Dysplasia Detected in Patients With Serrated Epithelial Change Is Frequently Associated With an Invisible or Flat Endoscopic Appearance, Nonconventional Dysplastic Features, and Advanced Neoplasia
Dorukhan Bahceci,Lindsay Alpert,Tanner Storozuk,Xiaoyan Liao,Masato Yozu,Maria Westerhoff,Bence P Kővári,Gregory Y Lauwers,Won-Tak Choi
DOI: https://doi.org/10.1097/PAS.0000000000002271
2024-10-01
Abstract:The significance of serrated epithelial change (SEC), defined as endoscopically invisible hyperplastic polyp (HP)-like mucosal change identified in patients with inflammatory bowel disease (IBD), remains unclear. Although some studies reported an increased risk of synchronous and/or metachronous colorectal neoplasia in patients with SEC, including advanced neoplasia (high-grade dysplasia or colorectal cancer), the development of SEC is not significantly associated with increased colonic inflammation. This contrasts with the reported positive correlation between increased colonic inflammation and the risk of colorectal neoplasia in ulcerative colitis, arguing against the notion that SEC may represent a form of dysplasia. As such, this study aimed to characterize the features of synchronous and metachronous dysplasia detected in patients with SEC to identify factors contributing to the increased risk of colorectal neoplasia, including advanced neoplasia, observed in a subset of these patients. Clinicopathologic features of 46 IBD patients with SEC (n=109) and synchronous and/or metachronous dysplasia (n=153) were analyzed. All dysplastic lesions were subtyped as either conventional or nonconventional dysplasia. As controls, 45 IBD patients with endoscopically visible or polypoid HP (n=75) and synchronous and/or metachronous dysplasia (n=87) were analyzed. The SEC group included 28 (61%) men and 18 (39%) women with a mean age of 58 years and a long history of IBD (mean duration: 23 years). The majority of patients (n=34; 74%) had ulcerative colitis, and 12 (26%) had Crohn's disease. Thirty-nine (85%) patients had a history of pancolitis, and 2 (4%) had concomitant primary sclerosing cholangitis. Twenty-seven (59%) patients had multifocal SEC. SEC was predominantly found in the left colon (n=52; 48%) and rectum (n=34; 31%). Dysplasia in the SEC group was often endoscopically invisible or flat (n=42; 27%) and demonstrated nonconventional dysplastic features (n=49; 32%). Six nonconventional subtypes were identified in the SEC group, including 17 (11%) dysplasia with increased Paneth cell differentiation, 12 (8%) hypermucinous dysplasia, 8 (5%) crypt cell dysplasia, 7 (5%) goblet cell deficient dysplasia, 3 (2%) sessile serrated lesion-like dysplasia, and 2 (1%) traditional serrated adenoma-like dysplasia. Advanced neoplasia was detected in 11 (24%) patients. The SEC group was more likely to have nonconventional dysplasia (32%, P <0.001), invisible/flat dysplasia (27%, P <0.001), and advanced neoplasia (24%, P <0.001) than the control group (7%, 2%, and 0%, respectively). High-risk nonconventional subtypes (ie, hypermucinous, crypt cell, and goblet cell deficient dysplasias) accounted for 18% of all dysplastic lesions in the SEC group, which were not seen in the control group ( P <0.001). The SEC group (n=35; 76%) also had a higher rate of concordance between the location of SEC and the area of synchronous/metachronous dysplasia than the control group (n=22; 49%) ( P =0.007). In conclusion, dysplasia detected in patients with SEC is often endoscopically invisible/flat (27%), nonconventional (32%, including the high-risk subtypes), and found in the same colonic segment as SEC (76%), which may in part explain why some patients with SEC are associated with an increased risk of colorectal neoplasia, including advanced neoplasia. The finding of SEC may warrant a careful follow-up colonoscopy with increased random biopsy sampling, especially in the segment of colon with SEC.