Today, adjuncts to detection during colonoscopy on the ASGE SUTAB tip of the week. In this tip, I'm going to give an overview and update on devices and techniques to improve detection during colonoscopy, but beginning with the basics, high level detection begins with a colonoscopist who has a detailed understanding of what subtle disease in the colon looks like and doesn't overlook it when it appears on the TV screen. So individuals with ADRs of 50% high level detection of seriated class lesions, they understand subtlety and both groups of precancerous lesions, the conventional adenomas in the seriated class have part of their spectrum as being very flat and it's even more so in the seriated class. I think actually individual conventional adenomas of the same or comparable size to seriated class lesions are considerably more dangerous. The risk of cancer in lesions of same size is seven times higher in conventional adenomas than in the seriated class, but seriated class lesions contribute disproportionately to interval cancers because they're missed more often. So for both groups of lesions, we have to begin with a colonoscopist who has this keen understanding and looking carefully for subtle disease. So how do you become a master of detecting flat disease? Well, the colonoscopist has to be very tuned into recognizing extremely subtle changes in the normal shape as well as the texture and the color of the mucosa and disruption of the normal vascular pattern and fold structure. And so here is an extremely flat adenoma seen better after injection. And then there's the exquisitely flat SSL probably missed more often than adenomas. As I mentioned, it's already appeared on the screen, but we haven't focused on it yet. If there's a mucous gap, of course, that makes it easier, but in many cases, we have to rely on being tuned into very subtle differences in the surface texture and color of the mucosa. You can see it better after injection, probably a true Paris 2B lesion. I think most of the lesions that I would consider to be true Paris 2B lesions are SSLs. Error detection then starts with this knowledge and appreciation of the spectrum of subtle disease. And if you don't have that, don't feel that you have that, I think looking at lots of pictures of subtle lesions is the best way to improve your awareness. So if you have this excellent understanding of disease spectrum, you can then get extremely high ADRs, 45 to 50% in primary screening populations, 65 to 70% in FIT positive populations by just combining that knowledge with great equipment, meaning high definition, white light colonoscopes. That's the minimum and meticulous technique. You don't need anything fancy, no special devices or tools. This has been shown to be adequate, sufficient, and on multiple occasions by many endoscopists, both in academic and community practice to get absolute top notch ADRs. And that has to be kept in mind when you go into a discussion about special devices. Now, of course, you've also got to have split or same day bowel preparations and you've got to measure ADR. You're never going to know if you're any good. And fortunately, the process of measurement and reporting has been shown several times to itself improve detection. The basics of good withdrawal technique haven't changed. It's looking behind all the folds, cleaning up and distending adequately. And in the next video tip, I'm actually going to focus on this often overlooked element of achieving and maintaining adequate distension throughout withdrawal. One of the tools that I use now consistently in all screening surveillance and diagnostic examinations is double examination of the right colon. And we do this because we've repeatedly seen that interval cancers occur disproportionately more in the right colon. And we know that when you examine twice, no matter where you are in the colon, that that is an effective way to improve your yield of detection. So we're using good basic technique here. We're cleaning up, we're getting the right colon adequately distended. And the time consuming part is really the process of sticking the tip of the scope in between all of the hostile folds and doing that examination. We've seen several important studies, one from Minnesota, one from the New Hampshire registry showing that it actually takes eight to nine minutes in normal colons of examination time to maximize detection and cancer prevention. Once we get back to the Patek Fletcher, we go back down to the CCAM randomized control trial show. It doesn't really matter whether you do a second examination in the forward view or retroflexion. Here we're doing in retroflexion, come back to the hepatic Fletcher and then unwind the scope and continue the examination, but two right colon examinations in just about everybody. What about devices for increasing ADR? Should you use them? Are they worth the cost? There are two goals that they can have. One is exposing more mucosa and the other one is highlighting flat lesions, picking up subtle disease. For exposing more mucosa, if you look across the literature, you can always find individual studies that