With the use of high-definition colonoscopies as well as greater attention to quality indicators, such as adenoma detection rates, more diminutive (<6 mm) polyps are being detected and removed.1 Only a small percentage of these polyps have been shown to have advanced pathology.1 However, the higher number of detected polyps has increased the overall cost of the procedure and colorectal cancer (CRC) screening, largely because of an increased number of specimens being sent for pathology. One proposed strategy to decrease cost is the endoscopic differentiation of polyps, or optical diagnosis, which may eliminate the need for pathologic interpretation of diminutive polyps.
There are 2 clinical applications of optical diagnosis that have been proposed: “resect-and-discard” and “diagnose-and-leave.”2 In the “resect-and-discard” strategy, the endoscopist makes an empiric diagnosis of the polyp as adenoma or hyperplastic. The endoscopist then resects the lesion but does not submit the specimen to pathology. In the “diagnose-and-leave” strategy, diminutive (≤5 mm) rectosigmoid polyps that are determined to be hyperplastic are not resected.
There are advantages for these strategies, as well as possible pitfalls. One clear advantage would be potential cost savings. One study has estimated that the strategy could save 33 million dollars per year in the United States.3 Another advantage is that the physician could provide the patient with the surveillance interval recommendations on the day of the colonoscopy. This would improve communication as well as possibly increase compliance with surveillance. One concern has been the acceptance of optical diagnosis by patients and physicians in the United States. However, a survey of over 400 colonoscopy patients found that 2/3 would accept this strategy.4 In another survey, 61% (64/105) of gastroenterologists stated that they would consider adopting this strategy in their practice.5 These limited data suggest that there might be acceptance in the United States for optical diagnostic strategies, but further study is warranted.
The American Society for Gastrointestinal Endoscopy (ASGE) Technology Committee published a systematic review and meta-analysis designed to evaluate endoscopic technologies for the optical diagnosis of diminutive polyps.2 This paper examined whether new evaluation tools, such as Narrow Band Imaging (NBI), can meet the previously established quality metrics previously published in the ASGE Preservation and Incorporation of Valuable endoscopic Innovations (PIVI) thresholds for adoption of new technologies into clinical practice.6 In particular, the ASGE PIVI recommended that when a new technology is used for “resect-and-discard” for polyps ≤5 mm, there has to be a 90% agreement with surveillance guidelines when the results with optical diagnosis are compared with results obtained using the histologic diagnosis. With respect to “diagnose-and-leave” for polyps≤5 mm, the ASGE recommended that the technology afford a 90% negative predictive value (NPV) for adenomatous histology. The ASGE meta-analyses supported the concept that optical diagnosis with NBI can meet these thresholds. Some important factors that were identified as predictors of higher NPV or agreement with surveillance intervals included exams performed at academic centers (versus community centers) as well as those performed by endoscopists with experience (versus novices). An additional important predictor was the confidence level of the endoscopist as dichotomized into high and low confidence. Diagnoses made with higher confidence as compared with lower confidence resulted in a higher agreement with surveillance intervals and with a higher NPV for adenomatous polyps.
The use of high-definition colonoscopies with NBI has allowed endoscopists to differentiate between adenomas and hyperplastic polyps (HPs) based on endoscopic appearance alone. NBI uses a narrowed wavelength light source that optimizes hemoglobin light absorption and highlights the surface as well as the morphologic appearance of adenomas due to the changes in pattern and size of small blood vessels in adenomatous tissue. The criteria used to distinguish between adenomatous and HPs are known as the NBI International Colorectal Endoscopic (NICE) classification.7 One criterion involves examining the color of the polyp. HPs are typically the same color or lighter than the background mucosa, whereas adenomas have a tendency to appear brown with NBI. In another criterion examining the vascular pattern, HPs may have no vessels or may have isolated lacy vessels, whereas adenomas are more likely to have brown vessels surrounding white structures. The third criterion examines the surface pattern and includes characterization of the pits as observed in the Kudo classification.8 HPs tend to have dark or white spots of uniform size similar to the Type II pattern in the Kudo classification, or a homogenous absence of pattern. Adenomas exhibit Type III or IV, which are oval, tubular, or branched white structures surrounded by brown vessels. Using the NICE criteria, studies have achieved results that are acceptable within the PIVI thresholds.7,9 One salient issue is whether sessile serrated adenomas/polyps (SSA/Ps) can be diagnosed endoscopically. The serrated pathway may be associated with a large proportion of all CRCs and may play a role in interval cancers, or those tumors diagnosed between routine colonoscopies. Thus, optical diagnosis of these polyps may have a significant impact for endoscopists’ practices.
