Predictors of longitudinal seizure outcomes after epilepsy surgery in childhood

There is a paucity of data on longitudinal seizure outcome of children undergoing epilepsy surgery. All children (n = 132) who underwent resective epilepsy surgery from January 1998 to December 2015 were identified. Relevant clinical, neurophysiological, imaging, surgical and seizure outcome data were extracted. Multivariable logistic regression analysis and Kaplan-Meier survival with Cox proportional hazard modelling were performed. The mean age at surgery was 7.8 years (range 0.2-17.9). 71% were seizure-free at a mean follow up of 5.3 ± 2.7 years. Of those who were seizure-free, 65 patients were able to completely wean off anti- seizure medications successfully. Using survival analysis, the probability of Engel Class I outcome at one year after surgery was 81% (95% confidence interval [CI] 87%-75%). This dropped to 73% at two years (95% CI 81%-65%), 58% at five years (95% CI 67.8%-48%), and 47% at ten years. Proportional hazard modelling showed that the presence of moderate to severe developmental disability (HR 6.5; p = 0.02) and lack of complete resection (HR 0.4; p = 0.02) maintain association as negative predictors of seizure-free outcome. Our study demonstrates favorable long-term seizure control following pediatric epilepsy surgery and highlights important predictors of seizure outcome guiding case selection and counseling of expectations prior to surgery.