RESUMO
BACKGROUND: Ketogenic diet (KD) is an effective treatment for pharmacoresistant epilepsy: more than half of the patients show a greater than 50% reduction in seizures. OBJECTIVE: To identify clinical or electroencephalogram (EEG) variables predicting the response to KD. METHODS: Clinical and EEG data were retrospectively analysed from 50 consecutive patients treated by KD for severe, pharmacoresistant epilepsy. Most of the patients (70%) had retarded mental and motor development. RESULTS: Three months after the start of the KD two-thirds (33) of the patients were responders (had a more than 50% reduction in seizure frequency). The presence of epileptiform EEG discharges in the temporal region correlated with an unfavourable response (P = 0.03). The presence of bilateral synchronous epileptiform discharges, and the presence of complex partial seizures approached significance but all other variables did not. CONCLUSIONS: Our results further support that KD is efficient in a wide variety of epileptic patients with a broad range of EEG features. However, patients with epileptiform discharges in the temporal region are less likely to achieve therapeutic response.