false
Catalog
ASGE Annual Postgraduate Course: Clinical Challeng ...
ASGE Weekend Endoscopy - Colonoscopy, Colon Cancer ...
ASGE Weekend Endoscopy - Colonoscopy, Colon Cancer and Artificial Intelligence | April 2022
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
Hi, good morning, everyone. This is Prateek Sharma, and welcome from the ASG studios and headquarters. This is ASGE Weekend Endoscopy Live, live from the IT&T Center, brought to you by Medtronic. This is one of our latest offerings from the ASGE in keeping up with endoscopy education. This is going to be a series of quarterly events, again, brought to you live from the IT&T Center on different topics. Before we get started with today's program, which is entitled Colonoscopy, Colon Cancer, and Artificial Intelligence, I'd just like to go around and introduce to you our star cast faculty members and panelists. To my right, the current president of the ASGE, welcome Doug Rex. Doug is a professor of medicine at the Indiana University School of Medicine, and again, both Doug and I are here at the ASGE studios in Downers Grove, so welcome, Doug. Glad to be with you, Prateek, and everybody. Excellent. And then joining us on Zoom live is Jennifer Christie. Hi, Jennifer. Hi, good morning. And Jennifer is a professor of medicine and the clinical chief at Emory University in Atlanta. And then finally, you'll be seeing live very shortly Dr. Alessandro Repicci, who is professor of medicine and chief of endoscopy and gastroenterology at Humanitas Hospital in Milan, Italy. So this series is a little bit different than what all of you are used to in the past. So we'll be live at the IT&T Center recording this, bringing it to you live, and then this will involve today a series of live cases involving colonoscopy, polypectomy, as well as artificial intelligence, along with a few video cases that Doug has kindly brought on today so as to show you some tips and tricks of the trade in colonoscopy, a procedure that is near and dear to us and something that we do almost every day in our clinical practice. For all of you joining us virtually, again, in the Zoom Q&A box, just type in your questions. So we'd like to get to those questions. We'll be having a very nice and lively discussion with our panelists. So I'd encourage you all to do that. The way this hour will be structured as, as all of you are grabbing your coffees at home in your pajamas while Doug and I have this tie and suit on. So hopefully next time we'll have some pajamas too, Doug. But we are starting off strong with this. So we'll start off with a live case brought to you by Cesare Hassan and Alessandro Rapici and his team, followed by a Q&A session discussing colonoscopy in general. So any question you have related to colonoscopy, how simple, how complicated it is, please send it to us on the Q&A function, and we'll try to get to as many as we can during the day. After that, we will go on to some video cases that Doug will present, again, followed by a discussion and just some wrap-up comments. So if we are ready in Milan, Italy, let's try to go live there. And again, this is a case related to detection of colon polyps. So I do see Alessandro and Cesare. Alessandro is the smart-looking guy on the left of you, and Cesare, we can barely see just his glasses there. Hi, Cesare. So welcome, guys, to the first version of ASG Weekend Live. Alessandro and Cesare, over to you. Hey, thanks, Pratik. Welcome from Milan as well. Good morning, everybody. Good afternoon for us. So let me say thank you so much to you guys. Thanks to the ASG leaders for having us in this innovative program by ASG. Thanks also to Dr. Christie, and thanks to Medtronic for supporting this event. Just a few seconds to thank my entire team, because here it's late afternoon, and everybody should be home playing with kids rather than they are here working with us. So I will leave to Cesare. We will present a case of a relatively young guy with polyposis syndrome. So Cesare, this is to you now. Yeah, thanks, Alessandro. Thanks, Pratik and Doug. This is indeed a 47-year-old male with a myoid-associated polyposis, and this is a surveillance colonoscopy, also to decide whether the patient should have surgery or should continue with polypathy. This is the Fudifilm Eluxeo Tower, and this is the old way, Alessandro, to do a colonoscopy. You have this image. You need to spot here whether there is one polyp, more polyp, and then we can switch on the new era of diagnostic colonoscopy. Let me get the endoscopy picture. Yeah, perfect. Thank you. Yeah. And then suddenly you have all of these bounding boxes that come up and suggest you where a polyp is. So this is the system CAD-E by V.I. Linus that was tested here in two landmark studies. Please consider that today we have more than 10,000 patients randomized. But this image is extremely interesting. Have a look now. You have one polyp that is detected in any frame, and some other polyp, like for this one, that is detected only in some frame. For instance, here you see this denomatous flat lesion. So Cesare, I understand that because AIG is approved for detection, you're nicely showing detection, but you're also showing that the machine does almost immediate tissue recognition or characterization. Can you talk also about characterization? Yes, Alessandro. Here in Europe, we have already approved the system of characterization in white light. And this is a complete novelty, because for me, it is this the way of doing characterization, that is to switch in a blue light. Or maybe I should use a zoom like here in order to detect the pit. But now, with artificial intelligence, there is no need to switch in blue light. And what does this mean? It means that the machine will always be before us in making the diagnosis. So our role is to look at the diagnosis in white light of the machine. It may take some time. You need to keep the image very stable. And then you switch to BLI. And you need only to confirm whether or not the machine is right, both in terms of diagnosis and in terms of a confidence of level. And in my view, the machine will not affect so much the diagnosis, but it will affect the level of confidence. In other words, I think this is an adenoma. The machine says this is an adenoma. Now, I have a very strong feeling that this is an adenoma. Cesare, a very nice demonstration. Can you go back a little bit to detection? So incorporating this in the practice, what