Epilepsy & Behavior 32 (2014) 1–8

Contents lists available at ScienceDirect

Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh

Review

The concept of symptomatic epilepsy and the complexities of assigning cause in epilepsy Simon Shorvon ⁎ UCL Institute of Neurology, Box 5, National Hospital for Neurology and Neurosurgery, Queen Square, London WC1N 3BG, UK

a r t i c l e

i n f o

Article history: Received 2 November 2013 Revised 16 December 2013 Accepted 21 December 2013 Available online 22 January 2014 Keywords: Symptomatic epilepsy Idiopathic epilepsy Provoked epilepsy Causes of epilepsy

a b s t r a c t The concept of symptomatic epilepsy and the difficulties in assigning cause in epilepsy are described. A historical review is given, emphasizing aspects of the history which are relevant today. The historical review is divided into three approximately semicentenial periods (1860–1910, 1910–1960, 1960–present). A definition of symptomatic epilepsy and this is followed by listing of causes of symptomatic epilepsy. The fact that not all the causes of idiopathic epilepsy are genetic is discussed. A category of provoked epilepsy is proposed. The complexities in assigning cause include the following: the multifactorial nature of epilepsy, the distinction between remote and proximate causes, the role of nongenetic factors in idiopathic epilepsy, the role of investigation in determining the range of causes, the fact that not all symptomatic epilepsy is acquired, the nosological position of provoked epilepsy and the view of epilepsy as a process, and the differentiation of new-onset and established epilepsy. The newly proposed ILAE classification of epilepsy and its changes in terminologies and the difficulties in the concept of acute symptomatic epilepsy are discussed, including the inconsistencies and gray areas and the distinction between idiopathic, symptomatic, and provoked epilepsies. Points to be considered in future work are listed. © 2013 Elsevier Inc. All rights reserved.

This paper reviews several aspects of the concept of ‘symptomatic epilepsy’ and more broadly the issues of attributing cause in epilepsy. A historical overview will first be given, listing areas which are important to our current understanding of ‘causation’ in epilepsy. The listing of causes and the problems of classification will then be covered. The question of the nosological position of provoking factors will be discussed. Finally, a critique of the concept of acute symptomatic epilepsy will be given. The concept of symptomatic epilepsy is not as simple as at first sight it might appear, and the purpose of this paper is to draw attention to the inherent complexities inherent in assigning causation in epilepsy.

1. Historical perspective The modern history of symptomatic epilepsy can be divided conveniently into three periods.

1.1.1. The distinction between predisposing and exciting causes All authors of the period, without exception, considered epileptic seizures to have two distinct causal components: an innate predisposition (a predisposing cause; a diathesis) and a precipitating (exciting) cause. Although different authors used the terms differently, the predisposition was considered by most to be largely inherited and the exciting causes to be external or provoking factors (the position of structural and congenital causes was ambiguous). This dichotomy has been arguably unjustly neglected in recent times. The analogy of gunpowder and the match was often used (for instance, by Sieveking [1]), even before Jackson's famous definition of the epileptic seizure as a ‘discharge’. Spratling [2], the doyen of American epileptologists, even attempted to explain the contributions of the exciting and predisposing causes mathematically, writing that: if it took 100 points to induce a seizure in an individual, a predisposition could contribute 60 points and an exciting cause 40 points, whereas if the predisposition contributed only 40 points, it would require an exciting cause to have 60 points in order to reach the ‘seizure point’. This concept is also the basis of the ‘seizure threshold’ widely referred to today.

1.1. 1860–1910 Many concepts of etiology of epilepsy were formulated between about 1860 and 1910, and several are worthy of special mention for their insights which are relevant to us today. ⁎ Fax: +44 207 6762155. E-mail address: [email protected]. 1525-5050/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yebeh.2013.12.025

1.1.2. The introduction of the term ‘symptomatic epilepsy’ The first use of the term ‘symptomatic epilepsy’ that I find was by John Russell Reynolds in 1861 [3], who classified epilepsy into 4 categories (idiopathic, eccentric (syn: sympathetic), diathetic, and symptomatic). Symptomatic epilepsy was defined as epilepsy in which convulsions are due to “more or less contiguous structural disease of the brain. Thus, an intracranial tumour, a chronic inflammatory condition of

2

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8

the meninges, softening or disintegration of the brain substance, or any other structural change in the nervous centres … may set up that peculiar interstitial or molecular change which is the immediate cause of convulsion.” This is a definition which is the basis of that used today although the eccentric and diathetic epilepsies would now be included within the symptomatic category. It should be pointed out too that Reynolds did not fall into the mistake of considering all symptomatic epilepsies to be acquired epilepsy, a mistake still sometimes perpetrated today.

1.1.5. The theory that ‘seizures beget seizures’: epilepsy as a process Another unique contribution made by Gowers [6] to “cause” in epilepsy is worthy of mention, and his theory was as follows:

1.1.3. The distinction between proximate and remote causes In the 19th century, a distinction was sometimes made between proximate and remote causes. This important distinction was expounded in its fullest form by Hughlings Jackson in 1874 (see Taylor [4]), who in this was following the lead of his mentor Russell Reynolds. He considered the proximate cause to be the actual cellular disturbance at the epileptic focus, and the remote causes to be those which triggered this disturbance (such as brain tumors, stroke, and infection). He viewed the epileptic seizure as the explosive release of abnormal energy (just as gunpowder can store energy that is liberated when firing the gun). He considered that the reason for the abnormal levels of stored energy was ‘abnormal nutrition’, and it was this physiological abnormality which he considered the proximate cause. It is worth quoting him at length:

This concept (“seizures beget seizures”) was also widely accepted in the time of Gowers. Gowers and others, therefore, made a major distinction between new-onset epilepsy (which was reversible with active therapy) and chronic epilepsy (which was largely incurable). Epilepsy was seen as a process, and the cause of the seizures was the maturation of this process. The molecular science has demonstrated a range of changes which might underlie this process, and this topic continues to excite debate.

