Abstracts

RATIONALE: To assess hemispheric distribution of language function by fMRI in patients with intractable left temporal lobe epilepsy (TLE) who underwent Wada test, and to evaluate whether left hemisphere dysfunction due to epilepsy could be associated with change in language lateralization.
METHODS: 10 right-handed TLE patients (5 males, age 22-48y) and 10 right-handed healthy controls (4 males, age 24-3y) were scanned in a 1.5T GE MR scanner while performing an auditory-verb-generation task. Two regions of interest (ROIs) were defined: Broca[ssquote]s area (inferior frontal gyrus, IFG) and Wernicke[ssquote]s area (posterior superior temporal gyrus, STG). The number of activated voxels in each ROI was subject to multiple analysis of variance (ANOVA). In addition, an individual Lateralization Index ((LI=(L-R)/(L+R)) was calculated based on number of activated voxels within each ROI. In addition, all patients underwent a Wada test.
RESULTS: In all patients, Wada test revealed left language lateralization. In both patients and healthy control groups, fMRI showed an overall left lateralization based on number of activated voxels and LI measure. However, lateralization was overall less pronounced for patients relative to controls (2-way interaction of Hemisphere by Group, p[lt]0.05). Measuring regional lateralization demonstrated that for both groups Broca[ssquote]s area was more lateralized to the left than Wernicke[ssquote]s area. However, patients showed less of a difference in activation between Broca[ssquote]s area and the homologous area on the right, and almost no difference between Wernicke[ssquote]s area and its right hemisphere counterpart (3-way interaction of Region by Hemisphere by Group, F(1.18)=5.0, p[lt]0.05). The reduced lateralization of Broca[ssquote]s area in patients was mainly due to reduced left hemisphere activation (planned comparison p[lt]0.05).
CONCLUSIONS: 1) fMRI revealed more language lateralization in Broca[ssquote]s than in Wernicke[ssquote]s area in all subjects. Thus, lateralization by Wada test was more consistent with fMRI measures obtained in IFG than with STG.
2.)Left temporal epileptic focus was associated with overall less language representation in the left hemisphere. Surprisingly, this change in laterality was mainly due to reduced activation in the left Broca[ssquote]s area. This finding suggest greater suceptability of the left frontal language region to intractable, long-standing left temporal lobe epilepsy.