I have the pleasure to introduce Dr. Offman, who is the William T. Butler Endowed Chair of the Distinguished Faculty, Professor of Medicine and Chief of the Gastroenterology Division of Baylor College of Medicine. He is well-published and knows so much about the complex polyp. We've asked him to talk on what to resect and when to pass on resection. So Dr. Offman. Thank you so much for the introduction. And these are my disclosures. So all of us get excited about resecting polyp, removing it, but actually the most important step is that topic, when to do it, when not to do it. And I'm sorry if this will appear academic, we're going to talk about a lot of numbers and data and lymph node invasion, but so important that we know that in everyday practice. So if we look at any process of cancer in the GI tract, the early malignancy, and we can see here in particular in the colon, the lymph channel or lymphatic channels can go all the way up to the submucosa. So you might be able to remove a tumor like this one here, but the big question would be, do we have tumor cells in the lymph node, what we call lymph node metastasis, LNM. Because resecting this and having recurrence two to three years later with liver mets, for example, would be a disaster. So the question would be when we can do resection safely or when we send our patient to surgery. So the surgery is not only about resection, it's about lymph node dissection. But I'm going to tell you something not related to this lecture. One day we will be able to do lymph node dissection by endoscopy, and maybe this talk will be totally different about what to select for these lesions. But as of now, since we're not doing lymph node dissection yet, at least in human, we'll do it in animal models, then we will have to avoid doing lesions that they have high risk to go into the lymph node. So I wish every time we have a polyps, we can see a graph like that, that can tell me how deep the tumor is and whether it's going to lymph node or not. Because we cannot see how deep it is, we only see the surface. So we have to learn to work on the surface pattern to predict lymphovascular invasion and lymph node metastasis. But also we take biopsies and biopsies can help us. So we'll start first with a multi-society task force. It's one of the best paper came in the last few years on evaluating the polyp. I really encourage all of you to read it. It was published in 2020. I feel it is so important. I almost feel like we have to teach our fellow this paper. And in that paper, that's part of it about the Paris classification. And the Paris classification, it's a group they met in Paris to decide on morphology of all GI lesions, stomach, esophagus. And the idea is to have one terminology based on looking at the lesion morphology. It appeared complex, but it's very, very simple in its simplest form. It could get complicated. But basically, you can have raised lesions or flat lesions or depressed lesion. And this will be 0.1, 0.2, 0.3. And after in each one of them, there's variation. The flat lesion could be bedonculated or sessile, the raised one. And then the flat could be slightly elevated or really flat or slightly depressed. And they will go 2A, 2B, 2C. Or you can have completely depressed ulcerated lesion is C. So that looks good. You can see here that's bedonculated. You can call that maybe 1P. If it is a little bit sessile, you can call it 1S. If you have slight elevation like this one, that would be 2A. That's clearly depressed lesion. You can see like it's so depressed. That will be type 3 was the exception of. It has also raised area. So if you have ulceration without raised margin, this classification would work very well. But what happened in a lesion like that? We have really depressed area and a raised area. It's actually a combination of both. So then you will get to some of the complex nemaculture, which is 2A, 2C, or 2C, 2A. I'm just showing you that you can see these things. The idea is anytime you have a depression, it is good. Anytime you have a depression, it's not good. It means like you have submucosa invasion most of the time. But also you have to be careful about how the depression looks like. If it's so deep that it's going to submucosa and have a heap of mucosa like exactly this lesion here, that could be Paris 2C, 2A. But sometimes you have slightly depressed area within raised area. And when you see that, that's not as high risk at that lesion. So how deep it is is very important. So from that Paris classification original study, they discovered that anything that has Paris A2 or Paris 2C, the high risk of lymphovascular invasion is 61%. It's very high. So then what's in the conclusion here? If you see that, do not resect it. Anything looks like that is bad. So remember this picture. Remember this image. That's what we don't want to resect. Then let's talk