IMG-25. Glioma induced intratumoral and peritumoral functional network alterations correlate with pathology, brain network dysfunction, clinical symptoms, and prognosis

Abstract Gliomas interact dynamically with brain networks, disrupting normal connectivity while integrating into neural circuits, which may influence symptoms and outcomes. This study examined 306 glioma patients from two centers using resting-state fMRI and individualized parcellation to characterize intratumoral and peritumoral functional networks. Tumors were segmented into core and peritumoral regions, and functional connectivity was assessed with respect to the Yeo 17 networks. The ratio of intratumoral functional regions to core tumor volume (RIFR) and peritumoral tumor network ratio (RPTR) were quantified and analyzed in relation to glioma pathology, molecular alterations, clinical symptoms, and outcomes. Intratumoral functional regions were found in all subtypes, with higher RIFR associated with lower tumor grade and IDH mutations. Conversely, patients with lower RIFR had more severe cognitive and language deficits. Higher RIFR was also linked to disruptions in network features, including increased local efficiency and clustering coefficient but decreased global efficiency. Survival analysis revealed that patients with higher RIFR and RPTR had significantly better outcomes. Prognostic models incorporating these functional metrics achieved high predictive accuracy for two-year survival (AUC=0.88-0.90 in external validation). Our findings highlight that intratumoral and peritumoral functional networks are not merely passive bystanders but active players reflecting glioma biology and network remodeling, with implications for cognitive and language outcomes. This integrative approach to individualized functional mapping underscores the potential of functional connectivity markers in refining glioma prognosis and advancing personalized treatment strategies.