Abbreviations:
ADR (adenoma detection rate), ASGE (American Society for Gastrointestinal Endoscopy), FICE (Fujinon intelligent chromoendoscopy system), NBI (narrow-band imaging), PVIV (preservation and incorporation of valuable endoscopic innovations), SSP (sessile serrated polyp), TSA (traditional serrated adenoma)In this issue of Gastrointestinal Endoscopy, Wallace et al
1
report a randomized comparison of the Olympus 180 and 190 colonoscopes for predicting the pathologic character of colorectal polyps. The 190 is the new version of the Olympus colonoscope, and its new features include pushbutton-operated optical magnification, which increases magnification from the ×30 provided by the 180 system to about ×65. By comparison, some magnification colonoscopes, including instruments made by Olympus, provide ×100 to ×125 magnification. Therefore, the level of magnification provided by the 190 colonoscope is intermediate between the 180 colonoscope and the highest-magnification colonoscopes available. The 190 magnified image provided no advantage over 180 imaging for accuracy of prediction of polyp pathologic lesions. In addition, there was no benefit from the use of narrow-band imaging (NBI) compared with white light. Accuracy was higher when polyps were interpreted with high confidence.Wallace et al
1
also tested whether the 180 and 190 colonoscopes provided accuracy that met the American Society for Gastrointestinal Endoscopy (ASGE) Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) criteria for real-time interpretation of diminutive colorectal polyp pathologic lesions. Three years ago, the ASGE proposed 2 clinical uses for real-time endoscopic prediction of pathologic polyp lesions (Table 1) and set thresholds for accuracy that technologies should achieve to be used as alternative treatment paradigms to the current management of these clinical problems based on pathologic assessment.2
The endpoint at which polyps have their pathologic character predicted by endoscopic criteria and are then discarded without submission to pathologic assessment is commonly called “resect and discard” (Table 1). Wallace et al1
found that the 180 and 190 imaging colonoscopes each allowed accuracy that exceeded the PIVI thresholds for both clinical endpoints described in Table 1.Table 1The American Society for Gastrointestinal Endoscopy Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) paradigms for alternative management strategy for diminutive polyps detected at colonoscopy
Adapted from 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;73:419-22.
| Paradigm 1. Resect and discard. A technology should allow ≥90% agreement between intervals selected for postpolypectomy surveillance based on (1) high-confidence endoscopic prediction of pathologic lesions for those ≤5 mm followed by resection and no submission to pathologic assessment combined with pathologic assessment for all lesions >5 mm and those <5 mm interpreted endoscopically with low confidence and (2) intervals selected based on pathologic assessment of all polyps. When this agreement threshold is met, the technology is an appropriate tool for a resect-and-discard paradigm as an alternative to submission of all diminutive polyps for pathologic assessment. |
| Paradigm 2. Leave distal colon hyperplastic polyps in place. A technology should allow a negative predictive value ≥90%, when predictions are made with high confidence, for adenomas ≤5 mm in the rectosigmoid colon. When this threshold is met, then the technology can be used to leave distal colon polyps in place without resection or biopsy. |
Comparison of the current results with those of previous studies
Two previous studies, 1 randomized
3
and 1 not,4
have compared the 180 and 190 colonoscopes for prediction of polyp pathologic lesions. Both studies found that the 190 system had comparable accuracy to the 180 system,3
, 4
(ie, the same conclusion reached by Wallace et al). Unlike Wallace et al, both prior studies found that the 190 resulted in more polyps being interpreted with high confidence, which could result in relative cost savings compared with the 180 colonoscope in a “resect and discard” policy. The nonrandomized study found that NBI had superior accuracy to white light for both the 180 and 190 systems,4
which was not addressed in the other randomized trial3
and was unlike the result of Wallace et al. Most previous studies have found that NBI permits better accuracy than white light.5
Like Wallace et al, both prior studies found that both the 180 and 190 colonoscopes allowed both PIVI thresholds to be reached.3
, 4
There is now a large literature reporting the accuracy of various technologies for prediction of a polyp as a pathologic lesion. In a meta-analysis, Wanders et al
6
identified 91 studies addressing the issue, including 56 of NBI, 10 of the Pentax i-SCAN, 14 of the Fujinon Intelligent Chromoendoscopy system (FICE), 11 of autofluorescence, and 11 of confocal laser microscopy. They concluded that all of the technologies were adequate for routine practice, although confocal laser microscopy had slightly higher sensitivity, and autofluorescence had lower specificity. In another meta-analysis, McGill et al7
concluded that NBI provides accuracy that meets both the ASGE PIVI thresholds. Recently, advocates of resect and discard have been disappointed to see studies from community-based endoscopists in which the ASGE PIVI thresholds were not met.8
, 9
However, the study by Wallace et al adds to those from academic practice showing that NBI allows high accuracy in polyp differentiation and for the PIVI thresholds to be met.10
, 11
What types of information can endoscopic imaging and pathologists provide about colorectal polyps?
