Purpose: To distinguish various types of childhood severe cryptogenic/idiopathic generalised epilepsy on the basis of reproducible diagnostic criteria, using multiple correspondence analysis (MCA). Methods: We applied MCA to a series of 72 children with no evidence of brain damage, starting epilepsy between 1 and 10 years, with two or more types of generalised seizures. We excluded patients with infantile spasms or typical absences. MCA was performed on all clinical and EEG parameters, first throughout follow-up, then restricted to the first year of the disease. Results: When including all follow-up variables, there were three groups: (1) Thirty-seven children with male predominance, familial history of epilepsy, simple febrile convulsions, massive myoclonus, tonic-clonic fits. Outcome was favourable, with no seizures and mildly affected cognitive functions. Interictal EEG showed short sequences of irregular 3-Hz spike-waves. (2) In 18 children, clinical characteristics were similar to those of the first group at the early stage, but 95% exhibited myoclonic status and vibratory tonic seizures, with persisting seizures on follow-up. EEG showed long sequences of generalised irregular spike and slow waves. Those two groups meet the characteristics of childhood onset myoclonic-astatic epilepsy (MAE) with respectively, favourable and unfavourable outcome. (3) Eleven children had later onset, atypical absences, tonic and partial seizures, and no myoclonus, or vibratory tonic seizures. All had mental retardation and persisting seizures. EEG showed long sequences of slow spike-wave activity and half the patients had spike and slow wave foci. These patients met the major characteristics of Lennox-Gastaut syndrome. Initial parameters failed to distinguish the first two groups, but Lennox-Gastaut syndrome (the third group) was distinct from both groups of myoclonic astatic epilepsy from the onset. Within MAE groups combined, clinical and EEG risk factors for mental retardation could be identified. Conclusion: It is possible to validate statistically the distinction between discrete epileptic syndromes. Myoclonic astatic epilepsy is therefore distinct from Lennox-Gas taut syndrome, and the distinction appears from the first year of the disorder. (C) 1999 Elsevier Science B.V. All rights reserved.
