The Neuropsychiatry of Epilepsy

This page intentionally left blank

The Neuropsychiatry of Epilepsy

Michael Trimble and Bettina Schmitz have assembled a multinational team of experts to review the most recent findings which explore the interface between epilepsy and behaviour disorders. They begin by looking at the classifications available and examine how adequate they are for defining the subtleties of behavioural changes in patients with neurological disorders. Coverage is broad-ranging, from related cognitive problems and the biological underpinnings, to clinical aspects, including pseudoseizures and treatment issues. There has been a great deal of research in this area over recent years, but limited published reviews. This timely book covers the practical implications of ongoing research, and offers both a diagnostic and a management perspective. It will be essential reading for all professionals engaged in the treatment of epileptic patients. Michael Trimble is Professor of Behavioural Neurology at the Institute of Neurology in

London. Bettina Schmitz is based in the Epilepsy Research Group of the Deparment of Neurology,

Humboldt University, Berlin.

The Neuropsychiatry of

Epilepsy Edited by

Michael Trimble and

Bettina Schmitz

   Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge  , United Kingdom Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521813747 © Cambridge University Press 2002 This book is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2002 - isbn-13 978-0-511-07326-7 eBook (EBL) - isbn-10 0-511-07326-7 eBook (EBL) - isbn-13 978-0-521-81374-7 hardback - isbn-10 0-521-81374-3 hardback - isbn-13 978-0-521-00516-6 paperback -  paperback isbn-10 0-521-00516-7 Cambridge University Press has no responsibility for the persistence or accuracy of s for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Every effort has been made in preparing this book to provide accurate and up-to-date information which is in accord with accepted standards and practice at the time of publication. Nevertheless, the authors, editors and publishers can make no warranties that the information contained herein is totally free from error, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publisher therefore disclaim all liability for direct or consequential damages resulting from the use of material contained in this book. Readers are strongly advised to pay careful attention to information provided by the manufacturer of any drugs or equipment that they plan to use.

Contents

List of contributors

Part I 1

page viii

Background Introduction

3

M.R. Trimble and B. Schmitz

2

Neuropsychiatric disorders in epilepsy – epidemiology and classification

5

E.S. Krishnamoorthy

3

Limbic connectivity: anatomical substrates of behavioural disturbances in epilepsy

18

J. Engel Jr, C. Wilson and F. Lopez-Rodriquez

Part II 4

Clinical aspects The psychiatry of idiopathic generalized epilepsy

41

Dieter Janz

5

Epilepsy and learning disorders

62

Cesare Maria Cornaggia and Giuseppe Gobbi

6

Subtle cognitive and behavioural effects of epilepsy

70

Frank M.C. Besag

7

Aggression and epilepsy

81

L. Tebartz van Elst

8

Epilepsy and suicide: a neuropsychiatric analysis

107

Dietrich Blumer

9

Postictal psychoses, revisited Kousuke Kanemoto

v

117

vi

Contents

Part III

Cognitive aspects

10

Dementia and epilepsy

135

Stephen W. Brown

11

The risk of cognitive decline in patients with refractory temporal lobe epilepsy

152

Hennric Jokeit and Alois Ebner

12

Behavioural and neuropsychological aspects of frontal lobe epilepsy

164

Christoph Helmstaedter

Part IV

Nonepileptic attacks

13

Epilepsy, dissociation and nonepileptic seizures

189

Richard J. Brown

14

Psychobiology of psychogenic pseudoseizures

210

J. Chris Sackellares and D. Kalogjera-Sackellares

15

Epilepsy and panic disorder

226

Howard A. Ring and Nuri Gene-Cos

Part V

Treatment complications

16

The effects of antiepileptic drugs on behaviour

241

Bettina Schmitz

17

Antiepileptic drug treatment and epileptic seizures – effects on cognitive function

256

Albert P. Aldenkamp

18

Psychiatric effects of surgery for temporal lobe epilepsy

266

Steffi Koch-Stoecker

19

Vagus nerve stimulation and mood

283

Christian E. Elger and Christian Hoppe

Part VI

Treatment

20

On the use of psychotropic drugs in patients with seizure disorder M.R. Trimble and Anke Hensiek

299

vii

Contents

21

The role of psychotherapy in the treatment of epilepsies

313

Martin Schöndienst

22

Choosing measures to assess quality of life (QOL) in epilepsy

323

Caroline E. Selai, Katja Elstner and M.R. Trimble

Index

343

Contributors

Albert P. Aldenkamp Secretary General Instituut voor Epilepsiebestrijding Meer en Bosch Postbus 21 2100 AA Heemstede The Netherlands Frank M.C. Besag Learning Disability Service Bedfordshire & Luton Community NHS Trust Twinwoods Health Resource Centre Milton Road Clapham Bedfordshire MK41 6AT, UK Dietrich Blumer University of Tennessee College of Medicine Department of Psychiatry 135 North Pauline Memphis TN38105, USA Stephen W. Brown Cornwall Healthcare Trust Unit 10 Bodmin Business Park Harleigh Road Bodmin Cornwall PL31 1AH, UK

viii

Richard J. Brown National Hospitals for Neurology and Neurosurgery Institute of Neurology Queen Square London WC1N 3BG, UK Cesare Maria Cornaggia Clinical Psychiatry University of Milano Bicocca San Gerardo Hospital Via Donizetti 106 Mona, Italy Alois Ebner Epilepsy Centre Bethel Epilepsy Surgery Program Maraweg 21 D33617 Bielefeld, Germany Christian E. Elger Department of Epileptology University of Bonn Sigmund-Freud-Strasse 25 Bonn, FRG-53105 Katja Elstner Raymond Way Neuropsychiatry Unit Institute of Neurology Queen Square London WC1N 3BG, UK

ix

List of contributors Jerome Engel Chief Division of Epilepsy and Clinical Neurophysiology Department of Neurology Reed Neurological Research Centre UCLA School of Medicine 710 Westwood Plaza Los Angeles CA 90095-1769, USA Nuri Gene-Cos St. Bartholomew’s & The Royal London School of Medicine London E1 1BB, UK Giuseppe Gobbi Department of Neuropsychiatry Maggiore Hospital Bologna, Italy Christoph Helmstaedter Universitätsklinik fur Epileptologie Sigmund-Freud-Str. 25 Universität Bonn 53105 Bonn, Germany Anke Hensiek Department of Neurology Addenbrooke’s Hospital P.O. Box 165 Hills Road Cambridge CB2 2HE, UK Christian Hoppe Department of Epileptology University of Bonn Sigmund-Freud-Strasse 25 FRG-53105 Bonn Germany

Dieter Janz Burgunder Strasse 8 14129 Berlin, Germany Hennric Jokeit Schweizerisches Epilepsie-Zentrum Neuropsychologie Bleulerstrasse 60 CH-8008 Zurich Switzerland Dalma Kalogjera-Sackellares Department of Neuroscience University of Florida Florida, USA Kousuke Kanemoto Department of Neuropsychiatry Aichi Medical University 21 Yazako-karimata, Nagakute 480-1195, Aichi, Japan Steffi Koch-Stoecker Klinik Mara I Maraweg 21 D33617 Bielefeld, Germany E.S. Krishnamoorthy National Hospitals for Neurology and Neurosurgery Institute of Neurology Queen Square London WC1N 3BG, UK Howard A. Ring Academic Department of Psychological Medicine Third Floor, Alexandra Wing Royal London Hospital Whitechapel London E1 1BB, UK

x

List of contributors F. Lopez-Rodriquez Department of Psychiatry and Biobehavioural Sciences UCLA School of Medicine 710 Westwood Plaza Los Angeles CA 90095-1769 USA J. Chris Sackellares College of Medicine Department of Neurosciences PO Box 100244 Gainesville Florida 32610-0244, USA Bettina Schmitz Neurologische Klinik und Poliklinik Charité, Virchow Klinikum Humboldt-Universität Augustenburger Platz 1 13353 Berlin, Germany Martin Schöndienst Epilepsie-Zentrum Bethel Maraweg 21 33617 Bielefeld, Germany

Caroline E. Selai 126 Scott Ellis Gardens London NW8 9HG, UK Ludger Tebartz van Elst Department of Psychiatry and Psychotherapy Albert-Ludwigs-University Haupstr.5 79104 Freiburg, Germany Michael R. Trimble National Hospitals for Neurology and Neurosurgery Institute of Neurology Queen Square London WC1N 3BG, UK C. Wilson Department of Neurology UCLA School of Medicine 710 Westwood Plaza Los Angeles CA 90095-1769, USA

Part I

Background

1

Introduction M.R. Trimble1 and B. Schmitz2 1 2

Institute of Neurology, London, UK Humboldt University, Berlin, Germany

In 1997, Dr Pete Engel, President of the International League Against Epilepsy, invited us to establish a commission on psychobiology, and this book represents one of the achievements of the commission’s work. The task of the commission was to explore the interface between epilepsy and behaviour disorders, from a biological and social point of view. To these ends a number of subcommissions were established, and their task was to explore the existing knowledge base of the discipline, to suggest research strategies for interventions and to educate both patients and carers about aspects of epilepsy which a number of people consider to have been neglected. The present book is divided into several parts, covering a spectrum of clinical topics which have been of concern to the commission. Some old chestnuts, for example the interictal psychoses of epilepsy, have not been allocated specific chapters, and a number of other areas, particularly relating to learning disability, cognitive decline, dissociative attacks and vagus nerve stimulation have been included. It is hoped that by expanding upon the literature on some of these less-well-discussed aspects of psychobiology in epilepsy further interest will be stimulated, leading to both intellectual discussion and research endeavours. We start our text with an introduction to the classification of psychiatric disorders in epilepsy. The point is made that existing classifications used in psychiatry such as the DSM–IV are quite inadequate when it comes to dealing with the subtleties of the behaviour changes of patients with neurological disorders. We then discuss the biological underpinnings of some behaviour problems, in terms of exploring the limbic system and related structures that are affected by the process of epilepsy and which are also related to behavioural disorders. The part on clinical aspects explores in particular the problems of learning disability, and introduces the important area of state-dependent learning disabilities: patients with cognitive deficits that can be profoundly reversed by appropriate treatment strategies. Other important areas covered include the ever-controversial topic of aggression, the importance of suicide, and the group of psychoses that occur postictally. 3

4

M.R. Trimble and B. Schmitz

The next part looks further at cognitive problems in patients with epilepsy, examining whether the concept of dementia is relevant, discussing the question as to whether or not there is cognitive decline in patients with various types of epilepsy over time, and also examining the issue of frontal lobe epilepsies. The latter have been well defined from the seizure point of view in recent years, but the behavioural and cognitive associations have yet to be clarified. We make no apology for including a section on nonepileptic seizures. The fact that many patients who are diagnosed as having epilepsy do not have epilepsy, but have some form of nonepileptic attack disorder (pseudoseizure) is now well recognized. This problem has been around for centuries, and such eminent neurologists as Charcot have spent some considerable time attempting to differentiate between nonepileptic and epileptic seizures. However, this often still proves difficult. We still have inadequate information as to the mechanisms for the development of nonepileptic attack disorder, and these, and the possible biological associations, are taken up in this section. The final sections deal with treatments and their side effects. Of importance in this section are the references to surgery, not only temporal lobe resection, but also more recent advances such as vagus nerve stimulation. The beneficial and negative psychiatric consequences of these treatments are at the present time being actively explored, and some early work is presented here. However, in the context of psychobiology, our treatment strategies must go beyond medication and surgical interventions, and we include a discussion of psychodynamic principles in relationship to the management of epilepsy, and also a chapter on quality of life. We, the editors, hope that the book will enliven the debate about the links between epilepsy and behaviour, an area which is often not well discussed, partly because of some worry that any association between psychiatry and epilepsy may stigmatize patients with epilepsy even more than they already are. However, the problems that we have identified are a reality not only in the clinic, but also for patients and carers themselves. It is difficult to define treatment and management strategies if problems are ignored, and so our intention is to enliven this area with these up-to-date reviews on behavioural and cognitive problems in epilepsy, and their social and biological underpinnings.

2

Neuropsychiatric disorders in epilepsy – epidemiology and classification E.S. Krishnamoorthy Institute of Neurology, London, UK

Introduction The association between epilepsy and psychiatric disorders has a long and chequered history. For centuries seizures were considered to be a form of demonic possession. Beginning late in the nineteenth century, considerable attention has been directed towards cataloguing, describing and understanding disorders at the interface between epilepsy and psychiatry, particularly by European neurologists and psychiatrists. However, it is only in the past few decades that any attention has been paid to the epidemiology of these disorders. Similarly, aside from some early attempts by European physicians, there have been no efforts to develop an operational classification of psychiatric disorders in epilepsy (Schmitz and Trimble, 1992 for a review). The paucity of epidemiological research at this interface, and the failure to develop an operational international system of classification, is in stark contrast with developments both in epilepsy per se, and in mental health research. The epidemiology of epilepsy has been well studied in many countries and considerable data (both descriptive and analytical) are now available. Indeed, epilepsy has been subject to the gamut of epidemiological research including cross-sectional, casecontrol and cohort studies (Hauser, 1998). Similarly, operational international systems of classifying epilepsy and its syndromes have been developed both by the Commission on Classification and Terminology of the International League Against Epilepsy (1989), and the World Health Organization (1967), and are used by epileptologists around the world. Impressive developments have also taken place in the field of mental health epidemiology. Efforts by the World Health Organization’s Division of Mental Health, and other pioneering organizations around the world, have led to a significant understanding of the epidemiology of psychiatric disorders. This has led to the development of universally accepted classificatory systems in psychiatry, such as the Diagnostic and Statistical Manual now in its fourth edition (DSM–IV; American 5

6

E.S. Krishnamoorthy

Psychiatric Association, 1994), and the mental disorders component of the International Classification of Diseases, now in its tenth edition (ICD–10; World Health Organization, 1992). The commonly held conviction among epileptologists and neuropsychiatrists is that psychiatric comorbidity is not only common in epilepsy, but that distinct and unique forms of psychopathology are prevalent (Krishnamoorthy 2000, 2001). In the past three decades attention has been directed towards discrete forms of psychopathology in epilepsy such as the temporal lobe personality, interictal and postictal psychosis, and interictal dysphoric disorder (Bear and Fedio, 1977; Blumer 1995, 2000). This combined with the observation of similarities in behaviour during seizures and in psychopathological states has strengthened the notion of an affinity between epilepsy and psychiatric disorder. Yet, the evidence that psychiatric disorders are overrepresented in epilepsy is far from convincing, with conflicting results in different studies. In this chapter the epidemiology of psychiatric comorbidity in adult, non-learning-disabled patients with epilepsy will be reviewed. There is a considerable literature on children and the learning disabled that is being addressed elsewhere in this book (Chapters 5 and 6). Some ideas on how psychiatric disorders in epilepsy may be classified, and the work of the subcommission on classification of the International League Against Epilepsy – the Commission on Epilepsy and Psychobiology – in this regard, will also be discussed. Epidemiology A majority of studies in this area has been hospital- and institution-based. While the contribution of these studies to the current understanding of psychopathology in epilepsy has been invaluable, the strong selection bias in these studies does make the extrapolation of their findings to the majority of patients with epilepsy, who live in the community, difficult. There have been some population-based studies of psychiatric comorbidity that are summarized here. Most studies have been cross-sectional and some have compared cases with controls. By and large, save one or two exceptions, these studies have generated crude estimates of prevalence, rather than more specific epidemiological indices. Population-based studies of psychiatric comorbidity in epilepsy

One of the earliest investigations to be carried out was that of Pond and Bidwell (1959/60), who surveyed patients from 14 doctor’s surgeries in the south-east of England. They found that 29% of 245 patients had psychological disorders of sufficient severity to seek treatment, i.e. conspicuous morbidity. The main criticism levelled against this study is its use of a social worker rather than a trained mental

7

Classification

health professional, and a lack of standardized techniques to assess patients with epilepsy for psychiatric comorbidity. The strength of this study, however, lies in its recognizing, four decades ago, the importance of an epidemiological approach. Gudmundsson (1966) personally surveyed 987 patients with epilepsy living in Iceland and reported that 512 (52%) had personality changes of various kinds. Of these 271 (27.5%) were described as ixoid, 73 (7.4%) as ixothymic and 168 (17.0%) as neurotic. More men were ixoid and more women neurotic. While Gudmundsson, unlike Pond, personally examined every subject, the clinical terminology and classification used have few parallels today, and no attempts were made to reduce bias. However, the high proportion of subjects with behavioural changes in this community-based population is striking and worthy of note. Edeh and Toone (1987) conducted a survey in doctor’s surgeries in south London. They interviewed 88 adult patients with epilepsy drawn from doctor’s surgeries in the area, using the Clinical Interview Schedule, and reported that 48% emerged as psychiatric cases. They also found that while patients with temporal lobe epilepsy (TLE) and focal non-TLE did not differ in terms of psychiatric morbidity, both groups were significantly more impaired than patients with primary generalized epilepsy. The techniques of ascertainment used in this study are commendable. Subjects with epilepsy underwent both CT scans and EEG tests, in confirmation of their diagnosis. The study also used a validated instrument for common mental disorder, the CIS-R (Lewis et al., 1992). In criticism, however, it must be said that the study failed to examine matched population-based controls, psychopathology specific to epilepsy was not examined, and while cases with psychosis were identified, no validated diagnostic instrument for psychosis was administered, the CIS-R being a validated instrument for common mental disorder alone. Cockerell et al. (1996) conducted a nation-wide survey in the UK of acute psychological disorders (APD) in patients with epilepsy using the British Neurological Surveillance Unit. Sixty-four incident cases were ascertained over a period of one year. Thirty-one were considered to have APD due to ictal or postictal activity and 33 were interictal. In 30% of cases the APD was reported by the referring physician to be secondary to the prescription of an antiepileptic drug (AED). The drugs most commonly implicated were carbamazepine, lamotrigine and vigabatrin. The broad psychiatric categories diagnosed included delirium (25%), schizophreniform (31%), affective (30%), delusional (5%) and other disorders (9%). The findings of this study are of interest as it gives us crude incidence figures of acute psychiatric disorder in epilepsy and highlights the importance of antiepileptic drugs in precipitating comorbid psychiatric illness in epilepsy. However, as the study used a reporting system, rather than a population-based cohort, the results cannot be used to generate population-based incidence figures, or be generalized.

8

E.S. Krishnamoorthy

Jalava and Sillanpaa (1996) examined a prospective population-based cohort (mean follow-up of 35 years) of patients with epilepsy since childhood, for comorbid somatic, psychosomatic and psychiatric disorders. In comparison with random controls, patients with epilepsy had a fourfold risk of psychiatric disorders or combinations of somatic, psychosomatic and/or psychiatric disorders. Thus patients with childhood-onset epilepsy demonstrated a higher risk for psychiatric or psychosomatic disorders and this appeared to be related to epilepsy and not AED administration. This is perhaps the only cohort study of psychiatric comorbidity in epilepsy and the findings have great relevance. The results clearly indicate that subjects with epilepsy are at higher risk of developing comorbid psychiatric illness, when compared with population-based controls, and indicate the need for greater provision for psychiatric treatment in primary care settings for epilepsy. However, as individual cases were not ascertained in any systematic way, it is possible that the findings do not represent the true extent of comorbidity, with subtle nevertheless disabling forms of psychopathology, or those not requiring medical attention or admission, being missed. This is of relevance, as subtle forms of psychopathology that often do not meet conventional diagnostic criteria may be overrepresented in epilepsy. Bredkjaer et al. (1998) conducted a record-linkage study in Denmark between a sample of people with epilepsy from the National Patient Register and from the Danish Psychiatric Register. They found that the incidence of nonorganic nonaffective psychoses including personality disorders that were broadly within the schizophrenia spectrum was significantly increased for both men and women with epilepsy, even after excluding all people diagnosed as suffering from a learning disability or substance misuse. The standardized incidence ratio was significantly increased for the entire schizophrenia spectrum (P⬍10–8), nonaffective psychosis (P⬍10–8) and schizophrenia alone (P⬍0.0001). In the absence of long-term prospective data, this study based on national registers provides evidence that disorders in the schizophrenia spectrum are clearly overrepresented in epilepsy. The study enabled the calculation of more sophisticated epidemiological indices, such as standardized incidence ratio, that have not been estimated in previous studies. However, the methodological limitations of reliance upon a case-register, i.e. the lack of standardization of ascertainment methods, both for epilepsy and psychoses, and the exclusion of more subtle cases, or those not requiring admission, do apply here. Stefansson et al. (1998) conducted a case-control study comparing the prevalence of nonorganic psychiatric disorders among patients with epilepsy and controls with other somatic diseases, both groups being of normal intelligence. The two groups were drawn from a disability register of the State Social Security Institute in Iceland. In this way, 241 index cases meeting inclusion criteria were

9

Classification

identified, and the ratio between subject (epilepsy) and control (somatic illness) cases was 1:2. Psychiatric diagnosis was present among 35% of cases as compared with 30% of controls, the difference not being statistically significant. Psychiatric disorders were, however, significantly more common in men with epilepsy, but not in women, the difference being due to a significantly higher rate of psychosis, particularly schizophrenia or paranoid states, among men. Some large hospital-based studies

Currie et al. (1971) surveyed 666 patients recorded to have features of temporal lobe epilepsy in the hospital diagnostic index and the records of the neurology, neurosurgery and EEG departments. They found 375 (56%) to be normal, 127 (19%) to be anxious, 71 (11%) to be depressed, 47 (7%) to be aggressive, 41 (6%) to be obsessive and 38 (6%) to have a severe disturbance of affect. Smith et al. (1986) studied 622 patients in the USA in a nation-wide cooperative study spanning 10 Veterans Administration Medical centres, using a battery of neuropsychological testing procedures. The majority of patients was not on anticonvulsant drugs at the time of initial testing, and the few who were had subtherapeutic levels on measurement. They found that patients with epilepsy scored significantly and consistently below the level of the 74 control subjects on all but three behavioural measures. Differences reaching statistical significance were found on tests of motor function (finger tapping, pegboard, colour naming), cognitive-attention (digit symbol, discrimination reaction time, word fluency) and subtests of the Profile of Mood States (tension, depression, vigour and confusion). These they felt provided a profile of behavioural characteristics of unmedicated patients with epilepsy. Gureje (1991) evaluated 204 unselected patients with epilepsy attending a neurological clinic using the Clinical Interview Schedule (Goldberg, 1972); 37% emerged as psychiatric cases. Of these 53% had a neurosis, 29% had a psychosis and 7% were diagnosed to have a personality disorder. Mendez et al. (1993) conducted a retrospective investigation of neurology clinic attenders. They found that interictal schizophrenic disorders occurred in 149 (9.25%) of 1611 patients with epilepsy as compared to only 23 (1.06%) of 2167 patients with migraine. They went on in the latter part of the study to compare 62 epilepsy and schizophrenia patients with 62 patients who had epilepsy alone on 6 seizure variables, and 62 patients with schizophrenia alone on 10 psychosis variables. The epilepsy and schizophrenia group was found to have a later age of onset of epilepsy with more complex partial seizures, more patients with auras and fewer patients with generalized epilepsy. Except for increased suicidal behaviour, patients with epilepsy did not differ from controls on psychosis variables; however, psychotic symptoms often emerged with increased seizure activity. They felt that the

10

E.S. Krishnamoorthy

data supported a distinct association of schizophrenic disorders with epilepsy, particularly with seizures emanating from the temporal limbic system. Manchanda et al. (1996) studied 300 consecutive patients refractive to treatment, admitted for evaluation of their candidature for epilepsy surgery over a 6-year period. Of these, 231 had a temporal lobe focus, 43 had a nontemporal lobe focus and 26 had generalized and multifocal seizure onset; 142 (47.3%) emerged as psychiatric cases based on DSM–III–R criteria. A principal Axis I diagnosis was made in 88 (29.3%). Anxiety disorders (10.7%) and schizophrenia (4.3%) were the most common Axis I diagnoses. Dependent and avoidant personality traits were frequent (18%) although patients rarely fulfilled criteria for a personality disorder. Are psychiatric disorders commoner in epilepsy? This question needs to be addressed from a public health perspective. Were psychiatric disorders to be commoner in patients with epilepsy, specific mental health resources would need to be created in the community for this patient group. On the other hand were there no excess in psychiatric comorbidity, when patients with epilepsy were compared with other illness groups, matched for age, sex and disability, and normal controls, such resources would not be required. Here we shall examine the evidence, to see if depression and psychosis are commoner in epilepsy. A majority of studies has shown depression to be common in epilepsy. Many of these have employed the Minnesota Multiphasic Personality Inventory (MMPI). Whitman et al. (1984) used a MMPI sequential diagnostic system (Goldberg, 1972) to reanalyse 87 published profiles of patients with epilepsy, other neurological disorders and chronic physical illnesses, encompassing a total of 2786 patients. This included 10 studies of epilepsy encompassing a total of 809 subjects. They found that patients with epilepsy were at higher risk of psychopathology than normal controls. However, no difference was found between people with epilepsy and those with other chronic disorders, or between people with TLE and those with generalized epilepsy. A similar investigation was also reported by Dodrill and Batzel (1986), who found that patients with epilepsy demonstrated more psychopathology than normal controls and patients with other neurological disorders, but that there were no differences in rates of psychopathology between TLE and other forms of epilepsy. Investigations using other instruments such as the Present State Examination have also shown a higher prevalence of depression in epilepsy (Standage and Fenton, 1975). However, other investigations have failed to demonstrate an increased prevalence of depression in epilepsy. For a review of these studies and a discussion of the phenomenology of depression in epilepsy see Lambert and Robertson (1999).

11

Classification

Of all the different psychiatric disorders in epilepsy, it is psychosis for which there is considerable evidence of overrepresentation. The prevalence of psychosis in epilepsy is reported to be in the order of 4% (see Manchanda et al., 1996 for example), sometimes rising as high as 10%. Psychotic disorders are 10 times more common in epilepsy than in the general population, and this is borne out in welldesigned population-based cohort and case-control studies, reviewed here. For a detailed review of studies of psychosis and of the nature and phenomenology of the epileptic psychosis, see Trimble (1991). Another reason for the difficulty in answering this question is the selection bias in most studies mentioned above. In this chapter we have deliberately concentrated on population-based studies. Reviewing these (Table 2.1) it is apparent that there is a considerable degree of psychiatric comorbidity in epilepsy, even in well-designed cohort studies (see Jalava and Sillanpaa, 1996 for example). However, while the evidence for a higher prevalence of psychotic disorders stands out, both in cohort studies (Bredjkaer et al., 1998) and nested case-control studies (Stefannson et al., 1998), the evidence for other psychiatric disorders, while present (see Jalava and Sillanpaa, 1996 for example), is contradictory and not as compelling. The classification of psychiatric disorders in epilepsy The classification of psychiatric disorders in epilepsy has always been controversial. There are two main schools of thought. The first is that the existing systems of classification in psychiatry, the current being the ICD–10 and DSM–IV, in its fourth edition, have made adequate provision for ‘organic’ conditions like epilepsy, and further subsystems of classification would only add to their complexity. The second, most often voiced by neuropsychiatrists with an interest in epilepsy, is that the existing systems of classification are hopelessly inadequate as far as neurological disorders in general and epilepsy specifically are concerned (Krishnamoorthy, in press). One recurrent theme in reviewing the literature about psychiatric disorders in epilepsy is that the failure to identify an excess of psychopathology is due more to the instruments used (generic to mental disorder and not specific to mental disorders in epilepsy), rather than a true finding. It was this that led Bear and Fedio (1977) to develop their own instrument, and conduct a study of psychopathology in patients with temporal lobe epilepsy. The traits that they looked for were those identified by Gastaut, and later Geschwind, who described the constellation of personality traits that characterize patients with temporal lobe epilepsy, including hypergraphia, hyposexuality, religiosity and emotional viscosity. The study by Bear and Fedio (1977) showed that while the MMPI failed to identify differences between patients with TLE and other patient groups, the differences were all too apparent when the responses to the instrument they developed were

12

E.S. Krishnamoorthy

Table 2.1. Important epidemiological studies of neuropsychiatric comorbidity in epilepsy

Investigators (country)

Results

Comments

1960

Pond and Bidwell (UK)

29% of 245 patients had significant morbidity

Study in 14 doctor’s surgeries Conducted by psychiatric social worker Instruments not standardized

1966

Gudmundsson (Iceland)

52% of 987 patients had personality changes

Personal survey by expert Instruments and diagnosis not standardized

1987

Edeh and Toone (UK)

48% of 88 patients emerged as cases

Primary care-based Sophisticated case ascertainment Standard instruments but not epilepsy-specific

1996

Cockerell et al. (UK)

64 incident cases of acute psychological disorder on AED institution

Nation-wide survey Relied on reporting system Crude data on incidence – cannot be generalized

1996

Jalava and Sillanpaa (Finland)

Patients with epilepsy – fourfold risk of somatic, psychosomatic and/ or psychiatric disorder in combination compared with population-based controls Results related to epilepsy and not antiepileptic drug administration

Prospective cohort study with 35year follow-up (only cohort study to date) Results clearly indicate that subjects with epilepsy are at higher risk of developing comorbid psychiatric illness

1998

Bredjkaer et al. (Denmark)

Incidence of schizophrenia spectrum psychoses significantly increased for both men and women with epilepsy Standardized incidence ratio for the entire schizophrenia spectrum (P⬍10–8), nonaffective psychosis (P⬍10–8) and schizophrenia alone (P⬍0.0001)

Record linkage study between a sample of people with epilepsy from the National Patient Register and the Danish Psychiatric Register Enabled the calculation of sophisticated epidemiological indices not estimated in previous studies

1998

Stefansson et al. (Iceland)

Psychiatric diagnosis in 35% of 241 epilepsy cases as compared with 30% of controls, the difference not being statistically significant Significantly higher rate of schizophrenia among men

Patients with epilepsy, and controls with other somatic diseases, both groups being of normal intelligence drawn from a disability register of the State Social Security Institute

Year

13

Classification

analysed. A number of studies have been conducted since then with conflicting results (for reviews Shetty and Trimble, 1997; Trimble, 1991). More recently, Blumer (1995, 2000) has drawn attention to the mood disorder that is seen in patients with refractory epilepsy, particularly TLE that may occur in conjunction with these personality traits. This interictal dysphoric disorder (IDD) of epilepsy is described as being polymorphic, and characterized by a constellation of eight symptoms, typically of short duration and occurring in different permutations and combinations at different times. Blumer argues that the personality traits seen in these patients, such as a serious demeanour, deliberate speech, an ethical and spiritual orientation, especially when subtle, can be positive attributes. However, when the IDD coexists, there may be a paroxysmal venting of angry affects not normally characteristic of the person, followed by a sense of remorse. This he contends can be the source of significant disability. A modified version of the Bear–Fedio scale, the Neurobehavioural Inventory (NBI; Blumer, 1995) is reportedly sensitive to this disorder, helping to identify some of its typical features. However, as the psychometric properties of this instrument have not been tested, and population-based studies have not been carried out, the validity of this diagnosis has not been established. Other differences have also been reported to characterize psychopathology in epilepsy and to differentiate it from psychopathology in general. The interictal psychosis of epilepsy is reported in many studies to be characterized by the preservation of affect, religiosity and paranoid ideation, rather than the undifferentiated, or hebephrenic picture seen in schizophrenia. Slater first observed this in his landmark study at the National Hospital, Queen Square (Slater and Beard, 1963). Subsequently, many others have commented on these findings (Trimble, 1991) and indeed, they have to some extent been borne out in the population-based studies reviewed here (Stefannson et al., 1998). Further, while psychiatric classifications such as the ICD and DSM tend to subsume epilepsy under the ‘organic’ umbrella, thus limiting the ability of such classification to be versatile, and clinically relevant, current classificatory systems in epilepsy pay little or no attention to psychopathology. Besides, there is no place in existing systems of classification for the psychiatric disorders that are reported to be specific to epilepsy. A distinct classificatory system enables clearer phenomenological descriptions of these disorders. It also lends itself more easily to empirical testing. A convincing argument can therefore be made for a distinct classificatory system, and this is currently the subject of a discussion document of the ILAE Commission on Epilepsy and Psychobiology (Krishnamoorthy et al., in preparation). As reviewed here, the neuropsychiatric disorders specific to epilepsy comprise the gamut of neuropsychiatry. Included are the so-called organic mental disorders,

14

E.S. Krishnamoorthy

such as postictal confusional states and complex partial status with psychopathological manifestations; personality changes (the Gastaut–Geschwind syndrome of temporal lobe epilepsy, and the labile personality of juvenile myoclonic epilepsy); a spectrum of psychoses with varying intensity, features and manifestations depending on the temporal relationship with seizure(s) (Trimble, 1991); and a spectrum of neuroses with predominantly affective features (Blumer 1995, 2000). These disorders are also inexorably linked to their relationship with the seizure(s) per se (preictal, postictal, interictal and perhaps periictal); their relationship to the EEG (for example forced normalization of Landolt and alternative psychosis of Tellenbach; Krishnamoorthy and Trimble, 1999); and their relationship to antiepileptic drug (AED) therapy (the AED-induced neuropsychiatric disorders; Trimble, 1998). The inclusion of nonepileptic attack disorder (NEAD) in such a classification is rather more controversial, as there is a growing understanding that NEAD is, in a number of people, the manifestation of a much wider spectrum of psychopathology than that specific to epilepsy (Brown and Trimble, 2000). Any classificatory system, to be viable, will need to take all these factors into consideration. Further, it is important to acknowledge that patients with epilepsy could like anyone, especially patients with chronic medical conditions, have comorbid psychiatric disorders that match existing descriptions in the ICD–10 and DSM–IV. It would serve little purpose to try and reclassify these disorders when associated with epilepsy. The judgement about whether to record the illness in a given patient as a comorbid disorder or as an epilepsy-specific disorder would be best left to the clinician dealing with that individual case. It also goes without saying that such a classificatory system should link closely with the ILAE Classification of Epilepsies and Epileptic Syndromes. While there is little doubt that classifications grounded in aetiology and pathophysiology are an ideal that we must aspire for in the long term, our understanding of causation and its mechanisms in psychiatry, even the neuropsychiatry of epilepsy, is fairly rudimentary, and much ground needs to be covered before we can move with any certainty towards an aetiological model. Aetiologically based systems of classification also require specialized knowledge and access to supportive investigative techniques. Both of these are unavailable in a number of settings, particularly in the developing world. Thus, at least at present, classificatory systems that aim to be culture-free and acceptable across the board, would do well to adopt a descriptive approach, based on a good history and clear clinical descriptions. Such descriptive approaches mirror good clinical practice around the world, and make few demands in terms of specialist expertise or investigation. An ideal classification would be one that is simple, user-friendly, grounded in good history taking and concise clinical descriptions, links with existing systems of classification both in epilepsy and psychiatry, and one which is applicable across

15

Classification

the board. Such a classificatory system will be helped by its application in good prospective population-based research leading eventually to its being operational and valid in the years to come. Undoubtedly, as our understanding of aetiology and pathophysiology in epilepsy and neuropsychiatry improves, one would expect a radical alteration in such descriptive classificatory systems, as we are beginning to see in other neurological and indeed neuropsychiatric disorders. Until then, however, it would perhaps be wise for us to continue in the classical descriptive tradition of good clinical medicine. Conclusions • Psychiatric disorders are common in epilepsy, and encompass the spectrum of conditions, from those that are a direct consequence of epileptogenic activity, to others that are simply comorbid. • There is considerable evidence from epidemiological research to suggest that the psychoses are greatly overrepresented in epilepsy; the evidence for an overrepresentation of other psychiatric disorders is less compelling. • While hospital-based data indicate the presence of epilepsy-specific psychopathology, this has never been examined in the epidemiological setting. Further, instruments such as the NBI that are supposedly sensitive to epilepsy-specific psychopathology have not been validated in this setting. • Existing classificatory systems, both in epilepsy and in mental health, are inadequate, with regard to neuropsychiatric disorders in epilepsy. Efforts to build consensus in this regard, and to develop a user-friendly classificatory system are currently being made by the ILAE Commission on Epilepsy and Psychobiology. • To be relevant and applicable, a classificatory system must not be demanding in terms of specialist knowledge or investigation. A descriptive approach that mirrors good clinical practice is therefore recommended. Such a classificatory system must also link in with existing systems in epilepsy and mental health. • Systematic population-based research using reliable methods of ascertainment, and controls matched for age, sex, disability and ethnicity, based on the ILAE classification of neuropsychiatric disorders in epilepsy, with long-term follow-up of such cohorts, must be conducted in the future.

R E F E R E N C ES American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (Fourth Revision) DSM–IV. Washington, DC: American Psychiatric Association.

16

E.S. Krishnamoorthy Bear, D.M. and Fedio, P. (1977). Quantitative analysis of interictal behaviour in temporal lobe epilepsy. Arch Neurol, 34, 454–67. Blumer, D. (1995). Personality disorders in epilepsy. In Neuropsychiatry of Personality Disorders, ed. J.J. Ratey, pp. 230–63. Boston: Blackwell Science. Blumer, D. (2000). Dysphoric disorders and paroxysmal effects: recognition and treatment of epilepsy-related psychiatric disorders. Harv Rev Psychiatry, 8, 8–17. Bredkjaer, S.R., Mortensen, P.B. and Parnas, J. (1998). Epilepsy and non-organic non-affective psychosis: National Epidemiologic Study. Br J Psychiatry, 172, 235–8. Brown, R.J. and Trimble, M.R. (2000). Dissociative psychopathology, non-epileptic seizures and neurology (Editorial). J Neurol Neurosurg Psychiatry, 69, 285–91. Cockerell, O.C., Moriarty, J., Trimble, M.R., Sander, J.W.A.S. and Shorvon, S.D. (1996). Acute psychological disorders in patients with epilepsy: a nation-wide study. Epilepsy Res, 25, 119–31. Commission on Classification and Terminology of the ILAE (1989). Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia, 30, 389–99. Currie, S., Heathfield, K.W.G., Henson, R.A. and Scott, D.F. (1971). Clinical course and prognosis of temporal lobe epilepsy – a survey of 666 patients. Brain, 94, 173–90. Dodrill, C.B. and Batzell, L.W. (1986). Interictal behavioural features of patients with epilepsy. Epilepsia, 27 (Suppl. 2), S64–S76. Edeh, J. and Toone, B. (1987). Relationship between interictal psychopathology and the type of epilepsy. Br J Psychiatry, 151, 95–101. Goldberg, L.R. (1972). Man versus mean: the exploitation of group profiles for the construction of diagnostic classification systems. J Abnorm Psychol, 79, 121–31. Gudmundsson, G. (1966). Epilepsy in Iceland – a clinical and epidemiological investigation. Acta Neurol Scand, Suppl 25, 43, 64–90. Gureje, O. (1991). Interictal psychopathology in epilepsy – prevalence and pattern in a Nigerian clinic. Br J Psychiatry, 158, 700–5. Hauser, A.W. (1998). Incidence and prevalence. In Epilepsy – A Comprehensive Textbook, ed. J.R. Engel and T.A. Pedley TA. New York: Lippincott-Raven. Jalava, M. and Sillanpaa, M. (1996). Concurrent illnesses in adults with childhood-onset epilepsy: a population-based 35-year follow-up study. Epilepsia, 37, 1155–63. Krishnamoorthy, E.S. and Trimble, M.R. (1999). Forced normalization – clinical and therapeutic relevance. Epilepsia, 40 (Suppl. 10), S57–64. Krishnamoorthy, E.S. (2000). An approach to classifying neuropsychiatric disorders in epilepsy. Editorial. Epilepsy Behav, 1, 373–7. Krishnamoorthy, E.S. (2001). Psychiatric issues in epilepsy. Curr Opinion Neurol, 14, 217–24. Lambert, M. and Robertson, M.M. (1999). Depression in epilepsy. Etiology, phenomenology, and treatment. Epilepsia, 40 (Suppl. 10), S21–47. Lewis, G., Pelosi, A.J., Araya, R. and Dunn, G. (1992). Measuring psychiatric disorder in the community: a standardised assessment for use by lay interviewers. Psychol Med, 22, 465–86. Manchanda, R., Schaefer, B., McLachlan, R. et al. (1996). Psychiatric disorders in candidates for surgery for epilepsy. J Neurol Neurosurg Psychiatry, 61, 82–9. Mendez, M.F., Grau, R., Doss, R.C. and Taylor, J.L. (1993). Schizophrenia in epilepsy: seizure and psychosis variables. Neurology, 43, 1073–7.

17

Classification Pond, D.A. and Bidwell, B.H. (1959/60). A survey of epilepsy in fourteen general practices II. Social and psychological aspects. Epilepsia, 1, 285–99. Schmitz, B. and Trimble, M.R. (1992). Epileptic equivalents in psychiatry: some 19th century views. Acta Neurol Scand, 86 (Suppl. 140), 122–6. Shetty, T. and Trimble, M.R. (1997). The Bear Fedio Inventory: twenty years on. J Epilepsy, 10, 254–62. Slater, E. and Beard, A.W. (1963). The schizophrenia-like psychoses of epilepsy. Br J Psychiatry, 109, 95–150. Smith, D.B., Craft, B.R., Collins, J., Mattson, R.H., Cramer, J.A. and the VA Co-operative Study Group 118 (1986). Behavioural characteristics of epilepsy patients compared with normal controls. Epilepsia, 27, 760–8. Standage, K.F. and Fenton, G.W. (1975). Psychiatric symptom profiles of patients with epilepsy: a controlled investigation. Psychol Med, 5, 152–60. Stefansson, S.B., Olafsson, E. and Hauser, A.W. (1998). Psychiatric morbidity in epilepsy: a case controlled study of adults receiving disability benefits. J Neurol Neurosurg Psychiatry, 64, 238–41. Trimble, M.R. (1991). The Psychoses of Epilepsy. New York: Raven Press. Trimble, M.R. (1998). New antiepileptic drugs and psychopathology. Neuropsychobiology, 38, 149–51. Whitman, S., Hermann, B.P. and Gordon, A.C. (1984). Psychopathology in epilepsy: how great is the risk? Biol Psychiatry, 19, 213–36. World Health Organisation (1992). The International Classification of Mental and Behavioural Disorders, Tenth Edition (ICD–10). Geneva: WHO.

3

Limbic connectivity: anatomical substrates of behavioural disturbances in epilepsy J. Engel Jr, C. Wilson and F. Lopez-Rodriguez UCLA School of Medicine, Los Angeles, USA

Introduction Behavioural disturbances associated with epilepsy are due, in part, to a number of important psychological and social factors; however, it is useful to acknowledge that there are neurobiological factors as well. Not only does identification of organic bases for behavioural aberrations help reduce the stigma they invariably engender, but also it provides a rational scientific approach to prevent or reverse a significant cause of disability experienced by persons with epilepsy. Cognitive and psychiatric disturbances associated with epilepsy may be due to the same underlying pathological process that causes the epileptic condition, and there is evidence that behavioural impairment may be related to the pathophysiological nature of the underlying lesion, its effect on development and its location (Engel et al., 1986; Engel and Shewmon, 1991). With respect to location, mesial temporal and other limbic epileptogenic lesions are most likely to be associated with behavioural disturbances (Engel et al., 1986; Engel and Shewmon, 1991). Temporal lobe epilepsy is the most common form of human epilepsy, and hippocampal sclerosis is the most common human epileptogenic lesion (Engel, 1998). Furthermore, mesial temporal lobe epilepsy with hippocampal sclerosis may be the most refractory to antiepileptic drugs (Engel, 1998). Consequently, limbic system dysfunction is undoubtedly the most important cause of behavioural disturbances associated with epilepsy. Epileptogenicity itself is known to cause enduring changes in brain function and structure in a number of ways which can result in altered behaviour. For instance, postnatally or in early childhood, abnormal recurrent synaptic bombardment of distant projection areas by repeated ictal and interictal epileptiform discharges can adversely affect the development of normal neuronal integration (Engel and Shewmon, 1991; Morrell et al., 1995). In the mature brain such abnormal synaptic activity can induce plastic changes through kindling mechanisms (Corcoran and Moshé, 1998; Engel et al., 1986), and the naturally occur18

19

Anatomical substrates of behavioural disturbances

ring homeostatic seizure-suppressing mechanisms which maintain the interictal state may have adverse consequences on normal neuronal function (Engel et al., 1986; Engel and Shewmon, 1991). In addition, epileptic seizures are known to disrupt sleep patterns (Shouse et al., 1997) and endocrine function (Herzog, 1997), which can alter behaviour; and severe epileptic activity results in neuronal cell death and consequent synaptic reorganization (Sloviter and Damiano, 1981). Given the importance of the limbic system in the development and maintenance of cognition, emotion, motivation and other essential behaviours, it is not surprising that epileptogenic lesions capable of disrupting limbic connectivity are associated with a great variety of clinically significant behavioural abnormalities. The limbic system The term ‘limbic system’ was coined by Maclean (1952) to designate a series of structures originally delineated by Broca (1878) and Papez (1937), plus the amygdala and its connections, which were believed to play a crucial role in mediating the exchange of information between the thinking brain (cortex) and the more primal animal brain (diencephalon and brain stem). Broca first described the limbic lobe as the area of brain making up the rim of the cortex, including the hippocampus, cingulate cortex and certain frontal lobe structures. Papez more precisely defined circuits that connected these structures, including the septal area, mamillary bodies, and parts of the thalamus and hypothalamus. The addition of the amygdala and its connections by Maclean provided the functional substrate for human emotional behaviours now attributed to the limbic system (Figure 3.1). The hippocampus, amygdala and their projection fields are central to the concept of the limbic system, and also are the most epileptogenic regions of the mammalian brain. The perirhinal and piriform regions of the parahippocampal cortex, along with the central nucleus of the amygdala, are the most rapidly kindled structures in the brain (McIntyre and Plant, 1993), and an area of deep piriform complex, area tempesta, has the lowest threshold in the brain to seizures induced by local application of antagonists to gamma amino butyric acid (GABA), the primary inhibitory neurotransmitter in the brain (Piredda and Gale, 1985). The limbic system, therefore, is uniquely susceptible to the development of epileptiform abnormalities, not only as a result of hippocampal sclerosis, but due to any irritating disturbance. Furthermore, distant epileptogenic lesions which preferentially project to mesial temporal structures also have a high propensity to induce mesial temporal epileptiform activity resulting in both temporal lobe seizures and their behavioural consequences (Williamson and Engel, 1997).

20

J. Engel Jr et al.

SMA

Cingulate

Frontomedial

Fornix

Anterior cingulate Thalamus Fronto-lateral

Hypothalamus septum MB

Fronto-orbital

Hipp.

NA

Ento. Parietal

Parahipp.

Post lateral temporal

Mid-brain tegm. reticular formation

Inferotemporal and anterior temporal

Insula

CC

VHC

DHC

AC

Contralateral hippocampal formation Second generalization

Figure 3.1.

Diagrammatic representation of the human limbic system and its afferent and efferent connections. Note the major connections between the hippocampus, entorhinal cortex and amygdala (NA), and the components of the Papez circuit, described in the text. SMA, supplementary motor area; HIPP, hippocampus; ENTO, entorhinal cortex; PARAHIPP, parahippocampal gyrus; CC, corpus callosum; VHC, ventral hippocampal commissure; DHC, dorsal hippocampal commissure; AC, anterior commissure. (With permission, adapted from Wieser, 1988.)

21

Anatomical substrates of behavioural disturbances Basket cell Regio superior (CA1) Lateral entorhinal cortex

Schaffer collateral

Perforant path

To septum Regio inferior (CA3)

Figure 3.2.

Mossy fibres

Medial entorhinal cortex Buried blade of fascia dentata

Exposed blade of fascia dentata

Illustration of the intrinsic circuitry of the hippocampal formation. The primary focus of this illustration is the trisynaptic circuit. The first synapse consists of the mossy fibre to CA3 pyramidal cell connection derived from the granule cells of the dentate gyrus. The second synapse is the connection of the CA3 cell output via the Schaffer collaterals to CA1, and the third, not shown, is the CA1 axonal output to subiculum and entorhinal cortex. Entorhinal cortex also provides the initial input to the dentate gyrus. (With permission from O’Keefe and Nadel, 1978.)

Hippocampus Normal structure

The hippocampus is perhaps the most studied region of the mammalian brain, due in part to its highly organized laminar structure, and its propensity for plastic change, such as long-term potentiation and long-term depression, which may underlie mechanisms of normal learning and memory (Gloor, 1997; Schwartzkroin and McIntyre, 1997). The hippocampal formation consists of the dentate gyrus and the hippocampus proper or cornu ammonis (CA), which is divided into CA1, CA2 and CA3 (Figure 3.2). The major excitatory hippocampal input is glutaminergic from the entorhinal cortex via the perforant path, which terminates on the distal two-thirds of granule cell dendrites in the dentate gyrus. The medial portion of the entorhinal cortex projects predominantly to the middle third

22

J. Engel Jr et al.

of the granule cell dendritic layer (middle molecular layer), while the lateral entorhinal cortex projections, which also contain opioids, terminate on the distal third of granule cell dendrites (outer molecular layer). These latter projections may play a role in epileptogenicity due to the opioid disinhibitory effect on granule cell excitability (Xie et al., 1992). The neuronal structure of the dentate gyrus provides the substrates for strong feed-forward and feedback inhibition which resists the propagation of epileptiform activity, leading to its designation as the ‘dentate gate’ (Lothman et al., 1991). The classical excitatory trisynaptic pathway through the hippocampus is glutamatergic and utilizes NMDA and non-NMDA receptors, while inhibitory interneurons are GABAergic. The pathway begins with the dentate granule cell axons, called mossy fibres, which project through the dentate hilus to area CA3. Mossy fibres contain two types of terminals: the first type are the large, complex mossy terminals which give the fibres their name, and the second type are small terminals and filopodia extending from the mossy terminals. The primary target of the mossy terminals are the CA3 pyramidal cells, with which they form multiple excitatory synapses, while the primary target of the small terminals and filopodia are GABAergic interneurons. The ratio of dentate granule axon terminals contacting GABAergic interneurons to those terminals contacting pyramidal and mossy cells (excitatory neurons in the dentate hilus) has been estimated at approximately 5:1, which means that the major postsynaptic targets of granule cells are inhibitory neurons (Acsády et al., 1998). Although the large proportion of granule cell output which activates interneurons has a significant inhibitory effect, the mossy terminals exciting CA3 pyramidal cells are quite powerful. Because CA3 pyramidal cell intrinsic connections support recurrent excitation, focused discharge of specific CA3 pyramidal cells can occur. Excitation is further supported by granule cell activation of mossy cells in the dentate hilus, which in turn project back onto the proximal one-third of the granule cell dendrites (inner molecular layer). This excitatory feedback loop is offset by similar inhibitory feedback loops via a variety of hilar inhibitory neurons. CA3 pyramidal cells have a unique propensity to burst synchronously, thus excitatory epileptiform input that traverses the dentate gate activates synchronously bursting pacemaker cells in CA3, giving rise to interictal spikes (Wong and Traub, 1983). The second segment of the hippocampal trisynaptic pathway consists of the Schaffer collaterals, CA3 pyramidal cell axons which terminate on CA1 pyramidal neurons. Under epileptogenic conditions, synchronous bursting in CA3 can produce continuous discharges in CA1, the usual site of ictal onset in the hippocampal slice preparation. CA2 is a transition zone that does not appear to participate in the transmission of epileptiform discharges in the nonsclerotic hippocampus. However, pyramidal cells in this region are relatively spared in the process of neuronal loss that characterizes hippocampal sclerosis (Mathern et al.,

23

Anatomical substrates of behavioural disturbances

1997), and these remaining neurons may play a more important role in mesial temporal lobe epilepsy (Williamson and Spencer, 1994). The third segment of the hippocampal trisynaptic pathway is its primary output, from CA1 to the adjacent subiculum. The subiculum in turn projects back to the entorhinal cortex, completing a closed hippocampal excitatory loop. This pathway is not absolutely unidirectional, however; there are also significant connections from entorhinal cortex to CA1, and some fibres even reach CA3. The subiculum predominantly projects to perirhinal and piriform cortex, but also to the mamillary bodies. The perirhinal cortex has connections with frontal motor cortical areas, whereas the piriform cortex projects reciprocally to olfactory areas and the brain stem. These projections are predominantly responsible for mediating the clinical manifestations of temporal lobe seizures; ictal discharges confined to the hippocampus may have no overt signs and symptoms (Sperling and O’Connor, 1990). Parahippocampal areas may also become an important site of seizure generation; success in eliminating seizures by amygdalohippocampectomy, performed to treat medically intractable temporal lobe epilepsy, appears to depend upon the amount of parahippocampal tissue removed (Siegel et al., 1990). CA1 axons also project via the fimbria and fornix along the traditional Papez circuit to the septal area, midline thalamus, amygdala, hypothalamus and brain stem autonomic centres. The amygdala, hypothalamic and brain stem projections mediate aspects of emotional behaviours, and the hippocampus may play a role in the adreno–hypothalamic–pituitary axis, in view of the fact that adrenal steroid receptors are abundant in the hippocampus (Gould et al., 1991). The hippocampus also receives direct cholinergic and GABAergic input from the septal area, which mediates the classical hippocampal theta rhythm, and these projections may have disinhibitory as well as inhibitory influences on epileptic activity. With respect to brain stem afferents, direct noradrenergic input to the hippocampus from the locus coeruleus can be excitatory (Madison and Nicoll, 1986), while serotonergic input from the raphé nucleus and dopaminergic input from the substantia nigra and ventral tegmental area are predominantly inhibitory (Andrade and Nicoll, 1987), but biogenic amine effects can be varied, depending on the postsynaptic target. Indirect afferent influences on the hippocampus, via entorhinal cortex, derive from a large area of neocortex, including frontal limbic regions responsible for goal-directed behaviour, cingulate cortex influencing memory, all sensory cortical areas for integration of polymodal sensory information, and perirhinal and piriform cortex, as well as from amygdala. Hippocampal sclerosis

Much is known about the structural and functional abnormalities that exist in the epileptic hippocampus, due in part to the ready availability of human tissue from

24

J. Engel Jr et al.

epilepsy surgery programmes, and in part to numerous animal models of this condition; however, the cause of these abnormalities, and the reasons why these abnormalities should result in spontaneous seizures, are not well understood. The characteristic finding in hippocampal sclerosis is cell loss, most prominent in the dentate hilus and CA1, but it is also marked in CA3 and, to a lesser extent, the dentate gyrus and subiculum (Mathern et al., 1997). Although specific hilar interneurons are lost, namely those containing somatostatin and neuropeptide Y, in general there is relative preservation of hippocampal inhibitory interneurons, compared to principal cells (Babb, 1992). The greatest proportion of hilar cell loss is from mossy cell death, which results in loss of excitatory inputs to dentate granule cell proximal dendrites, and consequent mossy fibre sprouting. These collateral granule cell axons, which can easily be seen with Timm’s stain, reinnervate the inner molecular layer, but most likely end on inhibitory interneurons, as well as on granule cell dendrites (Figure 3.3). There is also evidence of inhibitory interneuron sprouting (Babb, 1992). Electrophysiological studies in patients have confirmed that inhibition is actually increased, interictally, in the epileptiform sclerotic hippocampus (Wilson et al., 1998). This enhanced inhibition may act to suppress seizure generation; however, it has been postulated that such enhanced inhibition together with enhanced excitation, brought about by pathological neuronal reorganization, results in a predisposition to abnormal neuronal hypersynchronization (Figure 3.4). In addition, single input fibres sprout to innervate wider targets, and reduction of the dendritic domain places more excitatory input closer to the soma and axon hillock. Hypersynchronization underlies interictal epileptiform EEG spikes, but is also the hallmark of the typically hypersynchronous hippocampal ictal EEG onset (Velasco et al., 2000). Consequently, enhanced inhibition in the sclerotic hippocampus may contribute to its epileptogenicity (Engel et al., 1997). During the process of epileptogenesis, once the local anatomical substrate for abnormal hypersynchrony has been established, these hypersynchronous discharges may act to kindle distant structures responsible for the manifestation of ictal behaviour. Amygdala The amygdala has been implicated in a number of important primal functions, including fear, aggression, learning, reproduction and appetitive drive. In contrast to the hippocampus, it consists of a number of loosely organized subnuclei with diverse neurons and cytoarchitecture, which have been divided into an olfactory group, a centromedial group and a basolateral group (Gloor, 1997; Schwartzkroin and McIntyre, 1997). The olfactory group has predominant reciprocal connections with the hippocampus, piriform cortex and olfactory bulb, with less strong connections to the septal area, brain stem nuclei, hypothalamus, thalamus and the

25

Anatomical substrates of behavioural disturbances

Normal

FD

MML

IML

SG

Hilus

OML

Mossy fibre reinnervation

FD

CML

Figure 3.3.

MML

IML

SG

Hilus

Schematic diagram showing the reciprocal innervation between principal neurons of the hilus (squares), and dentate granule cells of the hippocampal formation (circles). In the normal situation shown above, granule cell axons (mossy fibres) synapse on principal neurons of the hilus and CA3, while hilar neurons make synapses on proximal dendrites of the granule cells. The reinnervation that occurs with neuronal loss, such as occurs with hippocampal sclerosis, is shown below. When hilar input to granule cells is decreased, the mossy fibres of surviving granule cells sprout and make contact with the vacated postsynaptic membrane, creating monosynaptic recurrent excitatory circuits. FD, fascia dentata; OML, outer molecular layer; MML, middle molecular layer; IML, inner molecular layer; SG, stratum granulosum. (With permission from Engel, 1990.)

26

J. Engel Jr et al.

Figure 3.4.

Reciprocal innervation of a hypothetical neuronal system is shown schematically above. This is the typical synaptic organization of neocortex and hippocampus. Excitatory afferent input terminates on the dendrites of principal neurons (filled triangles). Axon collaterals from these principal neurons terminate on inhibitory interneurons (empty circles), which in turn make hyperpolarizing synapses on the soma of the same, and adjacent (not shown), principal neurons. With cell loss, a number of synaptic reorganizations are likely to occur, as shown schematically on the right. Fewer afferent input fibres sprout to innervate more principal neurons, predisposing to hypersynchronization. Because the dendrites of principal neurons are shorter, these excitatory influences are closer to the axon hillock and more likely to induce neuronal firing. Neuronal excitability is further increased by the establishment of monosynaptic excitatory recurrent circuits, as also shown in Figure 3.3. Although not definitively demonstrated, it is possible that inhibitory interneurons sprout terminals to produce more powerful, and/or more extensive, recurrent inhibitory influences, further enhancing the potential for hypersynchronization. (With permission from Engel, 1990.)

nucleus accumbens of the basal forebrain. The centromedial group has reciprocal connections with extensive areas of the brain stem, including the periaqueductal grey, but also with the entorhinal cortex, hippocampus, hypothalamus, thalamus and striatum. The basolateral group has strong reciprocal connections with somatosensory and motor cortex, with only weak brain stem connections. It is likely, therefore, that the centromedial nuclei, particularly the central nucleus of the amygdala, are most important in mediating affective ictal and interictal symptoms observed with amygdala involvement in the propagation of epileptiform discharges.

27

Anatomical substrates of behavioural disturbances

Interhemispheric limbic connections Due to the enormous expansion of the human cerebral cortex in comparison with that of subprimate mammals, the limbic structures, with the exception of the cingulate gyrus, have moved far from the midline. Interhemispheric connections are necessarily increased in length and changed in trajectory by this major reorganization (Wilson, 1995). Three pathways exist for interhemispheric communication between limbic areas: corpus callosum, anterior commissure and hippocampal commissure. Corpus callosum

The corpus callosum carries fibres from the cingulate gyrus, but also from the medial orbital frontal cortex, the posterior two-thirds of the temporal cortex and the posterior parahippocampal gyrus (Demeter et al., 1990). As a component of the Papez circuit, cingulate cortex is in a position to relay output activity from limbic structures to the contralateral hemisphere, as well as back to the ipsilateral entorhinal cortex. The callosal connections of the orbital frontal cortex are significant for interhemispheric limbic interaction because of the major projection this area receives from the amygdala (Amaral et al., 1984). Anterior commissure

In primates, the anterior commissure carries fibres from olfactory and other areas of the orbital frontal cortex, piriform cortex, the anterior third of the temporal cortex and posterior parahippocampal gyrus, and a few sparse fibres from the cortical and deep subnuclei of the amygdala. For the most part, the widespread cortical projections of the amygdala are ipsilateral only, and the small interhemispheric projection which exists terminates in contralateral temporal cortex, not in contralateral amygdala (Demeter et al., 1990). Hippocampal commissure

Phylogenetically, commissural fibres have constituted a major group of extrinsic connections joining the two hippocampi and their subfields, but the connections are substantially reduced in the human brain. There are two components of the interhemispheric pathway: the ventral hippocampal commissure, which carries fibres from the hippocampal formation, and the dorsal hippocampal commissure, which carries fibres from the entorhinal cortex, presubiculum and parahippocampal gyrus. In the subhuman primate, the ventral hippocampal commissure is greatly reduced and limited to the anterior hippocampus, while the dorsal commissure carries primarily presubicular and parahippocampal gyrus fibres to the contralateral entorhinal cortex (Amaral et al., 1984). In the human brain, dissection of

28

J. Engel Jr et al.

autopsy specimens has revealed the presence of a dorsal hippocampal commissure, but not the ventral (Gloor et al., 1993), and electrophysiology studies (Wilson et al., 1990, 1991) provide additional functional evidence that hippocampal connections are absent in humans. The relative reduction of commissural interchange from lower animal to primate may provide the basis for the hemispheric specialization unique to the human brain, particularly relating to differences in verbal versus spatial memory performance, which is lateralized to the left and right temporal lobes, respectively (Amaral et al., 1984; Demeter et al., 1985). The absence of any significant interhemispheric amygdala connections suggests that emotion and affect may have similarly lateralized or hemispherically specialized components (Demeter et al., 1990), which has important implications for patients with temporal lobe seizures. Interictal and ictal behavioural and emotional manifestations of mesial temporal lobe epilepsy can be very different depending on whether the epileptogenic abnormality is on the right or left. Possible anatomic substrates of some behavioural disturbances associated with epilepsy The behavioural disturbances most associated with epilepsy include depression and other affective disorders and personality traits, schizophrenia, aggressivity and anxiety. Affective disturbances

The prevalence of affective disorders is higher among epileptic patients than the general population, or neurological patients with the same degree of disability (Mendez et al., 1986). No neurobiological mechanism has successfully explained this correlation between affective disorders and epilepsy. However, several hypotheses have been proposed. Reciprocal connections between limbic structures and those brain stem nuclei with biogenic amine and opioid projections to the forebrain provide ample pathways for the mediation of depression and other abnormalities of affect. In particular, endogenous opioids are known to be elevated during seizures (Vindrola et al., 1981), and mu-opiate-receptor binding is increased in the human epileptic temporal lobe (Frost et al., 1988). It is postulated that opioid peptides act as endogenous euphorogens, which may explain why electroconvulsive treatment is an effective therapy for depression (Kline et al., 1977). Conceivably, patients with frequent limbic seizures develop a dependence on high levels of circulating opioid peptides, and the appearance of affective symptoms reflects a form of withdrawal when seizures do not regularly recur (Engel et al., 1991). Indeed, transient depression lasting

29

Anatomical substrates of behavioural disturbances

a few months is commonly observed in patients who become suddenly seizure-free following successful surgical treatment (Glosser et al., 2000; Krahn et al., 1996). Depression is more common among patients with complex partial seizures than with generalized seizures (Devinsky and Vasquez, 1993). This finding is not surprising, because complex partial seizures originate from those limbic structures in which dysfunction has been strongly correlated with depression. For example, hippocampal damage after stress, induced by elevated levels of cortisol, has been proposed as a possible mechanism for depression (Stokes, 1995). Affective symptoms have been associated with dysfunction in a circuit that travels from the anterior cingulate and medial orbito-frontal cortices to the ventral striatum (nucleus accumbens and olfactory tubercle), the ventral pallidum, the medial dorsal nucleus of the thalamus, and back to the anterior cingulate and medial orbito-frontal cortices (Alexander et al., 1990; Ketter et al., 1996). Abnormalities in the components of this circuit have also been described with complex partial seizures (Bromfield et al., 1992; Savic et al., 1997). A common finding among patients with a primary affective disturbance, as well as in epileptic patients with depression, is decreased metabolic activity of the prefrontal cortex (Bromfield et al., 1992; Rogers et al., 1998). Whether the finding of decreased prefrontal metabolism is a consequence of the depression in patients with complex partial seizures, or alternatively, complex partial seizures have secondarily affected the activity of the prefrontal cortex, with a consequent development of depression, is not clear. However, based on the existence of strong connections between the mesial temporal structures and the prefrontal cortex, this second hypothesis is anatomically and functionally plausible (Drevets, 1998). Abnormalities of metabolism in the amygdala and anterior cingulate have also been repeatedly reported to be associated with depression (Drevets, 1998; Mayberg et al., 1994) and temporal lobe epilepsy (Henry et al., 1993a,b). Therefore, complex partial seizures could induce functional changes in limbic areas that overlap with those structures involved in the manifestations of affective psychopathology. Schizophrenia

The dopaminergic hypothesis of schizophrenia originated initially from two indirect findings: except for clozapine, all effective antipsychotics are blockers of D2 dopamine (DA) receptors, and stimulants that increase DA release in the limbic system frequently induce psychosis (Farde, 1997). This hypothesis has been applied mostly to the positive symptoms of schizophrenia such as hallucinations and delusions, because typical antipsychotic medications, which act almost exclusively on D2 receptors, are not very efficient in treating negative symptoms and thought disorganization. This last fact, together with the delay in action of antipsychotic medications (although D2 antagonism is immediate), put the dopaminergic theory into

30

J. Engel Jr et al.

question (Goldstein and Deutch, 1992). However, recent studies have shown that schizophrenic patients experience increased DA release after amphetamine challenge, when compared with control subjects, and this hyperactivity of DA transmission was found to be associated with activation of psychotic symptomatology (Laruelle and Abi-Dargham, 1999). In cats, kindling of the dopaminergic ventral tegmental area of the brain stem, which projects to the nucleus accumbens of the basal forebrain, does not produce seizures, but alters behaviour in a way that has been interpreted to resemble aspects of schizophrenia (Stevens, 1973). The nucleus accumbens and the ventral tegmental area both receive input from the mesial temporal limbic system, and are thus subject to kindling-like afferent input from seizures originating from the hippocampus or perihippocampal regions. An enduring effect of limbic seizures on dopaminergic function has been demonstrated experimentally by the fact that amygdala kindling in cats enhances methamphetamine-induced stereotypy, suggesting upregulation of DA receptors (Sato, 1983), although the location of these receptors has not been identified. Furthermore, disturbances of limbic structures have been associated with schizophrenia (Weinberger et al., 1992). Pathological studies of brains of schizophrenic patients revealed a disorganization of the typical laminar structure of the hippocampus (Kovelman and Scheibel, 1984). Studies from discordant monozygotic twins (i.e., one twin is affected by schizophrenia whereas the other is healthy), have shown that the schizophrenic twin had smaller hippocampi, and that the size of the hippocampi correlated with hypofrontality (assessed by blood-flow studies during performance of the Wisconsin Card Sorting Test) in the affected twins (Weinberger et al., 1992). This finding is consistent with the possibility that pathological abnormalities in the hippocampus might result in a widespread disruption of other corticolimbic structures, such as the prefrontal cortex, with a consequent development of schizophrenic symptoms. Aggression

Kindling is known to induce postictal hyperreactivity (which has been mistaken for aggression), as well as other postictal behaviours which can all be modulated by pretreatment with opioid agonists and antagonists (Caldecott-Hazard and Engel, 1987). Some of these behaviours may be direct opioid effects, while others may represent withdrawal phenomena. Mesial limbic structures project to the brain stem periaqueductal grey, a major source of forebrain opioid projections, and this area has been implicated in an experimental model of aggression in cats termed ‘defensive rage’. Electrical stimulation of this area, and of the lateral hypothalamus, produces a stereotypic ‘Halloween cat’ response, with piloerection, arched back, pupillary dilation and attack behaviour, lasting only as long as the duration of the

31

Anatomical substrates of behavioural disturbances

stimulus (Bandler, 1988). This behaviour can be modulated by lesions, or stimulation, of hippocampus and amygdala. Production of experimental temporal lobe epilepsy in cats, by hippocampal injection of kainic acid which causes a lesion similar to hippocampal sclerosis, has a profound effect on this behavioural model (Engel et al., 1991; Griffith et al., 1987). Immediately following kainic acid injection, cats experience continuous limbic seizures lasting many hours, and during this time also exhibit spontaneous defensive rage. This is followed by a latent period when the damaged hippocampus undergoes epileptogenic neuronal reorganization and ultimately gives rise to spontaneous limbic seizures. During the latent period, which can last from weeks to months, there are no spontaneous seizures and behaviour is normal. Once spontaneous seizures occur, however, not only is the threshold to electrical stimulation-induced defensive rage lowered, but cats also exhibit defensive rage in response to environmental stimuli, in the absence of periaqueductal grey or lateral hypothalamic electrical stimulation. Furthermore, the defensive rage reaction can be elicited only when the cat is approached contralateral to the epileptic hippocampus (Figure 3.5). Clearly, hippocampal epileptogenesis has induced distant structural changes, presumably involving amygdala, hypothalamus and periaqueductal grey, resulting in enduring aberrant behaviour. These experiments should not be interpreted to suggest that patients with mesial temporal lobe epilepsy are susceptible to defensive rage or other aggressive behaviours. Defensive rage is a hardwired species-specific behaviour of the cat, and the brain stem and diencephalic structures responsible for its manifestation undoubtedly come under strong neocortical control in the human. A variety of human correlates, other than rage, might be anticipated, including impulsivity, insecurity, paranoia and perhaps even depression. Anxiety

The amygdala is believed to provide an affective tone to perceptions and experiences, and human lesion and neuroimaging studies have demonstrated a role for the amygdala in verbal and nonverbal recognition of emotions (Adolphs et al., 1998; Hariri et al., 2000), as well as in fear conditioning (Furmark et al., 1997) and provoked anxiety (Davidson et al., 1999). In animals, there is also extensive evidence to support a role for the amygdala in fear and fear conditioning (LeDoux, 1993). Other structures that have been found to be repeatedly correlated with anxiety are the orbito-frontal cortex, anterior insula and anterior cingulate (Malizia, 1999). The projections of the central nucleus of the amygdala to the locus coeruleus and hypothalamus appear to be responsible for the arousal and autonomic responses associated with fear and anxiety (Goldstein et al., 1996) which, together with the

32

J. Engel Jr et al.

Figure 3.5.

Demonstration of lateralized hypereactivity. Photographs were taken from a videotape recording of a cat with a chronic, kainic acid-induced epileptogenic lesion in the left hippocampus. Top: when the cat was approached on the left (ipsilateral) side, it responded by rubbing up against the offered hand and purring. Bottom: when an attempt was made to touch the cat on the right (contralateral) side, it assumed a defensive posture (ears back, piloerection, and pupillary dilatation), hissed, and struck at the offered hand with the right paw. (With permission from Engel et al., 1991.)

33

Anatomical substrates of behavioural disturbances

well-established function of the amygdala in emotion recognition and fear conditioning, imply a key role of the amygdala in the integration of behavioural and neuroendocrine components of the stress response. Amygdala-kindled rats show behaviours suggestive of increased fear, such as decreased exploration in the open field (Depaulis et al., 1997). Consequently, there are ample anatomical and pathophysiological substrates that can be invoked to explain both ictal and interictal anxiety as well as panic disorder in patients with temporal lobe epilepsy. Acknowledgements Original research reported by the author was supported in part by Grants NS02808, NS-15654, NS-33310, and GM-24839, from the National Institutes of Health, and Contract DE-AC03-76-SF00012 from the Department of Energy.

R E F E R E N C ES Acsády, L., Kamondi, A., Sik, A., Freund, T. and Buzsáki, G. (1998). GABAergic cells are the major postsynaptic targets of mossy fibers in the rat hippocampus. J Neurosci, 18, 3386–403. Adolphs, R., Tranel, D. and Damasio, A.R. (1998). The human amygdala in social judgment [see comments]. Nature, 393, 470–4. Alexander, G.E., Crutcher, M.D. and DeLong, M.R. (1990). Basal ganglia-thalamocortical circuits: parallel substrates for motor, oculomotor, “prefrontal” and “limbic” functions. Prog Brain Res, 85, 119–46. Amaral, D.G., Insausti, R. and Cowan, W.M. (1984). The commissural connections of the monkey hippocampal formation. J Comp Neurol, 224, 307–36. Andrade, R. and Nicoll, R.A. (1987). Pharmacologically distinct actions of serotonin on single pyramidal neurones of the rat hippocampus recorded in vitro. J Physiol, 394, 99–124. Babb, T.L. (1992). GABA neurons, synapses and inhibition in human hippocampal epilepsy. In Epilepsy and Inhibition, ed. E.-J. Speckmann and M.J. Gutnick, pp. 375–97. Munich: Urban and Schwarzenberg Press. Bandler, R. (1988). Brain mechanisms of aggression as revealed by electrical and chemical stimulation: suggestion of a central role for the midbrain periaqueductal gray region. In Progress in Psychobiology, ed. A. Epstein & A. Morrison, pp. 67–153. New York: Academic Press. Broca, P. (1878). Anatomie comparée des circonvolutions cerebrales. Le grand lobe limbique et la scissure limbique dans le serie des mammiferes. Rev Anthropol, Ser 2, 1, 385–498. Bromfield, E.B., Altshuler, L., Leiderman, D.B. et al. (1992). Cerebral metabolism and depression in patients with complex partial seizures. Arch Neurol, 49, 617–23. Caldecott-Hazard, S. and Engel, J. Jr (1987). Limbic postictal events: Anatomical substrates and opioid receptor involvement. Prog Neuropsychopharmacol Biol Psychiatry, 11, 389–418. Corcoran, M.E. and Moshé, S.L. (Eds) (1998). Kindling V. New York: Plenum Press.

34

J. Engel Jr et al. Davidson, R.J., Abercrombie, H., Nitschke, J.B. and Putnam, K. (1999). Regional brain function, emotion and disorders of emotion. Curr Opin Neurobiol, 9, 228–34. Demeter, S., Rosene, D.L. and Van Hoesen, G.W. (1985). Interhemispheric pathways of the hippocampal formation. Presubiculum, entorhinal and posterior hippocampal cortices in the rhesus monkey: the structure and function of the hippocampal commissures. J Comp Neurol, 233, 30–47. Demeter, S., Rosene, D.L. and Van Hoesen, G.W. (1990). Fields of origins and pathways of the interhemispheric commissures in the temporal lobe of macaques. J Comp Neurol, 302, 29. Depaulis, A., Helfer, V., Deransart, C. and Marescaux, C. (1997). Anxiogenic-like consequences in animal models of complex partial seizures. Neurosci Biobehav Rev, 21, 767–74. Devinsky, O. and Vazquez, B. (1993). Behavioral changes associated with epilepsy. Neurol Clin, 11, 127–49. Drevets, W.C. (1998). Functional neuroimaging studies of depression: the anatomy of melancholia. Annu Rev Med, 49, 341–61. Engel, J. Jr (1990). Functional explorations of the human epileptic brain and their therapeutic implications. Electroencephalogr Clin Neurophysiol, 76, 296–316. Engel, J. Jr (1998). Etiology as a risk factor for medically refractory epilepsy: a case for early surgical intervention. Neurology, 51, 1243–4. Engel, J. Jr and Shewmon, D.A. (1991). Impact of the kindling phenomenon on clinical epileptology. In Kindling and Synaptic Plasticity: The Legacy of Graham Goddard, ed. F. Morrell, pp. 195–210. Cambridge, MA: Birkhäuser Boston. Engel, J. Jr, Caldecott-Hazard, S. and Bandler, R. (1986). Neurobiology of behavior: anatomic and physiologic implications related to epilepsy. Epilepsia, 27(Suppl. 2), S3–S11. Engel, J. Jr, Bandler, R., Griffith, N.C. and Caldecott-Hazard, S. (1991). Neurobiological evidence for epilepsy-induced interictal disturbances. In Advances in Neurology, ed. D. Smith, D. Treiman & M. Trimble, pp. 97–111. New York: Raven Press. Vol. 55. Engel, J. Jr, Dichter, M. and Schwartzkroin, P. (1997). Basic mechanisms of human epilepsy. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 499–512. Philadelphia, New York: Lippincott-Raven. Farde, L. (1997). Brain imaging of schizophrenia – the dopamine hypothesis. Schizophr Res, 28, 157–62. Frost, J.J., Mayberg, H.S., Fisher, R.S. et al. (1988). Mu-opiate receptors measured by positron emission tomography are increased in temporal lobe epilepsy. Ann Neurol, 23, 231–7. Furmark, T., Fischer, H., Wik, G., Larsson, M. and Fredrikson, M. (1997). The amygdala and individual differences in human fear conditioning. Neuroreport, 8, 3957–60. Gloor, P. (1997). The Temporal Lobe and the Limbic System. New York: Oxford University Press. Gloor, P., Salanova, V., Olivier, A. and Quesney, L.F. (1993). The human dorsal hippocampal commissure: an anatomically identifiable and functional pathway. Brain, 116, 1249–73. Glosser, G., Zwil, A.S., Glosser, D.S., O’Connor, M.J. and Sperling, M.R. (2000). Psychiatric aspects of temporal lobe epilepsy before and after anterior temporal lobectomy [In Process Citation]. J Neurol Neurosurg Psychiatry, 68, 53–8. Goldstein, M. and Deutch, A.Y. (1992). Dopaminergic mechanisms in the pathogenesis of schizophrenia. FASEB J, 6, 2413–21.

35

Anatomical substrates of behavioural disturbances Goldstein, L.E., Rasmusson, A.M., Bunney, B.S. and Roth, R.H. (1996). Role of the amygdala in the coordination of behavioral, neuroendocrine, and prefrontal cortical monoamine responses to psychological stress in the rat. J Neurosci, 16, 4787–98. Gould, E., Woolley, C.S and McEwen, B.S. (1991). The hippocampal formation: morphological changes induced by thyroid, gonadal, and adrenal hormones. Psychoneuroendocrinology, 16, 67–84. Griffith, N., Engel, J. Jr and Bandler, R. (1987). Ictal and enduring interictal disturbances in emotional behaviour in an animal model of temporal lobe epilepsy. Brain Res, 400, 360–4. Hariri, A.R., Bookheimer, S.Y. and Mazziotta, J.C. (2000). Modulating emotional responses: effects of a neocortical network on the limbic system. Neuroreport, 11, 43–8. Henry, T.R., Engel, J. Jr and Mazziotta, J.C. (1993a). Clinical evaluation of interictal fluorine-18fluorodeoxyglucose PET in partial epilepsy. J Nucl Med, 34, 1892–8. Henry, T.R., Mazziota, J.C. and Engel, J. Jr (1993b). Interictal metabolic anatomy of mesial temporal lobe epilepsy. Arch Neurol, 50, 582–9. Herzog, A.G. (1997). Disorders of reproduction and fertility. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr & T.A. Pedley, pp. 2013–19. Philadelphia, New York: Lippincott-Raven. Ketter, T.A., George, M.S., Kimbrell, T.A., Benson, B.E. and Post, R.M. (1996). Functional brain imaging, limbic function, and affective disorders. Neuroscientist, 2, 55–65. Kline, N.S., Li, C.H., Lehmann, H.E., Lajtha, A., Laski, E. and Cooper, T. (1977). Beta-endorphineinduced changes in schizophrenic and depressed patients, Arch Gen Psychiatry, 34, 1111–13. Kovelman, J.A. and Scheibel, A.B. (1984). A neurohistological correlate of schizophrenia. Biol Psychiatry, 19, 1601–21. Krahn, L.E., Rummans, T.A. and Peterson, G.C. (1996). Psychiatric implications of surgical treatment of epilepsy. Mayo Clin Proc, 71, 1201–4. Laruelle, M. and Abi-Dargham, A. (1999). Dopamine as the wind of the psychotic fire: new evidence from brain imaging studies. J Psychopharmacol, 13, 358–71. LeDoux, J.E. (1993). Emotional memory systems in the brain. Behav Brain Res, 58, 69–79. Lothman, E.W., Bertram, E.H. III and Stringer, J.L. (1991). Functional anatomy of hippocampal seizures. Prog Neurobiol, 37, 1–82. Maclean, P.D. (1952). Some psychiatric implications of physiological studies on frontotemporal portion of limbic system (visceral brain). EEG Clin Neurophysiol, 4, 407–18. Madison, D.V. and Nicoll, R.A. (1986). Actions of noradrenaline recorded intracellularly in rat hippocampal CA1 pyramidal neurones, in vitro. J Physiol (Lond), 372, 221–44. Malizia, A.L. (1999). What do brain imaging studies tell us about anxiety disorders? J Psychopharmacol, 13, 372–8. Mathern, G.W., Babb, T.L. and Armstrong, D.L. (1997). Hippocampal sclerosis. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr & T.A. Pedley, pp. 133–55. Philadelphia, New York: Lippincott-Raven. Mayberg, H.S, Lewis, P.J., Regenold, W. and Wagner, H.N.J. (1994). Paralimbic hypoperfusion in unipolar depression. J Nucl Med, 35, 929–34. McIntyre, D.C. and Plant, J.R. (1993). Long-lasting changes in the origin of spontaneous discharges from amygdala-kindled rats: piriform vs. perirhinal cortex in vitro. Brain Res, 624, 268–76.

36

J. Engel Jr et al. Mendez, M.F., Cummings, J.L. and Benson, F. (1986). Depression in epilepsy: significance and phenomenology. Arch Neurol, 43, 766–70. Morrell, F., Whisler, W.W., Smith, M.C. et al. (1995). Landau-Kleffner syndrome: treatment with subpial intracortical transection. Brain, 118, 1529–46. O’Keefe, J. and Nadel, L. (1978). The Hippocampus as a Cognitive Map. Oxford: Clarendon Press. Papez, J.W. (1937). A proposed mechanism of emotion. Arch Neurol Psychiatry, 38, 725–43. Piredda, S. and Gale, K. (1985). A crucial epileptogenic site in the deep prepiriform cortex. Nature, 317, 623–25. Rogers, M.A., Bradshaw, J.L., Pantelis, C. and Phillips, J.G. (1998). Frontostriatal deficits in unipolar major depression. Brain Res Bull, 47, 297–310. Sato, M. (1983). Long-lasting hypersensitivity to methamphetamine following amygdaloid kindling in cats: the relationship between limbic epilepsy and the psychotic state. Biol Psychiatry, 18, 525–36. Savic, I., Altshuler, L., Baxter, L. and Engel, J. Jr (1997). Pattern of interictal hypometabolism in PET scans with fludeoxyglucose F18 reflects prior seizure types in patients with mesial temporal lobe seizures. Arch Neurol, 54, 129–36. Schwartzkroin, P.A. and McIntyre, D.C. (1997). Limbic anatomy and physiology. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 323–40. Philadelphia, New York: Lippincott-Raven. Shouse, M.N., Martins da Silva, A. and Sammaritano, M. (1997). Sleep. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 1929–42. Philadelphia, New York: Lippincott-Raven. Siegel, A.M., Wieser, H.G., Wichmann, W. and Yasargil, G.M. (1990). Relationships between MRimaged total amount of tissue removed, resection scores of specific mesiobasal limbic subcompartments and clinical outcome following selective amygdalohippocampectomy. Epilepsy Res, 6, 56–65. Sloviter, R.S. and Damiano, B.P. (1981). Sustained electrical stimulation of the perforant path duplicates kainate-induced electrophysiological effect and hippocampal damage in rats. Neurosci Lett, 24, 279–84. Sperling, M.R. and O’Connor, M.J. (1990). Auras and subclinical seizures: characteristics and prognostic significance. Ann Neurol, 28, 320–8. Stevens, J.R. (1973). An anatomy of schizophrenia. Arch Gen Psychiatry, 29, 177–89. Stokes, P.E. (1995). The potential role of excessive cortisol induced by HPA hyperfunction in the pathogenesis of depression. Eur Neuropsychopharmacol, 5, 77–82. Velasco, A.L., Wilson, C.L., Babb, T.L. and Engel, J. Jr (2000). Functional and anatomic correlates of two frequently observed temporal lobe seizure-onset patterns. Neural Plasticity, 7, 49–63. Vindrola, O., Briones, R., Asai, M. and Fernandez-Guardiola, A. (1981). Brain content of leu 5and met 5-enkephalin changes independently during the development of kindling in the rat. Neurosci Lett, 26, 125–30. Weinberger, D.R. Berman, K.F., Suddath, R. and Torrey, E.F. (1992). Evidence of dysfunction of a prefrontal-limbic network in schizophrenia: a magnetic resonance imaging and regional cerebral blood flow study of discordant monozygotic twins. Am J Psychiatry, 149, 890–7.

37

Anatomical substrates of behavioural disturbances Wieser, H.G. (1988). Human limbic seizures: EEG studies, origin, and patterns of spread. In Anatomy of Epileptogenesis, ed. B.S. Meldrum, J.A. Ferrendelli and H.G. Wieser, pp. 127–38. London: John Libbey. Williamson, A. and Spencer, D.D. (1994). Electrophysiological characterization of CA2 pyramidal cells from epileptic humans. Hippocampus, 2, 226–37. Williamson, P.D. and Engel, J. Jr (1997). Complex partial seizures. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 557–66. Philadelphia, New York: LippincottRaven. Wilson, C.L. (1995). Functional pathways underlying ipsilateral and contralateral spread of temporal lobe seizures. In Epilepsy and the Corpus Callosum, ed. A. Reeves, Vol. II, pp. 153–73. New York: Plenum Press. Wilson, C.L., Isokawa-Akesson, M., Babb, T.L. and Crandall, P.H. (1990). Functional connections in the human temporal lobe: I. Analysis of limbic system pathways using neuronal activity evoked by electrical stimulation. Exp Brain Res, 82, 279–92. Wilson, C.L., Isokawa, M., Babb, T.L., Crandall, P.H., Levesque, M.F. and Engel, J. Jr (1991). Functional connections in the human temporal lobe: II. Evidence for a loss of functional linkage between contralateral limbic structures. Exp Brain Res, 85, 174–87. Wilson, C.L., Khan, S.U., Engel, J. Jr, Isokawa, M., Babb, T.L. and Behnke, E.J. (1998). Paired pulse suppression and facilitation in human epileptogenic hippocampal formation. Epilepsy Res, 31, 211–30. Wong, R.K.S. and Traub, R.D. (1983). Synchronized burst discharge in disinhibited hippocampal slice. I. Initiation in CA2-CA3 region. J Neurophysiol, 49, 442–58. Xie, C.W., Morrisett, R.A. and Lewis, D.V. (1992). Mu opioid-receptor mediated modulation of synaptic currents in the dentate granule cells of rat hippocampus. J Neurophysiol, 68, 1113–20.

Part II

Clinical aspects

4

The psychiatry of idiopathic generalized epilepsy Dieter Janz Burgunder Strasse 8, Berlin, Germany

‘Temporal Lobe Epilepsy (TLE) is associated with an increase in psychiatric morbidity compared to other types of epilepsy.’ This hypothesis is central to the discussion of the relationship between epilepsy and psychopathology, according to the statement of Anne Stub-Naylor, a young Danish doctor, in her thesis on ‘Epilepsy and psychiatric disorder – a comorbidity study’ (1996). And she continues: ‘Despite the scarce evidence it appears to have become a dogma that such a relationship exists. In “contemporary” medicine especially three “schools” have nourished this view, represented by Elliot Slater, David Bear and Michael Trimble’. Indeed, the description of schizophrenia-like psychoses in patients with TLE (Slater et al., 1963) – has encouraged the expectation of finding a biological-based explanation for schizophrenia. For this reason, the discussions on psychiatric disorder and especially those on psychoses in the case of epilepsy have long been dominated by the ‘limbic hypothesis’. It seems that the time has come to take a critical look at the above thesis, because the understanding that other forms of epilepsy also show a clinical profile and a biological quality is gaining ground. To have a stake in biological psychiatry, the psychiatric aspects should be taken into consideration in regard to other forms and syndromes of epilepsy as well. For the following reasons, the syndromes of idiopathic generalized epilepsy (IGE) are especially suitable for consideration. Grouped together they are at least as frequent as the localization-related temporal and extra-temporal epilepsies. Phenomenologically and electroencephalographically they are well defined; therefore they can easily be distinguished from focal epilepsies and symptomatic generalized epilepsies. The coexistence of generalized and focal epilepsies is extremely rare, although transitions from one to the other are possible due to a long and severe course of disease, and they might well be relevant for the cause of psychiatric disorder. The comparison of subsyndromes within families shows that the various syndromes of idiopathic generalized epilepsy make up one genetic entity, 41

42

D. Janz

which differs clearly from genetic focal epilepsies in view of their phenotype (Janz, 1997) as well as from the molecular-genetic view (Sander et al., 2000). According to the Commission on Classification and Terminology of the International League against Epilepsy (1989), within IGE belong – apart from ‘benign neonatal convulsions’ and ‘benign myoclonic epilepsy in infancy’, which are irrelevant for our topic – childhood absence epilepsy (CAE), juvenile absence epilepsy (JAE), juvenile myoclonic epilepsy (JME) and epilepsy with grand mal seizures (GTCS) on awakening (GMA). For these four syndromes of ‘IGE of childhood and adolescence’ electroencephalographic regular or fast generalized spike-wave discharges are the typical EEG patterns. When scanning the literature for the frequency of psychiatric disorders in various types of epilepsy, difficulties of classifying become obvious. The earlier literature does not recognize the modern syndrome classification. Yet, in some recent studies, especially those from the psychiatric side, the occurrence of generalized seizures is equated with the supposition of generalized epilepsy (Méndez et al., 1993). Epilepsy with exclusively grand mal seizures (TCS) can be taken for IGE, even without EEG, if it concerns an epilepsy with grand mal predominantly on awakening (GMA). But such details are still frequently neglected, although the International Classification has existed for 12 years. Psychosis Looking at the most serious, yet rarest, psychiatric disorder, the psychoses, and paying attention to their frequency in generalized epilepsies in comparison to focal epilepsies, most authors of the relevant literature (Table 4.1), found that psychoses in focal epilepsies are slightly more frequent than in generalized epilepsies, but the difference proved to be significant in only one study (Onuma, 1983). Two studies using the International Classification did not find the frequencies of psychoses in IGE and in TLE to be different (Schmitz and Wolf, 1991, 1995; Sengoku et al., 1997). With regard to specific epileptic syndromes, CAE was significantly associated with the presence of psychosis. CAE occurred in 7/28 (25%) of patients with psychosis compared with 55/669 (8%) of patients without psychosis (Schmitz and Wolf, 1991). Relation to seizure type

With regard to specific seizure types, absences in combination with GTCS are associated with an increased risk of psychosis. With this seizure combination psychosis occurred in 11.3% (7/62) of patients in comparison with only 1% (1/91) of patients with myoclonic jerks in combination with GTCS (Schmitz, 1988). This figure is in good correspondence with findings of Genton et al. (2000) who reported 5 (2.9%)

43

Psychiatry of idiopathic generalized epilepsy

Table 4.1. Controlled studies on the frequency of psychosis in different types of epilepsy

Studies

Type of psychosis/population

Results

Gastaut (1956)

Mixed

GE, TLE, non-TLE: not different

Small et al. (1962)

Schizophrenia

TLE, 24%; non-TLE, 12%: not significantly different

Stevens (1966)

History of admission to psychiatric hospital

GE, 29%; TLE, 31%; FOC non-TLE, 8%

Mignone et al. (1970)

MMPI: schizophrenia score

Psychomotor vs. nonpsychomotor: not different

Bruens (1971)

Mixed

GE, 3.3%; SIM FOC, 0%; COM FOC, 3.5%; GE⫹TLE, 19%

Standage and Fenton (1975)

Present State Examination

TLE vs. non-TLE: not different

Shukla et al. (1979)

Schizophrenia

GE, 4%; TLE, 17%

Onuma (1983)

Paranoid symptoms

PGE, 3.5%; TLE, 9.2%; non-TLE, 2.5%: PGE vs. TLE, significantly different

Sengoku et al. (1983)

Mixed, epilepsy institution

GE, 4.3%; TLE, 6.0%; non-TLE, 2.0%: GE vs. TLE not significantly different

Schmitz (1988)

Mixed, Department of Neurology

IGE, 3.5%; FOC,4.2%; GE⫹FOC, 4%: not significantly different

Sengoku et al. (1997)

Mixed, Department of Psychiatry

IGE, 19.4%; TLE, 15.2%: not significantly different

Bredkjaer et al. (1998)

Nonorganic Nonaffective psychosis

Psychomotor epilepsy 3.57 SIR, petit mal epilepsy 2.3 SIR

Notes: GE, generalized epilepsies; PGE, primary GE; IGE, idiopathic GE; TLE, temporal lobe epilepsies; non-TLE, nontemporal lobe epilepsies; FOC, focal epilepsies; SIM FOC, epilepsies with simple focal seizures; COM FOC, epilepsies with complex partial seizures; SIR, standardized incidence ratios; MMPI, Minnesota Multiphasic Personality Inventory.

cases with psychotic disorders among 170 patients with JME. Unfortunately, we do not know the influence that other seizure variables, such as the length of the disease, or frequency and intensity of the seizures, may have on these differences between the various IGE-syndromes. In their retrospective studies of 697 and 611 consecutive patients with unselected epilepsies (Wolf, 1976; Schmitz, 1988), Schmitz and Wolf (1995) found a ‘puzzling’ relation between absences and complex focal seizures regarding an increased risk of psychosis. Contrary to this, patients with only GTCS had no increased risk of

44

D. Janz

psychosis. They speculated therefore that both seizure types testify to the ability of a patient’s brain to produce epileptic events that are widespread but still limited. If prolonged but strongly limited seizure discharge is indeed present during epileptic psychoses of many types, as has been proposed, a brain’s ability to produce such a discharge would be a prerequisite for psychosis. The brains of patients seem to have learned the compromise ‘in between’ if they cannot produce or give an epileptic allor-nothing response. Ictal psychosis

The classical example of psychotic states due to ‘epileptic events that are widespread but still limited’ is the nonconvulsive status epilepticus. The symptoms which are common to both types of nonconvulsive status, the so-called absence- (or petit mal-) status and the focal status, are disturbance of consciousness and incomplete or complete amnesia. This psychotic appearance of absence status is completed by lack of responsitivity and stuporous behaviour. Absence status is by far more frequent in IGE than its analogue, the complex focal status, in focal epilepsies. Gibbs and Gibbs (1952) and Lennox (1960) reported a history of petit mal status in 27 (2.6%) of their 1039 petit mal patients. We have seen it in 15 (5%) of our approximately 300 patients with pyknoleptic petit mal (Janz, 1998), a percentage which corresponds to the 5.8% of Loiseau and Cohadon (1970). Dalby (1969) has reported petit mal status in 21 (6.2%) of his series of 346 patients with 3-Hz spikewave and 15 (9.3%) of the 160 patients with a history of petit mal. Absence status which lasts for hours or days usually occurs only after adolescence. As isolated events, they may even occur in middle or later age (Jaffé, 1962; Lee, 1985). According to its clear clinical and electroencephalographic picture, the almost negative interictal psychopathology, and the almost positive electric activity, the absence status can be characterized as a specific ictal psychosis of IGE. Postictal psychosis

Furthermore, postictal and interictal psychoses are to be distinguished. Postictal psychoses are clearly demarcated by a lucid interval from seizures themselves and from immediate postictal confusion. The phenomenology is pleomorphic, with fluctuating combinations of delirium, delusions, hallucinations, thought disorders, or/and affective change. Postictal psychosis is not specific for a type of epilepsy. Dongier (1959/60) found a preponderance of generalized epilepsies in cases of postictal psychosis. Others have found a higher rate among patients with localizationrelated epilepsies (Logsdail and Toone, 1988; Savard et al., 1991). Devinsky et al. (1995) noted in their case-control study of 20 patients with postictal psychoses no significant difference between the percentage of patients with focal or primary generalized epilepsy in cases and controls. It seems as if long duration of epilepsy

45

Psychiatry of idiopathic generalized epilepsy

(Kanemoto et al., 1996) and high number of lifetime GTCS (Devinsky et al., 1995) contribute to the pathogenesis of postictal psychosis. Interictal psychosis

Interictal psychotic episodes show a paranoid hallucinatory symptomatology. According to some authors the delusions are less systematized than in true schizophrenia (Janzarik, 1955; Tellenbach, 1965), personality deterioration is lacking (Slater et al., 1963) and thought disorder unusual (Korzeniowski, 1965). Unfortunately, there are no controlled data on the frequency of acute and chronic interictal psychoses comparing IGE and TLE. According to Sengoku et al. (1997) chronic psychoses can only be found in patients with TLE. Moreover, only in TLE do psychoses relate to the frequency of seizures. Correspondingly, Schmitz (1992) found that five of six chronic psychoses occurred in patients with TLE, but she says ‘also this relation was not significant due to small numbers’ and ‘that psychoses in TLE tend to be more ‘severe’ than those in generalized epilepsy’. Unfortunately a comparison of symptomatology in Sengoku et al.’s (1997) study cannot be made for evaluation of interictal psychoses, as the semiological rating is not defined. The preponderance of paranoid symptoms in TLE psychoses is surprising, as is the frequent occurrence of ‘perplexed behaviour’, which is not defined in the text. Similarly to Wolf (1982), Sengoku et al. (1997) have noted an association between visual hallucinations and generalized epilepsies, whilst other authors have also noticed auditory hallucinations in psychoses in TLE. Unfortunately there is also no systematic comparison of the psychopathological contents of interictal psychoses in both forms of epilepsy, but only impressions. Thus Sengoku et al. (1983) wrote that with regard to the contents of psychosis, the episodic paranoid states were most commonly found in the TLE group. In the generalized epilepsy (GE) group, however, the types of psychosis were more varied. Wolf (1982) noted an association between olfactory hallucinations and TLE and between visual hallucinations and generalized epilepsies, suggesting that the form of the psychosis is influenced by symptoms of the epileptic disorder. Alternative psychosis with forced normalization of the EEG

Some of the interictal psychoses – according to Schmitz (1988) 11%, to Wolf (1985) 15%, to Dongier (1959/60) 25% – belong to the group of psychiatric disorders with so-called forced normalization of the EEG (Landolt, 1955, 1958, 1963). With this term Landolt described the condition that with spontaneous or, more frequently, drug-induced disappearance of clinical and subclinical seizure manifestations, psychiatric symptoms appear. These may be (Table 4.2) prepsychotic dysphoria characterized by insomnia, anxiety, feelings of oppression and social withdrawal or may proceed to psychotic symptoms characterized by hallucinatory paranoid states.

