It's great to see so many friendly faces in the crowd. My name is Raj Basu. I am at Children's Health Care of Atlanta. And we are very excited to have this highly motivational session put together for you. We have three excellent speakers who probably don't need any introduction amongst our friendly crowd, but I'm going to do it anyway. I do appreciate that everyone is here on Monday afternoon when you could be doing other things. We're going to go in successive order. The speakers today, we've taken great effort not to have too much overlap, but naturally because of the topics, there is a little bit. We're going to try to save a couple of minutes after each one for questions. But if there isn't time, then we'll definitely have some time afterwards. So today's session is about emerging extracorporeal therapies in the PICU. And our three speakers come from around the country, and we're very excited to have them. Our first one is my dear friend, Dr. Stuart Goldstein, who comes from Cincinnati. He's the director of the Center for Acute Care and Nephrology. He's a professor of pediatrics and director of the Heart Institute as well. He's going to be talking about renal replacement therapy on ECMO. Dr. Goldstein? Thanks Dr. Basu. Thanks Raj. And it's really an honor to be here and share the stage with you and with Dr. Eric Kahn and Dr. Carcillo. And I hear on Twitter right now, hashtag proud renal nerd is trending. So I'm one, Dr. Natalia Stansky put that out there. So I think we're a bunch of renal nerds, even if I'm the only nephrologist of the group here. So I'm going to talk a little bit about renal replacement therapy in ECMO. Here are my disclosures, and none of them really have much to do with this topic, although I do consult for every CRT machine manufacturer that has a device in the United States. So you may take that for what you will, but I'm not going to mention their names. So we have to think about ECMO and CRT because you have very high flow on ECMO, and it becomes a bit of a challenge to do CRT in ECMO. We have to see positive access pressures when you're post-pump, and so we need to think about creating a negative pressure if the machine is not built to accept positive pressure. So it always amazes me we have these $25,000 to $50,000 machines, and we're sitting around with a $1.50 C-clamp to try to create a negative pressure, but that's why there is a C-clamp on the back of every one of our machines. Circuit flow affects ECMO flow, and we have to think about that, but I have to say there's a little bit of the sky is falling there. And if you do not have a program where the ECMO specialist is comfortable with CRT, blamed dialysis is the number one ICD-9 code in our hospital at 666, and everything that's bad on an ECMO circuit gets attributed to CRT. It's amazing that apparently air bubbles the sides of helium balloons get into the ECMO circuit on CRT. I've never seen that, but the real consideration here is to help people feel comfortable, especially if they've never done it because this machine is doing things that people don't understand. Anticoagulation is actually not necessarily a big deal because of the anticoagulation of the ECMO circuit itself, but there are occasional times that I have to think about doing regional citrate anticoagulation. And if CRT introduced air into the ECMO line, what would be the danger to the patient? That is a big deal, and so you have to think about where you're putting your CRT circuit. This should not ever be something you decide at the bedside. We all have simulation labs. There are places that have simulation labs. There are people who are experts in the room, and we should know ahead of time where we're going to place a CRT machine. But I want to get back to something that is actually more near and dear to my heart, which is acute kidney injury, because the definitions that ELSO has used in the past for acute kidney injury are really quite outdated, and so we have to think about those in context. I want to spend some time talking about acute kidney injury because the NICU and in ECMO are kind of the last bastions of getting into up-to-date—no trademark there, sorry—getting up-to-date with AKI. So prior to now almost two decades ago, there were 30 definitions of AKI in the published literature, and if you can't describe or define a syndrome or disease, then there's no way you're going to prove its outcome. In addition, all AKI was created equal. There was no dimensionality. Patients either had acute renal failure or they didn't, and we don't do that with any other disease process. We grade IVH. We stage cancers, heart failure, and so we were really lagging behind. Then this paper, which really changed everything in the field of acute kidney injury from Glern Chertow, where they asked a simple question, what rise in serum creatinine in a hospitalized adult Brigham and Women's Hospital was independently associated with mortality? And if you look here, a 0.3 milligram per deciliter rise in serum creatinine was independently associated with mortality. Multivariable odds ratio of four, controlling for age, ethnicity, New York Heart Classification Score, diabetes, and COPD. The area under the curve was 0.84, better than all these exciting biomarkers that Raj and I are looking at. 0.3 milligram per deciliter rise. That's what we used to think was the error of the lab measurement itself. And I will tell you, still at Cincinnati Children's, there are people who say, well, that's just a change in a number. Still having to convince people that 0.3 rise is important. And as you go up, the increase in creatinine, the odds ratio increases even more. This association is stronger than any Framingham risk factor for mortality in adults. This also is demonstrated in children. Some work from the P-Rifle criteria, which Dr. Erikan was first author on that paper where we looked at a staged multidimensional classification system and demonstrated that a doubling of creatinine was independently associated with mortality. And there were now numerous studies that have validated that. Then the AKI criteria from Rifle and P-Rifle turned into the Akin definition. So it's one of the progeny of Rifle. And here's where we have our friend 0.3. And then the KDGO criteria, Kidney Disease Improving Global Outcomes criteria, came about in 2012 where this harmonized Rifle, P-Rifle, and Akin, where as of today, February 18, 2020, this is how we define acute kidney injury. There will be a new KDGO consensus conference likely in the next year where we will revisit this. But when you look at this definition and compare it to Akin and KDGO or P-Rifle, you can see, yes, the epidemiology and the association with mortality in the ICU, and this is work by Scott Sutherland out of Stanford, there is a graded response with each stage of acute kidney injury. And the length of stay both in the ICU and outside the ICU is also very evident. So even though there's a slight difference in epidemiology, what Scott and we argued in that paper was we should use KDGO because this allows us to compare data with our adult colleagues. It's uniform. And in fact, John Kellum and Dana Fuhrman have a very nice paper looking at young adult