PATH-26. INTEGRATED MOLECULAR AND CLINICAL ANALYSIS OF BRAF-MUTATED GLIOMA IN ADULTS

Abstract

Abstract INTRODUCTION BRAF mutations in glioma have been recognized as a significant driver of disease in pediatric low-grade glioma, but the implications of BRAF alterations on disease trajectory and response to treatment are unknown in adult glioma. Here, we characterize a multi-institutional cohort of adults with BRAF-altered glioma. METHODS We identified patients aged ≥ 18 years with glioma containing BRAF alteration on sequencing in multi-institutional cohorts (Dana-Farber/Brigham Cancer Center, Johns Hopkins Hospital, GENIE, TCGA). BRAF mutations were grouped into three previously-defined classes: I (RAS-independent/dimerization-independent), II (RAS-independent/dimerization-dependent), III (RAS-dependent/dimerization-dependent). RESULTS We identified 198 adults with BRAF-altered glioma (median age 42 years, range 18-85 years), including 17 WHO grade I, 33 grade II, 26 grade III, 114 grade IV, and 8 others. Glioblastoma (n=115) predominated, followed by grade II-III astrocytoma (n=31), pleomorphic xanthoastrocytoma (n=18), pilocytic astrocytoma (n=12), oligodendroglioma (n=7), or other glioma (n=15). Class I mutations (BRAFV600E) comprised 49% (n=97), class II mutations 4.5% (n=9) and class III mutations 4.5% (n=9). Rearrangements occurred in 9.6% (n=19), of which canonical fusions comprised 4% (n=8), occurring primarily in pilocytic astrocytomas; copy number gains affected 13% (n=26), and other alterations 19% (n=38). All pleomorphic xanthoastrocytomas and 43.5% of glioblastoma harbored BRAFV600E. Mutation of NF1, an inhibitor of RAS/ERK signaling, was significantly associated with class II/III BRAF mutations (64.3%) and not observed in BRAFV600E-altered glioma (0%, n=62; p< 0.0001). BRAFV600E was the sole alteration in a subset of glioma with class I mutations, while others harbored multiple genomic alterations. Among 45 patients with glioblastoma and detailed clinicopathologic data, 7 received RAF-targeted therapy, with variable clinical response. Overall survival was 26 months with BRAFV600E, 21 months with class II/III mutation, and 33 months with other BRAF-alteration. CONCLUSIONS This large cohort of BRAF-altered adult glioma demonstrates a broad range of alterations, with implications for treatment sensitivity and patient survival.