Abstracts

Rationale: Temporal lobe epilepsy (TLE) is the most common form of drug refractory epilepsy treated with surgery. While temporal lobectomy is superior to medical management, a large portion of patients go on to have their seizures recur after surgery. This study aims to characterize outcomes at a large epilepsy center to describe patients at highest risk for seizure recurrence after temporal lobectomy. Methods: A single-surgeon, single-institution cohort of temporal lobe epilepsy treated with anteromedial temporal lobectomy was retrospectively analyzed over a period between May 2011 – March 2018. Inclusion criteria were preoperative diagnosis of temporal lobe epilepsy, resectional surgery, and at least 1 year of follow up. Survival analysis was performed to analyze for predictors of Engel class I seizure freedom. Charts were reviewed for pathologic and radiologic findings. Results: Over 6.8 years at a tertiary referral center for epilepsy, 104 patients treated with temporal lobectomy by a single surgeon had at least 1 year of follow up. The most common pathologic diagnosis was mesial temporal sclerosis (n = 46), followed by non-specific gliosis (n = 27), malformations of cortical development (n = 26), neoplasm (n = 6), vascular malformation (n = 4), and 1 unspecified calcified lesion.Six of these patients had dual pathology (MCD plus MTS or neoplasm). Long term intracranial electroencephalography (icEEG) was carried out in 48 patients. Mean follow up was 2.78 years. Overall probability of Engel class I seizure-freedom at 1 year was 63%. The strongest pathologic predictor of poor outcome was non-specific gliosis (P = 0.03; Hazard Ratio 2.08 [95% Confidence Interval: 1.06 - 4.11]). Of the 27 patients showing gliosis on pathology, 19 were studied with long term icEEG. Radiologic findings of the gliosis patients included 7 normal MRI’s, 7 encephaloceles, 6 mentions of asymmetric hippocampi, 6 subtle T2 changes of uncertain significance, and 1 patient with prior craniotomy for tumor resection. Conclusions: In this single-surgeon, single-center cohort of patients who underwent surgery for temporal lobe epilepsy and had at least 1-year follow up, seizure-recurrence clustered in patients with non-specific gliosis on pathology. This sub-group sharing non-specific histologic characterization could either represent a failure to resect the true epileptogenic substrate or be indicative of a more diffuse, ill-defined pathologic process. This potentially important subgroup of pathology requires more detailed characterization for treatments to be effectively targeted toward them. Next steps will include electrographic characterization of seizure spread. Funding: No funding