Although epilepsy is an ancient disease, described in writings of ancient Mesopotamia and India,¹² recognition that there are many forms of epilepsy is a fairly recent development. Petit mal and focal seizures were distinguished from grand mal seizures by nineteenth century neurologists,^3,9 and Hughlings Jackson noted that there could be many causes of epilepsy¹⁶; however, there was no universally accepted classification of epilepsy syndromes until the late twentieth century. Even in 1970, when the International League Against Epilepsy (ILAE) proposed the first international classification of epileptic seizures, the accompanying classification of the epilepsies merely divided them broadly into partial and generalized types (Table 1).¹⁴ This classification further distinguished between primary generalized epilepsies, which were associated with epilepsy alone, and secondary generalized epilepsies, in which the seizures were symptoms of an identifiable cerebral disorder that itself could produce other signs and symptoms as well. At the time, it was thought that all partial seizures were symptomatic.

With the increasing description of reasonably well defined epilepsy syndromes, the ILAE made its first attempt to organize them into a coherent classification in 1985.⁴ The 1985 classification retained the partial/generalized dichotomy but deleted the term “partial” because it seemed an inappropriate term for diseases and created misunderstandings when secondarily generalized tonic–clonic seizures were the only seizure type. The possible term “focal” was not chosen because the idiopathic epilepsies of childhood with focal seizures do not present stable foci but may have seizures coming from alternate sites and sides. “Localization related” was proposed instead as the preferred term. The primary/secondary dichotomy was retained, but these two groups were now referred to as idiopathic and symptomatic because “secondary generalized epilepsies” had become confused with “secondarily generalized seizures.” The term “idiopathic,” from the Greek word “idios,” meaning self, referred to epilepsy as the disorder itself, or epilepsy sui generis. This classification recognized the fact that some localization-related epilepsies are idiopathic and, for the first time, listed specific syndromes within each of the categories. For generalized epilepsies, a category of “idiopathic and/or symptomatic” was included, as was a category of “epilepsies and syndromes undetermined as to whether they are focal or generalized,” together with a category of special syndromes. The International Classification of Epilepsies and Epileptic Syndromes was again revised in 1989, and this is the classification in current use⁵ (Table 2).

The 1989 classification retained the localization-related/generalized and idiopathic/symptomatic dichotomies but introduced the term “cryptogenic” to define conditions in which the cause of the disorder is “hidden or occult.”⁵ The term “cryptogenic epilepsies” was intended to be used for conditions presumed to be symptomatic but without definitive evidence for the etiology. The use of this term became ambiguous, however, as a result of a 1993 report of the ILAE Commission on Epidemiology and Prognosis in which it was recommended that the term “cryptogenic epilepsies” be used to define “patients who do not conform to the criteria for the symptomatic or idiopathic categories.”⁶ As a result, there is considerable confusion regarding the use of the term “cryptogenic,” and a more recent report of the ILAE Task Force on Classification and Terminology recommended that this term be discarded and replaced by “probably symptomatic” when this is the intent, or “unknown as to whether idiopathic or symptomatic” when this is the intent.⁷ It should also be noted that, whereas the ILAE classification is a taxonomic classification,¹⁸ the problem addressed by the term “cryptogenic” is more relevant to a diagnostic scheme intended to describe individual patients. The task force also made several other recommendations. Because the term “localization related” is cumbersome and not used by everybody, it was recommended that it be replaced with the commonly used term “focal,” with the understanding that focal epilepsies do not usually result from a small, discrete cluster of epileptogenic neurons. The term “focal,” however, is particularly inappropriate in this respect for the idiopathic localization-related epilepsies, which are distributed disorders. The words “convulsion” and “convulsive” were also felt to be imprecise, and it was recommended that they not be used to define specific syndromes.

Much of the 1989 classification of the epilepsies is derived from initial work done in children and adolescents, that is, age groups in which highly distinct and diverse forms of epilepsy occur. The use of epilepsy syndromes in children and adolescents has had tremendous utility for clinical as well as research purposes. Large-scale community-based and specialized centers–based studies have demonstrated that, within this young age group, 50% to 60% of children can be assigned a specific syndromic diagnosis.^2,11,17 Remaining cases carry
diagnoses of less well-defined and less specific entities, namely symptomatic or cryptogenic focal epilepsy and undetermined (essentially unclassified) epilepsy. The focal epilepsies are generally defined based on localization from semiology, electroencephalogram (EEG), or neuroimaging and by the presence or absence of an identified underlying cause. Many entities identified within the focal epilepsies do not conform to current concepts of what is meant by the term syndrome in the same way that, for example, childhood absence epilepsy or West syndrome do. The vast majority of adults have these latter forms of focal epilepsy and only about 10% or 15% meet the diagnostic criteria for other highly specific epilepsy syndromes.¹¹ This had led some to criticize the 1989 classification for attempting to include all epilepsies in a rigid classification structure. An alternative proposal maintains that it is sufficient to define individual patients by describing their ictal phenomenology in detail.¹³ This is again relevant to diagnostic schemes or manuals and not to taxonomy, which only makes sense when it includes all recognized conditions, be they rare or frequent. Taxonomic classifications, however, need to be accompanied by diagnostic manuals.

An alternative or supplementary approach taken by the ILAE Task Force is to create a list of epileptic seizure types as diagnostic entities that, by themselves, have etiologic, therapeutic, and prognostic implications.^8,19 These seizure types, based on pathophysiologic mechanisms and anatomic substrates, information that was inadequately available to be used in the 1981 ILAE Classification of Epileptic Seizures, can be used to determine the diagnostic evaluation, treatment, and prognosis in patients for whom a syndrome diagnosis is not possible.