‘The confusion of two things physiology and pathology under one (pathology) leads to confusion in considering “causes”. Thus, for example, we hear it epigrammatically said that chorea is “only a symptom” and may depend on many causes. This is possibly true of pathological causation; in other words it may be granted that various abnormal nutritive processes may lead to that functional change in grey matter which, when established, admits occasional excessive discharge. But physiologically, that is to say, from the point of view of Function, there is but one cause of chorea – viz. instability of nerve tissue. Similarly in any epilepsy, there is but “one cause” physiologically speaking – viz. the instability of the grey matter, but an unknown number of causes if we mean pathological processes leading to that instability.’ Jackson defined abnormal physiology in the narrow and specific meaning of: ‘the departure of the healthy function of nerve tissue. That function is to store up and to expend force … in epilepsy, the cells store up large quantities and discharge abundantly on very slight provocation: there is what I call instability or what is otherwise spoken of as increased excitability’. Our present concepts of cause currently to an extent fail fully to appreciate this very important insight. Causal classifications of epilepsy would be very different if we thought in terms of physiological causal mechanisms, rather than pathologies. 1.1.4. The concept of the neurological trait Another most interesting aspect of the concept of ‘cause’ in the late nineteenth century, but one which is very relevant, is the concept of the neurological trait (syn: neurological taint, neuropathic trait). This concept was almost universally accepted at the time (although interestingly not by Jackson). According to this theory, a range of conditions including epilepsy were inherited together, and linked to this was the view that as the ‘trait’ was passed from generation to generation, and that the inherited tendency became more severe (the theory of ‘degeneration’). Gowers, for instance, in 1888, wrote that 40% of his 2400 cases showed evidence of the neuropathic trait [5]. This concept has been effectively reinvented today with the recognition of neuropsychiatric ‘comorbidities’ and their bidirectional nature. In the nineteenth century, the conditions were not seen as ‘comorbidities’ but as manifestations of the same underlying causes. It seems likely that this is in fact the case, with perhaps similar defects in biochemical or developmental pathways resulting in a variety of different disorders. This too reflects an orientation towards causal mechanism (proximate) rather than downstream (remote) causal pathologies.

‘The malady is self-perpetuating; when one attack has occurred, whether as the result of an immediate excitant or not, others follow either without any immediate cause, or after some very trifling disturbance. The search for the causes of epilepsy must thus be chiefly an investigation into the conditions with precede the occurrence of the first fit.’

Of course, other concepts of cause in this early period have not stood the test of time — including concepts of reflex causation, autointoxication, constipation, masturbation, and so on. 1.2. 1910–1960 This was a period when much work focused for the first time on the symptomatic causes of epilepsy. 1.2.1. The dependence of cause on investigatory methods At the start of this period, there was far more interest in the heredity of epilepsy than in the symptomatic causes, but this was to change as the century advanced. This was partly due to the catastrophic consequences of hereditarian theories which lead to eugenics and then to the sterilization and then mass extermination of people with mental handicap and epilepsy, which were based on and justified by contemporary hereditarian science and medicine. A second reason of course for the refocusing of interest towards symptomatic epilepsy was the advent of new investigatory modalities such as neuroimaging, electroencephalography, and advances in neuropathology, which all helped uncover cerebral structural defects associated with epilepsy. Advances in neuroimaging included the application of X-ray (discovered in 1898 but utilized in neurology widely only after 1910) and then ventriculography (1918), and cerebral angiography (1927). These techniques were strongly developed in the early postwar years and especially for the detection of vascular and tumoral lesions. One of the first to write in detail about the impact of the early changes was Walter Dandy, who wrote in 1932 [7]: ‘Epilepsy is always regarded as an idiopathic disease. The theories of its causation are indeed so numerous as to reflect seriously upon any exclusive stand concerning its etiology or pathology. However, the writer is confident that there is now assembled from experimental, pathologic, clinical and surgical studies a sufficient number of unquestioned facts to place epilepsy unequivocally upon a pathologic instead of idiopathic basis …. the fundamental conception that in every case of epilepsy there is a lesion of the brain can no longer admit of doubt ….’ Dandy recognized 17 categories of brain lesions causing epilepsy (Table 1), a list which seems peculiar today but which, nevertheless, put symptomatic epilepsy back into focus. In the early postwar years, the major methodological advance in the field of epilepsy was of course the EEG (introduced into clinical practice in 1940). In terms of etiology, EEG was used to detect structural abnormalities, such as brain tumor or infection, a role now completely superseded by modern neuroimaging. It was EEG too which led to the recognition of the importance of hippocampal sclerosis as a prominent cause of epilepsy and the underlying pathology in many cases of

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8 Table 1 The seventeen categories of brain lesions causing symptomatic epilepsy listed by Dandy [7]. Congenital malformation and maldevelopment, either general or focal Tumors Abscesses Tubercles Gummata Aneurysms Syphilis with or without demonstrable gummata or vascular occlusions Areas of cerebral degeneration and calcification Depressed fractures Hamartomata Foreign bodies Injuries from trauma at birth or subsequently (focal or general) Connective tissue formation after trauma Atrophy of the brain after trauma Thrombosis and embolism Cerebral arteriosclerosis Sequelae of obscure inflammatory processes including encephalitis

