Abstracts

The success of epilepsy surgery hinges on the accurate identification of a seizure focus. Relatively few studies have examined the potential of functional magnetic resonance imaging (fMRI) in lateralizing the side of seizure focus. The purpose of this study was to compare the clinical utility of two different fMRI hippocampal activation protocols and Wada memory testing in identifying the side of seizure focus in patients with temporal lobe epilepsy (TLE). Twenty-seven TLE patients (15 left [amp] 12 right) who underwent Wada memory testing and preoperative fMRI were included in this study. Wada memory testing closely modeled the procedures of Loring et al. (1992). Wada memory asymmetries (WMAs) were derived by subtracting memory scores obtained by the right hemisphere (inject left) from scores obtained by the left hemisphere (inject right). Whole brain fMRI was conducted using two activation protocols that have previously been shown to produce bilateral medial temporal lobe activation in healthy individuals. In the definition naming (DN) task, patients heard a series of spoken definitions and were required to name aloud the concept to which the definition refers. In the scene encoding (SE) task, patients viewed photographs and had to decide whether they represented indoor or outdoor scenes. Appropriate control tasks were used for each activation paradigm. Region of interest (ROI) volumes were created for the anterior and posterior hippocampus. Laterality indexes (LI), reflecting the interhemispheric difference between voxel counts in the left and right ROI[apos]s, were calculated. WMA scores were significantly correlated with both the anterior (r = .665, p [lt] .001) and posterior (r = .433, p = .026) DN fMRI LIs. No significant correlations were found between WMA scores and the SE task or between the DN and SE task (p[apos]s [gt] .05). Using the anterior hippocampal ROI and an optimal LI cutoff score to predict side of seizure focus, the percent of patients correctly classified was: 78% (21/27) with the DN LI, 63% (17/27) with the SE LI, and 85% (22/27) with the WMA score. A Chi-square analysis revealed both the DN LI and WMA score were significantly better than chance (p [lt] .05) in predicting side of seizure focus whereas the SE LI was not (p [gt].05). Consistent with previous findings, our results show that WMA scores are predictive of side of seizure focus in TLE patients. This is the first study to use a fMRI DN paradigm for lateralizing relative memory functions in pre-operative TLE patients. This task appears to be a better measure of lateralized memory functioning than a visual scene encoding task and is relatively equivalent to Wada memory testing in its ability to accurately predict side of seizure focus. (Supported by NINDS RO1 NS35929)