Good morning. First I would like to thank the SCCM and SPENIC for giving me the opportunity to talk about ECMO. Candidacy in patients with liver disease is stretching too far. I have no disclosure or any conflicts of interest except the fact that I think that transplant medicine is extremely fascinating. At the end of my talk, I hope that you will have a better understanding on which patients with liver disease are candidates for ECMO and how this can be done safely. We'll start with a little bit of mythology. After Prometheus gave fire to humans, Zeus bonded him to a rock where an eagle would come every day and eat a piece of his liver. However, that liver will regenerate overnight, causing the eternal torment. While the liver can regenerate and advances in science allowed for temporary support for different organs such as dialysis for kidneys or ECMO for cardiovascular and pulmonary function, currently there is no system or substitute for the immune metabolic and sustaining function of the liver other than transplantation. We care about patients with end-stage liver disease because only in the United States, each year 600 children are added to the waiting list for liver transplantations. More than half would wait for an offer for more than one year, with 1 in 20 children dying while waiting for an organ. They are critical year requiring a recurrent admission to the intensive care units while receiving life-saving treatments and therapies such as mechanical ventilation, blood pressure monitoring, and medications plus dialysis. We all have taken care of those patients, and we know that they are in multi-organ failure. And to me, it's very relevant that we have those kids with big bellies that look like a peach, but they are a ticking bomb, as we can call. Pre-transplant, there is profound vasodilatation that leads to high cardiac output as well as distributive shock physiology. Hypoxia and fluid load will cause respiratory failure. During transplant, patient will experience massive fluid shifts and bleeding. There is increased risk for air emboli during dissection while patients are at high risk for cardiac arrest and reperfusion. Most operatively, we must focus on making sure that hemodynamics and oxygenation are appropriate to assure great function of the graft. If those complications are caused by reversible causes, all are potential ECMO indication per ELSO guideline. The question is who is an ECMO candidate when, according to the same guidelines, acute liver failure along with trauma and immune disorders, diseases are relative contraindication for ECMO. I mentioned in trauma and immune disorders, I both put the patient at risk for bleeding and sepsis, similar with patients receiving liver transplantation. Over the years, with improvement in technology and patient care, we were able to achieve similar outcomes for this patient population as we achieve for pediatric ECMO. A little bit of history, I think it's worth mentioning as well because it's relevant to the topic. In green, I have ECMO history and brown, it's transplantation. In 1950, the first heart-lung machine and membrane oxygenator were developed, followed by first successful ECMO in a neonate with respiratory failure in 1972. The first human liver transplantation was performed in 1967 by Stasel, the father of liver transplantation. Unfortunately, that baby died secondary to major bleeding during surgery. Major achievements in transplantation occurred in 1976 when cyclosporine was developed and we were able to prevent ejection. In 84, Dr. Stasel performed the first liver transplantation when ECMO was used successfully to save the patient. Since, more ECMO and liver transplantation were performed with improved outcomes. In the United States, 600 patients received liver transplantation in 2021, according with the transplant registry. We have about 10 to 15% of those have liver failure. Unfortunately, there is still 2.4 patients dying on the waiting list are too sick to be transplanted. According with literature, the survival rate for children receiving liver transplantation needing ECMO support is between 56 and 68%. But also live dashboard, each year, 3,000 to 4,000 cardiac ECMO cases are performed with a 55 survival rate. It is not a surprise that there is limited scientific evidence for use of ECMO in liver transplantation. PubMed search gives approximately 50 manuscripts on ECMO for liver transplantation with majority being case reports, case series, and systematic reviews. There are some papers that I thought are most relevant to our talk here. In summary, the number of patients per study vary with up to 34 patients for the larger study. As expected, VBA ECMO was used for hepatopulmonary syndrome and RDS, while VA ECMO was used for cardiac indications. Early ECMO initiation led to better outcomes with a survival to discharge rate approximately of 60%. Each day between liver transplantation and ECMO initiation, there were 3.5 days added to the ECMO run. Sepsis, multi-organ failure, as well as bleeding were major cause for death in this patient population. Most patients experienced at least one complication during ECMO run, and it is common to have more than two complications per run. Bleeding is the most common, while anticoagulation practice varied from no anticoagulation to full anticoagulation. In addition to bleeding, related to anticoagulation, patients are at risk for ischemic stroke and hypoxia. Channel replacement is very important, especially in VBA ECMO, when it can migrate and can cause hepatic infraction. As patients are on immunosuppression, there is increased risk for infection in sepsis. Now let's compare if complications for liver transplantation and ECMO are different from complications on ECMO or CAMRS. First, we are going to look at coagulopathy. Again, brown refers to liver transplantation, green to ECMO. In the pre-transplant phase, a balanced coagulopathy helps preventing spontaneous bleeding, as both clotting and fibrinolysis factors are impaired equally. Immediately post-transplantation, all coagulation and fibrinolysis factors are deficient. And thrombosis is more likely related to surgical variants or techniques, anastomosis-size presence of graft for the vasculature. Both post-surgical bleeding and thrombosis incidence are under 10%. How do I go back? Acute liver failure and ECMO coagulopathy are similar in the fact that both have impairment in clotting, as well as fibrinolysis, but the mechanism is different, leading to more bleeding and thrombosis on ECMO. According with VAT and all, bleeding occurs in 70% of ECMO cases, while thrombosis is up to 37.5% in that patient population. Neurodevelopment and liver failure in transplant area are of interest, as well as both patient population are at risk for developing cognitive motor and language