3

temporal lobe epilepsy (1956; see Shorvon [8]). Advances in clinical chemistry and the understanding of metabolic disease in the prewar and postwar years also identified some of the inherited metabolic causes of epilepsy, a paradigm being the identification of phenylketonuria (PKU) by Ivar Asbjørn Følling in 1934 (see Christ [9]). 1.2.2. The multifactorial nature of causation of epilepsy The most important writer on the causes of epilepsy in this period was WG Lennox. Although Lennox was primarily interested in idiopathic epilepsy, his greatest contribution, in my view, to the theory of causation was to rediscover and update the nineteenth century concept of the multifactorial nature of etiology [10]. He recognized that there was, in most cases, a combination of genetic acquired and precipitating causes. His famous analogies of the ‘river’ and ‘reservoir’ are shown in Fig. 1. Despite this fact, he tended to divide epilepsy into genetic, acquired, and sympathetic categories (in the latter category, following Reynolds), based on what he considered their predominant cause. His acquired

Fig. 1. a. The multifactorial concept of epilepsy — the analogy of the river of causal factors from Lennox and Lennox [10]. “The genetic watershed is represented as three generations: parents, grandparents, and great grandparents [e.g., at A, a paternal grandmother has epilepsy]. A confluence of transmitted traits follows into (and through) the patient …. In addition to these branching streams, there is an independent stream which rises in a lake (the uterus). The outlet is the birth canal and below that are contributing streams: infections [e.g., at B, a viral encephalitis], brain trauma from diverse sources, brain tumor, and circulatory disorders. This side stream enters the main stream at the patient level and combines with the genetic influences which had travelled through three generations to make him have epilepsy. There is then a third stream which enters below the confluence of the two main streams. This represents transient conditions which may precipitate certain seizures in a person already having epilepsy or “all set” to be. This evoking circumstance may be physiologic (say at C, hypoglycemia) or emotional (say at D, a broken wedding engagement)”. b. The epileptic threshold — the analogy of the reservoir from Lennox and Lennox. “Causes may be represented as the sources of a reservoir. At the bottom is the already present volume of water, which represents the person's predisposition, a fundamental cause. However, the reservoir is supplied also by streams which represent the contributory conditions, such as lesions of the brain acquired since conception, certain disorders of bodily function, and emotional disturbances. Periodic overflow of the bank represents a seizure”.

4

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8

epilepsies were subdivided into four categories: development, intrauterine misfortunes, paranatal epilepsy, and postnatal epilepsy. He included developmental defects but realized that these often had strong genetic links (“it would be the channel that joins the genetic river system to the uterine lake”). Lennox used the term hereditary organic epilepsy to describe inherited genetic conditions which result in structural disorders of the brain (he cites mongolism, tuberous sclerosis, and ‘various malformations of development’). 1.3. 1960–present It is of course difficult to take a historical perspective on a period that is so close, but two aspects are of obvious importance. 1.3.1. The new structural neuroimaging modalities The major clinical advances in relation to symptomatic epilepsy have been the extraordinary developments in CT and structural MRI, which have revealed in great detail the structural cerebral abnormalities underlying epilepsy. These have transformed neurology. The most interesting advances in the field of epilepsy have been the detection of infection, cortical dysplasia and developmental anomalies, hippocampal sclerosis, cerebral tumors and degenerative changes. 1.3.2. The impact of molecular science and contemporary clinical genetics Neurochemistry and molecular genetics have revealed the metabolic basis of many inherited metabolic disorders. The remarkable discoveries in clinical genetics, since the recognition of DNA and its method of replication, and the publication of the human genome have allowed great advances to be made in the understanding of the underlying genetic causes of symptomatic epilepsy. These have led to elucidation of the single gene or Mendelian disorders (at least 250) which have epilepsy as part of their phenotype. Where genetics has failed to make much impact, however, is in the identification of cause in ‘idiopathic epilepsies’, and possible reasons for this are discussed below. The most remarkable finding in relation to the causes of idiopathic epilepsy is that where single genes have been found to underpin rare cases of pure epilepsy, these genes have often coded for ion channels, and, thus, these epilepsies are examples of channelopathies. The molecular mechanisms of epilepsy (Hughlings Jackson's proximate causes) have been the focus, since the 1960s, of a huge amount of experimental research. It is not possible to review this here, but much has been learnt on neurotransmitter, mitochondrial, receptor, channel function, and the physiological mechanisms and molecular genetics and pharmacology of epileptic seizures. One feels that it is here where paradigm shifts will be made, but it must be admitted that, currently, knowledge is incomplete. Where there is least knowledge is probably in understanding how the known structural diseases (i.e., remote causes) result in the molecular changes of epilepsy (i.e., proximate causes). 2. Symptomatic epilepsy and its causes SA Kinnier Wilson [11], in his magisterial textbook of neurology published in 1940, devoted 75 pages to epilepsy and wrote the following:

