Concomitant Comedications and Survival With First-Line Pembrolizumab in Advanced Non–Small-Cell Lung Cancer

Importance Antibiotics, steroids, and proton pump inhibitors (PPIs) are suspected to decrease the efficacy of immunotherapy. Objective To explore the association of comedications with overall survival (OS) in patients with advanced non–small-cell lung cancer (NSCLC). Design, Setting, and Participants This nationwide retrospective cohort study used target trial emulations of patients newly diagnosed with NSCLC from January 2015 to December 2022, identified from the French national health care database. Eligible patients were treated with pembrolizumab in a first-line setting and alive 2 months after initiating pembrolizumab. Exclusion criteria included hospitalization for infectious disease, autoimmune disorders, or peptic ulcer disease. Exposures Antibiotic and PPI exposures were defined as at least 2 prescriptions 60 days before to 42 days after pembrolizumab start. Steroid exposure was defined as at least 2 prescriptions 30 days before to 30 days after pembrolizumab start. Main Outcomes and Measures The primary outcome was OS. Patients exposed were compared with those without exposure, using inverse probability of treatment weighting (IPTW) to adjust for confounding. Results Between January 2015, and December 2022, 41 529 patients were treated with first line pembrolizumab for advanced disease (27 826 male [67.0%]; median [range] age, 65 [19-97] years; 14 835 [35.7%] treated with pembrolizumab alone; 26 694 [64.3%] treated with pembrolizumab plus chemotherapy). At treatment initiation, 12 898 (41.9%) patients were exposed to antibiotics, 18 210 (59.1%) to steroids, and 16 783 (53.7%) to PPIs. After IPTW, antibiotics (except for macrolide and penicillin) were associated with shorter OS (hazard ratio [HR], 1.08; 95% CI, 1.05-1.12; P < .001), but it varied by antibiotic type. Steroids were not associated with OS (HR, 0.98; 95% CI, 0.95-1.02; P = .37); however, there was an interaction with pembrolizumab regimen (ie, pembrolizumab alone or with chemotherapy) ( P for interaction < .001), and there was a dose-dependent association according to daily prednisone-equivalent dose ( P for trend < .001). Steroids were associated with worse OS when prescribed at doses greater than 20 mg per day for pembrolizumab alone ( P for trend = .005) and greater than 30 mg per day for pembrolizumab combined with chemotherapy ( P for trend < .001). PPIs were associated with worse OS (HR, 1.13; 95% CI, 1.10-1.17; P < 001). Conclusions and Relevance In this cohort study of patients with advanced NSCLC treated with pembrolizumab, exposure to some classes of antibiotics, to steroids (>20 mg per day of prednisone equivalent), and to PPIs was associated with worse OS, indicating that comedications should be monitored carefully in patients with immunotherapy.