Discrepancies in cardiovascular toxicity of immune checkpoint inhibitors: evidence from trials and real-world data
H D M Helena Michalopoulou, A P Amalia Papanikolopoulou, F K S Foivos Konstantinos Stamatis, K K Konstantinos Kyriakoulis, J V John Vathiotis, N S Nikolaos Syrigos, P A Polyxeni Alexiou, K T Konstantinos Thomopoulos, A C H Andriani CharpidouAbstract
Background
Cardiovascular immune-related adverse events (CV-irAEs) are increasingly recognized as clinically significant complications of immune checkpoint inhibitors (ICIs), yet data derived from clinical trials may underrepresent their true incidence. This study aimed to assess the real-world burden of CV-irAEs among patients with malignancies treated with ICIs and to explore discrepancies between observational evidence and prior trial-based meta-analyses.
Methods
A systematic literature review and meta-analysis were conducted in accordance with PRISMA guidelines, focusing on observational studies that included adult cancer patients receiving ICIs. Searches were performed in the PubMed database. A single-arm meta-analysis using the metaprop command in Stata v16 was used to estimate pooled incidence rates for a range of CV-irAEs, including myocarditis, pericardial disease, arrhythmias, cardiac failure, ischemic heart disease, valvular heart disease, arterial disease, venous thromboembolism, and Takotsubo cardiomyopathy.
Results
A total of 42 eligible studies were included. The median time to onset of CV-irAEs was 119 days, with an interquartile range of 53 to 180 days.
The pooled incidence of overall CV-irAEs was 8% (95% confidence interval [CI]: 6% to 10%). The most frequent events were arrhythmias with a pooled incidence of 18% (95% CI: 10% to 27%), followed by myocarditis at 11% (95% CI: 5% to 18%), and cardiac failure at 8% (95% CI: 2% to 15%). Ischemic heart disease occurred in 6% of patients (95% CI: 3% to 11%), pericardial disease in 5% (95% CI: 1% to 10%), and arterial disease in 5% (95% CI: 1% to 12%). Venous thromboembolism was observed in 3% of cases (95% CI: 0% to 8%). Cardiovascular-related mortality was estimated at 12% (95% CI: 4% to 22%).
Conclusion
This meta-analysis of real-world observational studies reveals a substantially higher incidence of CV-irAEs compared to estimates derived from clinical trial populations. The increased rates may be attributed to longer follow-up periods, broader patient inclusion criteria, and heterogeneous definitions of cardiotoxicity. These findings suggest that CV-irAEs represent an underappreciated but emerging clinical challenge in the era of immuno-oncology. Enhanced cardiovascular monitoring, risk stratification, and multidisciplinary management strategies are warranted to mitigate morbidity and mortality associated with ICI-related cardiotoxicity. Prospective studies and harmonized reporting standards are critical for better characterizing long-term outcomes and guiding therapeutic decision-making.