Effect of the oral microbiota, blood metabolome, and inflammatory proteins on oral cavity cancer: A bidirectional two-sample Mendelian randomization study and mediation analysis

Abstract

Objective

Oral Cavity Cancer (OCC) pathogenesis is complex, extending beyond traditional risk factors. While observational studies link oral microbiome dysbiosis, metabolic disturbances, and inflammation to OCC, inherent confounding limits causal inference regarding the putative 'microbiome-metabolite-inflammation' axis in OCC. Establishing causality is crucial.

Methods

We employed a two-sample Mendelian randomization (MR) framework using large-scale GWAS data to address this gap. We systematically evaluated causal effects of 43 oral microbial taxa, 1400 diverse circulating metabolites, and 91 inflammatory proteins on OCC risk. We performed univariable MR (UVMR) for direct effects, multivariable MR (MVMR) adjusting for interactions, and mediation MR dissecting causal pathways.

Results

UVMR identified protective effects for Clostridiales (OR = 0.89) and Rothia sp. ASV0016 (OR = 0.91), and increased risk for Bacteroidales (OR = 1.09). Furthermore, 60 metabolites (e.g., glycohyocholate increasing risk; 16α-hydroxy DHEAS-3-sulfate decreasing risk) and two proteins (Cystatin D increasing risk, OR = 1.26; MCP-1 decreasing risk, OR = 0.69) showed causal links to OCC. Crucially, mediation analyses indicated protective microbial effects were partially mediated via specific metabolites, including 5α-androstan-3α,17β-diol disulfate (Clostridiales) and carboxyethyl-GABA (Rothia sp.).

Conclusions

This study provides robust genetic evidence supporting causal roles for specific oral microbes and metabolites in OCC etiology. It offers mechanistic insights into the 'oral microbiome-host metabolism' axis, providing a basis for novel microbiome/metabolite-based biomarkers for early detection and risk assessment, and identifying potential preventative or therapeutic targets.