Abstract
BACKGROUND: Non-small cell lung cancer brain metastases (NSCLCBM) have dismal outcome. Immune checkpoint inhibitor (ICI) has promising activity in lung cancer; however, limited data exists on impact of steroids on efficacy of ICI in NSCLCBM. METHOD: We reviewed 120 NSCLCBM patients treated with ICI (2012–2017) at our center. Fifty-nine patients who received at-least 2 cycles of ICI after diagnosis of NSCLCBM with at least one follow-up MRI were included for analysis, others were excluded as they have received ICI before BM or did not have follow-up MRI. All patients had prior chemotherapy and received ICI (nivolumab/pembrolizumab) in second or third line. Overall survival (OS) was calculated from date of ICI therapy to date of death or last follow-up. Progression free survival (PFS) was calculated from date of ICI to date of progression. OS was estimated by Kaplan-Meier method and analyzed by Cox proportional hazards model. RESULTS: Median age was 61 years (39–77 years), median KPS of 90, 59% were females, 49 were adenocarcinoma (4 EGFR mutations, 2 ALK rearrangement). Forty-one patients had supratentorial lesions, 7 had infratentorial and 11 had both. Fifty-four patients underwent stereotactic-radiosurgery, 4 had whole brain radiation-therapy and 30 patients received steroids (28 at baseline before start of ICI and 2 patients within 2 cycles of ICI). Twenty-two patients received dexamethasone (2–8 mg/day), 8 received prednisone (5–20 mg/day). Median OS was 18.9 months and PFS was 9.9 months. Steroid use was associated with decreased PFS, 5.0 months vs 13.8 months (p value = 0.02). No significant difference was seen in OS with steroid treatment. On multivariable analysis, age at diagnosis significantly predicted OS (p = .034). Greater number of baseline intracranial lesions (p < .01) and concurrent steroids (p < .0005) correlated with decreased PFS. CONCLUSION: Steroid use in NSCLCBM is associated with decreased PFS with ICI.