time stamps on each figure show the progression of time. EKG rhythm strips illustrate the lengthening between complexes and the peaking of T waves.

may likely play a less signiἀcant role in our approach to SUDEP understanding, education, and avoidance.

References Bird, J. M., K. A. T. Dembny, D. Sandeman, and S. Butler. 1997. Sudden unexplained death in epilepsy: An intracranially monitored case. Epilepsia 38 (s11): 52–56. Birnbach, C. D., A. J. Wilensky, and C. B. Dodril. 1991. Predictors of early mortality and sudden death in epilepsy: A multidisciplinary approach. J Epilepsy 4: 11–17. Dasheiff, R. M. 1991. Sudden unexpected death in epilepsy and its relationship to sudden cardiac death. J Clin Neurophysiol 8: 216–222. Dasheiff, R. M., and L. J. Dickinson. 1986. Sudden unexpected death following a seizure in an epileptic patient: A case report. Arch Neurol 43: 194–196. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33: 310–316. Epilepsy Foundation. 2009. http://www.epilepsyfoundation.org/epilepsyusa/news/Travolta.cfm (acÂ�Â� cessed February 8, 2009). Fisher, R. S., B. G. Vickrey, P. Gibson et al. 2000. The impact of epilepsy from the patient’s perspective I. Descriptions and subjective perceptions. Epilepsy Res 41 (1): 39–51. Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67: 439–444. Langan, Y., and L. Nashef. 2003. Sudden unexpected death in epilepsy (SUDEP). ACNR 2 (6): 6–8.

220 Sudden Death in Epilepsy: Forensic and Clinical Issues Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Lee, M. A. 1998. EEG Video recording of sudden unexpected death in epilepsy (SUDEP). Epilepsia 39 (s6): 123–124. Leestma, J. E. 1990. Sudden unexpected death associated with seizures: A pathological review. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 61–88. New York, NY: Marcel Dekker. Leestma, J. E., J. F. Annegers, M. J. Brodie et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. McLean, B. N., and S. Wimalaratna. 2007. Sudden death in epilepsy recorded in ambulatory EEG. J Neurol Neurosurg Psychiatry 78: 1395–1397. Montague, B. T., J. R. Ouellette, and G. K. Buller. 2008. Retrospective review of the frequency of ECG changes in hyperkalemia. Clin J Am Soc Nephrol 3(2): 324–330. Nashef, L., D. R. Fish, S. Garner, J. W. Sander, and S. D. Shorvon. 1995. Sudden death in epilepsy: A study of incidence in a young cohort with epilepsy and learning difficulty. Epilepsia 36: 1187–1194. Neidemeyer, E., and F. L. Da Silva. 1987. Electroencephalography, Basic Principles, Clinical Applications and Related Fields, 2nd ed., 385. Baltimore, MD: Urban and Schanarzenberg. Nilsson, L., B. Y. Farahmand, P. G. Persson et al. 1999. Risk factors for sudden unexpected death in epilepsy: A case-controlled study. Lancet 13: 888–893. Nowack, W. J. 2006. Psychiatric disorders associated with epilepsy. http://www.emedicine.medscape╉ .com/article/1186336 (accessed May 12, 2009). Opeskin, K., and S. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12: 456–464. Perez, G. O., J. R. Oster, and C. A. Vaamonde. 1981. Serum potassium concentration in acidemic states. Nephron 27 (4–5): 233–243. Phillips, L. A. 2008. Death in epilepsy monitoring unit raises questions about safety policies and practice standards. Neurology Today 8 (16): 1–15. Purves, S. J., M. Wilson-Young, and V. P. Sweeney. 1992. Sudden death in epilepsy: Single case report with video-EEG documentation. Epilepsia 33 (Sl3): 123. Racoosin, J. A., J. Feeney, G. Burkhart et al. 2001. Mortality in antiepileptic drug development programs. Neurology 56: 514–519. Sander, J. W., Y. M. Hart, A. L. Johnson, and S. D. Shorvon. 1990. National General Practice Study of Epilepsy: Newly diagnosed epileptic seizures in a general population. Lancet 336: 1267–1271. Sperling, M. R. 2001. Sudden unexplained death in epilepsy. Epilepsy Curr 1 (1): 21–23. Sperling, M. R. 2004. The consequences of uncontrolled epilepsy. CNS Spectr 9: 98–101, 106–109. Sperling, M. R., H. Feldman, J. Kinman et al. 1999. Seizure control and mortality in epilepsy. Ann Neurol 46: 45–50. Spitz, M. C. 1998. Injuries and death as a consequence of seizures in people with epilepsy. Epilepsia 39: 904–907. Spitz, M. C., J. A. Towbin, D. Shantz, and L. E. Adler. 1994. Risk factors for burns as a consequence of seizures in people with epilepsy. Epilepsia 35: 764–767. Stollberger, C., and J. Finsterer. 2004. Cardiorespiratory ἀndings in sudden unexplained/unexpected death in epilepsy (SUDEP). Epilepsy Res 59: 51–60. Tellez-Zenteno. J. F., L. H. Ronquillo, and S. Weibe. 2005. Sudden unexpected death in epilepsy: Evidence-based analysis of incidence and risk factors. Epilepsy Res 65: 101–115. Tennis, P., T. B. Cole, J. F. Annegers et al. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36: 29–36. Walczak, T. S., I. E. Leppik, M. D’Amelio et al. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56: 519–525.

Intractable Epilepsy in the Setting of Malformations of Cortical Development as a Mechanism for SUDEP

15

Lara Jehi Imad Najm

Contents 15.1 Introduction 15.2 SUDEP: Epidemiology and Risk Factors 15.3 Proposed Mechanisms of SUDEP 15.3.1 Pulmonary Pathophysiology 15.3.1.1 Clinical Evidence 15.3.1.2 Experimental Evidence 15.3.2 Cardiac Pathophysiology 15.3.2.1 Clinical Evidence 15.3.2.2 Experimental Evidence 15.4 Central Autonomic and Respiratory Control 15.5 Malformations of Cortical Development and SUDEP 15.5.1 Classiἀcation of MCD 15.5.2 Neuroimaging of MCD 15.5.3 Relevance of MCD Classiἀcation and Imaging to SUDEP 15.6 Localization of MCD and SUDEP 15.6.1 Case Reports 15.6.2 Cleveland Clinic Epilepsy Center Experience 15.7 Mechanisms of Epileptogenicity in MCD and SUDEP 15.7.1 Localized Disruption of Excitatory and Inhibitory Neurotransmission 15.7.1.1 Experimental Evidence 15.7.1.2 Clinical Evidence 15.7.1.3 Relevance to SUDEP 15.7.2 Diffuse Disruption of Normal Neural Circuitry 15.7.2.1 Experimental Evidence 15.7.2.2 Clinical Evidence 15.7.2.3 Relevance to SUDEP 15.8 Conclusion References

221

222 222 222 222 223 223 223 223 224 225 226 226 227 227 228 228 229 229 229 229 230 230 230 230 231 231 231 232

222 Sudden Death in Epilepsy: Forensic and Clinical Issues

15.1â•…Introduction Sudden unexpected death in epilepsy (SUDEP) is currently accepted as the most important epilepsy-related mode of death, and is the leading cause of death in chronic uncontrolled epilepsy (Jehi and Najm 2008; Tomson 2000). Despite signiἀcantly increased interest in SUDEP over the past few decades, the exact mechanisms leading to its occurrence remain unknown (Nashef et al. 2007; Pedley and Hauser 2002; Schraeder et al. 2009; Tomson et al. 2008a). Furthermore, despite higher rates of SUDEP observed in patients with structural brain abnormalities (Monte et al. 2007), little is known about how, or even if, speciἀc epilepsy etiologies and brain pathologies interact with other potential triggers leading up to sudden death in a given epilepsy patient. Speciἀcally, the role played by malformations of cortical development (MCD), a major cause of intractable epilepsy, remains unknown. This chapter will ἀrst briefly review general concepts related to SUDEP and its proposed mechanisms, outline basic concepts pertaining to intractable epilepsy in MCD, and then focus on how those two topics—intractable epilepsy in MCD and SUDEP—may be related. The discussion will be based on a review of SUDEP occurrences among a cohort of patients evaluated at Cleveland Clinic Epilepsy Center over a 15-year period, and on a review of the currently available literature.

15.2â•… SUDEP: Epidemiology and Risk Factors SUDEP is most often deἀned as the sudden, unexpected, witnessed or unwitnessed, nontraumatic, and nondrowning death of patients with epilepsy with or without evidence of a seizure, excluding documented status epilepticus, and in whom postmortem examination does not reveal a structural or toxicological cause for death (Nashef et al. 2007). Estimates of its incidence range from 0.7 to 1.3 cases per 1000 patient years in large cohorts of patients with epilepsy (Nilsson et al. 1997; Tennis et al. 1995), and from 3.5 to 9.3 cases per 1000 patient years in anticonvulsant drug registries, medical device registries, and epilepsy surgery programs (Leestma et al. 1997; Nashef et al. 1995; Tomson et al. 2008b). Several potential risk factors for SUDEP have been investigated with conflicting ἀndings (Jehi and Najm 2008). Consistently identiἀed risk factors include young age, early onset of seizures, refractoriness of epilepsy, the presence of generalized tonic–clonic seizures, male sex, and being in bed at the time of death (Langan et al. 2005; Monte et al. 2007; Nashef et al. 2007; Tomson et al. 2008b). Weaker risk factors include being in the prone position at the time of death, having one or more subtherapeutic blood levels of anticonvulsant medication, having a structural brain lesion, and being asleep (Monte et al. 2007; Tomson et al. 2008b). At any rate, the current consensus is that SUDEP is primarily a seizure-related occurrence, with patients having poorly controlled epilepsy and frequent generalized tonic–clonic seizures being particularly vulnerable (Jehi and Najm 2008; Tomson et al. 2008b) (see Table 15.1).

15.3â•… Proposed Mechanisms of SUDEP 15.3.1â•… Pulmonary Pathophysiology The two major proposed respiratory mechanisms of SUDEP are central apnea and acute neurogenic pulmonary edema.

Intractable Epilepsy in the Setting of MCD as a Mechanism for SUDEP

223

Table 15.1â•… Summary of the General Mechanisms Thought to Contribute to SUDEP Respiratory mechanisms â•… Central apnea â•… Pulmonary edema Cardiac mechanisms â•… Arrhythmia â•… Asystole Effects of long-standing seizure disorder â•… Altered autonomic function â•… Structural heart change

15.3.1.1â•…Clinical Evidence In a prospective study of epilepsy patients undergoing a video-EEG evaluation, central apnea lasting at least 10 s was observed postictally in 40% of the recorded seizures (Nashef et al. 1996); otherwise healthy young epilepsy patients have been reported to develop central apnea immediately following complex partial seizures (Blum et al. 2000; Jehi and Najm 2008). Apnea might also represent the only ictal symptom of temporal lobe seizures, especially in children (Lee et al. 1999; Singh et al. 1993). Pulmonary edema is frequently found in SUDEP patients at autopsy (Terrence et al. 1981) and is known to occur in other neurological disorders affecting the central neurorespiratory control centers such as head trauma and subarachnoid hemorrhage. 15.3.1.2â•…Experimental Evidence In a sheep animal model of SUDEP, one third of animals died from hypoventilation and had associated pulmonary edema at autopsy (Johnston et al. 1995). DBA/2 mice are another proposed SUDEP model because they exhibit respiratory arrest after audiogenic seizures (Tupal and Faingold 2006), and where sudden death was preventable by oxygenation without any change in seizure severity (Venit et al. 2004). 15.3.2â•…Cardiac Pathophysiology The most signiἀcant and widely discussed cardiac mechanism of SUDEP is cardiac arrhythmia precipitated by seizure discharges acting via the autonomic nervous system (Jehi and Najm 2008; Nei et al. 2004; Tomson et al. 2008b). 15.3.2.1â•…Clinical Evidence A wide spectrum of cardiac arrhythmias, such as ictal asystole, atrial ἀbrillation, repolarization abnormalities, and bundle branch blocks, has been reported during seizures (Blumhardt et al. 1986; Galimberti et al. 1996; Leung et al. 2006; Nei et al. 2004; Opherk et al. 2002). Ictal cardiac arrhythmias occurred in 42% of hospitalized epilepsy patients in one study, with the most common being an irregular series of abrupt rate changes toward the end of the electroencephalographic (EEG) seizure discharge (Blumhardt et al. 1986). In another study, analysis of R–R intervals during the ἀrst 10-s period of EEG discharge showed a signiἀcant early heart rate increase in 49% of seizures and an early heart rate

224 Sudden Death in Epilepsy: Forensic and Clinical Issues

reduction in 25.5% (Galimberti et al. 1996) (Figures 15.1 and 15.2). Certain clinical seizure characteristics have been correlated with the occurrence of ictal electrocardiographic (ECG) abnormalities. While one study found that mean seizure duration was longer in patients with ECG abnormalities than in those without such changes (Nei et al. 2000), others observed that ictal ECG abnormalities occurred more often and were more severe in generalized tonic–clonic seizures relative to complex partial seizures (Nei et al. 2000, 2004; Opherk et al. 2002). Those same clinical seizure characteristics were correlated with a higher risk of SUDEP (Langan et al. 2005), suggesting an interrelation between seizure semiology, ECG abnormalities, and SUDEP. 15.3.2.2â•…Experimental Evidence Electrical brain stimulation of the limbic system and insular cortex has repeatedly been shown to provoke heart rate changes, including bradycardia, tachycardia, and asystole (Leung et al. 2006). Some studies have even suggested a lateralized influence of the insulae on cardiovascular autonomic control with intraoperative stimulation of the left posterior insula eliciting a cardioinhibitory response and hypotension, and stimulation of the right anterior insula eliciting tachycardia and hypertension (Jehi and Najm 2008). Other studies have suggested a localization-related influence of the limbic system on cardiovascular responses with stimulation of the amygdala alone being insufficient to produce the ictal

Fp1–F7

50 µV

F7–SP1 SP1–T7 T7–P7 P7–O1 Fp2–F8 F8–SP2 SP2–T8 T8–P8 P8–O2 SP1–SP2 TP9–TP10 Fp1–F3 F3–C3 C3–P3 P3–O1 Fp2–F4 F4–C4 C4–P4 P4–O2 EKG1–EKG2 50 µV

Figure 15.1╇ Asystole documented during a left hemispheric seizure (10-s epoch) in a patient with a normal MRI, and subsequently pathologically proven malformation of cortical development. The arrow identifies movement artifact secondary to patient fall.

Intractable Epilepsy in the Setting of MCD as a Mechanism for SUDEP

225

EEG EMG (masseter) EMG (biceps) Pupils Electrodermogram C. H. R. B.

R. H. B. C. Time (min)

Interictal

Preictal

Ictal

Enuresis Extinction Postictal

Aura

Ictal

Postictal

Recuperation

Interictal

Figure 15.2╇ Autonomic changes during a motor seizure. Clear increases in the heart rate

(H) and blood pressure (B) correspond to a suppression in the respiratory rate (R) during the ictal phase. Autonomic changes continue well after the ictal event is finished. C, cystogram. (Reprinted From Gastaut, H., and Broughton, R., Epileptic Seizures: Electrographic Features, Diagnosis, and Treatment, Springfield, IL: Charles C Thomas Publisher, 1972: 28. With permission.)

tachycardia so commonly seen in epileptic seizures, suggesting that cortical involvement is essential for the increase in HR (Keilson et al. 1987). Such cortical stimulation-induced HR changes may explain how massive seizure-related discharges can affect the cardiac rhythm during the seizure itself. There is, however, also evidence of a baseline epilepsy-related autonomic dysfunction. A recent study showed a pronounced reduction in cardiac metaiodobenzylguanidine (MIBG) uptake in patients who had ictal asystole compared to other epileptic and nonepileptic controls, suggesting a postganglionic cardiac catecholamine disturbance. The authors then propose that epilepsy-related impaired sympathetic cardiac innervation limits adjustment and heart rate modulation and may thus increase the risk of asystole and, ultimately, SUDEP (Kerling et al. 2008).

15.4â•…Central Autonomic and Respiratory Control A tight, interconnected network exists throughout the neuraxis to control various elements of the cardiovascular autonomic system. A solid understanding of this network provides

226 Sudden Death in Epilepsy: Forensic and Clinical Issues

useful insights for consideration of a cardiac pathophysiology of SUDEP (Jehi and Najm 2008). Key components of the central cortical control of autonomic functions include the insula, the anterior cingulate gyrus, and the ventromedial prefrontal cortex. The insula represents the primary viscerosensory cortex, while the cingulate gyrus and prefrontal cortices form the premotor autonomic region. At the subcortical level, the hypothalamus provides the interface with the endocrine stimuli and triggers corresponding autonomic responses to maintain homeostasis. The amygdala, an integral component of the limbic system linking the cortical and subcortical centers, mediates the autonomic response to emotions. In addition to playing a key role in autonomic control, the insula, amygdala, cingulate gyrus, and prefrontal cortex also represent the most common foci of partial epilepsy, a concept that we will elaborate on further later and that may explain the frequent observation of autonomic changes in relation to epileptic seizures (Leung et al. 2006). Although central apnea has been observed with focal epileptiform activity alone (Lee et al. 1999), a more accepted hypothesis is that neurotransmitters mediating the brain’s own seizure-terminating mechanism could also be inhibiting respiratory centers in the brainstem and causing postictal apnea (Jehi and Najm 2008). Understanding the concepts of central autonomic pathways and respiratory control will facilitate the discussion of possible mechanisms of SUDEP in MCD. In the subsequent sections of this chapter, we will discuss how the general concepts of SUDEP discussed so far apply speciἀcally to intractable epilepsy in MCD patients.

15.5â•…Malformations of Cortical Development and SUDEP Alzheimer and Rahcke recognized the presence of aberrant cortical lamination in autopsies of patients with a history of chronic epilepsy almost a century ago. It was, however, the detailed report published by Taylor et al. (1971) that has since raised awareness of the role of misshapen dysmorphic (dysplastic) neurons in the setting of cortical architectural disorganization (both columnar and laminar) as the pathological substrate in some patients with drug-resistant epilepsy. The report identiἀed the possible role of disorientated and giant neurons (and balloon cells with eccentric nuclei) in temporal cortex resected from patients with temporal lobe epilepsy. In the past 20 years, studies showed that MCDs include a broad spectrum of architectural anomalies, such as cortical laminar disorganization, neuronal heterotopia in the subcortical white matter, the persistence of neurons in the superἀcial cortical layer (layer I of the neocortex), clustering of neurons in the gray matter, nodular heterotopia, and the presence of aberrant neurons such as giant neurons and balloon cells (Fauser et al. 2006; Lawson et al. 2005; Palmini et al. 2004; Tassi et al. 2002). The relationship between epilepsy and MCD in general, and focal cortical dysplasia (FCD) in particular, has been well established. In fact, 8–12% of cases with intractable epilepsy are attributed to MCD, whereas 14–26% of surgically treated cases have MCD (Tassi et al. 2002). 15.5.1â•…Classification of MCD MCDs are due to abnormalities in neuronal migration, proliferation, and/or differentiation that result in four distinct pathological subtypes: IA, IB, IIA, and IIB (Widdess-Walsh et al. 2005). Those various subtypes have different microscopic and imaging characteristics, as well as distinct outcomes with epilepsy surgery. Type I is characterized by a lack of

Intractable Epilepsy in the Setting of MCD as a Mechanism for SUDEP

227

dysmorphic neurons or balloon cells (abnormal cellular elements with a thin membrane; pale, glassy, and eosinophilic cytoplasm, eccentric nucleus, usually of increased size compared with gemistocytic astrocytes).. Type IA is characterized by patchy and isolated architectural abnormalities, mainly consisting of dyslamination and columnar disorganization, accompanied, or not, by other abnormalities of mild MCD. Type IB is characterized by architectural disorganization intermixed with giant or immature, but not dysmorphic, neurons. Most of these mild abnormalities defy in vivo imaging recognition. The only available evidence is from patients undergoing epilepsy surgery in whom such mild abnormalities were the only histopathological ἀnding. This suggests that at least some patients with type IA/B FCD can have medically refractory epilepsy (Palmini et al. 2004). Type II or Taylor-type FCD show dysmorphic neurons without (IIA) or with balloon cells (IIB) in the setting of diffuse cortical architectural disorganization. These are the focal lesions most commonly identiἀed on MRI. However, studies showed various degrees of imaging abnormalities in patients with Taylor-type FCD ranging from thickening of the cortical mantle to focal and severe fluid-attenuated inversion recovery (FLAIR) signal abnormalities (Palmini et al. 2004). Figure 15.3 illustrates the various histopathological ἀndings of those four MCD subtypes. 15.5.2â•…Neuroimaging of MCD MRI can be either normal (most often in type I), despite the use of high-resolution techniques, or may demonstrate one or more of the following characteristics: (1) focal areas of increased cortical thickness; (2) blurring of the cortex (gray)/white matter junction; (3)€ increased signal on T2-weighted proton density or FLAIR sequences (more likely to occur in balloon cell-containing lesions); and (4) extension of cortical tissue with increased FLAIR signal from the surface to the periventricular region (transmantle dysplasia) (Palmini et al. 2004; Ruggieri et al. 2004; Tassi et al. 2002; Widdess-Walsh et al. 2005). 15.5.3â•… Relevance of MCD Classification and Imaging to SUDEP Several studies have shown that MRI and histopathological ἀndings in MCD correlate with seizure outcome following resective surgery for intractable epilepsy, with better outcomes

Normal

Type I

Type IIA

Type IIB

Figure 15.3╇ Cresyl echt violet staining of sections representing normal neocortex and type I,

type IIA, and type IIB malformations of cortical development; scale bar, 100 µm. (Palmini, A. et al., Neurology, 62 (6 Suppl 3), S2–S8, 2004. With permission.)

228 Sudden Death in Epilepsy: Forensic and Clinical Issues

observed in patients with clear MRI lesions, and more severe histopathological changes (mainly type IIB), as opposed to those with normal MRI (types IA and IB) (Jeha et al. 2007; Tassi et al. 2002; Widdess-Walsh et al. 2005). The poorer outcome following surgery in nonlesional MCD cases is mainly attributed to difficulties localizing the epileptogenic focus and/or its incomplete resection due to more diffuse cytoarchitectural changes that may be invisible on MRI (Jeha et al. 2007). This is relevant in a discussion about SUDEP because persistent seizures are the main risk factor for SUDEP, as extensively discussed earlier in this chapter (Jehi and Najm 2008; Tomson 2000; Tomson et al. 2008b). In fact, it has been shown in a recent prospective study that persistent seizures speciἀcally following failed epilepsy surgery carry a signiἀcantly high risk of sudden death, estimated at 6.3 cases per 1000 patient years (Sperling et al. 2005). As such, patients with intractable epilepsy, MCD, and normal imaging may be particularly more vulnerable to SUDEP.

15.6â•…Localization of MCD and SUDEP Several studies showed a predisposition of speciἀc MCD pathological subtypes to localize to certain brain regions. In a review of 145 cases of MCD operated on for intractable epilepsy at Cleveland Clinic Epilepsy Center between 1990 and 2002, we found that pathological subtypes IIA and IIB were predominantly frontal in location and had a more severe epilepsy syndrome than patients with subtypes IA and IB. Patients with subtype IA FCD had less severe, later onset epilepsy that was predominantly located in the temporal lobe (Widdess-Walsh et al. 2005). This is similar to ἀndings of another review of 52 MCD surgical cases of intractable epilepsy where patients with architectural dysplasia alone (type IA) had lower seizure frequency than those with Taylor-type dysplasia, and the epileptogenic zone was mainly in the temporal lobe, while in patients with Taylor-type dysplasia, the epileptogenic zone was mainly extratemporal, predominantly frontal (Tassi et al. 2002). Beyond these anatomical distributions of MCD in relation to histopathology, it remains€to be shown whether MCDs have a speciἀc affinity to localize in brain regions particularly relevant in the proposed mechanisms of SUDEP, such as the insula, cingulate gyrus, orbitoÂ�frontal regions, or amygdalae. We will review now the limited available data concerning this point. 15.6.1â•…Case Reports A few case reports have indeed documented ictal cardiac or respiratory changes in patients with MCD involving brain regions thought to be part of the central autonomic control centers. Known examples include one case report of a three-and-a-half-year-old child with episodic sinus bradycardia during habitual seizures and prolonged interictal discharges due to FCD in the anterior two-thirds of the insula and the inferior frontal cortex (Seeck et al. 2003). In another case report of a 30-year-old man who was found dead on arrival to the hospital following an hour-and-a-half of complex partial seizure, post mortem examination showed bilateral occipital frontal polymicrogyria in the brain and chronic interstitial and perivascular ἀbrosis in the heart without previous vascular risk factors, a ἀnding which the authors attributed to possibly chronic repetition of seizures (Ribacoba Montero et al. 2002).

Intractable Epilepsy in the Setting of MCD as a Mechanism for SUDEP

229

15.6.2â•…Cleveland Clinic Epilepsy Center Experience We recently reviewed 3841 patients evaluated with prolonged video-EEG monitoring in our epilepsy monitoring unit between 1990 and 2005 and identiἀed 301 mortalities (Jehi et al., in preparation). Of 223 mortalities evaluated so far, 29 died of SUDEP, including six who had MCD. Of these six cases, four had dual pathology with associated hippocampal sclerosis affecting their mesial temporal structures (three with type IA MCD, including two in their temporal lobe, and another in his lateral temporal and inferior frontal lobes; and one with MRI-visible dysplasia affecting his posterior temporal and angular gyrus), one had posterior left temporal lobe closed lip schizencephaly, and one had a temporoparietal ictal onset zone on invasive EEG with normal imaging. Five of the above patients had epilepsy surgery (some even up to three times), but continued to have seizures postoperatively. The epileptogenic zone thus localized to the peri-insular region in all of our MCD cases with SUDEP suggests a direct anatomical disruption of baseline physiologic autonomic centers in the insula as a possible mechanism of SUDEP. Although the observation is clear, a direct relationship between the suspected anatomical location of the FCD and epileptogenicity in the insular region and SUDEP is difficult to prove. An alternative hypothesis in this group may be that the increased risk of SUDEP is simply due to continuous seizures following a failed epilepsy surgery.

15.7â•…Mechanisms of Epileptogenicity in MCD and SUDEP Multiple mechanisms of epileptogenicity in MCD have been proposed, mainly revolving around two central concepts: (1) localized disruption of excitatory and inhibitory neurotransmission and (2) diffuse disruption of normal neural circuitry. We will now elaborate on those two general principles and discuss how they might be related to SUDEP. 15.7.1â•…Localized Disruption of Excitatory and Inhibitory Neurotransmission Dysplastic lesions have a high degree of intrinsic epileptogenicity: In one study, up to 67% of patients with MCD manifested continuous or frequent rhythmic epileptogenic discharges recorded directly from their cortical dysplastic lesions during intraoperative electrocorticography, as opposed to only 2.5% of patients with intractable partial epilepsy associated with other types of structural lesions (Palmini et al. 1995). Data collected through immunocytochemical and clinical studies support an increase in excitatory amino acid neurotransmission and an overall decrease in intralesional and perilesional inhibition as a signiἀcant contributor to this high degree of intrinsic epileptogenicity (Avoli et al. 1999; Najm et al. 2004; Palmini et al. 1995, 2004). 15.7.1.1â•…Experimental Evidence Ferrer et al. (1992) ἀrst showed abnormalities in the morphology and distribution of localcircuit (inhibitory) neurons in FCD, and hypothesized they may have a pivotal role in the appearance and prolongation of electrical discharges. Since then, Najm et al. (2000) correlated signiἀcantly higher immunoreactivity of the excitatory N-methyl-d-aspartate (NMDA) receptor (NR) 2A/B in both the dysplastic somata and all their dendritic processes

230 Sudden Death in Epilepsy: Forensic and Clinical Issues

with in vivo epileptic activity recorded through subdural EEG, whereas White et al. (2001) observed reduced levels of the inhibitory gamma aminobutyric acid A (GABA-A) receptors alpha1 and alpha2 mRNA in both dysplastic neurons and giant cells compared to control neurons. Neurotransmission changes in the dysplastic cortex extend, however, beyond the classical excitatory and inhibitory NMDA and GABA systems. Trottier et al. (1996) found evidence of serotonergic hyperinnervation and altered patterns of the catecholaminergic innervation in dysplastic cortex of epilepsy patients with MCD, as opposed to tissue from patients with cryptogenic neocortical epilepsy and normal controls (Trottier et al. 1996). 15.7.1.2â•…Clinical Evidence Most patients diagnosed by imaging studies as having lesions identiἀed as type IIA/B FCD have medically intractable partial epilepsy, with frequently disabling motor and secondary generalized seizures (Fauser et al. 2006; Lawson et al. 2005; Palmini et al. 2004; Tassi et al. 2002; Widdess-Walsh et al. 2005). Many patients have a history of status epilepticus, including epilepsia partialis continua, and scalp EEG and acute electrocorticography often show continuous spiking or other highly epileptogenic patterns, attesting to some type of re-entrant excitatory circuitry unopposed by faulty inhibition (Avoli et al. 1999; Ferrer et al. 1992; Palmini et al. 1995). 15.7.1.3â•…Relevance to SUDEP It is reasonable to hypothesize that the previously discussed disturbances in excitatory versus inhibitory balance leading to epileptogenicity of dysplastic tissue may also translate to disturbances in the normal sympathetic versus parasympathetic balance if the MCD happens to involve cortical regions crucial for central autonomic control. There is however no information available currently on the status of either NMDA or GABA receptors in SUDEP victims. Data presented previously supporting a role for serotonin in the modulation of sudden death induced by audiogenic seizures in DBA/2 mice via causing respiratory arrest (Tupal and Faingold 2006) may be related to the altered serotonergic pathways described speciἀcally in MCD (Trottier et al. 1996), but further investigation is needed. 15.7.2â•…Diffuse Disruption of Normal Neural Circuitry This hypothesis is based on the concept that rather than intrinsic epileptogenicity of the dysplastic lesion itself, the main mechanism of epilepsy, in the context of MCD, is disruption of neuronal circuitry extending far beyond the lesion itself. Following this idea, focal epileptogenesis associated with MCD, as opposed with postnatally acquired lesions such as those due to tumors or trauma, is best conceptualized as a disorder of widespread and patchy disturbance of cortical networks. This developmental perspective implies that the epileptogenic region in MCD is rarely discrete, even in patients with focal dysplasia, and may include remote cortical or subcortical areas (Duchowny et al. 2000). 15.7.2.1â•…Experimental Evidence Subtle structural abnormalities are seen beyond the clearly visible dysplastic lesion in many patients with MCD. Measurements of cerebral surface area and volume reveal widespread and unusually extensive anatomic changes throughout extralesional gray and subcortical white matter in a high proportion of patients with FCD (Sisodiya et al. 1997). Similarly, prenatal damage during critical maturational stages of primates resulted in anomalous

Intractable Epilepsy in the Setting of MCD as a Mechanism for SUDEP

231

sulcation and reorganization at remote cortical sites in both cerebral hemispheres (Goldman 1978). Functional neuroimaging abnormalities on proton MR spectroscopy and flumazenil PET studies of benzodiazepine receptors in FCD reveal abnormal activation patterns extending far beyond the lesional boundary (Richardson et al. 1996). Ictal SPECT in tuberous sclerosis complex also shows increased blood flow in the perituberal penumbra, suggesting dysfunction beyond the tuber (Koh et al. 1999). 15.7.2.2â•…Clinical Evidence Neurological deἀcits in patients with MCD are often more extensive than expected based on the extent of the lesion visible on MRI. Besides epilepsy, patients with MCD often suffer from cognitive impairment, and are prone to attention and behavioral problems and autism. Verbal IQ scores are decreased even in patients with very focal lesions of the dominant cerebral hemisphere (Duchowny et al. 2000). These ἀndings may be attributed to dysfunction beyond the lesion itself. However, there are other factors to consider. Younger age of seizure onset and larger lesions are associated with diminished cognitive outcome. Epileptogenic activity and medication also make it difficult to separate cause from effect. However, not all patients with frequent seizures and chronic medications exhibit cognitive disturbance (Duchowny et al. 2000; Fauser et al. 2006; Lawson et al. 2005) and the extent of the MRI lesion itself does not always correlate with cognitive outcome suggesting an additional functional mechanism to the disturbances seen with MCD besides the structural disturbance visible on routine imaging. 15.7.2.3â•…Relevance to SUDEP Parallels can be drawn between the dysfunctional connectivity alluded to previously in MCD, and the disruptions in the sympathetic and parasympathetic networks felt to be contributing to SUDEP. This issue needs, however, to be further investigated.

15.8â•…Conclusion SUDEP is the leading cause of death in chronic uncontrolled epilepsy and is a devastating complication, usually occurring in young, otherwise healthy individuals with persistent seizures or poor compliance to antiepileptic medications. Despite signiἀcantly increased interest in SUDEP over the past few decades, the exact mechanisms leading to its occurrence remain unknown. On the other hand, MCD have been increasingly recognized as a common and very signiἀcant cause of epilepsy that still remains difficult to control with epilepsy surgery. As such, MCDs may provide a unique substrate for further study of SUDEP mechanisms and possible preventative interventions. A lot remains unknown though, and areas of further research may include: 1. Determination of the incidence of SUDEP in MCD speciἀcally, as compared to other epilepsy etiologies, to evaluate whether this patient population is particularly more vulnerable to sudden death. 2. Investigation of the autonomic function in patients with MCD as compared to those with other structural abnormalities and epilepsy to clarify whether strategically located MCD can lead to distant autonomic dysfunction, possibly contributing to cardiac arrhythmias hypothesized in SUDEP.

232 Sudden Death in Epilepsy: Forensic and Clinical Issues

3. Characterization of the immunocytochemical characteristics of brain tissue from patients with SUDEP and MCD, especially excitatory and inhibitory pathways, and serotonergic pathways to investigate further the potential for common mechanisms between intractable epilepsy in MCD and SUDEP.

References Avoli, M., A. Bernasconi, D. Mattia, A. Olivier, and G. G. Hwa. 1999. Epileptiform discharges in the human dysplastic neocortex: In vitro physiology and pharmacology. Ann Neurol 46 (6): 816–826. Blum, A. S., J. R. Ives, A. L. Goldberger, I. C. Al-Aweel, K. B. Krishnamurthy, F. W. Drislane, and D.€L.€Schomer. 2000. Oxygen desaturations triggered by partial seizures: Implications for cardiopulmonary instability in epilepsy. Epilepsia 41 (5): 536–541. Blumhardt, L. D., P. E. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1 (8489): 1051–1056. Duchowny, M., P. Jayakar, and B. Levin. 2000. Aberrant neural circuits in malformations of cortical development and focal epilepsy. Neurology 55 (3): 423–428. Fauser, S., H. J. Huppertz, T. Bast, K. Strobl, G. Pantazis, D. M. Altenmueller, B. Feil et al. 2006. Clinical characteristics in focal cortical dysplasia: A retrospective evaluation in a series of 120 patients. Brain 129 (Pt 7): 1907–1916. Ferrer, I., M. Pineda, M. Tallada, B. Oliver, A. Russi, L. Oller, R. Noboa, M. J. Zujar, and S. Alcantara. 1992. Abnormal local-circuit neurons in epilepsia partialis continua associated with focal cortical dysplasia. Acta Neuropathol 83 (6): 647–652. Galimberti, C. A., E. Marchioni, F. Barzizza, R. Manni, I. Sartori, and A. Tartara. 1996. Partial epileptic seizures of different origin variably affect cardiac rhythm. Epilepsia 37 (8): 742–747. Goldman, P. S. 1978. Neuronal plasticity in primate telencephalon: Anomalous projections induced by prenatal removal of frontal cortex. Science 202 (4369): 768–770. Jeha, L. E., I. Najm, W. Bingaman, D. Dinner, P. Widdess-Walsh, and H. Luders. 2007. Surgical outcome and prognostic factors of frontal lobe epilepsy surgery. Brain 130 (Pt 2): 574–584. Jehi, L., and I. M. Najm. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Cleve Clin J Med 75 (Suppl 2): S66–S70. Johnston S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37 (4): 531–537. Keilson, M. J., W. A. Hauser, J. P. Magrill, and M. Goldman. 1987. ECG abnormalities in patients with epilepsy. Neurology 37 (10): 1624–1626. Kerling, F., I. Blumcke, and H. Stefan. 2008. Pitfalls in diagnosing limbic encephalitis—A case report. Acta Neurol Scand 118 (5): 339–342. Koh, S., P. Jayakar, T. Resnick, L. Alvarez, R. E. Liit, and M. Duchowny. 1999. The localizing value of ictal SPECT in children with tuberous sclerosis complex and refractory partial epilepsy. Epileptic Disord 1 (1): 41–46. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64 (7): 1131–1133. Lawson, J. A., S. Birchansky, E. Pacheco, P. Jayakar, T. J. Resnick, P. Dean, and M. S. Duchowny. 2005. Distinct clinicopathologic subtypes of cortical dysplasia of Taylor. Neurology 64 (1): 55–61. Lee, H. W., S. B. Hong, W. S. Tae, D. W. Seo, and S. E. Kim. 1999. Partial seizures manifesting as apnea only in an adult. Epilepsia 40 (12): 1828–1831. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earl. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38 (1): 47–55. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9 (1): 19–30.

Intractable Epilepsy in the Setting of MCD as a Mechanism for SUDEP

233

Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16 (1): 1–7. Najm, I., Z. Ying, T. Babb, P. B. Crino, R. Macdonald, G. W. Mathern, and R. Spreaἀco. 2004. Mechanisms of epileptogenicity in cortical dysplasias. Neurology 62 (6 Suppl 3): S9–S13. Najm, I. M., Z. Ying, T. Babb, A. Mohamed, J. Hadam, E. LaPresto, E. Wyllie et al. 2000. Epileptogenicity correlated with increased N-methyl-d-aspartate receptor subunit NR2A/B in human focal cortical dysplasia. Epilepsia 41 (8): 971–976. Nashef, L., D. R. Fish, J. W. Sander, and S. D. Shorvon. 1995. Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral centre. J Neurol Neurosurg Psychiatry 58 (4): 462–464. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300. Nashef, L., N. Hindocha, and A. Makoff. 2007. Risk factors in sudden death in epilepsy (SUDEP): The quest for mechanisms. Epilepsia 48 (5): 859–871. Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Nilsson, L., T. Tomson, B. Y. Farahmand, V. Diwan, and P. G. Persson. 1997. Cause-speciἀc mortality in epilepsy: A cohort study of more than 9,000 patients once hospitalized for epilepsy. Epilepsia 38 (10): 1062–1068. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52 (2): 117–127. Palmini, A., A. Gambardella, F. Andermann, F. Dubeau, J. C. da Costa, A. Olivier, D. Tampieri et al. 1995. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 37 (4): 476–487. Palmini, A., I. Najm, G. Avanzini, T. Babb, R. Guerrini, N. Foldvary-Schaefer, G. Jackson et al. 2004. Terminology and classiἀcation of the cortical dysplasias. Neurology 62 (6 Suppl 3): S2–S8. Pedley, T. A., and W. A. Hauser. 2002. Sudden death in epilepsy: A wake-up call for management. Lancet 359 (9320): 1790–1791. Ribacoba Montero, R., C. Garcia Pravia, A. Astudillo, and J. Salas Puig. 2002. Bilateral fronto-occipital polymicrogyria and epilepsy. Seizure 11 (Suppl A): 298–302. Richardson, M. P., M. J. Koepp, D. J. Brooks, D. R. Fish, and J. S. Duncan. 1996. Benzodiazepine receptors in focal epilepsy with cortical dysgenesis: An 11C-flumazenil PET study. Ann Neurol 40 (2): 188–198. Ruggieri, P. M., I. Najm, R. Bronen, M. Campos, F. Cendes, J. S. Duncan, H. G. Weiser, and W. H. Theodore. 2004. Neuroimaging of the cortical dysplasias. Neurology 62 (6 Suppl 3): S27–S29. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Seeck, M., S. Zaim, V. Chaves-Vischer, O. Blanke, M. Maeder-Ingvar, M. Weissert, and E. Roulet. 2003. Ictal bradycardia in a young child with focal cortical dysplasia in the right insular cortex. Eur J Paediatr Neurol 7 (4): 177–181. Singh, B., A. al Shahwan, and S. M. al Deeb. 1993. Partial seizures presenting as life-threatening apnea. Epilepsia 34 (5): 901–903. Sisodiya, S. M., N. Moran, S. L. Free, N. D. Kitchen, J. M. Stevens, W. F. Harkness, D. R. Fish, and S. D. Shorvon. 1997. Correlation of widespread preoperative magnetic resonance imaging changes with unsuccessful surgery for hippocampal sclerosis. Ann Neurol 41 (4): 490–496. Sperling, M. R., A. Harris, M. Nei, J. D. Liporace, and M. J. O’Connor. 2005. Mortality after epilepsy surgery. Epilepsia 46 (Suppl 11): 49–53.

234 Sudden Death in Epilepsy: Forensic and Clinical Issues Tassi, L., N. Colombo, R. Garbelli, S. Francione, G. Lo Russo, R. Mai, F. Cardinale et al. 2002. Focal cortical dysplasia: Neuropathological subtypes, EEG, neuroimaging, and surgical outcome. Brain 125 (Pt 8): 1719–1732. Taylor, D. C., M. A. Falconer, C. J. Bruton, and J. A. Corsellis. 1971. Focal dysplasia of the cerebral cortex in epilepsy. J Neurol Neurosurg Psychiatry 34 (4): 369–387. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36 (1): 29–36. Terrence, C. F., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann Neurol 9 (5): 458–464. Tomson, T. 2000. Mortality in epilepsy. J Neurol 247 (1): 15–21. Tomson, T., L. Nashef, and P. Ryvlin. 2008a. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7 (11): 1021–1031. Tomson, T., L. Nashef, and P. Ryvlin. 2008b. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7 (11): 1021–1031. Trottier, S., B. Evrard, J. P. Vignal, J. M. Scarabin, and P. Chauvel. 1996. The serotonergic innervation of the cerebral cortex in man and its changes in focal cortical dysplasia. Epilepsy Res 25 (2): 79–106. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. Venit, E. L., B. D. Shepard, and T. N. Seyfried. 2004. Oxygenation prevents sudden death in seizureprone mice. Epilepsia 45 (8): 993–996. White, R., Y. Hua, B. Scheithauer, D. R. Lynch, E. P. Henske, and P. B. Crino. 2001. Selective alterations in glutamate and GABA receptor subunit mRNA expression in dysplastic neurons and giant cells of cortical tubers. Ann Neurol 49 (1): 67–78. Widdess-Walsh, P., C. Kellinghaus, L. Jeha, P. Kotagal, R. Prayson, W. Bingaman, and I. M. Najm. 2005. Electro-clinical and imaging characteristics of focal cortical dysplasia: Correlation with pathological subtypes. Epilepsy Res 67 (1–2): 25–33.

16

Neurogenic Cardiac Arrhythmias Howan Leung Anne Y. Y. Chan

Contents 16.1 Introduction 16.2 What Can We Learn from Electrical Brain Stimulation Studies about the Cortical Control of the Autonomic System? 16.3 What Can We Learn from Functional Magnetic Resonance Imaging Studies Demonstrating Cortical Control of the Autonomic System? 16.4 Illustrating Ictal Bradyarrhythmia and Asystole with Scalp EEG Data 16.5 Illustrating Ictal Bradyarrhythmia and Asystole with Intracranial EEG Data (Diagram 2) 16.6 Can Ictal Bradyarrhythmia Enlighten Us about the Various Mechanisms of Neurogenic Cardiac Arrhythmia? 16.7 Could There Be a Link between Ictal Bradyarrhythmia and SUDEP? 16.8 What Lies in the Future for Researchers? Acknowledgments References

235 236 238 239 240 241 244 247 247 248

16.1╅Introduction The analysis of abnormal changes in heart rate is not only a matter of concern for cardi� ologists, but also a topic of importance among neurologists. Through the study of patients with epilepsy, which is characterized by recurrent abnormal and excessive discharge of a set of neurons in the brain (Commission on Epidemiology and Prognosis 1993), cardiac arrhythmias of neurogenic origin may be examined. In fact, it is a recognized clinical observation that during seizures there might be an increase or decrease in the rate of cardiac rhythm. Through the functional study of the brain during investigations and work-up of epilepsy patients, the cortical control of cardiac rhythmic activities may be uncovered. Blumhardt et al. (1986) analyzed the electrocardiograms (ECGs) of 26 patients with temporal lobe epilepsy and showed that ictal cardiac arrhythmias occurred in 42% of patients. Ictal bradyarrhythmia and ictal asystole, in particular, have received much attention recently because of the postulation that autonomic manifestations of seizures could be one of the mechanisms underlying sudden unexpected death in epilepsy (SUDEP), although the evidence supporting this remains fragmented. This short chapter aims to review the potential role of cardiac arrhythmia in the pathogenesis of SUDEP by examining the data pertaining to the cortical control of the autonomic system, the clinical observation of potentially life-threatening ictal bradyarrhythmia, and various proposed mechanisms for SUDEP. 235

236 Sudden Death in Epilepsy: Forensic and Clinical Issues

16.2â•…What Can We Learn from Electrical Brain Stimulation Studies about the Cortical Control of the Autonomic System? The central representation of autonomic states has long suspected, as stipulated by the James–Lange theory, that arousal and emotion were closely related (James 1894). More information about the representation of the autonomic system above the level of the brain stem was reported more than 50 years ago beginning with studies in primates and other animals (Kaada 1951). The limbic structures are thought to be the principal mediators of this function and the candidate areas include cingulate gyrus, amygdala, and insular and orbitofrontal cortex. In Kaada’s experiments, stimulation of the anterior portion of the cingulate gyrus in monkeys showed inhibition and arrest of respiration. Penἀeld and Jasper (1954) carried out experiments in humans in the 1950s and showed that stimulation of the cingulate gyrus in its anterior and inferior portion produced apnea, which could not be overcome even with a conscious effort. In addition, stimulation of the uncus and of the right anterior margin of insular cortex also produced a similar effect on respiration. The cortical influence on cardiovascular responses has been likewise demonstrated in various animal studies (Kaada 1951; Ward 1948; Burns and Wyss 1985; Chefer et al. 1997; Healy and Peck 1997) and in humans. This was demonstrated by delivering electrical stimulation to psychotic patients before bilateral fractional ablation of the cortex (Pool and Ransohoff 1949). A total of 12 patients were tested with bilateral cingulate stimulation, of which three cases had heart rate elevation, seven had heart rate reduction, and two cases showed no change. The authors did not explain what factors determined the direction of change nor whether the change in heart rate was unidirectional in any given patient. Van Buren (1961) performed electric stimulation to right orbitofrontal and right mesial temporal areas in 11 patients before pituitary surgery. The stimulation of the right orbitofrontal area was associated with a slight drop in heart rate, whereas stimulation of the right mesial temporal region produced a mixture of increase and decrease in heart rate. In an often-quoted study by Oppenheimer et al. (1992), electrical stimulation of the insular regions of ἀve epileptic patients (three with origin on the right side and two on the left side) was associated with changes in both heart rate and blood pressure. Electrical stimulation with 5 to 10 V was carried out with a pulse duration of 2 ms at 40 Hz. They found that bradycardia and depressor responses were more frequently produced than tachycardia when the left insular cortex was stimulated. The converse was noted when the right insular cortex was stimulated. Of interest, the amplitude of the bradycardia was greater on stimulation of the left posterior insular cortex compared with the left anterior insular cortex. This study was able to demonstrate that there is a left–right difference in the cortical control of heart rate (the implications of this ἀnding for SUDEP will be discussed in Section 16.7). Electrical stimulations are sometimes performed as part of the presurgical evaluation of medically refractory epilepsy, particularly in determining cortical eloquent areas and sensorimotor areas, or for the reproduction of habitual seizures. The amount of electrical stimulation given in those situations may vary from 1 to 10 mA depending on the apparatus used and the interpatient variation of parenchymal impedance. The frequency of stimulation may be higher for stimulation of eloquent areas but lower for motor areas to minimize discomfort. The duration of electrical stimulation is usually short (e.g., <10 s) to avoid the buildup of prolonged afterdischarges. In one study, electrical stimulation was delivered to a patient at 3 mA, 50 Hz for 7 s to the mesial and basal parts of the left anterior temporal lobe

Neurogenic Cardiac Arrhythmias

237

with a high-grade atrioventricular (AV) block being observed (Altenmuller et al. 2004). During this stimulation process, an epigastric aura was also triggered and epileptiform discharges localized at the left anterior temporal lobe, left amygdala, and the left anterior hippocampus were found. A case of electrical stimulation of the cingulate cortex with transient asystole was also reported during a presurgical evaluation of refractory epilepsy and the clinical ἀnding was entirely incidental (Leung et al. 2007). The patient was righthanded (the signiἀcance of the dominant hemisphere on this case illustration remained entirely speculative although in many studies this piece of information was missing) and had complex partial seizures with a semiology consisting of staring, motor arrest, and drop attacks. Magnetic resonance imaging (MRI) showed a suspected left frontal cortical dysplasia, whereas surface electroencephalogram (EEG) only revealed bilateral frontal epileptiform discharges at onset of seizures. The patient then underwent implantation of bilateral frontal strip electrodes and left interhemispheric electrodes. During the electrical stimulation event, which was carried out in the areas of interhemispheric electrodes near the left anterior cingulate cortex, at 1 mA, with pulse width 3 ms, 50 Hz for 5 s, cardiac asystole of 5 s duration was observed. The amount of electrical stimulation was only 1 mA as the procedure usually involved a gradual buildup and/or titration of the current while observing the clinical effect of the stimulation. Stimulation of the adjacent areas (still within the cingulate cortex) in a separate trial reproduced symptoms of aura and motor arrest. Note that electrical stimulation studies involving human subjects are often limited by the relatively small number of subjects (in many instances, case reports only) and the limited areas of stimulation owing to the need to conform to either an operative procedure, or to a speciἀc array of intracranial electrode placement in keeping with a preoperative seizure localization hypothesis. The presence of pathology may also shift the network responsible for the autonomic control of heart rate to an atypical location (e.g., in individuals with

Right insular cortex

Left insular cortex Left basotemporal region Right anterior cingulate

Right uncus

Right orbitofrontal cortex

Left anterior cingulate

Figure 16.1╇ Areas of brain producing effects on heart rate based on human cortical stimula-

tion studies.

238 Sudden Death in Epilepsy: Forensic and Clinical Issues

insular glioma, the “speculated” area influencing autonomic function may progress to an area outside the insular region as a result of tumor growth). The hand dominance of the testing subjects was likewise seldom mentioned, apart from the study of Oppenheimer et al. (1992). Figure 16.1 indicates, based on previous studies of human electrical stimulation, the various regions of brain producing effects on heart rate.

16.3â•…What Can We Learn from Functional Magnetic Resonance Imaging Studies Demonstrating Cortical Control of the Autonomic System? Recent advances in functional neuroimaging techniques have enabled researchers to evaluate the cortical control of the autonomic system by noninvasive means. The anterior cingulate cortex has received intensive research in this respect. Anatomically pyramidal neurons in anterior cingulate cortex bear connection with hypothalamus (Ongur et al. 1998), periaqueductal gray (Ongur et al. 1998) and pontine gray matter (Vilensky and van Hoesen 1981) as well as frontal and parietal areas (Allman et al. 2001). It is therefore well suited to the functional notion of integrating behavior within the autonomic system. Critchley et al. (2003) demonstrated that during effortful cognitive and motor behavior the dorsal anterior cingulate cortex supports the generation of associated autonomic states of cardiovascular arousal. Subjects performed alternating 2-min blocks of paced cognitive “n-back” and isometric handgrip exercise conditions in counterbalanced repetitions during functional MRI scanning. In six healthy volunteers, the authors located areas of interest in relation to sympathetic (low-frequency band) and parasympathetic (high-frequency band) control. One example from the results showed regional activity changes for parasympathetic control in the left cingulate and left paracentral/superior parietal lobule. On the other hand, regional activity changes for sympathetic control were found in right cingulate, right medial temporal lobe, and right inferior parietal lobule. The insular regions according to this study did not show much lateralization tendency as bilateral insular cortices have increased low-frequency power (sympathetic control). This study also looked at three patients with cingulate lesions—one with bilateral lesions, one with a predominantly leftsided lesion, and one with a predominantly right-sided lesion. They all had blunted heart rate variability but the second patient showed slightly more impairment of the parasympathetic control of the heart rate variability while performing mental tasks. However, in the authors’ conclusion, only modulation of the sympathetic system by the anterior cingulate cortex was emphasized. In a study by Matthews et al. (2004), 18 subjects performed a counting Stroop task twice, one during functional MRI (fMRI) scanning and another during heart rate recording. Stroop task was a frontal-lobe function-speciἀc psychological test and it originally consisted of the presentation of subjects with words of colors (e.g., “red”) printed in a different color (e.g., “green”) while the subject was asked to name the color of the word. It was also thought to be able to elicit heart rate variability changes. In this modiἀed version of a Stroop task, namely, a “counting” Stroop task, subjects had to report the number of rows of a block of numbers presented on a screen. A congruent stimulus consisted of presentation of a block of numbers that were numerically equivalent to the number of rows; an incongruent stimulus consisted of presentation of a block of numbers that were numerically different from the number of rows. The test may also be given as a “fast” run or “slow” run

Neurogenic Cardiac Arrhythmias

239

depending on the speed of presentation. With the imaging results, the authors reported that the dorsal anterior cingulate cortex was activated during incongruent conditions relative to the congruent conditions. Peak high-frequency heart rate variability over a “fast” run signiἀcantly correlated with activation in the left ventral anterior cingulate cortex during congruent trials as well as incongruent trials. This result was in keeping with the previous fMRI study in that the anterior cingulate cortex is part of the central autonomic network, and it also lent support for the involvement of anterior cingulate cortex in the modulation of the parasympathetic component of the autonomic system. In addition, it gave evidence to support the ἀne anatomical subdivision of the anterior cingulate cortex, and the association of the parasympathetic modulation with the left cerebral hemisphere. In both fMRI studies, it must be appreciated that the paradigm in use was not the same, and moreover, in the latter study, ECG monitoring had to be performed outside the scanning machine thus precluding real-time monitoring of heart rate. Consideration was not given in either study as to the hand dominance of the subjects, and this can make the inference on lateralization clinically less useful.

16.4â•…Illustrating Ictal Bradyarrhythmia and Asystole with Scalp EEG Data As early as 1906 there had been reports of cardiac standstill during seizures (Russell 1906). The term “ictal bradycardia syndrome” (Reeves et al. 1996) was subsequently coined. Tinuper et al. (2001) reviewed 60 cases from the literature and reported 3 more cases of ictal bradyarrhythmia. Taking into account all types of ictal bradyarrhythmia and using scalp EEGs for localization of seizure onset, the authors found that temporal lobe epilepsy accounted for 67% of cases whereas frontal lobe epilepsy accounted for 33% (Tinuper et al. 2001) (Figure 16.2). Analysis of seizures with multiple simultaneous onsets showed that an additional 13% may arise from the orbitofrontal cortex. When all cases were included, slightly more had a left-sided seizure onset (n = 26) than a right-sided onset (n = 19). Ictal asystole was much less common, and was reported in an equal number (n = 9) of left- and right-sided seizure-onset cases. Hand-dominance was, however, not reported in the data. A hospital record review of 1244 patient ἀles found only 5 cases of ictal asystole (Rocamora et al. 2003). One patient had bifrontal onset, two had left temporal onset, and two had left frontal onset. This ἀnding suggested the same left–right paradigm described in previous studies. Interestingly, in three cases a direct or indirect participation of the left insular cortex was also suspected on MRI or metabolic scans. The ECG analysis showed Right frontopolar regions

Right temporal regions

Left temporal basal regions

Figure 16.2╇ Areas of brain producing ictal bradyarrhythmia based on intracranial EEG data.

240 Sudden Death in Epilepsy: Forensic and Clinical Issues

that cardiac asystole was produced by sinus arrest that may be preceded by sinus bradycardia. The evidence may point toward a vagal-mediated mechanism although alternative explanations such as synchronization of cardiac autonomic neural discharge with epileptogenic activity were also pointed out by the authors (Lathers et al. 1983; O’Rourke and Lathers 2010). In the largest series of ictal bradycardia, which was not included in the aforementioned review (Tinuper et al. 2001), 13 patients with ictal bradycardia were diagnosed and all were associated with temporal lobe epilepsy. The authors did not ἀnd a consistent lateralizing value for ictal bradycardia, as right-side onset was observed in 7, left-side onset was observed in 5, and nonlateralizing onset was observed in 1. The authors also reported that a difference in electrographic involvement may be observed at the time of ictal bradycardia compared with that observed at seizure onset, and, in fact, bilateral involvement was more common. Hand dominance was also not reported in this study (Britton et al. 2006). In a prospective long-term study of 20 patients using implantable loop recorders, heart rhythm with a total of 220,000 patient hours of recording was obtained over 24 months (Rugg-Gunn et al. 2004). The cohort was drawn from a patient population with highly refractory epilepsy from a tertiary referral center, and one patient already had a history of peri-ictal arrest, although no patient had existing cardiac disease. Ictal bradycardia was observed relatively frequently—eight such events were recorded in seven patients. Four patients were found to have periods of asystole sufficient to warrant cardiac pacemaker implantation. Three of the seven patients with ictal bradycardia had left-side onset seizures, two had right-side onset seizures, and two either bilateral or undetermined. Early morning bradycardia was found in two patients, raising the possibility of underlying interictal abnormal heart rate variability. Interpretation of these ἀndings, however, was limited by the lack of real-time EEG recording during the asystolic events, and at other times when seizure activities were suspected, recording was only set off by either a patient-activated button or preset heart rate limits. In another recent study, a database search was performed of 6825 patients undergoing long-term video EEG and 0.27% of patients demonstrated ictal asystole. Eight patients had temporal epilepsy, two had extratemporal epilepsy, and none had generalized epilepsy. In 8 out of 16 recorded events, seizures were associated with a sudden atonia >40 s into the typical semiology of a complex partial seizure. However, EEG changes were only typical of cerebral hypoperfusion and no clinical predisposing factors can be identiἀed (Schuele et al. 2007).

16.5â•…Illustrating Ictal Bradyarrhythmia and Asystole with Intracranial EEG Data (Diagram 2) Ictal bradyarrhythmia captured on invasive EEG monitoring can give invaluable information, although the event is actually one that occurs rarely. Among the few reported cases (Altenmuller et al. 2004; Broglin and Bancaud 1991; Munari et al. 1995; Devinsky et al. 1997; Manitius-Robeck et al. 1998; Kahane et al. 1999; Rossetti et al. 2005), only four provided sufficient clinical information enabling correlation between ictal EEG and onset of bradyarrhythmia. In one of these reports, a patient with right hippocampal atrophy was described who underwent presurgical evaluation with placement of left temporal subdural grid and strip electrodes, right temporal subdural electrodes, and right orbitofrontal and

Neurogenic Cardiac Arrhythmias

241

frontal electrodes. Subsequent evaluations led to left temporal resective surgery rather than the right. An 8-s period of asystole took place after the onset of seizure and the intracranial recording registered a left temporal onset, although the asystole did not start until the seizure had spread to the right temporal lobe (Devinsky et al. 1997). In another case, a patient with hypothalamic hamartoma was implanted with 10 electrodes exploring the right frontal, central, and temporal cortices, two electrodes within the left frontal lobe, and one implanted within the lesion. The patient had dacrystic (i.e., crying seizures with intracranial EEG monitoring showing epileptiform activities beginning from and remaining localized in the hamartoma). An episode of ictal asystole, however, was registered with the onset of epileptiform discharges in the right frontocentral and temporal neocortical areas rather than the hypothalamic mass itself (Kahane et al. 1999). It was noteworthy that this patient was left-handed. In another report, a patient had foci of encephalomalacia in the right inferior frontal gyrus, temporal pole, and anterior parietal lobe. Invasive monitoring was carried out with bilateral temporal depth electrodes and right dorsolateral frontoparietal subdural strips. Asystole occurred in a seizure episode with left temporal seizure onset, and sharing similarity with the ἀrst case report, the actual phenomenon of asystole only occurred after an interhemispheric spread to opposite mesial structures. The proposed route of spread to the opposite side was thought to be via the anterior commissure or the corpus callosum. Although no intracranial electrode was placed in the neocortical portions of the left temporal lobe, the surface EEG showed rhythmic activity over the left lateral temporal region before involvement of the opposite mesial structure so that involvement of the left insular region before interhemispheric spread remained possible. Intracranial electrodes conἀrmed that the right frontal lobe was spared (Rossetti et al. 2005). In the last of the four reports, a patient with left temporal lobe epilepsy was implanted with left lateral and basal temporal subdural grid (x1) and strips (x2) with an additional left temporal depth electrode. Further evaluation eventually led to left temporal lobe surgery, the pathology of which was shown to be glioneuronal hamartoma. High-grade AV blocks were observed with left temporal seizure onset that was recorded before intracranial electrode placement. A left basal and anterior lateral temporal lobe seizure onset was thought to be present during the ictal cardiac event as this was subsequently inferred from the intracranial recording. Epileptic discharges were thought to be localized during the event of AV block on the basis of scalp EEG ἀndings, although the time lag from seizure onset and the placement of intracranial recording at a separate time may still raise doubts about a possible spread. Electrical stimulation to this region also reproduced AV block once, as described previously in the section about human brain stimulation studies (Altenmuller et al. 2004).

16.6â•…Can Ictal Bradyarrhythmia Enlighten Us about the Various Mechanisms of Neurogenic Cardiac Arrhythmia? Factors predisposing an epilepsy patient to ictal bradyarrhythmia remain unknown. It is tempting to suspect that certain seizure mechanisms may predispose to ictal bradyarrhythmia and there might be additional pathological factors contributing to this, such as an aberrant condition in the mediators of cortical autonomic control (e.g., the vagus nerve), concomitant antiepileptic drugs thought to be arrhythmogenic, imbalance between the

242 Sudden Death in Epilepsy: Forensic and Clinical Issues

sympathetic and parasympathetic divisions of the autonomic nervous system, and/or a cardiac substrate (i.e., underlying disease process affecting the cardiac functions). The observation of bradyarrhythmia associated with temporal and frontal lobe onset seizures may lend support to the hypothesis that activation of these regions is part of the mechanism, and in particular, the operculo–insulo–mesiotemporal–orbital complex (Mufson and Mesulam 1982). However, none of the intracranial reports of ictal bradyarrhythmia gave information about involvement of the left insular cortex as the electrodes are seldom placed in this area in clinical practice. Selection bias is also possible, given the fact that there is a preponderance of temporal and frontal epilepsies in specialized centers. The concept of recognizing a pattern of seizure spread during ictal bradyarrhythmia is important, as our analysis of intracranial EEG data showed that seizures beginning from the central autonomic network on the dominant hemisphere were present in the episodes with ictal bradycardia. In addition, a spread to the contralateral central autonomic network was frequently witnessed. The timing of onset of bradycardia in relation to seizure onset was supporting evidence for this. We suspect that the cerebral dominance may alter the “internal wiring” of the central autonomic network, hence a patient with left cerebral dominance may be observed with a left-to-right spread, and vice versa for a patient with right cerebral dominance, although in the intracranial reports there was no patient with bilateral cerebral dominance. This idea of cerebral dominance may be mirrored by the fact that language representation can be lateralized to either hemisphere or bilaterally during the patient’s developmental process. However, we did not know exactly what effect the lesions had on the mechanisms producing bradycardia, as the spreading of epileptiform discharges to the opposite hemisphere cannot be easily correlated with whether the lesion was ipsilateral or contralateral to the seizure onset using the intracranial data. In the only case with hypothalamic hamartoma, there was no seizure spread required for the process and the bradycardia was registered at the onset of seizure. In the case where the lesion was on the same side as seizure onset, no spreading was noted (Altenmuller et al. 2004). In the two cases where the lesion was contralateral to the seizure onset, spreading can be witnessed (Devinsky et al. 1997; Rossetti et al. 2005). We also suspect that the presence of a lesion may disrupt the original central autonomic network, keeping the speciἀc network required for the production of bradycardia more localized than it otherwise would be (e.g., only localized to one part of the central autonomic network, or simply one cerebral hemisphere). The inception of bradycardia may also be viewed as the indirect result of epileptiform discharges, and hence a less localization-speciἀc mechanism. In this respect, bradycardia may be viewed as a release phenomenon comparable to the release mechanisms observed in many mesial temporal seizures, such as chewing and epigastric sensation, which are also vegetative in nature. Under this precinct, the generation of the release mechanism did not have to have a one-to-one transmission relationship with the ictal epileptiform discharges. Regarding the issue of lateralization of seizures in ictal bradyarrhythmia, the basic science data as outlined above may lend support for this to occur at ἀrst glance, despite the existence of some inconsistencies. There was also an additional study investigating heart rate during a Wada test (Zamrini et al. 1990) with a total of 25 patients, using a repeatedmeasures analysis of variance design. (A Wada test is a commonly performed procedure in the presurgical evaluation of epilepsy in which intracarotid injections of amobarbital are given and the subjects tested for language and memory. The test indicates which of

Neurogenic Cardiac Arrhythmias

243

the two cerebral hemispheres is the dominant hemisphere and it may give an approximation to the effect of surgery after lobectomy.) The authors suggested that left carotid artery infusion was associated with tachycardia whereas right carotid artery infusion was noted with bradycardia. However, these subjects were not patients reported to have any ictal bradycardia. In practice, as we have seen with the observations in ictal bradycardia recorded with intracranial EEGs, the pattern of seizure spread, hand dominance, and the presence of lesion contribute to the lateralization issue, making the simplistic left/right dichotomy less likely for one to predict. Mediators of the cortical signals for the autonomic system should also be considered. The parasympathetic system and the vagus nerves are essential in this respect. The functional distribution of the efferent ἀbers of the vagi also demonstrates laterality, although it is not entirely clear how this may relate to the laterality observed in the cortical representation of cardiac autonomic control. The left vagus nerve carries fewer efferent ἀbers to the ventricle, which is why traditionally vagus nerve stimulation is performed on the left side. (Fewer efferent ἀbers may mean that electrical stimulation of the vagus nerve produces less vagal-mediated parasympathetic activation and/or bradycardia.) Using information about patients in whom vagus nerve stimulators were implanted, it was observed that neither left-sided nor right-sided vagal stimulation gave rise to frequent bradycardia or asystole (McGregor et al. 2005; Handforth et al. 1998). However, one case report during intraoperative implantation of left vagus stimulator did show persistent bradycardia and the author (Asconape et al. 1999) proposed, among other speculations, that the intensity of the vagus stimulation, or some idiosyncratic mechanism at work, may contribute to this. It was not known whether during a seizure such intensity or amplitude of vagus nerve activation can be achieved or if an ictal bradyarrhythmia occurs at all. Some authorities may view that both sympathetic and parasympathetic systems are at work during ictal bradyarrhythmia, owing to the synchronization of the cardiac autonomic neural discharge with epileptogenic activity, the so-called “lockstep” phenomenon (Lathers et al. 1983). In this theory, the effect may vary from animal to animal or patient to patient, so sometimes there might be excessive sympathetic stimulation and at other times only parasympathetic nervous system dominance, or an imbalance between the two divisions. Previous studies (Stauffer et al. 1989, 1990) and O’Rourke and Lathers (2010) reported precipitous mean arterial blood pressure changes correlating with unstable lockstep phenomenon. Four possible mechanisms through which lockstep phenomenon may be related to arrhythmia and sudden death in persons with epilepsy were postulated: (1) excessive sympathetic stimulation of a heart that is already electrically unstable due to prior damage (Jay and Leestma 1981); (2) a nonuniform discharge in the postganglionic cardiac sympathetic nerve branches (Lathers et al. 1977, 1978); (3) the parasympathetic nervous system causing sinus arrest and bradycardia during seizures (Kiok et al. 1986; Lathers and Schraeder 1982); and (4) the associated precipitous changes in blood pressure per se. A coexisting pathological cardiac substrate is, and has always been, a potential confounding factor in determining the mechanism underlying ictal bradyarrhythmia. Studies on refractory epilepsy patients showed that echocardiographic abnormalities can be found in up to 9% of patients (Tigaran 2002). Among the ἀve patients observed to have ictal asystole in one study mentioned earlier, two had underlying cardiac disease that was reported by the authors to be a history of myocardial infarction in one and a history of an aberrant complex in the other (Rocamora et al. 2003). Another school of thought proposes that long-standing epilepsy may alter the neuronal network system on the cardiac tissues. One

244 Sudden Death in Epilepsy: Forensic and Clinical Issues

study found altered postganglionic cardiac sympathetic innervation with a predominantly parasympathetic cardiac activity in patients with temporal lobe epilepsy by using MIBGSPECT (123-I metaiodobenzylguanidine single photon emission computed tomography) (Druschky et al. 2001). Others argue that there may be a common pathology that leads to both seizure and arrhythmia, such as a channelopathy (e.g., mutations of genes encoding nicotinic acetylcholine receptors) that may underlie many forms of inherited idiopathic generalized epilepsy (Hiroso et al. 2005). The potential arrhythmogenic effect of antiepileptic drugs should also be considered. It is well documented that carbamazepine can slow the AV conduction and increase the sympathetic tone in the autonomic nervous system (Isojarvi et al. 1998) and decrease heart rate variability (Persson et al. 2003). One study demonstrated that abrupt withdrawal of carbamazepine can lead to enhanced sympathetic activity in sleep (Hennessy et al. 2001). Other studies have reported conflicting results (Kenneback et al. 1992, 1997). In the clinical cases discussed above, not all patients were receiving carbamazepine when ictal bradyarrhythmia occurred. However, the contributory ἀndings for other antiepileptic drugs are not easy to decipher because most, if not all, of those patients in whom ictal bradyarrhythmia occurred had been put on multiple drugs rather than a single one. Another caveat in the discussion of ictal bradyarrhythmia is the careful ascertainment of the time sequence between the seizure event and the cardiac arrhythmia. A neurocardiogenic syncope in which the bradyarrhythmic event precedes the clinical attack will be a differential diagnosis for all patients who have ictal bradyarrhythmia.

16.7â•…Could There Be a Link between Ictal Bradyarrhythmia and SUDEP? The incidence of SUDEP varies between 1 in 200 and 1 in 1000 (Lhatoo and Sander 2002). A widely accepted deἀnition of SUDEP is “a sudden unexpected nonaccidental death in an individual with epilepsy, with or without evidence of a seizure having occurred, excluding status epilepticus, where autopsy does not reveal an anatomical or toxicological cause of death” (Nashef 1997). Another deἀnition consisting of six criteria was put forward by an expert panel in 1997 (Leestma et al. 1997). In all these deἀnitions, a “deἀnite” case can only be reached if all criteria are satisἀed, but where postmortem data are lacking, the term “probable” should be used. Many risk factors have been stratiἀed for SUDEP and on face value they may not always be consistent (Langan and Sander 1999; Kloster and Engelskjon 1999; Birnbach et al. 1991; George and Davis 1998; Opeskin and Berkovic 2003; Schnabel et al. 2000; Shields et al. 2002; Opeskin et al. 2000; Jick et al. 1992; McKee and Bodἀs 2000; Nilsson et al. 1999; Timmings 1993; Walczak et al. 2001). But on the other hand, these risk factors for SUDEP may depend on the type of controls used, and therefore may be regarded as complementary (Tellez-Zenteno et al. 2005). For instance, in the studies in which non-SUDEP deaths were used as controls, seizure preceding death, subtherapeutic drug levels, and patient-found-in-bed were considered the most consistent factors (Kloster and Engelskjon 1999; Birnbach et al. 1991; George and Davis 1998; Opeskin and Berkovic 2003; Schnabel et al. 2000; Shields et al. 2002; Opeskin et al. 2000). These studies may be clinically useful in terms of ascertaining the peri-SUDEP circumstances of the patients. In studies that used persons living with epilepsy as controls, the main risk factors were high seizure frequency, high number of antiepileptic drugs, youth (but not children), and

Neurogenic Cardiac Arrhythmias

245

long duration of epilepsy (Jick et al. 1992; McKee and Bodἀsh 2000; Nilsson et al. 1999; Timmings 1993; Walczak et al. 2001). These studies may be better in terms of determining enduring, long-term factors affecting SUDEP. Although risk factors located in epidemiological studies may not offer detailed mechanistic explanations for the mechanism of SUDEP, they may nonetheless provide lines of thinking to support or refute proposed mechanisms of SUDEP. Central to this issue is that seizures do have an important role to play in SUDEP. Research into the pathogenic mechanism of SUDEP has proposed several paradigms: (1) cortical control of autonomic system, in terms of ictal bradyarrhythmia; (2) primary cardiac disease; (3) arrhythmogenic side effects of antiepileptic drugs; and (4) respiratory€mechanisms (e.g., the contribution of peri-ictal apnea and hypoxia in generalized tonic–clonic seizures is well recognized and pulmonary edema to varying extent has been reported in patients with SUDEP), mainly with retrospective ἀndings, particularly autopsy cases (Stöllberger and Finsterer 2004, Lathers and Schraeder 2010). It is not difficult for one to decipher the overlap between the mechanisms underlying ictal bradyarrhythmia and SUDEP in this regard owing to the suggestion that the cortical control of the autonomic system is a prevailing underlying explanation. The discussions about the pathogenic mechanisms of ictal bradyarrhythmia may very well apply to SUDEP, but this is made on the assumption that ictal bradyarrhythmia actually occurs during a SUDEP episode. Epidemiological data suggesting that a seizure may occur shortly before SUDEP bear witness to the possibility of ictal bradyarrhythmia having taken place. In fact, in the study in which the highest ἀgure was quoted, signs of seizures starting the ἀnal event were observed in 67% of SUDEP patients, all due to generalized motor seizures (Kloster and Engelskjon 1999). If the patient was found to have SUDEP in bed or during sleep, then it may well be possible that during such times the patient was in a physiological state that predisposed the patient to SUDEP, such as reduced sympathetic tone to counteract ictal bradyarrhythmia or increased autonomic instability leading to extreme fluctuations in heart rate. More discussion regarding sleep and SUDEP will be given in another chapter in this book (Chapter 23; Hughes and Sato 2010). However, in our analysis of patients with ictal bradyarrhythmia (both scalp and intracranial EEG data), there was insufficient information to indicate whether ictal bradyarrhythmia may occur more often during sleep. Another useful clinical method would be to look for the characteristics in a cohort with both SUDEP and ictal arrhythmia. In one recent study (Nei et al. 2004) involving 21 patients with deἀnite and probable SUDEP, ictal cardiac repolarization and rhythm abnormalities were found to occur in 56% of cases, although only 16 out of the 21 patients had continuous ECG data recorded with video EEG and the analysis was understandably performed in retrospect. The rhythm abnormalities may range from atrial ἀbrillation, through ventricular premature depolarizations, to junctional escape. No overt ictal asystole and no association with laterality were found. Authors from this study stipulated that seizures from sleep can cause sudden and extreme fluctuations in autonomic tone, which can trigger lethal cardiac arrhythmia, including bradycardia. In one recent SUDEP review, the association between cardiac dysfunction and SUDEP was not substantiated (Tellez-Zenteno et al. 2005). Thus, we may see that the clinical data were not in perfect agreement with the mechanistic explanation of SUDEP using ictal bradyarrhythmia alone, and there is a possibility that individual variability can be important and different patients may have different mechanisms (Lathers 1982, 2008). Intrinsic cardiac diseases among SUDEP cases were suspected often by virtue of autopsy analyses. These were made on histological grounds such as focal myocarditis or

246 Sudden Death in Epilepsy: Forensic and Clinical Issues

hypertrophic or dilative cardiomyopathy or other problems in the conduction system (Cohle et al. 2002) although the quality of specimen analysis might be less than desirable due to the advanced stage of death. Speciἀc ion channel and preexcitation syndromes were also suspected by researchers, and, for instance, the long QT syndromes may be due to mutations of sodium and potassium cardiac ion channels and the short QT syndrome may be due to missense mutation in the potassium ion channel IKrHERG (de la Grandmaison 2006). Such pathology may, however, not be readily demonstrable at autopsy, making retrospective analysis of this theory rather difficult (Davies 1999). There was speculation that such genetic mutation may underlie both an epilepsy syndrome and a cardiac syndrome although the exact disease entity remains unknown (Hiroso et al. 2005). Moreover, these entities may comprise theoretically both tachycardia and bradycardia syndrome. Epidemiological studies did not identify as yet intrinsic cardiac disease as a risk factor (Tellez-Zenteno et al. 2005) and cardiac enzymes (such as troponin T) during seizure were previously shown not to be elevated ictally (Woodruff et al. 2003). Therefore, the contribution of an intrinsic cardiac disease to SUDEP, with or without direct reference to ictal bradyarrhythmia, remains unknown from a clinical point of view. The speciἀc association of carbamazepine with SUDEP has been previously investigated. An audit of an epilepsy clinic in Wales showed that carbamazepine was used in higher proportion among the 14 SUDEP patients than among the general epilepsy patients attending the clinic (Timmings 1998). However, confounding with other variables had not been removed. In a Norwegian study, carbamazepine was also shown to be in small excess among the SUDEP cases but polytherapy precluded worthwhile analysis (Kloster and Engelskjon 1999). The arrhythmogenic properties of carbamazepine had been already discussed, but once again, the contribution of carbamazepine to SUDEP with or without a linkage to ictal bradyarrhythmia is still open to debate. Please see a separate chapter in this book regarding the risk of SUDEP with pharmaceutical drugs (Chapter 51; Tomson 2010). Respiratory mechanisms have been reported to be compromised during an ictal event using polysomnography in 20 of 47 clinical seizures according to one study (Nashef et al. 1996). Central apnea was most often observed but obstructive apnea was also present. What is more, bradycardia was shown to be associated with apnea during the same time. In two reports featuring witnessed SUDEP cases, seizure associated respiratory embarrassment was a prominent observation (Nashef et al. 1998; Langan et al. 2000). The relationship between ictal bradyarrhythmia and ictal apnea lies at the proposition that either may be mediated by a similar set of central autonomic networks as suggested by previous electrical stimulation studies (Penἀeld and Jasper 1954). However, apnea due to noncentral influence may not be explained by this. In studies examining the risk factors for SUDEP, the proportion of patients found dead in a prone position ranged from 42% to 71% (Kloster and Engelskjon 1999; Nashef et al. 1996; Earnest et al. 1992). It was postulated that in such a position ventilation may be easily compromised, such as bringing about collapse of upper airways, particularly if seizure had happened. Pathological examination of SUDEP cases also revealed the common occurrence of pulmonary edema in the autopsy that pointed toward the involvement of apnea in peri-SUDEP circumstances (Black and Graham 2002). In the only case of SUDEP occurring in an intracranially monitored patient, the death occurred during a seizure with right mesial temporal onset with the EEG becoming flat after 16 s in the right hemisphere. The left hemisphere showed spike discharges for a further 8 s before ceasing suddenly as well. However, there was no respiratory or ECG recording

Neurogenic Cardiac Arrhythmias

247

although pulse artifact was found for a further 120 s at a rate of 46 s. During the event the patient was lying face down although no apparent evidence of asphyxia was noted. Bird et al. (1997) proposed that the mechanism of SUDEP in this case was entirely cerebrogenic. Therefore, the mechanism underlying SUDEP may be a multifactorial one, with contribution of seizure playing a major role, and with the central influence of cardiac and respiratory functions giving rise to potentially life-threatening events, with or without extra predisposition from intrinsically altered physiological parameters such as a cardiac substrate (Amlie 1997), or drugs, together with certain peri-SUDEP events, such as a prone position or during sleep, which ultimately brings about the demise of the patient.

16.8â•…What Lies in the Future for Researchers? We have reexamined the biological plausibility of the cortical control of the autonomic system in the explanation of ictal bradyarrhythmia. Although the evidence from scientiἀc studies is in keeping with this general notion, further analysis to allow for a clear dissection of the mechanism is not readily available, such as the left-right paradigm. Further extrapolation of ictal bradyarrhythmia to a mechanistic explanation for SUDEP has remained elusive. The missing links are (1) clinical evidence of common factors shared by ictal bradyarrhythmic patients and SUDEP patients, (2) evidence of arrhythmia from epidemiological studies as a risk factor for SUDEP, and (3) ascertaining the importance of ictal bradyarrhythmia in SUDEP with regard to other proposed mechanisms including apnea and intrinsic cardiac abnormalities. It may well be possible that SUDEP has an underlying mechanism attributable to multiple causes rather than a single, unifying factor. From the seizure mechanistic perspective, and also based on data from electrical stimulation and functional imaging studies, it might be logical to speciἀcally examine cases in which intracranial EEG monitoring of the left insular region took place. However, there is ethical concern in putting an intracranial electrode near the insular region simply to look for seizure spread rather than origin. In addition, in individual patients with intracranial recording showing seizure onset from the insular region, the presence of lesion may alter the normal physiological location of the autonomic cortical pathways. From the clinical point of view, it would be logical to look again at the underlying seizure mechanism and location of seizure onset and spread in each and every case of SUDEP. Given the increasing utilization of presurgical work-up for refractory epilepsy patients, the number of SUDEP cases with previous intracranial recording may be increasingly found. Comparison of SUDEP cohorts with ictal bradyarrhythmia patients and exploration of other alternative mechanisms for SUDEP may be potential areas for further research.

Acknowledgments We thank the research team at the Division of Neurology, Department of Medicine and Therapeutics, Chinese University of Hong Kong. We are grateful to Professor Lawrence K. S. Wong, the chair professor in Neurology at the Chinese University of Hong Kong, for guidance over our research interests over the years. We are also most indebted to our collaborators at the Department of Epileptology, University of Bonn, Germany, and to Professor Christian Elger for his most kind and generous offers of training opportunities

248 Sudden Death in Epilepsy: Forensic and Clinical Issues

during our visits to the University of Bonn. We are grateful to Drs. Vincent Ip and Lisa Au for proofreading the manuscript.

References Allman, J. M., A. Hakeem, J. M. Erwin, and E. Nimchinsky. 2001. The anterior cingulate cortex: The evolution of an interface between emotion and cognition. Ann N Y Acad Sci 935: 107–117. Altenmuller, D. M., M. Zehender, and A. Schulze-Bonhage. 2004. High grade atrioventricular block triggered by spontaneous and stimulation induced epileptic activity in the left temporal lobe. Epilepsia 45 (12): 1640–1644. Amlie, J. P. 1997. QT dispersion and sudden cardiac death. Eur Heart J 18: 189–190. Asconape, J. J., D. D. Moore, D. P. Zipes, L. M. Hartman, W. H. Duffell. 1999. Bradycardia and asystole with the use of vagus nerve stimulation for the treatment of epilepsy: A rare complication of intraoperative device testing. Epilepsia 40 (10): 1452–1454. Bird, J. M., K. A. T. Dembny, D. Sandeman, and S. Butler. 1997. Sudden unexplained death in epilepsy: An intracranially monitored case. Epilepsia 38 (suppl. 11): S52–S56. Birnbach, C. D., A. J. Wilensky, and C. B. Dodril. 1991. Predictors of early mortality and sudden death in epilepsy: A multidisciplinary approach. J Epilepsy 4: 11–17. Black, M., and D. I. Graham. 2002. Sudden unexplained death in adults caused by intracranial pathology. J Clin Pathol 55: 44–50. Blumhardt, L. D., P. E. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1 (8489): 1051–1056. Britton, J. W., G. R. Ghearing, E. E. Benarroch, and G. D. Gascino. 2006. The ictal bradycardia syndrome: Localization and lateralization. Epilepsia 47 (4): 737–744. Broglin, D., and J. Bancaud. 1991. Manifestations neurovegetatives au cours des crises partiÂ� elles du lobe temporal. In: Crises epileptiques et epilepsies du lobe temporal, 69–96. Gentilly: Documentation Medicale Labaz. Burns, S. M., and J. M. Wyss. 1985. The involvement of the anterior cingulate cortex in blood pressure control. Brain Res 340: 71–77. Chefer, S. I., M. I. Talan, and B. T. Engel. 1997. Central neural correlates of learned heart rate control during exercise: Central command demystiἀed. J Appl Physiol 83 (5): 1448–1453. Cohle, S. D., M. P. Suarez-Mier, and B. Aguilera. 2002. Sudden death resulting from lesions of the cardiac conduction system. Am J Forensic Med Pathol 23: 83–89. Commission on Epidemiology and Prognosis, International League Against Epilepsy. 1993. Guidelines for epidemiologic studies on epilepsy. Epilepsia 34 (4): 592–596. Critchley, H. D., C. J. Mathias, O. Josephs, J. O’Doherty, J. Zanini, B. K. Dewar, L. Cipolotti, T. Shallice, and R. J. Dolan. 2003. Human cingulate cortex and autonomic control: Converging neuroimaging and clinical evidence. Brain 126: 2139–2152. Davies, M. J. 1999. The investigation of sudden cardiac death. Histopathology 34: 93–98. de la Grandmaison, G. L. 2006. Is there progress in the autopsy diagnosis of sudden unexpected death in adults? Forensic Sci Int 156: 138–144. Devinsky, O., S. Pacia, and G. Tatambhotla. 1997. Bradycardia and asystole induced by partial seizures: A case report and literature review. Neurology 48: 1712–1714. Druschky, A., M. J. Hiz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, B. Neundorfer, and S. Oppenheimer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by (123I)metaiodobenzylguanidine-SPECT. Brain 124: 2372–2382. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33: 310–316. George, J. R., and G. G. Davis. 1998. Comparison of antiepileptic drug levels in different cases of sudden death. J Forensic Sci 43: 598–603.

Neurogenic Cardiac Arrhythmias

249

Handforth, A., C. M. DeGiorgio, S. C. Schachter, B. M. Uthman, D. K. Naritoku, E. S. Tecoma, T. R. Henry, et al. 1998. Vagus nerve stimulation therapy for partial onset seizures: a randomized, active control trial. Neurology 51: 48–55. Healy, B., and J. Peck. 1997. Bradycardia induced from stimulation of the left versus right central nucleus of the amygdala. Epilepsy Res 28: 101–104. Hennessy, M. J., M. G. Tighe, C. D. Binnie, and L. Nashef. 2001. Sudden withdrawal of carbamaÂ� zepine increases cardiac sympathetic activity in sleep. Neurology 57 (9): 1650–1654. Hiroso, S., A. Mitsudome, M. Okada, and S. Kaneko. 2005. Genetics of idiopathic epilepsies. Epilepsia 46 (suppl. 1): 38–43. Hughes, J. D. and S. Sato. 2010. Sudden death in epilepsy: Relationship to the sleep–wake circadian cycle and fractal physiology. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 23. Boca Raton: CRC€Press. Isojarvi, J. I., H. Ansakorpi, K. Suominen, U. Tolonen, M. Repo, and V. V. Myllyla. 1998. Interictal cardiovascular autonomic responses in patients with epilepsy. Epilepsia 39: 420–426. James, W. 1894. Physical basis of emotion. Psychol Rev I: 516–529. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. A comprehensive review of the literature and proposed mechanisms. Acta Neurol Scand Suppl 82: 1–66. Jick, S. S., T. B. Cole, A. Mesher, and P. J. H. Tennis. 1992. Sudden unexplained death in young persons with primary epilepsy. Pharmacoepidemiol Drug Saf 1: 59–64. Kaada, B. R. 1951. Somatomotor, autonomic and electrocorticographic responses to electrical stimulation of rhinencephalic and other structures in primates, cat and dog: A study of responses from the limbic, subcallosal, orbitoinsular, piriform and temporal cortex, hippocampus fornix and amygdala. Acta Physiol Scandinav Suppl 24 (83): 1–262. Kahane, P., M. Di Leo, D. Hoffmann, and C. Munari. 1999. Ictal bradycardia in a patient with a hypothalamic hamartoma: A stereo-EEG study. Epilepsia 40: 522–527. Kenneback, G., L. Bergfeldt, T. Tomson, E. Spina, and O. Edhag. 1992. Carbamazepine induced bradycardia—A problem in general or only in susceptible patients? A 24 h long term electrocardiogram study. Epilepsy Res 13 (2): 141–145. Kenneback, G., M. Ericson, T. Tomson, and L. Bergfeldt. 1997. Changes in arrhythmia proἀle and heart rate variability during abrupt withdrawal of antiepileptic drugs. Implications for sudden death. Seizure 6 (5): 369–375. Kiok, M. C., C. F. Terrence, G. H. Fromm, and S. Lavine. 1986. Sinus arrest in epilepsy. Neurology 36 (1): 115–116 Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurourg Psychiatry 67: 439–444. Langan, Y., L. Nashef, and J. W. A. S. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Langan, Y., and J. W. A. S. Sander. 1999. Patients with epilepsy who die suddenly do not always have cardiac disease. Arch Neurol 56 (2): 249. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and non-uniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Non-uniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 2010. Antiepileptic drugs beneἀt/risk clinical pharmacology: Possible role in cause and/or prevention of SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press.

250 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., F. L. Weiner, and P. L. Schraeder. 1983. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lock-step phenomenon. Circ Res 31: 630A. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earn. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. Leung, H., K. Schindler, P. Kwan, and C. Elger. 2007. Asystole induced by electrical stimulation of the left cingulated gyrus. Epileptic Disord 9 (1): 77–81. Lhatoo, S. D., and J. W. Sander. 2002. Sudden unexpected death in epilepsy. Hong Kong Med J 8 (5): 354–358. Manitius-Robeck, S., P. Schuler, H. Feistel, G. Platsch, and H. Stefan. 1998. Ictal syncopes. Cardiac sympathetic innervation disorder as the etiology? (German) Der Nervenarzt 69: 712–716. Matthews, S. C., M. P. Paulus, A. N. Simmons, R. A. Nelesen, and J. E. Dimsdale. 2004. Functional subdivisions within anterior cingulate cortex and their relationship to autonomic nervous system function. NeuroImage 22: 1151–1156. McGregor, A., J. Wheless, and J. Baumgartner. 2005. Right sided vagus nerve stimulation as a treatment for refractory epilepsy in humans. Epilepsia 46 (1): 91–96. McKee, J. R., and J. W. Bodἀsh. 2000. Sudden unexpected death in epilepsy in adults with mental retardation. Am J Ment Retard 105: 235. Mufson, E. J., and M. M. Mesulam. 1982. Insula of the old world monkey: Part II. Afferent cortical input and comments on function. J Comp Neurol 212: 23–37. Munari, C., L. Tassi, and M. Di Leo, P. Kahane, D. Hoffman, S. Francione, and P. Quarato. 1995. Video-stereo-electroencephalographic investigation of orbitofrontal cortex: Ictal electroclinical patterns. In Epilepsy and the Functional Anatomy of the Frontal Lobe, ed. H. H. Jasper, S. Riggio, and P. S. Goldman-Rakic, 273–795. New York, NY: Raven Press. Nashef, L. 1997. Sudden unexpected death in epilepsy: Terminology and deἀnitions. Epilepsia 38: S6–S8. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300. Nashef, L., S. Garner, J. Sander, D. Fish, and S. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64 (3): 349–352. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 13: 888–893. Nei, M., R. T. Ho, B. W. Abdou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45(4): 338–345. Ongur, D., X. An, and J. L. Price. 1998. Prefrontal cortical projections to the hypothalamus in macaque monkeys. J Comp Neurol 401: 480–505. Opeskin, K., and S. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12: 456–464. Opeskin, K., A. S. Harvey, S. M. Cordner, and S. F. Berkovic. 2000. Sudden unexpected death in epilepsy in Victoria. J Clin Neurosci 7: 34–37. Oppenheimer, M. B., A. Gelb, J. P. Girvin, and V. C. Hachinski. 1992. Cardiovascular effects of human insular cortex stimulation. Neurology 42: 1727–1732. O’Rourke, D., and C. M. Lathers. 2010. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 31. Boca Raton: CRC Press. Penἀeld, W., and H. Jasper. 1954. Epilepsy and the Functional Anatomy of the Human Brain. London: J. and A. Churchill Ltd (whole chapter). Persson, H., M. Ericson, and T. Tomson. 2003. Carbamazepine affects autonomic cardiac control in patients with newly diagnosed epilepsy. Epilepsy Res 57 (1): 69–75.

Neurogenic Cardiac Arrhythmias

251

Pool, J. L., and J. Ransohoff. 1949. Autonomic effects on stimulating the rostral portion of the cingulate gyri in man. J Neurophysiol 12: 385–392. Reeves, A. L., K. E. Nollet, D. W. Klass. F. W. Sharbrough, and E. L. So. 1996. The ictal bradycardia syndrome. Epilepsia 37: 983–987. Rocamora, R., M. Kurthen, L. Lickfett, J. von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44 (2): 179–185. Rossetti, A. O., B. A. Dworetzky, J. R. Madsen, O. Golub, J. A. Beckman, and B. E. Broomἀeld. 2005. Ictal asystole with convulsive syncope mimicking secondary generalization: A depth electrode study. J Neurol Neurosurg Psychiatry 76: 885–887. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364: 2212–2219. Russell, A. E. 1906. Cessation of the pulse during the onset of epileptic ἀts. Lancet 2: 152–154. Schnabel, R., M. Beblo, and T. W. May. 2000. Is geomagnetic activity a risk factor for sudden unexplained death in epilepsies? Neurology 54: 903–908. Schuele, S. U., A. C. Bermeo, A. V. Alexopoulos, E. R. Locatelli, R. C. Burgess, and N. FoldvarySchaefer. 2007. Video-electrographic and clinical features in patients with ictal asystole. Neurology 69 (5): 434–441. Shields, L. B., D. M. Hunsaker, J. C. Hunsaker 3rd, and J. C. Parker Jr. 2002. Sudden unexpected death in epilepsy: Neuropathologic ἀndings. Am J Forensic Med Pathol 23: 307–310. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72: 340–345. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1990. Relationship of the lock-step phenomenon and precipitous changes in blood pressure. In Epilepsy and Sudden Death, ed. C. M. Lathers and P.€L. Schraeder, 221–238. New York, NY: Marcel Dekker. Stöllberger, C., and J. Finsterer. 2004. Cardiorespiratory ἀndings in sudden unexplained/unexpected death in epilepsy (SUDEP). Epilepsy Res 59: 51–60. Stöllberger, C., and J. Finsterer. 2010. Cardiac and pulmonary risk factors and pathomechanisms of sudden unexplained death in epilepsy patients (SUDEP). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 42. New York, NY: Taylor & Francis. Tellez-Zenteno, J. F., L. H. Hernandez Ronquillo, and S. Wiebe. 2005. Sudden unexpected death in€epilepsy: Evidence based analysis of incidence and risk factors. Epilepsy Res 65: 101–115. Tigaran, S. 2002. Cardiac abnormalities in patients with refractory epilepsy. Acta Neurol Scand Suppl 177: 9–32. Timmings, P. L. 1993. Sudden unexpected death in epilepsy: A local audit. Seizure 2: 287–290. Timmings, P. L. 1998. Sudden unexpected death in epilepsy: Is carbamazepine implicated? Seizure 7: 289–291. Tinuper, P., F. Bisulli, R. Cerullu, R. Carcangiu, C. Marini, G. Pierangeli, and P. Cortelli. 2001. Ictal bradycardia in partial epileptic seizures. Autonomic investigation in three cases and literature review. Brain 124: 2361–2371. Tomson, T. 2010. Compliance with antiepileptic drug treatment and the risk of sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C.€M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 51. Boca Raton: CRC Press. Van Buren, J. M. 1961. Sensory, motor and autonomic effects of mesial temporal stimulation in man. J Neurosurg 18: 273–288. Vilensky, J. A., and G. W. van Hoesen. 1981. Corticopontine projections from the cingulate cortex in the rhesus monkey. Brain Res 205: 391–395. Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. C. Gascino, J. F. Anngers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56: 519–525. Ward, A. A. 1948. The cingular gyrus: Area 24. J Neurophysiol 11: 13–34.

252 Sudden Death in Epilepsy: Forensic and Clinical Issues Woodruff, B. K., J. W. Britton, S. Tigaran, G. D. Gascino, M. F. Burritt, J. P. McConnell, J. Ravkilde et al. 2003. Cardiac troponin levels following monitored epileptic seizures. Neurology 60: 1690–1692. Zamrini, E. Y., K. J. Meador, D. W. Loring, F. T. Nichols, G. P. Lee, W. O. Tomson. 1990. Unilateral cerebral inactivation produces differential left/right heart rate responses. Neurology 40: 1408–1411.

17

Stress and SUDEP Claire M. Lathers Paul L. Schraeder

Contents 17.1 Introduction 17.2 Stress-Related Risk Factors 17.3 Positive Life Events as a Stress: A Case Report of SUDEP References

253 254 260 262

17.1â•…Introduction Heart–brain interactions function during normal daily routine actions and during times of stress. The occurrence of stress itself is a powerful change initiator and may trigger transient ischemia and acute coronary syndrome in some persons (Pickworth et al. 1990; Lathers and Schraeder 2006; Lathers et al. 2008; Soufer and Burg 2007). These individuals are at increased risk for recurrent cardiac events and early death. Psychosocial stress can become an acute trigger of myocardial infarction in patients with preexisting coronary artery disease. Stress, via actions on the central and autonomic nervous systems, may produce a cascade of physiologic responses in individuals at risk that may lead to myocardial ischemia, ventricular ἀbrillation, plaque rupture, or coronary thrombosis (Krantz et al. 1996). Use of simultaneous single-photon emission computed tomography imaging with technetium99m tetrofosmin myocardial perfusion imaging and transthoracic echocardiography was done at rest and during mental stress induced in patients with stable coronary artery disease (Shah et al. 2006). It was concluded that C-reactive protein levels may be a risk marker for mental stress-induced myocardial ischemia. The role of C-reactive protein levels in stress associated with interictal and ictal seizure activity and the development of cardiac arrhythmias and/or death is unknown and should be examined. Nevertheless, today we still do not understand the pathophysiology of mental stress-induced ischemia, what diagnostic tests are needed to identify susceptible persons, nor how to develop risk stratiἀcation algorithms to be applied in the clinical workplace. Research is needed to understand the brain–heart relationship during the occurrence of mental stress that underlies the cognitive and emotional aspects of mental stress as the distinct patterns of brain activity occurring during mental stress trigger silent myocardial ischemia (Soufer and Burg 2007). Cardiac function itself may be adversely changed during an episode of acute emotional stress (Ziegelstein 2007). Left ventricular contractile dysfunction, myocardial ischemia, and/or cardiac arrhythmias have been demonstrated to be triggered by acute emotional stress. These events may be transient but have damaging and/or fatal consequences. The understanding of the anatomical substrate and physiological pathways involved in this heart–brain interaction are not clear, but new data obtained with functional neuroimaging suggest asymmetric brain activity are important in making the heart more susceptible 253

254 Sudden Death in Epilepsy: Forensic and Clinical Issues

to ventricular arrhythmias. Lateralization of cerebral activity during emotional stress may stimulate the heart asymmetrically and produce areas of inhomogeneous repolarization that create electrical instability and facilitate development of cardiac arrhythmias. Ziegelstein (2007) states that patients with ischemic heart disease, who do survive an episode of sudden cardiac death in the setting of acute emotional stress, should be treated with a beta blocker. Nonpharmacologic methods to manage the effects of stress in persons with or without coronary artery disease include social support, relaxation therapy, yoga, meditation, controlled slow breathing, and biofeedback. Clues for effective treatment of mental and emotional stress associated with heart– brain interactions exist. As early as 1993, the pathophysiologic effects of mental stress appear to involve alterations in both myocardial oxygen demand and supply (Merz et al. 1993). Intense negative emotion, such as hostility, and heightened cardiovascular reactivity are associated with occurrence of this type of ischemia. Thus, if persons at risk are taught to recognize these factors, interventional training may help protect them from unwanted consequences. Using traditional anti-ischemic therapy, such as beta blockers and vasodilators, has been shown to reduce mental stress–triggered ischemia in coronary artery disease. Both behavioral and psychosocial interventions, such as decreasing environmental stress via use of social support, alteration of stress perception by behavioral training, and altered physiologic reaction to stress through physical training were discussed as therapeutic options. Posttraumatic stress disorder, a psychiatric disorder that develops after a psychological trauma generally triggered by a situation perceived by the person experiencing the event as one that deeply threatens his/her life or integrity, is thought to be triggered by complex neurobiological changes. Kozaric-Kovacic (2008) reports that selective serotonin-reuptake inhibitors are the ἀrst line of treatment for posttraumatic stress disorders. These agents are more effective than noradrenalin-reuptake inhibitors or tricyclic antidepressants. Antipsychotic drugs, especially the atypical ones, are effective in posttraumatic stress disorder patients with psychotic characteristics or refractoriness to other drugs, reducing the overall overreaction to stress. Monoamine oxidase inhibitors have not been clearly identiἀed as beneἀcial. Serotonin agonists and antagonists, new antidepressants that are dual inhibitors of serotonin and noradrenaline reuptake, anticonvulsants, and opiate antagonists may be used. Additional rigorous clinical trials are needed to establish use, efficacy, tolerability, and safety in treating this pharmacotherapeutic disorder. The occurrence of acute postictal psychiatric symptoms are well recognized, but, fortunately, relatively uncommon, and most likely manifest after partial complex seizures. These symptoms may mimic anxiety, depression, or an acute psychotic disorder (Kanner et al. 1996). Violent behavior can also be an uncommon postictal manifestation. When such episodes of€ violence occur, they are not well organized, planned, or speciἀcally directed. Directed aggression is rare, but when postictal aggressive behavior—whether undirected or seemingly directed—does occur, the patient can be at risk for an aggressive police response including total body restraints. Mendez (1998) described a patient with directed postictal aggression.

17.2â•…Stress-Related Risk Factors The mechanisms involved in emotion-associated seizure activation are not known but multiple factors may explain the role of emotion. These factors include activation of neural networks, sleep deprivation, noncompliance, alcohol use, and hyperventilation. Increased

Stress and SUDEP

255

risk of seizure recurrence is also associated with use of antidepressants and/or neuroleptics, drugs that may lower the seizure threshold. Poor seizure control is a major risk factor for SUDEP (Dominian et al. 1963; Mattson 1991; Nashef et al. 1995; Sperling et al. 1999). Acute stress caused by emotions such as fear may cause sudden death in persons with coronary heart disease. Chronic stress also contributes to development of long-term coronary disease (Ackerman et al. 2001). Stress management in standard cardiac rehabilitation programs is important. The distinction between type A versus type B personalities as risk correlates is overly simplistic. Psychosocial factors of depression, hostility, social isolation, anxiety, anger, and other stresses are related to increased cardiac death and illness in all groups with coronary heart disease (Buselli and Stuart 1999) and these psychosocial risk factors may beneἀt from a biopsychosocial model of intervention. Death can result from acute stress related to disasters. Sudden deaths related to atherosclerotic coronary artery disease increased ἀvefold on the day of the Los Angeles earthquake in 1994 (Leor et al. 1996). Acute mental stress induced experimentally may be associated with ST-segment deviation and wall-

Heart –Brain Int erplay : A 53-Year-Old Wo man Reco v ering fro m Mit ral Valv e Repair Anxiety, panic, fatigue, and depression can have a profound adverse effect on physical well-being. This case illustrates how these adverse psychological symptoms interfered with recovery from mitral valve surgery. A 53-year-old woman had an uncomplicated mitral valve repair, after which she refused to use the incentive spirometer, ambulate, or even sit in a chair. She experienced unexplained episodes of shortness of breath and tachycardia. At 4 weeks after operation, the ejection fraction, which was 50% pre-op, remained at 40%. She was complaining of more post-op pain and fatigue than was usual for this type of successful surgery. The patient experienced tearful episodes, and reluctantly was interviewed by a psychiatrist. While the initial diagnosis was adjustment disorder with anxious features, subsequent evaluation established a diagnosis of depression. Ultimately, the patient agreed to take an SSRI, citalopram. After several weeks of treatment, she had a profound improvement in her mental and physical states, with she and her family concurring that she had returned to her “old self” (Callahan et al. 2008). Discussio n This case illustrates how postoperative cardiac rehabilitation can be signiἀcantly affected by an adverse psychological state. Although the pathophysiological mechanism of sudden death in epilepsy (SUDEP) is not established, acute cardiac dysfunction is thought to be a major factor. Cardiologists have long known about the physical risks of adverse mental states in cardiac patients including unexpected sudden cardiac death. The role of stress and other psychological disturbances require consideration as possible risk factors associated with SUDEP. However, in contrast to the interest cardiologists have manifested in psychogenic risks associated with cardiac sudden death, little research has been undertaken on the subject of psychological issues relative to the risk of SUDEP.

256 Sudden Death in Epilepsy: Forensic and Clinical Issues

motion abnormalities (Rozanski et al. 1988). Mental stress may be a greater risk factor than eÂ�xercise-induced ischemia in increasing the rate of fatal and nonfatal cardiac events (Jiang et al. 1996). Psychosocial stress management treatment in a cardiac rehabilitation program does reduce cardiac related mortality and morbidity (Linden et al. 1996). Distress at the family level and/or stress at work are strong predictors for developing stress-related disorders and need intervention (Anderberg 2001). Management of stress factors in coronary artery disease may also help to decrease the risk of sudden death. While depression and stress are major risk factors in sudden cardiac death, it is not known if the same risks apply to SUDEP. Having epilepsy in and of itself is stress-producing and stress does increase the frequency of seizures. The uncertainty of when a seizure can occur, the consequences of having a seizure on employment status and driving privileges are stress-producing circumstances. Both depression and anxiety are symptoms associated with epilepsy (Trimble and Perez 1980; Blumer 1992). Earnest et al. (1992) suggested there was a role for acutely stressful circumstances as a possible contributor near the time of death in a case control study of the metropolitan Denver area. The strong interest of cardiologists in adverse emotional states as a risk of sudden cardiac death implies that there is a need to investigate this issue more thoroughly in persons with epilepsy. Activation of stress-responsive systems during depressive episodes may contribute to metabolic risk factors and imbalance of the autonomic heart regulation (Blumer 1992; Linden et al. 1996; Deuschle and Lederbogen 2002). Clearly, cerebral activity can have a profound effect upon the autonomic regulation of cardiac function. Lathers et al. (1977, 1978) found that nonuniform postganglionic cardiac sympathetic neural discharge is capable of triggering cardiac arrhythmias in the manner described by Han and Moe (1964) (i.e., nonuniform cardiac repolarization). These authors noted that the aberrant nonuniform neural discharge was associated with cardiac arrhythmias triggered by abrupt coronary occlusion of the left anterior descending coronary artery to produce cardiac ischemia mimicking events occurring in the sudden death heart attack victim. The nonuniform postganglionic cardiac sympathetic neural activity also occurs with cardiac changes associated with ouabain-induced toxicity characterized by cardiac arrhythmias and death. In both animal models the potentially damaging role of adrenal catecholamines in the production of arrhythmias and death was discussed. Synchronization of brain electrical activity with both cardiac sympathetic and vagal neural discharge was also identiἀed by those working in Dr. Lathers’ laboratory using the cat model (Lathers et al. 1987; O’Rourke and Lathers 1990; Dodd-O and Lathers 1990; Stauffer et al. 1989; Lathers et al. 2010). This autonomic neural discharge synchronized with cerebral ictal and interictal discharges was termed the lockstep phenomenon and was hypothesized to be one mechanism contributing to the development of cardiac arrhythmias and/or sudden death associated with both interictal and ictal discharges in persons with epilepsy found dead in a sudden, unexpected manner. Subsequently, Davis and Natelson (1993) focused on brain–heart interactions and the neurocardiology of arrhythmia and sudden cardiac death, with an emphasis on the nervous system direction of the events leading to cardiac damage associated with raising catecholamine levels in experimental and clinical entities of stroke, epilepsy, and environmental stress. Autonomic sympathetic and parasympathetic cardiac neuronal dysfunctions are associated with interictal as well as ictal epileptiform discharges and with cardiac arrhythmias (Lathers and Schraeder 1982). Both types of epileptogenic activity are associated with temporal lobe epilepsy, autonomic dysregulation, and predominant sympathetic overactivity

Stress and SUDEP

257

Case Repo rt o f Po st ict al Ag g ressi v e Behav io r A 37-year-old man had complex partial seizures, consisting of an olfactory aura followed by alteration of consciousness, since age ἀve, with occasional tonic–clonic seizures. He also had a history of depression and bipolar affective disorder. During the postictal period, at times he would feel a sense of being threatened and having been harmed, and would focus his aggressive response by attacking any individual in his immediate environment, often resulting in physical injury to that person. The patient’s postictal confusion would remit usually after one hour, but the fear of harm and sense of being threatened would last for 24 hours, after which he felt great remorse. He had been charged with aggravated assault several times. His MRI was unremarkable, but sleep-deprived EEG manifested left anterior temporal spikes and sharp waves. The aggressive postictal episodes ended with control of his seizures resulting from the addition of carbamazepine to his original regimen of valproic acid and sertraline. Discussio n by M endez The author describes aggressive acts as direct consequence of seizures, occurring directly during seizures or postictally (Mendez 1998). He also emphasizes that postÂ� ictal violence is most commonly resistive behavior during the postictal delirium and is associated with attempts at restraint. He also discussed the observation that seemingly violent automatisms such as flailing or spitting can occur during complex partial seizures, and that secondary violent automatisms can be behavioral responses to ictal fear, hallucinations, or other disagreeable seizure related experiences. Discussio n by L at hers and Schraeder Another issue to consider in relationship to the occurrence of putative ictal or postictal violent behavior is that of the possible induction of an excited delirium syndrome when attempts are made to restrain the behavior of an individual who is manifesting actual or seemingly violent behavior. As deἀned by DiMaio and DiMaio (2006), excited delirium syndrome “involves the sudden death of an individual, during or following an episode of excited delirium, in which an autopsy fails to reveal evidence of sufficient trauma or natural disease to explain the death. In virtually all cases, the episode of excited delirium is terminated by a struggle with police or medical personnel, and the use of physical restraint. Typically, within a few to several minutes following cessation of the struggle, the individual is noted to be in cardiopulmonary arrest. Attempts at resuscitation are usually unsuccessful.” During the state of delirium, there are varying transient disturbances of consciousness and cognition with disorientation, disorganized and inconsistent thought processes, inability to distinguish reality from hallucinations, speech disturbance, and disorientation to time, place, and person. Deaths occur most commonly in individuals who have abused stimulants such as cocaine and methamphetamine, but also in persons with endogenous mental disease who have not used these drugs. The majority of deaths occur between the ages of 17 and 35. Although the mechanism of death in these individuals is not deἀned, DiMaio and DiMaio (2006) concluded that stimulation of the sympathetic nervous system causes release of norepinephrine at the synapses and in combination

258 Sudden Death in Epilepsy: Forensic and Clinical Issues

with epinephrine into the bloodstream from the adrenals. This response then results in subsequent increase in myocyte activity and oxygen demand in combination with decreased myocardial blood flow secondary to coronary artery constriction. We need to be aware of the potential for induction of this potentially fatal state of agitation when attempts are made to restrain persons with epilepsy who manifest ictal or postictal agitation or seemingly violent behavior. This is a highly stressful state and in combination with the history of epilepsy could be contributory to the occurrence of SUDEP in these individuals. Since both ictal and postictal states are self limited, it is imperative for family members, police, and emergency care personnel to understand that watchful observation to keep the affected individual out of harm’s way, rather than the high-risk intervention of physical restraint, is the most appropriate intervention.

(Hilz et al. 2002). Surgical treatment for temporal lobe epilepsy reduced sympathetic cardiomodulation and decreased baroreflex sensitivity (i.e., decreased the influence of sympathetically mediated tachyarrhythmias and excessive bradycardiac counterregulation). These factors are thought to contribute to the risk of SUDEP and thus the temporal lobe surgery itself appears to be one method to reduce and/or eliminate some risk factors associated with SUDEP (Hilz et al. 2002; Burgerman et al. 1995). Parasympathetic nervous system activity also regulates cardiac rhythm (Richter 1957; Talman 1985). Stimulation of the vagus nerve or application of acetylcholine to the SA node slows or abolishes depolarization of sinus node ἀbers (Schwartz et al. 1976; West et al. 1956). Decreased depolarization and shifts of membrane threshold potentials change the SA node rate, causing sinus bradycardia. A similar mechanism occurs in the AV node (Kralios and Millar 1981). ECG changes and cardiac muscle necrosis result from stimulation of the efferent limb of the sympathetic nervous system and by stimulation of the aortic arch and carotid baroreceptors regulated by the autonomic nervous system reflex activity (Pavlov 1951; Samuels 1997; Zavodskaya et al. 1980; Natelson et al. 1998). Sympathetic stimulation causes a sudden release of norepinephrine from cardiac nerve endings into the heart muscle. This leads to microscopic changes in the form of cardiac myocyte necrosis, cardiac dysfunction, and arrhythmias (Schwartz et al. 1976; West et al. 1956). Stimulation of cardiac sympathetic nerves accelerates sinoatrial depolarizations and shortens the cycle length of ἀring of the sinus node. Natelson et al. (1998) found pathologic changes in the form of irreversible perivascular and interstitial ἀbrosis and myocyte vacuolization in hearts of persons with epilepsy who died suddenly. These lesions occurred mostly in the subendocardium. Thus, it is possible they also resulted from sympathetic nerve catecholamine release with consequent cardiac arrhythmias and repolarization changes that predispose a patient to a form of cardiac damage known as myoἀbrillar degeneration or contraction band necrosis and possible sudden death. This lesion is associated with four types of etiologies: stress plus or minus steroids, catecholamine infusion, nervous system stimulation, and reperfusion. Sympathetic overactivity, with secondary catecholamine toxicity, is the common factor in all four etiologies. Samuels (1997) concludes that all forms of sudden death are based on the anatomic connection between the nervous system and the heart and lungs.

Stress and SUDEP

259

Exposure of rats to stress initially resulted in structural changes of stereotypical cardiac contraction bands, regardless of the type of stress factor, and differed only in severity. At a later stage, contractures were gradually replaced by cytolytic injuries and also did not depend on the type of stress. In the case of early predominance of myocytolysis in combination with excessive contracture injuries that led to rapid death, a genetically determined predisposition was proposed as an explanation (Zavodskaya et al. 1980). Since emotional stressors impact on the autonomic nervous system, this ensures that psychogenic factors are important in leading to cardiac dysrhythmias and to coronary and noncoronary sudden death syndromes in humans and other species (Bohus and Korte 2000). There is a causal relationship among depression, stress, and increased cardiac mortality. Depression may be a state of prolonged negative arousal or mental stress associated with a measurably higher risk of fatal cardiac events. Stress may also result in histopathological changes in a previously normal heart while genetically determined and/or acquired dysfunction of the opioidegic, GABAegic, cholinergic, adenosinergic, and other transmitter/modulator systems may interact to predispose to arrhythmias and sudden death (Tulner and den Boer 2000). The influence of psychological factors on autonomic, neuroendocrine, and immune systems is complex and exerts adverse effects on cardiac function. The amygdala is involved in integration of autonomic responses to emotional stimuli (Cechetto 2000). There is a common pattern of sympathetic representation in the medial prefrontal cortex, insular cortex, ventromedial temporal lobe, and ventral hippocampal region (Westerhaus and Loewy 2001). The ventromedial temporal lobe regions studied included the central, basomedial posterior, and lateral amygdaloid transition area and posterior medial cortical amygdaloid nucleus. This latter anatomical substrate for sympathetic control is of considerable theoretical importance. The amygdala has an important role in expression of emotional behaviors while it integrates autonomic responses to emotional stimuli under conditions of fear and anxiety. The amygdala is important in cardiovascular control within the limbic system, having reciprocal connections with the insular cortex and direct projections to other autonomic control centers in the hypothalamus, pons, and medulla. Other important brain–heart connections exist. Limbic cortex activity associated with an emotionally charged stimulus occurs with cardiac neural changes, resulting in intense autonomic stimulation of both sympathetic and parasympathetic neurons. This, in turn, may result in sudden stress-related death. It is assumed that the hypothalamic and brainstem structures are involved secondarily and that stress activates the hypothalamic–pituitary– adrenal axis. Prolactin levels are elevated after seizures and are used as an indicator that a seizure has actually occurred. The prolactin level is most consistently elevated after generalized tonic–clonic and complex partial seizures, and less predictably after a simple partial seizure. Frontal lobe, absence, myoclonic, and akinetic seizures are not generally associated with elevation of the prolactin level (Pritchard 1997). This observation raised the possibility of the frequent occurrence of a stress response during and after a seizure as a risk factor for SUDEP. An alternative explanation is that the hypothalamic-pituitary axis is directly simulated by the epileptiform activity. In either case, the occurrence of acute autonomic and/or neuroendocrine dysfunction appears to put the patient at risk. Elevated prolactin levels at necropsy were examined as a marker of antemortem stress (Jones and Hallworth 1999), but postmortem prolactin values differed according to the cause of death, with higher values in postoperative deaths and in the chronically ill.

260 Sudden Death in Epilepsy: Forensic and Clinical Issues

In the study of Wannamaker and Booker (1998), patients with epilepsy have identiἀed the common stressors of fear, worry, frustration, and anger as trigger factors. The seizure event usually does not occur immediately in association with the stressor. Animal models suggest that injured populations of neurons surround and interact with an epileptogenic focus. This may cause the focus to function independently when heightened brain excitability triggers neuronal activity in this network. As discussed by Homan (1998), patients often associate increased stress with increased seizure activity. Interventions to reduce stress include the use of behavioral techniques such as biofeedback, relaxation, and desensitization, and pharmacological agents such as psychotherapeutic or benzodiazepine drugs. Physical stress associated with elevated body temperatures induced by infection also triggers seizures and antipyretic drugs are recommended for prophylactic use. Adjunctive management of the patient with seizure and stress results in improvement in the control of seizures and overall quality of life (Moffett and Scott 1984; Fenwick 1995). Education of patients about the importance of drug compliance and stress management techniques ultimately will decrease the need for high therapeutic antiepileptic drug levels while decreasing the occurrence of dose-related side effects. This will also result in an improved lifestyle. Physicians must individualize recommendations for intervention and use referrals to a clinical psychologist with expertise in stress reduction. Both internal and external stressors must be decreased. Treatment with tranquilizers, antidepressants, or neuroleptics is used with counseling. The long-acting anxiolytics clorazepate and clonazepam are relatively effective as antiepileptic drugs in selected patients. Paranoia, thought disorder, hallucinations, and extreme agitation require concurrent psychiatric consultation in addition to treatment with neuroleptics. Care must be exerted when managing the patient since a rapid change in the levels of neuoroleptics, high doses, or induction of drowsiness may worsen the occurrence of seizures. Reduction in agitation, thought disorder, and hallucinations usually exerts a calming effect and contributes to lowering the stress level and to restoring the patient’s sense of well-being (Lathers and Schraeder 2006). As can be concluded from the above discussion, most life events that may be implicated as risk factors for sudden death result in psychological responses that are regarded as negative. The following brief case history raises the possibility that, at least in some persons with epilepsy, an intensely positive event may also be a risk factor for SUDEP.

17.3â•…Positive Life Events as a Stress: A Case Report of SUDEP Many diseases that can affect autonomic balance exhibit patterns of temporal variation during circadian, seasonal, reproductive, and life span cycles that remain unexplained. Termination of organisms during senescence, achieved by emergence of autonomic imbalance and other systemic dysfunction has been examined from a Darwinian perspective. This variation in autonomic balance and disease symptoms of epilepsy has yet to be carefully studied. Cardiac neural control may be estimated by frequency domain characterization of R–R interval variations and this technique is a clinical tool to examine the role of autonomic dysfunction in the pathophysiology of sudden cardiac death (Molgaard et al. 1994). Simultaneous examination of the quality of life and changes in heart rate variability of patients immediately after acute myocardial infarction showed that survivors exhibited heart rate variability within the ἀrst 3 days that was signiἀcantly higher than nonsurvivors and had developed a clear circadian pattern after 3 weeks (Kummell et al. 1993). The authors

Stress and SUDEP

261

SUDEP AFT ER RECEIVING V ERY GOO D NEWS An 18-year-old black female high school student had a history of infrequent (less than once yearly) generalized tonic–clonic seizures that were controlled with moderate doses of carbamazepine with therapeutic blood levels. She was an outstanding student, and during her senior year was offered admission to an Ivy League school with a full scholarship. Her social circumstances were modest in that her father worked as a municipal trash collector and she was the ἀrst family member to go to college. Shortly thereafter, her distraught parents notiἀed her neurologist (PLS) that their daughter was found dead in bed. As no postmortem was performed, and no other cause of death was evident, the diagnosis of probable SUDEP was applied. This case may demonstrate that intensely positive surprise life events can produce as much of a stress as negative events. concluded circadian patterns of heart rate and heart rate variability may be assessed meaningfully immediately post acute myocardial infarction and may ἀnd common expression in changes in sympathovagal balance. Future studies of persons with epilepsy who may be at risk for SUDEP could examine sympathovagal balance using 24-hour ambulatory BP monitoring to examine circadian blood pressure variation, power spectral analysis of R–R interval oscillation to measure autonomic function, and measurement of QTc-d and QTc intervals to monitor cardiac depolarization time. These techniques were employed in a 1-year study designed to examine sympathovagal balance, nighttime blood pressure and QT intervals (Esposito et al. 2003). Sustained weight loss eliminated the diastolic night time drop in blood pressure and sympathetic overactivity detected in normotensive obese women. This weight loss may reduce the cardiovascular risk in obese women. One could speculate that as the blood pressure falls during sleep, there may be, in some persons with epilepsy and autonomic dysfunction, an increased sympathetic discharge that results in the triggering of cardiac arrhythmias and/or sudden death. A study of the role of sympathovagal interaction in diurnal variations of QT interval suggested that although change in sympathovagal balance was responsible for diurnal variation in QT interval, the enhanced sympathetic activity in the day was a major determinate of the phenomenon (Murakawa et al. 1992). Examination of heart rate variability circadian patterns and effect on the QT interval dispersion in healthy subjects (Bilan et al. 2005) was studied. Multiple regression analysis revealed relations between mean QTd and R–R as well as mean QTd and high frequency after adjustment for periods, correlations were only observed during morning hours. The authors concluded that sympathovagal balance, as reflected in heart rate variability, and not the tone of both autonomic components that affects QTd variability. These data suggest that in persons with epilepsy, the sympathovagal balance, as reflected in heart rate variability, should be examined for changes in QTd variability as a risk factor for sudden death during both awake and sleep cycles. In Brugada syndrome, ventricular ἀbrillation occurs mainly during sleep, and Brugada ECG signs are intensiἀed by parasympathomimetic drugs (Mizumaki et al. 2004). Spontaneous augmentation of ST elevation in daily life was demonstrated along with an increase in vagal activity. The ST elevation was increased more in those patients with Brugada syndrome related ventricular ἀbrillation than in those without ventricular ἀbrillation under similar vagal tone. It may be that some patients with epilepsy at certain

262 Sudden Death in Epilepsy: Forensic and Clinical Issues

times exhibit a sympathovagal balance that is dominated by the parasympathetic nervous system and these patients may die in asystole. Most of the time, we consider stress to be associated only with adverse or negative circumstances. However, unexpected good news can also be a stress in a positive sense, especially if it is the fulἀllment of a heretofore seemingly out-of-reach quest. An explanation of why such an association could occur is speculative. However, while it is common knowledge that negative live events such as death, divorce, loss of a job, and so forth, can result in an increase in sympathetic parameters such as heart rate and blood pressure, relief from an adverse circumstance or the occurrence of a positive life event can result in more prominence of parasympathetic parameters such as decreased heart rate and blood pressure (see discussion above). One could speculate that in individuals who have the potential for an excessive parasympathetic response, the intervention of epileptiform discharges could augment this tendency to the point of asystole. Thus, when information about a SUDEP victim’s circumstances prior to the demise is solicited, one should consider eliciting the possibility of an intensely positive event as a risk factor.

References Ackerman, M. J., D. J. Tester, and D. J. Driscoll. 2001. Molecular autopsy of sudden unexplained death in the young. Am J Forensic Med Pathol 22 (2): 105–111. Anderberg, U. M. 2001. Stress-related syndromes—Contemporary illnesses. Lakartidningen 98 (51–52): 5860–5863. Basso, C., D. Corrado, and G. Thiene. 1999. Cardiovascular causes of sudden death in young individuals including athletes. Cardiol Rev 7 (3): 127–135. Bilan, A., A. Witczak, R. Palusinski, A. Ignatowicz, and J. Hanzlik. 2005. Circadian rhythm of the QT interval dispersion in healthy subjects. Correlation with heart rate variability circadian pattern. J Electrocardiol 38 (1): 36–42. Blumer, D. 1992. Postictal depression: signiἀcance for treatment of the neurobehavioral disorder of epilepsy. J Epilepsy 5: 214–219. Bohus, B., and S. M. Korte. 2000. Stress, the hypothalamo–pituitary–adrenal system and the heart. In The Nervous System and the Heart, ed. G. Ter Horst, chap. 6. Totowa, NJ: Humana Press. Burgerman, R. S., M. R. Sperling, J. A. French, A. J. Saykin, and M. J. O’Connor. 1995. Comparison of Mesial Versus Neocortical Onset Temporal Lobe Seizures: Neurodiagnostic Findings and Surgical Outcome. New York, NY: Raven Press. Buselli, E. F., and E. M. Stuart. 1999. Influence of psychosocial factors and biopsychosocial interventions on outcomes after myocardial infarction. J Cardiovasc Nurs 13 (3): 60–72. Callahan, T. D., U. Khokhar, L. Pozuelo, and J. B. Young. 2008. Case study in heart–brain interplay: A 53-year-old woman recovering from mitral valve repair. Cleve Clin J Med 75 (Suppl 2): S10–S14. Cechetto, D. 2000. Neuropathology and cardiovascular regulation. In The Nervous System and the Heart, ed. G. Ter Horst, chap. 4. Totowa, NJ: Humana Press. Davis, A. M., and B. H. Natelson. 1993. Brain–heart interactions. The neurocardiology of arrhythmia and sudden cardiac death. Tex Heart Inst J 20 (3): 158–169. Deuschle, M., and F. Lederbogen. 2002. Depression and cardiovascular disease: pathogenetic factors of the stress concept. Fortschr Neurol Psychiatr 70 (5): 268–275. DiMaio, T. G., and V. J. M. DiMaio. 2006. Excited Delirium Syndrome. Boca Raton: CRC Press. Dodd-O, J. H., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in phenobarbital-pretreated cats. In Epilepsy and Sudden Death, chap. 13. New York, NY: Marcel Dekker. Dominian, J., E. A. Serafetinides, and M. Dewhurst. 1963. A follow-up study of late-onset epilepsy: II. Psychiatric and social ἀndings. Br Med J 1 (5328): 431–435.

Stress and SUDEP

263

Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33 (2): 310–316. Elfawal, M. A. 2000. Sudden unexplained death syndrome. Med Sci Law 40 (1): 45–51. Esposito, K., R. Marfella, P. Gualdiero, C. Carusone, A. Pontillo, G. Giugliano, G. Nicoletti, and D. Giugliano. 2003. Sympathovagal balance, nighttime blood pressure, and QT intervals in normotensive obese women. Obes Res 11 (5): 653–659. Fenwick, P. 1995. The basis of behavioral treatment in seizure control. Epilepsia 36: 46–50. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hilz, M. J., O. Devinsky, W. Doyle, A. Mauerer, and M. Dutsch. 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125 (Pt 5): 985–995. Homan, R. W. 1998. Adjunctive and combination therapy. Chapter 114. In Epilepsy. A Comprehensive Textbook, ed. J. J. Engel, T. A. Pedley, J. Aicardi, M. Dichter, U. Heinemann, W. Moshe, R. Porer, and D. Taylor. Philadelphia, PA: Lippincott-Raven. Jiang, W., M. Babyak, D. S. Krantz, R. A. Waugh, R. E. Coleman, M. M. Hanson, D. J. Frid et al. 1996. Mental stress–induced myocardial ischemia and cardiac events. JAMA 275 (21): 1651–1656. Jones, T. J., and M. J. Hallworth. 1999. Postmortem prolactin as a marker of antemortem stress. J Clin Pathol 52 (10): 749–751. Kanner, A. M., S. Stagno, P. Kotagal, and H. H. Morris. 1996. Postictal psychiatric events during prolonged video-electroencephalographic monitoring studies. Arch Neurol 53 (3): 258–263. Kozaric-Kovacic, D. 2008. Psychopharmacotherapy of posttraumatic stress disorder. Croat Med J 49 (4): 459–475. Kralios, F. A., and C. K. Millar. 1981. Sympathetic neural effects on regional atrial recovery properties and cardiac rhythm. Am J Physiol 240 (4): H590–H596. Krantz, D. S., W. J. Kop, H. T. Santiago, and J. S. Gottdiener. 1996. Mental stress as a trigger of myocardial ischemia and infarction. Cardiol Clin 14 (2): 271–287. Kummell, H. C., P. Van Leeuwen, C. Heckmann, P. Engelke, G. Kesting, H. Kremer, and A. Becher. 1993. Quality of life and circadian variation of heart rate and heart rate variability in short-term survivors and nonsurvivors after acute myocardial infarction. Clin Cardiol 16 (11): 776–782. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: the lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., and P. L. Schraeder. 2010. Animal model for sudden unexpected death in persons with epilepsy. Chapter 28 in Sudden Death in Epilepsy: Forensic and Clinical Issues, eds., C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Leor, J., W. K. Poole, and R. A. Kloner. 1996. Sudden cardiac death triggered by an earthquake. N Engl J Med 334 (7): 413–419. Linden, W., C. Stossel, and J. Maurice. 1996. Psychosocial interventions for patients with coronary artery disease: A meta-analysis. Arch Intern Med 156 (7): 745–752. Mattson, R. H. 1991. Emotional effects on seizure occurrence. Adv Neurol 55: 453–460.

264 Sudden Death in Epilepsy: Forensic and Clinical Issues Mendez, M. F. 1998. Postictal violence and epilepsy. Psychosomatics 39 (5): 478–480. Merz, C. N., D. S. Krantz, and A. Rozanski. 1993. Mental stress and myocardial ischemia. Correlates and potential interventions. Tex Heart Inst J 20 (3): 152–157. Mizumaki, K., A. Fujiki, T. Tsuneda, M. Sakabe, K. Nishida, M. Sugao, and H. Inoue. 2004. Vagal activity modulates spontaneous augmentation of ST elevation in the daily life of patients with Brugada syndrome. J Cardiovasc Electrophysiol 15 (6): 667–673. Moffett, A., and D. F. Scott. 1984. Stress and epilepsy: The value of a benzodiazepine–lorazepam. J Neurol Neurosurg Psychiatry 47 (2): 165–167. Molgaard, H., K. Hermansen, and P. Bjerregaard. 1994. Spectral components of short-term RR interval variability in healthy subjects and effects of risk factors. Eur Heart J 15 (9): 1174–1183. Murakawa, Y., H. Inoue, A. Nozaki, and T. Sugimoto. 1992. Role of sympathovagal interaction in diurnal variation of QT interval. Am J Cardiol 69 (4): 339–343. Nashef, L., D. R. Fish, J. W. Sander, and S. D. Shorvon. 1995. Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral centre. J Neurol Neurosurg Psychiatry 58 (4): 462–464. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55 (6): 857–860. O’Rourke, D. K., and C. M. Lathers. 1990. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, chap. 15. New York, NY: Marcel Dekker. Owada, M., Y. Aizawa, K. Kurihara, N. Tanabe, T. Aizaki, and T. Izumi. 1999. Risk factors and triggers of sudden death in the working generation: An autopsy proven case-control study. Tohoku J Exp Med 189 (4): 245–258. Pavlov, I. 1951. The Complete Collection of Works, Vol. 1. Moscow: ANSSR. Pritchard, P. B. 1997. Hormone changes in epilepsy. In Epilepsy. A Comprehensive Textbook, ed. J. J. Engel and T. A. Pedley, chap. 188. Philadelphia, PA: Lippincott-Raven. Pickworth, W. B., J. Gerard-Ciminara, and C. M. Lathers. 1990. Stress, arrhythmias, and seizures. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, chap. 22. New York, NY: Marcel Dekker. Richter, C. P. 1957. On the phenomenon of sudden death in animals and man. Psychosom Med 19 (3): 191–198. Rozanski, A., C. N. Bairey, D. S. Krantz, J. Friedman, K. J. Resser, M. Morell, S. Hilton-Chalfen, L. Hestrin, J. Bietendorf, and D. S. Berman. 1988. Mental stress and the induction of silent myocardial ischemia in patients with coronary artery disease. N Engl J Med 318 (16): 1005–1012. Samuels, M. 1997. Voodoo death revisited: The modern lessons of neurocardiology. The Neurologist 3: 293–304. Schwartz, P. J., H. L. Stone, and A. M. Brown. 1976. Effects of unilateral stellate ganglion blockade on the arrhythmias associated with coronary occlusion. Am Heart J 92 (5): 589–599. Selye, H. 1958. The Chemcal Prevention of Cardiac Necrosis. New York, NY: Ronald Press. Shah, R., M. M. Burg, A. Vashist, D. Collins, J. Liu, F. Jadbabaie, B. Graeber, C. Earley, R. Lampert, and R. Soufer. 2006. C-reactive protein and vulnerability to mental stress-induced myocardial ischemia. Mol Med 12 (11–12): 269–274. Soufer, R., and M. M. Burg. 2007. The heart–brain interaction during emotionally provoked myocardial ischemia: Implications of cortical hyperactivation in CAD and gender interactions. Cleve Clin J Med 74 Suppl 1: S59–S62. Sperling, M. R., H. Feldman, J. Kinman, J. D. Liporace, and M. J. O’Connor. 1999. Seizure control and mortality in epilepsy. Ann Neurol 46 (1): 45–50. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345.

Stress and SUDEP

265

Talman, W. T. 1985. Cardiovascular regulation and lesions of the central nervous system. Ann Neurol 18 (1): 1–13. Trimble, M. R., and M. M. Perez. 1980. Quantiἀcation of psychopathology in adult patients with epilepsy. In Epilepsy and Behavior, ed. B. Kulig, H. Meinardi, and G. Stores. Lisse: Swets & Zeitlinger. Tulner, D. M., and H. A. den Boer. 2000. Heart disease and cognitive/neuropsychiatric disorders. In The Nervous System and the Heart, ed. G. Ter Horst, chap. 15. Totowa, NJ: Humana Press. Wannamaker, B. B., and H. E. Booker. 1998. Treatment of provoked seizures. In Epilepsy. A Comprehensive Textbook, ed. Engel, T. A. Pedley, J. Aicardi, M. Dichter, U. Heinemann, S. Moshe, R. Porer, and D. Taylor, chap. 119. Philadelphia, PA: Lippincott-Raven. West, T. C., G. Falk, and P. Cervoni. 1956. Drug alteration of transmembrane potentials in atrial pacemaker cells. J Pharmacol Exp Ther 117 (2): 245–252. Westerhaus, M. J., and A. D. Loewy. 2001. Central representation of the sympathetic nervous system in the cerebral cortex. Brain Res 903 (1–2): 117–127. Zavodskaya, I. S., E. V. Moreua, and N. A. Novikova. 1980. Neurogenic Heart Lesions. Oxford: Pergamon. Ziegelstein, R. C. 2007. Acute emotional stress and cardiac arrhythmias. JAMA 298 (3): 324–329.

18

Genetics of Sudden Death in Epilepsy Neeti Ghali Lina Nashef

Contents 18.1 The Genetics of Epilepsy 18.2 Mortality in Speciἀc Syndromes 18.2.1 15q Inversion/Duplicaton [inv dup(15)] 18.2.2 Rett Syndrome 18.2.3 Other Examples of Syndromes with Multisystem Involvement 18.2.4 SCN1A-Related Disorders 18.3 SUDEP and Idiopathic Epilepsy 18.4 SUDEP in Genetic Epilepsies: How Does It Occur? 18.5 Genetic Susceptibility to SCD and SIDS 18.6 Conclusion References

267 269 269 270 270 271 272 275 276 278 278

18.1â•… The Genetics of Epilepsy Idiopathic epilepsies account for a substantial proportion of all epilepsies and are considered to be largely genetically determined (Steinlein 2008). The inheritance of these epilepsies may be Mendelian, whereby a single identiἀed mutation results in epilepsy and/or febrile seizures. However, there is often interfamilial and intrafamilial variability in the clinical phenotype, suggesting an effect of other genetic variants. The inheritance is more often non-Mendelian or complex, whereby the phenotype is thought to be determined by several more minor genetic defects as well as environmental effects. Epilepsies inherited in a complex or multifactorial manner will arise when a chance combination of certain susceptibility alleles come together with sufficient effect in the individual to push neuronal hyperexcitability over the seizure threshold (Mulley et al. 2005), but where each susceptibility allele alone is insufficient to cause seizures. Susceptibility genes with minor effect have hitherto been much harder to identify, and most genes conἀrmed to be implicated in idiopathic epilepsy have thus far been Mendelian. Almost all mutations identiἀed in idiopathic Mendelian epilepsies, mostly in large pedigrees, are known to be in ion channel genes or genes interacting with ion channel genes (see Table 18.1). While the broad phenotype is largely determined by a mutation in a major gene, other genes with minor effects as well as environmental factors may modulate its expression. The consequences of these modiἀer effects are twofold: incomplete penetrance, whereby not all carriers of the mutation are clinically affected, and variable phenotypic expression, whereby within a family, factors such as age of onset, type, severity and frequency of seizures, response to antiepileptic drug (AED) treatment, and duration of 267

268 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 18.1â•… Genes Found to Be Factors in Epilepsy Monogenic Epilepsies

Gene

Reference

Benign familial neonatal–infantile seizures

KCNQ2 KCNQ3 SCN2A

GEFS+, SMEI, FS

SCN1A SCN1B GABRG2 GABRG2 CLCN2 GABRA1 LGI1

Singh et al. 1998 Charlier et al. 1998 Heron et al. 2002; Striano et al. 2006 Claes et al. 2001; Escayg et al. 2000 Wallace et al. 1998 Harkin et al. 2002 Baulac et al. 2001 Haug et al. 2003 Cossette et al. 2002 Kalachikov et al. 2002

CHRNA4 CHRNB2 CHRNA2

Steinlein et al. 1995 De Fusco et al. 2000 Aridon et al. 2006

Benign familial neonatal seizures

GEFS+ GEFS+, SMEI, FS Absence epilepsy and FS IGE JME Autosomal dominant lateral temporal lobe epilepsy Autosomal dominant nocturnal frontal lobe epilepsy

Note: GEFS+, generalized epilepsy with febrile seizures plus; SMEI, severe myoclonic epilepsy of infancy; FS, febrile seizures; IGE, idiopathic generalized epilepsy; JME, juvenile myoclonic epilepsy.

the epileptic disorder may be different. Mechanisms for incomplete penetrance are largely unknown. Some believe that if there are mutations in more than one ion channel gene, the individual will have a more severe phenotype (Kearney et al. 2006). However, of interest is an animal study whereby mutations in two different ion channel genes that would individually result in opposing excitability defects resulted in nullifying the epilepsy phenotype in mice (Glasscock et al. 2007). In addition, there may be some genotype–phenotype correlation with interfamilial variability, whereby the severity of the disease relates to the type of mutation in a speciἀc gene; a nonsense mutation resulting in a stop codon, for example, may lead to a more severe phenotype. For all these reasons, it can sometimes be difficult to distinguish between monogenic epilepsies and those inherited in a complex manner. Genetic mutations may also result in symptomatic epilepsies as a result of a cortical malformation (e.g., a mutation in the GPR56 gene results in bilateral frontoparietal polymicrogyria that often presents with seizures). Mutations in the MECP2 gene result in Rett syndrome, an X-linked syndrome presenting with seizures, ataxia, and severe learning difficulties. Epilepsy in the context of learning difficulties may be a result of singlegene disorders such as Rett syndrome or due to a chromosomal abnormality such as 1p36 microdeletion syndrome, ring 20 and ring 14 syndromes, or other chromosomal aberrations such as inversion-duplication 15. The genes responsible for the epilepsy in these chromosomal syndromes are at present largely unknown. The development of higher resolution chromosomal studies such as comparative genomic hybridization using arrays may help delineate this and identify further epilepsy genes. With the advent of array technology, submicroscopic deletions and duplications are now being identiἀed. Microdeletions of 15q13.3 have been reported to be associated with epilepsy and learning difficulties (Sharp et al. 2008). Furthermore, microdeletions of 15q13.3 have been associated with idiopathic generalized epilepsy (IGE) (Helbig et al. 2009). Genomic imbalances in these regions in the form of microduplications have been found to be associated with autistic spectrum

Genetics of Sudden Death in Epilepsy

269

disorder (Miller et al. 2009) and schizophrenia (International Schizophrenia Consortium 2008). In both single-gene disorders and chromosomal disorders, the epilepsy may not always be a consistent feature of a syndrome and the phenotype and severity often varies. In some genetic syndromes, there may be an underlying brain abnormality, whereas in others, the brain appears developmentally normal. A classiἀcation of diseases frequently associated with epileptic seizures or syndromes is available in Engel’s (2001) proposed diagnostic scheme. A United Kingdom Workshop in 2006 concluded that although epilepsy genetics has advanced signiἀcantly over the past decade, this is merely the beginning. Truly characterizing the full extent of genomic variation related to multifactorial or complex epilepsy remains very much in the future (Sisodiya et al. 2007). It was noted, however, that an impact on clinical practice has already been observed. For example, as a result of the identiἀcation of SCN1A mutations in Mendelian families with a phenotype of generalized epilepsy with€ febrile seizures plus (GEFS+), sporadic cases of Dravet syndrome or severe€ myoÂ� clonic€epilepsy of infancy (SMEI) were also found to carry SCN1A mutations. These cases have been shown to respond well to stiripentol, which is considered to enhance central GABA transmission by increasing duration of channel opening (Quilichini et al. 2006) and a meta-analysis has shown signiἀcant improvement in seizure control (Kassai et al. 2008). It is hoped that this beneἀcial effect will translate into improved developmental outcome in what is currently a disorder with an extremely poor prognosis and demonstrates the key role of research into basic science for these disorders (Mullen and Scheffer 2009).

18.2â•… Mortality in Specific Syndromes A family history of an increased risk of sudden death, where looked for, has generally not been reported in sudden unexpected death in epilepsy (SUDEP) cohorts or in a case control study of SUDEP (Nashef et al. 1998; Langan et al. 2005). However, this should not be taken to exclude the possibility of a familial tendency in a subgroup. Most epidemiological studies include all epilepsies and not only those with a likely genetic basis. Studies may not be designed to identify increased risk of syncope, sudden infant death syndrome (SIDS), or sudden cardiac death (SCD), particularly if inheritance is complex or if penetrance is reduced. Data are also very limited when it comes to mortality studies in epilepsy in speciἀc syndromes, and much of what is discussed below does not have a ἀrm basis. Nevertheless, there is a suggestion that mortality rates may differ between epileptic syndromes over and above that expected from the severity of the epilepsy, as in the examples below. 18.2.1â•… 15q Inversion/Duplicaton [inv dup(15)] This chromosomal anomaly is caused by the presence of a supernumerary chromosome 15 and results in tetrasomy of the 15q11-q13 region. This region is also involved in both Angelman and Prader–Willi syndromes and overlaps with recently reported microdeletions and microduplications referred to above. Inv dup(15) syndrome is characterized by hypotonia, minor dysmorphic features, moderate to severe learning difficulties, autistic spectrum disorder, and seizures. Seizure types include spasms, atypical absences, and tonic and atonic seizures. The electroencephalograph (EEG) shows atypical hypsarhythmia, with large amplitude diffuse slow spike waves and/or multifocal abnormalities. Case

270 Sudden Death in Epilepsy: Forensic and Clinical Issues

reports indicate a variable phenotype. Mosaic inv dup (15) has been identiἀed in a healthy child (Loitzsch and Bartsch 2006), while a child with mild generalized epilepsy and a developmental disorder was found to have a large inv dup (15) (Chifari et al. 2002). Most case reports describe a phenotype where seizures are a signiἀcant and consistent problem. The IDEAS support group for inv dup (15) (Isodicentric 15 Exchange, Advocacy & Support) released a physician advisory update alerting of the risk of sudden, unexpected death in this group of patients, suggesting that the risk is in the order of 1% per year (IsoDicentric 15 Exchange, Advocacy & Support 2009). Six cases are reported on the Web site and the€mechÂ� anism is unknown in all cases with each individual dying in bed during the night, presumably while asleep. According to the Web site, “Five of the six young people had recognized seizure disorders. One had no recent seizures, and the remaining three had seizures that were described as well controlled at the time of death, and one had not had a seizure for more than a month.” While caution is required in assuming an increased risk without the data being published in a peer-reviewed publication, it is clear that the advisory clinicians felt there was cause for concern. At the time of writing, two sudden unexpected deaths in individuals with Inv dup(15) syndrome have been published (Hogart et al. 2009). Further studies on this rare chromosomal anomaly need to be carried out to conἀrm if there is a higher-than-expected rate of sudden unexpected death in these patients. 18.2.2â•…Rett Syndrome Epilepsy is a manifestation of Rett syndrome with partial and generalized seizures being reported in 50% to 90% of cases. Sudden death with no preceding symptoms is a recognized problem associated with Rett syndrome and seizures may be a partial explanation although the precise etiology is not always understood (Byard 2006). Sudden death has been reported in 22% to 26% of cases compared to 2.3% in the general population of the same age (Byard 2006). Several studies have been carried out demonstrating a prolongation of QTc interval in patients with Rett syndrome, the pathogenesis of which is unknown (Acampa and Guideri 2006). Brainstem dysfunction resulting in cardiac autonomic dysregulation is also described as being characterized by disturbed breathing and heart rate during sleep. Some studies have demonstrated labile breathing patterns and a reduction in cardiac vagal tone, indicating brainstem immaturity (Julu et al. 2001). Other support for autonomic dysregulation stems from the observation of decreased heart rate variability and sinus bradycardia (Axelrod et al. 2006). Therefore, the increase in sudden death may partly be related to the autonomic disturbance as well as perhaps to the seizures (Byard 2006). Defective autonomic nervous system control and cardiac arrhythmias relate more to functional problems than any defects demonstrated at postmortem (Byard, 2006). The pathophysiology of sudden death in Rett syndrome is an example where a single gene mutation may result in increased susceptibility to sudden death by a number of different mechanisms. 18.2.3â•…Other Examples of Syndromes with Multisystem Involvement There are other examples of syndromes with multisystem involvement, including epilepsy and increased mortality. Lafora disease, an inherited progressive myoclonic epilepsy characterized by intractable epilepsy in association with progressive neurological and cognitive deἀcit due to a mutation in the EPM2A gene on chromosome 6q24 is also associated with

Genetics of Sudden Death in Epilepsy

271

premature death from a variety of causes, including sudden death (Wick and Byard 2006). The same applies to the mitochondrial mutation 3243A>G resulting in mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes (MELAS), which increases the risk of death among carriers and their ἀrst-degree maternal relatives. Death was due to a variety of causes including metabolic disturbances, cardiac involvement, status epilepticus, and sudden death associated with epilepsy or diabetes (Majamaa-Voltti et al. 2008). These examples emphasize the need for mortality studies in speciἀc syndromes and the limitations of “all inclusive” mortality studies in epilepsy. 18.2.4â•… SCN1A-Related Disorders The vast majority of mutations identiἀed in the SCN1A gene have been in sporadic cases of SMEI. SMEI is characterized by normal early development followed by the onset of feversensitive and refractory epilepsy with GTCS or unilateral seizures beginning within the ἀrst year of life. Most mutations identiἀed are nonsense mutations although some are missense mutations. Approximately 5% to 10% of families with GEFS+, an idiopathic familial epilepsy syndrome characterized by febrile seizures extending beyond 6 years with or without afebrile seizures, have also been shown to have SCN1A missense mutations (Marini et al. 2007), although other genes (GABRG2, SCN1B) are also implicated (Helbig et al. 2008). In addition, SCN1A mutations have been identiἀed in familial hemiplegic migraine. Mortality rates in SMEI are observed to be higher than in some other epilepsies, somewhere in the order of 15% versus 5% (Kassai et al. 2008), although controlled prospective studies have not been carried out and would require multicenter collaboration due to the rarity of SMEI. In Dravet’s series of 63 patients with a clinical diagnosis of SMEI and a mean age at follow-up of 11 years 4 months, 10 died, 2 of whom were sudden unexplained deaths. One death was unclassiἀed, while other causes of death included drowning, accidents, infection, and status epilepticus. In a Japanese series of clinically diagnosed SMEI or borderline SMEI, 12 of 85 patients died. Seven patients died from status epilepticus, one from severe infection, and one by accidental drowning. Three died suddenly of unknown causes (Oguni et al. 2001). In a further study of clinical cases of SMEI, two patients died; one from an unknown cause (Caraballo and Fejerman 2006). Accurate person years of follow-up are generally not provided to allow for incidence of sudden death, standardized mortality rates, and conἀdence intervals to be calculated. Mortality rates in GEFS+ families have not been reported to be higher than expected from patients with epilepsy, although this has not been systematically examined. We described a family with GEFS+ and a novel SCN1A mutation with two SUDEP cases (Hindocha et al. 2008) in individuals with more severe epilepsy. Although DNA samples were unavailable from the SUDEP cases to conἀrm their carrier status of the SCN1A mutation, their epilepsy phenotype was consistent with the familial diagnosis of GEFS+. To exclude as far as possible an alternative cause for the sudden death in these individuals, a ἀrst-degree relative of each SUDEP case was screened for the most common cardiac channelopathy mutations with negative results. SUDEP incidence in this family was found to be 7/1000 (95% conἀdence interval, 1–25), compared with an incidence of less than 1/1000, in population-based epilepsy cohorts. The wide conἀdence intervals suggest that the two deaths could still have occurred by chance. Nevertheless, the two SUDEP cases occurred in individuals with uncontrolled epilepsy and there was no other family history of sudden premature death. These two cases raise the possibility of an increased genetic predisposition to sudden death in people with

272 Sudden Death in Epilepsy: Forensic and Clinical Issues

SCN1A mutations in the setting of uncontrolled seizures. Possible mechanisms are discussed below. While initial characterization of ion channels presumed that these proteins were more localized to a speciἀc tissue, it is being increasingly recognized that this is not the case. Tissues are now thought to be mosaic for these proteins, composed of many isoforms, each expressed in different proportions (Haufe et al. 2007). Although SCN1A may primarily be expressed in the brain, several studies have shown that Nav1.1 (SCN1A gene product) is present in various regions of the heart in rat and mouse (Rogart et al. 1989; Dhar et al. 2001; Marionneau et al. 2005), in rabbit neonate (Baruscotti et al. 1997), and in dog (Haufe et al. 2005). There is good evidence for a role for Nav1.1 in pacemaker function of the sinoatrial (SA) node, but a lack of expression in atrial muscle (Tellez et al. 2006). In mice, Nav1.1 (but not Nav1.5, a SCN5A gene product) was detected in the SA node, and moreover, when brain-type Na+ channels were selectively blocked, signiἀcantly reduced spontaneous heart rate and greater heart rate variability were observed (Maier et al. 2003). A role for Nav1.1 in pacemaker activity in the mouse SA node (Lei et al. 2004) and rat SA node (Du et al. 2007) was conἀrmed in other studies. In contrast, evidence for a role for Nav1.1 in ventricular function is contradictory. Nav1.1 has been detected in mouse ventricular myocytes (Maier et al. 2002) and when brain-type Na+ channels are blocked, ventricular function is reduced, suggesting a role in excitation-contraction coupling (Maier et al. 2002). However, another similar study in rat ventricular myocytes demonstrated no reduction in ventricular function (Brette and Orchard 2006). SCN1A mutations may also result in dysfunction in the brainstem, resulting in alteration in autonomic function and thereby theoretically predisposing to sudden death. SCN1A mutations have been identiἀed in familial hemiplegic migraine. The pathophysiology of migraine implicates the brainstem, whereby brain imaging studies have established reproducible changes in the brain (Goadsby 2007; Goadsby and Hargreaves 2008). In addition, a study describing a missense mutation in a case of SMEI has suggested possible dysfunction of the brainstem in this disorder (Kimura et al. 2005). Two brothers with SMEI were found to have a missense mutation and their father, also a carrier of the mutation, had experienced two simple FSs before the age of 4. Both siblings had a deranged sleep-wake cycle after late infancy, postulated to be due to the dysfunction of aminergic neurons in the brainstem. Studies in rat models have also shown good expression of SCN1A in the brainstem (Gong et al. 1999).

18.3â•… SUDEP and Idiopathic Epilepsy Although most SUDEP cases are associated with more intractable and usually focal or symptomatic epilepsies, cases with IGEs with a history of generalized tonic seizures are nevertheless well represented in SUDEP cohorts. This was evident in an early study by one of the authors (L.N.) where 9 of 26 SUDEP cases were classiἀed as having idiopathic primary generalized epilepsy (see Chapter 58, this book). This cohort was largely identiἀed through the self-help group Epilepsy Bereaved and not through specialized services. Of these IGE cases, one had reportedly never been treated, and one had reportedly discontinued medication independently. Another, with juvenile myoclonic epilepsy in remission€on valproic acid, had been independently considering medication reduction, but it is not known if this had taken place. Two others had only ever been treated with carbamazepine or phenytoin.

Genetics of Sudden Death in Epilepsy

273

Two, both with a positive photoparoxysmal response, were found dead near visual display units, in one case a computer game was on and the person was found dead with a bitten tongue; the other was watching a video at the time of death. One had discontinued medication, while the other had only had a small number of convulsions in similar settings previously, and as already referred to, had never been treated (Nashef et al. 1998). In Langan et al.’s (2005) case control study, no association was found with syndromic diagnosis, but information on this variable was incomplete. In Nilsson et al.’s (1999) study, no increase in risk was associated with any particular type of epilepsy, but a lowered risk was associated with localization-related symptomatic epilepsy compared with generalized idiopathic epilepsy, especially among men, but numbers were small and conἀdence intervals wide. Genton and Gelisse (2001) reported three cases of premature death in juvenile myoÂ� clonic epilepsy. The report was useful in both highlighting this as a potential problem€and in providing person years of follow-up. Among 170 consecutive juvenile myoclonic epilepsy cases, three female patients died prematurely without autopsy being performed. One case had anorexia nervosa and died from severe aspiration pneumonia after provoked vomiting. The second had a history of psychosis and a case of IGE in her family. She had uncontrolled epilepsy and was found, aged 16, cyanotic and unconscious one morning “in the toilets of her institution” and died before resuscitation. The third had uncontrolled epilepsy, was also on neuroleptics, had a borderline personality disorder, a history of alcoholism, and low compliance. She was found dead at home at the age of 42 years. The authors concluded that severe mental disorders were risk factors for unexpected death in juvenile myoclonic epilepsy with a death ratio, if only two of these three cases are considered, of 0.9/1000. Aurlien et al. (2007) reported on four consecutive female SUDEP cases aged 25, 16, 37, and 24 years with idiopathic epilepsy treated with lamotrigine monotherapy. One case was unclassiἀed (auras suggested focal epilepsy but MRI was normal and EEG showed bilateral synchronous epileptogenic activity) and the other three were idiopathic generalized (one juvenile myoclonic epilepsy with photosensitivity, another with concomitant diabetes). The authors considered four possible explanations for this observation: an insufficient effect of lamotrigine leading to fatal seizures, a direct effect of lamotrigine on vital functions, such as cardiac rhythmicity (given that lamotrigine inhibits the cardiac rapid delayed rectiἀer potassium ion current (Ikr)), a combination of drug-induced effects and seizures, or coincidence. It is interesting that although they had not systematically identiἀed all SUDEP cases in their area during this period, these four cases were the only ones they were aware of among their outpatients and represented all SUDEP patients reported by the department pathologist. Also of interest was the control of the epilepsy. Seizure frequency on last outpatient hospital visit in cases 1 through 4, respectively, was: 1.5 simple partial seizures/ month; seizure free for 6 months; seizure free for 7 months; two seizures during the last week (previously seizure-free for 3 months). At autopsy, case 1 did not have AED levels performed, case 2 did not have detectable blood concentration of lamotrigine (previously 7 µmol/l), case 3 had a level of 15 µmol/l, previously 24.4, and one case had a lower postmortem lamotrigine (3.2 µmol/l) compared with the last antemortem level (27.5 µmol/l) when on combination therapy with carbamazepine. The authors considered the possibility that “there may be subgroups of patients with idiopathic epilepsy and generalized tonic– clonic seizures treated with lamotrigine that are at an increased risk of SUDEP.” They cited reports of a greater risk of drug-induced torsade in females, who also have an increased prevalence of symptoms of congenital long QT syndrome (LQTS) and an increase in episodes of supraventricular tachycardia in the perimenstrual period in susceptible patients.

274 Sudden Death in Epilepsy: Forensic and Clinical Issues

These careful observations yet again highlight the importance of mortality studies in speciἀc syndromes and suggest a greater risk of SUDEP in those with idiopathic epilepsy, even when the epilepsy is not severe. Which of the possible explanations suggested by the authors are true remain unknown. Note that a limited number of other studies (Timmings 1998; Langan et al. 2005; for review, see Rugg-Gunn and Nashef 2009) suggested that carbamazepine might also increase risk to a small extent again through an effect on cardiac function. In Aurlien et al.’s (2007) series, there is also possible selection bias in terms of preferential prescribing of lamotrigine to females with idiopathic epilepsy of childbearing age, whereas valproate, although particularly effective, may have been avoided because of potential teratogenicity. Theoretically, a particular AED or AED combination could have a detrimental effect by either increasing risk of sudden death through a variety of mechanisms or by giving insufficient protection for the seizure disorder. One angle that has not been explored but perhaps suggested by absent or lower levels in some of these cases (not withstanding reservations about postmortem blood levels), is a possible differential SUDEP risk associated with AED withdrawal, through an effect on seizure severity or autonomic function. This potential AED factor may have relevance if there is nonadherence to treatment or abrupt withdrawal. It is generally considered, for example, that the full effect of starting and stopping valproate in IGE cases is not all immediate with a delayed effect often observed. Thus, it is possible that effective control may be lost less quickly with valproate than when some other drugs are omitted. The same group (Aurlien et al. 2009) published a later follow up report on case 1 when autopsy DNA sequencing of LQTS-associated genes revealed a novel missense mutation in the SCN5A gene coding for the cardiac sodium channel, voltage-gated, type V alpha subunit. They discussed whether the mutation may explain both the epilepsy and the sudden death and the possible effect of lamotrigine on cardiac ion channel function. This is a particularly interesting report as it provides evidence in support of the same genetic mutation giving rise to both epilepsy and susceptibility to cardiac death. We have studied a small pedigree of an otherwise well young woman with juvenile myoclonic epilepsy who died suddenly while on lamotrigine, having been previously fully controlled on valproate, but who changed medication because of weight gain and potential teratogenicity, and whose control was not as good on lamotrigine as it had been on valproate. As can be seen in Figure 18.1, her mother had undiagnosed blackouts and her brother died of SIDS. LQT channel mutation screen in this pedigree, however, was negative. The pedigree raises the possibility of wider overlap presentations, as yet unexplored. Two other individuals with idiopathic epilepsy and cardiac arrhythmias have also so far been negative on screening for the most common LQTS and Brugada syndrome (BS) gene mutations (70% and 20%, respectively). Recently, it has been demonstrated that individuals with mutations in KCNH2 seen in LQTS are more likely to have a personal history of seizures than other subtypes of LQTS, raising the possibility that the KCNH2-encoded potassium channel could confer susceptibility to seizures (Johnson et al. 2009). Also of particular interest is the SCN1B gene previously implicated in GEFS+, absence epilepsy and temporal lobe epilepsy, and now in BS (Watanabe et al. 2008). This is discussed in more detail below. In our view, the available, though limited, evidence raises the possibility that in idiopathic epilepsy, with a history of generalized tonic clonic seizures, SUDEP may occur more often than in other syndromes with comparable epilepsy severity. At present, this is only a

Genetics of Sudden Death in Epilepsy

275

• Age of seizure onset—15 years 1.1

1.2

1.3 1.4 Late-onset epilepsy

II.I

III.1

II.2 Syncopal episodes

III.2

Seizures/epilepsy

III.3 III.4 SUDEP SIDS

• Myoclonic jerks and rare generalized tonic–clonic seizure • Idiopathic uncontrolled epilepsy at time of death • Found dead in bed • Clear bite mark on tip and left side of tongue found at postmortem supporting probability of epileptic seizure

DNA screened for the most common LQTS and BS gene mutations (account for 75% of all LQTS and BS). Mutations in these genes have been excluded as far as possible, as a cause of the sudden death.

Figure 18.1╇ SUDEP case study.

suggestion, but it needs to be explored further. AED choices may be particularly important in this subgroup as suggested by Aurlien et al. (2007). It is important in this context, however, to stress the potential for misdiagnosis of LQT as epilepsy, and to emphasize the need for clinicians to be constantly vigilant in ensuring diagnostic accuracy and in screening for alternative or concomitant diagnosis in those diagnosed with epilepsy.

18.4â•… SUDEP in Genetic Epilepsies: How Does It Occur? In SUDEP, different risk factors and mechanisms may operate with a ἀnal common pathway of cardiorespiratory compromise. Respiratory compromise and hypoxia occur frequently in seizures, particularly if convulsive, and the severity may be influenced by position and airway during postictal coma, particularly in the absence of someone capable of giving assistance. Cardiac changes occur during seizures and although sinus tachycardia is most commonly observed during seizures, ictal sinus arrest also occurs as do rare malignant dysrhythmias. Functional cardiac changes such as apical ballooning may also occur (Stöllberger and Finsterer 2004). In discussing genetic predisposition, an increased susceptibility to sudden death may be due to cardiac mechanisms, reflecting underlying processes common to neurological and cardiac functions, autonomic function, or brainstem control of respiration. Etiological factors underlying SUDEP are likely to be heterogeneous, in much the same way as in sudden unexpected death (SUD) and SIDS (see below). Given the association between intractable epilepsy and SUDEP, one can postulate that in some cases of idiopathic epilepsy a primarily “neuronal” mutation, if it can also cause a predisposition to cardiac arrhythmias, may manifest as sudden death in persons in whom the epilepsy is uncontrolled. In others there may be an unrelated coexisting “mild” susceptibility to SCD that would manifest itself in the presence of uncontrolled seizures (Nashef et al. 2007). In our 2007 review, we discussed the possibility of overlap between susceptibility to SCD and idiopathic epilepsy and that further genetic and epidemiological studies are needed (Nashef et al. 2007). However, at the time there was no bridging evidence to

276 Sudden Death in Epilepsy: Forensic and Clinical Issues

support this hypothesis. Since then there have been reports as above, which suggest that ion channel gene mutations already known to have more than a single pathological role, as demonstrated, for example, by LQTS with deafness, can cause both epilepsy and LQTS or other inherited susceptibility to cardiac dysrhythmias, although such reports are currently uncommon. Animal models provide evidence in support of this hypothesis as in the case of the Ca2+ release channel ryanodine receptor 2 (RyR2) required for excitation-Â�contraction coupling in the heart- and expressed in the brain. Mutations in RyR2, which result in “leaky” RyR2 channels, have been linked to exercise-induced SCD and catecholaminergic polymorphic ventricular tachycardia. Mice heterozygous for the R2474S mutation in Ryr2 exhibited spontaneous generalized tonic–clonic seizures (without cardiac arrhythmias), exercise-induced ventricular arrhythmias, and SCD (Lehnart et al. 2008). Treatment with a compound inhibiting the channel leak prevented cardiac arrhythmias and raised the seizure threshold. The authors proposed that this was a combined neurocardiac disorder.

18.5â•… Genetic Susceptibility to SCD and SIDS In considering possible genetic predisposition to SUDEP, it may useful to briefly review the evidence of genetic susceptibility to SCD and SIDS. At least 4% of sudden deaths are unexplained at autopsy and on average a quarter of these may be due to inherited cardiac disease (Behr et al. 2008). These ἀgures will vary according to age group (Rodriguez-Calvo et al. 2008). Diagnosis is crucial as close relatives may be at potential risk of also having a fatal cardiac event. While the genetic causes of SCD also include structural abnormalities (e.g., inherited cardiomyopathies), a proportion (Saenen and Vrints 2008) are a result of primary arrhythmogenic disorders, also known as cardiac channelopathies, such as LQTS, short QT syndrome, BS, and catecholaminergic polymorphic ventricular tachycardia. Most cases of LQTS and BS are inherited from a parent who may or may not show clinical symptoms (Morales et al. 2008). In a study of autopsies in cases of patients aged 5 to 35 years who died suddenly from cardiac causes, 29% were arrhythmia-related SCD (Puranik et al. 2005). Of the cardiac cases, SCD in a ἀrst-degree relative was reported in only 4.5% of cases with a low yield of signiἀcant positive family history from these cases. The diagnostic yield of investigating relatives of individuals who had died suddenly with no explanation has also been examined. In one study (Tan et al. 2005), 43 families with one sudden unexpected death victim who had died under the age of 40 were investigated. Seven of the 43 families studied (16%) were found clinically to have LQTS or BS, diagnoses that were conἀrmed with molecular techniques in four of these seven families (9%). A slightly larger study showed that 21 of 57 families (37%) were clinically identiἀed to have either deἀnite or probable LQTS or BS. Molecular conἀrmation was made in six of the 57 families (11%) (Behr et al. 2008). Once molecular conἀrmation is achieved, testing may be offered to appropriate relatives. One study suggests that clinical evaluation alone is no longer appropriate to exclude a diagnosis of LQTS, as penetrance may be very low (around 25%) (Priori et al. 1998). Congenital LQTS is mostly inherited in an autosomal dominant fashion (RomanoWard syndrome) and up until now, mutations in 12 genes have been identiἀed to result in LQTS (KCNQ1, KCNH2, SCN5A, ANK2, KCNE1, KCNE2, KCNJ2, CACNA1C, CAV3, SCN4B, AKAP9, and α-1-syntrophin) (Van Norstrand and Ackerman 2009). All of these genes are expressed primarily in cardiac tissue and most of these genes encode an ion

Genetics of Sudden Death in Epilepsy

277

channel protein. Those that do not have been found to affect the sodium or potassium channel (Rodriguez-Calvo et al. 2008). BS is characterized by episodes of VF and in approximately 20% of cases a positive family history of unexplained sudden death is present (Miura et al. 2008). BS is genetically heterogeneous and mutations in the sodium channel gene SCN5A account for 20% to 30% of cases (Van Norstrand and Ackerman 2009). More recently, a mutation in SCN1B, primarily a neuronal voltage-gated sodium channel and implicated in GEFS+ (Wallace et al. 1998; Singh et al. 1999) but also expressed in the heart, has been implicated in BS (Watanabe et al. 2008). This is an exciting discovery, as it is an example of a gene having a pathological effect on the heart and the brain. Mutations in SCN1B have been previously reported in a family with early onset absence epilepsy without the occurrence of febrile seizures (Audenaert et al. 2003) and temporal lobe epilepsy even without febrile seizures (Scheffer et al. 2007). Cardiac abnormalities have not been commented on in these families. β1-null mice exhibit a severe seizure disorder and premature death (Chen et al. 2004). In addition, these mice exhibit bradycardia and prolonged rate-corrected QT intervals (Lopez-Santiago et al. 2007). These changes suggest that this protein has a possible role in the heart. Other genes such as CACNA1C, CACNB2b, and GPD1-L, known to modulate sodium channel function, are also implicated in BS. Since the advent of genome-wide association studies, research examining genetic susceptibility to various disorders has developed signiἀcantly. Single nucleotide polymorphisms (SNPs) in the LQT genes have been identiἀed as demonstrating a genetic propensity to developing increased QT interval/cardiac arrhythmias/SCD, not related to a single gene defect but because of multiple low-risk alleles (Pfeufer et al. 2005; Gouas et al. 2005). However, replication data at present is lacking (Schulze-Bahr 2006). Of interest in this context is the variant allele (S1103Y) of the cardiac sodium channel gene SCN5A, which (1) has a subtle effect on risk in African Americans (13% of whom are carriers), manifesting if there are additional acquired risk factors, with most carriers never having an arrhythmia (Splawski et al. 2002), (2) has a strong effect on risk in Caucasians where it is very rare (Chen et al. 2002), and (3) is associated, if homozygous, with SIDS in AfricanAmericans (Plant et al. 2006). Using similar techniques, genetic susceptibility to SIDS has been examined (WeeseMayer et al. 2007). Around 5% to 10% of SIDS cases (9.5% by Arnestad et al. 2007) are linked to ion channelopathies (Tester and Ackerman 2005). From an 18-year study (Schwartz et al. 1998), an increase in the QT interval in the ἀrst week of life was established in individuals who went on to die from SIDS. SNPs in common cardiac channel genes may confer risk to sudden death at any age including infancy, even prenatally. Several studies have identiἀed differences between infants with SIDS and control infants with gene polymorphisms in SCN5A (Plant et al. 2006). Further research has identiἀed other genetic factors predisposing to SIDS. For example, decreased serotonin (5HT) receptor binding in the 5HT pathways of the medulla has been identiἀed in a small number of SIDS cases (Paterson et al. 2006; see Paterson’s chapter 5). In addition, association between the serotonin transporter gene (5-HTT) and SIDS has been demonstrated (Opdal and Rognum, 2004), suggesting that serotonin may play a regulatory role in SIDS. Of interest in this context is the protective effect of fluoxetine, a selective serotonin reuptake inhibitor, in reducing ictal respiratory arrest in DBA/2 mice with audiogenic seizures at doses that did not reduce seizure severity (Tupal and Faingold 2006). Polymorphisms in genes involved in inflammatory and infectious processes such as interleukin-10 (IL-10) gene have been shown to be associated with both SIDS and infectious

278 Sudden Death in Epilepsy: Forensic and Clinical Issues

death (Korachi et al. 2004) such as pneumonia and Epstein-Barr virus (Weese-Mayer et al. 2007). In addition, genes pertinent to the development of the autonomic nervous system are also being studied because of reports of autonomic dysregulation in SIDS victims and polymorphisms have been found to be associated in several genes (PHOX2a, RET, ECE1, TLX3, EN1) (Weese-Mayer et al. 2004).

18.6â•… Conclusion While genetic predisposition may be less prominent than the severity of the epilepsy, treatment-related factors, and supervision in many cases of SUDEP, it is still a potentially very important area requiring further research, particularly in idiopathic epilepsy. Future research in this area should include epidemiological studies looking at mortality in speciἀc syndromes, as well as studies of possible overlap of epilepsy and syncope in idiopathic epilepsy and in LQT cohorts. Clinicians need to be more aware, not only of the potential for misdiagnosis but also for potential clinical overlap. There is also an urgent need to collect DNA and clinical details in SUDEP cases, initially to carry out screening for genetic mutations in selected cases, but also to carry out sufficiently powered association studies. This is a difficult but not insurmountable task; one possible source is DNA from dried blood spots from Guthrie cards taken from newborns in the United Kingdom. These are stored for many years and can be retrospectively retrieved to carry out molecular studies (Skinner et al. 2004).

References Acampa, M., F. Guideri. 2006. Cardiac disease and Rett syndrome. Arch Dis Child 91: 440–443. Aridon, P., C. Marini, C. Di Resta, E. Brilli, M. De Fusco, F. Politi, E. Parrini et al. 2006. Increased sensitivity of the neuronal nicotinic receptor alpha 2 subunit causes familial epilepsy with nocturnal wandering and ictal fear. Am J Hum Genet 79: 342–350. Arnestad, M., L. Crotti, T. O. Rognum, R. Insolia, M. Pedrazzini, C. Ferrandi, A. Vege et al. 2007. Prevalence of long-QT syndrome gene variants in sudden infant death syndrome. Circulation 115: 361–367. Audenaert, D., L. Claes, B. Ceulemans, A. Lofgren, C. Van Broeckhoven, and P. De Jonghe. 2003. A deletion in SCN1B is associated with febrile seizures and early-onset absence epilepsy. Neurology 61: 854–856. Aurlien, D., E. Tauboll, and L. Gjerstad. 2007. Lamotrigine in idiopathic epilepsy—Increased risk of cardiac death? Acta Neurol Scand 115: 199–203. Aurlien, D., T. P. Leren, E. Tauboll, and L. Gjerstad. 2009. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18: 158–160. Axelrod, F. B., G. G. Chelimsky, and D. E. Weese-Mayer. 2006. Pediatric autonomic disorders. Pediatrics 118: 309–321. Baruscotti, M., R. Westenbroek, W. A. Catterall, D. DiFrancesco, and R. B. Robinson. 1997. The newborn rabbit sino-atrial node expresses a neuronal type I-like Na+ channel. J Physiol 498 (Pt 3): 641–648. Baulac, S., G. Huberfeld, and I. Gourἀnkel-An et al. 2001. First genetic evidence of GABA(A) receptor dysfunction in epilepsy: A mutation in the gamma2-subunit gene. Nat Genet 28: 46–48. Behr, E. R., C. Dalageorgou, M. Christiansen, P. Syrris, S. Huges, M. T. Tome Esteban, E. Rowland, S. Jeffery, and W. J. McKenna. 2008. Sudden arrhythmic death syndrome: Familial evaluation identiἀes inheritable heart disease in the majority of families. Eur Heart J 29: 1670–1680.

Genetics of Sudden Death in Epilepsy

279

Brette, F., and C. H. Orchard. 2006. No apparent requirement for neuronal sodium channels in excitation-Â�contraction coupling in rat ventricular myocytes. Circ Res 98: 667–674. Byard, R. W. 2006. Forensic issues and possible mechanisms of sudden death in Rett syndrome. J Clin Forensic Med 13: 96–99. Caraballo, R. H., and N. Fejerman. 2006. Dravet syndrome: A study of 53 patients. Epilepsy Res 70 (S1): S231–S238. Charlier, C., N. A. Singh, S. G. Ryan, T. B. Lewis, B. E. Reus, R. J. Leach, and M. Leppert. 1998. A pore mutation in a novel KQT-like potassium channel gene in an idiopathic epilepsy family. Nat Genet 18: 53–55. Chen, C., R. E. Westenbroek, X. Xu, C. A. Edwards, D. R. Sorenson, Y. Chen, D. P. McEwen et al. 2004. Mice lacking sodium channel beta1 subunits display defects in neuronal excitability, sodium channel expression, and nodal architecture. J Neurosci 24: 4030–4042. Chen, S., M. K. Chung, D. Martin, R. Rozich, P. J. Tchou, and Q. Wang. 2002. SNP S1103Y in the cardiac sodium channel gene SCN5A is associated with cardiac arrhythmias and sudden death in a white family. J Med Genet 39: 913–915. Chifari, R., R. Guerrini, M. Pierluigi, S. Cavani, V. Sgro, M. Elia, R. Canger et al. 2002. Mild generalized epilepsy and developmental disorder associated with large inv dup(15). Epilepsia 43: 1096–1100. Claes, L., J. Del Favero, B. Ceulemans, L. Lagae, C. Van Broeckhoven, and P. De Jonghe. 2001. De novo mutations in the sodium-channel gene SCN1A cause severe myoclonic epilepsy of infancy. Am J Hum Genet 68: 1327–1332. Cossette, P., L. Liu, K. Brisebois, H. Dong, A. Lortie, M. Vanasse, J. M. Saint-Hilaire et al. 2002. Mutation of GABRA1 in an autosomal dominant form of juvenile myoclonic epilepsy. Nat Genet 31: 184–189. De Fusco, M., A. Becchetti, A. Patrignani, G. Annesi, A. Gambardella, A. Quattrone, A. Ballabio, E. Wanke, and G. Casari. 2000. The nicotinic receptor beta 2 subunit is mutant in nocturnal frontal lobe epilepsy. Nat Genet 26: 275–276. Dhar, M. J., C. Chen, I. Rivolta, H. Abriel, R. Malhotra, L. N. Mattei, F. C. Brosius,€R.€S.€Kass,€and L.€ L.€ Isom. 2001. Characterization of sodium channel alpha- and beta-subunits in rat and mouse cardiac myocytes. Circulation 103: 1303–1310. Du, Y., X. Huang, T. Wang et al. 2007. Downregulation of neuronal sodium channel subunits Nav1.1 and Nav1.6 in the sinoatrial node from volume-overloaded heart failure rat. Pflugers Arch 454€(3): 451–459. Engel, Jr., J. 2001. A proposed diagnostic scheme for people with epileptic seizures and with epilepsy: Report of the ILAE Task Force on Classiἀcation and Terminology. Epilepsia 42: 796–803. Escayg, A., B. T. MacDonald, M. H. Meisler et al. 2000. Mutations of SCN1A, encoding a neuronal sodium channel, in two families with GEFS+2. Nat Genet 24: 343–345. Genton, P., and P. Gelisse. 2001. Premature death in juvenile myoclonic epilepsy. Acta Neurol Scand 104: 125–129. Glasscock, E., J. Qian, J. W. Yoo, and J. L. Noebels. 2007. Masking epilepsy by combining two epilepsy genes. Nat Neurosci 10: 1554–1558. Goadsby, P. J. 2007. Recent advances in understanding migraine mechanisms, molecules and therapeutics. Trends Mol Med 13: 39–44. Goadsby, P. J., and R. Hargreaves. 2008. Refractory migraine and chronic migraine: Pathophysiological mechanisms. Headache 48: 1399–1405. Gong, B., K. J. Rhodes, Z. Bekele-Arcuri, and J. S. Trimmer. 1999. Type I and type II Na(+) channel alpha-subunit polypeptides exhibit distinct spatial and temporal patterning, and association with auxiliary subunits in rat brain. J Comp Neurol 412: 342–352. Gouas, L., V. Nicaud, M. Berthet et al. 2005. Association of KCNQ1, KCNE1, KCNH2 and SCN5A polymorphisms with QTc interval length in a healthy population. Eur J Hum Genet 13: 1213–1322.

280 Sudden Death in Epilepsy: Forensic and Clinical Issues Harkin, L. A., D. N. Bowser, L. M. Dibbens et al. 2002. Truncation of the GABA(A)-receptor gamma2 subunit in a family with generalized epilepsy with febrile seizures plus. Am J Hum Genet 70: 530–536. Haufe, V., J. M. Cordeiro, T. Zimmer et al. 2005. Contribution of neuronal sodium channels to the cardiac fast sodium current INa is greater in dog heart Purkinje ἀbers than in ventricles. Cardiovasc Res 65: 117–127. Haufe, V., C. Chamberland, and R. Dumaine. 2007. The promiscuous nature of the cardiac sodium current. J Mol Cell Cardiol 42: 469–477. Haug, K., M. Warnstedt, A. K. Alekov et al. 2003. Mutations in CLCN2 encoding a voltage-gated chloride channel are associated with idiopathic generalized epilepsies. Nat Genet 33: 527–532. Helbig, I., I. E. Scheffer, J. C. Mulley, and S. F. Berkovic. 2008. Navigating the channels and beyond: Unravelling the genetics of the epilepsies. Lancet Neurol 7: 231–245. Helbig, I., H. C. Mefford, A. J. Sharp et al. 2009. 15q13.3 microdeletions increase risk of idiopathic generalized epilepsy. Nat Genet 41: 160–162. Heron, S. E., K. M. Crossland, E. Andermann et al. 2002. Sodium-channel defects in benign familial neonatal–infantile seizures. Lancet 360: 851–852. Hindocha, N., L. Nashef, F. Elmslie et al. 2008. Two cases of sudden unexpected death in epilepsy in a GEFS+ family with an SCN1A mutation. Epilepsia 49: 360–365. Hogart, A., K. N. Leung, N. J. Wang et al. 2009. Chromosome 15q11-13 duplication syndrome brain reveals epigenetic alterations in gene expression not predicted from copy number. J Med Genet 46: 86–93. International Schizophrenia Consortium. 2008. Rare chromosomal deletions and duplications increase risk of schizophrenia. Nature 455: 237–241. IsoDicentric 15 Exchange, Advocacy & Support 2009. Johnson, J. N., N. Hofman, C. M. Haglund, G. D. Cascino, A. A. M. Wilde, and M. J. Ackerman. 2009. Identiἀcation of a possible pathogenic link between congenital long QT syndrome and epilepsy. Neurology 72: 224–231. Julu, P. O., A. M. Kerr, F. Apartopoulos et al. 2001. Characterisation of breathing and associated central autonomic dysfunction in the Rett disorder. Arch Dis Child 85: 29–37. Kalachikov, S., O. Evgrafov, B. Ross et al. 2002. Mutations in LGI1 cause autosomal-dominant partial epilepsy with auditory features. Nat Genet 30: 335–341. Kassai, B., C. Chiron, S. Augier et al. 2008. Severe myoclonic epilepsy in infancy: A systematic review and a meta-analysis of individual patient data. Epilepsia 49: 343–348. Kearney, J. A., Y. Yang, B. Beyer et al. 2006. Severe epilepsy resulting from genetic interaction between Scn2a and Kcnq2. Hum Mol Genet 15: 1043–1048. Kimura, K., T. Sugawara, E. Mazaki-Miyazaki et al. 2005. A missense mutation in SCN1A in brothers with severe myoclonic epilepsy in infancy (SMEI) inherited from a father with febrile seizures. Brain Dev 27: 424–430. Korachi, M., V. Pravica, A. J. Barson, I. V. Hutchinson, and D. B. Drucker. 2004. Interleukin 10 genotype as a risk factor for sudden infant death syndrome: determination of IL-10 genotype from wax-embedded postmortem samples. FEMS Immunol Med Microbiol 42: 125–129. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Lehnart, S. E., M. Mongillo, A. Bellinger et al. 2008. Leaky Ca2+ release channel/ryanodine receptor 2 causes seizures and sudden cardiac death in mice. J Clin Invest 118: 2230–2245. Lei, M., S. A. Jones, J. Liu et al. 2004. Requirement of neuronal- and cardiac-type sodium channels for murine sinoatrial node pacemaking. J Physiol 559: 835–848. Loitzsch, A., and O. Bartsch. 2006. Healthy 12-year-old boy with mosaic inv dup(15)(q13). Am J Med Genet A 140: 640–643. Lopez-Santiago, L. F., L. S. Meadows, S. J. Ernst et al. 2007. Sodium channel Scn1b null mice exhibit prolonged QT and RR intervals. J Mol Cell Cardiol 43: 636–647.

Genetics of Sudden Death in Epilepsy

281

Maier, S. K., R. E. Westenbroek, K. A. Schenkman, E. O. Feigl, T. Scheuer, and W. A. Catterall. 2002. An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart. Proc Natl Acad Sci U S A 99: 4073–4078. Maier, S. K., R. E. Westenbroek, T. T. Yamanushi et al. 2003. An unexpected requirement for braintype sodium channels for control of heart rate in the mouse sinoatrial node. Proc Natl Acad Sci U S A 100: 3507–3512. Majamaa-Voltti, K., J. Turkka, M. L. Kortelainen, H. Huikuri, and K. Majamaa. 2008. Causes of death in pedigrees with the 3243A>G mutation in mitochondrial DNA. J Neurol Neurosurg Psychiatry 79: 209–211. Marini, C., D. Mei, T. Temudo et al. 2007. Idiopathic epilepsies with seizures precipitated by fever and SCN1A abnormalities. Epilepsia 48: 1678–1685. Marionneau, C., B. Couette, J. Liu et al. 2005. Speciἀc pattern of ionic channel gene expression associated with pacemaker activity in the mouse heart. J Physiol 562: 223–234. Miller, D. T., Y. Shen, L. A. Weiss et al. 2009. Microdeletion/duplication at 15q13.2q13.3 among individuals with features of autism and other neuropsychiatirc disorders. J Med Genet 46€(4): 242–248. Miura, D., K. Nakamura, and T. Ohe. 2008. Update on genetic analysis in Brugada syndrome. Heart Rhythm 5: 1495–1496. Morales, A., J. Cowan, J. Dagua, and R. E. Hershberger. 2008. Family history: An essential tool for cardiovascular genetic medicine. Congest Heart Fail 14: 37–45. Mullen, S. A., and I. E. Scheffer. 2009. Translational research in epilepsy genetics: Sodium channels in man to interneuronopathy in mouse. Arch Neurol 66: 21–26. Mulley, J. C., I. E. Scheffer, S. Petrou, L. M. Dibbens, S. F. Berkovic, and L. A. Harkin. 2005. SCN1A mutations and epilepsy. Hum Mutat 25: 535–542. Nashef, L., S. Garner, J. W. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64: 349–352. Nashef, L., N. Hindocha, and A. Makoff. 2007. Risk factors in sudden death in epilepsy (SUDEP): The quest for mechanisms. Epilepsia 48: 859–871. Nashef, L., and L. Sahlholdt. 2010. Bereavement and sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, chap. 58. Boca Raton, FL: CRC Press. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 353: 888–893. Oguni, H., K. Hayashi, Y. Awaya, Y. Fukuyama, and M. Osawa. 2001. Severe myoclonic epilepsy in infants—A review based on the Tokyo Women’s Medical University series of 84 cases. Brain Dev 23: 736–748. Opdal, S. H., and T. O. Rognum. 2004. The sudden infant death syndrome gene: Does it exist? Pediatrics 114: e506–e512. Paterson, D. S. 2010. Medullary serotonergic abnormalities in sudden infant death syndrome: Implications in sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues. Chapter 5, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. New York, NY: Taylor & Francis. Paterson, D. S., F. L. Trachtenberg, E. G. Thompson et al. 2006. Multiple serotonergic brainstem abnormalities in sudden infant death syndrome. JAMA 296: 2124–2132. Pfeufer, A., S. Jalilzadeh, S. Perz et al. 2005. Common variants in myocardial ion channel genes modify the QT interval in the general population: Results from the KORA study. Circ Res 96: 693–701. Plant, L. D., P. N. Bowers, Q. Liu et al. 2006. A common cardiac sodium channel variant associated with sudden infant death in African Americans, SCN5A S1103Y. J Clin Invest 116: 430–435. Priori, S. G., and C. Napolitano. 1998. Impact of molecular biology on diagnosis and management of arrhythmogenic disorders. Rev Port Cardiol 17: 323–327.

282 Sudden Death in Epilepsy: Forensic and Clinical Issues Puranik, R., C. K. Chow, J. A. Duflou, M. J. Kilborn, and M. A. McGuire. 2005. Sudden death in the young. Heart Rhythm 2: 1277–1282. Quilichini, P. P., C. Chiron, Y. Ben Ari, and H. Gozlan. 2006. Stiripentol, a putative antiepileptic drug, enhances the duration of opening of GABA-A receptor channels. Epilepsia 47 (4): 704–716. Rodriguez-Calvo, M. S., M. Brion, C. Allegue, L. Concheiro, and A. Carracedo. 2008. Molecular genetics of sudden cardiac death. Forensic Sci Int 182: 1–12. Rogart, R. B., L. L. Cribbs, L. K. Muglia, D. D. Kephart, and M. W. Kaiser. 1989. Molecular cloning of a putative tetrodotoxin-resistant rat heart Na+ channel isoform. Proc Natl Acad Sci U S A 86: 8170–8174. Rugg-Gunn, F. J., and L. Nashef. 2009. Sudden unexpected death in epilepsy, 211–240. In Blue Book of Neurology: The Epilepsies, ed. S. D. Shorvon and T. Pedley. Philadelphia, PA: Saunders. Saenen, J. B., and C. J. Vrints. 2008. Molecular aspects of the congenital and acquired long QT syndrome: Clinical implications. J Mol Cell Cardiol 44: 633–646. Scheffer, I. E., L. A. Harkin, B. E. Grinton et al. 2007. Temporal lobe epilepsy and GEFS+ phenotypes associated with SCN1B mutations. Brain 130: 100–109. Schulze-Bahr, E. 2006. Mutation detection beyond: an example from congenital long QT syndrome. Heart Rhythm 3: 822–823. Schwartz, P. J., M. Stramba-Badiale, A. Segantini et al. 1998. Prolongation of the QT interval and the sudden infant death syndrome. N Engl J Med 338: 1709–1714. Sharp, A. J., H. C. Mefford, K. Li et al. 2008. A recurrent 15q13.3 microdeletion syndrome associated with mental retardation and seizures. Nat Genet 40: 322–328. Singh, N. A., C. Charlier, D. Stauffer et al. 1998. A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nat Genet 18: 25–29. Singh, R., I. E. Scheffer, K. Crossland, and S. F. Berkovic. 1999. Generalized epilepsy with febrile seizures plus: A common childhood-onset genetic epilepsy syndrome. Ann Neurol 45: 75–81. Sisodiya, S., J. H. Cross, I. Blumcke et al. 2007. Genetics of epilepsy: Epilepsy Research Foundation Workshop Report. Epileptic Disord 9: 194–236. Skinner, J. R., B. Chong, M. Fawkner, D. R. Webster, and M. Hegde. 2004. Use of the newborn screening card to deἀne cause of death in a 12-year-old diagnosed with epilepsy. J Paediatr Child Health 40: 651–653. Splawski, I., K. W. Timothy, M. Tateyama et al. 2002. Variant of SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science 297: 1333–1336. Steinlein, O. K., J. C. Mulley, P. Propping et al. 1995. A missense mutation in the neuronal nicotinic acetylcholine receptor alpha 4 subunit is associated with autosomal dominant nocturnal frontal lobe epilepsy. Nat Genet 11: 201–203. Steinlein, O. K. 2008. Genetics and epilepsy. Dialogues Clin Neurosci 10: 29–38. Stöllberger, C., and J. Finsterer. 2004. Cardiorespiratory ἀndings in sudden unexplained/unexpected death in epilepsy (SUDEP). Epilepsy Res 59 (1): 51–60. Striano, P., L. Bordo, M. L. Lispi et al. 2006. A novel SCN2A mutation in family with benign familial infantile seizures. Epilepsia 47: 218–220. Sudden Death in Chromosome 15q11-q13 Duplication Syndrome. 2009. http://www.idic15.org/ dr_physicianadvisory.php (accessed April 3, 2009). Tan, H. L., N. Hofman, I. M. van Langen, A. C. van der Wal, and A. A. Wilde. 2005. Sudden unexplained death: Heritability and diagnostic yield of cardiological and genetic examination in surviving relatives. Circulation 112: 207–213. Tellez, J. O., H. Dobrzynski, I. D. Greener et al. 2006. Differential expression of ion channel transcripts in atrial muscle and sinoatrial node in rabbit. Circ Res 99: 1384–1393. Tester, D. J., and M. J. Ackerman. 2005. Sudden infant death syndrome: How signiἀcant are the cardiac channelopathies? Cardiovasc Res 67: 388–396. Timmings, P. L. 1998. Sudden unexpected death in epilepsy: Is carbamazepine implicated? Seizure 7: 289.

Genetics of Sudden Death in Epilepsy

283

Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47: 21–26. Van Norstrand, D. W., and M. J. Ackerman. 2009. Sudden infant death syndrome: Do ion channels play a role? Heart Rhythm 6: 272–278. Wallace, R. H., D. W. Wang, R. Singh et al. 1998. Febrile seizures and generalized epilepsy associated with a mutation in the Na+-channel beta1 subunit gene SCN1B. Nat Genet 19: 366–370. Watanabe, H., T. T. Koopmann, S. Le Scouarnec et al. 2008. Sodium channel beta1 subunit mutations associated with Brugada syndrome and cardiac conduction disease in humans. J Clin Invest 118: 2260–2268. Weese-Mayer, D. E., E. M. Berry-Kravis, L. Zhou et al. 2004. Sudden infant death syndrome: Casecontrol frequency differences at genes pertinent to early autonomic nervous system embryologic development. Pediatr Res 56: 391–395. Wick, R., and R. W. Byard. 2006. Mechanisms of unexpected and/or sudden death in Lafora disease. Forensic Sci Int 163: 144–147.

Cardiac Channelopathies and Sudden Death Benito Herreros

19

Contents 19.1 Introduction 19.1.1 Sudden Cardiac Death in Subjects with a Structurally Normal Heart 19.1.2 Ion Channels and Channelopathies 19.1.3 The Cardiac Action Potential 19.1.4 Inherited Primary Arrhythmia Syndromes 19.2 Long QT Syndrome 19.2.1 Clinical Manifestations 19.2.2 Genetic Background and Pathophysiology 19.2.3 Risk Stratiἀcation and Management 19.3 Brugada Syndrome 19.3.1 Clinical Manifestations 19.3.2 Genetic Background and Pathophysiology 19.3.3 Risk Stratiἀcation and Management 19.4 Catecholaminergic Polymorphic Ventricular Tachycardia 19.4.1 Clinical Manifestations 19.4.2 Genetic Background and Pathophysiology 19.4.3 Management 19.5 Short QT Syndrome 19.5.1 Clinical Manifestations 19.5.2 Genetic Background and Pathophysiology 19.5.3 Management 19.6 Considerations about Cardiac Channelopathies, Epilepsy, and SUDEP 19.6.1 Brugada ECG and Epilepsy 19.6.2 LQT and Epilepsy 19.6.3 The Hypothesis of a Common Channelopathy 19.6.4 SUDEP and Channelopathies References

285 286 286 287 287 289 289 290 291 291 292 293 294 295 295 295 296 296 296 297 297 297 298 298 299 299 299

19.1â•…Introduction Cardiac channelopathies are genetically determined diseases that may cause sudden arrhythmic death in young subjects with a structurally normal heart. Abnormal cellular electrophysiology related to dysfunctional ion channels is the underlying substrate. Some types of idiopathic epilepsy are also caused by ion channel dysfunction in the neuronal cell membrane. This shared pathophysiologic mechanism lets us hypothesize about an arrhythmic cause for some cases of sudden death in epilepsy. In this chapter, cardiac 285

286 Sudden Death in Epilepsy: Forensic and Clinical Issues

channelopathies related to sudden death are reviewed and data supporting a potential relationship with epilepsy are exposed in the last section. 19.1.1â•… Sudden Cardiac Death in Subjects with a Structurally Normal Heart Sudden cardiac death (SCD) is deἀned as a natural death from cardiac causes, heralded by abrupt loss of consciousness within 1 hour of the onset of an acute change in cardiovascular status. Preexisting heart disease may or may not be present, but the time and mode of death are unexpected. It is likely that ventricular ἀbrillation (VF), or ventricular tachycardia (VT) deteriorating to VF, is the initiating event in most cardiac arrests (Myerburg and Castellanos 2008). These arrhythmias most often occur in the presence of a cardiac structural abnormality that is responsible for disturbances in impulse initiation and conduction. Ischemic heart disease is the most common cause, followed by other structural heart diseases such as hypertrophic, dilated, or arrhythmogenic right ventricular cardiomyopathy, but malignant arrhythmias leading to SCD can also happen in some patients with a structurally normal heart. In several postmortem series of cardiac arrest victims, no structural abnormality was found in 5% to 8% of cases (Priori et al. 2003). This occurrence used to be designated as “idiopathic VF.” The etiology of SCD in these cases may be a primary electrical disorder of a genetic basis (and thus can be inherited) that predisposes to rhythm alteration. In fact, SCD without structural heart disease is proportionally more common at younger ages, in contrast with the adult age, in which the main cause of SCD, by far, is coronary artery disease (Myerburg 2001) (Table 19.1). During the past 15 years, with the help of molecular biology, the substrate of idiopathic VF has been better deἀned and genetically determined abnormalities of proteins that control the electrical activity of the heart have been demonstrated to cause cardiac arrest in the structurally intact heart (Priori et al. 2008). 19.1.2â•…Ion Channels and Channelopathies Each normal heartbeat is initiated by a pulse of electrical excitation that begins in a group of specialized pacemaker cells and subsequently spreads throughout the heart. This electrical impulse is made possible by the electrochemical gradient that exists across the cell membrane of the cardiomyocyte. At rest, the electrochemical potential inside the cell is negative with respect to the outside due to energy-consuming processes, such as the Na+/K+ ATPase, which maintain the ionic gradients. The resting cell membrane is relatively permeable to K+, but much less so to Na+ and Ca2+. During electrical excitation, the membrane becomes permeable to Na+ ions and the electrochemical gradient reverses (depolarization). Next, Ca2+ ions move into the cell to activate the contractile process. Finally, the negative

Table 19.1â•… Etiology of Sudden Cardiac Death Coronary artery disease Cardiomyopathies Valvular, inflammatory, and inἀltrative cardiac diseases Normal hearts/idiopathic VF

80% 10%–15% ±5% ?%

Source: Myerburg, R.J., J Cardiovasc Electrophysiol, 12, 369–381, 2001. With permission.

Cardiac Channelopathies and Sudden Death

287

membrane potential is restored, mainly due to the efflux of K+, and thus the wave of depolarization is self-limiting (Marbán 2002). All these changes in permeability are accomplished by the opening and closing of ion channels that are speciἀc for the individual ions. Ion channels are transmembrane proteins that allow the passage of ions from the interior of the cell to outside and vice versa (Figure 19.1). The movements of ions are passive down their respective concentration gradients. Some channels are tissue-speciἀc (e.g., Na+ channels have several isoforms with different expression in the heart, skeletal muscle, central, or peripheral nervous system), whereas others are widely distributed throughout the body. Ion channels are basic to the processes of electrical signaling and excitation essential to the functioning of the heart. Furthermore, for the normal function of any given ion channel, not only its protein subunits (alpha, beta), but also multiple other gene products with various functions (e.g., phosphorylation, assembly, posttranslational modiἀcation, anchoring units) are necessary. Mutation in any of these genes can cause ion channel dysfunction, or channelopathy, which may provoke cardiac arrhythmias and SCD as the most severe clinical expression (Roden et al. 2002). This chapter reviews the most common inherited arrhythmogenic diseases that may cause sudden death in subjects with a structurally normal heart. As most of them are due to dysfunction of ion channels participating in the cardiomyocyte excitation, it is important to review the cardiac action potential and its involved ionic currents. These are summarized in Figure 19.1. 19.1.3â•… The Cardiac Action Potential The cardiac action potential is a graphic representation of the stereotypical voltage changes against time, which follow one another when an appropriate voltage stimulus reaches the cardiomyocyte (Figure 19.1). Thus, the action potential is a reflection of the electrical activity of a single cardiac cell. A sharp depolarizing upstroke (phase 0) is the result of rapid inflow of Na+ ions through voltage-dependent Na+ channels. Phase 1 reflects the activation of a transient outward current (Ito), mostly carried by K+, which produces a notch or early repolarization soon after the initial depolarizing upstroke. During the plateau (phase 2), the net influx of Ca2+ is balanced by the efflux of K+ through several K+ channels (IKr, IKs, IK1, Ito). Phase 3 is the downslope after the plateau, representing the late depolarization, due to the chemical forces, that favors the efflux of K+ predominating over the electrostatic forces that would favor its influx through the same channels. Phase 4 represents the resting potential, during which there tends to be a net diffusion (efflux) of K+ in the direction of its concentration gradient through IK1 channels (Berne and Levy 1992). 19.1.4â•…Inherited Primary Arrhythmia Syndromes A number of genes are associated with inherited arrhythmia syndromes that predispose to sudden death in individuals with a structurally normal heart (Table 19.2). Studies in large cohorts suggest that cardiac channelopathies could be responsible for 35% of sudden deaths in the young and in 9% of cases of sudden infant death syndrome (SIDS) (Schwartz and Crotti 2007). Recently, some authors have proposed a classiἀcation of inherited primary arrhythmia syndromes according to their pathophysiologic basis; that is, into Na+ channelopathies, K+ channelopathies, and so forth (Lehnart et al. 2007; Wilde 2008). This chapter uses the traditional clinical nomenclature to review the four main inherited arrhythmia

288 Sudden Death in Epilepsy: Forensic and Clinical Issues

(a)

Na+

K+ (b)

K+

Ca2+

Na+

Ca2+

Probable gene SCN5A

INa ICa, L

CACNA1c

INa/Ca

NCX1

0

IK1

KCNJ2

It01

KCND

It02

?

IKr

KCNH2/KCNE2

IKs

KCNQ1/KCNE1

IKp

KCNK?

Figure 19.1╇ Ion channels underlie cardiac excitability. (a) The key ion channels (and an elec-

trogenic transporter) in cardiac cells. K+ channels mediate K+ efflux from the cell; Na+ channels and Ca 2+ channels mediate Na+ and Ca 2+ influx, respectively. The Na+/Ca 2+ exchanger transports three Na+ ions for each Ca 2+ ion across the surface membrane. (b) Ionic currents and genes underlying the cardiac action potential. Top, depolarizing currents as functions of time, and their corresponding genes; center, a ventricular action potential; bottom, repolarizing currents and their corresponding genes. (From Marbán, E., Nature, 415, 213–218, 2002. With permission.)

Cardiac Channelopathies and Sudden Death

289

Table 19.2â•… Genes Involved in Arrhythmogenic Cardiac Channelopathies Related to SCD Syndrome

Subtype

Long QT syndrome

LQT1 LQT2 LQT3 LQT4

KCNQ1 KCNH2 SCN5A ANK2

IKs K+ channel α subunit IKr K+ channel α subunit Na+ channel α subunit Anchoring protein ankyrin B

LQT5 LQT6 LQT7 LQT8 LQT9 LQT10 LQT11

KCNE1 KCNE2 KCNJ2 CACNA1c CAV3 SCN4B AKAP9

Brugada syndrome

BrS1 BrS2

SCN5A GPD1L

Catecholaminergic VT

CPVT1 CPVT2 SQT1 SQT2 SQT3

RyR2 CASQ2 KCNH2 KCNQ1 KCNJ2

IKs K+ channel β subunit IKr K+ channel β subunit IK1 K+ channel α subunit ICa,L Ca2+ channel α subunit Caveolin 3 Na+ channel β4 subunit Yotiao (A-kinase anchoring protein) Na+ channel α subunit Glycerol-3-P dehydrogenase 1 like protein Cardiac ryanodine receptor Cardiac calsequestrin IKr K+ channel α subunit IKs K+ channel α subunit IK1 K+ channel α subunit

Short QT syndrome

Gene

Protein

Effect IKs loss of function IKr loss of function INa gain of function Reduction of several ionic currents IKs loss of function IKr loss of function IK1 loss of function ICa,L gain of function INa gain of function INa gain of function IKs loss of function INa loss of function INa loss of function Citoplasmic Ca2+ overload Citoplasmic Ca2+ overload IKr gain of function IKs gain of function IK1 gain of function

syndromes which, to the moment, have been shown to be responsible for sudden death: long QT syndrome (LQTS), Brugada syndrome (BrS), short QT syndrome (SQTS), and catecholaminergic polymorphic ventricular tachycardia (CPVT). There are other inherited arrhythmic syndromes not necessarily linked to SCD that are responsible for atrial arrhythmias, sick sinus syndrome, and progressive cardiac conduction disease (PCCD).

19.2â•… Long QT Syndrome The congenital form of the LQTS is characterized by prolongation of the QT interval on the ECG and susceptibility to malignant ventricular arrhythmias. Two major forms have been described: one inherited as an autosomal dominant trait (Romano–Ward syndrome), and the other (more rare) transmitted as an autosomal recessive trait (Jervell and Lange– Nielsen syndrome) with associated neurosensorial deafness in affected individuals. 19.2.1â•… Clinical Manifestations The characteristic ECG in LQTS shows a prolonged QT interval, often with T-wave morphological abnormalities, and occasionally polymorphic VT of the torsades de pointes type. Clinical manifestations are recurrent syncope or fainting and sudden death, often precipitated by stress (fear, loud noise, sudden awakening), but they can also occur at rest (Schwartz et al. 2001). Typical onset of symptoms occurs in the ἀrst two decades of life. In the neonatal period, LQTS can be responsible for a certain number of SIDS (Arnestad et al.

290 Sudden Death in Epilepsy: Forensic and Clinical Issues

2007). Since the clinical presentation varies greatly, ranging from asymptomatic mutation carriers with normal QT interval to typical cases with marked QT prolongation and recurrent syncope, diagnostic criteria have been developed with a scoring system that allows classiἀcation of patients into categories of low, intermediate, or high probability of having LQTS (Schwartz et al. 1993). With the advent of the DNA diagnosis, these scoring systems have been shown to have less sensitivity than measurement of QTc alone (Napolitano et al. 2003; Hofman et al. 2007) for the detection of mutation carriers. Furthermore, an important proportion of mutation carriers have a normal QT interval (Priori et al. 2003), so molecular diagnosis has become part of the routine clinical management of LQTS. 19.2.2â•… Genetic Background and Pathophysiology At the time of writing, 11 genes have been proven to be or are thought to be associated with LQTS (Lehnart et al. 2007) (Table 19.2). However, LQT1, LQT2, and LQT3 variants account for more than 90% of all genotyped LQTS patients, whereas the remaining genes are responsible for a minority of cases (Priori et al. 2008). Most of these genes encode cardiac ion channels subunits. If not, they modulate ionic currents. In patients with LQT1, there is loss of function of the IKs potassium channel due to a mutation in the gene KCNQ1, which encodes the alpha subunit of IKs channel. The result is a reduced K+ outward current in phases 2 and 3 of the action potential, leading to an abnormal repolarization and reduced rate-dependent shortening of the action potential (Roden et al. 2002). Mutations in KCNQ1 resulting in IKs loss of function accounts for 40% to 50% of all genotyped LQT mutations (Splawski et al. 2000; Tester et al. 2005). The same channel IKs exhibits loss of function in patients with LQT5, but is due to mutation in gene KCNE1, which encodes its beta subunit. This variant is much less frequent than LQT1, but it is noteworthy that both genes KCNQ1 and KCNE1 are also expressed in the inner ear, where K+ channel function contributes to production of endolymph. That is the reason why mutations in both genes may cause the Jervell and Lange–Nielsen syndrome (LQT and bilateral deafness) in homozygous subjects (Neyroud et al. 1997; Schulze-Bahr et al. 1997). Patients with LQT2 exhibit loss of function of the IKr potassium channel due to a mutation in gene KCNH2, which encodes the alpha subunit of IKr channel. These mutations represent 35% to 45% of the genotyped LQTS mutations. The beta subunit of IKr channel is encoded by KCNE2, whose mutations are the substrate for the rarer LQT6 (Abbot et al. 1999). Whereas K+ channels mutations involved in LQTS produce a loss of function effect, a gain of function mutation in SCN5A encoding the alpha subunit of the cardiac Na+ channel is the substrate for the variant LQT3. As a consequence, persistent inward Na+ current during phase 2 (plateau) prolongs the action potential. LQT3 accounts for 2% to 8% of all LQTS patients (Splawski et al. 2000; Tester et al. 2005). Overall, penetrance of LQTS is 60%, but it is not the same for speciἀc variants. In LQT1 the proportion of mutation carriers without QTc prolongation is as high as 36%, in LQT2 it is 19%, and in LQT3 it is 10% (Priori et al. 2003). There are some speciἀc phenotypic features in the different variants. Patients with LQT1 have increased risk of developing VT during exercise (75%, especially swimming) or emotional stress (15%). IKs current is activated by fast heart rates and by catecholamines, so under these circumstances, LQT1 patients cannot shorten their QT interval appropriately. Accordingly, beta blocker therapy should be most effective in LQT1 patients (Schwartz et al. 2001). The ECG in LQT1 is characterized by long T wave duration (Figure 19.2). In LQT2 patients, cardiac events

Cardiac Channelopathies and Sudden Death LQT3

LQT2

291 LQT1

II aVF

V5

Figure 19.2╇ Electrocardiographic distinctive features in long QT syndrome variants. Left

panel: note the long ST segment with late onset of T wave in an LQT3 patient. Center panel: low amplitude T waves in an LQT2 patient. Right panel: early onset of broad-based T waves in an LQT1 patient. (Modified from Moss, A. J. et al., Circulation, 92 (10), 2929–2934, 1995. With permission.)

are often triggered by emotions, sudden loud noises, and acute arousal (37%). They are also at€risk during sleep or at rest (63%) but not during exercise. Their ECGs show small (low-amplitude), notched, or biphasic T waves (Figure 19.2). Finally, LQT3 patients suffer cardiac events usually at night, sleep, or rest (80%), and occasionally with emotions (15%) or during exercise (5%) (Schwartz et al. 2001). Typical ECG in LQT3 shows a flat, long ST segment with late onset of a narrow, peaked T wave (Figure 19.2). 19.2.3â•…Risk Stratification and Management Highest risk patients are those who have QTc ≥ 500 ms. Other ECG markers of bad prognosis are T wave alternans and torsades de pointes VT. Those who have survived after a cardiac arrest, and those who have recurrent syncope despite beta blockers, usually receive an implantable cardioverter deἀbrillator (ICD) (Zareba et al. 2003). On the other hand, the increasing knowledge of clinical evolution in genotyped patients has demonstrated that the different LQTS variants have distinct prognoses. LQT2 and LQT3 patients have a lower event-free survival than LQT1. Gender has influence in LQT2 (higher risk in females) and LQT3 patients (higher risk in males), but not in LQT1 (Priori et al. 2003). Once a diagnosis of LQTS has been made, lifestyle modiἀcation is recommended Â�(e.€g., avoidance of swimming among LQT1 patients, no use of loud alarm clocks in LQT2, prohibition of drugs with QT prolonging effects). Experts recommend beta blocker therapy in all symptomatic patients and in asymptomatic patients younger than 40 years with a clearly prolonged QTc (in which risk for SCD if untreated is around 13%) (Priori et al. 2003). Implantation of a deἀbrillator is recommended after a resuscitated cardiac arrest and in those patients who experience syncope or VT under beta blocker therapy (Zipes et al. 2006). Also, the recognition of the speciἀc arrhythmic risk of the different LQTS variants will probably have an increasing impact on clinical management.

19.3â•… Brugada Syndrome BrS is characterized by an ST segment elevation in the right precordial ECG leads and a high incidence of SCD in patients with structurally normal hearts. Sudden death is

292 Sudden Death in Epilepsy: Forensic and Clinical Issues

provoked by ventricular arrhythmias (polymorphic VT leading to ventricular ἀbrillation) that can also produce syncope when they are self-limited. Inheritance of BrS occurs via an autosomal dominant mode of transmission. 19.3.1â•… Clinical Manifestations Characteristic Brugada ECG shows a coved ST segment elevation ≥2 mm followed by a negative T wave in more than one right precordial lead (V1–V3, Figure 19.3). This is the only accepted diagnostic pattern (the so-called type 1), although two more ECG types have been described: type 2 has a saddleback ST elevation with a J point ≥ 2 mm, a trough ≥ 1 mm, and a positive/biphasic T wave; and in type 3, ST elevation is <1 mm. Both types 2 and 3 are only diagnostic if conversion to type 1 is observed, either spontaneously or after sodium channel blocker administration. In fact, class Ic antiarrhythmic drugs (ajmaline, flecainide, propafenone, pilsicainide) are used as a diagnostic tool to unmask concealed forms, as the diagnosis of BrS is often made difficult because of the intermittent nature of the ECG pattern. If a BrS patient exhibits a normal or nondiagnostic (types 2 or 3) ECG, an intravenous dose of an Ic antiarrhythmic drug can exacerbate cardiac sodium channel dysfunction so that his or her ECG may temporarily turn into diagnostic, type 1 Brugada ECG. To make the deἀnitive diagnosis of BrS, a type I ECG must be recorded in conjunction with one of the following criteria: documented VF, polymorphic VT, a family history of SCD at <45 years old, type 1 ECG in family members, inducibility of VT with programmed electrical stimulation, and syncope or nocturnal agonal respiration (Antzelevitch et al. 2005). 5/2/99 type 1

13/2/99 type 2 type 3

V1

V2 V3 V4

V5 1 mV

V6 500 ms

Figure 19.3╇ Precordial leads of a resuscitated patient with Brugada syndrome. Note the dynamic changes in the course of 1 week. All three patterns are exhibited (see text for details). Arrows denote the J wave. (From Wilde, A. A. M. et al., Circulation, 106 (19), 2514–2519, 2002. With permission.)

Cardiac Channelopathies and Sudden Death

293

The mean age at diagnosis is about 40 years, and the male/female ratio in symptomatic individuals is 8:1 (Priori et al. 2008). BrS is more prevalent in Southeast Asia, in probable relationship to genetic factors. In fact, BrS has been proven to be the same entity that sudden unexplained nocturnal death syndrome (SUNDS), which is the leading cause of death in young, healthy Thai men (Nademanee et al. 1997; Vatta et al. 2002). VF and sudden death in BrS usually occurs at rest and at night (Figure 19.4). It is thought that circadian variation of sympathovagal balance, hormones, and metabolic factors may contribute to this circadian pattern (Mizumaki et al. 2004). Approximately 20% of patients with BrS develop supraventricular arrhythmias, mostly atrial ἀbrillation (AF). It is thought that the substrate responsible for the development of ventricular arrhythmias also may contribute to arrhythmogenesis in the atria of the heart (Francis and Antzelevitch 2008). In fact, the incidence of AF seems to be higher in patients with VF inducibility (Bordachar et al. 2004) and in patients with spontaneous episodes of VF (Kusano et al. 2008). 19.3.2â•… Genetic Background and Pathophysiology The ἀrst identiἀed gene whose mutation provoked BrS was SCN5A, which encodes the alpha subunit of the cardiac Na+ channel (Chen et al. 1998). Since then, more than 100 mutations in this gene have been described, producing a loss of function in the cardiac INa and, as a consequence, BrS (Antzelevitch 2007). However, only 20% to 25% of patients with BrS carry an SCN5A mutation (Priori et al. 2002), with a higher incidence in familial than in sporadic cases (Schulze-Bahr et al. 2003). This means that the genetic basis remains unclear for the majority of patients. On the other hand, in addition to the mutations of the Na+ channel itself, mutations of genes that modulate Na+ channel function are also associated with some cases of BrS. A second locus on chromosome 3, close but apart from the SCN5A locus, encodes GPD1-L, a protein that seems to affect trafficking of the cardiac Na+

Figure 19.4╇ Stored electrogram from an ICD implanted in a patient with Brugada syndrome.

At 03:39 a.m., rapid VT starts and degenerates into VF. The device delivers a shock and interrupts the arrhythmia. The patient was sleeping and did not notice the event. He was surprised when he was told what had happened, during a routine follow-up visit at the ICD clinic.

294 Sudden Death in Epilepsy: Forensic and Clinical Issues

channel to the cell surface, and mutation of this gene has been related to reduced INa and BrS (London et al. 2007). The supposed mechanism why patients with BrS are prone to ventricular malignant arrhythmias has been called phase 2 reentry (Lukas and Antzelevitch 1996). In normal subjects, epicardial cardiomyocites display a characteristic “notch” in phase 1 of the action potential, due to a large Ito current; but endocardial cells do not, so there is a little transmural voltage gradient during phase 1. In BrS patients, the pathologically reduced INa exaggerates this phenomenon, as the unopposed Ito results in accentuation of the epicardial cell’s notch. The transmural voltage gradient then extends to the plateau phase, and its ECG correlate is the ST segment elevation (Antzelevitch et al. 2003). Further INa reduction may shorten the action potential duration in some epicardial cells, adding a transmural dispersion of repolarization. Conduction of the action potential dome from sites where it is maintained to sites where it is lost (phase 2 reentry) may lead to a very closely coupled extrasystole in this transmural vulnerable window, triggering VT or VF. Loss of function mutations in SCN5A gene have also been related to other heart rhythm disturbances, such as PCCD, sinus node dysfunction, and atrial standstill. As previously stated, gain of function mutations in SCN5A produce LQT3 syndrome. Although these various arrhythmia syndromes were originally considered separate entities, recent evidence indicates more overlap in clinical presentation and biophysical defects of associated mutant channels than previously appreciated. Various SCN5A mutations are now known to present with mixed phenotypes, a presentation that has become known as overlap syndrome of cardiac sodium channelopathy (Remme et al. 2008). Today, the clinical impact of genetic testing in BrS is limited. It may help identify silent carriers for presymptomatic diagnosis of family members and genetic counseling, but based on current knowledge, it does not contribute to risk stratiἀcation (Zipes et al. 2006). 19.3.3â•…Risk Stratification and Management In BrS patients with high risk of SCD, ICD implantation is the only therapeutic option. There is general agreement that BrS patients who have suffered an episode of cardiac arrest have to be implanted with an ICD. There is also opinion in favor of efficacy of ICDs in patients with a spontaneous type I ECG who have suffered syncope or documented ventricular arrhythmia (Zipes et al. 2006). According to the second consensus document (Antzelevitch et al. 2005), seizure and nocturnal agonal respiration are considered as an equivalent of syncope with regard to therapeutic decision-making in BrS patients. Those patients in whom the type I ECG is only detected after a sodium channel blocking drug is administered are considered to be at lower risk by some authors (Priori et al. 2008), although the consensus document recommends ICD implantation should syncope (or its equivalents) occur. The most debated issue is risk stratiἀcation of asymptomatic individuals, especially regarding the role of programmed ventricular stimulation in search of ventricular arrhythmia inducibility. Early studies reported that inducible patients were at higher risk of SCD (Brugada et al. 2003), but in other series no association was found between inducibility and arrhythmic events (Eckardt et al. 2005). The latter is also the conclusion of two recent meta-analyses (Gehi et al. 2006; Paul et al. 2007). This lack of predictability of programmed ventricular stimulation, along with the high rate of ICD complications in BrS patients (Sacher et al. 2006) highlight the need for better risk assessment factors and alternative therapeutic options. Drugs with an Ito blocking effect, such as quinidine, are recognized as an effective

Cardiac Channelopathies and Sudden Death

295

treatment of BrS by the current practice guidelines (Zipes et al. 2006), but at the moment, ICD is the only accepted therapy for the prevention of SCD.

19.4â•… Catecholaminergic Polymorphic Ventricular Tachycardia CPVT is an inherited arrhythmia syndrome that produces VT and SCD in young patients with a structurally normal heart, typically related to physical or emotional stress. Genetic abnormalities in Ca2+ homeostasis inside the cardiomyocytes are the molecular substrate of this entity. 19.4.1â•… Clinical Manifestations The syndrome was originally described in the 1970s as a cause of ventricular arrhythmiamediated syncope in children (Coumel et al. 1978). The mean age at the onset of symptoms is 12 years (Priori et al. 2008). Syncope, triggered by exercise or acute emotion, is the typical symptom of CPVT, but up to 50% of patients can suffer a cardiac arrest as the ἀrst manifestation of this syndrome. Family history of SCD in individuals younger than 40 years is frequent. The resting ECG of these patients does not show any speciἀc abnormality, although sinus bradycardia is reported in some patients. The most characteristic feature of CPVT is the exercise dependence of the ventricular arrhythmias. During progressive exercise, ventricular ectopic beats appear at a given “threshold” heart rate. As the exercise protocol progresses, ventricular ectopy gets more frequent and complex, and progressively longer runs of nonsustained VT develop until it becomes sustained. The most characteristic form of CPVT is bidirectional VT (Figure 19.5), although irregular polymorphic VT is not infrequent. After stopping exercise, arrhythmias disappear progressively in the reverse order. Also, atrial ectopy and runs of supraventricular tachycardia during exercise are not uncommon. Untreated patients have a high risk of SCD (Leenhardt et al. 1995). 19.4.2â•… Genetic Background and Pathophysiology CPVT is caused by mutations of genes encoding for proteins responsible for the control of intracellular Ca2+. These mutations alter Ca2+ release from the sarcoplasmic reticulum of the cardiomyocyte. During the plateau phase of the action potential, Ca2+ enters the cell through voltage-gated L-type Ca2+ channels and serves as a trigger for a massive release of Ca2+ from the sarcoplasmic reticulum, in a process known as calcium-induced calcium release. The channels through which Ca2+ exits the sarcoplasmic reticulum are called cardiac ryanodine

Figure 19.5╇ Bidirectional ventricular tachycardia in a patient diagnosed with CPVT. Note

the alternate QRS complexes with opposite direction. (Modified from Mohamed, U. et al., J€Cardiovasc Electrophysiol 18, 791–797, 2007. With permission.)

296 Sudden Death in Epilepsy: Forensic and Clinical Issues

receptor (RyR2). Some mutations of the RyR2 gene encode defective channels that, in the presence of sympathic stimulation, behave with a gain of function of the Ca2+ release from the sarcoplasmic reticulum (Lehnart et al. 2004). The resulting Ca2+ overload would induce delayed afterdepolarizations and subsequent arrhythmia by a mechanism of triggered activity, in a similar fashion to digitalis toxicity (whose characteristic ventricular arrhythmia is also bidirectional VT) (Priori et al. 2001). Mutations of RyR2 gene are transmitted with an autosomal dominant pattern. More than 60 RyR2–CPVT mutations have been reported up to date, and account for 60% of CPVT patients (Priori et al. 2008). There is a second variant of CPVT, inherited as an autosomal recessive trait, which accounts for 1% to 2% of CPVT cases. Its molecular substrate is dysfunction of the cardiac calsequestrin (CASQ2), a protein that serves as the major Ca2+ reservoir within the sarcoplasmic reticulum of cardiomyocites (Mohamed et al. 2007). 19.4.3â•… Management Beta blockers are the drug of choice in CPVT. All patients with a clinical diagnosis of CPVT and also asymptomatic RyR2 mutation carriers should be treated, even in the absence of documented ventricular arrhythmias. Survivors of a cardiac arrest and patients who still develop VT despite beta blocker therapy should be implanted with an ICD (Zipes et al. 2006).

19.5â•… Short QT Syndrome SQTS is best deἀned as an inheritable, primary electrical heart disease that is characterized by (1) an abnormally short QT interval and (2) paroxysmal atrial and/or ventricular tachyarrhythmias resulting from an accelerated cardiac (atrial and ventricular) repolarization due to congenital (genetically heterogeneous) cardiac channelopathies (Gussak and Bjerregaard 2005). 19.5.1â•… Clinical Manifestations SQTS was ἀrst described by Gussak et al. (2000), and the number of reported patients is small. Most patients have a family history of SCD (Gaita et al. 2003). Possible symptoms range from palpitations and syncope to SCD, provoked by AF or ventricular malignant arrhythmias, respectively. The age of reported SCD victims vary between a few months to 70 years, and cases of SIDS have been reported in which typical mutations of SQTS have been found (Rhodes et al. 2006). The hallmark of SQTS is a short QT interval, associated with other characteristics as detailed in Figure 19.6: impaired adjustment of the QT length to changes in heart rate; very short or no ST segment; tall, peaked T waves; and occasionally, depression of P–R interval. An absolute QT value below 320 ms should be considered as a possible SQTS. Due to the scarce rate dependence of QT in this syndrome, QTc will fail to reflect the magnitude of QT shortening at elevated heart rates, leading to a false negative diagnosis. Secondary causes of QT shortening (hypercalcemia, hyperkalemia, hyperthermia, acidosis, digoxin therapy) must be excluded before diagnosing SQTS. SQTS should be suspected in all patients with a short QT interval (less than 350 ms), lone AF, primary VF, family history of SCD, or family history of SQTS (Gussak et al. 2005).

Cardiac Channelopathies and Sudden Death

297

Figure 19.6╇ ECG from a patient with SQTS. Note the short QT interval with tall, peaked T waves in precordial leads. (Adapted from Gaita et al., Circulation 108, 965–970, 2003. With permission.)

19.5.2â•… Genetic Background and Pathophysiology Three different genes with mutations causing SQTS have been identiἀed. All of them involve K+ channels, and, as one could expect, gain of function K+ currents are the result of the different mutations responsible for this syndrome (Table 19.2). The three SQTS genes are involved, with opposite gene mutation effects, in LQTS. Unlike with LQTS, today there are not enough data to link the different mutations to any speciἀc symptoms (no genotype–phenotype correlation); nevertheless a nomenclature of three different variants (SQT1, SQT2, and SQT3) based on the genetic ἀndings has been proposed (Lehnart et al. 2007). Currently, there is no consensus on the arrhythmogenic mechanisms in SQTS. The different K+ channel mutations lead to a shortening of the action potential duration, and, not surprisingly, when patients have undergone electrophysiological study, typical ἀndings are short atrial and ventricular refractoriness along with easy atrial and VF inducibility (Gaita et al. 2003). 19.5.3â•… Management Given the reduced available cohort of patients, the management of patients with SQTS is still poorly deἀned. No speciἀc recommendations are stated in the current practice guidelines for prevention of SCD (Zipes et al. 2006). Treatment with quinidine can result in normalization of the QT, prolongation of ventricular refractoriness and turning previously inducible VF into noninducible VF (Gaita et al. 2004). Nevertheless, experts recommend cardiac deἀbrillator implantation in view of the apparent high risk for SCD of these patients (Gussak et al. 2005).

19.6â•…Considerations about Cardiac Channelopathies, Epilepsy, and SUDEP From a clinical point of view, cardiac channelopathies and epilepsy may manifest with similar symptoms. A convulsive syncope, originated by transient cerebral hypoperfusion related to a self-limited cardiac arrhythmia, may be difficult to distinguish from an epileptic seizure. In both cases the patient (who very often is a child or a young patient) may

298 Sudden Death in Epilepsy: Forensic and Clinical Issues

suffer a loss of consciousness, with tonic–clonic convulsive movements and stertorous breathing that lasts for seconds to minutes. These resemblances have lead in many cases to misdiagnosis of a primary cardiac arrhythmic disorder as epilepsy. This misdiagnosis is well documented in the medical literature with LQTS (Davis et al. 1998), BrS (Skinner et al. 2007), and CPVT (Lucet et al. 1983). These categories have been erroneously labeled as “epilepsy” in a number of patients, as is reflected in several case reports. Although some clinical features may help to distinguish between an “arrhythmic” seizure and an epileptic seizure (McKean et al. 2006), this distinction is not always easy if we only take into account the patient’s symptoms. Even if we have an ECG from a patient, the differential diagnosis may be difficult due to the possibility of displaying a normal ECG despite carrying a primary electrical heart disorder. This is the reason why detailed patient history and witness account, followed by physical examination and the appropriate use of diagnostic testing, is of paramount importance when a young patient presents with a ἀrst episode of seizure. The opposite setting—epilepsy misdiagnosed as a cardiac arrhythmia—would happen if heart rhythm alterations are recorded in the ECG during an epileptic seizure, or if the patient interictal ECG displays any abnormality that may lead us to think about an arrhythmic disorder as the cause for his or her symptoms. The occurrence of sinus tachycardia, bradycardia, asystole, and other ECG alterations during an epileptic seizure is well documented (Rugg-Gunn et al. 2004) and attributed to the autonomic influence on cardiovascular function during an ictal episode (Freeman 2006). However, this is not a frequent cause of epilepsy misdiagnosed as cardiac arrhythmia (Kothari 1990). More recently, some reports of patients with epilepsy have been misdiagnosed as a cardiac channelopathy because of exhibiting typical ECG features (Girard et al. 2008). These observations let us hypothesize about possible pathophysiologic relationships between epilepsy and an abnormal ECG and are discussed in the following sections. 19.6.1â•… Brugada ECG and Epilepsy Recently, the case of a woman with a Brugada ECG and recurrent seizures has been reported, in which the simultaneous video-EEG and cardiac telemetry monitoring allowed her physicians to discard a cardiac arrhythmia as the cause for her symptoms, conἀrming a diagnosis of epilepsy with a Brugada ECG (Girard et al. 2008). Another case with a high suspicion of coincidence of Brugada ECG and epilepsy (not demonstrated because of the absence of ECG recording during symptoms) has been previously published (Fauchier et al. 2000). These and other similar cases lead us to think that there can be a common pathophysiologic substrate to explain this clinical picture (epilepsy and a Brugada ECG). 19.6.2â•…LQT and Epilepsy LQTS has been reported as a frequent cause of erroneous epilepsy diagnosis (Davis et al. 1998). But there are studies in “real” epileptic children that demonstrate that the QT interval is prolonged temporarily after a seizure (Kandler et al. 2005). The signiἀcance of these ἀndings is not known, but it could be owing to transient changes in autonomic tone (Freeman 2006). More recently, analysis of the prevalence of seizures in the different variants of a cohort with a genetically conἀrmed diagnosis of LQTS concluded that a personal history of seizures was much more common in LQTS2 than in any other variant.

Cardiac Channelopathies and Sudden Death

299

The authors infer that KCNH2 mutations may confer susceptibility for seizure activity (Johnson et al. 2009). 19.6.3â•… The Hypothesis of a Common Channelopathy Dysfunction of both K+ and Na+ channels is the substrate for some genetic epileptic syndromes (Graves 2006). The coincidence in a given patient of a characteristic ECG suggestive of a cardiac channelopathy (e.g., Brugada or LQT syndromes) along with a certain diagnosis of epilepsy, raise the hypothesis of a common explanation for the cardiac and neurologic phenotype owing to the same channel dysfunction. In the case of Na+ channels, it has been proved that neuronal isoforms are expressed in the heart (Haufe et al. 2007), and cardiac isoforms are expressed in the nervous system (Donahue et al. 2000). Another striking clinical observation is the fact that in BrS there are two triggers that can unmask the typical ECG and also induce arrhythmia: sleep and fever (Smits et al. 2005), which are also triggers for seizures in patients with several types of epilepsy (Frucht et al. 2000). These modulating factors, which are shared by BrS and idiopathic epilepsy, point to some common pathophysiologic mechanism. 19.6.4â•… SUDEP and Channelopathies Patients with epilepsy may die unexpectedly, with or without evidence of a seizure, and postmortem examination may not reveal a structural or toxicological cause for death. Although case-control studies have identiἀed certain risk factors, little is known about the exact mechanism for death. The well-known heart rhythm disturbances that may occur during an epileptic seizure have been previously suggested as a cause for sudden unexpected death in epilepsy (SUDEP) (Nei et al. 2000; Opherk et al. 2002). The possibility of an underlying channelopathy that makes the patient prone to both epilepsy and arrhythmia leading to SCD is also supported by some published data. One is the postmortem ἀnding of SCN5A mutations (characteristic of LQTS3 and BrS) in patients with SUDEP and a previous diagnosis of idiopathic epilepsy (Aurlien et al. 2008). Another is the occurrence of several cases of SUDEP in a family with a diagnosis of GEFS+ (a type of genetic idiopathic epilepsy) and a novel mutation in SCN1A gene (one of the neuronal Na+ channel isoforms, which is also expressed in the heart) (Hindocha et al. 2008). SUDEP is probably a heterogeneous syndrome with multiple causes, as discussed in this book, and cardiac arrhythmia is thought to be one of them. Further clinical, epidemiological, and genetic studies are needed to go deeply into its underlying pathophysiologic mechanisms.

References Abbott, G. W., F. Sesti, I. Splawski et al. 1999. MiRP1 forms IKr potassium channels with HERG and is associated with cardiac arrhythmia. Cell 97: 175–187. Antzelevitch, C. 2007. Genetic basis of Brugada syndrome. Heart Rhythm 4: 756–757. Antzelevitch, C., P. Brugada, M. Borggrefe et al. 2005. Brugada syndrome. Report of the Second Consensus Conference. Circulation 111: 659–670. Antzelevitch, C., P. Brugada, J. Brugada, R. Brugada, J. A. Towbin, and K. Nademanee. 2003. Brugada syndrome: 1992–2002. A historical perspective. JACC 41: 1665–1671. Arnestad, M., L. Crotti, T. O. Rognum et al. 2007. Prevalence of long-QT syndrome gene variants in sudden infant death syndrome. Circulation 115: 361–367.

300 Sudden Death in Epilepsy: Forensic and Clinical Issues Aurlien, D., T. P. Leren, E. Taubøll, and L. Gjerstad. 2008. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18 (2): 158–160 (Aug 25, Epub ahead of print). Berne, R. M., and M. N. Levy. 1992. Cardiovascular Physiology. St. Louis, MO: Mosby-Year Book. Bordachar, P., S. Reuter, S. Garrigue et al. 2004. Incidence, clinical implications and prognosis of atrial arrhythmias in Brugada syndrome. Eur Heart J 25: 879–884. Brugada, J., R. Brugada, and P. Brugada. 2003. Determinants of sudden cardiac death in individuals with the electrocardiographic pattern of Brugada syndrome and no previous cardiac arrest. Circulation 108: 3092–3096. Chen, Q., G. E. Kirsch, D. Zhang et al. 1998. Genetic basis and molecular mechanism for idiopathic ventricular ἀbrillation. Nature 392: 293–296. Coumel, P., J. Fidelle, V. Lucet, P. Attuel, and Y. Bouvrain. 1978. Catecholaminergic induced severe ventricular arrhythmias with Adams–Stokes syndrome in children: Report of four cases. Br Heart J 40: 28–37. Davis, A. M., J. L. Wilkinson. 1998. The long QT syndrome and seizures in childhood. J Paediatr Child Health 34: 410–411. Donahue, L. M., P. W. Coates, V. H. Lee, D. C. Ippensen, S. E. Arze, and S. E. Poduslo. 2000. The cardiac sodium channel mRNA is expressed in the developing and adult rat and human brain. Brain Res 887: 335–343. Eckardt, L., V. Probst, J. P. Smits et al. 2005. Long-term prognosis of individuals with right precordial ST-segment-elevation Brugada syndrome. Circulation 111: 257–263. Fauchier, L., D. Babuty, and P. Cosnay. 2000. Epilepsy, Brugada syndrome and the risk of sudden unexpected death. J Neurol 247: 643–644. Francis, J., and C. Antzelevitch. 2008. Atrial ἀbrillation and Brugada syndrome. J Am Coll Cardiol 51: 1149–1153. Freeman, R. 2006. Cardiovascular manifestations of autonomic epilepsy. Clin Auton Res 16: 12–17. Frucht, M. M., M. Quigg, C. Schwaner, and N. B. Fountain. 2000. Distribution of seizure precipitants among epilepsy syndromes. Epilepsia 41: 1534–1539. Gaita, F., C. Giustetto, F. Bianchi et al. 2003. Short QT syndrome. A familial cause of sudden death. Circulation 108: 965–970. Gaita, F., C. Giustetto, F. Bianchi et al. 2004. Short QT syndrome: Pharmacological treatment. J Am Coll Cardiol 43: 1494–1499. Gehi, A. K., T. D. Duong, L. D. Metz, J. A. Gomes, and D. Mehta. 2006. Risk stratiἀcation of individuals with the Brugada electrocardiogram: a meta-analysis. J Cardiovasc Electrophysiol 17: 577–583. Girard, S., A. Cadena, and Y. Acharya. 2008. Woman with Brugada syndrome and epilepsy: A unifying diagnosis? South Med J 101: 1150–1153. Graves, T. D. 2006. Ion channels and epilepsy. Q J Med 99: 201–217. Gussak, I., and P. Bjerregaard. 2005. Short QT syndrome—5 years of progress. J Electrocardiol 38: 375–377. Gussak, I., P. Brugada, J. Brugada et al. 2000. Idiopathic short QT interval: A new clinical syndrome? Cardiology 94: 99–102. Haufe, V., C. Chamberland, and R. Dumaine. 2007. The promiscuous nature of the cardiac sodium current. J Mol Cell Cardiol 42: 469–477. Hindocha, N., L. Nashef, F. Elmslie et al. 2008. Two cases of sudden unexpected death in epilepsy in a GEFS+ family with an SCN1A mutation. Epilepsia 49: 360–365. Hofman, N., A. A. M. Wilde, S. Kääb et al. 2007. Diagnostic criteria for congenital long QT syndrome in the era of molecular genetics: Do we need a scoring system? Eur Heart J 28: 575–580. Johnson, J. N., N. Hofman, C. M. Haglund, G. D. Cascino, A. A. Wilde, and M. J. Ackerman. 2009. Identiἀcation of a possible pathogenic link between congenital long QT syndrome and epilepsy. Neurology 72: 224–231. Kandler, L., A. Fiedler, K. Scheer, F. Wild, U. Frick, and P. Schneider. 2005. Early post-convulsive prolongation of QT time in children. Acta Paediatr 94: 124–137.

Cardiac Channelopathies and Sudden Death

301

Kothari, S. S. 1990. When epilepsy masquerades as heart disease. Postgrad Med 88: 167–171. Kusano, K. F., M. Taniyama, K. Nakamura et al. 2008. Atrial ἀbrillation in patients with Brugada syndrome: Relationships of gene mutation, electrophysiology, and clinical backgrounds. J Am Coll Cardiol 51: 1169–1175. Leenhardt, A., V. Lucet, I. Denjoy, F. Grau, D. D. Ngoc, and P. Coumel. 1995. Catecholaminergic polymorphic ventricular tachycardia in children. A 7 year follow up of 21 patients. Circulation 91: 1512–1519. Lehnart, S. E., M. J. Ackerman, D. W. Benson et al. 2007. Inherited arrhythmias: A National Heart, Lung, and Blood Institute and Office of Rare Diseases workshop consensus report about the diagnosis, phenotyping, molecular mechanisms, and therapeutic approaches for primary cardiomyopathies of gene mutations affecting ion channel function. Circulation 116: 2325–2345. Lehnart, S. E., X. H. Wehrens, P. J. Laitinen et al. 2004. Sudden death in familial polymorphic ventricular tachycardia associated with calcium release channel (ryanodine receptor leak). Circulation 109: 3208–3214. London, B., M. Michalec, H. Mehdi et al. 2007. Mutation in glycerol-3-phosphate dehydrogenase 1 like gene (GPD1-L) decreases cardiac Na+ current and causes inherited arrhythmias. Circulation 116: 2260–2268. Lucet, V., J. Fidelle, D. Do N’Goc, M. C. Toumieux, and P. Coupel. 1983. Catecholaminergic polymorphic ventricular tachycardia in children. Differential diagnosis of epilepsy. Presse Med 12: 102. Lukas, A., and C. Antzelevitch. 1996. Phase 2 reentry as a mechanism of initiation of circus movement reentry in canine epicardium exposed to simulated ischemia: The antiarrhythmic effects of 4-aminopyridine. Cardiovasc Res 32: 593–603. Marbán, E. 2002. Cardiac channelopathies. Nature 415: 213–218. McKean, A., C. Vaughan, and N. Delanty. 2006. Seizure versus syncope. Lancet Neurol 5: 171–180. Mizumaki, K., A. Fujiki, T. Tsuneda et al. 2004. Vagal activity modulates spontaneous augmentation of ST elevation in the daily life of patients with Brugada syndrome. J Cardiovasc Electrophysiol 15: 667–673. Mohamed, U., C. Napolitano, and S. G. Priori. 2007. Molecular and electrophysiological bases of catecholaminergic polymorphic ventricular tachycardia. J Cardiovasc Electrophysiol 18: 791–797. Moss, A. J., W. Zareba, J. Benhorin et al. 1995. ECG T-wave patterns in genetically distinct forms of the hereditary long QT syndrome. Circulation 92 (10): 2929–2934. Myerburg, R. J. 2001. Sudden cardiac death: Exploring the limits of our knowledge. J Cardiovasc Electrophysiol 12: 369–381. Myerburg, R. J., and A. Castellanos. 2008. Cardiac arrest and sudden cardiac death. In Braunwald’s Heart Disease: A Textbook of Cardiovascular Medicine, 8th ed., ed. P. Libby, R. O. Bonow, D. L. Mann, and D. P. Zipes, 933–974. Philadelphia, PA: Saunders Elsevier. Nademanee, K., G. Veerakul, S. Nimmannit et al. 1997. Arrhythmogenic marker for the sudden unexplained death syndrome in Thai men. Circulation 96: 2595–2600. Napolitano, C., S. G. Priori, P. J. Schwartz et al. 2003. Value and accuracy of clinical diagnosis criteria for the long QT syndrome in the era of molecular diagnosis. Circulation 108: IV–363. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41: 542–548. Neyroud, N., F. Tesson, and I. Denjoy. 1997. A novel mutation in the potassium channel KVLQT1 causes the Jervell and Lange-Nielsen cardioauditory syndrome. Nat Genet 15: 186–189. Opherk, C., J. Coromilas, and L. J. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52: 117–127. Paul, M., J. Gerss, E. Schulze-Bahr et al. 2007. Role of programmed ventricular stimulation in patients with Brugada syndrome: A meta-analysis of worldwide published data. Eur Heart J 28: 2126–2133. Priori, S. G., E. Aliot, C. Blomstrom-Lundqvist et al. 2003. Update of the guidelines on sudden cardiac death of the European Society of Cardiology. Eur Heart J 24: 13–15.

302 Sudden Death in Epilepsy: Forensic and Clinical Issues Priori, S. G., C. Napolitano, and P. J. Schwartz. 2008. Genetics of cardiac arrhythmias. In Braunwald’s Heart Disease: A Textbook of Cardiovascular Medicine, 8th ed., ed. P. Libby, R. O. Bonow, D. L. Mann, and D. P. Zipes, 101–109. Philadelphia, PA: Saunders Elsevier. Priori, S. G., C. Napolitano, N. Tiso et al. 2001. Mutations in the cardiac ryanodine receptor gene (hRyR2) underlie catecholaminergic polymorphic ventricular tachycardia. Circulation 103: 196–200. Priori, S. G., C. Napolitano, M. Gasparini et al. 2002. Natural history of Brugada syndrome. Insights for risk stratiἀcation and management. Circulation 105: 1342–1347. Priori, S. G., P. J. Schwartz, C. Napolitano et al. 2003. Risk stratiἀcation in the long-QT syndrome. N Engl J Med 348: 1866–1874. Remme, C. A., A. A. Wilde, and C. R. Bezzina. 2008. Cardiac sodium channel overlap syndromes: Different faces of SCN5A mutations. Trends Cardiovasc Med 18: 78–87. Rhodes, T. E., L. Crotti, M. Arnestad et al. 2006. Gain of function KCNQ1 mutation associated with sudden infant death syndrome. Heart Rhythm 3(abstr suppl): S2. Roden, D. M., J. R. Balser, A. L. George Jr. et al. 2002. Cardiac ion channels. Annu Rev Physiol 64: 431–475. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364: 2212–2219. Sacher, F., V. Probst, Y. Iesaka et al. 2006. Outcome after implantation of a cardioverter–deἀbrillator in patients with Brugada syndrome: A multicenter study. Circulation 114: 2317–2324. Schulze-Bahr, E., L. Eckardt, G. Breithardt et al. 2003. Sodium channel gene (SCN5A) mutations in 44 index patients with Brugada syndrome: Different incidences in familial and sporadic disease. Hum Mutat 21: 651–652. Schulze-Bahr, E., Q. Wang, H. Wedekind et al. 1997. KCNE1 mutations cause Jervell and Lange– Nielsen syndrome. Nat Genet 17: 267–268. Schwartz, P. J., and L. Crotti. 2007. Can a message from the dead save lives? J Am Coll Cardiol 49: 247–249. Schwartz, P. J., A. J. Moss, G. M. Vincent, and R. S. Crampton. 1993. Diagnostic criteria for the long QT syndrome: An update. Circulation 88: 782–784. Schwartz, P. J., S. G. Priori, C. Spazzolini et al. 2001. Genotype–phenotype correlation in the long-QT syndrome: Gene-speciἀc triggers for life-threatening arrhythmias. Circulation 103: 89–95. Skinner, J. R., S. K. C. Chung, C. A. Nel et al. 2007. Brugada syndrome masquerading as febrile seizures. Pediatrics 119: e1206–e1211. Smits, J. P. P., and A. A. M. Wilde. 2005. Genotype-phenotype relationship in the Brugada syndrome. In The Brugada Syndrome: From Bench to Bedside, ed. C. Antzelevitch, P. Brugada, J. Brugada, and R. Brugada, 140–148. Malden: Blackwell Publishing. Splawski, I., J. Shen, K. W. Timothy et al. 2000. Spectrum of mutations in long-QT syndrome genes. KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 102: 1178–1185. Tester, D. J., M. L. Will, C. M. Haglund, and M. J. Ackerman. 2005. Compendium of cardiac channel mutations in 541 consecutive unrelated patients referred for long QT syndrome genetic testing. Heart Rhythm 2: 507–517. Vatta, M., R. Dumaine, G. Varghese et al. 2002. Genetic and biophysical basis of sudden unexplained nocturnal death syndrome (SUNDS), a disease allelic to Brugada syndrome. Hum Mol Genet 11: 337–345. Wilde, A. A. M., C. Antzelevitch, M. Borggrefe et al. 2002. Proposed diagnostic criteria for the Brugada syndrome: Consensus report. Circulation 106(19): 2514–2519. Wilde, A. A. M. 2008. Channelopathies in children and adults. PACE 31: S41–S45. Zareba, W., A. J. Moss, J. P. Daubert, W. J. Hall, J. L. Robinson, and M. Andrews. 2003. Implantable cardioverter deἀbrillator in high-risk long QT syndrome patients. J Cardiovasc Electrophysiol 14: 337–341. Zipes, D. P., A. J. Camm, M. Borggrefe et al. 2006. ACC/AHA/ESC 2006 Guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death. Circulation 114: e385–e484.

Sodium Channel Dysfunction Common Physiopathologic Mechanism Associated with Sudden Death ECG Abnormalities in Brugada Syndrome and Some Types of Epilepsy: Case Histories

20

Claire M. Lathers Paul L. Schraeder Michael W. Bungo

Contents 20.1 Introduction 20.2 Case Histories 20.3 Discussion References

303 304 306 308

20.1â•…Introduction When evaluating a patient presenting with Brugada electrocardiogram and seizures, the physician must differentiate between two possible causes: solely cardiac arrhythmias and/ or also a problem of idiopathic epilepsy. Some patients with Brugada syndrome with associated ECG abnormalities and other patients with some types of epilepsy may exhibit a common pathophysiologic mechanism associated with sodium channel dysfunction. Sodium channel dysfunction may be one of several risks and mechanisms for sudden death in cardiac and epileptic patients. Risks and mechanisms for sudden death in both cardiac and epileptic patients was recently reviewed by Lathers et al. in a comprehensive article entitled “The Mystery of Sudden Death: Mechanisms for Risks” (Lathers et al. 2008a). In a subsequent discussion Lathers et al. (2008b) raised unanswered questions and addressed needed clinical studies to evaluate the value of lifestyle management (Scorza et al. 2008). We have written the following discussion of two case histories to expand and highlight risks and mechanisms for sudden death in both cardiac disease and in persons with epilepsy. Facts described by Herreros (2010, Chapter 19), Lathers et al. (2010, Chapter 1), and Walczak (2010, Chapter 12) address risks for sudden unexpected death in epilepsy (SUDEP) and are used here to allow the readers to glean greater insight into the problem of sudden death and the potential role of sodium channel dysfunction. 303

304 Sudden Death in Epilepsy: Forensic and Clinical Issues

20.2â•…Case Histories Case 1â•… Epilepsy and Brug ada Sy ndro me in a Man wit h€a€Hist o ry o f Generaliz ed T o nic Clo nic Seiz ures A 24-year-old man had an EEG-conἀrmed diagnosis of generalized tonic–clonic epilepsy since age 7, at times associated with tongue biting and urinary incontinence. He was treated with phenobarbitol and diazepam. Since age 17 he also had some seizures that seemed to be related to alcohol abuse. At age 24 he was observed by his wife to have a syncopal episode without generalized clinical seizure activity. An ECG found right bundle branch block and ST segment elevation in V2–V3, ἀndings consistent with Brugada syndrome. All other cardiac studies, Doppler echocardiography, radionuclide tomography, coronary angiography, contrast ventriculography, cardiac MRI, 24-hour ambulatory ECG, and electrophysiological studies were normal. After implantation of a cardioverter–deἀbrillator, no syncope, no seizure, nor any appropriate therapy by the deἀbrillator occurred. Discussio n by F auchier et al. (2000) Fauchier et al. (2000) make the point that epilepsy may occur concurrently with a primary electrophysiological abnormality of the heart in the form of Brugada syndrome (i.e., defects of the cardiac sodium channel gene SCN5, associated with a high risk of cardiac sudden death from ventricular arrhythmias). They note that, as in SUDEP, there is young male predominance. In persons with a diagnosis of Brugada syndrome an implantable deἀbrillator seems to be the intervention of choice. Cases treated only with amiodarone and/or beta blockers were not protected against sudden death, with a death rate of 26% during follow-up. An emphasis was placed on the fact that persons with cardiac syncope usually do not experience tongue biting unless a hypoperfusion related secondary seizure occurs. Finally, they comment that the previous ECGs of SUDEP victims should be reviewed, when available, in order to help in determining whether Brugada syndrome was a factor in contributing to the death. Discussio n by L at hers, Schraeder, and Bung o This is one of the ἀrst case reports to document the concurrence of epilepsy and Brugada syndrome. However, the authors did not discuss the physiological connection in the form of sodium channel dysfunction that is extant in both entities. One of the most important clinical points that Fauchier et al. (2000) made was the need to review the previous ECGs of victims of SUDEP to look for evidence of subtle cardiac abnormalities that either could have been the primary disorder that was mimicking seizures, or that were concurrent with a bona ἀde seizure disorder. However, the next obvious step, namely, considering doing screening ECGs in all persons with a diagnosis of seizure disorder should be addressed. The presence of a primary diagnosis of Brugada syndrome as the explanation for seizures that are related to syncope has obvious implications for intervention in the form of implantation of a deἀbrillator, while the ἀnding of coexisting Brugada syndrome and epilepsy would suggest that the patient should also be continued on antiepileptic medication. The authors did

Sodium Channel Dysfunction

305

not clarify whether the patient in their case report was continued on antiepileptic medication. Constructed from Fauchier, L., D. Babuty, and P. Cosnay. 2000. Epilepsy, Brugada syndrome and the risk of SUDEP. J Neurol 247: 643–644.

Case 2â•… A Wo man wit h Brug ada Sy ndro me and Epilepsy A 40-year-old woman had recurrent episodes of dizziness followed by remitting episodes of hearing loss. It was suspected that the episodes of dizziness might be seizures and that the transient hearing loss might be a postictal phenomenon. An ECG on admission found ST elevation in V1 through V3 consistent with a diagnosis of Brugada syndrome. She was placed on long-term EEG and ECG monitoring. Several episodes of jerking movements and confusion followed by hearing loss were observed on video. As there were no apparent EEG or telemetry abnormalities, a diagnosis of pseudoseizures was entertained. It was determined that the patient should be weaned off of her antiepileptic medications of phenytoin, lamotrigine, levetiracetam, and topiramate. During the process of being weaned off of lamotrigine, she experienced a tonic–clonic seizure accompanied by generalized polyspike and wave activity. The EEG discharges persisted for 24 hours but resolved after the lamotrigine dosage was increased. No cardiac arrhythmia was observed on telemetry, and, in fact, the ECG changes consistent with Brugada syndrome improved but did not resolve as the phenÂ� ytoin dose was reduced. The patient suffered no further attacks and she was maintained on lamotrigine and low-dose phenytoin. At no time during her hospital stay were any arrythmias observed. Discussio n by G irard et al. (2008) Girard et al. (2008) hypothesized that their patient may have a presumably genetic underlying systemic sodium channel dysfunction. They acknowledge that both Brugada syndrome and epilepsy are associated with the risk of sudden death. They also note that even though seizures may occur in persons with Brugada syndrome, they are secondary to arrhythmia-related cerebral hypoperfusion. Discussio n by L at hers, Schraeder, and Bung o This case is another example of speculation that there may be a group of patients with systemic sodium channelopathies that have a potential for having both Brugada syndrome and epilepsy. However, as in this case, the Brugada component appeared to be clinically silent while the epilepsy was clinically manifest. An important question to consider is whether having the ECG markers of Brugada syndrome, without evidence of related ventricular arrhythmias, increases the risk of a person with epilepsy having a seizure-related fatal cardiac event. As discussed in the case described by Fauchier et al. (2000), the question can be raised about the possible need for implantation of a deἀbrillator in a patient with a heretofore clinically silent set of ECG ἀndings

306 Sudden Death in Epilepsy: Forensic and Clinical Issues

consistent with Brugada syndrome. Another question raised by the observation that the ECG changes in the V1 through V3 leads improved with a decrease in the dose of phenytoin is that of whether certain antiepileptic drugs place patients at greater risk for cardiogenic ventricular arrhythmias. If this turns out to be the case, parameters for what drugs should be avoided must be developed. Constructed from Girard, S., A. Cadena, and Y. Acharya. 2008. Woman with Brugada syndrome and epilepsy: A unifying diagnosis? South Med J 101: 1150–1153.

20.3â•…Discussion The two cases above illustrate that some patients will exhibit both Brugada syndrome and some types of epilepsy produced by sodium channel dysfunction. The risk factors for sudden death must be deἀned and linked with mechanisms for death (Lathers et al. 2008a, 2008b; Herreros, 2010, Chapter 19). Brugada syndrome is produced by a mutation in gene SCN5A that encodes the alpha subunit of cardiac sodium channel (Antzelevitch et al. 2005). Herreros et al. (2010, Chapter 19) note that some epileptic syndromes (Graves 2006) are due to different mutations in genes encoding alpha subunits of neuronal sodium channels (SCN1A, SCN2A) or in the beta subunit (SCN1B), common for both cardiac and neuronal isoforms. Additional clinical, electrocardiographic, and genetic studies are needed to improve risk stratiἀcation for a given patient and to determine if there is a relationship among sodium channel dysfunction, Brugada ECG, and idiopathic epilepsy. The response of the two patients presented in the cases above appears to support the possibility that a common pathophysiologic mechanism is associated with sodium channel dysfunction and may be common to ECG abnormalities of Brugada syndrome and some types of epilepsy. Many more patients must be screened to conἀrm this. Prospective electrocardiographic and genetic studies of patients must be done to determine how to identify which patients with epilepsy are at risk for SUDEP and which patients with cardiac disease are at risk of sudden cardiac death. The review of Lathers et al. (2008a) addressed the possible overlapping mechanisms that may apply to the risk of SUDEP and in cardiac disease. We explored the interaction between the central and peripheral autonomic nervous systems and the cardiopulmonary systems. A discussion of the potential interactive role of genetically determined subtle cardiac risk factors for arrhythmias with predisposition for seizure related cardiac arrhythmias was included. Possible mechanisms that are operant in producing both epileptogenic and cardiogenic arrhythmias were addressed. Speculation about potential preventive measures to minimize the risk of both SUDEP and sudden cardiac death was provided. We noted that although the mechanisms involved in SUDEP are not known, and many published epidemiological studies and review articles speculate about these mechanisms, relatively few basic research studies have addressed the question of mechanism. While epilepsy-Â�associated cardiac arrhythmia is believed to be a primary factor in the mechanism of SUDEP, respiratory factors and hypoxia, and psychological factors and mechanisms were also addressed by Lathers et al. (2008a). Lathers and Schraeder (1990) discuss in their book, Epilepsy and Sudden Death, that when evaluating persons with epilepsy who may be at risk for sudden death and considering potential causes of the problem, it is always a question of, “Which is the cart and which

Sodium Channel Dysfunction

307

is the horse?” When evaluating a given patient, simultaneous monitoring of the EEG and the ECG over a deἀned period of time is recommended since some patients with epilepsy are misdiagnosed as cardiac patients while some patients diagnosed with cardiac problems actually have epilepsy. The connection between cardiac and central nervous system mechanisms for sudden death, whether in SUDEP or in cardiac sudden death, strongly deἀnes the proposition that a pathological interaction between brain and heart underlies the occurrence of deaths in some, but not all, of the patients in both groups (Lathers et al. 2008a, 2008b). There is crossover between groups as exempliἀed by patients who present with epilepsy that masquerades as cardiac arrhythmias while other patients present with heart disease that masquerades as epilepsy (Schraeder et al. 1983; Kothari 1990; Ficker et al. 1998; Narang et al. 2005; Woodley et al. 1977; Williams and Frenneaux 2007; Leung et al. 2006; P-Codrea Tigeran et al. 2005; Natelson et al. 1998). More recently, Skinner et al.€(2007) suggested that Brugada syndrome may masquerade as febrile seizures. In 2008, Lathers et al. (2008a) stated, “Since to our knowledge no studies are extant in persons with epilepsy that look for genetic mutations that predispose to potentially fatal cardiac arrhythmias, we can only speculate about the possibility that epileptiform discharge related cardiac autonomic neural imbalance would aggravate the predisposition to perturbations of cardiac autonomic balance and cardiac rhythm regulation in patients with epilepsy who have such a genetic mutation.” Despite detailed characterizations of the mutated ion channels at the molecular level, we do not understand which individual mutations may lead to arrhythmias and sudden death. Soliciting a detailed family history of cardiac related deaths in relatives of persons with epilepsy could be an important ἀrst step in identifying cardiac risk factors in persons with epilepsy who might be susceptible to cortical discharge induced destabilization of cardiac neural regulation. Patients with a positive family history would be an important area for future investigation into prevention. Brugada syndrome is another distinct clinical entity that is associated with increased risk for sudden death. It is thought to be an autosomally inherited disorder affecting the cardiac sodium channel that predisposes to ST segment elevation in right precordial leads and malignant ventricular arrhythmias. Meregalli et al. (2005) suggest that Brugada syndrome is not a monofactorial disease, but rather one with various pathophysiological mechanisms in the individual patients at risk. This latter consideration raises again the possibility that ictally related disruption of cardiac neural regulation might be a factor in persons with epilepsy who also have a predisposition for Burgada syndrome. The electrocardiogram pattern of RBBB and the ST segment elevation in leads V1 through V3 (Brugada syndrome) is associated with high risk of sudden death in patients with a “normal” heart (Ackerman et al. 2004). The natural history of the disease is not well established nor is the approach to stratifying these patients according to risk. Furthermore, some patients do not consistently exhibit the characteristic ECG pattern, which may be apparent only intermittently or only after provocation with ajmaline or intravenous flecainide (Ali et al. 2006). Molecular genetics may explain some mechanisms that underlie sudden cardiac death in young persons with structurally normal hearts. Genetic mutations affecting cardiac ion channels may disrupt the balance of currents in the action potential, thus predisposing to malignant ventricular tachycardias (Bezzina et al. 2006; Dumaine et al. 1996). The cardiac sodium channel gene, SCN5A, is involved in two arrhythmogenic diseases—Brugada syndrome and one form of the long QT syndrome (LQT3). At present, it is thought that Brugada syndrome mutations reduce sodium channel activity current while LQT3 mutations are associated with the opposite effect (Ackerman et al. 2004). Both of these conditions result in an electrical gradient between the endocardium

308 Sudden Death in Epilepsy: Forensic and Clinical Issues

and the epicardium setting the substrate for dispersion of repolarization and the genesis of arrhythmia (Shimizu and Antzelevitch 1999). Studies to date illustrate how subtle changes in channel biophysics exert signiἀcant and distinct effects and provide evidence of the importance of our understanding molecular changes in order to better treat cardiac disease. There may be patients with a typical Brugada ECG and nocturnal seizures who, after implantation of a deἀbrillator, are then found to exhibit a nonarrhythmia cause for their seizures. It may be that in some instances both Brugada syndrome and some types of epilepsy are produced by sodium channel dysfunction. Review of known clinical observations suggest a common link: reversible Brugada ECG pattern and drug toxicity induced seizures, shared modulation factors for clinical expression Brugada syndrome and some syndromes associated with epilepsy, and cases of sudden unexpected death among persons with epilepsy. Mechanistic questions must be raised: how much of sudden death in cardiac and epileptic patients is due to dysfunction of the sodium channel? It is relevant to speculate the potential risk of sudden death by mixing the predisposition to disorders of the heart and/or disorders of the brain. How many patient groups are mixed, possessing both epileptic and cardiac dysfunction? Epidemiology studies must be designed to tease out the differences between cardiac versus seizure factors. The role of drugs in cause or prevention of sudden death is an important clinical pharmacology question (Lathers and Schraeder 1995, 2002; Lathers et al. 1993, 2003a, 2003b; Schraeder and Lathers 1995). Lathers et al. (2008a) note that drugs that prolong the QT interval are associated with increased risk of SUD (Reingardiene and Vilcinskaite 2007). Antiepileptic drug polytherapy, frequent dose changes, and high carbamazine levels are risk factors for SUDEP in unstable severe epilepsy (Walczak 2003). The risk of SUDEP can be decreased by aggressive treatment of tonic–clonic seizures with as few antiepileptic drugs as necessary to achieve complete control. The risk of SUDEP rates in patients on lamotrigine, gabapentin, topiramate, tigabine, and zonisamide appears to be similar to those on long-available standard antiepileptic drugs (Lathers and Schraeder 2002). Newer antiepileptic drugs may or may not induce sedation and may minimize noncompliance by reducing side effects of lethargy and cognitive impairment while improving tolerability with comedication (Lathers et al. 2003a). Difficulty in achieving therapeutic dosage with some newer antiepileptic drugs such as Topirimate, because of side effects, raises the question of whether these agents are “better” than the older and most side-effect-prone antiepileptic drug, phenobarbital (Lathers et al. 2003a). Prospective studies are needed to determine whether we can identify persons with epilepsy who are the greatest at-risk candidates for SUDEP and to identify the antiepileptic drug or combination of antiepileptic drugs to provide the best protection for a given patient, without, on the other hand, increasing the possibility of antiepileptic drug-related adverse cardiac changes.

References Ackerman, M. J., I. Splawski, J. C. Makielski et al. 2004. Spectrum and prevalence of cardiac sodium channel variants among black, white, Asians, and Hispanic individuals: Implications for arrhythmogenic susceptibility and Brugada/long QT syndrome genetic testing. Heart Rhythm 1: 608–609. Ali, I. A., C. A. Iliescu, and R. W. Smalling. 2006. Brugada syndrome in a black man with seizures and incontinence. Texas Heart Inst J 33: 273–274.

