Age-associated clinical patterns in paediatric myocarditis: left ventricular function as the key early risk stratifier
Authors
- Franziska Seidel
- Cynthia Huber
- Bernd Opgen-Rhein
- Martin Boehne
- Johannes Weickmann
- Tobias Hecht
- Felix Anderheiden
- Katja Reineker
- Axel Rentzsch
- Gesa Wiegand
- Dorothée Böcker
- Daniela Kiski
- Marcus Fischer
- Wiebke Frede
- Bettina Ruf
- Konstantin Papakostas
- Gleb Tarusinov
- Inga Voges
- Michael Kaestner
- Kerstin Gummel
- Assunta Schiebel
- Jennifer Kramp
- Ulrike Herberg
- Gunter Kerst
- Bardo Wannenmacher
- Eva Gatzweiler
- Walter Knirsch
- Nele Rolfs
- Eicke Schwanzkopf
- Tania Garfias-Veitl
- Thomas Pickardt
- André Rudolph
- Jens Böhmer
- Tim Friede
- Stephan Schubert
- Daniel Messroghli
Journal
- ESC Heart Failure
Citation
- ESC Heart Fail
Abstract
BACKGROUND AND AIMS: In myocarditis, early identification of high-risk patients is essential, robust prognostic markers are lacking. We aimed to identify early clinical risk stratifier at index hospitalization for major adverse cardiovascular events (MACE). METHODS: We analysed 725 paediatric patients with possible/definite myocarditis enrolled in the prospective multicentre MYKKE registry (2013-2024). MACE included mechanical circulatory support (MCS), heart transplantation, or death during initial admission. Model-based recursive partitioning (MOB) and bootstrap-based logistic regression with multiple imputation were applied for risk stratifier selection out of 20 baseline variables. The model was externally validated in the Swedish SWEDCON- cohort (n = 243). Discrimination was assessed by the area under the curve (AUC). RESULTS: Among 725 patients (median age 14.3 years [IQR 4.0-16.4]; 33% female), 88 (12.2%) experienced MACE (MCS 10.9%, transplantation 4.3%, death 2.2%). MOB identified left ventricular ejection fraction (LVEF) ≤ 39% as the only stable risk stratifier, delineating a "heart failure type" in younger children (1.9 years; LVEF 24%), and a "chest pain type" in adolescents with preserved LVEF (15.6 years; LVEF 59%). LVEF was confirmed as the sole robust risk stratifier (OR per 1% increase: 0.90 ; 95% CI 0.881-0.917; p < 0.01; AUC 0.90, 95% CI 0.896-0.903). In SWEDCON (7.4% MACE), discrimination was maintained (AUC 0.894, 95% CI 0.80-0.991). CONCLUSION: Reduced LVEF at presentation was the only reproducible and externally validated risk stratifier of MACE during index hospitalization. Two age-associated clinical patterns were observed: a high-risk "heart failure type" in younger children and low-risk "chest pain type" in adolescents.