Hello and welcome. My name is Dr. Jennifer Christie, and I am Professor of Medicine at the Emory University School of Medicine. I'm also ASGE Vice President. We are so excited to deliver the 2021 AASLD-ASGE endoscopy course to you today in this virtual format. I want to thank my colleagues, Drs. Jawad Ahmad and Mark Gromsky for their collaboration. We also want to thank the ASGE and AASLD staff for collaborating and putting this wonderful course together presented to you today. Then lastly, thank you for joining us and participating in these discussions as we discuss ways to better manage our patients with liver disease through endoscopic and other approaches. Our first session will be focused on a patient who has cirrhosis and presenting to the emergency room with hematemesis. Our first speaker will be Dr. Arun Sanhall. Dr. Sanhall is Professor of Medicine and Chair of the Division of Gastroenterology, Hepatology, and Nutrition at Virginia Commonwealth University. His research and clinical interests include portal hypertension, bile acid regulation, as well as iron and calcium absorption. Dr. Sanhall will discuss EGD for upper GI bleed, the right time and place. Our second speaker will be Dr. Virginia Hernandez-Ghia. Dr. Hernandez-Ghia is a professor at the Clinic Hospital Barcelona. Her interests include focusing on therapeutic targets for patients with liver fibrosis. She will discuss tips for tips in cirrhotic patients with upper GI bleed. Our last speaker for this session will be Dr. Lauren Nephew. Dr. Nephew is Assistant Professor of Medicine at Indiana University School of Medicine. Her research interests includes understanding barriers to liver transplant in vulnerable populations and understanding disparities in the care of patients with hepatocellular carcinoma. Dr. Nephew will discuss disparities in access and outcomes of emergent EGD for upper GI bleeding. Thank you to our speakers and we look forward to this very exciting discussion. Good afternoon, ladies and gentlemen. It's a great pleasure for me to talk to you today about EGD in upper GI bleeding, the right time and place. Here are my disclosures. Now, we all recognize and dread this call from the emergency room, which is a 53-year-old male with cirrhosis and a male of 25 has just showed up and is throwing up blood. Because what we worry about is variceal hemorrhage, which can be a life-threatening event. In this setting, we not only have to control the bleeding, but we have to control a number of other factors that's going to allow us to keep our patient alive. Specifically, our goals of management are to prevent exsanguination both by hemodynamic support and by achievement of hemostasis, prevention and treatment of complications, both due to bleeding and from the therapies that we provide, and to treat the underlying liver disease and related complications and comorbidities. Very importantly, to prevent and treat sepsis and prevent the slide into multi-organ dysfunction. In this setting, we have to now ask the question, what does endoscopy provide to the care paradigm? Clearly, it does contribute to bleeding management by providing a diagnosis. Really, the dichotomy here is variceal versus non-variceal bleeding, and then the ability to provide hemostatic therapy directly while you're doing the diagnostic test. But there are some practical considerations about the timing of endoscopy, endoscopy in the time of COVID, support systems that need to be brought into place, and issues related to intubation, need for intubation, concurrent hemodynamic instability, and how to manage that, and coagulopathy, etc. Let's work our way through some practical aspects of how we deal with all of this. One of the first issue is, when the patient comes in is to decide whether a endoscopic intervention is indeed indicated, because there's bleeding and then there's bleeding. There are a number of scoring systems of which the Glasgow Blatchford score is probably one of the more validated ones, but there are others such as the Rockall score, etc. These have been all looked at simultaneously to predict the need for hospital-based intervention or 30-day mortality in patients coming in with acute upper GI bleeding. And many of them actually do show that these perform relatively well. So in this particular study, they looked at a number of these and found that the Glasgow Blatchford score was particularly good at identifying the need for hospital-based intervention or mortality. And what you see over here is based on the different scores, as well as the Rockall score, that these all have high sensitivity, but their specificity is quite modest. So these probably have good negative predictive value, and so good at ruling out the need for these interventions. But once patients come in, the value of this is probably more iffy. So the principal utility of these is that they provide a high negative predictive value. So that leads to the next question. If somebody does need an intervention and you just got your call at 2 a.m., the question now is, scope now, or do you wait till the morning? So there are many arguments for doing it right away. There's an emotional need that we all feel that if somebody's bleeding, we need to stop the bleeding. Also, the longer and the more the patient bleeds, the more blood products will be needed. And blood products can have their own set of problems. The risk of complications may increase if you delay therapy. And you worry about transplant candidates becoming non-transplantable because something bad happened to the patient and now they are in multi-organ failure. On the other hand, resources have to be mobilized, adding cost to care. Safety of EGD in unstable patients is still an open question. We don't want to kill the patient while trying to help them. And is there evidence really to support better outcomes with early EGD? So here's a randomized controlled trial of urgent versus early endoscopy in patients coming in with upper GI bleed from Hong Kong that was published in New England Journal last year. 516 patients, and they were randomized to get a scope within six hours or between six and 24 hours. Now, they did not demonstrate any clear cut difference in the cumulative probability of death or in the cumulative probability of further bleeding. However, it is important to remember that a very small number of these patients actually had variceal bleeding. So that still remains an open question. So when you look at a meta-analysis of the studies looking at early endoscopy versus delayed endoscopy, well, I don't really want to use the word delayed, but at a later endoscopy, no clear cut evidence for benefit really emerges. And so, as you can see, even with the pooled odds ratio, it intersects the unit line. Now, this is an interesting study that was summarized by Lauren Lane in this very nice review that I would recommend everybody to look up. And this really looks at the association of timing of endoscopy and mortality in hospital patients with hemodynamic instability after correction of confounding variables. So this is actually a Danish study. And here, what you see is that there's a bimodal distribution of outcomes. First, if you scope too early, the outcomes rates are high, but then there's a sweet spot between 10 and 20 hours, which probably has given the team enough time to stabilize the patient. But then if you delay beyond 20 hours, again, it starts rising. It's important here, again, that peptic ulcer bleeds were the principal source of bleeds and the potential benefits of stabilizing patients before EGD is demonstrated here, but it's also important to note that most patients died from comorbid conditions rather than exsanguination. So there are some clinical factors that should guide decision-making in those with cirrhosis. First, the outcomes may vary based on whether bleeding is the sole decompensating event in an otherwise stable cirrhotic or is bleeding occurring in the background of someone with profound liver failure, ascites, and encephalopathy. The same bleed, the same severity of bleed in someone who has ascites, encephalopathy, and albumin of 1.5 and a baseline MELD score of 30 has a very different prognostic profile compared to the stable cirrhotic patient who shows up for the first time with a variceal bleed. Other factors that determine outcomes, a high HVPEG, and more than 80% of child PUC patients meet this threshold, by the way. MELD score, higher MELD scores, presence of HCC and presence of portal vein thrombosis. So when you have these high-risk settings, you may want to intervene a little bit sooner rather than waiting. But it is really important to stabilize the patient before rushing in. You don't want to take someone with systolic blood pressure at 60 and then try to scope them while they're throwing up blood with an unprotected airway. So you have to sort of take care of all of that stuff. And this is sort of summarized in this ASG guideline as well. Now, in addition to endoscopy, we talked about stabilizing. So part of the stabilization is the need for blood transfusion. And one of the big shifts that has occurred in the last decade is that we have become much more restrictive in how we use blood transfusions. So this is a meta-analysis showing that a restrictive approach to transfusion, limiting keeping the hemoglobin above eight or nine generally does better than to continuously replace every drop of blood that a person vomits out and try to keep the hemoglobin at higher levels. And the threshold for transfusion generally is now considered around a hemoglobin of seven, et cetera. Another thing that has emerged is what can we do to minimize the need for blood products because blood products have their own profile of adverse events. So this is a study that was published last year from India. We're using thromboelastography versus standard of care approach for blood product use in non-variceal bleeds in the setting of cirrhosis. And where they used TEG and TEG-based approach to use of fresh frozen plasma and cryoprecipitates, et cetera. And similarly, they had a a priori defined plan for standard of care. And what they found over here is that the need for FFP could be cut by half. The need for platelets could be reduced and the need for cryoprecipitates could also be reduced. So overall, you could diminish the need for blood products, but this really did not have a survival advantage. Now, the other issue that we often will encounter is the patient who is on an antiplatelet agent. As we evaluate patients for transplant and we are seeing more and more patients with NASH who are coming to transplant, they undergo cardiac evaluation and we discovered that they have advanced three vessel disease. Many of these patients undergo stenting in preparation for becoming transplant candidates. Once that happens, when they come in with the upper GI bleed, the question always arises what to do and how to approach these patients and should platelet transfusions be given if there's already on antiplatelet agents and they are now bleeding. So these are data that were published from a cohort study in CGH in 2017. And they make the case over here that the odds ratio for platelet transfusions when they present with upper GI bleeding for recurrent GI bleeding actually did not seem to, none of these really support that these antiplatelet transfusions made a big difference. And you will see that the confidence limits intersect one for all of these different clinically important events. Importantly, even the rate of myocardial events or MACE was not significantly different whether they got platelet transfusions or not. Antibiotics on the other hand do make a difference particularly in the cirrhotic patient with upper GI bleeding. And that's one of the things that should be and is part of the standard of care and is and should be a quality metric in this setting. What about PPIs? So meta-analysis failed to show that at 30 days there is a mortality benefit or a re-bleeding benefit. And certainly the UK NICE guidelines do not recommend routine PPI although there are other guidelines that support it based on somewhat decreased need for endoscopic intervention where the hazard ratio drops to 0.68. People also use prokinetic agents to improve gastric visualization. And this is summarized over here in this meta-analysis which shows that it reduced the need for second look endoscopy. And so this is often used as well. So there are a lot of stuff that you have to do beyond just making a decision to scope and when to scope the patient in order to have the best outcomes for your endoscopy. So in this patient that we first described stabilized with fluids, bacterial cells, octreotide antibiotics and EGDs performed after erythromycin to clear the stomach. So in patients with cirrhosis presenting with upper GI bleed, as we mentioned before the principal dichotomy is variceal versus non-variceal bleed and we also think about it as liver related versus non-liver related. Note that both liver and non-related sources of bleeding can lead to life-threatening hemorrhage. So it's really what they have different therapeutic approaches as we'll talk about in a moment. So this is a meta-analysis of variceal bleeding shows that you need both pharmacological and endoscopic therapy and is superior to monotherapy alone. And so that sort of standard of care is in every practice guideline and should be part of your basic paradigm. We also know that patients who have advanced liver disease have a higher mortality and likelihood of failure to medical therapy. And this has been shown for child view as well as with MELD. Mortality risk for people with MELD score less than 11 is relatively small 2%, whereas those with a MELD over 18, it rises to over 35%. So this also gives you some idea about what are the type of patients that you may wanna move rapidly towards tips or a more definitive form of therapy after you've done the initial stabilization. I'm not gonna talk about tips because there's a whole separate talk about it. But moving on, in addition to variceal bleeding, we may sometimes just see oozing of blood and we see severe portal gastropathy where the entire mucosa is sort of oozing blood. And there are anecdotal reports of using hemo spray to manage these acutely. And of course you can use beta blockers and tips in the long-term for control of recurrent bleeding or need for transfusions from portal gastropathy. What about gastric varices? Gastric varices can be classified as esophageal varices and gastroesophageal varices in continuity with esophageal varices along lesser or greater curve or isolated gastric varices which are often associated with spontaneous splenorenal shunts. So with IGVs, which can bleed at lower portal pressures than esophageal varices and bleed more severely and repeatedly. Pharmacologic treatment with beta blockers more for prevention, but for when somebody is actively bleeding. Sinoacrolat has probably emerged as a leading way of endoscopically managing this and is somewhat superior to beta blockers in this setting as shown on the left. Although this does not necessarily translate into a clear-cut survival advantage. Now, BRTO is another option for these patients. And there's some evidence that BRTO may be even better than Sinoacrolat, but BRTO is not available at a minute's notice everywhere. And so certainly Sinoacrolat provides the endoscopist an opportunity to intervene and slow down the bleeding and stabilize the patient further before a more definitive treatment is performed. Now, when you do glue injection, there is increasing evidence that if you do it, put on a Doppler probe and document cessation of flow in the varices after glue injection, you can particularly increase the efficacy. There's a small study published in gastroenterology a couple of years ago from Southern California, further suggesting that this may reduce re-bleeding rates, but we need more data from multiple centers. Now, what about non-variceal sources of bleeding? So epinephrine combined with bipolar electrocoagulation heater probe or hemoclips are sort of the garden variety standard of care that we provide. The decision to apply endoscopic treatment is based on the forest classification. This is shown on the slide. I won't bore you by reading it, but we know that 20 to 25% of upper GI bleeds due to ulcers meet this criteria. And particularly for people who are spurting blood, oozing, actively bleeding, or have non-bleeding visible vessels, the application of hemoclips is becoming much more mainstream and is mainstream. What about hemospray? This is sort of the new kid on the block, but it is currently used mainly as a secondary treatment after you've applied your first-line treatments. And if findings after initial therapy show that you have some continued bleeding, and that is sort of the current AGI clinical practice update guideline as well. Now, once you're finished with all of this, there is a post-endoscopy checklist, which is also extremely important, because it is not only important to do the endoscopy at the right time, in the right setting, provide the right therapy, but also to make sure that once you pull the scope out, all of the things you've done to that point do not create a problem for the patient afterwards. So very important for continuous assessment of mental status. Airway protection, particularly for those whose mental status is such that they cannot maintain their airway and are at high risk for aspiration. Hemodynamic stabilization and monitoring. Monitoring of urine output. Once the patient has been stable for a period of time, you can restart feeding. The literature suggests that all of the trials done for octreotide, which is what's used in the US and other drugs elsewhere, that you continue them for five days. In reality, typically when patients are moved out of the ICU, many of these drugs are discontinued. Antibiotics are extremely important. Please do not forget this. This is part of your standard of care, likely to be a performance metric. And then for those on anticoagulants, you have to do risk-benefit analysis, particularly, I would wait a certain period of time to feel comfortable that you have applied your therapy, if you're going to induce ulceration of the mucosa that it's well under healing before you restart anticoagulation or antiplatelet agents in people who have had significant bleeding. Sepsis is a critical part of the story. Patients who get into trouble and go into multi-organ failure and die often become septic. So not only is the role of prophylactic antibiotics important, all the usual steps that we take to minimize infection and hospital-acquired infections need to be in play after you've done your procedure. So with that, here are my take-home messages. Early risk assessment is important. You got to stabilize the patient. Evaluate comorbidities and other complications which are concomitant threats to life and manage these aggressively, particularly infections. Endoscopy is part of the management plan and is not the management plan and should be integrated with the overall management of the patient. This is sometimes hard to teach our ICU colleagues and our surgical colleagues, but this is something that is worth reiterating. Endoscopic therapies should be followed by risk assessment for outcomes and a plan for managing re-bleeding that is communicated to all team members managing the patient so that in this multidisciplinary environment in which we manage these patients, everybody's on the same page about how we're going to approach the patient and how we're going to treat the patient with pre-specified criteria for escalation of care that will eventually hopefully translate into the best outcomes for our patients. I thank you for your attention. Well, first of all, I would like to thank the organizers for giving me the opportunity being here with you today, speaking about one of my favorite topics, that is tips placement. They asked me to give you what I consider are the tips for tips placement, and I think there are like four very important things. To whom are we placing the tips? When? Who is placing it? And how do we do that technique? So before starting, I want to really highlight what are the indications for tips in patients with variceal bleeding. So there is no indication in primary prophylaxis, but if we have a patient with an acute variceal bleeding, we can place a tips with the aim of controlling the bleeding, because probably the patient is bleeding massively, and this is what we call rescue tips. Or we can place a tips just to prevent failure or to prevent the bleeding and death in a patient who is stable, and this is what we call preemptive tips. And sometimes the patient is bleeding and we're bleeding again, we follow up, we may place a tips, another rescue tips to prevent further bleeding. This one is not going to be in an emergency setting, but it's still a rescue tip. So usually when we are facing the typical patient needing a rescue tips is like here in the picture, you have a very sick patient, most of the time intubated under vasoactive drugs, and then sometimes with a balloon tamponade or a esophageal stand, because these tips rescue or salvage tips, it's a tips that we place in a patient with a massive uncontrolled bleeding, or in a patient after failure arises, and most of the time is after a bridge therapy with balloon tamponade or esophageal stand. So here the question is, are all the patients candidates to rescue tips? So patients admitted with a variceal bleeding needing a rescue tips, they can be stratified according to ACLF, because we know that ACLF grade, it really impacts the six weeks mortality. But the recent data from the UK have shown that even these very sick patients with ACLF, they benefit from tips placement. And the same data have been recently published in a cohort of patients from France and Spain. And here in this study, it seems that patients with ACLF 3B grade may be placing a tips in those patients may be futile, because they have a very high mortality. But in this paper, when we try to identify the prognostic factors that may help you stratifying the risk of the patient's levels of lactate and the male score, are able to really identify those patients at high risk of death, in whom the placement of the rescue tips may be futile. And this is a lactate levels equal or more than 12 and or a male equal or more than 30. Still in this setting, the mortality is very high. And the six weeks survival rate in this specific setting was 58% the whole series. Something different is when we speak about preemptive tips, because here the scenario is a little bit different. So we had our patients and acute variceal bleeding, we did everything in the general management preventing complications, we did the endoscopy with variceal band ligation, we started on vasoactive drugs, and we have everything controlled. However, if the bleeding episode is controlled up to 20% of these patients will experience failure to control bleeding or re-bleeding, and they may need rescue tips. So this is why we try to prevent failure, placing a preemptive or an early tips. So this is a tip that we place in these high risk patients that we know that they are a very high risk of presenting complications. We do that after establishing the patient after treatment with hemostatic treatment with vasoactive drugs and EBL. And it has to be early because we have to anticipate failure and we have to place that before the failure arises. So we have like many scores able to identify patients at risk of poor outcome, but we only have two validated in guiding therapy that are the HPPG and the childhood score combining with the active bleeding at the endoscopy. I'm sure that you are all aware of the multiple trials that we have so far in the setting of preemptive tips in variceal bleeding, and we have observational studies, we have randomized clinical trials, and the definition of high risk has been a little bit different. In the first study, it was an hemodynamic measurement who was defining the risk. But in the rest of the studies, we have clinical data, childhood plus active bleeding at the endoscopy. So in all of these studies with different details, but preemptive tips was able to improve outcome in these patients. However, the question here is, do all the patients have the same benefit? If we restrict to the child C population in all these studies, they included child C patients with less than 14 points and the survival benefit was seen in all these studies. However, what happened with patients with 14 and 15 points, it's tips always futile in this population, and we have this data coming from France, showing that in this specific population, if you have a liver transplant coming after the tips placement, tips can help you to control the bleeding and maintain the patient alive while you are preparing everything for the liver transplant and can be also considered in this kind of patients. What happened in the child B population? We have been arguing a lot in the last years, whether they benefit or they do not benefit. So if we take a look at one of the Chinese studies performing preemptive tips in patients with child A, B, and C, you can see here how the patients in the class A, they do not benefit from tips, while the child C, the benefit is so clear. And what happened in the child B class? Probably, or if you mix all the patients with and without active bleeding at the initial endoscopy, you don't see any benefit, but when the outputs stratify those patients according to the presence or not of active bleeding, they could see how the benefit was concentrated in the child B plus active bleeding. However, in the European observational study, we were not able to detect these differences and probably because we only have 19 patients in this category treated with preemptive tips. So this is why in Barcelona, we decided to run a meta-analysis of individual patient data, and we collected data from all the published studies at that moment, and we included more than 1,300 patients and 310 patients treated with preemptive tips. And we analyzed the survival according to the childhood in the whole population, we could demonstrate the benefit in survival, and also when we stratify in child C or child B patients. However, if you take a look at the Kaplan-Leyer curves, you will see how the child could be less active bleeding, then the separation of the two curves is not as wide as in the child C patient, and this is why we decided to study this population a little bit further, and we could identify that the survival benefit concentrates in those patients with more than seven points of child B with active bleeding, and now we have an agreement, and those are the patients that are in the child B population that we will treat with preemptive tips with the final aim of improving outcome and improving survival. But can we identify in these preemptive tips setting the population of patients that do benefit from the preemptive tips, and we came with this score able to differentiate three different classes according to prognosis, good prognosis, intermediate prognosis, and poor prognosis, taking into account age, the childhood, and creatinine, but in all the classes, preemptive tips was able to improve survival, so if your patient is fulfilling the high-risk criteria, you should go for a preemptive tips, and the same thing happens if your patient fulfilling the high-risk criteria has a high hypervigilance level or has an ACLF, because those patients, they also benefit in terms of survival of this strategy of placing a preemptive tips. So where are the contraindications for tips placement, or where are those patients in whom we won't place a tip? So we have absolute and relative contraindication, and this is something that has been changing over time as well. So of course, if you have a patient with a decompensated congestive health failure or an uncontrolled systemic infection or sepsis, you want to consider this patient a candidate for a preemptive tips. However, there are like other relative contraindications that deserves a careful study of your patients to see if this patient may be a good candidate or not. The most important one probably is cardiac decompensation, and we have this paper coming from France demonstrating that the levels of the BNP or pro-BNP can help you a lot in taking the decision, and that together with an echocardiography can help you in identifying those patients at high risk of having cardiac decompensation, and here you have the parameters that were associated in this study with severe cardiac decompensation during follow-up after tips placement. What about hepatic encephalopathy? This is probably the most severe complication of tips, and I'm here showing you this data submitted for publication right now, evaluating in the setting of the preemptive tips, evaluating patients fulfilling the high risk criteria that had or were not hepatic encephalopathy at admission, and in this population, even those patients, high risk patients with hepatic encephalopathy at admission, they really benefit from placing tips, and tips placement improves survival and is not associated with a high rate of hepatic encephalopathy during follow-up, and we should also have this into consideration when we are deciding, and I hope to have all the data available in the near future. Another contraindication can be portal vein cavernoma, and even if it was an absolute contraindication before, I think that nowadays what we really need is to have a nice study of the vascular tree and to really study the better approach to perform a tips if the patient has a real indication for tips. So, you can do different approaches, not only the transjugular, but also transplenic or transhepatic approach, and I'm showing you when tips place in my institution in a patient with a very