Dr. Professor Barkan is a professor in the Department of Medicine at McGill University. He is the director of the endoscopy unit as well as leading the endoscopy training there and also being the chief quality officer. And we're delighted to have him speak on the non-paracel upper GI bleeding, what you need to know in 2021. Welcome. Good afternoon, everyone. I'd like to thank the organizers for the privilege of speaking to you today. We're going to talk about non-paracel upper GI bleeding, and particularly what we need to know in 2021. These are my conflicts of interest disclosures. So what I'll do is I'll use two of the international guidelines that have come forward lately. The first was the annals published in 2019. And more recently, you may have seen them, the ACG non-paracel upper GI bleeding guidelines that were just published. I would like to acknowledge Lauren Lane, who headed the group, as well as John Saltzman, who's here somewhere in the audience, Mireme Martel, my collaborator, and Grigoris Lantiadis, who was our great methodologist. Combined, they did an amazing, amazing job. And I'm just the spokesman, spokesperson for the group. So overall, non-paracel upper GI bleeding mortality used to be about 10%, probably has gone down to about 2% in some of the larger national cohorts, and has decreased therefore over the last 10 years. Mortality is principally related to comorbid illnesses, and those used to be mainly respiratory and cardiac, but more recently, more so malignancy and chronic kidney disease, presumably because of platelet dysfunction, at least in part. Management is based on endoscopic hemostasis, which I remind you is the cornerstone of therapy, and PPIs should be considered an adjuvant to endoscopic therapy once that has been done successfully. So the question is, where do we stand in 2021 with the specific focus on the endoscopic management, but highlights some key issues leading to endoscopic endoscopy? I'll be talking mainly about peptic ulcer bleeding, but during the question session, we can discuss it further. We can see some of the other etiologies as well. So if you don't know the Glasgow Blachert score, you probably should. It's demonstrated here. So you see based on BUN, hemoglobin, systolic blood pressure, as well as how the patient presents, so based on the pulse, whether they have the presence of melina, syncope, and associated comorbid conditions. Note that just the presence of melina is already one point. This is important because when I show you the first guideline, you'll understand that the patient has to be perfect with no warning signs except for the presence of melina if you want to try and send them home. So who to admit versus send home? And we suggest that when patients present to the emergency room with an upper GI bleeding, the patient can be classified as very low risk if there's less than a 1% chance of having a false negative classification. And that is performed adequately with a Blachert score of 0 or 1. So if you want to use that score, by all means. And this has been shown to decrease the need for hospitalization in about 20% to 30% of all patients presenting with non-variceal upper GI bleeding or variceal. This is prior to scoping. Who to transfuse? This is an old message. We suggest a restrictive policy of RBC transfusion with a threshold of 7 grams per deciliter, aiming for 7 to 9 grams per deciliter. This is based on five randomized studies, almost 2,000 patients that showed a decrease in all-cause mortality as well as further bleeding without any nefarious increases in complications, either cardiac, cardiovascular, peripheral vascular, or need for additional interventions. With regards to ProConnect therapy, it's also an older message. We suggest an infusion of erythromycin before endoscopy in patients with upper GI bleeding. This is now based on eight studies in almost 600 patients. It's unclear what the optimal patient selection is, but usually, for sure, if patients have presented vomiting up red blood, absolutely. This results in improvement, adequate gastric mucosal visualization, as well as decreasing the need for repeating an endoscopy in order to find the bleeding lesion. It is also associated with a reduced length of stay, but there are no other increased advantages across the board. With regards to pre-endoscopic PPI therapy, this is a change in what had been previously suggested. We could not reach a recommendation for or against pre-endoscopic PPI therapy for patients with upper GI bleeding. This is important because often, the emergentologist will come in and say, oh, rest assured, the PPI drip is running. You can take your time to see the patient, or you can take your time to see the patient, but PPI drip does very little. In fact, a meta-analysis we did specifically for the guidelines of the three studies showed that there was a decrease in the high-risk lesions that are seen when you eventually scope the patient, but no actual improvements in other outcomes. There are some cost-effectiveness analyses that have looked at this and suggest perhaps if you are going to use it, you should use it if it's going to take a while to scope the patient, and if you're pretty confident the patient is bleeding from a non-variceal etiology. So when to endoscope? This is an old message with new data. We suggest that patients admitted to or under observation in hospital for upper GI bleeding undergo endoscopy within 24 hours of presentation. The evidence for that is that potential economic benefit with reduced length of stay, as well as a possible decrease in mortality and need for surgery in some