Electrocardiographic predictors of major adverse cardiovascular events in immune checkpoint inhibitor-associated myocarditis.

backgroundImmune checkpoint inhibitor (ICI)–associated myocarditis carries a high risk of mortality, yet electrocardiographic (ECG) predictors of adverse outcomes remain unclear. The present study aimed to evaluate the predictive value of clinical and admission ECG parameters for major adverse cardiovascular events (MACE).

methodsWe retrospectively analyzed 60 consecutive patients with ICI myocarditis. Patients were stratified according to the occurrence of major adverse cardiovascular events (MACE), defined as myocarditis-related cardiogenic shock, high-degree atrioventricular block requiring temporary or permanent pacing, or sustained ventricular tachycardia(VT)/ventricular fibrillation(VF). Admission ECGs were systematically adjudicated for conduction and repolarization abnormalities, and an ECG burden score (0–6) was derived. Univariable and multivariable Firth penalized Cox regression models were used to identify predictors of MACE. Model performance was evaluated using discrimination and reclassification metrics across nested models: M1 (clinical variables), M2 (M1 + biomarkers and echocardiography), and M3 (M2 + ECG burden score).

resultsAmong the 60 patients with myocarditis, MACE occurred in 21 patients (35%) patients. Compared with the non-MACE group, QTc prolongation (p = 0.011), QRS ≥ 120 ms((p = 0.002), Atrioventricular block (p = 0.007), bundle branch block (p = 0.005), ST-segment abnormalities (p = 0.003), and T-wave changes (p = 0.004) at the time of admission were more common in the MACE group. In multivariable Firth penalized Cox analysis, severe myocarditis, QTc prolongation, Atrioventricular block, QRS ≥ 120 ms, ST-segment abnormalities, T-wave changes, and higher ECG burden score independently predicted MACE. Sensitivity analyses excluding early MACE or in-hospital deaths confirmed robustness of results. Addition of ECG variables significantly improved model performance (C-index 0.726 [M1] vs. 0.860 [M2] vs. 0.941 [M3]), with significant net reclassification and integrated discrimination gains.

conclusionAdmission ECG abnormalities are powerful, independent predictors of adverse outcomes in ICI myocarditis and provide substantial incremental prognostic value beyond clinical and biomarker/echocardiographic features. Systematic ECG assessment and integration into risk stratification should be prioritized for early triage and monitoring in this high-risk population.