extensive thrombosis, what we did was a transplenic approach, and from there, advancing a white wire until the portal vein, and once we're in the portal vein, we can open a balloon, an angioplasty balloon in the portal vein track, and then combining a transjugular approach, we can puncture the angioplasty balloon, and with this technique, we know that we are connecting the two approaches, and after that, we can place the stem and end the procedure, putting a stem from the jugular vein up to, in this case, the splenic vein, because this was a patient with a very extensive portal vein thrombosis and refractory bleeding. Selecting the patient is very important, but another important point to be successful with your tips is when to place the tips, and this is especially important in the setting of the pre-emptive tips, also called early tips, and why it's so important. If we take a look at this, and now all the study from the 80s, evaluating the natural history of patients after abrasion bleeding, you can see how the bleeding and also mortality concentrates in the first days, in the first 72 hours, especially in the first 24, 48 hours, so if we really want to place the tips, a pre-emptive tips with the aim of preventing failure, we have to do that as soon as possible, meaning that if it's possible, it has to be placed in the first 24 hours to prevent the bleeding, and of course mortality, and the same data has been reproduced in the most recent studies, where you can see in this meta-analysis of individual data how still the mortality concentrates at the beginning of the Kaplan-Meier curve. And who has to place the tips? It's the task for who, and I think that the tips have to be placed by the most experienced person that you have, and I really like this American paper evaluating more than 5,000 patients, and what the authors could demonstrate is that those centers placing more than 20 tips a year, it would have an impact on survival, meaning that we should transfer our patients to a center of expertise, and probably it makes no sense that a center that is placing two tips a year takes the risk of doing these techniques if you have next to you another center with more expertise, and I think the tips have to be performed by the people that really have the expertise, and if you are in a center that you do not have expertise or there is no possibility to place the tips, you must pre-establish protocols to transfer the patient as soon as possible to centers of expertise. And another very important tip to be successful is how do we place the tips? How do we proceed with the technique? And just a few things to remind you that it's very important, the technique is very important, and I think that everyone would agree that doing ultrasound-guided puncture is the safest way to proceed, and here I'm showing you a picture at my institution where a combination of the ultrasound with the hemodynamist during the procedure makes things easier and quicker, and the kind of stand that we are placing is also important, and now we are all aware of the covered stand because they are associated with better outcome and also less dysfunction rates, and it's very important where to place the tips, especially in those patients that are candidates for liver transplant, and we do not want to jeopardize the liver transplant. So this is what I'm showing here in this figure is the covered stand that you know that has this uncovered part. So the way of placing it is that this uncovered part has to begin at the entrance of the portal vein into the liver, and in doing this you allow perfusion of the intrapartic branches through the non-covered part, and another important point is to place the proximal end here at the confluence of the hepatic vein and the inferior vena cava vein, and you have to really recover the whole hepatic vein because this is a frequent site of dysfunction, and we want to cover all the vein to make sure that these tips work good during follow-up. So tips placement is very important, not too long, not too far into the confluence, and not too into the atrium, and always consider that you should not make more difficult the work of the surgeon in the setting of the liver transplant. Another important consideration is to minimally, to go back the GPP the minimum as possible, because we don't want to place a stand and over dilate this stand because we know that it's associated with complications and especially hepatic encephalopathy during follow-up, but if we really want to reduce a little bit the pressure, we need to measure them in a very accurate setting and if you look here, you see how sedation, that usually the patients are under general anesthesia or deep sedation where when we are placing the tips, and here you can see how it really interferes with the pressure and sometimes it's too difficult to get a real value, an accurate value, and look here how it is the normal value when the patient is under stable conditions, so make sure that you take your measurement without the impact of sedation in your patient. What we do in Barcelona is like placing the tips and doing a minimal dilatation the day of the stand and then 24-42 hours after, once the patient is more dynamically stable, not intubated anymore, not under vasoactive drugs, then we make a revision of the stand and we adjust the pressure because we know that doing that in this specific setting is the optimal time to measure GPP and we know that if under this condition we reduce the PPG below 12, this is associated with a good outcome. Preventing hypertension complications and a few words about the hepatic encephalopathy, that it's probably the most feared complication of tips and we have now this French trial evaluating the role of rufaximin in preventing hepatic encephalopathy and what the author could show is that starting rufaximin whenever possible two weeks before the tips placement and continuing more than five months can really decrease the probability of having a hepatic encephalopathy and we should consider that in our patients. But then it's not only like placing the stand and putting the patient on rufaximin, patients with tips they deserve a very close follow-up and we have to follow them up very closely in our clinical, in our outpatient office and we should be, especially if the indication was ascites, a close management of the diuretics after the tips is needed and we should, a way of monitoring and detecting a stent dysfunction, it's doing an ultrasound and I'm showing here what is the protocol in my unit, what we do here in Barcelona is after one month after placing the tips what we do is an hemodynamic revision to know that the tips is patent and to see what is the pressure gradient that the patient has at this moment, and then combining with a Doppler ultrasound to see how these hemodynamic findings correlate with the findings in the Doppler ultrasound, and this is going to be the basal value, and we will be comparing during follow-up with this value at one month, and then we will take a look at the velocity in the portal vein, and if the velocity goes down, or there is changes in the direction of the flow, then we will have to suspect a dysfunction and go for a more dynamic revision, but usually the Doppler ultrasound, it's not, and it's really accurate in identifying dysfunction. So, with all that I've said, my tips for successful tips are that good selection of the candidate, you have to really stratify the risk and evaluate the potential contraindication to make sure that you select the right patients for this very life-saving treatment. If you are considering apprehensive tips, you have to do that as early as possible, and all the patients who are admitted with a varicial hemorrhage, they have to be stratified, the risk has to be stratified, and if you cannot place a tip in your center, you should derive in an urgent way the patient to a referral center. The techniques require a lot of planning and making sure that you do not jeopardize a future liver transplant, you're avoiding standover dilatation, it's an important point, and if we are like, if we want to really decrease the portal pressure, because this is what is associated with the better outcome of our patient, we have to measure that very accurately, and those patients, of course, they need a close follow-up, and you have to manage the complications after the tips placement, and also be aware of any sign of dysfunction that you may address, and trying to resolve in a more dynamic revision. So, thank you very much for your attention, and I would like also to thank my multidisciplinary team, because, of course, placing a tip is a multidisciplinary procedure. Thank you very much, and I will be more than happy to take any questions that you may have. Hello, my name is Lauren Nephew, and I am a gastroenterologist and transplant hepatologist at the Indiana University School of Medicine. I'm also a health services researcher, and I study disparities in access for vulnerable populations to treatments for end-stage liver disease, so I'm really excited and honored to be here to share with you all about this important topic, and so we're going to discuss disparities in access and outcomes of emergent EGD for upper GI bleeding, and I have no disclosures. So, our objectives for today will be as following. We will first make sure that we're on the same page about some concepts that we'll talk about, so we'll talk about health equity and health disparity. I'll define those so that we can have a common ground as we move forward to talk about the specific topic of emergent EGD in patients with cirrhosis, and then we'll move on and we'll review the data on disparities in access to emergent EGD in patients with end-stage liver disease, and then finally, we'll talk about potential steps. What do we do when we identify a disparity like this? How can we achieve equity and access to emergent EGD for our patient population? So, let's start off with some concepts, and so I love these drawings. They are simple, but I think they really illustrate the points and the difference between equality and health equity. So, equality says that we should give everyone an equal opportunity to achieve a certain goal, and so you see the little boy in the red jumper on the left of the tree. He's got a ladder and a bunch of apples on his side of the tree, and the tree is actually leaning in his direction, so he's having no problem accessing the things that he needs to access, in this case, the apple, and then you have the little boy in the blue jumper on the right. He's got three apples only on his side of the tree, and the tree is leaning away from him, yet he has the same size ladder as the little boy in the red jumper, and so equality would say that we give everybody the same opportunity to access health care. We open up hospitals to everyone, and then everyone should be able to achieve good health, right? Well, even if he has the same ladder as the little boy with the red jumper, the little boy with the blue jumper is at a disadvantage, right? He doesn't have equal opportunity in his neighborhood. He doesn't have weekend endoscopy. He doesn't have university health systems, and so even though those doors may be open to him, how is he going to ever access them, and so if we're going to move away from talking simply about equality, then we need to think about equity in health care, and so equity says that we create custom tools that address inequity, and so we can see that the tree is bent in the opposite direction for the little boy on the right side of the tree. We can see that he has fewer apples, and so rather than just giving him the same opportunity, we're going to give him a taller ladder. We're going to figure out a solution to the fact that he has very few apples on his side of the tree, and so the goal in health care is to really achieve health equity and try to create custom tools so that our patients can have access to emergent endoscopy and the other treatments that are very specific that we offer, and ultimately, we want to achieve a place of justice, and justice would say the tree is standing upright, and there are equal apples on both sides of the tree, and we no longer have to create custom tools, but for now, let's work towards maybe health equity, so I think it's also in addition to us understanding health equity that we understand and define what a health disparity is, so a health disparity, and this is a definition that I like taken from the American Journal of Medicine. It's not just any health difference. It's a disparity. It's something that's systematic and plausibly avoidable, and so when we're thinking about emergent EGD, it's not just that there's a difference in the procedure. It's that there's a difference that should have been able to be avoided. It's systematic. There's something going on that's not right about this, and that creates a health disparity, so it's not just a generic health difference that happens just to happen. It's something that happens and arises, and it's the setting of often intentional or unintentional discrimination, and so our goal to achieve health equity is to try to eliminate these disparities and create custom tools that lessen the burden of health care disparities. Now, why do we want to focus particularly on access to emergent EGD or access to medical care at all, and the real reason to really focus on this is that health outcomes are determined in large part by having access to medical care. I know many of us, including myself, spend a lot of time focusing on clinical factors. How much blood do they need? How advanced is their liver disease? Did they get bands? Did they have any prophylactic therapy when they arrived, and all of these clinical factors are very important, but it's also equally important to health outcomes to look at some of these other factors, and so this is a figure taken from the CDC website on social determinants of health, and it's really a composite of a decade of research looking at the impact of factors other than clinical factors on health outcomes, and you can see that medical care and the social and structural determinants of health play a large role in how our patients will do, and so it's important that we think about these issues of access if we want our patients to ultimately to do well. Understanding the disparities in access to procedures was even an issue. It really happened in the 90s, and it really happened and began a lot in the cardiac literature, and so this was a landmark study published in the New England Journal. It had over 400,000 patients, and this was a VA study published in the New England Journal looking at cardiovascular disease and chest pain and access to cardiology procedures, and what was nice about this paper, other than it being a very large sample size, was that the VA is diverse. It's a national cohort, and you're able to control for some confounders like insurance status, which is not really an issue in the VA system, practice patterns, which tend to be fairly consistent within the VA, also financial incentives for physicians that may drive whether or not procedures are offered or given, really not an issue in the VA system, and so it's really a nice system to study disparities and really kind of narrow in on race and not some of these other confounders, and so you can look at this table and look at the adjusted odds of procedure use among White compared to Black patients, and you can see that White patients were significantly more likely to obtain all of these advanced procedures, cardiac catheterization, CABG, in all of these different diagnosis groups, even when controlling for coexisting previous conditions, age, diagnosis, and other confounders, and so this work really paved the way for us thinking about in access to these procedures that people need to do well. This report, which is a famous report, is Unequal Treatment. It's the Institute of Medicine report. It came out in 2002. That study from the New England Journal looking at the cardiac procedures was in 1993, but by 2002, the medical community was beginning to realize that there were racial disparities across all of their health care. It wasn't just in cardiology. It wasn't just in GI. This was everywhere, and this report was groundbreaking because not only did it recognize the health disparities were pervasive, but it recognized that while they tended to diminish control for socioeconomic status, they didn't always go away, so controlling for socioeconomic status and access-related factors were not the only reasons for racial disparities in access to procedures or other outcomes, but there was something else going on, and I think it's important for all of us to recognize and to think about that as we try to come up with solutions to access to some of the procedures and care that we offer. So, moving on to talk specifically about the question at hand, so advances in endoscopic and medical management for esophageal variceal bleeding have led to really significant improvement in survival, and so now with, you know, octreotide and antibiotics, people are doing better after variceal bleeds. However, it's important to note in our sickest patients and our child C patients that there still remains approximately a 30 percent mortality from this complication, and it's important to understand that disparities often or these types of outcomes often impact the most vulnerable, and so it's important for us to understand whether or not emergent EGD and outcomes in this patient population might be worse in racial and ethnic minorities. So, when I began to think about this talk, I knew very well of the paper by Lynn published in 2007 addressing this topic, and I said that'll be certainly something that I cover, but I was sure that there must be other data in this area that I just wasn't completely familiar with and that I would find it when I began to prepare. Well, to my surprise, this is the only study in our literature that I could find that discusses this specific question of disparities by race and emergent EGD in patients with cirrhosis, so we'll talk about the study in a bit of detail. So, this was a study that used a national inpatient sample. This is a national data set of hospitals around the country. It allows us to get a representative understanding of inpatient outcomes. They looked at patients from 1998 to 2003, so although the study was published in 2007, the data that they used was a bit older. To be included in this study, you needed to have cirrhosis, and in addition to cirrhosis, you also needed to have a portal hypertensive complication, so these are patients with advanced liver disease, and when we look at this table that describes their presenting characteristics by race and ethnicity for our complication in question, which is variceal bleeding, you can see that about between eight and ten percent of the cohort presented for their inpatient admission with a diagnosis of variceal bleeding. There's no p-values in this table because if you've done work with the national inpatient sample, the sample sizes are so large that any difference is considered significant. I'm not sure that the difference between white, black, and Hispanic patients here is clinically meaningful. There's a difference less than one percent, so in general, eight to ten percent of patients are presenting with this complication, so they asked the question of the patients who are presenting with variceal bleed, what was the rate of endoscopy during their inpatient stay, and they found that it was really similar between the races and ethnicities, so 71.8 percent of white patients underwent endoscopy sometime during their hospital admission, and that was really similar between black, Hispanic, and Asian patients, and that was not statistically significantly different. However, they were also interested in the timing to endoscopy, so they looked at the proportion of patients who were admitted with bleeding who experienced a delay in upper endoscopy, and they considered a delay greater than 24 hours, and you can see in the bar graph on the right of the slide that black patients had a greater likelihood of having delayed endoscopy, with 23 percent of black patients having delayed endoscopy compared to 17 percent of non-Hispanic white patients, and you can see in the other ethnic groups, Hispanic patients and Asian and Pacific Islander patients, and in a group that is classified as other, that there were no significant difference between white patients, so this was a disparity was really only seen in the black cohort. They went on to look at racial disparities in hospital mortality, and they found that the odds of death for black patients compared to white patients was 1.12 for black patients and 0.83 for Hispanic patients, so black patients were significantly more likely to die in the hospital than white patients, and this disparity was not seen in the Hispanic patient population. They wanted to look specifically at that odds of death among people who had variceal bleeding, and the odds ratio was actually higher, so black patients, the odds of death was 1.72 compared to the 1.12 in the whole cohort, so certainly a disparity in inpatient mortality that was worse among the black patients with variceal bleeding, back in the question of whether or not delayed endoscopy played any role in that worse mortality. So this study is, you know, a large study of thousands of patients. The cohort is diverse. It's nationally representative hospitals across our country. I think some of the limitations currently, it's now an old cohort from 1998 to 2003. They don't really control for changes over time, so this is a five-year time period. Presumably, disparities are getting worse or getting less, and really, they just provide us with an aggregate odds ratio, and we really don't get to see how these disparities may be improving or worsening over time. They are unable to control for MELD score. If you're familiar with the data from NIS, it does not include labs. However, they do control for decompensations as well as the severity of a patient's comorbidity by using the Charleston index, and unlike, you know, like most studies, they're observational studies. They're unable to really identify the mechanism, so there's this delay in endoscopy for Black patients, but why is that? That's not something that they can really explore here. It certainly presents an opportunity for future directions. So when we're trying to think about mechanisms for racial disparities and survival, racial disparities and access to procedures, this is a framework provided by the Institute of Medicine report from 2002, and so we recognize that there's a health difference, as you can see by the figure on the left, but what components are contributing to the health difference? There are things that are related to a disparity, and there are things that are not, so clinical appropriateness, need, patient preferences are not really what we're getting at when we're talking about health disparity, right? We're looking for plausibly avoidable systematic issues that may result or come about as a result of bias or discrimination, so we're looking at those bottom two boxes. We're looking for system issues, discrimination, bias issues, and we want to control for some of those top issues. So factors, were they able to really control or think about these issues in the wind paper on the table? On the right, you can see, yes, they do think about clinical appropriateness. All the people in the study had a varicea, bleeding, and cirrhosis, so they were able to address that. Disease severity and patient preference, again, which are not necessarily things that are directly driving a disparity, but are important things to control for. They are able to control for disease severity somewhat by controlling for decompensations. Patient preference, they do control for somewhat in the sense that they look at overall endoscopy over the whole stay, and there's no significant difference by race, which would suggest that Black patients are certainly willing to have an endoscopy. It just doesn't seem to be happening in a timely fashion. Health system factors, they're less able to control for that. They're certainly in control for location, but they can't control for endoscopy availability and other system factors that may lead to disparities that we're seeing, and they're not able to control for any provider bias or any bias anywhere in the healthcare system that may contribute to this difference, and so we really want to think about these health system factors and bias factors when we're thinking about the disparity and how we might intervene there. So to move forward in this important question of access to really critical procedures for our patient population, I think we need to understand if this disparity still exists in a contemporary cohort, so that cohort is almost 20 years old now, and so we need to understand are things getting better or things getting worse? What are the trends? Where are we now on this important question? And then we need prospective studies. If we establish that there indeed is still a disparity, why is it there? What are the mechanisms? Is there bias that we need to identify as physician practice? We do need to explore patient preference. It might not be a reason that kind of gets at the disparity, but it certainly gets at a health difference if patients have less preference to get endoscopy or have to have their endoscopy delayed because they're nervous about it. We want to explore clinical and system barriers to EGD. Do patients of color need more transfusions? Do they have to hold their anticoagulation? Do they have more hyponatremia that has to be managed before they can get to endoscopy? So these are clinical and system barriers that are important to understand, and then once we can do that, we can develop interventions to remove barriers to emerging EGD for our patients. So in conclusion, health disparities are systematic. They're plausibly avoidable health differences that occur according to race or in other disadvantaged groups and demographics that can be now identified across all of healthcare, and removing barriers to access moves us closer to health equity, and that's the goal to get health equity and health justice. Specifically, there appears to be some racial disparities in timing to emerging EGD and survival for Black patients with esophageal varicose oblivion. However, we need more contemporary studies and prospective studies to better understand the mechanism of disparity so that we might develop interventions to eliminate those disparities or at least reduce them. I'd like to thank you for this opportunity to discuss this really important topic. Hi everybody, thanks for joining us. This is session two of the 2021 AASLD-ASGE endoscopy course. It's my pleasure to present the case for this session. This case is a 27-year-old male who has a history of ulcerative colitis on infliximab for that. Recently, due to symptoms of pruritus, he was found to have elevated liver tests. On laboratory evaluation, his alkaline phosphatase was 479, his ALT was 301, his AST was 183, and his total bilirubin was 4.6. He had an MRI MRCP performed. As you can see, he has some intra and also extra hepatic bile duct irregularities. And he also has what appears to be a dominant left hepatic duct stricture, which in this image has the arrowhead with upstream left hepatic duct segmental dilation. With a suspicion for PSC with potential dominant bile duct stricture, the plan was to proceed to ERCP. I'm going to introduce our speakers today. We're very happy and very pleased and fortunate to have Dr. Adam Slivka and Dr. Sumira Ilias with us speaking today. We have three talks in this session. The first talk will be titled Diagnosis, Utilizing Chalangioscopy in Indeterminate Biliary Strictures. And we have Dr. Adam Slivka from UPMC who's presenting that talk. Second talk is titled Beyond Cytology, FISH and NGS for Indeterminate Biliary Strictures. Dr. Sumira Ilias from Mayo Clinic Rochester will be presenting that talk. And finally, I'll be presenting our final talk of this session, which is Infection Prevention in ERCP, Innovations in Duodenoscope Design. We look forward to an excellent session. Hello, everybody. I'd like to thank the organizers of this AASLD and ASGE endoscopy course for inviting me as well as the two societies. My name is Adam Slivka and I'm from the University of Pittsburgh Medical Center. The springboard for this case is a 27-year-old male with PSC and a new dominant stricture on MRCP. And I've been asked to talk about utilizing chalangioscopy in indeterminate strictures. As an outline, I'm going to define and describe the workup of indeterminate biliary strictures. I'm going to review the role of chalangioscopy both as a visual impression, what I'll call the optical biopsy, as well as directing tissue acquisition or other modalities to a stricture, and I'll highlight the difficulties of PSC. I'll use a few cases to illuminate certain points. What is an indeterminate biliary stricture? Well, I think if you think about it as a bile duct stricture without an associated mass on cross-sectional imaging that occurs without a prior biliary intervention, that's a pretty good diagnosis. And really, we can break them down into benign and malignant, and the malignancy is largely chalangiocorsonoma. And chalangiocorsonoma remains very difficult to diagnose because of its desmoplastic nature and the low mitotic rate. When we do tissue sampling during ERCP, all of the techniques we use have been associated with low sensitivity and high specificity. What does that mean? That means there's no false positives if your specificity is 100%. And our opportunities as clinicians are to improve sensitivity. This is very problematic because the nonmalignant diagnosis, including PSC, like in our