Abstract
OBJECTIVE: To identify predictors of all-grade, grade ≥ 3, and onset time of immune-related adverse events (irAEs) in cancer patients undergoing immune checkpoint inhibitors (ICIs) therapy. METHODS: This retrospective analysis included cancer patients treated with ICIs at Chongqing Medical University Second Affiliated Hospital from 2018 to 2024. Logistic regression and Cox regression analyses were used to identify predictors of all-grade and grade ≥ 3 irAEs and the time of irAE onset. RESULTS: Among the 3,795 patients analyzed, 1,101 (29.0%) developed all-grade irAEs, and 175 (4.6%) experienced grade ≥ 3 irAEs. Multivariate logistic regression revealed that female (OR = 1.37, p < 0.001), combination therapy (OR = 1.87, p < 0.001), pre-existing autoimmune diseases (AIDs) (OR = 5.15, p < 0.001), pre-existing cirrhosis (OR = 1.34, p = 0.001), antibiotic use during ICIs treatment (OR = 1.51, p < 0.001), and a higher baseline prognostic nutritional index (PNI) (OR = 1.23, p = 0.01) were significant predictors for the development of all-grade irAEs. The predictors for grade ≥ 3 irAEs included age ≥ 60 (OR = 1.49, p = 0.023) and pre-existing AIDs (OR = 2.09, p = 0.005), For the onset time, predictors included female (HR = 1.26, p = 0.001), combination therapy (HR = 1.80, p < 0.001), pre-existing AIDs (HR = 2.25, p < 0.001), and pre-existing infection (HR = 1.20, p = 0.008). CONCLUSIONS: Females, combination therapy, pre-existing AIDs and cirrhosis, antibiotics, and a higher baseline PNI are associated with a higher risk of developing all-grade irAEs. Those aged ≥ 60 and with pre-existing AIDs face a higher risk of severe irAEs. Females, undergoing combination therapy, with pre-existing AIDs and infection generally experience a shorter time to irAEs onset. Multicentric prospective studies are warranted to validate these findings.