Abstracts

The current accepted method for determining an individual patient[apos]s language laterality is the intracarotid amytal (Wada) test. Various patient characterisitics, such as handedness, age of onset, and gender have been investigated to examine if they contribute to language lateralization. Findings have varied across studies, but most studies have addressed these factors individually, not in combination, or have used small sample sizes.
This study used data from 426 patients who had a Wada test and for whom there was adequate data from each injection to classify the patient[apos]s language laterality, and for whom complete data was available for the independent variables side of focus, site of focus (anterior temporal, posterior temporal, frontal, parietal), handedness, age of onset, sex, and etiology (epilepsy, tumor, AVM).
The dependent variable was language laterality. Language laterality was classified in five categories: Left Only, L [gt] R, L = R, R [gt] L, or Right Only, based on the pattern of intact or disrupted expressive and receptive language during the amytal injection ([quot]Classification A[quot]). Two other classificatiion schemes were used. [quot]Classification B[quot] clustered as [quot]bilateral[quot] patients labeled as L [gt] R, L = R, and R [gt] L. [quot]Classification C[quot] clustered as [quot]bilateral[quot] only the L = R group, placing the L [gt] R group into [quot]left dominant[quot] and the R [gt] L group into [quot]right dominant[quot]. This provided three opportunities to test the predictor variables.
Stepwise multiple regression was used to determine if an independent variable or a combination of independent variables could predict the dependent variable. Variables were enetered into the stepwise multiple regression in the order from greatest contribution of varaince to least. Variables ceased to be entered when an additional variable did not account for a significant amount of the variance.
Of the 12 predictor variables, only 4 were found to significantly account for variance in language laterality. These variables, and the additional amount of variance accounted for, were: Handedness (13.4%), Side of Focus (3.5%), Posterior Temporal Site (1.8%), and Age of Onset (0.9%), for a total 19.7% of the variance. The other variables were not found to account for any additional significant variance. Re-categorization of the language laterality results using the [quot]B[quot] or [quot]C[quot] categorization approaches yielded the same sequence, though the overall amount of variance accounted for was less, at 17.5% and 15%, respectively.
Based on these findings, it seems difficult to predict language representation, and reasonable to continue to perform Wada tests.
[Supported by: An NINDS RO1 Award (NS36007) [quot]An electronic data warehouse for epilepsy research[quot] from Human Brain Program (HBP) to STCW and an NLM R29 Award (LM 06300) [quot]A multimodality neuroimaging database systems for outcomes research[quot] to KSH. ]