Optical diagnosis of diminutive colorectal polyps
S. Sakata,S. Naidu,A. Stevenson,D. Hewett
DOI: https://doi.org/10.1111/ans.13531
IF: 1.7
2017-05-01
ANZ Journal of Surgery
Abstract:Colonoscopy has had a positive impact on colorectal cancer incidence and mortality, but with technological advances, the financial burden on health systems is increasing because more polyps are now detected, resected and referred for pathological assessment. ‘Optical diagnosis’ is a new management paradigm for diminutive colorectal polyps (≤5 mm) with two clinical endpoints: ‘resect and discard’, the endpoint at which diminutive adenomas are resected and discarded without pathological confirmation and ‘diagnose and leave’, the endpoint at which hyperplastic polyps in the rectosigmoid are left without resection. The overall risk of cancer in diminutive polyps is negligible, and only 2% of diminutive adenomas have advanced histologic features of high-grade dysplasia or villous histology. Histopathology of diminutive polyps is thus primarily used to assign patients into highand low-risk categories for post-polypectomy surveillance. Optical diagnosis allows for real-time assessment of all detected diminutive polyps, determining surveillance intervals on the procedure day, as well as reduction in polypectomy-related complications. Perhaps, the most convincing argument for optical diagnosis is the potential to contribute towards a substantial global reduction in healthcare costs. As diminutive colorectal polyps comprise more than 80% of the detected polyps, optical assessment has been estimated to result in the upfront savings of more than one billion dollars annually in the United States alone. Also, in a recent Australian study, endoscopists who were experts in narrow band imaging (NBI) demonstrated a potential cost savings of $319 per patient if optical diagnosis was adopted in their tertiary centre. In 2011, the American Society for Gastrointestinal Endoscopy (ASGE) established two Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) threshold criteria for optical diagnosis to be supported as an alternate management paradigm for standard practice. Endoscopists have to achieve a minimum of (i) 90% negative predictive value for adenomas and (ii) 90% accuracy in predicting appropriate surveillance intervals. In late 2014, a direct meta-analysis was conducted by the ASGE to determine if PIVI threshold criteria were achieved with the current endoscopic technologies. This study showed that NBI (Olympus, Tokyo, Japan) was the only technology that provided experts with high-confidence optical assessments that fulfilled both PIVI criteria. Recent guidelines from the European Society of Gastrointestinal Endoscopy (ESGE) have also supported the use of optical diagnosis in appropriately credentialed endoscopists. By using a narrowed wavelength light source to optimize haemoglobin light absorption, NBI highlights the characteristic pattern and size of microvessels in the mucosa and submucosa of neoplastic lesions, and makes these lesions appear darker than the surrounding normal mucosa. As no study is yet to demonstrate a considerable increase in adenoma detection with NBI versus conventional white light colonoscopy, the true value of NBI lies in its application within the paradigm of optical diagnosis. It must be emphasized that NBI cannot evaluate individual cellular architecture to detect cytologic dysplasia, and this technology assists in the optical diagnosis of lesions based on their macroscopic features that are associated with histology. Also, NBI is unable to differentiate between polyps of the serrated class (although this is no different from the wide inter-observer variation in the histopathologic diagnosis of sessile serrated polyps and hyperplastic polyps). Diminutive hyperplastic polyps of the serrated class can be targeted by optical diagnosis because they populate the rectosigmoid, and are not considered premalignant. Validated tools such as the NBI International Colorectal Endoscopic (NICE) classification provide endoscopists with a simple and efficient strategy to make high-confidence predictions of hyperplastic polyps (type 1 lesions), adenomas (type 2 lesions) and deep submucosal invasive cancer (type 3 lesions). An excellent resource of photographs detailing the key features of diminutive neoplastic and non-neoplastic polyps, classified by NICE and other imaging criteria is now available. We recommend that Australasian endoscopists consider the value of NBI by making real-time assessments of diminutive colorectal polyps with histopathological correlation. Although the ASGE and ESGE support the potential use of optical diagnosis for experts in NBI, this paradigm is not yet endorsed in Australasia, so we therefore appreciate why local endoscopists may choose to not use image-enhanced modalities in their routine practice. Nevertheless, we recognize the potential for optical diagnosis technologies to alleviate the rising costs of healthcare and we recommend the investigation and clarification of critical system factors before optical diagnosis of diminutive lesions is adopted in Australasia (Table S1). We encourage Australasian endoscopists, pathologists and their patients to be part of this worldwide discussion.