patients with acute kidney injury. So what about AKI, renal replacement therapy, and ECMO? Well, there are three main sources of data. Single center reviews, which is how we all start with where we get the details and we're able to control some variables. The ELSO registry, obviously incredibly important. But early on, ELSO defined acute kidney injury as getting to greater than 1.5 milligrams per deciliter in terms of serum creatinine, didn't really talk about change. And severe AKI was three milligram per deciliter. Especially when you're talking about children, we're not talking about rifle R, I, or F, we're talking about rifle Q or stage 12. This is severe, severe acute kidney injury. If I got into a fight with Raj, he'd win, so he would take out both my kidneys. In that fight, my creatinine would go up by two in 24 hours. So that's really a significant loss of kidney function. And then a number of our colleagues developed the KIDMO group, kidney interventions during membrane oxygenation. And they've started to give us data that are more relevant and contemporary. So what about this? Oh, sorry, I'm going backwards. So if you look at single center data and ELSO publications, anywhere from 10 to 71% of patients, depending on the definition, have AKI, a major organ illness that is not the heart or the lung. Pediatric data, 12 to 30%. So one in eight to one in three patients develop acute kidney injury in older data. And if you look here from ELSO, you can see that a creatinine rise to 1.5 to 3, only half of the patients survive, whereas three quarters of patients survived in the neonatal respiratory ECMO cohort. Gets worse to a creatinine of three. If dialysis is required, the survival is 40%. And if you get into neonatal cardiac, the association's worse. Now, the problem is that people look at these data and say, well, if I put a patient on dialysis, then their mortality is going to be worse. So I'm not going to do that, and I'm going to wait too long. And that still exists today. And so people have to understand that these associations are not causation. But we still come up against this at time. And the number one reason now that people will say that, well, if I put a patient on dialysis, they are not going to recover kidney function on ECMO or off of ECMO, and I will show you data that those are not true. So the fact that AKI is bad and the association with renal replacement therapy and mortality is bad should not prevent us from starting renal replacement therapy, should make us think like intensivists and start earlier. If we look at pediatric cardiac data, again, from the ALSO registry, these data are still pretty bleak from not that long ago. So then a group started using more common or more relevant recent definition, the RIFLE criteria. And if you look at the RIFLE criteria, look at the prevalence of acute kidney injury. Eight out of 10, four out of five patients have a major organ dysfunction that is not the heart and not the lungs. If you look at RRT and critically ill patients on ECMO, what was the reason? Well, the ALSO database, again, as fantastic as a resource as it is, has not been very kidney specific. Fluid overload, AKI, electrolyte imbalance. And then cumulative fluid overload and failure to return to dry weight was associated with higher mortality and a prolonged ECMO run. So waiting for a patient to become fluid overloaded, just like we have seen in every other aspect of pediatric AKI and CRT, is probably not the right thing to do. Prevention of fluid overload may improve outcomes. Now we don't have a randomized trial to prove that yet, but the data always go the same way and I'll discuss that a little bit later. So David Askenazy and this group looked at AKI in non-cardiac patients from the last decade. Am I allowed to say the last decade or is it now two decades ago? Is 2020 the new decade? I don't know the answer. We're going to say two decades ago. So non-cardiac patients. And what they found, again, AKI and renal replacement therapy, lower survival. The patient didn't have AKI. They had better survival. This should not tell us to not start renal replacement therapy. But if a patient didn't have AKI but they received renal replacement therapy, they did better than a patient who had AKI who did not receive renal replacement therapy. Now one of the things that I'm always asked is, how do you know this patient needs renal replacement therapy? I think, I hope, that they are going to get better. So if you're really good at predicting that, please publish that because we do not have good data and hope is not a strategy. If every patient you do a lumbar puncture on has bacterial meningitis, you're not doing enough lumbar punctures. If every patient you send to an MRI has an intracranial mass, you're not doing enough. And I guarantee you every patient you put on ECMO didn't need to be put on ECMO. And every patient you intubate doesn't need to be intubated. But you're looking at patterns and you're saying this patient is high risk and the patient is already on ECMO. Doing CRT is pretty easy. But look at this. Patients who are on RT who did not have AKI did better than patients who had AKI who are not on renal replacement therapy. Same thing is true, although the difference is a little bit smaller for the pediatric cohort. These are data from the Kidmo group trying to understand current practice from 65 centers. One quarter of centers say they don't, other than SCUF, don't provide renal replacement therapy when the rates of mortality are so high, 43% for fluid overload, 16%. So these are the early adopters for fluid overload prevention. And if we look at a US versus non-US location, seems to be a little bit different. Maybe the non-US folks are a little bit more progressive with respect to treating AKI. And if we look again at different indications, there may be a little bit of a difference here. Maybe the cardiologists, I dare say, may be a little bit more progressive. Now if we look at the mode employed, SCUF and CVVH, that's kind of interesting to me because I'd actually like to redo this data because a lot of people conflate CVVH with CRRT. So I want to make sure that when you're talking about CVVH, you're talking about convective clearance. And you can look, are people using the filter only or a separate machine? And with CVVH, there seems to be a little bit more of the use of a separate machine. And that's important because you're going to have much better control of fluid, much more accurate fluid management. And I'll show you some data from Jordan Simmons later about that. So let's get to fluid overload in this group out of Michigan. A single center review from July 2006 to October 2010 from Michigan, hemofiltration through 2008. So that's just using the inline filter and then CVVH. 203 ECMO runs, a quarter of them or so had CRRT, so real bonafide CRRT. Here you can see the reasons. And they had 33 neonates. Fluid overload was the indication in the vast majority to start pain replacement therapy. They had a 58% survival. And CRRT, 34%. But then filter versus CVVH, they found that things started to improve for them. The median initiation of fluid overload was very high, higher than what we would consider to be standard practice off of ECMO, 25% in survivors, 38% in non-survivors, kind of similar to what we had in our first CRRT paper many years ago out of Texas Children's Hospital. So again, prevention of fluid overload may improve outcomes. But I don't have a randomized study to show you that. How am I going to do that? I'm not going to go to the family in bed space one and say, hey, we do really well at CRRT. Your child's on ECMO. We're going to start at 10% fluid overload. And then we're going to go to the family in bed space two and say, we're going to let your kid get to 30% fluid overload to prove a point. We can't do that anymore. So the need for a randomized trial cannot exist at these levels anymore. The association is always there. And there's nothing wrong with putting associative data into context. The degree to fluid overload at CRRT initiation, again, in the cardiac group was significantly higher, 38% versus 14%. Why do we want to wait and let a patient get to 38% fluid overload? 