Optical diagnosis for diminutive colorectal polyps: time to implement?

In the current issue of Endoscopy, Hassan et al. present a revised optical diagnosis classification (BASIC) in conjunction with blue-light imaging (BLI) for optical diagnosis of colorectal polyps [1]. Using this methodology, the accuracy of optical diagnosis in distinguishing between adenomatous and hyperplastic colorectal polyps was 90.3 % and superior to that when using white-light imaging.

The study demonstrates that, when using modern endoscopic imaging technology, expert endoscopists are well able to differentiate between adenomatous and non-adenomatous polyps. The study also points out the importance of different imaging descriptors and how to combine these in order to improve the accuracy of optical diagnosis. The study is limited by the fact that a video library was used for optical diagnosis and clinical studies using BLI with BASIC will have to confirm the results. Ideally such studies should include endoscopists who are not dedicated experts in optical polyp diagnosis in order to better reflect the outcomes that can be expected in routine clinical practice. Previous studies have shown that the transfer of optical diagnosis into general practice can indeed decrease accuracy and lower the surveillance interval agreement compared with pathology to below the quality benchmark proposed by the ESGE and ASGE. In the past, this discrepancy between experts and non-experts using optical diagnosis has stood in the way of the widespread clinical implementation of optical polyp diagnosis.

“The study demonstrates that, when using modern endoscopic imaging technology, expert endoscopists are well able to differentiate between adenomatous and non-adenomatous polyps.”

Being able to implement optical polyp diagnosis into routine clinical practice remains an important endeavor as it would greatly reduce colonoscopy-associated costs. Adequately trained endoscopists using modern endoscopic video-optics and classification systems can reliably distinguish adenomas from non-adenomas as the study shows. The difference in diagnostic accuracy between expert endoscopists and routine practice is often mentioned as a roadblock standing in the way of widespread clinical implementation. However, there are actually arguments that it is time to implement optical diagnosis for routine colonoscopy practice and that this would be feasible, beneficial, and safe:

Firstly, the proposed 90 % quality benchmark for surveillance interval agreement when using optical diagnosis instead of pathology is based on the assumption that, in clinical reality, after a polyp has been sent for pathology assessment, the assignment of the post-polypectomy surveillance interval is done correctly. A recent meta-analysis has shown that the surveillance interval in North American studies after the detection of low risk adenomas is in accordance with the guidelines in only 45 % of cases and even lower (37 %) after the detection of hyperplastic polyps [2]. In Europe, the surveillance interval assignments after detection of low risk lesions conforms to the guideline recommendations in only 24 % of cases [2]. Incomplete or incorrect pathology follow-up likely contributes in large part to the assigned post-polypectomy surveillance intervals that do not conform to guidelines. Optical diagnosis provides endoscopists with an immediate diagnosis at the time of colonoscopy. This might be a distinct advantage over pathology that has not previously been considered when discussing the clinical implementation of optical diagnosis. The use of optical diagnosis in routine clinical practice might actually outperform guideline conformity for the assignment of surveillance interval compared with the real-life pathology-based approach, even if the optical polyp diagnosis is performed with less accuracy than that which can be achieved by experts.

Secondly, the negative effect of an incorrect optical polyp diagnosis to determine the surveillance interval might actually be very low. A recent study has shown that individual incorrect optical polyp diagnosis did not have any negative impact on determining the final surveillance interval in 74 % of patients [3]. For the majority of patients, the next surveillance interval is based either on a low risk situation or will be determined by synchronous larger polyps, a history of adenomas, or family history of colorectal cancer. Because most surveillance guidelines are moving towards 10-year intervals for patients in whom no polyps, hyperplastic polyps, and/or low risk adenomas are found, sending diminutive polyps to pathology is becoming almost obsolete as the risk for findings that would result in a surveillance interval other than 10 years (e. g. high grade dysplasia and/or serrated pathology) remains rare for diminutive polyps. The risk for an incorrect surveillance interval assignment when using optical diagnosis for diminutive polyps is consequently rare. We should, however, strengthen the ability of endoscopists to detect the predictors of high grade dysplasia and serrated pathology when they are using optical diagnosis.

Thirdly, from a global cost perspective, implementing resect and discard is becoming increasingly important. As Hassan et al. point out in their introduction, the improvement in endoscopic imaging quality and adjunct techniques has led to a steep increase in the detection rates of colorectal polyps [1]. Studies using the latest technology have shown adenoma detection rates (ADRs) above 50 % and polyp detection rates (PDRs) above 80 %, while mainly diminutive low risk adenomas and/or hyperplastic polyps are found. Sending all of these polyps for histopathology examination is putting an increasing cost burden on the medical system and the previous estimates of potential cost savings through optical diagnosis are likely to be underestimated when considering the increase in ADR and PDR. Non-neoplastic polyps make up about 50 % of small polyps, advanced pathology is rare, and cancer remains an exceedingly rare finding [4] [5] [6]. Furthermore, the majority of small colorectal adenomas have slow growth patterns and a low risk for malignant transformation, and the number of non-progressive adenomas is significant [7] [8] [9].

Taken together these factors speak for the safety of implementing optical diagnosis into clinical practice. The low risk of doing harm by missing advanced pathology when using optical diagnosis stands in stark contrast to its huge cost-saving potential. A vital aspect of providing good healthcare is not using inappropriate or overly expensive modalities. Replacing pathology examination for diminutive colorectal polyps with optical diagnosis makes sense in many ways and the presented study provides another important piece of information to support this. Along with additional studies that refine imaging modalities and classification systems, the discussion about clinical implementation will hopefully gain momentum as the benefits of optical diagnosis for diminutive polyps likely outweigh the risks, even if the accuracy would presumably be a bit lower than what we see in expert studies.

• References

• 1 Hassan C, Bisschops R, Bhandari P. et al. Predictive rules for optical diagnosis of <10-mm colorectal polyps based on a dedicated software. Endoscopy 2019; 52: 52-60
• 2 Djinbachian R, Dube AJ, Durand M. et al. Adherence to post-polypectomy surveillance guidelines: a systematic review and meta-analysis. Endoscopy 2019; 51: 673-683
• 3 von Renteln D, Kaltenbach T, Rastogi A. et al. Simplifying resect and discard strategies for real-time assessment of diminutive colorectal polyps. Clin Gastroenterol Hepatol 2018; 16: 706-714
• 4 Lieberman D, Moravec M, Holub J. et al. Polyp size and advanced histology in patients undergoing colonoscopy screening: implications for CT colonography. Gastroenterology 2008; 135: 1100-1105
• 5 Butterly LF, Chase MP, Pohl H. et al. Prevalence of clinically important histology in small adenomas. Clin Gastroenterol Hepatol 2006; 4: 343-348
• 6 Vu HT, Sayuk GS, Gupta N. et al. Patient preferences of a resect and discard paradigm. Gastrointest Endosc 2015; 82: 381-384 e1
• 7 Hofstad B, Vatn MH, Andersen SN. et al. Growth of colorectal polyps: redetection and evaluation of unresected polyps for a period of three years. Gut 1996; 39: 449-456
• 8 Mizuno K, Suzuki Y, Takeuchi M. et al. Natural history of diminutive colorectal polyps: long-term prospective observation by colonoscopy. Dig Endosc 2014; 26 (Suppl. 02) 84-89
• 9 Loeve F, Brown ML, Boer R. et al. Endoscopic colorectal cancer screening: a cost-saving analysis. J Natl Cancer Inst 2000; 92: 557-563