46

D. Janz

Table 4.2. Psychiatric syndromes in 44 cases of forced normalization observed in 36 patients

Paranoid hallucinatory psychosis Prepsychotic dysphoria Hysterical episode Hypochondrial episode Depressive episode Manic episode Twilight state Depersonalization

19 9 5 3 2 2 1 1

Source: Wolf (1984).

The phenomenon of an inverse relationship between well-being on one hand and seizure and specific discharges in the EEG on the other occurs with both generalized and focal epilepsy. Landolt’s first observations were with temporal lobe epilepsy (1955). He felt himself supported by Gibbs (1951), who from his experience with Phenurone drew the conclusion that ‘the epileptic and psychiatric components of psychomotor epilepsy are physiologically antithetic’. Later, with the advent of the succinimides, the focus turned towards the generalized epilepsies. The fundamental study of Tellenbach (1965) gives a detailed description of – as he called it – alternative psychoses of a paranoid type in 12 cases, 9 with IGE, in 4 cases in combination with pyknoleptic (childhood) absences. According to Wolf and Trimble (1985) alternative psychoses occurred three times more frequently in IGE than in epilepsies with complex focal seizures and in IGE have been the predominant type of psychoses. Nearly the same relationship holds true in the series of Schmitz (1988) (Table 4.3). In a subgroup – patients with IGE with absences beyond childhood – it may be seen in as many as 8% of patients, depending on the medication given (Wolf, 1986). According to Wolf ’s opinion (1986), in this group we seem to know some details of the pathogenesis. He mentioned in this respect drug influences, as for instance the sleep deprivation and increase of light sleep with ethosuximide, and social status, particularly regarding the frequent vocational disintegration. He points to the observations of Tellenbach (1965) of the frequent familial conflict situations, and on the patient’s personality being affected by having to cope with disabling drug effects. Tellenbach has argued that patients with awakening epilepsy are particularly sensitive to a drug-induced process of slowing down, the feeling of ‘being constrained’, of being ‘immured’ because of their personality trait of extraversion.

47

Psychiatry of idiopathic generalized epilepsy

Table 4.3. Psychosis and related syndromes in patients of a neurological outpatient clinic (N ⫽ 611)

Syndrome groups

Generalized epilepsies (n⫽237)

Focal epilepsies (n⫽270)

Epilepsies with generalized and focal features (n⫽46)

Unclassified epilepsies (n⫽58)

Ictal and peri-ictal Forced normalization Intoxicated Drug withdrawal Unrelated to epilepsy Multifactorial and unknown

3 4 1 — 3 2

9 1 2 1 4 1

2 — — — 1 1

1 — — — — 1

Source: Schmitz (1988).

Anything which threatens their stability will immediately and recklessly be expressed. In the number of conditions which in the case of ‘forced normalization of the EEG’ (forcierter Normalisierung des EEG) leads to psychiatric disorders, Tellenbach includes special personality traits of patients with awakening epilepsy. What did the scholars understand by this in 1965, and what do we know about it currently? Personality ‘The heart of the matter’

When, in the beginning of the 1950s, the clinical picture of awakening epilepsy became clear to me, including that of all epilepsies with seizures mainly after awakening, i.e. today’s IGE, I became more and more interested in the conditions which led to the triggering of seizures. According to Niedermeyer (1996), in IGE ‘the role of arousal is the heart of the matter’. The period after awakening, at whatever time of awakening, is the most critical biological situation in any variety of IGE, both for the appearing of generalized spike wave-discharges and for the manifestation of the generalized tonic-clonic seizures, the absences and the myoclonic jerks. Therefore the early-chosen term ‘epilepsy on awakening’ (Janz, 1953, 1962) focuses the attention to the crucial point which is pathophysiologically common to all forms of IGE. It has been noted that all seizures in IGE were frequently precipitated by external factors such as emotional and physical stress especially in association with lack of sleep. These factors seem to have little influence on seizures in other forms of epilepsy, as shown in comparison with sleep epilepsy (SE) (Table 4.4).

48

D. Janz

Table 4.4. Seizure precipitants in sleep epilepsy (SE) (n ⫽127) and in awakening epilepsy (AE) (n⫽ 90)

First seizures (%)

Following seizures (%)

Precipitant

SE

AE

SE

AE

Sleep deprivation Alcohol use Abrupt wakening Strong emotion Physical stress Menarche

5 1 — 2 3 2*

25 10 12 — 9 6*

6 3 — — 4 —

42 23 28 — 13 —

Note: * Female patients only. Source: Janz (1953).

In JME in particular, we have reported unusual lack of sleep and sudden awakening, and/or excessive alcohol consumption, as the most common precipitating factors (Janz and Christian, 1957). Other triggering factors were less common. Pedersen and Petersen (1998) registered, although retrospectively, alcohol consumption in 51%, stress in 70% and sleep deprivation in 84% of patients as the most frequent precipitating factors. Lifestyle

Most patients are not in a position to sleep as much as they would like to every day; they are forced to get up earlier than they would like to, and therefore suffer from a chronic sleep deficit. Additional stress forces them to pass their limits, and seizures result. Everyone who knows patients with epilepsy on awakening can produce numerous examples of seizures occurring after late nights spent in preparing for examinations, attending parties, watching television or preparing to leave early for holidays. Some of the causes are so extreme that one is tempted to say that it is no wonder that seizures occur. However, the patients describe the events as unavoidable, and one tends to pardon these excesses at first. One becomes suspicious when seizures continue to occur after repetitions of the same or similar events. It is surprising that the patients often do not avoid the situations which provoke seizures, and it is not clear why the patients are unable to learn from experience. One gets the impression that the patients are unable to draw conclusions from their own negative experiences and to change their lifestyles, and understands Niedermeyer’s

49

Psychiatry of idiopathic generalized epilepsy

Table 4.5. Modes of falling asleep and awakening in patients with awakening epilepsy (AE) and in patients with sleep epilepsy (SE)

AE

SE

n

%

n

%

Falling asleep

Quick, unmolested Delayed, difficult

19 28

40 60

36 6

86 14

Awakening

Quick, unmolested Delayed, difficult

11 50

18 82

32 15

68 32

Source: Janz (1953).

(1996) moaning: ‘Unfortunately, some patients are incorrigible and all exhortations fall on deaf ears’. Sleeping habits

Of all the precipitating factors, sleep deprivation is certainly the most important, and it would appear that other factors become effective in the presence of sleep deprivation. The question arises what is the reason for this relation: is it a faulty behaviour of the patients, is it a particular biological weakness or probably both? Patients with IGE, particularly those with JME, tend to awake slowly and with difficulty and remain sleepy for some time afterwards (Table 4.5) (Janz and Christian, 1957; Janz, 1985, 2000). Nearly every patient can report that he or she is difficult to arouse, and would love to stay in bed longer, that ‘sleep still sits in their limbs’, that they initially act ‘automatically’, ‘like in a dream’, or ‘are numb’. Many have adjusted to this by sleeping longer and having breakfast in bed. Others who cannot afford this, dunk their heads in cold water or drink strong coffee. All emphasize that ‘they need much sleep’ and most claim that they sleep very deeply. However, in our experience, these patients do not admit or do not recognize that they often retire too late and that, therefore, they often do not get enough sleep, i.e. that they easily suffer from lack of sleep. Interviews by questionnaire which have recently been carried out in two locations in Berlin indicate that there is a significant difference in well-being during day-time between the patients with JME and those with TLE (Pung, 2000) (Table 4.6). JME patients feel less efficient in the morning than TLE patients. The fact that perception of well-being in the evening met only partly but not significantly with expectations, might depend on a lack of acceptance or on an adaptation to social conventions. It is useful to complete the interview of the patients together with their near relatives or friends, because JME patients particularly tend to adapt their answers to conventional attitudes.

50

D. Janz

Table 4.6. Answers of 20 patients with juvenile myoclonic epilepsy (JME) and 20 patients with temporal lobe epilepsy (TLE)

Question

Time period

JME

TLE

16–21.00 h 10–16.00 h 8–10.00 h 5–8.00 h

3 16 1 —

1 12 6 1

Morning Evening

7 13

14 6

‘What time of the day do you feel at top?’

‘Are you more a morning type or an evening type?’

Source: Pung (2000).

Personality traits in patients with epilepsy on awakening Phenomenology

Patients with IGE, especially those with JME, seem to lack a certain degree of selfcontrol. This is evident in the fact that they seem to neglect their physical needs and their health. They are quite likable. They are not hypochondriacs, like many patients with temporal lobe epilepsy, and do not behave hysterically, but it is important to recognize that these patients fail to avoid situations which might provoke seizures, and also forget to take their medications or even discontinue treatment. This lack of self-control goes with a tendency to deny problems and conflicts, and besides a shortlasting depression postictally they tend to take seizures, symptoms and complaints lightly, some even believing that they can avoid the disease by not taking their tablets. These patients, particularly if teenagers, often become obstinate, claiming something to the effect that ‘I would not consider going to bed at regular hours, abstaining from alcohol and living like a monk. I am not a child anymore. If I can’t do what I want, then I might as well hang myself ’. This is an expression of an immature and childish attitude which must be met with comprehension and firmness rather than authoritarian exhortations. Another characteristic that one often encounters in patients with IGE is their impressionability. They comprehend and judge quickly, and they are versatile and adaptable as well as impressionable, which facilitates contact with the physician but endangers the continuity of the therapeutic relationship. Characteristics such as impressionability, openness and awareness to the point of being easily distracted make these patients appear alert and intelligent, but cause them problems when tasks requiring resoluteness, patience and perseverance are concerned.

51

Psychiatry of idiopathic generalized epilepsy

Patients with IGE are somewhat unstable and tend to lose their emotional balance easily. They may be moody, though these phases do not last long. Lack of self-confidence leads them to underestimate themselves though overconfidence may also be observed. Since my first description (1953), I have pointed out repeatedly (1962, 1969, 1974, 1989, 1998) that patients with this epileptic syndrome are, as a rule, very different from what had long been considered the typical personality of ‘genuine epileptics’ – that is, slow, fussy, viscous, circumstantial and irritable, utterly pedantic and obstinate, prone to hypochondria and practically incapable of changing the subject of conversation. Some of these traits are now grouped together as the ‘Geschwind syndrome’ (Benson, 1991; Blumer, 1999) and associated with TLE. Patients with IGE are inclined to the opposite – more unstable, less reliable, inconsiderate and neglectful of duties and self-interest, more ready to yield to the slightest temptation even against better judgement, insensitive to future consequences. They behave sometimes, as Tellenbach put it, like an ‘adult child’ or as Pellock (1999) described JME patients, as ‘perpetual adolescents’. Investigations

Peter Wolf (1985) wrote The clinical observation of Janz was followed by a psychological investigation of Leder (1967), who studied a group of 34 patients with GMA and 55 with GMS. From these, two groups of 10 persons each were selected for statistical evaluation. These groups were matched for sex, age, seizure frequency, association with minor seizures, intelligence, absence of etiological clues and social status. They were investigated with two personality tests, the Rorschach and Szondi tests, and several significant differences between the groups were found. Their interpretation describes patients with GMA as extroverts who have difficulty in recognizing the limits of their own person in relation to the external world. Often they have little ability to suppress, to contradict and to renounce conflicts; tensions and disinclinations are usually momentarily disposed often by denial. These patients will follow simultaneously the most divergent and irreconcilable aims without being aware of any difficulties.

Wolf remarks that the Leder investigation, which has been confirmed by similar studies in Japan (Aoki and Kawai, 1982; Kawai and Aoki, 1983; Yamashita et al., 1984), has been criticized for its methodology, since many psychologists have reservations about Rorschach and Szondi tests. However, his description will remind unprejudiced readers of many of their own patients with IGE whose lack of discipline is often an obstacle to successful therapy. Their unstable sleep behaviour may precipitate seizures, and sleep withdrawal is often the cause of the first seizure. Therefore, it seems possible that unstable personalities of this kind are common in GMA because such a personality is a factor which favours the manifestation of a predisposition to seizures which may often remain latent in people with more regular

52

D. Janz

lifestyles. It is the opinion of Peter Wolf that comparative psychological investigations of patients and predisposed but unaffected relatives could clarify this question. Personality in patients with pyknolepsy (childhood absence epilepsy) Phenomenology

In the case of two subsyndromes of IGE, namely in pyknolepsy or CAE and in JME, the description of personality traits has its own tradition. Before the era of EEG, the observation that children with pyknolepsy are usually ‘lively to over-lively, mentally active and markedly intelligent’ (lebhaft bis überlebhaft, geistig regsam und ausgesprochen intelligent; Rosenthal 1935), has often served as an argument that due to this unepileptic behaviour, pyknolepsy is not an epileptic disease. But even later, when the EEG had given an explanation for the true nature of the absences, the friendliness and brightness of the children was emphasized, and they were often called ‘model children’ (Bridge, 1949), even though their positive traits could not hide the critical ones. Bridge, for example, noted that children with pyknolepsy are overlively, never stop in order to relax and show no sign of fatigue when playing with other children. At home and at school, they are very keen to fulfil their duties due to their high sensitivity to criticism and scolding. As a consequence, these children strain their energies excessively. I have myself pointed out that often, in the course of their further development, the parents’ expectations are disappointed (Janz, 1955, 1998). The increasing demands at school during their education reveal that the children’s liveliness and cleverness hide a lack of profoundness and perseverance. The performance at school decreases. The parents complain that their children do not perform any better than average pupils even though they used to be among the best, and they explain this as follows: ‘despite their intelligence they are too light-headed’, ‘they are not attentive’ and ‘they have too many other things in their heads’. What once seemed like keen interest and brightness, then reveals itself to be a lack of concentration and carelessness, the liveliness is nervousness, and the quick perception a lack of profoundness. This metamorphosis from the intellectual miraculous child to the average pupil or even a ‘loser’ reflects a frequent misjudgement by the adults. These observations of the psychological characteristics and the development of children with pyknolepsy or childhood absence epilepsy show an important proximity to the generally described psychological structure of awakening epilepsy. Investigations

For this reason the experimental–psychological investigations of absence epilepsy which Mirsky and his group have pursued since the 1960s are of special interest. They have found in absence epilepsy deficits in sustained auditory and visual attention that are distinct from those seen in other seizure disorders (Mirsky et al., 1995). The event-related potential studies indicate that an impaired sensory input, seen

53

Psychiatry of idiopathic generalized epilepsy 1.0 0.5 Normal controls (n ⫽76)

Standard scores

0.0 ⫺0.5

Absence seizures (n ⫽76)

⫺1.0

Complex partial seizures (n ⫽7)

⫺1.5 ⫺2.0 ⫺2.5 ⫺3.0

Shift Figure 4.1.

Focus/Execute

Sustain

Encode

Component scores extracted by principal components analysis of attention tests on normal control subjects (white columns), patients with absence seizures (black columns) and complex partial seizures (cross-hatched columns). The asterisks appearing below the columns indicate that the group so designated differed significantly from the other two groups on that component. The absence group performed significantly worse than the other two groups on the sustain component, whereas the complex partial group was impaired, in comparison with the other groups, on the shift and focus/execute components. (Reproduced with permission from Mirsky et al., 1991.)

primarily in the auditory modality, not only occurs during spike-wave bursts, but during interictal periods as well. The clinical impression of increased distractibility in absence patients is psychologically explained by a significant score on a test assessing the sustaining function of attention, whereas other components of the complex process of attention like ‘shift’ or ‘focus/execute’ are significantly more impaired in patients with complex focal seizures (Mirsky et al., 1991) (Figure 4.1). Levav (1991) has investigated the question posed by Wolf (1985), as to whether the manifestations of IGE are associated with personality traits, that healthy relatives do not show. She investigated the attention performance of children affected with absence epilepsy and that of their first-degree relatives. She found significant impairment in tests of sustained attention in comparison with unaffected siblings; moreover, particularly in female probands. That mothers tended to perform more poorly than fathers, which is in accordance with the known maternally transmitted influence on seizure susceptibility (Ottmann et al., 1988) gave rise to the assumption that measures of attention may serve as markers of vulnerability to the disorder (Mirsky et al., 1995). This hypothesis needs investigating in relation to subclinical seizure activity.

54

D. Janz

Personality traits in patients with juvenile myoclonic epilepsy Phenomenology

With regard to the behavioural phenomenology of patients with JME, we described the observations in our first article (1957) as follows: Their psychological style – as was described for awakening epileptics – is very frequently characterized by unsteadiness, lack of discipline, hedonism and indifference towards their disease, in contrast to ‘typical epileptic conduct’. Since patients with impulsive petit mal (nowadays called JME) behave rather similar we feel that it is possible to establish a typology. Only one patient was mentally retarded. Most were intellectually average, and none was particularly gifted. But since they were all quick to learn and judge, flexible and adaptable, school and professional or occupational training were easy for them. But they promise more than they deliver. Of 19 men whose occupational training was known to us in some detail, ten descended to the level of unskilled laborers after they had been employees with business training, craftsmen, skilled workers and students. Although the psychological and social effects of their disease itself were certainly responsible for this to some extent, especially since the onset of the condition occurs very often during the period of occupational training, insufficient motivation and endurance remain possible explanations for this negative development. The conduct of these patients often has an unfavorable influence on therapy. Although the patients are ready to swear that they have done everything that they were instructed to do, they frequently fail to appear at follow-up visits or to take their medications regularly. Many handle themselves with great assurance and demanding, but they may also be decidedly mistrustful and shy, fearful and inhibited. Their labile feeling of self worth also leads them to be both eager to help, to invite, to give, on the one hand and to be able to react in an exaggeratedly sensitive way on the other hand. Their mood changes rapidly and frequently. This makes their contact both charming and difficult. They are easy to encourage and discourage, they are gullible and unreliable. Their suggestibility makes contacts easy but makes trust difficult. This personality profile plays along a scale from a likable nonchalance or timidity, through a psychoasthenic syndrome to the extremes represented by sensitive or reckless psychopathy behaviour (Janz and Christian, 1957). Investigations

These impressions, gained only by observations, were completed in 1976 by psychological investigations in Denmark (Bech et al., 1976, 1977; Lund et al., 1976; Reintoft et al., 1976; Simonsen et al., 1976). In comparison to patients with idiopathic grand mal epilepsy no difference could be found concerning social background, intelligence and education. However, patients with JME more often than controls tended to experience a social decline, difficulties in finding contacts and a feeling of being discriminated against. Signs of ‘character-neurosis’, that were only diagnosed in patients with JME, were associated to the psychological test findings of ‘substability’ and ‘subvalidity’. These results have been integrated as signs for ‘instability’ and ‘psychasthenie’. Since the introduction of JME to the Anglo-Saxon scientific orbit through Asconapé and Penry (1984) and Delgado-Escueta and Enrile-Bascal (1984), many

55

Psychiatry of idiopathic generalized epilepsy

papers on JME have appeared, but comments on the psychological profile have remained scarce and controversial. Psychiatric disorders in JME have been reported in comparison with TLE to be relatively high (Vasquez et al., 1993) or relatively low but higher than in controls (Perini et al., 1996). According to the latter authors, who performed psychological tests, JME patients only showed one significant difference from controls, by having high scores on a trait anxiety scale. At the occasion of the 1989 meeting of the American Epilepsy Association in Boston, I was asked by Kiffin Penry to outline before the young epileptologists the personality profile of patients with JME from my view as a clinician. At the end of my speech I raised two questions to the audience: 1. Is it possible to verify this behavioural pattern with psychological tests? 2. Is it possible to correlate this behavioural pattern with discrete brain structures? And I cautiously gave the following answer: Some of these characteristics such as limited rational self-control, suggestibility, distractibility and indifference to physical needs suggests involvement of frontal regions of the brain. This conclusion is compatible with the fact of accentuated bilateral frontal or fronto-central spike-wave activity in this type of epilepsy. Also morphological findings of bilateral cortical microdysgenesis which is probably more pronounced in the frontal regions (Meencke, 1985; Meencke and Janz, 1984, 1985) correlate with the described psychological syndrome, which therefore one might designate as a mild frontal lobe syndrome.

Since then two groups have studied this correlation with specific frontal psychological tests. Swartz et al. (1994) examined memory function in 9 patients with JME, 15 patients with frontal lobe epilepsy (FLE) and 15 controls. The key test was the delayed match to sample, which they refer to as a measure of a primary or working memory. The JME group performed intermediate between the controls and the patients with clear frontal pathology. The authors explained their findings as related to impairment in selective attention. In a second study with identical tasks they performed a FDG positron emission tomography (PET) comparing JME patients and normal controls (Swartz et al., 1996). Controls showed activation in the dorso-lateral prefrontal area while in JME patients there was a decrease in this area. Instead, JME patients showed increased activations in the brain stem, the medio-temporal region and the lateral orbital cortex. The authors explained this as an attempt to compensate for the failure of activation in the orbito-frontal region. Devinsky et al. (1997) compared the scores of a number of frontal tasks taken from 16 patients with JME with those from 15 patients with TLE, matched for IQ. The results showed evidence for a pattern of impaired frontal functioning in patients with JME. This was significantly different from the frontal functioning in patients with TLE. ‘It is the higher order cognitive functions of planning, reasoning

56

D. Janz

and flexibility of thought that are compromised, and to lesser extent, the ability to focus attention and inhibit habitual but inappropriate responses’ (Devinsky et al., 1997). Taken together, these data do suggest subtle but nonetheless real frontal impairments of psychological functions in people with JME (Trimble, 2000). This raises some questions: Are there essential differences in the personality traits of different subsyndromes of IGE? According to the information we have up to now, the differences seem to be less important than compared with focal epilepsies, especially TLE. If we assume certain psychological characteristics, what is then the common basic disorder? Is it a disturbance of the function of attention which is noted in the sustained attention test and in the visual working memory test ? What is the relation between the assumed psychological basic disturbance and the circadian sleep–waking cycle and maybe even the ultradian rest–activity changes? And finally, what is the relation between the assumed disturbance of a cognitive performance – that of continuous attention – and a biological performance – that of a rhythmic change of rest and activity – in view of the development of those two performances during adolescence? I think that such questions help best to prevent the affected from being stigmatized. If we see that these patients often behave like ‘adult children’ (Tellenbach, 1965), ‘perpetual adolescents’ (Pellock, 1999), or better, in some way like ‘still adolescents’, then we can find ways to understand their weaknesses and to help them effectively.

R E F E R E N C ES Aoki, K. and Kawai, I. (1982). Awake epilepsy and sleep epilepsy: a psychological study with Rohrschach test. Rohrschach Res, 24, 101–17. Asconapé, J. and Penry, J.K. (1984). Some clinical and EEG aspects of benign juvenile myoclonic epilepsy. Epilepsia, 25, 108–14. Bech, P., Kjaersgard Pedersen, K.K., Simonsen, N. and Lund, M. (1977). Personality traits in epilepsy. In Epilepsy, the Eighth International Symposium, ed. J.K. Penry, pp. 257–263. New York: Raven Press. Bech, P., Kjaersgard Pedersen, K., Simonsen, N. and Lund, M. (1976). Personality traits in epilepsy. A multidimensional study of personal traits ad modum Sjöbring. Acta Neurol Scand, 54, 348–58. Benson, D.F. (1991). The Geschwind syndrome. In Advances in Neurology, ed. D. Smith, D. Treiman and M. Trimble, Vol. 55, pp. 421–41. New York: Raven Press.

57

Psychiatry of idiopathic generalized epilepsy Blumer, D. (1999). Evidence supporting the temporal lobe epilepsy personality syndrome. Neurology, 53 (Suppl. 2), S9–12. Bredkjaer, S.R., Mortensen, P.B. and Parnas, J. (1998). Epilepsy and non-organic non-affective psychosis. National epidemiological study. Br J Psychiatry, 172, 235–8. Bridge, E.M. (1949). Epilepsy and Convulsive Disorders in Children. New York: McGraw-Hill. Bruens, H.J. (1971). Psychoses in epilepsy. Psychiatr Neurol Neurochir, 174–92. Commission on Classification and Terminology of the International League Against Epilepsy (1989). Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia, 30, 389–99. Dalby, M.A. (1969). Epilepsy and 3 per second spike and wave rhythms; a clinical, electroencephalographic, and prognostic analysis of 346 patients. Acta Neurol Scand, 45 (Suppl. 40), 1–183. Delgado-Escueta, A.V. and Enrile-Bascal, F. (1984). Juvenile myoclonic epilepsy of Janz. Neurology, 34, 285–94. Devinsky, O., Abramson, H., Alper, K. et al. (1995). Postictal psychosis: a case control series of 20 patients and 150 controls. Epilepsy Res, 20, 247–53. Devinsky, O., Gershengorn, J., Brown, E., Perrine, K., Vasquez, B. and Luciano, D. (1997). Frontal functions in juvenile myoclonic epilepsy. Neurol Neuropsychol Behav Neurol, 10, 243–6. Dongier, S. (1959/60). Statistical study of clinical and electroencephalographic manifestations of 536 psychotic episodes occurring in 516 epileptics between clinical seizures. Epilepsia, 1, 117–42. Gastaut, H. (1956). Colloque de Marseille, 15–19 Octobre 1956. Compte rendu du colloque sur l’étude électroniques en dehors des crises cliniques. Revue Neurol, 95, 587–616. Genton, P., Gélisse, P. and Thomas, P. (2000). Juvenile myoclonic epilepsy today: current definition and limits. In Juvenile Myoclonic Epilepsy: The Janz Syndrome, ed. B. Schmitz and T. Sander, pp. 11–32. Petersfield, UK and Philadelphia, USA: Wrightson Biomedical Publishing. Gibbs, F.A. (1951). Ictal and non-ictal psychiatric disorders in temporal lobe epilepsy. J Nerv Ment Dis, 113, 522–8. Gibbs, F.A. and Gibbs, E.L. (1952). Atlas of Electroencephalography. Cambridge, MA: AddisonWesley. Jaffé, R. (1962). Ictal behaviour disturbance as the only manifestations disorder: Case report. J Nerv Ment Dis, 135, 470–5. Janz, D. (1953). “Aufwach”-epilepsien (als ausdruck einer den “nacht”- oder “schlaf ” – epilepsien gegenüberstehenden verlaufsform epileptischer erkrankungen. Arch Psychiat Nervenkr, 191, 73–98. Janz, D. (1955). Die klinische Stellung der Pyknolepsie. Dtsch med Wchschr, 80, 1392–4, 1399–400. Janz, D. (1962). The grand mal epilepsies and the sleeping–waking cycle. Epilepsia, 3, 69–109. Janz, D. (1969). Die Epilepsien. Spezielle Pathologie und Therapie. Stuttgart: Thieme (1998), 2. unv Aufl. Janz, D. (1985). Epilepsy with impulsive petit mal (juvenile myoclonic epilepsy). Acta Neurol Scand, 72, 449–59.

58

D. Janz Janz, D. (1989). Juvenile myoclonic epilepsy. Epilepsy with impulsive petit mal. Cleveland Clin J Med, 56 (Suppl.), 523–33. Janz, D. (1997). The idiopathic generalized epilepsies of adolescence with childhood and juvenile age of onset. Epilepsia, 38, 4–11. Janz, D. (2000). Epilepsy with grand mal on awakening and sleep-waking cycle. Clin Neurophysiol, 111 (Suppl. 2), S103–10. Janz, D. and Christian, W. (1957). Inpulsiv-Petit mal. Dtsch Z Nervenheilk (J Neurol), 176, 346–86. English (1994) in Idiopathic Generalized Epilepsies, ed. A. Malafosse, P. Genton, E. Hirsch, C. Marescaux, D. Broglin and R. Bernasconi, pp. 229–51. London: John Libbey. Janzarik, W. (1955). Der wahn schizophrener prägung in den psychotischen episoden der epileptiker und die schizophrene wahnwahrnehmung. Fortschr Neurol Psychiat, 23, 533–46. Kanemoto, K., Kawasaki, J. and Kawai, I. (1996). Postictal psychosis: a comparison with acute interictal and chronic psychoses. Epilepsia, 37, 551–6. Kawai, I. and Aoki, K. (1983). Primary generalized epilepsy and temporal lobe epilepsy: a psychological study using Rohrschach test. Folia Psychiat Neurol Jap, 37, 245–51. Korzeniowski, L. (1965). Les problemes diagnostiques concernant les psychoses paranoiaques schizophreniformes en épilepsie. Ann Méd-psychol, 123, 35–42. Landolt, H. (1955). Über verstimmungen, dämmerzustände und schizophrene zustandsbilder bei epilepsie. Ergebnisse klinischer und elektroencephalographischer untersuchungen. Schweiz Arch Neurol Psychiatr, 76, 313–21. Landolt, H. (1958). Serial electroencephalographic investigations during psychotic episodes in epileptic patients and during schizophrenic attacks. In Lectures on Epilepsy, ed. A.M. Lorentz de Haas. Folia Psychiatr Neurol Neurochir, (Suppl. 4), pp. 91–133. Amsterdam: Elsevier. Landolt, H. (1963). Die dämmer- und verstimmungszustände bei epilepsie und ihre elektroencephalographie. Dtsch Z Nervenheilk, 185, 411–30. Leder, A. (1967). Zur psychopathologie der schlaf- und aufwachepilepsie (eine psychodiagnostische untersuchung). Nervenarzt, 38, 434–42. Lee, S.I. (1985). Nonconvulsive status epilepticus: Ictal confusion in later life. Arch Neurol, 42, 778–81. Lennox, W.G. (1960). Epilepsy and Related Disorders. Boston, MA: Little & Brown. Levav, M.L. (1991). Attention performance in children affected with absence epilepsy and their first-degree relatives. Doctoral dissertation. University of Maryland, College Park. Logsdail, S.J. and Toone, B.K. (1988). Postictal psychosis: a clinical and phenomenological description. Br J Psychiatry, 152, 246–52. Loiseau, P. and Cohadon, F. (1970). Le Petit Mal et ses Frontiéres. Paris: Masson. Lund, M., Reintoft, H. and Simonsen, N. (1976). Eine kontrollierte soziologische und psychologische untersuchung von patienten mit juveniler myoklonischer epilepsie. Nervenarzt, 47, 708–12. Meencke, H.-J. (1985). Neuron density in the molecular layer of the frontal cortex in primary generalized epilepsy. Epilepsia, 26, 450–4. Meencke, H.-J. and Janz, D. (1984). The significance of microdysgenesis in primary generalized epilepsies: a study of eight cases. Epilepsia, 25, 8–21.

59

Psychiatry of idiopathic generalized epilepsy Meencke, H.-J. and Janz, D. (1985). The significance of microdysgenesis in primary generalized epilepsy: an answer to the considerations of Lyon and Gastaut. Epilepsia, 26, 368–71. Méndez, M.F., Doss, R.C., Taylor, J.L. and Arguello, R. (1993). Relationship of seizure variables to personality disorders in epilepsy. J Neuropsychiatry Clin Neurosci, 5, 283–6. Mignone, R.J., Donelly, E.F. and Sadowsky, D. (1970). Psychological and neurological comparison of psychomotor epileptic patients. Epilepsia, 11, 345–59. Mirsky, A.F., Anthony, B.J., Duncan, C.C., Ahearn, M.B. and Kellam, S.G. (1991). Analysis of the elements of attention: a neuropsychological approach. Neuropsychol Rev, 2, 109–45. Mirsky, A.F., Duncan, C.C. and Levav, M.L. (1995). Neuropsychological and psychophysiological aspects of absence epilepsy. In Typical Absences and Related Syndromes, ed. J.S. Duncan and C.P. Panayiotopoulos, pp. 112–19. London: Churchill Communications. Niedermeyer, E. (1996). Primary (idiopathic) generalized epilepsy and underlying mechanisms. Clin Electroencephalogr, 27, 1–21. Onuma, T. (1983). Limbic lobe epilepsy with paranoid symptoms: Analysis of clinical features and psychological tests. Folia Psychiat Neurol Jpn, 37, 253–7. Ottmann, R., Annegers, J.F., Hauser, W.A. and Kurland, L.T. (1988). Higher risk of seizures in offspring of mothers than of fathers with epilepsy. Am J Hum Genet, 43, 257–64, Pedersen, S.B. and Petersen, K.A. (1998). Juvenile myoclonic epilepsy: Clinical and EEG features. Acta Neurol Scand, 97, 160–3. Pellock, J.M. (1999). JME: The elusive epilepsy. Perspect Pediat Neurol, 1, 1–6. Perini, G.I., Tosin, C., Carraro, C. et al. (1996). Interictal mood and personality disorders in temporal lobe epilepsy and juvenile myoclonic epilepsy. J Neurol Neurosurg Psychiatry, 61, 601–5. Pung, T. (2002). Schlaf-wachrhythmus und persönlichkeitbei juveniler myoklonischer epilepsie und patienten mit temporaler epilepsie. Thesis, Humboldt Universität Berlin. Reintoft, H., Simonsen, N. and Lund, M. (1976). A controlled sociological study of juvenile myoclonic epilepsy. In Epileptology, ed. D. Janz, pp. 48–50. Stuttgart: Thieme. Rosenthal, C. (1935). Die gehäuften kleinen Anfälle des Kindesalters (Pyknolepsie). Ergebn inn Med Kinderheilk, 48, 77–124. Sander, T., Schulz, H., Saar, K. et al. (2000). Genome search for susceptibility loci of common idiopathic generalised epilepsies. Hum Mol Genet, 9, 1465–72. Savard, G., Andermann, F., Olivier, A. and Remillard, G.M. (1991). Postictal psychosis after partial complex seizures: a multiple case study. Epilepsia, 32, 225–31. Schmitz, B. (1988). Psychosen bei epilepsie. Eine epidemiologische untersuchung. Thesis, Freie Universität, Berlin. Schmitz, B. (1992). Psychosis and epilepsy: the link to the temporal lobe. In The Temporal Lobes and the Limbic System, ed. M.R. Trimble and T.G. Bolwig, pp. 149–67. Petersfield, UK and Bristol, PA, USA: Wrightson Biomedical Publishing. Schmitz, B. and Wolf, P. (1991). Psychosis in epilepsy. In Epilepsy and Behaviour. Frontiers of Clinical Neuroscience, ed. O. Devinsky and W.H. Theodore, Vol. 12, pp. 97–128. New York: Wiley-Liss. Schmitz, B. and Wolf, P. (1995). Psychosis in epilepsy: frequency and risk factors. J Epilepsy, 8, 295–305.

60

D. Janz Sengoku, A., Yagi, K., Seino, M. and Wada, T. (1983). Risk of occurrence and psychosis in relation to the types of epilepsies and epileptic seizures. Folia Psychiat Neurol Jpn, 37, 221–6. Sengoku, A., Toichi, M. and Murai, T. (1997). Comparison of psychotic states in patients with idiopathic generalized epilepsy and temporal lobe epilepsy. Epilepsia, 38 (Suppl. 6), 22–5. Shukla, G.D., Srivastava, O.N., Katiyar, B.C. et al. (1979). Psychiatric manifestations in temporal lobe epilepsy: A controlled study. Br J Psychiatry, 135, 411–17. Simonsen, N., Mollgaard, V. and Lund, M. (1976). A controlled clinical and electroencephalographical study of myoclonic epilepsy (Impulsiv-Petit-Mal): preliminary report. In Epileptology, ed. D. Janz, pp. 41–8. Stuttgart: Thieme. Slater, E., Beard, A.W. and Glithero, E. (1963). The schizophrenia-like psychoses of epilepsy. Br J Psychiatry, 109, 95–150. Small, J.G., Milstein, V. and Stevens, J.R. (1962). Are psychomotor epilepsies different? Arch Neurol, 7, 187–94. Standage, K.F. and Fenton, G.W. (1975). Psychiatric symptom profiles of patients with epilepsy: a controlled investigation. Psychol Med, 5, 152–60. Stevens, J.R. (1966). Psychiatric implications of psychomotor epilepsy. Arch Gen Psychiatry, 14, 461–71. Stub-Naylor, A. (1996). Epilepsy and Psychiatric Disorder – A Comorbidity Study. Copenhagen: FADL Publishers. Swartz, B.E., Halgren, E., Simpkins, F. and Syndulko, K. (1994). Primary memory in patients with frontal and primary generalized epilepsy. J Epilepsy, 7, 232–41. Swartz, B.E., Simpkins, F., Halgren, E. et al. (1996). Visual working memory in primary generalized epilepsy. An FDG-PET study. Neurology, 47, 1203–12. Tellenbach, H. (1965). Epilepsie als anfallsleiden und als psychose. Über alternative psychosen paranoider prägung bei “forcierter normalisierung” (LANDOLT) des elektroencephalogramms epileptischer. Nervenarzt, 36, 190–202. Trimble, M. (2000). Cognitive and personality profiles in patients with juvenile myoclonic epilepsy. In Juvenile Myoclonic Epilepsy: The Janz Syndrome, ed. B. Schmitz and T. Sander, pp. 101–9. Petersfield, UK and Philadelphia, USA: Wrightson Biomedical Publishing. Vazquez, B., Devinsky, O., Luciano, D., Alper, K. and Perrine, K. (1993). Juvenile myoclonic epilepsy: clinical features and factors related to misdiagnosis. J Epilepsy, 6, 233–8. Wolf, P. (1976). Psychosen bei Epilepsie. Ihre Bedingungen und Wechselbeziehungen zu Anfällen. Habilitationsschrift. Berlin: Freie Universität. Wolf, P. (1982). Halluzinationen in Rahmen epileptischer Psychosen. In Halluzinationen bei Epilepsien und ihre Differentialdiagnose, ed. K. Karbowski, pp. 58–66. Bern: Huber. Wolf, P. (1984). The clinical syndromes of forced normalization. Folia Psychiatr Neurol Jpn, 38, 187–92. Wolf, P. (1985). Epilepsy with grand mal on awakening. In Epileptic Syndromes in Infancy, Childhood and Adolescence, 2nd edn (1992), ed. J. Roger, M. Bureau, Ch. Dravet, F.E. Dreifuss, A. Perret and P. Wolf, pp. 329–41. London: John Libbey. Wolf, P. (1986). Forced normalization. In Aspects of Epilepsy and Psychiatry, ed. M.R. Trimble and T.G. Bolwig, pp. 101–12. Chichester: John Wiley & Sons.

61

Psychiatry of idiopathic generalized epilepsy Wolf, P. and Trimble, M.R. (1985). Biological antagonism and epileptic psychosis. Br J Psychiatry, 146, 272–6. Yamashita, K., Aoki, K. and Kawai, I. (1984). Primary generalized epilepsies and temporal lobe epilepsies: a psychological study using P–F study. J Jap Clin Psychology, 2, 38–43.

5

Epilepsy and learning disorders Cesare Maria Cornaggia1 and Giuseppe Gobbi2 1 2

University of Milano Bicocca, Milan, Italy Maggiore Hospital, Bologna, Italy

Introduction In the practice of epilepsy, the association between epilepsy and learning disorder is very important, even if sometimes underestimated. A correct and early diagnosis of a learning disorder in a person with epileptic seizures will often determine future development and prognosis. Definitions Cognitive function defines the capacity of the human brain to process all information coming from the outside and internal world of the individual, and program ongoing behaviour (Aldenkamp and Bronswijk, 1999). This capacity involves the ability to remain in contact with the outside world (through the function of vigilance), to select and focus information (through the function of attention), and to memorize data (through the function of memory). In this way, cognitive function gives humans the opportunity of becoming aware of themselves and to solve problems – something that we also call intelligence. The latter is the generic capacity of using all of the elements of thinking necessary to recognize, plan and solve new problems in a directed and correct manner. More than one cortical area of the human brain is involved in cognitive function and related processes, and impaired cognitive function has been observed in the presence of a lesion or stable dysfunction in the temporal, frontal or parietal lobes of the dominant and nondominant hemispheres. An impairment in cognitive function may be seen as a reduction in the capacity of learning of children or reduction in the intellectual ability of adults. We normally refer to this as learning disability in children (which, at least in some countries, is different from mental retardation), and deterioration or dementia in adults or people who have already acquired mental capacity. From the terminological point of view, learning disorder is different from mental retardation. Actually, in some countries, such as the UK, learning disability is used 62

63

Epilepsy and learning disorders

for defining both a mental retardation and the learning disorders. But, in the Diagnostic and Statistical Manual of Mental Disorders (DSM–IV; American Psychiatric Association, 1994) learning disorders (DSM–IV 315.00–315.09) include a significant disturbance in academic achievement or daily living activities that require reading, mathematical or writing skills. In contrast, mental retardation is defined as a not necessarily lifelong disorder characterized by the combination of a low cognitive ability and diminished social and adaptive competence (DSM–IV 317–319) due to significantly subaverage intellectual functioning, with an onset before the age of 18 years. Actually, in most parts of the world, mental retardation defines a situation involving an abnormal IQ, and a learning disorder one involving a normal IQ. On the other hand, there are some doubts about the definition of mental retardation, especially in childhood epilepsy. In childhood epilepsy there is a risk of diagnosing the presence of mental retardation and failing to identify patients who are only affected by a specific learning disorder. Thus, it is known that an epileptic focus may lead to the dysfunction of a given specific performance, that coincides with the cortical area involved, without causing mental retardation. But, at a developmental age, when cognitive performance is progressing, a dysfunction in a specific cognitive area may cause a disequilibrium of the complex system of mental (cognitive) development, causing an apparent mental retardation, that may completely resolve if this specific cognitive dysfunction is adequately treated. The term ‘retardation’ itself may suggest that it is possible to make up the leeway, but we know that in many or perhaps the majority of cases of so-called mental retardation this is not to be expected, especially in the presence of brain lesions or some special epileptic syndromes. In some countries (including Italy), in fact, the term mental deficiency is used instead of mental retardation for defining situations assuming a definite deficit. This is the case, also, with the term disability. Disability means a restriction of the ability to perform an activity and it is generally considered to be stable and irreversible, at least in some countries such as Italy. In the case of epilepsy it could be more appropriate to use the word disorder because this opens up the possibility that the loss of capacity may be transitory or reversible, so-called state dependency as discussed by Besag (1989, 1995, Chapter 6). Learning disorders are linked to various types of other behaviour disturbances that must be separately classified: for example, the Attention Deficit/Hyperactivity Disorder (ADH; DSM–IV 314.00–314.01), which may or may not be associated with behavioural or mood disturbances (DSM–IV 312.8). We also know that learning disorders are associated with other disorders, such as conduct disorders, major depressive disorders, dysthymic disorders and opposition defiant disorder, in 10 to 25% of the cases.

64

C.M. Cornaggia and G. Gobbi

Learning and behavioural disorders are linked both in childhood and in adults. Although it is true that behavioural or mood disorders may be seen as a symptomatic expression of a cognitive impairment during epilepsy, it is also true that behavioural or mood disorders may lead to reduced learning ability. Prevalence There is no doubt that a disproportionate percentage of children with epilepsy have learning and behavioural problems (Bourgeois, 1998), and children with behavioural disturbances are certainly a minority in respect to the number of those with learning disorders. The concept that epilepsy is invariably associated with progressive intellectual deterioration both in children and in the adult population, which prevailed during the first half of the twentieth century (Bourgeois, 1998), cannot be accepted any more. In fact, a large proportion of children with epilepsy have some schooling difficulties (Ross et al., 1980, Sillanpää et al., 1999), but fewer than 1% attend special epilepsy schools. Nakken and Brodtkorb (1999) have recently reported that about 20% of the epilepsy population are mentally retarded (Forsgren et al., 1990, 1996), and about 20% of mentally retarded people have epilepsy (Wester, 1982). In reality, only a small subgroup of children with epilepsy shows stable decreases in IQ, as recently shown by Besag (1995), who demonstrated that the observation of a declining IQ at any given moment does not necessarily indicate a loss of skills, but simply implies that the child is not developing at the expected rate. According to some authors, this lag between mental and chronological age is more pronounced in older children, and in children with an earlier seizure onset, a higher lifetime total number of seizures, or multiple seizure types (Seidenberg et al., 1986). There is some evidence indicating that children aged less than ten years at the time of assessment gain less than those with an age at onset of more than ten years (Neyens et al., 1999). Causes of learning disorders There are three specific ways in which epilepsy and learning disorders may be related. First, epilepsy and learning disorders are the result of brain damage and/or permanent brain dysfunction. Secondly, epilepsy may cause brain damage or permanent brain dysfunction, which in turn leads to a learning disorder. Status epilepticus would be the classical example, as there is no doubt that epilepsy can cause deterioration in children experiencing prolonged bouts of status epilepticus. Thirdly, epilepsy may cause learning disorders directly, without brain damage or permanent brain dysfunction, such as in the case of continuous epileptiform EEG

65

Epilepsy and learning disorders

discharges or subtle seizures, and there is also some evidence that certain ‘sources of seizure discharge’ are associated with selective cognitive deficits (Stores, 1985) for example the Landau–Kleffner syndrome. Classification of learning disorders Learning disorders occurring during epilepsy can be classified into two categories: state-dependent (potentially treatable and reversible) and permanent. The prevalence of state-dependent or permanent learning disorders is not satisfactorily known. Brain damage or stable brain dysfunction is mainly associated with permanent learning disorders. Epilepsy itself or the medication used to treat it may lead to state-dependent learning disorders. The picture is further complicated by associated mood disorders or psychoses, a low level of self-perception and expectation and reduced learning opportunities. Regarding epilepsy itself, various situations can be recognized as leading to a state-dependent learning disorder. • Direct effects of seizures. There is evidence that single complex partial and secondarily generalized seizures are associated with neuronal damage (Rabinowicz et al., 1996), and that brain extracellular glutamate may build up to neurotoxic levels in partial seizures (During and Spencer, 1993). Perhaps, not surprisingly, a reduction in seizure frequency over time is associated with an improvement in cognitive functioning (Seidenberg et al., 1981), and the presence of discharges affects discrete scholastic performances (Kasteleijin-Nolst Trenite et al., 1988). • Special epileptic syndromes. Typical examples are West syndrome (WS) and Lennox–Gastaut syndrome (Gokyigit and Caliskan, 1995; Oguni et al., 1996). • Ictal changes. Nonconvulsive status epilepticus, which may continue for months or years is an example. • Peri-ictal changes. This situation refers to patients with so many seizures every day that they do not have time to recover from one to another. Such patients are effectively in a continuous postictal state, which may be misinterpreted as a permanent learning disorder. • The presence of frequent subtle seizures and the direct effects of interictal EEG epileptiform discharges. Children with epilepsy who show epileptiform discharges during IQ tests do less well than those who do not (Siebelink et al., 1988). The Transitory Cognitive Impairment (TCI) due to the direct effects of frequent interictal EEG epileptiform discharges or frequent subtle seizures, is a typical example. If these are very frequent, they may represent a nonconvulsive status epilepticus. In fact, state-dependent learning disorders may be a consequence of TCI, associated with either focal or generalized specific epileptiform EEG discharges. In 1939

66

C.M. Cornaggia and G. Gobbi

Schwab demonstrated that generalized spike-wave discharges without any apparent clinical concomitants may be associated with a simultaneous decline in cognitive function. TCI was also reported in 1957 by Kooi and Hovey, who studied a case involving focal discharges. These subclinical or ‘larval’ discharges are accompanied by a TCI which, if not recognizable from the patient’s spontaneous behaviour, can be detected by appropriate psychological tests during EEG monitoring (Binnie, 1979). TCI is most readily demonstrated by difficult tasks and during generalized regular spike-wave discharges lasting more than 3 seconds, but it can also be found with briefer and focal discharges. In this last case, the errors are more specific: leftsided focal discharges are more likely to produce errors in verbal tasks, whereas right-sided discharges are associated with the impaired handling of non-verbal material. This means that TCI is not only a consequence of global inattention, and it has been hypothesized that the errors may be due to a working memory disorder (Aarts et al., 1984). The effect of a TCI accompanying subclinical EEG discharges on everyday functioning is still uncertain and needs further examination, but there is experimental evidence that subclinical discharges may be accompanied by a disruption of educational skills in children, and the AED suppression of discharges is concomitant with cognitive improvement. It has been suggested that, if allowed to continue for a long time, state-dependent learning disorders may produce permanent learning disorders. Strong evidence for this comes from some rare models of epilepsy, e.g. Landau–Kleffner syndrome and continuous spike-wave in slow wave sleep (CSWS). Certainly, the extent to which subtle seizure manifestations affect learning and behaviour remains unknown. This suggests that there are major controversial questions to be asked regarding the importance of subtle manifestations of epilepsy on learning and behaviour. Further, it is possible that some children diagnosed as having ADHD or autism may actually be affected by undiagnosed manifestations of epilepsy. It is certainly a common experience to observe that many normally underestimated or undiagnosed severe epileptic conditions, such as some frontal lobe epilepsies with frequent specific EEG activity (including rapid activity lasting between one and one-and-a-half seconds), are subsequently associated with the onset of severe behavioural disturbances (sometimes with psychotic symptoms), or schizophrenia-like or autism-like syndromes. The existence of situations such as those reported above underlines the need to clarify the causative role of epilepsy or the presence of nonconvulsive EEG epileptiform discharges in childhood and adult behavioural problems. We must not be afraid to use the EEG in children and in adults showing behavioural or learning disturbances. Similarly, we believe that the issue of intervening to ameliorate interictal EEG discharges has to be reconsidered. Of course, that needs care and has to be

67

Epilepsy and learning disorders

evaluated in each single patient, since it is difficult to ensure that the drugs themselves do not disrupt cognitive processing. The role of treatments The data available in the literature indicate that the administration of antiepileptic drugs (AEDs) can impair cognitive function and affect the behaviour of children and adults, but the extent and severity of this is still unknown. Moreover, AEDs may cause state-dependent learning disorders, but evidence is anecdotal. Some recent studies have compared the efficacy and presence of the cognitive and behavioural side effects of various AEDs. For example, it has been found that the efficacy of vigabatrin, lamotrigine and gabapentin is similar, but that vigabatrin is associated with a higher incidence of behavioural problems (Bhaumik et al., 1997). On the other hand, an overall improvement in attention and school performance has been reported in a group of children treated with vigabatrin monotherapy in respect to a carbamazepine monotherapy control group (Gobbi et al., 1999). The results of other studies support the hypothesis that the cognitive profile of lamotrigine is similar to that of carbamazepine (Aldenkamp et al., 1997). In general, an AED may improve learning by reducing the number of EEG discharges or the frequency of seizures but, at the same time, it may impair learning as a result of its side effects – sleepiness, slowed reactions, attention deficit and so on. It is important not to forget clinical observations and manage each single case separately. Finally, learning disorders need to be considered in the preoperative evaluation and postoperative rehabilitation of patients with epilepsy. In the preoperative phase, it is important to consider the effect that the surgical ablation of the area of the epileptic focus may have on the processes of learning. The ablation of an epileptogenic area generally produces positive effects because it eliminates the negative consequences of the epileptic discharges that go from this to other parts of the brain. However, if the area of the epileptogenic focus has not completely lost its own primary function, the same ablation may lead to a functional deficit. In this case, even a satisfactory improvement in seizure frequency may lead to very negative postoperative results. Postoperative rehabilitation must take into account the developmental level of the individual, and may take weeks or months of intensive specialist input but, provided that this is carefully planned and provided, the outcome can be very good. Conclusions Tentatively, we may summarize the actual picture as follows. Epilepsy may be associated with learning disorders, and also behavioural or mood disorders. Behaviour

68

C.M. Cornaggia and G. Gobbi

and learning are often linked; behavioural changes may result from temporary or permanent cognitive deficits. Learning disorders in epilepsy may be state dependent or permanent; state-dependent learning disorders may produce permanent learning disorder. The prevalence of both conditions is not known. Strong evidence now shows that very frequent overnight and day-time epileptiform discharges may eventually result in permanent learning disorder. Epilepsy itself, AEDs and social factors may influence the onset of learning disorders. Although a large percentage of children with epilepsy have school difficulties, only a minority shows a loss of skills or a decreasing IQ. For children in whom actual loss of skills occurs, it is essential to look for a specific cause, and to plan an ad hoc management programme.

R E F E R E N C ES Aarts, J.H.P., Binnie, D.J., Smit, A.M. et al. (1984). Selective cognitive impairment during focal and generalized epileptiform EEG activity. Brain, 107, 293–380. Aldenkamp, A.P. (1997). Effect of seizures and epileptiform discharges on cognitive function Epilepsia, 38 (Suppl. 1), 52–5. Aldenkamp, A.P. and Bronswijk, K. (1999). Cognitive side effects as an outcome measure in antiepileptic drug treatment: the current debate. In Epilepsy and Mental Retardation, ed. M. Sillanpaa, L. Gram, S.I. Johannessen and T. Tomson, pp. 135–146. Petersfield: Wrightson Biomedica Publishing. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (Fourth edition) (DSM–IV). Washington, DC: APA. Besag, F.M.C. (1989). Epilepsy, learning and behaviour. Educ Child Psychol, 6. Besag. F.M.C. (1995). Epilepsy, learning, and behavior in childhood. Epilepsia, 36 (Suppl. 1), 58–63. Bhaumik, S., Branford, D., Duggirala, C. and Ismail, I.A. (1997). A naturalistic study of the use of vigabatrin, lamotrigine and gabapentin in adults with learning disabilities. Seizure, 6, 127–33. Binnie, C.D. (1979). Direction of transitory cognitive impairment during epileptiform EEG discharges: problems in clinical practice. In Epilepsy and Behaviour, ed. B.M. Kulig, H. Meinardi and G. Stores. Lisse: Swets & Zeitlinger. Bourgeois, B.F.R.D. (1998). Antiepileptic drugs, learning, and behavior in childhood epilepsy. Epilepsia, 39, 913–21. During, M.J. and Spencer, D.D. (1993). Extracellular hippocampal glutamate and spontaneous seizure in the conscious human brain. Lancet, 341, 1607–10. Forsgren, L., Edvinsson, S.-O., Blomquist, H.K., Heijbel, I. and Sidenvall, R. (1990). Epilepsy in a population of mentally retarded children and adults. Epilepsy Res, 6, 234–48. Forsgren, L., Bucht, G., Eriksson, S. and Bergmark, L. (1996). Incidence and clinical characterization of unprovoked seizures in adults: a prospective population-based study. Epilepsia, 37, 224–9.

69

Epilepsy and learning disorders Gobbi, G., Pini, A., Bertani, G. et al. (1999). Prospective study of first-line vigabatrin monotherapy in childhood partial epilepsies. Epilepsy Res , 35, 29–37. Gokyigit, A. and Caliskan, A. (1995). Diffuse Spike–Wave status of nine year duration without behavioural change or intellectual decline: Epilepsia, 36, 210–13. Kasteleijin-Nolst Trenite, D.G., Bakker, D.J., Binnie, C.D., Buerman, A. and Van Raaij, M. (1988). Psychological effects of subclinical epileptiform EEG discharges. I Scholastic skills. Epilepsy Res, 2, 111–16. Kooi, K.A. and Hovey, H.B. (1957). Alterations in mental functions and paroxysmal cerebral activity. Arch Neurol Psychiatry, 1978, 264–71. Nakken, K.A. and Brodtkorb, E. (1999). Epilepsy services for the patient with mental retardation in Norway. In Epilepsy and Mental Retardation, ed. M. Sillanpää, L. Gram, S.I. Johannessen and T. Tomson, pp. 193–200. Petersfield: Wrightson Biomedical Publishing. Neyens, L.G.J., Aldenkamp, A.P. and Meinardi, H.M. (1999). Prospective follow-up of intellectual development in children with a recent onset of epilepsy. Epilepsy Res, 34, 85–90. Oguni, H., Hayashi, K. and Osawa, M. (1996). Long-term prognosis of Lennox–Gastaut syndrome. Epilepsia, 37 (Suppl. 3), 44–7. Rabinowicz, A.L., Carreale, J., Buotros, R.B., Coulwell, W.T., Henderson, C.W. and DeGiorgio, C.M. (1996). Neuron-specific enolase is increased after single seizures during inpatient videoEEG monitoring. Epilepsia, 37, 122–5. Ross, E.M., Peckham, C.S. and West, P.B. (1980). Epilepsy in childhood: findings from the National Child Development Study. Br Med J, 1, 207–10. Schwab, R.S. (1939). A method of measuring consciousness on petit-mal epilepsy. J Nerv Ment Dis, 89, 690–1. Seidenberg, M., O’Leary, D.S., Berent, S. and Boll, T. (1981). Changes in seizure frequency and test-retest scores on Wechler adult intelligence scale. Epilepsia, 22, 75–83. Seidenberg, M., Beck, N. and Geisser, M. (1986). Academic achievement of children with epilepsy. Epilepsia, 27, 753–9. Siebelink, B.M., Bakker, D.J., Binnie, C.D. and Kasteleijn-Nolst Trenite, D.G. (1988). Psychological effects of subclinical epileptiform EEG discharges in children. II general intelligence tests. Epilepsy Res, 2, 117–21. Sillanpää, M., Gram, L., Johannessen, S.I. and Tomson, T. (1999). Epilepsy and mental retardation. Petersfield: Wrightson Biomedical Publishing. Stores, G. (1985). Clinical and EEG evaluation of seizures and seizures-like disorders. J Am Acad Child Psychiatry, 24, 10–16. Wester, K. (1982). Epilepsi hos psykik utviklingshemmede. Tidsskr Nor Laegeforen, 102, 225–8.

6

Subtle cognitive and behavioural effects of epilepsy Frank M.C. Besag Bedfordshire and Luton Community NHS Trust, Bedford, UK

Introduction Both cognitive and behavioural problems are common in people with epilepsy. Epidemiological studies have indicated that around 50% of children with epilepsy have some schooling difficulties and that many have behavioural problems (Besag et al., 1999; Pazzaglia and Frank-Pazzaglia, 1976; Ross et al., 1980; Sillanpää, 1992). The National Child Development Study in the UK (Ross et al., 1980) revealed that only 67% of children with epilepsy were attending a mainstream school at age 11. Using this broad-brush measure it was evident that a large proportion of the children with epilepsy had significant schooling problems. The work of Pazzaglia and Pazzaglia in Cesena, Italy, (Pazzaglia and Frank-Pazzaglia, 1976) also confirmed that a high proportion of children were underachieving at school. The excellent epidemiological studies of Sillanpää (Sillanpää, 1992) in Finland showed that 31.4% of this unselected sample of children had mental retardation. In a recent study carried out in the London Borough of Lambeth Besag et al. (1999) surveyed 127 children with epilepsy. The parents perceived that the children had schooling problems or learning disability in 65% and 48% were disturbed on Rutter Behavioural Scales. Sixty-two per cent of the scores indicated a significant impact on quality of life which was moderate or great in 35%. These studies suggest that schooling and behavioural difficulties constitute a major problem in childhood epilepsy. There is a lack of data, however, to indicate the causes of the problems encountered, either in children or adults. A systematic framework for assessing behaviour The current author has suggested that, when faced with an individual who has epilepsy and behavioural problems, the best approach is to use a systematic framework for assessing the possible cause or causes of the behavioural disturbance (Besag et al., 1989a; Besag, 1995). This framework consists of five main categories: the epi70

71

Subtle effects of epilepsy

lepsy itself, treatment of the epilepsy, reactions to the epilepsy, associated brain damage/dysfunction and causes equally applicable to people without epilepsy. The first category, the epilepsy itself, may be broken down into the peri-ictal phenomena of prodrome, aura, automatism and postictal changes, interictal psychoses, focal discharges and frequent absence seizures. This list is not necessarily comprehensive. Examples of the way in which the epilepsy itself may affect behaviour have been provided elsewhere(Besag et al., 1989a). A subtle cognitive or behavioural effect of epilepsy may be defined as an effect that is not immediately obviously attributable to an epileptic seizure. This does not imply that the epileptic activity itself is necessarily ‘subtle’. For example, prodromal mood change preceding a tonic-clonic seizure could be regarded as a subtle effect of the epilepsy, although the seizure itself is not at all subtle. On the other hand, the subtle cognitive and behavioural effects of epilepsy may be the result of subtle seizure activity. In this context, subtle seizure activity is taken to mean a clinical manifestation of epileptiform activity that is not immediately obviously identifiable as a seizure. Absence seizures and many complex partial seizures would fall into this category as would transitory cognitive impairment. There is a growing realization of the fact that epilepsy may affect cognition and behaviour in a variety of ways (Aldenkamp, 1997; Deonna, 1995; Stores and Hart, 1975; Stores, 1978; Trimble, 1988). There is often an overlap between cognitive and behavioural effects of epileptiform activity. Frequent frontal discharges may lead to a high degree of social disinhibition and consequent behavioural disturbance. A teenager had a long history of gross behavioural disturbance. The EEG showed very frequent left frontal discharges, typically occurring every second. He underwent a left frontal lobectomy, following which the epileptiform discharges were abolished, his behavioural disturbance resolved and he returned to his pleasant premorbid personality.

Frequent left temporal discharges may be associated with aggressive behaviour. There is evidence from the literature suggesting that young men who have left temporal discharges and subsequently undergo a left temporal lobectomy show improvements both in seizure control and in behaviour (Falconer, 1973). Further examples of both cognitive and behavioural disturbance that can result from subtle manifestations of epilepsy will be discussed in the next section. State-dependent cognitive impairment It is important to distinguish permanent cognitive impairment on one hand from state-dependent cognitive impairment on the other. The concept of permanent

72

F.M.C. Besag

cognitive impairment is readily understood. This may arise from a wide variety of causes of permanent brain damage or dysfunction that may be prenatal, perinatal or postnatal. The concept of state-dependent cognitive impairment is less widely acknowledged (Besag, 1994). What is state-dependent cognitive impairment? State-dependent cognitive impairment may be defined as cognitive impairment that depends on current factors affecting the individual, for example antiepileptic medication or epileptic activity, that are not necessarily permanent. State-dependent cognitive impairment is potentially reversible and treatable. Failure to treat state-dependent cognitive impairment is failure to provide an adequate service to the patient. Some might put the case even more strongly, suggesting that failure to recognize and treat statedependent cognitive impairment is a reflection on professional competence. The first step in managing state-dependent cognitive impairment is to think of the diagnosis. Regrettably, professionals often fail to take this first step. The result is that the condition is neither recognized nor treated. State-dependent cognitive impairment may be divided into two broad categories: drug-induced and epilepsyinduced. As already indicated, the epilepsy itself may cause state-dependent cognitive impairment in a number of different ways. Frequent absence seizures

Frequent absence seizures, by interrupting awareness, can affect both cognitive performance and behaviour. People who are having frequent absence seizures may present as having withdrawn behaviour, fragmented thought processes which may be mistaken for a psychosis, attention-deficit disorder with motor overactivity or, if the frequency of the seizures is variable, attention-seeking behaviour. The last of these behaviours tends to be seen when the person emerges from a bout of very frequent absence seizures. It is almost as if the child is ‘making up for lost time’ in being badly behaved when he has the opportunity of doing so, having been unable to misbehave when the absence seizures were very frequent. In some cases the absence seizures may become so frequent as to amount to nonconvulsive status epilepticus. The person is effectively cut off from his or her surroundings because he or she does not have the opportunity to function adequately between the absence seizures. Complex partial seizure status epilepticus may also cause a continuous state of behavioural and cognitive change until the status epilepticus is either treated or resolves spontaneously. It is clear that absence seizures can affect both performance and confidence in a major way, particularly if they are occurring frequently. It is not possible to count absence seizures by simple observation alone. To overcome this difficulty, an automatic spike-wave monitor, the ‘Monolog’ was developed (Besag et al., 1989b). The ‘Monolog’ is so-called because it monitors and logs spike-wave episodes automat-

73

Subtle effects of epilepsy

ically. Two recording electrodes and a reference electrode are attached to the scalp. The leads are brought down to a small solid-state recording device, which is in a belt-worn pouch. This device provides three pieces of hard data: the total duration of the spikewave episodes over the period of recording, the number of spike-wave episodes and, by playing the solid state circuitry into a chart recorder, the position of each spike-wave episode in time. In addition, there is an optional plug-in lapel badge with a flashing light and bleeper, activated by the presence of the spike-wave episodes. The teacher or carer is alerted by the flashing light/bleeper and he or she will realize that the child is having a spike-wave episode; it may be appropriate for the teacher/carer to repeat what has been said to the child during that episode. The light and bleeper also serve to alert the teachers and carers to the high frequency of absence seizures occurring in some children. Because absence seizures are often quite subtle in their manifestation, the teacher may be surprised to note how often the badge indicates that spike-wave discharges are occurring. This in itself may result in an alteration of attitude by the teacher who may have an increased inclination to repeat information if the child has not obviously taken it in on the first occasion. Using this monitor it has been possible to show that some children have not only hundreds but thousands of spike-wave episodes daily. The device also confirms whether treatment is effective or not. A 13-year-old boy with a history of epilepsy was reported as having reached the stage at which the epilepsy was not a problem for him. He was having relatively infrequent overt seizures. However, he presented as a withdrawn child who would not join group activities. He preferred to sit in the corner of the room sucking his thumb. In the clinic the examiner uttered the child’s name on four occasions in succession. On the first three occasions there was no response because he was momentarily unconscious in a subtle absence seizure. On the fourth occasion, when he was not having an absence seizure, he responded promptly and was more than willing to attend to whatever task the examiner asked him to do. Prolonged EEG monitoring revealed that around three thousand spike-wave episodes occurred daily. He responded very well to treatment.

Some individuals have both more obvious seizures, such as tonic-clonic or tonic seizures and absence seizures. The antiepileptic medication may improve both the obvious and the subtle seizures. However, in some cases the obvious seizures may not be affected by medication while the subtle seizures are greatly reduced. Such individuals may become bright, alert and in control of their lives as a result of the medication change, even though the obvious seizures have not been reduced in frequency. When patients are assessed in the outpatient clinic the doctor will typically ask how many seizures have been recorded or will consult the patient’s seizure diary. If there has been no reduction in the obvious seizures the doctor may choose to discontinue the medication with which the patient has been treated and try another

74

F.M.C. Besag

(a) Overt seizures (a) in 3 months

120

4000

(b) Spike-wave events (b) in 24 hours

Number of events

Number of seizures

100 80 60 40

3000

2000

1000

20 0

0 Baseline

1–3

4–6

7–9

Months Figure 6.1.

Baseline

2

4

8

10

Weeks

The effect of lamotrigine administration on number of overt seizures (a) and spike-wave events (b) experienced by Subject A.

antiepileptic drug. However, on this basis the doctor might stop a drug that has been highly effective in treating subtle absence seizures, even though there has been no major effect on the obvious seizures. An example is provided in Figures 6.1 and 6.2. In Case A there has been a reduction both in the obvious seizures and in the spike-wave episodes with the addition of lamotrigine. In Case B, however, the obvious seizures have not decreased. The spike-wave episodes, on the other hand, have been reduced by over 2000 per day. The parents of this teenager were delighted with the transformation. He was able to relate much more readily to the world around him and appeared much more alert. It is important not to discontinue medication on the basis that obvious seizures have not responded if the individual has had a good response in terms of reduction of subtle seizure activity. Transitory cognitive impairment A number of papers describing this phenomenon have been published by Binnie and coworkers (Aarts et al., 1984; Marston et al., 1993). The basic concept is that an epileptiform discharge that does not appear, on simple observation, to be manifesting as a seizure, may nevertheless cause a transitory impairment of cognitive function. It has been shown that discharges on the left side may impair language function and on the right side may impair visuo-spatial skills. The quality of the impairment is not always straightforward. For example, a child who is having epi-

75

Subtle effects of epilepsy (b) Spike-wave events in 24 hours

120

3000

100

2500 Number of events

Number of seizures

(a) Overt seizures in 3 months

80 60 40

2000 1500 1000 500

20

0

0 Baseline

Figure 6.2.

1–3

4–6 Months

2

7–9 Baseline

4

6

10

32

48

Weeks

The effect of lamotrigine administration on number of overt seizures (a) and spike-wave events (b) experienced by Subject B.

leptiform discharges during reading may pause or may read even more quickly but make additional mistakes when discharges are occurring. Binnie and his co-workers have shown that reduction of epileptiform discharges can result in improvement of psychosocial functioning (Marston et al., 1993), although this study was confounded by the fact that obvious seizures were improved as well as the discharges causing the transitory cognitive impairment. Observation of video tapes taken during testing of young people who have transitory cognitive impairment leaves no doubt about the fact that they find their lapses in performance irritating and frustrating. It is highly likely that this phenomenon affects self-esteem and self-confidence. This implies that, although the phenomenon itself is transitory, there may be an ongoing negative effect on the attitude and behaviour of the individual. Frequent localized discharges Reference has already been made to frequent frontal discharges and frequent left temporal discharges affecting behaviour. Localized discharges can also affect ongoing cognitive performance. This effect is more than just transitory if the discharges are frequent. A girl had a history of a benign left-sided temporal lobe tumour. She had very frequent epileptiform discharges arising from the left temporal lobe. She underwent left temporal lobectomy at 13 years of age. The epileptiform discharges disappeared after the neurosurgery. In a single year her speech development improved from around the 4-year level to the 7-year

76

F.M.C. Besag level. The temporal lobectomy released the remaining brain from the effect of very frequent epileptiform discharges, allowing rapid progress to be made.

Frequent hemispheric discharges The extreme example of frequent localized discharges is provided by individuals who have an abnormal hemisphere which is the source of such discharges. A number of different pathological conditions can give rise to this phenomenon, including hemimegalencephaly, a porencephalic cyst and the unilateral cerebral atrophy that accompanies Rasmussen’s encephalitis. These individuals often improve markedly after hemispherectomy. Not only are the seizures and epileptiform discharges abolished but the behaviour also improves greatly(Goodman, 1986). If there is a gross structural abnormality that is causing very frequent epileptiform discharges then early surgery should be considered. If the lesion responsible has been present from a very early age, for example if it is a congenital porencephalic cyst, then the outcome is likely to be good. A dilemma arises in the case of the child who develops Rasmussen’s encephalitis that is causing progressive hemiatrophy of one hemisphere. The function subserved by the affected hemisphere is likely to worsen and the longer the operation is delayed the less opportunity there will be for the healthy hemisphere to take over function. However, hemispherectomy usually results in loss of finger function on the affected side and there is a tendency to delay the operation until it is clear that the finger function has already been lost, implying that the surgery will not impair function further. Is it appropriate to wait? Because brain plasticity is greatest when the child is young, there is a better chance that the remaining hemisphere will take over function if the surgery is performed early. There is a strong argument, in such cases, for carrying out the surgery sooner rather than later if the assessment indicates that the disease in the affected hemisphere is progressive. Postictal state-dependent cognitive impairment At first sight it may appear trivial to discuss postictal cognitive impairment because this is such an obvious phenomenon. However, it is not always as obvious as one might imagine. An apparently dementing teenager sat rocking in his chair, not knowing where he was, barely able to carry out a rudimentary conversation and unable to perform his skills of daily living. He had been accepted as a boarding pupil in a special residential centre for children with epilepsy and other needs. The staff felt that he had been placed inappropriately. They commented that he could not learn and that he should not have been accepted. At that

77

Subtle effects of epilepsy stage he was having between three and five seizures a day. Following a medication review he became seizure-free. He was then fully orientated and began progressing very well in his educational programmes. Both his parents and the staff were delighted with his progress. The staff then commented how appropriate the placement was. They had previously assumed that his cognitive impairment was permanent whereas, in fact, a large component of the cognitive impairment was state-dependent, treatable and reversible. When he had been having frequent daytime seizures he had not had the chance to recover from one seizure before he had another. He was in a constant postictal state. When he emerged from this constant postictal state he was able to learn again and progressed rapidly.

It is important to be aware of the possibility of reversible, state-dependent cognitive impairment in an individual who is in a constant postictal state from frequent seizures. Reducing the seizure frequency can improve cognitive function markedly. Overnight video-telemetry has revealed that some people have very frequent unsuspected and unobserved nocturnal seizures. These seizures may be brief, silent and easily missed, even by awake night staff. In a series of 15 patients examined by the author and his coworkers, large numbers of nocturnal seizures were recorded by video-telemetry, that had been unobserved by awake night staff. In one case over 200 brief nocturnal tonic seizures were recorded. Some individuals are woken by the seizures. Frequent nocturnal seizures may affect daytime performance not only because of the direct after-effects of the seizures themselves but also because they may cause a very broken night’s sleep.

Electrical status epilepticus of slow wave sleep The Landau–Kleffner syndrome of acquired epileptic aphasia is classically associated with electrical status epilepticus of slow wave sleep (ESES) (Beaumanoir, 1992; Landau and Kleffner, 1957). However, a series of six cases of ESES examined by the author and colleagues confirmed that language is not necessarily the function that is impaired. Although some of these children clearly had language impairment in association with the ESES, others did not. A boy with a right congenital porencephalic cyst and an accompanying left hemiparesis had ESES. He had good language skills but his visuo-spatial skills became increasingly impaired. He was unable to find his way from his bedroom to the bathroom at home. Following neurosurgery his visuo-spatial skills improved greatly.

The Landau–Kleffner syndrome provides an important model of state-dependent cognitive impairment. Some of the older textbooks suggested that antiepileptic treatment was only of value in treating the seizures but would not affect the cognitive impairment. Such statements were probably based on ‘burnt-out’ cases in which the damage had become permanent. It is quite clear that early, energetic treatment

78

F.M.C. Besag

may improve the cognitive impairment in some children with Landau–Kleffner syndrome. The implication is that not only may cognition be improved by early treatment but that permanent impairment may be avoided. Treatments include antiepileptic medication such as sodium valproate and lamotrigine, steroid treatment and neurosurgery. Multiple subpial transection, pioneered by Morrell (Morrell et al., 1989), has been particularly effective in some cases. This can abolish the ESES and allow the language function to return. The results obtained from multiple subpial transection in this condition have made it clear that past statements that antiepileptic treatment could not help cognitive loss were wrong. It is worth repeating that early treatment may not only result in a return of these skills but may also prevent long-term cognitive impairment. The issues raised by ESES have led to the suggestion that any child who loses skills should be investigated fully and, if another cause is not found, investigations should include overnight EEG monitoring. It should be noted, in this context, that it is said that around 25% of the children with Landau–Kleffner syndrome do not have a previous history of obvious seizures (Beaumanoir, 1992). Preventing cognitive and behavioural problems In the examples just given, it was emphasized that some children with ESES may present with impaired cognition. They often also present with markedly disturbed behaviour. Both the cognitive and behavioural disturbance are reversible with early treatment. It has become evident that a number of children who present with markedly autistic features have unsuspected epileptiform discharges either during the day or at night (ESES or continuous spike-wave in slow wave sleep – CSWS). Treating these children allows them to emerge from their apparently autistic state. However, follow-up of teenagers who had very frequent absence seizures in childhood has indicated that some longer-term problems in social interaction or behaviour may be evident. Some teenagers who have had very frequent epileptiform EEG discharges earlier in life may subsequently still have autistic features when the epileptiform discharges are no longer present. It would appear that these young people have been unable to interact adequately with the world around them during a critical developmental phase because of the frequent epileptiform discharges at that time. If they have not had the opportunity to develop two-way social interaction during this developmental phase they may present with a very Asperger-like picture during teenage years. This suggests that early treatment of the epileptiform discharges, so as to allow the child to gain fully from the early developmental years, may have avoided these social impairments. In summary, it appears that there is now an increasing body of evidence to

79

Subtle effects of epilepsy

suggest that frequent epileptiform discharges should not be allowed to continue for very long periods because they might cause not only permanent cognitive impairment but also permanent social/behavioural problems. Conclusions Systematic assessment of cognitive and behavioural problems often allows the cause to be found and rational management to be provided. It is very important to consider the possibility of state-dependent cognitive impairment, especially in the child who has lost skills. Unless another obvious cause is found, the investigations should include overnight EEG monitoring. If frequent epileptiform discharges are found in association with cognitive or behavioural problems then early, energetic treatment should be initiated. Such treatment may not only improve the impairments at the time but may also prevent permanent cognitive or social impairments.

R E F E R E N C ES Aarts, J.H., Binnie, C.D., Smit, A.M. and Wilkins, A.J. (1984). Selective cognitive impairment during focal and generalized epileptiform EEG activity. Brain, 107, 293–308. Aldenkamp, A.P. (1997). Effect of seizures and epileptiform discharges on cognitive function. Epilepsia, 38 (Suppl. 1), S52–5. Beaumanoir, A. (1992). The Landau–Kleffner Syndrome. In Epileptic Syndromes in Infancy, Childhood and Adolescence, ed. J. Roger, M. Bureau, Ch. Dravet, F.E. Dreifuss, A. Perret and P. Wolf, pp. 231–43. London: John Libbey. Besag, F.M.C. (1994). Epilepsy, education and the role of mental handicap. In Epilepsy, ed. E.M. Ross and R.C. Woody, pp. 561–83. Baillieres Clin Paediatr Int Practice Res, 2, 561–83. Besag, F.M.C. (1995). Epilepsy, learning, and behavior in childhood. Epilepsia, 36, S58–63. Besag, F.M.C., Loney, G., Waudby, E., Fowler, M. and Brooks, N. (1989a). A multidisciplinary approach to epilepsy, learning difficulties and behavioural problems. Educational Child Psychol, 6, 18–24. Besag, F.M., Mills, M., Wardale, F., Andrew, C.M. and Craggs, M.D. (1989b). The validation of a new ambulatory spike and wave monitor. Electroencephalogr Clin Neurophysiol, 73, 157–61. Besag, F., O’Neill, C. and Ross, E. (1999). A comparison between children with epilepsy in an inner-city region and those within a special centre, using measures of educational difficulty, behavioural problems and quality of life. Epilepsia, 40, 243. Deonna, T. (1995). Cognitive and behavioral disturbances as epileptic manifestations in children: an overview. Semin Pediatr Neurol, 2, 254–60. Falconer, M.A. (1973). Reversibility by temporal-lobe resection of the behavioral abnormalities of temporal-lobe epilepsy. N Eng J Med, 289, 451–5. Goodman, R. (1986). Hemispherectomy and its alternatives in the treatment of intractable epilepsy in patients with infantile hemiplegia. Dev Med Child Neurol, 28, 251–8.

80

F.M.C. Besag Landau, W.M. and Kleffner, F.R. (1957). Syndrome of acquired aphasia with convulsive disorder in children. Neurology, 7, 523–30. Marston, D., Besag, F., Binnie, C.D. and Fowler, M. (1993). Effects of transitory cognitive impairment on psychosocial functioning of children with epilepsy: a therapeutic trial. Dev Med Child Neurol, 35, 574–81. Morrell, F., Whisler, W.W. and Bleck, T.P. (1989). Multiple subpial transection: a new approach to the surgical treatment of focal epilepsy [see comments]. J Neurosurg, 70, 231–9. Pazzaglia, P. and Frank-Pazzaglia, L. (1976). Record in grade school of pupils with epilepsy: an epidemiological study. Epilepsia, 17, 361–6. Ross, E.M., Peckham, C.S., West, P.B. and Butler, N.R. (1980). Epilepsy in childhood: findings from the National Child Development Study. Br Med J, 280, 207–10. Sillanpää, M. (1992). Epilepsy in children: prevalence, disability, and handicap. Epilepsia, 33, 444–9. Stores, G. (1978). School-children with epilepsy at risk for learning and behaviour problems. Dev Med Child Neurol, 20, 502–8. Stores, G. and Hart, J.A. (1975). Proceedings: reading skills of children with generalized and focal epilepsy attending ordinary school. Electroencephalogr Clin Neurophysiol, 39, 429–30. Trimble, M.R. (1988). Cognitive hazards of seizure disorders. Epilepsia, 29 (Suppl. 1), S19–24.

7

Aggression and epilepsy L. Tebartz van Elst Institute of Neurology, London, UK and Albert-Ludwigs-University, Freiburg, Germany

Introduction Human aggression is an important social and clinical problem (Fenwick, 1986; Saver et al., 1996; Swartz et al., 1998; Trimble, 1996). One difficulty of studying aggression is its phenomenological and probably neurobiological heterogeneity, leading to difficulties in assessment and classification. An important distinction has emerged between the terms violence and aggression. Following Treiman, violence has been defined as forceful infliction of abuse or damage onto another individual or object, but which is not necessarily the result of intentional aggression. In contrast, aggression is considered an offensive action directed toward another individual or object with the intent to harm, threaten or control (Treiman, 1991). But even this distinction leaves researchers with the problem of assessing intentionality, which in clinical practice often is impossible. Therefore, it is important to define phenomenological criteria of the specific behavioural syndrome of interest before embarking on research or discussion into the neurobiology and psychology of aggression. Classification of aggression In the animal kingdom aggressive behaviour has been classified according to the context in which it can be observed. Table 7.1 summarizes different subtypes of aggressive behaviour in animals following Moyer (1987). In contrast to animal behaviour, human behaviour is less preformed and it depends less on external cues. Thus a simple transfer of this classification to human aggressive behaviour is not valid (Kalin, 1999). Nevertheless, there is general agreement that at least two different phenomenological and neurobiological subtypes of aggressive behaviour can be differentiated in humans: predatory and defensive aggression (Goldstein, 1974; Kalin, 1999; Moyer, 1987; Mungas, 1983; Vitiello and Stoff, 1997). Predatory aggression is characterized phenomenologically as a well-structured and goal-directed behaviour performed in an emotionally calm and concentrated state of mind. In contrast, defensive aggression is typically being seen in the context 81

82

L. Tebartz van Elst

Table 7.1. Subtypes of animal aggression

Aggression type

Example of animal behaviour

Behavioural characteristic

Predatory aggression

Predator kills prey

Calm, concentrated, goal directed, well structured

Intermale (territorial) aggression

Fight for supremacy of two lions

Arousal, concentrated, goal directed, well structured

Maternal aggression

Swan protects offspring from predator

Arousal, concentrated, goal directed, well structured

Sex-related aggression

Mantis kills male after copulation

Calm, behavioural stereotypes

Fear-induced aggression

Flight–Fight–Reaction Gnu fights predator

Arousal, vocalization, less well structured, diverse behavioural pattern, reactive behaviour

Source: Moyer (1987).

of high emotional arousal associated with vocalizations and signs of fear or anger. The behaviour itself is less structured and defensive (Valzelli, 1981). Apart from the planned and goal-directed aggression often conducted by persons with antisocial personality disorder (predatory aggression), most forms of human aggression are thought to be a reaction toward a perceived threat, be it adequate or not (Albert et al., 1993). Obviously, the perception as to whether a stimulus is threatening or not is decisive in the information processing leading to the aggressive behaviour. Clinical relevance of aggression Aggressive behaviour can be observed in the context of different medical, neurological and psychiatric disorders and diseases. It is a common problem in patients with mental retardation, possibly due to impaired social perception or deficits in expressing personal needs (Barratt et al., 1997; Gunn, 1977; Kligman and Goldberg, 1975; Saver et al., 1996). Aggressive behaviour in the context of organic brain disease like frontal or hypothalamic brain tumours, neuro-degenerative disease, delirium or drug abuse is often malstructured, defensive and tends to occur in the context of states of confusion and diffuse emotional arousal. Goal-directed and well-planned acts of aggression can occur on the background of psychiatric disorders like psychosis with delusional states, attention-deficit hyperactivity disorder (ADHA) or bipolar disorder. It is frequently observed in patients with antisocial personality disorder (APD) where it is part of the characteristic trait-like behaviour (Barratt et al., 1997; Miller et al., 1997; Stein et al., 1993).

83

Aggression and epilepsy

Table 7.2. Functional relevance of different brain structures for aggressive behaviour

Brain structure

Assumed function

Frontal lobe

Inhibitory function Suppression of aggressive behavioural drive

Amygdala and limbic circuits

Emotional evaluation of multimodal sensory and cognitive input; emotional drive and arousal

Hypothalamus

Control of brain stem behavioural programs; regulation of internal environment; coordination of behavioural programs and internal environment in flight–fight situations

Brain stem structures i.e. periaqueductal grey

Evolutionary performed behavioural flight–fight programs

The only clinical syndrome of aggression that has been included in an international classificatory system is intermittent explosive disorder (IED) according to the guidelines of DSM–IV (American Psychiatric Association, 1994). IED is characterized by several discrete episodes of failure to resist aggressive impulses that result in serious assaultive acts or destruction of property. The behaviour is out of proportion to any precipitating psychosocial stressor and is not due to substance abuse, another mental disorder like personality disorder, any other first axis psychiatric disorder or a general medical condition like head trauma or neuro-degenerative diseases. The phenomenological criteria are basically those of episodic dyscontrol, a psychopathological entity that has been described by many authors in the past (Bach-Y-Rita et al., 1971). Neurobiology of aggression Various brain structures have been implicated in the mediation of aggressive behaviour in animals and humans, the most important being the periaqueductal grey (Behbehani, 1995; Brandao et al., 1994), the hypothalamus (Andy and Jurko, 1972), the amygdala and associated limbic structures (Aggleton, 1993; Dicks et al., 1969; Halgren, 1992; Kling and Brothers, 1992; Rolls, 1992), and the frontal lobes (Damasio et al., 1990; Miller et al., 1997; Raine et al., 1998). While an understanding of the complex interplay of these different brain circuits in regulating different aggressive behaviours is still poor, the first elements of a functional anatomy of aggression can be outlined (see Table 7.2). In animals, brain stem structures like the periaqueductal grey are crucial for the activation of evolutionary preformed behavioural programs like attacking or defensive behaviour

84

L. Tebartz van Elst

(Behbehani, 1995; Brandao et al., 1994). These structures are controlled by higher neuronal centres in the hypothalamus (Bhatnagar and Dallman, 1998; Van de Poll and Van Goozen, 1992) which in addition to controlling these behavioural brainstem programs adjust the internal endocrinological and immunological environment to aggressive behaviour in flight–fight situations (Luo, 1998; Reis, 1969; Shaikh, 1997; Zanchetti, 1968). Frontal lobe functions are known to be crucial for the ability to suppress behavioural impulses. Consequently patients with frontal lobe lesions lose their ability to suppress aggressive impulses and thus might present with severe aggressive and violent psychopathology (Damasio et al., 1990; Krakowski and Czobor, 1997; Petty et al., 1996; Stein et al., 1993). Within the network of critical brain structures for aggressive behaviour, the amygdala are thought to play a crucial role for the mediation of fear-induced aggression, a subtype of defensive aggression (Aggleton, 1993; Charney and Deutch, 1996; Gallagher and Chiba, 1996; LeDoux, 1995). They receive extensive input from various levels of sensory information processing and project to most of the other critical brain structures i.e. the brain stem, hypothalamus, thalamus and frontal lobe (Alheid et al., 1995; Amaral et al., 1992). From a neurophysiological point of view they are in a predestined position for the affective evaluation of multimodal sensory input. Thus pathology within the circuits affecting the amygdala might lead to mental states where the misinterpretation of sensory input as threatening leads to aggressive outbursts. In agreement with this assumption, electrical stimulation of the amygdala can lead to experiences like fear, anxiety or anger (Chapman et al., 1954; Gloor et al., 1982) and lesioning of the amygdala severely impairs fear conditioning in animals (Davis et al., 1994) and humans (LaBar et al., 1995). Furthermore, in an open retrospective study of 481 cases of bilateral amygdalotomies performed for the control of conservatively untreatable aggressiveness moderate to excellent improvement of aggressive behaviour was reported in 70–76% (Ramamurthi, 1988). Aggression and epilepsy The relationship between epilepsy and aggressive behaviour is a particularly controversial issue (Geschwind, 1975). While in patients with episodic affective aggression, a history of epilepsy is reported to be more common (Bach-Y-Rita et al., 1971; Elliot, 1982) most of the community-based studies did not find an increased prevalence of aggressive behaviour in patients with epilepsy (Kligman and Goldberg, 1975; Lishman, 1998). The prevalence of aggression in epilepsy in general, not regarding the specific epileptic subsyndrome, varies between 4.8% (Rodin, 1973) and 50% (Gastaut et al.,

85

Aggression and epilepsy

1955). In a large survey of 666 patients with TLE, Currie et al. (1971) reported aggression in 7% of the patients. Falconer reviewed 100 patients from London’s Maudsley Hospital referred for temporal lobectomy and found a prevalence of outbursts of aggressive behaviour in 27% of their patients (Falconer, 1973). However, these studies were hampered by selection bias and the real prevalence of aggressive behaviour in epilepsy remains controversial (Lishman, 1998). In epilepsy three different types of aggressive behaviours should be distinguished on the basis of their relationship toward the seizures: ictal, postictal and interictal aggression. Ictal aggression occurs with extreme rarity (Gunn, 1971; Saver et al., 1996). In a large survey of several thousand seizures documented on videotelemetry an incidence of about 1/1000 seizures was found (Delgado-Escueta et al., 1981). In ictal aggression hostile and verbal or physical aggressive behaviour is often directed towards nearby objects or persons and may be provoked or not. The patients are usually amnestic for these aggressive episodes and express remorse for their behaviour (Devinsky and Bear, 1984; Fenwick, 1989). Postictal aggression is more common than ictal aggression but it is still believed to be rare (Treiman, 1991). It often occurs following a cluster of complex partial seizures or very severe secondary generalized seizures. Ictal pain or dysphoria may predispose individuals to the development of postictal aggressive behaviour (Gerard, 1998). Postictal aggression is frequently observed in the context of postictal confusional states or postictal psychosis but it also occurs without any signs of delusion or hallucination (Kanemoto, 1999; Lancman, 1999; Szabo and Lancman, 1996). If postictal aggression is part of a postictal confusional state the disruptive behaviour immediately follows the seizure without a lucid interval. The violent behaviour tends to be resistive, poorly structured and patients usually are very aroused, angry and fearful (Kanemoto, 1999; Lancman, 1999). Postictal psychosis follows a cluster of complex partial and secondary generalized seizures after a lucid interval, which might vary in duration between hours up to days (Fenwick, 1986; Kanemoto, 1996; Logsdail and Toone, 1988; Trimble, 1991). Aggressive behaviour in the context of a delusional state may be wellstructured and goal-directed and even though patients often feel angry and aroused, they may appear calm and concentrated to the observer (Kanemoto, 1999; Szabo and Lancman, 1996). Kanemoto et al. (1999) recently pointed out that welldirected and self-destructive behaviour was not a feature of epileptic psychosis in general but the hallmark of postictal psychosis in particular. Finally, interictal aggression can be seen in the context of an antisocial personality disorder which, in turn, might be the consequence of the sometimes difficult psychosocial background of patients with epilepsy. It might also be part of a psychotic episode (Logsdail and Toone, 1988) but as Kanemoto (1999) pointed out this is rare compared with postictal psychosis.

86

L. Tebartz van Elst

Interictal aggression is frequently seen in patients with epilepsy and mental handicap, but in these patients aggressive behaviour is often a result of poor social and communication competence in expressing personal needs, and very rarely results in severe violence (Gunn, 1977). Apart from that, an interictal syndrome of episodic affective aggression, independent of observable ictal activity, major psychiatric disorder or antisocial personality disorder, is well described and has been referred to as episodic dyscontrol (Bach-Y-Rita et al., 1971; Elliott, 1984; Leicester, 1982; Maletzky, 1973; Ratner and Shapiro, 1979; Stone et al., 1986). Episodic dyscontrol is characterized by several discrete episodes of extreme unprovoked arousal and rage resulting in severe aggressive and violent behaviour. The behaviour is out of proportion to any precipitating psychosocial stressor and patients often feel remorse for their deeds. The phenomenological criteria are those of DSM–IV intermittent explosive disorder (IED) described above (Elliott, 1984). Because of the emotional arousal typically seen in episodic dyscontrol, this behavioural syndrome has to be classified as a subtype of affective, defensive aggression. The high level of arousal with signs of anxiety or fear is also the reason why amygdala pathology is thought to contribute to this behavioural syndrome (Elliot, 1992; Fenwick, 1986; Trimble et al., 1996). Intermittent explosive disorder in epilepsy In the past, few studies have addressed the relationship between different psychobiological factors like brain pathology, IQ and demographic background, and aggression in epilepsy. While Rodin found more evidence of organic brain disease (Rodin, 1973), and Falconer (1973) reported an increased incidence of mesial temporal lobe sclerosis in aggressive patients with temporal lobe epilepsy (TLE), Herzberg and Fenwick did not find any relationship between specific electroencephalography (EEG) or computerized tomography (CT) pathology and aggression in patients with TLE (Herzberg and Fenwick, 1988). All three studies found a relationship between low IQ and aggression and two reported an association between male sex and aggression (Falconer, 1973; Rodin, 1973). Since none of these studies used modern magentic resonance imaging (MRI) techniques to assess the relationship between brain pathology and aggressive behaviour in patients with epilepsy, in two recent projects we studied this relationship using quantitative MRI (Tebartz van Elst et al., 2000a,b; Woermann et al., 2000). The aims of our studies were first, to investigate amygdala pathology, and second, to assess cortical abnormalities in patients suffering from TLE and additional affective aggression, specifically IED. The most common pathology underlying TLE in general is hippocampal sclerosis (HS) often in the context of mesial temporal sclerosis (MTS) (Gloor, 1991;

87

Aggression and epilepsy

Margerison and Corsellis, 1966; Wieser, 1983). A radiological in vivo diagnosis of mesial temporal lobe sclerosis is possible by demonstrating atrophy of the mesial temporal lobe structures on T1-weighted anatomical MRI images and increased signal on conventional spin echo T2 MRI sequences (Duncan et al., 1996; Jackson et al., 1990; Woermann et al., 1998). Since HS seems to be diffuse rather than focal in most of the cases (Kim et al., 1995) an involvement of the amygdala by the pathological process underlying hippocampal sclerosis might be expected, and indeed is reported in the literature (Hudson et al., 1993; Miller et al., 1994). In vivo identification of amygdala sclerosis by measuring the amygdala T2 relaxation time has been reported in patients with TLE (Kalviainen et al., 1997; Van Paesschen et al., 1996). Furthermore amygdala volumetry has been validated as a reliable method (Cendes et al., 1993; Kalviainen et al., 1997; Soininen et al., 1994; Watson et al., 1992). In our study we hypothesized that, in patients with TLE and intermittent affective aggression, amygdala sclerosis in the context of hippocampal sclerosis would be more common as compared to control patients. Furthermore, we wanted to test if there is an association between aggression on the one hand and hippocampal sclerosis, low IQ, and poor social adjustment on the other hand in patients with TLE. In a second step we analysed cortical grey matter abnormalities in these patients in order to investigate frontal lobe pathology, since from a phenomenological point of view episodic dyscontrol or IED can be characterized as a hyperarousal–dyscontrol syndrome. Amygdala pathology in patients with TLE and IED

Twenty-five patients with TLE and IED diagnosed according to the DSM–IV criteria described above and 25 control patients with TLE without any psychopathology were recruited from a tertiary referral centre (National Hospital for Neurology and Neurosurgery and the associated Chalfont Centre for Epilepsy). The clinical syndrome of interest was defined as complex partial seizures with a semiology, EEG and MRI findings compatible with TLE. On the basis of the discharge summaries patients with TLE with and without a history of aggression were identified, contacted and seen by a neuropsychiatrist (LTVE). Patients with extratemporal or generalized epilepsy were excluded as were those with a history of mental handicap or psychoses. Patients with TLE with and without a history of IED diagnosed according to DSM–IV criteria were included in the study. A neurological and psychiatric history and examination were obtained, as well as routine EEG investigations and neuropsychological investigations. Full (FIQ), verbal (VIQ), and performance IQ (PIQ) were measured and patients with a FIQ below 70 were excluded from the study to avoid selection bias. All patients were asked to fill in

88

L. Tebartz van Elst

the Beck Depression Inventory (BDI–13) and the State Trait Anxiety Inventory (STAI). Both questionnaires are self-rating instruments for depression and anxiety respectively (Thomson, 1989a, b). In order to assess aggression, carers were asked to fill in the Social Dysfunction and Aggression Scale (SDAS–21), a well validated and recommended questionnaire (European Rating Aggression Group, 1992; Mak and de Konning, 1995). Twenty healthy volunteers, matched for age and sex were scanned and measured twice in order to assess the reliability of the volumetric measurements. The MRI images were obtained at the Chalfont Centre for Epilepsy on a 1.5 T GE Signa scanner (GE Medical Systems, Milwaukee, USA) using a T1–weighted inversion-recovery prepared volume acquisition (IRSPGR: TI/TR/TE/flip⫽ 450/15/4.2/20; 12⫻1.5 mm thick contiguous coronal slices; matrix 256⫻192, 24 cm⫻18 cm FOV), and a conventional spin echo sequence (TR/TE1/TE2/NEX 2000/30/120/1, 256⫻192 matrix, 24⫻18 cm FOV, 5 mm slices) for computation of T2 values. Volumetric measurements were performed using a locally developed interactive software program MRreg (Lemieux et al., 1998; Moran et al., 1999) following the established protocol described by Watson et al. (1992). The rater (LTVE) was blind to the subject grouping. The amygdala volumes were corrected for total brain size by division by the intracranial volume. Intrarater reliability figures were carefully assessed and proved to be satisfactory. Amygdala atrophy was defined as a volume smaller than 3 standard deviations (SD) below the average amygdala volume of the control group. Amygdala T2 values were measured using DispImage image analysis software (Plummer, 1992) by placing the largest possible elliptic region of interest within the amygdala while avoiding anatomical boundaries. Amygdala T2 signals were defined as pathological if they were greater than 2 SD above the mean of the control population. Details of the methodology have been published elsewhere (Tebartz van Elst et al., 1999, 2000b). The demographic data of both groups are summarized in Table 7.3. The two patient groups were matched for age, sex, demographic background, duration of epilepsy and seizure severity. There was no significant group difference regarding the history of birth complications, febrile convulsions or status epilepticus. However, the incidence of encephalitic brain disease (Fisher’s Exact Test: P⫽0.05) and left-handedness (Chi-square Test: P⬍0.05) was significantly increased in aggressive patients. There was less right-sided focal EEG abnormality and more bilateral EEG abnormality in the aggressive group compared to nonaggressive patients with TLE. Left- as well as right-sided hippocampal sclerosis was significantly less common in patients with TLE and IED. Other left temporal pathology, including three patients with amygdala pathology (amygdala sclerosis, amygdala glioma, amygdala DNT), two patients with multiple small temporal infarctions and two patients with diffuse left temporal atrophy of unknown origin, was significantly more common in patients with TLE plus IED (see Table 7.4).

