PL01.5.A ASSOCIATION BETWEEN RESTING-STATE-NETWORK FUNCTIONAL CONNECTIVITY AND SURVIVAL IN RECURRENT GLIOMAS DIAGNOSED USING FET PET

Abstract BACKGROUND Functional connectivity between glioma-infiltrated brain tissue and remote cortical regions has been linked to overall survival (OS) in patients with newly diagnosed glioblastoma. Here, we studied the association of functional connectivity between tumor and resting-state networks and OS in patients with recurrent gliomas. PATIENTS AND METHODS The study included 82 patients (age range, 26-81 years; median ECOG performance score, 0) with recurrent gliomas (CNS WHO grade 4 glioblastoma, n=57; CNS WHO grade 3 and 4 astrocytoma, n=12; CNS WHO grade 2 and 3 oligodendroglioma, n=13) who had undergone first-line therapy by current guidelines. In addition to the RANO criteria for anatomical MRI, diagnosis of glioma recurrence was confirmed by amino acid PET using the tracer O-(2-[18F]-fluoroethyl)-L-tyrosine (FET). FET PET and resting-state functional MRI were simultaneously acquired on a 3T hybrid PET/MR scanner. A seed-based dual regression analysis was performed to determine the functional connectivity between metabolically active recurrent tumors and a set of seven canonical resting state networks. RESULTS Average functional connectivity was significantly higher (p=0.028) in oligodendrogliomas (z-score, 6.9±1.9) and astrocytomas (z-score, 5.5±1.3) than in glioblastomas (z-score, 4.9±2.6) and was inversely related to the tumor grade (p=0.031). Univariate Cox regression analysis revealed a significant positive association between OS and tumor connectivity with the visual, somatomotor, and dorsal attention network (all p8.3, n=21) had a median OS of 31.4 months, while those with the lowest functional connectivity (z-score