A real-world retrospective study of incidence and associated factors of endocrine adverse events related to PD-1/PD-L1 inhibitors

BACKGROUND: The adverse events (AEs) related to immune checkpoint inhibitors (ICIs) have been mostly described in clinical trials, however, such trials are restricted to selection criteria and the results cannot wholly represent the real-world setting. We aimed to evaluate the real-world endocrine AEs associated with programmed death receptor-1/programmed death ligand-1 (PD-1/PD-L1) inhibitors in Chinese population. METHODS: This retrospective study included cancer patients who were treated with PD-1/PD-L1 inhibitors between January 2018 and December 2020 at Xinqiao Hospital, the Third Military Medical University. The information of 581 patients was reviewed, and data on clinical characteristics, PD-1/PD-L1 use, occurrence of endocrine AEs, and response to PD-1 blockade treatment were collated. The definition of endocrine AEs relied on diagnostic tests. Fisher’s exact test or Pearson’s chi-squared test was used to analyze the associations between endocrine variables and several categorical variables. Multivariate analyses were performed using a logistic regression model. RESULTS: Endocrine AEs were observed in 116 of the 581 patients (20.0%). The median time to onset of endocrine AEs was approximately 12 weeks. Pembrolizumab was associated with a significantly higher incidence of endocrine AEs compared to other anti-PD-1 agents (38.5%; P=0.0002); PD-1/PD-L1 inhibitor treatment combined with antiangiogenic therapy or with two other therapies (chemotherapy and antiangiogenic therapy) was associated with a significantly increased occurrence of endocrine AEs, compared to PD-1 blockade treatment alone (41.2%; P=0.015), both based on multivariate analysis. Patients who developed endocrine AEs had significantly higher overall response rates (ORRs; 33.3% vs. 23.1%, P=0.045) and disease control rates (DCRs; 91.1% vs. 79.1%, P=0.008) compared to patients without endocrine AEs. In multivariate analysis, endocrine AEs remained an independent factor for both ORR (OR: 1.764, 95% CI: 1.052–2.957, P=0.031) and DCR (OR: 2.896, 95% CI: 1.324–6.332, P=0.008) after adjusting for the confounding factors. CONCLUSIONS: A real-world Chinese population receiving PD-1/PD-L1 treatment, pembrolizumab administrated and triple therapy treatment modalities had a higher incidence of endocrine AEs. Patients who developed endocrine AEs demonstrated a favorable response to PD-l blockade treatment.