Background Management of advanced non-small cell lung carcinoma (NSCLC) has been transformed by PD-1/PD-L1 immune checkpoint inhibitors (ICI), with FDA approvals in 2015 (second-line) and 2016 (first-line). Despite ~40% of NSCLC patients developing brain metastases, these patients were disproportionately excluded from the pioneering ICI trials. Thus herein we evaluate the overall survival (OS) associated with ICI in NSCLC brain metastases nationally.

Methods Patients newly-diagnosed with stage 4 NSCLC, including brain metastases, from 2010–2016 were identified from the National Cancer Database (comprising >70% of all newly-diagnosed cancers in the U.S.) Landmark survival analysis was used to address immortal time bias. Post-approval, median time from diagnosis to ICI was 58 days, and this timepoint was selected for all landmark survival analyses (OS estimated by Kaplan-Meier technique, and compared by logrank test and multivariable Cox regression) and for multivariable logistic regression to identify predictors of ICI utilization.

Results 50,858 patients presented with advanced NSCLC that involved the brain: representing 27.6% of all newly-diagnosed stage 4 cases. Following initial FDA approvals in 2015, ICI use in brain metastasis patients rose from 7.2% in 2015 to 12.7% in 2016. OS for NSCLC brain metastasis patients diagnosed post-approval (i.e. 2015, median 6.3 months, 95% [confidence interval] CI: 6.0–6.6) was substantially better than those diagnosed pre-approval (median 5.5 months, 95%CI: 5.4–5.7, p<0.001) and, in fact, than those diagnosed in 2014 (median 5.9 months, 95%CI: 5.6–6.1, p=0.002). Among patients diagnosed post-approval (in 2015, n=7,431), ICI receipt demonstrated substantially improved OS in landmark survival analyses (median 13.8 months, 95%CI: 12.2–15.1; vs. 8.5 months, 95%CI: 8.3–8.9, p<0.001) – benefits which persisted in multivariable landmark survival analyses (hazard ratio [HR] 0.83, 95%CI: 0.71–0.96, p=0.02), independent of patient characteristics, other therapies, and extracranial disease. For patients diagnosed post-approval, who reached the landmark timepoint, ICI receipt was independent of patient demographics, socioeconomic status, and hospital type—with the exception of Medicaid-insured patients, who were less likely than privately insured patients to receive ICI (OR 0.77, 95%CI: 0.60–0.97, p=0.03).

Conclusions Nationally, the use of ICI for NSCLC brain metastasis patients is increasing, generally without significant socioeconomic barriers. Brain metastasis patients diagnosed in the post-approval second-line ICI era (2015) demonstrated significantly better OS than patients diagnosed pre-approval and even than patients diagnosed only in 2014. ICI was associated with a >60% relative increase in median OS. Together our findings from a real-world population demonstrate that the dramatic OS benefits of ICIs for advanced NSCLC also extended to brain metastasis patients.