The current study by Vleugels et al10 examined optical diagnosis of diminutive polyps using data that were prospectively collected from 3271 colonoscopies performed by 25 gastroenterologists and 11 senior gastroenterology residents between January 2011 and December 2014. Each endoscopist provided an optical diagnosis (SSA/P, adenoma, or HP) for all observed polyps. The investigators did not use a standard protocol for diagnosing polyps, and NBI technology was employed by endoscopists at their discretion. The study outcomes analyzed included accuracy of optical diagnosis, agreement of surveillance intervals, and NPVs for diminutive rectosigmoid neoplastic histology based on the optical diagnosis of diminutive polyps as compared with histopathology.
The authors observed that optical diagnosis of adenomas was made correctly in 80.0% [confidence interval (CI), 78.2-81.8] of cases, but diminutive SSA/Ps were identified correctly only in 24.4% (CI, 18.5-30.3) of cases. In addition, the agreement for surveillance intervals was much lower for patients with diminutive SSA/Ps as compared with the intervals for those without diminutive SSPs (53.3% vs. 78.1%, respectively, P<0.01). The overall NPV for neoplastic histology of rectosigmoid polyps was 79.7% (CI, 76.0-83.4). The presence of diminutive SSA/Ps was associated with a lower NPV as compared with that for patients who did not have SSA/Ps (84.3% vs. 50.0%, respectively, P<0.01).
What do these results tell us? Use of optical diagnosis for diminutive SSA/Ps may yield results for surveillance intervals that have a less than 90% (53.3%) agreement compared with those determined by actual histopathology. Thus, endoscopists, who were aware of NICE criteria, made optical diagnoses that failed to meet PIVI standards for exams with SSA/Ps. This may not be surprising as SSA/Ps are biologically different than HPs and adenomas. One study examined the ability of endoscopists to distinguish SSA/Ps from HPs and adenomas using the NICE criteria.11 When using NICE characteristics for HPs, each individual criterion was observed in 38% to 42% of SSAs, 66% to 67% of HPs, and 15% to 20% of adenomas (P<0.001 for each criterion). Conversely, NICE criteria for adenomas were observed in 57% to 67% of SSAs, 33% to 34% of HPs, and 80% to 84% of adenomas (P<0.001 for each). Thus, the SSA/Ps had features that were neither consistent with HPs nor consistent with adenomas.
To address this issue, some of the investigators of the current article have previously published criteria to aid endoscopists in optically diagnosing SSA/Ps.12 In short, the criteria, known as Workgroup serrated polyps and Polyposis (WASP) classification, can be used to diagnose SSA/Ps. Application involves the use of the NICE criteria to initially classify the lesions as adenomas or HPs. The 4 WASP criteria are then applied to determine whether there are features consistent with that for SSA/Ps. If 2 of the 4 WASP criteria are present then the lesion is considered an SSA/P: (1) a clouded surface; (2) indistinctive borders; (3) irregular shape; or (4) dark spots inside the crypts. Using the WASP criteria, the same investigators observed in another previously published study of 50 polyps that the NPV with high confidence of diminutive neoplastic lesions (adenomas and SSA/Ps together) was 0.91 (95% CI, 0.83-0.96).13 Thus, whereas the NICE criteria can distinguish between adenomas and HPs, the WASP criteria can be used in the optical diagnosis of SSA/Ps.
One interesting finding of the current study with regard to conventional adenomas is that the overall pooled NPV for diminutive polyps in the rectosigmoid for adenomatous polyps was 79.7% (CI, 76.0-83.4), short of the PIVI recommended 90%. In addition, the agreement with surveillance intervals was 75.2% (CI, 72.9-77.6), which falls short of the 90% threshold. Does this mean that optical diagnosis for conventional adenomas is not feasible in a community setting? There are several important issues that need to be recognized when interpreting the results of the current study. The optical diagnoses in the current study were not stratified by confidence levels for the endoscopic assessment of the polyps, high versus low. This factor was identified as a strong predictor in the meta-analysis in the PIVI paper that examined the success of optical diagnosis in meeting thresholds.2 Another important factor in the study was that the endoscopists did not use a standard protocol, nor were they required to use NBI. In addition, the endoscopists did not receive any formal training with respect to optical diagnosis. Therefore, it is possible that the results could have been greatly improved with formal NICE training, use of a standardized protocol for NBI use, and if the endoscopists stratified the confidence of the prediction into high versus low.
