Early mortality-associated factors in pediatric fulminant myocarditis patients treated with venoarterial extracorporeal membrane oxygenation: a single-center retrospective cohort study

Background: Fulminant myocarditis (FM) is a life-threatening pediatric condition that may progress rapidly to cardiogenic shock and multiorgan dysfunction. This study evaluated early clinical and laboratory factors associated with in-hospital mortality among pediatric FM patients treated with venoarterial extracorporeal membrane oxygenation (VA-ECMO). Methods: We retrospectively reviewed 44 consecutive children (<18 years) with FM treated with VA-ECMO from 2018 to 2025. Continuous variables are presented as median (interquartile range) and compared using the Mann-Whitney U test. Because only 7 deaths occurred, regression analyses were prespecified as exploratory. Least absolute shrinkage and selection operator (LASSO) was used for variable screening, followed by univariable and exploratory 3-variable logistic models. Receiver operating characteristic (ROC) analysis was used to describe discrimination of selected admission markers. Results: Thirty-seven patients survived to discharge and seven died in hospital, yielding an in-hospital survival rate of 84.1%. Compared with survivors, non-survivors had lower admission systolic blood pressure [73 (64–81) vs. 90 (83–105) mmHg, P=0.003], lower left ventricular ejection fraction (LVEF) [31.0% (29.5–34.0%) vs. 49.0% (40.0–55.0%), P=0.002], and higher blood urea nitrogen (BUN) [14.2 (9.7–15.1) vs. 7.3 (5.8–8.7) mmol/L, P=0.01]. In univariable logistic models, systolic blood pressure (SBP), LVEF, and BUN were associated with in-hospital mortality; however, estimates were imprecise after exploratory adjustment because of the limited number of events. The areas under the curve (AUCs) for SBP, LVEF, and BUN were 0.86, 0.87, and 0.79, respectively. Conclusions: Admission SBP, LVEF, and BUN were early markers associated with in-hospital mortality in this single-center cohort of pediatric FM patients treated with VA-ECMO. These findings should be interpreted as hypothesis-generating rather than definitive prognostic evidence and require validation in larger multicenter cohorts.