Prognostic role of functional neuroimaging after multilobar resection in patients with localization-related epilepsy

Abstract

To investigate the usage of functional neuroimaging as a prognostic tool for seizure recurrence and long-term outcomes in patients with multilobar resection, we recruited 90 patients who received multilobar resections between 1995 and 2013 with at least 1-year follow-up (mean 8.0 years). All patients were monitored using intracranial electroencephalography (EEG) after pre-surgical evaluation. Clinical data (demographics, electrophysiology, and neuroimaging) were reviewed retrospectively. Surgical outcomes were evaluated at 1, 2, 5 years after surgery, and at the end of the study. After 1 year, 56 patients (62.2%) became Engel class I and at the last follow-up, 47 patients (52.2%) remained seizure-free. Furthermore, non-localized 18F-fluorodeoxyglucose positron emission tomography (PET), identifying hypometabolic areas not concordant with ictal onset zones, significantly correlated with seizure recurrence after 1 year. Non-lesional magnetic resonance imaging (MRI) and leftsided resection correlated with poor outcomes. In the last follow-up, non-localized PET and left-sided resection significantly correlated with seizure recurrence. Both localized PET and ictal-interictal SPECT subtraction co-registered to MR (SISCOM) predicted good surgical outcomes in the last follow-up (69.2%, Engel I). This study suggests that PET and SISCOM may predict postoperative outcomes for patients after multilobar epilepsy and shows comparable long-term surgical outcomes after multilobar resection.