The incidence of cardiac adverse events is low with immune checkpoint inhibitor (ICI) treatment, administered either as a single agent or as a dual agent combination, and is not statistically different to that observed with non-ICI treatment for patients with diverse cancer types. These findings from a large meta-analysis of randomised clinical trials were presented at the ESMO Immuno-Oncology Virtual Congress 2020, held from 9 to 12 December 2020.

Elisa Agostinetto of the Academic Trials Promoting Team, Institut Jules Bordet in Brussels, Belgium and colleagues remarked that the risk of cardiotoxicity had been poorly investigated, especially whether ICI combinations increase cardiotoxicity compared to treatment with ICI monotherapy, thus prompting this meta-analysis. 

The primary objective of this study was to evaluate the cardiotoxicity with ICI therapy as compared to other cancer treatments, and the second objective was to determine the cardiotoxicity of dual agent ICI immunotherapy combinations versus single agent ICI.

Dr Agostinetto and colleagues conducted this systematic review and meta-analysis according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines that was registered in the PROSPERO database (CRD42020183524). The investigators performed a systematic search of the PubMed, MEDLINE, and EMBASE databases, as well as conference proceedings up to 30 June 2020. They identified all randomised clinical trials referencing all solid tumour types comparing an ICI against other treatments to satisfy the primary objective and also identified dual agent ICI compared to single agent ICI to meet the secondary objective. Pooled risk ratios (RR) with 95% confidence intervals (CI) for cardiotoxicity events were calculated using random effect models. Two separate and distinct meta-analyses were performed to address the primary and secondary objectives.

The analysis comprised 80 studies with 35,337 patients, of which 66 studies with 34,664 patients were focused towards the primary objective and 14 studies with 673 patients supported the secondary objective.

No statistically significant differences in the frequency of cardiac adverse events (AEs) were observed between the ICI and non-ICI treatment groups, nor between the dual ICI and single agent ICI treatment groups.

Overall, any cardiac AEs occurred in ≤4% of patients and myocarditis in ≤0.2% of patients (Table 1).


Table 1. Number of cardiac adverse events per number of patients, pooled incidence, and relative risk-ratios with 95% confidence intervals in the immune-checkpoint inhibitors (ICI) group vs non-ICI group and in the dual-ICI group vs single-ICI group.

© Elisa Agostinetto.

Specifically, the pooled incidence of any cardiac AE was 3.78% in the cohort treated with an ICI compared to 3.40% in the group receiving other, non-ICI treatment (RR 1.14; 95% CI 0.88-1.48).

With ICI versus non-ICI treatments, the pooled incidence for arrhythmias was 1.79 versus 1.49, respectively (RR 1.32; 95% CI 0.94-1.84). The pooled incidence for several other cardiotoxicities was less than 1.00 in either treatment group with lower RR, including myocarditis (0.12 vs 0.01; RR 1.11), myocardial infarction (0.41 vs 0.27; RR 1.19), pericarditis (0.51 vs 0.22; RR 1.14), heart failure (0.43 vs 0.63; RR 0.61), valvular disease (0 vs 0.03; RR 0.63), cardiac arrest (0.24 vs 0.09; RR 1.23), and cardiac death (0.33 vs 0.21; RR 1.07).

Regarding the secondary objective comparing cardiotoxicity in patients treated with an ICI combination versus those receiving single ICI therapies, the pooled incidence for any cardiac AE was 2.87 versus 0.40, respectively (RR 1.91; 95% CI 0.52-7.01). Again, the highest incidence was observed for arrhythmias of 2.48 with dual ICI versus 0 for single ICI treatment (RR 1.65; 95% CI 0.40-6.89). The incidence regarding all other cardiotoxicity was less than 1 in either treatment arm: myocarditis (0.17 vs 0; RR 1.10), myocardial infarction (0.50 vs 0; RR 0.98), pericarditis (0 vs 0.49; RR 0.67), heart failure (0.38 vs 0; RR 1.04), valvular disease (0 vs 0; RR 0.79), cardiac arrest (0 vs 0; RR 0.79), and cardiac death (0.27 vs 0; RR 1.28).

Of note, these observations remained consistent across subgroup analyses by tumour type, setting of disease, treatment line, and type of treatment where no statistically significant differences were found between the ICI versus non-ICI treatment or between combination versus single ICI therapy.

This study is the largest meta-analysis to date of cardiotoxicity induced by ICI,  investigating not only myocarditis events, which are known to have a potential immune-related aetiology, but also a broader range of cardiac AEs including myocardial infarction, pericarditis, heart failure, arrhythmias, valvular disease, cardiac arrest and cardiac death.

Nonetheless, the authors specify that not all studies included in the meta-analyses provided complete data about cardiac AEs among participants. Several studies presented only AEs occurring above a specified incidence, which might have ranged from 1% to 20%. This could favour underreporting of rare AEs, like cardiac events, and could mask the real incidence of this toxicity. Standardisation of AEs reporting should be a priority in clinical trials.


Importantly, the authors concluded that the use of an immune checkpoint inhibitor as a single agent or in combination regimens was not associated with increased risk of cardiotoxicity. However, despite the apparent cardiac safety of ICI, investigators of clinical trials should be strongly encouraged to report cardiac AEs systematically and as completely as possible.

No external funding was disclosed. 


41P – Agostinetto E, Eiger D, Lambertini M, et al. Cardiotoxicity of immune checkpoint inhibitors: A meta-analysis of randomized clinical trials. ESMO Immuno-Oncology Virtual Congress 2020 (9-12 December).