impairments, according with multiple studies. Same for ECMO development, the patient develops an array of neurologic complications, up to 34% having acute neurologic impairment or discharge that improves over time. Lastly, in terms of immunosuppression, over time, we were able to achieve similar survive in patients with lung transplantation, heart transplantation, as well as immune-mediated disorder-resilient ECMO. Also it's worth mentioning that tycholimus pharmacokinetics parameters are not altered, but ECMO, which is really important as rejection is very scary and we do not want that to happen. There are three ways to deploy ECMO in liver transplantation. We can do elective ECMO, ECMO as bridge for transplantation, as well as emergent ECMO for acute organ failure. Now I would like to take a few minutes to present some data from Stanford and how we use ECMO in this patient population. First case, we use ECMO as a preventative measure in a six-month-old baby with biliary atresia, FECASI, with severe pulmonary, hepatopulmonary syndrome. She was calibrated for VBA ECMO during induction as she developed severe hypoxia. Prior to all liver transplantation, we have a multidisciplinary huddle to discuss the case, and for this patient, it was decided to preventively initiate ECMO if hypoxia occurs. She was maintained on ECMO for four days, extubated shortly after, and she is alive, having some global development delay, especially fine motor that is improving. We use VA ECMO as bridge for transplantation in a four-month-old with unknown acute liver failure that developed grafnum function after his transplant. He was transplanted with a split liver 33 hours after ECMO convalescent and 60 hours after his first transplant. His ECMO run during transplantation was relatively smooth and required only a circuit change, giving his massive blood transfusion. He received approximately three times his blood volume. This was possible as all the details and scenarios were discussed during the preoperative huddle. He is alive, having great function, and only mild receptive language disorder. We had three patients for which we used emergent ECMO for acute organ failure with mixed results. First patient was convalescent for hypoxic cardiac arrest. His run was 25 days, and it was the longest for us. And he had major complications that require daily procedure at the bedside, including portal vein thrombectomy, bilateral thoracotomies, wound changes, bronchoscopy. He also was discharged home three months after his transplant with intact neurologic exam, and he has normal graft function six years out. Second patient was transplanted with a very critical state, secondary to severe sepsis, and continued to be very unstable after his surgery. He was on EVA ECMO for 20 days, and unfortunately, during his ECMO run, he had an air emergency that led to hypoxia and progressive severe neurologic injury. He is alive and had normal graft function eight years after his transplant. The last patient was placed on ECMO for multi-organ failure, secondary to acute pancreatitis and sepsis that developed after his second transplant. Unfortunately, he did not survive, and care was withdrawn by family. In addition, we had three ECMO runs for cardiac indication for combined liver, heart-liver transplants. The survival rate for Stanford for isolated liver transplant is about 70%, and early and anticipated ECMO was the way to go and led to shorter runs and less complications. So for the last part of the talk, I will focus on how can we improve outcomes for ECMO used for pediatric patients with end-stage liver disease. Both from experience and available literature, I believe that anticipation is the most important factor. Adequate case selection and preparation for ECMO initiation is the first step in promoting good outcomes. We need clear standards for initiation and management. To do so, we need to have deep understanding of ECMO and transplant physiology. While on ECMO, appropriate neurologic monitoring and bleeding assessments are extremely important, those are major complications. One should consider no anticoagulation if there is risk for bleeding. An important aspect of PQ is to assure rehabilitation with PT or T, initiate early and perform daily, having counseling, nutrition, and, of course, PQ up for all those patients. QI work is part of both transplant and ECMO programs and are required for accreditation. Each program should have well-designed and implemented projects that will focus on preoperative care of the liver transplant patients. Development of transplant ICU will create an environment that will promote advancing the care of this fragile patient population by developing more protocols, promoting education, and specialized pediatric nursing care with streamlined patient care. I cannot stress enough how important open and direct communication with all services is here. In conclusion, ECMO can save lives, and it can be done safely. It should be done case by case and initiated for reversible causes of hypoxia and cardiac failure. Early ECMO, preventive ECMO to achieve better outcomes is desired, and each program should have standard protocols for initiation and management that will focus on preventing complication. Hospitals must have all systems and structure to support ECMO deployment in difficult cases. We need to share our knowledge, and we should not forgive that the fact that we can do it, and unnecessary mandatory should not do it. And I would like to thank and also acknowledge all my teams that I couldn't do the work without. Thank you. Great talk, Mihaela. So we're going to switch gears now. From ECMO, we're going to go to kidneys and the interaction with the liver. So AKI in liver disease has always been a matter of debate, and we discuss a lot what is what in acute kidney injury and liver disease. So I thought we're going to talk about the short and long-term outcomes in children who have got AKI. Nothing to disclose. So in the next 20 minutes, we're going to talk about AKI in different kinds of liver failure, where it's acute, it's acute and chronic, it's cirrhotics, and very importantly, post-transplant group. We're going to define AKI, AKD, and CKD in this cohort of patients, and we talk about outcomes. We're going to talk about short-term outcomes, which might include your mortality, which is within ICU at three months, six months, and a year, and we're going to talk about the long-term effects, which is development of chronic kidney disease, end-stage kidney disease, or mortality thereafter, and of course, dialysis dependence. And in the last part of my talk, we'll talk about outcomes in a special group of AKI in liver disease, which is hepatorenal syndrome. But we have to remember that most data, I know most of us who are sitting in this room are pediatricians, but