89

Aggression and epilepsy

Table 7.3. Demographic and historical data of patient groups with temporal lobe epilepsy (TLE), with TLE and intermitttent explosive disorder (IED) or without IED (TLE alone) (n ⫽ 25 in each group)

Age (in years) [range] Sex: f/m Work: number unemployed Living: number living independently Income: number on social support Social: number living in stable relationship Therapy: monotherapy – polytherapy Mean duration of TLE (in years) [range] Mean seizure frequency (estimated frequency per month) [range] Birth complications Febrile convulsions Status epilepticus Left-handed History of encephalitis

TLE and IED

TLE alone

30.1 [18–49] 8/17 21 14 18 8 3–22 22.4 [5–45] 13.4 [0.5–60]

33.8 [19–56] 10/15 15 10 16 8 3–22 24.5 [7–46] 21 [1.5–190]

9 5 0 7 5

7 9 2 1 0

Source: Tebartz van Elst et al. (2000a).

Table 7.4. Radiological MRI pathology on visual assessment

Overall significance: P⫽0.002

No pathology

Right hemisphere

Left hemisphere

Bilateral hemispheres

TLE and IED TLE alone Significance

10 6

1 8 **

4 10 **

3 1

Other left temporal pathology 7 0 **

Notes: TLE, temporal lobe epilepsy; IED, intermittent explosive disorder. Other left temporal pathology: amygdala sclerosis⫽1, amygdala glioma⫽1, amygdala DNT⫽ 1; multiple small temporal infarctions⫽2; generalized left temporal atrophy⫽2. Closed test procedure: *⫽P⬍0.5, **⫽P⬍0.01). Source: Tebartz van Elst et al. (2000b).

90

L. Tebartz van Elst

There was no evidence of an increased amygdala T2 relaxation time in the aggressive group. Using the definition of amygdala T2 pathology as a T2 time greater than 2 SD above the mean of the control group, there was no significant group difference in amygdala pathology. A group comparison did not reveal a significant overall difference in amygdala volumes (right side: aggressive⫽1893 mm3, SD⫽435 mm3; nonaggressive⫽1909 mm3, SD⫽231 mm3; left side: aggressive⫽1840 mm3, SD⫽398 mm3; nonaggressive⫽1868 mm3, SD⫽290 mm3). However, in the aggressive patients, a subgroup of five patients (20%) showed amygdala atrophy as compared to only one in the nonaggressive group (Chi-square: P⫽0.04). Two patients had left-sided amygdala atrophy, two had bilateral atrophy and the only patient who exhibited right-sided amygdala atrophy was left-handed. An increased incidence of encephalitis (Chisquare Test: P⬍0.05; Fisher’s Exact Test: P⫽0.1) was the only clinical feature that distinguished patients with amygdala atrophy from those with normal amygdala volumes. Since there was no overlap between the five patients with severe amygdala atrophy and those seven patients with other amygdala pathology, in a total number of 12 out of 25 aggressive patients there was some evidence of amygdala pathology as compared to only one in the nonaggressive group (see Figure 7.1). There was a highly significant group difference in IQ figures with the verbal IQ (VIQ), the performance IQ (PIQ) and hence the full IQ (FIQ) all being lower in the aggressive group (Table 7.5). The verbal IQ differed more than the performance IQ. As expected there was a highly significant group difference in the Social Dysfunction and Aggression Scores (SDAS 9, SDAS 21) since this was the criterion for group definition (Table 7.5). There was a significant group difference in BDI and STAI scores with the aggressive group rating much higher in depression (P⬍0.05) state (P⬍0.05) and trait anxiety (P⬍0.01). Cortical abnormalities in patients with TLE and IED

In order to detect possible subtle cortical brain pathology, that was not present on visual assessment of the MRI scans, in a second approach we analysed the same groups of patients using a voxel-by-voxel-comparison of cortical grey matter. After automated segmentation of cerebral grey matter from T1-weighted MRI, the objective technique of statistical parametric mapping (SPM) was applied to study the patient groups described above and comparison made with 35 healthy control subjects (Woermann, 1999). Both TLE patient groups were compared with each other and with the control subjects on a voxel-by-voxel basis for increases and decreases of grey matter. SPM 96 characterizes significant regional differences in image parameters, while allowing for global differences to be taken into account. Details of the methodology are published elsewhere (Woermann, 1999; Woermann et al., 1999). The resulting significant differences through the 3D image space were dis-

91

Aggression and epilepsy

Figure 7.1.

Amygdala pathology in patients with TLE with and without IED. AGG ⫽ aggressive.

played collapsed into three orthogonal planes (‘glass brain’, see Figure 7.2). Regions of significant difference were overlaid on normalized T1-weighted images to facilitate correlation with anatomy (see Figure 7.3). In patients with TLE with IED compared as a group with healthy control subjects, reductions of grey matter were found over large areas of the left extratemporal neocortex, maximal in the left frontal neocortex; one maximum difference projection had a Z score of 5.67 at Talairach coordinates x⫽58, y⫽36, z ⫽9 mm (left anterior frontolateral cortex), the other a Z score of 4.78 in a more posterior left frontal lobe location (Talairach coordinates x⫽66, y⫽0, z⫽28 mm, Figures 7.2 and 7.3). Patients with TLE who did not have IED showed no significant decrease of cortical grey matter compared with control subjects. Patients with TLE with IED also had reduction of left frontal grey matter, compared with patients with TLE without IED, although less marked than when compared with control subjects (Z score⫽3.49 at Talairach coordinates x⫽66, y⫽2, z⫽26 mm). The SPM-based

92

L. Tebartz van Elst

Table 7.5. Neuropsychological and psychometric parameters

Variable

TLE and IED

TLE alone

Group comparison

Mean FIQ (SD) Mean VIQ (SD) Mean PIQ (SD) Mean BDI (SD) Mean S-STAI Mean T-STAI Mean SDAS-9 (SD) Mean SDAS-21 (SD)

80.6 (8.5) 81.0 (8.6) 83.3 (11.8) 8.8 (4.8) 44.1 (14.9) 50.1 (10.1) 14.9 (7.5) 30.9 (12)

93.0 (13.9) 93.8 (14.1) 94.7 (15.1) 4.2 (5.8) 32.1 (10.9) 34.7 (11.6) 0.4 (0.7) 3.1 (4.5)

** (P⬍0.000) ** (P⬍0.000) ** (P⬍0.005) ** (P⬍0.007) ** (P⬍0.005) ** (P⬍0.000) ** (P⬍0.000) ** (P⬍0.000)

Notes: TLE, temporal lobe epilepsy; IED, intermittent explosive disorder; FIQ, full IQ; VIQ, verbal IQ; PIQ, performance IQ; BDI, Beck Depression Inventory; STAI, State Trait Anxiety Inventory; SDAS, Social Dysfunction and Aggression Scale; SD, standard deviation. *⫽P⬍0.5, **⫽P⬍0.01 after Bonferroni correction. Source: Tebartz van Elst et al. (2000a).

voxel-wise correlation of SDAS scores and automatically segmented grey matter in all patients with TLE showed the left frontal grey matter area as negatively correlated with these scores which expressed social consequences of interictal affective aggression (Z score 3.65 at Talairach coordinates x⫽66, y⫽2, z⫽26 mm). Age, scores of depression and anxiety, IQ measures, or scores of verbal fluency did not significantly correlate with specific decreases in grey matter in all patients with TLE. Dual brain pathology in patients with affective aggression and epilepsy In our studies we have demonstrated that amygdala-related brain pathology could be recognized in about half of the patients with TLE and IED. Contrary to our hypothesis however, there was no evidence for increased mesial temporal lobe sclerosis or amygdala sclerosis in the aggressive patient group. Brain pathology in patients with epilepsy and aggression was more diverse in nature and more diffuse in distribution. Interestingly, we found an increased prevalence of a history of encephalitis in patients with epilepsy plus aggression. Encephalitis in the past might have been a pathogenic element for the more diffuse and widespread pathology in the temporal lobes seen in our aggressive patients. Furthermore the increased prevalence of left-handedness in our aggressive patients may indicate early brain pathology like encephalitis affecting the left hemisphere (i.e. lateralization of dominance to the right hemisphere).

93

Aggression and epilepsy

100

80

60

40

20 TLE with IED

Figure 7.2.

controls

‘Glass brain’ view of decreased grey matter in 24 patients with TLE with episodes of affective aggression compared with 35 control subjects; displayed after correction for multiple comparisons (Woermann et al., 2000).

Eleven of twelve patients with amygdala-related brain pathology displayed this pathology on the left-hand side and the only patient with right-sided amygdala atrophy alone was left-handed. Thus the dominant hemisphere seems to play a more important role in the mediation of affective aggression than the nondominant hemisphere. The finding of frontal cortical grey matter loss was clearly lateralized too. Patients with TLE plus IED displayed highly significant left frontal grey matter loss that correlated with the SDAS scores. This, together with the left lateralized finding of amygdala-related brain pathology, could support a theory of dual brain pathology in patients with IED (see Figure 7.4): pathology within the amygdala or amygdaloid circuits might result in hyperarousal states where patients become angry and

94

L. Tebartz van Elst

Figure 7.3.

Area of decreased grey matter (overlaid on normalized T1-weighted MRI) comparing 24 patients with TLE with episodes of affective aggression with 24 patients with TLE without such episodes. At the location of maximum difference (Talairach coordinates: x⫽⫺66, y⫽⫺2, z⫽26 mm) the Z-score was 3.49 (Woermann et al., 2000).

aroused without a sufficient external stimulus (hyperarousal syndrome). This dysfunctional arousal resulting in aggressive behavioural impulses normally can be suppressed by learned behavioural rules. However, associated with additional frontal lobe pathology, the capacity of the affected patients to suppress behavioural impulses arising from the ‘emotional brain’ is limited and this leaves the patients vulnerable to the development of hyperarousal–dyscontrol syndromes i.e. episodic dyscontrol or intermittent explosive disorder. This theory is in line with earlier functional imaging and MR spectroscopy studies showing a reduced prefrontal glucose metabolism in murderers and significantly lower neuronal markers in the frontal lobes of repetitively violent patients with learning disabilities, although without clear lateralizing effects (Raine et al., 1997, 1998). From the fact that in our patient sample frontal as well as limbic brain pathology was clearly lateralized to the left i.e. the dominant hemisphere, one might conclude that only dual brain pathology of the dominant hemisphere results in hyperarousal–dyscontrol syndromes while a crossover dual brain pathology with

95

Aggression and epilepsy

= hyperarousal

(Experience) Evaluation

= dyscontrol

Cognitive Primary cortex

Unimodal cortex

Polymodal cortex

Frontal cortex

Hypothalamus

Thalamus

Hipp.

Sensory input

Amygdala

Brainstem

Emotional

Input (Perception)

Figure 7.4.

Output (Behaviour)

Dual brain pathology in episodic dyscontrol – a theory of intermittent explosive disorder as hyperarousal–dyscontrol syndrome. Hipp, hippocampus.

pathology of the right frontal lobe or the right amygdala does not produce phenomenological states like this. However, this assumption is very speculative and further investigations, possibly comparing patient groups with left- and right-sided dual brain pathology may help clarify this question. There is a controversy regarding the importance of hemispheric specialization for aggressive behaviour (Bear, 1983; Nachson, 1991) with the majority of studies pointing to a more important role of the left hemisphere (Saver et al., 1996). Our MRI and neuropsychological findings (i.e. particularly low verbal IQ) support this assumption, which has been reported earlier (Herzberg and Fenwick, 1988). It is important to note that in our sample there was a strong link between aggression and high levels of depression and anxiety, confirming other reports of such an association in the nonpsychiatrically ill population (Bjork et al., 1997). It seems plausible that high levels of anxiety result in states of hyperarousal that might be facilitated by amygdala pathology as suggested by other authors (Cendes et al., 1994). Regarding the relationship between depression and aggression there are only few and nonconclusive reports (Braconnier and Jeanneau, 1997). Our findings point to a clear association between depression and IED in TLE. Social and psychological aspects of aggression in epilepsy Even though our studies concentrated on research into the neurobiological basis of aggressive behaviour in patients with epilepsy, there is no question about the critical

96

L. Tebartz van Elst

role of social and psychological intervening variables for the development of aggressive behaviour in general, be it in the context of epilepsy or not. Social disadvantage, prejudice, poor housing, poverty, poor communication skills, all are factors that make hyperarousal states and states of discontentment and anger more likely and thus might increase the probability of aggressive behaviour. On the other hand a precise and correct diagnosis always is the first step in managing difficult behaviour, and sociological, psychological and neurobiological views of the same problem are not necessarily contradictory. For example, we were able to demonstrate a close link between episodic dyscontrol, reduced IQ, depression and anxiety. Even though disentangling the complex interaction between these different psychobiological elements is very difficult, it nevertheless may suggest a possible way of treatment, irrespective of which of these elements is the most important one from an aetiological point of view. Treatment of aggression in epilepsy If aggression is a problem in the clinical management of patients with epilepsy the most important point is to establish a correct diagnosis (Figure 7.5). A careful neurological, psychiatric and medical history and examination should be performed to answer the following questions: 1. Is there any medical condition that contributes to the aggressive behaviour such as endocrinological or immunological diseases? Is there any medication that might contribute to the aggressive behaviour? 2. What is the correct neurological diagnosis? Are there any other cerebral problems in addition to the epilepsy? 3. Are there any psychiatric diagnoses which possibly are independent of the epilepsy, like bipolar disease or antisocial personality disorder? If the epilepsy started early in life it is in fact often impossible to establish if, for example, a clinical picture that fulfils the criteria for an antisocial personality disorder has to be judged as independent of the organic brain disease indicated by the epilepsy or alternatively is an organic personality disorder. 4. Finally, a careful behavioural analysis and possibly video-telemetry should clarify if the aggressive behaviour is ictal, postictal or interictal and whether it occurs in the context of altered states of consciousness or psychosis. Following syndromatic and possibly nosological diagnosis, treatment should be causal if possible i.e. intervening medical problems like endocrinological disorders should be treated adequately. Neurological syndromes like the epilepsy itself should be treated effectively but, as with all drugs that influence cerebral functioning, the question as to whether anticonvulsants might contribute to the behavioural problem should be considered. For example benzodiazepines are well known to have paradoxical effects in a minority of patients and may cause states of arousal and aggression (Binder, 1987; Daderman, 1999; Marcus, 1995; Sheth, 1994). Likewise phenobarbital is well known to cause behavioural problems with aggres-

97

Aggression and epilepsy

Medical history and exam

Neurological history and exam

Psychiatric history and exam

Behavioural analysis and diagnosis of aggressive syndrome (ictal, postictal, interictal)

Neuropsychiatric diagnosis

Treatment of medical, neurological and psychiatric condition (look for depression and anxiety)

Figure 7.5.

Possibly avoid tricyclic antidepressants

Possibly avoid convulsant antipsychotics

Consider paradox side effects

Consider psychotherapy, social support

Prophylactic interventions:

Psychotherapy, social support, lithium, SSRIs, antipsychotics

Management of acute aggression:

Benzodiazepines, antipsychotics

Causal treatment

Symptomatic treatment

Therapeutic guidelines for the treatment of aggression in patients with epilepsy.

sion in a considerable subgroup of patients with epilepsy often with learning disability (File, 1990). Besides, in individuals any given drug might have different effects than those described in large groups and thus a careful behavioural analysis of the sequence of events is the only way to establish any possible side effect, for example of antiepileptic drugs. Care should be taken to establish signs of depression or anxiety, since a close link between these psychopathological states and affective aggression in epilepsy has been established. Both should be treated medically and with psychotherapy at the same time (Goldstein, 1997; Lorenzen, 1973). Behavioural therapy in particular in patients with epilepsy and learning disability has been proven to be very effective (Davis, 1984; Holzapfel, 1998; Rapport, 1983). In the medical treatment of depression in patients with epilepsy, SSRIs or other new antidepressants like venlafaxine should be preferred to the old tricyclic antidepressants (TCA) since the latter are more likely to provoke seizures (Blumer, 1997; Lambert, 1999). In fact, an anticonvulsant effect of SSRIs is well documented in animal models of epilepsy (Browning, 1997; Lu, 1998; Pasini, 1996; Wada, 1995) and is also described in humans (Favale, 1995). Following treatment of all medical, neurological and psychiatric conditions that may or may not contribute to the aggressive psychopathology a symptomatic

98

L. Tebartz van Elst

treatment of the aggression is mandatory. However, this depends on whether the aggression is an ictal, postictal or interictal phenomenon. Ictal aggression does not need symptomatic treatment but one might consider interrupting a nonconvulsive status for example with benzodiazepines. Apart from that, a patient who displays agitation and aggression during a seizure should not be restricted since defensive violence is more common in such situations and the aggressive behaviour is self-limited, as is the seizure. The same is true for postictal confusional states, and even aggressive behaviour in the context of postictal psychosis is self-limited. However, if the aggression is severe and disturbing or self-harming, medical treatment with benodiazepines like diazepam or clobazan and/or antipsychotics, for example risperidone, olanzapine or quetiapine should be given. Good seizure control, avoiding severe complex partial or secondary generalized seizures, is the best prophylactic intervention since postictal confusional and psychotic states are more common after severe seizures. In interictal aggression, prophylactic and acute symptomatic treatment of aggression and agitation should be differentiated. For the treatment of acute hyperarousal–dyscontrol syndromes a combination of benzodiazepines like diazepam and antipsychotics like haloperidol or sulpiride are still the most effective and safest interventions. In cases of interictal psychosis however, the antipsychotic medication should eventually be switched to one of the atypical antipsychotic agents with little proconvulsant potential like risperidone, olanzapine or quetiapine, since these drugs are better tolerated. A good control of the psychosis is the best way to prevent aggression if it is part of the psychosis. In the case of interictal aggressive syndromes like IED there is no well-established medical prophylactic therapy; however, there are many anecdotal reports of effective use of substances like lithium, valproate, carbamazepine, antipsychotics, betablockers, clonidine and even psychostimulants (Fava, 1997; Griffith, 1985; Yudofsky, 1990). However, since there are hardly any well-conducted systematic treatment studies at the moment the medical treatment still is very experimental and single-case driven. Nevertheless, in the light of the very severe burden that is put on patients and their relatives and caregivers by the sometimes devastating behavioural episodes, a systematic trial of these agents seems to justified in special cases. The use of cognitive–behaviour therapy for anger management should also be considered, either as a first-choice treatment, or combined with psychotropic drugs. Conclusion Research into the neurobiological basis of aggression is still hampered by the difficulty in defining phenomenological and nosological homogeneous study groups.

99

Aggression and epilepsy

Nevertheless, it is important to develop a more precise understanding of the complex interplay of social, psychological and neurobiological factors all contributing to the aggressive and violent behaviour. Aggression in epilepsy is rare. However, if it occurs it imposes an immense burden on the patient, relatives and caregivers. Clinically, it is crucial to first of all establish a correct diagnosis. Intervening medical, neurological and psychiatric disorders, in particular depression and anxiety, should be recognized and treated adequately. A correct syndromatic diagnosis of the aggressive syndrome and its relation to the seizures should be made. Causal treatment should aim at any underlying medical, neurological or psychiatric disorders. In symptomatic treatment of aggressive outbursts, treatment of acute aggression and prophylactic treatment should be differentiated. In acute arousal states, if possible, patients should be left alone if aggression is a symptom of a seizure since restrictive aggression is the most common form in this situation. In cases of postictal or interictal aggression, it is important to establish a possible relationship to postictal or interictal psychosis and treat adequately. For interictal syndromes of affective aggression like IED (episodic dyscontrol), there are no generally established treatment protocols. However, tentative treatment with different substances as mentioned above seems justified in single cases. Furthermore, psychotherapeutic and social interventions may help prevent arousal states and thus can be very effective.

R E F E R E N C ES Aggleton, J.P. (1993). The contribution of the amygdala to normal and abnormal emotional states. Trends Neurosci, 16, 328–33. Albert, D.J., Walsh, M.L. and Jonik, R.H. (1993). Aggression in humans: what is its biological foundation? Neurosci Biobehav Rev 17, 405–25. Alheid, G.F., De Olmos, J.S. and Beltramino, C.A. (1995). Amygdala and extended amygdala. In The Rat Nervous System, ed. G Paxinos, 495–578. San Diego: Academic Press. Amaral, D.G., Price, J.L., Pitkänen, A. and Carmichael, S.T. (1992). Anatomical organisation of the primate amygdaloid complex. In The Amygdala: Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp. 1–66. New York: Wiley-Liss. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (Fourth Revision) (DSM–IV). Washington, DC: APA. Andy, O.J. and Jurko, M.F. (1972). Hyperresponsive syndrome. In Psychosurgery, ed. E. Hitchcock, L. Laitinen and K. Vaernet, pp. 117–26. Springfield, ILL: Charles C. Thomas. Bach-Y-Rita, G., Lion, J.R., Climent, C.E. and Ervin, F.R. (1971). Episodic dyscontrol: a study of 130 violent patients. Am J Psychiatry, 127, 1473–8. Barratt, E.S., Stanford, M.S., Kent, T.A. and Felthous, A. (1997). Neuropsychological and cognitive psychophysiological substrates of impulsive aggression. Biol Psychiatry, 41, 1045–61.

100

L. Tebartz van Elst Bear, D. (1983). Hemispheric specialisation and the neurology of emotion. Arch Neurol, 40, 195–202. Behbehani, M.M. (1995). Functional characteristics of the midbrain periaquaeductal gray. Prog Neurobiol, 46, 573–605. Bhatnagar, S. and Dallman, M. (1998). Neuroanatomical basis for facilitation of hypothalamicpituitary-adrenal responses to a novel stressor after chronic stress. Neuroscience, 84, 1025–39. Binder, R.L. (1987). Three case reports of behavioral disinhibition with clonazepam. Gen Hosp Psychiatry, 9, 151–3. Bjork, J.M., Dougherty, D.M. and Moeller, F.G. (1997). A positive correlation between selfratings of depression and laboratory-measured aggression. Psychiatry Res, 69, 33–8. Blumer, D. (1997). Antidepressant and double antidepressant treatment for the affective disorder of epilepsy. J Clin Psychiatry, 58, 3–11. Braconnier, A. and Jeanneau, A. (1997). Anxiety, aggression, agitation and depression: psychopathologic aspects. Encephale, 23, 43–7. Brandao, M.L., Cardoso, S.H., Melo, L.L., Motta, V. and Coimbra, N.C. (1994). Neural substrate of defensive behavior in the midbrain tectum. Neurosci Biobehav Rev, 18, 339–46. Browning, R.A. (1997). Enhancement of the anticonvulsant effect of fluoxetine following blockade of 5-HT1A receptors. Eur J Pharmacol, 336, 1–6. Cendes, F., Andermann, F., Gloor, P. et al. (1993). MRI volumetric measurement of amygdala and hippocampus in temporal lobe epilepsy. Neurology, 43, 719–25. Cendes, F., Andermann, F., Gloor, P. et al. (1994). Relationship between atrophy of the amygdala and ictal fear in temporal lobe epilepsy. Brain, 117, 739–46. Chapman, W.P., Schroeder, H.R., Geyer, G. et al. (1954). Physiological evidence concerning importance of amygdala nuclear region in the integration of circulatory functions and emotion in man. Science, 120, 949–54. Charney, D.S. and Deutch, A. (1996). A functional neuroanatomy of anxiety and fear: implications for the pathophysiology and treatment of anxiety disorders. Crit Rev Neurobiol, 10, 419–46. Currie, S., Heathfield, W., Henson, R. and Scott, D. (1971). Clinical course and prognosis of temporal lobe epilepsy: a survey of 666 patients. Brain, 94, 173–90. Daderman, A.M. (1999). Flunitrazepam (Rohypnol) abuse in combination with alcohol causes premeditated, grievous violence in male juvenile offenders. J Am Acad Psychiatry Law, 27, 83–99. Damasio, A.R., Tranel, D. and Damasio, H. (1990). Individuals with sociopathic behavior caused by frontal damage fail to respond autonomically to social stimuli. Behav Brain Res, 41, 81–94. Davis, G.R. (1984). Cognitive-behavioral treatment of depressed affect among epileptics: preliminary findings. J Clin Psychology, 40, 930–5. Davis, M., Rainnie, D. and Cassell, M. (1994). Neurotransmission in the rat amygdala related to fear and anxiety. Trends Neurosci, 17, 208–14. Delgado-Escueta, A.V., Mattson, R.H., King, L. et al. (1981). Special report. The nature of aggression during epileptic seizures. N Engl J Med, 305, 711–16. Devinsky, O. and Bear, D. (1984). Varieties of aggressive behavior in temporal lobe epilepsy. Am J Psychiatry, 141, 651–6.

101

Aggression and epilepsy Dicks, P., Myers, R.E. and Kling, A. (1969). Uncus and amygdala lesions: effects on social behavior in the free ranging rhesus monkey. Science, 165, 69–71. Duncan, J.S., Bartlett, P. and Barker, G.J. (1996). Technique for measuring hippocampal T2 relaxation time. Am J Neuroradiol, 17, 1805–10. Elliot, F.A. (1982). Neurological findings in adult minimal brain dysfunction and the dyscontrol syndrome. J Nerv Ment Dis, 170, 680–7. Elliott, F.A. (1984). The episodic dyscontrol syndrome and aggression. Neurol Clin, 2, 113–25. Elliot, F.A. (1992). Violence – the neurologic contribution: an overview. Arch Neurol, 49, 595–603. European Rating Aggression Group (1992). Social dysfunction and aggression scale (SDAS–21) in generalised aggression and in aggressive attacks: a validity and reliability study. Int J Methods Psychiatric Res, 2, 15–29. Falconer, M.A. (1973). Reversibility by temporal-lobe resection of the behavioral abnormalities of temporal-lobe epilepsy. N Engl J Med, 289, 451–5. Fava, M. (1997). Psychopharmacologic treatment of pathologic aggression. Psychiatr Clin N Am, 20, 427–51. Favale, E. (1995). Anticonvulsant effect of fluoxetine in humans [see comments]. Neurology, 45, 1926–7. Fenwick, P.B.C. (1986). Aggression and epilepsy. In Epilepsy and Psychiatry, ed. M.R. Trimble and T. Bolwig, pp. 31–60. Chichester: John Wiley & Sons. Fenwick, P. (1989). The nature and management of aggression in epilepsy. J Neuropsychiatry Clin Neurosci, 1, 418–25. File, S.E. (1990). Changes in seizure threshold and aggression during chronic treatment with three anticonvulsants and on drug withdrawal. Psychopharmacology, 100, 237–42. Gallagher, M. and Chiba, A.A. (1996). The amygdala and emotion. Curr Opin Neurobiol, 6, 221–7. Gastaut, H., Morrin, G. and Lesevre, N. (1955). Etudes du compartment des epileptiques psychomoteur dans L’interval de leurs crises. Ann Med Psychol, 113, 1–29. Gerard, M.E. (1998). Subacute postictal aggression [see comments]. Neurology, 50, 384–8. Geschwind, N. (1975). The clinical setting of aggression in temporal lobe epilepsy. In Neural Basis of Violence and Aggression, ed. W.S. Fields and W.H. Sweet, pp. 273–81. St. Louis, MI: Warren H. Green. Gloor, P. (1991). Mesial temporal sclerosis: historical background and an overview from a modern perspective. In Epilepsy Surgery, ed. H.O. Luders, pp. 689–703. New York: Raven Press. Gloor, P., Olivier, A., Quesney, L.F., Andermann, F. and Horowitz, S. (1982). The role of the limbic system in experiential phenomena of temporal lobe epilepsy. Ann Neurol, 12, 129–44. Goldstein, L.H. (1997). Effectiveness of psychological interventions for people with poorly controlled epilepsy. J Neurol Neurosurg Psychiatry, 63, 137–42. Goldstein, M. (1974). Brain research and violent behavior: a summary and evaluation of the status of biomedical research on brain and aggressive violent behavior. Arch Neurol, 30, 1–35. Griffith, J.L. (1985). Treatment of episodic behavioral disorders with rapidly absorbed benzodiazepines. J Nerv Ment Dis, 173, 312–15.

102

L. Tebartz van Elst Gunn, J. (1971). Epilepsy, automatism, and crime. Lancet, 1, 1173–6. Gunn, J. (1977). Criminal behaviour and mental disorder. Br J Psychiatry, 130, 317–29. Halgren, E. (1992). Emotional neurophysiology of the amygdala within the context of human cognition. In The Amygdala: Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp. 191–228. New York: Wiley-Liss. Herzberg, J.L. and Fenwick, P.B .(1988). The aetiology of aggression in temporal-lobe epilepsy. Br J Psychiatry, 153, 50–5. Holzapfel, S. (1998). Behavioral psychophysiological intervention in a mentally retarded epileptic patient with brain lesion. Appl Psychophysiol Biofeedback, 23, 189–202. Hudson, L.P., Munoz, D.G., Miller, L., McLachlan, R.S., Girvin, J.P. and Blume, W.T. (1993). Amygdala sclerosis in temporal lobe epilepsy. Ann Neurol, 33, 622–31. Jackson, G.D., Berkovic, S.F., Tress, B.M. et al. (1990). Hippocampal sclerosis can be reliably detected by magnetic resonnance imaging. Neurology, 40, 1869–75. Kalin, N.H. (1999). Primate models to understand human aggression. J Clin Psychiatry, 60 (Suppl. 15), 29–32. Kalviainen, R., Salmenpera, T., Partanen, K., Vainio, P., Riekkinen, P.S. and Pitkanen, A. (1997). MRI volumetry and T2 relaxometry of the amygdala in newly diagnosed and chronic temporal lobe epilepsy. Epilepsy Res, 28, 39–50. Kanemoto, K. (1996). Postictal psychosis: a comparison with acute interictal and chronic psychoses. Epilepsia, 37, 551–6. Kanemoto, K. (1999). Violence and epilepsy: a close relation between violence and postictal psychosis. Epilepsia, 40, 107–9. Kim, J.H., Tien, R.D., Felsberg, G.J. et al. (1995). Fast spin-echo MR in hippocampal sclerosis: correlation with pathology and surgery. Am J Neuroradiol, 16, 627–36. Kligman, D. and Goldberg, D.A. (1975). Temporal lobe epilepsy and aggression. J Nerv Ment Dis, 160, 324–41. Kling, A.S. and Brothers, L.A. (1992). The amygdala and social behavior. In The Amygdala: Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp. 353–77. New York: Wiley-Liss. Krakowski, M. and Czobor, P. (1997). Violence in psychiatric patients: the role of psychosis, frontal lobe impairment, and ward turmoil. Compr Psychiatry, 38, 230–6. LaBar, K.S., Ledoux, J.E., Spencer, D.D. and Phelps, E.A. (1995). Impaired fear conditioning following unilateral temporal lobectomy in humans. J Neurosci, 15, 6846–55. Lambert, M.V. (1999). Depression in epilepsy: etiology, phenomenology, and treatment. Epilepsia, 40 (Suppl. 10), S21–47. Lancman, M. (1999). Psychosis and peri-ictal confusional states. Neurology, 53, S33–8. LeDoux, J.E. (1995). Emotions: clues from the brain. Annu Rev Psychol, 46, 209–35. Leicester, J. (1982). Temper tantrums, epilepsy and episodic dyscontrol. Br J Psychiatry, 141, 262–6. Lemieux, L., Wieshmann, U.C., Moran, N.F., Fish, D.R. and Shorvon, S.D. (1998). The detection and significance of subtle changes in mixed-signal brain lesions by serial MRI scan matching and spatial normalisation. Med Image Analysis, 2, 227–42.

103

Aggression and epilepsy Lishman, W.A. (1998). Organic Psychiatry – The Psychological Consequences of Cerebral Disorder. Oxford: Blackwell Science. Logsdail, S.J. and Toone, B.K. (1988). Post-ictal psychoses. A clinical and phenomenological description. Br J Psychiatry, 152, 246–52. Lorenzen, D. (1973). Initial results of behavior therapy in Psychiatric Regional Hospital Weinsberg. [In German]. Nervenarzt, 44, 423–7. Lu, K.T. (1998). Endogenous serotonin inhibits epileptiform activity in rat hippocampal CA1 neurons via 5–hydroxytryptamine 1A receptor activation. Neuroscience, 86, 729–37. Luo, B. (1998). Cholecystokinin B receptors in the periaqueductal gray potentiate defensive rage behavior elicited from the medial hypothalamus of the cat. Brain Res, 796, 27–37. Mak, M. and De Konning, P. (1995). Clinical research in aggressive patients: pitfalls in study design and measurements of aggression. Progr Neuropsychopharmacol Biol Psychiatry, 19, 993–1017. Maletzky, B.M. (1973). The episodic dyscontrol syndrome. Dis Nerv System, 34, 178–85. Marcus, A. (1995). Benzodiazepine administration induces exogenic psychosis: a case of child abuse. Child Abuse Neglect, 19, 833–6. Margerison, J.H. and Corsellis, J. (1966). Epilepsy and the temporal lobe: a clinical, electroencephalographic and neuropathological study of the brain in epilepsy, with particular reference to the temporal lobes. Brain, 89, 499–530. Miller, B.L., Darby, A., Benson, D.F., Cummings, J.L. and Miller, M.H. (1997). Aggressive, and socially disruptive and antisocial behaviour associated with fronto-temporal dementia. Br J Psychiatry, 170, 150–5. Miller, L.A., McLachlan, R.S., Bouwer, M.S., Hudson, L.P. and Munoz, D.G. (1994). Amygdala sclerosis: preoperative indicators and outcome after temporal lobectomy. J Neurol, Neurosurg Psychiatry, 57, 1099–105. Moran, N.F., Lemieux, L., Maudgil, D.D., Kitchen, N.D., Fish, D.R. and Shorvon, S.D. (1999). Analysis of temporal lobe resections in MR images. Epilepsia, 40, 1077–84. Moyer, K.E. (1987). Violence and Aggression – A Physiological Perspective. New York: Paragon House Publishers. Mungas, D. (1983). An empirical analysis of specific syndromes of violent behavior. J Nerv Ment Dis, 171, 354–61. Nachson, I. (1991). Neuropsychology of violent behavior: controversial issues and new developments in the study of hemisphere function. In Neuropsychology of Aggression, ed. J.S. Miller, pp. 93–116. Boston: Kluwer Academic Publishers. Pasini, A. (1996). The anticonvulsant action of fluoxetine in substantia nigra is dependent upon endogenous serotonin. Brain Res, 724, 84–8. Petty, R.G., Bonner, D., Mouratoglou, V. and Silverman, M. (1996). Acute frontal lobe syndrome and dyscontrol associated with bilateral caudate nucleus infarctions. Br J Psychiatry, 168, 237–40. Plummer, D.L. (1992). DispImage, a display and analysis tool for medical images. Revues Neuroradiol, 5, 489–95. Raine, A., Buchsbaum, M. and Lacasse, L. (1997). Brain abnormalities in murderers indicated by positron emission tomography. Biol Psychiatry, 42, 495–508.

104

L. Tebartz van Elst Raine, A., Stoddard, J., Bihrle, S. and Buchsbaum, M. (1998). Prefrontal glucose deficits in murderers lacking psychosocial deprivation. Neuropsychiatry, Neuropsychol Behav Neurol, 11, 1–7. Ramamurthi, B. (1988). Stereotactic operation in behaviour disorders. Amygdalotomy and hypothalamotomy. Acta Neurochirurgica – Suppl, 44, 152–7. Rapport, M.D. (1983). Carbamazepine and behavior therapy for aggressive behavior. Treatment of a mentally retarded, postencephalitic adolescent with seizure disorder. Behav Modification, 7, 255–65. Ratner, R.A. and Shapiro, D. (1979). The episodic dyscontrol syndrome and criminal responsibility. Bull Am Acad Psychiatry Law, 7, 422–31. Reis, D.J. (1969). Brain norepinephrine: evidence that neuronal release is essential for sham rage behavior following brainstem transection in cat. Proc Nat Acad Sci USA, 64, 108–12. Rodin, E.A. (1973). Psychomotor epilepsy and aggressive behavior. Arch Gen Psychiatry, 28, 210–13. Rolls, E.T. (1992). Neurophysiology and functions of the primate amygdala. In The Amygdala: Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction, ed. J.P. Aggleton, pp. 143–65. New York: Wiley-Liss. Saver, J.L., Salloway, S.P., Devinsky, O. and Bear, D.M. (1996). Neuropsychiatry of Aggression. In Neuropsychiatry, ed. B.S. Fogel, R.B. Schiffer and S.M. Rao, pp. 523–48. Baltimore: Williams & Wilkins. Shaikh, M.B. (1997). Serotonin 5-HT1A and 5-HT2/1C receptors in the midbrain periaqueductal gray differentially modulate defensive rage behavior elicited from the medial hypothalamus of the cat. Brain Res, 765, 198–207. Sheth, R.D. (1994). Aggression in children treated with clobazam for epilepsy. Clin Neuropharmacol, 17, 332–7. Soininen, H.S., Partanen, K., Pitkanen, A. et al. (1994). Volumetric MRI analysis of the amygdala and the hippocampus in subjects with age-associated memory impairment: correlation to visual and verbal memory. Neurology, 44, 1660–8. Stein, D.J., Hollander, E., Cohen, L. et al. (1993). Neuropsychiatric impairment in impulsive personality disorders. Psychiatry Res, 48, 257–66. Stone, J.L., McDaniel, K.D., Hughes, J.R. and Hermann, B.P. (1986). Episodic dyscontrol disorder and paroxysmal EEG abnormalities: successful treatment with carbamazepine. Biol Psychiatry, 21, 208–12. Swartz, M.S., Swanson, J.W., Hiday, V.A., Borum, R., Wagner, H.R. and Burns, B.J. (1998). Violence and severe mental illness: the effects of substance abuse and nonadherence to medication. Am J Psychiatry, 155, 226–31. Szabo, C.A. and Lancman, M.S.S. (1996). Postictal psychosis: a review. Neuropsychiatry, Neuropsychol Behav Neurol, 9, 258–64. Tebartz Van Elst, L., Woermann, F., Lemieux, L. and Trimble, M.R. (1999). Amygdala enlargement in dysthymia – A volumetric study of patients with temporal lobe epilepsy. Biol Psychiatry, 46, 1614–23. Tebartz Van Elst, L., Woermann, F., Lemieux, L., Thompson, P.J. and Trimble, M.R. (2000a). Affective aggression in patients with temporal lobe epilepsy – a quantitative magnetic resonance study of the amygdala. Brain, 123, 234–43.

105

Aggression and epilepsy Tebartz Van Elst, L., Woermann, F., Lemieux, L. and Trimble, M.R. (2000b). Increased amygdala volumes in female and depressed patients with temporal lobe epilepsy. Neurosci Lett, 281, 103–106. Thomson, C. (1989a). Affective disorders. In The Instruments of Psychiatric Research, ed. C. Thomson, pp. 34–46. New York: John Wiley & Sons. Thomson, C. (1989b). Anxiety. In The Instruments of Psychiatric Research, ed. C. Thomson, pp. 47–54. New York: John Wiley & Sons. Treiman, D.M. (1991). Psychobiology of ictal aggression. In Advances in Neurology, ed. D. Smith, D. Treiman and M.R. Trimble, Vol. 55, pp. 341–56. New York: Raven Press. Trimble, M.R. (1991). The Psychoses of Epilepsy. New York: Raven Press. Trimble, M.R. (1996) Biological Psychiatry. Chichester: John Wiley & Sons. Trimble, M.R., Ring, H.A. and Schmitz, B. (1996). Neuropsychiatric aspects of epilepsy. In Neuropsychiatry, ed. B.S. Fogel, R.B. Schiffer and S.M. Rao, pp. 771–803. Baltimore: Williams & Wilkins. Valzelli, L. (1981). Psychobiology of Aggression and Violence. New York: Raven Press. Van de Poll, N.E. and Van Goozen, S.H.M. (1992). Hypothalamic involvement in sexuality and hostility: comparative psychological aspects. Progr Brain Res, 93, 343–61. Van Paesschen, W., Connelly, A., Johnson, C.L. and Duncan, J.S. (1996). The amygdala and intractable temporal lobe epilepsy: a quantitative magnetic resonance imaging study [published erratum: Neurology (1997) 48, 1751]. Neurology, 47, 1021–31. Vitiello, B. and Stoff, D.M. (1997). Subtypes of aggression and their relevance to child psychiatry. J Am Acad Child Adolesc Psychiatry, 36, 307–15. Wada, Y. (1995). Prolonged but not acute fluoxetine administration produces its inhibitory effect on hippocampal seizures in rats. Psychopharmacology, 118, 305–9. Watson, C., Andermann, F., Gloor, P. et al. (1992). Anatomic basis of amygdaloid and hippocampal volume measurement by magnetic resonance imaging. Neurology, 42, 1743–50. Wieser, H.G. (1983). Electroclinical Features of the Psychomotor Seizures. Stuttgart: Fisher Butterworths. Woermann, F.G. (1999). Voxel-by-voxel comparison of automatically segmented cerebral gray matter – a rater-independent comparison of structural MRI in patients with epilepsy. NeuroImage, 10, 373–84. Woermann, F.G., Barker, G.J., Birnie, K.D., Meencke, H.J. and Duncan, J.S. (1998). Regional changes in hippocampal T2 relaxation and volume: a quantitative magnetic resonance imaging study of hippocampal sclerosis. J Neurol Neurosurg Psychiatry, 65, 656–64. Woermann, F.G., Tebartz Van Elst, L., Free, S.L., Thompson, P.J., Trimble, M.R. and Duncan, J.S. (1999). Reduction of frontal neocortical grey matter associated with affective aggression in patients with temporal lobe epilepsy – an objective voxel-by-voxel analysis of automatically segmented MRI. Epilepsia, 40, 205–6. (Abstract.) Woermann, F., Tebartz Van Elst, L., Koepp, M.J. et al. (2000). Reduction of frontal neocortical grey matter associated with affective aggression in patients with temporal lobe epilepsy. An objective voxel-by-voxel analysis of automatically segmented MRI. J Neurol Neurosurg Psychiatry, 68, 162–9.

106

L. Tebartz van Elst Yudofsky, S.C. (1990). Pharmacologic management of aggression in the elderly. J Clin Psychiatry, 51 (Suppl.), 22–28. Zanchetti, A. (1968). Reflex and brain stem inhibition of sham rage behaviour. Progr Brain Res, 22, 195–205.

8

Epilepsy and suicide: a neuropsychiatric analysis Dietrich Blumer University of Tennessee College of Medicine, Memphis, USA

Introduction According to eight reports, death by suicide occurs in 5% of patients with epilepsy, compared with 1.4% in the general population (Matthews and Barabas, 1981). Based on four reports, the four- to fivefold increase in suicides among patients with epilepsy over the rate in the general population is magnified to approximately 25–fold among patients with temporal lobe epilepsy (Barraclough, 1987). In a Danish study covering 14 years (Henriksen et al., 1970), 164 of 2763 patients with epilepsy died (an excess mortality rate of 273% compared to the number of deaths expected in the general population). While epilepsy was the cause of death in 26%, suicide was the cause of death in 20% (an excess mortality rate of 300%) at an average age at death of 32 years. For studies of an association between epilepsy and suicide to be statistically valid, the cohort must be standardized by age and sex, according to Stenager and Stenager (1992); most studies have not met these standards. The same authors pointed out that the variety of forms of epilepsy necessitates an examination of each syndrome singly for risk of suicide. Their second point is very significant, because the data available indicate that suicide is associated particularly with temporal lobe epilepsy. Suicide and attempted suicide Murphy (1994) points out that it is important to distinguish between suicide and attempted suicide. Differences between epileptic patients who attempt suicide and those who complete the act are not established. In the general population, suicides are three times as likely to be men as women, while attempted suicides are predominantly women at a 2:1 ratio. Suicides are about equally frequent before and after age 40, and the victims are most commonly suffering from a major psychiatric illness; attempted suicides are mostly under age 40 and are less likely to be suffering from one of the psychiatric illnesses most commonly associated with suicide. 107

108

D. Blumer

Attempted suicide is about 10 times more frequent than suicide. The lifetime risk of completed suicide after a suicide attempt is estimated to be slightly above 10%. Suicide is carried out by effective means, while attempted suicides tend to use what is at hand (80–90% of the attempts are by overdose). Several reports document the high frequency of suicide attempts by overdose among epileptic patients. The incidence of self-poisoning in epilepsy has been reported as sevenfold that of the general population (Mackay, 1979). Sixty-five percent of suicide attempts were carried out by ingesting antiepileptic drugs, 60% of which were barbiturates (Hawton et al., 1980). The psychotoxic role of the barbiturates in suicide attempts is emphasized by two further studies. In a population of 126 children and adolescents admitted to an emergency room for suicide attempt, nine had epilepsy – a 15-fold increase compared with the prevalence of epilepsy in the same age group. Of those nine epileptic children and adolescents, eight had been treated with phenobarbital (Brent, 1986). A much higher prevalence of suicidal ideation (47% vs. 4%) was noted among patients treated with phenobarbital compared with those treated with carbamazepine (Brent et al., 1987). The generally difficult psychosocial circumstances of patients with chronic epilepsy have been considered the leading factor responsible for their elevated suicide rate, more important even than the presence of psychiatric illness or the availability of drugs (Editorial, 1980). However, in general, psychiatric illness has been identified as the nearly universal antecedent of suicide (Murphy, 1994). Mendez et al. (1989) studied the causative factors for suicide attempts by overdose in epilepsy and concluded that interictal psychopathological factors were of primary importance. A comparison of suicide attempts among patients with epilepsy and comparably handicapped controls with other chronic disabilities found that 30% of those with epilepsy had attempted suicide as compared with 7% of the controls (Mendez et al., 1986). Psychosocial circumstances cannot be considered a sufficient precipitant for suicide attempts or suicide. In his review of the topic, Diehl (1986) ranked the risk factors for suicide in epilepsy as follows: (1) psychiatric disorders (psychotic episodes, dysphoric episodes, twilight states, personality disorders); (2) relatively young males (ages 25–49 years); (3) generalized and temporal lobe seizures (with brain lesions); (4) prolonged duration of the seizure disorder and inadequate therapy; (5) personal, social or occupational difficulties; and (6) availability of large amounts of antiepileptic drugs. While a good number of papers list statistical data on the topic, there is a regrettable paucity of neuropsychiatric case reports that allow for a better understanding of the psychopathology and pathogenesis specific to suicide in epilepsy and that might point at methods to prevent the fatal outcome.

109

Epilepsy and suicide

Series of suicides providing neuropsychiatric data Taylor and Marsh (1977) reported on the occurrence of suicide among 193 patients who had undergone temporal lobectomy and who were followed for 5 to 24 years. Of 37 deaths, nine were by suicide (six poisoned themselves). Including an additional six patients who died in unclear circumstances would have raised the suicide rate observed to 50-fold of that expected (Taylor, 1987). Five of the nine who definitely had committed suicide had been rendered seizure-free by the surgery. The mental state of the victims was not described. The paper is of exceptional interest for the statistics of suicide in epilepsy but is of limited value for our understanding of the neuropsychiatric problem. Mendez and Doss (1992) reported on the psychiatric aspects of four patients who died by suicide out of 1611 patients with epilepsy followed in a neurology clinic over a period of 8 years: two male patients with chronic psychosis and good seizure control; one male patient with brief psychotic episodes associated with confusion and increased bitemporal spikes and diffuse slowing on EEG in the absence of seizures; and one female patient with profound ictal and postictal depression who committed suicide after three witnessed staring spells. The patient with brief psychotic episodes and one of the patients with chronic psychosis experienced voices commanding them to commit suicide. All four patients had suffered from complex partial seizures since childhood. All four patients committed suicide by medication overdose. Of our entire population treated for epilepsy at the Epi-Care Center in Memphis over the past 12 years (10739 patients), five patients have committed suicide. All had a history of early onset (mean age 9.5 years) of longstanding complex partial seizures (mean duration 29 years), with very high (often daily) seizure frequency in all but one. Suicide occurred in all patients after a short interval (mean 13 months) of having obtained full control of seizures for the first time by temporal lobectomy (three patients), medication (one patient), or vagus nerve stimulation (one patient). Three patients had a previous history of suicidal moods or suicide attempts, but in three of the five, the suicidal act was precipitate and not anticipated at the time. Four were male and one was female. None of the deaths was caused by overdose: one by hanging, one by drowning and three by gunshot. Role of the interictal and peri-ictal psychopathology in suicide The premodern psychiatrists who established the basis for our modern classification of mental disorders noted that specific mental changes were associated with epilepsy. They had the advantage of observing institutionalized patients with chronic epilepsy over prolonged periods and were familiar with the characteristic

110

D. Blumer

intermittent and pleomorphic changes that have eluded the modern cross-sectional psychological assessment of patients with epilepsy. Modern assessments are usually carried out with methods that have been validated for use in patients who do not have epilepsy and that are insensitive to the task. Premodern psychiatrists had arrived at the concept of the dysphoric disorder as the most common psychiatric disorder of epilepsy, a distinct disorder that has only recently been rediscovered (Blumer et al., 1995). One of its key symptoms is associated with the episodic suicidal moods of patients with chronic epilepsy, i.e. mesial temporal lobe epilepsy (Blumer, 2000; Blumer et al., 1995; Gastaut, 1956). Kraepelin (1923) precisely described the intermittent dysphoric disorder of patients with epilepsy. Dysphoric episodes present with depressive moods (‘very frequently with utter disgust of life and suicidal bent’), irritability, anxiety, headaches, insomnia or at times with euphoric moods. These polysymptomatic dysphoric episodes occur without external triggers with rapid onset and termination and recur fairly regularly in a uniform manner in the absence of clouding of consciousness. Dysphoric symptoms can be observed as prodromes of an attack or in the aftermath of an attack, but they most commonly appear as phenomena independent of the seizures, with a frequency varying from every few days to every few months. A patient just awakens one day dysphoric, or the dysphoria develops insidiously through the course of a day. As a rule, the dysphoric state lasts from 1 to 2 days but might dissipate after just a few hours. Based on our own observations, we have added anergia and phobic fears to Kraepelin’s six key symptoms of the dysphoric disorder and have defined it by the presence of at least three of the eight key symptoms, each present to a troublesome degree. We have noted an average of five key symptoms among our patients with dysphoric disorder (Blumer et al., 1995). The risk of suicide in patients with epilepsy is primarily associated with the episodes of intense depressive mood that occur during the interictal phase of patients with a dysphoric disorder, and suicide in epilepsy tends to occur in a precipitate manner. As already noted by Kraepelin, the dysphoric symptoms also tend to occur peri-ictally, during the prodrome or in the aftermath of a seizure. The postictal phase in particular may be associated with marked depressive mood (Blumer, 1992). A high suicidal risk has been observed in patients who experience ictal depressive mood that extends into the postictal phase for a period of 1 hour to 3 days. Williams (1956) reported 21 such cases among his 100 patients with ictal emotional experience, and 5 of the 21 patients made suicide attempts during their postictal phase. As noted by Kraepelin, interictal psychoses tend to develop among patients with interictal dysphoric disorder (Blumer et al., 2000; Kraepelin, 1923). The dysphoric disorder persists during the psychotic state, and intense depressive moods may occur in the course of an interictal psychosis. The presence of the hallucina-

111

Epilepsy and suicide

tion of voices commanding patients to kill themselves represents a particular suicidal risk. The psychopathology of four patients (Mendez and Doss, 1992) who committed suicide is not reported in detail beyond the psychotic episodes in three of them and the ictal and postictal depression in the fourth patient; the researchers state only that two of those with psychosis also experienced depressive episodes. Four of the five patients from the Epi-Care series had longstanding dysphoric disorders. The fifth patient who had been found free of dysphoric symptoms earlier was not examined during the 3-month interval from finally (after 20 years) becoming seizure-free to his death by suicide; it was learned, however, that he had begun to experience episodes of rage, symptomatic of a de novo dysphoric disorder. Pathogenesis of the psychiatric disorders of epilepsy Suicide in epilepsy appears to be caused by the interictal and sometimes by postictal psychopathology. Since the early reports by Gibbs (1951), Landolt (1953), and Hill (1953), evidence has been increasing that the interictal psychopathology tends to emerge or to worsen upon improvement of the epilepsy as measured by seizure frequency and EEG abnormalities. There is persuasive evidence that the psychiatric disorders of epilepsy may result from the inhibitory activity that develops in reaction to the excessive excitatory activity of the chronic seizure disorder, as postulated by Stevens (1975) and Engel (1989; Chapter 3). The precise nature of the seizure-suppressing mechanisms is insufficiently understood, and the phenomenon is usually referred to as forced normalization (Landolt, 1955, 1958, 1963). The evidence of a linkage of the psychiatric changes to inhibitory mechanisms can be summarized as follows: 1. The development of interictal dysphoric and psychotic disorders is delayed following the onset of epilepsy (Gastaut et al., 1955; Slater and Beard, 1963) as inhibitory mechanisms become increasingly established; this situation accords with the particular linkage of the psychiatric disorders of epilepsy with its most prominent chronic form – mesial temporal lobe epilepsy. 2. Upon decrease and particularly upon full control of seizures, dysphoric symptoms and psychosis tend to be exacerbated or to emerge de novo. 3. Psychiatric changes emerge also at times when acute exacerbation of the seizure activity engages an enhanced inhibitory response, commonly in the prodromal and postictal phases, and rarely upon increased seizure activity as a seeming opposite of forced normalization (Blumer et al., 2000). 4. There is a delayed phasing-out of the psychiatric changes after surgical elimination of the epileptogenic zone, presumably with only gradual fading of the inhibitory mechanism (Blumer et al., 1998).

112

D. Blumer

It is hypothesized that the dysphoric and psychotic symptoms, as well as the suicidality of epilepsy, are related, as are the seizures themselves, to the homeostatic ebb and flow of excitatory and inhibitory influences. All patients of the two series of suicides providing neuropsychiatric details (Mendez and Doss, 1992; Epi-Care patients noted above) had an onset of epilepsy in early life, with a mean duration of the seizure disorder of 25 and 29 years, respectively. This interval exceeds the mean interval from onset of epilepsy to the manifestation of interictal psychosis, reported as 14 years (Slater and Beard, 1963). Two of the three patients who committed suicide in a psychotic state were under good seizure control. The third patient showed the rare finding (Demers-Desrosiers et al., 1978) of what seems the opposite of forced normalization: psychotic episodes coinciding with the presence of increased electroencephalographic epileptiform potentials in the absence of seizures, presumably resulting from the forceful engagement of inhibitory mechanisms in response to enhanced subclinical seizure activity (Blumer et al., 2000). The increased suicide risk among patients with ictal and persistent postictal depression (as in the fourth case of Mendez and Doss, 1992) was previously documented by Williams (1956). In our Epi-Care series, the striking finding of all five suicides, which occurred shortly after a longstanding seizure disorder had become controlled, points to the risk of worsening (or de novo appearance) of a marked dysphoric disorder once predominance of inhibitory mechanisms is established. It must be noted again that five of the nine patients from the British series who committed suicide after temporal lobectomy had been rendered seizure-free (Taylor and Marsh, 1977). Preventing suicide in epilepsy Preventing suicide in epilepsy patients consists of effectively treating both the dysphoric disorder and the psychosis of the interictal phase (Blumer, 1997; Blumer et al., 2000; Blumer and Zielinski, 1988). We now treat both the patients with suicidal dysphoric moods and those with interictal psychoses with double antidepressant medication, enhanced if necessary with an atypical neuroleptic drug, e.g. with the combination of 100–150 mg imipramine, 20–40 mg paroxetine and 2–4 mg risperidone daily. The same treatment has been effective for patients with severe postictal depressive mood, although we have not had the occasion to treat a patient with ictal depression and suicidal intensity of the postictal phase. The dysphoric disorder is endogenous, and psychotherapy without pharmacotherapy leaves the patient with suicidal moods at risk. The bias against using antidepressants for the psychiatric disorders of epilepsy on the grounds that they may lower the seizure threshold (McConnell and Duncan,

113

Epilepsy and suicide

1998) is erroneous on both empirical and theoretical grounds. As our experience over some 15 years has shown, the modest amounts of antidepressant medication required do not increase seizure frequency in patients with chronic epilepsy whose interictal dysphoric disorder or psychosis indicates the presence of marked inhibition (Blumer, 1997; Blumer et al., 2000; Blumer and Zielinski, 1988). Gastaut et al. (1955) have pointed out that patients with temporal lobe epilepsy (in contrast to those with primary generalized epilepsy) show a higher interictal seizure threshold than do individuals without epilepsy. The proconvulsant effect of antidepressants does not provoke seizures but may serve to mitigate the psychotoxic effect of excessive inhibition. Avoiding antidepressants for a patient with suicidal dysphoric moods may have fatal consequences. The 10739 patients with epilepsy seen at the Epi-Care Center for the past 12 years were evaluated and treated by a comprehensive team consisting of a neurologist, neurosurgeon, psychiatrist and neuropsychologist. The five suicides among this population represent a much lower fatality rate if compared with the population reported by Mendez and Doss (1992). This decreased rate may result from the ready availability of a psychiatrist for treatment of every patient with dysphoric disorder or psychosis. There are limitations to suicide prevention, even if one is alerted to the risk. Of our patients who committed suicide, two were noncompliant with our treatment and two, for geographic reasons, were followed elsewhere prior to their suicide. In retrospect, our fifth patient, who had become dysphoric after control of his seizures by medication, should have been brought in for treatment upon the emergence of an episode of rage that preceded his suicidal act. In our experience, timely psychopharmacological treatment of the dysphoric disorder or of an interictal psychosis can usually prevent a suicidal outcome of the epileptic disorder. Conclusion Further neuropsychiatric studies of patients with epilepsy who commit suicide need to be made available before one can draw more definite conclusions about all the modes of suicide in epilepsy. The current review allows a number of conclusions, similar to those reached earlier by Diehl (1986), but more specific in nature. Suicide in epilepsy results from the psychiatric disorder of temporal lobe epilepsy; that is, from a severe dysphoric disorder, from interictal psychosis (associated with preceding and concomitant dysphoric disorder, and at times with command hallucinations), or from a severe postictal depressive state. These psychiatric disorders develop gradually as seizure-suppressing mechanisms become established or, at times, upon acute engagement of the inhibition. Suicide in epilepsy has increased with our improved ability to suppress seizures.

114

D. Blumer

Patients with early onset of temporal lobe epilepsy and prolonged duration of the illness (more than 20 years) are at particular risk of suicide once their seizures are suppressed. Males in the age range of 30–50 are more at risk. Treatment with barbiturates, availability of drugs, loss of loved ones or of jobs, and other difficult psychosocial predicaments are not aetiological factors but may contribute to the suicide risk. Suicide in epilepsy tends to occur precipitately during a ‘fit of melancholy’ (as van Gogh described the depressive mood of his dysphoric episodes) and is often not anticipated. However, there are usually warnings that precede a suicide. Upon the occurrence of episodes of suicidal moods, prompt intervention is required with psychotropic medication, chiefly of the antidepressant type, and with careful follow-up that includes adjusting the psychotropic medication as needed. Although transient suicidal moods among epileptic patients were observed frequently by premodern psychiatrists, completed suicide was not often reported (Delay et al., 1957). Deaths from seizures have been markedly decreased by our progress in seizure control but may be surpassed by now in numbers by deaths from suicide. Our ability to suppress seizures must become paired with our ability to treat the psychiatric consequences of improved seizure control. Acknowledgement The study of the patients seen at the Epi-Care Center was made possible by the collaboration of Drs Keith Davies, Bruce Hermann, Georgia Montouris and Allen Wyler.

R E F E R E N C ES Barraclough, B.M. (1987). The suicide rate of epilepsy. Acta Psychiatr Scand, 76, 339–45. Blumer, D. (1992). Postictal depression: significance for the treatment of the neurobehavioral disorder of epilepsy. J Epilepsy, 5, 214–19. Blumer, D. (1997). Antidepressant and double antidepressant treatment for the affective disorder of epilepsy. J Clin Psychiatry, 58, 3–11. Blumer, D. (2000). Dysphoric disorders and paroxysmal affects: recognition and treatment of epilepsy-related psychiatric disorders. Harvard Rev Psychiatry, 8, 8–17. Blumer, D. and Zielinski, J. (1988). Pharmacologic treatment of psychiatric disorders associated with epilepsy. J Epilepsy, 1, 135–50. Blumer, D., Montouris, G. and Hermann, B. (1995). Psychiatric morbidity in seizure patients on a neurodiagnostic monitoring unit. J Neuropsychiatry Clin Neurosci, 7, 445–56. Blumer, D., Wakhlu, S., Davies, K. and Hermann, B. (1998). Psychiatric outcome of temporal lobectomy for epilepsy: incidence and treatment of psychiatric complications. Epilepsia, 39, 478–86.

115

Epilepsy and suicide Blumer, D., Wakhlu, S., Montouris, G. and Wyler, A. (2000). Treatment of the interictal psychoses. J Clin Psychiatry, 61, 110–22. Brent, D.A. (1986). Overrepresentation of epileptics in a consecutive series of suicide attempts seen at a children’s hospital, 1978–1983. J Am Acad Child Psychiatry, 25, 242–6. Brent, D.A., Crumrine, P.K., Varma, R.R., Allan, M. and Allman, C. (1987). Phenobarbital treatment and major depressive disorder in children with epilepsy. Pediatrics, 80, 909–17. Delay, J., Deniker, P. and Barande, R. (1957). Le suicide des épileptiques. Encéphale, 46, 401–36. Demers-Desrosiers, L.A., Nestoros, J.N. and Vaillancourt, P. (1978). Acute psychosis precipitated by withdrawal of anticonvulsant medication. Am J Psychiatry, 135, 981–2. Diehl, L.W. (1986). Epilepsie und Suizid. Psychiat Neurol Med Psychol, 38, 625–33. Editorial (1980). Suicide and epilepsy. Br Med J, 281, 530. Engel, J. (1989). Seizures and epilepsy. Philadelphia: Davis. Gastaut, H. (1956). État actuel des connaissances sur l’anatomie pathologique des épilepsies. Acta Neurol Psychiatr Belg, 56, 5–20. Gastaut, H., Morin, G. and Lesèvre, N. (1955). Étude du comportement des épileptiques psychomoteurs dans l’intervalle de leurs crises: les troubles de l’activité globale et de la sociabilité. Ann Med Psychol, 113, 1–27. Gibbs, F.A. (1951). Ictal and non-ictal psychiatric disorders in temporal lobe epilepsy. J Nerv Ment Dis, 113, 522–8. Hawton, K., Fagg, J. and Marsack, P. (1980). Association between epilepsy and attempted suicide. J Neurol Neurosurg Psychiatry, 43, 168–70. Henriksen, B., Juul-Jensen, P.P. and Lund, M. (1970). The mortality of epileptics. In Life Assurance Medicine, ed. R.D.C. Brackenridge, pp. 139–48. London: Pitman. Hill, D. (1953). Psychiatric disorders of epilepsy. Med Press, 229, 473–5. Kraepelin, E. (1923). Psychiatrie, 8th edn. Leipzig: Barth. Landolt, H. (1953). Some clinical electroencephalographical correlations in epileptic psychoses (twilight states) [abstract]. Electronencephalogr Clin Neurophysiol, 5, 121. Landolt, H. (1955). Über Verstimmungen, Dämmerzustände und schizophrene Zustandsbilder bei Epilepsie: Ergebnisse klinischer und electroenzephalographischer Untersuchungen. Schweiz Arch Neurol Psychiatr, 76, 313–21. Landolt, H. (1958). Serial electroencephalographic investigations during psychotic episodes in epileptic patients and during schizophrenic attacks. In Lectures on Epilepsy, ed. A.M. Lorentz de Haas, pp. 91–133. Amsterdam: Elsevier. Landolt, H. (1963). Die Dämmer- und Verstimmungszustände bei Epilepsie und ihre Electroenzephalographie. Dtsch Z Nervenheilkunde, 185, 411–30. Mackay, A. (1979). Self-poisoning – a complication of epilepsy. Br J Psychiatry, 134, 277–82. McConnell, H.W. and Duncan, D. (1998). Treatment of psychiatric comorbidity in epilepsy. In Psychiatric Comorbidity in Epilepsy: Basic Mechanisms, Diagnosis, and Treatment, ed. H.W. McConnell and P.J. Snyder, pp. 245–361. Washington, DC: American Psychiatric Press. Matthews, W.S. and Barabas, G. (1981). Suicide and epilepsy: a review of the literature. Psychosomatics, 22, 515–24. Mendez, M.F. and Doss, R.C. (1992). Ictal and psychiatric aspects of suicide in epileptic patients. Int J Psychiatry Med, 22, 231–7.

116

D. Blumer Mendez, M.F., Cummings, J.L. and Benson, D.F. (1986). Depression in epilepsy. Significance and phenomenology. Arch Neurol, 43, 766–70. Mendez, M.F., Lanska, D.J., Manon-Espaillat, R. and Burnstine, T.H. (1989). Causative factors for suicide attempts by overdose in epileptics. Arch Neurol, 46, 1065–8. Murphy, G.E. (1994). Suicide and attempted suicide. In The Medical Basis of Psychiatry, ed. G. Winokur and P.J. Clayton, pp. 529–44. Philadelphia: W.B. Saunders. Slater, E. and Beard, A.W. (1963). The schizophrenia-like psychoses of epilepsy. Br J Psychiatry, 109, 95–112. Stenager, E.N. and Stenager, E. (1992). Suicide and patients with neurologic diseases. Methodologic problems. Arch Neurol, 49, 1296–303. Stevens, J.R. (1975). Interictal clinical manifestations of complex partial seizures. In Advances in Neurology, Vol. 11: Complex Partial Seizures and their Treatment, ed. J.K. Penry and D.D. Daly, pp. 85–112. New York: Raven. Taylor, D.C. (1987). Psychiatric and social issues in measuring the input and outcome of epilepsy surgery. In Surgical Treatment of the Epilepsies, ed. J. Engel Jr, pp. 485–503. New York: Raven Press. Taylor, D.C. and Marsh, S.M. (1977). Implications of long-term follow-up studies in epilepsy: with a note on the cause of death. In Epilepsy, the Eighth International Symposium, ed. J.K. Penry, pp. 27–35. New York: Raven Press. Williams, D. (1956). The structure of emotions reflected in epileptic experiences. Brain, 79, 29–67.

9

Postictal psychoses, revisited Kousuke Kanemoto Aichi Medical University, Aichi, Japan

Historical background Except for the immediate effects of a seizure on mental function, such as complex partial status epilepticus and postictal confusion, modern epileptic psychoses can be categorized into three main types: chronic, acute interictal and postictal psychoses. In 1953, Landolt stressed a seesaw relationship between epileptic seizures and psychoses, and proposed the concept of forced normalization. In 1963, Slater made a rather comprehensive report on chronic psychoses in patients with epilepsy (Slater and Beard, 1963). In contrast, it was as late as 1988 before the concept of postictal psychoses was revived by Logsdail and Toone. This delay in conceptual formation is all the more peculiar, when considering the very old root of the concept of postictal psychosis. In 1860, a French psychiatrist, Farlet, classified epileptic psychoses into three categories: transient peri-ictal, chronic and true epileptic psychosis (Farlet, 1860/1961). As there was a lack of strict distinctions between preictal, intraictal and postictal events at that time, it is not easy to compare Farlet’s classification to those of the present. While transient peri-ictal psychosis overlaps postictal confusion and the meaning of chronic psychosis is evident, Farlet’s true epileptic psychosis has no clear counterpart in modern classifications. Farlet assigned extreme psychomotor agitation as well as extraordinarily aggressive and self-destructive behaviour to his unique classification. As I will discuss later, these are salient psychopathological traits of postictal psychoses. Indeed, John Hughlings Jackson (1875), a pioneer of modern epileptology, stressed that the true epileptic psychosis described by Farlet often followed clusters of seizures. However, authors of subsequent psychiatric literature have confused Farlet’s true epileptic psychosis with other types of epileptic psychosis and even postictal confusion. While the literature has misunderstood the true nature of the epileptic psychosis of Farlet for more than 100 years (Kolb and Brodie, 1982; Taylor, 1972), that in the domain of epileptology has gradually forgotten it. As a result, and with the advent of antiepileptic drugs, the alternative psychosis of Landolt (1963) became very popular, obscuring postictal psychosis completely (Mendez et al., 1993; Onuma et al., 1995). 117

118

K. Kanemoto

This longstanding neglect, however, is being rapidly rectified, resulting in several review articles (Lancman, 1999; Sachdev, 1998). Episodes of postictal psychosis can no longer be ignored, as the intensive seizure monitoring under the diminution or discontinuation of antiepileptic drugs currently practiced has dramatically increased the chances for those who are involved in epileptic surgery to observe postictal psychosis directly (Devinsky et al., 1995; Kanner et al., 1996). In this chapter, we pay special attention to the natural history of postictal psychoses outside of the seizure monitoring unit, in the belief that such a patient sampling method would present in-vivo postictal psychoses, instead of the quasi-in-vitro ones produced in the intensive care unit. Following a description of two representative cases of postictal psychoses clinical data concerning postictal psychosis are reviewed. Case reports CASE 1 A 28-year-old kimono shop manager had a 20-year history of paroxysmal fearful feelings of being left alone. At age 11, complex partial seizures began to follow these moments of fear. As his age advanced, his seizures increased in intensity as well as frequency, despite maximum drug therapy. At age 26, the first manifest postictal mental derangement occurred, after several bouts of complex partial seizures. One day after this cluster of seizures, he struck his father, the owner of the kimono shop, who had asked whether he was all right, out of uncontrollable rage. This peculiar dysphoric state lasted for a week, during which he was continuously prone to violent behaviour over minor matters. He reported that alien ideas had invaded him and that opposing thoughts battled each other during this state. Afterwards, he could recall perfectly the details of his violent behaviour, but could not understand why he had acted in such a manner. Such postictal episodes and repeated violent behaviours tormented those around him, and recurred every 2 months before admission. An MRI revealed a marked asymmetry of the hippocampi (the left side was smaller than the right) with a lower signal intensity from the left hippocampus on a reversed T2 condition. Although he was righthanded, a 60 mg injection of Amytal into the right, but not the left, carotid artery caused the patient to become aphasic. Ictal EEG recordings, including those with depth, plus subdural electrodes, unanimously suggested that the left hippocampus was the origin of both the simple and complex partial seizures. Subsequently, a left inferior lobectomy with a hippocampoamygdalotomy was performed. The resected tissue revealed Ammon’s horn sclerosis. On awakening from anaesthesia, the patient showed a peculiar dysphoria, just as he had during his postictal psychoses, which spontaneously disappeared within 3 days. However, one month after the lobectomy, he became increasingly euphoric and elevated in mood. His speech was loud, rapid and difficult to interrupt, while full of jokes, puns and other amusing irreverence. He even exposed his genitalia in public during his elevated moods and courted several copatients. Since he sang throughout the night, we needed strong sedatives to put

119

Postictal psychoses, revisited him to sleep. This manic state, lasting for 4 weeks, gradually turned into a depressive one. Two months after the operation, he felt quite upset and lost interest in all activities. He became so agitated that he could not stand still for a moment, and walked around restlessly. He complained of slowed thinking and difficulties in making decisions. After treatment with tricyclic antidepressants, his mood improved steadily, but only gradually, over half a year. Six months after the operation, he set out to do his previous work and his relations with colleagues improved dramatically. A year after the operation, his mood was stabilized without the help of antidepressants and he was accepted once again as the manager of the kimono shop. He has been completely seizure-free for 4 years postoperatively. Moreover, neither the previous dysphoric episodes nor the mood disorders have recurred. CASE 2 At age 3, a 38-year-old housewife suffered from an episode of prolonged febrile convulsions, which lasted for more than 30 minutes and resulted in transient paralysis of the left upper extremities. At age 12, the first episode of complex partial seizures occurred, in which she unknowingly handed over an examination paper to a classmate who happened to be sitting next to her. After that, every time during a seizure, she would unconsciously reiterate the same phrase; ‘He’s coming to collect my examination paper. What should I do?’. At age 15, paroxysmal feelings of a peculiar familiarity began to precede the complex partial seizures, during which she felt that the atmosphere of her environment suddenly changed and it seemed as if she had dissolved into the immediate surroundings. Her seizures remained uncontrollable despite intensive medication. As the patient’s age advanced, she joined a local religious sect as a devoted member and became increasingly eccentric. The first manifest postictal psychosis occurred after bouts of complex partial seizures at age 29. After an intervening 36-hour lucid interval, she would rapidly become more and more elevated in mood, with loud rapid speech that was difficult to interrupt. She would change subjects kaleidoscopically from one to another. She screamed to her husband repeatedly, ‘I love you, darling’, and hugged and kissed him in public in a sensual manner. Three days after the cluster of seizures, the euphoric state culminated in agitated exaltation. She said, ‘I am directly feeling all that is happening in every corner of the world through the palpitating movement of my teeth. The circular movements of my teeth are synchronized with the circular movements of the world. Through nerves in my teeth, I can sense the future 2000 years from now’. Because of extreme psychomotor agitation, a short stay in the psychiatric ward as well as potent sedatives were required. This state disappeared completely within 10 days. Throughout the episode, her orientation and memory remained intact. After several episodes of such postictal psychotic states, she agreed to surgery. An MRI revealed a marked asymmetry of the hippocampi (the left side was smaller than the right) with a lower signal intensity from the left hippocampus on a reversed T2 condition. Although she was right-handed, her dominant language side proved to be right. Ictal EEG recordings unanimously suggested that the left hippocampus was the origin of both the simple and complex partial seizures. During the course of intensive seizure monitoring, postictal psychosis recurred once after a cluster of complex partial seizures. Subsequently,

120

K. Kanemoto a left inferior lobectomy with a hippocampoamygdalotomy was performed and the resected tissue revealed Ammon’s horn sclerosis. She recovered steadily without complications. A year after the operation, the patient began to work as a manager of a Japanese restaurant. She completely lost interest in the religious activity as if exorcised. She has been completely seizure-free for 7 years postoperatively and no episodes of postictal psychosis have recurred.

Clinical studies In our work, we re-examined all of the outpatient cases from 1984 to 1999 at the Kansai Regional Epilepsy Center who were known to have had epilepsy with psychotic episodes (n⫽177). Epilepsy and seizure classifications were based on definitions proposed by the International League against Epilepsy (Commission on Classification and Terminology of the International League Against Epilepsy, 1981; 1989). In our study, psychosis was defined according to the following ICD–10 criteria: the presence of hallucinations, delusions, or a limited number of severe abnormalities of behaviour, such as gross excitement and overactivity, and catatonic behaviour (World Health Organization, 1992). However, we excluded psychomotor retardation from the original definition. Ictal psychotic episodes directly corresponding to ictal epileptiform discharge, such as nonconvulsive status epilepticus, were also excluded from psychotic episodes. Postictal psychosis was defined as one that occurred within 7 days after the last generalized tonic-clonic seizures or clusters of complex partial seizures. We excluded those patients who exhibited postictal psychosis only during or immediately after intensive seizure monitoring. In this study, all episodes of psychosis, except for postictal psychosis, were included in the category of interictal psychosis. Patients with continuous hallucinations or delusions without remission were regarded as having chronic psychosis. Patients with interictal psychosis but without chronic psychosis were included with the acute interictal psychosis cases. We assessed the psychopathological features of psychotic episodes according to a modified scale for assessing positive symptoms (SAPS; Andreasen, 1984), for which the test procedure details have been described in a previous study (Kanemoto et al., 1996a). Statistical analyses were made with Chi-square tests, with Yates’ modification for small numbers. Incidence

Fifty-one (2%) out of 2905 patients with epilepsy treated at Kansai Regional Epilepsy Center experienced postictal psychoses that were not artificially induced (Table 9.1). It is difficult to compare our data with previous studies, as they are either multiple case reports (Lancman et al., 1994; Levin, 1952; Logsdail and Toone, 1988; Savard et al., 1991; Umbricht et al., 1995) or based on observations during

121

Postictal psychoses, revisited

Table 9.1. Incidence of epileptic psychosis

Patients with psychotic disorder Postictal psychosis* Acute interictal psychosis* Chronic psychosis*

177 (6%) 51 (2%) 75 (3%) 57 (2%)

Patients without psychotic disorder

2728 (94%)

Total

2905 (100%)

Note: * Six patients experienced both postictal and interictal psychoses.

the seizure monitoring in preparation for epilepsy surgery (Devinsky et al., 1995; Kanner et al., 1996). However, our finding seems to have a certain reliability, because the prevalence of interictal psychosis in patients with epilepsy in the current study (5%) agreed well with that seen in other specialized epilepsy clinics (4–9%) (Edeh and Toone, 1987; Mendez et al., 1993; Onuma et al., 1995). The prevalence of postictal psychosis in patients with temporal lobe epilepsy (11%) proved to be far higher than that in the general epilepsy population. Kanner et al. (1996) reported that the annual incidence of postictal psychiatric events, including postictal psychosis, at their monitoring unit was 7.8%. Seven out of the 13 patients in their series had their first-ever postictal psychiatric event during the monitoring study, therefore, at most only 4% experienced a truly spontaneous postictal psychiatric event. Considering that their study was limited to patients with symptomatic localization-related epilepsy, this figure is approximate to ours. Lucid interval, recurrence and duration of postictal psychosis

Postictal psychosis has long been confused with the clouded consciousness observed following complex partial seizures. Recent studies, including ours, have proven that postictal psychosis cannot be reduced to a mere extension of postictal confusions. The primary argument in support of this is twofold: (1) there is preserved orientation and memory seen during the episodes, and (2) lucid intervals occur between the end of the seizures and the start of postictal psychoses (Kanemoto et al., 1996a; Levin, 1952; Logsdail and Toone, 1988; Savard et al., 1991). It is very difficult to explain the delayed manifestation of positive symptoms after a short period of normality as a psychic equivalent of Todd’s paralysis, since a mere exhaustion of the nervous system would have steadily recovered without reversion, just as in postictal confusion. In 18 out of 51 patients, we confirmed the presence of a lucid interval, which lasted from 1 to 3 days in 15 out of the 18 (83%) (Table 9.2). This agrees well with previous studies (Kanemoto et al., 1996b; Kanner

122

K. Kanemoto

Table 9.2. Presence of a lucid interval in 18 of 51 patients observed

Less than 24 hours 1 day 2 days 3 days 4–7 days Undetermined

1 4 6 5 2 33

Table 9.3. Duration of the postictal psychosis

12–24 hours 1–7 days 1 week to 1 month Longer than 1 month Undetermined

4 16 13 3 15

et al., 1996; Logsdail and Toone, 1988; Savard et al., 1991). In comparison with acute interictal psychosis, the duration of postictal psychosis is relatively short. In more than half of the patients, psychotic episodes disappeared within a week, and psychotic states lasting for more than one week but less than a month were seen in another one third (Table 9.3). In a certain proportion of patients, postictal psychosis has been shown to have a tendency to recur (Kanner et al., 1996; Lancman et al., 1994). In our series, 49% of the patients with postictal psychosis experienced one or more recurrences. Patient background

In view of the common features that patients with interictal and postictal psychosis share, such as a comparatively long latent period between epilepsy and psychosis onset (longer than 10 years; Table 9.4) and a close association with temporal lobe epilepsy, the prevailing view has been that interictal and postictal psychoses are probably similar (Savard et al., 1991). However, we have found that postictal and interictal psychoses differ in several fundamental demographic data. First, age at epilepsy onset was significantly younger in patients with interictal psychosis than in those with postictal psychosis. Second, the latent period between psychosis and epilepsy onset was still longer in postictal than interictal psychosis. Third, the proportion of those patients with reduced intelligence quotient (IQ) was significantly

123

Postictal psychoses, revisited

Table 9.4. Patient background

Demographic dataa

Postictal psychosis (n⫽45) (SD)

Interictal psychosis (n⫽126) (SD)

M/F Age (year) Age at epilepsy onset (year) Duration of epilepsy (year) Age at psychosis onset (year) Latent period Low intelligenceb

29/16 41.4 (12.5) 16.0 (11.3) 25.6 (14.1) 34.5 (10.3) 18.1 (10.7) 4/45

78/48 32.2 (9.8) 11.2 (6.2) 20.5 (10.8) 24.8 (7.3) 13.2 (8.5) 39/126

Notes: There was a highly significant difference between the groups (P⬍0.005) in all items except for sex. b Full Scale IQ⬍70. a

higher in patients with interictal psychosis than in those with postictal psychosis. These data agree well with the report of Umbricht et al. (1995), which confirmed lower IQ and younger age at onset of epilepsy in interictal psychosis than in postictal psychosis. Association with temporal lobe epilepsy

Our data support the suggestion of a majority of previous studies that psychosis in epilepsy might be preferentially associated with the temporal lobe (Bruens, 1971; Edeh and Toone, 1987; Gibbs, 1951; Perez and Trimble, 1980; Toone et al., 1980). However, a direct comparison between postictal and interictal psychosis revealed that the close link between psychotic episodes and temporal lobe epilepsy was significantly more remarkable in postictal than in interictal psychosis (Table 9.5). Our data provided additional supportive evidence in this respect. First, patients with postictal psychosis exhibited a temporal lobe pathology significantly more often on an MRI than those with interictal psychosis. Second, electroencephalographically, generalized spikes and waves were significantly more often recorded in interictal psychosis than in postictal psychosis patients (Table 9.6). Furthermore, as for seizures, complex partial seizures were more closely associated with postictal psychosis than interictal psychosis (Table 9.7). These data support the view that the involvement of the limbic structure is almost always indicated in cases of postictal psychosis and that postictal psychosis, in contrast to the alternative psychosis of Landolt, almost never occurs in idiopathic generalized epilepsy.

124

K. Kanemoto

Table 9.5. Epilepsy classification

Idiopathic localization-related epilepsy Symptomatic localization-related epilepsy Temporal lobe epilepsyb Extra-temporal lobe epilepsy Idiopathic generalized epilepsy Symptomatic/cryptogenic generalized epilepsy Others

Postictal psychosis (n⫽45a)

Interictal psychosis (n⫽126a)

No psychotic episode (n⫽2728)

—

—

42

39 10 — — 2

74 35 5 7 11

422 857 308 369 730

Notes: a Six patients with both postictal and interictal psychoses were excluded. b Chi-square⫽5.14, statistically significant difference (p⬍0.05).

Table 9.6. Laterality and localization of the laboratory findings

Postictal psychosis (n⫽45)

Interictal psychosis (n⫽126)

EEG findings Temporal foci Extra-temporal foci Sidedness (L/R) Diffuse SWC

33 8 11/18 1

78 15 43/41 21b

MRI localization Temporala Extra-temporala Sidedness (L/R)

16 4 12/9

25b 14 34/20

Notes: a Patients with both temporal and extra-temporal pathology were excluded. b Statistically significant difference (P⬍0.05). SWC, spike and wave complex.

125

Postictal psychoses, revisited

Table 9.7. Seizures in patients with postictal or interictal psychoses

SPS CPS GTC Minor GS

Postictal psychosis (n⫽45)

Interictal psychosis (n⫽126)

25 37 36 0

79 84a 97 9

Notes: Statistically significant difference (P⬍0.05). SPS, simple partial seizures; CPS, complex partial seizures; GTC, generalized tonic-clonic seizures; minor GS, minor generalized seizures including absence, generalized myoclonic seizure and generalized tonic seizure. a

Table 9.8. Simple partial seizures in localization-related epilepsy

Autonomic Déjà vu Motor Elementary visual

Postictal psychosis (n⫽43)

Interictal psychosis (n⫽103)

16 10 4 0

40 10a 15 11

Note: a Statistically significant difference (P⬍0.05).

In a previous study based on MRI results (Kanemoto et al., 1996b), we suggested that medial temporal lesions with additional neocortical involvement were especially closely linked with postictal psychosis. We have also stressed the close association of postictal psychosis with psychic auras, such as déjà vu and ictal fear (Kanemoto et al., 1996). In the present study as well, we confirmed a statistically significant predominance of déjà vu (Table 9.8). Considering the results of a cortical stimulation study conducted by Gloor et al. (1982), which demonstrated involvement of the lateral as well as medial temporal structure in the genesis of déjà vu, this predominance of déjà vu aura agrees with the finding obtained from our MRI study. Although Savard et al. (1991) suggested that the high incidence of ictal fear in patients with postictal psychic aura was proof of the similarity between postictal and interictal psychoses, the current study contradicts this, demonstrating that such psychic auras were a salient feature of patients with postictal but not interictal psychosis. Furthermore, several authors have pointed out that bilateral interictal epileptiform discharges predisposed patients to develop postictal psychosis (Savard et al., 1991; Umbricht et al., 1995).

126

K. Kanemoto

Table 9.9. Psychopathological findings (Modified SAPS)a

Delusion of perception Delusions of reference Persecutory delusions Verbal hallucinations Visual hallucination Grandiose delusions Religious delusions Pressure of speech Illusion of familiarity Mental diplopia

Postictal psychosis (n⫽45)

Interictal psychosis (n⫽126)

0 3 5 3 9 12 10 22 13 8

37 113 111 82 2 1 3 1 1 1

Notes: a Only features that demonstrated significant difference (P⬍0.005) are listed. SAPS, Scale for the Assessment of Positive Symptoms (Andreason, 1984).

Psychopathological features

The most striking differences between interictal and postictal psychoses lay in the domain of psychopathological phenomenology (Table 9.9). In a series presented by Logsdail and Toone (1988), only one of 14 patients had primary delusions or thought disorders (7%), whereas as many as nine exhibited a markedly abnormal mood (64%). Our previous study, comparing the psychopathological features of postictal psychoses with those of interictal psychoses, supported their data. The first-rank symptoms of Schneider, such as delusions of perception and voice commenting, occurred significantly less often in postictal psychoses than in acute interictal psychoses, whereas sexual indiscretions, religious delusions, and grandiosity, often in the setting of an elevated mood, were observed in postictal psychosis five times more often than acute interictal psychosis. Illusions of familiarity, mental diplopia, and feelings of impending death, which Jackson and Stewart (1899) described as hallmarks of the dreamy state, occurred almost exclusively in postictal psychosis. The frequent occurrence of grandiose delusions and religious delusions, in a setting of markedly elevated moods and feelings of mystic fusion of the body with the universe, characterizes the psychopathology of postictal psychosis. In the present extended series of patients, we were able to amplify these findings. In Table 9.7, psychopathological traits that were proved to show a highly significant difference are listed. Violence and postictal psychosis

Violent behaviour elicited in the course of postictal psychosis deserves a special comment. The argument against the view that epilepsy is closely related to a libera-

127

Postictal psychoses, revisited

tion of aggressive impulses has marked modern epileptology, with the result that epileptologists have almost succeeded in dismissing this old view. However, in the course of our investigation of postictal psychosis, the sporadic episodes of abrupt violent behaviour that we observed impressed us greatly. In a previous study (Kanemoto et al., 1999), we compared violent attacks during episodes of postictal psychosis, acute interictal psychosis, and postictal confusion immediately following complex partial seizures in patients with temporal lobe epilepsy (TLE), and confirmed that severe violent confrontational behaviour towards surrounding people with impending danger occurred only rarely during the postictal confusions as previous studies have also pointed out (Ashford et al., 1980; Rodin, 1973; Treiman, 1991). In contrast, patients proved to be quite prone to violent behaviour during episodes of postictal psychosis. Recently, Gerard et al. (1998) reported six cases who were identified as having subacute postictal aggression, which supports our postulation that stresses a close link between postictal psychosis and violent behaviour. Postoperative de novo manic depressive illness Hill et al. (1957) were one of the first to recognize that depression could occur after a temporal lobectomy. In a series evaluated by Taylor (1972), five patients committed suicide. In another follow-up study, Taylor and Marsh (1977) reported that the mortality rate during the first 2 years postoperatively was twice as high as that in any subsequent 2-year period. Further, in a Danish series investigated by Jensen and Larsen (1979), all suicide attempts occurred within the first postoperative month. A literature search failed to find any descriptions of postoperative hypomanic or manic states, except for our own recent report (Kanemoto et al., 1998). However, we were able to confirm the presence of a substantial number of cases with postoperative transient manic or hypomanic states, and a close relationship between postoperative mood disorder and preoperative history of postictal psychoses. Considering the intrinsic interrelatedness of postictal psychosis with dramatic affective changes (Kanemoto et al., 1996a; Logsdail and Toone, 1988; Savard et al., 1991), this preponderance of postoperative mood changes among patients with preoperative postictal psychosis was all the more instructive. As Trimble (1991) has warned, in view of the high incidence of suicide during the first few years after a temporal lobectomy alone, the need for continuing psychiatric observations of patients who receive an operation, especially when they have a history of postictal psychosis prior to surgery, is apparent. Treatment Treatment for postictal psychosis should be directed at two different stages. First, once an episode of postictal psychosis appears, a direct shortening or alleviation of

128

K. Kanemoto

postictal psychosis should be attempted. While Kanner et al. (1996) recommend dopamine blockers, Lancman et al. (1994) advise the use of benzodiazepines and sedation with chloral hydrate. However, this difference in opinion is more apparent than real. In a typical case, postictal psychosis begins with an initial hypomanic state, which develops rapidly into a psychotic state with marked psychomotor agitation within 12–48 hours. If we succeed in making patients sleep amply during the initial hypomanic state, appearance of frank psychosis could be nipped in the bud. In this way, a certain proportion of the postictal psychosis could be prevented at the stage of the hypomanic state, especially in the seizure monitoring unit, where trained psychiatrists could recognize the initial signs without delay. Indeed, postictal psychosis is selflimited and lasts at most only for several days in a majority of cases. However, some patients tend to commit violent behaviours or even suicide during the episodes. Therefore, we strongly recommend prompt arrest of this state. Second, in contrast to the alternative psychosis of Landolt, efforts to reduce complex partial and generalized tonic-clonic seizures could prevent recurrence of postictal psychosis in a substantial number of patients. Because the direction of the treatment is opposite at times as a function of alternative and postictal psychosis, the differential diagnosis between these states is all the more important. Conclusion In conclusion, a concept of postictal psychosis is certainly contributing to the elucidation of several longstanding neurobehavioural problems in patients with epilepsy that have caused much controversy. Considering the conspicuous affective nature of postictal psychosis as well as the preponderance of postoperative mood disorders in patients with a preoperative history of postictal psychosis, further investigations of this state could reveal underlying neurochemical changes which could serve as a model for bipolar mood disorders. However, there remains much controversy on the pathophysiological mechanism underlying postictal psychosis (Wieser et al., 1985; Wolf, 1991). Although So et al. (1990) confirmed increased spiking activity in the medial temporal region with depth-EEG recording during an episode of postictal psychosis, they were sceptical about the causal relationship between this finding and occurrence of frank psychosis, since increased spiking is commonly observed as a postictal phenomenon in patients without psychosis. In a similar study using depth EEG, Mathern et al. (1995) failed to confirm increased postictal spike activity. Possibly, postictal psychosis should be further subdivided. While some cases, in which peculiar feelings such as mental diplopia or déjà vu are prominent with the underlying tone of dysphoria, might be attributed to subclinical limbic status epilepticus such that Wieser et al. (1985) confirmed with a depth-

129

Postictal psychoses, revisited

EEG study, other cases in the setting of markedly elevated moods would not be directly related to seizure activity but caused by some alterations in pathways of neurotransmitters. In this regard, the serotoninergic mechanism should also be considered in view of the marked affective nature of the postictal psychosis along with dopaminergic hypersensitivity (Kanner et al., 1996; Logsdail and Toone, 1988; Savard et al., 1991; So et al., 1990) as well as a GABA-mediated mechanism (Ring et al., 1994; Szabo et al., 1996). Certainly, postictal psychosis deserves further attention.

R E F E R E N C ES Andreasen, N.C. (1984). Scale for the Assessment of Positive Symptoms (SAPS). Iowa: The University of Iowa. Ashford, J.W., Schulz, S.C. and Walsh, G.O. (1980). Violent automatism in a partial complex seizure. Arch Neurol, 37, 120–2. Bruens, J.H. (1971). Psychoses in epilepsy. Psychiatr Neurol Neurochir, 74, 174–92. Commission on Classification and Terminology of the International League Against Epilepsy (1981). Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia, 11, 102–13. Commission on Classification and Terminology of the International League Against Epilepsy (1989). Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia, 30, 389–99. Devinsky, O., Abramson, H., Alper, K. et al. (1995). Postictal psychosis: a case control series of 20 patients and 150 controls. Epilepsy Res, 20, 247–53. Edeh, J. and Toone, B. (1987). Relationship between interictal psychopathology and the type of epilepsy. Br J Psychiatry, 151, 95–101. Farlet, J. (1860/1961). De l’état des épileptiques. Arch Gén Méd, 16, 661–99; 17, 461–91; 18, 423–43. Gerard, M.E., Spitz, M.C., Towbin, J.A. and Shantz, D. (1998). Subacute postictal aggression. Neurology, 50, 384–8. Gibbs, F.A. (1951). Ictal and non-ictal psychiatric disorders in temporal lobe epilepsy. J Nerv Ment Dis, 113, 522–8. Gloor, P., Olivier, A. and Quensney, L.F. (1982). The role of the limbic system in experiential phenomena of temporal lobe epilepsy. Ann Neurol, 12, 129–44. Hill, D., Pond, D.W., Mitchell, W. and Falconer, M.A. (1957). Personality changes following temporal lobectomy for epilepsy. J Ment Sci, 103, 18–27. Jackson, J.H. (1875). On temporary mental disorders after epileptic paroxysms. West Riding Lunatic Asylum Med Rep, 5, 105–29. Jackson, J.H. and Stewart, P. (1899). Epileptic attack with a warning of a crude sensation of smell and intellectual aura (dreamy state) in a patient who has symptoms pointing to gross organic disease of right temporo-sphenoidal lobe. Brain, 22, 534–49.

130

K. Kanemoto Jensen, I. and Larsen, J.K. (1979). Mental aspects of temporal lobe epilepsy – follow-up of 74 patients after resection of a temporal lobe. J Neurol Neurosurg Psychiatry, 42, 256–65. Kanemoto, K., Kawasaki, J. and Kawai, I. (1996a). Postictal psychoses: a comparison with acute interictal and chronic psychoses. Epilepsia, 37, 551–6. Kanemoto, K., Takeuchi, J., Kawasaki, J. and Kawai, I. (1996b). Characteristics of temporal lobe epilepsy with mesial temporal sclerosis, with special reference to psychotic episodes. Neurology, 47, 1199–203. Kanemoto, K., Kawasaki, J. and Mori, E. (1998). Postictal psychosis as a risk factor for mood disorders after temporal lobe surgery. J Neurol Neurosurg Psychiatry, 65, 587–9. Kanemoto, K., Kawasaki, J. and Mori, E. (1999). Violence and epilepsy: a close relation between violence and postictal psychosis. Epilepsia, 40, 107–9. Kanner, A.M., Stagno, S., Kotagal, P. and Morris, H.H. (1996). Postictal psychiatric events during prolonged video-electroencephalographic monitoring studies. Arch Neurol, 53, 258–63. Kolb, L.C. and Brodie, H.K.H. (1982). Modern Clinical Psychiatry. Philadelphia: W.B. Saunders. Lancman, M.E. (1999). Psychosis and peri-ictal confusional states. Neurology, 53 (Suppl. 2), S33–8 Lancman, M.E., Craven, W.J., Asconapé, J.J. and Penry, J.K. (1994). Clinical management of recurrent postictal psychosis. J Epilepsy, 7, 47–51. Landolt, H. (1953). Einige klinisch-elektroencephalographische Korrelationen bei epileptischen Dämmerzuständen. Nervenarzt, 24, 479. Landolt, H. (1963). Die dammer- und verstimmungszustande bei epilepsie und ihre Elektroencephalographie. Dtsch Z Nervenheilkunde, 185, 411–30. Levin, S. (1952). Epileptic clouded states. A review of 52 cases. J Nerv Ment Dis, 116, 215–25. Logsdail, S.J. and Toone, B.K. (1988). Postictal psychoses. A clinical and phenomenological description. Br J Psychiatry, 152, 246–52. Mathern, G.W., Pretorius, J.K., Babb, T.L. and Quinn, B. (1995). Unilateral hippocampal mossy fiber sprouting and bilateral asymmetric neuron loss with episodic postictal psychosis. J Neurosurg, 82, 228–33. Mendez, M.F., Grau, R., Doss, R.C. and Taylor, J.L. (1993). Schizophrenia in epilepsy: seizure and psychosis variables. Neurology, 43, 1073–7. Onuma, T., Adachi, N., Ishida, S., Katou, M. and Uesugi, S. (1995). Prevalence and annual incidence of psychoses in patients with epilepsy. Epilepsia, 36 (Suppl. 3), S218 Perez, M.M. and Trimble, M.R. (1980). Epileptic psychosis-diagnostic comparison with process schizophrenia. Br J Psychiatry, 141, 256–61. Ring, H.A., Trimble, M.R., Costa, D.C., Moriarty, J., Verhoeff, N.P. and Ell, P.J. (1994). Striatal dopamine receptor binding in epileptic psychosis. Biol Psychiatry, 35, 375–80. Rodin, E.A. (1973). Psychomotor epilepsy and aggressive behavior. Arch Gen Psychiatry, 28, 210–13. Sachdev, P. (1998). Schizophrenia-like psychosis and epilepsy: the status of the association. Am J Psychiatry, 155, 325–36. Savard, G., Andermann, F., Olivier, A. and Remillard, G.M. (1991). Postictal psychosis after partial complex seizures: a multiple case study. Epilepsia, 32, 225–31.