Sodium Channel Dysfunction

309

Antzelevitch, C., P. Brugada, M. Borggrefe et al. 2005. Brugada syndrome. Report of the Second Consensus Conference. Circulation 111: 659–670. Bezzina, C. R., W. Shimizu, P. Yang et al. 2006. Common sodium channel promoter haplotype Asian subjects underlies variability in cardiac conduction. Circulation 24 (113): 338–344. Dumaine, R., Q. Wang, M. T. Keating et al. 1996. Multiple mechanisms of Na+ channel-linked long-QT syndrome. Circ Res 78: 916–924. Fauchier, L., D. Babuty, and P. Cosnay. 2000. Epilepsy, Brugada syndrome and the risk of SUDEP. J Neurol 247: 643–644. Ficker, D. M., G. D. Cascino, and I. P. Clements. 1998. Cardiac asystole masquerading as temporal lobe epilepsy. Mayo Clin Proc 73: 784–786. Girard, S., A. Cadena, and Y. Acharya. 2008. Woman with Brugada syndrome and epilepsy: A unifying diagnosis? South Med J 101: 1150–1153. Graves, T. D. 2006. Ion channels in epilepsy. Q J Med 99: 201–217. Herreros, B. 2010. Cardiac channelopathies and sudden death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 19. Boca Raton, FL: CRC Press. Kothari, S. S. 1990. When epilepsy masquerades as heart disease. Postgrad Med 88: 167–170. Lathers, C. M., and P. L. Schraeder. 1990. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 1995. Experience-based teaching of therapeutics and clinical pharmacol antiepileptic drugs. J Clin Pharmacol 35: 573–587. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42: 123–136. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital upon autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33: 837–844. Lathers, C. M., S. A. Koehler, C. W. Wecht, and P. L. Schraeder. 2003a. Antiepileptic drug levels in autopsy cases of sudden, unexpected deaths in persons with epilepsy in Allegheny County Pennsylvania in 2001. J Clin Pharmacol 43. Lathers, C. M., P. L. Schraeder, and H. G. Claycamp. 2003b. Clinical pharmacology of topiramate vs lamotrigine vs phenobarbital: Comparison of efficacy and side effects using odds ratios. J Clin Pharmacol 43: 491–503. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008a. The mystery of sudden death: mechanisms for risks. Epilepsy Behav 12: 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Reply. Unanswered questions: Studies needed. Epilepsy Behav 13: 265–269. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Neurocardiologic mechanistic risk factors in sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 1. Boca Raton, FL: CRC Press. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9: 19–30. Meregalli, P. G., A. A. Wilde, and H. L. Tan. 2005. Pathophysiolgical mechanisms of Brugada syndrome: Depolarization disorder, repolarization disorder, or more? Cardiovasc Res 67: 367–378. Narang, M., T. Dua, S. Gomber, and V. Mahajan. 2005. Cardiac syncope from occult transposition masquerading as convulsive epilepsy. Eur J Paediatr Neurol 9: 361–362. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55: 857–860. P-Codrea Tigeran, S., S. Dalager-Pedersen, U. Baandrup, M. Dam M, and A. Vesterby-Charles. 2005. Sudden unexpected death in epilepsy: Is death by seizures a cardiac disease? Am J Forensic Med Pathol 26: 99–105. Reingardiene, D., and J. Vilcinskaite. 2007. QT-c prolonging drugs and the risk of sudden death. Medicina (Kaunas) 43: 347–453.

310 Sudden Death in Epilepsy: Forensic and Clinical Issues Schraeder, P. L., and C. M. Lathers. 1995. Clinical pharmacology of antiepileptic drug use: ‘Clinical pearls about the perils of patty.’ A Clinical Pharmacology Problem Solving (CPPS) Unit. J Clin Pharmacol 35: 1120–1135. Schraeder, P. L., R. Pontzer, and T. R. Engle. 1983. A case of being scared to death. Arch Intern Med 143: 1793–1794. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13: 263–264. Shimizu, W., and C. Antzelevitch. 1999. Cellular basis for long QT, transmural dispersion of repolarization, and torsade de pointes in the long QT syndrome. J Electrocardiol 32 (Suppl): 177–184. Skinner, J. R., S. K. C. Chung, C. A. Nel et al. 2007. Brugada syndrome masquerading as febrile seizures. Pediatrics 119: 1206–1211. Walczak, T. 2003. Do antiepileptic drugs play a role in sudden unexpected death in epilepsy? Drug Saf 26: 673–683. Walczak, T. 2010. Risk factors for sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chap. 12. Boca Raton, FL: CRC Press. Williams, L., and M. Frenneaux. 2007. Syncope in hypertrophic cardiomyopathy mechanism and consequences for treatment. Europace 9 (9): 817–822. Woodley, D., W. Chambers, H. Starke, B. Dzindzio, and A. D. Forker. 1977. Intermittent complete atrioventricular block masquerading as epilepsy in the mitral valve prolapse syndrome. Chest 72: 369–372.

Not Seizure but Syncope Saumya Sharma Trieu Ho Bharat K. Kantharia

21

Contents 21.1 Pathophysiology of Syncope 21.2 Classiἀcation of Syncope 21.2.1 Neurally Mediated Reflex Syncope 21.2.2 Cardiac Syncope 21.2.3 Neurologic Syncope 21.2.4 Metabolic Syncope 21.2.5 Psychogenic Syncope 21.2.6 Orthostatic Syncope 21.3 Investigations 21.3.1 History and Physical Examination 21.3.2 Electrocardiogram 21.3.3 Echocardiography 21.3.4 Prolonged EGG Monitoring 21.3.5 Tilt Table Testing 21.3.6 Electrophysiologic Testing 21.3.7 Miscellaneous Tests 21.4 Syncope and Epilepsy 21.5 Conclusion References

312 312 312 314 316 316 316 316 316 317 317 318 318 319 320 320 320 321 321

Syncope is deἀned as an acute transient loss of consciousness and postural tone with spontaneous and complete recovery (Demaksian and Lamb 1958; Kapoor 2000; Soteriades et al. 2002; Nair et al. 2003; Grubb 2005). The Framingham Heart Study, which examined 7814 individuals, reported the incidence of a ἀrst report of syncope of 6.2 per 100 personyears follow-up (Soteriades et al. 2002). Syncope is also a costly problem with an estimated annual cost of treatment of $800 million (Nyman et al. 1999), and has a signiἀcant impact on the morbidity and mortality; cardiac syncope in particular is associated with higher mortality rates than the general population (Soteriades et al. 2002). Patients with recurrent syncope have a poor quality of life (Rose et al. 2000). Since there are numerous conditions that may lead to syncope, many patients rather than going through a focused set of tests often undergo extensive and expensive investigations including brain scans, carotid Doppler studies, and electroencephalography routinely. The purpose of this chapter is to discuss different causes of syncope and methods to establish the diagnosis.

311

312 Sudden Death in Epilepsy: Forensic and Clinical Issues

21.1â•… Pathophysiology of Syncope The metabolism of the brain is dependent on cerebral perfusion. A sudden interruption of cerebral blood flow for 6 to 8 s is sufficient to cause syncope (Kapoor 2000). During tilt table testing, a decrease in systolic blood pressure (BP) to 60 mm Hg has been shown to be associated with loss of consciousness (Kapoor 2000). Hypoperfusion of the ascending reticular activating system in the brainstem, cerebral cortex, or both is the ultimate underlying mechanism for loss of consciousness (Kapoor 2000). There are several physiologic regulatory mechanisms that maintain cerebral blood flow (Nirkko and Baumgartner 2006): (1) autonomic regulation of heart rate, cardiac contractility, and systemic vascular resistance; (2) cerebrovascular autoregulation mediated by local metabolic factors (e.g., blood pressure, pO2, pCO2); and (3) vascular volume regulated by renal function and circulating hormones (e.g., aldosterone and vasopressin). Failure of these regulatory mechanisms to maintain the lower limits of cerebral perfusion is what leads to syncope.

21.2â•… Classification of Syncope In one-third of cases the cause of syncope can be determined with a good history and physical examination and a 12-lead electrocardiogram (Schnipper and Kapoor 2001; Ammirati et al. 2000; Alboni et al. 2001; Sarasin et al. 2001). In the remaining two-thirds, some pointed investigations should be considered. However, even after detailed investigations, the cause of syncope cannot be determined in one-third of patients (Schnipper and Kapoor 2001; Ammirati et al. 2000; Alboni et al. 2001; Sarasin et al. 2001). The causes of syncope are listed in Table 21.1. In patients with recurrent syncope, neurally mediated reflex syncope and orthostatic syncope are most common followed closely by cardiac syncope (Ammirati et al. 2000; Alboni et al. 2001; Sarasin et al. 2001). However, it is important to understand that the distribution in the cause of syncope varies with age. In young patients, neurally mediated reflex syncope is most common. In older patients, orthostatic hypotension, medication-related syncope, and cardiac causes of syncope are more common. There are features in the history and physical examination that help in determining the etiology of syncope. First, syncope must be differentiated from “nonsyncopal conditions.” In elderly patients, falls can occur with amnesia of the event. Obtaining information from a witness to the event, when possible, often helps in the diagnosis. Identifying history of cardiac disease, family history of cardiac disease, and medications may also be useful. There are also distinct clinical features that may be useful in the differentiation of the causes of syncope. Identiἀcation of triggers to the event, prodromal symptoms, history of convulsive jerking and/or incontinence during the episode, and the speed of recovery from the event are extremely important in diagnosing the cause of syncope. 21.2.1â•… Neurally Mediated Reflex Syncope There are many reflex-mediated syncope syndromes, all of which have a common response limb of reflex vasodilation and bradycardia due to enhanced vagal stimulation and withdrawal of sympathetic tone. What differs among these syndromes is the trigger associated with this reflex response and loss of consciousness (i.e., cough, micturation, or pain).

Not Seizure but Syncope

313

Table 21.1â•… Causes of Syncope Neurally mediated reflex syncope Vasovagal syncope Trigger associated syncope (e.g., micturation) Carotid sinus hypersensitivity syndrome Cardiac syncope Disorders of cardiac rhythm — op18 Tachyarrhymias Ventricular arrhythmia Supraventricular arrhythmia — op18 Bradyarrhythmias Sinus node dysfunction AV node/His–Purkinje system dysfunction Disorders of cardiac structure — op18 Aortic stenosis — Hypertrophic cardiomyopathy — Mitral stenosis — Atrial myxoma — Pulmonary stenosis — Tetrology of Fallot — Intracardiac shunts with Eisenneger’s syndrome — Right ventricular strain secondary to primary or secondary pulmonary hypertension Neurologic syncope Transient ischemic attacks (posterior circulation) Ictal bradycardia Intermittent obstructive hydrocephalus Migraine Orthostatic syncope Autonomic neuropathy (e.g., diabetes mellitus, amyloidosis) Drugs (e.g., anithypertensives) Metabolic syncope Hypoglycemia or hypocalcemia Psychogenic syncope Panic disorder

Vasovagal syncope, the most common cause of syncope, is a form of neurally mediated reflex syncope in which a speciἀc trigger cannot be reliably identiἀed. There are certain clinical features that distinguish neurally mediated reflex syncope from seizures and cardiac syncope. As mentioned earlier, there are often certain situations or speciἀc triggers that elicit neurally mediated reflex syncope. Vasovagal syncope can be in response to prolonged standing, hot crowded environments, emotional trauma, or pain. In certain susceptible individuals, swallowing, micturation, or coughing can provoke syncope. Neurally mediated reflex syncope almost never occurs when the patient is supine. Very frequently, there is a prodromal syndrome (presyncope) that warns the patient of impending syncope. Patients can describe nausea, lightheadedness, flushing, tingling, sweating, blackening of vision, and muffled hearing. There may be a dissociative state prior to frank syncope, during which patients are aware and can hear ongoing events but their vision is dark and they cannot respond. This prodrome can last between 1 and 5 min. Because of this prodromal syndrome, patients frequently sit or lay down and thus serious injury to the head and face is often avoided. Witnesses to the syncopal event may describe

314 Sudden Death in Epilepsy: Forensic and Clinical Issues

the patient as pale, sweaty, and cold. In rare cases, neurally mediated syncope is associated with brief convulsive jerking or urinary incontinence. However, tonic–clonic activity and tongue biting is very uncommon. Recovery from the event occurs once the patient is recumbent and takes several seconds to minutes. Unlike cardiac syncope where recovery is almost immediate, patients with neurally mediated reflex syncope may be confused for several seconds after they regain consciousness and may feel fatigued for several minutes afterward (Kapoor 2000; Schnipper and Kapoor 2001; Brignole et al. 2001). In contrast to seizures where there is prolonged postictal confusion, patients of neurally mediated reflex syncope become oriented within seconds of regaining consciousness. Carotid sinus hypersensitivity is a subtype of neurally mediated syncope and rarely manifests in young individuals. However, elderly patients frequently present with dizziness, syncope, or falls precipitated by head movements or with tight constrictive neckwear. Gentle pressure applied to the carotid pulse just below the angle of the jaw for 5 to 10 s stimulates carotid sinus baroreceptors located in the internal carotid artery above the bifurcation of the common carotid artery, resulting in enhanced vagal activity and withdrawal of sympathetic tone. A sinus pause for greater than 3 s and a drop in systolic blood pressure of greater than 50 mm Hg associated with symptoms is diagnostic. Carotid sinus massage should be performed in the supine and upright position as well (Morillo et al. 1999). Although persistent neurologic complications of carotid sinus massage are rare, carotid sinus massage should be avoided in patients with transient ischemic attacks, strokes, or carotid bruits. Unfortunately, carotid sinus hypersensitivity is a common ἀnding in asymptomatic elderly individuals. Other causes of syncope should be excluded before an attempt to diagnose carotid sinus hypersensitivity with carotid sinus stimulation. 21.2.2â•… Cardiac Syncope Cardiac causes of syncope account for 10% to 20% of syncopal episodes, making it the second most common cause of syncope. Cardiac syncope can be divided into disorders of cardiac rhythm or cardiac structure (Table 21.1). Cardiac arrhythmias are, by far, more common causes of syncope than cardiac structural disorders. Cardiac rhythm disorders can be further subdivided into tachyarrhythmias and bradyarrhythmias. Tachyarrhythmias can be ventricular or supraventricular in origin. In patients with prior myocardial infarction, syncope is commonly due to scar-related ventricular tachycardia. These patients are usually middle-aged or older, may have depressed ventricular function and multiple comorbid conditions (i.e., diabetes, peripheral vascular disease, renal disease). They are at very high risk for sudden cardiac death. In patients with structurally normal hearts, genetic disorders of cardiac ion channels, such as long QT syndrome or Brugada syndrome, are associated with syncope due to ventricular arrhythmias. Interestingly, 10% of patients with long QT syndrome present with episodes that are mistaken for generalized seizure (Ackerman 1998). Patients with genetic channelopathy disorders may have characteristic electrocardiographic ἀndings. For example, in case of Brugada syndrome, the electrocardiogram may show atypical right bundle branch block pattern with ST segment elevation in V1 to V3 leads at baseline or when provoked after procainamide infusion (Figure 21.1). Electrolyte disorders (e.g., hypokalemia, hypomagnesemia) and drugs (e.g., quinidine) may also cause ventricular arrhythmias presenting as syncope. Patients with supraventricular tachycardias more often present with palpitations than syncope. Bradyarrhythmias can be divided

Not Seizure but Syncope

315

II

aVL

V2

V5

III

aVF

V3

V6

II

Figure 21.1╇ Electrocardiogram in a patient with Brugada syndrome. Atypical right bundle branch pattern with characteristic ST segment elevation (following intravenous infusion of procainamide) in the precordial leads V1 to V3 is shown.

into disorders of cardiac impulse generation or its conduction. Both are usually due to degenerative changes in the conduction system of the heart that occur with aging. Sinus node dysfunction often presents with alternating supraventricular tachycardia (usually atrial ἀbrillation) and bradycardia (e.g., tachy-brady syndrome). In these patients, spontaneous termination of the supraventricular arrhythmia results in long sinus pauses that cause syncope. Disorders of impulse conduction occur within the atrioventricular node or the His–Purkinje system. Disorders of the His–Purkinje system have a more malignant course with higher progression to complete heart block and unreliable subsidiary escape rhythm. Electrocardiographic manifestations of bradyarrhythmias are common and include sinoatrial exit block, bifascicular block, and alternating left and right bundle branch block. Disorders of cardiac structure such as severe aortic stenosis, hypertrophic cardiomyopathy, left atrial myxoma, and mitral stenosis may also present with syncope. Cardiac syncope is usually described as a sudden loss of consciousness without any discernible trigger and which can occur in any posture. Prodromal symptoms are rare. Although brief episodes of palpitations or chest pain can be described, they are uncommon. Although brief convulsive jerking is common, bladder/bowel incontinence or lateral tongue biting is very rare. Recovery is very quick form the episode with no fatigue or confusion after the event. Because there is no prodromal syndrome, patients frequently injure themselves during the syncopal event. Witnesses to the event may describe the patient as falling suddenly and being cold and pale. Exertional syncope should prompt an evaluation for a cardiac structural problem like aortic stenosis or hypertrophic cardiomyopathy. Patients with a history of myocardial infarction, congestive heart failure, or family history of sudden cardiac death require urgent assessment by a cardiologist.

316 Sudden Death in Epilepsy: Forensic and Clinical Issues

21.2.3â•… Neurologic Syncope Neurologic causes of syncope are uncommon, accounting for less than 10% of all cases of syncope. The majority of patients with neurologic syncope are found to have seizure rather than true syncope. Seizure-induced arrhythmias are rare causes of true syncope. The most common arrhythmia during seizures is sinus tachycardia, which usually has little clinical signiἀcance. Arnold-Chiari malformations or colloid cysts in the third ventricle can result in intermittent obstructive hydrocephalus that causes headaches prior to loss of consciousness. Transient ischemic attacks involving the posterior circulation may present with vertigo, ataxia, diplopia, and rarely, syncope. Migraines can rarely cause syncope. Typical migraine symptoms are associated with gradual loss of consciousness. Basilar artery migraine presents with diplopia, visual blackening, vertigo, and prolonged syncope. Bradyarrhythmias seen in the ictal state are rare and are usually associated with partial seizures. It is postulated that ictal bradycardia may be a mechanism for sudden unexplained death in epilepsy patients (SUDEP; Britton 2004). 21.2.4â•… Metabolic Syncope Loss of consciousness may occur from metabolic disturbance with hypoglycemia being the most common metabolic cause of syncope. It is usually associated with altered consciousness prior to syncope, which may progress to convulsions in a patient with insulin-requiring diabetes mellitus. Syncopal symptoms are promptly reversed with administration of glucose. Hypocalcemia may present with episodes of parasthesias, carpopedal spasm, and syncope. 21.2.5â•… Psychogenic Syncope Psychogenic causes of syncope are being recognized with greater frequency (Ammirati et€al. 2000; Alboni et al. 2001; Sarasin et al. 2001). Panic disorders can result in hyperventilationinduced hypocapnia and cerebral vasoconstriction. This cause of syncope is associated with facial and limb parasthesias, carpopedal spasm, palpitations, and anxiety. 21.2.6â•… Orthostatic Syncope The mechanism of orthostatic syncope is autonomic dysfunction resulting in impaired vasoconstriction in response to postural changes. A drop in systolic blood pressure by greater than 20 mm Hg with seconds or minutes of standing is diagnostic. Orthostatic syncope is a disorder most common in the elderly, in patients with diabetics with autonomic neuropathy as occurs due to diabetes mellitus, alcohol, or amyloidosis. Certain pharmacological agents (e.g., antihypertensive drugs) may also cause orthostatic syncope. Orthostatic syncope occurs within seconds or minutes of being upright and can be diagnosed easily by orthostatic blood pressure measurements.

21.3â•…Investigations Identiἀcation of the cause of syncope can be challenging and costly because syncope is most often sporadic and infrequent. A presumptive diagnosis can be made based on

Not Seizure but Syncope

317

history, physical examination, and electrocardiogram. Further diagnostic tests only occasionally assist in determining the etiology. Nonetheless, selected diagnostic tests should be ordered to conἀrm the diagnosis. 21.3.1â•… History and Physical Examination The history, including a witness account, is the most important component of the diagnostic evaluation (Ammirati et al. 2000; Alboni et al. 2001; Sarasin et al. 2001). Several studies have demonstrated that the probable cause of syncope can be determined by history and physical alone (Ammirati et al. 2000; Alboni et al. 2001; Sarasin et al. 2001). Initially, the clinician must conἀrm that a true syncopal event has occurred, excluding “nonsyncopal” events (e.g., falls, alcohol intoxication, hyperventilation, and other psychiatric disorders). The history should then focus on triggers of the episode, prodromal symptoms, characteristics of the event itself (from a witness), and recovery from loss of consciousness. Obtaining a history of structural heart disease, a detailed medication history, and family history of sudden cardiac death is critical for diagnosing potentially life-threatening conditions. Arrhythmic causes of syncope are suggested by less than 5 s of prodromal symptoms, sudden recovery, male gender, less than three syncopal episodes, and increased age. Neurally mediated reflex syncope is suggested by blurred vision, nausea, warmth, diaphoresis, and fatigue after syncope. The clinical history is also valuable in distinguishing seizure from syncope. There are features such as prolonged lack of orientation after the event, lack of pallor during the episode, frothing at the mouth, duration of unconsciousness of greater than a few minutes, aching muscles, and lateral tongue biting strongly indicate seizure as the diagnosis. Vertebral basilar insufficiency should be suggested if syncope is associated with diplopia, tinnitus, vertigo, dysarthria, or focal weakness. Migraine-mediated syncope is associated with occurrence of headache, along with other symptoms such as scintillating scotoma and nausea. Orthostatic vital signs should be performed on all patients with syncope, especially the elderly. A drop in systolic BP by 20 mm Hg or a drop in diastolic BP by 10 mm Hg within 3 min of standing upright conἀrms orthostatic syncope (Grubb 2005). Blood pressure measurements in both arms may help diagnose subclavian steal syndrome. A complete cardiovascular examination should be done to determine the presence of structural heart disease. The presence of any abnormalities on neurologic may indicate cerebrovascular disease. In elderly patients with syncope, carotid sinus hypersensitivity may€be suggested by gentle carotid sinus massage demonstrating a 3-second pause and drop in systolic BP by 50 mm Hg that is associated with symptoms. 21.3.2â•… Electrocardiogram A standard 12-lead electrocardiogram is cheap, safe, establishes the cause of syncope in 5% of cases, and suggests an etiology in another 5% of patients (Linzer et al. 1997a, 1997b). Therefore, an electrocardiogram should be considered a standard part of the diagnostic evaluation of syncope. Most patients with syncope, however, have a normal electrocardiogram. A normal electrocardiogram indicates a low likelihood of cardiac syncope and is associated with an excellent prognosis, particularly in young patients with syncope (Brignole et al. 2001). Important observations to be noted on the electrocardiogram include varying degrees on conduction block (e.g., Mobitz II atrioventricular block, bifascicular block),

318 Sudden Death in Epilepsy: Forensic and Clinical Issues

evidence of structural heart disease (e.g., pathalogical Q waves), and predisposing conditions to arrhythmias (e.g., Wolff-Parkinson-White syndrome, Brugada syndrome, long QT syndrome). The presence of left bundle branch block has been identiἀed as a strong predictor of arrhythmic cause of syncope in clinical studies of patients with recurrent syncope who underwent prolonged electrocardiographic monitoring (Sud et al. 2009). 21.3.3â•… Echocardiography The main value of echocardiography is to establish the presence or absence of structural heart disease in a patient with syncope. Although echocardiography is safe and routinely performed in the evaluation of syncope, the diagnostic yield in patients with a normal physical and electrocardiogram is extremely low. Therefore, the current recommendation is to perform echocardiography if there is suspicion of cardiac syncope or structural heart disease based on history, physical examination, or electrocardiogram (Brignole et al. 2001). As mentioned earlier, the presence of prior myocardial infarction, cardiomyopathy, or valvular heart disease greatly increases the likelihood that syncope is due to a life-threatening ventricular arrhythmia. Untreated patients with syncope and structural heart disease have a 10% to 20% annual risk of sudden cardiac death (Kapoor 2000). In Sudden Cardiac Death Heart Failure Trial (SCD-HeFT) syncope was associated with increased mortality regardless of treatment arm (placebo, amiodarone, or cardiac deἀbrillator) and predicted total (hazard ratio 1.41, 95% conἀdence interval 1.13–1.76), and cardiovascular death (hazard ratio 1.55, 95% conἀdence interval 1.19–2.02) (Olshansky et al. 2008). Thus, patients with unexplained syncope and structural heart disease should be referred urgently to a cardiologist or cardiac electrophysiologist. 21.3.4â•… Prolonged EGG Monitoring Obtaining an electrocardiogram during the syncopal event is the gold standard for the diagnosis of cardiac syncope. Holter monitoring is a form of continuous electrocardiographic monitoring usually performed for 24 to 48 h. Because syncopal episodes are usually infrequent and sporadic, the diagnostic yield of Holter monitoring is low (Linzer et al. 1997a, 1997b). Holter monitoring is recommended for patients with daily episodes of syncope or presyncope. Transtelephonic event monitors are small recording devices that are worn continuously by the patient and can be activated by the patient. Some devices (i.e., loop recorders) are capable of capturing both retrospective (up to 4 min) and prospective (up to 1 min) electrocardiogram recordings when activated. These loop recorders are therefore preferred over Holter monitors to diagnose cardiac arrhythmic syncope. In order to be diagnostic, patients’ symptoms must correlate precisely with the rhythm abnormality as asymptomatic sinus bradycardia or second-degree AV block during sleep may be present in normal healthy individuals. In many patients, episodes of syncope are sometimes extremely rare, occurring once or twice a year. Event monitors, which are usually worn for as long as 30 days, are unlikely to reveal a diagnosis in these patients. Small implantable loop recorders can record a single or multiple lead electrocardiogram for 18 to 24 months. These devices can be implanted under the skin in the subcutaneous tissue of the anterior chest wall. The device can be triggered by the patient with a handheld activator or triggered automatically based on programmable parameters. Prolonged monitoring with implantable loop recorders improves the diagnostic yield for unexplained syncope. In a Canadian study, prolonged monitoring

Not Seizure but Syncope

319

with implantable loop recorders was more likely to result in a diagnosis than was with the conventional testing (55% versus 19%, P = 0.001) (Krahn et al. 2001). The implantable loop recorders are extremely useful for prolonged monitoring and detection of cardiac arrhythmias if the clinical suspicion for cardiac arrhythmias causing syncope remains high in spite of negative or nondiagnostic invasive electrophysiology study. Figure 21.2 shows retrieved episode data from an implantable loop recorder during a syncopal episode in a patient who had a prior negative invasive cardiac electrophysiology study. An underlying rhythm of a rapid monomorphic ventricular tachycardia at 176 beats/min during the episode conἀrmed cardiac arrhythmia as a culprit cause of syncope in this patient. 21.3.5â•… Tilt Table Testing Upright tilt table testing provides diagnostic evidence indicating a susceptibility to neurally mediated reflex syncope and is considered the gold standard for establishing a diagnosis. The test involves observing the patient strapped on a specially designed table that tilts at an angle of 60° to 80°. Although there is no one standardized tilt table testing protocol, in general the patients are observed in the tilted position for 30 to 45 min initially without administering pharmacological provocative agents (e.g., intravenous isoproterenol or sublingual nitroglycerine) for symptoms and hemodynamic changes associated with near or full-blown syncope, and if necessary, subsequently after. The sensitivity of the test is about€50% without provocative agents and approximately 65% with these drugs. When an abnormal response occurs, especially without any pharmacologic provocation, the speciἀcity of the test is about 90%. A tilt table test is considered positive only if the patients symptoms are reproduced, accompanied by a drop in arterial BP >20 mm Hg (vasodepressor response), decrease in heart rate >10% from baseline (cardioinhibitory response), or both (mixed response). Most patients with vasovagal syncope have a mixed response. In general, tilt table testing is reserved for patients with recurrent syncope or a single high-risk syncope episode when there is no evidence of structural heart disease or when other causes of syncope have been excluded. In patients with a single episode of uncomplicated syncope where the clinical picture is typical for neurally mediated reflex syncope, tilt table testing is unnecessary. Furthermore, tilt table testing is not useful in establishing the diagnosis of neurally mediated reflex syncope with a speciἀc trigger (e.g., micturation) (Brignole et al. 2001; Schnipper and Kapoor 2001; Grubb 2005).

Figure 21.2╇ Retrieved episode data of syncopal event in a patient with unexplained syncope in

whom an implantable loop recorder was inserted for a prolonged monitoring of cardiac rhythm. The tracing shows an initial three beats of sinus rhythm followed by a premature ventricular complex and a sinus beat before the onset of a rapid ventricular tachycardia at 176 beats/min (mean cycle length of 340 ms).

320 Sudden Death in Epilepsy: Forensic and Clinical Issues

21.3.6â•… Electrophysiologic Testing The electrophysiology study is an invasive study in which electrode-tipped catheters are placed in the heart in speciἀc locations and stimulation protocols are performed to assess the cardiac electrical system. In general, electrophysiology study is indicated when syncope is associated with structural heart disease. The role of electrophysiology study in recurrent unexplained syncope with negative tilt table test and no structural heart disease is not ἀrmly established. In patients with unexplained syncope and structural heart disease (e.g., prior myocardial infarction) induction of ventricular tachycardia indicates a poor prognosis (up to 30% mortality in 3 years) (Gouello et al. 1992). Ventricular tachycardia is the most common abnormality revealed by electrophysiology study. Since recent guidelines recommend implantable cardioverter deἀbrillator implantation for primary prevention of ventricular arrhythmia in patients with severe left ventricular systolic dysfunction (ejection fraction <30–35%), as such, electrophysiology study is rarely performed for the diagnosis of syncope in patients who fulἀll these criteria for ICD implantation. However, for patients with syncope and structural heart disease with preserved left ventricular systolic function, electrophysiology study has prognostic value as a negative electrophysiology study in these patients and suggests a low risk for sudden cardiac death (Calkins 2004). For the assessment of sinus node dysfunction, electrophysiology study has limited value. Although an abnormal sinus node recovery time indicates sinus node dysfunction, a normal sinus node function during electrophysiology study does not exclude bradyarrhythmia as the cause of syncope (Gouello et al. 1992). Abnormalities of impulse conduction during electrophysiology study that establish a probable cause of syncope are (1) delay in the infra-His conduction system at baseline (HV interval >100 ms), and/or (2) block below the His conduction system (Gouello et al. 1992). 21.3.7â•… Miscellaneous Tests Transient ischemic attacks from carotid atherosclerosis do not cause loss of consciousness. Therefore, carotid ultrasonography is not indicated in the evaluation of syncope. Blood tests rarely assist in the diagnosis of syncope, unless a metabolic etiology is strongly suspected (e.g., hypoglycemia). Brain imaging with computerized tomography (CT) and magnetic resonance imaging (MRI) are usually unnecessary and only yields a diagnosis when there is a focal neurologic deἀcit or a witnessed seizure. If there are no signs and symptoms of seizure, electroencephalography is not useful in the diagnosis of syncope (Britton 2004). Simultaneous electroencephalography and electrocardiography (especially with video monitoring) may help to diagnose frequent episodes that cannot be distinguished as syncope or seizure (Britton 2004). Myocardial ischemia is an unlikely cause of syncope, especially in the absence of angina or exertional symptoms. Therefore, stress testing and cardiac catheterization should be reserved for patients with syncope associated with exertion when the suspicion for coronary artery disease is high.

21.4â•… Syncope and Epilepsy There are reports of patients with recurrent syncope who develop cardiac asystole, transient AV block, and severe sinus bradycardia (heart rate <30 beats/min) preceded by partial

Not Seizure but Syncope

321

seizures often originating from the temporal lobe (Kouakam et al. 2009). In patients with recurrent traumatic/convulsive syncope, ictal bradycardia syndrome should be suspected. Ictal bradycardia is a rare but clinically signiἀcant ἀnding and may predispose a patient to SUDEP (Britton 2004). Delayed loss of muscle tone (which is uncommon in patients with temporal lobe epilepsy) was a common feature of ictal bradycardia (Schuele et al. 2007). Cerebral hypoperfusion during bradycardia or asystole is postulated to be€ the mechanÂ� ism€for sudden late atonia. Patients with epilepsy who present with sudden falls and collapse late in the course of a typical seizure warrant investigation. Also note that as many as 20% to 30% patients diagnosed to have epilepsy do not have epilepsy, and that many seizure-like attacks have cardiovascular cause. In these situations, strategies of videoelectroencephalography and electrocardiography monitoring or a multidisciplinary apÂ�Â� proach that involves a head-up tilt table test and carotid sinus massage during continuous electrocardioÂ�graphy,€electroencephalography, and blood pressure monitoring are helpful in establishing the diagnosis (Zaidi et al. 2000; Schuele et al. 2007). Patients with ictal bradycardia, in addition to anticonvulsant medication, may require pacemaker implantation (Strzelczyk et al. 2008; Novy et al. 2009).

21.5â•… Conclusion Syncope is a common and important clinical problem that poses several challenges. An accurate clinical history focusing on triggers of the episode, prodromal symptoms, characteristics of the event itself, and rate of recovery from loss of consciousness usually allows the clinician to make a presumptive diagnosis. Physical examination and a routine electrocardiogram can help establish the presence of structural heart disease that makes cardiac syncope more likely. The selection of investigation from this point depends on the presumptive diagnosis and the frequency of the episodes. An echocardiogram can be performed to conἀrm suspected structural heart disease. Ambulatory electrocardiogram monitoring with an event/loop recorder should be performed if recurrent syncope is frequent. Implantable loop recorders can be utilized for 18 to 24 months if syncope is very infrequent. Tilt table testing may be useful in patients with unexplained recurrent syncope or high-risk syncope in the absence of structural heart disease. Invasive cardiac electrophysiology testing is recommended for patients with suspected cardiac syncope and structural heart disease who do not meet primary prevention guidelines for implantable cardioverter deἀbrillator implantation. In patients with epilepsy, syncope may be related to ictal bradycardia which is considered to be a risk factor for SUDEP. Video-electroencephalography and electrocardiography monitoring may be useful in establishing the diagnosis in these cases.

References Ackerman, M. J. 1998. The long QT syndrome: Ion channel diseases of the heart. Mayo Clin Proc 73 (3): 250–269. Alboni, P., M. Brignole, C. Menozzi, A. Raviele, A. Del Rosso, M. Dinelli, A. Solano, and N. Bottoni. 2001. Diagnostic value of history in patients with syncope with or without heart disease. J Am Coll Cardiol 37 (7): 1921–1928.

322 Sudden Death in Epilepsy: Forensic and Clinical Issues Ammirati, F., F. Colivicchi, and M. Santini. 2000. Diagnosing syncope in clinical practice. ImplementaÂ� tion of a simpliἀed diagnostic algorithm in a multicentre prospective trial—The OESIL 2 study (Osservatorio Epidemiologico della Sincope nel Lazio). Eur Heart J 21 (11): 935–940. Brignole, M., P. Alboni, D. Benditt, L. Bergfeldt, J. J. Blanc, P. E. Bloch Thomsen, J. G. van Dijk et al. 2001. Guidelines on management (diagnosis and treatment) of syncope. Eur Heart J 22 (15): 1256–1306. Britton, J. W. 2004. Syncope and seizures—Differential diagnosis and evaluation. Clin Auton Res 14 (3): 148–159. Calkins, H. 2004. Syncope. In Cardiac Electrophysiology: From Cell to Bedside, ed. D. Zipes and J. Jalife. Philadelphia, PA: Saunders. Demaksian, G., and L. E. Lamb. 1958. Syncope in a population of healthy adults. JAMA 168: 1200–1207. Gouello, J. P., J. Victor, and A. Tadei. 1992. Natural history of syncope of undetermined origin with inconclusive electrophysiologic examination. Arch Mal Coeur Vaiss 85 (3): 297–302. Grubb, B. P. 2005. Neurocardiogenic syncope and related disorders of orthostatic intolerance. Circulation 111 (22): 2997–3006. Kapoor, W. N. 2000. Syncope. N Engl J Med 343 (25): 1856–1862. Kouakam, C., C. Daems, L. Guedon-Moreau, A. Delval, D. Lacroix, P. Derambure, and S. Kacet. 2009. Recurrent unexplained syncope may have a cerebral origin: Report of 10 cases of arrhythmogenic epilepsy. Arch Cardiovasc Dis 102 (5): 397–407. Krahn, A. D., G. J. Klein, R. Yee, and A. C. Skanes. 2001. Randomized assessment of syncope trial: Conventional diagnostic testing versus a prolonged monitoring strategy. Circulation 104 (1): 46–51. Linzer, M., E. H. Yang, N. A. Estes III, P. Wang, V. R. Vorperian, and W. N. Kapoor. 1997a. Diagnosing syncope: Part 1. Value of history, physical examination, and electrocardiography. Clinical Efficacy Assessment Project of the American College of Physicians. Ann Intern Med 126 (12): 989–996. Linzer, M., E. H. Yang, N. A. Estes III, P. Wang, V. R. Vorperian, and W. N. Kapoor. 1997b. Diagnosing syncope: Part 2. Unexplained syncope. Clinical Efficacy Assessment Project of the American College of Physicians. Ann Intern Med 127 (1): 76–86. Morillo, C. A., M. E. Camacho, M. A. Wood, D. M. Gilligan, and K. A. Ellenbogen. 1999. Diagnostic utility of mechanical, pharmacological and orthostatic stimulation of the carotid sinus in patients with unexplained syncope. J Am Coll Cardiol 34 (5): 1587–1594. Nair, N., F. A. Padder, and B. K. Kantharia. 2003. Pathophysiology and management of neurocardiogenic syncope. Am J Manag Care 9 (4): 327–334; quiz 335–336. Nirkko, A. C., and R. W. Baumgartner. 2006. Syncope. Front Neurol Neurosci 21: 239–250. Novy, J., A. Carruzzo, P. Pascale, M. Maeder-Ingvar, D. Genne, E. Pruvot, P. A. Despland, and A. O. Rossetti. 2009. Ictal bradycardia and asystole: An uncommon cause of syncope. Int J Cardiol 133 (3): e90–e93. Nyman, J. A., A. D. Krahn, P. C. Bland, S. Griffiths, and V. Manda. 1999. The costs of recurrent syncope of unknown origin in elderly patients. Pacing Clin Electrophysiol 22 (9): 1386–1394. Olshansky, B., J. E. Poole, G. Johnson, J. Anderson, A. S. Hellkamp, D. Packer, D. B. Mark, K. L. Lee, G. H. Bardy, for the CN-SCD-HeFT Investigators. 2008. Syncope predicts the outcome of cardiomyopathy patients: analysis of the SCD-HeFT study. J Am Coll Cardiol 51 (13): 1277–1282. Rose, M. S., M. L. Koshman, S. Spreng, and R. Sheldon. 2000. The relationship between healthrelated quality of life and frequency of spells in patients with syncope. J Clin Epidemiol 53 (12): 1209–1216. Sarasin, F. P., M. Louis-Simonet, D. Carballo, S. Slama, A. Rajeswaran, J. T. Metzger, C. Lovis, P. F. Unger, and A. F. Junod. 2001. Prospective evaluation of patients with syncope: A populationbased study. Am J Med 111 (3): 177–184. Schnipper, J. L., and W. N. Kapoor. 2001. Diagnostic evaluation and management of patients with syncope. Med Clin North Am 85 (2): 423–456, xi.

Not Seizure but Syncope

323

Schuele, S. U., A. C. Bermeo, A. V. Alexopoulos, E. R. Locatelli, R. C. Burgess, D. S. Dinner, and N. Foldvary-Schaefer. 2007. Video-electrographic and clinical features in patients with ictal asystole. Neurology 69 (5): 434–441. Soteriades, E. S., J. C. Evans, M. G. Larson, M. H. Chen, L. Chen, E. J. Benjamin, and D. Levy. 2002. Incidence and prognosis of syncope. N Engl J Med 347 (12): 878–885. Strzelczyk, A., S. Bauer, S. Knake, W. H. Oertel, H. M. Hamer, and F. Rosenow. 2008. Ictal asystole in temporal lobe epilepsy before and after pacemaker implantation. Epileptic Disord 10 (1): 39–44. Sud, S., G. J. Klein, A. C. Skanes, L. J. Gula, R. Yee, and A. D. Krahn. 2009. Predicting the cause of syncope from clinical history in patients undergoing prolonged monitoring. Heart Rhythm 6 (2): 238–243. Zaidi, A., P. Clough, P. Cooper, B. Scheepers, and A. P. Fitzpatrick. 2000. Misdiagnosis of epilepsy: Many seizure-like attacks have a cardiovascular cause. J Am Coll Cardiol 36 (1): 181–184.

Syncope, Seizures, and SUDEP Case Histories Claire M. Lathers Paul L. Schraeder Michael W. Bungo

22

Contents 22.1 Case Histories 22.2 Discussion References

325 330 330

22.1╅ Case Histories Various other risk factors exist for sudden€death and include arrhythmogenic factors, respiratory factors such as hypoxia, and psychological factors (Lathers et al. 2008). Overlapping mechanisms may apply to the risk of sudden death occurring in epilepsy and in cardiac disease. There is a potential interactive role of genetically determined subtle cardiac risk factors for arrhythmias with a predisposition for seizure-related cardiac arrhythmias. Fatal syncopal events may masquerade as sudden unexpected death in epilepsy (SUDEP) but this possibility has yet to be investigated. Syncope results from acute cerebral hypoperfusion and must be differentiated from epileptic seizures in cases where convulsive activity is observed during the syncopal episode. The following cases briefly explore the interactions of disordered electrical potentials in the brain and disordered electrical potentials in the heart. So often the clinician is presented with cases of sudden collapse. At times there are obvious cardiac etiologies; at times there are obvious central nervous system (CNS) etiologies; at times there are interactive cause and effect with cardiac disorders leading to CNS effects or visa versa; however, all too often there are subtle interplays with genetic predispositions, environmental interaction, therapeutic interventions, or unknown organic disease. The consequences are extreme, and therefore a thorough investigation and understanding of the mechanisms leading to sudden death are of paramount importance. The fact that such a large number of these instances occur in young people only underscores the need for us to explore these issues further.

325

326 Sudden Death in Epilepsy: Forensic and Clinical Issues

Co ng enit al Lo ng Q T Sy ndro me, Recurrent €Sy nco pe, and Sudden Deat h at Y o ung A g e A 10-year-old girl presented with recurrent episodes of syncope while swimming. She was diagnosed with type 1 LQTS. A 36-year-old asymptomatic male was unexpectedly diagnosed with type 2 LQTS. Familial history revealed audio-triggered syncope and sudden death at a young age. A 16-year-old male teenager died in his sleep. Postmortem testing revealed type 3 LQTS. Co mment s by A k k erhuis, Baars,€Marcelis, Ak k erhusi, and Wilde A leading cause of sudden death at a young age is congenital LQTS. Gene mutation for encoding myocardial ion channel proteins lead to prolonged QT interval and abnormal ST-T segments demonstrated in 12-lead ECG. Syncope or sudden cardiac death due to ventricular tachyarrhythmias may occur. Genotype-speciἀc differences in ECG-abnormalities and triggers for cardiac events may differentiate the type of LQTS and trigger initiation of genotype-speciἀc treatment before the results of genetic testing are known. Genetic testing to determine the genetic substrate, genotype-speciἀc treatment and possibility of treatment with an implantable cardioverter-Â�deἀbrillatory all contribute to improved prognosis of patients with LQTS. The authors recommend that young patients with unexplained recurrent syncope after speciἀc stimuli and those with atypical forms of epilepsy undergo a cardiological evaluation. Co mment s by L at hers, Schraeder, and Bung o Readers are also referred to another case of the video monitoring of long QT syndrome (LQTS)–related aborted sudden death by Rossenbacker et al. (2007). The congenital LQTS is an alteration in cardiac polarization that results in an increased risk of cardiac arrhythmias in young people. As a result of hypoperfusion, syncope, often with convulsive activity occur, so that up to 10% of cases may have an erroneous diagnosis of epilepsy. Typically, the clinical events are associated with preceding physical exertion with initiation of ventricular tachycardia, including torsades de pointes. The high mortality of LQTS and the potential for prevention with an implanted deἀbrillator emphasize the importance of accurate diagnosis from the clinical history of exercise-related loss of consciousness and ECG recording (MedinaVillanueva et al. 2002). The overlap of LQTS with epilepsy occurs when the patient presents with what appears to be a convulsive seizure. The risk of death resulting from an incorrect diagnosis is obvious. However, the observation that up to 10% of these patients, who are at risk for premature demise, are diagnosed as having epilepsy raises two questions. The ἀrst is whether some persons with epilepsy whose deaths are attributed to SUDEP in fact had LQTS. The second is whether some persons with epilepsy may also have LQTS, predisposing them for neurogenically induced arrhythmias. It has become common knowledge the long QT interval syndrome is familial and that family members of an index case should be screened with ECG and genetic testing. Research needs to be undertaken to address the question of whether

Syncope, Seizures, and SUDEP

327

there needs to be screening of family members of victims of SUDEP for long QT and other genetically determined ion channel disorders. Case summarized from Akkerhuis, J. M., H. F. Baars, C. L. Marcelis, K. M. Akkerhusi, and A. A. Wilde. 2007. Congentital long QT-syndrome: The cause of recurrent syncope and sudden death at a young age. Ned Tijdschr Geneeskd 151: 2357–2364.

New SCN5A Mut at io n in a SUDEP V ict im wit h Idio pat hic Epilepsy A 17-year-old teenage girl experienced symptoms of déjà vu, a feeling of fear, and the sensation of an electric current in her chest, accompanied by palpitations and flushing. These events lasted for a few seconds with some impairment of awareness. Subsequently, she experienced generalized seizures and some myoclonic events. She was not diagnosed as having epilepsy until age 24. Her EEG demonstrated bursts of€ bilateral spike and wave. No ECG was performed. She was treated with carbaÂ� mazepine monotherapy, with subsequent therapy changes to valproic acid and lamotrigine. Because of side effects of tremor and tiredness, the valproic acid was tapered and the patient maintained on lamotrigine monotherapy of 100 mg/day. She was found dead in bed at age 25, and because of a negative autopsy, was determined to have died of SUDEP. Molecular genetic analysis found the patient to be heterozygous for DNA sequencing of (LQTS) associated new missense mutation of the SCN5A gene. This gene codes for the cardiac sodium channel, voltage-gated, type V, alpha subunit. Co mment s by A urlien, Leren, Taubo ll, and Gjerst and Ion channel mutations involving the SCN5A gene are associated with Brugada syndrome and LQTS. Idiopathic epilepsy is also thought to be associated with ion channelopathies. The cardiac SCN5A mutation detected in this case may well have been a factor in explaining the death of this patient, possibly in combination with a terminal seizure. Confounding this association is the use of lamotrigine, a drug that has the effect of blocking sodium channel function with widening of the QRS complex and right-axis deviation. While the putative mechanism of death in this case may have been multifactorial (i.e., seizures, predisposition to cardiac channelopathy related arrhythmia, and effects of medication), to date, there is little clinical evidence directly connecting cardiac ion channel dysfunction to those associated with epilepsy. More research needs to be undertaken to investigate whether certain persons with epilepsy have a predisposition to cardiac arrhythmias that is the result of an ion channel gene mutation that affects both the heart and the brain. Co mment s by L at hers, Schraeder, and Bung o The issue of a common ion channel dysfunction that may underlie both the predisposition to seizures and cardiac arrhythmias is an important potential area for investigation. However, the potential role for inherited ion channel dysfunction as a

328 Sudden Death in Epilepsy: Forensic and Clinical Issues

factor in the occurrence of seizures may be operative in primary generalized epilepsy (Mulley et al. 2003). The patient in this case had both complex partial and generalized seizures, with the latter being those overwhelmingly associated with the risk of SUDEP. While the patient had temporal lobe symptoms (i.e., déjà vu, feelings of fear, and electric sensations), the fact that she also had myoclonic events in combination with generalized tonic–clonic seizures raises the possibility that she may have had two types of epilepsy, namely localization-related temporal lobe seizures with or without secondary generalization, and juvenile myoclonic epilepsy. The phenotype of the latter may be determined by a number of genes on chromosome 6p including a locus in the human leukocyte antigen region and in some instances on 1p (Berkokovic 1998). The potential for interactions between genetic defects that affect ion channel function in the heart and the brain needs to be the subject of future research, especially when combined with the potential for antiepileptic drug (AED)-related effects at the ion channel level. Case summarized from Aurlien, D., T. P. Leren, E. Tauboll, and L. Gjerstand. 2008. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18: 158–160.

Ict al Asy st o le in Tempo ral Lo be Epilepsy befo re and aft er Pacemak er Implant at io n A 41-year-old male presented with refractory partial seizures resulting in syncope leading to severe head trauma. Presurgical video-EEG monitoring demonstrated two episodes of ictal bradycardia followed by asystole and syncope. Implantation of a cardiac pacemaker provided a seizure-free, syncope-free 9-month follow-up period. Co mment s by S t rz elcz y k , Bauer,€Knak e,€Oert el, Hamer, and Ro seno w Ictal bradycardia or asystole in patients with epilepsy presenting with ictal falls may be a factor in some cases of SUDEP. This case documents that cardiac pacemaker implantation in addition to continuation of AEDs may optimize seizure control while preventing ictal syncope and/or trauma associated with a fall. Co mment s by L at hers, Schraeder, and Bung o When treating persons with epilepsy and a known history of ictal falls and/or ictal bradycardia or asystole, one must also consider the psychological stress related to fear of falling and injury consequent to having a seizure. One case study (Schraeder et al. 1983) described a young athletic male who was listening to his minister’s sermon describing gory details of how martyrs were tortured for their beliefs, passed out and exhibited what appeared to be generalized tonic–clonic seizures. Subsequent repetition of the offending verbal passages, with EEG monitoring, conἀrmed that the clinical seizures were the consequence of psychogenically induced asystole lasting more than 30 s with resultant electrocerebral silence. No epileptiform activity occurred on the EEG. This patient became seizure-free after implantation of a cardiac pacemaker

Syncope, Seizures, and SUDEP

329

and emphasizes the strong contribution of cerebral output on cardiac function. Stress itself may be a risk factor for sudden death in persons with epilepsy (Lathers and Schraeder 2006) and is discussed in Chapter 17. Carinci et al. (2007) note that the clinical distinction between cardiovascular and epileptic causes of loss of consciousness may be difficult to discern and becomes a diagnostic challenge when a primary epileptic seizure secondarily causes asystole. The ictal bradycardia syndrome refers to correlation of epilepsy with severe bradycardia or asystole, both of which may be a mechanism of SUDEP. Nonetheless, Carinci et al. (2007) conclude that asystole induced by partial seizures is a rare cause of syncope. Patients with potentially fatal mechanisms of syncope may be diagnosed inaccurately to have seizures and persons with epilepsy may be at risk for ictal arrhythmias as a cause for sudden death (Al Aloul et al. 2007). It is clinically important to accurately diagnose patients in both categories. Al Aloul et al. (2007) cited the fact that the emergence of Brugada pattern on the electrocardiogram in response to administration of class IA or IC antiarrhythmic agents has been routinely recognized as a means of inducing concealed Brugada syndrome, which is a risk factor for sudden death. They reported a case in which the patient demonstrated induced Brugada pattern after phenytoin, a class IB antiarrhythmic agent, was administered. Phenytoin levels were supratherapeutic. The authors recommend that all patients with supraÂ� therapeutic phenytoin levels be evaluated for emergence of the Brugada pattern in the electrocardiogram. Britton and Benarrouch (2006) also emphasize that, although pathophysiologically distinct, syncope and seizures often exhibit the same clinical phenomena, making it difficult to diagnose clinically the cause of a patient’s seizurelike activity. Furthermore, in some patients, both seizure and syncope may coexist. Syncope may be associated with seizure-like motor manifestations, and seizure may be complicated by cardiac arrhythmia and syncope. Combined EEG/ECG telemetry€is necessary to establish the accurate diagnosis and allow evaluation of neuroanatomic circuitry involved in the production of the cardiovascular manifestations of seizures. Case summarized from Strzelczyk, A., S. Bauer, C. Knake, W. H. Oertel, H. M. Hamer, and F. Rosenow. 2008. Ictal asystole in temporal lobe epilepsy before and after pacemaker implantation. Epileptic Disord 10: 39–44.

Sudden Deat h o f Pat sy C ust is: Geo rg e Washing t o n o n Sudden Unex plained Deat h in Epilepsy Patsy Custis, the stepdaughter of George Washington, suffered from what, by€descripÂ� tion, most likely was convulsive epilepsy that began at age 6. The seizures were uncontrolled by treatment of the time that included bleeding, purging, mercury, and cinchona along with other decoctions. She was taken to the healing baths in what is now Berkeley Springs, WV, and was made to wear an iron ring, since such a ring was thought to protect against seizures. All of these interventions, unsurprisingly, were of no avail. Despite being described by George Washington as being in “better health and spirits” than in the past, on June 19, 1773, at age 17, she died in the afternoon within 2 min of “one of her usual ἀts,” “with scarcely a sigh.”

330 Sudden Death in Epilepsy: Forensic and Clinical Issues

Co mment s by D o hert y ( 2004) Patsy’s demise met the criteria for probable SUDEP as established by Leestma and Koenig (1968) and Jay and Leestma (1981). In summary, she had epilepsy, was in a reasonably good state of health, the fatal event was sudden with or without a concurrent seizure, but not status epilepticus, it occurred during normal activity, and no other obvious explanation for the sudden death was extant. Her death is one of the ἀrst documented cases of SUDEP, as described by her stepfather, George Washington. Co mment s by L at hers, Schraeder, and Bung o George Washington clearly was a witness to the fatal event since he described her as having “one of her usual ἀts,” also noting that she manifested “scarcely a sigh” at the time of her demise, a description that implies apnea. From the distance of almost a quarter of a millennium, Patsy seemed to have met the published criteria for probable SUDEP. Although presently we are not much farther along in our understanding of the mechanism of death in SUDEP than we were in the eighteenth century (a circumstance that, hopefully, will be improved with research in the near future), the witnessed description is consistent with seizure-related apnea as the probable mechanism in this case in that George Washington observed his stepdaughter as having “scarcely a sigh” at the time of her demise. While having had uncontrolled convulsive seizures since childhood put her into a higher risk category, the possibility of the risk factor of an underlying nonprogressive brain disease is unknown. While she also may have been exposed to changes of medications, this risk factor would presumably be operative only in the circumstance of medications known to have efficacy in the treatment of epilepsy. While being female had a slight mitigating effect of risk of SUDEP, she was a young adult, and as mentioned earlier, had uncontrolled seizures. SUDEP is neither a function of the victims’ social nor socioeconomic status. Case summarized from DeToledo, J. C., M. B. DeToledo, and M. R. Lowe. 1999. Epilepsy and sudden death: Notes from George Washington’s diaries on the illness and death of Martha Parke-Custis (1756–1773). Epilepsia 40: 1835–1836.

22.2â•…Discussion These cases illustrate the spectrum of factors that may correlate with risk of sudden death. These include cardiac channelopathies, acute cerebral hypoperfusion, and presumed cardiogenic bradycardiac episodes that only bilaterally were found to be ictally induced. Nonetheless, the latter individuals required treatments for both seizures and asystole (i.e., AEDs and/or a pacemaker).

References Akkerhuis, J. M., H. F. Baars, C. L. Marcelis, K. M. Akkerhusi, and A. A. Wilde. 2007. Congentital long QT-syndrome: The cause of recurrent syncope and sudden death at a young age. Ned Tijdschr Geneeskd 151: 2357–2364.

Syncope, Seizures, and SUDEP

331

Al Aloul, B., A. S. Adabag, M. A. Houghland, and V. Tholakanahalli. 2007. Brugada pattern electrocardiogram associated with supratherapeutic phenytoin levels and the risk of sudden death. Pacing Clin Electrophysiol 30: 713–715. Aurlien, D., T. P. Leren, E. Tauboll, and L. Gjerstand. 2008. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18: 158–160. Berkokovic, S. F. 1998. Genetics of epilepsy syndromes. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr. and T. A. Pedley, 217–224. Philadelphia, PA: Lippincott Raven. Britton, J. W., and E. Benarroch. 2006. Seizures and syncope: Anatomic basis and diagnostic considerations. Clin Auton Res 16: 18–28. Carinci, V., G. Barbato, A. Baldrati, and G. Di Pasquale. 2007. Asystole induced by partial seizures: A rare cause of syncope. Pacing Clin Electrophysiol 30: 1416–1419. DeToledo, J. C., M. B. DeToledo, and M. R. Lowe. 1999. Epilepsy and sudden death: Notes from George Washington’s diaries on the illness and death of Martha Parke-Custis (1756–1773). Epilepsia 40: 1835–1836. Doherty, M. J. 2004. The sudden death of Patsy Custis, or George Washington on sudden unexplained death in epilepsy. Epilepsy Behav 5: 598–600. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. Acts Neurol Scand s82, 63: 1–66. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. Mystery of sudden death: Mechanisms for Risks. Epilepsy Behav 12: 3–24. Leestma, J. E., and K. K. Koenig. 1968. Sudden death and phenothiazines. Arch Gen Psychiatry 18: 137–148. Medina-Villanueva, A., C. Rey-Galan, A. Concha-Torre, and J. R. Gutierrez-Martinez. 2002. Long QT syndrome presented as epilepsy. Rev Neurol 35: 346–348. Mulley, J. C., I. E. Scheffer, S. Petrou, and S. F. Berkovic. 2003. Channelopathies as a genetic cause of epilepsy. Curr Opin Neurol 16: 171–176. Rossenbacker, T., D. Nuyens, W. Van Paesschen, and H. Heidbuchel. 2007. Epilepsy? Video monitoring of long QT syndrome-related aborted sudden death. Heart Rhythm 4: 1366–1367. Strzelczyk, A., S. Bauer, C. Knake, W. H. Oertel, H. M. Hamer, and F. Rosenow. 2008. Ictal asystole in temporal lobe epilepsy before and after pacemaker implantation. Epileptic Disord 10: 39–44.

Sudden Death in Epilepsy Relationship to the Sleep– Wake Circadian Cycle and Fractal Physiology

23

John D. Hughes Susumu Sato

Contents 23.1 Introduction 23.2 Circadian Physiology 23.3 The Sleep–Wake Circadian Cycle 23.4 Seizures and the Sleep–Wake Cycle 23.5 Cardiovascular Physiological Activity during Sleep 23.6 Heart Rate Variability and Fractal Physiology 23.7 A Proposed Pathophysiology of SUDEP 23.8 Conclusion References

333 334 334 334 337 339 340 342 342

23.1â•… Introduction Most case series of patients experiencing sudden death in the setting of epilepsy document a signiἀcant preponderance of cases occurring during sleep, especially nocturnal sleep (Langan et al. 2005; Kloster and Engelskjon 1999). However, the mechanisms underlying this strong tendency have gone relatively unexplored when considering the etiology of sudden unexpected death in epilepsy (SUDEP). In general, three main mechanisms have been proposed to explain SUDEP: (1) a seizure poses an insurmountable autonomic stress to the heart resulting in an arrhythmia, (2) the epileptogenic process itself deranges central autonomic network function substantially, thus rendering the heart more vulnerable to arrhythmogenesis, and (3) a prolonged postictal apnea results in death. Of course, these are not mutually exclusive explanations, and they may all play a role in SUDEP, with potentially additive effects in a given patient. The ἀrst two mechanisms appear to be potentially highly synergistic. In this chapter, we will review aspects of the sleep–wake cycle and other circadian factors that are potentially relevant to the pathogenesis of SUDEP, with emphasis on the role of dysfunctional central autonomic cardiac regulation. Additionally, we will propose a unifying theory of the pathogenesis of SUDEP based on the reviewed evidence. Unfortunately, because of the very limited research in this area, much of our discussion will be speculative, but we hope it may stimulate much needed research.

333

334 Sudden Death in Epilepsy: Forensic and Clinical Issues

23.2â•… Circadian Physiology The central nervous system generates an endogenous circadian rhythm of approximately 24 h that originates from the suprachiasmatic nucleus (SCN) of the hypothalamus (Moore 2007) and affects virtually every organ in the body; as a result, many physiological phenomena exhibit circadian fluctuations to some degree. These rhythms tend to be entrained to the environmental light–dark cycle, moderated via the retinohypothalamic tract, which projects information about environmental light to the suprachiasmatic nucleus. The sleep– wake cycle is generally entrained to this SCN generated circadian cycle but may have a 24-h circadian cycle of its own, independent of the suprachiasmatic nucleus-generated cycle, but rather determined by various social and professional factors (i.e., shift work) or due to a pathological “phase shift.” Signiἀcant changes in diverse physiological systems take place during the three different components of the sleep–wake cycle: wakefulness, nonrapid eye movement (NREM) sleep, and rapid eye movement (REM) sleep. Of particular interest with regard to SUDEP are the sleep-related changes that take place in the brain that play a role in the occurrence of seizures, as well as changes that take place in the autonomic nervous system that might play a role in cardiac arrhythmogenesis. These will be highlighted in the discussion that follows.

23.3â•… The Sleep–Wake Circadian Cycle Sleep consists of a series of cycles within the circadian cycle, called ultradian cycles. Each cycle lasts approximately 90 min and starts with a series of deepening NREM stages culminating in slow-wave sleep (stages 1 to 3). NREM sleep consists of relatively synchronized brain activity (including sleep spindles, K-complexes, and delta activity) and relatively regular, invariant activity of respiratory and other organ systems. NREM sleep may also be characterized by the presence or absence of a “cyclic alternating pattern” characterized by periodic (approximately every 20–40 s), brief “microarousals” (cyclic alternating pattern phase A) associated with increased autonomic activity, interspersed with periods free of microarousals (cyclic alternating pattern phase B) (Terzano and Parinno 2000; Halasz et al. 2004). Each ultradian cycle, of which there are normally 4 to 6 per night, theoretically ends with a period of REM sleep, which consists of tonic periods of “desynchronized” cortical activity and muscle atonia (except extraocular and respiratory muscles), and phasic periods of REMs, muscle twitches, and blood pressure variations. In general, NREM periods are longer with greater quantity of slow-wave sleep during ultradian cycles early in the night and REM periods become longer with little or no slow-wave sleep in cycles later in the night. In practice, there is signiἀcant individual variability in the number of cycles and occurrence and duration of stages within a cycle. Overall, NREM sleep comprises about 80% of sleep time, although it may comprise a greater percentage in various epileptic populations, who experience reduced REM sleep.

23.4â•… Seizures and the Sleep–Wake Cycle An awareness of a relationship between the sleep–wake cycle and the occurrence of seizures dates back to antiquity and has been the subject of intense research in the past several

Sudden Death in Epilepsy

335

decades and of several recent reviews (Foldvary-Schafer and Grigg-Damberger 2006; Malow 2005). Sleep has traditionally been viewed as an epileptogenic state, due in general to the thalamocortical synchrony that may predispose to seizure generation. A number of early studies (Janz 1962) divided seizures into those occurring during sleep (nocturnal), those occurring while awake (diurnal), and those occurring randomly (diffuse). One early formulation was that idiopathic generalized seizures (of genetic origin) tended to occur while waking (often within a short time of waking from sleep, during a presumed period of sleep inertia with some relative retained cortical synchrony), acquired partial seizures occurred predominantly while asleep, and generalized seizures in the setting of diffuse brain pathologies occurred rather randomly. Numerous studies in recent years have demonstrated that the relationship is not quite so straightforward, and while there are some epilepsy syndromes that demonstrate a profound sleep or waking predominance (such as juvenile myoclonic epilepsy with seizures shortly after awakening or autosomal dominant frontal lobe epilepsy with seizures exclusively during sleep), many seizure types defy strict sleep–wake cycle component occurrence categorization. Of particular interest with regard to SUDEP is the relationship of partial seizures that exhibit pathology in central (cortical) autonomic structures (medial temporal lobe, orbitofrontal cortex, cingulate gyrus, insular cortex). All of these structures may be implicated in the epileptogenic network involved with complex partial seizures in the setting of medial temporal lobe epilepsy, though they certainly may be the origin of partial seizures independently. In fact, diffuse cortical and subcortical limbic and autonomic centers tend to be involved in focal seizures originating in the limbic system (Mraovitch and Calando 1999). All of these structures not only tend to produce autonomic phenomena when stimulated exogenously and during seizures, but have been potentially implicated in cardiac arrhythmogenicity related to seizures (Lathers 2008). This topic is the subject of other comprehensive reviews in this book. Studies attempting to address the pattern of occurrence of partial complex seizures€of medial temporal lobe origin in terms of the sleep–wake cycle have demonstrated conflicting results. Despite the conventional wisdom dating back to and before Janz (1962) that partial seizures and in particular complex partial seizures have a signiἀcant sleep state predominance, studies by Quigg (2000) have demonstrated just the opposite, namely, that complex partial seizures in temporal lobe epilepsy are concentrated during late afternoon wakefulness and occur in the same circadian phase in human and experimental rat models. While other studies have conἀrmed the conventional wisdom (Herman et al. 2001) demonstrating a nocturnal sleep predominance, the preponderance of evidence seems to be accumulating for a late afternoon predominance for complex partial seizure occurrence (Durazzo et al. 2008). However, one aspect of the circadian timing of medial temporal lobe epilepsy that is clear is that secondary generalization tends to occur almost exclusively during nocturnal sleep (Jobst et al. 2001). Quigg (2000) has discussed the possible seizure-protecting effect€of melatonin during the night (though the data on the role of melatonin as€a€proconvulsant versus an anticonvulsant is somewhat conflicting). Additionally, varying circadian sleep– wake cycle levels of neurotransmitters may play a role. A recent study reported signiἀcantly increased receptor sensitivity to the neuropeptide orexin (important for the maintenance of arousal and with minimal brain activity during sleep) in the epileptic hippocampus. The authors speculated on the potential epileptogenicity of orexin in medial temporal lobe seizures (Morales et al. 2008). Even if there is a circadian protective effect on the occurrence of nocturnal partial complex seizures, it seems that if a seizure does occur during nocturnal sleep, it tends to propagate more extensively and often secondarily generalizes.

336 Sudden Death in Epilepsy: Forensic and Clinical Issues

We previously alluded to the profound synchronizing effect of NREM sleep on cortical activity as measured with the EEG. Speciἀcally, NREM sleep consists of highly synchronous activity in the form of sleep spindles (12–14 Hz), and K-complexes (which represent the particularly synchronized activity of diffuse cortical regions in producing a transient 0.5- to 2-s biphasic potential) on a background of theta and delta activity. During slow wave sleep, the delta activity may intermittently appear be quite regular and rhythmic. However, it is well accepted that even in the absence of obvious synchronicity or rhythmicity to visual inspection, cortical activity of lower frequencies (theta, delta) requires signiἀcantly greater synchronization over signiἀcantly larger areas of cortex than faster frequencies (Buzsaki and Draguhn 2004). Niedermeyer (2008) and others have written about the potential of K-complexes in particular to induce interictal or even ictal activity as a result of its profound synchronizing effect. Sleep spindles, which represent highly synchronized activity in thalamocortical systems, are theorized to be “transformed” into generalized 3-per-second spike-wave complexes in models of absence epilepsy (Kostopoulos 2000). More recently, Steriade and colleagues (1993) have described a slow (<1 Hz) oscillation throughout NREM sleep, which becomes more rhythmic as sleep deepens and which consists of a rhythmic alternation of membrane potential (membrane “bistability”). A period of signiἀcant membrane depolarization, highly synchronized throughout much of cortex and involving both cortical pyramidal cells and interneurons, alternates with a period of cellular quiescence that they have referred to as a period of “disfacilitation,” consisting of a hyperpolarized membrane potential with increased membrane resistance and an absence of cortical synaptic activity. Although, like spindles, the slow oscillation is present diffusely in corticothalamic systems, the slow oscillation is believed to be intrinsically generated in cortex with coherence across large areas of cortex, as the slow oscillation occurs in deafferented cortical slabs. The highly synchronous and precipitous onset of the depolarized state from a state of diffuse cellular quiescence of the hyperpolarized state creates the K-complex of the surface EEG, and presents an unprecedented epileptogenetic potential (Steriade and Amzica 1998). Indeed, it has demonstrated in a model of seizures occurring in a thalamically deafferented region of cat neocortex during NREM sleep that spike-wave activity emerges with cellular evidence of a paroxysmal depolarzing shift directly from the activity of the slow oscillation (Steriade and Contreras 1998; Steriade et al. 1998). The generation and propagation of epileptiform activity has also been shown to be lowered in the hippocampus during the slow oscillation (Nazer and Dickson 2009). Additionally, the physiological ripple rhythm (150–250 Hz), which originates in the hippocampus during NREM sleep (Sirota et al. 2003; Buzsaki 1998) with involvement of neocortex as well during that sleep state (Grenier et al. 2001) and is postulated to be involved in memory consolidation, appears to be the forerunner of the pathological fast ripple (250–500 Hz) activity that is instrumental in the initiation of partial seizures (Foffani et al. 2007; Engel et al. 2009). Finally, Terzano and Parinno (2000) have emphasized the propensity for epileptiform activity to occur during the synchronized activity of the cyclic alternating pattern phase A, which may consist of serial K-complexes and bursts of “hypersynchronous” delta activity, among other patterns. This group describes a tremendous preponderance of interictal and ictal activity beginning during the cyclic alternating pattern phase A microarousals in a wide variety of seizure types (Parino et al. 2006). Finally, infraslow oscillations in the frequency range of 0.02 to 0.2 Hz, a frequency range not detected with standard alternating current electroencephalography instrumentation, have been described during NREM sleep. This slow cyclic modulation of cortical excitability appears to be associated

Sudden Death in Epilepsy

337

with the production of interictal epileptic activity (Vanhatalo et al. 2004). The physiological basis of infraslow oscillations has not been well characterized and whether infraslow oscillations and the cyclic alternating pattern are a manifestation of the same underlying physiological mechanism is unclear. With this discussion as a background, it is not surprising that interictal discharges are signiἀcantly more frequent in the relatively synchronized states of drowsiness and NREM sleep in most seizure types and partial complex seizures of medial temporal lobe origin are no exception (Clemens et al. 2005). Although temporal lobe seizures appear to have a late afternoon predominance, there is nonetheless an overwhelming preponderance of interictal discharges during NREM sleep as opposed to the waking state. Furthermore, interictal discharges propagate to a much wider territory of cortex during NREM sleep, often with contralateral involvement of discharges that are strictly unilateral during wakefulness (Nita et al. 2007). Also, as mentioned, seizures are much more likely to secondarily generalize during NREM sleep. It may be that the extensive interareal hypersynchrony present during NREM sleep is particularly conducive to the development of interictal discharges regardless of seizure type, to the development of certain types of generalized ictal activity, such as with 3-per-second spike wave activity, and to secondary generalization of partial seizures, but not to the development of some types of partial seizures. Such profound synchrony may not be as crucial for the interictal–ictal transformation that may take place at a much more focal level in temporal lobe epilepsy. REM sleep, with its desynchronized cortical activity, is felt to be relatively suppressive with regard to epileptiform activity. When interictal discharges do occur, they tend to be more focal with less propagation than during NREM sleep or even during wakefulness. Ictal transformation is much less frequent during REM than NREM sleep, though it can occur (Malow 2005).

23.5â•… Cardiovascular Physiological Activity during Sleep In general, NREM sleep is characterized by an increase in parasympathetic cardiac activity (vagal dominance) with little or no change in sympathetic activity (Parmeggiani and Morrison 1993). As a result, NREM sleep is characterized by a reduced heart rate and reduced blood pressure as compared with wakefulness. A decrease in the gain or responsiveness of the sinoatrial baroreflex, which normally counteracts an increase in blood pressure with a decrease in heart rate, is felt to account for the paradoxical simultaneous fall in blood pressure and heart rate. This resetting of the baroreflex is potentially relevant given the role of this reflex in buffering abrupt increases in blood pressure related to surges in sympathetic activity. There is little overall change in stroke volume or peripheral vascular resistance during NREM sleep; the reduced blood pressure is primarily related to the fall in heart rate. Additionally, NREM sleep can be characterized by transient increases in blood pressure and heart rate associated with the microarousals of the cyclic alternating pattern (Ferri et al. 2000; Ferini-Strambi et al. 2000). These changes are generally felt to be related to surges in sympathetic outflow related to overall heightened somatic and autonomic system activity with the arousal. Such surges rely on buffering by the baroreflex to limit their magnitude. However, the baroreflex, which has a decreased gain in NREM sleep, is felt to be rapidly reset or even “overwhelmed” by the arousal effect of the cyclic alternating pattern phase A (Murali et al. 2003; Iellamo et al. 2004). Interestingly, a recent report

338 Sudden Death in Epilepsy: Forensic and Clinical Issues

demonstrated decreased baroreflex function in chronic temporal lobe epilepsy patients (Dutsch et al. 2006). Tonic REM sleep is also characterized by a vagal, parasympathetic predominance, apparently associated with a mild decrease in sympathetic activity compared with NREM sleep, resulting in an even more profound vagal dominance. As a result, the heart rate and blood pressure fall below NREM sleep levels during tonic REM sleep. REM sleep generally is characterized by the lowest blood pressures during the 24-h circadian cycle (Coote 1982). Additionally, sinus pauses may accompany the bradycardia of REM sleep. Indeed, otherwise healthy adults may experience periods of asystole during REM sleep requiring the use of a pacemaker (Guilleminault et al. 1984). However, the phasic component of REM sleep is characterized by signiἀcant surges in sympathetic activity that may be accompanied by a decrease in parasympathetic activity. This results in signiἀcant transient increases in blood pressure and heart rate. Peripheral vascular resistance is overall relatively unchanged, with vasoconstriction in muscle vasculature offset by vasodilatation in renal and mesenteric vasculature. This autonomic pattern of relative lability or “instability” seems to be mediated predominantly by higher central nervous system structures, which tend to “override” local and medullary reflexes (Silvani and Lenzi 2005; Silvani 2008). This surge in sympathetic activity in the setting of tonic low parasympathetic activity predisposes REM sleep to the occurrence of ventricular ectopy (Garcia-Touchard et al. 2007). An additional aspect of autonomic function, heart rate variability, is of particular relevance to the topic of sudden death. Heart rate variability refers to the relative constancy or “variability” in the R–R interval from one QRS complex to the next in the cardiac cycle (De Jong and Randall 2005). Conventional (stochastic) heart rate variability analysis can be performed in the time or the frequency domain. Time-domain analysis basically determines the standard deviation of the R–R intervals in a period of time (or related measures); frequency analysis is performed with a fast Fourier transform to determine the range and power of frequency components of the R–R intervals. This analysis generally involves separate analyses of low-frequency and high-frequency components. The high-frequency band is felt to reflect predominantly parasympathetic influences, which are more rapid, including the vagally dominated baroreflex. The low-frequency band reflects both sympathetic and parasympathetic activity, perhaps with a modest sympathetic predominance. The high freâ•›quency/low frequency ratio therefore approximates the parasympathetic/sympathetic ratio of activities. A sympathetic predominance is felt to predispose to ventricular arrhythmias, which many authorities feel to be commonly implicated in sudden death. A signiἀcant parasympathetic predominance, however, can also predispose to bradyarrhythmias or sinus arrest, which can occur in REM sleep (Janssens et al. 2007). It has been suggested that “sympathovagal balance” is important for cardiac health. However, Skinner (1993) has emphasized that sympathovagal balance can also be pathological and can predispose to arrhythmias if the tone of both systems is high. He has described this scenario in NREM sleep in cats with an increase in ventricular premature contractions (although in general, the vagal predominance of NREM sleep has been felt to suppress premature ventricular contractions and ventricular arrhythmias). Skinner (1993) has also described heightened tone in both components of the autonomic nervous system in the setting of stress with minimal change in heart rate leading to an increase in premature ventricular contractions. It is now well known that a decrease in the various measures of heart rate variability are indicative of impaired cardiac health in congestive heart failure patients and are predictive

Sudden Death in Epilepsy

339

of negative outcomes. Decreased heart rate variability has also been shown to be predictive of poor cardiac outcomes in other patient populations as well, including those with several neurological disorders (Korpelainen et al. 1997). Furthermore, heart rate variability has been demonstrated to be signiἀcantly decreased in temporal lobe epilepsy patients (Ansakorpi et al. 2002), with this effect more pronounced nocturnally “reflecting a suppression of circadian heart rate variability dynamics” (Ronkainen et al. 2005; Persson et al. 2007). One study in children with partial seizures showed that heart rate variability is especially impaired at night (Ferri et al. 2002). Finally, a recent study showed that patients with frontal lobe epilepsy, in whom seizures occur almost exclusively at night, have a unique pattern of decreased heart rate variability, with lower parasympathetic drive (Harnod et al. 2009).

23.6â•… Heart Rate Variability and Fractal Physiology Recent years have seen an increasing interest in the application of techniques of fractal physiology to heart rate variability data (see Goldberger et al. 1990 or West and Goldberger 1987 for an introduction to fractal physiology). Fractal phenomena tend to demonstrate the phenomenon of self-similarity. Structures that demonstrate fractal qualities in the spatial domain tend to look similar no matter what the scale (i.e., the same spatial patterns occur across all spatial scales). They are said to be “scale-free.” Phenomena that change over time can also exhibit fractal qualities in the temporal domain. The patterns of change from interval to interval are independent of the time scale. Such systems tend to obey a power law in the temporal domain, such that a relationship between two variables (a and f╛╛) that change over time can be characterized by the equation a = Af╛╛α, wherein a is proportional to f raised to some particular power, α (power law). When α approaches a value of −1, so that the relationship approximates 1/f (inverse power law), the system is said to be governed by 1/f statistics (Shlesinger 1987). This 1/f activity is the deἀning feature of mathematically deἀned “complex systems,” which are systems made up of many interacting parts and the system as a whole is organized to optimize adaptation to changing environmental influences. Complex systems are not functioning at equilibrium and are not designed to maintain some speciἀc “set” point. In fact, such systems are said to operate “far from equilibrium” (Peng et al. 1994). The traditional notion that the ultimate goal of biological systems is the maintenance of homeostasis appears in need of some modiἀcation. In fact, true homeostasis is eschewed by healthy biological systems. Deviations from baseline function in a scale-invariant manner appear to be the sign of a healthy biological system (Goldberger et al. 2002a, 2002b; Lipsitz 2002). To casual visual inspection, temporal data characterized by scale-invariant 1/f dynamics generally looks like nothing but random “noise.” The data generated by such a system has been referred to as “pink noise” to differentiate it from truly random “white noise” and from Brownian motion or “brown noise.” Closer examination of such data is revealing. Whereas data comprising typical white noise exhibits continuous relatively small but variable fluctuations from some baseline level (preserving homeostasis in the traditional sense), brown noise demonstrates only large relatively persistent deviations from a baseline over time. Pink noise characterized by 1/f dynamics demonstrates both types of temporal fluctuations. Therefore, its time series is temporally scale-invariant. Fluctuations on a small time scale (seconds) are approximately mirrored by proportional fluctuations on longer time scales (hours or days). Heart rate data exhibit such 1/f temporal dynamics (Perkiomaki et al. 2005). A 10-s fluctuation in heart rate observed every few minutes might be magniἀed by a

340 Sudden Death in Epilepsy: Forensic and Clinical Issues

similar 10-min fluctuation every several hours (with the 10-s fluctuations superimposed). For a comprehensive review of 1/f dynamics in biological systems, see Gisiger (2001). Although not all of the variability of heart rate data can be attributed to fractal physiological mechanisms, clearly a substantial component of it can be. Indeed, heart rate variability in a healthy heart exhibits temporal self-similarity over a time scale of seconds to many hours (Saul et al. 1988). Additionally, the fractal component of heart rate variability appears to be mediated primarily if not exclusively by the central autonomic system as opposed to the intrinsic cardiac nervous system. Indeed, the suprachiasmatic nucleus has been implicated in regulation of fractal cardiac physiology; bilateral lesions of suprachiasmatic nucleus eliminate fractal heart rate regulation (Hu et al. 1988a, 1988b). While a transplanted heart exhibits intrinsic rhythmicity and even some minimal degree of nonfractal heart rate variability, true fractal heart rate variability does not develop for approximately 6 months, following adequate reinnervation with host autonomic ἀbers to impart an influence from the central autonomic system. Some disease processes with signiἀcant systemic cardiovascular pathology may lose their heart rate variability due to impaired effector function such as blood vessel wall compliance (Malpas 2002), but that would not apply to the fractal component of heart rate variability. What is the signiἀcance of fractal physiology? As we alluded to above, systems that exhibit 1/f dynamics seem especially suited to adapt to physiological challenges on many time scales. An optimized ability of the neurocardiac system to meet physiological challenges while maintaining the normal coordinated activity of its electrical conduction system seems to be conferred by fractal dynamics. Congestive heart failure patients and several other patient groups demonstrate a loss of fractal physiology in heart rate variability data that predicts poor cardiac outcomes including death secondary to cardiac arrhythmias. In fact, recent evidence has shown that impoverished fractal physiology is a far better predictor of future ventricular arrhythmias in patients with left ventricular dysfunction than analysis of traditional heart rate variability data (Makikallio et al. 2001). With regard to SUDEP, it has now been demonstrated that refractory temporal lobe epilepsy patients, a group of patients at particular risk of SUDEP in association with secondarily generalized seizures, not only demonstrate reduced heart rate variability by conventional measures, but a signiἀcantly reduced component of fractal heart rate variation dynamics. Additionally, the fractal component of heart rate variability demonstrates a clear circadian variation, such that there is a decrease in the magnitude of fractal dynamics of heart rate variability during NREM sleep, with preservation in REM sleep (Ivanov 2000; Bunde et al. 2000; Aoyagi et al. 2007; Togo and Yamamoto 2005). Interestingly, the “normal” departure from 1/f dynamics in NREM sleep deviates toward white noise whereas it deviates toward brown noise in waking patients with primary autonomic failure, a disorder, which like NREM sleep, is characterized by a parasympathetic predominance (Aoyagi et al. 2007).

23.7â•… A Proposed Pathophysiology of SUDEP We synthesize the information discussed so far in the chapter in the following way. The epileptic network in temporal lobe epilepsy patients, especially chronic refractory temporal lobe epilepsy patients at higher risk for SUDEP, induces changes in the central cortical autonomic function necessary to impart fractal dynamics on the heart, disturbing or

Sudden Death in Epilepsy

341

eliminating this fractal phenomenon. Preliminary evidence suggests this derangement is more pronounced at night, in the setting of a preexisting physiological reduction in fractal dynamics during NREM sleep. Fractal autonomic dynamics are felt to be necessary to proÂ� tect the heart in the setting of physiological challenges on various time scales (e.g., rapid intense surges as well as sustained increases in sympathetic activity). A seizure, especially a generalized tonic–clonic event, with its associated rapid increase in sympathetic activity, is one such challenge (most witnessed events of SUDEP have been associated with generalized tonic–clonic seizures). The sympathetic activity seen in the lockstep phenomenon, which refers to an increase in sympathetic activity time-locked to interictal discharges (Lathers et al. 1987; Stauffer et al. 1989, 1990; Dodd-O and Lathers 1990; O’Rourke€and Lathers 1990; Lathers and Schraeder 2010a, Chapter 28), is another. As we described, due to its multiphenomenal synchronizing affects, NREM sleep results in a signiἀcant increase in interâ•›ictal discharges in temporal lobe epilepsy (as well as most other seizure types) and produces€almost all secondarily generalized seizures in patients with temporal lobe epilepsy. Additionally, although partial complex seizures have a diurnal rather than nocturnal predominance, 60% of such seizures associated with electrocardiographic abnormalities actually occurred during NREM sleep (Nei et al. 2000). Of course, not all patients succumb to SUDEP at night and while asleep. We propose that the reduced or absent fractal physiÂ� ology of temporal lobe epilepsy patients will render them more vulnerable to an arrhythÂ� mia€throughout the circadian cycle, though with a signiἀcantly greater propensity during nocturnal sleep, with its inherent autonomic instability, and with even greater likelihood in the setting of the poorly “buffered” heart rate surge associated with a generalized tonic– clonic seizure. Carbamazepine discontinuation has been associated with SUDEP. A recent study demonstrated that sudden carbamazepine withdrawal leads to an increase in nocturnal sympathetic activity (Hennessy et al. 2001), which may be ineffectively buffered in SUDEP patients. Nei and colleagues (2004) reviewed heart rate data from seizure patients including a subset that ultimately succumbed to SUDEP. They observed greater blood pressure fluctuations during seizures occurring in nocturnal sleep than daytime seizures with signiἀcantly greater fluctuations in SUDEP versus non-SUDEP patients. Therefore, the neurocardiac system is less capable of buffering these rapid changes at night, with the SUDEP patients apparently even less adept at this buffering mechanism. This deranged physiology will also render such patients more susceptible to the development of a ventricular arrhythmia or even a serious bradyarrhythmia due to the more modest autonomic effects of a partial complex seizure (as opposed to a generalized tonic–clonic seizure) during the day or€simply due to various autonomic challenges that patients will face during their daily lives. Stress is one such challenge. Recently, the role of emotional stress in the etiology of SUDEP has been emphasized (Lathers and Schraeder 2006; 2010b, Chapter 17). Even mild stress has been noted to decrease fractal heart rate variability to some degree (Hoshikawa and Yamamoto 1997). Therefore, theoretically, stress could simultaneously increase sympathetic tone and decrease the fractal dynamics necessary to adapt to this change in tone and could have an additive effect on the preexisting heart rate variability pathophysiology in temporal lobe epilepsy patients. Furthermore, chronic stress and anxiety may increase the percentage of time that patients experience the cyclic alternating pattern during NREM sleep (Terzano and Parrino 2000). This alteration will increase the number of transient sympathetic challenges (to up to several hundred per night), or at another time scale, it simply increases their repetitive distribution throughout the night (persistence of recurrent

342 Sudden Death in Epilepsy: Forensic and Clinical Issues

challenges) or the average level of sympathetic activity throughout the night. Additionally, these arousals are much more likely to simultaneously activate an epileptiform discharge (Niedermeyer 2008) that may add its own deleterious autonomic effect via the lockstep or related phenomena. For the same reason, REM sleep, with its signiἀcant autonomic instability, would also pose a signiἀcant autonomic challenge to such patients, and, despite its relative epileptiform suppressing effect, could potentially be an especially arrhythmogenic period in patients with impaired fractal physiology. Interestingly, one recent study found that the interictal spike rate was signiἀcantly higher during REM sleep in patients with localized epilepsy who had a history of secondary generalization, a population at higher risk for SUDEP (Clemens et al. 2005).

23.8â•… Conclusion We regard the preceding discussion as a unifying theory for SUDEP occurrence in patients who have a deranged central autonomic system due to structural or functional changes in their epileptic networks or perhaps even due to nonepileptic central autonomic nervous system functional changes in association with more widespread cerebral pathology that gives rise to partial or symptomatic generalized seizures. The majority of SUDEP victims have structural brain pathology, usually nonprogressive, of some sort. Patients with idiopathic generalized epilepsies appear less likely to develop SUDEP. Of note, newly diagnosed epileptic patients were found not to demonstrate a difference in heart rate variability versus controls. This ἀnding suggests that chronic lesions, especially evolving plastic epileptic networks, experience a gradual reorganization in the central autonomic network over time with loss of complexity and the fractal dynamics it produces. Such a mechanism, with resultant reduced buffering of autonomic fluctuations especially in sleep, in combination with a dramatically increased quantity and ἀeld of distribution of interictal epileptiform discharges and secondarily generalized seizures in medial temporal lobe epilepsy patients due to the profound synchronizing effects of NREM sleep, could account for or certainly contribute to the nocturnal sleep state predominance of SUDEP cases. Current research is attempting to address mechanisms to induce or restore nonlinear fractal dynamics in biological systems lacking such characteristics in order to increase adaptability and resilience in the face of environmental challenges (Schiff et al. 1994; In et al. 1995). Such research may prove critical to the development of future strategies to prevent sudden death in the setting of epilepsy.

References Ansakorpi, H., J. T. Korpelainen, and H. V. Huikuri et al. 2002. Heart rate dynamics in refractory and well controlled temporal lobe epilepsy. J Neurol Neurosurg Psychiatr 72: 26–30. Aoyagi, N., Z. R. Struzik, and K. Kiyaon et al. 2007. Autonomic imbalance and breakdown of longrange dependence in healthy heart rate. Methods Inf Med 46: 174–178. Bunde, A., S. Havlin, and J. W. Kantelhardt et al. 2000. Correlated and uncorrelated regions in heartrate fluctuations during sleep. Phys Rev Lett 85: 3736–3739. Buzsaki, G. 1998. Memory consolidation during sleep: A neurophysiological perspective. J Sleep Res 7 (Suppl. 1): 17–23. Buzsaki, G., and A. Draguhn. 2004. Neuronal oscillations in cortical networks. Science 304 (5679): 1926–1929.

Sudden Death in Epilepsy

343

Clemens, Z., J. Jansky, and B. Clemens et al. 2005. Factors affecting spiking related to sleep and wake states in temporal lobe epilepsy. Seizure 14: 52–57. Coote, J. H. 1982. Respiratory and circulatory control during sleep. J Exp Bio1 100: 223–244. De Jong, M. J., and D. C. Randall. 2005. Heart rate variability analysis in the assessment of autonomic function in heart failure. J Cardiovasc Nurs 20 (3): 186–195. Dodd-O, J. M., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in phenobarbitol-pretreated cats. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 199–220. New York, NY: Marcel Dekker. Durazzo, T. S., S. S. Spencer, and R. B. Duckrow et al. 2008. Temporal distributions of seizure occurrence from various epileptogenic regions. Neurology 70: 1265–1271. Dutsch, M., M. J. Hilz, and O. Devinsky. 2006. Impaired baroreflex function in temporal lobe epilepsy. J Neurol 253: 1300–1308. Engel Jr., J., A. Bragin, and R. Staba et al. 2009. High-frequency oscillations: What is normal and what is not? Epilepsia 50 (4): 598–604. Ferini-Strambi, L., A. Bianchi, and M. Zucconi et al. 2000. The impact of the cyclic alternating pattern on heart rate variability during sleep in healthy adults. Clin Neurophysiol 111: 99–101. Ferri, R., L. Parrino, and A. Smerieri et al. 2000. Cyclic alternating pattern and spectral analysis of heart rate variability during normal sleep. J Sleep Res 9: 13–18. Ferri, R., L. Curzi-Dascalova, and A. Arzimanoglou et al. 2002. Heart rate variability during sleep in children with partial epilepsy. J Sleep Res 11: 153–160. Foldvary-Schaefer, N., and M. Grigg-Damberger. 2006. Sleep and Epilepsy: What we know, don’t know and need to know. J Clin Neurophysiol 23 (1): 4–20. Foffani, G., Y. G. Uzcategui, and B. Gal et al. 2007. Reduced spike-timing reliability correlates with the emergence of fast ripples in the rat epileptic hippocampus. Neuron 55 (6): 930–941. Garcia-Touchard, A.,V. K. Somers, and T. Kara. 2007. Ventricular ectopy during REM€sleep:€ImplicaÂ� tions for nocturnal sudden cardiac death. Nat Clin Pract Card 4 (5): 284–288. Gisiger, T. 2001. Scale invariance in biology: Coincidence or footprint of a universal mechanism. Biol Rev 76: 161–209. Goldberger, A. L., D. R. Rigney, and B. J. West. 1990. Chaos and fractals in human physiology. Sci Am. February: 43–49. Goldberger, A. L., L. A. N. Amaral, and J. M. Hausdorff et al. 2002a. Fractal dynamics in physiology: Alterations with disease and aging. Proc Natl Acad Sci U S A 99 (Suppl. 1): 2466–2472. Goldberger, A. L., C. K. Peng, and L. A. Lipsitz. 2002b. What is physiologic complexity and how does it change with aging and disease? Neurobiol Aging 23: 23–26. Grenier, F., I. Timofeev, and M. Steriade. 2001. Focal synchronization of ripples (80–200 Hz) in neocortex and their neuronal correlates. J Neurophysiol 86 (4): 1884–1898. Guilleminault, C., P. Pool, and J. Motta et al. 1984. Sinus arrest during REM sleep in young adults. New Engl J Med 311: 1006–1010. Halasz, P., M. Terzano, and L. Parrino et al. 2004. The nature of arousal in sleep. J Sleep Res 13: 1–23. Harnod, T., C. C. H. Yang, and Y. L. Hsin et al. 2009. Heart rate variability in patients with frontal lobe epilepsy. Seizure 18: 21–25. Hennessy, M. J., G. M. Tighe, and C. D. Binnie et al. 2001. Sudden withdrawal of carbamazepine increases sudden cardiac sympathetic activity in sleep. Neurology 57: 1650–1654. Herman, S. T., T. S. Walczak, and C. W. Bazil. 2001. Distribution of partial seizures during the sleep– wake cycle: Differences by seizure onset site. Neurology 56 (11): 1453–1459. Hoshikawa, Y., and Y. Yamamoto. 1997. Effects of Stroop color–word conflict on autonomic motor system responses. Am J Physiol–Heart C 272: H1113–H1121. Hu, K., F. A J. L. Scheer, and R. M. Buijs et al. 2008a. The circadian pacemaker generates similar circadian rhythms in the fractal structure of heart rate in humans and rats. Cardiovasc Res 80: 62–68. Hu, K., F. A. J. L. Scheer, and R. M. Buijs et al. 2008b. The endogenous circadian pacemaker imparts a scale-invariant pattern of heart of heart rate fluctuations across time scales spanning minutes to 24 hours. J Biol Rhythm 23: 265–273.

344 Sudden Death in Epilepsy: Forensic and Clinical Issues Janssens, W., R. Willems, and D. Pevernaqie et al. 2007. REM sleep-related brady-arrhythmia syndrome. Sleep Breath 11 (3): 195–199. Iellamo, F., F. Placidi, and M. G. Marciani et al. 2004. Baroreflex buffering of sympathetic activation during sleep: Evidence from autonomic assessment of sleep macroarchitecture and microarchitecture. Hypertension 43: 814–819. In, V. V., S. E. Mahan, and W. L. Ditto et al. 1995. Experimental maintenance of chaos. Phys Rev Lett 74 (22): 4420–4423. Ivanov, P. C. 2007. Scale-invariant aspects of cardiac dynamics: Observing sleep stages and circadian phases. IEEE Eng Med Biol Mag 26 (6): 33–37. Janz, D. 1962. The grand mal epilepsies and the sleeping–waking cycle. Epilepsia 3: 69–109. Jobst, B. C., P. D. Williamson, and T. B. Neuschwander et al. 2001. Secondarily generalized seizures in mesial temporal epilepsy: Clinical characteristics, lateralizing signs and association with the sleep–wake cycle. Epilepsia 42 (1): 1279–1287. Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg 67: 439–444. Korpelainen, J. T., K. A. Sotaniemi, and H. V. Huikuri et al. 1997. Circadian rhythm of heart rate variability is reversibly abolished in ischemic stroke. Stroke 28: 2150–2154. Kostopoulos, G. K. 2000. Spike-and-wave discharges of absence seizures as a transformation of sleep spindles: The continuing development of a hypothesis. Clin Neurophysiol 111: (Suppl. 2): S27–S38. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case control study of SUDEP. Neurology 64: 1131–1133. Lathers, C. M. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., and P. L. Schraeder. 2010a. Animal model for sudden unexpected death in persons with epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 28. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 2010b. Stress and SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 17. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and F. W. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lipsitz, L. A. 2002. Dynamics of stability: The physiological basis of functional health and frailty. J€Gerontol 57A: B115–B125. Makikallio, T. H., H. V. Huikuri, and A. Makikallio et al. 2001. Prediction of sudden cardiac death by fractal analysis of heart rate variability in elderly subjects. J Am Coll Cardiol 37: 1395–1402. Malow, B. A. 2005. Sleep and epilepsy. Neurol Clin 23 (4): 1127–1147. Malpas, S. C. 2002. Neural influences on cardiovascular variability: possibilities and pitfalls. Am J Physiol–Heart C 282: H6–H20. Morales, A., C. Bonnet, and N. Bourgoin. 2006. Unexpected expression of orexin-B in basal conditions and increased levels in the rat hippocampus during pilocarpine-induced epileptogenesis. Brain Res 1109: 164–175. Moore, R. Y. 2007. Suprachiasmatic nucleus in sleep–wake regulation. Sleep Med 8: S27–S33. Mraovitch, S., and Y. Calando. 1999. Interactions between limbic, thalamo-striatal-cortical and central autonomic pathways during epileptic seizure progression. J Comp Neurol 411: 145–161. Murali, N. S., A. Svatikova, and V. K. Somers. 2003. Cardiovascular physiology and sleep. Front Biosci 8: s636–s652. Nazer, F., and F. N. Dickson. 2009. The slow oscillation facilitates epileptiform events in the hippocampus. J Neurophysiol 102 (3): 1880–1889. Nei, M., R. T. Ho, and B. W. Abou-Khalil. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345.

Sudden Death in Epilepsy

345

Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Niedermeyer, E. 2008. Epileptiform K complexes. Am J Electroneurodiagnostic Technol 48 (1): 48–51. Nita, D. A., Y. Cisse, and I. Timofeev et al. 2007. Waking-sleep modulation of paroxysmal activities induced by partial cortical deafferentation. Cerebr Cortex 17: 272–283. O’Rourke, D. K., and C. M. Lathers. 1990. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocortogram of the cat. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 239–260. New York, NY: Marcel Dekker. Parino, L., P. Halasz, and C. A. Tassinari et al. 2006. CAP, epilepsy and motor events during sleep: The unifying role of arousal. Sleep Med Rev 10: 267–285. Parmeggiani, P. L., and A. R. Morrison. 1990. Alterations in autonomic function during sleep. In Central Regulation of Autonomic Functions, ed. A. D. Lowey and K. M. Spyer, 367–390. New York, NY: Oxford University Press. Peng, C. K., S. V. Buldyrev, and J. M. Hausdorff et al. 1994. Non-equilibrium dynamics as an indispensable characteristic of a healthy biological system. Integr Phys Behav Sci 29: 283–293. Perkiomaki, J. S., T. H. Makikallio, and H. V. Huikuri et al. 2005. Fractal and complexity measures of heart rate variability. Clin Exp Hypertens 2–3: 149–158. Persson, H., E. Kumlien, and M. Ericson et al. 2007. Circadian variation in heart-rate variability in localization-related epilepsy. Epilepsia 48 (5): 917–922. Quigg, M. 2000. Circadian rhythms: Interactions with seizures and epilepsy. Epilepsy Res 42: 43–55. Ronkainen, E., H. Ansakorpi, and H. V. Huikuri et al. 2005. Suppressed circadian heart rate dynamics in temporal lobe epilepsy. J Neuro Neurosurg Psychiatr 76: 1382–1386. Saul, J. P., P. Albrecht, and R. D. Berger et al. 1988. Analysis of long term heart rate variability: Methods, 1/f scaling and implications. Comput Cardiol 14: 419–422. Schiff, S. L., K. Jerger, and D. H. Duong et al. 1994. Controlling chaos in the brain. Nature 370 (6491): 615–620. Shlesinger, M. F. 1987. Fractal time and 1/f noise in complex systems. Ann N Y Acad Sci 504: 214–228. Sirota, A., J. Csicsvari, and D. Buhl et al. 2003. Communication between neocortex and hippocampus during sleep in rodents. Proc Natl Acad Sci U S A 100 (4): 2065–2069. Skinner, J. E. 1993. Neurocardiology: Brain mechanisms underlying fatal cardiac arrhythmias. Neurol Clinics 11 (2): 325–351. Silvani, A. 2008. Physiological sleep-dependent changes in arterial blood pressure: Central autonomic commands and baroreflex control. Clin Exp Pharmacol Physiol 35: 987–994. Silvani, A., and P. Lenzi P. 2005. Reflex cardiovascular control in sleep. In The Physiologic Nature of Sleep, ed. P. L. Parmeggiani and R. A. Velluti, 323–350. London: Imperial College Press. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroenphalogr Clin Neurophysiol 72 (4): 340–345. Stauffer, A. Z., J. M. Dodd-O, and C. M. Lathers. 1990. Relationship of the lockstep phenomenon and precipitous changes in blood pressure. In Epilepsy and Sudden Death, ed. C. M. Lathers and P.€L. Schraeder, 221–238. New York, NY: Marcel Dekker. Steriade, M., and F. Amzica. 1998. Slow sleep oscillation, rhythmic K-complexes, and their paroxysmal developments. J Sleep Res 7 (Suppl. 1): 30–35. Steriade, M., and D. Contreras. 1998. Spike-wave complexes and fast components of cortically generated seizures. I. Role of the neocortex and thalamus. J Neurophysiol 80 (3): 1439–1455. Steriade, M., A. Nunez, and F. Amzica. 1993. A novel slow (<1 Hz) oscillation of neocortical neurons in vivo: Depolarizing and hyperpolarizing components. J Neurosci 13 (8): 3252–3265. Steriade, M., F. Amzica, and D. Neckelman et al. 1998. Spike-wave complexes and fast components of cortically generated seizures: II. Extra- and intracellular patterns. J Neurophysiol 80 (3): 1456–1479.

346 Sudden Death in Epilepsy: Forensic and Clinical Issues Terzano, M. G., and L. Parrino. 2000. Origin and signiἀcance of the cyclic alternating pattern (CAP). Sleep Med Rev 4 (1): 101–123. Togo, F., and Y. Yamamoto. 2000. Decreased fractal component of human heart rate variability during non-REM sleep. Am J Physiol Heart C 280: H17–H21. Vanhatalo, S., J. M. Palva, M. D. Holmes et al. 2004. Infraslow oscillations modulate excitability and interictal epileptic activity in the human cortex during sleep. Proc Natl Acad Sci U S A 101 (14): 5053–5057. West, B. J., and A. L. Goldberger. 1987. Physiology in fractal dimensions. Am Sci 75: 354–364.

SUDEP Medicolegal and Clinical Experiences

24

Braxton B. Wannamaker Contents 24.1 Childhood SUDEP 24.2 Medicolegal Experiences 24.3 Verbal Autopsy References

347 349 355 357

The treatment of persons with epilepsy is a most interesting, challenging, and rewarding experience in neurology. Our newest patients are often young and vibrant. Our oldest patients are most curious. In this most appreciative population the saddest moment in our practices is the loss of a life suddenly and unexpectedly. More often than not, sudden death occurs in the younger and less well-controlled patient. The boundaries of risk, however, are not limited by age and the phenomena of sudden unexpected death in epilepsy (SUDEP) may include children (Neuspiel and Kuller 1985; Kurtz et al. 1998; Donner et al. 2001), poorly controlled mostly but not always, compliant and noncompliant, patients who are depressed, or have elevated moods, poor or wealthy, any race, highly educated or mentally retarded. The recognition and application of risk factors do not provide security for the patient or physician. The best minds have considered drug effects, autonomic dysfunction, neuropathology, cardiovascular pathophysiology, psychological and genetic factors (Lathers and Schraeder 2009; So 2006) and yet the etiology and mechanisms for SUDEP are not established and elude all.

24.1â•… Childhood SUDEP That SUDEP occurs in children is evident in review of older literature (Wannamaker 1990) and in more recent literature (Leestma et al. 1997; Earnest et al. 1992; Harvey et al. 1993; Nashef et al. 1995). There are some features which indicate that adult SUDEP and childhood SUDEP are possibly two separate phenomena distinguished by factors other than age. Incidence data for SUDEP in children and in adults are quite dissimilar. SUDEP in children is mostly observed only in those with underlying neurological deἀcits. Children with nonsymptomatic or idiopathic (primary) epilepsies are generally spared from SUDEP. While SUDEP has become better recognized in the neurological community as a cause of death, SUDEP remains a large and poorly understood black box. A review of epidemiological information (Tomson et al. 2008) reveals incidence estimates ranging from about 1/1000 patient-years (community-based) to 9/1000 patient-years (presurgical cohort). These data are highly dependent on the population studied. Available crude SUDEP incidence rates limited to cross-sectional studies of children are 0.2/1000 patient-years (Donner et al. 2001) and 0.43/1000 patient-years (Weber et al. 2005). In a 347

348 Sudden Death in Epilepsy: Forensic and Clinical Issues

cohort of disabled children and young adults (Nashef et al. 1995) the incidence rate was 3.4/1000 patient-years. In one-half of the SUDEP cases the age at death was greater than 19 years. A prospective study (Walczak et al. 2001) of 4578 patients enrolled from three large epilepsy centers yielded a SUDEP incidence of 1.21/1000 patient-years. In this study SUDEP (n = 17) accounted for 18% of all deaths (n = 111). SUDEP did not occur in persons less than 20 years of age. In two prospective studies of children with epilepsy SUDEP was not observed (Callenbach et al. 2001) in one study and occurred in only one patient (Camἀeld et al. 2002) and that death occurred at age 21. Thus, there appears to be an overall lower incidence of SUDEP in children compared to adults, which is approximately 5- to 10-fold depending on populations. Information from prevalent cases of deaths in childhood (coroner’s records, death certiἀcates, etc.) tells us about those children whose onset of epilepsy was in childhood and who died in childhood (Harvey et al. 1993; Donner et al. 2001). Prospective studies may tell us if epilepsy began in childhood and if death occurred in childhood (Camἀeld et al. 2002; Callenbach et al. 2001) or adulthood (Walczak et al. 2001). Personal experience (unpublished data) with 15 SUDEP cases included no deaths in children. Yet, in 13 of 15 (87%) patients epilepsy began in childhood. Studies of deaths including SUDEP in children with epilepsy reveal important differences from adults. Most deaths occur in those children who also have neurological deἀcits (Harvey et al. 1993; Camἀeld et al. 2002; Callenbach et al. 2001). Symptomatic epilepsies were identiἀed in 74% of SUDEP cases in one series (Donner et al. 2001), 82% (Harvey et€ al. 1993) and 100% (Weber et al. 2005) in other series. Furthermore, chronic neuropathological changes were found at autopsy in 52% of SUDEP cases (Donner et al. 2001). In adult SUDEP cases Walzcak et al. (2001) found that 29% had epileptogenic structural lesions that were not signiἀcantly different from controls at 24%. Those children who die of SUDEP may be sufficiently different in that they are at risk for reasons distinctly different than the adult who dies of SUDEP. It is also reasonable to consider that the mechanism of SUDEP in the child who likely has a disordered central nervous system is not the same mechanism for SUDEP in the adult with a normally developed brain. Power spectral analysis of heart rate variability (HRV) demonstrated that the autonomic heart ἀndings in adults (Tomson et al. 1998) also could be found in children with epilepsy (Yang et al. 2001). This same group (Harnod et al. 2008) subsequently showed lower HRV resulting from parasympathetic or vagal reduction. Unlike the adult with sympathetic up-regulation (Evrengul et al. 2005), their changes suggest a potentially different mechanism of SUDEP in children. Additionally, El-Sayed et al. (2007) have studied additional measures (reflex tests, echocardiography, time domain analysis variables, urinary catecholamine metabolites) of interictal cardioregulatory mechanisms in children (mean age 10.4 years, range 6–18 years) with partial and generalized epilepsies. Mild to severe overall autonomic nervous system dysfunction was found in 54% of subjects. All patients with uncontrolled epilepsy had varying degrees of autonomic dysfunction (neuropathy). Thus for the pediatric population, SUDEP is much less frequent than for adults. SUDEP occurs only rarely in pediatric epilepsy patients who are otherwise normal and healthy. Are these pediatric cases different? Do risk factors transcend age groups? What are the relevant risk factors during childhood and when are those factors valid or predictive? Is childhood protective against SUDEP? Do mechanisms for SUDEP have any speciἀc determinates based on age at onset, age at death, type of epilepsy, or underlying neurological deἀcits? Hopefully, further studies will give answers to these questions.

SUDEP

349

24.2â•… Medicolegal Experiences Legal issues beset all physicians in clinical practice. Perper (1980) delineated the most common medicolegal problems in epilepsy and included as important medical responsibilities the determination of cause and manner of death. Notably, this forensic pathologist (Perper 1980) reviewed the difficulties encountered in the recognition and examination of cases when death was “sudden and unexpected.” Sudden and unexpected death is now appreciated by neurologists as a common occurrence in patients with epilepsy. However, this phenomenon is relatively uncommon (and unrecognized) among total deaths. That it occurs unexpectedly, is associated with a fatal outcome, has no clearly identiἀed cause, and above all is devastating to family members may create a fertile substrate for anger, depression, accusation, condemnation, and frustration. The legal system may be called into play in these circumstances. Thus, it is relevant to present several actual case histories of SUDEP that were associated with medicolegal importance. The author provided either consultation for defense attorneys or served as an expert witness in these cases. Case 1â•… Co nsent fo r SUDEP A 28-year-old mother and housewife returned to the university to complete her degree. She had been treated for generalized epilepsy and for a number of years had taken various medications. Even when well controlled at times she would have neurocognitive adverse events with relatively low doses of medication. She was followed carefully and regularly by her university physician but compliance was an issue at times. Most recently she was on lamotrigine with excellent seizure control. She independently decided to withdraw her medication because of concern for drowsiness or cognitive adverse events that might interfere with her academic career. She admitted to her epileptologist that she was no longer on AEDs. Shortly thereafter she died. Death was attributed to SUDEP. Her long-time university attending physician was sued by her husband. The lawyer for the plaintiff alleged that the physician had failed to advise his patient of the consequences of her being off medication. There are two primary issues raised in this case. How do we best document our consent to treat? Are all adverse outcomes in that consent to be discussed including SUDEP? Other important issues for SUDEP are whether to discuss, the timing of discussion, and the mode of presentation. The treating physician in this instance was a nationally recognized epilepsy specialist in a university clinic and he contended that within the continuum of his care he had given this patient essentially all information related to treatment and the consequences of treatment including advice about SUDEP. However, this was not speciἀcally delineated in the otherwise impeccable medical records. Written documentation would have gone a long way in avoiding extensive inquiries or preventing formal legal claims. It would seem that in this setting (university clinic) and with this speciἀc patient population, which was difficult to control and intermittently noncompliant, essentially all of this physician’s patients and families should be advised about SUDEP as well as the many other causes of death, accidents, and injuries. This should occur early in their relationship as the patients and families in these tertiary clinics are not new to epilepsy and its consequences and the patients are likely at higher risk for SUDEP. The information about

350 Sudden Death in Epilepsy: Forensic and Clinical Issues

SUDEP should be given in a straightforward manner and include risk factors speciἀc for that patient. In practices of typically lower risk patients, information should be available, given to a patient, and discussed at their request. As best as can be ascertained, the relative risks for each speciἀc patient would provide the most relevant information and perhaps be reassuring in most instances. The issue of requirement to discuss SUDEP with patients remains controversial. In the United Kingdom, guidelines (National Health Service 2004) have been recommended that patients with epilepsy and their families and/or caregivers should be given information and have access to information on SUDEP. A survey of neurologists in the United Kingdom (Morton et al. 2006) has shown that only 5% discussed SUDEP with all patients, 26% with a majority, and 61% with a few. There appeared to be a better rate of compliance (50%) with epilepsy nurse specialists who discussed SUDEP with most of their patients (Lewis et al. 2008). As an implied requirement to disclose SUDEP to all patients poses an issue of standard of care, Beran et al. (2004) conducted a cohort controlled comparative crossmatched study in an outpatient epilepsy clinic. Patients with similar risks to their SUDEP victims were identiἀed. There was no risk factor in the patients that could be modiἀed so as to alter an outcome of SUDEP. Thus the physician could not be held liable for negligence based on failure to disclose information about SUDEP should it occur. These investigators also pose the converse that in providing information about SUDEP in the absence of a patient request (“right not to know”) could adversely affect the patient’s quality of life (Beran 2006). It is clearly held by this group that the duty of care requires clear and full discussion should the patient seek the information. My experiences and patient population of more difficult patients lead me to advise patients on as many issues about epilepsy that I possibly can. It is an ongoing process. Timing and patient status will moderate the process. I have no indication that I have impaired quality of life by providing unsolicited pertinent information. It is my contention that the patients who most actively participate in their treatment are those who are the best informed (about their condition). This position assumes that I will at times infringe on their “right not to know.” The case discussed above went to mediation as is now the trend (requirement) in the judicial system. One of the mediators stated that any case resulting in death should include some reward for the plaintiff. Despite the prejudicial view of this one mediator the defendant physician was dismissed without trial after remaining steadfast in his position. No award was made to the plaintiff and the case was not pursued further.

Case 2â•… Cause o f SUDEP A 27-year-old woman had complex partial seizures and secondarily generalized tonic–Â�clonic seizures beginning in her early twenties. Seizures occurred infrequently yet at times resulted in social embarrassment or job-related issues. As well, there were some adverse reactions to medication in the past. Her primary care physician (an internist) provided most of her care with intermittent neurological consultations. Phenytoin had resulted in gingival hyperplasia, which was unacceptable to her. After 2 years of no seizures, an attempt to taper off phenytoin was undertaken. Two generalized tonic–clonic seizures occurred. Her medical regimen was eventually modiἀed

SUDEP

351

to carbamazepine. Later she did well with low doses. She remained self-sufficient and commanded reasonable employment and pay. Subsequently, she moved to another large city and married. She was initially referred to a highly recommended neurologist who continued her therapy without change. With carbamazepine at 50 mg in the morning and 100 mg at bedtime her random level was about 3.3 mg/dL carbaÂ� mazepine. At this point she was 9 years seizure-free. After some alleged inappropriate sexually oriented comments by the physician she was transferred to a capable private practice neurologist. This neurologist saw her only once and he complied with her request to taper off the carbamazepine as she wanted to become pregnant. Discussion about antiepileptic drugs and pregnancy had occurred with all of her neurologists including the most recent one. It was decided that slow taper and discontinuation of carbamazepine was appropriate. Both patient and physician agreed. SUDEP was not speciἀcally cited. The dose of 50 mg/day was reduced on a weekly basis. The patient awoke in the morning after her ἀnal dose of medicine was taken at bedtime, and saw her husband off to work. He called home at about 11:00 a .m. and there was no answer. Fearing a seizure, he called in for emergency medical services. His wife was found by emergency medical technicians across her bed unconscious, pulseless, and mildly cyanotic. Resuscitation efforts were made and a heart rhythm resumed. Although admitted to the hospital in coma and supported intensively, she survived only 20 h before life support was withdrawn. Autopsy demonstrated a tongue bite and no other evidence for cause of death. The pathologist concluded that she had had a seizure and that SUDEP was the cause of death. The distraught husband sued the physician citing that he was responsible for the cause of death, SUDEP. The plaintiff’s attorney brought in an expert witness who claimed that the cause of death was SUDEP. The medical examiner had extensive experience with death in patients with epilepsy. She reaffirmed that the cause of death was SUDEP as she had indicated on the death certiἀcate. The defense attorney then brought forth expert neurological witnesses who also agreed that SUDEP was the cause of death. Furthermore, the defense attorney had the medical examiner and expert defense neurologist also testify that the cause of SUDEP was unknown. The defense then argued that the defending physician could not be incriminated in any role in causing SUDEP since no one knows the cause of SUDEP. Hence, the jury found for the defendant. The legal approach to this case was formed in logic and based on principles of common law. Obfuscating views about putative mechanisms, risk factors, and the cause of SUDEP were removed from discussion. Clearly, in the eyes of the defense attorney there is no known cause for SUDEP. This logic apparently satisἀed the jury. The additional issues in this case focused on whether consent for drug withdrawal included information about SUDEP and what was the prognosis for recurrence of seizures after drug withdrawal. At the time of trial, the husband admitted that he knew of the plan of action taken around drug withdrawal. He did not attend the medical appointment with the victim and was less clear about his understanding of death being a consequence of a seizure. There was no speciἀc written documentation in the medical notes. Based on studies at the time of this case (Callaghan et al. 1988; Medical Research Council Antiepileptic Drug Withdrawal Study Group 1991; Specchio et al. 2002), the decision for

352 Sudden Death in Epilepsy: Forensic and Clinical Issues

drug withdrawal was not unreasonable although the history of one prior failure at withdrawal 9 years earlier is somewhat disconcerting. Case 3â•… Murder by S UDEP A 39-year-old housewife (JW) living in a base housing facility of a U.S. military compound in South Carolina was assaulted in 1973. Her husband had left home in the early morning. Later that morning their 4-year-old son found his mother unconscious and bleeding from the head. A hammer with bloodstains was found in an adjoining room. While there were no eyewitnesses, six sets of ἀngerprints were found in the home. The mother survived the attack and lived another 17 years with dysphasia and epileptic seizures. About 9 months after the assault she was communicating by gestures and was understood through her husband’s interpretations. In this manner she was able to describe the assault to agents of the Federal Bureau of Investigation (FBI). One of two young boys had entered the home after her husband left. She encountered the older boy leaving the bedroom (where the bloody hammer had been found). The FBI later determined that 22 boys had been absent from or late for the local school on this day of the attack. The U.S. Attorney’s Office refused to allow ἀngerprinting of all the truant children. The investigation was closed. Nine years later, a prisoner (M. Chase) who was incarcerated in Oklahoma on an armed robbery conviction contacted the FBI and confessed to assaulting a woman in a military housing area near Beaufort, SC. His apparent motivation for the confession was to bargain for a transfer out of the Oklahoma prison. He was serving his sentence under protective custody because he had provided information that led to a shakedown of that prison. He feared for his life. The confession, including the details, matched the known facts of the assault on JW. Yet, the United States declined to prosecute the matter even after Chase offered to waive any expired statute of limitations for assault. Sixteen years after the 1973 assault and 7½ years after the confession by Chase, JW€was found dead in her Florida home on January 17, 1990. The medical examiner considered her medical history and concluded that she had died of an epileptic seizure. On April 10, 1991, Chase was indicted for ἀrst-degree murder. He was tried in U.S. District Court (4th Circuit) and convicted by a jury in January 1992 (United States v. Chase 1994). The prosecuting attorney felt strongly that his expert witness, Dr. Jan Leestma, clearly educated the jury about SUDEP and its importance in this case. The assault led to brain injury and consequently to epilepsy. The epilepsy caused her death by SUDEP. Chase was sentenced to life imprisonment. Then a series of legal twists ensued during the appeals process. The district court held that sentencing guidelines were inappropriate because the crime was committed on the day on which the assault occurred, not the date of death; the attack occurred before the sentencing guidelines were in effect. Therefore, Chase appealed his conviction. The attorneys for Chase contended that (1) his indictment and trial were barred by the “year and a day” rule, (2)€by the 5-year federal statute of limitations for assault (yet no limitations for offenses punishable by death), and (3) that his Fifth Amendment right of due process of law was

SUDEP

353

violated by the government’s delay in bringing him to trial, and when the trial court committed reversible error by not granting him a new trial when it was discovered that jurors considered information coming into possession from outside the trial proceeding. The district court considered the three arguments; only the ἀrst argument was sufficient to bar his prosecution. The “year and a day” rule has been recognized by the United States Supreme Court (1891) and is applicable to federal processing for murder. The district court thereby applied it to this case and reversed the conviction of Chase. The examining pathologist concluded that the victim had died from an epileptic seizure. SUDEP was suggested by the knowledgeable expert witness, thereby tying the epilepsy and death to the assault and brain injury although the victim had been dead for some time (possibly 2 days) before being found. It is logical from the pathophysiological facts that death could be attributed to the brain injury which then led to epilepsy and death in the course of that illness. The protracted continuum of this medical problem does not appear to ἀt into a valid legal framework. The legal system has argued the relevance of the “year and a day” rule in other circumstances and it has not been overturned in more than 100 years. The sometimes long period required before brain injury results in epilepsy could be problematic for other situations if this rule applies. SUDEP linked to the epilepsy from traumatic brain injury due to vehicular accidents or due to military activity pose intriguing issues. Had this victim died of SUDEP within the year of the assault, the outcome for the defendant would have been most likely quite different. Case 4â•… Freedo m by S UDEP This case dates to 1997 and centers on a small rural community of about 2000 people at the base of the Appalachian Mountains. One of its lifelong residents, AR, a part-time television repairman, had created a presence of paranoid and litigious behavior. He had sued various citizens for apparently minor disagreements. He seldom interacted with the community and lived secluded in his dilapidated home. There was no telephone and no running water in the house. The yard was unkempt and strewn with old appliances and vehicles. Following a lawsuit and countersuit with the sheriff, AR had to give up his old van for restitution. In doing so, he could no longer earn a living. AR had married 30 years earlier a pretty, petite woman (Virginia), who was then 18 years old. He was 10 years her senior. She also remained essentially a total recluse. Her family reported that she was seldom seen since the marriage. The family contended that she was being restrained against her will by her husband. Citizens postulated that she was locked in the basement (which did not exist) or that she had long since died. The couple had no children. In 1997, AR reported to the 911 operator that his wife would not awaken. Strangely, he had driven out of town and then back to ἀnd a pay telephone. He passed a ἀre station but did not stop as he had had a prior dispute with them. The pay telephone that he found was immediately adjacent to the 911 call center. The 911 dispatcher later indicated that the caller had been calm and dispassionate in reporting the possible death of his wife.

354 Sudden Death in Epilepsy: Forensic and Clinical Issues

When the authorities arrived at the home, they found Virginia was dead in bed. The newly posted coroner reportedly found an emaciated, disheveled, and partially clothed woman in bed. There were marks on the left wrist suggesting a restraint and scatted facial periorbital petechiae. It was concluded that she had been deceased for about 8 h, as rigor mortis had begun. Blood for toxicology studies was drawn from a neck vein. The body was sent to the state forensic pathologist. At that examination bruises at the base of the neck were also seen. There was a lower lip abrasion. Periorbital and facial petechiae were seen. There was soft tissue injury over the anterior neck and evidence of prior needle puncture (for toxicology specimens). The throat and neck did not reveal any other soft tissue damage or fractures of the hyoid or larynx. Pulmonary edema and congestion were present. The pathologist concluded that the cause of death was asphyxia. The manner or death was homicide. AR was arraigned on charges of murder. The victim had a medical history of epilepsy with onset in early childhood. The medical records and hospital records revealed recurrent tonic–clonic seizures despite therapy with phenytoin and phenobarbital. She had been hospitalized on two occasions with status epilepticus between the ages of 10 and 12 years. During her married life, she continued to have generalized tonic–clonic seizures at the rate of one every month or two. AR also described episodes of staring that lasted 30 s to 2 min. Several years before her death, the victim had discontinued her antiepileptic drugs, in the belief that God would provide the best care and because the drugs apparently were not effective. There was no substantial change in her seizure frequency or severity after discontinuation. The defense lawyer for AR visited his home. The walls were obscured with hundreds of handwritten notes and letters. Some were notes of daily activities or news events. Many were “love letters” to AR and all were signed by the victim. These ἀndings were of unclear signiἀcance to the attorney until the phenomena were placed in the context of epilepsy and hypergraphia. A handwriting expert conἀrmed that Virginia had written the letters. As such, this suggested that no letters to AR were forced or forged and these were indeed “love letters” supporting the idea that there was no motivation for AR to kill “the only person who loved him.” Neurological expert witness testimony brought forth the concept that epilepsy was the most probable cause of Virginia’s death. The clinical setting of uncontrolled epilepsy, death in bed, and a recent convulsion 12 h before death, all indicated that the mode of death was most probably sudden unexpected death in epilepsy. The physical ἀndings of facial and periorbital petechiae were consistent with a recent and severe seizure and not with asphyxiation as alleged by the prosecution. The additional ἀndings of a bruised lip and pulmonary edema and congestion are common for SUDEP. The supporting evidence for murder that included a delay in calling for help with a matter-of-fact demeanor was attributed to AR’s unusual, bizarre, and suspicious behavior. The neck wounds were attributed to the inexperienced coroner having taken blood samples from the jugular vein. The “restraint” mark on the forearm was likely related to her wrist watch band being pushed up and some swelling. The “love letters” were excellent supporting evidence for their real relationship. Thus, the manner of death, SUDEP, was provided, and motivation for murder was eliminated. The jury acquitted AR.

SUDEP

355

The role of SUDEP in each of these cases has been pivotal, yet used in different ways. In the ἀrst case SUDEP was used as the evidence in the claim that the physician allegedly caused injury to the patient. The physician had no control over the patient’s decision to discontinue medication and had fulἀlled his duty of care in offering her appropriate therapy and advice. Both plaintiff and defendant in case 2 accepted the occurrence of SUDEP as a well-recognized phenomenon. Indeed, the defending attorney ampliἀed the issue of SUDEP in the cause of death only to proceed with the formulation of the defense that as the cause of SUDEP was not known, the treating physician could not be held responsible for causing death by his medical actions. Although the third case was remanded on appeal by a technical issue of time, the link between assault and death through SUDEP was conceivable and accepted by the jury when it found the defendant guilty of murder. In the ἀnal case SUDEP was presented by the defense as the correct mode of death rather than the alleged murder by strangulation. The medical evidence offering an alternative and most probable cause of death as presented by the defense was plausible and the jury found for the defendant. The autopsy report actually was consistent with SUDEP suggesting that the pathologist in 1997 was not familiar with the diagnosis of SUDEP.

24.3â•… Verbal Autopsy A widely accepted deἀnition of SUDEP is given by Nashef (1997): Sudden, unexpected, witnessed or unwitnessed, nontraumatic, and nondrowning death in patients with epilepsy with or without evidence for a seizure, and excluding documented status epilepticus, in which postmortem examination does not reveal a (structural or toxicological) cause for death. Functional criteria that are used in case determination of SUDEP (Leestma et al. 1997) include (1) a diagnosis of epilepsy, (2) unexpected death while in good health, (3) death was sudden (within minutes), (4) death occurred during normal activities and benign circumstances, (5) an obvious medical cause of death was not found, and (6) the death was not the direct result of the seizure or status epilepticus. Those deaths considered to be “deἀâ•›nite” SUDEP meet all criteria and have a postmortem report. “Probable” SUDEPs meet all criteria but lack a postmortem. Possible SUDEPs include cases in which SUDEP cannot be excluded. Often it is the part of the deἀnition and criteria that require mortem examination that creates a dilemma. Substantial problems exist today for obtaining an autopsy as well as creating death certiἀcates. These limitations complicate the need to get meaningful information for scientiἀc, epidemiological, and legal evaluations (Aspray 2005; Bell et al. 2004; Schraeder et al. 2006). Coroners or medical examiners resort to the postmortem examination or autopsy when a person dies unexpectedly and without apparent cause or is “found” dead. However, their decision can be discretionary. In making medical examiners and coroners more familiar with SUDEP another problem is created. As SUDEP becomes€a recognized mode of death, the examiners sometimes will make a presumptive step and€forgo an autopsy if family members or an observer report a medical history of active epilepsy. The rate of autopsy has declined drastically in the past few decades and autopsies are performed in less than 10% of deaths (Start et al. 1993; Ayoub and Chow 2008). It appears that in the United States this decline in routine autopsies may be related to (1)€ costs, (2)€ limâ•›ited reimbursement, (3) increased diagnostic conἀdence based on improved premortem diagnostic and technical capabilities, laboratory, and surgical pathological studies,

356 Sudden Death in Epilepsy: Forensic and Clinical Issues

(4) lack of physician requests/interests, and (5) a litigious atmosphere. In addition, speciἀc inhibitions for autopsy in epilepsy could include need for speciἀc protocols for epilepsy pathology (Black and Graham 2002a), no available and speciἀc research protocols, and limited involvement at the time of death by the neurologist/epileptologist who typically is not immediately informed of death. These situations hamper efforts to gather the best possible medical and epidemiological information about SUDEP. The autopsy that has mostly an exclusionary role (Ranson and Emmett 2004) was the cornerstone in establishing the diagnosis of SUDEP in three of the above described cases. There are numerous conditions that can be associated with sudden death (Wannamaker 1990) and in the most likely (coronary artery disease) and most unlikely (brain infection or tumor) circumstances (Black and Graham 2002b). There are discrepancies between premortem and postmortem ἀndings and diagnoses when autopsies are performed and reports (Battle et al. 1987) for major diagnoses with adverse impact are about 13%. There has been a decline in the diagnostic error rate with time to a level of about 9% (Shojania et al. 2003). Many of the autopsies that are conducted in the United States are at the request of the coroner or medical examiner due to the circumstances of death. Thus, there is a higher rate of autopsy in the cases of SUDEP or presumed SUDEP as is the situation in our practice experience (Table 24.1). In the speciἀc area of epilepsy one would anticipate that most patients and victims of SUDEP have had an array of diagnostic premortem studies which would eliminate many clinical and pathological errors. It must also be pointed out that despite the utility of our recent and very sophisticated diagnostic armamentarium the diagnostic error rate revealed by autopsy has changed only slightly over the past decades (Shojania et al. 2003). Autopsy remains a valuable examination in epilepsy and SUDEP; postmortem examinations may yield recently acquired coexistent reasons for death, demonstrate noteworthy microscopic ἀndings, provide histological information for diagnoses and classiἀcation and be the source for genetic studies and molecular biological studies. A complete autopsy may also uncover the presence and extent of extracranial changes related to comorbidities, which are a relatively unexplored area of epilepsy. Death certiἀcates inconsistently delineate cause of death in epilepsy (Bell et al. 2004) and SUDEP (Schraeder et al. 2006, 2009). Furthermore, there is no diagnostic classiἀcation code (ICD-9) for SUDEP. Consistent with these studies, the mention of epilepsy or epilepsy-associated phenomena (seizure, status epilepticus, convulsion, etc.) was not recorded on death certiἀcates in 2 of our 15 SUDEP cases. Indeed, only 26.2% of 386 death certiἀcates from our cohort recorded epilepsy or related phenomena in either Part I (cause) or Part II (other signiἀcant conditions). Based on full diagnostic evaluation and long-term care of patients in our practice, a comfortable conclusion is that there have been 15 SUDEP victims though only 5 of the 15 (33.3%) had autopsies. Thus, 10 SUDEP victims technically are relegated to the probable SUDEP category. In 14 of 39 possible or probable SUDEP Table 24.1â•… Autopsy Rates by Death Certificates (394) from an Epilepsy Practice Cohort (2979 Patients with Epilepsy) for SUDEP and Probable and Possible SUDEP Group SUDEP Probable and possible SUDEP Epilepsy, other Total

Number

Autopsy

Percent (%)

15 39 332 386

5 14 28 47

33.0 35.9 8.4 12.2

SUDEP

357

cases as suggested by death certiἀcate review, autopsy was done (35.9%). The drawback to the necessary autopsy as a requirement for affirming SUDEP continues to result in underestimation of the total extent of SUDEP deἀnite, probable, or possible. Even at this time the deἀnition of SUDEP and its recording (death certiἀcation) for public health issues is such that large-scale ascertainment of data for incidence information in the general population is difficult, if not impossible, to obtain. The currently available information is variable and population-speciἀc. In most instances where patients are under the care of an epileptologist, a clinical analysis to delineate deἀnite SUDEP or probable SUDEP should be sufficient. What are clearly needed are systems for improved and immediate surveillance and for good communications among families, medical examiners/coroners, and treating epileptologists at the time of death. Verbal autopsy for epilepsy deaths has been proposed as an alternative solution (Aspray 2005; Lathers and Schraeder 2009) for the dilemma of declining autopsy rates, limited utility of death certiἀcation, and public health policy and epidemiological studies in Third World areas. Our experience of discussing cases with families near the time of death has been favorable and clinically conἀrmatory. It is not clear whether the verbal autopsy would suffice in certain situations including when the “examiner” is unfamiliar with the families or observers, or when SUDEP is not considered as a potential cause of death, as in case 4. Familiarization with the patient, on the other hand, could lend itself to a prejudicial assessment and overstate speciἀc causes. Validation of this proposed procedure for epilepsy deaths will be necessary. In a fashion, coroners are already making some decisions based on verbal inquiries. From a medicolegal standpoint, the value of verbal autopsy would probably be restricted. Possibly, if conducted by an expert epileptologist, its strength would be greater. In summary, SUDEP has entered the arena of medicolegal interest. Our medical responsibilities require that neurologists provide the best possible information in the determination of cause of death and thus must be quite familiar with this potential mode of death. While the literature cites many risk factors for SUDEP, the actions of neurologists regrettably cannot prevent or alter SUDEP as an outcome at this time. This point creates a premise that renders the physician to a very circumscribed duty to tell or not to tell as in case 1 and case 2. Furthermore, the causal relationships for a SUDEP occurrence are not established. Hence, a complaint of causality could not support the action made against the physician in case 3. Recognition of SUDEP by the neurologist as the most probable cause of death prevented a grave major injustice (case 4). Information required to make a determination of cause and manner of death in epilepsy is facilitated by the best possible information gathered in very close proximity to the death. The postmortem examination is the gold standard and especially in medicolegal cases and yet must always be viewed with full disclosure of the clinical data. Case 4 is a notable example of autopsy results being erroneously interpreted. Autopsies have been utilized less frequently than in the past and verbal autopsies are offered as an alternative. This may be a feasible undertaking as it is in essence done every day by epileptologists when autopsies are not granted or pursued. However, validation in epilepsy deaths must be accomplished.

References Aspray, T. J. 2005. The use of verbal autopsy in attributing cause of death from epilepsy. Epilepsia 46 (Suppl. 11): 15–17.

358 Sudden Death in Epilepsy: Forensic and Clinical Issues Ayoub, T., and J. Chow. 2008. The conventional autopsy in modern medicine. J R Soc Med 101 (4): 177–181. Battle, R. M., D. Pathak, C. G. Humble, C. R. Key, P. R. Vanatta, R. B. Hill, and R. E. Anderson. 1987. Factors influencing discrepancies between premortem and postmortem diagnoses. JAMA 258 (3): 339–344. Bell, G. S., A. Gaitatzis, A. L. Johnson, and J. W. Sander. 2004. Predictive value of death certiἀcation in the case ascertainment of epilepsy. J Neurol Neurosurg Psychiatry 75 (12): 1756–1758. Beran, R. G. 2006. SUDEP—To discuss or not discuss: That is the question. Lancet Neurol 5 (6): 464–465. Beran, R. G., S. Weber, R. Sungaran, N. Venn, and A. Hung. 2004. Review of the legal obligations of the doctor to discuss sudden unexplained death in epilepsy (SUDEP)—A cohort controlled comparative cross-matched study in an outpatient epilepsy clinic. Seizure 13 (7): 523–528. Black, M., and D. I. Graham. 2002a. Sudden death in epilepsy. Curr Diag Pathol 8: 365–372. Black, M., and D. I. Graham. 2002b. Sudden unexplained death in adults caused by intracranial pathology. J Clin Pathol 55 (1): 44–50. Callaghan, N., A. Garrett, and T. Goggin. 1988. Withdrawal of anticonvulsant drugs in patients free of seizures for two years. A prospective study. N Engl J Med 318 (15): 942–946. Callenbach, P. M., R. G. Westendorp, A. T. Geerts, W. F. Arts, E. A. Peeters, C. A. van Donselaar, A. C. Peters, H. Stroink, and O. F. Brouwer. 2001. Mortality risk in children with epilepsy: The Dutch study of epilepsy in childhood. Pediatrics 107 (6): 1259–1263. Camἀeld, C. S., P. R. Camἀeld, and P. J. Veugelers. 2002. Death in children with epilepsy: A population-based study. Lancet 359 (9321): 1891–1895. Donner, E. J., C. R. Smith, and O. C. Snead III. 2001. Sudden unexplained death in children with epilepsy. Neurology 57 (3): 430–434. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33 (2): 310–316. El-Sayed, H. L., A. A. Kotby, H. Y. Tomoum, E. S. El-Hadidi, S. E. El Behery, and A. M. El-Ganzory. 2007. Non-invasive assessment of cardioregulatory autonomic functions in children with epilepsy. Acta Neurol Scand 115 (6): 377–384. Evrengul, H., H. Tanriverdi, D. Dursunoglu, A. Kaftan, O. Kuru, U. Unlu, and M. Kilic. 2005. Time and frequency domain analyses of heart rate variability in patients with epilepsy. Epilepsy Res 63 (2–3): 131–139. Harnod, T., C. C. Yang, Y. L. Hsin, K. R. Shieh, P. J. Wang, and T. B. Kuo. 2008. Heart rate variability in children with refractory generalized epilepsy. Seizure 17 (4): 297–301. Harvey, A. S., T. Nolan, and J. B. Carlin. 1993. Community-based study of mortality in children with epilepsy. Epilepsia 34 (4): 597–603. Kurtz, Z., P. Tookey, and E. Ross. 1998. Epilepsy in young people: 23 year follow up of the British national child development study. Brit Med J 316 (7128): 339–342. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies. Epilepsy Behav 14: 573–576. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earl. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38 (1): 47–55. Lewis, S., S. Higgins, and M. Goodwin. 2008. Informing patients about sudden unexpected death in epilepsy: A survey of specialist nurses. Br J Neurosci Nursing 4: 30–34. Morton, B., A. Richardson, and S. Duncan. 2006. Sudden unexpected death in epilepsy (SUDEP): Don’t ask, don’t tell? J Neurol Neurosur Ps 77 (2): 199–202. Nashef, L. 1997. Sudden death in epilepsy: Terminology and deἀnitions. Epilepsia 38 (Suppl. 11): S6–S8. Nashef, L., D. R. Fish, S. Garner, J. W. Sander, and S. D. Shorvon. 1995. Sudden death in epilepsy: A study of incidence in a young cohort with epilepsy and learning difficulty. Epilepsia 36 (12): 1187–1194.

SUDEP

359

Neuspiel, D. R., and L. H. Kuller. 1985. Sudden and unexpected natural death in childhood and adolescence. JAMA 254 (10): 1321–1325. Perper, J. A. 1980. Medicolegal implications of epilepsy. Med Trial Tech Q 27 (1): 9–36. Medical Research Council Antiepileptic Drug Withdrawal Study Group. 1991. Randomised study of antiepileptic drug withdrawal in patients in remission. Lancet 337 (8751): 1175–1180. National Health Service. 2004. Clinical Guideline 20: Epilepsy in Adults and Children. Available from http://www.nice.org.uk/nicemedia/pdf/CG020NICEguideline.pdf. Ranson, D. L., and S. L. Emmett. 2004. Exclusionary causes of death: Sudden unexpected death in epilepsy. J Law Med 11 (4): 414–416. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Shojania, K. G., E. C. Burton, K. M. McDonald, and L. Goldman. 2003. Changes in rates of autopsydetected diagnostic errors over time: A systematic review. JAMA 289 (21): 2849–2856. So, E. L. 2006. Demystifying sudden unexplained death in epilepsy—Are we close? Epilepsia 47 (Suppl 1): 87–92. Specchio, L. M., L. Tramacere, A. La Neve, and E. Beghi. 2002. Discontinuing antiepileptic drugs in patients who are seizure free on monotherapy. J Neurol Neurosurg Psychiatr 72 (1): 22–25. Start, R. D., T. A. McCulloch, E. W. Benbow, I. Lauder, and J. C. Underwood. 1993. Clinical necropsy rates during the 1980s: The continued decline. J Pathol 171 (1): 63–66. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol. 7 (11): 1021–1031. Tomson, T., M. Ericson, C. Ihrman, and L. E. Lindblad. 1998. Heart rate variability in patients with epilepsy. Epilepsy Res 30 (1): 77–83. United States Court of Appeals for the Fourth Circuit. 1994. United States v Chase. Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. D. Cascino, J. F. Annegers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56 (4): 519–525. Wannamaker, B. B. 1990. A perspective on death of persons with epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Weber, P., R. Bubl, U. Blauenstein, B. U. Tillmann, and J. Lutschg. 2005. Sudden unexplained death in children with epilepsy: A cohort study with an eighteen-year follow-up. Acta Paediatr 94 (5): 564–567. Yang, T. F., T. T. Wong, K. P. Chang, S. Y. Kwan, W. Y. Kuo, Y. C. Lee, and T. B. Kuo. 2001. Power spectrum analysis of heart rate variability in children with epilepsy. Childs Nerv Syst 17 (10): 602–606.

SUDEP Animal Models Mechanisms of Risks

II

Sudden Death Animal Models to Study Nervous System Sites of Action for Disease and Pharmacological Intervention

25

Claire M. Lathers

Contents 25.1 Introduction 25.2 In Vitro Animal Models 25.2.1 Rat/Mouse Hippocampal Slices 25.2.2 Isolated Autonomic Ganglia 25.2.3 Langendorff Preparation to Study Isolated Cat, Rat, or Mouse Hearts 25.2.4 Canine Purkinje Fiber Preparation 25.3 In Vivo Animal Models 25.3.1 Canine and Feline Models to Induce Myocardial Infarction, Arrhythmias, and Sudden Death 25.3.2 Cat Postganglionic Cardiac Sympathetic, Preganglionic Splanchnic Sympathetic, and Vagal Nerve Recordings Simultaneously with EKG Recordings after Acute Coronary Occlusion or Ouabain ToxicityInduced Arrhythmias and/or Death 25.3.3 Cat Postganglionic Cardiac Sympathetic Nerve Recordings Simultaneously with Synchronized EKG and EEG Recordings 25.3.4 Conscious Sheep and Cardiac Sympathetic Nerve Activity 25.3.5 Neurogenic Cardiac Arrhythmias in Anesthetized Rabbits 25.3.6 Cat Triceps Sura Muscle Preparation, Afferent and Efferent Electrophysiological Recordings and/or C1 Spinal Cord Transection 25.3.7 C1 Spinal Cord Transection or Bilateral Adrenal Vein Ligation Effect on Thioridazine-Induced Arrhythmia and Death in the Anesthetized Cat 25.3.8 C1 Spinal Cord Transection or 6-Hydroxydopamine or Intracerebroventricular Treatment in Anesthetized Cats 25.3.8.1 Intracerebroventricular Anticonvulsant and Antiarrhythmic Pharmacologic Actions 25.3.8.2 Central Hippocampal Penicillin Induced Epileptogenic Activity in Anesthetized Cats 25.3.9 Intraosseous Drugs in a Cardiac Arrest Model during Resuscitation in Anesthetized Swine 25.3.10 Intraosseous Drug Administration of Anticonvulsant Epilepsy Drugs in Anesthetized Swine 363

364 365 365 367 368 370 370 370

372 373 374 375 375 376 377 378 378 379 380

364 Sudden Death in Epilepsy: Forensic and Clinical Issues

25.3.11 Sympathetic Nerve Sprouting and Electrical Remodeling in Dogs: Mechanisms of Sudden Cardiac Death 25.3.12 Left and Right Stellate Ganglion Infusions in Dogs with Myocardial Infarction and Complete Heart Block 25.3.13 Spontaneous Stellate Ganglion Nerve Activity and Ventricular Arrhythmia and/or Sudden Death in a Canine Model 25.3.14 Cardiac Remodeling Post Myocardial Infarction in Mice 25.3.15 Digenic Mouse Model: Combination of Two Epilepsy Genes to Mask Epilepsy 25.3.16 Primate Models for Seizure and SUDEP 25.3.17 Other Animal Models 25.4 Conclusions References

381 381 382 382 383 383 384 386 386

25.1â•…Introduction How does one unravel the mystery of sudden unexpected death in persons with epilepsy (SUDEP)? Physicians must identify the persons with epilepsy who are at risk for SUDEP and use all available preventive medical and life style measures, gleaning as much information as possible about the deaths of SUDEP victims by talking with the medical examiners and coroners, reviewing autopsy reports (Schraeder et al. 2006, 2009, 2010), and obtaining verbal autopsy information provided by family members and close friends of the victim. The latter technique will help to ἀll in details missing from the physical autopsy or when no autopsy is done (Lathers and Schraeder 2009). It is also important to utilize in vivo and in vitro models of SUDEP in order to investigate risk factors, mechanisms, and preventive measures (Lathers et al. 2008; Scorza et al. 2008). Animal models allow us to focus on the details of one or more of the contributing mechanisms of risk for SUDEP. Use of experimental animals allows us to examine the contributing mechanisms of risk for SUDEP, with functional positive and negative feedback systems operating to maintain the normal physiology before experimental modiἀcations, and then to study the effect of dysfunction induced by experimental manipulation of the intact physiological system. Techniques may be applied to various animal models to glean information about the molecular and genetic mechanisms of risks for SUDEP. The various animal models described in this chapter are relevant to the induction of cardiac arrhythmias by various mechanisms and shed light and provide understanding of the problem of sudden death, whether of cardiac or epileptogenic origin. Application of the principles of clinical pharmacology when using animal models will provide two beneἀcial effects for persons at risk of sudden death. First, these models will provide the foundation for understanding the factors involved in the origin of arrhythmogenesis, and in the future will contribute to development of new categories of drugs that have both antiarrhythmic and antiepileptic effects. Second, application of the principles of clinical pharmacology when using animal models will allow development of techniques for better drug administration (e.g., introduction of the intraosseous route for pediatric seizing patients in whom a traditional intravenous line cannot be rapidly established in the emergency room in a life and death situation) (Spivey et al. 1985; Lathers et al. 1989c).

Sudden Death: Animal Models to Study Nervous System

365

Table 25.1â•… Selected Action Sites for Disease and Drugs 1. Central or peripheral autonomic nervous systems, including afferent and efferent neural activity 2. Superior cervical ganglia 3. Preganglionic nerve 4. Stellate ganglia 5. Postganglionic nerve 6. Baroreceptors 7. Heart 8. Preganglionic splanchnic nerve 9. Adrenal glands, release of circulating catecholamines

Application of clinical pharmacology principles to animal models will provide a better understanding of the mechanisms of the origin of cardiac arrhythmias and changes in the nervous system associated with epileptogenic activity and ultimately guiding development of preventative pharmacological measures. Nervous system sites of action for disease and/or pharmacological intervention are numerous. Animal models have been designed to mimic diseases such as coronary occlusion and myocardial or digitalis toxicity. The symptoms observed in both models are associated with the development of cardiac arrhythmias and/or sudden death (Lathers 1981). Cardiac arrhythmias that may trigger sudden death in patients originate from three primary sites of action, namely, directly in the heart, in the peripheral autonomic nervous system, or in the central autonomic nervous system and from combinations of some or all of these factors (Lathers et al. 1977, 1978). The central and peripheral autonomic nervous systems are connected, unless a patient has had a cardiac transplant. Below, various animal models are presented and the information included for each method is followed by possible applications to the problem of SUDEP. The discussion is a brief overview of models that can be used to study mechanisms of action for initiation of sudden death, whether for cardiac sudden death or sudden death in persons with epilepsy. Information obtained using these models, with appropriate modiἀcation, will provide insights into the mechanisms of sudden death. The data obtained from animal models will not only indicate physiological sites of action for disease but will also indicate pharmacological sites of action for antiepileptic drugs (AEDs) or antiarrhythmic drugs to prevent sudden death, as described in Table 25.1.

25.2â•…In Vitro Animal Models 25.2.1â•…Rat/Mouse Hippocampal Slices Synaptic plasticity is a long-lasting change in the efficacy of synaptic transmission resulting from patterned activities of the presynaptic nerve (Alkadhi et al. 2005). One type of synaptic plasticity is designated long-term potentiation, and is an activity-dependent marked increase in synaptic efficacy. In the central nervous system area of the hippocampus, longterm potentiation is thought to be a cellular correlate of attention, learning, and memory (Sarvey et al. 1989), studying synaptic transmission and synchronous activity in in vitro hippocampal slices obtained from rats (Sarvey et al. 1989; Dahl and Sarvey 1989; Lathers and Sarvey 1989). The hippocampal slice preparation is a tool to study the pharmacological

366 Sudden Death in Epilepsy: Forensic and Clinical Issues

effects of various drugs because the slice is viable and stable for several hours, allowing known concentrations of drugs to be added to the bathing medium and then washed out. The effects of norepinephrine and the beta blocking agent propranolol on long-term potentiation were examined. Norepinephrine (20 µM for 30 min) induced a long-lasting potentiation modiἀed by propranolol (Figure 25.1). Norepinephrine induced an activity-independent long-lasting depression of synaptic transmission in the lateral perforant path input to dentate granule cells, whereas highfrequency stimulation induced activity-dependent long-term potentiation. Bramham et al. (1997) investigated the role of endogenous activation of beta-adrenergic receptors in long(a)

R1 R2 FISSURE

R3

S1 Lat PP

Molecular layer

(b)

S2

Med PP

MEDIAL

LATERAL

1

240 µm

2

120 µm

3

0 µm

3a

0 µm

2 mV 2 ms

Figure 25.1╇ Scheme of a hippocampal rat slice showing the dentate gyrus and projections

of the lateral and medial PPs, the respective stimulating electrodes (S1 and S2) and recording electrodes (R1, R2, R3) in the outer molecular (R1), midmolecular (R2), and granule cell (R3) layers. (From Dahl, D., and J. M. Sarvey, Proc Natl Acad Sci U S A, 86 (12), 4776–4780, 1989. With permission.)

Sudden Death: Animal Models to Study Nervous System

367

term potentiation of the lateral and medial perforant paths under conditions affording selective stimulation of these pathways in the rat hippocampal slice. Propranolol (1 µM), a beta-receptor antagonist, blocked long-term potentiation induction of both lateral and medial perforant path-evoked ἀeld excitatory postsynaptic potentials. It was concluded there is a broad requirement for norepinephrine in different types of synaptic plasticity, including activity-independent depression and activity-dependent long-term potentiation in the lateral perforant path. A concentration-dependent long-lasting potentiation of the evoked population spike in the dentate gyrus of rat hippocampal slices was elicited by bath application of the GABAB receptor agonist baclofen (Burgard and Sarvey 1991). High concentration baclofen also produced a loss of inhibition manifest as the appearance of epileptiform, multiple-evoked population spikes. Data suggested that baclofen produces a selective disinhibitory effect in the granule cell layer of the dentate gyrus by inhibiting the activity of GABAergic interneurons. High concentrations of baclofen appeared to produce a long-lasting potentiation most likely due to a loss of inhibition. Lea and Sarvey (2003) reported modulation of epileptiform burst frequency by the metabotropic glutamate receptor subtype mGluR3. This receptor modulates high potassium (10 mM), low calcium (0.5 mN) induced spontaneous epileptic burst activity in acute rat hippocampal slice dentate granule cells. Activation of the group II metabotropic glutamate receptor subtype 3 induced an increase in spontaneous burst duration while inhibition reversibly reduced the spontaneous burst frequency. The number of spikes per burst was not altered. Nevertheless, this type of neuronal activity does induce spontaneous epileptiform burst frequency. This method is one model that may be explored to further understand mechanisms for and prevention of spontaneous epileptiform activity in the hippocampus. Recent data indicate that a deἀcit in glutamatergic synaptic transmission recorded in hippocampal slices obtained from aged mice, but not GABA-mediated synchronous network activity, may be coupled with alterations in synchronous network activity that could lead to deἀcient information process (Brown et al. 2005). This type of hippocampal activity may also be altered by epileptic discharges and/or AEDs administered to patients with epilepsy resulting in an impairment of cognitive function. Deἀnitive studies need to be conducted to evaluate this possibility. 25.2.2â•…Isolated Autonomic Ganglia Isolated superior cervical ganglia and stellate ganglia have been studied. The stellate ganglia and/or left stellate nerve activity become target sites for drug action and/or disease and may be studied in vitro (McIsaac 1978; Alkadhi and McIsaac 1973, 1974) and in vivo (Alzoubi et al. 2010). New studies are needed to examine the role of this site of action for both the induction and/or prevention of SUDEP. In vitro experiments are needed to examine the effects of anticonvulsant drugs. Data obtained in vitro should be compared with those obtained in vivo (see discussion below). A phenomenon similar to long-term potentiation in the hippocampus was identiἀed in sympathetic ganglia even earlier than its deἀnition in the hippocampus (Alkadhi et al. 2005). Ganglionic long-term potentiation of the nicotinic pathway is a long-lasting increase in synaptic effectiveness induced in autonomic ganglia after a brief train of relatively high-frequency stimulation of the preganglionic nerve and occurs in mammalian, amphibian, and avian species. Stress is a risk factor for sudden cardiac death and sudden death in persons with epilepsy (Pickworth et al. 1990; Lathers

368 Sudden Death in Epilepsy: Forensic and Clinical Issues

and Schraeder 2006). Sustained enhancement in ganglionic transmission occurs in chronic mental stress and may affect activity of autonomic functions of heart rate, arrhythmias, and blood pressure€(Alkadhi and Alzoubi 2007). Superior cervical ganglia from rats that developed hypertension as a result of chronic psychosocial stress express ganglionic longterm potentiation in vivo (Alzoubi et al. 2008). Synaptic plasticity in sympathetic ganglia may involve a molecular cascade similar to that of long-term potentiation in the hippocampal CA1 region. Future studies should be done to examine molecular changes in levels of phosphorylated CAMKII, total CaMKII, nitric oxide synthase, and calmodulin in superior cervical ganglia obtained from animals studied in models in which epilepsy is induced. Exposure to adrenergic agonists, as well as neuroactive peptides and cyclic nucleotides may produce long-lasting increases in synaptic effectiveness. Thus, one would predict that a beta-blocking agent would prevent this long-lasting increase in synaptic effectiveness and could exhibit an antiepileptic activity for beta blockers via an action on the preganglionic nerve and/or the autonomic ganglia. Understanding how to manipulate the mechanisms of induction and maintenance of ganglionic long-term potentiation in persons with epilepsy who may be at risk for sudden death is most important since the ganglionic long-term potentiation response is dependent on serotonin for both its induction and maintenance (Alkadhi et al. 2005). See Chapters 41 and 17, respectively (Faingold 2010; Patterson 2010), of this book for discussions of the possible role of serotonin in SUDEP or sudden infant death. The aging process itself is thought to involve a decrease in the ability of humans and animals to respond to stress (Alzoubi et al. 2008). Old animals have an exaggerated sympathetic activity associated with increased morbidity and mortality. In vivo expression of long-term potentiation in the superior cervical ganglion of aged animals may contribute to the moderate hypertension observed in aged subjects. Alzoubi et al. (2008) found elevations in molecular signals required for expression of long-term potentiation in sympathetic ganglia in obese Zucker rats in vivo. They also demonstrated that synaptic plasticity in sympathetic ganglia may involve a molecular cascade similar to long-term potentiation of the brain hippocampal area CA1. Additional studies of the role of these molecular signals in animal models of epilepsy are required to obtain a better understanding of the molecular mechanisms and of how to manipulate them to prevent sudden death. Long-term depression is a use-dependent decrease in synaptic efficacy and is a type of synaptic plasticity related to cognitive function in the central nervous system (Alkadhi et al. 2008). Long-term depression has been demonstrated in the rat superior cervical ganglia and suggests that expression of ganglionic long-term depression involves activation of 5-HT(3) receptors. 25.2.3╅Langendorff Preparation to Study Isolated Cat, Rat, or Mouse Hearts Langendorff established the isolated perfused mammalian heart preparation in 1897, a method still valid today (Skrzypiec-Spring et al. 2007). Crystalloid perfusates or blood enter the heart via a cannula in the ascending aorta at constant pressure or constant flow. Leaflets of the aortic valve are closed by the retrograde flow in the aorta. This action allows all of the perfusate to enter the coronary arteries via the ostia at the aortic root. The perfusate passes through the coronary circulation and drains into the right atrium via the coronary sinus. The method allows direct measurement of cardiac contractile function and coronary flow without interference from changes in the systemic circulation (Grover and

Sudden Death: Animal Models to Study Nervous System

369

Singh 2007). Use of this method provided the basis for understanding heart physiology, including the roles of temperature, oxygen and calcium ions for heart contractile function, origin of cardiac electrical activity in the atrium, the negative chronotropic effect of vagus stimulation, and chemical transmission of impulses in the vagus nerve by acetylcholine (Skrzypiec-Spring et al. 2007). The preparation is used to study ischemia-reperfusion injuries, cell-based therapy, donor heart preservation for transplant, and may be used to study vascular reactivity (Reichelt et al. 2009). Snyder et al. (1994) examined the decrease in norepinephrine release from cardiac adrenergic nerve terminal after ischemia and reperfusion whereas Vasilets et al. (1989) prevented reperfusion-induced arrhythmias by Na+/ H+ exchange block. Hearts with the right cardiac sympathetic nerve intact can be isolated and perfused and used to study the effect of beta blockers such as propranolol on presynaptic beta2 receptor mediated response in the rat heart (Mortimer et al. 1991), the role of calcium in adrenergic neurochemical transmission in the aging heart (Roberts et al. 1990), and gender differences (Tumer et al. 1992). Lathers et al. (1981, 1982) used in vivo animal data and compared/contrasted ἀndings obtained in the Langendorff preparations in rats to study the contributory role of the autonomic nervous system to cardiac arrhythmias and/or death. The central and peripheral autonomic nervous system inputs are not present in the Langendorff isolated beating heart preparation. Therefore, this preparation is used to examine effects of drugs on the heart, independent of autonomic neural input and independent of circulating substances such as catecholamines, both of which may trigger cardiac arrhythmias and/or death in in vivo preparations. Langendorff hearts were obtained from animals denervated chemically by pretreatment with reserpine, bretylium, or 6-hydroxydopamine or from rats whose hearts were surgically denervated and the effect of arrhythmic doses of digitalis glycosides were studied. In cats, pharmacologic denervation with bretylium pretreatment (2 μg/kg/min, i.v., until death) or with surgical denervation 2 weeks prior to 6-hydoxydopamine (20 mg/kg, i.v., 3 or 14 days before ouabain infusion to toxicity and death revealed that the protective action of only bretylium versus ouabain toxicity was eliminated by pretreatment with 6-OH dopamine (Lathers et al. 1982). Neither surgical denervation 2 weeks prior to the experiment or 6-OH dopamine given 3 or 14 days before ouabain infusion protected against the arrhythmogenic actions of ouabain. Since 6-OH dopamine and surgical denervation prevented the action of bretylium on ouabain-induced ventricular arrhythmia, bretylium action appears to be on the adrenergic nerve terminal. Bretylium, which acts on the adrenergic nerve terminal to leave it structurally intact but not functional and protected against ouabain-induced arrhythmia and death, differed from the effect of procedures that cause degeneration of the adrenergic nerve terminal (i.e., 6-OH dopamine and surgical denervation). The data suggests that for the protective effect of sympathectomy against ouabain-induced arrhythmia to develop, the adrenergic nerve terminal must be present, although not functional, as far as adrenergic neurotransmission is concerned. Neither a change in heart rate nor blood pressure appeared to be a factor in the protective effect of bretylium. Thus, comparison of data from these various animal models using both in vitro and in vivo data allowed conclusions to be made regarding the role of the actual heart rate in the initiation of cardiac arrhythmias and/or sudden death. In another laboratory, mean heart rates were compared in vivo and in isolated ex vivo preparation (Langendorff preparation) using hearts from rats with epilepsy (Colugnati et al. 2005). Differences occurred in the mean heart rate in vivo, but no differences were found in the heart rate in ex vivo, suggesting a central nervous system modulation on

370 Sudden Death in Epilepsy: Forensic and Clinical Issues

the heart that could lead to SUDEP. This conclusion was based on the fact that since the Langendorff preparation is an isolated heart preparation, it is devoid of input from the central nervous system. Thus, use of various in vitro and in vivo experimental designs will allow clariἀcation of the role of autonomic innervation and the influence of the central and peripheral nervous system in relation to epileptogenic activity, heart rate, arrhythmias, and sudden death. 25.2.4â•… Canine Purkinje Fiber Preparation Chlorpromazine, 10 μg/mL in Tyrode’s solution decreased Vmax and time to 50% repolarization (Lipka and Lathers 1987). No consistent change in the time to 95% repolarization was found. Chlorpromazine action on calcium mediated slow response produced by canine Purkinje ἀbers superfused with 12 mM KCl and 0.2 mg/L isoproterenol in Tyrode’s solution demonstrated that chlorpromazine prolonged the upstroke of the slow response and increased the duration of the stimulus necessary to produce the slow response. Data were interpreted to mean that chlorpromazine alters cardiac conduction to induce arrhythmia. The pharmacological agent may be acting to initiate arrhythmia by altering the cardiac slow response (Lipka et al. 1988).

25.3â•…In Vivo Animal Models 25.3.1â•…Canine and Feline Models to Induce Myocardial Infarction, Arrhythmias, and Sudden Death Dog and cat models are used to study myocardial infarction, arrhythmias, and sudden death since they induce sequelae similar to those observed in humans. Mechanisms of arrhythmia development, the role of imbalance in autonomic neural discharge, altered electrophysiology of the myocardium, size and location of ischemic and/or infracted areas, biochemical and molecular changes, presence of coronary collaterals, pathophysiology of myocardial perfusion, and drug effects on the sequalae of coronary occlusion may be examined in various in vivo animal models. Myocardial infarction can be induced readily in dogs and cats (Lathers 1981). Many techniques are used to occlude coronary arteries to produce symptoms comparable to those observed in humans. See Table 25.2. Use of these models contributed to progress made in the diagnosis and treatment of human myocardial infarction. The technique of acute coronary ligation has been used for more than 100 years (Cohnheim and von Schultess-Rechberg 1981; Lathers 1981). Many research laboratories used this model in the 1960s and 1970s (Ceremuzynski et al. 1969; Gillis 1971; Lathers et€al. 1977, 1978; Bissett et al. 1979) and have continued to use this model in the subsequent Table 25.2â•… Techniques for Producing Myocardial Infarction in Dogs and Cats 1. Acute coronary ligation in anesthetized, thoracotomized animals 2. Two-stage coronary ligation in anesthetized, thoracotomized animals 3. Gradual coronary occlusion using various types of materials implanted in anesthetized, thoracotomized animals 4. Use of selective catheterization to deliver emboli to coronary arteries in unanesthetized, closed-chest dogs Source: From Lathers C. M. Induced disease. Myocardial infarction in dogs and cats. In Mammalian Models for Research on Aging, 224–228. Washington, DC: National Academy Press, 1981.

Sudden Death: Animal Models to Study Nervous System

371

years (Lathers 1975, 1979, 1980a, 1980b, 1981; Lathers et al. 1976, 1978, 1988a; Spivey and Lathers 1985). In contrast to acute coronary ligation, the two-stage coronary occlusion method developed in the dog by Harris (1950) results in a larger number of animals surviving occlusion. First, a partial coronary occlusion is done, followed by complete coronary occlusion in an anesthetized, thoracotomized dog. Twenty-four hours after the dog has recovered from anesthesia, various types of experiments may be initiated (Harris et al. 1971; Gillis et al. 1973; Heng et al. 1978; Reynolds et al. 1979; Ritchie et al. 1979; Hashimoto et al. 1979). Techniques of gradual occlusion are used because they allow development of collateral circulation, mimicking the prompt development of collaterals when human coronary blood flow is restricted in advanced coronary atherosclerosis (Baroldi et al. 1956). One limitation of all the above methods of coronary ligation is that thoracotomy, done to expose and manipulate the coronary artery, alters neural and lymphatic pathways while opening the pericardium and surgical follow-up procedures are more difficult. The resulting infarction is dependent on many factors, including the size of the artery, the site and speed of occlusion, anatomy and preexisting state of other major arteries and myocardium, distribution, and extent of collateral circulation (Jobe 1967), and whether anesthesia has been used. The anesthesia itself may exert a protective influence (Bland and Lowenstein 1976) and decrease the probability of fatal ventricular ἀbrillation. Closed-chest techniques of occlusion have evolved and may employ selective catheterization to deliver multiple small emboli to the coronary arteries using lycopodium spores, glass or plastic beads, and clots (Guzman et al. 1962; Agress et al. 1952). As one would predict, these techniques did produce obstruction of small vessels but the changes were unpredictable and dissimilar to lesions observed in humans. Injection of autologous clots into the left anterior descending coronary arteries of dogs has been used to induce embolization until infarction or arrhythmia developed (Baumstark et al. 1978). Likewise, use of wire conductors to induce thrombus formation with electrical or thermal energy was reported (Salazar 1961). Roswell et al. (1965) tried the infusion of ADP into major coronary arteries to produce occlusive platelet aggregation and myocardial infarction. All of these techniques made it difficult to regulate the size and location of the obstruction. Thus the technique of coronary latex microsphere (25-µm diameter) embolization was developed to study postinfarction arrhythmias. These latex microspheres were used to examine pharmacological alterations of different parts of the coronary circulation to further elucidate the hemodynamic relationships between collateral and nutritive microcirculation (Wichmann et al. 1978) changes in collateral blood flow and development of myocardial necrosis (Reimer and Jennings 1979) and the time events of ischemic cell death and ability of pharmacological agents to limit the size of the infarction (Jugdutt et al. 1979). Data using these techniques especially provided information about changes occurring early after the occurrence of infarction (i.e., often in the time interval when patients do not survive long enough to reach a hospital). Using dog and cat models, techniques have been developed to study the ability of pharmacological agents to limit the area of necrosis postinfarction (Maroko et al. 1972; Reimer et al. 1973; Lucchesi et al. 1976; Powell et al. 1976) and to develop new techniques to measure size. Drugs such as practolol (Libby et al. 1973; Marshall and Parratt 1974; Lathers€et€al. 1976), methylprednisolone (Spath et al. 1974; Busuttil et al. 1975; Lathers 1979), and metoprolol (Sivam and Seth 1978; Lathers 1980b) were studied early on. Techniques to measure infarct size included 99m technetium pyrophosphate scans (Bonte et al. 1974; Willerson et al. 1979), thallium-201 myocardial perfusion imaging (Ritchie et al. 1979),

372 Sudden Death in Epilepsy: Forensic and Clinical Issues

indium-111 gamma-emitting radionuclide labeling (Thakur et al. 1979), iodine-131 labeled antibody (Fab) 2 fragment imaging (Khaw et al. 1978a, 1978b), computerized axial tomography€ scans (Siemers et al. 1978), intravenous administration of diatrizoate meglumine and sodium (Renograἀn-76) (Higgins et al. 1979) reduced nicotinamide adenine dinucleotide fluorescence photography (Barlow and Chance 1976), serial creatinine phosphokinase technique (Shell et al. 1971) nitro-blue tetrazolium test (Nachlas and Shnitka 1963), and two-Â�dimensional echocardiography (Meltzer et al. 1979). Both dogs and cats have been studied to evaluate the role of the autonomic nervous system in the production of acute coronary occlusion-induced arrhythmias (Constantin 1963; Malliani et al. 1969; Gillis 1971; Rotman et al. 1972; Webb et al. 1972; Levitt et al. 1976; Gillis et al. 1976). These species have also been used to examine mechanisms by which antiarrhythmic agents act on autonomic neural discharge (Lathers 1975; Levitt et al. 1976; Kupersmith 1976). Altered sympathetic neural discharge to the heart and/or an imbalance between the sympathetic and parasympathetic nervous system is thought to be involved in the production of ventricular arrhythmias after occlusion (Webb et al. 1972; Levitt et al. 1976; Schwartz et al. 1976; Lathers et al. 1977, 1978; Wehrmacher et al. 1979). Other factors involved in cardiac susceptibility to ventricular arrhythmias include extent and intensity of ischemia, severity of metabolic alteration within the ischemic area, the perfusion gradient between the nonischemic and ischemic cardiac muscle, and extent of coronary vascular collateralization (Corday et al. 1977). The autonomic nervous system directly affects the ἀrst three factors. Pharmacological or surgical interventions that correct regional ischemia to decrease extent of ischemic injury may also decrease the incidence of ventricular ἀbrillation. Lown et al. (Lown and Wolf 1971; Lown et al. 1977) reviewed the data supporting the role of altered sympathetic discharge and acute myocardial ischemia in sudden cardiac death after myocardial infarction. At that time, a number of experiments were ongoing in Dr. Lathers’ laboratory (Lathers et al. 1977, 1978), resulting in an animal model to monitor neural discharge in two or three postganglionic cardiac nerves simultaneously (see below). Nerve activity was monitored before and after initiation of arrhythmias and/or death. Several years later, this model was adapted to monitor EEG and the role of cardiac neural discharge in this animal model of sudden death associated with interictal and ictal activity. 25.3.2â•…Cat Postganglionic Cardiac Sympathetic, Preganglionic Splanchnic Sympathetic, and Vagal Nerve Recordings Simultaneously with EKG Recordings after Acute Coronary Occlusion or Ouabain Toxicity-Induced Arrhythmias and/or Death In two in vivo animal models, altered postganglionic cardiac sympathetic discharge induced cardiac arrhythmias and/or death (Lathers et al. 1977, 1978). Acute occlusion of the left anterior descending coronary artery in cats anesthetized with alpha-chloralose and pretreated with atropine induced arrhythmia within 3 min. Some of these animals died in ventricular ἀbrillation (Lathers et al. 1978). In the animals with arrhythmia, postganglionic cardiac neural discharge became nonuniform (i.e., spontaneous discharge increased in nine nerves, decreased in ἀve nerves, and showed no change in one nerve). The nonuniform neural discharge was associated with development of arrhythmia after occlusion. In some cats, neural discharge did not change within the ἀrst 3 min after coronary artery occlusion and arrhythmia did not occur. In a different set of animals, ouabain toxicity was initiated

Sudden Death: Animal Models to Study Nervous System

373

by giving bolus injections of ouabain (25 μg/kg, i.v.) every 15 min until death. Development of ouabain-induced arrhythmia was also accompanied by a nonuniform pattern in the neural discharge (Lathers et al. 1977). This nonuniform neural discharge may alter ventricular excitation and conduction to produce arrhythmia in the manner described by Han and Moe (1964). They demonstrated nonuniform recovery of excitability in cardiac ventricular muscle. The nonuniform postganglionic cardiac sympathetic neural discharge may be, in part, one mechanism involved in the production of coronary occlusion or ouabain-induced arrhythmias. In a different series of studies, Lathers (1980a) examined the effect of pretreatment with the beta-blocking agent timolol (5 mg/kg, i.v., infused at a rate of 0.5 mg/kg/min for 10 min). Postganglionic cardiac and preganglionic splanchnic sympathetic and vagal neural discharge, ouabain-induced arrhythmia, heart rate and mean arterial blood pressure were monitored. The ἀrst bolus injection of ouabain was given 15 min after the timolol infusion. Timolol increased the time to ouabain-induced arrhythmia and death from 23 ± 3 to 48 ± 7 and 46 ± 3 to 76 ± 9 min, respectively (p < 0.05). Heart rate and mean arterial blood pressure decreased from 137 ± 4 to 104 ± 6 beats/min and 133 ± 6 to 103 ± 7 mm Hg, respectively (p < 0.05). Ouabain did not reverse the decreases. The infusion of timolol did not signiἀcantly alter the neural activity monitored from the vagus and the postganglionic cardiac and preganglionic splanchnic sympathetic nerves. Ouabain after timolol did not alter splanchnic nor vagal discharge. Postganglionic cardiac sympathetic neural discharge exhibited both increases and decreases (i.e., a nonuniform neural discharge, with development of ouabain-induced arrhythmia). The ouabain-induced nonuniformity did not occur in animals pretreated with timolol. The protective effect of timolol may be due, in part, to prevention of the nonuniform postganglionic cardiac sympathetic neural discharge and to prevention of ouabain-induced increases in vagal discharge. Establishment of beta-blockade and a direct negative inotropic action may also contribute to the antiarrhythmic action of timolol. This animal model (Lathers et al. 1977, 1978) was modiἀed to simultaneously monitor EEG to develop a new animal model for sudden explained death in persons with epilepsy (Lathers and Schraeder 1982; Schraeder and Lathers 1983). 25.3.3â•…Cat Postganglionic Cardiac Sympathetic Nerve Recordings Simultaneously with Synchronized EKG and EEG Recordings A new animal model was developed to examine mechanisms for risk of sudden death in persons with epilepsy (Lathers and Schraeder 1982). Epileptic activity was induced with pentylenetetrazol. Electrocorticogram (ECoG), postganglionic cardiac sympathetic discharge, lead II EKG, and mean arterial blood pressure were monitored. Autonomic dysfunction was characterized by several ἀndings: the autonomic cardiac nerves did not always respond in a predictable manner to changes in blood pressure; a marked increase in variability of the mean autonomic cardiac sympathetic and parasympathetic neural discharge occurred; and a very large increase occurred in the variability of the discharge rate of parasympathetic nerves after 50 mg/kg pentylenetetrazol but did not develop until 100€mg/kg in the sympathetic nerves. While pentylenetetrazol (10 mg/kg) induced interictal epileptogenic activity, at higher doses the duration of ictal activity increased and the interictal discharges, if present, were of shorter duration. Autonomic imbalance occurred with both interictal and ictal discharge. Interictal epileptogenic activity was associated with nonuniform autonomic cardiac postganglionic neural discharge and by an imbalance of postganglionic cardiac sympathetic and vagal discharges causing cardiac arrhythmias

374 Sudden Death in Epilepsy: Forensic and Clinical Issues

(ventricular ἀbrillation or asystole) and/or sudden death. This model is discussed in depth in Chapter 28 (Lathers and Schraeder, 2010). Using this animal model, the lockstep phenomenon was detected. The lockstep phenomenon is the occurrence of postganglionic cardiac sympathetic and vagal discharges intermittently synchronizing one for one with both ictal and interictal cortical discharges in a time-locked fashion. The incidence of lockstep was less often detected with cardiac vagal neural discharge (Lathers and Schraeder 1987; Lathers et al. 1987; Stauffer et al. 1989, 1990). Premature ventricular contractions, ST/T changes, conduction blocks, and precipitous changes in blood pressure occurred concomitantly with the onset of interictal activity associated with the lockstep phenomenon. The duration of each lockstep phenomenon pattern was examined. Autonomic dysfunction with epileptogenic activity causing cardiac arrhythmias has been postulated as a cause of sudden unexplained death in persons with epilepsy. The lockstep phenomenon and associated patterns as related to precipitous changes in mean arterial pressure and/or occurrence of arrhythmias are indicators of changes in autonomic function. One potential mechanism for SUDEP is considered to be development of the lockstep phenomenon. Understanding the autonomic changes, as indicated by postganglionic cardiac sympathetic and vagal discharges associated with the lockstep phenomenon, may contribute to the prevention of sudden death in persons with epilepsy. See Chapter 28 for details of the lockstep phenomenon (Lathers and Schraeder 2010). 25.3.4â•… Conscious Sheep and Cardiac Sympathetic Nerve Activity The above animal models utilized anesthetized cats. Currently a conscious sheep model has been developed to examine cardiac sympathetic nerve activity and ventricular ἀbrillation during acute myocardial infarction (Jardine et al. 2005, 2007). Electrodes were glued into the thoracic cardiac nerves, allowing conscious cardiac sympathetic nerve activity recordings to be obtained before and after myocardial infarction. Mean arterial blood pressure and heart rate were also recorded. An early increase in cardiac sympathetic nerve activity burst size indexes occurred before 60 min postmyocardial infarction, mediated by an excitatory sympathetic reflex (Jardine et al. 2007). The authors concluded this mechanism was important in the genesis of ventricular ἀbrillation or sustained ventricular tachycardia. In another study, cardiac sympathetic nerve activity increased within 1 h of the onset of myocardial infarction (Jardine et al. 2005). Cardiac sympathetic nerve activity burst frequency increased from baseline 2 h postmyocardial infarction and remained elevated for 2 days. Cardiac sympathetic nerve activity burst area also increased and was sustained for 7 days after myocardial infarction. Baroreflex slopes for pulse interval and cardiac sympathetic nerve activity did not change. In a third study (Charles et al. 2008), urocortin 1, an entity that works with the sympathetic nervous system to participate in cardiac and circulatory regulation, was studied. Urocortin 1 reduced cardiac sympathetic nerve activity, with signiἀcant decreases in a dose-related response for burst frequency, burst incidence, and burst area when compared with time-matched control data. Urocortin 1 induced signiἀcant rises in heart rate and cardiac output and reduced peripheral resistance with no effect on mean arterial blood pressure. It was recommended that urocortin 1 be evaluated as a potential therapeutic application in acute myocardial injury and heart disease. Lathers et al. (Lathers et al. 1987; Lathers and Schraeder 1987) found that the lockstep phenomenon is associated with interictal and ictal activity and the occurrence of cardiac arrhythmias and/or death. The conscious sheep model should be studied after implanting chronic EEG

Sudden Death: Animal Models to Study Nervous System

375

electrodes and monitoring before and after induction of epileptogenic activity with an agent such as pentylenetetrazol. Potential new anticonvulsant drugs should also be evaluated in this model. 25.3.5â•…Neurogenic Cardiac Arrhythmias in Anesthetized Rabbits Another animal model used to study the role of the sympathetic nervous system in the genesis of neurogenic cardiac arrhythmias is an in vivo halothane-anesthetized rabbit model (Poisson et al. 2000). Chronic bipolar electrodes can be implanted within the posterior hypothalamus of halothane-anesthetized rabbits. Every 15 days, after three 10-min interval control electrical stimulations, the effects of techniques to induce interictal and ictal epileptogenic activity with the intravenous or central administration of pentylenetetrazol or the central application of penicillin or the induction of cardiac arrhythmias via acute coronary occlusion, can be studied. In the rabbit model, the effects of procedures or pharmacological agents may be studied on parameters of blood pressure, heart rate, and EKG. EKG effects to be monitored include the number of extrasystoles and abnormal complexes and whether a pharmacological agent exhibits antiarrhythmic activity. This model can also be used to examine the action of AEDs on the EEG and EKG and the associated effects, if any, as antiarrhythmic agents. 25.3.6â•…Cat Triceps Sura Muscle Preparation, Afferent and Efferent Electrophysiological Recordings and/or C1 Spinal Cord Transection Zipes (2008) emphasized that substantial evidence exists for a neural component in sudden cardiac death. Since sympathetic nerve injury promotes cardiac arrhythmia and sudden death, one method to protect against death is to modulate autonomic tone to decrease risk of ventricular arrhythmias. Indeed, spinal cord stimulation provides protection against ventricular tachycardia/ventricular ἀbrillation when tested in animal models of postinfarction heart failure. Stimulation of the thoracic spinal cord may be one treatment for refractory angina. One animal model to examine the effect of spinal afferent and efferent changes is the cat triceps sura muscle preparation. When developing new anticonvulsant drugs, one must ask the question of whether the AEDs alter motor coordination, and if so, is the action occurring via an action in the spinal cord and/or in the central nervous system and/or via a peripheral action? Observations of AED effects in unanesthetized animals, while also obtaining blood levels of the new AEDs, will reveal if mild to marked motor incoordination is elicited by the investigational AED and will demonstrate at what blood level this effect occurs. Sir John Eccles and coauthors (1975a, 1975b, 1975c, 1975d) and Armstrong et al. (1968) used various electrophysiological versions of an elegant animal model, the cat triceps sura muscle preparation, to study the electrophysiological effects of drugs on afferent and efferent (spinal reflexes) and their central actions. This anesthetized cat model was used by Lathers and Smith (1976) to study pharmacological action of drugs such as ethanol on muscle spindle afferent activity and on the mono- and polysynaptic spinal reflexes. Mean phasic and static discharge frequency responses were examined. Supramaximal single shocks to the muscle nerve allows one to record drug induced changes from control in the mean amplitudes of the mono- and polysynaptic reflexes in spinal animals. Using this model one could determine if spindle afferent excitation is involved in the motor effects

376 Sudden Death in Epilepsy: Forensic and Clinical Issues

of new anticonvulsant drugs. Likewise, use of this animal model in a separate series of studies in spinal animals (C1 spinal cord transection) will demonstrate if new anticonvulsant drugs induce action(s) occurring at peripheral and spinal sites and/or if AED actions are modiἀed by supraspinal sites in terms of impairment of skeletomotor function. Data obtained from the use of this animal model can be embellished by simultaneously recording EEG changes. The model may be studied in cats with and without C1 spinal section to separate central versus peripheral effects. Likewise, this cat model was used by Lathers and Smith (1976) to examine changes in the burst patterns of spinal afferent nerve discharge by measuring interspike intervals before and after administration of drugs such as succinylcholine and ethanol. Future studies using this animal model are needed and will contribute to our knowledge of central, spinal, and peripheral actions of AEDs. 25.3.7â•…C1 Spinal Cord Transection or Bilateral Adrenal Vein Ligation Effect on Thioridazine-Induced Arrhythmia and Death in the Anesthetized Cat Some psychoactive agents may induce autonomic dysfunction, cardiac arrhythmias, and/or sudden death. To explore whether the cardiotoxic and/or sudden death actions of the phenothiazine drug thioridazine is due to an action on the central nervous system or via release of catecholamines from the adrenal glands, the effect of C1 spinal cord transaction or bilateral adrenal vein ligation may be studied using various experimental designs. Thioridazine 1 mg/kg/min was infused intravenously into three groups of cats: (1) thioridazine only, (2) after bilateral adrenal ligation, and (3) after spinal cord section at the atlanto-occipital junction (C1) (Lathers et al. 1986). Times to arrhythmia and death with thioridazine alone or after bilateral adrenal ligation or after spinal cord section were not signiἀcantly different. Thus, the data suggests that neither adrenomedullary catecholamines nor the central sympathetic component above C1 plays a signiἀcant role in acute thioridazine-induced arrhythmia. The action of thioridazine to induce arrhythmia in spite of transection of the spinal cord or bilateral adrenal vein ligation suggests that its cardiotoxicity is a result of a direct myocardial effect. Thioridazine depressed blood pressure without producing the sustained reflex tachycardia normally seen with hypotension, indicating this drug may modify the baroreceptor reflex arc. In the future, with the addition of EEG electrodes, this animal model may be used to examine the effect of C1 spinal cord transection to separate central, from spinal, from peripheral effects in animal models with epileptogenic activity. Also, this cat model, with simultaneous recording of the EEG, may be used to explore the role of adrenal catecholamines in the initiation of cardiac arrhythmias and/or death associated with epileptogenic activity. The effect of another phenothiazine drug, chlorpromazine, was studied using two different models (Lathers and Lipka 1986). Dial-urethane anesthetized cats were studied to determine times to ouabain-induced arrhythmia and death. Data were compared with data obtained in dial-urethane cats with no ouabain. Doses of chlorpromazine (5, 10, 20, 30, 40, or 60 mg/kg, i.v.) were found to be neither arrhythmogenic nor antiarrhythmic in the ouabain toxicity model. In animals receiving just chlorpromazine at a rate of 1 mg/kg/min (i.v.), bilateral adrenal vein ligation to eliminate the action of adrenal catecholamines decreased, rather than increased, the dose of chlorpromazine necessary to produce arrhythmia and death. Thus, adrenal catecholamines did not appear to contribute to chlorpromazine-induced arrhythmia although the procedure of bilateral adrenal vein

Sudden Death: Animal Models to Study Nervous System

377

ligation was deleterious in combination with chlorpromazine. In all experimental models, chlorpromazine did depress the blood pressure but did not produce the reflex tachycardia normally observed with hypotension, suggesting chlorpromazine, like thiordazine, may alter baroreceptor reflex arc activity. Chlorpromazine-induced death occurred via cardiovascular collapse. Since chlorpromazine modiἀes the autonomic parameters of blood pressure, heart rate, and cardiac electrophysiology, sudden unexplained death in persons taking this medication may be due to drug-induced arrhythmias. 25.3.8â•…C1 Spinal Cord Transection or 6-Hydroxydopamine or Intracerebroventricular Treatment in Anesthetized Cats In another series of experiments, the role of the central nervous system in the production of phenothiazine-induced arrhythmia and death versus direct drug action on the heart was examined (Lipka et al. 1988). Spinal cords were transected at the atlanto-occipital junction prior to the 1 mg/kg/min, intravenous infusion of chlorpromazine or thioridazine. No protection against drug-induced arrhythmia or death was found. These data were compared to data obtained in other cats in which 6-OH-dopamine was utilized to develop an in situ denervated heart preparation. The 6-OH-dopamine was administered prior to intravenous injection of atropine and infusion of chlorpromazine, 1 mg/kg/min. In the preparations treated with 6-OH-dopamine, no protection against chlorpromazine-induced arrhythmia or death occurred. In another group of animals, 0.5 mg chlorpromazine was administered intracerebroventricularly to alpha-chloralose anesthetized cats. This central administration of chlorpromazine did not induce arrhythmia or death, although the blood pressure was initially decreased. It was concluded that neither chlorpromazine nor thioridazine appeared to produce arrhythmia or death via a central locus. These phenothiazine drugs may be acting directly on myocardial conduction to produce arrhythmia and death (Lipka et al. 1988; Lipka 1987). In the future, with the addition of EEG electrodes to cats, the model may be used to differentiate central nervous sites of action from direct cardiac actions in animals with interictal and ictal activity, with and without the occurrence of arrhythmias or death. Comparison of these in vivo data with those obtained using in vitro canine Purkinje ἀbers, although two different species, again suggested that chlorpromazine acts on cÂ�ardiac conduction to induce arrhythmia and may trigger arrhythmias by altering the cardiac slow response (Lipka 1987). Classically, data obtained in animal models is examined for evidence that may be provided for use when treating humans. Lipka and Lathers (1987) discuss the fact that major tranquilizers as well as antidepressant agents have been associated with clinical seizures, although the incidence is low if the doses are therapeutic. Animal model data indicates phenothiazines act as anticonvulsant drugs at higher doses and are convulsants at lower doses. Antidepressants exert anticonvulsant actions at low doses and convulsant action with high doses. Obviously drugs of lower seizure production potential should be substituted for those with greater potential in treating epileptic patients for psychiatric disorders. The problem of a possible comorbid role for psychotropic drugs in sudden death in epileptic persons must be considered. Some psychoactive agents have been associated with sudden death as well as cardiac arrhythmia and seizure production. Therefore, administration of psychoactive agents to a person with epilepsy should be approached with caution. Psychoactive agents that have a minimal risk of altering cardiac rhythm or seizure threshold should be used if such drugs are required.

378 Sudden Death in Epilepsy: Forensic and Clinical Issues

25.3.8.1â•…I ntracerebroventricular Anticonvulsant and Antiarrhythmic Pharmacologic Actions Alpha-chloralose anesthetized cats received pentylenetetrazol, 10 and/or 20 mg, intraÂ� cerebroventricularly to induced interictal and ictal activity (Lathers et al. 1989a). Timolol (10, 100, 500 mg/kg) was administered intracerebroventricularly or doses of 1, 5, 10, and/or 20 mg/kg were given intravenously. The central dosing of timolol reversed the epileptiform activity of pentylenetetrazol in the brain and suppressed associated increases in blood pressure, heart rate, and cardiac arrhythmias. Kraras et al. (1987) hypothesized that since d-alanine2 met-enkephalin induced a centrally mediated vasopressor response and attenuation of the baroreceptor reflex in conscious cats (Yukimura et al. 1981), possibly leading to autonomic imbalance, the latter may precipitate arrhythmias and sudden death in persons with epilepsy. They concluded that resolution of the question of whether enkephalins elicit epileptogenic activity and autonomic dysfunction via inhibition of GABA release is important since an understanding of this mechanism should eventually allow for the design of pharmacologic agents that prevent epileptogenic activity, autonomic dysfunction, and the associated risk of SUDEP. Studies were conducted to determine if d-ALA2 methionine enkephalinamide (DAME) contributed, at least in part, to initiation of autonomic dysfunction. Lathers and Schraeder (1982) established autonomic dysfunction associated with epileptogenic activity induced by pentylenetetrazol and Vindrola et al. (1983) reported increased d-ALA2 methionineenkephalinamide levels in rat brain after pentylenetetrazol-induced epileptogenic activity. A series of experiments studied the effects of intracerebroventricular d-ALA2 methionine enkephalinamide and naloxone in the cat to determine what changes occurred in autonomic cardiovascular parameters (Lathers et al. 1988b). The data suggest that DAME may induce epileptogenic activity and cardiovascular changes via an action on central opiate receptors. The authors hypothesized that increased levels of DAME inhibit the release of gamma aminobutyric acid and this then increases vagal bradycardia and hypotension. The latter event causes an imbalance in peripheral autonomic cardiac neural discharge that may initiate arrhythmias and/or sudden unexplained death in the individual with epilepsy. Additional studies are needed to explore this hypothesis and possible mechanism to determine the potential role in cause of sudden death (Kraras et al. 1987). It is of interest to note that a few years later, Asai et al. (1994) found that valproic acid induced rapid changes in the levels of striatal methionine-enkephalin levels in rat brain and suggested there may be an association with its anticonvulsant activity. The role of enkephalins in the induction of seizures and SUDEP and how drugs may modify this effect needs further study. 25.3.8.2â•…Central Hippocampal Penicillin Induced Epileptogenic Activity in Anesthetized Cats Focal hippocampal administration of penicillin G in cats is another animal model to study autonomic dysfunction in blood pressure and heart rate and rhythm associated with induced epileptiform activity (Lathers and Schraeder 1990; Lathers et al. 1993). The delay in onset of epileptiform activity at the site of injection ranged from 1 s to 16 min, and consisted of interictal discharges or ictal discharges. Blood pressure and heart rate increased signiἀcantly from control (p < 0.05) with the onset of epileptiform activity. Electrocardiogram changes were found and included P–R interval changes, increased P-wave amplitude, QRS cÂ�omplex€changes, T-wave inversion, and ST elevation. Phenobarbital (20 mg/kg, i.v.) suppressed epileptogenic activity and depressed the blood pressure and heart rate below control

Sudden Death: Animal Models to Study Nervous System

379

levels (p€< 0.05). In another series of experiments, penicillin G injected into the right hippocampus of cats produced epileptiform activity and increased the blood pressure and the heart rate signiἀcantly from control levels (p < 0.05). Phenobarbital (20 mg/kg, i.v., and 40 mg/kg,€i.v.) also prevented the penicillin-induced epileptiform activity. Phenobarbital (40€mg/kg, i.v.) reversed the effect of penicillin on blood pressure and heart rate, to levels below control (p < 0.05). Phenobarbital decreased both epileptiform activity and autonomic dysfunction. Changes in autonomic function induced by hippocampal penicillin were similar to that induced by the intravenous administration of pentylenetetrazol. 25.3.9â•…Intraosseous Drugs in a Cardiac Arrest Model during Resuscitation in Anesthetized Swine Obtaining venous access is a difficult problem faced by a physician caring for the pediÂ� atric€patient in cardiac arrest. The intraosseous route (through the bone via an 18-gauge spinal needle placed in the right proximal tibia) was established as a rapid and effective alternative for venous access in a cardiac arrest model (Spivey et al. 1985). Domestic swine (15–26 kg) were anesthetized with ketamine 20 mg/kg (i.m.) and alpha-chloralose 25 mg/ kg (i.v.), given gallamine (4 mg/kg, i.v.) to prevent muscle fasciculations, and ventilated with a respirator. Catheters were placed in the right ventricle, left ventricle, and femoral arteries for mean arterial blood pressure recordings and blood pH sampling every 2 min. Ventricular ἀbrillation and cardiac arrest was induced by endocardial stimulation with a Grass S88 stimulator. Five-minute postarrest resuscitation was initiated with a mechanical resuscitator. Ten-minute postarrest sodium bicarbonate 1 mEq/kg was administered by the peripheral intravenous, central, or intraosseous route. Blood pH was sampled every 2€min for 30 min from the right ventricle, left ventricle, and femoral artery. Analysis of variance revealed central and intraosseous routes were signiἀcantly different (p < 0.05) from the peripheral group. All three groups were signiἀcantly different (p < 0.05) from control. Pathology studies revealed only minor damage to bone when sodium bicarbonate was administered intraosseously (Lathers et al. 1989c). In a different series of experiments using the same methods, ventricular ἀbrillation was induced in anesthetized domestic swine (Lathers et al. 1989b). After the cardiac arrest was induced, postarrest resuscitation€was initiated 5 min later, and at 10 min after arrest sodium bicarbonate 1 mEq/kg was€given via the peripheral intravenous, central, or intraosseous route. Controls did not receive bicarbonate. Catecholamine samples were taken from the femoral artery every 2€min. Two-way analysis of variance did not reveal any difference in mean arterial pressures in the four groups. In all groups blood pH from the femoral artery demonstrated a respiratory alkalosis, peaking at pH 7.48 5 min after initiation of mechanical resuscitation. In groups receiving sodium bicarbonate, respiratory alkalosis peaked at pH 7.77 ± 0.09 central and pH 7.65 ± 0.06 peripheral 2 min postinfusion and at pH 7.71 ± 0.06 intraosseous 8 min postinfusion. Analysis of variance revealed the central and intraosseous routes were signiἀcantly different (p < 0.05) from the peripheral group. All three groups were different (p < 0.05) from control. Plasma epinephrine and norepinephrine concentrations at 0, 6, 10, 12, 20, and 30 min postarrest in the control group were: 3 and 10, 94 and 327, 119 and 329, 92 and 234, 33 and 135, and 127 and 62 ng/mL, respectively. All three groups receiving sodium bicarbonate demonstrated similar patterns and were not different from control. The value of the rapid onset of the intraosseous route versus the peripheral route was conἀrmed. Today, the use of the intraosseous route as an alternative venous access for drug

380 Sudden Death in Epilepsy: Forensic and Clinical Issues

administration has greatly increased and is used in every emergency room in the United States. Modiἀcation of this cardiac arrest/resuscitation intraosseous model in anesthetized swine by adding EEG recording electrodes was also studied in Dr. Lathers’ laboratory. The studies examined the effect of anticonvulsant drugs administered via the intraosseous route for rapid treatment/control of convulsing pediatric patients in the emergency room when a traditional intravenous line cannot be established (see discussion below). 25.3.10â•…Intraosseous Drug Administration of Anticonvulsant Epilepsy Drugs in Anesthetized Swine Intravenous access for AED administration may be difficult and time-consuming in an actively seizing infant. The intraosseous route was established to be a rapid and effective alternative for diazepam administration (Spivey et al. 1987). Domestic swine were anesthetized with ketamine (20 mg/kg, i.m.) and alpha-chloralose (80 mg/kg, i.v.) and given gallamine (4 mg/kg, i.v.) to prevent muscle fasciculations. Tracheostomies were performed and the animals were ventilated with a respirator. The left femoral vein was cannulated and pentylenetetrazol (100 mg/kg) was injected to elicit epileptogenic activity in swine that had undergone craniotomies for electrocortical recording. Diazepam (0.1 mg/kg) was administered intravenously or intraosseously; control animals received no drug. Epileptogenic activity was suppressed below control levels within 1 min in the intravenous group and within 2 min in the intraosseous group. Two-way analysis of variance did not show a signiἀcant difference between the intravenous and intraosseous routes. Both were different (p€< 0.05) when compared to control. There was no signiἀcant difference in plasma diazepam levels between the two groups at 1, 2, 5, 10, 15, and 20 min. The validity of the use of the intraosseous route to control seizures in pediatric patients when an intravenous line cannot be established was established by this study. dl -Propranolol, a beta-adrenoceptor antagonist, exhibits antiepileptic activity in various animal seizure models. Another study assessed the efficacy of intraosseous propranÂ�olol in suppressing pentylenetetrazol-induced seizure activity in pigs (Jim et al. 1989). Domestic swine were prepared for recordings of arterial blood pressure, ECG, and electrocortical activity. Seizure activity was induced by pentylenetetrazol (100 mg/kg, i.v.). Sixty seconds after onset of seizure activity, animals either received no drug (control) or propranÂ�olol (intravenous or intraosseous). A transient increase (16.3–50.0%) in the mean arterial blood pressure was observed following pentylenetetrazol administration. Both intraosseous and intravenous propranolol signiἀcantly suppressed seizure duration (s/min interval) 1€min following drug administration; seizure duration control, 36.3 ± 4.8; intravenous propranÂ� olol, 12.3 ± 5.1; intraosseous propranolol, 18.3 ± 6.0. Intravenous and intraosseous propranolol produced a maximal decrease of 32% to 38% in basal heart rate and reduced the transient increase in mean arterial blood pressure elicited by pentylenetetrazol, with no signiἀcant effect on basal mean arterial blood pressure. Another series compared the effects of propranolol with diazepam (Lathers and Sarvey 1989). Sixty seconds after onset of epileptogenic activity, the animals received saline or diazepam (0.1 mg/kg) or propranÂ� olol (2.5 mg/kg) intravenously or via the intraosseous route. Both diazepam and propranÂ� olol were effective in suppressing epileptogenic activity via intravenous or intraosseous routes. Thus, the intraosseous route is a rapid and effective alternative route for administration of AEDs when an intravenous route cannot be readily established. In a third series of experiments by Jim et al. (1989), 60 s after onset of seizure activity,€animals received no

Sudden Death: Animal Models to Study Nervous System

381

drug (control)€or€lorazepam (1.0 mg/kg) administered intravenous or intraosseous. Both intraosseous and intravenous lorazepam signiἀcantly suppressed duration of seizure activity (s/min interval) within 1 min following administration: duration of seizure activity control, 46.2 ± 3.6; intravenous lorazepam, 25.0 ± 5.1; intraosseous (18-gauge) lorazepam, 27.6 ± 6.0; intraosseous (13-gauge) lorazepam, 24.0 ± 2.4. Seizure activity was essentially abolished at 1 min following lorazepam infusion. Both intravenous and intraosseous lorazÂ� epam did not have signiἀcant effects on basal heart rate and mean arterial blood pressure. The data demonstrates that in swine the intraosseous route is an effective alternative venous access for lorazepam administration and the size of the spinal needles used did not affect the antiepileptic efficacy of the intraosseous infusion of lorazepam. 25.3.11â•…Sympathetic Nerve Sprouting and Electrical Remodeling in Dogs: Mechanisms of Sudden Cardiac Death A positive correlation between nerve density and a clinical history of ventricular arrhythmia has been found in studies of cardiac nerves in explanted native hearts of transplant recipients. Sympathetic stimulation is important in the generation of sudden cardiac death. Chen et al. (2001) hypothesized that myocardial infarction results in nerve injury, followed by sympathetic nerve sprouting and regional (heterogeneous) myocardial hyperinnervation. Coupling between augmented sympathetic nerve sprouting with electrically remodeled myocardium results in ventricular tachycardia, ventricular ἀbrillation, and sudden cardiac death. If nerve sprouting can be modiἀed, arrhythmia control may result. Nerve growth factor augmented myocardial sympathetic nerve sprouting. This hypothesis was based on studies in dogs with complete atrio-ventricular block and myocardial infarction in which the nerve growth factor infusion to the left stellate ganglion facilitated development of ventricular tachycardia, ventricular ἀbrillation, and sudden cardiac death. Electrical stimulation is a method to elicit nerve sprouting (Swissa et al. 2004). Subthreshold electrical stimulation of the left stellate ganglia induced cardiac nerve sprouting and sympathetic hyperinnervation and facilitated development of this canine model of ventricular arrhythmia and sudden cardiac death. 25.3.12â•…Left and Right Stellate Ganglion Infusions in Dogs with Myocardial Infarction and Complete Heart Block In vivo the stellate ganglia and/or left stellate nerve activity become target sites for drug action and/or disease. A high incidence of sudden cardiac death occurs in dogs with myocardial infarction, complete AV block, and nerve growth factor infusion into the left stellate ganglion (Zhou et al. 2001). Nerve growth factor infusion into the right stellate ganglion is antiarrhythmic, in contrast to the proarrhythmic effect observed when infused into the left stellate ganglion (Zhou et al. 2005). The effect on the QT intervals was studied after osmotic pump infusion of nerve growth factor over a 5-week period to the stellate ganglia (Zhou et al. 2001). Heart rhythm and QT and R–R intervals were monitored via implantable cardiac deἀbrillator ECG recordings. A time-dependent increase of QTc intervals occurred in the left stellate ganglia animals and a time-dependent decrease of QTc intervals in the right stellate ganglia dogs. Nerve growth factor infusion to the left and right stellate ganglia causes left and right ventricular sympathetic nerve sprouting and hyperinnervation, respectively. NGF infusion to the left stellate ganglion in dogs with myocardial infarct and

382 Sudden Death in Epilepsy: Forensic and Clinical Issues

complete AV block increase QT interval and incidence of ventricular tachycardia, ventricular ἀbrillation, and sudden cardiac death while infusion to the right stellate ganglion shortened the QT interval and reduced the incidence of ventricular tachycardia. The QT interval prolongation was causally related to the occurrence of ventricular arrhythmia in dogs with nerve sprouting, myocardial infarction, and complete AV block. Nerve growth factor infusion into the left stellate ganglia signiἀcantly increased beta (3)-AR immunoreactivity in dogs and signiἀcantly decreased beta (3)-AR immunoreactivity when infused into the right stellate ganglia when control dogs were compared with dogs with myocardial infarction and complete AV block (Zhou et al. 2005). Thus nerve growth factor infusion into the right versus the left stellate ganglia induced differential beta-adrenoceptor expression in the left ventricular myocardium. 25.3.13â•…Spontaneous Stellate Ganglion Nerve Activity and Ventricular Arrhythmia and/or Sudden Death in a Canine Model Simultaneous continuous long-term recordings of left stellate ganglion nerve activity in dog with nerve growth factor infusion to the left stellate ganglion, AV block, and myocardial infarction included low-amplitude burst discharge activity and high-amplitude spike discharge activity and both were associated with an increase in heart rate (Zhou et al. 2008). Most ventricular tachycardia (86.3%) and sudden death were preceded within 15€s by either of the two types of neural discharge. The closer the time to onset of ventricular tachycardia, the more neural discharge occurred. Much of the high-amplitude neural discharge was followed by ventricular arrhythmia (21%) or by changes in the QRS morphology (65%). Actual electrical stimulation of the left stellate ganglia increased transmural heterogeneity of repolarization (T peak-end intervals) and induced either ventricular tachycardia or ἀbrillation. Increase synaptogenesis and nerve sprouting was conἀrmed after nerve growth factor infusion to the left stellate ganglion. Experiments are needed in which in vivo animal models of sudden death are used and spontaneous stellate ganglion nerve activity is monitored to correlate changes in arrhythmias and/or the occurrence of sudden death. Lathers et al. (Lathers and Schraeder 1987; Stauffer et al. 1989) reported changes in the burst patterns of postganglionic cardiac sympathetic neural burst discharge patterns accelerated with changes in EEG and/or arrhythmias and/or death and termed this the lockstep phenomenon. Similar changes seem to occur in spontaneous discharge of the stellate ganglion nerves. 25.3.14â•… Cardiac Remodeling Post Myocardial Infarction in Mice Another experimental model of myocardial infarction uses mice undergoing ligation of the left coronary artery and then treated for 30 days postinfarction with vehicle or with a transforming growth factor-beta, a key cytokine that both initiates and terminates tissue repair (Ellmers et al. 2008). Sustained production of transforming growth factor-beta underlies development of tissue ἀbrosis, particularly after myocardial infarction. This progressive ventricular remodeling results in a deterioration of cardiac function. An orally active speciἀc inhibitor of the transforming growth factor-beta receptor 1 (SD-208) was demonstrated to reduce mean arterial pressure. It also produces a trend for reduced ventricular and renal gene expression of transforming growth factor-beta activated kinase-1, a downstream modulator of transforming growth factor-beat signaling. It caused a

Sudden Death: Animal Models to Study Nervous System

383

signiἀcant decrease in collagen 1 and a marked decrease in cardiac mass. Decreased circulating levels of plasma renin activity and downregulating components of cardiac and renal renin–angiotensin system (angiotensinogen, angiotensin converting enzyme and angiotensin II type I receptor) occurred. The data indicates that block of the transforming growth factor-beta signaling pathway did produce a signiἀcant amelioration of the deleterious cardiac remodeling postinfarction. Application of this mouse model for sudden death allows one the option to use genetically altered mice to study various molecular aspects of sudden death. In addition, application of simultaneous EEG monitoring will reveal associated changes in epileptogenic activity and may provide a better understanding of cardiac changes in persons with epilepsy who have previously experienced arrhythmias and whose hearts are found on autopsy to exhibit increased mass. 25.3.15â•…Digenic Mouse Model: Combination of Two Epilepsy Genes to Mask Epilepsy Glasscock et al. (2007) combined two epilepsy-associated ion channel mutations with mutually opposing excitability defects and overlapping subcellular localization to generate a digenic mouse mode of human idiopathic epilepsy. Increasing membrane excitability occurred by removing Shaker-like K+ channel, which are encoded by the Kcna1 gene. This masked the absence of epilepsy caused by a P/Q-type Ca(2+) channelopathy due to a missense mutation in the Cacna1a gene. A decrease in the network excitability via impairment of the function of Cacna1a Ca(2+) channels attenuated limbic seizures and sudden death in Kcna1-null mice. The authors recommend additional experiments be done to improve the accuracy of genetic risk assessment of this complex disease. 25.3.16â•… Primate Models for Seizure and SUDEP That primates exhibit seizures has been reported as early as 1918 by Sherrington for the adult Macaca fulignosus (CS 1918). Autopsy revealed no gross lesions but histological studies were not done. Reports of epilepsy are reported to occur in macaca, red, and patas monkeys, and in lemurs and baboons. Emotion or noise may trigger the response. Epileptic symptoms may be observed in monkeys, less frequently in chimpanzees and orangutans, and can be Jacksonian or generalized tonic–clonic seizures (van Bogaert and Innes 1962). A few unexpected deaths may occur in convulsive seizures and have been reported for monkeys in which seizures had not been observed previously during life. Autopsy often revealed purulent or nonpurulent meningoencephalitis. Cases of unexplained cachexia may be associated with rare epileptic seizures and lymphocytic meningoencephalitis. Some cases of clinical epilepsy may also exhibit paraplegia not explained by pathologic ἀndings. Epilepsies in monkeys do occur after cranial fractures. Meningoencephalitic scars of variable localization and age may be noted in some monkeys. In monkeys that die from acute clinical disease, it may be impossible to relate age or signiἀcance of the location of meninoencephalitic scars. The exact trigger factors for spontaneous epilepsies in monkeys or apes are not known. Very recently, Szabo et al. (2009) examined causes of death in 46 epileptic baboons and 78 nonepileptic controls because the baboon is a model of primary generalized epilepsy. A complete pathologic examination was conducted at the Southwest Foundation for Biomedical Research in San Antonio. Baboons with seizures died at a younger age than

384 Sudden Death in Epilepsy: Forensic and Clinical Issues

controls (p < 0.001). Almost all of the epileptic baboons that died suddenly without an apparent cause were found to have pulmonary congestion or edema but not evidence of trauma, systemic illness, or heart disease when compared to nine control baboons (12%) (p < 0.001). Most of the control baboons demonstrated evidence of a concurrent illness. Serosanginous bronchial secretions were found in 15 of the seizure baboons (58%), but only in three controls (4%) (p < 0.001). Chronic multifocal ἀbrotic changes in myocardium were noted in three (12%) of the seizure baboons and in one control baboon. The authors conclude that untreated seizures appear to reduce the life expectancy of captive baboons and that sudden unexpected death in epilepsy may be a common cause of natural death in epileptic baboons. When speculating on the results of the autopsies, one must be very careful to note that the ἀndings of pulmonary congestion and/or edema may still have occurred secondarily to the occurrence of a cardiac arrhythmia that would not be detected upon autopsy. Thus one may not 100% conclude that respiratory changes were the initial trigger for the death event. Furthermore, the ἀnding of multifocal ἀbrotic changes in the myocardium of three of the seizure baboons suggests that cardiac arrhythmias may indeed have occurred. A future study with a larger number of baboons is needed. One may speculate that the stress of captivity may exert a negative effect on some of the baboons to a greater effect than on others. This question could be answered by conducting a comparative study of these animals with those found living in the wild. Nevertheless, to be able to conduct a postmortem examination of these baboons does helps us to focus on mammals in the species phylum very close to humans. It will be of help to expand the number of animals in the study of Szabo et al. (2009). 25.3.17â•…Other Animal Models Due to space limitations, other animal models have not been discussed. For example, Goodman et al. (1990) have studied the acute cardiovascular response during kindled seizures in rats (Goodman et al. 1990, 2010). The reader is presented with a few ἀnal, abbreviated facts about different models and study ἀndings and is encouraged to examine this information when designing laboratory studies to delineate mechanisms and prevention of arrhythmias and/or sudden death in persons with epilepsy. One unresolved question is the role of serotonin in the initiation of arrhythmia and sudden death. The actions of serotonin (5-hydroxytryptamine) are multifactoral and appear to vary in different parts of the central nervous system. The effect of serotonin on excitatory postsynaptic currents has been studied using whole-cell recordings from superἀcial dorsal horn neurons in neonatal rat spinal cord slices (Hori et al. 1996). It was concluded that serotonin can induce a longlasting facilitation of evoked excitatory postsynaptic currents and spontaneous release of excitatory transmitter at superἀcial dorsal horn synapses of the rat spinal cord. This effect appears to be modulated by the protein kinase C inhibitor, calphostin C, which appears to link to the serotonin 2 receptor and to directly activate the exocytotic machinery. Wholecell patch-clamp recordings from projection neurons and interneurons of the rat basolateral amygdala have been done to examine the effects of serotonergic modulation of neurotransmission. Rainnie (1999) concluded acute serotonin release directly activates GABAergic interneurons of the basolateral almygdala, via activation of serotonin 2 receptors and increases the frequency of inhibitory synaptic events in projection neurons. Chronic serotonin release, or high levels of serotonin, reduce the excitatory drive onto interneurons

Sudden Death: Animal Models to Study Nervous System

385

and may act as a feedback mechanism to prevent excess inhibition within the nucleus. The entorinal cortex is thought to be involved in the generation of temporal lobe epilepsy (Deng and Lei 2008). The entorhinal cortex receives serotonergic innervations from the raphe nuclei in the brain stem. Cellular and molecular mechanisms involved in facilitation of GABAergic transmission and depression of epileptic activity in the superἀcial layers of the entorhinal cortex were studied. Serotonin was found to increase GABA release in rat entorhinal cortex by inhibiting interneuron TASK-3 K+ channels. The authors concluded that serotonin mediated depression of neuronal excitability and increased GABA release contribute to its antiepileptic effects in the entorhinal cortex. Studies are needed to identify not only the role of serotonin in sudden death but also other transmitters such as GABA, catecholamines, neuropeptides, and so forth. Tupal and Faingold (2006) and Faingold et al. (2010, Chapter 41) have published evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Future studies are needed to identify sites of action for disease and/or pharmacological interventions. The above discussion, although not all inclusive, presents in vitro and in vitro models that can be used to study mechanisms of action for initiation of sudden death and physiological sites for AEDs or antirrhythmic drugs to act to prevent sudden death. The discussion provides insights into potential in vitro and in vivo models, which when used in conjunction with molecular methods and genetic animal models, will ultimately lead to a better understanding of sudden death and how to prevent sudden death with selected use of current or yet-to-be developed pharmacological agents. Study of genetically designed mice with congenital and acquired long QT syndrome will prove a better understanding of sudden cardiac death originating from ventricular arrhythmogenesis, one of the major causes of mortality in the developed world (Killeen et al. 2008). Readers are also referred to additional relevant animal models found in published papers and in this book. Speciἀcally, Bealer et al. (2010) has examined chronic alterations in cardiac sympathovagal balance induced by status epilepticus in rats (Metcalf et al. 2009). Saito et al. (2006) discuss repeatable focal seizure suppression using a rat preparation to examine the effect of seizure activity after urethane anesthesia and reversible carotid artery occlusion. McCloskey et al. (2006) used stereological methods to examine the robust size and stability of ectopic hilar granule cells after pilocarpine-induced status epilepticus in rats. Discussing the neurobiology of epilepsy, Scharfman (2007) notes, using temporal lobe epilepsy as an example, that genes, developmental mechanisms, and neuronal plasticity play a major role in creating a state of underlying hyperexcitability. However, the critical control points or emergence of chronic seizures in temporal lobe epilepsy and their persistence, frequency, and severity have yet to be clearly understood. Misonou and coworkers (2004) have examined the biochemical regulation of ion channel localization and phosphorylation by neuronal activity in the kainite model of continuous seizures in rats as it relates to excitatory neurotransmission and the intrinsic excitability of pyramidal neurons. Nutritional and lifestyle factors [including factors such as stress (Lathers and Schraeder 2006)] have been suggested to play a role in sudden death (Scorza et al. 2008). Lathers et al. (2008) proposed mechanistic factors in SUDEP, listing risk categories of arrhythmogenic factors, respiratory factors and hypoxia, and psychological factors and mechanisms for risks associated with each category. Please see Tables 1.1 through 1.8 in Chapter€1. Scorza et al. (2008) noted clariἀcation of risk factors and establishment of the mechanism of SUDEP are important to establish pÂ�reventative measures for SUDEP. They emphasized the need to strive for full seizure cÂ�ontrol and the importance of encouraging patients with epilepsy

386 Sudden Death in Epilepsy: Forensic and Clinical Issues

worldwide to receive nonmedical, lifestyle-modifying interventions that have generally accepted public health beneἀts, even though there is as yet no consensus that they may or may not prevent sudden death. Both animal studies and clinical studies are needed to deἀnitely address risk factors of omega-3 fatty acids, cold temperatures, exercise, and heart rate to the development of cardiac arrhythmias and/or sudden death in persons with epilepsy. The reader is referred to Lathers et al. (2008), Scorza et al. (2008), and Lathers et al. (2010) for an in-depth discussion of risk factors for SUDEP. Prospective studies of patients need to be done to determine how we can identify which persons with epilepsy are at risk for SUDEP.

25.4â•… Conclusions Valuable insights are obtained from both in vitro and in vivo animal models, just as valuable data is obtained from using animal models in many different species to examine a clinical problem observed in humans. When both in vitro and in vivo models are used, one must evaluate the data obtained from an in vitro model and then compare results obtained from in vivo models to clearly understand how to interpret the composite data. Use of multiple animal models will shed light on mechanisms and risks associated with the problem of sudden unexplained death in persons with epilepsy by revealing physiological sites of actions for drugs to act. Use of multiple animal models will also demonstrate beneἀcial effects of pharmacological interventions. Researchers must use the best combination of research tools available to them to elucidate the mechanisms of risk for sudden death whether in persons with epilepsy or in persons with heart disease. Of course, one must realize that, in general, no one animal model for disease completely mimics the symptoms and ἀndings of the disease state in humans. When possible, risk factors and all supporting medical information must be obtained for persons with epilepsy or persons with cardiac disease that are thought to be potential candidates for sudden death in an effort to prevent the sudden death event.

References Agress C. M., M. J. Rosenberg, H. I. Jacobs, M. J. Binder A. Schneiderman, and W. G. Clark. 1952. Protracted shock in the closed-chest dog following coronary remobilization with graded microspheres. Am J Physiol 170: 536–549. Alkadhi, K. A., K. H. Alzoubi, and A. M. Aleisa. 2005. Plasticity of synaptic transmission in autonomic ganglia. Prog Neurobiol 75 (2): 83–108. Alkadhi, K. A., and K. H. Alzoubi. 2007. Role of long-term potentiation of sympathetic ganglia (gLTP) in hypertension. Clin Exp Hypertens 29: 267–286. Alkadhi, K. A., and R. J. McIsaac. 1973. Non-nicotinic transmission during ganglionic block with chlorisondamine and nicotine. Eur J Pharmacol 24 (1): 78–85. Alkadhi, K. A., and R. J. McIsaac. 1974. Effect of preganglionic nerve stimulation on sensitivity of the superior cervical ganglion to nicotinic blocking agents. Br J Pharmacol 51 (4): 533–539. Alkadhi, K. A., R. S. Al-Hijailan, and K. H. Alzoubi. 2008. Long-term depression in the superior cervical ganglion of the rat. Brain Res 1234: 25–31. Alzoubi, K. H., A. M. Aleisa, and K. A. Alkadhi. 2008. Expression of gLTP in sympathetic ganglia of obese Zucker rats in vivo: Molecular evidence. J Mol Neurosci 35 (3): 297–306.

Sudden Death: Animal Models to Study Nervous System

387

Alzoubi, K. H., A. M. Aleisa, and K. A. Alkadhi. 2010. In vivo expression of ganglionic long-term potentiation in superior cervical ganglia from hypertensive aged rats. Neurobiol Aging 31 (5): 805–812. Armstrong, D. M., J. C. Eccles, R. J. Harvey, and P. B. Matthews. 1968. Responses in the dorsal accessory olive of the cat to stimulation of hind limb afferents. J Physiol 194 (1): 125–145. Asai, M., E. Talavera, A. Massarini, M. Zubieta, and O. Vindrola. 1994. Valproic acid-induced rapid changes of met-enkephalin levels in rat brain. Probable association with abstinence behavior and anticonvulsant activity. Neuropeptides 27 (3): 203–210. Barlow, C. H., and B. Chance. 1976. Ischemic areas in perfused rat hearts: Measurement by NADH fluorescence photography. Science 193 (4256): 909–910. Baroldi, G. O., O. Mantero, and G. Scomazzoni. 1956. The collaterals of the coronary arteries in the normal pathologic hearts. Circ Res 4: 223–229. Baumstark, A. E., D. C. Levin, and M. C. Fishbein. 1978. Experimental myocardial infarction in dogs with normal coronary arteries. Angiographic resolution of coronary arterial emboli. Radiology 128 (1): 31–35. Bealer, S. L., C. S. Metcalf, J. G. Little, M. Vatta, A. Brewster, and A. Anderson. 2010. Sympathetic nervous system dysregulation of cardiac function and myocyte potassium channel remodeling in rodent seizure models: Candidate mechanisms for SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Bissett, J. K., J. W. Watson, J. A. Scovil, N. Schmidt, J. R. McConnell, and J. Kane. 1979. Changes in myocardial refractory periods following ischemia in the porcine heart. J Electrocardiol 12 (1): 35–40. Bland, J. H., and E. Lowenstein. 1976. Halothane-induced decrease in experimental myocardial isÂ�chemia in the non-failing canine heart. Anesthesiology 45 (3): 287–293. Bonte, F. J., R. W. Parkey, K. D. Graham, J. Moore, and E. M. Stokely. 1974. A new method for radionuclide imaging of myocardial infarcts. Radiology 110 (2): 473–474. Bramham, C. R., K. Bacher-Svendsen, and J. M. Sarvey. 1997. LTP in the lateral perforant path is beta-adrenergic receptor-dependent. Neuroreport 8 (3): 719–724. Brown, J. T., J. C. Richardson, G. L. Collingridge, A. D. Randall, and C. H. Davies. 2005. Synaptic transmission and synchronous activity is disrupted in hippocampal slices taken from aged TAS10 mice. Hippocampus 15 (1): 110–117. Burgard, E. C., and J. M. Sarvey. 1991. Long-lasting potentiation and epileptiform activity produced by GABAB receptor activation in the dentate gyrus of rat hippocampal slice. J Neurosci 11 (5): 1198–1209. Busuttil, R. W., W. J. George, and R. L. Hewitt. 1975. Protective effect of methylprednisolone on the heart during ischemic arrest. J Thorac Cardiovasc Surg 70 (6): 955–965. Ceremuzynski, L., J. Staszewska-Barczak, and K. Herbaczynska-Cedro. 1969. Cardiac rhythm disturbances and the release of catecholamines after acute coronary occlusion in dogs. Cardiovasc Res 3 (2): 190–197. Charles, C. J., D. L. Jardine, M. G. Nicholls, M. T. Rademaker, and A. M. Richards. 2008. Urocortin 1 exhibits potent inhibition of cardiac sympathetic nerve activity in conscious sheep. J Hypertens 26 (1): 53–60. Chen, P. S., L. S. Chen, J. M. Cao, B. Shariἀ, H. S. Karagueuzian, and M. C. Fishbein. 2001. Sympathetic nerve sprouting, electrical remodeling and the mechanisms of sudden cardiac death. Cardiovasc Res 50 (2): 409–416. Cohnheim, J, and A. von Schultess-Rechberg. 1981. Uber de folgen der Kranzarterienschliessung fur das Herz. Virchow’s Arch Pathol Anat 85: 503. Colugnati, D. B., P. A. Gomes, R. M. Arida, M. de Albuquerque, R. M. Cysneiros, E. A. Cavalheiro, and F. A. Scorza. 2005. Analysis of cardiac parameters in animals with epilepsy: Possible cause of sudden death? Arq Neuropsiquiatr 63 (4): 1035–1041.

388 Sudden Death in Epilepsy: Forensic and Clinical Issues Constantin, L. 1963. Extracardiac factors contributing to the hypotension during coronary occlusion. Am J Cardiol 11: 205–217. Corday, E., M. K. Heng, S. Meerbaum, T. W. Lang, J. C. Farcot, J. Osher, and K. Hashimoto. 1977. Derangements of myocardial metabolism preceding onset of ventricular ἀbrillation after coronary occlusion. Am J Cardiol 39 (6): 880–889. Sherrington, C. S. 1918. Stimulation of the motor cortex in a monkey subject to epileptiform seizures. Brain 41: 48–49. Dahl, D., and J. M. Sarvey. 1989. Norepinephrine induces pathway-speciἀc long-lasting potentiation and depression in the hippocampal dentate gyrus. Proc Natl Acad Sci U S A 86 (12): 4776–4780. Deng, P. Y., and S. Lei. 2008. Serotonin increases GABA release in rat entorhinal cortex by inhibiting interneuron TASK-3 K+ channels. Mol Cell Neurosci 39 (2): 273–284. Eccles, J. C., I. Rosen, P. Scheid, and H. Taborikova. 1975a. The differential effect of cooling on responses of cerebellar cortex. J Physiol 249 (1): 119–138. Eccles, J. C., P. Scheid, and H. Taborikova. 1975b. Responses of red nucleus neurons to antidromic and synaptic activation. J Neurophysiol 38 (4): 947–964. Eccles, J. C., R. A. Nicoll, W. F. Schwarz, H. Taborikova, and T. J. Willey. 1975c. Reticulospinal neurons with and without monosynaptic inputs from cerebellar nuclei. J Neurophysiol 38 (3): 513–530. Eccles, J. C., T. Rantucci, P. Scheid, and H. Taborikova. 1975d. Somatotopic studies on red nucleus: Spinal projection level and respective receptive ἀelds. J Neurophysiol 38 (4): 965–980. Ellmers, L. J., N. J. Scott, S. Medicherla, A. P. Pilbrow, P. G. Bridgman, T. G. Yandle, A. M. Richards, A. A. Protter, and V. A. Cameron. 2008. Transforming growth factor-beta blockade downÂ�regulates the renin–angiotensin system and modiἀes cardiac remodeling after myocardial infarction. Endocrinology 149 (11): 5828–5834. Faingold, C. L., S. Tupal, Y. Mhaskar, and V. Uteshev. 2010. DBA mice as models of sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 41. Boca Raton: CRC Press. Gillis, R. A. 1971. Role of the nervous system in the arrhythmias produced by coronary occlusion in the cat. Am Heart J 81 (5): 677–684. Gillis, R. A., F. H. Levine, H. Thibodeaux, A. Raines, and F. G. Standaert. 1973. Comparison of methyllidocaine and lidocaine on arrhythmias produced by coronary occlusion in the dog. Circulation 47 (4): 697–703. Gillis, R. A., P. B. Corr, D. G. Pace, D. E. Evans, J. DiMicco, and D. L. Pearle. 1976. Role of the nervous system in experimentally induced arrhythmias. Cardiology 61 (1): 37–49. Glasscock, E., J. Qian, J. W. Yoo, and J. L. Noebels. 2007. Masking epilepsy by combining two epilepsy genes. Nat Neurosci 10 (12): 1554–1558. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1990. Acute cardiovascular response during kindled seizures. In Epilepsy and Sudden Death. ed. C. M. Lathers and P. Schraeder, Chapter 11. New York, NY: Marcel Dekker. Goodman J. H., R. W. Homan, and I. L. Crawford. 2010. Acute cardiovascular response during kindled seizures. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 40. Boca Raton: CRC Press. Grover, G. J., and R. Singh. 2007. The isolated, perfused pseudo-working heart model. Methods Mol Med 139: 145–150. Guzman, S. V., E. Swenson, and M. Jones. 1962. Intercoronary reflex. Demonstration by coronary angiography. Circ Res 10: 739–745. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Harris, A. S. 1950. Delayed development of ventricular ectopic rhythms following experimental coronary occlusion. Circulation 1 (6): 1318–1328. Harris, A. S., H. Otero, and A. J. Bocage. 1971. The induction of arrhythmias by sympathetic activity before and after occlusion of a coronary artery in the canine heart. J Electrocardiol 4 (1): 34–43.

Sudden Death: Animal Models to Study Nervous System

389

Hashimoto, K., T. Tsukada, H. Matsuda, Y. Nakagawa, and S. Imai. 1979. Antiarrhythmic effects of bupuranolol against canine ventricular arrhythmias induced by halothane-adrenaline or twostage coronary ligation. J Cardiovasc Pharmacol 1 (2): 205–217. Heng, M. K., R. M. Norris, T. Peter, H. D. Nisbet, and B. N. Singh. 1978. The effect of glucose-insulinpotassium on experimental myocardial infarction in the dog. Cardiovasc Res 12 (7): 429–435. Higgins, C. B., M. Sovak, W. Schmidt, and P. T. Siemers. 1979. Differential accumulation of radiopaque contrast material in acute myocardial infarction. Am J Cardiol 43 (1): 47–51. Hori, Y., K. Endo, and T. Takahashi. 1996. Long-lasting synaptic facilitation induced by serotonin in superἀcial dorsal horn neurones of the rat spinal cord. J Physiol 492 (Pt 3): 867–876. Jardine, D. L., C. J. Charles, C. M. Frampton, and A. M. Richards. 2007. Cardiac sympathetic nerve activity and ventricular ἀbrillation during acute myocardial infarction in a conscious sheep model. Am J Physiol Heart Circ Physiol 293 (1): H433–H439. Jardine, D. L., C. J. Charles, R. K. Ashton, S. I. Bennett, M. Whitehead, C. M. Frampton, and M. G. Nicholls. 2005. Increased cardiac sympathetic nerve activity following acute myocardial infarction in a sheep model. J Physiol 565 (Pt. 1): 325–333. Jim, K. F., C. M. Lathers, V. L. Farris, L. F. Pratt, and W. H. Spivey. 1989. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous lorazepam in pigs. Epilepsia 30 (4): 480–486. Jobe, C. L. 1967. Selection and development of animal models of myocardial infarction. Paper read at the Symposium on Animal Models for Biomedical Research, July 10, 1967, Dallas, TX. Jugdutt, B. I., L. C. Becker, and G. M. Hutchins. 1979. Early changes in collateral blood flow during myocardial infarction in conscious dogs. Am J Physiol 237 (3): H371–H380. Khaw, B. A., G. A. Beller, and E. Haber. 1978a. Experimental myocardial infarct imaging following intravenous administration of iodine-131 labeled antibody (Fab′)2 fragments speciἀc for cardiac myosin. Circulation 57 (4): 743–750. Khaw, B. A., H. K. Gold, R. C. Leinbach, J. T. Fallon, W. Strauss, G. M. Pohost, and E. Haber. 1978b. Early imaging of experimental myocardial infarction by intracoronary administration of 131Ilabelled anticardiac myosin (Fab′)2 fragments. Circulation 58 (6): 1137–1142. Killeen, M. J., G. Thomas, I. N. Sabir, A. A. Grace, and C. L. Huang. 2008. Mouse models of human arrhythmia syndromes. Acta Physiol (Oxf) 192 (4): 455–469. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. Kupersmith, J. 1976. Antiarrhythmic drugs: Changing concepts. Am J Cardiol 38 (1): 119–121. Lathers, C. M. 1975. Effect of practolol and stalol on ouabrain-induced nonuniform adrenergic nerve activity. Pharmacologist 34: 745. Lathers, C. M. 1979. The effect of methylprednisolone on autonomic neural discharge associated with acute coronary occlusion in the cat. Pharmacologist 21: 201. Lathers, C. M. 1980a. Effect of timolol on autonomic neural discharge associated with ouabaininduced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106. Lathers, C. M. 1980b. The effect of metoprolol on coronary occlusion-induced arrhythmia and autonomic neural discharge. Fed Proc 39: 771. Lathers, C. M. 1981. Induced disease. Myocardial infarction in dogs and cats. In Mammalian Models for Research on Aging, 224–228. Washington, DC: National Academy Press. Lathers, C. M., and J. M. Sarvey. 1989. Effect of the beta blocking agent propranolol on longer term potentiation induced by norepinephrine in rat hippocampal slices. Research conducted in Dr. Sarvey’s laboratory. Lathers, C. M., G. J. Kelliher, J. Roberts, A. J. Teres, and A. B. Beasley. 1976. Effect of practolol and coronary anatomy on occlusion-induced arrhythmia and the associated nonuniform neural discharge. Clin Res 24: 617. Lathers, C. M., and P. L. Schraeder. 1990. Arrhythmias associated with epileptogenic activity elicited by penicillin. Chapter 9. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker.

390 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders. ed. L. Sher, Chapter 13. Hauppauge, NY: Nova Science. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies. Epilepsy Behav 14: 573–576. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin, J. C. Krusz, G. J. Kelliher, and J. Roberts. 1981. The action of reserpine, 6-hydroxydopamine, and bretylium on digitalis-induced cardiotoxicity. Eur J Pharmacol 76 (4): 371–379. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin, J. C. Krusz, G. J. Kelliher, P. B. Goldberg, and J. Roberts. 1982. Role of the adrenergic nerve terminal in digitalis-induced cardiac toxicity: A study of the effects of pharmacological and surgical denervation. J Cardiovasc Pharmacol 4 (1): 91–98. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., A. Z. Stauffer, N. Tumer, C. M. Kraras, and B. D. Goldman. 1989a. Anticonvulsant and antiarrhythmic actions of the beta blocking agent timolol. Epilepsy Res 4 (1): 42–54. Lathers, C. M., K. F. Jim, W. B. High, W. H. Spivey, W. D. Matthews, and T. Ho. 1989b. An investigation of the pathological and physiological effects of intraosseous sodium bicarbonate in pigs. J€Clin Pharmacol 29 (4): 354–359. Lathers, C. M., K. F. Jim, and W. H. Spivey. 1989c. A comparison of intraosseous and intravenous routes of administration for antiseizure agents. Epilepsia 30 (4): 472–479. Lathers, C. M., and L. J. Lipka. 1986. Chlorpromazine: Cardiac arrhythmogenicity in the cat. Life Sci 38 (6): 521–538. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital on autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33 (9): 837–844. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., R. F. Flax, and L. J. Lipka. 1986. The effect of C1 spinal cord transection or bilateral adrenal vein ligation on thioridazine-induced arrhythmia and death in the cat. J Clin Pharmacol 26 (7): 515–523. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988a. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28 (8): 736–745. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988b. The effect of intracerebroventricular d-ALA2 methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43 (26): 2287–2298. Lathers, C. M., and C. M. Smith. 1976. Ethanol effects on muscle spindle afferent activity and spinal reflexes. J Pharmacol Exp Ther 197 (1): 126–134. Lea, IV, P. M., and J. M. Sarvey. 2003. Modulation of epileptiform burst frequency by the metabotropic glutamate receptor subtype mGluR3. Epilepsy Res 53 (3): 207–215.

Sudden Death: Animal Models to Study Nervous System

391

Levitt, B., N. Cagin, J. Kleid, J. Somberg, and R. Gillis. 1976. Role of the nervous system in the genesis of cardiac rhythm disorders. Am J Cardiol 37 (7): 1111–1113. Libby, P., P. R. Maroko, J. W. Covell, C. I. Malloch, J. Ross Jr., and E. Braunwald. 1973. Effect of practolol on the extent of myocardial ischaemic injury after experimental coronary occlusion and its effects of ventricular function in the normal and ischaemic heart. Cardiovasc Res 7 (2): 167–173. Lipka, L. J. 1987. Mechanisms involved in the production of cardiac arrhythmias by chlorpromazine and other psychoactive agents. Ph.D. Thesis. Research conducted in Dr. Lathers’ Laboratory. Pharmacology, Medical College of Pennsylvania, Philadelphia, PA. Lipka, L. J., C. M. Lathers, and J. Roberts. 1988. Does chlorpromazine produce cardiac arrhythmia via the central nervous system? J Clin Pharmacol 28 (11): 968–983. Lipka, L. J., and C. M. Lathers. 1987. Psychoactive agents, seizure production, and sudden death in epilepsy. J Clin Pharmacol 27 (3): 169–183. Lown, B., and M. Wolf. 1971. Approaches to sudden death from coronary heart disease. Circulation 44 (1): 130–142. Lown, B., R. L. Verrier, and S. H. Rabinowitz. 1977. Neural and psychologic mechanisms and the problem of sudden cardiac death. Am J Cardiol 39 (6): 890–902. Lucchesi, B. R., W. E. Burmeister, T. E. Lomas, and G. D. Abrams. 1976. Ischemic changes in the canine heart as affected by the dimethyl quaternary analog of propranolol, UM-272 (SC-27761). J Pharmacol Exp Ther 199 (2): 310–328. Malliani, A., P. J. Schwartz, and A. Zanchetti. 1969. A sympathetic reflex elicited by experimental coronary occlusion. Am J Physiol 217 (3): 703–709. Maroko, P. R., P. Libby, C. M. Bloor, B. E. Sobel, and E. Braunwald. 1972. Reduction by hyaluronidase of myocardial necrosis following coronary artery occlusion. Circulation. 46 (3): 430–437. Marshall, R. J., and J. R. Parratt. 1974. Proceedings: The effects of practolol in the early stages of experimental myocardial infarction. Br J Pharmacol 52 (1): 124. McCloskey, D. P., T. M. Hintz, J. P. Pierce, and H. E. Scharfman. 2006. Stereological methods reveal the robust size and stability of ectopic hilar granule cells after pilocarpine-induced status epilepticus in the adult rat. Eur J Neurosci 24 (8): 2203–2210. McIsaac, R. J. 1978. Post-tetanic enhancement of stimulus-induced muscarinic after discharge in the rat superior cervical ganglion. J Pharmacol Exp Ther 207 (1): 72–82. Meltzer, R. S., J. N. Woythaler, A. J. Buda, J. C. Griffin, W. D. Harrison, R. P. Martin, D. C. Harrison, and R. L. Popp. 1979. Two dimensional echocardiographic quantiἀcation of infarct size alteration by pharmacologic agents. Am J Cardiol 44 (2): 257–262. Metcalf, C. S., P. B. Radwanski, and S. L. Bealer. 2009. Status epilepticus produces chronic alterations in cardiac sympathovagal balance. Epilepsia 50 (4): 747–754. Misonou, H., D. P. Mohapatra, E. W. Park, V. Leung, D. Zhen, K. Misonou, A. E. Anderson, and J. S. Trimmer. 2004. Regulation of ion channel localization and phosphorylation by neuronal activity. Nat Neurosci 7 (7): 711–718. Nachlas, M. M., and T. K. Shnitka. 1963. Macroscopic identiἀcation of early myocardial infarcts by alterations in dehydrogenase activity. Am J Pathol 42: 379–405. Mortimer, M. L., N. Tumer, M. D. Johnson, and J. Roberts. 1991. Effect of age on presynaptic beta2 receptor mediated responses in the rat heart. Mech Ageing Dev 59 (1–2): 17–25. Paterson, D. S. 2010. Medullary serotonergic abnormalities in sudden infant death syndrome: Implications in SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 5. Boca Raton: CRC Press. Pickworth, W. B., J. Gerard-Ciminara, and C. M. Lathers. 1990. Stress, arrhythmias, and seizures. Chapter 22. In Epilepsy and Sudden Death. ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Poisson, D., M. O. Christen, and F. Sannajust. 2000. Protective effects of I(1)-antihypertensive agent moxonidine against neurogenic cardiac arrhythmias in halothane-anesthetized rabbits. J€Pharmacol Exp Ther 293 (3): 929–938.

392 Sudden Death in Epilepsy: Forensic and Clinical Issues Powell, Jr., W. J., D. R. DiBona, J. Flores, and A. Leaf. 1976. The protective effect of hyperosmotic mannitol in myocardial ischemia and necrosis. Circulation 54 (4): 603–615. Rainnie, D. G. 1999. Serotonergic modulation of neurotransmission in the rat basolateral amygdala. J Neurophysiol 82 (1): 69–85. Reichelt, M. E., L. Willems, B. A. Hack, J. N. Peart, and J. P. Headrick. 2009. Cardiac and coronary function in the Langendorff-perfused mouse heart model. Exp Physiol 94 (1): 54–70. Reimer, K. A., M. M. Rasmussen, and R. B. Jennings. 1973. Reduction by propranolol of myocardial necrosis following temporary coronary artery occlusion in dogs. Circ Res 33 (3): 353–363. Reimer, K. A., and R. B. Jennings. 1979. The changing anatomic reference base of evolving myocardial infarction. Underestimation of myocardial collateral blood flow and overestimation of experimental anatomic infarct size due to tissue edema, hemorrhage and acute inflammation. Circulation 60 (4): 866–876. Reynolds, R. D., G. J. Kelliher, D. M. Ritchie, J. Roberts, and A. B. Beasley. 1979. Comparison of the arrhythmogenic effect of myocardial infarction in the cat and dog. Cardiovasc Res 13 (3): 152–159. Ritchie, D. M., G. J. Kelliher, A. MaLathersillan, W. Fasolak, J. Roberts, and S. Mansukhani. 1979. The cat as a model for myocardial infarction. Cardiovasc Res 13 (4): 199–206. Roberts, J., M. L. Mortimer, P. J. Ryan, M. D. Johnson, and N. Tumer. 1990. Role of calcium in adrenergic neurochemical transmission in the aging heart. J Pharmacol Exp Ther 253 (3): 957–964. Roswell, H. C., J. F. Mustard, M. A. Packham, and W. J. Dodds. 1965. The hemostatic mechanism and its role in cardiovascular disease of swine. Paper read at The Swine in Biomedical Research, July 19–22, Richland, WA. Rotman, M., G. S. Wagner, and A. G. Wallace. 1972. Bradyarrhythmias in acute myocardial infarction. Circulation 45 (3): 703–722. Saito, T., K. Sakamoto, K. Koizumi, and M. Stewart. 2006. Repeatable focal seizure suppression: A rat preparation to study consequences of seizure activity based on urethane anesthesia and reversible carotid artery occlusion. J Neurosci Methods 155 (2): 241–250. Salazar, A. E. 1961. Experimental myocardial infarction. Induction of coronary thrombosis in the intact closed-chest dog. Circ Res 9: 1351–1356. Sarvey, J. M., E. C. Burgard, and G. Decker. 1989. Long-term potentiation: Studies in the hippocampal slice. J Neurosci Methods 28 (1–2): 109–124. Scharfman, H. E. 2007. The neurobiology of epilepsy. Curr Neurol Neurosci Rep 7 (4): 348–354. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Schraeder, P. L., E. L. So, and C. M. Lathers. 2010. Forensic case identiἀcation. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 6. Boca Raton: CRC Press. Schwartz, P. J., H. L. Stone, and A. M. Brown. 1976. Effects of unilateral stellate ganglion blockade on the arrhythmias associated with coronary occlusion. Am Heart J 92 (5): 589–599. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13 (1): 263–264; author reply 265. Shell, W. E., J. K. Kjekshus, and B. E. Sobel. 1971. Quantitative assessment of the extent of myocardial infarction in the conscious dog by means of analysis of serial changes in serum creatine phosphokinase activity. J Clin Invest 50 (12): 2614–2625. Siemers, P. T., C. B. Higgins, W. Schmidt, W. Ashburn, and P. Hagan. 1978. Detection, quantitation and contrast enhancement of myocardial infarction utilizing computerized axial tomography: Comparison with histochemical staining and 99mTc-pyrophosphate imaging. Invest Radiol 13€(2): 103–109.

Sudden Death: Animal Models to Study Nervous System

393

Sivam, S. P., and S. D. Seth. 1978. Metoprolol—A new cardioselective beta adrenoceptor antagonist in experimental cardiac arrhythmias. Indian J Med Res 68: 176–182. Skrzypiec-Spring, M., B. Grotthus, A. Szelag, and R. Schulz. 2007. Isolated heart perfusion according to Langendorff—Still viable in the new millennium. J Pharmacol Toxicol Methods 55 (2): 113–126. Snyder, D. L., E. Gao, M. D. Johnson, and J. Roberts. 1994. Decrease in norepinephrine release from cardiac adrenergic nerve terminals after ischemia and reperfusion. Ann N Y Acad Sci 723: 389–391. Spath, Jr., J. A., D. L. Lane, and A. M. Lefer. 1974. Protective action of methylprednisolone on the myocardium during experimental myocardial ischemia in the cat. Circ Res 35 (1): 44–51. Spivey, W. H., and C. M. Lathers. 1985. Effect of timolol on the sympathetic nervous system in coronary occlusion in cats. Ann Emerg Med 14 (10): 939–944. Spivey, W. H., C. M. Lathers, D. R. Malone, H. D. Unger, S. Bhat, R. N. McNamara, J. Schoffstall, and N. Tumer. 1985. Comparison of intraosseous, central, and peripheral routes of sodium bicarbonate administration during CPR in pigs. Ann Emerg Med 14 (12): 1135–1140. Spivey, W. H., H. D. Unger, C. M. Lathers, and R. M. McNamara. 1987. Intraosseous diazepam suppression of pentylenetetrazol-induced epileptogenic activity in pigs. Ann Emerg Med 16 (2): 156–159. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Stauffer, AZ, J. Dodd-O, and C. M. Lathers. 1990. Relationship of the lockstep phenomenon and precipitous changes in blood pressure. Chapter 14. In Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Swissa, M., S. Zhou, I. Gonzalez-Gomez, C. M. Chang, A. C. Lai, A. W. Cates, M. C. Fishbein, H.€S. Karagueuzian, P. S. Chen, and L. S. Chen. 2004. Long-term subthreshold electrical stimulation of the left stellate ganglion and a canine model of sudden cardiac death. J Am Coll Cardiol 43€(5): 858–864. Szabo, C. A., K. D. Knape, M. M. Leland, J. Feldman, K. J. McCoy, G. B. Hubbard, and J. T. Williams. 2009. Mortality in captive baboons with seizures: A new model for SUDEP? Epilepsia 50 (8): 1995–1998. Thakur, M. L., A. Gottschalk, and B. L. Zaret. 1979. Imaging experimental myocardial infarction with indium-111-labeled autologous leukocytes: Effects of infarct age and residual regional myocardial blood flow. Circulation 60 (2): 297–305. Tumer, N., M. L. Mortimer, and J. Roberts. 1992. Gender differences in the effect of age on adrenergic neurotransmission in the heart. Exp Gerontol 27 (3): 301–307. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. van Bogaert, L., and Innes J. R. M. 1962. Neurologic diseases of apes and monkeys. Chapter IV. In Comparative Neuropathology, ed. J. R. M. Innes and L. Z. Saunders. New York, NY: Academic Press. Vasilets, L. A., V. P. Mokh, and B. I. Khodorov. 1989. Prevention of reperfusion-induced arrhythmia by Na+/H+ exchange block. Kardiologiia 29 (2): 91–94. Vindrola, O., M. Asai, M. Zubieta, and G. Linares. 1983. Brain content of immunoreactive [Leu5] enkephalin and [Met5]enkephalin after pentylenetetrazol-induced convulsions. Eur J Pharmacol 90 (1): 85–89. Webb, S. W., A. A. Adgey, and J. F. Pantridge. 1972. Autonomic disturbance at onset of acute myocardial infarction. Br Med J 3 (5818): 89–92. Wehrmacher, W. H., J. V. Talano, M. P. Kaye, and W. C. Randall. 1979. The unbalanced heart. Animal models of cardiac dysrhythmias. Cardiology 64 (2): 65–74. Wichmann, J., R. Loser, H. P. Diemer, and W. Lochner. 1978. Pharmacological alterations of coronary collateral circulation: Implication to the steal-phenomenon. Pflugers Arch 373 (3): 219–224.

394 Sudden Death in Epilepsy: Forensic and Clinical Issues Willerson, J. T., R. W. Parkey, L. M. Buja, and F. J. Bonte. 1979. Technetium-99m stannous pyrophosphate ‘hot spot’ imaging to detect acute myocardial infarcts. Cardiovasc Clin 10 (2): 139–148. Yukimura, T., G. Stock, H. Stumpf, T. Unger, and D. Ganten. 1981. Effects of [dAla2]-methionineenkephalin on blood pressure, heart rate, and baroreceptor reflex sensitivity in conscious cats. Hypertension 3 (5): 528–533. Zhou, S., B. C. Jung, A. Y. Tan, V. Q. Trang, G. Gholmieh, S. W. Han, S. F. Lin, M. C. Fishbein, P.€S. Chen, and L. S. Chen. 2008. Spontaneous stellate ganglion nerve activity and ventricular arrhythmia in a canine model of sudden death. Heart Rhythm 5 (1): 131–139. Zhou, S., J. M. Cao, Z. D. Tebb, T. Ohara, H. L. Huang, C. Omichi, M. H. Lee, et al. 2001. Modulation of QT interval by cardiac sympathetic nerve sprouting and the mechanisms of ventricular arrhythmia in a canine model of sudden cardiac death. J Cardiovasc Electrophysiol 12 (9): 1068–1073. Zhou, S., O. Paz, J. M. Cao, K. Asotra, N. N. Chai, C. Wang, L. S. Chen, M. C. Fishbein, B. Shariἀ, and P. S. Chen. 2005. Differential beta-adrenoceptor expression induced by nerve growth factor infusion into the canine right and left stellate ganglia. Heart Rhythm 2 (12): 1347–1355. Zipes, D. P. 2008. Heart-brain interactions in cardiac arrhythmias: Role of the autonomic nervous system. Cleve Clin J Med 75 (Suppl. 2): S94–S96.

Synaptic Plasticity of Autonomic Ganglia Role of Chronic Stress and Implication in Cardiovascular Diseases and Sudden Death

26

Karim A. Alkadhi Karem H. Alzoubi

Contents 26.1 Introduction 26.2 Activity-Dependent Ganglionic Long-Term Potentiation of the Nicotinic Pathway 26.3 Characteristics of Ganglionic Long-Term Potentiation 26.4 Role of Ganglionic Long-Term Potentiation in the Pathophysiologies of the Cardiovascular System 26.5 Blood Pressure and Long-Term Potentiation of Sympathetic Ganglia 26.6 Summary and Concluding Remarks References

395 396 397 404 405 415 416

26.1â•…Introduction Activity-dependent long-lasting potentiation in neural tissues originates from studies using sympathetic ganglia obtained from rodents and cats (Volle 1966; Dunant and Dolivo 1968). However, it took nearly a decade after the term “long-term potentiation” was coined by Bliss and Lomo (1973) in the brain hippocampal formation to describe a similar long-term potentiation in the superior cervical sympathetic ganglia (Brown and McAfee 1982; Briggs et al. 1985b). Later, ganglionic long-term potentiation was demonstrated in sympathetic ganglia from various animal species in vitro and in situ (Brown and McAfee 1982; Briggs et al. 1985a, 1985b; Briggs and McAfee 1988; Briggs et al. 1988; Alonso-deFlorida et al. 1991; Bachoo and Polosa 1992; Weinreich et al. 1995). Another decade later, the existence of a very similar longterm potentiation in the avian parasympathetic ciliary ganglion was reported (Scott and Bennett 1993b). In 1996, serotonin was identiἀed as the neurotransmitter required for induction of long-term potentiation in the rat superior cervical ganglion (Alkadhi et al. 1996). In addition to principal neurons, sympathetic ganglia contain small intensely fluorescent cells that may contain dopamine, serotonin (Egan et al. 1987), substance P, endogenous opiates, or histamine (Eranko et al. 1986 for review). The function of the small intensely fluorescent cells is not well understood, but they are generally considered as interneurons because some of these cells have been identiἀed as receiving synaptic input and making synaptic connections with principal neurons (Case and Matthews 1986). The major synaptic 395

396 Sudden Death in Epilepsy: Forensic and Clinical Issues

transmission pathway in sympathetic ganglia takes place through activation of nicotinic acetylcholine receptors. A multiplicity of preganglionic and postganglionic receptors to different neurotransmitters and neuromodulators has been identiἀed in the ganglia. These receptors, which include muscarinic cholinergic, serotonergic, adrenergic, dopaminergic, histaminergic, angiotensin, and opiate receptors, are perhaps involved in modulation of ganglionic transmission. Repetitive, relatively high-frequency stimulation of the preganglionic nerve results in two dissimilar and unrelated responses—an initial, transient large increase in ganglionic response, lasting 3–4 min, known as posttetanic potentiation (Martin and Pilar 1964; Magleby and Zengel 1975; Alkadhi et al. 1996). This is followed by ganglionic long-term potentiation—a smaller, stable increase in synaptic strength with a duration of up to 3 h (Bennett 1994; Lin and Bennett 1994; Brain and Bennett 1995; Alkadhi et al. 1996). The expression of ganglionic long-term potentiation may be the culmination of a set of events, arising both in the postsynaptic and presynaptic regions, and involving several enzymes, modulators, and second messengers. In the central nervous system, long-term potentiation is widely believed to be a cellular correlate of memory; however, the function of ganglionic long-term potentiation is unclear. It may be that repetitive activity invariably results in marked potentiation of synaptic transmission in any neural circuit. However, expression of long-term potentiation in autonomic ganglia as a result of hyperactivity of the CNS can result in unwanted pathophysiological changes in the cardiovascular system.

26.2╅Activity-Dependent Ganglionic Long-Term Potentiation of the Nicotinic Pathway The ganglionic long-term potentiation of the nicotinic pathway has been reported in different autonomic ganglia from various animal species (Table 26.1). Ganglionic long-term potentiation can be evoked by a brief period of high-frequency stimulation (20€Hz for 20 s) of the preganglionic nerve and can be measured in vitro by intracellular or extracellular recording techniques (Briggs and McAfee 1988). Additionally, ganglionic long-term potentiation has been measured in ganglia in anesthetized animals in situ (Alonso-deFlorida et al. 1991; Bachoo et al. 1992; Bachoo and Polosa 1992). Table 26.1╅ Ganglionic Long-Term Potentiation of the Nicotinic Pathway as Reported in Various Animal Species Animal Species

Speciἀc Ganglia

Rat

Superior cervical ganglion

Cat Guinea pig

Superior cervical, lumbar, and stellate ganglia Superior cervical ganglion

Chick Bullfrog

Parasympathetic ciliary ganglion Sympathetic ganglia

Reference Studies Brown and McAfee 1982; Briggs and McAfee 1988; Alkadhi et al. 1996; Alzoubi et al. 2004; Alkadhi et al. 2005a, 2005b; Alkadhi and Alzoubi 2007; Alzoubi et al. 2008a, 2008b, 2010; Alzoubi and Alkadhi 2009 Alonso-deFlorida et al. 1991; Bachoo and Polosa 1991; Morales et al. 1994 Weinreich et al. 1995; Cavalcante de Albuquerque et al. 1996 Scott and Bennett 1993b Koyano et al. 1985; Kuba and Kumamoto 1986; Kumamoto and Kuba 1986; Minota et al. 1991

Synaptic Plasticity of Autonomic Ganglia

397

26.3â•… Characteristics of Ganglionic Long-Term Potentiation The involvement of adrenergic, cholinergic, dopaminergic, or adenosine receptors in the induction of ganglionic long-term potentiation has been ruled out (Briggs et al. 1985a, 1985b; Bachoo and Polosa 1991; Scott and Bennett 1993a, 1993b; Alkadhi et al. 1996; Hogan et al. 1998). Recent evidence strongly suggests that long-term potentiation in the rat sympathetic ganglia is strictly dependent on activation of serotonin 5-HT3 receptor (Figure 26.1, Table 26.2; Alkadhi et al. 1996, 2001b; Gerges et al. 2002; Alkadhi et al. 2005a, Table 26.2â•… Effects of Various Serotonergic Drugs on Compound Action Potential Compound Action Potential

Seratonergic Drugs Serotonin (10–20 μM) Serotonin (30 μM and above) Fluoxetine (10 μM) 8-Hydroxydipropylaminotetralin (5 μM) R-(+)-Dimethoxy-4iodoamphetamine (1 μM) Ketanserin (3 μM) 1-m-(Chlorophenyl)biguanide (1€μM) Tropisetron (5 µM) Ondansetron (5 µM) 3-Tropanyl-3,5-dichlorobenzoate (MDL 72222, 0.5 μM) m-CPBG (1 μM) + MDL 72222 (0 5 μM) Control 1-m-(Chlorophenyl)biguanide (1€μM) R-(+)-Dimethoxy-4iodoamphetamine (1 μM)

Pharmacological Mode of Action

Posttetanic Potentiation

Untreated Ganglia Agonist No change Agonist No change Serotonin uptake No change inhibitor Serotonin 5-HT1A No change receptor agonist Serotonin 5-HT2 No change receptor agonist Serotonin 5-HT2 No change receptor antagonist Serotonin 5-HT3 No change receptor agonist Serotonin 5-HT3 No change receptor antagonist Serotonin 5-HT3 No change receptor antagonist Serotonin 5-HT3 No change receptor antagonist As above No change Reserpine Pretreated No change Serotonin 5-HT3 receptor agonist Serotonin 5-HT2 receptor agonist

Ganglionic Long-Term Potentiation Induced In Vitro

Ganglionic Long-Term Potentiation Induced In Vivo

Increase No change Increase

Not tested Not tested Not tested

Reduce

Not tested

Increase

Not tested

Increase

Not tested

Increase

Not tested

Reduce

Reduce

Reduce

Reduce

Reduce

Reduce

Increase

Not tested

Reduce

Reduce

No change

Increase

Not tested

No change

Reduce

Not tested

Sources: Alkadhi et al., J Physiol, 496 (Pt 2), 479–489, 1996; Alkadhi et al., Exp Biol Med (Maywood), 226: 1024– 1030; Gerges et al., Neuropharmacology, 43, 1070–1076, 2002; Alkadhi et al., Prog Neurobiol, 75, 83–108, 2005a; Alkadhi et al., Brain Res, 1234, 25–31, 2005b; Alkadhi, K., and Alzoubi, K., Clin Exp Hypertens, 29, 267–286, 2007; Alzoubi et al., J Mol Neurosci, 35, 297–306, 2008a; Alzoubi et al., J Mol Neurosci, 35, 201–209, 2008b; Alzoubi et al., Neurobiol Aging, 31, 805–812, 2010; Alkadhi et al., J Neurosci, 21, 3515– 3520, 2001b; Gerges et al., Neuropharmacology, 43, 1070–1076, 2002; Alkadhi et al., Neurobiol Dis, 20, 849–857, 2005b. With permission.

398 Sudden Death in Epilepsy: Forensic and Clinical Issues 160

% of Control spike

140

120

100 HFS 1

80

50

HFS 2

75

100

m-CPBG (1 µM)

125

150

175

200

Time (min)

Figure 26.1╇ Expression of ganglionic long-term potentiation requires a combination of highfrequency stimulation and activation of serotonin 5-HT3 receptor. High-frequency stimulation (arrowhead 1) of monoamine-depleted ganglia (from rats treated with reserpine). Rats were reserpinized with a single intraperitoneal dose of reserpine (3 mg/kg; given approximately 24 h before excision of ganglia) failed to induce ganglionic long-term potentiation but a second train (arrowhead 2) in the presence of the serotonin 5-HT3 receptor agonist m-chlorophenylbiguanide (m-CPBG), evoked a robust ganglionic long-term potentiation. Each point is the mean ± SEM of four ganglia from reserpine-treated animals. (From Alkadhi, K. and K. Alzoubi, Clin Exp Hypertens, 29, 267–286, 2007. With permission.)

2005b; Alkadhi and Alzoubi 2007; Alzoubi et al. 2008a, 2010). The site of origin of ganglionic long-term potentiation may be presynaptic, postsynaptic, or through a combination of presynaptic and postsynaptic mechanisms, depending on the animal species and/ or experimental methods used (Kumamoto and Kuba 1983; Koyano et al. 1985; Kuba and Kumamoto 1990). Both repetitive stimulation and serotonin are required for induction. High-frequency stimulation of preganglionic nerve is required, but not sufficient, for induction of ganglionic long-term potentiation. In addition to preganglionic nerve high-frequency stimulation, both induction and maintenance of ganglionic long-term potentiation require activation of serotonin 5-HT3 receptor, presumed to be released from an intraganglionic structure, probably in small intensely fluorescent cells (Alkadhi et al. 1996). Similarly, activation of serotonin 5-HT3 receptor is necessary but insufficient to evoke ganglionic longterm potentiation. Furthermore, several lines of evidence suggest that carbon monoxide is required for the induction phase (see Figure 26.2a; Alkadhi et al. 2001a) and nitric oxide

Synaptic Plasticity of Autonomic Ganglia

399

(a) 200

180

% of Control spike

160

140

120

100 ZnPP (10 µM) HFS 2 Carbon monoxide

HFS 1

(b)

80 50

75

100

125

150

175

200

Time (min)

200

180

% of Control spike

160 -NOARG (50 µM)

140

Wash

120

100 HFS 80 20

40

60

80 Time (min)

100

120

140

Figure 26.2╇ Carbon monoxide and nitric oxide are involved in the cascade of events leading to the expression of ganglionic long-term potentiation. (a) The first high-frequency stimulation (arrowhead 1) failed to induce ganglionic long-term potentiation in the presence of the heme oxygenase-II inhibitor ZnPP (to block production of CO), but when CO was bubbled into the solution a second high-frequency stimulation (arrowhead 2) readily induced ganglionic longterm potentiation. Each point is the mean ± SEM of five ganglia. (b) Inhibition of nitric oxide synthase with l-NOARG during the maintenance phase of ganglionic long-term potentiation caused reversible inhibition of ganglionic long-term potentiation. Each point is the mean ± SEM of four ganglia. (From Alkadhi, K. and K. Alzoubi, Clin Exp Hypertens, 29, 267–286, 2007. With permission.)