most data has been, unfortunately, extrapolated from adults, and children do not, unfortunately, have this data. This is just to show you different kinds of liver, which we see in pediatric liver intensive care, an architecturally normal liver. You have a completely destroyed architecture here of a cirrhotic liver, and you've got a regenerating liver, as Mihaila mentioned in her talk, liver can regenerate. So these three livers have got a very different pathophysiology, the clinical trajectory, and the long-term and short-term outcome when AKI develops in these patients. So AKI in liver disease does not mean AKI in liver disease as a whole. AKI in acute liver failure, different from AKI in ACLF, is different from AKI post-transplantation. This is the largest study which was published in New England Journal of Medicine. This is the AKI called the AWARE study. We had about 6,000, 7,000 patients, and what we found was that the incidence of any stage AKI was about 26%. When we had a look at the severe AKI, we found it was 11%, which means one in 10 children admitted to PICUs had severe AKI defined as stage 2, stage 3. But if you look at the outcomes here, as the stage of AKI becomes more, your survival decreases. So higher the stage of AKI, lower was the survival. So then we thought, let's have a deep dive into those patients within the AWARE subset which have got a liver disease. So that's what we did. We looked at these patients with the light blue bars. This is our liver disease. The dark blue bars are the rest of the AWARE cohort, and you would see here whether it is stage 1 AKI or stage 2 or stage 3 AKI, you see the difference. What is the incidence? 59%, i.e. 3 out of 5 children with liver disease will have AKI as compared to about 26.4% in the AWARE cohort. Look at the mortality, 5.5% in liver disease patients, 3.4% in the rest of the AWARE cohort. So starting off with AKI in acute liver failure, we at King's have just done this study looking at the prevalence, risk factors, and outcomes of severe AKI in patients with acute. I'm underlining the word acute because this does not include cirrhotics, it doesn't include patients with ACLF, and our aim was to look at different things, including outcome. But very importantly, in addition to looking at the mortality, in addition to looking at the length of stay, we wanted to look at how many children with acute liver failure with the same severity of illness and the same supportive measures survive without transplantation when they have an AKI versus when they do not have an AKI. Basically, what we want to know is what is the effect of AKI on the spontaneous regeneration of the liver. And this is what we found, 125 patients with acute liver failure. We excluded 31 because they were started on CRRT exclusively with four liver causes and you can see any stage AKI 56.4%. Severe AKI 39.4%. So two out of five children with acute liver failure end up having severe AKI. But what's very important is look at day 7 here. You have reached day 7 and you have still got a considerable number of children who have got AKI. Now whether this AKI then becomes AKD or becomes CKD we do not know. If you look at the various outcomes, you see here the number of children with no AKI 17% required transplant whereas with AKI 35% required transplant. And very interestingly if you look at the children who survived with their own livers, 71% with no AKI did not basically, they did not survive and others did. So basically what we concluded was that children who have got severe AKI tend to have less regeneration of the liver and therefore tend to have more transplantation. Again if you look at the mortality here, severe AKI versus non-severe AKI, you see there is a difference between mortality after controlling for confounders. So our conclusion was the significant number of children with acute liver failure develop AKI. AKI in ALF is associated with increased length of stay, prolonged duration and worst outcomes and AKI does seem to affect the spontaneous regeneration of the liver. Going to the second category of patient, the cirrhotics. This is a diagram which shows you a natural progression of patient with liver disease. You develop a liver disease, you go into compensated cirrhosis, you go into decompensated cirrhosis, you develop hypertension, you develop all the complications. If you have compensated cirrhosis, the risk of AKI is almost the same as general population. But if you develop portal hypertension complications, your risk of AKI increases exponentially. This is just to give you a concept of ACLF, what acute and chronic liver failure is. You have a child or an adult who is sitting with a chronic liver disease, gets hit by a precipitating event, rapid progression, multi-organ failure and within this multi-organ failure, you have AKI as a core component. And if you treat them well, there is a chance of potential reversibility. But here, higher the stage of AKI within ACLF, higher is the mortality, lower is the risk this patient actually will get a successful transplantation. Now how do you define AKI in cirrhosis? The International Club of Ophthalmologists took almost the same definition as the KD Go guys said and said we're going to look at serum creatinine rise within 48 hours and we are going to look at is it two times, is it three times or you're requiring renal replacement therapy. But the difference is stage 1 is also called stage 1A and stage 1B. If you satisfy the criteria of stage 1, but your serum creatinine does not go up beyond 1.5 mg per deciliter, you have 1A. If it goes beyond 1.5 mg per deciliter, it is 1B. And if you look at the staging correlation, how does the staging correlate to mortality, you will see here higher the stage and lower is your survival. 88% with no AKI and 31% if you've got stage 3 AKI. So one third the survival if you have AKI. Now what is the prevalence? You can see here that this is a cohort from Italy, 53% of patients with cirrhosis developed AKI. And if you look at the mortality, it is about four times more in patients who developed AKI who had cirrhosis. What are the causes of AKI? We always, I know when liver and kidney, we start talking about the first thing comes to mind, is it hepatorenal? Actually, the most common cause is hyperperfusion, hypovolemia, which could be due to gastrointestinal losses because of renal losses. Then you talk about intra-abdominal hypertension because of massive ascites. And then you talk about hepatorenal syndrome. And you can see here in this cohort of patients, about half of the patients had hypovolemia as the cause of AKI, and about one third had hepatorenal syndrome. Now why is it important for us to say, is it hypovolemia, is it acute tubular injury, or is it hepatorenal? Look at the survival here. If you've got a hypovolemia, your survival