case for this talk, this session, as well as autoimmune cholangiopathy or fibroinflammatory biliary strictures are often associated with a lot of cellular atypia. So this creates challenges, and the cytopathologists have evolved their criteria over decades such that specificity is unity. If they call cancer, you have cancer, you can proceed with surgery or chemotherapy. However, if they say atypia or negative for cancer, these results demand more tissue, more time, more expense, and the occasional diagnostic surgery, which is really what we're trying to avoid. If sensitivity is unity, cancer can be ruled out definitively, alternative diagnoses and treatments entertained. But we're dealing with a lethal diagnosis. What's acceptable in terms of sensitivity and specificity? We want perfect tests, and we're far from it at this point. If you look at the cartoon on the right, you can see how do we get tissue when we're doing ERCPs, largely with brush cytology you see on the left and little biopsy forceps that you see on the right. There is also an FNA needle that can be used at ERCP, but it's really fallen out of favor. These techniques are all associated with low sensitivity, high specificity, and the more you do, the higher the sensitivity is. So multimodal sampling increased sensitivity. And there's enough data out there now that really forceps are best for bile duct cancer. And again, that's because the desmoplastic nature of the tumor. Some people are tempted to lump highly suspicious cytology along with positive cytology for cancer. And you can get away with that in certain circumstances, but not in PSC. Because of the severe atypia, highly suspicious is not cancer. And the sensitivity is about 50% in our large published series. This is what brush cytology looks like on the left, which is cytology and tissue biopsy, which is histology on the right. Sensitivity in the literature varies from 8% to 60%. And specificity is 96% to 100%. Really should be close to 100%. So what are the strategies we use to improve sensitivity? We talked about multimodal tissue sampling, serum tumor markers, someone will talk about specifically CA99, improvement in the analysis of tissue, largely molecular diagnostics, which will be addressed in the next talk, new imaging modalities and new endoscopic techniques. And I'll focus on cholangioscopy. This was originally designed over 30 years ago as mother and baby scopes that required two operators. And for the past decade, we've been using a digital single operator cholangioscope, which is disposable and costs about $2,500 per use. There are reusable digital cholidocuscopes that have been developed. They've been recalled, they're hard to clean, they're very fragile, and they're back in development. All these scopes are about 10 French in diameter, which makes accessing strictures above strictures typically seen in PSC patients somewhat problematic. How good are we at looking at a stricture directly and saying, oh, this is cancer? Well, malignant features have been identified, scoring systems have been designed, but the validation isn't there yet. And there's huge inter-observer variability, which I'll show you. The above panel A and B is a picture of a bile duct stricture taken with a reusable scope, beautiful large images, you can see narrow band imaging. This particular case, you can see the cystic duct takeoff. There was a stone lodged there, and once the stone was cleared, this stricture was seen. Just looking at it, I would say it looks like a benign stricture compared to the stricture below, which was taken with the single operator disposable cholangioscope, where you see villus features, almost a fish egg appearance, and a little biopsy forceps sticking out there doing biopsies of a villus tumor. This is another case that we did not that long ago, an indeterminate stricture in the mid-bile duct, ERCP, an aberrant right posterior system, which is important for surgical staging. And this is what the stricture looked like. When I first looked at it, I thought it was benign. I thought there might be some enlarged tortuous vessels, which we call tumor vessels. Once we got through the stricture, it looks benign. Overall, I kind of got the sense that it was a benign stricture, and the biopsies just were read as atypical epithelium with abundant inflammation, and the differential diagnosis was between reactive changes and dysplasia. This patient, of course, did have a cancer, which we diagnosed using molecular methods. This is what the tissue that we get looks like, and you can see the denuded biliary epithelium with a lot of inflammation, and you can see some nuclear atypia, but not enough for the pathologist to call cancer. How do we do, as a group, diagnosing cancer with visual changes? If you just look at the top three lines, this was Dr. Yang Chen's series that I participated in back in 2011, a multicenter study with the first iteration of the single operator cholangioscope, and in the category of intrinsic malignancy, which is bile duct cancers, cholangiocarcinomas, when we just looked at the ERCP images, we had a sensitivity of 56%. We would guess what we were looking at. If we looked at the stricture with a cholangioscope, we thought it was cancer, and our sensitivity was 84%, but when we tried to prove it with directed biopsies through the single operator cholangioscope, our sensitivity was only 65%, which is really not adequate. I then got involved with three large studies of these visual characteristics to see how good we are at guessing what's malignant and not, and this is probably the most important slide. The first panel study was put together with the legacy fiber optic single operator cholangioscope. Amrita Sethi at Columbia put this study together. This was just looking at 27 video clips by nine blinded experts, and we used the original classification of ETOI, four features, surface features, vascular features, features of the lesion, and then a guess at the diagnosis. There were 16 benign and 11 malignant cases. Our accuracy was 50%, and if you look at the Kappa statistics for the four criteria, they were awful. They were slight in all cases. Then came digital single operator cholangioscopes, and we got together again. We added five experts. At this time, Michelle Kahela at Cornell took the reins, and we thought, all right, if four criteria didn't work, let's make nine criteria, so we modified ETOI. We had six benign, 15 malignant, and our accuracy went up to 70%, which was a step in the right direction, but again, the Kappa statistics of inter-observer agreement were awful, slight and fair for most of the nine criteria that you see listed on this slide. Just last year, we tried to simplify things. We got together again, looked at 50 video clips using a new classification called the Mendoza criteria, which were now back to five criteria. Unfortunately, we only had nine benign cases and 41 malignant out of the 50, and that may have skewed our results, but we got our accuracy up to 77%, which still isn't acceptable. However, look at our Kappa statistics. We had pretty good inter-observer agreement for all but one of the criteria, and so I think we're beginning to head in the right direction, although I'll admit this may be all a training bias. This is a study comparing fluoroscopic biopsies and cholangioscopic biopsies that I did many years ago, almost a decade ago, and this is a retrospective review. This is not a prospective randomized study. You can stick biopsy forceps right up in the duct. We had 110 specimens from 89 patients, and if you look at the flow diagram, you can see the breakdown of who had some had both fluoroscopic and cholangioscopic, and some had one or the other, and when you look at the results, we had 81 standard biopsies and 29 patients with single operator cholangioscope. We tended to do more fragments because it was a lot easier when we did fluoroscopic. There was a mean of four compared to three, which was a significant difference. Obviously, significantly bigger pieces of tissue. However, insufficient for diagnosis was not significant between cholangioscopic biopsies and standard fluoroscopic biopsies. The breakdown of benign and malignant cases was the same. The sensitivity trended to be better with the larger tissue obtained with fluoroscopic biopsies, but this didn't reach significance. Again, we had 100% specificity in both groups, and the accuracy trended to be better with fluoroscopic biopsies, but that didn't reach significance. So it doesn't seem that cholangioscopes in this large series help with obtaining tissue with biopsies, and this was also confirmed in a more recent study out of the Mayo Clinic that looked at cholangioscopic biopsies in detecting cholangiocarcinoma when compared with cytology in fish, where they thought they did a little better with cholangioscopic biopsies, but you'll see the data is not impressive, but not in patients with PSC, important for this talk. This is a retrospective cohort of 92 patients, including a third of them with PSC. There were 41 cholangiocarcinomas. The overall sensitivity of brushing was 45. When you added fish, it went up to 57. You'll hear about fish in the next talk. When you did fish plus cholangioscopic biopsies, 71%, and fish plus fluoro was 65%. Not a big difference, but a trend towards improvement, but there was no improvement in the cholangioscopic biopsies compared to brush cytology in the PSC population, and importantly, no differences in adverse events. I'm going to spend a few minutes talking about confocal microscopy, which is another way that we can bring a microscope into the patient and evaluate strictures. It came from work in astronomy. We deliver laser light through an optical probe, and we can put that probe either through a cholangioscope or directly through a catheter during ERCPs and do microscopy. When we did a video review, a consensus video review of 89 patients, we had a sensitivity of 98%. We were very excited to be able to detect 39 malignant patients out of 40 in real time. However, the specificity, when you compare it with tissue, was quite low at 67%. We also studied whether we should do cholangioscopy. We put the probe through a catheter in 37 patients and through a cholangioscope in 52, and it really didn't make a difference in terms of the sensitivity or specificity. There was a trend to higher specificity with the cholangioscope that did not reach statistical significance. So much like biopsies, you can do it either way. In order to improve the specificity, we came up with malignant, with benign criteria as opposed to just malignant criteria, and then we validated that in real life with real patients on over 100 patients. What you can see here was our sensitivity decreased, our specificity increased, and the overall accuracy was unchanged. But if you look at the third panel, if we did confocal with ERCP and tissue sampling, we were up close to 90% accuracy. So improving, getting there, but not there. The reason I bring up PCLE is Raj Shah did a spinoff study. He was one of the original investigators, and he did a spinoff study from the University of Colorado. We participated in this study just looking at PSC and dominant strictures. We used the Miami and Paris criteria. We also confirmed malignancy with fish and or cytology and histology. We lumped in high-grade dysplasia in this study, which turned out to be reasonable in this subgroup of patients. There were two technical failures. 63 strictures were evaluated. Seven patients had cholangio, and what you can see is the sensitivity for brushing in PSC is very low. For fish, it's about 60%. For biopsy, it's about 50%, and the sensitivity and specificity for PCLE alone was 85 and 73, which was very similar to the non-PSC patients. So this study is important to show that PCLE does not degrade in the setting of PSC like other modalities have, but it's still not good enough. What are the complications if you decide to do choledoscopy or cholangioscopy? Similar to ERCP. In the Chen study, about 7.5%, there was a suggestion of an increased risk of cholangitis, largely because many of these patients had a stent before the cholangioscopy. They had their biliary tree contaminated, and then when you occlude the stricture with the scope and you're pumping in a liquid to clear your view, you can induce cholangiovenous reflux, but that did not reach statistical significance, and a very large study published at the same time looking at over 2,000 ERCPs and 169 single-operator cholangioscopes found a very similar complication rate of 7.7% for all patients, and single-operator cholangioscopy was not an independent risk factor for complications. I want to show you one case before we close. This is a patient that was referred to us with recurrent cholangitis. He had dilated tortuous cystic ducts up in the left lobe of the liver seen on cross-sectional imaging and on the composite MRCP view, and at the outside ERCP, they were pulling mucus out of his bile duct. In this case, we would not have been able to get biopsy forceps up into the left duct without the use of a choledoscope, which we put up there, and you can see in panel D, there's a mass up there and arising from a villus-appearing biliary epithelium, and the biopsy showed mucinous neoplasm with at least high-grade dysplasia. This patient went for a robotic-assisted left hepatectomy, had an oncocytic adenocarcinoma arising in a biliary IPMN with extensive high-grade dysplasia and a PT1A tumor. So he was cured from this operation. Interestingly, when we did molecular markers on this tumor, we found a very unique fusion abnormality, which had only previously been reported in fibro lamellar HCC. We then went back and looked at all of our biliary oncocytic adenocarcinomas arising in IPMN in the bile duct and in the pancreas, which didn't take the traditional KRAS-GNAS pathways, and we found this fusion in all of them. We published that in Gastro along with colleagues at Mayo Clinic and Johns Hopkins. So, in summary, colidocoscopy can be a useful adjunct to evaluate indeterminate biliary strictures. It is safe to perform. Optical diagnosis of cancer is improving, but is subject to significant inter-observer variability. It can be used to direct biopsies, but it's not required, and its role in PSC is really limited and should be determined on a case-by-case basis. I want to thank you for your time, and I hope you have a great meeting. Hello, everyone. My name is Samara Elias. I'm a transplant hepatologist at the Mayo Clinic in Rochester, Minnesota, and it's my pleasure today to give this talk titled Beyond Cytology, FISH, and NGS for Indeterminate Biliary Strictures. I have no disclosures relevant to this talk. The learning objectives are to review causes of indeterminate biliary strictures, review the management of high-grade biliary stricture in the absence of PSC, including the role of cytology and FISH, review management of high-grade biliary strictures in a PSC patient, and then finish off with emerging diagnostic modalities including NGS and cell-free DNA. So is this stricture benign or malignant? Well, that's the million-dollar question, and that is why it is important to have diagnostic modalities that can help us distinguish between the two. Before I review the role of these emerging diagnostic modalities, I would like to review indeterminate biliary strictures and some of the notable etiologies that we encounter in our practice, as that will help provide a framework for understanding the current modalities and their potential limitations and the benefits of some of the emerging modalities that they may have. So what are indeterminate biliary strictures? These are strictures in which testing is non-diagnostic. Majority of indeterminate biliary strictures are malignant. However, up to 30% are benign, and the distinction between the two is important as it clearly impacts management since there is a risk of potentially missing a malignant stricture at an early stage. Alternatively, there is a risk of resecting a benign stricture. Biliary strictures can be classified into four broad categories. The most frequent benign causes are iatrogenic and are related to bile duct injury that may occur during a polycystectomy or a liver transplant. Vascular causes include ischemic insults to the biliary tree. Since the bile ducts receive their blood supply from the hepatic artery, in contrast to the liver, which has a dual blood supply, hepatic artery thrombosis tends to have a greater impact on the bile ducts and can result in ischemic flangeopathy. Amongst the inflammatory causes, the notable ones that pose a diagnostic challenge are primary sclerosing cholangitis and IgG4-related sclerosing cholangitis. So I will spend some time going over how to distinguish a stricture that occurs in the setting of these two and the role that the current and emerging diagnostic modalities play in that distinction. While the differential of benign causes is broad, the primary differential for malignant biliary strictures is flangiocarcinoma, and in the case of a distal stricture, pancreatic adenocarcinoma should be considered. The current diagnostic armamentarium for a biliary stricture includes cross-sectional imaging. So on an MRI-MRCP, it can help answer the following questions. Is the stricture isolated? Is there a mass lesion? Is there a lobar atrophy? Is there a vascular encasement? Any of these features would be concerning for a malignancy. In terms of serum tests, an IgG4 level and a carbohydrate antigen 19-9 or CA-19-9 level can be helpful, but there are instances we have to exercise caution when interpreting these levels, and I'll review that a little bit later on. Of course, ERC plays an essential role not only from a therapeutic standpoint, but also diagnostic as it allows for acquisition of biliary brushing for cytology and other testing. PSC is a fibroinflammatory disorder of the bile ducts that is often associated with inflammatory bowel disease. The lifetime risk of development of flangiocarcinoma in PSC patients is about 10%, and PSC strictures tend to be multifocal rather than isolated, so that can be helpful when we're dealing with an isolated biliary stricture. What do we do when we do have an isolated, asymptomatic biliary tract stricture? There are two main questions that should be asked. Is it malignant, and is it IgG4-related? Diagnosing IgG4 sclerosing cholangitis is based on a number of criteria, including histology, imaging, serology with elevated IgG4 levels, and involvement of other organs. Serology with elevated IgG4 levels, and involvement of other organs, including the pancreas, retroperitoneal fibrosis, and response to corticosteroids. Despite these criteria, diagnosis can still be challenging, and it requires an index of suspicion. So, this is a case of a 46-year-old gentleman who had jaundice with spontaneous resolution, MRI without a mask, CN-99 was normal, cytology and fish were negative for malignancy, and the IgG4 level was undetectable. A trial of steroids that led to resolution of the stricture, as you can see on the right-hand side of your screen. So, it's important to remember that IgG4 levels can be normal in approximately 30% of patients with IgG4 disease. On the other hand, IgG4 levels can be elevated in the absence of IgG4 disease. So, for instance, IgG4 levels can be elevated in approximately 10% of patients with PSC, and they can also be elevated in cholangiopersonoma. So, what about a high-grade biliary stricture that is IgG4 unrelated? If there's a high-grade stricture, the primary question is whether it is malignant, and the workup to distinguish that will depend upon whether there is underlying PSC. High-grade strictures can occur commonly in PSC in the absence of malignancy. However, in patients without underlying PSC or any other obvious risk factor, the odds are that a high-grade stricture is likely malignant. Therefore, a higher index of suspicion is needed to diagnose malignancy in the absence of PSC, and accordingly, the diagnostic algorithms are different for the two scenarios. So, as I mentioned, you have the higher index of suspicion for malignancy, the diagnostic algorithm for high-grade stricture in a non-PSC patient. It is different. It is simpler. And so, I'll come back to this algorithm once we have reviewed some of these diagnostic tests. This is a proposed algorithm for management of a high-grade stricture in a PSC patient. This algorithm is from the forthcoming ASLB-PSC-CCA guidelines. It's preliminary as these have not yet been approved by the governing board. Now, as you can see, in the setting of PSC, the diagnosis and decision-making and the algorithm is more complex. In either case, whether this is PSC-related or non-PSC-related, ERC plays an important role as it allows for therapeutic intervention as well as the acquisition of biliary brush strings for cytologic diagnosis. Now, before I review these algorithms in detail, let's review some of the currently available diagnostic modalities, primarily cytology and FISH, that we employ and their role in these algorithms. There are four possible results when biliary cytology is obtained. Atypical cytology is frequently encountered in PSC and by itself should not raise concern. This is primarily due to the presence of biliary inflammation. Suspicious cytology is also not diagnostic of malignancy. However, it is a concerning finding as approximately 30 to 40 percent of PSC patients without a mass lesion who have suspicious cytology may ultimately be diagnosed with CCA. For a non-PSC patient with suspicious cytology, the index of suspicion for malignancy should be much higher as they typically do not have this confounding underlying inflammation. Positive cytology is diagnostic of CCA. The chief disadvantage of biliary cytology is its limited sensitivity, 10 to 40 percent across different studies, and the potential for false negative results. Factors contributing to this include the desmoplastic, post-cellular nature of CCAs, which can reside in areas that are difficult to access. Therefore, the absence of a positive cytology does not rule out malignancy. Moreover, the presence of inflammation secondary to PSC or associated infection can pose a dilemma on cytologic evaluation as reactive cells can mimic cancer cells. The primary advantage of biliary cytology is its high specificity, close to 100 percent, and so it remains the gold standard. The other cytologic technique employed in the diagnosis of pancreatic biliary malignancy is fluorescence in situ hybridization, or FISH. This method uses fluorescently labeled DNA probes that hybridize to one of four chromosomal locations, which you can see on the left-hand side of the screen. FISH results can be categorized as follows, normal or disomy or FISH polysomy, which indicates the presence of four or more epithelial cells with two or more signals in two or more of those four loci on the left-hand side of your screen. FISH polysomy indicates duplication of more than one chromosome, and it is a marker of chromosomal instability, which is a hallmark of cancer. FISH has enhanced sensitivity and similar specificity for CCA diagnosis compared to conventional cytology. This was shown in a study published in Gastroenterology in 2015 that employed a pancreatic biliary-specific FISH probe set that had the four loci that I showed on the prior slide, and the sensitivity of FISH for detection of malignancy was 65% compared to 18% for conventional cytology. It is important to interpret FISH results in the context of each patient scenario, particularly for PSB patients. Factors that should be considered include serial or multifocal polysomy, presence of suspicious cytology, and elevated CNNT9 levels. I'll review these in greater detail in the next few slides. CNNT9 at a cutoff level of 129 had a sensitivity of close to 80% and a specificity of about 99% for diagnosis of CCA and PSB patients. This was shown in a study published in 2005. In a subsequent study, when a serum CCA19 level of greater than 129 was combined with the presence of FISH polysomy, this was predictive of cancer. In this study, which you can see on the right-hand side of your screen, patients who had this combination all went on to develop cancer, the majority within two years. When interpreting CNNT9 levels, it's important to remember that there are a number of causes besides malignancy that can cause an elevation of CNNT9 in PSB patients. These include jaundice and cholangitis, and levels can normalize after stenting. So let's come back to the diagnostic algorithm for a high-grade biliary restrictor in a patient without underlying PSB. As I indicated earlier, this algorithm is simpler than the one for PSC patients, because the index of suspicion for malignancy in this context is higher. So let's look at the extremes first. So biopsy or cytology are positive, that confirms the diagnosis of CCA. On the other end of the spectrum, if biopsy, cytology, or FISH are negative, given the higher index of suspicion for malignancy in this setting, a closer follow-up with an ERCP with brushings in three months is warranted. If cytology is suspicious or FISH polysomy is present, then this is probable CCA, or if the CNNT9 level is greater than 129 in the absence of unstented jaundice, then it's also probable CCA. Similar to a non-PSC patient, positive biopsy or cytology confirms the diagnosis of CCA in a PSC patient with a high-grade restrictor. However, on the other end of the spectrum, since the index of suspicion is lower in this setting, if biopsy or cytology or FISH are negative, then we would recommend MRI, MRCP in 6 to 12 months rather than the shorter follow-up for non-PSC patients in the algorithm on the prior slide. Now coming to the middle scenarios, well, these are a bit more challenging. If cytology is negative, but FISH polysomy is present, ERC should be repeated in three months. If there is serial FISH polysomy, then this is probable CCA. If there is serial FISH polysomy, then this is probable CCA. So let's look at the data behind serial polysomy and risk of CC and PSC patients, as that will help provide a rationale behind that recommendation. So in a retrospective study of PSC patients who underwent ERCP with brushings, 9 out of 13 patients, so 69% with a subsequent polysomy result, which we would refer to as serial polysomy, they were diagnosed with CCA compared with only 3 of 17 patients with subsequent non-polysomy. The three-year cumulative incidence of CCA was 75% among those who had serial polysomy. So this just reinforces the importance of confirming polysomy over time, particularly if FISH polysomy is found in the absence of other features of CCA in a PSC patient. Multifocal polysomy is also concerning. So this is polysomy that is found in multiple areas of the biliary tree that may have been brushed. In a study of 371 PSC patients, those who had multifocal polysomy were more likely to be diagnosed with CCA in contrast to those who had either normal FISH or unifocal polysomy. And compared with patients with unifocal polysomy, those with multifocal polysomy were more likely to have suspicious cytology and develop serial polysomy. If a PSC patient with a high-grade stricture has suspicious cytology and FISH is normal, so that's the blue box to our right, then ERC should be repeated in three months. If FISH remains negative, then follow-up in a six-month period would be reasonable as suspicious cytology by itself is not diagnostic of malignancy. On the other hand, suspicious cytology plus FISH polysomy is probable CCA, so that's the box to our left. So let's look at the data behind this proposed guidance. In a study of 102 PSC patients without a mass lesion who underwent ERC with brushings for cytology, only 40% of patients with suspicious cytology developed cancer. However, patients who had suspicious cytology and FISH polysomy, the survival was significantly lower compared to the ones with suspicious cytology alone. So these data indicate that although suspicious cytology by itself doesn't diagnose malignancy, the presence of suspicious cytology with FISH polysomy is a probable malignancy, particularly if the FISH is confirmed over time. So in the prior study that I showed with serial polysomy, 11 of 18 patients with FISH polysomy who also had suspicious cytology went on to be diagnosed with cancer, and of the ones with serial FISH polysomy and suspicious cytology, 80% went on to be diagnosed with cancer. So moving on to emerging diagnostic modalities, these include next-generation sequencing or NGS, cell-free DNA and circulating tumor cells, extracellular vesicles and bile, amongst others. Next-generation sequencing is massive parallel sequencing that allows probing of large panels of genes to identify relatively rare mutations that may be present in a small fraction of DNA templates. There are a few studies of NGS on bile duct brushings. In this study, NGS was conducted on post-cytocentrifuge bile duct brushings from 100 consecutive patients who had brushings. 