400 Sudden Death in Epilepsy: Forensic and Clinical Issues

for the maintenance phase of ganglionic long-term potentiation (Figure 26.2b; Alkadhi and Altememi 1997; Altememi and Alkadhi 1999). It appears that high-frequency stimulation causes the release of a number of messengers, including serotonin and carbon monoxide, which bring about the expression of ganglionic long-term potentiation. The exact function of serotonin 5-HT3 receptor in the expression of ganglionic long-term potentiation is not yet clear. It has been shown that activation of the presynaptic serotonin 5-HT3 receptor, which is known to be robustly permeable to calcium (Ronde and Nichols 1998), induces increases in intracellular calcium levels of individual nerve terminals in rat brain (Nichols and Mollard 1996). Thus, it is possible that the activation of this nonspeciἀc cationic channel–Â�receptor complex provides a conduit for a focused influx of calcium that elevates the intracellular levels of this ion to that required for activation of upstream enzymes including protein kinase C, calmodulin, and calcium-calmodulin kinase II that are essential for the expression of ganglionic long-term potentiation (Alzoubi et al. 2008b). Molecules in the signaling transduction pathways involved in the induction and maintenance of ganglionic long-term potentiation. The role of a number of putative intracellular signaling molecules necessary for the expression of long-term potentiation in various autonomic ganglia has been investigated using pharmacological and electrophysiological methods (reviewed by Alkadhi et al. 2005a). Only recently, however, has direct measurement of levels and activities of these molecules been carried out. Calmodulin and calcium-calmodulin–dependent kinase II. Given the dependence of ganglionic long-term potentiation induction and maintenance on calcium, it is reasonable to assume that calcium-calmodulin–dependent kinase II (CaMKII) is important for both phases of ganglionic long-term potentiation (Bennett 1994). Calmodulin inhibitors, trifluoperazine, calmidazolium, and W-7, block long-term potentiation induction in sympathetic ganglia (Minota et al. 1991; Alzoubi and Alkadhi 2009), without affecting the amplitude and the quantal content of fast excitatory postsynaptic potential (Minota et al. 1991). In the chick ciliary ganglia as well, W-7 blocks ganglionic long-term potentiation induction without affecting basal ganglionic transmission (Scott and Bennett 1993b). Calcium interacts with calmodulin to form a calcium–calmodulin complex. This complex activates CaMKII by phosphorylation, an essential process for long-term potentiation induction in ciliary ganglion (Scott and Bennett 1993b). In the rat superior cervical ganglia, Western blot analysis shows that the protein levels of calmodulin and phosphorylated and total CaMKII are markedly elevated when measured 1 h after induction of ganglionic long-term potentiation (Alzoubi et al. 2008b). Thus, it is likely that ganglionic long-term potentiation results from a sustained activation of CaMKII, which in turn might activate other target molecules involved in ganglionic long-term potentiation expression such as the neural form of nitric oxide synthase. Regulated by the calcium–calmodulin complex (Bredt and Snyder 1990; Sheng et al. 1992) via phosphorylation of CaMKII (Bredt and Snyder 1990; Nakane et al. 1991; Bredt et al. 1992), neural nitric oxide synthase protein levels are also found to be elevated 1 h after induction of ganglionic long-term potentiation (Alzoubi et al. 2008b). Nitric oxide. Nitric oxide, produced by neural nitric oxide synthase, is required for maintenance of long-term potentiation in the rat superior cervical ganglion and chick ciliary ganglion (Sheng et al. 1993; Lin and Bennett 1994; Alkadhi and Altememi 1997; Altememi and Alkadhi 1999). The existence of nitric oxide synthase in rat preganglionic neurons (Blottner and Baumgarten 1992; Grozdanovic et al. 1992; Valtschanoff et al. 1992; Wu and Dun 1995) and within preganglionic terminals of sympathetic ganglia is

Synaptic Plasticity of Autonomic Ganglia

401

well documented (Dun et al. 1993; Morris et al. 1993; Sheng et al. 1993; Saito et al. 1994; Okamura et al. 1995; Klimaschewski et al. 1996; Mazet et al. 1996). In rat superior cervical ganglia in which long-term potentiation was expressed by high-frequency stimulation, a signiἀcant elevation of neural nitric oxide synthase protein levels was detected (Alzoubi et al. 2008b). In the same preparation, nitric oxide synthase inhibitors l -Nâ•›G-nitroarginine (l -NOARG) and l -Nâ•›G-arginine methyl ester (l -NAME) completely blocked the expression of ganglionic long-term potentiation, without affecting posttetanic potentiation (Alkadhi and Altememi 1997; Altememi and Alkadhi 1999). Similarly, inhibitors of nitric oxide synthase blocked long-term potentiation in the chick ciliary ganglion (Scott and Bennett 1993a). The inactive d-enantiomers of those inhibitors had no effect on ganglionic long-term potentiation, and wash out of the nitric oxide synthase inhibitor l -NAME with a solution containing the nitric oxide precursor l -arginine reversed the inhibitory actions on ganglionic long-term potentiation, indicating the importance of nitric oxide for the maintenance phase of ganglionic long-term potentiation and not the induction phase (Altememi and Alkadhi 1999). This also indicates that there are different mechanisms for the induction phase and maintenance phase of ganglionic long-term potentiation (Alkadhi et al. 2001a). In rat superior cervical ganglion pretreated with sodium nitroprusside (a nitric oxide donor, thus bypassing nitric oxide synthase), l -NAME failed to inhibit ganglionic long-term potentiation, conἀrming the involvement of nitric oxide in ganglionic long-term potentiation (Altememi and Alkadhi 1999). In the rat superior cervical ganglion, the nitric oxide donor sodium nitroprusside produced prolonged potentiation of ganglionic transmission in the absence of high-frequency stimulation (Alkadhi et al. 2005a). When subjected to subsequent high-frequency stimulation, sodium nitroprusside–Â� potentiated ganglia failed to show additional enhancement, most likely due to saturation (Alkadhi et al. 2005a). Similar ἀndings using the nitric oxide donors, sodium azide and sodium nitroprusside, have been reported in the chick ciliary ganglia and rat superior cervical ganglion (Quenzer et al. 1980a, 1980b; Ariano et al. 1982; Ando et al. 1983; Sheng et al. 1992; Southam et al. 1996). Similarly, following potentiation of transmission by sodium nitroprusside in chick ciliary ganglia, high-Â�frequency stimulation failed to induce further potentiation, suggesting that sodium nitroprusside and high-frequency stimulation induce long-lasting potentiation through the same mechanism. In the chick ciliary ganglion, intracellular recording revealed that the nitric oxide synthase inhibitor, l -NAME applied 10–20 min before high-frequency stimulation, decreased the amplitude of excitaÂ� tory postsynaptic potential by 70%, while the nitric oxide donor sodium nitroprusside increased the amplitude of excitatory postsynaptic potential by 30% due to an increase in the quantal content (Lin and Bennett 1994). Cyclic nucleotide–protein kinase A pathway. Synaptic activity increases cyclic adenoÂ�sine monophosphate (cAMP) levels in sympathetic ganglia (McAfee et al. 1980). An increase in cAMP leads to activation of protein kinase A, which in turn phosphorylates target molecules, such as calcium-activated potassium channels, leading to increased calcium concentration in the presynaptic nerve terminal (Kuba et al. 1981; Kuba and Kumamoto 1990). cAMP-dependent protein kinase A has been identiἀed in ciliary ganglia (Lengyel et al. 1996). In the chick ciliary ganglion, 8-bromo-cAMP enhances synaptic transmission (Scott and Bennett 1993a). In agreement, we showed that treatment of rat superior cervical ganglion with 8-bromo-cAMP causes marked, gradual increase in ganglionic transmission (Alkadhi et al. 2005a). Subsequent high-frequency stimulation fails to evoke ganglionic long-term potentiation, suggesting that ganglionic long-term potentiation and 8-bromo-

402 Sudden Death in Epilepsy: Forensic and Clinical Issues

cAMP–induced potentiation likely share the same cellular cascade of events. Moreover, when 8-bromo-cAMP is superfused on rat superior cervical ganglion, it increases the magnitude of already expressed ganglionic long-term potentiation (Alkadhi et al. 2005a). Cyclic-guanine monophosphate (cGMP) is produced when nitric oxide activates guanylyl cyclase in the cascade of events leading to expression of ganglionic long-term potentiation. Activation of guanylyl cyclase leads to the accumulation of cGMP, which inhibits phosphodiesterase III (Beavo et al. 1994), thus preventing breakdown of cAMP (Maurice and Haslam 1990). The resulting elevated level of cAMP activates protein kinase A, which in turn increases the phosphorylation of a calcium-activated potassium channel in the nerve terminal membrane, leading to a reduction in the opening of the channel, accumulation of potassium ions, and consequent depolarization of the terminal (Cetiner and Bennett 1993). As a result, the incoming action potential duration will be enhanced, leading to increased calcium influx, and enhanced transmitter release and potentiation of ganglionic transmission (Scott and Bennett 1993a). This sequence of events is similar to that proposed for the effect of the serotonin on Aplysia sensory neurons, where serotonin causes an increase in the duration of nerve terminal action potential and subsequent potentiation of neurotransmitter release (Kandel and Hawkins 1992). Additionally, the guanylyl cyclase inhibitor, LY 83583 (5 µM) blocks basal transmission as well as induction of ganglionic long-term potentiation (Alzoubi and Alkadhi 2009). Treatment of rat superior cervical ganglion with the membrane-permeable 8-bromo-cGMP produces a gradual and prolonged enhancement of transmission that occludes subsequent high-frequency stimulation–induced ganglionic long-term potentiation (Alkadhi et al. 2005a). When superfused on rat superior cervical ganglion during the maintenance phase of established ganglionic long-term potentiation, 8-bromo-cGMP caused additional potentiation of ganglionic long-term potentiation (Alkadhi et al. 2005a). Treatment with 8-bromo-cGMP or 8-bromo-cAMP increased the quantal content of the excitatory postsynaptic potential without changing the quantal size in the chick ciliary ganglion (Lin and Bennett 1994). Protein kinase C–carbon monoxide pathway. Another player in the ganglionic longterm potentiation cascade is protein kinase C. Perfusion of rat superior cervical ganglion with the protein kinase C enhancer, phorbol esters, produced long-term potentiation of the nicotinic transmission that can be blocked with the protein kinase C antagonist, H-7 (Bachoo et al. 1992). Western blot analysis showed marked increases in the protein levels of protein kinase Cγ and protein kinase Cβ isoforms in rat superior cervical ganglion during the maintenance phase 1 h after induction of ganglionic long-term potentiation by highfrequency stimulation (Alzoubi et al. 2008b), indicating a possible role for protein kinase C in the expression of ganglionic long-term potentiation. Its has been proposed that as a consequence of activation of the calcium permeable serotonin 5-HT3 receptor–channel complex, calcium enters speciἀc synaptic regions where it activates protein kinase C, leading to the phosphorylation and activation of the enzyme heme oxygenase-2 (Dore et al. 1999), which produces carbon monoxide required for the induction of ganglionic longterm potentiation (Alkadhi et al. 2001a). Treatment of ganglia with either the carbon monoxide scavenger oxyhemoglobin or heme oxygenase-2 inhibitor, zinc protoporphyrin, completely and irreversibly prevented induction of ganglionic long-term potentiation (Alkadhi et al. 2001a). This strongly suggests that carbon monoxide is required for ganglionic long-term potentiation induction. Neither the posttetanic potentiation nor short-term potentiation is affected by treatment with oxyhemoglobin or zinc protoporphyrin (Alkadhi et al. 2005a), indicating that the

Synaptic Plasticity of Autonomic Ganglia

403

requirement for carbon monoxide is speciἀc only for the long-term potentiation form of synaptic plasticity in the ganglion (Alkadhi et al. 2001a). Furthermore, basal ganglionic transmission does not seem to require carbon monoxide, as neither zinc protoporphyrin nor exogenous carbon monoxide has any signiἀcant effect on the compound action potential (Alkadhi et al. 2001a). A similar retrograde messenger role for carbon monoxide has been suggested for the induction of long-term potentiation in the hippocampus (Stevens and Wang 1993; Zhuo et al. 1993; Ikegaya et al. 1994; Poss et al. 1995). The lack of effect of membrane-impermeable carbon monoxide scavenger, oxyhemoglobin, and heme oxygenase-2 inhibitor, zinc protoporphyrin, on the already established long-term potentiation in superior cervical ganglia, precludes the involvement of carbon monoxide in the maintenance phase of ganglionic long-term potentiation (Alkadhi et al. 2001a). This is consistent with the hypothesis that the role of carbon monoxide as a retrograde synaptic messenger would be limited to the period up to the end of the highfrequency stimulation required to induce ganglionic long-term potentiation. In support of the role of carbon monoxide, Western blot analysis shows no change in heme oxygenase-2 protein levels when measured 1 h after induction of ganglionic long-term potentiation by high-frequency stimulation (Alzoubi et al. 2008b). It is important to note that exogenous carbon monoxide in sympathetic ganglia (Alkadhi et al. 2001a), as in hippocampal slices (Zhuo et al. 1993), is necessary but not sufficient to evoke long-term potentiation; therefore, it must be accompanied by tetanic activation of the presynaptic nerve ἀbers to produce enhancement of synaptic transmission. In contrast, exogenous nitric oxide or nitric oxide donors applied to central nervous system or autonomic ganglia neurons produces prolonged enhancement of the synaptic transmission without applying presynaptic highfrequency stimulation (Bohme et al. 1991; Scott and Bennett 1993a; Southam et al. 1996). Other forms of synaptic plasticity in autonomic ganglia. Synaptic transmission in the autonomic ganglia, as in the central nervous system, may be modulated by a number of short-term and long-lasting processes, which may lead to a decrease or an increase in synaptic strength of varying durations. Multiple forms of synaptic plasticity are known to occur in autonomic ganglia, but they have not been sufficiently studied. In the superior cervical ganglion, in addition to posttetanic potentiation and ganglionic long-term potentiation, there is evidence for the presence of long-term depression and short-term potentiation. The characteristics of short-term potentiation and its distinction from ganglionic long-term potentiation have been reviewed (Alkadhi et al. 2005a). One study reported drug-induced long-lasting depression of transmission in the cat€ sympathetic superior cervical ganglion due to exposure to the opioid peptide met-╉ enkephalin. This long-lasting depression of transmission was blocked by naloxone or phorbol esters or by high-frequency stimulation of preganglionic nerve (Zhang et al. 1996). The depressant effect of either prolonged stimulation (40 Hz for 20 min) or phorbol esters is prevented by the protein kinase C inhibitor H-7, but not by the calmodulin inhibitor W-7 or the protein kinase A inhibitor HA 1004, indicating the involvement of protein kinase C in this form of long-term depression (Zhang et al. 1996). In addition, the expression of this long-term potentiation is calcium dependent and not affected by protein synthesis inhibitors (Zhang et al. 1996). Additionally, expression of this response is independent of acetylcholine since it can be evoked during complete blockade of muscarinic and nicotinic ganglionic transmission. We have recently published the ἀrst report of long-term depression in the isolated superior cervical ganglion of the rat (Alkadhi et al. 2008). In that model, low-frequency stimulation (3–5 Hz/15 min) of the preganglionic nerve produced

404 Sudden Death in Epilepsy: Forensic and Clinical Issues

a long-lasting, signiἀcant decrease in the amplitude of the extracellularly recorded postganglionic compound action potential. Pretreatment of ganglia with the serotonin 5-HT3 receptor antagonist tropisetron completely prevented the induction of ganglionic long-term potentiation. Moreover, treatment of ganglia with the serotonin 5-HT3 receptor antagonist MDL 72222 during the maintenance phase of established ganglionic long-term depression (1 h after low-frequency stimulation) antagonized the low-frequency stimulation–induced depression (Alkadhi et al. 2008). Inhibition of nitric oxide synthase with l -NOARG, applied before or after low-frequency stimulation, failed to affect the expression of ganglionic long-term potentiation (Alkadhi et al. 2008). Additionally, pretreatment with the protein synthesis inhibitor emetine totally prevented the expression of ganglionic long-term depression. However, inhibition of protein phosphatase with cantharidin did not interfere with the expression of ganglionic long-term depression. These results indicate the presence of long-term depression in the rat superior cervical ganglion and suggest that expression of this response involves activation of serotonin 5-HT3 receptor (Alkadhi et al. 2008).

26.4â•…Role of Ganglionic Long-Term Potentiation in the Pathophysiologies of the Cardiovascular System In vivo expression of long-term potentiation in autonomic ganglia is expected to enhance tonic efferent impulses to a multiplicity of neuroeffector organs, including the heart, blood vessels, and glands, which would modify the normal functions of these organs. Until recently, the consequences of expression of long-term potentiation in autonomic ganglia have not been studied. Enhanced activity of the sympathetic nervous system may be responsible for the development and/or aggravation of stress-induced hypertension. Autonomic dysregulation is also a well-established risk factor for cardiac arrhythmias, a major factor in the pathological mechanisms of sudden death. Cardiovascular centers in the brain stem are modulated by higher centers in the hippocampus, amygdala, and frontal lobe of the brain. This pathway has a forceful effect on autonomic functions (Nauta 1964; Naftel and Hardy 1997). Since stress perception involves frontal cortex participation, stress could alter the cardiovascular functions by influencing the autonomic regulatory systems (Szilagyi 1991). Recently, we presented evidence that links expression of long-term potentiation in sympathetic ganglia to the development or aggravation of hypertension in animal models. Continuous increase in sympathetic outflow to ganglia, induced by mental stress, could provide the repeated high-frequency presynaptic activity required for ganglionic longterm potentiation induction, which results in a sustained elevation of sympathetic tone to blood vessels, leading to or contributing to hypertension (Alkadhi et al. 2001b; Gerges et al. 2002; Alkadhi et al. 2005a, 2005b; Alkadhi and Alzoubi 2007; Alzoubi et al. 2008a, 2008b, 2010). This sustained elevation of sympathetic tone to blood vessels and the heart muscle can also lead to increased heart rate (clinically presents as palpitation) and/or arrhythmias leading to death. The most discussed mechanism for sudden unexplained death in epilepsy (SUDEP) is cardiac arrhythmia triggered by the intense activity of the central nervous system (i.e., seizure) channeled through the autonomic nervous system (Jehi and Najm 2008). Cardiac and respiratory dysfunctions have been described interictally. The increased incidence of SUDEP may be related to dysregulation in cardiac autonomic control. It is conceivable that repetitive efferent discharge from the CNS could trigger expression of long-lasting

Synaptic Plasticity of Autonomic Ganglia

405

ganglionic long-term potentiation that might eventually lead to cardiac arrhythmias, which could result in SUDEP.

26.5â•…Blood Pressure and Long-Term Potentiation of Sympathetic Ganglia Recently, the physiological or pathophysiological consequences of the expression of longterm potentiation in autonomic ganglia have been investigated. Like other forms of longterm potentiation in the central nervous system, ganglionic long-term potentiation is an enduring enhancement of the strength of synaptic transmission; therefore, in vivo expression of this form of synaptic plasticity in autonomic ganglia is expected to increase tonic efferent impulses to neuroeffector organs in the animals. Evidence that associates expression of ganglionic long-term potentiation with induction or aggravation of hypertension has been recently published (see Figures 26.3 and 26.5; Alkadhi et al. 2001b; Gerges et al. 2002; Alkadhi et al. 2005a, 2005b; Alkadhi and Alzoubi 2007; Alzoubi et al. 2008a, 2008b, 2010). Stress-induced hypertension. Hyperactivity of the sympathetic cardiovascular control is believed to contribute to hypertension in patients. Psychosocial stress is correlated with the onset and aggravation of ischemic heart disease, and is known to produce a greater increase in blood pressure in patients with labile hypertension than in normotensive subjects (Esler et al. 1977; Boone 1991; McEwen 1998). Although stress-induced hypertension normalizes within a few days of relieving the stress, prolonged mild–moderate hypertension may contribute to atherosclerotic cardiovascular diseases (Kannel et al. 1999) and

160

Stressed

Control

100 80

0

10

20

30

7

120

140 120

(b)

40

50

60

Days of stress

160

Systolic blood pressure (mm Hg)

Systolic blood pressure (mm Hg)

(a)

80 40 0

(c)

5 3 1

Stress Control

Male

Stress Control

Female

Male Female

Figure 26.3╇ Stress-induced hypertension. (a) Time-course of development of hypertension after initiation of the psychosocial stress procedure and reversal of hypertension on termination of stress. Blood pressure was measured by tail-cuff plethysmography. Each point is the mean ± SD from 10 stressed and 5 unstressed (control) male rats. Points between the two (*) are significantly (p < 0.05, t test) different than corresponding points of control rats. Some error lines are masked by signs. (b) Comparison of the magnitude of steady state stress-induced hypertension in male and female Wistar rats shows no significant difference (p > 0.05; unpaired t test) between the two groups. Each bar represents the mean ± SD from 10 stressed and 5 control male or female rats. (c) Male and female rats responded to stress at a similar rate. No significant (p > 0.05; unpaired t test) difference was seen in the number of days of stress that produced a steady state hypertension in male and female rats. Each bar is the mean ± SEM from 16 male and 10 female rats.

406 Sudden Death in Epilepsy: Forensic and Clinical Issues

cognitive impairment (Gerges et al. 2004; Aleisa et al. 2006). It has been demonstrated that both genetic and stress-induced experimental hypertension involve a signiἀcant neural component that contributes to the development and maintenance of this disorder (Mark 1996). Enhanced activity of the sympathetic nervous system may be responsible for the development and/or aggravation of stress-induced hypertension. Psychosocial stress is also associated with sustained hypertension and increased risk of coronary heart disease resulting from enhanced activation of the autonomic nervous system (Siegrist 2001). There is compelling evidence for the enhanced influence of the sympathetic nervous system on cardiovascular function in young mildly hypertensive humans (Egan et al. 1987). It is likely that stress-induced sustained increase in central sympathetic outflow to ganglia provides the repeated high-frequency presynaptic activity required to express long-term potentiation in sympathetic ganglia, which leads to a sustained increase in sympathetic tone to blood vessels, causing hypertension (see Figure 26.6; Alkadhi et al. 2005a; Alkadhi and Alzoubi 2007). It is uncertain what function, if any, stress-induced elevation of blood pressure serves. Perhaps it plays an important role in an adaptive mechanism to compensate for reduced blood perfusion reported in certain areas of the brain and other organs in stress situations (Shapiro et al. 2000; Ito et al. 2003). Regulation of cardiovascular function in the brain stem region is modulated by higher centers in the hippocampus, amygdala, and frontal lobe of the brain. This pathway is believed to have a profound effect on autonomic functions (Nauta 1964; Naftel and Hardy 1997). Since stress perception requires frontal cortex participation, stress may alter the cardiovascular functions by influencing the autonomic regulatory systems (Szilagyi 1991). The ability of the frontal cortex to alter the cardiovascular system response in stressed animals has been demonstrated where the blockade of the frontocortical brain stem pathways prevents the appearance of lethal arrhythmias in psychosocially stressed pigs (Skinner and Reed 1981). Animals develop reversible hypertension within 5–6 days of initiating psychosocial stress (see Figures 26.3 and 26.5a; Szilagyi 1991; Alkadhi et al. 2005b). There is no signiἀcant difference between male and female rats in the magnitude of the stress-induced increase in blood pressure or the number of days taken up by stress to induce steady-state hypertension (see Figures 26.3b and 26.3c; Alkadhi et al. 2005b). Pretreatment of rats with serotonin 5-HT3 receptor antagonists, which inhibit ganglionic long-term potentiation expression, prevents the development of hypertension (Alkadhi et al. 2005b). This indirectly implies that elevation of blood pressure by stress is a consequence of expression of long-€term potentiation in sympathetic ganglia. Additionally, electrophysiological evidence obtained in sympathetic ganglia excised from stressed rat provides further support for the concept that these ganglia have expressed ganglionic long-term potentiation in vivo (Figure€26.4). A shift of the input–output curve of stressed animals’ ganglia to the left side of that of unstressed normotensive controls (Figure 26.4a; Alkadhi et al. 2005b) suggests expression of long-term potentiation (Johnston and Wu 1995). The ganglionic long-term potentiation inhibitors (serotonin 5-HT3 receptor antagonists, e.g., ondansetron) reduce basal ganglionic transmission in ganglia from stress-hypertensive animal without affecting that of ganglia from normotensive controls, indicating the presence of long-term potentiation in ganglia from hypertensive animals (Figure 26.4b; Alkadhi et al. 2005b). In addition, high-frequency stimulation that evokes long-term potentiation in normal ganglia has no signiἀcant effect on ganglia excised from stress-hypertensive animals, indicating that in vivo expressed ganglionic long-term potentiation has occluded the effect of stimulation

(a)

0.15

0.00

180

Control Psych stress Swim stress

Control

2.5 5.0 7.5 Input (stimulus intensity)

(c)

*

160

100

0.10 0.05

(b)

110 Stressed

407

Spike amplitude, % of Baseline

0.20

Spike amplitude, % of Baseline

Output (spike amplitude, mv)

Synaptic Plasticity of Autonomic Ganglia

0.5 mV 20 ms

140

90

Control

*

120

80

100

Ondansetron (0.5 µM)

70

0

20 40 Time (min)

60

80 30

Stressed 40

60 50 Time (min)

70

Figure 26.4╇ Electrophysiological evidence of in vivo expression of gLTPlong-term potentia-

tion in stressed rat ganglia. (a) Input/output curves of synaptic transmission in ganglia isolated from stressed (open circles) and unstressed (control, closed squares) rats. Note the parallel left side shift of I/Oinput/output curves of ganglia from stressed rats indicating enhanced transmission. Each point in each group is the mean ± SEM form 6–7 male rats. (b) Ondansetron (serotonin 5-HT3 receptor antagonist) treatment (0.5 µM, black horizontal bar) decreased baseline compound action potential CAP of ganglia isolated from stressed (psychosocial or forced swim stress) but not of those isolated from control rats. (c) High frequency stimulation (HFS; 20€Hz/20€s, arrowhead) of the preganglionic nerves evoked robust long-term potentiation (gLTP) in ganglia excised from control rats. In a similar series of experiments, using identical protocol, HFS of preganglionic nerves of ganglia excised from stressed rats produced very small enhancement of transmission. Insets: records of CAP compound action potential of ganglia from stressed and control rats taken at times indicated on the graph. Calibrations bars apply to records of all groups. (From Alkadhi, K. and K. Alzoubi, Clin Exp Hypertens, 29, 267–286, 2007. With permission.)€

(Figure 26.4c; Alkadhi et al. 2005b; Alzoubi et al. 2008b). This is because, like other forms of long-term potentiation in the brain, ganglionic long-term potentiation is saturable and once it is maximally expressed, no further potentiation can be induced by high-frequency stimulation (Alkadhi et al. 1996). Two different stress paradigms, psychosocial stress and forced swim, administered concurrently produce an increase in blood pressure, but not greater than the increase produced by either stressor alone. This suggests that stress hypertension is due to a saturable mechanism and further supports the role of ganglionic long-term potentiation in stress-induced hypertension (Alkadhi et al. 2005b). In addition, immunoblot analysis revealed increased protein levels of signaling molecules important for expression of long-term potentiation in ganglia isolated from€stress-�hypertensive animals to a similar extent of that in normal rat ganglia in which ganglionic long-term potentiation was expressed by high-frequency stimulation in vitro, suggesting in vivo expression of long-term potentiation in stress ganglia (Alzoubi€et€al. 2008b). Finally, pharmacological inhibitors of calmodulin and guanylyl cyclase showed inhibition of basal synaptic responses in ganglia isolated from stressed, but not from control (normal), rats. This indicates in vivo potentiation of responses under stress conditions, which further supports the in vivo expression of long-term potentiation in ganglia during chronic stress (Alzoubi and Alkadhi 2009). Phosphorylated CaMKII, required for long-term potentiation, continues to be active even after intracellular calcium concentration returns to basal levels, until dephosphorylation

408 Sudden Death in Epilepsy: Forensic and Clinical Issues (a)

160

Control (unstressed) Stressed Stressed-treated

Systolic blood pressure (mm Hg)

150 140 130 120 110 100

Tropisetron (2.5 mg/kg/day)

90 80

Systolic blood pressure (mm Hg)

260

Psychosocial stress 0

5

10

Days

15

20

25

(b)

220

SHR

180 140

WKY

100 ICS (2.5 mg/kg) 60 10

20

Days

30

40

Figure 26.5╇ (a) The serotonin 5-HT3 receptor antagonist tropisetron (2.5 mg/kg/day) com-

pletely blocked established psychosocial hypertension. The same treatment did not affect blood pressure in normotensive rats. (b) Chronic treatment of rats with the 5-HT3 receptor antagonist tropisetron (ICS, 2.5 mg/kg, ip twice daily for 10 days, black bar) resulted in a marked decrease in blood pressure of the spontaneously hypertensive rats (SHR) but not Wistar Kyoto (WKY) rats.

by protein phosphatases, such as calcineurin (Fukunaga et al. 1996; Wang and Kelly 1996; Fukunaga and Miyamoto 2000). In the hippocampal dentate gyrus area, reduction in calcineurin protein levels were reported in chronically stressed rats in which long-term potentiation of the dentate gyrus remained intact (Gerges et al. 2003). Similarly, ganglia from stressed rats, and those from normal rats in which ganglionic long-term potentiation was expressed in vitro by high-frequency stimulation, showed a signiἀcant reduction in calcineurin protein levels (Alzoubi et al. 2008b). This reduction of calcineurin protein results in curtailed dephosphorylation, thus providing an adequate level of phosphorylated CaMKII to sustain ganglionic long-term potentiation (Gerges et al. 2003; Alzoubi et al. 2008b).

Synaptic Plasticity of Autonomic Ganglia

Systolic blood pressure (mm Hg)

140

409

(c)

Lean Zucker Obese Zucker Wistar

130 *

120

* *

110 100 90 –30

Ondansetron (0.5 mg/kg) –20

–10

0

10

20

30

Days

Figure 26.5 (Continued)╇ (c) Chronic treatment of obese Zucker (OZR), lean Zucker (LZR) and Wistar rats with the 5-HT3 receptor antagonist, ondansetron. Ondansetron was administered for 5 days (0.5 mg/kg, black bar) resulting in a decrease in blood pressure of the OZRs but not LZRs or Wistar rats. Each point in each series is the mean ± SEM from 4 to 7 rats. In all three graphs, blood pressure was measured by the tail-cuff plethysmography. *Indicates significant difference (p < 0.05, paired t test) from its own baseline value as well as from controls (ANOVA). (From Alkadhi and Alzoubi, Clin Exp Hypertens, 29, 267–286, 2007. With permission.)

Stress-induced component of hypertension in spontaneously hypertensive rats. The spontaneously hypertensive rat has exaggerated cardiovascular responses to external stressful stimuli (Hallback and Folkow 1974; Sanders and Gray 1997). Exaggerated sympathetic nerve activity contributes to the development and maintenance of hypertension in humans (Anderson et al. 1989; Esler et al. 1991; Matsukawa et al. 1993; Schobel et al. 1996), spontaneously hypertensive rats, and borderline hypertensive rats (Judy et al. 1976; Bunag and Takeda 1979; Koepke et al. 1987; Magee and Schoἀeld 1992; Huang and Leenen 1996; Mark 1996; Fuchs et al. 1998). The contribution of psychosocial stress to the elevated blood pressure in the spontaneously hypertensive rat is suggested by the ἀnding that socially isolated spontaneously hypertensive rats exhibit lower mean arterial blood pressure than those kept in groups (Hallback 1975). Expression of long-term potentiation in sympathetic ganglia of spontaneously hypertensive rats is indicated by the reported modiἀcation of ganglionic transmission (Yarowsky and Weinreich 1985; Magee and Schoἀeld 1992). Increased acetylcholine release (Magee and Schoἀeld 1992) and a shortening of spike frequency adaptation in spontaneously hypertensive rat ganglia (Yarowsky and Weinreich 1985) indicate both pre- and postsynaptic changes. The contribution of ganglionic long-term potentiation to hypertension in the spontaneously hypertensive rat is suggested by chronic treatment with the serotonin 5-HT3 receptor antagonist tropisetron, which decreases blood pressure in the spontaneously hypertensive rat, but not in their normotensive control Wistar Kyoto rats (see Figure 26.5b; Alkadhi et al. 2001b). Similar results obtained with a quaternary derivative of tropisetron support a site of action of the drug that is outside the central nervous system (Alkadhi et al. 2001b). This decrease in the blood pressure of spontaneously hypertensive rats during treatment with the quaternary tropisetron is most probably due to inhibition of the serotonin 5-HT3 receptor in sympathetic ganglia because it is unlikely that the quaternary

410 Sudden Death in Epilepsy: Forensic and Clinical Issues

Repetitive CNS activity (stress, seizures) in certain brain structures, particularly the frontal lobe, hippocampus, and amygdala

Cardiovascular centers (brain stem)

P LT f g HT3 ts) o rs 5- nis o to bi onin tag i h t In ero r an (s pto ce re

Blood vessels

Hypertension

Sympathetic ganglia Enhanced activity (GLTP)

Heart

Arrhythmia

Figure 26.6╇ In the whole animal, chronic mental stress provides the repetitive stimulation

required for expression of ganglionic long-term potentiation (gLTP) in sympathetic ganglia, leading to increased peripheral resistance and hypertension. (From Alkadhi, K. and K. Alzoubi, Clin Exp Hypertens, 29, 267–286, 2007. With permission.)

form of the drug can cross the blood–brain barrier to act centrally. The drug, however, may still gain access to the brain at the circumventricular organs, particularly the area postrema. The area postrema is known to have numerous serotonin 5-HT3 receptor binding sites and a strong association with the nucleus tractus solitarius of the medulla oblongata, which contains the cardiovascular control centers (Kilpatrick et al. 1987). However, this possibility is also unlikely inasmuch as the same drug does not decrease blood pressure in normotensive animals. As in the psychosocial stress rat model of hypertension, electrophysiological evidence indicates enhanced ganglionic transmission in superior cervical ganglia isolated from

Synaptic Plasticity of Autonomic Ganglia

411

spontaneously hypertensive rats, which strongly suggests that ganglionic long-term potentiation had been expressed in vivo as a contributing neurogenic factor to hypertension in spontaneously hypertensive rats. First, inhibition of basal ganglionic transmission by ganglionic long-term potentiation inhibitors (serotonin 5-HT3 receptor antagonists) in ganglia isolated from spontaneously hypertensive rats, but not in those from Wistar Kyoto or Wistar rat ganglia (Alkadhi et al. 2001b); second, failure of high-frequency stimulation to express long-term potentiation in ganglia isolated from spontaneously hypertensive rats, indicating occlusion of the response due to saturation (Alkadhi et al. 2001b); and, ἀnally, a shift of input–output curves of ganglia from spontaneously hypertensive rats to the left side of those from young (normotensive) spontaneously hypertensive rats or from Wistar Kyoto rats (Magee and Schoἀeld 1992) indicated expression of ganglionic long-term potentiation (Johnston and Wu 1995). Intercellular recording from single neurons of the superior cervical ganglia of spontaneously hypertensive rats showed increased amplitude of fast excitatory postsynaptic potential and excitatory postsynaptic current. This increase in fast excitatory postsynaptic potential and excitatory postsynaptic current has been shown to be due to increased neurotransmitter release from presynaptic nerve terminals (Magee and Schoἀeld 1994). Thus hypertension in the spontaneously hypertensive rats has a neurogenic component that is due to expression of long-term potentiation in sympathetic ganglia, most probably induced by psychosocial stress. Stress-induced component of hypertension in obese Zucker rats. Most researchers reported moderate hypertension in obese Zucker rats (Kurtz et al. 1989; Walker et al. 1992; Turner et al. 1995; Alonso-Galicia et al. 1996; Pacak et al. 1996; Pamidimukkala and Jandhyala 1996; Van Zwieten et al. 1996; Gerges et al. 2002; Alzoubi et al. 2008a; but see Kasiske et al. 1992). Hypertension in obese Zucker rats is not due to a change in vascular reactivity because morphological characteristics and vascular reactivity of resistance vessels are not different in obese Zucker rats compared to that of lean Zucker rats (Cox and Kikta 1992; Turner et al. 1995; Van Zwieten et al. 1996). Obese, insulin-resistant patients maintain higher levels of insulin and leptin, which lead to increased sympathetic activity (Steinberg 1999). Such increase in central sympathetic activity has been also demonstrated in obese Zucker rats (Arase et al. 1989; Morgan et al. 1995; Alonso-Galicia et al. 1996; Dubois et al. 1996; Pacak et al. 1996; Pamidimukkala and Jandhyala 1996; Tonello et al. 1999; Carlson et al. 2000; Fang et al. 2000). Pharmacological and electrophysiological evidence similar to that reported in the spontaneously hypertensive rats and psychosocial stress-hypertensive rats supports the in vivo expression of long-term potentiation in ganglia from obese Zucker rats. Chronic treatment with the serotonin 5-HT3 receptor antagonist ondansetron results in reversible reduction of blood pressure in obese Zucker rats, but not in lean Zucker rats or Wistar rats (see Figure 26.5c; Gerges et al. 2002). Additionally, basal (non-stimulated) ganglionic transmission can be inhibited by ondansetron in ganglia isolated from obese Zucker rats, but not from lean Zucker rats or Wistar rats. Third, high-frequency stimulation failed to induce long-term potentiation in ganglia isolated from obese Zucker rats, while robust long-term potentiation was expressed in ganglia from lean Zucker rats and Wistar rats indicating occlusion of ganglionic long-term potentiation due to saturation and suggesting in vivo expression of long-term potentiation in obese Zucker rats ganglia (Gerges et al. 2002; Alzoubi et al. 2008a). Finally, the increased ratio of phosphorylated to total CaMKII along with increased protein levels of phosphorylated CaMKII, protein kinase Cγ, neural nitric oxide synthase, and heme oxygenase-2 in ganglia from obese Zucker rats compared

412 Sudden Death in Epilepsy: Forensic and Clinical Issues

to those of lean Zucker rats indicates the in vivo expression of long-term potentiation in sympathetic ganglia of obese rats (Alzoubi et al. 2008a). Similar to ganglia from stresshypertensive rats (Alzoubi et al. 2008b), we have recently reported a reduction in calcineurin protein levels in ganglia from obese Zucker rats compared to that of ganglia from lean Zucker rats. It is possible that the reduction in calcineurin protein levels seen in obese Zucker rats is the result of protracted expression of long-term potentiation in sympathetic ganglia from obese Zucker rats, to allow the required adequate levels of phosphorylated CaMKII by curtailing its dephosphorylation (Alzoubi et al. 2008a). Hypertension in advanced age. Aging is often viewed as a progressive decline in physiological competence with a corresponding inability to adapt to stressful stimuli (McCarty 1986). Old animals show increased morbidity and mortality during prolonged exposure to stressful stimuli (McCarty 1986). Thus, the neural effects of aging could be exacerbated by stress (Porter and Landἀeld 1998; Rodriguez-Capote et al. 1998; Butterἀeld et al. 1999; Butterἀeld et al. 2001). Like stress, aging is a risk factor for hypertension (Weber et al. 1989; McNeil and Silagy 1991; Scarpace and Lowenthal 1994) because old animals develop exaggerated sympathetic activity (Anderson et al. 1989; Weber 1989; Weber et al. 1989; Yamada et al. 1989; Hajduczok et al. 1991a; Hajduczok et al. 1991b; Ng et al. 1993; Scarpace and Lowenthal 1994; Esler et al. 1995a, 1995b; Thor et al. 1999). In superior cervical ganglia from old rats, tyrosine hydroxylase and choline transferÂ� ase activities are increased (Reis et al. 1977; Partanen et al. 1985). Moreover, in reserpinedepleted ganglia, the recovery rate of catecholamines is slower in aged rats, and this slow recovery rate in aged ganglia is due to higher catecholamine secretion rate rather than reÂ�duced synthesis (Partanen et al. 1985). Together, these results indicate enhancement of neurotransmitter synthesis and release in aged rats ganglia (Partanen et al. 1985). Increased sympathetic activity is causally involved in a variety of cardiovascular diseases, including hypertension, heart failure, and ventricular arrhythmias, all of which have increased incidence with aging (Esler and Kaye 2000). Studies of the hippocampus as a target of stress have revealed a considerable degree of structural plasticity and remodeling to counteract stress effects (McEwen 2002 for review). A similar effect might be occurring in autonomic ganglia of stressed or aged animals. Sympathetic ganglia from stressed or aged animals express ganglionic long-term potentiation, which is most probably responsible for their elevated blood pressure (Alkadhi et al. 2005a, 2005b; Alkadhi and Alzoubi 2007; Alzoubi et al. 2008a, 2010). In old rats superior cervical ganglia, c-FOS expression (a marker of activation of cAMP-response element binding protein, which is a major factor in late phase long-term potentiation expression and memory consolidation in the hippocampus) is enhanced. This suggests that c-FOS or cAMP response element binding protein might be involved in genetic events leading to adaptive changes in neural activity (Yang and Koistinaho 1990) such as synaptic plasticity changes. Electrophysiological and biochemical evidence show that mildly hypertensive aged rats have expressed ganglionic long-term potentiation in vivo (Alzoubi et al. 2010). First, a shift of input–output curve of ganglia from aged rats to the left side of input–output curve of ganglia from adult rats indicates expression of ganglionic long-term potentiation. Second, failure of high-frequency stimulation to induce long-term potentiation in ganglia isolated from aged rats indicates occlusion, due to saturation and suggests in vivo expression of long-term potentiation in these ganglia. Third, in vitro inhibition of basal ganglionic transmission by blockers of ganglionic long-term potentiation (serotonin 5-HT3

Synaptic Plasticity of Autonomic Ganglia

413

receptor antagonists) is observed in ganglia isolated from aged rats but not in those from adult rats. Finally, immunoblot analysis revealed that protein levels of signaling molecules required for long-term potentiation expression such as CaMKII (phosphorylated and total) are elevated in ganglia isolated from aged rats compared to those in adult ones. Protein levels of calcineurin, which dephosphorylates phosphorylated CaMKII, were reduced in ganglia isolated from aged rats, probably as a compensatory mechanism to allow prolonged phosphorylation of CaMKII (Alzoubi et al. 2010). Together, these results demonstrate in vivo expression of long-term potentiation in sympathetic ganglia of aged animal, which may contribute to the moderate hypertension often seen in aged subjects. Ouabain-induced arrhythmia, hypertension, and sudden death. Ouabain-induced toxicity was ἀrst reported by Han and Moe (1964), where toxic doses of ouabain, sympathetic nerve stimulation, or coronary occlusion led to electrical instability in the heart that predisposed animals for ventricular arrhythmia (reviewed in Lathers 2002). Thereafter, it has been shown (Lathers et al. 1977, 1978) that ouabain toxicity was associated with non-Â�uniform sympathetic nerves discharges, leading to cardiac arrhythmia in a similar manner to that reported initially by Han and Moe (1964). These non-uniform sympathetic neural discharge was also associated with sudden death induced by coronary occlusion (Lathers et al. 1978). Recently, ouabain or a closely related ouabain-like factor has been identiἀed in human and animal circulation (Ferrandi et al. 1997; Schneider and Jackisch 1998; Kawamura et al. 1999; Bagrov and Shapiro 2008). The level of endogenous ouabain is increased in a large segment of patients with primary hypertension (Manunta et al. 1999) and seems to be dependent on adrenocortical function (Boulanger et al. 1993; Laredo et al. 1994). Action of exogenous ouabain on the hypothalamus causes enhancement of sympathetic nerve activity and elevation of blood pressure (Hamlyn and Manunta 1992; Yuan et al. 1993; Huang et al. 1994; Manunta et al. 1994; Huang and Leenen 1996). In superior cervical ganglia of chronically ouabain-treated rats, the maintenance phase of high-frequency stimulationevoked ganglionic long-term potentiation is prolonged. This prolongation of ganglionic long-term potentiation, which is dependent on ouabain levels in plasma, correlates well with the blood pressure magnitude in these rats (Aileru et al. 2001). Similar prolongation of the ganglionic long-term potentiation maintenance phase is seen in rats that have been inbred for seven generations, based on their increased sensitivity to ouabain-induced blood pressure elevation (Aileru et al. 2001). Thus, altered sympathetic ganglionic function is directly linked to hypertension secondary to elevated ouabain levels or secondary to a genetically determined increase of ouabain sensitivity. Since centrally administered ouabain-binding antibody fragment or angiotensin II type1 receptor antagonist can block increased sympathetic activity and blood pressure elevation in ouabain-treated animals, ouabain could indirectly modulate, or even generate, ganglionic long-term potentiation by centrally affecting the sympathetic output to the preganglionic nerve (Huang et al. 1994; Huang and Leenen 1994, 1996). Alternatively, ouabain could induce changes in ganglionic long-term potentiation by affecting the release and/or the half life of other synaptic factors like serotonin (Aileru et al. 2001), which is necessary for induction and maintenance of ganglionic long-term potentiation (Alkadhi et al. 1996, 2001b; Gerges et al. 2002). Although evidence does not exclude the possibility that long-term changes in blood pressure as being the trigger for changes in synaptic plasticity, recent ἀndings suggest that ganglionic long-term potentiation expression leads to hypertension. It has been shown that long-term potentiation is not altered in the superior cervical ganglion of the hypertensive

414 Sudden Death in Epilepsy: Forensic and Clinical Issues

transgenic (mREN-2) 27 rat, which expresses an exogenous renin gene and has elevated blood pressure (200 mm Hg), suggesting that hypertension does not lead to alteration in synaptic plasticity (Aileru et al. 2004). In other words, although (mREN-2) 27 rats were hypertensive, they did not express ganglionic long-term potentiation. Therefore, it may be concluded that hypertension per se does not induce the expression of ganglionic long-term potentiation, whereas expression of ganglionic long-term potentiation leads to neurogenic hypertension. Autonomic dysfunction and SUDEP. Epilepsy patients seem to differ in autonomic function from the general population, and these differences may be relevant to SUDEP (Lathers and Schraeder 1982; Lathers et al. 1987; Drake et al. 1998; Berilgen et al. 2004). Studies in animal models of epilepsy indicated that autonomic cardiac nerves did not always respond in a predictable manner to changes in blood pressure. Additionally, a marked increase in variability of mean cardiac autonomic neural discharge and the individual parasympathetic and sympathetic discharge were reported (Lathers and Schraeder 1982, 1987). Moreover, cardiac sympathetic neural discharge and cardiac arrhythmias were associated with interictal and ictal activity, and may be contributing factors to sudden death in patients with epilepsy who usually show no apparent seizure activity at the time of death (Lathers and Schraeder 1982, 1987; Lathers et al. 1987; Schraeder and Lathers 1989). In temporal lobe epilepsy, interictal and ictal discharge was linked to autonomic dysfunction with prominent sympathetic overactivity (Hilz et al. 2002). Surgical treatment of temporal lobe epilepsy was suggested as a possible method to reduce the risk for SUDEP (Burgerman et al. 1995; Hilz et al. 2002). This procedure leads to reduced sympathetic cardiomodulation and baroreflex sensitivity, thus reducing sympathetically mediated tachyarrhythmias and excessive bradycardic counter-regulation. Therefore, autonomic dysfunction with predominating sympathetic overstimulation may be an important factor in the pathophysiology of SUDEP. While the exact pathophysiological causes of SUDEP remain unknown, it is very probable that cardiac arrhythmia during and between seizures, or transmission of epileptic activity to the heart via the autonomic nervous system potentially plays a major role in sudden death (Garaizar 2000; Scorza et al. 2008). Several studies have implicated cardiovascular diseases such as arrhythmia as risk factors for SUDEP (Drake et al. 1993; Garaizar 2000; Nei et al. 2000; Darbin et al. 2002; Lathers and Schraeder 2002; Opherk et al. 2002; Nei et al. 2004). For example, patients with refractory temporal lobe epilepsy show greater cardiovascular dysfunction, which may be a relative risk for SUDEP, compared to those with well-controlled temporal lobe epilepsy (Ansakorpi et al. 2002). Moreover, evidence indicates increased autonomic stimulation is associated with seizures in SUDEP victims, compared with a clinically similar group of patients with refractory epilepsy (Nei et al. 2004). Autonomic dysfunction in epilepsy not only affects cardiac function but also affects the pulmonary system, which is postulated as a major factor in SUDEP. As shown by videoelectroencephalograph monitoring units, life-threatening period or death are most often due to pulmonary dysfunction (Ficker 2000; So et al. 2000; Swallow et al. 2002). Pulmonary apnea (Nashef et al. 1996; So et al. 2000) and neurogenic pulmonary edema (Swallow et al. 2002) are well documented in SUDEP. Pulmonary edema may lead to elevated pulmonary vascular pressure as a result of increased sympathetic activity, which causes a combination of pulmonary vasoconstriction and increased left arterial pressure related to systemic hypertension (Devinsky 2004). Since one of the risk factors in SUDEP is high seizure frequency (Walczak et al. 2001), it is likely that the frequent bouts of increased sympathetic

Synaptic Plasticity of Autonomic Ganglia

415

efferent discharge to ganglia cause expression of long-term potentiation in sympathetic ganglia. Such prolonged increase in synaptic strength of these ganglia could be a major contributing factor in SUDEP. Relation of synaptic plasticity in autonomic ganglia to SUDEP—the stress factor. Mental, physical, and psychological stressors are risk factors for SUDEP. One study reported about 21% of SUDEP cases to have suffered certain form of stress such as homelessness, physical trauma, recent relocation, travel, and/or psychiatric stressors, including suicide of a friend and the rape of a daughter (Lear-Kaul et al. 2005). Owada et al. (1999), who reported on autopsies of sudden death cases in Japan from May 1994 to February 1998, showed that the risk factors for sudden death included long-term stress, history of heart disease, hypertension, autonomic disturbance, and short-term stress. Both short-term stress and autonomic disturbance increased the risk of sudden death due to coronary artery disease, while long-term stress was associated with increased risk of sudden death due to acute cardiac dysfunction. Therefore, the authors’ recommendations were to identify subjective symptoms and intervene to alleviate such stress and most probably reduce the risk of sudden death (see Lathers and Schraeder 2006, for review). Whether such risks would extend to SUDEP is still a mater of debate. However, stress is considered a risk factor for seizure activity, which is well known as a major risk for SUDEP (e.g., Dominian et al. 1963; Nashef et al. 1995). Recent evidence established the involvement of psychosocial stress and stress-prone conditions such as obesity and aging in the occurrence of sympathetic overactivity in the autonomic nervous system in the form of in vivo expression of long-term potentiation in sympathetic ganglia. These stress or stress-prone condition–induced sympathetic overactivity or increased sympathetic tone were implicated in the occurrence of cardiovascular dysregulation leading to the development of systemic hypertension in animals (Alkadhi et al. 2001b; Gerges et al. 2002; Alkadhi et al. 2005a, 2005b; Alkadhi and Alzoubi 2007; Alzoubi et al. 2008a, 2008b, 2010; Alzoubi and Alkadhi 2009). Cardiovascular dysregulation can also be involved in increasing the heart rate and subsequent facilitation of arrhythmias, a potential major factor in SUDEP. Additionally, sympathetic overactivity can lead to pulmonary vasoconstriction, which along with systemic hypertension may precipitate pulmonary edema and apnea; both are major pulmonary symptoms that occur during the course of SUDEP. Seizure-induced long-term potentiation in sympathetic ganglia as a putative risk factor in SUDEP can be reduced or eliminated by blocking ganglionic long-term potentiation. Pharmacologic inhibitors of ganglionic long-term potentiation (serotonin 5-HT3 receptor antagonists) would be beneἀcial for reducing the risk of SUDEP through reducing the possibility of both cardiovascular dysregulation that leads to arrhythmias, and the pulmonary vasoconstriction along with systemic hypertension that lead to pulmonary symptoms (Figure 26.6).

26.6â•… Summary and Concluding Remarks Accumulating evidence suggests that stress-induced hypertension is a consequence of the expression of long-term potentiation in sympathetic ganglia and that this neurogenic increase in blood pressure contributes to hypertension in a variety of animal models of hypertension. Reduction of baseline ganglionic transmission by serotonin 5-HT3 receptor antagonists in ganglia excised from hypertensive animals without affecting that of ganglia

416 Sudden Death in Epilepsy: Forensic and Clinical Issues

from normotensive controls suggests prior in vivo expression of long-term potentiation in ganglia from hypertensive animals. Other evidence in support of the role of ganglionic long-term potentiation in stress hypertension includes (1) a left side shift of input–output curves of ganglia from hypertensive animals indicates expression of ganglionic long-term potentiation; (2) normalization or reduction in blood pressure upon treatment of hyperÂ� tensive animals with ganglionic long-term potentiation inhibitors (serotonin 5-HT3 receptor antagonists); (3) failure of high-frequency stimulation to express long-term potentiation in ganglia isolated from hypertensive animals indicating occlusion due to saturation and suggesting in vivo expression of long-term potentiation in these ganglia; (4) increased protein levels of signaling molecules important for long-term potentiation in ganglia isolated from stress or stress prone-hypertensive animals in a similar manner to those seen in the in vitro stimulated normal ganglia; and (5) the inhibition of basal synaptic responses in ganglia isolated from stress-hypertensive, but not in those from normal animals, by superfusion with pharmacological inhibitors of signaling molecules required for ganglionic long-term potentiation, indicating the in vivo expression of long-term potentiation in ganglia from stress-hypertensive rats. Therefore, there is strong evidence for the expression of long-term potentiation in sympathetic ganglia, which contributes to the development of high blood pressure in animal models of hypertension. High seizure frequency may also induce long-term potentiation in autonomic ganglia. It is likely that the continuous increase in sympathetic outflow to ganglia, induced by mental stress or frequent seizures, could provide the repeated high-frequency presynaptic activity required for ganglionic long-term potentiation induction, which results in a sustained elevation of sympathetic tone to blood vessels and the heart muscle. This elevated sympathetic tone (sympathetic overactivity) could increase heart rate, precipitate arrhythmias, and increase the risk for pulmonary apnea leading to SUDEP.

References Aileru, A. A., A. De Albuquerque, J. M. Hamlyn et al. 2001. Synaptic plasticity in sympathetic ganglia from acquired and inherited forms of ouabain-dependent hypertension. Am J Physiol Regul Integr Comp Physiol 281: R635–R644. Aileru, A. A., E. Logan, M. Callahan, C. M. Ferrario, D. Ganten, and D. I. Diz. 2004. Alterations in sympathetic ganglionic transmission in response to angiotensin II in (mRen2)27 transgenic rats. Hypertension 43: 270–275. Aleisa, A. M., K. H. Alzoubi, N. Z. Gerges, and K. A. Alkadhi. 2006. Nicotine blocks stress-induced impairment of spatial memory and long-term potentiation of the hippocampal CA1 region. Int J Neuropsychopharmacol 9: 417–426. Alkadhi, K., and K. Alzoubi. 2007. Role of long-term potentiation of sympathetic ganglia (gLTP) in hypertension. Clin Exp Hypertens 29: 267–286. Alkadhi, K. A., and G. F. Altememi. 1997. Nitric oxide mediates long-term potentiation in rat superior cervical ganglion. Brain Res 753: 315–317. Alkadhi, K. A., K. H. Alzoubi, and A. M. Aleisa. 2005a. Plasticity of synaptic transmission in autonomic ganglia. Prog Neurobiol 75: 83–108. Alkadhi, K. A., R. S. Al-Hijailan, and K. H. Alzoubi. 2008. Long-term depression in the superior cervical ganglion of the rat. Brain Res 1234: 25–31. Alkadhi, K. A., D. Salgado-Commissariat, Y. H. Hogan, and S. B. Akpaudo. 1996. Induction and maintenance of ganglionic long-term potentiation require activation of 5-hydroxytryptamine (5-HT3) receptors. J Physiol 496 (Pt 2): 479–489.

Synaptic Plasticity of Autonomic Ganglia

417

Alkadhi, K. A., R. S. Al-Hijailan, K. Malik, and Y. H. Hogan. 2001a. Retrograde carbon monoxide is required for induction of long-term potentiation in rat superior cervical ganglion. J Neurosci 21: 3515–3520. Alkadhi, K. A., S. A. Otoom, F. L. Tanner, D. Sockwell, and Y. H. Hogan. 2001b. Inhibition of ganglionic long-term potentiation decreases blood pressure in spontaneously hypertensive rats. Exp Biol Med (Maywood) 226: 1024–1030. Alkadhi, K. A., K. H. Alzoubi, A. M. Aleisa, F. L. Tanner, and A. S. Nimer. 2005b. Psychosocial stressinduced hypertension results from in vivo expression of long-term potentiation in rat sympathetic ganglia. Neurobiol Dis 20: 849–857. Alonso-deFlorida, F., M. A. Morales, and A. A. Minzoni. 1991. Modulated long-term potentiation in the cat superior cervical ganglion in vivo. Brain Res 544: 203–210. Alonso-Galicia, M., M. W. Brands, D. H. Zappe, and J. E. Hall. 1996. Hypertension in obese Zucker rats. Role of angiotensin II and adrenergic activity. Hypertension 28: 1047–1054. Altememi, G. F., and K. A. Alkadhi. 1999. Nitric oxide is required for the maintenance but not initiation of ganglionic long-term potentiation. Neuroscience 94: 897–902. Alzoubi, K. H., and K. A. Alkadhi. 2009. Calmodulin and guanylyl cyclase inhibitors block the in vivo expression of gLTP in sympathetic ganglia from chronically stressed rats. Neurosci Res 63: 95–99. Alzoubi, K. H., A. M. Aleisa, and K. A. Alkadhi. 2008a. Expression of gLTP in sympathetic ganglia of obese Zucker rats in vivo: Molecular evidence. J Mol Neurosci 35: 297–306. Alzoubi, K. H., A. M. Aleisa, and K. A. Alkadhi. 2008b. Expression of gLTP in sympathetic ganglia from stress-hypertensive rats: Molecular evidence. J Mol Neurosci 35: 201–209. Alzoubi, K. H., A. M. Aleisa, and K. A. Alkadhi. 2010. In vivo expression of ganglionic long-term potentiation in superior cervical ganglia from hypertensive aged rats. Neurobiol Aging 31: 805–812. Alzoubi, K. H., A. S. Bedawi, A. M. Aleisa, and K. A. Alkadhi. 2004. Hypothyroidism impairs longterm potentiation in sympathetic ganglia: Electrophysiologic and molecular studies. J Neurosci Res 78: 393–402. Anderson, E. A., C. A. Sinkey, W. J. Lawton, and A. L. Mark. 1989. Elevated sympathetic nerve activity in borderline hypertensive humans. Evidence from direct intraneural recordings. Hypertension 14: 177–183. Ando, M., T. Nanba, M. Okuya, and Y. Nagata. 1983. Effect of axotomy on the cyclic GMP increase induced by preganglionic stimulation and high extracellular K+ concentration in superior cervical sympathetic ganglion of the rat. Brain Res 267: 313–321. Ansakorpi, H., J. T. Korpelainen, H. V. Huikuri, U. Tolonen, V. V. Myllyla, and J. I. Isojarvi. 2002. Heart rate dynamics in refractory and well controlled temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 72: 26–30. Arase, K., N. S. Shargill, and G. A. Bray. 1989. Effects of corticotropin releasing factor on genetically obese (fatty) rats. Physiol Behav 45: 565–570. Ariano, M. A., J. A. Lewicki, H. J. Brandwein, and F. Murad. 1982. Immunohistochemical localization of guanylate cyclase within neurons of rat brain. Proc Natl Acad Sci U S A 79: 1316–1320. Bachoo, M., and C. Polosa. 1991. Long-term potentiation of nicotinic transmission by a heterosynaptic mechanism in the stellate ganglion of the cat. J Neurophysiol 65: 639–647. Bachoo, M., and C. Polosa. 1992. Preganglionic axons from the third thoracic spinal segment fail to induce long-term potentiation in the superior cervical ganglion of the cat. Can J Physiol Pharmacol 70 (Suppl): S27–S31. Bachoo, M., T. Heppner, J. Fiekers, and C. Polosa. 1992. A role for protein kinase C in long term potentiation of nicotinic transmission in the superior cervical ganglion of the rat. Brain Res 585: 299–302. Bagrov, A. Y., and J. I. Shapiro. 2008. Endogenous digitalis: Pathophysiologic roles and therapeutic applications. Nat Clin Pract Nephrol 4: 378–392. Beavo, J. A., M. Conti, and R. J. Heaslip. 1994. Multiple cyclic nucleotide phosphodiesterases. Mol Pharmacol 46: 399–405.

418 Sudden Death in Epilepsy: Forensic and Clinical Issues Bennett, M. R. 1994. Nitric oxide release and long term potentiation at synapses in autonomic ganglia. Gen Pharmacol 25: 1541–1551. Berilgen, M. S., T. Sari, S. Bulut, and B. Mungen. 2004. Effects of epilepsy on autonomic nervous system and respiratory function tests. Epilepsy Behav 5: 513–516. Bliss, T. V., and T. Lomo. 1973. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J Physiol 232: 331–356. Blottner, D., and H. G. Baumgarten. 1992. Nitric oxide synthetase (NOS)-containing sympathoadrenal cholinergic neurons of the rat IML-cell column: Evidence from histochemistry, immunohistochemistry, and retrograde labeling. J Comp Neurol 316: 45–55. Bohme, G. A., C. Bon, J. M. Stutzmann, A. Doble, and J. C. Blanchard. 1991. Possible involvement of nitric oxide in long-term potentiation. Eur J Pharmacol 199: 379–381. Boone, J. L. 1991. Stress and hypertension. Prim Care 18: 623–649. Boulanger, B. R., M. P. Lilly, J. M. Hamlyn, J. Laredo, D. Shurtleff, and D. S. Gann. 1993. Ouabain is secreted by the adrenal gland in awake dogs. Am J Physiol 264: E413–E419. Brain, K. L., and M. R. Bennett. 1995. Calcium in the nerve terminals of chick ciliary ganglia during facilitation, augmentation and potentiation. J Physiol 489 (Pt 3): 637–648. Bredt, D. S., and S. H. Snyder. 1990. Isolation of nitric oxide synthetase, a calmodulin-requiring enzyme. Proc Natl Acad Sci U S A 87: 682–685. Bredt, D. S., C. D. Ferris, and S. H. Snyder. 1992. Nitric oxide synthase regulatory sites. Phosphorylation by cyclic AMP-dependent protein kinase, protein kinase C, and calcium/calmodulin protein kinase; identiἀcation of flavin and calmodulin binding sites. J Biol Chem 267: 10976–10981. Briggs, C. A., and D. A. McAfee. 1988. Long-term potentiation at nicotinic synapses in the rat superior cervical ganglion. J Physiol 404: 129–144. Briggs, C. A., D. A. McAfee, and R. E. McCaman. 1985a. Long-term potentiation of synaptic acetylcholine release in the superior cervical ganglion of the rat. J Physiol 363: 181–190. Briggs, C. A., T. H. Brown, and D. A. McAfee. 1985b. Neurophysiology and pharmacology of longterm potentiation in the rat sympathetic ganglion. J Physiol 359: 503–521. Briggs, C. A., D. A. McAfee, and R. E. McCaman. 1988. Long-term regulation of synaptic acetylcholine release and nicotinic transmission: The role of cyclic AMP. Br J Pharmacol 93: 399–411. Brown, T. H., and D. A. McAfee. 1982. Long-term synaptic potentiation in the superior cervical ganglion. Science 215: 1411–1413. Bunag, R. D., and K. Takeda. 1979. Sympathetic hyperresponsiveness to hypothalamic stimulation in young hypertensive rats. Am J Physiol 237: R39–R44. Burgerman, R. S., M. R. Sperling, J. A. French, A. J. Saykin, and M. J. O’Connor. 1995. Comparison of mesial versus neocortical onset temporal lobe seizures: Neurodiagnostic ἀndings and surgical outcome. Epilepsia 36: 662–670. Butterἀeld, D. A., B. J. Howard, and M. A. LaFontaine. 2001. Brain oxidative stress in animal models of accelerated aging and the age-related neurodegenerative disorders, Alzheimer’s disease and Huntington’s disease. Curr Med Chem 8: 815–828. Butterἀeld, D. A., B. Howard, S. Yatin et al. 1999. Elevated oxidative stress in models of normal brain aging and Alzheimer’s disease. Life Sci 65: 1883–1892. Carlson, S. H., J. Shelton, C. R. White, and J. M. Wyss. 2000. Elevated sympathetic activity contributes to hypertension and salt sensitivity in diabetic obese Zucker rats. Hypertension 35: 403–408. Case, C. P., and M. R. Matthews. 1986. Incoming synapses and size of small granule-containing cells in a rat sympathetic ganglion after post-ganglionic axotomy. J Physiol 374: 33–71. Cavalcante de Albuquerque, A. A., J. H. Leal-Cardoso, and D. Weinreich. 1996. Antigen-induced synaptic plasticity in sympathetic ganglia from actively and passively sensitized guinea-pigs. J Auton Nerv Syst 61: 139–144. Cetiner, M., and M. R. Bennett. 1993. Nitric oxide modulation of calcium-activated potassium channels in postganglionic neurones of avian cultured ciliary ganglia. Br J Pharmacol 110: 995–1002. Cox, R. H., and D. C. Kikta. 1992. Age-related changes in thoracic aorta of obese Zucker rats. Am J Physiol 262: H1548–H1556.

Synaptic Plasticity of Autonomic Ganglia

419

Darbin, O., D. J. Casebeer, and D. K. Naritoku. 2002. Cardiac dysrhythmia associated with the immediate postictal state after maximal electroshock in freely moving rat. Epilepsia 43: 336–341. Devinsky, O. 2004. Effects of seizures on autonomic and cardiovascular function. Epilepsy Curr 4: 43–46. Dominian, J., E. A. Serafetinides, and M. Dewhurst. 1963. A follow-up study of late-onset epilepsy. II. Psychiatric and social ἀndings. Br Med J 1: 431–435. Dore, S., M. Takahashi, C. D. Ferris et al. 1999. Bilirubin, formed by activation of heme oxygenase-2, protects neurons against oxidative stress injury. Proc Natl Acad Sci U S A 96: 2445–2450. Drake, M. E., C. R. Reider, and A. Kay. 1993. Electrocardiography in epilepsy patients without cardiac symptoms. Seizure 2: 63–65. Drake, Jr., M. E., J. M. Andrews, and C. M. Castleberry. 1998. Electrophysiologic assessment of autonomic function in epilepsy. Seizure 7: 91–96. Dubois, E. A., K. L. Kam, and G. A. Somsen. 1996. Cardiac iodine-123 metaiodobenzylguanidine uptake in animals with diabetes mellitus and/or hypertension. Eur J Nucl Med 23: 901–908. Dun, N. J., S. L. Dun, S. Y. Wu, and U. Forstermann. 1993. Nitric oxide synthase immunoreactivity in rat superior cervical ganglia and adrenal glands. Neurosci Lett 158: 51–54. Dunant, Y., and M. Dolivo. 1968. Plasticity of synaptic functions in the excised sympathetic ganglion of the rat. Brain Res 10: 271–273. Egan, B., R. Panis, A. Hinderliter, J. Schork, and S. Julius. 1987. Mechanism of increased alpha adrenergic vasoconstriction in human essential hypertension. J Clin Invest 80: 812–817. Eranko, L., H. Paivarinta, S. Soinila, and O. Happola. 1986. Transmitters and modulators in the superior cervical ganglion of the rat. Med Biol 64: 75–83. Esler, M., and D. Kaye. 2000. Sympathetic nervous system activation in essential hypertension, cardiac failure and psychosomatic heart disease. J Cardiovasc Pharmacol 35: S1–S7. Esler, M., C. Ferrier, G. Lambert, G. Eisenhofer, H. Cox, and G. Jennings. 1991. Biochemical evidence of sympathetic hyperactivity in human hypertension. Hypertension 17: III29–III35. Esler, M., S. Julius, A. Zweifler, O. Randall, E. Harburg, H. Gardiner, and V. DeQuattro. 1977. Mild high-renin essential hypertension. Neurogenic human hypertension? N Engl J Med 296: 405–411. Esler, M. D., J. M. Thompson, and D. M. Kaye et al. 1995a. Effects of aging on the responsiveness of the human cardiac sympathetic nerves to stressors. Circulation 91: 351–358. Esler, M. D., A. G. Turner, D. M. Kaye et al. 1995. Aging effects on human sympathetic neuronal function. Am J Physiol 268: R278–R285. Fang, Z., S. H. Carlson, N. Peng, and J. M. Wyss. 2000. Circadian rhythm of plasma sodium is disrupted in spontaneously hypertensive rats fed a high-NaCl diet. Am J Physiol Regul Integr Comp Physiol 278: R1490–R1495. Ferrandi, M., P. Manunta, S. Balzan, J. M. Hamlyn, G. Bianchi, and P. Ferrari. 1997. Ouabain-like factor quantiἀcation in mammalian tissues and plasma: Comparison of two independent assays. Hypertension 30: 886–896. Ficker, D. M. 2000. Sudden unexplained death and injury in epilepsy. Epilepsia 41 (Suppl 2): S7–S12. Fuchs, L. C., A. M. Hoque, and N. L. Clarke. 1998. Vascular and hemodynamic effects of behavioral stress in borderline hypertensive and Wistar-Kyoto rats. Am J Physiol 274: R375–R382. Fukunaga, K., and E. Miyamoto. 2000. A working model of CaM kinase II activity in hippocampal long-term potentiation and memory. Neurosci Res 38: 3–17. Fukunaga, K., D. Muller, and E. Miyamoto. 1996. CaM kinase II in long-term potentiation. Neurochem Int 28: 343–358. Garaizar, C. 2000. Sudden unexpected and unexplained death in epilepsy. Rev Neurol 31: 436–441. Gerges, N. Z., A. M. Aleisa, A. A. Alhaider, and K. A. Alkadhi. 2002. Reduction of elevated arterial blood pressure in obese Zucker rats by inhibition of ganglionic long-term potentiation. Neuropharmacology 43: 1070–1076.

420 Sudden Death in Epilepsy: Forensic and Clinical Issues Gerges, N. Z., A. M. Aleisa, L. A. Schwarz, and K. A. Alkadhi. 2003. Chronic psychosocial stress decreases calcineurin in the dentate gyrus: A possible mechanism for preservation of early ltp. Neuroscience 117: 869–874. Gerges, N. Z., K. H. Alzoubi, C. R. Park, D. M. Diamond, and K. A. Alkadhi. 2004. Adverse effect of the combination of hypothyroidism and chronic psychosocial stress on hippocampusÂ�dependent memory in rats. Behav Brain Res 155: 77–84. Grozdanovic, Z., H. G. Baumgarten, and G. Bruning. 1992. Histochemistry of NADPH-diaphorase, a marker for neuronal nitric oxide synthase, in the peripheral autonomic nervous system of the mouse. Neuroscience 48: 225–235. Hajduczok, G., M. W. Chapleau, and F. M. Abboud. 1991a. Increase in sympathetic activity with age. II. Role of impairment of cardiopulmonary baroreflexes. Am J Physiol 260: H1121–H1127. Hajduczok, G., M. W. Chapleau, S. L. Johnson, and F. M. Abboud. 1991b. Increase in sympathetic activity with age. I. Role of impairment of arterial baroreflexes. Am J Physiol 260: H1113–H1120. Hallback, M. 1975. Consequence of social isolation on blood pressure, cardiovascular reactivity and design in spontaneously hypertensive rats. Acta Physiol Scand 93: 455–465. Hallback, M., and B. Folkow. 1974. Cardiovascular responses to acute mental ‘stress’ in spontaneously hypertensive rats. Acta Physiol Scand 90: 684–698. Hamlyn, J. M., and P. Manunta. 1992. Ouabain, digitalis-like factors and hypertension. J Hypertens Suppl 10: S99–S111. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hilz, M. J., O. Devinsky, W. Doyle, A. Mauerer, and M. Dutsch. 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125: 985–995. Hogan, Y. H., R. Hawkins, and K. A. Alkadhi. 1998. Adenosine A1 receptor activation inhibits LTP in sympathetic ganglia. Brain Res 807: 19–28. Huang, B. S., and F. H. Leenen. 1994. Brain “ouabain” mediates the sympathoexcitatory and hypertensive effects of high sodium intake in Dahl salt-sensitive rats. Circ Res 74: 586–595. Huang, B. S., and F. H. Leenen. 1996. Sympathoexcitatory and pressor responses to increased brain sodium and ouabain are mediated via brain ANG II. Am J Physiol 270: H275–H280. Huang, B. S., X. Huang, E. Harmsen, and F. H. Leenen. 1994. Chronic central versus peripheral ouabain, blood pressure, and sympathetic activity in rats. Hypertension 23: 1087–1090. Ikegaya, Y., H. Saito, and N. Matsuki. 1994. Involvement of carbon monoxide in long-term potentiation in the dentate gyrus of anesthetized rats. Jpn J Pharmacol 64: 225–227. Ito, H., I. Kanno, J. Hatazawa, and S. Miura. 2003. Changes in human cerebral blood flow and myocardial blood flow during mental stress measured by dual positron emission tomography. Ann Nucl Med 17: 381–386. Johnston, D., and S. Wu. 1995. Cellular neurophysiology of learning and memory. In Foundations of Cellular Neurophysiology, 441–475. Cambridge, MA: MIT Press. Judy, W. V., A. M. Watanabe, D. P. Henry, H. R. Besch Jr., W. R. Murphy, and G. M. Hockel. 1976. Sympathetic nerve activity: Role in regulation of blood pressure in the spontaenously hypertensive rat. Circ Res 38: 21–29. Kandel, E. R., and R. D. Hawkins. 1992. The biological basis of learning and individuality. Sci Am 267: 78–86. Kannel, W. B., W. F. Peter, and M. D. Wilson. 1999. Cardiovascular risk factors and hypertension. In Hypertension Primer: The Essentials of High Blood Pressure, ed. J. L. Izzo and H. R. Black, 199–200. Philadelphia, PA: Lippincott Williams and Wilkins. Kasiske, B. L., M. P. O’Donnell, and W. F. Keane. 1992. The Zucker rat model of obesity, insulin resistance, hyperlipidemia, and renal injury. Hypertension 19: I110–I115. Kawamura, A., J. Guo, Y. Itagaki et al. 1999 On the structure of endogenous ouabain. Proc Natl Acad Sci U S A 96: 6654–6659. Kilpatrick, G. J., B. J. Jones, and M. B. Tyers. 1987. Identiἀcation and distribution of 5-HT3 receptors in rat brain using radioligand binding. Nature 330: 746–748.

Synaptic Plasticity of Autonomic Ganglia

421

Klimaschewski, L., W. Kummer, and C. Heym. 1996. Localization, regulation and functions of neurotransmitters and neuromodulators in cervical sympathetic ganglia. Microsc Res Tech 35: 44–68. Koepke, J., P. S. Jones, and G. F. DiBona. 1987. Alpha 2-adrenoceptors in amygdala control renal sympathetic nerve activity and renal function in conscious spontaneously hypertensive rats. Brain Res 404: 80–88. Koyano, K., K. Kuba, and S. Minota. 1985. Long-term potentiation of transmitter release induced by repetitive presynaptic activities in bull-frog sympathetic ganglia. J Physiol 359: 219–233. Kuba, K., and E. Kumamoto. 1986. Long-term potentiation of transmitter release induced by adrenaline in bull-frog sympathetic ganglia. J Physiol 374: 515–530. Kuba, K., and E. Kumamoto. 1990. Long-term potentiations in vertebrate synapses: A variety of cascades with common subprocesses. Prog Neurobiol 34: 197–269. Kuba, K., E. Kato, E. Kumamoto, K. Koketsu, and K. Hirai. 1981. Sustained potentiation of transmitter release by adrenaline and dibutyryl cyclic AMP in sympathetic ganglia. Nature 291: 654–656. Kumamoto, E., and K. Kuba. 1983. Sustained rise in ACh sensitivity of a sympathetic ganglion cell induced by postsynaptic electrical activities. Nature 305: 145–146. Kumamoto, E., and K. Kuba. 1986. Mechanism of long-term potentiation of transmitter release induced by adrenaline in bullfrog sympathetic ganglia. J Gen Physiol 87: 775–793. Kurtz, T. W., R. C. Morris, and H. A. Pershadsingh. 1989. The Zucker fatty rat as a genetic model of obesity and hypertension. Hypertension 13: 896–901. Laredo, J., B. P. Hamilton, and J. M. Hamlyn. 1994. Ouabain is secreted by bovine adrenocortical cells. Endocrinology 135: 794–797. Lathers, C. M. 2002. Clinical pharmacology of antimicrobial use in humans and animals. J Clin Pharmacol 42: 587–600. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42: 123–136. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57: 1058–1065. Lear-Kaul, K. C., L. Coughlin, and M. J. Dobersen. 2005. Sudden unexpected death in epilepsy: A retrospective study. Am J Forensic Med Pathol 26: 11–17. Lengyel, I., K. A. Nichol, A. T. Sim, M. R. Bennett, P. R. Dunkley, and J. A. Rostas. 1996. CharacterizaÂ� tion€ of protein kinase and phosphatase systems in chick ciliary ganglion. Neuroscience 70: 577–588. Lin, Y. Q., and M. R. Bennett. 1994. Nitric oxide modulation of quantal secretion in chick ciliary ganglia. J Physiol 481 (Pt 2): 385–394. Magee, J. C., and G. G. Schoἀeld. 1992 Neurotransmission through sympathetic ganglia of spontaneously hypertensive rats. Hypertension 20: 367–373. Magee, J. C., and G. G. Schoἀeld. 1994. Alterations of synaptic transmission in sympathetic ganglia of spontaneously hypertensive rats. Am J Physiol 267: R1397–R1407.

422 Sudden Death in Epilepsy: Forensic and Clinical Issues Magleby, K. L., and J. E. Zengel. 1975. A quantitative description of tetanic and post-tetanic potentiation of transmitter release at the frog neuromuscular junction. J Physiol 245: 183–208. Manunta, P., A. C. Rogowski, B. P. Hamilton, and J. M. Hamlyn. 1994. Ouabain-induced hypertension in the rat: Relationships among plasma and tissue ouabain and blood pressure. J Hypertens 12: 549–560. Manunta, P., P. Stella, R. Rivera et al. 1999. Left ventricular mass, stroke volume, and ouabain-like factor in essential hypertension. Hypertension 34: 450–456. Mark, A. L. 1996. The sympathetic nervous system in hypertension: A potential long-term regulator of arterial pressure. J Hypertens Suppl 14: S159–S165. Martin, A. R., and G. Pilar. 1964. Presynaptic and post-synaptic events during post-tetanic potentiation and facilitation in the avian ciliary ganglion. J Physiol 175: 17–30. Matsukawa, T., T. Mano, E. Gotoh, and M. Ishii. 1993. Elevated sympathetic nerve activity in patients with accelerated essential hypertension. J Clin Invest 92: 25–28. Maurice, D. H., and R. J. Haslam. 1990. Molecular basis of the synergistic inhibition of platelet function by nitrovasodilators and activators of adenylate cyclase: Inhibition of cyclic AMP breakdown by cyclic GMP. Mol Pharmacol 37: 671–681. Mazet, B. S., M. Miller, and J. H. Szurszewski. 1996. Electrophysiological effects of nitric oxide in mouse superior mesenteric ganglion. Am J Physiol 270: G324–G331. McAfee, D. A., B. K. Henon, G. J. Whiting, J. P. Horn, P. J. Yarowsky, and D. K. Turner. 1980. The action of cAMP and catecholamines in mammalian sympathetic ganglia. Fed Proc 39: 2997–3002. McCarty, R. 1986. Age-related alterations in sympathetic-adrenal medullary responses to stress. Gerontology 32: 172–183. McEwen, B. S. 1998. Protective and damaging effects of stress mediators. N Engl J Med 338: 171–179. McEwen, B. W. 2002. Sex, stress and the hippocampus: Allostasis, allostatic load and the aging process. Neurobiol Aging 23: 921–939. McNeil, J. J., and C. A. Silagy. 1991. Hypertension in the elderly: Epidemiology and pathophysiology. Cardiovasc Drugs Ther 4 (Suppl 6): 1197–1201. Minota, S., E. Kumamoto, O. Kitakoga, and K. Kuba. 1991. Long-term potentiation induced by a sustained rise in the intraterminal Ca2+ in bull-frog sympathetic ganglia. J Physiol 435: 421–438. Morales, M. A., M. Bachoo B. Collier, and C. Polosa. 1994. Pre- and postsynaptic components of nicotinic long-term potentiation in the superior cervical ganglion of the cat. J Neurophysiol 72: 819–824. Morgan, D. A., E. A. Anderson, and A. L. Mark. 1995. Renal sympathetic nerve activity is increased in obese Zucker rats. Hypertension 25: 834–838. Morris, R., E. Southam, S. R. Gittins, and J. Garthwaite. 1993. NADPH-diaphorase staining in autonomic and somatic cranial ganglia of the rat. Neuroreport 4: 62–64. Naftel, J. P., and S. G. Hardy. 1997. Visceral motor pathways. In Fundamental Neuroscience, ed. D. E. Haines, 417–429. New York, NY: Churchill-Livingstone. Nakane, M., J. Mitchell, U. Forstermann, and F. Murad. 1991. Phosphorylation by calcium calmodulin-dependent protein kinase II and protein kinase C modulates the activity of nitric oxide synthase. Biochem Biophys Res Commun 180: 1396–1402. Nashef, L., D. R. Fish, J. W. Sander, and S. D. Shorvon. 1995. Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral centre. J Neurol Neurosurg Psychiatry 58: 462–464. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60: 297–300. Nauta, W. J. T. 1964. The frontal granular cortex and behavior, ed. J. W. Warren and K. Akert, 372– 396. New York, NY: McGraw-Hill. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41: 542–548.

Synaptic Plasticity of Autonomic Ganglia

423

Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45: 338–345. Ng, A. V., R. Callister, D. G. Johnson, and D. R. Seals. 1993. Age and gender influence muscle sympathetic nerve activity at rest in healthy humans. Hypertension 21: 498–503. Nichols, R. A., and P. Mollard. 1996. Direct observation of serotonin 5-HT3 receptor-induced increases in calcium levels in individual brain nerve terminals. J Neurochem 67: 581–592. Okamura, H., K. Umehara, N. Tadaki, Y. Hisa, H. Esumi, and Y. Ibata. 1995. Sympathetic preganglionic neurons contain nitric oxide synthase and project to the superior cervical ganglion: Combined application of retrograde neuronal tracer and NADPH-diaphorase histochemistry. Brain Res Bull 36: 491–494. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52: 117–127. Owada, M., Y. Aizawa, K. Kurihara, N. Tanabe, T. Aizaki, and T. Izumi. 1999. Risk factors and triggers of sudden death in the working generation: An autopsy proven case–control study. Tohoku J Exp Med 189: 245–258. Pacak K., M. Palkovits, S. Makino, I. J. Kopin, and D. S. Goldstein. 1996. Brainstem hemisection decreases corticotropin-releasing hormone mRNA in the paraventricular nucleus but not in the central amygdaloid nucleus. J Neuroendocrinol 8: 543–551. Pamidimukkala, J., and B. S. Jandhyala. 1996. Evaluation of hemodynamics, vascular reactivity and baroreceptor compensation in the insulin resistant Zucker obese rats. Clin Exp Hypertens 18: 1089–1104. Partanen, M., S. B. Waller, E. D. London, and A. Hervonen. 1985. Indices of neurotransmitter synthesis and release in aging sympathetic nervous system. Neurobiol Aging 6: 227–232. Porter, N. M., and P. W. Landἀeld. 1998. Stress hormones and brain aging: Adding injury to insult? Nat Neurosci 1: 3–4. Poss, K. D., M. J. Thomas, A. K. Ebralidze, T. J. O’Dell, and S. Tonegawa. 1995. Hippocampal longterm potentiation is normal in heme oxygenase-2 mutant mice. Neuron 15: 867–873. Quenzer, L. F., B. A. Patterson, and R. L. Volle. 1980a. The cyclic nucleotide content of the rat superior cervical ganglion. J Pharmacol Exp Ther 215: 297–303. Quenzer, L. F., B. A. Patterson, and R. L. Volle. 1980b. K+-induced accumulation of guanosine 3′,5′monophosphate in sympathetic ganglia. J Neurochem 34: 1782–1784. Reis, D. J., R. A. Ross, and T. H. Joh. 1977. Changes in the activity and amounts of enzymes synthesizing catecholamines and acetylcholine in brain, adrenal medulla, and sympathetic ganglia of aged rat and mouse. Brain Res 136: 465–474. Rodriguez-Capote, K., E. Cespedes, R. Arencibia, and M. Gonzalez-Hoyuela. 1998. [Indicators of oxidative stress in aging rat brain. The effect of nerve growth factor]. Rev Neurol 27: 494–500. Ronde, P., and R. A. Nichols. 1998. High calcium permeability of serotonin 5-HT3 receptors on presynaptic nerve terminals from rat striatum. J Neurochem 70: 1094–1103. Saito, S., G. J. Kidd, B. D. Trapp et al. 1994. Rat spinal cord neurons contain nitric oxide synthase. Neuroscience 59: 447–456. Sanders, B. J., and M. J. Gray. 1997. Early environmental influences can attenuate the blood pressure response to acute stress in borderline hypertensive rats. Physiol Behav 61: 749–754. Scarpace, N. T., and D. T. Lowenthal. 1994. Sympathetic nervous system: Aging and exercise. South Med J 87: S42–S46. Schneider, H. P., and C. Jackisch. 1998. Potential beneἀts of estrogens and progestogens on breast cancer. Int J Fertil Womens Med 43: 278–285. Schobel, H. P., T. Fischer, K. Heuszer, H. Geiger, and R. E. Schmieder. 1996. Preeclampsia—a state of sympathetic overactivity. N Engl J Med 335: 1480–1485. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3: 55–62. Scorza, F. A., R. M. Cysneiros, R. M. Arida et al. 2008. Adult hippocampal neurogenesis and sudden unexpected death in epilepsy: Reality or just an attractive history? Med Hypotheses 71: 914–922.

424 Sudden Death in Epilepsy: Forensic and Clinical Issues Scott, T. R., and M. R. Bennett. 1993a. The effect of nitric oxide on the efficacy of synaptic transmission through the chick ciliary ganglion. Br J Pharmacol 110: 627–632. Scott, T. R., and M. R. Bennett. 1993b. The effect of ions and second messengers on long-term potentiation of chemical transmission in avian ciliary ganglia. Br J Pharmacol 110: 461–469. Shapiro, P. A., R. P. Sloan, E. Bagiella, J. P. Kuhl, S. Anjilvel, and J. J. Mann. 2000. Cerebral activation, hostility, and cardiovascular control during mental stress. J Psychosom Res 48: 485–491. Sheng, H., M. L. Hughes, F. Murad, and C. A. Briggs. 1992. Evidence that nitric oxide mediates the cyclic GMP response to synaptic activity in the rat superior cervical ganglion. Brain Res 597: 343–345. Sheng, H., G. D. Gagne, T. Matsumoto, M. F. Miller, U. Forstermann, and F. Murad. 1993. Nitric oxide synthase in bovine superior cervical ganglion. J Neurochem 61: 1120–1126. Siegrist, J. 2001. Psychosocial factors influencing development and course of coronary heart disease. Herz 26: 316–325. Skinner, J. E., and J. C. Reed. 1981. Blockade of frontocortical-brain stem pathway prevents ventricular ἀbrillation of ischemic heart. Am J Physiol 240: H156–H163. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41: 1494–1497. Southam, E., S. L. Charles, and J. Garthwaite. 1996. The nitric oxide-cyclic GMP pathway and synaptic plasticity in the rat superior cervical ganglion. Br J Pharmacol 119: 527–532. Steinber, O. H. 1999. Insulin resistance and hypertension. In: Hypertension Primer: The Essentials of High Blood Pressure, ed. J. L. Izzo and H. R. Black, 121–122. New York, NY: Lippincott Williams & Wilkins. Stevens, C. F., and Y. Wang. 1993. Reversal of long-term potentiation by inhibitors of haem oxygenase. Nature 364: 147–149. Swallow, R. A., C. E. Hillier, and P. E. Smith. 2002. Sudden unexplained death in epilepsy (SUDEP) following previous seizure-related pulmonary oedema: Case report and review of possible preventative treatment. Seizure 11: 446–448. Szilagyi, J. E. 1991. Psychosocial stress elevates blood pressure via an opioid dependent mechanism in normotensive rats. Clin Exp Hypertens A 13: 1383–1394. Thor, P. J., W. Kolasinska-Kloch, A. Pitala, A. Janik, B. Kopp, and W. Sibiga. 1999. The influence of aging on autonomic nervous system activity and gastric myoelectric activity in humans. Folia Med Cracov 40: 53–62. Tonello, C., A. Giordano, V. Cozzi et al. 1999. Role of sympathetic activity in controlling the expression of vascular endothelial growth factor in brown fat cells of lean and genetically obese rats. FEBS Lett 442: 167–172. Turner, N. C., C. Gudgeon, and N. Toseland. 1995. Effects of genetic hyperinsulinaemia on vascular reactivity, blood pressure, and renal structure in the Zucker rat. J Cardiovasc Pharmacol 26: 714–720. Valtschanoff, J. G., R. J. Weinberg, and A. Rustioni. 1992. NADPH diaphorase in the spinal cord of rats. J Comp Neurol 321: 209–222. Van Zwieten, P. A., K. L. Kam, A. J. Pijl, M. G. Hendriks, O. H. Beenen, and M. Pfaffendorf. 1996. Hypertensive diabetic rats in pharmacological studies. Pharmacol Res 33: 95–105. Volle, R. L. 1966. Enhancement of neuromuscular responses to acetylcholine and succinylcholine following repetitive stimulation of the motor nerve. Arch Int Pharmacodyn Ther 160: 284–293. Walker, C. D., K. A. Scribner, J. S. Stern, and M. F. Dallman. 1992. Obese Zucker (fa/fa) rats exhibit normal target sensitivity to corticosterone and increased drive to adrenocorticotropin during the diurnal trough. Endocrinology 131: 2629–2637. Wang, J. H., and P. T. Kelly. 1996. The balance between postsynaptic Ca(2+)-dependent protein kinase and phosphatase activities controlling synaptic strength. Learn Mem 3: 170–181. Weber, M. A. 1989. Sympathetic nervous system and hypertension. Therapeutic perspectives. Am J Hypertens 2: 147S–152S. Weber, M. A., J. M. Neutel, and D. G. Cheung. 1989. Hypertension in the aged: A pathophysiologic basis for treatment. Am J Cardiol 63: 25H–32H.

Synaptic Plasticity of Autonomic Ganglia

425

Weinreich, D., B. J. Undem, G. Taylor, and M. F. Barry. 1995. Antigen-induced long-term potentiation of nicotinic synaptic transmission in the superior cervical ganglion of the guinea pig. J Neurophysiol 73: 2004–2016. Wu, S. Y., and N. J. Dun. 1995. Calcium-activated release of nitric oxide potentiates excitatory synaptic potentials in immature rat sympathetic preganglionic neurons. J Neurophysiol 74: 2600–2603. Yamada, Y., E. Miyajima, O. Tochikubo, T. Matsukawa, and M. Ishii. 1989. Age-related changes in muscle sympathetic nerve activity in essential hypertension. Hypertension 13: 870–877. Yang, G. A., and J. Koistinaho. 1990. Age-related increase in c-fos expression in the sympathetic neurons of the rat superior cervical ganglion. Brain Res 533: 338–343. Yarowsky, P., and D. Weinreich. 1985. Loss of accommodation in sympathetic neurons from spontaneously hypertensive rats. Hypertension 7: 268–276. Yuan, C. M., P. Manunta, J. M. Hamlyn et al. 1993. Long-term ouabain administration produces hypertension in rats. Hypertension 22: 178–187. Zhang, C., M. Bachoo, and C. Polosa. 1996. Long-term depression of a sympathetic ganglionic response to opioids by prolonged synaptic activity and by phorbol esters. Brain Res 710: 1–10. Zhuo, M., S. A. Small, E. R. Kandel, and R. D. Hawkins. 1993. Nitric oxide and carbon monoxide produce activity-dependent long-term synaptic enhancement in hippocampus. Science 260: 1946–1950.

Animal Model for Sudden Cardiac Death Autonomic Cardiac Sympathetic Nonuniform Neural Discharge

27

Claire M. Lathers

Contents 27.1 Introduction 27.2 Animal Models to Study Autonomic Dysfunction, Induction of Arrhythmias, and/or Death 27.3 Discussion References

427 428 430 433

27.1â•…Introduction Electrical instability of the heart may be a unifying mechanism for cardiac arrhythmia and sudden death. Contributing factors in the pathogenesis of sudden cardiac death include changes in stress-related release of catecholamines (Lathers and Roberts 1980; Lathers et al. 1981, 1982); platelet aggregation; arteriosclerotic coronary artery disease; cardiac pathology including rare myocardial tumors, embolism, rupture, or abnormalities in the structure of the AV junction; changes in the autonomic neural control of the heart; and cardiac arrhythmias (James 1983). Potentially fatal changes in catecholamine release and the related consequence of alterations in the neural control of cardiac rhythm (Lathers et al. 1977a, 1977b, 1978) would not be detected at autopsy. In this chapter, the role of autonomic neural mechanisms in the development of cardiac arrhythmias in sudden death will be reviewed by comparing data obtained in two animal models: digitalis toxicity and coronary occlusion. The use of multiple, different animal models to study the clinical problem of sudden death in humans with epilepsy is most important to gain an understanding of the underlying pathophysiological mechanisms that are risk factors for sudden death. Often an animal model developed to study one clinical problem is later applied to a different clinical problem and provides insight into the second clinical problem. Such application has been discussed by Lathers in her 1981 article, “Models for Studying Sequelae to ‘Induced Myocardial Infarction’,” written at the request of the National Academy of Science for publication in Mammalian Models for Research on Aging (Lathers 1981). In this article, Lathers describes the use of an anesthetized cat to monitor postganglionic cardiac sympathetic neural discharge innervating the heart before and after abrupt occlusion of the left anterior descending coronary artery to mimic the effects of myocardial infarction sudden death in 427

428 Sudden Death in Epilepsy: Forensic and Clinical Issues

humans. This neural recording model was ἀrst used to explore mechanisms for induction of arrhythmias and sudden cardiac death by examining neural discharge before and after cardiac arrhythmias induced by toxic doses of ouabain administered to the anesthetized cat (Lathers et al. 1977b, 1978). The nonuniform sympathetic neural discharge recordings were hypothesized by Lathers et al. (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978; Lathers and Roberts 1980; Lathers 1980a) to be contributing to the development of arrhythmia and/ or sudden death via nonuniform recovery of excitability in ventricular muscle (Han and Moe 1964). Sympathetic innervation and its role in normal and abnormal cardiac function was discussed (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978; Lathers and Roberts 1980; Lathers et al. 1988a; Lathers and Roberts 1985; Lathers 1980b). This nonuniform neural discharge animal model developed by Lathers et al. (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978; Lathers 1980b) was applied by Lathers and Schraeder (1982) and Schraeder and Lathers (1983) to the problem of sudden death in persons with epilepsy.

27.2â•…Animal Models to Study Autonomic Dysfunction, Induction of Arrhythmias, and/or Death Lathers developed two different animal models to study the role of autonomic cardiac neural discharge in the production of cardiac arrhythmias and/or death. Anesthetized cats were studied by performing a right thoracotomy to allow recording of the neural discharge from postganglionic cardiac sympathetic nerves (Lathers 1980a; Lathers et al. 1974a, 1977a 1977b). Blood gases were maintained within the physiological range while the surgical procedure was ongoing. In the ἀrst animal model, ouabain was administered intravenously every 15 minutes until arrhythmias were elicited and continued until death occurred. Neural discharge in the minute before the onset of arrhythmia was increased in one nerve, decreased in another, and decreased to a lesser extent in the third nerve (reproduced with permission from Lathers 1980b; also published as Figure 27.1, upper graph). Immediately before ouabain-induced arrhythmia, neural discharge was slightly increased in one postganglionic cardiac sympathetic nerve and depressed in the other when a second animal was examined (reproduced with permission from Lathers 1980b; also published as Figure 1, lower graph, in Lathers and Schraeder 1990). This discharge pattern was designated “nonuniform” (Lathers 1980b; Lathers et al. 1974a, 1977a, 1977b). In contrast, when the neural discharge in each of three postganglionic sympathetic branches was increased, this was a “uniform” neural discharge (reproduced with permission from Lathers and Schraeder 1987; also published as Figure 2A, page 137 in Lathers and Schraeder 1990). Neural discharge was also designated “uniform” when all neural activity was decreased or unchanged. The uniform postganglionic cardiac sympathetic neural discharge was hypothesized to be necessary at the cardiac myocardial junctions to maintain normal electrical excitability and automaticity, i.e., normal sinus rhythm. In contrast, a nonuniform neural discharge occurred when activity in one sympathetic branch was increased and that in a second was decreased, while discharge was not altered in a third (Figure 27.2) (reproduced with permission from Lathers and Schraeder 1987; also published as Figure 2B, page 137 in Lathers and Schraeder 1990). Nonuniform neural discharge was hypothesized to be manifested in the heart as inhomogeneity of myocardial electrical excitability and conduction patterns, as demonstrated by Han and Moe (1964). They found that myocardial nonuniformity could cause ventricular arrhythmias, including death via ventricular ἀbrillation.

Sympathetic neural discharge (impulses/s) percent of control

Animal Model for Sudden Cardiac Death

250 200 150 100 50 0

Sympathetic neural discharge (impulses/s) percent of control

429

0 25

10

0 25

10

50

20

30 75

40 Time (min) Ouabain (µg/kg)

20

30 75

40

150 100 50 0

50

100

Time (min) Ouabain (µg/kg)

Figure 27.1╇ Effect of ouabain (25 μg/kg, i.v.) on postganglionic cardiac sympathetic neural

discharge (impulses/s). Data are graphed as a function of time in minutes. In upper graph, two postganglionic cardiac nerves were monitored in one cat; in lower graph, three nerves were monitored in another cat. (From Lathers, C. M., Eur J. Pharmacol, 64, 95–106, 1980. With permission.)

The second animal model developed by Lathers in her laboratory mimicked the cardiac arrhythmias occurring in victims of sudden death associated with an acute myocardial infarction. Abrupt acute coronary occlusion of the left anterior descending coronary artery was done in anesthetized cats. The resulting cardiac arrhythmias and death were also associated with a nonuniform cardiac sympathetic neural discharge (Lathers 1980a, 1981; Lathers et al. 1977a, 1977b, 1978). The associated arrhythmia was suppressed by cardiacselective beta-adrenergic blocking agents (Spivey and Lathers 1985; Lathers and Spivey 1987; Lathers et al. 1988a, 1988b). The data obtained in these animal models for digitalis toxicity and coronary occlusion raised the question of whether a new animal model of neural nonuniformity and arrhythmias could be developed to investigate experimental mechanisms of sudden unexplained death in persons with epilepsy. Six years after Lathers developed the above two animal models to study the role of the autonomic nervous system in causing arrhythmias and death due to digitalis toxicity and coronary occlusion, they were modiἀed to study epileptogenic activity and autonomic cardiac neural discharge in a

430 Sudden Death in Epilepsy: Forensic and Clinical Issues

(a) Postganglionic cardiac sympathetic branches 1 2 3

Uniform neural discharge

NSR

(b) Postganglionic cardiac sympathetic branches 1 2 3

Nonuniform neural 0 discharge VF

Figure 27.2╇ Postganglionic cardiac sympathetic neural discharge in three branches innervat-

ing the ventricle. (a) Nerve activity in all branches is enhanced (↑) above control and is designated nonuniform neural discharge. (b) Nerve activity is increased in one branch (↑), decreased in a second (↓), and shows no change in a third (○); this trend is designated a nonuniform neural discharge. (From Lathers C. M. and Schraeder P. L., J Clin Pharmacol, 27, 346–356, 1987. With permission.)

new animal model while working with Dr. Schraeder (Lathers and Schraeder 1982, 2010; Schraeder and Lathers 1983).

27.3â•…Discussion Numerous preclinical animal studies have been conducted as a model for cardiac sudden death. The animal studies presented in this chapter showed that sympathetic nerve stimulation or arrhythmic doses of ouabain or coronary occlusion increased temporal dispersion of recovery of ventricular myocardium to induce arrhythmia and/or sudden death. Cardiac

Animal Model for Sudden Cardiac Death

431

arrhythmias in the animal model for ouabain-included toxicity were associated with neural autonomic dysfunction (Lathers and Schraeder 1987; Lathers et al. 1977a, 1977b). Neural discharges were characterized by increases, decreases, or no change in the discharge of postganglionic cardiac sympathetic nerves monitored simultaneously, predisposing to cardiac arrhythmia and sudden death. Stimulation of the sympathetic ventrolateral cardiac nerve produced a shift in the origin of the pacemaker and tachyarrhythmias because the nerve is not uniformly distributed to the various regions of the heart but is localized to the atrioventricular junctional and ventricular regions (Randall et al. 1978). Such nonuniform distribution of sympathetic nerves also contributes to initiation of arrhythmia as a nonuniform neural discharge occurred. Sympathetic innervation and its role in normal and abnormal cardiac functions are discussed (Lathers et al. 1974a, 1974b, 1978; Lathers and Roberts 1980; Lathers 1980a; Lathers et al. 1986a, 1986b, 1988a, 1990). Application of this sudden death animal model to studies of epileptogenic activity and sudden death allowed my laboratory to examine physiology and pharmacology of these ἀndings in multiple animal models. The animal model described in the above studies was modiἀed, and, with the addition of EEG monitoring, led to the development of an animal model mimicking the sudden death phenomenon occurring in persons with epilepsy (Lathers and Schraeder 1980, 1982, 1990, 2010; Schraeder and Lathers 1980, 1983; Carnel et al. 1980). This modiἀed cat model explored epileptogenic activity and autonomic nerve and cardiopulmonary system relationships as an investigative tool to examine possible mechanisms of sudden unexpected death in epilepsy (SUDEP) and reported associated cardiac neural autonomic dysfunction and cardiac arrhythmias in an animal model of epilepsy. We found that subconvulsant, interictal discharges were associated with autonomic cardiac neural nonuniform discharge and cardiac arrhythmias. These epileptogenic studies established that, similar to ouabain-induced and/or coronary occlusion models of cardiac arrhythmia, neural activity was characterized by increases, decreases, or no change in the discharge of simultaneously recorded postganglionic cardiac sympathetic neural discharges in association with cortical seizure discharges. The observation of cardiac sympathetic neural nonuniform discharges in association with seizure discharges was congruent with those of Han and Moe (1964). The studies demonstrated that cardiac sympathetic neural disturbance, whether secondary to direct sympathetic nerve stimulation, ouabain toxicity, or coronary occlusion, increased temporal dispersion of recovery of ventricular excitability and led to an underlying electrical instability that predisposes the ventricular myocardium to arrhythmia. Some of these experiments also monitored vagal cardiac neural discharge with the sympathetic cardiac neural discharge, and dysfunction was evident in both divisions and/or in the time of occurrence of the dysfunctions and cardiac arrhythmias. Splanchnic nerve activity was monitored as an indicator of catecholamine release from the adrenal medulla. It was concluded that the occurrence of cardiac arrhythmias and/or death was associated with autonomic neural dysfunction within the sympathetic neural branches to the heart or within the parasympathetic nervous system or may be manifest as an imbalance between the two divisions of the autonomic nervous system, i.e., the parasympathetic and sympathetic nervous system (Lathers et al. 1977a, 1977b, 1978). In addition to the effect of ictal discharges on sympathetic and parasympathetic cardiac neural discharges, subconvulsant interictal cortical discharges induced by pentyleneÂ� tetrazol were also associated with autonomic cardiac neural nonuniform discharges and cardiac conduction and rhythm changes (Lathers and Schraeder 1980, 1982; Schraeder and Lathers 1980, 1983; Carnel et al. 1980). Subsequent studies examined the relationship of the

432 Sudden Death in Epilepsy: Forensic and Clinical Issues

effect of this interictal activity, arrhythmias, and death with altered autonomic nonuniform postganglionic cardiac and sympathetic and parasympathetic postganglionic cardiac neural discharge. Other experiments examined the effect of pretreatment with phenobarbital, ἀnding a delay in the onset time of interictal and ictal activity but no protective effect on the associated autonomic neural changes once the epileptiform discharges were established (Lathers et al. 1984; Carnel et al. 1985). The same observations were extant in a model of focal epilepsy utilizing injection of penicillin into the hippocampus of the cat (Tumer et al. 1985; Lathers and Schraeder 1990; Lathers et al. 1993). Another series of experiments (Lathers et al. 1987; Lathers and Schraeder 1990; Stauffer et al. 1989, 1990) explored the lockstep phenomenon, i.e., the synchronization of interictal cortical discharges with postganglionic cardiac sympathetic and parasympathetic neural discharges and changes of blood pressure, cardiac conduction, and rhythm, including the effects of phenobarbital on this phenomenon (Tumer et al. 1985; Lathers et al. 1985; Dodd-O and Lathers 1990). Timolol (intracerebroventricular injection) exhibited an anticonvulsant effect (Lathers et al. 1989). Subsequent experiments by Mameli et al. (1988), using hemispherectomized rats, induced epilepsy with penicillin applied to the rat hypothalamus. In this model, both interictal and ictal activity induced cardiac arrhythmias. These data were conἀrmatory of the data documenting the arrhythmogenic potential of epileptiform discharges obtained in the cat model (Lathers and Schraeder 1980, 1982; Schraeder and Lathers 1980, 1983; Carnel et al. 1980). Other experimental evidence supported the possibility of a role for prostaglandin E2€ and enkephalins in autonomic dysfunction characterized by nonuniform discharge (Suter and Lathers 1984; Kraras et al. 1987; Lathers et al. 1985, 1988c; Lathers 1990; Schwartz and Lathers 1990). The question of whether the modulation of presynaptic gamma aminoÂ� butyric acid (GABA) release by prostaglandin E2 could provide the explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias was raised (Suter and Lathers 1984). Lathers and colleagues (Kraras et al. 1987; Lathers et al. 1988c; Schwartz and Lathers 1990) inquired into whether enkephalins elicit epileptogenic activity by inhibiting the release of GABA, resulting in associated autonomic dysfunction and cardiac arrhythmias. A prolonged elevation of immunoreactive methionine (met)-enkephalin content in the septum, hypothalamus, amygdala, and hippocampus of rats occurred after pentylenetetrazol-induced convulsions (Vindrola et al. 1984) with increased concentrations of met-enkephalins associated with a greater percent inhibition of potassium-stimulated GABA release (Brennan et al. 1980). Snead and Bearden (1980) found that leucine-enkephalin in the central nervous system may induce epileptogenic activity. Thus, use of the modiἀed sudden death coronary occlusion animal model (Lathers et al. 1977a, 1977b, 1978) has provided answers to questions raised such as whether disease states alter function of this neural discharge and if the sympathetic postganglionic cardiac neural discharge represents one site for action of pharmacological agent to modify function by preventing the nonuniform neural discharge (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978; Lathers and Schraeder 1980; Lathers and Roberts 1980; Lathers 1980b; Lathers and Schraeder 1982, 1983). These two questions have been answered just recently for persons with epilepsy. A postmortem study has been conducted of postganglionic cardiac sympathy innervations in patients with chronic temporal lobe epilepsy. Druschky et al. (2001) found sympathetic dysfunction in the form or altered postganglionic cardiac sympathetic innervation in persons with chronic temporal lobe epilepsy, and this appears

Animal Model for Sudden Cardiac Death

433

to increase the risk of cardiac abnormalities and/or SUDEP. The exact role of innervation in arrhythmogenesis, as well as developmental and regulatory mechanisms determining density and pattern of cardiac sympathetic innervation, have been studied by Lathers et al. (1986a, 1986b, 1988b, 1990) and are discussed by Lathers and Levin (2010, Chapter 33). The postmortem clinical study of Druschky et al. (2001) conducted in humans conἀrms the numerous animal studies conducted by Lathers and colleagues over the years.

References Brennan, M. J., R. C. Cantrill, and B. A. Wylie. 1980. Modulation of synaptosomal GABA release by enkephalin. Life Sci 27 (12): 1097–1101. Carnel, S. B., Schraeder P. L., and C. M. Lathers. 1980. Autonomic dysfunction in epilepsy: II. Cardiovascular changes associated with pentylenetetrazol-induced seizures. Clin Res 28: 609A. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30 (4): 225–240. Dodd-O, J. H., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in phenobarbital-Â�pretreated cats. In Epilepsy and Sudden Death, Chapter 13. New York, NY: Marcel Dekker. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. James, T. N. 1983. Chance and sudden death. J Am Coll Cardiol 1 (1):164–183. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. Lathers, C. M. 1980a. Effect of timolol on autonomic neural discharge associated with ouabaininduced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106. Lathers, C. M. 1980b. The effect of metoprolol on coronary occlusion-induced arrhythmia and autonomic neural discharge. Fed Proc 39: 771. Lathers, C. M. 1981. Induced disease. Myocardial infarction in dogs and cats. In Mammalian Models for Research on Aging, 224–228. Washington, DC: National Academy Press. Lathers, C. M. 1990. Role of neuropeptides in the production of epileptogenic activity and arrhythmias. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and R. M. Levin. 2010. Animal model for sudden cardiac death. Sympathetic innervation and myocardial beta-receptor densities. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma, Chapter 33. Boca Raton: CRC Press. Lathers, C. M., and J. Roberts. 1980. Digitalis cardiotoxicity revisited. Life Sci 27 (19): 1713–1733. Lathers, C. M., and J. Roberts. 1985. Are the sympathetic neural effects of digoxin and quinidine involved in their action on cardiac rhythm? J Cardiovasc Pharmacol 7 (2): 350–360. Lathers, C. M., and W. H. Spivey. 1987. The effect of beta blockers on cardiac neural discharge associated with coronary occlusion in the cat. J Clin Pharmacol 27 (8): 582–592. Lathers, C. M., and P. L. Schraeder. 1980. Autonomic dysfunction in epilepsy: III. Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced seizures. Clin Res 28: 615A.

434 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 1990. Arrhythmias associated with epileptogenic activity elicited by penicillin. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, Chapter 9. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 2010. Animal model for sudden unexpected death in persons with epilepsy. In Sudden Death in Epilepsy. Forensic and Clinical Issues, Chapter 28. Boca Raton: CRC Press. Lathers, C. M., G. J. Kelliher, and J. Roberts. 1974a. Correlation of ouabain-induced arrhythmia and nonuniformity in cardiac accelerator nerves. Clin Res 22: 682A. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1974b. Relationship between the effect of ouabain on arrhythmia and interspike intervals (ISI) of cardiac accelerator nerves. Pharmacologist 16: 201. Lathers, C. M., K. M. Keller, J. Roberts, and A. B. Beasley. 1977a. Role of the adrenergic nervous system in arrhythmia produced by acute coronary artery occlusion. In Pathophysiology and Therapeutics of Myocardial Ischemia, ed. A. M. Lefer, G. J. Kelliher, and M. J. Rovetto. New York, NY: Spectrum. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977b. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin, J. C. Krusz, G. J. Kelliher, and J. Roberts. 1981. The action of reserpine, 6-hydroxydopamine, and bretylium on digitalis-induced cardiotoxicity. Eur J Pharmacol 76 (4): 371–379. Lathers, C. M., J. L. Gerard-Ciminera, S. I. Baskin, J. C. Krusz, G. J. Kelliher, P. B. Goldberg, and J.€Roberts. 1982. Role of the adrenergic nerve terminal in digitalis-induced cardiac toxicity: A study of the effects of pharmacological and surgical denervation. J Cardiovasc Pharmacol 4 (1): 91–98. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936. Lathers, C. M., N. Tumer, and C. M. Kraras. 1985. Cardiovascular and epileptogenic effects of penÂ� tylenetetrazol administered intracerebroventricularly in cats. Epilepsia 26: 520. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986a. Regional distribution of myocardial betaadrenoceptors in the cat. Eur J Pharmacol 130 (1–2): 111–117. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986b. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39 (22): 2121–2141. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., W. H. Spivey, and N. Tumer. 1988a. The effect of timolol given ἀve minutes after coronary occlusion on plasma catecholamines. J Clin Pharmacol 28 (4): 289–299. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988b. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28 (8): 736–745. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988c. The effect of intracerebroventricular d-ALA2 methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43 (26): 2287–2298.

Animal Model for Sudden Cardiac Death

435

Lathers, C. M., A. Z. Stauffer, N. Tumer, C. M. Kraras, and B. D. Goldman. 1989. Anticonvulsant and antiarrhythmic actions of the beta blocking agent timolol. Epilepsy Res 4 (1): 42–54. Lathers, C. M., W. H. Spivey, R. M. Levin, and N. Tumer. 1990. The effect of dilevalol on cardiac autonomic neural discharge, plasma catecholamines, and myocardial beta receptor density associated with coronary occlusion. J Clin Pharmacol 30 (3): 241–253. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital on autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33 (9): 837–844. Mameli, P., O. Mameli, E. Tolu, G. Padua, D. Giraudi, M. A. Caria, and F. Melis. 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29 (1): 74–82. Randall, W. C., J. X. Thomas, D. E. Euler, and G. L. Rosanski. 1978. Cardiac dysrhythmias associated with autonomic nervous system imbalance in the conscious dog. In Perspectives in Cardiovascular Research, ed. P. J. Schwartz, A. M. Brown, A. Malliani, and A. Zanchetti. New York, NY: Raven Press. Schraeder, P. L., and C. M. Lathers. 1980. Autonomic dysfuction in epilepsy: I. A proposed animal model for unexplained sudden death in epilepsy. Clin Res 28: 618A. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schwartz, R. D., and C. M. Lathers. 1990. GABA neurotransmission, epileptogenic activity, and cardiac arrhythmias. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, Chapter 17. New York, NY: Marcel Dekker. Snead, 3rd, O. C., and L. J. Bearden. 1980. Anticonvulsants speciἀc for petit mal antagonize epileptogenic effect of leucine enkephalin. Science 210 (4473): 1031–1033. Spivey, W. H., and C. M. Lathers. 1985. Effect of timolol on the sympathetic nervous system in coronary occlusion in cats. Ann Emerg Med 14 (10): 939–944. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1990. Relationship of the lockstep phenomenon and precipitous changes in blood pressure. In Epilepsy and Sudden Death, Chapter 14. New York, NY: Marcel Dekker. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15 (1): 15–30. Tumer, N., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital upon autonomic function and epileptogenic activity induced by hippocampal injection of penicillin in cats. Epilepsia 26: 520. Vindrola, O., M. Asai, M. Zubieta, E. Talavera, E. Rodriguez, and G. Linares. 1984. Pentylenetetrazol kindling produces a long-lasting elevation of IR-Met-enkephalin but not IR-Leu-enkephalin in rat brain. Brain Res 297 (1): 121–125.

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

28

Claire M. Lathers Paul L. Schraeder

Contents 28.1 28.2 28.3 28.4 28.5

Introduction Postmortem Pulmonary Findings Tissue Hypoxia, Hypercarbia, and Alterations in Acid–Base Balance Alterations in Cerebral Blood Flow Possible Mechanisms for Autonomic Dysfunction and Sudden Death in Epileptic Persons 28.5.1 The Lockstep Phenomenon 28.5.2 Simultaneous Electroencephalogram and ECG Recordings in Humans 28.5.3 Role of Stress in SUDEP 28.5.4 Central Biochemical Changes Associated with Epileptogenic Activity 28.6 Summary Acknowledgments References

437 447 447 447 447 447 451 452 453 456 457 457

28.1â•…Introduction Six years after Lathers developed two different animal models to study the role of the autonomic nervous system in arrhythmias and death due to digitalis toxicity and coronary occlusion (Lathers et al. 1974b, 1977a, 1977b; Lathers 1980a, 1981a; Lathers and Roberts 1980; Lathers 2010, Chapters 25 and 27, this book; Lathers et al. 2010a, Chapter 1, this book; Lathers and Levin 2010, Chapter 33, this book), these animal models were modiἀed to study epileptogenic activity and autonomic cardiac neural discharge in a new animal model in collaboration with Dr. Schraeder to explore the role of neurogenic cardiac arrhythmias in sudden unexplained death in epilepsy (SUDEP) (Lathers and Schraeder 1982; Schraeder and Lathers 1983). Our unique model was designed to examine autonomic cardiac neural discharge and arrhythmias and/or death associated with epileptogenic activity since paroxysmal autonomic dysfunction is one possible explanation for SUDEP. Our thinking was, in part, based on the ἀndings of Terrence et al. (1975), who listed three possible causes of SUDEP and noted one postmortem ἀnding (Table 28.1). To examine the role of altered autonomic function in SUDEP, we designed a study using anesthetized cats. Varying doses of the epileptogenic agent pentylenetetrazol were used. The pentylenetetrazol animal model of epilepsy is an established method to produce generalized interictal and ictal epileptogenic activity (Hahn 1960). Cardiovascular 437

438 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 28.1â•… Three Possible Causes of SUDEP and One Postmortem Finding 1. Exhaustion of the heart and failure of respiration 2. Fatal syncope caused by acute disruption of the brain stem cardiac and/or respiratory functions 3. Paroxysmal cardiac autonomic dysfunction Postmortem ἀnding: anticonvulsant drug levels in blood or other body fluids of victims is often either subtherapeutic or absent. Source: Terrance et al., Neurology, 25, 594–598, 1975. With permission.

changes, including blood pressure, heart rate, and electrocardiogram (ECG) changes, associated with interictal and ictal activity were studied to determine if alterations in interictal and ictal activity correlated with nonuniform discharge (Lathers et al. 1977a, 1977b, 1978; Lathers 1980a, 1981b, 2010; Lathers and Schraeder 2010a). Although most victims of SUDEP succumb during a generalized seizure, Lathers and Schraeder hypothesized that detection of arrhythmia with subconvulsant interictal activity would provide a possible explanation for those persons with epilepsy who die unexpectedly during minimal clinical seizure activity or even without observed seizure activity. By deἀnition, persons with SUDEP have no deἀning pathological abnormalities found on autopsy. This model also allowed us to examine cardiac autonomic neural discharge associated with ictal and interictal epileptogenic activity. Cats were anesthetized with alpha-chloralose and tracheostomies were inserted. The femoral arteries and veins were cannulated to monitor the mean arterial blood pressure and to administer drugs, respectively. Intermittent intravenous doses of gallamine (4 mg/ kg) maintained paralysis while the cats were on a small-animal respirator. The lead II ECG was simultaneously monitored. The postganglionic cardiac sympathetic and right cardiac vagal nerve branches were isolated, and electrical activity was recorded using the technique of Lathers et al. (1978). After a 10-min control period, each cat was administered intravenous pentylenetetrazol, which was given every 10 min in doses of 10, 20, 50, 100, 200, and 2000 mg/kg, intravenous. These doses elicited one of three categories of epileptogenic activity. The ἀrst, designated prolonged ictal activity, was characterized by polyspike discharges greater than 10 s in duration; the second was brief ictal activity, characterized by groups of polyspikes of less than 10 s duration; and the third, designated interictal activity, was characterized as spike or polyspike complexes occurring at a rate of about 1/s. Interictal activity was often interspersed between the prolonged and brief ictal discharges. All categories of epileptogenic activity were quantiἀed. Interictal activity was counted as spikes per minute, and both types of ictal activity were measured as duration of ictal discharges occurring in each minute. In eight of nine cats, 10 mg/kg pentylenetetrazol elicited only interictal activity. One of nine cats exhibited prolonged ictal activity while another manifested no epileptogenic activity. A pentylenetetrazol dose of 20 mg/kg elicited ictal activity in most animals. With increasing doses of pentylenetetrazol, the amount of interictal activity decreased and the duration of ictal activity increased. Maximum duration of ictal activity occurred after doses of 100 or 200 mg/kg pentylenetetrazol. Interictal activity elicited by 10 mg/kg pentylÂ�enetetrazol was associated with a brief but signiἀcant decrease in the mean heart rate lasting for 1 min; little or no change in heart rate was noted thereafter. Little change in heart rate occurred with the development of interictal and/or ictal activity developing after 20€mg/kg pentylenetetrazol. As epileptogenic activity increased following higher doses of

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

439

pentylenetetrazol, heart rate increased signiἀcantly. Mean arterial blood pressure signiἀcantly increased with occurrence of interictal discharge. This trend continued as epileptogenic activity increased. Although the initial injection of each dose of pentylenetetrazol caused a decrease in mean arterial blood pressure, the overall tendency of the blood pressure was to increase. The interictal activity elicited by 10 mg/kg pentylenetetrazol was associated with a variety of ECG changes, including alterations in P, T, and Q waves; the QRS complexes; and ventricular tachycardia. In one animal, premature ventricular contractions occurred. With increasing degrees of both interictal and ictal activity, the number of animals exhibiting the foregoing ECG changes increased and some additional animals exhibited premature atrial contractions. The total number of animals exhibiting premature ventricular contractions increased with increasing duration of epileptogenic activity. Lathers et al. (1983, 1987) ἀrst reported the ἀnding that cardiac sympathetic and vagal neural discharges were intermittently synchronized 1:1 with epileptogenic discharge, i.e., the lockstep phenomenon. Statistical analysis was performed to examine the relationship between the lockstep phenomenon and the occurrence of a given electrocorticogram change. We note that better electrophysiological techniques are now available to do the analysis of ECG changes. We also note that the statistical analysis used may not have been ideal to examine the relationship between the lockstep phenomenon and the occurrence of a given electrocorticogram change. Use of statistical analysis for small numbers would also be appropriate for future analysis (Lathers 2002; Louie et al. 2008; Beauchemin et al. 2008). The relationship between lockstep phenomenon and the occurrence of a given electrocorticogram change was analyzed. For each electrocorticogram abnormality, a separate 2 × 2 table was created for each nerve monitored in every cat. Every minute was classiἀed as either containing lockstep phenomenon or not and as exhibiting the electrocorticogram abnormality or not. The number of minutes with each of the four possible “yes/no” combinations were the entries in the 2 × 2 table. A separate 2 × 2 table was created for doses of 10 and 20 mg/kg intravenous and for doses of 50, 100, and 200 mg/kg intravenous pentylÂ� enetetrazol. Since most of the interictal activity occurred with 10 and 20 mg/kg pentylÂ� enetetrazol and most of the brief or prolonged ictal activity occurred after 50, 100, and 200 mg/kg pentylenetetrazol, the formation of these two groups allowed comparison of ECG changes associated with interictal and ictal activity. The phi coefficient was computed as a measure of the relationship between the two variables. A +1.0 phi value indicates a perfect relationship; a value of −1.0 indicates a perfect inverse relationship; and a value of 0€indicates no relationship. These phi values were averaged for all nerves monitored in each subject. Averages for each subject were then used as the unit of analysis. Although many of the individual 2 × 2 tables have extreme marginal totals, e.g., very few minutes containing a given abnormality, the average of several nerves for each subject was more reliable, especially when averaged across all subjects for reporting. For each 10-min interval, the presence or absence of the lockstep phenomenon was determined, i.e., when the neural burst discharge pattern of postganglionic cardiac sympathetic or cardiac vagal nerves began to be synchronized with cortical interictal and/or ictal discharges. A count was made of the number of intradose intervals with interictal lockstep found during the 10 min between each dose of pentylenetetrazol. Similar counts were performed for brief and prolonged ictal lockstep phenomenon. A count of the number of 10-min intervals with each type of abnormality in the electrocorticogram was also done. ECG abnormalities evaluated were changes in the P–R intervals, P waves, T waves, and QRS complexes; or the appearance of Q waves, premature atrial contractions, premature

440 Sudden Death in Epilepsy: Forensic and Clinical Issues

ventricular contractions, ventricular tachycardia, atrial ἀbrillation, ventricular ἀbrillation, asystole, or cardiovascular collapse. The number of intradose intervals in which lockstep phenomenon occurred was then correlated across all subjects with the number of intervals in which each electrocorticogram abnormality appeared. The correlations performed both excluded and included the 2000 mg/kg dose of pentylenetetrazol since this dose induced death in all but one cat, which died after 200 mg/kg pentylenetetrazol. Furthermore, if a variable was constant across subjects, or nearly so, it was not entered into the correlations. When reviewing statistical analysis of any data set, one must always evaluate if the statistical results “make biological sense.” For example, there are instances when an observation of change in an ECG occurs with and/or immediately after changes in the electroencephalogram (EEG) when these parameters are simultaneously monitored in one animal. However, when the data from a group of animals are statistically analyzed together, statistical signiἀcance is not found. Very large numbers of animals would have to be studied to ἀnd statistical signiἀcance, and this is not always feasible, nor is it possible, with the time-consuming and expensive nature of each in vivo animal study in our model. An example of statistical vs. biological ἀndings comparison occurred when we analyzed the ECG changes associated with the EEG ἀndings during the lockstep phenomenon. The P–R interval, T€wave, P wave, and QRS complex, and occurrence of premature ventricular contractions, were analyzed over each 15-min period occurring after a dose of pentylÂ� enetetrazol was administered. We did not have the means to analyze our data beat by beat to compare each heartbeat with the existing EEG pattern. When the dose of 2000 mg/kg of pentylenetetrazol was excluded, 40 correlations were generated, and of these only three were signiἀcant one-tailed. However, this number, in and of itself, is not substantially different from the two that would have been anticipated by chance; thus, it was concluded that no statistically signiἀcant relationship between the occurrence of the lockstep phenomenon and the electrocorticogram abnormalities was demonstrated for our data. When the dose of 2000 mg/kg was included, only 3 of 80 correlations were signiἀcant one-tailed in the predicted direction. Several correlations with atrial ἀbrillation and asystole would have attained signiἀcance, but they were negative, suggesting that the greater frequency of the lockstep phenomenon was related to less frequent occurrence of the electrocorticogram patterns. Thus, statistical analysis indicated that, in general, there was no statistically signiἀcant correlation between the occurrence of the lockstep phenomenon and the ECG changes “lumped” together for 15-min periods. However, while observing the oscilloscope and the polygraph recordings and listening to the audio of the neural discharges monitored simultaneously with the EEG and ECG, it was obvious that biological data revealed there correlations of the EEG changes and the autonomic cardiac nerve discharges occurring with changes in the ECG, including atrial ἀbrillation and asystole, in a given animal. These changes occurred during EEG changes in the manner of the lockstep phenomenon. Thus, these changes were immediately obvious to all viewing the oscilloscope and the polygraph recordings and listening to the audio of the neural discharges. Nevertheless, when data were statistically grouped together for 15-min periods for data obtained in animals that did and those that did not show the lockstep phenomenon, statistical signiἀcance was not found. This suggests statistically that the greater frequency of the lockstep phenomenon was related to less frequent occurrence of ECG patterns. Thus, although some individual animals exhibited ECG changes in the presence of the lockstep phenomenon, the majority did not. During statistical analysis, if a variable was constant across subjects, or nearly so, it was not entered into the correlations. In hindsight, we think that this should not have been

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

441

done. Consistent ECG changes occurring with EEG and neural discharge changes should have been analyzed. The above statistical analysis provides an important lesson for those working with biological models while also raising an important concern when looking for possible mechanistic risk factors for SUDEP in both animal and human studies. When correlating electrophysical phenomena with cause-and-effect sequences that attempt to explain transient events that result in fatal outcome, i.e., SUDEP or cardiac arrhythmias leading to fatality, one cannot depend entirely on the statistical analysis of data resulting from population-based studies. The reason for making this statement is that unique individual susceptibility for a potentially fatal sequence or combination of events may not hold for other individuals in the group. Hence, one must exercise caution in excluding individual phenomenological observations that appear not to be statistically signiἀcant. Rather, one should also consider a clinically based modality of analysis, namely, that the diagnosis is likely to be tied to the single observation that does not hold to the expected pattern. This statement means that when EEG abnormalities that have the potential to interfere with normal cardiac function occur, such disturbances do not manifest in the overwhelming number of individuals at risk. They occur only in such individuals who have a substrate, whether transient or built in genetically, for a potentially fatal event. This would explain why only a few, and not most, individuals with epilepsy succumb to SUDEP. At the time these data were analyzed, we did not have the more sophisticated approach discussed by Verrier and Schachter (2010) in Chapter 43. Their comment is that T-wave alternans, a beat-to-beat fluctuation in the morphology of the T wave on the ECG, is linked to susceptibility to malignant ventricular tachyarrhythmias. We do not know if there is any association between increased occurrences of T-wave alternans in the presence of the lockstep phenomenon, making this a potential subject for future investigation. Figure 28.1 illustrates data obtained in one cat. Interictal discharge was elicited by 10 mg/kg pentylenetetrazol and was associated with increased mean arterial blood pressure. The autonomic sympathetic response was atypical in that the neural discharge in one of the sympathetic nerves was decreased, while that in a second nerve was increased. The expected, predicted physiological response for all sympathetic neural discharge to be changed in the same direction (Berne and Levy 1967). Such a typical physiological sympathetic neural response elicited in our studies is depicted in Figure 28.2. In each experiment, a positive identiἀcation of the function of the cardiac postganglionic sympathetic nerves was performed. During the control period, before induction of epileptogenic activity and associated changes in the ECG, a test dose of 5 μg/kg of histamine was given intravenously. Histamine caused the blood pressure in each cat to decrease, with a concomitant increase in the autonomic sympathetic cardiac neural discharge recorded in both branches. As blood pressure returned to the pre-histamine level, sympathetic discharge in both branches decreased. This response was always obtained in all postganglionic sympathetic branches during the control period; however, as depicted in Figure 28.1, when the blood pressure increased with the onset of interictal activity after 10 mg/kg pentylenetetrazol, the neural discharge in only one sympathetic nerve decreased. Discharge in the second nerve increased. This is an example of autonomic dysfunction in the discharge of the cardiac postganglionic sympathetic nerves. The asterisk in Figure 28.1 indicates occurrence of autonomic sympathetic neural dysfunction that was also associated with an alteration in the T-wave amplitude and the P–R interval in the ECG. The results obtained in the experimental animal model provide the

442 Sudden Death in Epilepsy: Forensic and Clinical Issues

600

Neural discharge (impulses/s) % control

500

400

300

S1

200 V

Mean arterial blood pressure (mm Hg)

100

S2

80 0

120 100 80

Heart rate (beats/min)

0 200 150

Seizure Interictal activity (spikes/min)

100 0 100 80 60 40 20 0 PI BI 0

10

20

30

40

10

20

50

100

200

50 Time (min) 200 PTZ (mg/kg)

Figure 28.1╇ Autonomic cardiac neural discharge and epileptogenic activity as a function of time (minutes) for data obtained in one cat. Neural discharge (impulses/second, percentage of control) is illustrated for two postganglionic cardiac sympathetic branches (S1, S2) and for vagus nerve (V) in first graph. Mean arterial blood (mm Hg) and heart rate (bpm) are illustrated as a function of time in the second and third graphs from the top, respectively. Interictal activity in spikes/minute is illustrated in the fourth graph, and duration of ictal activity (prolonged ictal, solid boxes; brief ictal, crosshatched boxes) in seconds is seen in the bottom graph. Arrows along the abscissa indicate injections of pentylenetetrazol at 10, 20, 50, 100, 200, and 2000 mg/kg intravenous, which were given every 10 min. Asterisk indicates the first occurrence of change in ECG. (From Lathers, C. M., and Schraeder, P. L., Epilepsia, 23, 633–647, 1982. With permission.)

(impulse/s)

443

100 50

(impulse/s)

(impulse/s)

0

Mean arterial blood pressure (mm Hg)

Sympathetic neural discharge

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

50 25 0 100 50 0 200 100 1 min 0

Histamine (5 µg/kg)

Figure 28.2╇ Sympathetic neural discharge (upper two panels) and mean arterial blood pressure response monitored in one cat. Arrow indicates injection of histamine (5 μg/kg intravenous). (From Lathers, C. M. and Schraeder, P. L., J Clin Pharmacol, 27, 582–592, 1987. With permission.)

basis for one possible pathophysiological event to explain the sudden autonomic dysfunction in persons with epilepsy who had no observed clinical seizure and only seizures of minimal severity preceding their death (Hirsch and Martin 1971; Jay and Leestma 1981; Spratling 1902; Terrence et al. 1975; Iivanainen and Lehtinen 1979). Our data suggest that interictal nonconvulsive activity may be associated with cardiac neural dysfunction with consequent abnormalities of conduction and rhythm. This would make certain persons with epilepsy susceptible to developing fatal arrhythmias. The cardiac neural and cardiovascular alterations were also observed with ictal activity and at the time of death (Lathers and Schraeder 1987; Schraeder and Lathers 1989). Mean parasympathetic and sympathetic neural discharge for all animals was calculated. An increase in the standard deviation of mean parasympathetic neural discharge occurred with the onset of interictal discharge. The development of increasing degrees of epileptogenic activity was associated with a further increase in the standard deviation. By deἀnition, a large standard deviation of the mean indicates that there is a large range among the values included within the mean. Thus, the neural discharge in some parasympathetic nerves was increased above control, whereas in others it was decreased or showed little or no change. These changes demonstrated autonomic dysfunction within the parasympathetic division. A very large standard deviation was also seen in the mean sympathetic neural discharge, although it did not occur until after the large standard deviation developed in the mean parasympathetic

444 Sudden Death in Epilepsy: Forensic and Clinical Issues

discharge, demonstrating an imbalance between the two divisions of the autonomic nervous system. Thus there was autonomic dysfunction within both the parasympathetic and the sympathetic divisions of the autonomic nervous system and in an imbalance between the two divisions of the autonomic nervous system. Two earlier studies (Onuma 1957; Orihara 1952) examined subconvulsive or convulsive doses of pentylenetetrazol actions on cervical sympathetic and parasympathetic autonomic discharge. Increases and/or decreases in the neural discharge within both divisions of the autonomic nervous system occurred, just as recorded for the autonomic cardiac nerves in our studies. Onuma (1957) emphasized that sympathetic neural activity was hyperexcited and that increased ictal activity was associated with an imbalance between the sympathetic and parasympathetic discharge. Their evidence supports our ἀndings for the cardiac autonomic nerves. Schraeder and Celesia (1977) found that minimal subclinical interictal epileptogenic activity had a wide-ranging effect on cerebral function when they monitored activity from the auditory cortex in the cat. Lathers and Schraeder (1982) and Schraeder and Lathers (1983) theorized that if the recorded interictal activity found in our new animal model also had a similar effect on regions of the cerebrum involved in autonomic regulation, an autonomic imbalance in cardiac neural discharge and arrhythmias might occur. These changes could be one factor contributing to sudden unexplained death in the epileptic patient. We then raised the question of whether a drug could be found to eliminate the autonomic cardiac neural discharge dysfunction and the associated arrhythmias. Both phenÂ� ytoin (Gillis 1971; Roberts 1970; Schlosser et al. 1975) and chlordiazepoxide (Schallek and Zabransky 1966) depressed the cardiac sympathetic neural discharge and abolished associated cardiac arrhythmias (Evans and Gillis 1974, 1975; Gillis 1969, 1971; Gillis et al. 1972, 1974; Pace and Gillis 1976; Raines et al. 1970). We hypothesized the pharmacologic agent that would provide the best protection against autonomic dysfunction associated with the epileptogenic activity is one that exhibits anticonvulsant, antiarrhythmic, and cardiac neural depressant activity (Lathers and Schraeder 1982). Because phenobarbital exhibits both anticonvulsant and antiarrhythmic properties, we decided to examine this agent in our experimental model. Pentylenetetrazol, an analeptic agent, was developed to stimulate the central nervous system after barbiturate intoxication (Carnel 1982). It has been used as an EEG-activating agent in diagnosis of epilepsy, as a convulsant agent in therapy of schizophrenia (Hahn 1960), and as a screening agent to test for new antiepileptic drugs (Faingold and Berry 1973). Subthreshold doses result in increased cortical excitability without overt clinical seizure manifestations and correlates with interictal discharge in our animal model. Threshold doses of pentylenetetrazol produce motor activity characterized by forelimb clonus, resembling convulsion produce by electrical stimulation of brain with current of just threshold intensity (Franz 1980) and correlated with prolonged activity in our model. Large doses of pentylenetetrazol produce generalized asynchronous clonic movement followed by a tonic convulsion resembling supramaximum brain stimulation such that movements of limbs are flexion followed by extension (Franz 1980) and were equal to brief ictal activity in our model. Pentylenetetrazol action is on the central nervous system rather than on peripheral organs (Toman and Davis 1949; Hahn 1960). Therefore, changes in blood pressure, heart rate, ECG, and autonomic neural discharge is perceived to be mediated via a central mechanism and permitted evaluation of our hypothesis that alterations in central

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

445

activity produce centrally mediated autonomic dysfunction, which may explain sudden death in epileptics. In our model, the lethal dose of pentylenetetrazol was determined to be 2000 mg/kg. We needed to ἀnd the lethal dose to examine electrocorticogram, blood pressure, heart rate, ECG, and autonomic neural discharge correlating with activity just before death with the goal of evaluating our hypothesis that autonomic dysfunction contributes to sudden death. Phenobarbital is used to treat generalized tonic–clonic and simple partial seizures in humans. Its interaction with pentylenetetrazol has been studied by many using anesthetized preparations. Phenobarbital has been examined as a central nervous depressant whose action is antagonized by pentylenetetrazol (Hahn 1960) and as a protective agent vs. pentylenetetrazol-induced seizures (Faingold and Berry 1973). The dose of phenobarbital (20 mg/kg), mode of administration (intravenous), and the time interval allotted for stabilization of drug distribution within the body after the end of infusion (60 min) was selected after review of the literature and preliminary evaluation. Concentrations of phenobarbital that were administered intravenously by others (LaManna and Rosenthal 1975; Nakamura and Kurebe 1962; LaManna et al. 1977; Ito et al. 1977) ranged from 5 to 80 mg/kg in the cat. Peak blood levels were reached in a few minutes (LaManna et al. 1977; Nakamura and Kurebe 1962) and distribution between white and gray matter was stable after 1 h (Domek et al. 1960). In our study, the dose initially used was 30 mg/kg, intravenous, by slow infusion over 10 min. This dose produced a marked depression in blood pressure from 127/87 to 47/36 mm Hg. Since this dose interferes with cerebral perfusion (Guyton 1969; Dittmer and Grebe 1959), the dose was reduced to 20 mg/kg. Phenobarbital caused signiἀcant depression of postganglionic cardiac sympathetic neural discharge and a decrease in the mean arterial blood pressure before the administration of pentylenetetrazol. The direction of this change in the sympathetic neural discharge was the opposite of the predicted response to a decrease in blood pressure. Neural depression continued for the next 70 min. In contrast, phenobarbital did not signiἀcantly alter the parasympathetic neural discharge from control. Comparison of the parasympathetic neural discharge monitored in the animals pretreated with phenobarbital and in those receiving only pentylenetetrazol found no signiἀcant difference in the standard errors of the mean (Figure 28.3). The sympathetic neural discharge monitored after pentylenetetrazol in animals pretreated with phenobarbital did not differ from that monitored in cats given only pentylenetetrazol (Figure 28.4). Phenobarbital exhibited an anticonvulsant effect since it prevented the interictal subconvulsant epileptogenic activity elicited by 10 mg/kg pentylenetetrazol and shifted the dose–response curve to the right. The duration of ictal activity was decreased by phenobarbital. Phenobarbital pretreatment did not exhibit an antiarrhythmic effect. The incidence of cardiac arrhythmias and the types of arrhythmias were the same in animals with and without phenobarbital. Phenobarbital initially depressed the sympathetic discharge but did not modify parasympathetic neural discharge. As increasing doses of pentylÂ� enetetrazol induced epileptogenic activity, there was no signiἀcant difference between the two groups. Our data suggest that phenobarbital, at the dose studied, is not a useful agent for the prevention of autonomic neural dysfunction and cardiac arrhythmias associated with epileptogenic activity in this cat model. Since phenobarbital (20 mg/kg intravenous, 60 min before pentylenetetrazol) exhibited an anticonvulsant action but did not exhibit antiarrhythmic and neural depressant activity, this antiepileptic drug, at the dose studied,

446 Sudden Death in Epilepsy: Forensic and Clinical Issues PTZ alone (n=8 nerves, 7 cats)

Parasympathetic neural discharge (mean ± S.E., impulses/s) % control

Phenobarbital pretreatment (n=10 nerves, 9 cats)

200 150 100 50 0

0

10

20

30

40

50 Time (min)

10

20

50

100

200

2000 PTZ (mg/kg)

Figure 28.3╇ Mean ± SE for parasympathetic neural discharge monitored in seven cats (eight

nerves) receiving only pentylenetetrazol is depicted by a white area that became black when the values overlapped with those obtained in the phenobarbital group. Data obtained in only nine cats (10 nerves) pretreated with phenobarbital before administration of pentylenetetrazol are depicted by gray area. (From Lathers, C. M. and Schraeder, P. L., J Clin Pharmacol, 27, 346–356, 1987. With permission.)

Sympathetic neural discharge (mean ± S.E., impulses/s) % control

300

PTZ alone (n=18 nerves, 9 cats)

250

Phenobarbital pretreatment (n=17 nerves, 9 cats)

200 150 100 50 0

0

10

20

30

40

10

20

50

100

200

50

Time (min)

2000 PTZ (mg/kg)

Figure 28.4╇ Mean ± SE for sympathetic neural discharge monitored in nine cats (18 nerves)

receiving only pentylenetetrazol is depicted by white area; values in this group overlapping those obtained for the cats (17 nerves) pretreated with phenobarbital are designated by black area. Areas depicting only values from the phenobarbital group are designated by gray area. (From Lathers, C. M. and Schraeder, P. L., J Clin Pharmacol, 27, 346–356, 1987. With permission.)

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

447

does not appear to be a useful agent to prevent the autonomic dysfunction associated with epileptogenic activity in this animal model. Future studies are needed.

28.2â•… Postmortem Pulmonary Findings Postmortem, we observed pulmonary ἀndings of multiple areas of punctuate hemorrhages and large areas of gross hemorrhage and edema in animals dying after inducing epileptogenic activity, asystole, or ventricular ἀbrillation (Lathers et al. 1984; Carnel et al. 1985).

28.3â•…Tissue Hypoxia, Hypercarbia, and Alterations in Acid–Base Balance Tissue hypoxia and hypercarbia and alterations in acid–base balance may have contributed to the results in our model of experimental epilepsy. Acid–base balance was maintained within the physiological range before initiation of epileptogenic activity (Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1983, 1989).

28.4â•… Alterations in Cerebral Blood Flow Changes in cardiac function lead to alterations in cerebral blood flow, which in turn produce central hypoxia resulting in epileptogenic activity. Some patients exhibit changes in cardiovascular status preceding the onset of convulsions (Schott et al. 1977; Schraeder and Lathers 1983).

28.5â•…Possible Mechanisms for Autonomic Dysfunction and Sudden Death in Epileptic Persons 28.5.1â•… The Lockstep Phenomenon Lathers et al. (1983) and Lathers and Schraeder (1987) ἀrst reported the ἀnding that cardiac sympathetic and vagal neural discharges were intermittently synchronized 1:1 with epileptogenic discharge, i.e., the lockstep phenomenon. This relationship was designated as real when it was locked 1:1 and semilocked when the relationship was almost 1:1. The abnormal cardiac neural discharge and cardiac arrhythmias were associated with subconvulsant interictal activity. It was suggested that if sudden cardiac arrhythmias are a cause of sudden unexpected death, then the lockstep phenomenon may be a factor in the mechanism of unexplained death in persons with epilepsy who exhibited no overt seizure activity at the time of demise (Lathers et al. 1987). Stauffer et al. (1989) found that a higher mean proportion of time was spent in precipitous changes in blood pressure, i.e., <23 mm Hg in a 10-s interval, in association with an unstable lockstep phenomenon pattern. An unstable lockstep phenomenon pattern was deἀned as all time intervals of 10 s or more during which the lockstep phenomenon existed but the interspike intervals were not constant. Dodd-O and Lathers (1990) also noted that when stable lockstep phenomenon was lost, both precipitous mean arterial

448 Sudden Death in Epilepsy: Forensic and Clinical Issues

blood pressure changes and the incidence of ECG changes occurred more frequently. They suggested that development of the abnormal rhythmic activity of the unstable lockstep phenomenon may alter neurotransmitter release and initiate autonomic dysfunction, thereby having a possible contributory role in sudden unexplained death in epileptic persons. The importance of central neuronal outflow in altering the peripheral efferent discharge to the heart is well recognized as is the observation that efferent sympathetic discharge can initiate abnormalities in the ECG (Lathers et al. 1977a, 1977b, 1978). Development of lockstep phenomenon may precede occurrence of changes in the ECG or vice versa. All of these lockstep phenomenon studies (Lathers et al. 1983, 1987; Lathers and Schraeder 1990a; Dodd-O and Lathers 1990, Chapter 29, this book; Stauffer et al. 1989, 1990, Chapter 30, this book; O’Rourke and Lathers 1990, Chapter 31, this book; Lathers and Schraeder 2010a) suggest that at least four mechanisms can be postulated through which the lockstep phenomenon may be related to arrhythmia and SUDEP. Deἀnitive experiments should be done to verify the following possibilities as being operative as possible mechanisms of SUDEP: 1. Excessive stimulation of an electrically unstable heart previously damaged 2. Occurrence of nonuniform postganglionic cardiac sympathetic discharge or an imbalance between the sympathetic and parasympathetic neural innervation of the heart 3. Sinus arrest and bradycardia associated with seizures and induced by the parasympathetic nervous system over activity 4. Development of abnormally precipitous blood pressure changes. In Chapter 26, Alkadhi and Alzoubi (2010) note that Nei et al. (2000) found that persons who€were victims of SUDEP had premorbid evidence of increased cardiac autonomic stimulation, adding weight to the argument that cardiac arrhythmias during and between seizures, with transmission of epileptic activity to the heart through the autonomic nervous system, may play a signiἀcant role in SUDEP. Parasympathetic nervous system functions in many ways as the yang for the yin of sympathetic effect on the cardiovascular system. Nonetheless, excessive vagal tone consequent to epileptiform activity can result in the occurrence of asystole (Kiok et al. 1986; Lathers and Schraeder 1982). In addition, although a stress response (Pickworth et al. 1990; Lathers and Schraeder 2006a, 2010b, Chapter 17) is commonly associated with increased risk for sympathetically mediated cardiovascular dysfunction, parasympathetic-mediated response in the form of stress related asystole can also occur (Schraeder et al. 1983). Thus both branches of the autonomic nervous system may have potential roles in the mechanism of SUDEP. The lockstep phenomenon ἀrst described by Lathers and Schraeder (1987), Staufer et al. (1989, 1990), Dodd-O and Lathers (1990), and O’Rourke and Lathers (1990) was manifested predominately as cardiac neural sympathetic discharge that was locked to the occurrence of cerebral epileptiform events. However, these studies also found a less frequent, but nonetheless documented, occurrence of vagus nerve branch activity that was time locked to these same cerebral discharges. One question that could be raised by these observations is that of whether the parasympathetic events are potentially adverse or are they, in a sense, a counterpoint to the more prominent sympathetic lockstep phenomenon. Another question is that of why were only some experimental animals manifesting the parasympathetic lockstep phenomenon. Is this observation a hint that some, but not all, individuals have more likelihood of this cerebral discharge–locked parasympathetic output? If this is the case, are these individuals

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

449

at less risk for sympathetically mediated arrhythmias than other individuals? Or are those with parasympathetic lockstep phenomenon at more risk for asystole? We do not know the answer. One comment: although the lockstep phenomenon and ECG changes were not a statistically observable predictable sequential correlation, this is not unlike the unpredictability of fatal arrhythmias. One premature ventricular contraction may, at a given time, be fatal although the patient has experienced others and€not have a fatal outcome. There is a correlation with the nonuse of a seat belt while driving a car. As one drives through a four-way traffic intersection, it may be done safely for seven times without use of the seat belt, but, at the eighth intersection, a car running a red light may hit the car containing a person without a seat belt and a fatal event may occur. Stauffer et al. (1989, 1990) reported precipitous mean arterial blood pressure changes were correlated with unstable lockstep phenomenon (vide infra). Four possible mechanisms through which lockstep phenomenon may be related to arrhythmia and sudden death in persons with epilepsy were postulated: 1. Excessive sympathetic stimulation of a heart that is already electrically unstable due to prior damage (Jay and Leestma 1981). 2. A nonuniform (simultaneous increases, decreases, or no change) discharge in the postganglionic cardiac sympathetic nerve branches (Lathers et al. 1977b, 1978). An autonomic dysfunction, i.e., an imbalance between sympathetic and parasympathetic discharge contributes to occurrence of arrhythmias and/or death. 3. The parasympathetic nervous system causing sinus arrest and bradycardia during seizures (Kiok et al. 1986; Lathers and Schraeder 1982). 4. The associated precipitous blood pressure changes per se. Stimulation of the hypothalamus elicits blood pressure increases. Arrhythmias ocÂ�curred after but not during stimulation and resulted from a sudden surge of paraÂ�sympathetic activity reflexively evoked by the rapid increase in blood pressure (Evans and€Gillis 1978). The proposed four mechanisms that may lead to arrhythmias and sudden death are not mutually exclusive. It is possible that no single mechanism can explain all cases of sudden death in epilepsy. Some cases of sudden death may result from ventricular ἀbrillation related to a lowered ventricular ἀbrillation threshold associated with increased sympathetic discharge. Other cases may result from sinus arrest related to reflex paraÂ�sympathetic discharge evoked by precipitous blood pressure changes, especially if€there is cardiac damage produced by multiple prior episodes of excessive sympathetic stimulation. Polosa et al. (1969) reported ἀndings similar to the lockstep phenomenon described in this study. Spontaneous, rhythmic, synchronized activity of the sympathetic cervical nerve; the cervical phrenic nerve; and small strands of the sciatic nerve occurred after administration of picrotoxin. That complex interactions exist between afferent and efferent discharges at central and peripheral levels of the autonomic nervous system is known (Lathers and Smith 1976). What is not known is the exact mechanism of rhythmicity and synchronization in peripheral nerve activity, although central mechanisms are thought to predominate (Polosa et al. 1969, 1972). Activity of populations of preganglionic or postganglionic sympathetic nerves is synchronized into bursts or slow waves that are usually temporally related to the phases of the cardiac and respiratory cycles (Gebber 1980; Gebber and Barman 1980). Oscillations in the discharge of whole sympathetic nerve bundles depict the synchronized ἀring of a large population of individual preganglionic and postganglionic

450 Sudden Death in Epilepsy: Forensic and Clinical Issues

neuronal units. Factors that determine the number of spontaneously active units at any given moment and the discharge rate of an individual preganglionic sympathetic neuron include the level of the mean arterial blood pressure (via baroreceptor input), level of CO2 (via chemoreceptor input), excitability of the neurons, and degree of synchrony within the driving input to these neurons from the brain stem and from somatic and visceral inputs (Barman et al. 1984). In essence, there is no predictable sequential correlation or combination of these events to explain why and when fatal arrhythmias occur. Dodd-O and Lathers (1990) deἀned a “stable” lockstep phenomenon as one containing time periods of 10 or more seconds during which the time interval between sympathetic spikes was unchanged. If the interspike interval was constant but then changed for no more than one interspike interval, stability was considered to be maintained. Two categories of stable lockstep phenomenon were found. The ἀrst was one in which the repeated 2.8-s interval was present and the second was one in which the 2.8-s interval was absent. If the interspike interval was altered on two consecutive occasions, the stability of the lockstep phenomenon was considered to be lost. The term “unstable lockstep phenomenon” was used to describe all time intervals of 10 or more seconds during which the lockstep phenomenon existed but the criteria for stable lockstep phenomenon were absent. We found that when stable lockstep phenomenon was lost, precipitous mean arterial blood pressure changes occurred more frequently. This ἀnding supports the fact that changes in the mean arterial blood pressure alter the number of units and the discharge rate of individual preganglionic sympathetic nerves and could be a factor in destabilization of previously stable lockstep phenomenon. When stable lockstep phenomenon was lost, electrocorticogram changes occurred more frequently. In addition, speciἀc burst patterns resulting from stimulation of the preganglionic nerve innervating the cat stellate ganglia can increase the number of neurons discharging in the inferior cardiac nerve (Birks et al. 1981), supporting the conclusion that such burst patterns of neural input are a major presynaptic mechanism in the modulation of synaptic transmission in sympathetic ganglia. Thus the central neuronal outflow is an important factor in altering the peripheral efferent discharge to the heart, which can initiate abnormalities in the electrocorticogram. The fact that central neuronal outflow alters the peripheral efferent discharge to the heart, resulting in arrhythmias, may explain why the present study did not ἀnd a correlation in the occurrence of the lockstep phenomenon and the occurrence of abnormalities in the electrocorticogram. It may be that the development of the lockstep phenomenon preceded the occurrence of visible changes in the electrocorticogram or vice versa. Deἀnitive experiments must be done to verify this possibility. Clinical signiἀcance of the lack of correlation of the occurrence of the lockstep phenomenon and abnormalities in the electrocorticogram may be inferred from a discussion by Blumhardt and Oozeer (1982). They reported that in some patients with temporal lobe epilepsy, cardiac acceleration preceded the onset of recognizable rhythmic surface EEG seizure activity and suggested that this sequence may reflect the onset of electrical discharge in deep limbic circuits and in the connections of these structures with the autonomic nervous system. Their study did not ἀnd a direct correlation between the occurrence of electrocorticogram abnormalities and development of rhythmical neuronal activity, at least as observable using their recording techniques. Indeed, in some patients, arrhythmias were observed at times when there were no seizure discharges on the EEG. Blumhardt and Oozeer (1982) also noted that the autonomic effects of temporal lobe epilepsy on heart rate and rhythm may be more severe in untreated, younger patients. This ἀnding agrees

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

451

with the observation that young epileptic patients are at high risk of sudden unexplained death. Blumhardt and Oozeer (1982) noted that simultaneous EEG and electrocorticogram recordings from patients with established epilepsy demonstrate the occurrence of nonepileptic and epileptic events on the same tape. They emphasized that negative ἀndings obtained in some patients must be cautiously interpreted and the outcome established by long-term follow-up studies. We believe that this conclusion and the results of our experimental work herein described suggest the need for more sophisticated electrophysiological monitoring of brain regions involved in autonomic control. Various central neural pathways, including the hypothalamus and amygdala, are involved in the initiation of the lockstep phenomenon (see Table 1.1 in Lathers et al. 2010a). The data demonstrate that central autonomic centers within the hypothalamus can modify peripheral cardiac nerve impulses to the heart and produce arrhythmias. Evans and Gillis (1975) demonstrated that the cardiac neural dysfunction and arrhythmias occurring in association with epileptogenic activity is due, at least in part, to modiἀcation of centrally mediated neural control of cardiac rhythm. They stimulated the posterior region of the cat hypothalamus at 5-min intervals. Cardioacceleration resulted from increased postganglionic sympathetic neural discharge and caused an increase in blood pressure. In the control period, stimulation of the hypothalamus increased sympathetic discharge but did not elicit arrhythmia; i.e., this was a subarrhythmogenic stimulus. When stimulation was repeated in the presence of a low dose of ouabain, a neural depressant effect occurred but arrhythmia did not develop. When the hypothalamus was again stimulated in the presence of a higher dose of ouabain, there was an increase in sympathetic neural discharge and subsequent changes in the ECG. The role of the hypothalamus in SUDEP has yet to be studied in detail. 28.5.2â•…Simultaneous Electroencephalogram and ECG Recordings in Humans The clinical relevance of our data is emphasized by the ἀndings of McLeod and Jewitt (1978). Twenty-four-hour continuous ECG monitoring was used to screen 300 patients not selected for suspected epilepsy. Of these, 36% exhibited major arrhythmias, many of which were ominous. These arrhythmias were found in asymptomatic patients, an important observation since the ἀnding strengthens the hypothesis that neurally induced autonomic arrhythmias may be a contributory mechanism in SUDEP. Schott et al. (1977) studied patients with suspected idiopathic epilepsy, four of whom had abnormalities on routine 12-lead ECGs. None had any cardiac symptoms and only one had a transient focal abnormality in the EEG. During prolonged continuous electrocardiographic monitoring, these patients manifested cardiac arrhythmias. Seizures were eliminated with treatment of antiarrhythmic agents or installation of a pacemaker. It was concluded that treatable cardiac arrhythmias may underlie a putative diagnosis of epilepsy more often than is generally recognized. Physicians must consider the possibility that cardiac arrhythmias occasionally mimic epilepsy. See the discussion of the Burgada syndrome by Herreros (2010) and Lathers et al. (2010b) in this book. Based on their experimental animal ἀndings, Lathers et al. (1983) suggested that there may be interindividual variability in susceptibility to arrhythmias, such that a smaller, as yet unidentiἀed group of persons at risk is more susceptible to autonomic dysfunction induced by epileptogenic activity. We suggested that the autonomic dysfunction observed

452 Sudden Death in Epilepsy: Forensic and Clinical Issues

in association with even minimal epileptogenic activity produced cardiac arrhythmias that could contribute to SUDEP. That interindividual variability may be a factor in explaining susceptibility to arrhythmias was later documented in two studies conducted in humans. Blumhardt and Oozeer (1982) noted simultaneous EEG and ECG recordings from patients with established epilepsy who demonstrated the occurrence of nonepileptic and epileptic events on the same tape. They emphasized that negative ἀndings obtained in some patients must be cautiously interpreted, and long-term follow-up studies need to be conducted to conἀrm the outcome. Only a few patients had a clear ECG or EEG abnormality accounting for the symptoms of repeated episodes of disturbed consciousness for which no diagnosis had been established by routine methods. It was concluded that simultaneous recordings of ECG and EEG are of initial value in the clinical workup of patients with unexplained attacks. They emphasized that unequivocally positive diagnostic records will be obtained only in a minority of patients because attacks generally occur with such widely varying frequencies that a high detection rate cannot be expected, even if recordings are extended for periods of up to a week in duration. Nonetheless, their study found some patients with ECG and EEG abnormalities. Keilson et al. (1987) conducted ambulatory monitoring of ECG and EEG in 338 consecutive patients with epilepsy for 20–24 h, including an overnight sleep period, and detected high-risk (based on the historical potential for sudden death) cardiac arrhythmias in 18 (5.3%) patients. It was concluded that the incidence of serious cardiac arrhythmias predisposing to sudden death is not increased in patients with epilepsy but is similar to the incidence reported for the nonepileptic population. The ἀnding of 5.3% of epileptic persons with high-risk arrhythmias in this study is important since it showed that some epileptic persons are more susceptible to autonomic dysfunction and development of serious arrhythmias. Occurrence of these arrhythmias may place these particular persons with epilepsy at risk for sudden death. The identical incidence of cardiac arrhythmias in both groups suggests that adding epilepsy to benign arrhythmias may increase the risk of transformation to a mortal arrhythmia at some time. 28.5.3â•…Role of Stress in SUDEP One contributory risk factor for SUDEP is stress. Stress, in and of itself, may contribute to the development of arrhythmias and epileptogenic activity (Lathers and Schraeder 2010b, Chapter 17; Wannamaker and Booker 1998; Lathers and Schraeder 2006a). Induced psychological stress in animals may lead to myocardial degeneration and sudden death. Monkeys subjected to electric shock developed bradyarrhythmia and death with asystole (Corley et al. 1975). Pigs, on the other hand, developed ventricular tachyarrhythmias rather than asystole (Johansson et al. 1974). Stress induction in rats was found to be associated with microthrombic and myocardial necrosis (Haft 1979). Experimental coronary arterial occlusion in dogs (Corbalan et al. 1974) and pigs (Skinner et al. 1975) demonstrated that tachyarrhythmias occurred more readily when the animals were also subjected to the stress of an unfamiliar adverse environment when compared to a familiar or comfortable environment. The importance of environmental stress in the person with epilepsy is obvious since stress is an ongoing factor in their lives. Stress may contribute both to the induction of seizures (Feldman and Paul 1976; Friis and Lund 1974; Wannamaker and Booker 1998; Kanner et al. 2004; Lathers and Schraeder 2006a, 2010b; Lathers 2009) and to the induction of autonomic dysfunction centrally. These factors, in turn, may modify cardiac peripheral neural discharge and produce cardiac arrhythmias and/or fatalities.

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

453

When considering central mechanisms, such as stress, which may be involved in the production of arrhythmias, one must consider that a patient taking drugs other than antiepileptic agents may also be predisposed to the development of arrhythmias. For example, the use of phenothiazines is associated with the phenomenon of sudden unexplained death in psychiatric patients (Leestma and Koenig 1968). Furthermore, individuals with epilepsy have been treated with phenothiazines as part of the management of psychiatric symptoms. Several studies (Lathers and Lipka 1986; Lathers et al. 1986) examined the effect of chlorpromazine and thioridazine on the heart rate, blood pressure, and ECG in the cat. As the drugs were infused, the heart rate and mean arterial blood pressure both decreased. Associated changes in the ECG included a loss of the P wave, occurrence of premature ventricular contractions, and ventricular ἀbrillation. Thus, these agents can modify the autonomic parameters of blood pressure, heart rate, and ECG. The question of whether the potential for phenothiazines to induce cardiovascular autonomic dysfunction means that the use of phenothiazines in individuals is a factor in sudden unexplained death remains to be answered, although this category of drugs now is used infrequently. The reader is referred to a review of this topic (Lipka and Lathers 1987). In addition, please see discussion by Faingold et al. (2010, Chapter 41) of the beneἀcial effects of selective serotonin reuptake inhibitor (SSRI). Serotonin (5-HT) is a critically important neurotransmitter in brainstem respiratory centers, known to exert an important role in controlling normal respiration and in enhancing respiratory rate in response to elevated carbon dioxide levels. Studies conducted in Faingold’s laboratory (Tupal and Faingold 2006; Faingold et al. 2010) indicate that drugs that enhance the activation of 5-HT receptors directly or indirectly, including a selective serotonin reuptake inhibitor and selective 5-HT receptor agonists, reduce or block respiratory arrest in DBA/2 mice at doses that do not block seizures. By contrast, a 5-HT receptor antagonist will induce€ SUDEP in the small percentage of DBA/2 mice that exhibit tonic seizures without respiratory arrest. The importance of enhancing 5-HT neurotransmission in prevention of respiratory arrest is emphasized in this SUDEP model. When DBA/1 mice were given fluoxetine (15–70 mg/kg, i.p.), a selective serotonin reuptake inhibitor, signiἀcant reductions in respiratory arrest were seen at 45–70 mg/kg, requiring the highest dose for complete suppression. These data indicate that fluoxetine is also effective in blocking respiratory arrest in DBA/1 mice, but higher doses were required than those in the DBA/2 mice. 28.5.4â•…Central Biochemical Changes Associated with Epileptogenic Activity Central biochemical mechanisms may contribute to or explain, in part, development of cardiac arrhythmias in association with epileptogenic activity that may ultimately induce SUDEP. In the study of Lathers et al. (1988b), (d-Ala2)methionine enkephalinamide (500€μg/ kg) was given intracerebroventricularly to nine cats anesthetized with alpha-Â�chloralose. Epileptogenic activity and hypotension occurred in all animals (maximum decrease ranging from 6 to 46 mm Hg; duration of 6–35 min). The heart rate decreased in six cats, increased in two, and showed little or no change in one. The duration of heart rate changes varied from 18 to 76 min. Naloxone (100 μg/kg) was given intravenously to six cats after (d-Ala2)methionine enkephalinamide. Naloxone suppressed or abolished the epileptogenic activity in all six cats, reversed (d-Ala2)methionine enkephalinamide–induced hypotension, and increased the heart rate in three cats, decreased the heart rate in two, and produced no change in one. (d-Ala2)methionine enkephalinamide may produce epileptogenic

454 Sudden Death in Epilepsy: Forensic and Clinical Issues

activity and cardiovascular changes through an action on central opiate receptors. In addition, (d-Ala2)metenkephalin has been shown to produce a centrally mediated vasopressor response as well as attenuation of the baroreceptor reflex in conscious cats (Yukimura et al. 1981), possibly leading to autonomic imbalance. This latter effect may precipitate arrhythmias. Possible mechanisms involved in the development of cardiac arrhythmias and/or sudden unexplained death in some epileptic patients are summarized in Figure 28.5 (Lathers et al. 1988b). Additional details are also found in Kraras et al. (1987). Enkephalins injected into the central nervous system elicit seizure activity (Frenk et al. 1978; Snead and Bearden 1980). Pentylenetetrazol increased enkephalin levels in the amygdala and the hippocampal, septal, and hypothalamic areas (Vindrola et al. 1983). It may be that pentylenetetrazolinduced increases in central enkephalin levels within the autonomic center of the hypothalamus inhibited the release of gamma aminobutyric acid and thus led to the production of epileptogenic activity and autonomic dysfunction in the experiments of Lathers and Schraeder (1982). The central roles of gamma aminobutyric acid and neuropeptides are discussed in detail in Chapters 17 and 18 of our ἀrst book (Lathers and Schraeder 1990b). At present, the role of enkephalins in SUDEP should be further studied. Resolution of the question of whether enkephalins elicit epileptogenic activity and autonomic dysfunction via inhibition of gamma aminobutyric acid release is important since an understanding of this mechanism should eventually allow the design of pharmacologic agents to prevent the epileptogenic activity and autonomic dysfunction and possibly diminish PTZ centrally Rat

Enkephalins centrally Rat Cat Concentration of Central Enkephalins

Rat K+ evoked release of GABA and/or Ca2+ dependent mechanism GABA release and/or Frog K+ conductance GABA terminal indirect Ca2+ entry in the same nerve terminal Anesthetized Animals Conscious Animals Sympathetic and parasympathetic central neural outflow and enhancement of reflex-induced vagal bradycardia and blood pressure Cat Dog

Inhibition of Central GABA Release Monkeys Epileptogenic Activity

Blood pressure and heart rate due to attenuation of the vagal component of the baroreceptor reflex Cat

Imbalance in Peripheral Sympathetic and Parasympathetic Neural Discharge Cat Cat-PTZ Arrhythmias Cat Humans Hypothesized Sudden Death in the Epileptic Person

Figure 28.5╇ Hypothesized biochemical and autonomic mechanisms involved in development of cardiac arrhythmias and/or sudden unexplained death in persons with epilepsy. (From Lathers, C. M. et al., Life Sci, 43, 2287–2298, 1988. With permission.)

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

455

the risk of associated SUDEP. This possibility is emphasized in a recent work published a number of years after Suter and Lathers (1984), and Kraras et al. (1987) ἀrst raised questions about a possible role for prostaglandin E2 and enkephalins in autonomic dysfunction characterized by nonuniform discharge associated with seizure activity and cardiac arrhythmias. Dhir and Kulkarni (2006) noted that numerous studies have implicated prostaglandins as potential modulators in seizure activity. They reported that rofecoxib, a selective cyclooxygenase-2 inhibitor, potentiates the anticonvulsant activity of tiagabine against pentylenetetrazol-induced convulsions in mice. They proposed that rofecoxib or similar drugs may have a place as adjuvant therapy in reducing the risk of adverse cardiac events when used in combination with standard antiepileptic drugs in the treatment of epilepsy. Incremental doses of pentylenetetrazol in the animal model for SUDEP (Lathers and Schraeder 1980, 1982; Schraeder and Lathers 1980, 1983; Carnel et al. 1980) triggered decreased sympathetic cardiovascular modulation, and baroreflex sensitivity occur after temporal lobe epilepsy surgery (Hilz et al. 2002). This suggests that temporal lobe epilepsy surgery reduces the risk of sympathetically mediated tachyarrhythmias and excessive bradycardiac counter-regulation, both of which may be involved in the pathophysiology of SUDEP. Thus, successful temporal lobe surgery for epilepsy may contribute to a reduction in the risk of SUDEP (Devinsky 2004; Devinsky et al. 2005; Khoury et al. 2005). The amygdala is a nucleus of the temporal lobe that has complex interconnections with multiple cortical and brainstem regions. Since the amygdala is involved in the production of autonomic temporal lobe epilepsy phenomenology, its role in seizure-induced cardiac arrhythmias needs to be considered. The amygdala is involved in seizure-related autonomic disturbances, as demonstrated in the kindling seizure model. Kindled seizures in rats are associated with an abrupt 50% increase in mean arterial blood pressure. Superimposed on this change in blood pressure was a profound bradycardia characterized by a rate about half that recorded before stimulation. Changes in heart rate and blood pressure observed during amygdaloid-kindled seizures were similar to those observed during secondary spontaneous seizures. These effects apparently are independent of the kindling stimulus because stimulus-induced cardiovascular changes were not present at the beginning of the kindling process. Amygdaloid-kindled seizures activate both branches of the autonomic nervous system. The bradycardia was mediated by the parasympathetic system; the pressor response was caused by an increase in peripheral resistance due to alpha-adrenergic receptor activation. More important, these ἀndings show that kindling is a useful seizure model for future studies on the effect of seizures on cardiovascular function and possible mechanisms of seizure-related sudden unexplained death (Goodman et al. 1990, 1999). A dramatic decrease in the incidence of stage 5 seizures in fully kindled animals after preemptive low-frequency sine wave stimulation suggests that this may be an effective therapy for prevention of seizures in patients with epilepsy (Goodman et al. 2005). Another set of experiments (Gary-Bobo and Bonvallet 1977) demonstrated that selected populations of neurons within the amygdala would, when stimulated, result in the combination of bradycardia and apnea, or in tachycardia and hyperpnea. These observations raise the possibility of the involvement of the amygdala being a factor in ictally related apnea, bradycardia, and tachyarrhythmias. In the latter instance, it would be reasonable to assume that there could be an adverse effect in persons who have epilepsy in combination with some genetically determined subclinical cardiac potential for fatal arrhythmia. The study of Druschky et al. (2001) reported sympathetic dysfunction in the form of altered postganglionic cardiac sympathetic innervation in patients with chronic temporal

456 Sudden Death in Epilepsy: Forensic and Clinical Issues

lobe epilepsy and suggested that the altered postganglionic cardiac sympathetic innervation may increase the risk of cardiac abnormalities and/or SUDEP. Developmental and regulatory mechanisms determining density and pattern of cardiac sympathetic innervation are still unclear, as is the exact role of innervation in arrhythmogenesis. This clinical study of Druschky et al. (2001), conducted in humans, conἀrms the numerous animal studies in which postganglionic cardiac sympathetic neural discharge was monitored before and during the development of arrhythmias (Lathers and Levin 2010, Chapter 33). Arrhythmias were initiated in association with increasing epileptogenic activity (Lathers and Schraeder 1982), an abrupt occlusion of the left anterior descending coronary artery (Lathers et al. 1977a, 1977b; Lathers 1980a, 1981a) or by administering toxic doses of digitalis glycosides (Lathers et al. 1977b, 1978). The nonuniform sympathetic neural discharge recordings were hypothesized by Lathers and colleagues (Lathers et al. 1974a, 1974b, 1977a, 1977b, 1978, 1988a; Lathers 1980b; Lathers and Roberts 1985) to be contributing to the development of arrhythmia and/or sudden death via nonuniform recovery of excitability in ventricular muscle (Han and Moe 1964). Sympathetic innervation and its role in normal and abnormal cardiac function were discussed (Lathers 1974a, 1974b, 1977a, 1977b, 1978; Lathers et al. 1988a; Lathers 1980b; Lathers and Roberts 1985). Recent studies report that cardiac-speciἀc overexpression of Sema3a in transgenic mice (SemaTG) associated with reduced sympathetic innervation and attenuation of epicardial-to-endocardial innervation gradient. SemaTG mice demonstrated sudden death and susceptibility to ventricular tachycardia due to catecholamine supersensitivity and prolongation of the action potential duration. The authors conclude that appropriate cardiac Sema3a expression is needed for sympathetic innervation patterning and is critical for heart rate control (Ieda et al. 2007). In a different model of status epilepticus developed in unanesthetized, chronically instrumented sheep, sudden death and pulmonary edema occurred. Hypoventilation was also demonstrated in the sudden death group. Difference in peak left atrial and pulmonary artery pressures and in extravascular lung water were found. The authors support the use of the model of epileptic sudden death to study the role of central hypoventilation in the occurrence of sudden unexpected death (Simon et al. 1982; Johnston et al. 1995, 1997). The DBA/2 mouse has also been proposed as a SUDEP model because they exhibit respiratory arrest after audiogenic seizures. Respiratory arrest is also implicated in human SUDEP. Since respiratory mechanisms are modulated, in part, by serotonin, the effect of serotoninergic agents on respiratory arrest has been evaluated (Tupal and Faingold 2006). Since the data showed that fluoxetine, a selective serotonin reuptake inhibitor, signiἀcantly reduced the incidence of respiratory arrest, this drug is under consideration for evaluation for SUDEP prevention in the patient population most susceptible to SUDEP.

28.6â•… Summary In summary, multiple animal models of SUDEP are needed to help in delineating mechanisms of SUDEP in humans since no one animal model addresses all of the potential contributory risk factors that may be involved in SUDEP. It is likely that no one mechanism of SUDEP will explain the contributory risk factors involved in all patients with epilepsy who die suddenly and unexpectedly. In this chapter, we discuss the similarities in autonomic dysfunction associated with arrhythmias and death in animal models for

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

457

digitalis toxicity, myocardial infarction, psychotropic toxicity, and epileptogenic activity. Pertinent points include the fact that central autonomic dysfunction may be associated with either interictal or ictal epileptogenic activity. Changes in peripheral cardiac sympathetic or parasympathetic neural discharge are associated with the epileptogenic activity and may produce arrhythmia (Carnel et al. 1985; Kraras et al. 1987; Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1983, 1989; Lathers et al. 1984; Suter and Lathers 1984; Lathers et al. 2008; Scorza et al. 2008; Lathers 2009). When pentylenetetrazol (intravenous) was given to anesthetized cats, autonomic dysfunction was associated with both interictal and ictal epileptogenic activity. Autonomic dysfunction was manifested by the fact that autonomic cardiac nerves did not always respond in a predictable manner to changes in blood pressure; development of a marked increase in variability in mean autonomic cardiac nerve discharge; and appearance of a very large increase in the variability of the discharge rate of parasympathetic nerves ἀrst and in sympathetic discharge second. The altered autonomic cardiac nerve discharge was also associated with interictal epileptogenic activity and arrhythmias, which may contribute to SUDEP. It may well be that various changes—pathological, neurophysiological, biochemical, and pharmacological—at times interact in a manner that is unfortunate for the patients with epilepsy. As summarized by James (1983), given the fact that the nerves to the heart have such an important function in both normal and abnormal cardiac activity, it is surprising that they have received so little attention in postmortem studies. James suggested that the condition of the intracardiac nerves and the autonomic ganglia should be investigated routinely during autopsy. This has ἀnally been done by Druschky et al. (2001), with a deἀnitive role for changes in the postganglionic nerves in epilepsy being demonstrated postmortem. The SUDEP animal model described may be used in the future to screen for new anticonvulsant agents that act to prevent autonomic neural dysfunction and cardiac arrhythmias.

Acknowledgments The original research was funded by NIH grant BRSGRR-4518, a grant from the Epilepsy Foundation of America, and NIH grant HL13666.

References Alkadhi, K. A. and K. H. Alzoubi. 2010. Synaptic plasticity of autonomic ganglia: Role of chronic stress and implication in cardiovascular diseases and sudden death (Chapter 26). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Barman, S. M., G. L. Gebber, and F. R. Calaresu. 1984. Differential control of sympathetic nerve discharge by the brain stem. Am J Physiol 247 (3 Pt 2): R513–R519. Beauchemin, C. A., J. J. McSharry, G. L. Drusano, J. T. Nguyen, G. T. Went, R. M. Ribeiro, and A. S. Perelson. 2008. Modeling amantadine treatment of influenza A virus in vitro. J Theor Biol 254 (2): 439–451. Berne, R. M., and M. N. Levy. 1967. Interplay of central and peripheral factors in the regulation of the circulation (Chapter 12). In Cardiovascular Physiology, ed. R. Berne and M. Levy. St. Louis, MO: C. V. Mosby. Birks, R. I., W. Laskey, and C. Polosa. 1981. The effect of burst patterning of preganglionic input on the efficacy of transmission at the cat stellate ganglion. J Physiol 318: 531–539.

458 Sudden Death in Epilepsy: Forensic and Clinical Issues Blumhardt, L. D., and R. Oozeer. 1982. Simultaneous ambulatory monitoring of the EEG and ECG in patients with unexplained transient disturbance of consciousness. Paper read at the 4th International Gent 1981 Symposium on Ambulatory Monitoring and the Second Gent Workshop of Blood Pressure Variability, New York. Carnel, S. B. 1982. An animal model for autonomic dysfunction in, and the effect of phenobarbital pretreatment on, pentylenetetrazol-induced seizures in the cat. In Pharmacology, The Medical College of Pennsylvania, Philadelphia. Masters Thesis research conducted in Dr. Lathers’ laboratory. Carnel, S. B., Schraeder P. L., and C. M. Lathers. 1980. Autonomic dysfunction in epilepsy: II. Cardiovascular changes associated with pentylenetetrazol-induced seizures. Clin Res 28: 609A. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30 (4): 225–240. Corbalan, R., R. Verrier, and B. Lown. 1974. Psychological stress and ventricular arrhythmias during myocardial infarction in the conscious dog. Am J Cardiol 34 (6): 692–696. Corley, K. C., H. P. Mauck, and F. Shiel. 1975. Cardiac responses associated with “yoked-chair” shock avoidance in squirrel monkeys. Psychophysiology 12 (4): 439–444. Devinsky, O. 2004. Effects of seizures on autonomic and cardiovascular function. Epilepsy Curr 4 (2): 43–46. Devinsky, O., W. B. Barr, B. G. Vickrey, A. T. Berg, C. W. Bazil, S. V. Pacia, J. T. Langἀtt, et al. 2005. Changes in depression and anxiety after resective surgery for epilepsy. Neurology 65 (11): 1744–1749. Dhir, A., and S. K. Kulkarni. 2006. Rofecoxib, a selective cyclooxygenase-2 (COX-2) inhibitor potentiates the anticonvulsant activity of tiagabine against pentylenetetrazol-induced convulsions in mice. Inflammopharmacology 14 (5–6): 222–225. Dittmer, D. S., and R. M. Grebe, eds. 1959. Handbook of Circulation. Philadelphia, PA: W. B. Saunders. Dodd-O, J. H., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in phenobarbital-Â�pretreated cats (Chapter 13). In Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Domek, N. S., C. F. Barlow, and L. J. Roth. 1960. An ontogenetic study of phenobarbital-C-14 in cat brain. J Pharmacol Exp Ther 130: 285–293. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Evans, D. E., and R. A. Gillis. 1974. Effect of diphenylhydantoin and lidocaine on cardiac arrhythmias induced by hypothalamic stimulation. J Pharmacol Exp Ther 191 (3): 506–517. Evans, D. E., and R. A. Gillis. 1975. Effect of ouabain and its interaction with diphenylhydantoin on cardiac arrhythmias induced by hypothalamic stimulation. J Pharmacol Exp Ther 195 (3): 577–586. Evans, D. E., and R. A. Gillis. 1978. Reflux mechanisms involved in cardiac arrhythmias induced by hypothalamic stimulation. Am J Physiol 234 (2): H199–H209. Faingold, C. L., and C. A. Berry. 1973. Quantitative evaluation of the pentylenetetrazol-Â�anticonvulsant interaction on the EEG of the cat. Eur J Pharmacol 24 (3): 381–388. Faingold, C. L., S. Tupal, Y. Mhaskar, and V. Uteshev. 2010. DBA mice as models of sudden unexpected death in epilepsy (Chapter 41). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C.€M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Feldman, R. G., and N. L. Paul. 1976. Identity of emotional triggers in epilepsy. J Nerv Ment Dis 162 (5): 345–353. Franz, D. N. 1980. Central nervous system stimulants (Chapter 24). In The Pharmacological Basis of Therapeutics, ed. A. Gilman, L. Goodman, and D. Gilman. New York, NY: Macmillan.

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

459

Frenk, H., G. Urca, and J. C. Liebeskind. 1978. Epileptic properties of leucine- and methionineenkephalin: Comparison with morphine and reversibility by naloxone. Brain Res 147 (2): 327–337. Friis, M. L., and M. Lund. 1974. Stress convulsions. Arch Neurol 31 (3): 155–159. Gary-Bobo, E., and M. Bonvallet. 1977. Commissural projection to the amygdala through the ἀmbria fornix system in the cat. Exp Brain Res 27 (1): 61–70. Gebber, G. L. 1980. Central oscillators responsible for sympathetic nerve discharge. Am J Physiol 239 (2): H143–H155. Gebber, G. L., and S. M. Barman. 1980. Rhythmogenesis in the sympathetic nervous system. Fed Proc 39 (8): 2526–2530. Gillis, R. A. 1969. Cardiac sympathetic nerve activity: Changes induced by ouabain and propranolol. Science 166 (904): 508–510. Gillis, R. A. 1971. Role of the nervous system in the arrhythmias produced by coronary occlusion in the cat. Am Heart J 81 (5): 677–684. Gillis, R. A., A. Raines, Y. J. Sohn, B. Levitt, and F. G. Standaert. 1972. Neuroexcitatory effects of digitalis and their role in the development of cardiac arrhythmias. J Pharmacol Exp Ther 183 (1): 154–168. Gillis, R. A., H. Thibodeaux, and L. Barr. 1974. Antiarrhythmic properties of chlordiazepoxide. Circulation 49 (2): 272–282. Goodman, J. H., R. E. Berger, and T. K. Tcheng. 2005. Preemptive low-frequency stimulation decreases the incidence of amygdala-kindled seizures. Epilepsia 46 (1): 1–7. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1990. Kindled seizures elevate blood pressure and induce cardiac arrhythmias. Epilepsia 31 (5): 489–495. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1999. Kindled seizures activate both branches of the autonomic nervous system. Epilepsy Res 34 (2–3): 169–176. Guyton, A. C. 1969. The Textbook of Medical Physiology, 3rd ed. Philadelphia, PA: W. B. Saunders. Haft, J. I. 1979. Role of blood platelets in coronary artery disease. Am J Cardiol 43 (6): 1197–1206. Hahn, F. 1960. Analeptics. Pharmacol Rev 12: 447–530. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Herreros, B. 2010. Cardiac channelopathies and sudden death (Chapter 19). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Hilz, M. J., O. Devinsky, W. Doyle, A. Mauerer, and M. Dutsch. 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125 (Pt 5): 985–995. Hirsch, C. S., and D. L. Martin. 1971. Unexpected death in young epileptics. Neurology 21 (7): 682–690. Ieda, M., H. Kanazawa, K. Kimura, F. Hattori, Y. Ieda, M. Taniguchi, J. K. Lee, et al. 2007. Sema3a maintains normal heart rhythm through sympathetic innervation patterning. Nat Med 13 (5): 604–612. Iivanainen, M., and J. Lehtinen. 1979. Causes of death in institutionalized epileptics. Epilepsia 20 (5): 485–491. Ito, T., M. Hori, K. Yoshida, and M. Shimizu. 1977. Effect of anticonvulsants on thalamic afterdischarge in rats and cats. Jpn J Pharmacol 27 (6): 823–831. James, T. N. 1983. Chance and sudden death. J Am Coll Cardiol 1 (1): 164–183. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. A comprehensive review of the literature and proposed mechanisms. Acta Neurol Scand (Suppl. 82) 63: 1–66. Johansson, G., L. Jonsson, N. Lannek, L. Blomgren, P. Lindberg, and O. Poupa. 1974. Severe stress– cardiopathy in pigs. Am Heart J 87 (4): 451–457. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37 (4): 531–537.

460 Sudden Death in Epilepsy: Forensic and Clinical Issues Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42 (4): 588–594. Kanner, A. M., A. Soto, and H. Gross-Kanner. 2004. Prevalence and clinical characteristics of postictal psychiatric symptoms in partial epilepsy. Neurology 62 (5): 708–713. Keilson, M. J., W. A. Hauser, J. P. Magrill, and M. Goldman. 1987. ECG abnormalities in patients with epilepsy. Neurology 37 (10): 1624–1626. Khoury, J. S., R. S. Winokur, J. I. Tracy, and M. R. Sperling. 2005. Predicting seizure frequency after epilepsy surgery. Epilepsy Res 67 (3): 89–99. Kiok, M. C., C. F. Terrence, G. H. Fromm, and S. Lavine. 1986. Sinus arrest in epilepsy. Neurology 36 (1): 115–116. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. LaManna, J. C., G. Cordingley, and M. Rosenthal. 1977. Phenobarbital actions in vivo: Effects on extra cellular potassium activity and oxidative metabolism in cat cerebral cortex. J Pharmacol Exp Ther 200 (3): 560–569. LaManna, J. C., and M. Rosenthal. 1975. Effect of ouabain and phenobarbital on oxidative metabolic activity associated with spreading cortical depression in cats. Brain Res 88 (1): 145–149. Lathers, C. M. 1980a. Effect of timolol on autonomic neural discharge associated with ouabaininduced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106. Lathers, C. M. 1980b. The effect of metoprolol on coronary occlusion-induced arrhythmia and autonomic neural discharge. Fed Proc 39: 771. Lathers, C. M. 1981a. Induced disease. Myocardial infarction in dogs and cats. In Mammalian Models for Research on Aging, 224–228. Washington, DC: National Academy Press. Lathers, C. M. 1981b. Models for studying sequelae to induced myocardial infarction. In Mammalian Models for Research on Aging. Washington, DC: National Academy of Sciences Press. Lathers, C. M. 2002. Risk assessment in regulatory policy making for human and veterinary public health. J Clin Pharmacol 42 (8): 846–866. Lathers, C. M. 2009. Epilepsy and sudden death: Personal reflections and call for global action. Epilepsy Behav 15 (3): 269–277. Lathers, C. M. 2010. Sudden death: Animal models to study nervous system sites of action for disease and pharmacological intervention (Chapter 25). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., and C. M. Smith. 1976. Ethanol effects on muscle spindle afferent activity and spinal reflexes. J Pharmacol Exp Ther 197 (1): 126–134. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., G. J. Kelliher, and J. Roberts. 1974a. Correlation of ouabain-induced arrhythmia and nonuniformity in cardiac accelerator nerves. Clin Res 22: 682A. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1974b. Relationship between the effect of ourabain on arrhythmia and interspike intervals (ISI) of cardiac accelerator nerves. Pharmacologist 16:€201. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977a. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., and J. Roberts. 1980. Digitalis cardiotoxicity revisited. Life Sci 27 (19): 1713–1733. Lathers, C. M., and J. Roberts. 1985. Are the sympathetic neural effects of digoxin and quinidine involved in their action on cardiac rhythm? J Cardiovasc Pharmacol 7 (2): 350–360.

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

461

Lathers, C. M., K. M. Keller, J. Roberts, and A. B. Beasley. 1977b. Role of the adrenergic nervous sysÂ� tem in arrhythmia produced by acute coronary artery occlusion (Chapter 5).€In€Pathophysiology and Therapeutics of Myocardial Ischemia, ed. L. A., G. J. Kelliher, and M. Rovetto. New York, NY: Spectrum. Lathers, C. M., L. J. Lipka, and H. Klions. 1988a. Digitalis glycosides: A discussion of the similarities and differences in actions and existing controversies. Rev Clin Basic Pharm 7 (1–4): 1–108. Lathers, C. M., and L. J. Lipka. 1986. Chlorpromazine: Cardiac arrhythmogenicity in the cat. Life Sci 38 (6): 521–538. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988b. The effect of intracerebroventricular d-ALA2 methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43 (26): 2287–2298. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010a. Neurocardiologic mechanistic risk factors in sudden unexpected death in epilepsy (Chapter 1). In Sudden Death in Epilepsy. Forensic and Clinical Issues, ed. C. M. Lathers, P. L Schraeder, M.€W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010b. Sodium channel dysfunction: Common pathophysiological mechanism associated with sudden death ECG abnormalities in Brugada syndrome and some types of epilepsy. Case histories (Chapter 20). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936. Lathers, C. M., and P. L. Schraeder. 1980. Autonomic dysfunction in epilepsy: III. Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced seizures. Clin Res 28: 615A. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 1990a. Synchronized cardiac neural discharge and epileptogenic activity, the lock-step phenomenon: Lack of correlation with cardiac arrhythmias (Chapter 12). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder, eds. 1990b. Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 2006a. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., and P. L. Schraeder. 2010a. Arrhythmias associated with epileptogenic activity elicited by penicillin. Chapter 35 in Sudden Death in Epilepsy: Forensic and Clinical Issues. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 2010b. Stress and SUDEP (Chapter 17). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J.€Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., R. F. Flax, and L. J. Lipka. 1986. The effect of C1 spinal cord transection or bilateral adrenal vein ligation on thioridazine-induced arrhythmia and death in the cat. J Clin Pharmacol 26 (7): 515–523.

462 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and R. M. Levin. 2010. Animal model for sudden cardiac death. Sympathetic innervation and myocardial beta receptor densities (Chapter 33). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Lathers, C. M., F. L. Weiner, and P. L. Schraeder. 1983. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lock-step phenomenon. Circ Res 31: 630A. Leestma, J. E., and K. K. Koenig. 1968. Sudden death and phenothiazines. Arch Gen Psychiatr 18: 137–148. Lipka, L. J., and C. M. Lathers. 1987. Psychoactive agents, seizure production, and sudden death in epilepsy. J Clin Pharmacol 27 (3): 169–183. Louie, A., H. S. Heine, K. Kim, D. L. Brown, B. VanScoy, W. Liu, M. Kinzig-Schippers, F. Sorgel, and G. L. Drusano. 2008. Use of an in vitro pharmacodynamic model to derive a linezolid regimen that optimizes bacterial kill and prevents emergence of resistance in Bacillus anthracis. Antimicrob Agents Chemother 52 (7): 2486–2496. McLeod, A. A., and D. E. Jewitt. 1978. Role of 24-hour ambulatory electrocardiographic monitoring in a general hospital. Br Med J 1 (6121): 1197–1199. Nakamura, K., and M. Kurebe. 1962. Differential effects of antiepileptics on hippocampal and pallidal afterdischarges in cats. Jpn J Pharmacol 12: 180–190. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Onuma, T. 1957. Relationships of the predisposition to convulsions with the action potentials of the autonomic nerves and the brain: II. Changes in action potential of the autonomic nerves and the brain under conditions for increasing the predisposition to convulsion. Tohoku J Exp Med 65 (2–3): 121–129. Orihara, O. 1952. Comparative observation of the action potential of autonomic nerve with E.E.G. Tohoku J Exp Med 57 (1): 43–55. O’Rourke, D. K., and C. M. Lathers. 1990. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat (Chapter 15). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Pace, D. G., and R. A. Gillis. 1976. Neuroexcitatory effects of digoxin in the cat. J Pharmacol Exp Ther 199 (3):583–600. Pickworth, W. B., J. Gerard-Ciminara, and C. M. Lathers. 1990. Stress, arrhythmias, and seizures (Chapter 22). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Polosa, C., J. L. Teare, and I. Wyszogrodski. 1972. Slow rhythms of sympathetic discharge induced by convulsant drugs. Can J Physiol Pharmacol 50 (3): 188–194. Polosa, C., P. Rosenberg, A. Mannard, N. Wolkove, and I. Wyszogrodski. 1969. Oscillatory behavior of the sympathetic system induced by picrotoxin. Can J Physiol Pharmacol 47 (9): 815–826. Raines, A., B. Levitt, F. G. Standaert, and Y. J. Sohn. 1970. The influence of sympathetic nervous activity on the antiarrhythmic efficacy of diphenylhydantoin. Eur J Pharmacol 11 (3): 293–297. Roberts, J. 1970. The effect of diphenylhydantoin on the response to accelerator nerve stimulation. Proc Soc Exp Biol Med 134 (1): 274–280. Schallek, W., and F. Zabransky. 1966. Effects of psychotropic drugs on pressor responses to central and peripheral stimulation in cat. Arch Int Pharmacodyn Ther 161 (1): 126–131. Schlosser, W., S. Franco, and E. B. Sigg. 1975. Differential attenuation of somatovisceral and viscerosomatic reflexes by diazepam, phenobarbital and diphenylhydantoin. Neuropharmacology 14 (7): 525–531. Schott, G. D., A. A. McLeod, and D. E. Jewitt. 1977. Cardiac arrhythmias that masquerade as epilepsy. Br Med J 1 (6074): 1454–1457. Schraeder, P. L., and C. M. Lathers. 1980. Autonomic dysfuction in epilepsy: I. A proposed animal model for unexplained sudden death in epilepsy. Clin Res 28: 618A.

Animal Model for Sudden Unexpected Death in Persons with Epilepsy

463

Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3 (1): 55–62. Schraeder, P. L., and G. G. Celesia. 1977. The effects of epileptogenic activity on auditory evoked potentials in cats. Arch Neurol 34 (11): 677–682. Schraeder, P. L., R. Pontzer, and T. R. Engel. 1983. A case of being scared to death. Arch Intern Med 143 (9): 1793–1794. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13 (1): 263–4; author reply 265–269. Simon, R. P., L. L. Bayne, R. F. Tranbaugh, and F. R. Lewis. 1982. Elevated pulmonary lymph flow and protein content during status epilepticus in sheep. J Appl Physiol 52 (1): 91–95. Skinner, J. E., J. T. Lie, and M. L. Entman. 1975. Modiἀcation of ventricular ἀbrillation latency following coronary artery occlusion in the conscious pig. Circulation 51 (4): 656–667. Snead, 3rd, O. C., and L. J. Bearden. 1980. Anticonvulsants speciἀc for petit mal antagonize epileptogenic effect of leucine enkephalin. Science 210 (4473): 1031–1033. Spratling, W. P. 1902. The cause and manner of death in epilepsy. Med News 80: 1225–1227. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1990. Relationship of the lockstep phenomenon and precipitous changes in blood pressure (Chapter 14). In Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15 (1): 15–30. Terrence, Jr., C. F., H. M. Wisotzkey, and J. A. Perper. 1975. Unexpected, unexplained death in epileptic patients. Neurology 25 (6): 594–598. Toman, J. E. P., and J. P. Davis. 1949. The effects of drugs upon the electrical activity of the brain. Pharm Rev 1: 425–492. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. Verrier, R. L., and S. C. Schachter. 2010. Neurocardiac interactions in sudden unexpected death in epilepsy: Can ambulatory electrocardiogram-based assessment of autonomic function and T-wave alternans help to evaluate risk? (Chapter 43). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Vindrola, O., M. Asai, M. Zubieta, and G. Linares. 1983. Brain content of immunoreactive [Leu5] enkephalin and [Met5]enkephalin after pentylenetetrazol-induced convulsions. Eur J Pharmacol 90 (1): 85–89. Wannamaker, B. B., and H. E. Booker. 1998. Treatment of provoked seizures (Chapter 119). In Epilepsy. A Comprehensive Textbook, ed. J. J. Engel, T. A. Pedley, J. Aicardi, M. Dichter, U. Heinemann, S. Moshe, R. Porer, and D. Taylor. Philadelphia, PA: Lippincott-Raven. Yukimura, T., G. Stock, H. Stumpf, T. Unger, and D. Ganten. 1981. Effects of [d-Ala2]-methionineenkephalin on blood pressure, heart rate, and baroreceptor reflex sensitivity in conscious cats. Hypertension 3 (5): 528–533.

A Characterization of the Lockstep Phenomenon in Phenobarbital-Pretreated Cats

29

Jeffrey M. Dodd-O Claire M. Lathers

Contents 29.1 Introduction 29.2 Method 29.3 Results 29.4 Discussion 29.5 Summary Acknowledgments References

465 466 469 477 478 479 479

29.1â•…Introduction Sudden unexplained death (SUD) was deἀned by Jay and Leestma (1981) as “non-traumatic death occurring in an individual within minutes or hours of the onset of the ἀnal illness or ictus.” These patients are not previously known to be suffering from any illness that would normally be expected to cause sudden death, and no pathologic explanation for their death has been found. Up to a 13% incidence of SUD has been reported in persons with epilepsy, with the epileptic population most at risk for SUD being the young person with a mean age of 32 years (Jay and Leestma 1981; Krohn 1977). Many causes for SUD have been postulated, including autonomic dysfunction and its relation to epileptogenic discharge (Jay and Leestma 1981). Using diverse models, different investigators have shown evidence of intrinsic activity at various levels in the nervous system. Cortical rhythms, controlled by subcortical neurons (Kiloh et al. 1972b) thought by many to be located in the thalamus (Kiloh et al. 1972a), are the basis for the alpha (8–13 Hz), beta (20–22 Hz), delta (3–4 Hz), etc., rhythms of electroencephalography. Basar (1976) used stereotaxic procedures to demonstrate spontaneous activities from medial geniculate nuclei, inferior colliculus, mesencephalic reticular formation, and dorsal hippocampus. Numerous studies (Barman and Gebber 1980, 1981; Gebber and Barman 1981) suggest the existence of an inherent rhythm of sympathetic nerve discharge, possibly originating from the hypothalamus (Barman and Gebber 1982). The results of Gebber and Barman (1981, 1984) indicate that a temporal relationship exists between the intrinsic rhythms of the central and the autonomic nervous systems. Lathers et al. (1977, 1978) reported that changes in the rate of autonomic discharge from postganglionic cardiac sympathetic branches may contribute to cardiac dysrhythmias. Thus, it is quite plausible that the association between epileptogenic activity and autonomic 465

466 Sudden Death in Epilepsy: Forensic and Clinical Issues

dysfunction evidenced in both animal (Lathers and Schraeder 1982; Meldrum and Brierley 1973; Meldrum and Horton 1973; Wasterlain 1974) and human (Jay and Leestma 1981; Van Buren 1958) studies may be a manifestation of the disruption of a normal pattern of temporally related intrinsic cortical and autonomic discharges. Studies in this laboratory have demonstrated a temporal synchronization between electrocorticogram (ECoG) activity and intrathoracic cardiac postganglionic sympathetic discharge during both ictal and interictal epileptogenic states (Lathers et al. 1987). The purpose of this chapter is to describe this phenomenon in phenobarbital-pretreated cats undergoing epileptogenic activity induced by pentylenetetrazol (PTZ). This relationship between the central nervous system discharges and the autonomic nervous system may prove important in explaining the high incidence of SUD in epilepsy.

29.2â•…Method Nine cats were anesthetized with 80 mg/kg intravenous (i.v.) alpha-chloralose. Tracheostomy was performed, and the femoral artery and vein were cannulated. Ventilation was maintained using a small-animal respirator, with intravenous gallamine (4 mg/kg doses, intermittently) being used to maintain paralysis. Arterial blood gases were monitored, and ventilation was altered to maintain the pO2, pCO2, and pH values within an acceptable physiological range. A bilateral frontal craniectomy was performed and the dura was resected to record ECoG activity. A thoracotomy and right partial pneumonectomy were performed to expose the cardiac postganglionic and right cardiac vagal nerves near the heart. The former were identiἀed as sympathetic by their discharge response to blood pressure drop produced by intravenous injection of 5 µg/kg histamine (Lathers et al. 1978). These nerves were desheathed, and nerve activity was recorded in one or more small nerve branches. Mean arterial blood pressure, electrocardiogram (lead II of the ECG), heart rate, and rectal temperature were monitored continuously throughout the experiment. Rectal temperature was maintained between 37.5°C and 38.5°C. A 10-min control period was monitored before beginning infusion of phenobarbital (20 mg/kg i.v.) over 10 min. One hour after completing the phenobarbital infusion, six doses of PTZ (10, 20, 50, 100, 200, and 2000 mg/kg) were administered i.v. at 10-min intervals. The half-life of this drug in the cat is unknown (Knoll Laboratories, personal communications, 1979) but has been shown to be 1.4 h in the dog (Jun 1976). If one assumes the half-life to be similar in the cat, these doses of PTZ administered were probably cumulative. Epileptiform discharges were categorized in three degrees, according to duration of discharge. Polyspike discharges continuing for 10 s or longer were designated prolonged ictal. Repetitive polyspike bursts of less than 10-s duration interrupted by brief periods of baseline cerebral activity were classiἀed as brief ictal. These types of ictal activity are analogous to those seen in the EEG during clinical seizures. Interictal spikes were those bilateral discrete paroxysmal spikes and/or polyspike and wave discharges analogous to the nonictal epileptogenic activity seen routinely in the interictal EEG of patients with a seizure disorder. In distinguishing interictal discharges from brief ictal activity, spikes occurring more frequently than 3.3 per second were not counted as interictal discharges. This maximal rate was decided on after determining that each oscillation of the polygraph pen required at least 100 ms to occur. A 200-ms return to baseline activity was considered evidence

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats

467

distinguishing a series of consecutive interictal spikes from one continuous brief ictal discharge. Spikes occurring more frequently than 3.3 per second (or 300 ms between the beginning of any two consecutive spikes) were classiἀed as part of the same polyspike discharge activity. An interictal ECoG spike was considered to be time locked to a sympathetic discharge only if the latter began within 200 ms of the beginning of the interictal ECoG spike. Each brief ictal discharge was analyzed as a single unit along with its corresponding autonomic activity. The brief ictal and autonomic discharges were considered time locked only if the autonomic activity depicted a single polyspike discharge whose ἀrst spike began after the brief ictal discharge began, and whose last spike began before the end of the ἀnal spike composing the brief ictal discharge. When the ECoG displayed prolonged ictal activity, the autonomic and ECoG discharge during this time period was not considered to be time locked. When sympathetic and ECoG spikes were time locked for an uninterrupted time period of ≥10 s, the total duration of this event was measured. Time-locked discharges were considered to depict the lockstep phenomenon (LSP) when (1) at least two episodes of time-locked autonomic and epileptogenic activities occurred during the 10-s interval, and (2) sympathetic activity occurred more frequently than ECoG activity, or vice versa. In the latter criterion, no more than one discharge from the less frequent component was allowed to exist without being time locked. If the sympathetic and ECoG spikes were not consistently time locked over a period of at least 10 s (uninterrupted), these spikes were not considered to be exhibiting LSP. If the discharges were time locked for more than 10 s and then interrupted for 4.5 s or less, the time-locked discharges were classiἀed as uninterrupted LSP. Interruptions longer than 4.5 s indicated that LSP had ended. The amount of LSP was quantiἀed in terms of incidence and duration. To quantify incidence, the number of time-locked ECoG and autonomic spikes exhibiting LSP was determined for the 10-min interval following each dose of PTZ. Next, the total numbers of sympathetic and ECoG spikes were determined for each of the nine cats for each of the six doses of PTZ. Also listed were the number of LSP discharges per total number of sympathetic discharges (LSP/S) and ECoG discharges (LSP/E). The value LSP/S is a measure of the proportion of all sympathetic spikes that are locked to ECoG spikes under the conditions of LSP. Likewise, the value LSP/E is a measure of the proportion of all ECoG spikes that are locked to sympathetic spikes under the conditions of LSP. To quantify duration, the total time (seconds) each cat spent in or out of LSP was determined. This was further classiἀed to depict one of the following situations: 1. The duration of LSP during each minute following administration of PTZ. In this case, all doses of PTZ were grouped together and time spent in LSP was determined only as a function of the latency period following administration of PTZ. 2. The duration of LSP during each dose of PTZ administered. In this case, duration of LSP was determined for each dose of PTZ regardless of the delay between administration of the drug and the beginning of LSP. One-factor analysis of variance (ANOVA) with repeated measures on time compared the mean proportions of total time that cats displayed LSP following the administration of PTZ. Means were collapsed across doses. A post hoc Student Newman–Keuls test (α = 0.050) was performed when indicated.

468 Sudden Death in Epilepsy: Forensic and Clinical Issues

Analogous tests compared the observed frequency of LSP during each dose of PTZ. The mean proportions of total time that cats displayed LSP during each 10-min interval following the administration of each dose of PTZ were examined. To determine whether the observed incidence of LSP could be due to chance alone, two multiple regressions were performed. In these analyses, the dependent variables were LSP/S and LSP/E. The independent variables were the number of sympathetic and ECoG spikes. Each ECoG spike was associated with two other ECoG spikes, one preceding it by 2.8 s and one following it by 2.8 s, a repeated ECoG interval observed in all cats. The 2.8-s interval could contain other episodes of ECoG activity. Frequently, each episode of ECoG activity contained within this 2.8-s interval was itself associated with two other ECoG spikes, one preceding it by 2.8 s and one following it by 2.8 s. The number of spikes contained within this 2.8-s interval varied greatly. However, the 2.8-s repeated ECoG interval was not considered to be present whenever it contained more than one other ECoG discharge within these borders. Figure 29.1 shows a diagram of these criteria for classiἀcation of presence or absence of the repeated ECoG interval. The ordinate displays the amplitude of the spike, and the abscissa displays the time interval between spikes. Tracing I displays the basic 2.8-s interval. This interval is actually the duration of the latency period from the end of one spike to the beginning of a spike with which it is associated. All spikes labeled A are related by this 2.8-s latency period. In tracing II, each pair of A spikes related by a 2.8-s interval envelops another B spike. Each B spike is itself related to two other B spikes by a 2.8-s interval. In tracing III, each pair of A spikes related by a 2.8-s repeated ECoG interval envelops two additional (B and C) spikes. Tracing HI violates our criteria for categorization as the presence of repeated ECoG interval. Determination was made of the total time (seconds) the repeated ECoG interval was observed. If the repeated ECoG interval (deἀned above) was present before being

Tracing I A

A

A

A

A

A B

A B

A B

A B

A B

A

A B C

A B C

A B C

A B C

Tracing II

Tracing III B C 2.8 s

Figure 29.1╇ Repeated 2.8-s interval in the electrocorticogram. Tracing I, basic 2.8-s repeated interval. Tracing II, acceptable variation of 2.8-s repeated interval. Tracing III, unacceptable variation. A, B, and C, each separate subset of interrelated spikes.

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats

469

interrupted for 4.5 s or less, it was considered to be present without interruption, since interruptions longer than 4.5 s marked the end of the repeated ECoG interval. Sympathetic ECoG activity was classiἀed based on both LSP (presence versus absence) and the repeated ECoG interval (presence or absence). The four patterns are: (1) LSP present with repeated ECoG interval present, (2) LSP present with repeated ECoG interval absent, (3) LSP absent with repeated ECoG interval present, and (4) LSP absent with repeated ECoG interval absent. The cats were evaluated for all doses of LSP, excluding the 2000 mg/kg dose. If artifactrendered segments of either the sympathetic or the ECoG printout were unreadable, this segment was deleted. A paired t test was used to compare time spent in LSP with repeated ECoG interval present with time spent in LSP with repeated ECoG interval absent. A second paired t test was performed to compare time spent in LSP with repeated ECoG interval present with time spent with LSP absent with repeated ECoG interval present. An ANOVA with repeated measures on time and dose, and a post hoc Student Neuman– Keuls test (α = 0.050), were performed for each of the following six patterns of ECoGsympathetic activity: (1) LSP present with repeated ECoG interval present, (2) LSP present with repeated ECoG interval absent, (3) total LSP, (4) LSP absent with repeated ECoG interval present, (5) LSP absent with repeated ECoG interval absent, and (6) total LSP absent. Precipitous, rather than gradual, changes in mean arterial pressure were measured to highlight any possible changes in the character of LSP coincident with a change in the mean arterial blood pressure. The mean arterial blood pressure changes occurring in all nine cats were reviewed during the control, phenobarbital infusion, and PTZ treatment periods. Systolic changes of greater than 23 mm Hg over a 10-s interval were never seen during the control period; thus, changes greater than 23 mm Hg over a 10-s interval were deἀned as precipitous. A total of 89 such episodes were analyzed after administration of all doses of PTZ except 2000 mg/kg. This dosage led to death in all cats. The incidence of precipitous change in mean arterial blood pressure was analyzed using two different methods. Ninety-ἀve percent conἀdence intervals were constructed around the incidences of precipitous mean arterial blood pressure changes for each minute following the administration of all doses of PTZ. Ninety-ἀve percent conἀdence intervals were also constructed around the incidences of precipitous mean arterial blood pressure changes at each dose of PTZ. The ECG of each cat was analyzed during the control period to ἀnd any abnormality intrinsic to that particular cat. These were discarded, and any new abnormalities occurring after the administration of PTZ were evaluated. Each change was classiἀed according to the time interval and dose during which it occurred. The ECG parameters examined were (1) T wave changes, (2) P wave changes, (3) changes in the QRS complexes, (4) appearance of a Q wave, (5) premature ventricular contractions, and (6) ventricular tachycardia. Ninety-ἀve percent conἀdence intervals were constructed around the incidences of ECG changes for each minute following dosing with PTZ and for each dose of PTZ administered.

29.3â•…Results As depicted in Figure 29.2, no ECoG spikes were time locked to a sympathetic discharge in the control period (Figure 29.2, I) or in the period during the infusion of phenobarbital

470 Sudden Death in Epilepsy: Forensic and Clinical Issues I. Pre-Pb Symp ECoG II. Pb Symp ECoG

200 µV

III. PTZ

200 µV

Symp

ECoG 5s

Figure 29.2╇ Occurrence of sympathetic and ECoG discharges in a time-locked manner, i.e.,

lockstep phenomenon (LSP). Traces in I, II, and III are three different time periods from the same cat: I, during prephenobarbital control period; II, 9 min, 50 s into phenobarbital control period; III, 8 min, 50 s into 200 mg/kg PTZ dosage. Symp, cardiac sympathetic postganglionic neurons; ECoG, electrocorticogram. Horizontal calibration is 5 cm/s in all cases. Vertical calibration is 200 µV/cm in all cases. Mean arterial blood pressures (not shown)–ECG patterns–mean heart rates were (I) 120–normal–131, (II) 54–normal–112, and (III) 98–normal–152, respectively.

before the administration of PTZ (Figure 29.2, II). Interictal ECoG spikes were time locked to the sympathetic discharge in Figure 29.2 (III) and were designated LSP. The latency period varied but the ECoG discharge always preceded the corresponding sympathetic discharge. Sympathetic discharge was observed less frequently than ECoG discharge. However, in 93% of the time periods during which sympathetic activity was present, the proportion of these discharges that were time locked to ECoG spikes was 0.85 or greater. Further analysis using multiple regressions and bivariate correlations suggested that the incidence of observed LSP was limited more by the incidence of observed sympathetic discharge than by the incidence of observed ECoG discharge. Table 29.1 compares, for each cat, the percentage of time that LSP was present (with or without the repeated ECoG interval) with the percentage of time absent. In eight of nine cats, LSP was observed to be present during at least 55% of the experiment (LSP present being deἀned as LSP present with repeated ECoG interval present and with repeated ECoG interval absent). In all cats following all doses of PTZ, LSP was present 66% of the time on average. The proportion of time each cat demonstrated each pattern is shown in Table 29.1. As an example, cat 4 demonstrated the pattern LSP present with repeated ECoG interval

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats

471

Table 29.1â•… Contingency Table Displaying the Presence or Absence of the Lockstep Phenomenon versus the Presence or Absence of the Repeated Electrocortical Interval

Cat No.

LSP Present, Repeated ECoG Interval Present (%€Total for Cat)

LSP Present, Repeated ECoG Interval Absent (%€Total for Cat)

LSP Absent, Repeated ECoG Interval Present (%€Total for Cat)

LSP Absent, Repeated ECoG Interval Absent (%€Total for Cat)

1 2 3 4 5 6 7 8 9 Average

68 40 13 60 41 63 42 46 67 49

13 12 1 1 3 3 0 7 3 5

15 18 18 14 24 5 26 24 9 17

5 31 68 26 33 29 32 22 21 30

present for 60% of the time. This was determined by dividing the total number of seconds evaluated (3000 s) into the total time the pattern was observed (1801 s). Similarly, this cat demonstrated the pattern LSP present with repeated ECoG interval absent 1% of the time, the pattern LSP absent with repeated ECoG interval present 14% of the time, and the pattern LSP absent with repeated ECoG interval absent 26% of the time. Note that for the one cat in which LSP was not present during at least 55% of the experiment, the data contained long recording periods that were technically inadequate for quantitative description of the relationship between ECoG and sympathetic spikes. Observation of these periods in cat 8 revealed that they were usually composed of global increases in ECoG activity associated with global increases in sympathetic activity. Table 29.2 lists the incidences of sympathetic spikes (No. Symp) and ECoG spikes (No. ECoG) for the 10-min period following the administration of a dose of PTZ for each cat. Also listed, in decimal form, are measures of (1) the proportion of sympathetic spikes that occurred time locked to ECoG spikes (TL/S) and (2) the proportion of ECoG spikes that occurred time locked to sympathetic spikes (TL/E). In 40 of the 45 (89%) measured time intervals, ECoG discharges occurred more frequently than sympathetic discharges. Also, in 43 of 45 (93%) of the measured time periods, the ratio TL/S was greater than 0.85. This high incidence of ECoG activity occurring more frequently than sympathetic activity, combined with the high proportion (>0.85) of observed sympathetic spikes being time locked to ECoG spikes in 93% of the measured time periods, suggests that the incidence of observed sympathetic ἀrings was the limiting factor in the total number of times when time locked activity (and, secondarily, LSP) was observed. Further evidence that the incidence of LSP is limited by the frequency of sympathetic discharge was that (1) multiple regressions showed no signiἀcant predictor where TL/S is dependent; (2) multiple regressions showed TL/E was related to No. Symp and No. ECoG (multiple r = 0.86 signiἀcant to 0.00005); and (3) bivariate correlation of No. Symp (0.55, step 1, proportion of variance = 0.30) and No. ECoG (0.01, step 2, proportion of variance = 0.73) suggested that ECoG becomes a factor in increasing the predictability of ECoG time locked only after sympathetic activity has made its impact.

472 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 29.2â•… Proportion of Sympathetic and Electrocortical Spikes and Proportions of Each in the Lockstep Phenomenon PZT 10 mg/kg Cat. No. 1 2 3 4 5 6 7 8 9

PZT 20 mg/kg

No. Symp

No. ECoG

LSP S

LSP E

No. Symp

No. ECoG

LSP S

LSP E

No. Symp

No. ECoG

LSP S

LSP E

253 48 — — — — — 139 —

313 161 — — — — — 9 —

1.0000 0.9583 — — — — — 0.0504 —

0.8083 0.2857 — — — — — 0.7778 —

354 340 73 204 219 256 239 276 346

356 465 308 396 227 390 260 301 374

0.9972 0.9882 0.5753 0.5887 1.0000 0.9883 1.0000 0.9746 0.9942

0.9916 0.7225 0.1364 0.3682 0.9648 0.6477 0.9192 0.8937 0.9198

451 471 224 358 129 337 345 304 298

450 582 302 368 641 352 359 310 346

0.9956 0.9915 0.8036 0.9915 1.0000 0.9941 1.0000 0.9901 0.9933

0.9978 0.9193 0.5960 0.9511 0.2012 0.9517 0.9610 0.9710 0.8555

PZT 100 mg/kg Cat. No. 1 2 3 4 5 6 7 8 9

PZT 50 mg/kg

PZT 200 mg/kg

PZT 2000 mg/kg

No. Symp

No. ECoG

LSP S

LSP E

No. Symp

No. ECoG

LSP S

LSP E

No. Symp

No. ECoG

LSP S

LSP E

441 290 110 399 238 359 444 376 403

429 335 125 430 328 395 443 432 398

0.9728 1.0000 0.8727 1.0000 0.9958 0.9944 0.9752 0.9947 0.9876

1.0000 0.8657 0.7680 0.9279 0.7226 0.9038 0.9774 0.8657 1.0000

318 124 262 392 211 225 347 158 359

306 284 255 402 285 277 352 364 372

0.9245 0.9919 0.8588 0.9974 0.8720 0.9956 0.9337 1.0000 0.9972

0.9608 0.4331 0.8824 0.9726 0.6456 0.8087 0.9204 0.4341 0.9624

10 48 — — 22 7 — 18 46

10 48 — — 24 13 — 36 47

1.0000 1.0000 — — 0.9091 1.0000 — 1.0000 1.0000

1.0000 1.0000 — — 0.8333 0.5385 — 0.5000 0.9787

A certain stability to the neural activity and a high incidence of LSP were associated with the presence of the repeated 2.8-s interval in ECoG. Disruption of this rhythm was associated with a degeneration of LSP. When LSP was present, it was associated an average 74% of the time (p < 0.01) with a 2.8-s repeated ECoG interval (13,131 s of LSP present with repeated ECoG interval present divided by [13,131 + 4545] s of LSP present with and without a repeated ECoG interval present). Furthermore, when the repeated ECoG interval was present, LSP was absent 9% of the time (1268 s repeated ECoG interval present with LSP absent divided by [13,131 + 1268] s of repeated ECoG interval present with and without LSP present) (p < 0.01). When the repeated ECoG interval was absent, LSP was absent more frequently than present (average 17% LSP without ECoG versus average 30% LSP absent without ECoG). The relative prevalence (seconds) of each sympathetic-ECoG pattern following the administration of PTZ was examined. Figure 29.3 shows that as the time period after the administration of the most recent dose of PTZ increased: (1) the mean proportion of time that LSP was present with repeated ECoG interval present increased regularly (panel a); (2) the mean proportion of time that LSP was present with repeated ECoG interval absent decreased irregularly (panel b); and (3) the mean proportion of time that LSP was present with or without repeated ECoG interval present increased gradually (panel c). Similarly, panels d, e, and f of Figure 29.3 show that as the time period after the administration of the most recent dose of PTZ increased: (1) the mean proportion of time that LSP was absent with repeated ECoG interval present increased irregularly over the ἀrst 7 min, then decreased;

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats 0.6000

(a)

(d)

(b)

(e)

(c)

(f )

0.5000 0.4000

473

Subset 1 Subset 2 Subset 3

0.3000 0.2000

Incidence (mean proportion)

0.1000

0.6000 0.5000 0.4000 0.3000 0.2000 0.1000

0.7000 0.6000 0.5000 0.4000 0.3000 0.2000 0.1000

1 2 3 4 5 6 7 8 9 10

1 2 3 4 5 6 7 8 9 10

Time elapsed (min)

Figure 29.3╇ Histograms representing the results of the Student Newman–Keuls test for each of the six sympathetic-ECoG patterns as a function of time interval following the administration of PTZ. (a) LSP present with repeated 2.8-s ECoG interval present; (b) LSP present with repeated 2.8-s ECoG interval absent; (c) LSP present whether repeated 2.8-s ECoG interval is present or absent; (d) LSP absent with repeated 2.8-s interval present; (e) LSP absent with repeated 2.8-s ECoG interval absent; and (f) LSP absent whether repeated 2.8-s ECoG interval is present or absent. Ordinate displays the mean proportion (for nine cats) of each 1-min interval following administration of all doses of PTZ during which the sympathetic-ECoG pattern in question was observed. Abscissa displays the time-elapsed interval (in minutes) following administration of PTZ in all cats. Patterns within bars distinguish the subsets (minutes) having statistically comparable mean proportions of time during which sympathetic-ECoG pattern was observed. Subset 1 consists of 1-min intervals having statistically similar mean proportions of time during which sympathetic-ECoG pattern was observed. Subset 2 consists of 1-min intervals having statistically similar mean proportions of time during which sympathetic-ECoG pattern was observed. Subset 3 consists of 1-min intervals having mean proportions of time during which sympathetic-ECoG pattern was observed, which is not distinct from either of other two groups.

474 Sudden Death in Epilepsy: Forensic and Clinical Issues

these changes were not statistically signiἀcant; (2) the mean proportion of the time that LSP was absent with repeated ECoG interval absent decreased steadily over the ἀrst 5 min, then increased somewhat; minute 1 was signiἀcantly greater than all others; and (3) the mean proportion of time that LSP was absent with or without the repeated ECoG interval present decreased progressively, minutes 1 and 2 being signiἀcantly greater than the others. Overall, the presence of LSP was directly related to the duration of the time interval after the administration of the most recent dose of PTZ. This relationship is fairly consistent from minute to minute. Similarly, the presence of the repeated ECoG interval was directly related to the duration of the time interval after the administration of the most recent dose of PTZ. This relationship was somewhat inconsistent from minute to minute. Using repeated-measure ANOVA and Student Newman–Keuls, the relative prevalence of each sympathetic-ECoG pattern in all cats during the entire 10-min period following the administration of each dose of PTZ was examined. In one sympathetic-ECoG pattern, LSP absent with repeated ECoG interval present, the mean proportions of time that the pattern existed following each of the ἀve doses of PTZ were statistically comparable. Figure 29.4a shows the pattern LSP present with repeated ECoG interval present. Two distinct groups were separated, with prevalence following the ἀrst dose eliciting none or only minimal interictal epileptogenic activity, signiἀcantly less than following any of the subsequent three doses. Prevalence during the ἀnal dose, which induced almost no ictal activity and very little interictal activity, was not signiἀcantly different from either of the other groups (p < 0.05). In a similar manner, the other four ECoGsympathetic patterns showing varying prevalence are seen in Figure 29.4 as a function of dose of PTZ administered, i.e., as a function of increasing amounts of epileptogenic activity consisting of more ictal activity and less interictal discharges. In Figure 29.5a the incidence of precipitous changes in mean arterial blood pressure for epileptiform discharges induced by all doses of PTZ in all cats was recorded as a function of the time interval after the last administration of PTZ. It showed that a statistically greater (p < 0.05) incidence of these changes occurred during minute 1. It held fairly constant during minutes 2 through 10, inclusive. The incidence of mean arterial blood pressure changes for increasing amounts of epileptogenic activity induced by the increasing doses of PTZ displayed a bell-shaped distribution (Figure 29.5b) with a peak following PTZ at 50 mg/kg dose. Ninety-ἀve percent conἀdence intervals showed this incidence to be higher than that following either the 10 or 200 mg/kg dose of PTZ. The incidence of changes in the ECG decreased during the ἀrst 8 min inclusive following the administration of PTZ (Figure 29.6a). This difference was statistically signiἀcant (p€< 0.06) during minute 1 when compared to minute 7 or 8. The occurrence of changes in the ECG was least frequent following 10 mg/kg PTZ than following any other dose. This dose of PTZ elicited little or no interictal activity in most cats and almost no ictal activity. This was a statistically signiἀcant difference from the prevalence following all other doses except 50 mg/kg PTZ. This dose of PTZ elicited primarily ictal activity. The mean proportion of each minute during which LSP was present, and the mean proportion of each minute during which the repeated ECoG interval was observed, were directly proportional to the time interval elapsed after the administration of all doses of PTZ. When parameters of autonomic dysfunction were examined as a function of the time interval since the administration of all doses of PTZ, the incidence of blood pressure

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats 0.4000

475

(a) Subset 1

0.2000

Subset 2

(b)

(d)

(c)

(e)

0.8000

Incidence (mean proportion)

0.6000 0.4000

0.2000

0.1000 0.8000 0.6000

0.4000

0.2000

10 20 50 100 200 10 20 50 100 200 Dose of PTZ administered (mg, kg)

Figure 29.4╇ Histograms representing results of Student Newman–Keuls test for each of five sympathetic-ECoG patterns as a function of dose of PTZ administered. (a) LSP present with repeated 2.8-s ECoG interval present; (b) LSP present with repeated 2.8-s ECoG interval absent; (c) LSP present whether repeated 2.8-s ECOG interval is present or absent; (d) LSP absent with repeated 2.8-s ECoG interval absent; and (e) LSP absent whether repeated 2.8-s ECoG interval is present or absent. Ordinate displays the mean proportion (for nine cats) of the 10-min period following administration of a given dose of PTZ during which the sympathetic-ECoG pattern in question was observed. The abscissa displays dose of PTZ administered. Patterns within bars distinguish subsets (minutes) having statistically comparable mean proportions of time during which sympathetic-ECoG pattern was observed. Subset 1 consists of 1-min intervals having statistically similar mean proportions of time during which sympathetic-ECoG pattern was observed. Subset 2 consists of 1-min intervals having statistically similar mean proportions of time during which sympathetic-ECoG pattern was observed.

476 Sudden Death in Epilepsy: Forensic and Clinical Issues (a)

(b)

100 40

30

60

Incidence

Incidence

80

40 20

20

10

0 1

2

3

4

5

6

7

8

9

20

10

10

50

100

200

Dose (mg/kg)

Interval (min)

Figure 29.5╇ Total incidence of mean arterial blood pressure changes for all nine cats both as a function of time interval since administration of most recent doses of PZT and as a function of dose PZT administered. (a) Total incidence of mean arterial blood pressure changes for all cats versus minute time interval since most recent administration of PTZ. Vertical bars indicate 95% confidence intervals. No confidence interval can be determined for the values in minute 9. (b) Total incidence of mean arterial blood pressure changes for all cats versus dose of PTZ administered. Vertical bars indicate 95% confidence intervals.

(a)

(b)

100

200

80

160 Incidence

240

Incidence

120

60

120

40

80

20

40

1

2 3

4

5

6

7

Interval (min)

8

9

10

0

10

20

50

100

200

Dose (mg/kg)

Figure 29.6╇ Total incidence of ECG changes for all nine cats both as a function of time inter-

val since administration of most recent doses of PTZ and as a function of dose of PZT administered. (a) Total incidence of ECG changes for all cats versus minute interval since the most recent administration of PZT. Vertical bars indicate 95% confidence intervals. (b) Total incidence of ECG changes for all cats versus dose of PTZ administered. Vertical bars indicate 95% confidence intervals.

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats

477

change was signiἀcantly greater (95% conἀdence interval) during the ἀrst minute following the epileptogenic activity induced by the administration of all doses of PTZ than during any other minute. Furthermore, the blood pressure became more stable with increasing time after each dose. Likewise, the incidence of ECG changes decreased over the ἀrst 8 min following the administration of PTZ and was statistically greater following minute 1 than following minute 7 or 8, when epileptiform activity decreased. When this stable LSP was lost, both precipitous mean arterial blood pressure changes and the incidence of ECG changes occurred more frequently. Just as the mean proportion of time during which LSP was observed was statistically greater after the PTZ 50 mg/kg when maximal numbers of epileptiform discharges were induced, the incidence of mean arterial pressure change was signiἀcantly greater after this dose than after the 10 and 200 mg/kg PTZ doses, which were less epileptogenic than the 50 mg/kg dose. No relationship between incidence of ECG changes and LSP, as a function of PTZ dosage, could be found; rather, there seemed to be a correlation with the doses inducing the greatest degree of epileptiform discharge.

29.4â•…Discussion We examined the temporal synchronization between ECoG and sympathetic spikes that was observed in a PTZ animal model of epilepsy, pretreated with the antiepileptic drug phenobarbital. These temporally synchronized spikes were designated LSP when they occurred consistently over a period of 10 or more uninterrupted seconds. The administration of PTZ to phenobarbital-pretreated cats converted the normal baseline state of desynchronized sympathetic discharges and phenobarbital spindles into a state of hypersynchronized cortical and sympathetic activities. This laboratory has frequently used PTZ treatment of cats to create an experimental model of epileptogenic activity (Lathers et al. 1978, 1984, 1987; Schraeder and Lathers 1983). The dosing regimen has been developed to produce all degrees of subconvulsant and convulsant activities (Schraeder and Lathers 1983). In this way the full spectrum of cortical epileptiform activity can be examined. Phenobarbital was used by this laboratory as an anticonvulsant in this animal model. The possibility that this observed temporal synchronization of ECoG and sympathetic spikes represents artifact, originating from either outside or within the cat, was examined ἀrst. Possible mechanical artifact from the respirator or the 60-Hz alternating current used to power the polygraph was dismissed for two reasons. First, other experiments undertaken with the same equipment in the same room in the same time period produced no such artifact. Second, these potential sources would produce artifact at a constant rate throughout the entire experiment. Sympathetic spikes, ECoG spikes, and LSP were not observed at a constant rate throughout the entire experiment. In addition, neither sympathetic nor ECoG spikes were observed during either control period. The possibility of intrinsic artifact from cardiac or skeletal muscle contractions was considered. However, neither the ECoG nor sympathetic spikes bore relationship to the incidence of changes in the ECG. Furthermore, the use of gallamine effectively blocked spontaneous muscle contractions. Analysis with multiple regression showed that the timelocked occurrence of ECoG spikes and sympathetic spikes was not random. This analysis

478 Sudden Death in Epilepsy: Forensic and Clinical Issues

suggested that the incidence of LSP was directly related to the incidence of sympathetic spikes. Epileptogenic activity induced by PTZ appears to be necessary to allow this phenomenon to express itself. Although, in eight of the nine cats, LSP was present more often than absent (>55% of time in seconds) while the cat was under the influence of PTZ, this phenomenon was not found in any cat during the control period. Epileptogenic activity induced by PTZ is an experimental model of primary generalized epilepsy. The method of action of this drug, however, is uncertain (Stone 1972). There are three proposed methods of action of PTZ, which would allow the LSP to express itself: (1) spatial and temporal summation of neuronal discharges in a subcortical center producing a stimulus strong enough to overcome the cortical and ganglionic threshold (Hahn 1960); (2) increased synaptic recruitment, resulting in the ampliἀcation of subcortical stimuli along their path so that, upon reaching the cortex and sympathetic ganglion, they are capable of causing these neurons to discharge; and (3) increased irritability of all neurons so that subcortical impulses could stimulate cortical and ganglionic neurons (Hahn 1960). In each case, PTZ effectively creates a hyperirritable state of epileptogenic electrical activity present in the central and autonomic nervous systems. Although phenobarbital can act to minimize this irritability, the effect of this pharmacological agent in this study eventually was overcome by increased epileptogenic activity. The 2.8-s repeated ECoG interval, a type of latency period between the end of one spike and the beginning of another associated spike, appeared to convey a stabilizing effect on the presence of LSP. When this repeated ECoG interval was present, LSP was present a signiἀcantly greater percentage of time than when LSP was absent. When the repeated ECoG interval was not present, LSP was much less common and the distinct ECoG spikes often degenerated into prolonged ictal activity. Our analysis minimized the degree of association between the repeated ECoG interval and the presence of LSP. Our deἀnition of a repeated ECoG interval excludes periods when two or more additional ECoG spikes are contained within the 2.8-s interval, even if these additional ECoG spikes are time locked to sympathetic spikes. A more liberal deἀnition would result in a more frequent association between LSP and repeated ECoG interval. The mean proportion of each minute during which LSP was present was directly proportional to the time interval elapsed following the administration of PTZ. The direct relationship reflects the fact that the episodes of epileptogenic activity, and particularly prolonged ictal activity, are most frequently observed shortly after PTZ is administered. Similarly, the incidence of precipitous blood pressure change was signiἀcantly greater (p€<€.05) during the ἀrst minute following the administration of all doses of PTZ than during any other minute. Likewise, the incidence of ECG changes decreased over the ἀrst 8 min following administration of all doses of PTZ, with minute 1 containing a signiἀcantly higher incidence compared to minute 7 or 8. Thus, the possibility arises of a relationship between the presence of LSP and both a stable mean arterial pressure and a normally functioning autonomic nervous system.

29.5â•…Summary The concept of rhythmic neuronal activity being associated with normal central and autonomic nervous system activities is not new. Normal and abnormal cortical patterns are

The Lockstep Phenomenon in Phenobarbital-Pretreated Cats

479

the basis of the clinical use of electroencephalography. Sympathetic nerve discharge patterns have been characterized by others (Barman and Gebber 1980; Gebber and Barman 1980). A temporal relationship between discharge patterns of the central and autonomic nervous system has also been shown previously (Gebber and Barman 1981). Loss of this rhythmic stability can alter neurotransmitter release (Birks 1978; Birks et al. 1981) and initiate autonomic dysfunction (Lathers and Schraeder 1982; Lathers et al. 1977). This chapter characterizes LSP and supports the idea of closely related central and autonomic rhythmic activity, which is important in maintaining homeostasis. The data showed that when a stable LSP was lost, both precipitous mean arterial blood pressure changes and the incidence of ECG changes occurred more frequently. This suggests that the LSP, either by its mere presence or by the rhythm at which it occurs, may play a role in the origin of autonomic dysfunction, contributing to the rise of SUD in epilepsy.

Acknowledgments The authors gratefully acknowledge Dr. Adele Kaplan for her statistical expertise and critical review of the manuscript. We are indebted to Dr. Paul L. Schraeder for his discussions.

References Barman, S. M., and G. L. Gebber. 1980. Sympathetic nerve rhythm of brain stem origin. Am J Physiol (Regul Integr Comp Physiol): 239 (8) R42–R47. Barman, S. M., and G. L. Gebber. 1981. Brain stem neuronal types with activity patterns related to sympathetic nerve discharge. Am J Physiol (Regul Integr Comp Physiol) 242 (11): R335–R341. Barman, S. M., and G. L. Gebber. 1982. Hypothalamic neurons with activity patterns related to sympathetic nerve discharge. Am J Physiol (Regul Integr Comp Physiol) 242 (11): R34–R43. Barman, S. M., and G. L. Gebber. 1984. Spinal interneurons with sympathetic nerve-related activity. Am J Physiol (Regul Integr Comp Physiol) 247 (16): R761–R767. Basar, E. 1976. Abstract methods of general systems analysis. In Biophysical and Physiological Systems Analysis, ed. E. Basar, 23–47. Reading, MA: Addison-Wesley. Birks, R. I. 1978. Regulation by patterned preganglionic neural activity of transmitter stores in a sympathetic ganglion. J Physiol 280: 559–572. Birks, R. I., W. Laskey, and C. Polosa. 1981. The effect of burst patterning of preganglionic input on the efficacy of transmission at the cat stellate ganglion. J Physiol 318: 531–539. Gebber, G. L., and S. M. Barman. 1980. Basis for 2–6 cycle/s rhythm in sympathetic nerve discharge. Am J Physiol (Regul Integr Comp Physiol) 239 (8): R48–R56. Gebber, G. L., and S. M. Barman. 1981. Sympathetic-related activity of brain stem neurons in baroreceptor-Â�denervated cats. Am J Physiol (Regul Integr Comp Physiol): 240 (9): R348–R355. Hahn, F. 1960. Analeptics. Pharmacol Rev 12: 447–530. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. A comprehensive review of the literature and proposed mechanisms. Acta Neurol Scand 63 (Suppl. 82): 1–66. Jun, H. W. 1976. Pharmacokinetic studies of pentylenetetrazol in dogs. J Pharmacol Sci 65: 1038–1041. Kiloh, L. G., A. J. McComas, and J. W. Osselton. 1972a. The neural basis of the EEG. In Clinical Encephalography, ed. L. G. Kiloh, A. J. McComas, and J. W. Osselton, 21–34. London: Butterworths. Kiloh, L. G., A. J. McComas, and J. W. Osselton. 1972b. Normal ἀndings. In Clinical Encephalography, ed. L. G. Kiloh, A. J. McComas, and J. W. Osselton, 52–70. London: Butterworths. Krohn, W. 1977. Causes of death among epileptics. Epilepsia 4: 315–321.

480 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuni-form cardiac sympathetic nerve discharge. Mechanisms for coronary occlusion and digitalis-induced arrhythmia. Circulation 57: 1058–1065. Lathers, C. M., P. L. Schraeder, and S. B. Camel. 1984. Neural mechanism in cardiac arrhythmia associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34: 1919–1936. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Meldrum, B. S., and J. B. Brierley. 1973. Prolonged epileptic seizures in primates. Arch Neurol 28: 10–17. Meldrum, B. S., and R. W. Horton. 1973. Physiology of status epilepticus in primates. Arch Neurol 28: 1–9. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Stone, W. E. 1972. Systemic chemical convulsants and metabolic derangements. In Experimental Models of Epilepsy. A Manual for the Laboratory Worker, ed. D. P. Purpura, J. K. Perry, D. Tower, D. M. Woodbury, and R. Walter, 407–433. New York, NY: Raven Press. Van Buren, J. M. 1958. Some autonomic concomitants of ictal automatism. A study of temporal lobe attacks. Brain 81: 505–528. Wasterlain, C. G. 1974. Mortality and morbidity from serial studies. An experimental study. Epilepsia 15: 155–176.

Relationship of the Lockstep Phenomenon and Precipitous Changes in Blood Pressure

30

Amy Z. Stauffer Jeffrey M. Dodd-O Claire M. Lathers

Contents 30.1 30.2 30.3 30.4

Introduction Method Statistical Analysis Results 30.4.1 LSP Pattern versus Mean Proportion of Time Spent in Precipitous Blood€Pressure Changes 30.4.2 LSP Pattern versus Mean Proportion of Precipitous Blood Pressure Changes 30.5 Discussion 30.6 Summary Acknowledgments References

481 483 485 486 486 487 489 492 493 493

30.1â•…Introduction Sudden unexplained death (SUD) is deἀned as “nontraumatic death occurring in an individual within minutes or hours of the onset of the ἀnal illness or ictus” (Jay and Leestma 1981). No anatomic cause of death can be demonstrated at autopsy. The prevalence of SUD in persons with epilepsy has been estimated to be between 1 death in 525 epileptic persons and 1 per 2100 (Leestma et al. 1984). Autonomic dysfunction occurring in conjunction with epileptogenic activity and causing fatal cardiac arrhythmia and/or arrest has been postulated as an explanation for the increased prevalence of SUD in persons with epilepsy (Leestma et al. 1984). Lathers and Schraeder (Lathers and Schraeder 1982; Schraeder and Lathers 1983) demonstrated that autonomic dysfunction occurs during ictal and interictal epileptogenic activity induced by pentylenetetrazol (PTZ). Lathers et al. (1987) deἀned the lockstep phenomenon (LSP), during which postganglionic cardiac sympathetic neural discharge occurs as if time locked to electrocorticographic epileptiform discharges (Figure 30.1). The purpose of this study is to determine whether LSP in any of its patterns is related to sudden alterations in autonomic function as manifested by precipitous changes in mean arterial blood pressure. A better understanding of the LSP and how it correlates with changes in 481

(a)

0

5

(b)

Time in seconds

Vagal neural discharge

Postganglionic cardiac sympathic neural discharge

Electrocorticogram

Time in seconds

Vagal neural discharge

Postganglionic cardiac sympathic neural discharge

Electrocorticogram

482 Sudden Death in Epilepsy: Forensic and Clinical Issues

0

5

Figure 30.1╇ Electrocorticographic and autonomic cardiac neural discharge monitored in one cat. The records in each panel, from top to bottom, illustrate the ECoG and neural discharge for two sympathetic postganglionic and the vagal cardiac nerves, respectively. (a) Data obtained during the control period; (b) data recorded 43 s after the administration of PTZ of 10 mg/ kg (i.v.). (From Lathers, C. M., Electroencephalogr Clin Neurophysiol 67, 247–259, 1987. With permission.)

physiological parameters such as mean arterial blood pressure, cardiac neural discharge, and cardiac rate and rhythm will help to delineate mechanisms that may contribute to sudden death in the epileptic person. Such an understanding should then allow one to develop better drugs and/or combination of drugs to prevent the physiologic changes that may result in a fatal event.

Relationship of the LSP and Precipitous Changes in Blood Pressure

483

30.2â•… Method Nine cats were intravenously (i.v.) anesthetized with 80 mg/kg alpha-chloralose. TraÂ� cheostomies were performed and the femoral arteries and veins were cannulated. The animals were ventilated with a small-animal respirator, maintaining pO2, pco 2, and pH within an acceptable physiological range. Intermittent doses of gallamine (2 mg/kg i.v.) maintained paralysis. After bilateral frontal craniectomy, the dura was resected to allow placement of electrodes for recording of the electrocardiogram (ECoG). A thoracotomy and right partial pneumonectomy were performed to expose the cardiac postganglionic sympathetic and right cardiac vagal nerves. Sympathetic nerves were identiἀed by their response to a blood pressure drop induced by histamine (5 µg/kg i.v.). The activity of sympathetic nerves, ECoG, ECG (lead II), and the mean arterial blood pressure were recorded and stored on magnetic tape. Rectal temperature was maintained at 37.5–38.5°C. Each cat received 20 mg/kg intravenous phenobarbital infused over 10 min after a 10-min control period. One hour after completion of the infusion of phenobarbital, the animals received six doses (10, 20, 50, 100, 200, and 2000 mg/kg) of PTZ intravenously at 10-min intervals. Although the half-life of PTZ in the cat is unknown (Knoll Laboratories, personal communication), it has been shown to be 1.4 h in the dog (Jun 1976). If the assumption is made that the half-life of PTZ in the cat is similar to that in the dog, then the doses of PTZ were cumulative. Three categories of epileptiform discharges were deἀned according to duration. Polyspike discharges of 10 or more seconds were classiἀed as prolonged ictal. Repetitive polyspike discharges of less than 10-s duration that were interrupted by periods of depression of cerebral activity were classiἀed as brief ictal. Prolonged ictal and brief ictal activities are analogous to the activity seen in the EEG during clinical seizures. Discrete paroxysmal spikes and/or polyspike and wave discharges, which are analogous to the nonictal epileptogenic activity seen in the interictal EEG of patients with a seizure disorder, were classiἀed as interictal spikes. The term “lockstep phenomenon” is used to describe the occurrence of sympathetic postganglionic discharge and various types of epileptogenic ECoG activity in a timelocked fashion (Dodd-O and Lathers, this book, Chapter 29; Lathers et al. 1987). In the present experiment, the parameters used to deἀne the occurrence of LSP are those utilized by Dodd-O and Lathers. When an interictal ECoG spike occurred, it was said to be time locked to a sympathetic spike only if the sympathetic spike began within 200 ms of the beginning of the interictal ECoG spike. If the ECoG activity was brief ictal in type, it was considered as a single event and was analyzed with the sympathetic neural activity occurring during the corresponding time interval. The LSP was said to exist if and only if the ἀrst spike of the sympathetic polyspike discharge began after the corresponding brief ictal discharge began, and if the last spike of the sympathetic polyspike discharge began before the ἀnal spike of the brief ictal discharge ended. Time periods during which the ECoG displayed prolonged ictal activity were not considered to contain LSP. The total amounts of time that the LSP was present and absent were analyzed. The LSP was considered to be present only when the previously mentioned criteria were fulἀlled for an uninterrupted time interval of 10 or more seconds. During this 10-s interval, at least two episodes of time-locked sympathetic postganglionic and ECoG discharges had to exist. When the incidences of sympathetic and ECoG discharges were not equal over the 10-s interval, a closer look was taken at the characteristics of the less frequently occurring

484 Sudden Death in Epilepsy: Forensic and Clinical Issues

discharge. If more than one of the less frequently occurring discharges appeared alone, the LSP was considered to be absent. If the LSP was interrupted after being present for an uninterrupted time interval of greater than 10 s, the LSP was considered to have ended if the period of interruption was greater than 4.5 s. If, however, the interruption in LSP lasted 4.5 s or less, the LSP was considered to exist without interruption. The repeated 2.8-s ECoG interval is the interspike time interval found in a previous study to exist most frequently between ECoG spikes when LSP was present (Dodd-O and Lathers, this book, Chapter 29). The time interval is measured from the end of one ECoG spike to the beginning of another ECoG spike. Contained within this 2.8-s interval may be another ECoG spike. This third ECoG spike need not be related to either of the spikes that bound the 2.8-s interval in question. This third spike is, however, related to another ECoG spike that begins 2.8 s after the third ECoG spike ends. The duration of time (in seconds) the 2.8-s interval was present was determined as follows. To be considered to exist, the foregoing criteria had to be fulἀlled for at least three consecutive 2.8-s intervals. Once present, the 2.8-s interval was considered to be present without interruption only if it continued without any interval of disruption lasting more than 4.5 s. If the repeated 2.8-s interval was present but was then interrupted for more than 4.5 s, the repeated 2.8-s interval was said to have ended. In an effort to explore the importance of LSP stability in maintaining blood pressure, a more speciἀc classiἀcation of LSP was developed. First, four categories of LSP were created. The ἀrst of these, no LSP, included all of those time intervals in which the ECoGsympathetic discharge pattern was such that conditions for the existence of LSP were not fulἀlled. If the condition for LSP were present, the rate of LSP was classiἀed as either stable or unstable. The term stable LSP was used to describe all uninterrupted time periods of 10 or more seconds during which the time interval between sympathetic spikes was unchanged. If the interspace interval was constant but then changed for no more than one interspike interval, stability was considered to be maintained. Two categories of stable LSP were observed: (1) stable LSP in which the repeated 2.8-s interval was present and (2) stable LSP in which the repeated 2.8-s interval was absent. If the interspike interval was altered on two consecutive occasions, stability of LSP was considered to be lost. The term unstable LSP was used to describe all time intervals of 10 or more seconds during which LSP existed but the criteria for “stable LSP” were absent. The four categories (no LSP, stable LSP with 2.8-s interval, stable LSP without 2.8-s interval, and unstable LSP) were analyzed, and the unstable category was divided into unstable LSP with increasing rate and unstable LSP with decreasing rate. The term unstable LSP with increasing rate was used to describe periods of unstable LSP in which the duration of interspike interval decreased between consecutive spikes. The term “unstable LSP with decreasing rate” was used to describe those periods of unstable LSP in which the duration of the interspike interval increased between consecutive spikes. These two types of unstable LSP were analyzed along with the other three LSP patterns (no LSP, stable LSP with 2.8-s interval, and stable LSP without 2.8-s interval) to determine if direction, i.e., an increasing or decreasing rate, had any effect on blood pressure changes. In summary, the LSP patterns were:

1. LSP absent 2. Stable LSP with 2.8-s interval 3. Stable LSP without 2.8-s interval 4. Unstable LSP (a) with increasing rate and (b) with decreasing rate.

Relationship of the LSP and Precipitous Changes in Blood Pressure

485

Mean arterial blood pressure changes were reviewed. After examination of the control period and the PTZ treatment period, it was found that mean arterial blood pressure changes of 23 mm Hg over a 10-s interval never occurred during the control period. This rate of change, i.e., 23 mm Hg over a 10-s period, was termed a precipitous blood pressure change. The occurrence and incidence of such changes were examined in all cats after the administration of each dose of PTZ except for the dose of 2000 mg/kg. This latter dose led to death in all cats, and thus it was decided not to analyze characterization of this event with the rest of the experiment.

30.3â•… Statistical Analysis For all cats, the amount of time in seconds that was spent in each LSP pattern was recorded. The number of precipitous blood pressure changes was also determined. Since each precipitous blood pressure change occurred over a 10-s period, the amount of time in seconds that was spent in precipitous blood pressure changes is equal to the number of precipitous blood pressure changes multiplied by 10. For each LSP pattern (no LSP, stable LSP with 2.8-s interval, stable LSP without 2.8-s interval, and unstable LSP), the proportion of time spent in precipitous blood pressure changes was calculated by dividing the amount of time spent in precipitous blood pressure changes by the total amount of time spent in that LSP pattern. This was done for each cat. To determine if any of the LSP patterns were associated with a signiἀcantly higher mean proportion of time in precipitous blood pressure changes, a one-way repeated-measures analysis of variance (ANOVA) was performed. LSP pattern was used as the independent variable, and mean proportion of time spent in precipitous blood pressure changes was used as the dependent variable. The Biomedical Program Package was used. To determine which of the LSP patterns was signiἀcantly different from the others, a post hoc Newman– Keuls test was performed if a signiἀcance of the F-ratio (p < 0.05) was found. Because the observations for each cat were proportions, the variances and means of the comparison groups were not independent of one another. Therefore, square root transformation X = x + 10 was applied to all data points before analysis. This strategy is suggested by Winer (1971) to stabilize the variances. The log transformation, X′ = log (x + 1.0) was also applied (Winer 1971). Neither strategy made a signiἀcant impact on variance heterogeneity. Since equal-sized groups weaken the impact of variance heterogeneity on α, the untransformed values were used; however, Huynh–Feldt degrees of freedom were applied to correct for heterogeneity of variance. To determine if any signiἀcant differences existed between unstable LSP with increasing rate and unstable LSP with decreasing rate, the unstable LSP category was divided into unstable LSP with increasing rate and unstable LSP with decreasing rate. The previous analysis was repeated. The next question was whether a signiἀcantly different mean proportion of precipitous blood pressure changes occurred during any of the LSP patterns. For each LSP pattern, the number of precipitous blood pressure changes occurring with that LSP pattern was divided by the total number of precipitous blood pressure changes in that cat. To adjust for the fact that more time was spent in some LSP patterns than in others, this proportion was divided by the proportion of time spent in that LSP pattern, i.e., the amount of time spent in that LSP pattern divided by the total time analyzed for that cat. This adjustment was done separately

(

)

486 Sudden Death in Epilepsy: Forensic and Clinical Issues

for each of the nine cats. A one-way repeated-measures ANOVA was performed, using LSP pattern as the independent variable and proportion of precipitous blood pressure changes as the dependent variable. The unstable LSP category was then divided into unstable LSP with increasing rate and unstable LSP with decreasing rate. The analysis was repeated to determine if there was a signiἀcant difference between unstable LSP with increasing rate, unstable LSP with decreasing rate, LSP absent, stable LSP with a 2.8-s interval, and stable LSP without a 2.8-s interval in terms of mean proportion of precipitous blood pressure changes. The difference between this analysis and the previous one lies in the fact that this analysis was limited to the time during which there were precipitous blood pressure changes, whereas the previous analysis took into account all time analyzed for each cat.

30.4â•…Results 30.4.1â•…LSP Pattern versus Mean Proportion of Time Spent in Precipitous Blood Pressure Changes In comparing the mean proportion of time spent in precipitous blood pressure changes for each LSP pattern, the one-way repeated-measures ANOVA showed borderline signiἀcance of the F ratio (Huynh–Feldt probability, p = 0.056). This indicates that a signiἀcantly different mean proportion of time was spent in precipitous blood pressure changes in one or more of the LSP patterns. The post hoc Newman–Keuls test showed that the unstable LSP pattern was signiἀcantly higher than all of the other LSP patterns in terms of mean proportion of time spent in precipitous blood pressure changes (Figure 30.2).

Mean proportion of time spent in precipitous blood pressure changes

.14 .12 .10

*

Stable LSP 2.8-s interval Stable LSP without 2.8-s interval Unstable LSP No LSP * = p < 0.05 n = 9 cats

.08 .06 .04 .02 0

LSP pattern

Figure 30.2╇ Mean proportion of time spent in precipitous blood pressure changes as a func-

tion of the LSP pattern (n = 9 cats). The one-way repeated-measures ANOVA showed borderline significance of the F ratio (Huynh–Feldt probability, p = 0.056), indicating that a significantly different mean proportion of time was spent in precipitous pressure changes in one or more of the LSP patterns. The post hoc Newman–Keuls test showed that the unstable LSP pattern was significantly higher than all other patterns in terms of mean proportion of time spent in precipitous blood pressure changes.

Relationship of the LSP and Precipitous Changes in Blood Pressure

Stable LSP 2.8-s interval Stable LSP without 2.8-s interval Unstable LSP increasing rate Unstable LSP decreasing rate No LSP n = 9 cats

.14 Mean proportion of time spent in precipitous blood pressure changes

487

.12 .10 .08 .06 .04 .02 0

LSP pattern

Figure 30.3╇ Mean proportion of time spent in precipitous blood pressure changes as a function of LSP pattern (n = 9 cats). This analysis is identical to the one presented in Figure 30.2, except that the unstable LSP category is divided into unstable LSP with increasing rate and unstable LSP with decreasing rate. Although there were no statistically significant differences among the LSP patterns (Huynh–Feldt probability, p = 0.068), the observed pattern was that unstable LSP with increasing rate contributed the highest mean proportion of time in precipitous blood pressure changes.

When the unstable LSP category was divided into unstable LSP with increasing rate and unstable LSP with decreasing rate and the analysis was repeated, the corrected (Huynh–Feldt) probability was 0.068. The observed pattern was that unstable LSP with increasing rate contributed the highest mean proportion of time in precipitous blood pressure changes, but the difference was not great enough to achieve statistical signiἀcance (Figure 30.3). 30.4.2â•… LSP Pattern versus Mean Proportion of Precipitous Blood Pressure Changes When comparing the mean proportions of precipitous blood pressure changes occurring during each of the LSP patterns (no LSP, stable LSP with 2.8-s interval, stable LSP without 2.8-s interval, and unstable LSP), the one-way repeated-measures ANOVA was not signiἀcant (Huynh–Feldt probability, p = 0.0926). This indicated that there were no statistically signiἀcant differences among the LSP patterns in terms of mean proportion of precipitous blood pressure changes. However, the observation can be made (Figure 30.4) that the unstable LSP pattern had the highest mean proportion of precipitous blood pressure changes, although this difference was not statistically signiἀcant. When the unstable LSP category was divided into unstable LSP with increasing rate and unstable LSP with decreasing rate, the ANOVA was also not signiἀcant (Huynh–Feldt

Mean proportion of blood pressure changes corrected for time

488 Sudden Death in Epilepsy: Forensic and Clinical Issues Stable LSP 2.8-s interval Stable LSP without 2.8-s interval Unstable LSP

4

No LSP n = 9 cats

3 2 1 0

LSP pattern

Figure 30.4╇ Mean proportion of precipitous blood pressure changes (corrected for time) as a function of LSP pattern (n = 9 cats). The one-way repeated-measures ANOVA was not significant (Huynh–Feldt probability, p = 0.093), indicating that there were no statistically significant differences among the LSP patterns in terms of mean proportion of precipitous blood pressure changes. The observation can be made, however, that unstable LSP had the highest mean proportion of precipitous blood pressure changes.

Mean proportion of blood pressure changes corrected for time

probability, p = 0.2105). Although there were no statistically signiἀcant differences among the LSP patterns in terms of mean proportion of precipitous blood pressure changes, it can be observed that unstable LSP with increasing rate contributed the highest mean proportion of precipitous blood pressure changes. It was also found that the mean proportion of precipitous blood pressure changes for unstable LSP with decreasing rate was higher than the other three LSP patterns (Figure 30.5).

Stable LSP 2.8-s interval Stable LSP without 2.8-s interval Unstable LSP increasing rate Unstable LSP decreasing rate No LSP n = 9 cats

4 3 2 1 0

LSP pattern

Figure 30.5╇ Mean proportion of precipitous blood pressure changes (corrected for time) as a function of LSP pattern. Unstable LSP is divided into unstable LSP with increasing rate and unstable LSP with decreasing rate. The one-way repeated-measures ANOVA showed no statistically significant differences among the LSP patterns in terms of mean proportion of precipitous blood pressure changes (Huynh–Feldt probability, p = 0.2105), but it can be observed that unstable LSP with increasing rate contributed the highest mean proportion of precipitous blood pressure changes. It can also be seen that the mean proportion of precipitous blood pressure changes for unstable LSP with decreasing rate was higher than the other three patterns.

Relationship of the LSP and Precipitous Changes in Blood Pressure

489

30.5â•…Discussion It is important to consider the autonomic neuroanatomical pathways that may be the basis for the LSP, since epileptogenic activity originating in the cortex could be transmitted to the hypothalamus and to the brain stem to alter autonomic neural control of blood pressure and cardiac rate and rhythm (Figure 30.6). The hypothalamus exerts control over the autonomic nervous system; the anterior and medial hypothalamus regulate parasympathetic function and the posterior and lateral hypothalamus regulate sympathetic function. Direct

Motor cortex Prefrontal cortex

Orbitofrontal cortex Pyriform cortex

Pyramidal tract

Anterior olfactory nucleus

?

Septal area

Temporal lobe MFB (Medial forebrain bundia)

Cingulate gyrus Thalamus

Dorsomedial

Hippocampus

Anterior thalamic nuclei

Fornix

Mammillo-thalamic tract ?

MFB

Hypothalamus Mammillotegmental tract

Stria terminalis Ventral amygdalofugal pathway

MFB

Amygdala

Reticular formation

Midbrain

Multisynaptic pathway

Hypothalamospinal tract

Pons

A5

Medulla

Baroreceptors

Cardiovascular areas Other input from medullaa

Reticulospinal/ reticulobulbar tracts

Preganglionic neurons in spinal cord and brain stem Preganglionic nerves

Autonomic ganglia Postganglionic nerves

Heart and blood vessels a

Includes A1 Neurona, Raphe, Nucleus Tractus Solitarius, and C1 area

Figure 30.6╇ Numerous neuroanatomical pathways exist by which epileptogenic activity may

be transmitted to the autonomic nervous system, resulting in LSP and associated autonomic changes.

490 Sudden Death in Epilepsy: Forensic and Clinical Issues

projections exist from the hypothalamus to the preganglionic sympathetic neurons of the intermediolateral cell column and to the parasympathetic nuclei of the brain stem (Saper et al. 1976). The mammillotegmental tract connects the hypothalamus with the reticular formation, which contains multisynaptic descending pathways linking the hypothalamus with autonomic areas in the brain stem and spinal cord. The medullary reticular formation contains cardiovascular areas that can produce changes in heart rate and blood pressure. These areas produce their effects through reticulospinal connections to the intermediolateral cell column and through connections to preganglionic parasympathetic neurons (Willis and Grossman 1981). The intermediolateral cell column also receives input from the following regions of the medulla: A1 (norepinephrine-containing) neurons, C1 area (epinephrine-containing) neurons, raphe, and the nucleus tractus solitarius. The nucleus tractus solitarius receives afferent projections from the arterial baroreceptors. Aside from its medullary input, the intermediolateral cell column receives projections from the A5 area, a group of norepinephrine-containing neurons that is located in the pons (Natelson 1985). The hypothalamus is also connected to the brain stem by the medial forebrain bundle, which projects to the midbrain reticular formation, providing another pathway for descending control over the autonomic nervous system. The medial forebrain bundle connects the hypothalamus with structures in the forebrain, such as the septal area, anterior olfactory nucleus, and the pyriform cortex. The medial forebrain bundle also contains ἀbers from the fornix and the orbitofrontal cortex. The orbitofrontal cortex projects to the thalamus and the amygdala (Korner 1979), which is connected to the hypothalamus by the stria terminalis and the ventral amygdalofugal pathway. A close link between the hypothalamus and the limbic system exists through the Papez circuit. Papez (1937) postulated that information from the cortex travels by way of the cingulate gyrus to the hippocampus, which projects to the hypothalamus via the fornix. The mamillothalamic tract connects the mammillary bodies to the anterior nuclei of the thalamus. The circuit is completed by a pathway connecting the thalamus and the cingulate gyrus. Electrical stimulation of the hypothalamus or other structures in the Papez circuit results in autonomic responses. Autonomic responses also occur as a result of stimulation of other areas of the cortex. Wall and David (1951) described three cortical areas in monkeys in which blood pressure changes of greater than 10–20 mm Hg could be produced with electrical stimulation. The ἀrst of these areas is the sensorimotor cortex; the descending pathway from this area is independent of the hypothalamus and closely related to the pyramidal tract. A direct corticospinal pathway that is independent of the hypothalamus was also described by Landau (1953). The second area is the orbitofrontal cortex; the pathway that begins in this area passes through the hypothalamus. The third of the areas described by Wall and David is the anterior temporal lobe; the pathway from this area of the cortex is partially dependent on the hypothalamus and partially direct to the tegmentum and pons. Although it has not been established that neocorticohypothalamic connections exist in man, the possibility has been raised that the dorsomedial nucleus of the thalamus, which connects with the prefrontal cortex, has projections to the hypothalamus (Breusch 1984), through which epileptogenic activity originating in the cortex could reach the hypothalamus and brain stem, resulting in discharge from the autonomic nervous system and phenomena such as blood pressure changes and cardiac arrhythmia. Electrical stimulation of the cerebral cortex in man and animals results in autonomic responses such as changes in blood pressure (Chapman et al. 1950; Hoff and Green 1936), changes in heart rate, and dilatation of the pupil (Hoff and Green 1936). Blood pressure changes have been

Relationship of the LSP and Precipitous Changes in Blood Pressure

491

evoked through stimulation of various regions of the cortex, including tips of the temporal lobes (Chapman et al. 1950), motor cortex, premotor cortex, parietal cortex, and cingulate gyrus (Hoff and Green 1936). Furthermore, autonomic changes, including alterations in blood pressure, have been associated with epileptogenic activity. The precipitous blood pressure changes observed in our experiments were not consistent in terms of direction; there was a total of 46 increases and 48 decreases observed during the 60-min periods monitored in each of the nine cats. Experiments involving electrical stimulation of the cortex have shown similar blood pressure changes in terms of direction. Wall and David (1951) and Delgado (1960) found both increases and decreases in blood pressure upon stimulation of the cortex. Kaada et al. (1949) also elicited both increases and decreases in blood pressure with electrical stimulation of the cortex, as well as biphasic responses in which an increase in blood pressure was followed by a secondary decrease or a decrease in blood pressure was followed by a secondary increase. In our experiments, biphasic responses (deἀned for the purpose of this discussion as two precipitous blood pressure changes of opposite direction occurring within 10 s of each other) were seen nine times in four cats. Hoff and Green (1936) demonstrated pressor and depressor points in the cortex, which are located 2–4 mm from each other. This observation may help to explain the inconsistency of the direction of precipitous blood pressure changes, the relationship of blood pressure changes to unstable LSP, and why the direction of blood pressure changes was not related to the direction of unstable LSP (increasing rate or decreasing rate). If, for example, predominantly pressor areas were being stimulated by epileptogenic activity and the frequency of the epileptogenic activity increased, the blood pressure might increase; if the frequency of the epileptogenic activity decreased, the blood pressure might decrease. If, however, predominantly depressor areas were stimulated by the epileptogenic activity, the blood pressure might decrease with increasing epileptogenic activity and increase with decreasing epileptogenic activity. The question that remains is how the mean arterial blood pressure can decrease in the presence of an increased rate of postganglionic cardiac sympathetic discharge. In cats treated with ouabain, Lathers et al. (1977) demonstrated “nonuniformity” of postganglionic cardiac sympathetic discharge; i.e., some cardiac sympathetic branches showed increased activity, some showed decreased activity, and some branches showed no change in activity. This nonuniform discharge was associated with ouabain-induced arrhythmia (Lathers et al. 1977). Lathers and Schraeder (1982) found a similar nonuniform sympathetic discharge in cats treated with PTZ. If two or three branches of postganglionic cardiac sympathetic nerves exhibit this nonuniform discharge pattern, then the possibility exists that the discharge from the sympathetic nerves that innervate the blood vessels and control blood pressure may not be uniform. In this case, blood pressure could decrease although discharge in some cardiac sympathetic branches was increased. This discussion has concentrated mainly on epileptogenic activity originating in the cortex because one of the factors associated with SUD is the presence of structural lesions of the brain that are thought to cause seizures. In one study, autopsies of 60% of epileptic persons who died suddenly revealed structural lesions of the brain, including old contusions of the frontal and temporal lobes, brain tumors, cortical malformations, evidence of craniotomy, focal atrophy or hemiatrophy, Wernicke’s encephalopathy, and cryptic vascular malformation (Leestma et al. 1984). In another study (Freytag and Ingraham 1964), 63% of cases of SUD associated with epilepsy were found to have structural brain lesions.

492 Sudden Death in Epilepsy: Forensic and Clinical Issues

However, epileptic persons without such lesions may also die suddenly; furthermore, autonomic symptoms occur in association with generalized epilepsy. It is thought that the thalamus and the midbrain reticular formation are involved in generalized epilepsy, perhaps as a site of origin for the seizure activity (Kiloh et al. 1980). The thalamus, being part of the Papez circuit and having connections to the cortex, is capable of transmitting epileptogenic activity to the hypothalamus, resulting in autonomic manifestations. The reticular formation contains the multisynaptic pathways that connect the hypothalamus with the autonomic areas in the brain stem and spinal cord. Seizure activity occurring in the reticular formation could be transmitted to the autonomic areas of the brain stem and spinal cord by these pathways.

30.6â•…Summary The results of this study show that autonomic changes, i.e., precipitous blood pressure changes, are associated with the unstable LSP patterns. These ἀndings indicate that the LSP (and its patterns) should be investigated further to determine its relationship to cardiac arrhythmias and epilepsy-related SUD. At least three mechanisms can be postulated through which LSP may be related to arrhythmia and SUD in persons with epilepsy. The ἀrst of these is excessive sympathetic stimulation of a heart that is already electrically unstable due to prior damage. It is the opinion of Jay and Leestma (1981) that this is the case; they describe pathological changes in the myocardium that several investigators have found in patients with epilepsy who died suddenly. The pathological changes are consistent with repeated high levels of catecholamines and resemble those produced in experimental animals by sympathetic stimulation. Jay and Leestma have suggested that this damage to the heart provides a locus where fatal arrhythmias can begin when the heart is again stimulated by sympathetic discharge. The LSP may be the link between the epileptogenic activity in the brain and sympathetic stimulation of the heart. The second possible mechanism involves nonuniform discharge of the postganglionic sympathetic nerve branches, which is associated with arrhythmias caused by administration of ouabain (Lathers et al. 1977). As mentioned earlier, Lathers and Schraeder (1982) found a similar nonuniform sympathetic discharge pattern in cats treated with PTZ. In the latter study, the nonuniform cardiac neural discharge was associated with epileptogenic activity and changes in the autonomic parameters of mean arterial blood pressure and cardiac rhythm. It was suggested that these changes may contribute to SUD. The third mechanism is especially relevant to this study; precipitous blood pressure changes per se may be a factor in the development of arrhythmia in persons with epilepsy. In a study by Allen (1931), premature systolic arrhythmias followed increases in blood pressure induced by stimulation of the superior colliculus in rabbits. The arrhythmias were not observed when blood pressure was maintained at a constant level during stimulation of the superior colliculus; the conclusion was made that the arrhythmias could be attributed to the blood pressure changes. Evans and Gillis (1978) elicited blood pressure increases by stimulation of the hypothalamus and concluded that the arrhythmias that occurred after (not during) such stimulation were the result of a sudden surge of parasympathetic activity reflexly evoked by the increase in blood pressure. A case reported by Kiok et al. (1986) describes a 23-year-old man who had sinus arrest lasting up to 9 s, as well as bradycardia of 40 to 50 bpm, during clinically observed seizures.

