So good afternoon, everyone. We'll get this exciting session underway. Thank you for coming to the session, Master Regulator of the Liver and Critical Illness. My name is Dean Karvelas. I'm from the University of Alberta in Edmonton, Canada. It's my pleasure to introduce our first speaker, a good friend of mine. Ram Subramanian is the Medical Director of Liver Transplantation and the Head of Liver Critical Care Services at Emory University in Atlanta. And he's going to be giving the first talk this morning on, sir, I can't see the title of your talk, Ram, sorry. I'll let you. The Liver-Brain Connection. The Liver-Brain Connection. Good morning, everybody. Can you hear me? It's just a quick mic check. We're good? So the topic of conversation will be hepatic encephalopathy, and I have no disclosures regarding this presentation. So from an outline standpoint, number one, I want to focus on classifying liver failure. I think it's important, right off the bat, as you see a patient with liver failure, with advanced hepatic encephalopathy, the first decision point is that, is it acute liver failure, I'll use the acronym ALF, or is it acute and chronic liver failure, ACLF? We'll sort of tease it out initially. And then I want to go into pathophysiology of hepatic encephalopathy. And in that category, we'll take some time to distinguish between the hepatic encephalopathy of ALF and the hepatic encephalopathy ACLF from a pathophysiology standpoint. And then finally, we'll finish off with management. Given the differing pathophysiologies of HE in these two disease processes, how do we approach the management strategies in ALF versus the management strategies in ACLF? So that's the general outline for the next 20, 25 minutes. So let's talk about classification of liver failure. So acute liver failure is defined by acute hepatic dysfunction in the absence of chronic liver disease. And that's an important sort of point to be noted. So a classic example would be 25-year-old, tonic overdose, no chronic liver disease, presents with severe hepatic encephalopathy. So that's HE in the context of ALF. So this disease process is characterized by acute hepatic necrosis, typically severe coagulopathy, encephalopathy. And the one take-home message in this disease process is that you have the risk of developing intracranial hypertension and ICP elevations. And that's an important, again, take-home message there is ALF, hepatic encephalopathy, can lead to intracranial hypertension and herniation. In contrast, what you see more commonly, and I'll use the term ACLF synonymously with decompensated cirrhosis, in this disease state, this is characterized by decompensation of prior cirrhosis, long-standing cirrhosis. A classic example would be hepatitis C cirrhosis, NASH cirrhosis. And this disease process is characterized, as some of you may know, with multi-organ system dysfunction related to chronic portal hypertension. So the disease process is chronic. And the hepatic encephalopathy associated with an ACLF is typically not associated with elevated ICP. And so as you can see, the clinical scenario is such that you have to, right off the bat, when you see somebody with liver failure, please make sure, with encephalopathy, you make sure that are you dealing with ALF or are you dealing with ACLF? Because it's a big distinguishing characteristic right off the bat. Just a brief word, especially in the hepatology community, we use this grading score. I just wanted to share this with you, the Westhaven criteria. It's still used to this day in certain centers. And it's a subjective criteria, but stage one is characterized by mild confusion, slurred speech, and then stage four is coma. And you can fill in the gaps in between, but this is a grading scale that is used both, to some degree, in ALF and in ACLF, and it's a useful scale to use as you're communicating between providers. So now let's sort of go into the pathophysiology of hepatic encephalopathy. In part, and I emphasize in part, it's hyperammonemia. It's not the whole story, because especially in cirrhotics, you can have stone-cold normal ammonia levels, and it can be florally confused, requiring intubation. But ammonia is a driver of encephalopathy. But as you look at other potential mediators, there's ample evidence that a systemic inflammatory state of the surge response, if you will, and infectious triggers are also mediators of exacerbating hepatic encephalopathy, both in ALF and in ACLF. And then finally, other mediators, and this is one of them in particular that I'll talk about is endogenous benzodiazepines. These are potential mediators that can play an important role in cirrhotics, in acute and chronic liver failure. And I'll talk about albumin dialysis briefly, and the mechanism of response to albumin dialysis may be the removal of these specific endogenous benzodiazepines. So just to sort of summarize this slide, in addition to ammonia or hyperammonemia, there are other mediators of hepatic encephalopathy that you need to be concerned about as well. So now let's sort of go into ammonia metabolism. So ammonia is part of the story. It's not the whole story, but it does, talking about hyperammonemic states does give you a mechanistic rationale for why we see the difference in the nature of hepatic encephalopathy between ALF and ACLF. So let's just walk down this slide. So this is normal ammonia metabolism. So you have the gut, which is a generator of ammonia. It finds its way through the port, through the mesenteric venous drainage into the liver. In a normal liver, you have normal urea cycle function. You have ammonia fixation conversion to BUN, so urea in particular, and that then gets excreted through a functional renal system. So that is the normal pathway. So as these arrows indicate, there's minimal ammonia going into the systemic circulation or bypassing the liver because you have a functional urea cycle that then fixes the ammonia, if you will. So this is normal. And as you can see, there are no errors going into the brain, and that'll become important when we start talking about disease states in liver failure. Let's now switch gears and talk about hepatic dysfunction, what happens to the circuit when you have hepatic dysfunction. So you still have the ammonia load, if you will, coming out of the gut. And now you have a dysfunctional liver, which is not fixing the ammonia and converting it into urea. And so now these dotted lines become, indicate dysfunction. And then you have a higher systemic ammonia load going into the circulation. And one of the examples of extrahepatic tissues that can take over ammonia fixation is muscle. And so that's why I put this, sort of this barrier here, which protects the brain from seeing this ammonia load. So this is one example of an extrahepatic tissue that can upregulate ammonia fixing mechanisms in the setting of hepatic dysfunction. So now let's sort of take this construct and apply it to both ALF and ACLF. So in ACLF, imagine you got a nascerotic or a hepatitis C cirrhotic who slowly dwindled away the urea cycle over 20 years. As the fibrosis gets worse, converges to cirrhosis, decompensated cirrhosis. So you had the chronicity of time during this process. So because this urea cycle has been slowly denuded, you've had time for extrahepatic tissues like muscle to upregulate ammonia fixing mechanisms. So even though you have this ammonia bolus coming through, I've put this arrow in big because an ACLF muscle has time to upregulate over the years and couple it into glutamine. And the CNS circulation does not see much of this ammonia load because the muscle is now neuroprotective, if you will, from an ammonia standpoint. So this is the context of decompensated cirrhosis, ACLF. And so you can see the glutamine effect on the brain is much lower. Contrast this situation with acute liver failure where it's a 25-year-old, no chronic liver disease, decides to take a whole lot of time and all acutely fries his or her urea cycle. Suddenly, you haven't had the luxury of time to upregulate ammonia fixing mechanisms outside the brain, sorry, outside the liver. And so the muscle, for example, does not have time to fix ammonia. And so there is a CNS bolus, if you will, of hyperammonemia that hits the CNS circulation. And astrocytes, by the way, have glutamine synthetase. So they then start converting. And you can see this glutamine arrow is much bigger. And so this difference in ammonia fixing mechanisms is one potential explanation for why, as in ALF, as you start developing astrocyte swelling and you develop cytotoxic edema, that can then translate to ICP elevations and intracranial hypertension. So this mechanistic slide using the ammonia pathway and comparing ALF and ACLF is a useful construct to explain the clinical differences that we see between ALF and ACLF. All right, so now let's switch gears. And again, based on that differing pathophysiology, how does it affect your management? And let's start with acute liver failure. And again, remember, now we're dealing with somebody who's at risk for developing intracranial hypertension. So risk factors for intracranial hypertension, ALF, as we look at different studies, hyperammonemia is a big mediator. And there's good evidence that as your ammonia surges and sort of in the 100 to 150 range, that can then translate to increased risk of ICP elevations. And the other sort of take-home point here is that ammonia now becomes a target for therapy. If you decrease the ammonia in an ALF patient, that can translate to improved hepatic encephalopathy and minimizing the risk of herniation. Other sort of triggers, hyperacute liver failure, so there are different forms of ALF depending on how rapidly the disease process and acetaminophen is a good example of hyperacute liver failure. Going back to the theme that in addition to hyperammonemia, there are other mediators, such as sort of multi-organ failure, need for vasopressors, need for RRT, advanced hepatic encephalopathy and younger age. And the rationale for this is in a younger brain, you have less room to swell, if you will. There's not much cerebral atrophy. So when you compare a young patient to an older patient, there's a higher risk of developing intracranial hypertension for the same degree of astrocyte swelling, if you will. So these are identifiable risk factors for developing intracranial hypertension in ALF. And just to put this in one data slide, this is from 2007 in one study where you look at the y-axis as proportion free of intracranial hypertension. And this is day of admission of ALF. And this sort of illustrates the fact that arterial ammonia is a mediator. And the higher the ammonia level, and this is greater than 200, the lower the chances of being free of ICP elevations. So just one specific study that talks about the importance of hyperammonemia in ALF. So initial management strategies for advanced encephalopathy. So let's assume you got a tunnel overdose, ALF, grade three slash four encephalopathy. And this is, in my opinion, a immediate indication to think about elective endotracheal intubation for airway protection. And then once they get intubated, you're dealing with somebody with potential ICP surges, so it is important to maintain synchrony of the ventilator with low-dose sedation, so propofol, fentanyl, dexamethamidine are reasonable options to optimize patient ventilator synchrony. And then a couple of points here. Once the patient is intubated, the next step in the management algorithm is how do you gauge ICP elevations? And this practice has been an evolution. We used to put intracranial monitors in, intraparenchymals, subdurals, but over time, given the risk-benefit, and as you can see, it's challenging to have an INR of 11 and invite neurosurgery to drill a hole in the brain. So the general theme, the practice pattern has been to start thinking about non-invasive neuromonitoring strategies. With just a couple of, and we'll talk about a bit more, but one point to make is CT, head CT is insensitive to detect intracranial hypertension. So please don't reassure yourself that the absence of radiographic signs of cerebral edema on a CAT scan rules out pathologic intracranial hypertension. So there's just one caveat to that rule, so just be careful about that. And I'll mention some of the roles of non-invasive techniques that are being studied. Couple of sort of take-home messages here. So once you have advanced encephalopathy, you can assume they have elevated ICP, and these are some empiric strategies to target as far as management. So mean arterial pressure greater than 80. You're making the assumption that you have an ICP in the 20 or greater range, so the goal to induce therapeutic hypertension is to compensate for elevated ICP and maintain a cerebral perfusion pressure in the 60 range. Other sort of strategies, hypertonic saline. So inducing therapeutic hyponatremia with a goal target of 145 to 155, and there are a couple of studies that support this. The third strategy that some centers use is therapeutic hypothermia. And this, I put a question mark there because there's some conflicting data regarding the efficacy of hypothermia in the management of ICP elevations in ALF. But personally, our center does think about this strategy when we're having ICP elevations that are refractory to hypertonic saline in particular. So management of elevated ICP. Again, this is sort of a, depending on different centers, different strategies, but again, this is targeting decreasing ICP elevation. So some centers are using Manitoulin right off the bat, but we have moved quickly, very quickly to hypertonic saline. That's become a backbone of our therapy. Again, trying to target a serum sodium 145 to 155. I've mentioned hypothermia, which is external cooling works very sufficiently in these patients. And then finally, this is one take-home message I want to share with you. There is an evolving role of high-dose CRT. Going back to the point I'd made that hyperammonemia is a target for therapy, there is emerging a greater practice and this strategy is to use high-dose CRT, again, defined in some studies as high as 90 milligrams per kilogram of dialysate. That's almost six liters of dialysate, with a goal to decrease the ammonia level to 100 to 150 range. And this in our practice has worked really well. We have folks with INRs of 11, basically no hepatic