about totally different type of classification. In the colon, the polyps, unlike the esophagus and the stomach, they can grow in a funky way. We call them granular, laterally spreading tumor, which led us to think about these polyps as a different thing, LST, laterally spreading. It's almost like when you have your backyard and have the weeds growing everywhere, spreading on the ground. They don't grow like a tree. They just grow, spread everywhere. That's LST. And LST would be either granular or non-granular. But it's even more complex than that. You can have component of it granular and another component not granular. So in this picture, A, that's clearly granular. That's also granular. This one here is flat or soft. And this one here is also soft. It's not depressed, but it's non-granular, and it's spreading over a large area. You can see here the non-granular type of polyp have the highest risk of submucosa invasion. In this study, which was the landmark study, it was around 15% in the non-granular. And what does that tell us? Yes, you can remove this, but you cannot remove that. Or if you're going to remove that, you better remove it with ESD, not EMR. So within the granular type, you can see that you can have homogenous granular. Sometime you'll have a lesion. All the granulars looks the same. And sometime, like this one, heterogeneous with some large nodules. And we'll find that there's certain feature. If you have a nodule larger than 10 millimeter or large tumors in general, this one have high risk of submucosa invasion. But you still can at least remove them by ESD before deciding and sending them to surgery. With advance of imaging, we start to have the NBI, iSCAN, and Fuji also have their own system. And each system can have some sort of optical imaging filters that can highlight the lesion. And the first one came with a nice classification. And I don't want you to feel bored or annoyed by all this classification. Just take a picture of them and put them in your room when you do endoscopy. And you can look at them every day and do it. I do that. I don't remember any of them. I just print them, and I put them there. And if you keep using them every day, you will know how to use them. So in this classification, you can tell you is this hyperplastic adenoma or submucosa invasion. You can see here is a difference between the three polyps. This pattern here looks like more wider in color if we get closer to it. This one has these tubules, looks like honeycomb a little bit. And here you have this amorphous material. What's the problem in this nice classification that a lot of people don't like? You jump from adenoma into submucosa invasion without having the stage of high-grade dysplasia. So if you are using the nice classification, you end up having lesion. You don't know. They are not tubular adenoma, but they are not submucosa invasion. They are the in-between, and you don't have anything to help you with the in-between. So the Japanese met, and they decided to say, we're going to have our own Japan NBI expert team to address the limitation of the nice classification. And then they came with something called GNET. If you want to hang something in your room, put this one. I think it is a little bit more comprehensive. What they did is that they divided type 2 into two types, and type 2 is the adenoma. So you can see here type 1, we're talking about hyperplastic. And the hyperplastic is basically white spots. It is dark white spots, and it's very regular, and it looks like dots. So hyperplastic polyps are dots. But if you have a large area of dots in the right side of the colon, this is the salicylated adenoma. You still have to remove it, but you can know it is SSA. Type 2, again, it is the same tubular pattern, and you can see the tubular pattern. Before it goes to type 3, it can go into that stage, which is the high grade. And what happens is that you lose this uniformity, it becomes more irregular. The blood vessels become more tortuous, and darker, and thicker. This looks good in theory, but still some of the limitation is that differentiating type 2 from 3 could be hard. Once you go to type 3, you have lots of vascular pattern. That's one is bad. This one has deep submucosal invasion. This one has to go for surgery. So when you see that type, you send this patient for surgery. There's no need to do endoscopy. And most likely, you would biopsy this area, and you'll get your diagnosis. So now we talked about morphology and predicting lymphovascular invasion. You should also know that when we take a biopsy and remove a polyp, you might be in the same situation of knowing, should I remove this lesion, send it to surgery or not? Because you might remove it, and you get surprised by what's in the pathology. So that classification is a very nice one called Haggett classification. And it's a histological