Abstracts

Rationale: The most commonly described pathology in temporal lobe epilepsy (TLE) is hippocampal atrophy. However, there is evidence for more widespread damage in the mesial temporal lobe and the limbic network involving the amygdala, the entorhinal cortex, the fornix and the thalamus. In patients with TLE with hippocampal sclerosis (HS), analysis of diffusion tensor imaging (DTI) data provides some indirect radiologic evidence of microstructural changes. We investigated whether the volume changes in mesial TLE could be correlate with water diffusion abnormalities by means of voxel-based DTI analysis. Methods: In 12 left and 7 right mesial TLE (mTLE) patients with HS, we performed manual volumetric measurement of amygdala, hippocampus, fornix, thalamus and entorhinal cortex on both sides. Mean diffusivity (MD) and fractional anisotropy (FA) data of each patient were then quantified and analyzed by statistical correlation method using SPM8. In this study, we used voxel-based DTI analysis to detect water diffusion abnormalities correlated with regional volume reduction in mTLE patients with HS. Results: Increased MD and reduced FA value in the bilateral anterior thalamic nucleus were correlated with the regional volume reduction of left amygdala in left TLE patients (p<0.001). Increased MD and reduced FA value in the bilateral inferior frontal lobe and uncus were correlated with the regional volume reduction of left hippocampus, and reduced FA value of left amygdale was correlated with the regional volume reduction of ipsilateral hippocampus in same patients (p<0.001). In addition, decreased left thalamic volume had the significant correlation with reduced FA value of the prefrontal gyrus, cingulate gyrus and uncus in left TLE cases (p<0.001). But, those were not observed in right TLE patients. There was no significant region correlated between DTI parameters and the regional volume change of EC in both side of TLE patients. Conclusions: Our results demonstrate that water diffusion abnormalities of bilateral anterior thalamic nucleus is correlated with the left amygdala atrophy, but isn t correlated with the hippocampal atrophy. The finding presented here, namely, that the thalamus can be atrophic even in the absence of hippocampal atrophy, provides further evidence that structural damage in the anterior thalamic nucleus occur dependently from amygdala atrophy without hippocampal atrophy. This could be considered the mechanism of TLE without HS. Furthermore, this provides a possibility that TLE without HS is heterogenous or entity different from TLE with HS.