131

Postictal psychoses, revisited So, N.K., Savard, G., Andermann, A., Olivier, A. and Quensney, L.F. (1990). Acute postictal psychosis: a stereo EEG study. Epilepsia, 31, 188–93. Slater, E. and Beard, A.W. (1963). The schizophrenia-like psychoses of epilepsy. Br J Psychiatry, 109, 95–150. Szabo, C.A., Lancman, M.L. and Stagno, S. (1996). Postictal psychosis: a review. Neuropsychiatry, Neuropsychol Behav Neurol, 4, 258–64. Taylor, D.C. (1972). Mental state and temporal lobe epilepsy. A correlative account of 100 patients treated surgically. Epilepsia, 13, 727–65. Taylor, D.C and Marsh, S.M. (1977). Implication of long-term-follow-up in epilepsy: with a note on the cause of death. In Epilepsy: The 8th International Symposium, ed. J.K. Penry, pp. 27–34. New York: Raven Press. Toone, B.K., Garralda, M.E. and Ron, M.A. (1980). The psychoses of epilepsy and the functional psychoses. Br J Psychiatry, 137, 245–9. Treiman, D.M. (1991). Psychobiology of ictal aggression. In Advances in Neurology, Vol. 55, ed. D. Smith, D. Treiman and M.R. Trimble, pp. 341–56. New York: Raven Press. Trimble, M.R. (1991). The Psychoses of Epilepsy. New York: Raven Press. Umbricht, D., Degreef, G., Barr, W.B., Lieberman, J.A., Pollack, S. and Schaul, N. (1995). Postictal and chronic psychoses in patients with temporal lobe epilepsy. Am J Psychiatry, 152, 224–31. Wieser, H.G., Hailemariam, S. and Regard, M. (1985). Unilateral limbic epileptic status activity: stereo-EEG, behavioral, and cognitive data. Epilepsia, 26, 19–29. Wolf, P. (1991). Acute behavioral symptomatology at disappearance of epileptiform EEG abnormality: paradoxical or forced normalization. In Neurobehavioral Problems in Epilepsy, ed. D. Smith, D. Treiman and M.R. Trimble, pp. 127–42. New York: Raven Press. World Health Organization (1992). The International Classification of Mental and Behavioural Disorders: Clinical and Diagnostic Guidelines (Tenth Revision) (ICD–10). Geneva: WHO.

Part III

Cognitive aspects

10

Dementia and epilepsy Stephen W. Brown Plymouth Postgraduate Medical School, Plymouth, UK

Definitions and context As Lishman (1998) has pointed out, the term ‘dementia’ has two potential meanings in medical practice: first, it may refer to a group of specific diseases, and second, it is used to describe a clinical syndrome that can have many causes. Although diseases in the former group are characterized by irreversible decline in function, the latter includes conditions in which decline can be arrested, or in some cases reversed. Both ICD–10 (World Health Organization, 1992) and DSM–IV (American Psychiatric Association, 1994) offer detailed diagnostic criteria for the syndrome, with DSM–IV defining additional principles for diagnosing different varieties of dementia. Both of these, in attempting to ensure uniformity of populations that might be used for research purposes, adopt a ‘cookbook’ style of approach. In this chapter, the term is used in the second of the meanings described by Lishman. ‘Dementia’ is regarded as an acquired global impairment of intellect, memory and personality, which is independent of any impairment of consciousness. The symptoms of dementia are typically of long duration, usually progressive and often irreversible, but none of these latter features are essential to the concept. How the concept of epilepsy came to be linked historically with that of dementia, how these concepts became uncoupled, and how new connections came to be made between them, are components of a story that contains much of the history of psychiatry and of neurology. Historical aspects

Aretaeus, in the second century AD (quoted by Temkin, 1971) described people in whom epilepsy had become chronic as being ‘languid, spiritless, stupid, inhuman, unsociable, . . . not disposed to hold intercourse’ and ‘slow to learn, from torpidity of the understanding and of the senses’. There was therefore a view that a continuing propensity to seizures led to deterioration of those faculties that we would probably refer to as intellect and personality. According to Berrios (1995) the term 135

136

S.W. Brown

‘dementia’ acquired a medical connotation in the second half of the eighteenth century. He quotes the French encyclopaedia of 1765, which lists as causes: 1. damage to the brain caused by excessive usage, congenital causes or old age, 2. failure of the spirit, 3. small volume of the brain, 4. violent blows to the head causing brain damage, 5. incurable diseases such as epilepsy, or exposure to venoms . . . Dementia is difficult to cure as it is related to damage of brain fibres and nervous fluids; it becomes incurable in cases of congenital defect or old age . . . [otherwise] treatment must follow the cause. (Diderot and d’Alambert 1765)

A later concept, which was to carry much influence, was the theory of degeneracy, especially as that propounded by Morel (1857). He suggested that although mental disorders might have an external or environmental cause in the first instance, the person’s biological state is then modified, so that the disorder becomes hereditary. Following this, each generation displays an increasing degree of pathology until the line becomes extinguished in idiocy. Although degeneration was exemplified by the deterioration across generations, it could also take place within the person’s lifetime. Morel expressed a particular interest in epilepsy, and was responsible for suggesting that there could exist a masked form (epilepsie larvée), in which the main features were not seizures but insanity. Thus a link was formalized between epilepsy and deterioration. By the beginning of the twentieth century many European medical writers took this as received fact. Thus: The mental state of epileptics, as is well known, frequently presents deterioration, . . . mischievous restlessness and irritability in childhood may develop to vicious and even criminal tendencies in adult life. Every grade of intellectual defect may be met with, down to actual imbecility. (Gowers, 1885)

And The most numerous class of epileptics show, after the lapse of years, a slowly progressive dimming of the active perceptions of the mind, a loss of memory, a blunting of the affections, a permanent mental obtuseness which increases and grows, until, if the patient lives long enough, there is a more or less absolute annihilation of all the faculties. (Berkley, 1901)

Most authorities seemed to believe that decline was consequent upon seizures. While Gowers acknowledged that in a minority of cases deterioration is ‘the expression of a cerebral imperfection of which the epilepsy is another manifestation’, he also stated that in the majority of cases ‘the failure must be regarded as a consequence of the disease’. Henry Maudsley also considered that seizures had a direct effect: The mind is slowly weakened by the storms of fury through which it passes, and they sink finally into the apathy of dementia – a state of mere oblivion, in which they cease to hope or care more. (Maudsley, 1874)

However, during the twentieth century the view that epilepsy alone might cause intellectual deterioration has by and large been discredited, although replaced by

137

Dementia and epilepsy

observations regarding the effects of concomitant brain disease, recurrent head injury due to seizures, and undesirable effects of antiepileptic treatment. It is easy to dismiss nineteenth century writings as the consequence of prejudice-based, evidence-free speculation, but the question really has to be considered as to whether the patient population was different to that which we see today. Perhaps aspects of the natural history of seizure disorders have changed with time. A common cause for admission to an asylum in the late nineteenth century was general paralysis of the insane. Neurosyphilis was very common then, but is now rare. Seizures and dementia can both occur in this condition, which we now know to be a late complication of syphilis, but the aetiology was not fully recognized until the twentieth century. Indeed, speculation was rife about the possibility of psychosocial causation (in some ways analogous to the debate in the 1960s and 1970s concerning the cause of schizophrenia). Objectifying the cognitive consequences of seizure disorders was a task that had to wait until the advent of psychometrics. Origins of psychometry

Visitors and residents in nineteenth century London could attend an ‘Anthropometric Laboratory’ in South Kensington, and pay to have exact measurements made of their height, weight, breathing power, strength of pull and squeeze, colour sense and much else. This was the brainchild of Sir Francis Galton (1822–1911), a gentleman-scholar, explorer, meteorologist and cousin of Charles Darwin. Galton was interested in explaining the differences between individuals. He held views about the distribution of human abilities, and became convinced that heredity was of prime importance in defining them. He was mentor to Karl Pearson (1857–1936), whose name is well known to scientists for his significant contribution to the study of statistics, and who occupied the post of Galton Professor of Eugenics at University College London after Galton’s death. Both Galton’s and Pearson’s views about the normal distribution of human abilities were influential in the subsequent design of intelligence quotient (IQ) tests. When Binet, Wechsler and others were to develop standardized tests purporting to measure intelligence, these tests were adjusted in order to provide a bell-shaped curve of distribution of results, and tests tended to be constructed to demonstrate previous underlying assumptions rather than to discover new ones. When the first IQ tests were applied to people with epilepsy, there was sometimes evidence of intellectual deterioration on retesting (Dawson and Conn, 1929) although this was found to be part of a wider pattern of fluctuation in test–retest performance in epilepsy that could occur in either direction (Fetterman and Barnes, 1934; Patterson and Fonner, 1928; Sullivan and Gahagan, 1935). One study found that improvement in seizure control was related to improvement in test performance (Kugelmass et al., 1938) raising the possibility that observed fluctuations

138

S.W. Brown

in IQ could be related to seizure frequency. Some studies did suggest a slight decline in mean IQ with increasing duration of seizure disorder, but no very clear picture emerged (Brown and Reynolds, 1981). Such observations, although superficially interesting, were of limited value in developing a sophisticated model of the relationship between seizure disorders and cognitive function. This was to require developments from an initially parallel, but later converging discipline, that of neuropsychology. Origins of neuropsychology

The origins of neuropsychology lie in direct clinical observation. Two nineteenth century landmarks in the understanding of structural and functional relationships in the brain were the observations of Broca and Wernicke. In 1861 Paul Broca identified the third frontal convolution of the left hemisphere as an area that if damaged, would result in a specific impairment of expressive language. In 1874 Karl Wernicke described specific impairment of receptive language associated with damage to an area in the left hemisphere extending from the first temporal convolution into the parietal lobe. The recognition of clinically definable psychological syndromes related to discrete brain pathology continued to develop along with a greater understanding of cerebral localization. Modern neuropsychological testing involves specific observation of memory, language, verbal and nonverbal fluency together with visuo-spatial and motor abilities. These assessments complement tests of general cognitive functioning. Neuropsychological assessment is of course used as part of the work-up for epilepsy surgery, but has been an essential tool in studying the interaction between cognitive impairment and seizure-related variables (Goldstein 1997; Kälviainen et al., 1992; Piccirilli et al., 1994; Rugland, 1990). An issue in interpretation

In the process of scoring IQ tests in children, the raw test scores are subjected to adjustment, which takes into account the age of the subject. This then gives a result that is a comparison against the general population of the same age. As children grow older, the mean raw scores for various items will rise as a result of learning and normal development, but the adjustment will ensure that the mean population IQ remains the same, and that IQ scores of the population of the same age fit the expected bell-shaped curve. This could lead to a situation where an individual child improves test performance over a period of time and therefore improves raw test scores, but does not improve as much as the rest of the population overall, and therefore the IQ is seen to fall. This is of course an example of slow learning compared to the rest of the population, but it does not represent deterioration. The interpretation of serial IQ scores in children therefore requires attention to the raw score changes. Clinical and educational psychologists will routinely take this into

139

Dementia and epilepsy

account as part of their everyday practice, but the matter is perhaps worth drawing to the attention of other clinicians who may not be so familiar with the details of test construction, scoring and interpretation. Possible relationships As Gowers implied, dementia and epilepsy may both be consequences of the same underlying disorder, rather than one being a consequence of the other. Some specific clinical epilepsy syndromes are associated with acquired disorders of intellect and memory, although it is not clear whether seizure activity is responsible for the cognitive changes or whether, like the previous group, there is a shared aetiology. It is however known that individual seizures may result in a cognitive penalty, and that interictal epileptiform EEG discharges can sometimes disrupt cognitive functioning. Some antiepileptic drugs can also play a part. These potential relationships are considered in turn. Neurological disease in which both dementia and seizures occur

It is well recognized that symptomatic seizures occur in the context of dementia syndromes in older people (Forsgren et al., 1996). Breteler et al. (1995) found that people aged 50–75 with a diagnosis of epilepsy had an overall relative risk of 1.5 of subsequently developing a dementia, which they described as a moderately increased risk over the expected rate. People with Down’s syndrome are at particular risk of developing early Alzheimer’s disease, and this is frequently associated with seizures (Collacott, 1993; Lott and Lai, 1982). In people with mental retardation (learning disabilities) who do not have Down’s syndrome, the incidence and prevalence of dementia is higher than expected, and is often associated with poorly controlled epilepsy (Cooper, 1997). There are also isolated case reports showing an apparent association between seizures, dementia and a pathological lesion. It may well be that there are many such very rare or even one-off disorders (Yerby et al., 1986). One example of a rare condition in which the relationship between cerebral lesions, seizures and intellectual deterioration is clear is Sneddon syndrome. This is of presumed autoimmune origin and is characterized by cerebral infarction, livedo reticularis, hypertension, epilepsy and progressive dementia (O’Riordan et at., 1995; Stephens, 1992). One specific pathology that is associated with epilepsy and cognitive deterioration is the hypothalamic hamartoma (Kuzniecky et al., 1997). This can give rise to gelastic epilepsy in which brief, frequent and mechanical laughing seizures begin in infancy against a background of otherwise normal development. Other seizure types, together with cognitive deterioration, start to appear between the ages of 4 and 10 years (Berkovic et al., 1988).

140

S.W. Brown

Cognitive deterioration and worsening epilepsy also co-exist in Rasmussen syndrome (Rasmussen et al., 1958), a devastating progressive focal encephalitis that has hitherto eluded aetiological explanation. Hart and her colleagues have recently described some comorbidity with other cerebral pathologies (Hart et al., 1998). Another hopeful lead has been the discovery by Rogers et al. (1994) of circulating antibodies to the glutamate GLUR3 receptor in a patient with this condition who showed some response to removal of the same antibodies by recurrent plasma exchange. There are also a number of specific genetic syndromes in which seizures and dementia occur. Many show a Mendelian pattern, such as the autosomal dominant hereditary prion disorders (Prusiner, 1994) or the progressive myoclonus of Unverricht and Lundberg (Lehesjoki et al., 1992), which has a recessive mode of inheritance. The number of these syndromes continues to expand, as does our understanding of the pathological processes involved. Much more information will become available as the map of the human genome continues to be elaborated. Some specific epilepsy syndromes

Loss of previously acquired skills or abilities is a known adverse association with certain epilepsy syndromes, most notably West syndrome, Lennox–Gastaut syndrome (LGS), the syndrome of continuous spike-wave discharges in slow wave sleep (CSWS) and the Landau–Kleffner syndrome (LKS). At first sight, these conditions all share the feature of frequent epileptiform EEG discharges, and it would be tempting to suggest that such activity, by disrupting learning, interferes with cognitive development. Although this may be partly true, it has been shown that diffuse electrical status can occur without apparent cognitive consequence (Gokyigit and Caliskan, 1995). Besides, some authorities attribute at least part of the deterioration in LGS to treatment effects, both pharmacological and psychosocial (Genton and Dravet, 1997). LGS is also characterized in many people by frequently disabling drop attacks (Oguni et al., 1996) which may have cumulative adverse effects on cognitive function. Goldsmith et al. (2000) found that individuals with LGS with a normal cognitive outcome tended to have onset of LGS at a later age than those who deteriorated. EEG paroxysmal activity occurring during critical periods in development may account for some of the specific cognitive and behavioural changes seen in LKS and CSWS. Synaptic contacts that should have degenerated by apoptosis as part of normal development may instead be strengthened. If this occurs in the temporoparietal cortex, this may lead to the acquired aphasia of LKS. If this is predominantly frontal, higher cognitive and executive functions will be affected first, leading to the disintegrative psychotic-like presentation of CSWS (Roulet Perez et al., 1993; Smith, 1997). This is consistent with the observation that the earlier the onset, the

141

Dementia and epilepsy

worse the symptoms. It does not, however, explain the cause of the paroxysmal discharges. Cognitive effects of seizures

Seizures can cause disruption in cognitive functioning lasting for days after otherwise apparent recovery (Dodrill, 1986), and a similar phenomenon lasting for several hours has been described following other seizure types (Aldenkamp et al., 1992). Related to this is the observation that seizures occurring during a night’s sleep may affect learning ability the next day (Aldenkamp, 1995). Single complex partial or secondarily generalized seizures may be associated with neuronal damage (Rabinowicz et al., 1996) and in humans, brain extracellular glutamate builds up to potentially neurotoxic levels in partial seizures (During and Spencer, 1993). Some other seizure-related variables that have been described as adversely affecting cognitive function over time include: the presence of tonicclonic or complex partial seizures (Dodrill, 1986; Seidenberg et al., 1981), early onset of seizures (O’Leary et al., 1983, Strauss et al., 1995), the total number of seizures experienced (Dodrill, 1986; Rodin et al., 1986), and repeated nonconvulsive status epilepticus (Aldenkamp, 1997). It has recently been reported that repetitive head injury in young adults is associated with neocortical neurofibrillary tangles (Geddes et al., 1999). Whether this may be of clinical significance in people with epilepsy who suffer repeated head trauma as a consequence of their seizures remains to be seen. Critical periods in development may render cerebral functioning more susceptible to interference (Holmes, 1997; Johnston, 1996). Not surprisingly, improvement in seizure frequency is related to improvement in cognitive functioning (Seidenberg et al., 1981). Jambaque et al. (2000) found that children with infantile spasms due to tuberous sclerosis whose seizures became well controlled with vigabatrin showed significant improvement in cognition and behaviour, in contrast to the usual poor prognosis of this condition if seizures are not well controlled. Cognitive effects of EEG discharges

Subclinical epileptiform EEG discharges can affect cognitive processes both generally and specifically (Aarts et al., 1984). However, this may not be inevitable, and at least one case has been reported in which diffuse electrical status epilepticus occurred without demonstrable learning problems (Gokyigit and Caliskan, 1995). Nevertheless, other studies have suggested that EEG epileptiform discharges occurring during IQ testing can affect test results (Siebelink et al., 1988) and such discharges may also have a discreet effect on school performance (Kasteleijn-Nolst Trenite et al., 1988). This phenomenon has been called transitory cognitive impairment (TCI) and has been extensively studied. TCI can be demonstrated in about

142

S.W. Brown

half the cases in which testing is accompanied by discharges (Binnie, 1993). Although the presence of generalized bursts of 3-Hz spike-wave lasting three seconds or more is most likely to cause TCI, it can be demonstrated in some cases with discharges of shorter duration. Focal spikes can also produce TCI, those on the left being mainly associated with verbal errors and those on the right with nonverbal task impairment (Binnie and Marston, 1992). Bailet and Turk (2000) found that children with epilepsy and abnormal interictal EEGs scored lower than children with epilepsy with normal interictal EEGs on reading and spelling measures. They concluded that overall long-term risks of learning problems exist among children with epilepsy compared with controls. This is true even for those children with normal IQs and whose seizures are well controlled. This observation lends some support to the view previously expressed by other authors that in some cases TCI can result in deterioration in psychological and social functioning. It has been suggested that specific antiepileptic drug therapy aimed at reducing the interictal EEG discharges, and not just clinical seizures, may have a beneficial effect (Besag, 1995; Marston et al., 1993). The issue of state-dependent cognitive deterioration is discussed in more detail by Besag, Chapter 6. Treatment effects

The effects of antiepileptic treatment on cognitive function remain ill-understood despite much research activity, and the subject remains to some extent controversial. A well-reasoned account of how this topic should be tackled has recently been given by Aldenkamp and van Bronswijk (1999). In a previous review by Vermeulen and Aldenkamp (1995) the authors concluded that the assembled evidence so far ‘is hardly a basis for definitive statements’. The following is a synopsis of published reports of varying scientific rigour, which nevertheless probably represents the current consensus. It serves to complement the more extensive review of Aldenkamp, Chapter 17. Phenobarbitone can significantly impair learning ability (Calandre et al., 1990), and impairs attention and memory at even low therapeutic doses (Riva and Devoti, 1996). A dementia-like picture sometimes seen with phenytoin has been long recognized (Rosen, 1968); this may be related to the serum level of the drug (Matthews and Harley, 1975) or to phenytoin-induced folate deficiency (Neubauer, 1970). In the latter case, folate supplementation can result in improved cognitive performance (Froscher et al., 1995). Phenytoin has also been implicated in thiamine deficiency, resulting in discrete performance deficits in visuo-spatial analysis, visuo-motor speed and verbal abstracting ability. These improve with thiamine supplementation (Botez et al., 1993). Phenytoin has also been reported as causing a specific memory impairment (Butlin et al., 1984; Gillham et al., 1990), most

143

Dementia and epilepsy

marked with visual memory (Pulliainen and Jokelainen, 1994), and it can also have adverse effects on motor and mental speed (Aldenkamp et al., 1994). Phenytoin also diminishes the practice effect expected to be seen on retesting in some cases (Sabers et al., 1995). Some authors have reported carbamazepine to affect adversely memory (Forsythe et al., 1991) and psychomotor speed (Gillham et al., 1990), while others report both carbamazepine and valproate as having negligible effects (Prevey et al., 1996; Stores et al., 1992). Recently, reversible pseudodementia with cortical pseudoatrophy induced by valproate has been described (Guerrini et al., 1998; Papazian et al., 1995; Straussberg et al., 1998). There are relatively few published studies in this area that relate to the newer antiepileptic drugs. The evidence so far, such as it is, suggests that with standardized testing procedures, vigabatrin, lamotrigine and tiagabine have no deleterious effect on cognitive functioning, and in some cases may enhance performance, presumably by preventing adverse consequences of seizures (Banks and Beran, 1991; Dodrill et al., 1997; Kälviainen et al., 1996; Meador and Baker, 1997; Monaco, 1996; Provinciali et al., 1996; Sveinbjornsdottir et al., 1994). Psychosocial effects

It has been observed in a number of studies that children with epilepsy as a group underachieve academically compared to that expected from objective measures of their cognitive ability (Rutter et al., 1970). There are many possible reasons for this, which will include loss of time at school due to illness and hospital appointments, and denial of opportunity to take part in normal activities. It has also been shown (Long and Moore, 1979) that significant people in the lives of children with epilepsy (such as their parents) have lower expectations of the children’s achievements than they do of children without epilepsy. This may also have an effect on cognitive outcome. Some observations in temporal lobe epilepsy The author has for some years reported observations on some people with temporal lobe epilepsy who show evidence of acquired cognitive problems in the absence of the factors listed above (Brown, 1989, 1992, 1999; Brown and Abeyasinghe, 1984; Brown and Vaughan, 1988, 1990). It seems that such problems may occur more often in males than females, where there is a long history of poorly controlled epilepsy of temporal lobe origin. The neuropsychological profile in females is that of frontal and left temporal impairment; males may show impairment of more cortical areas, especially the parietal lobes. In all cases the epilepsy syndrome is temporal lobe epilepsy, and the EEG epileptic foci are temporal, not frontal. A full account is given in Brown (1999).

144

S.W. Brown

A syndrome is described seen in people who develop temporal lobe epilepsy before puberty, and who may come to the attention of psychologists or other behavioural professionals after puberty because of behaviour problems. The difficult behaviour does not typically persist into adult life. Intellectual decline seems to date from the late teenage years and is followed in most (but not all cases) by relative stability into adult life. It is not related to repeated minor head injuries caused by seizures, although some people (mainly male) may have some dysarthria and ataxia. We identified two broad syndromes. A frontal subcortical picture was seen, mainly in males with a prepubertal age of onset of epilepsy, where there is some disturbance around or just after the time of puberty in which a stepwise decline of intellectual functioning occurs. In addition, a cortical frontotemporal picture was also seen, more commonly in females. These sometimes had a later age of onset of epilepsy, and were more likely to acquire a psychiatric diagnosis. Other people with localization-related epilepsy and discrete neuropsychological deficits are also seen with different histories and who show no intellectual deterioration. Previous reports of metabolic changes in temporal lobe epilepsy have suggested that frontal lobe functioning can also be affected. Ingvar (1984) comments: . . . in temporal lobe epilepsy, the pathways from the deep temporal structures to the frontal lobes do not function normally. The temporal ‘gate’ to the frontal lobes may be closed. This may explain symptoms of psychopathology in this common type of focal epilepsy.

More recently, Mackenzie and Miller (1994) examined senile plaques in temporal lobectomy specimens and a control group. They found the age-related incidence of senile plaques was significantly greater in the temporal lobe epilepsy group, and concluded that this suggested some aspects of TLE have a positive influence on the formation of senile plaques. The exact significance of this finding is unclear. Temporal Gate hypothesis

We have proposed a possible mechanism to account for the above observations, which in deference to Ingvar (1984) we have called the Temporal Gate hypothesis (Brown, 1999). This may be summarized as follows: 1. Certain neuropsychological functions of the frontal lobes come to maturity during and after puberty, the process being influenced by the neuroendocrine axis. 2. This maturation of functioning is also dependent on frontal lobe input from an intact temporo-limbic system. 3. If seizures of temporal lobe origin occur before and during puberty, the temporolimbic input to the frontal lobes during this critical period of development may be disrupted. Such disruption might lead to a period of behavioural or adjust-

145

Dementia and epilepsy

ment problems during and after puberty. Also, because frontal lobe function does not develop in the normal way, this might lead to the observed neuropsychological effects. These effects may reach a plateau, while cases may exhibit further decline. 4. In our studies, the neuropsychological disruption so caused is sufficient to impair some previously acquired skills, such as reading. Some indirect, or collateral, support for the hypothesis comes from a study recently reported by Metz-Lutz et al. (1999) who investigated EEG and neuropsychological outcomes in children with an initial diagnosis of benign partial epilepsy. Their findings suggested to them that maturing cognitive functions subserved by a cortical area distant from the epileptic focus are nevertheless susceptible to interference by epileptic activity, and this may affect cognitive outcomes. In other neurological conditions associated with deterioration, the phenomenon of oxidative stress (the production of oxygen radicals beyond a threshold for proper antioxidant neutralization) has been implicated. These include Alzheimer’s disease (Sims, 1996), Parkinson’s disease (Jenner and Olanow, 1994), amyotrophic lateral sclerosis (Gorman et al., 1994), Pick’s disease (Castellani et al., 1995) and schizophrenia (Ramchand et al., 1996). Various intracellular messenger systems involving glutamate are implicated in oxidative radical production. These systems are involved in neuronal growth, differentiation and apoptosis (Michaelis, 1998). Glutamate is also known to play an important role in epilepsy. The author has observed substantial improvement in cognitive functioning in two patients (one male ‘subcortical’, one female ‘temporal’) after using lamotrigine, a glutamate release inhibitor (Meldrum, 1994). There are various other reports, generally of an anecdotal nature, of the positive effects of lamotrigine on cognitive functioning and scholastic ability (Buoni et al., 1998; Meador and Baker, 1997). One could speculate that oxidative stress plays a part in the maintenance of the acquired frontal syndrome deficits described above. Conclusions True intellectual deterioration is now known to be a relatively uncommon association of epilepsy. Where it occurs it is usually associated with specific causes. These include concomitant degenerative brain disease, some specific epilepsy syndromes of uncertain aetiology, the direct effects of seizures (especially tonic-clonic seizures), transitory cognitive impairment due to interictal EEG epileptiform discharges, and the consequences of some antiepileptic treatments. In some cases it is possible that a seizure disorder with temporal lobe focus, if present during a critical developmental period, may adversely influence maturation and maintenance of normal frontal lobe function, giving rise to a clinical picture of cognitive

146

S.W. Brown

impairment or decline. The potential role of oxidative stress in this process deserves further study. Jokeit and Ebner (1999), as explored in Chapter 11, have described ‘slow, ongoing deterioration’ in some people with temporal lobe epilepsy, and state (with an echo of Henry Maudsley’s words from 125 years previously, about the mind being weakened by the storms of fury through which it passes) ‘It is assumed that epilepsy-related noxious events and agents exhaust the compensatory capacity of brain functions’. Thus has our understanding come full circle, with still much left to provide a good seam for investigators to mine.

R E F E R E N C ES Aarts, J.H., Binnie, C.D., Smit, A.M. and Wilkins, A.J. (1984). Selective cognitive impairment during focal and generalized epileptiform EEG activity. Brain, 107, 293–308. Aldenkamp, A.P. (1995). The impact of epilepsy on cognitive development and learning behaviour. In Epilepsy in Children and Adolescents, ed. A.P. Aldenkamp, W.O. Renier, F.E. Dreifuss and T.P.B.M. Suurmeijer, pp. 225–38. Boca Raton, FL: CRC Press. Aldenkamp, A.P. (1997). Effect of seizures and epileptiform discharges on cognitive function. Epilepsia, 38 (Suppl. 1), S52–5. Aldenkamp, A.P. and van Bronswijk (1999). Cognitive side-effects as outcome measures in antiepileptic drug treatment: the current debate. In Epilepsy and Mental Retardation, ed. M. Sillanpaa et al., pp. 135–146. Petersfield: Wrightson Biomedical Publishing. Aldenkamp, A.P., Gutter, T. and Beun, A.M. (1992). The effect of seizure activity and paroxysmal electroencephalographic discharge on cognition. Acta Neurol Scand, 86, 111–22. Aldenkamp, A.P., Alpherts, W.C., Diepman, L., van’t Slot, B., Overweg, J. and Vermeulin, J. (1994). Cognitive side-effects of phenytoin compared with carbamazepine in patients with localization-related epilepsy. Epilepsy Res, 19, 37–43. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) (DSM–IV). Washington, DC: APA. Bailet, L.L. and Turk, W.R. (2000). The impact of childhood epilepsy on neurocognitive and behavioral performance: a prospective longitudinal study. Epilepsia, 41, 426–31. Banks, G.K. and Beran, R.G. (1991). Neuropsychological assessment in lamotrigine treated epileptic patients. Clin and Exp Neurol, 28, 230–7. Berkley, H.J. (1901). A Treatise on Mental Disease. London: Henry Kimpton. Berkovic, S.F., Andermann, F., Melanson, D., Ethier, R.E., Feindel, W. and Gloor, P. (1988). Hypothalamic hamartomas and ictal laughter: evolution of a characteristic epileptic syndrome and diagnostic value of magnetic resonance imaging. Ann Neurol, 23, 429–39. Berrios, G.E. (1995). Dementia. In A History of Clinical Psychiatry: The Origin and History of Psychiatric Disorders, ed. G.E. Berrios and R. Porter, pp. 34–51. London: The Athlone Press. Besag, F.M. (1995). The therapeutic dilemma: treating subtle seizures or indulging in electroencephalogram cosmetics? Semin Pediatr Neurol, 2, 261–8.

147

Dementia and epilepsy Binnie, C.D. (1993). Significance and management of transitory cognitive impairment due to subclinical EEG discharges in children. Brain Develop, 15, 23–30. Binnie, C.D. and Marston, D. (1992). Cognitive correlates of interictal discharges. Epilepsia, 33 (Suppl. 6), S11–17. Botez, M.I., Botez, T., Ross-Chouinard, A. and Lalonde, R. (1993). Thiamine and folate treatment of chronic epileptic patients: a controlled study with the Wechsler IQ scale. Epilepsy Res, 16, 157–63. Breteler, M.M., de Groot, R.R., van Romunde, L.K. and Hofman, A. (1995). Risk of dementia in patients with Parkinson’s disease, epilepsy, and severe head trauma: a register-based follow-up study. Am J Epidemiol, 142, 1300–5. Brown, S.W. (1989). Cognitive impairment in epilepsy. Educ Child Psychol, 6, 25–32. Brown, S.W. (1992). Organic brain syndromes in chronic epilepsy – what is the truth about epileptic dementia? Seizure, 1 (Suppl. A), S26/1. Brown, S.W. (1999). Epilepsy dementia: intellectual deterioration as a consequence of epileptic seizures. In Epilepsy and Mental Retardation, ed. M. Sillanpaa et al., pp. 115–34. Petersfield: Wrightson Biomedical Publishing. Brown, S.W. and Abeyasinghe, R. (1984). Is there an epileptic dementia? Acta Neurol Scand Suppl, 70, 233. Brown, S.W. and Reynolds, E.H. (1981). Cognitive impairment in epileptic patients. In Epilepsy and Psychiatry, ed. E.H. Reynolds and M.R. Trimble. London: Churchill Livingstone. Brown, S.W. and Vaughan, M. (1988). Dementia in epileptic patients. In Epilepsy, Behaviour and Cognitive Function, ed. M.R. Trimble and E.H. Reynolds, pp. 177–88. Chichester: John Wiley & Sons. Brown, S.W. and Vaughan, M. (1990). The nature of cognitive changes in people with chronic epilepsy. Acta Neurol Scand Suppl, 82, 13. Buoni, S., Grosso, S. and Fois, A. (1998). Lamotrigine treatment in childhood drug resistant epilepsy. J Child Neurol, 13, 163–7. Butlin, A.T., Danta, G. and Cook, M.L. (1984). Anticonvulsant effects on the memory performance of epileptics. Clin Exp Neurol, 20, 27–35. Calandre, E.P., Dominguez-Granados, R., Gomez-Rubio, M. and Molina-Font, J.A. (1990). Cognitive effects of long-term treatment with phenobarbital and valproic acid in school children. Acta Neurol Scand, 81, 504–6. Castellani, R., Smith, M.A., Richey, P.L., Kalaria, R., Gambetti, P. and Perry, G. (1995). Evidence for oxidative stress in Pick disease and corticobasal degeneration. Brain Res, 696, 268–71. Collacott, R.A. (1993). Epilepsy, dementia and adaptive behaviour in Down’s syndrome. J Intellect Disabil Res, 37, 153–60. Cooper, S.A. (1997). High prevalence of dementia among people with learning disabilities not attributable to Down’s syndrome. Psychol Med, 27, 609–16. Dawson, S. and Conn, J.C.N. (1929). The intelligence of epileptic children. Arch Dis Child, 4, 142–51. Diderot & d’Alambert (ed.) (1765). Encyclopédie ou Dictionnaire Raisonné des Sciences, des Arts et des Métieres, par une Societé de Gens de Lettres, Vol. 4, Paris: David Briasson; Le Breton: Durand (quoted in Berrios, 1995).

148

S.W. Brown Dodrill, C.B. (1986). Correlates of generalized tonic-clonic seizures with intellectual, neuropsychological, emotional and social function in patients with epilepsy. Epilepsia, 27, 399–411. Dodrill, C.B., Arnett, J.L., Sommerville, K.W. and Shu, V. (1997). Cognitive and quality of life effects of differing dosages of tiagabine in epilepsy. Neurology, 48, 1025–31. During, M.J. and Spencer, D.D. (1993). Extracellular hippocampal glutamate and spontaneous seizure in the conscious human brain. Lancet, 341, 1607–10. Fetterman, J. and Barnes, R.R. (1934). Serial studies of the intelligence of patients with epilepsy. Neuropsychologia, 7, 287–300. Forsgren, L., Bucht, G., Eriksson, S. and Bergmark, L. (1996). Incidence and clinical characterization of unprovoked seizures in adults: a prospective population-based study. Epilepsia, 37, 224–9. Forsythe, I., Butler, R., Berg, I. and McGuire, R. (1991). Cognitive impairment in new cases of epilepsy randomly assigned to carbamazepine, phenytoin and sodium valproate. Dev Med Child Neurol, 33, 524–34. Froscher, W., Maier, V., Laage, M. et al. (1995). Folate deficiency, anticonvulsant drugs, and psychiatric morbidity. Clin Neuropharmacol, 18, 165–82. Geddes, J.F., Vowles, G.H., Nicoll, J.A. and Revesz, T. (1999). Neuronal cytoskeletal changes are an early consequence of repetitive head injury. Acta Neuropathol, 98, 171–8. Genton, P. and Dravet, C. (1997). Lennox–Gastaut syndrome and other childhood epileptic encephalopathies. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 2355–66. Philadelphia, PA: Lippincott-Raven. Gillham, R.A., Williams, N., Wiedmann, K.D., Butler, E., Larkin, J.G. and Brodie, M.J. (1990). Cognitive function in adult epileptic patients established on anticonvulsant monotherapy. Epilepsy Res, 7, 219–25. Gokyigit, A. and Caliskan, A. (1995). Diffuse spike-wave status of 9-year duration without behavioral change or intellectual decline. Epilepsia, 36, 210–13. Goldsmith, I.L., Zupanc, M.L. and Buchhalter, J.R. (2000). Long-term seizure outcome in 74 patients with Lennox–Gastaut syndrome: effects of incorporating MRI head imaging in defining the cryptogenic subgroup. Epilepsia, 41, 395–9. Goldstein, L.H. (1997). Neuropsychological assessment. In The Clinical Psychologist’s Handbook of Epilepsy, ed. C. Cull and L.H. Goldstein, pp. 18–34. London: Routledge. Gorman, A.M., McGowan, A., O’Neill, C. and Cotter, T. (1994). Oxidative stress and apoptosis in neurodegeneration. J Neurol Sci, 139 (Suppl.), 45–52. Gowers, W.R. (1885). Epilepsy and Other Chronic Convulsive Diseases. New York: William Wood and Company. Guerrini, R., Belmonte, A., Canapicchi, R., Casalini, C. and Perucca, E. (1998). Reversible pseudoatrophy of the brain and mental deterioration associated with valproate treatment. Epilepsia, 39, 27–32. Hart, Y.M., Andermann, F., Robitaille, Y., Laxer, K.D., Rasmussen, T. and Davis, R. (1998). Double pathology in Rasmussen’s syndrome: a window on the etiology? Neurology, 50, 731–5. Holmes, G.L. (1997). Epilepsy in the developing brain: lessons from the laboratory and clinic. Epilepsia. 38, 12–30. Ingvar, D.H. (1984). Epilepsy related to cerebral blood flow and metabolism. Acta Psychiatr Scand Suppl, 313, 21–6.

149

Dementia and epilepsy Jambaque, I., Chiron, C., Dumas, C., Mumford, J. and Dulac, O. (2000). Mental and behavioural outcome of infantile epilepsy treated by vigabatrin in tuberous sclerosis patients. Epilepsy Res, 38, 151–60. Jenner, P. and Olanow, C.W. (1994). Oxidative stress and the pathogenesis of Parkinson’s disease. Neurology, 47 (Suppl. 3), S161–70. Johnston, M.V. (1996). Developmental aspects of epileptogenesis. Epilepsia, 37 (Suppl. 1), S2–9. Jokeit, H. and Ebner, A. (1999). Long term effects of refractory temporal lobe epilepsy on cognitive abilities: a cross sectional study. J Neurol Neurosurg Psychiatry, 67, 44–50. Kälviainen, R., Aikia, M., Helkala, EL., Mervaala, E. and Riekkinen, P.J. (1992). Memory and attention in newly diagnosed epileptic seizure disorder. Seizure, 1, 255–62. Kälviainen, R., Aikia, M., Mervaala, E., Saukkonen, A.M., Pitkänen, A. and Riekkinen, P.J. Sr (1996). Long-term cognitive and EEG effects of tiagabine in drug-resistant partial epilepsy. Epilepsy Res, 25, 291–7. Kasteleijn-Nolst Trenite, D.G., Bakker, D.J., Binnie, C.D., Buerman, A. and Van Raaij, M. (1988). Psychological effects of subclinical epileptiform EEG discharges. I. Scholastic skills. Epilepsy Res, 2, 111–16. Kugelmass, I.N., Poull, L.E. and Rudnick, J. (1938). Mental growth of epileptic children. Am J Dis Child, 55, 295–303. Kuzniecky, R., Guthrie, B., Mountz, J. et al. (1997). Intrinsic epileptogenesis of hypothalamic hamartomas in gelastic epilepsy. Ann Neurol, 42, 60–7. Lehesjoki, A.E., Koskiniemi, M., Pandolfo, M. et al. (1992). Linkage studies in progressive myoclonus epilepsy: Unverricht–Lundborg and Lafora’s diseases. Neurology, 42, 1545–50. Lishman, W.A. (1998). Organic Psychiatry: The Psychological Consequences of Cerebral Disorder, Third Edition. Oxford: Blackwell Science. Long, C.G. and Moore, J.R. (1979). Parental expectations for their epileptic children. J Child Psychol Psychiatry, 24: 299–312. Lott, I.T. and Lai, F. (1982). Dementia in Down’s syndrome: observations from a neurology clinic. Appl Res Ment Retardation, 3, 233–9. Mackenzie, I.R. and Miller, L.A. (1994). Senile plaques in temporal lobe epilepsy. Acta Neuropathol, 87, 504–10. Marston, D., Besag, F., Binnie, C.D. and Fowler, M. (1993). Effects of transitory cognitive impairment on psychosocial functioning of children with epilepsy: a therapeutic trial. Dev Med Child Neurol, 35, 574–81. Matthews, C.G. and Harley, J.P. (1975). Cognitive and motor-sensory performances in toxic and nontoxic epileptic subjects. Neurology, 25, 184–8. Maudsley, H. (1874). Responsibility in Mental Disease. New York: Appleton. Meador, K.J. and Baker, G.A. (1997). Behavioral and cognitive effects of lamotrigine. J Child Neurol, 12 (Suppl. 1), S44–7. Meldrum, B.S. (1994). The role of glutamate in epilepsy and other CNS disorders. Neurology, 44 (Suppl. 23). Metz-Lutz, M.N., Kleitz, C., de Saint, M.A., Massa, R., Hirsch, E. and Marescaux, C. (1999). Cognitive development in benign focal epilepsies of childhood. Dev Neurosci, 21, 182–90.

150

S.W. Brown Michaelis, E.K. (1998). Molecular biology of glutamate receptors in the central nervous system and their role in excitotoxicity, oxidative stress and aging. Progr Neurobiol, 54, 369–415. Monaco, F. (1996). Cognitive effects of vigabatrin: a review. Neurology, 47 (Suppl. 1), S6–11. Morel, B.A. (1857). Traité des Degénéréscences Physiques, Intellectuelles et Morales de l’Espéce Humaine et des Causes qui Produisent ces Varietés Maladive. Paris: Bailliére. Neubauer, C. (1970). Mental deterioration in epilepsy due to folate deficiency. Br Med J, 2, 759–61. Oguni, H., Hayashi, K. and Osawa, M. (1996). Long-term prognosis of Lennox–Gastaut syndrome. Epilepsia, 37 (Suppl. 3), 44–7. O’Leary, D.S., Lovell, M.R., Sackellares, J.C. et al. (1983). Effects of age of onset of partial and generalized seizures on neuropsychological performance in children. J Nerv Ment Dis, 171, 624–9. O’Riordan, J.I., Javed, M., Murphy, R. and Hutchinson, M. (1995). Sneddon’s syndrome: clinical course and outcome. Irish Med J, 88, 66–7. Papazian, O., Canizales, E., Alfonso, I., Archila, R., Duchowny, M. and Aicardi, J. (1995). Reversible dementia and apparent brain atrophy during valproate therapy. Ann Neurol, 38, 687–91. Patterson, H.A. and Fonner, D. (1928). Some observations on the intelligence quotient in epileptics. Psychiatr Quart, 31, 542–8. Piccirilli, M., D’Alessandro, P., Sciarma, T. et al. (1994). Attention problems in epilepsy: possible significance of the epileptogenic focus. Epilepsia, 35, 1091–6. Prevey, M.L., Delaney, R.C., Cramer, J.A., Cattanach, L., Collins, J.F. and Mattson, R.H. (1996). Effect of valproate on cognitive functioning. Comparison with carbamazepine. The Department of Veterans Affairs Epilepsy Cooperative Study 264 Group. Arch Neurol, 53, 1008–16. Provinciali, L., Bartolini, M., Mari, F., Del Pesce, M. and Ceravolo, M.G. (1996). Influence of vigabatrin on cognitive performances and behaviour in patients with drug-resistant epilepsy. Acta Neurol Scand, 94, 12–18. Prusiner, S.B. (1994). Biology and genetics of prion diseases. Annu Rev Microbiol, 48, 655–86. Pulliainen, V. and Jokelainen, M. (1994). Effects of phenytoin and carbamazepine on cognitive functions in newly diagnosed epileptic patients. Acta Neurol Scand, 89, 81–6. Rabinowicz, A.L., Correale, J., Boutros, R.B., Couldwell, W.T., Henderson, C.W. and DeGiorgio, C.M. (1996). Neuron-specific enolase is increased after single seizures during inpatient video/EEG monitoring. Epilepsia, 37, 122–5. Ramchand, C.N., Davies, J.I., Tresman, R.L., Griffiths, I.C. and Peet, M. (1996). Reduced susceptibility to oxidative damage of erythrocyte membranes from medicated schizophrenic patients. Prostaglandins Leukot Essent Fatty Acids, 55, 27–31. Rasmussen, T., Olszewski, J. and Lloyd-Smith, D. (1958). Focal seizures due to chronic localized encephalitis. Neurology, 8, 435–45. Riva, D. and Devoti, M. (1996). Discontinuation of phenobarbital in children: effects on neurocognitive behavior. Pediatr Neurol, 14, 36–40. Rodin, E.A., Schmaltz, S. and Twitty, G. (1986). Intellectual functions of patients with childhoodonset epilepsy. Dev Med Child Neurol, 28, 25–33.

151

Dementia and epilepsy Rogers, S.W., Andrews, P.I., Gahring, L.C. et al. (1994) Autoantibodies to glutamate receptor GluR3 in Rasmussen’s encephalitis. Science, 265, 648–51. Rosen, J.A. (1968). Dilantin dementia. Trans Am Neurol Assoc, 93, 273. Roulet Perez, E., Davidoff, V., Despland, P.A. and Deonna, T. (1993). Mental and behavioural deterioration of children with epilepsy and CSWS: acquired epileptic frontal syndrome. Dev Med Child Neurol, 35, 661–74. Rugland, A.L. (1990). Neuropsychological assessment of cognitive functioning in children with epilepsy. Epilepsia, 31 (Suppl. 4), S41–4. Rutter, M., Graham, P. and Yule, W. (1970). A neuropsychiatric study in childhood. Clinics in Developmental Medicine Nos. 35/36. London: Spastics International and Heineman Medical. Sabers, A., Moller, A., Dam, M. et al. (1995). Cognitive function and anticonvulsant therapy: effect of monotherapy in epilepsy. Acta Neurol Scand, 92, 19–27. Seidenberg, M., O’Leary, D.S., Berent, S. and Boll, T. (1981). Changes in seizure frequency and test-retest scores on the Wechsler Adult Intelligence Scale. Epilepsia, 22, 75–83. Siebelink, B.M., Bakker, D.J., Binnie, C.D. and Kasteleijn-Nolst Trenite, D.G. (1988). Psychological effects of subclinical epileptiform EEG discharges in children. II. General intelligence tests. Epilepsy Res, 2, 117–21. Sims, N.R. (1996). Energy metabolism, oxidative stress and neuronal degeneration in Alzheimer’s disease. Neurodegeneration, 5, 435–40. Smith, M.C. (1997). Landau–Kleffner syndrome and continuous spikes and waves during slow sleep. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 2367–77. Philadelphia, PA: Lippincott-Raven. Stephens, C.J. (1992). Sneddon’s syndrome. Clin Exp Rheumatol, 10, 489–92. Stores, G., Williams, P.L., Styles, E. and Zaiwalla, Z. (1992). Psychological effects of sodium valproate and carbamazepine in epilepsy. Arch Dis Child, 67, 1330–7. Strauss, E., Loring, D., Chelune, G. et al. (1995) Predicting cognitive impairment in epilepsy: findings from the Bozeman epilepsy consortium. J Clin Exp Neuropsychol, 17, 909–17. Straussberg, R., Kivity, S., Weitz, R., Harel, L. and Gadoth, N. (1998). Reversible cortical atrophy and cognitive decline induced by valproic acid. Eur J Paediatr Neurol, 2, 213–8. Sullivan, E.B. and Gahagan, L. (1935). On intelligence of epileptic children. Genet Soc Gen Psychol Monogr, 17, 309–75. Sveinbjornsdottir, S., Sander, J.W., Patsalos, P.N., Upton, D., Thompson, P.J. and Duncan, J.S. (1994). Neuropsychological effects of tiagabine, a potential new antiepileptic drug. Seizure, 3, 29–35. Temkin, O. (1971). The Falling Sickness, 2nd Edition revised. Baltimore: Johns Hopkins Press. Vermeulen, J. and Aldenkamp, A.P. (1995). Cognitive side-effects of chronic antiepileptic drug treatment: a review of 25 years of research. Epilepsy Res, 22, 65–95. World Health Organization (1992). The International Classification of Mental and Behavioural Disorders: Diagnostic Criteria for Research (Tenth Revision) (ICD–10). Geneva: WHO. Yerby, M.S., Shaw, C.M. and Watson, J.M. (1986). Progressive dementia and epilepsy in a young adult: unusual intraneuronal inclusions. Neurology, 36, 68–71.

11

The risk of cognitive decline in patients with refractory temporal lobe epilepsy Hennric Jokeit1 and Alois Ebner2 1 2

Swiss Epilepsy Centre, Zurich, Switzerland Epilepsy Centre Bethel, Bielefeld, Germany

Introduction Patients with refractory temporal lobe epilepsy (TLE) are at higher risk for mental and cognitive impairments than healthy controls (Hermann et al., 1987, 1997; Trimble, 1988). Typically patients with right-sided TLE are frequently impaired in visuo-spatial retention tasks; patients with left-sided TLE may exhibit deficits of verbal memory. Because of these frequent and prominent memory deficits it is sometimes neglected that many TLE patients perform below healthy control subjects on a variety of neuropsychological tests including intelligence measures (Hermann et al., 1997). The probable reason is that the temporal epileptogenic zone is not only malfunctioning but also adversely influences remote cerebral structures resulting in additional cognitive deficits (Engel et al., 1991; Lüders and Awad, 1991). One of our recent studies confirmed that assumption (Jokeit et al., 1997). We investigated 96 TLE patients by FDG-PET and neuropsychological assessment who had a corresponding unilateral temporal hypometabolism, left hemisphere speech dominance, full scale IQ of ⬎70 and no extra-temporal lesion on MRIs. The regional glucose metabolism was determined in each patient in homologous regions including prefrontal cortex. A multivariate analysis of variance revealed that the observed prefrontal metabolic disturbances, that are remote from the temporal epileptogenic zone, were associated with impaired intellectual abilities. Patients who demonstrated prefrontal metabolic disturbances performed worse on verbal as well as performance IQ measures than patients without prefrontal metabolic disturbances. Although patients who demonstrated prefrontal metabolic disturbances had an earlier epilepsy onset, the revealed association with cognitive impairment was unrelated to the age at onset. Nevertheless, age at epilepsy onset is a well-documented risk factor for cognitive impairment (Bourgeois et al., 1983; Glosser et al., 1997). It is assumed that an early epilepsy onset affects 152

153

Risk of cognitive decline in refractory TLE

considerably the maturation of brain functions and structures as well as the acquisition of complex knowledge and abilities. One of the most frequent questions asked by epilepsy patients and their relatives is whether seizures are destructive and contribute to progressive decline of intellectual abilities. Generally dementia is a very rare syndrome in TLE patients and does not represent the typical course of refractory TLE. However, a growing number of clinical and experimental studies suggest a slow but ongoing progression of symptoms with increasing duration of refractory TLE or total number of lifetime seizures. A long duration of intractable epilepsy is related to a considerable number of focal or generalized seizures, pathological interictal electric brain activity, chronic and transient metabolic disturbances (Arnold et al., 1996; Savic et al., 1997; Theodore et al., 1989) and chronic antiepileptic medication usually with high serum levels (Hermanns et al., 1996). It is suggested that these noxious factors may induce secondary neurophysiological and structural long-term changes (Beach et al., 1995; Ben-Ari and Represa, 1990; Bengzon et al., 1997; Jokeit et al., 1997; Marsh et al., 1997; Multani et al., 1994; Tasch et al., 1999; Theodore and Gaillard 1999; Theodore et al., 1999). A few neuropsychological studies have been aimed at elucidating this question, whether the cognitive abilities of patients deteriorate with an increasing duration of intractable epilepsy. But neither short-term longitudinal nor cross-sectional studies demonstrated a convincing relationship between psychometric intelligence and the duration of epilepsy in samples of adult patients (Bourgeois et al., 1983; Brown, 1996; Brown and Vaughan, 1988; Dodrill and Wilensky, 1992; Rodin et al., 1986; Seidenberg et al., 1981; Selwa et al., 1994; Strauss et al., 1995; Trimble, 1988). On one hand, the absence of an evident duration effect suggests that probably no dramatic cognitive changes occur within periods of some years in adult TLE patients. On the other hand, methodological restrictions, for example, a limited time range of longitudinal studies which rarely exceeds a decade, an undetected cohort bias in cross-sectional studies, or confounded variables might cover possible duration effects. Additionally, in studies with small sample sizes, the possibility of a Type II error is frequently neglected. The results of densiometric and volumetric cross-sectional studies in TLE patients demonstrate that differences in neuropsychological measures became significant only if patients differed in decades of the duration of epilepsy (Barr et al., 1997; Breier et al., 1997; Jokeit et al., 1999; Mathern et al., 1995a, b; Multani et al., 1994; Salmenperä et al., 1998). Moreover, these studies indicate that neuronal injury within and beyond the temporal lobes continues to occur with ongoing seizure activity in TLE patients. However, it is well known that the brain possesses a large degree of redundancy and plasticity, and has compensatory mechanisms that may prevent or postpone a cognitive decline due to small but ongoing brain damage (Calne et al., 1986; Lewin, 1980). Therefore, the

154

H. Jokeit and A. Ebner

individual brain reserve or spare capacity may also considerably influence the course of cognitive changes in patients with refractory TLE. Cross-sectional studies There are different approaches to infer cognitive changes along the time axis. From a scientific as well as methodological point of view, prospective longitudinal studies are generally superior to cross-sectional studies. Clinical observations and results of longitudinal studies suggest that there is, if any, no rapid cognitive decline in patients with refractory TLE. Consequently, longitudinal studies in TLE patients are confronted with the problem of probably small effect sizes. Hence, to provide statistical evidence, large samples of patients have to be observed over long periods of time, probably exceeding decades. In contrast, cross-sectional studies allow the recruitment of large sample sizes. However, the existence of a duration effect only can be inferred from interindividual differences in the duration of illness. Hence, the interpretation of duration effects strictly presupposes that the patients were recruited from the same population. Otherwise a cohort bias might considerably affect the results. Although conclusions drawn from cross-sectional studies are limited in some respects they may reveal trends that might be covered in longitudinal studies restricted by time and sample size. Also, dependent variables, usually intelligence measures, can be treated differentially as we demonstrate by the following studies that are based on independent samples. In the first study we considered differences between an estimated measure of former or premorbid intelligence and the current performance in an intelligence test as a function of duration of epilepsy. This seems to be the only way to infer individually a cognitive decline using psychometric instruments during a single neuropsychological investigation. In the second study we directly related the duration of epilepsy with the current IQ test results. However, this approach is only appropriate for sample studies. Neuropsychological investigations of a patient frequently include so-called intelligence trace tests to estimate the former intelligence in order to compare it with current test results. Several studies have shown that most patients with cerebral lesions or early dementia are unimpaired in intelligence trace tests, whereas the same patients do worse in standard IQ tests. The difference between the intelligence trace test and standard IQ tests is considered to be a measure of cognitive decline (Figure 11.1). If intellectual abilities deteriorate with increasing duration of epilepsy the difference between both measures should become larger. In our study (Jokeit et al., 2000) we used the passive vocabulary-intelligence test (MWT–B) as an intelligence trace test (Lehrl, 1995). In this test, patients have to identify a real word among four pseudo-words in rows of increasing difficulty.

Risk of cognitive decline in refractory TLE

IQ Difference

155

30 20 10 0 ⫺10 ⫺20 ⫺30 ⫺40 ⫺50 0

10

20

30

40

50

Duration of epilepsy (years) Figure 11.1. Difference values of the estimated WAIS-R full scale IQ minus MWT-B-IQ (vocabulary intelligence), as a function of the duration of epilepsy. Data of 37 patients with refractory temporal lobe epilepsy are shown. The linear regression function confirms a negative correlation between the difference value and the duration of epilepsy. (With permission from Jokeit et al. 2000.)

Patients with a long duration of epilepsy more frequently demonstrated a considerable difference between both IQs compared with patients with a shorter duration of TLE. In a larger study (Jokeit and Ebner, 1999), we examined the effects of duration of epilepsy on the performance in a standard IQ-test as an indicator of global cognitive abilities and integrity of higher brain functions. Furthermore, the influence of the variable education on psychometric intelligence and its interaction with the duration of epilepsy was investigated. Epidemiological studies identified that

156

H. Jokeit and A. Ebner

education as an indicator of brain reserve modifies the clinical expression of dementia and Alzheimer’s disease (Evans et al., 1997; Satz, 1993; Schmand et al., 1997; Stern et al., 1994; Zhang et al., 1990). We studied 209 consecutive patients of our epilepsy surgery programme with TLE who fulfilled the following criteria: seizures of unilateral temporal origin as demonstrated by continuous interictal and ictal video/EEG-monitoring, unilateral lesions within the temporal lobes as demonstrated by MRI, and Full Scale Intelligence Quotient (FSIQ) greater than 55 to exclude severely mentally impaired patients. Detailed information on clinical and demographic variables are given in the original study (Jokeit and Ebner, 1999). Patients underwent a neuropsychological evaluation designed for patients with TLE. To have a comprehensive psychometric measure of global cognitive abilities, which was well normed to age-matched healthy controls, we used the Full Scale Intelligence Quotient (FSIQ) from the German version of the Wechsler Adult Intelligence Scale – Revised (WAIS–R) (Tewes, 1991; Wechsler, 1981). The FSIQ was estimated from the subtests Information, Comprehension, Similarities, Digit Symbol, Picture Completion and Block Design. First a data set of 16 independent variables for each patient was submitted to a linear multiple regression analysis. The analyses on FSIQ revealed significant contributions by the variables education and duration of epilepsy. The equation explained 34.6% of total variance. No further variable contributed significantly. To reveal possible factor interactions and to exclude linear effects of covariates in a one-way ANOVA model, the continuous variable duration of epilepsy was recoded into three values: ⬍15 years⫽0, 15–30 years⫽1, and ⬎30 years of refractory epilepsy⫽2. The side of seizure origin was submitted as the second factor (left, right). The presence or absence of lesions beyond mesio-temporal structures in MRI scans was submitted as the third factor (0, 1). Education, age at epilepsy onset, ranked frequency of interictal epileptiform discharges, ranked frequency of habitual seizures, presence of secondarily generalized seizures in a patient’s history, ranked frequency of secondarily generalized seizures within the last year, serum level for carbamazepine, phenytoin and phenobarbital, and AED mono- or polytherapy were controlled as covariates. Only the covariate education was significantly related to FSIQ. After removing linear effects of covariates the factor duration of epilepsy was significant. The factors side of seizure onset and presence or absence of temporal lesions beyond mesio-temporal structures did not reach significance. No interactions reached significance. Post-hoc contrasts adjusted by covariates revealed that patients with a duration of epilepsy for more than 30 years performed worse than patients with less than 15 years of epilepsy and patients with 15–30 years of epilepsy. Patients with less than 15 years of epilepsy did not differ from patients with 15–30 years of epilepsy in FSIQ values.

157

Risk of cognitive decline in refractory TLE

Since it is known that variable education explains a considerable amount of variance of FSIQ values, we computed an ANOVA with factors duration of epilepsy (0, 1, 2) and education (low, high) to reveal possible interactions and to specify the effect of variable education. The educational level was dichotomized into low (patients who did not attend or did not finish at secondary school, Realschule in Germany, n⫽109) and high (patients who finished at least at secondary school, n⫽100). The side of seizure origin, presence or absence of lesions beyond mesiotemporal structures, age at epilepsy onset, ranked frequency of interictal epileptiform discharges, ranked frequency of habitual seizures, presence of generalized seizures in a patient’s history, ranked frequency of secondarily generalized seizures within the last year, serum level for carbamazepine, phenytoin and phenobarbital, and AED mono- or polytherapy were controlled as covariates. The ANOVA revealed effects of the factors education and duration of epilepsy on the FSIQ. The interaction between both factors did not reach significance. No covariate was significantly related to FSIQ. Figure 11.2 shows mean FSIQ values for both educational groups as a function of duration of TLE. Contrasts adjusted for covariates revealed that the mean FSIQ values of groups with less than 15 years and 15–30 years of TLE did not differ in patients with high educational attainment. However, patients with more than 30 years of TLE performed worse than patients with less than 15 years or 15–30 years of epilepsy. Patients with low educational attainment and less than 15 years of TLE performed better than patients with 15–30 years of TLE and patients with more than 30 years of TLE. But there was no significant difference between patients with 15–30 years and more than 30 years of TLE in the low education group. Although not confirmed by an ANOVA interaction, these contrasts suggest a duration-dependent difference in mean FSIQ values of patients with low and high educational attainment. The absence of a significant ANOVA interaction probably results from a similar decremental trend of the duration effect in low- and high-educated patients. Conclusions Both studies demonstrate that patients with a long history of intractable TLE were at higher risk of cognitive impairment than patients with a shorter duration of TLE. Interestingly, in patients with higher educational attainment, the mean FSIQ was stable for a longer duration of TLE than in less-educated patients. We revealed that the educational level and the duration of epilepsy were the best predictors for psychometric intelligence. Then we provided evidence that only a long duration of TLE (more than 30 years) was related to impaired psychometric intelligence in the total sample. The linear influence of the variables age at epilepsy onset, educational level of patients, patient’s serum level of first-line antiepileptic

Mean FSIQ

158

H. Jokeit and A. Ebner 120

110

100

90

80

low–high

low–high

low–high

15–30

⬍30

70 ⬍15

Duration of epilepsy (years)

Figure 11.2. Mean FSIQ values with 95% confidence intervals of patients with low (n⫽109) and high (n⫽100) educational attainment for groups with a duration of epilepsy ⬍15 years, 15–30 years and ⬎30 years. The factors education and duration of epilepsy were significant (P⬍0.01). Asterisks (* P⬍0.05; ** P⬍0.01; one tailed) indicate significant contrasts between adjacent duration groups adjusted for covariates. (With permission from Jokeit and Ebner, 1999.)

drugs, polypharmacy, frequency of habitual and secondarily generalised seizures and frequency of interictal epileptiform discharges on psychometric intelligence were statistically controlled. Age at testing controlled for duration of TLE was not significantly related to FSIQ. The factors side of TLE and presence or absence of lesions beyond the mesio-temporal structures did not show main effects or interactions. Therefore, the effect of duration of epilepsy cannot be attributed to those covariates and factors. The patient’s educational level captured the most amount of FSIQ variance. This could explain why in contrast to other studies (Strauss et al.,

159

Risk of cognitive decline in refractory TLE

1995) the remaining covariates (e.g. age at epilepsy onset) did not reach significance. It is reasonable to assume that human brains develop a functional reserve or have a spare capacity to cope with a stepwise neuronal loss by efficiency, redundancy, plasticity and reorganization (Lewin, 1980; Meier-Ruge et al., 1991; Stern et al., 1996). Studies of different degenerative brain disorders (e.g. Parkinson’s disease, vascular and Alzheimer’s dementia) suggest that a functional decline becomes apparent only if a certain amount of brain parenchyma is insulted (Boone et al., 1992; Hornykiewicz, 1988; Pasquier and Leys, 1997; Small et al., 1995; Tomlinson et al., 1970). A long duration of intractable TLE is related to a considerable number of focal or secondarily generalized seizures, pathological interictal electric brain activity, chronic and transient metabolic disturbances due to morphological lesions (Arnold et al., 1996), seizures (Savic et al., 1997) and antiepileptic medication (Theodore et al., 1989), often the chronic antiepileptic medication with usually high serum levels (Hermanns et al., 1996). It is suggested that each of these factors may separately adversely affect cognitive functioning (Dreifuss, 1992; Lesser et al., 1986). The presence of reactive microglia (Beach et al., 1995), reduced dendritic spine density, dendritic swellings (Multani et al., 1994) and senile plaques (Mackenzie and Miller, 1994) in adult temporal lobe specimens suggests that neuronal injury continues to occur with ongoing seizure activity in TLE patients. Multani et al. (1994) demonstrated a correlation between decreased dendritic spine density remote from the epileptogenic zone and duration of seizure history. In patients with mesial TLE densiometric techniques have revealed secondary declines in hippocampal neuron densities associated with long histories of habitual seizures (Mathern et al., 1995c, 1996). Recent studies have suggested a secondary decline of hippocampal volume and temporal lobe metabolism in refractory TLE patients (Barr et al., 1997; Breier et al., 1997; Jokeit et al., 1999; Salmenperä et al., 1998). Hermann et al. (1997) reported that patients with hippocampal sclerosis had more generalized cognitive impairment, a significantly longer history of intractable TLE and a lower educational level than TLE patients without significant hippocampal sclerosis. We assume that a cumulation of small neuro-degenerative effects of noxious neurochemical agents, abnormal brain electric events, and metabolic disturbances over decades of epilepsy, accompanied by ageing, may increase the probability that the functional brain reserve or spare capacity is exhausted at a younger age than expected (mean age of patients with TLE duration ⬎30 years was 44 years), and deterioration of cognitive functions may begin (Meier-Ruge et al., 1991; Mori et al., 1997; Stern et al., 1996). The extent of functional reserve and therefore the vulnerability of brain functions may vary considerably between people. It was suggested that higher educational attainment is related to a higher reserve against cognitive impairment due to

160

H. Jokeit and A. Ebner

stepwise ongoing brain injury (Satz, 1993; Timiras, 1995). Our results of analyses in patients with lower and higher educational attainment are in accordance with findings of epidemiological studies on dementia and Alzheimers’s disease (Evans et al., 1997; Satz, 1993; Schmand et al., 1997; Stern et al., 1994; Zhang et al., 1990). In patients with higher educational attainment the mean FSIQ was stable for a longer duration of TLE than in less-educated patients. Thus, higher educational attainment as an indicator of higher cognitive reserve might delay the onset of cognitive decline in patients with intractable TLE. The fact that we found significant effects in the whole sample analysis only in patients with a history of intractable epilepsy lasting longer than three decades may question conclusions drawn from negative findings of studies on adverse effects of duration of refractory TLE (Mackenzie et al., 1996; Selwa et al., 1994). Based on our results, only prospective long-term studies which exceed three decades might reveal the causes of a presumable decline in cognitive functioning of patients with intractable TLE. Such studies may solve the important question as to whether certain epileptic syndromes are progressive disorders (Girvin, 1992; Gloor, 1991; Lesser et al., 1986; Mathern et al., 1996; Sadzot 1997; Sutula et al., 1989). However, today wellcontrolled cross-sectional studies comparing different well-defined epileptic syndromes and including retest measures to compare different treatment strategies may help to isolate adverse factors and to identify patients with increased risk of cognitive decline and symptom progression.

R E F E R E N C ES Arnold, S., Schlaug, G., Niemann, H. et al. (1996). Topography of interictal glucose hypometabolism in unilateral mesiotemporal epilepsy. Neurology, 46, 1422–30. Barr, W.B., Ashtari, M. and Schaul, N. (1997). Bilateral reductions in hippocampal volume in adults with epilepsy and a history of febrile seizures. J Neurol Neurosurg Psychiatry, 63, 461–7. Beach, T.G., Woodhurst, W.B., MacDonald, D.B. and Jones, W.M. (1995). Reactive microglia in hippocampal sclerosis associated with human temporal lobe epilepsy. Neurosci Lett, 191, 27–30. Ben-Ari, Y. and Represa, A. (1990). Brief seizure episodes induce long-term potentiation and mossy fibre sprouting in the hippocamus. Trends Neurosci, 13, 312–18. Bengzon, J., Kokaia, Z., Elmer, E., Nanobashvili, A., Kokaia, M. and Lindvall, O. (1997). Apoptosis and proliferation of dentate gyrus neurons after single and intermittent limbic seizures. Proc Natl Acad Sci USA, 94, 10432–7. Boone, K.B., Miller, B.L., Lesser, I.M. et al. (1992). Neuropsychological correlates of white-matter lesions in healthy elderly subjects. Arch Neurol, 49, 549–54. Bourgeois, B.F.D., Prensky, A.L., Palkes, H.S., Talent, B.K. and Busch, S.G. (1983). Intelligence in epilepsy: a prospective study in children. Ann Neurol, 14, 438–44.

161

Risk of cognitive decline in refractory TLE Breier, J.L., Mullani, N.A., Thomas, A.B. et al. (1997). Effects of duration of epilepsy on the uncoupling of metabolism and blood flow in complex partial seizures. Neurology, 48, 1047–53. Brown, S.W. (1996). Epilepsy dementia: Intellectual deterioration as a consequence of epileptic seizures. Epilepsia, 37 (Suppl. 4), S122–3. Brown, S.W. and Vaughan, M. (1988). Dementia in epileptic patients. In Epilepsy, Behaviour and Cognitive Function, ed. M.R. Trimble and E.H. Reynolds, pp. 177–88. Chichester: John Wiley & Sons. Calne, D.B., Eisen, A., McGeer, E. and Spencer, P. (1986). Alzheimer’s disease, Parkinson’s disease, and motoneurone disease: abiotrophic interaction between ageing and environment? Lancet, 2 (8515), 1067–70. Dodrill, C.B. and Wilensky, A.J. (1992). Neuropsychological abilities before and after 5 years of stable antiepileptic drug therapy. Epilepsia, 33, 327–34. Dreifuss, E.F. (1992). Cognitive function – victim of disease or hostage to treatment? Epilepsia, 33 (Suppl. 1), S7–12. Engel, J. Jr, Brandler, R., Griffith, N.C. and Caldecott-Hazard, S. (1991). Neurobiological evidence for epilepsy-induced interictal disturbances. Adv Neurol, 55, 97–111. Evans, D.A., Hebert, L.E., Becket, L.A. et al. (1997). Education and other measures of socioeconomic status and risk of incident Alzheimer disease in a defined population of older persons. Arch Neurol, 54, 1399–405. Girvin, J.P. (1992). Is epilepsy a progressive disorder? J Epilepsy, 5, 94–104. Gloor, P. (1991). Mesial temporal sclerosis: historical background and an overview from a modern perspective. In Epilepsy Surgery, ed. H.O. Lüders, pp. 689–703. New York: Raven Press. Glosser, G., Cole, L.C., French, J.A., Saykin, A.J. and Sperling, M.R. (1997). Predictors of intellectual performance in adults with intractable temporal lobe epilepsy. J Int Neuropsychol Soc, 3, 252–9. Hermann, B.P., Wyler, A.R. and Richey, E.T. (1987). Epilepsy frontal lobe, and personality. Biol Psychiatry, 22, 1055–7. Hermann, B.P., Seidenberg, M., Schoenfeld, J. and Davies, K. (1997). Neuropsychological characteristics of the syndrome of mesial temporal lobe epilepsy. Arch Neurol, 54, 369–76. Hermanns, G., Noachtar, S., Tuxhorn, I., Holthausen, H., Ebner, A. and Wolf, P. (1996). Systematic testing of medical intractability for carbamazepine, phenytoin, and phenobarbital or primidone in monotherapy for patients considered for epilepsy surgery. Epilepsia, 37, 675–9. Hornykiewicz, O. (1988). Neurochemical pathology and the etiology of Parkinson disease. Mt Sinai J Med, 55, 11–20. Jokeit, H. and Ebner, A. (1999). Long term effects of refractory temporal lobe epilepsy on cognitive abilities: a cross sectional study. J Neurol Neurosurg Psychiatry, 67, 44–50. Jokeit, H., Seitz, R.J., Markowitsch, H.J., Neumann, N., Witte, O.W. and Ebner, A. (1997). Prefrontal asymmetric interictal glucose hypometabolism and cognitive impairment in patients with temporal lobe epilepsy. Brain, 120, 2283–94. Jokeit, H., Ebner, A., Arnold, S. et al. (1999). Bilateral depressions of hippocampal volume, glucose metabolism, and Wada hemispheric memory performance are related to the duration of mesial temporal lobe epilepsy. J Neurol, 246, 926–33.

162

H. Jokeit and A. Ebner Jokeit, H., Luerding, R. and Ebner, A. (2000). Cognitive impairment in temporal-lobe epilepsy. Lancet, 355, 1018–19. Lehrl, S. (1995). Mehrfachwahl-Wortschatz-Intelligenztest MWT-B [multiple-choice vocabulary intelligence test]. Balingen: Perimed-spitta. Lesser, R.P., Lüders, H., Wyllie, E., Dinner, D.S. and Morris, III H.H. (1986). Mental deterioration in epilepsy. Epilepsia, 27 (Suppl. 2), S105–23. Lewin, R. (1980). Is your brain really necessary? Science, 210, 1232–4. Lüders, H.O. and Awad, I. (1991). Conceptual considerations. In Epilepsy Surgery, ed. H.O. Lüders, pp. 51–62. New York: Raven Press. Mackenzie, I.R.A. and Miller, L.A. (1994). Senile plaques in temporal lobe epilepsy. Acta Neuropathol, 87, 504–10. Mackenzie, I.R.A., McLachlan, R.S., Kubu, C.S. and Miller, L.A. (1996). Prospective neuropsychological assessment of nondemented patients with biopsy proven senile plaques. Neurology, 46, 425–9. Marsh, L., Morrell, M.J., Shear, P.K. et al. (1997). Cortical and hippocampal volume deficits in temporal lobe epilepsy. Epilepsia, 38, 576–87. Mathern, G.W., Pretorius, J.K. and Babb, T.L. (1995a). Influence of the type of initial precipitating injury and at what age it occurs on course and outcome in patients with temporal lobe seizures. J Neurosurg, 82, 220–7. Mathern, G.W., Babb, T.L., Pretorius, J.K., Melendez, M. and Lévesque, M.F. (1995b). The pathophysiologic relationships between lesion pathology, intracranial ictal EEG onsets, and hippocampal neuron losses in temporal lobe epilepsy. Epilepsy Res, 21, 133–47. Mathern, G.W., Babb, T.L., Vickrey, B.G., Melendez, M. and Pretorius, J.K. (1995c). The clinical-pathogenic mechanism of hippocampal neuron loss and surgical outcomes in temporal lobe epilepsy. Brain, 118, 105–18. Mathern, G.W., Babb, T.L., Leite, J.P., Pretorius, J.K., Yeoman, K.M. and Kuhlman, P.A. (1996). The pathogenic and progressive features of chronic human hippocampal epilepsy. Epilepsy Res, 26, 151–61. Meier-Ruge, W., Hunziker, O. and Iwangoff, P. (1991). Senile dementia: a threshold phenomenon of normal aging? A contribution to the functional reserve hypothesis of the brain. Ann NY Acad Sci USA, 621, 104–18. Mori, E., Hirono, N., Yamashita, H. et al. (1997). Premorbid brain size as a determinant of reserve capacity against intellectual decline in Alzheimer’s disease. Am J Psychiatry, 154, 18–24. Multani, P., Myers, R.H., Blume, H.W., Schomer, D.L. and Sotrel, A. (1994). Neocortical dendritic pathology in human partial epilepsy: a quantitative Golgi study. Epilepsia, 35, 728–36. Pasquier, F. and Leys, D. (1997). Why are stroke patients prone to develop dementia? J Neurol, 244, 135–42. Rodin, E.A., Schmaltz, S. and Twitty, G. (1986). Intellectual functions of patients with childhood-onset epilepsy. Dev Med Child Neurol, 28, 25–33. Sadzot, B. (1997). Epilepsy: a progressive disease? Br Med J, 314, 391–2. Salmenperä, T., Kälviäinen, R., Partanen, K. and Pitkänen, A. (1998). Hippocampal damage caused by seizures in temporal lobe epilepsy. Lancet, 351, 35.

163

Risk of cognitive decline in refractory TLE Satz, P. (1993). Brain reserve capacity on symptom onset after brain injury: a formulation and review of evidence for threshold theory. Neuropsychology, 7, 273–95. Savic, I., Altshuler, L., Baxter, L. and Engel, J. Jr (1997). Pattern of interictal hypometabolism in PET scans with fludeoxyglucose F 18 reflects prior seizure types in patients with mesial temporal lobe seizures. Arch Neurol, 54, 129–36. Schmand, B., Smit, J.H., Geerlings, M.I. and Lindeboom, J. (1997). The effects of intelligence and education on the development of dementia. A test of the brain reserve hypothesis. Psychol Med, 27, 1337–44. Seidenberg, M., O’Leary, D.S., Berent, S. and Boll, T. (1981). Changes in seizure frequency and test-retest scores on the Wechsler Adult Intelligence Scale. Epilepsia, 22, 75–83. Selwa, L.M., Berent, S., Giordani, B., Henry, T.R., Buchtel, H.A. and Ross, D.A. (1994). Serial cognitive testing in temporal lobe epilepsy: longitudinal changes with medical and surgical therapies. Epilepsia, 35, 743–9. Small, G.W., Mazziota, J.C., Collins, M.T. et al. (1995). Apolipoprotein E type 4 allele and cerebral glucose metabolism in relatives at risk for familial Alzheimer disease. J Am Med Assoc, 273, 942–7. Stern, R.A., Silva, S.G., Chaisson, N. and Evans, D.L. (1996). Influence of cognitive reserve on neuropsychological functioning in asymptomatic human immunodeficiency virus-1 infection. Arch Neurol, 53, 148–53. Stern, Y., Gurland, B., Tatemichi, T.K., Tang, M.X., Wilder, D. and Mayeux, R. (1994). Influence of education and occupation on the incidence of Alzheimer’s disease. J Am Med Assoc, 271, 1004–10. Strauss, E., Loring, D., Chelune, G. et al. (1995). Predicting cognitive impairment in epilepsy: findings from the Bozeman Epilepsy Consortium. J Clin Exp Neuropsychol, 17, 909–17. Sutula, T., Cascino, G., Cavazos, J., Parada, I. and Ramirez, L. (1989). Mossy fiber synaptic reorganization in the epileptic human temporal lobe. Ann Neurol, 26, 321–30. Tasch, E., Cendes, F., Li, L.M., Dubeau, F., Andermann, F. and Arnold, D.L. (1999). Neuroimaging evidence of progressive neuronal loss and dysfunction in temporal lobe epilepsy. Ann Neurol, 45, 568–76. Tewes, U. (1991). Manual des Hamburg-Wechsler Intelligenztest für Erwachsene Revision 1991. Bern: Verlag Hans Huber. Theodore, W.H. and Gaillard, W.D. (1999). Association between hippocampal volume and epilepsy duration. Ann Neurol, 46, 800. Theodore, W.H., Bromfield, E. and Onorati, L. (1989). The effect of carbamazepine on cerebral glucose metabolism. Ann Neurol, 25, 516–20. Theodore, W.H., Bahita, S., Hatta, J. et al. (1999). Hippocampal atrophy, epilepsy duration, and febrile seizures in patients with partial seizures. Neurology, 52, 132–6. Timiras, P.S. (1995). Education, homeostasis, and longevity. Exp Gerontol, 30, 189–98. Tomlinson, B.E., Blessed, G. and Roth, M. (1970). Observations on the brains of demented old people. J Neurol Sci, 11, 205–42. Trimble, M.R. (1988). Cognitive hazards of seizure disorders. Epilepsia, 29 (Suppl. 1), S19–24. Wechsler, D. (1981). WAIS–R Manual. New York: The Psychological Corporation. Zhang, M., Katzman, R., Salmon, D. et al. (1990). The prevalence of dementia and Alzheimer’s disease in Shanghai, China: impact of age, gender, and education. Ann Neurol, 27, 428–37.

12

Behavioural and neuropsychological aspects of frontal lobe epilepsy Christoph Helmstaedter University Clinic of Epileptology Bonn, Bonn, Germany

Introduction To date, neurobiological interest in behaviour and epilepsy has been concerned primarily with temporal lobe epilepsies (TLE), and mesial temporal lobe epilepsies (mTLE) in particular. This concentration on TLE is mostly due to the fact that this type of epilepsy represents the majority of the focal epilepsies – in the Bonn series of patients with pharmacoresistant epilepsies this is about 80% – and that mTLE often forms an entity within the focal epilepsies regarding pathology (hippocampal sclerosis), a frequent history of febrile convulsions, an early onset of epilepsy, and memory problems as the prominent neuropsychological impairment. In TLE, the affected cerebral structures and epileptogenic region are mostly circumscribed, and structural pathology can be well quantified by quantitative MRI (T2 relaxometry and volumetry) or postoperative histopathological examinations of the resected specimen. Frequency, homogeneity and quantifiable pathology provide ideal prerequisites for the study of the functional and behavioural correlates of TLE. Great progress has been made during recent years, at least with respect to the neuropsychological and cognitive aspects of TLE. Recent developments in the field, however, show that it is well recognized that temporal lobe functioning involves more than memory and that its role in emotion and psychiatric symptoms is being rediscovered. The conditions we meet with frontal lobe epilepsy (FLE) are quite different. Frontal lobe epilepsies, in spite of the size of the frontal lobes, are less frequent than the temporal lobe epilepsies. In our own series, patients with FLE represent about 15% of the patients with pharmacoresistant epilepsies. Site and type of the underlying pathology are very heterogeneous. Furthermore, ictal and interictal clinicoelectric manifestations of FLE are infrequently localized, because multiple connections to most other brain areas enable fast and widely distributed propagation of epileptic activity. The functional correlates of frontal pathology in epilepsy are thus less well understood. 164

165

Behavioural and neuropsychological aspects of FLE

Frontal lobe and behaviour disorders Correspondent to its cytoarchitectonic structure, the frontal lobe is traditionally divided into two parts, which are important in two major areas. The posterior part controls motor movement and is subdivided into a premotor and a motor area, which control movement preparation and actual execution of movement respectively. The anterior part of the frontal lobe, the prefrontal cortex, is especially important in higher mental function, as in anticipation and planning, initiative, judgement, and in affect control, will power, and the determination of personality (Bechara et al., 1999; Raine et al., 2000). The prefrontal cortex can be further subdivided into the dorso-lateral cortex and the orbito-frontal cortex. This subdivision of the prefrontal cortex is still simple and it should be noted that the orbito-frontal cortex itself is a heterogeneous area connected with a wide range of other prefrontal, limbic, premotor, sensory areas and subcortical nuclei (Cavada et al., 2000). For the purpose of this article it is important to know that there is evidence that damage of the dorso-lateral part of the prefrontal cortex is more associated with impairment of executive functions and functions of working memory, whereas damage of the orbito-frontal cortex leads to impairment of the choice of behaviour, the establishment of emotional valences, and the evaluation and balancing of the past and future consequences of a given behaviour (Bechara et al., 2000; Rolls, 2000; Sarazin et al., 1998). Studies in common marmosets suggest a dissociation between the lateral and the orbital-medial division of the prefrontal cortex according to which the first selects and controls actions on the basis of higher-order rules while the latter controls different behaviour on the basis of lower-order rules (see Roberts and Wallis, 2000). The significance of the orbito-frontal cortex for social and interpersonal behaviour recently came into focus when Anderson et al. (1999) reported two patients, one with an accident at 15 months, the other with a frontal tumour resection at 16 months, both showing severe impairment of social and moral behaviour. Traditionally, behavioural dyscontrol has been attributed to temporo-limbic structures. Evidence for the involvement of the amygdala in aggression comes from human and animal stimulation studies, from activating and inhibiting effects of antiepileptic drugs on aggression, and recently also from direct correlations of amygdala volumes with aggression in patients with mesial epilepsies (Azouvi et al., 1999; Bearn and Gibson, 1998; Trimble and Van Elst, 1999; Van Elst et al., 2000). However, aggression associated with the amygdala seems more defensive than offensive in nature (Kalynchuk et al., 1999). Disinhibition phenomena, or a loss of impulse control as it is observed with frontal lesions, may be an additional prerequisite for showing impulsive aggressive behaviour. The orbito-frontal cortex as the border zone between the frontal lobe and the limbic system is thus critical for the linkage of the frontal and

166

C. Helmstaedter

limbic aspects of behavioural dyscontrol disorders. Another important area in this respect is the anterior-cingulate gyrus, which is also strongly connected to the amygdala, and whose damage has also been associated with deviant social behaviour and affect states (for overview see Devinsky et al., 1995). The finding of antisocial and aggressive behaviour with frontal lobe damage is not in itself new. A prominent and often cited example is the historic case of Phineas Gage, who after an accident with a severe frontal brain injury changed from a well-mannered man into an irresponsible and convention-neglecting person (Damasio et al., 1994; Harlow, 1868). New in the study of Anderson et al. (1999) is the finding that whether patients not only display severe behavioural disorders but also fail to see the moral of the behaviour depends on the age at the lesion onset (Dolan, 1999). Consequently, the orbito-frontal cortex seems not only important for behaviour control but also for the acquisition of social and interpersonal rules. It is important to note that irresponsible, aggressive and sociopathic behaviours can occur independent of intellectual abilities, which are often well preserved in frontal lesions. Other areas in which the orbital and medial prefrontal cortex are believed to play a central role are addictive behaviour, attention-deficit hyperactivity disorder, negative emotion and major depression (Drevets and Raichle, 1996; London et al., 2000; Northoff et al., 2000; Rubia et al., 2000). Davidson et al. (2000) propose a key role of the prefrontal cortex in the regulation of emotion in violent subjects and those predisposed to violence. A theoretical basis of the role of the prefrontal cortex in the interplay of cognition and emotion has been provided with Damasio’s ‘somatic marker’ hypothesis (Damasio, 1996). This proposes that responses to external stimuli do not rely on either conditioning processes or cognition alone, but on somatic ‘marker signals’ or autonomic response sets, which determine the conscious/unconscious connection between stimulus conditions, feelings and behaviour. Epilepsy and behaviour disorders Single case reports of behavioural and personality disorders in patients with severe brain lesions often appear dramatic. However, with respect to focal epilepsies, these reports nevertheless raise the question of whether there might be parallels in the behaviour when epilepsy affects the same brain regions. With the exception of rare cases of ictal aggression, postictal confusional states or psychosis (Marsh and Krauss, 2000), behaviour and personality disorders observed in patients with FLE appear less severe. Furthermore, as with TLE, one can hardly expect to find the prototypical ‘frontal epileptic personality’ or Wesensänderung. Personality is by definition more trait than state dependent and, particularly in epilepsy, it is quite difficult to determine whether a given behaviour has trait characteristics or not.

167

Behavioural and neuropsychological aspects of FLE

Table 12.1. Factors affecting cognitive and mood states in epilepsy

States of epilepsy preictal ictal postictal interictal (seizure free after successful surgery) Seizures frequency generalization nonconvulsive status epilepticus Epileptic dysfunction local versus distant effects Lesion e.g. alien tissues vs. migration and developmental disorders (confounded with different ages at lesion/epilepsy onset) extent, location, lateralization Antiepileptic drugs positive vs. negative psychotropic effects individual incompatibility drug-induced encephalopathy intoxication

Conclusions about persistence and continuity require follow-up observations with longer time intervals. In epilepsy several factors can be discerned, which can lead to more or less reversible changes in patients’ cognitive abilities and mood states (see Table 12.1). Furthermore, although we can now look back on a long history of successful epilepsy surgery, it is still not clear to what degree the fact of having seizures is a prerequisite of behaviour and mood disorders in epilepsy. The patient with epilepsy must always be seen in his state relative to seizures, e.g. whether he is ictal, postictal or interictal. According to recent findings with regard to seizure prediction by nonlinear measures of complexity loss as recorded by intracranial EEG, significant seizure-precipitating drops in complexity, i.e. synchronization, can be recorded long before the seizure starts (Elger and Lehnertz, 1998). Accordingly, one must assume also preictal states, which would fit well with patients’ reports of increased dysphoric mood and cognitive problems before the seizure starts. Finally, since many patients can now become seizure-free on a longterm basis by epilepsy surgery, one can suggest an additional state of wellcontrolled epilepsy after successful epilepsy surgery.

168

C. Helmstaedter

Epileptic activity can affect distant brain areas and cause cognitive and behaviour problems not related to the primary lesion or epileptogenic zone (Shulman, 1984). Notwithstanding seizures and epileptic activity, one must also differentiate the underlying pathologies, which can be more or less systemic, have different ages of onset in life, and thus have different effects on brain maturation and the development of cognition and personality. We must finally consider influences of often longstanding antiepileptic medication in these patients. Antiepileptic drugs may have positive or negative psychotropic side effects, and can show incompatibilities in the individual patient (Schmitz, 1999). Interactive effects of pathology, epilepsy and treatment must be considered. Apart from idiosyncratic actions, drugs can have different effects in lesion and nonlesion patients, and they may act differently dependent on seizure control. Taking this into consideration, it will be shown in the next sections that there is nevertheless evidence of specific behavioural abnormalities in patients with FLE, which can be interpreted within a theoretical framework of frontal lobe dysfunction. This will be outlined with the example of interictal behaviour as assessed by neuropsychological examination and self-report measures concerning quality of life, everyday activities, personality and psychiatric symptoms. In addition, seizure semiology and impairment during frontal lobe seizures and frontal nonconvulsive status epilepticus will be considered, to convey an idea of what the behavioural consequences of impaired frontal lobe functions in FLE might be. The neuropsychology of frontal lobe epilepsy The development of neuropsychology in FLE is probably best reflected by Brenda Milner’s (1995) description of her evaluation of Penfield’s patient K.M., the frontal counterpart of the temporal patient H.M. This patient had a penetrating head injury in 1928, developed seizures, and underwent surgery of the anterior parts of both frontal lobes. Surgery successfully controlled the seizures and led to improved behaviour as well as improved IQ. However, when re-evaluated with the newly developed Wisconsin Card Sorting test in 1962 he showed severe impairment in flexible categorical thinking and concept formation whilst his IQ still was average. This case exemplifies how much outcome interpretation depends on both the test sensitivity and test selection. Since then there have been surprisingly few attempts to apprehend the cognitive characteristics of patients with FLE with group studies (Shulman, 1984). Unfortunately, most of the data from Milner’s era stem from patients after surgery and thus tell us more about frontal lobe lesions than about FLE. Furthermore, most earlier studies focused on single functions, more or less following a rather monistic view of a frontal ‘central executive’ (Baddeley and Hitch, 1974). Major impair-

169

Behavioural and neuropsychological aspects of FLE

ments indicated by these studies are problems in concept formation, response inhibition (Milner, 1964), estimations (Smith and Milner, 1984), conditional associative learning (Petrides, 1985; Petrides and Milner, 1982), and profit from information provided in advance in choice reaction tasks (Alivisatos and Milner, 1989). Focusing on memory, Delaney et al. (1980) found no differences in measures of memory when nonoperated patients with unilateral frontal lobe foci were compared to healthy controls. The first of our own studies found that deficits in attention are the most significant problem in patients with FLE (Kemper et al., 1992). Later systematic group studies in nonresected patients with FLE followed the theoretical suggestion of different frontal subfunctions (Stuss and Benson, 1986) and met the requirements of the manifold frontal lobe pathology by the use of a broader range of tests. These addressed different aspects of attention, motor coordination, psychomotor speed, fluency, response inhibition, conceptual formation and shift, as well as planning, guessing or estimating. Between 1996 and 1999 Upton and Thompson published a series of five articles reporting different findings on neuropsychology in their sample of 74 subjects with FLE. Using a test battery with different measures of executive functions and motor skills, they come to the conclusion that patients with epilepsy show a deficit pattern which is similar to that found in frontal lobe dysfunction in general (Upton and Thompson, 1996a, b). As compared with patients with TLE, frontal patients show poorer motor coordination, guessing, estimation and response inhibition. Similarly, we found in 23 patients with FLE that cognitive problems could be diagnosed with a broad range of 10 ‘frontal’ tasks with about double as many test parameters. The great number of test parameters, however, turned out to be highly redundant and could be statistically reduced to four relatively independent functional areas, namely ‘psychomotor speed/attention’, ‘motor coordination’, ‘working memory’, and ‘response inhibition’. These four factors explained 70% of the total variance. When compared with patients with TLE, those with FLE were characterized by prominent impairment in motor skills and response inhibition (Helmstaedter et al., 1996). Problems in speed/attention and working memory were frequent but they appeared rather nonspecific since they were also observed in the temporal lobe group. This, however, does not necessarily contradict the assumption that these are frontal functions. An imaging study of Jokeit et al. (1997) showed in this respect that in patients with TLE, prefrontal metabolic asymmetries are evident which are associated with ‘frontal lobe measures’ and intelligence. In another of our own studies we addressed the cognitive consequences of frontal lobe surgery. We evaluated 33 patients pre- and postoperatively and were able to confirm the impairment pattern of impaired motor skills and response inhibition. We also showed that frontal surgery does not cause considerable additional damage

170

C. Helmstaedter

if eloquent cortex (SMA; motor and language area) is spared. However, when surgery included resection of the SMA the most prominent neuropsychological symptom besides neurological deficits directly after surgery was a SMA deficiency syndrome (impairment of initiation) with aphasia (speech arrest and transcortical aphasia) (Helmstaedter et al., 1998). Additional psychomotor slowing was observed in lobectomies as compared to lesionectomies. Looking closer at clinical variables which might explain the impairment pattern in nonresected patients, no consistent picture emerges. According to Upton and Thompson (1997a), seizure frequency and the duration of epilepsy have an effect on performance, but this appears to be a nonspecific effect rather than a consistent finding over different tests. With the exception of motor skills, which were spared in early right-sided FLE, no systematic effect of the assumed influence of the age at the onset of epilepsy on cognitive development could be concluded from their data (Upton and Thompson, 1997b). The impact of having epileptic seizures on cognition can be demonstrated by our postoperative findings, indicating that in seizure-free patients adjacent functions recovered after surgery. Comparable release effects have been also reported after temporal lobe surgery (Hermann et al., 1988). However, one should not go so far as to conclude that all deficits are due to epileptic dysfunction and are thus reversible, as has been suggested by Boone et al. (1988) in a single case report. In conclusion, with regards to the neuropsychological findings in FLE, it appears that indeed different frontal subfunctions can be differentiated. Nevertheless, the measures which characterize FLE have in common the demand of adequate response selection and initiation and response inhibition. This holds for tests which explicitly assess interferences and response inhibition, but also for testing of motor skills or working memory. Ending up again with a unique central executive function, one may hypothesize that the particular problem in FLE is the impairment of response selection/initiation/inhibition with varying emphasis depending on different functional areas. Which area is affected then depends on the type and the localization of the underlying lesion, including the possibility that symptoms are overshadowed by spreading epileptic dysfunction. Besides this, it is important to mention that the development of appropriate test instruments for the assessment of frontal lobe dysfunction is not yet complete and still represents a challenge for neuropsychologists. Most psychometric tests which allow quantification of test behaviour provide patients with a clear structure for behaviour, i.e. with test instructions, rules, time constraints etc. This enables the patient to behave in an ordered way and real problems with behaviour organization arising from frontal pathology are easily overlooked. If provided with greater freedom and demands on spontaneous interactions with complex situations, the same patient would otherwise reveal deficits. A possible solution to this dilemma

171

Behavioural and neuropsychological aspects of FLE

Table 12.2. Ictal frontal seizure semiology (n ⫽ 15)

Localization

Positive symptoms

Motor area

Nearly 1:1 manifestation of seizure activity in myoclonic and tonic or clonic motor activity

SMA (speech and motor area)

Tonic posturing

Premotor

Contraversive head and eye movements

Prefrontal (including gyrus cinguli)

Explosive and complex motor automatisms (including vocalizations); bizarre and hysterical behaviour; mood change Negative symptoms

Mesial propagation and secondary generalization

Loss of consciousness

Impaired executive control ‘Pathological excitation and disinhibition’

could be to design tasks which evoke spontaneous behaviour and decisions which have to be made by the patient, as has been done by Bechara et al. (1998) with the gambling task, by Goldberg and Podell (2000) with their Cognitive Bias Task, or by Upton and Thompson (1999) with their Twenty Questions Task. Ictal behaviour in frontal lobe seizures: ‘positive’ and ‘negative’ phenomena Like others before us, we recently analysed seizure phenomena in patients with FLE by video-EEG monitoring. The main purpose was to get hints from seizures for differential diagnosis. On the other hand, seizures can be studied in terms of transient dysfunctions which are more or less localized and seizure semiology; preserved functions, as well as impaired functions, can tell us something about the cerebral functional organization of cognition and consciousness. We studied ‘positive phenomena’ in terms of seizure semiology, and ‘negative phenomena’ in terms of impairment when patients were neuropsychologically tested during their seizures (Helmstaedter et al., unpublished data; Lux et al., 2000; Scherrmann and Elger, 1999). Ictal phenomena in frontal seizures are mostly positive phenomena (see Table 12.2). On the one hand, this means a nearly 1: 1 relationship between discharges and motor excitation when direct access to motor neurons is possible in primary motor area seizures, for example. On the other hand, this means release and disinhibition of

172

C. Helmstaedter

Table 12.3. Negative ictal symptoms in focal epilepsy (n ⫽ 116)

Location of seizure activity Percentage impaired when tested ictally

Frontal n⫽29

Right temporal n⫽21

Left temporal n⫽38

Bitemporal n⫽28

Orientation reflex Receptive speech (commands) Expressive speech Memory Consciousness

62 48 77 31 33

10 15 11 0 12

18 59 47 46 39

57 93 76 100 100

complex behaviours and behaviour chains when precentral areas are involved. Examples are posturing and contraversive movements in SMA and premotor seizures, and explosive, bizarre, and emotional unstable behaviours in prefrontal seizures including its mesial parts. Negative phenomena like loss of consciousness are commonly observed in seizures with mesial propagation and secondary generalization. For frontal seizures one can thus conclude that the prominent feature is impairment of executive control in terms of a pathological ‘hyper-excitation or disinhibition’. Neuropsychological examination of the cognitive impairment during seizures can provide additional insight into the ictal event. We performed ictal testing in 116 patients, most of them being candidates for epilepsy surgery. These patients underwent ictal examinations which included examination of orientation reflexes (verbal, nonverbal, tactile), expressive/receptive language (commands, naming repetition), nonverbal reception/expression (commands and imitation) and finally awareness and memory (interview after the seizure). Testing was performed by the video-EEG monitoring staff and started as soon as possible after seizure onset. Functions were tested hierarchically according to their complexity and testing was continued until the seizure ended. About half of the patients had implanted strip or depth electrodes for invasive EEG recordings. Table 12.3 shows the impairment pattern which results when the distribution of ictal EEG activity at the time of testing is considered. In comparison to lateralized and bilateral temporal lobe seizures frontal lobe seizures are characterized by prominent impairment of orientation reflexes and expressive speech, which are typical frontal functions. Receptive speech is often preserved. Patients can try, for example, to follow body commands even when they appear involved in excessive motor activity. In contrast to left and bitemporal seizures, consciousness (awareness of any kind) and memory for the test situation during the seizures is mostly preserved.

