Abstracts

Rationale: Memory disturbance is a major complicating feature of temporal lobe epilepsy (TLE). While hippocampal dysfunction undoubtedly contributes to memory problems in TLE, the influence of extrahippocampal limbic structures critical for memory performance has not been systematically examined. The goal of the present study was to elucidate how the structural organization underlying memory function differs among individuals with right and left TLE, as compared to control subjects.Methods: Structural MRI data (3T Philips) were obtained in 27 patients with drug-resistant TLE (right TLE=11; left TLE=16; Age=37.4 11.7 years; Gender=16 Females) and 20 healthy controls (Age=32.0 11.3 years; Gender=11 Females). Hippocampal subfields (CA1, CA2-3, CA4-dentate gyrus), presubiculum, subiculum, thalamus, and anterior and posterior cingulate cortex were derived with Freesurfer segmentation (http://surfer.nmr.mgh.harvard.edu). Mammillary bodies were manually segmented with atlas guided anatomical landmarks (intrarater ICC = 0.92). Verbal memory was assessed with the Rey Auditory Verbal Learning Test and visual memory was measured with the Rey Complex Figure Test. Stepwise multiple regression analyses were performed to determine which cortical structures best predicted verbal and visual memory functioning among the three groups. Results: Different anatomical profiles distinguish verbal and visual memory performances of right and left TLE groups from healthy controls (See Table). In healthy controls, bilateral hippocampal subfield and left cingulate cortex volumes were the best predictors of verbal and visual memory scores. In the left TLE group, the left mammillary body volume was significantly associated with immediate verbal memory (R^2=.213, p=.041), whereas the right mammillary body volume was significantly related to immediate visual memory (R^2=.215, p=.046). The left posterior cingulate cortex accounted for additional variance in delayed visual memory score for the left TLE group (R^2=.215, p=.040). In the right TLE group, the left presubiculum volumes predicted verbal memory performance (immediate memory R^2= .839, p<.001; delayed memory R^2=.437, p=.022), with the left CA1 hippocampal subfield also contributing to immediate verbal memory scores. Finally, the left anterior cingulate cortex volume predicted immediate visual memory scores in the right TLE group (R^2=.677, p=.002). Conclusions: Memory performance in individuals with TLE is associated with structural integrity of the limbic network outside of the hippocampus, while healthy controls memory functioning had a greater link to volumes within the hippocampus. Further, volumes of the mammillary bodies differentiated verbal and visual memory performances in the left TLE group, with left side volumes related to performances in verbal memory and right side volume associated with visual memory scores. These findings suggest that the memory network undergoes structural reorganization in TLE, with greater reliance on extrahippocampal structures, possibly to compensate for hippocampal deficits.