43 of these ultimately turned out to have malignancy, and NGS revealed oncogenic alterations in a majority of these patients with the most common alterations occurring in KRAFs and P53. When combined with cytology, NGS improved the sensitivity to 93% while maintaining specificity. The largest study to look at NGS in biliary blood brushings employed an NGS assay known as BILI-seq. This had a panel of 28 genes that are commonly altered in biliary tract malignancies. The diagnostic performance of BILI-seq was assessed in a prospective evaluation of 252 patients with bile duct strictures. In the overall cohort, the sensitivity and specificity of BILI-seq was 73% and 100% respectively. The combination of pathological evaluation and BILI-seq testing increased the sensitivity to 83% while maintaining the specificity. Therapeutically relevant genomic alterations were identified in 8% of patients. In patients with PSC, BILI-seq had an 83% sensitivity as compared with pathological evaluation, which had a sensitivity of only 8%. These results indicate that this is a promising approach. However, it remains to be seen how this compares to FISH and what role it will play in the diagnosis of biliary strictures and whether they're malignant or not in the clinical practice. There is also emerging data on cell-free DNA methylated markers in biliary brushings for detection of malignancy. In a clinical pilot, methylated DNA markers were blindly assayed on DNA extracted from archived brushing specimens from 12 patients with perihelior CCA, 4 patients with distal CCA, and as well as 18 controls. Following a biological validation, as well as 18 controls. Following a biological validation, candidate differentially methylated regions or DMRs were selected. These are shown here on your right-hand side. Across all the cases and controls, at one of these DMRs, EMX1 had a sensitivity of 100% and specificity of 89% for diagnosis of CCA. These are promising, albeit preliminary, results and they need to be validated in larger cohorts. So, key takeaways. Majority of indeterminate biliary strictures are malignant. A higher index of suspicion for diagnosis of malignancy is needed in patients with a PSC-unrelated biliary stricture. FISH has enhanced sensitivity and similar specificity for diagnosis of malignancy compared to conventional cytology. However, FISH results do need to be interpreted in the context of each patient scenario and factors such as CA-19-9 levels, the presence of suspicious cytology, the presence of serial polysomy, which is FISH polysomy confirmed over time. These all need to be taken into account. Emerging diagnostic modalities such as next-generation sequencing and cell-free DNA, they both promise. However, it remains to be seen what role they will play in the diagnosis of biliary strictures in the clinical practice. Thank you very much for your time. Hello, I'm Mark Gromsky. Today, I'll be talking with you all about infection prevention in ERCP, innovation in duodenoscope design. Disclosures are, I'm a consultant from Boston Scientific. Agenda today, we'll review the problem of infection transmission through duodenoscopes. We'll discuss the efficacy of high-level reprocessing, and we'll also discuss new advances in duodenoscope design to address infection prevention. And finally, we'll propose pathways for introducing novel duodenoscopes into your practice. First of all, we have to talk about the risks of ERCP in relation to infection. So, infection is a risk that everybody discusses with patients when they're consenting them for ERCP. It's a known risk, and it has been for decades. But we should really look at it further into two different categories. One is procedure-related, and the second is device-related. So, the procedure-related infection risks are the traditional infection risks that we thought about for many, many years. This is the risk of cholangitis or bacteremia or even sepsis from the procedure of ERCP itself. And this is incumbent on the intervention that we're doing, which is injecting contrast, interpreting, and performing therapeutic maneuvers within the biliary tract. It has to do, likely, with pressurized injection of contrast into the intrapedic ducts, and certainly incompletely drained biliary segments contribute to the risk of cholangitis and also bacteremia. Based on prior studies, there's about a 1 to 2 percent risk of cholangitis from an ERCP procedure. And cholecystitis, particularly when the cystic duct is occluded and bacteremia, are also known risks. Strategies to prevent this complication include using sterile radiographic contrast, limiting the volume of injection to the minimum necessary to achieve the diagnostic and therapeutic goals, and restoring bile flow as able as possible in the case of obstruction. And then in patients where we suspect there may be incomplete bile drainage or other appropriate patients, we should use antibiotic prophylaxis and some antibiotics after the procedure in select patients. So, that's the first group. The second group is a newer category that we're just starting to unravel more now, although we've known it for some time, and this is device-related. And so, this is really related to infection transmission from a contaminated tool, most often the duodenoscope itself, to the patient. And so, that's what I've been asked to talk about today is the problem of duodenoscope-associated infection. First of all, we have to know that all GI scopes are present in a dirty environment. The lumen of the GI tract is not sterile. And thus, for every GI procedure, there are commensal bacteria present. ERCP also commonly intervenes upon infectious milieu, and there's bacteria in the setting of acute cholangitis, biofilms on pre-existing biliary stents, etc. And we're intervening upon that and oftentimes making maneuvers with the scope and dragging things through the scope from that milieu. And we know that the potential for transmission of pathogenic organisms is a problem that's inherent of the current duodenoscope design. Multiple reports in the last number of years of drug-resistant pathogenic organisms that were transmitted via duodenoscopes have been published and publicized, and many of those were in association with outbreaks. I bring this table up just to show you, this is an old study from about 20 years ago, and it was a study to determine the bacteria present in the bile in patients suspected to have cholangitis and with indwelling biliary stents. And I just want to direct you not to the nitty gritty of the table, but just to the breadth of all of the organisms that grow out in patients who have cholangitis and who have biliary stents. As you can see, there are multiple gram-negative, gram-positive bacteria that have been isolated in the setting of ERCP. This, I show you a duodenoscope, and it's just a standard duodenoscope from our reusable fleet. And the reason I show it to you is to show you the elevator mechanism, that metallic lever on the duodenoscope is thought to be likely a source of many of the infection and contamination problems of duodenoscope. And as you can see, there are ridges, and also within the plastic, there's some dents and some scratches. And so we put duodenoscopes through a lot of work and we expect them to perform well every time. But some of the components, including the elevator mechanism and including the distal cap, likely put it at a disadvantage in cleaning it and reprocessing it in the traditional sense. I bring up some of the background of the history of contamination and infection transmission of duodenoscopes. In the last five to six years, there's been a lot of press about it, but actually this is a problem that was identified in Spurps and Starts many years ago. Here's a paper from 2004 that demonstrates a potential infection of Pseudomonas from an outbreak of duodenoscopes. And this is an early study from 1987 published in Gastroenterology that describes an infection resulting from the use of a contaminated duodenoscope. It's not a new problem, although the attention to it is new. And this is really what brought a lot of the press to this. 2014, 2015, we had a series of publications and communications in the medical press, publications and also the lay press regarding outbreaks of infection transmission from duodenoscopes. This is one that was published in JAMA. And then later in 2015, another paper, this one from Virginia Mason recording an outbreak associated with ERCP. Then in 2019, there was a New York Times article that really brought this into perspective amongst not only medical practitioners, but the public as well, and prompted an increased amount of pressure on the societies, on the physicians, on the reprocessing of these scopes. And also put into perspective, although it wasn't a perfectly researched and perfectly written article, it did put this into a broader sense. There's been an incredible amount of FDA communications, probably more so with regards to duodenoscopes than any other medical device. And the FDA has had a number of communications and other regulators on this topic. In 2013, as you can see in the upper left-hand corner, the CDC did alert the FDA to the potential association between multidrug-resistant bacteria and duodenoscopes. In August of 2015, there was an FDA safety communication that offered supplemental measures to enhance duodenoscope reprocessing. And this included culturing duodenoscopes after reprocessing, adding multiple or double high-level disinfection or liquid chemical sterilization or ethylene oxide sterilization as a supplemental measure to clean duodenoscopes. In October of 2015, the FDA ordered each of the U.S. reusable duodenoscope manufacturers to conduct mandatory post-marketing surveillance studies. Again, in February of 2018, there were a multi-society, the FDA, the CDC, and the ASM, offered standardized protocols for duodenoscope surveillance sampling and culturing. And then as you see in March of 2018, there were warning letters that stated that the post-marketing surveillance studies were not accruing adequate numbers. In December of 2018, there was an interim communication of those results. Again, in April of 2019, and then finally in August of 2019, there was a communication from the FDA that had some language of guidance to eventually transition to duodenoscopes that had innovative designs to combat infection risks. So one question we should ask is what is the goal of duodenoscope reprocessing? For reusable duodenoscopes, do we think that the rate of post-reprocessing surveillance cultures will be zero? Is that the goal? The goal has not been clearly defined. I think a goal of eliminating transmission of infection is certainly one, but that's a different animal and it's a very difficult thing to study in a rare circumstance like a transmission of an infection from a duodenoscope. I just want to draw attention to a study from our institution that was published this year where we prospectively evaluated duodenoscope reprocessing for two different of those heightened reprocessing methods. One was double high-level disinfection and the second was liquid chemical sterilization. The reason I bring this to attention is because over the course of the last four to five years, we've spent an incredible amount of time and effort and energy to try to improve our reprocessing methods at our institution. And we found this as an opportunity to study this in what we could consider a best case scenario where for a high-volume center, we do approximately 3,000 ERCPs a year. We have a dedicated staff that cleans duodenoscopes and we have an infection prevention nurse and we have a surveillance mechanism within our institution. So what we did is we prospectively segregated and randomized the scopes themselves for the duration of the study period to either get double high-level disinfection or liquid chemical sterilization. The details of the study aren't necessarily important. We had about 878 cultures that were drawn from those two groups. What I want to draw to your attention is that there are positive post reprocessing cultures in 1.9% of the cases. In double high-level disinfection, there were eight and liquid chemical sterilization, there were nine. There was no significant difference between the groups. Of interest, the high concern or the potential pathogenic organisms were even lower, 0.4% to 0.5%. And so in a best case scenario where an institution is very high volume, has specialized personnel, has been working hard to develop over time strategies and protocols to deal with this, these data show that we can likely get the post reprocessing contamination rates under 1% for high concern organisms. The next question that needs to be answered is, is it possible to completely eliminate this or completely eliminate the transmission of infection for reusable scopes? Next, I want to talk about what innovations have been done in the space of the duodenoscopes to address infection prevention. So as I said, in August of 2019, the FDA did recommend the adoption of innovative duodenoscope designs, likely with disposable components to address infection prevention. Innovative duodenoscopes designed present today include disposable distal tip attachments on reusable scopes or fully disposable duodenoscopes. Currently, in today's environment, there are six duodenoscopes that meet this definition. There are two from Pentax, one of which has a disposable end cap, and the other of which has a disposable elevator cap, which actually contains the majority of the elevator mechanism in the disposable cap. There's a Olympus product, the TJF Q190, that has a disposable end cap. The Fuji offering also has a disposable end cap. So there are four offerings with reusable scopes that have disposable end cap components. And then there are two scopes that are fully disposable scopes, which are you use once and you dispose of them. One, the Exalt Model D from Boston Scientific, and two, the A-scope Duodeno from Amdo. A few pictures of the disposable tip components. These are the three offerings that are reusable scopes and they have disposable end cap attachments. And as you can see, those distal attachments on the left, the middle, and the right, after the procedure is performed, those are stripped off and then the reusable scope is sent to the reprocessing technicians to clean. This is the product that has the elevator mechanism in the disposable cap. And then this is the Exalt Model D, and this is the A-scope Duodeno. These are the two fully disposable Duodeno scopes. Of note, we should mention that cost is an issue when considering new technology. And so there are some things that people should be aware if they are adopting the single use Duodeno scopes. There's a pass-through payment or a pass-through code that was effective starting July of 2020. And this is used to bill for single use Duodeno scopes for patients who are Medicare patients in the hospital outpatient setting. And so using that code, there will be an additional payment on top of the usual outpatient ERCP to the facility. And the idea behind this is it is meant to help to offset the cost of adopting that technology. Then just recently, there's a new technology add-on payment for inpatient Medicare beneficiaries. And this is effective of October 1st, 2021. And it provides additional payment on top of the DRG payment for the hospital inpatient admission. And those listed at the bottom are those codes. And so these are two codes which should be used by your institution if you're using a single use Duodeno scope. Now in the last one to two years, all six of these products have come out onto the market. And there's been a flourish and a flurry of innovation. And I'll have to say that the innovation has likely outpaced the data that we have for those. And so although there has been adoption of these innovative Duodeno scopes, we don't have the large volume of data that we have for ERCP using reusable Duodeno scopes, the same Duodeno scopes that we've been using for many, many years. And so we do have some unanswered questions. First of all, what's the technical performance of these innovative Duodeno scopes? We want to know for sure that there's no change in the complication profile, such as pancreatitis, using the new Duodeno scopes. What are the learning curve of the innovative Duodeno scopes? Although the reusable innovative scopes are pretty much just a modification of the usual scopes, the single use scopes are pretty much a new class. And so it may take some getting used to. We don't have data yet on the learning curve. For the reusable scopes that have the single use or disposable tip components, we still need data on the risk of transmission of infection. That's really what we're trying to prevent. Cost analysis will be informative and necessary to see if it's possible for institutions to adopt this technology. And environmental effect is also something that should be considered. In the last couple of minutes, I want to discuss how a practice should consider introducing novel Duodeno scopes. First of all, it should be thoughtful and it should be objective. Considerations that will likely go into the calculus include, have you had prior infection transmission events before? What's your current state of reusable reprocessing strategies for Duodeno scope? Do you use single high-level disinfection, double, ETO, etc.? How well are your technicians cleaning the scopes? What's the turnover of your technicians? And what type of infection prevention monitoring or screening do you have? What's the volume? Do you just do straightforward ERCPs or do you do complex ERCPs? Are you a referral center or a community center? Do you have transplant patients? Do you not? That all goes into the calculus. What's the experience of the ERCP doctors? And how many Duodeno scopes do you have in your fleet? Currently, we believe that probably an all-in model of any one new Duodeno scope design is probably not ideal. Why is that? Because we don't have the data and the experience yet. So, we would hate to invest an enormous amount of resources in something that hasn't necessarily been borne out. We've employed a staged integration plan where we use both old model reusable scopes, but also we're integrating the newer model innovative reusable and also single-use scopes. Data collection is important. How well are the new scopes performing? How well do your doctors like them? Are there any issues that are coming up? This is important to track. For us, we've created a guidance document that helps our endoscopists and our administration know when we're using certain types of scopes and to track that. And it's important to work with your revenue services to ensure that you're collecting for any of those additional payments you may be eligible for. If you do adopt single-use, likely an algorithm would be important for you to choose which ones to use in which settings. And so, the two groups that we've targeted with our institution is, number one, the patients that are at risk of having multi-drug resistant organisms in their pancreatobiliary tree. And number two, patients at risk for significant morbidity or mortality if they were to contract an infection. So, basically, patients that may put your duodenoscope at risk of becoming contaminated with bad bugs. And number two, patients that wouldn't do well if they came in contact with a scope. So, we came up with a three-tiered system for single-use. Top tier is recommended use of a fully disposable scope. Second tier is that we should consider, but not necessarily recommend, the use of a fully disposable single-use scope. And the third tier, where we discourage the use of a single-use scope. This is the algorithm that we came up with, which was recently published in GIE. And in the top tier, we recommend the use of single-use in patients that have had prior CRE infection anywhere in the blood, the urine, or the bile. Also, patients who have had a resistant organism in the bile on prior bile culture. We oftentimes, in patients with cholangitis, aspirate the bile to see what is in their biliary tree. And then patients who have had recurrent cholangitis, recurrent bouts of antibiotics, maybe harboring resistant organisms such as PSC, secondary sclerosing, cholangitis, et cetera. All of those patients, we recommend the use of a single-use scope in our institution. The second tier is where they're at risk of maybe not doing well if they come in contact with a bad bug, likely because of immune suppression. So these patients include the transplant patients, patients on immune suppression, patients on chemotherapy for malignancy or neutropenia, et cetera. And so in those patients, we recommend that the endoscopist consider whether it's in the best interest based on their suspected ERCP to use a single-use duodenoscope or not. And then probably the largest group, tier three, all other routine cases, in those cases, we're recommending that we use a reusable duodenoscope. So in summary, transmission of pathogenic organisms from duodenoscopes is a problem. It's a rare but a serious issue. Duodenoscope reprocessing is a complex task, but it's an important task. And based on the study that I showed you, there's unlikely to be any difference between the heightened reprocessing of double high-level disinfection and liquid chemical sterilization. The FDA has recommended the adoption of novel innovative duodenoscopes. There have been a number of novel innovative duodenoscopes designed and on the market. There are many questions that are left unanswered with regards to them, and we look forward to a lot of research in this area in the future. Likely a hybrid model where we're using some reusable and some single-use duodenoscopes may fit most moderate and high volume centers currently. I would like to thank everyone for your opportunity, your attention, and this is our team here at Indiana University. Have a great day. Hello, everybody. My name is Jawad Ahmad, and I'm from the ICANN School of Medicine at Mount Sinai in New York, and it's my distinct pleasure on behalf of my co-course directors, Jennifer Christie and Mark Gromsky, to introduce you to session three of the 2021 AASLD and ASGE endoscopy course. This session is entitled NAFLD, obesity and advanced fibrosis, surgical and endoscopic options. I have no disclosures. As you've seen in the prior sessions, we've made this case-based, and so this is a 49-year-old female with a history of NAFLD, diabetes and hypertension, and you see her for evaluation of abnormal LFTs. Her exam is notable for a BMI of 47 kilograms per meter squared, and routine laboratory evaluation shows an ALT of 85, alkaline phosphatase of 138, total willirubin of 1.4, INR 1.3, platelet count of 123, and a creatinine of 0.8. She's undergone ultrasound that shows a coarsened 18 centimeter liver, heterogeneous in appearance, 13 centimeter spleen, but no evidence of hepatocellular carcinoma. Upper GI endoscopy shows no evidence of esophageal or gastric varices, and her fibroscan is shown for you with an E of 13.5 kilopascals, so that's equivalent to F3 or F4 fibrosis, and her CAT score is also elevated at 324, indicative of significant steatosis. So to discuss this case, we have three very eminent speakers. So Bashir Taouli from the ICANN School of Medicine in Mount Sinai is going to discuss assessment of portal hypertension, invasive versus non-invasive approaches, and then Dr. Abudae from Mayo Clinic is going to discuss gastric balloons in naphrodi and cirrhosis. Are we ready for prime time? And then to end the session, Dr. Julie Heimbach, also from the Mayo Clinic, is going to discuss surgical approaches to naphrodi and cirrhosis before, during, or after liver cancer augmentation. Thank you. Hello, my name is Bashir Taouli, and I would like to discuss assessment of portal hypertension, invasive versus non-invasive approaches at the ASLD 2021 virtual meeting. Financial disclosures. So I would like to discuss first some background on portal hypertension, and then I would like to briefly discuss invasive diagnosis with hepatic venous pressure gradient measurement and endoscopy, and then I will spend more time on non-invasive methods, mostly allosography methods. So, by way of a background, portal hypertension is a frequent complication of chronic liver disease and cirrhosis. Although there are less common etiologies described, such as non-cirrhotic portal hypertension, extra-hepatic portal venous occlusion. Most patients with part hypertension have background liver disease. Pathophysiology is based on a combination of fibrosis, vascular modifications and hepatic cell activation, which increase vascular resistance at the sinusoidal level in the liver, which eventually has a downstream hemodynamic effect on the vasculature, which may create increased pressure in the splint-like circulation, including the portal vein and the splenic vein. And this will eventually lead to the hyperdynamic circulation and the development of portal systemic collaterals, which may increase the risk of GI variceal bleeding. Complications of portal hypertension are feared, such as the formation of varices and bleeding, which may lead to death, hepatic encephalopathy, ascites, hydrothorax, and other complications such as hepatorenal and hepatopulmonary syndrome, et cetera. It is very important to diagnose portal hypertension early on so that it can be treated and complications can be eventually prevented so that patient can eventually undergo transplant or other therapies. Invasive diagnosis of portal hypertension is typically based on indirect portal pressure measurements or hepatic venous pressure gradient measurement, which is typically done using a catheter through the IJ, the interal jugular vein that goes through all the way down to the right hepatic vein where you can do first a conventional venogram, and then you can measure the pressure, free pressure, and then watch pressure using a balloon placed in the right hepatic vein. And the pressure gradient represents the indirect measure of portal pressure. Typically, normal pressure should be less than five millimeters of mercury. Portal hypertension is defined as a pressure gradient over five millimeters mercury. Clinically significant portal hypertension over equal 10 carries a risk of esophageal varices, ascites, and decompensation. And severe portal hypertension defined by pressure over equal 12 carries a very high risk of bleeding. It is a very, not very, it's an invasive method that is not widely available. The other method to diagnose and directly portal hypertension is endoscopy, typically upper GI endoscopy, which may screen for varices and also grade them as small or large depending on the size, less or equal, over equal five millimeters. Also provides a risk of bleeding depending on the size and the aspect or the appearance of the varix. And also it is expensive and relatively invasive. So overall, these two methods can be eventually either avoided or performed only in patients who really need them. There's a, unfortunately, still a large number of patients who undergo unnecessary endoscopies showing no varices. These can be eventually avoided by combining non-invasive methods, which we'll discuss later on. Blood tests have been described in diagnosis of psoriasis and portal hypertension. Platelet count is used clinically, especially when combined with liver stiffness. It may have value in deciding whether a patient's need or not endoscopy. Other blood markers such as APRE, AST to ALT ratio and fibro tests, which have been validated in liver fibers staging services have not been validated in portal hypertension yet. Standard imaging may be used to at least diagnose and eventually assess severity of portal hypertension. You can look at a spinomegaly defined by chronic caudal size over 13cm typically measured with a CT or MRI. This is a sensitive finding, but not specific. And you can also have patients with portal hypertension without spinomegaly, process of ascites, and also partial systemic shunts. So the combination of all three have been described in a scoring system published in 2014 by Iren Manesh, which provided reasonable performance. As you can see in the table, the composite, I'm sorry, the Iren Manesh score provided reasonable performance of 0.77 for clinically significant portal hypertension and for portal hypertension. We have developed a simpler score at Mount Sinai by combining assessment of ascites, spondylic size and number of portal systemic or variceal sites. This provided also reasonable performance. However, this has not been validated in the larger series. Most of the literature is based on elastography methods, which include ultrasound elastography such as transient elastography and shear wave elastography, which is also subdivided in point shear wave elastography and 2D shear wave elastography or RFI methods and also magnetic resonance elastography. There are differences in the way the data is acquired, as you can see on the image, with a wider region of interest or excitation for 2D SWE and MR elastography, which provides almost whole organ coverage for the liver. These methods all have advantages and disadvantages, which I will not discuss in this talk. This is an example of a patient without poor hypertension who underwent hepatic venous pressure gradient measurement. This is actually a recent patient in our study where we performed 2D SWE, as you can see here, of the liver and spleen with 10 measurements where you get the median value of the liver stiffness and spleen stiffness. Patient also underwent TE or TEM, TE or FibroScan of the liver. As you can see here, the stiffness was 10.1 and of the spleen using a new module provided by a FibroScan device. So let's discuss the diagnostic value of liver stiffness measured with transient elastography. It has, as we know, liver stiffness with TE has excellent performance for diagnosing advanced fibrosis and cirrhosis. It has also been described to have excellent performance in diagnosing clinically significant poor hypertension with multiple papers showing high performance. Two meta-analyses reporting AUCs over equal 0.9. The Veno6 guidelines suggest that patients with a liver stiffness less than 20 and the normal tick count should not undergo endoscopy with, however, a 5% chance of missing esophageal varices requiring therapy. So it's not 100% accurate. Patients with liver stiffness over equal 20 or 25 can be considered to have clinically significant poor hypertension and should undergo a screening endoscopy. However, the performance of liver stiffness for prediction of varices is not as well established. With AUC, best AUC is described between 0.82 and 0.84. There are limitations associated with TE, which are the