Optical coherence tomography to identify intramucosal carcinoma and high-grade dysplasia in Barrett's esophagus

光学相干断层扫描用于识别巴雷特食管的黏膜内癌和高级别异型增生

阅读:2

Abstract

BACKGROUND & AIMS: Optical coherence tomography (OCT) is an optical technique that produces high-resolution images of the esophagus during endoscopy. OCT can distinguish specialized intestinal metaplasia (SIM) from squamous mucosa, but image criteria for differentiating intramucosal carcinoma (IMC) and high-grade dysplasia (HGD) from low-grade dysplasia (LGD), indeterminate-grade dysplasia (IGD), and SIM without dysplasia have not been validated. The purpose of this study was to establish OCT image characteristics of IMC and HGD in Barrett's esophagus. METHODS: Biopsy-correlated OCT images were acquired from patients with Barrett's esophagus undergoing endoscopic surveillance. Two pathologists rendered consensus diagnoses of the biopsy specimens. A blinded investigator reviewed the biopsy-correlated OCT images and scored each for surface maturation and gland architecture. For each image the scores were summed to determine an OCT "dysplasia index." RESULTS: A total of 177 biopsy-correlated images were analyzed. The corresponding histopathology diagnosis was IMC/HGD in 49 cases, LGD in 15, IGD in 8, SIM in 100, and gastric mucosa in 5. A significant relationship was found between a histopathologic diagnosis of IMC/HGD and scores for each image feature (dysplasia index [Spearman correlation coefficient, r = 0.50, P < .0001], surface maturation [r = 0.48, P < .0001], and gland architecture [r = 0.41, P < .0001]). When a dysplasia index threshold of >or=2 was used, the sensitivity and specificity for diagnosing IMC/HGD were 83% and 75%, respectively. CONCLUSIONS: An OCT image scoring system based on histopathologic characteristics has the potential to identify IMC and HGD in Barrett's esophagus.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。