ABSTRACT
BACKGROUND: How corticosteroid use affects treatment response to chemotherapy and immune-checkpoint inhibitors (CICPIs) remains unknown. We assessed how systemic corticosteroid exposure before CICPI modifies the effect of CICPI on outcomes among patients with metastatic non-small cell lung cancer (mNSCLC) or extensive-stage small-cell lung cancer (ES-SCLC). METHODS: We conducted a retrospective cohort study using electronic health records to examine patients with mNSCLC or ES-SCLC who received chemotherapy (CT) between 1st April 2015 and 31st January 2018 or CICPI between 1st February 2018 and 31st August 2019. We excluded those with an actionable driver mutation. Baseline corticosteroid use was defined as systemic corticosteroids within 28 days before the initiation of CT or CICPI, not including premedications. Coprimary outcomes included overall survival (OS), real-world progression (rwP), and real-world progression-free survival (rwPFS) in CICPI-treated corticosteroid users versus non-users. We used inverse probability of treatment weighting (IPW) to adjust for potential confounding. RESULTS: The cohort of 316 patients (median [interquartile range] age, 67 [61-73] years; 156 [49%] male) included 228 CT-treated and 88 CICPI-treated patients. After applying IPW, characteristics were well-balanced between the CT and CICPI groups, and steroid users and non-users. Using CT-treated steroid non-users as a common comparator, CICPI-treated steroid users were as likely as CICPI-treated steroid non-users to die (users IPW hazard ratio [HR] = 0.67, 95% CI = 0.35-1.28 versus non-users IPW-HR = 0.88, 95% C = I0.55-1.42; p = 0.49), have rwP (IPW-HR = 0.35, 95% C = I0.12-0.99 versus IPW-HR = 0.41, 95% C = I0.24-0.70; p = 0.77), or experience rwPFS (IPW-HR = 0.56, 95% C = I0.29-1.09 versus IPW-HR = 0.69, 95% CI0.46-1.03; p = 0.59). CONCLUSION: Corticosteroid use before CICPIs was not associated with worse outcomes, suggesting that corticosteroids should be used with CICPIs when indicated.
