Abstract
BACKGROUND: While concomitant medications can affect the efficacy of immune checkpoint inhibitors (ICIs), few studies have assessed associations of concomitant medications with the occurrence and profile of immune-related adverse events (irAEs). METHODS: This study assessed associations of concomitant medication (antibiotics/proton pump inhibitors (PPIs)/corticosteroids)-based risk model termed the "drug score" with survival and the occurrence and profile of irAEs in 851 patients with advanced cancer treated with ICIs (with or without other agents). The study also assessed the survival impact of the occurrence of irAEs, using a landmark analysis to minimize immortal time bias. Multivariable Cox proportional hazard analyses were conducted for progression-free survival (PFS) and overall survival (OS). RESULTS: The drug score classified patients into three risk groups, with significantly different PFS and OS. Notably, the score's predictive capability was better in patients treated with ICIs only than in those treated with ICIs plus other agents. The landmark analysis showed that patients who developed irAEs had significantly longer PFS and OS than those without irAEs. Generally, concomitant medications were negatively associated with the occurrence of irAEs, especially endocrine irAEs, whereas PPI use was positively associated with gastrointestinal irAEs, as an exception. CONCLUSIONS: Using a large pan-cancer cohort, the prognostic ability of the drug score was validated, as well as that of the occurrence of irAEs. The negative association between concomitant medications and irAE occurrence could be an indirect measure of the detrimental effect on the immune system induced by one or more concomitant drugs.