50 kilo kid, let's just call it 40% fluid overload. You can do the math. That's not the first space. That's not the second space. That's not the third space. That's outer space. Thank you. Thank you for laughing at my jokes. The second show is at 10 o'clock. Drinks are half price. So all right. That's the kind of fluid we should be taking, by the way, is those drinks. The data always go the same way. Giving more fluid does nothing except going into the tissues, into the liver, into the heart, into the lungs. Dr. Erican has a beautiful paper in patients not on ECMO, not on CRRT, that shows at 15% fluid overload, the oxygenation index starts to go up. That fluid is going into the lungs. And now you have ultrasound. So I like to have bets with people who are telling me at the bedside that this patient is third spacing. And I say, please put an ultrasound on their IJ, if not the size of Lake Michigan. I will buy you your favorite frosty beverage. In univariate analysis, again, there is an association between initiation and discontinuation of fluid overload and mortality. Each 1% increase in fluid overload, the odds of mortality increased by 4%. And again, at the end, change in fluid overload. So getting that fluid off is not associated with improved outcome. So it's really a preventive strategy. And we're pediatricians, or many of us here are pediatricians. So we think in a preventive mindset. Bicycle helmets. Those things in the electrical outlets to prevent kids from electrocuting themselves. We used to vaccinate in this country, but some of us still want to, right? So these data suggest that prevention of significant fluid overload is likely more effective at improving outcomes than attempting fluid removal. Exactly what we said in 2001 in that small pediatrics paper. We need to prevent fluid overload, not treat it. Now again, this is Matt Payton's paper, and I highly recommend this paper to you. I have it on speed dial on my phone. And I know the KidMoot group, I just spoke with Jeff Fleming, is looking at this issue. But we are often told, if I put this kid on ECMO and I put them on CRT, these kids will not recover their kidney function. That is not true. Almost all of the kids in this group out of Children's of Atlanta had recovery of kidney function if they survived. The ones that didn't had primary kidney disease. They had focal sclerosis, or membranoproliferative glomerulonephritis. They had an underlying immunological kidney disease. I have not had a patient in my, I hate to say it, 20 years of practice who went on CRT, on ECMO, who survived, who has end-stage kidney disease. So that is just fear and telling you, I'm going to hurt the kidneys. Dialysis does not hurt kidneys. Okay, so you have to decide how you want to do this, and there are a number of different ways. Again, here's the HEMO filter, and I'll give you one little pearl. In kids who are 5 kilos and under, we will use the inline HEMO filter and we'll do CVVHD. The reason is the machine that we use, and others have demonstrated this, patients will become thrombocytopenic if they're 5 kilos or under. It's clearly associated with the CRT membrane. It's probably just clumping, and so we'll do inline CRT. We had a serious adverse safety event a number of years ago doing CVVHD this way, where we had the lines reversed and we were pumping dialysis fluid into the baby, and no one was weighing the dialysis fluid, and we didn't discover it after three days, and this baby ultimately perished. So this became a huge root cause analysis at our hospital, and we have these lines traced twice daily. If you do CRT this way, you have to be very careful because something simple like this can harm a patient, and so we have a standardized protocol for that. In this configuration, CRT is done in parallel, and it works quite well, and we usually go post-pump pre-filter, but just do it the same way each time in your institution. There is no better way, but it shouldn't be that when I'm on, we do it one way and Dr. Erikan's on, we do it a different way. Just do it the same way. So finally, last slide, or two, these are data from Jordan Simmons, and what he did out of Seattle is they looked at the fluid accuracy of doing inline CRT with the hemo filter and dialysis fluid versus using a CRT machine, and in the solid line, you can see free-flow CRT, and there is a persistent accumulation of fluid in the patient, where in the integrated CRT, it's much more accurate. So we actually, with a urometer, when we do inline CRT, we actually measure the fluid, and we balance it, and we talk to our ECMO team about how inaccurate it is and how we handle that. So in conclusion, AKI in ECMO is just like AKI in other settings. It's independently associated with poor outcomes. AKI is bad. A modifiable aspect of AKI include early initiation of renal replacement therapy to prevent worsening fluid overload, just like in every other population at risk for AKI. Starting renal replacement therapy does not lead to ESRD. Starting renal replacement therapy does not lead to ESRD. And starting renal replacement therapy, no matter where you are, does not lead to ESRD. So please do it. Thank you very much. Thank you very much for that great talk, Dr. Goldstein. In the limit of the time, we're going to hold questions until the end. We appreciate your jokes. We'll make sure to tip the waitstaff. So I am really excited to introduce the next speaker, Dr. Aisha Arakan, who is the medical director of critical care nephrology and the director of inpatient dialysis at Texas Children's, who's going to be talking about extracorporeal liver support. All right. Thank you so very much, Dr. Basu. It's a pleasure to be here. It's a treat to be at SCCM, and thank you for the invitation. And we are going to shift gears. We're going to talk about something that is niche, that only exists, well, that probably exists in your institution, but you don't really think of it that way. And then the more novel but not novel, old technology but not really commonly applied technology probably doesn't really exist in many institutions. There's four North American centers that currently utilize it. So open your mind. We're going to try and frame why we