we can expect in terms of increases in ADR? This is the first question. The second one, there are a lot of discussion about the fact that AI just increases the detection of a small or diminutive polyps. Can you comment on that? Also, I would love to have comments from Doug and Pratik and Dr. Christy as well about the impact of AI in ADR, the relationship with small and diminutive polyps. OK, so first of all, our trials suggest a 30% increase in ADR, 55% decrease in the miss rate of adenoma. And this should translate every year in 7,000 cancers prevented in the United States. Is this unexpected? I feel no. Look at this image. You don't need a randomized trial to understand that this machine detects virtually any polyps. Of course, you can argue that you will have detected most of them, but the machine also. So now we are in two in detecting lesions. But just look at the flat lesion on the right of this. It's undetected. This means that we are not useless. We still need to interact with the machine. And if the machine has a false negative, we need to be the one to look at this lesion that was missed, for instance, this large lesion, and detect, look, this now was a false negative for some frame. Now it was a false negative. See? OK, now it is us to compensate the machine. But it's not just an AI driven. You may say, oh, my God, they need to detect only 20% of the polyp and the machine detect 80%. But we don't cure polyp. We cure patients. If this flat lesion is important for patient's sake, this is our contribution to the patient. Doug, Pradeep, can I have your thoughts about this? Yeah, I think that what Cesare is saying is that this is an adjunct to detection. So far in individual studies, we really have seen an increase in just the detection of diminutive polyps. But we have to remember that individual studies can only be powered for detection of adenomas overall. You can't really power them for the detection of advanced lesions. But we have now seen, as the number of studies increase, meta-analyses, and Cesare has been involved in one of these, Cesare and Ale, that have shown that also there is an increase in advanced adenomas. The device is only going to get better and better. But Cesare is right. Right now we have to look at it as an adjunct, an ancillary device that helps us detect. But we still have to look carefully ourselves. OK, Pradeep, I have a question to you. So most of the systems available, at least in Europe, have been trained mostly with pictures and video of small and diminutive polyps, because it was the most concern when we talk about adenoma misarray. Do you think we should train also with the larger flat lesion? Is this demanding from our side? Yes, I mean, Alessandro, that's an excellent point. And you're absolutely right. I think it has to be done on everything. We just can't say that this device or any device that comes up is only for small lesions or only for big lesions. It has to be very generalizable because the practice patterns are generalizable. You can't say it's just only for small practices, big practices. My question to Cesare and Ale to you, I mean, one of the questions which always comes up with using AI devices is who should have the final call and who's liable? Is it the machine or is it the endoscopist, right? Who does this all fall on at the end? So since you are using this so routinely, what's your answer to your patients? OK, I leave this to Cesare and I will give you my feeling as well. I have three concepts to tell you in the same time. First, we are using endocapha. Why? Because in a landmark forearm trial, endocapha was suggested to increase detection by exposing more mucosa. And look at here, we feel that endocapha help AI. Doug, I feel you show a 30% increase in the APC. That was the primary endpoint of your trial. And now, Alessandro and myself are performing this certain study where patients are randomized with AI versus AI plus endocapha. Secondly, I want to come back to the answer to the question of Alessandro. Are small polyps important? Yes, Alessandro. One every 20 polyps, small, harbor advanced neoplasia. Advanced adenomas start from the small subcentimetric size. They don't start from big polyps. So we need to remove 19 small polyps in nocus to remove one advanced small adenoma. Third question, blue. This system was trained in a case done with methylene blue. And you may see that this is probably the only engine that is able to exploit the staining of the blue to detect polyps. This may suggest in the future that dye can interact with artificial intelligence and may help both the detection or may have some effect on characterization. Of course, in both directions, because it's not the same for the machine than for us. So what help may be characterization of the man cannot necessarily help the machine. But dye, if appropriately trained, can help artificial intelligence. So Cesare, that was a very long answer to my short question. But I think we did get the message there. Jennifer, any thoughts from your side about this? I mean, things that we're not looking into as Ali and Cesare are getting ready for the second case as well. We can still look at these beautiful images as Professor Hassan is withdrawing. But what are some tricks and tips, Jennifer, that you have for us for improving adenoma detection rate? Maybe things you tell your fellows to do or you do yourself. Can you share those with us? Yeah, sure. Thanks, Prateek. So I think that one comment first is here, Cesare is using pretty much every, almost every technology available to increase detection and small polyps. And so I think that's very useful. But in terms of real life practice, if we had to choose one or two, which would those be? And I think those are things that we need to further understand and explore. As far as tricks or tips, I would say the first thing is the prep. I think the white light, the AI, the dye, all of that is dependent on the patient doing the best that he or she can to prepare so we could see. And then for our fellows. Go ahead, Jennifer. Yeah, sorry. And actually, I was going to ask a question about this, Alessandro. With the AI, how does prep impact the ability for the computer to detect polyps? OK, Jennifer, a question to you. So you didn't see, but Cesare pulled back the scope to the rectal simoid because I think he want to start talking about the resect and discuss strategy. There is a lot of discussion on that, and I think