Abstract
BACKGROUND: Traditional diagnostic methods have inherent limitations in the comprehensive assessment of the etiological spectrum and microecological characteristics of pediatric lower respiratory tract infections (LRTIs), particularly community-acquired pneumonia (CAP). Against this backdrop, the present study seeks to delineate the pathogen profile of children with LRTIs via targeted next-generation sequencing (tNGS), and further explore the associations between clinical manifestations, upper respiratory microbiome signatures and disease severity in pediatric CAP cases. METHODS: A retrospective, single-center study was conducted on 2299 children with suspected lower respiratory tract infections. Throat swab samples from all patients underwent tNGS for pathogen detection. For 1845 CAP patients (293 SCAP, 1552 non-severe CAP [nsCAP]), clinical data and tNGS results were analyzed. Statistical comparisons, correlation analyses, and multivariate logistic regression were performed to identify factors associated with SCAP. Microbial diversity (Shannon/Simpson indices) and relative abundance of detected species were also analyzed. RESULTS: Mycoplasma pneumoniae was the dominant atypical pathogen, with an outbreak peaking in July 2024. M. pneumoniae detection rate (35.8% vs. 8.9%, P<0.001) and relative abundance (RA) were significantly higher in SCAP than nsCAP patients and correlated positively with severity markers. Multivariate analysis identified M. pneumoniae positivity, older age, female sex, circulatory and metabolic diseases as independent risk factors for SCAP. In M. pneumoniae-negative patients, pathogens like Streptococcus pneumoniae and Haemophilus influenzae were more common in nsCAP. Upper respiratory microbial diversity was lower in SCAP patients. Increased RA of specific commensals like Schaalia odontolytica was a protective factor, while increased abundance of Stenotrophomonas maltophilia was a risk factor for SCAP. Compared to bronchoalveolar lavage fluid (BALF), throat swab tNGS showed high agreement for M. pneumoniae but higher detection of potential colonizers like H. influenzae. CONCLUSIONS: During the study period, M. pneumoniae was a key driver of SCAP in children. Beyond single-pathogen detection, decreased upper respiratory microbial diversity and shifts in colonizing bacteria abundances were associated with pneumonia severity, offering a new ecological perspective. Throat swab tNGS is valuable for pathogen screening. The associations between upper respiratory microbial features and severity highlight a potential ecological dimension of pneumonia pathogenesis.