Abstracts

Rationale: Epilepsy surgery is superior to medical therapy in refractory temporal lobe epilepsy; however, in many cases the decision algorithm for pre-operative evaluations, including invasive implantation design, is generally based on empirical and center-specific logistics. There are few data available describing stereo-EEG (SEEG) implantations and seizure localization in patients with temporal lobe epilepsy who have non-lesional MRI and normal fluorodeoxyglucose-PET (FDG-PET) metabolism. Methods: We included all patients with ictal semiology and/or scalp ictal EEG suggesting temporal lobe epilepsy, who had non-lesional brain MRI and normal FDG-PET, and then underwent SEEG evaluations at our institution. Patients with previous resective neurosurgeries were excluded. We described SEEG localization results and concordance of the scalp EEG and SEEG ictal onsets. Results: Among 137 SEEG patients screened, 20 patients met inclusion and exclusion criteria. In this patient cohort, the SEEG localized seizure onsets were concordant with scalp EEG ictal onsets on lobar level in 13/20 (65%) patients, 5/20 (25%) patient had discordant SEEG and scalp EEG ictal onsets, and in 2/20 (10%) scalp EEG ictal onset was classified as non-localizing and non-lateralizing. Within the concordant group, 7/13 patients had unilateral scalp EEG ictal onset and unilateral mesial temporal SEEG ictal onset, 2/13 patients had independent bilateral temporal scalp EEG seizures, and bilateral mesial temporal seizures captured with SEEG, 1/13 patient had bilateral temporal scalp EEG seizures, and bilateral neocortical temporal seizures recorded with SEEG, 1/13 - unilateral temporal scalp EEG and unilateral neocortical SEEG seizures, 2/13 - unilateral scalp EEG and unilateral mesial and neocortical temporal SEEG ictal onsets. Among the 5/20 patients who had discordant scalp EEG and SEEG localization, 4 patients had bilateral mesial temporal SEEG ictal onsets despite of unilateral temporal scalp EEG ictal findings, and one patient had bilateral fronto-central scalp EEG seizures, but unilateral mesial temporal ictal SEEG onset. Among two patients with non-localizing and non-lateralizing ictal scalp EEG, one patient had unilateral mesial temporal ictal onset localized by SEEG, the second patient had unilateral mesial and neocortical temporal, and extra temporal (ipsilateral superior frontal sulcus) ictal onsets captured with SEEG. Conclusions: Our study re-demonstrated that temporal lobe epilepsy could be a complex disease related to broad epileptogenicity within and outside the temporal lobe(s). Thus, we propose that in presence of non-lesional MRI and normal PET it requires bilateral invasive evaluation to localize the ictal onset zone(s). Ultimately, larger studies are needed to replicate these findings and create guidelines for the role and design of invasive evaluation in MRI and PET non-lesional temporal lobe epilepsy surgery candidates. Funding: No funding