Indigo Carmine Staining Increases the Yield for Dysplasia in Ulcerative Colitis among Inexperienced Endoscopists: Implementation of a Chromo-endoscopy Program

Abstract

Purpose: Previous studies have shown that chromoendoscopy increases the yield for dysplasia in ulcerative colitis (UC) among experienced endoscopists. The aim of this study was to determine if the yield of dysplasia in patients with longstanding UC could be increased among endoscopists with no experience in this technique through the introduction of a chromoendoscopy program. Methods: Patients with longstanding UC presenting for routine surveillance colonoscopy were enrolled at 3 sites with visualization of the colon by white light endoscopy (WLE) followed by indigo carmine staining (IC). Prior to study initiation, 6 endoscopists who performed the colonoscopies were instructed on chromoendoscopy technique and image interpretation. Lesion size, morphology (pedunculated, flat or sessile), location, histology and modality that detected the abnormality (WLE or IC) were recorded. Standard surveillance biopsies were also obtained. Incremental yield of IC staining for all dysplastic lesions and flat dysplastic lesions on a per-patient and per-lesion basis was calculated. We set a clinically important incremental increase in dysplasia detection with IC of over 20% for statistical comparisons. Results: Seventy-five patients were enrolled across three sites (Florida: 38, Arizona: 24, Minnesota: 13). A total of 157 lesions were identified (WLE: 78, IC: 79), the majority of which were non-dysplastic (85%). Non-dysplastic lesions consisted of 89 pseudopolyps (57%) and 45 hyperplastic polyps (29%). Under WLE, 7 patients (9.3%) had at least 1 dysplastic lesion and with IC, 9 additional patients (12%) had at least 1 dysplastic lesion resulting in 16 patients (21%) having a dysplastic lesion. Among the 75 patients, 10 dysplastic lesions were detected with WLE (LGD, 3; HGD, 2; adenoma, 5) and an additional 12 dysplastic lesions were found with IC (LGD, 3; adenoma, 9) resulting in 22 dysplastic lesions found. Only 1 flat dysplastic lesion was identified with WLE, yet an additional 7 flat dysplastic lesions were found with IC resulting in 8 found overall. For all dysplastic lesions detected through WLE and IC the median size was 4.5 mm. IC staining resulted in a significant increase in detection of all dysplastic lesions and flat dysplastic lesions well in excess of what would be expected from a repeat WLE (Table).Table: Increase in UC dysplasia detection with chromoendscopyConclusion: Implementation of chromoendoscopy among endoscopists previously inexperienced with this technique resulted in higher dysplasia detection compared to white light endoscopy alone in patients with longstanding UC. These rates are similar to the published literature from experienced centers and are encouraging for implementation of chromoendoscopy programs for UC surveillance.