having AI doing detection characterization, we may have more strength in adopting this in clinical practice. Jennifer, the question is for you. You think we're ready once the system is incorporated in your practice to go with the resect and discard strategy, or are we still far away from that? From resect and? Resect and discard. Resect and discard. OK, yeah. Um, I think we still need to understand the impact of these, you know, diminutive polyps on on all patient populations and whether they are, you know, not as concerning. And are we better able to risk stratify? I think one of the important points for AI is that it may help us to determine which ones we can actually resect and dispose of them in terms of stratification and the risk of those developing into birth and polyps and then what that follow up should be in those patients. And Jennifer, to your point, I think, you know, the bowel prep issue is extremely important. And, you know, it's true. So AI can only detect what it sees, right? And so if your bowel prep is really poor, you know, it's like trying to find a polyp underneath a pool of stool. I think it's extremely difficult even for the machine to do that. So your point's well taken. I think it's extremely important that the bowel prep remain the same. Now, just following up, Doug, on the question of resect and discard, and Cesare, excellent images there, by the way, of us looking at those small diminutive polyps. All of them are being analyzed as non-adnomas. And so, Doug, again, the question, I mean, we face this with virtual chromoendoscopy, dye-based chromoendoscopy, and the same question persists now for AI. So what should we be doing? I mean, are we there yet to just remove them and throw them away? No, we're not, because it's a really complex situation. I think we are there from the standpoint of the technology and from the medical concepts. But we have barriers, and those are the concerns. I think about medical legal risk, we have to figure out a way to handle that. I think our societies need to say not only that it's okay to do resect and discard, but that we should do resect and discard, because resect and discard makes a lot of sense from a cost standpoint. And there are some financial, well, there's no financial incentive for doing it, and there's some financial disincentives in some practices that actually own pathology services. So I think with any paradigm that we'd like to move forward, really, resect and discard is a paradigm that makes sense to move forward from a medical standpoint, but there are political, financial, medical legal considerations that can impede it. But this is a tool that makes it more feasible, I think gets us closer. And as Cesare is showing us now these beautiful pictures of what are probably hyperplastic lesions, not sessile serrated lesions in the rectosigmoid, this is related to resect and discard. This is the recognize and leave in place phenomenon that we talk about in the rectosigmoid that we're already, many of us are doing already, but it's going to give us additional confidence to when we see a small rectosigmoid lesion, say it's okay to leave this in place. Absolutely, Doug. This is the living CT and we did the study showing a 98% negative predictive value for rectosigmoid lesion and we hope that this may help, as you said, make it more clear. We can photodocument this lesion, we can give this to the patient and probably the patient is more happy than with what we are doing now, that is not to give any clear report unless we do only one biopsy to do histology. So the living CT strategy, in my view, is the way we can use to make optical diagnosis with AI enter in our field. Then a reset and discard will come, but at least let's start with the upper plastic polypane and the rectosigmoid that is non-adenoma. I do the picture, I give it to the patient, I keep it for my record and I feel this is the most simple and acceptable way to have living CT in. Just look at this and we can estimate a saving of 80 million per year, according to the Wikimori cost effectiveness paper in the US, that is anyway relevant and can become an incentive some time, some water to have AI. But this is an important point, you're the president of the society right now, so the problem you say that I see very clearly, so we need to convince people, gastroenterologists in the community practice, either in the US or in Europe, this is to be adopted. I think that having a massive adoption of AI can probably have a very positive effect, incremental effect, to convince people they have more confidence. It's like having a double check with the machine, which is very efficient, very fast, very quick, but also very precise and indicating what kind of astrology is the polyp. You think that the level of society, is there anything else that has to be done, both on the American and European side, to implement this in the practice? Yeah, myself, I think the big change we're going to have to make, Ali, is that we're going to have to move from the position of saying, resect and discard makes sense, and you can do it, to saying that it's the right thing to do and we should do it, because that's going to give us the backup. Because many of us, for example, we have to make this conversion at the level of our institutions. Our institutions often require us to take every piece of resected tissue from a patient and submit it to pathology. You have to go into your institution and convince them from a financial, from a legal standpoint, that this is okay. But I think the societies can help with that. But this is a major, major paradigm shift, and it requires, I think, mutual understanding and agreement on multiple levels to make it go forward. But yeah, we can help. The societies can help. Now, Doug, between the two is characterize and leave behind. Is that a much easier paradigm to switch to first, which we already are doing, as you mentioned, but still, you know, I think our membership, the majority of the endoscopists still don't feel comfortable doing that. So do you think that's the earlier thing we should adopt first and promote before going to resect and discard? Yeah, it's easier. And I think in reality, we're actually already doing recognize and leave in place. And AI can give us a very helpful backup check for that. In the U.S., what I tell people is a little bit different, I think, than what Cesare is saying. We actually say, check it and either biopsy it or