is 76%. If you have got hepatorenal, your survival drops down to 37%. And if you look at the odds ratio of dying here, it is about seven. You've got hepatorenal syndrome, the risk of death is about seven times more. And if you have a progression of AKI, you don't have AKI before, but then you progress during your stay in the ICU, you see that the mortality exponentially gets more. What are the pediatric specificities in these patients? The first thing is to remember that biliary atresia is the most common cause of chronic liver disease in children. Therefore, we cannot directly extrapolate data from adults. They've got varying chronic or alcoholic liver disease and other bits. And the reported prevalence of pediatric HRS AKI is less than 10%, but I firmly believe that this is under representation because look at any guidelines. Unfortunately, pediatric definition for HRS is not there. This is a study which we did in collaboration with Paris and the Netherlands, where we looked at 130 pediatric cirrhotic patients and looked at the correlation between different stages of AKI and the fact that they were having cirrhosis. You can see here, as your stage of AKI becomes more, even children have got a lower survival rate as compared to patients who did not have AKI. Now we talked about pre-renal component and we said hypovolemia, you give fluids, patients will benefit. We have hepatorenal syndrome, it again is a functional component, but you give fluid, things will not get better. And then of course you've got acute tubular injury. It looks like the diagram, they're all three distinct entities, but in reality, there is an overlap. There is an overlap between the functional component and the structural component. And if you look at HRS patients, you have casts in the urine as well, which actually shows you it has got a structural component. Now, is there a correlation between baseline GFR in deciding the prognosis of patients on the wait list? This was a study by Lim et al, and you can see here, as your GFR is going below 60, this is 60 you can see here, this is 40, this is 20, the risk of mortality is 2.5 times more when you have a GFR less than 20. So lower the GFR, higher is your mortality in these patients. But once the GFR crosses 60, you can see almost the mortality remains the same. We know MELD predicts the mortality, and in this particular study, they looked at various models. They looked at MELD, they looked at MELD, and plus they added your GFR. In the third model, they took out creatinine, and in the fourth model, they added in sodium. You can see here, when you have got GFR added, things get better. But when you remove the creatinine, things remain the same. Probably measured GFR, if you can, is better in predicting survival as compared to serum creatinine. Now what are the outcomes? The outcomes in AKI is known, but outcomes of AKI in patients who have got liver disease probably is not known. Look at the three different, four different categories. You've got all-cause mortality, early, late. You've got cardiovascular. You've got hospitalizations, which can take place either because of AKI or other causes, and very importantly, chronic kidney disease and end-stage kidney disease. This cartoon shows you how the AKI progresses and what different terminologies we use. The red is a normal kidney. The orange is acute kidney disease, and your green one is a chronic kidney disease. You get hit by a precipitating event. You get AKI. This AKI could be stage 1, 2, 3. If your AKI goes away in 48 hours, as you can see here, it's called transient AKI. If it persists from day 2 to day 7, it's called persisted AKI. And if it progresses beyond that, now that's a nebulous area. If it persists up to 90 days, we have coined a term called AKD, acute kidney disease. But if it persists more than 90 days, we call it chronic kidney disease. So you've got AKI, which is transient and persistent. You've got AKD. You've got CKD. What are the definitions? These are definitions. They're all dependent on the duration of renal dysfunction, less than 48 hours, 48 hours to 7 days, 7 days to 90 days, and beyond. That's how you define AKI, AKD, and CKD. Now people might say transient AKI. What is the effect of it? If you look at patients with transient AKI versus without AKI, there's a difference of 20% in survival. So any AKI which takes place does matter. Now AKD, this is a paper published by Patidar, and they looked at the incidence of AKD, and they also looked at 180 days, what happens to these patients with AKD. And you can see here that out of the patients, 6,000 patients they looked at, there were patients, one third of them who developed AKD. And the incidence of mortality both at 90 days and 180 days was tremendously high as compared to patients who did not have AKD. And if you look at the de novo CKD, how many patients with AKD go on to develop CKD, the hazard ratio was much higher in those patients who did not progress to AKD and just had persistent AKI. Now what about the probability of survival? This is a study from Italy again looking at AKD patients, so patients who had AKI beyond seven days. They looked at five year survival. You can see here 88.8% in patients who did not have AKD the survival, and the five year survival in those who had AKD was 34%. So almost a decrement in survival by about threefold if you have AKD. Does the pattern and trajectory of kidney injury matter? Does it matter you have AKI? Does it matter you have AKD? And you can see here very clearly that if you have mild AKI with full improvement, your hazard ratio of dying is only 1.1. But if you have severe AKI with minimal improvement, the risk of dying is threefold. And if you have got AKI on CKD, this is the cohort which will have the worst prognosis as compared to patients who have either AKI or got either CKD. So definitely the pattern and trajectory of AKI matters in prognosticating patients with cirrhosis. Now when kidneys recover, they recover with adaptive repair. When they do not recover, they have maladaptive repair. And these are some of the risk factors. And we know there's inflammation in cirrhosis. We know it's age which matters. And the kidney condition, HRS matters. If you've got HRS, the risk of maladaptation increases. And there are many risk factors which progress from AKI to end-stage kidney disease. And very importantly, if you develop AKI within the hospital setting and the severity of AKI, these are two determining factors whether this kidney disease is going to progress. Now this is a study from the Spanish group looking at the cirrhotic patients, 140 patients looked at AKI patients. And of the AKI patients who survived, about one-fourth of them developed cirrhosis. So if you look at this particular cartoon, you see patients who develop CKD here. They already