The overwhelming majority of colorectal polyps fall into 2 histologic categories: the conventional adenomas and the serrated class.
12
Conventional adenomas are uniformly dysplastic (classified as low or high grade) and are also characterized as tubular, tubulovillous, or villous. The serrated class includes the hyperplastic polyps, sessile serrated polyps (SSPs, also known as sessile serrated adenomas), and traditional serrated adenomas (TSAs). Endoscopic imaging is effective at separating lesions into the conventional adenoma versus the serrated class, but it is not capable at this time of subtyping conventional adenomas by dysplasia grade or villosity and similarly is not reliably capable of subdividing serrated lesions into hyperplastic polyps versus SSPs. This is particularly true of the pushbutton-operated, easy-to-use technologies that are standard equipment on commercial colonoscopes (NBI, FICE, and i-SCAN) and that rely on interpretation of surface structures including microvessels.13
, 14
An exception to this generality is the presence of large open pits on serrated lesions (type O pits), which is specific to but not sensitive for SSP.15
Of course, there are other features such as proximal location, large size, and mucus cap that favor SSP over hyperplastic polyp,16
, 17
but these are different considerations from the endoscopic prediction of a pathologic lesion from microscopic surface features. Endoscopic imaging is also good at identifying deep submucosal invasion by cancer in polyps.18
In many cases, cancer can be predicted from gross morphologic features such as ulceration, but deep submucosal invasion is also typically accompanied by disruption of the surface vascular pattern, so that the surface vessels and structures have an irregular or amorphous pattern. Thus, endoscopic imaging has both strengths and limitations.What is often not appreciated is that endoscopists and pathologists have similar limitations. Pathologists are quite reliable for separating conventional adenomas from the serrated class and for identifying cancer in polyps.
19
By contrast, the level of interobserver variation among pathologists in interpreting dysplasia grade and villous elements is so great that the British Society of Gastroenterology does not consider these factors in its postpolypectomy surveillance recommendations.20
The problem is accentuated to a remarkable degree in small adenomas, where high-grade dysplasia and villous elements are uncommon.21
Similarly, differentiation of SSP from hyperplastic polyp is unreliable even among expert pathologists,22
and in clinical practice there are pathologists who never read SSPs.23
TSA, the only uniformly dysplastic lesion in the serrated class, is so consistently misread in clinical practice as tubulovillous adenoma that anecdotally I encounter many endoscopists who have never seen TSA on a pathology report.The approach to these issues by the ASGE was to limit the target polyps for resect and discard and for leaving distal hyperplastic polyps in place to those 5 mm or smaller.
2
Polyps in that size range have a cancer risk near zero and have a low prevalence of the histologic features that both endoscopists and pathologists have difficulty recognizing: villous elements, high-grade dysplasia, and SSP components.What are the potential benefits of the PIVI polyp paradigms?
The main gains from the PIVI paradigms for diminutive polyp management are cost savings. The potential savings from resect and discard were estimated at more than $1 billion per year in the United States.
24
As high-definition colonoscopes25
and split-dose bowel preparation26
become commonplace the savings would increase. Leaving distal diminutive hyperplastic polyps in place saves the costs of both pathologic assessment and polypectomy, and it reduces patient risk, at least when endoscopists use electrocautery to resect diminutive polyps. Some have argued that resect and discard would allow colonoscopists to tell patients immediately when the next colonoscopy would occur, which might improve adherence to the follow-up examination, but this is speculative at this time.What are the risks of the PIVI polyp paradigms?
