gms | German Medical Science

Objective: Due to the inter-individual variance of language functional anatomy, risk prediction based on anatomical data alone is insufficient in language area related brain tumor surgery. Repetitive navigated transcranial magnetic stimulation (rnTMS) is an evolving tool for locating speech eloquent areas preoperatively. The aim of our study is to provide a reliable method for locating speech eloquent areas using rnTMS in order to improve patient counseling, surgical planning and neurological outcomes in speech eloquent tumors.

Methods: We prospectively examined 15 patients with brain tumors in speech-related areas, who underwent preoperative rnTMS examination and direct cortical stimulation (DCS) during awake surgery. In both cases their task was to promptly name 80 objects which were used for TMS mapping as well as for the intraoperative mapping. Stimulation spots were documented in the same MRI data set. TMS and DCS language positive spots were compared in respect to their spatial correspondence. Moreover, based on video recordings the TMS and DCS mappings were compared based on the type of error: five linguistic qualities were assessed: no response, semantics, hesitation, phonology, performance.

Results: The comparison of preoperative rnTMS and intraoperative DCS (ground truth) observations revealed a low sensitivity (37%) for positive DCS spots and a moderate specificity (76%) for predicting negative DCS spots. The highest overlap in respect to the type of linguistic quality was seen in hesitation errors (31% of TMS induced hesiations were reproduce intraoperatively by hesitations). All other error types showed a lower correlation ranging from 9% to 18%. In total 21% of the positive speech spots showed no overlap at all.

Conclusion: nrTMS speech mapping and direct cortical brain stimulation are sensitive tools for detecting speech eloquent areas. DCS is still considered the gold standard for the localization of speech function. However, analysis of the patterns of fiber connectivity to TMS positive spots may improve value and applicability of TMS in preoperative language mapping.