“Current opinion is veering round to the view that all epilepsies are symptomatic, inclusive of the variety [idiopathic epilepsy] whose basis still elude search … the cause will eventually be revealed.” Since then, the view that epilepsy is essentially always a ‘symptom’ of an underlying condition has been widely accepted, even if the underlying cause is hidden. Wilson also makes the first reference I can find to the term “cryptogenic” which he feels is preferable to refer to epilepsies of unknown cause. In the same book, it is stated in a memorable phrase that listing all the known causes of epilepsy “would be an act of supererogation”. However, advances in the study of symptomatic epilepsy have now reached a stage where a synoptical list of all (or almost all) potential causes can be made, and this was attempted in the recently published textbook, The Causes of Epilepsy [12], and in Epilepsia [13]. In these, an etiological classification of the causes of epilepsy was proposed which divided the epilepsies into the following: idiopathic, symptomatic, and provoked categories, listing all causes as subcategories, and also cryptogenic epilepsy. Definition (and its twin sister, classification) is always a difficult topic. In the evolution of thought since the above publications, modified definitions have been proposed and are shown in Table 2. The listing of causes within the category of symptomatic epilepsy is divided into genetic and acquired etiologies and further subdivided into 111 subcategories (Table 3). It must be admitted that this scheme has inconsistencies and gray areas, and the distinction between idiopathic, symptomatic, and provoked epilepsies is logically artificial. If one accepts that most cases have multifactorial causation, then listing under three separate categories makes no sense. However, it is necessary to meet the exigencies of clinical practice (for instance, for investigation) and to provide a shorthand and structure for clinical practice. This distinction between a theoretically rational and a practical clinical classification was well brought out by Hughlings Jackson in the well-known distinction he made between the classifications made by gardeners and botanists [4]. Also, idiopathic epilepsy may have subtle anatomical, synaptic, membrane, neurotransmitter, or network abnormalities, and many patients with provoked epilepsy have an increased susceptibility to epilepsy (either genetic or acquired). The distinction between idiopathic, provoked, and symptomatic epilepsies is, therefore, essentially one of degree. The other inherent conceptual and practical obstacles to providing a rational etiological classification of epilepsy are discussed below, and the scheme is presented in Table 3.

3. The causes of idiopathic epilepsy — not all genetic One very large mistake is to consider ‘idiopathic’ epilepsy to be entirely genetic in origin. This is not to say that genetic influences are not important, and of course every human behavior or thought is ‘genetic’ in the sense that they arise from an internal milieu. Furthermore, many of the idiopathic epilepsies probably do have very strong genetic influences. However, no gene having a major influence on the vast majority of cases of idiopathic epilepsy has yet been identified. It seems very likely that the genetic influences in idiopathic epilepsies probably are complex involving multiple genes and interactions between genes (epistatic) and between genes and the environment (epigenetic). The

Table 2 Definitions of etiological categories of epilepsy. Modified from Shorvon [13] Idiopathic epilepsy Symptomatic epilepsy Provoked epilepsy Cryptogenic epilepsy

An epilepsy of predominately genetic or presumed genetic origin and in which there is no gross neuroanatomic or neuropathologic abnormalities nor other relevant underlying diseases. An epilepsy predominately due to a gross neuroanatomical or neuropathological abnormality or a relevant systemic disease, which can be acquired or genetic in origin. An epilepsy in which a specific systemic or environmental factor is the predominant precipitant of the seizures. An epilepsy of presumed symptomatic nature in which the cause has not been identified.

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8

5

Table 3 Etiological classification of epilepsy. Modified from Shorvon [13]. Main category

Subcategory

Examples of etiology

Idiopathic epilepsy

Pure epilepsies due to single gene disorders

Benign familial neonatal convulsions, autosomal dominant nocturnal frontal lobe epilepsy, generalized epilepsy with febrile seizures plus, severe myoclonic epilepsy of childhood, benign adult familial myoclonic epilepsy Idiopathic generalized epilepsy (and its subtypes), benign partial epilepsies of childhood

Symptomatic epilepsy of predominately genetic or developmental causation

Symptomatic epilepsy of predominately acquired causation

Provoked epilepsy

Pure epilepsies with complex inheritance Progressive myoclonic epilepsies

Unverricht–Lundborg disease, dentato-rubro-pallido-luysian atrophy, Lafora body disease, mitochondrial cytopathy, sialidosis, neuronal ceroid lipofuscinosis, myoclonus renal failure syndrome Neurocutaneous syndromes Tuberous sclerosis, neurofibromatosis, Sturge–Weber syndrome Other single gene disorders Angelman syndrome, lysosomal disorders, neuroacanthocytosis, organic acidurias and peroxisomal disorders, prophyria, pyridoxine dependent epilepsy, Rett syndrome, urea cycle disorders, Wilson disease, disorders of cobalamin, and folate metabolism Disorders of chromosome Down syndrome, Fragile X syndrome, 4p-syndrome, isodicentric chromosome 15, ring structure chromosome 20 Developmental anomalies of Hemimegalencephaly, focal cortical dysplasia, agyria–pachygyria band spectrum, agenesis of corpus cerebral structure callosum, polymicrogyria, schizencephaly, periventricular nodular heterotopia, microcephaly, arachnoid cyst Hippocampal sclerosis Hippocampal sclerosis Perinatal and infantile causes Neonatal seizures, postneonatal seizures, cerebral palsy, vaccination and immunization Cerebral trauma Open head injury, closed head injury, neurosurgery, epilepsy after epilepsy surgery, nonaccidental head injury in infants Cerebral tumor Glioma, ganglioglioma and hamartoma, DNET, hypothalamic hamartoma, meningioma, secondary tumors Cerebral infection Viral meningitis and encephalitis, bacterial meningitis and abscess, malaria, neurocysticercosis, tuberculosis, HIV Cerebrovascular disorders Cerebral hemorrhage, cerebral infarction, degenerative vascular disease, arteriovenous malformation, cavernous hemangioma Cerebral immunologic disorders Rasmussen encephalitis, SLE and collagen vascular disorders, inflammatory and immunologic disorders Degenerative and other neurologic Alzheimer disease and other dementing disorders, multiple sclerosis and demyelinating disorders, conditions hydrocephalus and porencephaly Provoking factors Fever, menstrual cycle and catamenial epilepsy, sleep–wake cycle, metabolic and endocrine-induced seizures, drug-induced seizures, alcohol- and toxin-induced seizures Reflex epilepsies Photosensitive epilepsies, startle-induced epilepsies, reading epilepsy, auditory-induced epilepsy, eating epilepsy, hot water epilepsy

