Abstracts

Rationale: In temporal lobe epilepsy (TLE), language abilities may depend on both the structural and functional integrity of frontotemporal brain networks. Given the reported higher discordance between fMRI and the Wada language laterality results in patients with "atypical" (i.e., bilateral or right) language dominance, this study examined the relationship between multimodal brain imaging data and language performance in patients with typical and atypical language dominance. Methods: Twenty-four patients with TLE (14 left; LTLE and 10 right; RTLE) and 24 healthy controls completed structural MRI, diffusion tensor imaging (DTI), language functional MRI (fMRI), and neuropsychological measures of visual naming, category fluency (CF) and letter fluency (LF). For DTI, mean diffusivity (MD) and fractional anisotropy values of four frontotemporal white matter tracts were examined. For fMRI, the number of supra-threshold voxels (p < .005) was counted within the inferior frontal (F) and posterior temporal (T) language regions. Cortical thickness was generated for the same F and T regions. Laterality indices (LI = [L-R]/[L+R]) were then calculated between the right and left hemisphere variables for all MRI measures. Those who showed an fMRI LI below .2 were categorized as atypical language dominance. Results: Patents with TLE produced fewer words than healthy controls on LF (omnibus ANOVA p = .004). In addition, patients with LTLE performed lower than controls on visual naming and CF (p values < .05). MD values of the inferior fronto-occipital fasciculus (IFOF, p = .018) and uncinate fasiculus (p = .001) were significantly more left lateralized in patients with LTLE compared to healthy controls, indicating greater damage to the white matter in the left relative to right hemisphere. In terms of fMRI language lateralization, patients with LTLE (fMRI LI = .25) showed significantly reduced laterality in F activity relative to healthy controls (fMRI LI = .57, p = .02) indicating more bilateral language lateralization in LTLE. There were no group differences in cortical thickness of the F and T regions. In patients with LTLE who showed atypical language dominance, greater leftward laterality of MD of the inferior longitudinal fasciculus (ILF) and IFOF was associated with better performance in LF (r = .85, r = .84, respectively), while no such an association was found for individuals with typical language lateralization (see Figure). No other relationships among variables were found. Conclusions: Our results show that greater left-than-right MD of two frontotemporal fiber tracts is associated with better verbal fluency performance, but only in patients with atypical language dominance. Although the cross-sectional design does not allow an inference of causal effect, these data suggest a complex relationship between language reorganization and language performances in patients with LTLE and indicate that successful language performance may depend on concomitant ipsi and contralateral reorganization of structural and functional brain networks.