Classification of the epilepsies 2011

Abstract

The issue of classification of the epilepsies is an area that is capable of generating a great deal of passion. As usual, the heat is greatest when outcomes are dependent on opinion rather than data. The purpose of classification is to summarize and conceptualize very complex data. The reason we need to do this can be primarily considered to be narrative; that is, a classification system allows a sense of "knowing" and "recognition" that is not possible if each person is treated as the unique individual he/she actually is. Such a narrative needs to achieve, at least, the following goals: 1 Effective Communication—so that precision and commonality of description and language is used within the community of clinicians, researchers, and caregivers. 2 Education—so that learning the classification will teach us about the core biologic concepts of epilepsy. 3 Utility—so that natural classes and mechanisms can lead to clinical action and create specific challenges for research. Furthermore, any classification needs to be flexible. Ideally, it should be based on concepts of biology that are conceptually known, and allow for changes in specific instances as the biology is better understood. In other words, the framework of the classification should stand, and be founded, in biologic concepts. However, where we place particular recognizable disorders in this classification may change with advances in knowledge both of the basic biology that forms the framework for classification, but more importantly, in the biology of each recognizable epilepsy condition. Therefore, in any forward-looking classification system, many "disorders" will need to be initially placed in such a system based largely on educated guesses of the underlying biology. So, if we start with such an idea, the first question is whether the traditional way of classifying the epilepsies needs fixing. I first became involved in epilepsy in 1988, and was of the generation for which the classification system was "just there." I confess little joy of intellectual insight in learning the classification system, although I recognize that it achieved common language. The classification of seizures was helpful, but the next step, that bundled things into "syndromes," always seemed an intermediate step that stopped intellectual effort at recognition rather than understanding. In some ways, this system was neither fish nor fowl. I have a strong bias to meaningful biologic underpinnings of any classification system that will both challenge our diagnostic effort to lead us to a mechanistic understanding, and will add an additional framework for research to uncover these mechanisms. My opinion is that we now must move to an era that is about the mechanisms underlying epilepsy. At the present time, this effort is based on understanding of genetics, cellular function, structural abnormalities, and network function in the human brain. The real challenge to forming a classification is to decide the first level of division, and the simplifying assumptions that allow this division to work. For example, when the classification system for malformations of cortical development was being evolved, a decision was made that the classification would be primarily linked to an understanding of embryology, as this biologic knowledge was fundamental to brain development. Regardless of how one regarded the malformations seen in each individual, normal embryology is well understood, and so would provide the biologically based "containers" into which recognized entities could be placed. This classification system was a fundamental shift from the histopathologic classification in use at that time—a system that had little clinical utility when faced with magnetic resonance imaging (MRI)–revealed macroscopic abnormalities. One complex problem was that there are many abnormalities that occur in different points in development in most individuals. The principle of classifying disorders according to the very first identifiable developmental step that was abnormal, and ignoring all "downstream" events, allowed this classification to work, and led to a very functional and widely used classification system. Of course, other dimensions—such as genetics and mechanistic pathways—are critically important when understanding specific disorders. It is my belief that in a similar way, principles of biology rather than epistemology should underlie a modern classification system for the epilepsies. Applying this approach to the current argument, it is clear that the words "idiopathic,""symptomatic," and "cryptogenic" do not arise from biology, do not describe biology, have little inherent meaning, and (worse still) are used differently by different writers. I think it is clear from my argument above that we should reject these terms as first-order categories for classification. This rejection leads to a further major premise: that to change a word is to imply that there has been a change in thinking that stands behind that word. Understanding the reason for the change in the word should educate as to the new knowledge being applied to the epilepsy disorder. Some may be disappointed that Berg and Scheffer (2011) have made clear that the new proposal is not a classification scheme in itself, but an intermediate step between a previous system of classification and the evolution of a new system. Part of the problem in taking such an intermediate step is that it can be mapped to the older framework, allowing debates to arise about terms rather than concepts. However, with the introduction of new terminology, a first step has clearly been taken on a new path. For example, the term "idiopathic," which really is a "hand waving" term, has largely been replaced with the well-defined term "genetic epilepsy." All classification systems need to make simplifying assumptions to have value, and a disorder does not need to be exclusively genetic in order to be classified in this way. A future classification would hopefully go beyond this simplifying scheme and be based, at its first order, on the biologic mechanisms that these genes act through (e.g., cellular ion channel disorders and axon initial segment disorders). It seems to me that there is much in common between the two Commentary papers in this issue (Berg & Scheffer, 2011; Shorvon, 2011) at a conceptual level. If we accept that we need to change terminology to reflect a change in thinking, knowledge, and concepts (and I think we do), then a main issue for development seems to be in the "structural/metabolic" term, and the need to define the nature of "causes of epilepsy" that will provide the first level of description in the evolving classification system. Shorvon (2011) argues that we need to come up with a better framework to deal with these causes, and that it needs to be based on the underlying biology of both the initiating disease and the mechanism of the epilepsy process. I agree with this approach, and common ground could be found on dealing with this category as the next area of major emphasis. This work has to start with a decision about what is the fundamental biology that characterizes the causes and processes of epilepsy. A further problem that the classification system should deal with is the biology relating to consequences of epilepsy on the brain, regardless of initiating cause. It would be best to understand both the causes of seizures as well as the secondary consequences of seizures in order to have a mechanistic classification of the epilepsies—that is, the difference between a "disease" and the "cause of disease." By the criteria I set at the beginning, communication is enhanced with more precise definition of terms while at the same time removing ambiguity from variable interpretation of old terminology. Education is better because the new terminology goes directly to the nature of the biology of epilepsy and avoids "old scholar" terms (something that I still find awkward to teach). And utility is at least preserved during this step; because it leads naturally to the next step of a better biologic and mechanistic classification of epilepsy, it is likely to have unexpected benefits that accurate diagnosis and biologic understanding always bring. The properly constituted Classification Commission has moved us forward, and the discussion should now be about what biologic concepts should be the basis of the classification framework, and how we can focus work on understanding epilepsy as a disease (i.e., causes, processes, and consequences). Old concepts and terminology tightly held can act as sea anchors slowing the progress of discovery and advancement. Both the Berg & Scheffer and the Shorvon articles are striving for the same end. It is clear to me that we must give up old terminology not only because these words do not reflect current knowledge, but also as a symbol that things are changing. We need to clarify and modernize. Renewal is the only option, and the argument should no longer be focused on the use of words. We must now put our community efforts into the next steps. I have no conflicts of interest to disclose. I know and have worked with all of the authors that are referenced in this commentary. This work was supported by a National Health and Medical Research Council (Australia) practitioner fellowship award to the author. I confirm that I have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Keywords

Affiliated Institutions

• The University of Melbourne AU

Related Publications