leave it alone. And then don't put it on the report, because it's always possible that somebody could develop Which is even more illegal than doing that. Which is even more illegal, not legal, if you do not put in the report. Yeah, I disagree. I think if you think it's hyperplastic and it's in the rectal sigmoid, and it's five millimeters and smaller or less, then you should just say it's not there. You should consider it normal. Because what you do if you say that is you give somebody the opportunity to second guess you. Now, if you take a picture of it and a convincing picture, now you have a record of the decision that you made. And that's someplace that AI could get us to. In the absence of that, if I see a little ditzel or if I see 10 of them and I remove two of them, I make no mention of the other eight. And I think that's the best thing. I've seen plaintiffs' experts say that was an adenoma that you didn't remove. And that was the cause of an interval cancer. Of course, it's very unlikely that it was. But as long as there's no photograph, you leave that possibility open. But getting back to AI, I think AI gives us the potential, especially when we have an excellent, high-quality photograph that supports the decision that we make. And then that takes that medical legal risk away, largely, I think. So let's get to a question from Thomas, one of our viewers. So, Alessandro, if you want to get set up in your second room, he'll be looking here. I go to the other room in a few minutes, and we'll be ready. OK, thanks. Cesare and Alessandro, thank you very much for your beautiful case. And Cesare, congratulations. A beautiful demonstration of the AI device there. And for all of you who are joining us now, this is ASG Weekend Live, live from the IT&T Center. And again, being sponsored by Medtronic. So we are grateful to them for this support. The question from Thomas, Doug, probably for you, is that now that we've switched to, with the new guidelines, the 7 to 10-year recommendation for a small diminutive polyps, two or less, how about somebody with a positive family history? Does that still change, or does it remain the same? Right. So you always want to pick the shorter interval that's dictated by either the surveillance guideline or the screening guideline. So the screening guideline says that if you have a first-degree relative with cancer or advanced adenoma at age less than 60, or two first-degree relatives with some combination of those, that you should get screened by colonoscopy every five years. So in the setting of one to two tubular adenomas in somebody whose dad developed colon cancer at age 50, you would still go at a five-year interval, because that screening recommendation dictates five years, even if the examination were normal. On the other hand, you shouldn't go shorter than five years on the basis of one or two tubular adenomas. Jennifer, with the change in the age now that all the GI societies have endorsed, now starting at age 45 for screening, that's obviously opened up a lot of potential subjects who need screening. And of course, we have limited resources. So tell us about some plans you, let's say, are implementing at Emory. What are your suggestions? How do we get to them? And of course, folks don't come in for screening. So what are some ways we can get more patients or more subjects screened? Yeah, thank you. So ideally, we want to try to get everyone who is appropriate for screening for colonoscopy. However, as you said, the resources can be limited, and not everyone has access to colonoscopy. So we encourage our patients and our primary care providers to get the test done that is able to get done. So we support the FITS test. We support the Cologuard test. And we also provide resources for screening colonoscopy. And what we found is that the younger population has been extremely motivated to get screened. They want to get done, particularly with the recent death of major actors and actresses in our society. So we've actually seen a very good uptick. And also, we're expanding and trying to just make sure we have access. But education is really the key in making sure that folks know that there are several ways in which to get screened. However, we really try to make colonoscopy available to whomever we can. OK, thanks very much, Jennifer. Now we have Milan ready. I see Roberta there. Buongiorno, Roberta. Good to see you again. Thank you again on a Saturday afternoon for being live with us. So over to you, Roberta and Alessandro. Please tell us about the case. Oh, thank you. This is a screening colonoscopy patient has been referred to us for this relatively large lesion into the sebum. It was originally referred as a serrated side lesion. Indeed, looking to the lesion, it's clearly not a serrated histology, not a serrated pattern. So Roberta, do you want to comment on that? Yeah, me too. I was expecting the serrated lesion in the second. But here is what we have is just behind the valve. So the position is not very stable. But if we look and we focus deeply on the lesion using also chromendoscopy. At first with LCI also to check for margins, because it's very nice for detection. And then also mainly with VLI to check for vessel patterns and also glandular patterns. It looks like to me an adenoma, not a serrated lesion. So the first question is, OK, we have to remove it. But how to remove it and in which way? I would like to know your comments and your suggestions. This is a flat, probably two and a half centimeter lesion. Adenoma to split pattern, 3L in the cecum. What is your first suggestion? Can we go cold? It should go traditional piece mille MR. Right. Great question. I fully agree that this is an adenoma. The Paris classification is 2A. The nice pattern is 2. This is a granular lateral spreading lesion. It has a very low risk of cancer. That's important in the decision making process. I would say probably no more than a 1% risk of cancer. So I think EMR is the fastest and safest way to remove this. We have a couple of trials that are going on around the world right now, comparing hot EMR to cold EMR in this instance. But I must say that if you're choosing a lesion for cold EMR, the downside for adenomas, this is not true for sessile serrated lesions, is that the recurrence rate is clearly higher. This is probably a lesion that is a pretty good candidate or a reasonable candidate for cold EMR because there's not a lot of bulk to it, a very