Relationship of the LSP and Precipitous Changes in Blood Pressure

493

The authors suggest that the parasympathetic nervous system may be involved in the production of some cases of arrhythmia in epileptic persons. Although the blood pressure remained stable during the seizures in this particular case, the autonomic manifestations suggest parasympathetic involvement. Interestingly, the patient described in the report was found by computer tomography to have right temporal lobe atrophy and had subtherapeutic blood levels of anticonvulsant drugs. According to Jay and Leestma (1981), structural abnormalities of the brain and subtherapeutic levels of anticonvulsants are two factors associated with epilepsy-related SUD. The three mechanisms leading to arrhythmia and SUD that are outlined in this discussion are by no means mutually exclusive. It is quite possible that no single mechanism can explain all cases of SUD in epileptic persons. Perhaps some cases are caused by ventricular ἀbrillation related to a lower ventricular ἀbrillation threshold associated with increased cardiac sympathetic discharge, whereas other cases are caused by sinus arrest related to reflex parasympathetic discharge evoked by precipitous blood pressure changes (especially if there is some cardiac damage produced by prior sympathetic stimulation). The association of autonomic events with the LSP indicates that further investigation in this direction is warranted.

Acknowledgments The authors gratefully acknowledge Dr. Adele Kaplan for statistical analyses, and Carol Harwick, Darlene Spino, and Don Stauffer for typing the manuscript. Special thanks to Dr. Paul L. Schraeder for consultation. The research was funded by NIH grant BRSGRR04518. A.Z. Stauffer received a 1986 Medical Student Research Fellowship from the Epilepsy Foundation of America.

References Allen, W. F. 1931. An experimentally produced premature systolic arrhythmia (pulsus bigeminus) in rabbits. Am J Physiol 98: 344–351. Breusch, S. R. 1984. Anatomy of the human hypothalamus. In The Hypothalamus, ed. J. R. Givens, 13. Chicago, IL: Year Book Medical. Chapman, W. P., K. E. Livingston, and J. L. Papper. 1950. Effect upon blood pressure of electrical stimulation of tips of temporal lobes in man. J Neurophysiol 13: 65–11. Delgado, J. M. R. 1960. Circulating effects of cortical stimulation. Physiol Rev 40 (54): 146–171. Evans, D. E., and R. A. Gillis. 1978. Reflex mechanisms involved in cardiac arrhythmias induced by hypothalamic stimulation. Am J Physiol 234 (2): H199–H209. Freytag, J. R., and F. D. Ingraham. 1964. 295 medical autopsies in epileptics. Arch Pathol 78: 274–286. Hoff, E. C., and H. G. Green. 1936. Cardiovascular reactions induced by electrical stimulation of the cerebral cortex. Am J Physiol 117: 411–422. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. Acta Neurol Scand 63 (Suppl 82): 1–66. Jun, H. W. 1976. Pharmacokinetic studies of pentylenetetrazol in dogs. J Pharmacol Sci 65: 1038–1041. Kaada, B. R., K. H. Pribram, and J. A. Epstein. 1949. Respiratory and vascular responses in monkeys from temporal lobe pole, insula, orbital surface and cingulate gyrus. J Neurophysiol 12: 347–356.

494 Sudden Death in Epilepsy: Forensic and Clinical Issues Kiloh, L. G., A. J. McComas, and J. W. Osselton. 1980. Clinical Electroencephalography. London: Butterworths. Kiok, M. C., C. F. Terrence, G. H. Fromm, and S. Lavine. 1986. Sinus arrest in epilepsy. Neurology 36: 115–116. Korner, P. I. 1979. Central nervous control of autonomic cardiovascular function. In Handbook of Physiology—The Cardiovascular System, ed. R. M. Berne, N. Sperelaskis, and S. R. Geiger, Vol.€1,€691–739. Bethesda, MD: American Physiological Society. Landau, W. M. 1953. Autonomic responses mediated via the corticospinal tract. J Neurophysiol 16: 299–311. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–641. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 461–419. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25 (1): 84–88. Natelson, B. H. 1985. Neurocardiology: An interdisciplinary area for the 80’s. Arch Neurol 42: 178–184. Papez, J. W. 1937. A proposed mechanism of emotion. Arch Neurol Psychiatr 38: 725–743. Saper, C. B., A. D. Loewy, L. W. Swanson, and W. M. Cowan. 1976. Direct hypothalamoautonomic connections. Brain Res 117: 305–312. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Wall, P. D., and G. D. David. 1951. Three cerebral cortical systems affecting autonomic function. J Neurophysiol 14: 507–517. Willis, W. L., and R. G. Grossman. 1981. Medical Neurobiology, 405. St. Louis, MO: C. V. Mosby. Winer, B. J. 1971. Statistical Principles in Experimental Design, 2nd ed., 399. New York, NY: McGrawHill.

Interspike Interval Histogram Characterization of Synchronized Cardiac Sympathetic Neural Discharge and Epileptogenic Activity in the Electrocorticogram of the Cat

31

Daniel K. O’Rourke Claire M. Lathers

Contents 31.1 Introduction 31.2 Methods 31.3 Results 31.4 Discussion 31.5 Summary References

495 497 501 505 510 511

31.1â•…Introduction This is one of a series of articles (Lathers et al. 1987; also Chapters 29 and 30 of this book) attempting to characterize one potential mechanism in an animal model that may, in part, explain unexpected death in an epileptic patient. Because many of the mechanisms of death at the time of clinically observable seizures are known, our emphasis has been to determine one potential factor that may be a contributory mechanism behind unexplained interictal death. Our model has used the anesthetized cat infused intravenously with pentylenetetrazol (PTZ). This chemical is known to produce seizures (Hahn 1960; Lathers et al. 1984). It is important to note that most epileptic patients who die unexpectedly are found on autopsy to have cortical lesions, which probably served as seizure foci during life (Leestma et al. 1984). The lockstep phenomenon (LSP) theory may help explain sudden unexpected death in this population. It is deἀned by the occurrence of a one-to-one synchronization of the electrocorticogram (ECoG) discharge patterns with the discharge patterns of the peripheral autonomic nerves innervating the heart. An example of LSP is shown in Figure 31.1. We theorize that the LSP occurs when the oscillatory driver of the interictal cortical focus becomes linked to the oscillatory driver of the autonomic cardiac nerves. The actual cause of death is hypothesized to be dependent on the function of the neural discharge, which 495

(a)

5

0 (b)

Time in seconds

Vagal neural discharge

Postganglionic cardiac sympathic neural discharge

Electrocorticogram

Time in seconds

Vagal neural discharge

Postganglionic cardiac sympathic neural discharge

Electrocorticogram

496 Sudden Death in Epilepsy: Forensic and Clinical Issues

0

5

Figure 31.1╇ Electrocorticogram and two postganglionic cardiac sympathetic neural discharge

patterns monitored simultaneously. (a) Pattern of neural discharges recorded during the control period, that is, the period just before the first dose of PTZ administration. (b) Example of the lockstep phenomenon. ECoG neural burst discharge patterns are seen to be correlated one to one with those of both postganglionic cardiac sympathetic neural discharge patterns.

would then be driven at the rate of the interictal focus. Because these are autonomic nerves innervating the heart, the ἀnely tuned electrical depolarization system of the heart could be disturbed. This autonomic dysfunction may directly produce cardiac arrhythmias. These and other theories are discussed and analyzed in terms of the science of chaos. Chaos is a concept in physics that is applicable in modeling these observed natural phenomena and aids in the interpretation of our data. Chaos identiἀes pattern and regularity in seemingly

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

497

disorganized events (Winfree 1987). The science is most useful in predicting the activity of nonlinear, noncyclical functions, such as the plot of the intervals between depolarizations in an electroencephalogram (EEG) showing LSP or in the plot of ventricular depolarizations in atrial ἀbrillation monitored by an electrocardiogram (ECG). The principles of chaos help to substantiate the association of LSP and sudden unexpected death in epilepsy. Our goal in this series of papers is to demonstrate that the LSP occurs and then to characterize LSP and its effects. This study demonstrated that, with the use of the interspike interval (ISI) technique, the discharges noted in the ECoG are correlated in several important ways with the simultaneous discharges occurring in two postganglionic sympathetic nerves monitored in the same cat.

31.2â•… Methods The animal preparation used in this experiment has been described previously (Carnel et al. 1985; Dodd-O and Lathers, Chapter 13, this book; Lathers et al. 1977, 1984). Briefly, nine cats were intravenously (i.v.) anesthetized with 80 mg/kg i.v. α-chloralose. Gallamine 4 mg/kg (i.v.) was used to maintain paralysis. Leads were placed to monitor several postganglionic sympathetic nerves, the vagus nerve, the ECoG, and the ECG. A 20-mg/kg i.v. dose of phenobarbital administered over 10 min was followed by a 60-min period of stabilization before the experiment was begun. PTZ was administered intravenously at 10, 20, 50, 100, 200, and 2000 mg/kg with a 10-min interval between each dose. Differential amplitude discriminators were used to produce a better signal-to-noise ratio. The output from the differential amplitude discriminators was stored on magnetic tape and printed on a polygraph. The data on the magnetic tapes were played back and fed into a Nuclear Chicago model 7100 Data Retrieval computer, which analyzed the discharge patterns for the time between each depolarization (ISI). The tape was played from the time PTZ was ἀrst administered until the next dose was given. The summed ISIs were plotted as a histogram. An oscilloscope was used to monitor the patterns being displayed and to adjust the discriminator of the computer. A variable-pulse generator was used to standardize the discriminator and later to provide a standard peak on the ISI histogram. This was used to extrapolate the number of intervals in the unknown peaks by comparing the areas beneath the ISI histogram curve. The ISI histograms were characterized by recording the number of peaks, height, mode, and least and greatest intervals. The area of the peak was approximated by assuming a triangle formed by the highest point (the mode) and the two baseline extremes. The area was then used to calculate the number of occurrences of that interval in seconds under the curve by comparing its number to the area of a known curve. Graphs of these data are illustrated in Figure 31.2. A one-way analysis of variance (ANOVA) was used for the calculation of modes of the intervals. One cat was excluded from the computation of the statistical analysis because the ISIs showed no useful information. The signal-to-noise ratio was small enough that the discriminator on the computer could not be set to discern the interictal spikes from that of the background information. This appears to have been due to equipment failure of one of the second-stage ampliἀers. All statistics were computed excluding this cat.

498 Sudden Death in Epilepsy: Forensic and Clinical Issues Control (0 mg/kg PTZ) ECoG Intervals

100 50 0

SYMP 1 Intervals

0

1

2 (s)

3

4

0

1

2 (s)

3

4

0

1

2 (s)

3

4

100 50 0

SYMP 2 Intervals

100 50 0

(a)

ECoG Intervals

10 mg/kg PTZ 100 50 0

SYMP 1 Intervals

0

1

2 (s)

3

4

0

1

2 (s)

3

4

0

1

2 (s)

3

4

100 50 0

SYMP 2 Intervals

100 50 0

(b)

Figure 31.2╇ (a–g) Ten-minute ISI histograms of ECoG and two postganglionic sympathetics from representative cats.

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

ECoG Intervals

499

20 mg/kg PTZ 100 50 0

SYMP 1 Intervals

0

1

2 (s)

3

4

0

1

2 (s)

3

4

0

1

2 (s)

3

4

100 50 0

SYMP 2 Intervals

100 50 0

(c) ECoG Intervals

50 mg/kg PTZ 100 50 0

SYMP 1 Intervals

0

1

2 (s)

3

4

0

1

2 (s)

3

4

0

1

2 (s)

3

4

100 50 0

SYMP 2 Intervals

100 50 0

(d)

Figure 31.2 (Continued)╇

500 Sudden Death in Epilepsy: Forensic and Clinical Issues

ECoG

100 mg/kg PTZ 100

Intervals

50 0

SYMP 1

0

1

0

1

0

1

2

3

4

2

3

4

2

3

4

2

3

4

(s)2

3

4

2

3

4

(s)

100 50

Intervals

0

SYMP 2

(s)

100 50

Intervals

0 (e)

ECoG

(s)

200 mg/kg PTZ 100

Intervals

50 0

SYMP 1

0

1

0

1

0

1

(s)

100 50

Intervals

0

SYMP 2

100

Intervals

50 0

(f )

Figure 31.2 (Continued)

(s)

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

ECoG

501

2000 mg/kg PTZ 100

Intervals

50 0

SYMP 1

0

1

2 (s)

3

4

0

1

2 (s)

3

4

0

1

2 (s)

3

4

100 50

Intervals

0

SYMP 2

100 50

Intervals

0 (g)

Figure 31.2 (Continued)

31.3╅Results Three characteristic time intervals seen on the ISI histograms are summarized in Table 31.1. Their average modes were approximately the same and their lengths were constant with respect to PTZ dose administered (see Figures 31.9 through 31.11). The neural and ECoG discharge data produced in each experiment are shown in Fig� ure€31.1. We analyzed these data using the techniques discussed above and produced the ISIs shown in Figure 31.2. Note that as the dose of PTZ is increased, two characteristic peaks are seen. A third peak can be seen to occur at an interval of about 3 s. These peaks, which represent sums of intervals, were found to be statistically different with p < 0.01. In the control (Figure 31.2a) and at the lethal dose of 2000 mg/kg PTZ, there are no histogram€peaks seen because at these points the LSP does not occur. A peak at less than 0.1 s occurs on each of the histograms. This reflects the ictal activity that occurs immediately Table 31.1╅ Time Intervals for ECoG and Two Postganglionic Sympathetic Cardiac Neural Discharge Patterns Peak 1 2 3

ECoG

Symp 1

Symp 2

1.38 1.54 2.82

1.38 1.64 2.85

1.38 1.58 2.98

F value for one-way ANOVA with 2 degrees of freedom 1755a 177a 1399a â•›p < 0.01.

a

502 Sudden Death in Epilepsy: Forensic and Clinical Issues

Intervals

3500

1750

0

0

100 mg/kg PTZ

200

Figure 31.3╇ Summed ISIs for the ECoG vs. PTZ dose (eight cats included). The total number of time intervals that produced the three characteristic peaks of the ECoG ISI were summed and plotted against the dose of PTZ administered. The number of intervals increases rapidly with increasing dose of PTZ until 50 mg/kg of PTZ is reached. There is a plateau period between 50 and 100 mg/kg PTZ. The increase in the total number of intervals between 100 and 200 mg/ kg PTZ occurs only in the ECoG and represents no increase in the time spent in the lockstep phenomenon. Compare with Figures 31.4 and 31.5.

after a dose of PTZ is given and represents the ISI histogram in ictus. The data from one cat are shown here for illustration, but eight cats were used in the computations and statistical analysis. Figures 31.3, 31.4, and 31.5 should be reviewed together. They were constructed by adding the total number of intervals that occurred at any of the three characteristic time intervals and plotting this against the dose of PTZ. These graphs show a rapid incline and then a plateau. In physiologic terms, the LSP occurs at fairly characteristic time intervals, producing a similar number of intervals, independent of the dose of PTZ once the threshold dose was given. This is similar for the ECoG and for the two representative postganglionic sympathetic nerves. The number of intervals seen in each peak rose quickly with increasing PTZ dose and then reached a plateau. The implication is that the number of time intervals that constituted

Intervals

3500

1750

0

0

100 mg/kg PTZ

200

Figure 31.4╇ Summed ISIs for a postganglionic sympathetic nerve (#1) vs. PTZ dose (eight cats included). There is a sharp rise at low doses of PTZ and a plateau of PTZ doses greater than 50€mg/kg, which is probably the natural occurrence of this phenomenon.

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

503

Intervals

3500

1750

0

0

100 mg/kg PTZ

200

Figure 31.5╇ Summed ISIs for a postganglionic sympathetic nerve (#2) vs. PTZ dose (eight cats included). There is a sharp rise until 50 mg/kg PTZ then a dip to lower numbers of intervals with increasing doses of PTZ. Although the number of intervals is less than in Figures 31.3 and 31.4, the shape of the curve approximates that of the other postganglionic sympathetic cardiac nerve.

the three characteristic peaks of the ISI histogram rose until a certain plasma concentration was reached. Increasing the plasma concentration beyond this level did not increase the number of intervals at the three characteristic time intervals. The ECoG showed an increase in the number of intervals from 100 to 200 mg/kg PTZ. This was not associated with a concomitant rise in the number of intervals seen in either of the sympathetic nerves. There was, therefore, no further contribution to LSP for the increased number of intervals in this range. Figures 31.6 through 31.8 are similar to Figures 31.3 through 31.5 except that the individual histograms from each ISI plot are shown separately, summed over eight cats. This plot was used to demonstrate that each of the individual peaks follows the general trend of rising sharply to 50 mg/kg and then leveling to a plateau. In other words, when the LSP

Intervals

3500

1750

0

0

100 mg/kg PTZ

200

Figure 31.6╇ Summed ISIs for the three peaks seen on the ECoG vs. PTZ dose (eight cats

included). ECoG ISI number increases for each peak after the PTZ dose of 100 mg/kg PTZ. This trend is not repeated by the sympathetic nerves monitored. This increase represents intervals that do not appear to contribute to LSP. Average modes are 1.38 s (open square), 1.54 s (closed square), and 2.82 s (open circle).

504 Sudden Death in Epilepsy: Forensic and Clinical Issues

Intervals

2000

1000

0

0

100 mg/kg PTZ

200

Figure 31.7╇ Summed ISIs for the three peaks seen from one post-ganglionic sympathetic nerve

(#1) vs. PTZ dose (eight cats included). The plot with the 1.64-s average mode (closed square) curve represents what we surmise to be typical of LSP, that is, there is a rapid increase in the number of intervals representing LSP beginning. A plateau thereafter signifies the maintenance of LSP. The other average modes are 1.38 s (open square) and 2.85 s (open circle).

occurred, the number of intervals produced at each dose of PTZ remained fairly constant regardless of the dose of PTZ. Furthermore, it occurred at one of three characteristic time intervals. The LSP occurred at one of the shorter two time intervals more commonly than it did at the longer interval. Figures 31.9 through 31.11 are plots of modes of time intervals in seconds vs. dose of PTZ administered. Note that the length of the time interval in seconds is very nearly constant. A series of straight, nonintersecting, horizontal lines in this graph indicate that the modes of the ISI peaks are completely independent of PTZ dose. Note that the time interval with a mode at 3 s did not occur in every cat and at some doses of PTZ it did not occur at all, thus producing few data points.

Intervals

2000

1000

0

0

100 mg/kg PTZ

200

Figure 31.8╇ Summed ISIs for the three peaks seen from one postganglionic sympathetic nerve (#2) vs. PTZ dose (eight cats included). Number of intervals increases with increasing PTZ dose until a critical value of PTZ is reached. The plots then begin to plateau. Average modes are 1.38€s (open square), 1.58 s (closed square), and 2.98 s (open circle). Compare this graph with Figures 31.6 and 31.7.

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

505

Seconds

4

2

0

0

100 mg/kg PTZ

200

Figure 31.9╇ Mode of the ISI histogram in seconds for the three peaks seen on the ECoG vs.

PTZ dose (eight cats included). This plot demonstrates the consistency of the interval length in seconds of the lockstep phenomenon. Average modes are 1.38 s (open square), 1.54 s (closed circle), and 2.82 s (open circle).

31.4â•…Discussion Figure 31.12I is a representation of the anatomical location of the three oscillatory drivers of concern in LSP and cardiac arrhythmias: the cortex, the cardiac medullary center, and the sinoatrial (SA) node of the heart. In our experiment, the cardiac accelerator nerve was recorded to assess the partial output from the cardiac center in the medulla. The pathways involved in the brain between the cortex and the medullary centers were described by Stauffer et al. (Chapter 30, this book). Figure 31.12II depicts schematically the normal association of the oscillatory drivers in the brain: the interictal activity in the cortex, the cardiac medullary center in the medulla, and the SA node of the heart. Note that springs have been used to represent the interaction

Seconds

4

2

0

0

100 mg/kg PTZ

200

Figure 31.10╇ Mode of the ISI histogram in seconds for the three peaks seen on one postganglionic sympathetic nerve discharge pattern (#1) vs. PTZ dose (eight cats included). The average mode is very constant over all doses of PTZ. While the number of intervals changes with increasing PTZ dose, the length of the interval, as demonstrated here, does not change. Average modes are 1.38 s (open square), 1.64 s (closed circle), and 2.85 s (open circle).

506 Sudden Death in Epilepsy: Forensic and Clinical Issues

Seconds

4

2

0

0

100 mg/kg PTZ

200

Figure 31.11╇ Mode of the ISI histogram in seconds for the three peaks seen on one postgan-

glionic sympathetic nerve discharge pattern (# 2) vs. PTZ dose (eight cats included). Average modes are 1.38 s (open square), 1.58 s (closed square), and 2.98 s (open circle) (compare with Figures 31.9 and 31.10).

between the centers. The centers influence each other, but the depolarizations are not synchronized one to one. In Figure 31.12III, a solid bar is used to depict the association between the interictal focus and the cardiac medullary center during LSP because there is a one-to-one association between the depolarizations of these two centers. In other words, the interictal discharge is driving the cardiac medullary center. A spring still depicts the connection between the medullary centers and the heart because the association is never

I

II

III

Figure 31.12╇ A model for the interaction of the brain and heart in the lockstep phenomenon.

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

507

one to one. When the ἀrst two drivers in the brain move in synchrony, their effect is translated to the SA node to produce a change in the depolarization of the heart. A change in rate, rhythm, or aberration results. The ISI histograms in this study show the dichotomy of LSP, that is, it is either on or off. The discharge pattern characteristic of LSP, once present, was not altered appreciably by increasing the dose of PTZ until the lethal dose was given. This is demonstrated in Figures 31.9 through 31.11. The intervals were shown to be distinct from one another by a one-way ANOVA with p < 0.01. We suggest that the operative mechanism in the LSP may be that the cortex ἀres constantly, during both the ictal and interictal periods. In the ictal period, the interictal depolarizations overwhelm the nearby cells and thus the seizure discharges spread becoming clinically noticeable. In the interictal period, the cells may continue to depolarize. If these cells ἀre with an electrical potential strong enough and at the proper frequency to overtake another group of cells, then the rate of depolarization of the latter group of cells will be driven to seizure activity that will serve as the pacemaker for the second group of cells. In this experiment, the second group of cells was the nucleus of the sympathetic nerves located in the cardiac centers of the medulla. These nerves innervate the heart and exhibit both chronotropic and inotropic effects; excessive stimulation will cause cardiac arrhythmias (Randall et al. 1978). The spread of seizure activity to variable sites involves more than simply the intensity of the interictal discharge. Proximity to specialized tracts may also be of concern. During the ictal period, the intensity of the depolarization originating at the focus is strong enough to spread to areas of the brain that elicit clinically noticeable phenomena. In a generalized seizure, the spread is so extensive that the reticular activating system is involved and the patient loses consciousness. The former dichotomy proposed for the seizure focus—that is, it is intense enough either to cause a full-blown seizure or to be completely inoperative— seems incorrect. Interictal activity enables it to capture areas of the brain in close proximity by direct continuity and in distant areas by involvement of tracts that efficiently carry the impulses over long distances. The clinical expression of these dynamics is a link between the oscillatory drivers of the focus and that of, for instance, the medullary cardiac centers. Figure 31.12 diagrammatically shows our concept of the interaction among the cortex focus, the medullary cardiac center, and the heart. The SA node is a pacemaker that is influenced by the discharge of the parasympathetic and sympathetic autonomic nerves that innervate it; this neural connection between the medullary centers and the heart is represented by a spring. These centers have the ability to influence the activity of the SA node but not to directly drive it, just as a spring potentiates movement but does not directly drive it (Guyton 1981). In the normal state the rate of ἀring of the SA node, and thus the rate of the heart, and the depolarizations of the nerves reaching the heart will not be correlated one to one because the heart has its own automatic pacemaker. However, it appears that with the development of interictal and ictal discharge and the occurrence of LSP, cardiac arrhythmias may occur in an unpredictable manner. These arrhythmias may contribute to sudden death. Many systems in nature are linked in a similar fashion to the proposed LSP link. The phenomenon is called entrainment or mode locking by physicists. Gleick (1987) used the example of the relationship of the moon to the earth to demonstrate this point. In the orbit of the moon around the earth, one lunar surface is always presented to the earth. This is because the orbit of the moon is locked to the rotation of the earth. The rotation of the

508 Sudden Death in Epilepsy: Forensic and Clinical Issues

earth and the orbit of the moon are locked just as the oscillators of the epileptic focus and the medullary cardiac centers are locked one to one during the state of what we call the LSP, or what physicists call mode locking. The occurrence of a very regular oscillator in the brain is theoretically dangerous, regardless of its mechanism of known effect. The science called chaos, which has arisen in the past few decades, is of great use in explaining the phenomenon of linked drivers. Ary Goldberger, a prominent physiologist and chaostician, has written: “Fractal processes associated with scaled, broad-band spectra are ‘information-rich.’ Periodic states, in contrast, reflect narrow-band spectra and are deἀned by monotonous repetitive sequences, depleted of information content” (Goldberger et al. 1985). The fractal processes of which he speaks, ubiquitous in biological systems including the brain, are those systems that convey different information, depending on how closely you look at them. A commonly used example is that of the Mona Lisa. At a great distance, the outside borders of its frame form a rectangle, and this is all that is conveyed to the observer. At an intermediate distance, the beauty of the woman may be appreciated. At a closer distance, the mastery of the artist can be known in terms of each brush stroke. Each of the observations is unique, but together they form the treasured masterpiece. The electrical depolarizations of the brain are an example of such a fractal process. The complexity of a fractal process may at times depreciate into a simple periodic process representing decay of the system and a dramatic change. The key message conveyed by the application of the science of chaos is that simple systems, such as periodic ones, are easily perturbed and less able to return to the preperturbed state. Therefore, seeing a periodic rhythm in the brain, where there is normally rich complexity, implies a susceptibility to failure of the system, that is, death. The ultimate cause of sudden unexpected death in this population is thought to be cardiac because at autopsy no obvious cause of death can be found. The only two possible causes are failure of one of the two major electrical systems of the body, the brain and the heart. If the brain fails to send out the impulses from the respiratory centers, then respiratory failure will be the immediate cause of death. If the brain sends a message to the heart that causes it to enter a fatal arrhythmia, or if the heart enters an arrhythmia on its own, then once again sudden death will occur. It is probable that other associated factors play a role in the disturbance of the normal healthy state, such as a ἀxed lesion in the heart, which alone would not explain death and/or autonomic dysfunction, as has been published by several labs including ours (Stauffer et al., Chapter 30, this book; Van Buren 1958). Central respiratory failure and arrhythmia are the only obvious etiologies that would leave no signs at autopsy (Jay and Leestma, 1981). Arrhythmia is much more likely, given the ἀndings of this experiment. The perturbation of the electrical depolarization of the heart may have several mechanisms, any or several of which may be operational. Further investigation is necessary to sort out the causes. The difficulty in doing a study of this, however, may mandate empirical treatment. We must ἀrst have plausible mechanisms from which models can be built. The speciἀc mechanism of death in this population may be more complex than the sympathetic discharge rate, although even this theory has merit. An area of damage to the electrical stimulation system of the heart, that is, the His-Purkinje system, may be caused by continual stimulation of the beta receptors by the sympathetic nerves innervating these receptors; this stimulation might produce a ἀxed microscopic lesion, which alone would be harmless but sufficient to cause this portion of the myocardium to be less flexible in its response to other insults, for example, excessive sympathetic discharge. A

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

509

second possibility concerns the arrangement of the receptors in the ventricle. Beta receptors in the ventricles are arranged in a pattern such that their highest density is in the apex with a gradually decreasing density near the base (Lathers et al. 1986). Down-regulation of the beta receptors from continual sympathetic discharge might disturb this gradation, producing a potentially arrhythmic situation. A pattern of increased sympathetic depolarization could be the fatal step in a two-stage process. A third possibility is that of a forbidden sequence of sympathetic depolarizations, that is, a pattern of depolarizations from the sympathetics, which can interrupt the regular electrical depolarization. Winfree (1987) has begun to characterize these processes in terms of chaos modeling of the heart. More work needs to be performed to identify which of these theories or combination of theories is active in the epileptic patient who dies of sudden unexpected death. Mathematical models like those developed for the heart by Jose Jalife and colleagues (Salata and Jalife 1985) will facilitate our understanding and provide a basis for continued investigation. Our model for the demonstration of the interictal activity of LSP is, of course, only an approximation of the human with epilepsy. There are, however, the following important similarities. The focal lesions that have been demonstrated in the brains of patients who have died to interictal unexplained death have shown cortical lesions in as many as 60% of those patients autopsied (Leestma et al. 1984). Moreover, the cat has a more highly developed cortex than many other animals proposed for such models, for example, the rat, which makes the cat model more likely to represent human epilepsy. The inadequacies of any animal model clearly demonstrate why future research must be coupled with mathematical modeling to maintain good correlation and to add direction for future study. The possibility exists that our model produces the LSP in a way that does not simulate the natural phenomenon. First, epilepsy in our model was caused by intravenous administration of a chemical, PTZ, which has access to parts of the brain in proportion to their perfusion with blood. Although many areas of the cortex are exposed to the drug, probably one area eventually becomes the pacemaker of interictal discharge activity. The study used pretreatment with phenobarbital to dampen the epileptogenic activity of PTZ to allow a greater amount of time spent in interictal activity, that is, to prevent status epilepticus. This allowed us to maximize the time spent in the interictal period. Recall that our objective was to study the possible mechanism of interictal death, not ictal death. In a previous study (Lathers et al. 1987) that did not use phenobarbital, we showed that phenobarbital is not directly involved in the production of LSP. The ISI technique is a novel way of analyzing LSP. It is a rapid method of analyzing the complex discharge patterns generated by a nerve and the brain. When computing an ISI much information is lost; only an instantaneous picture is produced. It is rather like taking the ἀrst derivative of a function: the information obtained is valuable but it is now only a useful trend of the original, more complex function. Therefore, the ISI will never serve as the sole measure of LSP but rather as a very useful indirect measure of the occurrence of LSP. The computation of mean ISI and the calculation of the area from the ISI peaks were accomplished by approximating the patterns with a triangle. Several errors can be introduced in this operation. First, the area will almost always be underestimated; thus, it is more difficult to show statistical signiἀcance, making any results all the more valid. Second, the standard that was used to calculate the area-to-interval number was based on the area

510 Sudden Death in Epilepsy: Forensic and Clinical Issues

noted in the standard peak histograms, which were added to each ISI histogram. Because this area would also probably be underestimated, the calculated number of intervals would be greater than the actual number of intervals that created the histogram. We consider the actual difference to be negligible. Sudden unexplained death in the epileptic patient may prove to have many causes, but it seems important to understand the possible mechanisms that have been discussed here and to contemplate their treatment. It seems practical to think that the administration of a pharmacologic agent to an identiἀed population at risk might be an effective means of prevention. To go about proving this, we must ἀrst have a valid animal model. Second, we must be able to quantitate very closely the degree of LSP occurring, enabling the comparison of various pharmacological agents in the prevention of LSP. The technique of using ISIs to examine the characteristic intervals is a rapid-assessment technique that could be used to analyze the effect of a pharmacological agent. Several authors have used the ISI to characterize discharge patterns, for example, the action of ouabain on autonomic nerves (Lathers et al. 1977). The future application of this technique is intriguing. An ambulatory EEG monitor coupled with an ambulatory ECG monitor could be used to record the electrical events for 24 h. Given that the characteristic intervals for LSP have been identiἀed, a statistical analysis of the EEG using the ISI technique could rapidly identify these characteristic intervals and pinpoint the precise time at which they occurred. Next, the ECG could be analyzed for aberrations. A statistical correlation would be sought for those times found to be suspect in the EEG with the aberrations noted in the ECG. This would eliminate the need for direct recording of the postganglionic cardiac stimulator nerve, as has been done in this experiment, because it is quite impractical in the human.

31.5â•… Summary The association between epileptogenic activity in the ECoG and aberrations in cardiac activity was investigated by further characterizing the LSP. The pattern of neuronal discharges from several postganglionic cardiac sympathetic branches with simultaneous recordings of the ECoG and the ECG monitored in nine anesthetized cats in which epileptogenic activity was induced with PTZ was analyzed based on time intervals between action potentials in the ECoG. A similar analysis was made of the time intervals between action potentials in the nerves innervating the heart. Time intervals for both ECoG and cardiac sympathetic discharge were summed and plotted as ISI histograms using a data retrieval computer. The ISIs obtained for the ECoG and the cardiac accelerator nerve tracings were compared. Analysis and comparison of these ISIs demonstrated the occurrence of LSP. The intervals found to be most characteristic of the LSP when monitoring the ECoG were 1.38, 1.58, and 2.85 s. These intervals were found to be statistically different with p < 0.01. The implications of our ἀndings were discussed in terms of the science of chaos. The data suggest that the discharge patterns from the ECoG and the sympathetic nerves are synchronized. This aberrant discharge pattern may be associated with changes in the ECG, which may ultimately contribute to understanding the mechanism of sudden unexpected death in persons with epilepsy.

Neural Discharge and Epileptogenic Activity in the ECoG of the Cat

511

References Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbitol pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptic activity in the cat. Pharmacology 30: 225–240. Gleick, J. 1987. Chaos: Making a New Science. New York, NY: Viking. Goldberger, A. L., V. Bhargava, and B. J. West. 1985. Nonlinear dynamics of the heart beat. Physica 17D: 207–214. Guyton, A. C. 1981. Rhythmic excitation of the heart. In Textbook of Medical Physiology, ed. A. C. Guyton, 165–175. Philadelphia, PA: Saunders. Hahn, F. 1960. Analeptics. Pharmacol Rev 12: 447–530. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. Acta Neurol Scand 63 (Suppl. 82): 1–66. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbitol. Life Sci 34: 1919–1936. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986. Regional distribution of myocardial receptors. Eur J Pharmacol 130: 111–117. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Randall, W. C., J. X. Thomas, D. E. Euler, and G. L. Rosanski. 1978. Cardiac dysrhythmias associated with autonomic nervous system imbalance in the conscious dog. In Perspectives in Cardiovascular Research, vol. 2, Neural Mechanisms in Cardiac Arrhythmias, ed. P. J. Shwartz, A. M. Brown, A. Malliani, and A. Zanchetti, 123–138. New York, NY: Raven Press. Salata, J. J., and J. Jalife. 1985. “Fade” of hyperpolarizing response to vagal stimulation at the sinoÂ� atrial and atrioventricular nodes of the rabbit heart. Circ Res 56 (5): 718–727. Van Buren, J. M. 1958. Some autonomic concomitants of ictal automatism. Brain 81: 505–528. Winfree, A. T. 1987. When Time Breaks Down: The Three-Dimensional Dynamics of ElectroÂ�chemical Waves and Cardiac Arrhythmias. Princeton, NJ: Princeton University Press.

Power Spectral Analysis A Procedure for Assessing Autonomic Activity Related to Risk Factors for Sudden and Unexplained Death in Epilepsy

32

Stephen R. Quint John A. Messenheimer Michael B. Tennison

Contents 32.1 Introduction 32.2 Theory 32.2.1 Spectral Concepts 32.2.1.1 Spectral Decomposition of a Data Record 32.2.1.2 Stationarity of a Data Record 32.2.1.3 Aliasing 32.2.1.4 Windowing 32.2.1.5 Scaling of Spectral Data 32.2.2 Spectral Analysis of the ECG 32.2.2.1 Heart Rate versus Heart Period 32.2.2.2 Heart Rate Detectors 32.2.2.3 Spectral Analysis of the Heart Event Series 32.3 Procedures 32.3.1 Clinical and Experimental Protocol 32.3.1.1 Anticonvulsant Studies 32.3.1.2 Epilepsy Studies 32.3.2 Data Collection and Spectral Estimation 32.3.2.1 ECG Digitization and Heart Period Determination 32.3.2.2 Power Spectra Estimation 32.3.2.3 Data Evaluation and Presentation 32.4 Preliminary Findings 32.4.1 Anticonvulsant Effects 32.4.1.1 Carbamazepine 32.4.1.2 Phenytoin 32.4.2 Seizures 32.5 Discussion References

513

514 516 516 516 517 518 519 519 520 520 520 520 521 521 521 521 523 523 525 527 527 527 528 528 528 530 536

514 Sudden Death in Epilepsy: Forensic and Clinical Issues

32.1â•…Introduction Sudden death of unknown etiology has been an active area of research starting only in recent decades (DeSilva and Lown 1978; Falconer and Rajs 1976; Gordon et al. 1984, 1986; Guilleminault et al. 1984; Hirsch and Martin 1971; Jay and Leestma 1981; Kiok et al. 1986; Leestma et al. 1984; Myers et al. 1986; Neuspiel and Kuller 1985; Schraeder and Lathers 1983; Terrence et al. 1975). Sudden unexplained death (SUD) in epileptic persons comprises an inordinately large subpopulation of this group in relation to its occurrence in the general population (Jay and Leestma 1981; Neuspiel and Kuller 1985; Wannamaker 1985). Abnormalities in the autonomic nervous system, particularly in its regulatory function of the cardiovascular system, have been postulated as central to the causes of death (Gordon et al. 1984, 1986; Guilleminault et al. 1984; Hirsch and Martin 1971; Jay and Leestma 1981; Kiok et al. 1986; Leestma et al. 1984; Myers et al. 1986; Neuspiel and Kuller 1985; Schraeder and Lathers 1983; Terrence et al. 1975; Wannamaker 1985). The implications of autonomic involvement in SUD are particularly signiἀcant in epilepsy in that even minimal epileptogenic activity can cause profound alteration of autonomic activity and balance, with associated alterations in cardiorespiratory parameters including ECG changes, blood pressure, respiration, and vasomotor tone (Hirsch and Martin 1971; Lathers and Schraeder 1982; Schraeder and Lathers 1983; Wannamaker 1985). Electrocardiogram (ECG) effects commonly observed in experimental epilepsy include heart rate changes, arrhythmias, conduction blocks, altered ECG morphology, and QT interval changes (Lathers and Schraeder 1982, 1987; Lathers et al. 1987; Mameli et al. 1988; Schraeder and Lathers 1983). Heart rate changes without arrhythmias have been reported during seizures in patients with epilepsy (Blumhardt et al. 1986; Keilson et al. 1987; Marshall et al. 1983). In addition, many drugs used in the treatment of epilepsy have direct autonomic or cardiac effects. This may have some signiἀcance in the observation that in almost all cases of SUD in epileptics, the level of prescribed anticonvulsant is found to be subtherapeutic or absent (Hirsch and Martin 1971; Jay and Leestma 1981; Leestma et al. 1984; Neuspiel and Kuller 1985; Terrence et al. 1975). The incidence of SUD in epilepsy is not uncommon, with estimates varying from 1/1100 to 1/500 or more frequent (Jay and Leestma 1981; Leestma et al. 1984). Identiἀcation of this subpopulation of epileptics at risk for SUD may be possible through a detailed understanding of autonomic activity under circumstances peculiar to epilepsy. Of particular beneἀt would be a procedure that would enable quantiἀcation of autonomic activity dynamically, with a methodology not requiring a perturbation of the system for the purpose of measurement. The speciἀc autonomic nervous system dysfunction that may be a critical factor in the genesis of sudden death in epileptics is uncertain (Lathers and Schraeder 1982, 1987; Lathers et al. 1987; Mameli et al. 1988; Wannamaker 1985). Jay and Leestma (1981) postulate that this catastrophic and unpredictable event is a result of fatal ventricular arrhythmia due to sympathetic stimulation. A support of this hypothesis is the ἀnding of focal myocarditis-like lesions in the hearts of epileptic patients who experienced SUD (Falconer and Rajs 1976), as similar lesions have been produced in animal models by intense sympathetic stimulation of the heart. Cardiovascular homeostasis is maintained by both the sympathetic and parasympathetic portions of the autonomic nervous system, with their effects being modulated by the renin–angiotensin system. Valsalva’s maneuver, ocular compression, carotid sinus

Power Spectral Analysis

515

massage, and other transient perturbations to the cardiorespiratory control system have traditionally been used to evaluate the functional status of the autonomic nervous system. These techniques are cumbersome, perturb the system that is being assessed, and do not lend themselves to steady-state evaluation of autonomic function. In animal models, fluctuations in heart rate have been investigated by power spectral analysis and have been shown to correlate with the activity of the autonomic nervous system. Three peaks in the power spectrum have been described in dogs (Akselrod et al. 1981) to correlate with respiration (high frequency), the baroreceptor reflex (mid-frequency), and changes in vasomotor tone related to the thermoregulation system (low frequency). This distribution in spectral power, from a normal human subject, is illustrated in Figure 32.1. The high-frequency peak (centered at the frequency of respiration) is generally recognized to be mediated solely by the parasympathetic system (Eckberg 1983; McCabe et al. 1984; Pomeranz et al. 1985; Porges 1986), although sympathetic activity has been implicated in respiratory arrhythmias (Pagani et al. 1986). The mid-frequency peak, often referred to as the low-frequency peak in recent studies (Pagani et al. 1986; Pomeranz et al. 1985), is centered at approximately 0.1 Hz and is mediated at rest by the parasympathetic system alone in the supine position (Akselrod et al. 1981; Pagani et al. 1986; Pomeranz et al. 1985) and by the combined sympathetic–parasympathetic autonomic system in the sitting and

Representative spectrum

0.19 0.18 0.17 0.16 0.15 0.14 0.13 0.12 Power

0.11 0.1 0.09

Low

0.08

Mid

High

0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 0

0.2

0.4

0.6

Frequency

Figure 32.1╇ Power spectrum of HPV from a normal adult subject. The three peaks illustrate the frequency bands where spectral power is typically concentrated, which correspond to frequencies of oscillation in the heart period under autonomic control.

516 Sudden Death in Epilepsy: Forensic and Clinical Issues

standing postures (Pagani et al. 1986; Pomeranz et al. 1985). The mid-frequency peak is prominent in an upright posture and the high-frequency peak is augmented in the supine position (Baselli et al. 1987; Pagani et al. 1986; Pomeranz et al. 1985). Pagani et al. (1986) suggested a ratio of the mid-frequency to high-frequency components as an indicator of relative sympathetic–parasympathetic balance. The low-frequency peak (at approximately 0.03 Hz) has both sympathetic and parasympathetic components (Akselrod et al. 1981) and is attenuated by the renin–angiotensin system. Myers et al. (1986) found that the power in the low-frequency band was not signiἀcant in comparing groups of patients at risk and not at risk for sudden cardiac death, and they found this peak to be very sensitive to postural changes and other activity. This peak is often ignored (Pagani et al. 1986) or possibly included with the mid-frequency component (Pomeranz et al. 1985). Spectral analysis of heartbeat variability in several patient populations has been used to identify groups at risk for sudden infant death, cardiac patients at risk for sudden cardiac death, and patients with autonomic disorders, and to study the influence of vagal tone on early development (Baselli et al. 1987; Gordon et al. 1984, 1986; Nugent and Finley 1983; Pagani et al. 1986; Pomeranz et al. 1985; Porges 1986; Zwiener 1978). At present, this method has proven valuable in understanding the relationship of the autonomic system to normal and abnormal conditions rather than as predictive of patient outcome. We are unaware of the application of this method, other than our preliminary reports (Messenheimer et al. 1987; Quint et al. 1987; Tennison et al. 1987) to the study of autonomic function in epilepsy. Epileptic patients frequently undergo intensive monitoring with routine collection of ECG, electroencephalograph (EEG), and video data, providing an ideal opportunity to study the involvement of the autonomic nervous system in this disease. Heart period variability (HPV) data are therefore readily available for analysis and correlation with seizure type, sleep/wake state, and antiepileptic drug level. To the extent that the ECG is recorded in analog form in epileptic patients who suffer unexplained death, it is possible to use this method to investigate SUD in epilepsy.

32.2â•…Theory 32.2.1â•… Spectral Concepts There are many methods for estimation of the spectral content of a signal, including maximum likelihood, maximum entropy, spectral ἀltering procedures, and Fourier analysis. The most generally used of these is Fourier analysis, with the fast Fourier transform used as an efficient algorithm for computer determination of the discrete Fourier transform. 32.2.1.1â•… Spectral Decomposition of a Data Record The premise of Fourier transformation is that an arbitrary waveform may be represented by a sum of sinusoids or, equivalently, by a sum of complex exponentials. For waveforms of discrete variables, either ἀnite sequences of length N or periodic sequences with N elements in each period,

x(n) = x(0) + x(1) + … + x(N – 1)â•… 0 < n < N

the waveform may be decomposed into a sum of N complex exponentials (or sine and cosine waves) in TV multiple harmonics of the fundamental frequency.

Power Spectral Analysis

517

1 x (n) = N



N −1





∑ X(k)e ° 2Nπ  nk j

k =0

with

1 e j 2π f1nk = °cos(2π f1nk ) + j sin(2π f1nk ) 2

and fundamental frequency



f1 =

1 N

The fundamental frequency is the inverse of the duration of a ἀnite sequence, or one period of a periodic sequence. The weighting factor X(k) of the exponential for each harmonic (Fourier coefficient) gives the relative presence of this frequency in the waveform. The set of N Fourier coefficients, which represents the decomposition of the waveform into its spectral components, is generally made up of complex numbers that can be presented as real and imaginary numbers in a rectilinear coordinate system (a + jb) or as magnitude and phase in a polar coordinate system (r angle θ). The autospectrum is the magnitude€squared [r2 or (a2 + b2)], which is an estimate of the power present at each frequency component (or power spectrum). The estimate of autospectra differs from the true power spectrum because of the presence of both random and bias errors that are either present in the data or introduced through computation of the estimate due to ἀnite register-length effects. The mean value of the data is the zero frequency component in the corresponding spectral estimate. This term may be treated independently of the dynamic portion of the data, and it is advantageous to subtract this term before computation of the spectral estimate. The power spectrum determined from a single record of time data is termed an inconsistent estimate because the random error at each spectral point is constant regardless of the sample size. Each spectral component has 2 degrees of freedom, independent of the number of data points in the time record. Averaging the spectral estimates of sequential records at each frequency increases the degrees of freedom at each spectral point in proportion to the number of records averaged, so that the random error approaches zero for large sample sizes. This process produces a consistent estimate of the power spectrum. However, spectral averaging will only produce a reliable estimate of the true spectra if the data are stationary. 32.2.1.2â•… Stationarity of a Data Record Stationarity of a time record is satisἀed provided sequential records have the same averaged properties (all moments are equal). A data set is weakly stationary if its mean and autocorrelation function do not vary with time in the set. This condition can be met, in practice, only as a matter of degree. Although various tests are available to assess the stationarity of data sets, weak stationarity is often judged qualitatively by visually comparing sequential data records for transients and by comparing the Fourier transforms of the autocorrelation functions. Because heart rate is influenced by a multitude of inputs (exercise, anxiety,

518 Sudden Death in Epilepsy: Forensic and Clinical Issues Simulated instantaneous rate change

2.6 2.4 2.2 2

R –R Interval

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 (a)

20

40

60

80

100

120

R wave

Figure 32.2╇ (a) Illustration of the effect of an impulse on the power spectrum. Simulated heart period data with 128 beats at 1 beat/s, with a single 2-s interval between beats 63 and 64.

body position, etc.), it is essential to keep the environment as constant as possible during collection of heart period data to minimize nonstationarities. When transients and other nonstationarities appear in a data record, it is possible to remove them under some circumstances with ἀltering and other procedures (Bendat and Piersol 1986; McCabe et al. 1984; Porges 1986). More generally, however, the portion of the record containing the nonstationarities is omitted from the analysis (Akselrod et al. 1981; Baselli et al. 1987; Myers et al. 1986; Pagani et al. 1986). 32.2.1.3â•… Aliasing For the determination of the spectral content of a signal by discrete computational procedures, the data set representing the signal must be both ἀnite and represented by discrete numbers. A ἀnite discrete data set may be obtained from a continuous signal by sampling at a regular interval for a ἀnite number of representative values. The sample rate must be fast enough to capture all of the information in the data, or, in terms of the sample interval, the time between samples must be short enough to represent all of the variation in the data. Fluctuations present in the data at a frequency faster than half the sample frequency will be represented at a slower frequency (aliasing).

Power Spectral Analysis

519 Simulated instantaneous rate change

0.035

0.03

0.025

Power

0.02

0.015

0.01

0.005

0 0 (b)

0.2

0.4 Frequency

Figure 32.2 (Continued)╇ (b) Power spectrum of the unit sample impulse. The oscillatory effect at low frequency is due to the mean removal and windowing of the time sequence data.

32.2.1.4â•… Windowing Truncation of the data set at the end points introduces frequency components that may not properly represent the spectral content of the data. Windowing has the effect of tapering, rather than truncating, the mean-zero data to zero at both ends of the record. The resulting distortion of the data in the frequency domain corresponds to convolving the autospectrum of the heart period data with the autospectrum of the window. 32.2.1.5â•… Scaling of Spectral Data When presented in graph form, the autospectrum (or power spectrum) is commonly scaled logarithmically, linearly, or as a square root. Both square-root scaling (also called an amplitude spectrum) and logarithmic scaling of the autospectrum have the advantage of compressing a large range of data for presentation on a single graph. However, for visual interpretation, this compression causes moderate variations in the spectral data to become indistinct. Linear scaling of the autospectra is best suited for visual distinction of the amplitude variations of interest in HPV data. However, this may require matching the scale of the vertical axis to the actual data. Detailed presentations on the theory of discrete Fournier analysis and the analysis of random data can be found in the work of Oppenheim and Schafer (1975) and Bendat and Piersol (1986).

520 Sudden Death in Epilepsy: Forensic and Clinical Issues

32.2.2â•… Spectral Analysis of the ECG 32.2.2.1â•… Heart Rate versus Heart Period Because the dynamic activity of the autonomic nervous system is most directly reflected in the time between successive heartbeats, spectral analysis is generally performed on the fluctuations in time or frequency between successive beats rather than on the ECG waveform itself. Although many investigators use instantaneous heart rate (the inverse of heart period) as the independent variable, we use heart period, which is a direct measure of time. Analysis of either measure of fluctuation between beats yields qualitatively similar results. The time between successive beats of the heart, generally measured as the R–R interval, is a discrete random variable. Obviously, it is necessary to obtain a measure of this variable before further analysis is possible. 32.2.2.2â•…Heart Rate Detectors The most convenient method to obtain a discrete representation of heart rate is to sample the analog output of a clinical heart rate detector at a rate near the heart rate. There are several problems with this technique. First of all, the ECG is subject to noise from a variety of sources (Friesen et al. 1990), and long records of constant high signal-to-noise measurements are rare under all but the best of circumstances. The algorithm for detection of the R wave used in the clinical detector is generally unknown to the user, and all algorithms for heart event detection are subject to both missed and false beats (Friesen et al. 1990). Erroneous data, which introduce a positive or negative impulse to the data set, are then present in the sample heart period data for spectral analysis without a procedure for observation or correction of the error. An impulse (or scaled unit sample in discrete signal-Â�processing terminology) contributes equal power to all frequencies of the power spectrum (Figure 32.2). In generating an analog output for sampling, the heart rate detector introduces a data hold characteristic (generally ἀrst order) into the signal, with associated random and bias errors. Selection of the sample rate for the instantaneous heart rate output is another problem that can introduce error in the data. Sampling too slow will introduce aliasing and sampling too fast generates redundant data and ultimately reduces the signiἀcance in the data (less than 2 degrees of freedom per spectral point). Because the very characteristic of interest in the data is the dynamic variation of heart rate, there is no completely satisfactory solution to the problem of sample rate selection with this technique. 32.2.2.3â•…Spectral Analysis of the Heart Event Series As an event series, the heart period (or heart rate) data are in the proper format for discrete Fourier analysis. However, there are both practical and theoretical problems with estimating the power spectra directly from the event series. The practical difficulty is, again, the problem of reliable and veriἀable detection of heart period. A procedure is required for determining the R–R interval from long records of noise-contaminated ECG without missing or falsely detecting an event. By digitizing the ECG waveform, it is possible to use a detection algorithm to mark events and then rapidly review the marked waveform with an editing facility to correct errors. Spectral analysis of the discrete event series has the advantage of not introducing the errors related to converting the discrete data to an analog signal and then sampling it at a regular interval. In particular, the random and bias errors associated with estimation of regularly spaced samples are avoided; and because the data are discrete, there is no aliasing.

Power Spectral Analysis

521

However, this approach has two theoretical problems. First, the frequency dimensions are in cycles per event rather than in hertz. This variable may be converted to hertz by multiplying by the mean event rate, with an error related to the amount of variability in the data. This approximation of radian frequency to Hertz is generally considered reasonably good for heart event series (DeBoer et al. 1984; Mohn 1976), and this direct procedure for estimation of the spectral content of the heart (or rate) is widely used (Akselrod et al. 1981; Haddad et al. 1984). However, because of the indirect relationship of the event series to time, it is difficult to determine the cross-correlation or cross-spectra between this variable and other physiological measures not directly tied to the heart event. For investigations of the interaction between the heart period (or rate) and most other time-related variables, it is necessary to convert the event series to a time series with regularly occurring samples. Procedures for this conversion are presented by DeBoer et al. (1984) and Berger et al. (1986). An excellent presentation of theoretical considerations for spectral analysis of heart rate is given by Sayers (1980).

32.3â•… Procedures 32.3.1â•… Clinical and Experimental Protocol Data are collected from a variety of patients and experimental subjects. Recording of the ECG is always included in our epilepsy monitoring procedures. The ECG is recorded with two chest leads, placed to accentuate the QRS amplitude and minimize the T wave. Data from volunteer subjects were collected in the Clinical Research Unit of the North Carolina Memorial Hospital, in patients admitted for evaluation of epilepsy, and in epileptic patients monitored at home. The physiological signals monitored in addition to the ECG, as well as the recording procedures, vary as appropriate to the circumstances. 32.3.1.1â•… Anticonvulsant Studies In studying the effects of anticonvulsants on the autonomic activity in normal subjects, we record only the ECG through online computer digitization of this waveform. The ECG is recorded from subjects at rest in a standing or supine position in a quiet room. The ECGs recorded in each volunteer subject continuously for 10 min in the absence of the drug (control) and again after administration of the anticonvulsant, coincident with its peak concentration in the blood. The anticonvulsants we used are carbamazepine administered orally (600 mg) and phenytoin given intravenously in a single 250-mg dose. Anticonvulsant serum concentrations are obtained corresponding to the time that the data for HPV are collected. 32.3.1.2â•…Epilepsy Studies All inpatients and outpatients undergoing diagnostic or presurgical epilepsy monitoring are studied using a commercially available (Telefactor) 16-channel radio telemetry-based system (CCTV/EEG). Montages, which include an EEG channel and up to 15 channels of EEG, are recorded on videotape (Figure 32.3) along with a video image of the subject. Recorded seizures that are technically acceptable are selected and replayed, with EEG and ECG hard copy printed and the ECG digitized and stored by the computer system. Separate preictal, ictal, and postictal segments are then evaluated. The hard copy record is used to

50 µV

8. C4–P4

7. F4–C4

6. FP2–F4

5. T6–02

4. T4–T6

3. F8–T4

2. FP2–F8

1. ECG

1s

LLF = 1 Hz HLF = 70 Hz

211

522 Sudden Death in Epilepsy: Forensic and Clinical Issues

Figure 32.3╇ Strip chart tracing of the ECG and seven channels of EEG at the onset of an ictal episode in a patient with partial complex seizures localized to the right posterior temporal region (see case 1, Section 32.4.2). The electrographic ictal onset is marked with an arrow in channel 5. Marked changes in the rhythmicity of the ECG (see Figure 32.7) preceded the EEG changes by approximately 16 beats (10 s).

Power Spectral Analysis

523

relate the onset of seizure activity, as determined from the EEG and/or behavioral state noted from the video signal, to the corresponding location in the digitized record stored in a computer ἀle. A commercially available eight-channel recording system (Oxford), with one channel dedicated to ECG, is used in outpatient ambulatory monitoring. Selected segments from these cassette tapes are analyzed in a manner similar to the procedure used in CCTV/EEG monitoring. The absence of a video picture as a behavioral correlate to the ECG and fewer channels of EEG are disadvantages of this type of monitoring. We tested both the CCTV/ EEG and ambulatory recording systems and veriἀed that neither has tape speed fluctuations that contribute to the time measured between heartbeats (less than 1 ms/s). 32.3.2â•…Data Collection and Spectral Estimation We extract the heart period from the ECG through a process of digitizing the ECG, detecting the R wave of each QRS complex in the ECG, and measuring each successive R–R interval (Figure 32.4). Spectra are determined from stationary records of heart period data. 32.3.2.1â•… ECG Digitization and Heart Period Determination The amplified ECG is displayed on an oscilloscope (Tektronix R5113) along with€ a potentiometer-adjusted threshold level, either online from a Hewlett Packard 7830 A€isolated ECG monitor or off-line from video or cassette tape. This analog waveform is sampled using a Digital 11/73 computer system, with the position of the threshold crossing

Schmitt trigger level

ECG

Isolated ECG monitor

DC amplifier

Real time

Graphics display

Clock

A–D converter D–A converter

LSI - 11 computer

Analog

IBM - AT

tape

computer

Graphics

Graphics

printer

display

Figure 32.4╇ Schematic diagram showing the procedure for digitizing the ECG, detection of R–R intervals, and processing of the resultant heart period data.

524 Sudden Death in Epilepsy: Forensic and Clinical Issues

(a)

(b)

Figure 32.5╇ Digitization and threshold detection of the QRS complex, and R–R interval deter-

mination. (a) The continuously digitized ECG is displayed as five sequential traces on the digital storage scope, after which the screen is erased and the process repeated. Threshold crossings are marked with an arrow. (b) During processing, sequential 500-point records of the digitized ECG are recalled from disk and displayed on the digital scope with the threshold crossings moved to the local peak (R wave). A file is simultaneously created that contains the times between arrows (heart period).

of€each€QRS wave stored along with the digitized waveform (Figure 32.5a). During data processing, sequential 500-point records of the waveform are recalled from disk and rapidly displayed on a digital oscilloscope (Hatachi V131), with the threshold crossing points moved to the local peak (R wave) and marked with an arrow (Figure 32.5b). The number of points between arrows (R–R interval) is written to a ἀle along with the sample rate. A heart period sequential plot is generated from this ἀle and displayed on the terminal screen; it may also be printed (Figure 32.6). Missed beats and/or spurious noise erroneously marked as beats are easily recognized in this display as unit spikes of double or half height, respectively, and they are noted for editing. Individual 500-point records of the digitized ECG may be randomly accessed and displayed on the digital scope for editing purposes (Figure 32.5b), and arrows marking the R wave may be inserted, removed, or positioned within the record as appropriate. This editing feature is essential regardless of the R-wave detection technique (threshold crossing or pattern detection algorithm) due to the prevalence of many types of noise in the ECG even under the best of circumstances (Friesen et al. 1990).

Power Spectral Analysis

525 CLV01 (preictal, ictal, postictal)

0.8 0.75 0.7

R–R interval

0.65 E 0.6 0.55 0.5 0.45 0.4 0

0.2

0.4

0.6

0.8 1 (Thousands) R wave

1.2

1.4

1.6

1.8

Figure 32.6╇ Heart period plotted sequentially for the seizure in Figure 32.3 with the ictal

onset marked I and the end of the ictus marked E. These data represent the R–R interval for approximately 5 min preceding the ictus, 5 min of the ictus, and 5 min of the postictal period. Each period was divided into four records of 128 beats each, and the power spectra for these 16 records are presented in the three-dimensional plot in Figure 32.7.

The spike introduced by a single missed or falsely detected beat essentially contaminates all frequencies of the spectrum with erroneous power (Figure 32.2). 32.3.2.2â•… Power Spectra Estimation Once we are satisἀed that the heart period sequential plot accurately represents the R–R interval sequence of the ECG, the data are normalized by subtracting the mean of 128 consecutive R–R intervals from each heart period within that record and then dividing each by the mean. This data set is then multiplied by a modiἀed super-Gaussian window (P = 6, E = 0.0005), which tapers the beginning and end of the record to the mean. From the normalized meanzero data, spectra are generated, displayed, and printed for each sequential 128-beat record. We use linear scaling of the autospectra with the range of the vertical axis automatically matched to the data. Spectral records that are judged absent of signiἀcant nonstationarities, based on the time interval and spectral records, are averaged to form an estimate of the power spectrum. We consider an estimate determined from three to four records to represent a good compromise of the time required to maintain the subject in a constant environment, the nonstationarities that transiently appear in the ECG under these circumstances, and the need for averaging many records to obtain a reliable (consistent) estimate of the power spectrum.

526 Sudden Death in Epilepsy: Forensic and Clinical Issues

End

0.04 0.03 –0.00 0.01

Power

0.06

0.07

Ictus

0. 10 ch

0. 0.25

End Ictus

05 0. 0

0.01

0. 0

0.11 0.06 Epo ch

0. 0 Fr 9 eq 0. ue 14 nc y 0.1

8

–0.00

0.

23

0.01

Power 0.04 0.03

0.06

0.07

(a)

Ep o

06

0.20

0.

0.15

01

0 Fre .10 que ncy

15

0.05

0.

0.00

(b)

Figure 32.7╇ (a) Three-dimensional representation of the power spectra corresponding to the

heart period data in Figure 32.6. Power spectra for 16 records of 128 beats each, from the preictal through postictal periods, were smoothed across frequency and records to create this power spectra surface. (b) The electrographic ictal onset (marked ICTUS) followed the marked drop in HPV power by about 16 beats, and spectral power gradually recovered over approximately 20 beats after the electrographic end (marked END) of the seizure.

Power Spectral Analysis

527

32.3.2.3â•… Data Evaluation and Presentation Although the Digital 11/73 is an extraordinary computer for data acquisition and processing, there is a paucity of cost-effective and convenient software and hardware available for statistical analysis and generation of quality graphics. The R–R interval and power spectral data are therefore ported to an IBM-AT computer for statistical quantiἀcation, drawing of three-dimensional surfaces (Figure 32.7), and generation of publication-quality heart period and spectral plots (Figures 32.2 and 32.6). Spectral data are tested for signiἀcance (F test for equal variance within subjects, with nonparametric comparison of these results between subjects) in each of the three spectral bands that have been shown to characterize speciἀc autonomic activity. Our computer software system for collecting, processing, and evaluating HPV data is presented in detail sequence (Quint et al. 1989).

32.4â•… Preliminary Findings Both seizure activity and anticonvulsant therapy are postulated to be related to sudden and unexplained death in epilepsy through autonomic effects (Hirsch and Martin 1971; Jay and Leestma 1981; Leestma et al. 1984; Terrence et al. 1975). For this reason, we are interested in the effect of anticonvulsants and epileptogenesis on autonomic regulation. The interaction of anticonvulsant therapy and the basic mechanisms of epilepsy are almost certainly more complex than the effect of either alone on the autonomic nervous system. However, in studying their effects on autonomic regulation, we hope to understand the relationship of seizures, anticonvulsant therapy, and autonomic activity to sudden death. Through the noninvasive procedure of spectral analysis of HPV, autonomic activity may be observed indirectly through its relationship to the dynamic variability in the interval between successive heartbeats. These studies are in progress, and the data are presented as examples of application of the method rather than as evidence of a speciἀc effect of seizures or an anticonvulsant drug on autonomic activity. 32.4.1â•… Anticonvulsant Effects We have obtained preliminary HPV data from normal subjects receiving carbamazepine or phenytoin while participating in pharmacokinetic studies. Data from subjects (n = 6) receiving carbamazepine (600 mg orally four times daily) were collected at the predicted peak serum concentration. Ten minutes of ECG data were collected under controlled conditions with subjects supine and standing. Baseline measurements from each subject, obtained before the administration of the drug, were used as control data for each subject. Data from subjects receiving phenytoin (n = 12) were obtained from the supine position only. ECG data were collected at baseline, at the end of infusion of 250 mg of phenytoin, and again 1 h after the end of infusion. The baseline measurements for each subject were used as control data. Data from subjects receiving carbamazepine and phenytoin were examined by comparing the spectral data in each of the three major frequency bands between the control and drug condition.

528 Sudden Death in Epilepsy: Forensic and Clinical Issues

32.4.1.1â•… Carbamazepine Of the six subjects in the carbamazepine study, four had signiἀcant changes in one or more frequency bands. In the supine position, two subjects had signiἀcant decreases in power at all three frequency bands while one subject had a signiἀcant increase and one had a signiἀcant decrease limited to the high-frequency band (0.2–4.0 Hz). In the standing position, one subject had decreases at the low-frequency and mid-frequency bands while one subject had an increase at the mid-band and one had a decrease at the high-frequency band. Overall, signiἀcant changes were seen in 12 bands in four subjects. Of the 12 changes, 10 were decreases in power. 32.4.1.2â•… Phenytoin Data from the subjects receiving phenytoin were more variable. Signiἀcant changes occurred in 18 bands for 7 of the 12 subjects at the end of infusion and in 16 bands for 10 subjects 1 h after the end of infusion. Most of the changes were increases in power (14/18 at the end of infusion and 11/16 one hour after infusion). This is contrasted by reductions in power observed with carbamazepine. These results are preliminary and insufficient to draw meaningful conclusions. They do suggest, however, that phenytoin and carbamazepine have very different effects on HPV. Because the data for each drug were obtained under different circumstances, we cannot be certain that the differences in HPV are attributable to drugs alone. 32.4.2â•… Seizures In addition to the normal variation in HPV between subjects, there is a considerable difference in the effects on HPV exerted by seizures both within and between subjects. Because uninterrupted records are required for spectral analysis, it is generally difficult to follow HPV through the ictus with spectral techniques due to movement artifact associated with the seizure. Data are presented for three patients with partial complex or generalized seizures, in which the R–R interval was detectable in the ECG through the ictus, to illustrate the variety and similarities of the effects of seizures on HPV. Case 1 This patient had repeated partial complex seizures that were localized in onset to the right posterior temporal region. The seizures consisted of left head and eye deviation and hallucinations in a hemianopic left visual ἀeld. The patient was monitored with the CCTV/EEG system in the supine position. Figure 32.3 is a tracing of the ECG and selected channels of the EEG. The R–R interval sequential plot was determined for a period extending from approximately 5 min preictally, through the ictus (5 min), and for approximately 5 min postictally (Figure 32.6). Little variation was evident in the preseizure portion of the R–R interval plot, probably related to a combination of the patient’s postural position during recording (Pagani et al. 1986) and the cumulative effect of repeated seizures (Wannamaker 1985) or anticonvulsants (carbamazepine and phenytoin). As evident from this ἀgure, the mean heart period during the ictus did not differ from the preictal or postictal period. The power spectrum was determined for each successive 128-beat record from these data, in the manner described

Power Spectral Analysis

529

above, and these spectra are presented in a three-dimensional plot in Figure 32.7. During the ictal period, the power fell throughout the spectrum in this subject, with the low-frequency power loss most obvious because of its predominance in the preictal period. This effect actually preceded the onset of the ictus as detected in the EEG (Figure 32.3). Accompanying 3 of 10 seizures recorded in this patient were occasional sudden and brief heart rate changes in the postictal period (not seen in this record), which did not occur preictally and are generally not seen in normal subjects. These impulse rate changes had the appearance of those seen in the postictal period for case 3 (Figure 32.11).

Case 2 Partial complex seizures occurred approximately once per day in this patient, localized in the left temporal regions. Seizures were accompanied by aphasia and abdominal pain. Records of the ECG and EEG during seizure activity were obtained with home ambulatory monitoring. In one of these records, repeated instantaneous heart period changes were present in the ECG, as shown in the R–R interval sequential plot in Figure 32.8a. In this case (Figure 32.9), these events are characterized by a short R–R interval caused by a premature atrial contraction (PAC) followed by a compensatory long heart period, seen as a negative–positive impulse pair in Figure 32.8a. These impulses in the heart period record were suppressed during the ictus and gradually reappeared with increasing frequency after the seizure. Because there is a compensatory pause after each of these premature beats, it is possible to use the editing procedure to average each impulse pair and obtain a spectrum of the underlying HPV (Figure 8b). There is a slight shift in power from the low-frequency band preictally to the mid-frequency band postictally (Figure 32.10). Although HPV does appear to be reduced during the ictal period, the ictus in this case is too short for its spectrum to be determined with the present version of our software. The mean heart period is only slightly less during the seizure than in the preictal or postictal period. Because these heart period impulses were present at least an hour preceding this seizure and not present during most other seizures in this patient, we are uncertain if the PACs were related to this seizure. However, their suppression during the ictus and slow reappearance postictally clearly indicate some interaction between the seizure and the origination of the PACs.

Case 3 This 4-year-old patient experienced frequent generalized seizures, having developed a severe hypersensitivity reaction to phenobarbitol, phenytoin, and carbamazepine. As an interim measure before the start of valproic acid therapy, the child was maintained on a ketogenic diet and lorazepam. Eight to twelve generalized tonic–clonic

530 Sudden Death in Epilepsy: Forensic and Clinical Issues

seizures, 60–90 s in duration, were noted daily. ECG and EEG data were recorded in this patient, in a supine position, with CCTV monitoring from his hospital room. From the R–R interval plot, HPV is seen to gradually increase in the preictal period, greatly diminish in the ictal period, and become pronounced postictally (Figure 32.11). In the three-dimensional spectral plot of these data (Figure 32.12), this variability develops as increasing power in the low-frequency and mid-frequency bands in the preictal period, is reduced in these bands during the ictus, and is large in the low-frequency band postictally. A respiratory component (high-frequency band) also becomes prominent in the postictal period. The mean heart period during the ictus is considerably less than the preictal or postictal period in this case. Instantaneous rate changes appear in the R–R interval plot immediately after the seizure, followed by the appearance of large oscillations. These instantaneous rate changes are similar to those often present in the ictal and postictal periods for the patient in case 1 and differ from the negative/positive impulses in case 2 in that in case 3, they appear postictally and as single positive impulses only. These positive impulses are seen in the ECG to be isolated beats with extended time from the Q wave of one beat to the P wave of the next beat, not preceded by any type of premature beat. This sudden slowing suggests a strong parasympathetic influence, also indicated by the high-frequency peak. Although we did not monitor respiration, the large oscillation in the postictal period, which coincides with the very large low-frequency spectral peak (0.03 Hz) and the appearance of the respiratory peak, are probably due to periodic breathing with an envelope of 0.03 Hz.

32.5â•…Discussion The spectral changes in HPV that we have observed in relation to seizures are indicative of a signiἀcant change in cardiac autonomic regulation. Although recordings obtained during ictal episodes are often short or difficult to analyze because of movement noise and artifact, there appears to be a consistent decrease in power over the entire spectrum during ictal periods compared to either the preictal or postictal periods. At times, this power loss preceded the EEG ictal onset (Figure 32.7). This is in agreement with observations of Blumhardt et al. (1986) and Van Buren and Ajmone-Marsan (1960) that autonomic effects of seizures may be the very ἀrst manifestation of seizure activity. More variable changes observed in the postictal period include increases and deacreases in power in the different frequency bands compared to the preictal period. These variations in the postictal spectra were seen in analyzing repeated seizures within a single patient as well as between patients. This variability of seizure effects from patient to patient and even within the same patient is not surprising. Van Buren and Ajmone-Marsan (1960) documented the importance of the baseline autonomic state in determining the type and direction of changes in autonomic activity that accompanies a seizure. The consistently reduced power we observed in the ictal period was often associated with tachycardia. Keilson et al. (1987) and Blumhardt et al. (1986) reported sinus tachycardia as the most frequent accompaniment to ictal episodes. In addition to spectral and mean rate changes associated with seizures, we frequently observed precipitous or

Power Spectral Analysis

531

0.8

0.8

R–R (upper) 0.6 Time

0.6

0.4

0.4

Interval 0.2 (s) 0

0.2 0.540 1

0.541

0.506

0.502

0.501

0.507

0.548

2

3

4

5

6

7

1.154

0.8

2.308 Time (min)

3.388

4.460 5.529 (128 points per division)

6.611

0.554 7.781

8

0 8.965 0.8

R–R (upper) 0.6 Time

0.6

0.4

0.4

Interval 0.2 (s) 0

0.2 0.540 9

0.541

0.506

0.502

0.501

0.507

0.548

10

11

12

13

14

15

10.11

11.32

Time (min)

12.46

13.58

14.81

(continued from above)

16.10

0.554 17.40

16

0 18.72

HPav = 0.548 s

(a)

Figure 32.8╇ (a) Heart period plotted sequentially from approximately 9 min preictally through

9 min of the postictus (see case 2, Section 32.4.2), organized into 16 records of 128 beats each. Characterized by a short R–R interval, PACs appear as a downward impulse followed by an upward impulse (compensatory pause), except in the short ictal period (electrographic ictal onset approximately at minute 9 and ending at minute 9.7 in the record).

instantaneous heart slowing, rarely occurring in normal subjects, in the late ictal or postictal period. These episodes of transient heart slowing were associated with extended R–R and normal QT intervals in an otherwise unremarkable ECG. Keilson et al. (1987) and Blumhardt et al. (1986) similarly observed sudden transient changes in heart rate in some patients in both the ictal and postictal periods, not associated with cardiac arrhythmias. These short periods of bradycardia differed from those reported by Mameli et al. (1988) in an experimental epilepsy model using decerebrate rats, in which brief periods of heart slowing during ictal and interictal episodes were frequently observed in conjunction with cardiac arrhythmias. When a sequential plot of heart period is generated, with heart period deἀned as the time separating every systole whether of sinus node, atrial, or ventricular origin, there is a characteristic difference between records containing arrhythmias and those without. Extrasystoles, resulting from either PACs or PVCs, are generally followed by a compensatory pause. This pause is such that the total time from the beat preceding the extrasystole to the following beat is equivalent to two normal heart periods. In the heart period plot, the arrhythmia appears as a negative impulse followed immediately by a positive

532 Sudden Death in Epilepsy: Forensic and Clinical Issues 0.8 R–R (upper)

0.8

0.6 Time

0.6

0.4

0.4

Interval 0.2 (s) 0

0.2 0.540 1

0.541

0.506

0.502

0.501

0.507

0.548

2

3

4

5

6

7

1.154

0.8 R–R (upper)

2.308 Time (min)

3.388

4.460 5.529 (128 points per division)

6.611

0.554 7.781

8

0 8.965 0.8

0.6 Time

0.6

0.4

0.4

Interval 0.2 (s) 0

0.2 0.539 9

0.567

0.534

0.525

0.575

0.603

0.609

10

11

12

13

14

15

10.11

11.32

Time (min)

12.46

13.58

14.81

(continued from above)

16.10

0.619 17.40

16

0 18.72

HPav = 0.548 s

(b)

Figure 32.8 (Continued)╇ (b) Same heart period plot with negative/positive impulses due to

PACs averaged in preictal records 1, 3, and 4 and in postictal records 14 and 15. Because these PACs occur as nonphasic events (unrelated to rate regulation), they may be averaged to reveal the underlying rhythmicity of heart rate. The power spectra may then be determined without introducing broad-spectrum artifacts due to impulses in the data (see Figure 32.4).

(a)

(b)

Figure 32.9╇ Digitized ECG showing normal rhythm and a PAC from the preictal period in

Figure 32.8a (see case 2, Section 32.4.2). (a) Two QRS complexes from a normal section of ECG with the arrows marking the R waves separated by 568 ms. (b) Two QRS complexes with the P wave of the second beat occurring on the T wave of the first. The arrows are separated by 342 ms corresponding to the R–R interval marked by the arrow in (a).

Power Spectral Analysis

533

0.060

record #’s 1, 3, 4

Heart 0.045 Interval 0.030 Variation 0.015 (R–R) (power) 0 HPav = 0.514

0.2 0.4 0.6 Frequency of fluctuations

0.8 (Hz)

(a) record #’s 15, 16 0.04 Heart 0.03 Interval 0.02 Variation 0.01 (R–R) (power) 0 HPav = 0.614

0.2 0.4 0.6 Frequency of fluctuations (Hz)

0.8

(b)

Figure 32.10╇ Power spectra averages for heart period data from selected records in Figure

32.9b, with PACs removed: (a) preictal records 1, 3, and 4; (b) postictal records 15 and 16. From the preictal to the postictal periods, there is a modest reduction in power in the low-frequency band and a corresponding increase in power in the mid-frequency band. The ictal period was too short, in this case, to determine its spectral content.

impulse (Figure 32.8a), such that when this impulse pair is replaced by its average there is no evidence of a transient change in the heart period data (Figure 32.8b, records 1, 3, 4, 14, and 15). In records containing impulse (one beat) or transient (several beats) changes in heart period without the presence of arrhythmias (Figure 32.11), there is a fundamentally different process involved, and these cannot be removed from the record by averaging. Arrhythmias characterized by premature beats followed by a compensatory pause are not associated with a change in the autonomic influence to the sinus node, as neither mean rate nor HPV is altered. This is a nonphasic event, and the system that regulates rate is not involved in the extrasystole. This is in contrast to profound transient rate changes of the type seen in Figure 32.11, which are a consequence of the rate-regulating system and must be effected through a transient change in autonomic influence on the sinus node. This is

534 Sudden Death in Epilepsy: Forensic and Clinical Issues

1.00 R–R (upper)

1.00

0.75 Time

0.75

0.50

0.50

Interval 0.25 (s) 0

0.25 0.553 1

0.557

0.571

0.581

2

3

4

1.180

1.00 R–R (upper)

2.370 Time (min)

0.576

0.597

0.581

3.590 4.829 5 6.060 6 (128 points per division)

7

7.334

0.531 8.575

0 8 9.709 1.00

0.75

0.75

Time 0.50

0.50

Interval 0.25

0.25 (s) 0

0.359 9

0.341

0.450

0.550

0.595

0.557

0.538

10

11

12

13

14

15

10.47

11.20

Time (min)

12.16

13.33

14.60

(continued from above)

15.79

0.545 16.94

16

0 18.10

HPav = 0.530 s

Figure 32.11╇ Heart period plot from an ECG preceding, during, and after an ictal episode (see case 3, Section 32.4.2), organized as 16 consecutive records of 128 R–R intervals per record. The ictal onset occurs in record 8 with the precipitous fall in heart period, and the postictal period begins in record 11 with a gradual return of heart period to the preictal level. The highfrequency fluctuations in heart period in the ictal period (records 9 and 10) are due to error in precisely identifying the R wave in the ECG. Positive impulses, which appear in the early postictal period and correspond to brief profound heart slowing, represent phasic events in heart rate regulation.

a phasic event, equivalent to a phase shift in heart period with respect to other dynamic physiologic parameters. Lathers and Schraeder (1982), Schraeder and Lathers (1983), and Lathers et al. (1987) demonstrated an imbalance in sympathetic and parasympathetic activities in experimental epilepsy, both during ictal activity and with interictal spike activity. This autonomic imbalance was characterized by alterations in heart rate and mean arterial pressure with a disruption in the physiological relationship between these variables. Cardiac arrhythmias (both PACs and PVCs) and other ECG effects were also observed in relation to nonuniform cardiac sympathetic neural discharge associated with epileptogenic activity. Disruption in the physiological relation between mean arterial blood pressure and heart rate, along with ECG changes, also occurred with the lockstep phenomenon (synchronization between cortical, cardiac sympathetic, and to a lesser extent vagal discharges; Lathers et al. 1987). The general loss of power in the heart period spectrum with varying increases in heart rate that we have observed during seizures is in agreement with nonuniform autonomic discharge

Power Spectral Analysis

535

–0.00

0.04

Power 0.15 0.10

0.21

S300

29

1.30

0.

1.04

06

0.5 Freq 2 0 .78 uen cy

0 Ep .53 oc 0 h .7 6

0.26

0.

0.00

Figure 32.12╇ Three-dimensional surface plot from the 16 heart period records in Figure 32.11. The transition from the preictal to ictal period is represented by an average of the autospectra of records 7 and 9, and the transition from ictal to postictal period is represented by averaging the autospectra for records 10 and 12. Smoothing was performed across frequency and records (epochs) to generate this surface. The broad-spectrum low-amplitude power in the ictal period of this plot is artifact due to random error in locating the R wave in the ECG during seizures.

or increased activity with loss of rhythmicity in the cardiac sympathetic discharge. The most common ECG effect associated with seizures that we have observed, profound brief heart slowing in the late ictal and postictal periods, may occur through a mechanism similar to the lockstep phenomenon described by Lathers et al. (1987), although we have not seen EEG discharges in synchrony or arrhythmias with these remarkable transient events. These transient events could also result from short periods of nonuniform (dysfunctional) cardiac sympathetic activity during which the sinus node is dominated by parasympathetic input. It is important to recognize that not all changes in HPV that appear to be related to seizure activity can be detected or quantiἀed by spectral analysis. In many instances, and particularly in association with epilepsy, records of the R–R interval contain profound transient changes in heart period as well as shifts in the mean. It is often possible to remove these nonstationarities from the heart period record by ἀltering or other procedures and obtain the spectral distribution of the underlying variability, but important information may be discarded in the process. Transients in the record, especially in epilepsy, may be indicative of dysfunctional autonomic activity and should be analyzed in the time domain. To fully characterize and quantify changes in HPV in epilepsy, both frequency and time domain procedures should be used. We are currently evaluating several nonspectral techniques for use in quantifying transient features in heart period data.

536 Sudden Death in Epilepsy: Forensic and Clinical Issues

References Akselrod, S., D. Gordon, F. A. Ubel, D. C. Shannon, D. C. Barger, and R. J. Cohen. 1981. Power spectrum analysis of heart rate fluctuation: A quantitative probe of beat-to-beat cardiovascular control. Science 213: 220–222. Baselli, G., S. Cerutti, S. Civardi, F. Lombardi, A. Malliani, M. Merri, M. Pagini, and G. Rizzo. 1987. Heart rate variability signal processing: A quantitative approach as an aid to diagnosis in cardiovascular pathologies. Int J Bio-Med Comput 20: 51–70. Bendat, J. S., and A. G. Piersol. 1986. Random Data: Analysis and Measurement Procedures. New York, NY: Wiley. Berger, R. D., A. Akselrod, D. Gordon, and R. J. Cohen. 1986. An efficient algorithm for spectral analysis of heart rate variability. IEEE Trans Biomed Eng 33: 900–904. Blumhardt, L. D., P. E. M. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1: 1051–1056. DeBoer, R. W., J. M. Karemaker, and J. Strackee. 1984. Comparing spectra of a series of point events particularly for heart rate variability data. IEEE Trans Biomed Eng 31: 384–387. DeSilva, R., and J. Lown. 1978. Ventricular premature beats, stress and sudden death. Psychosomatics 19: 649–661. Eckberg, D. L. 1983. Human sinus arrhythmia as an index of vagal cardiac outflow. J Appl Physiol 54: 961–966. Falconer, B., and J. Rajs. 1976. Postmortem ἀndings of cardiac lesions in epileptics: A preliminary report. Forensic Sci 8: 63–71. Friesen, G. M., T. C. Jannett, M. Aἀfy, S. Yaltes, S. R. Quint, and H. T. Nagel. 1990. A comparison of the noise sensitivity of nine QRS detection algorithms. IEEE Trans Biomed Eng 37: 85–98. Gordon, D., R. J. Cohen, D. Kelly, S. Akselrod, and D. C. Shannon. 1984. Sudden infant death syndrome: Abnormalities in short term fluctuations in heart rate and respiratory activity. Pediatr Res 18: 921–926. Gordon, D., D. P. Southall, D. H. Kelly, A. Wilson, S. Akselrod, J. Richards, B. Kenet, R. Kenet, R. J. Cohen, and D. C. Shannon. 1986. Analysis of heart rate and respiratory patterns in sudden infant death syndrome victims and control infants. Pediatr Res 10: 680–684. Guilleminault, G., P. Pool, J. Motta, and A. M. Gillis. 1984. Sinus arrest during REM sleep in young adults. N Engl J Med 306: 1006–1010. Haddad, G. G., H. J. Jeng, S. H. Lee, and T. L. Lai. 1984. Rhythmic variations in R–R interval during sleep and wakefulness in puppies and dogs. Am J Physiol 247: H67–W3. Hirsch, C. S., and L. M. Martin. 1971. Unexpected death in young epileptics. Neurology 21: 682–690. Jay, W. J., and J. E. Leestma. 1981. Sudden death in epilepsy. Acta Neurol Scand 63 (Suppl. 82): 1–66. Keilson, M. J., W. A. Hauser, J. P. Magrill, and M. Goldman. 1987. ECG abnormalities in patients with epilepsy. Neurology 37: 1624–1626. Kiok, M. C., C. F. Terrence, G. H. Fromm, and S. Lavine. 1986. Sinus arrest in epilepsy. Neurology 36: 115–116. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Mameli, P., O. Mameli, E. Tolu, G. Padua, D. Giraudi, M. A. Caria, and F. Melis. 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29: 74–82.

Power Spectral Analysis

537

Marshall, D. W., B. F. Westmoreland, and F. W. Sharbrough. 1983. Ictal tachycardia during temporal lobe seizure. Mayo Clin Proc 58: 443–446. McCabe, P. M., B. G. Yongue, S. W. Porges, and P. K. Ackles. 1984. Changes in heart period, heart period variability, and a spectral analysis estimate of respiratory sinus arrhythmias during aortic nerve stimulation in rabbits. Psychophysiology 21: 149–158. Messenheimer, J. A., S. R. Quint, M. B. Tennison, M. C. Corcoran, and P. Keaney. 1987. Changes in heart period variability during repeated complex partial seizures. Epilepsia 28: 635. Mohn, R. K. 1976. Suggestions for the harmonic analysis of point process data. Comp Biomed Res 9: 521–530. Myers, G. A., G. J. Martin, N. M. Magid, P. S. Barnett, J. W. Schadd, J. S. Weiss, M. Lesch, and D.€H. Singer. 1986. Power spectral analysis of heart rate variability in sudden cardiac death: Comparison to other methods. IEEE Trans Biomed Eng 33: 1149–1156. Neuspiel, D. R., and L. H. Kuller. 1985. Sudden and unexpected natural death in childhood and adolescence. J Am Med Assoc 254: 1321–1325. Nugent, S. T., and J. P. Finley. 1983. Spectral analysis of periodic and normal breathing in infants. IEEE Trans Biomed Eng 30: 672–675. Oppenheim, A. V., and R. W. Schafer. 1975. Digital Signal Processing. Englewood Cliffs, NJ: Prentice Hall. Pagani, M., F. Lombardi, S. Guzzetti, O. Rimoldi, R. Furlan, P. Pizzinelli, G. Sandrone, et al. 1986. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympathovagal interaction in man and conscious dog. Circ Res 59: 178–193. Pomeranz, B., R. J. Macaulay, M. A. Caudill, I. Kutz, D. Adam, D. Gordon, K. Kilborn, et al. 1985. Assessment of autonomic functions in humans by heart rate spectral analysis. Am J Physiol 248: H151–H153. Porges, S. W. 1986. Respiratory sinus arrhythmia: Physiological basis, quantitative methods, and clinical implications. In Cardiorespiratory and Cardio-somatic Psychophysiology, ed. P. Grossman, K. Janssen, and D. Vaitl, 101–115. New York, NY: Plenum. Quint, S. R., J. A. Messenheimer, M. B. Tennison, and M. C. Corcoran. 1987. A procedure for assessing autonomic activity related to risk factors for sudden and unexplained death in epileptics. Epilepsia 25: 610. Quint, S. R., J. A. Messenheimer, M. B. Tennison, and H. T. Nagel. 1989. Assessing autonomic activity from the EKG related to seizure onset detection and localization. Proceedings of the Second IEEE Symposium on Computer-Based Medical Systems, 2–9. Sayers, B. 1980. Signal analysis of heart-rate variability. In The Study of Heart Rate Variability, ed. R.€I.€Kitney and O. Rompelman, 27–58. Oxford: Clarendon Press. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Tennison, M. B., S. R. Quint, J. A. Messenheimer, M. C. Corcoran, and P. Keaney. 1987. Autonomic effects of seizures assessed by power spectral analysis of heart period variability. Epilepsia 28: 612. Terrence, C. F., H. M. Wisotzkey, and J. A. Perper. 1975. Unexpected, unexplained death in epileptic patients. Neurology 25: 594–598. Van Buren, J. M., and C. Ajmone-Marsan. 1960. A correlation of autonomic and EEG components in temporal lobe epilepsy. Arch Neurol 3: 683–703. Wannamaker, B. B. 1985. Autonomic nervous system and epilepsy. Epilepsia 26 (Suppl. 1): 31–39. Zwiener, U. 1978. Spectral analyses of blood pressure, heart rate and respiration rhythms in different postures of healthy individuals and patients with neurovegetative disorders. Acta Biol Med Ger 57: 1461–1469.

Animal Model for Sudden Cardiac Death Sympathetic Innervation and Myocardial BetaReceptor Densities

33

Claire M. Lathers Robert M. Levin

Sudden unexpected death in persons with epilepsy had been discussed by Leestma et al. (1984) and more recently has been evaluated in terms of its risk factors, impact, mechanisms, and prevention (Lathers and Schraeder 1987, 1990, 2002; Lathers et al. 2008a, 2008b; Scorza et al. 2008; Jehi and Najm 2008). One animal model for sudden death reported nonuniform sympathetic neural discharge recordings that were hypothesized by Lathers et al. (Lathers 1980, 1981; Lathers et al. 1977a, 1977b, 1978) to be contributing to the development of arrhythmia and/or sudden death via nonuniform recovery of excitability in ventricular muscle (Han and Moe 1964). Sympathetic innervation and its role in initiation of abnormal cardiac function, that is, arrhythmias and/or sudden death, are now recognized. Discussed herein is an explanation for the mechanism of how the nonuniform sympathetic neural discharge produces cardiac arrhythmias and death. The regional difference in the beta-adrenoceptor densities reflects differences in postganglionic cardiac sympathetic innervation of the myocardium, and these site differences will vary the release of norepinephrine in the various sites of the heart. This modiἀes cardiac contractile function and may trigger cardiac arrhythmia and/or sudden death. Site difference, in part, is one component of mechanism(s) involved in sudden cardiac death. This animal technique needs to be explored by application to studies analyzing the mechanisms that contribute to sudden death in persons with epilepsy (SUDEP). Application of this animal model, with the techniques described below, to studies of epileptogenic activity, arrhythmias, and sudden death using the physiology and pharmacology based on the nonuniform cardiac sympathetic neural discharge animal model will result in the development of new animal models to apply molecular and genetic techniques that further explore mechanisms involved in SUDEP. Use of this animal model will allow studies of new antiepileptic agents on catecholamine effects associated with arrhythmias and interictal and€ictal activity and will determine whether the new antiepileptics alter cardiac beta-receptor density, the initiation of cardiac arrhythmias, and/or the occurrence of sudden death. Circulating catecholamines contribute to the initiation of arrhythmias (Ceremuzynski et al. 1969; Lathers et al. 1977a, 1977b, 1978, 1988a, 1989). Patients suffering from intractable seizures may harbor a risk of postictal catecholamine surge and catecholamine-induced myocardial dysfunction (Shimizu et al. 2008). Experiments should be designed to measure circulating catecholamines. The contribution of adrenal medullary catecholamines could be 539

540 Sudden Death in Epilepsy: Forensic and Clinical Issues

studied with and without bilateral adrenal ligation to prevent catecholamine release of catecholamines from the adrenal gland (Lathers et al. 1986a). Importantly, this animal model may be used to examine the antiepileptic effects of current and future beta-blocking drugs. In a series of animal experiments designed to explore sympathetic innervation and beta-receptor density in association with sudden cardiac death, Lathers and others (Lathers et al. 1986b, 1988b, 1990) examined sympathetic beta-receptor density gradient as a measure of cardiac sympathetic innervation in the cat heart. Beta-receptor density was determined by binding of 3H-dihydroalprenolol (Lathers et al. 1986b). The distribution is presented in Figure 33.1. The beta-receptor density of the right atria was signiἀcantly lower than in the left atria; the beta-receptor density of the right ventricle was signiἀcantly lower than the density of the left ventricle. The density of the receptors of the ventricles was higher than those of the atria. The beta-receptor density of the distal distribution of the left anterior descending coronary artery was signiἀcantly higher than that of the proximal distribution. These regional differences in beta-adrenoceptor densities are related to cardiac contractile strength of the different areas of the heart. Acute, abrupt, and irreversible coronary occlusion of the left anterior descending coronary artery did not signiἀcantly alter the beta-adrenergic receptor density and resulted in death with a mean time of 5.8 ± 3.6 min. The effect of timolol (5 mg/kg, PO, for 1, 2, or 8 weeks before acute, abrupt, and irreversible coronary occlusion) on the response of the cat heart to coronary occlusion was

200

2

Beta adrenoreceptor density (fmol/mg protein)

175

4

150 125 100

3

75 50

1

25 0

Left atria

Right Left Right Septum LADCA ALDCA atria ventricle ventricle

Figure 33.1╇ Regional distribution of myocardial beta-adrenoceptors in the cat heart. Each

bar is the mean of five individual samples from normal cats. 1 = significantly different from left atria; 2 = significantly different from left atria and right ventricle; 3 = significantly different from right atria and left ventricle; 4 = significantly different from LADCA; P < 0.05. LADCA, proximal distribution of the left anterior descending coronary artery; ALDCA, distal distribution of the left anterior descending coronary artery. (Adapted from Lathers et al., Eur J Pharmacol, 130 (1–2), 111–117, 1986.)

Animal Model for Sudden Cardiac Death

541

determined for the following parameters: postganglionic cardiac sympathetic neural discharge, blood pressure, heart rate, and beta-receptor density (Lathers et al. 1988a). After 1 week of timolol treatment, there were no signiἀcant changes in the betaadrenergic receptor density in any area. After 2 weeks, the densities of beta-receptors increased€ in€ the left atria, left ventricle, and septum. After 8 weeks of treatment, the receptor densities in all areas increased (Figure 33.2). Spearman rank correlation coefficients between dose and beta-receptor density revealed an increase (P < 0.05) for all heart areas. Heart rate did not vary before timolol and was decreased after all doses of timolol. Timolol increased the mean times to coronary occlusion-induced death. Timolol did not prevent postganglionic cardiac sympathetic neural discharge associated with arrhythmia. Timolol may increase beta-receptor density and decrease synaptic norepinephrine, causing a decreased release per cardiac sympathetic nerve impulse. Alternatively, molecules of timolol may accumulate in nerve endings and be released in greater concentrations at the receptors. This could explain the protection against coronary occlusion-induced arrhythmia and death (Lathers et al. 1988b). In a second study of timolol (5 mg/kg, PO, b.i.d., 7 or 14 days), the effects of coronary occlusion after 14 days of timolol pretreatment were determined (Lathers et al. 1986c) (Figure 33.3). In general, similar to the previous study, timolol pretreatment resulted in an increase in the beta-adrenergic receptor density of all areas of the heart; and coronary occlusion had no signiἀcant effect on beta-receptor density or on the response to timolol. In addition, similar to the previous studies, there were the same distribution differences among the areas of the heart. Again, the importance of the gradation of beta-receptors with increasing density from base to apex appeared to be its relationship with cardiac contractile function. Timolol decreased heart rate and blood pressure before occlusion. The mean times to arrhythmia and death were not signiἀcantly increased by any dosing regimen of timolol, although there was a trend for an increase in the time to death after 1 week of timolol pretreatment.

Beta adrenoreceptor density (fmol/mg protein)

300 Control 8 Weeks treatment

250

*

200 *

150 100

* *

50 0

*

Left atria

Right atria

Left ventricle

Right ventricle

Septum

Figure 33.2╇ Effect of 8 weeks of timolol treatment on beta-receptor density. Each bar is the mean ± SEM of four to five individual cats. *Significantly different from control (P < 0.05). (Adapted from Lathers et al., J Clin Pharmacol, 28 (8), 736–745, 1988.)

542 Sudden Death in Epilepsy: Forensic and Clinical Issues

Beta adrenoreceptor density (nontreated = 100%)

250 200

(a)

Left atria Right atria

*

*

150

*

*

100 50 0

Control

Occluded

Control timolol

Occluded timolol

Beta adrenoreceptor density (nontreated = 100%)

250 200

Left ventricle Right ventricle Septum

(b) *

* *

150

*

100 50 0

Control

Occluded

Control timolol

Occluded timolol

Figure 33.3╇ (a) Left and right atria. Each bar is the mean of four or five individual cats.

*Significantly different from untreated (P < 0.05). For comparative purposes, the data were normalized to nontreated = 100%. (b) Left and right ventricles and septum. Each bar is the mean of four or five individual cats. *Significantly different from untreated (P < 0.05). (Adapted from Lathers et al., Life Sci, 39 (22), 2121–2141, 1986.)

In addition, there was an increase in the time to arrhythmia and death after 14 days of pretreatment. When compared with data obtained in saline cats, chronic timolol treatment produced minimal changes in postganglionic cardiac sympathetic neural discharge. Timolol given chronically (PO) or acutely (5 mg/kg, IV, given 15 min before occlusion) also did not prevent the cardiac sympathetic discharge associated with the development of arrhythmia. The time to arrhythmia and death in the acutely treated cats was increased, but not signiἀcantly. Because cardiac sympathetic neural discharge increased as blood pressure dropped in the control period but did not increase after occlusion in the timolol-treated animals, the combination of timolol and occlusion may have modiἀed neural discharge via an action on the baroreceptor mechanism. That chronic administration of timolol produces an effect not present in cats in which only occlusion was done

Animal Model for Sudden Cardiac Death

543

is supported by the observation that chronic treatment produced an occlusion-induced decrease in beta-adrenergic receptor density. In a different animal model, the site of occlusion of the left anterior descending coronary artery was placed not at the origin but lower to decrease the extent of myocardial damage, increase survival time, and enable the study of new parameters, including norepinephrine and epinephrine concentrations post myocardial infarction. Cats anesthetized with alpha-chloralose received saline or dilevalol (1 mg/kg, IV) during a 10-min infusion (Lathers et al. 1990). Fifteen minutes later, the left anterior descending coronary artery was subjected to coronary occlusion 2 mm below its origin. Dilevalol prevented the increase in norepinephrine and epinephrine levels associated with arrhythmia and death. Postictal catecholamine surge and catecholamine-induced myocardial dysfunction have been reported to occur in persons with intractable seizure activity (Shimizu et al. 2008). It will be of interest in future studies to determine if dilevalol offers a protective effect in animal models for seizure by preventing or decreasing circulating catecholamines (Table 33.1). Numerous preclinical animal studies have been conducted as a model for cardiac sudden death and have led to use of this model, with addition of EEG monitoring, to develop an animal model mimicking the sudden death phenomena occurring in persons with epilepsy. The animal studies show that sympathetic nerve stimulation, arrhythmic doses of ouabain, or coronary occlusion increased temporal dispersion of recovery of ventricular Table 33.1â•… Effect of Dilevalol on the Response to Coronary Occlusion (CO) Experimental Group

Parameter

No coronary occlusion and saline or dilevalol

Beta-receptor densities

Coronary occlusion and saline or dilevalol Treated with only saline

Beta-receptor densities Mean times to arrhythmia and death Plasma norepinephrine and epinephrine

Treated with dilevalol

Mean times to arrhythmia and death Blood pressure, heart rate, prior CO BP and HR after CO Postganglionic cardiac sympathetic neural discharge (PCSND) immediately before CO PCSND immediately before arrhythmia Plasma norepinephrine and epinephrine

Response Regional differences in beta-receptor densities found for atria and ventricles and for areas within left ventricle Variation in beta-receptor density distribution is related to functional contractile differences CO did not modify the myocardial betareceptor density or regional distribution Arrhythmia: 5.8 ± 3.6 min (n = 3) Death: 5.4 (n = 2); three cats sacriἀced 6 h post-CO Increased ἀrst 5 min after CO; increase associated with arrhythmia. Values from 15 to 360 min post-CO did not differ from control Arrhythmia: 2.2 ± 0.8 (n = 3) Death: 75.9 ± 70.7 min (n = 4); one cat sacriἀced 6 h post-CO Both decreased Both decreased Decreased by 95% Signiἀcantly increased to 121% 3 min post-CO Prevented increases associated with arrhythmia and death

Source: Lathers, C. M., et al., J Clin Pharmacol, 30 (3), 241–253, 1990.

544 Sudden Death in Epilepsy: Forensic and Clinical Issues

excitability and led to an underlying electrical instability that predisposed the ventricular myocardium to arrhythmia and/or sudden death. Cardiac arrhythmias in an animal model for ouabain-induced toxicity were associated with neural autonomic dysfunction (Lathers et al. 1977a, 1977b, 1978). Neural discharges were characterized by increases, decreases, or no change in the discharge of postganglionic cardiac sympathetic nerves monitored simultaneously, predisposing to cardiac arrhythmia and sudden death. Stimulation of the sympathetic ventrolateral cardiac nerve produced a shift in the origin of the pacemaker and tachyarrhythmias because the nerve is not uniformly distributed to the various regions of the heart but is localized to the atrioventricular junctional and ventricular regions (Randall 1977, 1984). Such nonuniform distribution of sympathetic nerves also contributes to initiation of arrhythmia as a nonuniform neural discharge occurs. Sympathetic innervation and its role in normal and abnormal cardiac function require further investigation (Lathers et al. 1977a, 1977b, 1978). Application of this animal model to studies of epileptogenic activity and sudden death examined the physiology and pharmacology of the role in sympathetic innervation associated with autonomic cardiac neural nonuniform discharge, cardiac arrhythmias, and sudden death (Lathers and Schraeder 1982; Schraeder and Lathers 1983, 1995; Lathers et al. 1986a, 1988c, 1989). Use of this animal model has allowed questions to be raised such as whether disease states alter the function of this neural discharge and if the sympathetic postganglionic neural discharge represents one site of action for pharmacological agents to act by preventing the nonuniform neural discharge (Lathers et al. 1977a, 1977b, 1978; Lathers and Schraeder 1982; Schraeder and Lathers 1983). These questions have been answered just recently. A postmortem study of postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy has been conducted (Druschky et al. 2001). The study found sympathetic dysfunction in the form of altered postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy and suggested the altered postganglionic cardiac sympathetic innervation may increase risk of cardiac abnormalities and/or SUDEP. The exact role of innervation in arrhythmogenesis and developmental and regulatory mechanisms determining density and pattern of cardiac sympathetic innervation are still unclear. This clinical study of Druschky et al. (2001), conducted in humans, conἀrms the results and conclusions of the animal studies conducted by Lathers and Schraeder (Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1980), in which the relationship of postganglionic cardiac sympathetic neural discharge was associated with arrhythmias and/or sudden death. Thus, the nonuniform sympathetic cardiac neural discharge recordings were hypothesized by Lathers et al. (1977a, 1977b, 1978) to be contributing to the development of arrhythmia and/or sudden death via nonuniform recovery of excitability in ventricuÂ� lar€ muscle after coronary occlusion (Han and Moe 1964). Recent studies report that cardiac-speciἀc overexpression of Sema in transgenic mice (SemaTG) is associated with reduced sympathetic innervation and attenuation of epicardial-to-endocardial innervation gradient. SemaTG mice demonstrated sudden death and susceptibility to ventricular tachycardia due to catecholamine supersensitivity and prolongation of the action potential duration. In addition to the fact that the Sema3a(-I-) mice lacked a cardiac sympathetic innervation gradient, they also had stellate ganglia malformation associated with a marked sinus bradycardia due to the sympathetic dysfunction. The authors conclude that appropriate cardiac Sema3a expression is needed for sympathetic innervation patterning and is critical for heart rate control (Ieda et al. 2007). It is relevant to note here that

Animal Model for Sudden Cardiac Death

545

the sympathetic cardiac stellate ganglia itself is a potential site for action when designing drugs to minimize the risk of SUDEP and/or to prevent the development of speciἀc patterns of autonomic nerve discharges and cardiac arrhythmia (Ogawa et al. 2007; Alkadhi et al. 2010). The development and regulation of the cardiac sensory nervous system, measured by myocardial ischemia-induced c-Fos expression in dorsal root ganglia, are dependent on nerve growth factor synthesized in the heart. Nerve growth factor is critical for cardiac sensory innervation and rescues neuropathy in diabetic hearts (Ieda et al. 2006). Further studies are needed to determine the role of interventions in the function of the dorsal root ganglia and abnormal cardiac function via control of the autonomic nervous system. Recently, Ieda et al. (2008), as in the earlier studies conducted by Lathers et al. (Lathers and Schraeder 1982; Lathers et al. 1977a, 1977b, 1978), have raised the question of whether regulation of cardiac nerves is a “new paradigm” in the management of sudden cardiac death because the heart is extensively innervated and its performance is regulated by the autonomic nervous system. Innervation density is high in the subepicardium and the central conduction system. In diseased hearts, cardiac innervation density varies. This may lead to sudden cardiac death. After myocardial infarction, sympathetic denervation is followed by reinnervation within the heart, leading to unbalanced neural activation and lethal arrhythmia (Lathers et al. 1986b, 1988c, 1990). In the case of diabetic sensory neuropathy, silent mycocardial ischemia may occur, associated with loss of pain perception during myocardial ischemia, a major cause of sudden cardiac death in diabetes mellitus (Ieda et al. 2008). To date, molecular mechanisms underlying innervation density are not well understood. Ieda et al. (2008) have demonstrated that cardiac sympathetic innervation is determined by the balance of neural chemoattraction and chemorepulsion, both of which occur in the heart. Nerve growth factor, a potent chemoattractant, is synthesized by cardiomyoctyes and is induced by endothelin-l upregulation in the heart. In contrast, Sema3a, a neural chemorepellant, is expressed strongly in the trabecular layer in earlystage embryos and, at a lower level after birth, causes epicardial-to-endocardial transmural sympathetic innervation pattern. Cardiac nerve growth factor downregulation is a cause of diabetic neuropathy and nerve growth factor supplementation rescues silent myocardial ischemia in diabetic neuropathy. Both Sema3a-deἀcient and Sema3a-overexpressing mice showed sudden death or lethal arrhythmias due to disruption of innervation patterning (Ieda et al. 2008). All of these regulatory mechanisms involved in neural development in the heart and their critical roles in cardiac performance need to be examined to determine relevance to methods for future use to decrease the risk of SUDEP. Another aspect to look at is sexual differences in both innervation and receptor densities and sensitivities. Lujan and Dicarlo (2008) demonstrated clear differences in both sensitivity to induced arrhythmias and in the potency of beta-adrenergic blockade. Control female mice were less sensitive to coronary occlusion-induced arrhythmias and less sensitive to beta-adrenergic blockade than control male mice. These differences disappeared when the females were ovariectomized, thus showing the effects of estrogen on the cardiovascular system and beta-adrenergic receptors (Lujan and Dicarlo 2008; Lujan et al. 2007). Exercise is an additional factor in the relationship between cardiac arrhythmias and beta-receptors. In an experiment by Billman et al. (2006), dogs that received daily exercise were compared to sedentary dogs in their response to induced ventricular ἀbrillation and beta-adrenergic sensitivity to stimulation. These studies showed that exercise signiἀcantly

546 Sudden Death in Epilepsy: Forensic and Clinical Issues

reduced the cardiac response to isoproterenol and prevented beta stimulation from inducing ventricular ἀbrillation. There seems to be a direct correlation between beta-adrenergic receptor sensitivity and the level of arrhythmia and mortality after a coronary occlusion. The greater the sensitivity to beta-adrenergic agonists, whether from hormones, exercise, or genetics, the greater the level of arrhythmia and mortality (Lujan et al. 2007; Billman et al. 1997, 2006; Houle et al. 2001; Du et al. 2000). Target organ-derived neurotrophic factors regulate neuronal function and innervation density (Kimura et al. 2007). The expression of growth factors of endothelin-l, angiotensin II, and leukemia inhibitory factor is altered by cardiac hypertrophy. Cardiac hypertrophy is an independent predictor of cardiovascular morbidity and mortality, predisposing patients to heart failure, QT interval prolongation, and ventricular arrhythmias (Lathers et al. 2008b). In a rat model of pressure overload-induced cardiac hypertrophy of the right ventricle, new cardiac sympathetic nerves were found to express beta (3)-tubulin (axonal marker), GAP43 (growth-associated cone marker), and tyrosine hydroxylase, resulting in large increases in all three in only the right ventricle. Nerve growth factor was also upregulated. Norepinephrine and dopamine content was attenuated, whereas protein and kinase activity of tyrosine hydroxylase were downregulated in the right ventricle. Reuptake of 125Imetaiodobenzylguanidine and 3H-norepinephrine were decreased in the right ventricle. This latter fact is indicative of a functional downregulation in cardiac sympathetic nerves. The cardiac sympathetic nerves in hypertrophic right ventricles expressed highly polysialylated neural cell adhesion molecule, an immature neuron marker, as well as neonatal heart. Kimura et al. (2007) conclude pressure overload induced anatomical sympathetic hyperinnervation and simultaneously initiated deterioration of neuronal cellular function. Rejuvenation of cardiac sympathetic nerves as well as the hypertrophic cardiomyocytes also showed fetal form gene expression. In summary, Lathers et al. (1986b, 1988a, 1990) examined sympathetic beta-receptor density gradient as a cardiac measure that regional differences exist in the beta-adrenoceptor densities among the areas of the heart and that these differences are directly related to the cardiac muscular strength of the speciἀc area, that is, ventricles have a signiἀcantly greater density than the atria and that the left ventricle has a signiἀcantly greater density than the right ventricle. In association with the nonuniform sympathetic innervation of the different regions of the heart, the differential activation of the sympathetic innervation associated with the differential receptor densities relate directly to the arrhythmia and sudden death that accompany coronary occlusion. In addition, Lathers and colleagues (Lathers 1980, 1981; Lathers et al. 1977a, 1977b, 1978) theorized that pharmacological agents, such as betablockers, by preventing or modifying nonuniform neural discharge, would exhibit an antiarrhythmic action in the myocardial coronary occlusion model for sudden death (Lathers et al. 1977a). The previously described studies on timolol and dilevalol support their views. Support for the importance of beta-adrenergic receptors in sudden death after coronary occlusion comes from the referenced studies on hormone sensitivity, exercise, and genetics showing that increased sensitivity to beta-adrenergic agonists results in increased sensitivity to arrhythmias after coronary occlusion and increased incidence of sudden death. These early studies of Lathers et al. (Lathers 1980, 1981; Lathers et al. 1977a, 1977b, 1978) have been conἀrmed by the studies of Druschky et al. (2001), Ieda et al. (2006, 2007, 2008), and Kimura et al. (2007). Recently, Ieda et al. (2008) conἀrmed the conclusion of the earlier studies and discussions of Lathers et al. (Lathers et al. 1977a, 1977b, 1978; Lathers

Animal Model for Sudden Cardiac Death

547

and Schraeder 1982). These current studies have again raised the question of whether regulation of cardiac nerves is an important variable in the management of sudden cardiac death because the heart is extensively innervated and its performance is regulated by the autonomic nervous system. Innervation density is high in the subepicardium and in the central conduction system. In diseased hearts, cardiac innervation density varies. This may lead to sudden cardiac death. After myocardial infarction, sympathetic denervation is followed by reinnervation within the heart, leading to unbalanced neural activation and lethal arrhythmia (Lathers et al. 1986b, 1988a, 1990). One must ask the question: “Does regulation of cardiac neural discharge and associated cardiac receptors become a ‘new paradigm’ in the management of sudden death associated in some persons with epilepsy?” Future animal experiments are required to answer this.

References Alkadhi, K. A., and K. H. Alzoubi. 2010. Synaptic plasticity of autonomic ganglia: Role of chronic stress and implication in cardiovascular diseases and sudden death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Billman, G. E., L. C. Castillo, J. Hensley, C. M. Hohl, and R. A. Altschuld. 1997. Beta2-adrenergic receptor antagonists protect against ventricular ἀbrillation: In vivo and in vitro evidence for enhanced sensitivity to beta2-adrenergic stimulation in animals susceptible to sudden death. Circulation 96 (6): 1914–1922. Billman, G. E., M. Kukielka, R. Kelley, M. Moustafa-Bayoumi, and R. A. Altschuld. 2006. Endurance exercise training attenuates cardiac beta2-adrenoceptor responsiveness and prevents ventricular ἀbrillation in animals susceptible to sudden death. Am J Physiol Heart Circ Physiol 290 (6): H2590–H2599. Ceremuzynski, L., J. Staszewska-Barczak, and K. Herbaczynska-Cedro. 1969. Cardiac rhythm disturbances and the release of catecholamines after acute coronary occlusion in dogs. Cardiovasc Res 3 (2): 190–197. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Du, X. J., X. M. Gao, G. L. Jennings, A. M. Dart, and E. A. Woodcock. 2000. Preserved ventricular contractility in infarcted mouse heart overexpressing beta(2)-adrenergic receptors. Am J Physiol Heart Circ Physiol 279 (5): H2456–H2463. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Houle, M. S., R. A. Altschuld, and G. E. Billman. 2001. Enhanced in vivo and in vitro contractile responses to beta(2)-adrenergic receptor stimulation in dogs susceptible to lethal arrhythmias. J Appl Physiol 91 (4): 1627–1637. Ieda, M., H. Kanazawa, K. Kimura, F. Hattori, Y. Ieda, M. Taniguchi, J. K. Lee, et al. 2007. Sema3a maintains normal heart rhythm through sympathetic innervation patterning. Nat Med 13 (5): 604–612. Ieda, M., H. Kanazawa, Y. Ieda, K. Kimura, K. Matsumura, Y. Tomita, T. Yagi, et al. 2006. Nerve growth factor is critical for cardiac sensory innervation and rescues neuropathy in diabetic hearts. Circulation 114 (22): 2351–2363. Ieda, M., K. Kimura, H. Kanazawa, and K. Fukuda. 2008. Regulation of cardiac nerves: A new paradigm in the management of sudden cardiac death? Curr Med Chem 15 (17): 1731–1736. Jehi, L., and I. M. Najm. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Clevel Clin J Med 75 (Suppl 2): S66–S70.

548 Sudden Death in Epilepsy: Forensic and Clinical Issues Kimura, K., M. Ieda, H. Kanazawa, T. Yagi, M. Tsunoda, S. Ninomiya, H. Kurosawa, et al. 2007. Cardiac sympathetic rejuvenation: A link between nerve function and cardiac hypertrophy. Circ Res 100 (12): 1755–1764. Lathers, C. M. 1980. Effect of timolol on autonomic neural discharge associated with ouabainÂ�induced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106. Lathers, C. M. 1981. Induced disease. Myocardial infarction in dogs and cats. Paper read at the Mammalian Models for Research on Aging, at Washington, DC. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder, eds. 1990. Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42 (2): 123–136. Lathers, C. M., K. M. Keller, J. Roberts, and A. B. Beasley. 1977a. Chapter 5. Role of the adrenergic nervous system in arrhythmia produced by acute coronary artery occlusion. In Pathophysiology and Therapeutics of Myocardial Ischemia, ed. A. Lefer, G. J. Kelliher, and M. Rovetto. New York, NY: Spectrum. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977b. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., R. F. Flax, and L. J. Lipka. 1986a. The effect of C1 spinal cord transection or bilateral adrenal vein ligation on thioridazine-induced arrhythmia and death in the cat. J Clin Pharmacol 26 (7): 515–523. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986b. Regional distribution of myocardial betaadrenoceptors in the cat. Eur J Pharmacol 130 (1–2): 111–117. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986c. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39 (22): 2121–2141. Lathers, C. M., W. H. Spivey, and N. Tumer. 1988a. The effect of timolol given ἀve minutes after coronary occlusion on plasma catecholamines. J Clin Pharmacol 28 (4): 289–299. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988b. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28 (8): 736–745. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988c. The effect of intracerebroventricular d-ALA2 methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43 (26): 2287–2298. Lathers, C. M., N. Tumer, and J. M. Schoffstall. 1989. Plasma catecholamines, pH, and blood pressure during cardiac arrest in pigs. Resuscitation 18 (1): 59–74. Lathers, C. M., W. H. Spivey, R. M. Levin, and N. Tumer. 1990. The effect of dilevalol on cardiac autonomic neural discharge, plasma catecholamines, and myocardial beta receptor density associated with coronary occlusion. J Clin Pharmacol 30 (3): 241–253. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008a. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher, Chapter 13. Hauppauge, NY: Nova Science.

Animal Model for Sudden Cardiac Death

549

Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Lujan, H. L., and S. E. Dicarlo. 2008. Sex differences to myocardial ischemia and beta-adrenergic receptor blockade in conscious rats. Am J Physiol Heart Circ Physiol 294 (4): H1523–H1529. Lujan, H. L., V. J. Kramer, and S. E. DiCarlo. 2007. Sex influences the susceptibility to reperfusioninduced sustained ventricular tachycardia and beta-adrenergic receptor blockade in conscious rats. Am J Physiol Heart Circ Physiol 293 (5): H2799–H2808. Ogawa, M., S. Zhou, A. Y. Tan, J. Song, G. Gholmieh, M. C. Fishbein, H. Luo, et al. 2007. Left stellate ganglion and vagal nerve activity and cardiac arrhythmias in ambulatory dogs with pacinginduced congestive heart failure. J Am Coll Cardiol 50 (4): 335–343. Randall, W. C., ed. 1984. Nervous Control of Cardiovascular Function. New York, NY: Oxford University Press. Randall, W. C., ed. 1977. Neural Regulation of the Heart. New York: Oxford University Press. Schraeder, P. L., and C. M. Lathers. 1980. Autonomic dysfunction in epilepsy: I. A proposed animal model for unexplained sudden death in epilepsy. Clin Res 28: 618A. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., and C. M. Lathers. 1995. Clinical pharmacology of antiepileptic drug use: Clinical pearls about the perils of patty. J Clin Pharmacol 35 (12): 1120–1135. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13 (1): 263–264; author reply, 265–269. Shimizu, M., A. Kagawa, T. Takano, H. Masai, and Y. Miwa. 2008. Neurogenic stunned myocardium associated with status epileptics and postictal catecholamine surge. Intern Med 47 (4): 269–273.

Antiepileptic Activity of Beta-Blocking Agents Claire M. Lathers Kam F. Jim William H. Spivey Claire Kahn Kathleen Dolce William D. Matthews

34

Contents 34.1 Introduction 34.2 Method 34.2.1 Seizure Studies in Anesthetized Cats 34.2.2 Seizure Studies in Anesthetized Pigs 34.3 Results 34.3.1 Data Obtained in Anesthetized Cats 34.3.2 Data Obtained in Anesthetized Pigs 34.4 Discussion 34.5 Summary Acknowledgments References

551 552 552 553 553 553 559 561 563 564 564

34.1â•…Introduction The etiology of unexplained sudden death in epilepsy is unknown, but autonomic dysfunction is a frequently presented hypothesis (Hirsch and Martin 1971; Leestma et al. 1984). Autonomic neural dysfunction contributes to the development of cardiac arrhythmias (Gillis 1969; Lathers et al. 1977, 1978; Randall et al. 1978) and sudden death (Lown and Verrier 1978). Stimulation of the sympathetic ventrolateral cardiac nerve produces tachyarrhythmias (Randall et al. 1978). Sympathetic neural imbalance, or “neural nonuniformity,” is deἀned to exist when neural activity is recorded simultaneously from two or three postganglionic cardiac sympathetic nerves in the same cat and increases and/or decreases in activity develop (Lathers et al. 1977, 1978). The neural discharge is associated with arrhythmias; it is hypothesized that the nonuniform neural discharge induced nonuniform changes in excitability and conduction in the heart, producing arrhythmia, as reported by Han and Moe (1964). Autonomic sympathetic and parasympathetic cardiac neural dysfunction and cardiovascular abnormalities, for example, changes in blood pressure, heart rate, and rhythm, as well as alterations in cardiac depolarization and repolarization, are associated with interictal and ictal spike activity induced by pentylenetetrazol (PTZ) (Carnel et al. 1985; Lathers 551

552 Sudden Death in Epilepsy: Forensic and Clinical Issues

and Schraeder 1982; Lathers et al. 1984, 1985; Schraeder and Lathers 1983). The major action of PTZ is thought to be mediated via the central nervous system and not by direct stimulation of the peripheral systems (Toman and Davis 1949). Furthermore, Gremels (1931) and Hildebrandt (1937), quoted by Hahn (1960), concluded that PTZ did not exert a positive inotropic action on the heart. Thus, the cardiovascular changes observed using the PTZ model of epilepsy in our study appear to be due to autonomic dysfunction associated with epileptiform activity rather than to a direct action of PTZ on the heart and peripheral vascular system. To exclude the possibility that the autonomic dysfunction observed with the epileptiform activity in our previous studies was due to the peripheral effects of PTZ, we examined the action of intracerebroventricularly (i.c.v.) administered PTZ to induce epileptiform activity and autonomic dysfunction. It was hypothesized that if the autonomic dysfunction, including the development of arrhythmias, also occurred in humans, it might contribute to sudden unexplained death of epileptic patients. This study examined whether the beta-blocking agent timolol, given centrally and then peripherally, would eliminate the cardiac arrhythmias, epileptiform activity, and changes in the blood pressure and heart rate induced by PTZ intracerebroventricularly administered to anesthetized cats. To further examine the potential anticonvulsant action of beta-�blocking agents in a different animal model, the action of propranolol on epileptogenic activity induced by intravenously (i.v.) administered PTZ in anesthetized pigs was studied.

34.2â•…Method 34.2.1â•… Seizure Studies in Anesthetized Cats Twenty-four cats of either sex, weighing from 2.5 to 3.4 kg, were used. Animals were anesthetized with alpha-chloralose (80 mg/kg i.v.) and cannulated as described by Lathers and Schraeder (1982). Cats were positioned in a stereotaxic apparatus. Coordinates A +0.5, HD +8.5, and RL +4 (Snider and Niemer 1970) were used to locate the cannula through a burr hole at a 90° angle. PTZ (10 or 20 mg) was dissolved in 50 μL saline and injected by using a 26G spinal needle placed into the left lateral ventricle. The electrocorticogram (ECoG) was monitored using silver ball electrodes placed on the surface of the sigmoid gyri unilaterally after a bifrontal craniectomy and dural resection. A recording electrode was inserted into the right hippocampus (A +7, HD −6.0, and RL +11.8) using a concentric bipolar electrode. Four experimental groups of cats were examined. In the ἀrst, only 10 mg i.c.v. PTZ was administered (n = 8 cats). In the second, PTZ 10 and/or 20 mg i.c.v. was administered and followed by timolol 10, 100, 500 μg/kg i.c.v. and 1, 5, 10, and/or 20 mg/kg i.v. (n = 6 cats). Doses were given at 5-min intervals. In the third group, the pharmacological agent (d-Ala5) methionine-enkephalinamide or prostaglandin E2 was administered intracerebroventricularly before the administration of PTZ and timolol (n = 7). In the fourth group (n = 3 cats), the same doses of timolol were administered at different time intervals, that is, one cat received timolol every 10 min and a second cat received timolol every 15 min after the administration of PTZ. In the third cat, timolol doses were given every 15 min but no PTZ was administered. A one-factor repeated-measures analysis of variance (ANOVA) was performed to determine whether changes in the mean arterial blood pressure and heart rate after the

Antiepileptic Activity of Beta-Blocking Agents

553

intracerebroventricular injection of PTZ (n = 8 cats) were signiἀcantly different from the control. To verify that it would be appropriate to combine the data obtained in the experimental cats receiving other pharmacological agents before the administration of PTZ and timolol with those cats receiving only PTZ and timolol, a two-factor ANOVA with repeated measures on time and nonrepeated measures on group was used to analyze the effects of time course, group membership, and the interaction of these two variables on blood pressure and heart rate, separately. The time course reflects the effect over time of PTZ and timolol on blood pressure or heart rate. To determine whether the PTZ–timolol actions on heart rate and blood pressure varied from the actions of only PTZ, a two-way ANOVA was done in which the independent variables were the presence or absence of timolol and the time course. Simple effect tests and Tukey honestly signiἀcantly different (HSD) post hoc tests were used where appropriate. For the 13 cats receiving PTZ and timolol, the ECoG data for the left and right cortex and for the hippocampus were analyzed separately by dividing the number of episodes of epileptiform activity into categories of ≤10-s and >10-s duration. For each cat, the mean numbers of episodes for the two durations were calculated for the control and for each dose of PTZ and timolol. Means were calculated by combining the ECoG data from all cats. To determine if there were any statistically signiἀcant differences among the means, data were analyzed by a Friedman rank ANOVA followed by a Bonferroni-corrected Wilcoxon post hoc tests. The following time points were analyzed: control; PTZ 10 and 20 mg; and timolol 10 μg, 100 μg, 500 μg, and 1 mg/kg. 34.2.2â•… Seizure Studies in Anesthetized Pigs Domestic swine weighing 13–20 kg were anesthetized with ketamine (20 mg/kg i.v.) and alpha-chloralose (80 mg/kg i.v.). Animals were prepared as described by Spivey et al. (1987). After a 10-min equilibration period, seizure activity was induced by PTZ (100 mg/ kg i.v.). Sixty seconds after the onset of epileptogenic activity, the animals were treated with no drug (control group) or propranolol 2.5 mg/kg i.v. Seizure activity was monitored for 20€min. Plasma levels of propranolol were determined by drawing blood samples (8 mL) at 1, 2, 5, 10, 15, and 20 min after propranolol. The samples were centrifuged at 1000 g for min and the plasma (3 mL) was stored frozen at −20°C. The concentrations of propranolol in plasma were determined by a modiἀcation of the method of Albani et al. (1982). The procedure involves reverse-phase high-pressure liquid chromatographic resolution with fluorometric detection of propranolol and an internal standard, Carvedilol (1-(4-carbazolyloxy)-3-[2(2-methoxyphenoxy)ethylamino]-2-propranol).

34.3â•…Results 34.3.1â•…Data Obtained in Anesthetized Cats A one-factor repeated-measures ANOVA revealed that the increase in the control blood pressure of 103 ± 13 to 149 ± 13 mm Hg minutes after intracerebroventricular PTZ was signiἀcant, whereas the mean heart rate increase from 159 ± 17 to 162 ± 10 bpm 3 min after

554 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 34.1â•… Sequence in Which Incidence of the Epileptiform Activity and an Increase in Mean Arterial Blood Pressure and Heart Rate Occurred after Intracerebroventricular PTZa Precededb

Same time

Proceededc

No change

0 0 1 0

2 1 1 1

0 1 0 0

BP 5 HRd 5 BPe 11 HRe 12 d

a b

c

d e

Incidence indicates the number of cats of the 13 studied. Preceded indicates that the increase in blood pressure or heart rate occurred before the onset of epileptiform activity elicited by PTZ. Proceeded indicates that the increase in blood pressure or heart rate occurred after the onset of epileptiform activity elicited by PTZ. Seven cats receiving only PTZ; the ECoG activity was not recorded in the eighth cat. Thirteen cats receiving PTZ and timolol.

PTZ and to 175 ± 14 bpm 15 min after PTZ was not signiἀcant. The PTZ-induced increase in blood pressure and heart rate preceded the initiation of epileptiform activity in most of these cats (Table 34.1). This trend was also observed in the 13 cats receiving PTZ and then timolol (Table 34.2). During the occurrence of PTZ-induced epileptiform activity, premature ventricular contractions of 3–8 min were observed in three of the eight cats. Data obtained from one cat are illustrated in Figure 34.1. In panel a (control), the cerebral activity present was that associated with the induction of anesthesia. With the intracerebroventricular administration of 10 mg PTZ (panel b), there was induction of epileptiform activity, a slight increase in blood pressure, and a slight decrease in heart rate. When timolol (10 μg/kg i.c.v.) was administered (panel c), the epileptiform activity was diminished, although the blood pressure and heart rate values were still elevated above control. The intracerebroventricular administration of PTZ increased both the mean arterial blood pressure and the heart rate in six cats receiving only PTZ and timolol; no other pharmacological agents were administered (data not shown). When timolol was given to the six Table 34.2â•… Incidence of the Sequence in Which Suppression of Intracerebroventricular PTZ-Induced Epileptiform Activity and a Decrease in Mean Arterial Blood Pressure and Heart Rate Occurred after Timolola Partial Suppression Compared to Control

Total Suppression Compared to Control

Total Suppression Compared to Mean of Previous 5-min Interval

No No Precededb Same Proceededc Preceded Same Proceeded Change Preceded Same Proceeded Change BP 1 HR 0 a b

c

d

0 0

12 13

11 12

0 0

1 0

1 1

7 2

0 1

4 6d

2 5d

Incidence indicates the number of cats of the 13 studied. Preceded indicates that the decrease in blood pressure or heart rate occurred before the partial or total suppression of epileptiform activity. Proceeded indicates that the decrease in blood pressure or heart rate occurred after the partial or total suppression of epileptiform activity. Includes same cat, hippocampus no change; both left and right cortex proceeded total suppression of epileptogenic activity.

Electroencephalogram (µV)

Antiepileptic Activity of Beta-Blocking Agents

Mean arterial blood 200 pressure 1000 (mm Hg) Electrocardiogram (Lead II)

(a)

(b)

555

(c)

L. motor cortex R. motor cortex R. hippocampus 108

123

143

122

113

130

1s Control

1s PTZ 10 mg i.c.v.

Timolol 10 µg/kg i.c.v.

Figure 34.1╇ Effect of timolol on the ECoG and cardiovascular parameters associated with

intracerebroventricular PTZ-induced epileptiform activity in one cat. Represented in all three panels are ECoGs from the left and right motor cortex and right hippocampus, as well as mean arterial blood pressure, and electrocardiogram from top to bottom, respectively. (Reproduced from Lathers, C. M., et al., Epilepsy Res, 4, 42–54, 1989b. With permission.)

cats after the administration of PTZ, the mean arterial blood pressure and heart rate values began to decrease. Similar data were obtained when PTZ and timolol were administered to cats that had previously received other agents (n = 7 cats). In the cats receiving other pharmacological agents before the administration of PTZ, the blood pressure, heart rate, and ECoG activity had returned to baseline values before the administration to PTZ. No signiἀcant effect of previous drugs or interaction between previous drug and time was found. The lack of a signiἀcant effect indicated that the two groups of animals responded the same ways to the drug injections over time and permitted combination of the two groups for purposes of analysis (n = 13 cats; Figure 34.2). Thus, the administration of PTZ increased both mean arterial blood pressure and heart rate. When timolol was given intracerebroventricularly, the mean arterial blood pressure and heart rate values began to decrease. In the two-way ANOVA, both main effects and the interaction were signiἀcant for the heart rate. A comparison of the two experimental groups (PTZ i.c.v. and PTZ–timolol, n = 13 cats) at each experimental time showed that the timolol group was lower in the control period, sharply rising to a nonsigniἀcant difference after the second dose of PTZ and after the ἀrst and second doses of timolol. At higher doses of timolol, the timolol group was consistently signiἀcantly lower in heart rate than the group receiving only PTZ. Experimental times within groups, done as separate one-way repeated-measures ANOVAs, due to the different variances, showed no signiἀcant difference across time in the heart rate for the cats receiving no timolol and a sharp rise followed by a fall (Tukey HSD) in the cats receiving PTZ and increasing doses of timolol. The two-way ANOVA revealed that the group main effects on blood pressure were not signiἀcant. The time effect and the interactions were signiἀcant. When the experimental groups were compared at each experimental time, no differences were found until the two highest doses of timolol. The data in the control group were steady at the time equivalent to that when timolol would have been administered to the animals in the experimental group, whereas the timolol group exhibited a rapid fall in blood pressure. Consequently,

556 Sudden Death in Epilepsy: Forensic and Clinical Issues 180

11 11

160

11

140

Mean arterial blood pressure (mean ± SE mm Hg)

10

11

N = 13 cats

11 12 12 12

120

12 12

100 12

80

11 12 12 12 12

60

12

12 12

40

12

8

8

20 –10

6

2

Control

14

18

22

26

30

34

38

42

20 mg PTZ, i.c.v.

100 µg/kg 1 mg/kg 10 mg/kg Timolol, i.c.v. Timolol, i.c.v. Timolol, i.v. 10 µg/kg 500 µg/kg 5 mg/kg 20 mg/kg Timolol, i.c.v. Timolol, i.c.v. Timolol, i.v. Timolol, i.v.

10 mg PTZ, i.c.v.

(a)

10

Time (min)

180

Heart rate (beats/min) mean ± SE

N = 13 cats 160

11

11

11 1110

11

140

12

12 12

120

12 11

100

12 12 12 11 12 12 11 1212 12 8 8

80 –10 Control

2

6

10

20 mg PTZ, i.c.v.

10 mg PTZ, i.c.v.

14

18

22

26

30

34

38

42

100 µg/kg 1 mg/kg 10 mg/kg Timolol, i.c.v. Timolol, i.c.v. Timolol, i.v. 10 µg/kg 500 µg/kg 5 mg/kg 20 mg/kg Timolol, i.c.v. Timolol, i.c.v. Timolol, i.v. Timolol, i.v.

Time (min)

(b)

Figure 34.2╇ The effect of timolol on mean arterial blood pressure and heart rate changes

induced by PTZ administered intracerebroventricularly in 13 cats. In the upper graph, the mean arterial blood pressure is graphed as a function of time. The lower graph depicts the mean heart rate. The arrows along the abscissa indicate the administration of PTZ intracerebroventricularly and timolol intracerebroventricularly or intravenously. (Reproduced from Lathers, C.€M., et€al., Epilepsy Res, 4, 42–54, 1989b. With permission.)

Antiepileptic Activity of Beta-Blocking Agents

557

the latter group was signiἀcantly lower. Experimental times within each group showed no signiἀcant effect of time course on blood pressure in those cats receiving only PTZ, whereas in the cats receiving both PTZ and timolol, the PTZ increased the blood pressure and timolol reversed the effect of PTZ. In the two cats receiving the same doses of timolol at time intervals other than 5 min (i.e., 10 or 15 min), timolol suppressed the epileptogenic discharges and decreased the blood pressure and heart rate. In the one cat in which all doses of timolol were given every 15 min but no epileptogenic activity occurred since PTZ was not given, both the heart rate and blood pressure values were decreased, but the magnitude of the decrease was much less than that observed in the cats given PTZ. Although PTZ induced both interictal and ictal epileptiform discharges in all cats, most epileptiform activity exhibited durations of ≤10 s, that is, interictal and brief ictal activity (Table 34.3 and Figure 34.3). The dashed curve in Figure 34.3 depicts the epileptiform activity with a duration of ≤10 s obtained in the left cortex of the cats receiving only PTZ and no timolol. The mean number of episodes of epileptiform activity remained approximately the same for the 45-min period, the time equivalent to the entire experimental duration in the cats receiving both PTZ and increasing doses of timolol. The mean number of episodes of epileptiform activity lasting <10 s for the other three brain areas depicted in Figure 34.3 decreased with increasing doses of timolol. Timolol produced this decrease, although most cats in this group received two doses of PTZ, that is, 10 and 20 mg. In all 13 cats receiving timolol after PTZ, the timolol partially suppressed the epileptiform activity evident in the left and right cortices and in the hippocampus after the ἀrst dose of 10 μg. The mean times to partial suppression for the left and right cortices and for the hippocampus were 36 ± 11, 33 ± 11, and 50 ± 15 s, respectively, after 10 μg timolol. In 12 of the 13 cats, the time to partial suppression was the same for all three brain areas. In one cat, epileptiform activity was suppressed 27 s earlier in the left and right cortices when compared to the hippocampus. Higher doses of timolol, ranging from 1 to 20 mg, were required to induce a total suppression of the epileptiform activity. Table 34.4 shows the number of cats in which a given dose of timolol induced total suppression of epileptiform activity. Using the time of administration of the ἀrst dose of timolol 10 μg as zero, the Table 34.3â•… Mean Duration of Episodes of Epileptogenic Activity Lasting >10 s (seconds/minute; mean ± SE) Experimental Groups

Left Cortex

Right Cortex

Right Hippocampus

Control PTZ, 10 mg i.c.v. PTZ, 20 mg i.c.v. Timolol, 10 μg/kg i.c.v. Timolol, 100 μg/kg i.c.v. Timolol, 500 μg/kg i.c.v. Timolol, 1 mg/kg i.c.v. Timolol, 5 mg/kg i.v. Timolol, 10 mg/kg i.v.

0 0 0.31 ± 0.31b 0.50 ± 0.50b 0 0 0 0 0

0 0 0.54 ± 0.54b 0.45 ± 0.45b 0 0 0 0 0

0 1.21 ± 1.21b 1.00 ± 0.65 5.51 ± 4.63 4.22 ± 4.22b 0 4.62 ± 4.62b 0.89 ± 0.89b 0

a

Source: Lathers, C. M., et al., Epilepsy Res, 4, 42–54, 1989b. With permission. Deἀned as the mean ± SE for the 13 cats in which a mean 10-min control period was obtained. b Mean ± SE was calculated by averaging 12 zeros and 1 number, resulting in an SE value that is the same as the mean. a

558 Sudden Death in Epilepsy: Forensic and Clinical Issues 22

}

Left cortex Right cortex Cats received PTZ and timolol Hippocampus Left cortex in cats receiving PTZ only

Mean number of episodes epileptiform activity lasting ≤10 s (episodes/min; mean ± SE)

20 18 16 14 12 10 8 6 4 2 0

Control

5

10

15

20

25

30

35

30

45

PTZ PTZ Timolol Timolol Timolol Timolol Timolol Timolol Timolol 10 100 500 1 5 10 25 10 20 mg/kg mg/kg µg/kg µg/kg µg/kg mg/kg mg/kg mg/kg mg/kg

50

55

60

Figure 34.3╇ Epileptiform activity with a duration of ≤10 s obtained in cats treated with PTZ and timolol. For purposes of comparison, data obtained in cats receiving only PTZ are shown for the comparable experimental duration. The control values obtained in the cats receiving both PTZ and timolol are means obtained in the 10-min period before the first dose of PTZ. (Reproduced from Lathers, C. M., et al., Epilepsy Res, 4, 42–54, 1989b. With permission.)

mean times to total suppression of epileptiform activity in the left and right cortices and the hippocampus were 26 ± 2, 26 ± 2, and 27 ± 2 min, respectively. In six cats, the time to total suppression of epileptiform activity was the same in only three brain regions. In six of the other seven cats, the dose of timolol producing total suppression of epileptiform activity was the same for all three brain areas; the times to suppression varied by only a few seconds. In one cat the epileptiform activity was suppressed in the left and right cortices at a dose of 5 mg timolol, whereas a dose of 10 mg was required to elicit total suppression in the hippocampus. The administration of timolol decreased the duration of all types of epileptiform activity, that is, prolonged ictal (>10 s), brief ictal, and interictal (Table 34.3). Table 34.4 Number of Cats and Dose of Timolol Inducing Total Suppression of Epileptiform Activitya Number of Cats Doses of Timolol (mg)

1

5

10

20

Left cortex Right cortex Hippocampus

2 2 2

3 3 2

1 1 2

6 5 7

a

Epileptiform activity was not induced by PTZ in the left and right cortex of one cat and in the right cortex of a second animal.

Antiepileptic Activity of Beta-Blocking Agents

559

The Friedman ANOVAs were signiἀcant (p ≤ 0.0001) for all three areas of the brain. The Wilcoxon post hoc tests revealed the expected PTZ-induced increase in epileptiform activity from control. In the 10 cats receiving 20 mg PTZ, this dose was not different from control (for any of the three areas), although the means were higher than for 10 mg PTZ. In fact, none of the Bonferroni-corrected Wilcoxon tests were signiἀcant between 20 mg PTZ and any other dose, for any area. The ἀrst dose of timolol elicited slightly (not signiἀcant) higher rates of prolonged and brief ictal and interictal activity than 20 mg PTZ, after which a steady decrease in epileptiform activity occurred, becoming signiἀcant (one-tailed) by the dose of 1 mg/kg i.c.v. timolol. This trend occurred in all three areas of the brain. None of the Friedman ANOVAs or duration more than 10 s approached signiἀcance; this was anticipated because many subjects had no prolonged ictal episodes at any dose. Table 34.2 shows the sequence in which the mean arterial blood pressure and heart rate were depressed by timolol in relation to the time when timolol partially or totally suppressed the epileptiform activity. In 12 of 13 cats, the epileptiform activity was partially suppressed before the fall in mean arterial blood pressure, that is, the fall in this parameter followed partial suppression induced by 10 μg timolol. The heart rate decreased in all 13 cats after the epileptiform activity was partially suppressed by this dose of timolol. In almost all of the cats, the decrease in the mean arterial blood pressure and heart rate preceded total suppression of epileptogenic activity when the cardiovascular parameters were compared to control values. When mean arterial blood pressure and heart rate values at the time of total suppression of epileptogenic activity were compared to blood pressure and heart rate values in the preceding 5-min interval, in most cats (7/13) the blood pressure decreased before total suppression of the epileptiform activity by timolol. However, the decrease in the mean heart rate followed total suppression of epileptogenic activity in six cats. In ἀve cats, the heart rate did not change immediately before or after the occurrence of total suppression of epileptiform activity. 34.3.2â•…Data Obtained in Anesthetized Pigs A transient increase (16.3–50.0%) in the mean arterial blood pressure occurred after the PTZ administration. The elevated blood pressure gradually declined to the basal level within 10 min after PTZ. Intravenously administered propranolol signiἀcantly reduced this transient pressor response and returned the elevated blood pressure to the basal level at 2 min after drug infusion (Figure 34.4). There was no signiἀcant change in the basal heart rate after PTZ administration. However, a marked bradycardia was observed at 1 min after intravenous propranolol administration. Propranolol produced a maximal decrease of 32–38% in the basal heart rate; the bradycardia persisted throughout the experiment (Figure 34.5). Epileptogenic activity induced by PTZ was associated with the occurrence of premature ventricular contractions in some pigs, similar to those observed in anesthetized cats when administered PTZ. Figure 34.4 illustrates the duration of seizure activity elicited by PTZ over an experimental period of 20 min. The seizure activity was continuous from t = 0 to 1 min for both groups. Intravenous propranolol produced a signiἀcant reduction in the duration of seizure activity 1 min after drug infusion; the seizure durations (second per minute interval) were 36.3 ± 4.8 and 12.3 ± 5.1 for the control and intravenous groups, respectively. Animals treated with intravenous propranolol had reduced duration of seizure activity throughout the entire experiment when compared to the control animals.

560 Sudden Death in Epilepsy: Forensic and Clinical Issues

Duration of seizure activity (s/min interval)

60 50 40 30 20 10 0

0

5 Propranolol (2.5 mg/kg)

10 Time (min)

15

20

PTZ (100 mg/kg, i.v.)

Figure 34.4╇ The effect of intravenous propranolol (2.5 mg/kg) on seizure activity induced by PTZ (100 mg/kg i.v.) in pigs (n = 5–6). PTZ was given to induce seizure activity. Propranolol was administered 60 s after the onset of seizure activity. The seizure activity at time zero was determined from the seizure duration of 0- to 1-min interval. A significant suppression of seizure activity was observed at 1 min after propranolol administration. (Modified and reproduced from Lathers, C. M., et al., Epilepsia, 30, 473–479, 1989a. With permission.)

Propranolol µg/ml plasma

10.0

1.0

0.1

0

2

4

6

8

10

12

14

16

18

20

Minutes after the dose

Figure 34.5╇ Propranolol plasma concentrations versus time in pigs administered propranolol 2.5 mg/kg i.v. Values are means ± SD (n = 6).

Antiepileptic Activity of Beta-Blocking Agents

561

Plasma propranolol concentrations were determined from 1 to 20 min after the intravenous administration of 2.5 mg/kg. After a rapid fall from 6.04 ± 1.43 μg/mL at minute 1 to 1.69 ± 0.31 μg/mL at minute 5, a steady decline in plasma concentration was observed up to 20 min, with a mean half-life of 23.3 ± 4.8 min (Figure 34.5). The brevity of the experiment precluded the determination of further kinetic parameters.

34.4â•…Discussion Previous studies have shown that PTZ administered intravenously induced epileptiform activity associated with cardiac arrhythmias and changes in the mean arterial blood pressure and heart rate (Carnel et al. 1985; Lathers and Schraeder 1982; Lathers et al. 1984; Schraeder and Lathers 1983). The present study demonstrated that the central administration of PTZ produced similar effects within seconds of its intracerebroventricular injection. These data support the conclusion that the intravenous administration of PTZ has little direct effect on the heart in eliciting cardiac arrhythmias and that such arrhythmias are associated with the epileptiform discharges induced by PTZ. The central administration of timolol decreased mean arterial blood pressure and heart rate. In another study, timolol administered intravenously to anesthetized cats also decreased the mean arterial blood pressure and heart rate (Lathers 1980) and exhibited an antiarrhythmic action against ouabain-induced arrhythmias. Lathers et al. (1986) demonstrated that chronic oral dosing with timolol for 1 or 2 weeks increased the time, although not signiἀcantly, to arrhythmia induced by acute permanent occlusion of the left anterior descending coronary artery. In the present study, increasing doses of timolol administered intracerebroventricularly and intravenously not only signiἀcantly decreased the elevation of mean arterial blood pressure and heart rate but also decreased and subsequently abolished the incidence of cardiac arrhythmias associated with the epileptiform activity. Epileptiform activity elicited in the present study with a duration less than or equal to 10 s includes both interictal and brief ictal discharges. PTZ-induced bilateral interictal spike activity is indicative of increased cortical excitability often evident in the EEG records of epileptic individuals. PTZ-induced generalized asynchronous clonic movements followed by a tonic convulsion in which limb movements are flexion followed by extension are analogous to the brief ictal discharges. Motor activity characterized by forelimb clonus is analogous to the prolonged ictal activity, that is, duration greater than 10€s. In the animal model using the cat and intracerebroventricular injection of PTZ, most of the epileptiform activity elicited was interictal and/or brief ictal activity. It has been hypothesized that the cardiac arrhythmias associated with interictal activity could be one potential mechanism for sudden unexplained death in epileptic persons (Carnel et al. 1985; Lathers and Schraeder 1982; Lathers et al. 1984; Schraeder and Lathers 1983). That timolol either partially or totally abolishes the epileptiform activity in the shorter duration category suggests that it may be a useful therapeutic agent to suppress the interictal discharges associated with cardiac arrhythmias. The administration of PTZ elicited epileptiform activity that was followed by increases in blood pressure, heart rate, and cardiac arrhythmias. Exactly how these changes develop is unknown, but this laboratory (Kraras et al. 1987; Suter and Lathers 1984) has proposed a possible mechanism to explain how epileptogenic activity and autonomic dysfunction may occur in epileptic patients, resulting in fatal cardiovascular changes. PTZ, trauma,

562 Sudden Death in Epilepsy: Forensic and Clinical Issues

inhibition of prostaglandin transport across the blood–brain barrier, or altered synthesis or metabolism of central enkephalins may lead to increased central levels of PGE2 and/or enkephalins. The consequence of this is thought to be inhibition of central GABA release, epileptogenic activity, increased blood pressure and heart rate, increased sympathetic and parasympathetic central neural outflow, impaired or imbalanced cardiac sympathetic and parasympathetic discharge, and a resultant arrhythmia and/or death. In the present study, the central intracerebroventricular administration of timolol partially suppressed the epileptiform activity and subsequently decreased the blood pressure and heart rate values elevated by PTZ. It may be that timolol is interfering with the central actions of PGE2 or enkephalins to reverse their known capabilities to induce epileptiform activity (see Chapter 18). Additional experimental studies are required to verify this suggestion. Additional mechanisms to explain the anticonvulsant and antiarrhythmic actions of timolol are discussed later. It has been theorized that pharmacological agents capable of suppressing epileptiform activity and the sympathetic component of cardiac arrhythmias may be the best regimens to prevent interictal activity and the associated cardiac arrhythmias that may contribute to the production of sudden unexplained death in the epileptic person (Carnel et al. 1985; Lathers and Schraeder 1982; Lathers et al. 1984; Schraeder and Lathers 1983). The data obtained in the present study indicate that in the experimental setting the pharmacologic agent timolol possesses components of both of these capabilities. Blockade of cardiac beta receptors, a cardiac neurodepressant effect, and/or membrane depressant actions of betablocking agents are thought to contribute to the antiarrhythmic action of beta-blocking agents (Lathers and Spivey 1987). PTZ has been used to induce seizure activity in humans (Franz 1980; Van Buren 1958), study seizure mechanisms (Faingold and Berry 1973; Krall et al. 1978; Langeluddeke 1936; Swinyard 1972), examine autonomic dysfunction associated with epileptogenic activity (Lathers and Schraeder 1982; Onuma 1957; Orihara 1952; Schraeder and Lathers 1983; Van Buren 1958; Van Buren and Ajmone-Marsan 1960), and screen anticonvulsant agents (Carnel et al. 1985; Faingold and Berry 1973; Lathers et al. 1984). Because it is accepted that PTZ is a convulsive model and that many drugs capable of suppressing the PTZ-induced epileptiform activity are anticonvulsant agents, the results of this study suggest that timolol exhibited an anticonvulsant action. Although the data indicate that timolol can reverse the effects of PTZ on the brain, this does not necessarily mean that timolol has intrinsic “anticonvulsant” properties separate from an ability to reverse the effects of PTZ. To answer this question, additional studies must be done to determine whether timolol will protect against seizures induced in other experimental models of epilepsy. In particular, it would be important to evaluate the capability of timolol to suppress interictal discharges and cardiac arrhythmias elicited in other in vivo experimental models not involving PTZ. If timolol also suppresses both the interictal discharges and the arrhythmias in these experimental models, this would provide additional evidence to support the possibility that timolol may be an effective agent to use in epileptic patients to prevent sudden unexplained death. The concept that beta-blocking agents may possess anticonvulsant action is not new (Bose et al. 1963; Conway et al. 1978; Papanicolaou et al. 1982). The studies of Dashputra et al. (1985), Jaeger et al. (1979), Murmann et al. (1966), and Tocco et al. (1980) demonstrated that propranolol possesses anticonvulsant actions. Mueller and Dunwiddie (1983) showed that timolol selectively blocked the proconvulsant activity of 2-fluoro-norepinephrine and 1-isoproterenol in in vitro hippocampal slice preparations superfused with penicillin and

Antiepileptic Activity of Beta-Blocking Agents

563

elevated levels of potassium. Louis et al. (1982) reported that propranolol or timolol (0.25 μg/kg i.c.v.) produced an anticonvulsant action when PTZ was used to induce convulsions in rats. The anticonvulsant action of timolol reported here for the data obtained in swine is similar to the anticonvulsant action of diazepam when used in the same experimental model (Lathers et al. 1987; Spivey et al. 1987). The anticonvulsant action of beta-blocking agents is commonly ascribed to a membranestabilizing effect, although exceptions have been reported (Lints and Nyquist-Battie 1985). Other proposed anticonvulsant mechanisms include decreased central serotonergic (Conway et al. 1978) and monoamine oxidase activity (Bose et al. 1963). An additional possible antiepileptic mechanism of the beta-blocking agents may include beta-Â�adrenoceptor blockade, especially beta2 receptors in the central nervous system (Papanicolaou et al. 1982). Although norepinephrine is generally believed to be anticonvulsant, studies suggest that norepinephrine may exacerbate seizure activity via activation of beta receptors. The state of abnormal seizure susceptibility, but not severity, in genetically epilepsy-prone rats may be determined by norepinephrine deἀcits in the hypothalamus/thalamus (Dailey and Jobe 1986). Both severity and susceptibility can be determined by norepinephrine deἀcits in the telencephalon, midbrain, and pons medulla, whereas seizure severity but not susceptibility may be determined by norepinephrine abnormalities in the cerebellum. Noradrenergic effects may not be uniform throughout the hippocampus; thus, selective activation of alpha or beta receptors by norepinephrine in the brain areas such as the hippocampus might produce either anticonvulsant or proconvulsant effects, respectively (Mueller and Dunwiddie 1983). Beta-blocking agents can increase norepinephrine concentration in cerebral spinal fluid (Tackett et al. 1981) and potentiate the effects of exogenously administered norepinephrine on vas deferens contraction (Patil et al. 1968). If a similar action occurred in this study, the establishment of beta blockade with timolol would increase the central norepinephrine concentration. The increased norepinephrine activity at the central postsynaptic alpha1 receptor sites may account for the anticonvulsant effects of beta-blocking agents (Goldman et al. 1987). Thus, the protective mechanism for timolol against seizures induced by PTZ may be due to a selective blockade of seizure-inducing beta receptors, allowing available norepinephrine to stimulate the central alpha1 receptors that exert an anticonvulsant action. In addition to the possibility that the central alpha1 receptors may be involved in the anticonvulsant action of beta-blocking agents, the role of central postsynaptic alpha2 receptors must be evaluated. Activation of alpha2 receptors decreases the excitability of CA1 pyramidal neurons (Mueller et al. 1982). Clonidine and 1-m-norepinephrine are more selective for alpha2 than for alpha1 receptors and inhibit epileptiform activity at low concentrations; the alpha1 agonist 1-phenylephrine was ineffective at much higher concentrations. These data suggest that central postsynaptic alpha 2 receptors may play a greater role than the alpha1 receptors in the anticonvulsant action of timolol observed in the present study. Deἀnitive experiments will have to be done to conἀrm this possibility.

34.5â•… Summary The experiments in this study were designed to explore the ability of beta-blocking agents to suppress seizures induced by PTZ in two species: the cat and the pig. Cats were anesthetized with alpha-chloralose and PTZ (10–20 mg i.c.v.) was administered to elicit epileptiform

564 Sudden Death in Epilepsy: Forensic and Clinical Issues

activity, including both interictal and ictal discharges. Various doses of timolol (10, 100, 500 μg/kg i.c.v. and 1, 5, 10, and/or 20 mg/kg i.v.) were then administered at 5-min intervals to determine whether it suppressed the epileptiform activity. Mean arterial blood pressure increased after the administration of PTZ and was associated with the development of epileptiform activity. Heart rate also was increased after PTZ. All doses of timolol caused a decrease in the blood pressure and heart rate elevated by PTZ. The administration of timolol also suppressed the epileptiform activity. Similar ἀndings were obtained in cats that received the same doses of timolol administered at different time intervals. The data indicate that the central administration of timolol reverses the epileptiform activity of PTZ on the brain and suppresses the associated increases in blood pressure and heart rate. Domestic swine (13–20 kg) were prepared for recordings of arterial blood pressure, ECG, and electrocortical activity. Seizure activity was induced by PTZ (100 mg/kg i.v.). Sixty seconds after the onset of seizure activity, the animals received either no drug (control) or propranolol (2.5 mg/kg i.v.). A transient increase in the mean arterial blood pressure was observed after PTZ administration. Intravenous propranolol signiἀcantly suppressed the seizure duration (second per minute interval) at 1 min after drug administration; seizure duration control, 36.3 ± 4.8; i.v. propranolol, 12.3 ± 5.1. Intravenous propranolol also produced a maximal decrease of 32–38% in the basal heart rate and reduced the transient increase in mean arterial blood pressure elicited by PTZ, with no signiἀcant effect on the basal mean arterial blood pressure. Plasma propranolol levels were found to be 6.07 ± 1.43 μg/mL at 1 min after administration, falling to 1.10 ± 0.27 μg/mg over the following 19 min of the experiment. The data demonstrate that propranolol possesses anticonvulsant activity against PTZ-induced seizures in both the pig and in the cat.

Acknowledgments The study was funded by a grant from the Epilepsy Foundation of America and from the Ben Franklin Partnership Fund, a program of the Commonwealth of Pennsylvania. The authors would like to thank Valerie Farris, Larry Pratt, and Michele Spino for technical help and also for typing the manuscript, and Dr. Edward Gracely for statistical analyses.

References Albani, F., R. Riva, and A. Baruzzi. 1982. Simple and rapid determination of propranolol and its active metabolite, 4-hydroxypropranolol, in human plasma by liquid chromatography with fluorescence detection. J Chromatogr 228: 362–365. Bose, B. C., A. Q. Saiἀ, and S. Sharma. 1963. Studies on anticonvulsant and antiἀbrillatory drugs. Arch Int Pharmacodyn 146: 106–113. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 20: 225–240. Conway, J., D. T. Greenwood, and D. N. Middlemiss. 1978. Central nervous actions of β-adrenoceptor antagonists. Clin Sci Mol Med 54: 119–124. Dailey, J. W., and P. C. Jobe. 1986. Indices of noradrenergic function in the central nervous system of seizure-naive genetically epilepsy-prone rats. Epilepsy 27: 665–670. Dashputra, P. G., V. P. Patki, and T. J. Hemnani. 1985. Antiepileptic action of beta-adrenergic blocking drugs: Pronethal and propranolol. Mater Med Pol 2: 88–92.

Antiepileptic Activity of Beta-Blocking Agents

565

Faingold, C. L., and C. A. Berry. 1973. Quantitative evaluation of the pentyl-enetetrazol-anticonvulsant interaction on the EEG of the cat. Eur J Pharmacol 24: 381–388. Franz, D. N. 1980. The Pharmacological Basis of Therapeutics. New York, NY: Macmillan. Gillis, R. A. 1969. Cardiac sympathetic nerve activity: Changes induced by ouabain and propranolol. Science 166: 508–510. Goldman, B. D., A. Z. Stauffer, and C. M. Lathers. 1987. Beta blocking agents and the prevention of sudden unexpected death in the epileptic person: Possible mechanisms. Fed Proc 46: 705. Gremels, H. 1931. Uber die Einwirkung einiger zentral-erregender Mittel auf Atmung und Kreislauf. Arch Exp Pathol Pharmacol 162: 29–45. Hahn, F. 1960. Analeptics. Pharmacol Rev 12: 447–530. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hildebrandt, E. 1937. Pentamethylenetetrazol (Cardiazol). Handb Exp Pharmacol 5: 151–183. Hirsch, C. S., and D. L. Martin. 1971. Unexpected death in young epileptics. Neurology 21: 682–690. Jaeger, V., B. Esplin, and R. Capek. 1979. The anticonvulsant effects of propranolol and beta-Â�adrenergic blockade. Experientia 35: 8081. Krall, R. L., J. K. Perry, B. G. White, H. J. Kupferberg, and E. A. Swinyard. 1978. Antiepileptic drug development: II. Anticonvulsant drug screening. Epilepsia 19: 409–428. Kraras, C. M., N. Turner, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23: 19–31. Langeluddeke, A. 1936. Die diagnostische Bedeutung experimentell erzeugter Krampfe. Dtsch Med Wehnschr 62: 1588–1590. Lathers, C. M. 1980. Effect of timolol on autonomic neural discharge associated with ouabainÂ�induced arrhythmia. Eur J Pharmol 64: 95–106. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and W. H. Spivey. 1987. The effect of timolol, metoprolol, and practolol on postganglionic cardiac neural discharge associated with acute coronary occlusion-induced arrhythmia. J Clin Pharmacol 27: 582–592. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge. Circulation 57: 1058–1064. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34: 1919–1936. Lathers, C. M., N. Turner, and C. M. Kraras. 1985. Cardiovascular and epileptogenic effects of pentÂ� ylenetetrazol administered intracerebroventricularly in cats. Epilepsia 26: 520. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Turner, and R. M. Levin. 1986. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39: 2121–2141. Lathers, C. M., K. Jim, and W. H. Spivey. 1989a. A comparison of intraosseous and intravenous routes of administration for antiseizure agents. Epilepsia 30: 472–479. Lathers, C. M., A. Z. Stauffer, N. Turner, C. M. Kraras, and B. D. Goldman. 1989b. Anticonvulsant and antiarrhythmic actions of the beta blocking agent timolol. Epilepsy Res 4: 42–54. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 1: 84–88. Lints, C. E., and C. Nyquist-Battie. 1985. A possible role for beta-adrenergic receptors in the expression of audiogenic seizures. Pharmacol Biochem Behav 22: 711–716.

566 Sudden Death in Epilepsy: Forensic and Clinical Issues Louis, W. J., J. Papanicolaou, R. J. Summers, and F. J. E. Vajda. 1982. Role of central beta adrenoceptors in the control of pentylenetetrazol-induced convulsions in rats. Br J Pharmacol 75: 441–446. Lown, B., and R. L. Verrier. 1978. Neural factors and sudden death. In Perspectives in Cardiovascular Research, vol. 2, Neural Mechanisms in Cardiac Arrhythmias, ed. P. J. Schwartz, A. M. Brown, A. Malliani, and A. Zanchetti, 87–88. New York, NY: Raven Press. Mueller, A. L., and T. V. Dunwiddie. 1983. Anticonvulsant and proconvulsant actions of alpha- and beta-noradrenergic agonists on epileptiform activity in rat hippocampus in vitro. Epilepsia 24: 51–64. Mueller, A. L., B. J. Hoffer, and T. V. Dunwiddie. 1981. Noradrenergic responses in rat hippocampus: Evidence for mediation by alpha and beta receptors in the in vitro slice. Brain Res 214: 113–126. Murmann, W., L. Almirante, and M. Saccani-Gueli. 1966. Central nervous system effects of four beta-adrenergic receptor blocking agents. J Pharm Pharmacol 18: 317–318. Onuma, T. 1957. Relationships of the predisposition to convulsions with the action potentials of the autonomic nerves and the brain: II. Changes in action potential of the autonomic nerves and the brain under conditions for increasing the predisposition to convulsions. Tohoku J Exp Med 65: 121–129. Orihara, O. 1952. Comparative observations of the action potential of autonomic nerve with EEC. Tohoku J Exp Med 57: 43–54. Papanicolaou, J., F. J. Vajda, R. J. Summers, and W. J. Louis. 1982. Role of beta-adrenoreceptors in the anticonvulsant effect of propranolol on leptazol-induced convulsions in rats. J Pharm Pharmacol 34: 124–125. Patil, P. N., A. Tye, C. May, S. Hetey, and S. Miyagi. 1968. Steric aspects of adrenergic drugs: XI. Interactions of dibenamine and beta adrenergic blockers. J Pharmacol Exp Ther 163: 309–319. Randall, W. C., J. X. Thomas, D. E. Euler, and G. J. Rozanski. 1978. Cardiac dysrhythmias associated with autonomic nervous system imbalance in the conscious dog. In Perspectives in Cardiovascular Research, vol. 2, Neural Mechanisms in Cardiac Arrhythmias, ed. P. J. Schwartz, A. M. Brown, A. Malliani, and A. Zanchetti, 123–138. New York, NY: Raven Press. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Snider, R. S., and W. T. Neimer. 1970. A Stereotaxic Atlas of the Brain. Chicago, IL: University of Chicago Press. Spivey, W. H., H. D. Unger, C. M. Lathers, and R. M. McNamara. 1987. Intraosseous diazepam suppression of pentylenetetrazol-induced epileptogenic activity in pigs. Ann. Emergency Med. 16: 156–159. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15: 15–30. Swinyard, E. A. 1972. Assay of antiepileptic drug activity in experimental animals: Standard tests. In Anticonvulsant Drugs, International Encyclopedia of Pharmacology and Therapeutics, Section 19.1, 47–65. Oxford: Pergamon Press. Tackett, R. L., J. G. Webb, and P. J. Privitera. 1981. Cerebroventricular propranolol elevates cerebrospinal fluid norepinephrine and lowers blood pressure. Science 213: 911–913. Tocco, D. J., B. V. Clineschmidt, A. E. W. Duncan, F. A. Deluna, and J. R. Baer. 1980. Uptake of the beta-adrenergic blocking agents propranolol and timolol by rodent brains: Relationship to central pharmacological action. J Cardiovasc Res 2: 133–143. Toman, J. E. P., and J. P. Davis. 1949. The effects of drugs upon the electrical activity of the brain. Pharmacol Rev 1: 425–492. Van Buren, J. M. 1958. Some autonomic concomitants of ictal autonomism. Brain 81: 505–528. Van Buren, J. M., and C. Ajmone-Marsan. 1960. Correlations of autonomic and EEG components in temporal lobe epilepsy. Arch Neurol 3: 683–703.

Arrhythmias Associated with Epileptogenic Activity Elicited by Penicillin

35

Claire M. Lathers Paul L. Schraeder

Contents 35.1 Introduction 35.2 Method 35.3 Results 35.4 Discussion 35.5 Summary Acknowledgments References

567 567 568 571 574 574 574

35.1â•…Introduction The association of autonomic dysfunction, clinical epilepsy, and sudden unexplained death has been the subject of many studies (Leestma et al. 1984; Terrence et al. 1975). Lathers and Schraeder (1982) and Schraeder and Lathers (1983) observed autonomic dysfunction in cats after epileptogenic activity induced by pentylenetetrazol. A marked increase in variability in mean autonomic cardiac sympathetic and parasympathetic neural discharge was associated with the epileptogenic activity. It was hypothesized that if altered cardiac neural discharge also occurs in the patient with epilepsy, cardiac arrhythmias and sudden unexplained death may occur. The ideal agent to prevent these events should possess anticonvulsant, antiarrhythmic, and cardiac neural depressant properties. This study developed a new small-animal model to study autonomic dysfunction in association with epileptogenic activity produced by injecting penicillin into the hippocampus of the cat. Epileptogenic activity was monitored as it spread to the left and right hippocampi and cerebral cortices. Data were analyzed to determine whether changes in the autonomic parameters of mean arterial blood pressure and heart rate were associated with both the interictal and the ictal epileptogenic activity. Phenobarbital was administered to determine whether it suppressed the epileptogenic and arrhythmic activities.

35.2â•…Method The stereotaxic hippocampal injection of aqueous penicillin solution in 11 cats anesthetized with general anesthesia elicited both interictal and ictal activities. Cats were anesthetized intravenously (i.v.) with alpha-chloralose (80 mg/kg) and surgically prepared for 567

568 Sudden Death in Epilepsy: Forensic and Clinical Issues

monitoring the mean arterial blood pressure, lead II ECG, and for drug administration as described by Lathers and Schraeder (1982). A burr hole was made in the region of the posterior sylvian and posterior ectosylvian gyri bilaterally after the animal was placed into a stereotaxic head holder (David Kopf Instruments). A microcannula and a concentric bipolar recording electrode were inserted into the hippocampus using coordinates obtained from a stereotaxic atlas of the cat brain (Snider and Niemer 1961). Electrocorticographic recording electrodes were placed on the left and right (motor) cortices and the hippocampi. Penicillin was injected into the right hippocampus (coordinates A +7, HD −6.0, and RL +11.8). Motor cortex activity was recorded because of evidence that the frontal cortex is involved in cardiovascular regulation (Yingling and Skinner 1976). Epileptogenic activity, interictal and ictal spikes, was elicited by the right hippocampal injection of penicillin as an aqueous solution of 400,000 U/mL colored with methylene blue to verify postmortem the injection recording site. A microsyringe in stereotaxic carrier was used to inject 0.0025 mL penicillin (1000 U). The epileptogenic activity, quantiἀed in spikes per minute, was correlated with changes in mean arterial blood pressure, heart rate, and ECG. Either 20 or 40 mg/kg sodium phenobarbital (Elkins Sinn, Inc.), dissolved in 5 mL physiologic (0.9%) saline, was infused into the femoral vein at a rate of 0.5 mL/min, followed by a 2-min wash at the same rate with saline. Phenobarbital was administered after the injection of penicillin into the hippocampus. One-factor repeated-measures analysis of variance were run where the independent variable was time (every 4 min) and the dependent measure was either heart rate or blood pressure. This was repeated for both penicillin and phenobarbital, creating four separate analyses. When a signiἀcant F ratio (using the Huynh–Feldt correction to degrees of freedom) was obtained, the Newman–Keuls post hoc procedure for determining which pairwise comparisons were signiἀcant was run at α = 0.05 (Winer 1962). Analyses of variance were done using biomedical programs, subprogram P2V. Post hoc tests were accomplished using the Statistical Package for the Social Sciences, subprogram one way.

35.3â•…Results Changes in the electrocardiogram that were observed after the administration of 1000 U penicillin in 11 cats included T-wave inversion in ἀve, changes in the ST and P–R intervals in four and three, respectively, alterations in the QRS complex in four, and ST depression in two. The administration of phenobarbital 20 mg/kg (i.v.) to four additional cats or 40 mg/kg (i.v.) to nine additional cats abolished the penicillin-induced changes in the ECG. The ECG changes included alterations in the P and QRS waves, the appearance of a U wave, and premature ventricular contractions. At some experimental times, the premature ventricular contractions occurred before the appearance of the epileptogenic activity; at other times in the same cat, the premature ventricular contractions appeared just after the initiation of the epileptogenic activity. The changes in the ECG were not caused by anesthesia or by surgical stress because they were not observed in the control period. In the rare event that anesthesia initiated arrhythmias, the cat was not included in the study. Data from one cat are depicted in Figure 35.1. Penicillin (1000 U/μL) induced interictal activity in the left motor cortex and in the left and right hippocampi 1 min after administration (not shown). Ictal activity developed in the right hippocampus 5 min after the injection of penicillin and was associated with a 5-mm Hg increase in the mean arterial

Electroencephalogram (µV)

Mean arterial blood pressure (mm Hg)

0

100

200

6

5

1s Control

176

78

R. hippocampus

4

R. motor cortex

3

L. hippocampus

2

L. motor cortex

1

1s 5 min after penicillin injection

176

80

(b)

30 min after penicillin injection

246

133

(c)

115

1 h after penicillin injection

6 242

5

R. hippocampus

R. motor cortex

3

L. hippocampus

L. motor cortex 2

1

(d)

10 min after phenobarbital infusion

138

50

(e)

Figure 35.1╇ Changes in the electroencephalogram and cardiovascular function observed in one cat after the administration of penicillin (100 U/ μL) into the right hippocampus. Phenobarbital (40 mg/kg, i.v.) was administered 2 h after the administration of penicillin. (a) Left motor cortex; (b) left hippocampus; (c) right motor cortex; (d) right hippocampus; (e) blood pressure; and row 6, electrocardiogram. The numbers above the mean arterial blood pressure and electrocardiogram tracings indicate the blood pressure and heart rate values in millimeters of mercury and beats per minute, respectively. Labels on the panels indicate data obtained in the control period and the times at which the data were obtained after the administration of penicillin and phenobarbital 40 mg/kg (i.v.).

Electrocardiogram (lead II)

(a)

Arrhythmias Associated with Epileptogenic Activity Elicited by Penicillin 569

570 Sudden Death in Epilepsy: Forensic and Clinical Issues

blood pressure (column B). Interictal activity was evident in all four electrocorticograms at minute 12 (not shown). Thirty minutes after the administration of penicillin (column C), the mean arterial blood pressure increased to 133 mm Hg and the heart rate to 246 bpm. Interictal spikes were observed in both motor cortices and ictal discharge in both hippocampi 1 h (column D) and 2 h (not shown) after penicillin administration. Mean arterial blood pressure and heart rate values were still increased from control. Penicillin-induced interictal and ictal activities were suppressed at minute 5 of the infusion of phenobarbital (not shown) and were abolished at minute 10 after 40 mg/kg (i.v.) phenobarbital (column E). Penicillin-induced increases in blood pressure and heart rate were also reversed to values lower than the control. Blood pressure and heart rate values were slightly higher than the control 30 min after phenobarbital; no epileptogenic activity was apparent (not shown). The hippocampal injection of penicillin (1000 U/μL) in six cats increased the mean arterial blood pressure from a control of 92 ± 12 to 106 ± 10 mm Hg at 30 min after penicillin. Heart rate was increased from the control of 179 ± 15 to 199 ± 13 bpm at this time. The change in heart rate was signiἀcant (p < 0.05), but the change in blood pressure was not. The administration of phenobarbital (20 mg/kg i.v.) decreased mean arterial blood pressure from the pre-phenobarbital control of 115 ± 14 to 79 ± 14 mm Hg (p < 0.05) 22 min after phenobarbital. Heart rate was decreased from 204 ± 19 to 162 ± 12 bpm (p < 0.05) at this time. To determine whether a higher dose of phenobarbital (40 mg/kg, i.v.) would completely abolish the penicillin-induced epileptogenic activity, ἀve additional cats

Mean arterial blood pressure (mean ± SE; mm Hg)

160

N = 5 cats Cats

140 120 100 80 60

Heart rate (mean ± SE; beats/min)

40 0 250 200 150 100 0

0

10

20

30

Penicillin (1000 U, hippocampal)

40

50

60

70

80

90

100 110 120 Time (min)

Figure 35.2╇ Effect of hippocampal injection of penicillin (1000 U/μL) on mean arterial blood pressure (mm Hg) and heart rate (bpm). Data are graphed as a function of time in minutes and are expressed as the mean ± SE for another group of five cats. Asterisks indicate values that are significantly different from control.

Arrhythmias Associated with Epileptogenic Activity Elicited by Penicillin

571

received hippocampal injections of penicillin (1000 U/μL). The effect of penicillin on the mean arterial blood pressure and heart rate in these ἀve additional cats is depicted in Figure 35.2. Mean arterial blood pressure increased from the control of 105 ± 5 to 143 ± 13 mm Hg 52 min after the administration of penicillin; heart rate increased from 170 ± 22 to 218 ± 16 bpm (p < 0.05) 34 min after penicillin. When phenobarbital (40 mg/kg, i.v.) was administered to these ἀve cats, it decreased mean arterial blood pressure from the prephenobarbital control of 128 ± 14 to 56 ± 28 mm Hg and heart rate from 201 ± 15 to 117 ± 14 bpm (p < 0.05) at 12 and 22 min, respectively (Figure 35.3). Comparison of the effect of 20 and 40 mg/kg (i.v.) phenobarbital revealed that the magnitude of the decrease in blood pressure and heart rate after the larger dose was twice that of the lower dose.

35.4â•…Discussion This study showed that penicillin-induced hippocampal epileptogenic discharges were associated with increases in mean arterial blood pressure and heart rate and changes in the ECG. The dose of 20 mg/kg (i.v.) phenobarbital reversed the associated increases in blood 160 N = 5 cats

Mean arterial blood pressure (Mean ± SE; mm Hg)

140 120 100 80 60 40

Heart rate (mean ± SE; beats/min)

0 250 200 150 100 0

0

10

Phenobarbital (40 mg/kg; i.v.)

20 30 Time (min)

Figure 35.3╇ Effect of phenobarbital (40 mg/kg, i.v.) on penicillin-induced changes in blood

pressure (mm Hg) and heart rate (bpm). Data are graphed as a function of time in minutes and are expressed as the mean ± SE and were obtained in the same five cats depicted in Figure 35.2. Phenobarbital was given 2 h after the administration of penicillin. Asterisks indicate values that are significantly different from control.

572 Sudden Death in Epilepsy: Forensic and Clinical Issues

pressure and heart rate and attenuated the epileptogenic activity; 40 mg/kg (i.v.) phenobarbital completely abolished epileptogenic activity and reversed the associated changes in the cardiovascular parameters. This study demonstrated that the use of general anesthesia with hippocampal injections of aqueous penicillin elicited both interictal and ictal spike activities, in contrast to intracortical penicillin into the cerebral convexity, which does not usually progress to ictal activity in cats under general anesthesia (Prince 1972). Several of our preliminary experiments (unpublished observations), using alpha-chloralose anesthesia and cortical injection of penicillin, elicited only interictal spike activity in the primary focus and in the minor focus, that is, the homologous contralateral cortical area, within 3 to 10 min. Afterdischarges did not develop for as long as 7 h after the intracortical injection of penicillin. In contrast, when ether was used for induction of anesthesia, followed by local anesthesia, afterdischarges occurred within a mean time of 105 min after penicillin injection (Schraeder and Celesia 1977). Thus, the use of alpha-chloralose and cortical convexity injection of penicillin reliably elicits only interictal spikes and therefore is useful only in the examination of the effect of interictal discharge on autonomic cardiovascular function. In contrast, this study found that injection of penicillin into the hippocampus in animals anesthetized only with alpha-chloralose resulted in the progressive development of interictal to ictal discharges. Thus, the use of general anesthesia in the model using hippocampal injection of penicillin does not preclude investigation into the effects of ictal discharge. The signiἀcance of the data reported in this study is that interneuronal pathways connect the hippocampal area to the autonomic cardiovascular areas within the hypothalamus. The hippocampal formation, the amygdaloid complex, the septal region, the gyrus fornicatus, the piriform lobe, and the caudal orbital frontal cortex constitute the limbic forebrain structures (Nauta and Haymaker 1969). The fornix system forms the main efferent pathways from the hippocampus. Fiber systems originating in the limbic forebrain are among the most conspicuous afferents of the hypothalamus. Hippocampal afferents come from the medial septal nucleus and cingulate and the parahippocampal regions of the gyrus fornicatus. The amygdaloid complex is connected with the hypothalamus by the stria terminalis and the ventral amygdalofugal pathway. The septoamygdalar complex projects directly to the hippocampus (Swanson and Cowan 1979). The hypothalamus, at least in part, is then under cerebral cortical control in its influence on the maintenance of homeostasis by virtue of its neural relationships with both divisions of the autonomic nervous system and with both lobes of the pituitary gland. “When the connections of the septohippocampal complex are considered as a whole, the conclusion emerges that it essentially forms the gateway between the hypothalamus and the limbic cortical regions” (Swanson 1983). It is quite possible that with spread of interictal activity in the hippocampus to the hypothalamus, the subclinical epileptogenic activity alters the function of other areas of the brain, with resultant simultaneous changes in the autonomic control of mean arterial blood pressure, heart rate and rhythm, and cardiac neural discharge in the periphery. Furthermore, cardiovascular regulation is a function of neuronal activity in the cerebral cortex, the amygdala, and the medullary reticular formations. Cardioacceleratory and cardioinhibitory centers exist at these levels of the nervous system, with selective activation producing either increased or decreased heart rate. Vasopressor and vasodepressor centers also exist at these central sites and produce their effects through reticulospinal connections to the preganglionic sympathetic neurons of the intermediolateral cell column and through connections to preganglionic parasympathetic neurons. In addition to

Arrhythmias Associated with Epileptogenic Activity Elicited by Penicillin

573

descending input from higher centers, the cardiovascular centers also receive input from peripheral receptors, the most important of which are the baroreceptors of the carotid sinus and the aortic arch. Although peripheral autonomic dysfunction in cardiac autonomic nerves can precede the changes in the ECG associated with subconvulsant interictal discharge, interictal activity alone can also be associated with premature ventricular contractions (Lathers and Schraeder 1982). Thus, minimal epileptogenic activity (single spikes) can be associated with altered cardiac neural discharge and arrhythmias. Schraeder and Celesia (1977) reported that even minimal epileptogenic activity has a wide-ranging effect on cerebral functions monitored at the auditory cortex of the cat. If, then, subclinical epileptogenic activity were to likewise have an effect on other functions of the brain, that is, autonomic regulation, producing autonomic imbalance in cardiac autonomic neural discharge with subsequent arrhythmias, this type of activity could be a contributing factor to the mechanism of unexpected death in epilepsy. That this sequence could occur in the person with epilepsy is supported by the anatomical relationships among the cerebral cortex, the hippocampi, and the hypothalamus. The anatomical and physiological relationship of the frontal cortex with the hypothalamus is complex; studies stimulating the dorsolateral surface of both hemispheres have reported both a rise and a drop in blood pressure and increases or decreases in heart rate (Hoff et al. 1963). Cortical stimulation evoked dilatation of pupils, retraction of the nictitating membrane, piloerection, salivation, sweating, and gastric motility and secretion. The autonomic localization in the motor and premotor cortex corresponds closely with the somatotopic representation (Brooks and Koizumi 1974). The focal model of epilepsy used in this study produced cardiovascular changes that were similar to those produced by pentylenetetrazol-induced epileptogenic activity (Lathers and Schraeder 1982; Schraeder and Lathers 1983). These data support the hypothesis that focal epileptogenic activity can produce alterations in cardiac electrical activity making the heart susceptible to arrhythmias that may cause sudden unexplained death in epileptic persons. The clinical signiἀcance of the data obtained in this type of epileptogenic model is emphasized by Blumhardt et al. (1986). They reported that in some patients with temporal lobe epilepsy, cardiac acceleration preceded the onset of recognizable rhythmic surface EEG seizure activity; this may reflect the onset of electrical discharge in deep limbic circuits and the connections of these structures with the autonomic nervous system. Arrhythmias were observed at times when there were no seizure discharges on the EEG in some patients. Blumhardt and others also noted that the autonomic effects of temporal lobe epilepsy on heart rate and rhythm may be more severe in untreated younger patients. Their suggestion agrees with the observation that young epileptic patients are at high risk of sudden unexplained death. The present study used an animal model that allowed testing of a pharmacologic agent that possesses antiepileptic, antiarrhythmic, and neural depressant activity. Future use of this model should help in the development of better therapeutic regimens designed to eliminate the autonomic dysfunction and arrhythmias associated with epileptogenic activity and ultimately contribute to our understanding of the risk factors for sudden unexplained death in epilepsy. If the data indicate that the autonomic changes are secondary to seizures, the primary clinical therapeutic goal would be to use a pharmacologic agent with maximum anticonvulsant potency. However, if interictal activity is associated with a risk of autonomic dysfunction, questions must be raised about the current therapeutic goal for

574 Sudden Death in Epilepsy: Forensic and Clinical Issues

epilepsy, which is to suppress seizures but not interictal discharges. In the latter case, a new type of drug may be required.

35.5â•…Summary Penicillin-induced epileptogenic activity (1000 U/mL) was recorded bilaterally from the hippocampi and the motor cortices of 11 anesthetized cats. The onset of epileptogenic activity ranged from 1 s to 16 min. Epileptiform activity, consisting of interictal discharges (n = 3) or ictal discharges (n = 3), ἀrst occurred at the injection site, the right hippocampus. Blood pressure increased from 92 ± 12 (control) to 106 ± 10 at 30 min and 115 ± 10 mm Hg at 60 min after penicillin (p > 0.05). Heart rate increased from 179 ± 15 (control) to 194 ± 13 at 30 min and 216 ± 13 bpm 60 min after penicillin (p > 0.05). Maximum increases in blood pressure and heart rate were 55 ± 15 mm Hg and 59 ± 15 bpm, respectively (p < 0.05). ECG alterations included P–R interval changes, increased P-wave amplitude, QRS complex changes, T-wave inversion, ST elevation, and the appearance of premature ventricular contractions. Phenobarbital (20, mg/kg i.v.) diminished the epileptogenic activity and depressed the blood pressure to 79 ± 14 mm Hg at 23 min from 115 ± 14 mm Hg (10 min before phenobarbital; p < 0.05). Heart rate was decreased to 162 ± 12 from the pre-phenobarbital control of 204 ± 19 bpm (p > 0.05). To determine whether a higher dose of phenobarbital (40 mg/kg, i.v.) would completely abolish the penicillin-induced epileptogenic activity, ἀve additional cats received 1000 U/μL penicillin G sodium into the right hippocampus. In these cats the penicillin also produced epileptogenic activity and increased the blood pressure from 105 ± 5 to 143 ± 13 and the heart rate from the control 170 ± 22 to 218 ± 16 (p < 0.05). Phenobarbital (40 mg/kg, i.v.) signiἀcantly reversed the effect of penicillin on the blood pressure and heart rate. Blood pressure dropped from the pre-phenobarbital control of 128 ± 14 to 56 ± 18 mm Hg and heart rate dropped from 201 ± 15 to 117 ± 14 bpm (p < 0.05). This dose of phenobarbital also prevented the penicillininduced epileptogenic activity. Thus, phenobarbital diminished the epileptogenic activity and autonomic dysfunction induced by penicillin. The autonomic dysfunction and epileptogenic activity induced by the peripheral intravenous administration of pentylenetetrazol (Lathers and Schraeder 1982) are similar to those induced by the hippocampal injection of penicillin.

Acknowledgments This study was funded by the Epilepsy Foundation of America. The authors are indebted to Dr. Nihal Tumer, Valerie Farris, and Larry Pratt for technical help, Dr. Edward Gracely for statistical analyses, and Michele Spino for typing the manuscript.

References Blumhardt, L. D., P. E. M. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1: 1051–1056. Brooks, C. M., and K. Koizumi. 1974. The hypothalamus and control of integrative processes. In Medical Physiology, 13th ed., ed. V. B. Mountcastle, 813–836. St. Louis, MO: C. V. Mosby.

Arrhythmias Associated with Epileptogenic Activity Elicited by Penicillin

575

Hoff, E. C., J. F. Kell Jr., and M. N. Carrol Jr. 1963. Effects of cortical stimulation and lesions on cardiovascular function. Physiol Rev 43: 68–114. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Nauta, W. J. H., and W. Haymaker. 1969. Hypothalamic nuclei and ἀber connections. In The Hypothalamus, ed. W. Haymaker, E. Anderson, and W. J. H. Nauta, 136–209. Springἀeld, IL: Thomas. Prince, D. A. (1972). Topical convulsion drugs and metabolic antagonists. In Experimental Models of Epilepsy. A Manual for the Laboratory Worker, ed. D. P. Purpura, J. K. Peney, D. Tower, D. M. Woodbury, and R. Walter, 51–84. New York, NY: Raven Press. Schraeder, P. L., and G. G. Celesia. 1977. The effects of epileptogenic activity on auditory evoked potentials in cats. Arch Neurol 34: 677–682. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Snider, R. S., and W. T. Niemer. 1961. A Stereotaxic Atlas of the Cat Brain. Chicago, IL: University of Chicago Press. Swanson, L. W. 1983. The hippocampus and the concept of the limbic system. In Neurobiology of the Hippocampus, ed. W. Seifert, 3–20. London: Academic Press. Swanson, L. W., and W. M. Cowan. 1979. The connections of the septal region in the rat. J Comp Neurol 186: 621–656. Terrence, C. F., H. M. Wisotzkey, and J. A. Perper. 1975. Neurogenic pulmonary edema in unexpected, unexplained death in epileptic patients. Neurology 25: 594–598. Winer, B. 1962. Statistical Principles in Experimental Design, 2nd ed. New York, NY: McGraw-Hill. Yingling, C. D., and J. E. Skinner. 1976. Selective regulation of thalamic sensory relay nuclei by nucleus reticularis thalami. Electroencephalogr Clin Neurophysiol 41: 476–482.

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias

36

Claire M. Lathers

Contents 36.1 Introduction 36.1.1 PTZ-Induced Increased Concentrations of Central Enkephalins and Epileptogenic Activity 36.1.2 Central Neuropeptide-Induced Epileptogenic Activity 36.1.3 Enkephalin Modulation of GABA Release 36.1.4 Met-Enkephalin Modulation of Acetylcholine Release 36.2 Central Cardiovascular and Neurological Effects of Enkephalins 36.2.1 Action on Mean Arterial Blood Pressure, Heart Rate, and Brain Electrical Activity in Conscious Cats 36.2.2 Action on Mean Arterial Blood Pressure, Heart Rate, and Brain Electrical Activity in Anesthetized Animals 36.3 Discussion 36.4 Summary References

577 578 578 578 578 580 580 583 585 588 589

36.1â•…Introduction Pentylenetetrazol (PTZ)-induced interictal and ictal epileptogenic activity associated with autonomic dysfunction—that is, changes in autonomic cardiac neural discharge, mean arterial blood pressure, and heart rate and rhythm—has been reported in the cat (Schraeder and Lathers 1989; Lathers and Schraeder 1982). If the autonomic dysfunction, including the development of arrhythmias, also occurs in humans, it may be a contributory factor to sudden unexplained death in a person with epilepsy. Elevation of immunoreactive (IR) met-enkephalin content in the septum, hypothalamus, amygdala, and hippocampus of rats occurs after PTZ-induced convulsions (Vindrola et al. 1984). This elevation may ultimately change central sympathetic neural discharge to the heart, resulting in the development of arrhythmias. Indeed, numerous reports indicate that neuropeptides may produce epileptic seizures (Elazor et al. 1979; Frenk et al. 1978). Resolution of the question of whether enkephalins elicit epileptogenic activity and autonomic dysfunction via an action to inhibit the release of GABA (Brennan et al. 1980; Snead and Bearden 1980) is important because an understanding of this mechanism should eventually allow the design of pharmacologic agents to prevent the epileptogenic activity and autonomic dysfunction.

577

578 Sudden Death in Epilepsy: Forensic and Clinical Issues

36.1.1â•… P  TZ-Induced Increased Concentrations of Central Enkephalins and Epileptogenic Activity PTZ-induced kindling was associated with a long-lasting elevation in brain content of IR met-enkephalin in the septum, hypothalamus, amygdala, and hippocampus of rats (Vindrola et al. 1984). Kindling was produced by the administration of intraperitoneal injections of 40 mg/kg PTZ every 24 h for 10 days. The control group received an equivalent volume of saline on the same schedule. Every animal was observed for 1 h after each injection for the appearance of convulsions. IR met-enkephalin was quantiἀed in several brain areas 16 days after the last injection of PTZ in both the control and experimental groups. Additional rats received a PTZ dose on day 16 and were sacriἀced 1 and 24 h later. Brain tissue was prepared and enkephalin content was assayed by radioimmunoassay. A long-lasting elevation in amygdala, septum, hypothalamus, and hippocampus€ IR€ metÂ�enkephalin content occurred in animals subjected to kindling and sacriἀced 16 days after the last dose of PTZ. A decrease in IR met-enkephalin occurred 1 h after the PTZ-induced seizure but increased to newly elevated levels 24 h later. Thus, PTZ-induced kindling increased levels of enkephalins that were temporally related to the appearance of seizures. 36.1.2â•… Central Neuropeptide-Induced Epileptogenic Activity In addition to the ἀndings of increased brain concentrations of enkephalins after induction of seizures by PTZ, injection of these agents has been shown to elicit seizure activity. Speciἀcally, the intracerebroventricular injection (Urca and Frenk 1983) and hippocampal injection (Elazor et al. 1979) of leu-enkephalin induced seizures in rats and cats. Snead and Bearden (1980) also found that neuropeptides administered into the central nervous system induced epileptogenic activity in rats. Intraventricular leu-enkephalin produced a consistent dramatic paroxysmal electrical response within the ἀrst 60 s of administration (Figure 36.1), which persisted for up to 6 min. The enkephalin-induced paroxysms increased the 3- to 6-Hz band of the EEG spectrum. This indicated that enkephalin is directly involved in the production of epileptogenic activity (Snead and Bearden 1980). 36.1.3â•… Enkephalin Modulation of GABA Release Met-enkephalin inhibited K+ depolarization-induced release of 3H-GABA from rat synaptosomes in a dose-dependent fashion (Brennan et al. 1980). The concentration of metenkephalin that inhibited 50% of the K+-stimulated release was approximately 5 × 10−10 M (Figure 36.2). In every instance, the reduction of GABA release was prevented by naloxone, suggesting that met-enkephalin may interact with opiate receptors to modulate the release of GABA. 36.1.4â•… Met-Enkephalin Modulation of Acetylcholine Release Met-enkephalin reversibly and speciἀcally reduces the quantal content of acetylcholine release from peripheral nerve terminals in the frog cutaneous pectoris muscle by blocking voltage-dependent Ca+ channels (Bixby and Spitzer 1983). It is likely that met-enkephalin also blocks the release of Ca2+-dependent neurotransmitters, such as GABA, from central synapses. Bixby and Spitzer applied met-enkephalin (10–30 μM) by pressure ejection

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias 579

RF-RP LF-LP

50 µV

1s

RF-RP LF-LP 50 µV

1s RF-RP LF-LP

50 µV

50 µV

1s

1s

Figure 36.1╇ ECoG changes produced by 100 μg leu-enkephalin injected intracerebroventricularly. The insets show a faster time trace. RF and RP, right frontal and parietal leads; LF and LP, left frontal and parietal leads. The rat was immobile during this paroxysmal electrical activity. The first change in the ECoG occurred within 1 min of administration and was a paroxysm of spikes at a frequency of 7–9 Hz lasting 30–40 s. This tapered off to 2–3 Hz slow-wave activity, which was followed by 20–25 s of low-voltage fast activity. A few seconds of 1-Hz high-voltage slow-wave activity then built up to 25–30 s of low-voltage activity. Finally, a prolonged period of high-voltage single spikes occurred at a rate of 1 paroxysm/5 s. (Reproduced from Snead, O.C., and Bearden, L.J., Science, 210, 1031–1033, 1980. With permission.)

% Inhibition K+-stimulated GABA release

through a “puffer” pipette to the presynaptic terminals of frog neuromuscular junctions before compound nerve stimulation with a suction electrode. Met-enkephalin was also applied in the presence of 15 μM naloxone. Application of saline to the presynaptic terminals before stimulation served as a control. Finally, puffer application of enkephalin was used on the postsynaptic membrane, followed by both iontophoretic application of acetylcholine and nerve stimulation. Application of met-enkephalin to the presynaptic membrane led to a consistent decrease in the size of the evoked response (Figure 36.3a).

100 80 60 40 20 10–13 10–12 10–11 10–10 10–9 10–8 10–7 10–6 10–5 10–4 [Met-enkephalin] M

Figure 36.2╇ A dose–response curve for the inhibition of K+ (55 mM)-induced GABA release

by met-enkephalin. Each point is the mean percentage of inhibition ± SD (n = 5). Basal release of GABA was 7651 ± 980 dpm (n = 46). (From Brennan, M., et al., Life Sci, 27, 1097–1101, 1980. With permission.)

580 Sudden Death in Epilepsy: Forensic and Clinical Issues (a)

Normal

Enkephalin

Normal

4 ms (b)

Normal

10 mV

Normal

Enkephalin

10 mV 200 ms

Figure 36.3╇ (a) Compound nerve stimulation with a suction electrode elicits an endplate

potential that is reduced by focal puffer application of 20 μM met-enkephalin. Saline contained 0.8 mM Ca2+ and 8.0 mM Mg 2+. (b) Responses to iontophoretically applied acetylcholine; the amplitude of the response is increased by focal application of 20 μM met-enkephalin. Note that the response with enkephalin is longer and larger. (Reproduced from Bixby, L., and Spitzer, C., Nature, 301, 431–432, 1983. With permission.)

The decrease, which averaged 40%, was not seen when met-enkephalin was applied in the presence of 15 μM naloxone or when normal saline substituted for enkephalin. In addition, met-enkephalin not only did not reduce but also slightly increased the size of response to iontophoretically applied acetylcholine on the postsynaptic membrane (Figure 36.3b). Thus, the opiates appear to be exerting its effect presynaptically, possibly by blocking the Ca2+ channels. When these channels are blocked, there is decreased release of acetylcholine from nerve terminals in the frog cutaneous pectoris muscle (Bixby and Spitzer 1983). GABA release in the central nervous system has been reported to be Ca2+-dependent (DeBelleroche and Bradford 1972). Therefore, met-enkephalin may be able to inhibit GABA release in the central nervous system by preventing Ca2+ influx into the presynaptic nerve terminal in a manner similar to that by which met-enkephalin reduces the presynaptic release of acetylcholine.

36.2â•… Central Cardiovascular and Neurological Effects of Enkephalins 36.2.1â•…Action on Mean Arterial Blood Pressure, Heart Rate, and Brain Electrical Activity in Conscious Cats The administration of neuropeptides elicits cardiovascular changes as well as epileptogenic activity in conscious male cats (Schaz et al. 1980). After intracerebroventricular application of [(d-Ala 2) methionine-enkephalinamide (DAME)] at a dose of 425 nmol in the cat, arterial systolic and diastolic blood pressures increased, suggesting that (d-Ala2)met-enk may produce a centrally mediated vasopressor response (Figure 36.4). A small increase in heart rate occurred. The maximal cardiovascular response was seen 16 min after the

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias 581 (-Ala2)-Met-Enkephalin 425 nMol i.c.v. N = 4, n = 9 BP sys BP dio HR

16

32

64

178

40

40

30

30

20

20

10

10

∆ HR%

∆ Bp%

8

Time after injection (min)

Figure 36.4╇ Changes in blood pressure and heart rate (given as percent change of the corresponding control values) elicited by intracerebroventricular injection of DAME, 425 nmol, in freely moving cats. Data are expressed as means ± SEM and have been obtained at different time intervals after the injections of (d-Ala 2)met-enk. Nine experiments were performed in four cats. In control experiments, using intracerebroventricular injections of the same volumes of 0.9% NaCl, changes in arterial blood pressure did not exceed 5–10 mm Hg. Pretreatment values of blood pressure and heart rate were 92 ± 3 mm Hg and 145 ± 3 bpm, respectively. (Reproduced from Schaz, K., et al., Hypertension, 2 (4), 395–407, 1980. With permission.)

intracerebroventricular injection and was attenuated after 64 min. A 170-nmol dose of DAME had no effect on blood pressure and heart rate. A dose of 850 nmol increased arterial pressure and produced catatonia-like behavior during the ἀrst 30 min; this was followed by an excitatory behavior that lasted up to 2 h. Spike-wave complexes occurred within the amygdala and hippocampus after 850 nmol. The time course of electrical activity changes did not exactly follow the hemodynamic changes (Figure 36.5). Spike-wave complexes appeared 5 min after the changes in hemodynamic parameters and lasted only 40 min (Schaz et al. 1980). Yukimura et al. (1981) studied the intracerebroventricular injection of 5, 10, 25, 50, and 100 nmol DAME in cats. In additional experiments, 500 nmol naloxone was injected intracerebroventricularly 3 min before DAME to test the effects of the opioid antagonist on enkephalin action. DAME induced dose-dependent increases in systolic blood pressure and heart rate. Doses of 50 nmol or less produced a maximal increase in blood pressure within 15 min; heart rate did not reach maximum until 30 min. For more than 2 h after 50 and 100 nmol of DAME, sharp waves were seen in the hippocampus recording; theta activity was attenuated. Seizures were not observed. Naloxone given before the injection

582 Sudden Death in Epilepsy: Forensic and Clinical Issues

Am cent

100 µV

Hyphotal

100 µV

Hipp.

500 µV

Sups Gyr lat 200 µV HR beats/min BP mm Hg

240 180 150 50

Am cent

100 µV

Hyphotal

100 µV

Hipp.

500 µV

C

i.c.v. injection (-Ala2)-Met-Enkepahlin 850 nMol

Sups Gyr lat 200 µV

BP mm Hg

240 180 120 150 50

Am cent

100 µV

Hyphotal

100 µV

Hipp.

500 µV

HR beats/min

5’

10’

30’

60’

Sups Gyr lat 200 µV HR beats/min BP mm Hg

240 180 120 150 50

Figure 36.5╇ Recordings of the electrical activity of the central amygdala (Am. cent.), hypo-

thalamus (Hypothal.), hippocampus (Hipp.), and lateral suprasylvian gyrus (Subs. Gyr. lat.); of heart rate (HR), instantaneously recorded as intervals between two heartbeats; and of arterial blood pressure (BP) before, immediately after, and 5, 10, 30, and 60 min after intracerebroventricular injection of (d-Ala 2)met-enk 850 nmol in a freely moving cat. The paper speed can be seen by the continuous marks on the top of each panel: each point represents 1 s. Five minutes (5′) after application of the peptide, there is an increase in arterial pressure by approximately 20 mm Hg. At 10′, arterial pressure is markedly increased by 30 mm Hg, and in the subcortical recordings, there are hypersynchronous waves and spike-wave complexes. At 60 min after intracerebroventricular application of (d-Ala 2)met-enk, electrical recordings are not different from the control period; however, arterial blood pressure is still elevated by 35 mm Hg. (Reproduced from Schaz,€K., et al., Hypertension, 2 (4), 395–407, 1980. With permission.)

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias 583 Table 36.1â•… Effects of DAME on Blood Pressure, Heart Rate, and Baroreceptor Reflex Sensitivity in Conscious Catsa Treatment

Control

15 min

30 min

60 min

25 nmol DAME Naloxone + 25 nmol DAME 50 nmol DAME Naloxone + 50 nmol DAME

133 ± 8 133 ± 11 143 ± 9 136 ± 9

Systolic Blood Pressure (mm Hg) 154 ± 14* 149 ± 13 144 ± 11 124 ± 17 133 ± 18 118 ± 14 168 ± 15* 163 ± 12 148 ± 11 150 ± 10 153 ± 9 148 ± 21

134 ± 13 125 ± 18 129 ± 7 146 ± 25

25 nmol DAME Naloxone + 25 nmol DAME 50 nmol DAME Naloxone + 50 nmol DAME

154 ± 12 149 ± 24 147 ± 6 144 ± 5

Heart rate (bpm) 188 ± 13* 182 ± 12* 164 ± 21 158 ± 20 166 ± 8* 193 ± 9* 180 ± 20* 193 ± 29*

146 ± 12 172 ± 23 142 ± 6 149 ± 10

164 ± 10 158 ± 27 160 ± 9 146 ± 9

120 min

Source: Yukimura, J., et al., Hypertension, 3 (5), 528–533, 1981. Data are expressed as means ± SEM. *p < 0.05. a

of 25 nmol of DAME blocked all cardiovascular responses (Table 36.1). A 50-nmol dose blocked only the blood pressure responses; the heart rate increases and baroreceptor reflex attenuations were unaltered. The baroreceptor reflex was attenuated for 15–60 min after DAME; higher doses were effective for a longer time (Table 36.1). Naloxone administered before enkephalin injection produced no changes in central electrical activity. It was concluded that enkephalins may play a role in central mechanisms of cardiovascular control by interacting with opiate receptors in the brain (Yukimura et al. 1981). 36.2.2â•…Action on Mean Arterial Blood Pressure, Heart Rate, and Brain Electrical Activity in Anesthetized Animals The injection of DAME into the cisterna magna of anesthetized dogs induced a short period of moderate hypertension followed by a marked and prolonged decrease in blood pressure, heart rate, and splanchnic nerve discharge (Laubie et al. 1977). Intravenous (i.v.) naloxone produced a transient increase in all three parameters and antagonism of DAME, which was extended by a subsequent injection of naloxone (Figure 36.6). It was concluded that the opioid peptides may be involved in central cardiovascular control. The intracerebroventricular injection of DAME (500 μg/kg) has been shown to produce hypotension, bradycardia, and seizure activity when administered to anesthetized cats (Kraras et al. 1987; Lathers et al. 1988). In general, intravenous naloxone (100 μg/kg) reversed the effects of DAME on blood pressure and heart rate while eliminating seizure activity. Figure 36.7 illustrates the effect of DAME in one cat. Although the mean arterial blood pressure dropped from the control value of 108 to 93 mm Hg, the heart rate increased to 132 bpm 18 min after the administration of DAME compared to the control interval rate of 108 bpm (Figure 36.7b). The subsequent administration of 100 μg/kg (i.v.) naloxone to this cat then decreased the heart rate elevated by the administration of DAME; a further decrease in blood pressure occurred after naloxone. The epileptogenic activity induced by DAME was decreased but not abolished by naloxone (Figure 36.7c). DAME also produced brief ictal activity beginning several minutes postadministration in most cats. Brief ictal activity consists of repetitive bilateral bursts of polyspike activity,

584 Sudden Death in Epilepsy: Forensic and Clinical Issues

200 BP mm Hg

mean

0 200 0

20 s 220

72

180

HR beats/min splanchnic discharges 50 µV

500 ms

[-Ala2] met-enkephalin

control

500 µg/kg ic

20 min

naloxone 100 µg/kg, i.v.

Figure 36.6╇ The inhibitory effect of (d-Ala 2)met-enk (500 μg/kg) injected into the cisterna

108

200

93

86

150 100

Electrocorticogram (µV)

Mean arterial blood pressure (mm Hg)

magna on blood pressure, heart rate, and splanchnic neural discharges on a dog anesthetized with alpha-chloralose and the reversal produced by naloxone (100 μg/kg, i.v.). (Reproduced from Laubie, M., et al., Eur J Pharmacol, 46, 67–71, 1977. With permission.)

50 0

132

Heart rate (beats/min)

106

114

2 sec

0 Control

(a) T = 0 min DAME (500 µg/kg) i.c.v.

T = 18 min (b)

T = 4 min (c) post naxalone

T = 0 min naxalone (100 µg/kg) i.v.

Figure 36.7╇ Brief ictal activity produced by DAME and eliminated by naloxone in cats anesthetized with alpha-chloralose. Heart rate and blood pressure changes are also shown. (From Kraras, C. M., et al., Med Hypothesis, 23, 19–31, 1987. With permission.)

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias 585

each lasting less than 10 s, interspersed with brief periods of depression of cerebral activity. Administration of naloxone (100 μg/kg, i.v.) eliminated brief ictal activity in some cats within 4 min of its administration, whereas in other cats the seizure activity was somewhat depressed although still present. DAME-induced depression of heart rate and blood pressure was generally reversed by naloxone.

36.3â•…Discussion In the studies of Lathers and Schraeder (1982) and Schraeder and Lathers (1983), in the control period, the mean heart rate increased with a decrease in the mean arterial blood pressure in anesthetized cats. This relationship did not always occur with the development of epileptogenic activity induced by the intravenous administration of PTZ. The autonomic cardiac nerves did not always respond in a predictable manner to changes in blood pressure after the development of epileptogenic activity. In contrast, during the control period, all postganglionic cardiac sympathetic nerves exhibited an increased discharge as blood pressure dropped after the administration of a vasodilating test dose of histamine. Discharge in the parasympathetic cardiac nerves followed the changes in the mean arterial blood pressure. These relationships of cardiac neural discharges (sympathetic and parasympathetic) to blood pressure changes represent the normal physiological function (Bronk et al. 1936). With the development of interictal activity, the variability of mean neural discharge for the parasympathetic nerves began to increase, as demonstrated by an increase in the standard deviation. With greater degrees of epileptogenic activity, the standard deviation continued to increase; that is, the variability in the discharge among the parasympathetic nerves monitored became larger. A neural variability, again evidenced by a large standard deviation, also occurred in the mean postganglionic cardiac sympathetic discharge. The variability observed for the mean sympathetic discharge developed subsequent to that occurring in the parasympathetic discharge. Thus, autonomic cardiac neural dysfunction was observed within both divisions of the autonomic cardiac nervous systems and between the two divisions. The altered cardiac neural discharge was associated with minimal epileptogenic activity (i.e., interictal discharges) and the development of cardiac arrhythmias. The proposed mechanisms involved in the development of these arrhythmias and the possible role of enkephalins in the induction of sudden unexplained death in some epileptic patients are summarized in Figure 36.8 and are discussed below. The injection of enkephalins into the central nervous system elicits seizure activity (Frenk et al. 1978; Snead 1983). Intraperitoneal administration of PTZ produced increases in enkephalin content of the amygdala, striatum, and septum (Vindrola et al. 1983). An increase in the level of enkephalin in the amygdala may have initiated seizure activity because the amygdala is extensively interconnected with the hypothalamus; indeed, it is considered to have a higher-order modulating influence on the hypothalamus. Furthermore, almost any visceral or somatic activity, including cardiovascular and respiratory changes, elicited by stimulating the amygdala can also be elicited by stimulating various areas within the hypothalamus (Nolte 1981). Seizure activity originating in the amygdala may have induced changes in the discharge to the hypothalamus; disturbances in hypothalamic function may result in autonomic dysfunction. PTZ (intraperitoneal) has also been shown to induce an increase in enkephalin levels within the hypothalamus (Vindrola et al. 1984). Because the hypothalamus contains autonomic centers, it may be that the PTZ-induced increases

586 Sudden Death in Epilepsy: Forensic and Clinical Issues PTZ

Concentration of central enkephalin acting on oplate receptors K+-evoked release of GABA and/or, Ca+2 entry into presynaptic GABA nerve terminals and/or, K+ conductance in the GABA nerve terminal indirect Ca+2 entry in the same nerve terminal Anesthetized Animals

Conscious Animals

Sympathetic and parasympathetic central neuron outflow and enhancement of central reflex-induced vagal bradycardia and blood pressure

Inhibition of central GABA release epileptogenic activity

Blood pressure and heart rate due to attenuation of the vagal component of the baroreceptor reflex

Imbalance in peripheral sympathetic and parasympathetic neural discharge Arrhythmia

Sudden unexplained death in the epileptic person

Figure 36.8╇ Postulated mechanism by which central enkephalins could antagonize GABA, resulting in autonomic dysfunction, epileptogenic activity, and sudden death. (From Kraras, C.€M., et al., Med Hypothesis, 23, 19–31, 1987. With permission.)

in central enkephalin levels led to the production of epileptogenic activity and autonomic dysfunction in the experiments of Lathers and Schraeder (1982). A central mechanism by which increased concentrations of enkephalins inhibit K+-dependent GABA release may exist because met-enkephalin has been shown to inhibit the release of GABA from rat brain synaptosomes (Brennan et al. 1980). There is also evidence suggesting that increased concentrations of enkephalins directly decrease the entry of Ca2+ into the presynaptic GABA nerve terminals (Bixby and Spitzer 1983). Metenkephalin reduced the amount of acetylcholine released at the frog neuromuscular junction, most likely by blocking voltage-dependent Ca2+ channels in the presynaptic terminal. However, it is also possible that enkephalin reduced Ca2+ entry indirectly, by increasing K+ conductance in the terminal. Nevertheless, it may be that met-enkephalin acts within the central nervous system by interfering with the K+- and/or Ca2+-dependent mechanism of GABA release (Figure 36.8). Decreased GABA levels are thought to initiate epileptogenic activity (Krnjevic 1980; Ribak et al. 1979). Enkephalins may inhibit the release of GABA by acting on central opiate receptors. Inhibition of GABA release by met-enkephalin was prevented by administration of naloxone (Brennan et al. 1980). Pretreatment with intracerebroventricular naloxone also prevented DAME from inducing changes in blood pressure and heart rate as well as producing seizure activity in conscious cats (Yukimura et al. 1981). Administration of intravenous naloxone after DAME reversed the effects of DAME on heart rate and blood pressure (Laubie et al. 1977). Naloxone (i.v.) had the same action on heart rate and blood pressure and either eliminated or depressed DAME-induced seizure activity in anesthetized cats in the experiments of Kraras et al. (1987) and Lathers et al. (1988). Thus, it may be that

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias 587

enkephalins act on central opiate receptors to inhibit GABA release because the actions of these agents are blocked by opioid antagonists such as naloxone. Inhibition of GABA release in anesthetized cats produces increased sympathetic and parasympathetic neural outflow and the enhancement of central reflex–induced vagal bradycardia (DiMicco et al. 1979; Gillis et al. 1980). The resultant increased parasympathetic central outflow and enhancement of central reflex–induced vagal bradycardia via the enkephalin-induced inhibition of GABA release, as depicted in Figure 36.8, may explain the decrease in heart rate observed after the intracerebroventricular injection of DAME in the experiments of Kraras et al. (1987) and Lathers et al. (1988). The data of Gillis et al. (1980) suggested that an increase in blood pressure should occur in anesthetized cats after removal of the tonically active GABAergic system present in the brain. However, the intracerebroventricular injection of DAME in our experiments and the injection of DAME into the cisterna magna in anesthetized dogs (Laubie et al. 1977) to inhibit the release of GABA produced hypotension. We hypothesize that the unanticipated drop in blood pressure is due to the ability of the epileptogenic activity to produce autonomic dysfunction, as suggested by the studies of Lathers and Schraeder (1982) and Schraeder and Lathers (1983, 1989) and as indicated by the broken arrow in Figure 36.8. Altered central parasympathetic and sympathetic neural outflow may induce an imbalance within each division as well as an imbalance between both peripheral autonomic divisions that innervate the heart. This imbalance may result in the production of arrhythmia (Lathers et al. 1977, 1978) and/or sudden unexplained death (Carnel et al. 1985; Lathers and Schraeder 1982, 1987; Lathers et al. 1984, 1988; Suter and Lathers 1984). The cardiovascular effects of enkephalins in the central nervous system vary in a dose-dependent manner and according to the state of the animal, that is, conscious versus anesthetized. Met-enkephalin had no effect after intracerebroventricular injection in dogs anesthetized with alpha-chloralose; met-enkephalin has a half-life of several seconds and the lack of a visible effect may be due to its rapid inactivation. However, DAME, a synthetic analogue of met-enkephalin that is metabolically more stable than the natural peptide, produced prolonged hypotension and bradycardia in alpha-chloralose-Â�anesthetized dogs€when injected into the cisterna magna (Laubie et al. 1977). In alpha-Â�chloralose-Â�â•›anesthetized cats, the intracerebroventricular administration of DAME also produced a€ marked€ decrease in€blood pressure with a decrease in heart rate occurring in most animals (Kraras et al. 1987; Lathers et al. 1988). In contrast, a dose-dependent increase in blood pressure and heart rate was induced by administration of (d-Ala 2)met-enk in conscious cats (Schaz et al. 1980). Similar changes were reported by Yukimura et al. (1981) when they injected DAME intracerebroventricularly into conscious cats. Maximal increases in blood pressure occurred approximately 15 min before the greatest increase in heart rate. Blood pressure increased 20 min before the increase in heart rate occurred (Kraras et al. 1987; Lathers et al. 1988). The observation of differences in the physiological effects of DAME, depending on whether a conscious or anesthetized preparation is used, raises the question of whether the presence of the anesthetic agent alpha-chloralose explains the differences found in the two experimental models. Alpha-chloralose has been shown to be a good anesthetic agent for neurological studies because many reflexes, including the baroreceptors, are present and, in fact, enhanced (Clifford and Soma 1969). This anesthetic also causes minimal change in the amount of epinephrine present in the adrenal glands of cats and does not depress cardiac renal discharge (Clifford and Soma 1969; Cox et al. 1936). These data indicate

588 Sudden Death in Epilepsy: Forensic and Clinical Issues

that the baroreceptor mechanism was intact in the anesthetized animals receiving alphaÂ�chloralose and centrally administered DAME in the studies of Kraras et al. (1987), Lathers et al. (1988), and Laubie et al. (1977). The occurrence of epileptogenic activity in conscious cats began 5 min after the administration of (d-Ala 2)met-enk and ended before the changes in cardiovascular parameters (Schaz et al. 1980). Epileptogenic activity was also observed in anesthetized cats to which DAME was administered (Kraras et al. 1987; Lathers et al. 1988), although it began during the ἀrst several minutes after the administration of DAME and continued throughout the duration of the experiments. Because the epileptogenic activity began after the cardiovascular changes in the experiments of Schaz et al. (1980), they concluded that the epileptogenic activity was independent of the autonomic changes. However, there are studies that support the concept that autonomic cardiovascular changes may occur initially and be followed by the development of seizure activity. Indeed, seizure activity in patients was abolished when cardiac arrhythmias were eliminated with the insertion of pacemakers or with the initiation of antiarrhythmic agents (Schott et al. 1977). The possibility that cardiac arrhythmias may result in the development of seizure activity via impaired peripheral cardiac neural discharge going back to the central nervous system is depicted in Figure 36.8 by the heavy arrows beginning with arrhythmia. In addition, the studies of Lathers and Schraeder (1982) and Schraeder and Lathers (1983) found that epileptogenic activity may lead to cardiovascular dysfunction in animals and substantiate earlier observations made in humans. As early as 1941, Penἀeld and Erickson reported a patient with temporal lobe seizures and episodes of tachycardia. Additional studies by Mulder et al. (1954), Phizackerly et al. (1954), White et al. (1961), and Walsh et al. (1968) reported changes in the electrocardiogram in humans that were associated with epileptogenic activity. Thus, Figure 36.8 illustrates how epileptogenic activity may initiate an enhanced autonomic central neural outflow that impairs peripheral cardiac neural discharge and results in the production of arrhythmia. Further experiments are needed to determine whether enkephalins elicit epileptogenic activity and autonomic dysfunction in both conscious and anesthetized animal preparations. It will be important to determine whether enkephalins impair autonomic dysfunction via an inhibition of the release of GABA from the nerve terminal because delineation of this mechanism will allow the design of experiments to evaluate the ability of pharmacologic agents to prevent the enkephalin-induced epileptogenic activity and autonomic dysfunction. Suppression of progression to cardiac arrhythmias induced by the autonomic dysfunction should ultimately decrease the incidence of sudden unexplained death in the epileptic patient.

36.4â•… Summary Autonomic dysfunction, including arrhythmias, is often associated with epileptogenic activity. This study examines the potential role for enkephalins in this process. Brennan et al. (1980) reported a greater percentage of inhibition of K+-stimulated GABA release with increasing concentrations of met-enkephalin. Snead and Bearden (1980) found that leuenkephalin in the central nervous system may induce epileptogenic activity. In addition, DAME has been shown to produce a centrally mediated vasopressor response as well as attenuation of the baroreceptor reflex in conscious cats (Schaz et al. 1980) possibly leading

Role of Neuropeptides in the Production of Epileptogenic Activity and Arrhythmias 589

to autonomic imbalance. The latter may precipitate arrhythmias and be a contributor to sudden unexplained death in the epilepsy. Resolution of the question of whether enkephalins elicit epileptogenic activity and autonomic dysfunction via inhibition of GABA release is important because an understanding of this mechanism should eventually allow the design of pharmacologic agents to prevent the epileptogenic activity, autonomic dysfunction, and associated sudden death.

References Bixby, L., and C. Spitzer. 1983. Enkephalin reduces quantal content at the frog neuromuscular junction. Nature 301: 431–432. Brennan, M., R. C. Cantrill, and B. A. Wylie. 1980. Modulation of synaptosomal GABA release by enkephalin. Life Sci 27: 1097–1101. Bronk, D. W., R. Ferguson, and R. Margaria. 1936. The activity of the cardiac sympathetic center. Am J Physiol 117: 237–249. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity. Pharmacology 30: 225–240. Clifford, D. H., and L. R. Soma. 1969. Feline anesthesia. Fed Proc 28: 1479–1499. Cox, W. V., M. E. Lewiston, and H. F. Robertson. 1936. The effect of stellate ganglionectomy on the cardiac function of intact dogs (and its effect on the extent of myocardial infarction and on cardiac function following coronary artery occlusion). Am Heart J 12: 285–300. DeBelleroche, J., and J. Bradford. 1972. Metabolism of beds of mammalian cortical synaptosomes: Response to depolarizing influences. J Neurochem 19: 585–602. DiMicco, J., K. Gale, B. L. Hamilton, and R. A. Gillis. 1979. GABA receptor control of parasympathetic outflow to heart: Characterization and brainstem localization. Science 204: 1106–1109. Elazor, F., E. Motles, Y. Elv, and R. Simantov. 1979. Acute tolerance to the excitatory effect of enkephalin microinjections into hippocampus. Life Sci 24: 541–548. Frenk, H., G. Urca, and J. C. Leibeskind. 1978. Epileptic properties of leucine- and methionineÂ�enkephalin: Comparison with morphine and reversibility by naloxone. Brain Res 147: 327–337. Gillis, R. A., J. DiMicco, D. Williford, B. L. Hamilton, and K. Gale. 1980. Importance of the CNS GABAergic mechanisms in the regulation of cardiovascular function. Brain Res Bull 5 (Suppl. 2): 303–315. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of neuropeptides in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmias and sudden death in the epileptic patient? Med Hypothesis 23: 19–31. Krnjevic, K. 1980. Principles of synaptic transmission. In Advances in Neurology, vol. 27, Antiepileptic Drugs: Mechanisms of Action, ed. G. Glaser, J. Penry, and D. Woodbury, 127. New York, NY: Raven Press. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 27: 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 461–479. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmias. CirÂ� culation 57: 1058–1065.

590 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital. Life Sci 34: 1919–1936. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988. The effect of intracerebroventricular d-Ala2 methiÂ� onine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43: 2287–2298. Laubie, M., H. Schmitt, M. Vincent, and G. Remond. 1977. Central cardiovascular effects of morphinomimetic peptides in dogs. Eur J Pharmacol 46: 67–71. Mulder, D. W., D. Daly, and A. A. Bailey. 1954. Visceral epilepsy. Arch Intern Med 93: 481–493. Nolte, J. 1981. Olfactory and limbic systems. In The Human Brain: An Introduction to Its Functional Anatomy, ed. J. Lotz, 304. St. Louis, MO: C. V. Mosby. Penἀeld, W., and I. C. Erickson. 1941. Epilepsy and Cerebral Localization, 320–362. Springἀeld, IL: Thomas. Phizackerly, P. J. R., E. W. Poole, and C. W. M. Whitty. 1954. Sinoauricular heart block as an epileptic manifestation: A case report. Epilepsia 3: 89–91. Ribak, C. E., A. B. Harris, J. E. Vaughn, and E. Roberts. 1979. Inhibitory GABAergic nerve terminals decrease at sites of focal epilepsy. Science 205: 211–214. Schaz, K., G. Stock, W. Simon, K. Schlor, T. Unger, R. Rockhold, and D. Ganten. 1980. Enkephalin effects on blood pressure, heart rate and baroreceptor reflex. Hypertension 2 (4): 395–407. Schott, G. D., A. A. McLeod, and D. E. Jewitt. 1977. Cardiac arrhythmias that masquerade as epilepsy. Br Med J 1: 1454–1457. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal cardiovascular dysfunction and epileptogenic activity. Epilepsy Res 3: 55–62. Snead, O. C. 1983. Seizures induced by carbachol, morphine, leucine-enkephalin: A comparison. Ann Neurol 13: 445–451. Snead, O. C., and L. J. Bearden. 1980. Anticonvulsants speciἀc for petit mal antagonist epileptogenic effect of leucine enkephalin. Science 210: 1031–1033. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypothesis 15: 15–30. Urca, G., and H. Frenk. 1983. Intracerebral opiates block the epileptic effect of intracerebroventricular (I.C.V.) leucine-enkephalin. Brain Res 259: 103–110. Vindrola, O., M. Asai, M. Zubieta, and G. Linares. 1983. Brain content of immunoreactive (leu5) enkephalin and (met5)enkephalin after pentylenetetrazol-induced convulsions. Eur J Pharmacol 90: 85–89. Vindrola, O., M. Asai, M. Zubieta, E. Talavera, R. Rodriquez, and G. Linares. 1984. Pentylenetetrazol kindling produces a long lasting elevation of IR-met-enkephalin but not IR-leu-enkephalin in rat brain. Brain Res 297: 121–125. Walsh, G., W. Masland, and E. Goldensohn. 1968. Paroxysmal cerebral discharge associated with paroxysmal atrial tachycardia. Electroencephalogr Clin Neurophysiol 24: 187. White, P. T., P. Grant, J. Mosier, and A. Craig. 1961. Changes in cerebral dynamics associated with seizures. Neurology 11: 354–361. Yukimura, J., G. Stock, H. Stumpf, T. Unger, and D. Ganten. 1981. Effects of (d-Ala2)-methionineenkephalin in blood pressure, heart rate, and baro-receptor reflex sensitivity in conscious cats. Hypertension 3 (5): 528–533.

Sudden Epileptic Death in Experimental Animal Models Ombretta Mameli Marcello Alessandro Caria

37

Epilepsy is one of the most serious and most common neurological diseases. Although epidemiological studies provide evidence that 70–80% of patients who develop epilepsy go into remission after treatment, the remaining are often resistant to the common therapeutic treatments and continue to have seizures (Kwan and Sander 2004). In these patients, sudden unexplained epileptic death (SUDEP) is the most common cause of mortality related to seizures (Hauser et al. 1980; Neuspiel and Kuller 1985; Tennis et al. 1995; Nashef 1999; Nashef and Brown 1996; Leestma et al. 1997; Ficker et al. 1998; Annegers and Coan 1999; Sperling et al. 1999). The risk of SUDEP has been reported to be up to 24-fold greater than in the general population (Leestma et al. 1989; Nashef and Shorvon 1997; Schraeder et al. 2006), the overall incidence being 1:680, that is, 1:100 per year. Recent studies aimed at identifying the risk factors for SUDEP (Nilsson et al. 2001; Walczak et al. 2001; Opeskin and Berkovic 2003; Monte et al. 2007; Nashef et al. 2007) concluded that the most important risk factors are related to being a young adult male, having generalized tonic–clonic seizures, and lying in bed. Of great interest would be analysis of the results of autopsies that should be performed in all patients with a history of epilepsy according to a common standardized protocol for detailed macro- and microscopic analysis of the brain, the autonomic nervous system, the lungs, and the heart. Comparison of the incidence estimates for SUDEP is difficult. In fact, evidence for epileptic seizures immediately before death is reported in 24–80% of patients (Leestma et al. 1997; Langan et al. 2000). Furthermore, not all patients have postmortem examinations. Case ascertainment methods and source populations have varied, and different deἀnitions of SUDEP have been used (Tomson et al. 2005). To clarify this matter and drawing on the opinions of several specialists, SUDEP has been deἀned as a “sudden, unexpected, witnessed or unwitnessed, nontraumatic, and non drowning death in patients with epilepsy, with or without evidence of a seizure, and excluding documented status epilepticus, in which postmortem examination does not reveal a toxicological or anatomical cause of death” (Nashef 1997). Different physiopathological events may contribute to SUDEP, and its genesis is probably multifactorial and includes cardiac arrhythmias induced by epileptic seizures (Nei et al. 2000), neurogenic pulmonary edema (Smith and Matthay 1997), respiratory failure (O’Regan and Brown 2005), and asphyxia (Johnston et al. 1995). In a systematic report of patients undergoing simultaneous EEG and ECG monitoring for temporal lobe epilepsy, the most common ἀnding was sinus tachycardia, occurring in 92% of recordings (Blumhardt et al. 1986). Bradycardia was seen in only 4% of patients. Whether these arrhythmias are primary or secondary to other phenomena, including respiratory changes, is unknown. A recent review by Lathers et al. (2008) extensively analyzed the overlapping mechanisms that may enhance the risk of SUDEP in epilepsy and in cardiac disease. 591

592 Sudden Death in Epilepsy: Forensic and Clinical Issues

Cardiac arrhythmias during both seizures and interictal activity may result in heart failure caused by complete atria-ventricular conduction block (Wilder-Smith 1992). In a reported case of a patient with epilepsy who died unexpectedly while undergoing cardiac monitoring, a nonreversible malignant ventricular arrhythmia occurred, indicating that it was the likely arrhythmogenic cause of SUDEP (Dasheiff and Dickinson 1986). Respiratory complications such as obstruction of air pathways, central apnea, neurogenic pulmonary edema, and metabolic impairments are probably concurrent ἀnal events (Tomson 2000). However, the precise role of these factors in the pathogenesis of SUDEP has yet to be clariἀed despite a number of possibilities examined by epidemiological studies. A signiἀcant contribution to better understanding of the physiopathological mechanisms involved in SUDEP and experimental conἀrmation of clinical observations in human beings has been inferred from basic research studies in animals. Among the hypotheses explored is the involvement of the autonomic nervous system, as seizures may be preceded by autonomic symptoms that are also evident during seizure evolution (Venit et al. 2004; Johnson and Davidoff 1964; Lathers 1990; Lathers and Schraeder 1987; Lathers et al. 1987; Schraeder and Lathers 1983, 1989; Kalviainen et al. 1990; Toichi et al. 1998; Freeman 2006; Sathyaprabha et al. 2006). This fact may be dependent on the propagation of electric activity from the epileptic focus to the autonomic centres (Van Buren 1958). From the analysis of heart rate variability in humans, about 30 s before seizures begin, a signiἀcant reduction of parasympathetic tone along with a signiἀcant increase of sympathetic activity occurs (Novak et al. 1999). On the other hand, during epileptic convulsions in both animals (Doba et al. 1975; Benowitz et al. 1986) and humans (Smith and Matthay 1997), elevations of epinephrine and norepinephrine to potentially arrhythmogenic levels have been documented. However, although few cases of potentially fatal arrhythmias have been recorded in humans with epilepsy (Phizackerley et al. 1954; Liedholm and Gudjonsson 1992), the bulk of electrocardiographic recordings during seizures shows nothing more malignant than sinus tachycardia. In fact, some clinical studies conclude that ventricular arrhythmias are no more common in epileptic than in nonepileptic patients (Keilson et al. 1987). Early studies in a variety of species demonstrated that cardiac arrhythmias could be induced by stimulation of a number of areas in the diencephalon, mesencephalon, and medulla (Allen 1931; Dikshit 1934; Van Bogaert 1936; Boeles et al. 1957; Purpura et al. 1958; Fuster and Weinberg 1960; Weinberg and Fuster 1960; Manning and Peiss 1960; Parker et al. 1962; Ueda 1962; Attar et al. 1963; Melville et al. 1963; Hockman et al. 1966; Gunn et al. 1968; Hall et al. 1974; Lisander et al. 1975; Evans and Gillis 1974, 1978; De Riu 1983; McCown et al. 1984) and may be prevented by cooling the vagus nerve and even by ablation of the stellate ganglia (Manning and Peiss 1960). These experiments helped to clarify the autonomic and reflex mechanisms mediating the post-stimulation-induced arrhythmias elicited by brain stimulation and may explain the autonomic disturbances described in epileptic patients. Pathological studies have raised questions about arrhythmia as the cause for epileptic sudden death. Lathers and Schraeder (1982) and Schraeder and Lathers (1983) developed an experimental model of generalized epilepsy in the cat that used pentylenetetrazol to explore the hypothesis that an altered autonomic function may be one cause for unexplained sudden epileptic death. Their studies showed for the ἀrst time that altered sympathetic and parasympathetic cardiac neural discharges preceded the cardiac arrhythmias that occurred with the development of interictal activity and worsened with increasing degrees of ictal activity. An imbalance within and between sympathetic and parasympathetic cardiac

Sudden Epileptic Death in Experimental Animal Models

593

neural discharges was found to be associated with a signiἀcant disruption of the physiological relationships between heart rate and blood pressure. All animals died after 2000 mg/kg pentylenetetrazol of cardiovascular failure, asystole, or ventricular ἀbrillation. The authors concluded that these cardiac arrhythmias could be a major contributing factor in SUDEP even in those patients who showed only interictal spikes at the time of death and thus did not have overt seizure activity. Their hypothesis was consistent with that of Han and Moe (1964) who demonstrated that cardiac sympathetic disturbances secondary to direct sympathetic nerve stimulation increased temporal dispersion of recovery of ventricular excitability and led to an underlying electrical instability that predisposes the ventricular myocardium to arrhythmia. However, other studies showed that pretreatment with phenobarbital induced a delay of onset of paroxysmal activity, without, however, showing protective effects on the associated autonomic neural changes once the epileptiform discharges were established (Lathers et al. 1984; Carnel et al. 1985). Similar results have been also obtained in rats with epilepsy induced by pilocarpine (Colugnati et al. 2005). The cardiac sympathetic nerves are more often activated by epileptogenic events than are the cardiac parasympathetic nerves, with tachyarrhythmias being far more common clinically than epileptogenic bradycardia (Schernthaner et al. 1999; Leutmezer et al. 2003). Despite these observations, some evidence suggests that epileptogenic activation of the cardiac parasympathetic nerves, which is revealed by ictal bradyarrhythmias or cardiac asystole, might be involved in causing sudden death of epileptic patients (Nashef et al. 1996; Fuhr and Leppert 2000; Weinstein and Albertario 2000; Kelly et al. 2001; Seeck et al. 2001; Tinuper et al. 2001; Mondon et al. 2002). Preganglionic cardiac parasympathetic neurons are primarily located in the nucleus ambiguous and in the dorsal motor nucleus of the vagus; these neurons participate in the control of heart rate and other cardiac functions (Izzo et al. 1993; Standish et al. 1994; Standish et al. 1995; Taylor et al. 1999). Little is known about the ἀring pattern of preganglionic cardiac parasympathetic neurons during an epileptic attack. In a recent study (Wang et al. 2006), a fluorescent tracer was injected into the cardiac sac of newborn rats for retrograde labeling of the parasympathetic neurons in the nucleus ambiguous. Fluorescence-labeled NA neurons were further examined using a whole-cell patch-clamp method in medulla slices with a respiratory-like rhythm, and neurons with an inspiratory-related increase of the mixed inhibitory synaptic activity were identiἀed as preganglionic cardiac parasympathetic neurons. The authors demonstrated that blockade of the GABAergic and the glycinergic receptors in medulla slices evoked intermittent seizure-like ἀring (synchronized intensive firing of a mass of neurons) in the preganglionic cardiac parasympathetic neurons under a current-clamp conἀguration and evoked intermittent excitatory inward currents under a voltage-clamp conἀguration. Results of this study, therefore, provided new evidence that preganglionic parasympathetic neurons might ἀre in a seizure-like pattern of activity (Wang et al. 2006). Therefore, during an epileptic attack, these neurons may be responsible for neurogenic ictal bradyarrhythmias, cardiac asystole, or even sudden death in epileptic patients. Convulsive seizures triggered by maximal electroshock also induced a severe disruption of cardiac rhythm in rats. In particular, in the immediate postictal state, marked cardiac arrhythmia often appears, the duration of which relates to the seizure activity (Darbin et al. 2003). These data seem to support the hypothesis that cardiac arrhythmias may be an important risk factor for SUDEP in epileptic patients (Oppenheimer et al. 1990), although in the opinion of several clinicians, the role in human SUDEP remains uncertain (Keilson et al. 1987).

594 Sudden Death in Epilepsy: Forensic and Clinical Issues

Animal models of generalized epilepsy do not help, however, to localize the cortical structures that spread paroxysmal stimulation to forebrain areas involved with cardiovascular regulation that are able to desynchronize sympathovagal cardiac neural ἀring and possibly induce SUDEP. Furthermore, a few studies during human brain surgery have investigated the cardiovascular effect of cortical stimulation. In this regard, stimulation of the cingulate gyrus, tips of the temporal lobes, and orbitofrontal cortex induced both pressor and depressor responses (Chapman et al. 1949; Pool and Ransohoff 1949; Chapman et al. 1950; Delgado 1960), and more recently, superἀcial insular stimulation in patients undergoing surgery for intractable epilepsy-induced cardiovascular changes (Oppenheimer et al. 1991). Emotional stress has been shown to increase the frequency and the severity of ventricular ectopic beats in patients with or without ischemic cardiac disease, suggesting a role for cortical regions connected with the limbic structures (Lown et al. 1976; Lown and DeSilva 1978; Taggart et al. 1973; Lathers and Schraeder 2006). Experimental results in animals showed an area of cardiac representation within the posterior conἀnes of the rat€insular cortex (Oppenheimer and Cechetto 1990), in an area known to have profuse reciprocal connections with the limbic system (Zhang and Oppenheimer 2000). Using a novel technique of phasic microstimulation linked to the R wave of the ECG, Oppenheimer and Cechetto (1990) ἀrst€demonstrated a cardiac chronotropic map of the insula. They identiἀed sites generating pure tachycardia independent from other autonomic or respiratory effects within the rostral posterior insula. More caudal sites within this region generated bradycardia after phasic microstimulation. Moreover, prolonged phasic microstimulation within the rat insular cortex resulted in bradyarrhythmia, complete heart block, QT interval prolongation, ventricular ectopy, and asystolic death (Oppenheimer et al. 1991). These parasympathetic arrhythmias were accompanied by elevated plasma norepinephrine levels, increased sympathetic tone and myocytolisis, a form of cardiac damage of sympathetic neural origin, and subendocardial hemorrhages. On the other hand, left anterior insular damage by stroke is associated, in some patients, with signiἀcant tachyarrhythmias, and in the rat with a reduction in baroreflex sensitivity and resulting parasympathetic tone (Oppenheimer et al. 1996; Zhang et al. 1998). The rat insula receives taste information and gastrointestinal stimuli, respiratory afferents, and chemoreceptor and cardiovascular inputs organized in a viscerotopic fashion. This area has reciprocal connections with the parabrachial nucleus, the contralateral insula, adjacent cortical regions, the infralimbic cortex, the thalamus, the lateral hypothalamic area, and the amygdala. Evidence exists for lateralized effects, although nonmyelinated transcallosal pathways (both inhibitory and excitatory) linking the cardiovascular regions of the two insulae have been recently described (Zhang and Oppenheimer 2000). The left insular cortex is involved in the regulation of the vagal cardiac parasympathetic neuronal pool, and the right insular cortex regulates sympathetic neurones involved in the regulation of cardiac function and in the control of vascular resistance and blood pressure. This structure may therefore play a crucial role in brain–heart interaction and perhaps even additionally by its direct and reciprocal ipsilateral connections with the amygdala. The amygdala represents an important central cardiovascular control structure within the limbic system because it seems to provide the neural basis for cardiovascular responses to stressful stimuli (Cheung et al. 1997). In agreement with this latter consideration, preemptive low-frequency sine wave stimulation of amygdala-kindled animals induced a dramatic decrease in the incidence of stage 5 seizures in fully kindled

Sudden Epileptic Death in Experimental Animal Models

595

animals, demonstrating the possibility that low-frequency sine wave stimulation may be an effective therapy for prevention of seizures in patients with epilepsy (Goodman et al. 2005). The amygdala is a nucleus of the temporal lobe, showing complex interconnections with multiple cortical and brainstem regions involved in the production of autonomic temporal lobe epilepsy. An altered postganglionic cardiac sympathetic innervation that may increase the risk of cardiac abnormalities and/or SUDEP has been described in patients with chronic temporal epilepsy (Druschky et al. 2001). In fact, after temporal lobe epilepsy surgery, a reduction of sympathetic cardiovascular modulation has been demonstrated (Hilz et al. 2002). In a rat kindled seizure model, in which paroxysmal activity is similar to that described during secondary spontaneous seizures, amygdala involvement has been demonstrated in seizure-related autonomic disturbances that involved both branches of the autonomic nervous system (Goodman et al. 1990, 1999). Changes in heart rate and blood pressure were observed during amygdala-kindled seizures, but they were apparently independent of the kindling stimulus because they did not appear at the beginning of the kindling process. On the other hand, microdialysis experiments in conscious kindled rats showed decreases in extracellular concentrations of noradrenaline and dopamine levels in the dorsomedial nucleus of the hypothalamus (Goren et al. 2003), suggesting that autonomic changes in kindling require further studies to explain their relationship to SUDEP. Instead, after insular involvement by ischemic stroke or excitatory injury, increased immunostaining for neuropeptide Y, leucine-enkephalin, dynorphin, and neurotensin was seen in the amygdala of the damaged side (Allen et al. 1995; Cheung and Cechetto 1995; Cheung et al. 1995). Because these neurochemical changes in the amygdala were greatest 3 days after ischemic stroke and subsided by 10 days (Cheung et al. 1995), it can be hypothesized that they participate in mediating the cerebrogenic cardiovascular disturbances originating from the insula. The role of neuropeptides in the origin of arrhythmias and SUDEP in epileptic patients has been analyzed in some studies, and the results suggested that prostaglandin E2 and enkephalins may play a signiἀcant role in the genesis of autonomic dysfunction associated with seizure activity and cardiac arrhythmias (Suter and Lathers 1984; Kraras et al. 1987; Lathers et al. 1985, 1988; Lathers 1990; Schwartz and Lathers 1990). In rats, pentylenetetrazol kindling produces a long-lasting elevation of IR-Met-enkephalin in the septum, hypothalamus, amygdala, and the hippocampus after convulsions (Walczak et al. 2001) and is associated with a signiἀcant inhibition of potassium-stimulated GABA release (Brennan et al. 1980). The enkephalins may elicit epileptogenic activity and autonomic dysfunction by inhibiting GABA release. This possibility has been recently conἀrmed by a study in mice showing that a selective cyclooxygenase-2 inhibitor potentiates the anticonvulsant activity of tiagabine against pentylenetetrazol-induced convulsions (Dhir and Kulkarni 2006). Pathological studies in cases of SUDEP have raised questions about pulmonary complications, as autopsies in patients who died from epileptic sudden death revealed pulmonary edema (Terrence et al. 1981; Leestma et al. 1989; Earnest et al. 1992). In anesthetized and ventilated animals, Johnston et al. (1996) showed elevated left atrial and pulmonary vascular pressures that represent a possible mechanism for the pulmonary edema found in patients who have died from SUDEP. However, the pulmonary edema alone does not appear to be fatal because lung congestion requires some time to develop. In an interesting model of SUDEP in which pulmonary edema and sudden death were generated (Johnston et al. 1995), the authors suggested that arrhythmia was not an important

596 Sudden Death in Epilepsy: Forensic and Clinical Issues

factor in SUDEP as they demonstrated a primary role for hypoventilation in the etiology of epileptic sudden death. Their model of SUDEP used chronically monitored unanesthetized sheep with generalized tonic–clonic status epilepticus induced by bicuculline. Some animals died within 5 min after the induction of seizures, although no signiἀcant differences were found in the epileptic activity, in arrhythmias, and in plasma epinephrine and norepinephrine concentrations compared to the group living the longest. Furthermore, in no instance could the death of an animal be ascribed to a malignant rhythm. On the contrary, marked differences in ventilation were found between those convulsing animals that died suddenly and those that survived. A sudden drop in pO2, a marked elevation in pCO2, and a decline in arterial blood pH were observed in animals that died, whereas animals that survived maintained ventilation and oxygenation at levels close to baseline. Hypoventilation in these experiments, probably centrally induced and indistinguishable from apnea, was independent of preterminal arrhythmias or hypotension and was proposed as the cause of death. In agreement with this experimental ἀnding, hypoventilation and central apnea have been documented in several patients with seizures (Nashef and Shorvon 1997) or after stimulation involving various elements of the limbic system (Nelson and Ray 1968). However, in these experimental animals, respiratory function was not directly measured so that central and obstructive hypoventilation could not be distinguished. Extending their studies, Johnston et al. (1997) monitored awake epileptic sheep with tracheostomies that allowed measurements of airway flow and prevented upper airway obstruction. These experiments proved that central apnea and hypoventilation were causes of death in the sheep. Endogenous opioids that may be released during seizures have been implicated in the pathogenesis of this central hypoventilation (Ramabadran and Bansinath 1990) and could well account for the suppression of respiratory drive, although why this should happen is unclear (Darnell and Jay 1982). Changes in respiratory-modulated neural activities, consistent with obstructive and central apnea, have been conἀrmed during seizures in an in situ anesthetized rat preparation (St. John et al. 2006). This preparation has an advantage over in vivo preparations in that delivery of oxygen to the brain is not dependent on the lungs or the cardiovascular system. The EEG activity was recorded, as was activity of the hypoglossal, vagus, and phrenic nerves. The hypoglossal and vagus nerves innervate muscles of the upper airway and larynx, and the phrenic nerve innervates the diaphragm. Seizures were elicited by injections of penicillin into the parietal cortex or the carotid artery. Results showed that after elicitation of the seizures, activity of the hypoglossal and vagal nerves declined greatly, whereas phrenic activity was slightly affected. The authors concluded that such a differential depression of activity of nerves of the upper airway and larynx compared to that of the diaphragm would predispose to obstructive apnea in intact preparations. Moreover, given more time, even the activity of the phrenic nerve declined or ceased, resulting in changes that are known to characterize central apnea. The major conclusion of this study is that seizures may result in recurrent periods of obstructive and central apnea that may account for SUDEP. Recently, severe postictal laryngospasm, as observed in a patient with refractory epilepsy during monitoring, has been described as a potential mechanism for SUDEP (Tavee and Morris 2008). Partial seizures may be associated with prominent oxygen desaturation and ictal ventilatory dysfunction, which could play a role in certain cases of SUDEP in adult patients (Blum et al. 2000). Experimental studies conducted in DBA/2 mice that exhibit sudden death due to respiratory arrest (Collins 1972; Willott and Henry 1976) in the period

Sudden Epileptic Death in Experimental Animal Models

597

immediately after audiogenic seizure have shown that oxygenation prevents sudden death (Venit et al. 2004). The function of respiratory neurons is largely controlled by speciἀc neurotransmitters, among which serotonin has been implicated as modulating respiratory responses to hypoxia (Feldman et al. 2003; Mitchell et al. 2001). However, serotonin receptors have also been implicated in the modulation of seizures (Browning et al. 1997; Pericic et al. 2005) and serotonin agents (fluoxetine) reduced respiratory arrest in DBA/2 mice at doses that did not reduce seizure severity (Tupal and Faingold 2006). The physiopathological mechanisms of SUDEP have also been analyzed in a series of studies on focal epilepsy induced in hemispherectomized rats. In this experimental animal model, independent of cerebral cortex influence, cardioarrhythmogenic triggers may be activated by paroxysmal activity focally induced at hypothalamic, midbrain, and hindbrain levels by topical application of penicillin G (Mameli et al. 1988, 1990). The cardioarrhythmogenic potential of these epileptic foci decreased in the rostrocaudal direction; being more relevant the cardiovascular impairments observed as a consequence of the hypothalamic epileptic focus ἀring, particularly during simultaneous coactivation of the mesencephalic focus (Mameli et al. 1993). Cardiovascular impairments were characterized by sinus bradyarrhythmias, alterations of the repolarization and atriaventricular blocks of different degrees, and signiἀcant decreases in systolic blood pressure (Mameli et al. 1988, 1989, 1990, 1993) temporally correlated to a signiἀcant increase in spontaneous vagal nerve activity. The cardiac arrhythmias became life-threatening when blood gases and electrolytic parameters were simultaneously impaired (Mameli et al. 2001). However, when paroxysmal activity ended, the cardiovascular alterations always disappeared, and vagal nerve ἀring returned to basal values. This demonstrated that in supported ventilation, and in the absence of metabolic derangements, the cardioarrhythmogenic trigger activation was not sufficient to explain SUDEP. These experiments suggested to the authors that fatal evolution consequent to heart impairment was probably related not only to cardiac dysfunction of autonomic origin but also to concomitant metabolic derangement that most likely shared the same genesis. Further experiments, performed using the same experimental animal model, tested the existence of pulmonary complications during the activation of the central cardioarrhythmogenic triggers (Mameli et al. 2006). In the hemispherectomized and artiἀcially ventilated animals, the following parameters were simultaneously analyzed before, during, and after epileptic foci induced both at hypothalamic and mesencephalic levels: spontaneous electrical activity of the hypothalamic neurons, electrothalamogram, spontaneous multiunit vagal nerve ἀber activity, systemic artery blood pressure, pulmonary artery blood pressure, dynamic ECG, and blood gas analysis (pO2, pCO2, bicarbonate, sodium, potassium, Hgb concentrations, pH value, O2% saturation, and bases excess). Metabolic derangements developing throughout the experiment were not purposely adjusted, and only body temperature was kept constant. This study showed that after hypothalamic epileptic focus and mesencephalic epileptic foci, the paroxysmal activity induced, within a short latency, a signiἀcant increase of spontaneous vagal nerve ἀring that was strictly correlated to ECG impairments such as wandering pacemaker, biphasic or negative P waves, extrasystolia, A-V blocks, derangement of A-V conduction and recovery phase, bundle branch blocks, bradyarrhythmias, flattened T waves, and hypotension. Together with a parasympathetic hypertonicity, a concomitant functional imbalance of the orthosympathetic division also developed. When paroxysmal activity began, despite that vagal activity signiἀcantly increased and corresponding bradycardia was expected, heart rate remained around its basal value, sometimes even

598 Sudden Death in Epilepsy: Forensic and Clinical Issues

showing a tendency to increase (Figure 37.1c). Under the same circumstances, supraventricular and ventricular extrasystoles and ventricular complexes of high voltage appeared concomitantly, which also strongly suggested a possible increase in cardiac inotropism that was dependent on orthosympathetic involvement (Figure 37.2d). In the surviving animals (approximately 75%), when paroxysmal activity ended, vagal nerve activity and cardiovascular parameters returned to basal conditions. Macro- and microscopic examination of their lungs never showed any alterations of pulmonary parenchyma, pulmonary vessels, or the bronchial tree. In the deceased animals (approximately 25%) that had manifested interictal and ictal activity, spontaneous vagal nerve ἀring showed similar electrophysiological features, as well as a time course that overlapped the one observed in animals that survived. However, as previously observed (Mameli et al. 2001), the occurrence of cardiac arrhythmias was always accompanied by hyperkalemia and other metabolic derangements, followed by death, which occurred after 3 to 4 h of paroxysmal activity (Figures 37.1 and 37.2). Systemic hypotension was accompanied by signiἀcant pulmonary hypertension, similar to what was described by Johnston and coworkers (1996) in sheep, that further worsened during ictal activity (50% increase). Postmortem macroscopic examination of the lungs demonstrated no relevant alterations in any case. However, histological preparations demonstrated a slight alveolar and perivascular edema in the subinterstitial spaces of the arterial vessels and a concomitant edematous inἀltration in the alveolar and bronchial spaces. Finally, the bronchial tree was ἀlled with considerable intraluminal mucous secretions. Comparison between surviving and deceased animals during ictal activity showed signiἀcant differences in the deceased animals in heart rate, systolic and diastolic pressures, pulmonary artery pressure, potassium and bicarbonate concentrations, pH value, pO2, and pCO2. Regarding the pattern of ECG impairment, during interictal activity, no signiἀcant differences were observed, whereas during ictal activity, signiἀcant differences were observed in the incidence of sinoatrial blocks, bundle branch blocks, and T-wave impairment. In addition, signiἀcant differences were found in the histological patterns of the lungs, which were most likely caused by a state of pulmonary hypertension concomitant with paroxysmal activity and parasympathetic overflow (Figure 37.3). In the opinion of Mameli et al. (2001), a possible explanation of this observation was that at the level of the respiratory apparatus, the paroxysmal activity triggered an increase of the parasympathetic tone, inducing bronchoconstriction and increased mucous secretion. Bronchoconstriction determined, in turn, a reduction of alveolar ventilation with a consequent fall in pO2 as conἀrmed by blood gas data. Furthermore, the decrease in pO2 could have triggered a reflex constriction of pulmonary vessels. Although the mechanism of this phenomenon is still unknown, it is possible that it is induced by a local reflex whose functional signiἀcance is to shunt blood flow from hypoventilated to normally ventilated zones. In the deceased animals, the parasympathetic hypertonicity prejudiced the normal ventilation of the pulmonary parenchyma, although the animals were artiἀcially ventilated. The parasympathetic hypertone caused, therefore, a hypoxic condition that probably extended to the entire lung and in turn caused a reflex vessel constriction and pulmonary artery blood pressure increase. Pulmonary hypertension then resulted in perivascular edema and the edematous inἀltration in the alveolar and bronchial spaces observed in histological preparations (Mameli et al. 2006). Disruption of pulmonary function has been repeatedly reported both in animal and in human epilepsy (Nelson and Ray 1968; Terrence et al. 1975; Bayne and Simon 1981; Harper et

Sudden Epileptic Death in Experimental Animal Models

599

B

A 1

28583 counts

300

2

19724 counts

300

(Scale = 46.22 s/div)

(Scale = 46.22 s/div)

3

C

D

1

56372 counts

38374 counts 300

300

2

3

(Scale = 46.22 s/div)

(Scale = 46.22 s/div) a b c d

Figure 37.1╇ Simultaneous recordings of spontaneous vagal nerve activity and ECG in a deceased animal. Trace 1: spontaneous electrical activity of multiunit vagal nerve fibers recorded using tungsten in glass electrodes. Trace 2: frequency distribution histograms of the same activity constructed during 231.1 s analysis. Trace 3: electrocardiogram. All the events were simultaneously recorded in basal conditions (a) and after the activation of hypothalamic epileptic focus and mesencephalic epileptic foci at 60 (b), 120 (c), and 130 min (d), respectively. Traces a–d: impairment of the ECG, 130 min after hypothalamic epileptic focus and mesencephalic focus induction. The concomitant imbalance of the orthosympathetic function can be indirectly inferred by analyzing the parasympathetic activity. In fact, despite a reduction in vagal nerve firing (b) compared to basal values (a), heart rate did not increase and ventricular extrasystoles appeared. On the other hand, when the vagal nerve firing markedly increased (c) and a corresponding bradycardia was expected, the heart rate also remained unaffected. Bradyarrhythmias appeared only after 130 min of epileptic foci activity. (From Mameli, O., et al., Seizure, 10 (4), 269–278, 2001. With permission.)

600 Sudden Death in Epilepsy: Forensic and Clinical Issues A

1

B

2

3 2547 counts

100

5816 counts

100 50 50

50 4

(Scale = 45.15 s/div)

(Scale = 45.15 s/div)

5

C

D

1 2

3 100

9225 counts

11322 counts

50

50 4

100

(Scale = 45.15 s/div)

(Scale = 45.15 s/div)

5 (Scale = 500.9 ms/div)

Figure 37.2╇ Simultaneous recordings of electrothalamogram, vagal nerve firing, and ECG in

a deceased animal. Trace 1: electrothalamographic signals were recorded using a pair of silver ball electrodes positioned on the thalamic surface. Recordings were performed with a Grass 7P5 and 7DA polygraph. Trace 2: the same event simultaneously analyzed by a computer (Tecfen Computer Scope Analysis ISC-16 software). Trace 3: spontaneous electrical activity of multiunit vagal nerve fibers. Trace 4: frequency distribution histograms of the same activity during 231.1 s analysis. Trace 5: Electrocardiogram. All the events were simultaneously recorded in basal conditions (a) and after the activation of both hypothalamic epileptic focus and mesencephalic epileptic foci at 40 (b), 60 (c), and 120 min (d), respectively. The concomitant imbalance of the orthosympathetic division is shown by the appearance of ventricular complexes of high voltage (c–d). (From Mameli, O., et al., Seizure, 10 (4), 269–278, 2001. With permission.)

