ILAE Classification of the Epilepsies Position Paper PDF

ILAE POSITION PAPER ILAE classification of the epilepsies: Position paper of the ILAE Commission for Classification and Terminology 1,2,3 Ingrid E. Scheffer, 1Samuel Berkovic, 4Giuseppe Capovilla, 5Mary B. Connolly, 6 Jacqueline...

ILAE POSITION PAPER ILAE classification of the epilepsies: Position paper of the ILAE Commission for Classification and Terminology 1,2,3 Ingrid E. Scheffer, 1Samuel Berkovic, 4Giuseppe Capovilla, 5Mary B. Connolly, 6 Jacqueline French, 7Laura Guilhoto, 8,9Edouard Hirsch, 10Satish Jain, 11Gary W. Mathern, 12 Solomon L. Mosh
e, 13Douglas R. Nordli, 14Emilio Perucca, 15Torbj€ orn Tomson, 16 Samuel Wiebe, 17Yue-Hua Zhang, and 18,19Sameer M. Zuberi Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709 SUMMARY The International League Against Epilepsy (ILAE) Classification of the Epilep- sies has been updated to reflect our gain in understanding of the epilepsies and their underlying mechanisms following the major scientific advances that have taken place since the last ratified classification in 1989. As a critical tool for the practicing clinician, epilepsy classification must be relevant and dynamic to changes in thinking, yet robust and translatable to all areas of the globe. Its primary purpose is for diagnosis of patients, but it is also critical for epilepsy research, development of antiepileptic therapies, and communication around the world. The new classification originates from a draft document submitted for public comments in 2013, which was revised to incorporate extensive feedback from the international epilepsy community over several rounds of consultation. It presents three levels, starting with seizure type, where it assumes that the patient is having epileptic seizures as defined by Dr. Ingrid E. Scheffer the new 2017 ILAE Seizure Classification. After diagnosis of the seizure type, chairs the ILAE Task the next step is diagnosis of epilepsy type, including focal epilepsy, generalized Force on the epilepsy, combined generalized, and focal epilepsy, and also an unknown epi- Classification of the lepsy group. The third level is that of epilepsy syndrome, where a specific syn- Epilepsies. dromic diagnosis can be made. The new classification incorporates etiology along each stage, emphasizing the need to consider etiology at each step of diagnosis, as it often carries significant treatment implications. Etiology is bro- ken into six subgroups, selected because of their potential therapeutic conse- quences. New terminology is introduced such as developmental and epileptic encephalopathy. The term benign is replaced by the terms self-limited and Accepted January 21, 2017; Early View publication March 8, 2017. 1 Department of Medicine, The University of Melbourne, Austin Health, Melbourne, Victoria, Australia; 2Department of Paediatrics, Royal Children’s Hospital, The University of Melbourne, Melbourne, Victoria, Australia; 3Florey Institute, Melbourne, Victoria, Australia; 4Child Neuropsychiatry Department, Epilepsy Center, C. Poma Hospital, Mantova, Italy; 5Department of Pediatrics, British Columbia’s Children’s Hospital, University of British Columbia, Vancouver, British Columbia, Canada; 6Department of Neurology, NYU School of Medicine, New York, New York, U.S.A.; 7Department of Neurology, Federal University of S~ao Paulo; University Hospital, University of S~ao Paulo, S~ao Paulo, Brazil; 8University Hospital INSERM U 964,
Lyon, France; 10Indian Epilepsy Centre, New Delhi, India; 11Departments of Neurosurgery, Psychiatry and Biobehavioral Strasbourg, France; 9IDEE, Medicine, David Geffen School of Medicine, UCLA, Los Angeles, California, U.S.A.; 12Saul R. Korey Department of Neurology, Dominick P. Purpura Department of Neuroscience and Department of Pediatrics, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, New York, U.S.A.; 13Division of Neurology, Children’s Hospital Los Angeles, Los Angeles, California, U.S.A.; 14C. Mondino National Neurological Institute and Clinical Pharmacology Unit, University of Pavia, Pavia, Italy; 15Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden; 16Departments of Clinical Neurosciences and Community Health Sciences, University of Calgary, Calgary, Alberta, Canada; 17Department of Pediatrics, Peking University First Hospital, Beijing, China; 18Paediatric Neurosciences Research Group, Fraser of Allander Neurosciences Unit, Royal Hospital for Children, Glasgow, United Kingdom; and 19School of Medicine, University of Glasgow, Glasgow, United Kingdom Address correspondence to Ingrid E. Scheffer, Epilepsy Research Centre, 245 Burgundy St, Heidelberg, Vic. 3084, Australia. E-mail: [email protected] Wiley Periodicals, Inc. © 2017 International League Against Epilepsy 512  513 Classification of the Epilepsies pharmacoresponsive, to be used where appropriate. It is hoped that this new framework will assist in improving epilepsy care and research in the 21st century. KEY WORDS: Classification, Epilepsy syndromes, Terminology, Etiology. beyond the clinical domain to clinical and basic epilepsy research and to the development of novel therapies. Classifi- cation serves many purposes: providing a framework for Key Points understanding the type of seizures that the patient has, the other seizure types that are more likely to occur in that indi- The ILAE presents a revised framework for the Classi- vidual, the potential triggers for their seizures, and often fication of the Epilepsies, designed to work with the their prognosis. Classification also informs the risks of classification of seizure types comorbidities including learning difficulties, intellectual Levels of diagnosis: seizure type, epilepsy type (focal, disability, psychiatric features such as autism spectrum dis- generalized, combined generalized and focal, order, and mortality risk such as sudden unexpected death in unknown) and epilepsy syndrome epilepsy (SUDEP). It is notable that classification often An etiologic diagnosis should be considered from guides the selection of antiepileptic therapies. when the patient first presents, and at each step along Classification of the epilepsies has evolved dramatically the diagnostic pathway; a patient’s epilepsy may be since its inception in the 1960s.6–9 The many iterations in classified into more than one etiological category classification reflect advances in understanding phenotypic The term “benign” is replaced by the terms self-lim- patterns and underlying mechanisms, based on major contri- ited and pharmacoresponsive to be used where butions from clinical and basic research from around the appropriate world. These insights are incorporated into the many facets The term “developmental and epileptic encephalopa- of clinical care for patients and lead to progress in the devel- thy” can be applied in whole or in part where opment of innovative treatments, be they pharmacologic or appropriate dietary therapies, surgical approaches or device develop- ment. Classification will always be a dynamic process, iter- Ongoing