Hi, everyone. My name is Dean Karvelas. I practice critical care and hepatology at the University of Alberta in Edmonton, Canada, and on behalf of my two esteemed colleagues, we thank you for coming to this lunch symposium looking at the critical care management of acute kidney injury and multi-organ failure in patients with decompensated cirrhosis. The format of this is that I'll go through a brief introduction, and then there'll be three subsequent presentations, and then there'll be time for questions at the end. So just to kind of give you some background that when we talk about patients with cirrhosis, or these are patients with chronic liver disease that get fibrosis or scarring of the liver, where they can kind of progress from a compensated state where there's evidence of fibrosis and portal hypertension, where eventually they can develop decompensation with complications such as ascites or hepatic encephalopathy, and generally when those patients develop multi-organ failure, we call this acute on chronic liver failure. And this condition, ACLF, is characterized by an exaggerated inflammatory response, which can lead in many cases a very predictable way to multi-organ dysfunction, where it is associated with a high mortality in the absence of liver transplant. In some cases, there are a couple of different definitions that are out there, and some have tended to discriminate the natural history of decompensated cirrhosis with ACLF, potentially based on prognosis, and the potential for reversibility. So probably the two largest organizations, the European Association for the Study of Liver and the ASLD, agreed on a definition of ACLF that, just so we're comparing apples with apples, and also for the purposes of research studies to make sure that we're looking at more homogeneous populations, that we're talking about patients with decompensated cirrhosis with a decompensation in hepatic function, either secondary to liver injury or some systemic precipitant, culminating in multi-organ failure and associated with a high mortality at three months. So where we get some of these definitions of severity of ACLF, one landmark study was the Canonic Study that was published by Richard Moreau and the Cliff Group, that the simplest way to think about this, and it's similar to the way that intensivists think of critical care patients. We're familiar with the SOFA score, which has six organ failures. They were able to derive a liver-specific score called the Cliff SOFA, and then subsequently the Cliff C ACLF score, where essentially, as you developed one, two, or three or more organ failures, this was associated with a stepwise increase in mortality, to the point that when you had three or more organ failures, you were looking at over 80% mortality at 90 days in the absence of a liver transplant. One of the things that's important is that acute kidney injury is a common precipitant or common organ failure in ACLF, and it's also the added increment of having AKI has a significant impact on mortality in ACLF patients. And the reason why we're spending so much time talking about HRS is because there are many studies now that show that patients that develop hepatorenal syndrome have a high associated mortality in the setting of ACLF, and this comes from a variety of different studies. But the challenge that we run into is that we haven't really made a lot of headway with regards to reversibility in survival in patients with ACLF, and particularly hepatorenal syndrome, partly related to the fact that these patients have a high transplant-free mortality. So we're going to be delving into specific aspects of ACLF and the treatment of hepatorenal syndrome over the next three presentations. For the next 15 minutes or so, I'm just going to discuss the management of ACLF and multi-organ failure in the ICU and kind of look at a little bit more of a global perspective at ACLF. So as means of an outline, we'll review the indications for ICU transfer and when you might be consulted for patients with cirrhosis and multi-organ failure, describe volume status assessment and management in ACLF patients, particularly with sepsis and septic shock, review respiratory failure and ventilatory strategies in ACLF, talk a little bit about renal replacement therapy strategies, just building on what we've talked about already, and finally, understanding prognosis in ACLF with and without liver transplant. So as we kind of mentioned earlier, there are a couple of different definitions of ACLF, but can we just go back one slide, please, sorry? The most widely used definition is the European CLIF definition, and this is based on six organ failures that kind of derives its origins from the SOFA score, hepatic, renal, neurologic, coagulation, circulatory, and respiratory. And the main reasons why we get consulted to assess these patients and to bring them to the ICU primarily is for circulatory management with vasoconstricting agents, respiratory management with noninvasive or mechanical ventilation, hematologic support, and then finally, renal support, which we've talked about in detail. So when looking at circulatory failure, we