same as for liver fibrosis staging, which relate to failed cases, where in which case you can use an XL probe and also unreliable measurements. What about spleen stiffness, which can be measured with ultrasound elastography with TE, which has now recently provided a spleen stiffness module and also with SWE? So papers have shown or studies have shown, mostly from Europe, that spleen stiffness can be used as a possible biomarker of esophageal varices. When you look more carefully at the data for performance, for diagnosing performance for clinically significant poor hypertension, the data is a little bit conflicting. There is a, for example, a paper from 2016 showed that spleen stiffness correlated better with HVPG measurements with excellent AUC over 0.9 for predicting CHPH and varices, significantly better than liver stiffness, spleen size, and platelet count. The same authors actually published another paper the same year showing that spleen stiffness can also predict outcome with reasonable accuracy, including patient death and decompensation. There is a paper from Europe also showing that actually liver stiffness would perform better than spleen stiffness and also that SWE was performing better than TE. And finally, another paper from Spain showed equivalent AUCs for spleen stiffness and liver stiffness for detection of varices. So I think, overall, the jury is still out to say whether spleen stiffness has added value over liver stiffness when using ultrasound allostrography. Let's discuss now the role of MR allostrography. So we know that MR allostrography, when measuring liver stiffness, has shown excellent performance for predicting liver fibrosis stage, including moderate advanced fibrosis and cirrhosis with failure rate of about 3 to 5 percent, mostly depending based on the patients with liver iron deposition. Now, it is possible to measure spleen stiffness sometimes using a single driver placed in the right side of the patient where there's enough wave propagation to the left side. However, we prefer, in our experience, to perform what we call a dual driver acquisition where you're placing a liver and spleen driver on the left side at the same time, and this can be using vibration of 60 hertz. So these are some examples showing wave propagation in the liver and spleen with excellent image quality, showing allostrogram in the liver and spleen with high liver and spleen stiffness in this case. This is another example of a patient where we perform also liver and spleen stiffness measurement, showing high or increased stiffness in both organs. What is the performance of MR allostrography for diagnosing clinically significant poor hypertension? This is our experience based on a prospective NIH-funded study where we have done over 60 cases. We are reporting here approximately 45 cases with concomitant HVPG measurements, and you can see overall that the patient can see overall that the measurement of spleen stiffness using 3D MRE had the best performance with an AUC of 0.9 for diagnosing CHPH. 2D MRE of the spleen also had excellent performance, 0.884, and SWE of the spleen had fair performance with an AUC of 0.73. You can see on the right side that the distribution shows a very nice separation of patients with CHPH versus patients without CHPH using 3D MRE of the spleen, while there was more overlap using SWE of the liver and spleen. This is showing two examples of one patient without poor hypertension and a patient with clinically significant poor hypertension. You can see major differences in liver and spleen stiffness, and also in 3D MRE, you can see increased spleen stiffness using 3D MRE, and also higher stiffness in the liver in the patient with CHPH and higher spleen stiffness as well. What's interesting also is that sometimes the spleen size can be normal in patients with clinically significant poor hypertension. You can see on the right side, the patient with the normal-looking spleen, however, with increased spleen stiffness at 14 kilopascals compared to a patient with normal-looking spleen stiffness while the spleen was at 20 centimeters. I want to discuss also non-ellisography methods that can be used noninvasively to diagnose poor hypertension. One of them is the measurement of liver T1 mapping and also spleen T1 mapping. So T1 mapping is a quantitative measure of T1 relaxation, which is an inherent magnetic property of the liver. This can be measured at baseline, meaning without contrast injection. It's called native T1 or pre-contrast T1, and it can be also measured after injection of contrast, either regular contrast or using a liver-specific contrast, such as garaxetate or garaxetic acid. In our experience, in the same series, we have done this prospectively. You can see that the measurement of the native or pre-contrast liver T1 did not provide good enough AUC. However, you can sensitize this by measuring the changes using a liver-specific agent, and you can see that the measurement of liver T1 after injection of garaxetate at the hepatobiliary phase, 20 minutes after injection, has excellent area under the curve of 0.881 with a threshold of 265 milliseconds with excellent sensitivity and specificity. The Delta T1, which is the difference in T1 pre- and post-contrast, also had excellent sensitivity and specificity. You can see also that in the same study, corrected T1 based on multi-scan at baseline without contrast injection had also poor performance. Other parameters like spleen T1 or corrected T1 also, and spleen size and APRE had also poor to fair performance for diagnosing CHPH. So, of course, we need validation for this data, but this is really promising. These are two examples of patients with liver T1 mapping. So, on top is a patient without clinically significant part hypertension. This patient had, as you can see, major changes in liver T1 pre- and post-injection, and the liver T1 becomes extremely short because there is normal uptake of garaxetic acid in the hepatocyte, meaning there are normal functioning hepatocytes without poor hypertension, as opposed to the patient at the bottom where you can see small changes, but, however, not the same magnitude. And you can see that the liver T1 post-contrast injection at the hepatobiliary phase was at 441 in this case compared to 128 in the case without clinically significant poor hypertension. So, of course, we need validation of these results. We have been also looking at DCE MRI, or dynamic contrast-enhanced MRI, whereby you inject a contrast into the IV contrast, and you keep acquiring data over a very high-temperature resolution acquisition. And in this case, we have been using also garaxetate, and you can see major differences in liver uptake, which in a way mirrors what happens also with T1 mapping. This is kind of a different quantitative way of doing it. And you can see that a patient with that poor hypertension has normal high uptake of the contrast using DCE MRI with parameters quantified here. And as you can see, a patient with clinically significant poor hypertension has very slow uptake, almost like a plateau type of uptake with major differences in flow parameters between the two patients. This also needs validation. The disadvantage of this method is that it needs contrast injection, which could be sometimes contraindicated in patients with liver and kidney disease. I would like to discuss briefly also flow quantification of the splenic vasculature. This has been performed typically with Doppler ultrasound, where you can measure the flow in the portal vein, for example, in the splenic vein. The problem with the Doppler and the ultrasound is that it is operator-dependent, and the flow quantification can be dependent on the angulation of your selection of your vessel. So face contrast imaging has been around for a long time in MRI. It's a non-contrast technique, which is interesting, can be used in any patient, whether their kidney function is normal or not, and it can be used to measure not only flow, but also velocities in the portal vein. So the advantage of now face contrast is that you can take it to the next level, and you can do what we call 4D flow imaging of the hepatic vessels, where you're doing a volumetric acquisition of a 3D face contrast acquisition that is cardiac-triggered, and you can really get flow quantification based on a triggering of the cardiac cycle. This is a quite elegant way of not only visualizing the mixing of the flow, as you can see on the left side, between the SMV and the splenic vein. You can do particle tracing of the flow, but you can also measure flow in any vessel in the volume acquisitions. You can measure, of course, the portal vein, the hepatic artery, the IVC, the aorta, and other vessels such as the splenic vein, the celiac artery, et cetera. This can be done within a few minutes. The disadvantage is that it needs complex processing, and processing is making a lot of progress recently. This is an example of the patient. Pre-TIPS and post-TIPS, you can see nice visualization of the flow in the TIPS performed in this patient. To summarize, this was an overview of the currently used non-invasive techniques for diagnosis of portal hypertension. Overall, ultrasound and MR allostrography techniques are accurate methods and non-invasive methods for diagnosing clinically significant portal hypertension and severe portal hypertension, especially when combined with platelet count as based on the BFNO6 guidelines. There are limitations inherent to each technique, including failure and cost or availability for MR allostrography. There's still some conflicting data about splenic stiffness using ultrasound allostrography. However, there's very promising data based on MR allostrography of the spleen that needs to be validated. Of course, new directions would include validating spleen stiffness using the new TE platform and 3D MR allostrography, and also using some of these biomarkers to also predict the outcome in these patients, which can be really important, and also assess response to therapy in patients with portal hypertension. I thank you very much for listening, and I wish you a good meeting. Thank you. Dear colleagues, it's a pleasure and honor to talk to you today about the role of the intragastric balloon in naphthalene T and cirrhosis. The question is, is it ready for prime time? My name is Parham Obedeha. I am a consultant in gastroenterology and the director of advanced endoscopy, professor of medicine, and vice chair of innovation at Mayo Clinic in Rochester, Minnesota. These are my disclosures as it pertains to this talk. Before we talk about the technicality of the role of the intragastric balloon for naphthalene D management, it's very important to highlight the unmet need for this disease process. Currently, this disease afflicts anywhere between 13 to almost 30% of the world population, depending on the geographic distribution. And this disease goes hand in hand with metabolic syndrome, insulin resistance, type 2 diabetes, and the obesity pandemic. Our understanding of the pathophysiology of this disease has changed significantly over the past few years, and now we recognize the importance of the gut, adipose tissue, gut, liver access in the pathophysiology of the disease. And as such, targeted therapy to the GI tract and or adipose tissue and or improvement of insulin resistance are key therapeutic targets for this disease. So, in states of metabolic stress, there's a state of dysbiosis and gut leakiness, communicating with the visceral adipose tissue and the liver through the portal circulation or through a nerve signaling, resulting in inflammation in adipose tissue, increased lipolysis, increased inflammation within the liver, with the end product being a deposit of fat in hepatocytes, starting inflammation, mitochondrial damage, and starting the process of fibrosis, which eventually would lead to irreversible liver injury. This is again highlighted in this review we published in the Journal of Hepatology, emphasizing the importance of insulin resistance, gut dysbiosis, and metabolic endotoxemia, and bile acid circulations in the pathophysiology of NASH, and as a therapeutic target for this disease. As I stated, metabolic syndrome, insulin resistance, and NAFLD go hand in hand, so the prevalence of NAFLD and NASH among patients with type 2 diabetes is about 70%. Therefore, there's common pathophysiological pathways of these disorders, and you cannot treat NASH with directed liver targets, ignoring the milieu that promoted this pathophysiology with important contributions from the gut and from a state of insulin resistance. Therefore, in 2021, the targeted goals for the treatment of NAFLD should be to decrease caloric metabolic load, and to improve this metabolic endotoxemia, and to improve insulin resistance, and the best tried and true way to do that is through weight loss in excess of 10% total body weight loss, capitalizing on weight-dependent pathway, and or weight-independent pathways capitalizing on this gut, liver, and adipose tissue access that have been implicated in the success of bariatric surgeries, such as row-wide gastric bypass, in the management of this disease. To highlight the importance of this 10% total body weight loss, this is a graph from scale 0 to 100 showing fibrosis improvement and NASH resolution. You could see 10% total body weight loss far exceeds all the targeted therapies in achieving both endpoints of fibrosis improvement and NASH resolution. So, bingo, I don't need to talk anymore. We have a treatment which is 10% total body weight loss, but unfortunately, unfortunately is the vast majority of patients who are subjected to a high-intensity even lifestyle modification program fail to achieve this threshold of 10% total body weight loss. As you could see here, percent of patients achieving this even within the confines of a high-quality intensive lifestyle program is about 10%. Therefore, we know the target, which is 10% total body weight loss, but we don't have a non- procedural or medication-mediated way to get there with lifestyle interventions alone. Therefore, the spectrum of intervention should expand to allow patients with NAPLD to reach this critical endpoint of 10% total body weight loss for remission of their disease, and the spectrum should include bariatric endoscopic therapies such as the intragastric balloon or endoscopic sleep gastroplasty that, per design, they're anatomy-preserving, minimally invasive, enabling patients to reach this 10% total body weight loss. So, technical details and mechanism of action of how the intragastric balloon is implicated in proof of NAPLD. First, it's very important to highlight that the intragastric balloon is a technically easy procedure. It's placed like an orogastric tube filled with 600 or so cc's of saline. Sometimes we'll mix it with methylene blue. It's done as an outpatient. For placement, it could be done with conscious sedation. So, the procedure is technically easy. However, the treatment should be administered within the compound of a multidisciplinary program with close patient follow-up as there is a foreign body in the stomach that could be associated with adverse events that could be mitigated if patient has adequate follow-up. So, we talked about 10% total body weight loss, and that's an important mechanism of action for the intragastric balloon in NAPLD improvement, and you could see that you put the balloon in the stomach. You're activating satiety association pathways. This results in 10% total body weight loss in the average patient, which results in steatosis improvement, and with time, after you remove the insults of fat and inflammation in the liver, hopefully fibrosis regression. So, what's the data that the balloon is capable of reaching this threshold? The data comes from more than 20 plus years of experience with the intragastric balloon. This is a meta-analysis published by the ASGE Bariatric Endoscopy Task Force a few years back, and you could see at the six-month mark, there's about 13% total body weight loss. 12 months, there's about 10% total body weight loss, and if the patients are subjected to a comprehensive weight maintenance program, that weight loss could be maintained even over the long-term period. The balloon is associated with some accommodative symptoms like pain and nausea, GERD, in the early periods of accommodation to the balloon. However, incidents of serious adverse events such as migration, perforation, and death are quite rare. So, these are results from randomized controlled trials in the United States that looked at the different balloons. This is the adjustable balloon. This is the fluid-filled non-adjustable balloon. This is the double balloon, and this is the gas-filled balloon. These are the different sham or lifestyle program comparative arm. As a group, you could see that the intragastric balloons result in 10 or more percent total body weight loss compared to lifestyle interventions in randomized controlled trials. So, this is what's important for the patient, is this slide right here. If I'm a patient with natural D or NASH, what's my odds of getting 10% total body weight loss within the treatment period that I'm subjected to? Here, I plotted different lifestyle interventions, different medications, and bariatric surgery with the odds of reaching more than 10% total body weight loss, and you could see bariatric surgery is very effective in reaching this threshold, but also endoscopic bariatric options, such as the intragastric balloon, which improve the chances of the patient getting to 10% total body weight loss compared to lifestyle interventions significantly. There's been some reports from the FDA post-marketing about mortality associated with the intragastric balloon. It's very important to highlight that there is no causal association that has been proven. These are just reports that go to the new database and are reported to just raise awareness. However, if you look at randomized controlled trials, post-marketing surveillance trials of large numbers of patients, meta-analysis that included thousands of patients, and registries like the Brazilian registry that included more than 40,000 individuals, the risk of severe adverse events like mortality, perforation, bleeding requiring intervention is quite low. It doesn't mean it doesn't exist. It means in the compound of a multidisciplinary program where the patient is carefully monitored and these things are detected and symptoms are addressed early, these events should be quite rare, and the balloon should be a well-tolerated intervention for obesity and for the purpose of our talk for an alcoholic fatty liver disease. When we look at the impact of the balloon on liver parameters and metabolic control, this is meta-analysis of 10 randomized trials and more than 30 observational studies, including more than 5,000 patients. You could see that the balloon is associated with improved glucose homeostasis and insulin resistance. There's improvement in liver biochemical panels, and there is also improvement in NASH activity scores in the published literature associated with the balloon. We spent a lot of time talking about weight-dependent effect and improvement of NAFLD with intragastric balloon. However, the balloon does have significant and important weight-independent effect and highlighted in this green pathway in addition to its weight loss-dependent pathways. Especially with the fluid-filled gastric balloons, they do delay gastric emptying, and that's coming from Level 1 evidence. The delay in gastric emptying translates to decreased postprandial hyperglycemia, decreased postprandial insulin levels, which results in improved insulin sensitivity. The improved insulin sensitivity decreases insulin levels, which results in decreased adipostitial lipolysis, decreased de novo liver lipogenesis, changes in visceral adiposity, alteration in adipokines and cytokines, with the ultimate result of decreased insulin, which is fat in the liver, resolution of inflammation that ultimately translates to fibrosis regression as well. This is the data that showed that the balloon is associated with delays in gastric emptying. This is from randomized control US trial. This is gastric retention in a radiolabeled meal at baseline. You could see after balloon placement in blue, there's significant delay in gastric emptying that persists while the balloon is in place. But when you remove the balloon, this delay in gastric emptying reverts back to baseline. So, why is this delay in gastric emptying important? Because of these glucose homeostasis curves. In green, it's normal. You're eating a meal. There is an insulin response to deal with the ingested meal. In states of impaired glucose tolerance, the insulin peak is heightened and the area under the curve of the insulin response is also increased to deal with the state of insulin resistance, and that is exaggerated in states of type 2 diabetes as such. When you delay gastric emptying, the peak of the insulin response is decreased, and the area under the curve is also decreased, resulting in improved insulin resistance and improved insulin levels, leading to the improvement in NASH parameters. This has been shown in this paper published in Cell Metabolism. You have a delay in gastric emptying, and that delay in gastric emptying decreased postprandial glucose and postprandial insulin, thus capitalizing on weight-independent effect of ventrogastric balloon on insulin resistance and on non-alcoholic fatty liver disease. A lot of the studies I described to you were indirect evidence of improvement in liver parameters of NAPLD on biochemical testing. This is a prospective FDA IDE study that we conducted at Mayo Clinic to look at histological impact of the gastric balloon on liver parameter. The study enrolled patients with NASH and early fibrosis, so we did MREs on everybody to select this cohort, and we subjected them to intragastric balloon and impaired liver biopsies using endoscopic ultrasound at the time of balloon placement and at the time of balloon removal with a repeat of the MRI at six months as well. The primary endpoint is more than two points improvement in NASH activity scores on impaired liver biopsies and resolution of steatohepatitis. This was a short-term study, so we did not anticipate much impact on fibrosis because fibrosis takes time to regress. However, as we'll show you, we did see some good signals that fibrosis is also improving with the intragastric balloon even in the short term. And then we have the safety endpoints of incidence of serious adverse events. These are the exclusion or inclusion exclusion criteria, but we included individuals with body mass index 30 to 55, and again with NASH with early fibrosis on MRE who failed to achieve lifestyle and weight loss through lifestyle intervention alone. This is the study design that we alluded to, impaired liver biopsies before and after balloon improvement, biochemical liver testing, and liver histological assessment at baseline at six months. This is the baseline demographics of the cohort. You see body mass index at 44 average for the cohort. So, these are individuals with severe or class 3 obesity. Most of them had either diabetes or impaired glucose tolerance. As we suggested, this goes hand-in-hand with NASH. Their A1c was relatively under reasonable control with an average A1c of 7.4. MRE did a good job in selecting our patients. Most of them, actually all of them, had a NASH activity score above 2, and they had different stages of liver fibrosis, but not cirrhosis. The intragastric balloon was associated with significant improvement in total body weight loss. Again, the majority of the cohort reaching more than 10 percent total body weight loss, which is the threshold for improvement in natal DNA. Waist and hip circumference improved. Glucose hemostasis and insulin resistance also improved significantly. I told you this cohort started with reasonably controlled diabetes in some of them with an A1c of 7.4, but even with this cohort, we observed a 1.3 points improvement in hemoglobin A1c with the intragastric balloon. Liver biochemical panels also significantly improved. Liver stiffness on MRE improved. Percent fat fraction dramatically improved by more than 60 percent, and so as burden of visceral adiposity as measured by mesenteric fat. If you look at NASH activity scores, which is the primary endpoint of the study, 95 percent of the patients have an average decrease of 2.6 points on NAS scale. Eight percent had decreased by at least 2.65 percent had complete resolution of NASH. As I stated, this was a six-month study. We did not anticipate to see much improvement in fibrosis. However, we did see improvement in fibrosis in some patients on liver biopsies, and on MRE, which is more of a global assessment of the liver, we saw that the MR elastography scores improved in 75 percent of the cohort by an average of 0.7 kilopascals. This is NAFLD activity score before and after. You could see lots of color before to having lots of white, which is indicating resolution in NASH, and this is the pre- and post-area under the curve with NASH activity score. This is the fibrosis scores with 5 percent having normal fibrosis scores before the intragastric balloon going up to 43 percent after the intragastric balloon placement. So, in summary, 90 percent improved their NAFLD activity scores, 80 percent improved by more than 2 points, 50 percent improved in MRE-detected fibrosis by an average of 1.5 stage, and 50 percent reached endpoints approved by the FDA for NASH resolution and fibrosis. And based on these data, the FDA did issue a breakthrough designation of the intragastric balloon that's fluid-filled, which is the ORBERA, for management of NAFLD or for management of NASH with fibrosis, and this will allow us to hopefully work with the FDA to design a larger trial to pan this concept. Another study published in 26 patients in Libre International looked at fibrosis improvement and controlled attenuation parameter using FibroScan with intragastric balloon that did show significant improvement in these parameters as well, and no serious adverse events associated with the intragastric balloon treatment. So, what's the role of the intragastric balloon in NASH management in 2021? My opinion, it should play a cornerstone role in the management of the disease because the paradigm is as such. We need to put this disease in remission. That means we need a treatment that have weight loss-dependent and independent effects that gets NASH resolution or fat and inflammation outside of the river, and currently in 2021, surgery is very effective, endoscopy is effective, pharmacotherapy with some newer agents is effective, but lifestyle by itself is not that effective. Now, after you put the disease in remission, you break through that wall, then that's where we maximize lifestyle intervention because you remove the insult, the patient is engaged now, they see the fruits of their labor, and now you could add significant lifestyle and behavioral components, you add pharmacotherapies, and the role of surgery endoscopy in the weight maintenance space becomes less and less. This is, by the way, is highlighted in the practice guidelines put by the HDA, where patients with a chronic disease like obesity should be always in treatment for that disease because the disease does not go away. The treatment starts by intense weight loss intervention phase, followed by weight maintenance intervention phase, followed by weight regain prevention, and then obesity assessment, and you keep going with that cycle. So that's how a temporary device like an intracast balloon should play a major role in the treatment of an unmet need, like non-alcoholic fatty liver disease or sciatica hepatitis, by putting the patient in this long-term cycle of treatment that starts with aggressive weight loss to get the disease in remission, followed by weight maintenance. And we have multiple options for weight maintenance. Now we have good drugs that you could combine them with the intracast balloon in order to have a more durable long-term response for that effect. And as I told you, the HDA put in clinical guidelines to highlight this paradigm shift in how we should be managing obesity. And I would refer you to these guidelines. They summarize the evidence beautifully, and we should be taking the disease of obesity, metabolic disease, and more importantly for this talk, non-alcoholic fatty liver disease and sciatica hepatitis very seriously. And we need effective treatment modalities, such as the intracast balloon. So where do you see the future? The future is we have tried multiple directed drugs against the liver, and most of them have not panned out because the pathophysiology of the disease does not start in the liver. It's in the gut. It's in insulin resistance. Therefore, I think we need to be focusing on weight loss and targeting the gut to weight-dependent and weight-independent pathways, because that's going to result in significant percentage of patients with resolution of NASH. And then we could add either targeted small intestinal interventions like duodenal mucosal resurfacing or newer drugs like the GLP-1 agonist to enhance this response and maintain it for the longer term. And with that, I conclude and thank you for your kind attention. Hello, I'm Julie Heimbach, Professor of Surgery and the Director of the Transplant Center at Mayo Clinic in Rochester, Minnesota, and I'm delighted to speak to you today about surgical approaches to naphthalene cirrhosis, whether we should think about this before, during, or after transplantation. I have nothing to disclose. Worldwide, the epidemic of obesity has actually tripled in a very short time, actually since 1975, and as you can see now, in the United States, currently 42% of U.S. adults are obese with a BMI greater than 