function left, with ammonia presenting at 300, starting high-dose CRT and minimizing, keeping them in grade two encephalopathy as they hopefully await liver transplantation. So I just wanted to make sure that I shared this strategy with you because it's working really well for us and for many other centers as well. So please think about high-dose CRT with a goal to target hyperammonemia. Just to sort of move on from brain targeting, other management strategies in other organ systems, just a brief word. For cardiovascular, we spoke about norepinephrine or therapeutic hypertension, where norepinephrine is our first vasoactive agent of choice. Just a thought about IV access, if you have the option of not blocking both IJ sites to facilitate central CNS venous drainage. From a lung standpoint, again, not evidence-based yet, but I would consider a lower PEEP strategy. Remember that anything that increases the intrathoracic pressure will impede venous return. Minimizing AT tube suctioning, because every time you put the tracheal catheter in, you're potentially causing surges in ICP. And a very important maneuver, which you're only doing for VAP prophylaxis, is head of bed elevation. When you have that ALF patient, please make sure the simple maneuver of keeping the head of bed elevated is an important measure to prevent further ICP elevations. From a renal standpoint, another important take-home message. If you are thinking of starting RRT for any of the conventional indications, or for hyperammonemia, please use CRRT as the default choice of dialysis irrespective of the hemodynamic status. Normally we think about CRRT in hemodynamically compromised patients in the ICU, but this patient population has the increased risk of herniation if you subject them to HD, because of rapid volume shifts and electrolyte shifts. So CRRT should be the default motor choice irrespective of your hemodynamic status. And then just to go back to the hyperammonemia story, there is emerging observational data, as I mentioned before, suggesting the utility of CRRT in making hyperammonemia a specific indication for dialysis in these patients. So just to finish up the story on ALF, briefly about neuromonitoring techniques. So one thing that we use at our center is pupillometry. As a readout of sort of an example of non-invasive neuromonitoring, this needs to be validated. All these strategies need to be validated. And so there are definitely areas of future research. And then lastly, extracorporeal liver support. I'll briefly talk about MARS. Plasma exchange, as some of you may know, has been shown to improve survival in ALF. So that may have an evolving role. And then this is the Holy Grail, by artificial liver systems. Can you put hepatocytes in the extracorporeal circuit and salvage the liver? That's an area of active investigation. So far, no success yet. All right, in the next remaining five to seven minutes, let me switch gears and talk about hepatic encephalopathy in ACLF, in acute and chronic liver failure. So precipitance of hepatic encephalopathy, far and away, you have to think about a nuanced infection, but other triggers, GI bleeding, which can increase the ammonia load from the gut, hyponatremia, a classic trigger, and then preexisting tips is another classic trigger for HE in ACLF and cirrhotics. Another important point to be made is the limitation of ammonia levels in ruling in a diagnosis of HE in cirrhotics. You can have a totally normal ammonia level and have severe hepatic encephalopathy in cirrhotics. So on rounds, I don't even ask for an ammonia level. If you have a cirrhotic who's confused, they're getting lactulose, rifaximin, yeah. From a treatment standpoint in the ICU, elective intubation, if you need mechanical ventilation, there's almost auto-sedated, so minimal sedation strategies. And then contrast this sort of next couple of bullet points with the previous ALF story. There we were talking about osmotherapy and hypertonic psyllium. Here, it's almost a homeopathic cocktail. You're starting lactulose, you're starting flagyl, you're starting rifaximin, and you're reversing the precipitating factors, so antibiotics for spontaneous bacterial peritonitis, for example. Between these two strategies, they will eventually wake up in two to three days without any sequelae of intracranial hypertension. And then finally, something I'll share with you briefly is the efficacy of albumin dialysis. We've had very good success using albumin dialysis in reversing refractory hepatic encephalopathy in ACLF. And this is supported by RCT-level data. And I'll show you the brief slide regarding the albumin dialysis strategy. So this is the MARS, some of you may have heard about this system, this is an example of albumin dialysis. And just to sort of contrast this with CRRT, this is the MARS membrane. So as you contrast this with CRRT, a couple of big differences. Number one, you've changed the dialysate from an aqueous dialysate to an albumin solution. The second is you've changed the pore size of this membrane such that in addition to the circular low molecular weight substances, you have now the opportunity to extract albumin-bound toxins, such as endogenous benzodiazepines, which I've mentioned is one potential mediator for hepatic encephalopathy, across this membrane because you've created a diffusion gradient by switching the dialysate to an albumin solution. So that sort of in a nutshell is the mechanistic rationale for how this albumin dialysis system is able to extract specific albumin-bound toxins that will not come off with CRRT. So this is a interesting slide that delineates the differences in what substances are removed. So on the right side are the host of molecules that are removed with CRRT, and you can recognize BUN, you can recognize creatinine. Interestingly, ammonia falls in this category as well, if you crank up the dialysate enough. On the left side is a host of molecules that are albumin-bound that will not come off with CRRT that can suddenly start coming off without an albumin dialysis system that I just showed you because of those specific properties. And endogenous benzodiazepines is one specific class of molecules that's pertinent to this conversation regarding reversing hepatic encephalopathy in cirrhotics. And just to put this in a circuit, this is regular CRRT. You got one membrane, you got an aqueous dialysate. This is the Mars circuit. So just to briefly walk you through this, this is a similar dialysis catheter, but you have two membranes. The first membrane is the one that I showed you in expansion, which has an albumin dialysate going counter-current to the patient's blood. You have the transfer of both aqueous and albumin-bound toxins coming across this membrane. And then once this albumin effluent comes out, you sequentially dialyze it. You dialyze the dialysate. And it goes through a regular CRRT machine where you get rid of all your aqueous toxins. And then what's left here are those specific albumin-bound toxins that first find their way into a charcoal column and then an ion exchange column. And then this albumin dialysate is cleansed and sees the patient's blood again. So just to give you a snapshot of the circuit in this albumin dialysis system that is working very well for us as far as reversing refractory hepatic encephalopathy in cirrhotics. All right, so let me summarize the points. The etiology of hepatic encephalopathy in ALF and ACLF is multifactorial. And that's an important take-home message. The other point, as we mentioned, is HE in ALF is characterized by the risk of developing intracranial hypertension and cerebral edema. And therefore, you have to tailor your strategies to watch out for that complication. The chronicity of the disease process in cirrhosis in the context of ACLF permits the upregulation of extrahepatic tissues that can then fix ammonia and can be neuroprotective and therefore limit the risk of ICP elevations in cirrhotics. And finally, there is, I would stay tuned for evolving dialytic therapy. So HIDOCRT, albumin dialysis in particular. Thank you for your attention. Thank you. So I'll introduce the next speaker. Again, our friend Dean Karvelas, who is one of the wisest men, I think, in the field of liver critical care. I'm trying to find his exact title, if you will. But it's about bleeding and thrombosis, a hematologic conundrum. Please welcome Dr. Karvelas. Yeah. Thanks. Thanks a lot. Thanks. Thanks. Yeah, we all have our own microphone. So thanks, Ron. That was an excellent talk. So for the next 20 minutes, we're just going to switch focus a little bit on the challenges around coagulation in liver disease. So the objectives of this talk is I want to understand the alterations in hemostasis in both acute and acute on chronic liver failure. There are a few more similarities. Kind of review current tests, and in particular, the challenges with conventional coagulation tests in terms of looking at the risk of bleeding and thrombosis in both ALF and ACLF. And I also want to talk a little bit about some of the more kind of novel viscoelastic tests, which may give us further information, and understand the role, in fact, of anticoagulation in some patients that might be at risk of thromboembolic disease. So Ron kind of highlighted the difference between ALF and ACLF very nicely, that in one hand, we're talking about in ACLF patients with portal hypertension and a scarred liver, as opposed to acute liver failure, where you have a necrotic liver due to acute inflammation and the release of pro-inflammatory cytokines. So the first thing I want to mention that applies to both ALF and ACLF is there are a lot of misconceptions about hemostasis. And the goal of kind of the next kind of 18 minutes or so is to kind of dispel some of these myths, including the relation between INR and bleeding risk, and that all ALF and ACLF patients are at risk of bleeding, that patients are quote, unquote, auto-anticoagulated, which I will try to show you the opposite, and that blood product transfusion poses a minimal risk. We'll kind of go through all these things or dispel these myths in detail. So first, it kind of goes back to basics, where we're looking at the different stages of blood clotting. And you can see with this table that we've got kind of the hemostatic risks of bleeding on the left and the risks of thrombosis on the right. So we know that if we look at platelet aggregation to begin with, that cirrhotic patients and ALF patients can develop thrombocytopenia and have platelet dysfunction. However, we're also aware that the liver makes ADAMTS13, which is involved in the formation of von Willebrand's or the cleaving of von Willebrand's multipers. So patients with liver failure end up with these large von Willebrand's multiplers that can lead to more aggressive platelet aggregation. So in some ways, it's a very different regulated balance. And then when you look at the formation of clot or thrombin generation, in fact, we have the depleted levels of procoagulant factors such as 2, 5, 7, 9, and 10. But in contrast to this, patients with liver dysfunction and liver failure have elevated levels of factor VIII. And they also have depleted levels of protein C and protein S. And a lot of these things are not measured by something like the INR. And that's really the challenge, is that a lot of our conventional tests only tell us one side of this balance. And then finally, with clot breakdown or fibrinolysis, this is something that's very difficult to predict in a patient with liver dysfunction. And this has important implications, for example, with surgical procedures and particularly with liver transplantation. And this is really where the role of viscoelastic testing is actually quite important. And there are other things. So we know that it's a reset balance, but a delicate balance between bleeding and thrombosis in a patient with liver failure. And this can be tipped over one way, particularly to bleeding under certain physiological conditions. And this includes hemodynamic alterations due to portal hypertension, endothelial dysfunction. And in sepsis in particular, a lot of bacteria produce endogenous heparin-like substances. And finally, with renal failure or acute kidney injury can also complicate matters. So in terms of the data for what I'm telling you, a lot of this work has been done by Ton Lisman's group in Groningen in Holland. And this is a nice study that showed that while you have thrombocytopenia, which is very common, you get these. If you look at progression of liver dysfunction, this is doing it based on the child's Pew score. If you look at the larger amounts of von Willebrand's multimers, you can see that this also goes up in a stepwise faction. So in some ways, this actually compensates for the patient being thrombocytopenic. The other thing in terms of thrombin generation, this is a study from Armando Trapodi's group in Italy, which shows that really the most important thing, it's not the test that we do in isolation in the lab, but really the name of the game is thrombin generation. And he showed that in a group of cirrhotic patients with an elevated INR compared to controls, that thrombin generation was actually preserved. So probably the most challenging thing to predict in the absence of viscoelastic testing is fibrinolysis. And as you can see here, if you look at, this was another study from Ton Lisman's group that looked at child's Pew A, B, and C, or kind of progressing severity of cirrhosis and liver failure, that fibrinolysis really didn't correlate. And this is something that can be very difficult to predict in a livered patient without ancillary testing. So assessment of coagulation, I'm going to go through some of the conventional coagulation tests and their strengths and their flaws, and then also talk a little bit about the newer data around viscoelastic testing, including TAG and Rotem. So what this chart on the right really shows you is that this is looking at the prothrombin time, or INR, and liver bleeding time at the time of liver surgery. And what you can see is that there's really no correlation at all. And this is just to highlight that INR does not correlate with bleeding risk. And it's because it's insensitive for natural anticoagulant factors. And it's not really reflective of the true balance between pro and anticoagulant factors. And this