One risk is that a cancer in polyp 5 mm or smaller would be discarded and not recognized. If this cancer had deep submucosal invasion it might recur in the colon wall or a lymph node, with deadly consequences. Recent evidence suggests that the chance of cancer in a polyp 5 mm or smaller is near nil,
27
, 28
and the rare occurrence of the scenario just described does not alter the cost effectiveness of resect and discard.24
Malignant polyps are generally recognizable endoscopically by features such as ulceration, depression, and altered surface patterns, with the possible exception of cancer in some larger pedunculated polyps and nongranular laterally spreading tumors. In any case, it should be generally understood that the recognition of any atypical features in a diminutive polyp should be followed by submission to pathologic assessment.2
The other risk of resect and discard is assigning the “wrong” postpolypectomy interval. The goal of screening and surveillance colonoscopy is to prevent colorectal cancer. Factors that determine whether this outcome eventuates include the adenoma detection rate (ADR) of the endoscopist,
29
, 30
the effectiveness of the polypectomy technique,31
and the selected interval for the next colonoscopy. Of these, the clearest evidence of an effect on cancer prevention is ADR.29
, 30
The evidence for completeness of polypectomy is indirect,32
although the logic of the concept that at least larger polyps warrant complete resection seems sound. By contrast, whether the follow-up interval is “wrong” seems relatively less important. Surveillance recommendations are made by expert panels evaluating mostly observational studies (only the National Polyp Study was randomized), and panels from different countries examining the same evidence produce disparate recommendations.20
, 33
Recommendations on follow-up of serrated lesions are based on knowledge about molecular features and associations of the lesions with cancer, inasmuch as not even observational postpolypectomy studies are available.12
, 33
The fundamental flaw in postpolypectomy surveillance recommendations is that they do not account for the detection skills of the endoscopist. Endoscopists with higher ADRs bring patients back at shorter intervals, providing a sort of double protection from cancer. Endoscopists with lower ADRs bring patients back at longer intervals, multiplying the risk of cancer. Considered in this light, the quality of the baseline examination appears more important than the surveillance interval. Furthermore, recommended intervals could change as evidence emerges that the number of small adenomas does not affect the incidence of advanced lesions at follow-up,34
or at least the risk with 3 or 4 small adenomas is not different from the risk with 1 or 2.35
Another context that makes concerns over resect and discard appear overdone is CT colonography, which has poor sensitivity and specificity for lesions 5 mm or smaller. Radiologists are advised to not even report lesions 5 mm or smaller,
35
and this advice has been accepted throughout the medical community. Without this practice, CT colonography would be untenable for screening. An underpinning of the policy is the evidence that polyps 5 mm and smaller so rarely contain cancer and so rarely become cancer that they can be entirely ignored for at least 5 years. Bizarrely, we then debate the appropriateness of saving money by side-stepping pathologic assessment of these same innocuous lesions, after their pathologic character has already been estimated endoscopically with high confidence and high accuracy. This accuracy is comparable to that of pathologic assessment and far above the capacity of CT colonography to estimate a pathologic lesion, which is zero. In addition, in resect and discard, the lesions have been removed, thereby all but eliminating any chance that the lesion could harm the patient.Is the use of pathologic assessment inviolate as a diagnostic test for resected tissue?
The decision to perform a diagnostic test is based on several considerations. One is the characteristics of the test itself, including its accuracy, cost, and risk. If these factors are unfavorable and the clinical suspicion of the diagnosis is low or the diagnosis is either not life-threatening or clinically less important, it’s probably best to not perform the test. If the diagnosis is life-threatening, proceeding with the test may be best even when the clinical suspicion is low and there is some risk. Thus, the consequences of mesenteric ischemia are so devastating that any serious consideration of the diagnosis warrants testing. Another consideration is the pretest probability of the diagnosis. Diagnostic testing has its greatest value when the pretest probability is neither very high nor low (ie, there is real uncertainty about the diagnosis).
As a diagnostic test, pathologic assessment of diminutive polyps is left out of the considerations just described for diagnostic testing. Rather, we are told (by whom, one wonders?) that we must submit every piece of resected tissue to pathologic assessment. Whether diminutive polyps are conventional adenomas or hyperplastic lesions does not meet the criterion of a serious or life-threatening issue. In the resect and discard paradigm, polyps are discarded only when the endoscopic prediction of a pathologic lesion is made with high confidence.
2
Applying costly diagnostic tests to clinical decisions of limited importance when the pretest probability of the diagnosis is already high is usually considered a waste of resources (Fig. 1). There is no a priori reason why the same considerations that apply to other diagnostic tests should not apply to pathologic assessment of resected diminutive polyps.
Figure 1Usefulness of diagnostic tests in relation to reliability of clinical findings.
It’s interesting that we could estimate the pathologic character of diminutive polyps by endoscopic criteria and then ablate them (eg, with argon plasma coagulation) and there would be no controversy because we would eliminate the “discard” part of “resect and discard.” If there were adequate data on the rates of eradication and complication of ablative therapy for diminutive polyps, this might become a legitimate route to reduction of pathologic assessment costs. As it stands, cold snare polypectomy is both effective and safe,
36
so we pursue the course best for our patients, which is resect and discard.Where from here?
