Primary malignant tracheal tumor (PMTT) is a rare disease, accounting for 0.5% of all malignancies and 0.01-0.4% of lung cancer cases, with approximately 1 new case per 1,000,000 reported yearly (1, 2). Histologically, PMTT mainly includes squamous cell carcinoma (SCC), and salivary gland type carcinoma (SGC) (3). Although PMTT represents a small proportion of total lung cancer cases, a large number of patients suffer from PMTT. PMTT most commonly arise from the respiratory epithelium, mesenchymal structures, or salivary glands of the trachea (3). The possible factors affecting the occurrence and development of tracheal cancer include underlying genetics, smoking, air pollution and other environmental factors. However, the exact mechanism of PMTT development has not been determined.

The lung has been historically considered a sterile environment. However, recent advances in high-throughput sequencing have shed new light on this misconception with evidence that there are abundant and diverse bacterial communities in the lower respiratory tract (LRT) dominated by Bacteroidetes, Firmicutes, and Proteobacteria phyla (4, 5). Several studies have shown that microbiota of the human lung contribute significantly to respiratory health and disease (5–7). Specifically, alterations in the composition and diversity of respiratory tract microbiota are common in lung diseases such as asthma (8), lung fibrosis (9), lung cancers (7, 10–15), and chronic obstructive pulmonary disease (16). However, respiratory microbiota is relatively unstudied compared to gut microbiota, and the tracheal microbiome under disease conditions is less studied than that in healthy subjects or subjects with laryngotracheal stenosis (17, 18). The trachea is regularly exposed to air pollution, particulate matter, tobacco smoke, and microbes, all of which can either cause or contribute to microbial dysbiosis and inflammation, consequently impacting tumorigenic processes. Therefore, further exploration of the relationship between microbiota and PMTT is needed to better understand their underlying interactions.

To date, there has been no report of which we are aware describing the composition and diversity of respiratory tract microbiota in PMTT patients. In this study, we used 16s rRNA gene sequencing to characterize the respiratory tract microbiota in patients with PMTT and compared their biogeographic distribution in the upper respiratory tract (URT), oral cavity (OC) and LRT. We hypothesized that tracheal tumors would be associated with respiratory tract dysbiosis and microbial populations will exhibit changes in their relative abundance along the respiratory tract in patients with PMTT compared with that in non-malignant tracheal tumor (NMTT) patients. We then identified significant differences in airway microbiota composition and characteristics among patients with NMTT, SGC, and SCC, as well as between different sites in the respiratory tract. In addition, we looked for correlations between specific taxa and disease states at each site to screen for candidate bacterial biomarkers associated with PMTT. Our findings provide a basis for exploration of the role of microorganisms in PMTT, as well as potential biomarkers for further development in clinical diagnostics.

To our knowledge, this report represents the first comprehensive investigation of the microbial composition and diversity of the microbiome at four different biogeographic sites (saliva, throat, trachea, and distal alveoli) in patients with PMTT. We found that the LRT microbiota are affected by oral and throat microbiota and observed that Prevotella and Veillonella species decrease in relative abundance with increasing depth into the airway. Our results also indicate that alpha diversity is lower in SCC patients compared to that in the SGC and NMTT groups. In addition, the relative abundance of Prevotella is lower in the SCC group than that in the SGC and NMTT groups, while Alloprevotella is found in the highest abundance in LRT samples of the SGC group. These significant differences in microbial diversity, and taxa abundance associated with different anatomical sites and pathological types suggests an imbalance in the microbiota of the PMTT microenvironment and implies that some taxa may participate in the pathology of PMTT.