use extracorporeal therapy in liver failure, which is the topic of the talk that I'm assigned today, taking it from the rationale, trying to think about why the primary indications, what might be the expected benefits. And then I chose to focus on the available pediatric literature to highlight it, because if I showed you the adult data, you'd just leave, because it's negative. And then I'm going to summarize our center's sort of different way of approaching this very challenging condition. So liver failure for the intensivists is a perfect model for multiple organ failure, because the dying organ becomes an engine for damps, and the organ crosstalk leads to failure of almost any organ system that you can think about. And we always say the only established treatment, especially in acute liver failure, but also in chronic liver failure, is liver transplantation in pediatrics, and our goal is to get our patients there. And the current management will treat the cause if the cause is identified and is mostly supportive in managing complications and trying to bridge the patient, which is a good idea if you're bridging for 48 hours. But ladies and gentlemen, depending on where you practice in the world, that time interval might be more than four weeks. So very quickly, let me maybe remind those of us who are not from North America, United Network for Organ Sharing, UNOS, lists acute liver failure patients as status 1A in pediatrics, and older kids and acute or chronic liver failure patients get listed as status 1B. 1A patients, which is sort of like high-urgency liver transplantation, wait about seven days. This is measured in hours in Europe. And the 1B patients wait weeks. The point-based listed patients wait months. And let me give you an example from September of this past year. On that day that I accessed UNOS, there were over 100,000 patients awaiting transplantation of any organ, and the next most common, the second most common transplantation is liver, but there is only 10% of donors available. So what ends up happening is very soon, our liver failure patients look like this. They're on CRRT for renal replacement therapy. They're on some sort of augmented clearance for liver support, and maybe they're getting also plasmapheresis, plasma exchange for hematological support, and somewhere in there is the baby. And the idea of extracorporeal support is not new. Even though we're calling this session Novel Extracorporeal Therapies, the idea of extracorporeal therapy in liver failure almost dates as far back as the idea of extracorporeal therapy in renal failure. However, there is challenges associated with these systems that we're going to go into the details, especially regarding pediatric data. Here's where the problem starts. This is a typical filter, conventional hemodialysis filter, and small molecules such as glucose and urea and creatinine are filtered freely. The middle molecules require a little bit more augmentation, maybe an altered dialytic modality, maybe convective clearance to be able to be cleared. But the majority of the toxins that actually accumulate in liver failure are bound to albumin, and this filter is not permeable to albumin. Therefore, the protein-bound toxins are not able to be cleared effectively in patients who have liver failure, leading to an ineffective treatment if we only use conventional hemodialysis or hemodialysis-based therapies. On top of that, liver doesn't only do purification. When you have liver failure, you also have synthetic failure of the synthetic duties of the liver, and it's a huge immune organ. So the ideal liver support system should actually be supporting all of the three modalities, and it unfortunately does not exist. So what do we do? Like we do in pediatrics in most things, we make do. We use the available modalities and use CRIT for water-soluble, water-based clearance, and we use augmented clearance with other modalities in order to try to remove protein-bound toxins, and then we could do a factor replacement and maybe do aphoresis like we do in our center in order to be able to treat the coagulation perturbations and try and replete vitamins. The immunological function of the liver we cannot support, so we throw antibiotics at every patient, and the other functions that we don't even currently know about, we don't support. So no wonder this modality or the patchwork of modalities isn't always 100% successful. So these are the artificial, non-cellular, extracorporeal liver support modalities that I have listed for you. The ones that I highlight in red are the ones that I'm going to show some data on, pediatric data, because they actually have been used in children, and there is published data available. But we can't really go there before maybe spending a few minutes for hepatic encephalopathy. And when you talk about hepatic encephalopathy, ammonia follows for us for intensivists, right? Hepatic encephalopathy is incredibly important, especially for survival with a native liver. The lack of it almost predicts being able to survive without a transplantation. And in clinical practice, it's incredibly challenging to assess, especially if there's other organ failure that is requiring ICU level of care, where we frequently are intubating, supporting, sedating these patients, and confounding our own exams. So we use ammonia as a surrogate marker for hepatic encephalopathy. It's not perfect, but it does associate with outcomes. For example, admission-level ammonia, if it's greater than 200, seems to be associated with decreased survival overall, compared to a lower-level ammonia at ICU admission. And if the ammonia is high, but it can also be cleared with extracorporeal therapy, survival seems to be better compared to the patients who don't actually have a response to the extracorporeal therapy with ammonia decrease. This effect seems to be modified in liver failure and not necessarily emphasized as much in patients who have inborn errors. So clearly, ammonia is a surrogate for the underlying pathophysiology in this patient population. So we do supportive therapy, and we're focusing on the extracorporeal therapies in sort of graded fashion. So we've decided that we're going to do extracorporeal therapy. We could be doing CRRT. We could be augmenting it with TPE, or we could be using single-pass albumin dialysis or multi-pass albumin dialysis. And we'll visit them each now with data. So we're going to talk about Dr. Deep's data first, who is in the audience. And this is from King's College in the UK. And he very nicely presented elegantly his pediatric acute liver failure series, who required CRRT. In 165 patients with a pediatric acute liver failure, 45 patients received CRRT. And the outcome was such a discrepant for the patients who were on CRRT that from 2011, they've shifted their practice to actually do protocolized, routine, high-volume CRRT in patients. And the indications I've listed for you there, most of them are conventional. But there is a few unconventional indications as well, such as rising ammonia or hyponatremia. And the prescription is very standardized with some anticoagulation with prostacyclin or heparin or sometimes no anticoagulation. What he's shown is