efforts to refine the classification of the epilep- ative to the new insights gained through research and sies have been made by the International League Against improved understanding of this heterogeneous group of dis- Epilepsy (ILAE) almost since its inception in 1909 and eases. Its continued evolution into the future promises to gained special momentum in the early 1960s when new con- lead to further advances in patient care. cepts of classification were proposed by Henri Gastaut.1–3 Classification engenders passionate debate. This is partly Intense debate and acquisition of new knowledge in the next because it is built on the complex clinical constructs under- two decades led to the landmark 1985 ILAE “Classification pinning epilepsy diagnosis and partly because it is so critical of Epilepsies and Epileptic Syndromes,”4 which was soon to our daily practice. Classification has been based on expert followed by a revised version ratified by the ILAE General opinion drawing together epileptologists and related experts Assembly in 1989.5 The 1989 Classification has been highly from around the world. Although there is no doubt that the influential worldwide and has had a major impact on epi- desired endpoint is a scientifically based classification, our lepsy care and research. The work presented herein builds understanding is not sufficiently advanced to construct a on the efforts of many over more than a century; we classification on a scientifically rigorous basis.9 Thus cur- acknowledge their seminal contributions in the develop- rent proposals are based on a combination of the latest sci- ment of the classification of the epilepsies. entific understanding coupled with high-level expert Although many concepts outlined in the 1989 ILAE clas- opinion, including an extensive consultation with epilepsy sification remain valid to this day, it has become increas- professionals and the wider epilepsy community ingly clear that a revision is needed to account for worldwide. subsequent scientific discoveries that over the last few dec- When a patient presents with seizures, the clinician works ades have fundamentally changed our understanding of the through several critical steps in making a diagnosis. Before epilepsies as well as our approach to the diagnosis and man- attempting to classify a seizure, the physician must deter- agement of individuals with epilepsy. mine whether the paroxysmal event is indeed an epileptic Epilepsy classification is the key clinical tool in evaluat- seizure with a myriad of differential diagnoses being possi- ing an individual who is presenting with seizures. It influ- ble. These include convulsive syncope, parasomnias, move- ences every clinical consultation yet its impact stretches far ment disorders, and other nonepileptic events (https:// Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709 514 I. E. Scheffer et al. www.epilepsydiagnosis.org/epilepsy-imitators.html). This online inviting discussion (Supporting Information for Sch- diagnostic step is taken as already established at the point of effer et al.30). Avid community engagement and debate beginning to classify the patient’s epilepsy. occurred, with 128 comments received from 43 countries. In terms of epilepsy classification, the clinician starts by The response was so extensive and the feedback on impor- classifying the type of seizure. This is the subject of the tant concepts so conflicting that the panel in charge of companion paper on the new classification of seizure reviewing the public comments determined that further types.10 Then, the patient’s type of epilepsy needs to be clas- pubic engagement was necessary to ensure the highest pos- sified and, in many cases, a specific epilepsy syndrome diag- sible level of agreement. The roadmap followed by the nosis can be made. Just as importantly, strenuous attempts panel to solicit further engagement and to respond to feed- to identify the etiology of the patient’s epilepsy should be back from stakeholders is described in an article published made at each step in the diagnostic pathway. Classification in Epilepsia Open in 2016, which again invited feedback of seizure type and epilepsy type both take into account the from the global community.30 Further comments and opin- results of investigations such as electroencephalography ions were then deliberated and considered in finalizing the (EEG) and neuroimaging studies together with other studies present position document which defines the Classification exploring the underlying etiology of the epilepsy. Herein, of the Epilepsies in 2017. we present the first major Classification of the Epilepsies since the last ratified ILAE Classification in 1989. Classification of the Epilepsies The new Classification of the Epilepsies is a multilevel Methods classification, designed to cater to classifying epilepsy in In the past, ILAE position papers on fundamental matters different clinical environments (Fig. 1). This is in acknowl- such terminology, definition, and classification of seizures edgement of the wide variation in resources around the and epilepsy required ratification by the General Assembly world, meaning that different levels of classification will be through a vote by the representatives of the ILAE Chapters possible depending on the resources available to the clini- from around the world.5 This approach is no longer optimal, cian making the diagnosis. Where possible, a diagnosis at since it does not permit adequate engagement of the greatly all three levels should be sought as well as the etiology of expanded constituency of epilepsy experts around the world the individual’s epilepsy. and fails to exploit opportunities offered by impressive advances in communication tools. Seizure type Consequently, in 2013, the League set in place a new pro- The starting point of the Epilepsy classification frame- cess for the finalization and approval of position documents, work is the Seizure Type; it assumes that the clinician has that is, documents that reflect the ILAE position on topics already made a definite diagnosis of an epileptic seizure and that involve adoption of a common language or set of defini- is not meant to be a diagnostic algorithm to distinguish tions (e.g., defining epilepsy, classification).11 This process epileptic from nonepileptic events. The Seizure Type Clas- is highly iterative and involves initial production of the doc- sification is determined according to the new nomenclature ument by a group of experts selected by the League, posting in the accompanying paper.10 Seizures are classified into the document on the ILAE website, soliciting comments focal onset, generalized onset, and unknown onset. and criticism by all stakeholders, and appointing a separate In some settings, classification according to Seizure Type expert panel to review and incorporate the public comments. may be the maximum level possible for diagnosis as there This process takes place in parallel with the peer review may be no access to EEG, video. and imaging