SEARCH

SEARCH BY CITATION

I appreciate the opportunity to add my opinion to the discussion of the complex subject of epilepsy classification, as prompted by the recent International League Against Epilepsy (ILAE) Commission report (Berg et al., 2010). The accompanying articles in this issue, by Berg & Scheffer (2011) and Shorvon (2011) are commendable not only in clearly outlining the limitations of imperfect classification systems, but also in attempting to improve on a process that will continue to evolve with future scientific and clinical discoveries. In my view, the most important issue is whether the classification system should be based on defining epilepsy as a symptom/syndrome or a disease. Although the semiologic descriptions of seizure types and epilepsy syndromes have served as the primary basis for classification for decades, technologic advances, especially in genetics and neuroimaging, have identified the etiologies of many epilepsies, and thus have redefined them as specific pathophysiological entities or diseases. The reductionist nature of scientific progress tends to promote an etiologic classification, which eventually may completely replace the semiologic descriptions of disease. However, I would contend that these two levels of classification are not mutually exclusive, and in fact both are necessary to provide a complete categorization of epilepsy.

Seizure semiology and epilepsy syndromes represent the aspects of the patient that the clinician can most directly observe and characterize. Although the distinction between generalized and partial epilepsies has perhaps become less useful from a pathophysiologic standpoint, and the ILAE committee has recommended that this dichotomy be abandoned (Berg et al., 2010), there will always be great practical value for a clinician to classify a patient’s seizures as generalized or partial/focal—such as in distinguishing between absence and complex partial seizures. Furthermore, the classic “idiopathic” syndromes, such as childhood absence epilepsy and juvenile myoclonic epilepsy, are entities that are universally recognized and easily diagnosed, mostly just by history and bedside examination (e.g., hyperventilation) as well as supportive electroencephalographic evidence; these syndromic diagnoses carry meaningful therapeutic and prognostic implications for the clinician and patient. Therefore, even if a specific cause for every case of epilepsy could be identified, there would still be value in classifying epilepsy based on semiology and syndrome.

Although identification of etiology is of tremendous importance, etiology alone does not provide a complete or specific picture of epilepsy. As pointed out by Berg and Scheffer, the same epilepsy syndrome can be caused by numerous etiologies, such as Lennox-Gastaut syndrome secondary to a multitude of genetic, metabolic, and structural causes. Conversely, the same etiology can produce a diversity of epilepsy syndromes, such as SCNA1 mutations causing febrile seizures and Dravet syndrome, or tuberous sclerosis complex causing infantile spasms, Lennox-Gastaut syndrome, and isolated partial epilepsy/seizures. Therefore, identifying both the seizure semiology/epilepsy syndrome and the associated etiology has significant clinical relevance.

In addressing the dichotomy between symptoms and disease, the new ILAE classification has essentially mixed syndromes and causes together in a parallel structure, either resulting in use of one in exclusion of the other or encouraging a flexible, customized selection of various descriptive and etiologic features most relevant to a specific patient (Berg et al., 2010; Berg & Scheffer, 2011). Therefore, the new ILAE system intentionally lacks any uniform, standardized organization or hierarchy for classifying epilepsy. This system basically amounts to a glossary or database of descriptors and causes to pick and choose from. In contrast, the paper by Shorvon focuses primarily on etiologic classification, although epilepsy syndromes are still recognized within the older framework of “idiopathic” and “symptomatic” epilepsies (Shorvon, 2011). Given the importance of both semiologic/syndromic features and etiologies of epilepsy, I would propose a compromise between these two approaches, in the form of an obligatory two-tiered classification, so that all epilepsies are always categorized according to both semiology/syndrome and cause. In the first tier, the most specific epilepsy syndrome or seizure description is identified; this descriptor could be a classic electroclinical syndrome or a seizure type(s). In the second tier, the identified etiology(ies) or “unknown cause” is specified.

In this two-tiered system, all epilepsies would be clearly defined as: epilepsy syndrome X secondary to cause Y. For example, one patient may have “juvenile myoclonic epilepsy secondary to a GABRA1 mutation,” whereas another patient may have “juvenile myoclonic epilepsy secondary to a malformation of cortical development” (this type of epilepsy has not been definitively established, but has been hypothesized based on recent pathophysiologic findings by de Nijs et al., 2009). For epilepsies without an identified cause, yet another patient may simply have “juvenile myoclonic epilepsy secondary to unknown cause.” Note that in this scheme, no type of epilepsy would be considered as “presumed genetic,” but would simply be attributed either to a genetic or nongenetic cause if one has been clearly documented, or to “unknown cause”; similarly, the controversial terms “idiopathic,”“cryptogenic,” and “symptomatic” are not needed. In the many cases of nonsyndromic epilepsies, the most specific descriptive information about seizure type(s) should be used for the symptom-based classification; for example, “epilepsy with partial/focal seizures secondary to cause Y.” If the seizure type(s) are not well-defined, “epilepsy secondary to cause Y” would represent the simplest, most nonspecific semiologic classification. Similarly, the etiologic classification could be as simple or detailed as appropriate, and given that many epilepsies may be multifactorial, could also include multiple causes (e.g., epilepsy syndrome X secondary to traumatic brain injury and SCNA1 mutation).

If the importance and validity of this dual concept of epilepsy are accepted, the generation of an official two-tiered classification system would simply involve a reorganization of the existing systems. Tier 1 would consist of the electroclinical syndromes that have been previously recognized (Engel, 2006; Berg et al., 2010), in addition to a listing of accepted seizure types for characterizing nonsyndromic epilepsies. Tier 2 would consist of a list of symptomatic etiologies—including specific genetic, structural, and metabolic causes, as already identified by Berg et al., 2010 and Shorvon, 2011, in addition to the category “unknown cause.” This proposal for a two-tiered classification is not particularly novel, as some suggestions of a similar system can be found within the previous works. However, in contrast to the flexible, open-ended approach endorsed in the latest ILAE Commission report, this approach represents a more formalized attempt to adopt a rational, standardized content and organization for epilepsy classification. It is hoped that this two-tiered classification system would promote a relatively comprehensive documentation of both semiologic and etiologic factors that are clinically relevant to epilepsy patients, while maintaining flexibility to incorporate future scientific advances.

Acknowledgments

  1. Top of page
  2. Acknowledgments
  3. Disclosure
  4. References

I receive funding from the National Institutes of Health (R01 NS056872), the Citizens United for Research in Epilepsy, the McDonnell Center, and the Washington University/Pfizer Collaboration.

Disclosure

  1. Top of page
  2. Acknowledgments
  3. Disclosure
  4. References

I have no conflicts of interest to disclose. 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.

References

  1. Top of page
  2. Acknowledgments
  3. Disclosure
  4. References