Sudden Epileptic Death in Experimental Animal Models

601

al. 1984; James et al. 1991; Graham 1992; Hanning and Alexander-Williams 1995) to worsen the neurogenic arrhythmias, together with metabolic impairment. In agreement with the observations in sheep (Johnston et al. 1996, 1995, 1997), the ἀndings obtained in hemispherectomized rats (Mameli et al. 1988, 1989, 1990, 1993, 2001, 2006) showed that when considered alone, neurogenic arrhythmias are insufficient to cause the animals’ death. However, it is not clear why, given that the same experimental protocol was applied to all animals, only some of these developed clinical symptoms so severe that survival was unfailingly compromised. Because paroxysmal activity involved the same cerebral structures, it seems difficult to hypothesize that activation of a fatal trigger responsible for animal death acts only in some animals. A more convincing hypothesis to explain these ἀndings could be the existence of animals characterized by a low threshold of excitability, whose central structures are unable during epileptic seizure to maintain a balanced ratio between sympathetic and parasympathetic activities. In particular, orthosympathetic activation could paradoxically synergize the vagal influence on heart activity and transform a serious, but reversible, cardiac bradyarrhythmia into a fatal event. This hypothesis coÂ�Â� incides with the ἀnding of increased concentrations of circulating catecholamines during epileptic seizures to potentially arrhythmogenic levels in both animals (Johnston et al. 1995; Doba et al. 1975; Benowitz et al. 1986) and human beings (Simon et al. 1984). At a peripheral level, in the animals with a low threshold of excitability, these high epinephrine levels could activate the alpha-membrane receptors that are responsible for cellular potassium depletion (Williams et al. 1984) at muscular cell levels (Mauger et al. 1982), with a consequent increase in their plasma concentration. Stimulation of alpha-receptors impairs extrarenal potassium disposition, a speciἀc effect that can be reversed by alpha-blockade (Williams et al. 1984). Although alpha-receptor stimulation is known to account for a transient release of hepatic potassium reserves in response to an epinephrine infusion, an effect that occurs within minutes and lasts briefly (Craig and Mendell 1959; Ellis 1956; Vick et al. 1972; Sterns et al. 1981; Giugliano et al. 1979; Guerra and Kitabchi 1976; DeFronzo et al. 1980; Minaker and Rowe 1982), a sustained effect may be observed in a continuous stimulation, as occurs during seizures. Therefore, these effects on potassium availability probably result from a decreased net cellular uptake of potassium, which depends on an enhanced alpha-adrenergic activity in those animals with a low threshold of excitability. This condition would have worsened the vagal-mediated cardiac arrhythmias to a point capable of inducing their fatal evolution into SUDEP. The short latency of hyperkalemia observed in the deceased animals (Mameli et al. 2001, 2006) seems to support this hypothesis and provides evidence for a likely neurogenic extrarenal origin of the phenomenon. Further evidence is suggested in these experiments by the appearance of sharpened T waves that were always related to the increased potassium plasma concentration. These results agree with the severe changes in potassium plasma concentrations detected during generalized seizures in adolescent baboons (Mello et al. 1993). Finally, extending the above considerations to human epilepsy, it can be postulated that patients with a low threshold of excitability may also exist. In these subjects, the cardioarrhythmogenic triggers, activated by hypothalamic and mesencephalic epileptic foci (hypothalamic epileptic focus and mesencephalic focus), may represent the ἀnal common pathway of a storm of signals originating at the level of cortical epileptic foci. In fact, the extensive neural connections, originating in the cerebral cortex and particularly in the frontal and temporal lobes, converge on hypothalamic and brain stem structures (Papez

602 Sudden Death in Epilepsy: Forensic and Clinical Issues 350 300

(a)

250

Basal conditions

200

p = ns

150 100

Surviving Deceased

Deceased

ASP

ADP

PAP

MUAV

HEF -MEF

121±9.8

80±4.5

13.3±1.4

22.88±4.9

0.0

335.1±10.9

121.34±8.4

82±10

14.58±1.8

19.4±6.5

0.0

Na

+

HCO -

pH

pO2

pCO2

Hb

136.4±4.5

3.45±0.44

27.36±3.3

7.38 ±0.16

93.1±10.3

40.6±3.1

14.9±3.4

136 ±6

3.6±0.4

29.2±2.8

7.39±0.6

95 ±4

40.5 ±2

14.6 ±3.7

K

3

Hb

pCO2

pO2

pH

3

K+ HCO-

Na+

MUAV

HEF-MEF

HR

336.2±17.5

+

Surviving

ADP

t test

PAP

HR

0

ASP

50

350 300

(b) Interictal activity

p < 0.005

250

*

200 150

*p = ns

*

100

*

Surviving Deceased

Surviving Deceased

HR 325±4.3 238.3±43.7

ASP 114.5±5.8 109.3±6.9

ADP 76.6±4.6 64.4±5.3

MUAV 156.7±7 159.9±50.4

PAP 11.2±4.4 20.7±3.6

Hb

pCO2

pH

pO2

K+ HCO-3

Na+

MUAV

*

HEF-MEF

ADP

PAP

HR

0

ASP

50

HEF -MEF 0.097±0.02 0.102± 0.17

Na+

K+

pH 7.40±0.09

pCO2

3.59±0.24

HCO3-26.09±3.4

pO2

137.3±2.4

92.1±7.7

41.7±1.5

Hb 15.1±1.3

135 ±0.6

7.2±1.8

25.8 ±1.9

7.20±0.17

66.7±9.1

44.8 ±3.6

15.5 ±0.9

450 * 400 Figure 37.3╇ Main metabolic, cardiovascular, and respiratory parameters determined in (a) 350 = ns basal conditions and during (b) interictal and (c) ictal activities in surviving and*pdeceased ani300 + (c) p < 0.005 mals. All values are expressed as means ± standard deviation. Sodium (Na ), potassium (K+), 250 − bicarboÂ�nateIctal (HCO as milliequivalent per liter. Hemoglobin (Hb) activity 3 ), and pH values are expressed 200 values are expressed as grams per milliliter. arterial systolic (ASP), 150 pO2 and pCO2 pressures, * arterial dystolic pressures (PAP) are expressed as millimeters of * t test (ADP), and pulmonary artery 100 mercury. Heart rate (HR) values are expressed 50 as beats per minute. Multiunit electrical activity 0 of vagal nerve fibers (MUAV) as well as the electrical activity of hypothalamic epileptic focus

Deceased

325±4.3 238.3±43.7 +

114.5±5.8 109.3±6.9 +

156.7±7 159.9±50.4

0.097±0.02 0.102±0.17

Surviving

Na

K

pCO2

3.76±0.45

24.9±2.8

pH 7.36±0.2

pO2

136.4±6.2

88 ±9.3

42.4±2.7

Hb 15.8±0.7

Deceased

136.5±2

8.6 ±4

13.6 ±4

7.18 ±0.1

47.8 ±17

49.9 ±5.3

16.07 ±2.4

Hb

pCO2

pH

pO2

3

K+ HCO-

Na+

HEF-MEF

PAP

MUAV

11.2±4.4 20.7±3.6

76.6±4.6 64.4±5.3

HCO3-

ADP

HR

Surviving

ASP

* and mesencephalic focus cardioarrhythmogenic trigger values are expressed as the number of spikes per second. Statistical significance of differences between surviving and deceased animals was analyzed by paired t tests. (From Mameli, O., et al., Seizure, 15 (5), 275–287, 2006. With permission.) HR ASP ADP PAP MUAV HEF -MEF

ADP 76.6±4.6 64.4±5.3

MUAV 156.7±7 159.9±50.4

PAP 11.2±4.4 20.7±3.6

HCO -pO pH Sudden Epileptic Na Death in KExperimental Animal Models

Surviving Deceased

HR 325±4.3 238.3±43.7 +

ASP 114.5±5.8 109.3±6.9 +

44.8 ±3.6

15.5 ±0.9

*

* *

MUAV 156.7±7 159.9±50.4

PAP 11.2±4.4

20.7±3.6

HEF -MEF 0.097±0.02 0.102±0.17

Surviving

Na

K

pCO2

3.76±0.45

24.9±2.8

pH 7.36±0.2

pO2

136.4±6.2

88 ±9.3

42.4±2.7

Hb 15.8±0.7

Deceased

136.5±2

8.6 ±4

13.6 ±4

7.18 ±0.1

47.8 ±17

49.9 ±5.3

16.07 ±2.4

3

Hb

*

ADP 76.6±4.6 64.4±5.3

HCO -

*p = ns p < 0.005

pO2

t test

66.7±9.1

pCO2

(c) Ictal activity

7.20±0.17

603

3

450 400 350 300 250 200 150 100 50 0

HEF-MEF

25.8 ±1.9

Hb 15.1±1.3

PAP

7.2±1.8

41.7±1.5

MUAV

135 ±0.6

pCO2

92.1±7.7

ADP

26.09±3.4

7.40±0.09

2

3

3.59±0.24

HR

Deceased

+

137.3±2.4

ASP

+

Surviving

HEF -MEF 0.097±0.02 0.102± 0.17

pH

ASP 114.5±5.8 109.3±6.9

K+ HCO-

Deceased

HR 325±4.3 238.3±43.7

Na+

Surviving

Figure 37.3╇ (Continued)

1937; Wall and Davis 1951; Langan et al. 2000; Landau 1953; Walker 1966; Gray 1973; Saper et al. 1976; Korner 1979; Willis and Grossman 1981; Breusch 1984; Natelson 1985). Therefore, the possibility cannot be excluded that under certain circumstances the cortical signals activate the arrhythmogenic triggers simultaneously, thus inducing cardiopulmonary and metabolic impairments that in these patients with a low threshold of excitability may result in sudden death. As for the existence of individuals with a low threshold of excitability, it may be of some interest to consider studies that investigated genetic conditions related to SUDEP. In double-mutant mice that express GM3 as their major ganglioside, a “sudden death phenotype” has been described that was extremely susceptible to induction of lethal seizures by a sound stimulus (Kawai et al. 2001). The gangliosides are a family of glycosphingolipids that contain sialic acid and, although they are abundant on neuronal cell membranes, their speciἀc functions in the CNS remain largely undeἀned. Results of this study showed that these compounds play essential roles in the proper functioning of the CNS and that their absence may contribute to SUDEP susceptibility. In other transgenic mice, the Sema TG, characterized by cardiac-speciἀc overexpression of Sema3a, reduced sympathetic innervation and attenuation of epicardial-to-endocardial innervation gradient have been described. These mice show susceptibility to ventricular tachycardia due to catecholamine supersensitivity and prolongation of the action potential duration, which can terminate in SUDEP. Results of this study showed that appropriate cardiac Sema3a expression is needed for proper sympathetic heart rate control (Ieda et al. 2007). Another interesting hypothesis for the risk of SUDEP considers the neurogenesis process in the CNS. It has long been believed that in mammals, the origin of new neurons in most CNS regions was a process limited to embryogenesis (Hilz et al. 2002) because once development is complete, the progenitor cells that mature into neurons go through a differentiation process and become incapable of division. In contrast, a neurogenesis process in the CNS of adults has been described in several species such as crustaceans (Hilz et al. 2002), reptiles (Lopez-Garcia et al. 1988), amphibians (Polenov and Chetverukhin 1993),

604 Sudden Death in Epilepsy: Forensic and Clinical Issues

birds (Nottebohm 1989), rodents (Altman and Das 1965), primates (Eckenhoff and Rakic 1988), and human beings (Eriksson et al. 1998). In all the mammalian species already studied, including humans, the mitotically active progenitor cells, capable of generating new neurons in the adult phase, are located in speciἀc regions (Eriksson et al. 1998; Gould et al. 1998) such as the subventricular zone of the lateral ventricles and the dentate gyrus of the hippocampal formation. In adult mice, these cells migrate to the olfactory bulb, where they differentiate into a large variety of cell types, such as periglomerular neurons and interneurons, as well as astrocytes and oligodendrocytes (Lois and Alvarez-Buylla 1994). In the rat, these new cells proliferate and migrate continuously into the granular cell layer (Kuhn et al. 1996) where they develop a morphology typical of granule cells (Cameron et al. 1993), express neuronal differentiation markers (Kuhn et al. 1996), and extend their axons to the pathway of mossy ἀbers that project into the CA3 region of the hippocampus (Stanἀeld and Trice 1988). Several brain insults, including epileptogenic activity associated with epilepsy, are able to stimulate progenitor cell proliferation in the dentate gyrus (Parent 2007; Zhao et al. 2008). Using the pilocarpine model of temporal lobe epilepsy, for the ἀrst time in rats, this status epilepticus has been shown to cause a dramatic and prolonged increase in cell proliferation in the dentate subgranular proliferative zone (Parent et al. 1997). This observation has been conἀrmed in humans and adult rodent models of temporal lobe epilepsy (Parent et al. 2006; Mello et al. 1993), showing the possibility that these cells aberrantly migrated after the status epilepticus and were abnormally integrated. Resultant hyperexcitability may contribute to seizure generation and/or propagation (Parent et al. 1997, 2006; Dashtipour et al. 2001; Scharfman et al. 2000). In a recent study, aberrant neurogenesis was hypothesized to negatively influence the cardiovascular system of patients with epilepsy, leading to cardiac abnormalities and hence SUDEP (Scorza et al. 2008). The analysis of the literature reviewed in the present report shows some inconsistencies concerning risk factors for SUDEP. In our opinion, differences in data collected using experimental animal models could derive from the variability in study design. Moreover, the different behaviors displayed by the same brain structures could be related not only to differences between focal and generalized models but also to the intrinsic characteristics of the same structures when considered in different species, as well as to speciἀc epileptogenic characteristic of the chemicals used to induce the epileptic foci (Prince 1969, 1972; Mameli et al. 1999, 1991). With regard to penicillin-G epileptogenic activity, for instance, the sensitivity of nervous structures has been shown to signiἀcantly decrease in the rostrocaudal direction (Mameli et al. 2006). Moreover, at the bulbar level, where cardiorespiratory neurons are localized, penicillin G–induced discharges are not as severe as those induced at the mesencephalic level and disappear after midcollicular transection (Mameli et al. 1991), showing their dependence on rostral nervous structures (De Riu et al. 1994). Instead, in drug-induced generalized epilepsy, the simultaneous general involvement of all cortical and subcortical neuronal networks must be considered. These can be impaired directly and/or indirectly by the epileptogenic drugs as well as by endogenous opioids and neurotransmitters released during generalized seizures. In conclusion, the different models of experimental epilepsy used in animals to analyze the physiopathological mechanisms of SUDEP conἀrm the clinical events detected in patients as well as the overlapping mechanisms proposed to explain human SUDEP. They€show that in fatal conditions, the cortical signals simultaneously activate the central autonomic regulatory triggers, inducing cardiopulmonary and metabolic impairments€that in some subjects with a low threshold of excitability might result in a sudden death. This

Sudden Epileptic Death in Experimental Animal Models

605

is particularly true when the parasympathetic division is involved. In fact, its activation might be responsible for neurogenic ictal bradyarrhythmias, respiratory and metabolic impairments, cardiac asystole, or even SUD in epileptic patients.

References Allen, W. F. 1931. An experimentally produced premature systolic arrhythmia (pulsus bigeminus) in rabbits. Am J Physiol 98: 344–335. Allen, G. V., R. T. Cheung, and D. F. Cechetto. 1995. Neurochemical changes following occlusion of the middle cerebral artery in rats. Neuroscience 68 (4): 1037–1050. Altman, J., and G. D. Das. 1965. Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats. J Comp Neurol 124 (3): 319–335. Annegers, J. F., and S. P. Coan. 1999. SUDEP: Overview of deἀnitions and review of incidence data. Seizure 8 (6): 347–352. Attar, H. J., M. T. Gutierrez, S. Bellet, and J. R. Ravens. 1963. Effect of stimulation of hypothalamus and reticular activating system on production of cardiac arrhythmia. Circ Res 12: 14–21. Bayne, L. L., and R. P. Simon. 1981. Systemic and pulmonary vascular pressures during generalized seizures in sheep. Ann Neurol 10 (6): 566–569. Benowitz, N. L., R. P. Simon, and J. R. Copeland. 1986. Status epilepticus: Divergence of sympathetic activity and cardiovascular response. Ann Neurol 19 (2): 197–199. Blum, A. S., J. R. Ives, A. L. Goldberger, I. C. Al-Aweel, K. B. Krishnamurthy, F. W. Drislane, and D.€L.€Schomer. 2000. Oxygen desaturations triggered by partial seizures: Implications for cardiopulmonary instability in epilepsy. Epilepsia 41 (5): 536–541. Blumhardt, L. D., P. E. Smith, and L. Owen. 1986. Electrocardiographic accompaniments of temporal lobe epileptic seizures. Lancet 1 (8489): 1051–1056. Boeles, J. T., G. C. Korteweg, and J. Ten Cate. 1957. Influence of stimulation of some subcortical areas on electrocardiogram. J Neurophysiol 20 (1): 100–107. Breusch, S. R. 1984. Anatomy of the human hypothalamus. In The Hypothalamus, ed. J. R. Givens. Chicago, IL: Year Book Medical Publishers. Browning, R. A., A. V. Wood, M. A. Merrill, J. W. Dailey, and P. C. Jobe. 1997. Enhancement of the anticonvulsant effect of fluoxetine following blockade of 5-HT1A receptors. Eur J Pharmacol 336 (1): 1–6. Cameron, H. A., C. S. Woolley, B. S. McEwen, and E. Gould. 1993. Differentiation of newly born neurons and glia in the dentate gyrus of the adult rat. Neuroscience 56 (2): 337–344. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30 (4): 225–240. Chapman, W. P., R. B. Livingston, and K. E. Livingston. 1949. Frontal lobotomy and electrical stimulation of orbital surface of frontal lobes; effect on respiration and on blood pressure in man. Arch Neurol Psychiatry 62 (6): 701–716. Chapman, W. P., K. E. Livingston, and J. L. Poppen. 1950. Effect upon blood pressure of electrical stimulation of tips of temporal lobes in man. J Neurophysiol 13 (1): 65–71, illust. Cheung, R. T., and D. F. Cechetto. 1995. Neuropeptide changes following excitotoxic lesion of the insular cortex in rats. J Comp Neurol 362 (4): 535–550. Cheung, R. T., T. Diab, and D. F. Cechetto. 1995. Time-course of neuropeptide changes in periÂ�ischemic zone and amygdala following focal ischemia in rats. J Comp Neurol 360 (1): 101–120. Cheung, R. T., V. C. Hachinski, and D. F. Cechetto. 1997. Cardiovascular response to stress after middle cerebral artery occlusion in rats. Brain Res 747 (2): 181–188. Collins, R. L. 1972. Audiogenic seizures. In Experimental Models of Epilepsy: A Manual for the Laboratory Worker, ed. D. P. Purpura, J. K. Penry, D. B. Tower, D. M. Woodbury, and R. D. Walter. New York, NY: Raven Press.

606 Sudden Death in Epilepsy: Forensic and Clinical Issues Colugnati, D. B., P. A. Gomes, R. M. Arida, M. de Albuquerque, R. M. Cysneiros, E. A. Cavalheiro, and F. A. Scorza. 2005. Analysis of cardiac parameters in animals with epilepsy: Possible cause of sudden death? Arq Neuropsiquiatr 63 (4): 1035–1041. Craig, Jr., A. B., and P. L. Mendell. 1959. Blockade of hyperkalemia and hyperglycemia induced by epinephrine in frog liver and in cats. Am J Physiol 197 (1): 52–54. Darbin, O., D. Casebeer, and D. K. Naritoku. 2003. Effects of seizure severity and seizure repetition on postictal cardiac arrhythmia following maximal electroshock. Exp Neurol 181 (2): 327–331. Darnell, J. C., and S. J. Jay. 1982. Recurrent postictal pulmonary edema: A case report and review of the literature. Epilepsia 23 (1): 71–83. Dasheiff, R. M., and L. J. Dickinson. 1986. Sudden unexpected death of epileptic patient due to cardiac arrhythmia after seizure. Arch Neurol 43 (2): 194–196. Dashtipour, K., P. H. Tran, M. M. Okazaki, J. V. Nadler, and C. E. Ribak. 2001. Ultrastructural features and synaptic connections of hilar ectopic granule cells in the rat dentate gyrus are different from those of granule cells in the granule cell layer. Brain Res 890 (2): 261–271. De Riu, P. L. 1983. Cardiovascular effects induced by lateral hypothalamic stimulation in anesthetized cats. IRCS Med Sci 11: 32–33. De Riu, P. L., F. Melis, A. Becciu, and O. Mameli. 1994. Cerebellar and mesencephalic influence on bulbar penicillin-G epileptogenesis in rats. Physiol Behav 56 (4): 715–721. DeFronzo, R. A., P. Felig, E. Ferrannini, and J. Wahren. 1980. Effect of graded doses of insulin on splanchnic and peripheral potassium metabolism in man. Am J Physiol 238 (5): E421–E427. Delgado, J. M. 1960. Circulatory effects of cortical stimulation. Physiol Rev Suppl 4: 146–178. Dhir, A., and S. K. Kulkarni. 2006. Rofecoxib, a selective cyclooxygenase-2 (COX-2) inhibitor potentiates the anticonvulsant activity of tiagabine against pentylenetetrazol-induced convulsions in mice. Inflammopharmacology 14 (5–6): 222–225. Dikshit, B. B. 1934. The production of cardiac irregularities by excitation of the hypothalamic center. J Physiol 81: 382–394. Doba, N., H. R. Beresford, and D. J. Reis. 1975. Changes in regional blood flow and cardiodynamics associated with electrically and chemically induced epilepsy in cat. Brain Res 90 (1): 115–132. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33 (2): 310–316. Eckenhoff, M. F., and P. Rakic. 1988. Nature and fate of proliferative cells in the hippocampal dentate gyrus during the life span of the rhesus monkey. J Neurosci 8 (8): 2729–2747. Ellis, S. 1956. The metabolic effects of epinephrine and related amines. Pharmacol. Rev. 8 (4): 485–565. Eriksson, P. S., E. Perἀlieva, T. Bjork-Eriksson, A. M. Alborn, C. Nordborg, D. A. Peterson, and F. H. Gage. 1998. Neurogenesis in the adult human hippocampus. Nat Med 4 (11): 1313–1317. Evans, D. E., and R. A. Gillis. 1974. Effect of diphenylhydantoin and lidocaine on cardiac arrhythmias induced by hypothalamic stimulation. J Pharmacol Exp Ther 191 (3): 506–517. Evans, D. E., and R. A. Gillis. 1978. Reflux mechanisms involved in cardiac arrhythmias induced by hypothalamic stimulation. Am J Physiol 234 (2): H199–H209. Feldman, J. L., G. S. Mitchell, and E. E. Nattie. 2003. Breathing: Rhythmicity, plasticity, chemosensitivity. Annu Rev Neurosci 26: 239–266. Ficker, D. M., E. L. So, W. K. Shen, J. F. Annegers, P. C. O’Brien, G. D. Cascino, and P. G. Belau. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51 (5): 1270–1274. Freeman, R. 2006. Cardiovascular manifestations of autonomic epilepsy. Clin Auton Res 16 (1): 12–17.

Sudden Epileptic Death in Experimental Animal Models

607

Fuhr, P., and D. Leppert. 2000. Neuroimages. Cardiac arrest during partial seizure. Neurology 54 (10): 2026. Fuster, J. M., and S. J. Weinberg. 1960. Bioelectrical changes of the heart cycle induced by stimulation of diencephalic regions. Exp Neurol 2: 26–39. Graham, D. I. 1992. Hypoxia and vascular disorders. In Greenfield’s Neuropathology, ed. J. H. Adams and L. W. Duchen. London: Arnold. Giugliano, D., R. Torella, S. Sgambato, and F. D’Onofrio. 1979. Effects of alpha- and beta-adrenergic inhibition and somatostatin on plasma glucose, free fatty acids, insulin glucagon and growth hormone responses to prostaglandin E1 in man. J Clin Endocrinol Metab 48 (2): 302–308. Goodman, J. H., R. E. Berger, and T. K. Tcheng. 2005. Preemptive low-frequency stimulation decreases the incidence of amygdala-kindled seizures. Epilepsia 46 (1): 1–7. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1990. Kindled seizures elevate blood pressure and induce cardiac arrhythmias. Epilepsia 31 (5): 489–495. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1999. Kindled seizures activate both branches of the autonomic nervous system. Epilepsy Res 34 (2–3): 169–176. Goren, M. Z., R. Aker, H. R. Yananli, and F. Y. Onat. 2003. Extracellular concentrations of catecholamines and amino acids in the dorsomedial hypothalamus of kindled rats. A microdialysis study. Pharmacology 68 (4): 190–197. Gould, E., P. Tanapat, B. S. McEwen, G. Flugge, and E. Fuchs. 1998. Proliferation of granule cell precursors in the dentate gyrus of adult monkeys is diminished by stress. Proc Natl Acad Sci U S€A 95 (6): 3168–3171. Gray, H. 1973. In Gray’s Anatomy, Descriptive and Applied, ed. D. V. Davies and R. E. Coupland. London: Longman Greens. Guerra, S. M., and A. E. Kitabchi. 1976. Comparison of the effectiveness of various routes of insulin injection: Insulin levels and glucose response in normal subjects. J Clin Endocrinol Metab 42 (5): 869–874. Gunn, C. G., G. Sevelius, J. Puiggari, and F. K. Myers. 1968. Vagal cardiomotor mechanisms in the hindbrain of the dog and cat. Am J Physiol 214 (2): 258–262. Hall, R. E., H. D. Sybers, J. H. Greenhoot, and C. M. Bloor. 1974. Myocardial alterations following hypothalamic stimulation in the intact conscious dog. Am Heart J 88 (6): 770–776. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hanning, C. D., and J. M. Alexander-Williams. 1995. Pulse oximetry: A practical review. BMJ 311 (7001): 367–370. Harper, R. M., R. C. Frysinger, R. B. Trelease, and J. D. Marks. 1984. State-dependent alteration of respiratory cycle timing by stimulation of the central nucleus of the amygdala. Brain Res 306 (1–2): 1–8. Hauser, W. A., J. F. Annegers, and L. R. Elveback. 1980. Mortality in patients with epilepsy. Epilepsia 21 (4): 399–412. Hilz, M. J., O. Devinsky, W. Doyle, A. Mauerer, and M. Dutsch. 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125 (Pt 5): 985–995. Hockman, C. H., H. P. Mauck Jr., and E. C. Hoff. 1966. ECG changes resulting from cerebral stimulation: II. A spectrum of ventricular arrhythmias of sympathetic origin. Am Heart J 71 (5): 695–700. Ieda, M., H. Kanazawa, K. Kimura, F. Hattori, Y. Ieda, M. Taniguchi, J. K. Lee, K. Matsumura, Y. Tomita, S. Miyoshi, K. Shimoda, S. Makino, M. Sano, I. Kodama, S. Ogawa, and K. Fukuda. 2007. Sema3a maintains normal heart rhythm through sympathetic innervation patterning. Nat Med 13 (5): 604–612. Izzo, P. N., J. Deuchars, and K. M. Spyer. 1993. Localization of cardiac vagal preganglionic motoneurones in the rat: Immunocytochemical evidence of synaptic inputs containing 5-Â�hydroxytryptamine. J Comp Neurol 327 (4): 572–583.

608 Sudden Death in Epilepsy: Forensic and Clinical Issues James, M. R., H. Marshall, and M. Carew-McColl. 1991. Pulse oximetry during apparent tonic– clonic seizures. Lancet 337 (8738): 394–395. Johnson, L. C., and R. A. Davidoff. 1964. Autonomic changes during paroxysmal EEG activity. Electroencephalogr Clin Neurophysiol 17: 25–35. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37 (4): 531–537. Johnston, S. C., T. M. Darragh, and R. P. Simon. 1996. Postictal pulmonary edema requires pulmonary vascular pressure increases. Epilepsia 37 (5): 428–432. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42 (4): 588–594. Kalviainen, R., T. Keranen, J. Mustonen, E. Lansimies, and P. J. Riekkinen. 1990. Autonomic nervous system function in Baltic myoclonus epilepsy. Epilepsy Res 5 (3): 251–254. Kawai, H., M. L. Allende, R. Wada, M. Kono, K. Sango, C. Deng, T. Miyakawa, J. N. Crawley, N. Werth, U. Bierfreund, K. Sandhoff, and R. L. Proia. 2001. Mice expressing only monosialoganglioside GM3 exhibit lethal audiogenic seizures. J Biol Chem 276 (10): 6885–6888. Keilson, M. J., W. A. Hauser, J. P. Magrill, and M. Goldman. 1987. ECG abnormalities in patients with epilepsy. Neurology 37 (10): 1624–1626. Kelly, A. M., C. J. Porter, M. D. McGoon, R. E. Espinosa, M. J. Osborn, and D. L. Hayes. 2001. Breathholding spells associated with signiἀcant bradycardia: Successful treatment with permanent pacemaker implantation. Pediatrics 108 (3): 698–702. Korner, P. I. 1979. Central nervous control of autonomic cardiovascular function. In Handbook of Physiology: The Cardiovascular System, ed. R. M. Berne, N. Sperelaskis, and S. R. Geiger. Bethesda, MD: Am Physiol Soc. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. Kuhn, H. G., H. Dickinson-Anson, and F. H. Gage. 1996. Neurogenesis in the dentate gyrus of the adult rat: Age-related decrease of neuronal progenitor proliferation. J Neurosci 16 (6): 2027–2033. Kwan, P., and J. W. Sander. 2004. The natural history of epilepsy: An epidemiological view. J Neurol Neurosurg Psychiatry 75 (10): 1376–1381. Landau, M. W. 1953. Autonomic responses mediated via the corticospinal tract. J Neurophysiol 16 (3): 299–311. Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68 (2): 211–213. Lathers, C. M. 1990. Role of neuropeptides in the production of epileptogenic activity and arrhythmias. In Epilepsy and Sudden Death, Chap. 18, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259.

Sudden Epileptic Death in Experimental Animal Models

609

Lathers, C. M., N. Tumer, and C. M. Kraras. 1985. Cardiovascular and epileptogenic effects of pentylÂ� enetetrazol administered intracerebroventricularly in cats. Epilepsia 26: 520. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988. The effect of intracerebroventricular D-ALA2 methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43 (26): 2287–2298. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earl. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38 (1): 47–55. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26 (2): 195–203. Leutmezer, F., C. Schernthaner, S. Lurger, K. Potzelberger, and C. Baumgartner. 2003. ElectroÂ� cardiographic changes at the onset of epileptic seizures. Epilepsia 44 (3): 348–354. Liedholm, L. J., and O. Gudjonsson. 1992. Cardiac arrest due to partial epileptic seizures. Neurology 42 (4): 824–829. Lisander, B., B. Jaju, and S. C. Wang. 1975. CNS site of antiarrhythmic action of diphenylhydantoin (DPH) in the cat. Eur J Pharmacol 31 (1): 53–62. Lois, C., and A. Alvarez-Buylla. 1994. Long-distance neuronal migration in the adult mammalian brain. Science 264 (5162): 1145–1148. Lopez-Garcia, C., A. Molowny, J. M. Garcia-Verdugo, and I. Ferrer. 1988. Delayed postnatal neurogenesis in the cerebral cortex of lizards. Brain Res 471 (2): 167–174. Lown, B., and R. A. DeSilva. 1978. Roles of psychologic stress and autonomic nervous system changes in provocation of ventricular premature complexes. Am J Cardiol 41 (6): 979–985. Lown, B., J. V. Temte, P. Reich, C. Gaughan, Q. Regestein, and H. Hal. 1976. Basis for recurring ventricular ἀbrillation in the absence of coronary heart disease and its management. N Engl J Med 294 (12): 623–629. Mameli, O., M. A. Caria, F. Melis, C. Severino, C. Tavera, P. Mameli, and S. Mameli. 2001. Autonomic nervous system activity and life threatening arrhythmias in experimental epilepsy. Seizure 10 (4): 269–278. Mameli, O., M. A. Caria, A. Pintus, G. Padua, and S. Mameli. 2006. Sudden death in epilepsy: An experimental animal model. Seizure 15 (5): 275–287. Mameli, O., F. Melis, M. A. Caria, A. Solinas, S. Mameli, and P. L. De Riu. 1999. Epileptic discharge of cortical, subcortical and spinal neurons in penicillin induced experimental epilepsy. Arch Ital Biol 137 (1): 29–46. Mameli, O., F. Melis, D. Giraudi, M. Cualbu, S. Mameli, P. L. De Riu, and P. Mameli. 1993. The brainstem cardioarrhythmogenic triggers and their possible role in sudden epileptic death. Epilepsy Res 15 (3): 171–178. Mameli, O., F. Melis, E. Tolu, M. A. Caria, P. L. De Riu, and G. B. Azzena. 1991. Penicillin-induced paroxysmal activity in brainstem neurons. Physiol Behav 49 (3): 631–634. Mameli, P., O. Mameli, E. Tolu, D. Giraudi, G. Padua, F. Melis, and M. A. Caria. 1989. Investigation into whether neurogenic cardiovascular alterations are involved in sudden epileptic death. In Advances in Epileptology, ed. J. Manelis, E. Bental, J. N. Loeber, and F. E. Dreifuss. New York, NY: Raven Press. Mameli, P., O. Mameli, E. Tolu, G. Padua, D. Giraudi, M. A. Caria, and F. Melis. 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29 (1): 74–82. Mameli, O., P. Mameli, E. Tolu, G. Padua, F. Melis, M. A. Caria, and A. Virdis. 1990. Analysis of central cardioarrhythmogenic triggers in experimental epilepsy. Epilepsy Res 7 (3): 210–218. Manning, J. W., and C. N. Peiss. 1960. Cardiovascular responses to electrical stimulation in the diencephalon. Am J Physiol 198: 366–370. Mauger, J. P., F. Sladeczek, and J. Bockaert. 1982. Characteristics and metabolism of alpha 1 adrenergic receptors in a nonfusing muscle cell line. J Biol Chem 257 (2): 875–879. McCown, T. J., R. S. Greenwood, G. D. Frye, and G. R. Breese. 1984. Electrically elicited seizures from the inferior colliculus: A potential site for the genesis of epilepsy? Exp Neurol 86 (3): 527–542.

610 Sudden Death in Epilepsy: Forensic and Clinical Issues Mello, L. E., E. A. Cavalheiro, A. M. Tan, W. R. Kupfer, J. K. Pretorius, T. L. Babb, and D. M. Finch. 1993. Circuit mechanisms of seizures in the pilocarpine model of chronic epilepsy: Cell loss and mossy ἀber sprouting. Epilepsia 34 (6): 985–995. Melville, K. I., B. Blum, H. E. Shister, and M. D. Silver. 1963. Cardiac ischemic changes and arrhythmias induced by hypothalamic stimulation. Am J Cardiol 12: 781–791. Minaker, K. L., and J. W. Rowe. 1982. Potassium homeostasis during hyperinsulinemia: Effect of insulin level, beta-blockade, and age. Am J Physiol 242 (6): E373–E377. Mitchell, G. S., T. L. Baker, S. A. Nanda, D. D. Fuller, A. G. Zabka, B. A. Hodgeman, R. W. Bavis, K. J. Mack, and E. B. Olson Jr. 2001. Invited review: Intermittent hypoxia and respiratory plasticity. J Appl Physiol 90 (6): 2466–2475. Mondon, K., B. Charbonnier, C. Hommet, P. Corcia, A. Autret, and B. de Toffol. 2002. Ictal bradycardia followed by cardiac asystole: A case report. Epileptic Disord 4 (4): 261–264. Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16 (1): 1–7. Nashef, L. 1997. Sudden unexpected death in epilepsy: Terminology and deἀnitions. Epilepsia 38: S6–S8. Nashef, L. 1999. From mystery to prevention: Sudden unexpected death in epilepsy, time to move on. J Neurol Neurosurg Psychiatry 67 (4): 427. Nashef, L., and S. Brown. 1996. Epilepsy and sudden death. Lancet 348 (9038): 1324–1325. Nashef, L., N. Hindocha, and A. Makoff. 2007. Risk factors in sudden death in epilepsy (SUDEP): The quest for mechanisms. Epilepsia 48 (5): 859–871. Nashef, L., and S. D. Shorvon. 1997. Mortality in epilepsy. Epilepsia 38 (10): 1059–1061. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300. Natelson, B. H. 1985. Neurocardiology. An interdisciplinary area for the 80s. Arch Neurol 42 (2): 178–184. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Nelson, D. A., and C. D. Ray. 1968. Respiratory arrest from seizure discharges in limbic system. Report of cases. Arch Neurol 19 (2): 199–207. Neuspiel, D. R., and L. H. Kuller. 1985. Sudden and unexpected natural death in childhood and adolescence. JAMA 254 (10): 1321–1325. Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42 (5): 667–673. Nottebohm, F. 1989. From bird song to neurogenesis. Sci Am 260 (2): 74–79. Novak V, V., L. A. Reeves, P. Novak, A. P. Low, and W. F. Sharbrough. 1999. Time-frequency mapping of R–R interval during complex partial seizures of temporal lobe origin. J Auton Nerv Syst 77 (2–3): 195–202. Opeskin, K., and S. F. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12 (7): 456–464. Oppenheimer, S. M., and D. F. Cechetto. 1990. Cardiac chronotropic organization of the rat insular cortex. Brain Res 533 (1): 66–72. Oppenheimer, S. M., D. F. Cechetto, and V. C. Hachinski. 1990. Cerebrogenic cardiac arrhythmias. Cerebral electrocardiographic influences and their role in sudden death. Arch Neurol 47 (5): 513–519. Oppenheimer, S. M., G. Kedem, and W. M. Martin. 1996. Left-insular cortex lesions perturb cardiac autonomic tone in humans. Clin Auton Res 6 (3): 131–140. Oppenheimer, S. M., J. X. Wilson, C. Guiraudon, and D. F. Cechetto. 1991. Insular cortex stimulation produces lethal cardiac arrhythmias: A mechanism of sudden death? Brain Res 550 (1): 115–121.

Sudden Epileptic Death in Experimental Animal Models

611

O’Regan, M. E., and J. K. Brown. 2005. Abnormalities in cardiac and respiratory function observed during seizures in childhood. Dev Med Child Neurol 47 (1): 4–9. Papez, J. W. 1937. A proposed mechanism of emotion. Arch Neurol Psychiatry 38: 725–743. Parent, J. M. 2007. Adult neurogenesis in the intact and epileptic dentate gyrus. Prog Brain Res 163: 529–540. Parent, J. M., R. C. Elliott, S. J. Pleasure, N. M. Barbaro, and D. H. Lowenstein. 2006. Aberrant seizure-induced neurogenesis in experimental temporal lobe epilepsy. Ann Neurol 59 (1): 81–91. Parent, J. M., T. W. Yu, R. T. Leibowitz, D. H. Geschwind, R. S. Sloviter, and D. H. Lowenstein. 1997. Dentate granule cell neurogenesis is increased by seizures and contributes to aberrant network reorganization in the adult rat hippocampus. J Neurosci 17 (10): 3727–3738. Parker, Jr., I. T., C. G. Gunn, and T. N. Lynn. 1962. Experimental centrogenic arrhythmias. Clin Res 10: 179. Pericic, D., J. Lazic, and D. Svob Strac. 2005. Anticonvulsant effects of acute and repeated fluoxetine treatment in unstressed and stressed mice. Brain Res 1033 (1): 90–95. Phizackerley, P. J., E. W. Pool, and C. W. Whitty. 1954. Sino-auricular heart block as an epileptic manifestation; a case report. Epilepsia 3: 89–91. Polenov, A. L., and V. K. Chetverukhin. 1993. Ultrastructural radioautographic analysis of neurogenesis in the hypothalamus of the adult frog, Rana temporaria, with special reference to physiological regeneration of the preoptic nucleus: II. Types of neuronal cells produced. Cell Tissue Res 271 (2): 351–362. Pool, J. L., and J. Ransohoff. 1949. Autonomic effects on stimulating rostral portion of cingulate gyri in man. J Neurophysiol 12 (6): 385–392. Prince, D. A. 1969. Microelectrode studies of penicillin foci. In Basic Mechanisms of the Epilepsies, ed. H. H. Jasper, A. A. Ward, and A. Pope. Boston, MA: Little Brown. Prince, D. A. 1972. Topical convulsant drug and metabolic antagonists. In Experimental Model of Epilepsy. A Manual for the Laboratory Worker, ed. D. P. Purpura, J. K. Penry, D. B. Tower, D. M. Woodbury, and R. D. Walter. New York, NY: Raven Press. Purpura, D. P., J. L. Pool, E. M. Housepian, M. Girado, and S. A. Jacobson. 1958. Hypothermic potentiation of centrally induced cardiac irregularities. Anesthesiology 19 (1): 27–37. Ramabadran, K., and M. Bansinath. 1990. Endogenous opioid peptides and epilepsy. Int J Clin Pharmacol Ther Toxicol 28 (2): 47–62. Saper, C. B., A. D. Loewy, L. W. Swanson, and W. M. Cowan. 1976. Direct hypothalamo-autonomic connections. Brain Res 117 (2): 305–312. Sathyaprabha, T. N., P. Satishchandra, K. Netravathi, S. Sinha, K. Thennarasu, and T. R. Raju. 2006. Cardiac autonomic dysfunctions in chronic refractory epilepsy. Epilepsy Res 72 (1): 49–56. Scharfman, H. E., J. H. Goodman, and A. L. Sollas. 2000. Granule–like neurons at the hilar/CA3 border after status epilepticus and their synchrony with area CA3 pyramidal cells: Functional implications of seizure-induced neurogenesis. J Neurosci 20 (16): 6144–6158. Schernthaner, C., G. Lindinger, K. Potzelberger, K. Zeiler, and C. Baumgartner. 1999. Autonomic epilepsy—The influence of epileptic discharges on heart rate and rhythm. Wien Klin Wochenschr 111 (10): 392–401. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3 (1): 55–62. Schwartz, R. D., and C. M. Lathers. 1990. GABA neurotransmission, epileptogenic activity, and cardiac arrhythmias. In Epilepsy and Sudden Death, Chap. 17, C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker.

612 Sudden Death in Epilepsy: Forensic and Clinical Issues Scorza, F. A., R. M. Cysneiros, R. M. Arida, C. A. Scorza, A. C. de Almeida, B. Schmidt, and E. A. Cavalheiro. 2008. Adult hippocampal neurogenesis and sudden unexpected death in epilepsy: Reality or just an attractive history? Med Hypotheses 71 (6): 914–922. Seeck, M., O. Blanke, P. Jallon, F. Picard, and S. Zaim. 2001. Symptomatic postictal cardiac asystole in a young patient with partial seizures. Europace 3 (3): 247–252. Simon, R. P., M. J. Aminoff, and N. L. Benowitz. 1984. Changes in plasma catecholamines after tonic–clonic seizures. Neurology 34 (2): 255–257. Smith, W. S., and M. A. Matthay. 1997. Evidence for a hydrostatic mechanism in human neurogenic pulmonary edema. Chest 111 (5): 1326–1333. Sperling, M. R., H. Feldman, J. Kinman, J. D. Liporace, and M. J. O’Connor. 1999. Seizure control and mortality in epilepsy. Ann Neurol 46 (1): 45–50. St. John, W. M., A. H. Rudkin, G. L. Homes, and J. C. Leiter. 2006. Changes in respiratory-modulated neural activities, consistent with obstructive and central apnea, during ἀctive seizures in an in situ anaesthetized rat preparation. Epilepsy Res 70 (2–3): 218–228. Standish, A., L. W. Enquist, J. A. Escardo, and J. S. Schwaber. 1995. Central neuronal circuit innerÂ� vating the rat heart deἀned by transneuronal transport of pseudorabies virus. J Neurosci 15 (3€Pt 1): 1998–2012. Standish, A., L. W. Enquist, and J. S. Schwaber. 1994. Innervation of the heart and its central medullary origin deἀned by viral tracing. Science 263 (5144): 232–234. Stanἀeld, B. B., and J. E. Trice. 1988. Evidence that granule cells generated in the dentate gyrus of adult rats extend axonal projections. Exp Brain Res 72 (2): 399–406. Sterns, R. H., J. Guzzo, and P. U. Feig. 1981. The disposition of intravenous potassium in normal man: The role of insulin. Clin Sci (Lond) 61 (1): 23–28. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15 (1): 15–30. Taggart, P., M. Carruthers, and W. Somerville. 1973. Electrocardiogram, plasma catecholamines and lipids, and their modiἀcation by oxyprenolol when speaking before an audience. Lancet 2 (7825): 341–346. Tavee, J., and H. Morris, 3rd. 2008. Severe postictal laryngospasm as a potential mechanism for sudden unexpected death in epilepsy: A near-miss in an EMU. Epilepsia 49 (12): 2113–2117. Taylor, E. W., D. Jordan, and J. H. Coote. 1999. Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiol Rev 79 (3): 855–916. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36 (1): 29–36. Terrence, C. F., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann Neurol 9 (5): 458–464. Terrence, Jr., C. F., H. M. Wisotzkey, and J. A. Perper. 1975. Unexpected, unexplained death in epileptic patients. Neurology 25 (6): 594–598. Tinuper, P., F. Bisulli, A. Cerullo, R. Carcangiu, C. Marini, G. Pierangeli, and P. Cortelli. 2001. Ictal bradycardia in partial epileptic seizures: Autonomic investigation in three cases and literature review. Brain 124 (Pt 12): 2361–2371. Toichi, M., T. Murai, A. Sengoku, and K. Miyoshi. 1998. Interictal change in cardiac autonomic function associated with EEG abnormalities and clinical symptoms: A longitudinal study following acute deterioration in two patients with temporal lobe epilepsy. Psychiatry Clin Neurosci 52 (5): 499–505. Tomson, T. 2000. Mortality in epilepsy. J Neurol 247 (1): 15–21. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. Ueda, H. 1962. Arrhythmias produced by cerebral stimulation. Jpn Circ J 26: 225–230.

Sudden Epileptic Death in Experimental Animal Models

613

Van Bogaert, L. 1936. Hypothamaus et reacations cardiovascularies d’origine centrale. Arch Int Pharmacodyn 52: 137–174. Van Buren, J. M. 1958. Some autonomic concomitants of ictal automatism; a study of temporal lobe attacks. Brain 81 (4): 505–528. Venit, E. L., B. D. Shepard, and T. N. Seyfried. 2004. Oxygenation prevents sudden death in seizureprone mice. Epilepsia 45 (8): 993–996. Vick, R. L., E. P. Todd, and D. W. Luedke. 1972. Epinephrine-induced hypokalemia: Relation to liver and skeletal muscle. J Pharmacol Exp Ther 181 (1): 139–146. Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. D. Cascino, J. F. Annegers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56 (4): 519–525. Walker, A. E. 1966. Internal structure and afferent–efferent relations of the thalamus. In The Thalamus, ed. D. P. Purpura and M. D. Yahr. New York, NY: Columbia University Press. Wall, P. D., and G. D. Davis. 1951. Three cerebral cortical systems affecting autonomic function. J Neurophysiol 14 (6): 507–517. Wang, J., Y. Chen, K. Li, and L. Hou. 2006. Blockade of inhibitory neurotransmission evoked seizurelike ἀring of cardiac parasympathetic neurons in brainstem slices of newborn rats: Implications for sudden deaths in patients of epilepsy. Epilepsy Res 70 (2–3): 172–183. Weinberg, S. J., and J. M. Fuster. 1960. Electrocardiographic changes produced by localized hypothalamic stimulations. Ann Intern Med 53: 332–341. Weinstein, M. D., and C. Albertario. 2000. Cardiac asystole and bradycardia as a manifestation of left temporal lobe complex partial seizure. Ann Intern Med 132 (2): 165–166. Wilder-Smith, E. 1992. Complete atrio-ventricular conduction block during complex partial seizure. J Neurol Neurosurg Psychiatry 55 (8): 734–736. Williams, M. E., R. M. Rosa, P. Silva, R. S. Brown, and F. H. Epstein. 1984. Impairment of extrarenal potassium disposal by alpha-adrenergic stimulation. N Engl J Med 311 (3): 145–149. Willis, W. L., and R. G. Grossman. 1981. Medical Neurobiology. St. Louis, MO: Mosby. Willott, J. F., and K. R. Henry. 1976. Roles of anoxia and noise-induced hearing loss in the postictal refractory period for audiogenic seizures in mice. J Comp Physiol Psychol 90 (4): 373–381. Zhang, Z. H., and S. M. Oppenheimer. 2000. Baroreceptive and somatosensory convergent thalamic neurons project to the posterior insular cortex in the rat. Brain Res 861 (2): 241–256. Zhang, Z. H., S. Rashba, and S. M. Oppenheimer. 1998. Insular cortex lesions alter baroreceptor sensitivity in the urethane-anesthetized rat. Brain Res 813 (1): 73–81. Zhao, C., W. Deng, and F. H. Gage. 2008. Mechanisms and functional implications of adult neurogenesis. Cell 132 (4): 645–660.

Sympathetic Nervous System Dysregulation of Cardiac Function and Myocyte Potassium Channel Remodeling in Rodent Seizure Models Candidate Mechanisms for SUDEP

38

Steven L. Bealer Cameron S. Metcalf Jason G. Little Matteo Vatta Amy Brewster Anne E. Anderson

Contents 38.1 Introduction 38.2 Effects of Seizure Activity on Cardiac Sympathovagal Balance 38.3 Effects of Seizures on Cardiac Responses to Environmental Stress 38.4 Effects of Seizure Activity on Cardiac Ion Channels 38.5 Discussion References

615 617 620 621 622 623

38.1â•…Introduction Although the precise causes of SUDEP have not been completely deἀned, a number of studies indicate that cardiac ventricular abnormalities, and the resulting arrhythmias and sudden cardiac death may contribute (Dasheiff 1991; Lathers and Schraeder 2002; Leung et al. 2006; Nei et al. 2004; P-Codrea Tigaran et al. 2005). Indeed, it has been proposed that cumulative ischemic damage resulting from sympathetic nervous system activation during the repeated seizures characteristic of intractable epilepsy may progressively increase susceptibility to ventricular arrhythmias and SUDEP (McGugan 1999; Shimizu et al. 2008; Tigaran et al. 1997). The proposal that multiple seizures enhance the risk of cardiac-related death is supported by numerous reports that patients with uncontrolled, intractable seizures are at the greatest risk of SUDEP (Nilsson et al. 1999; Opeskin et al. 1999; Opeskin and Berkovic 2003; Tennis et al. 1995).

615

616 Sudden Death in Epilepsy: Forensic and Clinical Issues

Lethal ventricular arrhythmias can be induced by repeated, acute, transient elevations in sympathetic nervous system activity, as well as by chronic increases in sympathetic dominance of cardiac function, particularly acting on a background of cardiac structural changes (Anderson 2003; Campbell 1991). For example, a number of recent studies demonstrate that both the magnitude and duration of sympathetic nervous system responses to experimental stressors are enhanced in pathological conditions characterized by increased vulnerability to lethal arrhythmias, including postmyocardial infarction (CudnochJedrzejewska et al. 2007; Dobruch, Cudnoch-Jedrzejewska, and Szczepanska-Sadowska 2005), experimental obesity (D’Angelo et al. 2006), depression (Grippo et al. 2003, 2006), and hypertension (Giulumian et al. 1999; McDougall et al. 2005). Consequently, it is evident that both enhanced basal cardiac sympathetic nervous system tone, as well as large, transient increases in activation can increase risk of sudden cardiac death. Seizures in epilepsy are accompanied by intense stimulation of the sympathetic nervous system (Simon et al. 1984), with a resulting increase in both heart rate and blood pressure (Di Gennaro et al. 2004; Mayer et al. 2004; Rugg-Gunn et al. 2004). In addition to sympathetic nervous system activity during seizures, it is possible that responses to other environmental stressors are enhanced in patients with epilepsy. Although one recent epidemiological study reported mental stress as a risk factor for SUDEP (Lear-Kaul et al. 2005), the relationship between sympathetic nervous system activation in response to environmental stress and seizure activity has not been determined. Recently, Lathers and Schraeder (2006) elucidated the lack of knowledge in this area and the potential importance of evaluating the relationship between stress and SUDEP. The repeated activation of the sympathetic nervous system during seizures, particularly in conjunction with exaggerated autonomic responses to environmental stressors, may induce progressive cardiac deterioration and contribute to cardiac arrhythmic activity observed during the ictal and immediate postictal period (Rugg-Gunn et al. 2004), which could progress to SUDEP. However, the relationship between seizure activity, sympathetic nervous system tone, and reactivity, in control of the heart in patients with seizure disorders, has not been completely elucidated. The balance between the influence of the sympathetic (sympathetic nervous system) and parasympathetic components of the autonomic nervous system, known as sympathovagal balance or the vagal-sympathetic effect (sympathovagal balance) (Goldberger 1999), is a major regulator of cardiac function. Normally, parasympathetic nervous system (or vagal) tone predominates and reduces the proarrhythmic effects of the sympathetic nervous system activity (Anderson 2003), and activation of the sympathetic nervous system, which occurs during stress or exercise, is normally short-lived and highly regulated. However, pathological conditions such as myocardial infarction, congestive heart failure, and coronary artery disease are associated with chronic alterations in normal sympathovagal balance, characterized by dominance of sympathetic nervous system activity, and an associated increased risk of cardiac arrhythmias (La Rovere et al. 2001; La Rovere and Schwartz 1997; Stein and Kleiger 1999; Stein et al. 1994; Vanoli et al. 2008; Liao et al. 1996; Verrier and Antzelevitch 2004). Therefore, a shift in sympathovagal balance toward sympathetic nervous system dominance, either by increased sympathetic nervous system or decreased parasympathetic nervous system activity, can contribute to increased cardiac risk. Sympathovagal balance can be determined by comparing the normal, control heart rate to the intrinsic heart rate. Intrinsic heart rate is deἀned as the heart rate in the absence of autonomic influence and represents the intrinsic rate of the cardiac pacemaker. If the intrinsic heart rate is greater than the control heart rate, the heart is being predominantly

Sympathetic Nervous System Dysregulation of Cardiac Function

617

influenced by the parasympathetic nervous system. However, if the intrinsic heart rate is less than the control heart rate, the sympathetic nervous system is providing the dominant influence on cardiac rate. There are some indirect indications that sympathovagal balance may be altered in the direction of sympathetic nervous system dominance as a result of seizure activity. For example, heart rate variability, the spontaneous fluctuations in the intervals between heart beats, is decreased in patients with established epilepsy (Ansakorpi et al. 2002; Ronkainen et al. 2006; Tomson et al. 1998). Heart rate variability represents the interaction between the sympathetic nervous system and parasympathetic nervous system on the cardiac pacemaker (Task Force of the European Society of Cardiology the North American Society of Pacing Electrophysiology 1996; Stein and Kleiger 1999). Decreased heart rate variability is indicative of increased sympathetic dominance of cardiac control (Stein and Kleiger 1999; van Ravenswaaij-Arts et al. 1993), and is associated with cardiomyopathies that increase risk of sudden cardiac death including coronary artery disease (Krittayaphong et al. 1997), congestive heart failure (van Ravenswaaij-Arts et al. 1993), depression (Grippo et al. 2002), and following recovery myocardial infarction (Kleiger et al. 1987). Although heart rate variability is indicative of increased sympathetic dominance of cardiac function, the effects of seizures on sympathovagal balance have not been previously deἀned. As noted earlier, enhanced basal sympathetic nervous system tone and transient increases in catecholamine stimulation can be a facilitatory factor initiating ventricular arrhythmias, particularly when acting on abnormal cardiac tissue (Anderson 2003; Chen et al. 2007; Dorian 2005). However, the cardiac effects of seizures that provide the arrhythmogenic substrate for adrenergic facilitation of potentially lethal arrhythmias has not been identiἀed. One mechanism that increases susceptibility to sudden cardiac death is ion channel remodeling in cardiac myocytes. Ion channel remodeling, secondary to enhanced sympathetic nervous system activation, mimics what occurs in ion channel dysfunction caused by primary mutations in the genes encoding for the ion channel subunits or for channel interacting proteins (Schimpf et al. 2009; Ueda et al. 2008; Vatta et al. 2006; Wu et al. 2008). In diseases such as long QT syndrome, loss of function of potassium channels occurs due to genetic mutations or acquired inactivation subsequent to blood alkalosis, antibiotics, or other drug-induced inactivation. These ion channel defects caused by primary gene defects in ion channels or channel-interacting proteins deἀne the underlying diseases as a channelopathy (Schimpf et al. 2009). We propose that a similar remodeling may occur after acquired prolonged or recurrent seizures, which would predispose to sudden cardiac death. In summary, it is possible that repeated seizure activity may increase basal sympathetic nervous system dominance of cardiac function and enhance sympathetic nervous system responsiveness, inducing ion channel remodeling in cardiac myocytes. These effects would result in increased susceptibility to ventricular arrhythmias, which would contribute to SUDEP.

38.2â•… Effects of Seizure Activity on Cardiac Sympathovagal Balance To determine the effects of seizures on the relative sympathetic nervous system and parasympathetic nervous system control of cardiac function, we evaluated sympathovagal balance after recovery from a single, long-lasting seizure in male, Sprague–Dawley rats.

618 Sudden Death in Epilepsy: Forensic and Clinical Issues

Seizures were induced with lithium–pilocarpine injections and were terminated after 90 min (Metcalf et al. 2009). Either 1 or 2 weeks after seizures, cardiac sympathovagal balance (sympathovagal balance) was determined by comparing intrinsic heart rate to control heart rate in animals that had experienced seizures (Pilo) or control procedures (saline injection, Cont). Furthermore, in addition to evaluating cardiac function in the absence of all autonomic influences, the individual contributions of the sympathetic nervous system and parasympathetic nervous system were individually determined by measuring changes in heart rate after blockade of the sympathetic nervous system with atenolol and blockade of the parasympathetic nervous system with atropine (Goldberger 1999). All cardiovascular measures were obtained from conscious, unrestrained animals. The relationships between intrinsic heart rate and control heart rate, that is, the sympathovagal balance, which represents a measure of sympathovagal balance (Goldberger 1999), observed in pilocarpine-treated and control-treated rats 1 week (panel a) and 2 weeks (panel b) after treatment are shown in Figure 38.1. As mentioned earlier, a lower sympathovagal balance is indicative of increased sympathetic nervous system dominance of cardiac function. At both periods, the sympathovagal balance was signiἀcantly lower in animals that underwent seizures (Figure 38.1). This observation suggests seizure activity resulted in a long-lasting shift in sympathovagal control of cardiac function toward sympathetic nervous system dominance. However, increased sympathetic nervous system influence on cardiac function may result from enhanced sympathetic nervous system tone, decreased parasympathetic nervous system tone, or both. To elucidate the origin of the shift in cardiac sympathovagal balance toward sympathetic nervous system dominance, we evaluated sympathetic nervous system tone (i.e., the fall in heart rate in response to sympathetic nervous system blockade in the presence of parasympathetic nervous system blockade) and parasympathetic nervous system tone (i.e., the rise in heart rate in response to parasympathetic nervous system blockade in the presence of sympathetic nervous system blockade) (Goldberger 1999). The increase in heart rate in response to atropine (parasympathetic nervous system antagonist) in the presence of sympathetic nervous system blockade was signiἀcantly smaller in animals experiencing seizures (pilocarpine) (Figure 38.2) both 1 week (panel a) and 2 weeks (panel c) after treatment compared to control-treated rats. These data indicate that parasympathetic nervous system tone was signiἀcantly lower in animals undergoing seizures, compared to control rats, for at least 2 weeks after treatment. However, the decrease in

1.10 1.05 1.00 0.95 0.90 0.85 0.80

1.10 1.05 1.00 0.95 0.90 0.85 0.80

VSE (iHR/HR)

(b)

VSE (iHR/HR)

(a)

Cont

Pilo

Cont

Pilo

Figure 38.1╇ Vagal-sympathetic effect (sympathovagal balance) in animals undergoing con-

trol procedures (Cont) or seizure (SE) animals 1 week (a) and 2 weeks (b) after treatment. *p < 0.05, compared to Cont. (From Metcalf, C. S., et al., Epilepsia, 50 (4), 747–754, 2009. With permission.)

45 40 35 30 25 20 15 10 5 0

(a) 0 –25 –50 –75

–100

(c)

(d)

50

HR change (beats/min)

HR change (beats/min)

60

40 30 20 10 0

619

(b) HR change (beats/min)

HR change (beats/min)

Sympathetic Nervous System Dysregulation of Cardiac Function

Cont

Pilo

–10 –35 –60 –85 –110

Cont

Pilo

Figure 38.2╇ Changes in heart rate (heart rate) resulting from atropine after atenolol (parasympathetic nervous system tone) 1 week (a) and 2 weeks (c) after treatment. Changes in heart rate resulting from atenolol after atropine (sympathetic nervous system tone) 1 (b) and 2 (d) weeks after treatment. *p < 0.05 compared to Cont. (From Metcalf, C. S., et al., Epilepsia, 50 (4), 747–754, 2009. With permission.)

heart rate in response to atenolol (sympathetic nervous system antagonist) during parasympathetic nervous system blockade was equivalent between pilocarpine and control rats both 1 and 2 weeks after treatment (Figure 38.2, panels b and d, respectively). This ἀnding suggests that seizures did not alter basal cardiac sympathetic nervous system. Taken together, these data demonstrate that seizure activity resulted in a prolonged increase in sympathetic nervous system dominance of cardiac function, resulting primarily from decreased parasympathetic nervous system tone, with little or no alteration in sympathetic nervous system activity. This chronic increase in sympathetic nervous system influence on the heart after seizures provides support for the proposal that decreased heart rate variability observed in patients with established epilepsy (Ansakorpi et al. 2002; Ronkainen et al. 2006; Tomson et al. 1998) results from increased cardiac sympathetic nervous system tone (Stein and Kleiger 1999; van Ravenswaaij-Arts et al. 1993). In addition, consistent with the ἀndings in this animal model, recent clinical studies have reported patients with intractable epilepsies exhibited higher sympathetic nervous system tone, lower parasympathetic nervous system tone, and signiἀcant autonomic dysregulation (Mukherjee et al. 2009; Sathyaprabha et al. 2006). Because susceptibility to lethal ventricular arrhythmias is enhanced by chronic, sustained increases in sympathetic nervous system control of cardiac function (Airaksinen 1999; Anderson 2003; Chen et al. 2007), these data support the proposal that seizure-induced changes in basal sympathovagal balance predispose patients to SUDEP.

620 Sudden Death in Epilepsy: Forensic and Clinical Issues

38.3â•…Effects of Seizures on Cardiac Responses to€Environmental Stress In addition to the arrhythmogenic effects of chronic, basal sympathetic nervous system dominance on cardiac function, lethal ventricular arrhythmias can also be precipitated by more acute and transient elevations in sympathetic nervous system activity (Anderson 2003; Campbell 1991). Enhanced sympathetic nervous system responses to environmental stressors, and the resulting increase in catecholamine stimulation of the heart, is characteristic of cardiac pathologies that often result in sudden cardiac death (CudnochJedrzejewska et al. 2007; D’Angelo et al. 2006; Dobruch et al. 2005; Grippo et al. 2003, 2006; Giulumian et al. 1999; McDougall et al. 2005). Indeed, it has been proposed that enhanced responses to mental stress may be a risk factor for SUDEP (Lear-Kaul et al. 2005). This would suggest that in addition to the documented activation of the sympathetic nervous system (Simon et al. 1984) and associated increase in heart rate and blood pressure (Di Gennaro et al. 2004; Mayer et al. 2004; Rugg-Gunn et al. 2004) that occurs during seizures, patients with epilepsy may experience more intense sympathetic nervous system activation to non-seizure-related environmental stressors, further increasing the risk of lethal arrhythmias (Anderson 2003; Campbell 1991; Cudnoch-Jedrzejewska et al. 2007). We suggest that environmental stress produces enhanced sympathetic nervous system–mediated cardiac responses in patients with epilepsy that, in conjunction with basal sympathetic nervous system dominance and acute sympathetic nervous system activation during seizures, contributes to progressive deterioration in cardiac function and potentially results in SUDEP. To determine if seizures affect autonomic responses to stress, we measured changes in heart rate and blood pressure before and during administration of a moderate, environmental stressor in animals 10 days to 2 weeks after seizures induced by lithium– pilocarpine treatment. Blood pressure and heart rate were measured in rats subjected to the moderate stress of an air jet aimed at the animal’s head for 3 min. Figure 38.3 shows control heart rate recorded before administration of the stress (Pre), and the maximum heart rate measured during the 3 min of air jet administration (Stress) in animals undergoing seizures (Pilo) and in control-treated rats. As can be seen, there were no differences in basal heart rate between control and pilocarpine rats before administration of the air jet. Furthermore, although mean heart rate tended to increase in control rats in response to the stress, this tachycardia was not statistically signiἀcant. However, in rats that had experienced seizures, air jet stress increased heart rate signiἀcantly above both prestress levels and the values observed in control animals during stress. In contrast, blood pressure was not elevated by air jet stress in either group of animals (data not shown). These ἀndings suggest that seizure activity produces a chronic increase in cardiac sympathetic nervous system responses to environmental stressors, which is reflected in a signiἀcant tachycardia. This increased cardiac reactivity during stress may contribute to SUDEP in patients with recurrent seizures because a similar relationship between enhanced responses to stress and increased risk of lethal arrhythmias has been reported in other pathological conditions characterized by sudden cardiac death, such as post-myocardial infarction (Cudnoch-Jedrzejewska et al. 2007; Dobruch, Cudnoch-Jedrzejewska, and Szczepanska-Sadowska 2005), experimental obesity (D’Angelo et al. 2006), depression (Grippo et al. 2003, 2006), and hypertension (Giulumian et al. 1999; McDougall et al. 2005).

Sympathetic Nervous System Dysregulation of Cardiac Function 500

621

*#

Cont Heart rate (beats/min)

Pilo 450

400

350

0

Stress

Pre

Pre

Stress

Figure 38.3╇ Basal heart rate measured before stress (Pre), and the maximum heart rate

observed during 3 min of air jet stress (Stress) in control (Cont) rats and animals undergoing seizures induced with lithium–pilocarpine (Pilo). *p < 0.05 compared to Pilo-Pre; #p < 0.05 compared to Control-Stress.

38.4â•… Effects of Seizure Activity on Cardiac Ion Channels We have observed sudden death after kainate convulsant stimulation in Kv4.2 knockout mice (Barnwell et al. 2009). These mice do not have a cardiac phenotype at rest. However, after prolonged seizure stimulation with presumed ongoing sympathetic nervous system dominance based on the above ἀndings, these animals suffered sudden death. Given that Kv4.2 subunits contribute to channels that underlie the transient outward current (Ito,f ) encoded in rodent myocytes, one possible mechanism underlying sudden death in these animals after seizure stimulation is a lethal cardiac arrhythmia. Subsequently, we have 125 Optical density (% Control)

100 75 50 25 0

Control

KA

Kv4.2 Actin

Figure 38.4╇ Reduction in cardiac Kv4.2 channel protein levels after kainate-induced seizures.

Immunoblotting of normalized cardiac membranes prepared from animals with sham treatment (Control) vs. kainate-induced seizures (KA) was performed using Kv4.2 and actin (lane loading control) antibodies. *p < 0.05, kainate-treated animals (KA) compared to Control.

622 Sudden Death in Epilepsy: Forensic and Clinical Issues

begun to evaluate whether remodeling of Kv4 channels occurs in models of acquired seizures or epilepsy. As a ἀrst step in these studies, we evaluated Kv4.2 protein levels in rats after kainate- or pilocarpine-induced seizures. We found signiἀcant decreases in Kv4.2 channel levels in both of these models (Figure 38.4 for kainate model data; pilo model data not shown). Thus, we conclude that ion channel remodeling occurs after seizures and may contribute to sudden death in epilepsy. Although the link remains to be shown, one candidate mechanism underlying ion channel remodeling in acquired epilepsy is through altered sympathetic tone.

38.5â•…Discussion These studies have demonstrated that convulsant-induced seizures in rodents result in long-lasting alterations in autonomic control of cardiac function, sympathetic nervous system responses to stress, and cardiomyocyte potassium channels. Speciἀcally, seizures produce an increase in sympathetic nervous system dominance in cardiac control, resulting from diminished parasympathetic nervous system tone, enhanced sympathetic nervous system–induced tachycardia during environmental stress, and reduced expression of Kv4.2 channels in cardiomyocytes. These effects of seizures would be expected to increase susceptibility to ventricular arrhythmias and sudden cardiac death contributing to SUDEP. Lethal ventricular arrhythmias can be produced when a physiological facilitator interacts with abnormal anatomical and electrical substrates in cardiac tissue (Anderson 2003; Campbell 1991). A number of previous studies demonstrated that sympathetic nervous system dominance of cardiac function and enhanced sympathetic nervous system responses to environmental stimuli are associated with increased risk of sudden death in several cardiac pathologies including post-myocardial infarction (Cudnoch-Jedrzejewska et al. 2007; Dobruch, Cudnoch-Jedrzejewska, and Szczepanska-Sadowska 2005), experimental obesity (D’Angelo et al. 2006), depression (Grippo et al. 2003, 2006), and hypertension (Giulumian et al. 1999; McDougall et al. 2005). These data suggest that excessive catecholaminergic stimulation of cardiac tissue is an important facilitator of arrhythmogenesis in these conditions. Our studies extend these ἀndings by demonstrating that seizures that can result in sudden cardiac death are similarly associated with increases in sympathetic nervous system dominance of cardiac death and exaggerated responses to environmental stressors lasting well beyond the period of seizure activity. In addition, these studies suggest that seizure-induced decreases in Kv4.2 channels may provide the electrical substrate for arrhythmogenic activity and contribute to sudden cardiac death in epilepsy. It has previously been shown that severe reduction in channel function caused by the dominant negative W362F mutation leads to prolonged QT interval, enhanced dispersion of repolarization, and refractoriness (London et al. 2007). In addition, a reduction in Kv4.2 levels may also affect the transmural gradient between epicardium and endocardium, leading to an imbalance in homogeneity of cardiac depolarization and thereby providing a setup for risk of cardiac dysrhythmia. Cardiac ion channel remodeling has been previously associated with autonomic imbalance or primary genetic mutations, which are both important causes of cardiac arrhythmias; however, ion channel remodeling acquired in association with various cardiac pathologies is an additional important cause of cardiac arrhythmias (for review, see Shah et al. 2005). Experimentally induced tachycardia through rapid cardiac pacing has been

Sympathetic Nervous System Dysregulation of Cardiac Function

623

associated with K+ channel remodeling. Speciἀcally, levels of Kv4.2 and Kv4.3 decreased in myocardium after prolonged cardiac pacing in rats (Yamashita et al. 2000). Thus, a disruption of the ἀnely tuned balance of ion channels normally expressed in the heart, which may occur through a variety of candidate mechanisms, is a risk for cardiac arrhythmia. Future studies will no doubt begin to shed light on the role of cardiac ion channel remodeling and the link to altered sympathetic tone in epilepsy and SUDEP. In summary, these animal studies suggest that lethal ventricular arrhythmias that contribute to SUDEP may occur in response to chronic sympathetic nervous system dominance of cardiac function, combined with periodic, transient, and exaggerated catecholaminergic stimulation, acting on cardiac myocytes that are predisposed to arrhythmogenesis due to altered potassium channel function. These data support the proposal that single intense seizures and/or repeated seizure activity may increase the risk of SUDEP in patients with intractable epilepsy by (1) increasing exposure of cardiac tissue to an arrhythmogenic facilitator, catecholamines, and (2) providing the electrical substrate, decreased Kv4.2 channels, on which the substrate acts to induce arrhythmic activity. These data further indicate that patients with epilepsy who are at increased risk of SUDEP could beneἀt from cardioprotective agents shown to diminish the incidence of lethal arrhythmias in other pathological conditions. For example, beta-adrenergic antagonists prevent lethal arrhythmias and detrimental cardiac adaptations associated with several cardiomyopathies, including coronary artery disease, heart failure, and myocardial infarction (Adamson and Gilbert 2006; Dorian 2005; Hohnloser 2005).

References Adamson, P. B., and E. M. Gilbert. 2006. Reducing the risk of sudden death in heart failure with betablockers. J Card Fail 12 (9): 734–746. Airaksinen, K. E. 1999. Autonomic mechanisms and sudden death after abrupt coronary occlusion. Ann Med 31 (4): 240–245. Anderson, K. P. 2003. Sympathetic nervous system activity and ventricular tachyarrhythmias: Recent advances. Ann Noninvasive Electrocardiol 8 (1): 75–89. Ansakorpi, H., J. T. Korpelainen, H. V. Huikuri, U. Tolonen, V. V. Myllyla, and J. I. Isojarvi. 2002. Heart rate dynamics in refractory and well controlled temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 72 (1): 26–30. Barnwell, L. F., J. N. Lugo, W. L. Lee, S. E. Willis, S. J. Gertz, R. A. Hrachovy, and A. E. Anderson. 2009. Kv4.2 knockout mice demonstrate increased susceptibility to convulsant stimulation. Epilepsia 50: 1741–1751. Campbell, R. W. 1991. Predisposing factors for ventricular arrhythmias. J Cardiovasc Pharmacol 17 (Suppl 6): S9–S12. Chen, L. S., S. Zhou, M. C. Fishbein, and P. S. Chen. 2007. New perspectives on the role of autonomic nervous system in the genesis of arrhythmias. J Cardiovasc Electrophysiol 18 (1): 123–127. Cudnoch-Jedrzejewska, A., J. Dobruch, L. Puchalska, and E. Szczepanska-Sadowska. 2007. Interaction of AT1 receptors and V1a receptors-mediated effects in the central cardiovascular control during the post-infarct state. Regul Pept 142 (3): 86–94. D’Angelo, G., J. D. Mintz, J. E. Tidwell, A. M. Schreihofer, D. M. Pollock, and D. W. Stepp. 2006. Exaggerated cardiovascular stress responses and impaired beta-adrenergic-mediated pressor recovery in obese Zucker rats. Hypertension 48 (6): 1109–1115. Dasheiff, R. M. 1991. Sudden unexpected death in epilepsy: A series from an epilepsy surgery program and speculation on the relationship to sudden cardiac death. J Clin Neurophysiol 8 (2): 216–222.

624 Sudden Death in Epilepsy: Forensic and Clinical Issues Di Gennaro, G., P. P. Quarato, F. Sebastiano, V. Esposito, P. Onorati, L. G. Grammaldo, G. N. Meldolesi, et al. 2004. Ictal heart rate increase precedes EEG discharge in drug-resistant mesial temporal lobe seizures. Clin Neurophysiol 115 (5): 1169–1177. Dobruch, J., A. Cudnoch-Jedrzejewska, and E. Szczepanska-Sadowska. 2005. Enhanced involvement of brain vasopressin V1 receptors in cardiovascular responses to stress in rats with myocardial infarction. Stress 8 (4): 273–284. Dorian, P. 2005. Antiarrhythmic action of beta-blockers: Potential mechanisms. J Cardiovasc Pharmacol Ther 10 (Suppl 1): S15–S22. Giulumian, A. D., S. G. Clark, and L. C. Fuchs. 1999. Effect of behavioral stress on coronary artery relaxation altered with aging in BHR. Am J Physiol 276 (2 Pt 2): R435–R440. Goldberger, J. J. 1999. Sympathovagal balance: How should we measure it? Am J Physiol. 276 (4 Pt 2): H1273–H1280. Grippo, A. J., T. G. Beltz, and A. K. Johnson. 2003. Behavioral and cardiovascular changes in the chronic mild stress model of depression. Physiol Behav 78 (4–5): 703–710. Grippo, A. J., T. G. Beltz, R. M. Weiss, and A. K. Johnson. 2006. The effects of chronic fluoxetine treatment on chronic mild stress-induced cardiovascular changes and anhedonia. Biol Psychiatry 59 (4): 309–316. Grippo, A. J., J. A. Moffitt, and A. K. Johnson. 2002. Cardiovascular alterations and autonomic imbalance in an experimental model of depression. Am J Physiol Regul Integr Comp Physiol 282 (5): R1333–R1341. Hohnloser, S. H. 2005. Ventricular arrhythmias: Antiadrenergic therapy for the patient with coronary artery disease. J Cardiovasc Pharmacol Ther 10 (Suppl 1): S23–S31. Kleiger, R. E., J. P. Miller, J. T. Bigger Jr., and A. J. Moss. 1987. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 59 (4): 256–262. Krittayaphong, R., W. E. Cascio, K. C. Light, D. Sheffield, R. N. Golden, J. B. Finkel, G. Glekas, G.€G. Koch, and D. S. Sheps. 1997. Heart rate variability in patients with coronary artery disease: Differences in patients with higher and lower depression scores. Psychosom Med 59 (3): 231–235. La Rovere, M. T., G. D. Pinna, S. H. Hohnloser, F. I. Marcus, A. Mortara, R. Nohara, J. T. Bigger Jr., A. J. Camm, P. J. Schwartz, on behalf of the Autonomic Tone and Reflexes After Myocardial Infarction (ATRAMI) Investigators. 2001. Baroreflex sensitivity and heart rate variability in the identiἀcation of patients at risk for life-threatening arrhythmias: Implications for clinical trials. Circulation 103 (16): 2072–2077. La Rovere, M. T., and P. J. Schwartz. 1997. Baroreflex sensitivity as a cardiac and arrhythmia mortality risk stratiἀer. Pacing Clin Electrophysiol 20 (10 Pt 2): 2602–2613. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42 (2): 123–136. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lear-Kaul, K. C., L. Coughlin, and M. J. Dobersen. 2005. Sudden unexpected death in epilepsy: A retrospective study. Am J Forensic Med Pathol 26 (1): 11–17. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9 (1): 19–30. Liao, D., J. Cai, R. W. Barnes, H. A. Tyroler, P. Rautaharju, I. Holme, and G. Heiss. 1996. Association of cardiac autonomic function and the development of hypertension: The ARIC study. Am J Hypertens 9 (12 Pt 1): 1147–1156. London, B., L. C. Baker, P. Petkova-Kirova, J. M. Nerbonne, B. R. Choi, and G. Salama. 2007. Dispersion of repolarization and refractoriness are determinants of arrhythmia phenotype in transgenic mice with long QT. J Physiol 578 (Pt 1): 115–129. Mayer, H., F. Benninger, L. Urak, B. Plattner, J. Geldner, and M. Feucht. 2004. EKG abnormalities in children and adolescents with symptomatic temporal lobe epilepsy. Neurology 63 (2): 324–328.

Sympathetic Nervous System Dysregulation of Cardiac Function

625

McDougall, S. J., R. E. Widdop, and A. J. Lawrence. 2005. Differential gene expression in WKY and SHR brain following acute and chronic air-puff stress. Brain Res Mol Brain Res 133 (2): 329–336. McGugan, E. A. 1999. Sudden unexpected deaths in epileptics—A literature review. Scott Med J 44 (5): 137–139. Metcalf, C. S., P. B. Radwanski, and S. L. Bealer. 2009. Status epilepticus produces chronic alterations in cardiac sympathovagal balance. Epilepsia 50 (4): 747–754. Mukherjee, S., M. Tripathi, P. S. Chandra, R. Yadav, N. Choudhary, R. Sagar, R. Bhore, R. M. Pandey, and K. K. Deepak. 2009. Cardiovascular autonomic functions in well-controlled and intractable partial epilepsies. Epilepsy Res 85 (2–3): 261–269. Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 353 (9156): 888–893. Opeskin, K., M. P. Burke, S. M. Cordner, and S. F. Berkovic. 1999. Comparison of antiepileptic drug levels in sudden unexpected deaths in epilepsy with deaths from other causes. Epilepsia 40 (12): 1795–1798. Opeskin, K., and S. F. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12 (7): 456–464. P-Codrea Tigaran, S., S. Dalager-Pedersen, U. Baandrup, M. Dam, and A. Vesterby-Charles. 2005. Sudden unexpected death in epilepsy: Is death by seizures a cardiac disease? Am J Forensic Med Pathol 26 (2): 99–105. Ronkainen, E., J. T. Korpelainen, E. Heikkinen, V. V. Myllyla, H. V. Huikuri, and J. I. Isojarvi. 2006. Cardiac autonomic control in patients with refractory epilepsy before and during vagus nerve stimulation treatment: A one-year follow-up study. Epilepsia 47 (3): 556–562. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Sathyaprabha, T. N., P. Satishchandra, K. Netravathi, S. Sinha, K. Thennarasu, and T. R. Raju. 2006. Cardiac autonomic dysfunctions in chronic refractory epilepsy. Epilepsy Res 72 (1): 49–56. Schimpf, R., C. Veltmann, C. Wolpert, and M. Borggrefe. 2009. Channelopathies: Brugada syndrome, long QT syndrome, short QT syndrome, and CPVT. Herz 34 (4): 281–288. Shah, M., F. G. Akar, and G. F. Tomaselli. 2005. Molecular basis of arrhythmias. Circulation 112 (16): 2517–2529. Shimizu, M., A. Kagawa, T. Takano, H. Masai, and Y. Miwa. 2008. Neurogenic stunned myocardium associated with status epileptics and postictal catecholamine surge. Intern Med 47 (4): 269–273. Simon, R. P., M. J. Aminoff, and N. L. Benowitz. 1984. Changes in plasma catecholamines after tonic–clonic seizures. Neurology 34 (2): 255–257. Stein, P. K., M. S. Bosner, R. E. Kleiger, and B. M. Conger. 1994. Heart rate variability: A measure of cardiac autonomic tone. Am Heart J 127 (5): 1376–1381. Stein, P. K., and R. E. Kleiger. 1999. Insights from the study of heart rate variability. Annu Rev Med 50: 249–261. Task Force of the European Society of Cardiology the North American Society of Pacing Electrophysiology. 1996. Heart rate variability: Standards of measurement, physiological interpretation and clinical use. Circulation 93 (5): 1043–1065. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36 (1): 29–36. Tigaran, S., V. Rasmussen, M. Dam, S. Pedersen, H. Hogenhaven, and B. Friberg. 1997. ECG changes in epilepsy patients. Acta Neurol Scand 96 (2): 72–75. Tomson, T., M. Ericson, C. Ihrman, and L. E. Lindblad. 1998. Heart rate variability in patients with epilepsy. Epilepsy Res 30 (1): 77–83.

626 Sudden Death in Epilepsy: Forensic and Clinical Issues Ueda, K., C. Valdivia, A. Medeiros-Domingo, D. J. Tester, M. Vatta, G. Farrugia, M. J. Ackerman, and J. C. Makielski. 2008. Syntrophin mutation associated with long QT syndrome through activation of the nNOS-SCN5A macromolecular complex. Proc Natl Acad Sci U S A 105 (27): 9355–9360. van Ravenswaaij-Arts, C. M., L. A. Kollee, J. C. Hopman, G. B. Stoelinga, and H. P. van Geijn. 1993. Heart rate variability. Ann Intern Med 118 (6): 436–447. Vanoli, E., P. B. Adamson, R. D. Foreman, and P. J. Schwartz. 2008. Prediction of unexpected sudden death among healthy dogs by a novel marker of autonomic neural activity. Heart Rhythm 5 (2): 300–305. Vatta, M., M. J. Ackerman, B. Ye, J. C. Makielski, E. E. Ughanze, E. W. Taylor, D. J. Tester, R. C. Balijepalli, J. D. Foell, Z. Li, T. J. Kamp, and J. A. Towbin. 2006. Mutant caveolin-3 induces persistent late sodium current and is associated with long-QT syndrome. Circulation 114 (20): 2104–2112. Verrier, R. L., and C. Antzelevitch. 2004. Autonomic aspects of arrhythmogenesis: The enduring and the new. Curr Opin Cardiol 19 (1): 2–11. Wu, G., T. Ai, J. J. Kim, B. Mohapatra, Y. Xi, Z. Li, S. Abbasi, E. Purevjav, K. Samani, M. J. Ackerman, M. Qi, A. J. Moss, W. Shimizu, J. A. Towbin, J. Cheng, and M. Vatta. 2008. alpha-1-syntrophin mutation and the long-QT syndrome: A disease of sodium channel disruption. Circ Arrhythm Electrophysiol 1 (3): 193–201. Yamashita, T., Y. Murakawa, N. Hayami, E. Fukui, Y. Kasaoka, M. Inoue, and M. Omata. 2000. Shortterm effects of rapid pacing on mRNA level of voltage-dependent K(+) channels in rat atrium: Electrical remodeling in paroxysmal atrial tachycardia. Circulation 101 (16): 2007–2014.

The Urethane/Kainate Seizure Model as a Tool to Explore Physiology and Death Associated with Seizures

39

Mark Stewart

Contents 39.1 Introduction 39.2 Autonomic Consequences of Seizures 39.3 The Urethane/Kainate Model 39.4 Results Relevant to Sudden Death in Epilepsy 39.5 Quantitative Activity Differences 39.6 Pathways 39.7 Insights into the Mechanism for Sudden Death in Epilepsy 39.8 Closing Comments on This and Other Animal Models References

627 628 629 630 636 637 638 639 639

39.1â•…Introduction The anesthetized whole animal or acute preparation has been supplanted in recent decades by a choice of preparations that offer different advantages. Brain slices and cell cultures offer much better access to the neurons for physiological and pharmacological studies; and provided the questions addressed relate to relatively simple neuronal circuits, the preparations are superior to an intact animal preparation—the specimen visualization and recording advantages (intracellular, whole cell patch, calcium and voltage sensitive dyes, neuronal imaging) are signiἀcant. In preparations where the behavior is important (e.g., to observe convulsions or maze learning), the freely moving, freely behaving animal preparation is the only choice. The acute animal preparation in fact has signiἀcant advantages for neurophysiology and physiology studies. Among these is access to measurements and recordings that can be impossible in freely moving animals. In the whole animal preparation, one has access to the entire central nervous system circuitry, something that cannot be preserved in a brain slice or cell culture. The acute animal preparation has been reexamined by us for studies of the autonomic consequences of seizures because we need access to the entire physiology of the animal. One has options for a host of invasive measures or procedures and considerable control of an animal’s condition (e.g., ventilation, temperature, blood pressure, and even blood volume). It is the preparation to explore the limits of the physiology associated with functional or anatomical neuropathology. One can directly address questions such as, Can a seizure 627

628 Sudden Death in Epilepsy: Forensic and Clinical Issues

cause death? If so, how? If not, why not? Can autonomic nervous system activity itself cause death? These questions and many others can be explored in detail to address fundamental questions of physiology. In addition to its flexibility, the acute animal preparation offers efficiency that comes from the control of conditions such as seizure activity to eliminate issues of “capturing” the right conditions. The trade-off, of course, is that although the animal is intact, the conditions of the animal are always a subset of the conditions that would be observed during “captured” events in behaving animals. In short, the acute whole animal preparation is ideally suited for efficiently deἀning boundary conditions and making certain kinds of manipulations or measurements that would later guide studies of activity such as seizures in behaving animals.

39.2â•… Autonomic Consequences of Seizures In the context of sudden death in epilepsy, most would agree that the autonomic nervous system provides the link between seizures and death. Autonomic dysfunction during seizures can have serious clinical consequences (Devinsky 2004; Goodman et al. 2008). Changes in heart rate and rhythm and blood pressure can occur during complex partial seizures (Baumgartner et al. 2001; Lathers et al. 1998; Locatelli et al. 1999; Nei et al. 2000; Wilder-Smith 1992). Generalized tonic–clonic seizures are sometimes associated with severe increases in blood pressure and arrhythmias (Schraeder and Lathers 1989), and some individuals experience more ominous autonomic derangements such as marked bradycardia or asystole (Smith-Demps and Jagoda 1998; Cheung and Hachinski 2000; Mameli et al. 2001). The causes of sudden unexpected death in epilepsy (SUDEP) are suspected to involve cardiovascular or respiratory dysfunction provoked by seizures (Cheung and Hachinski 2000; Langan 2000; Liedholm and Gudjonsson 1992; Mameli et al. 2001; SmithDemps and Jagoda 1998; Tinuper et al. 2001). Animal experimentation has replicated features of human epilepsy and autonomic dysfunction associated with seizures (reviewed by Lathers and Schraeder 2006). Amygdalakindled seizures in rats coactivate sympathetic and parasympathetic systems and were shown to produce ictal hypertension and bradycardia (Goodman et al. 1990, 1999). Left insular cortex stimulation in rats caused degrees of heart block (Oppenheimer et al. 1991), resembling ἀndings from a study of epilepsy surgery patients (Oppenheimer et al. 1992; Hilz et al. 2001). In our own model (described in more detail below), left-sided limbic seizure activity was associated with a more pronounced vagus nerve activity increase and associated blood pressure decrease (Saito et al. 2006) than right-sided limbic seizure activity. Focal and generalized seizures in anesthetized cats (Schraeder and Lathers 1983; Lathers et al. 1998, 1993) were associated with cardiac sympathetic and parasympathetic activity that was intermittently synchronized with interictal spike activity (“lockstep phenomena”) (Lathers et al. 1987; Stauffer et al. 1989). Application of penicillin G to the hypothalamus and mesencephalic centers of hemispherectomized rats produced increases in vagus nerve activity and various cardiovascular disturbances (Mameli et al. 2001, 1993) including death (Mameli et al. 2001, 2006). Mice that exhibit audiogenic seizures frequently display respiratory arrest that can be fatal and prevented by increasing oxygen in the environment (Venit et al. 2004; Willott and Henry 1976). Hypoventilation was reported as an important correlate of death during seizures in a sheep model (Johnston et€al. 1997, 1995).

The Urethane/Kainate Seizure Model A

Urethane

B

Ketamine+xylazine

Hippo EEG Face EMG Leg EMG

Hippo EEG Face EMG Leg EMG

629

5s

1s

Figure 39.1╇ Urethane permits limbic cortical seizure activity with minimal motor convul-

sions. (A) Urethane-anesthetized rat (1.5 g/kg i.p. with intra-arterial (i.a.) supplements). Three 50-s sweeps of simultaneously recorded (a) EEG (dorsal hippocampus area CA1), (b) EMG from face (in the area of vibrissae), and (c) EMG from the lower leg (gastrocnemius and soleus muscles). Seizure activity after systemic kainic acid (10 mg/kg) is present throughout all three sweeps, with some changes in frequency components. Very low EMG activity levels from leg recordings (third trace in each set) did not change during seizures. Minimal EMG activity is apparent in the face EMG recordings, but rarely (indicated by asterisks in sweeps 2 and 3). (B) Ketamine and xylazine–anesthetized rat (60 mg/kg ketamine + 10 mg/kg xylazine i.p. with i.a. supplements). Similar set of three 50-s sweeps with simultaneously recorded dorsal hippocampal EEG, and EMG from face and hind leg. Seizure activity is more pronounced in some, but not all, of the time shown. Most significantly, there was substantial EMG activity in all recording locations. Thick black bar indicates time period for faster sweep shown below. Synchronous phasic activity in both the face and leg EMG electrodes is clearly apparent. Motor convulsions were stopped by Nembutal (10 mg i.a.). EEG spikes are clipped in the display. Calibrations: time, 5 s for both A and B; voltage, 400 µV for EEG A and B, 600 µV for EMG A face and B leg, 400 µV for EMG A leg, and 300 µV for EMG B face. (From Figure 2 in Saito, T., et al., J Neurosci Methods, 155, 241–50, 2006.)

39.3â•… The Urethane/Kainate Model We have developed an acute rat preparation based on urethane anesthesia and systemic or focal kainic acid to deἀne the autonomic consequences of seizure activity. Initially, we€compared anesthetics such as urethane and ketamine/xylazine combinations for acute seizure studies because these anesthetics permit several kinds of polysynaptic activity (Flecknell

630 Sudden Death in Epilepsy: Forensic and Clinical Issues

1996). Hippocampal theta rhythm is one example of a complex polysynaptic brain oscillation that is readily and well studied in urethane-anesthetized animals (Brankack et€al. 1993; Stewart and Fox 1990). In urethane-anesthetized rats, systemic kainic acid can cause repeated long episodes of seizure activity ranging in duration from 10 s to more than 1€min€(Saito et al. 2006; Sakamoto et al. 2008; Hotta et al. 2009a). Seizures are typically characterized by an initial period of high-frequency activity that progresses to slower, larger amplitude waves until the seizure episode terminates leaving low-voltage activity in the EEG until the next seizure episode. Remarkably, the EMG during seizures in urethaneanesthetized rats is essentially flat from all locations except the face, even during limbic status epilepticus (Saito et al. 2006). Animals breathe spontaneously, but motor convulsions during seizure activity are absent. By contrast, motor convulsions accompany electroencephalographic seizures in animals anesthetized with ketamine/xylazine (Saito et al. 2006). Signiἀcant EMG activity can be recorded from multiple locations, and rhythmic muscle contractions corresponded to periodic activity in the EEG. These differences are highlighted in Figure 39.1. The fact that seizure activity in the urethane/kainate model was conἀned to limbic cortical areas without spread to neocortical areas meant that we had access to the autonomic consequences of seizure activity in an animal preparation that was breathing spontaneously and could be mechanically stable enough for many kinds of recordings. There was no need to paralyze the animal. Titration of the urethane level and the amount of kainic acid ensured mechanical stability and offered some flexibility in the durations of seizure activity, ranging from very brief to frank limbic status epilepticus. The reason for the limbic cortical–neocortical disconnect is unclear. Although urethane has been used for studies of activity in limbic cortical structures (e.g., hippocampal theta rhythm), including cingulate cortex (Feenstra and Holsheimer 1979), urethane has been described as anticonvulsant in somatosensory cortex (Heltovics et al. 1995). There is evidence that urethane can suppress glutamate release at some cortical synapses (Moroni et al. 1981).

39.4â•…Results Relevant to Sudden Death in Epilepsy Using the urethane/kainate model, we have deἀned massive increases in parasympathetic and sympathetic outflow that occur during limbic cortical seizures and the cardiovascular consequences of these changes (Saito et al. 2006; Hotta et al. 2009a; Sakamoto et al. 2008; Stewart 2008, Hotta et al. 2009b). Increases were 8–10 times baseline rates for parasympathetic (vagus nerve) activity and somewhat less in sympathetic pre-€and postganglionic nerve recordings (Sakamoto et al. 2008; Hotta et al. 2009a). Peripheral nerve activity increases during seizures were greater than increases induced by nitroprusside or phenylephrine that produced mean arterial pressure changes of >50 mm Hg. Increases in c-fos expression were found in both sympathetic and parasympathetic medullary regions (as well as hypothalamic areas) (Sakamoto et al. 2008). Changes in nerve activity induced by blood pressure increases or decreases were smaller or absent during tests made in€the€seizure state, indicating that baroreceptor reflex function was impaired during seizures. The normal relations of cardiac sympathetic nerve activity to ventilation (Kollai and Koizumi 1979) were lost when ventilation rate changes were tested during limbic seizures (Hotta et al. 2009a). Both the direction of the nerve activity change and the magnitudes of changes

The Urethane/Kainate Seizure Model

631

were unpredictable during seizures. Acute cardiac dilatation accompanied the bradyarrhythmia that occurred during seizures or seizures plus hypercapnia/hypoxia (Sakamoto et al. 2008). Cardiac dilatation depended on autonomic coactivation (Hotta et al. 2009b) (Figures 39.2 through 39.4). A parallel, nonreciprocal activation of the sympathetic and parasympathetic outflow€to€the heart can occur physiologically. Koizumi and others showed such activity in response to hypoxia (Koizumi et al. 1986; Fukuda et al. 1989) and during hypothalamic stimulation€(Koizumi and Kollai 1981). Coactivation enhanced cardiac output (Koizumi Sweep 1

BP Hippo EEG

Seizure begins

Vagus Sweep 2

0.5 mV 2 mV 20 mm Hg

Sweep 3

Sweep 4 Seizure ends

1s

Figure 39.2╇ Vagus nerve activity during a complete limbic seizure episode. Sweeps from a single kainic acid–induced seizure in a urethane-anesthetized rat. Each individual seizure episode typically begins with a period of very high-frequency activity that progressively slows and synchronizes (making larger amplitude EEG waves). Slowing continues until the large amplitude activity stops. Each panel shows BP, dorsal hippocampal EEG, and vagus nerve activity. Note the changes in frequency components of the seizure as it progresses, and the increase in vagus nerve activity, especially during the middle portion (sweeps 2 and 3). Bursts in the vagus nerve recording associated with late slow spikes in the seizure episode are less apparent than those shown in Figure 39.3. BP increases as the seizure begins, and again as the seizure ends. As the vagus nerve activity increases, there is also a decrease in the pulse pressure. Calibrations (and major divisions) are x-axis (time) = 1 s; y-axis (top, BP) = 20 mm Hg, y-axis (middle, hippocampal EEG) = 2.0 mV, and y-axis (bottom, vagus nerve recording) = 0.5 mV. Note that major divisions in the BP records for sweeps 3 and 4 are 10 mm Hg. Triangles at left indicate 180 mm Hg. Horizontal lines at left edge of some sweeps are where traces were cropped by removing trace labels. Each sweep is 30 s long. (From Figure 1 in Sakamoto, K., et al., Epilepsia, 49, 982–996, 2008.)

632 Sudden Death in Epilepsy: Forensic and Clinical Issues Cardiac sympathetic nerve

First half of seizure episode

EEG ECG

Ventilation 0.2 mV

Second half of seizure episode 0.4 mV 2 mV 10 mm Hg 10 s

Figure 39.3╇ Cardiac sympathetic nerve activity during a limbic seizure episode. Top to bot-

tom: cardiac sympathetic multiunit nerve activity, EEG, ECG, and a record ventilator air movements. There is a modest increase in the cardiac sympathetic nerve activity during the seizure episode. The entire time period is split in half. The first part of the recording is shown above and the second part is shown below. The seizure episode begins approximately halfway through the top set of traces and ends approximately halfway through the bottom set of traces. Pre-seizure vs. seizure comparisons can be made by looking across the top row or comparing the first part of the top traces with the first part of the bottom traces. Similarly, seizure vs. postseizure comparisons can be made by comparing across the bottom set of traces or the second parts of the top and bottom. Calibrations are given for all traces and labeled in the bottom set of traces. (From Figure 1 in Hotta, H., et al., Epilepsia, 50, 923–927, 2009a.)

et€al.€1982).€During periods of intense parasympathetic activity, sympathetic coactivation helped to preserve blood pressure in the face of bradyarrhythmia and thereby facilitate venous return and better ventricular ἀlling (Hotta et al. 2009b). During severe vagus nerve stimulation (≥20 Hz), the baroreceptor reflex response was not sufficient to support ventricular ἀlling, evidenced by the small cavity size in echocardiographic images. Strong sympathetic coactivation, either with pharmacological activation of adrenergic receptors or indirect sympathetic activation via hypercapnia/hypoxia, improved blood pressure, which improved venous return and ventricular ἀlling. The drawback of sympathetic coactivation seems to be impaired contractility (Hotta et al. 2009b). Catecholamines are generally viewed as compounds that increase myocardial contractility but accelerate myocardial relaxation (lusitropy). On echocardiography, however, it appeared that the contractions were actually weaker, suggesting that there is an element of sympathetic enhancement of parasympathetic activity. This is consistent with ἀndings that a background of sympathetic activity intensiἀes cardiac responses to parasympathetic input (Randall et al. 2003) either directly at cardiac myocytes or via interactions within the intrinsic cardiac plexus (Ardell 2004).

The Urethane/Kainate Seizure Model

633

Parasympathetics 6 5 4

KA

KA

KA

3 2 1 0

PE

PE

NP

PE NP

NP

Vagus (all)

Vagus (left)

Vagus (right)

Sympathetics 6 5 4 3 2 1 0

NP

NP PE

PE Cerv gang (all)

KA

KA

KA

Cerv gang (left)

PE NP Cerv gang (right)

6 5 4 3 2

KA NP

1 0

PE Renal (left)

Figure 39.4╇ Relative activity in autonomic nerve recordings. Summary of activity changes using a normalized activity scale (1 = baseline, 2 = 2× baseline, 3 = 4× baseline, 4 = 6× baseline, . . . , 7 = 12× baseline and above) for all recorded nerves. Plots are for parasympathetics (vagus nerve) and sympathetics (cervical sympathetic ganglion, renal sympathetic nerve, grater splanchnic nerve). For the vagus nerve recordings and cervical sympathetic ganglia recordings, right or left side data are also shown. Some recordings from each of these groups did not have the side recorded, so they appear only in the pooled totals. Nitroprusside (NP) and phenylephrine (PE) tested before kainic acid (KA) are shown to the left of the KA data. NP and PE tests made during seizures are shown to the right of the KA data. Stars indicate statistically significant increases or decreases (pre vs. post for that particular treatment). p < 0.0001 for vagus nerve significance levels except PE pre/post for the right vagus nerve data where p < 0.0005. p < 0.0001 for sympathetic nerves and ganglia (all), KA pre/post for cervical sympathetic ganglion (all, left). p < 0.01 for all other significance levels. (Adapted from Table 1 in Sakamoto, K., et al., Epilepsia, 49, 982–996, 2008.)

634 Sudden Death in Epilepsy: Forensic and Clinical Issues BP ECG EEG Ventilation

BP ECG EEG Ventilation 1

2

3

4

5

Figure 39.5╇ “Flat line” EEG in a rat model of death during seizures. Urethane-anesthetized rats receiving kainic acid as a convulsant have recurring limbic cortical seizures. (From Saito, T., et al., J Neurosci Methods, 155, 241–250, 2006; Sakamoto, K., et al., Epilepsia, 49, 982–996, 2008. With permission.) Intense seizure activity or seizures combined with asphyxia produce a profound bradycardia and cardiac dilatation leading to hypoperfusion of the brain, the termination of ongoing seizure activity, and eventually a flatline EEG. The top panel shows the full experiment and specific times are highlighted below. The set of sweeps in each panel are, from top to bottom, arterial blood pressure, ECG, EEG, and a pressure transducer on the ventilator tubing. The output of the ventilator produces upward deflections, and spontaneous “pulling” against the closed air system produces downward deflections. The horizontal bar at the bottom of the top panel shows the time during which the ventilator was stopped. This is also apparent as the switch from positive to negative deflections on the ventilation record. The seizure continues for a short time after the ventilator has stopped (panel 2, bottom), but is arrested in less than 1 min (panel 3, bottom), and the EEG is a flat line by panel 4. Bradycardia and AV block are evident in panel 4. As the force of the attempts to inspire increased (larger negative deflections), there were larger movement artifacts in the ECG and EEG recordings (panel 4, bottom). The combination of restored ventilation and seizure suppression permitted recovery of cardiac performance and brain perfusion, with recovery of brain activity (in this case, additional seizures). Calibrations are BP, 20 mm Hg; ECG, 1 mV; EEG, 2 mV; ventilator, 20 mm Hg; and time, 5 s. Data were collected together with Dr. Harumi Hotta of the Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan. (Taken from Stewart, M., J Neurol Neurosurg Psychiatry, e-letter, February 12, 2008.)

The mechanism of death in these animals was deἀned through ECG, blood pressure, and echocardiographic measures to be profound mechanical dysfunction coupled with sinus bradycardia and AV nodal block leading to hypoperfusion of the brain and ἀnally hypoperfusion of the heart itself (Sakamoto et al. 2008). The physiological picture we ἀnd during seizures resembles the picture found during asphyxiation. The time scale for seizures and/or hypoxia and hypercapnia to impact the heart is seconds, and the development of cardiac dilatation with bradyarrhythmia and eventual death can occur on a time scale of minutes. These changes result from a massive combined activation of the parasympathetic

The Urethane/Kainate Seizure Model Baseline

635

During 2 Hz VNS 0.6

During 5 Hz VNS 0.6

0.7

During 50 Hz VNS

0.4

0.4

1 cm 0.2 s

Baseline

During 50 Hz VNS + re-breathing

0.5

0.8

0.9

Figure 39.6╇ Electrocardiogram and M-mode echocardiographic images during VNS. Top left is a series of beats in the baseline condition. Dots below the panel indicate the times for p waves in the ECG (gray trace toward the bottom of each panel). Low-frequency VNS (2, 5 Hz) is shown in the top row. To the right edge of each panel, the maximal ventricular cavity size is marked. A modest increase in cavity size is evident as the heart rate goes down. In the middle row are two examples of isolated beats during 50 Hz VNS. The smaller size of the left ventricular cavity€is evident and indicated to the right of each panel. VNS (50 Hz) during hypercapnia/hypoxia is associated with cardiac dilatation. A series of tiny vertical white dots at about one-fourth of the way along the sweep (between the left edge of the image and the QRS complex) is the distance calibration of 1 cm between successive dots. Horizontal distance between successive tick marks along the top of each panel is 200 ms. (From Hotta, H., et al., Epilepsia, 50, 923–927, 2009b.)

and sympathetic divisions of the autonomic nervous system that is driven by seizures and compounded by respiratory obstruction or impaired ventilation. One question is whether seizures alone are enough to cause death. In these experiments and in our own kindling studies in freely moving rats, seizures can clearly cause serious arrhythmias, but each episode has been transient and therefore not life-Â�threatening. Changing blood flow (e.g., carotid occlusion) and lowering blood oxygen content (e.g., respiratory obstruction) are both processes that can rapidly terminate seizure activity. Transient unilateral or bilateral carotid occlusion terminated seizure activity in the ipsilateral hemisphere or bilaterally in seconds (Saito et al. 2006). Seizures combined with respiratory obstruction or rebreathing air from a balloon caused cardiovascular dysfunction that terminated seizure activity on a time scale of many seconds to minutes (Sakamoto et al. 2008) (Figure 39.5). Because a purely seizure-driven autonomic “storm” would stop when the brain blood flow was compromised, we used vagus nerve stimulation (VNS) in urethane-anesthetized rats to explore the limits of cardiac dysfunction during autonomic overactivity (Hotta et al. 2009b). Based on our previous ἀndings, vagus nerve ἀring rates about 10 times over baseline can occur during seizures (in a combined sympathetic and parasympathetic activity surge) and the highest rates can be achieved during the combination of seizure

636 Sudden Death in Epilepsy: Forensic and Clinical Issues

and asphyxia, or asphyxia alone (Sakamoto et al. 2008). With VNS, we deἀned a baseline parasympathetic “tone” of about 2 Hz and showed that the most severe bradyarrhythmia was produced by 50 Hz VNS. The cardiac consequences of VNS were enhanced by decreasing body temperature, and we could achieve cardiac standstill for many seconds before ventricular escape rhythms developed in ventilated animals. We showed that blood pressure during VNS is relatively maintained during lower-frequency VNS, but collapses at frequencies ≥20 Hz to dramatically impair ventricular ἀlling. Absolutely critical was the ἀnding that heart rate and blood pressure returned to near baseline levels within several seconds after VNS stopped, even after 5-min stimulus trains. When additional sympathetic activity was produced by systemic alpha- or beta-receptor agonists or by hypercapnia, hypoxia, or hypercapnia plus hypoxia, there was a small improvement in mean arterial pressure, much better venous return leading to cardiac dilatation, but even worse cardiac performance (e.g., fractional shortening, ejection fraction) (Figure 39.6). Abrupt increases in parasympathetic activity on the order of 5 times the background parasympathetic tone were sufficient to produce transient bradyarrhythmias, and abrupt increases on the order of 20 times can produce cardiac standstill, sometimes accompanied by apnea. This can only be the mechanism for death when there is airway obstruction because cardiac performance recovery was always rapid after the parasympathetic discharge stopped and, in the case of seizure activity, the decreased cardiac output during intense parasympathetic activity cannot support seizure activity (Sakamoto et al. 2008; Stewart 2008; Saito et al. 2006). More severe or entirely different autonomic disturbances might result when seizures occur on the background of a damaged brain rather than on the background of a normal brain. Another possibility is that episodes of arrhythmia predispose to more serious arrhythmias (J. A. Armour, personal communication, 2008)

39.5â•… Quantitative Activity Differences The notion that autonomic dysfunction arises from differences in activity of the two hemispheres predates suggestions that asymmetric seizure activity may contribute to sudden death in epileptic patients. The “brain-laterality hypothesis” was proposed to account for cases of sudden cardiac death, wherein patients died during extreme stress (Lane and Jennings 1998; Galin 1974). The brain-laterality hypothesis was that emotional arousal could trigger ventricular ἀbrillation and sudden death by inducing a net lateralized imbalance in the sympathetic outflow to the heart. Hemispheric inactivation studies have supported this idea (Hachinski et al. 1992; Zamrini et al. 1990). Differences in activity in the two hemispheres have also been proposed as a contributor or cause for autonomic disturbances in seizure patients, including death (Hilz et al. 2001; Oppenheimer 2001). Controlled lateralized differences in activity are difficult to ἀnd or cause in studies of seizure activity because seizures can spread rapidly in the brain, and most examples of death during or after a seizure involve generalized seizure activity. A different kind of asymmetry involves differential activation of dorsal and ventral portions of the limbic cortices. Part of this notion comes from data that ventral hippocampal areas have a greater tendency to generate seizure activity than dorsal areas (Gilbert et al. 1985). The connectivity of hippocampal formation areas with brain regions having autonomic impact, for example, insular cortex, is also heaviest from ventral areas (Saper 1982a; Witter et al. 1988).

The Urethane/Kainate Seizure Model

637

Other studies suggest that the volume or extent of the cortical involvement, not lateralization of seizure activity, is important for signiἀcant heart rate changes (Epstein et al. 1992; Britton et al. 2006). The Epstein study not only describes predominantly heart rate increases in patients but also mentions patients who deviated from the majority, showing heart rate slowing and sinus pauses. Ictal bradycardia was found to most often occur in patients with temporal lobe seizures when the seizure activity was bilateral (Britton et al. 2006). Furthermore, the current data suggest that sudden death is not consistently associated with the most severe seizures (Goodman et al. 2008). In fact, in kindling animal models (Goodman et al. 1990; Kunitake and Stewart, unpublished data), arrhythmias were often seen early in the kindling progression (e.g., stage 2 seizures); however, none of these were fatal arrhythmias. Although we have seen autonomic nervous system activity differences depending on the hemisphere exhibiting seizure activity (Saito et al. 2006), our nerve recordings and c-fos data indicate that both divisions of the autonomic nervous system are strongly activated during kainic acid–induced limbic seizure activity in rats. Furthermore, our data indicate that death results from massive combined autonomic nervous system overactivity, with the overall result being a parasympathetic-dominated cardiovascular state. In our opinion, cardiovascular complications depend on the degree and duration of increased autonomic activity, and therefore likely depend on the extent and severity of the seizure plus associated respiratory/ventilation complications. Based on our data, the predominantly parasympathetic effect at the heart is due to quantitative differences between sympathetic and parasympathetic outflow. This aspect of€the physiology appears to be quite consistent, even across various seizure models and various animal species, including rats with chronic seizure conditions. That said, there are a number of reports of seizures associated with a predominantly sympathetic state, even in€the same animal preparation during continuous mechanical ventilation (Lathers and Schraeder 1982). What can account for such autonomic variability? We have shown that sympathetic and parasympathetic activity changes accompanying ventilation rate (Hotta et al. 2009a) or blood pressure (Sakamoto et al. 2008) changes can become unpredictable. These become signiἀcant sources for altering the nerve activity patterns that occur during a seizure. An additional source of variability can be speciἀc seizure activity patterns that result in speciἀc autonomic responses, but we are just beginning to explore speciἀc seizure activity patterns.

39.6â•… Pathways Neocortex (Allen et al. 1991; Saper 1982b, 2000; Yasui et al. 1991) and limbic cortices (Witter and Amaral 2004; Witter et al. 1989) are heavily connected with structures that constitute the autonomic nervous system or with regions that project into autonomic structures. The central nucleus of the amygdala and bed nucleus of the stria terminalis are major inputs to hypothalamic, pontine, and medullary structures that control preganglionic autonomic neurons of both the sympathetic and parasympathetic branches of the autonomic nervous system (Alheid et al. 1995; Loewy 1990). They are positioned to relay cortical activity into the autonomic nervous system (Cullian et al. 1993; Pitakanen et al. 2000; Westerhaus and Loewy 2001; Witter et al. 1989). Neocortex (Saper 1985, 2000) and limbic cortices, especially subiculum (Kishi et al. 2000; Witter et al. 1989; Wouterlood and Tuinhof 1992), also have

638 Sudden Death in Epilepsy: Forensic and Clinical Issues

direct projections to hypothalamus. These latter projections are likely the route by which seizure activity spreads from limbic cortical areas into autonomic areas in the urethane/ kainate model. Autonomic pathways are integrated in the hypothalamus and influence all visceral systems, including cardiovascular, genitourinary, endocrine, and digestive, as well as peripheral involuntary muscles such as pilomotor and pupillary sphincters. Projections of hypothalamic, pontine, and medullary neurons to preganglionic parasympathetic and sympathetic neurons controlling cardiac and respiratory function have been studied anatomically (Loewy et al. 1979; Moga et al. 1990, 1989; Saper 1979; Saper et al. 1976; Tucker and Saper 1985) and physiologically (Chamberlin and Saper 1992, 1998; Gebber et al. 1990, 1987; Gebber and Snyder 1970; Gebber et al. 1995; Spyer 1979; Yamashita et al. 1983). Hypothalamic cells project onto medullary structures (ventrolateral medulla and periaqueductal gray) that project directly into the intermediolateral horn of the spinal cord where the sympathetic preganglionic cell bodies are located. Activity of hypothalamic and medullary neurons correlates with sympathetic nerve activity (Barman and Gebber 1982, 1998; Barman et al. 1999, 2000, 1995). Some hypothalamic cells project directly into the spinal cord (Cechetto and Saper 1988). For the parasympathetic system, central nucleus of the amygdala (Loewy 1990) and hypothalamus (PVN and SON) (Yamashita et al. 1983) project directly into the dorsal motor nucleus of the vagus and the nucleus ambiguus, the sites of parasympathetic preganglionic cell bodies. Furthermore, the sympathetic and parasympathetic paths can influence each other. For example, efferents of the nucleus ambiguus and dorsal nucleus of the vagus (preganglionic parasympathetic neurons) project to the nucleus of the solitary tract that has a relayed projection to intermediolateral horn (sympathetic preganglionic neurons).

39.7â•…Insights into the Mechanism for Sudden Death in Epilepsy In our urethane/kainic acid rat model of seizures, the mechanism for death during autonomic overactivity was a profound bradycardia coupled with cardiac dilatation (mechanical weakness). However, some form of persistent respiratory obstruction was necessary for death because prolonged severe bradycardic episodes were associated with a “flatlining” of the EEG; that is, the seizure stops (Stewart 2008). When seizures are the only stimulus for the autonomic overactivity, impaired cardiac function suppresses seizure activity that eliminates the autonomic overactivity. Cardiac function is rapidly restored. This suggests that a seizure cannot by itself cause death via an autonomically mediated heart failure. Continued hypercapnia/hypoxia appears to be essential, ἀrst, to sustain autonomic overactivity after a seizure stops and, second, to ultimately cause heart failure from hypoxia. It is interesting that the picture we ἀnd during seizures resembles the picture found during asphyxiation. The respiratory distress that can occur during a seizure (e.g., airway mucous overproduction) is a form of asphyxiation and works pathophysiologically in the same direction as the seizures to drive the sympathetic and parasympathetic divisions simultaneously. Seizures drive both the parasympathetics and the sympathetics. Respiratory distress causes an increase in vagal activity because of the increase in arterial carbon dioxide, and it increases sympathetic activity (and blood pressure) that also favors an increase in vagal outflow via any functional baroreceptor activity. The combination is clearly rapidly lethal in our rats. Some “jumps” in parasympathetic outflow were likely large enough to make nodal block the initial or predominant ἀrst event. In the absence of

The Urethane/Kainate Seizure Model

639

sustained seizure activity or respiratory distress, a transient arrhythmia may be all that occurs. We have seen such transient arrhythmias in freely moving rats that were receiving kindling stimulation (Kunitake and Stewart, unpublished data). Death in our animal model does not resemble the vast majority of cases of sudden cardiac death. Although the literature spans some three decades, it is written almost universally that 80–85% of cases of sudden cardiac death are due to ventricular arrhythmias (reviewed by Kirchhof et al. 2006). Sudden death from fatal arrhythmia is most commonly a ventricular tachyarrhythmia. In some patients, a ventricular tachyarrhythmia is triggered by an acute myocardial ischemic event, and in other patients, a ventricular tachyarrhythmia is related to an anatomical substrate (e.g., scarring from a previous myocardial infarct) (Huikuri et al. 2001; Kiviniemi et al. 2007; Lombardi et al. 2001). Such structural cardiac changes may be part of the disease spectrum in patients with epilepsy, exposing them to increased risk of ventricular ἀbrillation. We have seen similar changes in our chronically epileptic rats, but intense seizure activity in these animals leads to the same parasympathetically predominated bradyarrhythmias. Although we have referred to the circumstances surrounding death in our animals as a parasympathetic “point-of-noÂ�return,” an alternative view of the massive parasympathetic outflow is that it is an attempt to resist a ἀbrillation state. Beyond cardiac changes that would alter the heart’s response to autonomic activity are central nervous system changes that could potentially alter the quantitative balances of activity discussed earlier. Seizures spreading through an altered autonomic nervous system could have any number of different consequences on the heart. Finally, some of these same changes in cardiac and neural activity can be produced by chronic anticonvulsant use. Studies in animal models will be critical to address the range of impact of seizures on cardiac function in chronically epileptic animals.

39.8â•… Closing Comments on This and Other Animal Models There are a number of interesting animal models that have been used to study seizures and their cardiorespiratory consequences. Unfortunately, we are still in the very early stages of understanding brain–heart interactions. Our model has the speciἀc advantage of seizures without motor convulsions, but also without chemical paralysis of the animal. This offers access to the full spectrum of cardiorespiratory consequences in an acute preparation. The general advantages of animal models, efficiency plus measurement and manipulation options, make the acute preparation a flexible tool that should be considered for a variety of epilepsy studies. Each preparation has certain advantages and disadvantages. The choice of model should start with its advantages for studies of particular questions, but most importantly, the flexibility of animal models cannot be overstated. These can literally be “tailored” to address questions from the clinical world that the animal investigators may have never considered.

References Alheid, G., J. De Olmos, and C. Beltramino. 1995. Amygdala and extended amygdala. In The Rat Nervous System, 2nd ed., ed. G. Paxinos, 495–578. San Diego, CA: Academic Press. Allen, G. V., C. B. Saper, K. M. Hurley, and D. F. Cechetto. 1991. Organization of visceral and limbic connections in the insular cortex of the rat. J Comp Neurol 311: 1–16.

640 Sudden Death in Epilepsy: Forensic and Clinical Issues Ardell, J. L. 2004. Intrathoracic neuronal regulation of cardiac function. In Basic and Clinical Neurocardiology, ed. J. A. Armour and J. L. Ardell, 118–152. New York, NY: Oxford University Press. Barman, S. M., and G. L. Gebber. 1982. Hypothalamic neurons with activity patterns related to sympathetic nerve discharge. Am J Physiol 242: R34–R43. Barman, S. M., and G. L. Gebber. 1998. Classiἀcation of caudal ventrolateral pontine neurons with sympathetic nerve-related activity. J Neurophysiol 80: 2433–2445. Barman, S. M., G. L. Gebber, and H. Kitchens. 1999. Rostral dorsolateral pontine neurons with sympathetic nerve-related activity. Am J Physiol 276: H401–H412. Barman, S. M., G. L. Gebber, and H. S. Orer. 2000. Medullary lateral tegmental ἀeld: An important source of basal sympathetic nerve discharge in the cat. Am J Physiol Regul Integr Comp Physiol 278: R995–R1004. Barman, S. M., H. S. Orer, and G. L. Gebber. 1995. Axonal projections of caudal ventrolateral medullary and medullary raphe neurons with activity correlated to the 10-Hz rhythm in sympathetic nerve discharge. J Neurophysiol 74: 2295–2308. Baumgartner, C., S. Lurger, and F. Leutmezer. 2001. Autonomic symptoms during epileptic seizures. Epileptic Disord 3: 103–116. Brankack, J., M. Stewart, and S. E. Fox. 1993. Current source density analysis of the hippocampal theta rhythm: Associated sustained potentials and candidate synaptic generators. Brain Res 615: 310–327. Britton, J. W., G. R. Ghearing, E. E. Benarroch, and G. D. Cascino. 2006. The ictal bradycardia syndrome: Localization and lateralization. Epilepsia 47: 737–744. Cechetto, D. F., and C. B. Saper. 1988. Neurochemical organization of the hypothalamic projection to the spinal cord in the rat. J Comp Neurol 272: 579–604. Chamberlin, N. L., and C. B. Saper. 1992. Topographic organization of cardiovascular responses to electrical and glutamate microstimulation of the parabrachial nucleus in the rat. J Comp Neurol 326: 245–262. Chamberlin, N. L., and C. B. Saper. 1998. A brainstem network mediating apneic reflexes in the rat. J Neurosci 18: 6048–6056. Cheung, R. T., and V. Hachinski. 2000. The insula and cerebrogenic sudden death. Arch Neurol 57: 1685–1688. Cullian, W. E., J. P. Herman, and S. J. Watson. 1993. Ventral subicular interaction with the hypothalamic paraventricular nucleus: Evidence for a relay in the bed nucleus of the stria terminalis. J Comp Neurol 332: 1–20. Devinsky, O. 2004. Effects of seizures on autonomic and cardiovascular function. Epilepsy Curr 4: 43–46. Epstein, M. A., M. R. Sperling, and M. J. O’Connor. 1992. Cardiac rhythm during temporal lobe seizures. Neurology 42: 50–53. Feenstra, B. W., and J. Holsheimer. 1979. Dipole-like neuronal sources of theta rhythm in dorsal hippocampus, dentate gyrus and cingulate cortex of the urethane-anesthetized rat. Electroencephalogr Clin Neurophysiol 47: 532–538. Flecknell, P. A. 1996. Laboratory Animal Anaesthesia: A Practical Introduction for Research Workers and Technicians. London: Academic Press. Fukuda, Y., A. Sato, A. Suzuki, and A. Trzebski. 1989. Autonomic nerve and cardiovascular responses to changing blood oxygen and carbon dioxide levels in the rat. J Auton Nerv Syst 28: 61–74. Galin, D. 1974. Implications for psychiatry of left and right cerebral specialization. A neurophysiological context for unconscious processes. Arch Gen Psychiatry 31: 572–583. Gebber, G. L., S. M. Barman, and B. Kocsis. 1990. Coherence of medullary unit activity and sympathetic nerve discharge. Am J Physiol 259: R561–R571. Gebber, G. L., S. M. Barman, and S. F. Morrison. 1987. Electrophysiological evidence for the modular organization of the reticular formation: Sympathetic controlling circuits. Brain Res 410: 106–110. Gebber, G. L., and D. W. Snyder. 1970. Hypothalamic control of baroreceptor reflexes. Am J Physiol 218: 124–131.

The Urethane/Kainate Seizure Model

641

Gebber, G. L., S. Zhong, and S. M. Barman. 1995. Synchronization of cardiac-related discharges of sympathetic nerves with inputs from widely separated spinal segments. Am J Physiol 268: R1472–R1483. Gilbert, M., R. J. Racine, and G. K. Smith. 1985. Epileptiform burst responses in ventral vs dorsal hippocampal slices. Brain Res 361: 389–391. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1990. Kindled seizures elevate blood pressure and induce cardiac arrhythmias. Epilepsia 31: 489–495. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1999. Kindled seizures activate both branches of the autonomic nervous system. Epilepsy Res 34: 169–176. Goodman, J. H., M. Stewart, and F. W. Drislane. 2008. Autonomic disturbances. In Epilepsy: A Comprehensive Textbook, 2nd ed., ed. J. Engel and T. A. Pedley, 1999–2005. Philadelphia, PA: Lippincott Williams and Wilkins. Hachinski, V. C., S. M. Oppenheimer, J. X. Wilson, C. Guiraudon, and D. F. Chetto. 1992. Asymmetry of sympathetic consequences of experimental stroke. Arch Neurol 49: 697–702. Heltovics, G., B. Boda, and M. Szente. 1995. Anticonvulsive effect of urethane on aminopyridineinduced epileptiform activity. Neuroreport 6: 577–580. Hilz, M. J., M. Dutsch, K. Perrine, P. K. Nelson, U. Rauhut, and O. Devinsky. 2001. Hemispheric influence on autonomic modulation and baroreflex sensitivity. Ann Neurol 49: 575–584. Hotta, H., K. Koizumi, and M. Stewart. 2009a. Cardiac sympathetic nerve activity during kainic acid–induced limbic cortical seizures in rats. Epilepsia 50: 923–927. Hotta., H., J. Lazar, R. Orman, et al. 2009b. Vagus nerve stimulation-induced bradyarrhythmias in rats. Auton Neurosci 151: 98–105. Huikuri, H. V., A. Castellanos, and R. J. Myerburg. 2001. Sudden death due to cardiac arrhythmias. N Engl J Med 345: 1473–1482. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37: 531–537. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42: 588–594. Kirchhof, P., G. Breithardt, and L. Eckardt. 2006. Primary prevention of sudden cardiac death. Heart 92: 1873–1878. Kishi, T., T. Tsumori, K. Ono, S. Yokota, H. Ishino, and Y. Yasui. 2000. Topographical organization of projections from the subiculum to the hypothalamus in the rat. J Comp Neurol 419: 205–222. Kiviniemi, A. M., M. P. Tulppo, D. Wichterle, et al. 2007. Novel spectral indexes of heart rate variability as predictors of sudden and non-sudden cardiac death after an acute myocardial infarction. Ann Med 39: 54–62. Koizumi, K., and M. Kollai. 1981. Control of reciprocal and non-reciprocal action of vagal and sympathetic efferents: Study of centrally induced reactions. J Auton Nerv Syst 3: 483–501. Koizumi, K., M. Kollai, and N. Turei. 1986. The physiology of cardiac innervation: Relationships between cardiac vagal and sympathetic nerve activities. In Neurovegetative Control Systems: Basic Functions, Integration, and Disorders: Proceedings of the XIXth International Congress of Neurovegetative Research, Berlin (West), FRG, March 1984, ed. H. P. Koepchen, C. M. Brooks, K. Koizumi, and conf. author, 161–171. Amsterdam, NY: International Congress of Neurovegetative Research. Koizumi, K., N. Terui, M. Kollai, and C. M. Brookis. 1982. Functional signiἀcance of coactivation of vagal and sympathetic cardiac nerves. Proc Natl Acad Sci U S A 79: 2116–2120. Kollai, M., and K. Koizumi. 1979. Reciprocal and non-reciprocal action of the vagal and sympathetic nerves innervating the heart. J Auton Nerv Syst 1: 33–52. Lane, R. D., and J. R. Jennings. 1998. Hemispheric asymmetry, autonomic asymmetry, and the problem of sudden cardiac death. In Brain Asymmetry, 2nd printing ed., ed. R. J. Davidson and K. Hugdahl, 271–304. Cambridge, MA: MIT Press. Langan, Y. 2000. Sudden unexpected death in epilepsy (SUDEP): Risk factors and case control studies. Seizure 9: 179–183.

642 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., P. L. Schraeder, and J. G. Boggs. 1998. Sudden unexplained death and autonomic dysfunction. In Epilepsy: A Comprehensive Textbook, vol. 2, chap. 183, ed. J. Engel and T. A. Pedley, 1943–1956. Philadelphia: Lippincott-Raven. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital on autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33: 837–844. Lathers, C. M., P. L. Schraeder, and F. W. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Liedholm, L. J., and O. Gudjonsson. 1992. Cardiac arrest due to partial epileptic seizures. Neurology 42: 824–829. Locatelli, E. R., J. P. Varghese, A. Shuabib, and S. J. Potolicchio. 1999. Cardiac asystole and bradycardia as a manifestation of left temporal lobe complex partial seizure. Ann Intern Med 130: 581–583. Loewy, A. D. 1990. Central autonomic pathways. In Central Regulation of Autonomic Functions, ed. A. D. Loewy and K. M. Spyer, 88–103. New York, NY: Oxford University Press. Loewy, A. D., S. McKellar, and C. B. Saper. 1979. Direct projections from the A5 catecholamine cell group to the intermediolateral cell column. Brain Res 174: 309–314. Lombardi, F., T. H. Makikallio, R. J. Myerburg, and H. V. Huikui. 2001. Sudden cardiac death: Role of heart rate variability to identify patients at risk. Cardiovasc Res 50: 210–217. Mameli, O., M. A. Caria, F. Melis et al. 2001. Autonomic nervous system activity and life threatening arrhythmias in experimental epilepsy. Seizure 10: 269–278. Mameli, O., M. A. Caria, A. Pintus, G. Padua and S. Mameli. 2006. Sudden death in epilepsy: An experimental animal model. Seizure 15: 275–287. Mameli, O., F. Melis, D. Giraudi et al. 1993. The brainstem cardioarrhythmogenic triggers and their possible role in sudden epileptic death. Epilepsy Res 15: 171–178. Moga, M. M., H. Herbert, K. M. Hurley, Y. Yasumi, T. S. Gray, and C. B. Saper. 1990. Organization of cortical, basal forebrain, and hypothalamic afferents to the parabrachial nucleus in the rat. J Comp Neurol 295: 624–661. Moga, M. M., C. Saper, and T. S. Gray. 1989. Bed nucleus of the stria terminalis: Cytoarchitecture, immunohistochemistry, and projection to the parabrachial nucleus in the rat. J Comp Neurol 283: 315–332. Moroni, F., R. Corradetti, F. Casementi, G. Moneti, and G. Pepue. 1981. The release of endogenous GABA and glutamate from the cerebral cortex in the rat. Naunyn Schmiedeberg’s Arch Pharmacol 316: 235–239. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41: 542–548. Oppenheimer, S. 2001. Forebrain lateralization and the cardiovascular correlates of epilepsy. Brain 124: 2345–2346. Oppenheimer, S. M., A. Gelb, J. P. Girvin, and V. C. Hachinski. 1992. Cardiovascular effects of human insular cortex stimulation. Neurology 42: 1727–1732. Oppenheimer, S. M., J. X. Wilson, C. Guiraudon, and D. F. Chechetto. 1991. Insular cortex stimulation produces lethal cardiac arrhythmias: A mechanism of sudden death? Brain Res 550: 115–121. Pitakanen, A., M. Pikkarainen, N. Nurminen, and A. Ylinen. 2000. Reciprocal connections between the amygdala and the hippocampal formation, perirhinal cortex, and postrhinal cortex in rat. A review. Ann N Y Acad Sci 911: 369–391. Randall, D. C., D. R. Brown, A. S. McGuirt, G. W. Thompson, J. A. Armour, and J. L. Ardell. 2003. Interactions within the intrinsic cardiac nervous system contribute to chronotropic regulation. Am J Physiol Regul Integr Comp Physiol 285: R1066–R1075.

The Urethane/Kainate Seizure Model

643

Saito, T., K. Sakamoto, K. Koizumi, and M. Stewart. 2006. Repeatable focal seizure suppression: A rat preparation to study consequences of seizure activity based on urethane anesthesia and reversible carotid artery occlusion. J Neurosci Methods 155: 241–250. Sakamoto, K., T. Saito, R. Orman, et al. 2008. Autonomic consequences of kainic acid–induced limbic cortical seizures in rats: Peripheral autonomic nerve activity, acute cardiovascular changes, and death. Epilepsia 49: 982–996. Saper, C. B. 1979. Anatomical substrates for the hypothalamic control of the autonomic nerÂ� vous system. In Integrative Functions of the Autonomic Nervous System: An Analysis of the Interrelationships and Interactions of the Sympathetic and Parasympathetic Divisions of the Autonomic System in the Control of Body Function, ed. C. M. Brooks, K. Koizumi, and A. Sato, 333–341. Tokyo: University of Tokyo Press. Saper, C. B. 1982a. Convergence of autonomic and limbic connections in the insular cortex of the rat. J Comp Neurol 210 (2): 163–173. Saper, C. B. 1982b. Reciprocal parabrachial–cortical connections in the rat. Brain Res 242 (1): 33–40. Saper, C. B. 1985. Organization of cerebral cortical afferent systems in the rat: II. Hypothalamocortical projections. J Comp Neurol 237: 21–46. Saper, C. B. 2000. Hypothalamic connections with the cerebral cortex. In Cognition, Emotion and Autonomic Responses: The Integrative Role of the Prefrontal Cortex and Limbic Structures, ed. H.€B. M. Uylings, G. G. Van Eden, J. P. C. De Bruin, M. G. P. Feenstra, and C. M. A. Pennartz, 39–48. Amsterdam: Elsevier. Saper, C. B., A. D. Loewy, L. W. Swanson, and W. M. Cowan. 1976. Direct hypothalamo-autonomic connections. Brain Res 117: 305–312. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3: 55–62. Smith-Demps, C., and A. Jagoda. 1998. A case of seizure-related bradycardia and asystole. Am J Emerg Med 16: 582–584. Spyer, K. M. 1979. Baroreceptor control of vagal preganglionic activity. In Integrative Functions of the Autonomic Nervous System: An Analysis of the Interrelationships and Interactions of the Sympathetic and Parasympathetic Divisions of the Autonomic System in the Control of Body Function, ed. C. M. Brooks, K. Koizumi, and A. Sato, 287–292. Tokyo: University of Tokyo Press. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72: 340–345. Stewart, M. 2008. Is an abrupt “cerebral electrical shutdown” during a seizure the mechanism of SUDEP? J Neurol Neurosurg Psychiatry. E letter. http://jnnp.bmj.com/cgi/eletters/78/12/1395#╉ 3299, accessed on 12 Feb 2008. Stewart, M., and S. E. Fox. 1990. Do septal neurons pace the hippocampal theta rhythm? Trends Neurosci 13: 163–168. Tinuper, P., F. Bisulli, A. Cerullo, et al. 2001. Ictal bradycardia in partial epileptic seizures: Autonomic investigation in three cases and literature review. Brain 124: 2361–2371. Tucker, D. C., and C. B. Saper. 1985. Speciἀcity of spinal projections from hypothalamic and brainstem areas which innervate sympathetic preganglionic neurons. Brain Res 360: 159–164. Venit, E. L., B. D. Shepard, and T. N. Seyfried. 2004. Oxygenation prevents sudden death in seizureprone mice. Epilepsia 45: 993–996. Westerhaus, M. J., and A. D. Loewy. 2001. Central representation of the sympathetic nervous system in the cerebral cortex. Brain Res 903: 117–127. Wilder-Smith, E. 1992. Complete atrio-ventricular conduction block during complex partial seizure. J Neurol Neurosurg Psychiatry 55: 734–736. Willott, J. F., and K. R. Henry. 1976. Roles of anoxia and noise-induced hearing loss in the postictal refractory period for audiogenic seizures in mice. J Comp Physiol Psychol 90: 373–381.

644 Sudden Death in Epilepsy: Forensic and Clinical Issues Witter, M. P., and D. G. Amaral. 2004. Hippocampal formation. In The Rat Nervous System, 3rd ed., ed. G. Paxinos, 635–704. San Diego, CA: Elsevier Academic Press. Witter, M. P., H. L. Groenewegen, F. H. Lopes Da Silva, and A. H. Lohman. 1989. Functional organization of the extrinsic and intrinsic circuitry of the parahippocampal region. Prog Neurobiol 33: 161–253. Witter, M. P., P. Room, H. L. Groenewegen, and A. H. Lohman. 1988. Reciprocal connections of the insular and piriform claustrum with limbic cortex: An anatomical study in the cat. Neuroscience 24: 519–539. Wouterlood, F. G., and R. Tuinhof. 1992. Subicular efferents to histaminergic neurons in the posterior hypothalamic region of the rat studied with PHA-L tracing combined with histidine decarboxylase immunocytochemistry. J Hirnforsch 33: 451–465. Yamashita, H., H. Kannah, K. Inenega, and K. Koizumi. 1983. Role of neurones in the supraoptic and paraventricular nuclei in cardiovascular control. Prog Brain Res 60: 459–468. Yasui, Y., C. D. Breder, C. B. Saper, and D. F. Cechetto. 1991. Autonomic responses and efferent pathways from the insular cortex in the rat. J Comp Neurol 303: 355–374. Zamrini, E. Y., K. J. Meador, D. W. Loring, et al. 1990. Unilateral cerebral inactivation produces differential left/right heart rate responses. Neurology 40: 1408–1411.

Acute Cardiovascular Response during Kindled Seizures Jeffrey H. Goodman Richard W. Homan IsAac L. Crawford

40

Contents 40.1 Introduction 40.2 Methods 40.3 Results 40.3.1 Effect on Blood Pressure and Heart Rate 40.3.2 Effect of Seizure on the Electrocardiogram 40.3.3 Seizure Dependency 40.3.4 Autonomic Changes in Non-Amygdaloid-Initiated Seizures 40.3.5 Cardiovascular Changes during Kindling Acquisition 40.4 Discussion 40.4.1 Role of the Amygdala 40.4.2 The Pressor Response 40.4.3 The Arrhythmogenic Response 40.4.4 The Cerebral Ischemic Response 40.4.5 Clinical Relevance 40.4.6 Kindling as a Model of Seizure-Induced Arrhythmias and Sudden Unexplained Death 40.5 Summary References

645 647 648 648 648 649 651 651 652 652 653 654 654 654 655 655 655

40.1â•…Introduction The high incidence of sudden unexplained death in the epileptic population not only poses a serious problem for the epileptic patient but is also a challenge to the clinician and an unsolved mystery for the basic scientist. A reasonable hypothesis is that if seizures disrupt autonomic regulation of the cardiovascular system, then the consequent generation of cardiac arrhythmias may underlie sudden unexplained death. Clearly, spontaneous seizures in man (Gilchrist 1985; Marshall et al. 1983; Pritchett et al. 1980) and experimentally induced seizures in animals (Doba et al. 1975; Lathers and Schraeder 1982) can be accompanied by cardiac arrhythmias. There is also a case report of sudden cardiac death immediately after a complex partial seizure (Dasheiff and Dickinson 1986). However, no direct evidence has been provided that deἀnitely links sudden unexplained death to seizure occurrence. A recent report by Keilson et al. (1987) revealed that the incidence of serious cardiac arrhythmias is no greater in epileptic patients than in the general population. Because there is 645

646 Sudden Death in Epilepsy: Forensic and Clinical Issues

no clinical marker that identiἀes those epileptic patients most at risk for sudden death, a greater research emphasis placed on basic experimental models of epilepsy may reveal new directions for clinical studies. Cardiovascular changes occur during several types of experimentally induced seizures: electroconvulsive treatment (Petito et al. 1977; Plum et al. 1968; Wasterlain 1974; Westergaard et al. 1978), pentylenetetrazol (Doba et al. 1975; Lathers and Schraeder 1982; Plum et al. 1968), bicuculline (Meldrum and Horton 1973), and penicillin (Lathers and Schraeder, this book, Chapter 10; Mameli et al. 1988). However, these models have inherent limitations because of the presence of anesthetics, paralytic agents with artiἀcial ventilation, or widespread exposure of the CNS to chemical convulsants or electrical current. In the kindling seizure model, repeated spaced presentations of an initially subconvulsive electrical stimulus eventually led to a permanent change in brain function (Goddard et al. 1969; Homan and Goodman 1988). This change is characterized by the occurrence of a generalized motor seizure each time the stimulus is presented. During the kindling process, there is a progressive increase in electrical afterdischarge activity (Figure 40.1), with spread of the afterdischarge to remote brain regions. Eventually, clinically apparent seizures develop, progressing through ἀve distinct behavioral stages (Racine 1972; Figure 40.2). Early stages in the kindling process (stages 1 and 2) are equivalent to partial seizures. Spread of the seizure ἀrst occurs at stage 3, whereas a stage 5 seizure is equivalent to a generalized tonic–clonic convulsion. This is a key feature of kindling because partial and generalized seizures can be examined in the same animal. A stage 5 generalized seizure usually occurs after 9–10 days of amygdaloid stimulation. To elicit spontaneous interictal spiking or spontaneous seizures requires 100–300 days of daily stimulation (Pinel and Rovner 1978). For this reason, these types of epileptiform activity are difficult to investigate in the kindling model. However, in contrast to chemical models, no drug is injected, and therefore one does not have to distinguish the direct from the nondirect effects of nonexperimental drugs. Thus far, few major morphological alterations have been found in the kindled focus, in contrast to injection of heavy metals Day 1

Day 2

1 mV 1s

Day 5

Gate artifact 400 μA, 60 Hz, 1 s BLA stimulation

Figure 40.1╇ Growth of seizure afterdischarge during the kindling process. The lower elec-

trograph for day 5 is a continuation of that shown before and after the fifth stimulus. (From Goodman, J. H., et al., Epilepsia, 31, 489–495, 1990. With permission.)

Acute Cardiovascular Response during Kindled Seizures

IV

647

V

III

II I

Figure 40.2╇ Behavioral stages of kindled seizures. The five stages are those identified by

Racine (1972).

or thermal injury to produce convulsions; both of the latter procedures result in neuronal loss and glial proliferation. The kindling model is well suited for an examination of the relationship between seizures and the cardiovascular system without the presence of drugs. Kindled seizures can be initiated in a discrete region of the brain of unanesthetized, unrestrained animals (Racine 1972, 1978). We have used this model to explore the acute cardiovascular changes that occur during amygdaloid-kindled seizures. We will review the results of our studies and also discuss the signiἀcance of these changes with respect to the potential value of kindling as a model of seizure-induced cardiac arrhythmias. Speculation on the relationship of our ἀndings to the clinical phenomenon of sudden unexplained death will also be presented.

40.2â•… Methods To examine the effect of kindled seizures on the cardiovascular system of the rat, male Sprague–Dawley rats (275–325 g) were anesthetized with ketamine (90 mg/kg) and xylazine (15 mg/kg) by intramuscular injection. Teflon-coated stainless steel electrodes (100€µm) were stereotaxically implanted in the basolateral amygdalae or in the olfactory bulbs according to previously reported methods used in our laboratory (Campbell and Crawford 1980; Crawford 1986; Walker et al. 1981). Each animal was allowed to recover a minimum of 1 week before the kindling process was initiated. During the kindling process, each animal was stimulated once daily for 1 s (400 µA, 60 Hz, 1-ms square wave, bipolar pulse) until three consecutive stage 5 generalized seizures occurred. A chronic femoral artery catheter ἀlled with heparinized saline was then surgically implanted to allow for the direct measurement of blood pressure in unanesthetized rats during the kindled seizure. After recovery from catheterization (a minimum of 24 h), ECG electrodes were placed across the chest; a kindled seizure was initiated; and changes in blood pressure, heart rate,

648 Sudden Death in Epilepsy: Forensic and Clinical Issues

and ECG were measured. Controls, with electrodes and catheter, were treated in a similar manner without undergoing the kindling process. In a separate group of rats, blood pressure and heart rate were measured during amygdaloid kindling acquisition. Amygdaloid electrodes and the arterial catheter were implanted as described above. These animals did not have ECG electrodes, so changes in the ECG were not measured. Because the arterial catheter remained patent for only 5 days, these animals were stimulated twice a day during the kindling process. The other stimulus parameters were the same as previously described. After each stimulus, changes in behavioral seizure score, afterdischarge duration, blood pressure, and heart rate were measured.

40.3â•…Results 40.3.1â•… Effect on Blood Pressure and Heart Rate Figure 40.3 shows the typical cardiovascular response during a stage 5 generalized amygdaloid-kindled seizure. The response was characterized by an abrupt increase in systolic and diastolic pressures, which lasted 20–30 s after initiation of the seizure. Superimposed on this change in pressure was a profound bradycardia occasionally accompanied by premature ventricular contractions. A summary of the effect of the seizure on mean arterial pressure (MAP) and heart rate is shown in Figure 40.4. MAP was signiἀcantly elevated for 20 s after initiation of the kindled seizure. During the ἀrst 10 s of the seizure, MAP approximated 150 mm Hg, whereas the mean heart rate decreased 50%. The increase in MAP was concomitant with increases in systolic, diastolic, and pulse pressures (Table 40.1). Changes in pressure and heart rate were greatest in the ἀrst 10–15 s of the seizure; however, both parameters returned to baseline before the end of the seizure afterdischarge. 40.3.2â•… Effect of Seizure on the Electrocardiogram The effect of the kindled seizure on the heart was determined by making simultaneous recordings of blood pressure and the ECG from animals during kindled seizures. During the seizure the bradycardia present in the pressure recording was consistently present in€the ECG of each animal. In most of the animals, additional irregularities were detected in the ECG during the bradycardia. Occasionally, premature ventricular contractions were also observed. These changes, as illustrated in Figure 40.5, included nonconducted P waves€and increases in the P–R interval, which are indicative of conduction heart block (Berne and Levy 1981). 200 mm Hg 150 100 50 0

Kindling stimulus

5s

Figure 40.3╇ Typical blood pressure response during an amygdaloid-kindled seizure. Arrow indicates when stimulus was delivered.

Acute Cardiovascular Response during Kindled Seizures

Mean arterial pressure (mean ± SEM) (mm Hg)

125

*

*

* *

100 75

0

5

N=9

Map

400

* *

*

300 200

Heart rate

10 15 20 Time from seizure onset (s)

30

Heart rate (mean ± SEM) (beats/min)

*

150

649

100

Figure 40.4╇ Graphic comparison of changes in mean arterial pressure (MAP) and heart rate during the first 30 s after initiation of stage 5 amygdaloid-kindled seizures (*p < 0.01). SEM, standard error of the mean. (From Goodman, J. H., et al., Epilepsia, 31, 489–495, 1990. With permission.)

40.3.3â•… Seizure Dependency The possibility existed that the cardiovascular response observed during the kindled seizure was simply the result of electrical stimulation of the amygdala, a known cardiovascular control center. To test this hypothesis, rats not previously kindled were stimulated with a single kindling train. In each instance the stimulus elicited a small seizure afterdischarge; however, no change in blood pressure or heart rate was detected. This suggests the cardiovascular changes that occurred in kindled animals in response to the same stimulus were seizure dependent. Electrically kindled seizures can be followed by postictal spiking. When this occurred in several animals, postictal spikes were accompanied by increases in blood pressure (Figure 40.6). These postictal spikes often develop into spontaneous seizures 2 or 3 min after the stimulus-induced seizure. These seizures, electrographically and behaviorally, mimicked stimulus-induced seizures. The cardiovascular responses during the secondary spontaneous seizures (Figure 40.7b) were qualitatively the same as those recorded during amygdaloid-kindled seizures (Figure 40.7a). Table 40.1â•… Effect of Amygdaloid-Kindled Seizures on Blood Pressure a Mean ± SEM (mm€Hg) Systolic pressure (mean ± SEM) Diastolic pressure (mean ± SEM) Pulse pressure (mean ± SEM)

Time from Seizure Initiation (s) n

Baseline

5

10

15

20

30

60

9

131.1 ± 1.4 85.0 ± 1.1 46.1 ± 1.0

192.8** ± 2.6 126.1** + 2.2 66.7** ± 2.3

197.8** ± 3.3 121.1** ± 1.5 76.7** ± 2.2

171.1** ± 1.5 103.9* ± 1.6 67.2** ± 1.6

163.3** ± 1.9 108.3** ± 2.2 55.0 ± 1.3

142.2 ± 1.4 90.6 ± 1.9 51.7 ± 1.0

131.4 ± 2.1 85.9 ± 1.5 45.6 ± 1.6

9 9

Source: Goodman, J. H., et al., Epilepsia, 31, 489–495, 1990. With permission. a One repeated-measure analysis of variance followed by Neuman–Keuls test, compared to baseline value. *p < 0.05; **p < 0.01.

650 Sudden Death in Epilepsy: Forensic and Clinical Issues

2 mV

ECG

Blood pressure 200 mm Hg 150 100 1s

50 0

EEG

1 mV

Stimulus

Figure 40.5╇ Electrograph of the effect of stage 5 amygdaloid-kindled seizure on the electro-

cardiogram (ECG) and blood pressure. Notice the isolated P wave (*) and increased P–R interval (small arrow) that correspond with irregular changes in the blood pressure recording and the afterdischarge in the EEG. The up and down arrows under the EEG indicate the beginning and end of the kindling stimulus. This is followed by a period of amplifier block. (From Goodman, J. H., et al., Epilepsia, 31, 489–495, 1990. With permission.) ECG

(a)

5 mV 180

BP

100 0

EEG 1 mV

ECG

(b)

5 mV 180

BP

100 0 EEG

1 mV 5s

Figure 40.6╇ Electrograph of blood pressure. EEG, electroencephalogram; ECG, electrocardiogram in a kindled rat before (a) and 3 min after a kindling stimulus (b). Note the transient rise in pressure after each interictal, epileptiform spike-wave complex.

Acute Cardiovascular Response during Kindled Seizures

(a)

Blood pressure 200 mm Hg 150 100 50 0

(b)

651

Amygdala Kindling stimulus

200 mm Hg 150 100 50 0

Spontaneous seizure

Amygdala

(c) 200 mm Hg 150 100 50 0

Olfactory bulb Kindling stimulus

5s

Figure 40.7╇ Effect of kindled seizures on blood pressure. (a) Change in blood pressure during

an amygdaloid-kindled stage 5 generalized seizure. (b) Change in blood pressure during a spontaneous generalized seizure that occurred 3 min after a stimulus-induced amygdaloid-kindled seizure. (c) Blood pressure response during a stage 5 generalized seizure initiated in the olfactory bulb.

40.3.4â•… Autonomic Changes in Non-Amygdaloid-Initiated Seizures To determine whether kindled seizure-induced autonomic changes were restricted to seizure initiation in the amygdala, we kindled animals in the olfactory bulb. Figure 40.7c shows the cardiovascular changes that occurred during an olfactory bulb–kindled seizure. These changes were qualitatively the same as those recorded during amygdaloid-kindled seizures. 40.3.5â•… Cardiovascular Changes during Kindling Acquisition The initial stimulus in the kindling process did not elicit a pressor response or a change in heart rate. However, as kindling was continued, several changes were observed. Figure 40.8 shows the progressive changes in the pressure and arrhythmic responses during the kindling process. The ἀrst cardiovascular change was a small rise in pressure with no change in heart rate. With subsequent stimulations the pressor response continued to increase. In contrast, the arrhythmic response developed at a different rate. The arrhythmia did not occur until after the sixth stimulation. Continued stimulation resulted in an increase in the duration and the severity of the arrhythmia while the magnitude of

652 Sudden Death in Epilepsy: Forensic and Clinical Issues Blood pressure 180 mm Hg Stimulus * 3

100 0 180

Stimulus * 5

100 0 180

Stimulus * 6

100 0 180

Stimulus * 7

20 s

100

5s

0 Kindling stimulus

Figure 40.8╇ Changes in blood pressure during kindling acquisition. Notice the delayed appearance of the arrhythmic response after the sixth and seventh stimulations. Arrows indicate when stimulus was delivered. (From Goodman, J. H., et al., Epilepsia, 31, 489–495, 1990. With permission.)

the pressor response remained the same. The cardiovascular response present after the seventh stimulation occurred during a stage 2 seizure, before seizure generalization had developed.

40.4â•…Discussion The results from the foregoing experiments identify an experimental seizure model that simultaneously evaluates altered cardiovascular function. The cardiovascular response during amygdaloid-kindled seizures was characterized by a large increase in blood pressure accompanied by a profound bradycardia. These changes were present during partial (Figure 40.8) and generalized seizures. It is important to recognize that these results were obtained from unanesthetized animals, thereby eliminating one of the confounding variables associated with other experimental seizure models. 40.4.1â•…Role of the Amygdala The amygdala is an important cardiovascular control center with numerous connections to brain stem (Hopkins and Holstege 1978; Schwaber et al. 1980; Takeuchi et al. 1982) and

Acute Cardiovascular Response during Kindled Seizures

653

limbic (De Olmos 1972) structures that regulate the autonomic nervous system. The connections to the hypothalamus through the stria terminals (De Olmos and Ingram 1972) and ventral amygdalofugal pathways (Millhouse 1969) are particularly relevant because these pathways mediate amygdala-induced changes in cardiovascular function (Faiers et al. 1975; Galosy et al. 1982; Hilton and Zbrozyna 1963). Stimulation of the amygdala in the rat (Faiers et al. 1975; Iwata et al. 1987; Le Doux et al. 1982; Mogenson and Calaresu 1973), cat (Heinemann et al. 1973; Stock et al. 1978), and monkey (Reis and Oliphant 1964) has been shown to affect blood pressure and heart rate. Speciἀc changes with basal amygdaloid stimulation in unanesthetized animals include a decrease in blood pressure with no change or an increase in heart rate (Heinemann et al. 1973; Stock et al. 1978) or a bradycardia followed by an increase in blood pressure (Reis and Oliphant 1964). With these observations in mind, a predictable response to an electrical stimulus in the amygdala of kindled rats would be an alteration of cardiovascular function. However, a single stimulus train in the amygdala of implanted control animals had no effect on blood pressure or heart rate. The difference between this result and those of past studies may be due to the use of different stimulus parameters, speciἀcally stimulus frequency and duration. The observation that a single stimulus train (1 s, 400 µA, 60 Hz) did not elicit cardiovascular changes in control animals suggests that the cardiovascular changes in kindled animals are seizure dependent. This is supported by the observations that postictal spikes are accompanied by an increase in pressure and that secondary spontaneous seizures, in the absence of a kindling stimulus, result in the same abnormal cardiovascular changes that occur during stimulus-induced seizures. Similar changes observed during olfactory bulb–kindled seizures suggest that this response is not limited to seizures initiated in the amygdala. 40.4.2â•… The Pressor Response The hypertension observed during generalized kindled seizures was similar to pressor changes reported to occur in several other experimental seizure models: ECT (Petito et al. 1977; Plum et al. 1968; Wasterlain 1974; Westergaard et al. 1978), PTZ (Doba et al. 1975; Lathers and Schraeder 1982; Plum et al. 1968), and bicuculline (Johansson and Nilsson 1977; Meldrum and Horton 1973). The pressor response during the kindled seizure was characterized by a rapid elevation in MAP to approximately 150 mm Hg. The normal response to an increase in MAP is an increase in vascular resistance, minimizing the increase in cerebral blood flow. However, increases in MAP of this magnitude were observed by Plum et al. (1968) during ECT- and PTZ-induced seizures to interrupt cerebral autoregulation, thereby causing cerebral blood flow to become pressure dependent. Such major MAP elevations can also compromise the blood–brain barrier (Johansson and Nilsson 1977; Suzuki et al. 1984; Westergaard et al. 1978). The sudden rise in blood pressure during the kindled seizure may be due to activation of sympathetic pathways. Plum et al. (1968) demonstrated that severing the spinal cord in dogs eliminated the increase in blood pressure and cerebral blood flow that occurred during seizures initiated with ECT or PTZ. A massive autonomic discharge has been reported to accompany generalized seizures (Meldrum et al. 1979), and in other seizure models, the seizure-induced rise in pressure can be prevented with ganglionic blockers (Meldrum 1976), alpha-adrenergic blockers, or chemical sympathectomy (Doba et al. 1975).

654 Sudden Death in Epilepsy: Forensic and Clinical Issues

40.4.3â•… The Arrhythmogenic Response It is not surprising that kindled seizures are accompanied by cardiac arrhythmias. There are several reports of arrhythmias associated with other types of experimentally induced seizures in animals (Delgado et al. 1960; Doba et al. 1975; Lathers and Schraeder 1982; Wasterlain 1974). These arrhythmias were probably due to abnormal autonomic activity induced by interictal or ictal seizure activity. The bradycardic response that occurred during the kindled seizure was probably mediated by the parasympathetic system. The irregularities observed in the ECG during the seizure indicate the bradycardia was due to a conduction heart block (Berne and Levy 1981). The parasympathetic system could be activated by two different mechanisms. The most obvious would be the activation of baroreceptor reflexes by the large pressor response (Frysinger et al. 1984). However, the parasympathetic system could also be activated centrally by the seizure. Lathers et al. (1987) demonstrated that cardiac arrhythmias associated with PTZ-induced interictal or ictal seizure activity result from autonomic dysfunction, that is, an imbalance in neural activity within and between sympathetic and parasympathetic nerves to the heart. Seizure-induced coactivation of the sympathetic and parasympathetic systems is another possibility (Doba et al. 1975). 40.4.4â•… The Cerebral Ischemic Response The hypertension and bradycardia that occurred during the kindled seizure resemble cardiovascular changes that occur in response to an increase in intracranial pressure, the Cushing response (Cushing 1902; Doba and Reis 1972; Hoff and Reis 1970), or to cerebral ischemia (Dampney et al. 1979; Guyton 1948). The rapid onset of the seizure-induced cardiovascular response makes it unlikely that these changes were the result of an increase in intracranial pressure or cerebral ischemia. However, the kindled changes may be mediated through the same central pathways. The cardiovascular changes during cerebral ischemia are mediated by coactivation of sympathetic and parasympathetic systems at the level of the lower brain stem (Kumada et al. 1979). The kindled seizure may activate the same pathways to induce changes in cardiovascular function. 40.4.5â•… Clinical Relevance The etiology of sudden death in epileptic patients remains undeἀned and may not be the same in all patients. Evidence suggests that it may be associated with events occurring at the time of a seizure and perhaps is precipitated by a seizure. The ἀnding that many patients who are found dead have subsequently been shown to have subtherapeutic anticonvulsant levels supports this hypothesis (Dasheiff and Dickinson 1986). Of particular importance is the possibility that the interaction between seizures and the cardiovascular system underlies the sudden unexplained death that has been estimated to occur in 5–17% of epileptic patients (Leestma et al. 1984). Our observations obtained during kindled seizures may be relevant to similar cardiovascular changes seen during seizures in epileptic patients. Hypertension has been observed during spontaneous seizures (Meyer et al. 1966; Van Buren 1958; Van Buren and Ajmone-Marsan 1960; White et al. 1961) and ECT (Elliot et al. 1982). There are also case reports of ictal tachycardia (Marshall et al. 1983; Metz et al. 1978; Pritchett et al. 1980) and bradycardia (Devinsky et al. 1986; Coulter 1984; Gilchrist 1985; Kiok et al. 1986).

Acute Cardiovascular Response during Kindled Seizures

655

40.4.6â•…K indling as a Model of Seizure-Induced Arrhythmias and Sudden Unexplained Death In the past, a good experimental model to study this relationship was lacking. Experimental seizure models such as ECT, PTZ, and bicuculline have inherent limitations because the presence of anesthetics, paralytic agents, or widespread exposure of the CNS to chemical convulsants. In addition, these are models of generalized seizures, which precludes the examination of the effect of partial seizures on the cardiovascular system. The kindling model does not have these limitations. However, it is difficult to study the effect of interictal seizure activity on cardiovascular function in kindled animals. To study these changes, the PTZ model in anesthetized cats has been useful (Carnel et al. 1985; Lathers and Schraeder 1982; Lathers et al. 1984, 1987). Kindled seizures are initiated in a discrete region of the brain of unanesthetized, unrestrained animals. Kindling also progresses through a series of distinct stages that allows for the examination of the effect of partial and generalized seizures on the cardiovascular system. In light of our results, the kindling seizure model appears to be well suited for studies that examine the relationship between seizures and the cardiovascular system. Several questions remain to be answered. Do seizures have a progressive effect on the cardiovascular system? Is the severity of the seizure related to the degree of cardiac involvement? Is the quality of the arrhythmic response dependent on the site of origin of the seizure? Can repeated kindled seizures increase the susceptibility of kindled animals to cardiovascular changes associated with minimal discharges, that is, interictal spikes? All of these questions have yet to be addressed in the kindling seizure model.

40.5â•… Summary The effect of seizures on cardiovascular function was evaluated during kindled seizures in conscious rats. This relationship was also examined during kindling acquisition. The typical cardiovascular response during a generalized kindled seizure consisted of a large increase in blood pressure accompanied by a profound bradycardia during the ἀrst 20–30€ s of the seizure. Similar changes in heart rate and blood pressure were observed during amygdaloid- and olfactory bulb–kindled seizures as well as secondary spontaneous siezures, suggesting these changes were seizure dependent but not limited to seizures initiated in the amygdala. These cardiovascular changes were also present during partial seizures early in the kindling process. These results suggest that kindling is a useful seizure model in which to study the underlying mechanism of seizure-induced arrhythmias and possibly the clinical phenomenon of sudden unexplained death.

References Berne, R. M., and M. N. Levy. 1981. Cardiovascular Physiology, 46. St. Louis, MO: C. V. Mosby. Campbell, G. A., and I. L. Crawford. 1980. A gated electronic switch for stimulation and recording with a single electrode. Brain Res Bull 5: 485–486. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. The effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity. Pharmacology 20: 225–240.

656 Sudden Death in Epilepsy: Forensic and Clinical Issues Coulter, D. 1984. Partial seizures with apnea and bradycardia. Arch Neurol 41: 173–174. Crawford, I. L. 1986. Relationship of glutamic acid and zinc to kindling of the rat amygdala: Afferent transmitter systems and excitability in a model of epilepsy. In Excitatory Amino Acids and Epilepsy, ed. R. Schwarcz and Y. Ben-Ari, 611–623. New York, NY: Plenum. Cushing, H. 1902. Some experimental and clinical observations concerning states of increased intracranial tension. Am J Med Sci 124: 375–400. Dampney, R. A. L., M. Kumada, and D. J. Reis. 1979. Central neural mechanisms of the cerebral ischemic response. Circ Res 45: 48–62. Dasheiff, R. M., and L. J. Dickinson. 1986. Sudden unexpected death of epileptic patient due to cardiac arrhythmia after seizure. Arch Neurol 43: 194–196. Delgado, J. M., L. Mihailovic, and M. Sevillano. 1960. Cardiovascular phenomena during seizure activity. J Nerv Ment Dis 130: 477–487. De Olmos, J. S. 1972. The amygdaloid projection ἀeld in the rat as studied with the cupric-silver method. In The Neurobiology of the Amygdala, ed. B. Eleftheriou, 145–204. New York, NY: Plenum. De Olmos, J. S., and W. R. Ingram. 1972. The projection ἀeld of the stria terminalis in the rat brain: An experimental study. J Comp Neurol 146: 303–334. Devinsky, O., B. H. Price, and S. I. Cohen. 1986. Cardiac manifestations of complex partial seizures. Am J Med 80: 195–202. Doba, N., and D. J. Reis. 1972. Localization within the lower brainstem of a receptive area mediating the pressure response to increased intracranial pressure (the Cushing response). Brain Res 47: 487–491. Doba, N., H. R. Beresford, and D. J. Reis. 1975. Changes in regional blood flow and cardiodynamics associated with electrically and chemically induced epilepsy in the cat. Brain Res 90: 115–132. Elliot, D., D. Linz, and J. Kane. 1982. Electroconvulsive therapy. Arch Intern Med 142: 919–981. Faiers, A. A., F. R. Calaresu, and G. J. Mogenson. 1975. Pathway mediating hypotension elicited by stimulation of the amygdala in the rat. Am J Physiol 228: 1358–1366. Frysinger, R. C., J. D. Marks, R. B. Trelease, V. L. Schechtman, and R. M. Harper. 1984. Sleep states attenuate the pressor response to central amygdala stimulation. Exp Neurol 83: 604–611. Galosy, R. A., I. L. Crawford, and M. E. Thompson. 1982. Behavioral stress and cardiovascular regulation: Neural mechanisms. In Circulation, Neurobiology, and Behavior, ed. O. A. Smith, R. A. Galosy, and S. M. Weiss, 109–120. New York, NY: Elsevier. Gilchrist, J. M. 1985. Arrhythmogenic seizures: Diagnosis by simultaneous EEG/ECG recording. Neurology 35: 1503–1506. Goddard, G. V., D. C. McIntyre, and C. K. Leech. 1969. A permanent change in brain function resulting from daily electrical stimulation. Exp Neurol 25: 295–300. Goodman, J. H., R. W. Homan, and I. L. Crawford. 1990. Kindled seizures elevate blood pressure and induce cardiac arrhythmias. Epilepsia 31: 489–495. Guyton, A. C. 1948. Acute hypertension in dogs with cerebral ischemia. Am J Physiol 154: 45–54. Heinemann, H., G. Stock, and H. Schaefer. 1973. Temporal correlation of responses in blood pressure and motor reaction under electrical stimulation of limbic structures in unanesthetized, unrestrained cats. Pflügers Arch 343: 27–40. Hilton, S. M., and A. W. Zbrozyna. 1963. Amygdaloid region for defense reactions and its efferent pathway to the brainstem. J Physiol 165: 160–173. Hoff, J. T., and D. J. Reis. 1970. Localization of the regions mediating the Cushing response in CNS of the cat. Arch Neurol 23: 228–240. Homan, R. W., and J. H. Goodman. 1988. Endurance of the kindling effect is independent of the degree of generalization. Brain Res 447: 404–406. Hopkins, D. A., and G. Holstege. 1978. Amygdaloid projections to the mesencephalon, pons and medulla oblongata in the cat. Exp Brain Res 32: 529–547.

Acute Cardiovascular Response during Kindled Seizures

657

Iwata, J., K. Chida, and J. E. LeDoux. 1987. Cardiovascular responses elicited by stimulation of neurons in the central amygdaloid nucleus in awake but not anesthetized rats resemble conditioned emotional responses. Brain Res 418: 183–188. Johansson, B., and B. Nilsson. 1977. The pathophysiology of the blood brain barrier dysfunction induced by severe hypercapnia and by epileptic brain activity. Acta Neuropathol 38: 153–158. Keilson, M. J., W. A. Hauser, J. P. Magrill, and M. Goldman. 1987. ECG abnormalities in patients with epilepsy. Neurology 37: 1624–1626. Kiok, M. C., C. F. Terrence, G. H. Fromm, and S. Lavier. 1986. Sinus arrest in epilepsy. Neurology 36: 115–116. Kumada, M., R. A. L. Dampney, and D. J. Reis. 1979. Profound hypotension and abolition of the vasomotor component of the cerebral ischemic response produced by restricted lesions of the medulla oblongata in rabbit. Circ Res 45: 63–10. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–641. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital. Life Sci 34: 1919–1936. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Le Doux, J. E., A. Delbo, L. W. Tucker, G. Harshἀeld, W. T. Talman, and D. J. Reis. 1982. Hierarchic organization of blood pressure responses during the expression of natural behaviors in rat: Mediation by sympathetic nerves. Exp Neurol 78: 121–133. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Mameli, P., O. Mameli, E. Tolu, G. Padua, D. Giraudi, M. A. Caria, and F. Melis. 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29: 74–82. Marshall, D. W., B. F. Westmoreland, and F. W. Sharbrough. 1983. Ictal tachycardia during temporal lobe seizures. Mayo Clin Proc 58: 443–446. Meldrum, B. S. 1976. Neuropathology and pathophysiology. In A Textbook of Epilepsy, ed. J. Laidlaw and A. Richens, 314–354. Edinburgh: Churchill-Livingstone. Meldrum, B. S., and R. W. Horton. 1973. Physiology of status epilepticus in primates. Arch Neurol 28: 1–9. Meldrum, B. S., R. W. Horton, S. R. Bloom, J. Butler, and J. Keenan. 1979. Endocrine factors and glucose metabolism during prolonged seizures in baboons. Epilepsia 20: 527–534. Metz, S. A., J. B. Halter, D. Porte, and R. P. Robertson. 1978. Autonomic epilepsy. Ann Intern Med 88: 189–193. Meyer, J. S., F. Gotch, and E. Favale. 1966. Cerebral metabolism during epileptic seizures in man. Electroencephalogr Clin Neurophysiol 21: 10–22. Millhouse, O. E. 1969. A Golgi study of the descending medial forebrain bundle. Brain Res 15: 341–363. Mogenson, G. J., and F. R. Calaresu. 1973. Cardiovascular responses to electrical stimulation of the amygdala in the rat. Exp Neurol 39: 166–180. Petito, C. K., J. A. Schaeffer, and F. Plum. 1977. Ultrastructural characteristics of the brain and blood brain barrier in experimental seizures. Brain Res 127: 251–267. Pinel, J., and L. Rovner. 1978. Experimental epileptogenesis: Kindling-induced epilepsy in rats. Exp Neurol 58: 190–202. Plum, F., J. B. Posner, and B. Troy. 1968. Cerebral metabolic and circulatory responses to induced convulsions in animals. Arch Neurol 18: 1–13. Pritchett, E. L. C., J. O. McNamara, and J. J. Gallagher. 1980. Arrhythmogenic epilepsy: An hypothesis. Am Heart J 100: 683–688.

658 Sudden Death in Epilepsy: Forensic and Clinical Issues Racine, R. J. 1972. Modiἀcation of seizure activity by electrical stimulation. II. Motor seizure. Electroencephalogr Clin Neurophysiol 32: 281–294. Racine, R. J. 1978. Kindling, the ἀrst decade. Neurosurgery 3: 234–252. Reis, D. J., and M. C. Oliphant. 1964. Bradycardia and tachycardia following electrical stimulation of the amygdaloid region in monkey. J Neurophysiol 27: 893–912. Schwaber, J. S., B. S. Kapp, and G. Higgins. 1980. The origin and extent of direct amygdala projections to the region of the dorsal motor nucleus of the vagus and the nucleus of the solitary tract. Neurosci Lett 20: 15–20. Stock, G., K. H. Schlor, H. Heidt, and J. Buss. 1978. Psychomotor behavior and cardiovascular patterns during stimulation of the amygdala. Pflügers Arch 376: 177–184. Suzuki, R., C. Nitsch, K. Fujiwara, and I. Klatzo. 1984. Regional changes in cerebral blood flow and blood brain barrier permeability during epileptic seizures and in acute hypertension in rabbits. J Cereb Blood Flow Metab 4: 96–102. Takeuchi, Y., J. H. Mclean, and D. A. Hopkins. 1982. Reciprocal connections between the amygdala and parabrachial nuclei: Ultrastructural demonstration by degeneration and axonal transport of horseradish peroxidase in the cat. Brain Res 239: 583–588. Van Buren, J. M. 1958. Some autonomic concomitants of ictal automatism. Brain 81: 505–528. Van Buren, J. M., and C. Hopkins. 1960. Correlations of autonomic and EEG components in temporal lobe epilepsy. Arch Neurol 3: 683–703. Walker, J. E., J. A. Mikeska, and I. L. Crawford. 1981. Cyclic nucleotides in the amygdala of the kindled rat. Brain Res Bull 6: 1–3. Wasterlain, C. G. 1974. Mortality and morbidity from serial seizures. Epilepsia 15: 155–176. Westergaard, E., M. N. Hertz, and T. G. Bolwig. 1978. Increased permeability to horseradish peroxidase across cerebral vessels evoked by electrically-induced seizures in the rat. Acta Neuropathol 41: 73–80. White, P. T., P. Grant, J. Mosier, and A. Craig. 1961. Changes in cerebral dynamics associated with seizures. Neurology 11: 354–361.

DBA Mice as Models of Sudden Unexpected Death in Epilepsy Carl L. Faingold Srinivasan Tupal Yashanad Mhaskar Victor V. Uteshev

41

Contents 41.1 41.2 41.3 41.4

Introduction Mechanisms of SUDEP—Cardiac Mechanisms of SUDEP—Respiratory SUDEP Animal Model in DBA/2 Mice 41.4.1 Duration of Susceptibility to Respiratory Arrest in DBA/2 Mice 41.4.2 Genetic Basis of Seizure Susceptibility in DBA/2 Mice 41.5 Regulation of Respiration by Serotonin 41.6 Expression of 5-HT Receptor Proteins and Transporter in C57BL/6J and DBA/2 Mice 41.7 Sudden Infant Death Syndrome and SUDEP 41.8 Regulation of Seizure Susceptibility by Serotonin 41.9 Enhanced Serotonin Action Reduces SUDEP in DBA/2 Mice 41.10 Reduced Serotonin Action Increases SUDEP in DBA/2 Mice: Serotonin Antagonists 41.11 DBA/1 Mice as a Chronic Model of SUDEP 41.12 Discussion Acknowledgments References

659 660 661 661 663 663 663 665 666 667 667 669 669 671 672 672

41.1â•…Introduction Published clinical observations indicate that a major proposed cause for sudden unexpected death in epilepsy (SUDEP) patients is respiratory arrest that is commonly associated with a generalized convulsive seizure. DBA mice were known to exhibit generalized convulsive seizures in response to high intensity acoustic stimuli that lead to respiratory arrest, whereas recent data indicate that the electrocardiogram of these mice remains active for several minutes beyond respiratory arrest onset. These data have led to the proposal that DBA/2 mice are a useful model of respiratory arrest–mediated SUDEP. We have documented that SUDEP in DBA/2 mice can be prevented in >90% of these mice by rapid intervention with mechanical respiratory support, allowing multiple testing of individual mice for investigating possible preventative measures.

659

660 Sudden Death in Epilepsy: Forensic and Clinical Issues

Serotonin (5-HT) is a critically important neurotransmitter in brainstem respiratory centers that is known to play an important role in controlling normal respiration and in enhancing respiratory rate in response to elevated carbon dioxide levels. Our studies indicate drugs that enhance the activation of 5-HT receptors directly or indirectly, including a selective serotonin reuptake inhibitor and selective 5-HT receptor agonists, reduce or block respiratory arrest in DBA/2 mice at doses that do not block seizures. By contrast, a 5-HT receptor antagonist will induce SUDEP in the small percentage of DBA/2 mice that exhibit tonic seizures without respiratory arrest. Our recent data indicate that the expression of speciἀc 5-HT2C (which is inhibitory), 5-HT3, and 5-HT4 receptor proteins in the rostral ventral medulla respiratory region of DBA/2 mice are signiἀcantly diminished, whereas expression of the excitatory 5-HT2B receptors is signiἀcantly enhanced compared to seizure-resistant control (C57BL/6J) mice. By contrast, no statistically signiἀcant difference in expression of serotonin transporter protein between control and DBA/2 mice is seen. DBA/2 mice exhibit a relatively short duration of consistent susceptibility to respiratory arrest (~7 days), but we have recently observed that a closely related strain, DBA/1 mice, exhibits a >4-week susceptibility to seizure-induced respiratory arrest. The susceptibility of DBA/1 mice to respiratory arrest can be blocked, both acutely and after semichronic (5-day) treatment, with a selective serotonin reuptake inhibitor (fluoxetine, Prozac), with the mice subsequently returning to respiratory arrest susceptibility. These ἀndings indicate the potential usefulness of DBA/1 and DBA/2 mice as models of human SUDEP and drugs, such as selective serotonin reuptake inhibitors and selective 5-HT receptor agonists, that have proven effective in preventing respiratory arrest in these mice have the potential to be effective preventative agents for human SUDEP.

41.2â•… Mechanisms of SUDEP—Cardiac SUDEP is a devastating consequence of epilepsy and has long been recognized as an important concern in both adult and pediatric patients with epilepsy, particularly those who experience generalized convulsive seizures (Langan et al. 2000; Donner et al. 2001; Nilsson et al. 2001; McGregor and Wheless 2006; Tomson et al. 2008). A case-control study concluded that SUDEP is a seizure-related phenomenon and that control of generalized convulsive seizure and nocturnal supervision may be protective against SUDEP (Langan et al. 2005). The overwhelming seizure pattern (>80%) exhibited by SUDEP patients before death was syndromes that involved a generalized convulsive seizure, particularly tonic–clonic seizures (Opeskin and Berkovic 2003; Tomson et al. 2008). Diminished respiratory function and irregular cardiac rhythm in patients and animals are commonly associated with generalized convulsive seizure (Schraeder and Lathers 1989; Nashef et al. 1996; Langan et al. 2000; So et al. 2000; Ryvlin et al. 2006; Bateman et al. 2008; Lathers et al. 2008). Evidence has been presented that implicates cardiac events, including arrhythmias and asystole, as a major cause for SUDEP with varying degrees of clinical validation (Nashef et al. 1995; Langan et al. 2000; Darbin et al. 2002; Davis and McGwin 2004; So 2006, 2008; St-John et al. 2006; Tomson et al. 2008).

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

661

41.3╇ Mechanisms of SUDEP—Respiratory The second major proposed cause for SUDEP is respiratory malfunction. Thus, in a number of witnessed cases of human SUDEP, respiratory difficulties were observed before death in ~75% of cases (Langan et al. 2000; Lear-Kaul et al. 2005). A recent report on human SUDEP indicated that death occurring in the postictal period primarily resulted from respiratory arrest (Lear-Kaul et al. 2005). SUDEP and near-SUDEP cases have been observed in epilepsy monitoring units [in the ongoing mortality in Epilepsy Monitoring Unit Study (MORTEMUS)], and a signiἀcant percentage of these cases exhibited respiratory difficulties (Ryvlin et al. 2006; Bateman et al. 2010). A recent study in an epilepsy monitoring unit indicated that signiἀcantly compromised respiratory function was found in most observed seizures, and patients exhibiting generalized convulsive seizure showed ~50% decline in respiration (Bateman et al. 2008; Seyal et al. 2010). Thus, the preponderance of clinical observations suggests an important role of respiratory difficulties during and after generalized convulsive seizure and respiratory arrest in observed cases of SUDEP and near-SUDEP cases. A recent report observed that two human SUDEP cases in an epilepsy monitoring unit that began with respiratory difficulties subsequently led to cardiac arrhythmias prior to death (Bateman et al. 2010), which is similar to what is seen in the DBA mouse models of SUDEP, as discussed below (Faingold et al. 2010a).

41.4â•… SUDEP Animal Model in DBA/2 Mice Animal studies of SUDEP had been hampered by the paucity of suitable models, but in recent years, the dilute brown non-Agouti (DBA/2) mouse has been proposed as a model of SUDEP (Venit et al. 2004; Tupal and Faingold 2006). DBA mice are genetically susceptible to generalized convulsive seizures evoked by acoustic stimulation (audiogenic seizures) that results in sudden death. Signiἀcant evidence indicates that the sudden death in the DBA/2 mice is caused by generalized convulsive seizure–associated respiratory arrest and that several other inbred rodent strains exhibit generalized convulsive seizure when exposed to intense acoustic stimulation. Susceptibility to audiogenic seizures can be induced in previously normal rodents with several types of treatments (Faingold 2002). In our initial study of 168 DBA/2 mice, behavioral responses to intense acoustic stimulation (122 dB SPL re: 0.0002 dyn/cm2) using broadband acoustic stimuli (electrical bell) were evaluated. All of the DBA/2 mice exhibited generalized convulsive seizures, which were followed by postictal depression of behavior in 85%. Speciἀc behaviors during audiogenic seizures included wild running (100% incidence), generalized clonus (100%), and tonic seizures culminating in tonic hind limb extension (85%). During the postictal period, most DBA/2 mice (75%) displayed respiratory arrest (Figure 41.1). Ten percent of DBA/2 mice exhibited tonic extension without respiratory arrest. The remaining (15%) did not exhibit tonic extension or respiratory arrest. Respiratory arrest in DBA/2 mice occurred almost exclusively after convulsive behavior (Tupal and Faingold 2006). After audiogenic seizures, a large proportion of DBA/2 mice exhibited sudden death unless respiration was supported (Willott and Henry 1976; Seyfried and Glaser 1985; Tupal and Faingold 2006). DBA/2 mice that exhibited respiratory arrest could be consistently resuscitated if resuscitation was initiated within 2–5 s after the ἀnal deep respiratory gasp (Tupal and Faingold

662 Sudden Death in Epilepsy: Forensic and Clinical Issues DBA/1 100

% Seizure + RA

80 60

DBA/2

40 20 0

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Age of mice (days)

Figure 41.1╇ Susceptibility of DBA/2 and DBA/1 mice to audiogenic seizures with respiratory

arrest changes differentially with age and audiogenic seizure experience. The two strains of DBA mice (DBA/2 and DBA/1) display differing patterns of susceptibility to respiratory arrest (RA) after audiogenic seizures. The susceptibility to respiratory arrest after audiogenic seizures was seen in 75% of DBA/2 mice on postnatal day 21, which declined progressively to 0% on day 35 due to progressive hearing loss. (Note: no change in respiratory arrest susceptibility in DBA/2 mice was noted with retest and not all tests are shown.) Each individual DBA/2 mouse that exhibited audiogenic seizures with respiratory arrest showed consistent susceptibility for ~7 days. The incidence of audiogenic seizures with respiratory arrest in DBA/1 mice increased from 33% on postnatal day 21 to 100% by the third test on day 23. If testing began at later ages, it still required three tests for DBA/1 mice to exhibit 100% susceptibility. Unlike the DBA/2 mice, the DBA/1 mice continued to display respiratory arrest susceptibility for >35 (up to 50) days.

2006). This involves placing the mouse in a supine position and placing a polyethylene tube connected to outflow of a rodent respirator over the nostrils of the mouse for <60€s by which time the mouse will breathe independently (Tupal and Faingold 2006). Each€DBA/2 mouse that€ exhibited respiratory arrest showed a stable pattern of seizure€behaviors, including respiratory arrest, at ages 21–25 days, but susceptibility to respiratory arrest declined€subsequently due in part to loss of hearing (Tupal and Faingold 2006). This SUDEP-like syndrome in DBA/2 mice occurs immediately after the generalized convulsive seizure. Evidence indicates that cessation of respiration and subsequent anoxia are likely to be the key lethality-inducing events because sudden death can be prevented in€ >90% of DBA/2 mice that are given prompt respiratory support by an experienced exÂ�perimenter (Tupal and Faingold 2006). However, each experimenter requires practice to reach this level of success. Thus, for one of the authors (S. T.), in the initial month of experimentation, only 75% of the DBA/2 mice were successfully resuscitated, but by the second month, the rate exceeded 90%, and by the seventh month, it approached 99%. Increasing oxygen supply in the vicinity of DBA/2 mice also increases survival rates (Venit et al. 2004). Recent data indicate that during respiratory arrest in DBA/2 mice when the animal is not breathing, the electrocardiogram remains active for 4–6 min with decreasing and increasing heart rates and abnormal rhythms that alternate before cessation of the electrocardiogram (Figure 41.2), suggesting that some cardiac function remains after respiratory arrest.

DBA Mice as Models of Sudden Unexpected Death in Epilepsy (a)

(b)

(c)

663

(d)

Figure 41.2╇ A typical example of electrocardiogram findings in DBA/2 mice under nonseizure (anesthetized) conditions and electrocardiogram changes associated with seizure-induced respiratory arrest. (a) Electrocardiogram of a DBA/2 mouse under ketamine/xylazine anesthesia (85/5 mg/kg). (b–d) Taken from a different DBA/2 mouse; electrocardiogram changes after an audiogenic generalized convulsive seizure. Reduced electrocardiogram rate immediately after respiratory arrest is shown in (b), whereas (c) shows the electrocardiogram rate and pattern ~1 min after respiratory arrest onset from the same mouse, illustrating evidence of abnormal cardiac rhythm. In d, the electrocardiogram can be seen to be nearly absent in the same mouse ~4 min after respiratory arrest onset. The electrocardiogram was completely absent subsequent to this trace (trace length = 5 s, amplitude in (a) = 0.5 mv, (b–d) = 0.4 mv, two forelimb needle electrodes plus ground configuration).

41.4.1â•… Duration of Susceptibility to Respiratory Arrest in DBA/2 Mice Seizure susceptibility is known to decrease with age in DBA/2 mice (Collins 1972; Turner and Willott 1998). When the seizure-inducing stimulus was presented daily at 21–25 days of age, resuscitated DBA/2 mice subsequently exhibit seizures ending with respiratory arrest each time (Tupal and Faingold 2006). However, after 25 days of age, the occurrence of respiratory arrest declined; and when the DBA/2 mice reached 29 days of age, a loss of seizure susceptibility occurred in ~30% of mice, and ~40% of the mice that remained susceptible to seizures no longer exhibited tonic extension with a greatly reduced incidence of respiratory arrest (to ~25%). The incidence of audiogenic seizures with respiratory arrest reached zero by 35 days of age (Figure 41.1). 41.4.2â•… Genetic Basis of Seizure Susceptibility in DBA/2 Mice Seizure susceptibility of DBA/2 mice is associated with the Asp gene family (Collins and Fuller 1968) and is influenced by three gene loci. One locus, Asp-1, is located on chromosome 12 between Ah and D12Nyu1; another locus, Asp-2, is on chromosome 4, located within an eight-centimorgan segment distal to b; and Asp-3 is linked to c on chromosome 7 (Neumann and Collins 1991, 1992). These three loci account for most of the heritable variation in susceptibility to audiogenic seizures in crosses of DBA/2 and C57BL/6J mice, and susceptibility to audiogenic seizures is influenced by genomic imprinting (Neumann and Collins 1991).

41.5â•…Regulation of Respiration by Serotonin Considerable basic research has established the rostral ventral medulla, including the nucleus of the solitary tract and the pre-Bötzinger complex of the brainstem, as principal

664 Sudden Death in Epilepsy: Forensic and Clinical Issues

centers of respiration control (Bonham et al. 2006; Kubin et al. 2006; Potts 2006). The pre-Bötzinger complex is a group of neurons that has been localized in the ventrolateral medulla that exerts critical control of respiratory rhythm, and the intrinsic pacemaker currents in these neurons are modulated by several neurotransmitters (Ramirez and Viemari 2005; Brisson-Thoby and Greer 2008), including 5-HT. The respiratory sensory afferents from the respiratory tract (lungs and airways) enter the brainstem via the vagus nerve and terminate in the nucleus of the solitary tract (Boscan et al. 2002; Potts 2002). The nucleus of the solitary tract receives and integrates afferents from the respiratory system and projects to the medullary respiratory nuclei (Richerson and Getting 1992; Li et al. 1999). Several thalamic, hypothalamic, cortical, medullary, pontine, and midbrain nuclei send modulatory projections to the brainstem respiratory centers (Terreberry and Neafsey 1987; Agarwal and Calaresu 1990; Otake et al. 1994). Sensory signals are also modulated within the brainstem, as the excitatory and inhibitory synaptic inputs to second-order neurons as well as neuronal properties exhibit plastic changes as a result of stimulation, neuronal activity, or action of neuromodulators, particularly 5-HT (Chen and Bonham 1998; Kline et al. 2002; Bonham et al. 2006). As noted above, 5-HT is one of the major brain neurotransmitters and is involved in generating and transmitting respiratory rhythm (Bonham 1995; Pena and Ramirez 2002; Wong -Riley and Liu 2005). Endogenous 5-HT is known to play an important role in modulating the activity of neuronal networks in the brainstem and is required for normal respiration (Al-Zubaidy et al. 1996). Speciἀc subtypes of 5-HT receptors are thought to be selectively relevant to control of respiration and to epilepsy (Carley and Radulovacki 1999; Richter et al. 2003; Toczek et al. 2003). Pharmacological, electrophysiological, and immunohistochemical studies indicate that expression of at least 7 of the 14 known 5-HT receptor subtypes occurs in the brainstem (Jordan 2005; Raul 2003). Of those subtypes, the subfamily of 5-HT2 receptors (i.e., 5-HT2A, 5-HT2B, and 5-HT2C) is especially relevant to epilepsy and seizures (see below). Although all three 5-HT2 receptors act via activation of Gq/11 receptors coupled to phospholipase C, their net effects are not identical and stimulation of 5-HT2 receptors could produce either excitation or inhibition. In the brainstem and speciἀcally in the nucleus of the solitary tract, activation of 5-HT2C receptors is usually inhibitory (Jordan 2005). By contrast, activation of 5-HT2A and 5-HT2B receptors exhibited predominantly excitatory effects (Jordan 2005). Consistent with this mechanism, the absence of 5-HT2C receptors in transgenic mice is associated with audiogenic seizure susceptibility that can result in death (Applegate and Tecott 1998). Vagus nerve stimulation is an effective antiepileptic therapy in patients (Theodore 2005; Nemeroff 2007) and has also been shown to increase the ἀring of 5-HT neurons in the dorsal raphe, which is mediated, in part, via projections from nucleus of the solitary tract (Dorr and Debonnel 2006). Modulation of respiratory network function by 5-HT may result, in part, from activation of 5-HT3 receptors (Kubin et al. 2006). Endogenous activation of 5-HT2A receptors is required for the generation of the respiratory rhythm in vitro (Pena and Ramirez 2002). Enhanced activation of 5-HT receptors after generalized convulsive seizure may be an effective approach to prevention of seizure-induced respiratory arrest. Activation of 5-HT receptors (particularly 5-HT4 receptors) (Manzke et al. 2003; Meyer et al. 2006) effectively controls respiratory disturbances, and activation of 5-HT pathways may be an important approach to treatment of life-threatening respiratory disorders in patients (Richter et al. 2003). Blockade of 5-HT1 and 5-HT2 receptors by the 5-HT antagonist, methysergide, decreased respiratory frequency in vivo and in vitro

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

665

(Bodineau et al. 2004). A selective serotonin reuptake inhibitor, paroxetine, signiἀcantly increased respiratory rates in rodents (Olsson et al. 2004).

41.6â•…Expression of 5-HT Receptor Proteins and Transporter in C57BL/6J and DBA/2 Mice The deἀcit in serotonergic activation in DBA/2 mice, which are susceptible to audiogenic seizures and respiratory arrest, may result from altered expression of speciἀc subtypes of 5-HT receptors or the 5-HT transporter in respiratory centers of the brainstem. To test this hypothesis, antibodies against 5-HT2B, 5-HT2C, 5-HT3B, and 5-HT4 receptor proteins and the 5-HT transporter were used in Western blot experiments to compare the expression levels of these proteins in the caudal brainstem of DBA/2 vs. control (C57BL/6J) mice. C57BL/6J mice are genetically related to DBA/2 mice but do not exhibit audiogenic seizure susceptibility. The presence of 5-HT2B, 5-HT2C, 5-HT3B, and 5-HT4 receptors in the brainstem has been previously observed, and 5-HT neurotransmission in brainstem respiratory networks is known to contribute to increasing frequency of respiration in response to elevated carbon dioxide levels in normal animals (Feldman et al. 2003; Di Giovanni et al. 2006; Benarroch 2007; Hodges and Richerson 2008). Selective serotonin reuptake inhibitors increase the frequency of respiratory rhythm in medullary brain slices (Doi and Ramirez 2008). Null mutant mice lacking serotonin 5-HT2C receptors are susceptible to generalized convulsive seizure induced by sound, and these audiogenic seizures also result in postictal respiratory arrest (Brennan et al. 1997), supporting the hypothesis that an alteration in the activation of speciἀc subtypes of 5-HT receptors may result in respiratory deἀcits. In our experiments, the tissue extracts from the caudal brainstem of C57BL/6J and DBA/2 mice were immunostained on the same blots. Normalized (to the amount of actin) densities of the receptor protein from the two strains of mice were compared. In each experiment, caudal brainstems of DBA/2 and C57BL/6J mice of equal age were collected. Possible changes in serotonergic activation in DBA/2 mice may result from reduced levels of speciἀc inhibitory 5-HT receptors whose deἀciency is associated with occurrence of seizures, for example, 5-HT2C receptors (Brennan et al. 1997; Applegate and Tecott 1998). Possible differences of expression levels of 5-HT receptor proteins (5-HT2B, 5-HT2C, 5-HT3B, and 5-HT4) in the caudal brainstem of DBA/2 and C57BL/6J (control) mice were compared (Figure 41.3 and Table 41.1). The results indicated that DBA/2 mice express€signiἀcantly lower levels of 5-HT2C, 5-HT3B, and 5-HT4 receptors than control mice. Signiἀcant differences, 26% (n = 10; p < 0.02), 21% (n = 9; p < 0.035), and 28% (n = 7; p < 0.025) in the expression of 5-HT2C, 5-HT3B, and 5-HT4 receptor proteins, respectively, were detected in the caudal brainstem of DBA/2 relative to control mice (Uteshev et al. 2010). These observations are consistent with the hypothesis that a deἀciency in serotonergic activation in the brainstem of DBA/2 mice contributes to susceptibility to audiogenic seizures and respiratory arrest. By contrast, excitatory 5-HT2B receptors were expressed at signiἀcantly higher (n = 9; p < 0.001) levels in DBA/2 mice (143%) than in C57BL/6J mice (100%). The signiἀcant differences in expression of 5-HT receptor subtypes in the brainstem of audiogenic seizure-non-susceptible mice vs. seizure-prone DBA/2 mice suggest an important role for 5-HT receptors (including excitatory 5-HT2B and inhibitory 5-HT2C) in susceptibility to seizures and respiratory arrest in this DBA/2 SUDEP model. Understanding the molecular

666 Sudden Death in Epilepsy: Forensic and Clinical Issues (a)

(b)

5-HT2B 5-HT2B

5-HT2C

C57 DBA/2

(56 kDa)

Actin (43 kDa)

100%

100 75 50 25 0

100%

100 Relative expression (%)

Relative expression (%)

125

Actin (43 kDa) 143.5±24.2% p < 0.01 n=7

150

C57 DBA/2

5-HT2C (46 kDa)

74.3±32.4% p < 0.02 n = 10

75 50 25 0 C57BL/6J DBA/2

C57BL/6J DBA/2

Figure 41.3╇ Western blot analysis of 5-HT2B and 5-HT2C receptor proteins in C57BL/6J and DBA/2 mice. In each set of experiments, the caudal portions of the brainstem, containing the respiratory centers, of three DBA/2 and three C57BL/6J (control) mice were collected for Western blot analysis. Fresh brain tissue from each group of mice was homogenized separately, and specific antibodies against inhibitory 5-HT2C and excitatory 5-HT2B were used for labeling. The results indicate that the caudal brainstems of DBA/2 mice express significantly greater amounts of 5-HT 2B (143.5 ± 24.2% of control mice (a) receptors, which are excitatory, and significantly lower amounts of 5-HT 2C (74.3 ± 32.4% of control mice (b) receptors, which are inhibitory, compared to control mice. These results are consistent with the behavioral differences between DBA/2 (seizure-Â�susceptible) and C57BL/6J (seizure-resistant) and the previous finding that 5-HT2C knockout mice exhibit susceptibility to audiogenic seizures and respiratory arrest. (Reprinted by permission from Macmillan Publishers Ltd: Brennan, T. J., et al. Nat Genet. 16:387–390, 1997. Copyright 1997.)

basis of susceptibility to SUDEP can assist in developing more selective 5HT-based treatments for prevention of SUDEP.

41.7╅ Sudden Infant Death Syndrome and SUDEP The relevance of serotonergic neurotransmission to respiration is also supported by the abnormalities in 5-HT neurotransmission, particularly polymorphisms in the 5-HT Table 41.1╅ Expression of Serotonergic Receptor and Transporter Proteins in the Caudal Brainstem of C57BL/6J and DBA/2 Mice C57BL/6J DBA/2 No. of groups (three€animals/group) p value

5-HT2B

5-HT2C

5-HT3B

5-HT4

Transporter

100 143.5 ± 24.2 n=9

100 74.3 ± 32.4 n = 10

100 78.5 ± 25.1 n=9

100 72.0 ± 24.2 n=7

100 116.0 ± 21.0 n=9

<0.001

<0.02

<0.035

<0.025

>0.05

Note: Statistically signiἀcant differences in the expression levels are marked in bold. The corresponding p values and the number of experiments are provided. Animals were grouped to increase the amount of proteins. Each group contained three animals of the same age.

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

667

transporter genes, in sudden infant death syndrome, a disorder of breathing in infants (Moon et al. 2007). The 5-HT abnormalities in sudden infant death syndrome are localized in the rostral ventral medullary regions that control respiration (Kinney et al. 2009). Genetically modiἀed mice with near-complete absence of central 5-HT neurons exhibit signiἀcant breathing defects (Hodges et al. 2008). Our recent protein assays indicated that the protein expression levels of the 5-HT transporter in control (C57BL/6J) and DBA/2 are not signiἀcantly different (Table 41.1). These data suggest that, although respiratory difficulties in sudden infant death syndrome and this SUDEP model are both associated with deἀcits in 5-HT neurotransmission, the pathophysiology of the 5-HT system differs in these two disorders.

41.8â•…Regulation of Seizure Susceptibility by Serotonin 5-HT is also implicated in control of seizures. Anticonvulsant effects in both humans and rat epilepsy models are seen with drugs that enhance the action of 5-HT, and a rat model of generalized epilepsy, including rats susceptible to audiogenic seizures, is associated with a profound loss of 5-HT throughout the brain (Welsh et al. 2002; Merrill et al. 2003). Selective serotonin reuptake inhibitors, such as fluoxetine, exert anticonvulsant effects in several seizure models, and diminished 5-HT neurotransmission is seen in genetic epilepsy models (Browning et al. 1997; Applegate and Tecott 1998; Hernandez et al. 2002). Reduced 5-HT1A receptor binding has been observed in certain forms of human epilepsy (Merlet et al. 2004; Meschaks et al. 2005). 5-HT2C receptor knockout mice exhibit susceptibility to audiogenic seizures; and these animals exhibit respiratory arrest if the seizure pattern includes tonic hind limb extension, which is considered to be indicative of a very severe convulsion (Heisler et al. 1998). A recent retrospective study indicates that the respiratory dysfunction observed in certain epilepsy patients during seizures was signiἀcantly lower in patients who were taking selective serotonin re-uptake inhibitors as compared to patients not taking these agents (Bateman et al. 2010b).

41.9â•… Enhanced Serotonin Action Reduces SUDEP in DBA/2 Mice The selective serotonin reuptake inhibitor, fluoxetine (Prozac), is known to elevate synaptic 5-HT levels (Malagie et al. 1995), and administration of this agent signiἀcantly reduced respiratory arrest after seizures in DBA/2 mice (Tupal and Faingold 2006). Thus, signiἀcant reductions in respiratory arrest incidence were observed in the DBA/2 mice that received fluoxetine (15–25 mg/kg, i.p., Figure 41.4) compared to saline (Tupal and Faingold 2006). A signiἀcant reduction in the severity of seizures and respiratory arrest suppression began to be observed at the 25 mg/kg dose, which was not present at lower doses. This ἀnding is consistent with previous studies that indicated that selective serotonin reuptake inhibitors, including fluoxetine, exert anticonvulsant effects except in toxic doses (Dailey and Naritoku 1996; Specchio et al. 2004; Merrill et al. 2005). Return to respiratory arrest susceptibility in DBA/2 mice after ending fluoxetine administration was observed in most animals by 72 h after drug administration ceased, which is consistent with the half-life of this agent (Holladay et al. 1998). Suppression of respiratory arrest in this SUDEP model occurred after fluoxetine administration in DBA/2 mice in doses that did not affect seizure

668 Sudden Death in Epilepsy: Forensic and Clinical Issues

Dose (mg/kg)

10

15

20

FLUOXETINE

RECOVERY POST-DRUG

CONTROL

CONTROL

FLUOXETINE RECOVERY POST-DRUG

0

RECOVERY POST-DRUG

FLUOXETINE DRUG

CONTROL

RECOVERY POST-DRUG

DRUG FLUOXETINE

50

#

#

*

CONTROL

% Respiratory arrest

100

25

Figure 41.4╇ Effect of administration of certain doses of a selective serotonin reuptake inhib-

itor, fluoxetine (Prozac), to reduce the incidence of susceptibility to respiratory arrest after audiogenic seizures in DBA/2 mice. Black bars, vehicle (normal saline control); dark gray bars, the incidence of respiratory arrest 30 min after fluoxetine [10 (n = 5), 15 (n = 10), 20 (n = 12), or 25 (n = 15) mg/kg, i.p.]; and gray bars, recovery of respiratory arrest susceptibility at 24–72 h. Significantly different from saline control. *p < 0.05; #p < 0.005 (Wilcoxon signed ranks test). n indicates the number of animals. (Modified from Tupal, S., and Faingold, C. L., Epilepsia, 47, 21–26, 2006. With permission.)

severity, indicating that the effect on respiratory arrest was not dependent on an anticonvulsant effect. The onset of the respiratory arrest suppression by fluoxetine began at 30 min, which is consistent with the time course of signiἀcant elevations in 5-HT levels that occur in several rodent brain areas (Malagie et al. 1995). These ἀndings support the importance of enhancing 5-HT neurotransmission in prevention of respiratory arrest in this SUDEP model. The major clinical use for fluoxetine is for the treatment of depression, and this agent and similar selective serotonin reuptake inhibitors are commonly used to treat epileptic patients who have comorbid depression (LaFrance et al. 2008). However, the respiratory arrest–suppressive effect in DBA/2 mice occurs much earlier than the several weeks of treatment often required for remission of depression clinically (Tupal and Faingold 2006). The time course of the effect of fluoxetine in DBA/2 mice is consistent with half-life of this agent in mice (Holladay et al. 1998) and the onset of the therapeutic effects of antidepressants, such as fluoxetine, in treating certain human sleep disorders, which also does not require weeks of treatment (Nishino and Mignot 1997). The onset of the antidepressant effect in humans is currently thought to be delayed because the increase in 5-HT levels results in neuroplastic brain changes that take time to develop (Racagni and Popoli 2008). Recent data indicate that treatment of DBA/2 mice with a 5-HT 2C agonist, mCPP (m-chlorophenyl piperazine, 10 mg/kg, i.p.), 30 min before testing resulted in a signiἀcant decrease in the incidence of respiratory arrest (Figure 41.5). Treatment of DBA/2 mice with another selective serotonin 5-HT2C receptor agonist, RO 60-0176 (50 mg/kg, i.p.), also reduced the incidence of respiratory arrest (Faingold et al. 2008). Interestingly, systemic administration of a 5-HT1A partial agonist (buspirone, 10 mg/kg i.p.) signiἀcantly reduced the incidence of SUDEP but only to 50%, but this agent did not reduce the incidence

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

669

20 0

* Saline

mCPP

*

RA

AGS

RA

40

AGS

60

RA

80

AGS

% Seizure behavior

100

Post-drug

Figure 41.5╇ Effect of administration of a selective serotonin receptor agonist, m-chlorophenyl piperazine (mCPP), on the incidence of seizure-associated behaviors, audiogenic seizures, and the subsequent respiratory arrest in DBA/2 mice. The 5-HT2C agonist (mCPP, 10 mg/kg, i.p.), but not saline (vehicle), blocked respiratory arrest (RA) in DBA/2 mice at 30 min, without reducing the incidence of audiogenic seizure susceptibility. Post-drug return of respiratory arrest susceptibility occurred at 24–72 h. Black bars, incidence of audiogenic seizures; gray bars, the incidence of respiratory arrest. Number of animals was four in each group. Significant difference from saline control. *p < 0.05 (Wilcoxon signed ranks test).

of respiratory arrest in DBA/2 mice any further despite doubling the dose, which is in keeping with its incomplete effect on 5-HT receptors. These ἀndings further support the importance of enhancing 5-HT neurotransmission in prevention of respiratory arrest in this SUDEP model. These data suggest that these novel serotonergic agonists that enhance 5-HT neurotransmission more selectively may potentially be useful in patients for prevention of SUDEP.

41.10â•…Reduced Serotonin Action Increases SUDEP in DBA/2 Mice: Serotonin Antagonists As noted above, ~10% of DBA/2 mice exhibit audiogenic seizures with tonic extension but do not exhibit respiratory arrest (Tupal and Faingold 2006). Cyproheptadine, a nonselective 5-HT receptor antagonist (1 or 2 mg/kg, i.p.), was administered to those DBA/2 mice that exhibit tonic extension but not respiratory arrest, and signiἀcant increases in the incidence of respiratory arrest after seizure were observed in these mice 30 min after injection (Tupal and Faingold 2006). A return to generalized convulsive seizure without respiratory arrest was observed 24 h after cyproheptadine treatment in these DBA/2 mice (Figure 41.6). These ἀndings further support the importance of 5-HT neurotransmission in control of respiratory arrest in this SUDEP model. These data suggest that the use of agents that reduce 5-HT neurotransmission may be problematic in epileptic patients.

41.11â•…DBA/1 Mice as a Chronic Model of SUDEP Susceptibility to respiratory arrest in DBA/2 mice lasts for approximately 1 week, which allows these mice to serve as an acute model of SUDEP. Recent work unexpectedly indicated

670 Sudden Death in Epilepsy: Forensic and Clinical Issues 100

*

80

Dose (mg/kg)

1

POST-DRUG

DRUG CYPRO

0

CONTROL

20

RECOVERY POST -DRUG

40

DRUG CYPRO

60 CONTROL

% Respiratory arrest

#

2

Figure 41.6╇ Effect of administration of a nonselective serotonin receptor inhibitor, cyproheptadine (CYPRO), to increase the incidence of susceptibility to respiratory arrest after audiogenic seizures in DBA/2 mice that did not initially exhibit respiratory arrest. Black bars, saline control group; dark gray bars, CYPRO group, the incidence of respiratory arrest 30 min after CYPRO (1–2 mg/kg, i.p., in 9 and 14 mice, respectively); and gray bars, post-drug recovery from respiratory arrest susceptibility at 24–72 h. Significantly different from saline control. *p < 0.01; #p < 0.005 (Wilcoxon signed ranks test). (Modified from Tupal, S., and Faingold, C. L., Epilepsia, 47, 21–26, 2006. With permission.)

that a closely related mouse strain, DBA/1 mice, exhibits respiratory arrest susceptibility lasting for >4 weeks (Figure 41.1) (Faingold et al. 2008). Thus, DBA/1 mice allow monitoring of more prolonged drug treatments for RA prevention, as would be necessary for patients. This development in the DBA/1 mice was unexpected because the information from the suppliers did not indicate that these mice were susceptible to audiogenic seizures. DBA/1 mice actually exhibited a higher incidence of audiogenic seizure susceptibility than DBA/2 mice. The seizures in the DBA/1 mice were followed by respiratory arrest, and unless given prompt respiratory support, the mice would exhibit sudden death. The DBA/1 mice exhibited tonic extension when exposed to intense acoustic stimulation, and respiratory arrest followed tonic extension in 100% of DBA/1 mice (n = 24) that received three seizures. The incidence of respiratory arrest with the ἀrst seizure was 33%, but the incidence rose to 100% with the third seizure of the DBA/1 mice (Figure 41.1). The initial studies also indicate that sudden death can be prevented in >90% of DBA/1 mice by rapid resuscitation by an experienced researcher. DBA/1 mice remained susceptible to seizures and respiratory arrest, which continued unabated for as long as 80 days of age, allowing these mice to serve as a chronic SUDEP model. The continued susceptibility to respiratory arrest was surprising because the DBA/2 mice lose their susceptibility to respiratory arrest and seizure severity, which declines by 29 days of age, as noted above. The duration of susceptibility to seizure-associated respiratory arrest in DBA/1 mice was signiἀcantly greater than in the DBA/2 mouse strain. This is likely because DBA/2 mice develop a hearing deἀcit at this age, whereas DBA/1 mice hearing remains intact for a much longer period. DBA/1 mice were given fluoxetine (15–70 mg/kg, i.p.) (Faingold et al. 2008), and significant reductions in respiratory arrest were seen at 45–70 mg/kg, requiring the highest dose for complete suppression. These data indicate that fluoxetine is also effective in blocking respiratory arrest in DBA/1 mice, but higher doses were required than those in the DBA/2

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

671

0

Pre-drug

Day 5 Fluox

Fluoxetine Fluoxetine

*

Saline Saline

Fluoxetine

Saline Saline

Fluoxetine Fluoxetine

50

Saline Saline

% Respiratory arrest

100

Post -drug

Figure 41.7╇ Effect of subchronic administration of a selective serotonin reuptake inhibitor,

fluoxetine (Prozac), given once daily for five consecutive days (20 mg/kg, i.p., per day) to significantly reduce the incidence of susceptibility to respiratory arrest after audiogenic seizures in DBA/1 mice (n = 9). Dark gray bars, vehicle (normal saline); black bars, after 5 days of fluoxetine. Return of respiratory arrest susceptibility began at 24–96 h. n, number of animals. Significantly different from saline control. *p < 0.002 (χ2 test).

mice. The reason for this difference in sensitivity is unknown. Subchronic (5-day) fluoxetine administration (20 mg/kg, i.p., once daily) to DBA/1 mice signiἀcantly reduced respiratory arrest incidence by day 5 (Figure 41.7). The respiratory arrest–suppressive effect was reversible 2–4 days after termination of the chronic treatment, as indicated by the fact that respiratory arrest susceptibility returned (Faingold et al. 2008, 2009). These ἀndings further support the importance of enhancing 5-HT neurotransmission in prevention of respiratory arrest in this SUDEP model. These data indicate that the two strains of DBA mice may be useful animal models of SUDEP and that alterations of 5-HT neurotransmission may be an important target for understanding and treating human SUDEP. However, the cellular and molecular origins of the 5-HT deἀciency in DBA mice are currently unknown and need to be evaluated. Finally, because many patients with epilepsy are being treated for comorbid depression with selective serotonin reuptake inhibitors, a thorough examination of existing case information ἀles and a prospective evaluation of patients undergoing seizures, particularly in epilepsy monitoring units, may provide data on the potential usefulness of agents that enhance the activity of 5-HT neurotransmission in the prevention of SUDEP in patients. Future directions of this research in DBA mice include evaluation of other more selective agents that enhance the action of 5-HT and determination of sites of action of these agents within the brain.

41.12â•…Discussion The results of these studies support the usefulness of DBA mice as models of human SUDEP due to respiratory arrest, which is a major observed cause of human SUDEP. DBA/2 mice are an acute model of SUDEP because their susceptibility to seizure-induced respiratory arrest occurs over a 1-week period, and they lose their susceptibility beginning at ~29 days

672 Sudden Death in Epilepsy: Forensic and Clinical Issues

of age, which is likely due to early hearing loss. Many DBA/1 mice require up to three daily seizures before they become susceptible to seizure-induced respiratory arrest, but these mice exhibit SUDEP susceptibility until at least 5 weeks of age because their hearing remains intact. There is no evidence that human susceptibility to SUDEP occurs only acutely or that it declines with age until much later in life. The cause of SUDEP in epileptic individuals may vary from patient to patient, and cardiac events remain a major potential cause. Agents that enhance the action of 5-HT prevent SUDEP without blocking seizures in both the acute and chronic SUDEP models, and agents that act to prolong 5-HT action by blocking reuptake, the selective serotonin reuptake inhibitors, are quite effective in blocking SUDEP in DBA mice. Future directions in this research include determining if new 5-HT agonists that act selectively on 5-HT receptor subtypes may be more effective or potent in blocking respiratory arrest because they would be expected to exert fewer adverse effects than the selective serotonin reuptake inhibitors. It would be useful to determine where in the brain the serotonergic agents are exerting their anti-SUDEP effects. It would also be useful to know if DBA mice display differences in levels of 5-HT, the enzymes that mediate 5-HT synthesis or metabolism, or altered 5-HT uptake mechanisms. Future work should determine whether differences in protein levels of 5-HT receptor subtypes in DBA/1 mice occur in a fashion similar to those seen in DBA/2 mice and if the difference in the effective doses of fluoxetine in the two SUDEP models are due to altered 5-HT receptors, 5-HT enzymes, or other differences. Many 5-HT reuptake inhibitors are widely used in patients to treat depression, including epileptic patients, and it would be worthwhile to examine if these agents can be used prophylactically to prevent SUDEP, as suggested by a recent retrospective study (Bateman et al. 2010b). The ἀnding that a 5-HT antagonist increased SUDEP susceptibility in DBA/2 mice also suggests that these agents, which are also used clinically for allergies and migraine headaches, should be avoided in patients who display the clinical characteristics associated with higher risk for SUDEP susceptibility.

Acknowledgments We thank Marcus Randall and Mannish Raisinghani for their parts in the experiments from our laboratories; Ronald Browning for critical comments on the manuscript; Diana Smith for her manuscript assistance; and the Citizens United for Research in Epilepsy (CURE) Collaborative Grant and the CURE Chris Donalty Grant award for funding our work on SUDEP.

References Agarwal, S. K., and F. R. Calaresu. 1990. Reciprocal connections between nucleus tractus solitaris and rostral ventrolateral medulla. Brain Research 523: 305–308. Al-Zubaidy, Z. A., R. L. Erickson, and J. J. Greer. 1996. Serotonergic and noradrenergic effects on respiratory neural discharge in the medullary slice preparation of neonatal rats. Pflugers Arch 431: 942–949. Applegate, C. D., and L. H. Tecott. 1998. Global increases in seizure susceptibility in mice lacking 5-HT2C receptors: A behavioral analysis. Exp Neurol 154: 522–530. Bateman, L. M., C. S. Li, and M. Seyal. 2008. Ictal hypoxemia in localization-related epilepsy: Analysis of incidence, severity and risk factors. Brain 131 (Pt 12): 3239–3245.

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

673

Bateman, L. M., M. Spitz, and M. Seyal. 2010a. Ictal hypoventilation contributes to cardiac arrhythmia and SUDEP: Report on two deaths in video-EEG-monitored patients. Epilepsia 51 (5): 916–920. Bateman, L. M., C. S. Li, T. C. Lin, and M. Seyal. 2010b. Serotonin reuptake inhibitors are associated with reduced severity of ictal hypoxemia in medically refractory partial epilepsy. Epilepsia [Epub ahead of print]. Benarroch, E. E. 2007. Brainstem respiratory control: Substrates of respiratory failure of multiple system atrophy. Mov Disord 22 (2): 155–161. Bodineau, L., F. Cayetanot, D. Marlot, T. Collin, F. Gros, and A. Frugiere. 2004. Endogenous 5-HT(1/2) systems and the newborn rat respiratory control. A comparative in vivo and in vitro study. Respir Physiol Neurobiol 141: 47–57. Bonham, A. C. 1995. Neurotransmitters in the CNS control of breathing. Respir Physiol 101: 219–230. Bonham, A. C., C. Y. Chen, S. Sekizawa, and J. P. Joad. 2006. Plasticity in the nucleus tractus solitarius and its influence on lung and airway reflexes. J Appl Physiol 101: 322–327. Boscan, P., A. E. Pickering, and J. F. Paton. 2002. The nucleus of the solitary tract: An integrating station for nociceptive and cardiorespiratory afferents. Exp Physiol 87: 259–266. Brennan, T. J., W. W. Seeley, M. Kilgard, C. E. Schreiner, and L. H. Tecott. 1997. Sound-induced seizures in serotonin 5-HT2c receptor mutant mice. Nat Genet 16: 387–390. Brisson-Thoby, M., and J. J. Greer. 2008. Anatomical and functional development of the pre-Â�Bötzinger complex in prenatal rodents. J Appl Physiol 104 (4): 1213–1219. Browning, R. A., A. V. Wood, M. A. Merill, J. W. Dailey, and P. C. Jobe. 1997. Enhancement of the anticonvulsant effect of fluoxetine following blockade of 5-HT1A receptors. Eur J Pharmacol 336: 1–6. Carley, D. W., and M. Radulovacki. 1999. Role of peripheral serotonin in the regulation of central sleep apneas in rats. Chest 115: 1397–1401. Chen, C. Y., and A. C. Bonham. 1998. Non-NMDA and NMDA receptors transmit area postrema input to aortic baroreceptor neurons in NTS. Am J Physiol 275: H1695–H1706. Collins, R. L. 1972. Audiogenic seizures. In Experimental Models of Epilepsy—A Manual for the Laboratory Worker, 347–372. New York, NY: Raven Press. Collins, R. L., and J. L. Fuller. 1968. Audiogenic seizure prone (asp): A gene affecting behavior in linkage group 8 of the mouse. Science 162: 1137–1139. Dailey, J. W., and D. K. Naritoku. 1996. Antidepressants and seizures: Clinical anecdotes overshadow neuroscience. Biochem Pharmacol 52: 1323–1329. Darbin, O., D. Casebeer, and D. K. Naritoku. 2003. Effects of seizure severity and seizure repetition on postictal cardiac arrhythmia following maximal electroshock. Exp Neurol 181: 327–331. Davis, G. G., and G. McGwin Jr. 2004. Comparison of heart mass in seizure patients dying of sudden unexplained death in epilepsy to sudden death due to some other cause. Am J Forensic Med Pathol 25: 23–28. Di Giovanni, G., V. Di Matteo, M. Pierucci, A. Benigno, and E. Esposito. 2006. Central serotonin2C receptor: From physiology to pathology. Curr Top Med Chem 6 (18): 1909–1925. Doi, A., and J. M. Ramirez. 2008. Neuromodulation and the orchestration of the respiratory rhythm. Respir Physiol Neurobiol 164 (1–2): 96–104. Donner, E. J., C. R. Smith, and O. C. Snead III. 2001. Sudden unexplained death in children with epilepsy. Neurology 57: 430–434. Dorr, A. E., and G. Debonnel. 2006. Effect of vagus nerve stimulation on serotonergic and noradrenergic transmission. J Pharmacol Exp Ther 318: 890–898. Faingold, C. L. 2002. Role of GABA abnormalities in the inferior colliculus pathophysiology—Â� Audiogenic seizures. Hear Res 168: 223–237. Faingold, C. L., M. Randall, S. Tupal, and V. V. Uteshev. 2008. SUDEP model in DBA mice: Longterm SUDEP susceptibility and modulation by serotonin (5-HT). Epilepsia 49 (10): 3.060. Faingold, C. L., M. Randall, and S. Tupal. 2009. DBA/1 mice are a useful model of SUDEP for evaluating chronic prevention treatments. Epilepsia 50: 3.087.

674 Sudden Death in Epilepsy: Forensic and Clinical Issues Faingold, C. L., M. Randall, and S. Tupal. 2010. DBA/1 mice exhibit chronic susceptibility to audiogenic seizures followed by sudden death associated with respiratory arrest. Epilepsy Behav 17 (4): 436–440. Feldman, J. L., G. S. Mitchell, and E. E. Nattie. 2003. Breathing: Rhythmicity, plasticity, chemosensitivity. Annu Rev Neurosci 26: 239–266. Heisler, L. K., H. M. Chu, and L. H. Tecott. 1998. Epilepsy and obesity in serotonin 5-HT2C receptor mutant mice. Annals of the New York Academy of Sciences 861: 74–78. Hernandez, E. J., P. A. Williams, and F. E. Dudek. 2002. Effects of fluoxetine and TFMPP on spontaneous seizures in rats with pilocarpine-induced epilepsy. Epilepsia 43: 1337–1345. Hodges, M. R., and G. B. Richerson. 2008. Contributions of 5-HT neurons to respiratory control: Neuromodulatory and trophic effects. Respir Physiol Neurobiol 164 (1–2): 222–232. Hodges, M. R., G. J. Tattersall, and M. B. Harris et al. 2008. Defects in breathing and thermoregulation in mice with near-complete absence of central serotonin neurons. J Neurosci 28 (10): 2495–2505. Holladay, J. W., M. J. Dewey, and S. D. Yoo. 1998. Pharmacokinetics and antidepressant activity of fluoxetine in transgenic mice with elevated serum alpha-1-acid glycoprotein levels. Drug Metab Dispos 26 (1): 20–24. Jordan, D. 2005. Vagal control of the heart: Central serotonergic (5-HT) mechanisms. Exp Physiol 90: 175–181. Kinney, H. C., G. B. Richerson, S. M. Dymecki, R. A. Darnall, and E. E. Nattie. 2009. The brainstem and serotonin in the sudden infant death syndrome. Annu Rev Pathol Mech Dis 4: 517–550. Kline, D. D., K. N. Takacs, E. Ficker, and D. L. Kunze. 2002. Dopamine modulates synaptic transmission in the nucleus of the solitary tract. J Neurophysiol 88: 2736–2744. Kubin, L., G. F. Alheid, E. J. Zuperku, and D. R. McCrimmon. 2006. Central pathways of pulmonary and lower airway vagal afferents. J Appl Physiol 101: 618–627. LaFrance, Jr., W. C., A. M. Kanner, and B. Hermann. 2008. Psychiatric comorbidities in epilepsy. Int Rev Neurobiol 83: 347–383. Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Langan, Y., L. Nahef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lear-Kaul, K. C., L. Coughlin, and M. J. Dobersen. 2005. Sudden unexpected death in epilepsy: A retrospective study. Am J Forensic Med Pathol 26: 11–17. Li, Z., K. F. Morris, D. M. Baekey, R. Shannon, and B. G. Lindsey. 1999. Multimodal medullary neurons and correlational linkages of the respiratory network. J Neurophysiol 82: 188–201. Malagie, I., A. C. Trillat, C. Jacquot, and A. M. Gardier. 1995. Effects of acute fluoxetine on extracellular serotonin levels in the raphe: An in vivo microdialysis study. Eur J Pharmacol 286: 213–217. Manzke, T., U. Guenther, and E. G. Ponimaskin et al. 2003. 5-HT4(a) receptors avert upload-induced breathing depression without loss of analgesia. Science 301: 226–229. McGregor, A., and J. Wheless. 2006. Pediatric experience with sudden unexplained death in epilepsy at a tertiary epilepsy center. J Child Neurol 21: 782–787. Merlet, I., K. Ostrowsky, and N. Costes et al. 2004. 5-HT1A receptor binding and intracerebral activity in temporal lobe epilepsy: An [18F]MPPF-PET study. Brain 127: 900–913. Merrill, M. A., R. W. Clough, P. C. Jobe, and R. A. Browning. 2003. Role of the superior colliculus and the intercollicular nucleus in the brainstem seizure circuitry of the genetically epilepsy-prone rat. Epilepsia 44: 305–314. Merrill, M. A., R. W. Clough, P. C. Jobe, and R. A. Browning. 2005. Brainstem seizure severity regulates forebrain seizure expression in the audiogenic kindling model. Epilepsia 46: 1380–1388.

DBA Mice as Models of Sudden Unexpected Death in Epilepsy

675

Meschaks, A., P. Lindstrom, C. Halldin, L. Farde, and I. Savic. 2005. Regional reductions in serotonin 1A receptor binding in juvenile myoclonic epilepsy. Arch Neurol 62: 946–950. Meyer, L. C., A. Fuller, and D. Mitchell. 2006. Zacopride and 8-OH-DPAT reverse opioid-induced respiratory depression and hypoxia but not catatonic immobilization in goats. Am J Physiol Regul Integr Comp Physiol 290: R405–R413. Moon, R. Y., R. S. Horne, and F. R. Hauck. 2007. Sudden infant death syndrome. Lancet 370 (9598): 1578–1587. Nashef, L., D. R. Fish, S. Garner, J. W. Sander, and S. D. Shorvon. 1995. Sudden death in epilepsy: A study of incidence in a young cohort with epilepsy and learning difficulty. Epilepsia 36: 1187–1194. Nashef, L., F. Walker, P. Allen, J. W. A. S. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60: 297–300. Nemeroff, C. B. 2007. The burden of severe depression: A review of diagnostic challenges and treatment alternatives. J Psychiatr Res 41 (3–4): 189–206. Neumann, P. E., and R. L. Collins. 1991. Genetic dissection of susceptibility to audiogenic seizures in inbred mice. Proc Natl Acad Sci U S A 88: 5408–5412. Neumann, P. E., and R. L. Collins. 1992. Conἀrmation of the influence of a chromosome 7 locus on susceptibility to audiogenic seizures. Mamm Genome 3: 250–253. Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42: 667–673. Nishino, S., and E. Mignot. 1997. Pharmacological aspects of human and canine narcolepsy. Prog Neurobiol 52: 27–78. Olsson, M., K. Annerbrink, J. Hedner, and E. Eriksson. 2004. Intracerebroventricular administration of the angiotensin II receptor antagonist saralasin reduces respiratory rate and tidal volume variability in freely moving Wistar rats. Psychoneuroendocrinology 29: 107–112. Opeskin, K., and S. F. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12: 456–464. Otake, K., D. J. Reis, and D. A. Ruggiero. 1994. Afferents to the midline thalamus issue collaterals to the nucleus tractus solitarii: An anatomical basis for thalamic and visceral reflex integration. J Neurosci 14: 5694–5707. Pena, F., and J. M. Ramirez. 2002. Endogenous activation of serotonin-2A receptors is required for respiratory rhythm generation in vitro. J Neurosci 22: 11055–11064. Potts, J. T. 2002. Neural circuits controlling cardiorespiratory responses: Baroreceptor and somatic afferents in the nucleus tractus solitarius. Clin Exp Pharmacol Physiol 29: 103–111. Potts, J. T. 2006. Inhibitory neurotransmission in the nucleus tractus solitaril: Implications for baroreflex resetting during exercise. Clin Exp Physiol 91: 59–72. Ramirez, J. M., and J. C. Viemari. 2005. Determinants of inspiratory activity. Respir Physiol Neurobiol 147: 145–157. Racagni, G., and M. Popoli. 2008. Cellular and molecular mechanisms in the long-term action of antidepressants. Dialogues Clin Neurosci 10 (4): 385–400. Raul, L. 2003. Serotonin2 receptors in the nucleus tractus solitarius: Characterization and role in the baroreceptor reflex arc. Cell Mol Neurobiol 23: 709–726. Richerson, G. B., and P. A. Getting. 1992. Medullary respiratory neurons in the guinea pig: Localization and ἀring patterns. Brain Res 591: 79–87. Richter, D. W., T. Manzke, B. Wilken, and E. Ponimaskin. 2003. Serotonin receptors: Guardians of stable breathing. Trends Mol Med 9: 542–548. Ryvlin, P., A. Montavont, and P. Kahane. 2006. Sudden unexpected death in epilepsy: From mechanisms to prevention. Curr Opin Neurol 19: 194–199. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3: 55–62.

676 Sudden Death in Epilepsy: Forensic and Clinical Issues Seyal, M., L. M. Bateman, T. E. Albertson, T. C. Lin, and C. S. Li. 2010. Respiratory changes with seizures in localization-related epilepsy: Analysis of periictal hypercapnia and airflow patterns. Epilepsia [Epub ahead of print]. Seyfried, T. N., and G. H. Glaser. 1985. A review of mouse mutants as genetic models of epilepsy. Epilepsia 26: 143–150. So, E. L. 2006. Demystifying sudden unexplained death in epilepsy—Are we close? Epilepsia 47 (Suppl 1): 87–92. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (Suppl 9): 93–98. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41: 1494–1497. Specchio, L. M., A. Iudice, and N. Specchio et al. 2004. Citalopram as treatment of depression in patients with epilepsy. Clin Neuropharmacol 27: 133–136. St-John, W. M., A. H. Rudkin, G. L. Homes, and J. C. Leiter. 2006. Changes in respiratory-modulated neural activities, consistent with obstructive and central apnea, during ἀctive seizures in an in situ anaesthetized rat preparation. Epilepsy Res 70: 218–228. Terreberry, R. R., and E. J. Neafsey. 1987. The rat medial frontal cortex projects directly to autonomic regions of the brainstem. Brain Res Bull 19: 639–651. Theodore, W. H. 2005. Brain stimulation for epilepsy. Nat Clin Pract Neurol 1: 64–65. Toczek, M. T., R. E. Carson, L. Lang et al. 2003. PET imaging of 5-HT1A receptor binding in patients with temporal lobe epilepsy. Neurology 60: 749–756. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7 (11): 1021–1031. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47: 21–26. Turner, J. G., and J. F. Willott. 1998. Exposure to an augmented acoustic environment alters auditory function in hearing-impaired DBA/2J mice. Hearing Research 118: 101–113. Uteshev, V. V., S. Tupal, Y. Mhaskar, and C. L. Faingold. 2010. Abnormal serotonin receptor expression in DBA/2 mice associated with susceptibility to sudden death due to respiratory arrest. Epilepsy Res 88 (2–3): 183–188. Venit, E. L., B. D. Shepard, and T. N. Seyfried. 2004. Oxygenation prevents sudden death in seizureprone mice. Epilepsia 45: 993–996. Welsh, J. P., D. G. Placantonakis, S. I. Warsetsky, R. G. Marquez, L. Bernstein, and S. A. Aicher. 2002. The serotonin hypothesis of myoclonus from the perspective of neuronal rhythmicity. Adv Neurol 89: 307–329. Willott, J. F., and K. R. Henry. 1976. Roles of anoxia and noise-induced hearing loss in the postictal refractory period for audiogenic seizures in mice. J Comp Physiol Psychol 90: 373–381. Wong-Riley, M. T., and Q. Liu. 2005. Neurochemical development of brain stem nuclei involved in the control of respiration. Respir Physiol Neurobiol 149: 83–98.

Clinical Issues of Sudden Death

III

Cardiac and Pulmonary Risk Factors and Pathomechanisms of Sudden Unexplained Death in Epilepsy Patients

42

Josef Finsterer Claudia Stöllberger

Contents 42.1 Introduction 42.2 Cardiac Risk Factors of SUDEP 42.2.1 Channelopathies 42.2.2 Cardiotoxic Drugs 42.2.3 Other Possible Cardiac Risk Factors 42.3 Cardiac Pathomechanisms 42.3.1 Ictal Arrhythmias 42.3.1.1 Bradyarrhythmias and Asystole 42.3.1.2 Tachyarrhythmias 42.3.1.3 Autonomic Mechanisms 42.3.2 Systolic Dysfunction 42.4 Pulmonary Risk Factors of SUDEP 42.5 Respiratory Pathomechanisms 42.6 Clinical Implications 42.6.1 Interictal 42.6.2 Ictal 42.6.3 Postictal 42.7 Conclusion References

679 680 680 681 682 683 684 684 685 685 686 686 687 687 687 688 688 688 689

42.1â•…Introduction In addition to general and neurological risk factors, there is increasing evidence that cardiac (Aurlien et al. 2009; Kerling et al. 2009; Pezzella et al. 2009; Strzelczyk et al. 2008) and pulmonary comorbidities are predisposing factors (Scorza et al. 2007; Tavee and Morris 2008) for sudden unexplained death in epilepsy (SUDEP). As with general risk factors (young age, male sex, Dravet syndrome, being in bed, cold temperature) and neurological risk factors [early-onset epilepsy (long duration of epilepsy), severity of epilepsy (presence of tonic–clonic seizures), poor seizure control, high seizure frequency, increased number of antiepileptic drugs, subtherapeutic levels of antiepileptic drugs, frequent modiἀcations of antiepileptics, noncompliance with antiepileptics, presence of aberrant neurogenesis], the 679

680 Sudden Death in Epilepsy: Forensic and Clinical Issues

relevance of cardiac and pulmonary risk factors has not been investigated by prospective observational follow-up studies or by intervention studies, and the evidence for speciἀc risk factors and pathomechanisms is still not established (Johnston and Smith 2007). The goal of this chapter is to highlight and discuss recent ἀndings and practical implications concerning potential cardiac and pulmonary risk factors and pathomechanisms of SUDEP.

42.2â•… Cardiac Risk Factors of SUDEP Although SUDEP is deἀned as sudden, unexpected, witnessed or unwitnessed, nondrowning death of epilepsy patients with or without evidence of a seizure and without structural or toxicological abnormality of the heart or lungs (Tomson et al. 2008), there may be premorbid conditions on the molecular or ultrastructural level that predispose to the occurrence of SUDEP. This is particularly relevant in explaining the phenomenon of most epilepsy patients surviving even severe seizures, whereas some die without overt premorbid conditions. Possible cardiac risk factors, which potentially predispose to the occurrence of SUDEP, are mutations in genes that encode for ion channels, which have not manifested clinically (channelopathies) (Hughes 2009); cardiotoxic drugs, including some antiepileptic drugs; and additional risk factors (Table 42.1). However, no general consensus has yet been reached and some authors even believe that there are no deἀnable risk factors for SUDEP at all (Vlooswijk et al. 2007). 42.2.1â•… Channelopathies Channelopathies are believed to be responsible for a number of diseases, including periodic paralysis, epilepsy, long-QT syndrome, and Brugada syndrome. Recently, some evidence has been provided that channelopathies may also play a pathogenetic role in SUDEP (Aurlien et al. 2009; Hindocha et al. 2008). In a family with autosomal dominant Table 42.1â•… Potential Cardiac and Pulmonary Risk Factors for SUDEP Possible Cardiac Risk Factors Channelopathies â•… SCNA1 mutation â•… SCNA5 mutation

(Hindocha et al. 2008; Nabbout 2008) (Aurlien et al. 2009)

Cardiotoxic Drugs â•… Carbamazepine â•… Lamotrigine

(Timmings 1998) (Aurlien et al. 2009)

Other Possible Risk Factors â•… Short/prolonged QTc Outside temperature below zero â•…Myocardial ἀbrosis from myocardial ischemia during seizures â•… Hyponatremia â•… Aberrant neurogenesis Lung â•… Latent muscular respiratory insufficiency

(The et al. 2007) (Colugnati et al. 2008; Sonoda et al. 2008) (P-Codrea Tigaran et al. 2005) (Ruis Ginez et al. 2007) (Scorza et al. 2008) (Personal communication)

Cardiac and Pulmonary Risk Factors and Pathomechanisms of SUDEP

681

generalized epilepsy with febrile seizures, the causative mutation was found in the SCN1A gene at 5600T>C (Hindocha et al. 2008). The mutation resulted in the amino acid exchange I1867T within the intracellular C-terminal domain. Two of the family members, a 23-yearold woman and a 19-year-old woman, had died of SUDEP. The 23-year-old woman had afebrile, generalized tonic–clonic seizures throughout her life and the 19-year-old woman had febrile, generalized tonic–clonic seizures, absences, and focal-onset seizures since she was 14 months old and afebrile generalized tonic–clonic seizures since she was 2 years old. Arguments for a causative role of the mutation in these two patients are that two SUDEP cases occurred within the same family and that the SCN1A gene product was also detected in the sinoatrial node of mice (Hindocha et al. 2008). The latter ἀnding suggests that triggering of arrhythmias by a seizure may be facilitated by the presence of ion channel mutations within the cardiac conduction system. The mutation may also affect brainstem control of respiration or other autonomic functions, as in patients with severe myoclonic epilepsy at infancy, which is also caused by SCN1A mutations (Kimura et al. 2005). In a 25-yearold patient with SUDEP, postmortem DNA analysis revealed a novel missense mutation in the cardiac voltage-gated sodium-channel alpha-subunit gene SCN5A (Aurlien et al. 2009). Because epilepsy in this patient was treated with lamotrigine, it could not be decided whether SUDEP was causally related to the SCN5A mutation or to the side effects of lamotrigine, which is known to inhibit cardiac potassium channels (Aurlien et al. 2009). 42.2.2â•… Cardiotoxic Drugs All drugs known to have a cardiotoxic effect may be potentially dangerous for patients at risk for SUDEP. Most investigations, however, were directed toward the arrhythmogenic side effects of antiepileptic drugs (Table 42.2). Unfortunately, these studies only investigated the side effects of antiepileptic drugs but not of other concomitant cardiotoxic Table 42.2â•… Antiepileptic Drugs and Their Potential Cardiotoxicity Phenobarbital Phenytoin Carbamazepine Oxcarbazepine Valproic acid Lorazepam Clobazam Clonazepam Felbamate Gabapentin Tiagabine Lamotrigine Topiramate Vigabatrin Levetiracetam Pregabalin Zonisamide Lacosamide

Arrhythmias

Negative Inotropy

No Yes Yes No No Yes Yes No No Yes No Yes Yes No No Yes No No

No Yes Yes No Yes No No No No No No No No No No Yes No No

682 Sudden Death in Epilepsy: Forensic and Clinical Issues

drugs, which may also have arrhythmogenic side effects, such as antibiotics or neuroleptics (Stöllberger and Finsterer 2004). Antibiotics, in which an arrhythmogenic effect has been reported, include penicillin (AV blocks I and II), ampicillin (QT prolongation), cefotaxime (arrhythmia), ciprofloxazine (QT prolongation), moxifloxacine (sinus tachycardia), sulfametoxazole (QT prolongation), erythromycine (prolonged depolarization, QT prolongation, torsades des pointes), and doxycycline (supraventricular tachycardia, sporadic Wenkebach block). Concerning the question of whether or not antiepileptic drugs€predisÂ� pose to SUDEP, the results are conflicting (Kloster and Engelskjøn 1999; Langan et al. 2005; Lathers and Schraeder 2002; Nilsson et al. 2001; Walczak 2003). However, it is generally accepted in most to studies that the risk of SUDEP increases with the increased number of antiepileptic drugs being taken by the patient (Nilsson et al. 1999; Walczak et al. 2001). Most data on cardiotoxicity of antiepileptic drugs are available for carbamazepine. Carbamazepine has been shown to slow atrioventricular conduction, to increase the sympathetic tone, and to suppress parasympathetic functions (Isojärvi et al. 1998). In addition, abrupt withdrawal of carbamazepine resulted in enhanced sympathetic activity in sleep, as evidenced by reduced heart rate variability (Hennessy et al. 2001). This effect was explained by a rebound to the withdrawal of the sedating, relaxing, and antidepressive effects of carbamazepine. Heart rate variability was also generally reduced during sleep in patients taking carbamazepine in another study (Persson et al. 2007) and increased when carbamaÂ� zepine was withdrawn (Lossius et al. 2007). In a series of 14 SUDEP patients, 85% had been treated with carbamazepine compared to 38% of the non-SUDEP cases (Timmings 1998). In three larger controlled studies, however, carbamazepine was used with equal frequency in SUDEP patients and controls (Walczak et al. 2001; Kloster and Engelskjøn 1999; Nilsson et al. 1999). A further study found frequent changes of carbamazepine dosages with concentrations above the upper therapeutic range to be a risk factor for SUDEP (Nilsson et al. 2001). Oral phenytoin rarely induced AV block (Stöllberger and Finsterer 2004), whereas phenobarbital and valproic acid were not arrhythmogenic (Walczak 2003) (Table 42.2). In studies associated with the development of new antiepileptic drugs and reviewing data on the risk of SUDEP, lamotrigine, gabapentine, tiagabine, topiramate, and zonisamide were not associated with an increased risk of SUDEP in the young epilepsy population (Leestma et al. 1997; Walczak 2003). In addition, in a study of 4700 patients on lamotrigine, 45 deaths occurred, of which 18 were classiἀed as SUDEP. The rate of SUDEP was not increased compared to young adults with severe epilepsy and taking antiepileptic drugs other than lamotrigine (Leestma et al. 1997). Recently, however, it has been reported that SUDEP occurred in four females with idiopathic epilepsy who were under a monotherapy with lamotrigine (Aurlien et al. 2007). Lamotrigine was suspected of being a risk factor because it inhibits the cardiac rapid delayed rectiἀer potassium ion current (Ikr), which has been shown to increase the risk of cardiac arrhythmias and sudden cardiac death (Aurlien et al. 2007). Arrhythmogenic side effects of other antiepileptic drugs have been studied only in small series. 42.2.3â•… Other Possible Cardiac Risk Factors Although every cardiac disease has the potential to be a risk factor of SUDEP, only a few were thought to increase the risk of epilepsy patients dying from SUDEP. One of these presumed risk factors is cold (outside temperatures below 0°C), which has been identiἀed as a risk factor for sudden cardiac death (Colugnati et al. 2008). In rats with epilepsy,

Cardiac and Pulmonary Risk Factors and Pathomechanisms of SUDEP

683

exposure to cold increased the heart rate, causing the authors to suggest that cold may be a cardiovascular risk factor for SUDEP (Sonoda et al. 2008). Another potential risk factor could be a short QTc interval. In a recent cohort study, it has been shown that QTc is shortened in patients with epilepsy, particularly in patients with cryptogenic epilepsy (The et al. 2007). Whether shortened QTc interval is indeed a risk factor for SUDEP, however, remains speculative. There is a lack of data indicating that the cardiac conduction system is affected in SUDEP patients. There was no morphological difference of the cardiac conduction system in an autopsy study of 10 SUDEP patients and 10 matched controls (Opeskin et al. 2000). However, in another autopsy study on 15 SUDEP patients, 6 showed ἀbrotic changes in the subendocardial and deep myocardium (P-Codrea Tigaran et al. 2005), which were attributed to epilepsy. Some authors also consider electrolyte imbalances, particularly hyponatremia, known to trigger arrhythmias, as a risk factor of SUDEP (Ruis Ginez et al. 2007). So far, there is no indication that arterial hypertension has any influence on the outcome of patients with uncontrolled epilepsy. Whether previous Takotsubo syndrome, also known as reversible left ventricular apical ballooning, is a potential risk factor of SUDEP is so far unknown because this question has not been addressed in an appropriate study, most likely because of the low number of patients with this abnormality. Takotsubo syndrome, also known as reversible left ventricular apical ballooning, is a stress-induced myocardial contraction failure due to supposed nonresponsiveness to adrenergic stimulation (Dorfman and Iskandrian 2009).

42.3â•… Cardiac Pathomechanisms Although a number of potential pathomechanism for SUDEP are discussed (Table 42.3), no ἀnal decision has been made as to whether a single or several pathomechanism(s) lead(s) Table 42.3â•… Potential Cardiac or Pulmonary Pathomechanisms of SUDEP Heart (Hitiris et al. 2007; Scorza et al. 2008) (Almansori et al. 2006; Devinsky et al. 1997; Leung et€al. 2006) High-grade AV block (Altenmüller et al. 2004) Ictal asystole (Carinci et al. 2007; Devinsky et al. 1997; Leung et al. 2006; Schuele et al. 2007; So and Sperling 2007) Tachycardia (Langan et al. 2000; Zaidi et al. 1997) Transmission of epileptic activity to the heart via the (Scorza et al. 2008) autonomic system Impaired sympathetic cardiac innervation (Kerling et al. 2009)

Arrhythmias during or between seizures Bradycardia

Postictal central apnea Postictal obstructive apnea Acute neurogenic pulmonary edema Postictal laryngospasm Ictal hypoxemia, hypercapnia

Lung (Jehi and Najm 2008; Nashef et al. 1996; Ryvlin et al. 2009) (Ryvlin et al. 2009) (Jehi and Najm 2008; Terrence et al. 1981) (Tavee and Morris 2008) (Bateman et al. 2008)

684 Sudden Death in Epilepsy: Forensic and Clinical Issues

to SUDEP (Bell and Sander 2006). In general, potential pathomechanism of SUDEP may be classiἀed as those in which a preexisting, subclinical pathology becomes symptomatic during the seizure and those in which the seizure induces a pathologic reaction in an otherwise healthy individual. Actually, the most widely discussed pathomechanism assumes preexisting cardiopulmonary abnormalities to manifest during tonic–clonic seizures. The most frequently suggested cardiac pathomechanism of SUDEP is induction of asystole, ventricular tachycardia, or ventricular ἀbrillation. There is no deἀnitive proof that cardiac arrhythmias during or between seizures or transmission of epileptic activity to the heart via the central and peripheral autonomic nervous system plays a potential pathogenetic role in SUDEP, but animal studies suggest autonomic dysfunction during seizures to contribute to the fatal outcome of such events (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Scorza et al. 2008). Other authors propose postictal central or obstructive apnea (respiratory arrest) to represent the most likely pathomechanism of SUDEP (Ryvlin et al. 2009). 42.3.1â•… Ictal Arrhythmias Epileptic seizures are potentially arrhythmogenic (So et al. 2000). In a previous study of 43 patients with intractable epilepsy and a total of 51 seizures, seizures were associated with ECG abnormalities such as asystole (n = 1), atrial ἀbrillation (n = 1), marked or moderate sinus arrhythmia (n = 6), supraventricular tachycardia (n = 1), atrial premature depolarization (n = 8), ventricular premature depolarization (n = 2), or bundle-branch block (n = 3) (Nei et al. 2000). ECG abnormalities in patients with intractable epilepsy were also reported by others (Zijlmans et al. 2002). In a study on 11 SUDEP patients and 11 matched controls with epilepsy, the QTc interval increased ictally in both groups (Tavernor et al. 1996). In rats, repeated electrical stimulation of the insula induced progressive heart blocks, resulting in escape rhythms, premature ventricular contraction, or death (Oppenheimer et al. 1991). In a study on 21 SUDEP patients, maximal ictal heart rate was higher than in controls and ictal ECG abnormalities occurred in 56% of the patients: atrial ἀbrillation (n = 2), premature ventricular depolarization (n = 2), sinus arrhythmia (n = 2), atrial premature depolarization (n = 2), junctional escape rhythm (n = 1), and ST elevation (n = 1) (Nei et al. 2004). The authors supposed these ECG abnormalities to be contributory to the occurrence of SUDEP. 42.3.1.1â•… Bradyarrhythmias and Asystole Bradyarrhythmias have been observed repeatedly during seizures and may be explained by the influence of the central autonomic network on cardiac impulse generation, although localization and lateralization issues need to be considered in the light of patterns of seizure spread, hand dominance, or presence of lesions (Leung et al. 2006). Epileptic seizures may particularly induce bradycardia and asystole (Carinci et al. 2007). During a study using loop recording over 2 years, 21% of the patients with intractable epilepsy had at least one ictal bradycardia (Rugg-Gunn et al. 2004). Altogether, 2% of the seizures were associated with bradycardia. Two thirds of the patients with ictal bradycardia had temporal lobe epilepsy and the remaining had frontal lobe epilepsy (Tinuper et al. 2001). Although the literature about asystole in epilepsy is replete with single case reports and small case series (Strzelczyk et al. 2008), there is a consensus that ictal asystole is a rare event and is reported to occur in only 0.27% of the epilepsy patients experiencing a seizure

Cardiac and Pulmonary Risk Factors and Pathomechanisms of SUDEP

685

under video-EEG monitoring (Schuele et al. 2007). In another video-EEG study, 0.4% of the investigated patients had ictal bradycardia or asystole (Rocamora et al. 2003). Asystole during or after seizures most frequently occurs in patients with temporal lobe epilepsy, rarely in patients with extratemporal lobe epilepsy, and not in patients with generalized epilepsy (Schuele et al. 2007). Asystole may be preceded by a loss of muscle tone (Schuele et al. 2007; So and Sperling 2007). Seizures and cardiac syncope are often clinically indistinguishable. Cerebral hypoxemia due to asystole or ventricular ἀbrillation may induce clinical features imitating partial or secondarily generalized tonic–clonic seizures, and thus may be misinterpreted as epilepsy. Because such cardiac abnormalities are potentially fatal, and because a cardiac syncope may be misdiagnosed as epilepsy, the frequency of SUDEP may be difficult to ascertain (Venkataraman et al. 2001). 42.3.1.2â•… Tachyarrhythmias Generally, tachyarrhythmias may be induced by ischemia, metabolic changes, electrolyte disturbances, myocardial ἀbrosis, or abnormalities of the cardiac conduction system. In a loop-recorder study on 20 patients with refractory epilepsy, 16 had ictal tachycardia (RuggGunn et al. 2004). In addition, patients with epilepsy have been shown to develop torsades de pointes from long-QT syndrome, atrioventricular nodal reentry tachycardia, or sinus arrest (Langan et al. 2000; Zaidi et al. 1997; Venkataraman et al. 2001). In another study, it turned out that 50% of seizures resulted in tachycardia within the ἀrst 10 s of EEG recordings. Overall, there are conflicting results about whether bradyarrhythmias, including asystole (Nabbout 2008; Rugg-Gunn et al. 2004), or tachyarrhythmias (Nashef et al. 2007) contribute more frequently to the occurrence of SUDEP. 42.3.1.3â•… Autonomic Mechanisms There are some indications that epileptogenic activity may induce impaired cardiac impulse generation or propagation abnormalities, being explained by the fact that cardiovascular responses, including heart rate variability, have a cortical representation in the insular cortex, cingulate gyrus, and prefrontal cortex (Jehi and Najm 2008; Leung et al. 2006). At the subcortical level, the hypothalamus manages autonomic function to maintain homeostasis. Cortical and subcortical autonomic centers are linked by the amygdala, an integral component of the limbic system, which mediates the autonomic response to emotions (Jehi and Najm 2008). Interestingly, the amygdala, gyrus cinguli, insular cortex, and frontopolar and frontoorbital regions are the most frequent foci of partial epileptic activity, explaining why autonomic disturbances are frequently associated with the occurrence of supraventricular tachycardia, sinus tachycardia, sinus bradycardia, sinus arrest, atrioventricular block, or asystole (Devinsky et al. 1997). Direct transmission of epileptogenic discharges to the central or peripheral sympathetic cardiovascular system has been conἀrmed in various animal studies (Lathers et al. 1987; Lathers and Schraeder 1982, 1987; Schraeder and Lathers 1983). Additional arguments for involvement of the autonomic nervous system in the pathogenesis of SUDEP have been raised. After successful temporal lobe epilepsy surgery, the risk of sympathetically mediated tachyarrhythmias was lessened (Hilz et al. 2002). Physiologic heart rate and blood pressure modulation, mediated by both the parasympathetic and sympathetic nervous systems, were reduced in patients with epilepsy (Isojärvi et al. 1998). Patients with temporal lobe epilepsy also had reduced low-frequency power

686 Sudden Death in Epilepsy: Forensic and Clinical Issues

(low-frequency power represents sympathetic activity and high-frequency power represents parasympathetic activity) on spectral analysis of 24 h-ECG (Tomson et al. 1998). Other studies found both power of low and high frequencies to be reduced in patients with refractory epilepsy (Ansakorpi et al. 2002). Heart rate increased more during sleep seizures than during wake seizures (Nei et al. 2004), possibly reflecting the common occurrence of SUDEP at night (Langan et al. 2005). Overall, there is increased acknowledgment of the role of the autonomic nervous system for “epileptogenic” arrhythmias in SUDEP. 42.3.2â•… Systolic Dysfunction Epileptic activity may not only induce arrhythmia but may also lead to systolic dysfunction. Autopsy of SUDEP patients has shown that hearts are heavier and more dilated compared to controls (Leestma et al. 1997, 1989) and that pulmonary edema is present in 50–86% of the SUDEP cases (Kloster and Engelskjøn 1999; Leestma et al. 1997, 1989). In a sheep model of status epilepticus, animals that died suddenly were found to have developed pulmonary hypertension, an enlarged left atrium, and pulmonary edema. Autonomic dysfunction with altered pulmonary vascular tone and/or cardiac dysfunction is thought to be responsible for pulmonary congestion and edema in SUDEP patients (Pezzella et al. 2009). Whether the pathomechanism of left ventricular dysfunction in epilepsy is the same as that in subarachnoid hemorrhage (Deibert et al. 2003) is unknown. Possibly, stress from seizures or subarachnoid hemorrhage triggers a Takotsubo syndrome, which is characterized not only by systolic dysfunction but also by severe arrhythmias (Stöllberger et al. 2005). Although Takotsubo syndrome has not been described thus far in SUDEP, it has been repeatedly described in patients with epilepsy (Legriel et al. 2008; Stöllberger et al. 2009).

42.4â•… Pulmonary Risk Factors of SUDEP So far, there is minimal evidence that any primary pulmonary disease could be a deἀnite risk factor for SUDEP. For example, it is unknown if patients with respiratory chain defects, which frequently present with muscle disease and epilepsy and may develop muscular respiratory insufficiency, have an increased risk not to survive a tonic–clonic seizure. These patients appear particularly endangered because they often have epilepsy and because their epilepsy is often difficult to treat. Despite this uncertainty about pulmonary risk factors, there are frequent reports about patients who develop severe pulmonary problems during or after seizures, such as ictal hypoxemia or hypercapnia (Bateman et al. 2008), apnea (Bell and Sander 2006; Jehi and Najm 2008; Ryvlin et al. 2009), acute neurogenic pulmonary edema (Jehi and Najm 2008), or postictal laryngospasm (Tavee and Morris 2008) (Table 42.1). In a prospective autopsy series on 52 SUDEP patients, 80% had pulmonary congestion and edema (Leestma et al. 1989). It is uncertain if these abnormalities are due to primary pulmonary, cardiac, or laryngeal mechanisms. It is unclear if only patients with previous lung disease, as opposed to previously healthy subjects, develop such problems. There is also little information available about the effects of a tonic–clonic seizure on the respiratory system in general. Do tonic–clonic seizures generally induce bronchospasm, or loss of tone of the muscles involved in respiration? Recent investigations have shown that at least the vital capacity, forced vital capacity, and forced expiratory volume within the

Cardiac and Pulmonary Risk Factors and Pathomechanisms of SUDEP

687

ἀrst second, are not signiἀcantly different between healthy subjects and epilepsy patients (Scorza et al. 2007).

42.5â•…Respiratory Pathomechanisms Suspected pulmonary pathomechanisms for SUDEP include ictal or postictal central or obstructive apnea (Table 42.3) (Pezzella et al. 2009; Ryvlin et al. 2009). However, seizureinduced apnea and asystole might promote each other through cardiorespiratory reflexes or cerebral hypoxia (Tomson et al. 2008). Seizure-associated respiratory compromise, due to hypoventilation or obstructive apnea, is a prominent feature in witnessed SUDEP cases (Langan et al. 2000; Nashef et al. 1998; So et al. 2000). Polysomnography demonstrated central as well as obstructive ictal apnea to be sometimes associated with bradycardia (Nashef et al. 1996). Ictal apnea was also reported in 60% of patients undergoing video-EEG monitoring (Nashef et al. 1996). In other studies, however, postictal pulmonary edema is a rare ἀnding in SUDEP patients (Swallow et al. 2002). Postictal apnea was also reported in a number of animal studies (So 2008). In one experimental animal study, one-third of the animals died of hypoventilation and showed pulmonary edema at autopsy (Johnston et al. 1995). A role for mechanical obstruction in SUDEP is substantiated by a study that reported SUDEP cases to have been found more often in the prone rather than in the supine position (Kloster and Engelskjøn 1999). The prone position may cause obstruction of the nose and mouth by exerting pressure against the bed clothing. In addition, the prone position may impede ventilation by decreasing vital capacity and tidal volume, particularly after exercise (Röggla and Röggla 1999). Other studies, however, found prone position only to be a weak risk factor for SUDEP (Monté et al. 2007). Severe postictal laryngospasm was reported in a single patient during video-EEG monitoring and assumed to be causal for SUDEP (Tavee and Morris 2008). SUDEP resulting from apnea could result from epileptogenic activity transmitted to the medullar respiratory centers resulting in suppression of brainstem respiratory activity (So 2008). Apnea could also result from cardiorespiratory reflexes triggered by the seizure (Nashef et al. 1996; So 2008). These reflexes are more intense in the young population compared to the elderly, possibly reflecting the higher incidence of SUDEP in young adults (So 2008). In addition, hypotonia of the respiratory muscles, laryngospasm, or bronchospasm may be further involved in seizure-related apnea. Overall, it is unknown whether patients who develop apnea during or after seizures have a higher risk of SUDEP than patients without apnea. Furthermore, no information is available about whether apnea during seizures occurs more often in patients with preexisting pulmonary diseases, such as chronic obstructive pulmonary disease, emphysema, bronchial asthma, in smokers or nonsmokers, or in patients without pulmonary disease.

42.6â•… Clinical Implications 42.6.1â•… Interictal As a consequence of the data so far available about cardiac and pulmonary risk factors and potential pathomechanisms of SUDEP, we propose that patients at risk for SUDEP

688 Sudden Death in Epilepsy: Forensic and Clinical Issues

should undergo comprehensive cardiac and pulmonary investigations (Finsterer and Stöllberger 2009). These evaluations should not only include individual and family history and physical examination for cardiac and pulmonary disease and basic cardiac and pulmonary instrumental investigations but also echocardiography, stress test, 24-h ECG or loop recording, lung function testing, otorhinolaryngological investigations, and polysomnography (Finsterer and Stöllberger 2009). Antiepileptic drugs or other drugs with known or presumed cardiac or pulmonary toxicity should, if possible and arguable, be replaced by nontoxic medication. In addition, interactions between antiepileptic drugs and drugs given for comorbidities should be considered. The family history should be directed toward eliciting histories of syncope, sudden cardiac death, or hereditary arrhythmias, such as Wolf–Parkinson–White syndrome, AV blocks, Brugada syndrome, hereditary atrial ἀbrillation, or short- and long-QT syndrome. In cases where these investigations discover premorbid cardiac or pulmonary disease, optimal antiepileptic, cardiac, and pulmonary treatment should be provided to these patients, particularly when new comorbidities appear. 42.6.2â•… Ictal Continuous ECG recording during video-EEG monitoring would be of additional help in clarifying the pathogenesis of SUDEP. Even better would be a continuous ictal and interictal ECG recording by long-term loop recording in patients at risk to develop SUDEP (Rugg-Gunn et al. 2004). Only with such interventions would ECG changes such as bradyarrhythmias, asystole, or ventricular tachycardia actually be documented during seizures. Such observations could indicate the need for the implantation of pacemakers or even implantable cardioverter deἀbrillators. 42.6.3â•… Postictal Patients at risk for SUDEP should speciἀcally undergo cardiac and pulmonary investigations immediately after a seizure, irrespective of its phenotype and clinical outcome. Acute postictal investigations should include immediate physical examination, blood chemical investigations (creatine kinase, electrolytes, troponin, glucose, brain natriuretic peptide), ECG, x-ray of the lung, and echocardiography. To detect Takotsubo syndrome, a postictal ECG should be evaluated for ST- and T-wave changes and the echocardiogram evaluated for systolic dysfunction. If Takotsubo syndrome is detected, the patient should be intensively observed for potentially life-threatening complications of this syndrome (Stöllberger et al. 2005).

42.7â•… Conclusion There is general agreement that more intensive ictal, interictal, and postictal investigations for clinical and subclinical cardiac and pulmonary disease in patients at risk for SUDEP are required. For this purpose, international organizations should provide a common deἀnition of patients at risk for SUDEP and these patients should then be identiἀed and systematically investigated for clinical and subclinical cardiac and pulmonary disease. More comprehensive cardiac and pulmonary investigations during seizures and

Cardiac and Pulmonary Risk Factors and Pathomechanisms of SUDEP

689

postictally, development of more non-cardiotoxic and non-pulmonary-toxic antiepileptic drugs, research on interactions between antiepileptic drugs and drugs taken for comorbidities, and adequate identiἀcation and treatment of cardiac and pulmonary disease and epilepsy may prevent the fatal outcome in patients at risk for SUDEP.

References Almansori, M., M. Ijaz, and S. N. Ahmed. 2006. Cerebral arrhythmia influencing cardiac rhythm: A case of ictal bradycardia. Seizure 15: 459–461. Altenmüller, D. M., M. Zehender, and A. Schulze-Bonhage. 2004. High-grade atrioventricular block triggered by spontaneous and stimulation-induced epileptic activity in the left temporal lobe. Epilepsia 45: 1640–1644. Ansakorpi, H., J. T. Korpelainen, H. V. Huikuri, U. Tolonen, V. V. Myllylä, and J. I. Isojärvi. 2002. Heart rate dynamics in refractory and well controlled temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 72: 26–30. Aurlien, D., T. P. Leren, E. Taubøll, and L. Gjerstad. 2009. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18: 158–160. Aurlien, D., E. Taubøll, and L. Gjerstad. 2007. Lamotrigine in idiopathic epilepsy—Increased risk of cardiac death? Acta Neurol Scan 115: 199–203. Bateman, L. M., C. S. Li, and M. Seyal. 2008. Ictal hypoxemia in localization-related epilepsy: Analysis of incidence, severity and risk factors. Brain 131: 3239–3245. Bell, G. S., and J. W. Sander. 2006. Sudden unexpected death in epilepsy. Risk factors, possible mechanisms and prevention: A reappraisal. Acta Neurol Taiwan 15: 72–83. Carinci, V., G. Barbato, A. Baldrati, and G. Di Pasquale. 2007. Asystole induced by partial seizures: A rare cause of syncope. Pacing Clin Electrophysiol 30: 1416–1419. Colugnati, D. B., R. M. Arida, S. L. Cravo, et al. 2008. Hibernating mammals in sudden cardiac death in epilepsy: What do they tell us? Med Hypotheses 70: 929–932. Deibert, E., B. Barzilai, A. C. Braverman, et al. 2003. Clinical signiἀcance of elevated troponin I levels in patients with nontraumatic subarachnoid hemorrhage. J Neurosur 98: 741–746. Devinsky, O., S. Pacia, and G. Tatambhotla. 1997. Bradycardia and asystole induced by partial seizures: A case report and literature review. Neurology 6: 1712–1714. Dorfman, T. A., and A. E. Iskandrian. 2009. Takotsubo cardiomyopathy: State-of-the-art review. J Nucl Cardiol 16: 122–134. Finsterer, J., and C. Stöllberger. 2009. Cardiopulmonary surveillance to prevent SUDEP. Lancet€Neurol 8: 131–132. Hennessy, M. J., M. G. Tighe, C. D. Binnie, and L. Nashef. 2001. Sudden withdrawal of carbaÂ� mazepine€increases cardiac sympathetic activity in sleep. Neurology 57: 1650–1654. Hilz, M. J., O. Devinsky, W. Doyle, A. Mauerer, and M. Dütsch. 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125: 985–995. Hindocha, N., L. Nashef, F. Elmslie, et al. 2008. Two cases of sudden unexpected death in epilepsy in a GEFS+ family with an SCN1A mutation. Epilepsia 49: 360–365. Hitiris, N., R. Mohanraj, J. Norrie, and M. J. Brodie. 2007. Mortality in epilepsy. Epilepsy Behav 10: 363–376. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14: 280–287. Isojärvi, J. I., H. Ansakorpi, K. Suominen, U. Tolonen, M. Repo, and V. V. Myllyla. 1998. Interictal cardiovascular autonomic responses in patients with epilepsy. Epilepsia 39: 420–426. Jehi, L., and I. M. Najm. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Cleve Clin J Med 75 (s2): S66–S70. Johnston, A., and P. Smith. 2007. Sudden unexpected death in epilepsy. Expert Rev Neurother 7: 1751–1761.

690 Sudden Death in Epilepsy: Forensic and Clinical Issues Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37: 531–537. Kerling, F., M. Dütsch, R. Linke, T. Kuwert, H. Stefan and M. J. Hilz. 2009. Relation between ictal asystole and cardiac sympathetic dysfunction shown by MIBG-SPECT. Acta Neurol Scand 120 (2): 123–129. Kimura, K., T. Sugawara, E. Mazaki-Miyazaki, et al. 2005. A missense mutation in SCN1A in brothers with severe myoclonic epilepsy in infancy (SMEI) inherited from a father with febrile seizures. Brain Dev 27: 424–430. Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67: 439–444. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Langan, Y., L. Nashef, and J. W. A. S. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42: 123–136. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Leestma, J. E., J. F. Annegers, M. J. Brodie, et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26: 195–203. Legriel, S., F. Bruneel, L. Dalle, et al. 2008. Recurrent takotsubo cardiomyopathy triggered by convulsive status epilepticus. Neurocrit Care 9: 118–121. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9: 19–30. Lossius, M. I., J. E. Erikssen, P. Mowinckel, P. Gulbrandsen, and L. Gjerstad. 2007. Changes in autonomic cardiac control in patients with epilepsy after discontinuation of antiepileptic drugs: A randomized controlled withdrawal study. Eur J Neurol 14: 1022–1028. Monté, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16: 1–7. Nabbout, R. 2008. Can SCN1A mutations account for SUDEP? Commentary on Hindocha et al. Epilepsia 49: 367–368. Nashef, L., N. Hindocha, and A. Makoff. 2007. Risk factors in sudden death in epilepsy (SUDEP): The quest for mechanisms. Epilepsia 48: 859–871. Nashef, L., S. Garner, J. W. A. S. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64: 349–352. Nashef, L., F. Walker, P. Allen, J. W. A. S. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60: 297–300. Nei, M., R. T. Ho, B. W. Abou-Khalil, et al. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45: 338–345. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41: 542–548.

Cardiac and Pulmonary Risk Factors and Pathomechanisms of SUDEP

691

Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42: 667–673. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 353: 888–893. Opeskin, K., A. Thomas, and S. F. Berkovic. 2000. Does cardiac conduction pathology contribute to sudden unexpected death in epilepsy? Epilepsy Res 40: 17–24. Oppenheimer, S. M., J. X. Wilson, C. Guiraudon, and D. F. Cechetto. 1991. Insular cortex stimulation produces lethal cardiac arrhythmias: A mechanism of sudden death? Brain Res 550: 115–121. P-Codrea Tigaran, S., S. Dalager-Pedersen, U. Baandrup, M. Dam, and A. Vesterby-Charles. 2005. Sudden unexpected death in epilepsy: Is death by seizures a cardiac disease? Am J Forensic Med Pathol 26: 99–105. Persson, H., E. Kumlien, M. Ericson, and T. Tomson. 2007. Circadian variation in heart-rate variability in localization-related epilepsy. Epilepsia 48: 917–922. Pezzella, M., P. Striano, C. Ciampa, L. Errichiello, P. Penza, and S. Striano. 2009. Severe pulmonary congestion in a near-miss at the ἀrst seizure: Further evidence for respiratory dysfunction in sudden unexpected death in epilepsy. Epilepsy Behav 14: 701–702. Rocamora, R., M. Kurthen, L. Lickfett, J. von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44: 179–185. Röggla, G., and M. Röggla. 1999. Death in a hobble restraint. Can Med Assoc J 161: 21. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364: 2212–2219. Ryvlin, P., T. Tomson, and A. Montavont. 2009. Excess mortality and sudden unexpected death in epilepsy. Presse Med 38 (6): 905–910. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Schuele, S. U., A. C. Bermeo, A. V. Alexopoulos, et al. 2007. Video-electrographic and clinical features in patients with ictal asystole. Neurology 69: 434–441. Scorza, F. A., R. M. Cysneiros, R. M. Arida, et al. 2008. Adult hippocampal neurogenesis and sudden unexpected death in epilepsy: Reality or just an attractive history? Med Hypotheses 71: 914–922. Scorza, F. A., A. M. Abreu, M. Albuquerque, et al. 2007. Quantiἀcation of respiratory parameters in patients with temporal lobe epilepsy. Arq Neuropsiquiatr 65: 450–453. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (s9): 93–98. So, N. K., and M. R. Sperling. 2007. Ictal asystole and SUDEP. Neurology 69: 423–424. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41: 1494–1497. Sonoda, E. Y., D. B. Colugnati, and C. A. Scorza. 2008. Is cold the new hot in sudden unexpected death in epilepsy? Effect of low temperature on heart rate of rats with epilepsy. Arq Neuropsiquiatr 66: 848–852. Stöllberger, C., J. O. Huber, B. Enzelsberger, and J. Finsterer. 2009. Fatal outcome related to postictal Takotsubo-syndrome with left ventricular rupture. Eur J Neurol 16: e116–e117. Stöllberger, C., J. Finsterer, and B. Schneider. 2005. Takotsubo-like left ventricular dysfunction: Clinical presentation, instrumental ἀndings, additional cardiac and non-cardiac diseases and potential pathomechanisms. Minerva Cardioangiol 53: 139–145. Stöllberger, C., and J. Finsterer. 2004. Cardiorespiratory ἀndings in sudden unexplained/unexpected death in epilepsy (SUDEP). Epilepsy Res 59: 51–60. Strzelczyk, A., S. Bauer, S. Knake, W. H. Oertel, H. M. Hamer, and F. Rosenow. 2008. Ictal asystole in€temporal lobe epilepsy before and after pacemaker implantation. Epileptic Disord 10: 39–44.

692 Sudden Death in Epilepsy: Forensic and Clinical Issues Swallow, R. A., C. E. Hillier, and P. E. Smith. 2002. Sudden unexplained death in epilepsy (SUDEP) following previous seizure-related pulmonary oedema: Case report and review of possible preventative treatment. Seizure 11: 446–448. Tavee, J., and H. Morris III. 2008. Severe postictal laryngospasm as a potential mechanism for sudden unexpected death in epilepsy: A near-miss in an EMU. Epilepsia 49: 2113–2117. Tavernor, S. J., S. W. Brown, R. M. Tavernor, and C. Gifford. 1996. Electrocardiograph QT lengthening associated with epileptiform EEG discharges—A role in sudden unexplained death in epilepsy? Seizure 5: 79–83. Terrence, C. F., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann Neurol 9: 458–464. The, H. S., H. L. Tan, C. Y. Loo, and A. A. Raymond. 2007. Short QTc in epilepsy patients without cardiac symptoms. Med J Malaysia 62: 104–108. Timmings, P. L. 1998. Sudden unexpected death in epilepsy: Is carbamazepine implicated? Seizure 7: 289–291. Tinuper, P., F. Bisulli, A. Cerullo, R. Carcangiu, C. Marini, G. Pierangeli, and P. Cortelli. 2001. Ictal bradycardia in partial epileptic seizures: Autonomic investigation in three cases and literature review. Brain 124: 2361–2371. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7: 1021–1031. Tomson, T., A. C. Sköld, P. Holmgen, L. Nilsson, and B. Danielsson. 1998. Postmortem changes in blood concentrations of phenytoin and carbamazepine: An experimental study. Ther Drug Monit 20: 309–312. Venkataraman, V., J. W. Wheless, L. J. Willmore, and H. Motookal. 2001. Idiopathic cardiac asystole presenting as an intractable adult onset partial seizure disorder. Seizure 10: 359–364. Vlooswijk, M. C., H. J. Majoie, M. C. De Krom, I. Y. Tan, and A. P. Aldenkamp. 2007. SUDEP in the Netherlands: A retrospective study in a tertiary referral center. Seizure 16: 153–159. Walczak, T. 2003. Do antiepileptic drugs play a role in sudden unexpected death in eplilepsy? Drug Safety 26: 673–683. Walczak, T. S., I. E. Leppik, M. D’Amelio, et al. 2001. Incidence and risk factors in sudden unexpected death in epilepsy. A prospective cohort study. Neurology 56: 519–525. Zaidi, A., L. Cotter, and A. Fitzpatrick. 1997. Sudden unexplained death (Letter). Lancet 349: 135. Zijlmans, M., D. Flanagan, and J. Gotman. 2002. Heart rate changes and ECG abnormalities during epileptic seizures: Prevalence and deἀnition of an objective clinical sign. Epilepsia 43: 847–854.