173

Behavioural and neuropsychological aspects of FLE

Table 12.4. Ictal symptoms in frontal nonconvulsive status epilepticus (n⫽ 5)

Performance

Impairment

Motor functions (including speech)

Generally reduced activity Rarely, automatisms (fumbling etc. . . .)

Orientation

Fluctuating

Executive functions (including language)

No self-initiated directed actions Increased perseverations Intrusions Apractic signs in object use and imitation

Reasoning

Problems with concept formation and shift (colour/form . . .)

Working memory

Only impaired when complex mental information processing is required

Emotion

Emotional instability (dysphoric) Impaired executive control ‘Pathological inhibition’

A very interesting behaviour and neuropsychological pattern of impairment can be observed in patients with frontal nonconvulsive status epilepticus. It is important to note that in contrast to a generalized tonic-clonic status, which is the repetition of the same seizure, the nature of frontal nonconvulsive status and frontal seizures is completely different. In contrast to frontal lobe seizures, seizure semiology of nonconvulsive status is dominated by negative seizure phenomena. Without EEG recording the epileptic nature of this state is easily overlooked, and patients appear rather strange, since they are slowed, dysphoric, morose and difficult. When neuropsychologically examined during the seizure we found in five cases consistently generally reduced activity, fluctuating orientation, reflexive and no self-initiated behaviour, perseverations, intrusions, apractic signs, problems to shift between tasks, impaired working memory on a higher cognitive level and emotional instability (see Table 12.4). In 1997 we described a single patient with a nonconvulsive status epilepticus who showed a generalized EEG pattern but focal cognitive deficits when neuropsychologically tested during this state. Today, with better diagnostic tools, we would probably be able to reinterpret this case as frontal nonconvulsive status (Bauer et al., 1997). Contrasting frontal seizures to frontal nonconvulsive status, one could interpret the latter in terms of an impaired executive control by pathological ‘hyperinhibition’. Impressive recovery to normal behaviour can be observed in these

174

C. Helmstaedter

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Figure 12.1. Preictal baseline measures and postictal course of verbal memory and decision times in patients with frontal and left or right temporal lobe epilepsy. The bars indicate performance of healthy subjects when tested repeatedly in the same intervals.

patients when the status is successfully cessated by injection of diazepam. This is one form of state-dependent cognitive impairment, as discussed in more detail in Chapter 6 by Besag. Postictally, patients, since they often do not lose consciousness during frontal lobe seizures, are quickly reoriented. Figure 12.1 shows the course of verbal memory and decision times in pre- and postictal memory testing after frontal lobe seizures as compared with left/ right temporal seizures and repeated testing in healthy controls. After lateralized temporal lobe seizures, material-specific memory impairment can be observed for at least one hour after complete reorientation. What is shown for left temporal patients in verbal memory in Figure 12.1 has its counterpart for right temporal patients in figural memory. As for frontal lobe seizures, it is remarkable that there is no postictal deterioration in memory nor significant slowing down of reaction times. However, when seizures secondarily generalize, lasting memory impairment can be observed following frontal seizures (Helmstaedter et al., 1994). We can conclude so far that from frontal lobe seizures a dysexecutive syndrome results, with mostly preserved awareness and consciousness, only reflexive but no self-initiated behaviour, and a seizure semiology, which is dominated by a state of hyperexcitation and disinhibition or hyperinhibition. This would confirm the impression from neuropsychological findings that the major problem in FLE is

175

Behavioural and neuropsychological aspects of FLE

appropriate response selection/initiation and inhibition of behaviour. A further differentiation according to lesions or foci within particular sites of the frontal lobes can be suggested but has not yet been proven. From a neuropsychological point of view it is still difficult to decide whether one central executive function or different executive functions should be assumed. As already mentioned a compromise is favoured at the moment, which suggests that the frontal subfunctions are constituted by similar processes of response selection/initiation and inhibition in different domains and modalities of behaviour. Behavioural correlates of frontal lobe epilepsy If we propose problems with behaviour selection/initiation and inhibition as a functional complex which is mainly affected in frontal lobe epilepsy, the obvious question is whether or not this dysfunction has a correlate in personality and behaviour. With respect to this question we applied several self-rating scales to a group of 95 patients with either frontal (n⫽18) or mesial temporal lobe epilepsy (n⫽77). Epilepsy groups were matched regarding sex, the age at the onset of epilepsy (mean 11 years) and the duration of epilepsy (mean 24 years). The tests in use were the BPSE ‘activity subscale’ assessing frequencies of activities (Helmstaedter and Elger, 1994); depression and anxiety were assessed by the Beck Depression Inventory (BDI; Beck et al., 1981) and the Zung Self Rating Anxiety Scale/SAS (Zung, 1971); personality was assessed by the Neo Five Factor Inventory, a German version of the NEO personality inventory (Costa and McCrea, 1992). Quality of life in epilepsy was assessed by a German modified QOLIE-10 (English version: Cramer et al., 1996); and finally we evaluated education and employment in order to add some objective data. Group comparisons considering localization and lateralization of epilepsy revealed only slight differences (Table 12.5). Patients with mTLE as a trend showed poorer mood and significantly increased anxiety scores, they described themselves as more active at home, less active with respect to outdoor cultural activities, and less open for experiences than patients with FLE. It is important to note that, when compared to normative data of healthy control subjects, the result regarding outdoor cultural activities must be interpreted in a way that patients with FLE are more active than the controls and patients with mTLE. Furthermore, when compared to data of a healthy control group, the neuroticism score of patients with mTLE and the conscientiousness score of patients with FLE appeared elevated. As regards quality of life (QOL), patients were categorized as having poor QOL when they showed scores below the 25th percentile. As is shown in Figure 12.2 patients with TLE generally tended to report poorer QOL than patients with FLE.

176

C. Helmstaedter

Table 12.5. Group comparisons considering localization and lateralization of epilepsy

Scale Mood (BDI SAS) Depression Anxiety

Activities (BPSE: activity subscale) Home activities Social activities Cultural activities

Personality (NEO FFI) Neuroticism Extraversion Open to experiences Agreeableness Conscientiousness

Group

Mean

SD

Significance

FLE mTLE FLE mTLE

7.6 11.1 29.7 35.9

7.3 9.0 7.9 7.4

n.s.

FLE mTLE FLE mTLE FLE mTLE

25.3 27.8 20.2 18.7 16.3 12.8

4.9 5.7 5.3 6.1 6.5 5.4

FLE mTLE FLE mTLE FLE mTLE FLE mTLE FLE mTLE

21.4 24.7 26.2 26.0 28.6 25.1 31.6 30.2 34.9 33.2

5.4 7.4 4.3 6.2 6.6 5.3 4.6 4.2 5.6 5.3

a

a

n.s. a

n.s. n.s. a

n.s. n.s.

Notes: FLE, frontal lobe epilepsy; mTLE, mesial temporal lobe epilepsy. For decriptions of scales used, see text. a Significantly different; n.s., not significantly different.

Impaired mood, memory problems and social limitations correspond well to the features of TLE found with the other instruments in this evaluation. Our current approach to behavioural problems and personality in patients with focal epilepsies is less led by classification systems, which may be useful in idiopathic psychiatric disorders. As already mentioned in the introduction, there is a long history of personality research in epilepsy and up to now no consistent features have been discerned. So far this has been explained by the multifactorial determination of psychiatric problems in patients with symptomatic epilepsies. As

177

Behavioural and neuropsychological aspects of FLE

Figure 12.2. Quality of Life (QOL) in frontal lobe epilepsy (FLE) as compared with mesial temporal lobe epilepsy (mTLE). Values ⬍25th percentile were considered as reflecting perception of impaired QOL. Asterisks indicate significant group differences in Chi-square testing. AED, antiepileptic drug.

far as psychometric approaches are concerned previous studies of temporal lobe epilepsy mostly used the MMPI (Rose et al., 1996) or more specifically the Bear–Fedio Inventory (Bear et al., 1982; Devinsky and Najjar, 1999). It is our daily experience that commonly used psychiatric scales or psychological personality inventories largely fail to reflect objectively what seems to the examiner clearly to be an epilepsy-related change in personality or a behaviour disorder. At the moment we are evaluating our own clinical personality inventory, which was empirically designed according to a collection of behavioural problems perceived by the clinical psychological staff in the University Hospital of Epileptology in Bonn, Germany (Helmstaedter et al., 2000a). For preliminary analysis the questionnaire was consecutively applied to 59 patients with TLE, 17 patients with FLE, 9 patients with parieto-occipital epilepsy and 44 healthy controls. It consists of 82 questions concerning 15 different behavioural domains. The answer style is a sixstepped frequency of occurrence rating with 1⫽‘occurs not at all’ and 6⫽‘occurs very frequently’. Second order factor analysis resulted in six factors, which were interpreted as follows: 1. ‘organic personality change’ with patients reporting communication problems, emotional lability, being indecisive, susceptible to interference, perseverative and hypoactive; 2. ‘depressed mood’ including depressive mood, reduced vitality, anxiety, and insensitivity; 3. ‘addiction and obsession’ including addiction to legal and illegal substances, compulsion, and obsession;

178

C. Helmstaedter

Scale

Impairment < 1 SD

Lateralization 37 right/48 left

Localization 59 T/17 F/ 9 P

Sex 59/70

Pathology 28 with AHS

1. Organic personality change

32%**

LEFT*

TLE/FLE*

W>M

/

2. Addiction/obsession

20%*

RIGHT*

FLE*

/

/

3. Depressed mood

28%

/

/

/

AHS (right)*

4. Extraversion

32%**

LEFT/RIGHT*

PLE*

/

/

5. Aggression

31%**

LEFT*

/

/

/

6. Adaptivity/hyperactivity

29%*

/

FLE*

/

/

*P < 0.05, **P < 0.01 (significantly different from control subjects)

Figure 12.3. Results obtained with the clinical personality questionnaire. TLE, temporal lobe epilepsy; FLE, frontal lobe epilepsy; PLE, parieto-occipital epilepsy.

4. ‘extraversion’ comprised sociability, curiosity and self-determined behaviour; 5. ‘aggression’ comprised aggression, sensation-seeking, nonadaptive behaviour and violence; 6. ‘hyperactivity and adaptivity’. When taking clinical data as well as sex as independent variables some interesting and comprehensible results could be obtained (Figure 12.3). The data first of all indicate that problems in the respective areas are evident in 20–30 % of patients. Organic personality changes are preferentially seen in left epilepsies of either origin, in women more than in men. Addiction and obsession are more frequent in right epilepsies and in frontal epilepsies in particular. Depressed mood is preferentially seen in patients with hippocampal sclerosis, a finding which is in line with one of our recent publications (Quiske et al., 2000). All patients and patients with parietal epilepsies in particular show reduced extraversion. Aggressive behaviour seems more frequent in left epilepsies, and patients with FLE show increased hyperactivity and adaptivity, which may parallel the finding of increased outdoor/cultural activities and openness for experiences. It is important to note that these results are preliminary and that larger control groups and validation studies are still required. However, the data indicate that the often-cited depressive mood is not the only behavioural problem in patients with focal epilepsy, and that apart from this there are specific behavioural aspects which appear related to localized and lateralized lesions or epileptic dysfunctions. Although no differences between patients with FLE and TLE were obtained, it is worth reporting the results with regard to the organic personality change scale in more detail. As shown in Figure 12.4 for selected items, about 20% of the patients report that they offend others; between 20 and 35% of the patients report problems

179

Behavioural and neuropsychological aspects of FLE Organic personality change

Offends others

18%

Receptive problems

23%

Perseverative

26%

Misunderstandings

33%

Circumstantial

34%

Irritates others

0%

51%

10%

20%

30%

% deviant ratings > 75% percentile as compared with healthy controls

Figure 12.4. Items extracted out of the ‘organic personality change’ scale of the clinical personality inventory. Bars represent the per cent of patients with focal epilepsies reporting increased problems in communication and interpersonal contact.

with reception, misunderstandings, that they were perceived as perseverative or circumstantial; and 50% report that their behaviour irritates others. This is similar to the ‘epileptic personality’, and taken together with the depressed mood, one might well think of the dysphoric and paroxysmal mood disorder as it has been proposed from a more psychiatric point of view (Blumer, 2000). Academic achievement and employment in frontal lobe epilepsy From patients with frontal lesions, it is well known that they may show unimpaired cognitive functions but nevertheless fail on everyday demands of job and career because of behavioural problems, unsteadiness, concentration problems, increased susceptibility to interference and problems with timing and planning. Subjective data may not detect behavioural problems because patients with frontal lobe lesions have been reported to underestimate their impairments. With school achievement and employment, however, we have indirect markers, which allow us to infer how far patients are adapted to everyday life. As indicated in Table 12.6 it is not the group with FLE but the one with mTLE which is less educated, and the job situation is comparable in both groups.

180

C. Helmstaedter

Table 12.6. Academic achievement, employment and epilepsy

Academic achievement level %

FLE (n⫽18) mTLE (n⫽83)

No regular school

Lowa (Hauptschule)

Medium (Realschule)

Higha (Gymnasium)

Percentage employed

17 10

22 54

22 21

39 15

68 59

Note: Chi-square: significant difference.

a

Trait or state As shown above, patients with FLE have behavioural disorders which appear very mild when compared with those reported in patients with frontal mass lesions. With respect to mood disorders, they appear less affected than patients with TLE, and they also show better academic achievement. The finding that hyperactivity, addiction and obsession might be a behavioural feature of FLE is of great interest, and can be discussed as reflecting frontal dysfunction in general and to be in line with the behaviour observed in neuropsychological examination and during seizures. The question which remains open, is how consistent the behaviour in focal epilepsies is over time. We can not yet give a conclusive answer to this question on the basis of long-term follow-up observations. The impact of epilepsy and seizures on behaviour, however, can be estimated by comparisons of patients who after surgery still have seizures with those who became completely seizure-free. We therefore analysed data from surgically and nonsurgically treated patients who participated in a longterm follow-up study, which was originally designed to show the cognitive development of these patients over time (Helmstaedter et al., 2000b). However, at the time of the long-term follow-up visit we also assessed depression by use of the BDI and quality of life by use of a German modified QOLIE–10. For the present purpose we extracted from the total database only the data of the patients with temporomesial epilepsy and hippocampal sclerosis as compared with those with FLE. Fiftyseven patients had mTLE with hippocampal sclerosis (27 had surgery, 20 were treated conservatively) and 30 patients had FLE (16 had surgery and 14 were treated conservatively). Taking depression and quality of life measures as the dependent variables in a multivariate analysis with consideration of surgery, localization and lateralization of epilepsy as independent variables, and age and the follow-up interval (mean 56 months; 2–10 years) as covariates, seizure outcome turned out to be the only significant predictor. Only 14% of the seizure-free patients in contrast to 51% of those who still had seizures showed elevated depression scores greater than

181

Behavioural and neuropsychological aspects of FLE

the cut-off score of 12 points. It should be noted that 14% is much less than the normally reported 30% of patients with focal epilepsy and depressive mood, and that 51% clearly exceed this number. Comparably, 45% of the seizure-free patients reported good quality of life with QOLIE–10, as compared to only 11% of the patients who continued to have seizures. Although these data are not follow-up data and although depression and quality of life represent only two facets of the whole range of behaviour, these data show quite impressively what a difference the presence or absence of seizures can make. The finding parallels recent findings in children who after successful epilepsy surgery show marked improvement in behaviour disorders (Lendt et al., 2000). Long-term follow-up studies on personality and behaviour disorders are thus badly needed to complete our understanding of the interaction between brain damage, epilepsy and behaviour. Conclusion We can conclude that in FLE, ‘frontal dysfunctions’ can be suggested which characteristically become evident in cognition, seizures and behaviour. The main common feature of the behavioural problems met in FLE concerns behaviour control in terms of response selection/initiation and inhibition. The domains in which these problems become apparent may vary with clinical features. Following our own findings, hyperactivity, conscientiousness, obsession and addiction can be seen as reflecting frontal lobe dysfunction in frontal lobe epilepsy. Depression, anxiety, neuroticism, cognitive (memory) impairment and social limitations in contrast seem more a feature of mTLE. However, methodological difficulties regarding the adequacy of the clinical measures in use as well as confounding effects of lesions, epileptic dysfunction, antiepileptic drugs and psychosocial status do not yet allow further distinctions such as can be made, for example, in neurobiological models about the frontal lobes and behaviour. Full-blown personality disorders are very rare in FLE and symptoms appear rather mild when compared with patients with mass lesions. As regards the state/trait discussion in epilepsy, the effects of seizure control indicate that a great part of the observed behavioural problems is indeed state-dependent. However, follow-up evaluations are badly needed to partial out the contribution of lesions and epileptic dysfunctions to behaviour disorders and to demonstrate how far these are reversible or not.

R E F E R E N C ES Alivisatos, B. and Milner, B. (1989). Effects of frontal or temporal lobectomy on the use of advance in information in a choice reaction time task. Neuropsychologia, 27, 495–503.

182

C. Helmstaedter Anderson, S.W., Bechara, A., Damasio, H. et al. (1999). Impairment of social and moral behavior related to early damage in human prefrontal cortex. Nat Neurosci, 2, 1032–7. Azouvi, P., Jokic, C., Attal, N., Denys, P., Markabi, S. and Bussel, B. (1999). Carbamazepine in agitation and aggressive behavior following severe closed head injury: results of an open trial. Brain Injury, 13, 797–804. Baddeley, A.D. and Hitch, G. (1974). Working memory. In Recent Advances in Motivation and Learning, Vol. 8, ed. G.A. Bower, pp. 47–90. New York: Academic Press. Bauer, J., Helmstaedter, C. and Elger, C.E. (1997). Nonconvulsive status epilepticus with generalized ‘fast activity’. Seizure, 6, 67–70. Bear, D., Levin, K., Blumer, D., Chetham, D. and Ryder, J. (1982). Interictal behaviour in hospitalized temporal lobe epileptics: relationship to idiopathic psychiatric syndromes. J Neurol Neurosurg Psychiatry, 45, 481–8. Bearn, R.G. and Gibson, R.J. (1998). Aggressive behavior in intellectually challenged patients with epilepsy treated with lamotrigine. Epilepsia, 39, 280–2. Bechara, A., Damasio, H., Damasio, A.R. and Lee, G.P. (1999). Different contribution of the human amygdala and ventromedial prefrontal cortex to decision making. J Neurosci, 19, 5473–81. Bechara, A., Damasio, H. and Damasio, A.R. (2000). Emotion, decision making and the orbitofrontal cortex. Cereb Cortex, 10, 295–307. Bechara, A., Damasio, H., Tranel, D. and Anderson, S.W. (1998). Dissociation of working memory from decision making within the human prefrontal cortex. J Neurosci, 18, 428–37. Beck, A.T., Rush, A.J., Shaw, B.F. and Emery, G. (1981). Kognitive Therapie der Depression. München: Urban und Schwarzenberg. Blumer, D. (2000). Dysphoric disorders and paroxysmal affects: recognition and treatment of epilepsy related disorders. Harv Rev Psychiatry, 8, 8–17. Boone, K.B., Miller, B.L., Rosenberg, L., Durazo, A., McIntyre, H. and Weil, M. (1988). Neuropsychological and behavioural abnormalities in an adolescent with frontal lobe seizures. Neurology, 38, 583–6. Cavada, C., Compañy, T., Tejedor, J., Cruz-Rizzolo, R.J. and Reinoso-Suárez, F. (2000). The anatomical connections of the macaque monkey orbitofrontal cortex. A review. Cereb Cortex, 10, 220–42. Costa, P.T. and McCrea, R.R. (1992). Revised NEO Personality Inventory (NEO PI-R) and the NEO Five Factor Inventory. Professional Manual. Odessa, FL: Psychological Assessment Resources. Cramer, J.A., Perrine, K., Devinsky, O. and Meador, K. (1996). A brief questionnaire to screen quality of life in epilepsy the QOLIE–10. Epilepsia, 37, 577–82. Damasio, A.R. (1996). The somatic marker hypothesis and the possible functions of the prefrontal cortex. Philos Trans R Soc Lond B Biol Sci, 351, 1413–20. Damasio, H., Grabowski, T., Frank, R., Galaburda, A.M. and Damasio, A.R. (1994). The return of Phineas Gage: clues about the brain from the skull of a famous patient. Science, 20/264, 1102–5. Davidson, R.J., Putnam, K.M. and Larson, C.L. (2000). Dysfunction in the neural circuitry of emotion regulation – a possible prelude to violence. Science, 289, 591–4. Delaney, R.C., Rosen, A.J., Mattson, R.H. and Novelly, R.A. (1980). Memory function in focal

183

Behavioural and neuropsychological aspects of FLE epilepsy: a comparison of non-surgical, unilateral temporal lobe and frontal lobe samples. Cortex, 16, 103–17. Devinsky, O. and Najjar, S. (1999). Evidence against the existence of a temporal lobe syndrome. Neurology, 53, S13–25. Devinsky, O., Morrell, M.J. and Vogt, B.A. (1995). Contributions of anterior cingulate cortex to behavior. Brain, 118, 279–306. Dolan, R. (1999). On the neurology of morals. Nature Neurosci, 2, 927–9. Drevets, W.C. and Raichle, E. (1996). Positron emission tomographic imaging studies in human emotional disorders. In The Cognitive Neurosciences, ed. M.S. Gazzaniga, pp. 1153–62. Massachusetts: MIT Press. Elger, C.E. and Lehnertz, K. (1998). Seizure prediction by non-linear time series analysis of brain electrical activity. Eur J Neurosci, 10, 786–9. Goldberg, E. and Podell, K. (2000). Adaptive decision making, ecological validity, and the frontal lobes. J Clin Exp Neuropsychol, 22, 56–68. Harlow, J.M. (1868). Recovery from passage of an iron bar through the head. Mass Med Soc Publ, 2, 327–46. Helmstaedter, C. and Elger, C.E. (1994). Cognitive-behavioral aspects of quality of life in presurgical patients with epilepsy. J Epilepsy, 7, 220–31. Helmstaedter, C., Elger, C.E. and Lendt, M. (1994). Postictal courses of cognitive deficits in focal epilepsies. Epilepsia, 35, 1073–8. Helmstaedter, C., Kemper, B. and Elger, C.E. (1996). Neuropsychological aspects of frontal lobe epilepsy. Neuropsychologia, 34, 399–406. Helmstaedter, C., Gleißner, U., Zentner, J. et al. (1998). Neuropsychological consequences of epilepsy surgery in frontal lobe epilepsy. Neuropsychologia, 36, 333–41. Helmstaedter, C., Gleißner, U. and Elger, C.E. (2000a). Clinical Personality Scales (CPS) in focal epilepsy: preliminary results. Epilepsia, 42 (Suppl.), 236. Helmstaedter, C., Kurthen, M., Lux, S. et al. (2000b). Long-term clinical, neuropsychological, and psychosocial follow-up in surgically and nonsurgically treated patients with drug-resistant temporal lobe epilepsy. Nervenarzt, 71, 629–42. Hermann, B.P., Wyler, A.R. and Richey, E.T. (1988). Wisconsin card sorting test performance in patients with complex partial seizures of temporal lobe origin. J Clin Exp Neuropsychol, 10, 467–76. Jokeit, H., Seitz, R.J., Markowitsch, H.J., Neumann, N., Witte, O.W. and Ebner, A. (1997). Prefrontal asymmetric interictal glucose hypometabolism and cognitive impairment in patients with temporal lobe epilepsy. Brain, 120, 2283–94. Kalynchuk, L.E., Pinel, J.P. and Treit, D. (1999). Characterization of the defensive nature of kindling-induced emotionality. Behav Neurosci, 113, 766–75. Kemper, B., Helmstaedter, C. and Elger, C.E. (1992). Kognitive Profile von prächirurgischen Patienten mit Frontal- und Temporallappenepilepsie. In Epilepsie ’91, ed. D. Scheffner, pp. 345–50. Reinbeck: Einhorn Presse Verlag. Lendt, M., Helmstaedter, C., Kuczaty, S., Schramm, J. and Elger, C.E. (2000). Behavioural disorders in children with epilepsy: early improvement after surgery. J Neurol Neurosurg Psychiatry, 69, 739–44.

184

C. Helmstaedter London, E.D., Ernst, M., Grant, S., Bonson, K. and Weinstein, A. (2000). Orbitofrontal cortex and human drug abuse: functional imaging. Cereb Cortex, 10, 334–42. Lux, S., Helmstaedter, C., Kurthen, M., Hartje, W. and Elger, C.E. (2000). Localizing value of ictal neuropsychological deficits in focal epilepsies. Epilepsia, 41 (Suppl. 7), 153. Marsh, L. and Krauss, G.L. (2000). Aggression and violence in patients with epilepsy. Epilepsy Behav, 1, 160–8. Milner, B. (1964). Some effects of frontal lobectomy in man. In The Frontal Granular Cortex, ed. J.M. Warren and K. Akert, pp. 313–34. New York: McGraw-Hill. Milner, B. (1995). Aspects of human frontal lobe function. In Advances in Neurology, Vol. 66, ed. H.H. Jasper, S. Riggio and P.S. Goldman-Rakic, pp. 67–84. New York: Raven Press. Northoff, G., Richter, A., Gessner, M. et al. (2000). Functional dissociation between medial and lateral prefrontal cortical spatiotemporal activation in negative and positive emotions: a combined fMRI/MEG study. Cereb Cortex, 10, 93–107. Petrides, M. (1985). Deficits on conditional associative learning tasks after frontal- and temporal lobe lesions in man. Neuropsychologia, 5, 601–14. Petrides, M. and Milner, B. (1982). Deficits on subject-ordered tasks after frontal and temporal lobe lesions in man. Neuropsychologia, 3, 249–62. Quiske, A., Helmstaedter, C., Lux, S. and Elger, C.E. (2000). Depression in patients with temporal lobe epilepsy is related to mesial temporal sclerosis. Epilepsy Res, 39, 121–5. Raine, A., Lencz, T., Bihrle, S., LaCasse, L. and Coletti, P. (2000). Reduced prefrontal gray matter volume and reduced autonomic activity in antisocial personality disorder. Arch Gen Psychiatry, 67, 119–27. Roberts, A.C. and Wallis, J.D. (2000). Inhibitory control and affective processing in the prefrontal cortex: neuropsychological studies in the common marmoset. Cereb Cortex, 10, 252–62. Rolls, E.T. (2000). The orbitofrontal cortex and reward. Cereb Cortex, 10, 284–94. Rose, K.J., Derry, P.A. and McLachlan, R.S. (1996). Neuroticism in temporal lobe epilepsy: assessment and implications for pre- and postoperative psychosocial adjustment. Epilepsia, 37, 484–91. Rubia, K., Overmeyer, S., Taylor, E. et al. (2000). Functional frontalization with age: mapping neurodevelopmental trajectories with fMRI. Neurosci Biobehav Rev, 24, 13–19. Sarazin, M., Pillon, B., Giannakopoulos, P., Rancurel, G., Samson, V. and Dubois, B. (1998). Clinicometabolic dissociation of cognitive functions and social behavior in frontal lobe lesions. Neurology, 51, 142–8. Scherrmann, J.M. and Elger, C.E. (1999). Generation of seizure phenomena in the frontal and temporal lobe. Epilepsia, 40 (Suppl. 7), 102. Schmitz, B. (1999). Psychiatric syndromes related to antiepileptic drugs. Epilepsia, 40 (Suppl. 10), 65–70. Shulman, M.B. (1984). The frontal lobes, epilepsy, and behavior. Epilepsy Behav, 1, 384–95. Smith, M.L. and Milner, B. (1984). Differential effects of frontal-lobe lesions on cognitive estimation and spatial memory. Neuropsychologia, 22, 697–705. Stuss, T.D. and Benson, T. (1986). The Frontal Lobes. New York: Raven Press. Trimble, M.R. and Van Elst, L.T. (1999). On some clinical implications of the ventral striatum and the extended amygdala. Investigations of aggression. Ann NY Acad Sci USA, 29, 638–44.

185

Behavioural and neuropsychological aspects of FLE Upton, D. and Thompson, P.J. (1996a). General neuropsychological characteristics of frontal lobe epilepsy. Epilepsy Res, 23, 169–77. Upton, D. and Thompson, P.J. (1996b). Epilepsy and the frontal lobes: neuropsychological characteristics. J Epilepsy, 9, 215–22. Upton, D. and Thompson, P.J. (1997a). Neuropsychological test performance in frontal-lobe epilepsy: the influence of aetiology, seizure type, seizure frequency and duration of disorder. Seizure, 6, 443–7. Upton, D. and Thompson, P.J. (1997b). Age at the onset and neuropsychological function in frontal lobe epilepsy. Epilepsia, 38, 1103–13. Upton, D. and Thompson, P.J. (1999). Twenty question tests and frontal lobe dysfunction. Arch Clin Neuropsychol, 14, 203–16. Van Elst, L.T., Woermann, F.G., Lemieux, L., Thompson, P.J. and Trimble, M.R. (2000). Affective aggression in patients with temporal lobe epilepsy: a quantitative MRI study of the amygdala. Brain, 123, 234–43. Zung, W.W.K. (1971). A rating instrument for anxiety disorders. Psychosomatics, 12, 371–9.

Part IV

Nonepileptic attacks

13

Epilepsy, dissociation and nonepileptic seizures Richard J. Brown Institute of Neurology, London, UK

Introduction Within the context of psychiatry and neurology, the term dissociation is extremely difficult to define satisfactorily. Since its introduction in the late nineteenth century, the term has been applied to a wide range of neurological, psychiatric and psychological phenomena. As a result, there is considerable confusion over what actually constitutes dissociation and the concept is frequently misapplied. This is particularly true within the field of epilepsy. Many epileptic phenomena have been labelled as dissociative, including the sensory, affective and cognitive features of partial seizures, behavioural automatisms, postictal amnesia and fugue (Thomas and Trimble, 1997). Similarly, certain psychiatric phenomena that mimic epileptic events, so-called ‘nonepileptic seizures’, have been identified as primarily dissociative in nature. Indeed, many authorities have argued that the dissociative nature of nonepileptic seizures could provide the basis for their conceptual and practical differentiation from ‘genuine’ epileptic events (Kuyk et al., 1999). If, however, dissociation is experienced both by individuals with epilepsy and by those with pseudo-epileptic seizures, how can its occurrence aid in the differential diagnosis of these conditions? In this chapter, I will explore what we mean by dissociation and how it relates to the phenomena of epilepsy and nonepileptic seizures. In this way, I aim to demonstrate that (i) many epileptic phenomena previously labelled as dissociative should not be regarded as dissociative at all; (ii) dissociation is a fundamental aspect of nonepileptic seizures; (iii) those epileptic phenomena that are genuinely dissociative involve a different type of dissociation to that involved in nonepileptic seizures; and (iv) the concept of dissociation needs to be used far more precisely if it is to help distinguish between epileptic and nonepileptic seizures. What is dissociation? Cardeña (1994) has described a useful taxonomy that captures the different ways in which the concept of dissociation has been used. According to this scheme, there 189

190

R.J. Brown

Table 13.1. Classification of dissociative disorders in ICD–10 and DSM–IV

ICD–10 dissociative (conversion) disorders

DSM–IV dissociative disorders

Dissociative amnesia Dissociative fugue Dissociative motor disorders Dissociative convulsions Dissociative anaesthesia and sensory loss Dissociative stupor Trance and possession disorders Mixed dissociative (conversion) disorders Other dissociative (conversion) disorders Dissociative (conversion) disorder, unspecified

Dissociative amnesia Dissociative fugue Dissociative identity disorder Depersonalization disorder Dissociative disorder not otherwise specified

are three major facets to the dissociation construct: (i) dissociation as nonconscious or nonintegrated mental modules or systems; (ii) dissociation as an alteration in consciousness; and (iii) dissociation as a defence mechanism. Dissociation as nonintegrated mental modules or systems

This is probably the most widely used definition of dissociation and is reflected in the so-called dissociative disorders categories of current psychiatric taxonomies. According to DSM–IV, ‘the essential feature of the dissociative disorders is a disruption in the usually integrated functions of consciousness, memory, identity, or perception of the environment’ (p. 477; American Psychiatric Association, 1994). Dissociation in this sense reflects the original meaning of the term désagrégation, offered by Janet as a label for the putative process whereby previously integrated memories can become inaccessible to consciousness and control behaviour in a way that the individual is unaware of (Janet, 1889, 1924). According to Janet, such a process is the basic psychopathological mechanism underlying so-called ‘hysterical’ symptoms, phenomena that have now been reclassified within the dissociative and somatoform disorders categories in DSM–IV and ICD–10. The DSM–IV dissociative disorders category encompasses dissociative amnesia, dissociative fugue, dissociative identity disorder (formerly multiple personality disorder), depersonalization disorder and dissociative disorder not otherwise specified (see Table 13.1). A slightly broader definition is offered in ICD–10, which identifies the loss of control over bodily movements as an additional dissociative phenomenon (World Health Organization, 1992). As such, the ICD–10 dissociative (conversion) disorders category encompasses dissociative amnesia, dissociative fugue, dissociative motor disorders, dissociative convulsions, dissociative anaesthesia and sensory

191

Seizures, epilepsy and dissociation

loss, dissociative stupor, mixed dissociative (conversion) disorders, other dissociative (conversion) disorders and dissociative (conversion) disorders, unspecified (see Table 13.1). The somatoform and dissociative disorders are related in that these phenomena are characterized by symptoms that, on the face of it, resemble those that occur in certain physical conditions, but which are presumed to be psychological in origin. Dissociative and somatoform phenomena differ in that the symptoms of the former resemble those of neurological illness, while the symptoms of the latter are more akin to those encountered in internal medicine. As Kihlstrom (1994) has cogently argued, all of the phenomena identified as dissociative disorders within DSM–IV and ICD–10 are linked by the fact that each has a temporary disruption in consciousness or volition as its primary defining feature. The differences between DSM–IV and ICD–10 in their classification of the dissociative and somatoform disorders are readily apparent. First, unlike ICD–10, DSM–IV places nonepileptic attacks in the somatoform rather than the dissociative disorders category, along with other so-called ‘conversion’ phenomena, such as unexplained motor and sensory symptoms, that are identified as dissociative in ICD–10. This difference is more practical than conceptual, with DSM–IV placing greater emphasis on the importance of excluding physical illness in the differential diagnosis of these phenomena (American Psychiatric Association, 1994). Second, unlike DSM–IV, ICD–10 does not identify depersonalization as a dissociative phenomenon, due to the lack of any significant loss of control over sensation, memory or movement in this condition, and its limited effect on personal identity. Third, DSM–IV identifies a distinct category for multiple personality disorder, relabelled dissociative identity disorder in the latest edition of this scheme. In contrast, ICD–10 places multiple personality disorder in the other dissociative (conversion) disorders category, reflecting controversy over whether this syndrome is iatrogenic or culturally bound to North America. Inconsistencies aside, both DSM–IV and ICD–10 explicitly state that physical conditions such as epilepsy should be excluded in the differential diagnosis of the dissociative and somatoform disorders. In addition to the dissociative and somatoform disorders, Cardeña (1994) identifies other pathological phenomena characterized by a lack of integration between mental modules or systems that are caused by neurological rather than psychiatric events. Blindsight, a rare condition in which the sufferer displays above-chance visual discrimination despite reporting a lack of visual experience, provides one example of how normally integrated functions can become dissociated through neurological damage. Many of the unusual behaviours often displayed by patients following commissurotomy also fall within this category, as do those exhibited by individuals suffering from hemi-neglect. In each of these cases, the dissociation is between the individual’s ongoing behaviour and their introspective verbal report. Neurological phenomena of this sort are important because they demonstrate the

192

R.J. Brown

distributed, semimodular (Fodor, 1983; Kosslyn, 1994) organization of brain, mind, consciousness and behaviour. Neurological dissociations such as blindsight are superficially analogous to those observed in psychiatric instances of dissociation, such as the preservation of implicit perception1 in the context of dissociative blindness (Kihlstrom, 1994). Neurological and psychiatric dissociation differ, however, in that the former is often permanent, reflecting irreversible damage to the underlying neurological subsystems in question (Kihlstrom, 1994). Psychiatric instances of dissociation, in contrast, are thought to be the product of an alteration in the parameters governing otherwise intact psychological functions; they are, therefore, reversible by definition. Relatedly, neurological and psychiatric dissociations differ in that, unlike the former, the latter involves symptoms (e.g. ‘glove’ anaesthesia) that need not, and typically do not, relate to the actual organization of the nervous system and its many distributed components. On these grounds, it is apparent the ‘dissociation’ in these cases is an entirely different phenomenon and the two must not be confused. The idea that normally integrated psychological processes can become temporarily dissociated and exist in isolation of one another has also been cited as the basis for other, less pathological, phenomena (Cardeña, 1994; Hilgard, 1977; Woody and Bowers, 1994). Many apparently ‘hypnotic’ phenomena fall within this category, including profound amnesia, the loss of perceptual experience and complex behaviours characterized by a sense of involuntariness, all of which can be temporarily produced by appropriate suggestions in certain individuals. The extent to which similar processes are involved in these phenomena and those displayed by individuals with dissociative psychopathology has been a matter of debate since the time of Janet. Conceptually, there are good grounds to assume a common mechanism in hypnotic and dissociative phenomena (Oakley, 1999) and recent functional imaging evidence provides some support for a link between the two (Halligan et al., 2000). According to Cardeña (1994), this particular definition of dissociation has also been inappropriately applied to a number of other normal psychological phenomena. Following Hilgard (1977), the execution of complex behaviours with only minimal conscious awareness, such as the action of driving a car whilst holding a conversation, has often been identified as a dissociative phenomenon. As Cardeña has pointed out, however, the dissociation label should not simply be applied to any behaviour or psychological process that, for whatever reason, occurs without full 1

Implicit perception is evidenced when an external stimulus produces psychological effects despite not being perceived consciously (Kihlstrom, 1994). The phenomenon of blindsight provides one such example. Implicit perception is akin to the concept of implicit memory, in which behaviour is influenced by learnt information that the individual cannot consciously recall.

193

Seizures, epilepsy and dissociation

awareness. Such a practice ignores the fact that, in many such cases, the individual can bring the apparently ‘dissociated’ process into awareness by an act of selective attention. Other such cases involve ‘dissociation’ between systems or processes that one would not normally expect to operate in an integrated fashion. According to Cardeña, mental modules or systems should only be regarded as truly dissociated from one another if their dissociation is (i) in contrast to a normal state of integration; and (ii) cannot be overcome by an act of will. Dissociation as an alteration in consciousness

A second usage of the dissociation concept refers to an altered state of consciousness characterized specifically by a disengagement from the self or the environment (Cardeña, 1994). As Cardeña has pointed out, this sense of the dissociation concept should not be applied to everyday phenomena, such as daydreaming and other states of distraction, where engagement with the environment is less than complete. Rather, it should be reserved specifically for states that are regarded by the experiencing individual as qualitatively different to their normal state of awareness. Although a number of different phenomena fall within the bounds of this definition (e.g. ‘trance’ and ‘possession’ states), probably the most commonly reported are depersonalization and derealization. In depersonalization, the individual experiences a profound feeling of detachment from their thoughts, perceptions, actions and emotions, often characterized by a sense of numbness or disembodiment. In derealization, the individual experiences an intact sense of self coupled with a feeling of detachment from the external environment, which often feels unreal or at a distance. Such feelings are extremely common, frequently occurring in the context of psychiatric illnesses such as depression and anxiety; they also occur as a circumscribed problem in their own right, such as in depersonalization disorder. Although DSM–IV identifies depersonalization disorder as a dissociative phenomenon, this condition clearly relates to a different sense of dissociation than that which applies to the other members of this category; this difference further justifies the separation of depersonalization disorder from the dissociative disorders category in ICD–10. Depersonalization and derealization are also found in certain drug states (e.g. those produced by marijuana, LSD and ketamine), neurological conditions such as temporal lobe epilepsy and can occur spontaneously in the context of stress or fatigue. Dissociation as a defence mechanism

Finally, dissociation has been described as a defence mechanism that protects the individual from potentially overwhelming pain or anxiety. In many respects, this account of dissociation is indistinguishable from the Freudian concept of repression (Erdelyi, 1985). This sense of the dissociation concept is typically used to

194

R.J. Brown

describe the psychological function served by the creation of a dissociated state of consciousness, or the dissociation of mental modules or systems from one another (Cardeña, 1994). By this view, exposure to a traumatic situation may trigger the compartmentalization of memories, which preserves psychological integrity by preventing the distressing material from entering consciousness after the event. Alternatively, such traumatic exposure may spontaneously elicit a depersonalized state that prevents extreme emotion from inhibiting an appropriate behavioural response. As such, this definition of dissociation may relate to either of the definitions described previously. In both cases, dissociation of this sort could either be an acute response to an isolated traumatic event or a trait-like characteristic acquired as a result of repeated exposure to trauma. Epilepsy and dissociation Although the differential diagnosis of ICD–10 and DSM–IV dissociative disorders explicitly requires the exclusion of symptoms with an identifiable neurological basis, many of the phenomena associated with epilepsy, particularly temporal lobe epilepsy, have been regarded as dissociative in nature (Devinsky et al., 1989; Good, 1993). Indeed, ICD–10 includes a specific category for dissociative disorders due to a general medical condition, which encompasses many of the symptoms exhibited by individuals with epilepsy. The absence of such a category from DSM–IV, however, reflects doubt concerning the value of attaching the dissociative label to these phenomena. In my view, such doubt is well justified in many cases. The notion that many epileptic phenomena can be regarded as dissociative is based, to a considerable extent, on the frequent occurrence of amnesia in epilepsy (Good, 1993). Individuals with complex partial or generalized seizures typically display profound amnesia for events occurring during the ictus. Moreover, certain people experience a postictal fugue state characterized by apparently purposeful behaviour for which they are subsequently amnesic, much like dissociative fugue. Despite their prima facie resemblance to dissociative phenomena, however, these events should not be regarded as episodes of dissociation (Good, 1993). Dissociative amnesia is characterized by an inability to retrieve information that has been learnt and is present in memory despite its inaccessibility (American Psychiatric Association, 1994). Amnesia for ictal events, in contrast, reflects a disruption in normal information processing, resulting from uncontrolled neural activity, that prevents the encoding of new material during the ictus. The ictal amnesia is not a product of a retrieval failure, therefore, but simply the absence of memories to retrieve. It is for this reason that this form of amnesia is irreversible, unlike most cases of dissociative amnesia (American Psychiatric Association, 1994).

195

Seizures, epilepsy and dissociation

Postictal fugue should not be regarded as a dissociative episode for similar reasons. Unlike dissociative fugue, postictal fugue is characterized by a disruption in consciousness associated with significant confusion and an abnormal EEG (Thomas and Trimble, 1997). The apparently purposeful behaviour displayed in postictal fugue is not dissociated from ongoing cognitive activity as it is in dissociative fugue; rather, it occurs in the relative absence of such activity. The inability to reverse the amnesia associated with postictal fugue serves as an illustration of this fact. The behavioural automatisms often observed in the context of complex partial seizures and regarded as a dissociative phenomenon by some (Good, 1993) are amenable to a similar interpretation. Like the behaviours exhibited during postictal fugue, ictal automatisms only occur in the context of a disruption in consciousness and disturbed behavioural control; genuinely dissociated behaviours are noteworthy because they occur despite an otherwise intact ability to control action2. In both cases, it is likely that these behaviours result from the uncontrolled activation of circumscribed motor programs by epileptic discharges in neural sites associated with behavioural control. The fact that such automatisms are particularly characteristic of seizures originating in the frontal lobes lends support to this view. Many of the phenomena associated with partial seizures originating in the temporal lobes, such as hallucinations, sensory and cognitive auras, déjà vu and déjà veçu should also be distinguished from true episodes of dissociation. Although hallucinations and auras have a phenomenology that departs from external reality, these phenomena involve the paradoxical integration of information within conscious awareness. As such, they may be more appropriately regarded as phenomena of association rather than dissociation (Kihlstrom, 1994). The phenomenology of epileptic hallucinations and auras probably originates in the activation of representational structures in the temporal lobes, either directly by seizure activity in representational networks, or indirectly through seizure-related stimulation of limbic structures such as the amygdala and anterior cingulate (Bancaud et al., 1994). Experiences of epileptic déjà vu and déjà veçu are also more associative than dissociative and may involve a similar neurophysiological process. For example, stimulation of the amygdala by seizure-related discharges could imbue current perceptions and cognitions with an unwarranted emotional colouring that may be experienced as a sense of having encountered the situation before (Bancaud et al., 1994; Sierra and Berrios, 1998). Certain phenomena associated with temporal lobe epilepsy can, however, be regarded as genuine examples of dissociation according to the scheme described by Cardeña (1994). Depersonalization and derealization commonly occur in the 2

‘Dissociated’ behaviours should be distinguished from normal automatic behaviours because the latter are generally in accord with system goals and can therefore be considered voluntary (Brown, 1999; see also Cardeña, 1994).

196

R.J. Brown

context of temporal lobe epilepsy and involve an alteration in consciousness characterized by dissociation from the self and/or the environment. According to Sierra and Berrios (1998), depersonalization and derealization are the products of a vestigial defence mechanism evolved to provide the optimum processing conditions for adaptive behaviour in the face of threat. By this view, extreme anxiety triggers an inhibitory response from the left prefrontal cortex that dampens output from the sympathetic nervous system, through inhibition of the amygdala and anterior cingulate. In turn, the right prefrontal cortex is activated by ascending arousal systems controlled by uninhibited amygdala circuits, generating further inhibition of the cingulate. As a result, the individual experiences a sense of vigilant alertness devoid of any emotional or cognitive content, a state that is ideally adapted for the control of action in the face of extreme and potentially debilitating danger. If this response is triggered in the absence of threat, however, the resulting sense of depersonalization and derealization can itself be highly unpleasant and incapacitating. Given the validity of this account, depersonalization and derealization in the context of temporal lobe epilepsy may be the result of seizure activity in the amygdala that prevents the emotional tagging of perceptual and cognitive information prior to its entry into conscious awareness3. It may be that this process is the product of transient disconnection between the amygdala and sensory areas resulting directly from seizure activity. Alternatively, it may reflect an indirect defensive response in the face of anxiety elicited by seizure-based stimulation of the amygdala. Intuitively, one suspects that the former is the more plausible possibility, although the latter cannot be ruled out a priori. Following this account of depersonalization and derealization, these phenomena can be regarded as dissociative in sense (ii) of the term; whether they should, in the context of epilepsy, be regarded as the result of a dissociative defence mechanism remains an empirical issue. The fact that few epileptic phenomena can be regarded as dissociative in any strict sense reflects widespread confusion over what actually constitutes dissociation. Although widely endorsed, the idea that any breakdown in memory, consciousness, identity, perception or behavioural control is dissociative overextends the term and diminishes its descriptive validity (Cardeña, 1994). Amnesia cannot be considered genuinely dissociative unless it involves an inability to retrieve intact information that should, under normal circumstances, be available for recall (Kihlstrom, 1994). Amnesia resulting from a failure to encode information, including that which occurs in the context of epilepsy, does not fall within this category. Loss of behavioural control can only be considered dissociative if it is within the context of an otherwise intact ability to control action. Seizure-related motor phe3

Such a process could also be responsible for the paradoxical sense of unfamiliarity that characterizes jamais vu, a phenomenon commonly observed in epilepsy.

197

Seizures, epilepsy and dissociation

nomena, including complex automatisms, are not dissociative because they occur only in the context of reduced behavioural control in general. Current psychiatric taxonomies do not make these distinctions clearly enough, relying instead on a purely descriptive approach that precludes precise classification based on the mechanisms underlying different phenomena. Nonepileptic seizures and dissociation Epileptologists frequently encounter patients who present with paroxysmal events that, despite resembling epileptic episodes, are actually nonepileptic. Indeed, as many as 50% of patients referred to specialist epilepsy centres may turn out not to have epilepsy (Francis and Baker, 1999). While some nonepileptic seizures may be attributable to physical causes other than epilepsy (see Gates and Erdahl, 1993), a demonstrable organic basis is absent in many such cases. Of these, some are attributable to an identifiable psychiatric illness, such as psychosis, that can produce seizure-like symptoms. In other cases, however, nonepileptic seizures4 occur as an isolated psychiatric problem in their own right. Identifying such cases represents a considerable challenge to neurologists working within this domain. At present, there are few, if any, reliable criteria for an inclusive diagnosis of nonepileptic attack disorder. As a result, current diagnostic practice is essentially based on the exclusion of epilepsy and other physical disorders (see Brown and Trimble, 2000). The concept of dissociation is particularly important in relation to nonepileptic attacks as it sheds light on both the mechanisms and, potentially, the differential diagnosis of these phenomena. Although nonepileptic seizures have a presentation as diverse as that associated with epilepsy itself, it is possible to impose some order on their general semiology. In a sample of 110 patients with well-documented nonepileptic attacks, Meierkord et al. (1991) report that approximately one-third of all cases involved a collapse with limpness, while two-thirds involved prominent motor activity such as limb thrashing. In this study, decreased responsivity to verbal stimulation was evident in three-quarters of all cases, while purposeful or semi-purposeful motor behaviour 4

Much has been written about the relative merits of the various terms used to describe these phenomena. Many have argued that terms such as ‘pseudoseizures’ are pejorative because they imply that the events in question are not subjectively compelling (Betts, 1990). In contrast, the neutral alternative term ‘nonepileptic seizures’ has been criticized for being imprecise, as it fails to distinguish between the many different types of paroxysmal events that are nonepileptic (Kuyk et al., 1999). My own view is that the term ‘somatoform seizures’ would be more appropriate than either of the above, as it is nonpejorative, has descriptive precision and emphasizes the importance of excluding physical illness in the differential diagnosis of these events. For the sake of descriptive continuity, however, the terms ‘nonepileptic seizures’ and ‘nonepileptic attacks’ will be adopted here; the term ‘nonepileptic attack disorder’ will be used to describe the condition characterized by these events. In the present context, it should be assumed that the term refers specifically to those events that cannot be attributed to either an organic cause or an identifiable psychiatric illness.

198

R.J. Brown

was observed in 44%. Just over two-thirds of all cases presented with stereotyped seizures. These observations are broadly consistent with those reported by Betts and Boden (1992). In this study, three types of ‘emotional’ nonepileptic attacks were identified in addition to those deliberately simulated for primary or secondary gain and those attributable to a recognizable psychiatric attack disorder (e.g. panic disorder). Approximately 21% of these individuals experienced so-called swoon attacks, in which the individual characteristically sinks to the floor and lies inert and unresponsive for the duration of the attack. Roughly 33% displayed so-called tantrum attacks, involving a sudden drop to the floor followed by screaming and thrashing of the limbs. Finally, 46% displayed so-called abreactive attacks, involving gasping, limb-thrashing, pelvic thrusting and stiffening of the body with back arching. In all cases, nonepileptic attacks involve a temporary loss of behavioural, sensory or cognitive control that occurs in the context of intact neuropsychological functioning, as evidenced by a normal EEG during the nonepileptic ictus. The absence of paroxysmal brain discharges serves as the principal feature that distinguishes nonepileptic from ‘genuine’ epileptic events. By itself, however, the EEG cannot provide a completely reliable basis for the identification of epileptic and nonepileptic seizures (Brown and Trimble, 2000), underlining the potential value of dissociation as a criterion for an inclusive diagnosis of nonepileptic attack disorder. Several converging lines of evidence indicate that these events involve a dissociative psychological mechanism (see Kuyk et al., 1997). In the first instance, nonepileptic seizures are commonly found in the context of other forms of dissociative psychopathology. Bowman (1993) and Bowman and Markand (1996) found that the vast majority of individuals with nonepileptic attacks meet criteria for DSM–IV dissociative disorders such as dissociative amnesia, identity disturbance and depersonalization. Post-traumatic stress disorder, commonly assumed to involve a dissociative mechanism, was also particularly common in this group of patients (Bowman, 1993; Bowman and Markand, 1996). Other studies have found that nonepileptic seizures frequently occur alongside other unexplained physical symptoms (Krishnamoorthy et al., 2001; Meierkord et al., 1991), suggesting that they may be one aspect of a broader tendency to express psychological distress somatically, so-called ‘somatization’ (Lipowski, 1968). A number of authorities have suggested that dissociation is an important aspect of this phenomenon also (Nemiah, 1991). Eating disorder symptoms, which have been linked to a dissociative process (Pettinati et al., 1985), also appear to be particularly common in patients with nonepileptic seizures (Krishnamoorthy et al., 2001). The frequent co-occurrence of dissociative psychopathology in patients with nonepileptic attacks appears to indicate a general propensity for dissociative exper-

199

Seizures, epilepsy and dissociation

iences in these individuals. Consistent with this notion is a recent study by Kuyk et al. (1999), showing that individuals with nonepileptic attacks display elevated levels of hypnotic susceptibility. In a related vein, nonepileptic attacks can, in many such individuals, be provoked using suggestion, placebo or hypnosis (Dericioglu et al., 1999). High hypnotic susceptibility is commonly found in patients with dissociative psychopathology (Frischholz et al., 1992; Pettinati et al., 1985; Spiegel et al., 1988), and a dissociative interpretation of hypnosis has been offered by a number of authorities (Hilgard, 1977; Woody and Bowers, 1994). Bowman (1993) also found that individuals with nonepileptic seizures yield elevated scores on the Dissociative Experiences Scale (DES; Bernstein and Putnam, 1986), a self-report measure assessing everyday occurrences of dissociation, compared with nonclinical controls. However, in a more recent study, Alper et al. (1997) found that DES scores are also elevated in patients with complex partial seizures (see also Devinsky et al., 1989); indeed, there was no significant difference in overall DES scores between these patients and a group with nonepileptic seizures. Nevertheless, both epileptic and nonepileptic groups scored higher on the DES than typically observed in nonclinical populations. In my view, this particular finding demonstrates the danger of conflating the various definitions of dissociation within a single measure such as the DES. As the DES treats dissociation as a unitary concept, it cannot differentiate between conditions that are characterized by different forms of dissociative phenomena, such as epilepsy and nonepileptic attack disorder. It is widely thought that traumatic experiences precipitate the development of dissociative symptoms, which serve a defensive function that protects the individual from extreme anxiety and psychological disintegration. Indeed, both DSM–IV and ICD–10 make an explicit link between traumatic events and the onset of dissociative symptoms. Moreover, a number of studies have found disproportionately high rates of physical, sexual and emotional abuse in patients with dissociative disorders (Chu and Dill, 1990; Irwin, 1994; Pribor et al., 1993). As such, evidence indicating an increased prevalence of traumatic experiences in individuals with nonepileptic seizures could be viewed as additional support for a dissociative interpretation of this phenomenon. To this end, Bowman (1993) found that 70% and 77% of her sample of 27 nonepileptic seizure patients had experienced physical or sexual abuse respectively. Similarly, Betts and Boden (1992) obtained positive sexual abuse histories from 54% of 96 patients with nonepileptic seizures. Evidence implicating high dissociative comorbidity, hypnotic susceptibility and exposure to trauma in individuals with nonepileptic seizures provides only indirect evidence for a dissociative interpretation of this phenomenon. Although such evidence suggests that a tendency to dissociate may be a common feature of these individuals, it does not constitute conclusive proof that nonepileptic attacks

200

R.J. Brown

are themselves dissociative. A recent study by Kuyk et al. (1999) places such an interpretation on a firmer footing. Like ‘genuine’ epileptic seizures, nonepileptic attacks are often associated with a dense amnesia for events occurring during the ictus. As I have hopefully demonstrated, genuine epileptic amnesia should not be considered a dissociative phenomenon because it arises from a seizure-related disruption in memory encoding rather than an inability to retrieve intact memory traces. However, Kuyk et al. (1999) have shown that the amnesia associated with nonepileptic attacks may actually be the product of such a retrieval deficit. Kuyk et al. (1999) compared a group of individuals with amnesia for events occurring during well-documented nonepileptic attacks with a group displaying amnesia following complex partial and generalized epileptic seizures. All subjects were hypnotized and given suggestions designed to facilitate the recovery of ictal events; the experimenter remained blind to group status at all times. Using a free-recall paradigm, 17 out of 20 patients with nonepileptic seizures recovered significant information concerning the designated attack; this information was verified by video recordings or third-party reports. In contrast, not one of the 17 patients with epilepsy retrieved information concerning their attack during hypnosis. Such a finding appears to demonstrate that, unlike that found in epilepsy, nonepileptic amnesia results from a process that prevents the individual from accessing memories successfully encoded during the attack. This apparent separation of intact memorial information from conscious awareness following a nonepileptic attack, coupled with the phenomenological character of these events, clearly identifies these phenomena as dissociative in sense (i) of the term. Dissociative mechanisms of nonepileptic seizures

If one assumes that nonepileptic seizures are a dissociative phenomenon, what mechanisms might be involved in the generation and maintenance of these events? Current theorizing in this domain is still very much dominated by nineteenth century ideas concerning the mechanisms underlying so-called ‘hysterical’5 phenomena. Indeed, the term dissociation originates in the work of Pierre Janet who proposed one of the earliest systematic accounts of the psychological mechanisms underlying hysteria. According to Janet (1889, 1924), a fundamental weakness in the hysterical individual’s mental character makes them susceptible to a breakdown in the normally integrated functions of consciousness when faced by environmental stress or trauma. As a result, organized sets of knowledge pertaining to the 5

In recent years, the term hysteria has fallen from favour due to its unwarranted and pejorative connotations. The multiple unexplained symptoms captured by this concept, of which nonepileptic seizures are one, are now subsumed within the somatoform and dissociative disorder categories of DSM–IV and ICD–10.

201

Seizures, epilepsy and dissociation

trauma may become ‘dissociated’ from the main body of consciousness, and may serve to take control of behaviour and experience if activated by environmental events. The automatic activation of these dissociated memories results in a hysterical reaction (or ‘somnambulism’) that, in some instances, takes the form of a nonepileptic attack. According to this view, two aspects of the hysterical individual’s psychophysiological make-up are responsible for the processes of dissociation and somnambulism. First, the hysterical individual possesses an abnormally high degree of suggestibility that allows ideas from the external environment to develop within them in the absence of their own effort or awareness. Second, the hysterical individual suffers from an attentional dysfunction or ‘retraction of the field of consciousness’ (Janet, 1924, p. 314) which prevents them from entertaining alternative states of mind, thereby accentuating their responsivity to external suggestion. Although a century old, the theoretical analysis of hysteria offered by Janet continues to influence theory and research concerning dissociation and the dissociative disorders. This influence is particularly evident in Hilgard’s (1977) neodissociation theory and its conceptual derivatives (Woody and Bowers, 1994). According to Hilgard, behaviour is controlled by the operation of a large set of functionally autonomous low-level cognitive control systems specialized for the execution of particular behavioural acts. Each control system comprises an organized set of well-learned behavioural and cognitive routines or schema6 developed following extensive experience with the environment; these control systems are hierarchically organized beneath an executive ego, a higher-level cognitive structure associated with volition and consciousness. The executive ego is responsible for selecting the appropriate cognitive control systems for any given task; however, in order to conserve the limited resources of the executive, once control systems are selected they are able to function with a considerable degree of autonomy from the ego. As such, they have become dissociated from the executive, with their continued operation occurring largely outside of awareness; in this way, well-learned behaviours can be performed effortlessly and concurrently. According to Hilgard, such processing dissociations are fundamental to human cognition, with the conscious representation of information being largely unnecessary for the execution of extremely complex behaviours. Hilgard’s theory provides an elegant account of the psychological mechanisms involved in dissociation. Although it was originally developed as a general account of the mechanisms involved in behavioural control, neodissociation theory has been most influential as an account of hypnotic behaviour. According to neodissociation theory, the hypnotic induction brings about a functional inhibition of the executive ego, causing it to fractionate into two separate elements (Kirsch and 6

In contemporary cognitive theory, schemas are organized knowledge structures that represent the sequence of actions and/or processing operations involved in a given behaviour e.g. the movements involved in controlling a car.

202

R.J. Brown

Figure 13.1. The neodissociation account of dissociative behaviours. (Adapted from Kirsch and Lynn, 1995.)

Lynn, 1995). Although one part of the fractionated executive continues to function as normal during hypnosis, the second part is concealed from awareness by the formation of an amnesic barrier. This part of the ego can exert behavioural control in the usual fashion but such control is prevented from representing itself in consciousness by the amnesia (see Figure 13.1). Hypnotic behaviours result from the selection of cognitive control systems via the part of the executive ego concealed by the amnesia; as the hypnotized individual is aware of only the resulting behaviour and not the process by which it was selected (due to the amnesia), they experience the execution of hypnotic suggestions as occurring involuntarily. The neodissociative model of hypnotic behaviours also provides the basis for an account of more pathological phenomena such as dissociative amnesia, fugue and multiple personality disorder (Kihlstrom, 1994; Spiegel and Cardeña, 1991). According to such a view, the formation of amnesic barriers within the executive ego is a common defensive response in the face of trauma. These barriers serve an adaptive function in that they protect the individual from experiencing potentially overwhelming negative affect associated with the traumatic event. However, pathological dissociative amnesia can arise if the barrier within the executive ego endures to the point where the memory loss itself becomes distressing or debilitating. In the case of dissociative fugue, the amnesic barrier conceals large tracts of autobiographical memory as well as the traumatic events themselves. Without access to this autobiographical information, the fugue sufferer not only reports amnesia but also

203

Seizures, epilepsy and dissociation

a profound loss of personal identity (Kihlstrom, 1994). Neodissociation theory also postulates that different parts of the executive ego, separated from each other by amnesic barriers, can be activated at different times (e.g. during traumatic events). If a nonprimary aspect of the ego is activated often enough or for a sufficient length of time, it may become associated with enough autobiographical information to constitute a separate identity in its own right. Such a process could allow for the development of multiple identities within a single individual, a phenomenon that characterizes dissociative identity disorder. Neodissociation theory can also account for dissociative phenomena characterized by a loss of behavioural control, such as nonepileptic attacks. By this view, dissociative behaviours are controlled by that part of the executive ego concealed beneath the amnesic barrier (Kirsch and Lynn, 1995). As such, neodissociation theory would identify nonepileptic attacks as essentially voluntary behaviours; however, they are perceived as involuntary because the individual is amnesic for the executive control involved in their execution (Kihlstrom, 1994). Hilgard’s neodissociation theory represents an important step forward in our understanding of the dissociation concept. Perhaps most importantly, it recasts Janet’s original ideas within a modern cognitive psychological framework that bridges the gap between pathological (e.g. dissociative amnesia) and nonpathological (e.g. hypnotic amnesia) instances of dissociation. While Janet regarded dissociation as an abnormal process provoked by stress in the constitutionally weak mind of the hysterical individual, neodissociation theory identifies dissociation as a normal psychological process that is utilized by the organism as a defence mechanism in the face of trauma. Pathological dissociation arises when this normally adaptive process becomes generalized to situations where such a response is inappropriate. Similarly, hypnotic suggestions capitalize on the dissociative nature of brain and mind to produce unusual and entertaining phenomena. Despite its popularity within psychiatry and psychology, there are a number of problems with neodissociation theory as an account of the mechanisms underlying the dissociative disorders. In the first instance, neodissociation theory assumes that pathological dissociation results from a defensive response to trauma that has become overgeneralized. However, not all individuals who display dissociative symptomatology have been exposed to trauma. Second, neodissociation theory assumes that pathological dissociation arises from the generation of an amnesic barrier. However, amnesia is not apparent in all cases of dissociative symptomatology. Many individuals with nonepileptic seizures, for example, are not amnesic for the loss of behavioural control that they experience during the ictus. Third, neodissociation theory assumes that similar mechanisms underlie both pathological and nonpathological dissociations, such as those observed during hypnosis.

204

R.J. Brown

However, as Woody and Bowers (1994) have pointed out, the ‘amnesic-barrier’ concept that forms the basis of neodissociation theory does not provide a good explanation of hypnotic behaviours and experiences. If such a notion were correct, it would imply that all hypnotic phenomena involve some degree of spontaneous amnesia. This is, in fact, extremely uncommon. One would also expect hypnotic (i.e. suggested) amnesia to be far more common than is actually the case. Moreover, many individuals are unable to experience hypnotic amnesia despite being responsive to other hypnotic suggestions. More recently, Woody and Bowers (1994) have offered an alternative account of the dissociative mechanisms underlying hypnotic behaviours and experiences. Intrinsic to neodissociation theory is the idea that dissociative behaviours are controlled by part of an executive system that is concealed beneath an amnesic barrier. In this sense, neodissociation theory is based on the idea of dissociated experience; that is, the individual experiences a loss of behavioural control that is, in fact, illusory7 (Woody and Bowers, 1994). Woody and Bowers (1994) have proposed an alternative account based on the concept of dissociated control. By this view, the dissociation is between the executive ego and the lower-level systems controlling behaviour, rather than within the executive itself. According to Woody and Bowers (1994), this dissociation between higher- and lower-level cognitive control results from an inhibition of the executive system due to the induction of hypnosis. In this model, dissociated behaviours are generated by the automatic activation of lowerlevel behavioural routines by environmental cues (see Figure 13.2). As such, dissociated behaviours are genuinely involuntary, because they are not controlled by the executive ego. Although originally developed to account for the dissociative phenomena observed in hypnosis, dissociated control theory could also be applied to more pathological instances of dissociation, such as nonepileptic attacks. By this view, nonepileptic attacks would be the product of a learned behavioural routine triggered automatically (i.e. without input from the executive control system) by cues from the environment. The apparently involuntary and stereotyped nature of nonepileptic attacks (see Meierkord et al., 1991) certainly appears to indicate that this phenomenon is controlled by the activation of a low-level behavioural routine or schema. If this were the case, the semiology of any given nonepileptic attack would correspond to the nature of the behavioural representation underlying it. In addition to behavioural features (e.g. those representing the movements involved in the seizure), this schema could also involve a cognitive control component that serves to inhibit the operation of the executive system and the information that is passed 7

This is based on the notion that any behaviour controlled by the executive ego is volitional, regardless of whether the executive is concealed beneath an amnesic barrier. See Brown (1999) for a fuller discussion.

205

Seizures, epilepsy and dissociation

Figure 13.2. The dissociated control theory account of dissociative behaviours. (Adapted from Kirsch and Lynn, 1995.)

to it by lower-level systems. This could account not only for the behavioural aspect of nonepileptic seizures but also for the loss of behavioural control and disruption in subjective awareness that characterizes many such events. Such an account of the mechanisms involved in nonepileptic seizures raises two important questions. First, what are the origins of the behavioural routines underlying these events? The fact that individuals with nonepileptic events have often been exposed to epilepsy (either in themselves or others) suggests that some sort of behavioural modelling is involved in this process. Alternatively, it may be that nonepileptic attacks are a learned behavioural analogue of a biological reaction such as the ‘sham death’ reflex (Kretschmer, 1926), perhaps triggered previously as an acute response to threat. Second, what are the environmental cues that trigger the nonepileptic event? If one assumes that nonepileptic seizures serve something of a defensive function, it is likely that a common precipitant of nonepileptic attacks will be feelings of anxiety, or thoughts, images or memories associated with such feelings. In principle, however, triggers could be anything that is associated in memory with the behavioural representation underlying the attack; these are likely to vary from individual to individual depending largely on the circumstances surrounding the initial occurrence of the attacks. A dissociated control account of nonepileptic seizures has all the advantages of neodissociation theory and fewer problems. Like neodissociation theory, it can account for the basic features of nonepileptic seizures, but does so without recourse

206

R.J. Brown

to the much-criticized amnesic barrier notion (Brown, 1999; Woody and Bowers, 1994). Moreover, unlike neodissociation theory, a dissociated control account of nonepileptic seizures would identify them as genuinely involuntary behaviours. Such an idea is more intuitively appealing than the neodissociation theory view that dissociative behaviours are volitional but erroneously perceived as involuntary. Indeed, the idea that an individual can execute a behaviour ‘deliberately’ but not be aware of it (even though they may be paying it full attention) seems something of a contradiction in terms. One of the problems with dissociated control theory in its current form is that the direct activation of low-level cognitive systems by environmental cues is actually a fundamental aspect of routine behavioural control (Brown, 1999; Kirsch and Lynn, 1997). This is well illustrated by everyday ‘action-slips’ (Reason, 1979), such as dialling a familiar but out-of-date telephone number despite being aware that it is no longer valid. As such, one need not assume that executive inhibition is necessary for the occurrence of dissociative behaviours. This is entirely consistent with the present account of nonepileptic attacks, which posits that executive inhibition occurs after the activation of a low-level behavioural schema (i.e. as part of the schema itself), not before it. Summary Dissociation is a complex and multifaceted concept that is frequently misapplied within the field of epilepsy. In this chapter, I have explored the various components of the dissociation concept and how they relate to the phenomena of epilepsy and nonepileptic seizures. I have demonstrated why many epileptic phenomena often thought to be instances of dissociation, such as amnesia, postictal fugue, behavioural automatisms, auras and hallucinations, should not be regarded as dissociative at all. I have also argued, however, that depersonalization and derealization occurring in the context of epilepsy can be regarded as genuinely dissociative, in the sense that they involve an altered state of consciousness characterized by disengagement from the self or environment. I have also presented evidence indicating that nonepileptic attacks should be considered a dissociative phenomenon, in this case involving a temporary disruption in behavioural control and subjective awareness despite intact neuropsychological functioning. Although certain aspects of epilepsy are dissociative, therefore, they do not involve the same type of dissociation as that underlying nonepileptic attacks. If the dissociation concept is to prove useful in this area, much greater precision, both conceptual and methodological, is required. Researchers and clinicians should be explicit about which definition of dissociation they are referring to and efforts should be made to construct measures of dissociation that are, unlike the

207

Seizures, epilepsy and dissociation

DES, ‘phenomenon-pure’. Assessing the reversibility of nonepileptic amnesia, which may prove invaluable as an aid to differential diagnosis in this area, provides one illustration of the potential utility of such an endeavour.

R E F E R E N C ES Alper, K., Devinsky, O., Perrine, K. et al. (1997). Dissociation in epilepsy and conversion nonepileptic seizures. Epilepsia, 38, 991–7. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) (DSM–IV). Washington, DC: APA. Bancaud, J., Brunet-Bourgin, F., Chavel, P. and Halgren, E. (1994). Anatomical origin of déjà vu and vivid memories in human temporal lobe epilepsy. Brain, 127, 71–90. Bernstein, E. and Putnam, F.W. (1986). Development, reliability and validity of a dissociation scale. J Nerv Ment Dis, 174, 727–35. Betts, T. (1990). Pseudoseizures: seizures that are not epilepsy. Lancet, 336, 163–4. Betts, T. and Boden, S. (1992). Diagnosis, management and prognosis of 128 patients with nonepileptic attack disorder. Part I. Seizure, 1, 19–26. Bowman, E.S. (1993). Etiology and clinical course of pseudoseizures: relationship to trauma, depression and dissociation. Psychosomatics, 4, 333–42. Bowman, E.S. and Markand, O.N. (1996). Psychodynamics and psychiatric diagnoses of pseudoseizure patients. Am J Psychiatry, 153, 57–63. Brown, R.J. (1999). An Integrative Cognitive Theory of Suggestion and Hypnosis. PhD thesis, University College London. Brown, R.J. and Trimble, M.R. (2000). Dissociative psychopathology, non-epileptic seizures and neurology. J Neurol Neurosurg Psychiatry, 69, 285–91. Cardeña, E. (1994). The domain of dissociation. In Dissociation. Clinical and Theoretical Perspectives, ed. S.J. Lynn and J.W. Rhue, pp. 15–31. New York: Guilford Press. Chu, J.A. and Dill, D.L. (1990). Dissociative symptoms in relation to childhood physical and sexual abuse. Am J Psychiatry, 147, 887–92. Dericioglu, N., Saygi, S. and Ciger, A. (1999). The value of provocation methods in patients suspected of having non-epileptic seizures. Seizure, 8, 152–6. Devinsky, O., Putnam, F., Grafman, J., Bromfield, E. and Theodore, W.H. (1989). Dissociative states and epilepsy. Neurology, 39, 835–40. Erdelyi, M.H. (1985). Psychoanalysis: Freud’s Cognitive Psychology. New York: W.H. Freeman. Fodor, J.A. (1983). The Modularity of Mind. Cambridge, MA: MIT Press. Francis, P. and Baker, G.A. (1999). Non-epileptic attack disorder (NEAD): a comprehensive review. Seizure, 8, 53–61. Frischholz, E.J., Lipman, L.S., Braun, B.G. and Sachs, R. (1992). Psychopathology, hypnotizability, and dissociation. Am J Psychiatry, 149, 1521–5. Gates, J.R. and Erdahl, P. (1993). Classification of non-epileptic events. In Non-Epileptic Seizures, ed. J.R. Gates and A.J. Rowan, pp. 21–30. Oxford: Butterworth–Heinemann.

208

R.J. Brown Good, M.I. (1993). The concept of an organic dissociative syndrome: what is the evidence? Harv Rev Psychiatry, 1, 145–57. Halligan, P.W., Athwal, B.S., Oakley, D.A. and Frackowiak, R.S.J. (2000). Imaging hypnotic paralysis: implications for conversion hysteria. Lancet, 355, 986–7. Hilgard, E.R. (1977). Divided Consciousness: Multiple Controls in Human Thought and Action. New York: John Wiley & Sons. Irwin, H.J. (1994). Proneness to dissociation and traumatic childhood events. J Nerv Ment Dis, 8, 456–60. Janet, P. (1889). L’automatisme Psychologique. Paris: Felix Alcan. Janet, P. (1924). The Major Symptoms of Hysteria (2nd edition). New York: Macmillan. Kihlstrom, J.F. (1994). One hundred years of hysteria. In Dissociation: Clinical and Theoretical Perspectives, ed. S.J. Lynn and J.W. Rhue, pp. 365–94. New York: Guilford Press. Kirsch, I. and Lynn, S.J. (1995). The altered state of hypnosis. Am Psychologist, 50, 846–58. Kirsch, I. and Lynn, S.J. (1997). Hypnotic involuntariness and the automaticity of everyday life. Am J Clin Hypnosis, 40, 329–48. Kosslyn, S.M. (1994). Image and Brain. Cambridge, MA: MIT Press. Kretschmer, E. (1926). Hysteria. New York: Nervous and Mental Disease Publishing Co. Krishnamoorthy, E.S., Brown, R.J. and Trimble, M.R. (2001). Personality and psychopathology in non-epileptic attack disorder (NEAD): a prospective study, Epilepsy Behav, 2, 418–22. Kuyk, J., van Dyck, R. and Spinhoven, P. (1997). The case for a dissociative interpretation of pseudoepileptic seizures. J Nerv Ment Dis, 184, 468–74. Kuyk, J., Spinhoven, P. and van Dyck, R. (1999). Hypnotic recall: a positive criterion in the differential diagnosis between epileptic and pseudoepileptic seizures. Epilepsia, 40, 485–91. Lipowski, Z.J. (1968). Review of consultation psychiatry and psychosomatic medicine. III. Theoretical issues. Psychosom Med, 30, 395–422. Meierkord, H., Will, B., Fish, D. and Shorvon, S. (1991). The clinical features and prognosis of pseudoseizures diagnosed using video-EEG telemetry. Neurology, 41, 1643–6. Nemiah, J.C. (1991). Dissociation, conversion and somatization. In American Psychiatric Press Annual Review of Psychiatry, ed. A. Tasman and S. Goldfinger, Vol. 10, pp. 248–60. Washington, DC: APA Press. Oakley, D.A. (1999). Hypnosis and conversion hysteria: A unifying model. Cognitive Neuropsychiatry, 4, 243–65. Pettinati, H.M., Horne, R.L. and Staats, J.M. (1985). Hypnotizability in patients with anorexia nervosa and bulimia. Arch Gen Psychiatry, 42, 1014–16. Pribor, E.E., Yutzi, S.H., Dean, T.J. and Wetzel, R.D. (1993). Briquet’s syndrome, dissociation, and abuse. Am J Psychiatry, 150, 1507–11. Reason, J.T. (1979). Actions not as planned. In Aspects of Consciousness, Vol. 1: Psychological Issues, ed. G. Underwood and R. Stevens, pp. 67–89. London: Academic Press. Sierra, M. and Berrios, G.E. (1998). Depersonalization: neurobiological perspectives. Biol Psychiatry, 44, 898–908. Spiegel, D. and Cardeña, E. (1991). Disintegrated experience: the dissociative disorders revisited. J Abnorm Psychol, 100, 366–78.

209

Seizures, epilepsy and dissociation Spiegel, D., Hunt, T. and Dondershine, H.E. (1988). Dissociation and hypnotizability in posttraumatic stress disorder. Am J Psychiatry, 145, 301–5. Thomas, L. and Trimble, M.R. (1997). Dissociative disorders. In Epilepsy: A Comprehensive Textbook, ed. J. Engel and T.A. Pedley, pp. 2775–84. Philadelphia: Lippincott–Raven Publishers. Woody, E.Z. and Bowers, K.S. (1994). A frontal assault on dissociated control. In Dissociation: Clinical and Theoretical Perspectives, ed. S.J. Lynn and J.W. Rhue, pp. 52–79. New York: Guilford Press. World Health Organization (1992). The International Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines (Tenth Revision) (ICD–10). Geneva: WHO.

14

Psychobiology of psychogenic pseudoseizures J. Chris Sackellares and Dalma Kalogjera-Sackellares Malcolm Randall VA Medical Center and University of Florida, Florida, USA

Introduction Psychogenic pseudoseizures are paroxysmal events which mimic epileptic seizures. While patients suffering from these symptoms are referred to neurologists because they are mistakenly believed to have epilepsy, neurologists consider the underlying disorder to be of psychological aetiology. Our observations on the nature of pseudoseizures are based on a carefully studied sample of 100 patients with pseudoseizures evaluated over a period of 5 years at the University of Michigan Medical Center. The patients were evaluated by intensive neurological, clinical psychological and/or neuropsychological investigations. Psychodynamic psychotherapy was performed in a portion of this group. The results of these efforts have provided us with insights into the psychopathology of the disorder (Kalogjera-Sackellares, 1995; Kalogjera-Sackellares and Sackellares, 1997a, b, 1999). In addition, we discovered evidence to suggest that neurological disturbances may play a role in the pathophysiology of the disorder (Kalogjera-Sackellares and Sackellares, 1999). A recurrent theme among our patients was trauma. This included physical as well as emotional trauma (Kalogjera-Sackellares, 1995; Kalogjera-Sackellares and Sackellares, 1999). The prevalence of trauma, as well as the post-traumatic character of many of the symptoms, do not fit with the current diagnostic framework of the somatoform disorders (Guggenheim and Smith, 1995) under which pseudoseizures are typically considered. The concept of somatoform disorders obscures the operative impact of both the emotional as well as physical trauma which are cornerstones of our approach to this disorder. Our own observations regarding the psychopathology as well as the possibility of neurological disturbance led us to an extensive review of the literature on pseudoseizures and hysteria. Among the richest sources of information on the subject are the case descriptions by the nineteenth century clinical investigators, Paul Briquet (a psychiatrist) and Jean-Martin Charcot (a neurologist). Briquet wrote extensively on the subject of hysterical seizures (a historic term for pseudoseizures) in his classic book on hys210

211

Psychobiology of psychogenic pseudoseizures

teria (Briquet, 1859). Charcot’s case descriptions as well as his comments on the subjects have been captured for posterity in the form of published lectures (Charcot, 1879; Goetz, 1987; Harris, 1991). These clinicians contributed much to the foundations of both neurology and psychiatry. As we came to learn from the review of their work, both believed that hysteria resulted from a neurological disorder, but failed to find neuroanatomical abnormalities to support this view. As a result, neurological investigations into this disorder all but ceased. The emphasis shifted to investigations into the psychopathology of the disorder. These psychiatric investigations occurred under the historic impact of the pioneering insights of Janet, Freud and others. Our analysis of the literature was aided considerably by the translation of Charcot by G. Siegerson (Charcot, 1879), as well as the more recent translations (Goetz, 1987; Harris, 1991). Analysis of Briquet, on the other hand, was largely based on direct translations (by DKS) of his text (Briquet, 1859) because no English translation of his work was available. In this chapter, we will review the historical events that led to the commonly held view that pseudoseizures are solely grounded in psychopathology, in the absence of detectable neurological disorders. In reviewing the works of Charcot and Briquet, it is apparent that the importance of trauma in the development of hysteria was recognized by both Briquet (1859) and Charcot (1879; Harris, 1991). However, focusing on a series of well-described cases of Charcot, we find that it was emotional trauma, not physical trauma that he emphasized. Yet, we will provide ample evidence that, in Charcot’s cases, physical trauma (particularly head injury) may have played an equally important role. In addition to our review of historical cases, we will discuss the results of recent studies supporting the concept that neurobiological disturbances interact with traumatic environmental stresses, leading to the development of psychogenic pseudoseizures. Historically, psychogenic pseudoseizures have been considered to be a classic manifestation of hysteria. Until the emergence of the modern view of hysteria in the nineteenth century, hysteria was generally believed to result from disturbances in the reproductive system, specifically the uterus. Nineteenth century investigators, particularly Pierre Briquet, and Jean-Martin Charcot, rejected the aetiological role of the uterus. In addition to making this historic contribution, they furthermore hypothesized that hysteria arose from disturbances in the central nervous system (Briquet, 1859; Goetz, 1987; Harris, 1991). Thus, both of these clinical investigators believed hysteria to be a neurological disorder (although Briquet’s contribution to the conceptualization of hysteria as a disorder of the central nervous system is rarely acknowledged and has only recently been discussed at some length in the context of the ‘forgotten avantgarde of neuroscience’ (Mai, 1982; Mai and Merskey, 1981). First Briquet (Mai, 1982; Mai and Merskey, 1981), and later Charcot, attempted to identify lesions in the nervous system to account for the

212

J.C. Sackellares and D. Kalogjera-Sackellares

symptoms of hysteria. They were unable to find relevant anatomical lesions in their patients. However, their quest was limited by the histological methods of the times. In response to the absence of identifiable lesions, both men postulated physiological or ‘dynamical’ neurological disturbances to account for the functional disturbances and physical findings they observed in their patients. As noted earlier, Charcot emphasized the importance of trauma in patients with hysteria. However, he focused upon the emotional trauma, or ‘shock’ rather than physical injury to the nervous system. Review of his cases fully substantiates the importance of shock and emotional trauma in these unfortunate individuals (Harris, 1991). However, his case reports also describe physical trauma, and in particular, head injury as well as debilitating illnesses in the vast majority of his cases. While Charcot described these physical insults, he did not postulate a causal relationship between head injury and hysteria, presumably because of the absence of neurological lesions in postmortem tissue. Given the obvious presence of emotional trauma and psychological symptoms among hysterical patients and the failure to find neurological lesions, it was natural that subsequent clinical investigators, led by Janet and Freud, would focus upon psychodynamics. The success of psychological investigations and failure of biological investigations led neurologists to abandon the subject of hysteria. The primary interest of neurologists became limited to the differentiation of hysteria from ‘true’ neurological disorders. Apart from recognizing the importance of emotional trauma in his patients, Charcot was unable to provide further insight into the aetiology of hysteria. However, he did provide extraordinarily useful descriptions of findings on neurological examination as well as vivid description of hysterical seizures. His extraordinarily detailed studies of the phenomenology of hysterical patients can be found in the lectures chronicled by his students (Charcot, 1879; Harris, 1991). Based on these observations, Charcot described four phases of the archetypal hysterical seizure: (1) a prodromal phase, (2) an epileptoid phase, (3) a phase characterized by bizarre postures and (4) an emotional phase. Less severe or elaborate seizures could occur. These modified attacks consisted of only parts of the full-blown archetypal seizures. The study of these phases was aided by drawings that captured the important details of hysterical seizures (see Figures 14.1 and 14.2). A large sample of these drawings was published in a book by Richer (1885). Richer’s drawings captured the motor manifestations of pseudoseizures (for example, see Figure 14.1). Even more impressive was his success in rendering facial expressions which capture the intense emotional and experiential aspects of certain phases of the hysterical seizure. An example is illustrated in Figure 14.2. The question arises as to whether today’s patient with psychogenic pseudoseizures suffers from the same malady as the hysteria cases of Charcot’s time. Given the

213

Psychobiology of psychogenic pseudoseizures

Figure 14.1. Reproduction of a drawing by Richer depicting a patient experiencing the epileptoid phase of a hysterical seizure (from Richer, 1885).

high incidence of nonepileptic seizures (‘hystero-epilepsy’) in hysterical patients of the nineteenth century, it seems reasonable to assume that Charcot was describing the same clinical entity as that underlying today’s psychogenic pseudoseizure. Citing Briquet, Charcot stated that hysterical seizures occurred in all but 25% of patients with hysteria (Harris, 1991). Thus, nonepileptic seizures have been considered a typical manifestation of hysteria from the beginning of modern neurology. In the nineteenth century, hysteria was grouped along with other disorders which had no known anatomical lesion. Examples of such disorders included epilepsy and chorea, neither of which were associated with neuropathological lesions demonstrable with the histological techniques of the time. Disorders considered to be neurological, yet without a demonstrable lesion, were classified as ‘neuroses’. Paradoxically, while the term ‘neurosis’ has persisted, its modern meaning, in contrast to that of nineteenth century clinical investigators, indicates psychological causation. The modern vestige of this early attempt at classification is a tendency among many clinicians to consider nonepileptic seizures as ‘functional’ rather than

214

J.C. Sackellares and D. Kalogjera-Sackellares

Figure 14.2. Reproduction of the ‘period of passionate attitude’ (the emotional phase), the fourth stage of a classical hysterical seizure.

‘neurological’ or ‘organic’ in aetiology. The implication is that nonepileptic pseudoseizures do not have a biological substrate. However, careful reading of Charcot’s lectures suggests that he used the term ‘functional’ to mean a disorder of the nervous system that is not associated with demonstrable lesions of the nervous system. Today, with knowledge gained from more powerful investigative tools, we have extended the concept of organic to include a number of toxic, metabolic, degenerative and genetic disorders that are not associated with physical lesions. In clinical practice today, the term ‘functional’, applied to a symptom or finding on examination, is taken to mean that the cause is ‘psychological’ as opposed to biological in nature. It

215

Psychobiology of psychogenic pseudoseizures

is generally accepted that many disorders once considered to be ‘functional’ are due to biological causes (e.g. Gilles de la Tourette syndrome, schizophrenia, bipolar affective disorder). In an analogous manner, it is possible that a biological substrate for pseudoseizures will be identified, using today’s sophisticated investigative tools. Charcot meticulously applied the prevailing clinical–anatomical method, a method which he helped perfect, to the investigation of patients with hysteria. The clinical–anatomical method involved the correlation of clinical signs and symptoms to anatomical lesions in the nervous system. His application of this method to the study of the progressive muscular atrophies, amyotrophic lateral sclerosis, and multiple sclerosis resulted in major contributions to the understanding of those disorders. Such cases were classified as organic because they were associated with localized lesions of the nervous system. Such lesions were detectable by visual inspection of autopsy material. More detailed localization and description of these organic lesions was made possible by skilful use of the microscope, a tool that had been recently introduced into medical science. In contrast to his success with the organic neurological disorders, Charcot was unable to unravel the mystery of hysteria through the application of the clinical–anatomical method. Hysterical patients presented a plethora of observable physical findings on neurological examination. In addition to seizures, these findings included blindness, diplopia, limb weakness, sensory loss, tremors and involuntary movements. However, these findings occurred in patterns that differed from those found in patients with observable anatomical lesions of the nervous system. In some cases, the pattern of visual disturbance, weakness, or sensory loss conflicted with the rules of functional neuroanatomy known at the time. More importantly, the profound clinical abnormalities were not associated with detectable lesions in the nervous system. Charcot was convinced that the clinical findings of hysterical patients were due to appropriately located anatomical abnormalities. However, in the absence of observable structural lesions, he referred to these lesions as ‘dynamical’ lesions (Harris, 1991). Although the clinical findings in hysterical patients differed from those of the organic neurological disorders, Charcot was able to demonstrate classical clinical findings that have become the hallmark of hysteria. These findings include constricted visual fields, monocular diplopia, monoplegia, strange contractures and hemianaesthesia. The motor weakness observed in hysterics was not associated with the reflex abnormalities which were seen in patients with demonstrable lesions of the nervous system. Even today, such findings are taken as evidence for the presence of hysteria, although modern psychiatrists prefer to apply more modern terms, such as ‘conversion disorder’. Charcot preferred not to offer theories to explain neurological disorders, generally preferring to confine his discussions to empirical findings. However, his observations

216

J.C. Sackellares and D. Kalogjera-Sackellares

of the effects of hypnosis and suggestion in hysterics served as a stimulus for the subsequent development of the psychological theories pioneered by Janet and Freud (Goetz, 1987). The clinical histories he provided during his lectures frequently described neurological as well as psychiatric illness in family members. His comments suggest that he considered heredity an important factor in the development of hysteria. Although he did not speculate about this issue, his discussions indicate that various neurological symptoms appearing in the same family represented different clinical manifestations of an underlying hereditary predisposition (Harris, 1991). Charcot’s lectures included descriptions of traumatic events in his female patients with hysterical seizures (Charcot, 1879); however, we noted that it was in his male hysterics that he emphasized the history of antecedent trauma and previous illness (Harris, 1991). In his lectures on seven cases of male hysteria (in all of whom pseudoseizures figured prominently), he emphasized the history of antecedent trauma and its relationship to the onset of symptoms. During his lectures, his comments focused upon psychological trauma. However, our careful review of his lectures on seven male patients with hysteria reveals numerous references to significant closed head injuries. Indeed, he describes closed head injuries in four of the seven cases described. A fifth patient lost consciousness as a result of an accident that caused severe blood loss. Three were either physically abused or attacked. Antecedent debilitating illnesses occurred in five cases. Emotional trauma is described in all seven cases. Paternal alcohol abuse was described in four cases, but there is only one patient with a history of past alcohol abuse. Charcot’s emphasis on the role of emotional trauma and not physical trauma is understandable, given the intolerable psychological and social stresses patients experienced. What is less clear is that he did not consider the impact of the physical trauma which was equally apparent in his patients. Consistent with the historic heritage discussed in this chapter, a number of contemporary authors have rediscovered serious traumatic experiences in patients with psychogenic pseudoseizures (Alper et. al., 1993; Arnold and Privitera, 1996; Bowman, 1993; Cartmill and Betts, 1992; Gross, 1982; Harden, 1997; KalogjeraSackellares, 1995). This recurrent theme of trauma figured so prominently in the histories of pseudoseizure patients that it played a key role in the classification of pseudoseizure syndromes proposed by Kalogjera-Sackellares (1995). The continuity of historical and contemporary reference to trauma makes the study of trauma a natural starting point for investigations into the nature of pseudoseizures (Kalogjera-Sackellares and Sackellares, 1999). It must be considered relevant to the understanding of its causes and to the development of approaches to clinical management (Kalogjera-Sackellares, 1995). Although psychic trauma is a dominant factor in most cases of psychogenic pseudoseizures, physical trauma, often involving closed head injuries also is a

217

Psychobiology of psychogenic pseudoseizures

Table 14.1. Demographics of 100 patients with psychogenic pseudoseizures referred to the neuropsychology programme at the University of Michigan

Males Females Mean age Pure pseudoseizures Mixed epilepsy and pseudoseizures Abnormal neurological examination Structural brain abnormality on neuroimaging History of closed head injury History of substance abuse

20% 80% 33.7% 71% 29% 4% 8% 52% 13%

Source: Adapted with permission.

prominent part of the history in many patients. Several recent authors have commented on head injuries in their patients with psychogenic pseudoseizures. A high incidence of head trauma occurring during physical abuse or during accidents was reported by Bowman (1993). Similarly, Lancman et al. (1993) found many patients with head trauma associated with loss of consciousness. In addition, Westbrook et al. (1998) found that 32% of their patients with nonepileptic seizures had experienced an antecedent head trauma which was classified as minor in 91% of the cases. Upon reviewing the histories of 100 patients with psychogenic pseudoseizures evaluated at the University of Michigan (Kalogjera-Sackellares and Sackellares, 1999), we found a documented history of significant closed head injury in 52% (see Table 14.1). In most instances, the trauma would have been categorized as mild head injury. These patients sustained head injury as a result of motor vehicle accidents, accidents on the job, falls or physical assault. Often, the physical assault was part of a chronic pattern of physical abuse. In many instances, these injuries occurred during childhood. However, there are some cases in which there was no history of significant head trauma until adulthood. As in many of Charcot’s cases, in those cases with head trauma during adulthood, the injury preceded seizure onset by days to months and the patients usually ascribed a causal role to the head injury. The high incidence of head trauma in patients with psychogenic pseudoseizures raises the question as to whether the head trauma may have played some role in the pathogenesis of the disorder. Is it possible that mild brain injury can render a patient more vulnerable to psychogenic pseudoseizures? In our sample, other potential causes of brain injury were reported. A history of drug or alcohol abuse was found in 13% of patients (Table 14.1). Approximately 29% (Table 14.1) were

218

J.C. Sackellares and D. Kalogjera-Sackellares

found to have a prior or concurrent history of epilepsy. In another sample, we found evidence for active epilepsy in approximately 25% of patients with psychogenic pseudoseizures (Kalogjera-Sackellares and Sackellares, 1997a, b). In their case series, Wilkus et al. (1984) reported a high incidence of events that may have contributed to neurological impairment. Such events occurred in fully 80% of their sample. These events included head injury with loss of consciousness or other sequelae in 7 of their 25 cases; an infectious disorder with sequelae in 2 cases; prior brain surgery had occurred in 2 cases. Other events such as birth trauma, sun stroke, severe heat exhaustion, partial drowning or gas exposure also occurred in their patients. Further evidence of impaired brain function in patients with psychogenic pseudoseizures comes from studies of cognitive and intellectual function in these patients. Neuropsychological test batteries are employed in clinical practice to detect evidence of organic brain dysfunction. Therefore, it is natural to expect that patients with a purely ‘functional’ disorder would perform normally on these tests. Alternatively, one might expect the pattern of test performance to differ from patients with ‘organic’ brain disorders. However, several investigators have reported impaired neuropsychological performances in patients with pseudoseizures. In fact, several studies, reported by different investigators, have produced very similar results (Kalogjera-Sackellares and Sackellares, 1999). Wilkus et al. (1984; Wilkus and Dodrill, 1989) reported results from administration of an intensive clinical neuropsychological test battery in 25 patients with psychogenic pseudoseizures evaluated at the University of Washington. These investigations revealed that their sample of patients performed in the impaired range on 51.2% of the tests. Sackellares et al. (1985) analysed a sample of patients evaluated at the University of Virginia and University of Michigan. The authors found that ‘the cognitive performance of that group was less than would be expected for normal individuals of similar intelligence’. Binder et al. (1998) compared a group of patients with nonepileptic seizures with a group of patients with epilepsy and with normal control subjects on a battery of neuropsychological tests. They found that the two seizure groups performed significantly more poorly than normal controls. However, there were no significant differences between the two seizure groups on any items of the extensive test battery. They also found that performance on neuropsychological tests was more strongly correlated with measures on the MMPI/MMPI-2 in the nonepileptic seizure group. Based on these observations, the investigators concluded that neuropsychological tests do not discriminate between patients with epilepsy and those with nonepileptic seizures. These investigators concluded that impairment of neuropsychological performance was strongly related to emotional and psychosocial factors. In a more recent study, we examined intellectual and neuropsychological test performance in 53 patients with psychogenic pseudoseizures (Kalogjera-

219

Psychobiology of psychogenic pseudoseizures

Sackellares and Sackellares, 1999). The group comprised 44 patients with pseudoseizures who had no evidence of concomitant epilepsy, and no definitive historical data that would support the diagnosis of antecedent epilepsy. This subgroup was designated as the ‘pure pseudoseizure’ subgroup. In addition, there were nine patients with documented pseudoseizures and concomitant epilepsy, or a welldocumented history of epileptic seizures. This subgroup was designated as the ‘mixed pseudoseizures and epilepsy’ subgroup. In both subgroups, there was a preponderance of women (77% in the pure subgroup and 78% in the mixed subgroup). Intellectual functioning had been measured, using the WAIS–R in each of these patients. The Halstead–Reitan Neuropsychological Test Battery (Boll, 1981) was administered to all but one. These tests were performed as a part of standard clinical evaluation which included medical and neurological history, physical and neurological examinations, routine EEG, neuroimaging, neuropsychological evaluations, and in most cases, long-term EEG-video monitoring. Interestingly, there were no statistically significant differences in the two subgroups with respect to WAIS–R scores. The group as a whole (pure and mixed subgroups combined) revealed an interesting pattern of scores (see Figure 14.3). The mean verbal IQ, performance IQ, and full scale IQ all fell in the average range. Scores were highly variable and ranged from the mentally deficient to the very superior. However, the overall distribution was skewed toward the low end of the average range. Scores on the Halstead–Reitan Neuropsychological Test Battery were not significantly different for the pure and mixed subgroups. Of particular importance is the finding that more than 50% of the group as a whole performed in the impaired range on more than half of individual subtests of the battery (based on published cutoff scores of Jarvis and Barth (1984). The percentage of patients scoring in the impaired range on each of the Halstead–Reitan tests for which cutoff scores are available is shown in Table 14.2. The Halstead Impairment Index (HII) is a summary score for the Halstead–Reitan Neuropsychological Test Battery. The mean HII for the entire sample and for each subgroup was in the impaired range. However, individual HII scores varied substantially (Figure 14.4). Scores ranged from 0.1 (normal) to 1.0 (impaired performance on all the constituent tests used to calculate the HII). The HII fell within the moderately to severely impaired range (0.7 and higher, based on the criteria of Jarvis and Barth (1984)) in nearly half (48.8%) of the pure pseudoseizure group. This finding was particularly important given that this group had not been diagnosed with any defined neurological disorder and their symptoms had been assumed to result from purely psychological causes. This study (Kalogjera-Sackellares and Sackellares, 1999) involved a large sample of patients with well-documented psychogenic pseudoseizures. The intellectual and neuropsychological findings of this study are consistent with those of previous studies. These studies together underscore the prevalence of measurable

220

J.C. Sackellares and D. Kalogjera-Sackellares 16

Number of subjects

14 12 10 8 6 4 2 0 60–69

70–79

80–89

90–99

100–109

110–119 120–129 130–139

IQ range Verbal IQ Performance IQ Full scale IQ

Figure 14.3. Distribution of scores on the Wechsler Adult Intelligence Scale – Revised (WAIS–R) in a sample of 53 patients with psychogenic pseudoseizures. Scores range from the mentally deficient to the very superior range. However, the overall distribution was skewed toward the low end of the average range. (Adapted with permission.)

neuropsychological and cognitive deficits in patients with psychogenic pseudoseizures. Although some investigators have attributed these deficits to emotional factors (Binder et al., 1998), an equally plausible explanation is that many patients with pseudoseizures, with or without concomitant epilepsy, have undiagnosed neurological abnormalities. In the vast majority of cases, structural abnormalities are not reported on MRI and CT scans. In the vast majority of cases, neurological examinations do not reveal focal or lateralized abnormalities that fit patterns that are seen with lesions of the nervous system. In fact, focal or lateralized neurological findings do not fit characteristic patterns associated with anatomical lesions of the nervous system. However, this does not exclude the possibility of biological disturbances affecting brain function. Not all such disturbances cause gross anatomical changes detectable by routine CT or MRI scans. A surprising observation in our patient sample provides additional evidence that biological factors may play an important role in the pathogenesis of psychogenic

221

Psychobiology of psychogenic pseudoseizures

Table 14.2. Percentage of subjects scoring in the impaired range on Halstead–Reitan Neuropsychological Test Battery

Test Halstead Index Categories TPT-Total TPT-Memory TPT-Location Speech Sounds Perception Test Seashore Rhythm Test Trails A Time Trails B Time Finger tapping – dominant hand

Groups combined

Pure pseudoseizures

Mixed pseudoseizures and epilepsy

63% 67% 46% 18% 66% 60% 62% 24% 42% 98%

61% 73% 43% 14% 62% 59% 63% 21% 38% 98%

75% 33% 63% 38% 88% 63% 57% 38% 63% 100%

Source: Reprinted with permission.

pseudoseizures. We recently discovered the very interesting fact that approximately 30% of our patients with pure pseudoseizures are left-handed (KalogjeraSackellares and Sackellares, 2001). This is markedly increased in comparison with the 10% prevalence of left-handers reported in the general population (Hardyck and Petrinovich, 1977). This observation can be explained in one of three ways: (1) it is a chance observation, (2) left-handers are prone to developing psychogenic pseudoseizures, or (3) our sample contains pathological left handers. The probability of finding this high incidence of left-handers by chance alone is extremely low (P⫽0.000381) (see Kalogjera-Sackellares and Sackellares, 2001). That normal lefthanders are predisposed to developing some sort of functional disorder characterized by seizure-like activity seems quite unlikely. Given the neuropsychological and cognitive difficulties observed in patients with pseudoseizures, it is more likely that the over-representation of left-handers in our sample is due to the presence of some pathological left-handers (see Kalogjera-Sackellares and Sackellares, 2001). Pathological left-handers are individuals who were destined to be right-handers, but developed left-hand dominance due to an insult to the left frontal lobe prenatally, during infancy or early in childhood. In summary, we traditionally consider psychogenic pseudoseizures to be a manifestation of a purely psychological disorder. This view stems largely from the absence of gross neuroanatomical lesions that can explain the seizures or associated neurological symptoms and signs. Further, these patients manifest a plethora of

J.C. Sackellares and D. Kalogjera-Sackellares

Number of subjects

222

Impairment index Figure 14.4 Distribution of the Halstead Impairment Index (HII), the summary score for the Halstead–Reitan Neuropsychological Test Battery, administered to 52 of the same patients shown in Figure 14.3. A score of 0.1 is normal. Scores of 0.7 and above are in the moderately to severely impaired range. A score of 1.0 indicates performance in the pathological range on all constituent tests used to compute the HII. Note that individual scores ranged from normal to 1.0. However, 46.9% of patients in the entire group (pure pseudoseizures and mixed pseudoseizures and epilepsy subgroups combined) scored in the moderately to severely impaired range; and 48.8% of the pure group (n⫽44) scored in the moderately to severely impaired range. (Adapted with permission.)

emotional and personality disturbances (Bowman, 1993; Cartmill and Betts, 1992; Kalogjera-Sackellares, 1995; Kalogjera-Sackellares and Sackellares, 1997a, b; Roy, 1979; Vanderzant et al, 1986) which direct the clinician toward psychiatric issues. The importance of psychological factors is further underscored by the prevalence of traumatic emotional experiences in these patients. Yet, a persistent question is why these traumatic experiences trigger pseudoseizures and other pseudoneurological symptoms in some individuals, but not others. At least one possibility is that some individuals are biologically susceptible to the development of psychogenic pseudoseizures in response to traumatic emotional experiences.