Table 2 lists various steps and issues that must yet be addressed or overcome to make the ASGE PIVI polyp paradigms a reality. Although the ASGE has taken the lead in this area, progress is made fastest when our societies advocate in concert.
Table 2Remaining steps to make the Preservation and Incorporation of Valuable Endoscopic Innovations paradigms a clinical reality
|
Endoscope manufacturers should develop image storage systems that can memorialize the basis of an endoscopist’s decision and thereby provide medical and legal protection, and also document (by photography rather than histology) the ADRs that are now an essential part of colonoscopy.
The observation that academic physicians have thus far outperformed community physicians in real-time prediction of pathologic lesions indicates the need for better training tools and suggests that the 2 groups of physicians may be motivated differently. Academic physicians may be interested in succeeding in real-time prediction of pathologic lesions for the publications that follow training and investigation. These factors are of little interest to community endoscopists, who may need to see real-time prediction of pathologic lesions as a legitimate clinical practice, with an appropriate financial reward (see below), before they’ll invest the effort to master it.
A real and very substantial obstacle to resect and discard is that it asks endoscopists to train, expend effort and resources, and assume risk without compensation. One solution would be Current Procedural Terminology codes for real-time determinations, which are now in place for confocal laser microscopy and could be pursued for the other more practical modalities. Compensation is now paid to colonoscopists in Japan when they use NBI and other routine image-enhancing devices during endoscopy. Obtaining a fee-for-service payment for real-time endoscopic estimates of pathologic lesions in the United States would be a challenge, to say the least. Recently there has been talk of payment reforms for colonoscopy, such as reference payment and bundled payment. The American Gastroenterological Association proposed a bundled payment model, which included the costs of sedation and pathologic assessment.
37
The pathologic assessment part of the bundle is tricky because it requires detailed information regarding average pathologic assessment costs per colonoscopy in a given practice, and the bundle itself could disincentivize polyp detection, which runs counter to the current understanding of what prevents cancer. On the other hand, bundled payment would provide an important financial incentive for resect and discard.Is evidence enough to move resect and discard forward? In some academic practices the answer will be yes. New paradigms in colonoscopy, however, have generally moved forward when enough vested interests benefit. Screening colonoscopy itself moved forward (with almost no supporting evidence—that largely came later) in great part because it was attractive to all relevant physician groups and to patients. Endoscopist-directed propofol had plenty of supportive evidence but wasn’t aligned with the financial interests of any physician group. It thrives only in pockets in the United States and in countries like Switzerland and Germany, where a different set of financial incentives apply. The fate of resect and discard may be similarly linked to economic incentives.
Resect and discard is emerging as an evidence-based practice that could significantly reduce health care costs in the United States. Whether resect and discard will become a reality here will depend in part on whether and how effectively our societies endorse it and advocate for it. Furthermore, its future may depend on whether we and our societies find a reimbursement model in which colonoscopists have an appropriate and reasonable financial incentive to undertake the training for, and assume the risk of, performing resect and discard.
Disclosure
The author receives research support and consulting fees from Olympus America Corp.
References
- Accuracy of in vivo colorectal polyp discrimination by using dual-focus high-definition narrow-band imaging colonoscopy.Gastrointest Endosc. 2014; 80: 1072-1087
- 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
- The valid colonoscopy study: results of a multicenter prospective randomized controlled trial on real time colorectal polyp diagnosis using narrow band imaging (NBI).Gastrointest Endosc. 2012; 75 ([abstract]): AB151
- New colonoscope technology: impact on image capture and quality and on confidence and accuracy of endoscopy-based polyp discrimination.Endoscopy. 2014; 46: 172-178
- Systematic review of narrow-band imaging for the detection and differentiation of neoplastic and nonneoplastic lesions in the colon (with videos).Gastrointest Endosc. 2009; 69: 124-135
- Diagnostic performance of narrowed spectrum endoscopy, autofluorescence imaging, and confocal laser endomicroscopy for optical diagnosis of colonic polyps: a meta-analysis.Lancet Oncol. 2013; 14: 1337-1347