that the patients who receive CRRT have a significant survival benefit compared to the patients who don't receive CRRT. And the patients who are infants seem to be at higher risk of mortality. And the survivors are the ones, just like I showed you, are the ones who actually have a response in their ammonia. So if the ammonia decreases, it seems like the patients will do well. But it's not because we do things differently to the non-survivors because we have a grudge against them. We are doing the same thing to those patients, but they're not surviving. Either we're not able to keep up with their rate of ammonia production, or there is something else going on. And there is something else going on. Because as central as ammonia is to the pathogenesis of hepatic encephalopathy, something else going on. And there is something else going on. Because as central as ammonia is to the pathogenesis of hepatic encephalopathy, more and more, we understand that disease process to be an inflammatory disease, with cytokines actually de novo being produced in the brain. And if we do not address that portion of the disease pathobiology, then we're not going to be able to support our patient 100% effectively. So we could use routine ICU care, like we do, for any sort of cerebral protective measures. But we need to think about addressing the inflammatory disease process. And the way we choose to do this is through augmented clearance. Now, there is several ways that you can augment clearance on extracorporeal therapy. You can start with CRT, and then you can give high convective clearance, which leads to high volume hemofiltration. That dose that I'm listing you there, 80 milliliters per kilo per hour of ultrafiltration, is more than three times the dose that we would normally prescribe for the patients to put it into context for us. And this is the French group who reported on pediatric acute liver failure and their routine use of high volume hemofiltration. And they had about 55% survival, with some patients being bridged to recovery, which is really great in the realm of pediatrics. And they report improvement in all biomarkers of interest, such as creatinine and bilirubin. And hemodynamics improve, and hepatic encephalopathy grade improves. You're going to see a version of this in every modality that I show you. And then we're going to discuss about what the shortcomings are in a little bit. This is single-pass albumin dialysis. This is also from Europe, Single Center, case series of patients with acute liver failure who were treated with SPAD. Albumin gets added to the conventional dialysate solution and passes counter-current to the blood once in the CRIT filter, and that is spent. Hence why it's called single-pass, compared to the multiple-pass version that we're going to see in a minute. And these investigators also reported improvement in biomarkers of interest with bilirubin, bile acids decreasing, and ammonia improving as well. There is some concern always, especially with patients who have some coagulopathy who are going on extracorporeal therapy, whether they're going to be worsening their coagulation perturbations. And these investigators did report that the fibrinogen levels and the platelets decreased, but they were not statistically significant. And the mean arterial pressure remains stable throughout the treatment. Now, what we do in our institution is this, which is a version of multiple-pass albumin dialysis. It is the molecular adsorbent recirculating system that Baxter services vendors here in the States. It consists of four filters. There is one albumin filter. There is one anion exchanger. There is an activated carbon adsorber and a standard hemodialyzer. So let's magnify that and look at that a little more in detail. It's actually a pretty clever way of approaching and trying to solve this albumin wastage problem, but it's about the same principle as SPAD. This is the blood filter where the patient's blood is dialyzing against an albumin solution. And this is the conventional hemo filter where the albumin solution is dialyzing against the standard dialysate. The albumin filter is supposed to be filtering the protein-bound toxins, and the conventional filter is assisting removal for water-soluble toxins. And what happens in the middle is that by passing the albumin solution through an anion exchanger and a charcoal hemoperfuser, the toxins that are bound onto the albumin molecules that are trapped in that solution are being stripped. So albumin is regenerated to be used one more time, countercurrent to the blood filter against the patient's blood. So the molecular adsorbent recirculating system, the MARS system, can remove albumin-bound toxins and water-soluble toxins, but that also is albumin-tight. So it won't remove albumin. There is a little bit of albumin loss, but not clinically significant. It doesn't remove immunoglobulins, doesn't change clotting factors, or doesn't remove bigger proteins, such as hormone-binding proteins. And when you read the pediatric literature, you'll see the same thing that I just showed you on the CRRT, CVVH, and SPAD be repeated. All of the biomarkers of interest improve. There is a beneficial effect on hemodynamics. Hepatic encephalopathy improves. The intracranial pressure decreases. There is a little bit of contradictory reports on the effects on hematological parameters, with some thrombocytopenia being reported, but usually no serious increased bleeding events. But the survival benefit that we all search for evades us. And the reason for that is because pediatric ELS literature, especially as it pertains to MARS, is a study in clinical heterogeneity. There is some studies, a number of studies that I'm going to summarize to you in a second, but they're all low-quality, single-center, uncontrolled studies. And more importantly, the bulk of the data lives in the adult literature, but the case mix in the adult patients prevents us from extrapolating from that patient population. Pediatric liver failure is a completely different disease process compared to acute liver failure. Another confounder, like we tried to highlight, is the waitlist time, because depending on how long we need to support these patients, the need for extracorporeal therapy can actually additively increase over days. Some of the data comes from an older filter that is no longer used, and it wasn't very effective in terms of clearance, and the lack of standardization is a pervasive problem across the studies. Here are the available pediatric MARS studies. The first column that you see after the first author is the number of years that the patients were accumulated on, and the second column is the number of patients. So you'll see, in a given center, usually there's about one to two cases a year, and some of these cases were accumulated over a decade. The primary underlying diagnoses were very heterogeneous. The indications could be hepatic encephalopathy, but could be a biochemical abnormality, and the number of treatments varies significantly. The one at the bottom is our center, which is a US center. The others are European centers, mostly, except for one North American center, and you'll see that the number of