Hemispherectomy is an established effective procedure for medication-resistant epilepsy in children. However, there are patients who do not achieve an optimum outcome after surgery; possible causes of suboptimal results include the presence of bilateral independent epileptogenic foci.

     Magnetoencephalography (MEG) is an emerging noninvasive ancillary test that has been found useful in the management of lesional and non-lesional epilepsy but its role is ill-defined in hemispherectomy.

     Thirteen patients (9 male) were included in the study. Average age at time of surgery was 66 months.

     A retrospective chart review was performed on the children undergoing hemispherectomy at the Hospital for Sick Children who have had an MEG study as part of the presurgical evaluation.

     Seizure etiologies were Rasmussen’s encephalitis (6), hemimegalencephaly (2) and cortical dysplasia (5). In 8 patients, VEEG and MEG results were concordant in localizing the epileptogenic hemisphere. Two patients had a lateralizing VEEG but MEG showed bilateral spikes. Two patients had bilateral VEEG and MEG spikes. One patient was too sedated during MEG while under general anesthesia so no spikes were recorded. Seizure outcomes were Engel Class I (10), Class II (2), and Class IV (1). In 2 patients that had an outcome other than Engel Class I, the MEG clusters concentrated in the disconnected hemisphere. The third patient had bilateral clusters.

     This is a retrospective study.

     The presence of a lateralizing MEG correlated with good seizure outcomes.

     Further studies, particularly in cases of nonconcordant MEG findings are necessary to assess the potential added value of preoperative MEG.