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy Can Ambulatory Electrocardiogram-Based Assessment of Autonomic Function and T-Wave Alternans Help to Evaluate Risk?

43

Richard L. Verrier Steven C. Schachter

Contents 43.1 Scope of the Problem and General Features 43.2 Importance of Neurocardiac Interactions in Triggering Life-Threatening Arrhythmias 43.3 Neurocircuitry of Cardiac Rhythm Control 43.4 Autonomic Mechanisms in Arrhythmogenesis 43.4.1 Adrenergic Influences 43.4.2 Alpha-Adrenergic Receptors 43.4.3 Sympathetic–Parasympathetic Interactions 43.4.4 Baroreflexes and Arrhythmias 43.5 Behavioral Stress and Arrhythmias 43.6 Ambulatory Electrocardiogram-Based Tools for Evaluation of Autonomic Function 43.7 Ambulatory Electrocardiogram-Based T-Wave Alternans to Assess Cardiac Electrical Instability and Risk for Sudden Cardiac Death 43.8 Clinical Evidence of Ambulatory Electrocardiogram-Based T-Wave Alternans for Prediction of Sudden Cardiac Death 43.9 Update on Vagus Nerve Stimulation in Epilepsy 43.10 Conclusions and Implications References

693 694 696 697 697 699 699 700 700 701 702 703 705 705 706

43.1â•… Scope of the Problem and General Features Sudden unexpected death in epilepsy (SUDEP) accounts for nearly 17% of premature deaths among individuals with this condition (Lathers et al. 2008). Witnessed cases indicate that death is contemporaneous with seizure activity (Langan et al. 2005; Tomson 693

694 Sudden Death in Epilepsy: Forensic and Clinical Issues

et al. 2005; Kloster and Engelskjøn 1999). Cardiac dysfunction, primarily in the form of enhanced arrhythmia susceptibility, has been implicated as a critical factor (Opherk et al. 2002; Opeskin et al. 2000; Ryvlin et al. 2006). Whereas decreases in heart rate can occur during epileptic seizures, pronounced sinus tachycardia is the most common electrocardiographic abnormality. Other relatively severe cardiac rhythm, conduction, and repolarization abnormalities have been reported in association with seizures, including bradycardia, asystole, bundle-branch block, ST-segment changes indicative of myocardial ischemia, and T-wave inversion (Opherk et al. 2002; Opeskin et al. 2000; Nei et al. 2004; Tigaran et al. 2003; Jallon et al. 2004). When these abnormalities occur, the length of the seizure is typically prolonged (Zijlmans et al. 2002). The occurrence of bradycardia and asystole in some cases (Rugg-Gunn et al. 2004; Rocamora et al. 2003; Ryvlin et al. 2006) may reflect the indirect influences of seizure-related respiratory disturbances and hypoxia (Lathers et al. 2008) with prone position during sleep a possible factor (Kloster and Engelskjøn 1999). Identiἀcation of the influences contributing to enhanced myocardial risk in SUDEP has been elusive, although a number of plausible cardiac pathologic changes have been suggested. Importantly, SUDEP is not signiἀcantly associated with any speciἀc type of seizure (Annegers et al. 1984), but enhanced adrenergic state associated with seizures constitutes a major component of arrhythmia risk. The goals of this chapter are twofold. The ἀrst is to review the basic mechanisms of autonomic control of cardiac electrical function, which could potentially contribute to precipitation of SUDEP. The second is to draw attention to new ambulatory electrocardiogram-based tools for assessing autonomic function with heart rate turbulence and cardiac arrhythmia vulnerability with T-wave alternans.

43.2â•…Importance of Neurocardiac Interactions in Triggering Life-Threatening Arrhythmias A major theme in the ἀeld of sudden cardiac death research is that whereas autonomic factors exert potent influences on the stability of heart rhythm, precipitation of lifethreatening arrhythmias requires the coexistence of a vulnerable coronary vasculature and myocardial substrate (Figure 43.1). In the general population, enhanced vulnerability of the myocardium to arrhythmic death results from the presence of relatively advanced coronary artery disease due to atherosclerosis, which is an underlying condition in ~80% of cases. The mean age of sudden cardiac death victims is 60 years (Goldberger et al. 2008). However, patients with epilepsy who die suddenly are generally younger, in the range of 20 to 40 years of age (Tomson et al. 2005), and have ischemic heart disease in only 19% of cases (Annegers et al. 1984). Thus, coronary atherosclerosis does not appear to be a major predisposing condition for SUDEP. The hyperadrenergic state associated with seizures constitutes a major component of arrhythmia risk because it is capable not only of directly triggering rhythm disturbances but also of provoking changes such as cardiac ἀbrosis (Kloster and Engelskjøn 1999), perivascular and interstitial myocardial ἀbrosis (Natelson et al. 1998) (Figure 43.2), or cardiac conduction system abnormalities (Opeskin et al. 2000), which could constitute a vulnerable myocardial substrate for arrhythmia. Indeed, the incidence of myocardial infarction is increased in patients with epilepsy over the general population (Annegers et

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

695

Central nervous system and autonomic activity

Adrenergic and muscarinic receptors

Coronary vascular resistance

Platelet aggregability

Myocardial perfusion

Second messenger Myocardial electrical stability

Figure 43.1╇ Interaction between neural triggers and cardiovascular substrate during auto-

nomic activation. Stimulation of beta1-adrenergic receptors can decrease electrical stability directly as a result of changes in second messenger formation and alterations in ion fluxes. This deleterious influence is opposed by muscarinic receptor stimulation, which inhibits presynaptically the release of norepinephrine and opposes its action at the receptor level. Catecholamines may also alter myocardial perfusion by complex means, including alphareceptor stimulation of coronary vessels and platelets and by impairing diastolic perfusion time due to adrenergically mediated sinus tachycardia. (From Verrier, R.L., in Rosen, M.R., and Palti, Y., ed.,€Lethal Arrhythmias Resulting from Myocardial Ischemia and Infarction, Kluwer Academic Publishers, Boston, MA, 1988. With permission.)

al. 1984), and increased serum levels of cardiac troponin I, indicative of myocardial injury, have been documented after uncomplicated epileptic seizures (Hajsadeghi et al. 2009). The fact that neural influences play a major role in the stability of heart rhythm and that€a number of indicators of autonomic function reveal signiἀcant alterations provide a basis for considering the important role of neural triggering of arrhythmias during epileptic seizures. Increased cardiac sympathetic nerve activity during a seizure is indicated by surges in heart rates to greater than 150 beats/min (Opherk et al. 2002; Nei 2004) and by heart rate variability analysis (Evrengul et al. 2003; Ronkainen et al. 2006; Persson et al. 2007). The higher maximum heart rates and tachycardia associated with seizures are indicative of heightened adrenergic activity and are more frequent during sleep in SUDEP victims by comparison with other patients with epilepsy (Nei et al. 2004). This proἀle suggests a link to sleep-state related autonomic activity (Verrier and Harper 2005; Verrier and Mittleman 2005; Verrier and Josephson 2009). Although it is generally viewed that sleep is a protected state, bursts in cardiac sympathetic nerve activity have been reported in healthy individuals during rapid eye movement (REM) sleep (Somers et al. 1993) (Figure 43.3). These surges are concentrated in short, irregular periods, reach levels higher than during wakefulness, trigger intermittent increases in heart rate and blood pressure, and have been implicated in nocturnal ventricular arrhythmias and myocardial ischemia (Kales and Kales 1970). No sleep studies with electroencephalographic data have been published in patients with epilepsy, and the single report of nocturnal heart rate variability analysis results indicated no differences across the nocturnal period from healthy control subjects (Persson et al. 2007). However, detailed analyses of short-term autonomic activity especially during sleep may provide important clues regarding the risk of SUDEP.

696 Sudden Death in Epilepsy: Forensic and Clinical Issues

(a)

(b) B

(c)

(d)

Figure 43.2╇ Reversible (a, b) and irreversible (c, d) pathologic conditions found at autopsy in patients with epilepsy who died suddenly. (a) Myocytes with vacuolization; note in lower right the nucleus being displaced by vacuole (original magnification ×400). (b) Several longitudinally cut myocytes showing areas of vacuolization (original magnification ×400). (c) Diffuse interstitial fibrosis with multiple areas of myocyte replacement by connective tissue (original magnification ×40). (d) Perivascular fibrosis (original magnification ×100). All specimens were stained with hematoxylin–eosin. (From Natelson, B. H., et al., Arch Neurol, 55, 857, 1998. With€permission.)

43.3â•…Neurocircuitry of Cardiac Rhythm Control Regulation of cardiac neural activity is highly integrated and is achieved by circuitry at multiple levels (Figure 43.4). Higher brain centers operate through elaborate pathways between and within the hypothalamus and medullary cardiovascular regulatory sites. Baroreceptor mechanisms are integral to autonomic control of the cardiovascular system, as evidenced by heart rate variability and baroreceptor sensitivity testing of both cardiac patients and normal subjects. The intrinsic cardiac nerves and fat pads provide local neural coordination independent of higher brain centers (Armour 1999). Electrical remodeling of the myocardium by nerve growth and degeneration is newly recognized (Zhou et al. 2004). At the level of the myocardial cell, autonomic receptors influence G proteins to control ionic channels, pumps, and exchangers (Verrier and Antzelevitch 2004). Studies of behavioral state provide evidence that markers of arrhythmia vulnerability and autonomic

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy Awake

Stage 4 SNA

SNA

125 BP 0

125 0

REM

Stage 2 SNA 125 BP 0

697

K

Stage 3 SNA

SNA 125 BP 0

T

10 s

125 BP 0

Figure 43.3╇ Recordings of sympathetic nerve activity (SNA) and mean blood pressure (BP) in a single subject while awake and while in stages 2, 3, 4, and rapid eye movement sleep. As non– rapid eye movement sleep deepened (stages 2 through 4), sympathetic nerve activity gradually decreased and blood pressure (measured in millimeters of mercury) and variability in blood pressure were gradually reduced. Arousal stimuli elicited K complexes on the electroencephalogram (not shown), which were accompanied by increases in sympathetic nerve activity and blood pressure (indicated by the arrows, stage 2 sleep). In contrast to the changes during non– rapid eye movement sleep, heart rate, blood pressure, and blood pressure variability increased during rapid eye movement sleep, together with a profound increase in both the frequency and the amplitude of sympathetic nerve activity. There was a frequent association between rapid eye movement twitches (momentary periods of restoration of muscle tone, denoted by “T” on the tracing) and abrupt inhibition of sympathetic nerve discharge and increases in blood pressure. (From Somers, V. K., et al., N Engl J Med, 328, 303–307, 1993. With permission.)

parameters can be monitored noninvasively during emotional and physical stressors as well as sleep states to identify individuals at heightened risk of lethal cardiac arrhythmias (Kovach et al. 2001; Kop et al. 2005; Lampert et al. 2009).

43.4â•… Autonomic Mechanisms in Arrhythmogenesis 43.4.1â•… Adrenergic Influences It is well established that adrenergic inputs constitute the primary neural trigger for arrhythmias (Verrier and Antzelevitch 2004). Activation of sympathetic nerve structures, including the posterior hypothalamus or stellate ganglia, increases susceptibility to ventricular ἀbrillation. A striking surge in sympathetic nerve activity also occurs within a few minutes of induction of myocardial ischemia by left anterior descending coronary artery occlusion in experimental animals, as documented by direct nerve recording (Lombardi et al. 1983). This enhancement in sympathetic nerve activity is associated with a marked increase in susceptibility to ventricular ἀbrillation, as evidenced by a fall in ventricular

698 Sudden Death in Epilepsy: Forensic and Clinical Issues Integration Level

Cerebral cortex Hypothalamus

Central

Medulla Dorsal root ganglia

Spinal cord

Spinal

Stellate and mid-cervical ganglia

Cervical and Thoracic Ganglia

Interneuron Afferent

Sympathetic

Intrinsic Cardiac Nerves

Intrinsic cardiac ganglionated plexus Interneuron Afferent

Arterial baroreceptors

Sympathetic

Epinephrine β1

N. terminal

Parasymp.

AC Gs

M2 Gi

Acetylcholine

Myocardial Cell

AMP cAMP

Figure 43.4╇ Synthesis of new and present views on levels of integration important in neural control of cardiac electrical activity. More traditional concepts focused on afferent tracts (dashed lines) arising from myocardial nerve terminals and reflex receptors (e.g., baroreceptors) that are integrated centrally within hypothalamic and medullary cardiostimulatory and cardioinhibitory brain centers and on central modulation of sympathetic and parasympathetic outflow (solid lines) with little intermediary processing at the level of the spinal cord and within cervical and thoracic ganglia. More recent views incorporate additional levels of intricate processing within the extraspinal cervical and thoracic ganglia and within the cardiac ganglionic plexus, where recently described interneurons are envisioned to provide new levels of noncentral integration. Release of neurotransmitters from postganglionic sympathetic neurons is believed to enhance excitation in the sinoatrial node and myocardial cells through norepinephrine binding to beta1-receptors, which enhances adenyl cyclase (AC) activity through intermediary stimulatory G proteins (Gs). Increased parasympathectomy outflow enhances postganglionic release and binding of acetylcholine to muscarinic (M 2) receptors, and through coupled inhibitory G proteins (Gi), inhibits cyclic AMP production (cAMP). The latter alters electrogenesis and pacemaking activity by affecting the activity of specific membrane Na, K, and Ca channels. New levels of integration are shown superimposed on previous views and are emphasized here to highlight new possibilities for intervention. (From Lathrop, D. A. and Spooner, P. M., J Cardiovasc Electrophysiol, 12, 841–844, 2001. With permission.)

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

699

ἀbrillation threshold and by spontaneous occurrence of the arrhythmia. Stellectomy signiἀcantly blunts this occlusion-induced surge in vulnerability to ventricular ἀbrillation (Lombardi et al. 1983; Nearing et al. 1991). Enhanced sympathetic nerve activity increases cardiac vulnerability in the normal and ischemic heart by complex processes. The major indirect effects include impairment of oxygen supply/demand ratio due to increased cardiac metabolic activity, alpha-adrenergically mediated coronary vasoconstriction, especially in vessels with damaged endothelium, and changes in preload and afterload. The direct arrhythmogenic effects on cardiac electrophysiologic function, which are primarily mediated through beta1-adrenergic receptors, include derangements in impulse formation, conduction, repolarization alternans, and heterogeneity of repolarization (Janse and Wit 1989; Nearing and Verrier 2003). Increased levels of catecholamines stimulate beta-adrenergic receptors, which in turn alter adenylÂ� ate cyclase activity and intracellular calcium flux (Opie 2004; Verrier et al. 2009). These effects are probably mediated by the cyclic nucleotide and protein kinase regulatory cascade, which can alter spatial heterogeneity of calcium transients and consequently provoke T-wave alternans and dispersion of repolarization. The effects of increased intracellular calcium, with the potential for overload and impaired intracellular calcium cycling by the sarcoplasmic reticulum, may be compounded and become especially arrhythmogenic during concurrent myocardial ischemia, which further predisposes to intracellular calcium excess (Verrier et al. 2009). Cardiac beta1-adrenergic receptor blockade is capable of negating the proἀbrillatory effect of direct sympathetic nerve stimulation by an action at the neurocardiac effector junction. However, cardiac beta2-adrenergic receptors, which control vasodilation in the vasculature and bronchi (Brodde et al. 1991), do not appear to play a signiἀcant role in modulating ventricular excitable properties. 43.4.2â•… Alpha-Adrenergic Receptors In the normal heart, alpha-adrenergic receptor stimulation or blockade does not appear to affect ventricular electrical stability, as evidenced by the fact that administration of an alpha-adrenergic agonist such as phenylephrine or methoxamine does not influence excitable properties when the pressor response is controlled to prevent reflex changes in autonomic tone (Kowey et al. 1983). In the setting of myocardial ischemia, alpha-adrenergic blockade may alleviate coronary vasoconstriction and reduce platelet aggregability. Thus, alpha-adrenergic receptor activity exerts direct actions not only on myocardial excitable properties but also on platelet aggregability and coronary hemodynamic function. 43.4.3â•… Sympathetic–Parasympathetic Interactions Vagus nerve influences are contingent on the prevailing level of adrenergic tone. When sympathetic tone to the heart is augmented by thoracotomy, sympathetic nerve stimulation, myocardial ischemia, or catecholamine infusion, vagal activation exerts a protective effect on ventricular vulnerability. Vagus nerve stimulation is without effect on ventricular vulnerability when adrenergic input to the heart is ablated by beta-adrenergic blockade, a phenomenon termed accentuated antagonism (Verrier and Antzelevitch 2004). The basis for this antagonism of adrenergic effects is presynaptic inhibition of norepinephrine release from nerve endings and a muscarinically mediated action at the second messenger level,

700 Sudden Death in Epilepsy: Forensic and Clinical Issues

which attenuates the response to catecholamines at receptor sites. In addition, importantly, vagal influences provide indirect protection against ventricular ἀbrillation by reducing excess heart rates, which can otherwise increase ischemic insult by critically compromising diastolic perfusion time during acute myocardial ischemia. However, beneἀcial effects of vagus nerve activity may be annulled if profound bradycardia and hypotension ensue. Vagus nerve stimulation has been shown in experimental studies to protect against ventricular arrhythmias during myocardial ischemia. Myocardial infarction may damage nerve pathways by sustained reduction in supply of oxygen and nutrients to the nerve, thereby limiting the potential of the vagus nerve to be activated. Seizure activity causes local vasospasm and accumulation of microthrombi and could also thereby lead to impaired blood flow and damage to the nerve supply. Vanoli and colleagues (1991) demonstrated the antiἀbrillatory effect of vagus nerve stimulation during exercise-induced ischemia in canines with a healed myocardial infarction. Direct stimulation of the right cervical vagus through a chronically implanted electrode at 15 s after onset of exercise-induced acute myocardial ischemia reduced the incidence of ventricular ἀbrillation by 92%. This effect was only partly due to the attendant heart rate reduction, as in half of the animals, the antiarrhythmic efficacy of vagus nerve stimulation persisted despite maintenance of constant heart rate by atrial pacing. Vagus nerve stimulation in humans to elicit beneἀcial cardiac effects has recently been explored by Schwartz et al. (2008), who demonstrated that this procedure is safe and well tolerated in patients with heart failure, as indicated by some improvement in heart failure class and a decrease in left ventricular end-systolic volume. The exciting possibility that chronically implanted vagus nerve stimulation devices may suppress arrhythmias clinically remains to be explored. The potential for this form of therapy is underscored by the extensive experimental evidence and the clinical demonstration that vagomimetic maneuvers such as carotid sinus massage and administration of agents such as phenylephrine or edrophonium can terminate ventricular tachyarrhythmias (Waxman et al. 1994). 43.4.4â•… Baroreflexes and Arrhythmias The classic studies by Billman et al. (1982) drew attention to the importance of baroreceptor function in susceptibility to life-threatening arrhythmias associated with myocardial isÂ�Â� chemia and infarction. In their initial investigations in canines, they demonstrated that the more intense the baroreflex response was, the less vulnerable animals were to ventricular ἀbrillation during myocardial ischemia superimposed on prior myocardial infarction. The protective effect of the baroreceptor mechanism has been linked primarily to the antiἀbrillatory influence of vagus nerve activity. The latter effect improves diastolic coronary perfusion, minimizing the ischemic insult from coronary artery occlusion. The importance of baroreceptor sensitivity was subsequently documented in human subjects in whom baroreceptor function was evaluated with the pressor agent phenylephrine. La Rovere et al. (1998) demonstrated that following myocardial infarction, patients were less likely to experience sudden cardiac death if their baroreceptor function was preserved.

43.5â•… Behavioral Stress and Arrhythmias Because epilepsy is associated with intense emotions such as anxiety and depression (Lathers and Schraeder 2006), it is relevant to consider experimental evidence of the impact

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

701

of behavioral state on cardiac electrical stability. In early studies, we developed aversive behavioral conditioning paradigms and models eliciting natural emotions, notably anger and fear. Aversive conditioning of dogs in a Pavlovian sling with mild chest shock on 3 consecutive days resulted in a reduction in the repetitive extrasystole threshold greater than 30% during subsequent exposure to the environment without shock. The same paradigm elicited a threefold increase in the occurrence of spontaneous ventricular ἀbrillation when coronary artery occlusion was carried out in the aversive sling compared to the nonaversive cage environment (Lown et al. 1973). In dogs recovering from myocardial infarction, exposure to the aversive environment consistently elicited ventricular tachycardia for several days during the healing process. After this time, the animals continued to exhibit signs of behavioral stress in the aversive environment but no longer experienced ventricular arrhythmias, indicating that the arousal state required a substrate of cardiac electrical instability for the induction of rhythm disturbances. The stress-induced changes in cardiac excitable properties were largely obtunded by beta-adrenergic receptor blockade with propranolol or metoprolol. In a separate series of experiments, an experimental canine model was developed to emulate anger (Verrier et al. 1987; Kovach et al. 2001), which is the emotion most commonly associated with myocardial infarction and sudden death (Mittleman et al. 1995; Verrier and Mittleman 1996). A standardized food-access-denial paradigm provoked intense arousal and pronounced myocardial ischemia in territories of stenosed coronary vessels. The angerlike state also elicited severe repolarization abnormalities conducive to life-Â�threatening cardiac€arrhythmias.

43.6â•…Ambulatory Electrocardiogram-Based Tools for Evaluation of Autonomic Function Several studies have indicated that epileptic seizures are associated with surges in heart rate (Opherk et al. 2002; Nei et al. 2004) and reduced heart rate variability (Ronkainen et al. 2006; Persson et al. 2007; Evrengul et al. 2005), indicative of a hyperadrenergic state. This ἀnding is important because depressed heart rate variability is associated with heightened risk for sudden cardiac death in patients with ischemic heart disease (Lombardi 2002; Tulppo et al. 1996). Because depressed baroreceptor sensitivity is associated with heightened risk for arrhythmias, it will be crucial to establish whether baroreceptor reflexes might be altered by recurring epileptic seizures. Baroreceptor sensitivity can be noninvasively measured from ambulatory electrocardiograms by monitoring heart rate turbulence, a tool that has not been explored in patients with epilepsy. Heart rate turbulence refers to fluctuations of sinus-rhythm cycle length after a single ventricular premature beat. These effects are a direct function of baroreceptor responsiveness because reflex activation of the vagus nerve controls the pattern of sinus rhythm. Several studies conἀrm that in low-risk patients with cardiovascular disease, after a ventricular premature beat, sinus rhythm exhibits a characteristic pattern of early acceleration and subsequent deceleration. By contrast, patients at high risk for cardiovascular events exhibit essentially a flat, nonvarying response to the ventricular premature beat, indicating inability to activate vagus nerves and their cardioprotective effect. The method is an independent predictor of total mortality in patients with ischemic heart disease or heart failure (Schmidt et al. 1999; Bauer et al. 2008). Heart

702 Sudden Death in Epilepsy: Forensic and Clinical Issues

rate deceleration capacity, a related and even more comprehensive marker of autonomic control than heart rate turbulence, may be of considerable clinical value in assessing overall autonomic regulation of the heart in patients with diverse types of cardiovascular disease (Bauer et al. 2006) and could also be used to evaluate patients with epilepsy.

43.7â•…Ambulatory Electrocardiogram-Based T-Wave Alternans to Assess Cardiac Electrical Instability and Risk for Sudden Cardiac Death Extensive scientiἀc evidence spanning more than a decade points to a fundamental link between T-wave alternans, a beat-to-beat fluctuation in the morphology of the T wave of the electrocardiogram, and susceptibility to malignant ventricular tachyarrhythmias (VerÂ� rier et al. 2009). The basic principle is that T-wave alternans is an indicator of temporal– spatial heterogeneity of repolarization, a precondition for reentrant arrhythmias. T-wave alternans also detects derangements in intracellular calcium cycling, which is subject to the influences of both the autonomic nervous system and alterations in myocardial substrate, particularly ischemia and scar associated with infarction. Similar mechanisms may operate in association with cardiac ἀbrosis, contraction band necrosis, or other types of myocardial injury that are known to occur in patients with epilepsy. Diverse physiologic interventions have been shown to alter T-wave alternans magnitude in parallel with their influence on vulnerability to ventricular tachyarrhythmias. Speciἀcally, these include elevations in heart rate, coronary artery occlusion and reperfusion, and sympathetic nerve stimulation (Nearing et al. 1991). In the chronically instrumented canine model developed to emulate anger (Verrier et al. 1987; Kovach et al. 2001), Alternans (mV ms) Baseline

0.04

Occlusion

0.47

Anger

2.88

Anger and occlusion Ventricular fibrillation Lead: V4 No drug

3.97 1 mV 1s

Figure 43.5╇ Flow chart of the major components of the modified moving average method of

T-wave alternans analysis. The left ventricular electrocardiogram was obtained from a representative experiment in which coronary artery occlusion subsequently resulted in ventricular fibrillation. Electrocardiograms are filtered to reduce high-frequency noise and to remove baseline wander. Ventricular and supraventricular premature beats as well as beats with a high noise level are removed. The even beats in the sequence are then assigned to group A and the odd beats to group B. Modified moving average computed beats of types A and B are computed continuously. The nth computed beat is developed from the n − 1th electrocardiographic beat and the n − 1th computed beat. The amplitude of the effect of any one beat on the computed beat is also limited by the bounds on ΔA and ΔB. The alternans estimate is determined as the maximum absolute difference between A and B computed beats within the ST-segment and T-wave region. The output period can be adjusted as desired. (From Nearing, B. D. and Verrier,€R.€L., J€Appl Physiol, 92, 541–549, 2002. With permission.)

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

703

a pronounced increase in T-wave alternans was observed during behavioral arousal. A 3-min period of coronary artery occlusion potentiated arrhythmia risk, more than doubling the magnitude of anger-induced T-wave alternans (Figure 43.5). The stress-related effects were signiἀcantly lessened by metoprolol, further implicating a major role of beta1adrenergic receptors in sympathetic nerve induction of cardiac vulnerability and T-wave alternans (Kovach et al. 2001; Lathers et al. 2008). Vagus nerve stimulation, blockade of beta-adrenergic receptors, and sympathetic denervation, which reduce susceptibility to ventricular tachyarrhythmias, have been shown to decrease T-wave alternans magnitude (Verrier et al. 2009). These series of observations underscore the fundamental link between T-wave alternans and vulnerability to lethal arrhythmias, which underlies the utility of this parameter in assessing propensity for life-threatening ventricular arrhythmias.

43.8â•…Clinical Evidence of Ambulatory Electrocardiogram-Based T-Wave Alternans for Prediction of Sudden Cardiac Death A few years ago, we developed a time-domain method, termed Modified Moving Average analysis was developed to quantify T-wave alternans during both routine exercise stress testing and ambulatory electrocardiographic monitoring (Nearing and Verrier 2002). The technique is based on the powerful noise-rejection principle of recursive averaging (Figure

A B A B A B Noise reduction Baseline wander removal and separation of beats

A

A

B

A

B

B

Median beat B

Median beat A

Alternans measurement

Figure 43.6╇ Segments of electrocardiogram tracings with measurements of maximum T-wave alternans magnitude at baseline, during myocardial ischemia, during angerlike state, and with simultaneous angerlike state and myocardial ischemia in one dog that experienced ventricular fibrillation at 42 s after provocation of angerlike response was superimposed at 1 min of coronary artery occlusion. (From Kovach, J. A., et al., J Am Coll Cardiol, 37, 1719–1725, 2001. With permission.)

704 Sudden Death in Epilepsy: Forensic and Clinical Issues

43.6), and respiration and motion artifacts have been further reduced by cubic alignment and other ἀlters. Modiἀed moving average analysis computes T-wave alternans magnitude as the peak difference between A and B beats in an ABAB beat stream at any point within the JT interval of the electrocardiogram. The predictive capacity of this technique has been examined in consecutive patients referred for routine exercise stress testing enrolled in the Finnish Cardiovascular Study (FINCAVAS) (Nieminen et al. 2007; Minkkinen et al. 2009). Nieminen et al. (2007) found in multivariate analysis in this low-risk population that the relative risk of T-wave alterÂ�nans ≥65 µV for sudden cardiac death was 7.4 (95% CI, 2.8–19.4; P < 0.001), for cardiovascular mortality was 6.0 (95% CI, 2.8–12.8; P < 0.001), and for all-cause mortality was 3.3 (95% CI, 1.8–6.3; P = 0.001), indicating that T-wave alternans exhibits speciἀcity for sudden cardiac death. The modiἀed moving average method also proved suitable for analyzing T-wave alternans from ambulatory electrocardiographic monitoring records to investigate the effects of physical activity (Verrier et al. 2003), circadian factors (Verrier et al. 2003; Zanobetti et al. 2009), mental stress (Kop et al. 2004; Lampert et al. 2009), and sleep states in arrhythmogenesis in patients with diverse conditions including myocardial ischemia, infarction, and heart failure (Verrier et al. 2009). Recently, in a prospective trial, Sakaki et al. (2009) demonstrated the remarkable predictive capacity of T-wave alternans in ambulatory electrocardiograms in both ischeÂ�mic and nonischemic patients with left ventricular dysfunction (Figures 43.7 and 43.8). As mental stress has been implicated in seizure-induced risk, it is important to recognize that T-wave alternans has been shown to detect cardiac electrical instability induced by mental arithmetic and anger recall in patients with implantable cardioverter deἀbrillators. V5 (CM5) Max TWA 67 μV

V1 (NASA) Max TWA 98 μV

Figure 43.7╇ An example of a positive Modified Moving Average T-wave alternans test. Superimposed complexes from precordial leads V5 (CM5) and V1 (NASA) illustrate peak TWA. In this patient, TWA was determined to be positive because the peak TWA voltage was 98 μV in lead V1. (From Sakaki, K., et al., Heart Rhythm, 6, 332–337, 2009. With permission.)

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

705

100

Survival rate (%)

80 60

Hazard ratio = 15.9 95% Cl 6.7–37.8, P < 0.001

40

TWA negative

20

TWA positive

0 0

50

100

150

200

250

300

350

350

Time (days)

Figure 43.8╇ Event-free curves for cardiac mortality using peak voltage of Modified Moving Average T-wave alternans from 24-h ambulatory electrocardiogram recordings in either lead V1 or V5. (From Sakaki, K., et al., Heart Rhythm, 6, 332–337, 2009. With permission.)

Kop et al. (2004) and Lampert et al. (2009) reported that during these routine challenges, patients with implantable cardioverter deἀbrillators exhibited signiἀcant increases in T-wave alternans level compared to age-matched healthy control subjects. Furthermore, Lampert et al. (2009) determined that the rise in T-wave alternans level appears to coincide with enhanced risk for life-threatening arrhythmias. Testing patients with epilepsy with simple mental tasks such as mental arithmetic could be used to disclose latent cardiac electrical instability attributable to heightened adrenergic activity, behavioral state, or substrate changes, as manifested by elevated levels of T-wave alternans.

43.9â•…Update on Vagus Nerve Stimulation in Epilepsy Vagus nerve stimulation has been approved by the FDA for the adjunctive treatment of medically refractory partial seizures in patients older than 12 years since 1997 (Schachter 2002). Side effects related to stimulation are usually mild; almost always resolve with adjustment in the stimulation settings; and include hoarseness, throat pain, coughing, shortness of breath, and tingling. Although there are mixed reports of measurable autonomic effects of vagus nerve stimulation when used for epilepsy (Banzett et al. 1999; Binks et al. 2001; Schachter 2008), SUDEP rates halved after more than 2 years of treatment (Annegers et al. 2000). Therefore, the possibility exists but remains to be proven that vagus nerve stimulation may reduce cardiac arrhythmias and risk for sudden death, potentially mediated by the efferent cardiac vagus nerve (Vanoli et al. 1991; Verrier and Antzelevitch 2004). This potential beneἀt in patients with epilepsy is intriguing and unexplored but, given the scientiἀc rationale, deserves pursuit.

43.10â•… Conclusions and Implications Extensive evidence implicates the interaction between autonomic factors and a vulnerable myocardial substrate in predisposing patients to life-threatening cardiac arrhythmias. Adrenergic influences constitute the primary trigger, and vagus nerve stimulation can

706 Sudden Death in Epilepsy: Forensic and Clinical Issues

antagonize the arrhythmogenic effect of catecholamines to exert a cardioprotective action. Recently, new ambulatory electrocardiographic tools have been developed to assess autonomic function and vulnerability to cardiac arrhythmias, namely, heart rate turbulence and T-wave alternans, respectively. Although these methodologies have been extensively tested in populations with ischemic and nonischemic heart disease, their use in patients with epilepsy should be explored. Because it is likely that arrhythmic events result from an interaction between autonomic triggers and a vulnerable myocardial substrate, there is a great need for systematic longitudinal studies of electrocardiogram-based pathophysiologic markers including ST-segment changes, QRS interval duration, and R-wave changes and T-wave alternans in patients with epilepsy to provide insights into the factors leading to SUDEP. Increasing use of implantable loop recorders in patients with epilepsy promises to expand the opportunity to understand disturbances in autonomic function and cardiac rhythm associated with ictal events (Zaidi et al. 2000; Rugg-Gunn et al. 2004; Ronkainen et al. 2006). Thus, in the future, these parameters could not only play a role in advancing our understanding of the underlying cardiac pathologies responsible for SUDEP but also enable evaluation of cardiac antiarrhythmic as well as proarrhythmic effects of medical antiepileptic therapy and provide the basis for considering patients with epilepsy as potential candidates for antiarrhythmic therapies. Especially intriguing is the possibility that vagus nerve stimulation (Schachter 2009) could not only reduce seizure activity but may also protect the heart (Schwartz et al. 2008), as activation of this neural pathway may be antiarrhythmic. Ambulatory electrocardiogram-based T-wave alternans can detect vagus nerve influences on cardiac electrical instability, providing a tool for quantiἀcation of the potential cardioprotective effect of this potent intervention.

References Annegers, J. F., W. A. Hauser, and S. B. Shirts. 1984. Heart disease mortality and morbidity in patients with epilepsy. Epilepsia 25: 699–704. Annegers, J. F., S. P. Coan, W. A. Hauser, and J. Leestma. 2000. Epilepsy, vagal nerve stimulation by the NCP system, all-cause mortality, and sudden, unexpected, unexplained death. Epilepsia 41: 549–553. Armour, J. A. 1999. Myocardial ischaemia and the cardiac nervous system. Cardiovasc Res 41: 41–54. Banzett, R. B., A. Guz, D. Paydarfar et al. 1999. Cardiorespiratory variables and sensation during stimulation of the left vagus in patients with epilepsy. Epilepsy Res 35: 1–11. Bauer, A., J. W. Kantelhardt, P. Barthel et al. 2006. Deceleration capacity of heart rate as a predictor of mortality after myocardial infarction: Cohort study. Lancet 367: 1674–1681. Bauer, A., M. Malik, G. Schmidt et al. 2008. Heart rate turbulence: Standards of measurement, physiological interpretation, and clinical use: International Society for Holter and Noninvasive Electrocardiology Consensus. J Am Coll Cardiol 52: 1353–1365. Billman, G. E., P. J. Schwartz, and H. L. Stone. 1982. Baroreceptor reflex control of heart rate: A predictor of sudden cardiac death. Circulation 66: 874–880. Binks, A. P., D. Paydarfar, S. C. Schachter, A. Guz, and R. B. Banzett. 2001. High strength stimulation of the vagus nerve in awake humans: A lack of cardiorespiratory effects. Respirat Physiol 127: 125–133. Brodde, O. E. 1991. Beta1- and beta2-adrenoceptors in the human heart: Properties, function and alterations in chronic heart failure. Pharmacol Rev 43: 203–242.

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

707

Evrengul, H., H. Tanriverdi, D. Dursunoglu et al. 2005. Time and frequency domain analyses of heart rate variability in patients with epilepsy. Epilepsy Res 63: 131–139. Goldberger, J. J., M. E. Cain, S. H. Hohnloser et al. 2008. American Heart Association/American College of Cardiology Foundation/Heart Rhythm Society Scientiἀc Statement on Noninvasive Risk Stratiἀcation Techniques for Identifying Patients at Risk for Sudden Cardiac Death. A scientiἀc statement from the American Heart Association Council on Clinical Cardiology Committee on Electrocardiography and Arrhythmias and Council on Epidemiology and Prevention. Circulation 118: 1497–1518. Hajsadeghi, S., S. Afsharian, S. M. Fereshtehnejad, M. R. Keramati, and R. Mollahoseini. 2009. Serum levels of cardiac troponin I in patients with uncomplicated epileptic seizure. Arch Med Res 40: 24–28. Jallon, P. 2004. Mortality in patients with epilepsy. Curr Opin Neurol 17: 141–146. Janse, M. J., and A. L. Wit. 1989. Electrophysiological mechanisms of ventricular arrhythmias resulting from myocardial ischemia and infarction. Physiol Rev 69: 1049–1169. Kales, A., and J. D. Kales. 1970. Evaluation, diagnosis, and treatment of clinical conditions related to sleep. JAMA 213: 2229–2232. Kloster, R., and T. Engelskjøn. 1999. Sudden unexpected death in epilepsy (SUDEP): A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67: 439–444. Kop, W. J., D. S. Krantz, B. D. Nearing et al. 2004. Effects of acute mental stress and exercise on T-wave alternans in patients with implantable cardioverter deἀbrillators and controls. Circulation 109: 1864–1869. Kovach, J. A., B. D. Nearing, and R. L. Verrier. 2001. An anger-like behavioral state potentiates myocardial ischemia-induced T-wave alternans in canines. J Am Coll Cardiol 37: 1719–1725. Kowey, P. R., R. L. Verrier, and B. Lown. 1983. Effect of alpha-adrenergic receptor stimulation on ventricular electrical properties in the normal canine heart. Am Heart J 105: 366–371. La Rovere, M. T., J. T. Bigger Jr., F. I. Marcus et al. 1998. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet 351: 478–484. Lampert, R., V. Shusterman, M. M. Burg et al. 2009. Anger-induced T-wave alternans predicts future ventricular arrhythmias in patients with implantable cardioverter-deἀbrillators. J Am Coll Cardiol 53: 774–778. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case-control study of SUDEP. Neurology 64: 1131–1133. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Lathrop, D. A., and P. M. Spooner. 2001. On the neural connection. J Cardiovasc Electrophysiol 12: 841–844. Lombardi, F., R. L. Verrier, and B. Lown. 1983. Relationship between sympathetic neural activity, coronary dynamics, and vulnerability to ventricular ἀbrillation during myocardial ischemia and reperfusion. Am Heart J 105: 958–965. Lombardi, F. 2002. Clinical implications of present physiological understanding of HRV components. Card Electrophysiol Rev 6: 245–249. Lown, B., R. L. Verrier, and R. Corbalan. 1973. Psychologic stress and threshold for repetitive ventricular response. Science 182: 834–836. Minkkinen, M., M. Kähönen, J. Viik et al. 2009. Enhanced predictive power of quantitative TWA during routine exercise testing in the Finnish Cardiovascular Study. J Cardiovasc Electrophysiol 20: 408–415. Mittleman, M. A., M. Maclure, J. B. Sherwood et al. 1995. Triggering of acute myocardial infarction onset by episodes of anger. Circulation 92: 1720–1725. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55: 857.

708 Sudden Death in Epilepsy: Forensic and Clinical Issues Nearing, B. D., A. H. Huang, and R. L. Verrier. 1991. Dynamic tracking of cardiac vulnerability by complex demodulation of the T-wave. Science 252: 437–440. Nearing, B. D., and R. L. Verrier. 2002. Modiἀed moving average method for T-wave alternans analysis with high accuracy to predict ventricular ἀbrillation. J Appl Physiol 92: 541–549. Nearing, B. D., and R. L. Verrier. 2003. Tracking heightened cardiac electrical instability by computing interlead heterogeneity of T-wave morphology. J Appl Physiol 95: 2265–2272. Nei, M., R. T. Ho, and B. W. Abou-Khalil et al. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45: 338–345. Nieminen, T., T. Lehtimäki, J. Viik et al. 2007. T-wave alternans predicts mortality in a population undergoing a clinically indicated exercise test. Eur Heart J 28: 2332–2337. Opeskin, K., A. Thomas, and S. F. Berkovic. 2000. Does cardiac conduction pathology contribute to sudden unexpected death in epilepsy? Epilepsy Res 40: 17–24. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52: 117–127. Opie, L. H. 2004. Heart Physiology: From Cell to Circulation, 4th ed. Philadelphia, PA: Lippincott, Williams, and Wilkins. Persson, H., E. Kumlien, M. Ericson, and T. Tomson. 2007. Circadian variation in heart-rate variability in localization-related epilepsy. Epilepsia 48: 917–922. Rocamora, R., M. Kurthen, L. Lickfett, J. von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44: 179–185. Ronkainen, E., J. T. Korpelainen, E. Heikkinen et al. 2006. Cardiac autonomic control in patients with refractory epilepsy before and during vagus nerve stimulation treatment: A one-year follow-up study. Epilepsia 47: 556–562. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Ryvlin, P., A. Montavont, and P. Kahane. 2006. Sudden unexpected death in epilepsy: From mechanisms to prevention. Curr Opin Neurol 19: 194–199. Sakaki, K., T. Ikeda, Y. Miwa et al. 2009. Time-domain T-wave alternans measured from Holter electrocardiograms predicts cardiac mortality in patients with left ventricular dysfunction: A prospective study. Heart Rhythm 6: 332–337. Schachter, S. C. 2002. Vagus nerve stimulation: Where are we? Curr Opin Neurol 15: 201–206. Schachter, S. C. 2008. Review of “Effects of vagus nerve stimulation on cardiovascular regulation in patients with epilepsy.” J Watch Neurol 10: 52. Schachter, S. C. 2009. Vagal nerve stimulation. In The Treatment of Epilepsy, 3rd ed., ed. S. Shorvon, E. Perucca, and J. Engle, 1017–1023. London: Blackwell Publishing. Schmidt, G., M. Malik, P. Barthel et al. 1999. Heart-rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial infarction. Lancet 353: 1390–1396. Schwartz, P. J., G. De Ferrari, A. Sanzo et al. 2008. Long term vagal stimulation in patients with advanced heart failure: First experience in man. Eur J Heart Failure 10: 884–891. Somers, V. K., M. E. Dyken, A. L. Mark et al. 1993. Sympathetic nerve activity during sleep in normal subjects. N Engl J Med 328: 303–307. Tigaran, S., H. Molgaard, R. McClelland, M. Dam, and A. S. Jaffe. 2003. Evidence of cardiac ischemia during seizures in drug refractory epilepsy patients. Neurology 60: 492–495. Tomson, T., T. Walczak, M. Sillanpaa, and J. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61. Tulppo, M. P., T. H. Makikallio, T. E. S. Takala, T. Seppanen, and H. V. Huikuri. 1996. Quantitative beat-to-beat analysis of heart rate dynamics during exercise. Am J Physiol 271: H244–H252. Vanoli, E., G. M. De Ferrari, M. Strambaâ•‚Badiale et al. 1991. Vagal stimulation and prevention of sudden death in conscious dogs with a healed myocardial infarction. Circ Res 68: 1471–1481. Verrier, R. L., E. L. Hagestad, and B. Lown. 1987. Delayed myocardial ischemia induced by anger. Circulation 75: 249–254.

Neurocardiac Interactions in Sudden Unexpected Death in Epilepsy

709

Verrier, R. L. 1988. Central nervous system modulation of cardiac rhythm. In Lethal Arrhythmias Resulting from Myocardial Ischemia and Infarction, ed. M. R. Rosen and Y. Palti, 149–164. Boston, MA: Kluwer Academic Publishers. Verrier, R. L., and M. A. Mittleman. 1996. Life-threatening cardiovascular consequences of anger in patients with coronary heart disease. Cardiol Clin 14: 289–307. Verrier, R. L., B. D. Nearing, M. T. LaRovere et al. 2003. Ambulatory ECG-based tracking of T-wave alternans in post-myocardial infarction patients to assess risk of cardiac arrest or arrhythmic death. J Cardiovasc Electrophysiol 14: 705–711. Verrier, R. L., and C. A. Antzelevitch. 2004. Autonomic aspects of arrhythmogenesis: The enduring and the new. Curr Opin Cardiol 19: 2–11. Verrier, R. L., R. M. Harper, and J. A. Hobson. 2005. Central and autonomic mechanisms regulating cardiovascular function. In Principles and Practice of Sleep Medicine, 4th ed., eds. M. H. Kryger, T. Roth, and W. C. Dement, 192–202. Philadelphia, PA: WB Saunders. Verrier, R. L., and M. A. Mittleman. 2005. Sleep-related cardiac risk. In Principles and Practice of Sleep Medicine, 4th ed., ed. M. H. Kryger, T. Roth, and W. C. Dement, 1161–1170. Philadelphia, PA: WB Saunders. Verrier, R. L., and M. E. Josephson. 2009. Impact of sleep on arrhythmogenesis. Circ Arrhythmia Electrophysiol 2: 450–459. Verrier, R. L., K. Kumar, and B. D. Nearing. 2009. Basis for sudden cardiac death prediction by T-wave alternans from an integrative physiology perspective. Heart Rhythm 6 (3): 416–422. Waxman, M. B., D. Cameron, and R. W. Wald. 1994. Vagal activity and ventricular tachyarrhythmias. In Vagal Control of the Heart, eds. M. N. Levy, and P. J. Schwartz, 579–612. Mt. Kisco, NY: Futura. Zaidi, A., P. Clough, P. Cooper, B. Scheepers, and A. P. Fitzpatrick. 2000. Misdiagnosis of epilepsy: Many seizure-like attacks have a cardiovascular cause. J Am Coll Cardiol 36: 181–184. Zanobetti, A., P. H. Stone, F. E. Speizer et al. 2009. T-wave alternans, air pollution and traffic in highrisk subjects. Am J Cardiol 104 (5): 665–670. Zhou, S., L. S. Chen, Y. Miyauchi et al. 2004. Mechanisms of cardiac nerve sprouting after myocardial infarction in dogs. Circ Res 95: 76–83. Zijlmans, M., D. Flanagan, and J. Gotman. 2002. Heart rate changes and ECG abnormalities during epileptic seizures: prevalence and deἀnition of an objective clinical sign. Epilepsia 43: 847–854.

Arrhythmogenic, Respiratory, and Psychological Risk Factors for Sudden Unexpected Death and Epilepsy Case Histories

44

Claire M. Lathers

Contents 44.1 Introduction 44.2 Case Histories of SUDEP or “Near Miss” SUDEP 44.3 Discussion 44.3.1 Analysis of the Electrocardiogram Is Essential for All Patients Who Present with Seizures References

711 712 714 718 718

44.1â•…Introduction The classiἀcation of deaths in the context of SUDEP by coroners and medical examiners in the United States is underreported as per recent surveys conducted by Schraeder et al. (2006, 2009). In practice, the term SUDEP is not routinely used on death certiἀcate diagnoses. Thus, the incidence for SUDEP may actually be higher than reported. In addition, SUDEP is most often an unwitnessed event (Langan et al. 2000). These facts make the prevention of SUDEP more difficult. Physicians and caregivers responsible for treating persons with epilepsy must know the risk factors for SUDEP and evaluate these potential risks for each and every patient. Lathers et al. (2008a, 2008b) and Scorza et al. (2008) have recently summarized the proposed mechanistic factors in SUDEP with the three major risk categories of arrhythmogenic factors, respiratory factors and hypoxia, and psychological factors. The SUDEP case histories discussed by Langan et al. (2000) and the two cases of “near miss” sudden death in persons with epilepsy published by Tavee and Morris (2008) and by So et al. (2000) are presented below. The discussion emphasizes the mystery of sudden death and the difficulty of identifying given risk factors and associated mechanisms for risks (Lathers et al. 2008a, 2008b; Scorza et al. 2008). Lack of complete understanding of the risk factors and mechanisms of SUDEP make prevention of SUDEP more difficult. These case histories introduce the reader to problems associated with treating persons with epilepsy and problems associated with the objective of preventing the occurrence of SUDEP.

711

712 Sudden Death in Epilepsy: Forensic and Clinical Issues

44.2â•… Case Histories of SUDEP or “Near Miss” SUDEP A Series o f Wit nessed SUDEP Deat hs Fifteen witnessed deaths were included in a total group of 125 cases of SUDEP identiἀed by coroners, neurologists, and families of victims (Langan et al. 2000). Twelve of the deaths occurred in association with convulsive seizures. One victim exhibited a generalized seizure and then collapsed about 5 min later. Of the last two witnessed deaths, one victim died during what was thought to be a “probable postictal state” and one died after experiencing an aura. The witnesses reported that 12 of the 15 cases experienced respiratory difficulty. Co mment s by L ang an et al. (2000) Most of the sudden deaths in persons with epilepsy were not witnessed, that is, only 15 of 125 SUDEP cases. Most of the deaths that were witnessed occurred in conjunction with a seizure. Respiratory difficulty/compromise was reported to be the prominent problem observed. The authors suggest that repositioning of the patient and/or stimulation of respiration may be important to help prevent sudden deaths in persons with epilepsy. Co mment s by L at hers The possible mechanisms of SUDEP are several (see discussion and Tables 1.1, 1.2, and 1.4 in Chapter 1) and include central and obstructive apnea and cardiovascular, including cardiac arrhythmias. Caution must be exerted when concluding respiratory changes alone are the primary mechanism of death. At the ἀrst onset of the clinical problem, cardiac arrhythmias may be felt by the patient but may not be visually detected by a witness. “Invisible” cardiac arrhythmias may be initiated and then followed by “visible” respiratory distress. Therefore, in addition to repositioning the patient to ensure ease of respiration and/or stimulation of respiration, it is important, if possible, also to simultaneously monitor and medically support cardiac rate and rhythm. The answer to the question of “Which is the cart and which is the horse?” or “Which is the egg and which is the chicken?” will vary from patient to patient depending on the type of initiating clinical event. However, if an event is witnessed, it is of the upmost importance to support all vital systems, that is, respiratory, cardiac, and circulatory. If intravenous access is difficult to obtain in an adult, and even more so in a pediatric patient during a witnessed event, it is possible that the intraosseous route of drug administration would allow rapid access to the circulation to provide anticonvulsant drugs and/or drugs used to treat cardiac arrest (Jim et al. 1988, 1989; Lathers et al. 1989a, 1989b, 1989c; Schoffstall et al. 1989; Spivey et al. 1987a, 1987b). Cases summarized from Langan et al. (2000).

Arrhythmogenic, Respiratory, and Psychological Risk Factors for SUDEP

713

Po st ict al Lary ng o spasm as One€Po t ent ial Mechanism fo r SUDEP A 1-min generalized tonic–clonic seizure occurred as a 42-year-old male patient diagnosed with refractory epilepsy was being monitored in an epilepsy-monitoring unit (Tavee and Morris 2008). This episode was followed by persistent inspiratory stridor and cyanosis. Cardiac parameters remained stable during the event. However, the respiratory status declined rapidly, despite the administration of oxygen via bagvalve-mask. As the emergency code team proceeded with intubation, they noted severe laryngospasm as the endotracheal tube was being inserted. Resuscitation was successful. This case demonstrated that postictal laryngospasm may be one potential cause of sudden unexpected death in persons with epilepsy. Co mment s by L at hers As discussed by Abu-Shaweesh (2007), activation of the laryngeal mucosa results in apnea mediated through, and elicited via, electrical stimulation of the superior laryngeal nerve. The inhibitory reflex is thought to be involved in the pathogenesis of apnea of prematurity and sudden infant death syndrome. Theophylline and block of GABA(A) receptors attenuate the inhibitory reflex. Phrenic nerve response to increasing levels of superior laryngeal nerve stimulation was examined in ventilated, vagotomized, decerebrate, and paralyzed newborn piglets. Phrenic activity decreased with increased stimulation of the superior laryngeal nerve and resulted in apnea and hypotension with higher levels of stimulation. It was concluded that activation of adenosine A (2A) receptors enhances superior laryngeal nerve stimulationinduced apnea. This may occur via a GABAergic pathway. The authors hypothesize that superior laryngeal nerve stimulation may cause endogenous release of adenosine to activate A (2A) receptors on GABAergic neurons, resulting in release of GABA at inspiratory neurons and subsequent respiratory inhibition. Case summarized from Tavee and Morris (2008).

Po st ict al Cent ral Apnea as One€Po t ent ial Mechanism fo r SUDEP A 55-s convulsive seizure occurred in a 20-year-old woman as she underwent videoEEG monitoring (So et al. 2000). Persistent apnea then developed. Electrocardiogrammonitored rhythm was not altered for the ἀrst 10 s, then it gradually and progressively slowed and stopped 57 s later. Cardiorespiratory resuscitation was successful. No evidence of airway obstruction or pulmonary edema was noted. Co mment s by S o , Sam, and Lag erlund One previous cardiorespiratory arrest after a complex partial seizure without secondary generalization had been reported for this patient. So et al. (2000) note that although epileptic seizures may be associated with arrhythmogenic actions at the

714 Sudden Death in Epilepsy: Forensic and Clinical Issues

heart, in this patient the mechanism of marked central suppression of respiratory activity after seizures was clearly involved and almost resulted in sudden death. Co mment s by L at hers Psychological factors, including stress, are risk factors for SUDEP (Schraeder and Lathers 1983; Fenwick 1994; Lathers and Schraeder 2006; Stopper et al. 2007; Lathers et al. 2008a, 2008b; Scorza et al. 2008). One must address the issue of the stress experienced by some persons with epilepsy as they undergo video-EEG monitoring, especially because monitoring protocols often include reduction or withdrawal of antiepileptic drugs to increase the likelihood of occurrence of a clinical event. The stress of having the actual procedure itself may trigger seizures and the unwanted respiratory and/ or cardiac events. Fortunately, in this patient and in the patient of Tavee and Morris (2008), successful resuscitation occurred and SUDEP appears to have been prevented. Unlike the ἀnding of Tavee and Morris (2008) discussed above, So et al. (2000) do not mention any observation of severe laryngospasm as the endotracheal tube was being inserted. We do not know if severe laryngospasm did occur in this 20-year-old woman with epilepsy and the emergency code team did not detect it and/or if it did actually occur. In the future, it will be important for emergency teams to note if a given person with known epilepsy exhibits severe laryngospasm. In any event, these case histories highlight that both respiratory and cardiac changes do occur in persons with epilepsy. The timing of events such as seizures, respiratory and/or laryngospasm, and cardiac ECG changes does vary in different patients. The physician must consider risk factors for a given individual patient and protective procedures to prescribe to protect the person from future unwanted events that may result in SUDEP. The question must be asked as to whether a person ἀrst experienced seizures and respiratory events and then cardiac events or if the person experienced seizures and arrhythmia and then respiratory events. There is documented evidence in the literature to support both cardiac and respiratory events as initiating mechanisms of sudden death. Obviously, rapid reversal of these changes is essential, and the “availability of resuscitation methods on the spot where the victim is located” certainly increases the likelihood that SUDEP will be prevented. It is also important to note that interictal discharges, just like ictal discharges, have been reported to be associated with cardiac arrhythmias and/or respiratory changes and/or sudden death (Lathers and Schraeder 1982; Schraeder and Lathers 1983). The persons administering the EEG/ECG video monitoring must watch for the appearance of both interictal and ictal discharges as warning signs of unwanted events that may subsequently be triggered. Case summarized from So et al. (2000).

44.3â•…Discussion The evidence presented in these two “near miss” sudden death patient case histories highlight two potential mechanisms of SUDEP. In one person with epilepsy, severe laryngoÂ� spasm was detected as the endotracheal tube was being inserted, suggesting that postictal laryngospasm may be one potential cause of sudden unexpected death. In another person

Arrhythmogenic, Respiratory, and Psychological Risk Factors for SUDEP

715

with epilepsy, postictal central apnea occurred and suggested that this is another potential mechanism of sudden death (So et al. 2000). Clearly, there are different mechanisms of SUDEP involved in different patients. Lathers et al. (2008a) recently summarized proposed mechanistic factors in SUDEP with the three major risk categories of arrhythmogenic factors, respiratory factors and hypoxia, and psychological factors (see Tables 1.1 through 1.8 in Lathers et al. 2010, Chapter 1). The reader is also referred to Monte et al. (2007), Lathers and Schraeder (2010, Chapter 28), Walczak (2010, Chapter 12), Tomson (2010, Chapter 51), and Chapters 3, 4, and 13–19 in this book for further discussion about the risk factors and mechanisms of sudden death in persons with epilepsy. Strong risk factors for SUDEP include being a young adult male, having generalized tonic seizures, and lying in bed. Less vigorous risk factors include being in prone posture, having subtherapeutic AED blood levels, being in a bedroom (not in bed), and having a structural brain lesion. Nilsson et al. (2001) investigated the association between clinical variables and SUDEP in an effort to identify risk factors in a Swedish population and reported that 91% of the 57 SUDEP cases studied had autopsies performed. The associated risk factors for SUDEP were: 1. Higher number of seizures per year (a relative risk of 10.16 in patients having more than 50 seizures per year compared to those with up to two per year) 2. Increased number of antiepileptic drugs (9.89 for three antiepileptic drugs vs. monotherapy) 3. Early- vs. late-onset epilepsy (7.72) 4. Frequent changes in antiepileptic drug dosage vs. unchanged dosage (6.08) 5. Male gender The study found the association between SUDEP risk and early onset, and SUDEP risk and seizure frequency was weaker for females vs. frequent dosage changes, which had a stronger association in females. This study conἀrmed early age of onset of epilepsy and male gender as risk factors (Nilsson et al. 2001). In a more recent report (McHugh and Delanty 2008), the marginally lower incidence of epilepsy and unprovoked seizures in females is attributed to the fact that males have a greater exposure to risk factors for lesional epilepsy and acute symptomatic seizure. Females have a greater incidence of idiopathic generalized epilepsies that represent approximately 15–20% of all epilepsies. Of interest, common epilepsy syndromes such as mesial temporal sclerosis may differ between the genders with isolated auras more common among females and secondary seizure spread more likely to occur in males. Gender differences are observed in the incidence of status epilepticus (more common in men), the incidence of SUDEP, prognosis, and mortality (McHugh and Delanty 2008). Bateman et al. (2008) examined the incidence and severity of ictal hypoxemia in patients with localization-related epilepsy undergoing video-EEG telemetry. They measured seizure associated oxygen desaturation and hypoventilation. Pulse oximetry revealed oxygen desaturations below 90% in 33.2% of all 304 seizure events. The degree of desaturation was signiἀcantly correlated with seizure duration and with electrographic evidence of seizure spread to the contralateral hemisphere. Central apneas or hypopneas occurred with 50% of all seizures. Ictal hypoxemia occurred often in these patients with localization related epilepsy, and may be pronounced and prolonged, even if the seizures do not progress to generalized convulsions. End-tidal carbon dioxide increase occurred in oxygen desaturation and supports the assumption that ictal oxygen desaturation is a consequence of

716 Sudden Death in Epilepsy: Forensic and Clinical Issues

hypoventilation. Both ictal hypoxemia and hypercapnia may be contributing risk factors to SUDEP occurrence. Thus, mechanisms of SUDEP are various (Tables 1.1, 1.2, and 1.4 in Chapter 1) and include central and obstructive apnea and cardiovascular, including cardiac arrhythmias. Caution must be exerted when concluding that respiratory changes are the primary mechanism of death. In some victims, SUDEP has been attributed to a mechanism of cardiac arrhythmia (Lathers and Schraeder 1982, 1990; Schraeder and Lathers 1983; Drake et al. 1993). In some cases, the arrhythmias may be triggered by cerebral events. Data from the animal studies of Lathers and Schraeder evaluated ECG records obtained in animals before, during, and after epileptogenic activity was induced. Study results monitored intrathoracic cardiac sympathetic and vagal branch discharges and showed that discharge patterns varied widely with different degrees of epileptiform discharges. The cardiac sympathetic nerve discharges were time-locked to the bilateral cortical interictal spikes induced by pentylenetetrazol (Lathers et al. 1987; Stauffer et al. 1989). During interictal discharges, this time-locked event called the lockstep phenomenon waxed and waned among all of the monitored sympathetic branches. Latency between electrocorticogram spike and discharges in the cardiac sympathetic branches ranged from 30 to 100€ms. This suggests the cortical discharges preceded the cardiac neural discharges and were probably conducted via a multisynaptic pathway within the central nervous system. Only one of ἀve cats in which vagal branches were recorded manifested this phenomenon with interictal spikes. During the lockstep phenomenon, the ECG exhibited peaking, inversion, and biphasic T waves as well as P-wave changes consisting of peaking, flattening, and biphasic conἀguration; variable Q-wave changes and widened QRS intervals suggested neurologically mediated interference in myocardial conduction time. Premature ventricular contractions occurred in one animal in association with interictal spikes. The lockstep phenomenon was also signiἀcantly correlated with variability in mean arterial blood pressure. Resting ECGs obtained from 75 persons with epilepsy were reviewed (Drake et al. 1993). Ventricular rate, P–R interval, QRS duration, and QT interval corrected for heart rate were compared with normal ECGs obtained from age-matched patients with no cardiac or neurological disease. No potentially lethal arrhythmias were found in the persons with epilepsy. Those who appeared to be in a risk category for SUDEP did have more abnormal ECGs and the ventricular rate was faster than in other epileptic persons. Those with complex partial and secondary generalized seizures exhibited a faster ventricular rate than other epileptics. The QT interval was longer in patients with complex partial seizures vs. control or other epileptics while no changes in QRS duration or P–R interval were found. The authors conclude that resting ECG has low diagnostic yield in persons with epilepsy who do not exhibit cardiac symptoms. They suggest that factors predisposing to SUDEP may be associated with a relative increase in the resting heart rate. The relative increase in the QT interval with complex partial seizures may indicate some difference in cardiac excitability. This cardiac excitability possibly may be neurally mediated and could contribute to the occurrence of SUDEP. Nei et al. (2000) noted that SUDEP is responsible for excess mortality in those with refractory epilepsy and cardiac arrhythmias during seizures may be responsible. They examined the frequency of cardiac abnormalities during seizures in this patient group. Arrhythmias, repolarization abnormalities, and PR and QTc intervals were examined in the preictal, ictal, and postictal periods for one or more seizures per patient. Thirty-nine

Arrhythmogenic, Respiratory, and Psychological Risk Factors for SUDEP

717

percent (39%) or more exhibited one or more abnormalities of rhythm and/or repolarization during or immediately after seizures. Abnormalities included asystole, atrial ἀbrillation, mild or moderate sinus arrhythmia, supraventricular tachycardia, atrial premature depolarization, ventricular premature depolarization, and bundle-branch block. The duration of seizures was longer in persons with epilepsy and cardiac abnormalities. These abnormalities may be a contributing cause to SUDEP. These data obtained from persons with epilepsy document the ἀndings that Lathers and Schraeder (1982) and Schraeder and Lathers (1983) obtained in an animal model for SUDEP. Toth et al. (2008) examined the heart rate 6 h before and after seizure in 18 patients with focal epilepsy before epilepsy surgery. Studies were done for 2–10 days and 32 seizures were recorded. Heart rate was analyzed before and after seizures. Post-seizures heart rate was increased and was even higher 3 h later. One patient exhibited an ectopic cardiac rhythm after a generalized tonic–clonic seizure and none exhibited severe peri-ictal bradycardia. It was concluded that sympathetic activity increased while parasympathetic activity decreased after seizures. Because the changes persisted for a long time and may predict fatal arrhythmia, sudden death in epilepsy may be induced by cardiac arrhythmias connected with epileptic seizures. In some patients, seizures contribute to initiation of peripheral cardiac arrhythmias that may contribute to sudden death if not corrected. Such a case in which cerebral arrhythmia influenced cardiac rhythm was reported by Almansori et al. (2006). They noted that partial seizures of temporal origin may be associated with clinically signiἀcant tachycardia or bradycardia. Both ictal bradycardia and asystole have been designated as risk factors for SUDEP. Many patients exhibit symptomatic ictal bradycardia. In this case report, a patient being monitored via video EEG telemetry was diagnosed with asymptomatic ictal bradycardia. A cardiac pacemaker was implanted. As discussed by Zaidi et al. (2000), many seizure-like attacks have a cardiovascular cause and result in a misdiagnosis of epilepsy. Early on in the diagnosis, simple noninvasive cardiovascular evaluation with a head-up tilt test (Lathers et al. 1990, 1991) and carotid sinus massage during continuous electrocardiography, electroencephalography, and blood pressure monitoring should be done to identify patients with apparent epilepsy presenting with convulsive syncopal blackouts. Linzer et al. (1994) note that syncope and seizures are often clinically indistinguishable. Use of Holter monitoring, long-term ambulatory loop electrocardiographic recording, or tilt-table studies may be done to diagnose arrhythmic or neurally mediated syncope and to eliminate misdiagnosis of these patients as persons with epilepsy and treatment with anticonvulsant medications. Correct diagnosis is necessary to prevent potentially fatal events. An incorrect incidence of SUDEP may occur if undiagnosed cardiac syncope contributes to the documented increased sudden death rate in patients with presumed epilepsy. Likewise, Akhtar (Akhtar) recommends a simple noninvasive cardiovascular evaluation to diagnose convulsive syncope in children with apparent treatment-resistant epilepsy and to assess the extent of misdiagnosis of epilepsy in children. Long QT syndrome can masquerade as epilepsy. Hunt and Tang (2005) present the case of a female presenting to the emergency department with a generalized seizure. She had been previously diagnosed with epilepsy. Her ECG demonstrated QT prolongation secondary to bradycardia. A subsequent seizure documented the events were secondary to cerebral hypoperfusion during episodes of torsades de pointes. When the long QT syndrome does masquerade as epilepsy, the correct cardiac treatment is delayed and the patient is placed in a high risk of sudden cardiac death.

718 Sudden Death in Epilepsy: Forensic and Clinical Issues

44.3.1â•…A nalysis of the Electrocardiogram Is Essential for All Patients Who Present with Seizures In spite of the high proἀle of SUDEP, evidence for speciἀc risk factors and pathophysiology is not yet ἀrmly established. Inaccurate death certiἀcation, fewer postmortem examinations, and poor incident case reporting limit the value of epidemiological data on SUDEP (Johnston and Smith 2007). The importance of case histories is to build the “bigger medical picture” to understand the mechanisms of risk factors for sudden death in persons with epilepsy. Case histories also teach physicians and caregivers how to diagnose and how to treat patients with epilepsy who present with symptoms and risk factors for sudden death to prevent the occurrence of sudden death. Ultimately, we anticipate that the best medical/surgical practices will emerge to provide the best quality of life for those individuals who are at risk for SUDEP.

References Abu-Shaweesh, J. M. 2007. Activation of central adenosine A(2A) receptors enhances superior laryngeal nerve stimulation-induced apnea in piglets via a GABAergic pathway. J Appl Physiol 103 (4): 1205–1211. Akhtar, M. J. 2002. All seizures are not epilepsy: Many have a cardiovascular cause. J Pak Med Assoc 52 (3): 116–120. Almansori, M., M. Ijaz, and S. N. Ahmed. 2006. Cerebral arrhythmia influencing cardiac rhythm: A case of ictal bradycardia. Seizure 15 (6): 459–461. Bateman, L. M., C. S. Li, and M. Seyal. 2008. Ictal hypoxemia in localization-related epilepsy: Analysis of incidence, severity and risk factors. Brain 131 (Pt 12): 3239–3245. Drake, M. E., C. R. Reider, and A. Kay. 1993. Electrocardiography in epilepsy patients without cardiac symptoms. Seizure 2 (1): 63–65. Fenwick, P. 1994. The behavioral treatment of epilepsy generation and inhibition of seizures. Neurol Clin 12 (1): 175–202. Hunt, D. P., and K. Tang. 2005. Long QT syndrome presenting as epileptic seizures in an adult. Emerg Med J 22 (8): 600–601. Jim, K. F., C. M. Lathers, V. L. Farris, L. F. Pratt, and W. H. Spivey. 1989. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous lorazepam in pigs. Epilepsia 30 (4): 480–486. Jim, K. F., C. M. Lathers, W. H. Spivey, W. D. Matthews, C. Kahn, and K. Dolce. 1988. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous propranolol in pigs. J Clin Pharmacol 28 (12): 1106–1111. Johnston, A., and P. Smith. 2007. Sudden unexpected death in epilepsy. Expert Rev Neurother 7 (12): 1751–1761. Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68 (2): 211–213. Lathers, C. M., K. F. Jim, W. B. High, W. H. Spivey, W. D. Matthews, and T. Ho. 1989a. An investigation of the pathological and physiological effects of intraosseous sodium bicarbonate in pigs. J Clin Pharmacol 29 (4): 354–359. Lathers, C. M., K. F. Jim, and W. H. Spivey. 1989b. A comparison of intraosseous and intravenous routes of administration for antiseizure agents. Epilepsia 30 (4): 472–479. Lathers, C. M., N. Tumer, and J. M. Schoffstall. 1989c. Plasma catecholamines, pH, and blood pressure during cardiac arrest in pigs. Resuscitation 18 (1): 59–74. Lathers, C. M., P. H. Diamandis, J. M. Riddle, C. Mukai, K. F. Elton, M. W. Bungo, and J. B. Charles. 1990. Acute and intermediate cardiovascular responses to zero gravity and to fractional gravity levels induced by head-down or head-up tilt. J Clin Pharmacol 30 (6): 494–523.

Arrhythmogenic, Respiratory, and Psychological Risk Factors for SUDEP

719

Lathers, C. M., P. H. Diamandis, J. M. Riddle, C. Mukai, K. F. Elton, M. W. Bungo, and J. B. Charles. 1991. Orthostatic function during a stand test before and after head-up or head-down bed rest. J Clin Pharmacol 31 (10): 893–903. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008a. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Neurocardiologic mechanistic risk factors in sudden unexpected death in epilepsy. Chapter 1 in Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M.€W.€Bungo, and J. Leestma. Boca Raton: CRC Press. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Chapter 13. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher. Hauppauge, NY: Nova Science. Lathers, C. M., and P. L. Schraeder. 1990. Chapter 9. Arrhythmias associated with epileptogenic activity elicited by penicillin. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 2010. Antiepileptic drugs beneἀt/risk clinical pharmacology: Possible role in cause and/or prevention of SUDEP. Chapter 37 in Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Linzer, M., B. P. Grubb, S. Ho, L. Ramakrishnan, E. Bromἀeld, and N. A. Estes 3rd. 1994. CardioÂ� vascular causes of loss of consciousness in patients with presumed epilepsy: A cause of the increased sudden death rate in people with epilepsy? Am J Med 96 (2): 146–154. McHugh, J. C., and N. Delanty. 2008. Epidemiology and classiἀcation of epilepsy: Gender comparisons. Int Rev Neurobiol 83: 11–26. Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16 (1): 1–7. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42 (5): 667–673. Schoffstall, J. M., W. H. Spivey, S. Davidheiser, and C. M. Lathers. 1989. Intraosseous crystalloid and blood infusion in a swine model. J Trauma 29 (3): 384–387. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13 (1): 263–264; author reply 265–269. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41 (11): 1494–1497. Spivey, W. H., H. D. Unger, C. M. Lathers, and R. M. McNamara. 1987a. Intraosseous diazepam suppression of pentylenetetrazol-induced epileptogenic activity in pigs. Ann Emerg Med 16 (2): 156–159.

720 Sudden Death in Epilepsy: Forensic and Clinical Issues Spivey, W. H., H. D. Unger, R. M. McNamara, M. M. LaManna, T. Ho, and C. M. Lathers. 1987b. The effect of intraosseous sodium bicarbonate on bone in swine. Ann Emerg Med 16 (7): 773–776. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Stopper, M., T. Joska, M. M. Burg, W. P. Batsford, C. A. McPherson, D. Jain, and R. Lampert. 2007. Electrophysiologic characteristics of anger-triggered arrhythmias. Heart Rhythm 4 (3): 268–273. Tavee, J., and H. Morris 3rd. 2008. Severe postictal laryngospasm as a potential mechanism for sudden unexpected death in epilepsy: A near-miss in an EMU. Epilepsia 49 (12): 2113–2117. Tomson, T. 2010. Compliance with antiepileptic drug treatment and the risk of sudden unexpected death in epilepsy. Chapter 51 in Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Toth, V., L. Hejjel, Z. Kalmar, A. Fogarasi, T. Auer, C. Gyimesi, A. Szucs, and J. Janszky. 2008. Effect of epileptic seizures on the heart rate. Ideggyogy Sz 61 (5–6): 155–161. Walczak, T. 2010. Risk factors for sudden death in epilepsy. Chapter 12 in Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Zaidi, A., P. Clough, P. Cooper, B. Scheepers, and A. P. Fitzpatrick. 2000. Misdiagnosis of epilepsy: Many seizure-like attacks have a cardiovascular cause. J Am Coll Cardiol 36 (1): 181–184.

Sudden Arrhythmic Death Syndrome Underlying Cardiac Etiologies, Their Implications, and the Overlap with SUDEP

45

Paramdeep S. Dhillon Elijah R. Behr

Contents 45.1 Introduction 45.2 Illustrative Case 45.3 SADS: A Deἀnition and Characteristics 45.3.1 Etiology 45.3.1.1 The Channelopathies 45.3.2 Structural Heart Disease 45.3.2.1 Arrhythmogenic Right Ventricular Cardiomyopathy 45.3.2.2 Hypertrophic Cardiomyopathy 45.3.2.3 Other Structural Cardiac Diseases 45.4 The Familial Basis of SADS 45.5 Non-Mendelian Basis to SADS? 45.6 The Overlap between SUDEP and SADS 45.7 Conclusion References

721 722 724 724 724 730 731 731 732 733 733 733 735 735€

45.1â•…Introduction Approximately 300,000 adults die suddenly each year in the United States (Virmani et al. 2001) mainly as the result of fatal ventricular arrhythmias related to coronary artery and structural heart disease (Davies and Thomas 1984). Unexplained sudden death in children, adolescents, and young adults are much less common with an incidence of between 1.3 and 8.5 deaths per 100,000 patient-years (Liberthson 1996) equivalent to approximately 500 deaths annually in the UK (Behr et al. 2007), or 2500 deaths per annum in the United States. Cardiac fatalities in this younger group of individuals are caused predominantly by cardiomyopathic disease, although other heart diseases are recognized (Table 45.1) (Corrado et al. 2001; Fabre and Sheppard, 2006). Up to 30% of deaths remain unexplained despite a thorough postmortem examination and toxicology screen (Behr et al. 2003; Chugh et al. 2000; Corrado et al. 2001; Maron et al. 1996; Morentin et al. 2003; Puranik et al. 2005). Although the terms sudden adult death syndrome and sudden death syndrome have been used to describe unexpected death occurring within 1 h of symptoms (de la Grandmaison 721

722 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 45.1â•… Principal Causes of Sudden Death in Young Adults That May Be Identified at Autopsy Toxicology (Causes with a Predominant Genetic Basis Are Marked with Asterisks) Cardiomyopathy

Coronary artery disease Infective and inἀltrative Valvular heart disease Conduction disease Congenital heart disease Connective tissue disease Illicit drug use Unknown cause

Hypertrophic cardiomyopathy* ARVC* Dilated cardiomyopathy* Left ventricular non-compaction* Glycogen storage disease and mitochondrial myopathy* Premature coronary artery disease* Anomalous coronary arteries Focal myocarditis Cardiac sarcoidosis Mucoid degeneration of the mitral valve Lev-Lenègre’s disease Wolf–Parkinson–White syndrome Various Marfan’s syndrome* Cocaine, amphetamines Idiopathic ἀbrosis Idiopathic left ventricular hypertrophy

and Durigon 2002), the term sudden arrhythmic death syndrome (SADS) is now preferred to describe unexpected and unheralded deaths in adults as well as children (excluding infants) where no deἀnite cause of death can be found at postmortem (Behr et al. 2003; Bowker et al. 2003). Familial evaluation studies have shown that up to half of sudden unexplained deaths in young adults may have been the result of hereditary cardiac conditions including ion channel disease (Behr et al. 2003, 2008; Tan et al. 2005). These diseases are associated with mutations in genes coding for proteins that form or interact with cardiac ion channels that conduct sodium, potassium, and calcium ions (Ackerman 2004). The generation and propagation of the cardiac action potential are then disturbed, predisposing those affected to cardiac arrhythmias (Marban 2002). Epileptics are also prone to sudden unexpected death (Morentin and Alcaraz 2002), despite a normal heart at autopsy, and epilepsy-related cardiac rhythm abnormalities (Al-Hillawi et al. 1984; Lim et al. 1989; Rocamore et al. 2003; Tigaran et al. 2002) may offer a potential explanation to this phenomenon. In addition, ventricular arrhythmias and asystole may cause symptoms that mimic seizure disorders raising the possibility that the patient is not epileptic. In a recent study of patients diagnosed with the congenital long QT syndrome (LQTS), 7% of patients were at some stage managed with antiepileptic drugs (Johnson et al. 2009). The purpose of this chapter, therefore, is to describe the features of SADS and the underlying etiologies. In particular, we explore whether there are potential overlaps with the phenomenon of sudden unexpected death in epilepsy (SUDEP) and whether there may be common underlying mechanisms.

45.2â•…Illustrative Case A 25-year-old woman was referred by her neurologist. From the age of 15, she had suffered recurrent episodes of witnessed loss of consciousness. She would become vacant and

Sudden Arrhythmic Death Syndrome

723

unresponsive before falling to the floor and becoming rigid and motionless. She would then develop limb twitching, hyperventilation, and was unresponsive for approximately 1 min. On one occasion she bit her tongue and was incontinent of urine. There was no family history of epilepsy or personal history of neonatal trauma or febrile convulsion, although a male cousin had died in his sleep unexpectedly at age 32. She was diagnosed with complex partial and secondary generalized epilepsy but failed to respond to a variety of medications including lamotrogine, levitiracetam, and subsequently oxcarbazepine. Standard and sleep-deprived electroencephalography and brain magnetic resonance imaging did not demonstrate an abnormality. A 12-lead electrocardiogram (ECG) performed while on levitiracetam demonstrated QT interval prolongation with broad-based and flattened T-wave morphology that persisted despite discontinuation of the drug (Figure 45.1a). Apart from clear repolarization abnormality, all other cardiac (a)

(b)

Figure 45.1╇ (a) Twelve-lead ECG demonstrating QT interval prolongation with a broad-based and flattened T-wave morphology. (b) Polymorphic ventricular tachycardia.

724 Sudden Death in Epilepsy: Forensic and Clinical Issues

investigations were within normal limits. Further familial evaluation revealed that her mother had suffered a single seizure as a young women after childbirth. An ECG demonstrated mild QT interval prolongation. An implantable internal-loop recorder was able to demonstrate polymorphic ventricular tachycardia (Figure 45.1b) that correlated with her next episode of loss of consciousness. She was therefore diagnosed with the congenital LQTS and commenced on propranolol and her antiepileptic medication discontinued.

45.3â•… SADS: A Definition and Characteristics SADS is an umbrella term for unexpected and unexplained sudden death and has been deἀned for research purposes as a sudden death, age 4–64 years, last seen alive and well within 12 h of being found dead, no prior recorded cardiac disease, a normal full coroner’s postmortem, negative toxicology results, and a normal expert cardiac pathologist’s examination (Behr et al. 2003, 2008). Most of the victims are male, have antecedent symptoms (particularly syncope), and die in their sleep. A family history of sudden death can be elicited in 30% of cases (Behr et al. 2003, 2008). 45.3.1â•… Etiology The underlying etiologies of SADS, where identiἀable, have been shown to be predominantly inherited cardiac diseases. In particular, “molecular autopsy,” the postmortem study of unexplained sudden death victims, has established the presence of cardiac ion channel diseases or “channelopathies” (Tester and Ackerman 2007). In a British study of 57 families of SADS victims, such a diagnosis was established in 30 (53%) families after a comprehensive investigative algorithm (Figure 45.2) (Behr et al. 2008). This was an improvement over prior research from the same group that detected inherited heart disease in 7 of 32 families (22%) but relied upon a much more limited investigative protocol (Behr et al. 2003). Channelopathies were the most common cause, but subtle structural heart disease was also noted, in particular arrhythmogenic right ventricular cardiomyopathy (ARVC). Similar results were described in a Dutch study of 43 families with a high frequency of unexplained sudden deaths that showed that 17 (40%) deaths were due to inherited cardiac diseases (Tan et al. 2005). The main causes of SADS are shown in Figure 45.3. 45.3.1.1â•… The Channelopathies The cardiac ion channel diseases have several common features: a genetic etiology; an absence of structural disease; and the risk of arrhythmias causing syncope, seizures, and/ or sudden death. This group of diseases include LQTS, Brugada syndrome (BrS), and catecholaminergic polymorphic ventricular tachycardia that have been recognized as inherited causes of SADS (Behr et al. 2008; Tan et al. 2005; Tester and Ackerman 2006). Other more rare entities such as short QT syndrome (SQTS), progressive cardiac conduction defect (PCCD), and the recently described “early repolarization syndrome” may also play a role but have yet to be formally described in this group. 45.3.1.1.1â•… ἀ e Long QT Syndromeâ•… Congenital LQTS affects 1 in 5000 persons and is characterized by delayed myocardial repolarization that may manifest itself on the surface

Sudden Arrhythmic Death Syndrome

725

Proband evaluation: Mutation analysis in probands: “the molecular autopsy”

Sudden death Normal coroner’s autopsy Negative toxicology ± Normal expert pathologist’s assessment

NB concurrent process without delay to familial evaluation

Familial evaluation Initial relative evaluation: Historical assessment Physical examination Resting ECG 24 h ECG Exercise ECG Echocardiogram Normal heart

Normal ECG

Ajmaline test

Abnormal or equivocal cardiac morphology

Right ventricular lead repolarization changes

Cardiac MRI

Inherited heart disease diagnosed? Mutation analysis in relatives: If a relative is diagnosed or suspected of carrying an inherited cardiac disease, proceed to mutation analysis. If an unequivocal mutation is identified, offer familial cascade genetic testing and exclude noncarriers. If an equivocal mutation is detected, offer familial cascade testing and clinical evaluation.

Figure 45.2╇ An algorithm for investigating families with a SADS death. (Adapted from Behr, E. R., et al., Eur Heart J, 29, 1670–1680, 2008.)

ECG as QT interval prolongation and T-wave abnormalities (Figure 45.1a). Early afterdepolarizations may trigger ventricular arrhythmias (Roden 1993), speciἀcally polymorphic ventricular tachycardia (torsades de pointes) (Figure 45.1b) and ventricular ἀbrillation. If these are self-limiting, they result in syncope and/or secondary anoxic seizures, but if sustained, they result in sudden death in the setting of a structurally normal heart (Keating and Sanguinetti 2001). The Schwartz score (Schwartz et al. 1993) encompasses the accepted clinical criteria for the diagnosis of LQTS and relies on phenotypic features of risk such as the degree of QT

726 Sudden Death in Epilepsy: Forensic and Clinical Issues LQTS — definite 23%

(a) No diagnosis 47%

LQTS — possible 5%

Brugada Syndrome 9%

HCM 2% (b)

DCM 2%

LV noncompaction

ARVC 9%

3%

LQTS 21%

CPVT 14%

No diagnosis 65%

Figure 45.3╇ The causes of sudden arrhythmic death syndrome. CPVT, catecholaminergic

polymorphic ventricular tachycardia; LQTS, long QT syndrome; ARVC, arrhythmogenic right ventricular cardiomyopathy; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomyopathy. (a) Adapted from Behr et al., Eur Heart J, 29, 1670–1680, 2008. (b) Adapted from Tester, D.€J., and Ackerman, M. J., J Am Coll Cardiol, 49, 240–246, 2007.

Sudden Arrhythmic Death Syndrome

727

prolongation, other ECG abnormalities, documented ventricular arrhythmia, prior syncope, prior cardiac arrest, and a family history of sudden death and/or LQTS. A score of ≥4 is highly speciἀc for LQTS in an index case, although its sensitivity is low, particularly in relatives of the index case (Hofman et al. 2007). This is because LQTS is a genetically heterogeneous disease with variable penetrance, meaning that some patients demonstrate little or no QT prolongation on the 12-lead ECG, the forme fruste (Priori et al. 1999), until an additional event such as exposure to a QT-prolonging drug or electrolyte imbalance occurs (Yang et al. 2002). The natural history of LQTS patients varies greatly from sudden death in childhood to asymptomatic longevity even among members of the same family (Ackerman and Clapham 1997). The more common form of LQTS (99%) is autosomal dominant and known as the Romano–Ward syndrome (Romano et al. 1963; Ward 1964). The rarer autosomal recessive form, known as the Jervell and Lange–Nielsen syndrome, is associated with sensorineural deafness (Jervell and Lange-Nielsen 1957). The genetic basis of LQTS has become better understood over the past 15 years with the identiἀcation of disease-Â�associated mutations in 12 different genes linked to cardiac repolarization and depolarization (Table 45.2). These mutations lead either to a net reduction or “loss-of-function” of outward potassium rectifying currents (KCNQ1, KCNH2, KCNE1, KCNJ2, and KCNE2) or a net increase or “gainof-function” in the inward sodium current (SCN5A) or calcium currents (CACNA1C) (Noseworthy and Newton-Cheh 2008; Rodriguez-Calvo et al. 2008). Mutations of the ἀrst ἀve identiἀed genes (KCNQ1, KCNH2, SCN5A, KCNE1, and KCNE2) account for up to 70% of unrelated deἀnite LQTS probands undergoing genetic testing, with KCNQ1- and KCNH2-associated LQTS (LQT1 and LQT2, respectively) accounting for approximately 40–45% of genotyped patients each and SCN5A-associated disease (LQT3) being linked in 8–15% (Splawski et al. 2000; Tester et al. 2005). Also recognized are mutations in cellular

Table 45.2â•… Congenital LQTS: Subtypes, Their Genetic Basis, and Frequencies Subtype

Chromosome

Gene

Product

Population Frequency

Long QT syndrome 1 Long QT syndrome 2 Long QT syndrome 3 Long QT syndrome 4

11 7 3 4

KCNQ1 KCNH2 SCN5A ANKB

~30% ~30% ~5% ~2%

Long QT syndrome 5 Long QT syndrome 6 Long QT syndrome 7 Long QT syndrome 8 Long QT syndrome 9 Long QT syndrome 10 Long QT syndrome 11 Long QT syndrome 12

21 21 23 12 7 11 7 20

KNCE1 KNCE2 KCNJ2 CACNA1c CAV3 SCN4B AKAP9 SNTA1

IKs α subunit IKr α subunit INa α subunit Submembrane component IKs β subunit IKr β subunit IK1 Cav1.2 Caveola component INa β subunit AKAP9/yotiao a1-syntrophin

~2% ~2% – – ~2% – – –

Sources: Chen, L., et al., Proc Natl Acad Sci U S A, 104 (52), 20990–20995, 2007. Hofman, N., et al. Eur Heart J, 28 (11), 1399, 2007. Splawski, I., et al., Circulation, 102 (10), 1178–1185, 2000. Tester, D. J., et al., Heart Rhythm, 2 (5), 507–517, 2005. Ueda, K., et al., Proc Natl Acad Sci U S A, 105 (27), 9355–9360, 2008. Note: IKs, slow rectifying potassium current; IKr, rapid rectifying potassium current; INa, inward sodium current; IK1, Kir 2.1 inward rectifying current; Cav1.2, L-type calcium channel current.

728 Sudden Death in Epilepsy: Forensic and Clinical Issues

proteins required for cellular localization of ion channels, otherwise known as channel interacting proteins such as ankyrinB and caveolin3 (Rodriguez-Calvo et al. 2008). The incidence of sudden death or resuscitated cardiac arrest from birth to age 40 years ranges between 30% for KCNQ1 mutations and 46% in KCNH2 mutations (Priori et al. 2003). Interestingly, several phenotypical variations in terms of triggers for ventricular arrhythmias have been described. Swimming and exertion-related arrhythmias are associated with the LQT1 phenotype (Schwartz et al. 2001). In contrast, the LQTS3 phenotype is associated with cardiac events predominantly during sleep and rest. Auditory triggers may precipitate cardiac events, which also occur in the period after childbirth in LQT2 patients (Moss et al. 2002). Risk assessment of LQTS patients is predominantly based on symptoms (in particular syncope and prior cardiac arrest), gender, and the length of the corrected QT interval; a corrected QT interval >500 ms on the resting ECG carrying a higher risk (Hobbs et al. 2006; Sauer et al. 2007). The LQT1 genotype has been thought to be less hazardous than LQT2 and LQT3 (Priori et al. 2003), but the latest evidence from the international long QT registry does not support the utility of genetic locus as a marker for risk apart from in female adult LQT2 carriers (Goldenberg and Moss 2008). Beta-blocker therapy appears to be effective for patients with the LQT1 and LQT2 phenotypes, but is less effective for LQT3 patients. An implantable cardiac deἀbrillator (ICD) is indicated for survivors of cardiac arrest and those deemed particularly high risk for sudden death (Goldenberg et al. 2008). Surgical left cervicothoracic sympathetic denervation has been shown to reduce the mean annual number of LQTS-related cardiac events and ICD discharges in high-risk patients (Schwartz et al. 2004). 45.3.1.1.2â•… ἀ e Brugada Syndromeâ•… Brugada syndrome is an inherited condition that is characterized by the presence of coved ST elevation and “J point” elevation of at least 2 mm in at least two of the right precordial ECG leads (“type-1” ECG) (Figure 45.4) in the absence of cardiac structural disease (Antzelevitch et al. 2005; Brugada and Brugada 1992). The prevalence of the Brugada ECG pattern is between 0.05% in Europe and 0.6% in Japan (Hermida et al. 2000; Matsuo et al. 2004) and appears to increase with age (Brugada et al. 2001). BrS is thought to be the same disease as the “sudden unexpected nocturnal death syndrome” in Southeast Asia, also known as “Pokkuri” in Japan and “Lai Tai” in Thailand. It is characterized by ventricular arrhythmias, conduction abnormality, atrial arrhythmias, syncope, and sudden death. The diagnosis in an individual requires the presence of the Brugada ECG pattern with at least one of the recognized diagnostic criteria: syncope, prior cardiac arrest, documented or inducible polymorphic ventricular tachycardia or ventricular ἀbrillation, a family history of sudden death <45 years old, or type 1 Brugada pattern and/or nocturnal agonal respiration (Antzelevich et al. 2005). In some carriers, the 12-lead ECG may appear unremarkable with only transient occurrence of the type 1 Brugada pattern or the need for provocation testing with class I antiarrhythmics such as ajmaline, flecainide, procainamide, and pilsicainide to reveal it (Antzelevich et al. 2005). Thus, BrS exhibits variable expressivity and reduced penetrance. In addition, there are “mixed phenotypes,” where families have been described containing members with the BrS phenotype as well as sinus node disease, LQTS, SQTS, and/or progressive cardiac conduction defect (Hedley et al. 2009). BrS is an autosomal dominant disease and it has been associated with more than 100 mutations in seven different genes (Hedley et al. 2009). The most commonly affected gene

Sudden Arrhythmic Death Syndrome

729

Figure 45.4╇ Twelve-lead electrocardiogram demonstrating type 1 Brugada pattern.

is SCN5A encoding the α-subunit of the cardiac sodium channel Nav1.5 responsible for the inward sodium current INa. It is found in approximately 20% of cases and causes lossof-function by influencing the trafficking or the gating function of the sodium channel (Antzelevich 2007). The other genes involved account for smaller numbers and also appear to diminish the inward sodium and calcium currents (Table 45.3). Arrhythmias occur nocturnally, and at rest, but may also be triggered by fever and sodium channel blocking drugs such as class I antiarrhythmics and tricyclic antidepressants. Patients have a good prognosis if they are asymptomatic and their ECG requires provocation with a sodium channel blocker to reveal the BrS ECG pattern. Survivors of€cardiac arrest or those who have syncopal episodes are at a higher risk of recurrence or sudden death, but the mortality rate may reach 10% per annum (Brugada et al. 2003) Table 45.3â•… Gene Mutations Associated with BrS Chromosome 3 11 19 11 10 12 3

Gene

Protein

Product

SCN5A SCN3B SCN1B KCNE3 CACNB2 CACNA1C GPD1L

Nav1.5 Navβ3 Navβ MiRP2 Cavβ2 Cav1.2 G3PD1L

INa α subunit INa β subunit INa β subunit IKs/Ito β subunit ICa β subunit ICa α subunit INa α subunit interacting protein

Source: Hedley, P. L., et al., Hum Mutat 30 (9), 1256–1266, 2009. Note: INa, inward sodium current; IKs, slow rectifying potassium current; Ito, transient outward potassium current; ICa, L-type calcium channel current.

730 Sudden Death in Epilepsy: Forensic and Clinical Issues

whether symptomatic or not, and the only therapy that can reduce mortality is the ICD. Avoidance of sodium channel inhibitors and treatment of fever may reduce symptoms (Riera et al. 2007). 45.3.1.1.3â•… Catecholaminergic Polymorphic V entricular T achycardiaâ•… Catecholaminergic polymorphic ventricular tachycardia is an inherited disease with two genetic variants: an autosomal dominant trait caused by mutations in the ryanodine receptor (RyR2) gene (Laitinen et al. 2001; Priori et al. 2001) and a recessive form caused by mutations in the cardiac-speciἀc isoform of the calsequestrin gene (CASQ2) (Lahat et al. 2001). Mutations in these genes may cause CPVT by altering cellular calcium handling resulting in delayed afterdepolarization, a mechanism of triggered arrhythmias (Lehnart et al. 2004). Clinically, CPVT is characterized by exertion-induced polymorphic ventricular tachycardia (including “bidirectional VT” with alternating QRS axis on a beat-to-beat basis), ventricular ἀbrillation, syncope, and sudden death (Leenhardt et al. 1995). CPVT is particularly associated with swimming-triggered cardiac events and is highly lethal with untreated mortality rates of 30–50% by the age of 40 years (Choi et al. 2004). Beta-blockers can prevent arrhythmias (Kontula et al. 2005), but ICD therapy is recommended for those with life-threatening arrhythmias despite beta-blockade (Priori et al. 2002). 45.3.1.1.4â•…O thersâ•… Lev-Lenègre’s disease, or progressive cardiac conduction disease, is a rare condition that is associated with a defect in the SCN5A gene in some families (Schott et al. 1999). It may be associated with bradyarrhythmias requiring pacemaker implantation. The autosomal dominant SQTS (Gussak et al. 2000) is characterized by a short QT interval (corrected QT interval <300 ms) on the 12-lead ECG, with tall symmetric peaked T waves. Patients may present with palpitations, syncope, arrhythmias, and sudden death. The syndrome is attributed to gain-of-function mutations in potassium channel genes KCNH, KCNQ1, and KCNJ2 (Kir2.1, SQTS3) (Bellocq et al. 2004; Brugada et al. 2004; Priori et al. 2005), which shorten myocardial refractoriness. Quinidine has been reported to be effective in prolonging the QT interval in SQTS patients (Gaita et al. 2004), but ICD therapy is more effective (Yontar et al. 2008). Early repolarization ECG changes, represented by elevation of the J point (deflection at the QRS–ST junction) in the inferior and/or lateral leads, is more frequently associated with idiopathic ventricular ἀbrillation than age-matched controls (31% vs. 5%) (Haissaguerre et al. 2008). The underlying cellular mechanisms are not yet fully understood, but a mutation in the KCNJ8 gene has been recently associated with this syndrome (Haissaguerre et al. 2009). There are as yet no speciἀc treatments available apart from ICD implantation. 45.3.2â•… Structural Heart Disease SADS may also result from inherited structural heart diseases that are so subtle they are not noted at initial autopsy or even at expert autopsy. This has been described partÂ�icularÂ�ly€with ARVC, which can present in an early, concealed phase as sudden death but has also been described in cases where families have subsequently been diagnosed with hypertrophic cardiomyopathy, dilated cardiomyopathy, and left ventricular “nonÂ�compaction” (Behr et al. 2008; Tan et al. 2005). Atrioventricular accessory pathways responsible for the Wolff– Parkinson–White syndrome may cause sudden death because of rapid conduction of atrial arrhythmias resulting in ventricular ἀbrillation (Klein et€al.€1979).

Sudden Arrhythmic Death Syndrome

731

45.3.2.1â•… Arrhythmogenic Right Ventricular Cardiomyopathy The prevalence of ARVC varies from between 1 in 2000 to 5000, affecting predominantly men. It is characterized pathologically by progressive ἀbro-fatty replacement of the right ventricular myocardium with a predilection for the apical, inflow, and infundibular regions of the right ventricle; the so-called triangle of dysplasia. The phenotype may vary from widespread disease characterized by global right ventricular dilatation, wall thinning and aneurysms, to regional lesions with normal anatomy (Nava et al. 1992). ARVC commonly demonstrates incomplete penetrance and variable expression. It is becoming apparent that these changes may be seen in the left ventricle as well (Gallo et al. 1992). It is also clear that there may be difficulties in making a pathological diagnosis as epicardial fat without ἀbrosis is a common ἀnding in the Western world and not diagnostic of the condition (Bomma et al. 2004). ARVC is predominantly an autosomal dominant disease of cellular adhesion, affecting the desmosome components: junction plakoglobin (JUP); desmoplakin (DSP); plakophilin-2 (PKP2); desmoglein-2 (DSG2); and desmocollin-2 (DSC-2) (Gerull et al. 2004; McKoy et al. 2000; Rampazzo et al. 2002; Syrris et al. 2006). ARVC has also been linked to other genes unrelated to cell adhesion complex, such as the gene encoding for cardiac ryanodine receptor (RyR2) (Tiso et al. 2001), transmembrane protein 43 (TMEM43) (Merner et al. 2008), and the transforming growth factor-b3 gene (TGFb3) (Beffagna et al. 2005). The clinical diagnosis of ARVC in an index case requires that a number of established “task force” criteria are met: 12-lead ECG depolarization and repolarization abnormalities (Figure 45.5); structural abnormalities detected by cardiac imaging and biopsy histopathology; family history of sudden death and ARVC; the presence of symptoms and documented ventricular arrhythmias (McKenna et al. 1994). The natural history of ARVC is characterized by the progressive nature of the pathological process. This predisposes the carrier to ventricular arrhythmias at any time during the disease course, and myocardial loss results in ventricular dysfunction and heart failure. Ventricular arrhythmias can be managed with a combination of antiarrhythmics and ablation therapy, but if syncope, hemodynamically unstable ventricular arrhythmias or cardiac arrest are documented and/or disease is widespread with signiἀcant left ventricular involvement, then ICD therapy is recommended (Corrado et al. 2001). Progression to severe heart failure is uncommon but requires conventional optimal medical therapy. 45.3.2.2â•… Hypertrophic Cardiomyopathy Hypertrophic cardiomyopathy is a clinically heterogeneous, but relatively common, autosomal dominant genetic disease with a prevalence of 1:500 (Maron et al. 1995). The basis of the disease is mutations within genes encoding the contractile sarcomere, with more than 450 mutations 13 genes already described (Alcalai et al. 2008). Mutations that are most common include the genes encoding the beta-cardiac myosin heavy chain and the cardiac myosin binding protein C. Less common are mutations in cardiac troponoin T and I genes, as well as alpha-tropomyosin. Histopathological hallmarks include myocyte hypertrophy, myocardial disarray, increased interstitial ἀbrosis, and small vessel disease, or arterial dysplasia (Maron 2002). Electrophysiological abnormalities include prolonged refractoriness (Peters et al. 2000; Savelieva et al. 1999) and slow myocardial conduction velocity (Schumacher et al. 2005), both of which are prerequisites for reentrant ventricular arrhythmias.

732 Sudden Death in Epilepsy: Forensic and Clinical Issues

Figure 45.5╇ Twelve-lead electrocardiogram from a patient with arrhythmogenic right ventricular cardiomyopathy demonstrating T-wave abnormalities across the precordial leads.

Although sudden death may be the mode of presentation in some patients, many remain asymptomatic throughout the natural course of their lives. There is a great deal of phenotypical variability even within the same family such that the degree of hypertrophy and symptoms may differ greatly. Hypertrophic cardiomyopathy has an annual mortality rate of up to 2–6% in tertiary referral centers where the sickest patients are treated (McKenna and Camm 1989; McKenna and Deanἀeld 1984; Shah et al. 1974) but has a more benign prognosis in the wider community (Cannan et al. 1995; Kofflard et al. 1993). ICD therapy is the most effective form of therapy for ventricular arrhythmias. 45.3.2.3â•… Other Structural Cardiac Diseases Heriditary dilated cardiomyopathy, characterized by enlargement of the cardiac chambers, may be caused by a variety of autosomal dominant and recessive and X-linked conditions (Burkett and Hershberger 2005). There is a subgroup of patients with coexistent conduction system disease who are at high risk of ventricular arrhythmias and carry mutations of the LMNA gene encoding for the inner nuclear membrane protein lamin A/C (MacLeod et al. 2003). Ventricular noncompaction is an uncommon cardiomyopathy affecting mainly the left ventricle, and is characterized by a hypertrophy associated with deep trabeculations and with diminished systolic function, with or without associated ventricular dilation

Sudden Arrhythmic Death Syndrome

733

(Markiewicz-Loskot et al. 2006). Arrhythmias may occur, and it has been identiἀed as a cause of SADS (Behr et al. 2008).

45.4â•… The Familial Basis of SADS Most genetic heart diseases show autosomal dominant inheritance, meaning that the probability of other family members being affected is high. However, marked clinical heterogeneity and incomplete penetrance are common, which increases the difficulty of making a diagnosis. Familial evaluation after a SADS death requires a multidisciplinary approach to examination of the whole family that includes experienced cardiologists, clinical geneticists, pathologists, and genetic counselors (Ingles and Semsarian 2007), and an appropriate investigative algorithm is shown in Figure 45.2 (Behr et al. 2008). Attention should be focused on the patient’s symptoms, including syncope and seizures, as well as any additional members who have died unexpectedly. There is therefore a potential role for the neurologist with expertise in epilepsy in such a team. Once disease is diagnosed in a family, a diagnosis is provided for the SADS victim’s cause of death that in itself can help their grieving process. In addition, one can identify other family members at risk of sudden death and institute preventative measures. In one series, 51% required drug therapy and 11% an ICD implant (Behr et al. 2008).

45.5â•…Non-Mendelian Basis to SADS? There is a strong Mendelian genetic contribution to SADS as already described. There are, however, other less well described, and subtle, genetic variants that are more common in the general population (single nucleotide polymorphisms) and may predispose to sudden death (Noseworthy and Newton-Cheh 2008). One example is the SCN5A variant S1102Y, which has a 13.2% allelic frequency in the African-American general population and has been associated with up to an 8-fold increase in the risk of ventricular arrhythmias and sudden death and SADS (Burke et al. 2005; Splawski et al. 2002). A 30% increase in the risk of sudden death associated with a speciἀc SNP in the nitric oxide synthase 1 adaptor proÂ� tein gene, NOS1AP, has also been described in the white U.S. population (Kao et al. 2009). These variants alone are insufficient to cause arrhythmias and sudden death, but rather an additional “second-hit” to cardiac repolarization such as other functionally signiἀcant variants, ischemia, electrolyte disturbance, or pro-arrhythmic medication is required for arrhythmogenesis (Noseworthy and Newton-Cheh 2008).

45.6â•… The Overlap between SUDEP and SADS The possibility that the phenomenon of SUDEP overlaps with that of SADS is an intriguing concept given that both are unexplained at autopsy and arrhythmias may potentially link epilepsy to sudden death. Indeed, in one SADS autopsy series, 1 of 39 patients had epilepsy as a concurrent diagnosis and was recorded as an epilepsy-related death (Behr et al. 2007). The overlap may be a result of one or more of the following: (1) seizures may be arrhythmogenic and expose underlying cardiac disease that predisposes to sudden death;

734 Sudden Death in Epilepsy: Forensic and Clinical Issues

(2) epilepsy and/or antiepileptic medication may predispose individuals to sudden death without a seizure; (3) coexistent cardiac disease may cause secondary anoxic seizures; (4)€epilepsy and cardiac disease may result from the same genetic abnormality. Seizures may be arrhythmogenic in that they have been associated with a simultaÂ� neous marked increase in heart rate (Keilson et al. 1987; Nei et al. 2000; Opherk et al. 2002), atrioventricular nodal block (Devinsky et al. 1997; Rocamora et al. 2003; Tigaran et al. 2002), and even asystole (Rocamora et al. 2003). Relative prolongation of the mean corrected QT interval occurs during epileptiform EEG discharges compared to times without seizure in patients who later died of SUDEP compared to control patients without SUDEP (Tavernor et al. 1996). More marked repolarization abnormalities during or immediately after seizures have been described (Nei et al. 2000; Opherk et al. 2002; Tigaran et al. 2003; Zijlmans et al. 2002). This suggests that a transient period of increased risk from repolarization abnormality may occur during a seizure and predispose to arrhythmias. It has been speculated that the mechanism of these electrophysiological abnormalities, and predisposition to arrhythmia, may involve direct transmission of epileptogenic discharges onto the cardiovascular central and peripheral sympathetic system (Lathers and Schraeder 1987; Lathers et al. 1987). Focal epileptic activity in the amygdala, gyrus cinguli, insular cortex, or frontopolar and frontoorbital regions are known to be associated with alterations of heart rate, sinus and atrioventricular node dysfunction, and even supraventricular arrhythmias (Devinsky et al. 1997). Even subconvulsant interictal discharges have been shown to be associated with autonomic cardiac neural discharges and arrhythmias (Lathers et al. 1984). Furthermore, transient left ventricular dysfunction and repolarization abnormality may occur due to autonomic imbalance in response to other CNS diseases such as subarachnoid hemorrhage (Deibert et al. 2003). SUDEP does occur, however, in the absence of seizures, and it is a possibility that epilepsy or anticonvulsant therapy predisposes an individual to arrhythmic death. The latter option has been investigated without conclusive results (Lathers and Schraeder 2002; Walczak 2003). Epileptics, however, have higher resting heart rates and longer QT intervals than nonepileptics, although not outside the normal reference range (Drake et al. 1993). Whether these ἀndings are signiἀcant is unclear, but there have been several reports of epileptic patients with coexistent cardiac arrhythmias such as polymorphic ventricular tachycardia, atrioventricular nodal reentry tachycardia, sinus arrest, sinus bradycardia, and asystole (Langan et al. 2000; Linzer et al. 1994; Rugg-Gunn et al. 2004; Venkataraman et al. 2001). The possibility that seizures and arrhythmias are manifestations of the same genetic mutation affecting both neural and cardiac ionic currents must be considered. There are already examples of potential “crossover” phenotypes in LQTS. The Jervell and LangeNielsen syndrome is associated with congenital deafness because homozygosity or compound heterozygosity for KCNQ1 mutations inhibits endolymph production in the inner ear (Vetter et al. 1996). Similarly, LQT3 (SCN5A linked) is associated with an increased prevalence of gastrointestinal symptoms (Locke et al. 2006) possibly reflecting the expression of the Nav1.5 ion channel in human cardiovascular and gastrointestinal systems. Of more direct relevance is that patients with KCNH2 mutations (causing LQT2) are more likely to have either a personal or family history of seizures, or a history of prescription of antiepileptic drugs than patients with LQT1, LQT3 or the estimated background expected rate (Johnson et al. 2009). KCNH2-encoded potassium channels, found in hippocampal glia, are important in the regulation of extraneural potassium concentrations that when

Sudden Arrhythmic Death Syndrome

735

altered may be epileptogenic (Emmi et al. 2000; Janigro et al. 1997). Furthermore, KCNH2associated channels are known to be expressed in numerous other locations in the murine brain other than the hippocampus, potentially indicating that mutations in this gene may cause more widespread seizure activity (Papa et al. 2003). Other possible crossover syndromes include SCN1A (Hindocha et al. 2009) and SCN1B (Wallace et al. 1998) mutations. These encode the alpha and beta subunits, respectively, of the voltage-gated sodium ion channel Nav1.1. SCN1A is found in human neural tissue and is expressed in the murine, rat, and dog heart, especially pacemaker tissue. Human mutations have been associated with familial forms of febrile epilepsy, generalized seizures, and SUDEP (Hindocha et al. 2008), whereas selective block of murine neural Nav1.1 channels causes pacemaker dysfunction (Maier et al. 2003). This may be in keeping with observations of bradycardia during seizure activity. SCN1B mutations have similarly been detected in familial epilepsy disorders and have also been described in association with BrS. Therefore, overlap syndromes of ion channel disease that cause cardiac arrhythmia and epilepsy are likely and are potential etiologies for both SADS and SUDEP deaths. The prevalence of these potential familial overlap syndromes is likely to be low, but there may be a role for common variants in genes that affect cardiac repolarization and conduction in addition to neurological function, and so predispose carriers to epilepsy and arrhythmia in a non-Mendelian fashion.

45.7╅ Conclusion Cardiac genetic disorders are an important cause of sudden and unexpected death and may contribute to the phenomenon of SUDEP. Careful clinical evaluation of patients with seizure disorders may be able to identify those who demonstrate signs of cardiac disease, whether genetic or not, that causes secondary anoxic seizures and places them at risk of arrhythmic death. Evaluation of families of SUDEP victims and molecular autopsy of series of victims with retained postmortem DNA appears to be the next step in the study of€the hypothesis that inherited cardiac disease is an underlying etiology. In the light of more recent evidence of overlap between epilepsy and cardiac ion channel disease, this research should extend to the study of potential familial epileptic disorders that are likely to demonstrate incomplete penetrance in a similar way to inherited cardiac disease. In addition, the role for non-Mendelian inherited genetic risk for sudden death attributable to common variants requires genomic study in large numbers of well-characterized SUDEP and SADS cases. A comparison of both phenomena may reveal that their genetic differences may not be as great as once thought.

References Ackerman, M. J. 2004. Cardiac channelopathies: It’s in the genes. Nat Med 10 (5): 463–464. Ackerman, M. J., and D. E. Clapham. 1997. Ion channels—Basic science and clinical disease. N Engl J Med 336 (22): 1575–1586. Alcalai, R., J. G. Seidman, and C. E. Seidman. 2008. Genetic basis of hypertrophic cardiomyopathy: From bench to the clinics. J Cardiovasc Electrophysiol 19 (1): 104–110. Al-Hillawi, A. H., R. Hayward, and N. M. Johnson. 1984. Incidence of cardiac arrhythmias in patients taking slow release salbutamol and slow release terbutaline for asthma. Br Med J (Clin Res Ed) 288 (6414): 367.

736 Sudden Death in Epilepsy: Forensic and Clinical Issues Antzelevitch, C. 2007. Genetic basis of Brugada syndrome. Heart Rhythm 4: 756–757. Antzelevitch, C., P. Brugada, M. Borggrefe, J. Brugada, R. Brugada, D. Corrado, I. Gussak et al. 2005. Brugada syndrome: Report of the second consensus conference: Endorsed by the Heart Rhythm Society and the European Heart Rhythm Association. Circulation 111 (5): 659–670. Beffagna, G., G. Occhi, A. Nava, L. Vitiello, A. Ditadi, C. Basso, B. Bauce et al. 2005. Regulatory mutations in transforming growth factor-beta3 gene cause arrhythmogenic right ventricular cardiomyopathy type 1. Cardiovasc Res 65 (2): 366–373. Behr, E., D. A. Wood, M. Wright, P. Syrris, M. N. Sheppard, A. Casey, M. J. Davies, W. McKenna, and Arrhythmic Death Syndrome Steering Group. CN-Sudden. 2003. Cardiological assessment of ἀrst-degree relatives in sudden arrhythmic death syndrome. Lancet 362 (9394): 1457–1459. Behr, E. R., A. Casey, M. Sheppard, M. Wright, T. J. Bowker, M. J. Davies, W. J. McKenna, and D. A. Wood. 2007. Sudden arrhythmic death syndrome: A national survey of sudden unexplained cardiac death. Heart 93 (5): 601–605. Behr, E. R., C. Dalageorgou, M. Christiansen, P. Syrris, S. Hughes, M. T. Tome Esteban, E. Rowland, S. Jeffery, and W. J. McKenna. 2008. Sudden arrhythmic death syndrome: Familial evaluation identiἀes inheritable heart disease in the majority of families. Eur Heart J 29 (13): 1670–1680. Bellocq, C., A. C. van Ginneken, C. R. Bezzina, M. Alders, D. Escande, M. M. Mannens, I. Baro, and A. A. Wilde. 2004. Mutation in the KCNQ1 gene leading to the short QT-interval syndrome. Circulation 109 (20): 2394–2397. Bomma, C., J. Rutberg, H. Tandri, K. Nasir, A. Roguin, C. Tichnell, R. Rodriguez et al. 2004. Misdiagnosis of arrhythmogenic right ventricular dysplasia/cardiomyopathy. J Cardiovasc Electrophysiol 15 (3): 300–306. Bowker, T. J., D. A. Wood, M. J. Davies, M. N. Sheppard, N. R. Cary, J. D. Burton, D. R. Chambers et al. 2003. Sudden, unexpected cardiac or unexplained death in England: A national survey. Q J Med 96 (4): 269–279. Brugada, J., P. Brugada, and R. Brugada. 2001. Brugada syndrome: The syndrome of right bundle branch block, ST segment elevation in V1 to V3 and sudden death. Indian Pacing Electrophysiol J 1 (1): 6–11. Brugada, J., R. Brugada, and P. Brugada. 2003. Determinants of sudden cardiac death in individuals with the electrocardiographic pattern of Brugada syndrome and no previous cardiac arrest. Circulation 108 (25): 3092–3096. Brugada, P., and J. Brugada. 1992. Right bundle branch block, persistent ST segment elevation and sudden cardiac death: A distinct clinical and electrocardiographic syndrome. A multicenter report. J Am Coll Cardiol 20 (6): 1391–1396. Brugada, R., K. Hong, R. Dumaine, J. Cordeiro, F. Gaita, M. Borggrefe, T. M. Menendez et al. 2004. Sudden death associated with short-QT syndrome linked to mutations in HERG. Circulation 109 (1): 30–35. Burke, A., W. Creighton, E. Mont, L. Li, S. Hogan, R. Kutys, D. Fowler, and R. Virmani. 2005. Role of SCN5A Y1102 polymorphism in sudden cardiac death in blacks. Circulation 112 (6): 798–802. Burkett, E. L., and R. E. Hershberger. 2005. Clinical and genetic issues in familial dilated cardiomyÂ� opathy. J Am Coll Cardiol 45 (7): 969–981. Cannan, C. R., G. S. Reeder, K. R. Bailey, 3rd, L. J. Melton, and B. J. Gersh. 1995. Natural history of hypertrophic cardiomyopathy. A population-based study, 1976 through 1990. Circulation 92 (9): 2488–2495. Chen, L., M. L. Marquardt, D. J. Tester, K. J. Sampson, M. J. Ackerman, and R. S. Kass. 2007. Mutation of an A-kinase-anchoring protein causes long-QT syndrome. Proc Natl Acad Sci U S A 104 (52): 20990–20995. Choi, G., L. J. Kopplin, D. J. Tester, M. L. Will, C. M. Haglund, and M. J. Ackerman. 2004. Spectrum and frequency of cardiac channel defects in swimming-triggered arrhythmia syndromes. Circulation 110 (15): 2119–2124. Chugh, S. S., K. L. Kelly, and J. L. Titus. 2000. Sudden cardiac death with apparently normal heart. Circulation 102 (6): 649–654.

Sudden Arrhythmic Death Syndrome

737

Corrado, D., C. Basso, A. Nava, and G. Thiene. 2001. Arrhythmogenic right ventricular cardiomyÂ� opathy: Current diagnostic and management strategies. Cardiol Rev 9 (5): 259–265. Corrado, D., C. Basso, and G. Thiene. 2001. Sudden cardiac death in young people with apparently normal heart. Cardiovasc Res 50 (2): 399–408. Davies, M. J., and A. Thomas. 1984. Thrombosis and acute coronary-artery lesions in sudden cardiac ischemic death. N Engl J Med 310 (18): 1137–1140. de la Grandmaison, G. L., and M. Durigon. 2002. Sudden adult death: A medico-legal series of 77 cases between 1995 and 2000. Med Sci Law 42 (3): 225–232. Deibert, E., B. Barzilai, A. C. Braverman, D. F. Edwards, V. Aiyagari, R. Dacey, and M. Diringer. 2003. Clinical signiἀcance of elevated troponin I levels in patients with nontraumatic subarachnoid hemorrhage. J Neurosurg 98 (4): 741–746. Devinsky, O., S. Pacia, and G. Tatambhotla. 1997. Bradycardia and asystole induced by partial seizures: A case report and literature review. Neurology 48 (6): 1712–1714. Drake, M. E., C. R. Reider, and A. Kay. 1993. Electrocardiography in epilepsy patients without cardiac symptoms. Seizure 2 (1): 63–65. Emmi, A., H. J. Wenzel, P. A. Scwartzkroin, M. Taglialatela, P. Castaldo, L. Bianchi, J. Nerbonne, G. A. Robertson, and D. Janigro. 2000. Do glia have heart? Expression and functional role for ether-a-go-go currents in hippocampal astrocytes. J Neurosci 20 (10): 3915–3925. Fabre, A., and M. N. Sheppard. 2006. Sudden adult death syndrome and other non-ischaemic causes of sudden cardiac death. Heart 92: 316–320. Gaita, F., C. Giustetto, F. Bianchi, R. Schimpf, M. Haissaguerre, L. Calo, R. Brugada, C. Antzelevitch, M. Borggrefe, and C. Wolpert. 2004. Short QT syndrome: Pharmacological treatment. J Am Coll Cardiol 43 (8): 1494–1499. Gallo, P., G. d’Amati, and F. Pelliccia. 1992. Pathologic evidence of extensive left ventricular involvement in arrhythmogenic right ventricular cardiomyopathy. Hum Pathol 23 (8): 948–952. Gerull, B., A. Heuser, T. Wichter, M. Paul, C. T. Basson, D. A. McDermott, B. B. Lerman et al. 2004. Mutations in the desmosomal protein plakophilin-2 are common in arrhythmogenic right ventricular cardiomyopathy. Nat Genet 36 (11): 1162–1164. Goldenberg, I., and A. J. Moss. 2008. Long QT syndrome. J Am Coll Cardiol 51 (24): 2291–2300. Goldenberg, I., W. Zareba, and A. J. Moss. 2008. Long QT Syndrome. Curr Probl Cardiol 33 (11): 629–694. Gussak, I., P. Brugada, J. Brugada, R. S. Wright, S. L. Kopecky, B. R. Chaitman, and P. Bjerregaard. 2000. Idiopathic short QT interval: A new clinical syndrome? Cardiology 94 (2): 99–102. Haissaguerre, M., N. Derval, F. Sacher, L. Jesel, I. Deisenhofer, L. de Roy, J. L. Pasquie et al. 2008. Sudden cardiac arrest associated with early repolarization. N Engl J Med 358 (19): 2016–2023. Haissaguerre, M., S. Chatel, F. Sacher, R. Weerasooriya, V. Probst, G. Loussouarn, M. Horlitz et al. 2009. Ventricular ἀbrillation with prominent early repolarization associated with a rare variant of KCNJ8/KATP channel. J Cardiovasc Electrophysiol 20 (1): 93–98. Hedley, P. L., P. Jorgensen, S. Schlamowitz, J. Moolman-Smook, J. K. Kanters, V. A. Corἀeld, and M. Christiansen. 2009. The genetic basis of Brugada syndrome: A mutation update. Hum Mutat 30 (9): 1256–1266. Hermida, J. S., J. L. Lemoine, F. B. Aoun, G. Jarry, J. L. Rey, and J. C. Quiret. 2000. Prevalence of the Brugada syndrome in an apparently healthy population. Am J Cardiol 86 (1): 91–94. Hindocha, N., L. Nashef, F. Elmslie, R. Birch, S. Zuberi, A. Al-Chalabi, L. Crotti, P. J. Schwartz, and A. Makoff. 2008. Two cases of sudden unexpected death in epilepsy in a GEFS+ family with an SCN1A mutation. Epilepsia 49 (2): 360–365. Hindocha, N., R. Nabbout, F. Elmslie, A. Makoff, A. Al-Chalabi, and L. Nashef. 2009. A case report of a family with overlapping features of autosomal dominant febrile seizures and GEFS+. Epilepsia 50 (4): 937–942. Hobbs, J. B., D. R. Peterson, A. J. Moss, S. McNitt, W. Zareba, I. Goldenberg, M. Qi et al. 2006. Risk of aborted cardiac arrest or sudden cardiac death during adolescence in the long-QT syndrome. JAMA 296 (10): 1249–1254.

738 Sudden Death in Epilepsy: Forensic and Clinical Issues Hofman, N., A. A. Wilde, and H. L. Tan. 2007. Diagnostic criteria for congenital long QT syndrome in the era of molecular genetics: Do we need a scoring system? Eur Heart J 28 (11): 1399. Ingles, J., and C. Semsarian. 2007. Sudden cardiac death in the young: A clinical genetic approach. Intern Med J 37 (1): 32–37. Janigro, D., S. Gasparini, R. D’Ambrosio, G. McKhann 2nd, and D. DiFrancesco. 1997. Reduction of K+ uptake in glia prevents long-term depression maintenance and causes epileptiform activity. J Neurosci 17 (8): 2813–2824. Jervell, A., and F. Lange-Nielsen. 1957. Congenital deaf-mutism, functional heart disease with prolongation of the Q–T interval and sudden death. Am Heart J 54 (1): 59–68. Johnson, J. N., N. Hofman, C. M. Haglund, G. D. Cascino, A. A. Wilde, and M. J. Ackerman. 2009. Identiἀcation of a possible pathogenic link between congenital long QT syndrome and epilepsy. Neurology 72 (3): 224–231. Kao, W. H., D. E. Arking, W. Post, T. D. Rea, N. Sotoodehnia, R. J. Prineas, B. Bishe et al. 2009. Genetic variations in nitric oxide synthase 1 adaptor protein are associated with sudden cardiac death in US white community-based populations. Circulation 119 (7): 940–951. Keating, M. T., and M. C. Sanguinetti. 2001. Molecular and cellular mechanisms of cardiac arrhythmias. Cell 104 (4): 569–580. Keilson, M. J., W. A. Hauser, J. P. Magrill, and M. Goldman. 1987. ECG abnormalities in patients with epilepsy. Neurology 37 (10): 1624–1626. Klein, G. J., T. M. Bashore, T. D. Sellers, E. L. Pritchett, W. M. Smith, and J. J. Gallagher. 1979. VentricÂ� ular€ἀbrillation in the Wolff–Parkinson–White syndrome. N Engl J Med 301 (20): 1080–1085. Kofflard M. J., D. J. Waldstein, J. Vos, and F. J. ten Cate. 1993. Prognosis in hypertrophic cardiomyÂ� opathy observed in a large clinic population. Am J Cardiol 72 (12): 939–943. Kontula, K., P. J. Laitinen, A. Lehtonen, L. Toivonen, M. Viitasalo, and H. Swan. 2005. CatecholÂ� aminergic polymorphic ventricular tachycardia: Recent mechanistic insights. Cardiovasc Res 67 (3): 379–387. Lahat, H., E. Pras, T. Olender, N. Avidan, E. Ben-Asher, O. Man, E. Levy-Nissenbaum et al. 2001. A missense mutation in a highly conserved region of CASQ2 is associated with autosomal recessive catecholamine-induced polymorphic ventricular tachycardia in Bedouin families from Israel. Am J Hum Genet 69 (6): 1378–1384. Laitinen, P. J., K. M. Brown, K. Piippo, H. Swan, J. M. Devaney, B. Brahmbhatt, E. A. Donarum et al. 2001. Mutations of the cardiac ryanodine receptor (RyR2) gene in familial polymorphic ventricular tachycardia. Circulation 103 (4): 485–490. Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68 (2): 211–213. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42 (2): 123–136. Leenhardt, A., V. Lucet, I. Denjoy, F. Grau, D. D. Ngoc, and P. Coumel. 1995. Catecholaminergic polymorphic ventricular tachycardia in children. A 7-year follow-up of 21 patients. Circulation 91 (5): 1512–1519. Lehnart, S. E., X. H. Wehrens, P. J. Laitinen, S. R. Reiken, S. X. Deng, Z. Cheng, D. W. Landry, K. Kontula, H. Swan, and A. R. Marks. 2004. Sudden death in familial polymorphic ventricular tachycardia associated with calcium release channel (ryanodine receptor) leak. Circulation 109 (25): 3208–3214.

Sudden Arrhythmic Death Syndrome

739

Liberthson, R. R. 1996. Sudden death from cardiac causes in children and young adults. N Engl J Med 334 (16): 1039–1044. Lim, R., M. J. Walshaw, S. Saltissi, and C. R. Hind. 1989. Cardiac arrhythmias during acute exacerbations of chronic airflow limitation: Effect of fall in plasma potassium concentration induced by nebulised beta 2-agonist therapy. Postgrad Med J 65 (765): 449–452. Linzer, M., B. P. Grubb, S. Ho, L. Ramakrishnan, E. Bromἀeld, and N. A. Estes 3rd. 1994. Cardiovascular causes of loss of consciousness in patients with presumed epilepsy: A cause of the increased sudden death rate in people with epilepsy? Am J Med 96 (2): 146–154. Locke 3rd, G. R., M. J. Ackerman, A. R. Zinsmeister, P. Thapa, and G. Farrugia. 2006. Gastrointestinal symptoms in families of patients with an SCN5A-encoded cardiac channelopathy: Evidence of an intestinal channelopathy. Am J Gastroenterol 101 (6): 1299–1304. MacLeod, H. M., M. R. Culley, J. M. Huber, and E. M. McNally. 2003. Lamin A/C truncation in dilated cardiomyopathy with conduction disease. BMC Med Genet 4: 4. Maier, S. K., R. E. Westenbroek, T. T. Yamanushi, H. Dobrzynski, M. R. Boyett, W. A. Catterall, and T. Scheuer. 2003. An unexpected requirement for brain-type sodium channels for control of heart rate in the mouse sinoatrial node. Proc Natl Acad Sci U S A 100 (6): 3507–3512. Marban, E. 2002. Cardiac channelopathies. Nature 415 (6868): 213–218. Markiewicz-Loskot, G., E. Moric-Janiszewska, M. Loskot, L. Szydlowski, L. Weglarz, and A. Hollek. 2006. Isolated ventricular non-compaction: Clinical study and genetic review. Europace 8 (12): 1064–1067. Maron, B. J. 2002. Hypertrophic cardiomyopathy: A systematic review. JAMA 287 (10): 1308–1320. Maron, B. J., J. M. Gardin, J. M. Flack, S. S. Gidding, T. T. Kurosaki, and D. E. Bild. 1995. Prevalence of hypertrophic cardiomyopathy in a general population of young adults. Echocardiographic analysis of 4111 subjects in the CARDIA Study. Coronary Artery Risk Development in (Young) Adults. Circulation 92 (4): 785–789. Maron, B. J., J. Shirani, L. C. Poliac, R. Mathenge, W. C. Roberts, and F. O. Mueller. 1996. Sudden death in young competitive athletes. Clinical, demographic, and pathological proἀles. JAMA 276 (3): 199–204. Matsuo, K., M. Akahoshi, E. Nakashima, S. Seto, and K. Yano. 2004. Clinical characteristics of subjects with the Brugada-type electrocardiogram. J Cardiovasc Electrophysiol 15 (6): 653–657. McKenna, W. J., and A. J. Camm. 1989. Sudden death in hypertrophic cardiomyopathy. Assessment of patients at high risk. Circulation 80 (5): 1489–1492. McKenna, W. J., G. Thiene, A. Nava, F. Fontaliran, C. Blomstrom-Lundqvist, G. Fontaine, and F. Camerini. 1994. Diagnosis of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Task Force of the Working Group Myocardial and Pericardial Disease of the European Society of Cardiology and of the Scientiἀc Council on Cardiomyopathies of the International Society and Federation of Cardiology. Br Heart J 71 (3): 215–218. McKenna, W. J., and J. E. Deanἀeld. 1984. Hypertrophic cardiomyopathy: An important cause of sudden death. Arch Dis Child 59 (10): 971–975. McKoy, G., N. Protonotarios, A. Crosby, A. Tsatsopoulou, A. Anastasakis, A. Coonar, M. Norman, C. Baboonian, S. Jeffery, and W. J. McKenna. 2000. Identiἀcation of a deletion in plakoglobin in arrhythmogenic right ventricular cardiomyopathy with palmoplantar keratoderma and woolly hair (Naxos disease). Lancet 355 (9221): 2119–2124. Merner, N. D., K. A. Hodgkinson, A. F. Haywood, S. Connors, V. M. French, J. D. Drenckhahn, C. Kupprion et al. 2008. Arrhythmogenic right ventricular cardiomyopathy type 5 is a fully penetrant, lethal arrhythmic disorder caused by a missense mutation in the TMEM43 gene. Am J Hum Genet 82 (4): 809–821. Morentin, B., M. P. Suarez-Mier, and B. Aguilera. 2003. Sudden unexplained death among persons 1–35 years old. Forensic Sci Int 135 (3): 213–217. Morentin, B., and R. Alcaraz. 2002. Sudden unexpected death in epilepsy in children and adolescents. Rev Neurol 34 (5): 462–465.

740 Sudden Death in Epilepsy: Forensic and Clinical Issues Moss, A. J., W. Zareba, E. S. Kaufman, E. Gartman, D. R. Peterson, J. Benhorin, J. A. Towbin et al. 2002. Increased risk of arrhythmic events in long-QT syndrome with mutations in the pore region of the human ether-a-go-go-related gene potassium channel. Circulation 105 (7): 794–799. Nava, A., G. Thiene, B. Canciani, B. Martini, L. Daliento, G. Buja, and G. Fasoli. 1992. Clinical proἀle of concealed form of arrhythmogenic right ventricular cardiomyopathy presenting with apparently idiopathic ventricular arrhythmias. Int J Cardiol 35 (2): 195–206; discussion 207–209. Nei, M., R. T. Ho, and M. R. Sperling. 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41 (5): 542–548. Noseworthy, P. A., and C. Newton-Cheh. 2008. Genetic determinants of sudden cardiac death. Circulation 118 (18): 1854–1863. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: Analysis of influencing factors. Epilepsy Res 52 (2): 117–127. Papa, M., F. Boscia, A. Canitano, P. Castaldo, S. Sellitti, L. Annunziato, and M. Taglialatela. 2003. Expression pattern of the ether-a-gogo-related (ERG) K+ channel-encoding genes ERG1, ERG2, and ERG3 in the adult rat central nervous system. J Comp Neurol 466 (1): 119–135. Peters, S., H. Rust, M. Trummel, and A. Brattstrom. 2000. Familial hypertrophic cardiomyopathy associated with prolongation of the QT interval. Z Kardiol 89 (7): 624–629. Priori, S. G., C. Napolitano, M. Memmi, B. Colombi, F. Drago, M. Gasparini, L. DeSimone et al. 2002. Clinical and molecular characterization of patients with catecholaminergic polymorphic ventricular tachycardia. Circulation 106 (1): 69–74. Priori, S. G., C. Napolitano, N. Tiso, M. Memmi, G. Vignati, R. Bloise, V. Sorrentino, and G. A. Danieli. 2001. Mutations in the cardiac ryanodine receptor gene (hRyR2) underlie catecholaminergic polymorphic ventricular tachycardia. Circulation 103 (2): 196–200. Priori, S. G., C. Napolitano, and P. J. Schwartz. 1999. Low penetrance in the long-QT syndrome: Clinical impact. Circulation 99 (4): 529–533. Priori, S. G., P. J. Schwartz, C. Napolitano, R. Bloise, E. Ronchetti, M. Grillo, A. Vicentini et al. 2003. Risk stratiἀcation in the long-QT syndrome. N Engl J Med 348 (19): 1866–1874. Priori, S. G., S. V. Pandit, I. Rivolta, O. Berenfeld, E. Ronchetti, A. Dhamoon, C. Napolitano et al. 2005. A novel form of short QT syndrome (SQT3) is caused by a mutation in the KCNJ2 gene. Circ Res 96 (7): 800–807. Puranik, R., C. K. Chow, J. A. Duflou, M. J. Kilborn, and M. A. McGuire. 2005. Sudden death in the young. Heart Rhythm 2 (12): 1277–1282. Rampazzo, A., A. Nava, S. Malacrida, G. Beffagna, B. Bauce, V. Rossi, R. Zimbello et al. 2002. Mutation in human desmoplakin domain binding to plakoglobin causes a dominant form of arrhythmogenic right ventricular cardiomyopathy. Am J Hum Genet 71 (5): 1200–1206. Riera, A. R., L. Zhang, A. H. Uchida, E. Schapachnik, S. Dubner, and C. Ferreira. 2007. The management of Brugada syndrome patients. Cardiol J 14 (1): 97–106. Rocamora, R., M. Kurthen, L. Lickfett, J. Von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44 (2): 179–185. Roden, D. M. 1993. Early after-depolarizations and torsade de pointes: Implications for the control of cardiac arrhythmias by prolonging repolarization. Eur Heart J 14 (Suppl H): 56–61. Rodriguez-Calvo, M. S., M. Brion, C. Allegue, L. Concheiro, and A. Carracedo. 2008. Molecular genetics of sudden cardiac death. Forensic Sci Int 182 (1–3): 1–12. Romano, C., G. Gemme, and R. Pongiglione. 1963. Rare cardiac arrhythmias of the pediatric age: I. Repetitive paroxysmal tachycardia. Minerva Pediatr 15: 1155–1164. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Sauer, A. J., A. J. Moss, S. McNitt, D. R. Peterson, W. Zareba, J. L. Robinson, M. Qi et al. 2007. Long QT syndrome in adults. J Am Coll Cardiol 49 (3): 329–337. Savelieva, I., Y. G. Yap, G. Yi, X. H. Guo, K. Hnatkova, A. J. Camm, and M. Malik. 1999. Relation of ventricular repolarization to cardiac cycle length in normal subjects, hypertrophic cardiomyÂ� opathy, and patients with myocardial infarction. Clin Cardiol 22 (10): 649–654.

Sudden Arrhythmic Death Syndrome

741

Schott, J. J., C. Alshinawi, F. Kyndt, V. Probst, T. M. Hoorntje, M. Hulsbeek, A. A. Wilde, D. Escande, M. M. Mannens, and H. Le Marec. 1999. Cardiac conduction defects associate with mutations in SCN5A. Nat Genet 23 (1): 20–21. Schumacher, B., F. H. Gietzen, H. Neuser, J. Schummelfeder, M. Schneider, S. Kerber, R. Schimpf, C. Wolpert, and M. Borggrefe. 2005. Electrophysiological characteristics of septal hypertrophy in patients with hypertrophic obstructive cardiomyopathy and moderate to severe symptoms. Circulation 112 (14): 2096–2101. Schwartz, P. J., A. J. Moss, G. M. Vincent, and R. S. Crampton. 1993. Diagnostic criteria for the long QT syndrome. An update. Circulation 88 (2): 782–784. Schwartz, P. J., S. G. Priori, C. Spazzolini, A. J. Moss, G. M. Vincent, C. Napolitano, I. Denjoy et al. 2001. Genotype-phenotype correlation in the long-QT syndrome: Gene-speciἀc triggers for life-threatening arrhythmias. Circulation 103 (1): 89–95. Schwartz, P. J., S. G. Priori, M. Cerrone, C. Spazzolini, A. Odero, C. Napolitano, R. Bloise et al. 2004. Left cardiac sympathetic denervation in the management of high-risk patients affected by the long-QT syndrome. Circulation 109 (15): 1826–1833. Shah, P. M., A. G. Adelman, E. D. Wigle, F. L. Gobel, H. B. Burchell, T. Hardarson, R. Curiel, C. De La Calzada, C. M. Oakley, and J. F. Goodwin. 1974. The natural (and unnatural) history of hypertrophic obstructive cardiomyopathy. Circ Res 35 (2) (Suppl II): 179–195. Splawski, I., J. Shen, K. W. Timothy, M. H. Lehmann, S. Priori, J. L. Robinson, A. J. Moss et al. 2000. Spectrum of mutations in long-QT syndrome genes. KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 102 (10): 1178–1185. Splawski, I., K. W. Timothy, M. Tateyama, C. E. Clancy, A. Malhotra, A. H. Beggs, F. P. Cappuccio, G.€A. Sagnella, R. S. Kass, and M. T. Keating. 2002. Variant of SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science 297 (5585): 1333–1336. Syrris, P., D. Ward, A. Evans, A. Asimaki, E. Gandjbakhch, S. Sen-Chowdhry, and W. J. McKenna. 2006. Arrhythmogenic right ventricular dysplasia/cardiomyopathy associated with mutations in the desmosomal gene desmocollin-2. Am J Hum Genet 79 (5): 978–984. Tan, H. L., N. Hofman, I. M. van Langen, A. C. van der Wal, and A. A. Wilde. 2005. Sudden unexplained death: Heritability and diagnostic yield of cardiological and genetic examination in surviving relatives. Circulation 112 (2): 207–213. Tavernor, S. J., S. W. Brown, R. M. Tavernor, and C. Gifford. 1996. Electrocardiograph QT lengthening associated with epileptiform EEG discharges—a role in sudden unexplained death in epilepsy? Seizure 5: 79–83. Tester, D. J., and M. J. Ackerman. 2006. The role of molecular autopsy in unexplained sudden cardiac death. Curr Opin Cardiol 21 (3): 166–172. Tester, D. J., and M. J. Ackerman. 2007. Postmortem long QT syndrome genetic testing for sudden unexplained death in the young. J Am Coll Cardiol 49 (2): 240–246. Tester, D. J., M. L. Will, C. M. Haglund, and M. J. Ackerman. 2005. Compendium of cardiac channel mutations in 541 consecutive unrelated patients referred for long QT syndrome genetic testing. Heart Rhythm 2 (5): 507–517. Tigaran, S., H. Molgaard, and M. Dam. 2002. Atrio-ventricular block: A possible explanation of sudden unexpected death in epilepsy. Acta Neurol Scand 106 (4): 229–233. Tigaran, S., H. Molgaard, R. McClelland, M. Dam, and A. S. Jaffe. 2003. Evidence of cardiac ischemia during seizures in drug refractory epilepsy patients. Neurology 60 (3): 492–495. Tiso, N., D. A. Stephan, A. Nava, A. Bagattin, J. M. Devaney, F. Stanchi, G. Larderet, B. Brahmbhatt et al. 2001. Identiἀcation of mutations in the cardiac ryanodine receptor gene in families affected with arrhythmogenic right ventricular cardiomyopathy type 2 (ARVD2). Hum Mol Genet 10 (3): 189–194. Ueda, K., C. Valdivia, A. Medeiros-Domingo, D. J. Tester, M. Vatta, G. Farrugia, M. J. Ackerman, and J. C. Makielski. 2008. Syntrophin mutation associated with long QT syndrome through activation of the nNOS–SCN5A macromolecular complex. Proc Natl Acad Sci U S A 105 (27): 9355–9360.

742 Sudden Death in Epilepsy: Forensic and Clinical Issues Venkataraman, V., J. W. Wheless, L. J. Willmore, and H. Motookal. 2001. Idiopathic cardiac asystole presenting as an intractable adult onset partial seizure disorder. Seizure 10 (5): 359–364. Vetter, D. E., J. R. Mann, P. Wangemann, J. Liu, K. J. McLaughlin, F. Lesage, D. C. Marcus, M. Lazdunski, S. F. Heinemann, and J. Barhanin. 1996. Inner ear defects induced by null mutation of the isk gene. Neuron 17 (6): 1251–1264. Virmani, R., A. P. Burke, and A. Farb. 2001. Sudden cardiac death. Cardiovasc Pathol 10 (5): 211–218. Walczak, T. 2003. Do antiepileptic drugs play a role in sudden unexpected death in epilepsy? Drug Saf 26 (10): 673–683. Wallace, R. H., D. W. Wang, R. Singh, I. E. Scheffer, A. L. George Jr., H. A. Phillips, K. Saar et al. 1998. Febrile seizures and generalized epilepsy associated with a mutation in the Na+-channel beta1 subunit gene SCN1B. Nat Genet 19 (4): 366–370. Ward, O. C. 1964. A new familial cardiac syndrome in children. J Ir Med Assoc 54: 103–106. Yang, P., H. Kanki, B. Drolet, T. Yang, J. Wei, P. C. Viswanathan, S. H. Hohnloser et al. 2002. Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes. Circulation 105 (16): 1943–1948. Yontar, O. C., K. Yalta, M. B. Yilmaz, and O. O. Turgut. 2008. Short QT syndrome: A very rare arrhythmogenic entity. Acta Cardiol 63 (5): 553–555. Zijlmans, M., D. Flanagan, and J. Gotman. 2002. Heart rate changes and ECG abnormalities during epileptic seizures: Prevalence and deἀnition of an objective clinical sign. Epilepsia 43 (8): 847–854.

Odds Ratios Study of Antiepileptic Drugs A Possible Approach to SUDEP Prevention?

46

Claire M. Lathers Paul L. Schraeder H. Gregg Claycamp

Contents 46.1 Why Is Compliance with Antiepileptic Drugs a Continuing Issue? 46.2 Clinical Pharmacology: Drugs as a Beneἀt and/or Risk in Sudden Unexpected Death in Epilepsy? 46.3 Odds Ratio Methods 46.4 Use of ORs to Examine Clinical Pharmacology of Topiramate vs. Lamotrigine vs. Phenobarbital: Comparison of Efficacy and Side Effects Using ORs 46.4.1 Efficacy ORs for Topiramate vs. Lamotrigine vs. Phenobarbital 46.4.2 Side Effects of ORs for Topiramate vs. Lamotrigine vs. Phenobarbital 46.4.3 Comparison of Costs for the Three AEDs 46.5 Discussion References

743 745 745 746 747 748 748 749 752

46.1â•…Why Is Compliance with Antiepileptic Drugs a Continuing Issue? Although the pharmacological nature of the role of antiepileptic medication in explaining the occurrence of sudden unexpected death and epilepsy (SUDEP) is not known, it is generally accepted that noncompliance with prescribed medication is one of the possible risk factors for SUDEP. The control of seizures by anticonvulsant medication is a function of having an optimal, stable therapeutic blood level of a drug that is effectively preventing the recurrence of seizures. Factors that can affect compliance with taking effective doses are the presence of side effects and the affordability of the medication. In response to the former issue, that is, side effects, if the patient ἀnds that the therapeutic level produces symptoms such as drowsiness, slowed thinking, depression, or balance problems, he or she may elect, without consulting a physician, to decrease the dose, resulting in seizures. Likewise, the cost of more expensive medication may result in taking fewer doses to extend the time between medication reἀlls, thereby prolonging the interval between payments. This chapter uses an example of how odds ratio analysis can address cost, side effects, and efficacy issues that should be more systematically addressed in helping to optimize the compliance to treatment of persons with epilepsy and achieving maximal seizure 743

744 Sudden Death in Epilepsy: Forensic and Clinical Issues

control, thereby helping to minimize the risk of SUDEP. This chapter and the subsequent ἀve chapters provide discussion that, in part, emphasizes the importance of antiepileptic drug (AED) compliance and factors that may have an adverse effect on patient compliance, such as AED side effects. Each health care giver must consider the efficacy, safety, and cost of a given AED when determining which drug is best for a given patient. In today’s health care environment, cost is a factor that must be considered. Statisticians and/or risk assessors may, after adjustment for covariates, calculate the odds ratio for efficacy parameters such as termination of status epilepticus within a given time after administration of the AED. Likewise, the rates of complications, such as sedation, respiratory or circulatory complication, hypotension, or cardiac dysrhythmias, may be evaluated by the odds ratios for safety/side effect occurrence. The Food and Drug Administration (FDA) requested that most new AEDs be evaluated for their effect on preventing sudden death in persons with epilepsy (Leestma et al. 1997). Regulatory response resulting from the consequent increased awareness of SUDEP began to occur in 1993, when the FDA focused attention of practitioners and pharmaceutical manufacturers on the question of whether use of anticonvulsant drugs contributes to or prevents sudden unexpected death in epileptic persons. The FDA-convened panel of scientists considered the prevalence of sudden unexpected death in patients involved in studies associated with developing new anticonvulsant drugs and reviewed data on the risk of sudden unexpected death in patients taking lamotrigine. The risk of SUDEP was no different from that found in the young population with epilepsy in general. Estimated SUDEP rates in patients receiving the new anticonvulsant drugs lamotrigine, gabapentin, topiramate, tigabine, and zonisamide were found to be similar to those in patients receiving standard anticonvulsant drugs, suggesting that SUDEP rates reflect population rates and not a speciἀc drug effect, ἀndings that also would indicate that there was no additional beneἀt to the prevention of SUDEP with use of the newer generation of drugs. The FDA required warning labels on the risk of SUDEP in association with the use of each of the above-mentioned drugs. Data obtained in animal models of epilepsy will result in a better understanding of how to optimize the use of old and new AEDs to prevent sudden death. A better understanding of the risk factors for sudden death in persons with epilepsy will also contribute to knowing how to optimize the use of old and new AEDs to prevent sudden death (Lathers and Schraeder 1990, 1995, 2002; Lathers et al. 1993, 1997; Schraeder and Lathers 1989, 1990). The question of making a comparison between the “new” and the “old” AEDs seems to have an obvious answer, namely, “Why even ask the question?” However, it is difficult to separate the marketing data related to the “new” drugs from the efficacy and side effect data in comparison to the “old” drugs. The facts are often obscured by the need to market proprietary drugs while still under patent protection. In addition, funding for pharmacological investigations favors studying the “new” agents. There are few incentives for investigators to conduct comparative efficacy studies with the old AEDs such as phenobarbital. Nonetheless, there are many patients who will beneἀt from the newer AEDs. However, whether the “new” AEDs have advantages over the “old” drugs is not established. No “new” drug stands out as having efficacy in treating seizures that is better than the “old” drugs. The difficulty in achieving therapeutic dosage because of side effects makes one consider whether these agents are “better” than the oldest and most familiar AED, phenobarbital. Thus, it is a useful exercise to compare two of the newer agents (i.e., topiramate and lamoÂ� trigine) with phenobarbital.

Odds Ratios Study of Antiepileptic Drugs

745

46.2â•…Clinical Pharmacology: Drugs as a Benefit and/or Risk in Sudden Unexpected Death in Epilepsy? Death in persons with epilepsy may be the consequence of natural causes such as infections, cardiac disease, cerebrovascular disease, etc., or unnatural causes, such as accidents, homicide, and suicide, which have no direct relationship to the disease of epilepsy (Lathers and Schraeder 2002). Direct causes of death related to epilepsy include status epilepticus, whereas indirect causes may include head trauma or drowning subsequent to a seizure. When death occurs suddenly and without explanation in persons with epilepsy, the term sudden unexpected unexplained death is used (SUDEP). Clinicians and research scientists are working to clarify the term unexplained. Numerous preclinical animal studies have been conducted as models for sudden death and have led to clinical studies in persons with epilepsy. Stimulation of the sympathetic ventrolateral cardiac nerve produced a shift in the origin of the pacemaker and tachyarrhythmias because the nerve is not uniformly distributed to the various regions of the heart but is localized to the atrioventricular junctional and ventricular regions (Lathers et al. 1987, 1988; Schraeder and Lathers 1989). Studies examining the physiology and pharmacology of this ἀnding in animal models found that subconvulsant, interictal discharge was associated with autonomic cardiac neural nonuniform discharge and cardiac arrhythmias. Lathers and Schraeder (1990) summarized the clinical problem of SUDEP, concluding that there was a paucity of clinical data addressing the mechanism of death; a conclusion that, unfortunately, is still valid (Lathers and Schraeder 2002).

46.3â•…Odds Ratio Methods The “odds” is the ratio of the probability that an event occurs (p) to the probability that the event does not occur (1 − p), or odds = p/(1 − p). The “odds ratio” is one of the most common and useful measures in epidemiology that has its origins in 2 × 2 study designs. The 2 × 2 design is typically represented as a matrix with two columns, “disease present” and “disease absent,” and two rows, “risk factor present” and “risk factor absent.” In epidemiological research designed to estimate the association of a disease outcome with an observed risk factor, the odds ratio (OR) is the ratio of the odds for the disease in the presence vs. the absence of the factor. The present study adapts OR methods to comparisons of the relative efficacies of various AEDs. Cramer et al. (1999) referred to the adapted OR as a likelihood of success ratio (LSR). Table 46.1 shows the design necessary to estimate the LSR for a drug and placebo in an epilepsy study. Using the design in Table 46.1, the LSR is estimated from the cell frequencies as OR or LSR as follows:



LSR ≈

a/b ad = , c /d cb

where a/b is an estimate of the odds of improvement under drug treatment and c/d is an estimate of the odds under placebo treatment. Sometimes the complete information from a given drug study is either not reported or is not readily available to the prescribing physician. For example, a published seizure

746 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 46.1â•… Design for Estimating the LSR Seizures Drug Placebo

Improved

Not Improved

a c (a + c)

b d (b + d)

Sums (a + b) (c + d) (a + b + c + d)

rate for patients receiving the drug against a placebo improvement rate is 15%. In such cases, an estimate of the odds might be obtained, using the notation of Table 46.1, from the following:

%Improved ≡



〈Odds〉 =



a × 100, (a + b)

%Improved . 100 − %Improved

Continuing the example of a 60% drug improvement rate and a 15% placebo rate, the OR or the LSR can be estimated from

LSR =

〈Odds〉drug 60 / (100 − 60) = 8.5. = 〈Odds〉 placebo 15 / (100 − 15)



Although these methods provide a simple “ἀrst-pass” estimate for the relative efficacies of drugs for a common treatment endpoint, the reader is cautioned that the methods are only approximate and have important underlying assumptions about the nature of the sampled populations (Selvin 1996). In addition, conἀdence intervals for the OR cannot be estimated without the frequencies of patients in each of the cells in Table 46.1. ORs or LSRs for phenobarbital, topiramate, and lamotrigine were calculated using data published in the literature identiἀed via a search of the World Wide Web and compared with data published in the Physicians’ Desk Reference (2002).

46.4â•…Use of ORs to Examine Clinical Pharmacology of Topiramate vs. Lamotrigine vs. Phenobarbital: Comparison of Efficacy and Side Effects Using ORs Health care givers must consider the efficacy, safety, and side effect proἀle of a given AED when determining which drug is best for a given patient (Lathers et al. 2003). We used the methodology of ORs to address whether “new” AEDs have advantages over “old” drugs using Web-based information access to answer a neurology/clinical pharmacology

Odds Ratios Study of Antiepileptic Drugs

747

Table 46.2â•… Comparisons of AEDs Using Data in Web-Based Search AED Topiramate Lamotrigine Phenobarbital

Improvement Rate (%)

Placebo Rate (%)a

OR

n

40.6–50 20.6–34 58

9.8–20 9.3–18 9.5–22

2.7–29.2 1.3–25.0 5.5–13.0

NA NA 101

ORb (Phenobarbital) 0.6–24.8 1.1–2–210.3

Source: Lathers, C. M., et al., J Clin Pharmacol, 43, 491–503, 2003. With permission. Numbers in boldface are assumed for the purposes of the comparison examples only. b OR (phenobarbital) is the ratio of the improvement rate of phenobarbital to either lamotrigine or topiramate. a

problem: to compare the efficacy and side effects, important factor affecting compliance, of topiramate vs. lamotrigine vs. phenobarbital using ORs. Cost of all three AEDs was compared. A number of new AEDs, including topiramate and lamotrigine, have been developed for chronic focal and secondarily generalized epileptic seizures. Efficacy of these drugs as anticonvulsants does not seem to be superior to that of traditional anticonvulsants such as phenobarbital. However, the advantage of the new drugs is a different spectrum of possible adverse events. Newer AEDs may or may not induce sedation with the expectation of minimizing noncompliance by reducing side effects of lethargy and cognitive impairment. The difficulty in achieving therapeutic dosage because of side effects makes one consider whether these agents are “better” than the older AEDs, the oldest and putatively most side effect–prone of these older AEDs being phenobarbital (Patsalos and Sander 1994; Uldall and Bucholt 1999; Erfurth and Kuhn 2000; Guberman et al. 1999). One advantage of the new AEDs is less frequent drug interactions, leading to improved tolerability with comedication. We compared two “new” AEDs, topiramate and lamotrigine, with phenobarbital by evaluating efficacies and side effects using relative ORs, a method commonly used in drug development research. Development of new algorithms and/or new knowledge will bring beneἀcial tools to all clinical pharmacologists. Table 46.2 summarizes the ORs of improvement in seizure control for topiramate, lamotrigine, and phenobarbital vs. placebo. Following these approximate methods, the results in Table 46.2 show that we might expect from 0.6 to 4.8 times greater efficacy of phenobarbital when compared to topiramate, and from 1.1 to 10.3 times greater efficacy of phenobarbital when compared to lamotrigine. Given the uncertainties in comparisons across many studies, the reasonable conclusion is that the efficacies of topiramate and lamotrigine are probably not signiἀcantly different from the efficacy of phenobarbital. Although the reader is cautioned about the need for more information to control for bias in study comparisons, these ἀrst-pass estimates suggest that phenobarbital is an equivalent choice of AEDs based on seizure improvement rates and the ORs. 46.4.1â•… Efficacy ORs for Topiramate vs. Lamotrigine vs. Phenobarbital In the absence of detailed study information and a rigorous meta-analysis of numerous studies, a “ἀrst-order” comparison of drugs can be made using improvement rates and the ORs. The more data available, the stronger the conclusion to be made from the analysis. For example, comparing the three AEDs for a reduction in seizures, using data published by the VA cooperative study (Smith et al. 1987) and other data by Harden (2001), the improvement rates ranged from 20.6% to 58%, depending on the AED (Table 46.2). Similarly, the

748 Sudden Death in Epilepsy: Forensic and Clinical Issues

placebo improvement rates vary from 9.3% to 18% for topiramate and lamotrigine studies tabulated in Harden (2001). For the purpose of this basic exercise, we might assume that the placebo rate, were it to have been reported for all the drugs, would be similar to the rates reported for most AED studies, that is, from 9% to 20%. Given the uncertainties in comparisons across many studies, the reasonable conclusion is that the efficacies of topiramate and lamotrigine are probably not signiἀcantly different from that of phenobarbital. 46.4.2â•… Side Effects of ORs for Topiramate vs. Lamotrigine vs. Phenobarbital Similar approximate methods can be used to compare adverse effect rates for the AEDs. For example, the adverse effect rates for sedation are compared in Table 46.3. The data for the comparisons are incomplete because the placebo rates of adverse effects were subtracted from the reported rates of sedation for topiramate and lamotrigine (Cramer et al. 1999), whereas the gross adverse effect rate was apparently summarized for phenobarbital (Cramer and Mattson 1995). To compare the adverse effects rate for the AEDs, we used two different phenobarbital rates (Table 46.3). Assuming that the placebo rate for phenobarbital is of similar magnitude (10%) as reported for the remaining AEDs, an OR for the experience of sedation compared to topiramate and, separately, lamotrigine can be estimated using the data from Cramer and Mattson (1995) (Table 46.3, phenobarbital review 1). The results of this exercise show that the sedation rates for topiramate compare equally with phenobarbital and are 2.1 to 4 times worse than the rate observed for lamotrigine. Actual clinical observations of patients suggest that the sedative effect of phenobarbital is equal to that of topiramate, whereas the sedation rate for both phenobarbital and topiramate is greater than that of lamotrigine. 46.4.3â•… Comparison of Costs for the Three AEDs The authors’ survey of retail pharmacies indicated that, on average, a 1-month supply of phenobarbital is approximately $10.00 vs. $177.00 for a brand name of topiramate and $137.00 for lamotrigine. Thus, topiramate and lamotrigine are 17.7 and 13.7 times more expensive than phenobarbital. If the patient tolerates phenobarbital with minimal side effects, it is a very cost-effective drug for the treatment of epilepsy over the long term.

Table 46.3â•… Comparison of Sedation Rates for the AEDs ORa AED Topiramate Lamotrigine Phenobarbital, review 1 Phenobarbital, review 2

Net Sedation

Net Odds

20b 7b 14–23c 1

0.25 0.075 0.16–0.30 0.01

Phenobarbital Review 1

Phenobarbital Review 2

0.64–1.2 2.1–4.0

25 7.5

OR (phenobarbital) is the ratio of the net odds of phenobarbital to either lamotrigine or topiramate. Source: Cramer, J., et al., Epilepsia, 40 (5), 590–600, 1999. c Assuming a 10% placebo rate as in the other AED studies. Source: Cramer, J., and Mattson, R.H., in Levy, R. H., et al., Antiepileptic Drugs, Raven Press, New York, NY, 1995. With permission. a

b

Odds Ratios Study of Antiepileptic Drugs

749

46.5â•…Discussion The clinician is often presented with a dilemma in that much of the data received concerning prescription practice in treating epilepsy come from pharmaceutical company representatives, courses at meetings sponsored by the various manufacturers, and published research that is funded and often edited by these same corporations. There are very little data comparing the “old” and the “new” AEDs. Most recent studies compare the efficacy and side effects of the various “new” agents. Few, if any, studies make direct comparisons between “old” and “new” agents. Thus, the clinician is left to his or her own means and experience when deciding what drug to use in a given patient. When one looks critically at the efficacy of the “old” drugs, the only variable that seems to be operative is that of whether the patient will tolerate the side effects of the agent. This factor is most important in that when the variable of intolerance to side effects is removed, all of the “old” drugs (i.e., phenytoin, carbamazepine, phenobarbital, primidone, and valproic acid) have relatively equivalent efficacy. If one then considers the “new” AEDs (i.e., gabapentin, tiagabine, topiramate, lamotrigine, levetiracetam, zonisamide), little data show that clinical efficacy is an improvement over the “old” drugs. In addition, the question of side effects has major importance when prescribing these drugs. When using these agents, it often takes 4 to 8 weeks to reach minimal therapeutic doses. Thus, to achieve efficacy with the minimal acute doserelated side effects, using the “new” and most of the “old” agents, a slow titration is necessary. Were one to give the patient a therapeutic dose at the start of treatment, most of the patients, because of CNS side effects, would refuse to use any of these agents, with the possible exception of gabapentin or phenytoin. To minimize noncompliance due to excessive side effects (primarily lethargy and cognitive impairment), the recommended starting daily doses are miniscule and are increased gradually to reach doses that are in the range of achieving efficacy. The time required to reach these doses can be 4 to 12 weeks or more. OR methodology was used in our study to compare efficacy and side effects. The rule of thumb when interpreting ORs is that the lower the OR, the greater the number of subjects required to conclude that there is a statistically signiἀcant difference between the drug vs. placebo or drug A vs. drug B. In general, an OR of 1.5 obtained in studies containing small numbers of subjects, as are often found in clinical trials, should be summarized as showing no signiἀcant difference. The data compared three AEDs for a reduction in seizures (i.e., typically deἀned as a 50% reduction in the seizure rate) and obtained improvement rates of 1.0, 0.52, and 1.4 for topiramate, lamotrigine, and phenobarbital, respectively. The OR for phenobarbital compared to topiramate was 1.4/1.0 = 1.4, and the OR for phenobarbital compared to lamotriÂ� gine was 1.4/0.52 = 2.7. The reader is cautioned about the need to obtain more information to control for bias in study comparisons. This ἀrst-pass estimate suggested that phenobarbital is a better choice of AED based on improvement rates. However, it is important to compare the conclusion made when more data are available to obtain a stronger comparison using OR methodology. Use of a range of data provided a range of estimates for relative efficacy by comparing phenobarbital vs. topiramate and phenobarbital vs. lamotrigine. This comparison resulted in the conclusion that phenobarbital is an equivalent choice of an AED based on improvement rates and the ORs. The “grand conclusion” is the age-old fact that more data obtained from clinical studies allow the health care giver to make a better medical decision when treating a patient.

750 Sudden Death in Epilepsy: Forensic and Clinical Issues

The newer AED drugs do not have the same clinical trial experience as the standard AEDs (i.e., phenobarbital, phenytoin, carbamazepine, valproic acid). This fact is further confounded by the lack of a clear understanding of the placebo effect. The placebo effect is very difficult to control for in drug efficacy trials. Data from clinical trials are difficult to compare because of the variability in the placebo response and the variability in the severity of the seizure disorders in the different patient population (Cramer et al. 1999). That placebos have a measurable therapeutic effect in some patients is a phenomenon that is difficult to explain. The interactions of mind, body, environment, and patient expectations result in the placebo effect. If this were an inactive process and only a statistical phenomenon, then the percentage of responders would be expected to be nearly identical across all studies. The intangible effect of the relationship between the patient and the physician is also a factor in the occurrence of a placebo effect. Indeed, before the advent of modern medicine, this relationship and the common use of placebos as therapy were all that physicians could offer most patients (Papakostas and Daras 2001). That newer AED drugs do not have the same clinical trial experience as the standard AEDs (i.e., phenobarbital, phenytoin, carbamazepine, valproic acid) is unfortunate. The treatment efficacy of the newer, more recently available AEDs is based on a smaller clinical trial experience and patient numbers than is the case for the standard AEDs, and thus comparison of efficacy is only relatively valid. Nevertheless, this information is all that the physician has available when making a decision to prescribe a medication. The use of OR methodology helps to address this deἀciency. This study compared two “new” AEDs, topiramate and lamotrigine, with phenobarbital by comparing the efficacies, side effects, and costs using relative ORs, a method commonly used in drug development research. Development of new algorithms and/or new knowledge will bring beneἀcial tools to all readers and for authors and reviewers of new animal or human drug applications submitted to the US FDA. For those patients with epilepsy who still exhibit epileptic seizures and/or unacceptable side effects after medication with an “older” AED, better clinical control may be possible with use of one of the newer AEDs. Efficacy and tolerability of AEDs depend on the physician being aware of the pharmacokinetic properties affecting drug absorption, distribution, biotransformation, and excretion. Finding the best dose allowing easy titration to reach the effective dose while minimizing liability to interactions with other drugs will improve patient compliance and long-term beneἀts from the clinical pharmacological agent (Harden 2001). Use of any AED that allows monotherapy in a given patient is a valuable factor to ensure compliance and ease of patient management for the health care giver (Brodie and Dichter 1996) that, hopefully, will contribute to understanding how best to improve compliance with AED use with the goal of diminishing the risk of SUDEP. Teaching Tak e Ho me Po int s Question 1: How does the physician/health care provider determine which antiepileptic drug should be used to initiate monotherapy in the treatment of epilepsy? Answer 1 1. The drug selected must be known to have efficacy in treating the patient’s seizure type.

Odds Ratios Study of Antiepileptic Drugs

751

2. The ability of a given patient to tolerate a speciἀc antiepileptic drug is the most important factor in determining whether the patient should continue on the medication. 3. All of the major antiepileptic drugs have a similar efficacy when used in patients who tolerate the medication. Therefore, the only important variable in determining which drug should be used as initial monotherapy to treat seizures is the tolerance of side effects. Question 2: How does the health care provider determine what category of antiepileptic drug is appropriate to treat the patient’s seizure type? Answer 2 For partial seizures (complex or simple) or generalized tonic/clonic seizures, all three of the drugs, phenobarbital, topiramate, and lamotrigine, are effective. Question 3: How do the ORs for phenobarbital, topiramate, and lamotrigine guide the care giver when choosing medications for use in initial monotherapy for the treatment of seizures? Answer 3 Side effects and cost are signiἀcant factors when making this determination since the absolute efficacy of either topiramate or lamotrigine is no better than phenobarbital. Question 4: As another example of how ORs have been used in decision making, we raise the question of the use, efficacy, and safety of two other commonly used antiepileptic drugs, lorazepam and diazepam. “Is the administration of benzodiazepines by paramedics an effective and safe treatment for out-of-hospital status epilepticus?” See Figure 46.1. Answer 4 A recent study published by Alldredge et al. (2001) addressed this question. A randomized, double-blind trial was conducted to evaluate intravenous benzodiazepines administered in this manner. Adults with prolonged, 5 min or longer, or repetitive generalized convulsive seizure were administered diazepam (5 mg), lorazepam (2 mg), or placebo. A second injection of the same dose was administered if the need was present. Status epilepticus had been terminated on arrival at the emergency room in more patients treated with lorazepam (59.1%) or diazepam (42.6%) than in patients receiving placebo (21.1%, p = 0.001). Adjustment for covariates was done and then the ORs for termination of status epileptics by the time of arrival in the lorazepam group as compared to the placebo group was 4.8 (95% conἀdence interval, 1.9–13.0). The OR was 1.9 (95% conἀdence interval, 0.8–4.4) in those receiving lorazepam vs. diazepam and 2.3 (95% conἀdence interval, 1.0–5.9) in the diazepam vs. placebo groups. Rates of complications were 10.6% for those receiving lorazepam, 10.3% in the diazepam group, and 22.5% for the placebo group. It was concluded that the benzodiazepines are safe and effective when given by paramedics for out-of-hospital status epilepticus in adults. Lorazepam was likely to be a better therapy than diazepam. The reader is encouraged to read the full published article for additional details about the use of ORs to evaluate antiepileptic drugs. Reproduced from Lathers et al., J Clin Pharmacol, 43, 491–503, 2003. With permission.

752 Sudden Death in Epilepsy: Forensic and Clinical Issues Relative OR 1.0

Drug A Drug B Drug C

0.8 0.5 0.3 Placebo

0.0

Drug

Relative sample size

Figure 46.1╇ Four-way plot of the key feature in a clinical trial result. (From Lathers, C. M., et€al., J Clin Pharmacol, 43, 491–503, 2003. With permission.)

References Alldredge, B. K., A. M. Gelb, S. M., Isaacs, M. D. Corry, F. Allen, S. Ulrich et al. 2001. A comparison of lorazepam, diazepam and placebo for treatment of out-of-hospital status epilepticus. N Engl J Med 345: 631–637. Brodie, M. J., and M. A. Dichter. 1996. Antiepileptic drugs. N Engl J Med 334: 168–175. Cramer, J., R. Fisher, E. Ben-Menachem, J. French, and R. Mattson. 1999. New antiepileptic drugs: Comparison of key clinical trials. Epilepsia 40 (5): 590–600. Cramer, J., and R. H. Mattson. 1995. Phenobarbital toxicity. In Antiepileptic Drugs, 4th ed., ed. R. H. Levy, R. H. Mattson, and B. S. Meldrum, 409–420. New York, NY: Raven Press. Erfurth, A., and G. Kuhn. 2000. Topiramate monotherapy in the maintenance treatment of bipolar I disorder: Effects on mood, weight and serum lipids. Neuropsychobiology 2 (S1): 50–51. Guberman, A. H., F. M. Besag, M. J. Brodie, et al. 1999. Lamotrigine-associated rash: Risk/beneἀt considerations in adults and children. Epilepsia 40: 985–991. Harden, C. L. 2001. Practical and pharmacokinetic considerations in the management of refractory partial epilepsy. In The Protocols Series: Guidelines from Experts in Epilepsy. Number 3, 11–13. Ghent, NY: Galen Press. Lathers, C. M., and P. L. Schraeder. 1995. Experience-based teaching of therapeutics and clinical pharmacology of antiepileptic drugs. J Clin Pharmacol 35: 573–587. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42: 123–136. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital upon autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33: 837–844. Lathers, C. M., N. Tumer, and C. M. Kraras. 1988. The effect of intracerebroventricular d-Ala2Â�methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43: 2287–2298. Lathers, C. M., P. L. Schraeder, and J. G. Boggs. 1997. Sudden unexplained death and autonomic dysfunction. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr. and T. A. Pedley, 1943–1955. Philadelphia, PA: Lippincott-Raven.

Odds Ratios Study of Antiepileptic Drugs

753

Lathers, C. M., P. L. Schraeder, and F. W. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lathers, C. M., P. L. Schraeder, and H. G. Claycamp. 2003. Clinical pharmacology of topiramate versus lamotrigine versus phenobarbitol: Comparison of efficacy and side effects using odds ratios. J Clin Pharmacol 43: 491–503. Leestma, J. E., J. F. Annegers, M. J. Brodie, et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical trial. Epilepsia 38: 47–55. Physicians’ Desk Reference. 2002. 56th ed. Montvale, NJ: Medical Economics Co. Papakostas, Y. G., and M. D. Daras. 2001. Placebos, placebo effect, and the response to the healing situation: The evolution of a concept. Epilepsia 42: 1614–1625. Patsalos, P. N., and J. W. Sander. 1994. Newer antiepileptic drugs: Towards an improved risk-beneἀt ratio. Drug Saf 11: 37–67. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal cardiovascular dysfunction and epileptogenic activity. Epilepsy Res 3: 55–62. Selvin, S. 1996. Statistical Analysis of Epidemiologic Data, 2nd ed. New York, NY: Oxford University Press. Smith, D. B., R. H. Mattson, J. A. Cramer, J. F. Collins, R. A. Novelly, and B. Craft. 1987. The VA Epilepsy Cooperative Study Group. Epilepsia 28 (S3): S50–S58. Uldall, P., and J. Buchholt. 1999. Clinical experiences with topiramate in children with intractable epilepsy. Eur J Paediatr Neurol 3 (3): 105–111.

Antiepileptic Drugs Benefit/ Risk Clinical Pharmacology Possible Role in Cause and/ or Prevention of SUDEP

47

Claire M. Lathers Paul L. Schraeder

Contents 47.1 Antiepileptic Drug Beneἀts vs. Epilepsy 47.2 Antiepileptic Drugs’ Unwanted Side€Effects and Drug Interactions 47.2.1 Phenytoin 47.2.2 Carbamazepine 47.3 Newer Antiepileptic Drugs 47.3.1 Lamotrigine 47.3.2 Zonisamide 47.3.3 Levetiracetam 47.3.4 Pregabalin 47.3.5 Retigabine 47.4 Pregnancy, Epilepsy, and Antiepileptic Drug Use 47.5 Elderly, Epilepsy, and Antiepileptic Drug Use 47.6 Cognitive and Behavior Actions of Antiepileptic Drugs 47.7 Cognitive Effects of Older vs. Newer Antiepileptic Drugs 47.7.1 Oxcarbazepine 47.7.2 Topiramate 47.7.3 Topiramate vs. Lamotrigine 47.7.4 Topiramate vs. Levetiracetam 47.7.5 Topiramate and Zonisamide vs. Tiagabine, Gabapentin, Lamotrigine, and Levetiracetam 47.8 Effects of Interictal Spikes on Cognitive Function 47.9 Generic Antiepileptic Drugs, Unwanted Side Effects, and Risk of SUDEP? 47.10 Possible Role of Antiepileptic Drugs in Causes of SUDEP 47.11 Do Antiepileptic Drugs Induce ECG Changes and SUDEP? References

757 757 758 759 781 781 762 763 764 764 764 765 766 767 768 768 768 769 769 769 772 775 775 782

The role of antiepileptic drugs in the cause and/or prevention of sudden death in epilepsy (SUDEP) is not known. Each patient, each antiepileptic drug, and the various combinations of antiepileptic drugs need to be examined carefully to determine how best to individualize the beneἀt/risk ratio in persons with epilepsy. An emphasis on personalized medical care in persons with epilepsy helps to minimize the risk of both unwanted antiepileptic drug 755

756 Sudden Death in Epilepsy: Forensic and Clinical Issues

side effects and seizures and will eventually be important in the prevention of SUDEP. The role of genetic typing in this process is unclear. Although antiepileptic drugs produce an essential beneἀcial effect in the treatment of epilepsy (Table 47.1), their use is also associated with unwanted side effects and/or drug pharmacokinetic interactions. Combination antiepileptic drug therapy is used for persons with epilepsy who do not respond to monotherapy (Perucca 2006). In addition, non-antiepileptic drug medications may be prescribed simultaneously to treat other concurrent diseases. Antiepileptic drugs such as carbamazepine, phenytoin, phenobarbital, and primidone may induce cytochrome P450 and glucuronyl transferase enzymes. This action of the antiepileptic drugs can reduce serum concentrations of drugs given simultaneously if they are metabolized by these enzyme systems. Drugs at risk of higher levels€when used in€combination with some other antiepileptic drugs include lamotrigine; tiagabine; several steroidal drugs; cyclosporin A; oral anticoagulants; and many cardiovascular, antineoÂ� plastic, and psychotropic agents. Although not an enzyme inhibitor, valproic acid may inhibit the metabolism of selected substrates such as phenobarbital and lamotrigine. The newer antiepileptic drugs are less likely to induce or inhibit the activity of cytochrome P450 and glucuronyl transferase enzymes. Oxcarbazepine, lamotrigine, felbamate, and topiramate (high dose) may increase the metabolism of oral contraceptive steroids, resulting in unwanted pregnancies. The newer antiepileptic drugs levetiracetam, gabapentin, Table 47.1â•… Summary of Studies by Seizure Type/Syndromea Category

Class of Study

Efficacy/Effectiveness

Seizure Type

I

II

III

Alphabetical Order/Level

Adults: partial onset

2

1

30

Children: partial onset

1

0

17

Elderly adults: partial onset

1

1

2

Adults: generalized onset tonic–clonic

0

0

23

Children: generalized onset tonic–clonic

0

0

14

Children: absence

0

0

6

Benign epilepsy with centrotemporal spikes

0

0

2

A: CBZ, PHT B: VPA C: GBP, LTG, OXC, PB, TPM, VGB A: OXC B: None C: CBZ, PB, PHT, TPM, VPA A: GBT, LTG B: None C: CBZ A: None B: None C: CBZ, LTG, OXC, PB, PHT, TPM, VPA A: None B: None C: CBZ, PB, PHT, TPM, VPA A: None B: None C: ESM, LTG, VPA A: None B: None C: CBZ, VPA

Source: Glauser, T., et al., Epilepsia, 47, 1094–1112, 2006. With permission. Note: CBZ, carbamazepine; EXM, ethosuximide; GBP, gabapentin; LTG, lamotrigine; OXC, oxcarbazepine; PB, phenobarbital; PHT, phenytoin; TPM, topiramate; VPA, valproic acid; VGB, vigabatrin. a Juvenile myoclonic epilepsy was not included in this table because no class I, II, or III efficacy/effectiveness studies were found.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

757

and pregabalin do not appear to be associated with unwanted drug pharmacokinetic interactions. Pharmacodynamic drug interaction effects also include the neurotoxic actions of concomitant use of antiepileptic drugs possessing similar mechanisms.

47.1â•… Antiepileptic Drug Benefits vs. Epilepsy The International League against Epilepsy recently published guidelines (Glauser et al. 2006) for the initial monotherapy treatment of seizures using evidenced-based analysis of antiepileptic drugs efficacy. After a review of all applicable articles from 1940 to mid2005, recommendations were based on evidence provided by four classes of studies. Class I required double-blind randomized controlled trials with 48 or more weeks of treatment without forced exit criteria, evidence of efficacy based on 24 or more weeks free from seizure or evidence of effectiveness deἀned as 48 or more weeks retention data, demonstration of superiority or 80% power to detect a 20% or less relative difference in efficacy/effectiveness compared with an adequate comparator, and appropriate statistical analysis. Class II studies met the class I criteria, with the exception of having either treatment duration of 24–47 weeks or, for noninferiority analysis, a power to only exclude a 21–30% relative difference. Class III studies included other randomized double-blind and open-label trials. Class IV studies were based on other evidence (e.g., expert opinion or case reports). The authors noted that there was a discouraging lack of class I and II studies in all seizure€type and epilepsy syndromes addressed: adults, children and elderly adults with€partial-onset seizures;€children with absence seizures; children with benign epilepsy€with centrotemporal spikes. In addition, no class I, II, or III randomized controlled trials exist for juvenile myoclonic epilepsy. The summary recommendations were based on the results of 50 randomized controlled trials and 7 meta-analyses, with what was described as a rigorous method of assessment, resulting in a summary of the class of study being applied equally to all seizure types/syndromes along with the antiepileptic drugs that were given a recommendation grade of A, B, or C. A summary table based on that published in the International League against Epilepsy guidelines follows.

47.2â•…Antiepileptic Drugs’ Unwanted Side€Effects and Drug Interactions Sodium valproate has been used successfully for over three decades primarily to treat idiopathic generalized epilepsy and juvenile myoclonic epilepsy. Data from pregnancy registers demonstrate that this antiepileptic drug is teratogenic and that its administration to mothers early in pregnancy may be associated with neural tube defects, neurodevelopmental delay, and autistic spectrum disorders in the children born to them. While prenatal administration of folic acid may reduce the chance of a neural tube defect, Duncan (2007) suggests there may be a pharmacogenetic component to the teratogenic and neurodevelopmental effects. As appropriate genetic screening becomes available in the future, we may be able to identify which females are at risk of giving birth to children with these developmental defects in order to avoid exposure to potentially teratogenic effects. At the other end of the age spectrum, a major concern is the effect of aging on the metabolism of antiepileptic drugs. The elderly population appears to be at a high or higher

758 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 47.2â•… Drug Interactions with Valproic Acid Metabolic Inhibitors Lamotrigine, phenobarbital, carbamapzine-10-11epoxide, lorazepam, nimodipine, zidovudine Increased valproic acid levels

Concomitant Use

Drugs that Induce Liver Enzymes

Coadministered

Phenytoin, diazepam, warfarin, amitriptyline, chlorpromazine

Phenytoin, phenobarbital, primidone, carbamazepine

Felbamate, stiripentol, aspirin, naproxen, phenylbutazone, isoniazid, fluoxetine, chlorpromazine

Increased valproic acid levels

Increased valproic acid clearance

Increased valproic acid levels

Source: Stephen, L. J., Drugs Aging, 20, 141–152, 2003. With permission.

risk for new onset of epilepsy than was previously assumed. Data from studies investigating the effect of aging on metabolism of valproic acid are inconsistent. For example, one study suggested that no change occurs in the volume of distribution and elimination halflife in the elderly when compared with younger patients, whereas another study found an increase in both parameters for the elderly. Total valproic acid clearance is similar in both populations. Valproic acid drug interactions are summarized in Table 47.2 (Stephen 2003). In general, most of the older patients require lower doses of valproic acid than younger adults to maintain a therapeutic blood level, although higher doses may be needed if valproic acid is prescribed with drugs that induce hepatic enzymes. The controlled-release preparation of valproic acid improves compliance in the elderly and may be more effective than the standard dosage form in eliminating seizures. Dose-dependent and idiosyncratic reactions may occur more frequently in the elderly, with common adverse effects including gastrointestinal symptoms and tremor. Slow-dose escalation and controlled-release preparations may also minimize these unwanted effects of valproic acid. Stephen (2003) concluded that clinical trials are needed in the elderly to understand more clearly how best to use valproic acid. Fukuoka (2004) compared the routine steady-state antiepileptic drugs’ serum concentration (Ct) to those with concomitant use of valproic acid, carbamazepine, zonisamide, phenobarbital, and phenytoin. In general, antiepileptic drug levels were decreased, but in a few cases (phenobarbital + valproic acid or carbamazepine or phenytoin), increased levels resulted (Table 47.3). 47.2.1â•… Phenytoin The nonlinear pharmacokinetics of phenytoin represents a challenge for the physician to determine a safe and effective administration regimen for patients of all ages (Bachmann and Belloto 1999). When using antiepileptic drugs in the elderly, one must consider that hepatic oxidation of phenytoin may be slowed and that the differential disposition kinetics affect clinical levels. Because of these two factors, coupled with the fact that most often the elderly are subject to polypharmacy, Bachmann and Belloto (1999) examined the extent to which age affects plasma protein binding of phenytoin, its hepatic metabolism, and its pharmacokinetic proἀle. The authors addressed the question of whether the elderly are subject to more frequent and/or more severe drug interactions when given phenytoin vs. interactions that occur in younger adults. Interestingly, they concluded that no special

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

759

Table 47.3â•… Steady-State Antiepileptic Drug Serum Concentrations Antiepileptic Drugs Valproic acid + carbamazepine Valproic acid + phenobarbital Valproic acid + phenytoin Carbamazepine + phenobarbital Carbamazepine +phenytoin Zonisamide + valproic acid Zonisamide + carbamazepine Zonisamide + phenobarbital Zonisamide + phenytoin Phenobarbital + valproic acid Phenobarbital + carbamazepine Phenobarbital + phenytoin Phenytoin + valproic acid Phenytoin + carbamazepine Phenytoin + zonisamide Phenytoin+ phenobarbital

Ct Level/Dose Ratios 0.81 0.88 0.83 0.77 0.71 0.87 0.85 0.85 0.80 1.47 1.18 1.19 0.89 0.91 0.90 0.84

Source: Fukuoka, N., Yakugaku Zashi, 124, 443–450, 2004. With permission.

attention needs to be paid to initiation of phenytoin administration in elderly patients taking multiple anticonvulsants, a conclusion that is in contrast to the practice of neurologists experienced in the care of elderly patients with epilepsy. For the elderly receiving phenytoin monotherapy, the initiation of any antiepileptic drugs should occur at doses lower than used for younger adults and blood concentrations should be monitored to evaluate individual Vmax and Km values to allow informed dosage adjustments. 47.2.2â•… Carbamazepine The effect of concurrent anticonvulsant therapy on free and total (carbamazepine) and carbamazepine-epoxide plasma levels was studied in 113 patients with epilepsy on phenobarbital, phenytoin, valproic acid, or combinations thereof (Ramsay et al. 1990). Polytherapy exhibited variable effects on free and total carbamazepine plasma levels compared to monotherapy. Coadministered phenytoin and the other three antiepileptic drugs decreased free and total carbamazepine plasma levels. However, no change was found with coadministration of valproic acid. For polytherapy vs. monotherapy, free and total carbamazepine-epoxide levels increased with coadministered valproic acid, less with coadministered phenobarbital, but exhibited no change with coadministered phenytoin or when coadministered with other antiepileptic drugs. None of the antiepileptic drugs studied decreased the protein binding of carbamazepine or carbamazepine-epoxide. Free and total carbamazepine-epoxide ratio tripled after administration of valproic acid or co-Â�antiepileptic drugs. No difference in half-life (t1/2), plasma clearance, or volume of distribution was found. Autoinduction could explain the changes observed after chronic administration of carbamazepine. In a more recent study, Fukuoka et al. (2003a, 2003b) analyzed the effects of concomitant antiepileptic drugs on serum concentrations of carbamazepine, given as ἀne granules/ tablets (Table 47.4). A total of 119 patients were given carbamazepine alone; and 91, 39,

760 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 47.4â•… Concomitant Antiepileptic Drug Effect on Serum Levels of Carbamazepine Antiepileptic Drugs Carbamazepine + phenobarbital Carbamazepine + phenytoin Carbamazepine + valproic acid Carbamazepine + zonisamide Carbamazepine + Primidone, Clonazepam, ethosuximide

Carbamazepine Ct = D/VECWa 0.770 0.710 No effect No effect Number of patients insufficient to detect effect

Source: Fukuoka, N., et al., Yakugaku Zasshi, 123, 35–42, 2003a. With permission. Note: Ct, concentration; D, daily dose; VECW, extracellular water volume. a Postulated model to show Ct is affected by each concomitant antiepileptic drugs at each deἀnite ratio. Linear polynomial expression used to convert to common logarithms and used for multiple regression analysis. When the ideal body weight or the extracellular water volume was used as the transforming factor, the level/ dose ratio was independent of the age and gender of the patient.

19, and 6 were coadministered one, two, three, and more than four different antiepileptic drugs, including phenobarbital, phenytoin, ethosuximide, and zonisamide. Both phenobarbital and phenytoin decreased serum concentrations of carbamazepine. Although carbamazepine may induce mild hyponatremia, severe hyponatremia in persons taking it as monotherapy is relatively uncommon. Factors that increase the risk of€hyponatremia are age >40 years, concomitant use of medications associated with the development of hyponatremia, female gender, menstruation, psychiatric conditions, surgery, and psychogenic polydipsia. Prevention of carbamazepine-induced hyponatremia includes an awareness of the risk factors, monitoring of the serum electrolyte levels, and education of the patients who take carbamazepine about restricting excessive free water ingestion and not restricting salt intake. Carbamazepine-induced acute hyponatremia and subsequent tonic–clonic seizures occurred in one patient after taking two times her normal evening dose (Kuz and Mansourian 2005). A preadmission carbamazepine concentration was 8.6 µg/mL and on admission was 11.3. The concentration fell to 5.6 the next day when carbamazepine was withheld. The serum sodium level was 122 mEq/L on admission. Infusion of NaCl 0.9% was initiated and the level rose to 136 mEq/L 24 h later. In a much larger study of hyponatremia in 451 patients treated with carbamazepine, 13.5% exhibited hyponatremia (Na+ ≤134 mEq/L) (Dong et al. 2005). The hyponatremia was severe in 2.8% of carbamazepine-treated persons. Of 97 oxcarbazepine-treated patients, hyponatremia was found in 29.9% and was considered to be severe in 12.4%. Advanced age was also a risk factor for hyponatremia. Studies have found that some anticonvulsant drugs are associated with changes in bone metabolism. Pack and Morrell (2004) noted that adult persons with epilepsy are at increased risk for osteopenia and osteoporosis because of altered bone metabolism associated with antiepileptic drugs. Increased fracture rates may be due to both the physical trauma risks of seizures combined with the adverse bone metabolism side effects of antiepileptic drugs. Phenytoin, phenobarbital, and primidone, all hepatic enzyme–inducing drugs, are commonly associated with decreased bone mineral density. The authors recommend prophylactic use of calcium and vitamin D and evaluation of bone mineral density after 5 years of antiepileptic drug treatment and before treatment in the case of postmenopausal women. Carbamazepine and oxcarbazepine were demonstrated to signiἀcantly lower 25-hydroxyvitamin D levels when compared with normal controls. The pattern of changes in other bone biomarkers was suggestive of secondary hyperparathyroidism.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

761

Mintzer et al. (2006) recommended that it may be prudent for patients on carbamazepine or oxcarbazepine to be prescribed 25-hydroxyvitamin D replacement. In addition, bone mineral content and/or density may be decreased in children with epilepsy. A review of published studies by Gissel (2007) found varying reports for phenytoin, phenobarbital, valproate, and carbamazepine. Longitudinal studies are needed to clarify the problem and to determine potential interventions in children with decreased bone mineral density. Both carbamazepine and oxcarbazepine reduced serum thyroid hormone concentrations in females diagnosed with epilepsy. Patients on valproic acid, an antiepileptic drug with no liver enzyme–inducing properties, demonstrated normal serum thyroid hormone and increased thyrotropin levels. Changes in both serum thyroid hormone and thyrotropin levels were reversible after withdrawal of the offending antiepileptic drugs (Vainionpaa et al. 2004). An earlier study found similar changes for carbamazepine and valproic acid. Phenytoin was also shown to decrease serum thyroid hormone levels, as was carbamaÂ� zepine; it was theorized that both drugs accelerated hepatic plasma clearance of these hormones because of antiepileptic drug–related induction of hepatic microsomal enzyme systems (Isojarvi et al. 1992).

47.3â•…Newer Antiepileptic Drugs These drugs include lamotrigine, zonisamide, oxcarbazepine, vigabatrin, gabapentin, felbamate, tiagabine, eterobarb, remacemide, stiripentol, topiramate, levetriacetam, pregabalin, and retigabine. 47.3.1â•…Lamotrigine Lamotrigine is chemically unrelated to carbamazepine, phenytoin, phenobarbital, and valproic acid. May et al. (1996) examined 588 blood samples from 302 patients and found that comedication exerted a signiἀcant effect on serum lamotrigine levels. Coadministration of valproic acid increased lamotrigine levels 2-fold when compared with lamotrigine monotherapy or on comedication with drugs other than valproic acid. Correlations of serum concentrations and doses of carbamazepine, phenobarbital, phenytoin, and valproic acid with the level/dose ratio of lamotrigine were only weak or not signiἀcant. As shown in Table 47.5, comedication with other antiepileptic drugs altered the lamotrigine concentration and must be considered when selecting the lamotrigine dosage. See below for Table 47.5â•… Lamotrigine Level Altered by Comedication with Other Antiepileptic Drugs Antiepileptic Drugs

Level/Dose Ratio (mg/mL/mg/kg)

Lamotrigine Lamotrigine + phenytoin Lamotrigine + phenobarbital Lamotrigine + carbamazepine Lamotrigine + valproic acid + phenytoin Lamotrigine + valproic acid + carbamazepine Lamotrigine + valproic acid + phenobarbital Lamotrigine + valproic acid Source: May, T. W., et al., Ther Drug Monit, 18, 523–531, 1996. With permission.

0.32 0.52 0.57 0.98 0.99 1.67 1.80 3.57

762 Sudden Death in Epilepsy: Forensic and Clinical Issues

discussion of patient management related to free and total serum antiepileptic drug levels to avoid the potential of SUDEP from subtherapeutic free levels (George and Davis 1998; Tomson et al. 1998). 47.3.2â•… Zonisamide Shinoda et al. (1996) recommended, when administering zonisamide concomitantly with other antiepileptic drugs, monitoring the plasma concentration of zonisamide to readjust its dosage. The effect of coadministered antiepileptic drugs on the serum concentrations of zonisamide was examined using samples obtained from 175 patients with epilepsy treated with oral administrations of zonisamide (Fukuoka et al. 2003a, 2003b). The extracellular water volume was used as a transforming factor to provide level/dose ratios independent of age and sex for the administration of zonisamide alone. The data indicate that the concomitant antiepileptic drugs studied signiἀcantly lowered the level/dose ratio (Table 47.6). Patsalos and Sander (1994) discussed 12 new antiepileptic drugs: lamotrigine, vigabatrin, gabapentin, oxcarbazepine, felbamate, tiagabine, eterobarb, zonisamide, remacemide, stiripentol, topiramate, and levetiracetam; oxcarbazepine, remacemide, and eterobarb€are prodrugs. Lamotrigine, along with felbamate, and stiripentol exhibit signiἀcant drug€interactions in contrast to vigabatrin, gabapentin, and topiramate, which are excreted mainly unchanged in urine and are therefore, most importantly, not susceptible to signiἀcant pharmacokinetic interactions. Note that newer data suggest that drug interactions can exert an effect on the actions of topiramate (Johannessen and Tomson 2006). Central nervous system effects occur with all of the drugs. It is less of a problem with gabapentin, remaÂ� cemide, and levetiracetam. Newer antiepileptic drugs including felbamate, gabapentin, lamotrigine, levetiracetam, oxcarbazepine, pregabalin, tiagabine, topiramate, vigabatrin, and zonisamide possess pharmacokinetic properties that are more predictable and thus these drugs may be safer than the older antiepileptic drugs phenytoin, carbamazepine, and valproic acid (Johannessen and Tomson 2006). Because of the interindividual variability in the pharmacokinetic properties and narrow therapeutic range for these older antiepileptic drugs, therapeutic monitoring should be used to determine if the serum levels are those predicted to produce an optimal response in the patient. The question has been raised as to whether monitoring of drug levels is necessary with the newer drugs. Johannessen and Tomson (2006) note that even with the newer antiepileptic drugs, monitoring is used to determine the individual reference concentrations based on intraindividual comparisons of drug serum concentrations. Thus, the therapeutic drug monitoring program may be used whether or not there is a well-deἀned therapeutic range. Because the newer antiepileptic Table 47.6â•… Effect of Antiepileptic Drug Coadministration on Zonisamide Levels Antiepileptic Drugs Zonisamide + phenytoin Zonisamide + phenobarbitala Zonisamide + valproic acida Zonisamide + carbamazepinea Zonisamide + phenytoina

Level/Dose Ratio[C(t)/Dose/Volume (Extracellular Cellular Water)] 0.76 0.849 0.865 0.846 0.804

Source: Shinoda, M., et al., Biol Pharm Bull, 19, 1090–1092, 1996. With permission. Assumed each value independent from one another and multiplicative.

a

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

763

Table 47.7â•… Therapeutic Drug Monitoring Recommendations for Antiepileptic Drugs Drug

Therapeutic Drug Monitoring Needed?

Valproic acid Phenytoin Carbamazepine Oxcarbazepine

Yes Yes Yes Yes

Felbamate, lamotrigine, tiagabine, zonisamide, oxcarbazepine Gabapentin

Yes

Pregabalin

Yes

Vigabatrin

Yes

Gabapentin, pregabalin, vigabatrin, topiramate, levetiracetam Topiramate

Yes

Levetiracetam Vigabatrin

Yes

Yes Yes Only to check compliance

Rationale Interindividual variability in pharmacokinetics Interindividual variability in pharmacokinetics Interindividual variability in pharmacokinetics Prodrug (inactive) so must measure 10-╉ monohydroxy oxcarbazepine, the active liver metabolite Metabolized so pharmacokinetic variability can be a problem

Eliminated via kidneys completely unchanged, pharmacokinetics less variable, more predictable Dose-dependent absorption increases pharmacokinetic variability Eliminated via kidneys completely unchanged, pharmacokinetics less variable, more predictable Eliminated via kidneys completely unchanged, pharmacokinetics less variable, more predictable Individual factors of age, pregnancy, renal function contribute to all renally excreted antiepileptic drugs Eliminated via kidneys mainly unchanged, drug interactions can affect topiramate concentrations markedly Eliminated via kidneys mainly unchanged Unique action mode means no clear relationship between drug concentration and clinical effect

Source: Johannessen, S. I., and Tomson, T., Clin Pharmacokinet, 45, 1061–1075, 2006. With permission.

drugs all possess different pharmacological properties, the value of therapeutic monitoring must be assessed individually. The authors conclude that, even for the newer antiepileptic drugs, therapeutic monitoring is likely to be useful in many clinical settings. A summary of their conclusions is presented (Table 47.7). 47.3.3â•…Levetiracetam Levetiracetam is a relatively new antiepileptic drug approved for use as an adjunct agent in partial-onset seizures in adults and children 4 years or older (Sirsi and Safdieh 2007). The most common adverse effects reported are central nervous system based, such as somnolence, asthenia, and dizziness. These side effects are detected soon after drug initiation and usually resolve without withdrawal of the antiepileptic drugs. Adverse behavioral effects are the most serious side effects recorded and are more common in children or patients with a prior psychiatric history. Dooley and Plosker (2000) also note that headache may occur with administration of levetiracetam. When levetiracetam was given as adjunctive

764 Sudden Death in Epilepsy: Forensic and Clinical Issues

therapy with other antiepileptic drugs, digoxin, warfarin, probenecid, or oral contraceptives, no adverse drug interactions were reported. 47.3.4â•… Pregabalin Pregabalin is a new antiepileptic drug for refractory epilepsy. Pregabalin is the pharmacologically active S-enantiomer of 3-aminomethyl-5-methylhexanoic acid (Warner and Figgitt 2005). The site of action is a presynaptic calcium channel site where it binds with high affinity and speciἀcity to voltage-gated calcium channel alpha (2)-delta proteins to modulate the neurotransmitter release in the central nervous system. It is thought to reduce excitatory neurotransmitter release caused by binding to the alpha (2)-delta protein to initiate allosteric modulation of P/Q-type voltage-gated calcium channels. This mechanism of action is similar to that of gabapentin. Adverse effects are dose related and most often include somnolence, dizziness, and ataxia (Beydoun et al. 2005; Hamandi and Sander 2006). Only with the highest dose of 600 mg/day did some patients report weight gain. 47.3.5â•…Retigabine Retigabine has been evaluated for efficacy and safety as an adjunctive therapy in patients with partial-onset seizures (Porter et al. 2007). Adjunctive therapy was well tolerated and reduced the frequency of partial-onset seizures in a dose-related response. The most common adverse effects were somnolence, dizziness, amnesia, abnormal thinking, abnormal gait, paresthesias, and diplopia.

47.4â•… Pregnancy, Epilepsy, and Antiepileptic Drug Use Most seizure disorders necessitate years, if not a lifetime, of antiepileptic drug use. For females of child-bearing age, unwanted interactions of the antiepileptic drugs with hormonal contraception drugs must be considered. Decreased lamotrigine serum concentrations occur during hormonal contraception and pregnancy. The potential for adverse effects on both a mother and the fetus necessitates the evaluation of antiepileptic drugs on fertility, pregnancy, delivery, the postpartum period, and teratogenicity (Steinhoff 2008). Harden and Leppik (2006) note that no evidence exists to suggest that the use of oral contraceptives per se increases seizure activity and that, for most oral contraceptives, the protective effect against pregnancy is not altered by use of noninducers of hepatic enzymes. However, oral contraceptive failure arises with antiepileptic drugs that induce enzyme cytochrome P450 3A4. These antiepileptic drugs include phenobarbital, carbamazepine, phenytoin, felbamate, topiramate, and oxcarbazepine. In particular, felbamate induces metabolism of the progesterone component of oral contraceptives, whereas topiramate induces metabolism of only the estrogen component. Lamotrigine appears to induce the metabolism of the progestin levonorgestrel; thus, to ensure maximal pregnancy prevention when taking an antiepileptic drug that induces cytochrome P450, the oral contraceptive should contain at least 50 µg of ethinyl estradiol and low-dose formulations should not be used. Lamotrigine levels appear to be reduced by 50% in the setting of oral contraceptive use. Thus, women with epilepsy taking lamotrigine must be monitored for seizures when oral contraceptives are started and for toxicity when the oral contraceptives are

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

765

discontinued. Doses need to be adjusted to maintain clinical stability. The free week of the oral contraceptive regimen is also a time when clinical toxicity may occur.

47.5â•… Elderly, Epilepsy, and Antiepileptic Drug Use The elderly population is at a high or higher risk for onset of epilepsy (Stephen 2003). In addition, the issue of drug–drug interactions is a clinical problem when treating geriatric Table 47.8â•… Drug Interactions in the Elderly Drug Action Antidepressants Antiepileptic drugs with enzyme-inducing properties Oral anticoagulants

Other drugs affected by enzyme inducers

Antiepileptic drugs

Cardiovascular drugs Potential for metabolic drug interactions Cardiovascular drugs

Enzyme-inducing antiepileptic drugs Newer antiepileptic drugs

Drug Name/Class

Interaction/Effect

Citalopram, escitalopram, venlafaxine, duloxetine, mirtazapine Carbamazepine, phenytoin, phenobarbital

Least potential for altering antiepileptic drug metabolism

Carbamazepine, phenytoin, phenobarbital Selective 5-HT reuptake inhibitors fluoxetine (especially), gemἀbrozil, fluvastatin, lovastatin Include cytochrome P450 3A4 substrates, such as the calcium channel blockers nimodipine, nilvadipine, nisoldipine, felodipine and the 3-OH-3-methylglutaryl coenzyme A reductase inhibitors atorvastatin, lovastatin, simvastatin Ticlopidine Coadministration of ticlopidine with carbamazepine, phenytoin

Decreased prothrombin time; however, an increased prothrombin time was also reported for phenytoin Increased risk of bleeding with oral anticoagulants Levels are affected

Diltiazem or verapamil Diltiazem Cholinesterase inhibitors, but with enzyme-inducing antiepileptic drugs N-methyl-d-aspartate receptor antagonist memantine Antihypertensives such as angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, hydrophilic beta blockers, thiazide diuretics Lipophilic beta-blockers Levetiracetam and gabapentin

Reduction in levels of donepezil, galantamine, warfarin

No alteration of ticlopidine metabolism Toxicity to carbamazepine and phenytoin induced Carbamazepine levels elevated Phenytoin toxicity induced Lower potential for metabolic drug interactions Theoretical risk of reduction in donepezil and galantamine levels Lower potential for metabolic drug interactions Lower potential for metabolic drug interactions Moderate risk that levels of lipophilic beta-blockers may be decreased Have a lower potential for drug interactions Not reported to alter enzyme activity

Source: Levy, R. H., and Collins, C., Int Rev Neurobiol, 81, 235–251, 2007. With permission.

766 Sudden Death in Epilepsy: Forensic and Clinical Issues

epilepsy patients. Because these individuals most often have other concurrent diseases such as cardiovascular and psychiatric disorders (e.g., dementia and depression), they are treated simultaneously with numerous other classes of drugs along with antiepileptics (Table 47.8, Levy and Collins 2007).

47.6â•… Cognitive and Behavior Actions of Antiepileptic Drugs Cognitive impairment, especially memory impairment, is a common complaint of patients with epilepsy (Lathers et al. 2003). Factors involved in producing the cognitive impairment include the underlying etiology of epilepsy, the physiologically disruptive effects of seizures themselves (Lathers and Schraeder 2003), and the unwanted central nervous system side effects of the antiepileptic drugs (Aldenkamp et al. 2003). Antiepileptic drugs may affect learning in children, driving ability in adults, and memory in elderly persons. Animal models and studies of humans indicate that prolonged seizures are associated with cellular injury and cognitive impairment in epilepsy. Status epilepticus is thought to cause cerebral injury and cognitive dysfunction if it is of long duration. However, electroconvulsive therapy studies do not support the concept that recurrent limited seizures alone produce a decline in cognitive function. Some prospective studies do not support the premise that cognitive impairment develops or progresses in a population of persons with epilepsy. If impairment is present, the origin is likely to be multifactorial with epileptiform discharges being a major factor, in addition to the effect of the seizures themselves and/or the antiepileptic drugs prescribed (Hoch et al. 1994). In addition to the role of ictal effects on cognitive function, one must also consider adverse effects of both interictal activity and the postictal state (Aldenkamp 1997). This is especially true when looking for an explanation of substantial cognitive impairments in children with subclinical epileptiform discharges or with infrequent subtle seizures. Engel et al. (1991) discussed the possibility that antiepileptic drugs may produce interictal behavioral disturbances in patients with epilepsy by indirect mechanisms. Aberrant behaviors due to medication-induced systemic disorders, neuroendocrine dysfunction, or REM sleep deἀcit may occur. Depression after successful treatment with drugs, as well as after surgery, may be associated with cessation of seizures. In examining the effect of seizure per se on interictal behavior, one must consider that some interictal behavioral disturbances may actually be secondary to subclinical ictal events. Brown (2006) summarized causes of intellectual deterioration in persons with epilepsy (Table 47.9). Aldenkamp and Bodde (2005) note that in addition to cognitive changes, behavioral impairments have also been reported to occur with even single seizures. If a given person has high seizure frequency, these impairments may accumulate and exert an unwanted impact on daily life. The risk of behavioral impairments is increased for some seizure types, such as secondary generalized seizures. For all epilepsy types, increased cognitive risk is associated with persistent or poorly controlled seizures. Cognitive impairments induced by seizures are reversible for most seizure types if the seizures are well controlled, but there may be a duration of seizure activity, which, if exceeded, may result in irreversible impairment, behooving the physician to make every effort to obtain complete and early seizure control. However, because almost all antiepileptic drugs may affect cognitive function, use of antiepileptic drug polypharmacy may aggravate the risk of cognitive impairment. Thus, clinicians need to not only control seizures but also to use a more comprehensive

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

767

Table 47.9â•… Causes of Cognitive Intellectual Deterioration in Persons with Epilepsy Concomitant degenerative neurological disease, of which epilepsy is a symptom. Direct effects of seizures, abnormal EEG on brain function. Traumatic brain injury secondary to seizures, including status epilepticus. Effect of antiepileptic drugs on cognitive function. Effect of surgery on cognitive function. Psychosocial sequelae of diagnosis. Various combinations of causes 2–6. Presence of epileptogenic lesion. Developmental factors: other brain areas may interact in different ways and at different times with partial seizures, including irritative and function-deἀcit zones. Abnormal activity at critical times may interact and disrupt pathways for frontal lobe and limbic system function, causing long-term cognitive deἀcits. Source: Brown, S., Epilepsia, 47S2, 19–23, 2006. With permission.

approach to prevent cognitive side effects of the seizures and/or antiepileptic drugs. A balance between the beneἀt of seizure control and the risk of cognitive and behavioral effects must be obtained. Brief screening tools such as the 15-min EpiTrack cognitive function evaluation have been developed to detect and track cognitive side effects of medications and/or the effect of seizures on attention and executive functions in patients with epilepsy (Lutz and Helmstaedter 2005). Most antiepileptic drugs can produce cognitive side effects including impaired attention, vigilance, and psychomotor speed. Although use of older antiepileptic drugs such as bromides, phenobarbital, and benzodiazepines is predictably associated with impaired cognitive effects (Meador 1994), in most patients, antiepileptic drug monotherapy with therapeutic anticonvulsant blood levels usually is associated with minor cognitive side effects. In general, the side effects increase with polypharmacy, rapid initiation, and with high doses and antiepileptic drug blood levels (Meador 2002). Although the magnitude of cognitive effects is relatively modest for therapeutic levels of most antiepileptic drugs, even this impairment can have clinical signiἀcance and impact negatively the patient’s quality of life. An ongoing risk–beneἀt analysis must be done for each patient to determine whether maximum seizure control has been attained with minimal unwanted cognitive side effects. Drane and Meador (1996) recommend baseline evaluation of mental function and repetition of the test when a change in cognitive performance is suspected. Cognitive side effects of phenytoin, carbamazepine, and valproate appear modest when dosages are within standard therapeutic ranges especially when polypharmacy is avoided. If a given patient requires polypharmacy or antiepileptic drug doses that exceed the standard therapeutic range for optimal seizure management, special attention should be paid to monitoring the patient for increased risk of cognitive side effects. Children and the elderly are at particular risk for cognitive effects. During the formative years of the intellectual development of a child, attention should be focused on the possibility that subtle attention or arousal deἀcits may contribute to cumulative deἀcits in learning or memory. If a patient is refractory or intolerant to antiepileptic drug therapy, referral for video EEG monitoring will conἀrm the seizure diagnosis and determine if the patient is a candidate for epilepsy surgery (Meador 2002).

47.7â•… Cognitive Effects of Older vs. Newer Antiepileptic Drugs The cognitive effects of the older antiepileptic drugs phenytoin, phenobarbital, valproic acid, and carbamazepine were compared with those of the newer antiepileptic drugs

768 Sudden Death in Epilepsy: Forensic and Clinical Issues

oxcarbazepine, vigabatrin, lamotrigine, zonisamide, gabapentin, tiagabine, topiramate, and levetiracetam (Brunbech and Sabers 2002). Altered cognition may reflect a chronic adverse effect of antiepileptic drugs. However, subjective complaints about cognitive deficits such as memory problems or attention may reflect subtle disturbances of cerebral function secondary to the epilepsy and its underlying etiology and other aspects of adverse effects not directly related to disturbances of speciἀc cognitive functions, that is, mood and anxiety disorders. In general, the new antiepileptic drugs, with a few exceptions, manifested fewer cognitive side effects than the older agents (Lathers et al. 2003). 47.7.1â•…Oxcarbazepine Aldenkamp et al. (2003) suggest oxcarbazepine appears to be associated with a relatively low risk of impaired cognitive function in healthy volunteers or in adults with newly diagnosed epilepsy. However, the PDR states that there is a 26% risk of cognitive side effects vs. 12% for placebo (see The Physician’s Desk Reference web site, www.pdr.net). 47.7.2â•… Topiramate Although topiramate is listed as a newer antiepileptic drug, it is an exception to the generalization that newer drugs have a low cognitive side effect proἀle because it exhibits a relatively high rate of cognitive adverse effects. Kockelmann et al. (2003) used a battery of neuropsychological tests to evaluate verbal fluency task, verbal (Wechsler’s digits) and spatial spans [Corsi block-tapping and Trail Making Test (parts A and B)], before and after withdrawal of these antiepileptic drugs. Twenty patients with epilepsy were studied and compared to a matched group of patients who had been tested and retested before and after reduction of antiepileptic drugs other than topiramate. After withdrawal of topiramate, there was signiἀcant improvement in frontal lobe function associated with verbal fluency and working memory. Cognitive performance was not correlated to current daily dosages/current blood serum levels of topiramate. The authors conclude that cognitive impairment due to topiramate seems to be easily overlooked and as a result is underestimated. Prescribing physicians should consider the possibility of cognitive effects when selecting an antiepileptic drug and should be able to recognize these unwanted cognitive effects. Lee et al. (2003) examined the effects of adjunctive topiramate on cognitive function in patients with epilepsy and found topiramate use was associated with a decline in fluency, attention/ concentration, processing speed, language skills, and perception. Working memory was affected, but retention was not altered. Kelly et al. (2002) demonstrated that topiramate was effective as an add-on therapy in learning-disabled people with difficult-to-control epilepsy. When topiramate was withdrawn, it was mainly because of cognitive side effects. 47.7.3â•… Topiramate vs. Lamotrigine Miller and Kustra (2005) examined the cognitive and behavioral effects of topiramate and€lamoÂ�trigine in healthy volunteers and found that lamotrigine produced signiἀcantly fewer untoward cognitive and behavioral effects compared to topiramate. The lack of difference in efficacy between topiramate and lamotrigine leads one to support the principle that prescription writing preference in this instance, and in others where different drugs have similar efficacy, should be based on the potential risk of side effects in individual patients.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

769

47.7.4â•… Topiramate vs. Levetiracetam Cognitive changes were studied in 30 consecutively recruited patients with focal epilepsy treated with levetiracetam and in 21 treated with topiramate (Gomer et al. 2007). Function was assessed before gradual initiation and after reaching steady state of the individual target dosage. Before titration, cognitive performance in the two patient groups did not differ signiἀcantly. After antiepileptic drug titration, those receiving levetiracetam demonstrated no change in cognitive performance, but those receiving topiramate demonstrated a decrease in the cognitive parameters of cognitive speed and verbal fluency and shortterm memory. Only topiramate impaired frontal lobe function. The relatively low risk of adverse effects on cognitive function by levetiracetam vs. topiramate offers the prescribing physician another treatment option when considering/selecting antiepileptic drugs. 47.7.5â•…Topiramate and Zonisamide vs. Tiagabine, Gabapentin, Lamotrigine, and Levetiracetam The incidence of treatment-emergent cognitive adverse events in placebo-controlled studies of carbonic anhydrase inhibitor antiepileptic drugs, topiramate and zonisamide, were compared with antiepileptic drugs that are not carbonic anhydrase inhibitors (Knudsen and Sokol 2005). Data from clinical placebo or controlled randomized add-on trials for each of the antiepileptic drugs were reviewed for concentration/memory impairment. Both topiramate and zonisamide produced a signiἀcant increase in impaired concentration/ memory incidence when compared to placebo. There was a lack of effect found for the noncarbonic anhydrase inhibitors tiagabine, gabapentin, lamotrigine, and levetiracetam vs. placebo. Carbonic anhydrase is an enzyme involved in gluconeogenesis, ion transport, and provision of HCO3−. Knudsen and Sokol theorized that carbonic anhydrase and HCO3− play a role in neuronal homeostasis, with carbonic anhydrase acting as a molecular switch in the development of synchronous gamma-frequency ἀring of hippocampal pyramidal cells. GABAergic excitation and associated gamma oscillations seem dependent on carbonic anhydrase activity, HCO3− availability, membrane pH gradients, and provision of ionized calcium. In the big picture, carbonic anhydrase may mediate some aspects of cognition. Mechanistically, carbonic anhydrase inhibitors may have an effect on the highvoltage activated voltage-sensitive calcium channels encoded by human alpha (1E) subunit (McNaughton et al. 2004). Whole cell recordings were done using HEK293 cells that express human alpha(1E)beta(3)-mediated calcium channels in a stable manner. Ethoxyzolamide, an anticonvulsant, inhibited these currents and produced an associated hyperpolarization shift in the steady-state inactivation proἀle. Acetazolamide and benzolamide produced 30–40% inhibition of alpha(1E)beta(3)-mediated Ca(2+) currents. Topiramate, another anticonvulsant with carbonic anhydrase–inhibiting action, inhibited these currents by ~68%. The authors suggest that this off-target activity of carbonic anhydrase inhibitors may contribute to some of their effects observed in both in vitro and in vivo use.

47.8â•… Effects of Interictal Spikes on Cognitive Function Subclinical epileptiform EEG discharges may have a subtle adverse effect on educational skills associated with reading, mental arithmetic, and manual dexterity. Kasteleijn-Nolst

770 Sudden Death in Epilepsy: Forensic and Clinical Issues

Trenite et al. (1988) demonstrated that the interaction between epileptiform discharges and cognitive function/deterioration is complex in that the nature and level of difficulty of the task in turn affects the rate of EEG discharge. Binnie and Marston (1992) reported a study of benign childhood epilepsy with Rolandic spikes that detected transitory cognitive impairment in most patients. However, they note that although it is not possible to claim that all persons with subclinical EEG discharges have transitory cognitive impairment that adversely alters psychosocial function, it was concluded that in those patients with epilepsy-related cognitive symptoms it may be possible to use antiepileptic drugs to try to decrease the frequency of subclinical discharges to treat the transitory cognitive impairment. Besag (1995) examined the therapeutic dilemma arising when treating subtle subclinical seizures in the absence of obvious clinical seizures and concluded that the extent to which epileptiform discharges cause temporary or permanent impairment will influence the physician’s decision of whether to treat with antiepileptic medication or surgery. Binnie (2003) noted that transitory cognitive impairment in children may be associated with behavioral disorders and suggested that the important practical issue is whether this occurrence impairs psychosocial function. He concluded that preliminary data suggested antiepileptic drugs used to treat transitory cognitive impairment do suppress discharges and were associated with signiἀcant improvement in psychosocial function. The unanswered question is whether the observed positive effects of levetiracetam in persons with epilepsy are the result of a direct psychotropic beneἀt or the consequence of its efficacy in alleviating seizures. Tremmel et al. (2006) examined the impact of subclinical discharges on memory functions, visuospatial and verbal short-term memory, in 40 seizure-free children with subclinical epileptiform discharges and were unable to conἀrm that there is any association between subclinical epileptiform discharges and error rates used to indicate transient cognitive impairment. Austin and Dunn (2002) reported that a high rate of mental health problems occurs in children with epilepsy. Four potential causal factors observed in cross-sectional studies include:

1. The seizures themselves 2. Poor child and family response to the condition 3. Antiepileptic drugs side effects 4. Neurological dysfunction that causes both seizures and behavioral problems

The authors note that prospective studies that provide the best information addressing the relative importance of these four factors in children with new-onset seizures suggest that two and three are not major factors. They also emphasize that transient cognitive impairment associated with interictal epileptiform discharges is a better alternative explanation for behavior problems. The type of epilepsy impacts on stability of cognitive functions, including educational achievement. Paroxysmal epileptiform activity, acute effects of seizures, and EEG discharges affect primarily transient mechanistic cognitive processes of alertness and mental speed. Aldenkamp and Arends (2004) emphasize the importance of early detection of the cognitive impact of seizure-related activity because subsequent treatment may prevent the detrimental impact on cognitive and educational development. Aldenkamp (2001) raised the question of when cognitive and behavior assessment should be done in clinical trials, noting the effects of uncontrolled seizures are often larger than the effects of drug

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

771

treatment; thus, adverse cognitive antiepileptic drug effects may be masked by beneἀcial effects of better seizure control. Aldenkamp used the drug withdrawal study as the optimal design to assess behavioral drug effects because the subjects are ἀrst tested while still on medication and then tested after withdrawal. A larger gain in the epilepsy group relative to the control groups is considered evidence supporting a reversible impairment, attributable to antiepileptic drug use. In a prospective, open, clinical comparative study, Aldenkamp et al. (2005) examined impaired cognitive activation occurring during, immediately before, and immediately after epochs with epileptiform EEG discharges of 3 s or longer. No statistically signiἀcant slowing was found for the periods of postdischarge or predischarge. The type of discharge was important and the negative effects on cognitive activation were found exclusively for generalized discharges. This effect was also detected in the interictal period, outside of the peridischarge periods, and was concluded to represent the more general effect of the type of epilepsy on cognitive activation. The authors report that although short generalized epileptiform EEG discharge (duration 4.14 s; SD 1.38) may cause impairment (slowing) of cognitive activation, such was not the case for focal discharges, even of long duration. They concluded that the risk of accumulating effects with long-lasting repercussions on highorder cognitive functions seems to be negligible. Interictal spikes, corresponding to a large intracellular depolarization with evoked action potentials, can mimic a miniseizure (Holmes et al. 2006), resulting in transitory cognitive impairment with the type of deἀcit dependent on where in the cortex the spike arises. Interictal spikes, especially if frequent and widespread, can impair cognitive abilities through interference with waking learning and memory, and memory consolidation during sleep. Subtle seizures, composed of brief alterations of consciousness with or without automatisms, may not be detected during daily life activities, but EEG/video recordings will detect the subclinical epileptiform discharges capable of negatively influencing performance as well as the occurrence of epileptiform discharge evoked by mental activities. Persons conducting intense, mentally demanding professional or occupational activities could perform their job well when being challenged with recurrent, brief subtle alterations of cognitive functions (Kasteleijn-Nolst Trenite and Vermeiren 2005). This clinical evidence of interictal discharges interfering with cortical processing conἀrm the experimental observations of Schraeder and Celesia (1977) that focal interictal discharges in the sylvian cortex of the cat interfere with the ability of the brain to process auditory input by increasing the duration of the auditory cortical refractory period, thereby interfering with the ability of the brain to process closely sequenced auditory input. Not only do interictal discharges have an adverse effect on higher cortical function, they also have been demonstrated experimentally by Lathers and others (Schraeder and Lathers 1982, 1983; Lathers and Schraeder 1987; Schraeder and Lathers 1989; Lathers and Schraeder 1990; Lathers et al. 2008a, 2008b, 2008d) to be associated with a nonuniform postganglionic cardiac neural discharge, increasing the likelihood of arrhythmias and/or death. Studies in humans have recently conἀrmed these animal data (Druschky et al. 2001; Hilz et al. 2003). A lockstep phenomenon was demonstrated in which abnormal paroxysmal discharges of the autonomic nerves involved in controlling cardiac rhythmicity were time locked to interictal and ictal cortical epileptiform discharges (Lathers et al. 1987; Stauffer et al. 1989, 1990; Dodd-O and Lathers 1990; O’Rourke and Lathers 1990). Interictal and ictal activities elicited by the stereotaxic hippocampal injection of aqueous penicillin solution in this animal model were associated with changes in the electrocardiogram (Lathers and Schraeder

772 Sudden Death in Epilepsy: Forensic and Clinical Issues

1990; Lathers et al. 1993). Recent studies in rat hippocampus indicate that thyrotropinÂ�releasing hormone (TRH), a tripeptide, exhibits antiepileptic effects and regulates cognition, arousal, sleep, locomotion, and mood (Deng et al. 2007). TRH is thought to increase the action potential ἀring frequency recorded from GABAergic interneurons in CA1 stratum radiatum and induced membrane depolarization, suggesting that TRH increases the excitability of interneurons to facilitate GABA release. This may occur via inhibition of resting two-pore K+ channels. It was concluded that the TRH-mediated increase in GABA release contributes to its antiepileptic effects. As understanding of mechanism of the role of TRH in the hippocampus and the consequent effect on cognition is gained, it is anticipated that new antiepileptic drugs for use in humans may be developed that will have fewer effects on cognition.

47.9â•…Generic Antiepileptic Drugs, Unwanted Side Effects, and Risk of SUDEP? Given the continual rise in the cost of health care, all providers must work to control drug costs, including antiepileptic drugs. Generic drugs offer a less expensive choice for antiepileptic drugs. Given the variable serum concentrations of antiepileptic drugs when coadministered with other antiepileptic drugs possessing different pharmacokinetic properties of antiepileptic drugs, as demonstrated in Tables 47.2 through 47.7, the physician must be careful to monitor both seizure control and antiepileptic drug levels in a given patient, especially when switching to a generic antiepileptic drug. The lack of bioequivalence data from comparisons among different generic forms of antiepileptic drugs has the potential to question the relatively broad criteria for bioequivalence with the branded drug. Differences in drug exposure, not only between brand name and generic substitutes but also among the speciἀc generic preparations of the same antiepileptic drugs, are clinically relevant. Plasma level monitoring is required not only when switching formulations, whether from brand name to generic, but also from one generic preparation to another to avoid loss of seizure control or emergence of side effects (Kramer et al. 2007). A number of years ago, one of the authors (PLS) pointed out to a state Medicaid office that if they required generic substitution of antiepileptic drugs, then they should at least allow the physicians to specify that the same generic antiepileptic drug brand be dispensed to the patient for each prescription renewal. Their circular argument for not allowing this simple therapeutically important requirement was that if a speciἀc brand were prescribed, it would then be considered to be a “brand name drug” and therefore would be in conflict with state regulations requiring generic substitution! Monitoring drug levels during a switch from one brand to another does increase the health care cost. Both the physician and the patient must be informed of and approve of a generic substitution or a switch between generics. Jobst and Holmes (2004) asked the question of whether the prescribing physician should consider switching patients based on cost. To answer this question, one must evaluate not just the drug acquisition cost but also both direct and indirect costs, such as obtaining blood levels every time the patient has been prescribed a different generic preparation from what was used to ἀll the prior prescription. When one evaluates only efficacy and not adverse events, the older antiepileptic drugs that are no longer under patent protection are certainly less costly than newer antiepileptic drugs. If the epilepsy is poorly controlled, the additional costs, such as frequent

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

773

visits to the physician, multiple antiepileptic drug blood levels, emergency room visits, and hospitalizations, may be enormous. Thus, the antiepileptic drugs that provide total seizure control will become the most cost-effective choice. When switching to generic formulations, the drug acquisition cost will decrease but other problems may occur, such as resultant bioavailability difference, possible loss of seizure control, and onset of toxicityrelated symptoms. Switching the patient to generic phenytoin and carbamazepine may be especially problematic (Jobst and Holmes 2004). The bottom line is that seizure control should never be sacriἀced, and thus selection of an antiepileptic drug must be based on the individual clinical situation over the long-term care period and not on short-term cost savings (Heaney and Sander 2007). Case Hist o ry : Generic Carbamaz epine That reduced serum concentrations and seizure exacerbation followed generic substitution of Tegretol is a well-established phenomenon. Gilman et al. (1993) found that carbamazepine toxicity resulting from generic substitution occurred in two 6-yearold children. Increases in the maximum serum carbamazepine concentrations, one of 22% and one of 41%, occurred. When serum concentrations were decreased in both patients, the two patients became asymptotic and residual effects did not occur. Quest io ns Raised by L at hers and Schraeder 1. Could increased carbamazepine levels be a risk factor for SUDEP? Possibly yes. As cited below, carbamazepine has been shown on Holter monitoring to produce intermittent complete atrioventricular block. At higher blood levels, carbamazepine can interfere with cardiac electrical conduction and increase the risk of conduction block. Devinsky et al. (1994) found that variations occurred in blood pressure and heart rate during orthostasis and cold face test. Variations in these parameters were higher in patients with epilepsy than in controls with or without carbamazepine. Persons with epilepsy exhibited higher initial increase in blood pressure and greater subsequent decreases in blood pressure than did nonmedicated controls during the cold fact test. Controls treated with carbamazepine demonstrated higher heart rate during orthostasis and cold face test with apnea than did those not on the drug. The carbamazepine levels correlated with baseline and orthostatic blood pressure and heart rate during deep breathing (sinus arrhythmia). The authors conclude that persons with epilepsy have greater blood pressure and heart rate variability and reactivity than controls, attributable in part to carbamazepine levels. 2. If carbamazepine levels had decreased as opposed to increasing, would decreased levels be a risk factor for SUDEP? If carbamazepine levels had decreased, there is a possibility of increased uncontrolled interictal discharges and/or seizures developing and these events may increase the chance of SUDEP via impaired autonomic cardiac sympathetic discharges/nonuniformity (Lathers and Schraeder 1982; Schraeder and Lathers 1983) (see Chapter 28 of this book for a full discussion). Aside from the generic question, there is also a formulation difference. Ficker et al. (2005) found that switching patients to an extended-release formulation of carbamazepine from an immediate-release formulation diminished the risk of adverse events and improved quality of life measures.

774 Sudden Death in Epilepsy: Forensic and Clinical Issues

Furthermore, switching to the extended-release formulation of carbamazepine was associated with improved seizure control. 3. Could carbamazepine-induced hyponatremia contribute to occurrence of SUDEP in some patients by lowering the threshold for arrhythmia? Hyponatrema lowers seizure threshold and increases frequency of seizures. Summarized from Gilman, J. T., et al., Neurol, 43, 2696–2697, 1993. Heaney and Sander (2007) emphasized that although savings will occur when switching from brand name to less expensive generic drugs, clinicians should be cautious because of the narrow therapeutic range of most antiepileptic drugs and the variable bioavailability of various preparations of the same drug. They should be especially cautious in switching among preparations of phenytoin because it is a drug that is metabolized via zero kinetics. Likewise, the clinical principles of prescribing include making only cautious and gradual changes in dosing. In addition, the health and socioeconomic impact of breakthrough seizures or toxicity and the need for long-term consistency of supply must be considered. It has not been established that the switch to the cheapest generic antiepileptic drugs offers drug budget savings that outweigh the potential risk to patient safety. Newer antiepileptic drugs, used as monotherapy, may be cost-effective in those patients exhibiting adverse events with older antiepileptic drugs, in those who failed to respond to the older drugs, or in those for whom such drugs are contraindicated. Wilby et al. (2005) made the following suggestions to address the clinical effectiveness, tolerability, and costeffectiveness of newer antiepileptic drugs: 1. More direct comparisons of the different antiepileptic drugs, in clinical trials, are needed to evaluate different treatment sequences with both monotherapy and adjunctive therapy. 2. The length of follow-up needs to be carefully determined. 3. Trials to recruit patients with either partial or generalized seizure must be designed. 4. Investigate both therapeutic effectiveness and cost-effectiveness in patients with generalized onset seizures. 5. Examine effectiveness in speciἀc population of patients with epilepsy. 6. Evaluate outcomes to use more stringent testing protocols. 7. Adopt a more consistent approach in assessing outcomes. 8. Assess quality of life within trials of epilepsy therapy using a preference base measure of outcomes that generate cost-effectiveness data. 9. Provide information on use of antiepileptic drugs in actual practice, including details of treatment sequences and doses. Compulsory generic switching of antiepileptic drugs was examined by Andermann et al. (2007) using a pharmacy claims database in Canada. Switchback rates from generic to branded antiepileptic drugs Lamictal (lamotrigine), Frisium (clobazam), and Depakene (valproic acid; divalproex) were determined for 1354 patients (403 monotherapy, 951 polytherapy). Of those prescribed generic lamotrigine, 12.9% switched back to Lamictal (11.7% monotherapy, 13.4% polytherapy). Switchback rates were ~20% for clobazam and

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

775

valproic acid. Switchback rates for antiepileptic drugs were substantially higher than for the non-antiepileptic drug used for long-term therapies of antihyperlipidemia and antidepressants. Signiἀcant increases in lamotrigine doses occurred after generic substitution if a switchback did not occur. The average number of codispensed antiepileptic drugs and non-Â�antiepileptic drugs signiἀcantly increase after generic lamotrigine. In general, it was concluded that there was poor acceptance of switching antiepileptic drugs to generic compounds. Data may indicate increased toxicity and/or loss of seizure control associated with generic antiepileptic drug products.

47.10â•… Possible Role of Antiepileptic Drugs in Causes of SUDEP It has been hypothesized (Lathers and Schraeder 1982) that neural autonomic dysfunction–initiated cardiac arrhythmia associated with epileptogenic activity may contribute to sudden unexpected death in epileptic persons (SUDEP) and that pharmacological agents exhibiting anticonvulsant, antiarrhythmic, and cardiac neural depressant properties might diminish the risk of SUDEP. However, some anticonvulsant agents such as phenytoin or carbamazepine may also be associated with an increased risk of cardiac arrhythmia. Summarized in the nine tables in chapter 10 of this book are pharmacological risk factors for SUDEP including drugs that prolong the QTc interval, class III antiarrhythmics, antimicrobial agents, antipsychotic drugs, antidepressant drugs, and antiepileptic drugs that may contribute to sudden death.

47.11â•…Do Antiepileptic Drugs Induce ECG Changes and SUDEP? In general, antiepileptic drugs do not appear to be associated with a statistically signiἀcant increased risk of SUDEP. Low drug levels due to variations in bioavailability, drug metabolism (see Tables 47.2 through 47.7), and/or compliance with the associated increased seizure occurrence may contribute to the risk of SUDEP. In some cases, higher drug levels may be a factor, possibly because toxic antiepileptic drug levels can be paradoxically associated with increased risk of seizure occurrence. However, there is a need to investigate further the role of carbamazepine use with increased risk of SUDEP when compared with newer antiepileptic drugs, that is, lamotrigine, gabapentin, topiramate, and tiagabine. For example, higher levels of carbamazepine may be associated with electrolyte imbalance, cardiac arrhythmias, AV block, aggravation of coronary artery disease, syncope, and congestive heart failure. Carbamazepine is an iminostilbene and a structural congener of the tricyclic antidepressant drug imipramine. Use of tricyclic antidepressants is associated with sudden death (Leestma et al. 1969). Although there are some circumstantial observations, there is not credible evidence that carbamazepine use is associated with an increased risk of SUDEP (Timmings 1998). Case Hist o ry : Lamo t rig ine-Induced QRS Pro lo ng at io n QRS prolongation may be a possible contributory risk factor for sudden death (Lathers et al. 2008c). Lamotrigine toxicity secondary to acute intentional or unintentional

776 Sudden Death in Epilepsy: Forensic and Clinical Issues

overdose may be, as Herold (2006) notes, a possible cause of QRS prolongation in a patient with a seizure disorder. This author described a 22-year-old woman who had two seizure-like episodes consisting of tonic–clonic activity of the upper extremities with no sphincter incontinence who presented to the emergency room. On examination, she was found to have horizontal nystagmus on lateral gaze and ataxia, with widening of the QRS complex and right-axis deviation on the ECG. There was no history of her taking more than her prescribed dose of lamotrigine (200 mg t.i.d.). A lamotrigine level of 57.8 µmol/L was detected (therapeutic range is 3.9–15.6 µmol/L). Because lamotrigine blocks sodium channels, sodium bicarbonate administered intravenously resulted in improvement of the QRS narrowing. The patient was also taking felbamate 600 mg t.i.d. and had no laboratory evidence of any other medication that could have produced QRS prolongation. The author suggested that a drug interaction may have been responsible for the toxic lamotrigine level. She left the emergency room against medical advice. Co mment s by L at hers and Schraeder If a drug inhibits the human cardiac delayed rectiἀer potassium current, prolongation of the cardiac QT interval may occur. This change is associated with a potentially fatal, polymorphic, ventricular tachycardia called torsades de pointes. Lamotrigine inhibits the human cardiac delayed rectiἀer potassium current in vitro, and it has been hypothesized that QT prolongation may contribute to the risk of SUDEP. Dixon et al. (2008) studied QT/QTc and found the QTc interval was not prolonged by therapeutic doses of lamotrigine (50–200 mg b.d.) in healthy subjects. The patient described by Herold (2006) had been taking a therapeutic dose of lamotrigine (200 mg t.i.d.) and was found to have a toxic level, which in all likelihood was the cause of the ECG changes. Summarized from Herold, T. J., CJEM, 8, 361–364, 2006.

Case Hist o ry : Carbamaz epine Int ermit t ent C o mplet e AV Blo ck A 66-year-old woman with epilepsy and no history of heart disease went to the hospital complaining of frequent episodes of sudden dizziness (Ide and Kamijo 2007). She had taken a low dose of carbamazepine, 200-mg daily, for 1 year. The carbamazepine level on admission was subtherapeutic at 4 µg/mL (therapeutic range is 6–12 µg/mL). Holter monitoring showed intermittent complete atrioventricular block of up to 10 s duration with ventricular escape. The patient reverted to a normal sinus rhythm after the carbamazepine was discontinued with substitution of valproic acid 600 mg daily. The authors note that this case demonstrates that although complete atrioventricular block may occur in association with high carbamazepine level, it may also occur long after initiation of carbamazepine therapy in an “older” woman even if the serum concentration of carbamazepine is subtherapeutic.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

777

Co mment s by L at hers and Schraeder This case report raises the possibility that some persons without any history of heart disease may have a genetic predisposition to atrioventricular block when exposed to carbamazepine and emphasizes the potential role for an as yet undeveloped genetic methodology to screen for patients who might be at risk for antiepileptic drug-related cardiac conduction disturbances or arrhythmias. Summarized from Ide, A. and Kamijo, Y., Intern Med, 46, 627–629, 2007. Both carbamazepine and oxycarbamazepine used as monotherapy were associated with reduced serum thyroid hormone concentrations in 78 girls diagnosed with epilepsy (Vainionpaa et al. 2004). An earlier study found similar changes for carbamazepine. Phenytoin was also shown to decrease serum thyroid hormone levels and, similar to carbamazepine, was theorized to accelerate hepatic plasma clearance of these hormones due to induction of hepatic microsomal enzyme systems by these antiepileptic drugs (Isojarvi et al. 1992). Nonetheless, because the subjects were clinically euthyroid, the implication of the decreased thyroid hormone levels is unknown. Increased QT interval dispersion (QTd) is an ECG parameter associated with malignant ventricular arrhythmias and sudden death (Backiner et al. 2008). QT dispersion corrected for heart rate is thought to be one predictor of cardiac death. Thyroid-stimulating hormone level in overt hypothyroidism is associated with increased QTd. Backiner et al. (2008) investigated QTc in subclinical hypothyroidism and monitored changes when thyroid-stimulating hormone levels were normalized with l -thyroxine. The prolonged QTc found in the females was corrected when thyroidÂ�stimulating hormone levels reach 10 mIU/L or more. Thus, although there is, as yet, no known clinical implications for the female patients with epilepsy taking carbamazepine, oxycarbamazepine, or phenytoin, it is likely that they should be observed for changes in QTd/QTc because prolonged QTc may place these patients at a greater risk for SUDEP. The Sy mpt o mat ic Seco ndary H y po nat remia Due t o C o mbined Treat ment wit h Carbamaz epine, Lamo t rig ine, and Venlafax ine: Risk o f SUDEP? A 37-year-old woman presented with an episode of syncope secondary to symptomatic hyponatremia. She had refractory epilepsy associated with exogenous depressive syndrome and was receiving combined treatment with carbamazepine, lamotrigine, and venlafaxine, an antidepressant. Co mment s by R uiz G ines et al. The hyponatremia was generated by inappropriate secretion of antidiuretic hormone. The electrolytic anomaly can result in secondary neurological and cardiovascular effects and may contribute to sudden death secondary to the risk of cardiac arrhythmia in persons with epilepsy. The authors recommend “strict ionic control” in patients requiring combined antiepileptic drugs and antidepressant treatment to avoid “paroxysmal vascular episodes” and to minimize the risk of SUDEP.

778 Sudden Death in Epilepsy: Forensic and Clinical Issues

Co mment s by L at hers and Schraeder Readers are referred to a paper by Kloster et al. (1998) that reported sudden death in two patients with epilepsy and the syndrome of inappropriate antidiuretic hormone. The two patients, one with complex partial and the other with secondarily generalized seizures, were both taking oxcarbazepine and vigabatrin. One of the two patients was also taking lamotrigine. The authors note that both oxcarbazepine and carbamazepine may cause inappropriate secretion of antidiuretic hormone and recommend review of SUDEP cases and the antiepileptic drugs prescribed to determine if an inappropriate secretion of antidiuretic hormone may have been involved in the mechanism of death. From Ruiz Gines, M. A., et al., An Med Interna, 24, 335–338, 2007. possible role of QTc-prolonging drugs is the subject of ongoing investigational interest (Reingardiene and Vicissitude 2007). Cardiac changes may be associated with increased risk of mortality. Lund and Gormsen (1985) reported patients with SUDEP treated with one or more of the antiepileptic drugs phenobarbital, phenytoin, and carbamazepine had subtherapeutic drug levels in half of the cases and potentially lethal high concentrations in one-third. The ἀndings were especially true for phenobarbital. They discussed the possibility that variations in bioavailability and in drug metabolism by enzyme induction may contribute to or may be a risk factor for SUDEP. The importance of consistent therapeutic blood levels and drug compliance in patients with epilepsy was emphasized, as was that of antiepileptic drug levels as a measure of drug compliance. The observation of Kenneback et al. (1997) that abrupt withdrawal of carbamazepine and phenytoin resulted in signiἀcant reduction in heart rate variability and an increase in ventricular automaticity suggest that maintaining stable antiepileptic drug levels and avoiding precipitous decreases may have importance in maintaining stable cardiac function. Physicians must be aware of the need to prevent fluctuations of concomitant antiepileptic drug levels. Schachter (1999) noted that the goal of treating persons with epilepsy has evolved from attaining complete control of seizures regardless of side effects to enabling the person to lead a lifestyle consistent with his capabilities and emphasized that the goal for pharmacologic therapy is to suppress seizures without side effects (Schachter 2002). One must always strive to achieve the best seizure control with the anticipation that this will minimize the risk of SUDEP. Jay and Leestma (1981) and Graves et al. (1988) reported low postmortem blood concentrations of antiepileptic drugs in cases of SUDEP, suggesting that, although noncompliance may play a role in SUDEP, postmortem phenytoin concentrations should be interpreted with caution because animal data indicate that antemortem antiepileptic drug whole blood levels were only 65% of the corresponding serum concentration, whereas postmortem whole blood levels were even lower at 35% of antemortem serum concentration. These ἀndings indicate that low postmortem whole blood concentrations do not necessarily imply poor compliance with treatment. Thus, it is incumbent upon medical examiners and coroners to be speciἀc about the source of postmortem antiepileptic drug level, that is, blood vs. serum. However, by implication, the absence of any antiepileptic drug levels postmortem would imply noncompliance. May et al. (1996) reported a postmortem decrease in anticonvulsant serum concentrations, especially for phenobarbital and phenytoin, whereas Opeskin et al. (1999) concluded that insufficient

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

779

data exist to support the concept that patients with SUDEP were less compliant with antiepileptic drug treatment than the control group. Tennis et al. (1995) reported SUDEP incidence was highest in males with a history of treatment with three or more antiepileptic drugs and four or more psychotropic drug prescriptions. In contrast, Timmings (1993) found no signiἀcant differences in the number of drugs, seizure frequency, duration of epilepsy, nor age between patients with SUDEP and a control group but suggested that carbamazepine treatment, either as monotherapy or combination therapy, may be associated with increased risk of SUDEP. Structurally, carbamazepine is a tricyclic compound, and although use of tricyclic antidepressants has also been reported to be associated with some increased risk of cardiac conduction and rhythm dysfunction, there is no link with an increased risk of sudden death related to use of carbamazepine nor tricyclic antidepressants (Lathers et al. 1986; Lathers and Lipka 1987; Lipka and Lathers 1987; Lathers and Leestma 1990). In addition, phenytoin has been shown to be associated with increased chance of disturbance in cardiac conduction and rhythm in some cases, but as with carbamazepine, there is no established link between these antiepileptic drugs and risk of SUDEP. Patient management requires closely monitoring the free as well as total serum antiepileptic drug levels to avoid any potentially adverse cardiac effects of supratherapeutic free levels. Clinical management of persons with epilepsy should emphasize use of as few antiepileptic drugs as possible, patient compliance with the drug(s) prescribed, monitoring of total and free antiepileptic drug levels, avoidance of abrupt withdrawal of antiepileptic drugs, and awareness of the risk of simultaneous use of psychotropic drugs with antiepileptic drugs. George and Davis (1998) concluded that the incidence of subtherapeutic antiepileptic drugs is signiἀcantly greater in patients dying as a direct result of epilepsy than in those dying of an unrelated cause. Although low postmortem antiepileptic drug concentrations have been reported in SUDEP and attributed to poor compliance, Tomson et al. (1998) suggest a different conclusion. Steady-state carbamazepine and phenytoin concentrations were achieved in rabbits before death and 72 h postmortem. Both serum and whole blood carbamazepine concentrations were comparable and stable up to 72 h postmortem. In contrast, phenytoin whole blood and serum levels 72 h postmortem were only 65% and 35%, respectively, of values obtained just before death. Thus, patients found postmortem with low levels of phenytoin may not have been poorly compliant. However, if none of the prescribed medication is found in postmortem blood samples, it is reasonable to say that the patient was noncompliant. If the blood and serum levels of phenytoin postmortem were at least 65% and 35% of premortem levels, respectively, then it would be reasonable to assume that the patient probably was compliant in taking the antiepileptic drugs. Almost all available antiepileptic drugs have been prescribed for victims before the occurrence of SUDEP, and current opinion assumes that the relative proportion of patients exhibiting SUDEP is the same for all antiepileptic drugs. However, because there is still the theoretical possibility of increased risk of SUDEP in association with the use of some antiepileptic drugs more than others, more research needs to be done to establish whether any selective difference in antiepileptic drug–related risk of SUDEP exists. Although multiple concurrent antiepileptic drugs prescribed for a given patient to control seizures may be associated with an increased risk for SUDEP, the use of multiple antiepileptic drugs is commonly associated with a history of seizures that is difficult to control. In addition, as mentioned earlier, other possible risk factors for SUDEP are thought to

780 Sudden Death in Epilepsy: Forensic and Clinical Issues

include poorly controlled seizures, early onset of epilepsy, and generalized tonic–clonic seizures (Ficker 2000). Nilsson et al. (1999) reported the relative risk of SUDEP increased with the number of seizures per year. The relative risk of SUDEP was up to 10-fold greater in patients with 50 seizures per year vs. those with no more than two per year. Risk increased with increased number of antiepileptic drugs taken concomitantly and was as high as 9.89 greater with three concomitant antiepileptic drugs vs. monotherapy. Early-onset vs. lateonset epilepsy and frequent changes of antiepileptic drug dosage when compared with patients on an unchanged dosage were also risk factors. Optimizing seizure control, while minimizing polypharmacy, may reduce the risk of SUDEP. An FDA-sponsored expert panel evaluated the prevalence of SUDEP in patients involved in studies associated with developing new antiepileptic drugs and reviewed data on the risk of SUDEP in patients taking lamotrigine. The risk of SUDEP was found to be no different from that in the young epilepsy population in general. Estimated SUDEP rates in patients receiving the new antiepileptic drugs lamotrigine, gabapentin, topiramate, tiagabine, and zonisamide were found to be similar to those in patients receiving standard antiepileptic drugs, suggesting SUDEP rates reflect population rates and not a speciἀc drug effect (Leestma et al. 1997). The FDA requires package insert statements on the risk of SUDEP in association with the use of each of the above-mentioned drugs (Lathers and Schraeder 2002). Garaizar (2000) stated that some physiologic data in humans during seizures implied that central apnea may occur, occasionally followed by asystole, whereas in other patients, primary cardiac arrhythmias of reflex neural origin have been detected. Pathologic studies at autopsy support the second hypothesis, suggesting cardiac ischemia due to repeated seizure-related neurogenic and/or catecholamine-induced vasospasm. Garaizar (2000) recomÂ�mends that this type of cardiomyopathy should be included in the differential diagnosis of the cardiomyopathies and stated that SUDEP may be causally associated with a sudden decrease in antiepileptic drug serum levels that may cause potential fatal cardiac arrhythmias. Lhatoo et al. (1999) also concluded that seizure control is probably very important in prevention of death in persons with epilepsy. The ἀnding of possible neurogenically mediated cardiac pathology indicates a need for more extensive investigation of this question. Prevention of SUDEP, to date, centers on effective seizure control with use of antiepileptic drugs and/or epilepsy surgery. The latter has been demonstrated to reduce SUDEP incidence in a number of studies. Additional research is needed to clarify underlying mechanisms of SUDEP to help us to understand why patients with refractory epilepsy face an elevated risk of sudden death, with rates as high as 1% per year. Jehi and Najm (2008) cite both pulmonary and cardiac pathophysiology as underlying mechanisms of SUDEP. The cardiac mechanism of interest focused on precipitation of arrhythmias by seizure discharge via the autonomic nervous system. Patients with frontal lobe epilepsy have faster interictal heart rates that are attributed to lower parasympathetic drive, contributing to the higher incidence of sudden death observed in this group of patients. Harnod et al. (2008) suggest the mechanism of decreased heart rate variability in patients with frontal lobe epilepsy may be different from that in patients with temporal lobe epilepsy. Toth et al. (2008) discuss sudden death of patients with epilepsy not responding to antiepileptic drug administration. To further explore the possibility that a common cause of death is seizure-related cardiac arrhythmias, they analyzed alteration of heart rate 6 h before and after seizures in preparation for epilepsy surgery. Video-EEG-ECG was done for 2–10 days,

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

781

during which interval 32 seizures occurred. Heart rate increased immediately after seizure and was signiἀcantly higher 3 h after seizures. No person exhibited severe peri-ictal bradycardias. In one patient, ectopic cardiac rhythm was demonstrated after generalized tonic–clonic seizure. The authors conclude that sympathetic activity increased while parasympathetic activity decreased after seizures. These changes lasted for a long time and predicted fatal arrhythmia. The data support the possibility that SUDEP can be induced by cardiac arrhythmias connected with epileptic seizures. For example, the uncommon, heritable disease of catecholaminergic polymorphic ventricular tachycardia may present with syncope or sudden cardiac death. Calcium homeostasis is regulated by two genes, the ryanodine receptor gene and the calsequestrin 2 (CASQ2) gene. Both of these genes have been implicated with catecholaminergic polymorphic ventricular tachycardia. De la Fuente et al. (2008) reported a young man presenting with exercise-induced syncope who was clinically diagnosed with catecholaminergic polymorphic ventricular tachycardia. Genetic analysis revealed two mutations, p.Y55C (c.164A>G) and p.P308L (c.923C>T), for the CASQ2 gene. Subsequent familial analysis indicated a compound heterozygous form of inheritance (De la Fuente et al. 2008). Lehnart et al. (2008) provided data demonstrating leaky Ca2+ release channel/ryanodine receptor 2 (RyR2) can trigger seizures and sudden cardiac death in mice. This receptor is required for excitation–contraction coupling in the heart. The receptor is also located in the brain. Mutations of this receptor are linked to exercise-induced sudden cardiac death catecholaminergic polymorphic ventricle tachycardia. Catecholaminergic polymorphic ventricular tachycardia–associated mutations in the receptor may decrease binding to a subunit [calstabin (FKBP12.6)] responsible for stabilizing the closed state of the Ca+ channel. Mice heterozygous for the R2474S mutation in Ryr2 mice exhibited spontaneous generalized tonic–clonic seizures in the absence of arrhythmia, exercise-induced ventricular arrhythmias, and sudden cardiac death. When treated with a compound to enhance the binding of calstabin2 to the mutant RyR2-speciἀc channel, they inhibited the channel leak, prevented cardiac arrhythmias, and increased the seizure threshold. It was concluded that catecholaminergic polymorphic ventricular tachycardia–associated mutant leaky Ryr2-R2474S channels in the brain can cause seizures in mice, independent of cardiac arrhythmias. Lehnart et al. (2008) propose that catecholaminergic polymorphic ventricular tachycardia is a combined neurocardiac disorder in which leaky RyR2 channels in the brain cause epilepsy and the same leaky channels in the heart cause exercise-induced sudden cardiac death. Finally, omega-3 polyunsaturated fatty acid (omega-3 fatty acids) dietary supplements appear to reduce sudden death after myocardial infarction (Dujardin et al. 2008). In a Langendorff-perfused rabbit heart preparation, the proarrhythmic effects of dofetilideinduced ventricular arrhythmias were examined. Hearts were obtained from animals pretreated with omega-3 fatty acids and compared with data obtained in control rabbits. Torsades de pointes occurred in ἀve of six control hearts and none were observed in the pretreated hearts. Dietary omega-3 fatty acids markedly reduced dofetilide-induced instability and dispersion of the cardiac action potential. Ultrafast sodium channel block by docosahexaenoic acid may explain the antiarrhythmic protection of dietary supplements of omega-3 fatty acids. Scorza et al (2008a) emphasized the value of omega-3 fatty acids to protect the heart. Lathers et al. (2008b) also discussed the possible beneἀt of dietary supplemental omega-3 fatty acids to reduce the incidence of sudden death in persons with epilepsy. It is well known that the beneἀcial effect of nutritional aspects of omega-3-fatty acids may decrease cardiac arrhythmias and sudden death in patients with cardiac disease.

782 Sudden Death in Epilepsy: Forensic and Clinical Issues

Does a nutritional deἀciency in omega-3 fatty acids become a risk factor for SUDEP? There is no answer for the question at this time. Whether omega-3 fatty acids deἀciency is a risk factor for the relatively young population prone to SUDEP is unknown and studies designed to address this question have yet to be done. The beneἀcial cardiovascular effect of omega-3 fatty acids in decreasing sudden death in cardiac patients may also be due, in part, to actions on cholesterol levels. Will this cardioprotective effect, independent of the effect on cholesterol levels, be of beneἀt for patients with epilepsy? Although at this time there is certainly nothing negative to be said for including omega-3 fatty acid–containing ἀsh in one’s diet to gain the nutritional beneἀts, Scorza et al. (2008b) emphasized that nutritional therapy, including omega-3 fatty acid supplementation, should never be a substitute for anticonvulsant medications. Collectively, all of the human and animal studies discussed above emphasize the importance of the clinical pharmacology question: “Are antiepileptic drugs a beneἀt and/ or a risk in sudden unexpected death in epilepsy?” (Lathers and Schraeder 2002). Clearly, each patient must be continually evaluated for his/her response to a given antiepileptic drug and/or combination of antiepileptic drugs. Personalized therapeutic interventions in persons with epilepsy are important to prevent unwanted antiepileptic drug side effects while optimally addressing the need to control the unwanted symptoms of epilepsy to minimize the risk of SUDEP.

References Aldenkamp, A. P. 2001. Cognitive and behavioral assessment in clinical trials: When should they be done? Epilepsy Res 45: 155–157. Aldenkamp, A. P. 1997. Effect of seizures and epileptiform discharges on cognitive function. Epilepsia 38 (S1): S52–S55. Aldenkamp, A., and J. Arends. 2004. The relative influence of epileptic EEG discharges, short nonconvulsive seizures, and type of epilepsy on cognitive function. Epilepsia 45: 54–63. Aldenkamp, A. P., J. Beitler, J. Arends, I. van der Linden, and L. Diepman. 2005. Acute effects of subclinical epileptiform EEG discharges on cognitive activation. Funct Neurol 20: 23–28. Aldenkamp, A. P., M. DeKrom, and R. Reijs. 2003. Newer antiepileptic drugs and cognitive issues. Epilepsia 44 (S4): 21–29. Aldenkamp, A. P., and N. Bodde. 2005. Behavior, cognition and epilepsy. Acta Neurol Scand Suppl 182: 19–25. Andermann, F., M. S. Duh, A. Goselin, and P. E. Paradis. 2007. Compulsory generic switching of antiepileptic drugs: High switchback rates to branded compounds compared with other drug classes. Epilepsia 48: 464–469. Austin, J. K., and D. W. Dunn. 2002. Progressive behavioral changes in children with epilepsy. Prog Brain Res 135: 419–427. Bachman, K. A., and R, J. Belloto. 1999. Differential kinetics of phenytoin in elderly patients. Drugs Aging 15: 235–250. Backiner, O., M. E. Ertorer, F. E. Haydardedeogiu, E. Bozkirli, N. B. Tutuncu, and N. G. Demirag. 2008. Subclinical hypothyroidism is characterized by increase QT interval dispersion among women. Med Princ Pract 17: 390–394. Besag, F. M. 1995. The therapeutic dilemma: Treating subtle seizures or indulging in electroencephalogram cosmetics? Semin Pediatr Neurol 2: 261–268. Beydoun, A. B., M. Uthman, A. R. Kugler, M. J. Greiner, L. E. Knapp, and E. A. Garofalo. 2005. Pregabalin 1008-009 Study Group. Safety and efficacy of two pregabalin regiments for add-on treatment or partial epilepsy. Neurology 64: 475–480.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

783

Binnie, C. D. 2003. Cognitive impairment during epileptiform discharges: Is it ever justiἀable to treat the EEG? Lancet Neurol 2: 725–730. Binnie, C. D., and D. Marston. 1992. Cognitive correlates of interictal discharges. Epilepsia 33 (S6): S11–S17. Brown, S. 2006. Deterioration. Epilepsia 47 (S2): 19–23. Brunbech, L., and A. Sabers. 2002. Effect of antiepileptic drugs on cognitive function in individuals with epilepsy: A comparative review of newer versus older agents. Drugs 62: 593–604. De la Fuente, S., I. M. Van Langen, A. V. Postma, H. Bikker, and A. Meijer. 2008. A case of catecholaminergic polymorphic ventricular tachycardia caused by two calsequestrin 2 mutations. Pacing Clin Electgrophysiol 31: 916–919. Deng, P. Y., J. E. Porter, H. S. Shin, and S. Lei. 2006. Thyrotropin-releasing hormone increases GABA release in rat hippocampus. J Physiol 577 (Pt2): 497–511. Devinsky, O., K. Perrine, and W. H. Theodore. 1994. Interictal autonomic nervous system function in patients with epilepsy. Epilepsia 35: 199–204. Dixon, R., S. Job, R. Oliver et al. 2008. Lamotrigine does not prolong QTc in a thorough QT/QTc study in healthy subjects. Br J Clin Pharmacol 66: 396–404. Dodd-O, J., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in phenobarbitalÂ�pretreated cats. In Epilepsy and Sudden Death, Chapter 13, 199–220. ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Dong, X., I. E. Leppik, J. White, and J. Rarick. 2005. Hyponatremia from oxcarbazepine and carbaÂ� mazepine. Neurology 65: 1976–1978. Dooley, M., and G. L. Plosker. 2000. Levetiracetam. A review of its adjunctive use in the management of partial onset seizures. Drug 60: 871–893. Drane, D. L., and K. J. Meador. 1996. Epilepsy, anticonvulsant drugs and cognition. Baillieres Clin Neurol 5: 877–885. Druschky, A., M. J. Hilz, P. Hopp et al. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Dujardin, K. S., B. Dumotier, M. David, M. Guizy, C. Valenzuela, and L. M. Hondeghem. 2008. Ultra fast sodium channel block by dietary ἀsh oil prevents dofetilide-induced ventricular arrhythmias in rabbit hearts. Am J Physiol Heart Circ Physiol 295 (4): H1414–H1421. iEpub 2008 Aug 1. Duncan, S. 2007. Teratogenesis of sodium valproate. Curr Opin Neurol 20: 175–180. Engel, J., R. Bandler, N. C. Griffith, and S. Caldecott-Hazard. 1991. Neurobiological evidence for epilepsy-induced interictal disturbances. Adv Neurol 55: 97–111. Ficker, D. M. 2000. Sudden unexplained death and injury in epilepsy. Epilepsia 41 (Suppl2): S7–S12. Ficker, D. M., M. Privitera, G. Krauss, A. Kanner, J. L. Moore, and T. Glauser. 2005. Improved tolerability and efficacy in epilepsy patients with extended-release carbamazepine. Neurology 65: 593–595. Fukuoka, N. 2004. Influence of coadministered antiepileptic drugs on serum antiepileptic drug concentrations in epileptic patients—Quantitative analysis based on suitable transforming factor. Yakugaku Zasshi 124: 443–450. Fukuoka, N., T. Tsukamoto, J. Uno, M. Kimura, and S. Morita. 2003a. Effects of concomitant antiepileptic drugs on serum carbamazepine concentration in epileptic patients: Quantitative analysis based on extracellular water volume as a transforming factor. Yakugaku Zasshi 123: 35–42. Fukuoka, N., T. Tsukamoto, J. Uno, M. Kimura, and S. Morita. 2003b. Influence of coadministered antiepileptic drugs on serum zonisamide concentrations in epileptic patients: Quantitative analysis based on suitable transforming factor. Biol Pharm Bull 26: 1734–1738. Garaizar, C. 2000. Sudden unexpected and unexplained death in epilepsy. Rev Neurol 31: 436–441. George, J. R., and G. G. Davis. 1998. Comparison of antiepileptic drug levels in different cases of sudden death. J Forensic Sci 43: 595–603. Gilman, J. T., L. A. Alvarez, and M. Duchowny. 1993. Carbamazepine toxicity resulting from generic substitution. Neurology 43: 2696–2697.

784 Sudden Death in Epilepsy: Forensic and Clinical Issues Gissel, T., C. S. Poulsen, and P. Vestergaard. 2007. Adverse effects of antiepileptic drugs on bone mineral density in children. Expert Opin Drug Saf 6: 267–278. Glauser, T., E. Ben-Menachem, B. Bourgeois et al. 2006. International League against Epilepsy (ILAE). Treatment guidelines: Evidence-based analysis of antiepileptic drug efficacy and effectiveness as initial monotherapy for epileptic seizures and syndromes. Epilepsia 47: 1094–1112. Gomer, B., K. Wagner, L. Frings et al. 2007. The influence of antiepileptic drugs on cognition: A comparison of levetiracetam with topiramate. Epilepsy Behav 10: 486–494. Graves, N. M., G. B. Holmes, and I. E. Leppik. 1988. Compliant populations: Variability in serum concentrations. Epilepsy Res Suppl 1: 91–99. Hamandi, K., and J. W. Sander. 2006. Pregabalin: A new antiepileptic drug for refractory epilepsy. Seizure 15: 73–78. Harden, C. L., and I. Leppik. 2006. Optimizing therapy of seizures in women who use oral contraceptives. Neurology 67 (12 S 4): S56–S58. Harnod, T., C. C. Yang, Y. L. Hsin, P. J. Wang, K. R. Shieh and T. B. Kuo. 2009. Heart rate variability in patients with frontal lobe epilepsy. Seizure 18: 21–25. Hilz, M. J., G. Platsch, K. Druschky et al. 2003. Outcome of epilepsy surgery correlates with sympathetic modulation and neuroimaging of the heart. J Neurol Sci 15;21 (6): 153–162. Heaney, D. C., and J. W. Sander. 2007. Antiepileptic drugs: Generic versus branded treatments. Lancet Neurol 6: 465–468. Herold, T. J. 2006. Lamotrigine as a possible cause of QRS prolongation in a patient with known seizure disorder. CJEM 8: 361–364. Hoch, D. B., R. A. Hill, and K. H. Oas. 1994. Epilepsy and mental decline. Neurol Clin 12: 101–113. Holmes, G. L., and P. P. Lenck-Santini. 2006. Role of interictal epileptiform abnormalities in cognitive impairment. Epilepsy Behav 8: 504–514. Ide, A., and Y. Kamijo. 2007. Intermittent complete atrioventricular block after long term low dose carbamazepine therapy with a serum concentration less than the therapeutic level. Intern Med 46: 627–629. Isojarvi, J. I., A. J. Parkainen, and V. V. Myllyla. 1992. Thyroid function with antiepileptic drugs. Epilepsia 33: 142–148. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. A comprehensive review of the literature and proposed mechanisms. Acta Neurol Scand Suppl 82: 1–66. Jehi, L., and I. M. Namj. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Cleve Clin J Med 75 S2: S66–S70. Jobst, B. C., and G. L. Holmes. 2004. Prescribing antiepileptic drugs: Should patients be switched on the basis of cost? CNS Drugs 18: 617–628. Johannessen, S. I., and T. Tomson. 2006. Pharmacokinetic variability of newer antiepileptic drugs: When is monitoring needed? Clin Pharmacokinet 45: 1061–1075. Kasteleijn-Nolst Trenite D. G., D. J. Bakker, C. D. Binnie, A. Buerman and M. Van Raaij. 1988. Psychological effects of subclinical epileptiform EEG discharges: I. Scholastic skills. Epilepsy Res 2: 111–116. Kasteleijn-Nolst Trenite, D. G., and R. Vermeiren. 2005. The impact of subclinical epileptiform discharges on complex tasks and cognition: Relevance for aircrew and air traffic controllers. Epilepsy Behav 6: 31–34. Kelly, K., L. J. Stephen, G. J. Sills, and M. J. Brodie. 2002. Topiramate in patients with learning disability and refractory epilepsy. Epilepsia 43: 399–402. Kenneback, G., M. Ericson, T. Tomson, and L. Bergfeldt. 1997. Changes in arrhythmia proἀle and heart rate variability during abrupt withdrawal of antiepileptic drugs. Implications for sudden death. Seizure 6: 369–375. Kloster, R., H. C. Borresen, and P. Hoff-Olsen. 1998. Sudden death in two patients with epilepsy and the syndrome of inappropriate antidiuretic hormone secretion (SIADH). Seizure 7: 419–420.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

785

Knudsen, J. F., and G. Sokol. 2006. Correlates of neurocognitive changes: Role of carbonic anhydrase. J Clin Pharmacol Annual meeting abstracts. Kockelmann, E., C. E. Elger, and C. Helmstaedter. 2003. Signiἀcant improvement in frontal lobe associated neuropsychological functions after withdrawal of topiramate in epilepsy patients. Epilepsy Res 54: 171–178. Kramer, G., A. Biraben, M. Carreno et al. 2007. Current approaches to the use of generic antiepileptic drugs. Epilepsy Behav 11: 46–52. Kuz, G. M., and A. Manssourian. 2005. Carbamapepine-induced hyponatremia: Assessment of risk factors. Ann Pharmacother 39: 1943–1946. Lathers, C. M. 1981. Models for studying sequelae to “Induced myocardial infarction.” In Mammalian Models for Research on Aging, 224–228. Washington, DC: National Academy Press. Lathers, C. M., R. F. Flax, and L. J. Lipka. 1986. The effect of C1 spinal cord transaction or bilateral adrenal vein ligation on thioridazine-induced arrhythmia and death in the cat. J Clin Pharmacol 26: 515–523. Lathers, C. M., and J. E. Leestma. 1990. Psychoactive agents, epilepsy, arrhythmia, and sudden death. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 23, 447–484. New York, NY: Marcel Dekker. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986. Regional distribution of myocardial betaÂ�adrenoceptors in the cat. Eur J Pharmacol 130: 111–117. Lathers, C. M., and L. J. Lipka. 1987. Cardiac arrhythmia, sudden death, and psychoactive agents. J Clin Pharmacol 27: 1–14. Lathers, C. M., and L. J. Lipka. 1986. Chlorpromazine: Cardiac arrhythmogenicity in the cat. Life Sci 38: 521–538. Lathers, C. M., and P. L. Schraeder. 1990. Arrhythmias associated with epileptogenic activity elicited by penicillin. In Epilepsy and Sudden Death, ed. C. M Lathers and P. L. Schraeder, Chapter 10. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42: 123–136. Lathers, C. M., and P. L. Schraeder. 1990. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9: 236–242. Lathers, C. M., P. L. Schraeder, and M. Bungo. 2008a. Mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12: 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Reply. Unanswered questions: SUDEP studies needed. Epilepsy Behav 13: 265–269. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008c. Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders: Role of Stress and Psychosocial Influences, ed. Leo Sher, Chapter 13. New York, NY: Nova Science Publishers, Inc. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008d. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher. New York, NY: Nova Biomedical Books. Lathers, C. M., P. L. Schraeder, and H. G. Claycamp. 2003. Clinical pharmacology of topiramate versus lamotrigine versus Phenobarbital: Comparison of efficacy and side effects using odds ratios. J Clin Pharmacol 43: 491–503. Lathers, C. M., P. L. Schraeder, and N. Tumer. 1993. The effect of phenobarbital on autonomic function and epileptogenic activity induced by the hippocampal injection of penicillin in cats. J Clin Pharmacol 33: 837–844.

