Twenty-nine patients met inclusion criteria. Of 28 (97%) patients who underwent testing for viruses, polymerase chain reaction was used in 24 of 28 (86% of cases), and 16 of 24 (67%) detected a virus. Pathogens were rhinovirus (6),
influenza A/B (4), respiratory syncytial virus (RSV) (3), coronavirus (3), parvovirus B19 (2), adenovirus (2), and coxsackie B5 virus, enterovirus, and parainfluenza virus type 2 in one case each. Six (21%) patients had no pathogen detected but imaging and other laboratory test results were compatible with
myocarditis. Age 0-2 years was associated with RSV,
influenza A/B, coronavirus, and enteroviruses (P < .001). Twenty-one patients (72%) experienced full clinical recovery. Three patients (10%) required
venoarterial extracorporeal membrane oxygenation (VA-ECMO), and all 3 recovered. Three others (10%) required and underwent successful
cardiac transplantation without complications. Two patients (7%) died 9-10 days after hospitalization (1 had RSV and 1 had
influenza A/B). Two other patients presented with complete
atrioventricular block; 1 case (rhinovirus) resolved spontaneously, and 1 (coronavirus) resolved after support with VA-ECMO. Age <2 years, female sex, lower ejection fraction at admission, and greater initial and peak levels of
brain natriuretic peptide were significant predictors of critical outcomes (use of VA-ECMO, listing for
cardiac transplantation, and death).
CONCLUSIONS: Viral
nucleic acid-based testing revealed a wider spectrum of viruses that could be associated with
myocarditis in children than previously reported and traditionally anticipated. A predilection of certain pathogens in the very young patients was observed. Whether the observed range of viral agents reflects an undercurrent of change in viral etiology or viral detection methods is unclear, but the wider spectrum of viral pathogens found underscores the usefulness of polymerase chain reaction testing to explore possible viral etiologies of
myocarditis in children.