treatments actually are a reflection of how long they've been treated in the coming slide in a second. So one of the things that really was difficult for me when I first took on the challenge of trying to create an ELS program at Texas Children's was try to understand how to dose ELS, because as a nephrologist, I know dialysis needs a dose. Just like any other prescription, it needs a dose. When we look at the literature, and we say one thing works, and the other thing doesn't work, we really need to understand what we're talking about in terms of what modality it is, what dose is being prescribed, and how it is being utilized. And the lack of a protocol in many centers that have used this technology makes it very difficult to make comparisons across patients, let alone compare studies across centers. So this is a slide that I'd like to call, none of these things are like any other, because this is the same studies that I just showed you that are now broken down into the technical prescriptive elements of the therapies, where the modalities are different, the blood flow that is prescribed is different, the dialysate rate is very, very different, and more importantly, biggest confounder is the number of days that the patient needs to be supported on is very few in most centers, except for where we practice, which is North America. So I'm going to show you this one as well. This is the relief study, which is the biggest acute on chronic liver failure hepatic encephalopathy study that was done in adults. It was a randomized control trial. These are the exclusion criteria, which read like our inclusion criteria. So all of our patients would have been excluded from the study. This negative study doesn't really apply to us, because we are dealing with a completely different population. Even when investigators try to compare modalities head to head, the improvement or lack of improvement cannot really be quantified or understood in the absence of really understanding what the prescription is. One word about the therapeutic plasma exchange, which really is probably the one that would check the boxes if you don't have access to a specialized system, this is the big randomized control trial in acute liver failure in adults from Europe, multicenter, where there was a significant improvement in survival in patients who received high volume plasma exchange, especially in those who were not liver transplant candidates. And there is many case series, especially from the Far East, because of a dearth of donors there as well. One that I want to highlight here is this next one, where there is 17 infants that were actually bridged to transplantation or recovery with a CRRT combined with TPE, where the investigators reported improvement in all biomarkers. One caveat that we need to keep in mind is that because TPE is an intermittent therapy, when the timing of the sampling is done in terms of biomarkers is very important. Right at the end of TPE, when you sample, you probably are sampling the plasma that you just replaced the patient's plasma with, which is basically the blood bank plasma. So what do we do? We do, this is a reminder to you that none of the systems actually check anything, check all the boxes. So this is what we do in Texas Children's Hospital. We combine them all. We use CRRT for hyperemia. We use TPE for coagulopathy, medical refractory. And then we use MARS for hepatic encephalopathy. We published our protocol for our indications recently, well, I guess two years ago now, for the first 15 patients. These are, I'm happy to share it with you. These are our indications and our protocol. We have modified it slightly now. We used to do eight-hour treatments. Now we do continuous treatments. And the first 15 patients that we reported on were a combination of acute liver failure and acute on chronic liver failure patients. And you're gonna have to trust me when I tell you they were very sick. And the hepatic encephalopathy grade really improved in the patients who survived and didn't improve in the patients who did not survive, like many other investigators have reported. And compared to the two-day wait list that most European, or two-day wait until high-urgency liver transplantation that most European colleagues report, ours waited for 22 days. We're now actually up to 30 patients, so I should update this slide. Our overall survival is about 77% with good renal recovery, despite having prolonged renal transplantation. So the takeaway points is that extracorporeal therapy is frequently required in critical care, a critical liver failure, and no perfect modality exists. But protocolized high-volume hemofiltration and protocolized CRRT seems beneficial. And in the absence of dedicated equipment, augmented clearance with CRRT and TPE seems the way to go. However, most places will not really have enough clinical volume to be able to generate personal expertise. So a standardized approach is probably the best way to go, and newer platforms are being developed, so they'll be coming soon. And with that, I thank you. I apologize, I went over, I know. Thank you. Thank you very much, Aisha. So I'm gonna take just a quick moment and make sure everyone's aware that tomorrow is a very important day for all of us, and I want to thank you all for being here. Thank you. I hope everyone's aware that tomorrow afternoon we have a pediatric section meeting. Everyone's invited. I hope you can come and read and see A over at the Hyatt. So it is my great honor to introduce the next speaker who doesn't need an introduction, really, but in this room of pediatric practitioners where we are innovators, and we are people who are not like the stodgy nihilists who are adult practitioners. One of the greatest innovators we have in our field is standing here with us. It's Dr. Joseph Carcillo, who's a professor of pediatrics and critical care medicine at Pittsburgh. Thank you so much for joining us. You're so kind. Someone's gonna have to help me learn how to use the mouse. That's a novel innovation. I'm limiting my talk to blood purification today, and this is to show you the only meta-analysis that has looked at blood purification shows there's two therapies that are associated with improved survival. The upper therapy is hemoperfusion. The lower therapy is plasma exchange. The first half of the talk will be about plasma exchange, and then the second about hemoperfusion. I apologize to our international members, but there is no really international database that I could find, but in the U.S., we have the PHIS database, and this comes out of Children's Hospital of Atlanta. Dr. McCracken had out, and just to show you the use for severe sepsis in the U.S., ECMO's 4.2%, CRT is 5.3%, and plasma exchange, 4.2%, and also tandem therapies are common. 