Background: Treatments for melanoma increasingly involve the use of immune checkpoint inhibitors (ICIs), which are associated with cardiac toxicities. Our study aims to examine predictors of myocardial infarction (MI) in melanoma patients receiving treatment. Methods: The United Healthcare (Optum Clinformatics) Closed Claims + Lab Results Database (01/2007 – 03/2021) was used to examine factors potentially associated with the occurrence of hospitalized MI in melanoma patients. Adult patients were included if they: 1) had ≥12 months of continuous insurance coverage at baseline; 2) had ≥2 outpatient or ≥1 inpatient claims with an International Classification of Diseases (ICD) code indicative of melanoma; 3) received ≥1 type of cancer treatment (surgery, chemotherapy, radiotherapy, ICI, or targeted therapy); 4) had a claim with a melanoma ICD code within 5 days before the first cancer treatment; and 5) had no history of MI at baseline. Baseline was defined as the year prior to the first date of treatment; MI occurrence was identified from inpatient claims. We conducted a time-to-event (MI) analysis using a multivariate Cox proportional hazards regression model, with backward selection and death as a competing risk. Results: A total of 42,101 patients (58.8% male; median age of 68 yrs at baseline) were included in the study. Of these, 954 patients (2.3%) experienced a MI, and 5,578 (13.2%) died before having a MI. Compared with patients who underwent surgery only, those who received multiple treatments with an ICI were more likely to have a MI (hazard ratio [HR] = 1.25, 95% confidence interval [CI]: 1.01-1.55; Table). Men had a higher risk than women (HR = 1.47, 95% CI: 1.27-1.71), as did older people (compared with 65-74-yr-olds, HRs were 0.24, 0.49, 1.41, and 2.06 for <55-, 55-64-, 75-84-, and ≥85-yr-olds, respectively; all P<.05). Being Asian and having certain comorbidities at baseline were also associated with a higher MI risk (Table). Conclusions: In this retrospective cohort analysis of patients with melanoma, we observed an increased risk of MI among patients who received multiple treatments with ICIs, compared with surgery only. More research is needed to further elucidate predictors of MI events related to cancer treatment.

*The Cox model included all covariates in the table, as well as age, sex, and geographic region (death was the competing risk).