classification, and it talks about mainly bedonculated polyp. Because a lot of us see them, they're large bedonculated polyp, they're easy to remove, but then you might end up with cancer in them. And then you go to the tumor board, and they ask you how far you should go, and what's the risk of lymphoid invasion. You should know this, if the level 1 is a very slight risk, if you go all the way to level 4, which means the adenocarcinoma going to submucosa, level 4 should go for surgery, even if you have curative resection. We measure the stack, we measure how far the tumor is from the stack and how many millimeters. But there's always risk of lymphovascular invasion and lymph node metastasis that you have to discuss with your patient. And those discussions are not easy, because you can tell them, I completely removed the polyp, but you have 5% to 10% risk that you might have tumor recurrence, and then I would like you to remove this part of your colon with all the lymph node, just in case, and they go for it, and they come back negative. So why is it exposed into the surgery? Because we don't know, because we don't have a mechanism yet to know if there is a tumor in the lymph node. So you have to have this discussion with them, that I will send you to surgery, but most likely can come back negative. So SM classification, we divide the submucosa into three parts, SM1, 2, and 3, and the deeper the lesion in the submucosa, the higher risk of lymphovascular invasion, and you can see here it reached 35% in SM3. How did we know that? Well, we knew that when patients were sent for surgery, and they removed the entire colon, and they looked at how deep the tumor is, and how much of the tumor in the lymph node, and they came up with this. So how can I do that by endoscopy? Well, most likely if you're doing EMR, you're not going to be able to get this data. You have to be able to do ESD, and measuring how deep the tumor is in the submucosa, and you have to have a good pathologist who is even familiar with measuring the submucosa and ESD specimen, which is another story. So in general, you can remember these two numbers for submucosa invasion, and it will be like most likely not happen in benonculated lesion if it is less than 3,000 micromillimeter, but in non-benonculated lesion, it is 1,000. So remember in the stomach is 500, and the colon is 1,000. It also depends on the location of the colon. There's many studies showing that the distal you come, the higher risk of lymph node invasion. So this rectal lesion are actually the highest one to have lymphovascular metastasis compared to right side colonic lesion. So having all of this in your mind when you assist a patient is very important. So in general, what are the predictor of lymph node metastasis? And this will be something in pathology. So let's say you find a lesion, you're not sure if you're going to send it for endoscopy or for surgery, you biopsy it, and then it come back with something like mucinous adenocarcinoma or poorly differentiated adenocarcinoma, and I get this a lot. Can you do ESD in poorly differentiated adenocarcinoma? No, I'm not going to do it because even if I'm able to remove the tumor, there's a higher risk of lymphovascular metastasis. So that's a patient will go for surgery. Anytime you have in the biopsy lymphovascular invasion. So let's say you did an EMR and the EMR showed early tumor but it showed lymphovascular LVI. If you have a venous metastasis, if you have tumor budding, if you have mucinous adenocarcinoma, all this lesion should be sent for surgery. So we're going to go back to the U.S. multi-society task force and I'll take you back to the part of where they talk about the management. They talked about if you suspect that there's no invasion, then you're good. Smaller lesion, you can do cold or hot snare pulpectomy. Larger one, you can do EMR. But if you suspect submucosal invasion, you have to think about one or two things. Is this minimal or deep? Deep like this ulcerated type 3 one we saw before that you have no structure. This one has to go, confirm with biopsy, go for surgery. But if you have early submucosal invasion, did you see what the recommendation said here? EMR or ESD of complete resection is feasible in phase. What does it mean? I don't know what it means. There's no clear distinction here. We do not have guidelines in the United States to differentiate between EMR and ESD. But this will change very soon because the ASGE finally will have one for the esophageal and gastric. Hopefully, you're going to have one for colonic. But the reality is in the United States, there's no guidelines tell you when you do EMR or when you do ESD. So you'll have to go to other sources. And one of them, this is a proposed criteria published in 2015. It's published in Red Journal. And basically, it was talking about look at tumor size and