Preoperative clinical, EEG, and imaging findings do not predict seizure outcome following temporal lobectomy in childhood.

Abstract

Although certain clinical, electroencephalographic (EEG), magnetic resonance imaging (MRI), and pathologic findings in adults with intractable temporal lobe epilepsy predict seizure outcome following temporal lobectomy, predictors of seizure outcome have not been studied systematically in pediatric temporal lobectomy series. We retrospectively analyzed preoperative clinical, EEG, and neuroimaging findings with reference to seizure outcome (seizure free or non-seizure free) in 33 children (mean age, 9.3 years) who underwent tailored temporal lobe resections for intractable temporal lobe epilepsy. Trends were apparent with (1) younger age at seizure onset, younger age at surgery, shorter duration of epilepsy, localized unilateral temporal lesions on MRI, and right-sided surgery more frequently associated with a seizure-free outcome, and (2) significant prior history, daily preoperative seizures, generalized motor seizures, mental retardation, and localized unilateral temporal epileptiform EEG activity more frequently associated with a non-seizure-free outcome. However, none of these findings, alone or in combination, correlated with postoperative seizure status at a statistically significant level. Submitting the four variables generally considered to be most predictive of favorable outcome (ie, normal intelligence, unilateral ictal and interictal EEG discharges, and focal temporal MRI lesion) to a multiple-cutoff procedure did not predict seizure freedom. Our data indicate that predictors of outcome of temporal lobectomy in adults may not apply in children, perhaps due to inherent neurobiologic differences in the etiology and expression of temporal lobe epilepsy, and should therefore not be used as sole determinants of surgical candidacy in children.