I'm going to ask Amit Bhatt, who's at the Cleveland Clinic. We've been talking ESD, and he's one of the innovators in thinking about where this goes in the future. So Amit will start us off. The idea here is stop, ask questions. This is more of a discussion. Good morning, and thank you very much, probably, to Amitabh for having me come here and present this case. So unlike most ESD lectures, this one's actually not on technique and performing the procedure, but rather on the decision-making before and after resection. So this is a case of a 79-year-old man who was incidentally diagnosed with distal esophageal cancer after he underwent an upper endoscopy for a persistent cough. He has coronary artery disease, had a cabbage, a current heavy smoker with COPD, and he has moderate aortic stenosis. He was actually staged and then sent to our cardiothoracic surgeons. So the endo picture is on the left-hand side and shows a ulcerated distal esophageal tumor. An ultrasound was performed, and the person who performed the ultrasound staged the tumor as T1BN1, and you can see the lymph node on the top of that EOS screen. And they also had a PET scan where the tumor lit up brightly on the PET scan. Fortunately, he's not a good candidate, and our surgeons come to us and ask us, is there a role for endoscopic resection in this patient? So go back to that picture. You guys, when you see that picture, what do you think about endoscopic, if you were looking at just the picture endoscopic? What features are there that would make you... Use your mic, yeah. It's ulcerated, irregular borders, and depressed. All features that we're told not to do anything endoscopically, because it's likely submucosally invasive. And there's a node on ultrasound, so... And okay, so there's a node on ultrasound. How much weight, Amitabh, do you give to the findings of the PET scan and the ultrasound in this situation? Yeah, so I'm going to bring up that the node's not lighting up. We rely on ultrasound characteristics, but we know that they're not totally accurate. And Amit's probably going to point out that maybe some of our endoscopic criteria are not 100%. So if you had a FNA, maybe, but FNA is also challenging, because you go through the cancer sometimes to do FNA. So I don't think it can be 100%, but yeah, I'd worry about that lymph node. So Amit or Amitabh, do you want to talk to the trainees? What are the characteristics you look for, a metastatic lymph node, and how accurate is that in the studies? That's a gestalt of what do you... You guys have started doing U.S. What are the criterias of this lymph node that are bothersome? Apocrylic, round-shaped, well-defined, and associated with the mass. Those are all sort of worrisome features. How accurate is that, is what Prevlin's asked. About 80% or so from old—I mean, we don't even get the series anymore because everyone gets neoadjuvant chemotherapy, so who knows what our accuracy is. About 80% maybe. So Amitabh, were you part of those original studies? I know it was Tom Rice and Zuccaro who did the criteria for sort of positive malignant lymph nodes and esophageal cancer, but it's been a very long time since they've been updated. Yeah, because we don't have the ability to update it. And remember, they did not correlate histology with the note they saw. No one's ever done that. So I think this is great. It sort of brings up some of the flaws here with the information that we have coming in before resection. Amit, I have a question. Sure. So the stage 1B, like how comfortable are you? Was this your EUS? This was not my EUS. Do you really think—I mean, looking at that thing, does that look—does that really look like a 1? You know, I just—I think, you know, there's a lot of—you know, when you look at something, I think we're—I bet we're pretty good at coming up with whether—I mean, if someone sent that to me for endoscopic resection, I'd be a little worried. And to give that 1P piece of information, the lymph node's on the opposite wall of the tumor, so—which means you are able to pass a needle through the esophagus. Mohamed, what do you think? How would you address this? I would say the lymph node is large enough and I'm not going through the tumor. It might be worthwhile of an FNA. If it's close to 1 centimeter, I'd probably do an FNA on it. And of course, I think you're going to come to the conclusion this is a multidisciplinary tumor board review. So— I'd send it to Dr. Bhatt. So— So— Amit, I have a question. Sure. Can we see that lesion in a live form when your peristaltic wave is going through that? And I think that is a key part, actually. This picture gives you the idea of an ulcerated depressed lesion, and we'll see it later on to show how it looks when it actually comes out. It's actually not a depressed lesion. It was just that moment that it was captured like this. So Amit, the other comment I just wanted to make for the fellows, which I always tell my fellow is, so we're all talking about should we FNA this node? So here's a patient who's not surgical who has a node where it's obviously going to be important if that node is positive because are you going to really endoscopically resect something if there's nodal involvement? Will that really actually change the patient's outcome? Whereas if it's someone who's surgical, then it maybe matters less because there's a lot of patients getting neoadjuvant therapy instead of just going