A comparative study of oral and gut microbiota for differentiating benign pulmonary nodules from lung cancer

The differentiation between benign pulmonary nodules and early-stage lung cancer remains a significant clinical challenge due to the lack of highly specific non-invasive biomarkers. While the gut-lung axis has emerged as a critical player in pulmonary diseases, the concurrent evolution of oral and gut microbiota across the spectrum from benign nodules to lung cancer remains poorly characterized. This study aimed to investigate the compositional and dynamic changes in oral and gut microbiota across different stages of lung carcinogenesis, with particular emphasis on distinguishing tumor-specific alterations from nodule-related changes by incorporating benign nodules as a critical control group.

In this retrospective case-control study, we enrolled 111 participants, including healthy controls (C, n = 36), individuals with benign pulmonary nodules (B, n = 18), treatment-naive early-stage lung cancer patients (EL, n = 20), and treatment-naive advanced-stage lung cancer patients (AL, n = 37). Saliva and fecal samples were collected for 16S rRNA gene sequencing (V3-V4 region). Microbial diversity, community composition, and differential taxa were analyzed using QIIME, LEfSe, and trend tests. Diagnostic efficacy of candidate markers was evaluated by ROC curve analysis.

Lung cancer was associated with significant remodeling of both oral and gut microbiota. While salivary α-diversity remained stable between groups, gut microbial richness (Chao1 index) was significantly increased in lung cancer patients compared to non-tumor individuals (P = 0.009). β-diversity analysis revealed distinct separation in community structure between non-tumor and lung cancer groups for both saliva and feces (P < 0.01). Trend analysis across the C→ B→ EL→AL continuum demonstrated significant gradient changes: salivary Pasteurellaceae (primarily Haemophilus) and Enterobacterales progressively decreased, while Actinobacteriota, Rothia mucilaginosa, and Prevotella oris significantly increased with disease progression. In the gut, Veillonellaceae and Dialister gradually declined, whereas Bacteroides coprocola exhibited a strong positive correlation with disease stage (ρ = 0.84, P < 0.001). The benign nodule group displayed an intermediate microbial phenotype. Notably, R. mucilaginosa and Haemophilus in saliva achieved AUCs of 0.851 and 0.884, respectively, in distinguishing non-tumor individuals from lung cancer patients, while gut B. coprocola yielded an AUC of 0.901, with maintained discriminatory performance in differentiating benign nodules from early-stage lung cancer.

Lung cancer progression is accompanied by stage-specific remodeling of the oral-gut microecology. The incorporation of benign nodule controls enables the identification of tumor-specific microbial alterations with high diagnostic efficacy. Salivary Rothia mucilaginosa and gut Bacteroides coprocola represent promising non-invasive biomarkers for early lung cancer detection and benign-malignant discrimination, warranting further validation in multicenter prospective cohorts.