have the black dots here. They already have a low EGFR as compared to patients who did not develop CKD. And if you look at their presentation in the next 12 months, that EGFR never recovers in patients with CKD. Whereas in patients who do not develop CKD, the EGFR nicely recovers. So they concluded one-fourth of patients who present with AKI will develop CKD. And you will never get a recovery to baseline of the EGFR in those patients. What about pediatrics? Do we know ACLF in pediatrics? This is a study from All India Institute of Medical Sciences in Delhi, where they looked at about 600 patients who had chronic liver disease. They took only ACLF patients, which were about 84, and about, I would say, 22.6% of the patients developed AKI in this cohort of patients. They clubbed together death and liver transplantation within three months of developing AKI as a poor prognostic factor. One-third of these patients had hepatorenal syndrome. One-third had sepsis. And the remaining had nephrotoxic drugs. And you can see here that the hazard ratio for death or liver transplantation in children who have ACLF with AKI is about 7.7%. So if you are a child, you have AKI, you have ACLF, the risk of dying is about 7.7%. And this is King's College Hospital data for ACLF exclusively secondary to biliary atresia. And you will find here that the survival of ACLF is much lower than the patients who actually got a transplantation just with chronic decompensation, not with ACLF, but independently hepatorenal syndrome. And the fact that they received CRRT was a poor prognostic factor. The last group of patients, the post-transplantation group, we know a lot of things take place pre-transplantation. Pre-transplant AKI leads to post-transplantation AKI. But a lot of things happen intraoperatively. Hemodynamic changes take place. And you can see here within the first six months, the GFR decreases by 30% to 50% of an average. And by six months, kind of stability goes. And post-transplantation, you expose these patients to nephrotoxic antibiotics, nephrotoxic calcineurin inhibitors. And all these things lead to post-transplantation AKI. Now what are the risk factors? You can see here that if you compare this in these 1,400 patients where they looked at mortality, higher the stage of AKI, lower was the survival in these patients. And if you received renal replacement therapy before liver transplantation or develop AKI post-transplantation, that was a very high risk factor for mortality. So AKI post-transplantation can have long-lasting consequences. What is the incidence of CKD? About 60%. You can see here is a cumulative incidence of death, GFR less than 60, and the graft loss. And you can see as it's about six years, your risk of CKD is about 60%. So after five years of transplantation, if you've got an AKI, your risk of developing CKD is about 60%. So monitor these patients for kidney health and for liver health. Does the severity of GFR matter? And you can see here as your GFR is getting lower and lower, your chance of survival is getting lower as well. And your GFR more than 60, your survival is 91%. Your GFR less than 30, it's about 30%. So meta-analysis shows seven-fold increase in mortality in the wait list if there's renal dysfunction. Similarly, creatinine clearance, very important. Lower your creatinine clearance, higher is the mortality, both within 30 days and at two years. So creatinine clearance, EGFR matters. So lower the EGFR or the creatinine clearance, lower will be the survival, both short and long-term. What about hepatorenal syndrome? We know that it causes mortality. It does have an effect on mortality. That's short-term. What about the long-term impact of patients who responded to terliprasine versus who did not respond to terliprasine? You can see here that the risk of developing chronic kidney disease in patients who respond to terliprasine, the survival is much better as compared to patients who did not respond to terliprasine. So response to terliprasine, a key factor in determining the short-term outcomes as well as the long-term development of CKD. So response to terliprasine, very important. Now, do we have a tool to measure properly the GFR in patients with liver disease because there are lots of confounding factors here. So in this particular patient, they looked at GFR measurement and something known as GRAIL, which is Global Refiltration Rate in Patients with Liver Disease. And you can see here, sorry, I'm going to go back. You can see here when you compare the red one is the GRAIL, the comparison of GRAIL measured versus estimated, the difference is much lower. So probably GRAIL is the way forward. I'm going to quickly skip. So this is my last slide. In summary, we say kidney dysfunction matters in liver disease. You need to understand which liver disease are you dealing with. There's a spectrum, AKI, AKD, CKD, and there is a pre-transplantation kidney dysfunction which leads to post-transplantation problems as well. And what you need to do is to have a holistic picture in mind. Patient susceptibilities, kidney health, liver health, and make sure you have a holistic assessment of the patient when they come to you in your nephrology or hepatology follow-up post-transplantation. With this, I end my talk. Thank you. So our last talk for this session is by Elizabeth. She's from North Carolina. She's a pharmacist there. She's going to talk to us about liver aspects of sedation and analgesia. Thank you for the invitation to speak today. I really have no disclosures in relation to this presentation, and we're going to be talking about special considerations in sedation and analgesia in children with liver failure. The first thing I'm going to talk about, which I know is painful for most people, is pharmacokinetic changes that occur when you get liver failure. But I think they're very important, because if you understand what occurs in the liver, you can understand whether certain drugs can be used in acute liver failure versus cirrhosis, because the changes are different. The other thing to talk about briefly, which I'm just going to mention, for other drugs that other than sedation and analgesia, because most of these drugs don't have them, but a lot of the antibiotics we use in other drugs, will have a child Pugh classification A, B, or C. And many of the references we use, pediatric lexicon, will tell you, depending on your child Pugh score, how you should dose that medication in liver failure. But we're not going to talk about that specifically today. There is a MELD score, but most drugs are actually adjusted based on a child Pugh and not on a MELD. But you could, there are a few drugs that have MELD recommendations. So, if you think about clearance through the liver, a lot of that is dependent on our hepatic blood flow. We do know