low risk of cancer. It's granular. It probably will have very little submucosal fibrosis. It should be easy to cut through. So I think if this patient I saw in our randomized trial, I would feel comfortable going either way. But in my own practice, I usually go hot still because I like that very low recurrence rate now that we get with hot EMR. But I would not quibble either way, Ale, with you guys choosing cold or hot. I have a question to Jennifer. Jennifer, the lesion is located into the segum just right next to the ileocecal valve. Is this an additional point in favor of going cold in order to reduce the risk of perforation and risk of delayed bleeding? It is to you and Roberta as well. Jennifer. Yeah, I definitely think that because of that location, it probably makes sense to go cold. Now, honestly, because of just what I'm used to doing, I, like Doug, will probably do this. I would lift it and do it hot. But certainly, you know, evidence shows that it may be safe to do it cold. Okay, good. Yeah, that was exactly my comment because we have factors coming from the morphology and appearance of the polyp. So to decide which treatment is the best and which method suits best for the removal of this polyp. But other factors is the location, the position, and some factors deriving from how to approach the polyp. And the fact is that we are in the secum, so we know that the wall is very thin. So the risk of mucosal, deep mucosal injury from the electrocardiogram is very high. So to me, it's better to remove it by cold. Good, so let's go cold. In the meantime, I have a question to Doug. There is a recent paper in gastroenterology from Helmut Messmann Group showing that also you can do this underwater and doing underwater you have a shorter operational time, a little bit bigger pieces, reduce the number of pieces. So is underwater another option on the table, Doug, or you're not considering since it's in the secum? I think underwater is an option. I like underwater the best for lesions that are in the, you know, under 30 millimeters where I think that I really want to get it out on block. And I would, my feeling would be that with this lesion being granular and it's low risk of cancer, I don't think we have to try incredibly hard to get this out on block. But if we did go underwater, we definitely would have a better chance of, you know, using a big snare and taking it out in one piece. So if this, for example, was a non-granular lesion or it had a bit of pseudo depression in it, I would be much more tempted to go at it underwater, which of course we do with electrocautery. Okay, good. Are there locations that you prefer underwater over traditional EMR, the location of the polyp rather than just the morphology that you just described? I think the location, especially sometimes if you feel like you're going to have a very hard time accessing the lesion, you know, with injection, sometimes you'll be surprised when you take the gas out of the lumen and fill the lumen with water that the thing is actually more accessible underwater. And underwater is often very nice in the sigmoid when you have a pedunculated lesion that you're having a hard time getting the snare just where you want it. Because the patient's in the left lateral decubitus position and when you fill the sigmoid with water, it fills very easily and a pedunculated lesion will actually float up into the water and it makes it very easy. I think that... It's a great question, Prateek. I think location can be a factor. Yeah, I think that also Roberta is nicely showing like a combination of underwater and non-underwater can help in catching more tissue when you are in a difficult position with the scope. And also I think that going underwater sometimes gives you a better clear view of the margins. So this is another potential... No, Roberta, we saw your initial injection which was wonderful. I mean, you raised it so nicely. Can you just tell us some tips as to how you perform your injection and how much do you inject? How do you start proximally and then go... We missed that part, but it was really nice. Can you tell us? Yeah. In this particular case, I decided to start injecting in the oral side because I wanted the lesion get lifted to me because of the difficult of positioning because I am behind the valve. So I want some helping in having the lesion in a better position than to place my snare and to try to catch the proper tissue. And then I'm applying, as in this case, water to check for margins. And then when I'm sure, I have to close the snare, as in this case. If it's too much, I have to open, as in this case, and I pull the snare to me to grab the tissue. I personally, and I disagree, I mean, you're the experts, but would have gone hot. Partly because of the recurrence. Doug, do we know in such polyps, is there a higher recurrence rate with cold versus hot EMR? Do we have the evidence? Are we still waiting for it? And so we should stick with hot until that happens. We don't have a randomized controlled trial for adenomas. We have a lot of observational evidence for serrated lesions that the recurrence rate is quite low. But for adenomas, the recurrence rate is substantially higher. And we know this from observational data, including from Cyrus Paraka's group. And Cyrus is actually just the grand champion of cold resection. One thing I really like, Roberta, is that you're using the water jet a lot during the resection, at least it looks like you are, pushing water in there. And I really like that. It expands the submucosal space and cleans the base for you, helps you really define things as you're working. So, yeah, go ahead. You know, Doug, excellent point. I have a question to you, because we're using a 10 millimeter dedicated cold snare. And of course, this is our practice. In general, for this lesion, and for this approach, the risk of perforation is very low. If I remember well, you did a report, a single case of perforation with a cold resection. What is your feeling in terms of risk? Are we running less risk in this or not? Oh, the risk is unbelievable. So when we do cold resection, we basically are trading this fabulously low risk of a complication, either perforation or delayed hemorrhage. It's almost non-existent in the absence of anticoagulation. So we don't have to clip this. If we removed it with electrocautery, it would meet all the criteria for clipping. Larger than 20 millimeters, I would say, I