Predictors for seizure relapse in children with epilepsy after antiepileptic drug withdrawal: case-control study

Background Epilepsy is defined as a neurologic condition caused by a pathological brain condition. Epilepsy patients who have stopped treatment and are seizure-free for two years are considered to have achieved complete remission. Relapse occurs when seizures return after anti-epileptic drug (AED) withdrawal. Several studies reported that frequent seizures, prolonged duration to control seizures, number of AEDs consumed, and abnormal electroencephalography (EEG) found during AEDs tapered, were reported as predictors of relapse. Methods This retrospective, case-control study was carried out from 2012 – 2016 using multisite medical record evaluation, followed by interviews and EEG examinations. The case group included children with epilepsy who had seizure relapse, while the control group included children with complete remission of epilepsy. Bivariate and multivariate analyses were performed to identify predictors of relapse. Results Relapse predictors in bivariate analysis were symptomatic etiology epilepsy (OR 5.000; 95%CI 2.345 to 10.660; P<0.001), time to seizure control ?1 year (OR 3.689; 95%CI 1.493 to 9.116; P=0.003), and worsened EEG evolution at pre-withdrawal compared to EEG at the time of diagnosis (OR 2.310; 95%CI 1.132 to 4.717; P=0.021). Statistically significant relapse predictors in multivariate analysis were symptomatic etiology epilepsy (OR 4.384; 95%CI 1.985 to 9.681; P<0.001) and slow (?1 year) time to seizure control (OR 4.355; 95%CI 1.753 to 10.817; P 0.002). Conclusion Symptomatic etiology epilepsy and time to seizure control ?1 year are independent predictive factors for seizure relapse in children with epilepsy. Therefore, children with these conditions require a longer period and careful, gradual dose reduction before antiepileptic drugs withdrawal.