11.9% of patients on ECMO get plasma exchanged compared to 1.8% of ECMO patients who get CRT and ECMO. After Dr. Goldstein's beautiful talk, I'm sure it will be up to 10, maybe 20%. The mortality, just so you can see, of patients who get plasma exchange is 21%. CRT, 45%, and ECMO, 49.5%. That's according to that. How do I get this to work? Trong, just click here, I guess? Yeah. Okay. You've already met our second recipient. In 2012, Angelica Hale was a precocious four-year-old who loved to sing, but after several days of feeling quite unwell, Angelica's parents took her to see her pediatrician for fever and cough, then difficulty in breathing, and eventually diarrhea and abdominal pain. On presentation to the emergency department, she was thrombocytopenic, leukopenic, and had an elevated protein, reactive C protein. She tested positive for influenza. Upon admission, her respiratory status worsened, and two days later, she was in septic shock. She was then transferred to the Children's Healthcare of Atlanta for further care. Her status continued to deteriorate. Due to acute kidney injury and thrombocytopenia, yet lack of schistocytes on blood smear, the diagnosis of atypical hemolytic uremic syndrome, complicating sepsis, seemed likely. Plasma exchange for thrombocytopenia, associated with multiple organ failure and continuous renal replacement therapy, was initiated. Angelica was supported on VVACMO for 10 days and received plasma exchange for five of them. Although initially receiving enteral nutrition, development of pancreatitis led to parenteral nutritional support. Throughout her ICU PICU course, her family participated in daily rounds with her care team. She received physical therapy at all stages of her illness, and as you've seen, she eventually made a full recovery. It is now my pleasure to welcome back to the stage Angelica Hale, along with her family and members of the ICU care team from Children's Healthcare of Atlanta. with no proof anyone could hear in our hearts a hopeful song we barely understood no we are not afraid although we know there's oh sorry okay so we just completed this this is a US study of 10 centers called the phenotyping multiple organ failure and sepsis and approximately 10% of children have TAMOF or thrombocytopenia associated multiple organ failure like Angelica who was saved at Children's Hospital Atlanta for Dr. Goldstein 92% of these patients are on CRT so your your patients already are on the CRT machine they already have extracorporeal support and it's really not that big deal to do plasma exchange if you choose to. The diagnosis is thrombocytopenia AKI so basically it's what Dr. Goldstein said and Dr. Arakan said plus thrombocytopenia you should see an increased LDH that's indicative of endotheliopathy and hemolysis you can have a normal or an abnormal PT or INR you don't have to have abnormal it includes DIC thrombotic thrombocytopenic purpura and thrombotic microangiopathy you should have multiple organ failure and you want to remove the underlying cause you start steroids because oftentimes it's antibody related and then plasma exchange protocol the plasma exchange protocol is similar to what you do for TTP one and a half times volumes the first day then one volume each day afterwards you have to stand at the bedside when you do it they can become hypocalcemic from the citrate load we give calcium chloride 10 per kilo to teenagers 20 per kilo to children and then 30 per kilo to infants or babies and then start an infusion at that same amount per hour we check ionized calciums every 10 to 15 minutes if it's below one we give calcium. Dr. Nguyen in the audience here from Texas Children's showed what the path of physiology of this is you see ultra large von Willebrand fragment multiples clogging the micro vessels of the brain at autopsy and of the kidney. Dr. Nguyen also showed that when you do plasma exchange you restore Adams 13 activity to break down these ultra large von Willebrand fragment multiples Adams 13 used to be called von Willebrand cleaving protease. The easiest thing to do when you do plasma exchange is to look at the changing color over time so you'll see dark color and then it'll become increasingly clearer very much like changing the oil it makes it very difficult to do randomized trials if you choose to do a randomized trial you have you have to use an opaque bag over the bags we've made the mistake now twice of trying to do randomized trials and lost equipoise because of that. The evidence right now for using it comes from two registry trials first we'll talk about the Turkish trial done by Dr. Sevketoglu. You can see the PILOD score on the y-axis and the days on plasma exchange on the x-axis and what you can see is that the PILOD score returned to normal faster with plasma exchange and without. The length of stay if you choose to do plasma exchange in these patients was 10 days. I'm sorry the length of stay for the plasma exchange group in this without plasma exchange was 10 days it was 23 days with it so your hospital length of stay was longer it doesn't shorten your hospital length of stay but it improves your survival so what they found statistically was a value of 0.006. When they looked at the unit variables that were associated and put into multivariate analysis still if you have a higher prism or PILOD score that is a risk factor for death but the other independent risk factor is receiving or not receiving plasma exchange. In this case the odds ratio if you didn't receive plasma exchange was 1.8 if we were to switch that to the ratio of mortality with receiving it it's approximately 0.48. This is the study done in the United States by Dr. Fortenberry's group which presented that child yesterday that was our opening ceremony and here you can see that the group that got plasma exchange has a significant reduction in PILOD over time but the PILOD score does not change in the group that didn't get plasma exchange. They did not see a difference in the two groups in overall mortality but when they looked at unit variable analysis they found that PILOD, presence of ECMO, CRT and MRSA were associated with higher mortality and when doing propensity analysis with TPE using standard multivariable analysis using modified Poisson regression as well as propensity score adjusted with propensity score specified as a continuous covariate as well as propensity score adjusted with inverse probability of treatment weighting in each case the use of plasma exchange reduced your mortality by over 50%. So this is because in Turkey they tended to give plasma exchange to the less sick patients and in the U.S. we tend to give it to the more sick patients. Now I'm now going to show you two studies where people have used it on ECMO. One is the University of Michigan which in the interest of time I won't show the whole study but this is their circuit set up as Dr. Goldstein said you can go low pressure circuit you could hook it up on the venous side or high pressure on the arterial side. I would call the University of Michigan or you can call us and we'll help you through it. I don't want to just say do it this way in such a short period of time. We are now a pediatric intensive care unit and cardiac intensive care unit society and so I want to just show any cardio intensivists and the cardiac intensivist good. So I just want to show you that pretty soon in the U.S. it's going to become standard of care that if you have a cardiac patient whose thrombocytopenia with multiple organ failure they're going to be plate and they're on ECMO that they're going to be placed on plasma exchange because they were able to