location. And if you have non-granular LSD and they are 2 to 3 centimeters, just do ESD. For granular one, we know they are benign. You are good with EMR unless you get to more than 3 centimeters. If you have residual tumor and they are large, do ESD. If you have rectal carcinoid, obviously, if it is more than 1 centimeter, you should do ESD. But this is not a guideline. This was a proposed criteria by the authors. Nobody in the United States worked on validating this criteria yet, and I think we should. But the Japanese, they do have criteria for conic ESD. And I would like to go through it. So the first one in it is in-block resection is not feasible with EMR. So this is a mindset that is completely different than our mindset. What they're saying here, EMR should be done only for in-block resection. It's so hard in the United States to convince people to do everything in-block. So for Japanese, they say if you are not able to do the lesion in-block, you should send it for ESD. Will we be able to reach that at a certain point in time? Maybe when we train enough people. I don't see that for the next 10 years. So this criteria cannot apply here in the United States. And it doesn't make sense, because a lot of these patients have benign lesions, and we should not advocate for ESD in every clonic polyp. Because a lot of these granular polyps are benign, and doing EMR would be more than enough. So I feel like this criteria is a little bit strict. Even in my, like, I love ESD, I still think it's too much. So what other criteria do they have? Non-granular LSD, they really think that all of them should be done by ESD. Codobaron type 4. And we did not talk about codobaron. The code of classification is using chromoendoscopy. And a lot of time now, people use virtual chromoendoscopy. So the GNET and NICE classification is better. Anytime you suspect SMV invasion, anytime you have a depressed type of tumor or lesion, or if you suspect submucosal fibrosis. Or if you're removing a polyp in the setting of ulcerative colitis, or if you have recurrence of early cancer. So these are their proposed criteria. I think it is good, except the first one, you know, I feel it's a little bit too much. So what the European indication, the European Society, believe it or not, they did have guidelines since 2015 about chronic ESD. And they recommended it for depressed morphology, irregular or non-granular surface pattern, especially if the lesion is more than 2 centimeters. So what we are getting here? It is more into anytime you suspect early submucosal invasion, anytime you have a slightly depressed lesion, and anytime the lesion is a little bit larger than usual, maybe you should do ESD and not EMR. But if you see like frank depression with loss of vascular margin, this really deep submucosal invasion should send patient for surgery. So this is some example of patient I did ESD on in my practice. And that's like lateral spreading tumor. You can see here multiple nodule. One of them is very dominant large nodule. This is another tumor. We're going to discuss it right now. This is a fibrotic lesion. This is appendiceal polyp. This lesion is ulcerative colitis, as you can see. So we do see this patient in the United States, and we can, most of our patients will benefit from this. So before I move on, and I have 11 minutes, I'm going to show you this lesion, and you can tell me, can you predict what the pathology would be? So first in LST, how will you guys classify this? So this lesion has two components, that part and that part. So it's mixed, mixed granular and non-granular. Actually the front part is non-granular or have slight granular, and then the last part is more granular. So this lesion could be mixed. This is how it looked at the end. Did you see any depressed area there? I don't see any depression, but there is, if I stop the video here, I don't like this area. This soft, flat area. It's not, you know, that's what we call non-granular. So you have big granular area and non-granular area, and we saw earlier, if you have a mixed type, it's around 3% to 5% risk of lymphovascular metastasis. So you want to see what the pathology was? Okay. So it's left descending colon, ESD, tubular villus adenoma, high-grade dysplasia, cauterized debridal margin negative for cancer. So that's one. What about this one here? Let's look at this one. Tell me, what's your impression, first impression about looking at a lesion like that? What do you think? Anyone can say your opinion. This is true. This looks bad. So why do we think it is bad? What's in this lesion makes us... Yeah, it has multiple depressed area. It is bulky. It's very bulky and large. It's like four or five centimeters, as you can see. And it's not really slight. It is just like very bulk lesion over large area, but you can still see some adenoma, right? This is the tubular adenomatous