Mendelian randomization analysis to investigate the gut microbiome in oral and oropharyngeal cancer

Background Evidence supports an observational association between the gut microbiome and susceptibility to extraintestinal cancers, but the causal relationship of this association remains unclear. Methods To identify the specific causal gut microbiota of oral and oropharyngeal cancer, we performed two-sample Mendelian randomization (MR) analysis of gut microbiota on oral and oropharyngeal cancer using a fixed-effects inverse-variance-weighted model. Gut microbiota across five different taxonomical levels from the MiBioGen genome-wide association study (GWAS) were used as exposures. Oral cancer, oropharyngeal cancer and a combination of the two cancers defined from three separate data sources were used as the outcomes. Odds ratios (ORs) and 95% confidence intervals (CIs) for disease per standard deviation (SD) higher abundance of microbiome. Results & Conclusions There was little evidence for a causal effect of gut microbiota on oral and oropharyngeal cancer when using a genome-wide p-value threshold for selecting instruments. Secondary analyses using a more lenient p-value threshold indicated that there were 90 causal relationships between 58 different microbial features but that sensitivity analyses suggested that these were possibly affected by violations of MR assumptions and were not consistent across MR methodologies or data sources and therefore are also to unlikely reflect causation. These findings provide new insights into gut microbiota-mediated oral and oropharyngeal cancers and warrant further investigation.