Back

Microbial Cell-Free DNA Sequencing for Diagnosing Lung Infections in Immunocompromised Patients

Brick, K.; Jarman, K. H.; Nolte, F.; Vilfan, I. A.; Vaughn, M. L.; Zielinska, A.; Portnoy, V.; Perkins, B. A.; Bercovici, S.; Blauwkamp, T. A.; Goldman, J. D.; Hill, J. A.; Bergin, S. P.

2025-11-27 infectious diseases
10.1101/2025.11.24.25340574 medRxiv
Show abstract

RationaleRapid and comprehensive detection of respiratory pathogens is essential for immunocompromised patients with suspected pneumonia. Conventional diagnostic methods are limited in scope, often slow, and lack sensitivity, leading to reduced diagnostic yield. ObjectivesTo analytically and clinically validate a microbial cell-free DNA (mcfDNA) sequencing assay for bronchoalveolar lavage fluid (BALF) that enables broad and sensitive detection of pathogens causing pneumonia in immunocompromised patients. MethodsThe BALF-mcfDNA assay detects more than 500 bacteria, DNA viruses, fungi, and parasites. Analytical validation used 324 contrived samples containing DNA from 11 representative organisms and 1,294 in-silico samples. Clinical validation compared BALF-mcfDNA sequencing with usual care (UC) testing in 118 immunocompromised adults with suspected pneumonia. An independent expert panel adjudicated the etiologic diagnoses. Measurements and Main ResultsBALF-mcfDNA testing demonstrated high inclusivity (97%), exclusivity (96%), and precision (coefficient of variation <15%), with linear quantification across four orders of magnitude. Limits of detection were consistent across taxa, GC content, and genome size. In clinical evaluation, BALF-mcfDNA sequencing identified probable pneumonia pathogens in 34% of participants versus 24% for UC testing, increasing diagnostic yield by 43% (57% when combined with UC). In 24% of concordant cases, mcfDNA sequencing provided additional clinically relevant information by identifying new or species-level pathogens. ConclusionsBALF-mcfDNA sequencing enables sensitive, comprehensive, and rapid pathogen detection from BALF specimens, outperforming conventional methods and improving etiologic diagnosis of pneumonia in immunocompromised hosts.

Published in Respiratory Research · training set

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.