Abstracts

Rationale: Patients with drug-resistant temporal lobe epilepsy (TLE) should be considered for epilepsy surgery. Standard treatment for TLE is an anterior temporal lobectomy (ATL), but minimally invasive surgery has become a viable alternative, such as stereotactic laser amygdalohippocampotomy (SLAH). Predictors of surgery outcome after ATL have been identified, but less is known about predictors of outcome after SLAH. The most consistent predictor of good outcome after SLAH is the presence of mesial temporal sclerosis (MTS). Using a cohort of 101 patients who underwent SLAH at Emory University, we hypothesized that clinical variables other than MTS would independently predict outcomes.

Methods: Predictors of good outcome after ATL (Fitzgerald et al., 2021) include (1) pre-operative seizure frequency, (2) absence of generalized tonic-clonic seizures (GTCs), (3) presence of MTS, (4) unitemporal interictal epileptiform discharges (IEDs), and (5) well localized seizures. We collected these variables, as well as (6) unitemporal PET abnormality, (7) gender, (8) age of onset, and (9) age at surgery. We then used univariate logistic regression to see how well each clinical variable could predict surgical outcome (Engel I or not). We then compared two methods for performing multivariate logistic regression. First, we performed forward selection of the most significant variable, adding variables which remained significant (at p< 0.05).  Second, we performed backwards selection, removing the least significant variable and selecting the combination with the optimal Akaike Information Criterion (AIC)._x000D_