PROXIMAL POLYP LOCATION IS ASSOCIATED WITH HIGHER INCIDENCE OF POST COLONOSCOPY COLORECTAL CANCER: ANALYSIS OF THE MINNESOTA COLON CANCER CONTROL STUDY

Background: Despite reports indicating that polyps located proximal to the splenic flexure have higher recurrence and increased association with post colonoscopy colorectal (CRC) adenocarcinoma, the role of adenoma location on surveillance recommendations remains unclear. The aim of the current study was to analyze outcomes of patients with proximal vs distal colonic adenomas and association with post colonoscopy colorectal adenocarcinoma among participants of the Minnesota Colon Cancer Control Study (MCCCS).
Methods: The MCCS randomized 46,551 patients 50-80 years to usual care, annual or biennial screening with fecal occult–blood testing. Screening was performed from 1976-1982 and 1986-1992. Positive FOBT was followed by colonoscopy. Participants were followed through 20 years for CRC incidence. The analysis was restricted to participants who underwent a colonoscopy and were found to have at least one adenomatous/serrated polyp. Patients were classified as having a proximal polyp if they had at least one lesion proximal to the splenic flexure. We calculated stabilized inverse probability weighting (IPW) propensity scores through a logistic regression model for polyp location with main effects for aspirin usage, smoking history, and body mass index (BMI) at randomization, age at exam, sex, any polyps with diameter ≥1cm, villous polyps, high grade dysplasia, and study treatment arm. We estimated the weighted sub-distribution hazard ratio (SHR) of proximal polyps on CRC-incidence with the competing risk of death.
Results: A total of 2,295 patients met inclusion criteria (Table 1). The majority were males (63%) with a mean age at randomization of 62 years, and at first polyp of 69 years. Amongst them, the mean BMI was 26. Thirty seven percent did not routinely use aspirin and 41% were smokers. There was a high rate of advanced adenomas: 44% had at least one polyp ≥1 cm, 35% had a component of villous histology, and 5% were found with high grade dysplasia. At 20 years, 87 patients had a CRC diagnosis. In adjusted analyses, proximal adenomas had a significantly higher risk of developing a post colonoscopy colorectal adenocarcinoma (SHR =1.63, 1.05-2.53; Figure 1). Additionally, among patients who developed an adenocarcinoma, proximal adenomas were more likely to be associated with proximal CRC (94% of all proximal adenomas with subsequent CRC) than distal polyps (55%).
Conclusions: In this large trial investigating the impact of CRC screening, patients with a proximal adenoma were found to have >60% higher rates of having a metachronous adenocarcinoma. These findings may have important implications in clinical practice, suggesting that polyp location should be considered in the determination of appropriate surveillance time intervals after endoscopic polypectomy.