Discrepancies in the results obtained by different studies could be attributable to differences in the number of cases, ethnicity, gender, smoking history, age, and inclusion criteria. We enrolled patients from our clinics in Shanghai and Beijing who originated from provinces throughout China to avoid geographical selection bias and employed strict inclusion and exclusion criteria to minimize the effects of confounding factors. Contaminant-controlled analysis framework in our study allowed the accurate detection of microbial DNA for sequence. We found that the airway and oral microbiota associated with tracheal tumors are primarily composed of four phyla, which is consistent with findings of previous studies (4, 27). Unlike NMTT and SGC patients, the abundance of Proteobacteria in the LRT is significantly elevated in SCC patients, while the other phyla have significantly lower relative abundance. This effect could be at least partially explained by impaired local immunity at the lesion site due to smoking, which causes massive proliferation of pathogenic bacteria. It should be noted that Proteobacteria, such as Pseudomonas aeruginosa, are reportedly involved in chronic lung disease, lung fibrosis, and asthma (28–30).

Studies in lung cancer patients and animal models have demonstrated tumor-associated dysregulation of airway microbiota, which can in turn impact cancer progression by shaping conditions in the tumor microenvironment and modulating the activity of tumor-infiltrating immune cells (10–15, 31). Our previous study found that lung cancer-associated microbiota profile is distinct from that found in healthy controls, and the altered cancer-associated microbiota is not restricted to tumor tissue (10). These results suggested considerable spatial heterogeneity in bacterial community composition within the lungs of lung cancer patients. However, the role of airway microbiota in PMTT remains largely unknown. The respiratory tract is a complex system that is exposed to a large number of airborne microorganisms and is inhabited by niche-specific communities of microbes. Moreover, the oral microbiota shape the respiratory microbiome (32), which may contribute to the carcinogenesis of PMTT via mucosal dispersion and microaspiration of oral bacteria and their metabolites to the respiratory tract. Dysbiosis of oral microbiota is closely associated with cancer, including lung, esophageal, and colorectal cancer (33–37). For instance, the abundance of the periodontal pathogen Porphyromonas gingivalis is accompanied by an elevated risk of esophageal squamous cell carcinoma (33). Other work has suggested that Peptococcus and Centipeda species in saliva and mucosal Subdoligranulum species could serve as drivers of tumorigenesis of colorectal cancer (34). Dysbiosis of the salivary microbiome is also associated with non-smoking lung cancer and correlated with systemic inflammatory marker (35).

In this study, we found that the microbiota in the OC and TS could explain 8.5%-17.6% of the variance of the lower respiratory tract microbiota. Interestingly, we found that the microbiota of OC and throat exert a statistically significant effect on microbiota of the trachea but not on distal alveoli in SGC patients. However, in SCC patients, the microbiota of the OC and throat significantly affect the tracheal and distal alveolar microbiota, and similar results were also observed in NMTT patients. SGC are most commonly found in oral cancer patients, but is rare in lung cancer patients. Although rare, tracheal SGC also originates from tumorigenesis of the minor salivary glands located in the mucous membranes of the head and neck. Tracheal SGC microbiota can be thus affected by the oral microbiota originating in the adjacent oral cavity or from submucosal salivary glands that extend down into the trachea.

In healthy individuals, the lung microbiome is generally believed to be inoculated by OC bacteria, and that the community is maintained by a balance between immigration, colonization, and competition processes (32). However, this balance is apparently disturbed during several lung diseases (5). This likelihood is supported by our finding of decreased alpha diversity in SCC patients compared with that in SGC and NMTT subjects. However, we found no consistent changes in microbial diversity of distal lung alveoli and oral cavity among different carcinoma subtypes. Reduced microbial diversity has been considered a feature of some cancers, such as esophageal adenocarcinoma, gastric carcinoma, and colorectal cancer. Our previous study showed that decreased microbial diversity in the distal lung alveoli is associated with increased risk of lung cancer (10).

