Abstracts

RATIONALE: Patients undergoing temporal lobectomy for medically-refractory seizures may be at risk for memory decline following surgery. We examined how frequently memory decline occurred and determined the relative contributions demographic and neuropsychological factors made to predicting the extent of memory change following temporal lobe resection.
METHODS: 72 patients were administered the Wechsler Memory Scale-Third Edition (WMS-III) before and after surgery. 34 underwent left (LTL) and 38 underwent right (RTL) temporal lobectomy. The percentage of LTL and RTL patients showing a decline in their Auditory Immediate (AI), Auditory Delayed (AD), Visual Immediate (VI) and Visual Delayed (VD) memory was determined. A series of regression analyses investigated the relative importance of side of surgery, discrepancy between auditory and visual memory at baseline, baseline memory score, age at seizure onset, gender, and handedness in predicting pre- to post-surgery memory change.
RESULTS: Following LTL, a higher percentage of patients demonstrated declines in auditory (68% AI, 62% AD) than visual (47% VI, 59% VD) memory. Conversely, a greater percentage of RTL patients showed declines in visual (61% VI, 50% VD) than auditory (32% AI, 42% AD) memory.
Strong auditory relative to visual memory at baseline was associated with greater decline in auditory memory; this baseline discrepancy was the best predictor of change in AI and AD scores. Baseline auditory memory scores did not improve the prediction of outcome once this discrepancy was taken into account. The second best predictor of auditory memory outcome was side of surgery, with LTL patients at greater risk for decline than RTL patients. Finally, an earlier age of onset corresponded to less decline in AI scores.
The best predictor of VD memory decline was higher visual relative to auditory memory at baseline. In contrast, baseline VI scores accounted for more variance in VI change than the auditory/visual discrepancy. Side of surgery did not significantly contribute to predicting change in visual memory, but left-handers demonstrated more decline in VI scores than their right-handed peers.
CONCLUSIONS: Results indicate that both LTL and RTL patients are at risk for memory decline following surgery. Relative, rather than absolute, memory performances at baseline were generally better predictors of memory outcome, and side of surgery was important for predicting auditory, but not visual, memory decline. These results provide information that could play a critical role in clinical decision making when considering the relative merits and risks of surgical intervention.