Background: Viral infections are a major cause of inflammatory heart disease, but conventional polymerase chain reaction (PCR) often fails to detect the causative pathogens, limiting diagnostic and therapeutic decisions. We investigated whether targeted metagenomic next-generation sequencing (NGS) improves virus detection in endomyocardial biopsies (EMB) compared to standard PCR, and examined the clinical implications of undetected viral infections in heart failure.
Methods: EMB samples from 108 patients with unexplained heart failure underwent histologic and immunohistochemical analysis to assess myocardial inflammation. All samples were tested for common cardiotropic viruses using PCR. Targeted metagenomic NGS was performed in 36 PCR-positive and 72 PCR-negative cases. Virus prevalence and transcriptional activity were compared with PCR results and correlated with myocardial inflammation and left ventricular ejection fraction (LVEF).
Results: NGS identified all 45 viruses previously detected by PCR and revealed 31 additional viral genomes in PCR-positive patients, increasing diagnostic yield by 69%. In PCR-negative patients, NGS identified viral genomes in 56% of cases, uncovering a significant number of previously undiagnosed infections. Frequently detected viruses included parvovirus B19, Epstein-Barr virus, human herpesvirus 6, cytomegalovirus, and adenovirus, as well as less expected agents such as herpesvirus 7/8, adenovirus-associated virus, and pegivirus C. Transcriptionally active parvovirus B19 was more often detected by NGS than PCR (31% vs. 14%). Patients with NGS-confirmed viral infections showed significantly reduced LVEF compared to virus-negative individuals.
Conclusions: Targeted metagenomic sequencing substantially improves virus detection in EMB samples and reveals clinically relevant infections missed by PCR. These results support the integration of NGS into diagnostic workflows for virus-associated heart failure in order to better guide clinical management.