observational studies. But the question was in the very sick patients, is that in fact so? And we have new data, which I'll share with you in a second. First, you have to scope the patient. Please keep in mind that the sickest patients need adequate resuscitation. This is a large cohort published from Larson and colleagues that shows that amongst the sickest patients, ASA scores three to five, you have a U-shaped mortality curve, suggesting that if you scope the patient early, within six hours, there is actually an increased risk of a negative event for the patient. The patients need adequate resuscitation before being scoped. This is the actual new piece of evidence, specifically looking at high-risk patients with regards to timing of endoscopy. And what you can see here is a large randomized trial of very sick patients, Blatchford scores over 12, and they randomized patients to either get an urgent endoscopy or the so-called early endoscopy. You can see that the patients bled from bleeding ulcers, varices, other lesions, the type of patients we would all see. The time to endoscopy on average was 10 hours in the urgent group, and the average time in the so-called early group was actually over 24 hours. It was almost 25 hours. Randomized 516 patients, and the results are no differences in 30-day mortality, no difference in 30 bleeding at 30 days, even though there was an increased finding of high-risk peptic ulcer bleeding lesions and more endoscopic hemostasis performed when you scoped the patients early. And this parallels a little bit what's seen in lower GI bleeding and what's been seen before in other similar studies in upper GI bleeding. Somehow we find more lesions if you go in earlier, but treating those lesions don't seem to translate into improvement in outcomes. So even amongst the sick patients, we recommend scoping within 24 hours. Now, whom to treat an endoscopy? Well, we all know that we treat the patients with high-risk lesions, the active spurting, the oozing, and a non-bleeding visible vessel. Where this guideline is a little bit different is that we could not reach a recommendation for or against endoscopic therapy in patients with upper GI bleeding due to ulcers exhibiting adherent clots resistant to vigorous irrigation. Usually, we recommend about five minutes of vigorous irrigation. And what do we call a clot as opposed to an adherent vessel? Usually, if I look at the base and if I know where I want to do therapy, then that you can think is a non-visible vessel, non-bleeding visible vessel. If, however, it's a diffuse clot and you're not sure where to go ahead and start endoscopic therapy, that's what I use clinically to decide that it's an adherent clot, then try and do continuous irrigation and proceed. And you can treat those patients with PPI alone or do endoscopic therapy followed by high-dose PPIs. Coming now to the different types of endoscopic therapies, there's not a huge amount new here, but perhaps it's been looked at with additional data and gone through the grade grinder of recommendations. So you can see that the bipolar electrocoagulation, heater probe, and injections of absolute ethanol, which we have mainly from Asian studies, in these we have the best evidence, so-called moderate certainty of evidence, which translates in grade terms into a strong recommendation. So we use the word recommended as opposed to suggested. If you look at the data using through-the-scope clips, argon plasma coagulation, and soft monopolar electrocoagulation, the certainty of evidence is actually lower. It's very low to low, yielding only a conditional recommendation. This is why we use the words we suggest. And finally, epinephrine, the role of epinephrine, it should not be used alone, but rather in combination with a second hemostatic modality. This is an old message, but we have additional data present there that allow us to make it now a strong recommendation. So let me just speak about a few of these briefly. So we do recommend endoscopic therapy for hemostasis when using the bipolar electrocoagulation, the heater probe, and absolute alcohol. I remind you that if you're going to use a bipolar electrocoagulation mode, use a large probe, 3.2 millimeters, fixed maximal pressure at settings of 15 watts or 30 joules, if you use the heater probe, and you apply for 8 to 10 seconds. You really need to push on that for a prolonged duration of time. Sclerosine injection, which we don't use as much in North America or Western Europe, aliquots of 0.1 to 0.2 mils per injection are used with a maximal of 1 to 2 mils to minimize serious tissue injury. Now if we look at, as we said before, you look through the scope clips, argon plasma coagulation and soft monopolar coagulation. Yes, go ahead and use them, but the evidence is weaker. With regards to clips, there is no studies comparing it to no therapy. Compared to epinephrine, there is a decreased benefit, and clips versus thermal, unfortunately, we had no differences on a very limited study. Looking at APC, just to remind you again, the flow settings, as you can see here, 1 to 2 liters per minute, power settings of 40 to 70 watts for duodenal and gastric ulcers, and a distance from the mucosa about 2 to 10 millimeters. Finally, the relatively new kid on the block is a soft monopolar electrocoagulation. Of the five randomized trials, only one was pertinent to present-day usual recommended approaches, and you can see this is different than the traditional monopolar electrocoagulation that we did not recommend because of tissue injury. Using a soft coagulation tip, you can use the grasper, grasp the lesion