studies. In conducted by the journal to which the document is submit- other cases, there may simply be too little information avail- ted for publication (http://www.ilae.org/Visitors/Docume able to be able to make a higher level diagnosis, such as nts/Guideline-PublPolicy-2013Aug.pdf). when a patient has only had a single seizure. In the case of the revised Classification, a first proposal that preceded implementation of the procedure outlined Epilepsy type above was published by the ILAE Commission on Classifi- The second level is that of Epilepsy Type and assumes cation and Terminology in 2010.9 The emphasis was on that the patient has a diagnosis of epilepsy based on the employing transparent terminology, where words mean 2014 definition.31 The Epilepsy Type level includes a new what they say. The 2010 publication triggered extensive dis- category of “Combined Generalized and Focal Epilepsy” in cussion and commentaries.12–29 A new Commission on addition to the well-established Generalized Epilepsy Classification and Terminology was subsequently and Focal Epilepsies. It also includes an Unknown cate- appointed by the ILAE Executive and tasked to produce a gory. Many epilepsies will include multiple types of revised Classification through the procedure outlined for seizures. ILAE position documents. The Commission submitted the For a diagnosis of Generalized Epilepsy, the patient initial document in 2013, and the document was posted would typically show generalized spike-wave activity on Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709  515 Classification of the Epilepsies Figure 1. Framework for classification of the epilepsies. *Denotes onset of seizure. Epilepsia ILAE EEG. Individuals with generalized epilepsies may have a activity on EEG who cannot be classified into a known epi- range of seizure types including absence, myoclonic, atonic, lepsy syndrome but has a clear-cut diagnosis of Generalized tonic, and tonic–clonic seizures. The diagnosis of general- Epilepsy; or the less common scenario of a 20-year-old ized epilepsy is made on clinical grounds, supported by the woman with both focal impaired awareness seizures finding of typical interictal EEG discharges. Caution needs and absence seizures with both focal discharges and gener- to be exercised for a patient with generalized tonic–clonic alized spike wave on EEG recordings and normal MRI, who seizures and a normal EEG. In this case, supportive evi- would therefore have a diagnosis of Combined Generalized dence would need to be present to make a diagnosis of gen- and Focal Epilepsy. eralized epilepsy, such as myoclonic jerks or a relevant The term “Unknown” is used to denote where it is family history. understood that the patient has Epilepsy but the clinician Focal Epilepsies include unifocal and multifocal disor- is unable to determine if the Epilepsy Type is focal or gen- ders as well as seizures involving one hemisphere. A range eralized because there is insufficient information avail- of seizure types can be seen including focal aware seizures, able. This may be for a variety of reasons. There may be focal impaired awareness seizures, focal motor seizures, no access to EEG, or the EEG studies may have been focal non-motor seizures, and focal to bilateral tonic–clonic uninformative, for example, normal. If the Seizure Type(s) seizures. The interictal EEG typically shows focal epilepti- are unknown, then the Epilepsy Type may be unknown for form discharges, but the diagnosis is made on clinical similar reasons, although the two may not always be grounds, supported by EEG findings. concordant. For example, the patient may have had sev- The new group of Combined Generalized and Focal eral symmetrical tonic–clonic seizures without focal fea- Epilepsies exists, as there are patients who have both gener- tures and normal EEG recordings. Thus the onset of the alized and focal seizures. The diagnosis is made on clinical seizures is unknown and the person has an unknown grounds, supported by EEG findings. Ictal recordings are epilepsy type. helpful but not essential. The interictal EEG may show both generalized spike-wave and focal epileptiform discharges, Epilepsy syndrome but epileptiform activity is not required for the diagnosis. The third level is an Epilepsy Syndrome diagnosis. An Common examples in which both types of seizures occur epilepsy syndrome refers to a cluster of features incorporat- are Dravet syndrome and Lennox-Gastaut syndrome. ing seizure types, EEG, and imaging features that tend to The Epilepsy type may also be the final level of diagnosis occur together. It often has age-dependent features such as achievable where the clinician is unable to make an Epi- age at onset and remission (where applicable), seizure trig- lepsy Syndrome diagnosis. Examples include the following: gers, diurnal variation, and sometimes prognosis.4,5 It may the common situation of a child or adult with nonlesional also have distinctive comorbidities such as intellectual and temporal lobe epilepsy who has Focal Epilepsy with no psychiatric dysfunction, together with specific findings on known etiology; a 5-year-old child presenting with general- EEG and imaging studies. It may have associated etiologic, ized tonic–clonic seizures and generalized spike-wave prognostic, and treatment implications. It is important to Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709 516 I. E. Scheffer et al. note that an epilepsy syndrome does not have a one-to-one used where the clinician is comfortable with invoking a correlation with an etiologic diagnosis and serves a different genetic etiology. purpose such as guiding management. There are many well- recognized syndromes, such as childhood absence epilepsy, Self-limited focal epilepsies West syndrome, and Dravet syndrome, although it should There are several self-limited focal epilepsies, typically be noted that there has never been a formal classification of beginning in childhood. The most common is self-limited syndromes by the ILAE.9 The recently developed educa- epilepsy with centrotemporal spikes, formerly called “be- tional ILAE website, epilepsydiagnosis.org, provides an nign epilepsy with centrotemporal spikes.” Others included excellent resource to understand the parameters for diagno- in this broad group are the self-limited occipital epilepsies sis, review videos of seizure types and the EEG features of of childhood, with the early-onset form described by many established syndromes, and has been devised as a Panayiotopoulos and the late-onset form by Gastaut.33 teaching tool. Other self-limited frontal lobe,34 temporal35, and parietal lobe epilepsies36 have been described with some beginning Idiopathic Generalized Epilepsies in adolescence and even adult life. Within the Generalized Epilepsies is the well-recognized and common subgroup of the Idiopathic Generalized Etiology Epilepsies (IGEs). The IGEs encompass four well-estab- From the moment that the patient presents with a first lished epilepsy syndromes: Childhood Absence Epilepsy, epileptic