Compartmentalization of Host Multi-Omics Responses in Lower Respiratory Tract Infection: Discordance Between Airway-Local and Systemic Molecular Profiles—A Narrative Review

Lower respiratory tract infection (LRTI) remains a leading cause of infection-related mortality. Peripheral blood multi-omics shows promise for pathogen discrimination and severity prediction, yet how well blood reflects airway-local responses in LRTI other than coronavirus disease 2019 (COVID-19) is unclear. We searched PubMed, Web of Science Core Collection, and Europe PMC (January 2021 to August 2026); after de-duplication and dual screening, 29 of 817 unique records were included (Layer A: ≥2 host functional omics layers, n = 23; Layer B: single omics platform with paired airway–blood specimens, n = 6), plus four COVID-19 benchmarks. Blood-based multi-omics distinguished bacterial, viral, and Mycoplasma pneumoniae (M. pneumoniae) LRTI; severity models performed promisingly within their discovery cohorts. In the two studies reporting airway-versus-blood protein inventories, the airway showed more differentially expressed proteins (DEPs) than blood (airway-to-blood DEP count ratios ≈ 3.6–9.3; descriptive study-specific measures), and cross-compartment correlations were predominantly weak (r ≈ −0.3 to +0.3) in the only paired viral LRTI study; interferon-stimulated gene 15 (ISG15; r = 0.79) was a notable exception. Paired evidence derives mostly from a pediatric cohort, hospital-acquired pneumonia, and acute respiratory distress syndrome (ARDS); adult community-acquired LRTI remains under-represented. Current evidence suggests that peripheral blood provides an incomplete, context-dependent representation of airway-local responses; closing this gap requires adult longitudinal paired cohorts, non-invasive airway sampling, improved integration algorithms, and multimodal data fusion.