know there are well-defined pathophysiological mechanisms that lead to circulatory dysfunction in cirrhosis patients, and this kind of develops primarily as a result of portal hypertension and portal systemic shunting, which leads to shunting of vasodilatory molecules such as nitric oxide, which causes both splenctic and systemic vasodilatation. And generally, this leads to a decrease in systemic vascular resistance, and for a period of time, the patient is able to compensate by increasing their cardiac output. And for this reason, if you were to do bedside echocardiography or point-of-care ultrasound on these patients, a lot of them tend to look hyperdynamic. But after a period of time, you can only compensate so much, and you develop a state of relative arterial hypovolemia, which leads to upregulation of the renal angiotensin system, the sympathetic nervous system, and the release of antidiuretic hormone. And in a sense, this is what leads to hyponatremia through sodium and water retention, and eventually renal vasoconstriction and hepatorenal syndrome. So one of the first things to mention is that one of the challenges that we deal with is volume overload, not just from an AKI perspective, but from a multi-organ dysfunction perspective. And it's always important to remember that fluid is a drug, and there's well-described literature in the general critical care population that has showed that hypervolemia or volume overload is associated with end-organ complications. And one of the particular ones that we deal with is intra-abdominal hypertension. So one of the cautionary tales, and what it really highlights, is using albumin in well-defined settings. And those settings are hepatorenal syndrome for replacement after large-volume paracentesis, and in the setting of spontaneous bacterial peritonitis. But this study that was done by Dr. China and colleagues in Britain that was published recently in the New England Journal postulated essentially randomizing patients to transfusing albumin to a target serum albumin level of 35 versus standard of care. And as you can see here, while the patients in the albumin arm got significantly more albumin, there was no difference in most of their primary outcomes, specifically mortality, renal dysfunction, or infection. More so, what it was associated with was an increased rate of pulmonary edema. And this highlights something that Jody mentioned with regards to the confirmed trial, that over-transfusing patients with albumin can lead to pulmonary complications. The other thing to mention is that there is sometimes a reticence to use crystalloid as a resuscitation fluid in patients with acute on chronic liver failure. Generally, based on the literature at present, it is reasonable to use this as our standard resuscitation fluid, and as mentioned before, to consider albumin for well-defined indications. And this was highlighted by a study by Rocky Mywall and the group at ILBS in New Delhi, where they randomized patients to either albumin or plasmolyte. And while they found that in the albumin group, patients were more likely to attain a mean arterial pressure of 65, which is what is recommended for ACLF patients without hepatorenal syndrome. There wasn't really any statistically significant difference in any of the other significant endpoints. And particularly, they found that the 20% albumin group was associated with higher rates of pulmonary complications. And it should be noted, although they mentioned that there was a reduced need for renal replacement therapy, that finding actually was not statistically significant. So we know that there's been changes in the ICU over the last several years with regards to assessment of cardiac output and volume status. And thankfully, there's been a shift to the use of dynamic techniques, such as point of care ultrasound and bedside echocardiography, which gives us a lot of information. And we're relying less on static techniques, such as central venous pressure and other factors. And this is very important in patients with ACLF, because often, it can be very difficult to get accurate assessments of interabdominal hemodynamics in patients that have, for example, ascites or a significant amount of third spacing of fluid. So the second question that comes up with circulatory management is, we know that targeting a higher mean arterial pressure is associated with reversal of HRS in patients with HRS-AKI. But what about everybody else? Should we be using a target of a MAP of 65 or targeting a higher MAP for patients without AKI? And the short answer is probably a MAP of 65 in the absence of AKI is OK. And this was highlighted recently in another study that came from Raki Maiwal's group in Delhi, where they stratified patients or randomized them either to a higher MAP of 80 to 85 versus kind of the standard of care, which was a MAP of 65. And they didn't notice any statistically significant difference in hard endpoints. So at this point, generally, what we advocate for is performing early baseline assessment of volume and perfusion status and cardiovascular function in ACLF patients. In terms of the backbone of