Oral Microbiome and Subsequent Risk of Head and Neck Squamous Cell Cancer

Abstract

Importance: The oral microbiota may be involved in development of head and neck squamous cell cancer (HNSCC), yet current evidence is largely limited to bacterial 16S amplicon sequencing or small retrospective case-control studies.

Objective: To test whether oral bacterial and fungal microbiomes are associated with subsequent risk of HNSCC development.

Design, setting, and participants: Prospective nested case-control study among participants providing oral samples in 3 epidemiological cohorts, the American Cancer Society Cancer Prevention Study II Nutrition Cohort, the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial, and the Southern Community Cohort Study. Two hundred thirty-six patients who prospectively developed HNSCC were identified during a mean (SD) of 5.1 (3.6) years of follow-up. Control participants who remained HNSCC free were selected by 2:1 frequency matching on cohort, age, sex, race and ethnicity, and time since oral sample collection. Data analysis was conducted in 2023.

Exposures: Characterization of the oral bacterial microbiome using whole-genome shotgun sequencing and the oral fungal microbiome using internal transcribed spacer sequencing. Association of bacterial and fungal taxa with HNSCC was assessed by analysis of compositions of microbiomes with bias correction. Association with red and orange oral pathogen complexes was tested by logistic regression. A microbial risk score for HNSCC risk was calculated from risk-associated microbiota.

Main outcomes and measures: The primary outcome was HNSCC incidence.

Results: The study included 236 HNSCC case participants with a mean (SD) age of 60.9 (9.5) years and 24.6% women during a mean of 5.1 (3.6) years of follow-up, and 485 matched control participants. Overall microbiome diversity at baseline was not related to subsequent HNSCC risk; however 13 oral bacterial species were found to be differentially associated with development of HNSCC. The species included the newly identified Prevotella salivae, Streptococcus sanguinis, and Leptotrichia species, as well as several species belonging to beta and gamma Proteobacteria. The red/orange periodontal pathogen complex was moderately associated with HNSCC risk (odds ratio, 1.06 per 1 SD; 95% CI, 1.00-1.12). A 1-SD increase in microbial risk score (created based on 22 bacteria) was associated with a 50% increase in HNSCC risk (multivariate odds ratio, 1.50; 95% CI, 1.21-1.85). No fungal taxa associated with HNSCC risk were identified.

Conclusions and relevance: This case-control study yielded compelling evidence that oral bacteria are a risk factor for HNSCC development. The identified bacteria and bacterial complexes hold promise, along with other risk factors, to identify high-risk individuals for personalized prevention of HNSCC.

Figure 1.. Oral Bacteria and Risk of HNSCC

A, Fold changes and 95% CIs for the association between selected taxa and risk of head and neck squamous cell cancer (HNSCC) from random-effects meta-analysis of analysis of compositions of microbiomes with bias correction abundance estimates of the 3 cohorts (American Cancer Society Cancer Prevention Study II Nutrition Cohort [ACS-CPS-II]; Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial [PLCO]; and Southern Community Cohort Study [SCCS]). Complete results for all bacteria are shown in eTable 4 in Supplement 1. B, Fold changes and 95% CIs for the association between selected taxa and risk of HNSCC for the 3 individual cohorts. C, Odds ratios and 95% CIs for the association between a priori periodontal pathogen complexes and risk of HNSCC from random-effects meta-analysis of logistic regression estimates for the 3 cohorts. The red complex includes Treponema denticola, Porphyromonas gingivalis, and Tannerella forsythia and the orange complex includes Fusobacterium nucleatum, Prevotella intermedia, Prevotella nigrescens, Eubacterium nodatum, Campylobacter showae, and Campylobacter gracilis. D, Odds ratios and 95% CIs for the association between red and orange periodontal pathogen complexes and risk of HNSCC for the 3 cohorts. All models were adjusted for age, sex, race and ethnicity, smoking status, number of cigarettes per day for those reporting ever smoking, alcohol consumption status, grams of ethanol consumed per day for those reporting alcohol consumption, and oral human papillomavirus 16 status as covariates. Fold change values greater than 1 indicate enrichment in HNSCC case participants, while values less than 1 indicate enrichment in control participants. The size of the points indicates relative abundance.

Figure 2.. Cladogram Representation of Oral Microbiome Species Associated With HNSCC

The color of nodes and branches represents fold changes in head and neck squamous cell cancer (HNSCC) case participants compared with control participants, ranging from blue to red. A red branch indicates a taxon enriched in case participants with HNSCC and a blue branch indicates a taxon enriched in control participants. Thirteen differentially abundant species are highlighted and labeled on the outer ring. Oral bacterial periodontal pathogenic complexes are highlighted with yellow shading and labeled in the top left corner of the cladogram. The intensity of magenta shading corresponds to the relative abundance of bacterial species. A total of 320 species are included in the cladogram, representing taxa with at least 2 sequences in at least 10% of participants and with mean relative abundance of 0.001% or greater. The cladogram was created using GraPhlAn.

Figure 3.. Oral Fungi and Risk of HNSCC

A, Fold changes and 95% CIs for the association between fungal taxa and risk of head and neck squamous cell cancer (HNSCC) from random-effects meta-analysis of analysis of compositions of microbiomes with bias correction abundance estimates of the 3 cohorts (American Cancer Society Cancer Prevention Study II Nutrition Cohort [ACS-CPS-II]; Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial [PLCO]; and Southern Community Cohort Study [SCCS]). Complete results for all fungi detected are shown in eTable 5 in Supplement 1. B, Fold changes and 95% CIs for the association between fungal taxa and risk of HNSCC for the 3 individual cohorts. All models were adjusted for age, sex, race and ethnicity, smoking status, number of cigarettes per day for those reporting ever smoking, alcohol consumption status, grams of ethanol consumed per day for those reporting alcohol consumption, and oral human papillomavirus 6 status as covariates. Fold change values greater than 1 indicate enrichment in HNSCC case participants, while values less than 1 indicate enrichment in control participants. The figure presents results for all Candida species and for other fungi of 1% or greater abundance. Relative abundance and fold changes with 95% CIs for fungal taxa are shown next to the plot.