Abstracts

Rationale: Rationale : Temporal lobe epilepsy with hippocampal sclerosis is a chronic disease which affects neural networks way beyond hippocampal sclerosis. The latter changes have received much recent interest with the advent of diffusion tensor imaging (DTI). Several DTI-studies showed alterations of large white matter tracts, most notably in left mesial temporal sclerosis (LMTS). The way in which this affects deafferentation of specific cortical structures has so far only indirectly been examined. Our study intended to compare whole brain connectomes of patients with left and right MTS with healthy control subjects in order to delineate deafferentation of major cortical and subcortical structures. Methods: We have gathered a large data base on a homogeneous group of patients with unilateral sclerosis, comprising 20 patients with right and 19 patients with left sided lesions and 28 matched control subjects. We recorded DTI sequences with 50 directions and performed whole-brain probabilistic fiber tracking using the MRtrix software package. Based on the Destrieux atlas implemented in freesurfer, 164 regions of interest in cortical and subcortical gray matter were segmented and coregistered with the DTI maps. We then computed individual graphs based on these 164 nodes. The edges between two regions of the graph were weighted by the average FA along the selected tracts, the latter weight representing the connectivity strength. Finally, we determined the connectivity strength in the patient's groups as compared to control subjects were determined with three complementary analyses : individual comparison, group permutation tests and Network Based Statistics. Results: The three lines of analysis yielded a homogeneous picture of deafferentation in both left and right MTS. The decrease of connectivity with respect to control subjects was much more pronounced in left than in right MTS. Furthermore, the left MTS pattern was strongly lateralized to the ipsilateral temporal lobe with a special impact on temporal cortex and temporal pole, hippocampus and collateral sulcus. Conversely, connectivity in right MTS was globally less affected and essentially focused on bilateral hippocampus and collateral sulcus with a right-sided predominance. Conclusions: Connectome analysis in left and right MTS corroborates a clear cut decrease in connectivity strength in left MTS, which is moreover strongly lateralized and focused on temporal lobe and hippocampus. Right MTS shows lesser and more bilateral effects. Yet, deafferentation in both patients groups targets hippocampus and collateral sulcus, a subregion often altered in MTS (Kim et al., 2008, Bernhardt et al., 2008).