pattern, but the more you go to the middle here, you lose it. So it looks like there's something here in the middle. So let's see how... You can even see the extent of how large this area. This was like two-thirds circumference. That's a very large lesion. It's not a small area. All right. Do you want to know what this was? All right. This was well-differentiated adenocarcinoma. So it's going in the submucosa into 108 micron. But clear margin and deep margin is also clear. Can you see how much clear margin you have? So that's a cautery artifact here. That's a submucosa. That's where the tumor is going into the submucosa. And that's a clear margin more than 1,000 micromillimeter. That's when ESD become advantageous, because really, if you can get so deep in submucosa that you can measure how far you are, then you can tell this patient, and I did this case two years ago. He still did not have any recurrence. You can tell a patient like that, that if you have your tumor only 108 micron into submucosa, that's very superficial. The risk of lymphovascular invasion is less maybe than 2%. It is up to you. But I think we had a curative resection, but people take risk differently. And he chose not to have a surgery. All right. So you cannot do ESD right away. You know, I want to say that ESD is also hard, and you have to predict how hard ESD is, because this other factor. So when we talk about doing this and doing that, we also have to know, is it easy to do? So this was a nice study, and it shows it, because it's highlight. It can help you know how hard ESD is. So if you have any tumor larger than 3 centimeter, and if you have anything, the circumference more than 5 centimeter, and it's a 2 third of the circumference, sorry, or tumor larger in size, or tumor in the flexure, like hepatic flexures is harder. And it give all of these points, as you can see here. So one point for everything. So non-granular, one point. Anything close to the intake line, one point. Anything more than 2 third, two point. You're going to get a score how easy and how very difficult it is. So if you collect four, then it become hard. So if you have a lesion larger than 5 centimeter, which is around the hepatic flexure, it's most likely going to be very difficult ESD. You don't want to start doing that after getting training. This is not a good idea. So then knowing what to choose, because what I'm finding, one of the biggest problem of adopting ESD in United States is the lesion selection. So I train people, and the first thing they decide to do is large right side chronic lesion that they spend eight hour on, and the patient end up going to the OR. And then they end up having a negative experience and don't want to try it again. And then we say the technique is unsafe. So we have to change how we think about this. We have to start first with easier lesion. And that's the ESG, European Society recommendation. They recommend, again, it's unsupervised ESD. No matter how much you are an expert or so good at what you're doing, you cannot just bring a knife and start doing it, or take one course and do it. So at least have somebody to help you at the beginning. You should go into a curriculum and learn what you should learn, things like we're doing right now. That's a great curriculum that we're having today. And later on, once you learn and work in animal model, you can start with rectal lesion. Lesion that have, you should know that you shouldn't start in cases which has submucosal fibrosis. They are harder. Start with a smaller lesion, and do not do chronic ESD until you have experience in other cases. But that does not apply in United States. In United States, you have to start with rectal cases from the beginning, because that's the only way to train our providers, is to start with rectal ESD. So when to consider surgical resection? And that's a review, again, polydifferentiated tumor, syndrome carcinoma, mucinous carcinoma, deep submucosal invasion, any time you have LVI, or tumor budding. And that came from the Korean guidelines. So if you told me what to send for surgery, that's what I sent for surgery. But wait a second. I did not talk about location of the tumor. Where is IC valve and appendiceal polyps? Well, in the past, this used to go for surgery, too. And I would say some of this lesion should go for surgery. But I want to propose for you now, paradigm shift, that even this lesion now can be removed by endoscopy, but not by, like, I wouldn't say universal. So this is a lesion that's totally involving the appendix, right? We can do FTR, but I have to be honest with you. I don't like to do FTR, because I cannot control the margin. And when you start doing ESD, you feel like you don't want to be involved in techniques that does not guarantee R0 resection. So I would rather go around the lesion and define my margin before I do the