Treatment outcomes in melanoma with brain metastases in the era of immunotherapy: An analysis of the National Cancer Database.

Abstract

e14513 Background: Brain metastases (BM) are a major cause of morbidity and mortality in metastatic melanoma. Radiotherapy (RT) and/or surgery are the standard of care for treating brain metastases. The introduction of immune checkpoint inhibitors has drastically improved disease outcomes since approved in 2011 and recent trials have exhibited intracranial activity. We aim to assess the added survival benefit of immuno-oncologic agents (IO) for melanoma with BM in National Cancer Database. Methods: We queried the database from 2011 to 2015 for Stage IV melanoma with BM. RT treatments were divided into stereotactic radiosurgery (SRS), whole brain radiotherapy (WBRT), or no brain radiation; those with insufficient documentation were excluded. Patients were further divided into receipt of IO or no receipt of IO; those without documentation were excluded. Concurrent (CC) treatment was defined as IO begun before or during RT treatment; sequential (SQ) therapy was defined as IO begun after RT. Survival analyses were performed by Kaplan-Meier method and log-rank test. Results: Among 8,561 metastatic melanoma patients with documentation of metastatic sites, 3,008 (35.1%) had documented BM. Median overall survival (mOS) for all stage IV melanoma was 10.5 months with worse survival for those with BM compared to those without (mOS 6.0 v 14.7 m; p < 0.01). Among 3,008 cases with survival data available, 2,762 met criteria for survival analysis of BM treatments. Table summarizes treatment outcomes between cohorts by RT and IO treatment categories. IO + SRS had improved survival compared to all other treatment groups (mOS 19.9 m; p < 0.01). No difference in survival between SRS and IO alone was observed (mOS 10.9 v 11.5 m; p = 0.56). WBRT as single modality did not appear to improve survival over no therapy (mOS 4.2 v 3.3 m; p = 0.35). 436 cases had documented time course between IO and RT. IO was given SQ with RT in 69% of SRS (n = 158/230) and 67% of WBRT (n = 137/206) treated patients. In SRS + IO cohort, survival was longer for SQ compared to CC treatment (23.5 v 13.1 m; p < 0.01). Survival was similar in WBRT for SQ v CC groups (6.8 v 7.9 m; p = 0.62). Conclusions: Superior survival is realized for those who receive IO in addition to SRS for BM. IO and SRS as single modality have similar outcomes, while WBRT seems to add little survival benefit to IO or no therapy. [Table: see text]