30. What does this mean for the liver? Well, of course, we all know that obesity is associated with fatty liver disease, and for the very vast majority of patients, this can be simple steatosis, which then may progress in a small number of patients to inflammation, which is NASH, and actually only a minority progress to advanced fibrosis, which is when we start thinking about transplant. What we also know is that fibrosis regression is possible if we can achieve a greater than 10% total body weight loss. What does this mean for transplant? Well, what is shown here on the left of this graph is the U.S. population incidence of obesity, and as you can see, that is really rising rapidly, as we've just demonstrated on the last slide. Over this very short period of time, from 1995 to 2015, you can also see the dramatic rise of the liver transplant registrations for the condition of NASH, and we actually also see overlaid on this same graph is the incidence of diabetes in the U.S. population, which is also obviously on the rise. What does this mean for post-transplant outcomes? Well, actually, though initial SRTR reports showed inferior outcomes for obese patients, more recent data, looking from 2004 to 2011, comparing those with BMI less than 18 to those 18 to 45, and then those with extreme obesity greater than 45 BMI, actually only BMI at less than 18.5 is associated with worse survival, keeping in mind, however, that these are highly selected patients, most notably selected for the absence of cardiovascular disease, but again, evidence that patients could be safely selected with obesity to undergo transplant and have satisfactory outcomes. However, looking at the impact on the long-term for obesity, a multi-center study in Australia, looking at the combination of patients with both obesity and diabetes, demonstrated a worse outcome at five years post-liver transplant. This was true only for those with the combination of obesity and diabetes, whereas those that had either just obesity or just diabetes, or neither of these things, with similar outcomes out to 10 years post-transplant. So what can be done about that? Well, this is a very elegant study, looking at the impact of weight loss on liver fibrosis, and it's 45 patients who had a paired biopsy serially, and those who are able to obtain greater than 10% total body weight loss, as you can see from biopsy one to biopsy two over five years, those with that greater than 10% total body weight loss, all actually demonstrating even those with stage four fibrosis, some regression of that fibrosis, whereas those that had between zero and 10% or weight gain, primarily demonstrating a progression of fibrosis. This has also been demonstrated in this series, which specifically looked just at those patients with NAFLD who had bariatric surgery. This is a large cohort, 64 patients who had a paired biopsy, and they had NASH resolution in 84%, and 70% of these had improvement in fibrosis. Those who had persistent NASH had no improvement in fibrosis, and that cohort also was demonstrating less weight loss. You can see the baseline, and then the improvement, so those with F0, no fibrosis, being higher, almost 60% of the population, which is really remarkable. We also know that bariatric surgery is efficacious in the long term. This is a large cohort of patients followed for 12 years. This is not patients with fatty liver disease. This is just patients who underwent Roux-en-Y gastric bypass, shown on the top here, with 418 at the baseline, and then followed out to 12 years. We see not only did the patients have that good weight loss early after the surgery, but this is maintained for 12 years in a majority of patients, obviously not every patient, but more than 50% maintaining that weight loss. We also see a 95% reduction in new onset diabetes at 12 years, as well as, quite remarkably, a 51% of the patients that had diabetes at the time of surgery resolved their diabetes and had that diabetes still resolved at 12 years. This really demonstrates the long-term efficacy of Roux-en-Y gastric bypass in this cohort. The comparator group are those that were actually referred, but were not approved by their insurance company. You can really see, essentially, no change in their weight in that cohort, other than those that were able to cross over, which are the ones that are shown in the gray. What about bariatric surgery in patients with cirrhosis? I previously showed data for patients with NAFLD. This is done in a very small number of studies. You can see here five studies between 13 and 23 patients. The laparoscopic sleeve gastric or laparoscopic Roux-en-Y gastric bypass was performed in these five studies. Again, these are quite small, between 13 and 23 patients. All reported a longer OR time and some increased complication rate, but what they were able to conclude was that bariatric surgery was overall safe and effective with achieving their weight loss in selected patients with compensated child's acerosis. Well, child's acerosis are typically not the patients that we consider for transplant. What about patients we might be thinking about for transplant? Is there any role of bariatric surgery in this cohort? Well, in fact, this actually was looked at at UCSF for 32 patients who were referred for transplant. Now, notably, these patients had a lower MELD than most patients that are being considered for transplant with MELD of 12, range was 11 to 13 in this study. The majority were child's A, but it did include some patients with child's B, including some with varices. The mean BMI was 45. They had no deaths in this cohort of 32 patients. They had one patient with a leak. Of this initial 32 that underwent the sleeve, 21 were ultimately listed and 14, only 14 actually were ultimately transplanted, which is notable to consider. However, we do see the efficacy of the weight loss in this cohort. This analysis is from Texas, again, single center analysis with 78 patients. It gets to this question of potentially what may happen to patients while they're waiting for transplant when they've undergone prior bariatric surgery. This actually was a cohort that came to the center having already had their bariatric surgery, subsequently then referred for liver transplant, matched to 156 patients who did not have bariatric surgery prior to coming to the center. They were matched for age and MELD and etiology of liver disease. They had a similar metabolic profile in the two groups with a similar BMI, similar diabetes and hypertension rate. In fact, the death and delisting was found to be higher than those with prior bariatric surgery, so really not making it to transplant in 33% of those that were listed, so initially found to be suitable, but then ultimately did not make it to transplant compared to only 10% being delisted when they did not have had that history of prior bariatric surgery. Transplant rate was also lower in those with prior bariatric surgery at 49% compared to 65%. So, the overall intention to treat from the time of listing to one year post liver transplant is shown to be lower in those with prior bariatric surgery coming for liver transplant referral compared to those coming for liver transplant referral alone without that prior bariatric surgery. You can see that here in this Tapper-Meier curve. Another analysis of patients with prior bariatric surgery being considered for liver transplant, although this one really does not look at waitlist outcomes, rather it just focuses on post-transplant outcomes. It's a pretty large cohort, 33 patients, single center analysis from the University of Minnesota who were then matched 3 to 1 by age, meld, etiology, BMI sex, and date of transplant, and this is a rather broad cohort all the way back to 1987 up to 2017. They found a similar length of stay, similar ICU stay, and actually similar complication rate, and as you can see in the Tapper-Meier post-transplant outcomes quite similar in those two groups of patients, so this is retrospective, but it's a big number, and their overall conclusion was that you can achieve similar post-liver transplant survival at 1 in 3 years. Again, there is obviously the opportunity to select the patients carefully, but that bariatric surgery alone does not necessarily mean that the patient would be guaranteed to have a worse outcome. But we do need to consider bariatric surgery in patients with cirrhosis and the possibility of what's called publication bias, and we typically don't like to write about things, especially surgical things, that don't work out very well, so we can get around publication bias by using large data sets, and this is a large data set, the nationwide inpatient sample between 1998 and 2007 looking at patients, so this is across the U.S. cohort of patients. They were able to identify those with bariatric surgery and decompensated cirrhosis, just 62 patients compared with nearly 4,000 who had cirrhosis that was compensated or no cirrhosis, which was 670,000. The diagnosis of ascites or varices was required to be classified as decompensated. In those with decompensated cirrhosis, they did see an in-hospital mortality of 16% for an elective surgery, which is clearly unacceptable, versus 0.9 for cirrhosis and 0.3 for no cirrhosis. The length of stay was also longer, and so this was published in 2011, and actually, a similar analysis was just redone now on a cohort of patients, again, using the nationwide inpatient sample from 2004 to 2016 to see if now, maybe in this more recent era, things are going better, but unfortunately, things appear to be going worse with the in-hospital mortality at 1.8 in patients with compensated cirrhosis compared to 0.2 with no cirrhosis and 22% suffering in-hospital mortality with decompensated cirrhosis who underwent an elective procedure, which, again, is clearly unacceptable. Interestingly, with this data, they also demonstrated higher complication rates in deaths for low-volume centers, and interestingly also, after 2013, they saw a decrease in the rate of rewind gastric bypass and an increase in the rate of sleeve, and concurrent with that, an improvement in the post-operative survival. This is an analysis which is not in patients with cirrhosis, but it is a randomized study, and it does specifically look at that type of surgery and whether this could have any impact, and this is rewi gastric bypass versus sleeve gastrectomy, again, randomized patients, so 66 with NAFLD, and they underwent either a sleeve or rewi. Patients with a rewi were noted to have more elevated INR and a lower albumin at one month versus sleeve gastrectomy, although this resolved by one year, and when we look at one year, basically, the patients had the same excess weight loss and the overall same outcome, but what they concluded was that patients with NASH who were undergoing rewi gastric bypass were more susceptible to early transient liver dysfunction versus sleeve gastrectomy, so it's interesting that potentially sleeve gastrectomy is a better operation for these patients with more susceptible liver. What about post-transplant? So, we've talked about pre-transplant and how it's not possible to do bariatric surgery in those with decompensated cirrhosis, so then you could do the transplant and think about bariatric surgery after transplant. What would happen? Well, this is actually a report of open rewi gastric bypass performed post-transplant on just seven patients, so this is a small series, but really, unfortunately, all the series are relatively small, followed after transplant at 2.6 years, and with the main follow-up then of five years after the surgery, for these seven patients, they were able to show effective weight loss, but unfortunately, two patients died in the first one year and one had to have a reversal of the surgery. Now, this is in the era of hepatitis C and lack of effective therapy, but still three very adverse outcomes in a cohort of just seven patients undergoing rewi gastric bypass after liver transplant. A more recent cohort, 2014 to 2018, single center, 15 patients now undergoing sleeve gastrectomy. This is a median of 2.2 years post-liver transplant, similar to the last analysis. They had basically, essentially, the same outcomes as the patients who did not have a liver transplant prior to their sleeve gastrectomy, with the exception of a little bit more blood loss in the liver transplant group, otherwise, operative time, length of stay trending towards being longer for the liver transplant patients, but really, otherwise, very similar outcomes. So, is there another option? Well, actually, there is also a consideration of a combined liver transplant with sleeve gastrectomy, and this is actually the approach we've adopted at Mayo Clinic, where we first enroll the patients in a non-invasive medical weight loss, and for those patients who are not successful with a non-invasive medical weight loss to not achieve their goal weight, but they have now a high enough meld to achieve transplant, we offer a combined gastric sleeve at the time of liver transplant. We've opted for the gastric sleeve over the Roux-en-Y gastric bypass because of the lack of malabsorption, and because the weight loss is slower, so it's more gradual, which may be favorable in a patient who is more medically complex, and potentially also favorable in a patient with liver issues, as demonstrated in the previous randomized study. This is also technically more straightforward. This is an interoperative assessment of a patient who is undergoing a liver transplant, as you can see, also with obesity, and then this would be the newly transplanted liver and the sleeve gastrectomy performed. And we've now recently published our outcomes following a long term, so this is all of the patients who had more than three years of follow-up, and what we can see here is that those in the red, who at the time of listing had a significantly higher VMI, it was actually 48 compared to those at the time of listing who were at 40, and able to achieve that VMI goal under 35, so they did not have the sleeve gastrectomy, whereas those who started at a higher level underwent medical weight loss but could not achieve that acceptable weight loss had the combined surgery. And we can see following those patients at the long term, VMI continuing to be maintained at that lower amount for the combined patients with a higher incidence of weight regain in those that had the sleeve gastrectomy. Looking at the percent excess body weight loss, again, at the time of listing, and then at the time of transplant, by definition, those who did not have the sleeve had more weight loss, but following transplant, unfortunately susceptible then to that weight gain. And the key point is, really, this is for 29 patients with a combined liver transplant sleeve gastrectomy who were followed for three years compared to 36 who had the liver transplant alone. 29% of those with the liver transplant alone could maintain that greater than 10% total body weight loss, which, again, is needed to maintain or improve the fibrosis, whereas 100% of the liver transplant sleeve gastrectomy patients maintain that greater than 10% total body weight loss. We also looked at the other metabolic parameters, such as diabetes and hypertension, and we found, again, significant improvements in these other parameters of the metabolic syndrome in patients with combined liver transplant and sleeve gastrectomy compared to those with liver transplant alone. From a practical standpoint, the key would be to have a standardized approach that we offer to all patients rather than intermittently differing recommendations with specific nutritional activity and weight loss goals rather than simply telling the patient to, you know, go lose weight and come back when you've been successful. We really try to give a standardized approach with very specific follow-up goals for all patients. We also know that it's critical to follow very closely after transplant because of the possibility of reflux, which is quite significant following transplant and sleeve gastrectomy. This may be potentially related to their other comorbidities like portal hypertension or potentially something related to medications, but this is something that does need to be managed. We also see the risk of excess weight loss, so we need to follow these patients closely and also regain in some of the patients. From a technical standpoint, the distribution of the weight and the presence of ascites are really important in the ability to attain exposure for the surgery. So the patient on the left with ascites and the lower distribution of their weight compared to the patient on the right whose obesity is central, sub-fascial, really more challenging to attain the necessary exposure for transplant. So this actually is a systematic review trying to address the exact question that I was asked to cover in the talk today, whether bariatric surgery should be before, during, or after liver transplant. And this systematic review was able to find 19 studies about this question, eight prior to liver transplant, two concurrent, and nine post-liver transplant. However, unfortunately, in the systematic review, nearly all of the studies are not all of the studies are relatively small series or case reports. And there is, of course, a risk of publication bias. There are variable endpoints, and there's three distinctly different populations. And really, the bottom line conclusion from the systematic review is it's too early to really be able to perform a systematic review. And thus, from the standpoint of looking at it systematically, there's not a way to make an optimal recommendation. And this actually is a single center analysis of the exact same question. Unfortunately, this is not prospective. This is all retrospective data where they looked at their experience in patients before, during, or after liver transplant. So between 2013 and 2016, they looked at all the patients who had bariatric surgery and who also were noted to either be waiting for a liver transplant or already had a liver transplant. So again, they had three different patient populations. So in the pre-transplant setting, they had 10 that had bariatric surgery, and then they had three that had the combined surgery. This is a single center from Israel. And then they had five that had that liver transplant or had bariatric surgery after liver transplant. For the pre-group of 10 patients, they did have two patients who experienced decompensation and death. They had several patients who actually stabilized and didn't need a transplant. And then they had several patients that proceeded to transplant. For the three that were concurrent, they had two that went without complication and one that had a bile leak, but that was easily managed. In the post-group, notably, in those five patients, actually, they performed both rewide gastric bypass and sleeve gastrectomy. And the way they made the decision was really looking at the standard indication. So if the patient had no contraindication for sleeve gastrectomy, they would proceed to sleeve gastrectomy. But if they had severe reflux or large hiatal hernia, this is a reason to think about rewind gastric bypass. So reviewing the AGA clinical practice update on bariatric surgery and cirrhosis. And this is a really small figure, but because this is recorded, you can likely go back and look at it and zoom it in even. But what you can see here, basically, is looking at patients with cirrhosis and obesity. The goal, again, is to achieve that total body weight loss greater than 10% so that we can potentially achieve some improvement. If they have compensated cirrhosis and they don't, meaning no portal hypertension, and if they're not sarcopenia, you really do have to understand if they have alcohol use as a comorbid condition because there are significant contributions to problems in patients with bariatric surgery who have excess alcohol after their bariatric surgery. But if we've managed that as well, they could be considered for bariatric surgery, in which case laparoscopic sleeve gastrectomy is likely the most optimal choice for these patients. That does, in a cirrhotic patient, allow you to continue to access the whole elementary tract. So in the case they do develop portal hypertension and, for example, need endoscopic therapy for varices, you'd still easily be able to do that. And post-transplant, still able to access the bile duct if you need to for any kind of revision. If they have decompensated cirrhosis, the question is, are they a liver transplant candidate? If they are, they could be considered for bariatric surgery concurrent with or following liver transplant. If they're not a liver transplant candidate, definitely don't want to think about surgery in that group, but you have to continue with specialized nutritional management activity and other best medical therapy for these patients. The key takeaway is greater than 10% total body weight loss does result in improved fibrosis and NAFLD, and that bariatric surgery does provide durable weight loss and improves fibrosis in NAFLD patients. Selected patients with compensated cirrhosis may be suitable for bariatric surgery. Bariatric surgery in decompensated cirrhosis is only feasible concurrent with or after liver transplant. Thank you very much for your attention today. Hello, everybody. My name is Jawad Ahmad. I'm a professor of medicine at the ICANN School of Medicine at Mount Sinai. And it's my pleasure to give this talk, the approach to early bile leaks, case for surgery as part of the 2021 AASLD-ASGE endoscopy course. This is session four, bilious JP drainage after liver transplantation, operative or endoscopic approach. I'm actually a last minute substitute and I have no disclosures. Just as a proviso, I'm actually not a surgeon. I'm a therapeutic endoscopist as well as a transplant hepatologist, but I put my surgical gown and hat on and I'll try and approach this from the point of view of a surgeon. So we'll start with this case. So this is a 59-year-old male who underwent liver transplantation for alcoholic liver disease and hepatocellular carcinoma four days ago. He's on Prograf, Cellcept and Prednisone and developed some low-grade fever and bilious output in his JP drain. His laboratory studies are shown there, mildly elevated white cell count at 13.8, hemoglobin is 9.7, ALT is 291, but that's actually improving since the transplant, since he had a little reperfusion injury. Total bilirubin is 3.4, which is also improving. He's still in the intensive care unit. His ultrasound shows patent vessels, but because of the bilious drainage, he undergoes ERCP and all of these images are real patients that I've taken care of. So you can see the cholangiogram on the right-hand side. You can see the staples are still in place. The scope has been inserted. The scope is 12 millimeters in diameter, so that gives you some idea of dimensions. You can see that a cannula and a wire have been inserted, and this is the recipient duct. You see contrast. The duct is around about four millimeters. You see duct filling what looks like the donor duct, but there is extravasation and filling of the inferior surface of the liver. So this is the outline of the inferior surface of the liver. So there is extravasation likely coming from here. This is the area of the anastomosis. So this is an anastomotic leak. The patient underwent stenting. So here we go. There's a large leak, as we've said, and this is a picture taken a little bit later on, and you can see the amount of contrast that's filling here. And again, this is the inferior surface of the liver. A wire has been inserted, and you can see the proximal end of the wire here. This actually is the donor duct, and on the subsequent picture over here, you can see that I've placed a stent across the anastomosis with the proximal end inside the donor duct. Patient actually did well, but there was still quite a large amount of biliary drainage, and the decision was taken to take him back to the operating room two days later, where this leak was seen. But in fact, the donor duct here, the distal end of it, looked pretty ischemic with not very good back bleeding, so they decided to perform a Roux-en-Y hepaticogenostomy, and the patient's now doing well one year post-transplant. Just as an outline for this talk, we'll talk about the definition of what's an early bowel leak, but really, to be honest, there isn't a standard definition. What are the risk factors for a bowel leak, and does that impact upon what you're going to do? Is there morbidity and mortality after a bowel leak? And then we'll talk about surgery versus endoscopic management, although there is very limited data comparing the two modes of treatment. So as I mentioned, what is an early bowel leak? There's no standard definition. The way I think about it, it can occur anywhere from post-operative day zero or one to post-operative day 30, and a lot of the studies, when you look at them, this is the kind of cutoff they use is 30 days, which usually means the patient's still in the hospital and never left. Typically, the very early leaks are going to be within the first week. The patient's still in the intensive care unit or at the very least in the step-down unit. When you look at the incidence of leaks, again, it depends which study you read. It's anywhere from 5% to 25%, particularly from earlier studies in the 90s. When you look at live donor transplants and in the pediatric population, the risk is essentially double, and A2 All study, which I'll show you a couple of slides later on, has some data on this. When you do have an early bowel leak, it's almost always a technical issue. So this is something that surgeons want to really take ownership of because when they see a leak, it's usually something that happened in the operating room, although I'll show you in some subsequent slides. It's also possible that there's a donor risk factor, a recipient risk factor, or perhaps some issue with the vasculature that perhaps led to this leak. Diagnosis is easy if the drains are still in place. There shouldn't be bile in JP drains after surgery. So if there is, particularly if there's a large amount, then by definition, there's a bile leak. The labs are not usually helpful. This is occurring very early on. Patients just got transplanted. Maybe they have a reperfusion injury. Maybe they were very sick heading into transplant, so they're very jaundiced. They're already in renal failure. So the labs don't usually help you. It's usually the fact that there's bile in the drain. Now, if the drains have already been removed, a clinical change can sometimes suggest this, a high white count or some suggestion that the patient has a collection in the abdomen. Since in those patients where the drains have been removed, you'll end up having to do imaging, and the imaging is likely to show a collection. The question is, is this a routine post-op collection? Is this a hematoma versus a bilious collection or a biloma? A HIDA scan might be helpful, but in general, this is a clinical diagnosis, but most of the time for this particular talk where we're talking about early bile leaks, there will be drains still in place, so you'll see bile, and that's an easy diagnosis to make. I'll show you some pictures in the next few slides of where would the leak be occurring. The most common is obviously the duct-to-duct anastomosis, but you can occasionally see cystic duct stump leaks and also cut surface. So remember, with live donor transplants, it's essentially a partial graft, so there are going to be cut surfaces, and there can be leaks from the cut surface. I'll show you one slide on T-tube insertion and T-tube leaks. This has gone out of fashion, but occasionally still used. So a T-tube is placed to have access to the willow tree and then has to be removed, and this is associated with a leak at the time of T-tube removal, so typically that occurs two to three months after the surgery. So again, some cases that I've been involved in here is an anastomotic leak. You can see on the left-hand side, a scope's been inserted. You see a wire inside the recipient duct and then entering into the donor system just up here, but you can see a large amount of extravasation where the orange arrow is pointing towards it. A little bit better picture a little bit later on, and you can see that there's a kind of a gap here at the anastomosis. Enteropatics are filling. A wire is inside the donor segment, but extravasation occurring. So this was a 68-year-old male, and he was actually three weeks out from transplant, still in the hospital because of the prolonged postoperative course complications, and this healed after I placed the stent, but we'll come back to this case a little bit later. This is a cut surface leak. So here again, you can see the scope. You can see the biliary tree over here on the left-hand side of the picture with contrast being inserted, and way out here over on the left or the right side of the picture, but on the left of the patient, you can see extravasation where the orange arrow is pointing. This was actually a 37-year-old female who was 10 days post-op. She actually had received a split liver graft. So this is a patient that received the graft where the left lateral segment had been taken, presumably for a baby, and the rest of the liver was inserted, but you can see extravasation from the surface where they cut the left lateral segment, and placing a stent across the anastomosis here into the donor segment healed this up very nicely. And this is a cystic duct leak, and the surgeons kid with me that this must have been the fellow. So this is actually a surgical fellow leak because the attending surgeon can't possibly have done this. So you can see here, scope's been inserted. There are still staples in place. A cannula and wire have been inserted, contrast