223

Psychobiology of psychogenic pseudoseizures

We suggest that, at least in some individuals, the biological susceptibility may result from neurological impairment. As our previous discussion indicates, a number of investigators have reported a high incidence of closed head injury and other neurological insults in their patients. Several investigators have reported impaired performance on tests of neuropsychological and cognitive function. Finally, we have found an over representation of left-handers in our pseudoseizure patient sample (Kalogjera-Sackellares and Sackellares, 2001). This overrepresentation of left-handers suggests one of two possibilities. First, it is possible that left-handers subjected to certain neurological insults or severe emotional trauma may be predisposed to psychogenic pseudoseizures. A second, and more plausible explanation is the presence of pathological left-handers in the sample. If the additional left-handers are pathological left-handers, it suggests that some patients may have experienced neurological insults to the left hemisphere prior to the establishment of hand dominance (Kalogjera-Sackellares and Sackellares, 2001). Taken together, all of these observations suggest that biological factors play an important aetiological role in the development of psychogenic pseudoseizures. This possibility in no way negates the role of emotionally traumatic experiences or other environmental factors in the development of psychiatric disorders characterized by pseudoseizures. Nonetheless, it may help to explain why all individuals experiencing traumatic life events do not develop pseudoseizures. We feel that it is extremely important to consider the strong possibility that psychobiological factors may play an important role in the pathophysiology of epilepsy. Because pseudoseizures have long been assumed to stem from purely ‘functional’ causes, there have been essentially no investigations into the neurobiology of this disorder. This disorder causes as much disability as medically intractable epilepsy. Yet, little progress has been made in the exploration of new therapeutic interventions. Psychotherapy can be beneficial to many of these patients (Kalogjera-Sackellares, 1995). However, not all respond, and many do not have access to psychotherapy. To our knowledge, there have been no organized scientific investigations into the use of psychopharmacology in the treatment of this disorder.

R E F E R E N C ES Alper, K., Devinsky, O., Perrine, K., Vazquez, B. and Luciano, D. (1993). Nonepileptic seizures and childhood sexual and physical abuse. Neurology, 43, 1950–3. Arnold, L.M. and Privitera, M.D. (1996). Psychopathology and trauma in epileptic and psychogenic seizure patients. Psychosomatics, 37, 438–43.

224

J.C. Sackellares and D. Kalogjera-Sackellares Binder, L.M., Kindermann, S.S., Heaton, R.K. and Salinsky, M.C. (1998). Neuropsychologic impairment in patients with nonepileptic seizures. Arch Clin Neuropsychology, 13, 513–22. Boll, T.J. (1981). The Halstead–Reitan Neuropsychology Test Battery. In Handbook of Clinical Neuropsychology, ed. S.B. Filskov and T.J. Boll, pp. 577–607. New York: John Wiley & Sons. Bowman, S. (1993). Etiology and clinical course of pseudoseizures – relationship to trauma, depression and dissociation. Psychosomatics, 34, 333–42. Briquet, P. (1859). Traité Clinique et Thérapeutique de L’hystérie. Paris: J.B. Bailliere. Cartmill, A. and Betts, T. (1992). Seizure behavior in a patient with post-traumatic stress disorder following rape. Notes on the etiology of pseudoseizures. Seizure, 1, 33–6. Charcot, J.M. (1879). Lectures on the Diseases of the Nervous System Delivered at La Salpetriére. Translated by G. Siegerson. Philadelphia: Henry O. Lea. Goetz, C.G. (1987). Charcot the Clinician: The Tuesday Lessons. New York: Raven Press. Gross, M. (1982). Incest and hysterical seizures. Med Hypnoanalysis, 3, 146–52. Guggenheim, F.G. and Smith, G.R. (1995). Somatoform disorders. In Comprehensive Textbook of Psychiatry, Sixth Edition, Vol. 1, ed. H.I. Kaplan and B.J. Sadock, pp. 1251–70. Baltimore: Williams & Wilkins. Harden, C.L. (1997). Pseudoseizures and dissociative disorders: a common mechanism involving traumatic experiences. Seizure, 6, 151–5. Hardyck, C. and Petrinovich, L.F. (1977). Left-handedness. Psychol Bull, 84, 385–404. Harris, R. (ed.) (1991) Charcot, J.M.: Clinical Lectures on Diseases of the Nervous System. London: Tavistock/Routledge. Jarvis, P.E. and Barth, J.T. (1984). The Halstead–Reitan Test Battery: an Interpretive Guide. Odessa, FL: Psychological Assessment Resources Inc. Kalogjera-Sackellares, D. (1995). Psychological disturbances in patients with pseudoseizures. In Psychological Disturbances in Epilepsy, ed. J.C. Sackellares and S. Berent, pp. 191–217. Boston: Butterworth–Heinemann. Kalogjera-Sackellares, D. and Sackellares, J.C. (1997a). Personality profiles of patients with pseudoseizures. Seizure, 6, 1–7. Kalogjera-Sackellares, D. and Sackellares, J.C. (1997b). Analysis of MMPI patterns in patients with psychogenic pseudoseizures. Seizure, 6, 419–27. Kalogjera-Sackellares, D. and Sackellares, J.C. (1999). Intellectual and neuropsychological features in patients with psychogenic pseudoseizures. Psychiatry Res, 86, 73–84. Kalogjera-Sackellares, D. and Sackellares, J.C. (2001). Impaired motor function in patients with psychogenic pseudoseizures. Epilepsia, 42, 1600–6. Lancman, M.E., Brotherton, T.A., Asconape, J.J. and Penry, J.K. (1993). Psychogenic seizures in adults: a longitudinal analysis. Seizure, 2, 281–6. Mai, F.M. (1982). The forgotten avant-garde. Trends Neurosci, 5, 67–8. Mai, F.M. and Merskey, H. (1980). Briquet’s treatise on hysteria. Arch Gen Psychiatry, 37, 1401–5. Mai, F.M. and Merskey, H. (1981). Briquet’s concept of hysteria: an historical perspective. Can J Psychiatry, 26, 57–63. Richer, P. (1885). Etudes Cliniques sur la Grande Hystérie ou l’Hystéroépilépsie. Paris: Delahaye et LeCrosnier. Roy, A. (1979). Hysterical seizures. Arch Neurol, 36, 447.

225

Psychobiology of psychogenic pseudoseizures Sackellares, J.C., Giordani, B., Berent, S. et al. (1985). Patients with pseudoseizures: intellectual and cognitive performance. Neurology, 35, 116–19. Vanderzant, C.W., Giordani, B., Berent, S. Dreifuss, F.E. and Sackellares, J.C. (1986). Personality of patients with pseudoseizures. Neurology, 36, 664–8. Westbrook, L.E., Devinsky, O. and Geocadin, R. (1998). Nonepileptic seizures after head injury. Epilepsia, 39, 978–82. Wilkus, R.J., Dodrill, C.B. and Thompson, P.M. (1984). Intensive EEG monitoring and psychological studies of patients with pseudoepileptic seizures. Epilepsia, 25, 100–7. Wilkus, R.J. and Dodrill, C.B. (1989). Factors affecting the outcome of MMPI and neuropsychological assessments of psychogenic and epileptic seizure patients. Epilepsia, 30, 339–47.

15

Epilepsy and panic disorder Howard A. Ring and Nuri Gene-Cos St Bartholomew’s and the Royal London School of Medicine, London, UK

Introduction Epilepsy is a neurological condition associated with well-defined abnormalities of brain electroencephalographic (EEG) activity. It is generally treated with antiepileptic agents and occasionally with resective neurosurgery. Panic disorder (PD) is a psychiatric condition mostly treated with antidepressant medication and cognitive–behaviour therapy (CBT). Why then should the conditions in any way be considered together in the same chapter? This paper will initially point out clinical similarities and differences and will then discuss the question of whether there are any pathophysiological similarities between the two conditions. Epilepsy

Epilepsy is a complex condition characterized by recurrent episodes of paroxysmal disturbance of normal brain functioning. For a secure diagnosis of epilepsy to be made there needs to be evidence that these paroxysms of disturbance involve both disruptions of ongoing behaviour and abnormalities of brain electroencephalographic (EEG) activity recognized as epileptiform. The clinical manifestations of epilepsy depend on the site of origin, or focus of the seizure, the pattern, manner and extent of propagation of epileptiform electrical activity through the brain and the aetiology of the epilepsy. From the point of view of the person with epilepsy, the subjective experience of repeated seizures is generally relatively similar, or stereotyped, although the actual nature of the experience is highly variable between individuals. Hence some may experience hardly any awareness of a seizure, even if they collapse to the ground unconscious and proceed to suffer a sustained tonic-clonic convulsion. Other people with epilepsy may experience emotional experiences such as fear, unreality or déjà vu which, although perhaps not associated with any behaviours that an external observer would recognize as abnormal, nevertheless may be very disruptive and unpleasant. The current International League against Epilepsy (ILAE) classification of seizures is based on EEG findings and seizure phenomenology. The two largest groups 226

227

Epilepsy and panic disorder

of seizure types are the partial seizures, which have a focal onset, and the primary generalized seizures, which do not have any identifiable focal origin. When considering the relationship between epilepsy and PD, whether with respect to differential diagnosis or putative commonalities in pathophysiology, it is the partial seizures which are of most relevance. Panic disorder

In both the ICD–10 and DSM–IV diagnostic classifications PD is considered as an anxiety disorder. Although these classificatory systems do not represent the last word in mechanistic understanding of behavioural disorders, it is clear from the inclusion of PD within the anxiety disorder/neurotic disorder grouping that the general view is that PD has a psychological rather than a biological aetiology. However, whilst it is clear that the core subjective experience of PD is one of extreme fear, this does not in itself prove that the disorder is simply an extreme end of a continuum that starts with mild anxiety. Panic attacks may (in common with epileptic seizures) be described as paroxysmal events. They are discrete periods of intense fear or emotional discomfort, accompanied by a range of somatic symptoms including palpitations, trembling, a feeling of shortness of breath (which may be associated with hyperventilation), sweating, feelings of choking and psychological symptoms including depersonalization, fear of losing control and fear of dying. The attacks occur spontaneously, without warning, and although they may occur in situations in which they have previously occurred, when the patient is concerned that an event may happen, they may also occur unexpectedly. Individual panic attacks are self-limiting although estimates of duration vary. Retrospective estimates by sufferers suggest an average duration of between 10 and 20 minutes. However, a prospective study reported considerably longer attacks, with a mean of between 15 and 50 minutes (Taylor et al., 1986). Whilst some people experience attacks accompanied by most if not all of the associated symptoms described above, in others there may be very few experienced apart from paroxysmal fear and anxiety. The frequency of attacks varies, both between individuals, and over time within individuals, from several attacks in a day to only occasional attacks over a whole year. The clinical diagnosis of PD is characterized by panic attacks, avoidance of situations in which previous panic attacks have occurred and ongoing worry regarding the possibility of future attacks. However, these recurrent attacks of extreme fear and a feeling of impending death or disaster are not restricted to any particular environmental setting or set of circumstances. In addition, it is important to note that although patients with PD worry about having further panic attacks, this worry is of a lower magnitude than the emotions experienced during an attack.

228

H.A. Ring and N. Gene-Cos

There is good evidence, based on clinical accounts, that PD is not a homogeneous disorder. In some people pure PD exists with panic attacks in the absence of any other psychopathology. However, sizeable proportions of those with PD are comorbid for agoraphobia or depression or both. It has been reported that women with PD are more likely to report depression, anxiety or agoraphobic avoidance than men with this diagnosis (Chambless and Mason, 1986). Increased anxiety and depression in affected women was also noted by Oie et al. (1990). Whilst women may show greater agoraphobic avoidance, men may increase their alcohol intake to cope with their symptoms. Although these differences in the associated symptoms and behaviours between the manifestations of PD in men and women exist, the core features of panic attacks appear to be relatively similar across the sexes. Hence, although in therapeutic terms it is clearly important to consider the whole syndrome with which the patients present, in mechanistic terms it may be that both sexes share a common aetiology of panic attacks, but that cultural and behavioural responses to these attacks then determine the sex differences in associated symptoms. In a large American epidemiological study, the National Comorbidity Survey, it was found that the prevalence of PD was greatest in the age range 15–25 (Eaton et al., 1994). However, separating data from males and females revealed that whilst peak age of onset was within this range for males, for females it was older. This study also noted that people with less than 12 years of education were up to ten times more likely to suffer from PD than those with more than 16 years of education (i.e. including a college education). Perhaps surprisingly, in recent years it has become clear that panic attacks are associated with an increased risk of attempted and completed suicide (Markowitz et al., 1989). The rate of suicide attempts is reported as 20% in those with PD and 12% in those with panic attacks but without the avoidance of panic-inducing situations and anxiety regarding future attacks that contribute to the diagnosis of PD. In this context it is also noted that there is an increased risk of suicide and attempted suicide in people with epilepsy, as discussed by Blumer (Chapter 8). Potential pathophysiological commonalities Biochemical pathophysiology

Although there are various models of neurochemical disturbance underlying seizure genesis and also various models proposed to underlie PD, in both conditions aetiological disruptions of the GABA system have been proposed. It is well established that GABA exerts an inhibitory control on neuronal excitability by its rapid action on Cl⫺ channels. The mechanisms of several antiepileptic drugs, such as phenobarbitone, vigabatrin and benzodiazepines, involve enhancing GABAergic activity in various ways.

229

Epilepsy and panic disorder

As yet, no specific defect in GABA functioning has been identified in patients with PD. However, GABA is indirectly implicated in PD through benzodiazepines. Benzodiazepine receptor agonists produce neuronal inhibition via benzodiazepine receptor modulation of GABAa receptor mechanisms, leading to, amongst other effects, anxiolysis, muscle relaxation and sedation. Benzodiazepine antagonists occupy the benzodiazepine receptor site without producing pharmacological effects, while inverse agonists, such as beta-carboline, are anxiogenic and proconvulsant (Katz et al., 1993). Clinical data show that high potency benzodiazepine agonists, such as alprazolam and clonazepam, have marked antipanic effects. Moreover, other studies have suggested subsensitivity of benzodiazepine receptors in these patients (Eison, 1990). Possibly in relation to this, flumazenil, a benzodiazepine antagonist, has been reported to be panicogenic in patients with PD but not in healthy controls. Electrophysiology

A variety of abnormal electrophysiological findings have been reported in groups of patients with PD in comparison with healthy control subjects. EEG studies

Patients with panic attacks have been reported to have an increased amount of paroxysmal EEG activity (Hughes, 1996), with this occurring up to four times more often than is the case in patients with a depressive illness. Temporal lobe abnormalities have been highlighted in brain electrical activity mapping (BEAM) studies in patients with panic attacks. However, other studies have failed to detect any EEG changes resembling epileptiform activity in people with PD. Stein and Uhde (1989) evaluated a group of 35 medication-free patients with PD (Research Diagnostic Criteria). The EEGs were performed over 45 to 75 minutes by using a 21-channel scalp EEG. In addition, 31 patients had an EEG performed with additional use of nasopharyngeal or anterior temporal leads. Twenty-two patients had been sleep-deprived for 24 hours before the EEG, and recordings were performed during drowsiness or light sleep whenever possible. In all patients, EEGs were obtained during a 2-minute period of hyperventilation and in response to photic stimulation. Patients were divided into two groups: 15 with psychosensory symptoms and 20 without psychosensory symptoms. Their results showed that EEG abnormalities of any type were infrequent, occurring in a total of 5 (14%) of the 35 patients. None of these abnormalities suggested the presence of an epileptiform disturbance but were nonspecific in nature. One patient experienced a severe panic attack during his EEG, yet his EEG recording was normal. Moreover, the authors found no significant association between the presence or absence of EEG abnormalities and the presence or absence of prominent

230

H.A. Ring and N. Gene-Cos

psychosensory symptoms. However, they concluded that given the technical limitations of surface EEG recordings, their findings cannot exclude the possibility that PD and complex partial seizures share common pathophysiological mechanisms or sites of dysfunction. Their findings suggest that although it is not likely that PD is an epileptiform disorder, temporal lobe and limbic structures may play a major role in the pathophysiology of panic. In agreement with Stein and Uhde’s work, Lepola et al. (1990) reported normal EEG findings in the majority of a group of 54 patients with PD who were investigated using extensive EEG recordings and computerized tomography (CT) scan. Fifteen (28%) had previously been treated for temporal lobe epilepsy or another neurological condition. The vast majority of patients exhibited normal EEG and CT findings. Only in 13 (24%) patients did the EEG show increased slow-wave activity, whilst CT scans revealed incidental abnormalities in 6 (20%) of the 30 patients so investigated. The authors commented that, although they did not use a control group to compare the findings with, neither the EEG nor CT showed any focal abnormalities related to PD itself. The situation is slightly different when patients with what have been described as ‘atypical’ panic attacks are studied. Edlund et al. (1987) described a series of six patients who presented with atypical PA involving hostility, irritability, severe derealization, and social withdrawal. All the patients underwent standard EEG recordings. None of the patients had clear temporal lobe epilepsy but most had minor and nonspecific temporal EEG abnormalities. Weilburg et al. (1995) studied 15 subjects who met DSM criteria for panic attack but who also had atypical features including at least one of the following: sensory distortions, change in level of consciousness, aphasia, focal paraesthesia, altered sense of body position, hallucinations, sudden shifts in mood, headache or autonomic changes. These subjects underwent prolonged ambulatory EEG monitoring which included sphenoidal recordings. Eleven of their subjects were thus recorded during the course of at least one, and in three subjects multiple, panic attacks. In 45% (5/11, including the three recorded during multiple attacks) the clinical symptoms were associated with focal paroxysmal EEG changes. However, even in those who on some occasions had abnormal EEG changes associated with a panic attack, this only occurred in a proportion (with an average of 35%) of their recorded attacks. However, as the authors acknowledge, first, their results may not be applicable to those with panic attacks without atypical features and second, that at least a proportion of their patients, although not meeting diagnostic criteria for epilepsy, may nevertheless have been manifesting atypical ictal activity accounting for the atypical (or possibly the typical) features of panic. Overall, if these studies demonstrate anything, it is that there may be a grey area in which some symptoms associated with panic attacks are associated with abnor-

231

Epilepsy and panic disorder

mal EEG activity. This does not mean to say that the whole episode is driven by electrophysiological disturbances, but it does raise the possibility that there is, in at least a proportion of people with the symptoms of panic attack, some detectable pathophysiological change in brain activity. Feelings of derealization and depersonalization occur relatively frequently in people with PD and are also accepted to occur from time to time in people with temporal lobe epilepsy. Although less common in those with epilepsy, when these symptoms do develop they tend to be experienced as more robust phenomena. Interestingly, there is some evidence that there are electrophysiological differences between those with PD whose symptoms include derealization or depersonalization and those who do not experience these phenomena. Locatelli et al. (1993) investigated computerized EEG activity derived from the temporal lobes (F7, T3, T5, F8, T4, T6) in 30 healthy subjects and 37 patients with PD (DSM–III–R; American Psychiatric Association, 1987) (with or without agoraphobia), in a resting condition and also in an odour stimulation condition designed to activate temporal lobe structures. The patients with PD were divided into two groups: 17 with depersonalization and/or derealization during their panic attacks and 20 without depersonalization and/or derealization. Patients with PD without depersonalization or derealization and healthy controls showed an increase of fast activity (beta 2: 18–30 Hz) and a decrease of slow activities (delta 2: 2–4 Hz; theta 1: 4–6 Hz) independent of odour stimulation. PD patients with depersonalization and/or derealization showed an increase of slow activity (delta 1: 1–2 Hz; delta 2: 2–4 Hz) and bilateral lack of responsiveness in the fast alpha (alpha 2: 10–12 Hz) frequency band during odour stimulation. The authors suggested that the EEG changes during odour stimulation (a temporal lobe activation task) could be interpreted as the orienting reaction to the activating procedure; and that this appears to be different depending on whether or not patients have temporal lobe symptomatology (depersonalization/derealization). These findings indicate that PD patients with temporal symptoms respond to procedures activating their temporal regions with hypersynchronization of electrical activity. The increase of delta activity can be seen as a lowering of the sensitivity threshold of the deep temporal regions, supporting the hypothesis of temporal lobe involvement in patients with PD and temporal lobe symptomatology. These findings also demonstrate that this subset of patients with PD have an abnormality of temporal lobe electrophysiology that in certain circumstances produces clinical symptoms and also, though not necessarily at the same time, show a tendency towards abnormal synchronization of electrical activity. In this context it is noted that hypersynchronous EEG activity may be a feature of ictal EEG discharges. Indeed, in human temporal lobe epilepsy hippocampal hypersynchronous discharges are present and may evolve into a recruiting rhythm leading to propagation of ictal activity beyond the site of onset (Engel, 1998).

232

H.A. Ring and N. Gene-Cos

Event-related potentials

Event-related potentials (ERP) are changes in electrical brain activity that provide a neurophysiological reflection of information processing. They are derived from averaged EEG recordings made whilst subjects undergo repeated presentations of various stimuli in a variety of experimental paradigms. The study of ERP components recorded from subjects whilst they perform cognitive tasks enables the assessment of cerebral information processing with millisecond resolution (Pfefferbaum et al., 1995). The P3a, occurring about 300 milliseconds after the stimulus, is associated with the orienting of attention. It is elicited by irrelevant novel sounds in a sequence of repetitive standard tones. It is generated by centres in the frontal lobes and the hippocampi (Alho et al., 1998; Knight and Nakada, 1998). It has been reported (Clark et al., 1996) that patients with PD, compared with normal controls, have increased peak amplitude fronto-central P3a responses to all tones (not just to novel sounds). The authors suggest that the presence of a large P3a in PD patients might indicate an abnormal cognitive response to processes that otherwise would have been dealt with automatically. PD patients appeared to apply unnecessary attention to the processing of stimuli that should have been filtered out at an earlier processing stage, engaging conscious attention unnecessarily. As P3a is normally not seen in active attention tasks, as it is swamped by the task-related P3b, its presence in PD would indicate, as well as an excess of stimuli processing, a failure to reduce their response to these stimuli after repeated presentation. The P3a normally habituates with repeated stimulus presentation and it may be that there is reduced habituation in PD. If this was to be confirmed it may explain why PD patients became excessively aroused in environments such as crowds or supermarkets, where there is a high level of irrelevant stimuli. The enlarged P3a to taskrelevant stimuli is characteristic of activity that would be expected in reaction to novel, task-irrelevant events and is consistent with specific, functional pathology involving the prefrontal–limbic pathways. The mismatch negativity (MMN), is a relatively early ERP that is considered to reflect the earliest cortical event in cognitive processing of auditory stimuli (Tiitinen et al., 1994), reflecting the preconscious processing of unexpected auditory stimuli. The main sites of MMN generation are in the superior temporal cortices. It is elicited in the laboratory as a response to infrequent stimuli in sequences of frequent homogeneous stimuli. This potential characteristically occurs about 150 milliseconds after the stimulus and can be elicited by changes in simple tones, complex stimuli or components of speech such as phonemes (Naatanen, 1992). In recent studies we have investigated MMN in age and sex-matched groups consisting of 10 patients with panic attacks and PD, 9 patients with epilepsy and 10 normal controls. The results are displayed in Table 15.1. It is noted that whilst MMN parameters differ significantly from controls in a number of sites, in general the

233

Epilepsy and panic disorder

Table 15.1. Mismatch negativity (MMN) results in patients with epilepsy (n⫽9), PD (n⫽10) and in a normal control group (n⫽10)

MMN parameters

Epilepsy group

Panic group

Control group

Onset latency (ms) Duration (ms) Fz (␮V/ms) F4 (␮V/ms) Cz (␮V/ms) Pz (␮V/ms)

⫺66 ⫺88.4 ⫺54a ⫺46a, c ⫺26a ⫺32

⫺767 ⫺775.2b ⫺136 ⫺134c ⫺112 7⫺63

⫺772 ⫺110b ⫺180a ⫺179a ⫺177a ⫺110

Notes: Group differences (P⬍0.05): a epilepsy vs. controls; b panic vs. controls; c epilepsy vs. panic. The MMN was recorded from four electrode sites (Fz, F4, Cz, Pz) defined using the standard 10–20 system. The data in the table demonstrate that the patients with panic disorder showed significantly shorter duration of the MMN potential than the control group. Considering the MMN amplitude (measured as ␮V/ms at Fz, F4, Cz and Pz), it is noted that the patients with epilepsy had significantly smaller amplitude MMNs than controls at the three more anterior electrodes and a significantly smaller amplitude MMN than the PD patients at the electrode nearest the right anterior temporal region.

results for the patients with panic disorder lie between those observed in the epilepsy and the control groups. These results suggest that whilst epilepsy and panic disorder do not share the same electrophysiological abnormalities, nevertheless there are disturbances in temporal lobe electrophysiology in patients with PD. Structural and functional imaging

Fontaine et al. (1990) carried out MRI scans in a group of 30 patients (age between 20 and 40 years) with PD (DSM–III criteria) and 20 matched controls. All patients had been treated with clonazepam for up to 3 months and the MRIs were done when the patients had significantly improved from their anxiety symptoms; moreover, all patients took an additional 2 mg of clonazepam in the hours before the MRI took place. In contrast, none of the controls were on clonazepam. The main finding of this study was the increased incidence of focal abnormalities in the right mesiotemporal area in the PD group. There were a variety of circumscribed highsignal lesions in the white matter which were detected by the MRI as well as asymmetric atrophy of the temporal lobes. The authors emphasized that their findings may be relevant to panic and phobic disorders as not only were limbic structures involved but both the parahippocampal gyrus and the hippocampal formation play a major role in receiving input from the association areas for all sensory modalities.

234

H.A. Ring and N. Gene-Cos

Moreover, these structures could initiate a marked defensive response through the septo-amygdalar complex and brain stem structures. Fontaine et al. (1990) concluded that although they observed an increased incidence of focal neuroanatomical changes in the temporal lobes, it was unclear whether these abnormalities were related to any genetic predisposition to PD. Lucey et al. (1997) compared regional cerebral blood flow (rCBF), using single photon emission tomography, in three groups of patients, 15 patients with PD and agoraphobia, 16 patients with post-traumatic stress disorder (PTSD) and 15 patients with obsessive compulsive disorder (OCD). Their main finding was a reduction in caudate and superior frontal cortical perfusion in both OCD and PTSD groups compared with PD and healthy controls. The caudate reduction correlated negatively with depression (Beck Depression Inventory) and with the PTSD syndrome severity (Impact of Events scale). No differences were found in temporal lobes. Experimental models of anxiety

In a fMRI activation study in normal volunteers it has been demonstrated that the amygdala is involved in conditioning and extinction of fear responses in a fashion similar to that previously observed in experimental animals (LaBar et al., 1998). There is evidence from studies in normal volunteers that abnormal patterns of limbic activity may result in symptoms resembling both features of temporal lobe complex partial seizures and features of panic attacks. In one study, intravenous injections of procaine resulted in a range of subjective experiences including emotional, somatic and visceral experiences often similar to those experienced in the auras of temporal lobe epilepsy as well as resulting in the development of panic attacks in four out of ten subjects. These experiences also included; euphoria, anxiety, depression, fear and derealization. Positron emission tomography (PET) scanning of the subjects during this experiment revealed that all these experiences, described as ‘powerful and overwhelming’, were associated with increased activity in anterior limbic and paralimbic regions (Servan-Schreiber et al., 1998). Both these authors and Ketter et al. (1996), using a similar paradigm, noted that procaine-related activation of the left amygdala was positively correlated with symptoms of fear and negatively correlated with feelings of calmness or euphoria. In this context it is interesting to note that it is left temporal more than right temporal lobe epilepsy that is more associated with the development of negative affective states in people with epilepsy. Amygdala activation (in this case bilaterally) has also been observed following cholecystokinin tetrapeptide-induced anxiety in normal volunteers (Benkelfat et al., 1995). Hence there is good evidence that neurochemical activation of the amygdala and surrounding structures is associated with brief but extreme experiences of panic and anxiety. Similarly, it is well established that seizure activity may originate in the amygdala in some patients with epilepsy. It has further been reported that seizures originat-

235

Epilepsy and panic disorder

ing in the region of the amygdala are particularly associated with subjective emotional experiences that resemble aspects of PD. The biological significance of phenomenological similarities between PD and epilepsy It has been reported that ictal fear occurs in between 5 and 15% of those with temporal lobe epilepsy. The experience may vary from mild uneasiness to intense feelings of panic and impending doom (Devinsky and Vazquez, 1993). Fear is a relatively common aura experience in patients with temporal lobe epilepsy, representing 10% of the experiences reported by Taylor and Lochery (1987). Williams (1956) investigated emotional phenomena in 2000 patients with epilepsy and found that 100 of them reported an emotion as part of the ‘epileptic experience’. As in the study by Taylor and Lochery (1987), the most commonly reported emotion was fear, occurring in Williams’ sample in 61% of the 100 patients with emotional phenomena. On some occasions this fear was quite pervasive, with psychic and somatic features. More recently, it has been reported that patients who experience ictal fear as part of their temporal lobe epilepsy are more likely to have previously suffered an anxiety disorder than those who have not experienced ictal fear (Devinsky and Vazquez, 1993). It is well established that fear may be provoked by activity in medial temporal structures. It is the most common experiential phenomenon produced by depth electrode brain stimulation in antero-medial temporal regions (Halgren et al., 1978). Hence there is evidence that abnormal electrical activity, both epileptiform and experimental, leads to a subjective experience of intense fear of sudden onset: a cardinal feature of panic. In experiments with rats, LeDoux et al. (1990) found that the lateral nucleus of the amygdala receives direct input from the sensory thalamus. By this pathway, the amygdala is able to detect aversive input and fear-conditioned stimuli even if sensory neocortical areas are disconnected, lesioned or ablated. As a consequence, it can quickly and automatically elicit autonomic, endocrine and motor fear responses even before the neocortex is able to build up a coherent representation of the triggering threat stimulus. Moreover, it transmits an alarm signal to the neocortex, which causes it to allocate its attentional resources to the current sensory input. This can be achieved, technically, by activating ascending neuromodulatory transmitter systems such as serotonin, acetylcholine and noradrenaline (Graeff et al., 1993). These neurotransmitters are presumed to affect the signal-to-noise ratio of neocortical processing (Robbins and Everitt, 1995), correlated with focal attention and conscious perception (Crick, 1994). LeDoux (1995) found that neocortical processing is necessary for discriminative conditioning as well as for the

236

H.A. Ring and N. Gene-Cos

extinction of conditioned fear responses and speculated that in the absence of input from primary sensory areas, potential fear information cannot be relayed to higher regions such as the prefrontal cortex. Thus when the activity of the amygdala is not sufficiently controlled and inhibited by these more discriminative modules, it may tend to give rise to false threat alarms. In summary, it seems that preattentive processing of potential threat tends to elicit false threat alarms that automatically activate both physiological fear responses as well as attention directed to the current sensory input (leading to conscious perception and analysis). However, in the intact brain, both of these effects can be modified and extinguished by more discriminative and more elaborated modes of processing (Windmann, 1998). Integrating the previous models into a theory, Windmann (1998) proposes what he calls the ‘false-alarm-theory of PD’. Symptoms of panic and pathological states of anxiety arise from the hyperfunctioning of a preattentive alarm system whose structural basis is closely related to the amygdala and its connections to ascending neuromodulatory transmitter systems. The hyperfunction results in a tendency to signal potential threat to the neocortex which is not adequately modified by more elaborated processing in patients with PD. Conclusions Some studies have found temporo-limbic abnormalities in subsets of people with PD or panic attacks. In general, there is no evidence of a specific relationship between epilepsy, more specifically complex partial seizures, and PD; indeed, in PD most of the EEG, structural and functional findings are nonspecific. However, there clearly is an overlap of phenomenology between the two conditions. A possible explanation is that PD is associated with disturbances of temporo-limbic function, but that these disturbances are not epileptiform or epileptogenic in nature. The clinical similarities between aspects of PD and complex partial seizures, of temporal lobe origin in particular, arise by virtue of the similarity in affected brain structures and the functions of these structures. If there is an overrepresentation of people with both temporal lobe epilepsy and PD, then it may be that both arise, though independently, from a common pathophysiological precursor.

R E F E R E N C ES Alho, K., Winkler, I., Escera, C. et al. (1998). Processing of novel sounds and frequency changes in the human auditory cortex: magnetoencephalographic recordings. Psychophysiology, 35, 211–24.

237

Epilepsy and panic disorder American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders and Related Health Problems (Third Edition, Revised) (DSM–III–R). Washington, DC: APA. Barraclough, B. (1981). Suicide and epilepsy. In Epilepsy and Psychiatry, ed. E.H. Reynolds and M.R. Trimble, pp. 72–6. Edinburgh: Churchill Livingstone. Benkelfat, C., Bradwejn, J., Meyer, E. et al. (1995). Functional neuroanatomy of CCK4-induced anxiety in normal healthy volunteers. Am J Psychiatry, 152, 1180–4. Chambless, D.L. and Mason, J. (1986). Sex, sex-role stereotyping and agoraphobia. Behav Res Ther, 24, 231–5. Clark, C.R., McFarlane, A.C., Weber, D.L. and Battersby, M. (1996). Enlarged frontal P300 to stimulus change in panic disorder. Biol Psychiatry, 39, 845–56. Crick, F. (1994). The Astonishing Hypothesis: The Scientific Search for the Soul. New York: Simon & Schuster. Devinsky, O. and Vazquez, B. (1993). Behavioural changes associated with epilepsy. Neurol Clin, 11, 127–49. Eaton, W.W., Kessler, R.C., Wittchen, H.U. and Magee, W.J. (1994). Panic and panic disorder in the United States. Am J Psychiatry, 151, 413–20. Edlund, M.J., Swann, A.C. and Clothier, J. (1987). Patients with panic attacks and abormal EEG results. Am J Psychiatry, 144, 508–9. Eison, M.S. (1990). Serotonin: a common neurobiologic substrate in anxiety and depression. J Clin Psychopharmacol, 10 (Suppl.), 26–30. Engel, J. Jr (1998). Research on the human brain in an epilepsy surgery setting. Epilepsy Res, 32, 1–11. Fontaine, R., Breton, G., Déry, R., Fontaine, S. and Elie, R. (1990). Temporal lobe abnormalities in panic disorder: an MRI study. Soc Biol Psychiatry, 27, 304–10. Graeff, F.G., Silveira, M.C., Nogueira, R.L., Audi, E.A. and Oliveira, R.M. (1993). Role of the amygdala and periaqueductal gray in anxiety and panic. Behav Brain Res, 58, 123–31. Halgren, E., Walter, R.D., Cherlow, D.G. and Crandall, P.H. (1978). Mental phenomena evoked by electrical stimulation of the human hippocampal formation and amygdala. Brain, 101, 83–117. Hughes, R.N. (1996). Drugs which induce anxiety: caffeine. New Zealand J Psychol, 25, 36–42. Katz, R.J., Landau, M., Lott, A. et al. (1993). Serotonergic (5–HT), medication of anxiety: therapeutic effects of serazepine in generalised anxiety disorder. Biol Psychiatry, 34, 41–4. Ketter, T.A., Andreason, P.J., George, M.S. et al. (1996). Anterior paralimbic mediation of procaine-induced emotional and psychosensory experiences. Arch Gen Psychiatry, 53, 59–69. Knight, R.T. and Nakada, T. (1998). Cortico-limbic circuits and novelty: a review of EEG and blood flow data. Rev Neurosci, 9, 57–70. LaBar, K.S., Gatenby, J.C., Gore, J.C., LeDoux, J.E. and Phelps, E.A. (1998). Human amygdala activation during conditioned fear acquisition and extinction: a mixed-trial fMRI study. Neuron, 20, 937–45. LeDoux, J.E. (1995). Emotion: clues from the brain. Ann Rev Psychol, 46, 209–35. LeDoux, J.E., Cicchetti, P., Xagoraris, A. and Romanski, L.M. (1990). The lateral amygdaloid nucleus: sensory interface of the amygdala in fear conditioning. J Neurosci, 10, 1062–9. Lepola, U., Nousiainen, U., Puranen, M., Riekkinen, R. and Rimón, R. (1990). EEG and CT findings in patients with panic disorder. Biol Psychiatry, 28, 721–7.

238

H.A. Ring and N. Gene-Cos Locatelli, M., Bellodi, L., Perna, G. and Scarone, S. (1993). EEG power modifications in panic disorder during a temporolimbic activation task: relationships with temporal lobe clinical symptomatology. J Neuropsychiatry Clin Neurosci, 5, 409–14. Lucey, J.V., Costa, D.C., Adshead, G. et al. (1997). Brain blood flow in anxiety disorders. Br J Psychiatry, 171, 346–50. Markowitz, J., Weissman, M.M., Ouellette, R. et al. (1989). Panic disorder and quality of life. Arch Gen Psychiatry, 46, 984–92. Mathews, W.S. and Barabas, G. (1981). Suicide and epilepsy: a review of the literature. Psychosomatics, 22, 515–24. Naatanen, R. (1992). Attention and Brain Function. Hillsdale, NJ: Lawrence Erlbaum Associates. Oei, T.P.S., Wanstall, K. and Evans, L. (1990). Sex differences in panic disorder with agoraphobia. J Anxiety Disord, 4, 317–24. Pfefferbaum, A., Roth, W.T. and Ford, J.M. (1995). Event-related potentials in the study of psychiatric disorders. Arch Gen Psychiatry, 52, 559–63. Robbins, T.W. and Everitt, B.J. (1995). Arousal systems and attention. In The Cognitive Neurosciences, ed. M. Gazzaniga, pp. 703–20. Cambridge, MA: MIT Press. Servan-Schreiber, D., Perlstein, W.M., Cohen, J.D. and Mintun, M. (1998). Selective pharmacological activation of limbic structures in human volunteers: a positron emission tomography study. J Neuropsychiatry Clin Neurosci, 10, 148–59. Stein, M.B. and Uhde, T.W. (1989). Infrequent occurrence of EEG abnormalities in panic disorder. Am J Psychiatry, 146, 517–20. Taylor, C.B., Sheikh, J., Agras, W.S. et al. (1986). Ambulatory heart rate changes in patients with panic attacks. Am J Psychiatry, 143, 478–82. Taylor, D.C. and Lochery, M. (1987). Temporal lobe epilepsy; origin and significance of simple and complex auras. J Neurol Neurosurg Psychiatry, 50, 673–81. Tiitinen, H., May, P., Reinikainen, K. and Naatanen, R. (1994). Attentive novelty detection in humans is governed by pre-attentive sensory memory. Nature, 372, 90–92. Weilburg, J.B., Schachter, S., Worth, J. et al. (1995). EEG Abnormalities in patients with atypical panic attacks. J Clin Psychiatry, 56, 358–62. Williams, D. (1956). The structure of emotions reflected in epileptic experiences. Brain, 79, 29–67. Windmann, S. (1998). Panic disorder from a monistic perspective: integrating neurobiological and psychological approaches. J Anxiety Disord, 12, 485–507.

Part V

Treatment complications

16

The effects of antiepileptic drugs on behaviour Bettina Schmitz Humboldt University, Berlin, Germany

Introduction All antiepileptic drugs (AEDs) may have effects on thinking, mood and behaviour in individual patients. These psychotropic effects are not simply idiosyncratic but depend on the drug’s anticonvulsive strength and the person’s biological and psychological predisposition. If a psychiatric disorder occurs in a patient with epilepsy this always has a multifactorial aetiology, anticonvulsant pharmacotherapy being only one of many risk factors. Because of the complex pathogenesis of psychiatric complications in epilepsy, aetiology-related epidemiological data are difficult to obtain, particularly since chronic effects of anticonvulsants are almost impossible to identify. Among a series of consecutive patients who developed either a schizophreniform psychosis or a major depression, peri-ictal syndromes predominated in psychotic patients, and interictal syndromes in depressed patients (Schmitz et al., 1999). Twenty-eight per cent of depressive episodes and 15% of psychotic episodes were attributed to drug treatment, including alternative syndromes, intoxication and withdrawal syndromes. However, these figures may not be representative for today’s praxis because the cases were largely collected prior to the introduction of new AEDs. In a more recent series from Japan, the percentage of AED-related psychoses was significantly higher; 40% of unselected cases had psychoses following a change of their AED regime (half of which occurred with the new anticonvulsant zonisamide, which is not as yet licensed in Europe (Matsuura, 1999)). The psychotropic effects of AEDs can be broken down into those which are negative and those which are positive or prophylactic. The major psychotropic effect of AEDs is via their antiepileptic properties. Therefore, the best prophylaxis of psychiatric complications is early and complete control of seizures. Unfortunately, even with the expanding repertoire of effective AEDs there are many patients in whom complete and lasting seizure control cannot be achieved immediately after first manifestation 241

242

B. Schmitz

of seizures. Patients whose seizures are difficult to treat are particularly at risk to develop psychiatric complications secondary to AED treatment. There are (at least) six questions in the context of severe behavioural adverse events of AEDs, most of which cannot be satisfactorily answered: 1. Drug-related incidence rates: How often and when in the course of treatment are complications to be expected when a specific drug is given? 2. Medication-related issues: What are the differences between monotherapy and polytherapy; and is there a relationship with titration rates and maximum dosages? 3. Psychopathology and outcome of the psychiatric reaction: Are there specific psychiatric syndromes; what is the prognosis, which therapeutic actions ought to be taken? 4. Neurological, epileptological and psychiatric risk factors: How can we identify vulnerable patients? 5. Is there a relationship with cognitive side effects; are there predictable positive psychotropic effects? 6. What are the underlying mechanisms? The aim of this chapter is an overview on the current knowledge on psychiatric effects of anticonvulsants. One of the major methodological problems of studies looking at the relationship between AED and mental state in epilepsy relate to classification and terminology. Unfortunately, diagnostic criteria for psychiatric side effects are neither defined nor standardized. The four major categories used in the literature are psychoses, affective syndromes, behavioural or personality disorders and encephalopathies. From a nosological point of view these categories are not distinct and not specific. The label ‘psychosis’ may stand for a chronic and schizophrenia-like illness, or a short lasting, transient and delirious episode, or a psychosis in the context of a severe, depressive illness. Data from drug trials are usually restricted to isolated psychopathological symptoms, such as nervousness, anxiety, depressed mood or abnormal thinking. The clinical significance and the broader psychiatric context of these remain unclear. An extrapolation towards a specific syndrome like depression would be inappropriate, but is nevertheless often done. Anxiety, for example, is a nonspecific affective symptom occurring with almost all psychiatric syndromes. Classical antiepileptic drugs With respect to the classical antiepileptic drugs there are almost no systematic data on their psychiatric side effects. Our knowledge is largely empirical, based on small case series or anecdotal reports (for a synopsis see Table 16.1). A number of studies have suggested a link between depression and treatment with primidone or phenobarbital both in adults and in children (Brent, 1986; Brent

243

Effects of antiepileptic drugs on behaviour

Table 16.1. Psychotropic effects of antiepileptic drugs

Positive psychotropic effects

Negative affective effects

Psychoses and other complications

Barbiturates, Primidone

—

Aggression, depression, withdrawal syndromes

ADHD in children

Benzodiazepines

Anxiolytic, sedative

Withdrawal syndromes

Disinhibition

Ethosuximide

—

Insomnia

Alternative psychoses

Phenytoin

—

—

Toxic schizophreniform psychoses, encephalopathy

Carbamazepine

Mood stabilizing, impulse control

Rarely mania and depression

—

Valproate

Mood stabilizing, antimanic

—

Acute and chronic encephalopathy

Vigabatrin

—

Aggression, depression, psychosis withdrawal syndromes

ADHD, encephalopathy, alternative psychoses

Lamotrigine

Mood stabilizing, antidepressive

Insomnia

Rarely psychoses

Felbamate

Stimulating?

Agitation

Psychoses possible

Gabapentin

Anxiolytic, antidepressive?

Rarely aggression in children

—

Tiagabine

—

Depression

Nonconvulsive status epilepticus

Topiramate

Mood stabilizing ??

Depression

Psychoses

Levetiracetam

—

—

—

Notes: ? represents minimal data. —, not applicable; ADHD, attention-deficit hyperactivity disorder.

et al., 1987; Robertson et al., 1987). In children a conduct disorder resembling the attention deficit hyperactivity disorder may be provoked by many antiepileptic drugs, the most frequently implicated drugs being, however, the barbiturates. Irritability and aggressive behaviour are side effects particularly often seen in mentally handicapped patients. Phenytoin may provoke schizophrenia-like psychoses at high serum levels (McDanal and Bolman, 1975). These psychoses are dose related, thus toxic syndromes, but they are not associated with cerebellar symptoms which are the most

244

B. Schmitz

common central nervous side effects of phenytoin. A chronic encephalopathy has also been described with phenytoin and has been referred to as ‘Dilantin dementia’ (Trimble and Reynolds, 1976). Psychoses typically following cessation of seizures and associated with a normalization of the EEG occur in 2% of children treated with ethosuximide. The risk for ‘forced normalization’ is higher (8%) in adolescents and adults treated with ethosuximide for persisting absence seizures (Wolf et al., 1984). Affective problems are rare complications of treatment with carbamazepine (Dalby, 1975). These are either depressive disorders or mania, the latter being explained as a paradoxical effect due to the antidepressant properties of carbamazepine which is chemically related to tricyclic antidepressants (Drake and Peruzzi, 1986). Rarely, valproate is associated with acute or chronic encephalopathies (Sackellares et al., 1979; Schöndienst and Wolf, 1992; Zaret and Cohen, 1986). These encephalopathies are related to dose and perhaps polytherapy and are reversible with dose reduction. New antiepileptic drugs In Europe, eight new antiepileptic drugs have been introduced in the last decade. With respect to the psychiatric side effects of these new AEDs, data exist particularly from premarketing studies. However, drug trials are designed to test antiepileptic efficacy and psychiatric adverse events are not systematically reported. Thus severity and psychopathological nature of behavioural problems remain obscure. Further, differences in patients included in trials do not allow comparisons of psychiatric risks of specific drugs, particularly since following the early vigabatrin experience patients with a psychiatric history were often excluded from trials. Finally, these data are not always entirely transparent to the interested epileptologist, at least not as soon and to an extent as one might wish. There are three aspects which need to be differentiated in the analysis of the psychotropic profiles of the new AEDs in epilepsy: the first relates to psychiatric side effects of new AEDs when used in epilepsy patients, the second to the evidence for positive psychotropic effects when these drugs are used in psychiatric patients, and the third relates to data on positive psychotropic effects of new AEDs when used in epilepsy patients. Psychiatric side effects of new antiepileptic drugs Vigabatrin

Of the newer agents vigabatrin has been the most studied, perhaps on account of its being the first to be tested and licensed. Shortly after the introduction of viga-

245

Effects of antiepileptic drugs on behaviour

batrin, a London group had published an incidence of significant psychiatric complications in 7% of treated patients (Sander et al., 1991). Thomas et al. (1996) have analysed case records of psychiatric complications, episodes of psychoses or major depression, reported to the manufacturer of vigabatrin. With respect to psychoses the authors identified three patterns: of a total of 28 psychotic patients, eleven had become seizure free with vigabatrin, six had a postictal psychosis following a cluster of seizures after initial seizure control, possibly related to tolerance, and two psychoses occurred after withdrawal of vigabatrin. In children, particularly in association with learning disabilities, the most common psychiatric side effects are agitation and excitation, hyperkinesia and aggression, a behaviour syndrome similar to that seen with barbiturates. In an early French study the incidence of behavioural disturbances was as high as 26% (Dulac et al., 1991). Since these early reports, the clinical significance of vigabatrin-associated behavioural problems has been a matter of controversy, prompting Ferrie to perform a meta-analysis of psychoses and severe behavioural reactions leading to drug discontinuation, in seven placebo-controlled European studies (Ferrie et al., 1996). The overall incidence of these complications was 3.4% in the vigabatrin group and 0.6% in the placebo group. Although this is a significant difference in statistical terms, the authors conclude that the risk for psychiatric complications is not increased with vigabatrin referring to published incidence rates of psychiatric disorders in epilepsy patients in general. Remarkably, the incidence rates seem to be rather different in different studies, ranging from 1% to 12% suggesting that either the risk is not the same for all patient groups or that the threshold to report psychiatric side effects is not the same among different investigators. Another meta-analysis on the risk for psychosis has only recently been published (Levinson and Devinsky, 1999). These authors used a uniform dictionary and translated psychopathological symptoms described in the investigator forms into standardized psychiatric terminology, which was then summarized into a syndromatic diagnosis. This analysis of American and non-American double-blind studies demonstrated that there is a significantly increased risk for psychosis and particularly for depression. Psychoses occurred in 2.5% of patients treated with vigabatrin compared to an incidence of 0.3% in the placebo group (P < 0.05), and depression occurred in 12.1% of patients treated with vigabatrin in contrast to only 3.5% in the placebo group (P < 0.001). Less than 2% of the patients were taken off vigabatrin suggesting that these were relatively mild manifestations of depression in the majority of patients. It is however important to recognize that the duration of the analysed studies was 3–4 months, later complications are therefore not included. It is a pity that this useful analysis only came out in 1999, since most epileptologists have almost stopped using vigabatrin because of the risk for visual field defects since 1998.

246

B. Schmitz

Psychiatric side effects are not restricted to patients with complicated epilepsies who receive vigabatrin as an add-on treatment. The monotherapy trials showed a significantly increased incidence of depressive disorders in 5% of vigabatrintreated patients as compared to only 1% in a carbamazepine-treated group (Chadwick, 1999). Lamotrigine

In contrast to vigabatrin, lamotrigine has early on gained a reputation of having positive psychotropic properties, improving both mood and cognitive functions. Severe psychiatric complications seem to be uncommon with lamotrigine, and psychosis and depression occurred only in very few cases in the trials (Fitton and Goa, 1995). Insomnia, which may be associated with irritability, anxiety or even hypomania, is the only significant psychiatric side effect, occurring in 6% of patients treated with lamotrigine in monotherapy, compared to 2% in patients treated with carbamazepine and 3% in patients treated with phenytoin (Brodie et al., 1995). When there were reports that carers complained that handicapped patients became more alert and demanding, this was interpreted as reflecting inadequate rehabilitation facilities rather than being a negative side effect (Binnie, 1997). Besag refers to this as a ‘release phenomenon’ (Besag, 2001). There are now, however, a number of reports that children with learning difficulties and adults with mental handicap develop behavioural problems such as aggression with lamotrigine (Beran and Gibson, 1998; Ettinger et al., 1998). More recently there have been reports on the induction of a reversible Gilles de la Tourette syndrome, which in some cases was accompanied by obsessive–compulsive symptoms (Lombroso, 1999). Felbamate

Felbamate is at present only used in a minority of patients, particularly with Lennox–Gastaut syndrome, due to its haematologic and hepatic toxicity. According to the manufacturer psychoses are rare, reported as serious adverse events in 0.02% of all patients treated in 1996 (Essex Pharma, personal communication). Felbamate may lead to increased alertness, inducing sleep problems and behavioural problems related to agitation in some patients, again, particularly in children with learning disabilities (McConnell et al., 1996). Ketter et al. (1999) have specifically investigated psychotropic effects of felbamate. They concluded from their study of 30 refractory epilepsy patients that the stimulating effects of felbamate may be beneficial or negative depending on preexisting psychopathology. Patients with baseline insomnia or anxiety experienced a deterioration in their psychic state, while other children improved.

247

Effects of antiepileptic drugs on behaviour

Gabapentin

Beyond somnolence, negative psychotropic effects have not been demonstrated in the controlled studies of gabapentin, which is a generally well-tolerated but also a relatively weak anticonvulsant. However, there are a number of studies suggesting that gabapentin may induce behavioural problems such as aggression in children with learning disabilities and adults with mental handicap (Lee et al., 1996; Tallian et al., 1996; Wolf et al., 1995). It is not clear whether this could be related to rapid titration. Tiagabine

A specific problem with tiagabine is the paradoxical provocation of de novo nonconvulsive status epilepticus due to a relatively narrow therapeutic window (Schapel and Chadwick, 1996). Therefore, EEG investigations are necessary when behavioural problems arise. Unfortunately, this complication was discovered following the initial trials and it is therefore not possible to know whether psychiatric complications in the trials were related to underlying epileptic activity or not. In the placebo-controlled add-on studies nervousness and depressed mood were both increased in the tiagabine group (Leppik, 1995) (12% vs. 3%, 5% vs. 1%). The incidence for serious adverse events presenting as psychosis was not significantly increased in the tiagabine group (2% vs. 1%). A total of 84 psychoses had been reported to the manufacturer in 1996. In 30 patients tiagabine was withdrawn, in 38 patients tiagabine was reduced, and in 16 cases tiagabine was continued at an unchanged dosage. The records of 19 patients with psychoses classified as serious adverse events have been further analysed (Schmitz, unpublished data). Psychoses occurred after a variable duration of treatment with tiagabine, with a mean of 267 days (range 13–606 days). The mean dosage was 48 mg/day (range 8–80 mg). Seven psychoses occurred postictally. Only one patient had an alternative psychosis following seizure control and one patient became psychotic following tiagabine withdrawal. There was no systematic EEG monitoring in these cases, so nonconvulsive status cannot be excluded and may have gone unrecognized in some cases. The psychoses were of a paranoid-hallucinatory type in 12 cases and duration was less than a week in 10, and less than a month in 4 cases. In summary, for psychoses with tiagabine there are no distinguishable patterns, as described for vigabatrin. Topiramate

Topiramate, promising ‘polytherapy with a single drug’ because of three different modes of action, is a highly effective anticonvulsant, working against a broad spectrum of seizure types. Topiramate also has a high rate of reported side effects. These might in part relate to rapid titration schemes in the earlier studies. However, there

248

B. Schmitz

is no evidence so far that severe psychiatric complications can be avoided by slow titration. An unusual idiosyncratic side effect of topiramate is amnesic or motor aphasia, and in the controlled trials 17–28 % of patients developed symptoms classified as ‘abnormal thinking’ (Janssen-Cilag, 1996). The rate of affective symptoms is clearly dose-dependent with an incidence of 9% and 19% with a daily dose of 200 mg and 1000 mg respectively in one clinical study (Janssen-Cilag, 1996). The high incidence of depressive syndromes with topiramate may be related to cognitive side effects, which are particularly common with this drug. But this relationship has not been studied. Trimble et al. (2000) have looked at the clinical data from a series of patients with psychoses and depression related to topiramate. In this analysis there was no specific pattern for the precipitation of psychoses. Interestingly, depressive syndromes were related to a complete control of seizures in five cases. They also suggested that, in contrast to vigabatrin, at least some of these psychoses were associated with high serum levels, and possible intoxication. Levetiracetam

Levetiracetam is the latest drug, which only recently has been launched in Europe. Data on psychiatric effects of levetiracetam are limited but a preliminary analysis suggests that there are no significant psychiatric risks associated with this drug (Trimble, 2000). Affective disorders were reported in 0.02% and psychosis in 0.007% of patients treated with levetiracetam in clinical trials. Mechanisms

There are a number of theoretical mechanisms linking antiepileptic drugs and psychiatric disorders. These are: (1) dose-related drug toxicity, (2) dose-unrelated or idiosyncratic effects in vulnerable individuals, (3) drug withdrawal and (4) effects related to antiepileptic efficacy (‘forced normalization’). The most important mechanisms both from an epidemiological and a theoretical point of view are idiosyncratic side effects and alternative syndromes associated with the phenomenon of forced normalization. Idiosyncratic, dose-unrelated effects

The mode of action is more or less known for the new substances, if this can be reliably concluded from animal models. In a simplistic scheme, the major antiepileptic mechanisms are either via the sodium channel or GABAergic or antiglutamatergic. It has been pointed out by Trimble that psychiatric problems and in particular depressive disorders are significantly increased with those AEDs which have strong GABAergic properties: vigabatrin, tiagabine and topiramate (Trimble, 1997); all three drugs had an increased rate of depressive symptomatology in placebo-controlled trials.

249

Effects of antiepileptic drugs on behaviour

This observation is not easy to explain but has been used as further evidence for the GABA hypothesis of depression. A number of clinical observations and experimental studies have shown that GABAergic mechanisms are involved in the pathogenesis of depression (Petty, 1995). Trimble’s hypothesis of a link between psychiatric complications and GABAergic mechanisms of AEDs has been extended by Ketter et al. (1994). These authors distinguish two categories of AED, the first being GABAergic with sedating, anxiolytic and antimanic properties. This category comprises barbiturates, benzodiazepines, valproate, vigabatrin, tiagabine and gabapentin. The second category comprises antiglutaminergic drugs which are claimed to have activating, anxiogenic and antidepressive effects: felbamate and lamotrigine. In this scheme topiramate holds an intermediate position because of its multiple mechanisms. The authors suggest that anticonvulsant drugs have different psychiatric effects depending on the preexisting mental status of patients. They predict that patients who are primarily activated may benefit from drugs which belong to the ‘sedating’ category and become worse with ‘activating’ drugs. On the other hand, patients who are primarily sedated would benefit from a drug from the ‘activating’ category, while the same patients would worsen with a ‘sedating’ anticonvulsant. Although this scheme might represent an oversimplification, taking the primary psychopathological status of patients into account explains the sometimes unexpected and seemingly paradoxical effects of some AED in individual patients. Therefore, the consideration of the patient’s preexisting mental state beyond the epileptic syndrome when choosing an AED is a useful approach which deserves further study. Forced normalization

The concept of forced normalization goes back to the publications of Heinrich Landolt, head of the Swiss epilepsy centre in Zurich from 1955 until 1971 (Landolt, 1958). Cases of forced normalization or alternative psychoses have been reported with all of the novel drugs but seem to be particularly common with vigabatrin. There are a number of reports with topiramate, very few with tigabine and lamotrigine, and only one case with gabapentin (Blumer, personal communication). There seems to be a link between the incidence of these alternative syndromes with a number of drugs which happen to be more efficacious than others, when accepting the results of the meta-analysis by Elferink and Van Zwieten Boot (1997). These authors analysed drug trials and compared AEDs by looking at the number of patients which are needed to be treated in order to find one responder. According to this analysis, vigabatrin and topiramate hold relatively good positions while gabapentin and lamotrigine appear to be less effective. The phenomenon of forced normalization is not restricted to drug-induced seizure control. It is likely that in patients who develop de novo psychosis following

250

B. Schmitz

epilepsy surgery, forced normalization plays a role, and recently a first case of an alternative psychosis secondary to vagus nerve stimulation has been published (Gatzonis et al., 2000). Positive psychotropic effects of antiepileptic drugs in psychiatric patients

The positive psychotropic properties of carbamazepine and valproate are well established. Both anticonvulsants are frequently used in psychiatric patients. Carbamazepine is indicated for the prophylaxis of bipolar disorder and the management of episodic dyscontrol, and valproate is particularly useful in the treatment of acute mania (Walden et al., 1998). The antidepressive effects of lamotrigine, which were suspected shortly after its introduction, have been confirmed in controlled studies of patients with bipolar and rapid cycling affective disorder (Calabrese et al., 1999; Kasumakar and Yatham, 1997). However, in its present formulation, lamotrigine is not useful in acute psychiatric disorders because of the slow titration rate. Gabapentin is increasingly being prescribed for an almost unlimited spectrum of psychiatric disorders (Letterman and Markowitz, 1999). This is largely based on positive case reports or small open studies. Controlled studies have established efficacy as yet only in subforms of anxiety disorders (social phobia) (Pande et al., 1999). Tiagabine has been used in only few patients with psychiatric problems, resulting in mixed, but mostly negative, results (Grunze et al., 1999). More promising are the preliminary experiences regarding topiramate. This drug is currently being intensively tested for various psychiatric indications (Marcotte, 1998). Data from open studies in mania and depression have shown positive results (Normann et al., 1999). As yet, there are no data on effects of levetiracetam in psychiatric disorders. Positive psychotropic effects of AEDs in epilepsy patients

Antiepileptic drugs are increasingly being marketed for a broad spectrum of psychiatric disorders. Unfortunately we cannot extrapolate from studies performed with primary psychiatric patients that there are positive psychotropic effects when these drugs are prescribed for epilepsy patients. Most studies in psychiatry are done with patients suffering from bipolar disorder, and this is a relatively uncommon diagnosis in patients with epilepsy. As an example, the well-established mood-stabilizing effects of carbamazepine have never been demonstrated in epilepsy. Given the high incidence of psychiatric comorbidity in epilepsy it would be very useful to know whether antiepileptic drugs have a positive influence on the psychic status of epilepsy patients beyond their influence on seizure activity. However, there is hardly any scientific evidence for this. The methodological problems of four

251

Effects of antiepileptic drugs on behaviour

studies which have looked at psychotropic effects of AED in epilepsy will be discussed below. Ketter et al. (1994) prospectively investigated psychopathology in 32 patients who were openly withdrawn from all antiepileptic drugs. Thirty-eight per cent developed moderate to severe psychopathology, the most prominent symptoms being anxiety and depression. According to the authors, and the complex statistics, this withdrawal emergent psychopathology was not fully explained by an increase in seizures, demographic factors or psychiatric history suggesting pharmacodynamic effects following drug discontinuation. This study is often quoted as evidence for positive psychotropic effects of AEDs. However, the patients were withdrawn from all AEDs within a short period of time in order to start a felbamate monotherapy trial, and were in a very specific anxiety-provoking situation. Therefore, these data should not be used to claim positive psychotropic effects of anticonvulsants. As mentioned before, the improvements in mood and alertness related to lamotrigine treatment are claimed to be independent from seizure control and have been described in many studies. Most of these studies are however uncontrolled, and the controlled data are difficult to interpret. For example, in the early quality of life study by Smith et al. (1993), patients improved only on two psychiatric scales (happiness and alertness), but not on the other four scales which measured self-esteem, mood, anxiety and depression. In another study, the effects of lamotrigine were compared to carbamazepine after 4 weeks of treatment (Gilham et al., 1996) resulting in significant differences on a self-report questionnaire with respect to psychological parameters such as dysphoria and worries. Patients treated with lamotrigine showed significantly better results on these scores than patients in the carbamazepine group. However, given the slow titration rate of lamotrigine in the first 4 weeks of treatment, this result probably reflects the adverse effect profile of carbamazepine rather than positive psychotropic effects of lamotrigine. A study by Harden et al. (1999) claims to demonstrate that gabapentin has a beneficial effect on mood in epilepsy. However, the evidence for this is very weak; the design of this study is open, and patients only improved on one scale, a clinicianrated questionnaire on dysthymia (Cornell), while patients did not improve on self-rating mood scales such as the Beck Depression Inventory. (At least, this weak positive psychotropic effect of gabapentin is likely to be independent from antiepileptic effects, since there were no significant differences with respect to seizure control compared to placebo.) In summary, there is clearly a need for further, well-designed studies (prospective, double-blind and placebo-controlled) looking specifically at the psychotropic effects of anticonvulsants in epilepsy patients.

252

B. Schmitz

Summary and clinical recommendations The risk for psychiatric complications of AEDs is likely to be linked to severity of epilepsy, polytherapy, rapid titration and high dosages of drugs. Patients with previous psychiatric problems or a familial predisposition seem to be specially prone to behavioural side effects. It is important to recognize patients at risk in order to inform them and their families about the possibility of psychiatric side effects, in order to apply a careful titration scheme and to make sure that patients are seen frequently. Psychiatric complications are mild and reversible in most cases, when recognised at an early stage. Risk factors for psychiatric complications are not a strict contraindication for any particular drug, and it is not always necessary to completely withdraw the responsible drug. Depending on the pathophysiology and the severity of the syndrome, a dose reduction or a comedication with a neuroleptic or antidepressive drug may be a good compromise. Psychiatric disorders in epilepsy have a multifactorial aetiology, pharmacotherapy being only one of many risk factors which are both biological and psychosocial. Among psychiatric adverse events of anticonvulsants, a variety of nonpsychotic behavioural problems are reported most commonly, followed by affective disorders, and psychosis being a relatively rare though severe complication. Psychotropic effects of anticonvulsants warrant further research because many relevant parameters related to pathomechanism, frequency, psychopathology and prognosis are not known. Behavioural side-effect profiles, both negative and positive psychotropic effects, ought to be considered in the choice of the optimal antiepileptic drug for an individual patient. Despite the extensive existing literature covering this topic, there is a need for more studies specifically devoted to the psychiatric effects of AEDs in patients with epilepsy. We need these studies in order to better identify patients at risk for severe behavioural reactions with specific drugs and also in order to identify patients who have a good chance to benefit from potentially positive psychotropic effects of AEDs.

R E F E R E N C ES Beran, R.G. and Gibson, R.J. (1998). Aggressive behaviour in intellectually challenged patients with epilepsy treated with lamotrigine. Epilepsia, 39, 280–2. Besag, F.M.C. (2001). Behavioural effects of the new anticonvulsants. Drug Safety, 24, 513–36. Binnie, D.B. (1997). Lamotrigine. In Epilepsy. A Comprehensive Textbook, ed. J. Engel and T.A. Pedley, pp. 1531–40. Philadelphia, New York: Lippincott–Raven.

253

Effects of antiepileptic drugs on behaviour Brent, D.A. (1986). Overrepresentation of epileptics in a consecutive series of suicide attempters seen at a children’s hospital, 1978–1983. J Am Acad Child Psychiatry, 25, 242–6. Brent, D.A., Crumrine, P.K., Varma, R.R., Allan, M. amd Allman, C. (1987). Phenobarbital treatment and major depressive disorder in children with epilepsy. Pediatrics, 80, 909–17. Brodie, M.J., Richens, A. and Yuen, A.W. (1995). Double-blind comparison of lamotrigine and carbamazepine in newly diagnosed epilepsy. UK Lamotrigine/Carbamazepine Monotherapy Trial Group. Lancet, 345, 476–9. Calabrese, J., Bowden, C., Sachs, G., Ascher, J., Monaghan, E. and Rudd, D. (1999). A double blind placebo controlled study of lamotrigine monotherapy in outpatients with bipolar I depression. J Clin Psychiatry, 60, 79–88. Chadwick, D. (1999). Safety and efficacy of vigabatrin and carbamazepine in newly diagnosed epilepsy: a multicentre randomised double-blind study. Vigabatrin European Monotherapy Study Group. Lancet, 354, 13–19. Dalby, M.A. (1975). Behavioral effects of carbamazepine. In Complex Partial Seizures and their Treatment. Advances in Neurology, Vol. 11, ed. J.K. Penry and D.D. Daly, pp. 331–43. New York: Raven Press. Drake, M.E. and Peruzzi, W.T. (1986). Manic state with carbamazepine therapy of seizures. J Natl Med Assoc, 78, 1105–7. Dulac, O., Chiron, D., Cusmai, R., Pajot, N., Beaumont, D. and Mondragon, S. (1991). Vigabatrin in childhood epilepsy. J Child Neurol, 6 (Suppl. 2), 30–7. Elferink, J.A. and Van Zwieten Boot, B.J. (1997). New antiepileptic drugs. Analysis based on number needed to treat shows differences between drugs studied. Br Med J, 314, 603. Ettinger, A.B., Weisbrot, D.M., Saracco, J., Dhoon, A., Kanner, A. and Devinsky, O. (1998). Positive and negative psychotropic effects of lamotrigine in patients with epilepsy and mental retardation. Epilepsia, 39, 874–7. Ferrie, C.D., Robinson, R.O. and Paniotopoulos, C.P. (1996). Psychotic and severe behavioural reactions with vigabatrin: a review. Acta Neurol Scand, 93, 1–8. Fitton, A. and Goa, K.L. (1995). Lamotrigine. Drugs, 50, 691–713. Gatzonis, S.D., Stamboulis, E. and Siafakas, A. (2000). Acute psychosis and EEG normalisation after vagus nerve stimulation. J Neurol Neurosurg Psychiatry, 69, 278–9. Gillham, R., Baker, G., Thompson, P. et al. (1996). Standardisation of a self-report questionnaire for use in evaluating cognitive, affective and behavioural side-effects of anti-epileptic drug treatments. Epilepsy Res, 24, 47–55. Grunze, H., Erfurth, A., Marcuse, A., Amann, B. and Walden, J. (1999). Tiagabine appears not to be efficacious in the treatment of acute mania. J Clin Psychiatry, 13, 194–9. Harden, C.L., Lazar, L.M., Pick, L.H. et al. (1999). A beneficial effect on mood in partial epilepsy patients treated with gabapentin. Epilepsia, 40, 1129–34. Janssen-Cilag (1996). Topiramate. Product monograph. Kasumakar, V. and Yatham, L.N. (1997). An open study of lamotrigine in refractory bipolar depression. Psychiatry Res, 19, 145–8. Ketter, T.A., Nalow, B.A., Flamini, R., White, S.R., Post, R.M. and Theodore, W.H. (1994). Anticonvulsant withdrawal-emergent psychopathology. Neurology, 44, 55–61.

254

B. Schmitz Ketter, T.A., Post, R.M. and Theodore, W.H. (1999). Positive and negative psychiatric effects of antiepileptic drugs in patients with seizure disorders. Neurology, 53 (Suppl. 2), 53–67. Landolt, H. (1958). Serial electroencephalographic investigations during psychotic episodes in epileptic patients and during schizophrenic attacks. In Lectures on Epilepsy, ed. A.M. Lorentz de Haas, pp. 91–131. Amsterdam: Elsevier. Lee, D.O., Steingard, R.J., Cesena, M., Helmers, S.L., Riviello, J.J. and Mikati, M.A. (1996). Behavioral side effects of gabapentin in children, Epilepsia, 37, 87–90. Leppik, E. (1995). Tiagabine: the safety landscape. Epilepsia, 36 (Suppl. 6), 10–13. Letterman, L. and Markowitz, J.S. (1999). Gabapentin: a review of published experience in the treatment of bipolar disorder and other psychiatric conditions. Pharmacotherapy, 19, 565–72. Levinson, D.F. and Devinsky, O. (1999). Psychiatric adverse events during vigabatrin therapy. Neurology, 53, 1503–11. Lombroso, C.T. (1999). Lamotrigine-induced tourettism. Neurology, 52, 1191–4. Marcotte, D. (1998). Use of topiramate, a new anti-epileptic as a mood stabilizer. J Affect Disord, 50, 245–51. Matsuura, M. (1999). Epileptic psychoses and anticonvulsant drug treatment. J Neurol Neurosurg Psychiatry, 67, 231–3. McConnell, H., Snyder, P.J., Duffy, J.D. et al. (1996). Neuropsychiatric side effects related to treatment with felbamate. J Neuropsychiatry Clin Neurosci, 8, 341–6. McDanal, C.E. and Bolman, W.M. (1975). Delayed idiosyncratic psychosis with diphenylhydantoin. J Am Med Assoc, 231, 1063. Normann, C., Langosch, J., Schaerer, L., Grunze, H. and Walden, J. (1999). Treatment of acute mania with topiramate. Am J Psychiatry, 156, 2014–15. Pande, A.C., Davidson, J.R., Jefferson, J.W. et al. (1999). Treatment of social phobia with gabapentin: a placebo-controlled study. J Clin Psychopharmacol, 19, 341–8. Petty, F. (1995). GABA and mood disorders: a brief review and hypothesis. J Affect Disord, 34, 275–81. Robertson, M.M., Trimble, M.R. and Townsend, H.R.A. (1987). Phenomenology of depression in epilepsy. Epilepsia, 28, 364–72. Sackellares, J.C., Lee, S.I. and Dreifuss, F.E. (1979). Stupor following administration of valproic acid to patients receiving other convulsant drugs. Epilepsia, 20, 697–703. Sander, J.W.A.S., Hart, E.M., Trimble, M.R. and Shorvon, S.D. (1991). Vigabatrin and psychosis. J Neurol Neurosurg Psychiatry, 54, 435–9. Schapel, G. and Chadwick, D. (1996). Tiagabine and non-convulsive status epilepticus. Seizure, 5, 153–6. Schmitz, B., Robertson, M. and Trimble, M.R. (1999). Depression and schizophrenia in epilepsy: social and biological risk factors. Epilepsy Res, 35, 59–68. Schöndienst, M. and Wolf, P. (1992). Zur möglichkeit neurotoxischer Spätwirkungen von valproinsäure. In Valproinsäure, ed. G. Krämer and M. Laub, pp. 259–65. Berlin: Springer. Smith, D., Baker, G., Davies, G., Dewey, M. and Chadwick, D. (1993). Outcomes of add-on treatment with lamotrigine in partial epilepsy. Epilepsia, 34, 312–22. Tallian, K.B., Nahata, M.C., Lo, W. and Tsao, C.Z. (1996). Gabapentin associated with aggressive behavior in pediatric patients with seizures. Epilepsia, 37, 501–2.

255

Effects of antiepileptic drugs on behaviour Thomas, L., Trimble, M.R., Schmitz, B. and Ring, H.A. (1996). Vigabatrin and behaviour disorders: a retrospective study. Epilepsy Res, 25, 21–7. Trimble, M.R. and Reynolds, E.H. (1976). Anticonvulsant drugs and mental symptoms: a review. Psychol Med, 6, 169–78. Trimble, M.R. (1997). Neuropsychiatric consequences of pharmacotherapy. In Epilepsy. A Comprehensive Textbook, ed. J. Engel and T.A. Pedley, pp. 2161–70. Philadelphia, New York: Lippincott–Raven. Trimble, M.R. (2000). Anticonvulsants and behaviour: The profile of new drugs with respect to psychosis and depression. Epilepsia, 41 (Suppl. Florence), 114. Trimble, M.R., Rusch, N., Betts, T. and Crawford, P.M. (2000). Psychiatric symptoms after therapy with new antiepileptic drugs: psychopathological and seizure related variables. Seizure, 9, 249–54. Walden, J., Normann, C., Langosch, J., Berger, M. and Grunze, H. (1998). Differential treatment of bipolar disorders with old and new antiepileptic drugs. Neuropsychobiology, 38, 181–4. Wolf, P., Inoue, Z., Röder-Wanner, U.U. and Tsai, J.J. (1984). Psychiatric complications of absence therapy and their relation to alteration of sleep. Epilepsia, 25, 56–9. Wolf, S.M., Shinnar, S., Kang, H., Balaban gil, K. and Moshé, S.L. (1995). Gabapentin toxicity in children manifesting as behavioral changes. Epilepsia, 36, 1203–5. Zaret, B.S. and Cohen, R.A. (1986). Reversible valproic acid-induced dementia: a case report. Epilepsia, 27, 234–40.

17

Antiepileptic drug treatment and epileptic seizures – effects on cognitive function Albert P. Aldenkamp University of Amsterdam, Amsterdam, the Netherlands

Introduction One of the consequences of epilepsy is impairment of cognitive function and memory impairments, mental slowing and attentional deficits are the most frequent disorders (Aldenkamp et al., 1995a; Dodson and Trimble, 1994). Sometimes such cognitive consequences are more debilitating for the individual patients than the seizures. The exact causes of cognitive impairment in epilepsy have not been explored fully, but clearly three factors are involved: aetiology, the seizures and the ‘central’ side effects of treatment (Dodson and Pellock, 1993). Aetiology-related factors are often clear, are separate from epilepsy or concern a small group; especially in the age-dependent encephalopathies, such as the West syndrome (Berg and Shinnar, 1997). We concentrate here on the cognitive effects of seizures and antiepileptic drug treatment. When evaluating these factors it is imperative to realize that in practice most cognitive problems have a multifactorial origin and the three aforementioned factors combined are responsible for most of the ‘make-up’ of a cognitive problem in an individual patient. Moreover the factors are related, which causes therapeutic dilemmas when seizure control can only be achieved with treatments that are associated with cognitive side effects. Nonetheless, these factors are discussed here separately and sequentially. In the last paragraph the overall impact of the factors combined is discussed. Cognitive side effects of antiepileptic drugs Treament of seizures requires antiepileptic drug (AED) treatment for the large majority of patients and may be accompanied by unwanted effects on cognitive function. Although the magnitude of such ‘cognitive side effects’ is generally considered to be mild to moderate for most of the AEDs, their impact may be substan256

257

Drug treatment, seizures and cognitive function

tial in some patients when critical functions are involved such as learning in children (Aldenkamp et al., 1995a) or driving capacities in adults (often requiring milliseconds precision); or when functions are impaired that are already vulnerable, such as memory function in the elderly (Trimble, 1987). Moreover, as the cognitive side effects represent the long-term outcome of AEDs, the effects may increase with prolonged therapy, which contributes to the impact on daily life functioning in refractory epilepsies (Committee on Drugs, 1985). The following topics are relevant to clinical practice: The combined effects of seizures and AEDs on cognitive function Undoubtedly many controversies concern the relative contribution of AED therapy, compared to the effect of seizure activity on cognition. Improved seizure control (when for example a new AED is added into the existing drug regime) may cause cognitive improvement that itself may camouflage genuine cognitive side effects of the new drug. In many situations it will thus be impossible to separate seizure effects from ‘genuine’ AED effects. Subjective patient complaints may enlarge this problem, as patients often believe that their cognitive problems are caused by ‘external’ factors, such as the drugs they have to take instead of by ‘internal’ factors, such as their own seizures. Habituation In most drugs ‘early’ side effects may occur, fortunately only for a short period, i.e. during the first few days or weeks of drug exposure. After this period normalization occurs, possibly due to the development of so-called positive tolerance or habituation (Kulig and Meinardi, 1977). Although little is known about how tolerance to the cognitive effects of AEDs develops, a failure to take this factor into account may lead to overestimation of the negative effects of drugs on cognition. In clinical assessment we should only conclude ‘cognitive side effects’ if these persist during long-term treatment. Subjective patient complaints Most of the studies use formal neuropsychological tests to assess the cognitive side effects of AEDs, although in clinical practice the opportunities for such assessment are usually very limited. Patient complaints, suggestive of problems in cognitive behaviour often represent the only available evidence of possible cognitive dysfunction. Nevertheless, it would be unwise to take all patient complaints that suggest cognitive dysfunction at face value. While some patients may have a clear insight into their own failures, others may have a poor understanding of their performance. Thus, although subjective cognitive complaints are an important factor to be considered, their use may generate discrepancies between patient complaints and the outcome of cognitive tests, a situation that, in our experience, occurs fairly regularly in clinical practice and is a frequent cause for controversies. When changing to another drug, patients may be ‘doing better but

258

A.P. Aldenkamp

feeling worse’, and vice versa, a topic that has been discussed since it was raised in 1890. A possible solution is to use standardized scales to assess the subjective impressions of patients. Several of these scales with proven validity now exist (e.g. the ABNAS neurotoxicity scale; Aldenkamp and Baker, 1997; Aldenkamp et al., 1995b). The relationship with serum levels The specificity of cognitive side effects of AEDs came back into debate recently when several studies failed to find earlier reported cognitive side effects of AEDs when serum concentrations were sufficiently controlled (Meador et al., 1990). When Dodrill and Troupin initiated their research on cognitive side effects of AEDs in 1977 (Dodrill and Troupin, 1977), they reported significant druginduced impairment in patients using phenytoin. However, no impairment remained, when they removed all patients with phenytoin serum levels exceeding the upper limit of the therapeutic reference interval (30 g/ml). This was done in a reanalysis of their original work and published about 15 years later (Dodrill and Troupin, 1991). For other types of AEDs the effects of higher serum levels seem to be milder, but were nonetheless also found for carbamazepine and for valproate. The implication of these findings is that some of the reported drug-induced cognitive impairments and differences between drugs actually may have been due to differences in drug concentrations during the study, rather than representing specific effects on cognition. Although the value of routine serum level monitoring in clinical treatment is criticized it may be helpful to control the serum levels when using high dosing if patients have cognitive complaints. In addition to the effect of dose, the pharmacokinetic properties of antiepileptic drugs may have an effect on cognitive functioning. In carbamazepine, transient cognitive deficits have been detected in relation to high peak serum levels (Aldenkamp et al., 1987). The pharmacokinetic profile of this drug, characterized by rapid and marked fluctuations in serum levels, may differentially affect test performance across short periods during the day. In drugs with large differences between trough and peak levels, a part of the cognitive assessment procedure should be performed repeatedly during the day, allowing a comparison of performance at peak levels with other periods. Generics and branded formulations of one drug It is sometimes suggested that different formulations of the same drug may have different effects on cognitive function (Crawford et al., 1996). This may especially be true for carbamazepine, due to differences in absorption rate and pharmacokinetics among the different formulations. In a recent study, we did not obtain significant differences in cognitive performance when patients were switched from the conventional branded form of carbamazepine to several

259

Drug treatment, seizures and cognitive function

generic formulations (Aldenkamp et al., 1998). Nonetheless, patients’ performance should be monitored when frequent swiches are made between different formulations of an antiepileptic drug. The specific effects of the AEDs on cognitive function have been assessed in a meta-analysis that evaluated 25 years of studies (Vermeulen and Aldenkamp, 1995): Polypharmacy shows most impact on cognitive function as opposed to monotherapy, irrespective of the type of AEDs included. Two drugs with mild cognitive effects may show potentiation of tolerability problems and thus yield serious cognitive impairment, when used in combination (Trimble, 1987). All established AEDs have ‘absolute’ cognitive side effects, i.e. all the investigated drugs have effects when compared to no treatment. These effects are larger for phenobarbitone (PHB) than for phenytoin (PHT), carbamazepine (CBZ) or valproate (VPA). But even the latter drugs, that are generally considered to be drugs with a safe cognitive profile, have cognitive effects, mostly resulting in a mild general psychomotor slowing (Aldenkamp et al., 1993). The respective differences between the four most investigated AEDs: PHB, PHT, CBZ and VPA, can be considered as small, when studied within a normal therapeutic dose, with the exception of the cognitive effects of PHB that has a less favourable cognitive profile when compared to PHT, VPA and CBZ. Possibly the most marked findings were that all AEDs have some cognitive effects and that the effects of PHT are more moderate than has been suggested previously (Meador et al., 1990). The cognitive side effects of each antiepileptic drug. Phenobarbitone is the only AED with specific absolute effects: when compared with no treatment memory functions are affected. In long-term treatment even impairment or delay of psychological development is reported, leading to intellectual deterioration. No information is available on dose-effects. Phenytoin has a much milder impact on cognitive function than had hitherto been suspected. When compared with ‘no treatment’ PHT-induced attentional deficit and mental slowing is observed but these are not markedly different from similar effects of CBZ and VPA. No correlation with dose was found. Carbamazepine has been compared with several drugs and some differences have been found in respect of VPA and PHT. There is some evidence that doseeffects may occur and that controlled-release formulations have a more favourable cognitive profile than conventional forms. No differences were found between different generic forms of CBZ. Valproate has mild psychomotor slowing as an absolute effect and has a similar cognitive profile when compared with CBZ and PHT. In contrast with CBZ, dose-relationships have not been obtained and there is no effect when

260

A.P. Aldenkamp

patients are switched from conventional to controlled-release formulations. For the newer antiepileptic drugs there is a suggestion of equivalence with CBZ and VPA for oxcarbazepine, gabapentin and vigabatrin (although mood effects have been reported for vigabatrin). For the other newer drugs, lamotrigine and tiagabine, no data from controlled studies have been published yet. Clinical anecdotal information suggests mood effects during titration of tiagabine and attentional-deficits in relation to treatment with lamotrigine in children. A recent study (Aldenkamp et al., 2000) showed cognitive impairment in topiramate related to dose escalation speed. Cognitive effects of seizures When evaluating the impact of seizures on cognition we should consider the timescale of such impairments; they may be short-term reactions (direct or acute effects of seizures) or they may persist over time. Somewhere in this time-factor there is the crucial issue of some aspects of reversibility versus irreversibility of the cognitive impairment. Some aspects of this are discussed by Besag (Chapter 6). Short-term or acute effects

Short-term or direct (acute) effects of single seizures are reported on several cognitive functions especially on alertness and short-term learning (Bornstein et al., 1988) even when the period with unconsciousness is short or when there is no or only partially disturbed consciousness. Most impact of the seizures, however, is through postictal effects, with impaired alertness sometimes over considerable periods (Dodrill, 1986). These postictal effects are more difficult to detect and therefore to acknowledge in the patient. Nevertheless, post-ictal effects may be detectable on attentional tests during days after a single tonic-clonic seizure and during several hours after other seizure types (Aldenkamp et al., 1996). In addition to the cognitive effects related to postictal confusion, some studies point to the disruptive effects of epileptic discharges on long-term potentiation (Moore et al., 1993), the physiological process important for learning and especially for stabilizing information in memory. It is thus not just the epileptic seizure (i.e. the actual ictal period) that causes cognitive impairment that explains the sometimes extended periods with cognitive impairment as an aftermath of a seizure, even in patients with short nonconvulsive seizures. This is probably also the reason that repeated nocturnal seizures may have cognitive effects. These are found to have detrimental effects on language functions, on memory and alertness, possibly also through the effects of disturbed sleep patterns (Snead and Saito, 1993). Seizure type must be taken into account, with most effects on cognitive function

261

Drug treatment, seizures and cognitive function

being noted with secondarily generalized tonic-clonic and complex partial seizures (Prevey et al., 1998). One might argue that if a single seizure causes cognitive impairment then more seizures will cause more severe cognitive deficits, an issue that is associated with seizure frequency. Indeed, seizure frequency has been successfully correlated to cognitive impairment in a number of studies (Aldenkamp and Alpherts, 1999). A particularly convincing case is the follow-up of identical twins concordant for incidence of epilepsy, revealing seizure frequency as the only factor associated with cognitive and educational problems (Dodrill and Troupin, 1976). This factor may also explain the sometimes remarkable improvements of cognitive function that are found after starting treatment in children with nonconvulsive seizures. Nonconvulsive seizures are by definition more difficult to detect than convulsive attacks, and there is an effect of having seizures for a considerable period of time before being clinically detected (Mandelbaum and Burack, 1997). Such periods with uncontrolled seizures may even cause deterioration of intelligence (Aldenkamp et al., 1996), although this may be reversible after initiation of treatment (Mandelbaum and Burack, 1997). The conclusion is therefore that single seizures may have an aftermath on cognitive dysfunction over sometimes extended periods (even with short seizures) and ongoing seizure activity may result in impressive cognitive effects due to the accumulating effects of the seizures. This is especially relevant in patients with difficultto-detect seizures as these seizures may accumulate to serious cognitive impairment, without recognition of the source of deterioration. Conversely, seizure control may cause cognitive improvements. Long-term (irreversible?) effects

The direct effects of seizures will thus dissolve in the majority of the patients after the seizures are adequately controlled. Although some experimental studies, using several animal models, found damage to the hippocampus and related limbic structures after a single prolonged generalized or limbic seizure (Meldrum, 1997), there is no convincing evidence for similar effects in humans (Holmes, 1991). Even with high seizure frequency most cognitive effects will therefore dissolve after seizure control is achieved (Hoch et al., 1994). Berg and Shinnar (1997) conclude in their review that epilepsy in humans is usually not a progressive disorder. The discussion about irreversibility of seizure-related cognitive effects thus concentrates on cases with ongoing seizure activity and consequently on factors such as seizure duration, number of seizures before treatment (Camfield et al., 1996), number of seizures during lifetime (Dodrill, 1986) or number of years with seizures (Jokeit and Ebner, 1999). This is very reminiscent of Gowers’ suggested mental decay (‘epileptic dementia’) as a consequence of the pathological long-term

262

A.P. Aldenkamp

sequelae of the seizures. In general it can be concluded that the severity of cognitive impairment is associated with the number of years in which seizures actually occurred (Bourgeois et al., 1983; Rodin et al., 1986). For secondarily generalized seizures, Dodrill (1986) obtained a clear relationship between cognitive deterioration and number of secondarily generalized tonic-clonic seizures during lifetime, with ⬎100 seizures as the crucial cut-off point. For complex partial seizures recent studies do not point to number of seizures but to a ‘time-window’ that allows for normalization of cognitive effects. In the study of Kotagal et al. (1987) irreversible effects were found after 5 years of continuing seizures. Most controlled studies, however, showed a longer ‘time-window’ and point to irreversible cognitive effects after periods of about 20 years with continuing complex partial seizures (Jokeit and Ebner, 1999; Chapter 11). Additionally, different results were reported for patients when reaching seizure remission after a short period with seizures (Seidenberg et al., 1981), revealing cognitive improvement, versus patients reaching remission after an extended period with seizures (Rodin et al., 1986). This latter group had lower outcomes on IQ tests and showed no improvement after remission. The conclusion is therefore that irreversible cognitive effects of seizures occur only as an effect of continuing seizures over longer periods. Some studies suggest the existence of a ‘time-window’: when seizure remission is achieved within this ‘time-window’, cognitive function will normalize to premorbid levels (when aetiology does not interfere). Outside this ‘time-window’, cognitive impairments may become irreversible. This ‘time-window’ has been set for secondarily generalized seizures to be within 100 seizures and for complex partial seizures at ⬎5 years with continuing seizures. This illustrates the need for achieving seizure control before such a ‘time-window’ for irreversibility is exceeded. Conclusion; seizures and drugs Single seizures have an aftermath in regard to cognitive impairment. Although this sometimes affects the individual over extended periods, these effects are generally reversible and dissolve, but may be substantial in patients with high seizure frequency. This may be especially important in patients with difficult-to-detect seizures, as the effect of undetected seizures may accumulate to serious cognitive impairment and even (temporary) deterioration of intelligence. Irreversible cognitive effects have been demonstrated in patients with recurrent seizures that persist over longer periods. On the other hand, all antiepileptic drugs also affect cognitive function to some extent, although this effect is much more limited compared to the effects of the seizures. However, when high dosing or adjunctive polytherapy is needed the balance

263

Drug treatment, seizures and cognitive function

between benefits and disadvantages may be negatively biased against drug treatment. Drug treatment therefore requires careful balancing in the attempt to reach maximal seizure control and avoidance of neurotoxic side effects. The relevance of cognitive impairment in this balance is illustrated in studies using ‘quality of life’ as an outcome measure. The integrity of cognitive function is highly correlated with the possibility of achieving important goals in life, such as satisfactory occupational opportunities and social relationships. Probably this relationship is mediated through an effect of cognitive impairment on education. Whenever cognitive impairments occur, even if these are temporary effects, they may affect educational progress and lead to restricted occupational opportunities in later life (Austin et al., 1999). Sillanpåå et al. (1998) studied the long-term prognosis of seizures, using a 30-year follow-up of 220 patients. The study showed a clear correlation between socio-economic status at endpoint and seizure remission. Whenever seizures continued, adverse social effects occurred and these tended to persist for extended periods of their life even when seizure remission was achieved in a later phase of life. The longer the period with seizures, the more impact on daily life was obtained, but the most crucial period appeared to be the early period immediately after onset of epilepsy, if the epilepsy had its onset in childhood. This confirms the aforementioned relationships and illustrates that early achievement of seizure remission is thus a crucial factor preventing the development of cognitive impairments and consequently adverse educational and social effects of epilepsy.

