Presence of Villous Histology in Otherwise Low-Risk Small Adenomas Does not Significantly Increase Risk of Advanced Neoplasia Upon Surveillance: 2017 Presidential Poster Award

Abstract

Introduction: The USMSTF recommends shortened surveillance intervals in the presence of any adenoma >10mm in size, >3 adenomas in number, high-grade dysplasia (HGD) or villous histology. The risk of metachronous advanced neoplasia (MAN) when villous histology or HGD is found in isolated small polyps has not been well described. We aimed to assess rates of MAN based on the presence of villous histology or HGD in small adenomas compared to small tubular adenomas in its absence. Methods: Colonoscopy and pathology reports from all patients found to have 1 or 2 adenomas whose maximum size was <10 mm from 2005 and 2011 at two medical centers were reviewed. These patients were stratified into three groups based on initial colonoscopy findings as follows (1) isolated low-risk adenoma [LRA], (2) small adenoma by size with presence of villous histology [small villous], and (3) small adenoma by size with presence of HGD [small HGD]. Patients with serrated lesions were excluded. Concordance in interpretation of villous histology in small polyps between two GI pathologists was assessed through random review of 50 cases with both pathologists blinded to the other's interpretation. Results: There were 3,185 patients with 1 or 2 small tubular adenomas found on index colonoscopy. The rate of villous histology was 2% (55/2,685) in adenomas <5mm compared to 12% (60/500) in adenomas 6-9mm in size, thus an overall rate of 3.6%. Of these, there were 790 patients with follow-up colonoscopy data available. The rate of MAN in the small villous group (9/79, 11.4%) is similar to the conventional LRA group (56/697, 8%; P=0.287). Conversely, the rate of MAN in the small HGD group (4/14, 28.6%) is significantly greater compared to the LRA group (P=0.024). Polyp size >10 mm was the predominant defining feature of AN at surveillance in all three groups. There was pathology concordance in 37 out of 50 cases with a Kappa score of 0.553 (SE=0.123). Conclusion: The rate of advanced neoplasia on follow-up in patients with small adenomas containing villous histology appears comparable to patients with low-risk adenomas in their absence. Villous histology appears to occur infrequently in diminutive polyps. Given the lack of evidence that isolated villous features portends significant risk and the variation in pathology interpretation, we question the utility of describing villous features in small polyps, especially when diminutive, and making surveillance recommendations based on this determination.Table: No Caption available.