resuscitation for shock, it still is balanced crystalloid solution similar to what we do for the general critical care general sepsis population, including lactate and Ringer's lactate and plasmolyte, and generally give albumin if there is evidence of a level one indication such as SBP, HRS, or large volume paracentesis, but not as a primary resuscitation fluid. In the case of vasoconstricting agents, norepinephrine, as mentioned before, outside of HRS-AKI continues to be first line. And as mentioned before, we're targeting a MAP of 65 in the absence of AKI. And as Jodi eloquently showed, in the presence of HRS-AKI, then we will target a higher MAP to try to lead to HRS reversal. One other thing to mention is to always rule out adrenal insufficiency. And you can consider a short course of hydrocortisone therapy if you are concerned about relative adrenal insufficiency or a patient is on significant vasopressors. We know that one of the biggest precipitants of ACLF is infection, and this has become – because in many senses cirrhosis patients are immunocompromised. So they – cirrhosis patients tend to lose the tight gap junctions in the gut, which leads to bacterial translocation. And because of portal systemic shunting due to scarring in the liver, quite often bacterial pathogens can evade the kind of standard surveillance of your immune system, including the reticuloendothelial system, which leads to, you know, seeding of sterile spaces such as the blood, leading to bacteremia, and also seeding peritoneal fluid, which leads to SBP. So we know currently as described in the surviving sepsis guidelines that speed is life. And early appropriate antimicrobial therapy is probably the most important thing. And this highlights even in the subpopulation of ACLF patients with SBP and septic shock that with each hour of delay in appropriate antimicrobial therapy led to an increase in mortality by almost 1.8 times. Okay, so looking at respiratory failure, it's important to remember in ACLF that there are causes of respiratory failure that are specific to cirrhosis, such as hepatic isothorax, hepato-pulmonary syndrome, and alpha-1 antitrypsin. And then there are nonspecific causes, which are standard things that we deal with in general critical care patients, such as pneumonia, thromboembolic disease, pulmonary edema, etc. So really there isn't a lot of directed literature with regards to ventilatory strategies, specifically in ACLF patients with lung injury. So really this is borrowed literature from the general critical care population where we use a lung protective strategy that you're all familiar with. The only thing that I will mention is that using higher PEEPs, you have to be careful in a patient with ACLF because they can be very preload dependent. And often if you really start with a very high PEEP, you can develop hypotension. So that's always one thing to keep in mind. Just a quick comment on variceal bleeding. This is another reason why we bring patients to the ICU often for airway protection to facilitate endoscopy and banding or gluing. That the main predictors of mortality in an ACLF patient with a variceal bleed have more to do with the severity of their underlying liver disease or complications of liver disease than it does with what the endoscopist actually sees. And it should be mentioned that there is data to show that turlopressin actually has been used as a vasoconstricting therapy in treating variceal bleeding to decrease the portal pressure gradient and has been actually associated with decreased mortality. Currently we're primarily using octreotide, which is a somatostatin analog, which is associated with a mortality benefit in combination with an endoscopy, which is what most of us do, but not on its own. And while it should be noted that there are selected patients that an early transjugular intrahepatic portosystemic shunt or creating a shunt to bypass the liver may be associated with decreased one-year mortality. You have to be very careful in patients with very severe liver disease, and most ACLF patients often have contraindications to TIPS because you might exacerbate their hepatic encephalopathy or precipitate further ongoing liver failure. One thing I want to mention on altered mental status is that in the ACLF patient, not everything is hepatic encephalopathy, and we know one of the biggest challenges is that somebody that comes in with alcohol-induced liver disease, is it alcohol withdrawal, is it hepatic encephalopathy, and the treatments can sometimes be different. One thing to mention is generally for the workup on the first presentation, we'll work up a variety of different differential diagnoses, you know, ruling out structural causes, other kind of metabolic causes, et cetera. But generally if you've honed down on the diagnosis of hepatic encephalopathy and somebody comes in with a repeat presentation, generally you don't have to do this kind of elaborate workup. Again, you can probably just do it the first time. So in terms of the management of these patients, generally we're getting called when they have difficulty with airway protection. West Haven criteria, hepatic-grade