Long-term seizure and psychosocial outcome in childhood epilepsy of unknown cause

Objective The purpose was to investigate long-term prognosis of epilepsy of unknown cause with onset between ages 2 and 16 in children without any major disability, by evaluation of a previously described prognostic model and long-term follow-up of a study on the impact of duration of initial antiseizure medication (ASM) treatment. Methods Patients included in a randomized controlled trial (RCT) of either one or three years of ASM therapy prior to withdrawal (if seizure-free for at least 6 months) were contacted after 29-35 years and asked to complete a survey. Potential prognostic factors were evaluated: duration of initial ASM treatment, seizure type, seizure frequency, and score in a prognostic model developed in the initial publication. Results One hundred and forty-nine subjects answered the questionnaire (response rate 65%). Seizure freedom without treatment was found in 110 responders (77%, 95%CI: 73-81). There was no significant difference in score in the prognostic model between responders with and without epilepsy at follow-up. Those with active epilepsy were unemployed significantly more often and perceived their mental health significantly more affected than those seizure-free without treatment. Conclusions Duration of initial ASM treatment was not associated with any difference in subsequent epilepsy risk. This indicates that the timing of withdrawal attempts is unlikely to alter the long-term prognosis of uncomplicated childhood epilepsy. The failure of the prognostic model from the initial study to predict long-term outcome argues that although prediction of relapse risk in the shorter term may be possible, the bearing of such models on long-term epilepsy risk is more questionable.