injected, and you can see extravasation occurring here. It looks like it's at the anastomosis, but in fact, you can just about make out there's another tube here. This is actually the cystic duct, and this picture shows it a little bit better because you can see the wire has been placed up inside, heading towards the left intrapanic system. So across the anastomosis, and this actually was the cystic duct, and up here, this is leakage from a cystic duct. And this was two days post-op. It healed by placing a stent across into the donor segment. Live donor is an interesting field because essentially the way I think about it, the ducts are just much smaller. So this is a 43-year-old female, three weeks post-op from a live donor transplant, and you can see the ducts just look smaller. So again, 12-millimeter scope. This is probably 4-millimeter, the recipient duct, and you can see the anastomosis is always much more proximal because the anastomosis is occurring up. This is a right lobe graft, so you can see there's an anterior and a posterior segment. The wire has been placed into the anterior, and you can see extravasation occurring right here at the anastomosis. This here, we're placing a stent across here into the anterior segment, but in fact, and as we'll talk about a little bit later, one of the problems with a leak is that you may end up with a stricture at the point of a leak. So this patient actually long-term, this area did not heal well and needed a Roux-en-Y hepatic jejunostomy. So what are the risk factors for leaks? So the term technical can mean many things. It, again, depends which studies you read, and particularly older studies talk about the type of suture that's used and the technique that's actually used for putting the ducts together. Mismatch can play a role, meaning there's a difference or discrepancy in size between the donor and the recipient ducts, sometimes as much as, you know, 4 millimeters on one side, 12 millimeters on the other side. So they need to try and match up the ducts to allow an anastomosis. The surgeon makes a difference, or at least the experience of the surgeon, and I'll show you some data from the A2 all-study with live donor to show you how that makes a difference. As we talked about already, the donor and the recipient can make a difference. And when we talk about the recipient, the hepatic artery, if they develop thrombosis or stenosis or there's disease of the hepatic artery, this can play a role in viral leaks. Donation after cardiac or circulatory death, so that's a DCD donor, or other donor factors in terms of the degree of anoxia can play a role in leading to a leak. Some surgeons will place a stent across the anastomosis at the time of surgery, and this is associated with leaking. T-tubes, we'll talk about in one of the subsequent slides, and then as we talked about already, live donor liver transplant, just essentially because the ducts are small and so the same applies to pediatric transplantation. So here's some data looking at the hepatic artery. So this is a paper from a few years ago now. So just to talk you through it, it was 43 patients that had an anastomotic viral leak after liver transplantation over a five-year period in a single center. And what they talked about were some were refractory, meaning that they required repeated intervention or ended up requiring surgery. So 17 out of the 43, so a little less than half, and you can see they had a worse outcome. So the survival curve is shown here. This is kind of a long period over 12 years, but the lines separate very early. So the black is patients that had treatment-responsive viral leaks, meaning they had a leak, but it responded to treatment. And the red is the refractory viral leaks, and you can see a big difference in the hazard ratio of almost four. So having a refractory viral leak was bad. And if they looked at the patients that had a refractory viral leak, hepatic artery disease was much more common. So overall, 50 versus 8%. And you can see here in the table below, they defined hepatic artery thrombosis, hepatic artery stenosis, or hepatic artery disease. And so this was the total, 43. But if you look at the two columns here, treatment-responsive versus refractory, and you can see that the percentage of patients that had problems with the artery was greater in the refractory viral leak group. And you can see they're all significant. So problems with the artery, if you identify that in a patient with a leak, really suggests that you're gonna need repeat intervention, even if you put a stent. And that suggests that perhaps surgery might be a better option, particularly as we mentioned in these cases, which are early on. What about donation after cardiac death? So this is a single center study again, but a little bit of a busy slide, but I'll go through it for you. So this is transplant using donation after cardiac death. So this, our two columns here, the column in the middle is patients that received the DCD graph, and the column over the right with the bigger numbers is the patients that received donation after brain death. And you can see the N is like 1,000 patients in total, about 5% were DCD graphs. And you can see that the biliary complication rate, this is all biliary complications, not just leak, was greater in patients that had a DCD. So you can see here, this is significant. But if you look at leaks alone, probably around two or three times a greater number of leaks in the patients that received a DCD. Strictures actually were a little bit more common, but actually not significantly more common. And the other problem with a DCD graph is ischemic cholangiopathy. And you can see a much greater risk in patients that received the DCD graph versus the DVD graph. And we'll talk about this a little bit later in terms of how many procedures were required per patient. So a DCD graph is a risk factor for a leak. So DCD graph, we talked about, and hepatic artery issues are two situations where you need to be a little bit concerned because the risk of a leak is going to be higher. One slide on a T-tube. This is essentially of historical interest. We published this almost 10, 12 years ago now in Pittsburgh. So here's a cartoon of a T-tube being inserted. This is the T-tube. It looks T-shaped. It's placed inside the duct. And you can see this comes out to the outside world. So you have access to the livery tree. The problem with this is at some point it needs to be removed, and typically it would be two or three months later. And there hopefully will be a track from here to the outside world so that when you remove this, it collapses on itself, that this hole will seal, but it doesn't always heal. So if you looked, again, we looked at many patients. The strongest risk factor for a leak with an odds ratio of 3.5 was the use of a T-tube. So, and when you look at when they occurred, at around about a mean of 102 days, that's when the time when the T-tube was removed. And this means that if you don't use a T-tube, the risk of a leak, particularly a later leak, is going to be less. Some data from A to O, just showing you that the surgeon is important. So here you go. This is on the left-hand side, the live donor recipient. And this is on the right-hand side, the live donor. So these are complications, biliary complications in the recipient and the donor in live donor transplant. And you can see that the first 20 cases as shown in this column, the complication rate is very high, but it gets less after you've done 20 cases, the surgeon gets better. And this is historical controls. And you can see that the bar leak I've highlighted here, it's almost double the rate of controls. In fact, triple the rate. If you look at in the first 20 cases, and it drops after the first 20 cases, meaning the surgeons are getting better, there's less risk of a leak, but still much greater than it would be for the regular transplant. And if you look at the donor, the risk of a leak in the donor is also pretty significant, 9%. So hepatic artery problems, DCD donor and live donor transplant, the leak rate is higher. Bar leaks do affect outcomes. So here's a graph as well as some data to show you that this is also again, lots of patients, almost 1000 patients. And you can see the dotted line are controls that did not have a leak. And the solid line is patients that had a leak and you can see the line separate pretty early. So having a leak is bad. This is patient survival. Leaks lead to patient and graph loss. And it affects also length of stay and 30 day admission rate. So here you can see the leaks are shown here in this column, 55 leaks compared to controls and the P values below. So there's increased length of stay, increased 30 day admission rate. And as we talked about already, hepatic artery thrombosis associated with leaks. As I showed you this patient earlier on, anastomotic leaks can cause strictures. So this again, we published years ago. This is exactly the same patient I showed you previously that had the leak after three weeks. You can see the leakage here. This is several months later. He's developed a pretty tight stricture at the anastomosis. And again, we published this a few years ago that a prior leak can lead to a stricture. And this has implications in terms of how many procedures are gonna be required for ERCP. So this is some data looking at how patients do if they get surgical revision as opposed to ERCP. So this again is a single center published very recently. 650 transplants done over five years. These were not DCDs, so these are all DBDs. 26 leaks, 69 strictures out of these 650 patients. We talked already about some of the risk factors, but recipient age in this particular paper and cold ischemia time is a risk factor for a leak. They favored surgical repair in this center. And the long-term outcome actually was favorable after they've repaired the leak. Some data also on strictures, which show that they're very favorable outcome after ERCP. So again, the table on the right here, and I'll just draw your attention to the middle section here, leak management. 19% of the patients were managed with stents, but the other 80% were managed with surgical, either revision of the anastomosis or conversion or revision of the ROO. And I can show you the survival curve, and you can see there really is no difference. So these patients, they had surgery to repair the leak and had good outcome after around three years compared to controls. One of the issues with ERCP is that even when you do a leak, there's going to be a requirement for more procedures. The stent's going to have to be removed, but also we talked about the development of stricture. So this paper looked at this, so this is 1,000 transplants, again, 55 leaks, 148 strictures. Majority of the leaks were early, and they repaired about half of them and about half they treated with ERCP. But the patients that required ERCP had an average of four, four and a half procedures to fix the biliary complication. And this is shown in graphical form here. You can see the solid line at the bottom is surgery and the dotted line here is ERCP. So you can see this is historical anyway that they were using PTCs. That number went down, so the ERCP number went up. But you can see it's still three, four procedures per patient to fix the complication. One of the things in live donor transplant, and to some extent, pediatric transplant, although there's less literature in this, is that ERCP is actually a little bit less successful. So this is published a few years ago now, 338 live donor transplants, the vast majority were duck to duck. They had 30 vial leaks. And of the 17 that underwent ERCP, 14 healed. So that's an 82% heal rate. But of those, more than half developed the subsequent stricture, and the other half developed the subsequent stricture. And of the ones that underwent a surgical hepatic or jejunosomy, they all healed. So this is another issue that live donor transplant, you can expect ERCP to be a little bit less successful. Shown there again in the table on the right. So I think, again, from an endoscopist point of view, I think management of vial leaks, I've put these two columns and I think this is reasonable. So medical ERCP management, small leaks early on can be managed just with drainage actually, but a stent will help. And I know Dr. Kote is going to talk about this in the next talk. ERCP is diagnostic, you end up doing it anyway, when there's a suspected vial leak. And because you're there, you place a stent anyway. I'm sure he'll talk about this in the next talk, but you need to put a stent. A sphincterotomy by itself is not enough. Success rate depends which study you read, but 80, 90% should be expected. Leaks can lead to strictures as we've talked about. And it does require multiple procedures. It's likely less successful in live donor transplant. I've not talked anything about the safety of ERCP and Dr. Shulman's going to talk about that. This is not a procedure without some risk. So particularly in a sick patient population, that needs to be factored into your consideration. The surgical approach, you can assess the abdomen, you can assess the artery, you can assess the liver, you can drain collections. The large leaks that are early, the surgical interventions prevents potential complications later. So, you know, you have bile bathing, the biliary tree, all the vasculature, the risk of developing a pseudoaneurysm, developing a mycotic aneurysm. Unfortunately, we've had a few disasters with patients developing infections from bile that's not been drained. And so this is something else to factor into your decision-making. Surgery is definitive. It's rare that you need to go back after you fix the leak with surgery, either revision of the anastomosis or with a ROO. So the success rate should be essentially almost 100%. So I'm going to sum up from the surgical aspect. I think early viral leaks, no doubt, they're a significant cause of morbidity after liver transplantation. We went through some of the risk factors, technical and donor-related, but if you ask surgeons if there's a leak identified within the first few days after a transplant, it's a technical issue. Even though, and I say this is a surgical tool, the majority of leaks can be managed by ERCP. If it's not a big leak and there's no other worrying factors, placing a stent should fix the problem. But I think surgery should be considered if there's a large leak and you identify it very early, it's usually better just to take the patient back to the operation room and fix it, because it will be a definitive fix and will prevent further problems. For those of the people listening in who are in the field will know that oftentimes if you don't fix a leak early, yeah, you can place a stent and the leak looks like it's healed, but it can lead to any number of complications, prolonged postoperative course, prolonged drains and sometimes going back to the operating room just to clean the patient out. If there's hepatic artery disease, stenosis, thrombosis, that can have its own issues in terms of requiring retransplant, but particularly if there's poor flow, that's another concern because the distal end of the donor duct can be ischemic. DCD organ already has problems with the biliary tree in terms of ischemic glangiopathy, but particularly a leak if it occurs in a patient with a DCD organ, another consideration to think about surgery. Live donor transplant, no doubt leaks are more common, get better as the surgeon gets more experienced, but the ducts are small, and again, something to consider because the ERCP is less successful and something that a surgeon can fix definitively. Thank you for your attention. Well, thank you, everybody. My name is Gregory Cote. I'm a professor of medicine at Oregon Health and Science University, and I'm excited to talk to you today about the approach to early bileaks and make a case for ERCP as our first-line approach. My only disclosure is that as a consultant for Olympus, and there'll be no relevant overlap with what I'll be speaking about today. Post-transplant leaks will occur in approximately 5% of cases, and most of these will occur within days or weeks of the transplant surgery itself. The vast majority of them will occur at the biliary anastomosis, although we'll discuss some of the other potential locations here momentarily. The stakes here are very high. This is a little bit more of a complicated conversation than that of a clinical trial, and it's a little bit more complicated than that of a clinical trial. It's a little bit more of a complicated conversation than that of a typical post-cholcystectomy leak. The primary reason for that is that early anastomotic complications are associated with a higher odds of graft failure. A leak is increased risk with a hazard ratio of 1.3, whereas a stricture alone is 1.5. When you combine a leak with a stricture, the hazard ratio, again, is 1.5 times as likely for graft failure than a comparable individual without a biliary anastomotic complication. You have several unique considerations when you're thinking about post-transplant leaks and strictures. First, you want to think about what is the type of biliary anastomosis that we're dealing with. The majority of anastomosis in an adult population will be a duct-to-duct anastomosis. This results in traditional biliary access, a pretty standard approach to ERCP, assuming the patient does not have altered foregut anatomy. Of course, ERCP in this configuration is associated with a risk of post-procedural pancreatitis, as well as the other usual complications associated with ERCP. The biliary anastomosis in this place is very close to the vascular anastomosis, and this becomes relevant in our risk evaluation, as we'll be discussing. Now, in an adult population for patients with PSC, or perhaps a retransplant, or for other reasons as determined by the surgeon, you may have a choledocojejunal anastomosis. This would be a typical anastomosis that you would see, for example, after something like a Whipple procedure as well. The advantage of this approach from the endoscopist standpoint is that you're no longer in continuity with the pancreatic ducts, so there's no risk of post-procedural pancreatitis. However, endoscopic access can be more challenging. The other caveat with this anatomic subtype is that the biliary anastomosis is not as proximal to the hepatic artery anastomosis, so some of the vascular complications may have a lower risk. Another thought here is thinking about was an internal stent left in place at the time of surgery or an external drain, such as the T-tube. Now, I'll make a note here that there really is very little to no evidence to support the use of universal T-tubes in preventing biliary complications. There have been three randomized controlled trials on this since 2010 that included 277 participants, and the odds of preventing an anastomotic complication, either a leak or a stricture, was not significantly different between the two populations. However, if you have a patient with an intact and patent endobiliary stent, so a stent that was placed internally at the time of surgery or a T-tube, and in spite of that, leaking is still occurring. While I don't have level one evidence to support this, it is logical from the endoscopic perspective to consider that the role of a second ERCP may be less efficacious. The patient already has a drain in place, and the likelihood of us being able to address a leak in that setting is probably not as successful as a patient who has no drain in place at the time of presentation. Not all bile leaks are the same. I mentioned that the majority of these will be occurring at the anastomosis, and I would assume that's the case unless there's some objective evidence to suggest otherwise. Usually, this is related to technical issues. At the time of the biliary anastomosis itself, the sutures may de-hiss, and often this is occurring in the setting of local ischemia. Now, there can also be leaks at the cystic duct stump, and remember, this may be at the donor or the recipient cystic duct stumps, and there may be both. There could be a broader issue of ischemia to the extrahepatic duct, and if that is suggested by cross-sectional imaging such as a contrast-enhanced CT, CT-NGO, or MRA, one would have to have pause in terms of the viability of the bile duct around the area that is leaking. Other potential culprits or locations for leak include the cut surface of the liver and aberrant ducts that could be injured at the time of the transplant itself. These are less common in the setting of a liver transplant than what we'll see, for example, after a cholecystectomy. You will occasionally encounter delayed leaks in the setting of a patient who may have had an external drain. The drain was removed, but in fact, the tract was immature, and so the patient then has a bile leak at that time. Now, early anastomotic leaks will lead to or be associated with other problems. I've already mentioned the issue of vascular complications. The two biggest issues to consider here would be hepatic artery thrombosis or portal vein thrombosis, and as a result of the inflammatory milieu around particularly the arterial anastomosis, one can run into issues with aneurysmal dilation and even catastrophic hemorrhage in certain cases. This is why the transplant surgeon is highly concerned with the presence of an uncontrolled leak. They want that leak to be controlled as quickly as possible. Of course, that's in the best interest of the patient as well. Other issues that can develop in association with the leak include renal failure due to the systemic inflammation and just overall stress on the body, secondary infections, so infection of biloma fluid or any other infectious complications that can occur in the area of the surgery, as well as rejection of the graft that is also associated with the presence of an early anastomotic leak. As I've mentioned, graft survival is associated with the presence of these biliary complications, so this isn't a simple plumbing issue to be taken care of and not thinking about these other consequences. On the left curve here, you'll see in blue patients who have no biliary complications, their chances of graft survival at five years is significantly higher than the red line, which are patients with any biliary complication. Breaking this down into a more granular subtype here, you again still have the patients with no complications in blue, but once you start to have the combination of leak or stricture or simply leak alone, you can see that these groups are at the highest risk of graft failure and many of those are occurring within the first one to two years from transplant. What are the potential clinical pathways? Well, you have a patient who you are suspecting a bile leak. First of all, sometimes these things are fairly obvious by the presence of abdominal drains that are clearly putting out bile. Other imaging studies can also confirm this, such as a HIDA scan. In the setting of a proven bile leak, you have the option of proceeding to an ERCP, which is what I'll advocate for here momentarily. You can also consider the placement of a percutaneous transhepatic bile duct drain. For patients who may have altered anatomy or are simply too ill to undergo a sedation that is required from ERCP, this may be a plan B. Then, of course, you can go straight to reoperation to clean the area out and try to reconstruct the biliary anastomosis. If the surgical option here were straightforward, this may be a more contingent debate. However, surgery in the setting of a patient who has an early bile leak after surgery is quite complex. The risks there are also not to be minimized. Now, if you're unsure of the bile leak, if the clinical picture is not convincing, I recommend proceeding with the HIDA scan, which would be your standard first-line diagnostic test or an ERCP. Again, in the interest of time, however, if your level of suspicion is high, this is one of the particular scenarios where one could advocate for a diagnostic ERCP. For example, if the patient, this is over the weekend, access to these other nuclear MR imaging studies is not available, a diagnostic ERCP, I think, would still be considered within the standard of care for this high-risk patient population. Now, generally speaking, in terms of grading leaks, they're graded in a dichotomous manner. If you have a high-grade leak, and I really don't demonstrate this in either of these fluoroscopic images, but you can see a significant amount of contrast that is already extravasated here in this right fluoroscopic image, generally speaking, a high-grade leak is a leak that you can identify before you begin to fill the intraepatic duct branches. A low-grade leak, as is shown here, is only identified after significant opacification of the intraepatic ducts. While this is an older study from the early 2000s, I still think it's actually one of the highest quality studies evaluating the role of endoscopic treatments for bile leaks. Now, this is a cohort of patients who had bile leak from multiple etiologies, so this is not a transplant-only population, but in this group of 204 patients, the majority of these were after cholecystectomy. As a result of that, the majority of the leaks were low-grade and found at the cystic duct stump. There were also Leuschka leaks and other locations for bile leaks, as shown here. What I like about this study is that the authors dichotomized the population into these two categories, a low-grade leak and a high-grade leak. Now, for the purposes of this presentation, let's assume that all anastomotic leaks are high-grade because of the nature of the surgery, the timeliness of it, etc. In this situation, the treatment for these patients in this earlier study includes what most of us would be considering standard of care. The intent of ERCP is to reduce any sort of pressure gradients associated with bile flow into the duodenum. In order to do that, we perform a biliary endoscopic sphincterotomy, or BES, as shown here, and we typically will place a stent into the bile duct. The standard first-line approach would be a 10 French bile duct stent if the duct is large enough to accommodate. Again, this is just based on flow dynamics, Laplace's law. The bigger the diameter of the stent, the more rapid flow you will have through it, and the greater likelihood you will have of equalizing pressure between the bile duct and the duodenum. In these high-grade leaks, you can see that the success rate was quite high, and even in the patients that did not resolve following the initial ERCP procedure, the majority of those that had a persistent leak resolved with a second ERCP. The results are even better in the low-grade category when you consider that in that population, only a biliary sphincterotomy was performed. Most of us err on the side of caution and will place stents in these patients to minimize the likelihood of a persistent leak. Now, when you look at cohorts of patients that have been presented following transplant, the success rate admittedly is a bit more variable. I've highlighted here three recent studies evaluating ERCP in patients presenting with early anastomotic leaks, and the success rate ranges from 60 to 90%. As I would estimate here, there's certainly a better than 50-50 chance of improving slash resolving the leak with the ERCP. Sometimes the ERCP can shed interesting or important information about the clinical course, such as findings of ischemic cholangiopathy, such as the presence of a very wide leak, very large bore leak. In that case, it can be helpful to the surgeon to understand that the likelihood of controlling that without surgery is going to be very low long-term. I've demonstrated the basic concept here. You can see the surgical staples on the abdominal wall. On the left side of this fluoroscopic image, there are some staples near the biliary anastomosis. You can see in this case, a plastic stent that has been placed across the area of both stricture and leak in this case. Again, the basic concept is to shuttle that bile down into the duodenum where it belongs and to reduce any, to the greatest extent possible, any pressure gradient between the bile duct and the duodenum. I won't show you any literature or examples of placing fully covered metallic stents. The advantage of fully covered metallic stents is that they are larger in diameter, so you are more likely to oppose the wall of the bile duct. These have clearly been shown to have comparable efficacy in treating strictures after liver transplant. They have not been compared head-to-head prospectively in resolving bile duct leaks. In my practice, I would typically reserve this for a patient who has a very high grade leak and I'm really feeling that we need that larger diameter stent to oppose the wall of the bile duct, but this is really just based on experience, not based on level one evidence. Now, the most important argument that favors ERCP in this conversation is that the data do not support taking these patients back to surgery first. Here we have two Kaplan-Meier curves. I'll direct you first to the curve on the left. So you have patients here that did not require any treatment for their bile leak, so it was low grade and managed conservatively versus managed by way of radiologic management. This is typically in this series was mostly percutaneous, but also endoscopic versus surgery first. You can see that the likelihood of graft survival was similar in those two populations. Similarly, when you isolate a population that was treated for stricture after post-transplant, endoscopic first versus surgery first, you can see that graft survival is fairly comparable between the two. In fact, in the early stages in the first year, actually early surgical intervention trended towards a higher rate of graft failure and the reason for