Relationship between Barrett's esophagus and colonic diseases: a role for colonoscopy in Barrett's surveillance

Background Given that risk factors for Barrett's carcinogenesis are predictive, appropriate management and surveillance of Barrett's esophagus (BE) may be provided. The presence of colorectal neoplasms (CRNs) is a possible predictor of the development of BE and the progression to esophageal adenocarcinoma (EAC). We evaluated the relationship between BE or EAC and colonic diseases, including neoplasms and diverticulosis. Methods Patients (N = 5606) who underwent both colonoscopy and esophagogastroduodenoscopy between January 2016 and December 2017 at three institutions were enrolled. The relationships between the presence of colonic diseases and BE or EAC and other clinical or endoscopic predictors of the presence of BE were investigated retrospectively. Results The prevalence of BE >= 1 cm and >= 3 cm in length was 13.0% and 0.52%, respectively. BE was closely related with the presence of colorectal adenoma (48.4% vs. 37.2% in non-BE; P < 0.001), adenocarcinoma (16.6% vs. 8.4%, P < 0.001) and colonic diverticulosis (CD) (34.1% vs. 29.3%, P < 0.001). In patients with long-segment BE, CRNs (79.3%, P < 0.001) and CD (48.2%, P = 0.038) were more common. EAC patients also had a statistically significantly higher incidence of CRNs than non-BE patients (87.5% vs. 45.6%, P = 0.027). Diverticulosis at the distal colon correlated significantly with EAC and BE (50.0%, P = 0.010 and 15.4%, P = 0.024, vs. 12.0% in non-BE). Multivariate analysis showed that CRNs (t = 8.55, P < 0.001), reflux esophagitis (t = 5.26, P < 0.001) and hiatal hernia (t = 11.68, P < 0.001) were predictors of BE. Conclusions The presence of CRNs was strongly associated with BE and EAC. Therefore, colonoscopy may be useful for establishing a strategy for the surveillance of BE.