Abstracts

RATIONALE: Factors predicting surgical outcome are important for presurgical planning and patient counseling. MRI evidence of mesial temporal sclerosis (MTS) is one of the best prognostic factors for good surgical outcome after temporal lobe resection. However, up to 30% of patients with MTS have persistent disabling seizures after surgery. We report our experience with using magnetoencephalography (MEG) to predict surgical outcome. METHODS: Seven consecutive patients with temporal lobe epilepsy and MTS documented by MRI were evaluated with a whole-head 122-channel biomagnetometer. Interictal spikes were identified by off-line analysis, and single-dipole spherical head source modeling algorithms were used for spike localization. MEG spikes were classified as either anteromedial or posterolateral in their temporal lobe location prior to any surgical intervention. All patients subsequently underwent intracranial EEG and/or surgical resection. The MEG spike locations were then correlated to surgical outcome based on the Engel Classification. Mean follow-up time is 16 months. RESULTS: Four of four patients with anteromedial MEG spikes achieved an Engel Class I/II outcome. One of three patients with posterolateral MEG spikes achieved an Engel Class I/II outcome. One patient with posterolateral MEG spikes had 10/10 seizures originating from the sclerotic hippocampus on intracranial EEG, underwent anterior temporal lobectomy and had a Class IV outcome. CONCLUSIONS: Anteromedial MEG spike locations may be associated with better surgical outcomes than posterolateral spikes. More patients are needed to validate this observation. If validated, this observation suggests electrophysiological differences exists between MTS patients with good versus poor outcomes.