often, we realize, is a challenge, because sometimes when we're discussing with our radiology colleagues getting percutaneous procedures ordered, often they'll be asking for this test. And even if it's mildly elevated, but the patient, in fact, is making thrombin appropriately, they're asking you to give blood product. And this is where a lot of these misconceptions come in, is that the INR was developed to monitor the administration of warfarin. It wasn't designed to be used in patients with liver disease. And it only assesses a small portion of overall hemostasis. And as mentioned before, there's no consistent relationship. So what test should you be using? Well, I think one of the most important tests that we order infrequently, but probably should order more often, is a fibrinogen. Because we know that cirrhotic patients have significant dysfibrinogenemia. And this has been clearly demonstrated in patients with a fibrinogen level of less than 1.5. So this is one take-home point, if you want to come up with a very simple cut-off. A fibrinogen level of 1.5 is usually associated with normal thrombin generation. And this can be given through fibrinogen substitutes, such as cryoprecipitate or a fibrinogen concentrate. The second thing is that what is a safe kind of platelet level? And there are two scenarios. There's the patient that's not bleeding. There's the patient that will be undergoing a procedure. And then there's the patient that's actively bleeding. And this is data that's come from a secondary data from this HALT-C, which was a large hepatitis C study. And one of the interesting things is a lot of these patients needed liver biopsy. And as you can see in this diagram on the left, that you really didn't see an increase in bleeding complications until you got to a platelet count of less than 50 to 60. So once again, this is a very practical cut-off of about a, generally, we know also based on in vitro data that thrombin generation is preserved with generally a platelet count of above 50 to 60. So taking it into the different context, which is acute liver failure, which Ram talked about in detail, what about bleeding complications in ALF? Because often, we get very concerned about the INR and ALF. But that's primarily from a prognostic point of view. We use the INR to determine whether the liver is improving or not. And in fact, if we give patients excess fresh frozen plasma, we can actually confound one of our most important prognostic markers. In this nice study that Todd Stravitz did looking at data from the US Acute Liver Failure Study Group, and this was from over 1,700 patients with a rare disease with ALF, what they demonstrated was actually that the INR level did not correlate or was not significantly associated with rates of bleeding complications. But what really did discriminate patients that had a bleeding event versus patients that didn't have a bleeding event was platelet count. And that's this diagram on the left. And this was statistically significant with a p-value of less than 0.001. So kind of some simple recommendations to take home, and this has been published in a couple of previous consensus guidelines. And this has also come out recently in our SCCM liver management guideline that was just published in Critical Care Medicine, is we recommend against the routine prophylactic use of fresh, frozen plasma that the INR does not adequately assess hemostasis. And a platelet count of 50 in the presence either of active bleeding or for procedures, generally a reasonable threshold to transfuse above. And also based on the Villanueva study, a hemoglobin of above seven is usually good. So I've mentioned indirectly a little bit about viscoelastic tests and thromboelastography. And really what this is is you're now graphically looking at clotting in different phases. This is a technique that's used very commonly in cardiac surgery and liver transplantation where really you wanna look at the patient's ability to make clot and to break down clot. And this is done in different phases. So one example here is looking at Teg with the R time, which is looking at the latency or the initiation of the clotting cascade. You have the kinetic time, which looks at the rate of initial clot formation. The alpha-lengel, which looks at fibrinogen formation and cross-linking. And then you also look at the maximal amplitude, which is the strength of clot. Now a lot of this stuff you can get in the first 10 or 15 minutes. And in some cases, at least you can tell if the patient has coagulation failure, which does happen in ACLF where you may need to transfuse that patient. The only thing I should caution you is that looking at fibrinolysis or clot breakdown generally takes more than 30 minutes. So if it's the decision in the operating theater of whether I'm gonna give an antifibrinolytic, then that can be a little bit more difficult and it does take more time. Just to mention as well is that also Teg and Rotem, the names of the parameters are slightly different, but generally they do reflect the same thing and these are proprietary. So one other study that Todd Stravitz did that showed very nicely is he looked at a cohort of acute liver failure patients and showed that if you went by their Teg parameters, that they were all essentially within the normal range. So one of the concerns we have, for example, is putting in an intracranial pressure monitor, which we don't do as often in acute liver failure. But what a lot of centers do now is if you do a Teg and the Teg shows that you have preserved clotting and coagulation, often you can get away with doing this without giving blood product, even if the INR is three and a half or four, if they generally have a platelet count above 50 in a normal looking Teg. So the other thing to mention is that, has there been any studies to show the benefit of using Teg and Rotem and there was a randomized study that was done from Lupe Garcia-Sal's group where they looked at 60 patients. So it's a pilot study where these were patients that needed some kind of percutaneous procedure, either it was a paracentesis or a liver biopsy or a chest drain and they were randomized either to getting Teg done and then based on the Teg, the decision was whether they needed blood products or not or the decision was made standard of care, which was without viscoelastic testing and generally this went based on platelet count, fibrinogen and unfortunately INR. And what you can see in this study is what they found is that the rate of bleeding complications was no different between the two groups, despite the fact that the Teg group received significantly less blood products. And we know that a lot of our blood banks and our coagulation medicine physicians are wanting us to be more judicious with the use of blood products. We know that there's also a risk of transfusion related complication and this shows likely that we're probably over transfusing ACLF and cirrhotic patients for procedures and this is something where clearly I think it's cost effective and we know that not all centers have access to Teg and Rotem in their conventional practice, but there probably certainly is a economic model for that. So when we look at what the bleeding risk is then with procedures, because there's a few different procedures that are done, so there are the non-invasive procedures such as liver biopsy and parainthrocentesis and I just wanna show another kind of study that was done by Sehgal and colleagues where not surprisingly they show that INR does not predict breeding complications with invasive procedures and it kind of builds onto what we were talking about before and they showed this systematic review of about seven different studies that looked at liver biopsy, transjugular biopsy and laparoscopy. And what's probably more important in terms of what's the bleeding risk of more in a surgical procedure in a patient with cirrhosis or with liver disease is it's the type of surgery. So in non-transplant surgery certainly if you're doing a valve or cardiac surgery there's a much higher bleeding risk compared to something like a hernia or a cholecystectomy. And if you look at parameters that predict bleeding complications during liver transplantation it actually has more to do with volume status with the central venous pressure and we know that's why our anesthesiologists like our patients dry when they're going for transplant, has to do with the ischemia time or one of the donor factors with the liver. And also if it's a difficult operation, if they've had a prior abdominal surgery or prior transplant then there's a lot of adhesions in the abdomen. So once again kind of the same thing again, recommendations that you'll see for minimally invasive and surgical procedures really the simple thing to remember is the platelet count of 50 and a fibrinogen level of 1.5 if you don't have viscoelastic testing readily available. Finally I just want to mention something that is also interesting is that patients with liver dysfunction can be hypercoagulable and this is one of the other reasons why you may be potentially putting a patient at risk by transfusing them. Rates of portal venous thrombosis, we know in cirrhotic patients, the incidence is up to about 25%, and part of this is a blood flow issue. But interestingly, if you look at rates of venous thromboembolism in hospitalized cirrhotic patients compared to controls, they're actually, the incidence is twofold higher than hospitalized patients without liver disease. And one of the places where we see this actually is in the intensive care unit, is if we look at the circuit life of a CRRT or a dialysis machine, that we see that CRRT circuit life is actually significantly lower in patients with liver disease. And in some cases, we either have to anticoagulate them with a small amount of heparin, or in some cases, for short course at least, you can use citrate. And also, the rate of ischemic stroke after you adjust for a variety of other factors in cirrhotic patients, in one study, was found to be 30% higher. So this was another kind of population-based study. So this is a pooled meta-analysis of 15 studies where they looked at almost 700,000 cirrhotic patients compared to 1.5 million control hospitalized patients. And you can see that the odds of a thromboembolic event was increased in cirrhotic patients with an odds ratio of 1.7. So this is a real phenomenon, and that ACLF cirrhotic patients can, in fact, have a hypercoagulable state. And to take this a step further, this group in Spain did a very elegant study, and this was obviously a non-hospitalized patient, so it's borrowed data. What they did is they randomized 70 cirrhotic patients to receive prophylactic low molecular weight heparin versus placebo. And at 48 weeks later, they not only noticed that the rate of portal vein thrombosis was significantly lower, but they actually had a decreased rate of hepatic decompensation and an overall improved survival. So we know that there is a subpopulation of ACLF and cirrhotic patients that are probably hypercoagulable, but probably where the work that needs to be done in the future is identifying which group this is. So the other thing to mention is the use of anticoagulation in these patients generally is associated with recanalization and the prevention of thrombosis progression. And then the one question that comes up is what about the concern about variceal bleeding? It was interesting that this study that was published in 2015 showed that if you compared patients that came into the hospital with a variceal bleed for endoscopy, and they compared 52 patients that were anticoagulated for a PVT or otherwise with 104 patients that were not, that they show that there was no significant difference in the rate of active bleeding or the failure to control bleeding or various other kind of hematological parameters. And more importantly, there was no difference in mortality or rescue therapy. And most likely at this point, what's driving complications or mortality in particularly a critically ill patient with a variceal bleed is their portal pressure and also the severity of their liver failure. So this is one of those things where we know that there is a group that we can probably safely anticoagulate. But I do admit that at this point, really in terms of personalized medicine and what is the appropriate population, this is the more difficult thing that we're dealing with. So probably what agent would you use? It would be low molecular weight heparin is probably the safest. And probably the population that we have the most data on is the patients that have a partial or occlusive portal vein thrombosis because we know if we can either initiate recanalization or prevent further progression of thrombosis, we can ensure that the patient potentially may be an anatomical candidate for liver transplantation. So in conclusion, patients with ALF and ACLF have derangements in both pro-hemostatic and anti-hemostatic factors creating a more unstable equilibrium. Conventional coagulation tests like INR should not be used as a predictor of bleeding or a guide to transfusion. Viscoelastic testing may play a larger role in the future in defining bleeding and thrombosis risk and the rational use of blood products. Thrombogeneration is generally preserved and a couple of simple targets to measure are a platelet count of 50 and a fibrinogen level of 1.5. And finally, patients with ALF and ACLF are not anticoagulated and should warrant consideration for anticoagulants and prophylaxis to prevent thrombosis. Thank you very much. So we'll move on to our next speaker, which is Bill Peppard, who's coming from the Medical College of Wisconsin. And he'll be talking about medical challenges with dosing and unusual infections in liver patients. Sorry, man. I'm like fussing with him. Thank you, Dean, and good morning. So today we're gonna talk a little bit about drug metabolism and liver failure. I don't have any relevant conflicts of interest. Now, it's important to note that there's not a lot of new literature in this area. So really what... Hello, can you hear me? Is that better? Okay. So what I'm gonna try to provide you some basic considerations about drug dosing, and then provide some practical application ideas from my real-world practice. We're gonna briefly describe the pharmacokinetics of the critically ill