straight to surgery. So it may not change what they do. So the important thing when you're taking care of patients is there's a lot of things we can do. The question is, should we do them? So you have to always say with everything you do, is this going to change ultimately what happens to the patient? Because if it doesn't, you shouldn't be monkeying around doing things that aren't going to actually change the patient's management. So I think the big learning thing, because I mean, our fellows, they always want to do the ERCP no matter what, right, because you're the fellow. You're not the attending. But you've got to remember that soon you're going to be responsible for all the bad things that happen when you do stuff. So just make sure that you're doing stuff for a good reason. So that's just my thought. And remember, the first endoscopy is done, probably the patient's not on PPI, because the indication is something very different. And that can fool you if you have inflammation from the reflux and you quiet it down. This lesion may look different after PPI treatment, just. Very astute. I'm on top. So what information is available to us prior to resection of esophageal cancer? And how good is that information? Can we depend upon it? So to understand this, you need to understand some basic principles of endoscopic resection of early cancer. So the main difference between endoscopic and surgical resection of a tumor is the absence of lymph node dissection with endoscopic technique. So endoscopic resection should only be considered in lesions that have a negligible risk of traveling to lymph nodes, or that that is less than its counterpart surgery for mortality and morbidity. So to give you an idea, if we're talking about esophageal cancer, the 30-day mortality after an esophagectomy is 4.2%. That means 4 out of 100 patients will be dead from the surgery alone. And the overall morbidity is close to 50%. People undergoing respiratory failures, leak, and sepsis. So you can see for many elderly patients with comorbid disease, a large interthoracic surgery might not be the best thing for them. So how do we define high-risk features? So what features do you see in an early tumor that increase the risk of it going to a lymph node? So that would be depth of invasion, if it was a poorly differentiated tumor, and that it showed lymphovascular invasion. So which one of these are available to us prior to resection? Schaefer? Theoretically, all of them, depending upon your pathology. So let's start off with pathology. So differentiation, most biopsies you'll see come back with adenocarcinoma. We'll say at least intermucosal adenocarcinoma. And the pathologist will not commit to a differentiation on preoperative biopsies. Sometimes you will see poorly differentiated pathology. But as differentiation is based on the percentage, and there's only a small amount of tissue in biopsies, it's unclear how accurate this is. And actually, me and Amitav are studying this right now, and it doesn't look too good. Lymphovascular invasion is generally not reported on endoscopic biopsies. So there's actually very little information you can gather before resection from the biopsy itself. So what about depth of invasion? An argument we see at our tumor board a lot of, please. Before you make a decision about the resection in these patients, how do you biopsy and where do you biopsy if you anticipate that you'll be bringing them back for resection? Does that make a difference? Yeah, absolutely. So I think we're always concerned in the world of endoscopic resection is these patients that come across, they've had five to seven biopsies. And some of these biopsy sites can induce submucosal fibrosis and make subsequent resection or even endoscopic diagnosis hard. So we generally want people to take a biopsy from the highest risk area that they see or the most advanced area they suspect, and just one single biopsy. So an argument we see at tumor board very commonly is, oh, the tumor lit up on PET scan. This must be an advanced tumor. Or the tumor didn't light up on PET scan, so it's not an advanced tumor. So we actually studied this, and we found that both T1A and T1B tumors can be PET positive and PET negative, and that the mean SUV did not help differentiate from one tumor or another. So this is not a very useful tool in helping us with depth of invasion. What about endoscopic ultrasound? Form it a lot. It's very good at differentiating T2 from T3 tumors, but in our recent multicenter study that we published in GIE, only 50% of T1B tumors were actually staged as T1B prior to resection. The majority of these were staged as T1A. So EOS ability to differentiate T1A from T1B is almost a flip of a coin. We don't have an LVI. Tumor differentiation is questionable. Depth of invasion between T1A and T1B is not very clear. And that really brings us to the mainstay of esophageal cancer endoscopic therapy is we perform staging endoscopic. Amit, let me interrupt you. Regarding the, you said T1A and B, and you're talking about a standard EOS or a mini probe? So we're talking about standard radial EOS, which is the most common form of EOS that is performed in the United States. In Japan, they do use mini probe EOS, especially underwater, and they may be able to get more information. But we don't have really data on esophageal adenocarcinoma and staging with mini probe unless you're aware of some. I'm