seizure, the clinician should be aiming to deter- Juvenile Absence Epilepsy, Juvenile Myoclonic Epilepsy mine the etiology of the patient’s epilepsy. A range of etio- and Generalized Tonic–Clonic Seizures Alone (formerly logic groups has been recognized, with emphasis on those known as Generalized Tonic–Clonic Seizures on Awaken- that have implications for treatment. Often the first investi- ing but modified in recognition that seizures can occur at gation carried out involves neuroimaging, ideally MRI any time of day). The intention to remove the term “idio- where available. This enables the clinician to decide if there pathic” from the nomenclature of Epilepsy Classification is a structural etiology for the patient’s epilepsy. The five was suggested, as its definition was “no known or suspected additional etiologic groups are genetic, infectious, meta- etiology other than possible hereditary predisposition.”4 bolic, and immune, as well as an unknown group (Fig. 1). A The Greek term “idios” refers to self, own, and personal, patient’s epilepsy may be classified into more than one etio- and is thus meant to reflect the genetic etiology without logic category; the etiologies are not hierarchical, and the explicitly saying so. Idiopathic may therefore be regarded importance given to the patient’s etiological group may as an imprecise term given our increasing recognition and depend on the circumstance. For instance, a patient with discovery of the genes involved in many epilepsies, includ- tuberous sclerosis has both a structural and a genetic etiol- ing those with monogenic (with inherited or de novo patho- ogy; the structural etiology is critical for epilepsy surgery, genic variants) or complex (polygenic with or without whereas the genetic etiology is key for genetic counseling environmental factors) inheritance. In addition, the word and consideration of novel therapies such as mammalian “genetic” may sometimes be wrongly interpreted as synony- target of rapamycin (mTOR) inhibitors. mous with “inherited.” It is therefore more meaningful to refer to this group of Structural etiology syndromes as Genetic Generalized Epilepsies (GGEs), The concept behind a structural etiology is that a struc- where the clinician feels there is sufficient evidence for this tural abnormality has a substantially increased risk of being classification. Such evidence is drawn from meticulous clin- associated with epilepsy based on appropriately designed ical research of the inheritance of these syndromes in twin studies.9 A structural etiology refers to abnormalities visible and family studies and does not mean that specific genetic on structural neuroimaging where the electroclinical assess- mutations have been identified. Indeed, it is currently rarely ment together with the imaging findings lead to a reasonable the case that the genetic mutation(s) causing a patient’s epi- inference that the imaging abnormality is the likely cause of lepsy has been determined, perhaps with the exception of the patient’s seizures. Structural etiologies may be acquired the infantile onset developmental and epileptic encephalo- such as stroke, trauma, and infection, or genetic such as pathies, where many patients have been shown to have a de many malformations of cortical development. Despite there novo pathogenic variant.32 being a genetic basis with such malformations, the structural There has been, however, considerable desire to retain correlate underpins the person’s epilepsy. Identification of a the term IGE. The Task Force has therefore decided that the subtle structural lesion requires appropriate MRI studies term IGE will be acceptable specifically for the group of using specific epilepsy protocols.37 four epilepsy syndromes: Childhood Absence Epilepsy, There are well-recognized associations within the epilep- Juvenile Absence Epilepsy, Juvenile Myoclonic Epilepsy, sies with a structural etiology. These include the relatively and Generalized Tonic–Clonic Seizures Alone. In individ- frequent finding of mesial temporal lobe seizures with hip- ual cases, the term Genetic Generalized Epilepsy may be pocampal sclerosis. Other key associations include gelastic Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709  517 Classification of the Epilepsies seizures with hypothalamic hamartoma, Rasmussen syn- spectrum associated with mutations of a specific gene is drome, and hemiconvulsion-hemiplegia-epilepsy. Recogni- critical information, as the finding of a mutation in a speci- tion of these associations is important to ensure that the fic gene may not, on its own, enable prediction of the out- patient’s imaging is carefully examined for a specific struc- come. Interpretation of its significance needs to be tural abnormality. This in turn highlights the need for con- considered in the context of the electroclinical presentation. sideration for epilepsy surgery should the patient fail Thus, to date, the majority of genes show phenotypic hetero- medical therapy. geneity and the majority of syndromes reveal genetic The underlying basis for a structural abnormality may be heterogeneity. genetic or acquired, or both. For example, polymicrogyria Where epilepsy follows complex inheritance, which may be secondary to mutations in genes such as GPR56, or implies multiple genes with/without an environmental con- acquired, secondary to intrauterine cytomegalovirus infec- tribution, susceptibility variants may be identified that con- tion.38 Acquired structural causes include hypoxic-ischemic tribute to causation but are insufficient alone to cause encephalopathy, trauma, infection, and stroke. Where a epilepsy.45,46 In this setting, there may be no family history structural etiology has a well-defined genetic basis such as of seizures because other family members do not have tuberous sclerosis complex, which is caused by mutations in enough epilepsy genetic variants to be affected. the genes TSC1 and TSC2 encoding hamartin and tuberin, It is important to note that genetic does not equate with respectively, both etiologic terms, structural and genetic can inherited. An increasing number of de novo mutations are be used. being identified in both severe and mild epilepsies.47–52 This means that the patient has a new mutation that has Genetic etiology arisen in him or her, and therefore is unlikely to have a fam- The concept of a genetic epilepsy is that it results directly ily history of seizures and has not inherited the genetic from a known or presumed genetic mutation in which sei- mutation. Nevertheless, this patient may now have a herita- zures are a core symptom of the disorder. The epilepsies in ble form of epilepsy. For example if the individual has a de which a genetic etiology has been implicated are quite novo dominant mutation, their offspring will have a 50% diverse and, in most cases, the underlying genes are not yet risk of inheriting the mutation. This does not necessarily known. mean that their children will have epilepsy, as its expression First, the inference of a genetic