Objective: Post-operative memory decline is one of the feared complications of temporal lobe epilepsy surgery. In the current study we investigate the relation between post-operative memory and postictal memory and the predictive value of postictal testing in temporal lobe epilepsy surgery patients.

Methods: 66 patients with pharmacoresistant unilateral left- or right-sided temporal lobe epilepsy (L-TLE, R-TLE), who subsequently had their medial temporal lobe structures resected, underwent interictal, postictal and post-operative (3-9 months) verbal and non- verbal memory evaluation. Verbal and non-verbal delayed recall (VDR and NVDR) rates were tested using Rey’s auditory verbal and visual learning tests (15 items repeated 5 times and recalled after 40 min). During postictal testing a 10-item sets were given 3 times and recalled after 30 min.

Results: VDR for interictal 72.3±22.7%, postictal 53.5±32.2%, post-operative 67.6±25.1%. Post-operative VDR was significantly lower in L-TLE (55.5±26.3% vs. 79.1±15.9%; p<0.001). A significant decline in postictal compared to interictal VDR was observed in L- TLE (68.2±24.9% to 45.2±34.5%; p=0.001), but not R-TLE patients (75.9±19.1% to 61.0±28.3%; p=0.07). Postictal VDR significantly predicted the post-operative VDR (B=0.263, p=0.01, CI [0.059; 0.467]). Postictal decline predicted a 33% post-operative decline with 70% sensitivity (negative predictive value NPV = 90%) in the whole group and a 26% decline with 78% sensitivity (NPV=92%) in patients with hippocampal sclerosis (n=40). We found no significant differences between NVDR scores.

Conclusions: Postictal verbal memory can lateralize the epilepsy focus. Postictal memory testing helps to predict post-operative memory and could have additional value in predicting moderate to severe verbal memory decline after epilepsy surgery.