that once we get cirrhosis and we start to get shunts, you can get decreased liver blood flow because of that intra-orhepatic shunting. Many of our drugs are bound to plasma proteins, and we know in acute liver, chronic liver failure, we get decreased amount of plasma proteins. And then we have intrinsic clearance, which is our cytochrome P450 system. We have drugs that fall into categories. They can either be high extraction, poor extraction, or intermediate extraction. And a good way I think of high extraction drugs, if I look at my PO dose, and it's three times the IV dose, that's probably a drug that's high extracted, if you think big picture. Otherwise, you know, you can find these numbers in the literature, but the drugs you are most concerned about in hepatic failure would be those ones that your IV and PO difference is significant. So those orally administered drugs normally pass through the portal vein and undergo that substantial metabolism prior to reaching the systemic circulation. So if we have shunts, now we're going to get a significant increase in the serum level of those medications, and those have to be considered. You're most likely to have shunts in cirrhosis, so if you just have acute liver failure, you actually may have no difference in the absorption of our oral medications, because we haven't had time for changes to occur yet. So you're going to be most concerned in those patients who go on to have cirrhosis. So that's, you just worry about bioavailability of those oral medications. We do have decreased production of drug-binding proteins, typos, sorry, I knew I'd miss one. And about 85% of our drugs bind to albumin, and about 15% bind to alpha-1 acid glycoprotein, mostly our basic drugs, propranolol, lidocaine, those types of things. But both of those are reduced in liver failure. So if you think about our drugs that we give, it's the free drug that gives us our clinical efficacy and has the ability to go into our tissues. So if you have less binding proteins for that same dose of medication, more is going to get into your tissues and the receptor, and you may have an amplified effect of medications, because only those unbound drugs can enter and leave the tissue compartments. It also may lead to faster clearance of those drugs if they're renally eliminated, because that drug can get into the renal system, if we don't yet have hepato-renal syndrome that we've just learned about. With progression of liver disease, changes in body composition, such as ascites, so extracellular fluids and edema, drugs that are now free drug and hydrophilic, can go into those tissues. And then decreased muscle mass can alter your volume and distribution. So if we think of how drugs are metabolized through the liver, this is the very basic Farrington version. We have phase one and phase two reactions. Phase one reactions are usually making something smaller. So we have oxidoreductive processes. That's our cytochrome P450. That's more effective by liver disease. And if you think about all that shunting we get, then we get hypoxia in the liver, and that kind of dings off our cytochrome P450 system. Cytochrome 2C19 seems to go first. And if we have moderate to severe liver disease, you're going to get decreased activity of all. But that first one to go is 2C19. And with severe liver disease, you'll get decrease in all of the cytochrome systems. These two reactions are our conjugating enzymes. So we're adding something to a drug to make it bigger. Both of these have the same purpose, to make something more hydrophilic so it can be eliminated by the kidney. But you still maintain glucuronidation up until the very end. And you have up-regulation of what we call UGT activity in the hepatocytes. And that actually preserves that glucuronidation pathway. So in mild to moderate liver disease, you're probably not going to lose that ability. But when we get to severe or cirrhosis, we're going to lose both. So then we have to think about hepatorenal syndrome, because many of our drugs, even if they're not hepatically eliminated, may be renally eliminated. So then you're going to have to consider altering our intervals in drugs that have renal disease. But we also have to consider we have indirect methods of measuring creatinine clearance. If we're using a serum creatinine, we may not have as much muscle mass. And so our serum creatinine is not going to rise as quickly. And so under production of creatinine, when you lose that muscle mass, we'll make what we use to estimate creatinine clearance inaccurate. And we will overestimate their renal function. So remember, the liver also produces creatine, which is our substrate for creatinine production. So muscle mass and the liver both assist in producing creatinine. And that's most likely to occur in cirrhotic patients. So chronic liver disease without significant fibrosis, drug pharmacokinetics are going to be somewhat unchanged and modified only to a small extent. So I'm kind of trying to display some of the myths that we talk about with hepatic failure. My attendings say, oh, we've got elevated liver function tests. I can't give the patient acetaminophen. Well, if you think about acetaminophen, remember, it's metabolized to glucuronide and sulfate activated. But then it's bound to that glutathione. So remember, that's the one enzyme we said is preserved in early hepatic failure. But patients who are known to have that low glutathione, alcoholics, they upregulate their alcohol dehydrogenase, which also increases that CYP2E1. So they're more likely to make more of that hepatic metabolite that's toxic just because that they have upregulated 2E1. Patients with chronic hepatitis C or non-alcoholic cirrhosis may also have less amount of glutathione. But what has been demonstrated with acetaminophen and liver disease? Most of these are adult studies, but I don't know how some of these got through IRBs. It's kind of interesting. The first one is 20 patients with chronic renal hepatic disease, and they gave them what we consider the maximum dose of Tylenol or acetaminophen, 4 grams per day versus placebo. They were on that for 13 days, the max dose for 13 days. This is why I'm like, really? You got this through IRB? And then they had crossover. So those who were on placebo were crossed over to drug, and there was no clinically significant changes in the clinical features of what we monitor, our bilirubin, our ALKFOS, bile acids, creatinine, albumin, or prothalbin time. A similar study was done in patients with hepatitis C, which was one of the categories we said might have less glutathione, but 3 grams per day, so slightly lower dose, versus placebo for 7 days, and again, no changes in any of our hepatic markers. When alcoholics were evaluated looking at that same max 4 grams per day for 48 hours, there was no