patient. We're gonna help you identify strategies that you can use to optimize drug dosing in this patient population. And more specifically, we're gonna try to select therapies that are appropriate for the critically ill patient, focusing on analgesias, sedatives, and antimicrobial therapies. So in critical illness, liver failure is a common phenomenon. Up to 54%, depending upon how it's categorized, as is acute kidney injury. Within the context of cirrhosis itself, both acute kidney injury and chronic kidney disease are very common. The take-home message here is that just because you have liver failure doesn't mean that the kidneys are functioning. You often have to consider those two organ systems together as they work together to metabolize and ultimately clear drug drugs. So as a pharmacist, I think a lot about pharmacokinetics. Pharmacokinetics, basically how the drug behaves in the body. And when we think about that, we think about absorption, distribution, metabolism, and elimination. And this is the core principle of most drug therapy. If we're lucky, we have pharmacotherapy outcome data, which focuses on both safety and efficacy. And ideally, we'd be able to consider both of those therapies together when figuring out how to dose drugs. So EDME in critical illness. We've got a lot of different variables to consider in this patient population. Often patients are hypoperfusing. They are volume overloaded. Their GI traction not functioning can significantly affect the absorption of drug therapy. You've got decreased perfusion to tissues. And then you've got the metabolism itself and the clearance to consider. And a lot of this is further amplified by acute liver failure within the setting of the critically ill patient. So we have to try to take all these variables in aggregate and figure out how a particular drug is going to behave. In addition to that, we have a lot of other organ systems to consider, and we have both acute and chronic diseases to consider. So what is the net effect of drug therapy? It's not very clear in many instances. So you really have to take an individualized approach to patient care. So metabolism is predominantly done via the liver, though for completeness sake, we also need to consider the GI tract, kidneys, lung, and brain. And we refer to a term called hepatic clearance, and that's the amount of blood that's cleared by the liver per unit time. Now, we do have an equation that helps us to explain this phenomenon. And hepatic clearance is the hepatic blood flow multiplied by the hepatic extraction ratio. But really the three major variables that we consider when we're thinking about drug therapy, hepatic blood flow, also protein binding, and the intrinsic ability of the liver to metabolize drug. And we have to figure out, based on the physicochemical properties of a given drug, how we weigh these variables in dosing considerations. So a little bit about the hepatic extraction ratio ranges from zero to one. Zero representing that the drug really is not significantly metabolized by the drug, whereas one represents the drug is significantly almost completely metabolized by drug therapy. And when we break it down by low, intermediate, and high, that helps us understand which variable we really need to be considering. So in low hepatic extraction ratio drugs, clearance is heavily dependent upon the enzymatic activity of the liver itself. High extraction ratio drugs, we're predominantly considering hepatic blood flow. And the intermediate is kind of a hybrid of the two, taking into consideration the intrinsic clearance of the liver itself. So I'm a visual learner, so this helps me understand and teach this particular concept. On the left, you've got a high extraction ratio drug. On the right, you have a low extraction ratio drug. So in a high extraction ratio drug, what you're gonna see in a liver failure patient compared to a regular patient is that your serum peak concentration is gonna be a lot higher. The bioavailability is gonna increase significantly, and your elimination is gonna be diminished. So in a drug like this, you're thinking about an empiric dose reduction, and you're also thinking about giving the drug less frequently. So that's in contrast to a low extraction ratio drug where your serum peak concentrations really aren't affected nearly as much, nor is your bioavailability, but your clearance is. So in this case, you might consider a normal dose, but you're going to be giving it less frequently. And there are some really nice review articles out there that try to explain this phenomenon and provide some nice summary tables. Gallagher and Beesbore put this review article together, really trying to focus in specifically on the drugs that are used in the ICU setting. So they've eliminated a lot of non-ICU static, and then they've gone far enough to recommend what sort of dosing strategies to employ based on the degree of protein binding that you have and the extraction ratio. So it's not necessary to commit all of these drug dosings to memory. If you can understand the principles behind it, you can come up with the dosing yourself, or if you wanna use a dosing table like this, that would be fine as well. So scoring tools are very common in liver failure, right? We try to objectify this disease state. When we talk about clearance in renal dysfunction, we've got the luxury of considering creatinine clearance. Things are not quite as clear cut in liver disease. So the main tools that we have at our disposal are CTP, which is broken into mild, moderate, and severe, or A, B, and C, respectively. And this gives us an idea of the degree of liver dysfunction. You've also got the MELD score, and this is really what we use in our transplant patient population at my institution. And this number can fluctuate from day to day. And while I'm not regularly calculating these myself, I am referring to them in the progress notes that my providers share with me. And this gives me a little bit of confidence in that I do know, based on limited data out there, that the MELD and the CTP do actually correlate with one another. More importantly, when you look at the study that evaluated 30 patients with midazolam, a drug that is heavily dependent on hepatic metabolism for its ultimate clearance from the body, compared to a control group of patients that have normal liver function, we can see that both the CTP and the MELD score correlate with clearance on unbrown drug. Okay, so as a pharmacist, this gives me a little bit of confidence to say that while there's not a perfect scoring tool that gauges drug clearance, this gets me in the right ballpark. It gets me thinking about what I need to be thinking about. So why is this all important? Why am I boring you with kinetic data? Well, there's a very good reason for that. Because you've got animal data when a drug comes to market, and then you've got healthy volunteer data, and then you've got the patients with the disease that we study the drug in. And usually that's the end of it. If we're really lucky, we've got outcome data looking at the critically ill patient population. But by and large, acute liver failure is an exclusion criteria for a lot of these trials, and that patient population is underrepresented. It's not uncommon to have patients with chronic liver disease in some of these trials, but that's a little bit of a different beast, as Ram had explained. There are significant differences in these patient populations. So in large part, we're extrapolating acute and acute and chronic liver failure from the data that we have in chronic liver failure patients. So how does this translate into clinical practice? Okay, well, we're going to focus on analgesics and sedatives to start with, and we're going to go over some basic principles. So when you look at this table, you can see that the most common analgesics that we use in the ICU often have hepatic components to their metabolism. They have active metabolites. Some of them have high extraction ratios and varying degrees of protein binding. So the basic principles of analgesia. Use an objective score, set goals, and use your tools, your drugs, to help get you there. Use multimodal therapy when you can. What we found is that this generally tends to have an opioid sparing effect, and in some instances, it actually improves pain control. This approach is endorsed by many different treatment guidelines that are out there. So when we talk about multimodal therapy, in my mind, the forefront is acetaminophen. We have to consider acetaminophen carefully in the setting of patients with liver disease because we do know that acetaminophen can be hepatotoxic. When you look at all the cases of acute liver failure, more than half of the time, it's associated with drug therapy. And of the drugs that cause it, about half the time, it is acetaminophen. So the general recommendation is that you can safely use two to three grams of acetaminophen per day, and it's effective for pain control and generally considered to be safe. There are some data suggesting that you can actually go up to four grams per day in the acute setting for short-term therapy, but you don't want to commit patients to long-term therapy with this. In general, the metabolic pathway is preserved even in significant liver disease. Non-steroidal anti-inflammatories, the other drug class that we think about when we want opioid sparing therapy. Generally, this is a drug class that you want to avoid in liver failure. It increases the risk of bleeding. It is associated with variceal bleeds. It actually can worsen renal function and routinely does. This is a basic principle that we apply when we try to mitigate the risk for hepatorenal syndrome. And it has a decreased diuretic effect. So we really try to avoid those. Now, this is a summary of analgesics in liver failure. It's important to recognize that this isn't from a treatment guideline. Okay, you're not gonna find this in a reference. This is kind of Bill's perspective on what he sees in practice when he synthesizes the data and how we implement care. So it's also important to recognize that in the early stages of our transplant program at my hospital, liver failure patients were in the medical ICU. And then post-transplant, they came to the surgical ICU. In recent years, we've changed our practice where the liver failure patients that are listed for transplantation are managed by the transplant service and they are admitted to the surgical ICU versus the medical ICU. So we see a lot of patients that are slightly different from the historical patients we saw. We saw a lot of patients that are acute on chronic liver failure that are listed for transplantation. Whereas previously, it was largely patients that had acute liver failure as a product of their acute illness or they were surgical oncology patients. So that's changed our approach to managing these patients a little bit. And because we see these patients in repetition, we've gained comfort with certain drugs. This isn't to suggest that you shouldn't use the drugs that I recommend or that not using the drug I recommend is wrong. Rather, this is what we've gotten comfortable with. It's important to recognize that essentially all these medications have some sort of hepatic metabolism. And by and large, what we've done is we've found therapies that we're comfortable with prescribing, we implement an empiric dose reduction, we use objective tools, and we titrate to effect. Now, Dwyer and colleagues put together a pretty nice algorithm. And I think this is fairly representative of what you can implement in your practice, right? So you determine the type of pain that you're dealing with. Is it acute or is it chronic? And your first line of therapy essentially is acetaminophen at a reduced dose. And they've gone far enough to say that the dosing could be based on body weight. You further differentiate the type of pain and oxycodone or tramadol is generally considered their first line of therapy. I would draw your attention to the dosing. In our practice, a normal oxycodone dose for acute pain in the perioperative setting or in the surgical patient population is five to 10 milligrams every three to four hours as needed for pain. When you look at the dosing strategy here, you're looking at a dose reduction of five milligrams. I would actually suggest two and a half milligrams is actually what we start with in our ICU. So two and a half to five milligrams every six hours. The frequency is considerably less frequent. And then if you go to IV, you're using fentanyl, 25 mics every six hours as needed as a first line therapy. I would also draw your attention to the normal dosing that we would use in our ICU for normal patients, 50 to 100 mics every one hour as needed. So not only are we using an empiric dose reduction, we're using it much, much less frequently. Going back to the basic PK principles that I described earlier in the presentation. Now we do recognize that there will be instances where you need continuous infusion and we tend to use fentanyl. You need to keep in mind that fentanyl, even under normal circumstances in patients without liver failure, you're going to see accumulation. It's a very lipophilic drug. And that accumulation will also depend upon the patient's body mass with underweight patients accruing or accumulating the most amount of drug. So even not in the setting of liver failure, you have to keep that in consideration. When you look at graph B, patients with liver failure or heart failure, you see lots of accumulation, right? And that's only after the first three hours, first three days, okay? So in my mind, you don't want to set a fentanyl infusion at 100 mics an hour like was done in this study and then turn around and walk away. You have to actively be evaluating the patient and probably do an empiric dose reduction on a daily basis to counteract the effects of the accumulation. Moving on to sedatives, again, you see the same things that we saw with our analgesics. Lots of hepatic metabolism, protein binding, high extraction ratios in some instances. And again, we're going back to using the most effective drug at the lowest dose possible to achieve your objectively determined goals using an objective scale. So we use lorazepam intermittently for patients in our ICU. Again, we use an empiric dose reduction. We avoid midazolam. We see