Background and objective Epilepsy surgery within the temporal lobe of the language dominant hemisphere bears the risk of postoperative verbal memory decline. As surgical procedures have become more tailored, the question has arisen, which type of resection within the temporal lobe is more favourable for memory outcome. Since the hippocampus (HC) is known to play an essential role for long-term memory, we examined whether HC sparing resections help to preserve verbal memory functions.

Methods We retrospectively analysed neuropsychological data (prior to and 1 year after surgery) of patients undergoing either HC sparing resections (HC-S, N=65) or resections including the hippocampus (HC-R, N=62).

Results Prior to surgery, the HC-R group showed worse memory performance as compared to HC-S patients. Both patient groups revealed further deterioration over time, but in verbal learning HC-R patients demonstrated a stronger decline. Predictors for verbal learning decline were left-sided surgery, better preoperative performance, higher age at surgery, hippocampus resection, and lower preoperative IQ. In patients with spared HC, resection of the left-sided parahippocampal gyrus was rather accompanied by a decline in verbal learning performance. For visual memory, better preoperative performance best predicted deterioration after surgery. Seizure outcome was comparable between the two groups (HC-S: 66%, HC-R: 65% Engel 1a).

Conclusions Temporal lobe resections within the language dominant hemisphere can be accompanied by a decline in verbal memory performance, even if the HC is spared. Yet, HC sparing surgery is associated with a benefit in verbal learning performance. These results can help when counselling patients prior to epilepsy surgery.