Cryptogenic epilepsies

expression of genes is influenced greatly by environmental factors (both internal and external). One important yet neglected aspect is the temporal dimension (i.e., the neurodevelopmental aspect). As development proceeds, the expression patterns of most genes change continuously, with different gene–gene and gene–environment interactions. All sorts of factors play a part in the developmental process, and the final phenotype can be influenced by very small perturbations in the process, genetic or not, especially when these occur at those critical times when major diversions in developmental pathways occur (choice points). There might also be a large contribution of ‘chance’ in this process [14]. It is for these reasons that the substitution of the term ‘idiopathic’ with ‘genetic’ as proposed in the recent ILAE classification seems misleading. Furthermore, in many parts of the world, labeling a condition as genetic carries enormous social stigma (and impairing the possibility of marriage and childbearing, for instance), and to do so when the genetic basis is only hypothetical seems entirely retrogressive. It will also cause confusion, especially in non-specialist medical practice and in non-medical settings (such as legal, regulatory and legislative settings). Conversely, it should be realized that there are many clearly identified genetic causes of symptomatic epilepsy (see Table 3). 4. The concept of provoked epilepsy The fact that seizures can be ‘provoked’ has of course been long recognized. The precipitation by light was recorded by Pliny and by Apuleius, and by the XIXth century, cases of epilepsy induced by stress, startle, sensory stimulation, shock, noise, sexual activity and masturbation, eating, reading and music, for instance, were well recorded. The ‘exciting’ causes of the 19th-Century epileptologists were often

provoking factors, and Hughlings Jackson and Gowers amongst others wrote at length about these factors and ways of mitigating their effects. Amongst the provoking factors discussed between 1860 and 1920 were as the following: dentition, menstruation, infections and fever, eye strain, alcohol and toxins, endocrine changes (especially thyroid disease), psychiatric difficulties, sleep disturbance, constipation, as well as infections and deficiency diseases (notably rickets). Some but not all of these have stood the test of time, and some led to extreme therapies including surgical procedures such as colectomy, castration, oophorectomy, and carotid ligation. In a study of 500 patients with drug-resistant epilepsy, Aird [15] concluded that in 17%, seizure-inducing factors made a significant contribution to the occurrence of seizures and that manipulating these factors in these cases could greatly improve seizure control. In my own clinics, a survey questionnaire of 104 adult patients revealed that 28% felt that 100% of their seizures were related to some precipitating factor, and a further 49% felt that 50–99% of their seizures were related to a precipitating factor. Stress, sleep deprivation, and fatigue were the most frequent precipitating factors involved (Ferlisi and Shorvon, submitted for publication). The most extreme examples in the spectrum of provoked epilepsies are the reflex epilepsies, but these are rare, and less reliable environmental triggers are far more common. In the etiological classification of epilepsies published in 2011 [16], provoked epilepsies were considered a separate category, only if there was no gross anatomical or pathological features present. This, however, is inconsistent, and it makes more sense to consider epilepsy to be provoked where seizure precipitants are the most obvious cause of seizures whether or not additional etiologies are present. This is similar to the concept of 'sympathetic epilepsy' as defined by Russell Reynolds and later by Lennox.

6

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8

There is no single mechanism of provocation. In some types of reflex epilepsy, there are genetic factors (and some are even monogenic) — hot water epilepsy, audiogenic seizures, and photosensitivity are examples. There are also obvious nongenetic factors, for instance, startle or touch in lesions in the prefrontal areas (interestingly first reported by Jackson and Gowers). In some models, subcortical structures are involved, as shown in the baboon experiments of Naquet, as well as cortical hyperexcitability. Hormonal aspects and stress also modulate hyperexcitability. There are also developmental aspects, for example, the age-related occurrence of startle- or touch-evoked seizures in patients with prefrontal lesions which often occurs in late childhood and adolescence and then disappears. Sometimes, the provoking factor can be so strong that any ‘inherent propensity’ is minimal (an example could be an ECT-induced or metrazol-induced seizure (see below)), and it is obviously far-fetched to categorize these cases as ‘epilepsy’, but this is an exceptional situation. 5. The complexities of assigning cause in epilepsy As will be clear from the above, there are complex issues to consider when attempting to assign cause in epilepsy, which can be summarized as follows (Table 4). 5.1. The multifactorial nature of cause in epilepsy Epilepsy is in the majority of cases multifactorial, and the result of genetic and acquired influences and provoking factors. Assignment to any single etiology is, therefore, to an extent arbitrary. It is best to consider etiologies as ‘causal factors’ rather than ‘causes’. This is an important issue in relation to the concept of symptomatic epilepsy where the relative importance of any ‘cause’ can be crucial clinically. In this situation, it would be logical to use risk factor methodologies (for instance, odd ratios) and derive ‘susceptibilities’ to provide a firm statistical basis of the strength of the ‘causal factor’. The estimation of the odds ratio of any particular causal factors gives a weighting of its importance. Thus, the development of epilepsy after open head injury has a high odds ratio, and it can be considered statistically likely that that ‘cause’ contributes a great deal to the development of epilepsy; whereas the development of epilepsy after a mild head injury, with a low odds ratio, can be considered not likely to contribute a major susceptibility. The problem with this approach is that it is on a population level, and for any individual patient, the weighting may depend on many other individual factors, and a ‘minor head injury’ in some patients may have greater significance than the population odds ratio might suggest. There is often no way of being sure about this in any individual case. Such an approach would also assist in dealing with ‘provoked’ epilepsy in the extreme circumstance, for instance, of an ECT-induced or metrazol-induced seizure where there may be no obvious ‘epileptic propensity’. 5.2. Remote vs. proximate causation A powerful way of categorizing or classifying causation would be to do so according to the mechanisms by which a downstream (remote) pathology, such as a tumor or stroke, will result in a seizure (i.e., by Table 4 Some of the difficulties of assigning cause in epilepsy. The multifactorial nature of cause in epilepsy Remote .v. proximate causation (the distinctions between cause or causal mechanism and the ‘neurological trait’) The causes of idiopathic epilepsy are not all genetic The role of investigation in defining the range of causes Symptomatic epilepsy is not the same as acquired epilepsy The nosological position of provoked epilepsy Epilepsy as a process: differences between new-onset and established epilepsy