thought it was about 25 millimeters or so in size. It is proximal to the splenic flexure. And if we remove it with electrocautery, we should clip it closed. When we do this cold, we don't have to clip. We don't have to essentially worry about complications. But the recurrence rate is significantly higher. I think one thing I do, Ali, is I actually go sometimes by my impression of the patient. If sometimes I like to go cold with a granular lesion that's very big that I know that I can't clip. If I trust very much that the patient's going to come back. Whereas if I'm concerned that I'll never see them again, They've sort of been pushed to come, they've canceled a couple of times. I want to go hot because I want that recurrence rate to be very low because I may never see the patient again. But the recurrence rate is related to size and when you get up to several centimeters in size, you are having recurrence rates for adenomas that are in the 50% and even higher range. Very nice paper from Cyrus Paraka's group on this in EIO a few months ago. Yeah, and also you're doing an excellent job in removing the hard tissue. Yeah, to reduce the recurrence I think it's important that you carefully look to the borders. You may cut a little bit of normal tissue all around, like thermal ablation recommended by Michael Bork for standard EMR. So she's carefully looking to all the borders. I think having done this very precisely, you can bring down the risk of the recurrence a lot. So it's also a matter of being precise in space, spending a little bit more time, but overall it took about a couple of minutes to complete the resection and I think we're done. So you see how far you can go. This is the muscle of the segum exposed. You can show by underwater, you see here, and it's incredible. There is no bleeding at all. So this is surprising me all the time. We are not using electrocaudary, but the amount of bleeding vessels, even a small oozing, it's almost nil. So Pradeep, do you want to comment on this? Yeah, no, I mean I think that's the major advantage of cold snare, right, is if you get any bleeding it's right there and you treat that and there's almost studies and expert hand shows almost zero delayed bleeding. So very nicely done. I mean, I'm really amazed. I mean, that's why you guys are the masters, Ali and Roberta, and you've shown us characterization, virtual chromoendoscopy, dye-based chromoendoscopy, underwater resection, and cold snare all within 10 or 15 minutes. You know, this ASG live weekend endoscopy is only for a one hour, unfortunately, so we have to move on. We could go on looking at this forever with you guys, which is great. Stay on. We have some cases with Doug now, so Ali and Roberta would love you. Thank you so much, guys. We are indebted that we keep following all of you. Thank you so much, guys. Oh, thanks, guys. Okay, so ciao from Milano. Okay, ciao. Grazie. So we move now, while you want to pull up your videos, Doug, we do have a slew of questions here. And Jennifer, I can ask you this first one, which is, you've done a colonoscopy, and this is a question by an anonymous attendee, saying that you've done a colonoscopy, which is negative. The patient comes in with rectal bleeding within one to two years. Do you repeat the colonoscopy, or do you just say it's probably related to hemorrhoids and not do it? Jennifer, any thoughts on that? Yeah, well, I think we do have literature to suggest that if the patient has had new onset rectal bleeding, so the colonoscopy was not done for rectal bleeding in the first place, and we found hemorrhoids, but new onset rectal bleeding, and they are two years out from their screening colonoscopy, I go ahead and do it. And certainly in someone who is older than the age of 40 or so, but my threshold has become very low for repeating the colonoscopy, because we know that young age, these, you know, colonoscopy can miss lesions to a small extent, but it can happen. And also that we're seeing early onset colorectal cancer. And again, if we can find something early and treat it, then we'll have much better outcomes. Okay, thanks, Jennifer. Doug, you have some video cases to show us, please. Yeah, so we've been talking about AI, and right now in the U.S., AI is available, it's what we call CAD-E, which is for detection, and detection is incredibly important. The new wave in quality over the next few years is going to be on the quality of resection, because we've seen when you look at video recordings of different endoscopists, that competency in resection varies about threefold, about the same that you typically see across a group of gastroenterologists in the same practice group in their ADRs. So this is kind of, I think, the new wave. And we really saw a nice demonstration just coming from Milano of cold resection in the cecum, which kind of, you know, as part of this cold revolution, it's like cold is taking over, Pratik, we hear that all the time, and especially for small lesions that are a centimeter or less in size. So I think something for us to consider is, are we doing cold really well? And so the keys to this, there's actually a scale that you can use called the CSPAT scale, cold snare polypectomy assessment tool, and it's worth looking up. And the keys to it are that we position the polyp correctly down in the five or six o'clock part of the endoscopic field, that we keep a good working distance from the polyp. So here's a lesion that's up around two o'clock. And you can see that we're rotating the scope so that we put the lesion down at about five or six o'clock. Many people, there's a tendency to work on lesions at eight o'clock or two o'clock. I think that's okay if the colon is turning in front of you, and you're very on fost to the lesion. But if you're working in a section that you're relatively tangential to the wall, like you are here, it's better to work at five or six o'clock, because you see the lesion throughout the resection. And then I think the next step is accurate snare placement. You want to place the snare so that you have a margin of normal tissue visible around the polyp. And we're looking at this little lesion, we're characterizing it with NBI. And now we're rotating a little bit to put it in the five or six o'clock position. And then we're going to place the snare. This again is a dedicated cold snare. And notice that margin of normal tissue that's around the polyp. Another key thing to learn is just the right