INNV-19. MULTIMODAL THERAPY FOR BRAF V600E MUTANT MELANOMA BRAIN METASTASES: CHARACTERIZATION OF CHRONOLOGICAL TREATMENT STRATEGIES AND ASSOCIATIONS WITH OVERALL SURVIVAL

Abstract BACKGROUND B-raf proto-oncogene (BRAF)-mutant melanoma brain metastases (BRAF MBM) are associated with poor prognosis. Amidst our large institutional experience treating patients with BRAF MBM with multimodality therapy, our aim was to identify disease and treatment-related prognostic features, accounting for longitudinal sequence of events, to inform clinical decision-making. METHODS We retrospectively reviewed electronic health records of 100 patients with primary cutaneous melanoma diagnosed with BRAF MBM. Demographic, disease, and clinical factors including treatment sequencing were tabulated. Descriptive statistics were used to summarize the study population and non-parametric tests were used to evaluate differences across groups. A multivariable Cox regression identified factors associated with overall survival (OS) following BRAF MBM diagnosis. RESULTS Median follow-up was 22 months and median OS was 33.5 months. Patients received an average [standard deviation (SD)] of 3.36 [1.06] unique therapies. After adjustment for covariates, greater number of brain metastases at initial diagnosis (hazard ratio (HR) 2.3 [1.14-4.62], p=0.01 for 3-10 metastases and HR 2.54 [1.01-6.36], p=0.04 for >10 metastases, respectively), receipt of combination IO and targeted therapy prior to BRAF MBM diagnosis (HR 2.91 [1.08-7.86], p=0.03), and receipt of no unique subsequent therapy after initial management strategy (HR 3.89 [1.48-10.16], p=0.005) were significantly associated with worse OS. CONCLUSIONS Despite heterogeneity in individual treatment sequences, patients with greater intracranial disease burden and those receiving combination targeted and immunotherapy prior to BRAF MBM diagnoses had the worst outcomes. Dynamic changes in therapy resistance may inform optimal treatment strategies for individuals with BRAF MBM.