resection. So you can see here, you will have a problem with lesions going into the appendix that sometime you have to remove the entire appendix. So after I had my appendectomy done by surgery, I decided maybe you can do that by endoscopy. So I kept watching a lot of their videos about how they do appendectomies. And so far, we had five transcecal appendectomy cases. And how you do it, you dissect into the mesoappendix, and you dissect the appendix out of the mesoappendix. But you need some traction device, and some of them you are going to be trained on today. So I don't want to talk much about these type of lesions. But I just want to say that it is possible that you can remove the appendix by endoscopy. I know another group in China that they're doing that routinely for a mucinous tumor. And you can see here, you can cut it from the base and close it. And most of these patients, they do very well. But I would not propose that this would be something we do every day. I'm just saying that endoscopic resection is advancing. And even tough locations, such as appendix and IC valve, we can really do the procedure there with no problem. You can see here the entire appendix out. You also have to learn in appendix that they have different length. Sometimes knowing the CT scan size is important, because like two weeks ago, I had to remove five-centimeter appendix. They are long. So you have to keep going with them in this appendix and know how long they are to get them out. So it's not like only a portion of the appendix. This is our experience in the appendiceal cases and also IC valve. So this type of lesion, honestly in the past, until now, if you want to send them to surgery, if you have a circumferential involvement of IC valve, you can send this patient to surgery. But you'll find also from the Japanese literature, a lot of data about taking care of this patient endoscopically and also in Europe. So with ESD, you can do dissect this lesion too, and you can dissect them from IC valve. But maybe it also depends on patient and case by case. If there is a patient who's younger and he has other polyps on the right side of the colon and then morbidly obese, yes, I will send them to surgery. But what I want to say and propose is that location of the polyp in the GI tract should not be a contraindication anymore. It might be a relative contraindication, but it's not a contraindication. I want to highlight here, if ever you work an IC valve, I suggest you do this technique by getting the mucosa of the terminal allium outside into the cecum and open it up. The reason is, if you do circumferential dissection of terminal allium, it may close up. And this patient will have a stretcher. And sometimes they only present with intestine obstruction, and they can go to emergent surgery. I had one patient like that, and I learned my lesson. After that, that any time I'd work in the terminal allium, I have to clip it to outside to the cecum to keep the IC valve open. So that's it. I think I finished just one minute ahead of time. Thank you. Can I make a comment? That was great. I love all the classification stuff. So one thing I tell my fellows to do is to put in the note, commit to what you think it's going to be, and test yourself. So if you think it's a success alterative, I don't know, but if you think it's got something, put it in the notes. And when you look at the path, you're, I don't know, you guys are busy. Did I think it was going to be this or that? And so I don't know what your thoughts are. I just find it helpful. Make a commitment to test yourself, to see if your eyes are seeing things that you're not. But what I'm noticing is that we are not applying the NPI classification to polyps, and a lot of people will impart adjustment with a polyp without documenting the NPI. So it would be a good habit, especially for fellows, to do that. And you'll be surprised. You'll be teaching your other attending. When you see you're doing it, you'll start doing it too. But I think even your images that are white light, I mean, you can, you know, we now have these monitors on booms that are right next to us. We have high-def scopes, high-def screens. So I think even without NPI, a lot of your pictures, you can clearly tell. So I think test yourself, and I don't know, I think it's helpful if it's in the note, because then, you know, you're reviewing path, you do so many colonoscopies, and I, you know, I was like, or even size-wise, when I first started doing big polyps, I was always like, was that really five centimeters? I mean, not that the path is going to completely correlate, but I think it's useful so that as you're doing it, you, you know, you learn from your own doing. So that's just my suggestion. Nice talk.

resect polyps

lymph nodes

Paris classification

GI lesions

lymphovascular invasion

EMR procedures

endoscopic removal