etiology may be based will depend on the penetrance of the mutation. solely on a family history of an autosomal dominant disor- Drilling down further, patients may be mosaic for a muta- der. For example, in the syndrome of Benign Familial tion. This means they have two populations of cells, with Neonatal Epilepsy, most families have mutations of one of one population having the mutation and the other having the the potassium channel genes, KCNQ2 or KCNQ3.39 Con- wild-type (normal) allele. Mosaicism may affect the severity versely, in the syndrome of Autosomal Dominant Nocturnal of their epilepsy, with lower mosaicism rates resulting in a Frontal Lobe epilepsy, the underlying mutation is known in milder severity of epilepsy, as shown in SCN1A studies.53 only a small proportion of individuals at this time.40 A genetic etiology does not exclude an environmental Second, a genetic etiology may be suggested by clinical contribution. It is well accepted that environmental factors research in populations with the same syndrome such as contribute to seizure disorders; for example, many individu- Childhood Absence Epilepsy or Juvenile Myoclonic Epi- als with epilepsy are more likely to have seizures with sleep lepsy. Evidence for a genetic basis comes from elegant stud- deprivation, stress, and illness. A genetic etiology refers to a ies such as Lennox’s twin studies in the 1950s and familial pathogenic variant (mutation) of significant effect in caus- aggregation studies.41,42 ing the individual’s epilepsy. Third, a molecular basis may have been identified and may implicate a single gene or copy number variant of Infectious etiology major effect. There is an increasing number of patients with The most common etiology worldwide is where epilepsy known genetic abnormalities causing both severe and mild occurs as a result of an infection.54 The concept of an infec- epilepsies. Molecular genetics has led to identification of tious etiology is that it directly results from a known infec- the causative mutation in a large number of epilepsy genes, tion in which seizures are a core symptom of the disorder. most frequently arising de novo, in 30–50% of infants with An infectious etiology refers to a patient with epilepsy, severe developmental and epileptic encephalopathies.32 rather than with seizures occurring in the setting of acute The best known example is Dravet syndrome in which infection such as meningitis or encephalitis. Common >80% of patients have a pathogenic variant of SCN1A. It is examples in specific regions of the world include neurocys- notable that a monogenic etiology may cause a spectrum of ticercosis, tuberculosis, HIV, cerebral malaria, subacute mild to severe epilepsies, such as SCN1A mutations, which sclerosing panencephalitis, cerebral toxoplasmosis, and are associated with Dravet syndrome and Genetic Epilepsy congenital infections such as Zika virus and cytomegalo- with Febrile Seizures Plus (GEFS+), and may have implica- virus. These infections sometimes have a structural corre- tions for treatment.43,44 Understanding the phenotypic late. An infectious etiology carries specific treatment Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709 518 I. E. Scheffer et al. implications. An infectious etiology may also refer to the difficulties to intellectual disability, to psychiatric features postinfectious development of epilepsy, such as viral such as autism spectrum disorders and depression, to psy- encephalitis leading to seizures in the aftermath of the acute chosocial concerns. In the more severe epilepsies, a com- infection. plex range of comorbidities may be seen, including motor deficits such as cerebral palsy or deterioration in gait, move- Metabolic etiology ment disorders, scoliosis, sleep, and gastrointestinal disor- A range of metabolic disorders is associated with epi- ders. Like etiology, it is important that the presence of lepsy. This area is expanding and a greater understanding of comorbidities be considered for every patient with epilepsy the phenotypic spectrum emerging. The concept of a meta- at each stage of classification, enabling early identification, bolic epilepsy is that it results directly from a known or pre- diagnosis, and appropriate management. sumed metabolic disorder in which seizures are a core symptom of the disorder. Metabolic causes refer to a well- delineated metabolic defect with manifestations or bio- New Terminology and chemical changes throughout the body such as porphyria, Definitions uremia, aminoacidopathies, or pyridoxine-dependent sei- Developmental and epileptic encephalopathies zures. In many cases, metabolic disorders will have a The term “epileptic encephalopathy” was redefined in the genetic defect. It is likely that most metabolic epilepsies will Berg et al.9 report as where the epileptic activity itself con- have a genetic basis, but some may be acquired such as cere- tributes to severe cognitive and behavioral impairments bral folate deficiency. The identification of specific meta- above and beyond what might be expected from the under- bolic causes of epilepsy is extremely important due to lying pathology alone (e.g., cortical malformation). Global implications for specific therapies and potential prevention or selective impairments can worsen over time. These of intellectual impairment. impairments can be seen along a spectrum of severity and across all epilepsies, and can occur at any age. Immune etiology The concept of the epileptic encephalopathy may be The concept of an immune epilepsy is that it results applicable to epilepsies at all ages and should be utilized directly from an immune disorder in which seizures are a more widely than just for the severe epilepsies with onset in core symptom of the disorder. A range of immune epilepsies infancy and childhood. Many epilepsy syndromes associ- has been recently recognized with characteristic presenta- ated with encephalopathy have a genetic etiology, such as tions in both adults and children.54 An immune etiology can West syndrome, where there is marked genetic heterogene- be conceptualized as where there is evidence of autoim- ity, and Epileptic encephalopathy with continuous spike- mune-mediated central nervous system inflammation. Diag- and-wave during sleep (CSWS), where the first genes have nosis of these autoimmune encephalitides is rapidly begun to emerge.32 Equally, such syndromes may have an increasing, particularly with greater access to antibody test- acquired cause such as hypoxic-ischemic encephalopathy or ing. Examples include anti-NMDA (N-methyl-D-aspartate) stroke, or may be associated with a malformation of cortical receptor encephalitis and anti-LGI1 encephalitis.55 With the development that may also have a genetic or acquired etiol- emergence of these entities, this etiologic subgroup ogy. deserves a specific category, particularly given the treat- The concept of an epileptic encephalopathy can also be ment implications with targeted immunotherapies. applied to single gene disorders, such as CDKL5 Unknown etiology