Furthermore, precise polyp diagnosis may not be necessary for surveillance recommendations. In a large trial examining the optical diagnosis of diminutive polyps only 83% of polyps were correctly diagnosed, even with high confidence.9 However, the overall agreement with the surveillance intervals was >90%.9 Thus, with modest optical diagnosis accuracy, an adequate agreement with surveillance guidelines was achieved in that study. This was due to the fact that many exams revealed other important synchronous lesions, such as advanced adenomas, which influenced the surveillance recommendations. In addition to these encouraging observations, there are some important factors that the endoscopists who adopt optical diagnosis need to be cognizant of in their practice. Much of the accuracy depends on the confidence level that the endoscopist has for the optical diagnosis. Another issue that needs to be recognized is that endoscopists need to be certain of the size and location of the polyp, rectosigmoid colon versus proximal. This may be challenging as it has been shown that endoscopic assessment of polyp size may not be accurate.14 Thus, there may be significant variability with regard to accuracy of optical diagnostic strategies among endoscopists.
In conclusion, the optical diagnosis of SSA/Ps and polyps in general has advantages for both patients and the health care system. However, factors such as diagnosis confidence, location, sizing, and how to employ NBI need to be resolved before tools such as NICE and WASP can become fully adopted for conventional adenomas and all serrated polyps. Furthermore, NICE criteria use NBI, which is only one type of optical imaging technology. Other image-enhancement technologies that are currently available need to be explored before a final optical diagnosis standard can be adopted.
1. Kahi CJ, Anderson JC, Waxman I, et al. High-definition chromocolonoscopy vs. high-definition white light colonoscopy for average-risk colorectal cancer screening. Am J Gastroenterol. 2010;105:1301–1307.
2. Abu Dayyeh BK, Thosani N, Konda V, et al. ASGE Technology Committee systematic review and meta-analysis assessing the ASGE PIVI thresholds for adopting real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc. 2015;81:502–516.
3. Hassan C, Pickhardt PJ, Rex DK. A resect and discard strategy would improve cost-effectiveness of colorectal cancer screening. Clin Gastroenterol Hepatol. 2010;8:865–869. 9 e1-3.
4. Rex DK, Patel NJ, Vemulapalli KC. A survey of patient acceptance of resect and discard for diminutive polyps. Gastrointest Endosc. 2015;82:376–380. e1.
5. Soudagar AS, Nguyen M, Bhatia A, et al. Are gastroenterologists willing to implement the “Predict, Resect, and Discard” Management Strategy for Diminutive Colorectal Polyps?: results from a National Survey. J Clin Gastroenterol. 2016;50:e45–e49.
6. Rex DK, Kahi C, O’Brien M, et al. The American Society for Gastrointestinal Endoscopy PIVI (Preservation and Incorporation of Valuable Endoscopic Innovations) on real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc. 2011;73:419–422.
7. Hewett DG, Kaltenbach T, Sano Y, et al. Validation of a simple classification system for endoscopic diagnosis of small colorectal polyps using narrow-band imaging. Gastroenterology. 2012;143:599–607. e1.
8. Wada Y, Kudo SE, Kashida H, et al. Diagnosis of colorectal lesions with the magnifying narrow-band imaging system. Gastrointest Endosc. 2009;70:522–531.
9. Pohl H, Bensen SP, Toor A, et al. Quality of optical diagnosis of diminutive polyps and associated factors. Endoscopy. 2016;48:817–822.
10. Vleugels JLA, IJspeert JEG, Hazewinkel Y, et al. Optical diagnosis of sessile serrated polyps; bottleneck for the optical diagnosis paradigm? J Clin Gastroenterol. 2017;51:426–432.
11. Kumar S, Fioritto A, Mitani A, et al. Optical biopsy of sessile serrated adenomas: do these lesions resemble hyperplastic polyps under narrow-band imaging? Gastrointest Endosc. 2013;78:902–909.
12. Hazewinkel Y, Lopez-Ceron M, East JE, et al. Endoscopic features of sessile serrated adenomas: validation by international experts using high-resolution white-light endoscopy and narrow-band imaging. Gastrointest Endosc. 2013;77:916–924.
13. Ijspeert JE, Bastiaansen BA, van Leerdam ME, et al. Development and validation of the WASP classification system for optical diagnosis of adenomas, hyperplastic polyps and sessile serrated adenomas/polyps. Gut. 2016;65:963–970.
14. Anderson BW, Smyrk TC, Anderson KS, et al. Endoscopic overestimation of colorectal polyp size. Gastrointest Endosc. 2016;83:201–208.