not aware of data. I do this because I do mini probes, and I know that when you use those in this setting, you'll probably get a better differentiation between these type of superficial tumors, especially that the frequency used is quite high. But yes, I'm aware there is no solid data to support that. But they're available even in the US to my knowledge. Oh, I think it's a great idea. Actually, we're trying to, we have a very old mini probe, and we're trying to update that for that exact reason, to help with this. So this patient underwent FNA with a lymph node, and it returned benign. And we did endoscopic submucosal tunneling technique that allows us to get underneath the tumor into the deep submucosa and remove the tumor in one complete piece. And as Raju was pointing out before, this is not actually a depressed lesion. This is actually more of a sessile lesion. And as you can see, it only has superficial ulceration. And we do put these patients on PPI beforehand to help that heal. So the pathology is now returned. It was actually just a T1A tumor. It was not invading into the submucosa. All the margins are negative. It's poorly differentiated pathology. The D240 staining comes back LVI positive, but it's completely resected. What's the next steps now? What's that? It's important. Any thoughts? The rate of recurrence is very high. OK. So I think this is becoming a much more common thing that we're being, we're now being asked to cut out a lot more marginal tumors, T1B tumors, and the real question is when you come back and you have this pathology, just to begin with, is this curative or non-curative pathology? Non-curative. So the non-curative factors are the poorly differentiated and the LVI. So those are high-risk features for lymph node metastasis. So you may want to expand what LVI is for us. Sure. So lymphovascular invasion, so invasion seen off the tumor into the little lymphatics and vascular area in the semicosal. So we know it's non-curative, it's high-risk, and we're doing more and more of these resections. So the question is what to do next. I think somebody who's a surgical candidate, these patients should go on to surgery. But I do wonder, at your institution, what are you doing with non-curative pathology? I agree with you. We see quite a bit of that. And I, at this point, I would commit this patient to endoscopic surveillance. I would not, not a surgical candidate, obviously. We discussed the tumor board. Our surgeons will tell us, will not resect. You continue to watch, bring them back to us when you need us on board. Radiation, chemo? No, not, I have not seen this pickup traction at all, especially in very borderline functional status like what we saw. I think this very much goes for multidisciplinary tumor boards. So your local practice may vary depending on what's available locally, even amongst cancer centers. What trials and stuff are available will vary from institution to institution. So for us, at MD Anderson, we have, they might get adjuvant chemotherapy or radiation therapy. There are trials now on immunotherapy as well. I've had more than one patient who have gotten adjuvant immunotherapy for stuff like this. To Amit's point, actually, something you alluded to, there's actually a difference as well between endoscopic surveillance and oncologic surveillance. So for a case like this, the major risk is oncologic surveillance, because the major risk is not necessarily local recurrence, but metastatic recurrence. Because if you have a lesion taken beautifully like this with clean margins, a clean endoscopic cut, the next time it shows up, it's probably not going to be at that location. The next time it shows up, it's going to be that regional lymph node along the same lymphatic chain, a skipped lesion further up the esophagus, or a distant node like a periaortic node, or the liver. So it's actually very important to have tumor board discussion, but also oncologic surveillance for follow-up for something like this. That's an excellent point. And a lot of the surveillance for these patients isn't actually to look for recurrence at the site of resection. It's actually to look at recurrence in the lymph nodes or metastatic disease. So a lot of these patients for us are EUS plus CAT scan surveillance moving forward. Amitabh, are you giving these guys chemo or radiation? Yeah. Tumor board depends on who gets it, what flavor of the day. So we're moving towards that as well. So this patient received adjuvant radiation, but we're also having patients who get adjuvant immunotherapy and chemotherapy as well. So this patient actually did well. No recurrence for two years. Are we done now, or is there any other important steps that need to be done? So what I've learned is one of the most important things in the treatment of esophageal cancer is actually RFA ablation of the residual Barrett segment. We've had a number of patients who's had their tumors resected now. Their RFA ablation was disrupted due to COVID. And fast forward three years, they're now coming in with metachromis cancers. So resecting the primary lesion, the job is not done. These patients need to come back, get RFA ablation until they achieve CRIM or CRD so they don't develop any further esophageal cancers. Thank you very much for your kind attention. Oh. Okay.

distal esophageal cancer

upper endoscopy

PET scan

endoscopic resection

lymph node

personalized treatment