Search for a command to run...
To the Editor: Fisher et al. (1) state, “Little common agreement exists on the definition of the terms seizure and epilepsy,” and they propose ILAE-endorsed definitions for these terms. Although their proposed definition of “seizure” is consistent with that which has been in use throughout the field for decades, their proposed definition of epilepsy is not. Fisher and colleagues (1) propose the following definition of epilepsy: “Epilepsy is a disorder of the brain characterized by an enduring predisposition to generate epileptic seizures and by the neurobiologic, cognitive, psychological, and social consequences of this condition.” The definition of epilepsy in Fisher's Table 1 (1) requires the occurrence of at least one epileptic seizure but not that the seizure be unprovoked. Although it may be helpful to consider diverse conditions (febrile seizure, acute symptomatic seizure, single unprovoked seizure, and epilepsy) within the context of studying the seizure disorders, it is not helpful to consider all of these conditions as epilepsy. The more restrictive definition of epilepsy (recurrent unprovoked seizures), adopted by the ILAE Commission on Epidemiology and Prognosis (2), is related to therapeutic, management, and counseling approaches and supported by epidemiologic studies of seizure disorders. Furthermore, this definition has been largely adopted in clinical practice and was instrumental in developing practice guidelines (3). The failure to clarify the concept of “enduring” is a problem with the proposed definition, and it is unclear how Fisher et al. (1) would define or make operational this term. Making operational “enduring alteration of the brain that increases the likelihood of future seizures (1)” would require a list of indicators of such an alteration. These, in turn, would have to be qualified and changed as knowledge increases. For clinical and scientific purposes, the operational criteria must be simple and robust. We suggest instead that the best way to know whether a person has an enduring alteration of the brain that increases the likelihood of future unprovoked seizures after a first seizure is the occurrence of a second unprovoked seizure. This new definition would reclassify many situations previously excluded from the term epilepsy in recent studies. Examples include a single provoked seizure secondary to a neurologic insult (e.g., stroke), a single provoked or unprovoked seizure in someone with depression or migraine, and a febrile seizure in a child with cerebral palsy, with an epileptiform EEG, or with febrile seizure recurrence. The all-inclusive definition proposed by Fisher et al. (1) is consistent with use before the emergence of the epidemiologic studies of seizure disorders and epilepsy over the past 60-year period. The exclusion of these conditions from the diagnosis of epilepsy was based on large, carefully conducted clinical and population-based studies. Most acute symptomatic seizures would be recategorized as epilepsy under the definition proposed by Fisher et al. (1). Acute symptomatic seizures have been defined as seizures in close temporal association with a transient CNS insult and presumed to be an acute manifestation of the insult. Although the risk of developing unprovoked seizure is higher in people with acute symptomatic seizures, in most, later seizures do not develop. Although the incidence of acute symptomatic seizure is similar to the incidence of epilepsy, the high early mortality and the protective effect of anticonvulsants on the development of acute symptomatic seizures dramatically distinguish this category of seizures from epilepsy. By the proposed definition (1), many children with febrile seizures, the most common convulsive disorder, would be reclassified as having epilepsy. This would be true for children with developmental delay, neurologic abnormalities, epileptiform EEG abnormalities, complex febrile seizure, and recurrent febrile seizure. Regardless of the presence of such factors, in most children with febrile seizure, later unprovoked seizures do not develop (4,5). Restricting the diagnostic labeling of epilepsy to the few who truly have recurrent unprovoked seizures would seem prudent. It is useful to study single unprovoked seizures within the context of epilepsy to better understand the underlying processes to increase the risk for the development of recurrent unprovoked seizures. Contrary to the proposed definition (1), the epidemiologic data on recurrence risks support separating single unprovoked seizure from recurrent unprovoked seizures (i.e., epilepsy). The recurrence risk is lower after a first unprovoked seizure (typically <50%) than the recurrence risk after a second unprovoked seizure for both children and adults (6,7), suggesting that the recurrence of unprovoked seizure or lack thereof delineates different entities. A major problem with the proposed definition (1), particularly for those with single seizure and with febrile seizure, is that labeling patients with only a single seizure as having epilepsy, when many will never experience another seizure, will cause unnecessary use of anticonvulsant drugs, increase stigma, and result in social and occupational limitation. This does not serve the needs of these patients and is inconsistent with epidemiologic data. The inclusion of associated conditions in the proposed definition (1) raises concerns on several levels. Although general agreement may exist that “for some people with epilepsy, behavioral disturbances such as interictal and postictal cognitive problems, can be part of the epileptic condition (1),” the definition as written seems to require these disturbances for the condition to be epilepsy. Thus a person with multiple unprovoked seizures and a likelihood of more would not have epilepsy by the definition of Fisher et al. unless one of these associated conditions also was present. This aspect of the proposal creates a new unnamed category that may be quite large—people who clearly have recurrent unprovoked seizures, but lack documentation of associated conditions. Even if the proposed behavioral component is accepted as an essential ingredient in the definition of epilepsy, it is unclear how this would be made operational. Other consequences ensue from this definition. The incidence of “epilepsy” will increase at least threefold, and the increase in prevalence will be greater, particularly in developing countries, which may provide political leverage. Undesired consequences of use of this definition will be the invalidation of prognostic studies, including those of mortality, long-term prognosis for seizure remission, and response to initial therapy. Contrary to the proposal of Fisher et al. (1), widespread acceptance of and agreement over the definitions of seizures and epilepsy are in general use in the field. We fail to see the advantages of the proposed definitions to the individual patient, to epilepsy as a condition, or to the study of epilepsy and the convulsive disorders. Maintaining a common language has been acknowledged in several ILAE Commission and Task Force reports as a prerequisite to communication and comparability of research from different groups. In addition, the medical, social, and emotional implications of epilepsy and seizures speak in favor of a separation between acute symptomatic seizures, febrile seizures, and unprovoked seizures and, for those with unprovoked seizures, between single and repeated episodes. To this end, the current definitions have been most successful. They are based on a process similar to the evidence-based approaches used for evaluating therapies and therapeutic policies. They may be subject to revision as new information comes to light, but this process should be respected. It does not appear that proposed definitions advance the field in any way.