Abstract
Observational studies found that phosphodiesterase 5 (PDE5) inhibitors use is linked to both increased and decreased risk of cancer; while the causal relationship remains unclear. To clarify whether PDE5 inhibitors medication may affect the risk of cancer, 2-sample cis-Mendelian randomisation (MR) analysis was therefore performed. Uncorrelated (linkage disequilibrium [LD] r(2) < 0.001) single-nucleotide polymorphisms (SNPs) in PDE5A gene associated (P < 5.0 × 10(-8)) with circulating levels of PDE5A protein identified from UKB-PPP were used as genetic instrument to mimic the action of PDE5 inhibition. Summary-level data for 22 types of cancer obtained from site-specific GWAS were analyzed in discovery stage (428,361 cancer cases) and then replicated in the FinnGen study (87,505 cancer cases). Inverse-variance weighted random-effects models were used as primary analysis. After multiple testing correction, genetically predicted, per-standard deviation (SD) decrease in PDE5A protein was associated with decreased risk of colorectal cancer with a pooled odds ratio (OR) of 0.80 (95% confidence interval [CI]: 0.75-0.86; P = 6.15 × 10(-11)). A significant MR association (OR = 0.48, 95% CI: 0.34-0.68; P = 4.70 × 10(-5)) with gastric cancer (GC) was also observed in combined analysis. There was little evidence to support associations between genetically proxied PDE5 inhibition and other 20 studied cancers. We found an protective effect of genetically proxied PDE5 inhibition on CRC and GC risk. Our drug target MR analyses provide genetic evidence in predicting long-term safety profiles of PDE5 inhibitors on cancer risk and highlight the potential of drug repurposing in CRC and GC.