Abstracts

Absence of hippocampal sclerosis (HS) indicates a more [lsquo]functionally adequate[rsquo] ipsilateral memory system and higher risk of verbal memory decline after dominant anterior temporal lobectomy (ATL). Patients vary in other ways (other MRI abnormalities, EEG focality, baseline verbal memory, Wada memory results, gender, age at first risk, age at surgery) that may reflect functional adequacy and capacity of non-resected areas to mediate recovery ([lsquo]functional reserve[rsquo]). We examined whether such factors predicted decline, beyond effects of HS, speech laterality and baseline memory. Patients had ATL in the Multicenter Study of Epilepsy Surgery, a prospective, observational study. MRIs were coded for presence/absence of non-incidental findings in ipsilateral and contralateral temporal and extra-temporal regions. Scalp ictal and interictal EEG was summarized as the presence vs. absence of exclusively focal, consistent abnormalities in the ipsilateral medial temporal lobe. Memory decline was change in Long Delay Free Recall on the California Verbal Learning Test at 1-2 years post-surgery. Variables were included in a multivariate prediction model only if a univariate relationship was suggested. 36/132 patients (27%) had MRI findings other than exclusively unilateral HS (77/132 (58%)) or normal MRI (19/132 (14%)). Of these, 21/36 (16% overall) had MRI findings outside the ipsilateral temporal lobe (e.g. atrophy, contralateral HS, developmental anomaly). EEG other than exclusively focal and consistent in the ipsilateral medial temporal lobe occurred in 30 % with exclusively unilateral HS on MRI and in 48% with other MRI findings. Speech laterality (p [lt] .01), early age at injury (p [lt] .05) and unilateral HS (p [lt] .01) were associated with memory decline in univariate analyses. Those with poor baseline memory tended to be less likely to decline (p=.09). Only unilateral HS and speech laterality were significant predictors in multivariate analyses (p [lt] .01). EEG and MRI abnormalities outside the ipsilateral temporal lobe were unrelated to verbal memory decline after ATL, suggesting that [lsquo]functional reserve[rsquo] (as reflected by these indices) plays a minor role in mediating memory changes in the majority of ATL patients. As in prior studies, memory decline was more strongly related to the functional adequacy of the ipsilateral medial temporal memory system. Failure of baseline memory function and WADA memory assessment to predict decline may reflect insensitivity of single memory measures, prevalence of HS in this sample and cross-center variability in Wada methods. (Supported by RO1 NS32375 (NINDS) )