The effect of immunotherapy on brain metastasis patients remains incompletely understood. Our goal was to evaluate its effect on survival, neurologic death, and patterns of failure after stereotactic radiosurgery (SRS) without prior whole-brain radiation therapy (WBRT) in patients with lung and melanoma primaries metastatic to the brain. We performed a retrospective analysis of 271 consecutive lung or melanoma patients treated with upfront SRS for brain metastases between 2013 and 2018. Of these patients, 101 (37%) received immunotherapy and 170 (63%) did not. Forty-three percent were treated with nivolumab. Thirty-seven percent were treated with pembrolizumab. Fifteen percent were treated with ipilimumab. One percent were treated with a combination of nivolumab and ipilimumab. One percent were treated with atezolizumab. Three percent were treated with another immunotherapy regimen. Survival was estimated by the Kaplan–Meier method and cumulative incidences of neurologic death, and local and distant brain failure were estimated using death as a competing risk. The median overall survival (OS) of patients treated with immunotherapy vs without was 15.9 (95% CI: 13.3 to 24.8) vs 6.1 (95% CI: 5.1 to 8.8) months (P < .01). The 1-year cumulative incidence of neurologic death was 9% in patients treated with immunotherapy vs 23% in those treated without (P = .01), while nonneurologic death was not significantly different (29% vs 41%, P = .51). Median brain metastasis velocity (BMV) did not differ between groups, and rates of salvage SRS and WBRT were similar. The use of immunotherapy in patients with lung cancer or melanoma metastatic to the brain treated with SRS is associated with improved OS and decreased incidence of neurologic death.

Background/Objectives: No prior studies have attempted to identify a biomarker for initial brain metastasis velocity (iBMV), with limited studies attempting to correlate genomic data with the development of brain metastases. Methods: Patients with non-small-cell lung cancer (NSCLC) who underwent next-generation sequencing (NGS) were identified in our departmental database. iBMV was calculated by dividing the number of BMs by the interval of time between primary cancer and BM diagnosis. Two-sample t-testing was used to identify mutations statistically associated with iBMV (p < 0.1). A value of +1 was assigned to each mutation with a positive association (“deleterious genes”), and a value of −1 to each with an inverse association (“protective genes”). The sum of these values was calculated to define iBMV risk scores of −1, 0 and 1. Pearson correlation test was used to determine the association between iBMV risk score and calculated iBMV, and a competing risk analysis assessed for death as a competing risk to the development of BMs. Results: A total of 312 patients were included in the analysis, 218 of whom (70%) developed brain metastases. “Deleterious genes” included ARID1A, BRAF, CDK4, GNAQ, MLH1, MSH6, PALB2, RAD51D, RB1 and TSC1; “protective genes” included ARAF, IDH1, MYC, and PTPN11. iBMV risk scores of 1, 0 and −1, predicted an 88%, 61% and 65% likelihood of developing a BM (p < 0.01). A competing risk analysis found a significant association between iBMV risk scores of 1 vs. 0 and 1 vs. −1, and the likelihood of developing a BM using death as a competing risk. Overall survival (OS) at 1 and 2 years for patients with iBMV risk scores of 1, 0 and −1 was 72% vs. 84% vs. 85% and 46% vs. 69% vs. 70% (p < 0.02). Conclusions: Development of a genomic signature for iBMV via non-invasive liquid biopsy appears feasible in NSCLC patients. Patients with a positive iBMV risk score were more likely to develop brain metastases. Validation of this signature could lead to a biomarker with the potential to guide treatment recommendations and surveillance schedules.