Randomized, controlled trial of standard-definition white-light, high-definition white-light, and narrow-band imaging colonoscopy for the detection of colon polyps and prediction of polyp histology

Abstract

Background: Missing adenomas and the inability to accurately differentiate between polyp histology remain the main limitations of standard-definition white-light (SD-WL) colonoscopy.

Objective: To compare the adenoma detection rates of SD-WL with those of high-definition white-light (HD-WL) and narrow-band imaging (NBI) as well as the accuracy of predicting polyp histology.

Design: Multicenter, prospective, randomized, controlled trial.

Setting: Two academic medical centers in the United States.

Patients: Subjects undergoing screening or surveillance colonoscopy.

Intervention: Subjects were randomized to undergo colonoscopy with one of the following: SD-WL, HD-WL, or NBI.

Main outcome measurements: The proportion of subjects detected with adenomas, adenomas detected per subject, and the accuracy of predicting polyp histology real time.

Results: A total of 630 subjects were included. The proportion of subjects with adenomas was 38.6% with SD-WL compared with 45.7% with HD-WL and 46.2% with NBI (P = .17 and P = .14, respectively). Adenomas detected per subject were 0.69 with SD-WL compared with 1.12 with HD-WL and 1.13 with NBI (P = .016 and P = .014, respectively). HD-WL and NBI detected more subjects with flat and right-sided adenomas compared with SD-WL (all P values <.005). NBI had a superior sensitivity (90%) and accuracy (82%) to predict adenomas compared with SD-WL and HD-WL (all P values <.005).

Limitations: Academic medical centers with experienced endoscopists.

Conclusions: There was no difference in the proportion of subjects with adenomas detected with SD-WL, HD-WL, and NBI. However, HD-WL and NBI detected significantly more adenomas per subject (>60%) compared with SD-WL. NBI had the highest accuracy in predicting adenomas in real time during colonoscopy. (

Clinical trial registration number: NCT 00614770.).

Trial registration: ClinicalTrials.gov NCT00614770.