that was presumably due to the fact that these were sicker patients that prompted a surgical first approach or perhaps the stress of the surgical intervention and delaying their overall recovery from their liver transplant. Now, there are negatives of ERCP. So first of all, the procedure may be technically unsuccessful. We may not be able to traverse the stricture or leak if it's very, very high grade. We do now have the ability to perform endoscopic ultrasound and try to access the bile to integrate, but if there's a complete dehiscence of the anastomosis, it's highly unlikely that we can handle this by way of endoscopic approaches alone. We can perform rendezvous procedures with our radiologists, so a percutaneous approach first, and then we can use their anti-grade guidewire to traverse an area of high-grade dehiscence. There may just be simply technical failure, inability to get to or into the major papilla. We obviously in certain cases can cause post-procedural pancreatitis, even with the best of hands, and this can further cause stress on the patient's overall recovery. Remember that when you are performing an ERCP, you are injecting contrast through an endoscope that has now been in a duodenum that has normal bacterial flora, so there is a chance of secondary infection. These patients should certainly be on prophylactic antibiotics and kept on antibiotics for several days post-procedure to make sure that we don't cause a secondary abscess formation and all of those vascular issues that I mentioned earlier in my talk. The procedure can also be complicated in rare cases from post-phankterotomy bleeding. There is an argument to try to avoid sphankterotomy in certain cases and just place a stent. I think the jury is still out on whether stent alone is sufficient, but there is certainly a compelling argument for that. Since the risk of bleeding in my estimation is relatively low, I feel like it's most important to provide the most optimal endoscopic and aggressive treatment for their bile leak first and foremost. Some of the late negatives of ERCP, by definition, when we place stents, we need to repeat the procedure to remove it and assess for stricture formation. If the patient develops ischemic cholangiopathy, we have potentially avoided surgery in the short term, but then surgery may be required later down the road for recalcitrant strictures. Retransplantation in these cases is very rarely done in those early phases because, again, the majority of these patients will recover, although it can be a tumultuous course. In summary, in the absence of a concomitant indication for surgery, such as having to repair a vascular and asthmatic issue, ERCP is a reasonable first step to address a post-operative bile leak. Consider the morbidity of early repeat surgery in this conversation and also make sure that the patient is aware that the stakes are high here. This isn't simple. While there's a leak, we're going to place a stent. You'll be fine. There's a lot of things that need to be monitored in those days following the procedure, such as secondary infections, such as vascular complications, and then also bearing in mind that when leaks present early, there may be delayed complications later, such as ischemia and recalcitrant strictures of the duct. Careful monitoring these patients by both the endoscopist as well as the transplant surgical team is critically important. Thank you very much for the opportunity to present. I wanted to thank the course organizers and the AASLD and ASGE for the invitation to speak today. I'm so honored to be here and to be presenting on this topic. As you heard, I will be speaking on optimizing the safety of ERCP in liver transplant recipients. Here are my disclosures. Before we get started, I just wanted to quickly review my outline for the talk as we will be covering a lot of ground. I thought it would be important to start with a quick historical review of liver transplantation and a bit about where we are with organ donation currently, including the changing patient population and the projected changes and indications over the next decade. I will then move on to a review of the most common surgical procedures, the most common surgical procedures, mostly focusing on the anatomy and how that will affect our procedural strategies for intervention. I will then spend the majority of the talk on ERCP with a brief discussion on indications, but spending most of the lecture discussing how we can optimize safety in the pre, intra, and post-procedure settings. Let's get started with a brief historical look at liver transplantation. Dr. Starzl performed the world's first human liver transplantation in 1963 and then the first successful liver transplantation in 1967. All of this was at the University of Colorado and Denver VA hospitals. He then joined the University of Pittsburgh Medical School in 1981 as a professor of surgery where he studied cyclosporine, which really transformed transplantation from an experimental procedure into one that gave hope. It was really his development of cyclosporine and in combination with steroids that offered a solution to organ rejection. He just died approximately five years ago. Other historical landmarks in this field include the creation of the National Organ Transplant Act in 1984. This prohibited buying and selling of organs and also established what we now know as UNOS, the National Transplant Registry and Allocation System, in addition to organ procurement organizations. Just to put this into perspective, the first small bowel transplant was in 1989, as was the first living donor liver transplant. The first hand transplant was in 1998 and the first face transplant in 2005. Several of the following graphs and slides are courtesy of Dr. Chris Sonnenday, the director of the liver transplant program at University of Michigan and also a very close friend. Here you can see the supply and demand on liver transplantation over the past two decades. What is striking to me about this graph is the fact that we are only just starting to see the waitlist decrease as more donor livers are harvested. But clearly, we have a long way to go. To address this organ shortage, there are some movement to use organs that have not been typically used such as those donated after cardiac death or with alternative preservation strategies. Also to identify alternative sources of donor organs, and this is where living donor liver transplantation has played a major role, and we will briefly talk about this later. This may also play a role in what types of biliary accomplishations we may start to see more of in the future as interventionalists. While numbers of patients undergoing transplant are slowly increasing, we are also seeing that the expected median survival of liver transplant recipients is also slowly increasing over time. Furthermore, the median age of adult transplant recipients is steadily increasing. As you can see in this graph. This graph perhaps reflects the most impressive change, which shows the evolution of indications for liver transplant over two decades and predicts where we will be in about five years from now. As you would expect, our hepatitis C treatment has tremendously improved. This is no longer the leading reason for transplantation, but instead it is NASH. This really reflects what we are seeing with the obesity epidemic within the United States. As you can see here, the prevalence of obesity has been increasing. From around 2000 to 2018, the prevalence of obesity increased from 30 percent to 42.4 percent. The prevalence of severe obesity has increased from 4.7 to 9.2 percent. The impact of obesity on morbidity, mortality, and healthcare costs is quite profound. The other thing to keep in mind is that endoscopic interventions should be carefully planned out in any patient with obesity. I just want to highlight a thought-provoking piece that we wrote after being asked to do an ERCP in a patient with a BMI of 93. Here we highlight the barriers, limitations, and risks that these patients experience for what would typically be seen as routine care. Everything ranging from sedation planning to preoperative optimization to whether or not the table weight is sufficient should be considered to safely perform endoscopy in patients with obesity. Next, we are going to move on to a brief review of common surgical procedures specifically focusing on anatomy. The two most common surgical procedures performed for liver transplant are colodocostomy, which is the anatomy that we see most of the time as interventionalists. This accounts for approximately three-quarters to nearly 90 percent of all liver transplants. The other anatomy is known as Roux-en-Y colodocojejunostomy. But as you can see, this accounts for far less than a quarter of patients that we will see in our practice. Starting with colodocostomy, this can be performed in either the conventional technique on the left or a piggyback technique in which the ductal anatomy reconstruction is identical for our purposes, but the venous outflow is different. This is a duct-to-duct reconstruction. This is the default method of reconstruction for liver transplant. This requires that all native bile ducts be healthy and without any distal obstruction or strictures. The benefit surgically is speed of reconstruction. But for us endoscopists, this anatomy provides considerable ease of interrogation and treatment specifically during ERCP. Here you can see that the duct-to-duct anastomosis can be reconstructed in a variety of ways, including either being sewn end-to-end or with a spatulated connection. There may have implications for what we see as endoscopists but are beyond the scope of this talk. Moving on to Roux-en-Y colodocojejunostomy. As you can see here, and this is indicated when the native duct is not healthy, conditions where this may be common is extrahepatic biliary disease such as PSC, biliary atresia, Corollis disease, or entrauma. This can also be useful if a severe size mismatch is acknowledged between the native and the donor bile ducts. The major advantages of this anatomy is clearly versatility, but there's some obvious disadvantages as well, including operative time, additional anastomosis and dissection required, and the fact that more invasive diagnostic procedures may be required such as PTC, where ERCP is not feasible or not able to be attempted. I just want to quickly round out the picture by showing you the anatomy for a living donor transplant patient, where the right lobe is typically removed and placed in the recipient. Here you can see the recipient and donor liver. What is really amazing about this is how much the donor liver regenerates over just a short time. In just two months, you can see considerable growth. Other surgical considerations that may affect biliary complications down the road include injury during harvest time, cold ischemia time, surgical technique, and also the integrity of the portal vein and the hepatic artery anastomosis. Now we're going to move on to the bulk of the talk, where we will be focusing on ERCP, including indications following liver transplant, and specifically on the pre, intra, and post-procedural considerations that we can take to really optimize safety in this patient population. What are the indications for ERCP following liver transplant? We know that biliary complications occur in up to 25 percent of patients, and the early incidence is estimated to be somewhere between 3-15 percent, while the late incidence is a bit higher, around 10-20 percent. These include a variety of different entities, including biliary strictures, bile leaks, choledocholithiasis, biliary casts, bilomas, mucous seals, hemobilia, etc. But the most common complications we see in practice, and those that I will spend the most of my time on, are anastomotic strictures and bile leaks. It is important to note that bile leaks are actually a strong predictor of the later development of anastomotic strictures, and also that biliary complications in general have been associated with graft loss and the need for retransplant and have also demonstrated to diminish overall survival. Oftentimes, these complications present as an acute rise in AST or ALT or bilirubin, and this would of course signal us that we should be worried about cholangitis or a bile duct stricture. However, your differential should be broad, including other phenomena like vascular thrombosis or stenosis, dehydration or sepsis, acute rejection, etc. Make sure that you work closely with your transplant surgical service and your transplant hepatologist to think carefully about the differential in any particular clinical scenario before jumping to an ERCP. Additionally, a slower rise in liver function tests may point more toward graft failure or acute rejection, or potentially decreased perfusion to the organ. Again, it is important to keep the clinical scenario in mind before jumping to a procedure. Another important historical clue into what may be causing the clinical presentation is what type of transplant the patient had. As you can see here, biliary complications are much more common in living donor liver transplants, and so this may help narrow your differential depending on your evaluation. Once you determine that the patient does in fact have a biliary complication, can you jump right into an ERCP? The short answer is only after you understand critical components of the patient's history. First, we need to know if we can reach the biliary anatomy reliably, i.e., does the patient have a duct-to-duct anastomosis or a hepatico-jeunal anastomosis, as that may require expertise that your center does not have, and we'll touch on that in a bit. We should also know the timing since surgery, as this may dictate what type of intervention or stent we could offer. Furthermore, it is important to know if we can reach and or bridge a structure based on cross-sectional imaging. The size of the ducts, the donor and recipient is also important when selecting a stent and the type of endoscopic therapy that will be offered may depend on all of the aforementioned data. It's important to know all of this before proceeding. Of course, before you pursue an ERCP, it's important to remember that antibiotic prophylaxis in this patient population is important. The ASGE recommends that antibiotics be administered before ERCP in patients who have had liver transplant. This has certainly been my personal practice. The question of whether we can reach the biliary anastomosis is critically important. The answer for duct-to-duct anastomosis is yes. But with Roux-en-Y Colodoco jejunostomies, the answer is oftentimes rarely, or without the appropriate resources in your hospital could be incredibly difficult. This is both a diagnostic and therapeutic challenge to first-line approach with traditional ERCP. To get to the hepatico-jejunal anastomosis, there can be angulated loops, adhesions due to previous complex surgeries, failure to identify and successfully intubate the correct limb to get to the anastomosis. Finally, when you reach the anastomosis, cannulation can also be challenging, and getting into the appropriate bile duct branches can be challenging. Device-assisted enteroscopy, including single or double balloon and spiral enteroscopy has allowed deeper small bowel intubation and access to bilioenteric anastomosis more easily than what we had known previously. One recent meta-analysis that included 155 studies and six relevant trials was analyzed and included a total of 132 patients. The pooled success rates among patients with liver transplant were enteroscopically somewhere around 80-100 percent, diagnostically 75-100 percent, and only about 67 percent therapeutic success with an overall success rate of somewhere between 60-100 depending on the study. But there's no doubt that these can be challenging. Even using device-assisted enteroscopy, finding the colodoco-jejunostomy can be challenging enough and takes considerable patience. Here you can see a recent case that we did, and I've intentionally left this video a little bit long and somewhat lengthy so that you can get a sense of how difficult this procedure can be. You can see our scope circling each fold of the jejunum several times, trying not to miss what you will ultimately see as a very hidden biliary orifice. Oftentimes with these procedures, I like to put a cap at the end of the scope because it's really much easier to pleat back the folds of the small bowel, and also to use a lot of water and not necessarily air or CO2. Here you can see finally a glimpse of the anastomosis and then eventually once you can secure it within your cap, you're able to actually access it and perform the ERCP. But even passing down instruments through these long thin scopes can be challenging, and eventually intubating and getting into the appropriate bile duct can be challenging. We will be focusing mostly on ERCP and duct-to-duct reconstruction, as this is the majority of what we see, and we can uniformly intervene upon this at various centers, independent of really your expertise with enteroscopy. We will start with anastomotic strictures, and as I mentioned before, these can occur early or late after transplant and the timing is important. We'll start with a discussion of early strictures. Early strictures are typically those that occur with a fresh anastomosis with post-operative edema. These you can typically stent, but oftentimes should not be stenting with dilation because that could blow out the anastomosis, and usually we use plastic stents for these. The biggest diameter possible is desired, and that's typically 10 French, and oftentimes we're performing sphincterotomy with these as well. When we think about late strictures, on the other hand, these most always involve balloon dilation, and the selection of the stent type, plastic or metal, may be appropriate depending on the location of the stenosis. But regardless, require multiple procedures, and plastic stents are commonly exchanged every one to three months, whereas covered metal stents can oftentimes stay in three to six months. As many of us are familiar, these stents come in all different sizes and lengths. Plastic stents that are used in the biliary tree really range from seven French to 10 French, and they can either be placed as a single stent or multiple stents, which we will get into shortly. The length is also variable and ranges from three to 15 centimeters. These can typically be placed down a variety of different endoscopes, including duodenoscopes, 1T wide-bore upper endoscopes, pediatric colonoscopes, or balloon endoscopes. Only fully covered metal stents play a role in these types of anastomotic strictures. Really uncovered stents have no role here. These metal stents range from eight millimeters to 10 millimeters in width, and four to 10 centimeters in length. Only a single metal stent is placed at once. These are traditionally placed either through a duodenoscope or a wide-bore 1T gastroscope. Although some of the newer iterations may be placed down a pediatric colonoscope. Here you can see some recent examples of these types of strictures. These can be quite tight as you can see in both of these two examples illustrated here. Sometimes these can be so tight and acutely angulated that the use of colodicoscopy can actually be very helpful. Here you can see a video of one such case. You saw a very tight angulated stricture on the cholangiogram. Here you can see us under direct visualization, taking a flexible ERCP guide wire and trying to direct it toward the opening. But the guide wire keeps bending at the level of the anastomosis. Here you can really see how this technique and technology has allowed us to advance a wire even through the tightest of strictures. Then once we are able to access beyond it, we advance a significant amount of guide wire through and eventually are able to remove the colodicoscope and over the guide wire place a long stent. Furthermore, as I previously mentioned, the use of multiple stenting with plastic stents has been quite successful. This was the first prospective demonstration of the efficacy and safety of sequential multi-stenting protocol. A key limitation of this study is the lack of a comparative group that was treated according to traditional stent exchange approach, but the data is still quite compelling. Here's a Kaplan-Meier curve of recurrent free survival in a population of patients who had multiple stents versus naive patients. This can be particularly helpful approach in liver donor related transplant in my experience, where the strictures can be much more proximal and challenging to reach and certainly not amenable to covered metal stent placement. Here's one such case. As you can see here, the stricture is very proximal and we started with a single plastic stent. Eventually over time, we increase that to four stents and eventually even five stents. Here you can see what that looks like during an ERCP. First, you have to place four flexible guide wires through the stricture. This can be quite challenging to do especially if it is tighter angulated. One method that I've used in situations like this is to actually advance a cytobrush catheter over your first wire and then you can remove the existing wire and use that second channel for a second guide wire that can be advanced alongside the existing wire. Here you can see that we already have three stents in and over that fourth wire, we're about to place the fourth stent. These should commonly be placed somewhat perpendicular to each other so that you don't advance a stent up the bile duct orifice because that could be problematic. In looking at a comparison of covered metal stents with multiple plastic stents, what you can see from this publication in GIE a few years ago is that covered metal stents were comparable for liver transplant and asthmatic stricture resolution and allowed for fewer procedures with a positive effect on the overall cost for that reason. But the duration of treatment with covered metal stents really still needs to be investigated. Furthermore, a meta-analysis involving metal versus multiple plastic stents investigated technical success, stricture resolution, recurrence, and complications. It was deemed that based on low-quality evidence overall, the authors could really not draw any reliable conclusions on the superiority of multiple plastic stents to covered metal stents. Even though shorter treatment times and fewer ERCP procedures supported the use of covered metal stents, whether one technique has well-defined advantages over the other really remains unclear from this data. There are many considerations when determining which is the appropriate choice of stent for a given patient and a given stricture. As I mentioned before, plastic stents are cheaper. They're also able to be delivered by a variety of different endoscopes that we alluded to earlier. However, they oftentimes do require multiple procedures and require being exchanged every 1-3 months so that the patient does not come back with cholangitis or obstruction. Metal stents and specifically only fully covered metal stents play a role here, may not be long enough to bridge the stricture in certain cases. The duct size and the luminal anatomy may really limit their use. But overall, these may require fewer procedures. One article that was published this month in liver transplantation was examining the use of what the authors called inside stents. These were plastic stents that were placed into the bile duct and did not require sphincterotomy or any long-lasting transpapillary component. The premise of the study that is while intraprocedural risks such as bleeding and perforation are well-described in ERCP, the creation of a sphincterotomy also disrupts the inherent sphincter function and potentially leads to increased duodenobiliary reflux, contributing to both stent occlusion and ascending cholangitis. Transpapillary stenting similarly disrupts the sphincter function. Potentially, the use of inside stents may obviate the need for either of these two interventions. This study did appear to decrease the incidence of cholangitis in their observational patient population. We also wrote an editorial on this manuscript and think it may guide exciting work moving forward. The endpoint of dilation and stenting is really not well-defined, but includes both fluoroscopic resolution and the ability to pass an inflated balloon catheter easily through a stricture. It is also important to keep in mind a differential when doing an ERCP with the assumption that a patient has an anastomotic stenosis. Other entities that can mimic the clinical presentation include choledocal lathiasis, biliary casts, and bile leaks. Here you can see a case of a patient I had with both an anastomotic stricture, but also several large stones proximal to the stricture. Given this concern, we elected to place a covered metal stent and then do cholangioscopy with electrohydraulic lithotripsy through the stent. Eventually, we were able to clear all of the stones. Here are some of the images of the stones within the stent and just proximal to the stent. Finally, our full clearance of his bile duct. Biliary casts are another phenomenon to be aware of. Morphologically, biliary casts are a similar shape to bile ducts, appearing as a hardened dark material in the biliary ductal system. Biliary casts can really prevent bile drainage resulting in biliary obstruction and also inducing biliary tract infections. Here you can see a short video of the removal of one of these casts. As you can see here, the cast is starting to come into view. Our biliary occlusion balloon is up into the bile duct and we're pulling back on it as this large cast is being delivered into the lumen of the duodenum. As you can see, it takes the shape of the actual bile duct itself. Finally, there are certain types of strictures which are non-anastomotic. These commonly occur as a result of ischemia or hepatic artery thrombosis, but other etiologies can also predispose to these. It is important to keep your differential broad. This is a nice schematic showing the classification of the anastomotic regions of the biliary tree affected by non-anastomotic biliary structures. Oftentimes, the location can really help us determine what the etiology is. Here's a cholangiogram that also nicely demonstrates this phenomenon, clearly in a patient with diffused ischemic biliary drainage damage after a liver transplant. Bioleaks are also complications that we will see commonly in our interventional practices. This is primarily a clinical diagnosis. The only true diagnostic test is really cholangiography. There are really no reliable serum laboratory indicators and sampling of actual fluid or even obtaining a HIDA may be helpful, but it is really not 100 percent sensitive or specific. Here you can see a cholangiogram demonstrating a small leak right at the level of the anastomosis. Just this past week, we had a patient with a distant history of a liver transplant, which was complicated by a late hepatic artery thrombus, who developed a much more impressive leak that you can see here on cross-sectional imaging. The patient underwent percutaneous strain placement for source control, but we also did an ERCP for transpapillary stenting. As you can see here, there's a massive leak at the biliary system. The most important part of bioleaks is really controlling the infection. Bilomas or bilia society should be drained and your surgical team should consider a washout if the clinical condition fails to improve. It is also important to establish control of the biliary leak, which may mean percutaneous drainage or ERCP, but also may vary by reconstruction and operative management really may be required early. It is important to ensure adequate healing and to monitor for the development of a stricture as leaks are a risk factor for the development of strictures as we mentioned before. In summary, when we think about performing an ERCP in a patient following liver transplant, there are several considerations that are required to really optimize safety. You should know the date and the type of the liver transplant, any surgical complications or other risk factors that the patient may have, their luminal anatomy, ductile and asthmatic anatomy, and of course, their platelets and coags which may be more deranged in this patient population. Furthermore, if the patient has had any prior endoscopic procedures, you should know what stents they currently have in place and make sure that a follow-up plan is well-coordinated and a plan for endoscopic failure. If endoscopy is not successful, there are non-ERCP options that exist as we've discussed, including a PTC and also a retransplant. In conclusion, the number of liver transplants continues to slowly increase with a changing patient population. Biliary complications are common, and endoscopists should be familiar with all types of surgical anatomy, and the decision to proceed with ERCP should really be guided by the anatomy and the expertise of your center. But prior to pursuing the timing of surgery, type of transplant, and surgical risk factors are important to know how to perform the ERCP safely. Of course, these decisions should all be guided by a comprehensive multidisciplinary team. Thank you very much. That's all for this session. I would like to thank all the participants for their participation and their attention. I'd also like to thank Dr. Jawad Ahmad and Dr. Jennifer Christie for co-directing this course with us this year. I think it's been a wonderful day. Thank you.

biliary strictures

PSC

cholangioscopy

diagnostic accuracy

ultrasound

MR elastography

portal hypertension

liver diseases

intragastric balloons

non-alcoholic fatty liver disease

obesity

liver transplant outcomes

bariatric surgery

NASH

fibrosis post-transplant