[Discontinuation of treatment in epileptic seizures in adolescence].

A retrospective survey on the frequency of seizure-recurrence after drug withdrawal in all forms of adolescent seizures has been undertaken to detect possible prognostic criteria. Patients were selected according to 3 criteria: first seizure between 11 and 19 years of age; out-patients seen by one of us between 1955 and 1979 within the year of onset of epilepsy; all patients followed for at least 5 years after the first seizure. Two-hundred and seventy two patients entered the study. One-hundred and four patients (49 per cent) relapsed, most of them (83 per cent) within the first year after drug withdrawal. In univariate analysis, partial seizures, a normal initial EEG, an isolated seizure, a short length of illness, a long seizure-free period, one seizure only during the first year were significantly linked to a low relapse rate. The probability of being seizure-free after drug withdrawal was about 73 per cent in partial epilepsy, 20 per cent in idiopathic generalized epilepsy, and 36 per cent in undetermined generalized epilepsy. In patients having had a single seizure a dramatic difference was noted according to the seizure type: 12 per cent of relapses after a partial seizure, 69 per cent after a generalized seizure. Neither the age of onset of epilepsy nor the presence of an etiological factor were significant variables in predicting the outcome. When all factors were analysed simultaneously with Cox's hazard function, the type of seizure and the initial EEG appeared to be the only two independent factors significant for a risk of relapse.(ABSTRACT TRUNCATED AT 250 WORDS)