change. But when they looked at them for out to 10 days, that's alcoholics, remember we said are the most likely ones to see changes with acetaminophen. There was changes in their indices of liver function. So we do know that the elimination half-life of acetaminophen doubles, or it can increase by 50 to 100, slow by 50 to 100 percent. So if we think of our adult and pediatric patients with various forms of hepatic failure, we should use a normal dose because you have to give enough drug to get effect, but you may prolong that interval knowing that initially, so I might write Q12 initially if I'm having a patient with significant fever or pain, but if at 12 hours or 8 hours they're still having issues, I can probably, knowing that Q6 for 13 days was fine, feel more comfortable using that medication. But due to the prolonged half-life, probably that initial order if you've got an adult-sized patient would be 650 P or IV every 8. I'm trying to max at that 2 gram per day that we've talked about. And then pediatric patients, a normal 15 per kilo dose, and maybe starting Q8 initially. But we would avoid it in our teenagers if you had them that were actually chronic alcoholics. I have seen, unfortunately, plenty of those, which is kind of sad. So you say, well, should I use acetaminophen or what about using an NSAID or a non-steroidal anti-inflammatory drug? Remember patients with severe liver disease, especially those with cirrhosis and ascites, have unstable renal hemodynamics. So even with a normal GFR and renal blood flow, renal blood flow is sensitive to any modest reduction in plasma volume. So if you think of how NSAIDs work, the impairment in renal function and any effect that we give effect on vasoconstrictive hormones is normally countermanded by the increased production of vasodilatory renal prostaglandins. But our NSAIDs decrease the ability of the kidney to produce vasodilatory renal prostaglandins. So therefore, they're going to decrease GFR and they're going to worsen our renal blood flow, which may actually worsen or enhance the development of acute renal failure. Those that are going to be most sensitive to the renal effects of NSAIDs are clearly those with ascites and sodium retention, where we already have maybe some decreased delivery of fluid to the kidney. And there is evidence that when NSAIDs have been given, there is decreased diuresis after administration of furosemide. So you're giving a drug that's actually you think you're giving it because it's safer than Tylenol, but it's actually more risky to the patient with hepatic dysfunction. And then we have to remember that once we get hepatic dysfunction, we get coagulation disorders. So if you think of platelets, NSAIDs decrease the stickiness of the platelet by changing the production of thromboxane A2. So the NSAIDs actually decrease that. And so you're going to increase the risk of having bleeding. So in conclusion, when just comparing NSAIDs to acetaminophen, when administering PRN pain meds or something for fever control, acetaminophen is much safer than using an NSAID. Naproxen is the only NSAID that we know reduction in metabolism is significant. Ibuprofen, you can give a normal dose. You could probably use Catorlac, but overall, you should always use acetaminophen over an NSAID in a patient to try to preserve their renal function. Opioids, obviously, we try to avoid in renal failure if we're thinking about pain because they can worsen our hepatic encephalopathy. They're going to alter. Now you don't know if it's because I gave them an opioid, which now can actually cross the blood-brain barrier more quickly because, remember, less protein binding, more free drugs, so it gets across the blood-brain barrier more rapidly. And so that's always a concern. But if we choose to use an opioid or we need one because we've had a surgery, we've done a procedure, we need something with more strength, there's two drugs we should just never use in hepatic failure. We probably shouldn't be using them anyway. But remember, codeine is a prodrug. And so if we don't even know, both of these drugs are CYP2D6, which is the one enzyme that could have a low expression of the enzyme, a high expression of the enzyme or medium. And so some people who are completely healthy get no pain relief from codeine or tramadol because they're a low expressor of the enzyme that takes that prodrug to active drug. But when we have liver failure, you may have worse production of the enzyme and actually not even make these prodrugs into active drug and get pain relief. So you probably should just not think of them as an active agent. This is just a short version because we're going to go into sedation kind of infusions in a minute, which is a little bit more important, I think, in the interest of our time. But oxytocodone, you need to decrease the dose by half and extend the interval. Morphine IV, again, the same dose for a clinical effect, but extend interval. But remember, we said with that shunting, you're going to get decreased the ability of that first pass through the liver. So we know that you get a fourfold increase with both morphine and hydromorphone when given orally. So you have to consider that if you're giving an oral dose to cut the dose by a quarter. And then PO or IV morphine, again, where you're going to accumulate the M3 and the N6 metabolites. So a single dose, completely fine. We're giving it for a procedure. We're getting pain relief. But when we're trying to give chronic medications, remember, morphine has active metabolites that can accumulate, especially in hepatorenal syndrome. And the same thing would be true of hydromorphone that has an active hydromorphone 3-glucuronide. These are all really eliminated metabolites that have longer half-lives than morphine itself or hydromorphone. So these are just big picture concepts to remember when you're using these medications. But IV morphine, hydromorphone are going to be completely safe for a single procedure. They're just going to last longer than what you're used to based on metabolism difference. Now remember, fentanyl is about 85% bound to albumin. So we're going to have more free drug, and it's going to cross the blood-brain barrier more readily. So you might have a little bit more CNS effect. It's a high-extraction medication. So clearance is going to be affected by those changes in hepatic blood flow. But pharmacokinetics have been demonstrated to be unaltered in patients with just biopsy-confirmed cirrhosis after just a single dose. But nobody has really done good studies on this drug, looking at it when we give it with a continuous infusion. So when we're using it for regular intermittent dosing or a loading dose prior to initiating an infusion, you want to give the same regular dose that you would use to. But knowing that it's a