Objective In the context of discussions on the optimal cognitive outcome of temporal lobe epilepsy surgery, and stimulated by recent reports on the beneficial effects of a selective subtemporal approach to memory function, this study evaluated the cognitive consequences of subtemporal versus transsylvian selective amygdalohippocampectomy (SAH) in patients with mesial temporal lobe epilepsy, taking verbal/figural memory and language functions into account.

Methods We contrasted cognitive outcomes of 26 subtemporal SAH patients with those observed in a transsylvian SAH control group. The surgical groups were pairwise matched with regard to clinical and demographic characteristics. Preoperative and 1 year postoperative memory and language evaluations served as within group factors, and surgical approach (transsylvian vs subtemporal) and side of surgery (right vs left) as between group factors.

Results Both surgical approaches caused decline in verbal memory to a similar degree. Differential effects were seen with regard to decline in verbal recognition memory (more affected by left transsylvian SAH) as well as in figural memory and verbal fluency (more affected by subtemporal SAH).

Interpretation Different from previous optimistic reports, this study demonstrates that subtemporal surgery, such as transsylvian surgery, poses similar risks for verbal memory. Differences between the approaches appear to reflect the effect of different collateral temporal lobe lesions due to the approach. Different cognitive outcomes across studies on the subtemporal approach are discussed as being in part due to study design and the chosen dependent functional measures.