- Narrow band imaging to differentiate neoplastic and non-neoplastic colorectal polyps in real time: a meta-analysis of diagnostic operating characteristics.Gut. 2013; 62: 1704-1713
- Real-time optical biopsy of colon polyps with narrow band imaging in community practice does not yet meet key thresholds for clinical decisions.Gastroenterology. 2013; 144: 81-91
- Endoscopic versus histological characterisation of polyps during screening colonoscopy.Gut. 2014; 63: 458-465
- Narrow-band imaging without optical magnification for histologic analysis of colorectal polyps.Gastroenterology. 2009; 136: 1174-1181
- Optical diagnosis of small colorectal polyps at routine colonoscopy (Detect InSpect ChAracterise Resect and Discard; DISCARD trial): a prospective cohort study.Lancet Oncol. 2009; 10: 1171-1178
- Serrated lesions of the colorectum: review and recommendations from an expert panel.Am J Gastroenterol. 2012; 107: 1315-1329
- Meshed capillary vessels by use of narrow-band imaging for differential diagnosis of small colorectal polyps.Gastrointest Endosc. 2009; 69: 278-283
- Validation of a simple classification system for endoscopic diagnosis of small colorectal polyps using narrow-band imaging.Gastroenterology. 2012; 143: 599-607
- A novel pit pattern identifies the precursor of colorectal cancer derived from sessile serrated adenoma.Am J Gastroenterol. 2012; 107: 460-469
- 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
- A morphologic analysis of sessile serrated polyps observed during routine colonoscopy (with video).Gastrointest Endosc. 2011; 74: 1360-1368
- Endoscopic prediction of deep submucosal invasive carcinoma: validation of the narrow-band imaging international colorectal endoscopic (NICE) classification.Gastrointest Endosc. 2013; 78: 625-632
- Variable pathologic interpretation of columnar lined esophagus by general pathologists in community practice.Gastrointest Endosc. 1999; 50: 23-26
- Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002).Gut. 2010; 59: 666-689
- Agreement in interpreting villous elements and dysplasia in adenomas less than one centimetre in size.Dig Liver Dis. 2013; 45: 1049-1055
- Reinterpretation of histology of proximal colon polyps called hyperplastic in 2001.World J Gastroenterol. 2009; 15: 3767-3770
- Endoscopic detection of proximal serrated lesions and pathologic identification of sessile serrated adenomas/polyps vary on the basis of center.Clin Gastroenterol Hepatol. 2014; 12: 1119-1126
- A quantitative assessment of the risks and cost savings of forgoing histologic examination of diminutive polyps.Endoscopy. 2011; 43: 683-691
- High definition colonoscopy detects colorectal polyps at a higher rate than standard white light colonoscopy.Clin Gastroenterol Hepatol. 2010; 8: 364-370
- Increased adenoma detection rate with system-wide implementation of a split-dose preparation for colonoscopy.Gastrointest Endosc. 2012; 76: 603-608
- Systematic review: distribution of advanced neoplasia according to polyp size at screening colonoscopy.Aliment Pharmacol Ther. 2010; 31: 210-217
- Prevalence of advanced histological features in diminutive and small colon polyps.Gastrointest Endosc. 2012; 75: 1022-1030
- Quality indicators for colonoscopy and the risk of interval cancer.N Engl J Med. 2010; 362: 1795-1803
- Adenoma detection rate and risk of colorectal cancer and death.N Engl J Med. 2014; 370: 1298-1306
- Incomplete polyp resection during colonoscopy: results of the Complete Adenoma Resection (CARE) study.Gastroenterology. 2013; 144: 74-80
- Analysis of colorectal cancer occurrence during surveillance colonoscopy in the Dietary Polyp Prevention Trial.Gastrointest Endosc. 2005; 61: 385-391
- Guidelines for colonoscopy surveillance after screening and polypectomy: a consensus update by the US Multi-Society Task Force on Colorectal Cancer.Gastroenterology. 2012; 143: 844-857
- Postpolypectomy colonoscopy surveillance guidelines: predictive accuracy for advanced adenoma at 4 years.Ann Intern Med. 2008; 148: 419-426
- Risk of advanced lesions at first follow-up colonoscopy in high-risk groups as defined by the United Kingdom post-polypectomy surveillance guideline: data from a single U.S. center.Gastrointest Endosc. 2014; 80: 299-306
- Removal of small colorectal polyps in anticoagulated patients: a prospective randomized comparison of cold snare and conventional polypectomy.Gastrointest Endosc. 2014; 79: 417-423
- A bundled payment framework for colonoscopy performed for colorectal cancer screening or surveillance.Gastroenterology. 2014; 146: 849-853
Article Info
Identification
Copyright
© 2014 American Society for Gastrointestinal Endoscopy. Published by Elsevier Inc. All rights reserved.
ScienceDirect
Access this article on ScienceDirectLinked Article
- Accuracy of in vivo colorectal polyp discrimination by using dual-focus high-definition narrow-band imaging colonoscopyGastrointestinal EndoscopyVol. 80Issue 6
02198-1&id=gr1.jpg){kind=link}