The genera Prevotella and Veillonella are among the most abundant bacteria in the respiratory tract of healthy individuals and are universal commensal colonizers of mucosal surfaces in a “healthy” lung microbiome (38–40). Recent work has shown that the relative abundance of Prevotella in BALF is increased in lung cancer patients (11). Here, we also observed that the relative abundance of Prevotella increased in the SCC group. Prevotella can adhere to epithelial cells and form a biofilm through their ability to modulate host immune response and metabolic activity (41). Alternatively, Segal et al. classified the microbiota of BALF samples from healthy subjects into two types: a supraglottic predominant taxa (SPT) pneumotype and a background predominant taxa (BPT) pneumotype (42, 43). Further studies revealed that characteristic SPT taxa, including Prevotella and Veillonella, were positively correlated with multiple cytokine levels, including Th17 cytokines such as IL-1α, IL-1β, IL-6, fractalkine, and IL-17, in both Th17 cells and neutrophils (43). Thus, URT-derived bacteria and/or their metabolites in the lung can regulate the tonic level of airway inflammation and Th17 immune activation (43). Tsay et al. found that lower airway dysbiosis induced by microaspiration of oral commensals affect lung tumorigenesis by promoting an IL-17 driven inflammatory phenotype (12). Dysregulation of local lung microbiota stimulated Myd88-dependent IL-1β and IL-23 production from myeloid cells, inducing proliferation and activation of Vγ6+Vδ1+ γδ T cells that produced IL-17 and other effector molecules to promote inflammation and tumorigenesis (31).

Among these URT bacteria, Veillonella is associated with increased mucosal inflammation in the airway, and is significantly more abundant in saliva and BALF samples from lung cancer patients than healthy control subjects (14, 15). Airway microbiota can regulate specific oncogenic pathways that directly drive carcinogenesis. Veillonella is also associated with transcriptomic changes in airway epithelial cells, especially with activation of the ERK/PI3K pathway relevant to lung cancer (13). In addition to Veillonella, our results reveal that Alloprevotella is found in the highest abundance in LRT samples of the SGC group, suggesting a strong association between SGC and predominant oral bacteria. Other studies have shown that Alloprevotella likely contributes to the enrichment for proinflammatory genes, has been reported to be progressively enriched along the order of negative control → precancerous lesions → oral cancer (37). Our results are thus consistent with previous studies and further indicate that specific bacterial taxa may serve as potential biomarkers and new therapeutic targets.

Some limitations of this study should be noted. First, for ethical reasons, we did not obtain PSB or BALF samples from healthy individuals as controls, so the use of specific bacteria to distinguish different tracheal tumors types may be unreliable in some cases, resulting in false positive correlations. Second, respiratory microbial communities undergo dynamic changes, and cross-sectional studies may not reflect a comprehensive landscape of microbial features across all stages of the disease. Third, this pilot study utilizes a small cohort and therefore requires a large-scale longitudinal cohort for validation. In addition, the sample size is too small to perform risk stratification or to predict prognosis.

In this study, we investigated the composition, diversity, and differences in respiratory tract microbiota at different sites in PMTT patients. This study uncovers novel and plausible correlations between dysbiosis in microbiota and PMTT. This work also identified bacterial genera that are unique to specific sites during SCC or SGC development and can therefore serve as candidate targets for PMTT diagnosis or intervention, although large sample studies are needed to validate their reliability as biomarkers.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/genbank/, PRJNA766788.

The studies involving human participants were reviewed and approved by the Ethics Committee of Ruijin Hospital Affiliated with Shanghai Jiaotong University and Emergency General Hospital. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

K-XL and H-XL: collection and assembly of data, data analysis and interpretation, writing the original draft, review, and editing. JZ, NZ, Y-ZZ, and M-MT: Sample and data collection, data analysis. J-MQ and H-WW conception and design and administrative support, review, and editing. All authors contributed to the article and approved the submitted version.

This work was supported by grants from Shanghai Key Laboratory of Emergency Prevention, Diagnosis and Treatment of Respiratory Infectious Disease (20dz2261100), the National Natural Science Foundation of China (81570066 and 81630001), Shanghai Top-Priority Clinical Key Disciplines Construction Project (2017ZZ02014), Shanghai Shenkang Hospital Development Center Clinical Science and Technology Innovation Project (SHDC12018102), National Key Research & Development Program of China (2018YFE0102400), National Innovative Research Team of High-level Local Universities in Shanghai, Shanghai Municipal Key Clinical Specialty (shslczdzk02202) and the Young and Middle-aged Backbone Talent training plan of Fujian Provincial Health Commission (2020GGA053).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.