Abstracts

Rationale: Epilepsy surgery provides patients with refractory epilepsy an opportunity for seizure freedom. Intracranial EEG, the gold standard of epileptogenic zone identification, is required prior to surgical resection or ablation when noninvasive data is inconclusive. Historically, patients with diffuse or poorly localized seizure onsets on scalp EEG were not considered ideal surgical candidates; extensive bilateral coverage with subdural strips was necessary, with poor coverage of deep regions. In contrast, stereotactic EEG (sEEG) allows for the sampling of bilateral hemispheres in three dimensions in a potentially less invasive fashion. This technique may allow evaluation for some patients who appear to be poor surgical candidates. We examined sEEG findings and surgical outcomes in patients with non-localizable seizure onsets on scalp EEG. Methods: Between June 2011 and May 2016, 61 patients with refractory epilepsy underwent sEEG at our comprehensive epilepsy center in order to determine whether a seizure focus could be identified and resected. Of these patients, 7 had poorly localized epilepsy, defined as non-localizable seizure onsets on scalp EEG in the absence of a spatially discrete candidate lesion on neuroimaging. Demographic information, seizure data, imaging data, presurgical hypotheses, sEEG electrode counts and locations, sEEG results, surgery performed, and surgical outcomes were collected and analyzed. Results: The mean patient age was 32 (range 17-47) years with a mean 20 (range 3-40) years of seizures and having failed a mean 7 (range 4-11) antiepileptic drugs. Five patients had normal MRIs; 3 patients had widespread unilateral or bilateral abnormalities. PET results were normal in 2 patients and discordant with MRI findings in 5 patients. SPECT scans were concordant in 2/6 and nondiagnostic in 4/6 scans obtained. A mean 10 (range 4-22) seizures were captured on scalp video EEG monitoring; results were non-localizable in all patients. All patients underwent sEEG using an average of 130 depth electrode contacts (range 60-200). Between 3 and 12 subdural strips were also placed in 5 patients. A median of 21 seizures (range 6-51) were captured. Seizure foci were identified in 6/7 patients: 3 frontal, 1 insular, and 2 mesial temporal; in 5/6 patients, onsets were seen in depth electrodes. Two patients had bilateral onsets (one bifrontal; one temporal and parietal). Two patients required additional depth electrode placement to confirm the epileptogenic zone; 1 patient required grid placement. All 6 patients with localizable onsets underwent resection or laser ablation. Of the original 7 patients, 5/7 (71%) [or 5/6 (83%) of those treated] were seizure free at median follow-up of 12 months (range 1-34 months). Conclusions: sEEG is a safe, effective method to identify seizure onset zones in patients with poorly localized scalp EEG onsets. Using sEEG data to identify the epileptogenic zone and guide resection or ablation can afford an excellent chance of seizure freedom to patients traditionally considered to be poor surgical candidates. Funding: None