Abstracts

Rationale: Hippocampal sclerosis (HS) patients show impaired executive functions and language. fMRI and WADA studies suggest language reorganization in epilepsy patients. These studies are limited by patient heterogeneity. Executive functions have been less studied with fMRI in epilepsy patients. We assessed brain activation patterns in patients with left and right (HS), in language-related executive function tasks in a phonologic and in a semantic decision-paradigmsMethods: We studied 23 left and 16 right HS patients and 22 healthy controls (C). All subjects were right-handed, aged between 18 and 55 years, with at least eight years of education. Patients underwent WADA testing, that documented left hemisphere dominance. Subjects performed a phonologic (rhyme) and a semantic (classes) decision-making. Behavioral data were recorded (two choice forced decision, via button pressing) synchronized with stimulus presentation and image acquisition. Data were acquired with 3T MRI scanner with compressed GRE EPI BOLD images, (40ACPC oriented slices, 3.3mm isotropic voxels, TR=2s, TE30ms FA90). Data were processed with FSL package (Oxford University, UK), submitted to motion correction, spatial and temporal filtering, and normalized to MNI152 space. Statistical inference was nonparametric. Between groups analysis was performed with ANOVA (LHSxRHSxC) with a p<0.05 significance level. Results: In the phonemic decision task fMRI activation pattern analysisfor the left HS group showed reduced BOLD response compared to controls in the left inferior and middle frontal, left posterior and lateral orbital, left superior temporal and left post-central gyri, as well as, left insula and bilateral inferior parietal lobule, while RHS showed decreased activation, compared to C in a large number of clusters, comprising bilateral superior temporal gyri, right inferior frontal, right supramarginal and angular gyri, left pre-central gyrus, precuneus, left hippocampus, bilateral pericalcarine cortex, bilateral cerebellum, right occipito-temporal cortical transition and right post-central gyrus. In the semantic decision task LHS showed reduced BOLD response in the left middle and inferior frontal gyri compared to C, while RHS showed reduced BOLD response compared to C in similar areas as in the phnemic decision task, with additional minor differences: BOLD effect was also reduced in the middle frontal gyri, and decreased BOLD effect was not noted in the temporal lobes, pericalcarine cortices and left hippocampus. Conclusions: Network disruption of language-based decision paradigms showed distinct patterns for LHS and RHS. While LHS only displayed reduced BOLD effect in traditional language areas, RHS failed to produce BOLD response in an extended network, which may be more related to executive dysfunction. Further analyses should be performed to investigate correlation of reduced activation patterns and performance in language and executive function tests in RHS and LHS.