PDF icon Download PDF
Endoscopy 2016; 48(09): 792-793
DOI: 10.1055/s-0042-112576
Editorial
© Georg Thieme Verlag KG Stuttgart · New York

Is optical diagnosis sufficiently established to safely discard diminutive polyps?

Athanasios D. Sioulas
1   Department of Interdisciplinary Endoscopy, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
,
Ioannis S. Papanikolaou
2   Hepatogastroenterology Unit, 2nd Department of Internal Medicine and Research Unit, “Attikon” University General Hospital, University of Athens, Greece
› Author Affiliations
Further Information

Publication History

Publication Date:
29 August 2016 (online)

Also available ateRef
   Permissions and Reprints
Preview

The current standard-of-care during colonoscopy is to resect all polyps that are identified and send them for histopathological examination. This enables subsequent determination of the optimal surveillance interval for each individual; however, significant costs associated with polypectomy, specimen analysis, and additional office visits arise from the aforementioned routine.

It has been shown that approximately 60 % of polyps found in primary screening colonoscopies are of diminutive size (≤ 5 mm) and harbor a remarkably low overall risk of advanced histological features or invasive cancer [1] [2] [3]. An accurate real-time assessment of the histology of these diminutive polyps by means of optical diagnosis may allow for a change in their management, which could substantially reduce the total cost of colonoscopy without compromising its efficacy in colorectal cancer prevention. In this setting, virtual chromoendoscopy technologies such as narrow-band imaging (NBI; Olympus, Tokyo, Japan), i-SCAN (Pentax, Tokyo, Japan), and Fujinon Intelligent Color Enhancement (FICE; Fujinon Inc., Saitama, Japan), as well as confocal laser endomicroscopy, have demonstrated remarkable accuracy in characterizing diminutive polyps [4] [5] [6] [7].

Could new technology alter the “resect – retrieve – send to histology” algorithm in our practice? Taking the growing body of relevant data into account, the American and the European Societies of Gastrointestinal Endoscopy (ASGE and ESGE) have recently endorsed the implementation of real-time optical diagnosis, suggesting a “resect-and-discard” strategy for diminutive polyps that are endoscopically interpreted as conventional adenomas and a “diagnose-and-leave” strategy for predicted non-neoplastic diminutive polyps in the rectosigmoid colon [8] [9]. Accordingly, the ASGE initiative to preserve and incorporate valuable endoscopic innovations (PIVI initiative) has set two quality thresholds before management based on optical diagnosis can be implemented. These include: (i) agreement between optical diagnosis-based and pathology-based surveillance intervals in at least 90 % of patients; and (ii) a negative predictive value (NPV) for the diagnosis of rectosigmoid diminutive adenomas of at least 90 % [10]. Moreover, both societies highlight that this method should be performed preferably by experienced and trained endoscopists in academic centers.

Undoubtedly, polyp optical biopsy and the resultant management strategies are challenging in our everyday endoscopy practice. Discarding the resected polyp or even leaving it in place certainly reduces pathology and, in some cases, polypectomy costs, while also minimizing a patient’s risk of adverse events associated with an eventually unnecessary polypectomy. Moreover, patients can often be provided with an immediate recommendation for surveillance interval, thereby avoiding the burden and cost of a subsequent office visit. These represent great advantages, especially in light of the increasing spending cuts in health care systems worldwide, and serve the need for highest quality care at the lowest possible expense.

On the other hand and beyond the endorsement of the gastroenterology societies, several concerns are raised that bring into question the idea of a universal adoption of such a management for diminutive colon polyps.

1 Appropriate training tools, metrics for performance evaluation, and periodical skill maintenance assessments are definitely needed to minimize operator dependency in optical diagnosis and to ensure a high degree of interobserver and intraobserver agreement. Interestingly, the study by Pohl and colleagues in the present issue of Endoscopy reports no association between previous experience in advanced imaging and an endoscopist’s adenoma detection rate and the quality of optical diagnosis. Furthermore, the authors show that this skill was maintained over time with only a brief intermittent refresher course. Nevertheless, they acknowledge that their results are of limited generalizability, because they reflect an ideal situation [11].

2 Given that most of the evidence is based on results from expert endoscopists in academic centers, optical diagnosis efficacy in the “real world” is at least questionable. There are several studies that have revealed differences in performance between expert and non-expert endoscopists, as well as between training modules and in vivo examinations [12] [13].

3 Medicolegal coverage and support needs to be clarified and become aligned with possible future changes in practice guidelines as a result of the mounting evidence.

4 Patients’ preferences and perceptions should not be neglected. This means that a detailed pre-procedural informed consent should be obtained that clearly states even the negligible risks of such a change in clinical practice. It is of interest that a significant proportion of patients have reported a willingness to pay out of their own pocket the costs for a pathology examination, if their insurance company declined to do so [14]. Ideally, the approval of whoever is paying (i. e. insurance companies) should also be obtained.

5 High quality photo-documentation of target lesions and standardized formal reporting of in vivo histological prediction are considered mandatory.

6 Finally, a criticism of optical diagnosis-based prediction is its inability to discriminate non-advanced and advanced neoplasia and to differentiate hyperplastic and non-hyperplastic serrated lesions. Both limitations may have considerable impact on reaching the quality benchmarks as set by the scientific societies.

In spite of the aforementioned limitations, the study by Pohl et al. published in this month’s issue of Endoscopy lends further support to the accumulating evidence which allows us to presume that the new paradigm could be added to our armamentarium for the management of diminutive colon polyps in the near future. Therefore, we may feel highly confident and be ready to change our plans relying solely on what appears in front of our eyes given the enthusiastic literature data. However, there is no room for over-confidence and, indeed, there are several pragmatic obstacles that need to be overcome first. Further studies are definitely welcome to elucidate the open issues before universal implementation of the new strategy can be recommended.