3E-03 - Statins and Risk of Gastrointestinal Cancer: A Mendelian Randomization Study

Background: Observational studies have suggested that statins may reduce the risk of gastrointestinal (GI) cancers. However, confounding factors and reverse causation limit the ability of conventional epidemiological methods to determine whether this association is causal. Mendelian randomization (MR), which uses genetic variants as proxies for an exposure, can provide less biased estimates of causality.

Objectives: To estimate the causal effect of statin therapy on GI cancer risk using a two-sample MR design.

Methods: A two-sample MR approach was employed using large-scale genome-wide association study (GWAS) datasets. Single nucleotide polymorphisms (SNPs) strongly associated with lower LDL cholesterol (p < 5.10^-8) were selected as instrumental variables to approximate statin therapy. For the outcome, we obtained GWAS summary statistics on GI cancer (encompassing cancers of the esophagus, stomach, liver, and colon). Genetic associations with GI cancer were harmonized to ensure consistent effect alleles. We then applied inverse-variance weighted MR as the primary analysis. Sensitivity analyses included MR-Egger regression, weighted median estimation, heterogeneity testing, and leave-one-out analyses to assess potential pleiotropy and outlier effects.

Results: Genetically proxied statin-induced LDL lowering showed no significant associations with any of the four GI cancer outcomes. Specifically, for esophageal cancer the aOR was 1.01 (95% CI, 0.92–1.11), for gastric cancer 1.02 (95% CI, 0.91–1.14), for liver cancer 0.98 (95% CI, 0.87–1.09), and for colon cancer 1.00 (95% CI, 0.93–1.07). Sensitivity analyses using MR-Egger and weighted median methods yielded similarly null findings, with Egger intercept tests indicating no strong evidence of directional pleiotropy (all p>0.05). Heterogeneity measures based on Cochran’s Q likewise showed no substantial inconsistency across SNP-specific estimates (p>0.05 for all), and leave-one-out analyses did not identify any single variant as disproportionately influential.

Conclusions: This study does not provide evidence that statin-mediated LDL cholesterol reduction causally lowers the risk of GI cancer. Observed protective associations in non-randomized epidemiological studies may be driven by confounding factors. Although further research with larger sample sizes and diverse populations is warranted, our findings suggest that any protective benefits of statins on GI cancer risk may be more limited than previously assumed. Future studies should investigate alternative biological pathways and incorporate individual-level genetic data to clarify the nuanced relationship between lipid metabolism and carcinogenesis.