Abstracts

Rationale: Temporal lobe epilepsy is one of the most common forms of epilepsy to undergo surgery for those who are medically refractory. Of those who undergo surgery, 50-80% have reported to be seizure free; however, 10-20% sustain little to no benefit. Recently, our center published an evaluation of surgical failures in patients who underwent anteromedial temporal resections(AMTR)(Andrews et al., JAMA Neurology 2018). We found that those who underwent intracranial EEG with seizure spread to another anatomic region in less than 10 seconds were more likely to have poor outcomes(P<.01). Identification of noninvasive markers with similar predictive value peri-operatively could aide in better designing an intracranial electrode study to provide more therapeutic options. Methods: We identified patients who underwent an AMTR or extended temporal lobe resection between 2011-2018 at Yale New Haven Hospital with a minimum of a one-year post surgery follow up. All patients must have undergone Video EEG monitoring with habitual seizures being captured, quantitative PET imaging, and resting state fMRI. Seizures detected by scalp EEG were evaluated for seizure onset along with location and time to spread. EEGs were reviewed by two epileptologists independently (PF and JP). Onset was defined as the first ictal rhythmic change seen on EEG. Location of ictal onset was classified as localized (within a focal region of the temporal lobe (anterior, mid or posterior), diffuse (involvement of the entire temporal lobe) or lateralized (two or more lobes). Ictal spread was defined as the identical ictal rhythm that evolved at seizure onset and seen in another location on scalp EEG. It was defined to be diffuse (entire temporal lobe), lateralized (two or more lobes), or contralateral (opposite hemisphere). If intracranial EEG was completed the data was reviewed for seizure onset and spread times to other regions. Quantitative PET was visually analyzed and hypometabolism was classified as focal(located in a region of the temporal lobe (medial or lateral), diffuse (entire temporal lobe) or lateralized (two or more lobes). Surgical outcome was classified using the ILAE system. Results: Eighteen patients were analyzed. 11 male (61%) and 7 female (39%) with a mean epilepsy duration of 21.7 years. 10 patients were seizure free with ILAE 1 classification, 3 had ILAE 2 and 5 had ILAE 3. Patients with seizure spread on scalp EEG occurring in less than 6 seconds did not achieve ILAE 1 classification or seizure freedom (p=0.006). Of those patients, all with ILAE 3 were seen to have seizure spread occur in less than 6 seconds. Ictal onset location and pattern at onset did not have any association with surgical outcome. A total of 9 patients underwent intracranial EEG monitoring. Of these patients, four had ILAE 3 classification and all had spread to another anatomic location seen in less than 10 seconds on intracranial EEG. This finding is congruent to our recent published findings on temporal lobe epilepsy surgical failures. Patients with diffuse hypometabolism seen visually on quantitative PET did not achieve seizure freedom (p=0.0001). Conclusions: 72% of our patients had good surgical outcome consisting of ILAE 1(10/18) and ILAE 2(3/18). Seizure spread on scalp EEG occurring in less than 6 seconds (p=0.0006) along with diffuse hypometabolism seen on quantitative PET (p=0.0001) were associated with poorer seizure control. Based on our current analysis, scalp EEG and quantitative PET may serve as noninvasive biomarkers in predicting surgical outcomes. Larger studies and perhaps resting state fMRI may validate and further elucidate these findings. Funding: Swebilius Foundation