Objective To evaluate the impact of glucocorticoids (GC) on the clinical efficacy of immune checkpoint inhibitors (ICI) in treating non-small cell lung cancer (NSCLC).

Methods The clinical data of 131 patients with advanced NSCLC who received ICI treatment in the Affiliated Hospital of Yangzhou University were retrospectively analyzed. They were divided into GC group (n=79) and non-GC group (n=52) according to whether they used GC (≥ 10 mg/d prednisone or equivalent GC) within 3 months before and after ICI treatment or not. The GC group was divided into four subgroups according to the following indications: non-cancer indications group, cancer indications group, immune related adverse events (irAEs) group and pre-chemotherapy treatment group. The correlation between GC usage and the clinical characteristics of patients was analyzed. The Kaplan-Meier survival curve was used to analyze the impact of GC use on overall survival (OS) and progression-free survival (PFS). The univariate and multivariate analysis based on Cox hazard proportional regression models were used to identify whether GC use was a prognostic factor.

Results No significant correlations were found between GC use and age (χ²=0.180, P=0.672), gender (χ²=3.179, P=0.075), smoking (χ²=0.579, P=0.447), pathological type (χ²=0.628, P=0.428), the score of United States Eastern Collaborative Group(ECOG) (χ²=0.074, P=0.785) score and treatment lines (χ²=1.853, P=0.173), while it was correlated with treatment strategies (χ²=3.998, P=0.046). The OS and PFS of the GC group were shorter when compared with the non-GC group, and the differences were statistically significant (P < 0.05). The multivariate analysis showed that GC use was an independent influencing factor for OS (HR=1.82, P=0.026) and PFS (HR=1.76, P=0.012). The subgroup analysis demonstrated that patients receiving GC for cancer related indications had shorter OS (P=0.006) and PFS (P=0.011) than those receiving GC for non-cancer indications, irAEs, and pre-chemotherapy treatment. The multivariate analysis showed that GC use for cancer related indications was an independent risk factor for OS (P=0.001) and PFS (P=0.005).

Conclusion The GC use is negatively correlated with the clinical efficacy of ICI in treating patients with advanced NSCLC, especially for those receiving GC for cancer-related indications. GC should be used cautiously in ICI-treated NSCLC patients.