lots of accumulation with that drug in hemodynamic changes. If you need a continuous infusion, we use propofol. There is a component of accumulation with propofol, but because the short half-life at baseline, it's still a readily titratable medication. We actually avoid dexamethamidine completely. The drug, in my opinion, is not readily titratable even in healthy individuals. The half-life is too long. I know that it is routinely titrated, but I don't think it's an optimal approach to therapy. And in liver failure, you should expect prolonged duration of therapy, prolonged half-life. If you look into Lexicomp, you're gonna see that it says, no specific dose adjustments provided by the manufacturer. However, clearance is reduced. And yeah, when you look at the early data, the phase one data, you can look at the different degrees of liver dysfunction and see how much lower the clearance is relative to a healthy individual, how much longer the half-life is and how the protein binding is affected. If you use it, you need to use very, very low doses. You need to use very infrequent dose adjustments and you need to be very, very proactive about monitoring for side effects. But my recommendation would be to avoid it altogether. Infections are very common in liver failure, right? And we don't have time to get into all the details that I would love to get into with this topic, but the most common infections you'll see are SBPV, bacterial translocation, UTI, pneumonia, and bacteremia, and cellulitis. In the ICU, the frequency of pneumonia is actually quite a bit higher. As far as pathogens go, you see a little bit of everything and your pathogens that are associated with infections shift a little bit from whether or not the infection is community-required versus nosocomial. So what does this all mean? Well, Jenna put together this nice algorithm that was published in AJHP a number of years ago that really does a nice job of, in my opinion, detailing the drug therapy that's out there. In this article, you're going to find dosing recommendations for drugs that don't have dosing recommendations in their package insert or in Lexicomp. And what they've done is they've synthesized all the pharmacokinetic data that are out there. There's a nice summary table. So if you understand the basic principles, you can look at the summary table and come up with dosing recommendations. But at the top of the algorithm, are you committing this patient to a short or long duration of therapy? If you're committing the patient to a short duration of therapy, you probably don't need to do a lot in terms of dose adjustment. But if we're committing the patient to long-term therapy, there's some other considerations we need to have in mind. So is there an alternative agent that is safer and doesn't have hepatic? If so, go without therapy, the path of least resistance. Are there data available for dosing guidelines? If so, great. But by and large, you're going to find that they're not. There's just not a lot of dosing recommendations of hepatic failure. The FDA reported that in a 10-year span from the late 90s to early 2000s, with 250 new drugs being submitted for approval, very, very small number of those actually came up with dosing recommendations. Now we've gotten better since the early 2000s, but by and large, there aren't good dosing recommendations out there in the package insert. And then is there a lot of toxicity associated? Do the drugs have narrow therapeutic windows? Is there toxicity associated with them that's a concentration or dose-dependent phenomenon? If so, you really need to think about therapeutic drug monitoring, and you need to think about dose adjustments. Up to this point, we've talked largely about dose reductions in drugs that are hepatically metabolized, right? The phenomenon in antibiotics is that we actually have a lot of water-soluble antibiotics, beta-lactams, for example. And these patients, by and large, are going to have increased volume and distribution from volume overload or SIT. So in this instance, we're actually talking about maybe doing an empiric dose for a loading dose and actually dosing aggressively, especially in the setting of renal replacement therapy, where there may still be some underlying residual renal function in addition to the renal replacement therapy. Considerations for drug monitoring. If the infection is severe, it makes sense. If the patient's not responding to conventional therapy, if the patient's high-risk for development of drug toxicity, neurotherapeutic window. And lastly, when doing so is not time- or cost-prohibitive. So for example, if it's a send-out drug that's gonna be expensive, a send-out level that's gonna be expensive and it's not gonna be back for four days for an eight-day regimen, probably doesn't make a lot of sense. You're not gonna be able to take action on that. That's meaningful. If it's in-house lab that you can get and it's a neurotherapeutic window, it probably makes a lot of sense to get that level. So here's an example of one of the tables from that study for future reference. I don't expect anybody to actually read this from where you're sitting, but know that this is a phenomenal reference. If you wanna take a second and jot down that citation if you haven't already, this is a really good resource. And then finally, as we wrap up my time, we're gonna talk about devices. Well, we're not really gonna talk about them. I'm just gonna give you a 10,000-foot overview. So 15 years ago, we relied heavily on dosing tables for renal replacement therapy. And they were okay for the time. They weren't optimal. We've learned a lot since then. But the PK data in that patient population is fairly well-established. And there's even outcome data in that patient population. We're confident in our drug dosing. ECMO, that's kind of the new thing right now, right? The PK data are emerging. There's a lot of really stellar pharmacists out there that are doing some pretty cool projects that I'm excited to read their work on. But there's not outcome data. It's all PK data. So we're gaining confidence in how to dose in that patient population. And then finally, devices, MARS, ELAD, as Ram referred to, right, that's the future. That's where we are, and that's where we're headed right now. That gets a little bit more complicated. You've got renal replacement therapy. You've got MARS and ELAD. You've got protein-binding issues to consider. You've got live hepatocytes. It's complicated. There is a paucity of data in this population, even the kinetic data. There's just not much out there. Outcome data, as far as I know, is virtually nonexistent. I hesitated to put none just because I didn't find it doesn't mean it's out there. But right now, we're making educated guesses on how to dose these patients. So in summation, patient's liver disease does demonstrate alterations in PK profiles of drugs. There's not a reliable method to figure out how to dose these patients. You have to individualize your care. And you want to select therapy that's minimally affected by drug metabolism. Short-acting agents make a lot of sense. Use the package insert for dosing if you have it. But routinely, that's not going to be the case. Establish goals. Use objective measures to get there. Utilize drug monitoring. And then look at the big picture. Weigh the benefit of toxicity versus treatment failure. And you know what? Use your pharmacist. We really enjoy looking into this stuff. It's a clinical challenge for us. And I know it can be time-consuming. But we're happy to help. With that, appreciate your time today. Thank you. What do you do when electrolytes get out of whack? Go ahead. Good morning, everybody. Thanks for the nice introduction, Ram. We'll be taking a look at electrolyte disturbances in our end-stage liver disease patients today. But let me start with declaring that I have no conflicts of interest to declare. And then let's take a quick look at what I would like to present to you today. We'll take a very brief look at tapeterinous syndrome, how we got there, as it's very closely linked to electrolyte disorders, on which we will spend a large amount of time throughout the talk. In particular, we'll be looking at hyponatremia. We will be looking at potassium disorders. And I'll be finishing off my talk with a quick look at acid-base disorders in these patients. Well, however, first of all, let's take a look how we are getting to the point that we need to talk about electrolyte disturbances in our patients with end-stage liver disease. Usually, as long as everything is compensated, things at least look fine on the outside. It gets a little more complicated and challenging when we are starting to look at patients with decompensated cirrhosis. When we are dealing with decompensated cirrhosis, the next or one of the main thing that triggers the rest of the cascade is the severe vasodilation in the splenic arterial and arterial system, which in turn leads to a decrease in the effective arterial blood volume. This decrease in the effective blood volume then triggers the activation of two main hormonal systems, meaning the renin, angiotensin, aldosterone system, and also the non-osmotic release of arginine vasopressin. And this non-osmotic release of arginine vasopressin, I would like you to keep in mind, as it's important for a lot of electrolyte disorders we'll be talking about over the next few minutes. The activation of these two systems, in particular, leads to the increased renal reabsorption of sodium and water, which in turn plays an important role in the formation of ascites. They are also important when it comes to worsening of renal function, because they promote intrarenal vasoconstriction, which usually leads to a reduction in the glomeral filtration rate, and then the patient is faced with what we call hepatorenal syndrome. In general, when we are looking at electrolytes and end-stage liver disease, we can note that they are fairly common in our patients. They can either be the result of the underlying disease process, or they can be the side effects of the treatment of what we are doing to our patient. As lined out before, in a compensated state with end-stage liver disease, everything on the outside looks fine, usually meaning electrolytes and acid-based status are not too much out of whack. However, this is a very delicate balance that is easily disrupted by just simply the progression of the disease, any kind of new infection, any dietary issues, and also, last but not least, by the things that we are doing to our patients. Now, becoming a little more specific, going into a little more detail, let's take a look at sodium disorders, in particular, hyponatremia, which is fairly common in our patients with end-stage liver disease. Nearly 20% of our patients present with sodium levels of less than 130, which is kind of the official definition for hyponatremia in patients with end-stage liver disease. However, the majority of patients has, the majority of patients with cirrhosis, I'm sorry, has sodium levels that are less than 135. Amongst all the forms of hyponatremia, the so-called hypervolemic hyponatremia is the most common form. This figure here kind of shows you the timeline of the events that I was just lining out before. As cirrhosis, or as end-stage liver disease progresses, you have a decrease in the effect of circulating blood volume with increasing activation of those humeral system. I was just referring to the renin-angiotensin-allosterone system and the non-osmotic release of arginine vasopressin. In the overall context of end-stage liver disease, hyponatremia is a very strong marker of the prognosis, especially when the patient otherwise has a low MELD score. However, hyponatremia is more of a marker of severity and complication, but it's not so much actually a mediator of the overall outcome. What do we now do when a patient with end-stage liver disease is presenting to us with hyponatremia? Well, although most cases are due to the excessive secretion of arginine vasopressin, which, as I've pointed out, leads to an increase in the renal sodium and water reabsorption, and then which, in turn, creates a positive free water balance that leads to a subsequent dilution of serum sodium, we still have to consider etiologies that are independent of the arginine vasopressin. And we also always have to carefully assess the volume status of our patients, as this will determine the further treatment. First, let's take a look at what are actually arginine vasopressin or ADH-independent etiologies in our liver patients. Well, that situation arises when our patients have a low consumption of solutes in the setting of a high or normal water intake, leading to hyponatremia. For our serotics, the classical picture in the setting of alcohol cirrhosis is the syndrome of biopotomania. Here, we have our patients with alcoholic cirrhosis who have a high intake of hypotonic alcoholic beverages, but lack sufficient nourishment, leading to the problem of the low consumption of solutes and intake of hypotonic solutions. The easiest way to find out about this and then treat appropriately is by checking urine osmolality, which is considered a direct marker of arginine vasopressin secretion. If you find a urine osmolality that is less than 100, you can be pretty sure that you have an AVP-independent mechanism, meaning you have a situation that has a low consumption of solutes with a high water intake. Here, the treatment is pretty straightforward. It consists of reinitiating the intake or consumption of solutes and decreasing or even ceasing the intake of hypotonic fluids. However, if you have an AVP-dependent hypotonic hyponatremia, your next step has to be the assessment of the volume status. As I already mentioned, the distinction between hypovolemic and hypervolemic hyponatremia is crucial because of therapeutic implications that come with it. Also, one thing to always keep in mind compared to a lot of other patients presenting with hyponatremia, the chronicity of hyponatremia in our patients with end-stage liver disease usually leads to the fact that the classic neurological symptoms like confusion, altered mental status, or even seizures can be absent a lot of times or do not occur unless very, very low sodium levels are reached. So when and how should we start treating patients with hyponatremia? In general, we