White-Light or Narrow-Band Imaging Colonoscopy in Surveillance of Ulcerative Colitis: A Prospective Multicenter Study.

BACKGROUND & AIMS: Early detection of neoplastic lesions is essential in patients with long-standing ulcerative colitis but the best technique of colonoscopy still is controversial. METHODS: We performed a prospective multicenter study in patients with long-standing ulcerative colitis. Two colonoscopies were performed in each patient within 3 weeks to 3 months. In white-light (WL) colonoscopy, stepwise random biopsy specimens (4 biopsy specimens every 10 cm), segmental random biopsies (2 biopsy specimens in 5 segments), and targeted biopsy specimens were taken. In NBI colonoscopy, segmental and targeted biopsy specimens were taken. The sequence of WL and NBI colonoscopy was randomized. RESULTS: In 36 of 159 patients enrolled (22.6%), 54 lesions with intraepithelial neoplasia (IN) were found (51 low-grade, 3 high-grade). In WL colonoscopy we found 11 IN in stepwise biopsy specimens, 4 in segmental biopsy specimens, and 15 in targeted biopsy specimens. In NBI colonoscopy 7 IN were detected in segmental biopsy specimens and 24 IN were detected in targeted biopsy specimens. Almost all IN were found with one technique alone (κ value of WL vs NBI, -0.86; P < .001). Statistically equivalent numbers of IN were found in NBI colonoscopy with targeted and segmental biopsy specimens as in WL colonoscopy with targeted and stepwise biopsy specimens, but with fewer biopsy specimens (11.9 vs 38.6 biopsy specimens, respectively; P < .001), and less withdrawal time was necessary (23 vs 13 min, respectively; P < .001). CONCLUSIONS: Stepwise biopsy specimens are indispensable in WL colonoscopy. The combination of targeted and segmental biopsy specimens in the NBI technique is as sensitive as targeted together with stepwise biopsy specimens in WL colonoscopy, but requires fewer biopsy specimens and less time. The highest sensitivity should be reached by combining the WL and NBI techniques by switching between the modes.