Infusion-related reactions from immune checkpoint inhibitors: A proportional and network meta-analysis.

38 Background: Immune checkpoint inhibitors (ICIs) have become the standard of care for the treatment of patients with solid tumors. Infusion-related reactions (IRRs) to ICIs have been reported in up to 20%, but the incidence varies among ICIs. Therefore, we aimed to report and compare the incidence of IRRs to each ICI therapy. Methods: We searched PubMed/MEDLINE, Embase, and Web of Science to identify phase 3 randomized controlled trials (RCTs) evaluating ICIs (CTLA-4, PD-1, PD-L1, and LAG-3 inhibitors) in solid tumors. We performed a random-effects model network meta-analysis to compare the odds ratio (OR) of IRRs by using RCTs comparing dual ICIs, ICI monotherapy, and placebo/observation. For a proportional meta-analysis, treatment arms evaluating ICI monotherapy, or dual ICIs, were selected to pool the incidence of IRRs. Results: A total of 25,250 patients from 47 phase 3 RCTs were included in the analysis. A network meta-analysis of treatment-related IRRs included 10 RCTs and showed that avelumab (OR 70.2, 95% confidence interval [CI]: 4.32-1140) and atezolizumab (40.72, 2.47-671.4) had a significantly higher risk of treatment-related IRRs compared with placebo/observation. Network ranking revealed that avelumab, atezolizumab, and relatlimab plus nivolumab were the top three therapies with risk of treatment-related IRRs. Proportional meta-analyses revealed that the pooled incidence of treatment-related IRRs was 5.39% for PD-1 plus CTLA-4 inhibitors, 8.25% for PD-L1 (avelumab: 13.1%, non-avelumab: 1.93%), 1.97% for CTLA-4, and 1.95% for PD-1 inhibitors. Results of immune-related IRRs were relatively consistent with those of treatment-related IRRs (Table). Conclusions: Avelumab has the highest risk of IRRs followed by atezolizumab and dual ICIs. This comparative study provides insight into the incidence of IRRs with ICI regimens. These results are useful in assessing which systemic therapies are responsible for IRRs, particularly when ICIs are combined with other agents. [Table: see text]