Background: Differences in biological and clinical characteristics of colorectal cancer (CRC) by subsite, and the reduced effectiveness of screening to reduce proximal colon cancer compared to distal CRC, highlights the need for research into factors associated with proximal colon cancer risk.
Methods: I investigated the association between polyp characteristics detected during flexible sigmoidoscopy or colonoscopy and proximal colon cancer incidence using trial and routine hospital data, respectively. I also examined associations between diet, lifestyle and medication use factors and proximal colon cancer incidence by undertaking a systematic review (medications only) and an exposure-wide association study (EWAS) in UK and European cohorts.
Results: The number of distal adenomas was positively associated with proximal colon cancer in both asymptomatic individuals and those referred for colonoscopy. Among the latter, multiple, tubulovillous/villous and proximal adenomas were associated with an increased risk of proximal colon cancer, whereas high-grade adenoma dysplasia was associated with increased distal CRC incidence. Of the medications studied, aspirin use was inversely associated with proximal colon cancer in the systematic review and among men in the EWAS, where no association with distal colon cancer was found. No dietary variables were found to be associated with proximal colon cancer in the EWAS, despite associations with distal colon and rectal cancer. Lifestyle factors, specifically body size and fatness and sedentary behaviour were positively associated with proximal colon cancer whereas higher levels of physical activity was protective. These lifestyle associations were not found for rectal cancer. All observed associations were found among men in both datasets but only among women in the European cohort with longer follow-up.
Conclusions: This thesis provides evidence of associations between clinical and behavioural factors for proximal colon cancer and of heterogeneity compared to other CRC subsites. These findings are important to inform screening and surveillance guidance and risk reduction strategies.