Association between baseline hyperplastic polyps and metachronous serrated lesions

Marc Monachese,Gautam Mankaney,Faris El-Khider,Carol Rouphael,Rocio Lopez,Carol A Burke
DOI: https://doi.org/10.1016/j.gie.2020.11.028
Abstract:Background and aims: Some data suggest that individuals with numerous, <10-mm, rectosigmoid hyperplastic polyps (HPs) are at average risk for the development of metachronous advanced adenomatous neoplasia. Guidelines suggest that these individuals do not need surveillance colonoscopy and should be followed akin to individuals with a normal colonoscopy. Less is known of the risk of metachronous neoplasia because of ≥1 HPs <10 mm proximal to the sigmoid colon. We compared the risk of metachronous neoplasia between individuals with small HPs and those with normal colonoscopy, specifically addressing the impact of location and number of HPs on risk. Methods: Colonoscopy and pathology reports from patients with ≥2 colonoscopies between 2004 and 2014 were reviewed. Exclusions included inpatients; age <40 or >75 years; and family or personal history of colorectal cancer, inflammatory bowel disease, previous colorectal surgery, or a previous colonoscopy with any adenoma, sessile serrated lesion (SSL), or HP ≥10 mm. The risk of metachronous neoplasia, including adenomas and SSLs, was compared in individuals with a normal index colonoscopy and those with <10-mm HPs stratified by location and number of HPs. Results: After exclusion, 1795 patients were included. At index colonoscopy, 82% (n = 1469) had a normal examination, 12% (219) had only 1, and 6% (107) had between 2 and 9 HPs <10 mm. Compared with patients with a normal index colonoscopy, patients with a proximal (odds ratio, 3.82; 95% confidence interval, 1.77-7.53) or distal HP (odds ratio, 2.23; 95% confidence interval, 1.18-4.00) had an increased risk of metachronous SSLs but not adenomas. Conclusions: Patients with small proximal and distal HPs are at increased risk of metachronous SSLs. These preliminary findings warrant consideration during surveillance recommendations and future studies in larger cohorts.
What problem does this paper attempt to address?