Characteristics and risk factors of immune-related adverse events in patients receiving immune checkpoint inhibitor combination therapy.

Objectives: To overcome the limitations of immune checkpoint inhibitor (ICI) monotherapy, combination strategies are increasingly adopted in clinical practice, which may further complicate the occurrence profile of immune-related adverse events (irAEs). This study aimed to investigate the characteristics, onset time, and risk factors of irAEs and endocrine-irAEs (EirAEs) in cancer patients receiving ICI combination therapy. Methods: Demographic data, treatment regimens, concomitant medications, and laboratory biomarkers were collected from 297 cancer patients, of whom 282 (94.95%) received ICI combination therapies. The incidence, severity, and onset time of irAEs were analyzed. Binary logistic regression was used to evaluate factors associated with irAEs and EirAEs, respectively. Model performance was assessed by the receiver operating characteristic (ROC) curve analysis. Results: IrAEs were seen in 113(38.05%) patients, with 4.38% experiencing grade≥3 irAEs. The most common irAEs were endocrine (12.12%), gastrointestinal (8.75%), and dermatological (6.73%) toxicities. For irAEs, ICI combined with chemotherapy accounted for the highest proportion (49.18%). For EirAEs, the proportion of ICI combined with targeted therapy was the highest(36.11%). Binary logistic regression analysis model revealed corticosteroid use was a protective factor for irAEs (OR = 0.511, 95%CI:0.296-0.884, P = 0.016), and age≥65 years was associated with lower risk (OR = 0.487, 95%CI:0.292-0.813, P = 0.006). Chemotherapy and targeted therapy did not show a significant impact on the overall incidence of irAEs. For EirAEs, combination with targeted therapy (OR = 2.888, 95%CI:1.186-7.033, P = 0.020) and higher baseline TSH levels (OR = 1.032, 95%CI:1.002-1.063, P = 0.038) were risk factors, while corticosteroid use was a protective factor(OR = 0.231, 95%CI:0.092-0.581, P = 0.002). The average time to onset of irAEs was 2.43 treatment cycles, with 85.25% emerging within 4 cycles, and endocrine toxicity exhibited delayed onset. Conclusion: Among patients predominantly treated with ICI combination therapy, irAEs were diverse and primarily low-grade, with the majority emerging within the first 4 treatment cycles. For irAEs, Age≥65 years was associated with lower risk, and corticosteroid use was a protective factor. For EirAEs, combination with targeted therapy and higher baseline TSH levels were risk factors, while corticosteroid use remained a protective factor. Onset time varied across different irAEs. These findings provide evidence to support risk-stratified monitoring and early intervention in clinical practice. Multicenter prospective studies are warranted for further validation.