R E F E R E N C ES Aldenkamp, A.P. and Alpherts, W.C.J. (1999). Psychological assessment. In Handbook of Clinical Neurology, ed. H. Meinardi and G. Bruyn, pp. 387–408. Aldenkamp, A.P. and Baker, G. (1997). The Neurotoxicity Scale-II; Results of a patient-based scale assessing neurotoxicity in patients with epilepsy. Epilepsy Res, 27, 165–73. Aldenkamp, A.P., Alpherts, W.C.J., Moerland, M.C., Ottevanger, N. and van Parijs, J.A.P. (1987). Controlled release carbamazepine: cognitive side effects in patients with epilepsy. Epilepsia, 28, 507–14. Aldenkamp, A.P., Alpherts, W.C.J., Blennow, G. et al. (1993). Withdrawal of antiepileptic medication; effects on cognitive function in children – the results of the multicentre ‘Holmfrid’ study. Neurology, 43, 41–51. Aldenkamp, A.P., Dreifuss, F.E. and Renier, W.O. (1995a). Epilepsy in Children and Adolescents. New York: CRC-Press Publishers. Aldenkamp, A.P., Baker, G., Pieters, M.S.M., Schoemaker, H.C., Cohen, A.F. and Schwabe, S. (1995b). The Neurotoxicity Scale; the validity of a patient-based scale, assessing neurotoxicity. Epilepsy Res, 20, 229–39.

264

A.P. Aldenkamp Aldenkamp, A.P., Overweg, J., Gutter, Th., Beun, A.M., Diepman, L. and Mulder, O.G. (1996). Effect of epilepsy, seizures and epileptiform EEG discharges on cognitive function. Acta Neurol Scand, 93, 253–9. Aldenkamp, A.P., Rentmeester, Th., Hulsman, J. et al. (1998). Pharmacokinetics and cognitive effects of carbamazepine formulations with different dissolution rates. Eur J Clin Pharmacol, 54, 185–92. Aldenkamp, A.P., Baker, G., Mulder, O.G. et al. (2000). A multicentre randomized clinical study to evaluate the effect on cognitive function of topiramate compared with valproate as add-on therapy to carbamazepine in patients with partial-onset seizures. Epilepsia, 41, 1167–78. Austin, J.K., Huberty, T.J., Huster, G.A. and Dunn, D.W. (1999). Does academic achievement in children with epilepsy change over time? Dev Med Child Neurol, 41, 473–9. Berg, A.T. and Shinnar, S. (1997). Do seizures beget seizures? An assessment of the clinical evidence in humans. J Clin Neurophysiol, 14, 102–110. Bornstein, R.A., Drake M.E. Jr and Pakalnis, A. (1988). WAIS-R factor structure in epileptic patients. Epilepsia, 29, 14–18. Bourgeois, B.F.D., Presky, A.L. and Palkes, H.S. (1983). Intelligence in epilepsy: a prospective study in children. Ann Neurol, 14, 438–44. Camfield. C., Camfield, P., Gordon, K. and Dooley, J. (1996). Does the number of seizures before treatment influence ease of control or remission of childhood epilepsy? Not if the number is 10 or less. Neurology, 46, 41–4. Committee on Drugs (1985). Behavioral and cognitive effects of anticonvulsant therapy. Pediatrics, 76, 644–7. Crawford, P., Hall, W.W., Chappell, B., Collings, J. and Stewart, A. (1996). Generic prescribing for epilepsy – is it safe? Seizure, 5, 1–5. Dodrill, C.B. (1986). Correlates of generalized tonic-clonic seizures with intellectual, neuropsychological, emotional, and social function in patients with epilepsy. Epilepsia, 27, 399–411. Dodrill, C.B. and Troupin, A.S. (1976). Seizures and adaptive abilities. A case of identical twins. Arch Neurol, 33, 604–7. Dodrill, C.B. and Troupin, A.S. (1977). Psychotropic effects of carbamazepine in epilepsy: a double-blind comparison with phenytoin. Neurology, 27, 1023–8. Dodrill, C.B. and Troupin, A.S. (1991). Neuropsychological effects of carbamazepine and phenytoin; a reanalysis. Neurology, 41, 141–3. Dodson, W.E. and Pellock, J.M. (1993). Pediatric Epilepsy: Diagnosis and Treatment. New York: Demos Publications. Dodson, W.E. and Trimble, M.R. (1994). Epilepsy and Quality of Life in Epilepsy. New York: Raven Press. Hoch, D.B., Hill, R.A. and Oas, K.H. (1994). Epilepsy and mental decline. Neurol Clin, 12, 101–3. Holmes, G.L. (1991). Do seizures cause brain damage? Epilepsia, 32 (Suppl. 5), S14–18. Jokeit, H. and Ebner, A. (1999). Long term effects of refractory temporal lobe epilepsy on cognitive abilities: a cross sectional study. J Neurol Neurosurg Psychiatry, 67, 44–50. Kotagal, P., Rothner, A.D., Erenberg, G., Cruse, R.P. and Wyllie, E. (1987). Complex partial seizures of childhood onset. A five year follow-up study. Arch Neurol, 44, 1177–80.

265

Drug treatment, seizures and cognitive function Kulig, B. and Meinardi, H. (1977). Effects of antiepileptic drugs on motor activity and learned behavior in the rat. In Advances in Epileptology, ed. H. Meinardi and A.J. Rowan, pp. 98–104. Amsterdam: Swets & Zeitlinger. Mandelbaum, D.E. and Burack, G.D. (1997). The effect of seizure type and medication on cognitive and behavioral functioning in children with idiopathic epilepsy. Dev Med Child Neurol, 39, 731–5. Meador, K.J.M., Loring, D.W., Huh, K. et al. (1990). Comparative cognitive effects of anticonvulsants. Neurology, 40, 391–4. Meldrum, B.S. (1997). First Alfred Meyer Memorial Lecture. Epileptic brain damage: a consequence and a cause of seizures. Neuropathol Appl Neurobiol, 23, 185–201. Moore, S.D., Barr, S.D. and Wilson, W.A. (1993). Seizure-like activity disrupts LTP in vivo. Neurosci Lett, 163, 117–19. Prevey, M.L., Delaney, R.C., Cramer, J.A. and Mattson, R.H. (1998). Complex partial and secondarily seizure patients: cognitive functioning prior to treatment with antiepileptic medication. VA Epilepsy Comparative Study 264 Group. Epilepsy Res, 30, 1–9. Rodin, E.A., Schmaltz, S. and Twitty, G. (1986). Intellectual functions of patients with childhoodonset epilepsy. Dev Med Child Neurol, 28, 25–33. Seidenberg, M., O’Leary, D.S. and Giordani, B. (1981). Test-retest changes of epilepsy patients: assessing the influence of practice effects. J Clin Neuropsychol, 3, 237–55. Sillanpåå, M., Jalava, M., Kaleva, O. and Shinnar, S. (1998). Long-term prognosis of seizures with onset in childhood. N Engl J Med, 338, 1715–22. Snead, O.S. and Saito, M. (1993). Encephalopatic epilepsy after infancy. In Pediatric Epilepsy: Diagnosis and Therapy, ed. W.E. Dodson and J.M. Pellock. New York: Demos Publications. Trimble, M.R. (1987). Anticonvulsant drugs and cognitive function: a review of the literature. Epilepsia, 28, 37–45. Vermeulen, J. and Aldenkamp, A.P. (1995). Cognitive side-effects of chronic antiepileptic drug treatment: a review of 25 years of research. Epilepsy Res, 22, 65–95.

18

Psychiatric effects of surgery for temporal lobe epilepsy Steffi Koch-Stoecker Klinik Mara I, Bielefeld, Germany

Introduction The cessation of epilepsy by surgical intervention – although it may be an end in itself – primarily aims at a reduction of injuries and embarrassment, in order to enable patients to do what they formerly were precluded from doing. It offers possibilities to live a more fulfilled and self-confident life. But ‘if the disasters and injuries are somehow self-imposed, if the embarrassment arises mainly from peculiarities of character, if the sense of stigmatization is a deeply held conviction . . . removing the epilepsy will not necessarily alter these conditions’ (Taylor et al., 1997). Those patients – although seizure-free – may stay in their disabled situation. The experience that the results of epilepsy surgery (ES) – even if successful from a neurological point of view – may be far from satisfying for special groups of patients leads to the following discussion. On the one hand, a growing interest in psychiatric comorbidity has led to better prognostic knowledge about negative outcomes. This led to a tendency to set up contraindications for surgery for those patients who suffered from the most severe psychiatric disorders, namely chronic psychoses, because this patient group continued to exhibit their psychotic features postoperatively, and the positive effects of the surgical intervention seemed doubtful. In principle, this strategy to exclude patients with predictable barriers to surgical success is reasonable, but it is still unclear which group of psychotic patients to preclude from surgery (see below). On the other hand, epileptologists started to consider more than just postoperative seizure frequency for the evaluation of surgical outcome. Evidence for this tendency is to be found in the widely accepted system of outcome classification by Engel et al. (1993). They stated that ‘some degree of consideration for the impact of residual seizures on quality of life is necessary’, which led to a differentiation between the two outcome categories, Class III: ‘Worthwhile improvement’ and Class IV: ‘No worthwhile improvement’. With that differentiation, Engel et al. explicitly went beyond the level of pure seizure-counting and introduced aspects of 266

267

Psychiatric effects of surgery for TLE

quality and subjective value: the same postoperative seizure frequency may be a worthwhile success for one patient but a negligible seizure reduction for another one. With that introduction of aspects of subjective value, Engel et al. have introduced a new direction into the outcome discussion which forces evaluators to take into account the individual experiential conditions and personal assessments of patients postoperatively. It therefore makes sense to go one step further and employ such a ‘worthwhile’-category also with respect to Engel’s outcome Class I, namely the seizure-free patients. Engel et al. (1993) emphasize the present lack of quantitative measures to distinguish between Class III and Class IV. Of course, quality of life research has developed some approaches to comprehend quality with quantitative methods, but solid answers to questions of individual success still need single case reconstructions, including, among others, an understanding of patients’ thoughts and emotions, of sorrows and fears, and of lucid and blind spots in social perception. It is a genuine psychiatric-psychotherapeutic task to conduct such reconstructions, and it requires time, expertise and costs. However, it can make a crucial difference to the indications for and the evaluation of surgery. With the help of the psychiatric concept of personality disorders, some basic predictors for subjective surgical success beyond the number of postoperative seizures can be found. In short, there are at least two aspects to the psychiatric evaluation of surgical candidates: assessment is not only useful for the prognosis of psychiatric complications, but also with respect to the estimation of the subjective value of surgery for individual patients. In the following I shall present results and ask questions concerning both aspects: what do we know about the occurrence of psychiatric disorders in the context of epilepsy surgery, and how can the analysis of individual personality traits be used to explain peculiarities and set up personal aims of patients around surgery? I want to propose how to integrate both aspects into psychiatric assessment in epilepsy centres. Presented data refer to surgical candidates with temporal lobe epilepsies only. Whenever mention is made of our own data (‘Bethel’), I refer to a psychiatric outcome study of the first 100 adult patients who had a temporal lobe resection in our epilepsy centre, Mara I, in Bethel, Bielefeld, whom I have followed up for at least 2 years (Koch-Stoecker, 2001). Psychiatric disorders in the context of temporal lobe resections In order to achieve relevant information about psychiatric diagnoses in the context of epilepsy surgery it is imperative to collect psychiatric data preoperatively and set them into relation with postoperative changes. Moreover, in order to ensure the comparability of results, it would be sensible for different epilepsy centres to

268

S. Koch-Stoecker

Table 18.1. Total psychiatric comorbidity in surgical candidates

Jensen and Larsen 1979: Polkey 1983: Naylor et al., 1994: Manchanda et al., 1996: Ring et al., 1998: Blumer et al., 1998: Glosser et al., 2000: Bethel:

⬎80% 43% 43% 47% 52% 57% 51% (psychiatric syndromes, Axis I DSM–III–R) 43% (psychiatric syndromes, Axis I DSM–III–R); 72% (psychiatric syndromes plus personality disorders)

Note: DSM-III-R; American Psychiatric Association (1987).

attempt an agreement about the diagnostic methods to be used. Those two basic demands were already formulated in the 1960s (Ferguson and Rayport, 1965), but are still unfulfilled today. What is lacking is the integration of psychiatrists into the epilepsy surgery units, who can routinely conduct a psychiatric assessment for all surgical candidates. Instead, in the literature, there are many retrospective analyses. These studies attempt to reconstruct psychiatric disorders only using patients’ notes, in which the reports of psychiatric case histories are often incomplete or even missing. This leads to unreliable results, which are especially susceptible to underestimations and false classifications of psychiatric disorders. Further, there are several studies using psychological questionnaires. Yet, although well constructed and standardized for specific disorders, these questionnaires often are not validadted for problems of patients with epilepsy, and in any case cannot substitute for clinical psychiatric diagnostics. In the following these limitations of some of the presented data are to be kept in mind. Total psychiatric morbidity in patients with temporal lobe surgery Results on total psychiatric comorbidity differ depending on the ways patients are referred to the centres, on the strategies of evaluation, and on the absence or presence of a psychiatric assessment, the latter leading to higher frequencies of psychiatric diagnoses. There are two notable trends. First, comorbidity in surgical candidates is surprisingly high, ranging between 43 and more than 80% (Table 18.1). The most plausible explanation of these findings lies in the fact that mesio-temporal structures which are disturbed in tempo-

269

Psychiatric effects of surgery for TLE

ral lobe epilepsy (TLE) are central to experiencing and processing emotions and are susceptible to the development of dysfunctional cellular connections, with an effect on behaviour. Second, the total amount of psychopathology diminishes only slightly after surgery: stable patients usually stay stable, some deteriorate, some – especially if seizure-free – improve (see below for more details). As for the prognosis of poor psychiatric outcome, in Bethel we have noted the existence of severe personality disorders as a potent indicator. Anhoury et al. (2000) found the presence of preoperative psychiatric disorders, bilateral independent spike discharges, and the size of surgical resections as predictors. Psychoses The term ‘psychosis’ designates – globally spoken – severe psychiatric syndromes, characterized by thoughts, feelings and actions that are incomprehensible for a neutral observer. Diagnoses are often based on spectacular symptoms such as delusions and hallucinations without any further diagnostic differentiation. Given this weak assessment basis, Savard (1991) found in a meta-analysis of diverse studies, preoperative rates of psychoses between 7 and 16%, and postoperatively between 10 and 28%. The importance of an exact classification of psychoses in the context of epilepsy is emphasized by Trimble and Schmitz (1997). They distinguish between ictal, postictal, peri-ictal, interictal and alternative psychoses. Except for interictal psychoses, which require neuroleptic treatment, all other psychoses in epilepsy require a regulation of the antiepileptic drugs as the first therapeutic intervention. Thus, clear diagnoses can save patients from referrals to psychiatric hospitals. Postictal psychoses

Poor diagnostic differentiation between the psychoses, especially between postictal and interictal ones, can have severe consequences for surgical candidates. Thus, without exact psychiatric classification, the already-mentioned tendency to exclude psychotic patients from surgery could mislead surgeons into regarding postictal psychoses as a contraindication for ES. ‘Mislead’ because patients with postictal psychoses can profit from surgery in two ways. If the resection is successful, they lose their seizures. However, in addition they will lose their directly seizure-related psychosis. For these reasons Fenwick (1994) has even suggested that postictal psychoses should be regarded as a psychiatric indication for ES. The neurological condition of these patients is, however, complicated and this may result in a less favourable seizure prognosis. For example, bitemporal (Savard et al., 1991) and extratemporal EEG discharges, clusters of seizures (Umbricht et

270

S. Koch-Stoecker

Table 18.2. Postictal psychoses

Incidence in TLE 4% (Kanemoto et al., 1996) Incidence in surgical candidates 18% (Umbricht et al., 1995); 13% (Kanemoto et al., 1998); 6% (Bethel) Psychiatric outcome Psychoses none (Bethel) Temporary mood disorders 60% (Kanemoto et al., 1998); 50% (Bethel) Seizure outcome (class I) 33%; further 33% after second resection (Bethel)

al., 1995), and nocturnal GTCS (Kanemoto et al., 1996) are all reported. This emphasizes the need for a comprehensive neuropsychiatric evaluation of such cases. The occurrence of postictal psychoses was quoted at 4% in a large study group of more than 800 patients with TLE (Kanemoto et al., 1996), but the incidence in surgical candidates is higher (between 6 and 18%). Whether seizure outcome is less favourable than in the total group needs further evaluation. Results on incidence and postsurgical course are shown in Table 18.2. Chronic interictal psychoses

Because of the widespread reservation about operating on chronic psychotic patients, due to the argument that the psychoses would continue to persist anyhow, and surgery would therefore not be profitable, the number of severe psychotic patients evaluated has become small in epilepsy surgery centres. However, the argument of the positive effects of a seizure reduction in psychotic patients is relevant, and Fenwick (1988) has argued that psychotic patients without seizures could be much ‘better off’ than patients with psychosis and seizures. Moreover there are reports of permanent remittance of psychoses after surgery (Jensen and Larsen, 1979), and enduring deteriorations on the other hand have not been found (Taylor, 1972). Nevertheless, surgical interventions in chronic psychotic patients are complicated. First, the comorbidity of psychosis and TLE may be evidence for an extended

271

Psychiatric effects of surgery for TLE

limbic dysfunction. Second we know that these patients are vulnerable and tend to experience acute exacerbations of their psychosis during stressful life events, like surgery. It is therefore imperative to provide special perioperative care and tailored postoperative rehabilitation settings for such patients (Krahn et al., 1996; Taylor, 1987) in order to prevent acute crises. It is recommended that discussion with such patients includes information about the indications and the aims of surgery but to differentiate between their epilepsy and their psychosis, including the information that the psychosis will most probably continue. With these special preparatory conditions, we have achieved worthwhile results in Bethel. All three chronic psychotic patients we operated on experienced acute psychotic exacerbations after surgery, but in the long run their psychoses became milder in all cases, going along with the seizure reduction. However, only one of the three patients has become seizure-free. Postoperative psychoses

One question is, whether or not there are typical de novo psychoses induced by ES. According to one position, postoperative psychoses primarily occur as so-called de novo postictal psychoses (Savard et al., 1998) in patients with persistent seizures, and thus are only indirectly connected with the surgical event. Another position is that surgery only has the function of a trigger that releases a manifest psychosis, which was already latent and might have found its preoperative expression in paranoid personality traits (Ferguson et al., 1993). However, there are still good arguments for the diagnostic entity ‘de novo psychosis’ as aetiologically linked to the surgical intervention. Mace and Trimble (1991) consider them to be an effect related to a nondominant hemisphere hypofunction, because they predominantly occur after right/nondominant resections. They further argue that the sudden inhibition of seizure activity through surgery may induce mechanisms parallel to those of ‘forced normalization’. Altogether there is no doubt about the occurrence of postoperative psychoses, but again most data concerning aetiology and predictors as well as clinical features have to be interpreted with caution, because of well-known methodological reasons (retrospective analyses, missing reliable preoperative information, problems of classification). Savard (1991) found a spectrum of frequencies between 0.5 and 21%, and Trimble (1992) between 3.8 and 35.7%, with a mean of 7.6%. Concerning morphology, several studies have suggested that gangliogliomas predispose to postoperative psychoses (Andermann et al., 1999; Bruton, 1988), but we did not find such a correlation in our centre. Nondominant temporal foci are frequent (Mace and Trimble, 1991; Bethel) in contrast to an excess of left-sided, dominant lesions noted for chronic psychoses in epilepsy.

272

S. Koch-Stoecker

Table 18.3. Postoperative ‘de novo’ psychoses

Frequency 0.5–21% (meta-analysis, Savard, 1991); 3.8–35.7% (meta-analysis, Trimble, 1992); 11% (Bethel) Morphology Gangliogliomas preferred (Bruton, 1988; Andermann et al., 1999) Laterality Nondominant epileptic focus (Mace and Trimble, 1991); ⬎80% nondominant (Bethel) Preoperative psychopathology 100% personality disorders (Bethel) Symptoms Starting with depressive symptoms, sleep disorders, going on with delusions (frequently after first seizure-relapse) Psychotic contents Coping with surgery/new psychosocial demands (Ferguson and Rayport, 1965) Long-term development Variable: some chronic, some free of psychosis after second resection, some remitting with neuroleptic treatment (Bethel)

Concerning preoperative psychopathology, we found that all patients had personality disorders before surgery (Koch-Stoecker, 1997), which indicates that surgery may be the critical event overwhelming the psychotic threshold in patients with already preoperatively weakened personality structures. Psychoses often start with symptoms of mood and sleep disturbance and then continue with delusions, which are frequently initiated by a seizure-relapse. The psychotic contents mainly relate to two themes (Ferguson and Rayport, 1965): (1) a psychotic structure of impressions of the surgical context (such as suspecting microchips or laser influence in the brain), (2) the apprehension of new psychosocial demands (such as the paranoid ideation that neighbours control the patient’s actions). The long-term course in some cases depends on patients’ compliance to take prescribed medications, in others on further treatment of epilepsy (e.g. reoperation). Many patients respond to neuroleptic medication, but in others there is a necessity for repeated treatment in psychiatric hospitals. These findings are summarized in Table 18.3. Affective disorders and anxiety As is the case for psychoses, it is also true for the affective disorders that the usual psychiatric diagnostic categories do not offer adequate classification for epilepsy

273

Psychiatric effects of surgery for TLE

patients. Typical constellations of symptoms allowing the diagnosis of a ‘major’ depressive episode are rare, dysthymic or ‘organic’ depressive states rather frequent. Blumer and Altshuler (1997) have attempted to introduce a useful classification category of affective disorders in patients with TLE, which they call ‘interictal dysphoric disorder’. Predominant features are, among others, depressive mood, paroxysmal irritability leading to outbursts of verbal aggressivity with consecutive feelings of shame, a sudden onset and a brief duration of only days. Blumer et al. (1998) found that those interictal dysphoric mood disorders – present in 57% of their patients – faded away after surgery in 20% of patients, 36% stayed stable with their dysphoric disorder, and 44% worsened after surgery, some with a remission after antidepressant treatment. There were also about 40% of the psychiatrically intact group, predominantly those who continued to have seizures, who developed their dysphoria after surgery. The enduring remittance of depressive symptoms depends on complete seizure relief (Blumer et al., 1998; Hermann and Wyler, 1989). This finding is supported by our results. Moreover we found that our preoperatively depressed patients showed differences in psychopathology after surgery related to laterality: dominant resections led to somatoform symptoms as surrogates of depression (headache, backache, etc.), while nondominant resected patients frequently had postoperative depressions. Emotional irritation and lability with sudden mood changes, uncertainty concerning the future, reduced stress tolerance etc. during the first months after ES are very typical (41% of all resected patients, Fraser 1988; 45% during the first 6 weeks, Ring et al., 1998). Additionally, circumscribed episodes of depression occur after ES. As early as in 1957 Hill et al. described their occurrence, being independent of seizure outcome, with a remission within the first 18 months after surgery. Because of their temporary character, Trimble (1992) designates them as ‘complications of surgery’. Their frequency is about 8–10% of surgically treated patients (Naylor et al., 1994). They occure more with nonlesional resections or mesio-temporal scleroses (Bruton, 1988), in nondominant resected patients (Fenwick et al., 1993; Bethel) and in preoperatively aggressive patients, who lose their aggressiveness after surgery and tend to develop depressions (Taylor, 1987). There are hints from one research group (Kanemoto et al., 1998) of correlations with dominant resections and with postictal psychoses before surgery. The occurrence of postoperative depression was found to be independent of seizure outcome (Hill et al., 1957), except for psychiatrically preoperatively intact patients in whom depression seems to be linked to seizure recurrence (Blumer et al., 1998). For an overview see Table 18.4. Mania seldom occurs in patients with TLE (Wolf, 1982). However, with respect to postsurgical outcome there are some hints for the occurrence of manic syndromes. Krahn et al. (1996) describe hypomanic states immediately after

274

S. Koch-Stoecker

Table 18.4. Episodes of postoperative depression

Duration Remission within 18 months (Hill et al., 1957) Aetiological hypothesis Process of scarring (Trimble, 1992) Frequency 8–10 % of resected patients (Bruton, 1988; Naylor et al., 1994; Bethel) Morphology Mesio-temporal sclerosis or nonlesional resections (Bruton, 1988) Laterality Nondominant resections (Fenwick et al., 1993; Bethel); dominant resections (Kanemoto et al., 1998) Psychiatric predictors Aggressivity leads to postoperative depression (Taylor, 1987); postictal psychoses leads to postoperative depression (Kanemoto et al., 1998) Seizure outcome Independent occurrence (Hill et al., 1957; Bethel)

surgery and Kanemoto et al. (1998) reported about 10% of the resected patients showing (hypo)manic episodes directly after surgery. It may well be that the incidence of manic disorders is usually underestimated because of two different reasons: (1) the differentiation between optimistic gladness after successful resection and symptomatic euphoria is difficult in some cases; (2) manic symptoms may have already vanished and may not be remembered at the time of the first postoperative evaluation, which in many centres takes place at 3 or 6 months after surgery. Symptoms of anxiety are very common in epilepsy patients, but their classification covers many problems of differentiation between fear of seizures, fear as a symptom of seizures, avoidant behaviours due to stigmatization, fear as a symptom of depression, and others. Accordingly, preoperative estimations of anxiety disorders in candidates for surgery vary between 10% (Manchanda et al., 1996) and 44% (Bladin, 1992). More than 2 years after surgery Koch-Weser et al. (1988) even found higher rates of anxiety than before surgery. Ring et al. (1998) reported a frequency of 42% of early postoperative symptoms of anxiety which had already diminished after 3 months. In the early weeks after surgery, an exact differentiation between the woven symptoms of irritability, anxiety, and mood fluctuation is not easy and might even be impossible. This time period deserves to be better evaluated from the psychopathological perspective.

275

Psychiatric effects of surgery for TLE

Table 18.5. Postoperative nonepileptic attacks

Frequency 10% (Glosser et al., 1999); 5% (Ney et al., 1998); 4% (Bethel) Preferred incidence Gender Women (Glosser et al., 1999; Bethel) Laterality Right (Glosser et al., 1999; Bethel); Left (Ney et al., 1998) Onset-time After adolescence (Glosser et al., 1999) Preoperative psychopathology High (Ney et al., 1998); Borderline personality disorders (Bethel) IQ Low (Ney et al., 1998) Operative complication rate High (Ney et al., 1998)

Nonepileptic attacks (nonepileptic seizures) Most centres are reluctant to operate on patients with epileptic seizures which occur in association with nonepileptic attacks. Even if the epilepsy is cured by surgery, there is a high probability of the dissociative attacks continuing. Therefore, only after nonepileptic attacks are well treated by psychotherapeutic interventions should surgery be considered (Henry and Drury, 1997). Concerning postoperative nonepileptic attack disorders (NEADs), there are some reports in early surveys on the psychiatric effects of epilepsy surgery (Ferguson and Rayport, 1965; Taylor, 1972). After a long period of scientific neglect, they recently are attracting attention again. Glosser et al. (1999) found NEADs in just under 10% of cases within a timeframe of 10 years after epilepsy surgery. They started within the first months after surgery, affected predominantly women, lateralization of resection was right temporal, and the seizure-onset was frequently after adolescence. Ney et al. (1998) found 5% postoperative NEADs, with left lateralization, a high rate of preoperative psychopathology, low IQ and high frequency of perioperative complications. We had 4% postoperative dissociative attacks, all of them right temporal resections, all of them with preoperative borderline personality disorders (Table 18.5).

276

S. Koch-Stoecker

While Glosser et al., interpret the incidence of nonepileptic attacks as somatoform disorders, we would regard them as phenomena of dissociation, which enable the patient to shut out conscious experience temporarily, in distinct situations of unbearable, overwhelming emotions by using mechanisms comparable to epileptic seizure activity. Personality disorders – a gateway to an individual understanding of patients After a period of 20 years of neglect, a growing scientific interest has again been directed to personality disorders during the last decade. One reason is our increasing knowledge about the neurobiological basis of behaviour. Questions about the demarcation of personality disorders from manifest psychiatric syndromes at one end and from normal variants of behaviour patterns at the other end have been discussed, as well as the aetiological components and the predictive value of personality disorders. According to current psychiatric theory, personality disorders represent enduring patterns of thoughts, emotions, and actions which differ considerably from expectations of sociocultural surroundings and lead to impairment and suffering. They usually become manifest during childhood and adolescence. Constitutional, biographic and experience-related conditions are discussed as aetiological factors. With respect to TLE, there has been an ongoing debate about the ‘epileptic personality’, which has been shown to be more harmful than helpful. Yet, there is much evidence that epilepsy patients, especially those with a mesiotemporal seizure focus, show behaviour disturbances, which could partly be seizure-related due to limbic system hyperactivity and interictal inhibitory mechanisms, partly linked to the brain lesion itself, and partly be due to the effects of antiepileptic drugs, etc. (Engel et al., 1991; Chapter 3). To recapitulate our main results on personality disorders in the surgical context: First, 60% of our patients with temporal lobe resections had personality disorders; second, about one-third of all patients with severe personality disorders suffered from postoperative psychiatric deteriorations; third, we had no new psychoses after surgery without preexisting personality disorder (paranoid features in most cases) and finally, we had no new dissociative attacks after surgery without preexisting personality disorders (all borderline type). These results have implications for our preoperative information to the patients. However, we can do more than inform patients about their disorders and warn them about postoperative complications, in order to fulfil the criteria for informed consent to operation. An analysis of the individual development of a personality disorder in each single case permits us to gain an insight into the complex structure of the internal affairs and subjective values of patients and provides hooks for psychotherapeutic interventions.

277

Psychiatric effects of surgery for TLE

How could the development of personality disorders and their neuronal basis be explained?

Many of our neuronal networks are constructed from birth by the repetitive use of cognition and emotion. They differ in complexity depending on the variety of different pathways in use. Optimal stimulation enhances the spectrum of possible reactions, and mechanisms like kindling facilitate the reactive choices. If there are severe limitations of the capacity or function of neuronal connectivity, the person will not always be able to respond appropriately to differentiated situational demands. Instead, they will repeatedly rely on use of available standard reaction types. Such stereotype reactions, provoked due to limitations in limbic connectivity could be the organic basis of personality disorders. Various factors could lead to such limitations. One may be a temporal lobe epilepsy itself, which provokes intermittent overexcitations within limbic structures, with the result of a disturbance in processing emotional reactions. It may perhaps lead to sudden unexplained experiences of fear. The seizure-induced kindling process of fear may then, as a generalizing reaction, facilitate avoidant behaviour and lead to an avoidant personality. The same behaviour strategy could develop as a reaction to punishment-induced fear, or it could be a consequence of feelings of inferiority in social communication due to severe memory deficits, etc. In any of these cases, each single behaviour of the avoidant strategy – however strange it may seem – is selected as the most adaptive of the available alternative reactions, which are reduced due to functional or structural limbic deficits. For that reason the persons themselves will not understand that their behaviour is judged as inconvenient or even as a psychiatric disorder. Beyond the problem of maladaptive behaviour itself, personality disorders involve a reduced stress tolerance and a heightened psychic vulnerability, as an additional result of the limitations due to dysfunctional neuronal connections. Thus it becomes evident that in so-called ‘stressful life events’ processing capacities are easily overwhelmed and the mental system breaks down, which often results in psychotic decompensations. For epilepsy patients with personality disorders, the context of surgery itself is a stressful event. This may facilitate neuronal excitation in unusual directions. In addition, and supporting the escalating process, the surgical disconnection of temporal structures forces other parts of the brain to take over functions during the time of scarring and healing. Thus the postoperative period is a double delicate time-span, involving changes in the cerebral mechanisms of excitation and inhibition. Such a model of interaction of psychosocial and neurobiological factors could be paradigmatic for the development of all psychoses: maladaptive schemata of action and behaviour, acquired by constitutional and/or experiential faults, are preconditions, which emerge as personality disorders. Under special emotional stress

278

S. Koch-Stoecker

conditions they easily run into overstimulation, become dysfunctional and end up in psychotic confusion. According to this model personality disorders change their status from a category of psychiatric diagnosis to meaningful developmental tracks. This may lead to a better understanding of the very special views and values of these patients. CASE REPORT A 40-year-old woman, whose husband has a going concern with the sale of cars, and who has two children aged 20 and 8, exhibits friendly manners without obvious problems in the neurological examination. She seems a bit worried especially when in contact with the nurses. After two seizures at the age of two, epilepsy started at 16, and worsened after her first pregnancy. She has about 6–8 seizures per month in two clusters about the time of ovulation and menstruation. The aura contains massive fear of dying. A right mesio-temporal sclerosis was diagnosed and she was operated on with an optimal prognosis. In fact she became seizure-free, except for some auras (Engel-classification: class I, category B). Now to the postoperative psychiatric situation: she had a severe major depression starting shortly after surgery. The use of antidepressants was limited because she refused to take them after only a few days. She had massive feelings of disgust concerning her husband. About half a year after surgery she recovered from her depression, but then started to throw all conventions overboard and showed manic symptoms. Within the following months she developed the delusional idea of having a love-affair with a neurologist at our centre. She went from manic symptoms to paranoid-hallucinatory experiences of being influenced through the internet in her thoughts and emotions. She left her home and was recently hospitalized against her will and put on neuroleptic drugs. Where are the hints from her biography? She was raised by her mother under poor social conditions with the message: we are poor but proud. Her mother then married again and the girl suddenly was confronted with an aggressive alcoholic stepfather, who had no appreciation for her self-confidence and broke her will. This was when she started to have seizures, which were accompanied by auras of massive fear. She left her home with an unstable conception of the world and immediately became pregnant and married. At that time she had a nervous breakdown and became a psychiatric inpatient. From then on, shortly after the start of her married life, she tried to hide her inner world and to fulfil the demands of a good housewife, showing a well-functioning smiling face to her husband. She also tried to hide her seizures and told people that she suffered from circulatory lability. She suspected neighbours envied her success and invested much energy in fulfilling their imagined expectations. She blamed herself for being uncontrolled and aggressive towards her children. Her expectations about the time after surgery were those of definitely getting the chance to be what she always believed she really was: strong, beautiful, self-confident, the way she had felt before the traumatizing experience with her stepfather. Wasn’t it predictable that the adjustment of her self-confidence would be difficult and that an overestimation of her own personal capacities could result, when the combined

279

Psychiatric effects of surgery for TLE seizure-induced and trauma-induced neuronal pathways of fear and caution were disconnected? That the surgical-induced imbalance of excitation and inhibition could lead to uncontrolled emotional discharges in her case, which had been bundled in the seizure activity before? Couldn’t we have foreseen that she would get into trouble with her husband, when she was seizure-free? That she would search for a dreamlike man who takes care of her and that this picture could easily be projected to the neurologist who set her free of seizures, which symbolized the negative time of her life? Wasn’t it predictable that she would try to free herself from barriers and flee from home, and that she had no real chance to escape except into psychiatry? All this happened. Maybe we could have saved her from at least some of these traps, if we had invested more time analysing the case-history before surgery and insisted on more transparency of her developmental needs, such as the narcissistic feeding by her mother, the unexpected traumatization by the stepfather with the consequence of starting epilepsy and a global distrust and suspicion against men, and later against everybody. Instead we only diagnosed her combined narcissistic and partly paranoid personality disorder and recommended psychotherapy, which did not happen.

A proposal for psychiatric assessment strategies It is neither necessary, nor economic, nor possible to carry out such an extensive analysis for every patient. It would make more sense to have a diagnostic screening for everybody and then decide who needs a detailed examination of organic, biographic and situational aspects of their personality in order to discover the individual traps and try to remove them before surgery, or at least to inform the patient about possible dangers. If therapeutic interventions take place, they must include three steps. First, to find out and accept the different limitations within the individual social and neuronal network. Second, to instruct the patient about the possibility of using different, more adaptive behaviour strategies. Finally, to train the patient to create new pathways and extinguish unsuitable old ones. Unfortunately these possibilities are still seldom attainable.

R E F E R E N C ES American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders (Third edition, revised) (DSM–III–R). Washington, DC: APA. Andermann, L.F., Savard, G., Meencke, H.J., McLachlan, R., Moshé, S. and Andermann, F. (1999). Psychosis after resection of ganglioglioma or DNET: evidence for an association. Epilepsia, 40, 83–7.

280

S. Koch-Stoecker Anhoury, S., Brown, R.J., Krishnamoorthy, E.S. and Trimble, M.R. (2000). Psychiatric outcome following temporal lobectomy: a predictive study. Epilepsia, 41, 1608–15. Bladin, P.F. (1992). Psychosocial difficulties and outcome after temporal lobectomy. Epilepsia, 33, 898–907. Blumer, D. and Altshuler, L.L. (1997). Affective disorders. In Epilepsy: A Comprehensive Textbook, ed. J. Engel Jr and T.A. Pedley, pp. 283–99. Philadelphia, New York: Lippincott–Raven. Blumer, D., Wakhlu, S., Davies, K. and Hermann, B.P. (1998). Psychiatric outcome of temporal lobectomy for epilepsy: incidence and treatment of psychiatric complications. Epilepsia, 39, 478–86. Bruton, C.J. (1988). The Neuropathology of Temporal Lobe Epilepsy. Maudsley Monographs 31. Oxford: Oxford University Press. Engel, J. Jr, Bandler, R., Griffith, N.C. and Caldecott-Hazard, S. (1991). Neurobiological evidence for epilepsy-induced interictal disturbances. In Advances in Neurology, Vol. 55, ed. D. Smith, D. Treiman and M.R. Trimble, pp. 97–111. New York: Raven Press. Engel J. Jr, Van Ness, P.C., Rasmussen, T.B. and Ojemann, L.M. (1993). Outcome with respect to epileptic seizures. In Surgical Treatment of the Epilepsies, Second edition, ed. J. Engel Jr, pp. 609–21. New York: Raven Press. Fenwick, P. (1988). Psychiatric assessment and temporal lobectomy. Acta Neurol Scand, 78 (Suppl. 117), 96–101. Fenwick, P. (1994). Psychiatric assessment and temporal lobectomy. In The Surgical Management of Epilepsy, ed. A.R. Wyler and B.P. Hermann, pp. 217–33. London: Butterworth–Heinemann. Fenwick, P., Blumer, D.P., Caplan, R., Savard, G. and Victoroff, J.I. (1993). Presurgical psychiatric assessment. In Surgical Treatment of the Epilepsies, Second edition, ed. J. Engel Jr, pp. 273–90. New York: Raven Press. Ferguson, S.M. and Rayport, M. (1965). The adjustment to living without epilepsy. J Nerv Ment Disease, 140, 26–37. Ferguson, S.M., Rayport, M., Blumer, D.P., Fenwick, P. and Taylor, D.C. (1993). Postoperative psychiatric changes. In Surgical Treatment of the Epilepsies, Second edition, ed. J. Engel Jr, pp. 649–61. New York: Raven Press. Fraser, R.T. (1988). Improving functional rehabilitation outcome following epilepsy surgery. Acta Neurol Scand, 78 (Suppl. 117), 122–8. Glosser, G., Roberts, D. and Glosser, D.S. (1999). Nonepileptic seizures after resective epilepsy surgery. Epilepsia, 40, 1750–4. Glosser, G., Zwil, A., Glosser, D.S., O’Connor, M.J. and Sperling, M.R. (2000). Psychiatric aspects of temporal lobe epilepsy before and after anterior temporal lobectomy. J Neurol Neurosurg Psychiatry, 68, 53–8. Henry, T.R. and Drury, I. (1997). Non-epileptic seizures in temporal lobectomy candidates with medically refractory seizures. Neurology, 48, 1374–82. Hermann, B.P. and Wyler, A.R. (1989). Depression, locus of control, and the effects of epilepsy surgery. Epilepsia, 30, 332–8. Hill, D., Pond, D.A., Mitchell. W. and Falconer, M.A. (1957). Personality changes following temporal lobectomy for epilepsy. J Ment Sci, 103, 18–27.

281

Psychiatric effects of surgery for TLE Jensen, I. and Larsen, J.K. (1979). Mental aspects of temporal lobe epilepsy. J Neurol Neurosurg Psychiatry, 42, 256–65. Kanemoto, K., Kawasaki, J. and Kawai, I. (1996). Postictal psychosis: a comparison with acute interictal and chronic psychoses. Epilepsia, 37, 551–6. Kanemoto, K., Kawasaki, J. and Mori, E. (1998). Postictal psychosis as a risk factor for mood disorders after temporal lobe surgery. J Neurol Neurosurg Psychiatry, 65, 587–9. Koch-Stoecker, S. (1997). Psychotische phänomene bei patienten mit operativ behandelten temporallappenepilepsien. Epilepsie-Blätter, 10, 32–5. Koch-Stoecker, S. (2001). Psychiatric outcome. In Epilepsy Surgery, Second edition, ed. H. Lüders and Y. Comair, pp. 837–44. Philadelphia: Lippincott Williams & Wilkins. Koch-Weser, M., Garron, D.C., Gilley, D.W. et al. (1988). Prevalence of psychologic disorders after surgical treatment of seizures. Arch Neurol, 45, 1308–11. Krahn, L.E., Rummans, T.A. and Peterson, G.C. (1996). Psychiatric implications of surgical treatment of epilepsy. Mayo Clin Proc, 71, 1201–4. Mace, C.J. and Trimble, M.R. (1991). Psychosis following temporal lobe surgery: a report of six cases. J Neurol Neurosurg Psychiatry, 54, 639–44. Manchanda, R., Schaefer, B., McLachlan, R.S. et al. (1996). Psychiatric disorders in candidates for surgery for epilepsy. J Neurol Neurosurg Psychiatry, 61, 82–9. Naylor, A.S., Rogvi-Hansen, B., Kessing, L. and Kruse-Larsen, C. (1994). Psychiatric morbidity after surgery for epilepsy. J Neurol Neurosurg Psychiatry, 57, 1375–81. Ney, G., Barr, W.B., Napolitano, C., Decker, R. and Schaul, N. (1998). New-onset psychogenic seizures after surgery for epilepsy. Arch Neurol, 55, 726–30. Polkey, C.E. (1983). Effects of anterior lobectomy apart from relief of seizures: a study of 40 patients. J R Soc Med, 76, 354–8. Ring, H.A., Moriarty, J. and Trimble, M.R. (1998). A prospective study of the early postsurgical psychiatric associations of epilepsy surgery. J Neurol Neurosurg Psychiatry, 64, 601–4. Savard, G. (1991). Psychosis and surgery of epilepsy. In Epilepsy Surgery, ed. H. Lüders, pp. 461–5. New York: Raven Press. Savard, G., Andermann, F., Olivier, A. and Remillard, G.M. (1991). Postictal psychosis after partial complex seizures: a multiple case study. Epilepsia, 32, 225–31. Savard, G., Andermann, L.F., Reutens, D. and Andermann, F. (1998). Epilepsy, surgical treatment and postoperative psychiatric complications: a re-evaluation of the evidence. In Forced Normalization and Alternative Psychoses of Epilepsy, ed. M.R. Trimble and B. Schmitz, pp. 179–92. Petersfield and Bristol: Wrightson Biomedical Publishing. Taylor, D.C. (1972). Mental state and temporal lobe epilepsy. A correlative account of 100 patients treated surgically. Epilepsia, 13, 727–65. Taylor, D.C. (1987). Psychiatric and social issues in measuring the input and outcome of epilepsy surgery. In Surgical Treatment of the Epilepsies, ed. J. Engel Jr, pp. 485–503. New York: Raven Press. Taylor, D.C., Neville, B.G.R. and Cross, J.H. (1997). New measures of outcome needed for the surgical treatment of epilepsy. Epilepsia, 38, 625–30. Trimble, M.R. (1992). Behavior changes following temporal lobectomy, with special reference to psychosis. J Neurol Neurosurg Psychiatry, 55, 89–91.

282

S. Koch-Stoecker Trimble, M.R. and Schmitz, B. (1997). The psychoses of epilepsy/schizophrenia. In Epilepsy: A Comprehensive Textbook, ed. J. Engel and T.A. Pedley, pp. 2071–81. Philadelphia, New York: Lippincott-Raven. Umbricht, D., Degreef, G., Barr, W.B., Lieberman, J.A., Pollack, S. and Schaul, N. (1995). Postictal and chronic psychoses in patients with temporal lobe epilepsy. Am J Psychiatry, 152, 224–31. Wolf, P. (1982). Manic episodes in epilepsy. In Advances in Epileptology: 8th Epilepsy International Symposium, ed. H. Akimoto, H. Kazamatsuri, M. Seino and A.A. Ward Jr, pp. 237–40. New York: Raven Press.

19

Vagus nerve stimulation and mood Christian E. Elger and Christian Hoppe Department of Epileptology, University of Bonn, Germany

Introduction When asked for their connotations with the vagus nerve, most medically trained people will think of the parasympathetic efferent tasks of cranial nerve X: heart rate modulation, regulation of ingestion/digestion, effects on lung functioning and so on. And these connotations are what gave this nerve its name: vagus, latin the wanderer. However, the vagus nerve is more a sensory than a motor nerve, since it has about 80% afferent but only 20% efferent fibres (Foley and DuBois, 1937). First reports on the cerebral effects of an electrical stimulation of the vagus were published by Bailey and Bremer in 1938. In the following decades, researchers revealed that vagus nerve stimulation (VNS) may influence surface EEG, suppress epileptic electrical activity in the brain, and even terminate seizures in animal models of epilepsy (Zabara, 1985, 1992). The first single-patient trial on VNS for treatment of epilepsy was set up in 1988 in the USA (Penry and Dean, 1990). The electrical stimulation of the left vagus nerve trunk by a totally implanted stimulation device (NCPTM-system, Cyberonics Inc.) was approved for treatment of drug-resistant epileptic seizures by the FDA in 1997 and by the European Community in 1994 (Schachter and Saper, 1998). The pulse generator is implanted into the chest wall, in a similar fashion to cardiac pacemakers, while the spiral platinum electrodes are attached to the left vagus nerve trunk below the cardiac branch. The pulse generator may be programmed telemetrically by a programming wand that is held over the generator and connected to a portable computer. At standard settings, the stimulator would deliver electrical pulses to the vagus every 5 minutes for about 30 seconds (pulse frequency: 20–30 Hz, pulse width: 250–500 ␮s). Alternatively, rapid cycles with 12 seconds off-stimulation time and 7 seconds onstimulation time may be programmed. The treatment usually gets started with an output currency of 0.25 mA after implantation which can be stepwise increased during the next weeks depending on the seizure outcome and adverse side effects (maximum: 3.5 mA). Several studies have shown effectiveness (Amar et al., 1999; Ben-Menachem et al., 1994; DeGiorgio 283

284

C.E. Elger and C. Hoppe

et al., 2000; Handforth et al., 1998), safety (Annegers et al., 1998; Fisher and Handforth, 1999; Ramsay et al., 1994), and sufficient cost–benefit ratios (Boon et al., 1999) of VNS for patients with intractable seizures. While the antiseizure effect of VNS is usually attributed to the ‘afferent’, that is, direct cerebral effects, the most common adverse effects such as hoarseness, cough, or throat paraesthesia, are supposed to result from the efferent portion of stimulation which is unavoidable since the entire nerve is stimulated. Regarding the issues reviewed here, one should keep in mind that VNS may achieve its effects by an efferent peripheral mechanism as well. Anxiety and depressive disorders are common psychiatric conditions in patients with epilepsy (Jacoby et al., 1996; Kohler et al., 1999). About one-third to one-half of patients score high on anxiety and depression self-report scales, but only onethird of the affected patients are recognized by general practitioners to have psychiatric problems (O’Donoghue et al., 1999). Depressive mood states and poor quality of life are part of a complex interplay of clinical measures (e.g. seizure frequency, seizure severity, epilepsy duration, age at onset) and psychosocial parameters (employment, marital status) (Jacoby et al., 1996; Roth et al., 1994; Smith et al., 1991). However, depression in epilepsy patients may not be fully accounted for by either clinical or psychosocial factors since biological mechanisms involved in epileptogenesis may also contribute to depression (Hermann et al., 1996; Schmitz et al., 1999). Therefore, seizure outcome is only one – even though leading – outcome measure of epilepsy treatment. Psychiatric aspects of epilepsy have to be considered and new drugs and methods, as for example VNS, have to be evaluated for their benefits regarding mood and quality of life as well as any effect on seizures. Mood improvements by VNS in epilepsy patients Reports from the early randomized controlled trials on VNS for epilepsy treatment (EO3, EO5) suggested improved quality of life in a majority of patients (BenMenachem et al., 1994; Handforth et al., 1998): At the 14-weeks follow-up, about 50–60% of the patients stated that their quality of life has improved since implantation. From a psychometric point of view, reliability and validity of these data were questionable and these reports had to be considered as preliminary. But they initiated some studies of this phenomenon which will be reviewed in the following section. Harden et al. (2000) recently published a study on VNS and mood in which 20 epilepsy patients under VNS and 20 control patients were enrolled (a nonrandomized, nonblinded, controlled clinical trial). Mood outcome measures were scores from standard psychiatric rating scales (Cornell Dysthymia Rating Scale; Mason et al., 1993; Hamilton Depression Rating Scale/Hamilton Anxiety Rating Scale; Hamilton, 1960) and from the Beck Depression Inventory (Beck, 1967), an estab-

285

Vagus nerve stimulation and mood

lished self-report questionnaire. Significant improvements of depression scores from baseline to follow-up were observed only in the VNS treatment group. A slight decrement in the score of the Hamilton Anxiety Rating Scale was not significant. No significant mood changes were observed within the control group. No betweengroup differences were obtained at baseline or follow-up and the authors failed to demonstrate a significant interaction effect (group ⫻ time) to confirm their hypothesis; only for the BDI score did the interaction effect approach significance (P⫽0.07). From the within-group changes which were limited to the VNS group, Harden et al. (2000) concluded that their findings, for the first time, showed improvements of depressed mood in epilepsy patients during VNS treatment by means of psychometrically evaluated measures. Elger et al. (2000) completed the EO3 study on seizure outcome of VNS, in which their unit participated with some patients in 1993, with a comprehensive psychiatric evaluation. This international multisite outcome study included 14 weeks of a randomized control trial (RCT). Patients were randomly assigned to a low or high stimulation condition (dose-effect study). In order to give patients from both groups the feeling of participating in an optimal treatment condition, patients from the low stimulation group were told that VNS has optimal effects when subjects are just able to recognize the signal. In contrast, patients from the high stimulation group were informed that VNS should be maximized according to subjective tolerability of adverse effects (with 1.75 mA as the predefined maximum). Seizure and psychiatric data were recorded 4 weeks before implantation (baseline) and at the 3- and 6-month follow-ups. Medication was unchanged during the entire duration of the study. From our unit, 11 patients with severe drug-resistant epileptic seizures enrolled in the EO3 study and agreed to participate in the psychiatric study as well. An experienced psychiatrist, to whom the stimulation conditions were masked, completed several standard rating scales, such as the Montgomery–Åsberg Depression Rating Scale (MADRS; Montgomery and Åsberg, 1979) and the Scale for the Assessment of Negative Symptoms (SANS; Andreasen, 1981). Figure 19.1 depicts improvements of depressive mood states during VNS treatment as revealed by the MADRS. Despite the small sample size, and consequently small statistical power, MADRS changes from baseline to the 6-month follow-up were significant in the total sample (nonparametrical ANOVA, Friedman tests: P⬍0.05; post hoc pairwise analysis by Wilcoxon test: P⬍0.05). Interestingly, patients from the high and low stimulation group tended to experience different courses of mood improvements: in the high stimulation group mood improvements appeared earlier and were already seen at the 3-month follow-up and sustained until the 6-month follow-up. Patients from the low stimulation group experienced significant mood improvements, particularly after finishing the RCT, that is, after output currencies were increased and

286

C.E. Elger and C. Hoppe

ns

ns

Low High Total

⫹P ⬍ 0.10; *P ⬍ 0.05

pre VNS

3 months post

6 months post

Figure 19.1. Improvement of depressed mood during VNS treatment: high (n⫽6) vs. low stimulation (n⫽5).

adapted to the other patients during the second ramp-up period. At the 3-month follow-up, group differences between the MADRS scores approached significance (Mann–Whitney test: P⬍0.10). This finding is consistent with the notion of a (partial) VNS dose–response relationship which may indicate specificity of the observed effects. In addition, data revealed a (partial) independence of mood improvements and the antiseizure effect of VNS: 9 of 11 patients presented mood responses whereas only 3 of 11 patients had about 50% reductions in seizure frequency, that is, mood improvements in 6 of 9 patients could not be attributed to improved seizure control. Both studies have to be regarded as preliminary and the most critical methodological aspects, such as a possible rater bias due to missing attempts of masking the experimental conditions in the Harden et al. (2000) study and the missing control group in the Elger et al. (2000) study, are carefully discussed in both papers. From a rigorous methodological point of view, both studies fail to definitely prove the specificity of the observed effect and to exclude a mere placebo effect. However, it is questionable whether this issue can be taken further: blinding is difficult in surgical treatments such as VNS; arranging an appropriate group of control patients is very questionable; masking VNS patients to their stimulation condition appears almost impossible; and it is unclear whether a placebo condition needs to include surgery as well. Furthermore, even a controlled dose–response study may be difficult to perform since today’s patients are probably well informed about the effects of low and high stimulation, due to comprehensive patient information.

287

Vagus nerve stimulation and mood

In an on-going study we are trying to circumvent these problems by using selfreport questionnaires: here patients are asked to evaluate their present mood states (e.g. last 4 weeks). At the follow-up, 6 months after implantation, they are presumed to be unable to remember their former answers to single items. Mood score changes can be compared to normative data as recorded during construction of the questionnaire, e.g. to exclude a mere regression to the mean effect. VNS as an antidepressant treatment in nonepileptic patients From the evidence reported above it is a small step to the evaluation of VNS for treatment of clinical depressions in nonepileptic patients. Recently, Rush et al. (2000) published their findings from a first single-arm study on this issue. On the basis of ethical considerations, only the most affected patients from different clinics in Dallas were included and provided with a vagus nerve stimulator (n⫽30). Inclusion criteria were a DSM–IV diagnosis of major depression disorder (MDD), bipolar I or bipolar II disorder (American Psychiatric Association, 1994). Patients had to be in a major depressive episode (MDE) which either was lasting ⱖ2 years or which was one of at least four MDEs during life. Finally, they had to have failed on at least two antidepressant medication treatments from different medication groups during the current MDE. The study design was as follows: the baseline period was up to 4 weeks before implantation. Implantation was followed by a 2-weeks recovery period with the stimulation device turned off. During these weeks patients were left unclear about whether the device was already turned on or not. During the next 2 weeks output currency was increased stepwise to the maximum tolerated level (ramp-up period). In the following 8 weeks stimulation was left unchanged (fixed-stimulation period). After implantation patients were evaluated for mood states every week. The main outcome score was the Hamilton Depression Rating Scale (28-item version, HDRS-28; Hamilton, 1960, 1967). A subsequent open clinical observation period lasted at least 9 months in all patients reported on. No effects were obtained during the first 2 weeks after implantation (recovery period with inactive stimulator). Using a ⱖ50% reduction in the HDRS-28 total score to define response, a 40% response rate was found at the end of the 8-weeks fixed stimulation period. Complete response (HDRS-28ⱕ10) was observed in 17% of the patients. Follow-up data from the open clinical observation period showed sustained mood improvement in all responders. In some cases response was further increased so that at the end of the reported observation period, 7 of 10 responders from the acute study demonstrated complete response (23% of the total sample). Since the depressions were very severe and chronic, a mere placebo effect was very unlikely to occur; even with the lack of a control group and despite the small

288

C.E. Elger and C. Hoppe

sample size these findings are intriguing and justify further investigations. From a clinical point of view, these data raise hope for a totally new approach in the treatment of severe clinical depressions, and a multisite double-blind randomized control trial is underway in the USA in which more than 200 severely depressed patients will enrol. Cerebral mechanisms of action A further development of the indications and the effectiveness of VNS depends on a comprehensive understanding of the mechanisms of action underlying VNS. Theoretical considerations have to rely on the anatomical conditions of the vagus (George et al., 2000; Rutecki, 1990), functional imaging studies (Henry et al., 1998, 1999; Vonck et al., 2000), studies on the neurochemical effects of VNS in epilepsy patients (Ben-Menachem et al., 1995; Hammond et al., 1992b; Naritoku et al., 1995; Walker et al., 1999), and behavioural studies on VNS (Clark et al., 1998, 1999). Most researchers assume direct cerebral mechanisms while possible peripheral parasympathetic effects are widely ignored. Several studies with different methodological approaches have demonstrated that VNS may bilaterally activate a wide range of brain regions, as to be expected from the anatomical relations of the vagus: the nucleus of the solitary tract (NTS) as the primary target of afferent vagal input; brain stem regions in the direct vicinity of NTS; thalamus, hypothalamus, amygdala, hippocampus and isocortex. These brain regions are particularly relevant in the context of both epileptogenesis and neuropsychiatric disorders. The details of activation remain somewhat unclear since different studies report slightly different effects, e.g. regarding a possible lateralization of the obtained activations. The behavioural studies of Clark et al. (1998, 1999) have demonstrated dose-dependent enhancement effects of VNS on retention and recognition performance in animals and epilepsy patients, and thus shown the functional impact of these activations. From a clinical point of view, there is the striking convergence of anticonvulsant and antidepressant treatments (Post et al., 1992; Trimble, 1998): several antiepileptic drugs (AEDs), for example carbamazepine (Okuma et al., 1973), valproate (Swann et al., 1997), gabapentin (Harden et al., 1999; Letterman and Markowitz, 1999), and lamotrigine (Calabrese et al., 1999; Suppes et al., 1999), have been shown to be effective also in depression. Vice versa, electroconvulsive therapy (ECT), as one of the most efficacious therapies for drug-resistant depressions (Olfson et al., 1998) is also known for its antiseizure effect (Regenold et al., 1998; Sackheim, 1999; Sackheim et al., 1983). Thus, one may assume biochemical mechanisms which underlie both epileptogenesis and mood disorders, and which are affected and ‘tuned’ by AEDs and ECT – and potentially VNS.

289

Vagus nerve stimulation and mood

Several mechanisms which could explain the anticonvulsant and/or the antidepressant effect of VNS have been proposed (George et al., 2000; Harden et al., 2000). Noradrenergic system

Naritoku et al. (1995) investigated regional c-Fos immunoreactivity and reported activation of the locus coeruleus (LC) by therapeutic VNS. By means of lesion studies in animal models of epilepsy, Krahl and coworkers have demonstrated that the antiseizure effect of injected norepinephrine depends on the vagus (Krahl et al., 2000) and that the seizure-attenuating effect of VNS is mediated by and totally depends on LC activity (Krahl et al., 1998). The locus coeruleus is the major origin of the noradrenergic system in the brain and has projections to brain regions which are involved in both mood regulation and epileptogenesis (e.g. thalamus, hippocampus, amygdala and isocortex). Norepinephrine has an inhibitory influence on postsynaptic neurons which may explain the antiseizure effect of activating the LC by VNS. At the same time, the noradrenergic system is involved in neuropsychiatric disorders like depression: one major effect of tricyclic antidepressant medication is to increment the ‘noradrenergic tone’ of the brain (Schatzberg and Schildkraut, 1996). Thus, roughly speaking, VNS is supposed to stimulate LC neurons, which in turn increase the delivery of norepinephrine which is supposed to be an endogenous antidepressant and anticonvulsant. Serotonergic system

Another hypothesis, with some evidence from neurochemical studies on VNS in epilepsy patients (Ben-Menachem et al., 1995), relies on the role of the serotonergic system in mood regulation. From the NTS there are direct connections to the raphe nuclei which are the main and nearly the only origin of the cerebral serotonergic system. Changes in the ‘serotonergic tone’ of the brain are clearly associated with mood changes. Cortical inhibition

From clinical experience with ECT, a third hypothesis may be derived: inhibition of cortical activity, as the major mechanism of ECT (Sackheim et al., 1996), may enhance depressive mood states and may improve seizure control. Since VNS seems to produce rather similar effects as ECT, one may speculate also that VNS increases the inhibitory influences on cortical tone. However, evidence confirming this speculation and replication of the findings from ECT is missing (Hammond et al., 1992a). GABA and glutamate

The possible role of GABA and glutamate transmission from the NTS was highlighted by an animal study of Walker et al. (1999). These authors tested four

290

C.E. Elger and C. Hoppe

substances for their potential to attenuate seizures in standard animal models of epilepsy (bicuculline methiodide, pentylenetetrazol) when they were microinjected into the mediocaudal NTS (mNTS): a GABA-A receptor agonist (muscimol), a GABA-a receptor antagonist (bicuculline methiodide), a glutamate receptor antagonist (kynurenate) or a local anaesthetic (lidocaine). The findings revealed that increased GABA or decreased glutamate transmission from the NTS reduces susceptibility for limbic seizures. This may provide a possible mechanism for seizure attenuation due to VNS as well as for other secondary limbic alterations which may affect mood regulation. Peripheral mechanisms Since the vagus is a mixed nerve, VNS always comprises a portion of efferent stimulation which may alter peripheral functions. For instance, it may cause hoarseness as the most common adverse effect of VNS. As a matter of course, any changes concerning mood or epileptic seizures must be due to cerebral changes. However, from a theoretical point of view one has good reasons to expect that peripheral changes induced by efferent VNS may in turn result in cerebral changes relevant for the issues discussed here. In mammalians the efferent branch of the vagus plays a decisive role in emotion regulation and expression (Porges 1997; Porges et al., 1994). Furthermore, the vagus is supposed to coordinate and protect the organism’s metabolic resources, e.g. by retarding heart rate more or less (‘vagal brake’) (Porges, 1995). This more theoretical view is confirmed by clinical observations in neuropsychiatric disorders such as depression and anxiety which reveal clear associations between mood and parasympathetic functions (Glassman, 1998; Lehofer et al., 1997, 1999). Diurnal mood variations in some depressed patients may be associated with parasympathetic activity (Rechlin et al., 1995). Regarding cardiac measures, it is noteworthy that there is no evidence for altered vagal tone in unmedicated clinical depressions but for increased sympathetic tone (heart rate) which may be due to increased anxiety in depressed patients (Lehofer et al., 1997; Yeragani et al., 1991). Interestingly, experimentally induced panic attacks (hyperventilation, sodium lactate administration) are accompanied by an attenuated vagal tone (George et al., 1989) suggesting that anxiety disorders may be even more susceptible for VNS treatment than depressions. Other authors have also suggested a linkage between vagal functions and anxiety disorders (Watkins et al., 1998). In animal experiments, one can transiently block efferent neural transmission by a lidocaine injection below the point of electrical stimulation (Brodin, 1985). In such an experiment, Clark et al. (1998) could show that effects on retention and recognition exclusively resulted from the afferent portion of VNS. Investigating the

291

Vagus nerve stimulation and mood

role of efferent vagal transmission in patients is difficult. One would have to record peripheral physiological measures and consider them as covariates during data analysis. Even if there is no evidence for general alterations of cardiac or gastrointestinal functions due to VNS in the sense of adverse effects (Ramsay et al., 1994), peripheral changes induced by VNS may be small and more difficult to register. In the Elger et al. (2000) study, mood improvements were particularly expressed in a reduction of negative symptoms as recorded by the Scale for the Assessment of Negative Symptoms (Andreasen, 1981) or by the anergia scale of the Brief Psychiatric Rating Scale (Overall and Gorham, 1962). We propose that negative symptoms and particularly anergia may be interpreted as a lack of energy in which the autonomic nervous system and particularly the vagus may be involved. Preliminary data of our on-going self-report questionnaire study suggest that VNS improves ‘anxiety’ and ‘unpleasant exertion’ as recorded by the Self-Rating Anxiety Scale (Zung, 1971) or the Befindlichkeits–Skala (Zerssen et al., 1970). In contrast, improvement of depressed mood, which was measured by the Beck Depression Inventory, appears to be a smaller effect. One has to consider that this self-report questionnaire particularly accounts for higher cognitive and emotional aspects of depression. Therefore, we assume that efferent VNS may contribute to mood improvements, first and more unspecifically, by tuning the basic autonomic balance and the vagal management of metabolic resources, or second and more specifically, by attenuating sympathetic tone and peripheral symptoms of anxiety. Finally, we would like to allude to some theoretical difficulties associated with the fact that the vagus is more part of a ‘system’ than two ‘one-way routes’: VNS has an impact on the entire vagal brain–periphery feedback loop and electrical stimulation affects signalling in both directions. A vagus under VNS may make the brain ‘think’ that peripheral functions have changed – even if they actually have not, that is, even if no objective changes can be revealed by psychophysiological measurements. Such a mechanism could be described as virtually peripheral. Conversely, VNS may distort or suggest commands coming from the brain which are to be transmitted to the periphery by the vagus. This virtually cerebral mechanism results in peripheral effects, as for example hoarseness. Studies on the alterations of neural transmission within the vagus as induced by VNS would be required. So far, the artificial stimulation of the vagal system by VNS – with its unphysiological duty cycles, output currents and pulse frequencies – has to be regarded as very coarse. In fact, some authors assume that this is the true reason why more serious cardiac side effects do not occur in patients under VNS (George et al., 2000). A better understanding of vagal neurotransmission will provide the basis for more subtle, more adaptive and hopefully even more effective brain stimulation techniques in the future. VNS is probably the promising beginning of this intriguing development and an important scientific tool for human research on these issues.

292

C.E. Elger and C. Hoppe

R E F E R E N C ES Amar, A.P., DeGiorgio, C.M., Tarver, W.B. and Apuzzo, M.L. (1999). Long-term multicenter experience with vagus nerve stimulation for intractable partial seizures: results of the XE5 trial. Stereotact Funct Neurosurg, 73, 104–8. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (Fourth edition) (DSM–IV). Washington, DC: APA. Andreasen, N.C. (1981). Scale for the Assessment of Negative Symptoms (SANS). Iowa City: University of Iowa. Annegers, J.F., Coan, S.P., Hauser, W.A., Leetsma, J., Duffell, W. and Tarver, B. (1998). Epilepsy, vagal nerve stimulation by the NCP system, mortality, and sudden, unexpected, unexplained death. Epilepsia, 39, 206–12. Bailey, P. and Bremer, F. (1938). A sensory cortical representation of the vagus nerve. J Neurophysiology, 1, 405–12. Beck, A.T. (1967). Depression: Clinical, Experimental, and Theoretical Aspects. New York: Hoeber. Ben-Menachem, E., Manon Espaillat, R., Ristanovic, R. et al. (1994). Vagus nerve stimulation for treatment of partial seizures, 1. A controlled study of effect on seizures (First International Vagus Nerve Stimulation Study Group). Epilepsia, 35, 616–26. Ben-Menachem, E., Hamberger, A., Hedner, T. et al. (1995). Effects of vagus nerve stimulation on amino acids and other metabolites in the CSF of patients with partial seizures. Epilepsy Res, 20, 221–7. Boon, P., Vonck, K., D’Have, M., O’Connor, S., Vandekerckhove, T. and De Reuck, J. (1999). Costbenefit of vagus nerve stimulation for refractory epilepsy. Acta Neurol Belg, 99, 275–80. Brodin, P. (1985). Differential inhibition of A, B and C fibres in the rat vagus nerve by lidocaine, eugenol and formaldehyde. Arch Oral Biol, 30, 477–80. Calabrese, J.R., Bowden, C.L., Sachs, G.S., Ascher, J.A., Monaghan, E. and Rudd, G.D. (1999). A double-blind placebo-controlled study of lamotrigine monotherapy in outpatients with bipolar I depression. Lamictal 602 Study Group. J Clin Psychiatr, 60, 79–88. Clark, K.B., Smith, D.C., Hassert, D.L., Browning, R.A., Naritoku, D.K. and Jensen, R.A. (1998). Posttraining electrical stimulation of vagal afferents with concomitant vagal efferent inactivation enhances memory storage processes in the rat. Neurobiol Learning Memory, 70, 364–73. Clark, K.B., Naritoku, D.K., Smith, D.C., Browning, R.A. and Jensen, R.A. (1999). Enhanced recognition memory following vagus nerve stimulation in human subjects. Nature Neurosci, 2, 94–8. DeGiorgio, C.M., Schachter, S.C., Handforth, A. et al. (2000). Prospective long-term study of vagus nerve stimulation for the treatment of refractory seizures. Epilepsia, 41, 1195–200. Elger, G., Hoppe, C., Falkai, P., Rush, A.J. and Elger, C.E. (2000). Vagus nerve stimulation is associated with mood improvements in epilepsy patients. Epilepsy Res, 42, 203–10. Fisher, R.S. and Handforth, A. (1999). Reassessment: Vagus nerve stimulation for epilepsy: A report of the Therapeutics and Technology Assessment Subcommittee for the American Academy of Neurology. Neurology, 53, 666–9. Foley, J.O. and DuBois, F. (1937). Quantitative studies of the vagus nerve in the cat. I. The ratio of sensory and motor studies. J Comp Neurol, 67, 49–67.

293

Vagus nerve stimulation and mood George, D.T., Nutt, D.J., Walker, W.V., Porges, S.W., Adinoff, B. and Linnoila, M. (1989). Lactate and hyperventilation substantially attenuate vagal tone in normal volunteers. Arch Gen Psychiatry, 46, 153–6. George, M.S., Sackheim, H.A., Rush, J.A. et al. (2000). Vagus nerve stimulation: a new tool for brain research and therapy. J Biol Psychiatry, 47, 287–95. Glassman, A.H. (1998). Depression, cardiac death, and the central nervous system. Neuropsychobiology, 37, 80–3. Hamilton, M. (1960). A rating scale for depression. J Neurol Neurosurg Psychiatry, 12, 371–9. Hamilton, M. (1967). Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol, 6, 278–96. Hammond, E.J., Uthman, B.M., Reid, S.A. and Wilder, B.J. (1992a). Electrophysiological studies of cervical vagus nerve stimulation in humans: I. EEG effects. Epilepsia, 33, 1013–20. Hammond, E.J., Uthman, B.M., Wilder, B.J. et al. (1992b). Neurochemical effects of vagus nerve stimulation in humans. Brain Res, 583, 300–3. Handforth, A., DeGiorgio, C.M., Schachter, S.C. et al. (1998). Vagus nerve stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology, 51, 48–55. Harden, C.L., Lazar, L.M., Pick, L.H. et al. (1999). A beneficial effect on mood in partial epilepsy patients treated with gabapentin. Epilepsia, 40, 1129–34. Harden, C.L., Pulver, M.C., Ravdin, L.D., Nikolov, B., Halper, J.P. and Labar, D.R. (2000). A pilot study of mood in epilepsy patients treated with vagus nerve stimulation. Epilepsy Behav, 1, 93–9. Henry, T.R., Bakay, R.A.E., Votaw, J.R. et al. (1998). Brain blood flow alterations induced by therapeutic vagus nerve stimulation in partial epilepsy: acute effects at high and low levels of stimulation. Epilepsia, 39, 983–90. Henry, T.R., Votaw, J.R., Pennell, P.B. et al. (1999). Acute blood flow changes and efficacy of vagus nerve stimulation in partial epilepsy. Neurology, 52, 1166–73. Hermann, B.P., Trenerry, M.R. and Colligan, R.C. (1996). Learned helplessness attributional style and depression in epilepsy (Bozeman Epilepsy Surgery Consortium). Epilepsia, 37, 680–6. Jacoby, A., Baker, G.A., Stehen, N., Potts, P. and Chadwick, D.W. (1996). The clinical course of epilepsy and its psychosocial correlates: findings from a UK Community study. Epilepsia, 37, 148–61. Kohler, C., Norstrand, J.A., Baltuch, G. et al. (1999). Depression in temporal lobe epilepsy before epilepsy surgery. Epilepsia, 40, 336–40. Krahl, S.E., Clark, K.B., Smith, D.C. and Browning, R.A. (1998). Locus coeruleus lesions suppress the seizure attenuating effects of vagus nerve stimulation. Epilepsia, 39, 709–14. Krahl, S.E., Senanayake, S.S. and Handforth, A. (2000). Seizure suppression by systemic epinephrine is mediated by the vagus nerve. Epilepsy Res, 38, 171–5. Lehofer, M., Moser, M., Hoehn-Saric, R. et al. (1997). Major depression and cardiac autonomic control. Biol Psychiatry, 42, 914–19. Lehofer, M., Moser, M., Hoehn-Saric, R. et al. (1999). Influence of age on the parasympatholytic property of tricyclic antidepressants. Psychiatry Res, 85, 199–207. Letterman, L. and Markowitz, J.S. (1999). Gabapentin: a review of published experience in the treatment of bipolar disorder and other psychiatric conditions. Pharmacotherapy, 19, 565–72.

294

C.E. Elger and C. Hoppe Mason, B., Kocsis, J., Leon, A. et al. (1993). Measurement of severity and treatment response in dysthymia. Psychiatr Ann, 23, 625–31. Montgomery, S.A. and Åsberg, M. (1979). A new depression scale designed to be sensitive to change. Br J Psychiatry, 134, 382–9. Naritoku, D.K., Terry, W.J. and Helfert, R.H. (1995). Regional induction of Fos immunoreactivity in the brain by anticonvulsant stimulation of the vagus nerve. Epilepsy Res, 22, 53–62. O’Donoghue, M.F., Goodridge, D.M., Redhead, K., Sander, J.W. and Duncan, J.S. (1999). Assessing the psychosocial consequences of epilepsy. A community-based study. Br J Gen Pract, 49, 211–14. Okuma, T., Kishimoto, A., Inoue, K., Matsumoto, H. and Ogura, A. (1973). Anti-manic and prophylactic effects of carbamazepine (Tegretol) on manic depressive psychosis. A preliminary report. Folia Psychiatr Neurol Jpn, 27, 283–97. Olfson, M., Marcus, S., Sackheim, H.A., Thophson, J. and Pincus, H.A. (1998). Use of ECT for the inpatient treatment of recurrent major depression. Am J Psychiatry, 155, 22–9. Overall, J.E. and Gorham, D.R. (1962). The Brief Psychiatric Rating Scale. Psychol Rep, 10, 799–812. Penry, J.K. and Dean, J.C. (1990). Prevention of intractable partial seizures by intermittent vagal stimulation in humans: preliminary results (Abstract). Epilepsy, 31 (Suppl.), S40–3. Porges, S.W. (1995). Orienting in a defensive world: mammalian modifications of our evolutionary heritage. A polyvagal theory. Psychophysiology, 32, 301–18. Porges, S.W. (1997 ). Emotion: an evolutionary by-product of the neural regulation of the autonomic nervous system. Ann New York Acad Sci, USA, 807, 62–77. Porges, S.W., Doussard-Roosevelt, J.A. and Maiti, A.K. (1994). Vagal tone and the physiological regulation of emotion. Monogr Soc Res Child Dev, 59, 167–86. Post, R.M., Weiss, S.R.B. and Chuang, D.M. (1992). Mechanisms of action of anticonvulsant in affective disorders: comparisons with lithium. J Clin Psychopharmacol, 12 (Suppl. 1), S23–35. Ramsay, R.E., Uthman, B.M., Augustinsson, L.E. et al. (1994). Vagus nerve stimulation for treatment of partial seizures, 2. Safety side effects and tolerability (First International Vagus Nerve Stimulation Study Group). Epilepsia, 35, 627–36. Rechlin, T., Weis, M. and Kaschka, W.P. (1995). Is diurnal variation of mood associated with parasympathetic activity? J Affect Disord, 34, 249–55. Regenold, W.T., Weintraub, D. and Taller, A. (1998). Electroconvulsive therapy for epilepsy and major depression. Am J Geriatr Psychiatry, 6, 180–3. Roth, D.L., Goode, K.T., Williams, V.L. and Faught, E. (1994). Physical exercise, stressful life experience and depression in adults with epilepsy. Epilepsia, 35, 1248–55. Rush, A.J., George, M.S., Sackheim, H.A. et al. (2000). Vagus nerve stimulation (VNS) for treatment-resistant depression, a multicenter study. J Biol Psychiatry, 47, 276–86. Rutecki, P. (1990). Anatomical, physiological, and theoretical basis for the antiepileptic effect of vagus nerve stimulation. Epilepsia, 31, S1–6. Sackheim, H.A. (1999). The anticonvulsant hypothesis of the mechanism of action of ECT, current status. J ECT, 15, 5–26. Sackheim, H.A., Decina, P., Malitz, S., Reesor, S.R. and Prohovnik, I. (1983). Anticonvulsant and

295

Vagus nerve stimulation and mood antidepressant properties of electroconvulsive therapy, a proposed mechanism of action. Biol Psychiatry, 18, 1301–10. Sackheim, H.A., Luber, B., Katzman, G.P. et al. (1996). The effects of electroconvulsive therapy on quantitative electroencephalograms, relationship to clinical outcome. Arch Gen Psychiatry, 53, 814–24. Schachter, S.C. and Saper, C.B. (1998). Vagus nerve stimulation. Epilepsia, 39, 677–86. Schatzberg, A.F. and Schildkraut, J.J. (1996). Recent studies on norepinephrine systems in mood disorders. In Psychopharmacology, 4th Generation of Progress, ed. F. Bloom and D. Kupfer, pp. pp. 911–20. New York: Raven Press. Schmitz, E.B., Robertson, M.M. and Trimble, M.R. (1999). Depression and schizophrenia in epilepsy, social and biological risk factors. Epilepsy Res, 35, 59–68. Smith, D.F., Baker, G.A., Dewey, M., Jacoby, A. and Chadwick, D.W. (1991). Seizure frequency, patient-perceived seizure severity and the psychosocial consequences of intractable epilepsy. Epilepsy Res, 9, 231–41. Suppes, T., Brown, E.S., McElroy, S.L. et al. (1999). Lamotrigine for the treatment of bipolar disorder, a clinical case series. J Affect Disord, 53, 95–8. Swann, A.C., Bowden, C.L., Morris, D. et al. (1997). Depression during mania. Treatment response to lithium or divalproex. Arch Gen Psychiatry, 54, 37–42. Trimble, M.R. (1998). New antiepileptic drugs and psychopathology. Neuropsychobiology, 38, 149–51. Vonck, K., Boon, P., Van Laere, K. et al. (2000). Acute single photon emission computed tomographic study of vagus nerve stimulation in refractory epilepsy. Epilepsia, 41, 601–9. Walker, B.R., Easton, A. and Gale, K. (1999). Regulation of limbic motor seizures by GABA and glutamate transmission in nucleus tractus solitarius. Epilepsia, 40, 1051–7. Watkins, L.L., Grossman, P., Krishnan, R. and Sherwood, A. (1998). Anxiety and vagal control of heart rate. Psychosom Med, 60, 498–502. Yeragani, V.K., Pohl, R., Balon, R. et al. (1991). Heart rate variability in patients with major depression. Psychiatry Res, 37, 35–46. Zabara, J. (1985). Peripheral control of hypersynchronous discharge in epilepsy. Electroencephalogr Clin Neurophysiol, 61s, S162. Zabara, J. (1992). Inhibition of experimental seizures in canines by repetitive vagal stimulation. Epilepsia, 33, 1005–12. Zerssen, D.v., Koeller, D.M. and Rey, E.R. (1970). Die befindlichkeits-skala Bf-S – ein einfaches instrument zur objektivierung von befindlichkeits-störungen insbesondere im rahmen von längsschnitt-untersuchungen. Arzneimittelforschung, 20, 915–18. Zung, W.W.K. (1971). A rating instrument for anxiety disorders. Psychosomatics, 12, 371–9.

Part VI

Treatment

20

On the use of psychotropic drugs in patients with seizure disorder M.R. Trimble1 and Anke Hensiek2 1 2

Institute of Neurology, London, UK Addenbrooke’s Hospital, Cambridge, UK

Introduction It is now accepted that many patients with epilepsy have psychiatric problems. Recent epidemiological evidence from selected clinics suggests that over 50% of patients may have a recognizable psychiatric disorder (Krishnamoorthy and Trimble, unpublished data). It is also known that many patients with epilepsy receive psychotropic drugs, sometimes, but not always, on account of their psychiatric symptoms. Thus, it has to be acknowledged that there is an overlap between anticonvulsant drugs and psychotropic agents, such that many of the former are known to have mood-regulating properties, while a number of the latter (for example, benzodiazepines) have anticonvulsant properties. A classification of psychotropic drugs currently in use is given in Table 20.1. The main part of this text relates to the prescription of antidepressant and antipsychotic drugs in patients with epilepsy, and then a brief comment will be made about some of the other agents. Antidepressant drugs seizures and epilepsy Ever since the introduction of tricyclic drugs into clinical practice, seizures have been recognized as a side effect. This has been reviewed on several occasions (Trimble, 1980, 1987). The position has changed in the last few years because of the introduction of a number of new antidepressant drugs, and following a brief review of the older literature, these will be discussed. A review of some early studies Animal data

A number of laboratory and clinical investigations were carried out to assess the effect of tricyclic drugs on seizures and the seizure threshold. The early laboratory 299

300

M.R. Trimble and A. Hensiek

Table 20.1. Classification of psychotic drugs

Antidepressants Antipsychotics Minor tranquillizers Mood stabilizers Psychostimulants Others (beta blockers etc.)

investigations essentially revealed the proconvulsant effect of these agents, and it emerged from these studies that clomipramine, amitriptyline and maprotiline were probably the most proconvulsant compounds (Trimble, 1987). Luchins et al. (1984) used spike activity in perfused guineapig hippocampal slices as an indication of epileptogenicity and reported the effect of a variety of antidepressants. Imipramine, amitriptyline, nortriptyline, desipramine and maprotiline generally increased spike activity, while viloxazine, protriptyline and trimipramine appeared to decrease neuronal excitability. Nomifensine, a drug no longer available, had a biphasic effect, increasing excitability initially, and then producing cessation of spikes. Similarly, doxepin produced a significant increase in excitability, and then significant decreases. In this model mianserin had no effect. Using the photosensitive baboon, Papio papio, Trimble et al. (1977) compared two tricyclic drugs, namely clomipramine and imipramine, and the quadricyclic maprotiline with the nontricyclic nomifensine. The first three all lowered the seizure threshold, while nomifensine had little effect and in some animals was anticonvulsant. One interesting drug examined in these early studies was viloxazine. In two animal models (Luchins et al., 1984; Meldrum et al., 1982) it was observed, if anything, to have a seizure-protective effect. These early animal studies therefore suggested the potential for both anticonvulsant and proconvulsant effects of these drugs; in some compounds the potential was a dose-related effect. Clinical data

The clinical data at that time came mainly from reports of government agencies such as the Committee for the Safety of Medicines (CSM), and from clinical trial data. Several reviews emphasized the poor quality of the available information (Edwards, 1985). However, from the clinical studies the highest reporting of seizures was with maprotiline and clomipramine, and the lowest reporting with protriptyline.

301

The use of psychotropic drugs in seizure disorder

The estimated risk of seizures with tricyclic drugs was around 0.06 to 0.1% (Burley, 1977; Jick et al., 1983) The incidence of seizures with imipramine was 0.7%, and with clomipramine 3.0%. Garvey and Tollefson (1987) drew attention to myoclonic seizures that could occur in relationship to tricyclic drug prescribing, and suggested that as many as 40% of patients reported some kind of myoclonic event after starting them. They calculated the frequency of the effect with the different drugs in descending order from maprotiline, trazodone, nortriptyline, desipramine, amitriptyline to imipramine. Doxepin was not associated with this effect. Other information that emerged from the early clinical studies was the low reporting of seizures with viloxazine (Edwards and Glen-Bott, 1984), and for proconvulsant drugs a relationship of seizure reporting to the therapeutic dose, with higher doses of the drug having a higher frequency of seizures, a clear relationship to overdose, and the relationship of seizures to the number of psychotropic drugs prescribed. The time relationship between commencing the drug and the seizures varies considerably between studies. The attack may occur from 24 hours to several weeks after starting an antidepressant, although early seizures (less than a week later) would seem to be associated with lower dosing schedules and possibly more patient-related factors that lower the seizure threshold (Trimble, 1980) than lateronset seizures. It was generally concluded that patients were more likely to have seizures (if they did not have epilepsy) if they had a family history of seizures, or a past history of relevant medical conditions such as a head injury or a cerebrovascular accident. Few of the above antidepressant drugs were tried in patients with epilepsy. Viloxazine was studied, but it was shown to easily lead to anticonvulsant drug toxicity, and was therefore not recommended in epilepsy (Pisani et al., 1984). Paradoxically, there were some clinical reports of tricyclic antidepressants being anticonvulsant. Ojemann et al. (1983) reported retrospective data, which suggested that doxepin improved seizure frequency in 15 of 19 patients who were prescribed the drug. This included a diminution of both partial and generalized tonic-clonic seizures. Conclusions from these earlier clinical studies were that, in general, antidepressant drugs were proconvulsant, although they were not all proconvulsant to the same degree, and some may not alter the seizure threshold or have a biphasic effect, sometimes revealing some anticonvulsant effects. Table 20.2 shows factors which were thought to be interlinked with lowering of the seizure threshold, emphasizing that this is a problem of the use of these drugs which is not confined only to epilepsy. There are many patients without epilepsy, who have a lowered seizure threshold, who are susceptible to psychotropic-induced seizures.

302

M.R. Trimble and A. Hensiek

Table 20.2. Factors that lower the seizure threshold

A family history of epilepsy Head injury, especially with a prolonged posttraumatic amnesia, or intracranial injury Neurological illness

Table 20.3. Newer antidepressants

Noradrenergic uptake inhibitors – reboxetine Serotonin–noradrenaline uptake inhibitors – venlafaxine Serotonin antagonist/reuptake inhibitor – nefazodone Noradrenaline selective serotonin uptake inhibitors – mirtazapine

Newer antidepressant drugs

There have been several developments of antidepressants since the tricyclic era. Some drugs have briefly been mentioned above, which were nontricyclic, such as mianserin, maprotiline and viloxazine. However, the major development in the last few years has been of agents that selectively inhibit reuptake and either noradrenaline, or serotonin, or both. Table 20.3 shows a list of the newer drugs. Table 20.4 gives a receptor profile and the epileptogenic potential of these compounds. In brief, the selective serotonin reuptake inhibitors (SSRIs) are represented by citalopram, fluoxetine, fluvoxamine, sertraline and paroxetine. Of these, citalopram is the most selective on serotonergic reuptake, inhibiting serotonin reuptake 3000 times more than noradrenaline uptake, and 22000 times more than dopamine (Noble and Benfield, 1997). In general, the SSRIs are better tolerated and safer in overdose compared to tricyclic drugs. The latest generation of antidepressants has been developed to derive their therapeutic benefits from tailor-made action at specific monoamine receptors and reuptake sites, in theory providing better efficacy and better tolerability (Feighner, 1999). Reboxetine is a selective noradrenergic reuptake inhibitor (NARI) with low affinity for histaminergic, cholinergic, dopaminergic and alpha-1 adrenergic receptors. It appears to be equally effective as the tricyclics in treating depression, and there is a suggestion that it may be more effective than fluoxetine (Montgomery, 1997). Venlafaxine is a serotonin–noradrenergic reuptake inhibitor (SNRI), which is similar to the earlier generation of antidepressants, but it does not interact with histaminergic or cholinergic receptors, thus diminishing side effects due to those receptor systems. Several studies have indicated equi-potentiability or superior

303

The use of psychotropic drugs in seizure disorder

Table 20.4. Receptor profile and epileptogenic potential of antidepressants

Action on receptor Drug

H1

M1

NA

5HT1

5HT2

5HT3

Seizures/epilepsy

Imipramine

—

—

⫹

⫹

⫹

⫹

0.1–4% In overdose: 3.8–8%

Paroxetine

䊊

䊊

䊊

⫹

⫹

⫹

Prolonged seizures during ECT In overdose: No seizures in 15 patients with maximum dose 850 mg

Sertraline

䊊

䊊

䊊

⫹

⫹

⫹

Rare reports of seizures secondary to SIADH In overdose: No seizures in 40 patients up to 8000 mg

Fluoxetine

䊊

䊊

䊊

⫹

⫹

⫹

⬍1/1000a

Citalopram

䊊

䊊

䊊

⫹

⫹

⫹

No worsening of epilepsy in 16 patients In overdose: 100 mg – 1.9 g: 18% seizures; ⬎1.9 g: 49%

Reboxetine

䊊

䊊/⫺

⫹

䊊

䊊

䊊

0.13%a

Venlafaxine

䊊

䊊

⫹

⫹

⫹

⫹

0.18%a In overdose: Seizures in dosages over 1000 mg

Nefazodone

䊊

䊊

䊊

⫹

—

⫹

No seizures in premarketing trials, since then rare reports of convulsionsa

Mirtazapine

—

䊊

⫹

⫹

—

—

⬍0.1%a

Notes: 䊊, no/negligible effect; ⫹, stimulation; —, blockade; a information from premarketing trials and product monograph. Receptors: H1, histamine; M1, muscarinergic; NA, noradrenaline; 5HT1, 5HT2, 5HT3, serotonin.

effectiveness with this compound compared with tricyclics (Burnett and Dinan, 1994). Nefazodone is a noradrenaline–serotonin reuptake inhibitor whose most potent action is blockade of 5HT2 postsynaptic receptors, leading to a dual mechanism of action on the serotonin system. Noradrenaline reuptake inhibition is only minimal, and there is no interaction with histamine or cholinergic receptors. Mirtazapine (a noradrenaline-specific serotoninergic antidepressant or NASSA) has a selective action at alpha-2 adrenoreceptors, and only at some serotonin receptor subtypes. Its actions are to increase noradrenergic and serotoninergic transmission by blocking the alpha-2 autoreceptors. However, because it also blocks 5HT2

304

M.R. Trimble and A. Hensiek

and 5HT3 receptors, the increased serotonin turnover only stimulates the 5HT1 receptors. Thus it enhances noradrenergic and 5HT1A-mediated serotonergic neurotransmission. It is free of muscarinic, alpha-1 adrenergic and 5HT2- and 5HT3-related side effects, but its effect on histamine receptors can cause sedation and increased appetite. Several studies have shown equal or superior efficiency of this compound compared with other antidepressants (Bremner, 1995). New antidepressant drugs and seizures

All of the SSRIs have been associated with seizures in clinical practice, although there is some evidence which suggests that they may have less seizure potential than the earlier agents. Krijzer et al. (1984), used freely moving rats implanted with subcortical electrodes. Almost all of the antidepressants tested caused epileptogenic EEG changes; mianserin was the most potent. However, fluvoxamine caused only minimal effects. None of the newer agents have been tested in such models and clinical information is largely derived from the clinical trials and postmarketing surveys. Generally the figures for seizure incidence given for all of these new compounds are less than for the tricyclics, the lowest figures being recorded so far for mirtazapine and nefadazone. Citalopram has been used in depression in patients with epilepsy; no change of seizure frequency was noted in 16 patients in an open study (Specchio et al., 1999). In another study, Hovorka et al. (2000) gave citalopram to 43 patients with epilepsy and comorbid depression, and assessed them over an 8-week period. Sixty-five per cent were judged to be responders to the antidepressant effect. No change of seizure frequency was noted, and no de novo generalized tonic-clonic seizures were observed. The SSRI most used in patients with epilepsy is paroxetine. Blumer (1997; Chapter 8) has reported on the effective use of paroxetine in the management of patients with what he refers to as the interictal dysphoric disorder of epilepsy. In his studies, paroxetine is often given in combination with a tricyclic antidepressant. He reports this combination to be safe and efficacious in this population. Exacerbation of seizures was not reported, and dysphoric symptoms resolved in the majority of his cases. We have observed three patients with persistent epilepsy who have become seizure-free on paroxetine. In one, the effect was short-lived, but in the other two the effect has been sustained. This in spite of no changes to the anticonvulsant prescriptions. The other SSRI recently evaluated in epilepsy is sertraline (Kanner et al., 2000). They prospectively evaluated the effect of this antidepressant in 100 consecutive patients with epilepsy and depression (n⫽97) or obsessive–compulsive disorder (n⫽3). They noted an increase in seizures following start of therapy in 6% of

305

The use of psychotropic drugs in seizure disorder

patients, patients being assessed from 0.2 to 38 months. Interestingly, the mean dose of sertraline in these six was lower than in the other patients. They reported that depressive symptoms resolved in 54% of patients, but also described (Blumer, 1997) the pleomorphic clinical picture of these patients, and the symptom differences from typical major affective disorder. Pharmacokinetic interactions

It has been emphasized for a long time now that serum anticonvulsant-level monitoring can be of value in obtaining good seizure control, and checking compliance. The administration of additional drugs can cause metabolic interactions, leading to either a fall or a rise in the anticonvulsant serum levels. This may lead to a recrudescence or a worsening of seizure frequency, or precipitate anticonvulsant toxicity. There are occasional but nevertheless important reports of interactions between tricyclic drugs and both phenytoin and carbamazepine, leading to toxicity. The case of viloxazine has also been noted above. The new generation of antidepressant drugs differs considerably in their ability to induce liver enzymes of the P450 system. Most psychotropic drugs are metabolized by four isoenzymes (CYP1A2, CYP2C, CYP2D6 and CYP3A4) (Monaco and Cicolin, 1999). The anticonvulsants mainly affect CYP3A4, there thus being some potential for pharmacokinetic interactions. In general, drugs which induce liver enzymes may lower the levels of antidepressant drugs, and this may have therapeutic consequences. Information on the effect of SSRIs on plasma levels of anticonvulsants are limited, although there are reports of carbamazepine toxicity in patients given fluoxetine (Dursan et al., 1993). Keller et al. (1997) looked for interactions between fluoxetine and carbamazepine in patients with epilepsy, but did not find any change of plasma levels over an observation period of 20 days. Fluoxetine has been associated with case histories of increased phenytoin (Jalil, 1992) and sodium valproate levels (Cruz-Flores et al., 1995). Sertraline has less of an influence on concomitant anticonvulsant levels, probably because it has little or no effect on the cytochrome P450 3A4 system. However, possible interactions between sertraline and lamotrigine have been suggested (Kaufman and Gerner, 1998). Paroxetine also does not inhibit the CYP3A4 system, and may not therefore provoke any interactions. The limited clinical data on 20 patients with epilepsy given this drug did not reveal any significant alteration of anticonvulsant levels (unpublished data). With regard to the newer non-SSRI agents, no information is available. Conclusions

The data on the use of antidepressants in epilepsy suggest the following. Nearly all of the tricyclic drugs are proconvulsant, but there are clinical reports in which, at

306

M.R. Trimble and A. Hensiek

Table 20.5. Antipsychotic drugs

Phenothiazines Butyrophenones Atypical Others (sulpiride; tetrabenazine)

least with some of them, an anticonvulsant effect has been noted. The reason for this paradoxical effect is unclear. In psychiatric practice, in recent years, there has been a move away from the use of tricyclic drugs, mainly on account of their other side effects and risk of death with overdose; in patients with a reduced seizure threshold, not necessarily having epilepsy, they should be avoided. Other drugs that are proconvulsant include maprotiline and mianserin. Of the newer generation of drugs, the SSRIs appear to provoke less in the way of seizures than the tricyclic drugs. It is a possibility that the even newer, more selective drugs provoke less in the way of seizures than the SSRIs, but more data on these compounds are needed. Metabolic interactions occur with some of these compounds, which may lead to anticonvulsant toxicity. Any change in patient-reported symptoms, or a deterioration of the affective disorder, that may suggest a toxic effect, need to be watched out for. Whichever drug is used it is advisable to start, if clinical needs will allow, at smaller doses, and increase the dose relatively slowly in order to avoid precipitation of potential seizures. Antipsychotic drugs A classification of the antipsychotic drugs is given in Table 20.5; as with antidepressant drugs, in recent years there have been several newer agents introduced into clinical practice. These essentially, with some exceptions, fall into the class of atypical antipsychotics. Table 20.6 gives a review of the receptor-binding profiles of a number of these agents. The classical neuroleptic drugs, such as chlorpromazine and haloperidol, antagonize dopamine D2 receptors. Essentially their clinical efficiency has been shown to correlate with inhibitory activity at these receptor subtypes. However, these drugs block dopamine receptors in the striatum leading to catalepsy in animal models, and unwanted extrapyramidal side effects in clinical practice. The new generation of antipsychotic drugs essentially fall into two categories; those that are clozapine related, which includes olanzapine and quetiapine, and others such as risperidone.

307

The use of psychotropic drugs in seizure disorder

Table 20.6. Receptor binding profiles of antipsychotics

Affinity for receptor (Ki in nmol/l) Drug

D1

D2

D4

␣1

␣2

H1

5HT2a

5hHT2c

M

Haloperidol Clozapine Risperidone Olanzapine Quetiapine

25 85 75 31 455

1 126 3 112 160

5 9 7 27 NA

46 7 2 19 7

360 8 3 228 87

⬎1000 ⬎ 6 ⬎ 155 ⬎ 7 ⬎ 11

78 12 0.6 4 220

⬎1000 ⬎ 8 ⬎ 26 ⬎ 11 ⬎ 615

⬎ 570 ⬎ 1.0 ⬎1000 ⬎ 2.1 ⬎ 56

Notes: Receptors: D1, D2, D4, dopamine; ␣1, ␣2, adrenergic; M, muscarinergic; 5HT2a, 5HT2c, serotonine; H, histamine; NA, not available. Source: Information from premarketing trials and product monographs.

