Association of distal adenoma and hyperplastic polyp characteristics with long-term proximal colon cancer risk: a secondary, observational analysis of data from the UK Flexible Sigmoidoscopy Screening Trial

Abstract

Objectives Colorectal cancer screening with flexible sigmoidoscopy focuses on the distal colorectum, but it is unclear which distal polyp characteristics are associated with future proximal colon cancer incidence. We examined associations between distal adenoma or hyperplastic polyp characteristics and long-term incident proximal colon cancer.Methods In secondary, observational analyses of UK Flexible Sigmoidoscopy Screening Trial data, we obtained data on the number and size of distal hyperplastic polyps (n=4872) and adenomas (n=4581), adenoma histology and dysplasia from endoscopy and pathology reports for screened asymptomatic participants. Adjusted HRs and 95% CIs for the association between distal polyp characteristics and proximal colon cancer incidence were estimated using multivariable Cox proportional hazard models.Results Over a median of 20.7 years of follow-up (IQR 16.5–21.7), 110 proximal colon cancers were diagnosed among participants with distal adenomas and 96 were diagnosed among those with only distal hyperplastic polyps detected at baseline. Larger adenoma size (6–9 mm vs ≤5 mm: HR 1.67 (95%CI: 1.07 to 2.59) and ≥10 mm vs ≤5 mm: HR 2.08 (95%CI: 0.98 to 4.43); p=0.037) and high-grade (vs low-grade) adenoma dysplasia (HR 2.82, 95% CI: 1.34 to 5.93; p=0.012) at baseline were positively associated with proximal colon cancer incidence. No associations were observed for distal adenoma number overall or histology, or the number or size of hyperplastic polyps and proximal colon cancer incidence.Conclusions We found some evidence that larger distal adenomas and those with high-grade dysplasia at baseline were positively associated with proximal colon cancer incidence. Larger studies are needed to confirm these findings.Trial registration number ISRCTN28352761.