Of Cabbages and Kings: Do We Really Need a Systematic Classification of Epilepsies?

Pappataci! che mai sento! “Pappataci”! What's this I hear? La ringrazio. Son contento. I thank you and I am happy Ma di grazia Pappataci But, for mercy's sake, Che vuol poi significar? What's the meaning of “pappataci”? L'Italiana in Algeri (G. Rossini) I hesitated before entering the arena of the classification debate, but decided to take the risk. The following views are personal, and not officially those of the ILAE President or Board. I shall consider the classifications of epileptic seizures and epileptic syndromes separately. It seems to be generally agreed that our present knowledge allows us to draw up a seizure taxonomy reflecting our “understanding of the significance of signs and symptoms”(1), based on the 1981 ICES and taking into consideration Hans Lüders's proposal (2). However, I do not agree with Peter Wolf's statement that the 1981 ICES (3) faithfully reflects seizures “unfolding over time.” It may be true for some types of seizure, such as Jacksonian and tonic-clonic seizures, whose typical time course is implicit in the definition, but it is not true for others. In particular, the sequence of events characterizing some partial seizures are not easily incorporated into the 1981 ICES (3) definitions, and Hans Lüders's classification (2,4) provides some helpful concepts for overcoming its limitations. The situation is more complicated when it comes to classifying epileptic syndromes or the epilepsies, as I prefer to call them. Past efforts to classify the epilepsies along the spectra of idiopathic-cryptogenic-symptomatic or partial-generalized categories (5) have advanced our clinical conceptualization. However, ongoing clinical observation and neurobiological studies mandate a revision of our taxonomy (6). One key question is how the use of the larger idiopathic category helps the clinician or advances our understanding. When I have diagnosed a patient as having childhood absence epilepsy, or a GEFS-plus syndrome, or any other type of epilepsy that is accepted as being idiopathic, what do I gain from putting my diagnosis in the box of the idiopathic syndromes? Idiopathic syndromes share several (but not all) common traits, but would it not be an advantage to be as specific as possible? General categories might be useful for certain purposes, for example, pharmacological trials, in which the authors might wish to give the percentages of the entrants with idiopathic, cryptogenic, or symptomatic epilepsy. Nevertheless, this aggregation of patients contains a trap; some papers give the strong impression that large categories mix “cabbages and kings.” How much more advantageous it would be for the reader to know how many and which patients were classified as having JME, how many as having BECT, and so on! Similar considerations apply to the utility of the partial (focal)-generalized categorization. We can argue for days about BECT or West syndrome being a partial (focal) or generalized epilepsy, but would either conclusion affect in any way our attitude toward a single patient or a group of patients? On the other hand, forcing every diagnosis into a category may lead to decisions that are simply based on theoretical assumptions. Temporal lobe epilepsy is associated with mesial temporal sclerosis, but current evidence does not prove that it depends on MTS, as would be implied if this diagnosis were put in the category of symptomatic epilepsies. Here again, whether or not an association is considered causative (thus determining a classification in the symptomatic or in the cryptogenic category, respectively) would not influence the clinical attitude and scientific approach to this disorder In his thoughtful paper (1), Peter Wolf stresses the concept that a classification reflects the state of our knowledge in a given field, and metaphorically highlights the difference between botanists' taxonomies and the lists made by the gardeners for utilitarian purposes. From this point of view, I doubt that the principles underlying the previous classification reflect our state of knowledge of “nature and nurture” in epilepsy. Genetic studies (7) are revealing previously unknown types of epilepsies, which sound to us like the pappataci did to Mustafa bey in Rossini's opera, but they have so far failed to identify the genetic basis of the most common, putatively genetic, idiopathic epilepsies. On the other hand, the pathogenetic mechanisms underlying symptomatic epilepsies seem to be much less clear than we thought, as studies on epilepsies associated with hippocampal sclerosis, cortical dysplasias, or genetically determined structural and biochemical defects are increasingly revealing to us. It is my opinion that we are losing the ground that the previous “botanists'” taxonomy of epilepsy was based upon, without a safety net of a new conceptual framework What are we then to do? If we cannot be botanists, let's be gardeners and cultivate the plants we have learned to identify and take care of, even if we do not yet know whether they belong to the thallophyte family or not. To step out of the metaphor, I think that all possible efforts should concentrate on: (a) revising the classification of seizures for which the present pathophysiological information is sufficient, in order to build the necessary conceptual framework; (b) reaching a consensus on a list of epilepsies (8) (or epilepsy types, or epileptic syndromes) for which natural history, constellation of signs and symptoms, prognosis, methods of evaluation and therapy are clear; and (c) creating denominations that clearly designate the distinctive characteristics of the individual syndromes while avoiding the eponyms that tends to deter “nonepileptologists” (but also epileptologists) from adopting the proposed denominations. How such an effort will lead to a systematic classification, cannot yet be elucidated. Personally, I do not think that continuing discussion and debate will upset too many epileptologists.