Intraoperative language mapping in multilingual patients with gliomas

Surgical removal of lesions located close to or in areas of the brains mediating speech requires the use of intraoperative techniques to localize which cortical areas and subcortical tracts have those functions. Localization of speech is particularly problematic in patients that are usually fl uent with different languages. It has been reported that in bilingual patients multiple and separate areas of the cortex mediate the different languages. Other investigations pointed out that in bilingual patients common areas of the brain are responsible for those functions. We report here five cases of patients harboring a left frontal glioma who were proficient with three to five different languages in which a multiple language brain mapping was undertaken during awake craniotomies for tumor removal. They were 3 males and 2 females, age ranging from 34 to 61 years. Language profi - ciency was tested by submitting patients to confrontation tests for each language during the pre-operative examination. Each language was tested serially starting from the fi rst acquired language. Language mapping was undertaken during asleep-awake craniotomy, by the use of an Ojemann cortical stimulator and by using the largest current that did not produce afterdischarge (from 3.5 to 6 mA) during counting and confrontation naming tests. Subcortical stimulation by using the same current threshold was also applied during tumor resection, in a back and forth fashion. Our data showed that each language has separate and distinct cortical centers. Cortical areas for fi rst acquired language had a larger cortical representation, whereas those for the secondly acquired languages were localized in more distinct cortical sites. Subcortical stimulations found tracts for the fi rst acquired language in 4 patients, whereas those for the other languages in 3 patients. Subcortical tracts for the fi rst language had a larger representation. Three patients experienced a decrease in fluency immediately after surgery, mainly affecting the fi st acquired language, which fully recovered in two patients in two months and partially in one. These findings support the existence of language-specifi c cortical sites and the concept that intraoperative mapping should be performed for all the languages the patient is fluent for, to maximally preserve functional language integrity. In addition, confrontation naming test is more accurate than counting for localization of functional areas.