can consider treating our patient when the serum sodium level goes to less than 130. And if we have ruled out AVP-independent hyponatremia in our patient with cirrhosis, the next step is going to be assessing whether our patients are hyper or hypovolemic. Our hypervolemic patients with hyponatremia usually present with ascites and edema. This is a lot of times triggered, or this requires the restriction of fluid intake, the discontinuation of diuretics, and if also present, the correction of hypokalemia, which we will talk about in a minute. If we get a response from that treatment, we should aim for serum sodium levels greater than 125, ideally trying to maintain the level somewhat above 130. If we do not get a response to this approach, some studies recommend the use of albumin infusions, although clear outcome data supporting that approach are lacking. But based on the data that's available, it is at least worth a try. If those efforts are also not successful, you still have no response. Of course, it requires a chat with your transplant hepatologist about further overall management and prioritization for liver transplant surgery. If you, on the other hand, have a patient that is hypovolemic, which usually can be seen in patients that are also hypotensive, orthostatic, you also need to correct the underlying cause. And here, once again, diuretics can be the problem. Patients will need fluid resuscitation, preferably with balanced electrolyte solution. And a lot of times, that is enough to correct the hyponatremia and to bring the sodium levels up to a point which is a lot safer and easily to maintain. If that is not working either, you can try some additional volume resuscitation with albumin infusion and see whether or not you are getting a response. A few things, irrespective of the underlying etiology and the classification, you want to keep in mind when you treat hyponatremia in patients with end-stage liver disease, the use of hypertonic saline should really be reserved for patients that are severely symptomatic and that are acutely symptomatic. And here, I mean, for example, the onset of seizures. Bole of 3 cc's per kick of hypertonic saline, of 3% hypertonic saline, is one of the approaches you can use. However, keep in mind that the long-term use of hypertonic saline actually worsens ascites and edema and is therefore not recommended. One of the other critical things to keep in mind, because it can go wrong at times, and it can go wrong pretty badly, is the risk of osmotic demyelinization if you raise your serum sodium levels too quickly in non-acute situation. You should really limit the raise of serum sodium to less than 8 millimolars over the course of 24 hours. I know a lot of times we talk about, and we haven't seen it, but based on my own experience over the past year, I can tell you, unfortunately, it happens if you don't pay attention to that. And the outcome is very devastating and frustrating to everybody involved. Based on available studies and available data today, there is really no benefit or no role in the use of VapTens. On one side, VapTens have been shown to raise your sodium levels. However, it comes at the cost of the increased risk of liver injury, something we don't really want in our patients with end-states liver disease. And also, keep in mind that when you do a simultaneous correction of hypokalemia, that because of the trans-cellular exchange of sodium and potassium in the kidney, you can actually get a further rise in the serum sodium concentration beyond what you were expecting just simply based on your previous measures. And talking about potassium, let's take a look with potassium in end-stage liver disease. And in end-stage liver disease patients, hypokalemia is actually fairly frequent. Hyperkalemia, on the other hand, is somewhat that is not that common. It usually only occurs at the later stages and often triggered or induced by potassium-sparing diuretics. Also, not only in patients with end-stage liver disease, but in general, to keep in mind that serum potassium levels are poor indicators of the total body potassium. The reason I'm mentioning this is because when you're overzealous with your correction of your potassium, and you don't know what the intracellular stores actually are, and you can get some bad surprises with rebound phenomena or sudden trans-cellular shifts that are unleashing potassium from intracellular stores. Why do patients with end-stage liver disease get hypokalemia? Well, if you take a look at the characteristics of cirrhosis, like vomiting, malnourishment, hypomagnesemia, diarrhea, and use of diuretics, those are all perfect stimulators or perfect events that can lead to renal and extrarenal wasting of potassium. So just looking at these things, it's surprising that our patients with end-stage liver disease actually develop hypokalemia. Also, one of the things to keep in mind is that hypokalemia, and that's why we want to treat it, hypokalemia can precipitate hepatic encephalopathy. And there are some interesting data that actually show that, as we found in our first talk, that ammonia is not everything in hepatic encephalopathy. But it's still important that the rise in ammonia level actually correlates with a decrease in serum potassium levels. So how do we deal with hypokalemia in patients with end-stage liver disease? If you are having a hard time correcting hypokalemia in these patients, you definitely want to check your magnesium levels, especially in patients with alcohol cirrhosis, as refractory hypokalemia can be linked to coinciding hypomagnesemia. You want to probably think about correcting hypokalemia even if the patient is asymptomatic, because a lot of our end-stage liver disease patients also show some degree of cardiomyopathy and are therefore prone to cardiovascular comorbidities, in particular, potentially life-threatening arrhythmias. However, when you start correcting hyperkalemia, you want to do it slowly. You don't want to overshoot. You don't want to get rebound hyperkalemia. And you also want to consider that, actually, in patients with end-stage liver disease, the buffering capacity for potassium loads is substantially diminished. The last part of my talk, I would like to take a quick look at acid-base status in end-stage liver disease. And as you can just see from these two simple cartoons here, the liver, both under healthy and under disease conditions, is greatly involved in maintaining acid-base status. The liver participates in ketogenesis. It is very important for albumin synthesis. And it's heavily involved in lactate metabolism, both under healthy and disease situation. In patients with end-stage liver disease, we will frequently notice hyperventilation. And when this hyperventilation is due to centrally-mediated causes, like ammonia-induced encephalopathy or the central activation of progesterone receptors, those patients will develop respiratory alkalosis based on the liver disease. And just of those few things I've mentioned and all the other things we know about our end-stage liver disease patients, you can imagine that there's a plethora of acid-base disturbances that could, at least theoretically, go wrong in our patient with end-stage liver disease. And I'm not going to go through all of these for the sake of time. But in general, metabolic or acid-base derangement, let them be acidosis or alkalosis, can be due to the disease process itself. Or they can be due to the treatment to what we do our patients. And in the next slide, I would actually like to show you how does the acid-base status in our ICU patients actually look like when they have end-stage liver disease. Well, here in this panel here, you see those non-serotic ICU patients who show a fairly balanced acid-base status. And as you can see, over time, as the cirrhosis, the end-stage liver disease, gets worse and we have more and more acute and chronic liver failure, you see that this seemingly well-balanced state changes to a net metabolic acidosis, which seems to be due to increased lactic acidosis and the presence of unmeasured ions. And this net metabolic acidosis cannot be compensated for by a respiratory alkalosis as we sometimes find in our patients with end-stage liver disease. And also, the degree of that net metabolic acidosis also is strongly associated with poor outcome in our cirrhotic patients. Let me summarize what I've shown you so far, what I've shown you today. I've pointed out that hypervolemic hyponatremia is the most common form of hyponatremia in patients with end-stage liver disease. When you correct hyponatremia, you have to do it slow due to the risk of osmotic demyelination syndrome. If you don't have any acute symptoms, you should limit your sodium. And with acute symptoms, I mean seizures or severely altered mental status, you have to limit your serum correction to an increase of less than 8 millimolar per 24 hours, at least in the beginning. Patients with end-stage liver disease also frequently show hypokalemia. Hypokalemia can be a precipitator of hepatic encephalopathy. And also, for cardiac reasons, it's certainly worthwhile considering to be treated. Similar to hyponatremia, hypokalemia also warrants a slow correction to avoid unwarranted side effects like rebound hypokalemia, for example. And last but not least, patients with end-stage liver disease, especially as the disease progresses, develop a net metabolic acidosis, which cannot be compensated for by respiratory alkalosis in these patients and is also strongly associated with poor outcome in these patients. And with this being said, I thank you very much for attention. And as I've been told, we'll be saving questions for later on. Thank you very much. Thank you. So thanks, Kai. And just to echo that, we'll have about 15 minutes for questions of all the speakers just at the end of the last talk. And our last talk is going to be presented by Dr. Jody Olson, who is from the University of Kansas. He's a transplant hepatologist and intensivist and associate professor. And he'll be speaking to us about the microbiome and liver failure. Thank you, Dean. And thank you to the organizers for inviting me to speak today. The topic of my talk is to discuss the microbiome as a potential target for therapeutic intervention in patients with chronic liver failure. I have no disclosures to report. The outline and my objectives today are to review the current data regarding the microbiome and what we know about it in chronic liver disease, define current treatment strategies that exist that currently target the microbiome and liver failure, and talk about future directions that may influence our ability to manage chronic liver conditions that's sort of woven throughout the talk. Microbiome research has exploded in the last several years. And there's a multitude of reasons for this. But chief among them is our ability to further characterize the microbiome in patients both in health and disease by utilizing molecular biology techniques, specifically 16S ribosomal sequencing to characterize bacteria that cannot be grown in culture. If you look at a PubMed search for the microbiome in 2008, it yielded about 525 publications. And this rocketed past 14,000 in 2019. When talking about the microbiome in health and disease, it's important to recognize that there's a huge micro burden that lives on animal hosts. The adult human consists of approximately 10 trillion eukaryotic cells. But there are approximately 100 trillion microbes that live on human beings. Therefore, there's about a 10 to 1 ratio of microbes to eukaryotic cells in the human cell mass. It's been recognized for many, many years that these microbes likely have a significant influence in health and disease. It's been stated before as early as 1977 that microbes profoundly influence some of the physiological processes of their animal host. Some of these normal mechanisms that help influence health in patients in normal state of health include alterations in glucose metabolism. They help liberate short chain fatty acids. They're important for biosynthesis of certain vitamins, especially things like vitamin K. And they also are important in bile salt metabolism in the intro hepatic circulation. Mutualism refers to the homeostasis that exists between hosts and the large divergent microbial population that lives within them. In the case of humans, we're really talking about the significant bulk of microbes living within the GI tract. And there's stringent mucosal compartmentalization that keeps microbes, even those with pathological capabilities, from causing problems. When they do escape, the first line of defense includes the mesenteric lymph node blockade. And the liver eventually can serve as a vascular firewall preventing pathologic organisms from entering into central circulation. Mutualism can be significantly disrupted in patients with advanced liver disease. When you think about how the liver is positioned in the body and as it relates to circulation, specifically with regard to the portal vein, you can see that the liver is positioned in a way that makes it uniquely influenced by the intestinal microbiome. When thinking about the liver-gut axis in both health and disease, it's important to recognize that in healthy hosts, we have numerous mechanisms that protect us from microbial invasion. But this can be severely disrupted in patients who developed advanced liver disease. Among these include the secretion of mucus within the intestinal cells, which provide a mucosal barrier. There are normal tight junctions between the mucosal epithelium and the intestinal tract, preventing translocation. The bacteria themselves are a more normal type of mixture of bacteria. And very few of these cells traverse the intestinal membranes and get into circulation. When they do, they're typically cleared by mesenteric lymph nodes. And rarely do they enter into the liver. When they do enter into the liver, the ability of Kupfer cells to clear them without resulting in significant over-inflammation is very well preserved. But this gets disturbed dramatically in patients with advanced liver disease and cirrhosis. First, we have a significant change in the microbial content of the GI lumen. Tight junctions are lost, allowing more ease of translocation of bacteria into pathogenic spaces. The mucosal barrier is often disrupted. And significant