and raise, or you can close it and use it almost like a bi-cap, if you will, and putting pressure on it at 50 to 80 watts with 1 to 2 seconds application, but again, very limited of the five studies. We could only use one to make the recommendation. With regards to hemostatic powders and gels, the group could not fully agree. That's why I have a star in here, but we do all agree that TC325 or hemo spray has a role to play for patients with active bleeding ulcers. Within the past, I've recommended for rescue therapy. In this case here, we now recommend as first-line therapy. Typically, as for those of you who have not used it, the delivery catheter has to be kept dry. You insert the tip 1 to 2 centimeters from the bleeding site and you do 1 to 2 second bursts until you cover the bleeding lesion. There's a non-inferiority randomized trial that will soon be published in the annals of internal medicine looking at 224 patients, including 130 ulcers, where they actually used hemo spray versus standard approaches and found that it was non-inferior to use a hemo spray as first-line therapy. The reason I have not one but two stars here is that there are methodological issues with this with more representation of malignant bleeding, which may actually benefit from this approach of only using hemo spray, at least in a short duration of time. We can discuss this if you have further questions. The panel nonetheless did recommend, because of the very high cost of this approach, to consider using other modalities initially on when treating patients with bleeding ulcers. What about the over-scope clips? We suggest over-the-scope clips as hemostatic therapy for patients who develop recurrent ulcer bleeding due to ulcers after previous successful endoscopic hemostasis. So we recommend it first as an approach in patients who have recurrent bleeding. This is, for those of you not familiar, a CAP device with a single clip that's placed on the distal tip of the endoscope. The bleeding lesion is approached. You suck the lesion or you draw it with a grasper and go ahead and deliver hopefully as much through the mucosa bite as you can provide. I'll show you a video on this shortly. That data was informed, or that recommendation was informed, from the evidence from a randomized trial published in Gastro in 2018 from Schmidt and colleagues, where they randomized patients to either the over-the-scope clip. These are patients who had recurrent bleeding versus standard therapy. And you can see that the primary endpoint, which was further bleeding, was lowered significantly so in the patients having the over-the-scope clip versus the traditional therapy, which was mainly clips and injection. No difference in secondary endpoints because the study was very small. More recently, there's been another randomized trial looking at over-the-scope clip from Dean Jensen, published in CGH last year. A very small study, but this one looked at using the over-the-scope clip as primary therapy for non-variceal bleeding and found benefits. However, there are a number of issues with the study, particularly with regard to generalizability and some with regard to internal validity, where we could not feel comfortable or we did not feel comfortable recommending it as a first-line therapy. At this point, I'm happy to discuss it further in questions period, if you have any. So this is the video I want to show you of the over-the-scope clip. I think I'm ready to set you for it. And the reason I chose this video is I think it highlights very nicely the transmural bite that you can grasp with this particular tool. This is a patient who actually bled, who had recurrent bleeding, who actually went to transarterial embolization. In fact, what you see there are the coils that were injected at the time of transarterial embolization, and then the patient bled again. So Amrita and colleagues brought the patient back. There's no bleeding right now, but that'll change shortly. Don't worry. And as you can see here, they're trying to initially grasp the actual bleeding lesion and bring it into the cap of the device, which did not work. And then they proceeded to try and actually suction the base of the bleeding ulcer into the cap in order to be able to deliver the transmural bite of the clip that you can see positioned here. And sure enough, a little bit like when you do a variceal banding, eventually they got the right plane. And luckily, the base of the ulcer was not too scarred that it could, in fact, be drawn into the full cap itself. And what you're going to see here is obviously going to start activating bleeding, which gives you an idea of the significant bleeding. That's the pancreatic adenoid artery in the posterior bulb of the duodenum that's doing its job when it bleeds, unfortunately. And then they delivered the transmural bite. And the reason I showed you here, this shows a nice family picture where you can see the transmural clip, but also the coils that are standing out because the clip is beyond them, illustrating the depth of clipping of the actual lesion itself. And I'll try and go to the next one, but I don't seem to... Is it possible to move to the next one for me? I don't seem to. Thanks. Now, before I finish, a couple more recommendations. What about post-PPI, post-endoscopic therapy? So we recommend high-dose PPI therapy for three days. This is sadly like religion. You have people who believe in the IV high-dose, and people who believe in the high-dose but not the continuous infusion, and they just never agree. The bottom line is that the best quality data are with the continuous infusion for three days, 