coma of three is kind of equivalent to a GCS of nine or less. So generally these are patients that we're bringing to the ICU for airway protection and to avoid aspiration and facilitate using medications like lactulose and rifaximin and PEG. And as mentioned before, generally we use an extensive workup on the first presentation. As has been mentioned before, treatment of hepatic encephalopathy, you know, lactulose has been well described in the hospitalized cirrhosis population. There are some cases in the ICU though where patients may have ileus or bowel obstruction or bowel distention where this might not be an appropriate medication, and you don't tend to get the bowel distention with polyethylene glycol. We probably need more literature to validate this. This is a good second-line agent. And I should mention too that while there isn't as much direct data of rifaximin in the ACLF population, which is a non-absorbable antibiotic, we tend to use it in the ACLF setting. With regards to medications, we prefer sedating medications with short half-lives like propofol and dexmedetomidine in patients that require it for mechanical ventilation, and we try to avoid the use of benzodiazepines as this can exacerbate HE. And that's one of the benefits of dexmedetomidine is that you can use it in a patient where you may be concerned about alcohol withdrawal, but also use it in a patient that does not have a secured airway, that you can use it in somebody that's not intubated. And finally, while in, you know, the use of ammonia levels is very challenging in cirrhosis and ACLF patients, it doesn't tend to correlate with daily fluctuations with hepatic encephalopathy grade, so we don't advocate for serial ammonia testing. Really the one benefit of ammonia is if the level is completely normal, it likely potentially excludes hepatic encephalopathy and means that you need to look for another cause of a patient's altered mental status with ACLF. Talking about renal replacement therapy, as kind of mentioned from a couple of the different guidelines, generally continuous renal replacement therapy is likely safer to initiate, particularly in patients with HRS, because these patients are often hemodynamically unstable and pressor dependence. You're probably all aware that there was the START-AKI study that came out in 2021 that looked at standard initiation versus accelerated initiation of continuous renal replacement therapy. We don't, to this point, have a specific study in the ACLF population, but this showed that there was no benefit to accelerated initiation of RRT. However, one of the exceptions would be in somebody that you're potentially bridging to transplant, that you want to treat things like volume overload and electrolyte abnormalities prior to getting them to operating theater. In those cases, it may be preferential to initiate renal replacement therapy early. The other question that comes up is how do we prevent clotting of the circuit? Most patients with ACLF on CRRT can be run without circuit anticoagulation. In those patients that are prothrombotic, you can probably use citrate regional anticoagulation for short periods of time, for a couple of days. If you need it for a more extended period of time, then you might be considering something like prostacyclin. And I do want to mention one subgroup of the START-AKI study that showed that, potentially, in critically ill patients that required RRT, that there was potentially a benefit to the use of a continuous mode versus intermittent hemo, although this needs to be validated in other patient populations. As I was kind of mentioning before, there is good physiological rationale to preoperative conditioning a patient, certainly prior to liver transplant, because we know when patients with ACLF undergo liver transplant, they're exposed to a lot of insults, including lactic acidosis from clamping of the large vessels during the enhypatic phase. They can get hyperkalemia because of reperfusion of the donor organ that may have been on cold storage. And there's also hemodynamic instability and coagulopathy, where they may require significant volumes of blood transfusions or crystalloid for resuscitation. And we know that if somebody is going into transplant with AKI or oliguria, they may not be able to mitigate this kind of metabolic derangement. So the question becomes, should we precondition them with CRRT prior to liver transplant, or should we be running it interoperatively during the operation itself? And I just want to mention that, essentially, there was a pilot study that came out of our institution that the short answer is that it did not show that there was a significant benefit with interoperative versus preoperative CRRT. And there was also a systematic review that came out recently that kind of confirmed the same thing. So really, it highlights it's kind of a personalized medicine approach, that interoperative CRRT is certainly not for everyone. This is a very difficult population to study. But what I would highlight is that there are probably some patients that might benefit, specifically those patients that have oliguric AKI that