Abstracts

Rationale: Due to advanced presurgical imaging methods, intracranial EEG studies are less and less utilized. However in many patients the seizure onset zone cannot be reliably identified without an intracranial study. We retrospectively examined a large cohort of patients who underwent an intracranial EEG study and present here their long-term seizure outcomes. Methods: At the Dartmouth Epilepsy Program, 223 intracranial investigations in 209 patients were performed since 1992. The clinical characteristics of intracranial electrode arrays, subsequent epilepsy surgeries, MRI findings and pathologies were tabulated in a standardized database. Outcomes in resective and non-resective cases after intracranial EEG evaluation were determined using Engel classification. Only patients with reliable follow up information for 12 months or more were included in the study. Decision for performing an intracranial EEG was based on discordant presurgical data or non-lesional epilepsy, during a multidisciplinary conference. Chi-square analysis was used for statistical analysis. Results: There were 183 intracranial investigations in 171 patients who had follow up periods of 12 months or more. Twelve patients had two intracranial studies. Sixty-six patients had non-lesional, MRI negative epilepsy. One hundred and thirteen patients had neocortical epilepsy. Implants included a mean of 72 (between 24 and 140 contacts) intracranial contacts. Forty-nine patients had interhemispheric electrodes and 81 implants included depth electrodes. The mean follow up period was 71.5 months (12 months-15 years). After 151 of the 183 intracranial EEG studies a resective surgical procedure was performed in 138 (80.7%) of the 171 patients. In patients with no resection (n=33), the seizure onset zone was either bilateral (n=18), non localizable (n=5), or in eloquent cortex (n=10). Forty-eight (28.1%) patients had frontal lobe seizure onset, 85 (49.7%) patients had temporal lobe epilepsy, and 26 (15.2%) patients had seizures arising from the posterior cortex. 52.9 % (n=73) of patients with resective procedures had an Engel Class I outcome, 17.4 % Class II, 15.9 % Class III and 13.8 % Class IV. If no resective procedure could be performed only 15.2 % had a Class I outcome (P < 0.001) with VNS implantation alone, callosotomy or best medical management. Fifteen patients had a VNS implanted after resection failed. Of the non-lesional patients (n=44) with resective procedures, 61.4% (n=27) had a Class I outcome. Of the 84 patients with neocortical epilepsy and a resective procedure, 36 (42.8%) had a Class I and 15 (17.4%) had a Class II outcome. Conclusions: Intracranial EEG recordings if aggressively pursued still have a major role in achieving excellent outcomes in epilepsy surgery.