amounts of bacteria can enter into the portal venous circulation, where they then cleared by liver macrophages. This results in a significant increase in inflammatory markers, which then can translocate into central circulation, resulting in an inflammatory response. We believe this is probably central to the development of the phenomenon known as acute on chronic liver failure, which is a significant deterioration of patients with well-stabilized chronic liver disease that results in intensive care admission. That same slide has been demonstrated. Some of the features in that previous slide have been demonstrated in experimental studies again, when mice are given pathogenic E. coli in gastric gavages, it's rare that microbes enter into the liver. And when the animals are sacrificed, the liver remains relatively sterile. Bacterial breakdown products can be isolated from the liver. But again, the mesenteric lymph nodes are primarily responsible for clearing bacteria in this situation, preventing them from ever causing problems. However, when pathogenic states develop, the liver now serves as a firewall, protecting the organism from bacteria getting into central circulation, where they can cause more problems. If bacteria are injected specifically into the portal venous system, we can see that then bacteria can be isolated from the liver, but they do not oftentimes cause systemic problems in patients that are otherwise healthy. When patients, or in this case mice, are given a chemical which disrupts the mucosal epithelium, we also see that there's a significant increase in the amount of bacteria that can find their way into the liver. Liver disease disrupts mutualism in patients who have normal flora that live within the gut. We can see that the amount of IgG to certain bacterial species, including E. coli, bacteroides, Bifidobacterium, and Klebsiella species increase dramatically when patients have cirrhosis. But now this becomes an important question. Are alterations in the microbiome what are influencing the development of liver disease? Does liver disease itself cause significant changes in the microbiome? Or I think, as we'll see in some other studies that I'm going to show you, it's probably a combination of both. There's no question that when the liver becomes diseased, this results in pathologic changes within the body that facilitate microbial translocation, et cetera. We will also see that alterations in the microbiome are likely influencers in the development of certain liver inflammation and pathogenic states. Dysbiosis is the term that is applied to significant changes in the microbial flora that live within different organisms. Now, this is a busy slide, but I'm just using it to demonstrate that because of our ability to now sequence bacteria that are not easy to grow in culture, we can see that there are significant differences in the microbial concentrations of different species, both in healthy and diseased states. So in the slide on the left, in the frame on the left, we can see healthy patients are represented in green, and diseased patients with cirrhosis are represented in red. And you can see that there are significant differences in the amount of certain bacterial species between health and disease states. This is such a powerful change that the microbiome has been used to predict patients or identify patients with certain disease states. And cirrhosis has been particularly robust as a condition in which the presence of disease can be accurately predicted by changes in microbial species. So in this study, looking at changes in microbial flora in different patients, those with known cirrhosis, known colorectal cancer, inflammatory bowel disease, obesity, and two different studies looking at patients with type 2 diabetes, you can see that the alteration in the microbiome very accurately predicted patients who had cirrhosis with an area under the curve of 0.945. I'm gonna switch now to talk about the microbiome and its influence in alcohol-related liver disease. So as everybody recognizes, not everybody who drinks excess alcohol develops cirrhosis or liver problems. In fact, among all heavy drinkers, the amount of people that develop cirrhosis is only between 20 and 30%. This also holds true for mice. And so we have identified certain mouse strains which are resistant to the effects of alcohol. And this serves as an effective model for examining the effects of alcohol and how we may modify alcohol-related disease. So two sets of mice, those that were sensitive to alcohol and those that are resistant to the effects of alcohol were fed initially with the Liber de Carli diet and then transitioned to an alcohol-containing diet. And as was expected, those that are resistant to the effects of alcohol developed no significant steatosis, but those that are sensitive to the effects of alcohol had marked steatosis of the liver as seen on histological specimens. We also see that there is a significant increase in the amount of triglyceride contained within the liver in the alcohol-sensitive mice. And as would also be expected, ALT levels rise in the mice that are susceptible to the effects of alcohol. Interestingly, but not surprisingly, we also see significant increases in certain inflammatory markers, including TNF-alpha, interleukin-1-beta, CCL-2. And as it relates to the microbiome, we see that there is a fall in bacteroides species in patients that are susceptible to, excuse me, there's a lower amount of bacteroides species in mice that are susceptible to the effects of alcohol. When looking at different studies that characterize the bacteria within these different hosts, we can see that mice that are sensitive to alcohol as outlined in green are significantly different than those that are outlined in red. But now, can the microbiome be targeted in the patients with, or in this particular case, mice who have alcohol-related liver disease? So in this particular study, two things were done. First, the mice who had resistance to alcohol effects, they took stool samples and transplanted these into the mice that were sensitive to alcohol. A second trial in an effort to modify the effects of, or to modify the microbiome was by feeding them pectin. We can see that after fecal transplant or after administration of pectin, the microbial flora resembled that of the resistant mice. Now, in the fecal transplant mice, the bacterial species were almost identical. In the pectin-treated group, the bacterial species were significantly different, but again, it served as a protective measure for protecting against the effects of alcohol. And this is where it gets very interesting. So in the mice who underwent fecal transplantation, mice who were previously sensitive to the effects of alcohol now became protected when they had fecal transplant from the resistant mice. So in this particular case, we see that the mice who were previously susceptible to the effects of alcohol, but who had a stool transplant from the resistant mice developed no significant alcohol-related liver damage. Similarly, they were protected against the development of triglyceride deposition. ALT levels remain normal. And similarly, the inflammatory markers did not increase and protected mice from the development of alcohol-related liver damage. Switching to a different disease state, I'll talk briefly about primary sclerosing cholangitis. Primary sclerosing cholangitis is a autoimmune disease that affects the bile ducts and results in significant liver damage, cirrhosis, and the need for liver transplantation, often in very young patients. This has also been a hot area of study whether or not the microbiome may be influencing the development of PSC. Remember that patients with PSC oftentimes have co-existent inflammatory bowel disease. Therefore, they have increased risk for having damaged intestinal epithelial layers. And it's believed that enteric microbiota may be influencing the development of inflammation within the cholangiocytes. But all patients with ulcerative colitis or Crohn's disease do not develop PSC. So the question is, are there immunogenic modifiers such as differences in HLA alleles, toll-like receptors, cytokine receptors, or others that prevent the development of an inflammatory liver disease in patients with ulcerative colitis? So again, we recognize that certain patients who have constant exposure to bacterial toxins do develop chronic biliary injury. They develop biliary cirrhosis, cholestasis, ductopenia, and several of these patients will ultimately develop cholangiocarcinoma. The microbiome as a potential target for therapy has been trialed in PSC, but with disappointing results. Oral antibiotic therapy has been used as a disease modifier in patients with PSC. And while it does result in improvements in biochemical markers such as alkaline phosphatase, AST and ALT, as of today, there have been no long-term studies that have demonstrated a significant modification of the disease course in PSC, nor has it resulted in significant improvement in survival. Switching again to another possible therapeutic target is hepatocellular carcinoma. Hepatocellular carcinoma is a devastating disease. It's currently the third leading cause of cancer death worldwide. It's the cancer with the fastest increasing incidence in both men and women in the United States. And while we were optimistic that with the wide availability of treatment for chronic hepatitis C, that we would see a decrease in hepatocellular carcinoma, we still see a rapid increase in incidence due to non-alcoholic fatty liver disease, which is now the fastest growing cause of hepatocellular carcinoma. The microbiome has been studied in patients with hepatocellular carcinoma. And again, we see a couple of different mechanisms as possible tools that are leading to the development of HCC in patients with chronic liver disease. One, again, is the phenomenon of the leaky gut where changes in portal hypertension result in disruption of the mucosal barriers that protect patients from bacterial translocation. This results in translocation of both bacteria and inflammatory mediators, resulting in stimulation of inflammation, which may serve as a trigger for the development of hepatocellular carcinoma. The presence of dysbiosis, again, where they have a significant change in the bacterial milieu of the intestines may result in change in metabolite production, again, resulting in changes in stimulation of hepatocytes and liver-derived stem cells, which may also influence the development of hepatocellular carcinoma. The microbiome has been thought as a potential therapeutic target, both as impeding the development of cancer by impairing the pathways that result in tumor promotion, but also as it modifies immunosurveillance, which may have tumor suppressive properties that may prevent the development of hepatocellular carcinoma in the first place, recognizing that patients with chronic liver disease are at high risk for developing HCC. Hepatic encephalopathy, as you've heard about from other presenters this morning, is also a known target of microbiome modification for therapy in chronic liver disease. As discussed previously, ammonia is central to the pathogenesis of hepatic encephalopathy. It's not the only mediator. Circulating ammonia level rises from the breakdown of ingested amino acids and from microbial flora. Ammonia is normally removed from circulation in healthy liver and skeletal muscle, and current treatments today are aimed at alterations in production of ammonia, typically nonabsorbable disaccharides, such as lactulose, and usage of antibiotics. The first major study to report on rifaximin was published in the New England Journal of Medicine in 2010. This demonstrated that in outpatients treated with combination therapy of lactulose plus rifaximin, there was a significant decrease in hepatic encephalopathy-related breakthrough events, and there was also a decrease in hospitalizations. So many of us started using this in our hospitalized patients without any data to support that, however, more recently, a study has been performed in hospitalized patients with overt hepatic encephalopathy. Again, when rifaximin is added to lactulose, there's a significant difference in the primary endpoint of complete reversal of hepatic encephalopathy. There were also differences, improvement in, there were significantly different improvements in the secondary endpoints of hospital mortality, as well as hospital length of stay in patients who were treated with rifaximin. So how does rifaximin work? This is a very broad-spectrum antibiotic that has activity against both gram-positive and gram-negative bacteria, both aerobic and anaerobic species. It's very similar in structure to rifampin, and it inhibits bacterial RNA synthesis by binding to the beta subunit of bacteria-DNA-dependent RNA polymerase. Interestingly, experimental studies by Dr. Jasmohan Bajaj demonstrated that there is minimal impact, even though this is a very broad-spectrum antibiotic, on the overall normal flora found within the intestines. This is a non-absorbed antibiotic. There's decreased concentration of coliforms within the upper GI tract, and there were some increases in health-promoting lactobacillus and bifidobacterium, and there was a reduction in the more deleterious proteobacteria in the colon. But importantly, compared to other antibiotics that have been trialed in hepatic encephalopathy, rifaximin does not select for resistant organism. And it absolutely alters the cytokine expression profiles in patients who are treated with rifaximin in chronic liver disease. It prevents stress-induced increase in mucosal inflammation. And so again, we can see that rifaximin does alter how patients behave, but it does not have a dramatic change in the overall flora of the intestine. So the take-home points from my talk this morning, the microbiome does have major impact in both health and disease. The liver is uniquely influenced by, but also influences the host microbiome in significant ways. Modification of the microbiome has already been a proven target for certain therapies in chronic liver conditions, including hepatic encephalopathy. But further studies are required to evaluate other microbiome-targeted therapies, which may be beneficial, such as the development of primary sclerosing cholangitis and hepatocellular carcinoma. Thank