are going to be a real challenge to get through transplant. In terms of taking it a step further, and there will be more presentations on this later on in the meeting, looking at extracorporeal liver support or other forms of blood purification in ACLF. To date, most of the studies have, unfortunately, been underwhelming. Currently, there is a study called the APACHE trial, which is looking at the role of plasma exchange in acute on chronic liver failure. And we're still waiting for the results of the study to be presented. Finally, I just want to mention that managing ACLF, in some ways, you can become a little bit nihilistic, because these patients, in the absence of transplant, quite often have a high mortality. But I just want to mention that there is importance in the work that we do. This was kind of a combined North American-European study. And what it showed is that those patients that come to the ICU, about 40% of them improved by at least one organ failure. And this can potentially have an impact, either in terms of getting them out to the ward, or in some settings, it might be making them a better candidate for liver transplantation. And I just want to mention that, you know, at this point, we don't have futility criteria to say when somebody is too sick for transplant. And this is also a topic for one of the presentations tomorrow. And part of the reason for this is that, you know, there are human decisions with regards to who we transplant and who we delist. And I just want to highlight this study that we published back in 2013. It was a multicenter Canadian study that showed that there was a significant difference, and this was before the CLIF-ACLF score, it was the regular SOFA score, that discriminated those people that were successful in getting a transplant versus those that died on the wait list. And there are studies from UNOS and from other places that show that the increasing number of organ failures increases your mortality quite dramatically on the wait list. However, those patients that do go on to get a transplant generally do okay. But really what it highlights, and we can't actually discriminate with them very well, at least with scores like CLIF-ACLF, and part of the reason for this is that we're only kind of doing hand-picked patients that we think would benefit. So there's the selection bias of the transplant team and the surgeon taking other people off the transplant list, but furthermore, certainly if you have four or more organ failures going for transplant, you're probably 35 and you don't have significant cardiopulmonary disease. And as mentioned before, this has also been showed from the big UNOS registry that actually showed that once you have four or more organ failures, your wait list mortality at 30 days is almost 95%. So really when you see data like this that shows an 80% survival at a year, that's a So if you're saying patients with four or more organ failures, 90% of them die on the wait list. That means it's really, you start with 100, you end up with 10 patients, and of those 10, 8 people are doing okay a year after transplant. So really what I'm highlighting is don't look at this graph in isolation. You have to look at the, it almost effectively is an intention to treat study. I do want to mention that with an ACLF patient, if we are able to improve their clinical status prior to transplant, this is generally a good thing, and it's been showed that if somebody comes in with ACLF grade 3 or 3 organ failures and they get better with ICU support, generally their natural history post-transplant, if they can be downstaged to a grade 0 to 2, they tend to do as well as somebody that was transplanted from the get-go with, for example, grade 1 or 2 ACLF. And I want to mention, too, that my two colleagues here can attest to this as well, is that transplanting high-risk patients with multi-organ failure, you're in the high-risk business, they are very resource-intensive. Generally these patients, certainly with more organ failures, they will be alive at a year, but they require often long ICU stays, a higher likelihood of needing renal replacement therapy, and also a higher rate of ICU readmission. So in summary, in terms of bridging the ACLF patient through liver transplant, consider ICU early on the ACLF patient who is potentially transplantable. Volume status, use dynamic tools like point-of-care ultrasound. Consider polyethylene glycol in ACLF patients with HE that have ileus or intra-abdominal hypertension. Renal replacement therapy in ACLF continuous modes may be preferable because of hemodynamic instability and potentially beneficial. With regards to preoperative or intraoperative renal replacement therapy, consider intraoperative renal replacement therapy if available in the patient is an uric-undergoing transplant. Transplant in ACLF is a resource-intensive procedure, but just remember there is a significant selection bias, which is a good thing, and weightless mortality is probably the most important outcome. Thank you very much.

acute kidney injury

multi-organ failure

decompensated cirrhosis

ACLF

CLIF-ACLF score

renal replacement therapy