Abstracts

Rationale: Temporal lobectomy is effective for many patients with refractory seizures. Outcomes have been evaluated extensively in the years immediately following surgery for epilepsy, but the literature on outcomes through 10 years of follow-up is less robust. Methods: Consecutive patients receiving an anterior medial temporal lobectomy without extratemporal resection by the senior author (DDS) from the years 2000 – 2015 were queried from the Yale Epilepsy Center database. Engel and ILAE outcomes were assigned by chart review for 1, 2, 3, 4, 5 and ≥10 year follow-ups. Pathology was determined by postoperative pathology report. Concordance of pre-operative studies (PET, EEG, MRI) was determined by consensus after review of clinical records by authors. Results: 114/127 patients had at least 1-year follow-up. The most common pathologic diagnosis was mesial temporal sclerosis (MTS) with 77, followed by neoplasms with 31.13 patients showed no distinct pathology other than gliosis. Less common pathology included malformations of cortical development (MCD), vascular malformations (VM) and infarct. Preliminary analysis shows freedom from disabling seizures in ≥75% through 10 years follow-up. Over 50% of patients whose seizures recurred after surgery recurred during post-operative year 1. Of patients who recurred in post-operative year 1, 27% lacked a defined lesion on pathology (gliosis). Kaplan-Meyer analysis of time to first disabling seizure shows a worse recurrence rate for patients lacking a defined lesion on pathology compared to those with a defined lesion (MTS, neoplasm, MCD, VM, or infarct) [P=0.024], as well as worse outcome for patients with discordant PET-EEG studies [P=0.048] and patients chosen to undergo intracranial EEG studies [P=0.0001]. Conclusions: In this retrospective, single-center, single surgeon consecutive series, mesial temporal sclerosis was the most common pathology treated with anterior medial temporal lobectomy for seizures. Preliminary analysis suggests that the majority of patients treated with anterior medial temporal lobectomy achieve stable freedom from disabling seizures through 10 years follow-up. Seizure recurrence was highest in the first postoperative year. Patients with non-lesional pathologies, discordant EEG-PET studies and those chosen for intracranial EEG studies are overrepresented among patients whose seizures recur. These predictors of post-resection seizure-recurrence are consistent with the notion that patients with distributed seizure networks are less amenable to surgery than those more easily localized to a lesion or by preoperative work-up. Funding: HHMI-CURE fellowship