you for your time. Thank you. So thanks, Jody, and thanks, everybody, for your patience. So we've got about 15 minutes for questions, if there's any questions from the floor. And we'll just get everybody, all the speakers have a microphone, mention which speaker, specifically, you're directing your question to. So microphone over here. Hi, my name is Sade from Salinas, California. My question is for Dr. Ram, the first presenter. So I just wanted to ask, so you did talk about the difference, the management of acute and chronic. And you also spoke about the difference of ICP elevation in acute and no ICP elevation in chronic. And you spoke about the management using reduction of ICP for the acute patients. For the acute patients, do you also use lactulose and rifaximin, in addition to ICP management using CRT, or you don't consider using those management techniques? Thank you for that question. So to my knowledge, there's no data supporting the use of lactulose or rifaximin in ALF, as in contrast to what you saw that Jody showed in ACLF with cirrhosis. Having said that, there is a reasonable argument to be made, and some of my colleagues do use, at our center use lactulose empirically, saying that if you're invoking ammonia, hyperamnemia, is it reasonable to use that? So the answer is I think it's an area of debate. My sort of counterargument, or not counterargument, but just the counterpoint to that, I think it's about a supply-demand mismatch. You have such a severe state of hyperamnemia in achilles bifida that I think you need something more aggressive than what you see in cirrhotics. So you may use lactulose. You could use rifaximin. You may get some decrease in ammonia in the ALF state. But is it enough to counteract with a higher sort of strategy that is more aggressive, such as hydroCRT? Yeah. Can you hear me? Yeah. I have both an apology and a question. I was supposed to be the moderator. I apologize for being late. But the question is predominantly for Dr. Kai Singbada. If anyone else has a different opinion, you can chime in. Frequently we have a patient with an ammonia, I mean a sodium, in the 120s, and then with a high MELD, and you have availability of an organ the next day. And you have what the next day? Sorry. Availability of an organ for transplant. So we always are stuck in the dilemma. I was curious to know what your approach is. This is one of the scenarios. I was referring to in my talk indirectly, when I worry about osmotic demyelination syndrome and when things can go wrong, especially when you have these like really sick patients going to the OR with a sodium of 120. Things don't go well in the OR. They receive tons and literally tons of bicarbonate pushes, and they come out with a sodium of 142 all of a sudden. Those are really, in quotes, frightening scenarios. And when we have these patients in the unit or when we have the situation in the unit, it gets very tricky to control the sodium. It usually always means you have to run D5 on the side, or the same issue arise when you have these patients. Put them on CRT or standard CRT solution. Have a sodium of 140. Patient has a sodium of 120. The only way to do it safely is really very close sodium checks, and always, or like almost always, you have to run D5 in the background just really to limit the increase. Those are very challenging situations, very demanding on the entire team. But at the same time, if you don't pay attention, things can go wrong, and you can easily lose a patient because of ODS if you don't pay attention to those rises in sodium. Thank God it doesn't happen that often. That's why people sometimes get relaxed about it. But when it happens, and as I said, it's my own experience over the past year or our own experience, it's very devastating for the entire team. Hi, my name is Fraser McKay. I work at Lahey Hospital Medical Center. We do a lot of liver transplants. We also take care of a lot of the pre-transplant population in the medical ICU. I've noticed that the use of not anticoagulation but DVT prophylaxis is pretty highly variable in a lot of our providers, both our APPs and our fellows and attendings. So I was wondering if I could kind of get some insight into recommendations from, I guess, any of you or all of you about what are your kind of thresholds for initiating VTE prophylaxis and what are kind of your hard stops in terms of, you know, whatever platelet-level INR, if anything. Thank you. So that's a very good question, and it is a very challenging question. And particularly in the critically ill population, we obviously don't want to tip the balance. And a lot of the studies I showed you, the challenge was a lot of these were in ambulatory cirrhotic patients that were in the community. So the first question is most of these patients need some form of DVT prophylaxis. So probably the least risk is your pneumatic compression stockings, and pretty well you can use that for anybody. So in the ICU population, there's not a whole lot of downside to that. The question of chemical prophylaxis or when would you use something like heparin or low molecular weight heparin, this is where you really have to look at the part of it relates to the stability of the patient. So for a ward cirrhotic patient waiting on the transplant list, commonly a lot of these patients in hospital will be on heparin. Once we start to really get concerned, as you said, it's really once again the platelet count below 50. So if you have a patient who's ambulatory with a platelet count below 50, you know, you really, that's somebody you probably would avoid it. And even in the ICU setting of the non-ambulatory patient, if the platelet count was less than 50, we would probably use pneumatic stockings. I hope that helps. Yeah. Thanks very much. I think we all agree. Yeah. Yeah. Hi, my name is Emilio Mazum from Virtual Health in New Jersey. My question is kind of a piggyback to that question, but in the patients, a real world scenario patient who has end stage liver disease comes in, has a pulmonary embolism. And INR, you know, what it's worth is 2.2. Creatinine is 2. Do you use Lovenox? Do you use Warfarin ultimately when you have to get them out of the hospital? And if you use Warfarin, how do you dose it when someone already has a baseline elevated INR? It's always been a quandary for us in terms of what to use in this scenario. So I guess this, can you guys hear me? Yeah, so this microphone is working. So that's another very good question. So at the end of the day, the thing to remember is before having an INR of 2.2, that patient, like I said, because INR was developed for Warfarin, we know that that patient isn't anticoagulated and needs to be anticoagulated for their pulmonary embolism. So I agree, the first question, I agree with low molecular weight heparin and renal dysfunction can be a challenge. So in some of those cases, we'll actually use unfractionated heparin to begin with and put them on a heparin infusion. I agree with you too that it is a challenge to titrate somebody if you are going to send them home on Warfarin. So it is technically easier to give them low molecular weight heparin. Also, that would be more commonly done in the setting. It's easier to follow a patient if they're waiting for a transplant or if they're obviously being followed by their generalist and a hepatologist and probably somebody from a coagulation clinic. Warfarin is used commonly in the outpatient community for things like portal vein thrombosis and DVT, partly because of ease of administration. And right now, we still target the same level between two and three. So it is a difficult situation. We really don't, unfortunately, unless you're doing, unless you're monitoring factor levels or doing TAG or Rotem, it's a very difficult question to deal with. Thank you. Without any basis for doing this, for my patients with chronic elevations INR, I usually try to target one point above what they were at. Again, there's no science behind that. Hi, my name is Sleepin. I'm a resident at St. Agnes Hospital in Baltimore. My question was for the second speaker. I'm sorry, I forgot your name. You were mentioning one of the challenges in cirrhotics with elevated INR was percutaneous procedures from our radiology colleagues. At your institution, do you have a protocol set in place or what kind of communication do you have with radiology colleagues? That's a very good question. And we actually have had, this was one of the things that was highlighted in our institution was actually overuse of blood products. And you may find in your same institution that you have these standard operating procedures that a radiologist is calling you with a manual that's 20 years old and says, I want an INR of less than 1.7. We know that even FFP, if you were to test this, will have an INR of 1.7. So this actually ended up being a collaborative approach where we met with kind of the, we had joint rounds with radiology and blood bank. And the idea was that we would start, really one of the ideas is that in blood bank, we can get viscoelastic testing. At the moment, we can't get it at the bedside. And the arguments is that if it would be, we've talked to blood bank, this is the patient, this is the scenario. Recommendation is not to give blood product. And generally, most of the time, they're reasonably compliant. And we all agree that sometimes it also depends on the clinician on the day. And we all understand that it's joint responsibility if there is a complication. So we've tried to educate our colleagues with data. The one thing I would mention too is that if at the end of the day, you do have to give blood product or if you are giving FFP, that the procedure has to be done at the time that product is given. So you don't order another blood test. You don't order another fibrinogen or platelet count because it's going to wear off after a period of time. So it's been an uphill struggle that I think everybody deals with. Thank you so much. Hello, Rachel Steinhorn from Boston, Massachusetts, addressing my question to the first speaker. You had mentioned using renal replacement therapy for hyperammonemia. When we have initiated it for kind of the same reason, one of the issues that our renal colleagues have brought up is the fact that ammonium can act as an osmotic force in the same way that sodium can, so that rapid shifts can lead to an increased risk for further cerebral edema secondary to those osmotic changes. Have you had to address that with your own cases, or do you have a regimen in place? So thanks for that question. So we have not had that concern from the nephrologists at our center. In fact, our moderator, Prem Kandahar here, is one of my colleagues who is a neurointensivist as well as a general intensivist who has been instrumental in coming up with a specific protocol to specifically address this issue. And actually, maybe I can have you go into this, yeah? So without my slide deck, well, the units have to kind of, but if you look at the units, it's micromole per liter versus milligram per liter. So the ammonia compared to BUN, for instance, is a fraction of BUN. So your ammonias have to be, like, sky high before you would have a dramatic osmotic contribution to it. So from that standpoint, for everyday encephalopathic patients with high ammonias, I'm not too worried about. But if you get an ammonia that's exorbitantly high, you're, like, in the 7 to 100 to 1,000 range, then maybe there's something to think about, maybe even go. But the caveat there is the quicker you bring down the ammonia, the less you have brain injury, so it's a tradeoff. There's no real easy answer for when it's exceptionally high. Sometimes you can use an osmotic swap, we call it. The analogy I use is fight osmo with osmo. So you can give mannitol to kind of as a placeholder for the osmolarity. But in every situation, you'll never have to do that. Usually the BUN is the bigger problem. Focus on the BUN. If they are rebounding, it's the BUN. It's the issue more than ammonia. That answers it. And I think one thing to remember, too, that if you look at acute liver failure over the last 20 years, the rates of cerebral edema have actually dropped very significantly. And we published a paper back in 2018 that we looked at about 1,000 patients with a multivariable model. And one of the things that was actually associated with this was if you put somebody on CRRT, the odds of death actually goes down by about, after adjusting for severity of illness, by almost 30%. What was happening probably in the 90s is we were using hemodialysis in these patients, and we were getting these rapid disequilibrium changes, and we were probably killing people. Hi, good morning. Brittany from Orlando Bay Medical Center. Having a family member that was affected by PSC and got a transplant, I was really interested about your microbiome talk, and I'm just curious if there's any post-transplant interventions or studies that are ongoing about how to prevent the recurrence of PSC with microbiome treatment. The biggest studies that were being performed were in antibiotic therapy, and there's still some clinical trials going on, but to date those have been pretty disappointing. And even though we refer to this as an autoimmune disease, a modification of the immune system post-transplant, immunotherapy, pre-transplant immunotherapy doesn't prevent recurrence. And the recurrence risk is really high, and it's very frustrating for patients and it's frustrating for us. PSC is a difficult disease to manage all the way around, and as of now there's lots and lots of interest on this, and there's lots of preliminary studies and animal studies going on looking at the microbiome, but today there's been no effective targeted therapy. And even though, like that slide I showed you, you can make some biochemistries normalize or improve, it does not seem to have an impact on long-term outcomes or disease progression, at least with the current data that's available. So I think this will bring our session to a close. We want to thank you for your participation and thank you to all the speakers.

drug metabolism

dosing considerations

liver failure

analgesics

sedatives

antimicrobial therapies

altered GI tract function

decreased perfusion

drug absorption

distribution

metabolism

elimination

individualized approach

hepatic blood flow

protein binding