Abstracts

Rationale: Patients who undergo left (dominant) temporal lobe resection (LTLR) for treatment of pharmacoresistant epilepsy are often at risk for postoperative declines in language and verbal memory. While this risk is often moderated by MRI evidence of left mesial temporal sclerosis (MTS), there are a subset of patients with left MTS who show postoperative declines in language and verbal memory. The objective of this study was to identify factors that predict cognitive decline after LTLR in patients with left MTS.Methods: 126 adults with left temporal lobe epilepsy and left MTS completed language (Boston Naming Test) and verbal memory (Wechsler Memory Scale-Third Edition) measures before and approximately 6 months after LTLR (including resection of mesial structures) as part of comprehensive neuropsychological evaluations. Patients with evidence of atypical language dominance or incomplete cognitive data were excluded. For each cognitive measure, patients were categorized into one of two groups (decline, no decline) using established reliable change indices (RCIs). First, the proportion of decliners was examined on each of the cognitive measures. Then, a series of binary logistic regression analyses were used to identify factors that predict cognitive decline following LTLR. The following variables were included as predictors in these analyses: bilateral MTS on MRI, age at seizure onset, age at preoperative assessment, and preoperative test performance.Results: 57 patients (45%) showed clinically meaningful decline in confrontation naming following LTLR. Age at seizure onset and preoperative naming performance predicted postoperative decline on the Boston Naming Test with correct classification in 69% of patients. Specifically, older age at onset and higher preoperative naming ability were associated with greater risk for naming decline. While not as prevalent, decline in verbal memory was also observed among a subset of patients in the study [Auditory Immediate Memory=31 patients (25%), Auditory Delayed Memory=11 patients (9%)]. Preoperative ability was the best predictor of postoperative decline in Auditory Immediate Memory (75% correctly classified) and Auditory Delayed Memory (92% correctly classified), again with higher scores conferring greater risk.Conclusions: Although MTS moderates cognitive risk following LTLR in many patients, there remain a sizeable proportion of patients who demonstrate postoperative decline in language and verbal memory. Preoperative cognitive ability and age at seizure onset are important factors to consider when assessing and counseling patients with left MTS about the risk for cognitive decline following LTLR.