encephalopathy and CHD2 encephalopathy. However, a Unknown means that the cause of the epilepsy is not yet single gene may cause an epileptic encephalopathy in some known. There remain many patients with epilepsy for whom individuals and a self-limited epilepsy in others; examples the cause is not known. In this category it is not possible to include SCN1A, SCN2A, SLC2A1, KCNQ2, KCNA2, and make a specific diagnosis apart from the basic electroclini- CHD2. In an epileptic encephalopathy, the abundant epilep- cal semiology such as frontal lobe epilepsy. The extent to tiform activity interferes with development resulting in cog- which a cause can be found depends on the extent of the nitive slowing and often regression, and sometimes is evaluation available to the patient. This differs across differ- associated with psychiatric and behavioral consequences. ent health care settings and countries and hopefully will The epileptiform activity can cause regression in an individ- improve over time in resource-poor countries. ual with normal development or preexisting developmental delay, who then shows developmental plateauing or regres- Comorbidities sion. A key component of the concept is that amelioration of There is increasing awareness that many of the epilepsies the epileptiform activity may have the potential to improve are associated with comorbidities such as learning, psycho- the developmental consequences of the disorder. This is a logical, and behavioral problems (Fig. 1, left hand vertical critical issue from a clinical perspective and one often mir- oval). These range in type and severity, from subtle learning rored in the observations of families and clinicians. Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709  519 Classification of the Epilepsies Many of these severe genetic disorders also have devel- In many instances where a genetic mutation of major opmental consequences arising directly from the effect of effect is identified, the terms “developmental and epileptic the genetic mutation, in addition to the effect of the frequent encephalopathy” may be subsumed by using the name of the epileptic activity on development. There are several ways in underlying condition. For example, many of the well-recog- which this may manifest. There may be preexisting develop- nized developmental and epileptic encephalopathies can mental delay, complicated by plateauing or regression with now be called by their gene name together with the word seizure onset or with prolonged seizures. In other disorders, encephalopathy, such as “STXBP1 encephalopathy” or developmental slowing may occur on a background of nor- “KCNQ2 encephalopathy.” This is particularly important mal development, with the slowing emerging prior to the when referring to a genetic disease where genes are associ- presence of frequent epileptic activity on EEG. A well- ated with both severe and self-limited, pharmacoresponsive known example is the relatively common encephalopathy epilepsies, such as KCNQ2 or SCN2A. Then the term of Dravet syndrome, in which developmental slowing or “encephalopathy” can be used to denote the severe form of regression occurs between 1 and 2 years of age, at a time the disease associated with developmental impairment. when epileptiform activity on EEG is typically not yet fre- quent. This suggests a developmental component in addi- Self-limited and pharmacoresponsive tion to an epileptic component, with both occurring With increasing recognition of the impact of these comor- secondary to the underlying sodium channel subunit gene bidities on an individual’s life, there has been considerable (SCN1A) mutation found in >80% of cases. In a third group, concern that the term “benign” underestimates this burden, the epilepsy may settle down relatively early in the child’s particularly in the milder epilepsy syndromes such as history, but the developmental consequences may remain benign epilepsy with centrotemporal spikes (BECTS) and profound as seen in some patients with KCNQ2 childhood absence epilepsy (CAE). Despite the gestalt of a encephalopathy or STXBP1 encephalopathy. These obser- benign syndrome, BECTS may be associated with transient vations, pertinent to many of the genetic encephalopathies, or long-lasting cognitive effects56,57 and CAE with signifi- suggest that a broadening of the terminology, where appro- cant psychosocial consequences such as increased risk of priate, to include the word “developmental,” acknowledges early pregnancy.58 that both aspects may be playing a role in the clinical pre- The Berg et al. report9 suggested new terms to distill the sentation. These concepts are crucial to understanding the elements implied in the term “benign.” Thus “benign,” as a disease process for both families and clinicians. descriptor for epilepsy, is replaced by both “self-limited” It is therefore suggested that the term “developmental and “pharmacoresponsive,” each replacing different com- and epileptic encephalopathy” be used where appropriate ponents of the meaning of benign. “Self-limited” refers to and can be applied to individuals of any age. This allows for the likely spontaneous resolution of a syndrome. “Pharma- the use of either or both descriptors: developmental coresponsive” means that the epilepsy syndrome will be encephalopathy where there is just developmental impair- likely to be controlled with appropriate antiepileptic ther- ment without frequent epileptic activity associated with apy. It is important to acknowledge, however, that there will regression or further slowing of development; epileptic be individuals with these syndromes who are not pharma- encephalopathy where there is no preexisting developmen- coresponsive. As noted previously, there is no formal ILAE tal delay and the genetic mutation is not thought to cause classification of syndromes; however, we expect the word slowing in its own right; and developmental and epileptic benign in time will be replaced in the names of specific syn- encephalopathy where both factors play a role. Often it may dromes. The terms “malignant” and “catastrophic” will no not be possible to disentangle whether the epileptic or devel- longer be used; they will be removed from the epilepsy lexi- opmental component is more important in contributing to a con because of their serious and devastating connotations. patient’s presentation. It is hoped that this new Classification of the Epilepsies Many patients with these disorders have been classified will serve the epilepsy community well, leading to previously as having “symptomatic generalized epilepsies”; improved diagnosis, understanding of etiology, and targeted however, this term will no longer be used as it was applied therapies to the patient’s disease. It is notable that even to a highly heterogeneous group of patients. This term has where the etiology is clearly defined, the underlying mecha- been applied to patients with developmental encephalopa- nism producing recurrent seizures still requires elucidation. thies and epilepsy (e.g., static intellectual disability and With significant advances in understanding the neurobiol- mild epilepsy), those with epileptic encephalopathies, those ogy of seizures and epileptic diseases, there have been with developmental and epileptic encephalopathies, as well major paradigm shifts in the concepts underpinning classifi- as some patients with generalized epilepsy or combined cation. This Classification is designed to mirror current generalized and focal epilepsy. The new classification will understanding, so that it is relevant to clinical practice as the allow more precise classification of these individuals’ preeminent tool for communication in both clinical and epilepsy. research domains. Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709 520 I. E. Scheffer et al. Disclosure of Conflict of References Interest 1. Gastaut H, Caveness WF, Landolt W, et al. A proposed international classification of epileptic seizures. Epilepsia 1964;5:297–306. Ingrid Scheffer received support from and/or has served as a paid con- 2. Gastaut H. Clinical and electroencephalographical classification of sultant for UCB, Eisai, Athena Diagnostics, GlaxoSmithKline, Transge- epileptic seizures. Epilepsia 1969;10(Suppl.):2–13. nomics, and Biocodex. She serves on the editorial boards of Neurology 3. Gastaut H. Classification of the epilepsies. Proposal for an international and Epileptic Disorders. She has received grants from the National classification. Epilepsia 1969;10(Suppl.):14–21. Health and Medical Research Council, Australian Research Council, 4. Commission on Classification and Terminology of the International National Institutes of Health (NIH), Human Research Council, Citizens League Against Epilepsy. Proposal for classification of epilepsies and United for Research in Epilepsy (CURE), the US Department of Defense, epileptic syndromes. Epilepsia 1985;26:268–278. and the March of Dimes. Samuel Berkovic discloses payments from 5. Commission on Classification and Terminology of the International UCB Pharma, Novartis Pharmaceuticals, Sanofi-Aventis, and Jansen League Against Epilepsy. Proposal for revised classification of epilep- Cilag for advisory board and educational activities, and a patent for sies and epileptic syndromes. Epilepsia 1989;30:389–399. SCN1A testing held by Bionomics Inc licensed to various diagnostic 6. Commission on Classification and Terminology of the International companies. Giuseppe Capovilla serves on the editorial board of the Euro- League Against Epilepsy. Proposal for revised clinical and electroen- pean Journal of Pediatric Neurology. Mary Connolly has received cephalographic classification of epileptic seizures. Epilepsia research grants and/or speakers honoraria from UCB, Novartis, Biocodex, 1981;22:489–501. Eisai, and Sage Therapeutics. All honoraria are donated to the Epilepsy 7. Engel J Jr. A proposed diagnostic scheme for people with epileptic sei- Research and Development Fund. She has also received research grants zures and with epilepsy: report of the ILAE Task Force on Classifica- from CIHR (Canadian Institute for Health Research) and The Alva Foun- tion and Terminology. Epilepsia 2001;42:796–803. dation. She is Co-Chair of the Canadian Paediatric Epilepsy Network. 8. Engel J. Report of the ILAE classification core group. Epilepsia Jacqueline French: The Epilepsy Study Consortium pays her university 2006;47:1558–1568. employer for her consultant time related to Acorda, Anavex, Brabant 9. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminology and con- Pharma, Bio-Pharm Solutions, Eisai Medical Research, GlaxoSmithK- cepts for organization of seizures and epilepsies: report of the ILAE line, GW Pharma, Impax, Johnson & Johnson, Marinus, Neusentis, Commission on Classification and Terminology, 2005–2009. Epilepsia Novartis, Roivant, Pfizer, Sage, Sunovion, SK Life Sciences, Supernus 2010;51:676–685. Pharmaceuticals, Takeda, UCB, Upsher-Smith, Ultragenyx, Vertex, 10. Fisher RS, Cross JH, French JA, et al. Operational classification of sei- Zogenix, Zynerba, and the Scientific Advisory Board for Anavex, UCB; zure types by the International League Against Epilepsy: Position grants and research from Acorda, Alexza, LCGH, Eisai Medical Paper of the ILAE Commission for Classification and Terminology. Research, Lundbeck, Pfizer, SK Life Sciences, UCB, Upsher-Smith, and Epilepsia 2017;58:522–530. Vertex; and grants from National Institute of Neurological Disorders and 11. Epilepsy ILa. Guidelines for publications from league commissions Stroke, Epilepsy Therapy Project, Epilepsy Research Foundation, Epi- and task forces, 2014. Available at:http://www.ilae.org/visitors/centre/ lepsy Study Consortium. She is on the editorial board of Lancet Neurol- guidelines.cfm. Accessed July 2, 2015. ogy, Neurology Today, and Epileptic Disorders, and is an Associate 12. Avanzini G. A sound conceptual framework for an epilepsy classifica- Editor of Epilepsia, for which she receives a fee. Laura Guilhoto: serves tion is still lacking. Epilepsia 2010;51:720–722. on the editorial board of Seizure—European Journal of Epilepsy. 13. Beghi E. New classification proposals for epilepsy: a real advancement Edouard Hirsch has received support from UCB, and/or has served as a in the nosography of the disease? Epilepsia 2011;52:1197–1198; dis- paid consultant for UCB, Eisai, and Bial. Satish Jain and Yue-Hua Zhang cussion 1205–1199. have no disclosures. Gary Mathern is partially supported by the Davies/ 14. Berg AT, Scheffer IE. New concepts in classification of the epilepsies: Crandall Chair for Epilepsy Research at UCLA and is co-editor in chief entering the 21st century. Epilepsia 2011;52:1058–1062. for Epilepsia and Epilepsia Open. He is also on the editorial board of 15. Berg AT, Scheffer IE. What is at stake in a classification? Epilepsia Neurology, and the Data Management Committee for NeuroPace, Inc. 2011;52:1205–1208. Solomon L. Mosh
e MD is the Charles Frost Chair in Neurosurgery and 16. Camfield P. Issues in epilepsy classification for population studies. Neurology and funded by grants from NIH NS43209 and Epilepsia 2012;53(Suppl. 2):10–13. 1U54NS100064-01, CURE Infantile Spasms Initiative, the US Depart- 17. Duncan JS. The evolving classification of seizures and epilepsies. ment of Defense (W81XWH-13-1-0180), the Heffer Family and the Epilepsia 2011;52:1204–1205; discussion 1205–1209. Segal Family Foundations, and the Abbe Goldstein/Joshua Lurie and 18. Engel J Jr. The etiologic classification of epilepsy. Epilepsia Laurie Marsh/Dan Levitz families. He serves as Associate Editor of Neu- 2011;52:1195–1197; discussion 1205–1209. robiology of Disease, and is on the editorial boards of Brain and Devel- 19. Ferrie CD. Terminology and organization of seizures and epilepsies: rad- opment, Pediatric Neurology, and Physiological Research. He receives ical changes not justified by new evidence. Epilepsia 2010;51:713–714. from Elsevier an annual compensation for his work as Associate Editor 20. Fisher RS. What is a classification essay? Epilepsia 2010;51:714–715. of Neurobiology of Disease and royalties from two books he co-edited. 