‘causal mechanism’ of a seizure; this was the basis of Hughling Jackson's concept of cause), but knowledge is not sufficiently advanced to attempt this yet. Nevertheless, it is likely that most future advances in the field of causation of epilepsy will be in the understanding of the molecular types of epilepsies (the ‘proximate’ causes) rather than of downstream (remote) causes. This would lead to a radically different approach to classification and is the sort of paradigm shift which epilepsy classification is sorely in need of. The recent discovery of some possible mechanisms underlying the bidirectional nature of comorbidities could also contribute to elucidating the proximate causes, and this exciting prospect might also rekindle interest in the concept of the neurological trait. 5.3. The causes of idiopathic epilepsy are not all genetic As noted above, idiopathic epilepsies are often considered to be ‘genetic’, even if a genetic cause is also not demonstrable. In this situation, again, it would seem best to treat any genetic factor exactly as one would a causal factor in ‘symptomatic epilepsy’ and assign an odds ratio. The problem is that in the vast majority of cases, no genetic factor has been identified, and it is, thus, impossible to determine any ‘susceptibility’. In this situation, the term ‘idiopathic’ is a useful one, and seems to me misleading to change the term to ‘genetic’, as has been recently proposed [16]. The fact that many symptomatic epilepsies have strong genetic influences is another reason not to equate ‘idiopathic’ with ‘genetic’ and to avoid the confusion this will cause. 5.4. The role of investigation in defining the range of causes The identification of cause in any individual of course depends on how thoroughly investigations have been carried out. As pointed out above, the range of ‘causes’ identified in clinical practice alters when new investigatory modalities become available, as happened, for instance, with EEG, neuroimaging, clinical chemistry, and molecular histochemistry. It is probably true to say now that all or almost all of the causal conditions of epilepsy which cause gross structural change, and all monogenic illnesses, have been identified, but the mechanisms of genetic influences and the proximate molecular mechanisms remain to be discovered. A new molecular or clinical genetic investigatory modality may have a big impact here. 5.5. Symptomatic epilepsy is not the same as acquired epilepsy Another point which is important to make is that not all symptomatic epilepsy is ‘acquired’ (or environmental). There are many congenital and innate causes of epilepsy which are developmental or genetic in origin and yet which belong in the ‘symptomatic’ category. These include, for instance, cortical dysplasias, neurocutaneous syndromes (e.g., tuberose sclerosis), monogenic diseases (e.g., Rett syndrome and Angelman syndrome), chromosomal disorders (e.g., ring chromosome 20 syndrome), and progressive myoclonic epilepsies (e.g., mitochondrial disease or the neuronal lipofuscinoses). 5.6. The nosological position of provoked epilepsy The place of provoking factors has been often ignored in classification of cause. In the catalog of causes listed above, a separate category of provoked epilepsy is included. This was done to emphasize that genetic, symptomatic, and provoking factors can all play a part in an individual case. If, for instance, a person with idiopathic generalized epilepsy has awakening seizures only after lack of sleep and not at other times, it is logical to consider the presence of the provoking factor as at least as important in terms of susceptibility as the putative genetic basis. Reflex epilepsies are less controversially included in this category of provoked seizures, although whether

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8

7

these should be considered as a separate discrete category or simply as part of the spectrum of provoked epilepsy is arguable.

does the ILAE no service, and I hope that these points will be further considered.

