P11.08.A resection of glioma from classic Broca's area is feasible due to peritumoral and distant functional language reorganization.

BACKGROUND
Historically, Broca’s area has been assigned a critical role in motor language functions. Although there has never been consensus about its precise anatomical definition, most experts consider the left pars triangularis and pars opercularis as the core of classic Broca’s area. Electrocortical mapping and functional MRI (fMRI) routinely find naming and speech sites in this region. However, case reports and small series have indicated that gliomas can often be resected from Broca’s area without new or permanent language deficits, suggesting that in these patients there is functional reorganization due to tumor growth. To test this hypothesis, we compared the organization of the language system in patients with a glioma that encompassed classic Broca’s area (‘Broca group’) to patients with a glioma outside of Broca’s area (‘non-Broca group’).
MATERIAL AND METHODS
We retrospectively selected 13 patients from a prospective dataset of consecutive glioma patients (2010-2024) that had infiltration of left pars triangularis (n=13) and pars opercularis (n=8). There was additional infiltration of the middle frontal gyrus (n=4), insula (n=4), superior temporal gyrus (n=1) or ventral premotor cortex (n=1). All of these Broca group patients were operated under awake conditions with electrical stimulation mapping to delineate motor and language areas. Standardized preoperative MR imaging for all glioma patients included a language fMRI.
RESULTS
Broca group patients had either an oligodendroglioma (7 low grade, 1 high grade) or astrocytoma (3 low grade, 2 high grade). Clinical debut was a seizure (n=10), mild language (n=1) or cognitive (n=2) deficits. In 4 patients there were mild clinical language deficits prior to surgery. In 11 patients there were temporary language deficits after surgery (4 due to seizures) that resolved over time to initial clinical performance. All Broca group patients had a language dominant left hemisphere as deducted from clinical symptomatology and intraoperative mapping results. In none of these patients, electrocortical mapping found functional (language) sites over the tumor part in the pars triangularis and pars opercularis, and these areas were consequently resected. fMRI language maps of Broca versus non-Broca patients (n=276) showed lower activation in ipsilesional inferior frontal gyrus but higher activation in contralesional inferior frontal gyrus.
CONCLUSION
Language is represented differently in patients with a glioma in Broca’s area as compared to non-Broca glioma patients. Peritumoral and distant functional reorganization likely explains why gross total resection in classic Broca’s area is feasible and safe.