Abstracts

Rationale: White matter (WM) abnormalities have been described in patients with AED refractory epilepsy, especially in temporal lobe epilepsy (TLE) associated with hippocampus sclerosis (HS) or TLE with normal MRI (TLE-NL). However, a WM alteration has not been adequately assessed in frontal lobe epilepsy with normal MRI (FLE-NL). To investigate the presence of WM abnormalities in TLE-HS, TLE-NL and FLE-NL relative to controls, by three different computational techniques: i) diffusion tensor imaging (DTI), ii) voxel based volumetry (VBV) and iii) voxel based morphometry (VBM).Methods: Ninety one subjects were included, 40 healthy controls (age 22-69; average 38), 17 TLE-HS (age 26-59; average 38), 17 TLE-NL (age 21-60; average 36) and 17 FLE-NL (age 18-59; average 35). There was no difference in gender proportion among groups. All MRIs were acquired on a Phillips 3T including a 3D T1 and a DTI with 32 directions. ExploreDTI was used for tensor calculation and fiber tractography of the inferior frontal occipital (IFO) and uncinate fasciculi (UF) due the anatomical relevance of these tracts, with a semiautomatic deterministic method (Fig.1). All tracts were visually checked and the average of the fractional anisotropy (FA) ipsilateral to the epileptogenic zone (defined by EEG recordings) were tested (ANCOVA; control vs. patient groups; age/sex as covariates). For the VBV analysis an in house script was used to calculate the ratio between WM volume (SPM8 segmentation) and total intracranial volume (TIV) for all subjects. Two sample T-tests were performed for group comparison. On the VBM analysis, each group had their images lateralized in order to align all hemispheres ipsilateral to the epileptogenic zone and two sample T-tests of control vs. patient groups were applied (p<0.01, FWE corrected).Results: DTI parameters for both IFO and UF tracts only differed for the TLE-HS (p values<0.0001 and 0.002 for IFO and UF respectively) compared to the control group (Table 1). The VBV analysis showed significant reduction of WM/TIV ratio in the TLE-HS (p=0.0004) and FLE-NL (p=0.003), but no difference for TLE-NL (p=0.06). VBM showed diffuse WM abnormalities in TLE-HS and FLE-NL patients with most significant atrophy observed at ipsilateral insula (Fig. 2-A) and inferior frontal gyrus, respectively (Fig. 2-B). In TLE-NL, the VBM showed only a focal WM reduction at the cerebellar tonsil (Fig. 2-C).Conclusions: Microstructural WM abnormalities (DTI analysis) in the IFO and UF tracts were observed in TLE-HS but not in TLE-NL and FLE-NL patients. However, macro-structural alterations (VBV and VBM analyses) were observed in both TLE-HS and FLE-NL patients in contrast to the TLE-NL patients that had no abnormalities. The different patterns of WM alterations in these groups of patients emphasize that a complex interaction of distinct factors might lead to WM atrophy in patients with refractory localization-related epilepsies.