Although clozapine has been available for many years, it was initially removed from clinical practice (except in some selected countries) on account of its potential to produce agranulocytosis. However, it was reintroduced as a model of an atypical antipsychotic. The term relates to the low potential of these compounds to cause extrapyramidal problems, and they also have minimal effects on serum prolactin levels. The mechanism of atypicality seems to relate to different receptor profiles. In general, the atypical antipsychotics occupy lower levels of D2 receptors than the classical antipsychotics (20–60% as opposed to 80–90%) (Kapur et al., 1999). One reason for their profile may be due to the rapid displacement of these agents from receptors by endogenous dopamine, on account of their being more loosely bound. The newer antipsychotic agents also have a lower relative affinity for striatal D2 receptors as opposed to limbic D2 receptors (dorsal vs. ventral striatum). Further, of all the newer agents, clozapine is the one that seems not to bind to the core of the nucleus accumbens. Since their introduction, antipsychotic drugs have been shown to be proconvulsant. Early animal models, using the photosensitive baboon Papio papio, suggested that there may be differences between the phenothiazine-derived agents, such as chlorpromazine, and the butyrophenones, represented for example by haloperidol and pimozide. Pimozide in particular seemed to have less of an effect on the seizure threshold. In clinical practice it has recently been problematic to prescribe because of the need to carry out ECG investigations before prescription. It is one of a

308

M.R. Trimble and A. Hensiek

growing number of drugs associated with prolonging the Q-T interval, with the possibility of being associated with cardiac complications. In general, the use of intramuscular preparations, such as fluphenazine decanoate, was not associated with any change in the frequency of reporting of seizures in patients with epilepsy who also had psychosis. As with the newer antidepressants, there is much less information about the effect of the atypical neuroleptics on the seizure threshold, with the single exception of clozapine. The latter was known to be proconvulsant from early studies, the seizures being a dose-related effect. The incidence of seizures rises to about 5% at doses of 600 mg, although EEG changes may be recorded at lower doses. The seizures are often myoclonic, but can be generalized tonic-clonic, or partial, depending on the individual patient. It is perhaps no coincidence that the drug which appears to be the most effective antipsychotic, namely clozapine, is also associated with a high frequency of seizures. The relationship of convulsive seizures to the relief of psychopathology is an integral part of psychiatric therapy, through ECT. It is often forgotten that the latter was introduced for the treatment of dementia praecox, and has clinically and theoretically important antipsychotic effects. There are some patients with epilepsy who are nonresponsive to neuroleptic drugs, and need clozapine. In particular there is a group of patients whose seizure frequency decreases or who become seizure-free, whose psychosis deteriorates in this setting. For them clozapine may be the drug of choice.

CASE REPORT A 30-year-old patient, diagnosed as having leucine-sensitive hypoglycaemia at 10 months of age developed epilepsy at the age of four. This typically presented with clusters of several episodes daily, lasting 3 or 4 days, recurring at monthly intervals. During her seizure she would have a typical aura with a feeling of fear and butterflies in her stomach lasting about 30 seconds. This was followed by a scream, and a generalized seizure, which would last about a minute. Prior to these seizures she would often have a prodrome of 2–3 days with a build-up of verbal and physical aggression. She had been treated with various medications, but at the age of 26 she was changed to sodium valproate, and her seizure frequency improved, indeed she became seizure-free. She had gradually developed a psychotic illness, but this dramatically deteriorated with resolution of the seizures and she was admitted twice to psychiatric hospitals under the Mental Health Act, aged 26 and 27. An EEG had revealed left anterior temporal abnormalities, and an MRI showed prominent ventricles. She had been prescribed several antipsychotic drugs, including the atypical antipsychotics risperidone and olanzapine. None of these were of any help in resolving her psychosis.

309

The use of psychotropic drugs in seizure disorder She was therefore started on clozapine, in gradually increasing doses. Her EEG was initially monitored. On clozapine, her EEG revealed more frequent sharp waves over the left temporal region, and she began to develop auras again, although had no complex partial or generalized seizures. The auras were simple partial attacks, and were of little concern to her. A dramatic improvement in her psychosis was noted, such that she is once again living independently in the community, with stable mood, infrequent auditory hallucinations, and with more insight into her paranoia. She remains on sodium valproate and clozapine (400 mg a day).

Figures for the incidence of seizures with the other atypical antipsychotics vary from a reporting of 0.1% of seizures in double-blind clinical trials of rispiridone, to 0.2–0.9% for olanzapine, and 9 out of 1710 cases for quetiapine. These latter figures come from the reporting of seizures in clinical trials, and do not necessarily reflect a direct cause–effect relationship between prescription of the drug and the seizure event. Pharmacokinetic interactions

The interactions between antipsychotic drugs and antiepileptic drugs have been even less studied than the antidepressants. Some psychotropics, such as haloperidol, mainly metabolize using the P450 system, others such as chlorpromazine use different liver mechanisms. However, decreases in the levels of some neuroleptics can occur in patients prescribed anticonvulsant drugs, and several studies have been carried out in patients with schizophrenia who have received both carbamazepine and a neuroleptic. Haloperidol levels can drop by up to 50% following coadministration of the antiepileptic (Arana et al., 1986). Clozapine and olanzapine primarily use the CYP1A2 isoenzyme, which may lead to interactions with some of the tricyclic antidepressants, and carbamazepine. Conclusions

As with antidepressants, further work needs to be done in the important area of managing patients with neurological disease, particularly epilepsy, with antipsychotic agents. At present, particularly in epilepsy, the tendency is away from using the more traditional neuroleptics, to using the atypical neuroleptics, for several reasons. The main one relates to the potential danger of the long-term development of extrapyramidal motor disorders, which are much less likely to occur with the atypical neuroleptics. The latter are mainly well tolerated by patients with epilepsy, and seizures are not usually a problem clinically. As emphasized, clozapine can be used in patients with epilepsy, particularly if the psychosis is proving intractable to treatment. The drug is introduced slowly, and the EEG monitored. Patients are warned that their seizure frequency may rise; however, at doses below about 600 mg/day, clinical

310

M.R. Trimble and A. Hensiek

problems have not been encountered. One particular caution relates to the development of agranulocytosis, and patients (in the UK at least) need to be placed on a special register, and also have regular haematological assessments. Further, it is a contraindication to prescribe clozapine at the same time as carbamazepine. Other psychotropic agents Patients with epilepsy are prescribed a variety of other psychotropics, the main ones being benzodiazepines, either as hypnotics or anticonvulsants, and lithium, a mood stabilizer. Benzodiazepines should be used in caution in patients with epilepsy, the main problem being the potential for a paradoxical increase in seizures, or withdrawal seizures on stopping the prescription. Further, some of these drugs have a potential for the development of dependency. There appear to be differences between the 1,5 and 1,4 benzodiazepines, the former being represented by clobazam. This drug was introduced initially as an anxiolytic, but was shown to have effective and sustained anticonvulsant properties. It is recommended as an adjunct treatment for the management of patients with intractible epilepsy, and may be particularly of value in patients with epilepsy with a high level of anxiety, who may also present with panic attacks. It is less cerebrotoxic than the 1,4 equivalents such as clonazepam, and is recognized to have inherent psychotropic properties. Clobazam is of particular value in patients with intermittent clusters of seizures (such as catamenial episodes), and for the supression of clusters of seizures. The latter are associated in some patients with postictal psychosis, and prevention of the cluster may well abort a potential psychosis. Ten milligrams given 4–6 hourly for 24–48 hours may be all that is required. Clobazam can also be given after the cluster, if any psychiatric symptoms seem to be developing, using a similar schedule. Lithium, which is also proconvulsant, can be used as a mood stabilizer in patients who have recurrent cyclical mood disorders, or recurrent outbursts of affective aggressive behaviours. Caution should be exercised when combining lithium with carbamazepine, as patients occasionally develop a cerebrotoxic syndrome. Monitoring of serum levels of lithium is mandatory, as is observing patients over time for the development of secondary complications of lithium therapy such as hypothyroidism, or diabetes insipidus. Conclusions Psychotic drugs are used with considerable frequency in patients with epilepsy, and used appropriately and cautiously they add considerably to management. However,

311

The use of psychotropic drugs in seizure disorder

like all CNS drugs, they have a variety of side effects, and the exacerbation or precipitation of seizures is important in this patient group. It is particularly relevant in patients who may have been seizure-free for a period of time, and who then go on to develop psychiatric disorders. Recent years have seen an expansion in the number of psychotropic drugs available, particularly with regard to the antidepressants and the antipsychotics. The newer developed agents generally seem to have a more favourable profile than older agents for use in patients with epilepsy.

R E F E R E N C ES Arana, G.W., Goff, D.C., Freidman, H. et al. (1986). Does carbamazepine-induced reduction in haloperidol plasma levels worsen psychotic symptoms? Am J Psychiatry, 143, 658–9. Blumer, D. (1997). Antidepressant and double antidepressant treatment for the affective disorder of epilepsy. J Clin Psychiatry, 58, 3–11. Bremner, J.D. (1995). A double blind comparison of ORG 3770, amitriptyline and placebo in major depression. J Clin Psychiatry, 56, 519–26. Burley, D.M. (1977). A brief note on the problem of epilepsy and antidepressant treatment. In Depression: The Biochemical and Physiological Role of Ludiomil, ed. A. Jewkes, pp. 202–3. Horsham: Ceiba. Burnett, F.E. and Dinan, T.G. (1994). The clinical effectiveness of venlafaxine in the treatment of depression. Rev Contemp Pharmacother, 9, 303–20. Cruz-Flores, S., Ghazala, R., Hyat, R. and Mirza, W. (1995). Valproaic toxicity with fluoxetine therapy. Missouri Med, 92, 296–7. Dursan, S.M., Natthew, V.W. and Reveley, M.A. (1993). Toxic serotonin syndrome after fluoxetine plus carbamazepine. Lancet, 342, 442–3. Edwards, J.D. and Glen-Bott, M. (1984). Does viloxazine have epileptogenic properties. J Neurol Neurosurg Psychiatry, 47, 960–4. Edwards, J.G. (1985). Antidepressants and seizures: epidemiological and clinical aspects. In The Psychopharmacology of Epilepsy, ed. M.R. Trimble, pp. 119–139. Chichester: John Wiley & Sons. Feighner, G.P. (1999). Mechanism of action of antidepressant medication. J Clin Psychiatry, 60, 4–11. Garvey, M.J. and Tollefson, G.D. (1987). Occurrence of myoclonus in patients treated with cyclic antidepressants. Arch Gen Psychiatry, 44, 269–72. Hovorka, J., Herman, E. and Nemcova, I. (2000). Treatment of interictal depression with citalopram in patients with epilepsy. Epilepsy Behav, 6, 444–8. Jalil, P. (1992). Toxic reaction following the combined administration of fluoxetine and phenytoin. J Neurol Neurosurg Psychiatry, 55, 412–13. Jick, H., Dinan, B.J., Hunter, J.R. et al. (1983). Tricyclic antidepressants and convulsions. J Clin Psychopharmacol, 3, 182–5.

312

M.R. Trimble and A. Hensiek Kanner, A.M., Kozak, A.M. and Frey, M. (2000). The use of sertraline in patients with epilepsy, is it safe? Epilepsy Behav, 1, 100–5. Kapur, S., Zipursky, R.B. and Remmington, G. (1999). Clinical and theoretical implications of 5HT2 and D2 receptor occupancy of clozapine, risperidone and olanzapine. Am J Psychiatry, 156, 286–93. Kaufman, K.R. and Gerner, R. (1998). Lamotrigine toxicity secondary to sertraline. Seizure, 7, 163–5. Keller, R.. Tortar Prolo, P., Ravizza, L. and Monaco, F. (1997). Interazioni farmacocinetiche tra fluoxetina e farmaci anti epilettica. Bolletino Lega Italiana Contro L’epilepsia, 99, 183–6. Krijzer, F., Snelder, M. and Bradford, D. (1984). Comparison of the proconvulsant properties of fluvoxamine and clovoxamine with other antidepressants in an animal model. Neuropsychobiology, 12, 249–54. Luchins, D.J., Oliver, A.P. and Wyatt, R.J. (1984). Seizures with antidepressant: an in vitro technique to assess relative risk. Epilepsia, 25, 25–32. Meldrum, B.S., Anlezark, G., Adam, H.K. and Greenwood, D.T. (1982). Anticonvulsant and proconvulsant properties and viloxazine hydrochloride. Psychopharmacology, 76, 212–17. Monaco, F. and Cicolin, A. (1999). Interactions between anticonvulsants and psychoactive drugs. Epilepsia, 40 (Suppl.), S71–6. Montgomery, S.A. (1997). Riboxetine: additional benefits to depressed patients. J Psychopharmacol, 11 (Suppl.), S9–15. Noble, S. and Benfield, P. (1997). Citalopram: a review of its pharmacology, clinical efficiency and tolerability in the treatment of depression. CNS Drugs, 8, 410–31. Ojemann, L.M., Friel, P.N., Trejow, J. and Dudley, D.L. (1983). Effect of doxepin on seizure frequency and depressed epileptic patients. Neurology, 33, 646–8. Pisani, F., Narvone, M.C., Fazio, A. et al. (1984). Increased serum carbamazepine levels by viloxazine in epileptic patients. Epilepsia, 25, 482–5. Specchio, L.M., La Neve, A., Spinelli, A. et al. (1999). Il trattamento antidepressevo con citalopram in pazienti con epilepssia. Bollettino Lega Italiana Crontro L’epilepsia, 99, 187–8. Trimble, M.R. (1980). New antidepressant drugs and the seizure threshold. Neuropharmacology, 19, 1227–8. Trimble, M.R. (1987). Antidepressant drugs seizures and epilepsy. In Quo Vadis. Montpelier: Sanofi Group. Trimble, M.R., Anlezark, G. and Meldrum, B. (1977). Seizure activity in photosensitive baboons following antidepressant drugs, and the role of serotoninergic mechanisms. Psychopharmacology, 51, 159–64.

21

The role of psychotherapy in the treatment of epilepsies Martin Schöndienst Epilepsie-Zentrum Bethel, Bielefeld, Germany

Introduction The use of psychotherapy in epilepsy is a complicated topic, because it always has to take into account the neurological dimension of the underlying epileptic disorder as well as the specific psychiatric vulnerability epilepsy engenders. As a result, the following cannot provide a simple set of treatment rules. Instead, it presents: 1. Introductory comments on the complex relation between (neurological) epileptological and psychotherapeutic approaches. 2. This is followed by four case reports to give some idea of how the general rules of diagnosis have to be applied in an individual way to produce a unique treatment in every single case. 3. Then, some ideas are presented on indications and contraindications. 4. Finally, findings from a posttreatment comparison of conventional inpatient treatment of epilepsy vs. treatment supplemented by psychotherapy are reported to help decide whether psychotherapy is in any way effective in treating epilepsy or is only well-intended. Despite the above-mentioned complexity, it is easy to gain an overview on psychotherapy in the treatment of epilepsy because there seems to be practically no recent literature on the topic. The indices of major modern textbooks on epileptology refer to psychotherapies exclusively in connection with pseudoseizures (Engel and Pedley, 1997) or give, at best, the terse reference that ‘this modality is largely neglected in the literature on epilepsy for several reasons’ (Stagno, 1993, p. 1154). Except for pseudoseizures, psychotherapeutic procedures have not been addressed in any relevant epileptological journals during the last 10 years. It is quite remarkable, and almost worth studying in its own right, that psychotherapy in the treatment of epilepsy has hardly ever been a subject of scientific research, even though there is no denying the frequency of psychological disturbances in such patients. In line with the situation in the literature, medical care reveals a disproportionately high number of secondary psychological disturbances 313

314

M. Schöndienst

in epilepsy patients but a disproportionately low level of psychotherapeutic assistance. This has less to do with the psychotherapists themselves, despite the frequent truism that psychotherapy could aggravate an epilepsy and is, therefore, contraindicated. It is far more the case that epileptology itself has remained conspicuously silent as far as demands for the development of psychotherapeutic concepts for its clients are concerned. To some extent, this may be due to the way in which epileptologists often view psychotherapists: namely, as little more than those who seem to have all the time at their disposal that one would wish for one’s own work. They also make comparatively little use of the available psychotherapeutic treatments compared to what are considered to be potent drug options. Finally, clinicians often view psychotherapists as persons who always only confirm what they know and anticipate already: particularly overprotective mothers, destructive fathers, strangulated affect in the patients or some kind of spectacular narratives from a patient’s prior life that only get in the way of the abstraction necessary in clinical work. Finally, there are a few major differences in the ways of thinking themselves that create major handicaps to the integration of epileptological and psychotherapeutic perspectives: whereas the epileptologist strives continuously to test whether external treatments such as new drugs or, if necessary, surgery should be applied, the psychotherapist focuses specifically on the action potentials of the patients themselves, be they in developing an individual technique for stopping seizures, improving insight or correcting problematic patterns of interpersonal interactions. Instead of relying on the positive, visible findings of neuro-imaging techniques, neurophysiology and neuropsychology, psychotherapy focuses on the gaps and contradictions that first make it possible to grasp defensive mechanisms such as projections, denials and isolations with all their intra- and interindividual consequences. In contrast to the usual and necessary practice in epileptology of advancing knowledge by tracing back the disorder increasingly more precisely to a circumscribed cause that is localized as accurately as possible, psychotherapy is concerned with how symptoms are embedded in a person’s individual life context. Hence, a psychotherapeutic diagnosis is based on the idea that by creating a suitable setting, cautious interventions and on-line observations of the interplay in the counselling session, the decisive problem areas for the patients will be revealed in interaction and also become accessible to modification. Furthermore, there is also a justifiable apprehension that focusing on, for example, defensive processes or (counter) transference might lead to the neglect of physical aspects that are just as essential for treatment. Each recognized dimension added to the relevant data field, in this case, the usual discourse on epileptology (i.e. case history, EEG and, not least, neuro-imaging techniques) multiplies the number

315

Psychotherapy in the treatment of epilepsies

of interactions that need to be taken into account, and produces not only a factual increase in knowledge but also an equally factual increase in potentially confusing complexity. A reluctance to tackle such a lack of transparency and a preference to narrow the field to neurological treatments has a certain cognitive rationality and is indubitably better than forcing oneself to take more demanding perspectives, particularly when they are not encouraged by the hospital or surgery framework. Nonetheless, epileptologically difficult treatment situations often emerge in which it is popular to talk about a ‘noncompliance’ or ‘pharmacoresistance’ that apparently cannot be explained further. These are situations in which the integration of a psychotherapeutic approach can bring about a decisive change in the course of treatment. However, it is harder to recognize such a need for psychotherapy in patients with epilepsy than in those without such an organic disease: many treatment problems are attributed too hastily to the epilepsy or to the seizures as such, to the medication or also to accompanying neuropsychological deficits rather than being conceived as problems accessible to psychotherapy. This neglects the problems due to either more or less unconscious conflicts (e.g. of identity, selfesteem or dependency) or so-called ego-structural deficits (e.g. a highly reduced perception of self and/or other; inadequate self-control; unstable affect; or immature defence mechanisms such as, in particular, denial or dissociation, neurotic or even psychotic projection and unstable attachment behaviour). In the following, four short case reports will be used to illustrate how epileptological, psychiatric, and, in the stricter sense, psychotherapeutic dimensions interrelate. F O U R C A S E R E P O R TS The first patient has an epilepsy with focal and generalized seizures plus perimenstrually peaking diffuse tonic-clonic seizures with onset at 15 years. Although her left-cephalic aura indicated a right-temporal focus, photosensitivity and corresponding spike-waves as well as a hereditary factor pointed to a generalized disorder. The patient was resistant to phenytoin and carbamazepine and was referred to us with a relatively high phenobarbital level. The admission interview was characterized by the recurring and piercingly expressed theme that her last physician had said he was referring her to us after telling her that he had ‘nothing left up his sleeve’. ‘Yes, and then, for a while, I had a doctor who was on television, and everybody gave him a lot of praise, and he almost cost me my life and treated me with valproic acid until I was in a coma.’ Another recurring phrase was, ‘I can’t fall into the open arms of a doctor.’ Hence, these and similar communications linked together major therapeutic, erotic and destructive ideas of reference in an indiscriminable, confusing and entangled manner.

316

M. Schöndienst We admitted her for inpatient treatment with the diagnosis of an agitated depression plus occasionally severe suicidal intent. Initially – after the experiences reported above – she must almost have thought we wanted to murder her when we proposed changing her medication to the combination of valproic acid and lamotrigine that is particularly promising for such mixed epilepsies. Against the background of such breakdown fantasies, the planned changeover was even more difficult because it also included an inpatient phenobarbital detoxification with all the risks of seizures in withdrawal. None of those involved were spared in any way. The patient was good at eliciting a number of strong reactions whose impact was bound to remain destructive as long as it was not understood as an actualization and externalization of her self- and self-esteem conflict and associated fears of doom and destruction. Although the wish for a less exhausting patient is understandable in such phases of treatment, it would only make the therapeutic relationship superficial or lead to a cessation of treatment. A few themes recurred during treatment: 1. Poisoning by the prior therapist along with the peevish reproach that, nowadays, it seemed that such poisonings were simply taken for granted. 2. The father’s working in the garden. 3. Buying expensive shoes. 4. Her mother’s own occasional seizure 20 years before. Although this list may seem absurd, its contents provided opportunities for symbolic understanding. In her accounts of weekend visits to her parents that always featured her father working in the garden, the patient found a way to leave behind all her bitterness and was at times so warm-heartedly humorous as to not only disclose the both decisive and psychosexually fixated relationship with her father but also enable her to become aware of this through a reflective self-distancing. The patient’s rather strange deliberations over whether to buy a wonderful pair of very expensive winter boots not only reflected her self-conflict ranging from her emerging identity plans and oppressive material restrictions, but it was also capable of being named as such. This was joined by dreams that made it possible for her to see the possibility of taking an intermediate position somewhere between delusions of grandeur and depression. At the same time, the mother started to recollect her single occasional seizure 30 years before that, as now threatening the daughter, had led her to give up a career as a school teacher while also suggesting a biological flaw in the family on which the daughter had fixated in anxious anticipation. One may consider that all this has little to do with the epilepsy but perhaps with a completely independent narcissistic neurosis. However, it is precisely the specific constellation of connections in each single case between (a) directly seizure-related breakdown experiences, (b) the narcissistic crises associated with every seizure for many patients, (c) the cumulative growth of resignation, (d) the fragility of self-esteem and (e) the ambivalence towards medicines that, like a disappointing object, are perceived simultaneously as an indispensable protection but also repeatedly as a failure and disappointment that makes it necessary for the

317

Psychotherapy in the treatment of epilepsies therapist to consider not only the epilepsy and its best possible pharmacotherapy, but also the ego-structural sequelae of both the seizures themselves and the medicines. Naturally, it is impossible to reach a final decision on whether the long-term reduction in anxiety and calming of this patient is due to her now being free from seizures for a complete year, to the withdrawal from phenobarbital, to the combined treatment with valproic acid and lamotrigine or to the processing of the above-mentioned (in this case, self- and oedipal) conflicts with the resulting increase in internal latitude. Whatever the case, the psychotherapeutic setting was certainly necessary to generate a tolerance for the difficult changeover without which it would have been impossible to contain a patient in such a precarius condition. After freedom from seizures had been achieved, I found it more than ironic when the patient told me how her mother had asked her impatiently during her menstruation whether she had had another seizure. After the second seizure-free menstruation, the mother surprised the patient one morning by telling her that she had had her first seizure for 30 years the night before, or at least she had woken up after biting her tongue. The next patient, a 28-year-old male with a right-hemisphere epilepsy and cerebral hemihypotrophy and somatosensory auras, frontal hypermotor and generalized tonic-clonic seizures had not, at the time of admission, left his parent’s home for more than 10 years because of his fear of having a seizure on the street. The drug changeover was accompanied by two half-hour psychotherapy sessions per week. A number of these sessions were characterized by the patient’s complaints over the way seizures prevented him from getting anywhere in life along with my personal tiredness in reaction to this that was almost impossible to control. After I had turned up late for several of our sessions, I realized how far the patient seemed to accept my tardiness with complete indifference. I mentioned this to him, and this transformed my role from a sacrosanct physician into that of an assailable other, leading to an incredible change: unexpectedly, I became the focus of very excessive demands. He simply thought that I or the hospital should help him to find not only a flat and a job but also, when possible, a mate. Hence, a regression to the level of grandiose infantile wishes had occurred. Several sessions pursued a kind of reality test of the patient’s wishes that led inevitably to disappointed anger and the necessary emotional counter-control of the therapist known in the literature as containing. After a hefty but brief depressive reaction, the patient found his way out of his regressive arrest and began to structure his future in small practical steps rather than getting caught up in grandiose desires. He worked out his own desensitization programme to overcome his fear of the streets that was soon a success. In addition, he managed to learn the necessary DC potential shifts in a biofeedback treatment so that although auras continue to occur daily, these have not turned into major seizures for more than 18 months. The next 30-year-old patient with a focal epilepsy and right-temporal lobe hypotrophy was admitted with such prior diagnoses as an ‘abnormal personality development with depressions and anxiety states’ and even a tentative diagnosis of ‘onset of psychosis with auditory hallucinations’. A diagnostic phase within a preoperative institute had taken a relatively turbulent course in interactive terms.

318

M. Schöndienst Alongside improving antiepileptic drug therapy, the 3.5-month treatment consisted initially in a very cautious approach to her strange and prolonged disturbances to perception and experience. For a long time, these led to a threatening atmosphere that was hard to understand. In the past, her so-called auditory (pseudo) hallucinations had led to the tentative diagnosis of a psychosis, and these were currently also the reason for major disintegration anxieties expressed in, for example, the sensation ‘then, I feel like a dictionary in which more and more pages are simply blank’ or ‘everything I want to think gets sucked out of my brain as if it was a dry sponge’. During psychotherapy, it took some time to verbalize the various episodic sensations to which she was exposed because of the difficulty of putting the quality of these experiences into words. It was also necessary to distinguish very slowly and carefully her auras from other episodes that had the character of anticipatory anxiety or mental disintegration anxiety. This also made it possible to calm down the patient’s own apprehension that she was facing a creeping psychotic disintegration, an apprehension that she would suddenly start to project on her communication partner with corresponding surges in anger when treatment first commenced. Hence, in this case, it was the verbalization of what was experienced in the auras that helped to make the repeated sudden affects and intensively conflictual relationship patterns accessible to a more relaxed self-observation. The patient developed more precise differentiations of affect, and the initially threatening quality of the treatment atmosphere gradually disappeared. The final 19-year-old patient was referred to us after spending 11 months in a psychiatric hospital for adolescents. For several years, she had been treated by an epileptic outpatient department for the prior diagnosis of a focal frontal epilepsy. Our own diagnosis, in contrast, was very clearly an idiopathic generalized epilepsy with myoclonic and generalized tonicclonic seizures. Correcting the diagnosis was important, because the administration of lamotrigine and valproic acid led immediately to a remarkable improvement in her seizures. Only isolated relapses occurred either when the patient was woken abruptly or when medication had not been taken. Nonetheless, such irregularities were very frequent because of her very irregular bedtimes as well as a failure to take medication. Encouraged by the impressive initial effect of changing her drugs, we thought that the disruptive gaps in compliance could be closed through simple advice. However, contact with the patient then became reduced: she appeared to draw back from her previous attentiveness into a dour, seemingly almost unreachable silence. After an exhausting processing that confronted major familial scotomization tendencies in her life history, it emerged that at the age of 10–14 years, that is, at the onset of her epilepsy, her mother had blocked any adequate pharmacotherapy due to her own fears of poisoning engendered by her personal psychotic development. Her father, in contrast, had advised her emphatically to take her medicine, but had been unable to assert himself. The reconstruction of this history in conjunction with the impression that the patient ‘drifted off’ precisely when attempting to discuss the background of her clearly self-injurious noncompliance revealed the intrapsychological conflict behind the biographical drama: by

319

Psychotherapy in the treatment of epilepsies switching apparently meaninglessly between adequate and inadequate health behaviour, the patient subconsciously maintained the inner relationship to both her father and her mother. Compliance would have meant the loss of the internal mother. It is precisely when such a meaning of noncompliance becomes recognized that compliance becomes conceivable and, indeed, unproblematic.

Discussion I would like to present some remarks on the setting for a psychotherapeutic approach to epilepsies. My first point addresses the apparently marginal element of a fixed timeframe, that is, the defined amount of time specified at the beginning of treatment on the basis of the initial psychostructural situation, for example, four times 20 minutes per week or two times 25 minutes per week. This changes time from something that the physician uses to attain certain goals into something that is primarily at the patient’s disposal, to be used or abused as he or she sees fit. The second point is embedded in the first, and concerns the patient’s right and need to start the discussion without being belittled by any kind of guiding questions on the choice of topic. The therapist is then no longer the knowledgeable one who already knows what the patient should think about, what should be recommended and what should be discouraged. The therapist becomes a companion in the patient’s process of self-discovery. A further point is the need to pay attention to methods for recognizing egostructural problems behind the more obvious surface symptoms. Hence, behind the first patient’s irritable depression lies the instability of her selfesteem regulation that she externalized in fantasies of being poisoned with medicines. In the second patient, the agoraphobic avoidance covers the tendency to regress to the level of passive infantile need. In the third patient, schizoid rejection of contacts hides her fear of psychotic disintegration as a consequence of her auras that she had previously been unable to identify. In the fourth patient, depressive withdrawal resulted from the unconscious conflict arising from her identification with both very hostile parents. Interventions have to be selected to fit the specific disorder, and they require a willingness in the therapist to participate emotionally in what is perceived. In the first patient, for example, this meant involvement in the anxiety associated with her seizures. In the second patient, the therapist had to counter his regression in a demanding way and risk his anger. In the third patient, it was necessary to recognize the fearful atmosphere of vague threat. In the fourth patient, this meant coming to terms with the seemingly (at first glance) inappropriate switches of affect. When collating the numerous case reports in which supplementing epilepsy treatment with psychotherapy has proved to be of lasting benefit, we have found

320

M. Schöndienst

again and again that the decisive element in these treatments is to reconstruct the remarkably blurred self–object borders in these patients. One could say that it is important for the therapist to take over various functions of a patient’s self (in the sense of taking on a helping ego function) and reconstruct them successively. This requires a correction of defence mechanisms, particularly of projection and regression, that is never without some emotional strain. The therapist also has to cope repeatedly with a certain form of balance. He or she has to know the specific strains on self-regulation from having epilepsy that develop not least from the loss of the availability of the self that is particular to the disease and through which seizures can also take on the character of narcissistic crises. In the first patient, this meant commencing practicable confrontations and demarcations instead of being trapped by both negative and positive idealizations of her images of her parents. In the second patient, this meant letting the archaic grandiose self be replaced by a more mature, disappointed self that was capable of formulating and implementing phase-appropriate limited goals. In the third patient, the concern was to overcome threatening projections and calm the prepsychotic disintegration anxiety and develop the self-reflective ability to discriminate between auras and depersonalizations. Questions of indication and contraindication Alongside obvious indications such as a marked anxiety, obsessive–compulsive or borderline disorder, we consider that the indication for a supplementary psychotherapy should always be examined in epilepsy patients when (a) an epilepsy treatment is difficult in itself, and particularly when (b) ‘noncompliance’ or (c) ‘pharmacoresistance’ are present. One particular impediment to recognizing the indication for psychotherapy is found in those underdiagnosed cases in which an epilepsy is accompanied by a posttraumatic stress disorder (PTSD) (Rosenberg et al., 2000). This does not just mean cases involving exceptionally violent biographical experiences. Recurrent severe seizures themselves may represent a directly traumatic experience in a stricter sense, particularly when – as in many frontal epilepsies – the patient retains a considerable degree of consciousness. Several of the characteristics of PTSD mentioned in DSM-IV (American Psychiatric Association, 1994) such as the ‘numbing of general responsivity, constriction of affect, re-experiencing the traumatic event or an exaggerated startle response’ can be considered characteristic of many epilepsies as well. This makes it all the more surprising that so little research has focused on the potentially traumatizing character of seizures. One reason for this may be the well-known fact in psychotraumatology that the more destructive stressful events have been, the more difficult they are to express in words. Contraindications

321

Psychotherapy in the treatment of epilepsies

Table 21.1. Follow-up of inpatient epilepsy treatments

Completely seizure-free

Without psychotherapeutic add-on

With psychotherapeutic add-on

⬎6 months after discharge n⫽43 5

⬎12 months after discharge n⫽34 11

Note: Exact Fisher test: P⫽0.025

are given particularly when professional and institutional preconditions are inadequate. Commencing treatment without the right setting (founded psychotherapy training, defined session times, discussions regarding the duration, form, and goals of treatment, supervision) borders on malpractice. Likewise, one should not commence psychotherapy with an epilepsy patient without being aware of the multitude of psychopathological phenomena that may occur in epilepsies. Finally, supplementary psychotherapy cannot be recommended in an inpatient treatment framework when this is not supported by the institution, and sound framing conditions are not provided. Effective or only well-intended?

In my opinion, the above considerations justify conceiving psychotherapy in an epilepsy centre within an interdisciplinary framework and practising it as a supplement to conventional treatment – hence, fashionably speaking, in the sense of an ‘add-on’. However, to examine the methodologically difficult question as to whether its effectiveness can be confirmed, we (the head of the nursing service, H. Welteke, and the author) carried out a follow-up study. Seizure-related outcomes were surveyed in 34 patients at least 12 months after discharge from inpatient treatment. Findings were classified with the adapted version of the so-called Engel Scale (Engel, 1987) for drug treatment histories (categories are: Class I, completely free of disabling seizures; Class II, rare disabling seizures (‘almost seizure-free’); Class III, quantifiable worthwhile improvement; and Class IV, no measurable improvement in life quality). Results showed that 11 (32.4%) former patients were completely free of seizures, a further 4 (11.8%) were almost seizure-free, 6 (17.6%) had experienced a worthwhile improvement, whereas the remaining 13 (38.2%) had to be classified to Engel’s Class IV. It is interesting to compare our findings with a parallel 6-month follow-up in the rest of the Bethel Epilepsy Centre (Pfäfflin and May, in press). This study reports a complete freedom from seizures in only 12.4% of patients. Table 21.1 compares the data.

322

M. Schöndienst

This suggests that the number of optimal treatment outcomes was significantly higher in the group with supplementary psychotherapy at an alpha level of 0.05. It can be concluded that there now is less need to justify provision of a psychotherapeutic add-on for epilepsy patients, but, instead, more need to justify why the potentials of such treatments often remain unexploited.

R E F E R E N C ES American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (fourth revision) (DSM–IV). Washington, DC: APA. Engel, J. Jr (1987). Outcome with respect to epileptic seizures. In Surgical Treatment of the Epilepsies, ed. J. Engel Jr. New York: Raven Press. Engel, J. and Pedley, T. (1997). Epilepsy –A Comprehensive Textbook, Vol. I–III. Philadelphia, New York: Lippincott-Raven. Pfäfflin, M. and May, T. (in press). Comprehensive care in an epilepsy clinic. In Comprehensive Care for People with Epilepsy, ed. R.T. Fraser, M. Pfäfflin, R. Thorbecke, U. Specht and P. Wolf. London: John Libbey. Rosenberg, H.J., Rosenberg, S., Williamson, P. and Wolford, G. (2000). A comparative study of trauma and posttraumatic stress disorder prevalence in epilepsy patients and psychogenic nonepileptic seizure patients. Epilepsia, 41, 447–52. Stagno, S.J. (1993). Psychiatric aspects of epilepsy. In The Treatment of Epilepsy: Principles and Practice, ed. E. Wyllie. Philadelphia, New York: Lea and Febiger.

22

Choosing measures to assess quality of life (QOL) in epilepsy Caroline E. Selai, Katja Elstner and M.R. Trimble Institute of Neurology, London, UK

Why assess quality of life? Whilst quality of life (QOL) measures have been developed for a number of reasons (Fitzpatrick et al., 1992), two basic aspects of health care underlie most of the questions that QOL appraisals set out to answer: outcome of treatment and cost. With increasingly sophisticated life-saving and life-prolonging medical interventions, and a range of options between alternative treatments, quality of life has emerged as an important outcome. Also, it is argued that no country in the world can afford to do all that it is technically possible to do to improve the health of its citizens and so the need has arisen for some system of setting priorities. Quality of life and other outcome data are informing health economic decisions and debate about the allocation of scarce resources. The field of QOL research is thriving, and much progress has been made in the last 10 years. Types of QOL measure There is no ‘gold standard’ for measuring QOL and the range of instruments available, or still undergoing development, is remarkable in terms of both quantity and heterogeneity. The range of categories of QOL/health status measures has been comprehensively reviewed elsewhere (Brooks, 1995). In brief, generic instruments cover a broad range of QOL domains in a single instrument. Their chief advantage is in facilitating comparisons among different disease groups. Disease-specific instruments reduce patient burden by including only relevant items for a particular illness but their main disadvantage is the lack of comparability of results with those from other disease groups. Health profiles provide separate scores for each of the dimensions of QOL, whereas a health index, a type of generic instrument, gives a single summary score, usually from 0 (death) to 1 (perfect health). 323

324

C.E. Selai, K. Elstner and M.R. Trimble

Another type of measure, developed within the economic tradition, is the utility measure. Whilst some decisions are taken for individual patients, others, such as those made by health policy makers, concern groups of patients. Here the focus is on society as a whole and the societal allocation of scarce resources. For this purpose, preference-weighted measures are required. Finally, some researchers have questioned the appropriateness of the fixed questionnaire to assess QOL. They argue that, since quality of life is a uniquely personal perception, most standardized measurements of QOL in the medical literature seem to aim at the wrong target (Gill, 1995). Their argument, that any technique to assess QOL should be tailored to the individual respondent, is discussed further below. Individualized, patient-driven QOL assessment techniques There is a fundamental tension in the measurement of QOL. Since what is deemed important for QOL is acknowledged to be subjective and idiosyncratic, differences being influenced by a variety of personal and cultural factors, an appraisal of QOL should strive to capture the individual’s subjectively appraised phenomenological experience. On the other hand, the hallmark of scientific measurement is reliable, ‘objective’, empirical data collection. Researchers have devised QOL measurement techniques at various stages of this ‘subjective/objective’ continuum and there now exist over 1000 instruments that have been developed taking a variety of approaches to measurement (Hedrick et al., 1996). The two poles of the qualitative–quantitative continuum have different strengths. It has been suggested that qualitative methods are more valid whilst quantitative methods are more reliable (Mays and Pope, 1996). The QOL literature advocates a robust and rigorous programme of instrument development and testing, and most QOL measures are developed within the psychometric tradition (Juniper et al., 1996; McDowell and Newell, 1987). However, since quality of life is uniquely personal, most standardized measurements of QOL are inappropriate (Gill, 1995). It is argued that quality of life can be suitably measured only by determining the opinions of patients and by supplementing (or replacing) the instruments developed by ‘experts’ (Gill and Feinstein, 1994). Scales developed within the psychometric tradition often omit items important to the beliefs and values of individual patients (Gill, 1995; Hunt, 1999) and the psychometric aim of internal reliability is in conflict with the goals of achieving comprehensiveness and content validity (Brazier and Deverill, 1999). In response to this, a number of ‘individualized’, patient-driven techniques have been developed whereby the patient can nominate items of importance to him/herself (Fraser et al., 1993; Geddes et al., 1990; Guyatt et al., 1987a, b; O’Boyle et al., 1993; Ruta et

325

Measures to assess quality of life in epilepsy

al., 1994; Tugwell et al., 1990). The literature on ‘individual’ QOL assessment techniques has recently been reviewed (Joyce et al., 1999). The choice of measure will depend upon the goal of the study; a common recommendation is to include both disease-specific and generic measures in an investigation. QOL measures for use in epilepsy

The history, development and current status of measures to assess health-related quality of life (HRQOL) in epilepsy have been comprehensively reviewed (Cramer, 1996; Hermann, 1995; Trimble and Dodson, 1994). It is well documented that the occurrence of seizures and stigma of epilepsy lead to a variety of physical, psychological and social problems which have an impact on many areas such as selfesteem, work, career prospects, family, leisure, (in)ability to drive and relationships (Cramer, 1996). Therapeutic outcome: change in seizure frequency Whilst approximately 70% of patients are well controlled on monotherapy, with standard antiepileptic drugs (AEDs), for the remaining 30% of patients polytherapy is considered. Surgery (mainly temporal lobectomy) is another option for some patients with intractable epilepsy. Drug trials

In clinical trials, a change in seizure frequency has traditionally been the main measure of efficacy. Those patients who experience a ⬎50% reduction in seizure frequency are described as ‘responders’ whilst all other participants are ‘nonresponders’ (Smith et al., 1995). QOL has rarely been an outcome measure in clinical trials of AEDs. Surgery

More stringent criteria have been adopted for the seizure-defined outcome after epilepsy surgery. In addition to seizure freedom, some researchers have chosen a 75% (Bladin, 1992; Hermann et al., 1992; Malgrem et al., 1997) and some a 90% reduction in seizures (McLachlan et al., 1997; Rose et al., 1996). The two studies reported here were designed to look at the relationship between changes in seizure frequency (clinically defined end-point) and changes in QOL. The data were collected from patients attending clinics at the National Hospital for Neurology, Queen Square. The main thrust of this chapter is to compare the sensitivity of the Quality of Life Assessment Schedule (QOLAS), an individualized,

326

C.E. Selai, K. Elstner and M.R. Trimble

patient-driven technique, to the EuroQol EQ-5D, a generic, single index which yields a single, summary score. Study 1: patients starting adjunctive AED therapy Drug study: subjects and methods

Patients attending a follow-up appointment at the outpatient epilepsy clinics at Queen Square were approached after their medical consultation. The patients recruited were about to start on one of the five antiepileptic drugs (vigabatrin, clobazam, lamotrigine, gabapentin and topiramate) as add-on therapy. Those willing to take part, and who gave informed consent, were offered the choice of a telephone interview at home as an alternative to a face-to-face interview and most patients chose this option. The timing of the interviews was: (i) baseline, (ii) 3 months from baseline and (iii) 6 months from baseline. Only the baseline and 6-months followup data are reported here. Seizure frequency and seizure severity were assessed using the National Hospital Seizure Severity Scale (O’Donoghue et al., 1996). Quality of Life (QOL)

QOL was measured by the QOLAS (Kendrick and Trimble, 1994) and the EuroQol instrument, the EQ-5D (Brooks, 1996; EuroQol Group, 1990). These are briefly described below. The QOLAS

The Quality of Life Assessment Schedule (QOLAS) is an individualized QOL assessment technique which is tailored to each individual patient, and is a revised version of the QoLASCA, a method originally based on the Repertory Grid Technique (Kendrick, 1997; Kendrick and Trimble, 1994; McGuire, 1991). The full QoLASCA technique was somewhat burdensome and the revised method (QOLAS) has been considerably streamlined. Two main aspects of the original theoretical work have been maintained: (i) the original emphasis (in order to assess therapeutic outcome) on a careful and comprehensive interview, recording items of importance to the patient in the patient’s own words; (ii) the idea that QOL is a function of the conceptual distance between ‘how I am now’ and ‘how I would like to be’, the gap between actuality and expectation. This is known in the medical literature as ‘Calman’s gap’, since Calman suggested that a key aim of medical care should be to narrow the gap between a patient’s hopes and expectations and what actually happens (Calman, 1994). The QOLAS has been used in studies of patients with epilepsy (Selai and Trimble, 1998; Selai et al., 2000), dementia (Selai et al., 2001) and Gilles de la Tourette syndrome (Elstner et al., 2001).

327

Measures to assess quality of life in epilepsy QOLAS interview

The QOLAS interview used in this study is as follows: 1. Introduction and rapport-building. 2. The respondent is invited to recount what is important for his/her QOL and ways in which their current health condition is affecting their QOL. Key constructs are extracted from this narrative. Prompting is sometimes required. 3. In total, ten ‘constructs’ are elicited, two for each of the following domains of QOL: physical, psychological, social, daily activities and cognitive functioning (or well-being). 4. The patient is asked to rate how much of a problem each of these is now on a 0–5 scale where 0⫽no problem; 1⫽very slight problem; 2⫽mild problem; 3⫽ moderate problem; 4⫽big problem and 5⫽it could not be worse. 5. The patient is asked to rate how much of a problem they would ‘like’ each of these to be ideally on a 0–5 scale as above. 6. At follow-up interview, the respondent’s individual constructs are read out to them and they are invited to re-rate each on the 0–5 scale for how much of a problem there is with each ‘now’. QOLAS scoring

(i) For each construct, the ‘like’ score is subtracted from the ‘now’, giving a score for the distance between expectation and reality. (ii) The scores, calculated in (i) above, for the two constructs per domain are summed to give a domain score out of ten. The total for each of the five domains is summed to give an overall QOLAS score out of 50. The EuroQol EQ-5D

The EQ-5D is a generic instrument for describing and evaluating health-related quality of life, developed to complement other forms of quality of life measure, and to generate a cardinal index of health, thus giving it considerable potential for use in economic evaluation (Brooks, 1996; EuroQol Group, 1990). It has three components, each providing separate data. In the first part, which yields a simple descriptive profile, the respondent rates his own health today on five questions, one for each of the dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Each question has three response options: no problems, some problem and extreme problems. This descriptive classification thus defines 243 possible health states. The respondent next rates his own health, today, on a visual analogue scale (VAS), calibrated from 0–100 with the end-points 100⫽best imaginable health state and 0⫽worst possible health state. Finally, valuations for each of the 243 health states have been obtained and so, according to how the respondent has rated himself on the descriptive profile, the corresponding utility value can be ascertained.

328

C.E. Selai, K. Elstner and M.R. Trimble

The EQ-5D has been used in a number of clinical studies, but in only one study in epilepsy, to our knowledge. In this study, the EQ-5D VAS was used in a comparison of four preference measures (Stavem, 1998) but data on the HRQOL of patients was not presented. Side effects, adverse events, defined as any epilepsy-related health event requiring urgent medical attention, and the reason for stopping medication were also recorded. The outcome criterion, chosen after review of the current literature, was ⱖ50% reduction in seizures. We do not report the relative performance of each individual drug. Drug study: statistics

The data were analysed by the chi-square statistic or paired t-tests, two-tailed, as appropriate. The QOLAS data were normally distributed, as were the VAS scores (although the latter were slightly skewed). The EQ-5D utility data were markedly skewed and so the nonparametric Wilcoxon signed-ranks test was performed. Drug study: results

A total of 146 patients who were about to start on one of the following five AEDs were recruited: vigabatrin, clobazam, lamotrigine, gabapentin and topiramate. These included 125 patients on whom complete data for 6 months were collected and 21 patients who failed to attend follow-up. The mean age of the patients was 37.2. Thirty-five per cent of patients were working or studying (at least part-time). Of the 125 patients, 15 started on vigabatrin (of which 10 were male), 20 on clobazam (8 male), 26 on lamotrigine (14 male), 17 started on gabapentin (8 male) and 47 on topiramate (28 male). Of the other 21 patients, 3 had started on vigabatrin, 4 on clobazam, 6 on gabapentin, 6 on lamotrigine and 2 on topiramate. Drug study: seizures at baseline

Almost half of the patients (49%) had experienced more than one type of seizure in the last month and 8% had experienced more than two seizure types. At baseline, most of our patients were classified as having severe epilepsy (Vickrey et al., 1995). A total of 115 patients (92%) reported having more than 12 seizures in the previous year, 6 patients (5%) were having from 2–12 seizures, no patients were seizure-free or having only auras. The remaining patients (3%) reported having seizures in their sleep and were unable to report seizure frequency. The percentage of patients in each outcome group experiencing convulsions at baseline was identical (both 46%). Drug study: seizure reduction

Table 22.1 summarizes the status of the patients at the 6-month follow-up. Forty-six patients (37%) had achieved 50% reduction in seizures (27 male; 19 female). There

329

Measures to assess quality of life in epilepsy

Table 22.1. Drug study: clinical status at 6-months follow-up

Status at 6 months

Patients (n⫽125)

%

Still on drug Experiencing side effectsa Experienced serious adverse eventsb 50% or more reduction in seizures

75 49 15 46

60 39 12 37

Did not attend follow-up interview

21

—

Notes: Side effects as reported by patients and attributed by them to the add-on therapy. b Serious adverse events are epilepsy-related events requiring urgent medical intervention. a

Table 22.2. Drug study: summary of outcome measures scores

Outcome measure

ⱖ50% reduction baseline mean ± (SD)

ⱖ50% reduction follow-up mean ± (SD)

⬍50% reduction baseline mean ± (SD)

⬍50% reduction follow-up mean ± (SD)

QOLAS EQ-VAS EQ-utility EQ-utility median

31.85 (10) 67.85 (23) 0.86 (0.20) 1

23a (11) 75b (17) 0.89 (0.16) 1

32.85 (8) 63.85 (19) 0.85 (0.16) 0.85

30.85 (11) 64.85 (20) 0.85 (0.19) 0.85

Notes: P⫽0.001, b P⫽0.02.

a

was no significant difference in age between those patients who achieved 50% seizure reduction (mean age⫽36 years) and those who did not (mean age⫽38 years). Drug study: quality of life measures

The three issues most frequently raised in each of the five QOLAS domains are reported in Appendix 22.1 (a full report on these qualitative data is available from the authors). Table 22.2 summarizes the quantitative QOLAS and EQ-5D results for this study. The group whose seizures were reduced by 50%, showed significantly lower QOLAS scores i.e. a significant improvement in HRQOL, at follow-up compared to baseline (t⫽6.18, P⬍0.001). Drug study: EQ-5D profile scores

Tables 22.3, 22.4 and 22.5 report descriptive EQ-5D data. Table 22.3 shows the baseline profile scores, for each of the five domains, for the whole group (n⫽125)

330

C.E. Selai, K. Elstner and M.R. Trimble

Table 22.3. Drug study: comparison of baseline EQ-5D profile data with UK norms. Percentage reporting ‘no problems’

EQ domain

Our study (n⫽125) %

UK survey (n⫽3395) %

Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

88 93 78 82 65

82 96 84 67 79

Table 22.4. Drug study: EQ-5D descriptive health profile data. ≥ 50% seizure reduction group: baseline and follow-up (n⫽46)

No problems

Some problems

Severe/extreme

EQ domain

t⫽1

t⫽2

t⫽1

t⫽2

t⫽1

t⫽2

Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

40 (87)a 43 (93) 35 (76) 36 (78) 36 (78)

39 (85) 44 (96) 38 (83) 39 (85) 29 (63)

6 (13) 3 (07) 9 (20) 10 (22) 8 (17)

7 (15) 2 (04) 8 (17) 7 (15) 16 (35)

0 (0) 0 (0) 2 (4) 0 (0) 2 (5)

0 (0) 0 (0) 0 (0) 0 (0) 1 (2)

Note: a Figures in parentheses are percentage of patients.

Table 22.5. Drug study: EQ-5D descriptive health profile data. < 50% seizure reduction group: baseline and follow-up (n⫽79)

No problems

Some problems

Severe/extreme

EQ domain

t⫽1

t⫽2

t⫽1

t⫽2

t⫽1

t⫽2

Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

70 (89)a 73 (92) 63 (80) 66 (84) 45 (57)

70 (89) 75 (95) 68 (86) 61 (77) 45 (57)

9 (11) 6 (08) 15 (19) 13 (16) 30 (38)

9 (11) 4 (5) 11 (14) 16 (20) 31 (39)

0 (0) 0 (0) 1 (1) 0 (0) 4 (5)

0 (0) 0 (0) 0 (0) 2 (3) 3 (4)

Note: a Figures in parentheses are percentage of patients.

331

Measures to assess quality of life in epilepsy

in comparison to UK normative data (Kind et al., 1998). This table shows the number of patients reporting ‘no problem’ for each of the EQ-5D domains. In our study, fewer patients reported ‘no problem’ in the domain ‘anxiety/depression’ compared with the UK survey. On the other hand, more patients reported ‘no problems’ with pain/discomfort than the UK survey. Table 22.4 shows descriptive data comparing baseline to follow-up for the group who achieved 50% seizure reduction (n⫽46). Table 22.5 shows descriptive data comparing baseline to followup for the group who did not achieve 50% reduction in seizures. At follow-up, there was no significant difference between the two outcome groups on the health profile. Drug study: EQ-5D VAS scores

Most patients queried the VAS. Forty-seven per cent of patients said they thought that ‘health’ did not include their epilepsy. These patients said that if the VAS was to include epilepsy the score would be up to 70 VAS points lower. Table 22.2 summarizes the EQ-VAS scores for the two groups of patients at baseline and follow-up. The group whose seizures were reduced by 50%, showed significantly higher VAS scores i.e. a significant improvement in HRQOL/health status (t⫽⫺2.48, P⬍0.02). Drug study: EQ-5D utility scores

Table 22.2 summarizes the EQ-utility scores for the two groups of patients at baseline and follow-up. The group whose seizures were reduced by 50% or more did not show a statistically significant improvement on the utility score (Wilcoxon Z⫽⫺0.470; P⫽0.64). Drug study: summary

The results of this study suggest that HRQOL improves in patients with severe epilepsy on adjunctive treatment who experience a 50% or greater seizure reduction. This is the first study to report health status using the EQ-5D in patients with epilepsy. The QOLAS and the EQ-VAS were sensitive to change but the EQ-5D profile and EQ-5D utility were not responsive. Compared with a large UK survey, a similar percentage of patients in our study reported any problem in each EQ-5D domain. Fewer patients in our study reported any pain/discomfort compared with the UK survey but more of our patients reported any problem with anxiety/depression. Study 2: epilepsy surgery Surgery: subjects and methods

A total of 145 patients undergoing evaluation for definitive treatment for intractable epilepsy were interviewed during their stay on the telemetry unit of the National Hospital, Queen Square. Quality of life was assessed using the Quality of

332

C.E. Selai, K. Elstner and M.R. Trimble

Life Assessment Schedule (QOLAS); the Epilepsy Surgery Inventory (ESI-55) and the EuroQol EQ-5D. QOL measures

The QOLAS and the EQ-5D have been described above. The Epilepsy Surgery Inventory (ESI-55)

The ESI-55 consists of the generic SF-36 plus a number of epilepsy-specific questions (Vickrey et al., 1992). The scoring produces 11 subscales (health, energy, QOL, social functioning, emotional functioning, cognitive functioning, roleemotional, role-memory, role-physical, physical function and pain) and three composite scores for physical health, mental health and role functioning. Surgery study: statistics

The data were analysed by the chi-square statistic or paired t-tests, two-tailed, as appropriate. The EQ-5D utility data were markedly skewed and so the nonparametric Wilcoxon signed ranks test, was performed. Criterion validity and internal reliability were assessed using Spearman’s rank correlations and the coefficient alpha. The differences between the QOLAS scores at baseline and follow-up were tested using the Wilcoxon signed ranks test. Epilepsy surgery: results Seizure reduction

A subgroup of 40 patients was interviewed at follow-up (mean time to followup⫽1 year). Of these 40 patients, 15 had not had surgery at follow-up interview and 25 were postsurgery. Sixteen (64%) of the surgical patients had left temporal lobe resection, four patients (16%) had right temporal lobe resection and 5 (20%) had extra temporal lobe resection. Of the 25 patients who had undergone surgery, 22 had ⱖ75% reduction in seizures and 3 patients did not. Of the 15 patients who had not gone on to have surgery, no patient achieved a ⱖ75% seizure reduction. These data are summarized in Table 22.6. The mean duration of epilepsy in years was 23.1 (SD 9.9) and the median was 24 years. The mean age was 32.8 years (SD 8.6) and the median 31.0 years. The mean follow-up period was 14.3 months (SD 8.7) and median 12.5 months. The mean age of onset of epilepsy was 9.6 years (SD 9.4) and the median was 8 years. Table 22.7 summarizes the patient outcome data. QOLAS scores

The results for the QOLAS at baseline and at follow-up are presented in Table 22.7. There was significant improvement in all QOLAS domains at follow-up in the operated group who achieved 75% or greater reduction in seizure frequency.

333

Measures to assess quality of life in epilepsy

Table 22.6. Epilepsy surgery: summary of patient data

Number of patients Male Female No surgery Surgery

⬍75% seizure reduction

ⱖ75% seizure reduction

18 8 10 15 3

22 7 15 0 22

Table 22.7. Epilepsy surgery: summary of outcome measures scores

Outcome measure

ⱖ75% reduction mean (SD) t⫽1

ⱖ75% reduction mean (SD) t⫽2

No surgery mean (SD) t⫽1

No surgery mean (SD) t⫽2

QOLAS EQ-VAS ESI-CMH ESI-CPH ESI-CRF

32.3 61.6 62.2 73.2 69.6

17.1a 76.6a 74.8b 82.9b 78.5b

31.3 64.6 59.9 65.6 56.5

29.3 69.1 63.4 71.1 67.4

(8.0) (20.3) (14.3) (14.0) (22.9)

(8.8) (15.6) (12.1) (11.6) (20.8)

(6.7) (17.4) (14.9) (26.8) (27.1)

(8.3) (13.6) (14.5) (18.3) (27.9)

Notes: ESI-CMH, ESI-55 composite mental health score; ESI-CPH, ESI-55 composite physical health score; ESI-CRF⫽ESI-55 composite role functional score. a P⫽0.05, b P⫽0.01.

ESI-55 composite scores

Table 22.7 summarizes the ESI-55 composite scores. At follow-up, there was a statistically significant improvement in QOL in comparison with baseline scores, on two of three ESI-55 composite scores: Composite Mental Health (CMH) t⫽⫺4.3; df⫽21, P⫽0.0001, 95% CI (⫺18.7; ⫺6.5); Composite Physical Health (CPH) t⫽⫺4.4, df⫽20, P⬍0.0001, 95% CI (⫺14.8; ⫺5.3). Although Composite Role Functioning (CRF) scores showed improvement at follow-up, they did not reach statistical significance. EQ-5D

Tables 22.8 and 22.9 show descriptive EQ-5D profile data at baseline and at followup. As with the drug study, most patients queried the EQ-5D visual analogue scale (VAS). Forty-two per cent of patients said they thought that ‘health’ did not include their epilepsy. Again, if the VAS was to include their epilepsy, they would

334

C.E. Selai, K. Elstner and M.R. Trimble

Table 22.8. Baseline EQ-5D profile data (n⫽22)

EQ domain

No problema

Some problems

Severe problems

Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

86 86 72 82 59

9 14 18 18 32

5 0 9 0 9

Note: Figures are percentage of patients.

a

Table 22.9. EQ-5D profile data at follow-up

EQ domain

No problemsa

Some problems

Severe problems

Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

90 100 89 85 80

10 0 11 15 20

0 0 0 0 0

Note: a Figures are percentage of patients.

have given a score up to 70 points lower on the VAS scale. For example, one patient said that his ‘health’ was excellent in general and he felt well on the day of the interview so he scored himself as 80. On reflection, he added that if the score was supposed to include his epilepsy and seizures, then he would have adjusted it to 30. Although we noted the qualitative data, we took the score each patient originally gave for their health since this is what the EQ-5D asks. Table 22.7 summarizes the EQ-5D VAS scores for the two groups of patients at baseline and follow-up. There was significant improvement at follow-up (t⫽⫺2.6, df⫽20, P⫽0.02, 95% CI (⫺26.0; ⫺2.8)). Table 22.10 shows the baseline profile scores for the whole group (n⫽145) in comparison with UK normative data (Kind et al., 1998). Fewer patients in our surgery study reported ‘no problem’ on most of the EQ-5D domains compared with the UK survey. A marked difference was observed for the ‘anxiety/depression’ domain with 49% of the surgery patients reporting ‘no problem’ compared to 79% of the UK population.

335

Measures to assess quality of life in epilepsy

Table 22.10. Surgery study: comparison of baseline EQ-5D profile data with UK norms. Percentage reporting ‘no problems’

EQ domain

Surgery study (n⫽145) %

UK survey (n⫽3395) %

Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

84 86 70 81 49

82 96 84 67 79

Table 22.11. Epilepsy surgery: summary of outcome measures scores

ⱖ75% reduction

EQ-Utility

t⫽1 t⫽2

No operation

Median

Mean (SD)

Median

Mean (SD)

0.85 1.00

0.81 (0.31) 0.91 (0.11)

0.85 0.85

0.77 (0.24) 0.90 (0.11)

EQ-5D utility scores

The EQ-5D utility scores at baseline and at follow-up were: mean 0.81 (median 0.85) and mean 0.91 (median 1.0) respectively (Table 22.11). The difference was not significant. Summary

The figures for the epilepsy surgery study are similar to the figures for the antiepileptic drug audit. The QOLAS, the EQ-VAS and 2/3 ESI-55 Composite Scores were sensitive to change as shown by statistically significant changes in scores. For the two patient groups there were differences in the EQ-5D profile at baseline but the percentages of problems experienced by the two groups at follow-up were similar. The EQ-5D utility scores showed improvement but the changes were not significantly different and this finding requires further discussion. Discussion The results of this study suggest that HRQOL improves both in patients with severe epilepsy on adjunctive treatment who experience a ⱖ50% seizure reduction and in patients who have undergone epilepsy surgery who achieved a ⱖ75% seizure

336

C.E. Selai, K. Elstner and M.R. Trimble

reduction. The QOLAS, the EQ-VAS and two out of three subscales of the ESI-55 were sensitive to change but the EQ-5D profile and EQ-5D utility were not responsive. There are two main possible explanations for our findings. Choice of clinical outcome

There is no simple relationship between seizure severity, seizure frequency and the consequences of epilepsy, and there is debate whether a reduction in (but not elimination of) seizures does lead to an improvement in HRQOL (Smith et al., 1995). Even if a new agent reduces the seizure frequency by ⱖ50%, patients will continue to experience seizures, and these may be unpredictable with severe ictal or postictal phenomena (Baker, 1995). It has been suggested that the ultimate goal of new antiepileptic drugs, therefore, should be seizure freedom (Walker and Sander, 1996). Patients about to undergo surgical treatment for epilepsy often have high, but not necessarily realistic, expectations of significant positive changes postoperatively (Baxendale and Thompson, 1996). The EQ-5D might not be sensitive to the relatively modest changes perceived after reduction in (but not elimination of) seizures. Ability of the QOLAS and the EQ-5D to detect change in HRQOL

The QOLAS asks patients to nominate the HRQOL topics of concern to them. They are able to both choose the items and, importantly, discuss them in their own language or idiom. The QOLAS therefore taps each patient’s perceived change in their own HRQOL rather than objectively verifiable changes in status e.g. role/social functioning. The EQ-5D, a generic instrument, might not have basic face validity for particular patient populations, and epilepsy in particular. Comparing our audit data with the UK norms, the only domain where the epilepsy patients reported significantly more problems was the anxiety/depression domain. Numerous studies, however, have shown that seizures and the stigma of epilepsy considerably impair HRQOL, and HRQOL is certainly poor in patients with intractable epilepsy, who are taking a cocktail of antiepileptic drugs, and who have been referred to a centre of tertiary referral such as Queen Square. In the surgery study, where the patients were interviewed in their hospital beds attached to video-telemetry monitoring equipment, more problems (compared to UK norms) were reported in most of the EQ-5D domains with a marked difference in percentages reported for the ‘anxiety/depression’ domain. Nevertheless, the utility scores for the patients who achieved ⱖ75% seizure reduction were not significantly different from baseline. For patients with intractable epilepsy, the occurrence of seizures and stigma of epilepsy lead to a variety of physical, psychological and social problems which have

337

Measures to assess quality of life in epilepsy

impact on many areas such as self-esteem, career prospects, family, leisure, (in)ability to drive and relationships (Cramer, 1996). All of our patients mentioned some problems due to epilepsy with at least one of: work; studies; career; housework; child-care; leisure; sports; relationships, family and social life, yet 78% of the audit patients (and 70% of the surgery patients being interviewed in their hospital beds) ticked ‘no problem’ with the EQ-5D Usual Activities domain at baseline. An explanation could be that the EQ-5D might be more useful for acute rather than chronic illness since it does not capture chronic problems to which the patient has adapted. This phenomenon of coping/adjustment resulting in the underreporting of problems on HRQOL instruments by patients with epilepsy has been previously discussed (Devinsky et al., 1995). Because the EQ-5D utility score is dependent upon the EQ-5D profile, and given the problems outlined above, it is not surprising that the utility scores were not sensitive to changes in seizure outcome. Most of our patients with epilepsy queried the VAS. The most common comment (47% of patients) was that ‘health does not include epilepsy’. Most of these patients said that, if the VAS was to include epilepsy, the score would be up to 70 points lower. Even though the VAS was sensitive to change, these qualitative data raise questions about the interpretation of the numerical data obtained from the VAS in this patient group. It has previously been reported that patients with epilepsy have difficulties completing visual analogue scales (Fallowfield, 1994). Conclusions The QOLAS, an ‘individualized’, patient-centred measure was demonstrated to be sensitive to change. The EQ-5D VAS, which is measuring some aspect of health-related HRQOL, is sensitive to change. Our data suggest, however, that the EQ-5D does not have face validity for patients with severe epilepsy. The EQ-5D profile, and, therefore, the EQ-5D utility scores, might not have validity for this patient group. We recommend that further research is required into the suitability of generic measures to assess QOL in patients with epilepsy. Meanwhile, we would urge all researchers to be cautious in their choice of QOL measure. Acknowledgements For assistance with recruitment of patients we gratefully acknowledge the assistance of: Prof. S. Shorvon, Prof. J. Duncan, Dr J.W.A. Sander, Dr D. Fish and Dr S. Smith. Thanks also to Dr M. O’Donoghue who provided training in the use of the National Hospital Seizure Severity Scale. An earlier version of this paper was presented at the EuroQol Plenary meeting in Hannover, 1–3 October 1998.

338

C.E. Selai, K. Elstner and M.R. Trimble

R E F E R E N C ES Baker, G. (1995). Health-related quality of life issues: optimizing patient outcomes. Neurology, 45 (Suppl. 2), S29–34. Baxendale, S.A. and Thompson, P.J. (1996). ‘If I didn’t have epilepsy . . .’: patient expectations of epilepsy surgery. J Epilepsy, 9, 274–81. Bladin, P.F. (1992). Psychosocial difficulties and outcome after temporal lobectomy. Epilepsia, 33, 898–907. Brazier, J. and Deverill, M. (1999). A checklist for judging preference-based measures of health related quality of life: learning from psychometrics. Health Econ, 8, 41–51. Brooks, R.G. (1995). Health Status Measurement: A Perspective on Change. Basingstoke: Macmillan. Brooks, R. (1996). EuroQol: The current state of play. Health Policy, 37, 53–72. Calman, K.C. (1994). Quality of life in cancer patients: an hypothesis. J Med Ethics, 10, 124–7. Cramer, J. (1996). Quality of life assessment for people with epilepsy. In Quality of Life and Pharmacoeconomics in Clinical Trials, Second edition, ed. B. Spilker, pp. 909–18. Philadelphia, New York: Lippincott-Raven. Devinsky, O., Vickrey, B.G., Cramer, J. et al. (1995). Development of the quality of life in epilepsy inventory. Epilepsia, 36, 1080–104. Elstner, K., Selai, C.E., Trimble, M.R. and Robertson, M.M. (2001). Quality of life of patients with Gilles de la Tourette syndrome. Acta Psychiatr Scand, 103, 52–9. EuroQol Group (1990). EuroQol – a new facility for the measurement of health-related quality of life. Health Policy, 16, 199–208. Fallowfield, L. (1994). An overview of quality of life measurements. In Epilepsy and the Quality of Life, ed. M.R. Trimble and W.E. Dodson, pp. 85–98. New York: Raven Press. Fitzpatrick, R., Fletcher, A.E., Gore, S.M., Jones, D.R., Spiegelhalter, D.J. and Cox, D.R. (1992). Quality of life measures in health care. I: applications and issues in assessment. Br Med J, 305, 1074–7. Fraser, S.C.A., Ramirez, A.J., Ebbs, S.R. et al. (1993). A daily diary card for quality of life measurement in advanced breast cancer trials. Br J Cancer, 67, 341–6. Geddes, D.M., Dones, L., Hill, E. et al. (1990). Quality of life during chemotherapy for small cell lung cancer: assessment and use of a daily diary card in a randomised trial. Eur J Cancer, 26, 484–92. Gill, T.M. (1995). Quality of life assessment: values and pitfalls. J R Soc Med, 88, 680–2. Gill, T.M. and Feinstein, A.R. (1994). A critical appraisal of the quality of quality of life measurements. J Am Med Assoc, 272, 619–26. Guyatt, G.H., Berman, L.B., Townsend, M., Pugsley, S.O. and Chambers, L.W. (1987a). A measure of quality of life for clinical trials in chronic lung disease. Thorax, 42, 773–8. Guyatt, G.H., Townsend, M., Pugsley, S.O. et al. (1987b). Bronchodilators in chronic air-flow limitation. Am Rev Resp Disorders, 135, 1069–74. Hedrick, S.C., Taeuber, R.C. and Erickson, P. (1996). On learning and understanding quality of life: a guide to information sources. In Quality of Life and Pharmacoeconomics in Clinical Trials, Second edition, ed. B. Spilker, pp. 59–64. Philadelphia: Lippincott-Raven.

339

Measures to assess quality of life in epilepsy Hermann, B.P. (1995). The evolution of health-related quality of life assessment in epilepsy. Qual Life Res, 4, 87–100. Hermann, B.P., Wyler, A.R and Somes, G. (1992). Preoperative psychological adjustment and surgical outcome are determinants of psychosocial status after anterior temporal lobectomy. J Neurol Neurosurg Psychiatry, 55, 491–6. Hunt, S. (1999). The researcher’s tale: a story of virtue lost and regained. In Individual Quality of Life: Approaches to Conceptualisation and Assessment, ed. C.R.B. Joyce, C.A. O’Boyle and H. McGee, pp. 225–32. Amsterdam: Harwood Academic Publishers. Joyce, C.R.B., O’Boyle, C.A. and McGee, H. (eds.) (1999). Individual Quality of Life: Approaches to Conceptualisation and Assessment. Amsterdam: Harwood Academic Publishers. Juniper, E.F., Guyatt, G.H. and Jaeschke, R. (1996). How to develop and validate a new healthrelated quality of life instrument. In Quality of Life and Pharmacoeconomics in Clinical Trials, Second edition, ed. B. Spilker, pp. 49–56. Philadelphia, New York: Lippincott-Raven. Kendrick, A. (1997). Quality of life. In The Clinical Psychologist’s Handbook of Epilepsy, ed. C. Cull and L.H. Goldstein, pp. 130–48. London: Routledge. Kendrick, A.M. and Trimble. M.R. (1994). Repertory grid in the assessment of quality of life in patients with epilepsy: the Quality of Life Assessment Schedule. In Epilepsy and the Quality of Life, ed. M.R. Trimble and W.E. Dodson, pp. 151–63, New York: Raven Press. Kind, P., Dolan P., Gudex, C. and Williams, A. (1998). Variations in population health status: results from a United Kingdom national questionnaire survey. B Med J, 316, 736–41. Malgrem, K., Sullivan, M., Ekstedt, G., Kullberg, G. and Kumlien, E. (1997). Health-related quality of life after epilepsy surgery: a Swedish multicentre study. Epilepsia, 38, 830–8. Mays, N. and Pope, C. (1996). Rigour and qualitative research. In Qualitative Research in Health Care, ed. N. Mays and C. Pope, pp. 10–19. London: BMJ Publishing Group. McDowell, I. and Newell, C. (1987). Measuring Health: A Guide to Rating Scales and Questionnaires. Oxford: Oxford University Press. McGuire, A.M. (1991). Quality of life in women with epilepsy. In Women and Epilepsy, ed. M.R. Trimble, pp. 13–30. Chichester: John Wiley & Sons. McLachlan, R.S., Rose, K.J., Derry, P.A., Bonnar, C., Blume, W.T. and Girvin, J.P. (1997). Healthrelated quality of life and seizure control in temporal lobe epilepsy. Ann Neurol, 41, 482–9. O’Boyle, C.A., McGee, H.M., Hickey, A. et al. (1993). The Schedule for the Evaluation of Individual Quality of Life (SEIQoL). Administration Manual. Dublin: Royal College of Surgeons in Ireland. O’Donoghue, M. F., Duncan, J.S. and Sander, J.W.A.S. (1996). The National Hospital Seizure Severity Scale: A further development of the Chalfont Seizure Severity Scale. Epilepsia, 37, 563–71. Rose, K.J., Derry, P.A. and McLachlan, R.S. (1996). Neuroticism in temporal lobe epilepsy: assessment and implications for pre- and postoperative psychosocial adjustment and health-related quality of life. Epilepsia, 37, 484–91. Ruta, D.A., Garratt, A.M., Leng, M., Russell, I.T. and MacDonald, L.M. (1994). A new approach to the measurement of quality of life: the patient-generated index (PGI). Med Care, 32, 1109–26. Selai, C.E. and Trimble, M.R. (1998). Adjunctive therapy in epilepsy with new antiepileptic drugs: is it of any value? Seizure, 7, 417–18.

340

C.E. Selai, K. Elstner and M.R. Trimble Selai, C.E., Elstner, K. and Trimble, M.R. (2000). Quality of life pre- and post epilepsy surgery. Epilepsy Res, 38, 67–74. Selai, C.E., Trimble, M.R., Rossor, M. and Harvey, R.J. (2001). Assessing quality of life (QOL) in dementia: the feasibility and validity of the Quality of Life Assessment Schedule (QOLAS). J Neuropsychol Rehab, 11, 219–43. Smith, D., Baker, G.A., Jacoby, A. and Chadwick, D.W. (1995). The contribution of the measurement of seizure severity to quality of life research. Qual Life Res, 4, 143–58. Stavem, K. (1998). Quality of life in epilepsy: comparison of four preference measures. Epilepsy Res, 29, 201–9. Trimble, M.R. and Dodson, W.E. (ed.) (1994). Epilepsy and Quality of Life. New York: Raven Press. Tugwell, P., Bombardier, C., Buchanan, W.W. et al. (1990). Methotrexate in rheumatoid arthritis: impact on quality of life assessed by traditional standard item and individualized patient preference health status questionnaires. Arch Intern Med, 150, 59–82. Vickrey, B.G., Hays, R.D., Graber, J., Rausch, R., Engel, J. and Brook, R.H. (1992). A healthrelated quality of life instrument for patients evaluated for epilepsy surgery. Med Care, 30, 299–319. Vickrey, B.G., Hays, R.D., Engel, J. et al. (1995). Outcome assessment for epilepsy surgery: the impact of measuring health-related quality of life. Ann Neurology, 37, 158–66. Walker, M.C. and Sander, J.W.A.S. (1996). The impact of new anti-epileptic drugs on the prognosis of epilepsy: seizure freedom should be the ultimate goal. Neurology, 46, 912–14.

Appendix 22.1. QOLAS qualitative data. The three most frequently nominated items per QOLAS domain

341

QOLAS domain

Number of times nominating

Physical Seizure-related (injury, incontinence) Tiredness Drug side effects

153 34 21

Psychological Depression Anger (‘why me..?’) Anxiety

65 57 42

Social Inability to drive Restricted social life/leisure Family

59 66 38

Work/economic Unable to work/keep job Interference with career/promotion Discrimination (job application)

49 63 28

Cognitive Memory Concentration impaired Thinking (difficulties with)

91 45 33

Index

Numbers in italics indicate tables and figures. abreactive attacks 198 absence seizures frequent 72–4 risk of psychosis 42, 43 see also childhood absence epilepsy (CAE) absence status 44 action-slips 206 affective disorders in epilepsy anatomic substrates 28–9 and temporal lobe surgery 272–4 see also depression in epilepsy; dysphoric disorders; mania agoraphobia 228, 317 aggression 81 classification 81–2 clinical relevance 82, 83 in epilepsy anatomic substrates 30–1 diagnosis 96, 97 postictal psychosis 126–7 prevalence 84–5 social and psychological aspects 95–6 treatment 96–8, 99 types of behaviours 85–6 neurobiology 83–4 see also intermittent explosive disorder (IED) alcohol consumption men with panic disorder 228 precipitation of JME 48 alprazolam 229 alternative psychoses 45, 47, 249–50, 269 amitriptyline 300 amnesia dissociative 194, 200, 202 ictal 194, 200 amnesic barriers 202, 204 amygdala 234 in anxiety 31, 33, 234 direct thalamic input 235 false threat alarms 235–6 in fear-induced aggression 84 organization 24, 25 pathology in TLE with IED 87–90, 91

343

place in the limbic system 19–20 seizures originating in 234–5 amygdalohippocampectomy 23 amygdalotomy, bilateral 84 animals, aggression subtypes 82 anterior commissure 27 Anthropometric Laboratory 137 anticonvulsant drugs see antiepileptic drugs (AEDs) antidepressant drugs 97 interactions with AEDs 305 newer drugs 302–4 and seizures early animal data 299–300 early clinical data 300–1 newer drugs 304–5 suicide prevention 112–13 see also vagus nerve stimulation (VNS): antidepressant effects antiepileptic drugs (AEDs) antiglutaminergic 249 cognitive side effects 67, 142–3, 256–7 combined effects with seizures 257 different drug formulations 258–9 habituation 257 meta-analysis of studies 259–60 patient complaints 257–8 related to serum levels 258 GABAergic 248–9 interactions with psychotropic drugs 305, 309 positive psychotropic effects 243 epilepsy patients 250–1 psychiatric patients 250 psychiatric side effects 7, 241, 243 classical drugs 242–4 classification/terminology 242 clinical recommendations 252 mechanisms 248–50 new drugs 244–8 and preexisting mental state 249 questions regarding 242 quality of life study 326–31 in suicide attempts 108 see also specific drugs

344

Index antipsychotic drugs atypical 307 case report (clozapine in epilepsy) 308–9 classical 306 classification 306 interactions with other drugs 309 newer drugs 306–7, 308 receptor binding profiles 307 and seizures 307–8 antisocial personality disorder (APD) 82 anxiety in epilepsy 10 aggression 95, 97 temporal lobe surgery 274 experimental models 234 role of the amygdala 31, 33, 234 susceptibility to VNS 290, 291 see also antiepileptic drugs (AEDs): psychiatric side effects; panic disorder (PD) area tempesta 19 attention deficit hyperactivity disorder 243 attention deficits in FLE 169 auras 125, 195 autistic features 78 awakening epilepsy see idiopathic generalized epilepsy (IGE) Bear–Fedio scale 11, 13 behavioural disturbances in epilepsy 166–8 anatomic substrates 18–19 affective disorders 28–9 aggression 30–1 anxiety 31, 33 schizophrenia 29–30 assessment framework 70–1 epidemiological studies 70 frequent absence seizures 72–4 and learning disorders 63–4 prevention 78–9 subtle 71 see also aggression; antiepileptic drugs (AEDs): psychiatric side effects; specific types of epilepsy benzodiazepines 229 and aggression 96, 98 for postictal psychoses 128 psychotropic effects 243 and seizures 310 beta-carboline 229 blindsight 191 brain functional reserve 159 Briquet, Paul 210, 211–12 Broca, Paul 138 CAE see childhood absence epilepsy Calman’s gap 326 carbamazepine and clozapine 310 cognitive effects 143, 258, 259 and lithium 310

pharmacokinetic interactions 305, 309 psychotropic effects 243, 244, 250, 251 cats, defensive rage 30–1, 32 character neurosis in JME 54 Charcot, Jean-Martin see hysteria: Charcot’s work childhood absence epilepsy (CAE) 42 attention deficits 52–3 frequency of psychosis 42 personality traits 52 see also absence seizures: frequent chlorpromazine 306 cingulate gyrus 27, 166 citalopram 302, 303, 304 classification of psychiatric disorders in epilepsy 11, 13–15 clobazam 310 clomipramine 300, 301 clonazepam 229 clozapine 307, 308, 309–10 cognitive decline in refractory TLE 152, 153 duration effects 153–4 cross-sectional studies 154–7 exhaustion of brain functional reserve 159 education and 156, 157, 158, 159–60 measurement 154, 155 prefrontal metabolic disturbances 152 profiles in males and females 143, 144 senile plaque formation 144 temporal gate hypothesis 144–5 cognitive effects of epilepsy effects of seizures see seizures: cognitive effects factors for consideration 167–8 frequent localized discharges 75–6 Landau–Kleffner syndrome 77–8, 140 postictal state-dependent impairment 76–7 prevention 78–9 psychosocial 143 and quality of life 263 subtle, defined 71 TLE see cognitive decline in refractory TLE transitory cognitive impairment 74–5, 141–2 see also antiepileptic drugs (AEDs): cognitive side effects; dementia and epilepsy; learning disorders in epilepsy cognitive function 62 complex partial seizures attention tests on patients 53 and depression 29 continuous spike-wave discharges in slow wave sleep (CSWS) 140 corpus callosum 27 cortical inhibition and VNS 289 Damasio’s somatic marker hypothesis 166 defensive aggression 81, 82, 84 defensive rage (cats) 30–1, 32 degeneracy theory 136 déjà vu 125, 195 dementia, definitions 135 dementia and epilepsy historical links 135–7

345

Index possible relationships 139 cognitive effects of EEG discharges 141–2 cognitive effects of seizures 141 neurological disease with dementia and seizures 139–40 specific epilepsy syndromes 140–1 treatment effects see antiepileptic drugs (AEDs): cognitive side effects see also cognitive effects of epilepsy dentate gyrus 21, 22 depersonalization 193, 195–6, 231 depersonalization disorder 193 depression in epilepsy 10 aggression 95, 97 anatomic substrates 28–9 postoperative 127, 272–4 suicide risk 110 TLE and FLE 176, 178, 180–1 GABA hypothesis 248–9 in panic disorder 228 see also antiepileptic drugs (AEDs): psychiatric side-effects; dysphoric disorders derealization 193, 195–6, 231 desipramine 300 discharges see epileptiform discharges dissociated control theory 204–6 dissociation classification of disorders 190–1, 190 confusion over concept 189, 196 and epilepsy 194–7 facets 189–90 alteration in consciousness 193 defence mechanism 193–4 nonintegrated mental modules 190–3 and nonepileptic seizures see nonepileptic seizures: dissociative mechanisms dissociative amnesia 194, 200, 202 dissociative fugue 202–3 dissociative identity disorder 203 dopaminergic hypothesis of schizophrenia 29–30 dorso-lateral cortex 165 Down’s syndrome 139 doxepin 300, 301 dysphoric disorders 13, 110–11, 112, 273 educational performance childhood absence epilepsy 52 epidemiological studies 70 FLE and mTLE 179–80 electrical status epilepticus of slow wave sleep (ESES) 77–8 electroconvulsive therapy (ECT) 28, 288, 289 entorhinal cortex 20, 21–2, 23 epidemiology epilepsy 5 learning difficulties 70 psychiatric comorbidity in epilepsy 6 AED-related 241 hospital-based studies 9–10

population-based studies 6–9, 12 prevalence 10–11 psychiatric disorders 5–6 epilepsy on awakening see idiopathic generalized epilepsy (IGE) epilepsy surgery inventory (ESI-55) 332 epileptiform discharges cognitive effects 74–5, 141–2 early treatment 78–9 frequent frontal 71 hemispheric 76 left temporal 71 localized 75–6 episodic dyscontrol see intermittent explosive disorder (IED) EQ-5D 327–8, 336–7 ESES see electrical status epilepticus of slow wave sleep ESI-55 (epilepsy surgery inventory) 332 ethosuximide 46, 243, 244 EuroQuol EQ-5D 327–8, 336–7 event-related potentials 232–3 fear see anxiety; panic disorder felbamate 243, 246 FLE see frontal lobe epilepsy (FLE) flumazenil 229 fluoxetine 302, 303, 305 fluphenazine 308 fluvoxamine 304 folate supplementation 142 forced normalization 45–7, 111, 249–50 frontal lobe and aggression 83, 84, 166 assessment test design 170, 171 and behaviour disorders 165–6 cognitive consequences of surgery 169–70 structure 165 frontal lobe epilepsy (FLE) 164 behaviour and personality 175, 176 clinical personality inventory 177–9 education and employment 179–80 quality of life 175, 177, 180–1 state/trait 180–1 neuropsychology 168–70 postictal reorientation 174 seizure analysis negative phenomena 171, 172–3 nonconvulsive status epilepticus 173, 174 positive phenomena 171, 172 fugue dissociative 202–3 postictal 194, 195 Full Scale Intelligence Quotient (FSIQ) 156 gabapentin 243, 247, 250, 251 Galton, Sir Francis 137 gamma aminobutyric acid (GABA) in depression 248–9 in panic disorder 228–9

346

Index gamma aminobutyric acid (GABA) (cont.) regulation of seizures 289–90 generalized tonic-clonic seizures (GTCS; grand mal) 42, 43–4 glutamate 145, 289–90 hallucinations 45, 110–11, 195 haloperidol 306, 307, 309 head trauma 141, 216, 217 health-related quality of life (HRQOL) see quality of life (QOL) hemispherectomy 76 hemispheric specialization 28, 95 Hilgard’s neodissociation theory 201–4 hippocampal commissure 27–8 hippocampal formation 21, 25 hippocampus place in the limbic system 19, 20 in schizophrenia 30 sclerosis 23–4, 25, 26, 86–7, 159 structure 21–3, 25 trisynaptic circuit 21, 22–3 hyperarousal syndrome 93–4 hypersynchronization 24, 26 hypnotic phenomena 192 dissociated control theory 204 neodissociation theory 201–2, 204 hypothalamic hamartomas 139 hypothalamus 83 hysteria Briquet’s work 210, 211–12 Charcot’s work 211 classical clinical findings 215 clinical–anatomical method 215 descriptions of seizures 212 importance of trauma 211, 212, 216 male hysteria 216 Janet’s work 200–1 nonepileptic seizures 212–14 Richer’s drawings 212, 213, 214 role of the uterus 211 see also nonepileptic seizures; psychogenic pseudoseizures ictal aggression 85, 98 ictal amnesia 194, 200 ictal fear 125, 235 ictal psychoses 44 idiopathic generalized epilepsy (IGE) 41–2 classification of syndromes 42 precipitating factors 47–8 lifestyle 48–9 sleep deprivation 48, 49–50 psychoses 47 frequency 42, 43 relation to seizure type 42, 43–4 types 44–7 see also childhood absence epilepsy (CAE); juvenile myoclonic epilepsy (JME) IED see intermittent explosive disorder (IED) imipramine 300, 301, 303

implicit perception 192 inhibitory interneurons 24, 26 intelligence 62 intelligence quotient see IQ intelligence trace tests 154 interictal aggression 85–6, 98 interictal dysphoric disorder see dysphoric disorders interictal psychoses 13, 45 associated seizures 125 chronic, and temporal lobe surgery 270–1 demographic data 122–3 in different types of epilepsy 124 in dysphoric disorder 110–11 pathogenesis 111–12 psychopathological features 126 treatment 98 intermittent explosive disorder (IED) 83, 86 studies in TLE aims and rationale 86–7 amygdala pathology 87–90, 91 cortical abnormalities 90–2, 93, 94 demographic data 88–9 dual brain pathology 92, 93, 94–5 neuropsychological and psychometric parameters 92 treatment 98 IQ and aggression 86, 90 early studies in epilepsy 137–8 effects of duration of TLE 154–8 interictal vs. postictal psychoses 122–3 interpretation of serial scores 138–9 Janet, Pierre 190, 200–1 juvenile myoclonic epilepsy (JME) 42 frequency of psychoses 42, 43 frontal impairment 55–6 personality traits 54–5 precipitating factors 48 sleep deprivation 48, 49–50 kainic acid 31 lamotrigine cognitive effects 143, 145 interaction with sertraline 305 psychiatric side effects 7, 246 psychotropic effects 243, 246, 250, 251 Landau–Kleffner syndrome 65, 77–8, 140 learning disorders in epilepsy causes 64–5 effects of subtle seizures 66 links with behavioural disorders 63–4 permanent 65, 66 prevalence 64, 70 state-dependent 65–6 terminology 62–3 treatment 67 see also cognitive effects of epilepsy left-handedness 92, 221, 223

347

Index Lennox–Gastaut syndrome (LGS) 140 levetiracetam 248 limbic system 19–20 amygdala see amygdala hippocampus see hippocampus interhemispheric connections 27–8 lithium 310 locus coeruleus activation by VNS 289 magnetic resonance imaging (MRI) studies 86, 87, 88, 233–4 mania postoperative 127, 273, 274 valproate treatment 250 maprotiline 300 mental retardation aggression 82 defined 63 dementia 139 in epilepsy 63, 64, 70 mesial temporal lobe epilepsies (mTLE) 164 mesial temporal sclerosis 86–7 mianserin 304 mirtazapine 303–4 mismatch negativity (MMN) 232, 233 Monolog 72–3 mossy fibres 21, 22, 24, 25 MRI see magnetic resonance imaging (MRI) studies multiple subpial transection 78 narcolepsy (sleep epilepsy) 48, 49 NBI (Neurobehavioural Inventory) 13 NEAD see nonepileptic attack disorder nefazodone 303, 304 neodissociation theory 201–4 Neurobehavioural Inventory (NBI) 13 neuropsychology, origins of 138 neurosyphilis 137 nomifensine 300 nonconvulsive status epilepticus 44, 65, 72, 173, 174 nonepileptic attack disorder (NEAD) 14, 197 postoperative 275 nonepileptic seizures 189, 197 dissociative mechanism (evidence) amnesia 200 dissociative comorbidity 198 hypnotic susceptibility 199 traumatic experiences 199 dissociative mechanisms dissociated control theory 204–6 Janet’s analysis of hysteria 200–1 neodissociation theory 201–4 presentations 197–8 and temporal lobe surgery 275–6 terminology 197 see also psychogenic pseudoseizures nontemporal lobe epilepsies (non-TLE), frequency of psychosis 42 noradrenergic system and VNS 289

norepinephrine 289 nortryptiline 300 olanzapine 98, 307, 309 opioids 22, 28 orbito-frontal cortex 165, 166 oxidative stress 145 P3a and P3b 232 panic disorder (PD) 226 aetiology 227 cerebral blood flow 234 diagnosis 227 EEG studies 229–30 atypical panic attacks 230 derealization/depersonalization symptoms 231 event-related potentials 232–3 epidemiology 228 experimental models 234 false-alarm theory 235–6 GABA system implicated 228–9 MRI 233–4 panic attacks 227, 230, 290 suicide risk 228 Papez circuit 19, 20 paroxetine 303, 304, 305 PD see panic disorder Pearson, Karl 137 periaqueductal grey 30, 83–4 permanent cognitive impairment 71–2 personality disorders 276 case report 278–9 development 277–8 and temporal lobe surgery 276, 277 personality traits 7, 10 clinical inventory 177–9 FLE 175, 176, 177–9 IGE 50–1 childhood absence epilepsy 52–3 juvenile myoclonic epilepsy 54–6 TLE 11, 13, 175, 176, 178–9 see also behavioural disturbances in epilepsy petit mal see absence seizures phenobarbital behavioural effects 96, 97 cognitive effects 142, 259 and depression 242 and suicide 108 phenytoin cognitive effects 142–3, 258, 259 interactions with tricyclic drugs 305 psychiatric complications 243–4 pimozide 307–8 positron emission tomography (PET) 55, 234 postictal aggression 85, 98 postictal fugue 194, 195 postictal psychoses 44–5 aggression 85, 126–7 case reports 118–20 clinical studies 120

348

Index postictal psychoses (cont.) clinical studies (cont.) association with TLE 123–4 demographic data 122–3 duration 122 incidence 120–1, 270 lucid intervals 121–2 psychic auras 125 psychopathological features 126 recurrences 122 seizures 125 historical background 117–18 pathophysiological mechanism 128 prevention 128, 310 and temporal lobe surgery 127, 269–70 treatment 98, 127–8 postoperative psychoses 127, 271–2 posttraumatic stress disorder (PTSD) 320 predatory aggression 81, 82 prefrontal cortex 29, 165–6 prepsychotic dysphoria 45 primidone 242, 243 procaine 234 protriptyline 300 pseudoseizures see psychogenic pseudoseizures psychiatric comorbidity in epilepsy classification of psychiatric disorders 11, 13–15 epidemiology see epidemiology: psychiatric comorbidity in epilepsy pathogenesis 111–12 temporal lobe surgery candidates 268–9 see also specific disorders psychiatric disorders and temporal lobe surgery 267–8 affective disorders 127, 272–4 anxiety 274 case report 278–9 evaluation of surgical outcome 266–7 limitations of presented data 268 nonepileptic attacks 275–6 personality disorders 276, 277 psychiatric assessment strategies 279 psychoses 266, 269 chronic interictal 270–1 postictal 269–70 postoperative 127, 271–2 total psychiatric comorbidity 268–9 psychiatric effects of anticonvulsants see antiepileptic drugs (AEDs): psychiatric side effects psychic auras 125, 195 psychogenic pseudoseizures 210 emotional trauma 222 evidence for biological factors 220, 221, 223 historical review see hysteria intellectual and neuropsychological test findings 218–19, 220–2 left-handedness 221, 223 patient demographics 217 physical trauma 216, 217 see also nonepileptic seizures

psychometry, origins of 137–8 psychoses in epilepsy AED-related 241, 243 see also specific antiepileptic drugs epidemiology 8, 9–10, 11, 120–1 forced normalization 45–7, 111, 249–50 ictal 44 interictal see interictal psychoses in men 9 postictal see postictal psychoses postoperative 127, 271–2 relation to seizure type 42, 43–4 see also psychiatric disorders and temporal lobe surgery: psychoses; schizophrenia psychotherapy in epilepsy case reports agoraphobic avoidance 317 agitated depression 315–17 depressive withdrawal 318–19 fear of psychotic disintegration 317–18 clinicians’ views of psychotherapists 314 contraindications 320–1 follow-up study 321–2 handicaps to integration 314–15 indications 320 a neglected area 313–14 therapist’s roles 319–20 time frames 319 psychotropic drugs antidepressant see antidepressant drugs antipsychotic see antipsychotic drugs classification 300 other agents 310 see also specific drugs pyknolepsy see childhood absence epilepsy (CAE) Quality of Life Assessment Schedule (QOLAS) 326–7, 336 quality of life (QOL) assessment epilepsy surgery inventory (ESI-55) 332 EuroQuol EQ-5D instrument 327–8, 336–7 individualized techniques 324–5 QOLAS 326–7, 336 qualitative vs. quantitative methods 324 reasons for 323 types of measure 323–4 and cognitive function in epilepsy 263 FLE 175, 177, 180–1 and seizure frequency 325, 336 drug study 326–31, 341 surgery study 331–5 TLE 175, 177, 180–1 quetiapine 98, 307, 309 Rasmussen syndrome 76, 140 reboxetine 302, 303 receptor profiles antidepressants 303 antipsychotics 307 risperidone 98, 307, 309

349

Index Schaffer collaterals 21, 22 schema 201 schizophrenia in epilepsy anatomic substrates 29–30 epidemiology 8, 9–10 in men 9 school performance see educational performance seizures absence see absence seizures and antipsychotic drugs 307–8 and benzodiazepines 310 cognitive effects 141 acute/short-term 260–1 frequent absence seizures 72–4 impairment in FLE 172, 173 long-term 261–2 neuronal damage 65 nocturnal seizures 77 complex partial 29 FLE see frontal lobe epilepsy (FLE): seizure analysis nonepileptic see hysteria; nonepileptic seizures; psychogenic pseudoseizures prediction 167 subjective experience of 226 subtle 71 type related to risk of psychosis 42, 43–4 see also antidepressant drugs: and seizures; quality of life (QOL): and seizure frequency selective serotonin reuptake inhibitors (SSRIs) 302 and seizures 304–5 see also specific drugs senile plaques 144 serotonergic system and VNS 289 sertraline 302, 303, 304–5 sleep deprivation and IGE 47–50 sleep epilepsy 48, 49 Sneddon syndrome 139 sodium valproate 78 somatic marker hypothesis 166 somatoform seizures see nonepileptic seizures; psychogenic pseudoseizures somnambulism 201 spike-wave monitoring 72–3 SSRIs see selective serotonin reuptake inhibitors (SSRIs) state-dependent cognitive impairment 71–2, 76–7 state-dependent learning disorders 65–6 statistical parametric mapping 90 status epilepticus 64, 72 ESES 77–8 nonconvulsive 44, 65, 72, 173, 174 subtle seizures 71 suicide in epilepsy 107 attempted 107–8 dysphoric disorder 110–11, 112 interictal and peri-ictal psychopathology 109–11 neuropsychiatric data 109 overdose 108 postoperative 127

prevention 112–13 risk factors 108, 114 seizure control and 111–12, 114 suicide and panic attacks 228 sulpiride 98 surgery bilateral amygdalotomy 84 effects on learning 67 evaluation of outcome 266–7 frontal lobe, cognitive consequences 169–70 hemispherectomy 76 multiple subpial transection 78 quality of life study 331–5 see also psychiatric disorders and temporal lobe surgery swoon attacks 198 tantrum attacks 198 TCI see transitory cognitive impairment (TCI) temporal lobe epilepsy (TLE) 18 behaviour and personality compared with FLE 175–81 traits 11, 13 cognitive decline see cognitive decline in refractory TLE ictal fear 235 memory deficits 152 mesial (mTLE) 164 postictal memory impairment 174 psychoses forced normalization 45–6 frequency 43, 45 interictal 45, 124 postictal 121, 123–4, 127 quality of life 175, 177 senile plaques 144 see also intermittent explosive disorder (IED): studies in TLE; suicide in epilepsy temporal lobe surgery, psychiatric disorders see psychiatric disorders and temporal lobe surgery thiamine deficiency 142 tiagabine cognitive effects 143 psychiatric side effects 247 psychotropic effects 243, 250 TLE see temporal lobe epilepsy (TLE) topiramate cognitive impairment 260 psychiatric side effects 247–8 psychotropic effects 243, 250 transitory cognitive impairment (TCI) 65–6, 74–5, 141–2 trauma dissociation response 193–4, 199 emotional 212, 222 head 141, 216, 217 and pseudoseizures/hysteria 210–12, 216–17, 222 trimipramine 300 twin studies, schizophrenia 30

350

Index vagus nerve stimulation (VNS) 283–4 antidepressant effects epilepsy patients 284–6 nonepileptic patients 287–8 anxiety disorders 290, 291 epilepsy treatment 283–4 mechanisms of action cerebral 288–90 peripheral 290–1 procedure 283 valproate cognitive effects 143, 259–60 encephalopathy 244 psychotropic effects 243, 250

venlafaxine 302, 303 vigabatrin cognitive effects 67, 143 psychiatric side effects 7, 243, 244–6 viloxazine 300, 301 violence defined 81 in postictal psychosis 126–7 see also aggression VNS see vagus nerve stimulation (VNS) vocabulary intelligence test (MWT-B) 154 Wernicke, Karl 138