Introduction Early detection and removal of premalignant colorectal polyps with a high potential to progress to invasive cancer is important for incidence reduction. However, there is evidence that cancers in the proximal colon tend to be detected later than other subsites resulting in more advanced stage at diagnosis and lower survival. This study examined which polyp characteristics were independently associated with proximal colon cancer incidence.

Methods Data were used from the All Adenomas study, which examined endoscopy and associated pathology data on ~30,000 individuals with at least one adenoma identified. Eligible participants underwent colonoscopy between 1984 and 2010 in one of 17 UK hospitals. Polyp characteristics at baseline colonoscopy, including number, size, histology, grade and location were obtained from the database. Cox regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for incidence of proximal colon adenocarcinoma. Time since baseline colonoscopy was used as the underlying time scale. HRs were mutually adjusted for polyp characteristics in addition to demographic- and colonoscopy-related confounders.

Results Of the 27,812 (42.4% female) participants included in the analysis, 227 (0.82%) developed proximal colon cancer during a median follow-up of 9 years. Cumulative incidence over 15 years was 1.4% (95% CI: 1.2% - 1.6%). Proximal colon cancer incidence was higher among participants with ≥1 adenoma in the proximal colon at baseline, either solely or in addition to distal adenomas, compared to patients with only distal adenomas (HR 1.95, 95% CI: 1.46 – 2.62). The risk was also higher among those with ≥3 adenomas compared to those with <3 adenomas at baseline (HR 1.47, 95% CI: 1.04 – 2.08) and those with adenomas ≥10 mm compared to those with adenomas <10 mm (HR 1.47, 95% CI: 1.07 – 2.01). Neither adenoma histology nor grade were independently associated with the outcome.

Conclusions Adenoma location, number and size are informative of subsequent proximal colon cancer. This study provides evidence needed to identify individuals at high risk for proximal colon cancer who would require post-polypectomy colonoscopy surveillance for the early detection and removal of cancer and precancerous lesions in this subsite.