Factors Associated With Immune Checkpoint Inhibitor–Related Myocarditis

Immune-related adverse events (irAE) can develop in patients treated with immune checkpoint inhibitors (ICIs). For example, ICI treatment can increase the risk of myocarditis as a lethal irAE,¹ with a mortality rate up to 50%.² However, the frequency of ICI-related myocarditis is low, and thus the detailed pathogenic mechanisms and risk factors remain unknown. General myocarditis and ICI-related myocarditis have distinct manifestations, suggesting different risk factors; however, the differences of risk factors in these 2 types are unclear.

Risk factors for adverse drug events can be determined in a real-world setting through risk assessment using a large-scale spontaneous reporting system covering patients of various backgrounds and observation areas. Here, we investigated the risk factors for ICI-related myocarditis using the US Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) database, an adverse event spontaneous report database from the United States.

As an observational study using an open access database (FAERS) with anonymized patient information, institutional review board review and approval were waived. Because it is impossible to identify individual patients, informed consent was not required.

Adverse event reports were downloaded from the FDA website, and FAERS data collected between July 2014 and June 2018 were analyzed. Dates of analysis were October 7, 2013, to June 30, 2018. Because FAERS includes duplicate reports, only the most recent report of a patient was used, as recommended by the FDA. Patients aged 0 to 100 years were included in this study. Myocarditis was defined according to the 7 preferred terms listed in the National Center for Biomedical Ontology Medical Dictionary for Regulatory Activities Terminology (ie, autoimmune myocarditis, eosinophilic myocarditis, hypersensitivity myocarditis, lupus myocarditis, myocarditis, myocarditis postinjection, and radiation myocarditis; http://bioportal.bioontology.org/ontologies/MEDDRA?p=classes&conceptid=10029548). Multiple logistic regression analysis, including age, sex, ICI use, and interactions as covariates for the risk of ICI-related myocarditis, was performed using R statistical software version 3.3.2 (R Foundation for Statistical Computing).

We identified 1 979 157 reports including 13 096 cases that received 5 different ICIs. Nivolumab was the most common ICI (n = 6029 of 13 096 [46.04%]). Reporting rates of myocarditis were significantly high in each ICI group (atezolizumab: odds ratio [OR] 6.38, [95% CI, 1.59-25.59]; P = .04; durvalumab: OR, 16.81 [95% CI, 4.18-67.69]; P = .007; ipilimumab alone: OR, 14.26 [7.85-25.88]; P < .001; nivolumab alone: OR, 19.09 [95% CI, 14.34-25.42]; P < .001; and pembrolizumab: OR, 15.70 [95% CI, 10.17-24.23]; P < .001; concomitant ipilimumab and nivolumab: OR, 30.26 [95% CI, 19.58-46.78]; P < .001) (Table 1).

As shown in Table 2, multiple logistic regression analysis showed that myocarditis risk was significantly associated with ICI use (OR, 9.66; 95% CI, 7.16-13.05; P < .001). Female sex or age of 75 years or older alone was not associated with an increase in the reported frequency of myocarditis; however, in the analysis of interaction with ICI, myocarditis risk was significantly higher in female patients (OR, 1.92; 95% CI, 1.24-2.97; P = .004) and patients 75 years or older (OR, 7.61; 95% CI, 4.29-13.50; P < .001). In addition, the combination of ICIs (ipilimumab and nivolumab) was also associated with an increased risk of myocarditis (OR, 1.93; 95% CI, 1.19-3.12; P = .008).

Our findings suggest that careful monitoring of ICI-related myocarditis is warranted for populations regarded as low-risk for general myocarditis,³ such as female patients and patients 75 years or older. A limitation of the present study is that the FAERS database is a self-reporting database that may include reporting bias and inaccurate reports. In addition, FAERS provides limited information regarding prior diseases, cancer type, and treatment history. Therefore, it is inappropriate to compare the risks of myocarditis among different ICIs based on the results of present study. In addition, the imbalance between 2 groups of irAE risk factors, such as autoimmune disease,⁴ may have contributed to the increased risk of ICI-related myocarditis in female patients and patients 75 years or older.

In previous randomized clinical trials, target patients included relatively healthy elderly patients, in contrast to those in the FAERS database, which might explain the increased risk of ICI-related myocarditis in this population. In addition, the sex differences on irAE outcomes remain controversial, and therefore further specific studies of this association are needed.^5,6 Nevertheless, the FAERS database enables rapid risk assessment, and the present results are useful for improved management of ICI-related myocarditis in clinical practice as well as provide helpful information for detailed future investigations.

Accepted for Publication: June 5, 2019.

Corresponding Author: Yoshito Zamami, PhD, Department of Clinical Pharmacology and Therapeutics, Tokushima University Graduate School of Biomedical Sciences, 3-18-15 Kuramoto, Tokushima 770-8503 (zamami@tokushima-u.ac.jp).

Published Online: August 22, 2019. doi:10.1001/jamaoncol.2019.3113

Author Contributions: Dr Zamami and Mr Niimura had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Zamami, Niimura, Okada, Koyama, Izawa-Ishizawa, Ishizawa.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Zamami, Okada.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Niimura, Okada, Koyama, Fukushima, Izawa-Ishizawa, Ishizawa.

Obtained funding: Zamami.

Administrative, technical, or material support: Zamami.

Supervision: Zamami.

Conflict of Interest Disclosures: None reported.

Funding/Support: This research was partially supported by a Platform Project for Supporting Drug Discovery and Life Science Research (Basis for Supporting Innovative Drug Discovery and Life Science Research [BINDS] from AMED under grant No. JP18am0101085). This research was supported by the Japan Research Foundation for Clinical Pharmacology Grant (No. 2018A10) and the Japan Society for the Promotion of Science (JSPS) KAKENHI (No. 18K06785). The authors also received support of crowdfunding launched by Otsucle (Organization for People With Universities) for development of a preventive drug against the adverse effects of anticancer agents (Tokushima, Japan).

Role of the Funder/Sponsor: The funders/sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We thank Mitsuhiro Goda, PhD (Tokushima University Hospital), Shunsuke Ishida, PhD (Tokushima University Hospital), and Kenshi Takechi, PhD (Tokushima University Hospital) for technical assistance; Masayuki Chuma, PhD (Tokushima University Hospital) for review of the study design; Masaki Yoshino, MS (Niigata Prefectural Cancer Center Hospital), Hirofumi Hamano, PhD (Tokushima University Hospital), Masaya Kanda (Tokushima University Hospital), Fuka Aizawa, PhD (Tokushima University Hospital), Takumi Sakurada, PhD (Tokushima University Hospital), for critical review of draft; Hiroaki Yanagawa, PhD (Tokushima University Hospital), Hiromichi Fujino, PhD (Tokushima University), Koichiro Tsuchiya, PhD (Tokushima University), and Yoshihiro Yamanishi, PhD (Kyushu Institute of Technology) for providing critical advice for this study. We would like to thank Editage (Editage, Tokyo, Japan) for English-language editing. None of the individual contributors received compensation beyond their usual salaries; Editage received compensation for editorial services rendered.