Abstracts

Rationale: MRI morphometry has consistently demonstrated that drug-resistant temporal lobe epilepsy (TLE) is associated with mesiotemporal and multilobar cortical atrophy (Bernhardt et al. 2010). Furthermore, diffusion MRI has shown white matter (WM) changes in numerous deep fibre tracts (Otte et al. 2012), particularly those close to the mesiotemporal lobe (Concha et al. 2012). So far, the WM immediately subjacent to the cortical mantle, i.e. the subcortical WM (SWM), has not been evaluated. Given its anatomical proximity to neocortical networks and role in maintaining long-range connectivity, the SWM may represent an important facet of TLE-related pathology. Our purpose was to map SWM diffusion properties via a surface-based framework and examine its relationship to hippocampal volume and cortical thickness.Methods: We studied 61 drug-resistant TLE patients (34±9 years, 46 males; 31/30 left/right TLE) and 38 healthy controls (30±7 years, 21 males) on a 3.0 Tesla MRI. Based on T1-weighted 3D-MPRAGE images (1x1x1mm3 isotropic voxels), we generated cortical surface models and measured cortical thickness. We computed a Laplacian potential field between the grey-white matter interface and the ventricles to guide the placement of subcortical surfaces at 2mm depth (Schüz & Braitenberg 2002). By mapping these surfaces into co-registered diffusion MRI (64 directions, b=1000 s/mm2, 2x2x2mm3 isotropic voxels), we sampled vertex-wise mean diffusivity (MD) and fractional anisotropy (FA). Hemisphere-specific measurements of left and right TLE patients were normalized with respect to controls, and sorted into ipsi- and contralateral to the seizure focus. Following correction for age and gender, we carried out vertex-wise uni- and multivariate linear models between patients and controls to detect differences in FA, MD, and their combination. We assessed correlations between diffusion indices, cortical thickness, and hippocampal volume (Kim et al. 2012). Findings were corrected at FWE<0.05.Results: Relative to controls, patients presented with increased MD and decreased FA of the ipsilateral limbic SWM, including the parahippocampal, anterior cingulate, lateral temporal, and orbitofrontal regions (Fig. 1). As diffusion abnormalities were fairly similar for FA and MD, we aggregated them into a multivariate feature map. Multivariate diffusion did not show a strong spatial or statistical relationship with cortical thinning; instead, it related to ipsilateral hippocampal volume in patients, but not controls (Fig. 2)Conclusions: We identified for the first time point-wise patterns of subcortical WM structural alterations in TLE. The predominantly ipsilateral diffusion alterations distributed within the limbic regions may reflect gliosis and/or demyelination (Garbelli et al. 2012), likely secondary to hippocampal deafferentation.