Abstracts

Rationale: In patients with intractable temporal lobe epilepsy (TLE), cross-sectional MRI studies have shown that longer duration of epilepsy is associated with decreased mesiotemporal volumes (Bernasconi et al., Neurology 2005; 65(2):223-8). Studies relating neocortical atrophy to the duration of epilepsy, however, have not yielded conclusive results (Moran et al., Brain 2001; 124(1):167-75). In longitudinal designs, sensitivity to detect subtle changes increases, as they take a subject as its own baseline. A previous study that assessed a heterogeneous sample of patients with newly diagnosed and chronic epilepsy of unknown duration failed to demonstrate significant neocortical atrophy (Liu et al., Epilepsia 2005; 46(9):1482-94). Our purpose was to map neocortical atrophy in a homogeneous sample of patients with medically intractable TLE patients using a longitudinal MRI-based cortical thickness analysis. Methods: We studied 18 patients (12 males, 30 ± 11 years) with medically intractable TLE. The seizure focus was left-sided in 8 patients and right-sided in 10. Duration was 18 ± 13 years. Based on MRI volumetric assessment, we classified patients into hippocampal atrophy (n=12) and normal hippocampal volume (n=6). Forty-two serial MR scans (3D T1-spoiled gradient echo sequence; TR = 18; TE = 10; voxel size = 1mm3) with at least 2 scans per subject were available. The mean interval between the first and last scan was 31 ± 21 months (range = 7 - 90 months). Cortical thickness analysis was performed using an unbiased surface-based deformation algorithm (Kim et al., NeuroImage 2005; 27(1):210-21). We fitted linear fixed-effects models on thickness data and tested for a negative effect of time from the baseline scan. Results: We found a progressive decrease in mean cortical thickness in the hemisphere ipsilateral (-0.016 ± 0.009 mm/year; t = -2.20, p < 0.02) and contralateral to the seizure focus (-0.022 ± 0.009 mm/year; t = -2.88, p < 0.01). Annual rates of cortical atrophy exceeded 0.05 mm/year in bilateral prefrontal, fronto-central, insular as well as contralateral temporal and posterior cingulate regions. Vertex-wise analysis revealed progressive cortical thinning (p < 0.005, uncorrected) in bilateral frontal (orbitofrontal and superior frontal), postcentral, and temporal (ipsilateral temporopolar and contralateral lateral temporal) areas, as well as in contralateral insular and posterior cingulate (p < 0.0001) regions (see Figure).