Hi, everyone. So we will begin our last session of the afternoon. Thank you for hanging in there with us. So the next talk is going to be on eradication of Barrett's once and for all, EMR, EST, RFA and cryotherapy. Okay, so again, same disclosures. And just an outline of my talk, we'll go over components of endoscopic therapy. I think in the last few talks, we've talked about how to screen, how to survey, how to find dysplasia. And now we'll talk about how to deal and treat dysplasia and early cancer. So we'll talk about resection, why and how, CAP-assisted EMR versus ESD, some data on comparative outcomes, talk about ablation, rationale and methods, and then radiofrequency ablation, cryotherapy, newer methods, briefly touch on outcomes, and then also end with recurrence, which is a problem we cannot ignore and we should be aware of. So endotherapy of Barrett's neoplasia, as you have just heard, consists of several components. So the first is detection and staging of neoplasia, followed by what I would call tissue acquisition, which is endoscopic resection, followed by elimination of intestinal metaplasia, which we achieve by ablation. And lastly, post-ablation surveillance to detect recurrence and hopefully treat it. And in the interest of time, I'm going to deal with the last three components here. So the primary goal of endoscopic therapy of Barrett's related neoplasia and dysplasia is is cancer prevention, which is accomplished by the removal and destruction of metaplastic and neoplastic epithelium, controlling the acid reflux, which leads to squamous re-epithelialization, leading to reduction of cancer risk. We can pair this with cancer treatment, which we accomplish with accurate staging with resection techniques, followed by ablation, which leads to squamous re-epithelialization, and then the reduction of cancer recurrence. So we achieve this with resection techniques, which provide us tissue for histology, allow accurate staging, and allow us, most importantly, to choose the right treatment for the right patient. We then follow this up with ablation techniques, which cause tissue necrosis, either with thermal technologies like radiofrequency ablation or cryotherapy, with liquid nitrogen, nitrous oxide. And all of this is predicated on identifying neoplasia and dysplasia in Barrett's. And as you have heard many times over today, this tends to be a challenge, because dysplasia or neoplasia in Barrett's is usually subtle. I have several pictures in front of you. These are not subtle picture. These are not subtle findings, which hopefully none of us are going to miss. Typically, this is how dysplasia and neoplasia appears in Barrett's, where it's relatively flat, subtle, requires us to have knowledge of how abnormal mucosa looks like, use high definition and high resolution endoscopy, use virtual chromoendoscopy, or dye-based chromoendoscopy. So once we see and identify these lesions, our goal is to endoscopically resect them. The first question we need to ask is why. The most logical answer is that when we resect instead of biopsy lesions, we achieve accurate histologic staging. Several studies have shown that if you do an endoscopic resection, you upstage histology in over 30% of cases that are referred to you. Pathologists also, not surprisingly, have less inter-observer variation when they are looking at larger chunks of tissue as opposed to biopsy pieces. And the bottom line is that it allows us accurate staging of T1 cancers. And we know from several studies that endoscopic resection is more accurate than endoscopic ultrasound in the T staging of esophageal, early esophageal cancer, particularly as we divide the T1 cancer into T1a, which is confined to the mucosa, and T1b, where there is invasion into the submucosa, because our treatment options for these two subtly different, hard to distinguish stages of early esophageal cancer are so radically different. We suggest endotherapy for T1a cancer, whereas we suggest esophagectomy for T1b cancer. Additionally, endoscopic resection specimens also provide us with prognostic information. And I'll go over this in a bit more detail. They provide us grade of differentiation. They provide us with the information on lymphovascular invasion. They allow us to evaluate our margins, particularly the deep margins. And of course, if our margins are negative, they also allow us to be therapeutic with the endoscopic resection. So the basic rationale is that we want to ensure appropriate treatment, right treatment for the right patient. So in that regard, when we see patients who have an early esophageal cancer arising in barretts, and how do we know a cancer is early? Typically, these cancers are, quote unquote, relatively asymptomatic. These patients are not presenting with dysphagia, change in appetite, change in weight. They are typically picked up either on surveillance, or they've had an endoscopy for some other reason, and the cancer is really stumbled upon. And when you see a patient with a T1 esophageal adenocarcinoma, my first step is typically to do an endoscopic ultrasound. Now, I just told you that endoscopic ultrasound is not great at distinguishing between T1A versus T1B. So I really don't do the endoscopic ultrasound to differentiate between these two, but really to exclude muscularis propria invasion, which you can do fairly well with endoscopic ultrasound, and more importantly, to exclude metastatic lymphadenopathy, because EUS can do that fairly well. If I see metastatic lymphadenopathy, then I know