127 Multimodality Word-Finding Distinctions in Pediatric Cortical Stimulation Mapping.

INTRODUCTION: Recently, auditory naming has become a part of cortical stimulation mapping (CSM) to provide a comprehensive language map prior to resection in epilepsy patients. Modality-specific language sites have been found using CSM in adult epilepsy patients, but research in the pediatric population is limited. Here, we demonstrate distinctions between visual and auditory modalities and identify where errors are most likely to occur in pediatric patients. METHODS: A series of 23 pediatric epilepsy patients (14 female; mean age 12.6 + 3.5 years, 6-18 years) underwent CSM by using visual (n = 23) and auditory (n = 16) naming paradigms. Mixed-effects logistic regression was used with the response being the indicator of an error on an individual trial. We adjusted for patient-specific effects using random effects, and included fixed effects for Region, Modality, interactions between Modality and Region, and age <14 years. RESULTS: No evidence was found to support differences in errors due to age. Statistically significant differences were found by Modality, with auditory naming having a 1.43 higher odds of identifying errors (95% confidence interval [CI], 1.10-1.87, P = .0085) compared with visual naming. Probabilities of visual and/or auditory naming errors varied significantly among regions (P = .002). The highest probability of detecting an error was in the PolMTG under Auditory and Visual modalities (0.48 and 0.39, respectively). CONCLUSION: The auditory naming modality is more sensitive in detecting errors than visual naming in pediatric epilepsy patients. The PolMTG shows the most naming errors regardless of age or modality, suggesting that CSM mapping with both modalities is necessary to obtain a comprehensive language map before resection.

Duke Scholars

Author Gerald Arthur Grant Neurosurgery

Author Merlise Clyde Statistical Science

Author Carrie Rebecca Muh Neurosurgery