Which cardiac parameters best predict the cardiovascular outcomes among patients with anti-PD-1 immunotherapy-induced myocardial injury?

AIM: To explore the association of cardiac parameters with different clinical outcomes in patients with anti-PD-1 immunotherapy-induced myocardial injury. METHODS AND RESULTS: We screened 3,848 patients who received anti-PD-1 immunotherapy from June 2018 to Oct 2021 at the Second Xiangya Hospital of Central South University. Among those patients, 134 patients were diagnosed with anti-PD-1 immunotherapy-induced myocardial injury. Twenty-four patients with cardiovascular symptoms were divided into the major adverse cardiac events (MACE) group, and 110 patients without cardiovascular symptoms were divided into the non-MACE group. We compared creatine kinase isozyme (CK-MB), high-sensitivity troponin T (hsTNT), N-terminal pro–B-type natriuretic peptide (NT-ProBNP), electrocardiography (ECG), and echocardiographic parameters between the two groups of patients. CK-MB, hsTNT, NT-proBNP [2,600.0 (1,317.00–7,950.00) vs. 472.9 (280.40–788.80), p ≤ 0.001], left ventricular end-diastolic diameter (LVEDd), left ventricular ejection fraction (LVEF) and QRS interval were significantly different. The receiver operating characteristic (ROC) curve was used to compare the accuracy of various indicators to predict the occurrence of MACE events. NT-ProBNP (area under the curve [AUC] 97.1) was the best predictor, followed by CK-MB (AUC = 94.1), LVEF (AUC = 83.4), LVEDd (AUC = 81.5), and other indicators. In the MACE group, 11/24 patients had experienced cardiogenic death by the end of follow-up. There were significant differences in the CK-MB, hsTNT, NT-proBNP, LVEDd, LVEF, and QRS intervals between the deceased patients and the survivors. The ROC curve shows that hsTNT is the most accurate marker for predicting cardiogenic death in the MACE group (AUC = 91.6). CONCLUSION: In patients with myocardial injury after PD-1 inhibitor treatment, NT-proBNP is the parameter of choice to predict the likelihood of developing cardiovascular symptoms, whereas, in symptomatic patients, hsTNT is the optimal parameter associated with the outcome of death compared with other cardiac parameters.