5.7. Epilepsy as a process: differences between new-onset and established epilepsies

7. Acute symptomatic seizures

There is considerable evidence of molecular changes occurring after the onset of epilepsy, which may in themselves contribute to the evolution of a new-onset epilepsy into an established or chronic epilepsy, as first suggested by Gowers in 1881 [6]. In such cases, the internal process can be considered at least in part as the ‘cause’ of the established epilepsy. The nature of these processes though at present is not clearly understood. In symptomatic epilepsies due to acute lesions, it is important to make the distinction between ‘early’ and ‘late’ seizures which are physiologically and clinically different (this is discussed below), and there is often too a ‘latent’ period which can extend for months or even years (after head injury, for instance) between the acute insult and the onset of late seizures. Clearly, there is a process of epileptogenesis which takes a significant time to develop. The physiological basis of this process is not known, and amongst current suggested mechanisms are changes in neuronal networks, synaptic formation, glial function, biochemical or cellular function, inflammation, and neogenesis. These may continue after the onset of epilepsy and contribute to the ‘cause’ of the symptoms. 6. Proposed new ILAE classification of epilepsy and changes of terminology In relation to etiology, the recent proposal of the ILAE [16] to provide a database of causes has been very welcome (and indeed Table 3 is exactly such a database). However, the new classification provides no conceptual change in relation to etiology, and this is a lost opportunity. Furthermore, there are a number of unresolved and contentious issues which reduce its value. There is not sufficient regard to the issues of multifactorial causation, to the risk factor (susceptibility) approach, or to the distinctions between proximate causation and remote causation. The importance of provoking factors (exciting causes) is ignored. There is no recognition of the distinction between causes in new-onset and established epilepsies with regard to the processes producing established epilepsy. The changing of terminology is another controversial aspect of the new classification as it relates to etiology [17]. Every time a classification system is changed, it causes confusion, not only amongst specialists but also particularly amongst those who are not specialists in the condition, and sometimes disastrously. It can take years for a new scheme to be fully adopted, and there can be nosological chaos during this time. This has happened before in epilepsy (the introduction of the classification in 1980) and widely in other branches of medicine. In my view, changes in terminology should be envisaged only if they are underpinned by a major change in conceptual basis, thus justifying the change. It is stated that the proposed changes are justified by the fact that the new terms ‘say it as it is’. This seems antihistorical and superficial and is also misleading. To replace the term ‘idiopathic’ with ‘genetic’ ignores the fact that the actual genetic basis is obscure in the vast majority of patients and that there are important environmental (epigenetic) and gene–gene (epistatic) and neurodevelopmental influences, the effect of chance, and also the fact that some symptomatic epilepsies are in fact genetic. It will also be unnecessarily stigmatizing in some cultures. To replace the term ‘symptomatic’ with ‘structural/metabolic’ excludes the many causes which are neither structural nor metabolic (for instance, those with an immunological, inflammatory, degenerative, toxic, or biochemical basis). To replace the term ‘cryptogenic’ with ‘unknown’ fails to recognize that ‘hidden’ is not the same as ‘unknown’ and looses a venerable time-honored term. This approach

Finally, the use of the term ‘acute symptomatic seizures’ is worthy of consideration. This term was first widely applied in 1970 in the landmark epidemiological work in Rochester, Minnesota [18]. It was devised to differentiate those seizures that occurred at the time of an acute insult, from seizures in ‘idiopathic’ epilepsy (defined as unprovoked seizures in which ‘no cause was found’ and not in its current usage as indicating a genetic cause), and ‘remote symptomatic’ seizures (the use of the term ‘remote’, meaning remote in time from the causative insult, in contrast to the Jacksonian usage of remote in location). The underlying premise is that there is something fundamentally different about these seizures and those of ‘genuine’ epilepsy, and it is indeed quite reasonable to differentiate the early seizures after a head injury from the later chronic posttraumatic seizures. The physiology of the two is also quite different, and early seizures can be caused by contusions, hemorrhage, metabolic change, endocrine change, hypotension, etc. — all mechanisms that have nothing in common with the late seizures of posttraumatic epilepsy. Similar considerations apply to the ‘early seizures’ in stroke and infection, for instance. The term then fell from fashion, but, recently, the Epidemiology Commission of the International League Against Epilepsy convened a subgroup to reconsider the definition of acute symptomatic seizures for epidemiological studies. This group has defined an acute symptomatic seizure as “a clinical seizure occurring at the time of a systemic insult or in close temporal association with a documented brain insult” [19]. The usage of the term though is problematic for two main reasons. The first problem with the term arises because within this category, two quite distinct types of seizures are lumped together: the ‘early seizures’ in acute destructive brain insults and also the seizures which result from reversible provoking factors. In the opinion of the author, it makes no sense to include both categories, which so utterly differ clinically, prognostically, physiologically, and pathologically. The early seizures, with associated brain damage and significant risk of recurrence, have no similarities to the seizures provoked by a reversible environmental trigger such as metabolic change or fever in which there is no acute brain damage and no risk of recurrence. Furthermore, in patients with epilepsy, the seizures precipitated by such factors have no known physiological differences from the patient's habitual seizures. In effect, these seizures are surely simply ‘provoked seizures’ and reflect the interplay of the provoking factor and the individual ‘seizure threshold’. The second problem with the current usage is the arbitrary nature of the criteria for inclusion and indeed of the list of included conditions. In the ILAE scheme, the period for categorizing seizures as ‘acute symptomatic’ is within one week of trauma and stroke but longer (without a time specified) for a subdural hematoma or infections. Some infections are included but not all. Alcohol withdrawal seizures are included but not alcohol-induced seizures. Seizures due to environmental triggers such as visual stimulation are not included, but those induced by hypoglycemia are. In the metabolic conditions, arbitrary cut-off levels are listed despite the fact that there is a marked individual variation in susceptibility and that the rate of change of metabolic factors is as important as the extent of change. In the original definition [18] (but not in the recent ILAE proposal [19]), the term even included initial seizures due to progressive lesions such as tumors which led to the diagnosis, which made no sense at all. In the author’s opinion, the term acute symptomatic seizure should be abandoned, or if it is to be retained, its use should be restricted to the physiologically distinct ‘early seizures’ after acute brain damage, which mechanistically are different both from any subsequent consequential chronic epilepsy and also from the seizures due to reversible provoking

