Abstracts

Rationale: To study the patterns of pre and post operative grey matter atrophy (GMA) and white matter atrophy (WMA) atrophy in patients with MTLE who became seizure free and those who did not. We aimed to evaluate evidences of brain plasticity after seizure control. Methods: We performed a whole brain VBM study, followed by a region of interest (ROI) analyses of 69 controls (mean age 34.2 years [SD 11.2]) and 67 patients (34.1 years [SD10.5]) with unilateral MTLE who underwent surgery (34 Engel IA (Seizure-free group), 23 Engel IB-IIA and 10 Engel III-IV (Failure group). The VBM-ROI analyses were used to confirm areas detected on the whole brain VBM. We used Mricro software (www.mricro.com) to preprocess and flip to the left side all MR scans with right hippocampal atrophy, to avoid right to left analysis cancellation. We created a new routine for SPM2 (www.fil.ion.ucl.ac.uk) to perform VBM analysis including individual masks for the surgical lacuna and used the software package MARSBAR (http://marsbar.sourceforge.net) to extract the mean GM and WM from pre defined regions of interest (ROIs), according to Automatic Anatomic Labeling (AAL) ROI Library. With these ROIs we can improve the statistical power over voxel-by-voxel analyses since we reduce the number of statistical comparisons. We used T-test and paired T-test. Results: Groups were balanced for age (p=0.95) and gender (p=0.7), with mean follow up of 60.2+-30.7months. Seizure-free group: On baseline MRI, there was significant GMA in the ipsilateral temporal lobe, hippocampus and in both frontal lobes (p<0.0001), occipital lobe and basal ganglia (p=0.02). WMA involved bilateral temporal, frontal, cingulum and parietal lobes (p<0.0001). After surgery, we identified areas with relative increase (compared to preoperative scans) of GM on contralateral basal nuclei (p=0.0001), temporal lobe (p=0.00001) and bilateral cingulum (p=0.00004). Areas with relative increase of WM were identified on bilateral frontal, cingulum, temporal (p<0.0001) and occipital lobes (p=0.001). Failure group: On baseline MRI, there was bilateral and widespread GMA on temporal lobes (p=0.003), bilateral hippocampus (p<0.0001), basal ganglia (p=0.001), frontal (p=0.002) and parietal lobes (p=0.001). The pattern of WMA was similarly widespread as in the seizure-free group, involving both hemispheres. All areas above were first detected on the whole brain VBM and confirmed on the ROI-VBM. After surgery we did not identify areas of relative increase of WM or GM in the whole brain VBM in the failure group. Conclusions: Our results suggest that different patterns of baseline GM atrophy between seizure-free group and failure group may be involved in the persistence of seizures after surgical treatment in MTLE. The postoperative significant increase of WM and GM in seizure-free group suggests that successful surgery may offer better chances of favorable brain plasticity in patients with MTLE.