Microbiome-host interactions in different pathomic subtypes of early-stage lung adenocarcinoma

BACKGROUND: Microbiome-host interactions in different pathological subtypes of the early-stage lung adenocarcinoma (LUAD) remain poorly understood. METHODS: We conducted histopathological and imaging analysis on a cohort of 106 Chinese patients with LUAD. Further deep analysis of the lung tissue microbiome, serum metabolome, and host transcriptome was performed. Using correlation analysis, microbial genetic information, and established metabolite-human gene pairs, we integrated multi-omics data to explore potential associations between the microbiome, microbial metabolites, and host gene expression. FINDINGS: We identified distinct groups based on malignancy severity. Massilia, Sphingomonas, Staphylococcus, and Brevundimonas may influence host gene expression through their metabolites, affecting the progression of LUAD. We used Mendelian Randomization(MR) and found suggestive evidence that the levels of key metabolites (serotonin, uric acid) are negatively associated with LUAD risk. INTERPRETATIONS: This study demonstrates that pathological images of early-stage LUAD cells can reflect potential clinical differences, and some microbial-metabolite-gene potential associations have been found in early-stage LUAD, which may affect the development of early-stage lung adenocarcinoma. FUNDING: This research was supported by grants from the National Natural Science Foundation of China (No. 82171931), the Guangdong Basic and Applied Basic Research Foundation Enterprise Joint Fund (2025).