Application of Mendelian randomization to explore metabolic pathways in colorectal cancer

We conducted a Mendelian randomization (MR) study to investigate the relationship between genetically predicted metabolites and the risk of colorectal cancer (CRC) and to explore the underlying pathways.

Genetic instruments for metabolite levels were selected based on data from 64 genome-wide association studies involving a total of 362 750 individuals. Using a two-sample MR approach, we assessed associations with CRC utilizing summary statistics from a meta-analysis of the UK Biobank and FinnGen. The primary analysis was conducted using the inverse-variance weighted method, with additional sensitivity analyses employing the median-weighted and MR-Egger methods to account for potential pleiotropy. The identified significant metabolites were further analyzed through enrichment analysis to explore the metabolic and lipid pathways involved.

Across the three MR methods, we identified 67 metabolites that were positively associated and 92 metabolites that were inversely associated with CRC risk. Among these, 7-methylguanine and creatinine are the most strongly connected metabolites. Enrichment analysis revealed that positively associated metabolites were significantly linked to pathways such as small nuclear ribonucleoprotein assembly, non-coding RNA metabolism, and gut-liver indole metabolism. In contrast, inversely associated metabolites were enriched in pathways related to the urea cycle, amino group metabolism, pyrimidine metabolism disorders, and pyrimidine catabolism.

This study highlights genetically predicted metabolites associated with CRC risk, suggesting that metabolite panels and lipid-based biomarkers have potential for CRC risk assessment and early detection. However, further standardization and extensive validation are necessary before its clinical application.