Interaction of cognitive aging and memory deficits related to epilepsy surgery

Abstract

Temporal lobe epilepsy surgery can cause significant memory impairment. This study was intended to examine whether surgery also could affect prognosis of memory in older age. Age regression of verbal memory was examined in 187 patients (before and 1 year after left temporal lobe surgery) and 264 healthy controls. Eighty patients underwent selective amygdalohippocampectomy, and 107 patients underwent anterior two-thirds temporal lobectomy. Amygdalohippocampectomy patients had mesiotemporal epilepsy; anterior two-thirds temporal lobectomy patients had more extramesial or diffuse seizure onset zones. Memory was assessed by word list learning for its more mesial (consolidation/retrieval) and more neocortical (learning) aspects. Patients showed significant preoperative memory impairment. Independent of seizure outcome and surgical approach, surgery had significant negative effects on learning and consolidation/retrieval. In the amygdalohippocampectomy group, preoperative and postoperative age regressions of learning and consolidation/retrieval were not different from those of controls. In the anterior two-thirds temporal lobectomy group, age regression of verbal learning became steeper after surgery, and consolidation/retrieval was negatively correlated with older age and later onset of epilepsy even before surgery. The data confirm that age regression of verbal memory in left temporal lobe epilepsy is similar to that in healthy controls. Both left anterior two-thirds temporal lobectomy and amygdalohippocampectomy worsen verbal learning and memory and bring patients closer to cognitive disability. Particularly in anterior two-thirds temporal lobectomy patients, surgery and reduced capacities for compensation cause acceleration of lifetime memory decline. The results support earlier and tailored epilepsy surgery and suggest that memory prognosis in older age should be considered if more extensive temporal resections would be inevitable.