Preoperative standardised neuropsychological testing plus MRI is superior to WADA testing in predicting verbal memory outcome in left TLE patients after removal of the hippocampus

Purpose: To identify the best predictors of postoperative verbal memory decline in patients undergoing left temporal lobe excisions for epilepsy.

Methods: A retrospective analysis was carried out on data from 59 patients with left mesial temporal lobe epilepsy (TLE) who underwent en-bloc resection. Memory was assessed by a standard word list-learning task (z-score for total number of words over 5 trials) administered before and after surgery. For the MR images, a new rating was devised to characterize the extent of hippocampal damage. Atrophy and signal change of the hippocampus were examined separately. Both changes were assessed on a scale: 0=no change 1=moderate atrophy/signal change, 2=severe atrophy/signal change. The average of these two factors was calculated to classify the degree (range 0 to 2) of mesiotemporal sclerosis (MTS). The Wada results included number of correctly identified items after ipsilateral injection. A multiple regression analysis was performed with the change in memory score as the dependent variable.

Results: Postoperatively, 35% of the patients showed a memory decline indicated by a z-score loss of more than one point. The regression analysis revealed that the pre-operative memory test score was the strongest predictor of memory change (p<0.05, adjusted R Square=0.341). Higher preoperative performance in the verbal memory task (verbal learning) was associated with a greater risk of verbal memory decline after surgery. The second significant predictor was the MTS classification (p<0.05, adjusted R Square=0.428). The greater the damage to hippocampus, the less decline seen following surgery. None of the other variables, including the Wada results, were significant predictors of memory outcome.

Conclusion: The Wada test results did add to the prediction of verbal memory decline in this sample of patients in whom temporal lobe resection was considered safe with regard to the risk of amnesia (i.e., adequate memory functioning following ipsilesional injection). Thus, the clinical data obtained routinely before surgery seems to be sufficient to predict risk of memory decline in the majority of TLE patients. However, our findings do not refute the usefulness of the Wada test for excluding from surgery those patients who are at risk for amnesia after temporal lobe resection.