786 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67: 247–259. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28: 736–745. Lathers, C. M., W. H. Spivey, R. M. Levin, and N. Tumer. 1990. The effect of dilevalol on cardiac autonomic neural discharge, plasma catecholamines, and myocardial beta receptor density associated with coronary occlusion. J Clin Pharmacol 30: 241–253. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39: 2121–2141. Lehnart, S. E., M. Mongiool, A. Bellinger et al. 2008. Leaky Ca2+ release channel/ryanodine receptor 2 causes seizures and sudden cardiac death in mice. J Clin Invest 118: 2230–2245. Lee, S., V. Sziklas, F. Andermann et al. 2003. The effects of adjunctive topiramate on cognitive function in patients with epilepsy. Epilepsia 44: 339–347. Leestma, J. E., J. F. Annegersm, M. J. Brodie et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38: 47–55. Levy, R. H., and C. Collins. 2007. Risk and predictability of drug interactions in the elderly. Int Rev Neurobiol 81: 235–251. Lhatoo, S. D., Y. Langan, and J. W. Sander. 1999. Sudden unexpected death in epilepsy. Postgrad Med J 75 (890): 706–709. Lipka, L. J., and C. M. Lathers. 1987. Psychoactive agents, seizure production, and sudden death in epilepsy. J Clin Pharmacol 27: 169–183. Lipka, L. J., C. M. Lathers, and J. Roberts. 1988. Does chlorpromazine produce cardiac arrhythmia via the central nervous system? J Clin Pharmacol 28: 968–983. Lund, A., and H. Gormsen. 1985. The role of antiepileptics in sudden death in epilepsy. Acta Neurol Scand 72: 444–446. Lutz, M. T., and C. Helmstaedter. 2005. EpiTrack: Tracking cognitive side effects of medication on attention and executive functions in patients with epilepsy. Epilepsy Behav 7: 708–714. May, T. W., B. Rambec, and U. Jurgens. 1996. Serum concentrations of lamotrigine in epileptic patients: The influence of dose and comedication. Ther Drug Monit 18: 523–531. McNaughton, N. C., C. H. Davies, and A. Randall. 2004. Inhibition of alpha(1E) Ca(2+) channels by carbonic anhydrase inhibitors. J Pharmacol Sci 95: 240–247. Meador, K. J. 2002. Cognitive outcomes and predictive factors in epilepsy. Neurology 58 (8 S5): S21–S26. Meador, K. J. 1994. Cognitive side effects of antiepileptic drugs. Can J Neurol Sci 21: s12–s16. Miller, J. M., and R. P. Kustra. 2005. Cognitive and behavioral effects of lamotrigine and topiramate in healthy volunteers. Neurology 64: 2108–2114. Mintzer, S., P. Boppana, J. Toquri, and A. DeSantis. 2006. Vitamin D levels and bone turnover in epilepsy patients taking carbamazepine or oxcarbazepine. Epilepsia 47: 1586. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 353: 888–893. Opeskin, K., M. P. Burke, S. M. Cordner, and S. F. Berkovic. 1999. Comparison of antiepileptic drug levels in sudden unexpected deaths in epilepsy with deaths from other causes. Epilepsia 40: 1795–1798. O’Rourke, D. K., and C. M. Lathers. 1990. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 15, 239–260. New York, NY: Marcel Dekker. Pack, A. M., and M. J. Morrell. 2004. Epilepsy and bone health in adults. Epilepsy Behav 5 (Suppl 2): S24–S29.

Antiepileptic Drugs Benefit/Risk Clinical Pharmacology

787

Patsalos, P. N., and J. W. Sander. 1994. Newer antiepileptic drugs. Towards an improved risk–beneἀt ratio. Drug Saf 11: 37–67. Perucca, E. 2006. Clinically relevant drug interactions with antiepileptic drugs. Br J Clin Pharmacol 61: 246–255. Physician’s Desk Reference web site, www.pdr.net. Porter, R. J., A. Partiot, R. Sachdeo, V. Nohria, and W. M. Alves. 2007. 205 Study Group. RandomÂ� ized,€ multicenter, dose-ranging trial of retigabine for partial-onset seizures. Neurology 68: 1197–1204. Ramsay, R. E., D. Q. McManus, A. Guterman et al. 1990. Carbamazepine metabolism in humans: Effect of concurrent anticonvulsant therapy. Ther Drug Monit 12: 235–241. Reingardiene, D., and J. Vilcinskaite. 2007. QTc-prolonging drugs and the risk of sudden death. Medicina (Kaunas) 43: 347–353. Ruiz Gines, M. A., S. Garcia Garcia, J. A. Ruiz Gines, E. Tze Kiong, and E. Ferrnadez Rodriquez. 2007. Symptomatic secondary hyponatremia due to combined treatment anticonvulsant and antidepressant: Risk of sudden death in epilepsy? An Med Interna 24: 335–338. Schachter, S. C. 1999. Antiepileptic drug therapy: General treatment principles and application for special patient populations. Epilepsia 40 (Suppl 9): S20–S25. Schachter, S. C. 2002. Drug-mediated antiepileptogenesis in humans. Neurology 59 (9s 5): S34–S35. Schraeder, P. L., and G. G. Celesia. 1977. The effects of epileptogenic activity on auditory evoked potentials in cats. Arch Neurol 34: 677–682. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32: 1371–1382. Schraeder, P. L., and C. M. Lathers. 1995. Clinical pharmacology of antiepileptic drug use: “Clinical pearls about the perils of Patty.” J Clin Pharmacol 35: 1120–1135. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3: 55–62. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008a. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13: 263–264. Scorza, F. A., R. M. Cysneiros, R. M. Arida, V. C. Terra-Bustamante, M. de Albuquerque, and E. A. Cavalheiro. 2008b. The other side of the coin: Beneἀciary effect of omega-3 fatty acids in sudden unexpected death in epilepsy. Epilepsy Behav 13 (2): 279–283. Shinoda, M., M. Akita, M. Hasegawa, T. Hasegawa, and T. Nabeshima. 1996. The necessity of adjusting the dosage of zonisamide when coadministered with other anti-epileptic drugs. Biol Pharm Bull 19: 1090–1092. Sirsi, D., and J. E. Safdieh. 2007. The safety of levetiracetam. Expert Opin Drug Saf 6: 241–250. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72: 340–345. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1990. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. In Epilepsy and Sudden Death, ed. C.€M. Lathers and P. L. Schraeder, Chapter 14, 221–238. New York, NY: Marcel Dekker. Stephen, L. J. 2003. Drug treatment of epilepsy in elderly people: Focus on valproic acid. Drugs Aging 20: 141–152. Steinhoff, B. J. 2008. Pregnancy, epilepsy, and anticonvulsants. Dialogues Clin Neurosci 10: 63–75. Vainionpaa, L. K., K. Mikkonen, J. Rattya et al. 2004. Thyroid function in girls with epilepsy with carbamazepine, oxcarbazepine, or valproate monotherapy and after withdrawal of medication. Epilepsia 45: 197–203. Tennis, P., T. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36: 29–36. Timmings, P. L. 1998. Sudden unexpected death in epilepsy: Is carbamazepine implicated? Seizure 7: 289–291.

788 Sudden Death in Epilepsy: Forensic and Clinical Issues Timmings, P. L. 1993. Sudden unexpected death in epilepsy: A local audit. Seizure 2: 287–290. Tomson, T., A. C. Sköld, P. Holmgen, L. Nilsson, and B. Danielsson. 1998. Postmortem changes in blood concentrations of phenytoin and carbamazepine: An experimental study. Ther Drug Monit 20: 309–312. Toth, V., L. Hejjel, Z. Kalmar et al. 2008. Effect of epileptic seizures on the heart rate. Ideggyogy Sz 61: 155–161. Tremmel, L., M. Holtmann, M. H. Schmidt, and U. Brandl. 2006. Do subclinical epileptiform discharges really affect short-term memory in children? Z Kinder Jugendpsychiatr Psychother 34: 139–148. Vainionpää, L. K., K. Mikkonen, J. Rättyä et al. 2004. Thyroid function in girls with epilepsy with carbamazepine, oxcarbazepine, or valproate monotherapy and after withdrawal of medication. Epilepsia 45: 197–203. Warner, G., and D. P. Figgitt. 2005. Pregabalin: As adjunctive treatment of partial seizures. CNS Drugs 19: 265–272. Wilby, J., A. Kainth, N. Hawkins et al. 2005. Clinical effectiveness, tolerability and cost-effectiveness of newer drugs for epilepsy in adults: A systematic review and economic evaluation. Health Technol Assess 9: 1–157, iii–iv.

Clinical Pharmacology and SUDEP Claire M. Lathers Paul L. Schraeder

48

Contents 48.1 Clinical Problem of Sudden Death in Persons with Epilepsy 48.2 Preclinical Animal Studies Leading to Clinical Studies in Persons with€Epilepsy 48.2.1 Animal Models of Ouabain-Induced or Coronary Occlusion–Induced Arrhythmia and Death 48.3 Clinical Studies in Persons with Epilepsy 48.4 Criteria for Sudden Unexplained Death in Epileptic Persons 48.5 Incidence of Sudden Unexpected Death in Epileptic Persons 48.6 Risk Factors for SUDEP 48.6.1 Low Drug Levels due to Variations in Bioavailability, Drug Metabolism, and/or Compliance? 48.6.2 Patients with Epilepsy Who Do Not Receive Antiepileptic Drug due to a Dispensing Error 48.7 Conclusions References

789 790 790 791 792 793 794 794 795 796 797

48.1â•… Clinical Problem of Sudden Death in Persons with Epilepsy Kuller (1996) and Kuller and Lilienfeld (1996) deἀned the disease process and percentage of cases of sudden and unexpected death in a general forensic autopsy population as follows: heart and/or aorta, 56.1%; respiratory, 14.5%; brain and meninges, 15.8%; digestive/urogenital, 8%; and miscellaneous, 9.5%. Rodin (1968) has stated that life ends earlier, in general, for the person with epilepsy. The mortality ratio is two to three times that found in the general population (Hauser et al. 1980; Kurtzke 1972). As summarized by Wannamaker (1990), death may be the consequence of natural or unnatural causes, such as accidents, homicide, and suicide, that have no relationship to epilepsy. Direct causes of death include status epilepticus, and indirect causes may be head trauma or drowning subsequent to a seizure. When death occurs suddenly and without explanation, the term sudden unexpected unexplained death is used. Sudden is deἀned as death occurring within 1 h. Leestma (1990) has discussed additional time frames as deἀnitions of sudden because forensic pathologists prefer more precise limits such as less than 1 h, less than 2 h, less than 12 h, and so on. Unexpected notes that death was not imminent. To date, unexpected implies that there are no symptoms or antecedent illness that would predict that death may be imminent. Most persons with epilepsy have many seizures without lethal outcome, and thus it is difficult for the neurologist to ascribe the unexpected, unexplained death of 789

790 Sudden Death in Epilepsy: Forensic and Clinical Issues

a patient with epilepsy to a single convulsion. Unexplained is a term that clinicians and research scientists are working to clarify.

48.2â•…Preclinical Animal Studies Leading to Clinical Studies in Persons with Epilepsy 48.2.1â•…A nimal Models of Ouabain-Induced or Coronary Occlusion–Induced Arrhythmia and Death In 1964, Han and Moe established that sympathetic nerve stimulation, ouabain toxicity, or coronary occlusion increased temporal dispersion of recovery of ventricular excitability and concluded that this led to an underlying electrical instability that predisposed the ventricular myocardium to arrhythmia. Lathers et al. (1974, 1977, 1978) ἀrst reported cardiac arrhythmias in an animal model for ouabain-induced toxicity that were associated with neural autonomic dysfunction. All studies established that the neural discharges were characterized by increases, decreases, or no change in the discharge of postganglionic cardiac sympathetic nerves monitored simultaneously in the same cat. Lathers and Schraeder (1982) concluded that the peripheral neural nonuniform sympathetic discharge initiated cardiac arrhythmia in the manner described by Han and Moe (1964). The conclusion was also supported by the ἀnding of Randall et al. (1968) that stimulation of the sympathetic ventrolateral cardiac nerve produced a shift in the origin of the pacemaker and tachyarrhythmias. This ἀnding was attributed to the anatomical observation that the nerve is not uniformly distributed to the various regions of the heart but is speciἀcally localized to the atrioventricular junctional and ventricular regions. This nonuniform distribution of sympathetic nerves would also contribute to initiation of arrhythmia if a nonuniform neural discharge occurred regardless of the underlying cause. Lathers et al. (1975) extended the ἀndings of the ouabain-induced nonuniform neural discharge associated with the arrhythmia model to an animal model of arrhythmia and sudden coronary death and reported that the nonuniform autonomic neural discharge was also associated with the occurrence of cardiac arrhythmias and/or sudden death induced by coronary occlusion demonstrating that differences in the ability of different digitalis glycosides to alter neural discharge existed. For example, the digitalis glycoside digoxin did not initiate a nonuniform neural discharge. Some pharmacological agents were found to modify the neural nonuniform discharge and arrhythmogenic activity of the heart, whereas others did not exert a beneἀcial effect. The beta-blocking agents practolol and solotalol (Lathers 1975), metoprolol, timolol (Lathers 1980a, 1980b), quinidine (Lathers et al. 1977), and procainamide (Lathers et al. 1976) did modify the nonuniform neural discharge. Pharmacological agents such as lidocaine (Lathers et al. 1977) and methylprednisolone (Lathers 1983) likewise did not exert an effect on the nonuniform neural discharge and/or the times to arrhythmia and death. Studying the animal model of ouabain-induced toxicity clearly established that appropriate doses of ouabain have a beneἀcial antiarrhythmic effect, whereas higher doses pose the risk of ouabain toxicity–induced arrhythmia and/or death. Just as differences exist in the ability of different glycosides to trigger nonuniform neural discharge, arrhythmia, and death (e.g., the polar molecule ouabain does and the nonpolar digoxin does not), one may theorize that differences exist in the changes associated with various diseases such as myocardial infarction or epilepsy or with different types of epileptogenic disorders. One

Clinical Pharmacology and SUDEP

791

could theorize that in an epileptic patient during a given aberrant epileptogenic discharge, the extent of the contribution of the nonuniform peripheral sympathetic discharge may vary so that during some events, the epilepsy patient may exhibit an arrhythmia, yet not continue on to a fatal event. The same patient may be at greater risk for sudden death in persons with epilepsy (SUDEP) during another event characterized by an efferent pattern of peripheral neural nonuniform discharge that is predisposed with a combination of other circumstances to actually elicit a “fatal event.” It is of interest, when considering persons with epileptogenic activity, to theorize that differences may exist in the status of the autonomic nervous system and its contribution to arrhythmia and/or death, just as differences were found in the effects of two digitalis glycosides, ouabain and digoxin. Furthermore, it may also be speculated that a given anticonvulsant drug at a given dose may be more beneἀcial in preventing SUDEP, an effect that perhaps may be determined in part by its interaction with the contribution of the autonomic nervous system at any given point in time. Thus, at some times, a potentially fatal event may be fortunately avoided, whereas at other times SUDEP may occur. The conclusions of all of the studies conducted in cats were reviewed in detail by Schraeder and Lathers (1989) and Lathers and Schraeder (1987). Lathers and Schraeder (1990) published the ἀrst book focusing on the clinical problem of epilepsy and sudden death. The ἀrst half of the book described the animal studies investigating possible mechanisms of SUDEP, whereas the second half of the book described the state of current clinical knowledge of SUDEP. All contributing authors acknowledged the paucity of clinical data addressing the mechanism of death. Bigger (1990) wrote in his forward to the book: “The current volume assembles a spectrum of international experts who discuss the latest experimental and clinical information available about sudden death and epilepsy. The contributors all struggle for insight into this problem, which resembles the ἀeld of sudden cardiac death 20 years ago while we seem a long way from understanding and controlling the problem of sudden death in epilepsy, the primary question is how shall we proceed toward those goals? Additional epidemiological inquiry should sharpen the focus on the high risk groups. We are on the threshold of major advances in addressing the problem of sudden death in epilepsy. The tools to advance our knowledge are at hand and we should energetically put them to use for the future beneἀt of patients with epilepsy. Knowledge gained from studies of sudden death in epileptics will very likely be useful for understanding sudden death in other situations as well.” To date, unlike the walls of Constantinople, the “threshold of major advances” has not been breached.

48.3â•…Clinical Studies in Persons with Epilepsy By the convening of a panel of experts to examine the question of what product labeling should be requested of the pharmaceutical industry, the FDA focused attention of practitioners and pharmaceutical manufacturers on the question of whether use of anticonvulsant drugs contributes to or prevents SUDEP. Consideration of the possibility of sudden unexpected death in epileptic persons when developing new anticonvulsant drugs was emphasized by the FDA-convened panel of scientists assigned to review data on the risk of sudden unexpected death in epileptic persons taking lamotrigine (Leestma et al. 1997). Other “new” anticonvulsant drugs, for example, gabapentin, topiramate, tigabine,

792 Sudden Death in Epilepsy: Forensic and Clinical Issues

and zonisamide, also have FDA-required warning labels with data on the risk of SUDEP in association with use of each drug. Another long-term beneἀcial effect of bringing SUDEP to the attention of epilepsy researchers has been the development of epidemiological studies devoted to this phenomenon. The original data from the Cook County coroner’s office have been supplemented by numerous other studies in various populations and have mostly conἀrmed the conclusions that SUDEP is a common cause of death in persons with epilepsy, with the risk increasing with increasing refractoriness of the seizure disorder. Some of the most recent studies found that the risk of sudden death in epilepsy was 24 times greater than in the general population (Ficker et al. 1998). The overall incident rate is 1:680/year but varies to some degree with the severity of seizure disorder with an incidence of 1:100 in populations most refractory to treatment (Langan et al. 1998). These epidemiological data conἀrm the conclusion that SUDEP is a risk facing any person with epilepsy.

48.4â•…Criteria for Sudden Unexplained Death in Epileptic Persons To avoid the problem of the lack of a standardized deἀnition of SUDEP associated with previous studies, Leestma et al. (1997) developed and then applied an algorithm for SUDEP in 1997. The criteria for SUDEP (deἀnite or highly probable) were as follows: 1. The subject had epilepsy, as deἀned by Gastaut (1973) and the World Health Organization (WHO): “a chronic disorder characterized by recurrent seizures due to excessive discharge of cerebral neurons.” Because all patients had had chronic and usually intractable epilepsy (according to their physicians who had prescribed one or more antiepileptic drugs for many years for their patients’ seizures), it was assumed that Gastaut/WHO criteria were met. 2. The subject died unexpectedly while in a reasonable state of health. 3. The fatal attack occurred suddenly. The complexity of this deἀnition was discussed. It was recognized that the ἀnal ictus must occur precipitously and unexpectedly but that death might not occur for several hours. Death may have occurred presumably from a seizure-associated cardiorespiratory arrest and its complications and not from status epilepticus. Sudden collapse and death may also have occurred without an observable seizure. 4. The death occurred during normal activities (e.g., at work, at home, in or around bed) in benign circumstances. 5. An obvious medical cause of death was not found. (An exception would be the presence of sudden cardiac arrhythmia, which may be related to the mechanism of SUDEP. Death in water if the victim does not show evidence of drowning may also be attributable to SUDEP.) 6. SUDEP was excluded in the presence of status epilepticus or acute trauma in the setting of a seizure. The classiἀcation of “possible SUDEP” was assigned when cases met most or all of these criteria for SUDEP but data suggested more than one possible cause of death [e.g., deaths associated with seizures while in the bath or swimming, or aspiration (conἀrmed or suspected) occurring concurrently with a seizure]. It was recognized that the classiἀcation of

Clinical Pharmacology and SUDEP

793

drowning deaths is difficult, and unless such deaths were witnessed or other information was available, a level of indeterminateness of classiἀcation was appropriate. The “other (non-SUDEP)” classiἀcation was assigned to cases in which the criteria for SUDEP (deἀnite or highly probable) or possible SUDEP were not met and an obvious other cause of death had been established. The classiἀcation “insufficient data” was assigned to cases that could not be properly interpreted because of missing or ambiguous data related to the circumstances of death or concurrent medical condition.

48.5â•…Incidence of Sudden Unexpected Death in Epileptic Persons Leestma et al. (1997) emphasized that even when using a well-deἀned population, confounding factors such as differences in the dose, duration of therapy, and concomitant medications must be addressed. For example, valproic acid was allowed as a concomitant medication in the UK-sponsored lamotrigine trials but was not allowed in those conducted in the United States. Although the resulting data indicated a lower SUDEP per patient-year of treatment in the lamotrigine database in the US clinical studies than that reported for the United Kingdom, it was not signiἀcantly different. Data obtained from those patients dying of SUDEP were compared with similarly treated patients who died of non-SUDEP causes. This analysis allowed assessment of how the data from these groups compared with data from the overall population. Males represented 52% of the total lamotrigine trial population, and men represented almost 75% of the SUDEP and possible SUDEP patients, suggesting that this observation might be of little signiἀcance because 63% of the non-SUDEP deaths also occurred in males. No differences were detected between the SUDEP group and the overall trial population in terms of the number of antiepileptic drugs or duration of epilepsy. No evidence was found to implicate high doses of lamotrigine or longer duration of exposure with any increase in the risk of SUDEP. Estimated SUDEP rates ranged from <1.5 to >9 in 1000 patient-years. Most studies investigated SUDEP rates in the general epilepsy population rather than in a population with severe epilepsy. Severe or intractable epilepsy (Tennis et al. 1995) as well as age in the 20- to 40-year range has been suggested to be one of the risk factors for SUDEP. Most patients in the lamotrigine clinical development program were categorized in the highrisk group for SUDEP. By deἀnition, they were refractory to currently available pharmacological treatments and exhibited more severe epilepsy than the general population of patients with epilepsy, with a mean of 6.6 seizures/week at baseline in controlled trials while receiving antiepileptic drug polytherapy. When compared to patients in the literature, the derived sudden death rates for patients in the lamotrigine clinical trial program compared favorably with the rate expected in young adults with severe epilepsy and with that reported in the gabapentin clinical trial population (Glaxowellcome 2000). Because the sudden death rate in the lamotrigine and gabapentin trials was low, even in patients with refractory epilepsy, the power to detect a relation between the severity of epilepsy, treatment, and a rare but catastrophic event of SUDEP was limited. Leestma et al. (1997) concluded that the rate of SUDEP in the lamotrigine trials was within the range of rates expected for this patient population. The rate of SUDEP did not appear to be affected by either lamotrigine dose or duration of therapy with lamotrigine. No evidence was found to suggest that lamotrigine alters the risk of SUDEP in patients

794 Sudden Death in Epilepsy: Forensic and Clinical Issues

with epilepsy. Fewer patients treated with lamotrigine died of SUDEP as compared with those treated with placebo. The lamotrigine SUDEP rate was found to be similar to that reported for a similar population of patients treated with gabapentin. This ἀnding suggests the need for additional investigation into the effect new antiepileptic drugs on the risk of SUDEP.

48.6â•…Risk Factors for SUDEP Wren et al. (2000) identiἀed the incidence, causes, and characteristics of sudden death in individuals ranging from 1 to 20 years of age. Their review of all deaths required obtaining death certiἀcates and additional information, where appropriate, from coroners, pediatricians, physicians, and pathologists. The study examined the resident population of one English health region from 1985 to 1994. It showed that in a population of 806,500 persons, medical causes explained 1017 (40%) deaths, 1236 (49%) were unnatural, and 270 (11%) were sudden. Of the sudden deaths, 142 had a previous diagnosis, with the most common being epilepsy for 49 patients (34%). Infections occurred in 17 (20%) and unsuspected cardiovascular abnormalities in 26 (30%). A total of 41 remained unexplained. It is unclear what proportion of the epilepsy-related deaths was attributable to SUDEP. Although no hard data are extant concerning the risk of SUDEP in persons with epilepsy, it is estimated that this phenomenon accounts for 10–15% of deaths in a general population of persons with epilepsy. In selected populations with risk factors such as mental retardation and frequent seizures, SUDEP may account for up to 22% of deaths in persons followed for many years (Chaney and Eyman 2000). Ficker (2000) concluded that risk factors for SUDEP might include poorly controlled seizures, early onset of epilepsy, and generalized tonic–clonic seizures. Of 57 SUDEP cases examined by Nilsson et al. (1999), 91% had undergone necropsy. It was found that the relative risk of SUDEP increased with the number of seizures per year. The estimated relative risk of SUDEP was 10.16 times greater in those patients with more than 50 seizures per year compared with those with up to two seizures per year. The risk also increased with increased number of antiepileptic drugs taken concomitantly, being as high as 9.89 with the use of three antiepileptic drugs compared with monotherapy. Other major risk factors were early-onset vs. late-onset epilepsy and frequent changes of antiepileptic drug dosage when compared with patients on an unchanged dosage. The associations between SUDEP risk and early onset and SUDEP risk and seizure frequency were weaker for female than for male patients, whereas frequent dose changes showed a stronger association in female patients. The authors suggest improvement of seizure control, and an attempt to avoid polytherapy may reduce the risk of SUDEP. 48.6.1â•…Low Drug Levels due to Variations in Bioavailability, Drug Metabolism, and/or Compliance? Lund and Gormsen (1985) examined the role of antiepileptics in sudden death in epilepsy and evaluated patients treated with one or more of the following anticonvulsants: phenobarbitone, phenytoin, and carbamazepine. They noted that variations in bioavailability and in drug metabolism by enzyme induction, resulting in low drug levels, might be contributing factors to the occurrence of SUDEP. They found subtherapeutic drugs levels

Clinical Pharmacology and SUDEP

795

in half of the cases and lethal concentrations, mainly of phenobarbitone, in one third of the cases. The results were interpreted to suggest that there were severe consequences of noncompliance in the treatment of epilepsy. The authors also emphasized the importance of the antiepileptic blood-drug level control as a monitor of compliance in patients with epilepsy, stressing the value of drug-blood level control. Graves et al. (1988) focused on the importance of preventing fluctuations of concomitant antiepileptic drugs in safety and efficacy trials and in patient care settings. Schachter et al. (1998) discussed observations of low postmortem blood concentrations of antiepileptic drugs in cases of SUDEP. These authors emphasized that although noncompliance may play a role in SUDEP, postmortem phenytoin concentrations should be interpreted with caution because animal data indicate that antemortem whole-blood concentrations were only 65% of the corresponding serum concentrations, and postmortem blood levels were even lower, being 35% of antemortem serum concentrations. The major implication of this study is that low postmortem concentrations do not necessarily imply poor compliance with treatment. In contrast, Opeskin et al. (1999) concluded that there were no data to support the idea that SUDEP patients were less compliant with antiepileptic drug treatment than the control group. They eliminated poor compliance with antiepileptic drug treatment as a risk factor for SUDEP. They also found no evidence that phenytoin or carbamazepine use was associated with a higher risk of SUDEP. May et al. (1999) also noted that the postmortem decrease in anticonvulsant serum concentrations, especially for phenobarbital and phenytoin, should be considered with caution to avoid misinterpretation in relation to the so-called “subtherapeutic” serum levels and noncompliance in relation to SUDEP.

48.6.2â•…Patients with Epilepsy Who Do Not Receive Antiepileptic Drug due to a Dispensing Error Clinical pharmacologists, neurologists, pharmacists, nurses, and all medical health professionals should also be aware of another issue that could contribute to the risks faced by persons with epilepsy, namely, the name of the drugs that they use. For example, the patent name for Lamictal (lamotrigine), an anticonvulsant, has been confused with Lamisil (terbinoἀne hydrochloride), an antifungal drug. Patients erroneously receiving either medication would be unnecessarily subjected to the risk of adverse events. In addition, patients with epilepsy who do not receive their antiepileptic drug because of a dispensing error would be inadequately treated and could experience serious consequences, including status epilepticus. Patients receiving Lamictal instead of Lamisil would not have the dose of Lamictal properly dispensed and would unnecessarily be subjected to a risk of potential side effects, including Stevens–Johnson syndrome. Clear communication of oral and written prescriptions will help avoid future dispensing errors. An earlier problem, related to drugs with similar names that could easily be confused when the pharmacist was dispensing, involved the antiepileptic drug clonazepam. One of the authors (PLS) had the experience many years ago of prescribing clonazepam only to ἀnd out that the patient was nearly in a coma associated with hypotension occurring when clonidine was accidentally dispensed by the pharmacist. This potential problem originated when “klonapin” was spelled “clonapin” as a brand name and was commonly confused with clonidine. This problem was simply corrected when the spelling of the name “clonapin” was changed from a “c” to a “k.”

796 Sudden Death in Epilepsy: Forensic and Clinical Issues

48.7â•… Conclusions In conclusion, the 1993 effort of the FDA (Leestma et al. 1997) subsequently focused attention of practitioners and pharmaceutical manufacturers on the question of whether use of anticonvulsant drugs contributes to or prevents sudden unexpected death in epileptic persons. Consideration of the possibility of sudden unexpected death in epileptic persons when developing new anticonvulsant drugs was emphasized by the FDA-convened panel of scientists assigned to review data on the risk of sudden unexpected death in epileptic persons in patients taking lamotrigine. The sudden unexpected death rate in persons with epilepsy in this cohort of patients was comparable to that expected in young persons with poorly controlled epilepsy. Estimated sudden unexpected death rates in patients receiving any of these newer anticonvulsant drugs are similar to those observed in patients receiving other anticonvulsant drugs. Cumulative evidence suggests that sudden unexpected death rates in epileptic persons reflect population rates and not a speciἀc drug effect. Five newer anticonvulsant drugs—lamotrigine, gabapentin, topiramate, tigabine, and zonisamide— have FDA-required warning labels with data on the risk of sudden unexpected death in epileptic persons in association with use of each drug. None of these ἀve newer anticonvulsant drugs have shown an associated increased risk of sudden unexpected death, nor did their use show a decreased risk. The status of clinical studies of SUDEP leaves much to be desired. Although some epidemiology studies have been done, many questions remain to be answered. 1. It is difficult to conduct an epidemiological study when lacking good postmortem examinations on persons with epilepsy who have died. 2. Even if postmortem studies have been done, there is a lack of standardized autopsy end points, including a lack of microscopic cardiac examinations emphasizing cardiac neural elements. 3. Most epidemiology studies of SUDEP lack a signiἀcant n to reach any deἀnitive conclusion regarding the risk factor for SUDEP. 4. Clinical pharmacological studies that determine the best use of drugs in terms of preventing SUDEP need to be designed. 5. The observation that poor seizure control is a risk factor for SUDEP should be an incentive for practitioners to work on improving patient compliance with antiepileptic drug use. 6. The fact that many other SUDEP victims have relatively low seizure frequency implies that leaving any seizure untreated places the patient at risk. Whenever a particular drug is selected to treat a given patient, consideration must be made of the fact that each pharmacological agent has the capability of providing the beneἀcial therapeutic effect if used in the correct dose range but may also produce unwanted, unexpected adverse effects (Lathers 1996; Lathers and Schraeder 1995). Some subpopulations may be more sensitive to a given drug than most of the patients. This problem is confounded in the epileptic population, especially because the individual may be at risk for sudden unexpected death from the disease state itself. Each person with epilepsy must be evaluated to determine if he or she may be at risk for SUDEP and to determine that the antiepileptic drug treatment regimen selected is the best therapy. It is essential that the drugs themselves do not place them at greater risk for SUDEP. At this time, the answer

Clinical Pharmacology and SUDEP

797

to the question of whether one or the other of the antiepileptic drugs, when used alone or in combination, have the effect of increasing or decreasing the risk of SUDEP is also unknown. One possible answer to the question posed appears to be that a given drug in a given patient with epilepsy during a given state of epileptogenic activity may be more beneἀcial or may lead to SUDEP, depending on the overall “mix” of these events during the given risk event. Good patient care is essential to provide continued regular evaluation and monitoring of each patient by his or her physician. Optimizing seizure control with antiepileptic drugs is the ἀrst line of defense in the prevention of SUDEP.

References Bigger, J. T. 1990. Foreword. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, v–vii. New York, NY: Marcel Dekker. Chaney, R. H., and R. K. Eyman. 2000. Patterns in mortality over 60 years among persons with mental retardation in a residential facility. Ment Retard 38: 289–293. Ficker, D. M. 2000. Sudden unexplained death and injury in epilepsy. Epilepsia 41 (S2): S7–S12. Ficker, D. M., E. L. So, W. K. Shen et al. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51: 1270–1274. Gastaut, H. 1973. Dictionary of Epilepsy. Geneva: World Health Organization. Glaxowellcome. 2000. Prescribing information for Lamictal (lamotrigine), 2000. In Physicians’ Desk Reference, 54th ed., 1208–1214. Montvale, NJ: Medical Economics. Graves, N. M., G. B. Holmes, and I. E. Leppik. 1988. Compliant populations: Variability in serum concentrations. Epilepsy Res Suppl 1: 91–99. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular vulnerability. Circ Res 14: 44–60. Hauser, W. A., J. F. Annegers, and L. R. Elveback. 1980. Mortality in patients with epilepsy. Epilepsia 21: 399–412. Kuller, L. 1966. Sudden and unexpected non-traumatic deaths in adults: Review of epidemiological and clinical studies. J Chron Dis 19: 1165–1192. Kuller, L., and A. Lilienfeld. 1966. Epidemiological study of sudden and unexpected deaths due to arteriosclerotic heart disease. Circulation 34: 1056–1068. Kurtzke, J. F. 1972. Mortality and Morbidity Data on Epilepsy. Pub. No. (NIH) 73-390. Washington, DC: Department of Health, Education and Welfare. Langan, Y., N. Nolan, and M. Hutchinson. 1998. The incidence of sudden unexpected death in epilepsy (SUDEP) in South Dublin and Wicklow. Seizure 7: 355–358. Lathers, C. M. 1980a. Effect of metoprolol on coronary occlusion–induced arrhythmia and autonomic neural discharge. Fed Proc 39: 771. Lathers, C. M. 1980b. Effect of timolol on autonomic neural discharge associated with ouabaininduced arrhythmia. Eur J Pharmacol 64: 95–106. Lathers, C. M. 1982. Lack of effect of methylprednisolone on cardiac neural discharge associated with coronary occlusion-induced arrhythmia and death. Eur J Pharmacol 85: 233–238. Lathers, C. M. 1983. Failure of methylprednisolone to prevent nonuniform cardiac accelerator nerve discharge associated with coronary occlusion-induced arrhythmia: Evidence against prostaglandin modulation of autonomic cardioaccelerator neural discharge in the anesthetized CAT. Med Hypothesis 10: 43–57. Lathers, C. M. 1996. Drug development: Role of academia, government, industry, and CRO’s. J Clin Pharmacol 36: 1–2. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism of coronary occlusion and digitalis-induced arrhythmia. Circulation 57: 1058–1064.

798 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and J. Roberts. 1985. Are the sympathetic neural effects of digoxin and quinidine involved in their action on cardiac rhythm? J Cardiovasc Pharmacol 7: 350–360. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1974. Relationship between the effect of ouabain on arrhythmia and interspike intervals (I.S.I.) of cardiac accelerator nerves. The Pharmacologist 16: 201. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1975. Relationship of arrhythmias to digoxin and ouabain action on cardiac accelerator nerves. The Pharmacologist 17: 261. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203: 467–479. Lathers, C. M., J. Roberts, G. J. Kelliher, and A. B. Beasley. 1975. Comparison of the neural component in coronary occlusion and digitalis-induced arrhythmia. Clin Res 23: 566A. Lathers, C. M., J. Roberts, G. J. Kelliher, and A. B. Beasley. 1976. A comparison of procainamide and lidocaine effects on nerve discharge associated with coronary occlusion-induced arrhythmia in the cat. The Pharmacologist 18: 169. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23: 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27: 346–356. Lathers, C. M., and P. L. Schraeder. 1990. Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 1995. Experience-based teaching of therapeutics and clinical pharmacology of antiepileptic drugs: Sudden unexplained death in epilepsy: Do antiepileptic drugs have a role? J Clin Pharmacol 35: 563–586. Lathers, C. M., A. J. Teres, A. B. Beasley, A. B. Malhotra, G. J. Kelliher, and J. Roberts. 1977. Cardiac accelerator nerve discharge and lidocaine blood levels after coronary occlusion. Fed Proc 36: 1002. Lathers, C. M., A. J. Teres, and G. J. Wetmore. 1977. Effect of quinidine on splanchnic and cardiac sympathetic neural discharge associated with digoxin-induced arrhythmia. Clin Res 25: 652A. Leestma, J. E. 1990. Sudden unexpected death associated with seizures: A pathological review. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 61–88. New York, NY: Marcel Dekker. Leestma, J. E., A. F. Annegers, and M. J. Brodie, et al. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical trial. Epilepsia 38: 47–55. Lund, A., and H. Gormsen. 1985. The role of antiepileptics in sudden death in epilepsy. Acta Neurol Scand 72: 444–446. May, T., U. Jurgens, B. Rambeck, and R. Schnabel. 1999. Comparison between premortem and postmortem serum concentrations of phenobarbital, phenytoin, carbamazepine and its 10,11epoxide metabolite in institutionalized patients with epilepsy. Epilepsy Res 33: 57–65. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case-control study. Lancet 13: 888–893. Opeskin, K., M. P. Burke, S. M. Cordner, and S. F. Berkovic. 1999. Comparison of antiepileptic drug levels in sudden unexpected deaths in epilepsy with deaths from other causes. Epilepsia 40: 1795–1798. Randall, W. C., M. Szentivanyi, J. B. Pace, J. S. Wechsler, and M. P. Kaye. 1968. Patterns of sympathetic nerve projections onto the canine heart. Circ Res 22: 315–323. Rodin, E. A. 1968. The Prognosis of Patients with Epilepsy, 326–329. Springἀeld, IL: Charles C Thomas. Schachter, S. C., G. W. Cramer, G. D. Thompson, R. J. Chaponis, M. A. Mendelson, and L. Lawhorne. 1998. An evaluation of antiepileptic drug therapy in nursing facilities. J Am Geriatr Soc 46: 1137–1141.

Clinical Pharmacology and SUDEP

799

Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3: 55–62. Tennis, P., T. H. B. Cole, J. F. Annegers, J. E. Leestma, M. McNutt, and A. Rajput. 1995. Cohort study of incidence of sudden unexplained death in persons with seizure disorder treated with antiepileptic drugs in Saskatchewan, Canada. Epilepsia 36 (1): 29–36. Wannamaker, B. B. 1990. A perspective on death of persons with epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 26–37. New York, NY: Marcel Dekker. Wren, C., J. J. O’Sullivan, and C. Wright. 2000. Sudden death in children and adolescents. Heart 83: 410–413.

Experience-Based Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs Sudden Unexplained Death in Epilepsy: Do Antiepileptic Drugs Have a Role?

49

Claire M. Lathers Paul L. Schraeder

Contents 49.1 49.2 49.3 49.4 49.5

Introduction Deἀnition and Classiἀcation of Epilepsy Epidemiology of Epilepsy Sudden, Unexpected, Unexplained Death in the General Population Sudden, Unexpected, Unexplained Death in the Epileptic Population 49.5.1 Cause of Death in Epilepsy 49.5.2 Cases of Sudden Death in Persons with Epilepsy 49.5.3 Cases of Near Sudden Death Associated with Seizures 49.5.4 Epidemiology of Sudden Death in Persons with Epilepsy 49.6 Possible Contributing Factors in the Pathogenesis of Cardiac Sudden Death 49.7 Possible Mechanisms for Autonomic Dysfunction and Sudden Death in Persons with Epilepsy 49.8 Some Clinical Pharmacologic Considerations in the Management of Epilepsy 49.8.1 Postmortem Toxicologic Findings in Victims of Sudden Unexplained Death in Epilepsy 49.8.2 The Ideal Pharmacologic Agent 49.8.3 Generic Substitutions of Antiepileptic Drugs 49.8.4 Federal Guidelines for Generic Antiepileptic Drugs 49.8.5 Consequences of Generic In Vivo Criteria 49.8.6 Generic Bioequivalence 49.8.7 Cases Illustrating Potential Problems Involved with the Use of Generic Antiepileptic Drugs 49.8.8 Drug Interactions 49.9 Conclusion References Appendix I. Self-Assessment Quiz

801

802 802 802 803 804 804 805 805 807 809 810 811 811 813 814 815 816 816 816 817 819 820 824

802 Sudden Death in Epilepsy: Forensic and Clinical Issues

49.1â•…Introduction Sudden unexplained death is an unfortunately frequent cause of mortality in young persons with epilepsy, accounting for up to 10% (Leestma 1990a) of all deaths in the epileptic population. Almost all persons with epilepsy take one or more antiepileptic drugs for many years. Yet many if not most victims of sudden unexplained death are found to have low or no measurable plasma levels of antiepileptic drugs. This latter observation suggests that low levels of antiepileptic drugs or possibly their withdrawal may predispose to sudden unexplained death. The risk/beneἀt of antiepileptic drugs in sudden unexplained death has recently become a concern of several investigators (American Academy of Neurology 1993). Although there are no ἀrm data, the appropriate plasma levels of antiepileptic drugs may exhibit a protective effect by decreasing the contribution of some of the factors, such as metabolism and bioavailability, that may be contributory to myocardial electrophysiologic instability and/or the associated sudden death.

49.2â•…Definition and Classification of Epilepsy Epilepsy has been deἀned by Gastaut (1973) as “chronic brain disorder of various etiologies characterized by recurrent seizures due to excessive discharge of cerebral neurons.” This deἀnition excludes single or occasional seizures that may be triggered by trauma, high fever, toxic states, syncopal attack, a ἀt of rage, somatic dysfunction, movement disorder, or behavioral reaction. In 1981, the International League against Epilepsy (1981) classiἀed epileptic seizures and syndromes into three major categories: (1) focal (partial or local); (2) generalized (convulsive or nonconvulsive); and (3) unclassiἀed. Abnormal electrical behavior of a limited population of neurons are focal seizures, whereas those involving large populations of neurons throughout the brain are designated as generalized. The latter group is subdivided into absence (petit mal), myoclonic, clonic, tonic, tonic–clonic, and atonic seizures, or combinations of the above. Seizures not included in the ἀrst two classiἀcations are designated as unclassiἀed. Repetitive or continuous seizures occurring in any category are termed status epilepticus.

49.3â•… Epidemiology of Epilepsy The incidence of epilepsy in the general population in the United States has been reported to vary from 30 to 54 in every 100,000 persons (Hauser and Kurland 1975; Kurland 1959). The likelihood of developing epilepsy is age related, with a gradually increasing prevalence with age until 50 years of age; thereafter, the number of cases begins to decline, and for the general population, begins to plateau (Schoenberg 1985). Hoffman (1988) reported that at least 73,000–131,000 individuals developed epilepsy in the United States in 1987. A lifetime prevalence rate for those with epilepsy that is clinically active and are living in the general population of major industrialized nations of the western hemisphere probably ranges between 0.5% and 0.78% (5.7–7.8/1000) (Zielinski 1982). Juul-Jensen and Foldspang (1983) concluded that approximately 1.2% of the population had epilepsy. The types of seizure disorders across the population are varied, with generalized tonic–clonic (grand mal) seizures being the most common in all age groups. Absence (petit mal) seizures are

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

803

100

Percentage of cases

80

60

40

20

0 <1

2–4

5–9

10–14

15–19

20–29

>30

Age group GTC

Absence

Psychomotor

Other

Figure 49.1╇ Types of seizure disorders by percentage of incidence by age group. Note the relative commonness of absence seizures in the younger individuals and its relative rarity in older persons. (Adapted from Parke-Davis Company, Epilepsy: Patterns of Disease. Special Report. Chicago, 1958; and Epilepsy Foundation of America. Basic Statistics on the Epilepsies. F. A. Davis, Philadelphia, PA, 1975.)

the second most common type in young children, but diminish in prevalence with age. Psychomotor (complex-partial) seizures assume the secondary position in adolescents and adults (Figure 49.1). Note that approximately 50% of those with epilepsy will exhibit more than one type of seizure (Lennox 1960).

49.4â•…Sudden, Unexpected, Unexplained Death in the General Population As recently reviewed by Leestma (1990b), the component of sudden death that is “unexpected” is more readily deἀned than the time frame of the phenomenon. Some deἀne sudden death as that occurring between 1 and 24 h; others as less than 1 h, less than 2 h, or less than 12 h; some refer only to that occurring in min; and others distinguish between “instant” death, occurring in seconds, and “rapid” death, occurring in a few seconds up to about 10 min (Leestma et al. 1997). Because there are numerous deἀnitions of the time frame of unexpected sudden death, physicians should be careful to report the time frame in those cases where it is known to ultimately allow a clearer understanding of the contributory mechanisms involved and how to prevent their occurrence.

804 Sudden Death in Epilepsy: Forensic and Clinical Issues

However, some forensic pathologists deemphasize the importance of the time interval and focus on the unexpected aspect of the death. The aspect of unexplained death adds still yet another dimension to the deἀnition of sudden death (Leestma et al. 1985). The focus of this discussion is to present hypotheses on the mechanism(s) of sudden, unexpected, unexplained death in persons with epilepsy, and to discuss pharmacologic management of persons with epilepsy to attempt to avoid sudden death. In general, somewhat less than 1% of the population dies in any given year. In Cook County, Illinois, only about half of all deaths (15,000–17,000 per year) are reported to the Medical Examiner’s Office because (1)€there was no physician in attendance who could or would execute a death certiἀcate; (2)€foul play or criminal activity might have been involved; (3) the individual died within 24 h after admission to a hospital or during childbirth, surgery, or some other treatment; (4) the individual died in an institution or while incarcerated; or (5) the individual died in an accident (Leestma et al. 1985; Leestma 1990a). Sudden deaths account for at least 20% of all deaths in most urban areas and between 450,000 and 500,000 each year in the entire United States (Leestma 1990a). Approximately 65% of all reported cases are judged to have died of “natural” causes, from homicide, 4–5% of suicide, and 20% from accidents. In approximately 5% of the deaths, no “manner” of death could be determined. Although many individuals die suddenly by accident, homicide, or suicide, they could hardly be regarded as having died suddenly and unexpectedly or without obvious explanation. Most of the unexpected nontraumatic death cases fall into the “death-by-natural-causes” category (Leestma 1990a; Leestma et al. 1985). Anatomic classiἀcations of the autopsy ἀndings for 20,981 patients who were declared to have died from sudden death included the following: heart and/or aorta, respiratory, brain and meninges, digestive/urogenital, and miscellaneous. The miscellaneous category included sudden death occurring during no event or associated with surgery, obstetric delivery, bronchoscopy or radiologic procedures, hormone-secreting tumors such as pheochromocytomas and carcinoids, hemorrhaging tumors, or fat embolism (Luke and Helpern 1968; Kuller et al. 1967, 1974).

49.5â•…Sudden, Unexpected, Unexplained Death in the Epileptic Population 49.5.1â•… Cause of Death in Epilepsy In 1868, Bacon (1868), the medical superintendent of Cambridge County Asylum, classiἀed the causes of death in epilepsy into four categories: (1) those arising from the longcontinued effects of the disease on the body; (2) deaths after a rapid succession of ἀts; (3) sudden deaths in a ἀt; and (4) accidents due to ἀts. Category 3 was further deἀned into three subsets: (1) asphyxia from the spasms; (2) mechanical suffocation; and (3) sudden loss of nervous power, due most probably to the state of the heart or its nerves. In 1902, Spratling (1902) reported that approximately 4% of deaths that occurred over a number of years in a large population of institutionalized epileptic individuals could be classiἀed as sudden, unexpected, and attributable to no cause on postmortem examination. Numerous recent articles have summarized sudden, unexpected, unexplained deaths in persons with epilepsy once demonstrable causes of status epilepticus, accidents, drowning,

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

805

drug overdoses, and intercurrent disease, etc., have been eliminated (Jay and Leestma 1981; Leestma et al. 1989). The phenomenon of sudden, unexpected, unexplained death in epileptic persons probably accounts for more than 10% of deaths in persons with epilepsy (Jay and Leestma 1981; Leestma et al. 1985; Leestma 1990a). The annual incidence of sudden death among individuals with epilepsy is at least 1 in every 500 to 1000, and perhaps higher (Jay and Leestma 1981). Annegers and Blakley (1990) reported the annual risk of sudden death to be 1 in 738, but noted that the annual risk for those older than 30 years of age was 1 in 371. They also noted that the risk varied by etiology of epilepsy: the annual risk was only 1 in 1000 in persons with idiopathic epilepsy compared with 1 in 100 for those with remote symptomatic epilepsy (i.e., symptoms caused, e.g., by a previous injury to the head as opposed to idiopathic epilepsy, in which there is no speciἀc structural or metabolic cause of the primary generalized epileptogenic activity).

49.5.2â•…Cases of Sudden Death in Persons with Epilepsy

Case 1 Terrence et al. (1981) described a sudden death occurring in a 19-year-old white male susceptible to grand-mal seizures since the age of 5 years. Before his death he was taking phenytoin, 100 mg, and phenobarbital, 30 mg, each three times a day. His seizures were infrequent and he was reported to go for days without taking his medications. On the evening before his death, the victim drank two small glasses of beer and returned home at 2:00 a .m. At 10:00 a .m., he was awakened to go to work; 15 min later he was found on the floor, deeply cyanotic, and in the midst of a grand-mal seizure. A paramedic team reported cardiopulmonary arrest and were unable to successfully resuscitate him. On autopsy, only pulmonary edema was noted. Plasma phenytoin and phenobarbital levels were subtherapeutic (5.8 and 3.2 μg/ml, respectively) and blood alcohol content was zero. Case 2 On a more personal level, one of the authors (PLS) treated a hospitalized woman who had just been stabilized at a therapeutic level of her usual anticonvulsant medication. As the neurologic team made rounds, the patient was told she was doing well and would be discharged later that day. While standing in the hall, they heard a noise in her room and found that she was in asystole. Resuscitation was unsuccessful.

49.5.3â•… Cases of Near Sudden Death Associated with Seizures The often difficult differentiation between seizures caused by cerebral epileptiform discharges and those resulting from cardiac arrhythmias or asystole, that is, syncopal seizures, is illustrated by the following two cases from one of the author’s (PLS) clinical experience.

806 Sudden Death in Epilepsy: Forensic and Clinical Issues

Case 3 A 45-year-old man was admitted to a hospital because of the onset of daily episodes of loss of consciousness without any evidence of seizures. While in the hospital, he had another such episode. An electrocardiogram (ECG) showed multiple premature ventricular beats that resolved once he regained consciousness. He was scheduled for a cardiologic workup, including invasive electrophysiologic studies. Further questioning of the patient revealed that he experienced the sensation of being far away from reality and had a strange taste in his mouth before each loss of consciousness. A prolonged electroencephalogram (EEG) was then obtained with video and ECG monitoring. During this study, he again experienced the sensation of dissociation from reality and had a bad taste in his mouth. The EEG at that time recorded left temporal spikes that proceeded to a bilateral, bitemporal electrographic seizure with onset of tachycardia with frequent premature ventricular beats. He was started on carbamazepine, achieving a serum level of 8 μg/ml and has had no more losses of consciousness. Case 4 The patient, a 14-year-old girl, had a history of generalized tonic–clonic seizures that started at the age of 10 years. She was placed on valproic acid, and a blood level of 55.0 μg/ml was achieved. A neurologic consultant was asked to see the patient because of recurrent seizures. The patient related that most of the events occurred when she saw her urologist for interval cystoscopy (she had a congenital urinary tract malformation) or when she was subjected to phlebotomy. The patient and her parents agreed that an EEG with ECG monitoring would be performed, during which the patient would undergo a phlebotomy. At the moment of skin penetration, the patient complained of nausea, followed by bradycardia progressing to asystole. After 13 s of asystole on ECG, the EEG became silent and a 15-s duration tonic–clonic seizure was observed. After a total of 40 s, the heart spontaneously started with a tachycardia of 120 beats/min. The patient awakened and was immediately alert and oriented. She was given an atropine inhaler to be used before any painful procedures and the valproic acid was discontinued. No more seizures have occurred in the 3-year period subsequent to the withdrawal of valproic acid. Case 3 illustrates that seemingly cardiogenic syncope can be the result of complex partial seizures. The history of symptoms compatible with a temporal lobe aura resulted in speciἀc studies being obtained that established the pathophysiologic mechanism of the cardiac arrhythmia. Case 4, in turn, demonstrates that reproducing the offending circumstances under appropriate monitoring conditions is necessary to establish a primary cardiovascular mechanism for “typical” tonic–clonic seizures. Further discussion of the spectrum of relationships between seizures and the cardiovascular system can be found in the work of Marshall et al. (1983) and Schraeder and Lathers (1989). Schraeder et al. (1983) reported a case involving a young man undergoing treatment with anticonvulsants who had tonic–clonic seizures. The seizures were the

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

807

result of asystole induced by the mental image of pain. This report and cases 3 and 4 described above emphasize the potential for complex interactions between brain and heart in situations in which the patient has bona ἀde clinical seizure activity, albeit cardiogenic or epileptogenic.

McLeod and Jewitt (1978) used 24-h continuous ECG monitoring and noted that a variety of arrhythmias appeared in asymptomatic patients. Schott et al. (1977) using the same monitoring system found that some patients classiἀed as having suspected idiopathic epilepsy and no cardiac symptoms (only one had a transient focal abnormality in the EEG) had serious cardiac arrhythmias that were detected with the continuous ECG monitoring. Treatment with antiarrhythmic agents or the installation of a pacemaker eliminated the seizures and led to the conclusion that arrhythmias associated with seizure may be misdiagnosed as epilepsy more often than is generally recognized. Howell and Blumhardt (1990) suggest that the diagnostic possibilities include anoxia-induced seizures secondary to the cardiac arrhythmia or perhaps an overlap in the symptoms of hindbrain ischemia and temporal lobe seizures. Ictal EEG recordings are required to separate these two mechanisms. 49.5.4â•… Epidemiology of Sudden Death in Persons with Epilepsy Characteristics of some individuals in a population of patients with epilepsy who died suddenly have been compiled by Leestma et al. (1984, 1985, 1989). The data are based on 500 cases in Chicago, Illinois, during a period of 10 years. The greatest number of cases occurred in the 31- to 40-year-old age group (Figure 49.2). As depicted in Table 49.1, the

60

31-40

40

0

51–60

>60

10

10–20

20

41–50

21–30

30

<10

Number of cases

50

Age groups in years

Figure 49.2╇ Age distribution by decade of 124 sudden unexplained deaths in epileptic persons (sudden unexplained death). The mean and median age is 32 years; the mode is 33 years. (Reproduced from Leestma, J. E., in Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. Marcel Dekker, New York, 1990. With permission.)

808 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 49.1â•… Characteristics of the SUDEP Population Observed over Nearly 10 Years in Cook County, Illinois, Office of the Medical Examiner Age range Average age Sex Race

Seizure history duration

Frequency of seizure

Type of seizure Circumstances of death

4 months to 85 years 31.4 years Male (74%); female (26%) Black male: 49% White male: 25% Black female: 11% White female: 15% More than a year, <5 years: 86% More than 5 years: 54% (Mean duration: 9.5 years; mode duration: 3 years) <1 per year: 9% 3–10 per year: 52% >1 per week: 39% Generalized tonic–clonic: 96–98% Dead in bed: 37% Dead in another room: 25% Dead on arrival at hospital: 22% Other circumstances: 12%

Source: Leestma, J. E., in Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, Marcel Dekker, New York, NY, 1990c. With permission.

mean age€was€31.4 years, and there was a greater prevalence in males than in females (74%) and in black males (49%) than in white males or black and white females. Of those with a seizure history for more than 1 year (86%), a frequency of seizures of 3–10 per year, and a history of having generalized tonic clonic seizures, 37% were found dead in be (Leestma 1990a). It was rare for death to occur in those experiencing less than one seizure per year. The medical history of the victims is summarized in Table 49.2.

Table 49.2â•… Medical History Obtained from Family, Friends, or Other Witnesses of Deceased SUDEP Victims Past Medical/Health History In apparent good health Old head injury Gunshot wound to head History of hypertension Anemia/sickle cell disease Asthma Diabetes Miscellaneous conditions Unknown

Percentage of Cases (%) 23 27 2 8 8 2 2 8 20

Source: Leestma, J. E., in Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. Marcel Dekker, New York, NY, 1990c. With permission.

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

809

49.6â•…Possible Contributing Factors in the Pathogenesis of Cardiac Sudden Death Sudden fatal syncope is thought to involve neurally mediated abnormality of cardiac or vascular function and may be associated with fright, strong emotion, stress, dreams and nightmares, shock, extreme pain, or sudden blows to the head, chest, abdomen, or extremities. Myocardial electrical instability has been suggested as one mechanism that might initiate cardiac arrhythmias and sudden death. Factors that have been suggested to contribute to the pathogenesis of cardiac sudden death may include changes in the autonomic neural control of the heart (Lathers et al. 1978); stress related release of catecholamines (Pickworth et al. 1990); cardiac arrhythmias; embolism, rupture, or abnormalities in the structure of the atrioventricular junction; changes in platelet aggregation; arteriosclerotic coronary artery disease; and cardiopathologic conditions, including rare myocardial tumors (James 1983). As discussed by Lathers and Schraeder (1987), alterations in the autonomic neural control of the rhythm of the heart and enhanced catecholamine release would not be detected at autopsy. Lathers and Schraeder (1982) and Schraeder and Lathers (1983) demonstrated in an animal model that changes in peripheral cardiac sympathetic or parasympathetic neural discharge are associated with interictal subconvulsant epileptiform discharges and cardiac arrhythmia. Autonomic dysfunction was manifested by (1) the fact that autonomic cardiac nerves did not always respond in a predictable manner to changes in blood pressure; (2) the development of a marked increase in variability in mean autonomic cardiac nerve discharge; and (3) the appearance of a very large increase in the variability of the nerve discharge rate of parasympathetic nerves primarily, and secondarily in the sympathetic discharge. It was suggested that the altered autonomic cardiac neural discharge associated with interictal epileptiform activity and arrhythmias may contribute to sudden death in epileptic persons. Additional studies have suggested that various central neural and biochemical changes in animal models could also be associated with alterations of peripheral autonomic neural control of the heart and interictal epileptiform activity, with resultant increased risk for sudden death in epileptic persons (Lathers et al. 1988; Kraras et al. 1987; Lathers and Schraeder 1990; Schwartz and Lathers 1990; Lathers 1990). In animal models of autonomic dysfunctions, postganglionic cardiac sympathetic neural discharge in the minute before onset of arrhythmia was simultaneously increased in one nerve, decreased in another, and/or increased or decreased in the third nerve (Lathers et al. 1974, 1978). Uniform neural discharge is deἀned as all neural activity being uniformly decreased or unchanged or increased; the uniform discharge was hypothesized to be necessary at the cardiac myocardial junction to maintain normal electrical excitability and automaticity, that is, normal sinus rhythm (Figure 49.3). Nonuniform neural discharge, that is, when neural activity is simultaneously increased, decreased, and/or unchanged among the various cardiac autonomic neural components, was hypothesized to be manifested in the heart as inhomogeneity of myocardial electrical excitability and conduction patterns. This was demonstrated by Han and Moe (1964) when they found that myocardial nonuniformity could cause ventricular arrhythmias, including ventricular ἀbrillation. These experimental animal data provide a possible pathophysiologic explanation for sudden autonomic dysfunction in individuals who had no observed clinical seizures and/ or only seizures of minimal severity preceding their demise (Jay and Leestma 1981; Hirsch and Martin 1971; Terrence et al. 1975). Thus, as found in animal experimental studies

810 Sudden Death in Epilepsy: Forensic and Clinical Issues

(a) Postganglionic cardiac sympathetic branches 1 2 3

Uniform neural discharge NSR

(b) Postganglionic cardiac sympathetic branches 1 2 3

Nonuniform neural discharge VF

Figure 49.3╇ Postganglionic cardiac sympathetic neural discharge in three branches innervating the ventricle. (a) Nerve activity in all branches is enhanced above control and is designated a uniform neural discharge. (b) Nerve activity is increased in one branch, decreased in a second, and shows no change in a third; this trend is designated a nonuniform neural discharge. (From Lathers, C. M., and Schraeder, P. L., J Clin Pharmacol, 27 (5), 346–356, 1987. With permission.)

(Lathers and Schraeder 1982), interictal nonconvulsive activity may be associated with cardiac nonuniform neural dysfunction, with consequent abnormalities of conduction and rhythm, making some persons with epilepsy susceptible to fatal arrhythmias.

49.7â•…Possible Mechanisms for Autonomic Dysfunction and Sudden Death in Persons with Epilepsy In an experimental animal model in which the occurrence of cardiac autonomic neural discharges were correlated with cerebral epileptiform discharges, a “lockstep phenomenon” was observed and was deἀned as the occurrence of cardiac sympathetic and vagal cardiac neural discharges intermittently synchronized with epileptogenic discharge (Lathers et al. 1983). The abnormal cardiac neural discharge and cardiac arrhythmias were associated

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

811

with subconvulsant interictal activity, and thus it is hypothesized that lockstep phenomenon may be a factor in the mechanism of unexplained death in persons with epilepsy who exhibited no overt seizure activity at the time of death (Lathers et al. 1983, 1987; Dodd-O and Lathers 1990). In experimental studies, an unstable lockstep phenomenon pattern with time intervals of 10 s or more during which lockstep phenomenon existed but with variable interspike intervals was correlated with a higher mean proportion of time spent in precipitous changes in blood pressure (Stauffer et al. 1989). Stable lockstep phenomenon occurred when interdischarge intervals were constant. With the disappearance of constant interdischarge intervals (stable lockstep phenomenon), precipitous changes in blood pressure and the incidence of ECG changes occurred more frequently (Stauffer et al. 1989). It was suggested that the development of the abnormal rhythmic activity of the unstable lockstep phenomenon may alter neurotransmitter release and initiate autonomic dysfunction, thereby raising the possibility of a contributory role in sudden unexplained death in epileptic persons. At least four mechanisms have been postulated through which lockstep phenomenon may play this role (Stauffer et al. 1989): (1) excessive stimulation of an electrically unstable heart previously damaged; (2) the occurrence of nonuniform postganglionic cardiac sympathetic discharge or an imbalance between the sympathetic and parasympathetic neural discharge to the heart; (3) sinus arrest and bradycardia associated with seizures and induced by the parasympathetic nervous system; and (4) the development of precipitous blood pressure changes.

49.8â•…Some Clinical Pharmacologic Considerations in the Management of Epilepsy 49.8.1â•…Postmortem Toxicologic Findings in Victims of Sudden Unexplained Death in Epilepsy Leestma (1990a) examined cases of sudden death in epileptic persons and listed the medications the individuals were taking at the time of death (Table 49.3). In approximately 23% of the cases, it was not determined whether the victim had ever taken or been prescribed anticonvulsant medication. In the other 77%, a number of combinations of drugs were used, 33% were taking phenytoin and phenobarbital, 25% were taking phenytoin only, Table 49.3â•… Antiepileptic Drugs Taken by SUDEP Victims Drug Phenytoin only Phenobarbital only Phenytoin and phenobarbital Phenytoin, phenobarbital, and other Phenytoin and carbamazepine Valproic acid and other Other combinations No known anticonvulsant

Percentage of Cases (%) 25 3 33 10 2 2 2 23

Source: Leestma, J. E., in Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, Marcel Dekker, New York, NY, 1990c. With permission.

812 Sudden Death in Epilepsy: Forensic and Clinical Issues

and 10% were taking phenytoin, phenobarbital, and other. Analysis of blood samples for ethanol (Chan et al. 1990), phenytoin, phenobarbital, opiates, tranquilizers, and analgesics showed that in 50% of the cases, no detectable levels of phenytoin and/or phenobarbital or any other anticonvulsants were present (Table 49.4). These data suggest extensive patient noncompliance with anticonvulsant therapy and drug interactions, or problems with product formulations. The therapeutic-range blood levels were assumed to be between 9 and 20 μg/ml for phenytoin and between 9 and 25 μg/ml for phenobarbital. Although this maximum therapeutic level of 24 μg/ml is quoted from Laidlaw and Richens (1982), Rall et al. (1990) refer to a therapeutic range of 10–35 μg/ml. In a few cases, primidone levels were detected: alone in the therapeutic or greater range in three victims, in subtherapeutic levels with a subtherapeutic concentration of phenytoin, and in one additional victim only was the level of both drugs considered to be within the therapeutic range. The possibility exists that subtherapeutic anticonvulsant levels, either chronic or recent (as in sudden death “cold turkey” withdrawal), could contribute to seizures, with consequent autonomic instability. The common observation of low or absent antiepileptic drug levels in victims of sudden death emphasizes the possibility that sudden autonomic instability could be the consequence of neuronal hyperexcitability associated with antiepileptic drug withdrawal (see discussion below). Leestma (1990c) noted that approximately half of the victims were taking or had previously taken numerous other drugs. Thirty-three percent were taking benzodiazepines, tricyclic antidepressants, and phenothiazines. Other categories of drugs included diuretics, antihistamines, disulἀram, or narcotic analgesics, but these were being taken by a small percentage of victims (3–5%). A history of substance abuse was conἀrmed in approximately half of the cases, with ethanol the most commonly abused substance (70%). Only three to ἀve of the victims were reported to use intravenous drugs or excessive amounts of caffeinated beverages. Approximately 5% of victims used cannabis. Autopsy analysis revealed that 38% of the victims showed virtually no signiἀcant pathologic abnormality considered to be responsible for sudden death. Conditions found included cerebral edema (32%), pulmonary emphysema cirrhosis or fatty liver (8.5%), chronic pancreatitis (3%), and miscellaneous. The hearts, lungs, and livers of the male victims were signiἀcantly heavier than was expected for these organs. Histologic examination of the heart did not reveal any obvious pathologic changes that would account for death. Histologic examination of the lungs and livers showed passive congestion, and both pulmonary edema and congestion were noted in the lungs. Leestma (1990c) concludes Table 49.4â•… Toxicological Results Obtained from Postmortem Blood Samples in 124€Victims of SUDEP Phenobarbital Levels Phenytoin Levels

None

Subtherapeutic

Therapeutic

None Subtherapeutic Therapeutic

50% 8.7% 7%

13% 9.6% 4.3%

7% 0% 0.9%

Source: Leestma, J. E., in Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder, Marcel Dekker, New York, NY, 1990c. With permission.

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

813

that because, in most cases, no obvious anatomic causes of death were found at autopsy among these patients with epilepsy who died unexpectedly and suddenly, it is reasonable to assume that death most probably involved an asystolic or arrhythmic cardiac event alone or in combination with a circulatory embarrassment, such as shock or acute respiratory failure. 49.8.2â•… The Ideal Pharmacologic Agent The “ideal pharmacologic agent,” which would be an effective antiepileptic drug and would also eliminate the autonomic cardiac neural discharge dysfunction and the associated arrhythmias, has not been discovered. Phenytoin (Gillis et al. 1971; Roberts 1970; Schlosser et al. 1975) and chlordiazepoxide depress cardiac sympathetic neural discharge and also possess antiarrhythmic properties (Gillis et al. 1971, 1972, 1974; Evans and Gillis 1974, 1975; Gillis 1969; Pace and Gillis 1976; Raines et al. 1970). Lathers and Schraeder (1982) hypothesized that the pharmacologic agent that would provide the best protection against the autonomic dysfunction associated with epileptogenic activity is one exhibiting the three properties of anticonvulsant, antiarrhythmic, and cardiac neural depressant activity. In one experimental animal model, phenobarbital depressed mean arterial blood pressure and sympathetic neural discharge but had no effect on the parasympathetic discharge before induction of epileptogenic activity. With the onset of epileptogenic activity, arrhythmias and neural dysfunction occurred in both divisions of the autonomic nervous system. Thus, it was concluded that phenobarbital is not an agent for the prevention of autonomic neural dysfunction and arrhythmias associated with epileptogenic activity in this animal model. It may be that no one drug will work in all persons with epilepsy who are at risk of sudden death, and that the ideal drug or drugs to prevent autonomic neural dysfunction and arrhythmias in a given person will need to be an individualized choice. Jay and Leestma (1981) and Leestma et al. (1984) have emphasized that another risk factor for sudden death may be low antiepileptic drug levels, as most of the victims in their studies had subtherapeutic or no detectable levels postmortem. This raises the possibility that the victims did not comply with treatment, placing them at increased risk of drug-withdrawal seizures. Negative or subtherapeutic drug levels of anticonvulsants have been noted in other reports. Hirsch and Martin (1971) noted this ἀnding at autopsy in more than 19 patients with epilepsy who died suddenly and who had been prescribed anticonvulsants before death. Terrence et al. (1975) also reported absent or subtherapeutic anticonvulsant levels in 37 cases for which autopsy was performed 4 years later. Seventeen had been prescribed phenytoin, but none of the victims had therapeutic levels. Only 3 of 24 victims who had been prescribed phenobarbital had therapeutic levels. In one exception to the above ἀnding, Lascelles et al. (1970) reported that 45% of the victims they studied had therapeutic blood levels. Nevertheless, Terrence et al. (1975) states that subtherapeutic or absent levels at death is most likely due to recent discontinuation or noncompliance with the prescribed drug regimen rather than due to the possibility that suboptimal dosages were consistently prescribed. The possibility of noncompliance ἀts with the clinical information available, that is, that the prime candidate for sudden death is a young, ambulatory patient with seemingly well-controlled seizures. These individuals may impulsively discontinue the anticonvulsant because of undesired side effects or because they think the infrequent occurrence of seizures will accommodate noncompliance.

814 Sudden Death in Epilepsy: Forensic and Clinical Issues

However, one must not overlook the fact that some sudden deaths in epilepsy occur in individuals with therapeutic levels of anticonvulsants (Lascelles et al. 1970; Terrence 1990). Obviously, patients should be warned to not “self-withdraw” anticonvulsants because such an act may be a risk factor for a lethal convulsive episode that is different from the seizures the patients survived in the past (Terrence 1983; Lund and Gormsen 1985). Wannamaker (1990) concludes that, as therapeutic administration and monitoring of antiepileptic drugs continue to improve, the causes of deaths attributable to seizures should diminish. 49.8.3â•… Generic Substitutions of Antiepileptic Drugs When considering selection of a pharmacologic agent for the treatment of an individual with epilepsy, one must consider the question of whether a generic drug should be used (American Academy of Neurology 1993). Although generic antiepileptic medications may offer signiἀcant cost reduction, there is evidence that some of the generic products provide variable biologic equivalence to the brand-name product for which they have been substituted (Nuwer et al. 1990; Andermann et al. 2007; Berg 2007). In addition, because of variation in bioavailability, different generic antiepileptic agents may not adequately substitute for each other or for the brand-name product. The two drugs for which these problems are best recognized are phenytoin and carbamazepine (Nuwer et al. 1990). The three factors most related with risk for nonequivalence include (1) low water solubility; (2) a narrow therapeutic range; and (3) nonlinear pharmacokinetics. All three of these risk factors are possessed by phenytoin, whereas only the ἀrst two are possessed by carbamazepine. The consequences of nonequivalence may include (1) decreased serum drug concentrations responsible for breakthrough seizures and consequent injuries, such as result from falls or motor vehicle accidents, or (2) increased concentrations resulting in toxicity producing discomfort, imbalance, or poor job performance. The bioavailability of a generic drug is compared to a brand-name drug by giving each agent to the same volunteers at different times. Serum collected over time and the serum concentration vs. time is graphed (Figure 49.4) (Nuwer et al. 1990) and depicts the pharmacokinetic values of area under the serum concentration curve (AUC), the maximum serum concentration after administering

Serum concentration

Tmax Cmax

AUC Time

Figure 49.4╇ Graphic presentation of data collected in a typical in vivo bioavailability study. AUC, area under the serum concentration curve; t, time between administration of the drug and attainment of the maximum serum concentration; C, maximum serum concentration after drug administration. (From Nuwer, M. R., et al., Neurology, 40 (11), 1647–1651, 1990. With permission.)

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

815

Table 49.5â•… Pharmacokinetics of Antiepilepsy Drugs Drug Carbamazepine Clonazepam Ethosuximide Phenobarbitala Phenytoinb Primadone Valproic acid

H2O Solubility

Narrow Range

Nonlinear Kinetics

Generic Available

Almost 0 Almost 0 150 ml/ml Very great 0.0014 ml/ml 0.6 ml/ml 1.3 ml/ml

Yes No Yes Yes Yes Yes Yes

No No No No Yes No Yes

Yes No No Yes Yes Yes Yes†

Source: Nuwer, M. R. et al., Neurology, 40 (11), 1647–1651, 1990. With permission. a Brand name phenobarbital (Luminal) is no longer marketed. b Generic formulations of extended-release sodium phenytoin have been withdrawn from the US market.

the drug (Cmax), and the time between giving the drug and attaining the maximum serum concentration (tmax). AUC is used to compare bioavailability of different formulations of the drug. If one dose is given, AUC measures the total amount of drug absorbed (DA) and the mean drug serum concentration over the time interval studied; that is, the mean serum concentration during a single dose study = AUC/time. AUC will also predict mean steadystate serum concentration during chronic administration (Css), since Css = (DA/dosing interval)/clearance (Nuwer et al. 1990). If a drug is easily soluble in water, it can generally be formulated in completely absorbed formulations. For instance, phenobarbital is completely water soluble, and the generic preparations of this agent have long been used with little or no problems. In contrast, a drug that is poorly soluble in water will necessitate the use of different formulations, which may vary in the inert ἀller used, the pressure used to form the tablet or capsule, and so on. These differences will result in clinically signiἀcant differences in dissolution rate and extent, bioavailability, and shelf life. The current Food and Drug Administration generic substitution guidelines provide for a wide range of variability in the bioavailability permitted and may cause problems, especially when drugs with nonlinear pharmacokinetics are used. This is true because a given percentage difference in bioavailability will result in an even larger percentage of difference in mean steady-state serum concentration. Table 49.5 summarizes these three risk factors for the antiepileptic drugs most commonly prescribed. Phenytoin is the pharmacologic agent for which many examples of problems have been reported when patients are switched to generic preparations. This agent possesses all three risk factors listed in the table. Carbamazepine is listed as possessing two of the risk factors. 49.8.4â•… Federal Guidelines for Generic Antiepileptic Drugs Federal guidelines for generic antiepileptic drugs (Food and Drug Administration 1981) to be marketed as “equivalent to brand-name product” require the manufacturer to conduct (1) an in vitro dissolution test comparing generic vs. speciἀed reference material and (2) an in vivo bioavailability study comparing generic vs. reference. Generic is deemed bioequivalent if (1) in vitro dissolution rate proἀles are similar and (2) in vivo results show that generic meets the following speciἀed criteria:

816 Sudden Death in Epilepsy: Forensic and Clinical Issues

• The 90% conἀdence interval for mean value of the generic drug relative to the reference drug falls within a range of 80–120% for (1) active drug ingredient concentration in plasma; (2) peak plasma levels (Cmax); (3) area under the curve (AUC). • In at least 80% of subjects given the generic drug, bioavailability is greater than 80% relative to the reference drug when each subject is used as his or her own comparison (80%/80% rule). • The power of the analysis of variance test to detect a 20% difference is 20% (difference in PK parameters will be detected 80% of the time with P = 0.05). 49.8.5â•… Consequences of Generic In Vivo Criteria If the conἀdence interval for the mean value of generic AUC is small, it is theoretically possible for the average patient to experience an almost increase in serum concentration when switched from a low-bioavailability generic formulation (80% of brand name) to a high-bioavailability (120% of brand name) generic formulation. Conversely, the average patient could have an almost 33% decrease in serum concentration if switched from a high-bioavailability generic formulation to a low-bioavailability generic formulation. Although on average there may be a difference in blood levels when a patient alternates between high and low bioavailability of generic phenytoin, some individuals could have as much as a 40% difference in blood levels and still meet federal guidelines for bioequivalence. 49.8.6â•… Generic Bioequivalence Generic bioequivalence is allowed even if it produces a widely varying bioavailability in some individuals, as long as mean −90% conἀdence interval values for AUC and other PK parameters fall within a range of ±20% compared with the brand name and if the 80%/80% rule is met. Regulations are not a guarantee that each individual will receive the same amount of antiepileptic drug when switching from a brand-name drug to a generic drug or from one generic drug to another. 49.8.7â•…Cases Illustrating Potential Problems Involved with the Use of Generic Antiepileptic Drugs

Case 5 A 35-year-old man with a 10-year history of posttraumatic tonic–clonic seizures, which were well controlled with a 15–1 μg/ml level of Dilantin (Parke Davis, Morris Plains, NJ) on 300 mg of Dilantin once daily, was found to have esophageal cancer. He required a gastric feeding tube and nutritional support with a proprietary parenteral nutrient formula. Because of absorption problems associated with tube feedings, a Dilantin dose of 500 mg once daily was necessary to maintain a trough level of 12 μg/ml and a peak level of 17 μg/ml. He was placed in a nursing facility where generic phenytoin was prescribed. Within 10 days, the patient developed lethargy and diplopia. His peak phenytoin level was 28 μg/ml on the same dose, that is, 500 mg once

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

817

daily. Changing the dose of generic phenytoin to 200 mg in the morning and 300 mg in the evening has lowered his phenytoin level to 15 μg/ml with abatement of his toxic symptoms. This case illustrates that with phenytoin, minor differences in absorption between preparations can result in major changes in phenytoin blood levels consequent to nonlinear pharmacokinetics. In this instance, substitution with the more rapidly absorbed preparation of generic phenytoin resulted in sufficiently increased absorption to produce toxic peak levels of the drug with attendant symptoms (Woodbury 1989). Case 6 The patient, a 32-year-old man, had a history of complex partial and generalized tonic–Â�clonic seizures that were well controlled with a speciἀc generic brand of phenÂ� ytoin 300 mg once daily. A trough level of 16 μg/ml was required to control his seizures. He was admitted to the hospital because of recurrent generalized tonic–clonic convulsions, at which time his serum level was 9 μg/ml. After being stabilized with intravenous phenytoin, he was able to relate that 2 weeks ago his pharmacist had dispensed a capsule form of his antiepileptic drug that looked different than usual. He was reassured that one form of phenytoin was equivalent to the other. With reinstitution of his previous brand of phenytoin, his serum level was maintained at 15 μg/ml on 300 mg daily. He remains seizure-free. This very commonly observed clinical problem is the consequence of the allowable difference in absorption between various preparations of the same drug. This patient had recurrence of seizures because the substituted phenytoin preparation was a form that had less bioavailability. In all probability, the ἀrst form of the drug could have been 40% better absorbed than the substituted capsules and yet been considered bioequivalent (Woodbury 1989). It is important to be aware that problems may occur in management of persons with epilepsy as a result of switching the patient from a generic preparation to the brand-name drug or vice versa, or when one generic drug of different bioavailability is substituted for another generic drug. By consistent use of the same preparation in a given patient, the physician can be reasonably assured of consistent therapeutic antiepileptic drug levels. 49.8.8â•…Drug Interactions One mechanism that places epileptic patients at risk stems from a lack of understanding of the complex drug interactions that can result from polypharmacy. Subtherapeutic levels may result from drug interactions with polypharmacy, as illustrated by the following cases. Case 7 A 29-year-old woman with a history of simple partial seizures with rare secondarily generalized tonic–clonic events was taking carbamazepine 300 mg three times a day, with a trough level of 9.8 μg/ml, to control her seizures. She had a history of an allergic

818 Sudden Death in Epilepsy: Forensic and Clinical Issues

reaction to phenytoin. Recently, she reported having symptoms of ataxia and diplopia when the carbamazepine dose was increased by 100 mg twice daily. After she had a single generalized tonic–clonic seizure, the dose of carbamazepine was reduced back to 300 mg three times daily, and treatment with phenobarbital 120 mg (at bedtime) was initiated. Three weeks later, the patient had a flurry of generalized tonic–clonic seizures and was hospitalized. Levels of phenobarbital and carbamazepine were 14 and 4.5 μg/ml, respectively, on admission. An ECG performed in the emergency room showed runs of ventricular tachycardia concurrent with a seizure. The patient developed severe recurrent seizures with a potentially life-threatening tachyarrhythmia when her carbamazepine level was inadvertently lowered consequent to induction of its metabolism by a modest dose of phenobarbital. The patient was in a situation in which the dose of phenobarbital was insufficient to produce a therapeutic level, but yet was sufficient to induce hepatic metabolism, lowering the carbamazepine to a subtherapeutic level. The reader is referred to Table 49.6 for a summary, to Hanstein (1989) for a comprehensive discussion of common drug interactions, and to Reynard and Smith (1991) for sources of regularly updated computerbased compilations of drug interactions. Another example of a drug interaction that may lead to subtherapeutic levels of an anticonvulsant may occur with the concomitant use of phenytoin and theophylline. This combination has been demonstrated to result in a decreased steady-state serum level of phenytoin. The absorption of phenytoin is thought to be inhibited by theophylline when the two agents are swallowed together (Hanstein 1989). Phenytoin also markedly increases the clearance of theophylline. The clinical relevance of this interaction is that the interaction may occur in half of the patients taking both drugs and may result in the loss of seizure control. If a rapidly absorbed theophylline formulation is taken 2 h before the dose of phenytoin, the effect of the interaction is decreased but not eliminated. One suggestion to circumvent this interaction is to use cromolyn for chronic asthma. If theophylline must be Table 49.6â•… Summary of Interactions of Commonly Used Antiepileptic Drugs First Drug Administered Phenytoin

Phenobarbital

Carbamazepine Valproic acid

Second Drug Administered

Effect of Plasma Levels of First Drug after Giving Second Drug

Carbamazepine Phenobarbital/primidone Valproic acid Carbamazepine Valproic acid Phenytoin Phenytoin Phenobarbital/primidone

↑ 0 ↓ 0 ↑ ↑ ↓ ↓

Carbamazepine Phenytoin Phenobarbital/primidone

↓ ↓ ↓

Source: Brown, T. R., in Epilepsy: Diagnosis and Management, ed. T. R. Brown and R. G. Feldman, Little Brown, Boston, MA, 1983. With permission. Note: ↑, increase; 0, stable; ↓, decrease.

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

819

added on a continuous basis, phenytoin levels should be monitored at 2-week intervals and the dose adjusted accordingly (Taylor et al. 1980). A third example of a more complex drug interaction occurring as a consequence of polypharmacy is illustrated by the following case.

Case 8 A 27-year-old woman with a history of complex partial seizures for the past 14 years was taking primidone 125 mg four times daily, with a therapeutic level of 9.5 μg/ml. She continued to have seizures every 2 or 3 days but was unable to tolerate a higher dose of primidone because of side effects (lethargy and impaired concentration). Treatment with phenytoin, 200 mg in the morning and 100 mg at bedtime, was added with marked improvement in seizure control. However, within a month, the patient became increasingly lethargic. Her physician lowered the dose of phenytoin to 100 mg twice daily and obtained antiepileptic drug levels. Within 4 days the patient again had an increase in seizure frequency. The antiepileptic drug levels detected by the clinical laboratory were as follows: phenytoin, 12 μg/ml; primidone, 8.0 μg/ml; and phenobarbital, 55 μg/ml. The drug interaction in this case demonstrates that the development of side effects after introduction of a new agent does not warrant the assumption that the new agent is the direct cause of the symptoms. In this instance, phenytoin stimulated conversion of primidone to phenobarbital, causing toxic levels to occur after several weeks. The appropriate action should have been to gradually lower the primidone dose to bring its metabolite, phenobarbital, below toxic levels rather than to lower the phenytoin dose and induce seizures because of a subtherapeutic phenytoin level. The frequency of pharmacokinetic interactions of antiepileptic drugs argues in favor of monotherapy as the ideal treatment approach. However, in persons with intractable epilepsy, it is often necessary to use two agents concurrently. Table 49.6 shows interactions of some commonly used antiepileptic drugs. When primidone is the ἀrst drug administered, it is unwise to use a second agent because of the potential for increased metabolic transformation to phenobarbital, with the subsequent risk of toxicity. Phenytoin and carbamazepine, in particular, have such an effect. Valproic acid should not be combined with primidone because of the potential for inhibition of the breakdown of phenobarbital with, again, the risk of barbiturate toxicity. Brown (1983) has presented an excellent discussion of this problem.

49.9â•… Conclusion The clinical pharmacology of antiepileptic drugs in treating epilepsy requires many considerations beyond the basic dose–response relationships. This article is not a comprehensive review of classic clinical pharmacology but rather an attempt to address what we believe to be some important facets of patient management, especially as they might apply to risks of autonomic dysfunction and the maintenance of dependable antiepileptic drug therapeutic levels.

820 Sudden Death in Epilepsy: Forensic and Clinical Issues

The clinical pharmacology of antiepileptic drugs in treating epilepsy requires many considerations beyond the basic dose–response relationships. This article is not a comprehensive review of classic clinical pharmacology but rather an attempt to address what we believe to be some important facets of patient management, especially as they might apply to risks of autonomic dysfunction and the maintenance of dependable antiepileptic drug therapeutic levels.

References Andermann, F., M. S. Duh, A. Gosselin, and P. E. Paradis. 2007. Compulsory generic switching of antiepileptic drugs: High switchback rates to branded compounds compared with other drug classes. Epilepsia 48 (3): 464–469. American Academy of Neurology. 1993. Paper read at the Satellite Symposium on Sudden Unexplained Death in Epilepsy, April 4, 1993, New York. Annegers, J. F., and S. A. Blakley. 1990. Patterns of overall and unexplained death mortality among persons with epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 39–52. New York, NY: Marcel Dekker. Bacon, G. M. 1868. On the modes of death in epilepsy. Lancet 1: 555–556. Berg, M. J. 2007. What’s the problem with generic antiepileptic drugs?: A call to action. Neurology 68 (16): 1245–1246. Brown, T. R. 1983. Pharmacologic principles of antiepileptic drug administration in epilepsy. In Epilepsy: Diagnosis and Management, ed. T. R. Brown and R. G. Feldman. Boston, MA: Little Brown. Chan, A. W. K., C. M. Lathers, and J. E. Leestma. 1990. Alcohol arrhythmias, seizures, and sudden death. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 19, 329–391. New York, NY: Marcel Dekker. Dodd-O, J. H., and C. M. Lathers. 1990. A characterization of the lockstep phenomenon in phenobarbital-pretreated cats (Chapter 13). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 199–220. New York, NY: Marcel Dekker. Epilepsy Foundation of America. 1975. Basic Statistics on the Epilepsies. Philadelphia, PA: F. A. Davis. Evans, D. E., and R. A. Gillis. 1974. Effect of diphenylhydantoin and lidocaine on cardiac arrhythmias induced by hypothalamic stimulation. J Pharmacol Exp Ther 191 (3): 506–517. Evans, D. E., and R. A. Gillis. 1975. Effect of ouabain and its interaction with diphenylhydantoin on cardiac arrhythmias induced by hypothalamic stimulation. J Pharmacol Exp Ther 195 (3): 577–586. Food and Drug Administration. 1981. New standards for phenytoin products. FDA Drug Bull 11 (1): 4. Gastaut, H. 1973. Dictionary of Epilepsy. Geneva: World Health Organization. Gillis, R. A. 1969. Cardiac sympathetic nerve activity: Changes induced by ouabain and propranolol. Science 166 (904): 508–510. Gillis, R. A., J. R. McClellan, T. S. Sauer, and F. G. Standaert. 1971. Depression of cardiac sympathetic nerve activity by diphenylhydantoin. J Pharmacol Exp Ther 179 (3): 599–610. Gillis, R. A., A. Raines, Y. J. Sohn, B. Levitt, and F. G. Standaert. 1972. Neuroexcitatory effects of digitalis and their role in the development of cardiac arrhythmias. J Pharmacol Exp Ther 183 (1): 154–168. Gillis, R. A., H. Thibodeaux, and L. Barr. 1974. Antiarrhythmic properties of chlordiazepoxide. Circulation 49 (2): 272–282. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Hanstein, P. O. 1989. Drug Interactions. Philadelphia, PA: Lea & Febiger.

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

821

Hauser, W. A., and L. T. Kurland. 1975. The epidemiology of epilepsy in Rochester, Minnesota, 1935 through 1967. Epilepsia 16 (1): 1–66. Hirsch, C. S., and D. L. Martin. 1971. Unexpected death in young epileptics. Neurology 21 (7): 682–690. Hoffman, M. S., ed. 1988. The World Almanac and Book of Facts. New York, NY: Ballantine. Howell, S. J., and L. D. Blumhardt. 1990. The role of EEG monitoring in the diagnosis of epilepsyrelated cardiac arrhythmias and of cardiac arrhythmias mimicking epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 7, 101–119. New York, NY: Marcel Dekker. International League Against Epilepsy. 1981. Proposal for revised clinical and electroencephalographic classiἀcation of epileptic seizures. Epilepsia 22: 489–501. James, T. N. 1983. Chance and sudden death. J Am Coll Cardiol 1 (1): 164–183. Jay, G. W., and J. E. Leestma. 1981. Sudden death in epilepsy. A comprehensive review of the literature and proposed mechanisms. Acta Neurol Scand. 63 (Suppl. 82): 1–66. Juul-Jensen, P., and A. Foldspang. 1983. Natural history of epileptic seizures. Epilepsia 24 (3): 297–312. Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. Kuller, L., A. Lilienfeld, and R. Fisher. 1967. An epidemiological study of sudden and unexpected deaths in adults. Medicine 46: 341–361. Kuller, L. H., J. A. Perper, M. Cooper, and R. Fisher. 1974. An epidemic of deaths attributed to fatty liver in Baltimore. Prev Med 3 (1): 61–79. Kurland, L. T. 1959. The incidence and prevalence of convulsive disorders in a small urban community. Epilepsia 1: 143–161. Laidlaw, J., and A. Richens. 1982. A Textbook of Epilepsy, 2nd ed. Edinburgh: Churchill-Livingstone. Lascelles, P. T., R. S. Kocen, and E. H. Reynolds. 1970. The distribution of plasma phenytoin levels in epileptic patients. J Neurol Neurosurg Psychiatry 33 (4): 501–505. Lathers, C. M. 1990. Role of neuropeptides in the production of epileptogenic activity and arrhythmias (Chapter 18). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 309– 327. New York, NY: Marcel Dekker. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1974. Relationship between the effect of ouabain on arrhythmia and interspike intervals (ISI) of cardiac accelerator nerves. Pharmacologist 16: 201. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J Clin Pharmacol 27 (5): 346–356. Lathers, C. M., and P. L. Schraeder. 1995. Experience-based teaching of therapeutics and clinical pharmacology of antiepileptic drugs. Sudden unexplained death in epilepsy: Do antiepileptic drugs have a role? J Clin Pharmacol 35 (6): 573–86; quiz 586–587. Lathers, C. M., and P. L. Schraeder. 1990. Arrhythmias associated with epileptogenic activity elicited by penicillin (Chapter 9). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. p. 309–327. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259.

822 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., N. Tumer, and C. M. Kraras. 1988. The effect of intracerebroventricular d-ALA2 methionine enkephalinamide and naloxone on cardiovascular parameters in the cat. Life Sci 43 (26): 2287–2298. Lathers, C. M., F. L. Weiner, and P. L. Schraeder. 1983. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lock-step phenomenon. Circ Res 31: 630A. Leestma, J. E. 1990a. Natural history of epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, Chapter 1, 1–26. New York, NY: Marcel Dekker. Leestma, J. E. 1990b. Sudden unexpected death associated with seizures: A pathological review (Chapter 5). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder, 61–88. New York, NY: Marcel Dekker. Leestma, J. E., J. F. Annegers, M. J. Brodie, S. Brown, P. Schraeder, D. Siscovick, B. B. Wannamaker, P. S. Tennis, M. A. Cierpial, and N. L. Earl. 1997. Sudden unexplained death in epilepsy: Observations from a large clinical development program. Epilepsia 38 (1): 47–55. Leestma, J. E., J. R. Hughes, S. S. Teas, and M. B. Kalelkar. 1985. Sudden epilepsy deaths and the forensic pathologist. Am J Foren Med Pathol 6: 215–218. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26 (2): 195–203. Lennox, W. G. 1960. Epilepsy and Related Disorders. Boston, MA: Little Brown. Luke, J. L., and Helpern M. 1968. Sudden unexpected death from natural causes in young adults. Arch Pathol 85: 10–17. Lund, A., and H. Gormsen. 1985. The role of antiepileptics in sudden death in epilepsy. Acta Neurol Scand 72 (4): 444–446. Marshall, D. W., B. F. Westmoreland, and F. W. Sharbrough. 1983. Ictal tachycardia during temporal lobe seizures. Mayo Clin Proc 58 (7): 443–446. McLeod, A. A., and D. E. Jewitt. 1978. Role of 24-hour ambulatory electrocardiographic monitoring in a general hospital. Br Med J 1 (6121): 1197–1199. Nuwer, M. R., T. R. Browne, W. E. Dodson, F. E. Dreifuss, J. Engel, Jr., I. E. Leppik, R. H. Mattson, J. Penry, D. M. Treiman, and B. J. Wilder. 1990. Generic substitutions for antiepileptic drugs. Neurology 40 (11): 1647–1651. Pace, D. G., and R. A. Gillis. 1976. Neuroexcitatory effects of digoxin in the cat. J Pharmacol Exp Ther 199 (3): 583–600. Parke-Davis Company. 1958. Epilepsy: Patterns of Disease. Special Report. Chicago, IL. Pickworth, W. B., J. Gerard-Ciminara, and C. M. Lathers. 1990. Stress, arrhythmias, and seizures (Chapter 22). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Raines, A., B. Levitt, F. G. Standaert, and Y. J. Sohn. 1970. The influence of sympathetic nervous activity on the antiarrhythmic efficacy of diphenylhydantoin. Eur J Pharmacol 11 (3): 293–297. Rall, T. W., L. S. Schleifer, A. S. Nies, and R. H. Levy. 1990. Drugs effective in the therapy of the epilepsies. In Goodman and Gillman’s the Pharmacological Basis of Therapeutics, ed. A. Gillman, T. Rall, A. Nies, and P. Taylor, Chapter 19, 436–462. New York, NY: Pergamon. Reynard, A. M., and C. M. Smith. 1991. Information and learning resources in pharmacology. In Textbook of Pharmacology, ed. C. M. Smith and A. M. Reynard. Philadelphia, PA: Saunders. Roberts, J. 1970. The effect of diphenylhydantoin on the response to accelerator nerve stimulation. Proc Soc Exp Biol Med 134 (1): 274–280. Schlosser, W., S. Franco, and E. B. Sigg. 1975. Differential attenuation of somatovisceral and viscerosomatic reflexes by diazepam, phenobarbital and diphenylhydantoin. Neuropharmacology 14 (7): 525–531. Schoenberg, B. S. 1985. Epidemiology of epilepsy. In The Epilepsies, ed. R. J. Porter and P. L. Morselli. London: Butterworths.

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs

823

Schott, G. D., A. A. McLeod, and D. E. Jewitt. 1977. Cardiac arrhythmias that masquerade as epilepsy. Br Med J 1 (6074): 1454–1457. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., and C. M. Lathers. 1989. Paroxysmal autonomic dysfunction, epileptogenic activity and sudden death. Epilepsy Res 3 (1): 55–62. Schraeder, P. L., R. Pontzer, and T. R. Engel. 1983. A case of being scared to death. Arch Intern Med 143 (9): 1793–1794. Schwartz, R. D., and C. M. Lathers. 1990. GABA neurotransmission, epileptogenic activity, and cardiac arrhythmias (Chapter 17). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Spratling, W. P. 1902. The cause and manner of death in epilepsy. Med News 80: 1225–1227. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Taylor, J. W., L. Hendeles, M. Weinberger, L. W. Lyon, R. Wyatt, and S. Riegelman. 1980. The interaction of phenytoin and theophylline. Drug Intell Clin Pharm 14: 638. Terrence, C. F. 1983. Unexpected, unexplained death of epileptic persons: Clinical correlation. Epilepsia 24: 515–516. Terrence, C. F. 1990. Unexpected, unexplained death of epileptic persons: clinical correlation (Chapter 6). In Epilepsy and Sudden Death, ed. C. M. Lather and P. L. Schraeder. New York, NY: Marcel Dekker. Terrence, C. F., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann Neurol 9 (5): 458–464. Terrence, C. F., Jr., H. M. Wisotzkey, and J. A. Perper. 1975. Unexpected, unexplained death in epileptic patients. Neurology 25 (6): 594–598. Therapeutics and technology subcommittee of the American Academy of Neurology. 1990. Assessment: Generic substitution for anti-epileptic medication. Neurology 40 (1641): 1643. Wannamaker, B. B. 1990. Perspectives on death of persons with epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Woodbury, O. M. 1989. Phenytoin: Absorption, distribution and excretion. In Antiepileptic Drugs, ed. R. Levy, R. Mattson, B. S. Meldrum, J. K. Penry, and F. F. Dreifus. New York, NY: Raven Press. Zielinski, J. J. 1982. Epidemiology. In A Textbook of Epilepsy, ed. J. Laidlaw and A. Richens. Edinburgh: Churchill Livingstone.

824 Sudden Death in Epilepsy: Forensic and Clinical Issues

Appendix I. Self-Assessment Quiz Select the best answer. 1. The incidence of epilepsy in the general United States population is: 1. 30–54/100,000 2. 100–154/100,000 3. 200–354/100,000 4. 450–500/100,000 [More than one answer may be correct for questions 2–9] 2. Which of the following fact(s) about epilepsy are correct? 1. The likelihood of developing epilepsy is not age related. 2. The prevalence rate of those with epilepsy living in the major industrialized nations of the western hemisphere ranges between 5.7-7.8/1000. 3. 10% of those with epilepsy will exhibit more than one type of seizure. 4. The deἀnition of epilepsy excludes a single or occasional seizure triggered by trauma, high fever, syncopal attack. 3. Sudden unexpected death in the general population 1. is related to anatomic causes in the heart and/or aorta as found in 6.1% of 20,981 autopsies. 2. accounts for between 450,000 and 500,000 deaths per year in the United States. 3. accounts for at least 20% of all deaths per year in the United States. 4. is not related to neurally mediated abnormalities of cardiac or vascular function. 4. Sudden unexpected death in the epileptic population: 1. annual incidence of 1/500–1000. 2. risk is 1/1000 in those with idiopathic epilepsy. 3. risk is 1/100 in those with remote symptomatic epilepsy. 4. risk is 1/10,000 in those with idiopathic epilepsy. 5. Characteristics of epileptics dying in a sudden, unexplained manner include (based on Leestma 1990b): 1. frequency of seizures of 3–10 per year. 2. greatest number of cases in the 31- to 40-year-old age group. 3. most likely to be found dead in bed. 4. greater prevalence in black males than in white males. 6. Antiepileptic medication(s) detected in epileptic patients dying of sudden death included (based on Leestma 1990b) are: 1. phenytoin and phenobarbital (33%) 2. phenytoin (5%) 3. phenytoin, phenobarbital, and others (10%) 4. subtherapeutic levels of anticonvulsant(s)

Teaching of Therapeutics and Clinical Pharmacology of Antiepileptic Drugs



825

7. Non-antiepileptic medications detected in epileptic patients dying of sudden death included (based on Leestma 1990b) are: 1. diuretics (2–5%) 2. ethanol (70%) 3. benzodiazepines 4. phenothiazines (80%)

8. Contributing factors in the pathogenesis of cardiac sudden death include: 1. changes in autonomic neural control of the heart 2. changes in platelet aggregation 3. cardiac arrhythmias 4. electrical stability of the heart 9. Postulated mechanisms for sudden death in epilepsy include: 1. development of precipitous blood pressure changes. 2. imbalance between sympathetic and parasympathetic neural discharge to the heart. 3. excessive stimulation of an electrically unstable heart previously damaged. 4. sinus arrest and bradycardia associated with epileptogenic activity. Select the one best answer for questions 10–12. 10. A 37-year-old woman taking carbamazepine 300 mg four times daily for complex seizures maintained a level of 8.8 μg/ml with only modest seizure control. Phenobarbital 120 mg hs was added. Within 2 weeks her seizure frequency increased. Antiepileptic drug levels were ordered. The expected result would be which of the following? 1. phenobarbital 55 μg/ml; carbamazepine 13.2 2. phenobarbital 15 μg/ml; carbamazepine 4.1 3. phenobarbital 15 μg/ml; carbamazepine 13.2 11. Generic forms of drugs are said to be equivalent if the bioavailability is what percentage of the brand-name product? 1. 100% 2. 90–110% 3. 85–115% 4. 80–120% 12. The substitution of one generic form of phenytoin for another in a patient with a stable serum level of 15 μg/ml can result in which of the following levels? 1. 15 μg/ml +/− same equivalent 2. 9 μg/ml +/− minus 40% 3. 25 μg/ml − saturation kinetics + 66% 4. 18 μg/ml − ἀrst order kinetics + 20% 5. All of the above.

826 Sudden Death in Epilepsy: Forensic and Clinical Issues

ANSWERS

1. 1. 2. 2, 4. 3. 1, 2, 3. 4. 1, 2, 3. 5. 1, 2, 3, 4. 6. 1, 2, 3, 4. 7. 1, 2, 3. 8. 1, 2, 3. 9. 1, 2, 3, 4. 10. 2. 11. 4. 12. 5.

Clinical Pharmacology of Antiepileptic Drug Use Clinical Pearls about the Perils of Patty

50

Paul L. Schraeder Claire M. Lathers

Contents 50.1 50.2 50.3 50.4

Introduction Episode I: Correct Answers for the Pretest Episode I Questions for Episode I 50.4.1 Discussion Notes for Questions in Episode I 50.5 Episode II: Correct Answers to the Pretest 50.6 Episode II 50.7 Questions for Episode II 50.8 Answers to the Questions for Episode II 50.9 Episode III: Correct Answers to the Pretest 50.10 Episode III 50.11 Questions for Episode III 50.11.1 Discussion Notes for Questions in Episode III 50.12 Episode IV: Correct Answers to the Pretest 50.13 Episode IV 50.14 Questions for Episode IV 50.14.1 Discussion Notes for Questions in Episode IV 50.15 Posttest Questions 50.16 Correct Answers to the Posttest References

829 831 832 832 832 833 833 834 834 835 836 837 837 838 839 839 840 840 842 844

In September 1992, the American College of Clinical Pharmacology sponsored the ἀrst annual teaching clinic in clinical pharmacology immediately after their annual meeting in Washington, DC. Claire M. Lathers, Hugh Burford, and Cedric M. Smith served as the faculty for the teaching clinic. The ἀnal product of the teaching clinic was four clinical pharmacology problem solving (CPPS) units. These CPPS units are case histories illustrating pertinent principles in clinical pharmacology written to be used as teaching and learning tools for those interested in clinical pharmacology. There is a strong emphasis on the interaction of the clinical problem and implications for rational selection of pharmacologic agents. Of the CPPS units written to date, three have been designed for use by fourth-year medical students, residents, or fellows in clinical pharmacology. Those taking the board 827

828 Sudden Death in Epilepsy: Forensic and Clinical Issues

examination in clinical pharmacology will ἀnd them useful while reviewing different topics in the ἀeld of clinical pharmacology. One of the CPPS units has been written for use by second-year medical students. This teaching exercise has bearing on SUDEP prevention as it is an attempt to address the complexity of antiepileptic drug therapeutics when it comes to maximizing seizure control. This prototypical learning tool has the goal of helping current and future health care professionals to avoid pitfalls in clinical pharmacology and therapeutics as apply to the optimal management of epilepsy, since the achievement of optimal seizure control is one of the most widely acknowledged measures that are thought to reduce the risk SUDEP. Each CPPS unit consists of congruent learning objectives, a pretest, four clinical scenarios that have been written for a speciἀc area in clinical pharmacology, and a posttest. The units may be used by one person for self-assessment or by a group of four students as an academic exercise in which information is exchanged among those working in this small conference-group format. Thus, students may take the CPPS unit and work through the material by themselves, using the pretest and posttest questions as a means of evaluating their level of knowledge in this topic. Alternatively, each CPPS unit may also be used by a group, using the learning format ἀrst established by the patient-oriented problem-solving (POPS) system in pharmacology, designed for laboratory sessions for second-year medical students enrolled in the Medical Pharmacology course. It generally requires a 2- or 3-hour period for a group to discuss the material in the CPPS units. When the CPPS unit is to be used by a group, the members should ἀrst read the learning objectives and take the pretest the night before the session. The four divisions of each CPPS unit should then be divided among those in the group. Each member will therefore be responsible for one of the four clinical scenarios and one-quarter of the answers to the questions in the pretest. Students must then go to the library and read any pertinent references, not limited to those included in the reference list cited within the CPPS unit (Evans et al. 1966; Dansky et al. 1992; Finnell et al. 1992; Rall et al. 1990; Hansten 1979; Lindhout 1992; Mikati et al. 1992; Theodore 1992; Yerby et al. 1992). The next day, under supervision of a faculty preceptor, the students must share the reference information that they have learned individually, the data in their particular episode, and the answers to the questions in the pretest that pertain to the assigned episode. This interaction should be in the manner of providing a consultation to the other members of the group to allow resolution of the clinical questions raised in the clinical scenarios. Thus, each person will present one-quarter of the clinical scenario in the CPPS unit to the group. Questions pertinent to this portion of the CPPS unit will be addressed by the entire group. On completion of the group discussion of all four sections of the CPPS unit, members of the group should answer the questions in the past test separately and by themselves, without referring to the material in the CPPS unit. The faculty member is also present to serve in the manner of a consultant and may wish to recommend additional textbooks, original or review journal articles, and other pertinent materials relevant to the topic being discussed. The faculty of the First Teaching Clinic in Clinical Pharmacology believe that the CPPS units make a very nice addition to the education of those interested in clinical pharmacology and complement the educational and teaching interests of the American College of Clinical Pharmacology and its official publication, the Journal of Clinical Pharmacology. We are pleased to present here the ἀrst CPPS unit to be completed (Schraeder and Lathers

Clinical Pharmacology of Antiepileptic Drug Use

829

1995). This CPPS unit is intended for use by fourth- or ἀfth-year pharmacy students; thirdand fourth-year medical students doing rotations in neurology or clinical pharmacology; and house staff teaching residents in neurology, pediatrics, internal medicine, or family practice or fellows in clinical pharmacology. Those taking the board examination in clinical pharmacology may also wish to use this CPPS unit for review and for self-assessment.

50.1â•…Introduction The prevalence of epilepsy is 1–2% of the general population in the United States, with a higher percentage experiencing one or more symptomatic seizures consequent to speciἀc circumstances, e.g., sleep deprivation, medications, and alcohol withdrawal. In contrast to circumstantial events, epilepsy is deἀned as the spontaneous recurrence of seizures. Although the clinical manifestations of seizures are as varied as the regional functions of the brain, there are two basic pathophysiologic mechanisms explaining the initiation and spread of seizures. Primary generalized seizures occur without any deἀnable cortical focus of origin, with almost instantaneous spread throughout the entire cerebrum. Partial seizures originate in the cerebral cortex and may remain localized or may spread to contiguous and/or distal areas of the brain. In general, antiepileptic drugs are categorized by their efficacy in controlling seizures within these two major classiἀcations. Antiepileptic drugs are used by persons with epilepsy on a long-term basis. Physicians must therefore be aware of the clinical pharmacologic risks associated with use of these medications, such as the potential for short- and long-term side effects, drug interactions, and teratogenicity. Physicians in training must also consider that the principles of antiepileptic drug use in epilepsy are applicable to many other categories of drugs used for treating a wide variety of chronic disorders. The relevance of these principles to the subject of this book is obvious when one considers that optimizing seizure control is one of the most important variables in SUDEP prevention. When you have completed this CPPS workbook, you should understand the following: (1) clinical therapeutics of antiepileptic drug use in children, adults, and pregnancy; (2)€management of febrile convulsions; (3) bioavailability implications of federal guidelines for generic antiepileptic drugs; (4) teratogenic risks associated with the use of antiepileptic drugs; (5) toxicity of long-term administration of antiepileptic drugs; and (6) potential drug interactions associated with the use of antiepileptic drugs. Selected general references are provided for use by students and faculty if additional information is required. Instructions. Please mark your answers to the following questions on this examination to facilitate later discussion and review. Roman numerals designate each separate student. 1I. Simple febrile convulsions: A. May occur in a normally developing child younger than 3 years old B. Have a focal onset of the seizure C. Are associated with persistent electroencephalographic (EEG) abnormalities D. Require long-term antiepileptic drug use E. Are not an inherited trait 2I. Which of the following is the primary disadvantage of using phenobarbital as an antiepileptic drug?

830 Sudden Death in Epilepsy: Forensic and Clinical Issues

A. Changes in physical appearance, i.e., excessive hair growth, gum hyperplasia, and coarsening of facial features B. Unpredictability of dosage of the liquid form C. Behavioral problems and learning disability D. Variable absorption from the gut or intestinal tract during infancy and early childhood E. Lack of a parenteral form 3II. Antiepileptic drug half-lives (t1/2) vary widely. If the t1/2 of phenytoin is approximately 30 hours in a given patient, the predicted time to reach a level of less than 2 μg/mL from a level of 15 μg/mL is: A. 45 hours B. 120 hours C. 90 hours D. 60 hours E. 150 hours 4II. Therapeutic levels of antiepileptic drugs must be maintained to ensure maximal protection against seizures. All but one of the following are major factors in maintaining such levels: A. Patient reliability in taking medication (compliance) B. Brand substitution C. Drug interactions D. Seizure type E. Drug dosage 5II. Complex partial (psychomotor) seizures are the most common type of focal epilepsy. All but one of the following antiepileptic drugs are useful in treating this disorder: A. Primidone B. Ethosuximide C. Phenytoin D. Carbamazepine E. Phenobarbital 6III. Which of the following statements about anticonvulsant drug interactions has not been documented? A. Concomitant use of phenytoin and isoniazid may result in phenytoin toxicity B. Concomitant use of phenobarbital and phenytoin lowers the phenytoin level C. Folic acid levels are lowered by phenytoin use D. Valproic acid displaces phenytoin from plasma protein binding E. Phenytoin increases the effectiveness of corticosteroid effect 7III. Known teratogenic effects of anticonvulsants include all of the following except: A. Short stature B. Neural tube defect (spina biἀda) C. Urogenital anomalies D. Ventricular septal defect E. Cleft lip and/or palate 8III. General considerations when prescribing an antiepileptic drug in someone anticipating pregnancy include all of the following except:

Clinical Pharmacology of Antiepileptic Drug Use

831

A. Monotherapy, i.e., using only one antiepileptic drug B. Using two antiepileptic drugs to take advantage of a beneἀcial drug interaction(s) C. Maintaining the lowest effective therapeutic serum level D. Using no antiepileptic drug E. Avoiding valproic acid 9IV. All of the following may be chronic toxicities associated with the use of the indicated antiepileptic drugs except: A. Phenobarbital: central nervous system sedation B. Phenytoin: osteoporosis C. Valproic acid: gastric irritability D. Carbamazepine: leukopenia E. Ethosuximide: paradoxical hyperactivity 10IV. Complications of long-term phenytoin use include all of the following except: A. Cerebellar ataxia and atrophy B. Peripheral neuropathy C. Low folic acid levels D. Low vitamin B12 levels E. Gingival hyperplasia

50.2â•… Episode I: Correct Answers for the Pretest You have the answers to some of the 10 pretest questions, and other members of your group have the remainder. This is designed to encourage a “consult and discussion” situation. 1I. A is correct. By deἀnition, an uncomplicated or simple febrile seizure occurs in a neurologically normal child. Often, there is a family history of other close relatives having febrile seizures as children. If any focal component is observed, i.e., a unilateral onset or a focal seizure, then underlying cerebral pathologic conditions are likely, making the event a complicated febrile seizure. Although the EEG may show some generalized slowing for a brief time after the seizure, a persistent abnormality is not expected. Finally, antiepileptic drugs are not used in treating simple febrile seizures but are indicated in managing complicated febrile seizures. 2I. The answer is C. The main risk in using phenobarbital in young children is the occurrence of behavioral problems such as hyperactivity and the known association, but not necessarily predictable occurrence, of learning disabilities in school. Both phenobarbital and phenytoin are available in parenteral forms, but carbamazepine and valproic acid are not. The availability of parenteral medication is important in treating prolonged recurrent seizures, such as status epilepticus. Phenobarbital elixir provides a very predictable dose per volume, in contrast to phenÂ� ytoin, which as a suspension is notoriously unpredictable in its dose per volume. The unpredictability of phenytoin absorption from the gut in very young children militates against its use. Cosmetic side effects are most common with the use of phenytoin, a drug that should be avoided in young women because of the risk of permanent facial hair growth and coarsened facial features. Gum hyperplasia is a reversible side effect of phenytoin.

832 Sudden Death in Epilepsy: Forensic and Clinical Issues

50.3â•… Episode I Patty Generica Preggo was born in 1957. Her health was normal until the age of 3 years when she had her ἀrst generalized tonic–clonic seizure in association with a high fever from Roseola. The seizure started in her left arm and progressed to a generalized event, lasting almost 0.5 hour. The patient was hospitalized and a lumbar puncture was performed, which showed normal cerebrospinal fluid pressure and spinal fluid. The seizure recurred twice in the next 24 hours. The patient was noted to have left arm weakness for 6 hours after the last seizure. By the second day, she was afebrile and has normal clinical ἀndings. She was discharged with instructions to her parents to use 30 mg phenobarbital elixir twice daily at the onset of any febrile illness, to be used only for the duration of the fever. Six weeks later, the patient had another febrile illness and experienced three more seizures similar to those described above. The patient was started on a 5 mg/kg oral daily dose of phenobarbital and sent home on this drug regimen.

50.4â•… Questions for Episode I 1.1. The ἀrst issue is the occurrence of a febrile convulsion. The primary question is whether this is a simple or a complex febrile seizure and whether antiepileptic drug use is necessary. 1.2. This episode raises questions of the efficacy of intermittent phenobarbital. A. What is the t1/2 of phenobarbital? B. Is intermittent therapy appropriate? 1.3. What is the standard dose of phenobarbital? 1.4. Should pediatric doses be prescribed on a “mg/kg” basis? 50.4.1â•…Discussion Notes for Questions in Episode I 1.1. It is common practice not to use antiepileptic drugs after a single uncomplicated febrile convulsion. However, points to be considered when treating this child include more than one seizure occurring within a 24-hour period, and whether there was a focal onset with Todd’s (postictal) paralysis afterward. This sequence tells us that there is a high probability of a focal problem in the cerebrum, making this a complicated rather than a simple febrile seizure. 1.2. The t1/2 of phenobarbital is 80–120 hours. Intermittent therapy is not appropriate because the achievement of therapy requires 5× t1/2, and this would be too long a time period for intermittent treatment to be effective. 1.3. The standard dose of phenobarbital is 3 to 6 mg/kg orally per day. The average dose is 5 mg/kg. With febrile illnesses, it may sometimes be appropriate to use intermittent rectal diazepam and the parenteral solution in patients with uncomplicated febrile convulsions. In this patient, it is appropriate to use long-term phenobarbital therapy, as the sequence of events followed that of complicated febrile seizures. 1.4. Note that although phenobarbital is usually prescribed in a mg/kg dosage, for children it is more theoretically appropriate to be prescribed on a “mg/m2” basis with the use of a nomogram.

Clinical Pharmacology of Antiepileptic Drug Use

833

50.5â•… Episode II: Correct Answers to the Pretest 3II. C is correct. The predicted t1/2 level will be 7.5 μg/mL at 30 hours, 3.75 μL/mL at 60 hours, and less than 2 μg/mL at 90 hours. However, if the phenytoin level were in the toxic range (30 μg/mL), zero-order saturation kinetics, which are unique for this antiepileptic drug, would be operative. In this situation, it would take many (often unpredictable) hours (more than 30) to reach half of the initial t1/2 blood level, i.e., 15 μg/mL. Once the therapeutic range (10–20 μg/mL) was reached, predictable (ἀrst-order) kinetics would take hold, and the above calculated decremental rate of change would occur. 4II. D is correct. The type of seizure a patient has is not a factor in maintaining a therapeutic antiepileptic drug level. Although it is true that complex partial seizures are more difficult to control than generalized tonic–clonic seizures or primarily generalized absence, this seizure type does not affect the drug level because patient compliance with therapeutic recommendations is probably the most important factor in seizure control. Every effort should be made to have patients understand the importance of maintaining a stable therapeutic serum level of the antiepileptic drug. Brand substitution is to be avoided because of the variability in absorption between different preparations of the same compound. Use of two or more antiepileptic drugs should be avoided because drug interaction with effects on liver metabolism or protein binding could result in either subtherapeutic or toxic levels. Finally, as long as a drug appropriate for the seizure type is being prescribed, it is very important to achieve the highest tolerated drug level in an individual patient before thinking of substituting another drug if seizure control is not achieved. 5II. B is correct. Ethosuximide is an agent used to treat primarily generalized absence (petit mal). It has little use in treating partial seizures. Although carbamazepine and phenÂ� ytoin are the two most widely used antiepileptic drugs for treatment of complex partial seizures, one must not forget that primidone and phenobarbital are also effective agents, with their sedative side effects being a more important limiting factor in their use than therapeutic efficacy. That sedative side effects are common but not universal and depend on individual susceptibility should also be kept in mind. Bias against one drug or the other should not interfere with trying all available potentially effective and relatively safe agents in persons with intractable seizures.

50.6â•… Episode II Over the next 7 years, Patty had only an occasional seizure, consisting of a staring spell with lip smacking. One July when she was attending camp (at 12 years of age), she started menstruating and subsequently had a complex partial seizure lasting 18 minutes. After it was determined that the phenobarbital blood level was 13 μg/mL, her dose was increased by 60 mg/day to a 120 mg once daily regimen. When Patty returned to school in September, her academic performance plummeted. Her phenobarbital level was 24 μg/mL. Her pediatrician decided to change from a 120 mg once a day regimen to a 30 mg four times daily regimen, with no improvement in performance. In October, Patty had three more complex partial seizures. Her physician added 300 mg phenytoin once a day. She was seizure free for the next 3 months; thus, treatment with phenobarbital was abruptly discontinued. She had another seizure 2 weeks later. The physician continued prescribing phenytoin, maintaining a blood level of 10 μg/mL.

834 Sudden Death in Epilepsy: Forensic and Clinical Issues

The patient’s insurance carrier, a health maintenance organization (HMO), required her to change physicians early the next year. Thereafter, she started having recurrent complex partial seizures. The new physician had prescribed generic phenytoin but noted that serum concentration was only 6 μg/mL. The physician suspected Patty of not complying with her treatment. At their insistence, the family was allowed to return to the original physician, who prescribed the original brand of phenytoin. Patty’s subsequent phenytoin level was 11 μg/mL, and she became seizure free.

50.7â•… Questions for Episode II 2.1. Patty had recurrent seizures at age 12 years. What are likely explanations for this breakthrough? 2.2. The change in Patty’s school performance was quite noticeable. What are common dose-related side effects of phenobarbital speciἀcally and other antiepileptic drugs in general? 2.3. Was the change in phenobarbital dosage from once daily to four times a day reasonable? What is the effect of the t1/2 on dosage frequency? 2.4. Patty had a seizure shortly after initiation of treatment with phenytoin. Would it have been appropriate to add another antiepileptic drug? 2.5. While following a stable dose regimen of phenytoin, Patty had another seizure and was found to have a low phenytoin level. What are the possible explanations for this change?

50.8â•… Answers to the Questions for Episode II 2.1. The most common explanation is that the patient forgot to take her phenobarbital, resulting in withdrawal seizures. Another possibility was that she had a growth spurt and needed a larger dose of phenobarbital. While at camp, she began menstruating. Seizures commonly occur in women near or during menstruation, i.e., the so-called catamenial epilepsy. In such individuals, 250 mg acetazolamide may be used twice daily or three times a day for several days before and during menses. Alternatively, she may have been staying up late around the campἀre, drinking excessive amounts of soft drinks containing caffeine, and not getting sufficient hours of sleep. It is important to solicit a history of militating circumstances to explain recurrence of seizures. 2.2. Phenobarbital is a central nervous system depressant that often alters mental performance at therapeutic levels. Other antiepileptic drugs may also have depressant effects. Common side effects related to high therapeutic levels of most antiepileptic drugs include incoordination and visual symptoms. 2.3. Changing the dosage schedule of phenobarbital will not affect the daytime level. The t1/2 of phenobarbital is so long that going to a four times a day dosage regimen will have no effect. In contrast, a drug with a shorter t1/2, such as carbamazepine, may require a frequent dosage schedule to avoid peak-level side effects resulting from high doses taken at long intervals, e.g., twice daily or three times a day.

Clinical Pharmacology of Antiepileptic Drug Use

835

2.4. Because a seizure occurred when the patient took phenytoin alone, the temptation would be to restart the phenobarbital and discontinue the phenytoin on the presumption that the phenytoin was ineffective as monotherapy. In reality, Patty had a phenobarbital withdrawal seizure because the drug was abruptly stopped without gradual downward titration. A seizure at 2 weeks was predictable because of the long (2 weeks) washout time for phenobarbital (t1/2 = 100 hours). Thus, the plasma level would be one-half of the therapeutic level at 100 hours, one-quarter of the therapeutic level at 200 hours, and one-eighth of the therapeutic level at 300 hours (i.e., 12, 6, and 3 μg/mL, respectively). 2.5. Federal guidelines for generic antiepileptic drugs require that the generic agent falls within a range of 80–120% for an active innovator product concentration in plasma, peak plasma levels (Cmax), and area under the curve (AUC). In at least 80% of subjects given the generic drug, bioavailability is more than 80% relative to the reference drug when each subject is used as his or her own comparison (80%/80% rule). In other words, the power of the analysis of variance test to detect a 20% difference is ~0.80 (20% difference in pharmacokinetic parameters will be detected 80% of the time with p = 0.05). The consequences of generic in vivo criteria include the fact that if the conἀdence interval for mean value of generic AUC is small, it is theoretically possible for the average patient to experience an almost 50% increase in serum concentration when switched from a low-bioavailability generic formulation (80% of brand name) to a high-bioavailability (120% of brand name) generic formulation. Conversely, the average patient could have an almost 33% decrease in serum concentration if switched from a high- to a low-Â�bioavailability generic formulation. Although, on average, there may be a 33% difference in blood levels when a patient alternates between high and low bioavailability of generic phenytoin, some individuals could have as much as a 40% difference in blood levels and still meet federal guidelines for bioequivalence. Although Patty could have been overdosed or underdosed, in this case she was underdosed with the generic preparation of phenytoin. Note that laws and healthcare plans for generic substitution vary from state to state. Requirements of HMOs and other healthcare plans will vary. Patients may receive one brand from a hospital formulary and then be given different brands by their pharmacists after discharge.

50.9â•… Episode III: Correct Answers to the Pretest 6III. The correct answer is E. Phenytoin will decrease the effectiveness of steroids presumably by increasing the metabolic breakdown of the steroids by the liver. This is particularly problematic in women who take anovulatory steroids, putting them at risk for unplanned pregnancies. Other drug interactions that can have signiἀcant effects on patients include the risk of toxic phenytoin levels when this drug is used with isoniazid due to competition with the metabolic breakdown of phenytoin. The opposite problem occurs when phenytoin and phenobarbital are concomitantly used, in that phenobarbital enhances the metabolic breakdown of Dilantin (Parke-Davis, Morris Plains, NJ) and many other drugs. Although lowered folic acid levels are commonly associated with use of Dilantin, the

836 Sudden Death in Epilepsy: Forensic and Clinical Issues

clinical signiἀcance is unclear. Finally, the displacement of protein-bound phenytoin by valproic acid is a potential problem; however, this is compensated for by a decrease in the total phenytoin level due to metabolic effects on phenytoin breakdown by valproic acid. 7III. The correct answer is A. No anticonvulsant drugs have been shown to cause short stature. However, the use of carbamazepine is associated with a 0.5–1.5% risk of pregnancies resulting in spina biἀda. Valproic acid also is associated with a 1.5–2.5% risk of pregnancies resulting in spina biἀda. Although rare, urogenital and gastrointestinal anomalies and ventricular septal defect have been associated with the use of various antiepileptic drugs. Whereas cleft lip and palate were once thought to be associated with the exclusive use of phenytoin, it is now known that these anomalies can also occur rarely with the use of the other anticonvulsants. 8III. The correct answer is B. Monotherapy is the safest approach to treating epilepsy in pregnant women. The risk of teratogenicity is increased with the use of multiple antiepileptic drugs and with maintenance of high therapeutic or toxic serum levels. Although the mechanism of teratogenicity is unclear, the increased metabolic breakdown products of the high levels and multiple drugs may result in increased concentration of potentially toxic metabolites, i.e., oxides, which are presumably teratogenic early in pregnancy. The conventional wisdom of the past militated against the use of phenytoin during pregnancy; however, recent studies suggest that phenytoin is probably no more hazardous than other drugs. More recently, a risk of neural tube defects in association with the use of valproic acid and carbamazepine has been recognized. Recent recommendations for prevention of these anomalies include the use of supplemental folic acid before and during pregnancy.

50.10â•… Episode III During the next 8 years, Patty had approximately one seizure per year while continuing phenytoin therapy, 100 mg three times a day. At the age of 21 years, Patty married her high school sweetheart. Because she and her husband were planning to have a family, she visited an obstetrician. On learning that Patty was taking phenytoin, the obstetrician advised her to change to valproic acid (25 mg/kg) and to discontinue phenytoin. One week after discontinuing phenytoin, she had a generalized tonic–clonic seizure. Her valproic acid level was found to be 51 μg/mL (therapeutic range, 50–100 μg/mL). Assuming that the recurrence of seizures was not explained by subtherapeutic valproic acid levels, he added phenobarbital at a dosage of 60 mg twice daily. Subsequent blood levels were 55 μg/mL for valproic acid and 38 μg/mL for phenobarbital. The seizures recurred, so 200 mg carbamazepine four times a day was added to the other two drugs. Patty then became pregnant. At the end of the ἀrst trimester, she returned to her obstetrician complaining of some gait ataxia. Her phenobarbital level was now 65 μg/mL, so the dose was decreased to 60 mg daily. Her symptoms improved, and 6 months later Patty gave birth to a 7-lb boy with mild spina biἀda. Fortunately, no neurologic deἀcits were found, and the defect was treated successfully by surgery. Patty continued taking phenytoin alone and remained seizure free for 1 year at a level of 11 μg/mL. Although Patty was taking birth control pills, she nevertheless became pregnant again. Her obstetrician told her that she must not have been taking her birth control medication. At the end of the ἀrst trimester, Patty was found to have a phenytoin level of 8 μg/mL, yet remained seizure free. Because of this subtherapeutic level, her physician increased her phenytoin dose to 200 mg twice daily.

Clinical Pharmacology of Antiepileptic Drug Use

837

She then reported having difficulty with balance, which persisted. Her total phenytoin level was 15 μg/mL. She delivered a healthy 8-lb girl at term.

50.11â•… Questions for Episode III 3.1. The patient was given 25 mg/kg valproic acid, resulting in a level of 51 μg/mL. At that point, a second drug, phenobarbital, was added because her seizures continued to occur. Was this an appropriate therapeutic sequence? 3.2. What is the likely mechanism of the seizure occurrence at this time? Was the addition of carbamazepine an appropriate response to the seizure? 3.3. Despite taking birth control pills, why did Patty get pregnant again? 3.4. During her second pregnancy, Patty was taking phenytoin alone. Because her phenytoin was found to be mildly subtherapeutic at 8 μg/mL (therapeutic range, 10–20 μg/mL), dosage was increased from 300 to 400 mg/day. Was this increase necessary? Patty’s total phenytoin level was 15 μg/mL when she was taking 400€mg/day. She developed problems with her balance at this moderate therapeutic level. Why? 50.11.1â•…Discussion Notes for Questions in Episode III 3.1. Valproic acid in adults is prescribed in incremental mg/kg doses ranging from 20 to 60 mg/kg (Figure 50.1). The therapeutic range is quite wide (50–150 μg/mL). It would have been appropriate to increase the dose to achieve a higher therapeutic level before adding a second drug. However, since Patty anticipated becoming pregnant, valÂ�proic acid could have been discontinued because of the risk of neural tube defect and another drug, such as phenobarbital, substituted to reduce the risk to the fetus. 3.2. The physician interpreted the seizure as the consequence of valproic acid being ineffective as monotherapy. The more likely cause of the seizure was the abrupt withdrawal of phenytoin. His response, i.e., adding a second drug (carbamazepine) to the valproic acid, was not appropriate, as concurrent use of two or more antiepileptic drugs is associated with a risk of fetal neural tube defect. Although both of these drugs have been known to produce spina biἀda, this problem is more likely to be associated with the use of valproic acid. In a postulated model of enhanced teratogenicity of speciἀc antiepileptic drug combinations, induction of monoxygenase by carbamazepine increases epoxide concentrations while valproic acid inhibits epoxide hydrolase. The result is increased intercellular concentrations of cytotoxic epoxides. 3.3. The pregnancy occurred because phenytoin increases the rate of metabolic breakdown of steroids, and thus decreased the effectiveness of the birth control pills. 3.4. Although during pregnancy, the total phenytoin level may drop below the therapeutic range, the proportion of free phenytoin will increase secondary to diminished protein binding. Thus, despite a decrease in the total phenytoin level, therapeutic efficacy is often maintained. The seemingly innocuous increase to 15 μg/mL in the circumstances of decreased protein binding would result in clinical toxicity from a high free phenytoin level. Total and free phenytoin levels should therefore be monitored regularly during pregnancy.

838 Sudden Death in Epilepsy: Forensic and Clinical Issues Percent adult dose

Surface area m2 2.0 Pounds to sq. meter (m2)

105

1.8

100 1.6

94 88

1.4

82 76 71

1.2

65 1.0

59 53 47

0.8

41 0.6

35 30

0.4

24 18

0.2

12 6 0

20

40

60 80 100 Body weight (lb)

120

140

160

Figure 50.1╇ Relations between body weight in pounds, body surface area, and adult dosage. Note that the adult dose is for a patient weighing about 140 lbs and with a body surface area of approximately 1.2 m squared. (Reproduced from Talbot, N. B., et al., Metabolic Homeostasis: A Syllabus for Those Concerned with the Care of Patients. Harvard University Press, Cambridge, MA, 1959. With permission.)

50.12â•… Episode IV: Correct Answers to the Pretest 9IV. The correct answer is E. The use of ethosuximide is not associated with paradoxical hyperactivity; this side effect is most commonly associated with the use of phenobarbital in children. Phenytoin is associated with an increased risk of osteoporosis, especially in postmenopausal women. Vitamin D and calcium supplements should be initiated in all women taking phenytoin who are approaching menopause. Valproic acid is a relatively strong acid and is associated with gastric irritability; it also may affect the hypothalamus, causing weight gain and amenorrhea. 10IV. D is correct. Low folic acid levels are common in patients taking phenytoin. There is no such correlation with levels of vitamin B12. Whether the low folic acid levels contribute to the occurrence of the low-grade peripheral neuropathy seen in association with long-term phenytoin use is unclear. However, there is an increasing body of evidence that suggests that supplementary folic acid use in women before and during pregnancy diminishes the likelihood of fetal developmental abnormalities. Cerebellar ataxia is commonly a

Clinical Pharmacology of Antiepileptic Drug Use

839

result of toxic phenytoin levels and is reversible when the serum drug level is lowered below the toxic range. It should be noted, however, that permanent parenchymatous cerebellar atrophy may occur in patients who have taken phenytoin for a decade or more. Gingival hyperplasia is probably one of the most common side effects of phenytoin use. It is aggravated by dental plaque formation and is prevented by vigorous dental hygiene measures (frequent flossing and brushing, use of antiplaque mouth wash, preventive dentistry). Once hyperplasia is present, preventive measures are of little value; at that point, the only way of reversing the gum changes is to discontinue use of phenytoin and substitute another antiepileptic drug.

50.13â•… Episode IV After the birth of her second child, Patty’s weight plummeted by 23 lb. Her unsteady gait had improved despite a total phenytoin level of 21 μg/mL. Nonetheless, her physician decreased her phenytoin dose to 100 mg three times a day. Patty had a grand mal seizure 1 week later. Her total phenytoin level was 11 μg/mL, and she was again accused of not complying with therapy. Patty subsequently had asthma and was given theophylline. She then contracted bronchitis and was treated with erythromycin. Because of the recurrence of seizures, phenytoin was increased to 200 mg twice daily; however, she experienced increased frequency of seizures nonetheless. Her physician discontinued erythromycin, and the seizures ended. The phenytoin dosage was decreased to 100 mg in the morning and 200 mg before bed. Patty remained seizure free for the next 11 years, during which time the results of three EEGs were normal. At the age of 35 years, despite a total level of 12 μg/mL and a free level of only 1 μg/mL, Patty had increasing difficulty with gait ataxia. A computed tomographic scan demonstrated cerebellar atrophy. There was no family history of ataxia, and Patty did not drink alcohol. Further examination showed diminished sensation in her distal extremities and absent deep tendon (myotatic) reflexes. Phenytoin treatment was tapered off over 3 months; since discontinuation, the patient has had no seizures.

50.14â•… Questions for Episode IV 4.1. Postpartum, without a decrease in her total dosage and in spite of an increase in her total phenytoin level, Patty’s balance continued to improve. Why? 4.2. Ignoring the clinical improvement in her toxic symptoms and relying only on her total phenytoin level, her physician decreased Patty’s phenytoin dosage to 100 mg three times daily. A 25% decrease in dosage resulted in a 50% decrease in her phenÂ� ytoin level. What happened? 4.3. Patty developed asthma and was started on theophylline and erythromycin. Why did her seizures increase in frequency? 4.4. At 35 years of age, despite low total and free therapeutic phenytoin levels, Patty’s gait became progressively ataxic. Why? 4.5. Patty also reported tingling in her toes and numbness in her feet. These symptoms combined with absent reflexes are diagnostic of a peripheral neuropathy. Where did this come from?

840 Sudden Death in Epilepsy: Forensic and Clinical Issues

50.14.1â•…Discussion Notes for Questions in Episode IV 4.1. Patty lost weight postpartum. Her phenytoin levels increased yet her toxic symptoms of ataxia, although still present, improved. The issue is that of increased protein binding of phenytoin postpartum, lowering the free phenytoin level. Free phenytoin levels should be evaluated along with total levels whenever additional drugs or changing metabolic circumstances (e.g., pregnancy, uremia) could affect the protein binding of the primary antiepileptic drug. 4.2. The physician inappropriately decreased the phenytoin dosage by 100 mg, resulting in a seizure. Zero-order kinetics in reverse resulted in halving the phenytoin level despite only a 25% decrease in dosage. The patient also had a higher percentage of protein binding than when she was pregnant, resulting in even less availability of phenytoin to the brain. 4.3. It is well known that toxic levels of xanthines cause seizures. The physician did not realize that erythromycin interferes with the metabolism of theophylline, resulting in toxic concentrations of theophylline, which lowered seizure threshold. 4.4. In some persons, long-term use of phenytoin can result in diffuse parenchymatous cerebellar damage, presumably from a long-term toxic effect of phenytoin. The mechanism of this association is unclear. Consideration should be given to using an alternative antiepileptic drug when nontoxic free and total levels of phenytoin are associated with slowly worsening ataxia in persons who take the drug for more than a decade. The clinical symptoms commonly improve, albeit slowly, after discontinuation of the offending drug. 4.5. Phenytoin also can cause chronic lowering of folic acid concentrations. Although this interaction may be a factor in explaining peripheral neuropathy associated with longterm phenytoin use, the actual mechanism is unclear.

50.15â•… Posttest Questions 1I. Antiepileptic drugs should be used in a child with febrile seizures except in the following circumstances: A. Onset of seizures after 4 years of age B. History of delays in development before the ἀrst seizure C. Focal component to the seizure D. Brief duration of seizure without neurologic deἀcits before or after the event E. Two or more seizures within 24 hours 2I. Useful characteristics of phenobarbital in the prevention of complicated febrile seizures include all of the following except: A. A short t1/2, allowing short-term treatment at the onset of a fever B. A t1/2 resulting in minimal interdose peak/trough serum level variability C. Once daily dosage D. The predictability of its dose-related side effects E. Its linear kinetics, resulting in a predictable dose response 3I. Pediatric dosing is most appropriately done using: A. mg/lb B. mg/kg C. mg/m2 as determined using a nomogram D. mg/m3 E. mg/kg from a nomogram

Clinical Pharmacology of Antiepileptic Drug Use

841

4II. Generic forms of drugs are said to be equivalent if the bioavailability is what percentage of the brand name product? A. 100% B. 90–110% C. 85–115% D. 80–120% E. 70–130% 5II. Phenobarbital is considered to be an effective and safe antiepileptic drug. However, problems concomitant with the use of phenobarbital include all of the following except: A. Cognitive impairment at therapeutic levels B. Withdrawal seizures when discontinued C. The necessity to administer it in frequent divided doses D. A long t1/2 resulting in a delay of dose response with dose changes E. Availability in a parenteral form 6III. Gynecologic complications and pitfalls of antiepileptic drug use include which of the following? A. Valproic acid–induced amenorrhea B. Phenobarbital-induced menorrhagia C. Augmentation of birth control efficacy by phenytoin D. Increased body hair from carbamazepine use E. Decreased birth control pill efficacy from phenobarbital 7III. The concomitant use of phenytoin and warfarin results in all but one of the following: A. Inhibition of parahydroxylation of phenytoin in the liver B. Increased metabolism of warfarin due to liver enzyme induction C. Prolongation of the prothrombin time D. Decreased anticoagulant effect E. Decreased serum levels of phenytoin 8III. Concomitant use of steroids and phenytoin can result in which one of the following? A. Pregnancy B. Decreased dosage requirement for corticosteroids C. Augmented anovulatory effect of birth control pills D. Prolonged t1/2 of corticosteroids E. Increased immunosuppressive effect of corticosteroids 9III. Valproic acid has complex interactions with other antiepileptic drugs. These interactions include which of the following? A. Lowered phenobarbital serum levels B. Raised total serum phenytoin levels C. Decreased percentage of free phenytoin D. Phenobarbital toxicity E. Valproic acid level is lowered by phenytoin 10IV. Phenytoin interacts with many other drugs. Which of the following is a known interaction? A. Enhances erythromycin metabolism B. Increases carbamazepine plasma levels C. Diminishes the efficacy of warfarin

842 Sudden Death in Epilepsy: Forensic and Clinical Issues

D. Diminishes the efficacy of theophylline E. Causes phenobarbital toxicity to occur 11IV. Long-term effects of phenytoin use include all of the following except: A. Low serum folic acid level B. Elevated vitamin B12 level C. Peripheral neuropathy D. Parenchymatous cerebellar atrophy E. Osteomalacia

50.16â•… Correct Answers to the Posttest 1I. D is correct. A febrile seizure must be differentiated between uncomplicated and complicated types. An uncomplicated febrile seizure occurs in a neurologically normal child who is usually younger than 4 years of age. A complicated febrile convulsion occurs in a neurologically impaired child, is of long duration, may have a focal component, and is often associated with ongoing EEG abnormalities long after the seizure is over. Anticonvulsants, usually phenobarbital, are recommended for use only in cases of complicated convulsions, as these children have a much greater likelihood of developing epilepsy, i.e., nonfebrile convulsions. 2I. A is correct. Phenobarbital has a very long t1/2 (80–120 hours), which militates against interval therapy. In effect, interval therapy is of no use in preventing seizures in association with fever as they often occur during the initial stages of the fever. The desirable characteristics of phenobarbital include a long t1/2, which allows for once-daily dosing (which improves patient compliance) with minimal peak/trough serum level variability and a predictable dose response in most persons when used as monotherapy. The drug dose-related side effects, i.e., drowsiness and impaired learning in some patients, are predictable and are often remedied by lowering the dose, or, if necessary, changing to an alternative agent. 3I. C is correct. Pediatric dosing is more complex than with adults. Particularly in very young children, using mg/m2 with a nomogram to determine dosage is desirable initially. However, to determine the optimal steady-state dose, the dose response is a function of the desired serum level. 4II. D is correct. According to federal standards, generic forms of drugs are said to be equivalent if the bioavailability is 80–120% of the steady-state level of the brand name form of the agent. 5II. C is correct. Despite its long t1/2 (80–120 hours), phenobarbital, and for that matter all antiepileptic drugs, can produce withdrawal seizures if discontinued abruptly. Drugs with a long t1/2, such as phenobarbital and phenytoin, have the advantage of a once- or twice-daily dosage schedule being sufficient to maintain a stable blood level throughout the day. However, in contrast to agents with a short t1/2, such as carbamazepine, longer intervals are needed between adjustments to ἀnd the new steady-state level, making dosing adjustments prolonged. It is of note that an increasing body of evidence suggests that any of the antiepileptic drugs, not just barbiturates, can cause cognitive impairment, and that individual susceptibility and tolerance is a very important factor in choice of drugs. 7III. The correct answer is C. Phenytoin and warfarin are both metabolized in the liver. Warfarin inhibits the breakdown of phenytoin by interfering with parahydroxylation.

Clinical Pharmacology of Antiepileptic Drug Use

843

Phenytoin stimulates the metabolism of warfarin due to enzyme induction. Phenytoin may consequently decrease the anticoagulant effect of warfarin, and warfarin may increase phenytoin levels to the point of clinical toxicity. 8III. The correct answer is A. Many women experience amenorrhea when taking valproic acid. Menses usually returns when the drug is discontinued. This side effect is not seen with the use of phenobarbital. Because phenytoin increases the metabolic breakdown of steroids, it will tend to diminish the efficacy of birth control pills. Phenytoin, not carbamazepine, can lead to hirsutism, which unfortunately is not reversible once treatment with the drug is stopped. Finally, phenobarbital is not known to decrease the efficacy of birth control pills. 9III. The correct answer is D. Valproic acid interferes with the metabolic breakdown of phenobarbital, resulting in elevation of phenobarbital levels. If valproic acid is added to the treatment program of a patient who has stable therapeutic levels of phenobarbital, the latter level is likely to rise to the toxic range. The effect of valproic acid on phenytoin is complex in that the free phenytoin level will tend to increase because of displacement from plasma protein, whereas the total level will tend to drop. There is little effect of phenytoin on serum levels of valproic acid. 10IV. The correct answer is C. Phenytoin tends to increase the metabolism of warfarin and thus decrease the efficacy of that drug. The cause of seizures in persons taking phenytoin, theophylline, and erythromycin is the result of the interference with metabolic breakdown of theophylline by erythromycin. Phenytoin has little effect on this interaction. The increased risk of seizures results from the marked elevation of the blood theophylline concentration. Finally, although there may be a slight effect of phenytoin on the phenobarbital concentration, there is little chance it will cause phenobarbital toxicity to occur. 11IV. The correct answer is B. There is little effect of phenytoin on metabolism of vitamin B12. However, serum folic acid levels are often low in persons taking phenytoin. This effect is not correlated with any clinical symptoms. Peripheral neuropathy is found only after many years of use and is usually rather mild. Parenchymatous cerebellar atrophy is a rare but incapacitating complication of long-term phenytoin use. The mechanism is unclear. Because the occurrence of osteomalacia is particularly problematic in postmenopausal women, the use of other antiepileptic drugs is preferable in this group of patients. Evaluation by Third- or Fourth-Year Medical Students of Clinical Pharmacology Problem Solving (CPPS) Unit Entitled Clinical Pharmacology of Antiepileptic Drug Use: “Clinical Pearls about the Perils of Patty” Paul L. Schraeder and Claire M. Lathers This unit (please check the appropriate opinion): 1. Met the six stated learning objectives 2. Taught me new material about the treatment of persons with epilepsy 3. Will alter my treatment of persons with epilepsy 4. Is relevant to my clinical experience in neurology 5. Is relevant to my clinical experience in clinical pharmacology

Agreeâ•… Disagreeâ•… No Opinion

844 Sudden Death in Epilepsy: Forensic and Clinical Issues

6. Satisἀed my expectations 7. The faculty discussion of this CPPS unit enhanced the material covered in the unit 8. The posttest questions accurately reflected the material discussed in the CPPS unit Please answer the following questions by ἀlling in the blanks. 9. If you answered no to question 2 above, asking whether the CPPS unit met the stated learning objectives, which one(s) of the six learning objectives were not met? 10. This CPPS unit may be used by a group of four students or by one student reading all four sections. Which format did you participate in, i.e., four students versus one student? 11. Which format do you think is the best for presentation of this material, i.e., four students versus one student? 12. Would you like to use additional CPPS units on different topics? If so, in what clinical area(s)? 13. What area of specialty does the faculty presenter work in (neurology, clinical pharmacology, pharmacology, obstetrics-gynecology, pharmacy, PharmD program, other)?

References Dansky, L. V., D. S. Rosenblatt, and E. Andermann. 1992. Mechanisms of teratogenesis: Folic acid and antiepileptic therapy. Neurology 42 (4 Suppl 5): 32–42. Evans, W. E., J. J. Shentag, and W. J. Jusko, eds. 1966. Applied Pharmacokinetics: Principles of Therapeutic Drug Monitoring, 2nd ed. Spokane, WA: Applied Therapeutics. Finnell, R. H., B. A. Buehler, B. M. Kerr, P. L. Ager, and R. H. Levy. 1992. Clinical and experimental studies linking oxidative metabolism to phenytoin-induced teratogenesis. Neurology 42 (4 Suppl 5): 25–31. Hansten, P. D. 1979. Drug Interactions: Clinical Significance of Drug–Drug Interactions and Drug Effects on Clinical Laboratory Results, 4th ed. Philadelphia, PA: Lea & Febiger. Lindhout, D. 1992. Pharmacogenetics and drug interactions: Role in antiepileptic-drug-induced teratogenesis. Neurology 42 (4 Suppl 5): 43–47. Mikati, M., N. Bassett, and S. Schachter. 1992. Double-blind randomized study comparing brandname and generic phenytoin monotherapy. Epilepsia 33 (2): 359–365. Rall, T. W., L. S. Schleifer, A. S. Nies, and R. H. Levy. 1990. Drugs effective in the therapy of the epilepsies. In Goodman and Gillman’s The Pharmacological Basis of Therapeutics, ed. A. Gillman, T. Rall, A. Nies, and P. Taylor. New York, NY: Pergamon. Schraeder, P. L., and C. M. Lathers. 1995. Clinical pharmacology of antiepileptic drug use: “clinical pearls about the perils of patty.” J Clin Pharmacol 35 (12): 1120–1135. Talbot N. B., R. H. Rickie, J. D. Crawford and E. S. Tagrin. 1959. Metabolic Homeostasis: A Sysllabus for Those Concerned with the Care of Patients. Cambridge, MA: Harvard University Press. Theodore, W. H. 1992. Rational use of antiepileptic drug levels. Pharmacol Ther 54 (3): 297–305. Yerby, M. S., P. N. Friel, and K. McCormick. 1992. Antiepileptic drug disposition during pregnancy. Neurology 42 (4 Suppl 5): 12–16.

Compliance with Antiepileptic Drug Treatment and the Risk of Sudden Unexpected Death in Epilepsy

51

Torbjörn Tomson

Contents 51.1 Introduction 51.2 What Is Compliance and Noncompliance? 51.3 How Can Compliance Be Assessed? 51.4 How Could Poor Compliance Relate to the Risk of SUDEP? 51.5 Postmortem Drug Levels 51.6 Other Methods to Assess Nonadherence 51.7 Conclusions References

845 845 846 846 847 849 849 850

51.1â•…Introduction The ultimate goal of all research on sudden unexpected death in epilepsy (SUDEP) is to ἀnd methods to reduce the risk of this most devastating consequence of epilepsy. As poorly controlled epilepsy appears to be associated with a particularly high risk of SUDEP (Tomson et al. 2008), and since drug treatment is a mainstay in epilepsy therapy, much attention has been given to the management of pharmacotherapy in epilepsy. Among drug-related risk factors that have been investigated are polytherapy with antiepileptic drugs, use of speciἀc anticonvulsants, and adherence to the prescribed antiepileptic drug regimen (Tomson et al. 2008). Such treatment-related risk factors are of particular interest as they may be amenable to changes that eventually prevent or reduce the risk of SUDEP. Poor compliance with antiepileptic drug treatment has been of special interest as this is a common cause of treatment failure (Cramer et al. 2002) and is also manifested as increased frequency of hospitalizations and emergency room admissions (Davis et al. 2008). Poor compliance has frequently been suggested as a risk factor for SUDEP (Téllez-Zenteno et al. 2005; Tomson et al. 2005; Monté et al. 2007). This chapter will discuss nonadherence to drug treatment and the available evidence that this may play a role in SUDEP.

51.2â•… What Is Compliance and Noncompliance? Compliance can be deἀned as “the extent to which a person’s behavior (in terms of mediÂ� cations, following diet, or executing lifestyle changes) coincides with medical or health advice” (Haynes 1979). Even conἀning the discussion to noncompliance with drug 845

846 Sudden Death in Epilepsy: Forensic and Clinical Issues

therapy,€this remains a major issue in medicine in general. It has been estimated that half of patients for whom appropriate therapy is prescribed fail to receive full beneἀt because of inadequate adherence to treatment (Haynes 1979). Epilepsy is no exception. A consensus document was published after the First International Workshop on Compliance in Epilepsy (Leppik and Schmidt 1988). It was suggested that compliance be categorized by three dimensions: (1) type of behavior (consistent overcompliers, consistent undercompliers, or those who are irregular in behavior); (2) extent of compliance (ranging from those who do not take the medication at all to those who take every dose as prescribed); and (3) whether the patient is intentionally noncompliant or not. Cramer and collaborators (2008) have stressed the importance of distinguishing between medication compliance and medication persistence, in which the latter refers to the act of continuing the treatment for the prescribed duration. Depending on methods used, criteria and deἀnitions, and type of population, estimates of noncompliance among epilepsy patients range from 20% to 65% (Leppik 1988; Tomson 1995; Davis et al. 2008). However, as suggested by the consensus document from the workshop, it is better to describe the extent of noncompliance using continuous variables rather than an oversimpliἀed either/or dichotomy.

51.3â•…How Can Compliance Be Assessed? Many different methods have been used to assess compliance in different epilepsy populations. These methods have been grouped in two categories: direct and indirect (Paschal et al. 2008). Direct measures involve determination of drug concentrations in plasma, saliva, and hair, whereas indirect methods utilize manual or electronic pill counts, self-reports, and medication reἀlls. All methods have their merits and limitations, although plasma drug level monitoring is more of a standard.

51.4â•…How Could Poor Compliance Relate to the Risk of SUDEP? Nonadherence to the prescribed antiepileptic drug treatment, noncompliance, has been claimed to be a major cause of treatment failure in epilepsy. Some early small-scale studies have indicated that unreliable drug intake and poor compliance could explain the majority of seizures in selected populations of patients with poor seizure control (Kutt et al. 1966; Cramer and Mattson 1991). A more recent questionnaire-based survey of 670 epilepsy patients in the United States found that 45% of patients reported a seizure after a missed dose (Cramer et al. 2002). Nonadherence to anticonvulsant treatment is thus associated with poor seizure control. It is therefore reasonable to assume that poor medication compliance also increases the risk of SUDEP since SUDEP, in most instances, occurs in the context of a seizure (Langan et al. 2000). Poor control of generalized tonic–clonic seizures has also been the strongest and most consistent risk factor in case–control studies (Tomson et al. 2008). In a retrospective cohort study, Faught and colleagues (2008) used Medicaid claims data to evaluate adherence to treatment in more than 33,000 patients with antiepileptic drug prescriptions. Nonadherence was associated with a more than 3-fold increase in mortality compared to adherence with a hazard ratio of 3.32 (95% conἀdence interval,

Compliance with Antiepileptic Drug Treatment

847

3.11–3.54). Unfortunately, these investigators analyzed all causes of mortality but not SUDEP speciἀcally. Noncompliance theoretically could increase the risk of SUDEP through mechanisms other than by increasing the risk of seizures. The role of autonomic cardiac control has been much discussed in relation to SUDEP. A decreased heart rate variability is used as a marker of autonomic function and is associated with risk of sudden death in clinical conditions other than epilepsy (Bigger et al. 1993). People with chronic epilepsy have also been shown to have decreased heart rate variability, and it has been speculated that this could be associated with increased risk of cardiac arrhythmias in relation to seizures and thus to SUDEP (Persson et al. 2007). Two studies have analyzed the effects of rapid withdrawal of antiepileptic drugs on heart rate variability (Kennebäck et al. 1997; Hennessy et al. 2001). In the ἀrst study, 10 patients on carbamazepine or phenytoin were studied in conjunction with abrupt drug withdrawal due to adverse effects (Kennebäck et al. 1997). A signiἀcant reduction in both time and frequency domains of heart rate variability was observed. In addition, 3 of the 10 patients had a 10-fold increase in ventricular premature beats. The second study noted increased sympathetic activity in sleep when carbamazepine was withdrawn during monitoring for epilepsy surgery work-up (Hennessy et al. 2001). Although the antiepileptic drug withdrawal in these studies was planned, it may mimic the situation in noncompliant patients and indicate that cardiac or autonomic effects can follow rapid changes in plasma levels of antiepileptic drugs.

51.5â•… Postmortem Drug Levels The notion that poor compliance is a major risk factor for SUDEP largely stems from€reports of postmortem concentrations of antiepileptic drugs in SUDEP victims.€“Subtherapeutic” or even undetectable drug levels were frequently found in SUDEP cases from the coroner’s office. The results of such studies are presented in Table 51.1. The proportion found to have such low or undetectable drug concentrations ranges from 65% to almost 100% of cases in uncontrolled studies (Terrence et al. 1975; Leestma et al. 1984, 1989; Earnest et al. 1992; Langan et al. 1998; Ficker et al. 1998; Lear-Kaul et al. 2005; Table 51.1). However, for several reasons, such results are difficult to interpret. First, postmortem drug concentrations may not be readily comparable to plasma concentrations obtained in live patients. The postmortem analysis is often made on whole blood rather than plasma. Additionally, the concentration of an anticonvulsant may change after death, e.g., by redistribution to other tissues than blood. Hence, it has been shown that such alterations can substantially reduce phenytoin concentrations after death (Tomson et al. 1998). Similar changes are likely to occur with some other antiepileptic drugs. Second, ἀndings of subtherapeutic postmortem plasma concentrations do not necessarily imply that the patient is either noncompliant or undertreated. The often quoted therapeutic ranges are in general poorly deἀned for antiepileptic drugs, and it is well established that a large proportion of epilepsy patients are well controlled at drug concentrations below these ranges (Patsalos et al. 2008). A more meaningful interpretation of postmortem drug levels in SUDEP victims therefore requires an appropriate control group in addition to taking into account the possible postmortem changes in concentrations. Few studies have included control populations, and the results are somewhat conflicting. In a retrospective study from the United States, George and Davis (1998) examined postmortem antiepileptic drug levels in 52 SUDEP

848 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 51.1â•… Studies of Postmortem Concentrations of Antiepileptic Drugs in SUDEP Cases Where Such Levels Were Measured Reference

SUDEP cases (n)

Observation antiepileptic drug levels in SUDEP

Terrence et al. 1975

37

19/37 “Not therapeutic,”15/37 undetectable 46/66 Subtherapeutic or undetectable 51/54 Subtherapeutic or undetectable 34/39 Subtherapeutic or undetectable 4/6 Subtherapeutic or absent 4/4 Subtherapeutic or absent 13/23 Subtherapeutic

Leestma et al. 1984

66

Leestma et al. 1989

54

Earnest et al. 1992

39

Langan et al. 1998 Ficker et al. 1998 Kloster and Engelskjon 1999

6 4 23

Opeskin et al. 1999

44

10/44 Subtherapeutic, 13/44 undetectable

George and Davis 1998

52

36/52 Subtherapeutic

Lear-Kaul et al. 2005

67

51/67 Subtherapeutic or absent

Controls None None None None None None 0/7 Subtherapeutic in non-SUDEP deaths 13/44 Subtherapeutic, 11/44 undetectable among non-SUDEP deaths 15/44 Subtherapeutic in non-SUDEP deaths None

cases and 44 deceased epilepsy controls whose deaths were considered to be unrelated to their epilepsy (e.g., ischemic heart disease, accidents including drowning, suicide, and homicide). Antiepileptic drug levels were found to be subtherapeutic in 69% of the SUDEP cases compared to 34% in the control population. Another retrospective study from the National Epilepsy Centre in Norway analyzed postmortem serum concentrations of antiepileptic drugs in 23 SUDEP cases and in seven epilepsy control patients who had died for reasons other than SUDEP (e.g., pneumonia, cardiac disease, status epilepticus, trauma, suicide, and drowning) (Kloster and Engelskjon 1999). Subtherapeutic concentrations were noted in 57% of SUDEP victims compared with none among the non-SUDEP deaths. A third retrospective controlled study analyzed coroner cases from Australia (Opeskin et al. 1999). There were 44 SUDEP cases and one control per SUDEP case. Controls were epilepsy patients who died of causes other than epilepsy (e.g., ischemic heart disease, accidents that could be due to poor seizure control, and suicide). Compared with the controls, the SUDEP cases showed no difference in the number with undetectable or subtherapeutic levels. The number with therapeutic concentrations was the same (21) in both groups. Hence, while the studies from the United States and Norway found low postmortem antiepileptic drug concentrations to be more common among SUDEP cases than controls, this was not the case in the study from Australia. Differences between the control groups could contribute to these apparently conflicting observations. Evidently, the composition of the control group is important. In this comparison, the controls are meant to represent drug concentrations in the average non-SUDEP epilepsy population. However, the controls are patients who have died for various reasons, and it may well be that depending on the cause of death, antiepileptic drug levels may be different from those of the general epilepsy population. For example, patient deaths in status epilepticus or seizure-related accidents may be because of low drug levels. In conclusion, even with inclusion of a control group, it is

Compliance with Antiepileptic Drug Treatment

849

difficult to assess the level of compliance with the prescribed treatment regimen using postmortem drug levels.

51.6â•…Other Methods to Assess Nonadherence The level of adherence to antiepileptic drug treatment in SUDEP cases has been assessed by methods other than measuring postmortem drug levels. In a case-control study from Minnesota, compliance was assessed by anticonvulsant levels at the last visit in 17 SUDEP cases and 67 living epilepsy controls (Walczak et al. 2001). Very few of the cases (6%) and controls (15%) had none of their anticonvulsant levels within the therapeutic range. All levels were therapeutic in 59% of the cases and 40% of the controls. Hence, compliance appeared to be high among SUDEP cases judging from drug levels at the last visit; however, this is based on a spot check on one single occasion per patient. A Swedish case-control study instead analyzed the within-patient variation in antiepileptic drug plasma concentrations during the last year before death in 57 SUDEP cases or the comparison period in 171 living epilepsy controls (Nilsson et al. 2001). Fluctuations in drug levels over time have been considered to reflect the level of noncompliance (Leppik 1988). No signiἀcant association was found between SUDEP risk and within-patient variation in drug levels over time. Thus, none of these studies indicates that noncompliance is more frequent among SUDEP cases than in living epilepsy controls. Williams and collaborators (2006) used a different approach. They compared hair antiepileptic drug concentration variability in 16 patients with SUDEP, 9 non-SUDEP epilepsy–related deaths, 31 epilepsy outpatients, and 38 epilepsy inpatients. Drug concentrations were measured in 1 cm hair segments (1 cm is assumed to represent 1 month). The coefficient of variation of the corrected mean hair concentration was used as an index of variability of an individual’s antiepileptic-drug-taking behavior. The observed variability of hair concentrations was greater in SUDEP cases than in epilepsy outpatients or inpatients, suggesting more variable antiepileptic drug ingestion over time. In a controlled prospective study from Australia, compliance was assessed by questionnaires to the patient’s physicians (Opeskin and Berkovic 2003). Doctors were asked to indicate whether compliance with antiepileptic drug treatment was good or poor based on the detail of medication usage and antemortem drug levels. Poor compliance was not considered to be more common among the 50 SUDEP cases than in the 50 epilepsy controls who had died from other causes.

51.7â•… Conclusions The available data on the level of adherence to prescribed anticonvulsant medication among SUDEP cases are conflicting irrespective of the method used to assess compliance. In particular, there is no consistent evidence from controlled studies that SUDEP victims are less compliant than the epilepsy patient in general; however, this does not mean that noncompliance cannot play a role in SUDEP. Medication compliance is likely to improve seizure control and thus probably reduce the risk of SUDEP, although ἀrm evidence is lacking.

850 Sudden Death in Epilepsy: Forensic and Clinical Issues

References Bigger, J. T., J. L. Fleiss, L. M. Rolnitzky, et al. 1993. Frequency domain measures of heart period variability to assess risk late after myocardial infarction. J Am Coll Cardiol 21(3): 729–736. Cramer, J. A., and R. H. Mattson. 1991. Monitoring compliance with antiepileptic drug therapy. In Patient Compliance in Medical Practice and Clinical Trial, ed. J. A. Cramer and B. Spilker, 123–127. New York, NY: Raven Press. Cramer, J. A., M. Glassman, and V. Rienzi. 2002. The relationship between poor medication compliance and seizures. Epilepsy Behav 3 (4): 338–342. Cramer, J. A., A. Roy, A. Burrell, et al. 2008. Medication compliance and persistence: Terminology and deἀnitions. Value Health 11: 44–47. Davis, K. L., S. D. Candrilli, and H. M. Edin. 2008. Prevalence and cost of non-adherence with antiepileptic drugs in an adult managed care population. Epilepsia 49: 446–454. Faught, E., M. S. Duh, J. R. Weiner, A. Guérin, and M. Cunnington. 2008. Nonadherence to antiepileptic drugs and increased mortality: Findings from the RANSOM Study. Neurology 71 (20): 1572–1578. Ficker, D. M., E. L. So, J. F. Annegers, P. C. O’Brien, G. D. Cascino, and P. G. Belau. 1998. Populationbased study of the incidence of sudden unexplained death in epilepsy. Neurology 51: 1270–1274. George, J. R., and G. G. Davis. 1998. Comparison of anti-epileptic drug levels in different cases of sudden death. J Forensic Sci 43: 598–603. Haynes, R. B. 1979. Introduction. In Compliance in Health Care, ed. R. B. Haynes, D. W. Taylor, and D. L. Sacket, 1–7. Baltimore, MD: Johns Hopkins University Press. Hennessy, M. J., M. G. Tighe, C. D. Binnie, and L. Nashef. 2001. Sudden withdrawal of carbaÂ� mazepine€increases cardiac sympathetic activity in sleep. Neurology 57 (9): 1650–1654. Kennebäck, G., M. Ericson, T. Tomson, and L. Bergfeldt. 1997. Changes in arrhythmia proἀle and heart rate variability during abrupt withdrawal of antiepileptic drugs. Implications for sudden death. Seizure 1997 6: 369–375. Kloster, R., and T. Engelskjon. 1999. Sudden unexpected death in epilepsy: A clinical perspective and a search for risk factors. J Neurol Neurosurg Psychiatry 67: 439–444. Kutt, H., J. Haynes, and F. McDowell. 1966. Some causes of ineffectiveness of diphenylhydantoin. Arch Neurol 14: 489–492. Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68: 211–213. Langan, Y., N. Nolan, and M. Hutchinson. 1998. The incidence of sudden unexpected death in epilepsy (SUDEP) in South Dublin and Wicklow. Seizure 7: 355–358. Lear-Kaul, K. C., L. Coughlin, and M. J. Dobersen. 2005. Sudden unexpected death in epilepsy, a retrospective study. Am J Forensic Med Pathol 26: 11–17. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26: 195–203. Leppik, I. E. 1988. Variability of phenytoin, carbamazepine, and valproate concentrations in a clinical population. In Compliance in Epilepsy, ed. D. Schmidt and I. E. Leppik, 85–90 (Epilepsy Res Suppl. 1). Amsterdam: Elsevier Science Publishers. Leppik, I. E., and D. Schmidt. 1988. Consensus statement on compliance in epilepsy. In Compliance in Epilepsy, ed. D. Schmidt and I. E. Leppik, 179–182 (Epilepsy Res Suppl. 1). Amsterdam: Elsevier Science Publishers. Monté, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. D. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systematic review. Seizure 16 (1): 1–7.

Compliance with Antiepileptic Drug Treatment

851

Nilsson, L., U. Bergman, V. K. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case–control study. Epilepsia 42: 667–673. Opeskin, K., and S. F. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12: 456–464. Opeskin, K., M. P. Burke, S. M. Cordner, and S. F. Berkovic. 1999. Comparison of antiepileptic drug levels in sudden unexpected death in epilepsy with death from other causes. Epilepsia 40: 1795–1798. Paschal, A. M., S. R. Hawley, T. St Romain, and E. Ablah. 2008. Measures of adherence to epilepsy treatment: Review of present practices and recommendations for future directions. Epilepsia 49 (7): 1115–1122. Patsalos, P. N., D. J. Berry, B. F. Bourgeois, et al. 2008. Antiepileptic drugs—best practice guidelines for therapeutic drug monitoring: A position paper by the subcommission on therapeutic drug monitoring, ILAE Commission on Therapeutic Strategies. Epilepsia 49 (7): 1239–1276. Persson, H., E. Kumlien, M. Ericson, and T. Tomson. 2007. Circadian variation in heart-rate variability in localization-related epilepsy. Epilepsia 48 (5): 917–922. Téllez-Zenteno, J. F., L. H. Ronquillo, and S. Wiebe. 2005. Sudden unexpected death in epilepsy: Evidence-based analysis of incidence and risk factors. Epilepsy Res 65: 101–115. Terrence, Jr., C. F., H. W. Wisotzkey, and J. A. Perper. 1975. Unexpected, unexplained death in epileptic patients. Neurology 25: 594–598. Tomson, T. 1995. Non-compliance and intractability of epilepsy. In Intractable Epilepsy, ed. S. I. Johannessen, L. Gram, M. Sillanpää, and T. Tomson, 93–103. Petersἀeld: Wrightson Biomedical Publ. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7 (11): 1021–1031. Tomson, T., A. C. Skold, P. Holmgen, L. Nilsson, and B. Danielsson. 1998. Postmortem changes in blood concentrations of phenytoin and carbamazepine: An experimental study. Ther Drug Monit 20: 309–312. Tomson, T., T. Walczak, M. Sillanpää, and J. W. A. S. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (S11): 54–61. Walczak, T. S., I. E. Leppik, M. D’Amelio, et al. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56: 519–525. Williams, J., C. Lawthom, F. Dunstan, et al. 2006. Variability of antiepileptic medication taking behaviour in sudden unexplained death in epilepsy: Hair analysis at autopsy. J Neurol Neurosurg Psychiatry 77: 481–484.

SUDEP Clinical Case Histories Typical and Atypical Paul L. Schraeder

52

Contents 52.1 Cases of SUDEP with a Premorbid Diagnosis of Nonepileptic Seizures 52.2 Discussion References

853 858 859€

52.1â•…Cases of SUDEP with a Premorbid Diagnosis of Nonepileptic Seizures The risk factors associated with sudden unexpected death in epilepsy (SUDEP) are discussed in detail throughout this book as determined by statistical analysis of population data. This brief chapter presents the history of individuals who succumbed, or nearly succumbed, to SUDEP with a succinct discussion after each that is an attempt to focus attention on both the common and the unique issues that can be gleaned from each case. These cases emphasize the potential role of individual risk factors associated with the occurrence of SUDEP. Obviously, not all issues related to risk factors are addressed; however, the compelling circumstances illustrated in these cases cover a relatively broad spectrum despite the small number of cases presented. A focused summary discussion is presented after each case and at the end of the chapter.

Case 1â•… A Case o f Put at iv e No nepilept ic Seiz ures A 35-year-old woman was found dead in bed by her parents. She had a long-standing psychiatric history with the occurrence of complex seizure-like events with a differential diagnosis of nonepileptic seizures. On one occasion, she had a nonepileptic seizure in her neurologist’s office consisting of sliding to the floor with bizarre asynchronous bilateral motor activity without loss of consciousness. This event occurred consequent to a recent, emotionally tense situation at her home. Complicating the history was the past observation of another type of event that consisted of some automatisms and postictal confusion. Multiple routine electroencephalograms (EEGs) over the years were unremarkable save for one that showed unequivocal isolated left temporal interictal discharges. The patient was placed on carbamazepine and was found to have consistently therapeutic levels. She was able to achieve enough insight to be able to separate the actual complex partial seizures from the nonepileptic events when discussing their occurrence with her neurologist. Presumably, as a result of the therapeutic carbamazepine levels, she recognized that the complex partial seizures were occurring very infrequently, but that she continued to have occasional nonepileptic 853

854 Sudden Death in Epilepsy: Forensic and Clinical Issues

events. After the retirement of her neurologist, the patient was seen at another center and subjected to several days of inpatient EEG monitoring during which multiple clinical events were recorded without any epileptiform activity being documented on any of the closed-circuit TV/EEG recordings. As a result, the patient was informed that her seizures were nonepileptic and advised to taper her antiepileptic medications. Several weeks later, her parents found her dead in bed. No autopsy was performed. Co mment s Persons with only nonepileptic seizures would not be at risk for SUDEP. However, the unfortunate reality is that 10–30% of persons who appear to have long established nonepileptic seizures also have epilepsy (Betts 1997). Thus, although the majority of persons with nonepileptic seizures do not have concurrent epilepsy, those who do have a concurrent bona ἀde seizure disorder, as illustrated by this case, can have a risk of SUDEP associated with withdrawal of antiepileptic drugs after the observation of multiple nonepileptic events during long-term EEG monitoring. The physician must consider all nuances of the history from the patient and the immediate family, as well as all prior EEG data before having conἀdence that the antiepileptic medication can be safely withdrawn, since discontinuation or change in antiepileptic medication is one of the acknowledged risk factors for SUDEP (So 2008). Most epileptologists have had the experience of treating patients with mixed diagnoses and are well aware of the importance of caution when considering the discontinuation of anticonvulsants. Case 2â•… A Case o f No nco mpliance as a SUDEP and Lit ig at io n Risk A 22-year-old woman who had a history of generalized tonic–clonic seizures since age 13 years was found dead in bed by her mother. Although she had auras consisting of a “strange feeling” before most seizures, there was no known history of complex partial seizures. An EEG and computed tomography scan shortly after the initial diagnosis were normal. The patient was placed on phenytoin with therapeutic levels, and after 3 years of being seizure free, the drug was tapered. A few months later, she had a recurrent seizure, was started on carbamazepine, and maintained reasonably good seizure control with no more than one seizure per year for over 5 years. The patient was seen gratis every 6 months by her neurologist, as she had no health insurance, with interval carbamazepine blood levels being within the therapeutic range. However, for the last couple of years before her death, she had only sporadically kept scheduled follow-up appointments. She did see her neurologist several days before her demise at which time a carbamazepine blood level was obtained and found to be within the therapeutic range previously maintained. A postmortem performed at the medical examiner’s office, including toxicology studies, found no cause of death. In contrast to the premortem blood levels, the postmortem carbamazepine level was less than 1.5 μg/mL, i.e., below the limit of laboratory measurability. The patient had been under stress recently as she had lost her job working in a pizza restaurant. The neurologist was sued by the patient’s family on the grounds the she had not been treated appropriately for her seizure disorder, resulting

SUDEP Clinical Case Histories

855

in her death. During the trial, which occurred over half a decade later, the patient’s primary care physician, who was also a close friend of the family, admitted that, early on, he had advised the family that the patient was not properly managed, resulting in her unexplained death. (Nonetheless, the primary care physician continued to refer patients to the defendant!) An epilepsy specialist, testifying as a defense expert, explained to the jury that sudden unexplained death can occur in association with a diagnosis of epilepsy, and that although the reason for its occurrence is unknown, a possible risk factor is the patient not being compliant with taking prescribed antiepileptic medication. Although the jury found the neurologist not guilty of any malpractice, the emotional toll of dealing with this prolonged legal process resulted in his leaving neurology practice and taking a nonclinical employment. Co mment s This case illustrates several issues that are associated with SUDEP. First, the patient had generalized tonic–clonic seizures, which is the seizure type overwhelmingly associated with a risk of SUDEP (Walczak et al. 2001). However, although most studies indicate the increased risk for SUDEP in association with increased seizure frequency, this patient demonstrated that although the seizures occurred infrequently, having a diagnosis of epilepsy, even with infrequent seizures, entails a risk for SUDEP. Second, the fact that the patient had postmortem evidence that indicated the likelihood of an apparent precipitous drop in her carbamazepine blood level to one that essentially indicated that there was none measurable indicates that acute noncompliance with taking her antiepileptic drug was a likely factor in producing a fatal seizure. While there is controversy relative to the reliability of using subtherapeutic anticonvulsant levels as a measure of compliance (Opeskin et al. 1999), that of postmortem carbaÂ� mazepine levels is better than for phenytoin levels (Tomson et al. 1998). Third, there is an evolving legal climate in which SUDEP cases can be the subject of litigation (Wanamaker, this book, Chapter 24). While the outcome of the litigation resulting from this death was appropriate from a legal and medical standpoint, the net result was to convince a caring and highly competent neurologist that the stresses associated with having to deal with a prolonged, complex legal process that had no basis in fact, made continuing in clinical practice no longer tenable, resulting in now being employed in a nonclinical position. Case 3â•… A Case o f Increasing S eiz ure Frequency , Chang e in Medicat io n, and Emo t io nal St ress A 20-year-old university student was found dead in bed. He had a history of generalized tonic–clonic seizures since age 10 years, although later there were also events during which he experienced partial disturbances of awareness after which he remained fatigued for the rest of the day. Imaging studies were unremarkable and the interictal EEGs showed some generalized epileptiform activity. The patient was placed on valproic acid with considerable improvement in his seizures. After he started attending university, seizure control worsened with generalized tonic–clonic events occurring as frequently as once per month. Nonetheless, with the support of

856 Sudden Death in Epilepsy: Forensic and Clinical Issues

his family and friends, he was able to complete his ἀrst year of studies with excellent grades. On several occasions not associated with seizure occurrence, he complained of feeling his heart beating rapidly for no apparent reason. The family physician thought that he might be having cardiac dysrhythmia; however, a baseline electrocardiogram was unrevealing, with the symptoms being attributed to anxiety. Although the patient continued doing well in school, despite increased seizure frequency to several generalized tonic–clonic seizures per month, he was increasingly withdrawn and depressed, in addition to being anxious about the prospect of not being able to maintain an independent lifestyle because of worsening seizure control. The patient’s neurologist, an internationally respected epilepsy specialist, decided to change his medication and started to carefully transition him over several weeks to treatment with carbamazepine. After a therapeutic level of carbamazepine was reached, the valproic acid dose was slowly diminished. In the midst of the medication transition, the patient died in his sleep. An autopsy found no cause for the death. Toxicological data were not available. Co mment s This case illustrates that increased frequency of generalized tonic–clonic seizure occurrence is a factor associated with increased risk of SUDEP. The patient was also in the midst of a change from one antiepileptic drug to another, using an appropriÂ� ate€gradual transition. Nonetheless, despite the obvious need to do so in this case, this case illustrates that change in antiepileptic medication, even when done gradually under supervision of a epileptologist, is another associated risk factor for SUDEP (Tomson et al. 2005; Walczak et al. 2001). The history of the patient having had several symptomatic events that were strongly suggestive of episodes of tachycardia raises the possibility that there may have been an underlying predisposition to tachyarrhythmia. The normal ECG does not exclude the possible diagnosis of an inherited cardiac arrhythmogenic sodium channelopathy (Brugada et al. 2003; Priori et al. 2003; Herreros, this book, Chapter 19). Such symptoms in anyone with epilepsy would be an indication for a cardiac evaluation with the goal of identifying individuals who are at risk for ictal aggravation of cardiac predisposition to potentially fatal dysrhythmias. In particular, determination of a genetic proἀle of predisposition for cardiac arrhythmias, e.g., Brugada and Long QT syndromes, would be reasonable (Herreros, this book, Chapter 19). Case 4â•… A Case o f Increased Seiz ure Frequency , Palpit at io ns, and Wo rry A 29-year-old man with a 2-year history of generalized tonic–clonic epilepsy was found dead in the family living room one evening where he had been watching latenight television. His seizures started several months after he had sustained a head injury with loss of consciousness for several minutes while working on an oil rig. Medical evaluation determined that he had no evidence of neurological sequellae. He returned to work later that week. Four months later, without warning, he had a generalized tonic–clonic seizure while at work. Magnetic resonance imaging and an EEG

SUDEP Clinical Case Histories

857

were unremarkable. He was placed on valproic acid but had another seizure several weeks later. He was laid off from his well-paying job as it was considered to be a too inherently dangerous work environment for someone with seizures. Despite attempts with dose modiἀcation and use of different antiepileptic medications, the patient continued to have at least one seizure per month. His wife noted that he was becoming more anxious and depressed about his inability to make a living for his family and told his physician that he was also very frustrated with the fact that, despite his being very compliant about taking his medications daily and on the prescribed schedule, the seizures persisted. An autopsy was not performed. Co mment s In a paradoxical sense, despite being compliant with taking his antiepileptic medication, the patient was becoming increasingly frustrated and emotionally distressed by the persistence of seizures. While there were not enough medical data available to the author to know what the blood levels were, we assume that they were therapeutic as the patient was being followed by a physician and historically had several interval blood tests without subsequent adjustment of dosage. The other obvious source of stress and depression was the fact that this previously independent person had, until the onset of his seizure disorder, a well-paying job and was able to support his family, which was not the case at the time of his demise. As is discussed by Lathers and Schraeder (2006), the potential risk of operant psychological issues should be considered in any patient with epilepsy. In contrast to the extensive interest in the role of psychological factors as a risk for sudden death in cardiovascular disease (Sher 2006), very little investigation has been undertaken looking at the role that such factors may have in SUDEP. Case 5â•… A Near Miss A 31-year-old female with a recent diagnosis of complex partial seizures was started on phenytoin. Her imaging studies were unremarkable and the EEG had left temporal interictal sharp waves. The seizure frequency improved once a therapeutic phenytoin level was achieved, but several weeks later she appeared at her neurologist’s office complaining of a facial rash and joint pains. A rheumatology consultation conἀrmed that the patient most probably had an acute lupus-like reaction to phenytoin. The patient was admitted to the hospital, started on carbamazepine, and the phenytoin was rapidly tapered. It was noted that the patient was becoming increasingly confused and withdrawn, progressing to a comatose state. An EEG manifested continuous generalized irregular spike and wave and sharp wave activity consistent with electrographic status epilepticus. The patient was placed in the intensive care unit and given intravenous (IV) diazepam without improvement in the EEG. She subsequently started having recurrent generalized tonic–clonic seizures. As this was a€time when the only other available IV antiepileptic drugs were phenobarbital and pentobarbital, the patient was intubated and IV phenobarbital was administered.€A€chest x-ray at the time of intubation was unremarkable. Once the phenobarbital level exceeded 40 μg/ mL, the generalized clinical seizures were controlled. The EEG, however, continued

858 Sudden Death in Epilepsy: Forensic and Clinical Issues

to manifest continuous generalized electrographic status epilepticus. Because the patient’s arterial pO2 was dropping, a repeat chest x-ray revealed the presence of diffuse pulmonary inἀltrates consistent with acute respiratory distress syndrome. No infection was found. The intensive care unit staff did not consider that aspiration was a likely explanation for the pulmonary changes as the patient was appropriately intubated before the changes occurred, which would have prevented aspiration. Over the next several days, with continued dosage of IV phenobarbital, the patient’s electrographic seizures ceased and the dose of phenobarbital was lowered to the point where the patient could function. The chest x-ray continued to manifest pulmonary inἀltrates and she was eventually discharged with a pO2 of 60 while using a supplementary oxygen source. Her pulmonary problem eventually improved to the point where supplementary oxygen was not needed. The phenobarbital continued to control her seizures and the lupus syndrome disappeared. Co mment s Although very rare (Chang and Sher 1991), the occurrence of an idiosyncratic side effect in the form a lupus-like syndrome secondary to phenytoin. Although this syndrome is milder than idiopathic systemic lupus erythematosis and does not usually involve the central nervous system, its occurrence is a reason to withdraw the drug. The subsequent seizures and status epilepticus were in all likelihood the result of transitioning the patient from phenytoin to carbamazepine. More relevant to the subject of this book, this case illustrates that severe respiratory compromise from neurogenic pulmonary changes can occur in association with seizures. The degree of severe compromise of pulmonary function in this case indicates that in addition to the possibility of central apnea (Nashef et al. 1996), potentially fatal acute seizurerelated pulmonary pathology can factor in SUDEP (Simon et al. 1982; Jehi and Najm 2008). Were the patient not in the intensive care unit, she could well have been a victim of unexplained death secondary to severe neurogenic pulmonary edema in association with a history of epilepsy rather than a near miss.

52.2â•…Discussion There are a number of obvious as well as subtle concerns raised by these cases around the issue of risk factors for SUDEP. The commonly associated factors include occurrence of generalized tonic–clonic seizures with poor control, poor compliance with subtherapeutic antiepileptic drug level, changes in antiepileptic drugs, being a young adult male, and early onset vs. late onset epilepsy (Nilsson et al. 2001; Walczak et al. 2001). Psychological issues involving stress can also be applied as a possible albeit less tangible risk factor (Lathers and Schraeder 2006). Only cases 2 and 3 had an autopsy, which illustrates the reality that there is underutilization of postmortem examination in cases of epilepsy-related death (Schraeder et al. 2006, 2009). The issue of the pitfalls of postmortem toxicological data is also illustrated. In case 1, while the nil postmortem level of carbamazepine in the face of a therapeutic level 48 hours before death emphasizes the importance of obtaining toxicological data, including antiepileptic drug level, the accuracy of such levels is still an unresolved

SUDEP Clinical Case Histories

859

issue (Tomson 2010, this book, Chapter 51). In contrast to phenytoin, the postmortem levels of carbamazepine have been found to be a more relatively reliable indicator of compliance (Tomson et al. 1998). On the other hand, although a postmortem was obtained in case 3, the lack of toxicological data indicates the importance of consistent postmortem protocols (Schraeder et al. 2009). In addition, two of the cases, 3 and 5, were involved in changing antiepileptic drugs. In one case, the transition was appropriately gradual, whereas in the other, it was presumably rapid and in all likelihood resulted in induction of a fatal seizure. All of the ἀve cases manifested generalized tonic–clonic seizures. In case 1, there were only some temporal lobe symptoms as an aura before the generalized seizures occurred, with no history of clinical complex partial seizures. In case 3, the predominant seizures were generalized tonic–clonic; however, the family gave a history suggesting complex partial events. In case 5, the primary diagnosis was that of complex partial epilepsy, with transition to generalized electrographic and, ultimately, generalized tonic–clonic status epilepticus. The possible role of psychological factors in progression to SUDEP has not been studied in any meaningfully systematic way (Lathers and Schraeder 2006). However, in four cases, there were indications of a recent increase in adverse psychosocial issues. In case 1, the patient had recently been told that her seizures were not real. In case 2, the patient had recently lost her job. Case 3 illustrates the frustration of an intelligent young university student when facing the prospect of a life of dependence rather than independence. The young man in case 4 was increasingly anxious and depressed about losing his job because of his seizures and a lack of seizure control despite every effort to comply with his prescribed treatment. The medical legal aspect of SUDEP is illustrated in case 2 in that the neurologist was sued over the unexpected death. Although the court system determined that the neurologist was not liable, the emotional toll on the defending neurologist resulting from a litigation process that lasted over half a decade, especially in the context of knowing that the best of care was made regularly available despite inability of the patient to pay, resulted in withdrawal of a caring, compassionate, and highly competent physician from clinical practice. Finally, some mechanistic affirmation is illustrated by case 5 in that the complex autonomic afflictions that can be associated with seizures can include not only cardiovascular disturbance, but also acute potentially fatal pulmonary changes. In summary, within only 5 patients described, there is documentation of many of the multiple and complex interactions that are extant in the ongoing effort of trying to get a handle on how to minimize the occurrence of SUDEP.

References Betts, T. 1997. Chapter 199: Psychiatric aspects of nonepileptic seizures. In Epilepsy: A Comprehensive Textbook, ed. J. J. Engel and T. A. Pedley. Philadelphia, CA: Lippincottt-Raven. Brugada, J., R. Brugada, and P. Brugada. 2003. Determinants of sudden cardiac death in individuals with the electrocardiographic pattern of Brugada syndrome and no previous cardiac arrest. Circulation 108 (25): 3092–3096. Chang, D. K. M., and N. H. Sher. 1991. Cutaneous reactions to anticonvulsants. Semin Neurol 12: 329–336. Jehi, L., and I. M. Najm. 2008. Sudden unexpected death in epilepsy: Impact, mechanisms, and prevention. Cleve Clin J Med 75 (Suppl 2): S66–S70.

860 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300. Nilsson, L., U. Bergman, V. Diwan, B. Y. Farahmand, P. G. Persson, and T. Tomson. 2001. Antiepileptic drug therapy and its management in sudden unexpected death in epilepsy: A case-control study. Epilepsia 42 (5): 667–673. Opeskin, K., M. P. Burke, S. M. Cordner, and S. F. Berkovic. 1999. Comparison of antiepileptic drug levels in sudden unexpected deaths in epilepsy with deaths from other causes. Epilepsia 40 (12): 1795–1798. Priori, S. G., P. J. Schwartz, C. Napolitano, R. Bloise, E. Ronchetti, M. Grillo, A. Vicentini, et al. 2003. Risk stratiἀcation in the long-QT syndrome. N Engl J Med 348 (19): 1866–1874. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Sher, L., ed. 2006. Psychological Factors and Cardiovascular Disorders. New York, NY: Nova Biomedical Books. Simon, R. P., L. L. Bayne, R. F. Tranbaugh, and F. R. Lewis. 1982. Elevated pulmonary lymph flow and protein content during status epilepticus in sheep. J Appl Physiol 52 (1): 91–95. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (Suppl 9): 93–98. Tomson, T., A. C. Skold, P. Holmgen, L. Nilsson, and B. Danielsson. 1998. Postmortem changes in blood concentrations of phenytoin and carbamazepine: An experimental study. Ther Drug Monit 20 (3): 309–312. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61. Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. D. Cascino, J. F. Annegers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56 (4): 519–525.

Cardiac Antiarrhythmic Agents Pharmacological Basis for Their Antiarrhythmic and Proarrhythmic Effects

53

Saumya Sharma Trieu Ho Bharat K. Kantharia

Contents 53.1 Basic Electrophysiologic Principles 53.2 Antiarrhythmic Drugs 53.2.1 Class Agents 53.2.2 Class II 53.2.3 Class III 53.2.4 Class IV 53.2.5 Miscellaneous Antiarrhythmic Drugs 53.3 Proarrythmia 53.4 Epilepsy and Proarrhythmia 53.5 Conclusion References

862 864 864 868 868 870 871 872 873 873 873

Cardiac electrical activity requires the movement of ions through ionic channels across the cell membrane of cardiac myocytes. Although all cardiac myocytes are capable of generating and conducting electrical activity, there are specialized structures in the heart that are responsible for impulse generation (e.g., sinus node) and impulse conduction (e.g., atrioventricular [AV] node and the His–Purkinje system). The functional properties of cardiac ion channels may be altered or modulated by many pharmacological agents. While druginduced modulation of cardiac electrical activity may be used to treat cardiac arrhythmias, it can also compromise conduction, amplify dispersion of refractoriness, and induce triggered activity, resulting in the development of cardiac arrhythmias (i.e., proarrhythmia). The purpose of this chapter is to review the pharmacology of drugs that alter cardiac electrophysiologic properties and highlight drugs commonly used to treat cardiac arrhythmias. Most of these drugs are classiἀed as per Vaughan Williams (1992) classiἀcation. Although certain drugs that are known for their proarrhythmic properties are discussed, a comprehensive review of drugs that cause proarrhythmia is beyond the scope of this chapter.

861

862 Sudden Death in Epilepsy: Forensic and Clinical Issues

53.1â•… Basic Electrophysiologic Principles The movement of Na+, K+, Ca2+, and Cl− ions across the cell membrane is what generates electrical activity in cardiac myocytes. Ionic channels are large proteins that span the cell membrane of cardiac myocytes and allow the flux of speciἀc ions passively “down” their electrochemical gradients. This is in contrast to ionic pumps, which counteract these downhill fluxes by “pumping” ions against their electrochemical gradient usually coupled with the hydrolysis of adenosine triphosphate (ATP) (e.g., Na-K-ATPase). Ionic channels may be ligand gated, voltage gated, or changed by mechanical stress. Conformational changes occur in these ionic channels that allow the channel to be open, closed, or inactivated. Certain drugs predominantly act on an ionic channel in only a speciἀc state (i.e., lidocaine acts on the Na+ channel in the activated state), which has important effects on the drug actions. There are ἀve phases of the cardiac action potential (Figure 53.1). The resting membrane potential of cardiac myocytes (phase 4) is between −50 and −95 mV depending on the cell type. Outward K+ current mediated by the inward rectifying K channel (IKl) is largely responsible for the resting membrane potential. Na-K-ATPase is an electrogenic pump that also contributes to the resting membrane potential by maintaining high extracellular Na+ concentrations and high intracellular K+ concentrations. Cardiac glycosides inhibit Na-K-ATPase, and the subsequent increase in intracellular Na+ is exchanged for Ca2+ through the Na-Ca exchanger. The subsequent increase in intracellular Ca2+ increases the contractility of cardiac myocytes. Phase 0 of the cardiac action potential is characterized by a sudden rapid increase in the membrane potential. This is an “all or none” response to a depolarizing stimulus, which reaches a threshold value of −70 to −65 mV. This rapid upstroke in the action potential is due to an increased Na+ conductance through voltage-gated fast Na+ channels and is seen predominantly in atrial, ventricular, and His-Purkinje cells. In contrast, sinus node and AV nodal tissue have very slow upstrokes, which are mediated by a slow inward Ca2+ current (Ica). This slow current mediated by Ca 2+ channels (IcaL) can be phosphorylated by cyclic adenosine monophosphate (cAMP)–dependent protein kinases to enhance conductance. In this manner, drugs such as β agonists (e.g., albuterol) and phosphodiesterase inhibitors (e.g., theophylline) increase cAMP concentrations and improve sinus node and Phase 1 Phase 2 Ito I Na/Ca ICa

+ 40

IKp

IK

Phase 3

Phase 0 Vm (mV)

INa IK1

–80

Phase 4 INa and INa/Ca

Ipump IK1

Figure 53.1╇ Ionic movements during phases of the cardiac action potential.

Cardiac Antiarrhythmic Agents

863

AV nodal function. On the other hand, among the Vaughan Williams (1992) classiἀcation of antiarrhythmic agents (Table 53.1), class I antiarrhythmic drugs block voltage-gated Na+ channels in atrial, ventricular, and His-Purkinje tissues with minimal effect on normal sinus node and AV node function. In phase 1, voltage-gated Na+ channels are inactivated, whereas transient outward K+ current (Ito) is activated. Interestingly, regional heterogeneity in Ito channel density may play a role in the genesis of ventricular arrhythmias. In phase 2 (plateau phase), there is an exchange between outward K+ and Cl− currents and inward Ca2+. The outward delayed rectiἀer K+ current, either a rapid (IKr) or a slow (IKs) component, is mediated by different K+ channels. Mutation in the K+ channel genes mediating IKr cause long QT syndrome 1, whereas long QT 2 syndrome is caused by mutations in the K+ channel modulating IKs. Phase 3 of the action potential is the phase of rapid repolarization. Slow deactivation of IcaL along with activation of both IKr and IKs results in a net decrease in intracellular positive charge resulting in rapid repolarization. Drugs such as erythromycin, quinidine, and terfenadine prolong repolarization by exerting an effect on IKr (observed on electrocardiogram [ECG] as a prolonged QT interval) and predispose to the development of torsades de pointes. As mentioned earlier, there are specialized cardiac structures responsible for impulse generation and conduction. The sinus and AV nodes have different electrophysiologic properties as compared to other cardiac tissues. These specialized tissues display spontaneous phase 4 depolarization. In other words, the resting membrane potential in these tissues gradually depolarizes and can reach the threshold by itself. This is responsible for normal automaticity. The discharge rate of the sinus node is faster than other cells with Table 53.1â•… Vaughan Williams Classification of Antiarrhythmic Drugs Class I – IA I – IB

Electrophysiologic Properties Sodium channel blockers: moderately slow conduction, moderately prolonged duration of action potential Sodium channel blockers: minimally slowed conduction, shortened duration of action potential

I – IC

Sodium channel blockers: markedly slowed conduction, minimal duration of action potential

II

β blockers

III

Potassium channel blockers

IV

Calcium channel blockers

Source: Vaughan Williams, E.M., J Cardiovasc Pharmacol 20 (Suppl 2), S1–S7, 1992.

Examples Quinidine Procainamide Diisopyramide Lidocaine Mexiletine Tocainamide Phenytoin Flecainide Encainide Propafenone Atenolol Metoprolol Acebutalol Pindolol Labetolol Carvedilol Amiodarone Bretylium Sotalol Ibutilide Verapamil Diltiazem

864 Sudden Death in Epilepsy: Forensic and Clinical Issues

automaticity and is thus the main pacemaker of the heart. The action potential of sinus and AV nodal cells is also different, displaying a slower upstroke of phase 0, which is mediated by slow Ca 2+ conductance rather than rapid Na+ conductance. The sinus node, AV node, and His bundle region are innervated with sympathetic and parasympathetic nerve ἀbers, allowing modulation of automaticity and AV conduction with varied physiologic conditions (e.g., exercise or sleep).

53.2â•… Antiarrhythmic Drugs The Vaughan Williams classiἀcation (class I–IV) is the most widely used classiἀcation system for antiarrhythmic drugs (Table 53.1) (Vaughan Williams 1992). Drugs are classiἀed according to which ionic channel or receptor they block in the heart. This classiἀcation system is limited because it is based on the effect of drugs in normal cardiac tissue at arbitrary concentrations. The mechanism of action of antiarrhythmic drugs is complex, depending on the tissue type, presence of disease, heart rate, membrane potential, and electrolyte balance. Furthermore, many antiarrhythmic drugs exert their effects on multiple ion channels and receptors or have active metabolites that exert different electrophysiologic properties. Most antiarrhythmic drugs demonstrate use dependence, which means that they act predominantly at faster heart rates or predominantly after depolarization (e.g., lidocaine). This may result from a drug binding greater to an ionic channel in the activated state. On the other hand, some drugs, such as sotalol, display reverse use dependence. These drugs exert their effects predominantly at slow heart rates when an ionic channel is in the resting state. Certain antiarrhythmic drugs demonstrate stereoselectivity. In other words, most drugs are a 50:50 mixture of two racemates (mirror-image structure), and one racemate may have different electrophysiologic properties than the other (e.g., l -sotalol has no β blocking properties). There are also genetically determined differences in the metabolism of certain antiarrhythmic drugs that account for differences in the response or side effects (e.g., slow acetylators are more likely to develop drug-induced lupus with procainamide). Proarrhythmia constitutes the most serious problem with antiarrhythmic drug therapy. As many as 5–10% of patients on antiarrhythmic drugs develop proarrhythmia (Naccarelli et al. 2001). Prolongation of repolarization, the development of triggered activity, and alterations in reentry pathways constitute the mechanism for proarrhythmia. Congestive heart failure, depressed left ventricular function, diuretic or digitalis use, and long pretreatment QT interval are risk factors for the development of proarrhythmia. 53.2.1â•… Class Agents These drugs block the fast voltage-gated Na+ channel and thus affect phase 0 of the action potential. This class of drugs is subdivided onto three subgroups based on the magnitude of their effect on cardiac conductance. Class 1A drugs have an intermediate effect on the Na+ channel and prolong the action potential duration. On the other hand, class 1B drugs have a minimal effect on Na+ channel in normal tissue but cause signiἀcant prolongation of conductance in depolarized tissues. Class 1C drugs cause signiἀcant block in the Na+ channel and can prolong conduction enough to be manifested on the ECG with prolongation of the QRS complex.

Cardiac Antiarrhythmic Agents

865

Class 1A drugs are quinidine, procainamide, and disopyramide. Although they demonstrate an intermediate effect on Na+ channel blockade, they prolong the action potential duration by blocking the K+ current IKr. Because of the risk of proarrhythmia, all class 1A drugs are contraindicated in patients with structural heart disease (i.e., myocardial infarction, cardiomyopathy, left ventricular hypertrophy) (Roden 1998). The only exception is disopyramide in patients with hypertrophic cardiomyopathy. Quinidine increases the threshold of excitability, prolongs conduction time, and decreases automaticity. It also has vagolytic and α-adrenergic receptor blocking activity and thus can cause hypotension by decreasing peripheral vascular resistance. Quinidine is highly bound to the plasma protein α1-acid glycoprotein, which increases in congestive heart failure. It is metabolized in the liver and 20% is excreted unchanged in the urine via P-glycoprotein (i.e., decrease digoxin clearance and increase digoxin levels). Congestive heart failure, liver disease, and renal disease increase quinidine concentrations (Grace and Camm 1998). Although quinidine is rarely used today, it can be used to control both atrial and ventricular arrhythmias. It is effective in converting atrial ἀbrillation or atrial flutter to normal sinus rhythm. When utilizing class 1 drugs for atrial flutter or ἀbrillation, it is important to use AV nodal blocking agents for rate control in conjunction because atrial rate may slow down or organize enough to be conducted 1:1 through the AV node. The limitations to quinidine use are its adverse effects. Common adverse effects are nausea, vomiting, abdominal pain, and Cinchonism (tinnitus, hearing loss, delirium). Quinidine can also cause some idiosyncratic reactions such as thrombocytopenia from quinidineplatelet antibodies, hemolytic anemia, and anaphylaxis. However, the most serious side effect of quinidine is the development of torsade de pointes. Torsade de pointes due to quinidine is not related to higher plasma concentrations. A meta-analysis of trials utilizing quinidine demonstrated increased mortality compared to placebo, presumably because of increased proarrhythmia (Grace and Camm 1998). Procainamide blocks INa as well as IKr. It is less anticholinergic compared to other class 1A drugs. Procainamide is metabolized to N-acetyl procainamide, which has different electrophysiologic properties than its parent compound, having no effect on Na+ conductance. N-acetyl procainamide blocks IKr and prolongs repolarization similar to class III antiarrhythmic drugs. Procainamide is metabolized in the liver to N-acetyl procainamide and excreted in the urine; therefore, N-acetyl procainamide levels need to be monitored in patients with depressed creatinine clearance. The rate of metabolism is genetically variable with over one-half of the population being rapid acetylators. Procainamide, like quinidine, can also be used for atrial and ventricular arrhythmias. It is effective in blocking conduction through accessory pathways and is one of the drugs of choice in patients with Wolff– Parkinson–White syndrome who present with atrial ἀbrillation. Procainamide is more effective than lidocaine in terminating sustained monomorphic ventricular tachycardia (Gorgels et al. 1996). Procainamide is also utilized as a diagnostic tool used to uncover Brugada type 1 ECG pattern in the electrophysiology laboratory (Brugada et al. 2000). Procainamide is better tolerated than quinidine. Procainamide-induced lupus characterized by fever, arthralgia, Raynaud’s phenomenon, pleuritis, pericarditis, pericardial effusion, and/or hepatomegally occurs in slow acetylators. It is associated with antihistone antibodies and is usually reversible. Agranulocytosis can also occur, and complete blood counts should be monitored periodically in patients on procainamide. Finally, prolongation of the QT interval can occur with risk for torsade de pointes due to accumulation of N-acetyl procainamide.

866 Sudden Death in Epilepsy: Forensic and Clinical Issues

Disopyramide has similar electrophysiologic actions as quinidine and procainamide. However, disopyramide has potent antimuscarinic action and negative inotropic effect on the myocardial contractility. About 50% of the drug is excreted unchanged and thus renal failure can increase plasma concentrations. The main use of disopyramide is to control atrial ἀbrillation in patients with hypertrophic cardiomyopathy. Because of its antimuscarinic effects, dry mouth, constipation, urinary retention, blurred vision, and closed angle glaucoma can occur. Disopyramide can depress myocardial contractility in patients with left ventricular systolic dysfunction to the extent of causing cardiovascular collapse. The class 1B antiarrhythmics are lidocaine, mexilitine, and phenytoin. They block the rapid Na channel with a rapid time constant of recovery from block. They decrease the slope of phase 4 diastolic depolarization and therefore increase the threshold of excitability and reduce automaticity. Their effects are greater at very fast heart rates or in depolarized tissue because they block the Na channel in the activated state. Lidocaine predominately affects Purkinje tissue and ventricular muscle, having little effect in the atrium or accessory pathways. It can suppress afterdepolarizations and decrease conduction in ischemic tissue, thereby preventing the development of ventricular ἀbrillation. However, in patients with sinus node dysfunction or His-Purkinje system disease, lidocaine can suppress or worsen sinus node automaticity and AV conduction, or suppress junctional or ventricular escape rhythms. Lidocaine is metabolized in the liver and must be administered parenterally due to extensive ἀrst-pass metabolism. When there is liver disease or decreased blood flow to the liver (e.g., congestive heart failure), lidocaine levels may accumulate. It is also bound in serum to α1-acid glycoprotein, which increases in congestive heart failure. Lidocaine is the indicated treatment of ventricular arrhythmias particularly in the setting of myocardial infarction. However, it is not indicated in post– myocardial infarction prophylaxis of ventricular arrhythmias because randomized trials have shown an increase in mortality (Sadowski et al. 1999). The most common adverse effects are dose-related central nervous system toxicity manifested by dizziness, parasthesia, confusion, delirium, and seizure. Occasionally, sinus node suppression and AV conduction disturbances due to His-Purkinje block may be seen. Plasma lidocaine levels should be monitored in patients with myocardial infarction and shock or severe ventricular dysfunction. Mexilitine is similar in structure and function to lidocaine, except that it can be adminÂ� istered orally. It is metabolized in the liver by the cytochrome P450 system.€Plasma€levels are increased in liver disease and with the concomitant administration of rifampicillin, phenytoin, and cimetidine. Mexilitine is not widely used as a single agent; however, it can be used in combination with other anitarrhythmic drugs (e.g., amiodarone, procainamide, disopyramide) for synergistic effects. Mexilitine has been shown to have some efficacy in patients with long QT 3 syndrome (Schwartz et al. 1995). Nausea, stomach cramps, tremor, dizziness, dysarthria, parasthesia, blurred vision, and ataxia are common side effects. Phenytoin is a drug commonly used for seizures. It has antiarrhythmic properties and has been shown to abolish automaticity in Purkinje tissue due to digitalis-induced delay in afterdepolarization. Therefore, phenytoin is indicated in digitalis-induced atrial and ventricular arrhythmias. Short-term side effects include nystagmus, ataxia, drowsiness, stupor, and coma. There are many long-term side effects of phenytoin, which are seen predominantly in patients using this drug to control seizures.

Cardiac Antiarrhythmic Agents

867

Class 1C drugs are flecainide and propafenone. They block the fast Na+ channel with a slow time constant for recovery from block. They also block K+ channels. They prolong action potential duration at faster heart rates in the atrium, making them useful in treatment of atrial ἀbrillation. The landmark Cardiac Arrhythmia Suppression Trial (CAST) and other subsequent trials showed that class 1C drugs, when used in the setting of structural heart disease, increase the risk of ventricular proarrhythmia (Roden 1998; Kuck et al. 2000; Wyse et al. 2001). Flecainide demonstrates marked-use-dependent block of the rapid Na+ channel. Although this effect is greater at faster heart rates, marked drug effects can occur even at physiologic heart rates. The Na+-channel-blocking effect of flecainide also vary depending on the type of cardiac tissue and the presence of disease. The electrical heterogeneity in the setting of structural heart disease promotes proarrhythmia. Flecainide only minimally prolongs the QT interval. Flecainide decreases myocardial systolic function and should not be used in patients with moderate or severe ventricular systolic dysfunction. It is administered orally and is excreted in the kidney unchanged (concentrations may increase in the setting of renal failure). Flecainide is used for the treatment of supraventricular tachycardia, paroxysmal atrial ἀbrillation, or atrial tachycardia. It is useful not only in preventing atrial ἀbrillation but also in acute conversion of atrial ἀbrillation to sinus rhythm. Highdose flecainide can be used to acutely convert atrial ἀbrillation to sinus rhythm in patients with rare episodes of paroxysmal atrial ἀbrillation using a “pill in the pocket” approach in which the patient takes the drug only when necessary and if there is no spontaneous conversion of the rhythm to sinus rhythm in a reasonable time as deemed appropriate by the patient (Alboni et al. 2004). The main adverse effect of flecainide is the ability to induce proarrhythmia. It may worsen conduction in patients with type II AV block or may cause sinus arrest in patients with sinus node dysfunction. In patients with structural heart disease, it may worsen ventricular arrhythmias and/or render these arrhythmias unresponsive to therapy. They increase mortality in patients with coronary artery disease and are contraindicated in this setting (Roden 1998). Some patients on flecainide complain of neurological side effects such as confusion or irritability due to its effect on the cortical cell membrane. Signiἀcant prolongation of QRS duration on a routine ECG is a sign of flecainide toxicity. Exercise stress tests should be done annually to monitor for exercise-induced proarrhythmia as well as for coronary artery disease. Propafenone has electrophysiologic action similar to flecainide, except that it is also a weak β blocker. It is absorbed orally and metabolized in the liver to 5-hydroxypropafenone, which has the same Na+ channel effects as the parent compound but no β blocker effects. There is marked genetic variability in the metabolism of propafenone. Approximately 7% of the general population are poor metabolizers for whom propafenone elimination is two to three times longer. These patients may develop bronchospasm from the β blocker properties of propafenone. The indications and contraindications of propafenone are the same as for flecainide. A single 600 mg dose of propafenone can be used as a pill in the pocket approach for the outpatient treatment of atrial ἀbrillation (Alboni et al. 2004). Propafenone can cause sinus node depression and AV block. Bronchospasm may occur in susceptible patients who are poor metabolizers of the drug. Propafenone commonly causes taste disturbance (i.e., metallic taste) and, although rarely, dizziness or blurred vision. The risk of proarrhythmia is thought to be less than flecainide. Nonetheless, all class 1C drugs are currently contraindicated in patients with structural heart disease.

868 Sudden Death in Epilepsy: Forensic and Clinical Issues

53.2.2â•… Class II Class II antiarrhythmics are drugs that block the β adrenergic receptor in the heart. There are two β receptors: β1 receptors, which are predominately located in the heart, and β2 receptors, which are located in the blood vessels and bronchi. Adrenergic neurotransmitters (e.g., norepinephrine) and hormones (e.g., epinephrine) bind to β1 receptors in the heart and activate signaling cascades leading to the production of cAMP from adenylate cyclase, which then activates cAMP-dependent protein kinases. These kinases phosphorylate the ICa-L channel and proteins that regulate calcium homeostasis in the myocyte (e.g., phosphlambin), ultimately leading to enhanced chronotropy and inotropy in the heart. β blockers antagonize the effect of the arrhythmogenic effects of catecholamines. They decrease sinus rate discharge and impair AV conduction in the heart. They have a modest effect on suppressing atrial and ventricular arrhythmias and increase the threshold of ventricular ἀbrillation particularly in ischemic myocardium. There are differences among β blockers. Some β blockers are cardioselective (predominately blocking β1 receptors), such as metoprolol, atenolol, carvedilol, and bisoprolol, while others in this class block both types of receptors, such as propanolol, timolol, and nadolol. Pindolol and acebutolol are two β blockers that demonstrate intrinsic sympathomimetic activity (i.e., partial agonists). Carvedilol and labetalol also block the α adrenergic receptors, making them potent vasodilators. Despite these differences, evidence from over a decade of studies on β blockers indicates that, as a class, they reduce the incidence of sudden cardiac death and prolong survival in patients with structural heart disease (Singh 1998). β blockers are indicated for both supraventricular and ventricular tachyarrhythmias. They are useful in controlling the ventricular rate in atrial ἀbrillation and atrial flutter. They may be used prophylactically to prevent atrial ἀbrillation in patients undergoing cardiac surgery (Connolly et al. 2003). A β blocker may terminate and prevent recurrence of AV nodal reentry tachycardia, AV reentry tachycardia utilizing an accessory pathway, and atrial trachycardias. In general, β blockers are useful in the treatment of ventricular arrhythmias associated with myocardial ischemia. Some idiopathic ventricular tachycardias from the right and left ventricular outflow tract may respond to β blockade. β blockers are usually well-tolerated drugs. However, some common adverse effects are hypotension, bradycardia, fatigue, worsening congestive heart failure, and bronchospasm. β blockers can worsen peripheral vascular disease, Raynuad’s phenomenon, and depression. In susceptible patients, β blockers can impair sexual function. β1 selective drugs may have a lower incidence of side effects. 53.2.3â•… Class III Class III antiarrhythmic drugs prolong action potential duration by lengthening the refractory period. These drugs primarily block the rapid component of the delayed rectiἀer K+ current (IKr), although some of the drugs in the class block other channels (e.g., amiodarone, ibutilide). Reentrant arrhythmias travel around a functional circuit for which the wavelength and the size of the circuit determine the excitable gap. For any reentrant arrhythmia to sustain itself, it must have an excitable gap of nondepolarized tissue. Prolongation of the action potential by class III drugs increases the wavelength and thereby diminishes or abolishes the excitable gap. However, prolongation of the refractory period increases the

Cardiac Antiarrhythmic Agents

869

QT interval, predisposing to torsades de pointes. Despite this proarrhythmic potential, class III drugs are safe to use in the setting of structural heart disease. Amiodarone is the most commonly used class III antiarrhythmic agent. It exhibits the electrophysiologic characteristics of all four classes of antiarrhythmic drugs. Because amiodarone contains two iodine moieties constituting 37% of the molecular weight, a single 200 mg dose of amiodarone contains 75 mg of iodine, which is >100 times the normal dietary intake. This has important effects of thyroid metabolism, leading to both hyperthyroidism and hypothyroidism. Furthermore, the actions of amiodarone are different when given acutely as an intravenous preparation compared to chronic oral administration. Amiodarone acutely inhibits Na+ and Ca 2+ currents in a use-dependent manner (i.e., faster heart rates and depolarized tissues). Inhibition of IKr is less prominent, and thus QT is minimally prolonged. AV nodal conduction is slowed due to β-blocking properties. Chronic administration of amiodarone results in inhibition of K+ currents (IKr and IKs); therefore, QT prolongation may be seen. AV nodal conduction is slowed and bradycardia may occur. There is decreased peripheral conversion of the T4 to T3, inhibition of T3 into myocytes, and binding of T3 to the nuclear receptors, which contributes to inhibition of Ito. Amiodarone has unique pharmacokinetics. It is highly lipid soluble and has a large volume of distribution. It accumulates in many tissues (e.g., fat, skin, liver, and lung). Several weeks of oral amiodarone administration are required to reach steady state (approximately 15 g). Elimination half life is longer than 30 days, and amiodarone levels can be detected in the blood even 9 months after discontinuation. It is metabolized in the liver to N-desethylamiodarone, which has similar electrophysiologic properties. The loading regimen for amiodarone can take up to 2–3 weeks, with up to 1600 mg/day in divided dose. Amiodarone is eliminated entirely in hepatic bile; therefore, no dose adjustment is needed for renal impairment. Amiodarone is approved by the Food and Drug Administration for the treatment of life-threatening ventricular arrhythmias. There are several trials demonstrating that amiodarone is effective in preventing cardiac arrest due to ventricular ἀbrillation (Randomized antiarrhythmic drug therapy in survivors of cardiac arrest [the CASCADE Study]. The CASCADE Investigators 1993; Doval et al. 1994; Cairns et al. 1997). A meta-analysis demonstrated improved mortality in patients treated with amiodarone after acute myocardial infarction or congestive heart failure. (Effect of prophylactic amiodarone on mortality after acute myocardial infarction and in congestive heart failure: Meta-analysis of individual data from 6500 patients in randomised trials. Amiodarone Trials Meta-Analysis Investigators 1997.) Amiodarone is effective in maintaining sinus rhythm in patients with atrial ἀbrillation. In fact, it is more effective than sotalol or propafenone (Roy et al. 2000). Amiodarone prophylaxis during cardiac surgery may prevent postoperative atrial ἀbrillation (Giri et al. 2001). The list of adverse effects from chronic amiodarone use is long. Almost every organ can be affected except for the kidney, and most adverse effects are dose dependent. Bradycardia from slowing sinus node conduction or worsening AV conduction is common. Pulmonary toxicity is a rare but potentially lethal side effect. It may present with a chronic interstitial pneumonitis or similar to the adult respiratory distress syndrome picture. Amiodarone pulmonary toxicity is difficult to diagnose with nonspeciἀc ἀndings on chest x-ray. Treatment of pulmonary toxicity consists of stopping amiodarone and supportive care. Hypothyroidism is very common but easily treated. Hyperthyroidism is rare

870 Sudden Death in Epilepsy: Forensic and Clinical Issues

but very difficult to treat. Stopping amiodarone does not immediately relieve thyrotoxicosis, and radioactive iodine cannot be used (the thyroid is saturated with iodine from the amiodarone). Thyroidectomy may be required. Blue–gray skin photosensitivity can occur, and sunscreen should be recommended to all patients on amiodarone. Corneal deposits, abnormal liver function tests, and peripheral neuropathy can occur but are reversible. Amiodarone can increase the deἀbrillation threshold of implantable cardiac deἀbrillators, and testing of the deἀbrillator should be done after several weeks on amiodarone. Amiodarone increases warfarin and digoxin levels, and doses of these drugs should be adjusted. Despite the effect on QT prolongation, the risk for torsade de pointes is low. This risk may be ampliἀed, however, if other QT prolonging drugs are given concomitantly. Patients on amiodarone should undergo semiannual monitoring of liver function tests, chest x-ray, amiodarone and N-desethylamiodarone levels, pulmonary function tests, and ophthalmological examination. Sotalol is a racemate of d-sotalol, which has pure class III electrophysiologic properties, and l -sotalol, which has additional β-blocking properties. Sotalol demonstrates reverse use dependence, having more profound activity at slower heart rates. It is administered orally and excreted in the urine unchanged. Doses need to be adjusted for renal insufficiency. Class II effects of sotalol do not start until the drug dose is 120 to 160 mg twice daily. Sotalol has been shown to decrease or prevent deἀbrillator shocks and can also lower the deἀbrillation threshold. Therefore, sotalol is considered a good ἀrst-choice drug to reduce deἀbrillator shocks in young patients. Sotalol can also be used to maintain sinus rhythm in patients with atrial ἀbrillation but is less effective than amiodarone or propafenone. The proarrhythmic risk of sotalol is considerable due to QT prolongation and torsade de pointes. Inpatient telemetric monitoring with daily ECGs to measure the QT interval for 48 to 72 h is recommended when starting sotalol. Bradycardia is common, and worsening exacerbation of congestive heart failure (due to the β blocker effect) can occur in patients with decompensated congestive heart failure. Dofetilide is a pure IKr blocking agent. This effect is more prominent in the atrium than in the ventricle. It prolongs the action potential duration by prolonging repolarization, and thus prolongs the QT interval. It is eliminated in the kidney, and dose needs to be adjusted for creatinine clearance. It is contraindicated in patients with a creatinine clearance of <20 ml/min. Like sotalol, initiation of dofetilide required inpatient hospitalization with telemetry and daily ECG. Special certiἀcation is required to administer dofetilide. Dose adjustment is required when QTc prolongation occurs greater than 50% from the baseline ECG. Dofetilide is indicated for the conversion of atrial ἀbrillation to sinus rhythm and the maintenance of sinus rhythm. It can be used in patients with structural heart disease. Dofetilide is well tolerated, and the major adverse effect is the risk for torsades de pointes. 53.2.4â•… Class IV Verapamil and diltiazem are the prototype of this class of antiarrhythmic agents. These drugs have electrophysiologic properties similar to β blockers but with important differences. They inhibit ICa-L , resulting in decreased slope of spontaneous phase 4 depolarization (reducing automaticity), and slow AV nodal conduction. They decrease the concentration of intracellular Ca2+ in myocytes, which may reduce triggered activity from early afterdepolarization. Reduction in intracellular Ca2+ also contributes to the signiἀcant negative

Cardiac Antiarrhythmic Agents

871

inotropic action of these drugs. Unlike β blockers, they do not directly counteract the effect of catecholamine excess and thus have no effect on reducing sudden cardiac death or ventricular arrhythmias. Verapamil and diltiazem are indicated for the control of ventricular rate in atrial ἀbrillation. They also are useful in terminating and preventing recurrences of atrial tachycardia, AV nodal reentry tachycardia, and AV reciprocating tachycardia utilizing an accessory pathway. Class IV drugs are not effective against ventricular arrhythmias, with the exception of some idiopathic ventricular tachycardia in patients with structurally normal hearts. These arrhythmias often originate from triggered activity in the right or left ventricular outflow tract or from left ventricular septum (e.g., verapamil-sensitive ventricular tachycardia). Calcium channel blockers can cause hypotension, bradycardia, asystole, and hemodynamic collapse, particularly in patients receiving β blockers as well. They should not be given to patients with advanced heart failure because of their negative inotropic effects. They should not be given in the setting of wide complex tachyarrhythmias for differentiating supraventricular tachycardia with aberrancy from ventricular tachycardia because of the signiἀcant risk of hemodynamic collapse. They are also contraindicated in atrial ἀbrillation and pre-excitation Wolff–Parkinson–White syndrome because of their potential to increase conduction over the accessory pathway and to induce ventricular ἀbrillation. 53.2.5â•… Miscellaneous Antiarrhythmic Drugs Adenosine is an endogenous purine nucleoside, which, when bound to its receptor on cardiac cells, activates K channels (IK Ach and IK Ado) similar to acetylcholine. This increase in K conductance results in hyperpolarized resting membrane potential in atrial tissue, AV node, and sinus node. Therefore, adenosine slows sinus node activity and blocks AV nodal conduction. Adenosine is rapidly removed from cardiac cells by degradation in the bloodstream or by reuptake into cells, resulting in very rapid elimination (1–6 s). Adenosine is administered as an intravenous bolus (6–18 mg). It is the ἀrst drug of choice for the termination of supraventricular tachycardias. It can induce transient block in the AV node in atrial flutter so that flutter wave can be recognized on the surface ECG. Although most ventricular arrhythmias are not affected by adenosine, rare idiopathic ventricular tachycardias originating from the outflow tract may respond. Approximately 10–15% of ectopic atrial tachycardias respond to adenosine. Transient flushing, chest pressure, and dyspnea occur with bolus administration. Transient complete AV block, sinus arrest, and symptomatic bradycardia are common occurrences. Because adenosine shortens the action potential duration in the atria, it may induce atrial ἀbrillation, which can be problematic for patients with Wolff–Parkinson–White syndrome. Digoxin is a cardiac glycoside that has been used for centuries for the management of congestive heart failure. Digoxin inhibits Na-K-ATPase, resulting in increased intracellular Na+ concentrations, which subsequently result in increased intracellular Ca levels through modulation of the Na–Ca exchanger. This increase in Ca2+ conductance into cardiac myocytes is responsible for the increased inotropic effects of digoxin. The electrophysiologic effects of digoxin result from augmentation of vagal tone, resulting in signiἀcant conduction slowing in the AV node. Therefore, the main use of digoxin is for ventricular rate control in atrial ἀbrillation or atrial flutter. Although other rate control agents (e.g., β blockers

872 Sudden Death in Epilepsy: Forensic and Clinical Issues

and Ca channel blockers) have gained favor over digoxin, it is still useful in patients with atrial ἀbrillation and congestive heart failure or when hypotension limits the use of other agents. It has no effects on the conversion of atrial ἀbrillation to normal sinus rhythm. Digoxin is also less useful in controlling the ventricular rate in atrial ἀbrillation when exercising. It can be administered orally or loaded intravenously for faster control of ventricular rate. It is excreted in the kidney unchanged and must be adjusted for renal failure. It has a narrow therapeutic window, and toxic dose can result in mental status changes, visual disturbances, nausea, vomiting, AV block, or sinus bradycardia. Because digoxin increases intracellular Ca2+ in atrial and ventricular myocytes, it can cause delayed afterdepolarizations. Delayed afterdepolarizations can induce arrhythmias characteristic to digoxin toxicity (i.e., atrial tachycardia with variable block, atrial ἀbrillation with junctional rhythm, bidirectional ventricular tachycardia). As previously mentioned, phenytoin can be used for digoxin-mediated proarrhythmia. However, digoxin-speciἀc antibody fragments should be given to treat life-threatening arrhythmias.

53.3â•… Proarrythmia Proarrhythmia is deἀned as the generation of new or the worsening of existing arrhythmias with drug therapy. The most common mechanism of proarrhythmia is prolongation of repolarization, which induces early afterdepolarizations and subsequent torsades de pointes. Torsades de pointes refers to a type of polymorphic ventricular tachycardia associated with a prolonged QT interval on ECG. Apart from the antiarrhythmic drugs, there is a whole variety of drugs, for example, macrolide antibiotics (erythomycin, clarithromycin), quinolone antibiotics (ciprofluoxacine), antidepressants (fluoxetine, parocetine, sertaline), and antiretroviral agents (ritonavir, amprenavir) that prolong QT interval and may cause torsades de pointes. There are many Web sites (www.Torsades.org) available that list and update these drugs. The most common cause of drug-induced prolonged QT interval is inhibition of IKr. However, drugs may block multiple ionic channels, causing complex alterations to the cardiac action potential. Other mechanisms for the development of proarrhythmia include delayed afterdepolarizations (e.g., digoxin), alterations in reentry to initiate ventricular arrhythmias (e.g., flecainide), and decreased automaticity and AV conduction block (e.g., β blockers). A complex interplay between the environment and genetic background drive the clinical manifestation of proarrhythmia. For example, single nucleotide polymorphisms in MiRP, a β subunit of IKr, predisposes to quinidine-mediated torsades de pointes. In approximately 10–15% of patients with acquired long QT syndrome, DNA variants of congenital long QT syndrome could be identiἀed (Yang et al. 2002). Heterozygous mutation of SCN5A (i.e., long QT syndrome 3) were identiἀed in African American patients at increased risk for torsades de pointes (Splawski et al. 2002). Therefore, there appears to be relatively common subclinical mutations that predispose to proarrhythmia. Genetic differences in hepatic cytochrome P450 enzymes may predispose to proarrhythmia by altering the metabolism of drugs. As previously mentioned, genetic differences in the metabolism of procainamide can lead to increased levels of its metabolite, N-acetyl procainamide, which inhibits IKr and can prolong the QT interval.

Cardiac Antiarrhythmic Agents

873

53.4â•… Epilepsy and Proarrhythmia Sudden unexplained death in epilepsy (SUDEP) is a well-recognized, although not wellunderstood, phenomenon. The mechanisms may involve the development of cardiac arrhythmias during or after a seizure. Utilizing video electroencephalography/ECG monitoring, the most common arrhythmia noted is sinus tachycardia (Britton 2004). However, QT prolongation, bundle branch block, bradycardia, and asystole have also been noted. These arrhythmias are thought to be the result of autonomic effects of seizures on central cardiac chronotropic modulation (Britton 2004). The role of drugs that may regulate autonomic function (i.e., β blockers) requires further investigation. Ictal bradycardia is less common, but probably clinically more signiἀcant. Carbamazepine has been implicated in ictal bradycardia. Phenytoin is commonly used to manage seizures and very rarely utilized as an antiarrhythmic agent. It has a speciἀc use in life-threatening ventricular arrhythmias caused by digitalis. No speciἀc proarrhythmic effect of phenytoin has been demonstrated. Lamotrigine is a newer anticonvulsant agent that has been shown to suppress K+ currents and potentially prolong the QT interval (Danielsson et al. 2005). There are rare case reports of SUDEP associated with this lamotrigine (Aurlien et al. 2007). However, no speciἀc drug or drug combinations have been implicated in SUDEP. Patients with epilepsy and coexisting cardiac arrhythmias may pose special challenges to clinicians. Carbamazepine is a powerful inducer of cytochrome P450 system (CYP3A4) and may increase the metabolism of many drugs used to control cardiac arrhythmias (e.g., calcium channel blockers such as verapamil and diltiazem, amiodarone, and quinidine). Phenytoin and phenobarbital also induce CYP3A4. Higher doses of drugs may be needed to control arrhythmia, and adjusting medication may result in proarrhythmia development. Careful monitoring of drug levels, serial ECGs, and ambulatory ECG monitoring during dose adjustments may be needed to ensure adequate control of arrhythmia and absence of proarrhythmia.

53.5â•… Conclusion The electrophysiologic properties of the heart are dependent on the movement of ions through ionic channels. Drugs that modulate these ionic channel activities can be used to treat cardiac arrhythmias; however, they also have the risk of producing new arrhythmias or worsening existing arrhythmia (i.e., proarrhythmia). Proarrhythmia is a serious clinical problem. The most common mechanism of proarrhythmia is prolongation of repolarization characterized by long QT interval on the ECG and risk for torsades de pointes. It is likely that a genetic susceptibility for proarrhythmia may exist, and a complex interaction between drug, genetic background, metabolism, and the environment determines the clinical manifestation of proarrhythmia.

References Alboni, P., G. L. Botto, N. Baldi, M. Luzi, V. Russo, L. Gianfranchi, P. Marchi et al. 2004. Outpatient treatment of recent-onset atrial ἀbrillation with the “pill-in-the-pocket” approach. N Engl J Med 351 (23): 2384–2391.

874 Sudden Death in Epilepsy: Forensic and Clinical Issues Aurlien, D., E. Tauboll, and L. Gjerstad. 2007. Lamotrigine in idiopathic epilepsy—increased risk of cardiac death? Acta Neurol Scand 115 (3): 199–203. Britton, J. W. 2004. Syncope and seizures—differential diagnosis and evaluation. Clin Auton Res 14 (3): 148–159. Brugada, R., J. Brugada, C. Antzelevitch, G. E. Kirsch, D. Potenza, J. A. Towbin, and P. Brugada. 2000. Sodium channel blockers identify risk for sudden death in patients with ST-segment elevation and right bundle branch block but structurally normal hearts. Circulation 101 (5): 510–515. Cairns, J. A., S. J. Connolly, R. Roberts, and M. Gent. 1997. Randomised trial of outcome after myocardial infarction in patients with frequent or repetitive ventricular premature depolarisations: CAMIAT. Canadian Amiodarone Myocardial Infarction Arrhythmia Trial Investigators. Lancet 349 (9053): 675–682. Connolly, S. J., I. Cybulsky, A. Lamy, R. S. Roberts, B. O’Brien, S. Carroll, E. Crystal, K. E. Thorpe, and M. Gent; Beta-Blocker Length of Stay (BLOS) Study. 2003. Double-blind, placebo-controlled, randomized trial of prophylactic metoprolol for reduction of hospital length of stay after heart surgery: The Beta-Blocker Length of Stay (BLOS) study. Am Heart J 145 (2): 226–232. Danielsson, B. R., K. Lansdell, L. Patmore, and T. Tomson. 2005. Effects of the antiepileptic drugs lamotrigine, topiramate and gabapentin on hERG potassium currents. Epilepsy Res 63 (1): 17–25. Doval, H. C., D. R. Nul, H. O. Grancelli, S. V. Perrone, G. R. Bortman, and R. Curiel. 1994. Randomised trial of low-dose amiodarone in severe congestive heart failure. Grupo de Estudio de la Sobrevida en la Insuἀciencia Cardiaca en Argentina (GESICA). Lancet 344 (8921): 493–498. Effect of prophylactic amiodarone on mortality after acute myocardial infarction and in congestive heart failure: Meta-analysis of individual data from 6500 patients in randomised trials. Amiodarone Trials Meta-Analysis Investigators. 1997. Lancet 350 (9089): 1417–1424. Giri, S., C. M. White, A. B. Dunn, K. Felton, L. Freeman-Bosco, P. Reddy, J. P. Tsikouris, H. A. Wilcox, and J. Kluger. 2001. Oral amiodarone for prevention of atrial ἀbrillation after open heart surgery, the Atrial Fibrillation Suppression Trial (AFIST): A randomised placebo-controlled trial. Lancet 357 (9259): 830–836. Gorgels, A. P., D. ool A. van den, A. Hofs, R. Mulleneers, J. L. Smeets, M. A. Vos, and H. J. Wellens. 1996. Comparison of procainamide and lidocaine in terminating sustained monomorphic ventricular tachycardia. Am J Cardiol 78 (1): 43–46. Grace, A. A., and A. J. Camm. 1998. Quinidine. N Engl J Med 338 (1): 35–45. Kuck, K. H., R. Cappato, J. Siebels, and R. Ruppel. 2000. Randomized comparison of antiarrhythmic drug therapy with implantable deἀbrillators in patients resuscitated from cardiac arrest: The Cardiac Arrest Study Hamburg (CASH). Circulation 102 (7): 748–754. Naccarelli, G. V., D. L. Wolbrette, and J. C. Luck. 2001. Proarrhythmia. Med Clin North Am 85 (2): 503–26, xii. Randomized antiarrhythmic drug therapy in survivors of cardiac arrest (the CASCADE study). The CASCADE Investigators. 1993. Am J Cardiol 72 (3): 280–287. Roden, D. M. 1998. Mechanisms and management of proarrhythmia. Am J Cardiol 82 (4A): 49I–57I. Roy, D., M. Talajic, P. Dorian, S. Connolly, M. J. Eisenberg, M. Green, T. Kus, et al. 2000. Amiodarone to prevent recurrence of atrial ἀbrillation. Canadian Trial of Atrial Fibrillation Investigators. N Engl J Med 342 (13): 913–920. Sadowski, Z. P., J. H. Alexander, B. Skrabucha, A. Dyduszynski, J. Kuch, E. Nartowicz, G. Swiatecka, D. F. Kong, and C. B. Granger. 1999. Multicenter randomized trial and a systematic overview of lidocaine in acute myocardial infarction. Am Heart J 137 (5): 792–798. Schwartz, P. J., S. G. Priori, E. H. Locati, C. Napolitano, F. Cantu, J. A. Towbin, M. T. Keating, H. Hammoude, A. M. Brown, and L. S. Chen. 1995. Long QT syndrome patients with mutations of the SCN5A and HERG genes have differential responses to Na+ channel blockade and to increases in heart rate. Implications for gene-speciἀc therapy. Circulation 92 (12): 3381–3386.

Cardiac Antiarrhythmic Agents

875

Singh, B. N. 1998. Antiarrhythmic drugs: A reorientation in light of recent developments in the control of disorders of rhythm. Am J Cardiol 81 (6A): 3D–13D. Splawski, I., K. W. Timothy, M. Tateyama, C. E. Clancy, A. Malhotra, A. H. Beggs, F. P. Cappuccio, G.€A. Sagnella, R. S. Kass, and M. T. Keating. 2002. Variant of SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science 297 (5585): 1333–1336. Vaughan Williams, E. M. 1992. The relevance of cellular to clinical electrophysiology in classifying antiarrhythmic actions. J Cardiovasc Pharmacol 20 (Suppl 2): S1–S7. Wyse, D. G., M. Talajic, G. E. Hafley, A. E. Buxton, L. B. Mitchell, T. K. Kus, D. L. Packer, et al. 2001. Antiarrhythmic drug therapy in the Multicenter UnSustained Tachycardia Trial (MUSTT): Drug testing and as-treated analysis. J Am Coll Cardiol 38 (2): 344–351. Yang, P., H. Kanki, B. Drolet, T. Yang, J. Wei, P. C. Viswanathan, S. H. Hohnloser, et al. 2002. Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes. Circulation 105 (16): 1943–1948.

54

Could Beta–Blocker Antiarrhythmic and Antiseizure Activity Help Prevent SUDEP? Claire M. Lathers

Contents 54.1 Animal Models to Study New and Existing Antiepileptic Drugs 54.2 Interventions to Prevent Sudden Death References

877 880 883

54.1â•… Animal Models to Study New and Existing Antiepileptic Drugs The pentylenetetrazol test is an acute seizure model for screening of antiepileptic drugs in animals. This model can identify drugs with varied mechanisms of action and broad clinical utility for treatment of epilepsy. The pentylenetetrazol intravenous or subcutaneous seizure test provides a threshold dose for induction of seizures in individual animals (Mandhane et al. 2007). This model was used in mice to compare seizure patterns and intra- and inter-animal variabilities. Anticonvulsant activities of antiepileptic drugs at nontoxic dose levels were evaluated in both the pentylenetetrazol and maximal electroshock seizure tests. Intra- and inter-animal variabilities in the seizure response were low in the intravenous pentylenetetrazol when compared with the subcutaneous pentylÂ� enetetrazol. Phenobarbital sodium, ethosuximide, sodium valproate, diazepam, tiagabine, oxcarbazepine, and zonisamide enhanced threshold or onset latency for clonus after both intravenous and subcutaneous pentylenetetrazol. Levetiracetam and pregabalin were active in the intravenous pentylenetetrazol test and exhibited no action after subcutaneous pentylenetetrazol. The action of drugs to protect from tonic extensor in the maximal electroshock seizure test and after intravenous pentylenetetrazol established that phenobarbital, phenytoin, carbamazepine, sodium valproate, gabapentin, oxcarbazepine, zonisamide, and pregabalin were effective in both tests. Ethosuximide, diazepam, and levetiracetam were effective after intravenous pentylenetetrazol test but not in the maximal electroshock seizure test. Lamotrigine and topiramate were effective in the maximal electroshock seizure test but not after intravenous pentylenetetrazol. Pentylenetetrazol has also been used in an animal model to induce seizures for the€study of a unique, alternate route for the administration of antiepileptic drugs€when€an€intraveÂ� nous line is not easily established. Intravenous access for administration of antiepileptic drugs can be time consuming and difficult in an infant during a seizure. The intraosseous route is an alternative means of vascular access for drug administration. Domestic swine (13–20 kg) were anesthetized with ketamine (20 mg/kg, i.m.) and alpha-chloralose (80€mg/â•›kg, i.v.) and given gallamine (4 mg/kg, i.v.) to prevent muscle€Â�fasciculations (Lathers 877

878 Sudden Death in Epilepsy: Forensic and Clinical Issues

et al. 1989c). Tracheostomies were performed, and the animals were ventilated with a Harvard respirator. The left femoral vein was cannulated, and pentylenetetrazol (100€mg/kg) was given to elicit epileptogenic activity. Sixty seconds after the onset of epileptogenic activity, the animals received saline or diazepam (0.1 mg/kg) or propranolol (2.5 mg/kg) given intravenously or via the intraosseous route (18-gauge spinal needle placed in the right proximal tibia). Both diazepam and propranolol effectively suppressed epileptogenic activity via both the intravenous or intraosseous routes. Thus, the intraosseous route was established to be a rapid and effective alternative route for the administration of antiepileptic drugs when an intravenous line cannot be readily established. This study established the validity of the intraosseous route to stop seizures and also demonstrated antiepileptic activity of the beta blocker propranolol. The fact that beta blocking agents may possess anticonvulsant action has been known for many years (Bose et al. 1963; Conway et al. 1978; Papanicolaou et al. 1982; Dashputra et al. 1985; Jaeger et al. 1979; Murmann et al. 1966). Tocco et al. (1980) also demonstrated that propranolol possesses anticonvulsant actions. Mueller and Dunwiddie (1983) showed that timolol selectively blocked the proconvulsant activity of 2-fluoronorepinephrine and l -isoproterenol in in vitro hippocampal slice preparations superfused with penicillin and elevated levels of potassium. Propranolol or timolol produced an anticonvulsant action when pentylenetetrazol induced convulsions in rats (Louis et al. 1982). The anticonvulsant action of timolol reported in the study of Lathers et al. (1990) is similar to the anticonvulsant action of diazepam when employed in the swine model (Lathers et al. 1989c; Spivey et al. 1987a). The anticonvulsant effect of propranolol has been studied in DBA/2 mice. DBA/2 mice have been proposed for use as an animal model for sudden unexpected death in epilepsy (SUDEP) (Faingold et al. 2010, Chapter 41). The laboratory of De Sarro et al. (2002) studied the ability of various beta blockers to antagonize tonic–clonic seizures in these mice. A summary of their data is presented in Table 54.1. The two enantiomers of propranolol antagonized the tonic–clonic seizures in these mice, with the (−) enantiomer approximately 1.5 times more potent than the (+) enantiomer. Propranolol was better than metopropol, and atenolol did not exhibit an effect. Both of the enantiomers of propranolol and metoprolol exhibited an additive anticonvulsant effect when given with diazepam, phenobarbital, lamotrigine, and valproate. Where appropriate, the therapeutic relevance of such drug combinations must be examined in persons with epilepsy. Fischer (2002) examined ss-propranolol and its two enantiomers in multiple screening models to characterize the anticonvulsant actions and to determine the molecular mechanism of action. The animal models and data are summarized in Table 54.2. It was concluded that propranolol and its two enantiomers exhibit anticonvulsant effects in models

Table 54.1â•… Ability of Beta Blocking Drugs to Antagonize Tonic–Clonic Seizures in€DBA/2€Mice Beta-Blocker

Ability to Antagonize Tonic–Clonic Seizures

Propranolol Metroprolol Atenolol

(−) Enantiomer was approximately 1.5 times more potent than the (+) enantiomer Less active than propranolol Did not affect audiogenic seizures

Source: Constructed from De Sarro, G., et al., Eur J Pharmacol, 442 (3), 205–213, 2002. With permission.

Could Beta–Blocker Antiarrhythmic and Antiseizure Activity Help Prevent SUDEP? 879 Table 54.2â•… Anticonvulsant Actions of Propranolol and Its Enantiomers in Screening Models Model Tonic electroshock seizures in mice Traditional maximal electroshock test Pentylenetetrazol seizure threshold test Unrestrained rats with chronically implanted electrodes dorsal hippocampus Amygdala-kindled rats Whole-cell patch-clamp experiments in cultured rat cardiomyocytes Mouse spinal cord neurons in culture

Findings for ss-Propranolol and Its Two Enantiomers Dose-dependently raised threshold for tonic electroshock seizure In combination with antiepileptics, signiἀcantly increased the anticonvulsant effectiveness of ss-propranolol and two enantiomers Effective anticonvulsant action (+/−) and (+) Propranolol were not effective in preventing clonic seizures Propranolol and (+) enantiomer equieffectively reduced duration of electrically evoked hippocampal afterdischarges and raised focal stimulation threshold Reduced seizure severity from generalized clonic–tonic to unilateral forelimb seizures (+) and (−) Propranolol depressed the fast inward sodium current in concentration and use-dependent manner Inhibited picrotoxin-induced burst ἀring activity

Source: Constructed from Fischer, W., Seizure, 11 (5), 285–302, 2002. With permission.

for generalized tonic–clonic and complex partial seizures, and this action was due to sodium channel blockade and not due to ss-adrenoceptor blocking action. Propranolol has also been demonstrated to exert anticonvulsant activity in a different animal model. The ability of propranolol to increase the threshold for lidocaine-induced tonic–clonic convulsions in awake, spontaneously breathing rats was studied (Nakamura et al. 2008). Propranolol was administered via two different sites, i.e., intracerebroventricular or intravenous routes to differentiate central from peripheral actions, respectively. For a discussion of the use of different routes of administration to examine drug action on peripheral or central sites, see the chapter by Lathers and Levin (2010, this book) discussing animal models. The results revealed that propranolol increased the threshold for lidocaine-induced convulsions by acting directly on the brain following intracerebroventricular administration. Animal data demonstrating anticonvulsant actions of beta blockers has also been demonstrated in humans. Studies have examined the beta blocker esmolol in humans. Once electroconvulsive therapy–induced seizures have stopped, esmolol reduces the elevated heart rate without initiating asystole or bradycardia (Robinson and Lighthall 2004). A review of the literature describing the effect of beta blockers on seizure duration and cardiovascular variables in patients undergoing electroconvulsive therapy in randomized placebo-controlled studies was conducted by van den Broek et al. (2008). This class of drugs is given to treat the tachycardia and high blood pressure that may occur during treatment with electroconvulsive therapy. The review demonstrated that esmolol was the most often administered drug and was deemed to be the beta blocker of choice for use during electroconvulsive therapy. Esmolol shortens seizure duration in a dose-dependent manner. It is also used to treat those developing prolonged hypertension or tachycardia. Labetalol is a possible alternative to esmolol but has a longer half-life. Landiolol has a short halflife, greater cardioselectivity, and a higher potency; however, its use is yet to be deἀned. Landiolol is an ultra-short-acting beta blocking agent that is highly selective for cardio beta receptors (Sasao et al. 2001). In rabbits, it was demonstrated to exhibit a slightly more

880 Sudden Death in Epilepsy: Forensic and Clinical Issues

potent negative chronotropic action than esmolol with less effect on blood pressure. It was studied in psychiatric patients before induction of anesthesia to elicit motor seizure activity during electroconvulsive therapy (Nomoto et al. 2006). Pretreatment with landiolol caused shorter motor seizure duration. The degree to tachycardia and rate pressure produced after electroconvulsive therapy was attenuated by landiolol. An earlier study demonstrated that a lower dose did not alter cognitive function, recovery from anesthesia, or seizure duration (Sakamoto et al. 2004). In a third study, landiolol was demonstrated to suppress heart rate elevation during electroconvulsive shock therapy without affecting blood pressure and cerebral blood flow velocity in the middle cerebral artery (Saito et al. 2005). The collective data reviewed demonstrate that beta blockers decrease seizure duration.

54.2â•…Interventions to Prevent Sudden Death The antiarrhythmic and anticonvulsant activity of beta blocking agents pertinent to adrenergic mechanisms of SUDEP presents an interesting challenge to the possible justiἀcation of trials testing the effects of beta blockers in SUDEP cases where resuscitation is possible. Lathers (1980) reported that timolol, a beta blocking agent approved only for use to treat glaucoma in humans, exhibited antiarrhythmic activity in the cat. This antiÂ�arrhythmic action was later conἀrmed in humans enrolled in the Beta-Blocker Heart Attack Trial (BHAT) trial (Barker et al. 2005). Lathers et al. (1989a) also administered timolol centrally and then peripherally to examine its effect on cardiac arrhythmias, epileptiform activity, blood pressure, and heart rate changes induced by pentylenetetrazol adminÂ� istered intracerebroventricularly to anesthetized cats. Increasing doses of timolol administered intracerebroventricularly and intravenously signiἀcantly decreased the elevation of mean arterial blood pressure and heart rate, and decreased and subsequently abolished the incidence of cardiac arrhythmias associated with the epilepiform activity. One may speculate that pentylenetetrazol, trauma, inhibition of prostaglandin transport across the blood–brain barrier, or altered synthesis or metabolism of central enkephalins may lead to increased central levels of prostaglandin E2 and/or enkephalins (Suter and Lathers 1984; Kraras et al. 1987). Increased central levels of prostaglandin E2 and/or enkephalins may inhibit central gamma aminobutyric acid release and the associated epileptogenic activity, increase blood pressure and heart rate, increase sympathetic and parasympathetic central neural outflow, impair or imbalance cardiac sympathetic and parasympathetic discharge, and cause arrhythmia and/or death. These factors may interact in various combinations in different susceptible individuals to initiate arrhythmias and/or death. It may be that the central intracerebroventricular administration of timolol partially suppressed the eÂ�pileptiform activity and subsequently decreased the blood pressure and heart rate values elevated by pentylenetetrazol. Timolol may interfere with the central actions of prostaglandin E2 or enkephalins to reverse their known capabilities to induce epileptiform activity. Experimental studies are required to verify this possibility. It has been theorized that pharmacological agents capable of suppressing epileptiform activity and the sympathetic component of cardiac arrhythmias may be a regimen to prevent interictal activity and cardiac arrhythmias that may contribute to production of SUDEP (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Carnel et al. 1985; Lathers et al. 1985). Data indicate that timolol possesses components of both of these capabilities. Blockade of cardiac beta 1 receptors, a cardiac neurodepressant effect, and/â•›or

Could Beta–Blocker Antiarrhythmic and Antiseizure Activity Help Prevent SUDEP? 881

membrane depressant actions of beta blocking agents are thought to contribute, at least in part, to the antiarrhythmic action of beta blocking agents (Lathers 1980; Lathers et al. 1977a, 1977b; Spivey and Lathers 1985; Lathers et al. 1986a, 1986b, 1990; Lathers and Spivey 1987; Lathers et al. 1988b, 1988c). Since pentylenetetrazol is an accepted convulsive model and since some drugs capable of suppressing pentylenetetrazol-induced epileptiform activity are anticonvulsants, the data suggest that timolol exhibited an anticonvulsant action. Although timolol reversed the effects of pentylenetetrazol on the brain, this does not necessarily mean that timolol has intrinsic “anticonvulsant” properties separate from an ability to reverse the effects of pentylenetetrazol. To address this, studies need to determine whether timolol and other beta blockers will protect against seizures induced in other experimental models of epilepsy and in persons with epilepsy. The capability of timolol to suppress interictal discharges and cardiac arrhythmias elicited in other in vivo experimental models not involving pentylenetetrazol should be examined. If timolol also suppresses interictal discharges and arrhythmias in these experimental models, this would provide additional evidence to support the suggestion that timolol may be an effective agent for use in epileptic patients to prevent SUDEP. Various mechanisms of beta blockers may explain their anticonvulsant action. The anticonvulsant action of beta blocking agents is commonly ascribed to a membrane-Â�stabilizing effect (Conway et al. 1978). Other proposed mechanisms include decreased central serotonergic (Conway et al. 1978) and monoamaine oxidase activity (Bose et al. 1963). One other possible antiepileptic mechanism of beta blockers may include beta-Â�adrenoceptor blockâ•›ade, especially beta2 receptors in the central nervous system (Papanicolaou et al. 1982). Although norepinephrine is generally believed to have an anticonvulsant effect, studies suggest that norepinephrine itself may exacerbate seizure activity via activation of beta receptors. In genetic epilepsy-prone rats, the actual state of abnormal seizure susceptibility but not severity may be determined by norepinephrine deἀcits in the hypothalamus/ thalamus (Dailey and Jobe 1986). Both severity and susceptibility can be determined by norepinephrine deἀcits in the telencephalon, midbrain, and pons–medulla, while seizure severity but no susceptibility may be determined by norepinephrine abnormalities in the cerebellum. The noradrenergic effect may not be uniform throughout the hippocampus; thus, selective activation of alpha or beta receptors by norepinephrine in brain areas such as the hippocampus may trigger either anticonvulsant or proconvulsant effects, respectively (Mueller and Dunwiddie 1983). Beta blocking agents can increase norepinephrine concentration in cerebral spinal fluid (Tackett et al. 1981) and potentiate effects of exogenously administered norepinephrine on vas deferens contractions (Patil 1968). The data of Lathers et al. (1990) suggest that establishment of beta blockade with timolol would increase central norepinephrine concentration; an increase in norepinephrine activity at the central postsynaptic alpha 1 receptor sites may account for the anticonvulsant effect of beta blocking agents (Goldman et al. 1987). Thus, timolol may protect against seizures induced by pentylenetetrazol, in part via a selective blockade of seizure-inducing beta receptors, allowing available norepinephrine to stimulate central alpha 1 receptors that exert an anticonvulsant action. The role of central postsynaptic alpha 2 receptors must also be evaluated. Activation of alpha 2 receptors decreases excitability of CA1 pyramidal neurons (Mueller and Dunwiddie 1983; Mueller et al. 1982). Clonidine and 1-m-norepinephrine are more selective for alpha 2 than for alpha 1 receptors and inhibit epileptiform activity at low concentrations; the alpha 1 agonist l -phenylephrine was ineffective at much higher concentrations. The data implicate that central postsynaptic alpha 2 receptors exert

882 Sudden Death in Epilepsy: Forensic and Clinical Issues

a greater role than alpha 1 receptors in the anticonvulsant action of timolol (Lathers et al. 1989a). Experiments are needed to conἀrm this. Swine are used as a model of cardiac arrest and resuscitation (Spivey et al. 1987a; Lathers et al. 1989c; Schoffstall et al. 1989; Jim et al. 1988, 1989). The heart of the swine, in terms of the number of collaterals, is more like the human heart and is a much better model for studying human disease than dog, cat, or rat hearts. For an in-depth discussion of cardiac differences among species and different pharmacologic reactions, the reader is referred to the review of Lathers et al. (1988a). The weight of the swine and the size of the tibia used in the intraosseous studies are similar to pediatric human subjects. Swine could be studied to address the question of the use of beta blockers in SUDEP cases where patients could be resuscitated. Intraosseous propranolol (Jim et al. 1988) or diazepam (Lathers et al. 1989c) or lorazepam (Jim et al. 1989) exerted anticonvulsant effects against pentylenetetrazolinduced convulsions. Since resuscitation was also performed, this animal model could be used to evaluate whether clinical trials should be designed to test the effect of beta blockers in SUDEP cases where patients are found in circumstances amenable to resuscitation efforts. Interventions to prevent sudden cardiac deaths are few. Depression and/or stress play a role in the course of coronary artery disease (Lathers and Schraeder 2006; Glassman and Shapiro 1998; Ziegelstein 2007a, 2007b; Lathers et al. 2008, 2010). The role of stress and sudden death in patients is important in that it gives a common explanation for adrenergicrelated arrhythmias, leading to death in different conditions such as epilepsy, heart failure, exercise-Â�induced death, and possibly sudden death in schizophrenia. In the presence of ischemic heart disease in those who have survived an episode of sudden cardiac death in the setting of emotional stress, the use of a beta blocker should be considered. Localization-related seizures involve the most common types of epilepsy in adults with focal and/or asymmetric epileptiform activity. Asymmetric brain activity may make the heart more susceptible to ventricular arrhythmias. Lateralization of cerebral activity during stress may stimulate the heart asymmetrically and produce areas of inhomogeneous repolarization that create electrical instability and facilitate the development of arrhythmias (Ziegelstein 2007a; Han and Moe 1964; Lathers et al. 1990, 1977b, 1978). In general, the therapeutic approach to ischemic heart disease related to congestive heart failure and left ventricular hypertrophy, especially in the post infarction period, includes beta blockers, converting enzyme inhibitors, and amiodarone (McAlister and Teo 1997; Kochs et al. 1993). Implantable deἀbrillators may beneἀt persons who survived cardiac arrest due to ventricular arrhythmias or those with poorly treated ventricular tachycardia (Merino 2001). Measures for the prevention of SUDEP are even more limited. Optimizing seizure control with a minimal number of antiepileptic drugs and/or with temporal lobectomy in appropriate patients are obvious interventions (Sperling et al. 1996). Use of nocturnal respiratory monitors may help (Langan et al. 2002). The prevention of sudden death in cardiac patients and prevention of sudden, unexpected death in persons with epilepsy (So 2006), await further deἀnition of the best combination of pharmacological and nonpharmacological measures. There is a need to investigate the potential clinical usefulness of beta blockers as a primary adjunctive anticonvulsant drug in humans. Many questions must be addressed for the possible clinical use of beta blockers in persons with epilepsy. Which beta blocker would be the best to study in persons with epilepsy: propranolol or a different beta blocker? Does the half life of the selected beta blocker provide protection for a sufficient time interval? What is the clinical dose of the beta blocker to be used to treat persons with epilepsy?

Could Beta–Blocker Antiarrhythmic and Antiseizure Activity Help Prevent SUDEP? 883

What is the clinical dose–response curve to prevent seizures? Not always is the exact full therapeutic dose of a drug used for a different therapeutic effect in a given species. For example, aspirin tablets (325 mg each) at a dose of 325 mg twice a day for a total of 650€mg every 4 h with not more than 12 tablets in 24 h is used to treat adult headaches and/or muscle aches and pain. A much lower dose of aspirin, 81 mg once a day, is used to prevent clotting and/or unwanted strokes. Propranolol is an example of a drug initially used to treat hypertension and used in an off-label manner to treat migraine headaches. There are many persons with epilepsy who also suffer from migraine. These patients might receive an added beneἀt from the use of propranolol as an adjunct drug to control their seizures. There is a need for clinical trials to sort out the answers to these questions.

References Barker, A. et al. Beta-Blocker Heart Attack Trial (BHAT) NCT00000492. National Heart Lung Blood Institute, 2005/06/23. 2005. Available from http://clinicaltrials.gov. Bose, B. C., A. Q. Saiἀ, and S. K. Sharma. 1963. Studies on anticonvulsant and antiἀbrillatory drugs. Arch Int Pharmacodyn Ther 146: 106–113. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30 (4): 225–240. Conway, J., D. T. Greenwood, and D. N. Middlemiss. 1978. Central nervous actions of beta-adrenoreceptor antagonists. Clin Sci Mol Med 54 (2): 119–124. Dailey, J. W., and P. C. Jobe. 1986. Indices of noradrenergic function in the central nervous system of seizure-naive genetically epilepsy-prone rats. Epilepsia 27 (6): 665–670. Dashputra, P. G., V. P. Patki, and T. J. Hemnani. 1985. Antiepileptic action of beta-adrenergic blocking drugs: Pronethalol and propranolol. Mater Med Pol 17 (2): 88–92. De Sarro, G., E. D. Di Paola, G. Ferreri, A. De Sarro, and W. Fischer. 2002. Influence of some betaadrenoceptor antagonists on the anticonvulsant potency of antiepileptic drugs against audiogenic seizures in DBA/2 mice. Eur J Pharmacol 442 (3): 205–213. Faingold, C. L., S. Tupal, Y. Mhaskar, and V. V. Uteshev. 2010. DBA mice as models of sudden unexpected death in epilepsy (Chapter 41). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. Schraeder, M. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Fischer, W. 2002. Anticonvulsant proἀle and mechanism of action of propranolol and its two enantiomers. Seizure 11 (5): 285–302. Glassman, A. H., and P. A. Shapiro. 1998. Depression and the course of coronary artery disease. Am J Psychiatry 155 (1): 4–11. Goldman, B. D., A. Z. Stauffer, and Lathers C. M. 1987. Beta blocking agents and the prevention of sudden unexpected death in the epileptic person: Possible mechanisms. Fed Proc 46: 705. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Jaeger, V., B. Esplin, and R. Capek. 1979. The anticonvulsant effects of propranolol and beta-Â�adrenergic blockade. Experientia 35 (1): 80–81. Jim, K. F., C. M. Lathers, V. L. Farris, L. F. Pratt, and W. H. Spivey. 1989. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous lorazepam in pigs. Epilepsia 30 (4): 480–486. Jim, K. F., C. M. Lathers, W. H. Spivey, W. D. Matthews, C. Kahn, and K. Dolce. 1988. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous propranolol in pigs. J Clin Pharmacol 28 (12): 1106–1111. Kochs, M., T. Eggeling, and V. Hombach. 1993. Pharmacological therapy in coronary heart disease: Prevention of life-threatening ventricular tachyarrhythmias and sudden cardiac death. Eur Heart J 14 (Suppl E): 107–119.

884 Sudden Death in Epilepsy: Forensic and Clinical Issues Kraras, C. M., N. Tumer, and C. M. Lathers. 1987. The role of enkephalins in the production of epileptogenic activity and autonomic dysfunction: Origin of arrhythmia and sudden death in the epileptic patient? Med Hypotheses 23 (1): 19–31. Langan, Y., L. Nashef, and J. W. Sander. 2002. Certiἀcation of deaths attributable to epilepsy. J Neurol Neurosurg Psychiatry 73 (6): 751–752. Lathers, C. M. 1980. Effect of timolol on autonomic neural discharge associated with ouabainÂ�induced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106. Lathers, C. M., K. F. Jim, W. B. High, W. H. Spivey, W. D. Matthews, and T. Ho. 1989b. An investigation of the pathological and physiological effects of intraosseous sodium bicarbonate in pigs. J Clin Pharmacol 29 (4): 354–359. Lathers, C. M., K. F. Jim, and W. H. Spivey. 1989c. A comparison of intraosseous and intravenous routes of administration for antiseizure agents. Epilepsia 30 (4): 472–479. Lathers, C. M., K. F. Jim, W. H. Spivey, C. Kahn, K. Dolce, and W. D. Matthews. 1990. Chapter 24: Antiepileptic activity of beta-blocking agents. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., K. M. Keller, J. Roberts, and A. B. Beasley. 1977a. Role of the adrenergic nervous system in arrhythmia produced by acute coronary artery occlusion. In Pathophysiology and Therapeutics of Myocardial Ischemia, ed. A. M. Lefer, G. J. Kelliher, and M. J. Rovetto. New York, NY: Spectrum. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., and R. M. Levin. 2010. Animal model for sudden cardiac death: Sympathetic innervation and myocardial beta receptor densities (Chapter 33). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986a. Regional distribution of myocardial betaadrenoceptors in the cat. Eur J Pharmacol 130 (1–2): 111–117. Lathers, C. M., L. J. Lipka, and H. Klions. 1988a. Digitalis glycosides: A discussion of the similarities and differences in actions and existing controversies. Rev Clin Basic Pharm 7 (1–4): 1–108. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977b. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., and P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Neurocardiologic mechanistic risk factors in sudden unexpected death in epilepsy (Chapter 1). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton: CRC Press. Lathers, C. M., and W. H. Spivey. 1987. The effect of beta blockers on cardiac neural discharge associated with coronary occlusion in the cat. J Clin Pharmacol 27 (8): 582–592. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988b. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28 (8): 736–745. Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986b. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39 (22): 2121–2141.

