Predictors of persistent extensive late gadolinium enhancement after acute myocarditis: a longitudinal cardiac magnetic resonance study
L Lorenzo Alves, J Goncalves, E Andrade, T Branco, B Viana, S Amorim, R Rodrigues, M VasconcelosAbstract
Background
Persistent late gadolinium enhancement (LGE) on follow-up cardiac magnetic resonance (CMR) is a frequent finding after acute myocarditis and reflects residual myocardial injury. Identifying determinants of persistent extensive LGE may improve risk assessment and longitudinal management, yet available data remain limited.
Purpose
To identify clinical, biochemical, electrocardiographic, echocardiographic and CMR-derived predictors of persistent extensive LGE on follow-up CMR in patients with acute myocarditis.
Methods
We performed a retrospective observational study including consecutive patients admitted with acute myocarditis to a tertiary referral centre between January 2016 and November 2025 who underwent baseline and follow-up CMR. Baseline assessment included clinical variables, biomarkers, electrocardiography, echocardiography, and detailed CMR phenotyping (myocardial oedema, LGE extent and distribution, and biventricular function). The primary endpoint was persistent extensive LGE on follow-up CMR, defined as involvement of ≥4 myocardial segments. Associations were explored using univariate logistic regression, and discriminatory performance was assessed using receiver operating characteristic (ROC) curve analysis.
Results
Among 54 patients with available follow-up CMR, persistent extensive LGE was observed in 13 (24%). Baseline clinical characteristics, electrocardiographic findings, and peak troponin or inflammatory markers did not differ significantly between patients with and without persistent extensive LGE. In contrast, CMR-derived parameters showed relevant associations. Patients with persistent extensive LGE had a higher burden of subepicardial LGE segments at baseline (median 4 [IQR 4–6] vs 3 [2–4], p=0.048) and a numerically greater total LGE burden. In univariate logistic regression, total LGE burden showed a consistent association with persistent extensive LGE (OR 1.43, 95% CI 0.93–2.22), although no single predictor reached statistical significance. ROC analysis demonstrated acceptable discrimination for the number of subepicardial LGE segments (AUC 0.70) and modest performance for total LGE burden (AUC 0.66), whereas biomarkers showed poor discrimination (AUC <0.61).
Conclusion
In acute myocarditis, baseline CMR features—particularly the extent and subepicardial distribution of LGE—are more informative than clinical presentation, biomarkers or echocardiography for identifying patients at risk of persistent extensive myocardial injury on follow-up. These findings reinforce the central role of comprehensive CMR phenotyping in the longitudinal evaluation of myocarditis.ROC curveFor image description, please refer to the figure legend and surrounding text.ROC analysisFor image description, please refer to the figure legend and surrounding text.