21. Guerrini R. Classification concepts and terminology: is clinical He received a consultant’s fee from Eisai and UCB. Douglas Nordli is description assertive and laboratory testing objective? Epilepsia funded from NIH (1-RO1-NS43209) and CURE. He is an Associate Edi- 2010;51:718–720. tor for UpToDate. Emilio Perucca has received speaker’s or consultancy 22. Jackson G. Classification of the epilepsies 2011. Epilepsia fees and/or research grants from the following pharmaceutical compa- 2011;52:1203–1204; discussion 1205–1209. nies: Eisai, Biopharm Solutions, GW Pharma, Mylan, Sanofi, SK Life 23. Moshe SL. In support of the ILAE Commission classification proposal. Sciences, Sun Pharma, Takeda, and UCB Pharma. Torbj€orn Tomson has Epilepsia 2011;52:1200–1201; discussion 1205–1209. received research grants and/or speakers honoraria to his institution from 24. Luders HO, Amina S, Baumgartner C, et al. Modern technology calls the following pharmaceutical companies: Eisai, GlaxoSmithKline, Novar- for a modern approach to classification of epileptic seizures and the tis, Bial, and UCB. He has also received research grants from CURE, epilepsies. Epilepsia 2012;53:405–411. Stockholm County Council, and EU (DG Sante). Samuel Wiebe has 25. Panayiotopoulos CP. The new ILAE report on terminology and con- received speaker’s or consultancy fees and/or research/educational grants cepts for the organization of epilepsies: critical review and contribu- from UCB, Electrocore, and Sunovion. Sameer Zuberi has received tion. Epilepsia 2012;53:399–404. research support and or speaker’ honoraria/consultancy fees from 26. Shinnar S. The new ILAE classification. Epilepsia 2010;51:715–717. Epilepsy Research UK, Dravet Syndrome UK, UCB Pharma, Yorkhill 27. Shorvon SD. The etiologic classification of epilepsy. Epilepsia Children’s Charity, GW Pharma, Brabant Pharma, and Zogenix. He is 2011;52:1052–1057. Editor-in-Chief of the European Journal of Paediatric Neurology. We 28. Wolf P. Much ado about nothing? Epilepsia 2010;51:717–718. confirm that we have read the Journal’s position on issues involved in 29. Wong M. Epilepsy is both a symptom and a disease: a proposal for a ethical publication and affirm that this report is consistent with those two-tiered classification system. Epilepsia 2011;52:1201–1203; dis- guidelines. cussion 1205–1209. Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709  521 Classification of the Epilepsies 30. Scheffer I, French J, Hirsch E, et al. Clasification of the epilepsies: 45. Helbig I, Mefford H, Sharp A, et al. 15q13.3 microdeletions increase new concepts for discussion and debate – special report of the ILAE risk of idiopathic generalized epilepsy. Nat Genet 2009;41:160–162. Classification Task Force of the Commission for Classification and 46. Dibbens LM, Mullen S, Helbig I, et al. Familial and sporadic 15q13.3 Terminology. Epilepsia Open 2016;1:37–44. microdeletions in idiopathic generalized epilepsy: precedent for disor- 31. Fisher RS, Acevedo C, Arzimanoglou A, et al. ILAE official report: a ders with complex inheritance. Hum Mol Genet 2009;18:3626–3631. practical clinical definition of epilepsy. Epilepsia 2014;55:475–482. 47. Claes L, Del-Favero J, Ceulemans B, et al. De novo mutations in the 32. McTague A, Howell KB, Cross JH, et al. The genetic landscape of the sodium-channel gene SCN1A cause severe myoclonic epilepsy of epileptic encephalopathies of infancy and childhood. Lancet Neurol infancy. Am J Hum Genet 2001;68:1327–1332. 2016;15:304–316. 48. Weckhuysen S, Mandelstam S, Suls A, et al. KCNQ2 encephalopathy: 33. Guerrini R, Pellacani S. Benign childhood focal epilepsies. Epilepsia emerging phenotype of a neonatal epileptic encephalopathy. Ann Neu- 2012;53(Suppl. 4):9–18. rol 2012;71:15–25. 34. Beaumanoir A, Nahory A. [Benign partial epilepsies: 11 cases of fron- 49. Depienne C, Bouteiller D, Keren B, et al. Sporadic infantile epileptic tal partial epilepsy with favorable prognosis]. Rev Electroencephalogr encephalopathy caused by mutations in PCDH19 resembles Dravet Neurophysiol Clin 1983;13:207–211. syndrome but mainly affects females. PLoS Genet 2009;5:e1000381. 35. Berkovic SF, McIntosh A, Howell RA, et al. Familial temporal lobe 50. Arsov T, Mullen SA, Rogers S, et al. Glucose transporter 1 deficiency epilepsy: a common disorder identified in twins. Ann Neurol in the idiopathic generalized epilepsies. Ann Neurol 2012;72:807–815. 1996;40:227–235. 51. Scheffer IE, Grinton BE, Heron SE, et al. PRRT2 phenotypic spectrum 36. Sveinbjornsdottir S, Duncan JS. Parietal and occipital lobe epilepsy: a includes sporadic and fever-related infantile seizures. Neurology review. Epilepsia 1993;34:493–521. 2012;79:2104–2108. 37. Gaillard WD, Chiron C, Cross JH, et al. Guidelines for imaging infants 52. Carvill GL, Heavin SB, Yendle SC, et al. Targeted resequencing in and children with recent-onset epilepsy. Epilepsia 2009;50:2147– epileptic encephalopathies identifies de novo mutations in CHD2 and 2153. SYNGAP1. Nat Genet 2013;45:825–830. 38. Guerrini R, Dobyns WB. Malformations of cortical development: clin- 53. Depienne C, Trouillard O, Gourfinkel-An I, et al. Mechanisms for ical features and genetic causes. Lancet Neurol 2014;13:710–726. variable expressivity of inherited SCN1A mutations causing Dravet 39. Grinton BE, Heron SE, Pelekanos JT, et al. Familial neonatal seizures syndrome. J Med Genet 2010;47:404–410. in 36 families: clinical and genetic features correlate with outcome. 54. Vezzani A, Fujinami RS, White HS, et al. Infections, inflammation Epilepsia 2015;56:1071–1080. and epilepsy. Acta Neuropathol 2016;131:211–234. 40. Tinuper P, Bisulli F, Cross JH, et al. Definition and diagnostic criteria 55. Lancaster E, Dalmau J. Neuronal autoantigens–pathogenesis, associ- of sleep-related hypermotor epilepsy. Neurology 2016;86:1834–1842. ated disorders and antibody testing. Nat Rev Neurol 2012;8:380–390. 41. Lennox WG. The genetics of epilepsy. Am J Psychiatry 1947;103:457– 56. Staden UE, Isaacs E, Boyd SG, et al. Language dysfunction in children 462. with rolandic epilepsy. Neuropediatrics 1998;29:242–248. 42. Lennox WG. The heredity of epilepsy as told by relatives and twins. J 57. Lillywhite LM, Saling MM, Harvey AS, et al. Neuropsychological and Am Med Assoc 1951;146:529–536. functional MRI studies provide converging evidence of anterior lan- 43. Brunklaus A, Dorris L, Ellis R, et al. The clinical utility of an SCN1A guage dysfunction in BECTS. Epilepsia 2009;50:2276–2284. genetic diagnosis in infantile-onset epilepsy. Dev Med Child Neurol 58. Wirrell EC, Camfield CS, Camfield PR, et al. Long-term psychosocial 2013;55:154–161. outcome in typical absence epilepsy. Sometimes a wolf in sheeps’ 44. Mullen SA, Scheffer IE. Translational research in epilepsy genetics: clothing. Arch Pediatr Adolesc Med 1997;151:152–158. sodium channels in man to interneuronopathy in mouse. Arch Neurol 2009;66:21–26. Epilepsia, 58(4):512–521, 2017 doi: 10.1111/epi.13709