Could Beta–Blocker Antiarrhythmic and Antiseizure Activity Help Prevent SUDEP? 885 Lathers, C. M., W. H. Spivey, and N. Tumer. 1988c. The effect of timolol given ἀve minutes after coronary occlusion on plasma catecholamines. J Clin Pharmacol 28 (4): 289–299. Lathers, C. M., A. Z. Stauffer, N. Tumer, C. M. Kraras, and B. D. Goldman. 1989a. Anticonvulsant and antiarrhythmic actions of the beta blocking agent timolol. Epilepsy Res 4 (1): 42–54. Lathers, C. M., N. Tumer, and C. M. Kraras. 1985. Cardiovascular and epileptogenic effects of pentÂ� ylenetetrazol administered intracerebroventricularly in cats. Epilepsia 26: 520. Louis, W. J., J. Papanicolaou, R. J. Summers, and F. J. Vajda. 1982. Role of central beta-Â�adrenoceptors in€ the control of pentylenetetrazol-induced convulsions in rats. Br J Pharmacol 75 (3): 441–446. Mandhane, S. N., K. Aavula, and T. Rajamannar. 2007. Timed pentylenetetrazol infusion test: A comparative analysis with s.c.PTZ and MES models of anticonvulsant screening in mice. Seizure 16 (7): 636–644. McAlister, F. A., and K. K. Teo. 1997. Antiarrhythmic therapies for the prevention of sudden cardiac death. Drugs 54 (2): 235–252. Merino, J. L. 2001. Mechanisms underlying ventricular arrhythmias in idiopathic dilated cardiomyÂ� opathy: Implications for management. Am J Cardiovasc Drugs 1 (2): 105–118. Mueller, A. L., M. R. Palmer, B. J. Hoffer, and T. V. Dunwiddie. 1982. Hippocampal noradrenergic responses in vivo and in vitro. Characterization of alpha and beta components. Naunyn Schmiedebergs Arch Pharmacol 318 (4): 259–266. Mueller, A. L., and T. V. Dunwiddie. 1983. Anticonvulsant and proconvulsant actions of alpha- and beta-noradrenergic agonists on epileptiform activity in rat hippocampus in vitro. Epilepsia 24 (1): 57–64. Murmann, W., L. Almirante, and M. Saccani-Guelἀ. 1966. Central nervous system effects of four beta-adrenergic receptor blocking agents. J Pharm Pharmacol 18 (5): 317–318. Nakamura, T., Y. Oda, R. Takahashi, K. Tanaka, I. Hase, and A. Asada. 2008. Propranolol increases the threshold for lidocaine-induced convulsions in awake rats: A direct effect on the brain. Anesth Analg 106 (5): 1450–1455. Nomoto, K., T. Suzuki, K. Serada, K. Oe, T. Yoshida, and S. Yamada. 2006. Effects of landiolol on hemodynamic response and seizure duration during electroconvulsive therapy. J Anesth 20 (3): 183–187. Papanicolaou, J., F. J. Vajda, R. J. Summers, and W. J. Louis. 1982. Role of beta-adrenoceptors in the anticonvulsant effect of propranolol on leptazol-induced convulsions in rats. J Pharm Pharmacol 34 (2): 124–125. Patil, P. N. 1968. Steric aspects of adrenergic drugs. 8. Optical isomers of beta adrenergic receptor antagonists. J Pharmacol Exp Ther 160 (2): 308–314. Robinson, M., and G. Lighthall. 2004. Asystole during successive electroconvulsive therapy sessions: A report of two cases. J Clin Anesth 16 (3): 210–213. Saito, S., F. Nishihara, T. Akihiro, K. Nishikawa, H. Obata, F. Goto, and N. Yuki. 2005. Landiolol and esmolol prevent tachycardia without altering cerebral blood flow. Can J Anaesth 52 (10): 1027–1034. Sakamoto, A., R. Ogawa, H. Suzuki, M. Kimura, Y. Okubo, and T. Fujiya. 2004. Landiolol attenuates acute hemodynamic responses but does not reduce seizure duration during maintenance electroconvulsive therapy. Psychiatry Clin Neurosci 58 (6): 630–635. Sasao, J., S. D. Tarver, J. D. Kindscher, C. Taneyama, K. T. Benson, and H. Goto. 2001. In rabbits, landiolol, a new ultra-short-acting beta-blocker, exerts a more potent negative chronotropic effect and less effect on blood pressure than esmolol. Can J Anaesth 48 (10): 985–989. Schoffstall, J. M., W. H. Spivey, S. Davidheiser, and C. M. Lathers. 1989. Intraosseous crystalloid and blood infusion in a swine model. J Trauma 29 (3): 384–387. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. So, E. L. 2006. Demystifying sudden unexplained death in epilepsy—are we close? Epilepsia 47 (Suppl€1): 87–92.

886 Sudden Death in Epilepsy: Forensic and Clinical Issues Sperling, M. R., M. J. O’Connor, A. J. Saykin, and C. Plummer. 1996. Temporal lobectomy for refractory epilepsy. JAMA 276 (6): 470–475. Spivey, W. H., and C. M. Lathers. 1985. Effect of timolol on the sympathetic nervous system in coronary occlusion in cats. Ann Emerg Med 14 (10): 939–944. Spivey, W. H., H. D. Unger, C. M. Lathers, and R. M. McNamara. 1987a. Intraosseous diazepam suppression of pentylenetetrazol-induced epileptogenic activity in pigs. Ann Emerg Med 16 (2): 156–159. Spivey, W. H., H. D. Unger, R. M. McNamara, M. M. LaManna, T. Ho, and C. M. Lathers. 1987b. The effect of intraosseous sodium bicarbonate on bone in swine. Ann Emerg Med 16 (7): 773–776. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15 (1): 15–30. Tackett, R. L., J. G. Webb, and P. J. Privitera. 1981. Cerebroventricular propranolol elevates cerebrospinal fluid norepinephrine and lowers blood pressure. Science 213 (4510): 911–913. Tocco, D. J., B. V. Clineschmidt, A. E. Duncan, F. A. deLuna, and J. E. Baer. 1980. Uptake of the betaadrenergic blocking agents propranolol and timolol by rodent brain: Relationship to central pharmacological actions. J Cardiovasc Pharmacol 2 (2): 133–143. van den Broek, W. W., T. H. Groenland, P. G. Mulder, A. Kusuma, T. K. Birkenhager, E. M. Pluijms, and J. A. Bruijn. 2008. Beta-blockers and electroconvulsive therapy: A review. Tijdschr Psychiatr 50 (4): 205–215. Ziegelstein, R. C. 2007a. Acute emotional stress and cardiac arrhythmias. JAMA 298 (3): 324–329. Ziegelstein, R. C. 2007b. Treatment of depression in patients with coronary artery disease. JAMA 297 (17): 1878–1879; author reply 1880.

55

Decision Analysis and Risk Management H. Gregg Claycamp

Contents 55.1 Introduction 55.2 Risk Management 55.3 Elements of Risky Decisions 55.4 Preferences for Outcomes and Probabilities 55.5 Uncertainties 55.6 Beneἀts and Quality-of-Life Objectives 55.7 Weighting the Quality of Life 55.8 Example Decision Tree 55.9 A Structured and Focused Way of Making Difficult Choices References

887 888 889 892 894 895 896 898 901 902

55.1â•…Introduction The fact that several treatment alternatives are often available for treating a disease increases the complexity of decision making for patients, physicians, and health care providers. It is conceivable, if not likely, that the alternative treatments identiἀed in the patient’s decision have not been compared in head-to-head randomized clinical trials and it might not be feasible to do so (Cohen and Neumann 2008). Even if the treatments had been compared directly, clinical trials are designed to measure efficacy and safety under optimal experimental conditions. For the patient, choosing among alternate therapies includes evaluation of possible consequences that are not necessarily reported in clinical trials. For example, differences in the long-term quality of life, alternative treatment costs, and even access to treatment are often weighted alongside with the overall chance of cure or control in a patient’s multiobjective decision making for choosing among treatment alternatives. Decision analysis and risk management provide logical frameworks for making decisions under these complex scenarios. This chapter reviews some of the basic decision analytical approaches that can be used in deciding among alternative treatments, and discusses the beneἀts and limitations among the various approaches. Decision analysis is at its best when providing a structure for resolving uncertainties, risk and value tradeoffs, and dealing with linked decisions over an extended period of treatment. Difficult decisions having many inputs and uncertain outcomes are typically a challenge for health care providers treating persons with epilepsy at risk of dying from sudden death. Deciding on an optimal course of treatment of patients at risk of sudden death from epilepsy (SUDEP) involves striking balances among possible efficacies weighted against possible side effects (Lathers et al. 2003) and is a decision-making process that can beneἀt from applied decision analysis. 887

888 Sudden Death in Epilepsy: Forensic and Clinical Issues

“Decision Analysis is the discipline comprising the philosophy, theory, methodology, and professional practice necessary to address important decisions in a formal manner.”* Decision analysis shares some methods with statistically based decision making used in, e.g., clinical trial and other research analyses. Statistical decision methods are intended to quantitatively support the acceptance or rejection of a hypothesis under well-deἀned endpoints and experimental designs. Statistically based decisions are commonly viewed as inherently objective perhaps because the methods are based on mathematics and probability theory. In contrast, decision analysis is generally descriptive or normative and lacks a “ἀrst-principles” mathematical or probabilistic theory. The inclusion of human behavior, economics, and (sometimes) social norms in a decision analysis suggests that, like risk management, decision analysis is more of a “postnormal science” than conventional statistics (Funtowicz and Ravetz 1992). High-quality randomized clinical trials are likely to remain as the best formal way to compare treatments (Song et al. 2003). However, these trials become too complex and costly to perform when multiple treatments are to be evaluated in one experimental design. Intermediate between statistically driven clinical trials and pure decision analytic designs are indirect statistical methods, such as meta-analysis, to simulate randomized, comparison clinical trials (Otoul et al. 2005). Meta-analysis sometimes uses individual records from separate clinical trials and then regroups these samples to simulate new groups in a comparison clinical trial. The simulated trial can then be used for statistical inference about the comparative safety and efficacy of the alternatives and sometimes prognostic modeling. These relatively formal meta-studies sometimes include statistically driven decisions and recommendations, ἀlling some of the gaps in information needed for treatment decisions by health care programs, physicians, and their patients (Song et al. 2003). Although providing useful information, meta-analyses are certainly not free of limitations and potential pitfalls, especially when groupings are made across population groups, patient ages, and other descriptive variables (Caldwell et al. 2005). Although less expensive and less complicated undertakings than the original clinical trials, meta-analytical methods and resources needed to execute a study are not widely accessible to physicians, pharmacologists, and their patients who seek decision support for complex, multiobjective decisions among treatment alternatives. The demand for basic tools that support treatment decisions has led to interest in decision analysis and related concepts from risk management. The two related areas of applied social and mathematical sciences, decision analysis and risk management, are increasingly applied to decision problems in which uncertainties and complexity abound. As complementary and sometimes interwoven practices, it is difficult to deἀne exact boundaries between decision analysis and risk management; thus, principles will be drawn from both for purposes of discussion.

55.2â•…Risk Management In its most general meaning, “risk” is the chance of loss of something of value, given exposure to a hazard. Risk management usually refers to systems and processes for controlling, mitigating, and monitoring risks. In some models, risk management is the stage of the overall process of risk analysis in which decisions are made about mitigating, accepting, * Wikipedia, http://en.wikipedia.org/wiki/Decision_analysis (accessed April 2009).

Decision Analysis and Risk Management

889

or controlling risks based on the information provided by risk assessments (National Research Council 1996). In other paradigms, risk management refers to the overall process that includes risk assessment, risk control, and risk communication (Claycamp 2007; ISO 31000). In personal life, organizations, and societies, risk management decisions typically evaluate the costs of the risk management decisions, personal organization or societal values, and possibly trade-offs among risks, given limitations on risk management resources. The evaluation of multiple risks and the resources to manage them means that risk management often includes a formal beneἀt–cost analysis. Given that risk management involves decisions about the need and approaches to controlling or mitigating risks, risk management often uses decision analysis to make better risk-based decisions (Paté-Cornell and Dillon 2006). Decisions about managing personal, organizational, and societal risks are frequently complex due to competing or overlapping objectives, different values, and the need for risk–risk trade-offs among the resources for managing risks. Decision analysis focuses on the elements of a decision process, while risk analysis and risk management provide theory and tools for identifying and characterizing risks. Both decision analysis and risk management draw from probability theory and they are both more Bayesian than frequentist in the application of statistical methods.* Although processes and tools for risk management and decision analysis are so intertwined that a cursory look might suggest that they are simply different perspectives on the same problem, high-quality risk management programs are in fact designed to anticipate problems before the need for a complete decision analysis arises (Paté-Cornell and Dillon 2004; Thompson 2002; Claycamp 2006, 2007). Risk and decision analysis have many applications across the pharmaceutical life cycle, from discovery research and clinical trials to individual patient decisions among therapies and even marketing withdrawal decisions (Food and Drug Administration 2006; Claycamp 2007). Two closely related uses of decision analysis are of interest in this discussion: its use in modeling or recommending various treatment options for populations of patients and the extension of such used for individual decisions among alternative therapies. These two are closely related, for the individual decision-making process is a consequence of the decision analysis used to recommend among various alternative therapies for a group of patients.

55.3â•… Elements of Risky Decisions After several decades of scholarship and practice in decisions analysis, Keeney (2004) estimated that 20% of decisions are “no-brainers,” 70% have small consequences, and, ultimately, 10% are worth thinking about. However, among the 10% of “worthy” decisions, Keeney estimated less than 1% of those decisions get systematic thought. Decisions are made in a “decision frame” or “decision context” that often includes a mixture of individual preferences, societal norms, or even an organization’s mission or business plan. A central purpose of decision analysis is to evaluate these “intangibles”—the influence they have in the decisions alongside objective statements of the decision (Keeney 1992, 2004). * “Bayesian” refers to a “degree of belief” formulation of probability opposed to a concept of probability based on frequencies of successful outcomes divided by the number of attempts. Although differing in fundamental theory, in practice, both Bayesian and frequentist schools have methodologies for projecting likelihoods of possible consequences of decisions.

890 Sudden Death in Epilepsy: Forensic and Clinical Issues

Furthermore, the process of thinking hard about a decision and working through the values can help elucidate hidden objectives in the decision (Keeney 1992). Although much of descriptive decision analysis is intuitive and derives from commonsense thinking about objectives, alternatives, and the consequences of decisions, few in treatment scenarios engage in systematic decision analysis. Yet, decisions that are worth thinking about systematically are typically the ones with high “life or death” stakes. Avoiding decision analysis simply because evaluating preferences (or utilities), judgments, probabilities, and risk tolerances is difficult leaves the outcome of the decision purely to chance and improves the chances that human tendencies to be “predictably irrational” will be realized (Ariely 2008). All decisions have at least two possible outcomes—hence, the need for decision. However, more precisely, decisions worth analyzing include • • • • •

A decision problem, or the decision context Objectives that specify what the decision maker(s) is trying to achieve Alternatives Consequences (clinical outcomes) Trade-offs—balancing the pros and cons of the various alternatives

Although the basic elements are commonly understood, formal analysis in complex decision problems typically includes identifying and quantifying the uncertainties affecting the decision, measuring the decision makers’ risk tolerance, and linking sequential decisions to evaluate the impact of decisions on the deἀnition of future decisions. There are many opportunities to apply decision analysis to selecting an optimal treatment from among alternatives. The decisions range from patient-level, evidence-based medicine to analyzing public policy-level decisions about the optimal choice of treatments for reimbursements (Man-Son-Hing et al. 2000; Vickers and Elkin 2006; Wilson et al. 2008). The formality of individual decision analysis likely to be a matter of individual skills and preferences and the formality of health care provider’s decision analysis is likely to be influenced by both business objectives and access to decision analysts. Public policy analysis depends on the agency, their regulatory or program mandates, and the assembled expertise for complex decision analysis. A common element among all of these scenarios for the use of decision analysis is that the quantitative tools and the analysts cannot replace the expertise of health care practitioners who are face-to-face with the patient. Rather, decision analysis is a collaboration that provides patients impacted by the decision(s) and their clinicians an analytical framework for comparing alternatives under uncertainty. A treatment decision can be modeled as either a decision among various treatment alternatives with binary (e.g., “cured” or “not cured”) or multilevel consequences (Figure 55.1). Similar to clinical trial settings, diagnostic decisions, or epidemiological constructs, the chance variable in the decision could be “diseased” or “not diseased,” and odds ratios are calculated (Vickers and Elkin 2006) in which the probabilities p1 and p2 in the illustration are identical. Alternatively, a known disease state might be the prior condition for decisions among treatment alternatives for which the uncertainties are the patient’s individual response to a particular treatment. The likelihood of a given response or outcome would be estimated from population-based studies or clinical trials and used in calculating the expected values along each branch of the decision tree (Man-Song-Hing et al. 2000; Vickers and Elkin 2006; van Dijk et al. 2008).

Decision Analysis and Risk Management Alternatives

Chance p1

L1

891 Outcomes x1

(1 – p1 )

x2

p2

x3

L2

(1–p 2)

x4

EV ( L1 ) = p1 x1 + (1 − p1 ) x2 EV ( L2 ) = p2 x3 + (1 − p2 ) x4

Figure 55.1╇ Typical decision tree representation of choice between two alternatives (L1 and L 2). Decision trees depict decision nodes as squares and chance nodes as circles. Expected values (EV) of each alternative are shown as linear combination of outcomes and respective probabilities. Here, chance nodes are simple binary probabilities. More complex decisions might have many possible levels of outcomes.

The uncertainties about the effectiveness and safety of alternatives, in addition to the patient’s preferences or utilities for various quality-of-life states complicate the analysis, creating a multiobjective decision (Figure 55.2, discussed below). Difficult decisions almost always derive much of their challenge from uncertainties because, like any prognostication, the outcomes for a speciἀc patient cannot be foretold. Both decision analysis and statistical modeling generate prognostic models for supporting decisions (e.g., Harrell et al. 1996; van Dijk et al. 2008; Whitney et al. 2008). Perhaps a key difference between classic decision analysis and statistically based decisions between alternative hypotheses is that Clinical trials; studies Safety

Decision analysis with multiple objectives Probabilities Prob(S|Ti )

T1 T2 T1

Efficacy

Probabilities Prob(E|Ti )

T2 T1

QALY

Costs

Probabilities Prob(Q|Ti ) (Probabilities) Prob(C|Ti )

T2 T1

...

T2

Weights: Maximize safety

k1

Maximize efficacy

k2

Maximize quality of life

k3

Minimize costs

k4

Figure 55.2╇ A multiobjective decision between treatment alternatives T1 and T2. In this model,

each objective is shown in a separate tree for clarity of presentation. Study information from safety, efficacy, and quality-adjusted life years (QALY) studies are likely to inform probabilities for chance nodes (circles). Information about costs might be in the form of probabilities or deterministic values. To trade off objectives for a solution to the decision problem, the decision maker assigns value weights, ki.

892 Sudden Death in Epilepsy: Forensic and Clinical Issues

probabilities of various alternatives in decision analysis were historically modeled as discrete values rather than distributions due to the challenges of multiobjective optimizations (Haimes 2009).

55.4â•… Preferences for Outcomes and Probabilities A major difference between statistical modeling in comparison to clinical trials and decision analysis for decisions among alternative treatments is that preferences for speciἀc outcomes or consequences are included as factors affecting the decision. The patient, as decision maker, chooses among recommended treatments under multiple objectives possibly including efficacy, safety, and the quality of life expected from the same decision problem.* In addition to personal experience and values, social norms and other “soft” factors that contribute to the patient’s objective function in a decision, product marketing strategies are intended to influence preference for the marketer’s product independently of actual differences in expected efficacy and side effects compared to alternative treatments. In the population-based decision for health care policies, economics in addition to efficacy, safety, and population preferences are evaluated. Ultimately, evaluating preferences for uncertain consequences under multiple objectives, in a constantly evolving science of medicine, are the reasons why these decisions are “hard” (Clemen 1996). Patient (or a population of patients) preferences for particular outcomes are expressed in utilities or value functions. Various questionnaires, games, and other elicitation strategies have been devised to quantify patients’ utility functions. Finding the parameters of a utility function is difficult to accomplish in most settings and it is likely to be further challenged by communication of quantitative scientiἀc information to a lay audience. Differential perception of risk and known flaws in human judgments of probabilities further complicates general solutions to modeling patient preferences (Tversky and Kahneman 1974; Slovic 2000). Typically, a power function is used to represent the decision maker’s preference for outcomes. Modeling of decision maker’s preferences has for decades taken its lead from Von Neumann and Morgenstern’s (1947) application of expected utility theory in game theory. Expected utility, as used by Von Neumann and Morgenstern, states that a decision can be described by the decision maker’s behavior of maximizing the expected utility following a simple set of rational axioms. Decisions by rational decision makers are modeled as a gamble between/among two or more alternatives. The theory describes a lottery between alternatives A and B, for which the probability for alternative A is p and that of B is (1 − p), or

L = pA + (1 – p)B.

More generally speaking, the alternatives can be expressed in a linear combination of the probabilities and their outcomes: L = ∑piAi, where pi is the probability of alternative Ai. The Von Neumann–Morgenstern theory identiἀed four axioms of rational decision making. Given simple lotteries (gambles), L1 and L2, two “weak order” axioms include:

* Drug approvals are ultimately group decision processes in which efficacy and safety are among the objectives in the decision whether to approve a drug.

Decision Analysis and Risk Management

893

• Completeness: for any two lotteries, either L1 ⩾ L2 or L2 ⩾ L1 • Transitivity: if L1 ⩾ L2 or L2 ⩾ L3 then L1 ⩾ L3 where “A > B” means “A is strictly preferred to B” and “A ⩾ B” means “A is preferred or equivalent to B.” The two remaining, complicated axioms include the • Archimedean axiom, in which if L1, L2, L3 ∈ Δ(X)such that L1 > L2 > L3, then there is an α,β ∈ (0,1) for combining the lotteries such that αL1 + (1 – α)L3 > L2 and L2 > βL1 + (1 – β)L3 • Independence axiom, in which for all L1, L2, L3 from the set of alternatives, and any α ∈ [0,1], then L1 ⩾ L2 if and only if αL1 + (1 – α)L3 ⩾ αL2 + (1 – α)L3 The difficulty with expected utility theory before Von Neumann and Morgenstern was the lack of a theory to explain the decision maker’s preferences (utilities) over the outcomes in a consistent and quantitatively meaningful way. For example, if the preference for one side-effect-free treatment over another is 5 “utils,” then how many utils is the preference for choosing a sports car over a sedan? Von Neumann and Morgenstern showed, counterintuitively, that a decision maker’s preferences are understood and modeled consistently if they are preferences over the lotteries (e.g., the probabilities), not the outcomes. Ultimately, preferences (utilities) over the outcomes, U(xi), can be inferred by gaining understanding of the decision maker’s preferences over the lotteries in a given decision context. Utility functions are not derived from ἀrst principles but are functions chosen empirically based on the preferences elicited from the decision makers themselves or are simply approximated using functions of the appropriate shape (Keeney 1992). Generally, there are three classes of utility function that capture the decision maker’s risk tolerance, including utility curves for • Risk averse decision makers of the general concave functional form, –e–αx • Risk neutral decision makers U(x) = x • Risk seeking decision makers, of the general convex form, e–αx A simple interpretation of risk tolerance in medical decision making is that the risk averse patient tightly holds a known outcome with known side effects that engages in a treatment and uncertain lottery to signiἀcantly improve the outcome but at risk of worse side effects. The relatively risk-seeking patient would be expected to give up the known outcome for even a small chance at signiἀcant improvements in the face of potentially worse side effects. Although health care providers have always had a major role in guiding patients through difficult decisions often having complicated trade-offs, increasingly accessible tools of decision analysis can serve as visual aids to help the patients comprehend differences that various trade-offs have in affecting the outcomes of their decisions. Health care providers have a role guiding patients directly but also in ensuring that the self-guided decision tools on internet health sites present objective information as input to the patient’s decision making. There are many other elements for modeling utilities of decision makers that are beyond this discussion. Except for a few modiἀcations and some related theories such as ambiguity as opposed to risk avoidance in stating preferences, expected utility theory has guided much of decision analysis for decades. However, perhaps one of the most important

894 Sudden Death in Epilepsy: Forensic and Clinical Issues

descriptive theories to further explain the decision maker’s judgments leading to preferences over the probabilities is prospect theory, in which it was shown that decision makers might be risk seeking at low probabilities of occurrence and risk averse at high probabilities (Kahneman and Tversky 1979; Tversky and Kahneman 1992). This work followed and was conἀrmed by many studies on judgments of probabilities and risk perception that show human difficulties in judging the probabilities of low-probability events and underestimate the probabilities of high probability events. Judgment flaws and the influence of decision making behaviors that run counter to classic, rational decision theory are discussed for general audiences in Predictably Irrational. The Hidden Forces that Shape Our Decisions (Ariely 2008). Outcomes or consequences in the discussion thus far have been univariate in nature. In reality, a disease treatment decision is unlikely to have only one objective, because the treatment comes with costs—both in terms of the charge for the treatment and in possible health consequences—and uncertain levels of beneἀcial outcome (efficacy). For instance, the decision maker’s objective might include maximizing the chance for life, minimizing the ἀnancial costs, and minimizing treatment-related discomfort. Clearly, there might be three or more objectives in this multiobjective decision (Figure 55.2). In reality, the decision maker must specify value weights or “scaling constants” (ki) for trading-off the objectives.* For example, a decision with two objectives and corresponding utilities, u1(x1) and u2(x2), has an overall utility, u(x1,x2), expressed as the linear combination (Keeney 1992),

u(x1,x2) = k1u1(x1) + k2u2(x2) + k3u1(x1)u2(x2).

If it is shown that additive independence of the utilities u1 and u2 exists, then the third term in the additive equation is omitted. There are several approaches commonly used to discover hidden objectives revealed by ἀnding incomplete independence of the component utilities. Generally, trade-off analysis, in which indifferences (equal utilities) are calculated among at least two terms, can reveal the presence of an underlying objective (Keeney 1992; Keeney and Raiffa 1993; Clemen 1996; Haimes 2009).

55.5â•…Uncertainties Outcomes in medical decision making are seldom known with certainty (Man-Son-Hing et al. 2000; Shakespeare et al. 2001). Uncertain states of nature (or chances in lotteries) include the probability of invasive disease revealed only after a decision to treat using surgery (Vickers and Elkin 2006) or the chance that a patient might not tolerate the chosen drug treatment. Choosing one treatment over another is an example of expected utility decision in which the probabilities over the consequences are different for the two treatments. Moreover, these are multiobjective decisions (Keeney and Raiffa 1993) for which objectives might include maximizing chances for a cure, minimizing chances of side effects, and minimizing time in inpatient care. Speciἀc measures for the attainment of * The k i are referred to a scaling constants to convey that they rescale sometimes differing independent objective scales to 0 → 1 for convenience and consistency of the analysis. The k i can more appropriately be referred to as “value weights” in a multiobjective value analysis, v(x1 ,x 2), where the individual terms are value functions as opposed to utility functions.

Decision Analysis and Risk Management

895

objectives (a.k.a. “attributes”) can be deἀned some of the attribute values can be informed using predictive models from population studies. For example, “cure” might be deἀned as “disease-free at 5 years after treatment” and scaled as a dichotomous variable in the set, [0,1]. Uncertainties in decision analysis are frequently modeled as discrete values. This simpliἀes the mathematical optimization problem for the quantitative analysis of the decision; however, in cases relying on population studies, it can overlook useful information. On the other hand, the value of including imperfect information studies has shown that, in some cases, modeling with symmetrical distributions can cancel the effect of including uncertainty (Morgan and Henrion 1990). Partly a result of difficulty in optimizing the decision under the continuous probability, some analyses use probability represented by discrete values for, e.g., the mid-range of a portion of the continuous distributions (Haimes 2009). Contemporary simulation approaches, such as Markov chain Monte Carlo methods, have greatly simpliἀed using more complex relationships. Thus, continuous distributions are appearing more commonly in decision models.

55.6â•… Benefits and Quality-of-Life Objectives Evaluation of the quality of life after treatment as an outcome of a treatment decision is useful for both population-based and individual decision analyses (Neumann et al. 2000; Hammitt 2002). Health policy analysts seek an understanding of how individuals value health as input for cost–beneἀt analyses (CBA) in health policy decisions. If valuation can be accomplished, then traditional CBA supports a simple decision rule: ἀnding that a net beneἀt for a health care program or treatment alternative exceeds its net costs means that the program should be adopted (or a treatment should be chosen). In the cost–beneἀt model, valuation of treatment alternatives for decision making requires a monetary value of life or, in other words, the economic production potential of “human capital.” The valuation exercise often generates controversy about the appropriate value of a life and also whether human preferences for health outcomes can be adequately modeled for the decision analysis (Hirth et al. 2000; Neumann et al. 2000, 2009; Hammitt 2002). Cost-effectiveness analysis circumvents some of the challenges of valuation in CBA, making it a preferred method to CBA for economic evaluation of health care policies (Neumann et al. 2000). Cost-effective analysis compares net costs for the net health beneἀts achieved for a speciἀc treatment compared to a speciἀc alternative treatment. The cost-per-effect ratio can be used to compare intermediate health states in terms that are relevant to speciἀc diseases and treatments, making it also useful for communicating health care policies or recommendations to clinicians in a given discipline (Mortimer and Segal 2008; Cohen and Neumann 2008). On the other hand, comparisons across diseases are difficult to accomplish using CEA because the intermediate outcomes of treatment for a speciἀc disease, e.g., cancer, can differ substantially from those in, e.g., epilepsy treatment. Furthermore, the relief from difficulties assigning monetary values to life in CBA comes at a sacriἀce of simplicity in a decision rule: there is no generally applicable decision rule in CEA. Although the direct valuation of life can be avoided in CEA, a measure of the relative beneἀt to at least intervals of lifetime after treatment is needed to close the beneἀts loop of a decision analysis. Policy makers and researchers sometimes use observational studies (Mortimer and Segal 2008) and willingness-to-pay methods (Hirth et al. 2000) to infer the

896 Sudden Death in Epilepsy: Forensic and Clinical Issues

dollar beneἀt per effective change in health status after a given treatment to derive a decision rules for deἀne problems. To model the relative beneἀt of treatment alternatives and the patient’s preferences over the various health outcomes as needed for CEA, the quality-adjusted life-year (QALY) has become a commonly used measure in health care evaluation studies (Neumann et al. 2000; Hirth et al. 2000; Brixner et al. 2009). In its simplest form, the patient’s preferences for different levels of quality of life generally assumes her utility function is zero for death and 1 for perfect health. A state of health quality that is judged to be a “fate worse than death” is sometimes given a negative quality weight by analysts. QALY analysis assumes that an individual moves through states of health over a lifetime and each of those states can be weighted with respect to the relative quality relative to death and perfect health. Although health status can improve or decline from one interval to the next, the summation over the remaining lifespan after treatment and without treatment is an integral measure of the beneἀt for the treatment of the disease. QALY is calculated simply as n

QALY =

∑q T , i i

i =1



where the quality of life during interval i is weighted as qi for the interval duration, Ti. For population-based analyses, the individual values can be summed across the population. As is the case for measures of preference or utility, QALYs need to be utility independent (Keeney 1992), risk neutral, and have constant proportional values for trade-off analysis. The latter concept means that a person would be willing to trade some fraction of their lifespan for an improvement in quality from one level to a preferred level fraction of lifespan would depend only on the differences in the level of QALYs and not the expected lifespan at the beginning of the period (Neumann et al. 2000). Risk neutrality means that the utilities are directly proportional to longevity for a ἀxed level of quality. However, if assumptions of utility independence (e.g., utility for the length of life and the quality of life) and constant proportionality hold, then QALYs can be modeled using general riskadjusted utility models.

55.7â•… Weighting the Quality of Life There is yet to emerge a single model for estimating weights (q) for QALY calculations. Direct questioning of patients is difficult for a variety of reasons, including the notion that patients who are ill over-weight the value of not having an existing condition. Sometimes psychometric tests are used to quantify relative health and mental status; however, these tests do not address the patient’s preferences for various health states—the fundamental core of a decision analysis. Additionally, prediction of QALYs from statistical analysis of descriptive health outcomes in, e.g., clinical trials remains challenging and suggests that algorithms need to ἀt speciἀc disease conditions (Mortimer and Segal 2008). Methods for scaling individual preferences for a health-related quality level in QALY analysis included direct elicitation using the standard gamble, time trade-off, visual analogue scaling, and person trade-off methods (Hammitt 2002). The standard gamble follows decision–theoretic modeling useful in everyday decisions (Hammond et al. 1999), in which

Decision Analysis and Risk Management

897

the respondent speciἀes the smallest chance of survival he or she could accept in a lottery where the alternative outcome is immediate death. This approach ἀts many life and death medical decisions but might be more difficult to apply in instances in which death is not a probable outcome for either lottery. For those instances, a time trade-off method might be preferable. All procedures are further complicated by the fact that the quality of life possibly changes from one period to the next (Figure 55.3). Time trade-off methods call for the patient to state the number of years in perfect health (q = 1) he or she considers equivalent to a speciἀc duration of chronic health condition. For example if the patient is indifferent between 10 years with a mild condition and 5 years of perfect health, then q = 5/10 = 0.5. The QALYs for either condition for which the patient is indifferent are 5 QALYs. The person trade-off method externalizes the preference framework by asking the respondent to consider the value of improving the health of groups in different health states (Hammitt 2002). Like the previous two methods, an indifference point is sought from which to estimate the QALY. For example, if the individual were asked to choose between extending the life of 100 healthy individuals for a year and extending the life of partially paralyzed individuals for a year, at what number (n) of partially paralyzed individuals would a decision maker be indifferent about the choice? The health-related quality level for living with partial paralysis would be 100/n. Another common approach for directly eliciting preferences for health outcomes is the visually adjusted scales. Visual methods call for asking the respondent to place various health outcomes along a linear scale from death to perfect health (Hammitt 2000; Neumann et al. 2000). Alternatively, preferences can be elicited using probability-based visualizations, such as those used in a comparison probability (gamble) trade-off techniques by Feldman-Stewart and colleagues (2007). Finally, willingness to pay (WTP) is an alternative approach to QALY for valuing health status in various intervals and health states after an event (Hirth et al. 2000; Hammitt 2002). WTP is often preferred to QALY in environmental risk analysis (Hammitt 2002). The essential key difference between WTP and QALY is that WTP is based on the amount of wealth (income) that an individual is willing to forego or pay to avoid a particular level of health decrement or, alternatively, “willing to pay” to improve the health statues to a Health related quality level (q)

1.0

0

0

Time after treatment

DC

DR

Figure 55.3╇ Hypothetical expected quality of life after disease treatment. Treatment alternative (dashed line) leads to longer life (DR) than “no treatment” option (solid line). Quality of life immediately after treatment is worse for treatment than control; however, the summed quality of life over periods—the QALYs—shows that treatment not only extends life but also increases the average.

898 Sudden Death in Epilepsy: Forensic and Clinical Issues

particular state of health in the intervals of interest. Although both QALY and WTP seek to support a clear beneἀts analysis and to avoid the controversy of directly valuing life, once incremental values are placed on either the WTP or QALYs, the value of a statistical life can be inferred from the data (Hammitt 2002; Alberini 2005).

55.8â•… Example Decision Tree Decisions can be analyzed to the degree of skills and interests of the decision maker and using tools that range from pencil and paper to sophisticated software. Various approaches at simplifying complex decisions and presenting the results visually have been published and some examples are given in Table 55.1 and Figure 55.4. The decision tree is often used because it can be drawn to illustrate all elements of the decision and provide a framework for calculations. This example will use a decision tree (Figure 55.5) and the discussion and comparison of odds ratio approaches in Lathers et al. (2003). Figure 55.5 shows a possible decision tree for the choice among antiepileptic drugs (AEDs), including both positive and negative consequences. The decision is modeled as the choice (square node) among alternative AEDs followed by the chance (circular nodes) of improvement, given the drug selection. The side effect of “sedation” is modeled as a subsequent chance node having probabilities of sedation that are also drug dependent. The numbers at the ends of the nodes are the outcomes or consequences. Figure 55.5 depicts a simplifying assumption that sedation is not a factor for the decision maker given a failure to improve after initiating treatments that do not achieve improvements. Although the onset of side effects can begin with the beginning of drug treatment, the depicted model assumes that the ineffective (and its side effects) treatment would cease in favor of an alternative. Moreover, the simpliἀed model in Figure 55.5 does not capture the fact that treatment regimen can be altered deliberately with duration of treatment, i.e., while titrating Table 55.1â•… Typical Decision Analysis and Visualizations for Treatment Decisions Authors

Method

Type of Presentation

Otoul et al. 2005

Two-way plots: odds ratios for withdrawal vs. QR responder

Visual

Vickers and Elkin 2006

“Decision curve.” Variation on receiver operator characteristics calculated as the net beneἀt against the probability of disease among patients Classic meta-analysis calculating odds ratios and conἀdence interval estimates Risk–beneἀt curves: adverse events and efficacy probability contours Pairwise comparisons of odds ratios Odds ratios in four directions Graphical decision analysis for studying perception

Visual and table analysis

Song et al. 2003 Shakespeare et al. 2001 Caldwell et al. 2005 Lathers et al. 2003 Feldman-Stewart 2007

Visual plots of odds ratios ± conἀdence intervals Visual—contours Table Visual—plot Visual qualitative presentation

Includes Uncertainties in Analysis? Yes; classic conἀdence interval estimates uncertainties Discrete and distributions for the disease uncertainty Yes; classic conἀdence intervals Yes—discrete in contour levels Yes No No

Decision Analysis and Risk Management Relative OR 1.0 0.8 0.5

899

Drug A Drug B Drug C

0.3 Placebo

0.0

Drug

Relative sample size

Figure 55.4╇ Four-way plot of a decision problem comparing three alternative drug treatments for epileptic seizures. The largest relative odds ratio for each two-way drug efficacy comparison is used as denominator for the “Relative OR” axis. Drug treatment improvement rate appears on right horizontal axis, placebo rate for the same study on left horizontal axis, and relative sample size on lower vertical axis. The larger the area under similar placebo rates, the better the comparison. (From Lathers, C. M. et al., J Clin Pharmacol, 43, 491–503, 2003. With permission.)

the patient’s tolerance for side effects. Although beyond the scope of this chapter, linked or multistage decisions, in which the outcome of one decision is an input to the next, can be represented visually and in basic calculations. For example, Hammond et al. (1999) discussed the features of linked decisions for everyday decision making. The calculations that the decision tree represents follow from the expected value formulae shown in Figure 55.1. Tracing backward on the topiramate branch, the expected value of the sedation lottery, given that improvement has occurred from the choice of topiramate, is the sum of the probabilities of the possible outcomes times the values of their respective outcomes or consequences. As discussed above, the outcomes may include a vector of beneἀcial and deleterious consequences, such as physical health, emotional wellbeing, loss/maintenance of income, and other factors. The consequences must be in the same value units, such as dollars, to complete the expected value calculations. Given that the branches of the example tree are identical in structure, the expected value (EV) of a branch for drug alternative a is given by:

{



}

 psed ,a w °T (1 + Qepi )/ 2 − ∆Qsed  − Ca − Csed ,a +      EVa = pI,a *   (1 − psed ,a )* wT (1 + Qepi )/ 2 

where pI,a = probability of improvement, given treatment with drug a psed,a = probability of sedation, given drug a T = duration in years (or remaining lifetime, if applicable) Qepi = quality loss per year due to disease (epilepsy)

900 Sudden Death in Epilepsy: Forensic and Clinical Issues Sedation Improvement

45%

Topiramate net outcome No sedation

Topiramate

55% 0 Sedation

Improvement

27%

7%

Lamotrigine net outcome No sedation

93%

Lamotrigine treatment –135 Neutral

AED

80%

Topiramate treatment –177 Neutral

Lamotrigine

20%

73%

AED choice 19% –2500 Phenobarbital net outcome Sedation

Improvement

58%

No sedation Phenobarbital

81% 5000

Phenobarbital treatment –100 Neutral

42%

Figure 55.5╇ Typical decision tree for the choice among three AEDs including sedation side effect. The example from Lathers et al. (2003) was used to develop the decision tree; thus, probabilities expressed as percentages along the branches are average response rates from Tables VI and VII in Lathers, C. M., et al. (2003). See text for assumptions used in developing the decision tree, here shown using Precision Tree Ⓡ software for ExcelⓇ.

ΔQsed = quality decrement (QALYs lost) due to sedation Ca = cost of treatment for drug alternative a Csed = cost of treating sedation w = value weight of a unit of quality adjusted life year The expected beneἀt of one drug treatment alternative compared to a second is the difference between the expected values of the branches, e.g., EVtopiramate – EVphenobarbital. For example, if the values for the equation are substituted from Tables VI and VII in Lathers et al. (2003) and we arbitrarily assign w = $75,000 per year; Qepi, = 0.3 per year; ΔQsed = 0.3; Ca€ = $177, $135, and $100 per month for topiramate, lamotrigine, and phenobarbital, respectively (Lathers et al. 2003); and T = 5 years, then the beneἀt of topiramate is about 24% less and the beneἀt of lamotrigine about 53% less than that for phenobarbital.

Decision Analysis and Risk Management

901

Of course, this example considers only seizure improvement, costs, and sedation rates. Including other side effects would change and possibly reverse the outcome. Once the branches of the decision tree have been calculated, derived measures in addition to net beneἀt are useful when comparing alternative treatments. In particular, the expected value of perfect information for a decision is useful in indentifying optimal strategies for obtaining additional information. This is the difference between the highest expected net beneἀt in the decision and the expected value of the decision, given the uncertainties. For the static, simple model shown in Figure 55.5, the expected value of perfect information for each branch is simply the difference between the inner terms in the branch equation above. However, the expected value of perfect information is maximally useful when the uncertainties of the probabilities and possibly other terms in the equation are uncertainty distributions. Then, the relative values of reducing uncertainty for each term can be estimated from the distributions (Clemen 1996; Ades et al. 2004; Willan 2007). The sources of the uncertainties are the clinical trial statistics (e.g., the improvement rate or probability ± error) and judgments about the appropriate form of the uncertainty distribution. Often, beta distributions in the range 0 to 1 are used for modeling uncertainty in probabilities along the branches (Ades et al. 2004). Finally, contemporary software for decision modeling affords decision analyses the efficient tools for incorporating uncertainty distributions in the calculations. In a Bayesian approach to modeling and analyzing the decision, the starting parameters, such as the branch probabilities, can be estimated from prior knowledge—the outcomes of clinical trials or public health follow-on studies such as those used by Lathers and colleagues (2003). Because the number of subjects in a given trial affect the relative uncertainty in the parameters, simulations of the decision using distributions of the decision variables will reflect the relative information or “strength” of the data supporting the overall calculation. Although computationally and mathematically more complex, the solutions of comparative therapies using full uncertainty analysis yields richer information for decision support than the back of the envelope methods. Again, the ultimate choice of a computation method for the decision analysis depends on the purpose of the decision (public health or individual?) and the needs of the decision maker for either rigorous or informal decision support.

55.9â•… A Structured and Focused Way of Making Difficult Choices This chapter presents some of basic elements of decision analysis. There are numerous advances in the practice of decision analysis that are used to scale decision analysis from personal decisions (Howard et al. 1999; Glasziou 1995) to high-impact health care policy (Devlin and Parkin 2004; Brixner et al. 2009) or even environmentally contentious decisions (Linkov et al. 2006). Across the diverse applications of decision analysis, the basic principles of linking objectives, chances, consequences (or outcomes), trade-offs, and risk tolerance in a structured thinking process is central to creating robust decision support and to selecting a single treatment from among many. Ultimately, there might be as many ways to approach analyzing the decision problem as there are personal, organizational, or societal values and preferences for beneἀcial health outcomes for balanced costs. The structured or “focused thinking” processes described by many (Keeney 1992; Clemen 1996; Linkov et al. 2006; Haimes 2009) provide ways to include the preferences, concerns,

902 Sudden Death in Epilepsy: Forensic and Clinical Issues

and values of the decision makers with the quantitative observations from clinical trials of safety and efficacy for optimum choices of treatment. This chapter emphasizes the importance of decision analysis in resolving uncertainties, risk and value trade-offs, and dealing with linked decisions over an extended period of treatment. Difficult decisions, depending on many variables and uncertainties that might affect outcomes, are a continuing challenge for health care providers treating persons risk for SUDEP. The optimal course of treatment of patients at risk for SUDEP necessitates establishing a balance among efficacies, side effects, and the possible occurrence of SUDEP (Lathers et al. 2003). Decision risk analysis as it applies to risk management is a process that can beneἀt in determining modiἀcation of risk factors for SUDEP.

References Ades, A. E., G. Lu, and K. Claxton. 2004. Expected values of sample information calculations in medical decision making. Med Decis Making 24: 207–226. Alberini, A. 2005. What is a life worth? Robustness of VLS values from contingent valuation surveys. Risk Anal 25: 783–800. Ariely, D. 2008. Predictably Irrational. The Hidden Forces that Shape Our Decisions. New York, NY: Harper-Collins. Brixner, D. I., A. P. Holtorf, P. J. Neumann, D. C. Malone, and J. B. Watkins. 2009. Standardizing quality assessment of observational studies for decision making in health care. J Manag Care Pharm 15: 275–283. Caldwell, D. M., A. E. Ades, and J. P. T. Higgins. 2005. Simultaneous comparison of multiple treatments: Combining direct and indirect evidence. BMJ 331: 897–900. Claycamp, H. G. 2007. Perspective on quality risk management of pharmaceutical quality. Drug Inf J 41: 353–367. Claycamp, H. G. 2006. Rapid beneἀt-risk assessments: No escape from expert judgments in risk management. Risk Anal 26: 135–146. Clemen, R. T. 1996. Making Hard Decisions. An Introduction to Decision Analysis. 2nd ed. Paciἀc Grove, CA: Duxbury Press. Cohen, J. T., and P. J. Neumann. 2008. Using decision analysis to better evaluate pediatric clinical guidelines. Health Affairs 27: 1467–1475. Devlin, N., and D. Parkin. 2004. Does NICE have a cost-effectiveness threshold and what other factors influence its decisions? A binary choice analysis. Health Econ 13: 4437–4452. Feldman-Stewart, D., M. D. Brundage, and V. Zotov. 2007. Further insight into the perception of quantitative information: Judgments of gist in treatment decisions. Med Decis Making 27: 34–43. Food and Drug Administration (FDA). 2006. Guidance for Industry Q9 Quality Risk Management. CenÂ� ter for Drug Evaluation and Research (CDER) and Center for Biologics Evaluation and Research (CBER). http://www.fda.gov/downloads/drugs/GuidanceComplianceRegulatoryInformation/ Guidances/ucm073511.pdf (Accessed February 1, 2009). Funtowicz, S. O., and J. R. Ravetz. 1992. Risk management as a postnormal science? Risk Anal 12:€95–97. Glasziou, P. P. 1995. An evidence based approach to individualizing treatment. BMJ 311: 1356–1359. Haimes, Y. Y. 2009. Risk Modeling, Assessment, and Management. New York, NY: John Wiley & Sons, Inc. Hammond, J. H., R. L. Keeney, and H. Raiffa. 1999. Smart Choices. New York, NY: Broadway Books. Hammitt, J. K. 2002. QALYs versus WTP. Risk Anal 22: 985–1001.

Decision Analysis and Risk Management

903

Harrell, Jr., F. E., K. L. Lee, and D. B. Mark. 1996. Multivariable prognostic models: Issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 15: 361–387. Hirth, R. A., M. E. Chernew, E. Miller, A. M. Fendrick, and W. G. Weissert. 2000. Willingness to pay for a quality-adjusted life year: In search of a standard. Med Decis Making 20: 332–342. Hocine, M. N., P. Tubert-Bittner, T. Moreau, M. Chavance, E. Varon, and D. Guillemot. 2007. Relativerisk ratio was a useful measure of differential association in cohort and case-series studies. J Clin Epidemiol 60: 361–365. Kahneman, D., and A. Tversky. 1979. Prospect theory: An analysis of decisions under risk. EconÂ� ometrica 47: 263–291. Keeney, R. 1992. Value-Focused Thinking. Cambridge, MA: Harvard University Press. Keeney, R. 2004. Making better decision makers. Decis Anal 1: 193–204. Keeney, R. L., and H. Raiffa. 1993. Decisions with Multiple Objectives: Preferences and Value Tradeoffs. Cambridge, UK: Cambridge University Press. Lathers, C. M., P. L. Schraeder, and H. G. Claycamp. 2003. Clinical pharmacology of topiramate versus lamotrigine versus phenobarbital: Comparison of efficacy and side effects using odds ratios. J Clin Pharmacol 43: 491–503. Linkov I., F. K. Satterstrom, G. Kiker, C. Batchelor, T. Bridges, and E. Ferguson. 2006. Comparative risk assessment to multi-criteria decision analysis and adaptive management: Recent developments and applications. Environ Int 32: 1072–1093. Man-Son-Hing, M., A. Laupacis, A. M. O’Connor, D. Colyle, R. Berquist, and F. McAlister. 2000. Patient preference-based treatments thresholds and recommendation: A comparison of decision-analytic modeling with the probability-tradeoff technique. Med Decis Making 20: 394–402. Morgan, M. G., and M. Henrion. 1990. Uncertainty. A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis. Cambridge, UK: Cambridge University Press. Mortimer, D., and L. Segal. 2008. Comparing the incomparable? A systematic review of competing techniques for converting descriptive measures of health status in QALY-weights. Med Decis Making 28: 66–89. Neumann, P. J., S. J. Goldie, and M. C. Weinstein. 2000. Preference-based measure in economic evaluation in health care. Annu Rev Public Health 21: 587–611. Neumann, P. J., P. D. Jacobson, and J. Palmer. 2009. Measuring the value of public health systems: The disconnect between health economists and public health practitioners. Am J Pub Health 98: 2173–2180. National Research Council. 1996. Understanding Risk: Informing Decisions in a Democratic Society, eds. P. C. Stern and H. V. Fineberg. Washington, DC: National Academy Press. Otoul, C., C. Arrigo, K. van Rijckevorsel, and J. A. French. 2005. Meta-analysis and indirect comparisons of levetiracetam with other second-generation antiepileptic drugs in partial epilepsy. Clin Neuropharmacol 28: 72–78. Paté-Cornell, M. E., and R. L. Dillon. 2006. The respective roles of risk and decision analyses in decision support. Decision Anal 3: 220–232. Shakespeare, T. P., V. J. Gebski, J. Veness, and J. Simes. 2001. Improving interpretation of clinical studies by use of conἀdence levels, clinical signiἀcance curves, and risk–beneἀt contours. Lancet 357: 1349–1353. Slutsky, J. R., and C. M. Clancy. 2009. AHRQ’s Effective Health Care Program: Why comparative effectiveness matters. Am J Med Qual 24: 67–70. Slovic, P. 2000. The Perception of Risk. London, UK: Earthscan Publications. Song, F., D. G. Altman, A. M. Glenny, and J. J. Deeks. 2003. Validity of indirect comparison for estimating efficacy of competing interventions: Empirical evidence from published meta-analyses. Br Med J 326: 472–477. Thompson, K. M. 2002. Variability and uncertainty meet risk management and risk communication. Risk Anal 22: 647–654.

904 Sudden Death in Epilepsy: Forensic and Clinical Issues Tversky, A., and D. Kahneman. 1974. Judgment under uncertainty: Heuristics and biases. Science 165: 1232–1238. Tversky, A., and D. Kahneman. 1992. Advances in prospect theory: Cumulative representation of uncertainty. J Risk Uncertainty 5: 297–323. van Dijk, M. R., E. W. Steyerberg, and J. D. F. Habbema. 2008. A decision-analytic approach to deἀne poor prognosis patients: A case study for non-seminomatous germ cell cancer patients. BMC Med Inform Decis Mak 8: 1–9. Willan, A. R. 2007. Clinical decision making and the expected value of information. Clin Trials 4: 279–285. Whitney, S. N., M. Holmes-Rovner, and C. H. Brody et al. 2008. Beyond shared decision making: An expanded typology of medical decisions. Med Decis Making 28: 699–705. Vickers, A. J., and E. B. Elkin. 2006. Decision curve analysis: A novel method for evaluating prediction models. Med Decis Making 26: 565–574. Von Neumann, J., and O. Morgenstern. 1947. Theory of Games and Economic Behavior, 2nd ed. Princeton, NJ: Princeton University Press. Wilson, E. C. F., S. J. Peacock, and D. Ruta. 2008. Priority setting in practice: What is the best way€to€compare costs and beneἀts? Health Economics. 17 Jun 2008 Publ online http://www3╉ .interscience.wiley.com/journal/5749/home.

Epilepsy Surgery and the Prevention of SUDEP Ryan S. Hays Michael R. Sperling

56

Contents 56.1 Introduction 56.2 Methods 56.2.1 Selected Studies 56.2.2 Study Design 56.2.3 Calculating Mortality 56.2.4 Duration of Follow-Up 56.3 Results 56.3.1 Effect of Postoperative Seizure Control on Mortality 56.3.2 Effect of Laterality of Surgery on Mortality 56.3.3 Effect of Gender of Surgical Patients on Mortality 56.3.4 Effect of Pathologic Diagnosis on Mortality 56.3.5 Effect of Cerebral Lobe on Mortality 56.3.6 Comparison of Resection/Subpial Transection with Callosal Section 56.4 Conclusion References

905 906 906 907 907 907 908 908 910 911 911 911 911 912 912

56.1â•…Introduction Epilepsy can occasionally be a fatal condition. People with epilepsy have mortality rates that are two to three times higher than the general population (Cockerell et al. 1994; Hauser et al. 1980; Lhatoo et al. 2001; Nilsson et al. 1997), and the cause of the excess mortality is likely related in part to the underlying etiology of epilepsy and in part to the seizures themselves. Sudden unexplained death in epilepsy (SUDEP) is a leading diagnosis, especially in younger individuals, with the risk of sudden death perhaps as much as 24 times higher than that of the general population (Ficker et al. 1998; Walczak et al. 2001). While the exact pathophysiologic mechanisms of SUDEP remain elusive, a number of risk factors have been identiἀed as being associated with an elevated risk of SUDEP (Tomson et al. 2008). These include developmental delay or mental retardation, subtherapeutic anticonvulsant levels, treatment with more than two antiepileptic medications, and uncontrolled seizures, particularly generalized tonic–clonic seizures (Hughes 2009; Strauss et al. 2003; Walczak et al. 2001). The potential pathophysiologic mechanisms of SUDEP, as well as the risk factors, are reviewed in detail elsewhere in this volume (Lathers et al. 2010;€ Walczak 2010, this book). Ultimately, for death to occur, there must be an acute disruption of autonomic function, with cardiac arrhythmias and respiratory arrest or€ hypoventilation likely to be the ἀnal common direct pathway for SUDEP. Given the 905

906 Sudden Death in Epilepsy: Forensic and Clinical Issues

strong direct association between seizures and SUDEP, one might theorize that preventing seizures might reduce the risk of SUDEP and other seizure-related causes of mortality. Consequently, some have studied how improving seizure control might affect mortality. There is some evidence that controlling seizures with medications can lower the risk of SUDEP (Nilsson et al. 1999; Strauss et al. 2003). Epilepsy surgery provides another means of controlling seizures in some patients. At least 35% of people with epilepsy do not fully respond to medical therapy (Kwan and Brodie 2000), and surgery often alleviates seizures when medications fail to control seizures (Engel 1996; Engel et al. 2003; Sperling et al. 1996; Téllez-Zenteno et al. 2005; Wiebe et al. 2001; Wieser et al. 2003). No prospectively randomized studies have been performed to examine the effect of surgery on mortality. Indeed, only one prospective randomized study of epilepsy surgery has ever been performed (Weibe et al. 2001). This was conducted in Canada, where the typical wait for surgery was approximately one year, and patients could be randomized to either remain in the surgery queue while continuing medical therapy for another year, or advance to the head of the line and immediately have surgery. This study found that the patients treated surgically had a signiἀcantly higher rate of seizure freedom compared to patients who were awaiting surgery additionally, patients treated surgically had a signiἀcantly higher quality of life. The report contains one other interesting observation. In the group randomized to late surgery, that is, continued medical therapy, one death occurred. In the patients randomized to immediate surgery, no one died. Only 80 patients were included in this trial; thus, a statistical analysis cannot be done. Nonetheless, it is intriguing that the single death occurred in the medically treated patients, while none were observed in surgical patients. Only a few nonrandomized studies have examined the effect of the surgical treatment on long-term mortality in epilepsy, and these suggest that seizure control is associated with reduction in mortality. This chapter will review this literature regarding the impact of epilepsy surgery on mortality.

56.2╅ Methods 56.2.1╅ Selected Studies A literature review of studies published in English that examined the long-term mortality of surgically treated patients with epilepsy yielded six key studies (Hennessy et al. 1999;€Nilsson et al. 2003; Salanova et al. 2002; Sperling et al. 1999, 2005; Stavem and Guldvog 2005). All of the studies rely on retrospective analyses of nonrandomized subjects. With the exception of one study that reported no impact of surgery on mortality, all other studies reported a positive effect of epilepsy surgery on the mortality rate and the incidence of SUDEP. The sole negative study compared patients treated surgically and medically in Norway from€1948 through 1988, many of whom had surgery before the advent of modern evaluation techniques (Stavem and Guldvog 2005). Most of their surgically treated patients continued to experience frequent seizures after surgery, and surgical morbidity was quite high. Consequently, the absence of a difference in death rates between medically and surgically treated patients likely reflects these underlying facts. The other reports include three studies performed at academic institutions in the United States, one multicenter report from Sweden utilizing the nationwide epilepsy registry, and one study from a tertiary hospital in London. One

Epilepsy Surgery and the Prevention of SUDEP

907

of the American studies (Sperling et al. 2005) updates an earlier report by the same group (Sperling et al. 1999), and the more recent study will be discussed in this chapter. 56.2.2â•… Study Design The nonrandomized studies tell a largely consistent story. The study by Nilsson and colleagues (2003) compared mortality rates in epilepsy patients with that of equally intractable patients who did not have surgery, either because they were deemed not to be surgical candidates, or were unwilling to have surgery, or died before surgery was offered. The other studies followed surgically treated patients and examined mortality in relation to postoperative seizure control. Both study designs are limited by necessity. The ἀrst design does not offer ideal control subjects for comparison with the surgical patients, since the composition of the medical group differs somewhat from the surgically treated group. The second design does not prospectively account for the possibility that patients who ultimately fail to respond to surgery differ in some fundamental way from those who respond. This shortcoming might be partly overcome by comparing mortality in patients who become seizure free with expected mortality in the general population (Sperling et al. 1999); however, this approach still lacks the robustness of a prospective, randomized controlled trial. However, the extant studies probably represent the best that can be done at present since it is not ethical to randomize some patients with intractable epilepsy to an unnecessary delay in operation. 56.2.3â•… Calculating Mortality In calculating the rate of death, two different measures were used. The standardized mortality ratio compares the mortality rate in a subject population with that of the general population. It is calculated by dividing the number of observed deaths in a particular cohort by the number of expected deaths in an age, gender, and perhaps racially matched general population. The standardized mortality ratio is helpful in comparing the all-cause mortality of patients with epilepsy or subgroups thereof with that of the general population. The second method relates the total number of deaths of a speciἀc cohort during a deἀned time period (usually per 1000 person-years of follow-up), thus allowing the mortality incidence to be tabulated. An expected number of deaths per thousand person-years can be determined for an age, gender, and ethnically matched general population. The death rate per thousand person-years of the speciἀc cohort can then be compared with the death rate in the general population, and an annualized risk of death estimated. 56.2.4â•…Duration of Follow-Up The studies varied in duration of follow-up and in frequency of assessment during that time. Most patients were followed on an annual basis postoperatively, and studies had a mean follow-up of 5–9 years. The Nilsson et al. (2003) study was limited by following each patient for 2 years after surgery and then relied on medical records of only selected patients for further detail. This design has the risk of underreporting the number of patients with recurrent postoperative seizures. Indeed, in that study, some of the patients who had been reported as being seizure free at the ἀxed 2-year interval were discovered to have had seizure relapses when the individual medical records were subsequently reviewed. The issue

908 Sudden Death in Epilepsy: Forensic and Clinical Issues

of follow-up duration is a complex one. New seizure recurrences happen with each passing year after surgery. Recurrence rates are low after the ἀrst few postoperative years, only 2–4% per year (Sperling et al. 2008). Since the bulk of postoperative seizures occur in the early years after surgery, one can still reasonably estimate mortality risk if seizure recurrence in the ἀrst few years after surgery is known. Moreover, use of survival analysis helps account for disparities in duration of follow-up. This issue is even more complicated, since the added risk of dying might mainly exist around the time that seizures occur. Analyses might subdivide time after surgery for each patient into seizure-free and non–seizure-free epochs to better estimate risk of mortality imposed by seizures. This is methodologically fraught, however, since methods for deἀning such epochs would be arbitrary. Moreover, all studies are limited by accuracy of subject reporting. Patients may either intentionally or unintentionally underreport seizures to their physicians. They may hide seizure recurrence to gain driving privileges, may be loathe to disappoint their physicians, or may be unaware of their seizures. Patients may also forget whether seizures have occurred if follow-up contact is infrequent. In these circumstances, the bias is to underreport seizures, and patients might inappropriately be registered as seizure free. On the other hand, other patients may report continued seizures after surgery when they may not be experiencing epileptic seizures but rather may have non-epileptic psychogenic seizures or other non-epileptic phenomena instead. While it is the authors’ experience that overreporting is probably less common than under-reporting seizures, no studies have been performed to evaluate the frequency of these confounders.

56.3â•…Results 56.3.1â•… Effect of Postoperative Seizure Control on Mortality Table 56.1 summarizes the pertinent ἀndings from the aforementioned studies. Even after controlling for any mortality directly related to the surgery itself, each study found that mortality rates were lower in the surgically treated patients. This effect was most striking in patients who stopped having seizures after surgery (Figure 56.1). While Nilsson et al. (2003) demonstrated no signiἀcant difference in the rate of SUDEP between surgical patients who remained seizure free versus surgical patients who had recurrent seizures, they demonstrated a signiἀcant difference in the standardized mortality ratio between surgical and medically treated patients. Surgical patients had a standardized mortality ratio of 4.9 with a SUDEP incidence of 2.4 per 1000 person-years, whereas medically treated patients had a standardized mortality ratio of 7.9 and SUDEP incidence of 6.3 per 1000 person-years. Salanova et al. (2002) demonstrated that surgical patients who became seizure free had a signiἀcantly lower mortality (standardized mortality ratio = 1.7, approaching that of the general population), while surgical patients with persistent seizures had a much higher mortality (standardized mortality ratio = 7.4). Finally, Sperling et al. (2005) demonstrated that the excess mortality associated with refractory epilepsy was eliminated when epilepsy surgery successfully abolished seizures. Patients with recurrent postoperative seizures had a mortality rate of 11.4 per 1000 person-years, and those without recurrent seizures had a rate of 0.85 per 1000 personyears (p = 0.001), which is the same as that of the general population. These studies all demonstrate a consistent trend. Patients who stop having seizures tend to have lower mortality than patients who continue to experience seizures, and mortality

Epilepsy Surgery and the Prevention of SUDEP

909

Table 56.1â•… Results of Selected Studies Study Years studied Surgery type

Number of surgical patients

Number of nonsurgical patients Surgical outcome, Engel class Percent of mortality due to SUDEP or possible SUDEP SUDEP incidence, surgical cohort

SMR, all-cause death, surgical cohort SUDEP incidence, nonsurgical cohort SMR, nonsurgical cohort

Sperling et al. 2005

Nilsson et al. 2003

1986–2000 89.4% Resection and/or subpial transection; 10.6% CC 583 (Includes 393 from the analysis published in 1999) N/A

1990–1998 84.2% Resection; 11.5% CC

1984–1999 100% Resection

Hennessy et al. 1999 1975–1995 100% Resection

215

299

212

N/A

N/A

44% Class I (50% in patients with focal surgery) 53%

49% Class I

69% Class I

40% Class I A/B; 24% class I C/D or class II

43%

55%

20%

6.3/1000 (based on 10 cases of€SUDEP per€1580 person-years)a

2.4/1000 person-years

2.2/1000 person-years (reported as 1/455 person-years)

Whole surgical cohort = 3.56 Persistent sz = 5.75 Sz-free = 0.45 N/A

Whole surgical cohort = 4.9 Persistent sz = 4.3 Sz-free = 3.8 6.3/1000 Person-years 7.9

4.0/1000 personyears (based on 6 unexplained or seizure-related deaths per 1514 person-years) Whole surgical cohort = 5.27 Persistent sz = 7.9

N/A

651 Operations in 596 patients

Salanova et al. 2002

Sz-free€ = 1.7 N/A N/A

Whole surgical cohort€= 4.5 Persistent sz = not reported Sz-free = not reported N/A N/A

Note: CC, corpus callosotomy; N/A, not applicable; sz, seizure; SMR, standardized mortality ratio. Class I: seizure free (no seizures with loss of awareness); class I A, B, C, and D are subtypes of seizure freedom (A, no seizures after surgery; B, simple partial seizures only; C, some postoperative seizures but none for 2 years; D, generalized seizures after drug withdrawal only); class II: rare seizures. a All cases of SUDEP occurred in patients with persistent seizures after surgery.

rates in seizure-free patients approach that of the general population. This latter ἀnding, although the studies are neither controlled nor randomized, most strongly suggests that epilepsy surgery reduces the excess mortality imposed by epilepsy. Untreated, these patients should have had a much higher death rate, and it seems reasonable to hypothesize that deaths were prevented by successful surgical treatment. Equally striking is the fact that patients with persistent seizures after surgery die at a rate similar to that observed in medically refractory patients treated at epilepsy centers (Nashef et al. 1995; Tomson 2000).

910 Sudden Death in Epilepsy: Forensic and Clinical Issues 1.00

Cumulative proportion surviving

0.95

+

0.90

(

+

(

(

((

()

(

+ + + ++++

0.85

+

0.80 0.75 0.70 0.65 0.60 0.55 0.50

0

2

4

6

8

10

12

14

16

Years

Figure 56.1╇ Survival as a function of postoperative seizure recurrence after surgery for the entire cohort. Top line shows survival in patients who did not recur (n = 258, 1 death); lower line shows survival in patients who had recurrent seizures (n = 325, 18 deaths). Survival was significantly better in patients who did not experience recurrent seizures after surgery. (From Sperling, M. R., et al., Epilepsia 46, 49–53, 2005. With permission.)

This was observed even when there was a substantial reduction in seizure frequency postoperatively; some of the patients who died had only a few seizures per year (Hennessy et al. 1999; Sperling et al. 1999, 2005). Hence, any reduction in mortality following surgery appears to require elimination of seizures. Whether the patient with rare,€sporadic seizures after surgery has excess mortality only around the time of seizure recurrence is unknown. This poses methodological problems when assessing the effect of surgery on mortality. Since most patients have at least one seizure after epilepsy surgery if they€live for many years (McIntosh et al. 2004; Sperling et al. 2008), a partial rather than complete reduction in the excess mortality from epilepsy might be anticipated. 56.3.2â•… Effect of Laterality of Surgery on Mortality While none of the studies found that the side of resection influenced mortality, two studÂ� ies€noted an increase in mortality and SUDEP in a speciἀc subgroup of patients that had right temporal lobe resections with a pathological determination of mesial temporal sclerosis or gliosis. Hennessy and colleagues (1999) reported that patients with right-sided mesial temporal sclerosis had a postoperative mortality of 1 per 53 person-years (standardized mortality ratio = 32.0), whereas those with left-sided mesial temporal sclerosis had a signiἀcantly lower mortality of 1 per 310 person-years (standardized mortality ratio = 3.3). Nilsson et al. (2003) also observed that patients with pathologically conἀrmed gliosis who underwent right temporal lobe resection had a 2.5-fold increase in the standardized mortality ratio compared to the left, and a SUDEP incidence of 8.3 per 1000 person-years from the€right€temporal€subgroup, compared to no SUDEP occurrences from the left temÂ� poral subgroup€with€pathologically conἀrmed gliosis. This ἀnding was not reproduced in the€studies€by€Salanova and colleagues (2002) or Sperling and colleagues (2005). Surgery

Epilepsy Surgery and the Prevention of SUDEP

911

rendered€fewer€right-sided surgical patients in the Hennessy et al. (1999) study seizure free; thus, the increased mortality may have been related to seizure control. Moreover, the number of deaths in both series was small. At present, there appears to be insufficient data to conclude that right-sided epilepsy poses greater risk of death and further study is needed. 56.3.3â•… Effect of Gender of Surgical Patients on Mortality In the study by Sperling et al. (2005), gender did not influence death rates when assessed in person-years. However, the standardized mortality ratio has been reported to be higher in women than men (Nilsson et al. 2003; Sperling et al. 2005). This elevation in the standardized mortality ratio is likely an artifact of the method by which the value is calculated. In the general population, women have lower mortality rates than men; thus, the denominator used in calculating the standardized mortality ratio is considerably smaller for women than men. Consequently, while the absolute mortality rates are similar in men and women, the higher standardized mortality ratio in women is an artifact of the methodology. 56.3.4â•… Effect of Pathologic Diagnosis on Mortality Patients with progressive lesions or malignant tumors as the etiology of the seizures were generally excluded from the reviewed studies. Only one study (Hennessy et al. 1999) found that almost all of the epilepsy-related deaths occurred in surgical patients with either hippocampal sclerosis or nonspeciἀc ἀndings. This ἀnding was not conἀrmed in the Salanova et al. (2002) study, and the other studies did not assess the relationship between pathology and mortality. Further study is needed, therefore, to determine how underlying pathology affects mortality in surgically treated patients with epilepsy. 56.3.5â•… Effect of Cerebral Lobe on Mortality Only two of the studies included epileptic lesions beyond the temporal lobe (Nilsson et al. 2003; Sperling et al. 2005). Neither found a difference in mortality rates between patients with temporal lobe resections and those with extra-temporal resections. However, studies with more patients are needed to conἀrm this ἀnding since the relatively small number of deaths could mask a type 2 error and not discover an effect of lobe on mortality. A multicenter French trial has been proposed to compare the mortality of patients with a purely temporal epileptogenic zone to that of patients with a “temporal plus” epileptogenic zone (Ryvlin et al. 2005). 56.3.6â•… Comparison of Resection/Subpial Transection with Callosal Section One study (Sperling et al. 2005) reported that patients who underwent an anterior or complete corpus callosotomy had a higher late mortality rate than patients who had surgery for a focal epilepsy (p = 0.001). Most of these patients had symptomatic generalized epilepsy and, presumably, had signiἀcant underlying neurological impairment. This, itself, is a risk€factor for excess mortality (Forsgren et al. 1996). Comparing survival among patients with recurrent seizures after either callosal section or focal surgery, a nonsigniἀcant trend was observed for increased mortality in the patients after callosal section (22.6 per 1000 person-years, p = 0.06).

912 Sudden Death in Epilepsy: Forensic and Clinical Issues

56.4â•… Conclusion In the modern era, epilepsy surgery has a low morbidity and eliminates seizures in many patients with medically refractory seizures. The studies reviewed above, while with limitations, suggest that successful epilepsy surgery reduces the excess mortality in epilepsy, and the mortality rates in seizure-free patients can approach that of the general population. Data from medically treated patients conἀrm that even rare seizures carry a signiἀcant risk of mortality. Half of the patients who die suddenly have infrequent seizures, and nearly one-third have fewer than three seizures per year (Hirsch and Martin 1971; Leestma et al. 1989). These ἀndings are replicated in the surgical studies. Since it is unlikely that large-scale randomized studies will ever be performed to address the effects of epilepsy surgery on mortality, future investigations will need to study speciἀc questions that can be answered in surgical populations. Does limbic epilepsy convey a greater mortality risk than extra-limbic epilepsy? Since surgical patients are at high risk for SUDEP, studies might be conducted to attempt to elucidate the pathology of SUDEP. Perhaps, detailed autonomic testing might be performed before surgery so that more can be learned should SUDEP occur. Changes in autonomic function might be assessed before and after surgery as well, to explore effects of surgery on various autonomic parameters. Electrocardiographic analyses thus far of patients who subsequently died of SUDEP have been provocative but limited in scope (Nei et al. 2004; Nei 2009). For the present, it seems appropriate to counsel patients with uncontrolled epilepsy that surgery may lessen their chance of dying and that the risks of recurrent seizures may exceed the risks of surgery.

References Cockerell, O. C., A. L. Johnson, J. W. Sander, Y. M. Hart, D. M. Goodridge, and S. D. Shorvon. 1994. Mortality from epilepsy: Results from a prospective population-based study. Lancet 344 (8927): 918–921. Engel, Jr., J. 1996. Surgery for seizures. N Engl J Med 334 (10): 647–652. Engel, Jr., J., S. Wiebe, and J. French et al. 2003. Quality Standards Subcommittee of the American Academy of Neurology; American Epilepsy Society; American Association of Neurological Surgeons. Practice parameter: temporal lobe and localized neocortical resections for epilepsy: Report of the Quality Standards Subcommittee of the American Academy of Neurology, in association with the American Epilepsy Society and the American Association of Neurological Surgeons. Neurology 60: 538–547. Ficker, D. M., E. L. So, W. K. Shen et al. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51 (5): 1270–1274. Forsgren, L., S. O. Edvinson, L. Nystrom, and H. K. Blomquist. 1996. Influence of epilepsy on mortality in mental retardation: An epidemiological study. Epilepsia 37: 956–963. Hauser, W. A., J. F. Annegers, and L. R. Elveback. 1980. Mortality in patients with epilepsy. Epilepsia 21 (4): 399–412. Hennessy, M. J., Y. Langan, R. D. Elwes, C. D. Binnie, C. E. Polkey, and L. Nashef. 1999. A study of mortality after temporal lobe epilepsy surgery. Neurology 53 (6): 1276–1283. Hirsch, C. S., and D. L. Martin. 1971. Unexpected death in young epileptics. Neurology 21 (7): 682–690. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14 (2): 280–287. Kwan, P., and M. Brodie. 2000. Early identiἀcation of refractory epilepsy. N Engl J Med 342 (5): 314–319.

Epilepsy Surgery and the Prevention of SUDEP

913

Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26: 195–203. Lhatoo, S. D., A. L. Johnson, D. M. Goodridge, B. K. MacDonald, J. W. Sander, and S. D. Shorvon. 2001. Mortality in epilepsy in the ἀrst 11 to 14 years after diagnosis: Multivariate analysis of a long-term, prospective, population-based cohort. Ann Neurol 49: 336–344. McIntosh, A. M., R. M. Kalnins, L. A. Mitchell, G. C. Fabinyi, R. S. Briellmann, and S. F. Berkovic. 2004. Temporal lobectomy: Long-term seizure outcome, late recurrence and risks for seizure recurrence. Brain 27 (Pt 9): 2018–2030. Nashef, L., D. R. Fish, J. W. A. S. Sander, and S. D. Shorvon. 1995 Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral centre. J Neurol Neurosurg Psychiatry 58: 462–464. Nei, M. 2009. Cardiac effects of seizures. Epilepsy Curr 9 (4): 91–95. Nei, M., R. T. Ho, B. W. Abou-Khalil et al. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45: 338–345. Nilsson, L., A. Ahlbom, B. Y. Farahmand, and T. Tomson. 2003. Mortality in a population-based cohort of epilepsy surgery patients. Epilepsia 44 (4): 575–581. Nilsson, L., B. Y. Farahmand, P. G. Persson, I. Thiblin, and T. Tomson. 1999. Risk factors for sudden unexpected death in epilepsy: A case–control study. Lancet 353 (9156): 888–893. Nilsson, L., T. Tomson, B. Y. Farahmand, V. Diwan, and P. G. Persson. 1997. Cause-speciἀc mortality in epilepsy: A cohort study of more than 9000 patients once hospitalized for epilepsy. Epilepsia 38 (10): 1062–1068. Ryvlin, P., A. Montavont, and P. Kahane. 2005. The impact of epilepsy surgery on mortality. Epileptic Disord 7 (S1): S39–S46. Salanova, V., O. Markand, and R. Worth. 2002. Temporal lobe epilepsy surgery: Outcome, complications, and late mortality rate in 215 patients. Epilepsia 43 (2): 170–174. Sperling, M. R., M. Nei, A. Zangaladze et al. 2008. Prognosis after late relapse following epilepsy surgery. Epilepsy Res 78 (1): 77–81. Sperling, M. R., A. Harris, M. Nei, J. D. Liporace, and M. J. O’Connor. 2005. Mortality after epilepsy surgery. Epilepsia 46 (Suppl 11): 49–53. Sperling, M. R., H. Feldman, J. Kinman, J. D. Liporace, and M. J. O’Connor. 1999. Seizure control and mortality in epilepsy. Ann Neurol 46 (1): 45–50. Sperling, M. R., M. J. O’Connor, A. J. Saykin, and C. Plummer. 1996. Temporal lobectomy for refractory epilepsy. JAMA 276 (6): 470–475. Stavem, K., and B. Guldvog. 2005. Long-term survival after epilepsy surgery compared with matched epilepsy controls and the general population. Epilepsy Res 63 (1): 67–75. Strauss, D. J., S. M. Day, R. M. Shavelle, and Y. W. Wu. 2003. Remote symptomatic epilepsy: Does seizure severity increase mortality? Neurology 60: 395–399. Téllez-Zenteno, J. F., R. Dhar, and S. Wiebe. 2005. Long-term seizure outcomes following epilepsy surgery: A systematic review and meta-analysis. Brain 128 (5): 1188–1198. Tomson, T. 2000. Mortality in epilepsy. J Neurol 247 (1): 15–21. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7: 1021–1031. Walczak, T. S., I. E. Leppik, M. D’Amelio et al. 2001. Incidence and risk factors in sudden unexpected death in epilepsy. A prospective cohort study. Neurology 56: 519–525. Wiebe, S., W. T. Blume, J. P. Girvin, and M. Eliasziw. 2001. Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporallobe epilepsy. N Engl J Med 345 (5): 311–318. Wieser, H. G., M. Ortega, A. Friedman, and Y. Yonekawa. 2003. Long-term seizure outcomes following amygdalohippocampectomy. J Neurosurg 98 (4): 751–763.

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP

57

Jane Hanna Rosemary Panelli

Contents 57.1 Introduction 57.2 SUDEP—Out of the Shadows 57.3 United Kingdom National Sentinel Clinical Audit of Epilepsy-Related Death (2002) 57.4 Fatal Accident Inquiry 57.5 Clinical Guidelines (2003, 2004) 57.6 Scottish Public Services Ombudsman—Case 200700075 (2009) 57.7 SUDEP—Assessment and Communication of Risk 57.7.1 The Evidence 57.7.2 Current Practice 57.8 Ethical, Legal, and Practical Arguments in Favor of Disclosure 57.8.1 Autonomy, Truthfulness, and Trust 57.8.2 Potential for Reduction in Fear 57.8.3 Adherence, Self-Management, and Shared Decisions 57.8.4 Disclosing a Material Risk—The Legal Right to Know 57.8.5 Reducing Trauma of Those Bereaved through SUDEP 57.9 Ethical, Legal, and Practical Arguments against Disclosure 57.9.1 Harms of Communication 57.9.2 Not All Patients Want to Share in Discussion and Decision Making 57.9.3 Denying the Patient’s Right Not to Know 57.10 SUDEP Information and Education—When and How? 57.10.1 Tell if Asked 57.10.2 High-Risk Patients Only 57.10.3 If a Patient Does Not Adhere to Treatment 57.10.4 Routine Discussion 57.10.5 Risk Communication 57.11 SUDEP and Bereavement 57.12 SUDEP Support Services 57.13 Summary References

915

916 916 917 917 918 918 919 920 920 921 921 922 922 923 924 924 924 925 926 926 926 926 927 928 929 929 930 931 932

916 Sudden Death in Epilepsy: Forensic and Clinical Issues

57.1â•…Introduction Sudden unexpected death in epilepsy (SUDEP) is slowly emerging from the shadows of neglect and medical uncertainty. As awareness of SUDEP spreads and research increases, risk factors are slowly being identiἀed although the causes remain elusive. Responses to SUDEP have included guidelines emphasizing quality medical care in epilepsy, not only to tackle SUDEP but also to enhance the well-being of all epilepsy patients. Few deny the importance of such a goal; however, it is apparent that when it comes to assessment and communication of SUDEP risks, there are divergent views on what deἀnes best practice. Ethical, legal, and practical arguments associated with this issue are all points of debate. Nevertheless, patients are increasingly included in the conversation about SUDEP and much remains to be learned about optimal timing and the method of this interaction. SUDEP support services have played a central role in the promotion of SUDEP awareness and action on behalf of people with epilepsy, and their friends and families. Activities include personal support and counseling for those directly affected by SUDEP, production of resource materials, training of health professionals, and facilitation of research, either directly or through fundraising. SUDEP advocacy groups have also provided speciἀc information and support to bereaved families around the time of the post mortem, thereby assisting the family while also facilitating the work of the forensic agency. This chapter discusses some of the developments in the SUDEP story over the past 15 years. The debate concerning legal and ethical aspects of the assessment and communication of SUDEP risk is explored, and options for the timing and method of disclosure are presented. The long-lasting effect of SUDEP on the bereaved is discussed, and the chapter concludes with a brief review of the role played by SUDEP support services.

57.2â•… SUDEP—Out of the Shadows SUDEP was well recognized in the early 20th century (Spratling 1904), along with identiἀed risks and recommendations for prevention; however, by the 1960s, a myth had gained hold in the medical literature that epilepsy was not fatal. “Patients with epilepsy had moved from asylums into the community and there was much less opportunity for observation. Risks from epilepsy were minimized then denied; that seizures could not be fatal became ‘common knowledge’ despite evidence to the contrary” (Nashef 1995). However, in the early 1990s, families bereaved by epilepsy began to ask questions of the scientiἀc community. In an era when modern medicine had mastered the techniques of organ transplants and in vitro fertilization, they began to ask why a young person in apparent good health, apart from epilepsy, would die unexpectedly with no explanation. In the United Kingdom the lack of medical knowledge and social support led bereaved families to join together for mutual support and to seek answers to the questions raised by SUDEP. In collaboration with the few clinicians who were interested in the issue, these motivated few pushed SUDEP onto the international scene as a serious topic for research and clinical practice. In 1996, the newly founded charity, Epilepsy Bereaved, convened the ἀrst ever international workshop on SUDEP. Those present identiἀed the need for research on all key aspects of SUDEP (Nashef et al. 1997). In addition, an information review concluded that SUDEP information was not easily accessible to patients at this time (Preston 1997). Today, the nature of the scientiἀc uncertainty on SUDEP has changed. While there are no

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 917

identiἀed causes, research has led to advances in understanding the risk factors associated with SUDEP. In addition, community action groups for SUDEP have multiplied.

57.3â•…United Kingdom National Sentinel Clinical Audit of Epilepsy-Related Death (2002) Between September 1999 and August 2000, a nationwide audit of epilepsy-related deaths was conducted in the United Kingdom. This National Sentinel Clinical Audit of EpilepsyRelated Death (Hanna et al. 2002) aimed to establish whether deἀciencies in the standard of clinical management, or in the overall health care package, could have contributed to the deaths. The audit found signiἀcant problems of access to epilepsy services and reviews as well as medication issues, concluding that there was signiἀcant potential for avoidance of death if these risks were addressed. Others concluded that if a similar audit were to be repeated in the United States, the results would probably be the same because of problems of access to services (Pedley and Hauser 2002). The audit motivated policy makers in the United Kingdom to develop government action plans on epilepsy (Department of Health 2003; Welsh Assembly Government 2009) and to fund the development of patient information on risk, involving patients who valued the inclusion of information on SUDEP.

57.4â•… Fatal Accident Inquiry In the same year as the audit of epilepsy-related deaths, a Fatal Accident Inquiry was held into the sudden unexpected death of a young woman, aged 17 years (Taylor 2002). This type of judicial inquiry is held in Scotland by a judicial office holder called a Sheriff where a death concerns the public interest. The purpose is to make recommendations to prevent future deaths. The deceased’s mother had died from epilepsy in 1988, and in 1991 the deceased presented with seizures. The family was reassured that the deceased suffered from benign focal seizures of childhood and she was discharged by the specialist to general practice on antiepileptic drugs. From 1991, she had four to ἀve seizures yearly, varying in frequency and severity, but there was no annual review or re-referral for specialist care. The court found a catalogue of failures to look after the deceased in a proper manner. These included a failure on the part of specialists to alert the general practitioner as to what circumstances required re-referral; a failure on the part of the general practice to prescribe appropriate levels of medication; a failure to reâ•‚refer the deceased when the seizures did not stop after 2€years; a failure to re-refer the deceased when the intensity, form, and duration of her seizures altered as the deceased matured; and a failure by the medical team to discuss with the deceased’s family the diagnosis, the attendant risks, and how these risks might be properly managed. The sheriff stated that given the association between seizures and SUDEP and the potential for control, that it was a “short step” to the view that if the€deceased had been€referred for review she might not have died. He determined that the family ought to have been informed that the deceased was suffering from epilepsy, the risks of SUDEP explained, and a discussion held on how her condition might be managed. The most important recommendation was considered to be the need for a personal care plan. The sheriff suggested that all the key issues would have been addressed if a care plan,

918 Sudden Death in Epilepsy: Forensic and Clinical Issues

“. . . shared or otherwise” had been produced, and “. . . it might have saved her life” (Taylor 2002). The sheriff was clear that for the purposes of the public inquiry, it was not necessary for there to be any scientiἀc certainty. Any legal judgment is determined by balance of probabilities, and in this case the concern was whether preventative measures “might” have saved a life. The sheriff was clear that information on SUDEP could be relevant to how proactive a family might be in probing decisions about treatment as well as informing discussions on how risks might be reduced. The sheriff accepted that the question of informing about SUDEP must be left to the discretion of the medical profession to form a view. In particular, he accepted there might be people of “an extreme disposition” where discussion might cause harm. Nevertheless, he said, “I do, however, accept that in the vast majority of cases there should be such a discussion” (Taylor 2002).

57.5â•… Clinical Guidelines (2003, 2004) Clinical guidelines in Scotland and those produced by the National Institute of Health and Clinical Excellence (NICE) for England and Wales now recommend discussion of SUDEP as part of general epilepsy information (Scottish Intercollegiate Guidelines Network 2003; Stokes et al. 2004). The NICE guideline sets out information to show why preventing seizures is important. It is based on the view that SUDEP risk can be minimized by optimizing seizure control and by families being aware of the potential consequences of nocturnal seizures. It also recommends tailored information on the individual’s risk of SUDEP as part of any counseling checklist for people with epilepsy, and their families and/or care givers.

57.6â•… Scottish Public Services Ombudsman—Case 200700075 (2009) In 2009, 7 years after the Scottish Fatal Accident Inquiry, ongoing concerns with epilepsy care in the United Kingdom were highlighted by another public investigation, carried out by the Scottish Public Services Ombudsman. The Ombudsman is an individual appointed by the Queen on the nomination of the Scottish Parliament (www.spso.org.uk/about-us). The investigation into the death of an 18-year-old woman upheld the complaint of the deceased’s family that there was a failure in patient-centered care following a withholding of information on SUDEP (Scottish Public Services Ombudsman 2009). Although the recommendations of the ombudsman are not legally binding, they are normally followed as their purpose is not only to provide justice for the individual but also to improve the delivery of public services. The deceased was due to leave home for the university when she was diagnosed with tonic–clonic seizures and prescribed antiepileptic drugs following a nocturnal seizure. On a second review some 5 months later, no further seizures were reported, and the medication was reduced because of concerns about side effects. The family had reason to be concerned that she was not adhering to treatment following her move to the university. The deceased died suddenly in her sleep shortly after, and toxicology testing at postmortem showed no evidence of antiepileptic drugs. A complaint was brought by the family that the deceased was not provided with sufἀcient information on her treatment and the risk of SUDEP to inform her decisions in self-

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 919

management of her condition. The complainant argued that information about SUDEP should have been given to the deceased, unless the clinician had good reason not to. The clinician argued that to give information could cause undue distress and there was nothing that the deceased could do to prevent her death. Therefore SUDEP information should only be given when there was good reason to do this. The ombudsman accepted that there was signiἀcant divergence in view between clinicians on this question and described these as the “proactive” approach supporting the complainant’s view and the “reactive” approach supporting the clinician’s view. Because of this divergence the ombudsman was not able to form any conclusion on the issue of modiἀcation of speciἀc SUDEP risks or on the issue of patient harm from disclosure of risk, and has urged that both are the subject of research as a matter of priority. Despite this the ombudsman upheld the complaint on the grounds that there was no evidence in the record to suggest the clinician had made an individualized decision to withhold information on SUDEP, and that this amounted to a failure to provide individualized patient care by the clinician and the board of the hospital. The ombudsman noted that the intention of the Fatal Accident Inquiry (Taylor 2002), the ethos of the health service, general professional guidance from the regulatory body, and the plain reading of clinical guidelines on the epilepsies pointed in the direction of information giving as the norm. She recognized that while deviation may be appropriate in some cases, it should be recorded. The ombudsman found that the signiἀcance of the particular guideline was not determined simply by the level of evidence supporting it, and in this case the decision to withhold the information should be recorded. Further ἀndings were a failure to provide written information regarding SUDEP and other aspects of the medical condition. The ombudsman noted the lack of research into the effect of SUDEP awareness on bereaved families and made mention of reports from support groups indicating that SUDEP information helps families to better deal with the tragedy. She commented that in the case under investigation, the family could not help but wonder: . . . how they might have altered the outcome if they had only known this information. Mrs C accepted that Miss C may still have died even if they had known about SUDEP, but because there is the suggestion of the chance she might not have, they are not able to ἀnd any peace. (Scottish Public Services Ombudsman 2009)

The Health Board acted on the complaint in respect of the provision of written epilepsy information to patients following diagnosis and the appointment of an epilepsy specialist nurse, but has not accepted the ombudsman’s ἀnding on the failure to evidence an individualized decision to withhold information on SUDEP. The ombudsman will be drawing the attention of the Scottish Government to the disagreement and has referred her report to the national guidance body in Scotland.

57.7â•… SUDEP—Assessment and Communication of Risk The report of the Scottish Public Services Ombudsman (2009) highlighted an ongoing lack of consensus on how SUDEP should be addressed in epilepsy management. Despite the increase of research data, the development of some clinical guidelines that include SUDEP (Scottish Intercollegiate Guidelines Network 2003; Stokes et al. 2004), and adjustments to practice that have occurred worldwide, inconsistencies in care persist. Factors influencing

920 Sudden Death in Epilepsy: Forensic and Clinical Issues

opinion include the perceived avoidability of SUDEP; the estimation of risk for individuals; and the ethical, legal, or practical issues that are related to patient disclosure. This section aims to summarize the evidence for prevention of SUDEP before addressing current practice and the main arguments in favor and against communication of risk. 57.7.1â•… The Evidence Although there is recognition that more research is needed on risk factors, there is now a strong body of evidence supporting the association between SUDEP and seizures, especially generalized tonic–clonic seizures. Despite some contradictory voices, a body of medical opinion has been building over a number of years that some SUDEP deaths are potentially avoidable (Faught et al. 2008; Hanna et al. 2002; Hitiris 2007; Hughes€2009;€Monte et al. 2007; Opeskin and Berkovic 2003; So 2006; So et al. 2009; Surges et al. 2009; Tomson et al. 2008). Recent reviews (Hughes 2009; Monte et al. 2007; Tomson et al. 2008) have concluded that there is evidence of varying degrees for risk factors that include seizures (especially generalized tonic–clonic seizures), young age, early onset of epilepsy, absence of treatment or nonadherence to antiepileptic drugs, polytherapy, the prone position, being in the bedroom, sleeping, and being male. The effect of nocturnal supervision was examined in only one case–control study but was found protective (Langan et al. 2005). It has been suggested that the precautionary approach of the sudden infant death syndrome (SIDS) risk reduction campaign is a suitable model for SUDEP (So et al. 2009). SIDS is a rare event with no one cause identiἀed, yet during the past two decades many countries around the world have launched successful public campaigns informing about its risk factors (Moon et al. 2007). A precautionary approach for SUDEP would support prevention strategies based on optimization of seizure control, effective antiepileptic drug treatment, supervision in appropriate cases, and information to patients and care givers. This precautionary approach is supported by a reasonable body of policy makers (Department of Health 2003; Stokes et al. 2004; Welsh Assembly Government 2009). Most recently a report of the American Epilepsy Society and the Epilepsy Foundation Joint Task Force on SUDEP supported the view that certain risk factors associated with SUDEP “may” be modiἀable— namely, uncontrolled seizures, subtherapeutic drug levels, and the number of antiepileptic drugs used. The report supported optimization of seizure control as a treatment goal. It also noted the need to raise awareness of SUDEP in both medical and lay communities and to establish frameworks for the guidance of future research (So et al. 2009). 57.7.2â•… Current Practice A survey of clinical practice asked neurologists in the United Kingdom if they told patients about SUDEP (Morton et al. 2006). The neurologists who replied reported a wide variation in practice. While 30.3% of respondents reported that they discussed SUDEP with the majority or all of their patients, 68.7% discussed with very few or none of their patients. Interestingly clinicians with a special interest in epilepsy were signiἀcantly more likely to discuss SUDEP. The paper does not quantify at any point what was meant by the terms “few” or “many” used in the questionnaire, and statistics given cannot be considered anything other than a general indicator (Scottish Public Services Ombudsman 2009). Those who support routine discussion of epilepsy-related death suggest that informing of the small risk of fatality associated with seizures sits appropriately with an early discussion of

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 921

risk (Cook 2005; Tomson et al. 2008). Such a discussion would not necessarily require the inclusion of the term SUDEP. Specialist nurses have been shown to be more likely than clinicians to discuss SUDEP. A recent survey in the United Kingdom found that 56% of specialist nurses discussed SUDEP with the majority of their patients. Most combined SUDEP with discussion of general and speciἀc risks of epilepsy (Lewis et al. 2008). No specialist nurse supported nondisclosure.

57.8â•… Ethical, Legal, and Practical Arguments in Favor of Disclosure The main arguments for disclosure are that it supports patient autonomy, informs choices on adherence and self-management, reduces fear by managing natural anxieties, and avoids the harm of false assurance. Research repeatedly tells us that people with epilepsy feel they receive insufficient information about their condition (Couldridge et al. 2001). A recent qualitative study found that patients felt a pressing need for information, including “. . . more concrete answers on SUDEP,” to discuss treatment options and medication in detail (Prinjha et al. 2005). This suggests that clinicians need to check and regularly revisit with patients how much information they would like about their condition. The perspective of care givers is also a consideration. Research shows that patient decision making involves an extended social context (Fraenkel and McGraw 2007); however, in any event, care givers may need accurate information in their own right. Families affected by SUDEP described a variety of roles associated with care such as keeping records of seizures and responding to seizures. Many considered that they needed more information on SUDEP as it pertains to these roles (Kennelly and Riesel 2002). One study found parents expected this (Gayatri et al. 2010). 57.8.1â•… Autonomy, Truthfulness, and Trust The ethical principle of patient autonomy involves the patient’s right to know about his or her own medical condition and prognosis. The American Epilepsy Society and the Epilepsy Foundation Joint Task Force on SUDEP considered discussion of SUDEP as consistent with the ethical principle of patient autonomy and also consistent with the need to accept that some persons with epilepsy have increased risks of morbidity and death (So et al. 2009). Examples of professional guidance based on the autonomy principle fully support the patient’s right to information about their condition. The amount of information given on a condition should depend on patient wishes, and need, for information. This should be determined by discussion with a patient and not be based on assumptions of what patients require (General Medical Council 2008). Provision of information is also important to the relationship of trust between doctor and patient. Where information on a risk is withheld and that risk materializes, it is natural that those affected by the consequences will seek to understand why information was not shared. They may experience the harm of false assurance. There is some limited evidence on the negative impact of false assurance in the context of medical screening, including public conἀdence and legal action (Petticrew et al. 2001); however, the impact of most seizures being presented as benign (apart from accidents and status epilepticus) has not been researched. Relatives bereaved through SUDEP frequently report, however, that their grief is exacerbated because epilepsy was presented in this way and it proved to be false

922 Sudden Death in Epilepsy: Forensic and Clinical Issues

(Kennelly and Riesel 2002). Further, the relationship of trust between the bereaved and the medical professional is more likely to be maintained where information withholding is patient-centered and withstands scrutiny. 57.8.2â•… Potential for Reduction in Fear A common assumption is that discussing SUDEP will create anxieties. In fact the only research on epilepsy and death anxiety comprises studies looking at the death anxiety associated generally with epilepsy. A study of 373 epilepsy patients found that approxÂ� imately two-thirds harbored fears of death from their next seizure (Mittan 1986). In a more€recent cross-sectional study of 92 patients having epilepsy for a minimum of 5 years, 56.52% of patients had either moderate or high death anxiety (Otoom et al. 2007). Not surprisingly, death anxiety was likely to be higher in patients with generalized epilepsy. Otoom et al. (2007) emphasize the importance of counseling patients to reduce anxiety. Anecdotal reports from clinicians who regularly talk about SUDEP mention the potential for reduction of fear by putting fears into perspective. Indeed, there is some evidence that this may be so. A case–control study on the beneἀts of a weekend educational program (SEE, known then as the Sepulveda Epilepsy Education program but more recently as Seizures and Epilepsy Education), which included discussion of risk of mortality, found a subsequent decrease in anxiety (Helgeson et al. 1990). The experience of this program was that lack of discussion of mortality led to adverse suppression of natural anxiety. Fear was also associated with overprotection and overcontrol of the person with epilepsy (Mittan 2005). The program was therefore concerned to ensure that patients and their care givers had a more realistic appreciation of the risk. 57.8.3â•… Adherence, Self-Management, and Shared Decisions Nonadherence to prescribed medication regimens in epilepsy is associated with a more than threefold increased risk of mortality (Faught et al. 2008). A recent review of all SUDEP studies since 1975 found 18 studies where subtherapeutic drug levels were signiἀcant and only 1 study that claimed that this factor was not signiἀcant (Hughes 2009). Research across chronic conditions has consistently found that lack of adherence is associated with patient doubts about their need for medication and concerns about side effects (Horne 2006). Patients with asthma, for example, were signiἀcantly more likely to endorse the need for regular medication if they shared the “medical view” of asthma as an “acute or chronic condition” with potentially serious consequences. If, on the other hand, patients saw their condition as chronic only, they were more likely to doubt the need for regular medication. The implications for asthma treatment were that it is not sufficient to advise a patient to take medication, but that a clear rationale is needed. This issue has not been researched in epilepsy. However, in a recent study, epilepsy specialist nurses perceived a positive association (Lewis et al. 2008). Other issues of self-management include reporting of seizures, avoiding triggers to seizures, lifestyle choices, and supervision. The risk of death following a seizure might be signiἀcant to some patients in balancing the risks and beneἀts of such behaviors. For example, a study of general practice found that 40% of patients who had anonymously reported seizures to a questionnaire, but not to their general practitioner, held a driving license. This suggests that patients perceive the beneἀts of concealment to outweigh what they understand as the risks (Dalrymple and Appleby 2000). Two deaths were reported

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 923

by bereaved families in people choosing not to visit a doctor because of potential loss of license (Kennelly and Riesel 2002). The severity of any potential harm is recognized as a key influence on health behavior and decision making (Weinstein 1999). If patients and care givers do not appreciate that avoidance of fatality is one motivation for treatment, it raises the question of whether this lack of information is signiἀcant to patient decisions about adherence and self-management. Chronic illness literature supports and promotes key themes of self-management and shared decision making (Department of Health 2001a; Lorig and Holman 2003; National Health Priority Action Council 2006; Whitney et al. 2008). Patient participation fulἀlls the ethical principle of patient autonomy, and when clinicians deny patients a choice that should be theirs, it causes harm (Katz 1984). In some countries supporting this approach, detailed professional guidance has been developed to mark the shift toward patients and doctors making decisions together and reflects exchange of information between patients and doctors as central to decision making (Nunes et al. 2009). This should include information on diagnosis, prognosis, and uncertainties, plus information on the potential beneἀts, risks, and burdens of treatment and of not treating (General Medical Council 2008). 57.8.4â•…Disclosing a Material Risk—The Legal Right to Know A legal duty to discuss a risk prima facie arises where there is a signiἀcant risk that would affect the judgment of the reasonable patient (Bolitho v City and Hackney Health Authority 1998; Pearce v United Bristol Healthcare NHS Trust 1999). Risks that are only remote possibilities can be regarded as material if the severity of the risk materializing is very serious. In the context of a SUDEP death, it could be argued that there is a material risk to be addressed where discussion might influence behavior, improve compliance, be relevant to issues of supervision, or where optimum treatment is not in place (Beran 2006). The courts in many countries will test the medical evidence offered by parties in litigation to reach their own conclusions on the magnitude and severity of risk, and the ease by which the risk might be avoided (Bolitho v City and Hackney Health Authority 1998; Rogers v Whitaker 1992; Videto v Kennedy 1981). The existence of guidelines, while not predictive of negligence, would be a relevant consideration in examining the rationality of withholding information. It has been argued that a claim would fail in a SUDEP case because of the need to prove causation between the actions of the physician and the death (Beran et al. 2004). However, the more recent authority of Chester v Afshar (2005) has extended the law of causation to include the scenario where risk is inherent in a condition and not caused by medical intervention. This would be a helpful legal authority in a SUDEP case. Although there has been no litigation into this question, a judicial determination was made in the Fatal Accident Inquiry (Taylor 2002). The inquiry concluded that patients should be told unless there is good reason not to. The sheriff did not specify precisely what such reasons would be, leaving this to the exercise of reasonable discretion by the medical practitioner. I stop short of recommending that the risk of sudden and unexplained death from epilepsy .€.€. should be discussed with every epilepsy sufferer or their family. . . . Discretion must be left with the medical profession to form a view. There may well be families who are of an extreme disposition or perhaps who have learning difficulties, where such a discussion would do more harm than good. I do, however, accept that in the vast majority of cases there should be such a discussion. (Taylor 2002)

924 Sudden Death in Epilepsy: Forensic and Clinical Issues

It is worth noting, however, that no actions for negligence have yet been reported in cases involving SUDEP deaths, and instead systems for investigation in the public interest have been activated by some families affected by SUDEP. The two Scottish cases detailed above both led to changes in service provision. Following the 2002 Fatal Accident Inquiry, an epilepsy clinic to review patients with epilepsy was established by the general practice and the 2009 Ombudsman’s Report makes reference to the creation of a new epilepsy specialist nurse post. 57.8.5â•…Reducing Trauma of Those Bereaved through SUDEP When SUDEP occurs and there has been no prior disclosure of risk, the justiἀcation for withholding information on grounds of causing anxiety or the lack of direct evidence of prevention is unlikely to withstand scrutiny by those who are suddenly bereaved. A relationship of trust with the particular medical team and with the broader medical profession may inevitably be placed under some strain at this time. A qualitative study involving 78 people affected by SUDEP reported that many found the lack of SUDEP information exacerbated their feelings of grief (Kennelly and Riesel 2002). As noted above, the aggravation of grief for families that lacked information before a SUDEP death was a point of discussion for the Scottish Public Services Ombudsman (2009) during a recent inquiry.

57.9â•… Ethical, Legal, and Practical Arguments against Disclosure The main arguments against disclosure are (1) that telling patients will cause anxiety for no beneἀt; (2) that there is no research on patient wishes in this area and that patients may not want to hear about SUDEP; and (3) that a clinician may be successfully sued for denying a patient “the right not to know” the risks associated with a condition. 57.9.1â•…Harms of Communication A therapeutic privilege exists in medical care to withhold information where it would cause harm to the patient. This concept is recognized by medical law as a defense to a legal action based on failure to disclose a material risk. Therapeutic privilege is normally conἀned to a psychiatric setting. What is clear from professional guidance and from legal precedents is that therapeutic privilege does not mean unfettered discretion to decide one way or another, but instead is conditional on a process of rational decision making. The General Medical Council of the United Kingdom (2008), for example, states that such information should not be withheld, and that it is necessary for making decisions unless the clinician believes the patient would be caused serious harm beyond being upset or refusing treatment. Any decision to withhold information should be recorded, justiἀed, and reviewed. The concern about the harm of raising patient anxiety is raised regularly in the medical literature on SUDEP as the reason for not providing information. Temporary anxiety in a patient is not normally viewed by the courts or ethical bodies as sufficient to constitute sensible medical grounds for withholding information (Deriche v Easling Hospital NHS Trust 2003). The only related research on this subject is a survey of neurologists in the United Kingdom where doctors were asked to report on how patients reacted to being told about SUDEP (Morton et al. 2006). About one-third of doctors did not respond to this

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 925

question, suggesting perhaps the difficulty of assessment. Some reported that telling such information to patients was difficult. One-third of respondents thought that information about SUDEP caused anxiety, but they were not asked how they assessed this, whether they followed up to check if anxieties were lasting, and whether the patient was offered or responded well to further information and support. Interestingly, doctors who discussed SUDEP with the majority of their patients were less likely to report negative reactions, suggesting, as the author recognized, a practice effect. The lack of evidence on patient harm from discussion of SUDEP was highlighted in the recent Scottish investigation. The Scottish Public Services Ombudsman (2009) declared that “much of the evidence in this area is . . . at best anecdotal and any reliance on an assumption about patients’ reactions must be tempered by the lack of actual hard evidence.” Outside of the epilepsy ἀeld, there is some evidence in the literature from which to draw, including a systematic review and a meta-analysis that provide some general indication of the relationship between bad news and anxiety. The evidence suggests that anxiety is a common and adaptive initial response to “at risk” notiἀcation, but it usually dissipates within a month (Marteau 2008; Shaw et al. 1999). Psychological theories of self-regulation describe the ways that humans maintain equilibrium while responding to threat (Taylor 1991). Clearly, fear unchecked can become negative and restrictive, but natural anxiety should not be assumed as harmful. 57.9.2â•…Not All Patients Want to Share in Discussion and Decision Making It is understood that patients vary in personality and coping styles, and therefore in their attitude to information and how they use it to navigate health issues (Andrewes et al. 1999; Politi et al. 2007). Some may not want to be involved in aspects of discussion or decision making with their health care providers. In the SUDEP debate, those who argue for the right not to know highlight differing patient-information-seeking behavior such as active searching (seekers), conscious blocking (avoiders), and a combination of styles (weavers) (Morton et al. 2006; Pinder 1990), expressing concern that where broad requirements to discuss SUDEP with all patients are mandatory, preference is therefore given to the needs of seekers above all others. But this need not be so. Where guidelines do exist, there are options for variation. Patients’ wishes are clearly highlighted in ethical literature on information giving, and clinical guidelines are always recognized as subject to clinical discretion where there is good reason not to follow a guideline. The General Medical Council (2008) advises doctors to provide patients with appropriate information, which should include an explanation of any risks to which they may attach signiἀcance. Doctors must not make assumptions about a patient’s understanding of risk or the importance they attach to different outcomes. They should discuss these issues with their patient. A clinician who probed patient wishes on information provision would have good reason not to inform, if the patients indicated that they did not want to have full information including risks of seizures; if the issue was regularly revisited; and if there was no material reason to override this, such as nonadherence to treatment. Pinder (1990) investigated information provision using Parkinson’s disease as the case study. She found that clinical practice on information giving was rarely determined by an accurate assessment of what the patient wanted, but was largely dictated by clinician assumptions as to what patients did or did not want to know. Clinicians were broadly divided into three groups of information-giving style: closed, open, or changing. Pinder concluded that

926 Sudden Death in Epilepsy: Forensic and Clinical Issues

identifying patient wishes necessitates proactive engagement by a clinician to discover what the patient wants to know, and ought to know, about their condition and its treatment. However, the medical literature suggests that many clinicians have taken speciἀc positions on when to discuss SUDEP, such as only if a patient asks, at the time of prescribing antiepileptic drugs, or only in cases with recognized risk factors (Morton et al. 2006). 57.9.3â•…Denying the Patient’s Right Not to Know Some authors contributing to the SUDEP debate have put forward the argument for a patient’s “right not to know” (Beran 2006; Black 2005), even suggesting that in some circumstances clinicians could be sued if they tell patients about SUDEP (Beran et al. 2004). While therapeutic privilege provides a defense to the legal right to know where there is good reason that discussing SUDEP with a patient will cause harm, the concept of a legal right not to know, although recently mooted in the context of routine HIV testing and genetics, is not one that has rooted yet in any established legal concept or legal authority.

57.10â•… SUDEP Information and Education—When and How? Where the decision is made to inform patients about SUDEP, the timing and method of communication continue to be points of debate. 57.10.1â•… Tell if Asked The literature on this question supports disclosure should a patient ask a question relevant to SUDEP (Beran et al. 2004; Black 2005; Brodie and Holmes 2008; Morton et al. 2006; So et al. 2009). However, this approach clearly places the onus of asking upon patients who may not know which questions to frame. Consequently, disclosure is likely to be restricted to those individuals who are already well informed and conἀdent enough to seek information. 57.10.2â•… High-Risk Patients Only It is generally agreed that SUDEP information should be provided to high-risk groups, such as those with poorly controlled seizures, allowing them the opportunity to participate in risk assessment (Brodie and Holmes 2008; Tomson et al. 2008; So 2006). This may include screening for cardiac interventions and discussion of options for supervision (Finsterer and Stollberger 2009; Tomson et al. 2008, 2009). Where patients have epilepsy and learning difficulties, professional guidelines recommend that health professionals be aware of the higher mortality risks and discuss these with the individuals affected, and their families and/or care givers (Stokes et al. 2004). For people with newly diagnosed or mild epilepsy, risk may be lower, but SUDEP deaths do occur in this group as noted by the Scottish Public Services Ombudsman (2009). Nevertheless the assumption has prevailed that SUDEP is a risk for only those with intractable epilepsy (Tomson et al. 2008), which, to some extent, ignores “. . . the larger group of individuals with better but not fully controlled epilepsy who have a lower, but nevertheless real, risk of SUDEP” (Tomson et al. 2008). In the clinical audit of epilepsy-related death in

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 927

the United Kingdom, a number of SUDEP cases were recorded in patients having only rare seizures. Looking at 158 adult patients who had attended specialist care, 5% were known to have had less than one seizure per year, 3% were people noted to have had a single seizure or the ἀrst in many years, and 4% were people considered to be seizure free. Twenty-ἀve percent of the deaths were in people whose records did not record the frequency of seizures, and 22% were in people where the records were not clear. Six percent of the patients who died were not treated with antiepileptic drugs. The audit also identiἀed some 45 people who died during the year who were never referred to secondary care or who died waiting for referral (Hanna et al. 2002). 57.10.3â•…If a Patient Does Not Adhere to Treatment There is general agreement that SUDEP discussion should take place with the nonadherent patient (Black 2005; Morton et al. 2006; So et al. 2009). The challenge is timely identiἀcation of nonadherence. Research in other chronic conditions has shown that it is often difἀcult for a doctor to know who will comply with treatments as prescribed and why (Horne 2006; Wertheimer and Santella 2003). Infrequent reviews limit monitoring. Further, during the consultation, patients often overestimate adherence so as not to disappoint (Horne 2006). Waiting until the patient is known to be nonadherent may be too late, or it may be impossible to change behaviors. The report of the Scottish Public Services Ombudsman (2009) is illustrative as the family reported that the deceased was not taking medications as prescribed, whereas the clinical record noted concerns by the patient about side effects, leading to a reduction in medication, but not problems with adherence. The issue of adherence is not straightforward (Horne 2006; Wertheimer and Santella 2003). A recent guideline for the United Kingdom suggests that medicine taking is “. . . a complex human behavior . . .â•›,” and that unwanted and unused medicines “. . . reflect inadequate communication between professionals and patients . . .â•›” (Nunes et al. 2009). The complexities of adherence are constantly underlined for those of us working with SUDEP. For example we regularly see a troubling scenario, aspects of which are reflected in the stories of David, Celine, and Peter presented in Sudden Unexpected Death in Epilepsy: A Global Conversation (2005). A young person dies, and in discussions with family and friends a picture emerges of a vibrant individual who did not want to have epilepsy. Seizures were embarrassing and a nuisance, but they believed coping well meant not to fuss, and friends were told not to worry. It appears that their doctors may have considered them to be wellcontrolled, adherent, informed patients. However, these patients and their families did not realize epilepsy could be fatal. Frequently, the bereaved families believe that the young adults would have handled their epilepsy just a little differently if they had known .€.€.€. and of course no one can answer this question. We know that issues of career, driving, and the adverse effects of medication are all questions for young people. They may choose to selfmanage in their own way, based on the understanding they have been given. This implies that some coping styles, while apparently successful, if uninformed, may in fact lead to greater vulnerability. Whether an individual patient is perceived to be of high or low risk, behavior cannot be predicted. Private decisions take place outside of the consulting room. Life changes occur, and people with epilepsy will make decisions based on the framework of knowledge they have been given, with some choices leading to fatal outcomes. For example, when a young woman unexpectedly becomes pregnant, she may decide to stop her epilepsy

928 Sudden Death in Epilepsy: Forensic and Clinical Issues

medication without consulting any doctor. National investigations of maternal deaths consistently raise concerns that many pregnant women with epilepsy are concerned about the side effects of medication, and there are a steady number of women who die each year who appear to have stopped medication, with and without the knowledge of their medical team. The importance of preconception counseling for women with epilepsy of child-bearing age is a key recommendation of the report into all maternal deaths released in the United Kingdom (Lewis 2007). Since not all pregnancies are planned, it is imperative that such information must be given at the ἀrst opportunity, with frank two-way communication underpinning true physician–patient concordance (Horne 2006) and hopefully engendering informed patient adherence. 57.10.4â•…Routine Discussion The National Guidelines for England and Wales support discussion about SUDEP following diagnosis, tailored to the individual’s relative risk (Stokes et al. 2004). Routine discussion is also recommended by a growing body of expert opinion (Brodie and Holmes 2008; So et al. 2009; Tomson et al. 2008). With evidence that maintenance of a stable antiepileptic drug regimen might reduce risk, it would be timely for discussion about risks and beneἀts of treatment to include the small risk of fatality from a seizure (Cook 2005; Faught et al. 2008; Hughes 2009; Tomson et al. 2008). In the words of a person with epilepsy, “for many, the decision to take medication is a huge one. If they are not aware of the dangers of seizures as well as the side-effects of medication I do not feel that their decisions are truly informed” (Kearton 2005). Some patients may prefer information in stages, while others may desire as much information as possible on the condition. Strategies for identifying individual preferences may be the use of counseling checklists with SUDEP included, or the provision of preconsultation written epilepsy information (Stokes et al. 2004). Where investigation and diagnosis are delayed following a seizure, although it may be considered inappropriate to mention SUDEP (So et al. 2009), it has been recommended that essential information include how to recognize seizures, ἀrst aid, and the importance of reporting further attacks. Some information on epilepsy routinely imparted to patients may create major disincentives for engagement with treatment. The concern over possible loss of a driving license, or unpleasant side effects, for example, may inhibit reporting of seizures or attendance at appointments. It is logical to fully inform patients about risks soon after diagnosis when there is possibly the greatest potential to communicate the imperative of aiming for seizure freedom through appropriate treatment and lifestyle choices. As time passes, if few seizures are experienced with no apparent harmful effects, there is the potential for patients to become blasé about seizures and more concerned about the daily inconvenience of treatment and lifestyle adaptation. Focusing on comprehensive epilepsy education as the framework for SUDEP discussions is a positive recommendation (So et al. 2009), although models and evaluation of such programs are limited. The Sepulveda Epilepsy Education project (now known as Seizures and Epilepsy Education) was one early interesting example, in that fear was seen as a key issue to tackle even before SUDEP became a topic of public discussion (Helgeson et al. 1990). However, development of such programs is lacking. Interestingly the SUDEP discussion has injected some urgency into discussion on how epilepsy education generally should take place, something which is long overdue (Prinjha et al. 2005).

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 929

It is understood that individuals vary in their approach to health information. Some seek it to cope well while others cope by avoidance (Andrewes et al. 1999; Pinder 1990; Politi et al. 2007). Clearly health information giving must aim to minimize anxiety and maximize the effectiveness of communication. This already occurs in epilepsy where risk communication is a core aspect of management. Rather than excluding SUDEP on the premise that this will avoid patient anxiety, why not include the information sensitively as part of the existing education? There may be no need to use the term SUDEP if the clinician can convey the substance of the risk. For example, the rare risk of dying from a seizure, not only through drowning and accidents but also following a seizure, can be mentioned (Cook 2005; Tomson et al. 2008). Additional information can be provided if patients request it. In some cases, gentle questioning of patients as to their fears may indicate whether questions lie unanswered. A patient’s attitude to information and involvement is likely to vary over time so that revisiting information topics is important (Pinder 1990). For those whose risk is low, a frank explanation of what is known may be reassuring (Mittan 2005; So et al. 2009). Individualization of information does not mean ἀltering or excluding information. SUDEP is in the public domain. An internet search will bring many hits, and not all offer rational, accurate material. The use of the internet grows daily (Escoffery et al. 2008; Fox 2007; Lu et al. 2005), and patients who discover SUDEP themselves may feel disappointed or even angry that their physician did not mention it. Trust may be damaged, and attempts to put the issue in context may no longer be as well received by the patient (Cook 2005). Where physicians routinely offer advice on SUDEP, they report few patient concerns, which suggests that with practice, providing such information to patients is successful (Morton et al. 2006). 57.10.5â•…Risk Communication If a message is to be meaningful to the recipient, it must be in a language that can be understood; however, in health care, this is not always so. For example, research literature on lay understanding of any health risks shows that the public has difficulty with odds and percentages (Weinstein 1999). Weinstein (1999) argues that for successful communication, the evidence points to the need for a basic set of core information rather than statistics. This includes the identity and severity of the potential harm, the likelihood of harm under various circumstances, and the possibility and difficulty of reducing that harm. One note of caution in risk communication is that with nearly all hazards, people at risk apply an optimistic bias (Weinstein 1999). False assurance can occur if risks are communicated in such a reassuring way that they are interpreted as meaning no risk at all. Research into epilepsy risk communication is urgently required, and there is a burgeoning ἀle of literature of risk communication from which to draw (McComas 2006).

57.11╅ SUDEP and Bereavement Communication of SUDEP risk rightly begins with patient well-being in mind. Neverthe� less,€where SUDEP occurs in families that are ignorant of the phenomenon, it multiplies the distress to the bereaved. The shock and sadness extends beyond the family, to friends, work colleagues, and even the associated health professionals (Cook 2005; Kennelly and Riesel

930 Sudden Death in Epilepsy: Forensic and Clinical Issues

2002). Bereavement affects people uniquely, but there is evidence that sudden bereavement can complicate the grieving process because people are unprepared. Early intervention can reduce associated morbidity (Raphael 1977; Yates et al. 1990). Little research has been carried out regarding SUDEP bereavement. However, studies into the impact of SIDS are informative. These studies found acute distress and long-lasting damage, particularly relationship conflict, difficulties with surviving children, and anxiety about future children becoming victims (Woodward et al. 1985). In both SUDEP and SIDS, the death is wholly unexpected, the cause is unknown, and families face the bewildering interventions of police, a postmortem examination, ongoing inquiry, and uncertainty. Some families experiencing SUDEP have other family members with epilepsy (Kennelly and Riesel 2002). Like SIDS families, they can fear a recurrence. Unlike SIDS, which is well publicized, SUDEP may create an additional burden on the grieving family in explaining the death to the police, family, friends, and the local community who are ignorant of SUDEP (Sudden Unexpected Death in Epilepsy: A Global Conversation 2005). The lack of research on the extended impact of SUDEP leaves a serious gap in our understanding and should be rectiἀed.

57.12â•… SUDEP Support Services Community-based SUDEP support services have been slow to appear internationally, with the lack of SUDEP awareness being a barrier to their development. The ἀrst SUDEP support service, Epilepsy Bereaved, was established in the United Kingdom in 1996 (www .sudep.org). The service, not dissimilar to SIDS support programs (Woodward et al. 1985), today offers information, listening, and support through a dedicated help line, group meetings, and memorial services. Resources are available for families and health professionals. Families receive a regular magazine and can become active in the work of the organization if this helps their grieving process. Families describe the service as a “lifeline.” In addition, Epilepsy Bereaved is funding and working with researchers in the United Kingdom, as well as raising public awareness of SUDEP and campaigning for more effective epilepsy and coroner’s services. Epilepsy guidelines in the United Kingdom recommend that health professionals contact families to offer their condolences, invite them to discuss the death, and offer referral to bereavement counseling and a SUDEP support group (Stokes et al. 2004). With respect to communications with doctors following an epilepsy death, a qualitative study found that bereaved people reported positively when the medical team had offered condolences, support, and information, and negatively when contact was not offered, delayed, or was defensive in nature (Kennelly and Riesel 2002). Families were grateful for the offer regardless of whether it was taken up. Internationally, few services exist to meet the needs of families affected by SUDEP, but activity is increasing with international networking seen as a key to effective use of experience and resources. The Irish Epilepsy Association (www.epilepsy.ie/) has worked with Epilepsy Bereaved to organize meetings for families. In Australia, epilepsy agencies support and link SUDEP families, organize a biennial memorial service, and develop resource€material. A SUDEP-related research fund has also been recently established (www.epilepsyaustralia╉ .net/). In Canada, SUDEP Aware holds meetings twice a year for bereaved families to connect and€share experiences, offers referral to counseling, and is a contact point for North American

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 931

families to network with other families (www.sudepaware.org/). In the United States, bereaved families have supported the development of a SUDEP-targeted research program through CURE (www.cureepilepsy.org/home.asp) and have also participated in the American Epilepsy Society and the Epilepsy Foundation Joint Task Force on SUDEP, and the National Institute of Health SUDEP symposium in 2008. Support agencies also have an important role to play in relation to forensic investigation of SUDEP. Through a collaborative approach, the agency can assist the work of the forensic institute by offering extended contact and counseling to families and friends of the bereaved. Where families are referred to a support agency immediately after a death, the rationale for a postmortem examination can be explained, and support can be offered by others who have themselves experienced the fear and trauma of facing such an ordeal. Newly bereaved families can be assisted to understand the longer-term beneἀt that may result through family participation in a thorough investigation (Lathers and Schraeder 2009). In the United Kingdom, Epilepsy Bereaved has worked with the Royal College of Pathologists, leading to national guidelines on the investigation of epilepsy-related deaths. In Australia, collaboration between the Epilepsy Foundation of Victoria and the Victorian Institute of Forensic Medicine has helped to raise awareness of SUDEP in health professionals and the community, and promote research including death scene investigation.

57.13â•… Summary An examination of the evolution of epilepsy care over the past 15 years reveals that SUDEP advocacy groups have been working collaboratively and effectively with health professionals to address questions previously unanswered by the scientiἀc community, and to facilitate signiἀcant improvements in epilepsy care. Their questions have prompted reÂ�examination of services when others could not despite years of effort (Hanna et al. 2002). As a lowerprevalence condition, epilepsy has struggled to attract attention and resources (Brown 1995; Chronic Disease Directors 2003). However, the SUDEP issue has harnessed community participation and leadership, which has proved to be a signiἀcant, positive force for change. The discussion of epilepsy has been elevated to the highest levels (All Party Parliamentary Group on Epilepsy 2007; Department of Health 2001b), and guidelines for improved care have emerged (Scottish Intercollegiate Guidelines Network 2003; Stokes et €l. 2004). Ongoing research is essential to evaluate progress and guide future improvements. Nevertheless, in the wake of the SUDEP debate, it would appear that epilepsy might ἀnally be coming out of the shadows. This chapter has presented the arguments for and against informing patients and care givers about SUDEP, and takes the position that for the vast majority of patients the arguments for informing are convincing. While the reduction of SUDEP may be partly influenced by quality medical management, where risk reduction is the core strategy, patient education and participation is integral to success. More research is urgently needed on the most effective way to do this. It is sometimes suggested that we do not hear the voices of epilepsy patients in the SUDEP discussion. While it might be true that their opinions are not often reflected in this debate through formal research publications, nevertheless in some countries many people with epilepsy are now well aware of SUDEP. They have heard of deaths in the media, lost friends to SUDEP, or learned about it through their epilepsy support agencies. Consequently,

932 Sudden Death in Epilepsy: Forensic and Clinical Issues

people with epilepsy are frequently supporters of SUDEP organizations. Our observations are that many people with epilepsy feel that it is right for them to know, and that, to some degree, it is considered empowering to be trusted with the information, which enables them to make informed choices about their epilepsy management. We would like to conclude our discussion with the words of one such young woman with epilepsy. I had uncontrolled seizures for many years never realizing that people could die from a seizure. . . . I had frequent seizures but I never considered this to be a serious condition—just something I had to acceptâ•›. . . Some of the things I discovered about epilepsy and its treatment were not easy to hear, and when a young man I knewâ•›. . . subsequently died as a result of a seizure, the issue of death really hit home. However, I don’t live in fear of death. Now that I know about epilepsy I make choices to take care of myself. I strongly believe people have the right to know about their condition. (Kearton 2005)

References All Party Parliamentary Group on Epilepsy. 2007. Wasted Money Wasted Lives: The Human and Economic Cost of Epilepsy in England. London: All Party Parliamentary Group on Epilepsy. Andrewes, D., K. Camp, C. Kilpatrick, and M. Cook. 1999. The assessment and treatment of concerns and anxiety in patients undergoing presurgical monitoring for epilepsy. Epilepsia 40 (11): 1535–1542. Beran, R. 2006. SUDEP—to discuss or not to discuss: That is the question. Lancet Neurology 5 (6): 464–465. Beran, R. G., S. Weber, R. Sungaran, and N. Venn. 2004. Review of the legal obligations of the doctor to discuss Sudden Unexplained Death in Epilepsy (SUDEP)—A cohort controlled comparative cross-matched study in an outpatient epilepsy clinic. Seizure 13 (7): 523–528. Black, A. 2005. SUDEP Beran whether to tell and when? Med Law 24 (1): 41–49. Bolitho v City and Hackney Health Authority. 1998. AC 232. Brodie, M. J., and G. L. Holmes. 2008. Should all patients be told about sudden unexpected death in epilepsy (SUDEP)? Pros and cons. Epilepsia 49 (S9): 99–101. Brown, S. W. 1995. What resources? Addressing the needs of the epilepsy community. Seizure 4 (3): 207–210. Chester v Afshar. 2005. 1 AC 134. Chronic Disease Directors. 2003. The role of public health in addressing lower prevalence chronic conâ•›ditions: The example of epilepsy. Atlanta, GA: Centers for Disease Control and Prevention. Cook, M. 2005. Reflecting on my clinical experience. In Sudden Unexpected Death in Epilepsy: A Global Conversation, ed. D. Chapman, B. Moss, R. Panelli, and R. Pollard. Camberwell: Epilepsy Australia & Epilepsy Bereaved. Couldridge, L., S. Kendall, and A. March. 2001. A systematic overview—A decade of research. The information and counselling needs of people with epilepsy. Seizure 10 (8): 605–614. Dalrymple, J, and J. Appleby. 2000. Cross sectional study of reporting of epileptic seizures to general practitioners. BMJ 320 (7227): 94–97. Department of Health. 2001a. The Expert Patient: A New Approach to Chronic Disease Management for the 21st Century. London: Department of Health. Department of Health. 2001b. Annual Report of the Chief Medical Officer: Epilepsy—Death in the Shadows: On the state of the public health. London: Department of Health. Department of Health. 2003. Improving Services for People with Epilepsy: Department of Health Action Plan in Response to the National Clinical Audit of Epilepsy-Related Death. London: Department of Health.

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 933 Deriche v Easling Hospital NHS Trust. 2003 EWHC 3104 (QB). Escoffery, C., C. Dilorio, K. Yeager, F. McCarty, E. Robinson, E. Reisinger, T. Henry, and A. Koganti. 2008. Use of computers and the internet for health information by patients with epilepsy. Epilepsy Behav 12 (1): 109–114. Faught, R. E., M. S. Duh, J. R. Weiner, A. Guerin, and M. C. Cunnington. 2008. Nonadherence to antiepileptic drugs and increased mortality: Findings from the RANSOM study. Neurology 71 (20): 1572–1578. Finsterer, J., and C. Stollberger. 2009. Cardiopulmonary surveillance to prevent SUDEP. Lancet Neurol 8 (2): 131–132. Fox, S. 2007. E-patients with a chronic disability or chronic disease. Washington: Pew internet and American life project. http://www.pewinternet.org/~/media//Files/Reports/2007/EPatients_ Chronic_Conditions_2007.pdf. Fraenkel, L., and S. McGraw. 2007. Participation in medical decision making: The patients’ perspective. Med Decis Making 27 (5): 533–538. General Medical Council. 2008. Consent: Patients and Doctors Making Decisions Together. UK: General Medical Council. Gaitatzis, A., A. L. Johnson, D. W. Chadwick, S. D. Shorvon, and J. W. Sander. 2004. Life expectancy in people with newly diagnosed epilepsy. Brain 127 (Pt 11): 2427–2432. Gayatri, N. A., M. C. Morrall, V. Jain, P. Kashyape, K. Pysden and C. Ferrie. 2010. Parental and physician beliefs regarding the provision and content of written sudden unexpected death in epilepsy (SUDEP) information. Epilepsia 51 (5): 777–782. Hanna, N. J., M. Black, and J. W. S. Sander et al. 2002. The National Sentinel Clinical Audit of EpilepsyRelated Death: Epilepsy—Death in the Shadows. Norwich: The Stationary Office. Helgeson, D. C., R. Mittan, S. Y. Tan, and S. Chayasirisobhon. 1990. Sepulveda Epilepsy Education: The efficacy of a psychoeducational treatment program in treating medical and psychosocial aspects of epilepsy. Epilepsia 31 (1): 75–82. Hitiris, N., S. Suratman, K. Kelly, L. J. Stephen, G. J. Sills, and M. J. Brodie. 2007. Sudden unexpected death in epilepsy: A search for risk factors. Epilepsy Behav 10 (1): 138–141. Horne, R. 2006. Compliance, adherences, and concordance: Implications for asthma treatment. Chest 130 (1 Suppl.): 65S–72S. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14 (2): 280–287. Katz, J. 1984. The Silent World of Doctor and Patient. New York, NY: Free Press. Kearton, M. 2005. Living with the risks. In Sudden Unexpected Death in Epilepsy: A Global ConÂ� versation, ed. D. Chapman, B. Moss, R. Panelli and R. Pollard. Camberwell: Epilepsy Australia & Epilepsy Bereaved. Kennelly, C., and J. Riesel. 2002. Sudden Death and Epilepsy—The Views and Experiences of Bereaved Relatives and Carers. UK: Epilepsy Bereaved. Langan, Y., L. Nashef, and J. W. Sander. 2005. Case–control study of SUDEP. Neurology 64 (7): 1131–1133. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies and SUDEP. Epilepsy Behav 14 (4): 573–576. Lewis, G. 2007. The conἀdential enquiry into maternal and child health (CEMACH). Saving mothers’ lives: Reviewing maternal deaths to make motherhood safer—2003–2005. The Seventh Report on Confidential Enquires into Maternal Deaths in the United Kingdom. London: CEMACH. Lewis, S., S. Higgins, and M. Goodwin. 2008. Informing patients about sudden unexpected death in epilepsy: A survey of specialist nurses. Br J Neurosci Nurs 4 (11): 30–34. Lorig, K. R., and H. R. Holman. 2003. Self-management education: History, deἀnition, outcomes, and mechanisms. Ann Behav Med 26 (1): 1–7. Lu, C., E. C. Wirrell, and M. Blackman. 2005. Where do families of children with epilepsy obtain their information? J Child Neurol 20 (11): 905–910. Marteau, T. 2008. Detection is not as harmful as it might seem. BMJ 336: 973–974.

934 Sudden Death in Epilepsy: Forensic and Clinical Issues McComas, K. A. 2006. Deἀning moments in risk communication: 1996–2005. J Health Commun 11 (1): 75–91. Mittan, R. 1986. Fear of seizures. In Psychopathology in Epilepsy: Social Dimensions, ed. S. Whitman and B. P. Hermann. New York, NY: Oxford University Press. Mittan, R. 2005. Managing fear. In Sudden Unexpected Death in Epilepsy: A Global Conversation, eds. D. Chapman, B. Moss, R. Panelli and R. Pollard. Camberwell: Epilepsy Australia & Epilepsy Bereaved. Monte, C. P., J. B. Arends, I. Y. Tan, A. P. Aldenkamp, M. Limburg, and M. C. de Krom. 2007. Sudden unexpected death in epilepsy patients: Risk factors. A systemic review. Seizure 16 (1): 1–7. Moon, R. Y., R. S. Horne, and F. R. Hauck. 2007. Sudden infant death syndrome. Lancet 370 (9598): 1578–1587. Morton, B., A. Richardson, and S. Duncan. 2006. Sudden unexpected death in epilepsy (SUDEP): Don’t ask don’t tell? J Neurol Neurosurg Psychiatry 77 (2): 199–202. Nashef, L. 1995. Sudden unexpected death in epilepsy: Incidence, circumstances and mechanisms. MD thesis, University of Bristol. Nashef, L., J. F. Annegers, and S. Brown. 1997. Introduction and overview. Epilepsia 38 (S11): S1–S2. National Health Priority Action Council. 2006. National Chronic Disease Strategy. Canberra: Australian Government Department of Health and Ageing. Nunes, V., J. Neilson, N. O’Flynn et al. 2009. Clinical Guidelines and Evidence Review for Medicines Adherence: Involving Patients in Decision about Prescribed Medicine and Supporting Adherence. London: National Collaborating Centre for Primary Care and Royal College of General Practitioners. Opeskin, K., and S. F. Berkovic. 2003. Risk factors for sudden unexpected death in epilepsy: A controlled prospective study based on coroners cases. Seizure 12 (7): 456–464. Otoom, S., A. Al-Jishi, A. Montgomery, M. Ghwanmeh, and A. Atoum. 2007. Death anxiety in patients with epilepsy. Seizure 16 (2): 142–146. Pearce v United Healthcare NHS Trust. 1999. PIQR P53. Pedley, T. A., and W. A. Hauser. 2002. Sudden death in epilepsy: A wake-up call for management. Lancet 359 (9320): 1790–1791. Petticrew, M., A. Sowden, and D. Lister-Sharp. 2001. False-negative results in screening programs: Medical, psychological and other implications. Int J Technol Assess Health Care 17 (2): 164–170. Pinder, R. 1990. What to expect: Information and the management of uncertainty in Parkinson’s disease. Disabil Soc 5 (1): 77–92. Politi, M. C., P. J. K. Han, and F. Nananda. 2007. Communicating the uncertainty of harms and beneἀts of medical interventions. Med Decis Making 27 (5): 681–695. Preston, J. 1997. Information of sudden deaths from epilepsy. Epilepsia 38 (S11): S72–S74. Prinjha, S., A. Chapple, A. Herxheimer, and A. McPherson. 2005. Many people with epilepsy want to know more: A qualitative study. Fam Pract 22 (4): 435–441. Raphael, B. 1977. Preventative intervention with the recently bereaved. Arch Gen Psychiatry 34 (12): 1450–1454. Riddle, M. C. 1980. A strategy for chronic disease. Lancet 2 (8197): 734–736. Rogers v Whitake. 1992. 175 CLR 479. Scottish Intercollegiate Guidelines Network. 2003. Diagnosis and Management of Epilepsy in Adults. Edinburgh: Scottish Intercollegiate Guidelines Network. Scottish Public Services Ombudsman. 2009. Case 200700075. Shaw, C., K. Abrams, and T. M. Marteau. 1999. Psychological impact of predicting individuals’ risks of illness: A systematic review. Soc Sci Med 49 (12): 1571–1598. So, E. L. 2006. Demystifying sudden unexplained death in epilepsy—Are we close? Epilepsia 47 (S1): 87–92.

Challenges in Overcoming Ethical, Legal, and Communication Barriers in SUDEP 935 So, E. L., J. Bainbridge, J. R. Buchhalter et al. 2009. Report of the American Epilepsy Society and the Epilepsy Foundation joint task force on sudden unexplained death in epilepsy. Epilepsia 50 (4): 917–922. Spratling, W. P. 1904. Prognosis. In Epilepsy and Its Treatment. Philadelphia, PA: W. B. Saunders. Stokes, T., E. J. Shaw, A. Juarez-Garcia, J. Camosso-Steἀnovic, and R. Baker. 2004. Clinical guidelines and evidence review for the epilepsies: Diagnosis and management in adults and children in primary and secondary care, CG20 full guideline. London: Royal College of General Practitioners. In Sudden Unexpected Death in Epilepsy: A Global Conversation. 2005. eds. D. Chapman, B. Moss, R. Panelli and R. Pollard. Camberwell: Epilepsy Australia and Epilepsy Bereaved. Surges, R., R. D. Thijs, H. L. Tan, and J. W. Sander. 2009. Sudden unexpected death in epilepsy: Risk factors and potential pathomechanisms. Nature Reviews Neurology 5: 492–504. Taylor, J. A. 2002. Determination of Sheriff James Taylor, Sheriff of the Sherifdom of Glasgow and Strathkelvin at Glasgow. Inquiry held under fatal accidents and sudden death inquiry (Scotland) Act 1976 into the death of Colette Marie Findlay. Taylor, S. E. 1991. Asymmetrical effects of positive and negative events: The mobilization–minimization hypothesis. Psychol Bull 110 (1): 67–85. Tomson, T., L. Nashef, and P. Ryvlin. 2008. Sudden unexpected death in epilepsy: Current knowledge and future directions. Lancet Neurol 7 (11): 1021–1031. Tomson, T., L. Nashef, and P. Ryvlin. 2009. Cardiopulmonary surveillance to prevent SUDEP: Authors’ reply. Lancet Neurol 8 (2): 132–133. Videto v Kennedy. 1981. 125 DLR (3d) 127. Weinstein, N. 1999. What does it mean to understand a risk? Evaluating risk comprehension. J Natl Cancer Inst Monogr 25: 15–20. Welsh Assembly Government. 2009. Designed for people with chronic conditions: Service DevelÂ� opment Directive—Epilepsy. Cardiff: Welsh Assembly Government. Wertheimer, A. I., and T. M. Santella. 2003. Medication compliance research: Still so far to go. The Journal of Applied Research in Clinical and Experimental Therapeutics 3 (3): 254–261. Whitney, S. N., M. Holmes-Rovner, and H. Brody et al. 2008. Beyond shared decision making: An expanded typology of medical decisions. Med Decis Making 28 (5): 699–705. Woodward, S., A. Pope, W. J. Robson, and O. Hagan. 1985. Bereavement counselling after sudden infant death. Br Med J (Clin Res Ed) 290 (6465): 363–365. Yates, D. W., G. Ellison, and S. McGuiness. 1990. Care of the suddenly bereaved. BMJ 301 (6742): 29–31.

Bereavement and Sudden Unexpected Death in Epilepsy Lina Nashef Lene Sahlholdt

58

Contents 58.1 Introduction 58.2 Grief 58.3 Grief Therapy 58.4 Bereavement in SUDEP: How Is It Different? 58.5 Self-Help Groups and Epilepsy Charities 58.6 Information, Physicians, and SUDEP 58.7 Publications by Bereaved Relatives in SUDEP 58.7.1 A Global Conversation 58.7.2 SUDEP: The Views and Experiences of Bereaved Relatives and Carers (Epilepsy Bereaved 2002) 58.8 Bereaved Relatives in the Medical Literature: An Interview Study 58.9 Conclusion References

937 937 938 938 939 939 939 939 939 940 941 942

58.1â•…Introduction Being sudden and unexpected, often affecting the young, sudden unexpected death in epilepsy (SUDEP) leaves much sorrow, pain, turmoil, and anguish in its wake. Communication between the bereaved and health services in such circumstances can be very helpful but can also be demanding for all concerned and needs to be handled with sensitivity. Regrettably, there has been little focus in the SUDEP medical literature on bereaved families or partners. This chapter will by necessity be concise, its intention being to draw attention to this area and to the need for further research. It will also include suggestions for management in this situation. Before discussing bereavement following SUDEP, grief following bereavement in general will be briefly considered.

58.2â•… Grief Grief is the complex psychic pain in the bereaved related to loss and longing for the dead person (Bonanno et al. 2007). The loss is often felt to be unfair and senseless. Grief is often mixed with many other feelings such as guilt and anger toward the dead person and may also be accompanied by fear of the future. Relatives grieve and look for comfort and consolation. They ask themselves many questions, which reflect feelings of guilt, disbelief, and/or anger. Why him/her? Why our family? Could we have done something 937

938 Sudden Death in Epilepsy: Forensic and Clinical Issues

to prevent it? Could the doctors have done more? Anger or mistrust toward the medical team responsible for the treatment may be a part of these mixed feelings. The most severe grief is often said to be after the death of a child, a parent, or a spouse. Grieving is often long lasting, with feelings of loss and longing often revived on special occasions for many years. Grieving may be accompanied by sleep disorders, attention problems, and depression, and thus may have some resemblance to aspects of depression or posttraumatic stress disorder. Long-lasting grief is not necessarily dysfunctional. Ways of grieving depend on personality and culture. Wijngaards-de Meij et al. (2007) found that grief is related to the “attachment style.” An insecure or weak attachment may produce longer and stronger grief than if the attachment is conἀdent. Most people, including clinicians, often understand grief as a step-by-step loosening of attachment or of bonds to the dead person in which the goal is to break the bonds, the assumption being that this provides the basis for the potential to “move on” with life. This may include, for example, sending good-bye letters to the deceased. However, there is no evidence for this step-by-step understanding of the grief process, although it is “common knowledge” and widely assumed (Wortman and Silver 2001). Newer evidence-based descriptive models offer a more complex view whereby the grief reaction is seen as a mixture in terms of focus not only on the loss but also on regeneration and restoration. It involves looking back and forward in a wave-like motion, an oscillation between concentration on loss and restoration of life. Breaking the bonds is not necessary for moving on (Stroebe and Schut 1999).

58.3â•… Grief Therapy While grief therapy may be seen currently as a way of acknowledging grief, referring all who lose a dear person for grief therapy is pointless as it is neither necessary nor effective in uncomplicated grieving. It is only helpful when asked for, if there are other risk factors, or if the grief is complicated, for example, by depression, unless the grieving person is a child who lost a parent. Grief therapy may, on the contrary, be both necessary and helpful if grief is complicated and otherwise would result in the bereaved being unable to cope with daily life (Stroebe et al. 2005, 2007).

58.4â•… Bereavement in SUDEP: How Is It Different? While bereavement following SUDEP shares features common to bereavement in general, there are a number of circumstances that add additional dimensions to some aspects of the grief reaction: the unexpected and sudden nature of SUDEP, which occurs in young and, apart from the epilepsy, often otherwise well individuals; the lack of abnormality on autopsy to explain the death; the lack of recognition of risk by the medical profession, at least until recently; the ongoing general lack of acknowledgement of the risk; and the dayto-day choices someone with epilepsy and their relatives might make in managing their condition and in leading their lives. Death may be the culmination of years of living with epilepsy, a chronic condition requiring major adjustment, where the requirements of safety need to be constantly reconciled with the desire for independence.

Bereavement and Sudden Unexpected Death in Epilepsy

939

58.5â•… Self-Help Groups and Epilepsy Charities This situation has led bereaved relatives in many countries, starting with “Epilepsy Bereaved” in the United Kingdom in 1993 (http://www.sudep.org), to form highly motivated self-help groups. In this, they have demonstrated commitment and strength of purpose, arising from their grief and fuelled by their dissatisfaction with the status quo. Their aims have been primarily to support and inform other bereaved relatives, and also to redress wider deἀciencies by increasing awareness of SUDEP, encouraging research as well as best medical and self-management. In doing so, self-help groups have provided a badly needed resource and backup for other bereaved relatives, often, but perhaps not always, welcomed by the medical profession or other epilepsy charities.

58.6â•…Information, Physicians, and SUDEP Although relevant to this discussion, it is outside the scope of this short chapter to address the ongoing debate among health workers and bereaved relatives regarding the need for physicians and nurses to improve information provision on risks associated with epilepsy, and how this is best provided, the role of guidelines, or legal aspects (Hanna and Panelli 2010, Chapter 57 in this book). Much of the expressed focus of differences in opinion can be taken to reflect limited research to guide best practice in this area and the need to make a distinction between the views and interests of the bereaved and those of the late relative or of people with epilepsy in general. There is, however, another dimension as yet unexpressed, but which may nevertheless be necessary to consider. While dwarfed by the strength of the grief reaction in relatives, we should also acknowledge the grief reaction in a health worker whose patient dies unexpectedly “on his or her watch.”

58.7â•… Publications by Bereaved Relatives in SUDEP The views of bereaved relatives are referred to in the two publications below, which are discussed in more detail in Chapter 57 (Hanna and Panelli 2010). 58.7.1â•… A Global Conversation In the publication A Global Conversation (Chapman et al. 2005), views of bereaved relatives were expressed and included the following: (a) not being informed of the risk, (b) believing that the dead relatives were not informed, (c) experiencing a taboo in the medical world, and (d) wishing they were informed. 58.7.2â•…SUDEP: The Views and Experiences of Bereaved Relatives and Carers (Epilepsy Bereaved 2002) Kennelly and Riesel (2002) conducted a qualitative study in parallel with the United Kingdom National Sentinel Clinical Audit of Epilepsy Related Death. A summary was published by Epilepsy Bereaved, although not to our knowledge, in a peer-reviewed journal. In this study, six focus groups were held in the United Kingdom to identify issues

940 Sudden Death in Epilepsy: Forensic and Clinical Issues

that bereaved relatives and carers considered important. In addition, 78 semistructured telephone interviews with bereaved relatives were carried out to explore important issues in further detail. The summary report concludes with a number of very useful broad and speciἀc recommendations about epilepsy care and response to bereavement, many in line with our recommendations below.

58.8â•…Bereaved Relatives in the Medical Literature: An Interview Study As has already been mentioned, there has been little focus in the medical literature on bereaved relatives. In 1993/1994, one of the authors (LN), and a colleague (S. Garner), interviewed 34 sets of self-referred bereaved relatives, carers, or partners of people with epilepsy-related deaths (Nashef et al. 1998), of whom 26 fulἀlled the deἀnition for SUDEP. In€addition to detailed assessment of epilepsy diagnosis, control, treatment, and circumstances of death, the semistructured interviews also touched on attitudes and perception of relatives/carers or partners. The publication arising from this research focused on the quantitative aspects of the research, including epilepsy classiἀcation, treatment history, and triggers for and evidence of a terminal seizure. Of note is that the SUDEP cases identiἀed did not all have severe epilepsy. Of the 26 cases, 8 had fewer than 10 generalized tonic–clonic seizures before SUDEP, 10 between 10 and 100, and 7 greater than 100 (1 unknown with continuing partial seizures but no tonic–clonic seizure reported for over 20 years). Five had a history of status epilepticus or serial seizures. Eleven had generalized epilepsy (9€of whom were idiopathic generalized), 10 localization-related epilepsy, and 5 undetermined. Although there was no control group, it was felt that in some, a number of facÂ�tors€relating to lifestyle and medical treatment could have influenced seizure control and, potentially, outcome. In addition, the personal and qualitative dimension was extremely enlightening. Many of the bereaved had witnessed severe epileptic seizures and had often been reassured by the medical profession of their benign nature, falsely as it turned out, in€their case. Yet, intuitively, many “knew” seizures were potentially dangerous and lived in fear of possible disaster with constant anxiety for their loved ones. Had risks associated with seizures been discussed with some of these relatives, such information would not have come as a surprise. Conversely, for another, the possibility that epilepsy could “kill” had€not€been considered at all. Feelings of guilt and blame were also evident, although not always explicit. Blame often focused on the medical profession and health services, particularly in relation to lack of information being made available during life. However, there€were also other concerns, relating to care, support, and communication during life and, to the bereaved, after death. In relation to care after death, examples of issues raised included being denied the opportunity to explore the environment or circumstances of death, not being offered the opportunity to ask questions or go over medical details, not being linked up with support services, lack of contact after death by the general practiÂ�tioner or the specialist, learning about SUDEP for the ἀrst time from the press, disagreement with death certiἀcation (e.g., status epilepticus entered as cause for death without supporting evidence), and SUDEP denial. During life, concerns related to medical matters are the lack of support, being discharged from specialist services too soon, issues relating to medication changes, not being referred to specialist services, and especially not acknowledging or appreciating the potential seriousness of the condition. Some felt that,

Bereavement and Sudden Unexpected Death in Epilepsy

941

had it been discussed, this would have allowed them to express their concerns and could have led them to be more proactive in managing the epilepsy and in pushing for more medical help and support for their relative. When the above research was being carried out, more than 15 years ago, bereaved relatives felt very isolated in their loss. The deaths occurred entirely unexpectedly and were apparently unexplained on a background of a chronic condition labeled as “benign,” but which required much support from the family. Many stated that they had been categorically told, or led to believe, that epilepsy could not be fatal. Almost all stated in retrospect that they would have preferred to know of the possibility of premature death, however remote. Many would have preferred to have been given the opportunity to acknowledge epilepsy as a serious condition rather than have intuitive fears dismissed, although, as already stated, many still harbored such fears despite reassurances to the contrary. Many regretted not being given the opportunity of making informed decisions regarding lifestyle and treatment. Counseling was not usually offered. Contact with the general practitioner or specialist was helpful when it took place after the death but was sometimes limited or absent (Nashef et al. 1998). At the time, we recommended that information about mortality and response to seizures be included in leaflets on epilepsy, and that in the event of SUDEP, the general practitioner (family doctor) makes early contact with bereaved relatives and the specialist writes with an open offer of a meeting. We also suggested that information about SUDEP, selfhelp groups, other supporting agencies, and counseling services be given as appropriate.

58.9â•… Conclusion There is more awareness now of the entity of SUDEP, and although the intense sense of isolation that characterized the feelings of bereaved relatives before may no longer be the same, it is likely that much of the above still applies. This remains a very difficult and tragic situation, sometimes, if not often, inadequately handled. On the basis of this study, we favor better information provision on risks associated with epilepsy during life, this being necessary for informed choice and optimizing treatment, self–management, and lifestyle. In the event of an epilepsy-related death, especially if sudden and unexpected, grief is a normal reaction and support should be offered. We recommend that health workers inform other members of the team about any death, and that letters of condolences are sent or contact made at an early opportunity, with an offer of an appointment for partners or relatives, both older and younger members of the family. Sympathy, support, and the opportunity to go over any aspects of the medical history or treatment are likely to be welcomed by relatives even if some chose not to, or are unable to, take up the offer immediately or at any stage: it is important to be open and prepared to accept all the mixed emotions of grief, anger, or aggression that the bereaved may be experiencing, including questioning one’s authority. Referral for appropriate counseling services and to self-help groups should also be offered. Offering psychological treatment—grief therapy—is a modern way of acknowledging grief. It is not always necessary but may be welcome and helpful to some. Although the above recommendations are based on experience in the United Kingdom, they are likely to have wider application. Needless to say, more research exploring current experiences of bereaved relatives and optimal management of bereavement following SUDEP in different settings is needed.

942 Sudden Death in Epilepsy: Forensic and Clinical Issues

References Bonanno, G. A., Y. Neria, A. Mancini, K. G. Coifman, B. Litz, and B. Insel. 2007. Is there more to complicated grief than depression and posttraumatic stress disorder ? A test of incremental validity. J Abnorm Psychol 116 (2): 342–351. Chapman, D., B. Moss, R. Panelli, and R. Pollard. 2005. Sudden Unexpected Death in Epilepsy—A Global Conversation. Camberwell, Australia: Epilepsy Australia Ltd. Kennelly, C., and J. Riesel. 2002. Sudden death and epilepsy. The views and experiences of bereaved relatives and carers. Epilepsy Bereaved, Oxford, U. K. Nashef, L., S. Garner, J. W. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64: 349–352. Stroebe, M., and H. Schut. 1999. The dual process model of coping with bereavement: Rationale and description. Death Studies 23: 197–224. Stroebe, M., H. Schut, and W. Stroebe. 2005. Grief work, disclosure and counselling: Do they help the bereaved? Clin Psychol Rev. 25: 395–414. Stroebe, M., H. Schut, and W. Stroebe. 2007. Health outcomes of bereavement. Lancet 370 (9603): 1960–1973. Wijngaards-de Meij, L., M. Stroebe, and H. Schut. 2007. Neuroticism and attachment insecurity as predictors of bereavement outcome. J Res Pers 41: 498–505. Wortman, C. B., and R. C. Silver. 2001. The myths of coping with loss revisited. In Handbook of Bereavement Research, ed. M. S. Stroebe, R. O. Hansson, W. Stroebe, and H. Schut, 405–429. Baltimore, MD: United Book Press.

SUDEP A Clinical and Communicative€Conundrum

59

Paul L. Schraeder Claire M. Lathers The risk of sudden unexpected death is associated with multiple disease categories, with the group most commonly at risk being that with atherosclerotic heart diseases (Lathers, Schraeder, and Bungo 2008a, 2008b; Kuller, Lilienfeld, and Fisher 1967). Other cardiac entities such as rheumatic valvular disease, hypertensive heart disease, congenital heart disease, myocarditis, and genetically determined arrhythmogenic channelopathies also put patients at risk for sudden death. In the pulmonary disease category, the commonly occurring examples of causes of sudden death are embolism and asthma. Also, it has long been known that persons with psychiatric disorders have a risk of sudden death, but whether the risk is associated with the psychiatric disorder itself or with the use of psychotropic drugs, or both, is not clear (Wendkos 1979). Others at risk for sudden death are members of ethnic-related entities, such as Filipino men who are victims of sudden death while sleeping, the phenomenon being called bangungut, as well as some refugee populations from Southeast Asia living in the United States (Wannamaker 1990). Voodoo deaths are another category that may have ethnic implications (Samuels 1997). Among children, sudden infant death syndrome is a well-recognized phenomenon (Paterson 2010, this book, Chapter 5). Finally, sudden unexpected/unexplained sudden death in epilepsy (SUDEP), which is the subject of this book, is one of the most common causes of death in persons with epilepsy, which, despite increasing interest in unraveling the mystery, still remains an enigma (Lathers, Schraeder, and Bungo 2008b). As mentioned above, the largest group at risk for sudden unexpected death is that of cardiac disease. The overall incidence in this group is 1/738, with a higher incidence of 1/371 in the over age 30 group (Annegers and Blakley 1990). However, epilepsy is likewise associated with a high risk of unexpected death (Tomson et al. 2005). In comparison, having idiopathic epilepsy overall has a 1/1000 yearly incidence risk. However, when one factors in decreasing degrees of seizure control, it is evident that the risk increases, resulting in persons with poorly controlled seizures having an incidence risk as high as 1/100 (Annegers and Blakley 1990). This conclusion was conἀrmed by Nashef et al. (1996). Persons with epilepsy who were seizure free for more than 2 years had a 1/2500 patient-year risk of unexpected death. These investigators found that in patients with poorly controlled active epilepsy, the risk was greater than 1/200 patient-years, and concluded that improved seizure control reduces the risk of unexpected death. Although the risk factors associated with SUDEP, as determined by statistical analysis of population data, are discussed in varying detail throughout this book, it is also important to personalize these risk factors with the clinical histories of individuals who succumbed to SUDEP. Obviously, not all issues are addressed, but the compelling circumstances illustrated in the cases discussed in different parts of the book certainly cover a broad spectrum of circumstances, all of which resulted in the same unfortunate end point of death. 943

944 Sudden Death in Epilepsy: Forensic and Clinical Issues

The causes of unexpected deaths in persons with epilepsy can be broken down arbitrarily into four categories. (1) Epilepsy may be a direct cause of death as exempliἀed by a person dying during uncontrolled status epilepticus; (2) epilepsy as an underlying cause would include circumstances such as drowning during a seizure or dying as the result of a seizure-related motor vehicle accident; (3) deaths in which epilepsy is present, but occurs as a secondary entity associated with a progressively fatal brain disorder such as a malignant tumor or intracerebral hemorrhage; and (4) the mysterious entity of sudden unexplained/ unexpected death in persons with epilepsy in which there is no evidence of any associated cause of death other than the history of epilepsy (Leestma 1990b). SUDEP is deἀned as “sudden, unexpected, witnessed or unwitnessed, nontraumatic and nondrowning death in patients with epilepsy, with or without evidence for a seizure and excluding documented status epilepticus, in which postmortem examination does not reveal a toxicologic or anatomic cause for death” (Nashef 1997). For many years, SUDEP was in many ways a risk associated with having epilepsy that was ignored and denied. One of the original long-term population-based research projects to investigate SUDEP was conducted by Leestma (Leestma 1990a, 1990b; Leestma et al. 1984). This study utilized the Cook County coroner’s office and was one of the ἀrst and the largest study that focused on thoroughly investigating all deaths in persons with a history of epilepsy in the greater Chicago region. The data from this project indicated that SUDEP accounted for at least 10% of epilepsy-related deaths, with an average age at time of death of 30+ years. SUDEP was found to be more common in males, with a male/ female ratio of 3:1. While higher seizure frequency was also a factor, in that over one-half of the victims had 3 to 10 seizures per year, two-ἀfths had as few as one to two seizures per year. The majority of SUDEP events occurred at home during sleep. In instances when a rescue team managed to arrive shortly after the onset of the fatal event, electrocardiograms showed ventricular ἀbrillation. More than 97% of the deaths (36/37) occurred at home, with most (82%) in the prone position. Of the four witnessed deaths, three victims had generalized tonic seizures and, interestingly, one had no observed clinical seizure at the time of death. Postmortem examination in the Cook County population found that in 50% of cases, none of the prescribed antiepileptic drugs were found in blood samples, a ἀnding that was consistent with noncompliance in the SUDEP group. Of interest was the observation that the hearts, livers, and lungs had heavier weights than expected. The cause of cardiomegaly is not known, and it is uncertain how it may relate to the mechanism of SUDEP. Histological examination found, on routine screening, that the heart had no obvious pathology to explain a cause of death. The livers had only passive congestion that was compatible with acute cardiac failure. The lungs revealed both pulmonary edema and congestion, with some cases showing increased protein content in the edema fluid, ἀndings that may be compatible with neurogenic pulmonary edema. Brain pathology was varied but consisted of no lesions that could be used to explain an acute cause of death. Static nonprogressive entities include sequellae of old trauma, stable hydrocephalus, old subdural hematomas, cortical malformations, etc., that were found in the brains were in all likelihood the underlying causes of the epilepsy and themselves were not the cause of death (Leestma et al. 1984, 1989). Comparable autopsy data were found in a population-based study in Denver, CO (Earnest et al. 1992). Subtherapeutic antiepileptic drug levels were evident in 92% of autopsies. Pulmonary congestion occurred in 86% of the lungs, 32% of the brains, and in 32% of the abdominal viscera. Static brain pathology that could be the

SUDEP

945

etiology of epilepsy, but not the direct cause of death, was found in 31% of the brains that were available for study. One of the more interesting histological ἀndings was the observation of interstitial myocardial ἀbrosis in 11% of the hearts, an observation that was conἀrmed by later studies of the hearts of SUDEP victims (Natelson et al. 1998). Based on the above and subsequent data (Nashef et al. 1996; Nashef and Shorvon 1997; Nashef et al. 1996; Nashef 1997), several associated risk factors have come into focus. These are being young (mean age, 32 years), being male (50% greater risk than female), sleeping (most deaths occur during sleep), generalized tonic–clonic epilepsy (usually with poor control, but not always), antiepileptic drug polypharmacy (with frequent dose and drug changes), and underlying static brain pathology that is often associated with mental retardation syndromes. Two other possible risk factors should also be discussed. The ἀrst of these, stress, is known to be a risk factor for sudden death in persons with a history of coronary artery diseases (Ong 2008). Examples of stressful circumstances associated with an increased risk of unexpected sudden death are threat of or actual loss of a person close to the individual, loss of status or self-esteem, and anticipation of or relief from sudden danger. Investigation of stress as a risk factor for SUDEP inexplicably has been mostly ignored. Having epilepsy itself is stressful. For example, there are associated fears of losing control in potentially embarrassing circumstances, social stigmata associated with public ignorance about epilepsy, as well as concerns about losing employment and one’s driver’s license (Lathers and Schraeder 2006). Persons with epilepsy appear to be more socially isolated, less likely to marry, more likely to be under- or unemployed, and are more likely to suffer from anxiety, depression, and poor self-esteem (Ring 1997). The second possible risk factor is the concurrent presence of inherited potentially arrhythmogenic channelopathies in persons with epilepsy. For example, an abnormal sodium channel gene SCN5A is found in Brugada syndrome, which consists of ST elevation, incomplete complete right bundle branch block, mutifocal ventricular tachycardia, and, most disturbingly, a usually normal baseline electrocardiogram (EKG) ἀnding. The same abnormal gene is also associated with the long QT syndrome, resulting in a torsade de pointe ventricular tachycardia. Both of these clinical entities are associated with a family history of cardiac disease (Lathers and Schraeder 2009) (Lathers, Schraeder, and Bungo 2008a). We do not know if the concurrent presence of epilepsy and a potentially arrhythmogenic inherited channelopathy increase the chance of SUDEP. Future studies need to be undertaken to screen family members of SUDEP victims to determine if such a link exists, for if it does, then a valid argument would exist for screening the highest risk group of persons with a history of epilepsy, namely those whose seizures are not under control. Needless to say, we do not yet understand how these two factors will ἀt in with those that are already established and those yet to be discovered. It is likely that SUDEP is associated with interactions of multiple risk factors (Lathers, Schraeder, and Bungo 2008a). While the mechanism of SUDEP is not known, the observation of clinical autonomic dysfunction in association with seizures mitigates in favor of a role for autonomically mediated acute disturbances having a role in explaining the occurrence of SUDEP. For example, Hilz et al. (2002) found cardiac single photon emission computed tomography evidence of decreased norepinethrine uptake in cardiac sympathetic nerves during seizures, and Opherk et al. (2002) described increased ictal heart rate, ST depression, and T wave inversion in association with seizures. However, in contradistinction to the observation of increased heart rate, Rugg-Gunn et al. (2004), using prolonged implanted EKG

946 Sudden Death in Epilepsy: Forensic and Clinical Issues

monitoring, observed ictal bradycardia, with 4/16 subjects requiring pacemaker implantation. Other instances of ictal bradycardia are well documented (Rocamora et al. 2003). Observations of autonomic changes consequent to cortical stimulation in persons with epilepsy have been extant for over half a century, allowing for longstanding speculation about the mechanism(s) of SUDEP. Penἀeld and Jasper (1954), who undertook systematic stimulation of multiple cortical areas in persons undergoing surgery for epilepsy while awake, under local anesthesia, documented a variety of autonomic responses to cortical stimulation. These included the following: observation of salivation with suprasylvian stimulation; gastrointestinal activation with nausea, abdominal discomfort, boborygmi, and a desire to defecate during insula stimulation; apnea with maintained voluntary ability to breath during anterior and inferior cingulate stimulation; irresistible apnea, i.e., inability to overcome apnea with voluntary effort, during rolandic or uncal region stimulation; and abdominal sensations, increased heart rate, flushing, and papillary changes during stimulation of the supplementary motor area. In animals’ stimulus of the amygdala,€Bonvallet and Bobo (1972) found neuronal groupings that indicated that cardiac and respiratory control were intertwined. These investigators demonstrated, for the ἀrst time, that within the amygdala, stimulation of one grouping of neurons resulted in bradycardia and respiratory deactivation, while stimulation of a different, nearby group resulted in tachycardia and respiratory activation. The complex interconnections between various cortical and subcortical structures, including the amygdala, hypothalamus, and cardiovascular/respiratory control centers in the brainstem, would give credence to the possibility of seizure-related potentially fatal cardiovascular and/or respiratory disruption contributing to the risk of SUDEP (Stollberger and Finsterer 2004). Ictal tachycardia is commonly observed during generalized tonic–clonic seizures. Marshall et al. (1983) found frequent episodes of tachycardia and premature ventricular contractions in association with complex partial seizures. Nei et al. (2004) reported increased ictal heart rates in persons who subsequently were victims of SUDEP, and suggested that potential SUDEP victims manifested increased autonomic activity with repolarization and rhythm abnormalities, especially during sleep. Ictal-related bradyarrhythmias seem to be associated most often with insular, cingulate, pyriform cortex, or amygdala activation. Of interest is the observation of associated episodes of apnea (Penἀeld and Jasper 1954) (Leung, Kwan, and Elger 2006). The association of cerebral discharges and disturbances of the normal balance between cardiac autonomic neural discharges, as illustrated by the lockstep phenomenon (Lathers, Schraeder, and Weiner 1987) (Lathers and Schraeder, 2010b, Chapter 28, this book), provides evidence that there is a direct relationship between ictal, and even interictal discharges, and disturbance of cardiac sympathetic and parasympathetic neural discharges that are involved in controlling normal cardiac conduction and rhythm. While the potential for adverse autonomic effects were alluded to above and has been little investigated since the study of Lathers and Schraeder (1990), Mameli et al. (1988) found in animal experiments that interictal discharges resulted in sinus arrest, supraventricular extra systoles, bradycardia, junctional rhythms, and associated drop in blood pressure. They also observed a very short latency between epileptiform discharges and the cardiovascular changes, implying that these were indeed neurogenic, not neurohumoral, events, and that the changes persisted beyond the duration of the interictal events. Respiratory factors should also be considered. It is well known that victims of SUDEP have increased lung weights compared to control (Leestma 1990c; Leestma et al. 1984), ἀndings that support the possible mechanistic role of seizure-induced or related neurogenic

SUDEP

947

pulmonary edema in SUDEP. While these human postmortem pulmonary ἀndings are supported by their occurrence in animal models of epilepsy, there is also evidence that seizure-related central apnea can occur concurrently or independently of the acute pulmonary changes (Johnston et al. 1995, 1997). The occurrence of central apnea in these series of experiments supports the data of Penἀeld and Jasper (1954) and Bonvallet and Bobo (1972), as discussed earlier in this chapter (vide supra). However, determination of whether any seizure-related observations are concomitant occurrences, potential risk factors, or potential mechanisms requires more clinical and animal-based investigation. Investigation of the potential role of psychological factors in association with risk of SUDEP has been for the most part ignored. This general lack of interest is in marked contrast to the longstanding acknowledgment of the association of psychological factors with risk for sudden death in persons with cardiac disease (Lathers and Schraeder 2006) (Lathers and Schraeder 2010a, Chapter 17, this book). The investigational scotoma related to the question of what role adverse psychological circumstances may have in SUDEP is difficult to explain. For example, it is well known that persons with epilepsy are more likely to be depressed than persons with other disorders (Mendez, Cummings, and Benson 1986). As anxiety and depression often occur concurrently, it is important to consider the reality of the negative societal issues that can be an ongoing stimulus for persons with epilepsy to be more likely to experience these symptoms than persons with other chronic disorders. The societal stigmata associated with having epilepsy impose a persisting impact on persons with epilepsy. Any neurologist committed to treating persons with epilepsy is well aware that these stigmata can result in discrimination in obtaining and keeping a job, the fear of embarrassment associated with having a seizure in public, and impaired mobility when not allowed to drive. Persons with epilepsy also face having associated psychological issues, such as depression (Mendez, Cummings, and Benson 1986), which may be aggravated by antiepileptic medication. In addition, another common source of anxiety in young parents is concern about the potential inheritability of epilepsy and the risks of maternal antiepileptic medication use during pregnancy. That psychological stress can have a potentially fatal adverse effect on some individuals is illustrated by the case of a 22-year-old male college student who came to the office because he had loss of consciousness associated with tonic–clonic seizures while attending church. At the time he lost consciousness, he was listening to a passage being read by his minister from Foxe’s Book of Martyrs, a book that described the sufferings of persons being persecuted and tortured because of their religious beliefs. The young man’s sister also described that he had lost consciousness in the past while listening to a friend describe to him the details of a recent hospitalization for major surgery. An electroencephalogram (EEG) was obtained, during which, unbeknownst to the patient, the minister arrived and recited the previously offending passage. At that point, the patient’s heart slowed, then ceased to beat. After 15 seconds of asystole, the EEG became isoelectric and generalized tonic–clonic motor activity occurred, after which the heart returned to a normal sinus rhythm and the EEG to alpha rhythm. Eventually, he had a pacemaker implanted and had no further events (Schraeder, Pontzer, and Engel 1983). Although this young man had been observed having generalized tonic–clonic seizure-like episodes, he did not have epilepsy. His sequence of events demonstrated how adverse emotional responses can adversely affect the autonomic regulation of cardiovascular homeostasis, resulting in potentially life-threatening changes. This case emphasizes the need for investigation of the relationship between stress and sudden death.

948 Sudden Death in Epilepsy: Forensic and Clinical Issues

Autonomic/homeostatic dysfunctions such as ictally related cardiac tachy- and bradyarrhythmias, pulmonary edema, and central apnea have each been long discussed as possible mechanisms of SUDEP. In the future, is seems quite possible that psychological stress may also need to addressed. Each potential mechanism has a following. It is possible to speculate that several combinations of factors that lead to SUDEP may be operant. These include (1) the occurrence of seizure-related central apnea plus neurogenic pulmonary edema leading to a respiratory death; (2) seizure-related hypoxia plus acute neurogenic pulmonary changes combined with acidosis resulting in a fatal arrhythmia; (3) the presence of an inherited channelopathy aggravated by an acute seizure-related autonomic dysfunction, resulting in a fatal arrhythmia; and (4) psychological stress adding to potentially fatal combinations of events, tipping the balance toward SUDEP in susceptible individuals. Future research needs to address a broad range of investigations. (1) At a clinical level, multicenter, prospective population-based autopsy studies will be the only method to determine the proἀle of persons at risk for SUDEP (Schraeder et al. 2006, 2009, 2010, Chapter 6, this book). Such studies should include detailed interviews of witnesses, family members, and health care providers (Lathers and Schraeder 2009), emphasizing a search for environmental, medical, and psychological risk-related data as well as detailed microscopic examination of the somatic and autonomic nervous system tissue and the heart (Lathers and Schraeder 2009). Additional investigation addressing the possible role of disturbed serotonin metabolism may also be important in the future (Paterson 2010). (2) There is also a need to determine whether a potential link exists between the subtle, genetic cardiac channelopathies and risk of SUDEP. Investigation should include detailed cardiac history, EKGs, and genetic screening for channelopathy genes in surviving family members. (3) Animal models need to be designed and used to investigate whether there is concomitance between seizure-related cardiovascular autonomic/respiratory dysfunction, induced stress, and sudden death. (4) Finally, there needs to be an investigation into possible preventive interventions such as sleeping position, prophylactic use of antiarrhythmic drugs, approaches to improving patient compliance with prescribed antiepileptic drugs, and stress management protocol for persons with epilepsy. As in most prevention efforts, while an absolute abolition of the occurrence of SUDEP would be desirable, it is unlikely. A measurable decrease in its risk of occurrence should be an achievable initial goal. The educational need around the issue of SUDEP should also to be addressed. A national survey of coroners and medical examiners was conducted to determine the extent of postmortem investigation in persons who died with a history of epilepsy and to ascertain how these officials approached the issue of SUDEP (Schraeder et al. 2006, 2009). This survey documented that while the overall quality of postmortem examination was sufἀciently thorough to determine the cause of death, in cases of epilepsy without a determinable cause, there seemed to be a dichotomous and paradoxical appreciation of SUDEP among those surveyed. The survey data found that while there was a general acknowledgment of the existence of SUDEP as a diagnosis, most of the officials acknowledged having reluctance to use it as a diagnosis in appropriate cases. This intellectual disconnect, even in medical examiners trained as forensic pathologists, impedes the collection of accurate national data on the prevalence of SUDEP. Relative to the education of patients with epilepsy and their families, in general, there is no consensus of how to approach the question of what patients and their families should know about the risk of SUDEP (Beran 2006). Clinical practice experience indicates that

SUDEP

949

there seems to be three approaches. The ἀrst is to say nothing, which is a variation of the “don’t ask, don’t tell” approach when dealing with uncomfortable issues; the second is to talk about SUDEP only if asked; the third is to tell those who are at increased risk, e.g., young males, those with poorly controlled generalized tonic–clonic seizures, persons tÂ�aking multiple antiepileptic drugs, especially when changes in medication is being contemplated, and those with localization related generalized tonic–clonic seizures; and ἀnally telling all patients. No one knows what the best approach is, but the elimination of the “don’t ask, don’t tell” approach should be the ἀrst effort. In conclusion, although the mechanism and prevention of SUDEP continue to be a clinical conundrum that will require future research to resolve, there is no justiἀcation for maintaining patients who are at risk for SUDEP and their families in a state of ignorance about its existence and the currently acknowledged risk factors.

References Annegers, J. F., and S. A. Blakley. 1990. Patterns of overall and unexplained death mortality among persons with epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Beran, R. G. 2006. SUDEP—To discuss or not discuss: that is the question. Lancet Neurol 5 (6): 464–465. Bonvallet, M., and E. G. Bobo. 1972. Changes in phrenic activity and heart rate elicited by localized stimulation of amygdala and adjacent structures. Electroencephalogr Clin Neurophysiol 32 (1): 1–16. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33 (2): 310–316. Hilz, M. J., O. Devinsky, W. Doyle, A. Mauerer, and M. Dutsch. 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125 (Pt 5): 985–995. Johnston, S. C., J. K. Horn, J. Valente, and R. P. Simon. 1995. The role of hypoventilation in a sheep model of epileptic sudden death. Ann Neurol 37 (4): 531–537. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42 (4): 588–594. Kuller, L., A. Lilienfeld, and R. Fisher. 1967. Am epidemiological study of sudden an unexpected deaths in adults. Medicine 46: 341–361. Lathers, C. M., and P. L. Schraeder. 1990. Chapter 12. Synchronized cardiac neural discharge and epileptogenic activity, the lock-step phenomenon: Lack of correlation with cardiac arrhythmias. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies. Epilepsy Behav 14: 573–576. Lathers, C. M., and P. L. Schraeder. 2010a. Stress and SUDEP (Chapter 17). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 2010b. Animal model for sudden unexpected death in persons with epilepsy (Chapter 28). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008a. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Chapter 13. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher. Hauppauge, NY: Nova Science.

950 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Leestma, J. E. 1990a. Sudden unexpected death associated with seizures: A pathological review. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E. 1990b. Natural history of epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E. 1990c. Sudden unexpected death associated with seizures: A pathological review (Chapter 5). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Leestma, J. E., M. B. Kalelkar, S. S. Teas, G. W. Jay, and J. R. Hughes. 1984. Sudden unexpected death associated with seizures: Analysis of 66 cases. Epilepsia 25: 84–88. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26 (2): 195–203. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9 (1): 19–30. Mameli, P., O. Mameli, E. Tolu, G. Padua, D. Giraudi, M. A. Caria, and F. Melis. 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29 (1): 74–82. Marshall, D. W., B. F. Westmoreland, and F. W. Sharbrough. 1983. Ictal tachycardia during temporal lobe seizures. Mayo Clin Proc 58 (7): 443–446. Mendez, M. F., J. L. Cummings, and D. F. Benson. 1986. Depression in epilepsy. Signiἀcance and phenomenology. Arch Neurol 43 (8): 766–770. Nashef, L. 1997. Sudden unexpected death in epilepsy: Terminology and deἀnitions. Epilepsia 38: S6–S8. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300. Nashef, L., and S. D. Shorvon. 1997. Mortality in epilepsy. Epilepsia 38 (10): 1059–1061. Natelson, B. H., R. V. Suarez, C. F. Terrence, and R. Turizo. 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch Neurol 55 (6): 857–860. Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. Ong, L. 2008. Negative affective disorders and arrhythmogenesis. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher. Hauppauge, NY: Nova Science. Opherk, C., J. Coromilas, and L. J. Hirsch. 2002. Heart rate and EKG changes in 102 seizures: analysis of influencing factors. Epilepsy Res 52 (2): 117–127. Paterson, D. S. 2010. Medullary serotonergic abnormalities in sudden infant death syndrome: Implications in SUDEP. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Penἀeld, W., and H. Jasper. 1954. Summary of clinical analysis and seizure patterns. In Epilepsy and the Functional Anatomy of the Human Brain, ed. W. Penἀeld and H. Jasper. Boston, MA: Little Brown. Ring, A. R. 1997. Other psychiatric illnesses. In Epilepsy: A Comprehensive Textbook, ed. J. J. Engel and T. A. Pedley. Philadelphia, PA: Lippincott-Raven. Rocamora, R., M. Kurthen, L. Lickfett, J. Von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44 (2): 179–185. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Samuels, M. 1997. Voodoo death revisited: The modern lessons of neurocardiology. The Neurologist 3: 293–304.

SUDEP

951

Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Schraeder, P. L., R. Pontzer, and T. R. Engel. 1983. A case of being scared to death. Arch Intern Med 143 (9): 1793–1794. Schraeder, P. L., E. L. So, and C. M. Lathers. 2010. Forensic case identiἀcation (Chapter 6). In Epilepsy in Sudden Death: Forensic and Clinical Issues. Boca Raton, FL: CRC Press. Stollberger, C., and J. Finsterer. 2004. Cardiorespiratory ἀndings in sudden unexplained/unexpected death in epilepsy (SUDEP). Epilepsy Res 59 (1): 51–60. Tomson, T., T. Walczak, M. Sillanpaa, and J. W. Sander. 2005. Sudden unexpected death in epilepsy: A review of incidence and risk factors. Epilepsia 46 (Suppl 11): 54–61. Wannamaker, B. B. 1990. Perspectives on death of persons with epilepsy. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York: Marcel Dekker. Wendkos, M. H. 1979. Sudden Death and Psychiatric Illness. New York, NY: SP Medical and Scientiἀc Books.

Epilepsy and SUDEP Lessons Learned: Scientific and Clinical Experience

60

Claire M. Lathers Paul L. Schraeder

Contents 60.1 Proposed Areas of Inquiry about SUDEP [106] References

954 961

Most likely, different mechanisms and/or different combinations of mechanisms are responsible for death in persons with epilepsy. Contributing mechanisms at peripheral sites include, in part, a direct action on the heart, indirect and/or direct actions on the lungs, and actions on the adrenal glands. In the discussion of our papers published over the years, we listed the following mechanisms thought to be contributing to sudden unexpected death in epilepsy (SUDEP) (Table 60.1). Based on their ἀndings in the above animal models, Lathers and Schraeder emphasized the importance of continuous electrocardiography (EKG) and electroencephalography (EEG) recordings in identifying persons with epilepsy at risk for sudden death. Despite this three-decade effort, discussion at the recent (November 2008) National Institutes of Health (NIH) SUDEP Workshop (NINCDS 2008) revealed that many physicians still do not concurrently examine heart and brain electrical activities when evaluating persons at risk for sudden death. It will take a global effort of multidisciplinary research collaboration among physicians and basic scientists to evaluate the importance of simultaneously monitoring sophisticated EKG and EEG activity in persons thought to be at risk for sudden death. Integration of neurology, cardiology, and clinical pharmacology is needed for the best diagnostic and treatment interventions for persons at risk for SUDEP. Lathers and Schraeder (1990) edited the ἀrst book on SUDEP, Epilepsy and Sudden Death, which summarized the “state of the art” experimental and clinical information available about sudden death and epilepsy. Today it still provides a guide for how basic researchers and clinicians should approach detection, understanding, and prevention of SUDEP. J. Thomas Bigger Jr., MD, Professor Medicine and Pharmacology, College of Physicians and Surgeons at Columbia University, succinctly summarized in the Foreword eight areas of inquiry that needed to be addressed to answer questions about SUDEP. The following is published with permission (Lathers 2009).

953

954 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 60.1â•… Summary: Mechanisms/Sites Associated with SUDEP 1. Nonuniform autonomic postganglionic cardiac sympathetic neural discharge related to the nonuniform beta sympathetic receptor locations in the heart just before development of arrhythmias associated with interictal and ictal activity (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers, Schraeder, and Carnel 1984; Carnel, Schraeder, and Lathers 1985; Lathers et al. 1986; Lathers and Spivey 1990; Lathers, Levin, and Spivey 1986; Lathers, Spivey, and Levin 1988). 2. Autonomic imbalance between sympathetic and parasympathetic neural discharges and variability within each autonomic division are associated with varying degrees of epileptogenic activity and can be associated with ventricular ἀbrillation or asystole (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers, Schraeder, and Carnel 1984; Carnel, Schraeder, and Lathers 1985). 3. Autonomic parameter dysfunction of heart rate and blood pressure occurs before development of interictal discharges and continues with ictal discharges (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers, Schraeder, and Carnel 1984; Carnel, Schraeder, and Lathers 1985; Lathers, Schraeder, and Weiner 1987; Stauffer, Dodd-O, and Lathers 1989, 1990; Dodd-O and Lathers 1990; O’Rourke and Lathers 1990; Lathers and Schraeder 1990). 4. Autonomic cardiac neural discharges are intermittently synchronized 1:1 with the epileptogenic discharge, i.e., the lockstep phenomenon (LSP) (Lathers, Schraeder, and Weiner 1987; Stauffer, Dodd-O, and Lathers 1989, 1990; Dodd-O and Lathers 1990; O’Rourke and Lathers 1990; Lathers and Schraeder 1990). 5. Epileptogenic activity may alter autonomic central (Chadwick, Jenner, and Reynolds 1975; Oishi, Suenaga, and Fukuda 1979; Mason and Corcoran 1979) or peripheral autonomic release of catecholamines (Ceremuzynski, Staszewska-Barczak, and Herbaczynska-Cedro 1969; Kelliher, Widmer, and Roberts 1975). 6. Multiple areas of pulmonary punctuate hemorrhages and gross hemorrhage and edema were found in animals dying after epileptogenic-activity-induced asystole or ventricular ἀbrillation (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers, Schraeder, and Carnel 1984). 7. Tissue hypoxia, hypercarbia, and alterations in acid–base balance may contribute to the results in experimental epilepsy (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers, Schraeder, and Carnel 1984; Carnel, Schraeder, and Lathers 1985). 8. Changes in cardiac function alter cerebral blood flow and may produce central hypoxia, resulting to epileptogenic activity (Schott, McLeod, and Jewitt 1977; Schraeder, Pontzer, and Engel 1983; Schwartz and Lathers 1990). 9. Modulation of the major central nervous system presynaptic inhibitory neurotransmitter gamma aminobutyric acid release by prostaglandin E2 may be an explanation for epileptogenic–activity-related dysfunction of autonomic cardiac neural discharge leading to arrhythmias (Suter and Lathers 1984; Schwartz and Lathers 1990). Mechanisms for interference of GABA neurotransmission may lead to initiation of arrhythmias and/or epileptogenic activity (Schwartz 1988; Rastogi and Ticku 1986) and sudden death (Suter and Lathers 1984; Killam and Bain 1957). 10. Beta blockers exhibited anticonvulsant activity, whether administered via the intraosseous route or intravenously (Lathers, Schraeder, and Bungo 2008a; Jim et al. 1988; Lathers and Jim 1990). Source: Lathers, C. M., Epilepsy Behav, 15, 269–277, 2009. With permission.

60.1â•… Proposed Areas of Inquiry about SUDEP [106] Dr. Thomas Bigger has stated in the preface of Epilepsy and Sudden Death (Lathers and Schraeder 1990) that “Although many pieces of the epilepsy–sudden death puzzle are in place, the total picture is not clear. . . . We seem a long way from understanding and controlling the problem of sudden death in epilepsy. How shall we proceed toward those goals?” The following list summarizes his view of what research needed to be done to investigate the mechanisms and prevention of SUDEP.

Epilepsy and SUDEP

955

1. Additional epidemiologic studies to sharpen the focus on the high-risk group. 2. Studies with ambulatory EEG, EKG, and respiratory recordings in high-risk individuals to capture the events before and during sudden death. 3. The role of the sympathetic nervous system could be explored more intensively with functional and biochemical evaluation of cardiac sympathetic nervous activity. 4. The anatomic distribution of cardiac sympathetic nerves in persons with epilepsy vs. controls could be clariἀed by imaging studies using biological markers taken up by sympathetic nerve terminals in the heart. 5. Studies in relevant animal models will permit a forward leap in our capability for generating relevant new knowledge and greatly accelerate the rate of progress in clarifying the pathophysiology of SUDEP. 6. Hypotheses based on animal data about effective treatments will naturally follow and can be pursued using clinical studies in persons with epilepsy. 7. Studies of the factors that govern compliance in epileptic patients should proceed now because the results could be applied immediately to maximizing control of seizure disorders and thereby decreasing the risk of SUDEP. We are on the threshold of major advances in the problem of sudden death in epilepsy. The tools to advance our knowledge are at hand and we should energetically put them to use for the future beneἀt of patients with epilepsy. Knowledge gained from studies of sudden death in epileptics will very likely be useful for understanding sudden death in other situations as well. (Bigger 1990)

Only some of Dr. Bigger’s suggested areas of investigation have now been addressed. 1. Additional epidemiology studies have been conducted (Earnest et al. 1992; Ficker et al. 1998; Walczak et al. 2001; Moran et al. 2004; Rugg-Gunn et al. 2004). Also see reviews by Lathers et al. (Lathers, Schraeder, and Bungo 2008a, 2008b; Scorza, Arida, and Cavalheiro 2008; Lathers, Schraeder, and Bungo 2010; Bacon 1868; Lathers and Schraeder 1990). 2. Studies of ambulatory recordings to capture recordings before and during sudden death are still needed today. Recordings of EEG, EKG, and respiration must still be done. Simultaneous recordings of these variables, and arterial oxygen saturation, during sudden death would permit us to evaluate the role of apnea or hypoxia. Postictal central apnea appears to be one potential mechanism for SUDEP. A 55-s convulsive seizure occurred in a 20-year-old woman as she underwent videoEEG monitoring (So, Sam, and Lagerlund 2000). Persistent apnea then developed. Electrocardiogram-monitored rhythm was not altered for the ἀrst 10 s, then it gradually and progressively slowed and stopped 57 s later. Cardiorespiratory resuscitation was successful. No evidence of airway obstruction or pulmonary edema was noted. One previous cardiorespiratory arrest after a complex partial seizure without secondary generalization had been reported for this patient. So et al. (2000) note that although epileptic seizures may be associated with arrhythmogenic actions at the heart, in this patient the mechanism of marked central suppression of respiratory activity after seizures was clearly involved and almost resulted in sudden death. The timing of events such as seizures, respiratory and/or laryngospasm, and cardiac EKG changes varies in different patients. The physician

956 Sudden Death in Epilepsy: Forensic and Clinical Issues

must consider risk factors for a given patient to recommend protective procedures to minimize the chance of unwanted events that may result in SUDEP. The question must be asked as to whether a person ἀrst experienced seizures and respiratory events and then cardiac events or if the person experienced seizures and arrhythmia and then respiratory events. There is documented evidence in the literature to support both cardiac and respiratory events as initiating mechanisms of sudden death. Obviously, rapid reversal of these changes is essential and the “availability of resuscitation methods on the spot where the victim is located” certainly increases the likelihood that SUDEP will be prevented. In a recent study, Bateman et al. (2008) examined the incidence and severity of ictal hypoxemia in patients with localization-related epilepsy undergoing video-EEG telemetry. They measured seizure-associated oxygen desaturation and hypoventilation. Pulse oximetry revealed oxygen desaturations below 90% in one-third of all 304 seizure events. The degree of desaturation was signiἀcantly correlated with seizure duration and with electrographic evidence of seizure spread to the contralateral hemisphere. Central apneas or hypopneas occurred with 50% of all seizures. Ictal hypoxemia occurred often in these patients with localization-related epilepsy, and may be pronounced and prolonged, even if the seizures do not progress to generalized convulsions. End tidal carbon dioxide increase occurred in oxygen desaturation and supports the assumption that ictal oxygen desaturation is a consequence of hypoventilation. Both ictal hypoxemia and hypercapnia may be contributing risk factors to SUDEP occurrence. 3. The role of the sympathetic nervous system could be explored more intensively with functional and biochemical evaluation of cardiac sympathetic nervous activity. Lathers and Levin (2010) have examined the distribution of cardiac beta receptors and found a correlation with the release of norepinephrine at sympathetic nerve terminals in the heart concurrent with the production of arrhythmia. Innervation density is high in the subepicardium and the central conduction system. In diseased hearts, cardiac innervation density varies. This may lead to sudden cardiac death. After myocardial infarction, sympathetic denervation is followed by reinnervation within the heart, leading to unbalanced neural activation and lethal arrhythmia (Lathers et al. 1986; Lathers and Spivey 1990; Lathers, Levin, and Spivey 1986; Lathers, Spivey, and Levin 1988). Recently, Ieda et al. (2008), as in the earlier studies by Lathers et al. (Lathers et al. 1977; Lathers, Roberts, and Kelliher 1977; Lathers et al. 1978; Lathers, 1980b, 1981, 1982; Lathers and Schraeder 1982; Schraeder and Lathers 1983), have raised the question of whether regulation of cardiac nerves is a “new paradigm” in the management of sudden cardiac death since the heart is extensively innervated and its performance is regulated by the autonomic nervous system. In the case of diabetic sensory neuropathy, silent myocardial ischemia may occur, associated with loss of pain perception during myocardial ischemia, a major cause of sudden cardiac death in diabetes mellitus (Ieda et al. 2006). To date, molecular mechanisms underlying innervation density are not well understood. Ieda et al. (2006) have demonstrated that cardiac sympathetic innervation is determined by the balance of neural chemoattraction and cheomorepulsion, both of which occur in the heart. Nerve growth factor, a potent chemoattractant, is synthesized by cardiomyoctyes and is induced by endothelin-l upregulation in the heart. In contrast, Sema3a, a neural chemorepellent, is expressed strongly in

Epilepsy and SUDEP

957

the trabecular layer in early-stage embyros and, at a lower level after birth, cause epipcardial-to-endocardial transmural sympathetic innervation pattern. Cardiac nerve growth factor downregulation is a cause of diabetic neuropathy, and nerve growth factor supplementation rescues silent myocardial ischemia in diabetic neuropathy. Both Sema3a-deἀcient and Sema3a-overexpressing mice showed sudden death or lethal arrhythmias due to disruption of innervation patterning (Ieda et al. 2007). All of these regulatory mechanisms involved in neural development in the heart and their critical roles in cardiac performance need to be examined to determine relevance to methods to decrease the risk of SUDEP. 4. The issue of using imaging studies to examine the anatomic distribution of cardiac sympathetic nerves in epileptics vs. normals has been answered recently. A postmortem imaging study, using 123I-metaiodobenzylguanidine single photon emission computed tomography (SPECT), has been conducted of postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy (Druschky et al. 2001), ἀnding sympathetic dysfunction in the form of altered postganglionic cardiac sympathetic innervation in patients with chronic temporal lobe epilepsy. These observations suggested that the altered postganglionic cardiac sympathetic innervation may increase risk of cardiac abnormalities and/or SUDEP. The exact role of innervation in arrhythmogenesis and developmental and regulatory mechanisms determining density and pattern of cardiac sympathetic innervation are still unclear. This clinical study of Druschky et al. (2001), conducted in humans, conἀrms the results and conclusions of the animal studies conducted by Lathers et al. (Lathers et al. 1977; Lathers, Roberts, and Kelliher 1977; Lathers et al. 1978; Lathers 1980a, 1981, 1982) in which the relationship of postganglionic cardiac sympathetic neural discharge was associated with arrhythmias and/or sudden death. Kerling et al. (2009) stated that since tachyarrhythmias are common during epileptic seizures while bradyarrhythmias or asystoles occur less frequently, they evaluated cardiac postganglionic denervation in patients with epilepsy to evaluate ictal asystole. SPECT examined 123I-meta-iodobenzylguanidine as a marker of postganglionic cardiac norepinephrine uptake. The pronounced reduction in cardiac SPECT uptake in asystolic patients indicated postganglionic cardiac catecholamine disturbance. Impaired sympathetic cardiac innervation limits adjustment and modulation of heart rate and may increase the risk of asystolic events and, eventually, sudden unexpected death in persons with epilepsy. The data of Kerling et al. (2009) support the ἀndings of Lathers, Levin, and Spivey (Lathers et al. 1986; Lathers and Spivey 1990; Lathers, Levin, and Spivey 1986; Lathers, Spivey, and Tumer 1988) and those of Han and Moe (1964). 5. Studies in relevant animal models. Many different relevant animal models for SUDEP are still needed to understand the pathophysiology of sudden death in epileptics and to hypothesize about effective treatments (Bigger 1990). The importance of using many different animal models to study SUDEP to glean an insight into the various mechanisms of risks and their contribution to the initiation of the death event is discussed in detail by Lathers (2010). Schwartz et al. (1995) concluded that delayed enhancement of GABAergic neurotransmission directly at the site of vulnerability after an ischemic event protects the neurons from death. This ἀnding

958 Sudden Death in Epilepsy: Forensic and Clinical Issues

should be explored to further study the effect of diazepam on GABA-mediated effects that may prevent ischemia-induced neuronal death and ultimately prevent the worsening of central neuronal communication due to epileptogenic activity. This may eventually contribute a protective central nervous system effect to make an individual less likely to be at risk for SUDEP. There are many more basic science questions to be raised and answered. The study of Wang et al. (2006), discussed above, must be expanded to examine antiepileptic or other categories of drugs and their effect on the intermittent seizure-like ἀring of cardiac parasympathetic neurons that may cause neurogenic ictal bradyarrhythmias, cardiac asystole, or sudden death. So (2008) emphasized the signiἀcance of using audiogenic seizure mice to study postictal respiratory arrest. Postictal respiratory arrest was induced by serotonin receptor inhibition and prevented by selective serotonin reuptake inhibitor drugs. The role of serotonin in SUDEP must be examined in future animal studies (Tupal and Faingold 2006). The reader is referred to Dr. Paterson’s chapter in this book. He has presented a very thorough discussion of medullary serotonergic low levels found in sudden infant death syndrome and the implications for SUDEP. This ἀnding of low brainstem serotonin levels raises the possibility of doing similar functional magnetic resonance imaging studies in adults with epilepsy vs. controls. 6. Hypotheses about effective treatments will naturally follow and can be pursued using animal models, small pilot studies in epileptic patients, and, ἀnally, largescale clinical trials (Bigger 1990). We have not yet fully addressed this area of inquiry about SUDEP. There are several questions to be asked postmortem (Schraeder et al. 2006; Schraeder et al. 2009; Lathers and Schraeder 2009) about the patient who is on an antiepileptic drug but still becomes a SUDEP victim. A very important clinical pharmacology question that must be asked is whether the patient was on the correct antiepileptic drug to control his particular type/ mixture of seizures. A second question to be asked is whether the correct categories of drugs have been prescribed. There may be a role for using beta blocking agents. When evaluating the role of drugs as protectors of life, clinical pharmacologists (Lathers and Schraeder 2002) caution us to remember that use of all drugs is a risk/beneἀt ratio evaluation (Lathers and Schraeder 2002). Thus, the use of antiepileptic drugs may not 100% protect the patient against sudden death. Celiker at al. (2009) report clinical experience of patients with catecholaminergic polymorphic ventricular tachycardia and concluded medical treatment with propranolol and verapamil may decrease the incidence of arrhythmia. If a person is still refractory, implantation of intracardiac deἀbrillators should be considered. Houle et al. (2001) reported enhanced in vivo and in vitro contractile responses to beta2-adrenergic receptor stimulation in those susceptible to lethal arrhythmias. Billman et al. (1997, 2006) found that beta2-adrenergic receptor antagonist protected against ventricular ἀbrillation, and endurance exercise training attenuates cardiac beta2-adrenoceptor responsiveness and prevents ventricular ἀbrillation in animals. There seems to be a direct correlation between beta adrenergic receptor sensitivity and the level of arrhythmia and mortality following a coronary occlusion. The greater the sensitivity to beta adrenergic agonists, whether from hormones, exercise, or genetics, the greater is the level of arrhythmia and mortality (Houle, Altschuld, and Billman 2001; Billman et al. 1997, 2006; Du

Epilepsy and SUDEP



959

et al. 2000). Whether we can learn from these data when considering persons with epilepsy at risk for SUDEP remains to be seen. Data suggest that beta blockers exert a protective effect against seizure induction and/or the development of cardiac arrhythmias with interictal and ictal activity (Lathers, Schraeder, and Bungo 2008a, 2008b; Scorza, Arida, and Cavalheiro 2008; Lathers, Schraeder, and Bungo 2010; Jim et al. 1988; Lathers and Jim 1990; Lathers and Schraeder 1990; Lathers et al. 1990a, 1990b). The question must be asked if persons thought to be at risk for SUDEP should be placed on a beta blocker in addition to the prescribed anticonvulsant(s). Beta blockers are also used to reduce stress, and persons with epilepsy generally are stressed by the disease and associated problems (Lathers and Schraeder 2006). 7. Studies of the factors that govern compliance in epileptic patients should proceed now (Bigger 1990). We have not yet fully addressed this area of inquiry relative to prevention of€ SUDEP. There are several questions to be asked postmortem (Schraeder et al. 2006, 2009; Lathers and Schraeder 2009) about the patient who is on an antiepileptic drug but still becomes a SUDEP victim. To avoid SUDEP, one must inquire family and friends as to whether the person was compliant (Lathers and Schraeder 2002). In 2003, Lathers et al. (Lathers, Koehler, Wecht, and Schraeder 2003a, 2003b) found low or no levels of antiepileptic drugs postmortem in persons with epilepsy deemed to have died of SUDEP. These data suggested that compliance is a problem in some postmortem studies of victims of SUDEP. In 2009, Hughes (2009) deemed the most important SUDEP risk factor to be nonÂ�compliance with antiepileptic medication. Ryvlin et al. (2009) found the risk of SUDEP is increased in patients who have poor compliance and exhibit nocturnal seizures and generalized tonic–clonic seizures. Although there are some questions about the reliability of postmortem antiepileptic drug levels (Tomson et al. 1998), it seems reasonable to state that non-compliance is an issue if very low or no drug levels are found at autopsy. However, compliance is not the only question to ask about victims of SUDEP. One must also ask if the correct dose of the antiepileptic drug was being used for a given patient to control his/her seizures. All agree that maintenance of an optimal therapeutic drug level for a given individual is crucial to avoid SUDEP.

Scientists, clinicians, and granting authorities must not allow another 20 years to go by with little additional data gleaned relative to lowering the risk of SUDEP (Lathers 2009). Future steps have been suggested by The American Epilepsy Society and the Epilepsy Foundation Joint Task Force on SUDEP (So et al. 2009). During the interim, before the answers have been obtained, it is most important to provide prompt and optimal control of seizures, especially generalized convulsive seizures, to prevent the occurrence of SUDEP. Assessment of the current knowledge about SUDEP concludes: 1. A need for multidisciplinary workshops to reἀne current lines of investigation and identify additional areas of research for mechanisms underlying SUDEP. 2. To conduct a survey of patients, families, and caregivers to identify effective means of education that will enhance participation in SUDEP research. 3. A campaign to emphasize the need for complete autopsy examinations for patients with suspected SUDEP.

960 Sudden Death in Epilepsy: Forensic and Clinical Issues Table 60.2â•… Summary Lessons Learned—Epilepsy and SUDEP: Global Focus Needed 1. Mechanistic risk factors for SUDEP, obtained in animal studies in our laboratory, included cardiac arrhythÂ�Â�mias and/or death associated with changes in autonomic cardiac postganglionic sympathetic neural discharge, cardiac parasympathetic neural discharge, and respiratory changes including multiple areas of punctuate hemorrhages and large areas of gross pulmonary hemorrhage and pulmonary edema (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers, Schraeder, and Carnel 1984; Carnel, Schraeder, and Lathers 1985). Both central and peripheral sites are involved in the pathophysiology of SUDEP (Table 60.1). 2. Many different relevant animal models for SUDEP are still needed (Lathers 2010) to understand the pathophysiology of sudden death in epileptics, hypothesize about effective treatments, develop pilot studies in persons with epilepsy, and, ἀnally, conduct conἀrmatory large-scale clinical trials. 3. “Think outside of the box” when evaluating an established animal model with potential for modiἀcaÂ� tion(s) to be used to address questions about the mechanism(s) of SUDEP. 4. The ἀeld of pharmacology/clinical pharmacology has much to offer as we work to improve compliance and to develop new antiepileptic drugs and/or apply new categories of drugs, such as beta blockers or selective serotonin reuptake inhibitors, to prevent and resolve the mystery of SUDEP (Lathers, Schraeder, and Bungo 2008a, 2008b; Scorza, Arida, and Cavalheiro 2008; Lathers, Schraeder, and Bungo 2010; Jim et al. 1988; Lathers et al. 1989; Lathers, Jim, and Spivey 1989; Jim et al. 1989; Lathers and Jim 1990; Lathers and Schraeder 1990; Lathers and Schraeder 2002). 5. Team work is needed among different multidisciplinary professionals working in the clinical settings and/or within a laboratory, among laboratories within the United States, and in laboratories located around the world to solve the global mystery of SUDEP. 6. Ambulatory simultaneous EKG and EEG telemetry monitoring of patients thought to be at risk for sudden death will help identify the possibility of between cardiac predisposition to potentially arrhythÂ� mogenic and brain epileptogenic triggers/causes to be treated or prevented, in an attempt to decrease the risk for SUDEP. Respiratory function monitoring is also needed (Bigger 1990). 7. Academic fellowships and competitions for medical students and postdoctoral fellows/residents and faculty will attract medical and graduate students and faculty to work in the ἀeld of SUDEP. 8. Grant funding is essential to move the SUDEP knowledge base forward. Academic administrative leaders are not interested in faculty and investigators addressing a problem that is not well funded at the national/international levels. 9. Pharmaceutical industry leaders are not interested in addressing a problem if the market for a new product is not a large one. Incentives must be offered to encourage them to examine the problem of SUDEP and to develop new drugs that may or may not have a large market. If the true incidence of SUDEP is established to be much higher than previously thought to be, by the correct use of the term SUDEP on autopsy reports and/or the use of verbal autopsies postmortem, then the market for new antiepileptic and/or other new drugs in different categories to treat SUDEP will be larger than presently determined. The Food and Drug Administration category of orphan drug development should be considered if necessary. 10. Leadership foundations of vision, knowledge, and courage are essential to address the global mystery of SUDEP. Use of a leadership philosophy foundation that provides strength primarily for research and teaching programs from faculty members and students and secondarily on the administration that must provide the innovative vision and approaches, facilities, and monies to support the needs of the faculty and students. The interaction of teaching and research is essential. While a student is learning how to conduct research, he must simultaneously learn how to become a teacher himself. The components of teaching, including excellent communication and writing skills, coupled with patience, form the€foundÂ� ation for communicating the ἀndings of research to the academic and research communities and to fundÂ�ing agencies. Today’s medical and graduate programs will not be able to train each student for each and every advance that will develop in his selected profession, including a focus on SUDEP. Therefore, the program must help today’s student understand the basic approach to survival in a rapidly changing technological, academic, and political environment. The best prepared student for the challenges of tomorrow will be one trained to be flexible, possessing the basic knowledge and tools and courage to adapt to the different career twists and turns encountered as modern technology rushes to the forefront with numerous new techniques, understandings, and thoughts foreign to us at this time in our lives. (continued)

Epilepsy and SUDEP

961

Table 60.2â•… (Continued) Today’s medial and graduate leadership must have the vision to provide a fertile and proper environment for teachers to work with academic freedom to teach today’s students how to become the self-learning student of tomorrow. While addressing issues of SUDEP, teach our students of today to become selflearners and leaders in the ἀeld for tomorrow’s solutions (Lathers 1995a, 1995b, 1994). Source: Lathers, C. M., Epilepsy Behav, 15, 269–277, 2009. With permission.

4. To secure infrastructure grants to fund a consortium of centers that will conduct prospect clinical and basic research studies to identify preventable risk factors and mechanisms underlying SUDEP. Additional symposia must be held to encourage discussion and “thinking out of the box” to solve the problem of SUDEP (Table 60.2). Around the world, all must work to introduce excitement and an intellectual interest in young students, scientists, and clinicians, with the goal of encouraging them to focus on epilepsy and sudden death. Only by encouraging an understanding of the importance of solving the mystery of SUDEP for young investigators will we be able to ἀnd answers that solve the mystery of sudden death (Lathers, Schraeder, and Bungo 2008a, 2008b; Scorza, Arida, and Cavalheiro 2008; Lathers, Schraeder, Bungo 2010; Lathers and Schraeder 1990).

References Bacon, G. M. 1868. On the modes of death in epilepsy. Lancet 1: 555–556. Bateman, L. M., C. S. Li, and M. Seyal. 2008. Ictal hypoxemia in localization-related epilepsy: Analysis of incidence, severity and risk factors. Brain 131 (Pt 12): 3239–3245. Bigger, J. T. 1990. Foreword. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Billman, G. E., L. C. Castillo, J. Hensley, C. M. Hohl, and R. A. Altschuld. 1997. Beta2-adrenergic receptor antagonists protect against ventricular ἀbrillation: In vivo and in vitro evidence for enhanced sensitivity to beta2-adrenergic stimulation in animals susceptible to sudden death. Circulation 96 (6): 1914–1922. Billman, G. E., M. Kukielka, R. Kelley, M. Moustafa-Bayoumi, and R. A. Altschuld. 2006. Endurance exercise training attenuates cardiac beta2-adrenoceptor responsiveness and prevents ventricular ἀbrillation in animals susceptible to sudden death. Am J Physiol Heart Circ Physiol 290 (6): H2590–H2599. Carnel, S. B., P. L. Schraeder, and C. M. Lathers. 1985. Effect of phenobarbital pretreatment on cardiac neural discharge and pentylenetetrazol-induced epileptogenic activity in the cat. Pharmacology 30 (4): 225–240. Celiker, A., I. Erdogan, T. Karagoz, and S. Ozer. 2009. Clinical experiences of patients with catecholaminergic polymorphic ventricular tachycardia. Cardiol Young 19 (1): 45–52. Ceremuzynski, L., J. Staszewska-Barczak, and K. Herbaczynska-Cedro. 1969. Cardiac rhythm disturbances and the release of catecholamines after acute coronary occlusion in dogs. Cardiovasc Res 3 (2): 190–197. Chadwick, D., P. Jenner, and E. H. Reynolds. 1975. Amines, anticonvulsants, and epilepsy. Lancet 1 (7905): 473–476.

962 Sudden Death in Epilepsy: Forensic and Clinical Issues Dodd-O, J. H., and C. M. Lathers. 1990. Chapter 13. A characterization of the lockstep phenomenon in phenobarbital-pretreated cats. In Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Druschky, A., M. J. Hilz, P. Hopp, G. Platsch, M. Radespiel-Troger, K. Druschky, T. Kuwert, H. Stefan, and B. Neundorfer. 2001. Interictal cardiac autonomic dysfunction in temporal lobe epilepsy demonstrated by [(123)I]metaiodobenzylguanidine-SPECT. Brain 124 (Pt 12): 2372–2382. Du, X. J., X. M. Gao, G. L. Jennings, A. M. Dart, and E. A. Woodcock. 2000. Preserved ventricular contractility in infarcted mouse heart overexpressing beta(2)-adrenergic receptors. Am J Physiol Heart Circ Physiol 279 (5): H2456–H2463. Earnest, M. P., G. E. Thomas, R. A. Eden, and K. F. Hossack. 1992. The sudden unexplained death syndrome in epilepsy: Demographic, clinical, and postmortem features. Epilepsia 33 (2): 310–316. Ficker, D. M., E. L. So, W. K. Shen, J. F. Annegers, P. C. O’Brien, G. D. Cascino, and P. G. Belau. 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51 (5): 1270–1274. Han, J., and G. K. Moe. 1964. Nonuniform recovery of excitability in ventricular muscle. Circ Res 14: 44–60. Houle, M. S., R. A. Altschuld, and G. E. Billman. 2001. Enhanced in vivo and in vitro contractile responses to beta(2)-adrenergic receptor stimulation in dogs susceptible to lethal arrhythmias. J Appl Physiol 91 (4): 1627–1637. Hughes, J. R. 2009. A review of sudden unexpected death in epilepsy: Prediction of patients at risk. Epilepsy Behav 14 (2): 280–287. Ieda, M., H. Kanazawa, K. Kimura, F. Hattori, Y. Ieda, M. Taniguchi, J. K. Lee, et al. 2007. Sema3a maintains normal heart rhythm through sympathetic innervation patterning. Nat Med 13 (5): 604–612. Ieda, M., H. Kanazawa, Y. Ieda, K. Kimura, K. Matsumura, Y. Tomita, T. Yagi, et al. 2006. Nerve growth factor is critical for cardiac sensory innervation and rescues neuropathy in diabetic hearts. Circulation 114 (22): 2351–2363. Ieda, M., K. Kimura, H. Kanazawa, and K. Fukuda. 2008. Regulation of cardiac nerves: A new paradigm in the management of sudden cardiac death? Curr Med Chem 15 (17): 1731–1736. Jim, K. F., C. M. Lathers, V. L. Farris, L. F. Pratt, and W. H. Spivey. 1989. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous lorazepam in pigs. Epilepsia 30 (4): 480–486. Jim, K. F., C. M. Lathers, W. H. Spivey, W. D. Matthews, C. Kahn, and K. Dolce. 1988. Suppression of pentylenetetrazol-elicited seizure activity by intraosseous propranolol in pigs. J Clin Pharmacol 28 (12): 1106–1111. Kelliher, G. J., C. Widmer, and J. Roberts. 1975. Influence of the adrenal medulla on cardiac rhythm disturbances following acute coronary artery occlusions. Recent Adv Stud Cardiac Struct Metab 10: 387–400. Kerling, F., M. Dutsch, R. Linke, T. Kuwert, H. Stefan, and M. J. Hilz. 2009. Relation between ictal asystole and cardiac sympathetic dysfunction shown by MIBG-SPECT. Acta Neurol Scand 120 (2): 123–129. Killam, K. F., and J. A. Bain. 1957. Convulsant hydrazides. I. In vitro and in vivo inhibition of vitamin B6 enzymes by convulsant hydrazides. J Pharmacol Exp Ther 119 (2): 255–262. Lathers, C. M. 1980a. Effect of timolol on autonomic neural discharge associated with ouabaininduced arrhythmia. Eur J Pharmacol 64 (2–3): 95–106. Lathers, C. M. 1980b. The effect of metoprolol on coronary occlusion-induced arrhythmia and autonomic neural discharge. Fed Proc 39: 771. Lathers, C. M. 1981. Induced disease. Myocardial infarction in dogs and cats. In Mammalian Models for Research on Aging, 224–228. Washington, DC: National Academy Press. Lathers, C. M. 1982. Lack of effect of methylprednisolone on cardiac neural discharge associated with coronary occlusion-induced arrhythmia and death. Eur J Pharmacol 85 (2): 233–238.

Epilepsy and SUDEP

963

Lathers, C. M. 2009. Epilepsy and sudden death: Personal reflections and call for global action. Epilepsy Behav 15 (3): 269–277. Lathers, C. M., G. J. Kelliher, J. Roberts, and A. B. Beasley. 1978. Nonuniform cardiac sympathetic nerve discharge: Mechanism for coronary occlusion and digitalis-induced arrhythmia. Circulation 57 (6): 1058–1065. Lathers, C. M., J. Roberts, and G. J. Kelliher. 1977. Correlation of ouabain-induced arrhythmia and nonuniformity in the histamine-evoked discharge of cardiac sympathetic nerves. J Pharmacol Exp Ther 203 (2): 467–479. Lathers, C. M., K. F. Jim, W. B. High, W. H. Spivey, W. D. Matthews, and T. Ho. 1989. An investigation of the pathological and physiological effects of intraosseous sodium bicarbonate in pigs. J Clin Pharmacol 29 (4): 354–359. Lathers, C. M., K. F. Jim, W. H. Spivey, C. Kahn, K. Dolce, and W. D. Matthews. 1990a. Chapter 24: Antiepileptic activity of beta-blocking agents. In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Lathers, C. M., K. F. Jim, and W. H. Spivey. 1989. A comparison of intraosseous and intravenous routes of administration for antiseizure agents. Epilepsia 30 (4): 472–479. Lathers, C. M., K. M. Keller, J. Roberts, and A. B. Beasley. 1977. Chapter 5. Role of the adrenergic nervous system in arrhythmia produced by acute coronary artery occlusion. In Pathophysiology and Therapeutics of Myocardial Ischemia, ed. A. M. Leffer, G. J. Kelliher and M. Rovetto. New York, NY: Spectrum. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008a. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and S. B. Carnel. 1984. Neural mechanisms in cardiac arrhythmias associated with epileptogenic activity: The effect of phenobarbital in the cat. Life Sci 34 (20): 1919–1936. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., and P. L. Schraeder. 2002. Clinical pharmacology: Drugs as a beneἀt and/or risk in sudden unexpected death in epilepsy? J Clin Pharmacol 42 (2): 123–136. Lathers, C. M., and P. L. Schraeder. 2006. Stress and sudden death. Epilepsy Behav 9 (2): 236–242. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008b. Chapter 13. Sudden death: Neurocardiologic mystery. In Psychological Factors and Cardiovascular Disorders, ed. L. Sher. Hauppauge, NY: Nova Science. Lathers, C. M., and P. L. Schraeder, eds. 1990. Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Lathers, C. M., R. M. Levin, and W. H. Spivey. 1986. Regional distribution of myocardial betaÂ�adrenoceptors in the cat. Eur J Pharmacol 130 (1–2): 111–117. Lathers, C. M., and R. M. Levin. 2010. Chapter 33: Animal model for sudden cardiac death:€Sympathetic innervation and myocardial beta-receptor densities. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., S. A. Koehler, C. H. Wecht, and P. L. Schraeder. 2003a. Forensic antiepileptic drug levels in 2001 autopsy cases of sudden, unexpected deaths in persons with epilepsy in Allegheny County Pennsylvania. Paper read at the FDA Science Forum, at Washington, DC. Lathers, C. M., S. A. Koehler, C. H. Wecht, and P. L. Schraeder. 2003b. Forensic antiepileptic drug levels in 2001 autopsy cases of sudden, unexpected deaths in persons with epilepsy in Allegheny County Pennsylvania. Paper read at the Annual Meeting of American College of Clinical Pharmacology, September, at Orlando, FL.

964 Sudden Death in Epilepsy: Forensic and Clinical Issues Lathers, C. M., W. H. Spivey, L. E. Suter, J. P. Lerner, N. Tumer, and R. M. Levin. 1986. The effect of acute and chronic administration of timolol on cardiac sympathetic neural discharge, arrhythmia, and beta adrenergic receptor density associated with coronary occlusion in the cat. Life Sci 39 (22): 2121–2141. Lathers, C. M., W. H. Spivey, and N. Tumer. 1988. The effect of timolol given ἀve minutes after coronary occlusion on plasma catecholamines. J Clin Pharmacol 28 (4): 289–299. Lathers, C. M., W. H. Spivey, R. M. Levin, and N. Tumer. 1990b. The effect of dilevalol on cardiac autonomic neural discharge, plasma catecholamines, and myocardial beta receptor density associated with coronary occlusion. J Clin Pharmacol 30 (3): 241–253. Lathers, C. M., W. H. Spivey, and R. M. Levin. 1988. The effect of chronic timolol in an animal model for myocardial infarction. J Clin Pharmacol 28 (8): 736–745. Lathers, C. M. 1994. Presidential Address. Paper read at the American College of Clinical Pharmacology (ACCP) Annual Meeting, December. Lathers, C. M. 1995a. Leadership needs nurturing environment. Times Union, 1995/10/22. Lathers, C. M. 1995b. Presidential Address, Albany College of Pharmacy. Paper read at the Profiles in Leadership Program, September 12, at Albany, NY. Lathers, C. M. 2010. Chapter 25: Sudden death: Animal models to study nervous system sites of action for disease and pharmacological intervention. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 2009. Verbal autopsies. Epilepsy Behav 14: 573–576. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Chapter 4. Unanswered questions: SUDEP studies needed. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Mason, S. T., and M. E. Corcoran. 1979. Catecholamines and convulsions. Brain Res 170 (3): 497–507. Moran, N. F., K. Poole, G. Bell, J. Solomon, S. Kendall, M. McCarthy, D. McCormick, L. Nashef, J. Sander, and S. D. Shorvon. 2004. Epilepsy in the United Kingdom: Seizure frequency and severity, anti-epileptic drug utilization and impact on life in 1652 people with epilepsy. Seizure 13 (6): 425–433. National Institute of Neurological and Communicative Disorders and Stroke (NINCDS). 2008. Workshop on Sudden Unexpected Death in Epilepsy, November 12–14, at NINCDS, Rockville, MD. Oishi, R., N. Suenaga, and T. Fukuda. 1979. Possible involvement of brainstem norepinephrine in pentylenetetrazol convulsions in rats. Pharmacol Biochem Behav 10 (1): 57–61. O’Rourke, D. K., and C. M. Lathers. 1990. Chapter 15. Interspike interval histogram characterization of synchronized cardiac sympathetic neural discharge and epileptogenic activity in the electrocorticogram of the cat. In Epilepsy and Sudden Death., ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Rastogi, S. K., and M. K. Ticku. 1986. Anticonvulsant proἀle of drugs which facilitate GABAergic transmission on convulsions mediated by a GABAergic mechanism. Neuropharmacology 25 (2): 175–185. Rugg-Gunn, F. J., R. J. Simister, M. Squirrell, D. R. Holdright, and J. S. Duncan. 2004. Cardiac arrhythmias in focal epilepsy: A prospective long-term study. Lancet 364 (9452): 2212–2219. Ryvlin, P., T. Tomson, and A. Montavont. 2009. Excess mortality and sudden unexpected death in epilepsy. Presse Med 38 (6): 905–910. Schott, G. D., A. A. McLeod, and D. E. Jewitt. 1977. Cardiac arrhythmias that masquerade as epilepsy. Br Med J 1 (6074): 1454–1457. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2006. Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy. Epilepsy Res 68 (2): 137–143.

Epilepsy and SUDEP

965

Schraeder, P. L., K. Delin, R. L. McClelland, and E. L. So. 2009. A nationwide survey of the extent of autopsy in sudden unexplained death in epilepsy. Am J Forensic Med Pathol 30 (2): 123–126. Schraeder, P. L., R. Pontzer, and T. R. Engel. 1983. A case of being scared to death. Arch Intern Med 143 (9): 1793–1794. Schwartz, R. D. 1988. The GABAA receptor-gated ion channel: Biochemical and pharmacological studies of structure and function. Biochem Pharmacol 37 (18): 3369–3375. Schwartz, R. D., X. Yu, M. R. Katzman, D. M. Hayden-Hixson, and J. M. Perry. 1995. Diazepam, given postischemia, protects selectively vulnerable neurons in the rat hippocampus and striatum. J Neurosci 15 (1 Pt 2): 529–539. Schwartz, R. D., and C. M. Lathers. 1990. GABA neurotransmission, epileptogenic activity, and cardiac arrhythmias (Chapter 17). In Epilepsy and Sudden Death, ed. C. M. Lathers and P. L. Schraeder. New York, NY: Marcel Dekker. Scorza, F. A., R. M. Arida, and E. A. Cavalheiro. 2008. Preventive measures for sudden cardiac death in epilepsy beyond therapies. Epilepsy Behav 13 (1): 263– 264; author reply 265–269. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (Suppl 9): 93–98. So, E. L., J. Bainbridge, J. R. Buchhalter, J. Donalty, E. J. Donner, A. Finucane, N. M. Graves, et al. 2009. Report of the American Epilepsy Society and the Epilepsy Foundation joint task force on sudden unexplained death in epilepsy. Epilepsia 50 (4): 917–922. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41 (11): 1494–1497. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1989. The relationship of the lock-step phenomenon and precipitous changes in mean arterial blood pressure. Electroencephalogr Clin Neurophysiol 72 (4): 340–345. Stauffer, A. Z., J. Dodd-O, and C. M. Lathers. 1990. Chapter 14. Relationship of the lockstep phenomenon and precipitous changes in blood pressure. In Epilepsy and Sudden Death. New York, NY: Marcel Dekker. Suter, L. E., and C. M. Lathers. 1984. Modulation of presynaptic gamma aminobutyric acid release by prostaglandin E2: Explanation for epileptogenic activity and dysfunction in autonomic cardiac neural discharge leading to arrhythmias? Med Hypotheses 15 (1): 15–30. Tomson, T., A. C. Skold, P. Holmgen, L. Nilsson, and B. Danielsson. 1998. Postmortem changes in blood concentrations of phenytoin and carbamazepine: An experimental study. Ther Drug Monit 20 (3): 309–312. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. Walczak, T. S., I. E. Leppik, M. D’Amelio, J. Rarick, E. So, P. Ahman, K. Ruggles, G. D. Cascino, J. F. Annegers, and W. A. Hauser. 2001. Incidence and risk factors in sudden unexpected death in epilepsy: A prospective cohort study. Neurology 56 (4): 519–525. Wang, J., Y. Chen, K. Li, and L. Hou. 2006. Blockade of inhibitory neurotransmission evoked seizurelike ἀring of cardiac parasympathetic neurons in brainstem slices of newborn rats: Implications for sudden deaths in patients of epilepsy. Epilepsy Res 70 (2–3): 172–183.

SUDEP A Mystery Yet to Be Solved Claire M. Lathers Paul L. Schraeder

61

Contents 61.1 Animal Models for SUDEP 61.2 Clariἀcation: Role of Cardiac vs. Respiratory Mechanisms in Actual Death Events 61.3 Need for Multidisciplinary Clinical and Basic Science Collaboration References

967 968 969 969

Sudden unexpected death in epilepsy (SUDEP), including risk factors for and causes and prevention of, is a mystery yet to be solved. We are certainly further along than a few years ago. However, the following discussion briefly presents areas where work is needed.

61.1â•… Animal Models for SUDEP “Think out of the box” when evaluating any established animal model with a potential for modiἀcation(s) to study the mechanism(s) of SUDEP. Multiple relevant animal models are needed to understand the pathophysiology of SUDEP, hypothesize about effective treatments, develop pilot studies in persons with epilepsy, and conduct conἀrmatory largescale clinical trials (Lathers 2009). The importance of using many different animal models to study SUDEP to glean an insight into the various mechanisms of risks and their contribution to the initiation of the death event is discussed in detail by Lathers and colleagues (Lathers and Levin 2010; Lathers, Schraeder, and Bungo 2010, Chapter 1). Schwartz et al. (1995) concluded that delayed enhancement of GABAergic neurotransmission directly at the site of vulnerability after an ischemic event protects the neurons from death. This ἀnding should be explored, for example, to further study the effect of diazepam on gammaaminobutyric acid (GABA)–mediated effects that may prevent ischemia-induced neuronal death and ultimately prevent the worsening of central neuronal communication due to epileptogenic activity. This may eventually contribute a protective central nervous system effect to make an individual less likely to be at risk for SUDEP. There are many more basic science questions to be raised and answered. Wang et al. (2006) provided supporting data for the lockstep phenomenon ἀnding of Lathers and colleagues (Lathers and Schraeder 1982; Schraeder and Lathers 1983; Lathers and Schraeder 2010, Chapter 28). Blockade of GABAergic and glycinergic receptors in medulla slices of newborn rats evoked intermittent seizure-like ἀring of cardiac parasympathetic neurons, suggesting that seizure-like pattern of ἀring during an epileptic attack may cause neurogenic ictal bradyarrhythmias, cardiac asystole, or even sudden death in persons with epilepsy. The study of Wang et€al. 967

968 Sudden Death in Epilepsy: Forensic and Clinical Issues

(2006) must be expanded to examine antiepileptic or other categories of drugs and their effect on the intermittent seizure-like ἀring of cardiac parasympathetic neurons that may cause neurogenic ictal bradyarrhythmias, cardiac asystole, or sudden death. So (2008) emphasized the signiἀcance of using audiogenic seizure mice to study postictal respiratory arrest. Postictal respiratory arrest was induced by serotonin receptor inhibition and prevented by selective serotonin reuptake inhibitor drugs. The role of serotonin in SUDEP must be examined in future animal studies (Tupal and Faingold 2006; Faingold, Tupal, and Uteshev 2010; Paterson 2010).

61.2â•…Clarification: Role of Cardiac vs. Respiratory Mechanisms in Actual Death Events The ongoing discussion of the differential role of cardiac vs. respiratory mechanisms as operant in SUDEP has to some degree resulted in an either/or discussion (Simon et al. 1982; Simon 1993; Simon et al. 1988; Nashef et al. 1996; Nashef and Ryvlin 2009; Lathers, Schraeder, and Weiner 1987; Lathers, Schraeder, and Bungo 2008; Schraeder and Lathers 1983; Leestma et al. 1989; Leung, Kwan, and Elger 2006; Rocamora et al. 2003). The fallacy of such a dichotomous approach is that it diminishes the obvious likelihood that an interactional relationship may exist at several levels. We know that epilepsy-related disturbances of respiratory regulation and that central nervous system–mediated apnea can both occur (So, Sam, and Lagerlund 2000; Johnston et al. 1997; Langan, Nashef, and Sander 2000; Lee et al. 1999; Sethi and Chhabra 2008). In addition, neurogenic pulmonary edema is also seen in association with seizures and is a common ἀnding on autopsy of SUDEP victims (Leestma et al. 1989; Terrence, Rao, and Perper 1981; Sedy et al. 2007). We also know that seizures can be associated with the occurrence of tachyarrhythmias, bradyarrhythmias/asystole, and cardiac conduction changes (Aurlien et al. 2009; Brugada, Brugada, and Brugada 2003; Drake, Reider, and Kay 1993; Espinosa et al. 2009; Ieda et al. 2008; Nei et al. 2004; Opeskin, Thomas, and Berkovic 2000). Therefore, it is self-evident that the occurrence of seizure-related apnea/pulmonary edema/hypoxia/acidosis may well have the potential to set the stage for acute cardiac rhythm disturbance. Thus, there is the possibility of a combination of ingredients making up a stew of possible mechanisms for SUDEP. If such were to be the case, it would be difficult to say which single putative mechanism or combination of mechanisms predominated in any given victim of SUDEP. We also know that there are inherited disturbances of the sodium channel (Brugada, Brugada, and Brugada 2003; Aurlien et al. 2009; Kornick et al. 2003; Herreros 2010, Chapter 19; Lathers, Schraeder, and Bungo 2010, Chapter 1; Lathers, Schraeder, and Bungo 2010, Chapter 20) that predispose to potentially fatal arrhythmias. Persons with Brugada syndrome are known to have seizure-like activity during episodes of potentially fatal cardiac syncope (Sharma, Ho, and Kantharia 2010; Lathers, Schraeder, and Bungo 2010, Chapter 20), raising the distinct possibility that at least some SUDEP victims may have had an incorrect diagnosis of epilepsy. On the other hand, there has been no effort to investigate persons with epilepsy to determine if there is a subgroup that has a previously unrecognized cardiac predisposition, such a latent Brugada syndrome, to arrhythmia. Likewise, more work needs to be done on investigating the possibility of persons with epilepsy having any pattern or patterns of breathing disturbance, e.g., transient episodes of apnea, during sleep (Hughes and Sato 2010, Chapter 23).

SUDEP

969

The fact that the most cases of SUDEP occur during or associated with sleep implies that the sleep state may incorporate a predisposing factor or factors in association with epilepsy. Nasheff et al. (1998) have found that the sleeping position of SUDEP victims may be a risk for suffocation and advocate that persons with poorly controlled seizures who live in an institutional environment have ongoing monitoring during sleep. The overall approach to what interventions during sleep may be the most helpful preventive measures for the vast majority of potential victims of SUDEP who have relatively infrequent seizures and who live at home or alone is not known. As with sudden infant death syndrome, the possibility that a simple intervention such as sleeping supine rather than prone might be of measurable preventive beneἀt deserves investigation. As one can see from this brief discussion, there is an obvious need for interdisciplinary clinical research among epileptologists, cardiologists, pulmonologists, and sleep specialists. Unraveling the mystery of the mechanisms (the plural is used purposely) of SUDEP will not be an easy task. Developing simple preventive interventions that may at least diminish the occurrence of SUDEP could result from studies that screen persons with epilepsy and family members for potential cardiac sodium channel abnormalities, symptoms of sleep disturbance, and even sleeping position.

61.3â•…Need for Multidisciplinary Clinical and Basic Science Collaboration A worldwide network of professionals must focus on basic scientiἀc research programs as well as clinical and epidemiology studies. Team work among different multidisciplinary professionals in clinical settings and within and among laboratories should address the global issues of SUDEP. The ἀelds of neurology, pharmacology, clinical pharmacology, cardiology, pulmonology, and sleep have much to offer as we work to improve compliance, develop new antiepileptic drugs, and apply different categories of drugs to resolve the mystery of SUDEP. Ambulatory simultaneous electrocardiographic and electroencephalographic telemetry monitoring of patients at risk for sudden death will help identify cardiac vs. epileptogenic triggers to decrease risk of SUDEP.

References Aurlien, D., T. P. Leren, E. Tauboll, and L. Gjerstad. 2009. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18 (2): 158–160. Brugada, J., R. Brugada, and P. Brugada. 2003. Determinants of sudden cardiac death in individuals with the electrocardiographic pattern of Brugada syndrome and no previous cardiac arrest. Circulation 108 (25): 3092–3096. Drake, M. E., C. R. Reider, and A. Kay. 1993. Electrocardiography in epilepsy patients without cardiac symptoms. Seizure 2 (1): 63–65. Espinosa, P. S., J. W. Lee, U. B. Tedrow, E. B. Bromἀeld, and B. A. Dworetzky. 2009. Sudden unexpected near death in epilepsy: Malignant arrhythmia from a partial seizure. Neurology 72 (19): 1702–1703. Faingold, C. L., S. Tupal, Y. Mhaskar, and V. V. Uteshev. 2010. Chapter 41. DBA mice as models of sudden unexpected death in epilepsy. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. Bungo and J. E. Leestma. Boca Raton, FL: CRC Press.

970 Sudden Death in Epilepsy: Forensic and Clinical Issues Herreros, B. 2010. Cardiac channelopathies and sudden death. In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 19. Boca Raton, FL: CRC Press. Hughes, J. R., and S. Sato. 2010. Sudden death in epilepsy: Relationship to the sleep-wake circadian cycle and fractal physiology. In Sudden Death in Epilepsy: Relationship to the Sleep-Wake Circadian Cycle and Fractal Physiology, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 23. Boca Raton, FL: CRC Press. Ieda, M., K. Kimura, H. Kanazawa, and K. Fukuda. 2008. Regulation of cardiac nerves: A new paradigm in the management of sudden cardiac death? Curr Med Chem 15 (17): 1731–1736. Johnston, S. C., R. Siedenberg, J. K. Min, E. H. Jerome, and K. D. Laxer. 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann Neurol 42 (4): 588–594. Kornick, C. A., M. J. Kilborn, J. Santiago-Palma, G. Schulman, H. T. Thaler, D. L. Keefe, A. N. Katchman et al. 2003. QTc interval prolongation associated with intravenous methadone. Pain 105 (3): 499–506. Langan, Y., L. Nashef, and J. W. Sander. 2000. Sudden unexpected death in epilepsy: A series of witnessed deaths. J Neurol Neurosurg Psychiatry 68 (2): 211–213. Lathers, C. M. 2009. Epilepsy and sudden death: Personal reflections and call for global action. Epilepsy Behav 15 (3): 269–277. Lathers, C. M., P. L. Schraeder, and F. L. Weiner. 1987. Synchronization of cardiac autonomic neural discharge with epileptogenic activity: The lockstep phenomenon. Electroencephalogr Clin Neurophysiol 67 (3): 247–259. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2008. The mystery of sudden death: Mechanisms for risks. Epilepsy Behav 12 (1): 3–24. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Sodium channel dysfunction: Common pathophysiological mechanism associated with sudden death ECG abnormalities in Brugada syndrome and some types of epilepsy. Case histories (Chapter 20). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., and P. L. Schraeder. 1982. Autonomic dysfunction in epilepsy: Characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23 (6): 633–647. Lathers, C. M., P. Schraeder et al. 2010. Animal model for sudden unexpected death in persons with epilepsy (Chapter 28). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., P. L. Schraeder, and M. W. Bungo. 2010. Neurocardiologic mechanistic risk factors in sudden unexpected death in epilepsy (Chapter 1). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Lathers, C. M., and R. M. Levin. 2010. Animal model for sudden cardiac death. Sympathetic innervation and myocardial beta receptor densities (Chapter 33). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma, Chapter 1. Boca Raton, FL: CRC Press. Lee, H. W., S. B. Hong, W. S. Tae, D. W. Seo, and S. E. Kim. 1999. Partial seizures manifesting as apnea only in an adult. Epilepsia 40 (12): 1828–1831. Leestma, J. E., T. Walczak, J. R. Hughes, M. B. Kalelkar, and S. S. Teas. 1989. A prospective study on sudden unexpected death in epilepsy. Ann Neurol 26 (2): 195–203. Leung, H., P. Kwan, and C. E. Elger. 2006. Finding the missing link between ictal bradyarrhythmia, ictal asystole, and sudden unexpected death in epilepsy. Epilepsy Behav 9 (1): 19–30. Nashef, L., F. Walker, P. Allen, J. W. Sander, S. D. Shorvon, and D. R. Fish. 1996. Apnoea and bradycardia during epileptic seizures: Relation to sudden death in epilepsy. J Neurol Neurosurg Psychiatry 60 (3): 297–300.

SUDEP

971

Nashef, L., and P. Ryvlin. 2009. Sudden unexpected death in epilepsy (SUDEP): Update and reflections. Neurol Clin 27 (4): 1063–1074. Nashef, L., S. Garner, J. W. Sander, D. R. Fish, and S. D. Shorvon. 1998. Circumstances of death in sudden death in epilepsy: Interviews of bereaved relatives. J Neurol Neurosurg Psychiatry 64 (3): 349–352. Nei, M., R. T. Ho, B. W. Abou-Khalil, F. W. Drislane, J. Liporace, A. Romeo, and M. R. Sperling. 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45 (4): 338–345. Opeskin, K., A. Thomas, and S. F. Berkovic. 2000. Does cardiac conduction pathology contribute to sudden unexpected death in epilepsy? Epilepsy Res 40 (1): 17–24. Paterson, D. S. 2010. Medullary serotonergic abnormalities in sudden infant death syndrome: Implications in SUDEP (Chapter 5). In Sudden Death in Epilepsy: Forensic and Clinical Issues, ed. C. M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton, FL: CRC Press. Rocamora, R., M. Kurthen, L. Lickfett, J. Von Oertzen, and C. E. Elger. 2003. Cardiac asystole in epilepsy: Clinical and neurophysiologic features. Epilepsia 44 (2): 179–185. Schraeder, P. L., and C. M. Lathers. 1983. Cardiac neural discharge and epileptogenic activity in the cat: An animal model for unexplained death. Life Sci 32 (12): 1371–1382. Schwartz, R. D., X. Yu, M. R. Katzman, D. M. Hayden-Hixson, and J. M. Perry. 1995. Diazepam, given postischemia, protects selectively vulnerable neurons in the rat hippocampus and striatum. J Neurosci 15 (1 Pt 2): 529–539. Sedy, J., J. Zicha, J. Kunes, P. Jendelova, and E. Sykova. 2008. Mechanisms of neurogenic pulmonary edema development. Physiol Res 57 (4): 499–506. Sethi, D., and A. Chhabra. 2008. Seizure disorder leading to apnea and bradycardia in a 9-year-old child in immediate postoperative period. Paediatr Anaesth 18 (12): 1211–1212. Sharma, S., T. Ho, and B. K. Kantharia. 2010. Not seizure but syncope (Chapter 21). In Sudden Death in Epilepsy: Forensic and Clinical Issues,€ed C.€M. Lathers, P. L. Schraeder, M. W. Bungo, and J. E. Leestma. Boca Raton: CRC Press. Simon, R. P. 1993. Neurogenic pulmonary edema. Neurol Clin 11 (2): 309–323. Simon, R. P., B. Graham, L. L. Bayne, and T. M. Darragh. 1988. Effect of pulmonary vascular pressure on lung lymph flow following seizures. Chest 93 (2): 386–389. Simon, R. P., L. L. Bayne, R. F. Tranbaugh, and F. R. Lewis. 1982. Elevated pulmonary lymph flow and protein content during status epilepticus in sheep. J Appl Physiol 52 (1): 91–95. So, E. L. 2008. What is known about the mechanisms underlying SUDEP? Epilepsia 49 (Suppl 9): 93–98. So, E. L., M. C. Sam, and T. L. Lagerlund. 2000. Postictal central apnea as a cause of SUDEP: Evidence from near-SUDEP incident. Epilepsia 41 (11): 1494–1497. Terrence, C. F., G. R. Rao, and J. A. Perper. 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann Neurol 9 (5): 458–464. Tupal, S., and C. L. Faingold. 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1): 21–26. Wang, J., Y. Chen, K. Li, and L. Hou. 2006. Blockade of inhibitory neurotransmission evoked seizurelike ἀring of cardiac parasympathetic neurons in brainstem slices of newborn rats: Implications for sudden deaths in patients of epilepsy. Epilepsy Res 70 (2–3): 172–183.

Forensic Evidence and Expert Witnesses Scientific Evidence: Getting It in and Keeping It Out

62

Thomas L. Bohan

Contents 62.1 62.2 62.3 62.4 62.5

Objectives Introduction Common Law vs. Statutory Law Tort Actions Forensic Evidence and Expert Witnesses 62.5.1 Frye vs. Daubert 62.5.2 Federal vs. State, State vs. State 62.5.3 Criminal vs. Civil 62.6 Conclusions and Some Practical Suggestions References

973 974 975 977 978 978 979 979 980 982

62.1â•…Objectives This chapter provides a primer of the American justice system* and then sets out within that framework the current rules (de facto as well as de jure) governing admission into evidence of expert testimony. It concludes with instructions of what to examine when one is facing a hearing to determine whether one will be allowed to testify as an expert witness on a medical subject matter. The rationale is to provide the reader of this book with not only a key to understanding trials he or she hears described second or third hand, but, more importantly, to prepare the reader for the eventuality that he or she will actually participate in a criminal or civil trial as an expert witness or consultant. It is in the latter capacity that he or she will be most focused on (1) helping introduce particular scientiἀc testimony into evidence at trial and (2) helping exclude that expert testimony that appears to be bogus. The U.S. justice system is easy to understand piecewise; it is its totality that can be quite overwhelming. Probably its most difficult aspect arises from the number of overlapping divisions to which it is subject, all of which must be taken into account in considering any particular trial. For openers, there is the usually-but-not-always-obvious distinction between criminal proceedings and civil proceedings. Having determined whether the trial is governed by criminal or civil rules of procedure is not enough. Within the United States,

* Although the major emphasis in this chapter is on U.S. trials, much of it will be applicable across all the countries sharing the common-law tradition (to be defined later in this chapter).

973

974 Sudden Death in Epilepsy: Forensic and Clinical Issues

there are more than 50 state* jurisdictions, with no assurance that any particular criminal or civil rule will be the same in any two of them. Although there is only one federal system, many rules regarding procedure and substance will vary from one to another of the 11 federal circuits.† Furthermore, when state civil actions are removed to federal court (under the latter’s “diversity” jurisdiction), the federal judge must adhere to state law. Nevertheless, there is at least the comfort that all federal courts must comply with Daubert and Kumho Tire (q.v.)‡ in determining whether particular expert testimony should be admitted into evidence. Of course, this is little comfort in the large scheme of things since essentially all trials occur in the state courts. Moreover, the postÂ�-Daubert, post-Kumho Tire track record in the federal courts reveal that some types of proffered expert testimony receive much greater scrutiny than others despite the universal reliability criterion imposed by Kumho Tire on all expert testimony. This irregular landscape will be explored in detail in the second half of this chapter. I will note here, however, that it is medical testimony that seems to be given practically a bye when it comes to the trial judges gatekeeper function, an ironic circumstance given that the seminal Daubert decision resulted in the exclusion of medical expert testimony.

62.2â•…Introduction For a good many years, disputes in the United States have by and large been settled through process of law rather than through dueling or lynching, a circumstance to reflect on when confronted with the whine that ours is a “litigious society.” Because of the widespread use of the legal system to settle disputes and to prosecute those accused of criminal behavior, there is probably no ἀeld of endeavor that does not at one time or another form the subject matter of courtroom proceedings. The practice of medicine is more likely than most ἀelds to be implicated in legal disputes, civil and criminal. It therefore behooves all who engage in, or hope to engage in, a medical career in the United States to familiarize themselves with the nature of civil and criminal trials in the courts of the United States. This chapter seeks to provide that familiarity, which should also serve well those who practice in Canada or, indeed, any English-speaking country. Civil litigation in English-speaking countries is dominated by what some call the “great common law tradition.” This includes the United States outside of Louisiana and Canada outside of Quebec (these exceptions having their legal systems rooted in the Napoleonic Code).§ Even in the United States and the other common-law countries, however, common law is only part of the story. Common law and statutory law work together to determine the operation of our legal system, especially that part of it involved in the resolution of civil disputes. The common law/statutory law distinction carries across all of the divisions * The District of Columbia, Puerto Rico, the U.S. Virgin Islands, etc., have state-like court systems. † The label “circuit” is a vestige from two centuries ago when these courts held trials in all of the federal districts within their circuits, to which they traveled throughout the year, accompanied by the attorneys who were to try the cases. These courts now sit in a fixed location, such as Boston for the First Circuit and New York City for the Second Circuit, and hear appeal from trials that have taken place in federal district courts within their territory. ‡ These cases will be discussed in this chapter. They are the governing decisions of the Supreme Court of the United States regarding expert testimony, issued in 1993 and 1999, respectively. § Further exceptions in each country include legal proceedings before Indian and First Nation courts, respectively.

Forensic Evidence and Expert Witnesses

975

in the American legal terrain, and therefore forms a good starting point for the present discussion.

62.3â•… Common Law vs. Statutory Law For those not schooled in jurisprudence and, perhaps more so for those who are schooled, common law can seem reminiscent of looking for guidance to scripture, the only difference being that, in the secular arena, the scripture is found not in a single tome but rather in tens of thousands of judicial decisions stretching back into ancient times. Supporting the appeal-to-scripture image is the practice of referring to appellate court decisions as being “handed down.” It is in the context of common law that one refers to “precedent.” To see what this means, consider a litigant in state court in Vermont. For the sake of deἀnitiveness, assume this litigant is a plaintiff in a civil suit, a party who brought a claim for damages against someone, the defendant in the action. In all legal actions except the most trivial, there will be a number of narrow issues* contested by the parties leading up to trial and even within the trial itself, issues that will require a speciἀc ruling from the court. In preparing to argue a particular issue, the plaintiff in Vermont looks for a precedent governing the issue. The goal in this quest will be to discover a governing precedent that is “on all fours”† with the issue in question and supports the plaintiff’s side of the argument regarding the issue. If such a precedent is found, the subsequent argument before the court reduces to simply directing the attention of the judge‡ to it. In a Vermont trial court, a governing precedent will be a decision rendered by the Vermont Supreme Court. Occasionally, a careless lawyer will make the serious mistake of citing a helpful decision by the governing appellate court, only to learn later (usually under the most embarrassing circumstances) that the decision had subsequently been overruled. One must not only ἀnd a helpful appellate decision, but also determine that it is still “good law.” Notice the use of the word “law” in this context. Despite strenuous complaints from one political faction and then another about the evil of “judges making law,” this is something that they have been doing for hundreds of years as an inherent and necessary component of legal systems based on common law. Returning to our Vermont plaintiff, it is reasonable to conclude that the hoped-for Vermont Supreme Court decision did not exist. One is rarely so lucky. Nevertheless, this will not be the end of the search, a well-argued decision from any respected appellate court supporting one’s side of an issue being of great use in arguing an issue, although not dispositive. Thus, the Vermont plaintiff looks to appellate-court rulings in other states, perhaps beginning with Vermont’s two companions in northern New England plus Massachusetts. Of course, once there is no hope of ἀnding a governing decision, there is no need to stick to state appellate courts. The federal appellate courts, that is, known as the Circuit Courts of Appeal, are a rich source of verbiage. One would expect the Vermont litigant to look ἀrst to decisions from the Court of Appeals for the Second Circuit, the circuit to which Vermont

* Contrary to creeping popular practice, the word “issue” here means “issue” and not “problem.” † “On all fours” is lawyer talk meaning that all relevant aspects of two situations are identical. ‡ When one speaks of issues to be decided, one is referring to interpretations of the law, a responsibility of the trial judge and not the jury.

976 Sudden Death in Epilepsy: Forensic and Clinical Issues

belongs. If there is no help there, the search widens, encompassing all the appellate courts, state and federal in the country. On rare occasions, the issue of interest may be so subtle or the underlying facts so unusual that no appellate court in the United States will have dealt with it. Obvious but worth stating is the fact that appellate decisions on an issue only exist if the issue has been appealed from a trial court. Only a small minority of trials give rise to appeals at all. The usual circumstance is that a party disappointed by the trial verdict appeals for a new trial based on his or her contention that the trial judge made an error in ruling on an issue raised at trial.* At any event, if there have been no appeals regarding the issue of interest to the Vermont plaintiff, he or she may then look to the small minority of trial decisions that have been published. Rather than argue empty-handed, a lawyer prefers to have something to point to, even if it is just a ruling by a trial judge. It has been observed that much legal work is driven by lawyers’ universal desire never to be the ἀrst person to argue a particular point of view on an issue. Indeed, if the issue is of particular importance, the party in Vermont state court may even look to judicial decisions and opinions from courts in other common-law countries. Although common law plays a small role in criminal matters, it dominates civil litigation. Even in that realm, however, statutory law (“black-letter law”) trumps common law wherever the two conflict. The origin of such conflict arises from legislatures from time to time enacting statutes to counter perceived hardships caused by common law. For example, under common law, you could not bring an action for wrongful death, regardless of how close the decedent was to you, how emotionally devastated you were by his or her death, and no matter how culpable the person who caused the death. Under common law, a claim for personal injury could only be brought by the person directly harmed. Since a dead person could not bring a claim, the cause of action died with the decedent. To remedy this harsh result, all the states have enacted wrongful-death statutes that permit recovery for the wrongful death of a family member, thereby overruling common law. While permitting the lawsuits for wrongful death, these statutes also limit the amount that can be recovered in such actions. Another example where statutes have been enacted to trump the rules of common-law actions lies in suits against the government. Such suits were forbidden across the board by the common-law prohibition of “claims against the sovereign.” It is only within the past 75 years in the United States that state and federal tort claims acts have been enacted to open the door somewhat. These tort claims acts narrowly deἀne the alleged governmental wrongs against a private person for which satisfaction can be sought. Only those speciἀed in the controlling act can form the basis for a lawsuit against the governmental body in question. As with wrongful-death statutes, tort claims acts place caps on the monetary recovery that can be obtained by plaintiffs prevailing in claims against the government. For example, the Maine tort claims act permits suits against the state for negligence in the course of highway repair but does not permit claims for bad highway design per se. Also, it limits successful plaintiffs to judgments not to exceed $300,000. From time to time, state legislatures will enact special legislation to waive the maximum recovery for a particular claim. * Most people have to go to law school before learning that appealing a decision to a higher court is not the same as a child appealing his mother’s opinion to his father, or vice versa. The appeals court does not retry the case but limits its consideration to specific errors by the trial judge alleged by the appellate party to have occurred and skewed the trial.

Forensic Evidence and Expert Witnesses

977

62.4â•… Tort Actions Thus far, one word that had not been deἀned is “tort.” A tort (from the same root as “torture”) is usually characterized as a private wrong; that is, a wrongful act—usually unintentional—that injures a private person. A crime is an intentional wrong against the public, although usually one member of the public bears the brunt of the injury to person or purse. An alleged crime may lead to criminal charges, culminating in a criminal trial in which the government is represented by a prosecutor and the accused is normally represented by a defense attorney. An alleged tort must be dealt with by the injured individual asserting a civil claim against the party alleged to have committed the tort. If a civil trial follows, the two individuals will normally be represented by a plaintiff’s attorney and a defense attorney, respectively. The criminal trial run to its conclusion will result in a verdict of “guilty” or “not guilty.”* A civil trial run to its conclusion will result in a “plaintiff’s verdict” or “defense verdict.”† Most civil actions involve the tort of negligence. The plaintiff asserts (and must prove) (1) that the defendant had a duty to the plaintiff, (2) that the defendant breached that duty through negligence, and (3) that the plaintiff was injured (personally or economically) as a result of that breach. Common law dominates the key components of these actions. Essentially no acts are deἀned as negligent per se, that is by statute. Therefore, the set of circumstances surrounding each alleged negligent act must be examined. Because these circumstances vary so widely from case to case, it is unlikely that the parties will be able to ἀnd a governing precedent. Although few cases go to trial without some issues being argued based on earlier appellate decisions, the ultimate issue of whether the defendant’s actions were negligent is not one of these.‡ The purpose of the trial is for the answer to this key question to be determined by the jury (the usual ἀnder of fact in the United States, but not in any other country that I am familiar with), based on the evidence laid before it. There are both common-law and statutory rules governing the nature of the evidence, factual and opinion (expert), that can be introduced, and they vary from jurisdiction to jurisdiction. For example, in negligence actions based on an alleged hazardous condition (slippery walkway, exposed high-voltage wires, etc.), it would seem reasonable that the plaintiff could introduce evidence that the defendant made post-injury repairs. However, many jurisdictions bar such evidence as being against public policy (because of its perceived tendency to discourage the defendant from remedying a hazard if he thinks it can be used against him). Some states did this through enacting statutes prohibiting the introduction of evidence of “subsequent repair,” whereas others did it through case law, decisions by their appellate courts. The result is the same.

* These verdicts establish the defendant’s status in the eyes of the law. They have no effect on whether the defendant was truly guilty or not guilty. Unfortunately, some judges lose track of this truth, as they fulminate at “failure to show remorse” and impose augmented sentences. † This is pretty basic material. However, I am moved to include it by the frequency with which I read newspaper stories announcing that a defendant in a civil suit has been found “guilty” or “cleared” of charges. ‡ For definitiveness, consider whether a merchant leaving the sidewalk bordering his shop icy for a period of 6 hours was thereby negligent. Does it depend on the prevailing temperature? The state of precipitation? Wind speed?

978 Sudden Death in Epilepsy: Forensic and Clinical Issues

62.5â•… Forensic Evidence and Expert Witnesses Although the previous paragraphs provide an essential background and context for a discussion of forensic evidence in the courtroom, it is nevertheless necessary to do some further parsing before getting into the meat of the matter. The matter is, of course, the rules governing the admission and exclusion of scientiἀc testimony at trial. 62.5.1â•… Frye vs. Daubert The modern era for scientiἀc evidence began in 1923. That was the year that an appellate court for the District of Columbia rendered a decision in a criminal case involving one Mr. Frye (Frye v. United States, 1923). Frye had contended that his defense in the court below had been wrongfully hampered by the trial judge’s refusal to allow into evidence testimony that he had “passed” a lie detector test that tended to support his claim of innocence. The appellate court’s decision affirming Frye’s conviction gave us the Frye standard: for scientiἀc testimony to be admitted into evidence, its profferer must establish that the testimony is based on a theory or technique that has general acceptance within the relevant scientiἀc community. Initially, Frye served as governing precedent only in the courts of the District of Columbia. However, over the subsequent 70 years, the majority of state high courts and federal circuit courts throughout the country adopted this standard (the “general acceptance” standard) as their own for both criminal and civil trials. Then, in 1993, following an appeal from a civil action in federal court, the Supreme Court of the United States ruled (Daubert v. Merrill-Dow Pharmaceuticals, Inc., 1993) that the general acceptance standard was too narrow and should not have been used to exclude the plaintiff’s expert testimony in the trial. However, in removing that particular barrier to admission, it replaced it with what it considered a more flexible one, namely that proffered scientiἀc testimony must be shown to be reliable if it was to be admitted into evidence. In ordering federal trial judges to be more vigilant gatekeepers when it came to the admission of scientiἀc evidence, the Supreme Court offered a number of suggestions to them as possible reliability criteria, questions to be posed. These suggested questions quickly became known as the Daubert factors. They can be summarized as follows. Has the technique or theory underlying the proffered testimony been tested and found to have a low error rate? Are there standards for the application of the underlying technique or theory? Has the underlying technique or theory been published in a peer-reviewed journal? Does the underlying technique or theory have general acceptance within the relevant scientiἀc community? Although the Daubert standard regarding reliability (although not the speciἀc suggestions) initially constituted federal common law, it subsequently became federal black-letter law when it was incorporated into Rule 702 of the Federal Rules of Evidence, the rule pertaining to expert testimony. Since the text of Daubert referred to scientiἀc evidence, there was a short-lived effort by civil and criminal attorneys, as well as prosecutors, to deἀne their expert testimony as nonscientiἀc, in the hope that this would relieve them the obligation to establish its reliability. Engineers were claiming that their work was nonscientiἀc, as were ἀre investigators,

Forensic Evidence and Expert Witnesses

979

medical experts, etc. It was rather a shameful episode, brought to an end by a subsequent Supreme Court decision. Kumho Tire (Kumho Tire v. Carmichael, 1999) holds that all expert testimony proffered under Federal Rule of Evidence 702, that is, all expert testimony offered in federal court,* must meet the reliability requirement of Daubert. This is consistent with the plenary coverage of FRE 702 as revised. 62.5.2â•… Federal vs. State, State vs. State Thus, for trials in any federal court, the profferer of any expert testimony must convince the trial judge that it is sufficiently reliable. The situation in the state trial courts is far more complicated. Although most state courts are governed by either the Frye or Daubert standard† as the result of rulings by the respective high courts, very few of them follow Kumho Tire. This is deἀnitely not a trivial distinction since, not being governed by the all-encompassing principle of the latter decision, the courts can narrowly deἀne the expert testimony that must pass a reliability test a la Daubert or Frye. Some states limit judicial scrutiny to testimony based on “cutting-edge science”; many distinguish between sciencebased testimony and experience-based testimony, requiring only the former to satisfy the reliability requirement. This distinction is obviously a lively concept when determining whether medical testimony is to be subject to preadmission reliability hearings (referred to as “Frye hearings” or “Daubert hearings” depending on the state). An extended discussion of the differences among the 50 states with respect to the admission of expert testimony can be found in Keierleber and Bohan (2005). 62.5.3â•… Criminal vs. Civil Although in neither state nor federal court should the admissibility of expert testimony be determined by whether the trial is civil or criminal, it seems to happen in actual practice. In particular, there appear to be more efforts to exclude testimony through a Frye or Daubert hearing or some other kind of pretrial hearing in civil trials than in criminal cases. One theory advanced to explain this asymmetry is that more resources are available to both sides in civil court than there are in criminal trials. There are also some handwaving arguments about Daubert having arisen from a civil trial rather than a criminal one being the reason its application is primarily limited to civil cases. It would be interesting to examine whether the Frye standard, having arisen in a criminal matter, was used predominantly in criminal trials. A more serious allegation, apparently backed up by facts, is that that expert testimony proffered by the prosecution in criminal cases in both state and federal trials seems to be exposed to a lower degree of scrutiny than does expert testimony proffered by the defense. Although economics may again be the explanation, criminal defendants generally not having the resources with which to mount a challenge that the prosecution does, the reaction to the few times that prosecution evidence has been excluded on Daubert grounds suggests that more is involved. In 2002, a federal district

* Interestingly, these do not include the courts of the District of Columbia, which still adhere to the Frye (general acceptance) standard. † There are some notable exceptions, especially Wisconsin, which purports to require only that the expert testimony be “relevant,” while explicitly eschewing any reliability tests. See, for example, Keierleber and Bohan (2005).

980 Sudden Death in Epilepsy: Forensic and Clinical Issues

trial judge in Pennsylvania took Daubert seriously when confronted with Federal Bureau of Investigation (FBI) ἀngerprint testimony purporting to link a latent partial print to a full-rolled print of the defendant. He opined that, contrary to everyone’s assumptions, there had never been any scientiἀc studies establishing the range over which the identiἀcation of latent prints was valid. “Adversarial testing in court is not .€.€. what the Supreme Court meant when it discussed testing as an admissibility factor.” (United States v. Llera Plaza, 2002.) He had no choice, he said, but to exclude testimony that the prints were made by the same person. A storm of protest from law enforcement and prosecutors’ offices erupted during the months between the original Daubert hearing and the judge’s recantation. The recantation was clothed by the judge announcing his belated realization that ἀngerprint identiἀcation was not a science. Rather, it was a “specialty” and therefore need not satisfy Daubert. Collaterally, he said, he now realized that the FBI ἀngerprint technicians never make mistakes when they match known and unknown prints. The comments he made when the 2004 Brandon Mayἀeld case* burst on the scene, showing his second opinion to be wishful thinking, remain unpublished.

62.6â•… Conclusions and Some Practical Suggestions For medical practitioners active in the legal system as advisors and as expert witnesses, it would appear that the good news and the bad news is the same. Judges are not scrutinizing medical expert testimony to the extent that this chapter’s analysis should suggest would be the case. On the one hand, this permits truly knowledgeable and objective medical experts to escape from the nonsensical rejections of their testimony that their counterparts in ἀelds such as engineering sometimes have to endure. On the other hand, it also permits truly awful medical testimony to be admitted into evidence in the face of solid arguments as to why it should not be. In the hope that trial judges are going to improve their ability to scrutinize medical testimony, I offer the following observations for medical practitioners who may be confronting Daubert and Frye hearings with increasing frequency, and who may more often be advising legal counsel on what questions to ask at such hearings for opposing experts. The key to maximizing one’s chance at succeeding at either function is to have a good idea of what is meant by “reliability” for the particular testimony in question. To take a simple example, viewers of CSI-type television probably have heard many times that a fractured hyoid bone in a cadaver indicates that the decedent was strangled. At a Daubert hearing for a pathologist opining a cause of death based on having observed such a fracture, legitimate questions would include a demand for peer-reviewed literature supporting this theory, including discussions of the frequency with which criminal strangulation and fractured hyoids occur and that of the occasions when a fractured hyoid occurs without criminal strangulation being involved. The questioning could legitimately explore how those frequencies were established; that is, how was it deἀnitely established in the studies giving rise to them as to whether criminal strangulation had taken place? If the basis for establishing for the sake of the studies that criminal strangulation had occurred was nothing more than confession by the suspect, it would be legitimate to attack that as a nonscientiἀc basis. * The FBI, using the full panoply of its “zero error” system, notoriously misidentified a latent print from the Madrid train bombing with an Oregon lawyer, Brandon Mayfield, a recent convert to Islam.

Forensic Evidence and Expert Witnesses

981

Indeed, any time that a pathology is deemed pathognomonic of a particular crime, questions of these types should be put to the proffered witness, and, in the absence of satisfactory (i.e., scientiἀcally valid) answers to them, that witness should not be allowed to testify. In the sense that there may be signiἀcant subsets of medical practitioners who accept the theory that a particular pathology is pathognomonic of a particular event, and yet the type of indicia of reliability just outlined does not exist, the Daubert standard is a much better one than the “general acceptance” Frye standard. For most of the years since Daubert was handed down by the U.S. Supreme Court, there have been databases maintained and available online that have reported cases dealing with expert testimony of all kinds. One present such source of information is the Daubert Tracker™. For those who wish to see how any particular issue has been treated in the past by one or another U.S. jurisdiction, or how a particular expert witness has fared at the hands of the courts, this is an invaluable source. One of its valuable features is that it is not limited to appellate decisions but also contains, although on a hit–and-miss basis, decisions issued by trial courts. I would go so far as to say that no one should attempt to write a paper on expert testimony, scientiἀc or otherwise, without consulting the Daubert Tracker. Although there is a fee for use, membership in some professional organizations such as the American Academy of Forensic Sciences provides a reduced rate. In closing, I turn to the overall theme of this book—sudden unexpected death in persons with epilepsy (SUDEP)—to provide an additional example of the legal context that medical specialists in the neurological sciences might encounter. Death while in custody occurs not infrequently in the United States, leading to the necessity of medical cause-ofdeath testimony at postmortem inquests. Often, disputes arise between the families of the decedent and the persons in charge of or overseeing the incarceration of the decedent. If the decedent had been a person with epilepsy, it can be expected that one argument of the latter group is that the death was just another unexpected death of a person with epilepsy. Adversarial litigation being what it is, it may also be expected that both sides of the argument will present medical testimony. It can also be expected that both sides will demand the exclusion of testimony from the expert medical witness for the other side. Depending on the state in question, these demands will lead to Daubert or Frye hearings, at which efforts will be made to show that the witness being questioned does not have a scientiἀc basis for his or her opinions regarding the likelihood or not of the prisoner having died a natural death. Both experts therefore determine in advance the scientiἀc basis for the opinions they are testifying to. Although at present, in many jurisdictions, they may get by simply by asserting their general expertise, their years in the ἀeld, their clinical experience, etc., this may not be the case. The expert attributing the prisoner’s death to SUDEP should be familiar with the known attributes such as age, general health, and so forth, of known cases of SUDEP. By the same token, the expert who is questioning the claim of natural death should be prepared to discuss the rate of SUDEP in persons possessing the characteristics of the decedent prisoner. If the probability is very small that a person with epilepsy will die during a speciἀc period of time as opposed to the probability that this will occur during the person’s lifetime—approximately 10%—the argument that SUDEP is a reasonable explanation for the prisoner’s death decreases accordingly. As can be seen from the above example and the earlier discussion, the legal system is an imperfect means of determining the truth. To the extent possible, scientiἀc issues should be resolved outside of the adversarial system. One excellent route to accomplish this resolution is the National Academy of Sciences, which, when requested and funded

982 Sudden Death in Epilepsy: Forensic and Clinical Issues

to study any topic will carry out its assignment with a committee broadly based in science and technology. One of many studies carried out by the National Academy of Sciences that has had a powerful impact on our legal system was its examination of forensic DNA. Another that is bound to have a far-reaching effect is the report it issued in early 2009: Forensic Science in the United States: A Path Forward (2009).

References Daubert v. Merrill-Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). Keierleber, J. A., and T. L. Bohan. 2005. Ten years after Daubert: The status of the states, J Forensic Sci€50 (5). Kumho Tire v. Carmichael, 526 U.S. 137 (1999). United States v. Llera Plaza, 188 F. Supp. 2d 549 (E.D. Pa. 2002). Strengthening Forensic Science in the United States: A Path Forward. 2009. Washington, DC: National Academies Press.