80 bolus followed by eight per hour, but that there are excellent data as well for a high-dose other, so a little bit less so. But the issue is, and the key is, for three days, because this is how long it takes to go from a high-risk lesion to a low-risk lesion. What about PPIs following endoscopic therapy? You have to follow me here. These are patients who've received the high-dose for three days, and when they go home, they should continue not single but twice daily PPI therapy for the first two weeks if they were high-risk bleeders. And this is based on one Korean study that suggested this. So they get the high-dose for three days if they're high-risk, they were high-risk bleeders, you double oral daily dose for the first two weeks, and then they can go to a single oral daily dose for the remainder of the treatment, which is usually four weeks for bleeding duodenal and eight weeks for bleeding gastric ulcers. What if a patient rebleeds? We've talked about the over-the-scope clips already, but in general, we recommend that patients who have recurrent bleeding from endoscopic therapy for a bleeding ulcer, that these patients undergo repeat endoscopic therapy rather than undergo surgery or transcatheter arterial embolization. This is based on an older study, a randomized trial comparing to surgery, and also on the fact that over three quarters of all patients treated endoscopically will, when you treat them a second time, will stop bleeding. There is a little proviso there. If you do go ahead and do a re-treatment, be careful if you've done thermal first time around, because if you do thermal second time around, there's some, not great data, but some data suggesting an increased risk of perforation in that context. So keep that in mind. What do we suggest? If the patient fails endoscopic therapy, well, we suggest that patients with bleeding ulcers who failed endoscopic therapy should then go to transarterial embolization. If you actually look at the studies, you'll find that surgery is better in preventing further bleeding. The problem is the cost that you pay in morbidity and side effects and complications. And with the development and availability of transcatheter arterial embolization today, this is the way we would recommend to go. Of course, all of these recommendations are based on available expertise. Now, before I finish, I do want to spend a couple of minutes. It's not endoscopic therapy, but you need to know and make sure that patients are well taken care of when they, in fact, go home. So first, these are all reminders, very brief, and there's not a huge amount new here. H. pylori eradication is superior to acid suppression. It's malpractice not to look for HP for a patient who's come in with bleeding, aldodenal, or gastric ulcer. Also, be aware of H. pylori testing. If you're going to test the patients for HP in the acute context of the ulcer bleed, be aware that there's up to a 55% risk of false negative tests. So that patient needs to have repeat testing outside the acute context of bleeding. And finally, in patients with previous ulcer bleeding who are receiving cardiovascular prophylaxis with either single or dual antiplatelet therapy, we suggest continuing the PPI treatment for as long as the patients will need to remain on that cardiovascular prophylaxis regimen. This is above and beyond looking and treating for H. pylori. And similarly, although there's less data for that, in patients with previous ulcer bleeding requiring continued cardiovascular prophylaxis but with anticoagulants, let it be vitamin K antagonists or DOACs, we also suggest using PPI therapy versus no PPI therapy for as long as those patients require the anticoagulation. Again, this is based on less data. So in conclusion, there are past recommendations and there are a few new ones. The risk score, think of the Blatchford scale, is excellent. If your patient has a score of 0 or 1, they can safely be sent home, discharged, and brought back for a scope in an elective setting, usually within a couple of days of the initial presentation to the emergency room. Transfusion practice, restrictive transfusion strategy with a threshold of 7, aiming for 7 to 9 grams per liter, per deciliter. I'm sorry, I work in different units in Canada and most of the rest of the world, but we won't go into that right now. And so just keep that in mind. And if your patient has cardiovascular disease, we recommend a threshold value of 10, but that's based on very few data right now. Timing of endoscopy, including the very sick patients, we recommend within 24 hours. Do not feel the need to be pushed into doing something earlier, particularly if the patient has not been adequately resuscitated. Do not treat with injection of epinephrine alone, an old message, and there are possible limitations in using a hemo spray alone as well. High-dose PPI therapy, I've mentioned already. Consider the over-the-scope clip if the patient has recurrent bleeding, and if that fails, then trans-arterial embolization. And in all cases, and I always finish my talks like this, but I think it is absolutely important to remind ourselves that you only perform procedures you're comfortable doing in the appropriate clinical setting with adequate support. And I would mention to you that we will be having, actually, ACG updated antithrombotic guidelines in early 2022 that you may want to look at. They will complement some of the messages I've given you today. Thank you very much for your attention.

non-variceal upper GI bleeding

endoscopic hemostasis

Glasgow Blatchford score

PPIs

endoscopic therapies

H. pylori eradication