Functional metagenomics highlights varied infection states with dynamics of pathogens and antibiotic resistance in lower respiratory tract infections.

BACKGROUND

Antimicrobial resistance (AMR) amongst pathogenic bacterial species poses significant challenges in treating infections of the lower respiratory tract (LRT), leading to higher hospitalization and mortality rates.

METHODS

Bronchoalveolar lavage fluid (BALF) from 84 clinically adjudicated LRTI patients were subjected to respiratory pathogen ID/AMR (RPIP) enrichment and sequencing followed by Explify and CZID seq data analysis to identify potential LRTI pathogens and associated AMR genes. Patients were categorized as LRTI-WP (with pneumonia) and LRTI-WoP (without pneumonia).

FINDINGS: mNGS achieved 100 % pathogen detection compared to 73 % through clinician-used BioFire panel. Predominant pathogens included Acinetobacter baumannii, Klebsiella pneumoniae along with detection of Aspergillus versicolor and Herpes simplex virus. Double and polymicrobial infections were captured, involving non-respiratory pathogens like Rothia mucilaginosa, Escherichia coli, and Moraxella osloensis. AMR detection highlighted macrolide (MPH; ERM) and Sulfonamide (SUL) rich resistome in 60 % of patients followed by extended spectrum beta lactamase (OXA) and tetracycline (TET). LRTI-WP showed high abundance of A. baumannii, majorly associated with MPH whereas K. pneumoniae with beta-lactams was comparable in both groups. Differences in clinical severity may stem from non-respiratory pathogens, newly recognized via mNGS. CZID seq pipeline validated and revealed additional microbes and AMR genes in the cohort.

INTERPRETATION: The prevalence of common pathogens like A. baumannii and K. pneumoniae along with the non-respiratory pathogens identified by RPIP-Explify and CZID seq provided an understanding to evaluate the LRTI. mNGS is crucial for precise pathogen and antibiotic resistance detection, vital for combating antibiotic resistance.