Abstracts

Rationale: Object naming declines are common after left anterior temporal lobectomy (L-ATL). In a previous study, we found that language lateralization on functional MRI (fMRI) and pre-operative naming scores were predictors of post-operative naming decline in a sample of 24 L-ATL patients (Sabsevitz et al. 2003). However, this study had a small sample size, limiting the power to examine multiple predictor variables. The purpose of the current study was to replicate this finding in a larger sample of L-ATL patients, and to develop a multivariate regression model for predicting naming outcome. Methods: Seventy-two L-ATL patients had pre- and post-operative neuropsychological testing and pre-operative fMRI language mapping using a semantic decision - tone decision contrast. The outcome measure of interest was the Boston Naming Test (BNT). fMRI laterality indexes (LIs) representing asymmetry of activation volumes were calculated for each patient in five regions of interest: Whole Hemisphere, Lateral Hemisphere, Frontal, Temporal and Angular. Correlations between fMRI LIs by region of interest and BNT change scores were calculated. A regression analysis was performed to examine the unique contributions of fMRI and clinical variables for predicting naming outcome. Results: Significant correlations were found between BNT change scores and Whole Hemisphere, Lateral Hemisphere, Frontal, Temporal and Angular LIs, with the highest correlation found between BNT change and Lateral LI (r= -.423, p < .001). Preoperative BNT scores correlated with post operative naming decline (r= -.274, p=.02). There was no correlation between BNT change scores and age at onset of epilepsy or age at surgery. Regression using baseline BNT score accounted for 7.5% of the variance in predicting naming outcome. fMRI predicted significant additional variance (p=.001) beyond what was accounted for by the preoperative naming score. fMRI Lateral LI accounted for an additional 14.4% of the variance. The regression equation that best modeled the data for post-operative BNT predicted score = 10.55+(B1*-0.29)+(B2*-0.11) where B1 = BNT pre-operative score and B2 = fMRI Lateral LI. Conclusions: Language lateralization determined using a fMRI semantic decision - tone decision contrast predicts risk for cognitive morbidity (naming decline) after L-ATL in this larger sample. Funding: Funded by: NINDS R01 NS35929