amount of deflation so that the snare grips the tissue well, but so that you don't get stuck on the submucosa, or at least you get stuck on it. We saw Roberta get stuck for a moment. She kind of backed off and regroup during that cold resection. So what do you do when you get stuck there? Do you just open it and jiggle the snare back and forth? What's the trick to, you know, do you have a video of that? Great question here. I think I've got some video of this because it happens a lot, right? So here's a lesion that's probably about eight millimeters. It's a serrated lesion. We've got the cap cold. And again, we're shooting for that accurate placement. And we got to make a judgment about should we deflate a little bit. And then initially we're going to get stuck. And what we're getting stuck on is some submucosa. And you saw me pull the polyp up to the scope and just pull it off. But some people don't like that. We call that snare stalling. And, you know, there's a little bit of concern that it's, I think it's kind of a bad form. But these cords, that white cord that you see there in the middle of the screen, a couple of studies have biopsied that cord. And one found that it never had polyp tissue. And the other one found that only in 2% of the time did it have actual polyp tissue. So that cord is usually stripped completely clean. But some people feel that when you get stuck, the thing is to loosen the snare a little bit, sort of lift up toward the lumen so that you're moving higher up that cord and then try to cut. Some people will saw a little bit. I don't think you should pull, you know, incredibly hard. But if you can pull it through and not pull too hard, I think it's okay. I do it all the time and never see any problem with it. How about that little nub that you see, Doug, when that happens is that submucosal tissue, right? Yes. Do you go ahead and remove that? No, that's a great... Or do you leave that behind? Leave that behind, Prateek, because that is just submucosal tissue. Here's another case of doing it. The tissue at the perimeter, and of course it's going to increase in size. It's going to increase in likelihood as the lesion gets bigger. Both of these lesions are close to a centimeter in size. I would say eight or nine millimeters, and so the size is going to affect it. But no, that thing right there in the middle, that we call that the submucosal cord, and you want to leave that alone. Now, I'll show... This is an example of a serrated lesion, and we've been talking about detection. And these lesions are missed more than adenomas. This lesion is probably about eight or nine millimeters in size, too. But you heard about cold resection of an adenoma in the cecum. Basically, all of the serrated lesions now can be removed using cold techniques. And so this lesion is probably about eight or nine millimeters. If it were 10, 15 millimeters, if it were too big for the snare, you could still get it, but you would piecemeal it out. And I think when you go... Let's say right now we're going to go at this lesion, and you'll notice that the snare is right around the perimeter of the lesion. And that's a 10 millimeters cold snare. If it were 12, 13 millimeters, we'd have to piecemeal it. I think it's better to take a nice border on both sides of it than try to cram something that's around 10 or 11 millimeters into the snare. When you're finished, of course, we get some immediate bleeding. We can just largely ignore that, tamponade it with some water jet, forcing the water into the submucosa there. So this technique has really revolutionized the removal of small polyps. And we mentioned that you can do it for one or two millimeter polyps. I see, critique in practice, a lot of people still removing polyps with cold forceps. And for lesions that are up to maybe two or three millimeters in size, that's probably okay, especially if you're using large capacity or even jumbo forceps. But above that, it's not as effective and it's not as efficient. So I really pretty much let cold snaring take over the removal of everything that's 10 millimeters or smaller. We just completed a randomized trial, just came out in gastrointestinal endoscopy, that cold snaring was effective for lesions up to 15 millimeters in size. I think cold snaring- And this was for SI lesions or for all lesions? All non-pedunculated lesions. But you raise a question, Pratik, which I have a video of here about pedunculated lesions, because we've seen a couple of papers now, here's a little pedunculated lesion. And a lot of people think, oh, I've got to remove this with electrocautery, but actually you don't. Up to about 10, 12 millimeters, it's quite safe to remove them cold. We just cut through the base of that without electrocautery, using a dedicated cold snare. Occasionally, you get a little bit of bleeding. You can, again, stop it with the water jet or push on it with the end of the scope. Or the best thing is just ignore it, go on about your business because it's going to stop in a minute or two. Right. The members who are tuned on here are thinking, well, this is Doug Rex doing it. But what do the guidelines say? Do they actually tell us that pedunculated lesions can be cold snared? No, they don't. And it's a good point. Touche. But I do think we're seeing literature. And I would say, I think for everything up to 10 millimeters, non-pedunculated or the occasional pedunculated lesion, it's okay to go cold. Now, there's a few questions related to cold snare. And Jennifer, chime in please for it is, what's the reason of this high recurrence rate after cold snare? Is it the depth that we're not getting enough? Or is it the margins we're not getting enough? Is it a combination of both? Or is there another reason? So it's a fabulous question. And I want to make the point that this recurrence rate is confined to adenomas that are 20 millimeters or larger. We are not seeing evidence that there's a problem with cold removal of lesions smaller than that, including the adenomas or for serrated lesions of any size. So this problem with recurrence has to do with EMR of conventional adenomas, like the one you saw from Italy, that are 20 millimeters or larger in size. And I think Pratik, it's the depth. We have seen a very nice study from Japan showing that when we resect cold, we are more likely to leave the muscularis mucosa intact. And therefore, some cells are surviving on there. The thing about serrated lesions