Abstracts

Rationale: Neurosurgical resection of brain tumors or epileptogenic foci often necessitates intraoperative functional mapping of proximal language areas in the awake state using electrical cortical stimulation (ECS) to reduce postoperative language deficits. In this procedure, pairs of electrodes are stimulated through subdural electrocorticographic (ECoG) grids that are placed over the cortical area of interest. Each ECS train is time-locked to a cued language task and repeated several times for each electrode pair. This requires a high level of compliance from the patient and several minutes for each pair of electrodes to complete. This limits the benefit of intra-operative mapping of language areas to those patients that are highly functional and able to endure this procedure. Recently, task-related electrocorticographic mapping (ECoG) and functional magnetic resonance imaging (fMRI) have been suggested as alternatives to ECS, however their integration into established neurosurgical procedures has not been explored yet. In this first study, we investigate how ECoG analysis can complement ECS and fMRI during awake resective neurosurgery in the OR.Methods: We investigated this question in a patient with new-onset seizures that were caused by a brain tumor in very close proximity to expressive language cortex (Broca’s area). The surgical treatment was comprised of two surgeries separated by a 7-day stay at the epilepsy-monitoring unit. Prior to the first surgery, the patient underwent a comprehensive pre-surgical evaluation including anatomic and functional MRI. During the first surgery, we implanted a 64-contact ECoG grid with 10-mm spaced electrodes to localize the epileptogenic foci. At the epilepsy-monitoring unit we used ECS and ECoG analysis to delineate the epileptogenic foci from functional language areas. During the second surgery, we implanted a high-density 64-contact ECoG grid with 3-mm spaced electrodes over the previously localized functional language areas. We reversed the anesthesia and the patient performed a simple cued verb generation task while we recorded and analyzed the ECoG signals in near real-time.Results: Our near real-time ECoG signal analysis in the OR further refined the boundary between pathologic tissue and functional language areas. The subsequent neurosurgical resection of the pathologic tissue used these boundaries. The treatment was considered clinically successful as the patient has not had any evidence of tumor recurrence over 28 months of surveillance and did not acquire any permanent aphasic deficit. Our post-hoc co-registration of the fMRI, ECoG and ECS revealed qualitative concordance across modalities (see Figure 1).Conclusions: Our results show that combining extra- with intra-operative functional mapping achieves an excellent surgical margin and postoperative functional outcome. This suggests that, in cases of pathology bordering eloquent cortex, a multi-modality approach, inclusive of novel ECoG methods, may serve to improve the precision of functional localization.