A Two-Sample Mendelian Randomisation Analysis of the Oral Microbiome and Oral/Oropharyngeal/Tongue Cancers

To assess the potential causal relationships of the oral microbiome with the risks of oral cancer, oropharyngeal cancer, and tongue cancer using two-sample Mendelian randomisation (MR) analysis, while distinguishing these from reverse causal effects of the cancers on microbial abundance.

Using single-nucleotide polymorphisms as instrumental variables, we applied the MR inverse-variance-weighted approach to evaluate the effects of the dorsal-tongue and salivary microbiomes on oral, oropharyngeal, and tongue cancers. Analyses were conducted with the R package TwoSampleMR, leveraging genome-wide association study (GWAS) summary statistics from CNGBdb, the FinnGen consortium, and other sources. Sensitivity, heterogeneity, and pleiotropy assessments were performed. Additionally, reverse MR sensitivity analyses were conducted to explore the possible causal influence of cancers on the oral microbiota.

Using a single nucleotide polymorphism (SNP) significance threshold of p 5 × 10-6, our large-scale MR study revealed genetically supported causal relationships between microbial taxa derived from saliva and the tongue and the risk of oral, oropharyngeal, and tongue cancers. Integrating these results, we found that both 's Veillonella_rogosae_mgs_2008' and 's unclassified_mgs_1048' conferred a reduced risk of oropharyngeal and tongue cancers. Sensitivity analyses based on heterogeneity tests and pleiotropy evaluations further corroborated the robustness of our findings, lending additional credibility to the conclusions.

This study leveraged large-scale publicly available genetic data and identified significant causal relationships between the oral microbiota and cancers of the oral cavity, oropharynx, and tongue. Reverse MR analyses indicated that oral and tongue cancers may in turn alter the abundance of specific oral microbes, suggesting a potential bidirectional causal loop. Future work should integrate metagenomic data to further validate these microbiota-cancer associations.