Multimodal protocol for awake craniotomy in language cortex tumour surgery.

BACKGROUND: Intra-operative neurophysiological language mapping has become an established procedure in patients operated on for tumours in the area of the language cortex. Awake cranial surgery has specific risks and patients are exposed to an increased physical and mental stress. The aim of the study was to establish an algorithm that enables tailoring the neurosurgical and anaesthetic techniques to the individual patient. METHOD: A total of 25 patients underwent awake craniotomy for intra-operative language mapping between 1999 and 2004. Following craniotomy under analgesia and sedation without rigid pin fixation of the head, cortical language mapping was performed in the fully co-operative patient. The results of functional magnetic resonance imaging and of cortical language mapping were incorporated into the 3D dataset for neuronavigation. Depending on the functional data and the individual operative risk tumour resection then proceeded either under conscious sedation with the option of subcortical language monitoring or under general anaesthesia. FINDINGS: After cortical language mapping patients are assigned to one of four groups: BACC (Berlin awake craniotomy criteria) I-IV. BACC I (9 patients): adequate functional data+operative risk not increased-->tumour resection in the awake patient; BACC II (4 patients): limited functional data+operative risk not increased-->tumour resection in the awake patient with the option of language monitoring as needed; BACC III (9 patients): adequate functional data+increased operative risk-->tumour resection under general anaesthesia using functional navigation; BACC IV (3 patients): limited functional data+increased operative risk-->tumour resection in the awake patient with the option of language monitoring as needed. We observed less adverse events in group BACC III. No permanent deterioration of language function occurred in this series. CONCLUSIONS: The multimodal protocol for awake craniotomy provides for tumour resection under general anaesthesia in selected patients using functional neuronavigation. Our experience with the algorithm suggests that it is a useful tool for preserving function in patients undergoing surgery of the language cortex while reducing the operative risk on an individual basis.