Colonic adenoma detection and differentiation – does technology improve efficacy?

One of my teachers outside of my own teaching hospital, while I was still a fellow and later on, was Mike Sivak, one of the former editors of Gastrointestinal Endoscopy, and a truly intellectual endoscopist. He kept telling me that the most influential endoscopic technique is polypectomy. It is polypectomy of precancerous lesions which makes sense of screening colonoscopy, although final proof or disproof of this assumption by a randomized trial will not be available before 2020 [1]. Nationwide screening programs have mostly involved primary stool tests, but a few countries have also included colonoscopy as a primary screening examination in their armamentarium. Based on the main purpose of screening colonoscopy, namely to detect and remove polyps as precursor lesions, there are two main challenges with regard to diagnostic colonoscopy: the detection of colonic adenomas, as represented by the so-called adenoma detection rate (ADR) [2], and more recently the differential diagnosis of colonic polyps into adenomas and hyperplastic polyps. The latter could pave the way to reducing the need for histopathologic analysis of removed polyps. The endoscopic differential diagnosis between adenomas (requiring follow-up earlier than 10 years) and hyperplastic polyps (requiring no specific follow-up) would then finally determine follow-up intervals after polypectomy. This so-called DISCARD strategy (Detect InSpect ChAracterise Resect, and Discard; [3]) would save on histopathology costs associated with screening. High standards have been set by both ASGE and ESGE with regard to this issue [4] [5].

It follows that several measures of quality control have been implemented, and that factors have been elucidated that contribute both to adenoma detection rate and to the accuracy of differential diagnosis. Adenoma detection rate has been shown to be a surrogate parameter for the outcome of screening colonoscopy, namely cancer prevention, since the rate of missed cancers has been correlated with adenoma detection rate [6] [7]. We have recently found several patient-, examiner-, and endoscope-related factors that determine adenoma detection rate, in a large prospective study from Berlin [8]. Interestingly, there is evidence from this and other trials [9] [10] that the transition from one instrument generation to the next one does not consistently increase ADR, neither are special features, such as narrow band imaging or similar techniques, effective in increasing ADR [11] [12] [13]. I suppose that endoscopists themselves may be the most influential factor with regard to training, knowledge, and attentiveness and capacity for very careful scrutiny, the impact of the latter two abilities shown, albeit indirectly, by a variety of studies on withdrawal times [14] [15] [16]. Closeness of scrutiny cannot be reliably measured as yet within public colonoscopy programs, although computerized measurement attempts have been made [17].

The other challenge, namely the endoscopic differential diagnosis of colonic polyps is a more complex one. The initial approach to making this differential diagnosis in vivo and leaving hyperplastic polyps in place while resecting only those looking like adenomas, has led to a plethora of literature reports with mostly splendid results [4], but has not been realized in clinical practice for a variety of reasons. By common consensus only very small hyperplastic polyps in the rectum are left in place, and this is seen as especially acceptable if there are multiple polyps. A “safer” approach with regard to endoscopic polyp differentiation may be to apply the DISCARD strategy mentioned above: assess these polyps by their endoscopic appearance, classify them as adenomas or hyperplastic polyps, resect them, and throw them away. Follow-up intervals would then be determined by the endoscopic characterization. This strategy [3] has been shown to save big money on histopathologic assessment and to be safe from a cancer prevention standpoint [18]. However, high accuracy standards have been set in a recent position paper from ASGE: namely, required accuracy of more than 90 % in the prediction of post-polypectomy surveillance intervals, and required negative predictive value of 90 % or more for adenomatous histology [4]. In addition, a high level of confidence is required to make this differential diagnosis, and if this cannot be reached in a proportion of cases, these are disregarded as far as endoscopic polyp differentiation is concerned (and often not included in the accuracy calculation in studies). I am not so sure that we should allow confidence levels to be introduced into diagnostics: while on the one hand this reflects the real-life doubt (we should not encourage an illusion of certainty), on the other hand it helps those who do not want to take responsibility for a decision and may increase the rate of “definite perhaps” cases. These could then provoke further testing (similarly to the “MRI after CT” or vice versa sequence in radiology) and keep the medical business going. Thus, confidence levels may tell more about the individual examiner’s psychological status than about the status of a given lesion.

Nevertheless, irrespective of the everlastingly subjective element in image diagnosis, high accuracy rates have been shown in quite a few studies from reference centers [3]. However, we and others have shown that these excellent results cannot be reproduced in the routine daily practice of screening and diagnostic colonoscopy [19] [20]. Thus, we need either to take out the subjective element by computerized polyp analysis [21, or, less ambitiously, to improve endoscope technology so that the examiner’s eye can more easily recognize polyp structures, either live or from later analysis of photographs. The immediate post-colonoscopy analysis of images taken on the computer during colonoscopy appears realistic. In our study cited above [20], however, we found that many of the photographs taken even under study conditions were felt to be inadequate for a reliable post hoc analysis by experienced examiners, and that the rate was greatly variable, with five different evaluators judging 5 % to 50 % of pictures as unsuitable for such a post hoc analysis. Subtle movements of colon, breathing, and endoscope may render sharp photography difficult, and sufficient time and dedication for perfect imaging are unlikely in the setting of a busy screening practice. Therefore, in this issue of Endoscopy, Doug Rex and his group (22) have focused on a new generation of HDTV colonoscopes with an inbuilt special feature that selects sharp images by using software algorithms that allow easier image capture. The authors showed that the image quality of the photographs taken was substantially improved (from 84 % to 98 %) but accuracy of polyp differentiation was not [22]. However the accuracy rates of this super-specialized colonoscopy research group were already pretty high. Thus, it would be interesting to know whether this finding would also apply to groups with lower accuracy values as published by us and others from routine clinical settings [19] [20]. In any case, better image capture also has multiple advantages that cannot be measured appropriately, especially in gastrointestinal luminal areas that move a lot more than the colon (e. g. the esophagogastric junction), and so, in addition to measurable differences in confidence about polyp differentiation, more advantages may emerge. The DISCARD discussion continues and will lead to further interesting studies and papers.

