Prostaglandin E2 Metabolite and Colorectal Cancer: Genetic Determinants and Lifestyle Modifiers

Abstract

Colorectal cancer (CRC) is a major global health burden. Urinary prostaglandin E2 metabolite (PGE-M) has been associated with CRC risk in observational studies. However, causal interpretation has been limited by potential confounding, selection bias, and reverse causation. In this dissertation, I examined the potential causal role of PGE-M in CRC development and the influence of modifiable lifestyle factors on PGE-M levels. In Aim 1, I conducted a genome-wide association study (GWAS) of urinary PGE-M levels among approximately 8,200 cancer-free participants from eight studies and identified two genome-wide significant loci represented by rs9302064 near the ABCC4 and rs1501101 in the EMCN genes previously implicated in PGE2 signaling and CRC development. In addition, I identified suggestive association of PGE-M levels with 54 variants across 13 loci in Asian participants, 80 variants across 11 loci in Europeans, and 145 variants across 22 loci in the combined analysis of both populations at P < 10-5. Mendelian randomization (MR) analyses using the PGE-M associated variants identified in Aim 1 and data from the largest CRC GWAS conducted to date (100,204 cases; 154,587 controls) provided a suggestive evidence for potential causal association between PGE-M levels and CRC risk. In Aim 2, I developed dietary and lifestyle scores predictive of urinary PGE-M using Elastic Net regression models. These dietary and lifestyle scores were significantly associated with CRC risk and mortality, especially among Asian men. Together, these findings provide new insights into a potential etiologic role of PGE-M in CRC and suggested that urinary PGE-M could be targeted for CRC prevention via lifestyle modifications.