8

S. Shorvon / Epilepsy & Behavior 32 (2014) 1–8

factors [20]. The term acute provoked seizure can be used for the latter category of seizures, as has been anyway in use for over 100 years. 8. Points to consider in future work It can be seen from the above that the concept of ‘symptomatic’ epilepsy is complex. In future work, the following points might be considered: 1. In considering cause in epilepsy, it would be better to refer to ‘causal factors’ and to use risk factor methodologies (odd ratios; i.e., susceptibilities) to provide a firm statistical basis of the strength of the ‘causal factor’. This might obviate the need to divide into categories (idiopathic/symptomatic, etc.). However, for the time being, although artificial, such categorization is necessary clinically to bring order (see, for instance, Hughlings Jackson's distinction between the botanist's and the gardener's type of classification). 2. A greater focus on the proximate causes (molecular causal factors) than the remote causes (downstream pathologies) might lead to a more rational classification than our current rather empirical schemes. This would be the sort of paradigm shift which would justify the adoption of new classification schemes. 3. A binary division of causes into idiopathic/symptomatic epilepsies fails to recognize the importance of ‘provoking’ factors (the ‘exciting’ factors of the nineteenth century). These factors are often as important a cause as any symptomatic or idiopathic causal factor. Although most epilepsies have contributions from genetic, symptomatic, and provoking causes, to have a separate category of ‘provoked’ epilepsy at least emphasizes the importance of the latter category. 4. It is now possible to list the remote causal factors in symptomatic epilepsy. It is likely that all or almost all of the structural and monogenic metabolic disorders have been recognized. However, it is also likely that when there are new investigatory modalities, especially in relation to molecular mechanisms, new ‘causes’ will be recognized. 5. Idiopathic epilepsy is not simply ‘genetic’. There are influences from neurodevelopment (the influence of the dimension of time) and chance and epistatic and epigenetic influences. These mechanisms require further investigation. 6. The term acute symptomatic seizure in its current form should be either abandoned or redefined. It should not include both the early seizures in acute brain insults (head injury, stroke, etc.) and seizures due to reversible provoking factors (fever, metabolic disturbance, drugs, toxins, etc.). If it is retained, a more consistent approach to defining criteria is needed. 7. In relation to etiology, the new ILAE classification scheme has not fully accounted for many of the complexities in defining cause. The renaming of idiopathic as genetic, symptomatic as structural/metabolic, and cryptogenic as unknown should be reconsidered.

Acknowledgment This paper is based on the Distinguished Epileptologist Lecture given at the 6th Cleveland International Epilepsy Symposium in May 2013. The invitation to give the lecture was from Dr. Samden Lhatoo, and I gratefully acknowledge his support and assistance. The ideas for this lecture and paper are partly reproduced from the author's contributions to the book, The Causes of Epilepsy [12], and other papers [11,21]. Disclosure I confirm that I have no conflict of interest to declare in relation to this paper. I confirm that I have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with these guidelines. References [1] Sieveking E. Epilepsy and epileptiform seizures: their causes, pathology and treatment. London: John Churchill; 1858. [2] Spratling W. Epilepsy and its treatment. London, Philadelphia, New York: WB Saunders & Co; 1904. [3] Reynolds JR. Epilepsy: its symptoms, treatment and relation to other chronic convulsive diseases. London: Churchill; 1861. [4] Taylor J. Selected writings of John Hughlings Jackson, vol. 1. London: Hodder and Stoughton; 1930 162–72. [5] Gowers W. A manual of diseases of the nervous system. London: Churchill; 1888. [6] Gowers W. Epilepsy and other chronic convulsive disorders. London: Churchill; 1881. [7] Dandy WE. The practice of surgery. The brain. In: Lewis D, editor. The practice of surgery, vol. XII. Connecticut: Prior WF; 1932. [8] Shorvon SD. An episode in the history of temporal lobe epilepsy: the quadrennial meeting of the ILAE in 1953. Epilepsia 2006;47:1288–91. [9] Christ SE. Asbjørn Følling and the discovery of phenylketonuria. J Hist Neurosci 2003;12:44–54. [10] Lennox WG, Lennox M. Epilepsy and related disorders. Boston: Little Brown; 1960. [11] Wilson SAK. Neurology. London: Edward Arnold; 1940. [12] Shorvon SD, Andermann F, Guerrini R, editors. The causes of epilepsy: common and uncommon causes in adults and children. Cambridge: Cambridge University Press; 2011. [13] Shorvon SD. The etiologic classification of epilepsy. Epilepsia 2011;52(6):1052–7. [14] Johnson MR, Shorvon SD. Heredity in epilepsy: neurodevelopment, comorbidity, and the neurological trait. Epilepsy Behav 2011;22(3):421–7. [15] Aird RB, Gordon NS. Some excitatory and inhibitory factors involved in the epileptic state. Brain Dev 1993;15:299–304. [16] Berg AT, Berkovic SF, Brodie MJ, Buchhalter J, Cross JH, van Emde Boas W, et al. Revised terminology and concepts for organization of seizures and epilepsies: report of the ILAE Commission on Classification and Terminology, 2005–2009. Epilepsia 2010;51:676–85. [17] Shorvon SD. New terminologies: the downsides. Epilepsia 2013;54(6):1134. [18] Annegers JF, Hauser WA, Lee J, Rocca WA. Incidence of acute symptomatic seizures in Rochester Minnesota, 1935–1984. Epilepsia 1995;36:327–33. [19] Hauser W, Beghi E, Carpi A, Fosgren L, Hesdorffer D, Malmgren K, et al. Recommendations for a definition of acute symptomatic seizure. Epilepsia 2010;51:671–5. [20] Shorvon S, Guerrini R. Acute symptomatic seizures—should we retain the term? Epilepsia 2010;51(4):722–3. [21] Shorvon SD. The causes of epilepsy: changing concepts of etiology of epilepsy over the past 150 years. Epilepsia 2011;52(6):1033–44.