Comparison of the diagnostic and prognostic value of criteria for immune checkpoint inhibitor related myocarditis

Myocarditis is a dreaded complication of immune-checkpoint inhibitor (ICI) therapy but challenging to diagnose. There are no published data comparing the two leading diagnostic criteria for ICI-related myocarditis (ICIrM) and their association with cardiovascular events. In this retrospective cohort study, we reviewed all patients who underwent ICI therapy and had cardiac troponin assessment for possible myocarditis across three tertiary institutions from 2011 to 2022. ICIrM was adjudicated by the Bonaca et al. criteria and the ESC-ICOS guidelines. A propensity matched control group was identified of patients treated with ICI without developing myocarditis. Baseline characteristics and long-term outcomes, including cardiac death, MACE (myocardial infarction, TIA/stroke, heart failure), and arrhythmias data were curated, and patients diagnosed with ICIrM by each criteria were compared to controls for cardiovascular events. A total of 59 patients (mean age was 73.1 ± 10.2 years, 60.1% male) were identified as having a diagnosis of ICIrM by Bonaca criteria (16 definite, 13 probable and 30 possible myocarditis). Forty-seven of these patients met the ESC-ICOS guidelines criteria, and all patients meeting either set of ICIrM criteria were treated with steroid therapy. At 3-year follow up, patients diagnosed with ICIrM by the Bonaca criteria had a high risk of cardiac mortality (HR 17.84, 95%CI 2.36-134.62, p = 0.005), MACE (HR 4.90, 95%CI 2.40-10.02, p < 0.001) and arrhythmias (HR 3.33, 95%CI 1.78–6.21, p < 0.001) when compared to matched controls. ICIrM by ESC-ICOS criteria was similarly predictive of cardiac mortality, MACE, and arrhythmias (HR 15.01, 95%CI 1.96-114.76, p = 0.009, HR 5.18, 95%CI 2.33–11.53, p < 0.001, and HR 3.41, 95%CI 1.73–6.70, p < 0.001 respectively). The ESC-ICOS guidelines were more restrictive than the Bonaca et al. criteria for the diagnosis of ICIrM but similar in terms of prognostic value.