endoscopic therapy is not an option, and I stop. On the other hand, if we see this, which I would call a lift sign, a good lift sign, and I see that the tumor is not invading the muscularis propria, then I know I'm likely dealing with an early esophageal cancer, and then I can proceed to endoscopic resection. Another reason we want to remove all nodular lesions is that we know ablative techniques like radiofrequency ablation are really superficially, they affect only the superficial layers of the mucosa. So Dr. Thotta showed you the layers of the esophagus, and radiofrequency ablation only goes down into the lamina propria. So if you have nodular areas that you miss, and you ablate, you're going to get superficial response with likely a possibility of deeper untreated advanced neoplasia, which can lead to subsquamous barrets or cancer. And of course, if you miss a cancer that is a T1B cancer, should have been sent for esophagectomy, then you are really choosing the inappropriate treatment for the patient. So this is another reason why we want to be resecting any visible lesions. So we talked about endoscopic resection histology giving you staging and prognostic information. So this is a high part, actually a low part view of an endoscopic resection specimen, and you have the mucosa here, you have the muscularis mucosa, and then you have the submucosal glands and fat. And on quick inspection, you might think, okay, the cancer is only in the mucosa, in the epithelium, and the lymphadenopathy risk is less than 2%. But when we look closer, we see that this cancer is poorly differentiated, and it actually has signet cell features, which is telling us that this is a more aggressive poorly differentiated cancer. And when you look even further, you see that there is tumor in the lymphatic vessels and in the blood vessels. So now you have lymph node, lymphovascular invasion, which increases the risk of metastatic lymphadenopathy. So again, you've changed your perspective of how you would view this cancer. This is another patient who has obvious tumor in the mucosa, this is a muscularis mucosa, and this is the submucosa. And again, you are seeing cancer along with desmoplasia in the submucosa, which tells you this is T1B disease with a likely 20 to 25% risk of metastatic lymphadenopathy. So we have understood how endoscopic, why endoscopic resection can be performed. Let's turn our attention to how. We can do this in one of two ways. We can do cap-assisted endoscopic resection, wherein you identify the lesion, you inject into the submucosa, and you excise the lesion. And you can do this with a couple of ways using cap-assisted endoscopic resection, either using a crescent snare and a hard cap. And this requires one resection at a time with each snare being used only once. Or you can do this with band ligation, where you can use devices which are very similar to band ligation devices and perform multiple resections and achieve a wide area of clearance. A newer modality is that of endoscopic submucosal dissection. And this is theoretically similar to cap-assisted endoscopic resection in the sense that you identify your lesion. This is a Paris type 2A, 2B kind of a lesion that Dr. Thotta had discussed. You mark the borders of the lesion. You inject at the margins. And in addition, in distinction from cap-assisted endoscopic resection, using a variety of knives, you make an incision, and you extend that all around the lesion. And then you dissect in the submucosal plane. And you can achieve end-block resection of a larger neoplastic lesion. So how do these two compare? There has only been one randomized control study in the literature. This was published now about five years ago. And they restricted their patients to those who had up to 3 centimeter size lesions with no evidence of massive submucosal infiltration, which usually means that there is no deeper excavated lesions. And again, they were, as to be expected, ESD took longer than cap EMR, had more adverse events in terms of perforation, mediastinitis, chest discomfort. And in terms of outcomes, they achieved a higher rate of R0 resection in the ESD arm. But when they looked at patients who underwent surgery or had complete remission of intestinal metaplasia or had recurrence, there was no significant difference. However, of note, the study, they only randomized 40 patients, 20 into each group. So they weren't really powered to assess some of these other outcomes. So this has been followed by ESD, as you know, has been popularized in Japan and several of the Western endoscopists have adopted this technology. And in GI endoscopy, we published two fairly decent sized papers looking at experience from Europe. This was 143 patient study, high end block, R0 resection rates, low adverse events, and also a similar series from the United States, a little bit smaller, but similar high rates of end block resection, R0 resection and adverse events. So in expert hands, this does well. And really in terms of its current status, these were some editorials that accompanied these papers. It is felt that ESD is safe, it is feasible in the hands of expert endoscopists and should likely be reserved for larger lesions, 15 to 20 millimeters or more, which may suggest the endoscopic appearance may suggest submucosal invasion. If you have scarring where cap assisted or band assisted endoscopic resection becomes more challenging, as you probably know. However, I would like to stress that EMR followed by ablation remains a very viable option because