high-extraction medication and we may not be clearing drug as fast, probably let's say if I normally would start my fentanyl infusion at a mic per kilo per hour, I'm probably going to start that patient at a half a mic per kilo per hour. The advantage of fentanyl, which is probably why it's the preferred opioid infusion for patients in hepatic failure, is it has no active metabolites. So versus morphine and hydromorphone that you have to worry about the accumulation of those active metabolites, that would not be a concern when we're using fentanyl. Rumi fentanyl, at most institutions, is restricted to the OR because of cost. But this is one time I have to say, okay, this may be our best agent. It's hydrolyzed by blood and tissue esterases. So it does not need the liver or the kidney to be eliminated. So no adjustments are needed for hepatic or renal dysfunction. So in a patient with significant hepatorenal syndrome, actually using Rumi fentanyl as a continuous infusion is probably our safest option. But if you're in a large hepatic liver center, that's something to discuss with your pharmacist or pharmacy and therapeutics because this is where cost might benefit your patient because if you stop another infusion and it takes a couple days for things to go away, that might leave them intubated longer and that's a cost benefit here. So although more expensive than fentanyl, it's more predictable clearance make it really the optimal choice for continuous infusion opioids. Benzodiazepines, as we all know, should be avoided for sedation because they have our highest association with delirium. But we do know that if we have to use a benzo for some reason, midazolam has active metabolites that are really eliminated. They actually can accumulate and they actually bind to the receptor differently and can make you hyper excitable. So then it makes you more difficult to sedate that patient. So midazolam should be, it's fine for a single dose for a procedure but should be avoided for long-term use. Lorazepam is diluted in propylene glycol that could accumulate in hepatic renal dysfunction, make the patient more acidotic, but it does have no active metabolites so if you needed to give something intermittently, probably more safe. And diazepam also has active metabolites and it does have some excipients in the product. So obviously any administered orally will have that increased bioavailability due to the decreased first pass. So if you were using, so if I had to give one of these, intranasal would be a preferred versus an oral sedation if I was doing something for a procedure because I'm not going to worry about that first pass. So if you need it for procedures, lorazepam has no active metabolites but probably any of the three for a once dose would be fine for a procedure. For status epilepticus, we would follow the drug of choice as lorazepam, followed by diazepam, and then we would give midazolam either intranasally or IM, but avoiding them for chronic sedation. Liver failure actually has been demonstrated in one, a couple papers not to be affected by liver failure. And there's one paper that looked at using continuous infusion and pripofol in patients with fulminant hepatic failure who had been deteriorating for cerebral edema. So they are at the very end stage of disease. And they did 10 mics per kilo per minute, an adult study here, and advanced until the dose was adequate to control ICP less than 15, and their medium dose was about 50 mics per kilo per minute, which is kind of what we think of as that higher adult dose, but a range of 31 to 89. No systemic complications, hypotension, acid-base abnormalities, or evidence of pripofol infusion syndrome. However, they do not tell us in this case series how many days patients receive pripofol. And remember, pripofol is in a lipid base, and triglycerides have to be eliminated through the liver. So if you choose to use pripofol, we need to make sure we're measuring triglycerides. But this really is a drug that needs probably more close research because it has an advantage there. Pripofol could be attractive also because if you're trying to assess your patient's neurostatus, you could actually stop it. It should have a fairly quick on and off, maybe not quite as quick as we're used to, but pretty close. In this one case report, they were waking patients up daily to check their ICPs. So in a pediatric patient where you might start a small baby at 100, you might start them at that 50 mic per kilo per minute, and advance is tolerated until you get the benefit that you need. And remember, adolescents, I might start them at that adult dose just because we don't know exactly what our target is going to be in this agent. But so far, appears to be safe and may be an attractive option for us. So in conclusion, we can say that acetaminophen is preferable. Oh, dexmedetomidine, I could not find any published data, but based on its metabolism, again, might be a very safe agent for us to use. Again, most of us do not load patients with dexmedetomidine. We wait for it to get to steady state. This might be one place where you might load the patient, infusing it over 10 to 20 minutes to make sure we don't get that bradycardia, only because there may be some slight decrease in its metabolism, which means it's going to take us longer to get to steady state. So you don't want somebody to keep bumping up the drip because we're not sedated. So a bolus dose might be appropriate here to make sure that we get our serum levels where we want them, and then start at our normal dose of 0.2 to 0.4, and titrate every six or so to make sure we're hitting target. If you need to give another loading dose if the sedation wears off, and over that same time period, that's probably safer. And that's just hypothetical me giving you my opinion without good published evidence. Acetaminophen is going to be preferred for mild to moderate pain in patients with hepatic failure. We want to avoid using NSAIDs because they can make our renal function worse. Fentanyl, rimi-fentanyl are preferred opioids for continuous infusion. Piperone may be preferred over morphine for PRN use due to the lack of active metabolites. Prepavol has predictable clearance in hepatic failure, but really seven cases is kind of small. So we just really need more evidence. And you really need to consider the phase of your patient. Are they acute hepatic failure? Are they in cirrhosis? Because in that acute phase, we probably haven't had a chance to see changes yet. Cirrhosis, we're going to see all the changes in first pass. Because you have to think cirrhosis versus non-cirrhosis. And I think I talked fast enough, and I have one minute to spare.

ECMO

liver disease

acute liver failure

bleeding risks

sepsis

pediatric ECMO

heart-lung machine

case studies

patient outcomes