A 1-Year Cross-sectional Inflammatory Bowel Disease Surveillance Colonoscopy Cohort Comparing High-definition White Light Endoscopy and Chromoendoscopy

Abstract

We sought to compare the dysplasia detection rate of high-definition white light endoscopy (HDWLE) with that of chromoendoscopy in patients with long-standing inflammatory bowel disease (IBD). This is a retrospective observational cohort of patients with IBD who underwent surveillance colonoscopy between October 1, 2016 and September 30, 2017. We assessed the association between dysplasia detection and multiple variables. A total of 808 unique colonoscopies were performed, of which 150 (18.6%) included chromoendoscopy. Primary sclerosing cholangitis was a comorbid diagnosis in 24.5% of patients. The performing endoscopist was an IBD specialist with 37.1% of patients and had >10 years’ experience with 64.9% of patients. Prior dysplasia had been seen in 245 (30.3%) patients: 102 (68.0%) and 143 (22.0%) among patients who had chromoendoscopy and HDWLE, respectively. Dysplasia in polyps was found in 129 procedures (15.1%). Among patients who had chromoendoscopy and HDWLE, polypoid dysplasia was identified in 50 (33.0%) and 79 (12.0%) patients, respectively, P < 0.01. Dysplasia in random biopsies was found in 39 patients (4.8%): 15 (10%) who had chromoendoscopy and 24 (3.6%) who had HDWLE (P < 0.001). On multivariate analysis, patient and disease characteristics significantly associated with an increased odds for polypoid dysplasia included older age at diagnosis (odds ratio [OR] = 1.3 per 10 years; 95% confidence interval [CI], 1.07-1.60), having an IBD physician endoscopist (OR = 1.6; 95% CI, 1.01-2.67), having an endoscopist with less than 10 years’ experience (OR = 1.8; 95% CI (1.16-2.89), and prior random dysplasia (OR = 4.2; 95% CI (1.93-9.17). Concomitant primary sclerosing cholangitis was significantly associated with random dysplasia (OR = 2.3; 95% CI, 1.02-5.07). After multivariate analysis adjusting for these variables, chromoendoscopy was no more likely to identify dysplasia than was HDWLE. Chromoendoscopy and HDWLE had a similar diagnostic yield for dysplasia detection in patients with chronic IBD-colitis after adjusting for multiple known risk factors.