we know that there are several studies which show that you can reach a 90, 95% rate of complete remission of intestinal metaplasia for cancers. And what we need to look at are CRIM and CRD rates and recurrence rates between these two technologies to see, do we really gain much by moving from cap assisted endoscopic resection to ESD? So this was one such paper that we published from the Mayo Clinic not too long ago, where we looked at histologic outcomes of patients who underwent cap assisted endoscopic resection versus ESD followed by ablation. And we looked at the rates of CRD, which is the absence of dysplasia and the rates of CRIM, which is the absence of intestinal metaplasia. And what we found was patients with ESD reach CRD somewhat sooner at two years than those who undergo cap assisted endoscopic resection, likely because we are taking out larger chunks of mucosa. But in terms of CRIM rates, they actually achieve the results at the same time. So more to come. It is unclear if recurrence rates are necessarily very different between ESD followed by ablation versus EMR followed by ablation. The Japanese have shown this in squamous cell cancer patients as well as in gastric cancer, but ablation is not applied to those patients. So Barrett's related neoplasia might be a slightly different disease. And I believe the jury is still out in that regard. So what should happen after you have endoscopic resection? So if you find an early cancer, a T1A cancer, disease confined to the mucosa, it is low risk, well differentiated, no LVI, proceed with endotherapy. On the other hand, if it is high risk, but still T1A, you may want to discuss surgery with the patient or at least follow the patient closer. On the other hand, if you have T1B cancer, the standard is to discuss esophagectomy. And as Dr. Thotta mentioned, in a low risk group, wherein you have moderate to well differentiated disease, it is superficially invasive and the deep margin is negative, you could make a case for following these patients carefully endoscopically. So what is the rationale for ablation of the residual Barrett segment after endoscopic resection? And the rationale is that if you only were to take out the cancer or the high grade and leave the rest of the Barretts behind, you are going to have a fairly high risk of recurrent or metacronous dysplasia of cancer. This was well shown in a randomized control study, wherein the recurrence rates in the ablation arm, and in this case, they use APC, was significantly lower than in the surveillance arm, in which 36% of patients actually develop dysplasia or neoplasia at 28 months. And there is also good randomized data to suggest that if you have remaining, if you have Barretts that is left behind, then you are better off ablating versus continuing to resect. This was a randomized study again, published in Gut, which showed that ablation can reach the elimination of intestinal metaplasia with fewer procedures, with fewer complications and less friction. So it's a more efficient way of reaching complete remission of intestinal metaplasia. And another point I'd like to make is that the endpoint of endoscopic therapy should really be complete remission of intestinal metaplasia. One should not stop at the absence of dysplasia and leave some non-dysplastic Barretts behind. So this was a systematic review and analysis that we published where we looked at rates of recurrence in patients who had reached CRIM versus those who had reached CRD. And as you can tell here right away, the incidence of recurrence was substantially higher if you only reach CRD. It's not zero if you reach CRIM, but it's about less than 50% of what you reach, what you get if you reach only CRD. So we feel CRIM should remain the goal of endoscopic therapy in dysplastic Barretts. So moving on from cancer and high-grade dysplasia to low-grade dysplasia, we know from several studies that the progression risk in low-grade dysplasia, and I should add that this is confirmed low-grade dysplasia, is two to three times that of non-dysplastic Barretts. And this has been shown in several studies from the Netherlands, Denmark, Northern Ireland, and now we have level one evidence to suggest that in low-grade dysplasia, which is confirmed, if you randomize to ablation versus surveillance, your risk of progression is substantially reduced with ablation. The only caveat from the study is that the risk of progression in those with low-grade was very high, almost approaching that of high-grade. This may have been something to do with the selection of these patients. So where do we stand in terms of management of Barretts low-grade dysplasia? The first thing is to confirm low-grade dysplasia diagnosis by a GI pathologist. As we know, if you look at 100 patients with low-grade dysplasia picked up in the community, almost 85% are going to be downgraded. You then consult with the patient, discuss ablation risks, benefits, adverse events. And I think this is, again, another critical step where an expert who has some experience in identifying prevalent dysplasia should perform a careful examination with biopsies, endoscopic resection, chromoendoscopy, and provide the patient with the option of either proceeding with ablation or surveillance every year. In our experience, and I think this has been borne out in the literature, if a patient with confirmed low-grade dysplasia is sent to an expert center, almost 20% to 25% are actually upstaged to high-grade dysplasia on expert evaluation. And this decision on which