Search

The purpose of this study is to evaluate an endoscopic trimodal imaging (ETMI) system (high resolution, autofluorescence, and NBI) in the detection and differentiation of colorectal adenomas.A prospective randomised trial of tandem colonoscopies was carried out using the Olympus XCF-FH260AZI system. Each colonic segment was examined twice for lesions, once with HRE and once with AFI, in random order per patient. All detected lesions were assessed with NBI for pit pattern and with AFI for colour. All lesions were removed and sent for histology. Any lesion identified on the second examination was considered as missed by the first examination. Outcome measures are adenoma miss rates of AFI and HRE, and diagnostic accuracy of NBI and AFI for differentiating neoplastic from non-neoplastic lesions.Ninety-four patients underwent colonoscopy with ETMI (47 in each group). Among 47 patients examined with AFI first, 31 adenomas in 15 patients were detected initially [detection rate 0.66 (0.52-0.75)]. Subsequent HRE inspection identified six additional adenomas. Among 47 patients examined with HRE first, 29 adenomas in 14 patients were detected initially [detection rate 0.62 (0.53-0.79)]. Successive AFI yielded seven additional adenomas. Adenoma miss rates of AFI and HRE were 14% and 16.2%, respectively (p = 0.29). Accuracy of AFI alone for differentiation was lower than NBI (63% vs. 80%, p 0.001). Combined use of AFI and NBI achieved improved accuracy for differentiation (84%), showing a trend for superiority compared with NBI alone (p = 0.064).AFI did not significantly reduce the adenoma miss rate compared with HRE. AFI alone had a disappointing accuracy for adenoma differentiation, which could be improved by combination of AFI and NBI.

The aim of this study is to assess the capability of endocytoscopy (ECS) in differentiating neoplastic from nonneoplastic lesions in the colorectum and to validate an ECS classification.Patients with colorectal polypoid and nonpolypoid lesionsor =10 mm were prospectively included. ECS classification subgrouped nonneoplastic (EC 0) and neoplastic (EC 1-3) lesions. Lesions were observed at super-magnified view (450x) before endoscopic resection. Blinded pathological assessment was obtained.Fifty-two lesions were examined in 49 patients (17 polypoid and 35 nonpolypoid). Final pathological diagnosis was normal mucosa or hyperplastic polyp in ten cases, low-grade adenoma in 29, high-grade adenoma in 11, and submucosal invasive cancer in two cases. Positive predictive values of each EC group were 100%, 93.1%, 90.1%, and 100%, respectively. ECS diagnosis correlated completely with pathology in the differentiation between neoplastic and nonneoplastic lesions.ECS enabled observation of colorectal lesion at a subcellular level in vivo. The classification of ECS images had a good correlation with the final pathological diagnosis. ECS was useful to differentiate between neoplastic and nonneoplastic lesions.