Pulmonary infections are the most common tissue-invasive infection in immunocompromised patients and can be caused by a wide range of common and opportunistic pathogens. Early etiologic diagnosis and initiation of effective treatment is critical to reduce morbidity and mortality but current tests remain negative in up to 60% of patients. Metagenomic DNA and RNA sequencing provides a hypothesis-free alternative for detection of any known pathogen with a single test. We have developed and validated a metagenomics-based test for respiratory pathogens that can generate final results within 48 hours (Explify Respiratory).
RNA and DNA from BAL samples were sequenced (~15 million sequencing reads/sample) and analyzed with a validated, diagnostic-grade, rapid data analysis software. Sensitivity and specificity for >200 respiratory pathogens were validated using >200 real and contrived BAL samples and thousands of virtual, in silico generated BAL samples. Validated protocols were then applied to banked BAL samples from immunocompromised children with pneumonia treated in the intensive care unit.
Test validation demonstrated high agreement with conventional microbiology tests (90.2% for bacterial, 94.1% for virus, 66.7% for fungal detection) and >98.8 accuracy and specificity with virtual BAL samples. Previously missed putative pathogens were identified in 18 of 41 immunocompromised children (44%) with life-threatening suspected pulmonary infections, including 7 of 11 children (64%) with fatal infections. The NGS test detected a single pathogen in 12 (63%), 2 pathogens in 5 (26%), and 4 pathogens in 1 (5%) patient.
Hypothesis-free, metagenomics-based pathogen detection can provide rapid etiologic diagnosis in seriously-ill children with negative conventional tests and may help early administration of targeted antibiotic therapy. Explify Respiratory is now available at our national reference laboratory.