is that they're so loosely connected to the submucosa that it tends to be very easy to cut below the muscularis mucosa. Adenomas tend to have a bit more fibrosis underneath them. And so here's an example. But it's a good point, and we don't fully know it. But what we were talking about of cutting the margin out completely with a cold snare, we don't know that that's going to fix the recurrence rate problem. I don't think it is for large adenomas. The reason I thought is because the soft tip coagulation does decrease the recurrence. So that's why I thought perhaps it's the margins which are also equally important. It could be. And right now, we don't know. We had a historically controlled study from Australia with hot EMR of cutting the margins out, and it didn't work. But snare tip soft coag did. So I think nobody understands that right now. But we would say that based on the snare tip soft coag studies for large adenomas, the recurrences do come in from the margin. But I don't think we know that for a cold resection. They could well be in the base. So we have a couple more minutes to go, believe it or not. An hour is almost up. So, Jennifer, one question is related to the amount of volume to be used during EMR. Do you have a set volume, Jennifer, or is more always necessary better? How do you decide that for injection? Yeah, I usually decide just based on how the palate responds to my injection. So I try not to go over 10 cc, but generally just how the palate responds and whether I feel comfortable or not with the injection. So I don't have a hard stopping volume, but rather, you know, case by case. Got you. Doug, any comments on using epinephrine within the mix, or do you just go just plain? So I think it's a matter of personal preference, and we just completed a randomized trial. It was just published in Gastrointestinal Endoscopy, where we found that using 1 to 200,000 epi for large EMRs, that patients had more abdominal discomfort afterwards. And I do think that there's an element of ischemia that lasts for an hour or two when you put a large volume of epi in. So I still occasionally use epi, but I'm sort of less in favor of it than I used to be. I think that's the downside of it. It does keep the field drier. Most times when you see a really large EMR defect that looks just perfectly blue, like it hasn't bled a drop, that was done with epinephrine. That's the epinephrine effect. Okay, great. Sorry, go ahead, Jennifer. Yeah, no, and I'm always concerned that it's giving me a false sense of security and that, you know, it's dry now, but maybe tomorrow, two days later, it won't be. Yeah, the ESD people say, Jennifer, that they don't like to use epi, some of them, because they want the bleeding to occur during the procedure. And immediate bleeding is unquestionably better for the patient than delayed bleeding. Yeah. Okay, just about post-recurrence follow-up. So the piecemeal EMR we saw from Roberta, they would probably bring the patient back in 6 to 12 months to look for recurrence. Let's say they find recurrence and they remove it. Does that still continue 6 to 12 months or is it earlier, is it later? So the standard recurrence, standard follow-up pattern after hot EMR has been 6 months after resection and then 12 months after that. So you do two follow-ups. I basically, recurrence or not, tend to follow the same pattern. So if we would come back at 6 months and there was a recurrence, I actually always treat the recurrences with electrocautery. We don't have randomized control trials on how to treat recurrences, but you're dealing with scar underneath there. And the good thing is recurrences are almost always small. And even if you had a huge lesion, you'll have a small recurrence. You can remove it with electrocautery, either the snare or avulsion. And then you can close that defect with clips and prevent delayed complications. And then we would still go on to a year. I think there's an increasing tendency and we've seen data from Michael Burke that even up to 30 millimeters, a lot of times we can do that first follow-up at a year. And I often now go straight to a year for the first follow-up. And if the base is clean, then we go directly to 3 years rather than that 6 months, 18 months, and then 3 years after that. Okay. It's 3 minutes after the hour. So we unfortunately have to wrap this up. I know there are a few questions we've not been able to get to. We'll try to respond that in the chat box. But again, Jennifer, thank you for joining us, you know, virtually from Atlanta. Great to see you. Doug, great to have you here. And of course, to our group in Milan for some wonderful live cases and to the audience for joining us with coffee on a Saturday morning. Again, on behalf of ASGE and Medtronic, I'd like to thank all the participants as well as our esteemed faculty. Just a couple of housekeeping things is send us our comments. If you have, you know, suggestions for future topics, let us know. And two upcoming AI events that I want all of you to be aware of. On the Friday of DDW, join us for AI workshop. And on August 20th, it will be our ASGE's fourth annual GI AI Summit in San Francisco. And also in another 3 months, look forward to another ASGE weekend live endoscopy, live from the IT&T Center. So again, thank you all very much.
Video Summary
In this video, Prateek Sharma introduces a live endoscopy event titled "Colonoscopy, Colon Cancer, and Artificial Intelligence" sponsored by Medtronic. The event features several faculty members and panelists, including Doug Rex, Jennifer Christie, and Alessandro Repicci. The discussion focuses on various aspects of colonoscopy, such as the use of artificial intelligence for polyp detection and characterization, resect and discard strategies, and the technique of cold snaring for polyp removal. The panelists share their insights and experiences, addressing questions from the audience. The goal of the event is to provide education and discuss advancements in endoscopy techniques to improve patient care and outcomes. The video also includes live demonstrations of colonoscopy procedures and highlights the importance of accurate detection, quality resection, and appropriate follow-up in colonoscopy practice.
Keywords
Colonoscopy
Colon Cancer
Artificial Intelligence
Medtronic
Live Endoscopy Event
Polyp Detection
Polyp Characterization
Resect and Discard Strategies
Cold Snaring
Patient Care
×
Please select your language
1
English