Entropy for Prediction of MACEs in Myocarditis: A Cardiac MRI–based Biomarker of Myocardial Tissue Heterogeneity

Background The prognostic value of left ventricular (LV) entropy in acute myocarditis is unclear. Purpose To assess the association of LV entropy, a measure of myocardial heterogeneity derived from cardiac MRI, with cardiovascular events in patients with acute myocarditis. Materials and Methods This retrospective study evaluated 492 patients with acute myocarditis who underwent cardiac MRI between March 2016 and March 2024 (discovery cohort: 319 patients from one center; external testing cohort: 173 patients from two centers). LV global entropy reflects the distribution of pixel signal intensities across the LV myocardium, whereas LV late gadolinium enhancement (LGE) entropy reflects myocardial heterogeneity in LGE areas. Patients were followed up after cardiac MRI for major adverse cardiovascular events (MACEs), including cardiac death, heart failure hospitalization, heart transplant, recurrent myocarditis, and sustained ventricular arrhythmia. Multivariable Cox proportional hazards regression was used to evaluate the association between LV entropy and MACEs, with adjustment for diabetes mellitus, log N-terminal pro-B-type natriuretic peptide, LV ejection fraction, LGE extent, and extracellular volume fraction; receiver operating characteristic and model performance analyses were performed. Results In the discovery cohort (mean age, 32 years ± 14; 191 male [59.9%]), 57 of 319 patients (17.9%) experienced MACEs during a median follow-up of 3.1 years. LV global entropy of at least 8.5 defined the high-risk group. Multivariable Cox regression showed that LV global entropy (hazard ratio, 3.33 [95% CI: 1.61, 6.87];