Alterations of the Human Lung and Gut Microbiomes in Non-Small Cell Lung Carcinomas and Distant Metastasis

Abstract

Non-small cell lung cancer (NSCLC) is the leading cause of cancer-related deaths worldwide. Although dysbiosis of the lung and gut microbiota have been associated with NSCLC, their relative contributions are unclear; in addition, their roles in distant metastasis (DM) are still illusive. We recruited in total 121 participants, including 87 newly diagnosed treatment-naive NSCLC patients of various stages and 34 healthy volunteers, and surveyed their fecal and sputum microbiota. We compared the microbial profiles between groups, identified microbial biomarkers, and generated machine learning models for distinguishing healthy individuals from patients with NSCLC and patients of various stages. We found significant perturbations of gut and sputum microbiota in patients with NSCLC and DM. A machine learning model combining both microbiota (combined model) performed better than an individual data set in patient stratification, with the highest area under the curve (AUC) value of 0.896. Sputum and gut microbiota both contributed to the combined model; in most cases, sputum-only models performed similar to the combined models. Several microbial biomarkers were shared by both microbiotas, indicating their similar roles at distinct body sites. Microbial biomarkers of distinct disease stages were mostly shared, suggesting biomarkers for DM could be acquired early. Furthermore, Pseudomonas aeruginosa, a species previously associated with wound infections, was significantly more abundant in brain metastasis, indicating that distinct types of DMs could have different microbes. Our results indicate that alterations of the sputum microbiota have stronger relationships with NSCLC and DM than the gut and strongly support the feasibility of metagenome-based noninvasive disease diagnosis and risk evaluation. (This study has been registered at ClinicalTrials.gov under registration no. NCT03454685). IMPORTANCE Our survey on gut and sputum microbiota revealed that both were significantly disturbed in non-small cell lung cancer (NSCLC) and associated with distant metastasis (DM) while only the sputum microbiota was associated with non-DM NSCLC. The lung microbiota could therefore have a stronger association with (and thus may contribute more to) disease development than the gut microbiota. Mathematic models using both microbiotas performed better in patient stratification than using individual microbiota. Sputum models, however, performed similar to the combined models, suggesting a convenient, noninvasive diagnostic for NSCLC. Microbial biomarkers of distinct disease stages were mostly shared, suggesting that the same set of microbes were underlying disease progression, and the signals for distant metastasis could be acquired at early stages of the disease. Our results strongly support the feasibility of noninvasive diagnosis of NSCLC, including distant metastasis, are of clinical importance, and should warrant further research on the underlying molecular mechanisms.

Keywords: NSCLC; brain metastasis; distant metastasis; gut microbiota; lung microbiota; machine learning; patient stratification.

Figures

FIG 1

Sputum and gut microbiota differed significantly in terms of alpha- and beta-diversities. (A) Numbers of sputum (red) and gut (blue) samples collected in this study and their distributions in healthy controls and distinct disease stage groups; CON, healthy controls; I_III, patients with stages of I to III; DM, patients with distant metastasis (also referred to as stage IV). Disease stages were assigned according to the 8th American Joint Committee on Cancer guidelines. (B) Comparisons of alpha-diversity and beta-diversity between the sputum and gut in healthy controls. Richness index (alpha-diversity; left) at the genus level was significantly lower in feces; principal-coordinate analysis (PCoA; right) based on Bray-Curtis distance at the genus level showed that the overall microbiota composition was different between fecal and sputum samples. Wilcoxon rank sum tests were used to compare between groups. (C) Comparisons of alpha-diversity (left) and beta-diversity (right) between sputum and gut in NSCLC patients (stages I to IV). Level of significance: ***, P < 0.001; **, P < 0.01; *, P < 0.05; NS, P ≥ 0.05.

FIG 2

Global alteration of the sputum microbiota was associated with NSCLC and distant metastasis (A, B), while the fecal microbiota was only significantly associated with the latter (C, D). (A) Significant differences were found in alpha-diversity between healthy controls and individuals with NSCLC. Richness index, Shannon index, evenness index and Simpson index at the genus level were significantly lower in patients than in healthy controls. A Wilcoxon rank sum test was used to compare between groups. Level of significance: ***, P < 0.001; **, P < 0.01; *, P < 0.05; NS, P ≥ 0.05. (B) Alpha-diversity of sputum dysbiosis in pairwise comparisons. The richness index (top left), Shannon index (top right), and evenness index (bottom left) are shown. The Shannon index and richness index were significantly lower in patients than in healthy controls. An ANOVA with a post hoc Tukey HSD test was used to compare between groups. Level of significance: ***, P < 0.001; **, P < 0.01; *, P < 0.05; NS, P ≥ 0.05. (C) Significant differences were found in beta-diversity between controls and individuals with NSCLC as well as between controls and I_III, controls and DM, and I_III and DM groups, indicating that dysbiosis of the sputum microbiota was associated with lung cancer development and metastasis. Conversely, applying similar analyses to fecal samples, no differences in alpha-diversities were apparent (D), but the beta-diversity in controls versus DM and I_III versus DM were different (P = 0.07), suggesting that fecal microbiota dysbiosis was associated with distal metastasis but not NSCLC.

FIG 3

Shared and distinct microbial biomarkers between subject groups in sputum (red) and fecal (blue) microbiota. Differentially abundant microbial biomarkers between subject groups were identified using LEfSe analyses; red and blue bar plots indicate LEfSe results for sputum and fecal microbiota, respectively. The relative abundance of 18 and 9 genera was significantly different between NSCLC and control groups in sputum (A) and feces (B), respectively. To identify biomarkers for specific disease stages, we compared neighboring groups along the disease progression in sputum (C) and feces (D). LDA, linear discriminant analysis.

FIG 4

The top-ranked genera of the mixed models shown in Table 2 for each disease stage. The genera were ranked by the robustness of 1,000 repeats; therefore, box plots were used here to demonstrate the means and distributions of these values. (A) Control versus NSCLC. (B) Control versus I_III. (C) I_III versus DM. (D) Control versus DM. Red boxes represent sputum-derived genera, and blue boxes represent gut-derived genera. Please consult Table 2 for details on the model performance. F, feces; S, sputum.

FIG 5

Patients with brain metastasis differed significantly from other distant metastasis in microbial profiles of the sputum and feces. (A) Numbers of sputum (red) and gut (blue) brain metastasis samples (left); BM, NSCLC patients in stage IV with brain metastasis; non-BM, stage IV NSCLC patients without brain metastasis. A principal-coordinate analysis showed differences in beta-diversity between BM and non-BM in sputum (middle) but not in the gut (right). LEfSe (left) analysis and Wilcoxon rank sum test (right) of differentially abundant microbial biomarkers between BM and non-BM in sputum (B) and the gut (C). Level of significance: ***, P < 0.001; **, P < 0.01; *, P < 0.05; NS, P ≥ 0.05. The star indicates that the genus Pseudomonas was significantly different in abundance.

FIG 6

Brain metastasis classification based on taxonomic profiles of the sputum, gut, and both. (A) The classification performance using the relative abundance of genera as the area under the ROC curve between BM and non-BM. (B) The top 20 genera important to the sputum model ranked by the median values of 1,000 repeats. Box plots were used here to demonstrate the medians and distributions of these values. The star indicates that the genus Pseudomonas was significantly different in abundance using LEfSe and Wilcoxon analysis.