• References

• 1 Kaminski MF, Bretthauer M, Zauber AG et al. The NordICC Study: rationale and design of a randomized trial on colonoscopy screening for colorectal cancer. Endoscopy 2012; 44: 695-702 (Epub 2012 Jun 21)
• 2 Bretagne JF, Ponchon T. Do we need to embrace adenoma detection rate as the main quality control parameter during colonoscopy?. Endoscopy 2008; 40: 523-528 (Epub 2008 May 8)
• 3 Ignjatovic A, East JE, Suzuki N et al. 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 (Epub 2009 Nov 10)
• 4 Rex DK, Kahi CJ, O’Brien M et al. The American Society for Gastrointestinal Endoscopy PIVI on real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc 2011; 73: 419-422
• 5 Rembacken B, Hassan C, Riemann JF et al. Quality in screening colonoscopy: position statement of the European Society of Gastrointestinal Endoscopy (ESGE). Endoscopy 2012; 44: 957-968 (Epub 2012 Sep 17)
• 6 Kaminski MF, Regula J, Kraszewska E et al. Quality indicators for colonoscopy and the risk of interval cancer. N Engl J Med 2010 2013; 362: 1795-1803
• 7 Corley DA, Jensen CD, Marks AR et al. Physician adenoma detection rate variability and subsequent colorectal cancer risk following a negative colonoscopy [abstract]. Gastroenterology 2013; 144: 2-S3 , abstract 9
• 8 Adler A, Wegscheider K, Lieberman D et al. Factors determining the quality of screening colonoscopy: a prospective study on adenoma detection rates, from 12,134 examinations (Berlin colonoscopy project 3, BECOP-3). Gut 2013; 62: 236-241 (Epub 2012 Mar 22)
• 9 Adler A, Aminalai A, Aschenbeck J et al. Latest generation, wide-angle, high-definition colonoscopes increase adenoma detection rate. Clin Gastroenterol Hepatol 2012; 10: 155-159 (Epub 2011 Nov 2)
• 10 Subramanian V, Mannath J, Hawkey CJ et al. High definition colonoscopy vs. standard video endoscopy for the detection of colonic polyps: a meta-analysis. Endoscopy 2011; 43: 499-505 (Epub 2011 Feb 28)
• 11 Adler A, Aschenbeck J, Yenerim T et al. Narrow-band versus white-light high definition television endoscopic imaging for screening colonoscopy: a prospective randomized trial. Gastroenterology 2009; 136: 410-416
• 12 Aminalai A, Rösch T, Aschenbeck J et al. Colonoscopy: a randomized comparison between FICE and conventional imaging (Berlin Colonoscopy Project 5, BECOP-5). Live image processing does not increase adenoma detection rate during colonoscopy. Am J Gastroenterol 2010; 105: 2383-2388 Epub 2010 Jul 13
• 13 Nagorni A, Bjelakovic G, Petrovic B. Narrow band imaging versus conventional white light colonoscopy for the detection of colorectal polyps. Cochrane Database Syst Rev 18.01.2012; 1: CD008361
• 14 Barclay RL, Vicari JJ, Doughty AS et al. Colonoscopic withdrawal times and adenoma detection during screening colonoscopy. N Engl J Med 2006; 355: 2533-2541
• 15 Simmons DT, Harewood GC, Baron TH et al. Impact of endoscopist with­drawal speed on polyp yield: implications for optimal colonoscopy withdrawal time. Aliment Pharmacol Ther 2006; 15: 965-971
• 16 Benson ME, Reichelderfer M, Said A et al. Variation in colonoscopic technique and adenoma detection rates at an academic gastroenterology unit. Dig Dis Sci 2010; 55: 166-171 (Epub 2009 Jan 21)
• 17 de Groen PC. Advanced systems to assess colonoscopy. Gastrointest Endosc Clin N Am 2010; 20: 699-716
• 18 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 .e1–3. Epub 2010 Jun 1
• 19 Ladabaum U, Fioritto A, Mitani A et al. 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 Epub 2012 Oct 3
• 20 Schachschal G, Mayr M, Treszl A et al. Endoscopic versus histological characterisation of polyps during screening colonoscopy. Gut 28.06.2013;
• 21 Tischendorf JJ, Gross S, Winograd R et al. Computer-aided classification of colorectal polyps based on vascular patterns: a pilot study. Endoscopy 2010; 42: 203-207 (Epub 2010 Jan 25)
• 22 Bade K, MacPhail ME, Johnson CS et al. New colonoscope technology: impact on speed of image capture, image quality, polyp differentiation, confidence, and accuracy. Endoscopy 2014; 46