vary. I think Endocuff has been the most consistent device. We did a non-inferiority study comparing it to Amplify and found that it was the same. The current version, Endocuff Vision, I think is also effective. Average gains in ADR are about 7%, which is clinically important. You'll remember from Doug Corley's study in the New England Journal, 2014. We have this rule that for each 1% gain in the ADR, there is a 3% drop in the risk of interval cancer and a 5% drop in the risk of fatal cancer. Those are significant gains. We showed in a randomized trial that you can examine the colon about two minutes less. You can do withdrawal about two minutes more quickly and still have gains in detection. Here's the Endocuff Vision in action, using the fingers to expose the mucosa on the proximal sides of folds and the hepatic flexure. That, of course, is how you can go faster because the work of colonoscopy without one of these devices is going back and forth across households trying to get the nose of the scope in between the folds. If you have this finger device, you feel like you can get to the point where you've seen everything faster and move on. That's how withdrawal time can be short. The big problem with the device, of course, is the cost. This is what has held up its use in ambulatory surgery centers and office practices. I'm not in that kind of practice, so I prefer to have this device or an equivalent on the scope. For highlighting, I want to mention electronic chromoendoscopy and artificial intelligence. We used to say that for detection during colonoscopy, electronic chromoendoscopy doesn't work, but now we have to say it does work. This is data from James East patient-level data meta-analysis showing gains in detection with NBI compared to white light for both the old 180 system, which was too dark, and the new 190 Xera 3 system, which has brighter illumination. You can see, especially if the bowel preparation is excellent, you get really substantial gains in detection. I'm not showing here data for, if you have Fuji scopes, the BLI or LCI systems, but they also both have been shown to produce gains in detection. Here we're going to switch from white light to NBI. You notice that the bowel preparation is excellent. To use NBI efficiently, you need excellent preparation, but with an excellent preparation, I think it's a great way to examine the colon. I used to just like to use white light all the time, and I still use a lot of white light, but sometimes I switch over to NBI when the prep is excellent. I think that the color contrast between both adenomas, that's a small adenoma, and the normal mucosa is greater for both adenomas and serrated lesions. Having said that about electronic chromoendoscopy, it looks like the dominant technology for highlighting during colonoscopy is going to be AI. These are the first five randomized controlled trials in AI. All of them produce significant gains in detection. The average gain in ADR was 11%. We've not seen any other technology of any type for improving detection with gains of that magnitude. So here is the GI Genius, the artificial intelligence device from Medtronic. This is the first device to get FDA approval in the United States. The CAD-E program puts a box around the lesion. You can see it's working here in NBI as well as white light, and there you can see it's almost immediate in its pickup of polyps. I think it helps to slow you down during the procedure. There are occasional false positives, but I think the slowing down is part of the reason that it's effective and it does a really good job serving as an adjunct to detection. So finally, I want to point out that when we talk about devices and improvement in detection, for example, when I say there's a 7% gain with endocuff vision or an 11% gain with artificial intelligence, that means that's the average of a large number of endoscopists who participated in the study. So really, it represents the average of a range of detection from possibly no benefit to really substantial benefit, and you can't see that typically from looking at the results of studies. We are trying in our detection studies now to report the results for individual examiners, and I just want to show as an example of this study on EndoRings, which is a device that has an analogous principle to that of the endocuff vision, and you can see in the bottom line, there's a statistically significant improvement in adenomas per colonoscopy with this device. But if you look at the individual examiner results, you can see that almost all the improvement results from a single endoscopist who had a tremendous gain in detection with the device, and for all the other endoscopists, there's very little or no gain with the device, and this points out the operator dependence of improvement with devices. So if you get the opportunity to try, to sample, to use these devices and measure whether they improve your performance, that's the understanding you're really after. Thanks and see you next week on the ASGE SuTab tip of the week.

colon abnormalities

detection

colonoscopy

withdrawal technique

adenoma detection rate