ORLANDO – Narrow-band imaging colonoscopy accurately predicts the histology of most diminutive polyps in the distal colon, according to findings from a multicenter prospective study.
Use of the technology, which has been promoted as a method for differentiating hypoplastic and adenomatous polyps, could reduce the costs associated with pathological examinations and polypectomy by 60% and 33%, respectively, Dr. Alessandro Repici reported at the annual Digestive Disease Week.
The findings suggest that narrow-band imaging colonoscopy allows high-confidence prediction of the histology of some polyps smaller than 5 mm and is ready for prime time, Dr. Repici said.
"According to our study, high-confidence narrow-band imaging characterization of polyps less than 5 mm was able to meet the criteria threshold for incorporating this kind of real-time histology analysis in clinical practice. This may result in significant reduction of costs in terms of reduction of polypectomy and pathology assessment, and we believe that, according to the present study, high-confidence narrow-band imaging prediction of diminutive polyps located in the distal colon appears ready to be incorporated in clinical practice," he said.
In 278 consecutive patients undergoing elective colonoscopy, a total of 574 polyps smaller than 10 mm, including 429 polyps smaller than 5 mm, were retrieved for histological analysis and were assigned a designation of high or low confidence; 60% of the diminutive polyps were adenomatous. Narrow-band imaging had a sensitivity of 90%, a specificity of 88%, a positive predictive value of 89%, a negative predictive value of 89% and an accuracy of 89% for high-confidence prediction of adenomatous histology in lesions smaller than 5 mm, said Dr. Repici of Istituto Clinico Humanitas, Milan.
Overall, narrow-band imaging had an 84% negative predictive value for diminutive polyps, but the negative predictive value was 92% for those in the distal colon. This met the 90% threshold established by the American Society for Gastrointestinal Endoscopy for adequacy, he said.
Furthermore, high-confidence characterization of polyps smaller than 5 mm predicted the correct surveillance interval in 92% and 99% of cases by American and European guidelines, respectively, he said.
On multivariate analysis, the only independent predictor of an incorrect characterization of polyp histology by narrow-band imaging was a low level of confidence (odds ratio, 3.8), Dr. Repici said.
Study subjects were adults with a mean age of 63 years who underwent elective colonoscopy between May 2011 and May 2012. More than half (58%) were men, and 22% had a family history of colorectal cancer. Participating endoscopists were required to pass a qualifying examination before the start of the study.
Dr. Repici has served on advisory committees or review panels for Boston Scientific and Almirall Prodesfarma, and has served as a consultant for Cosmo Technologies. He has received grant and/or research or other support from Ferring Pharmaceuticals, Olympus Europe, US Endoscopy, and Cook.
