Introduction: The 1989 International League Against Epilepsy (ILAE) classification of epilepsies has been used in many studies world-wide. In 2010, the ILAE proposed a new classification.
Aim: We evaluated the potential contribution of the 2010 classification of epilepsies compared to the previous one. We also analyzed the new version according to the principles of a good classification.
Development: The 2010 classification of epilepsies shows radical changes, both in terminology and taxonomic criteria, when compared to the 1989 classification. The new version lacks many of the desirable principles of a good classification system. The main criterion selected to classify the epilepsies (a mixture of syndromic specificity and etiology) divides the epilepsies in four groups which are not mutually exclusive. For instance, benign rolandic epilepsy should be included both under 'electroclinical syndromes' and 'epilepsies of unknown cause'. The division of clinical entities in 'electroclinical syndromes', 'constellations', and 'ill-defined syndromes' could have avoided such problem, although it appears to be clinically less relevant than 1989 classification criteria (localization and etiology). The ILAE has proposed an example of 'organization' of the epilepsies, instead of a true classification, and has encouraged the creation of different classifications for specific purposes. Should these ones were not linked to a previously established main classification they would represent a risk. Several international authorities have already disapproved the new proposals of the ILAE.
Conclusions: The attempts to replace the 1989 ILAE classification of epilepsies have been far from successful. An international debate on the subject might help to develop a new classification supported world-wide.