modality of treatment to use should really be a shared decision between the patient and the provider. So talking about switching gears and talking about methods of ablation, I think radiofrequency ablation is the most well-studied method of ablation. It is thermal. It is a contact ablation device. And the devices are available in circumferential as well as focal configurations. And as we just showed you, there is randomized controlled data to suggest its efficacy. Meta-analyses have shown you can reach CRIM in about 80% of patients and CRD in about 90% of patients. Cryotherapy has made enormous amounts of progress. It is emerging as an alternative to RFA, but still does not have level one randomized controlled level data. And you can do this either by spray cryotherapy, wherein liquid nitrogen, which is stored in this large console, is sprayed through a catheter over the dysplastic mucosa with a decompression catheter alongside. And the efficacy in several studies have been shown to be a little bit lower, CRD rates of 80%, CRIM rates of 60%. And the challenges here are that of visibility and dosimetry, but it has been shown to cause perhaps a little less ficturing and less pain. A newer modality is that of balloon cryotherapy, which consists of a handheld controller, which is connected to this catheter advanced through the biopsy channel of the endoscope and can be inflated into a balloon. And at the center of the balloon, you have a diffusing catheter, which sprays nitrous oxide and can be rotated in a clockwise fashion. And you can treat different quadrants of the mucosa. And given the fact that only a focal device is available, this can be used to treat shorter segments. And initial data from multicenter studies suggest CRD rates of more than 80% and CRM rates of more than 70%. And again, the benefits of pain and stricture seem to hold up here. And again, we talked about rates of remission with liquid nitrogen, as well as the multicenter US trial with balloon cryotherapy, CRD rates of about 90%, CRIM rates of about 82%. Low pain scores. Now, this was a paper we just published this year or late last year, where for the first time, we really compared histologic outcomes in patients with RFA and cryo-balloon and performed a propensity score matched analysis, 85 patients with cryo-balloon, 226 patients with radiofrequency ablations, RFA non-randomized, propensity score matched. And when we look at CRIM and CRD rates, the rates are not necessarily very different when you adjust for the length of the barrets and the presence or absence of prior endoscopic resection. Hybrid APC is another emerging technique, which modifies conventional APC by the addition of a sub-mucosal fluid cushion to reduce the depth of tissue damage, causes less stricturing, and you are able to mark and inject and then treat with APC at actually a slightly higher voltage energy level, scrape the necrotic tissue, and then retreat at 40 watts, and has been shown to have reasonable rates of endoscopic and histologic remission with a fairly low stricture rate. And this might be an option in RFA failure. And I believe there is a clinical trial and process in the United States. So we talked about indications. We talked about methods. Let's talk a little bit about outcomes. And in this instance, outcomes of endoscopic therapy, particularly for cancer and hybrid dysplasia have been shown to be similar between endoscopic therapy and esophagectomy with comparable overall 5-year survival. And this has led to a paradigm change wherein we now recommend endoscopic therapy for those with T1A, esophageal adenocarcinoma, or hybrid dysplasia. I alluded to this before, predictors of mortality, lymphovascular invasion, again, an important histologic variable that is reported in endoscopic resection specimens, deep margin, positivity, again, a predictor of mortality. I would be remiss if we didn't talk about recurrence. So I think we had a short discussion about this in the Q&A session as well, is that you do have recurrence after successful endoscopic therapy. Any Barrett's about 9.5% a year, Barrett's with dysplasia 2% a year, Barrett's with hybrid or cancer about 1% a year. So keep these figures in mind. The important thing is that if you detect these recurrences, 97% of them were treated endoscopically. And this was another paper where we looked at the timeline of recurrence. And we showed that almost three quarters of recurrences actually appear at the GE junction. And so it is very important for us to biopsy the GE junction. And in the esophagus, most of the recurrences are visible. And you can easily see them and sample them. And therefore, surveillance after CRIM is essential. And you should be biopsying the GE junction and distal esophagus separately. So in conclusion, endoscopic therapy is effective for the treatment of Barrett-related dysplasia. And neoplasia, it allows us not only cancer prevention and treatment, but we also should be thinking about surveillance for non-dysplastic Barrett's. Resection and ablation techniques are complementary. RFA is a first-line treatment after resection of visible lesions. Balloon cryo is effective and safe in its initial studies, and newer ablative modalities are being tested. And continued surveillance following successful ablation is important. And I would like to emphasize that CRIM should be the goal of endoscopic eradication therapy. Thank you.

Barrett's esophagus

endoscopic therapy

tissue acquisition

intestinal metaplasia

endoscopic resection

ablation methods