Mapping language in patients undergoing brain surgery

Itzhak Fried (University of California Los Angeles, Tel-Aviv University)

Rarely is the relationship between language and medicine more acute and critical than in the operating theatre, where the neurosurgical patient undergoing surgery at the vicinity of the brain language areas is kept awake and cooperative while brain mapping techniques are applied to delineate the individual language map for this patient. This procedure determines the ability to remove a brain lesion while preserving the language capabilities of the patient. We have developed a special anesthetic technique, asleep-awake-asleep (AAA) anesthesia, where the patient emerges out of general anesthesia during the operation to be able to perform complex language tasks awake.

The "gold standard" method of mapping cortical language areas, involves the application of electrical stimulation at the surface of the cortex while the patient, awake and cooperative, performs various language tasks. This method identifies sites where stimulation evokes errors in naming, reading, sentence comprehension, and other language tasks. Maps vary considerably among individuals, reflecting both inter-subject variability in cortical organization and reorganization due to the presence of a brain lesion. Mapping of bilingual patients yields different maps for different languages in the same individuals, but usually with some overlap.

While distribution is often wide spread within the language-dominant hemisphere, the following areas are particularly prone to yield stimulation-evoked disturbances in language functions: (1)The frontal operculum extending to dorsolateral frontal lobe (2)The region of posterior part of superior and middle temporal gyri (3) The posterior basal temporal region (4) The inferior parietal lobe. (5) The anterior supplementary motor area (SMA). Following electrical stimulation mapping, resection often proceeds as the patient performs continuous language tasks in order to detect any change in performance.

When these results are compared with mapping of language with functional magnetic resonance imaging (fMRI), discrepancy exists between the two methods. fMRI is reliable in identifying the hemisphere dominant for language function. However, often the map obtained with electrical stimulation is a subset of the map obtained by various fMRI measures. It appears that fMRI mapping of language function is useful as preoperative guide, pointing to the dominant hemisphere and to regions within it, which should be further explored with intraoperative electrical stimulation.