attain a 56% survival rate. Now in the U.S. if your mortality rate goes over 2% they shut down your center so you're going to get a lot of pressure from your cardiac and thoracic surgeons to consider this therapy. So please look at this article in Frontier Pediatrics. So on the x-axis negative means days before initiation and then positive is days after initiation and you can see here in these patients with three organ failure thrombocytopenia after cardiac surgery that the platelet count was 100,000 with TAMOF and then it continued to go down and then when they did plasma exchange it went up. This group thinks that if it's seven days you have a response it's good and if it's seven days you do not then they don't think you should continue. Here you see the same thing the number of organ failures went up before starting plasma exchange and then went down afterwards. The Turkish group also has done a study with plasma exchange for macrophage activation syndrome. Now macrophage activation syndrome can occur with TAMOF so you can have TAMOF patients with high ferritins or you can have TAMOF patients without high ferritins. So in the first Turkish study I showed you the ferritins were normal but in this one they're quite high. They're quite high. If you were in this study they did bone marrows on all the patients you do not have to do bone marrows you can just do ferritins to help you do this but this was just for the purpose of the study. This is Dr. De Merkel. 100% of the patients had fever, 96% cytopenia, 96% anemia, 91% thrombocytopenia so that's the TAMOF right. They had 87% hypertriglyceridemia, 100% hyperferritinemia, 100% hemophagocytosis so basically if your child has fever and hyperferritinemia you can consider this. In this population they showed that with five days of plasma exchange they normalized the ferritin level, they normalized the platelet count and they normalized the LDH so there's no need to go beyond five days with a therapy but when you're treating macrophage activation syndrome with TAMOF together it's not just plasma exchange. You also have to turn off the macrophage activation and whether you're on plasma exchange or not if you use dexamethasone or chemotherapy your mortality rate is 50%. If you use methylprednisone and IVIG they had no mortality rate. All the mortalities with dexamethasone and chemotherapy were from secondary infection. Now on to hemoperfusion. Most hemoperfusion is not done in the United States so the remainder of the talk is not a United States centric talk. These therapies are used in Europe but predominantly in Japan and there's a paper that just came out in our journal specifically proposing, the Japanese are proposing that we look at it in late onset neonatal sepsis versus early onset neonatal sepsis that we use it in early onset neonatal sepsis because the organisms that you get at birth are endotoxin producing with the exception of group B strep. In other words the gram negative rod infections. Now the way this works is it's not an exchange, you're hemoperfusing and the polymyxin is attached to the column. You can see there's been a wealth of experience in Japan. The research started in 1980. It was approved in the Japanese health insurance system in 1994. The European pilot in adults was done in 2005. The most recent study done in the U.S. with adults was a negative trial so there was no benefit. However in secondary analysis they believe it might be related to the amount of dosing as well as the level of endotoxin so there's a new trial being done in the U.S. and Europe in adults to address this. Now the Japanese published again if you look at our journal last month a small study six patients with early onset sepsis newborns and then six with late onset they showed no benefit with the late onset but benefit with the early onset all limited to their gram negative infections. So it's an emerging therapy for if there's anyone here who takes care of little babies something you consider. The trial that was done in Euphrates the reason that they want to reconsider it is there were 450 randomized and they took out the patients who had endotoxin activity between 0.6 to 0.9 rather than ones with high activity and when they did that subset analysis they did find a significant improvement. Now the other adhesion column that is available in Europe but not in the U.S. is Cytosorb. Cytosorb is advertised as removing cytokines. It's not specific for endotoxin. This is what that cartridge looks like. It's not available to us in the U.S. but it's available in Europe. It has a very interesting profile. They did one patient with MAS and it showed that with one pass they removed all the cytokine abnormalities. These are patients with sepsis and you can see it removes all the cytokines within 100 to 150 hours. There is one trial that's been done in septic shock patients and again similar to the plasma exchange studies we talked about it was a propensity analysis. 101 received Cytosorb versus 109 with CRT. When they looked at the analyzed mortality there were 67 and 49. There was no significant differences between the populations but in propensity analysis there was. So in conclusion we may be over time I suppose but available for any questions just grab me afterwards. Thank you. Thank you very much Dr. Arcillo. I know the Q&A is the only thing that stands between us and happy hour but I invite the three speakers to come up and I actually have a question to start for Dr. Goldstein. I'm going to be true or false. So true or false all patients who go on ECMO should immediately be started on some form of renal support therapy. Does that include scuff? It can include whatever you wanted to include other than vitamin lasix. Yeah yeah no absolutely. I think that volume control is there are only two things we can control in AKI. Avoiding nephrotoxins that patients don't need and then managing their fluid. And so I would think accurate fluid control and if scuff stands if scuff allows you to or is classified as CRT or renal replacement therapy then yes. If you're saying does every patient need to be on a CRT machine I would say no unless you're very good at it and you want to be more accurate. Again if you want to be very you have to be very careful with your scuff filter is not as accurate so you need to have the protocols in place. All right. Thanks for all three of you given kind of for giving great There goes that question. It's a little bit of a downer, isn't it? But just a little caveat there. I think we need to be careful before we assign that diagnosis. I agree. Next question with the lights coming back up. Thanks. Thank you. I would like to ask about TAMOF. Yeah, so we published a study that you can do CRT and plasma infusion. You just have to change your CRT filter if you're doing through plasma filtration every eight hours. The plasma filtration filter will pull out the ultra-large 1 milligram fragment multimers. The advantage to plasma exchange over plasma infusion in studies, for the interest of time I didn't show them to you all, is the fluid overload issue. So you need to give them 50 mLs per kilo per day of plasma. And if you're not doing that with CRT, you can just do plasma exchange alone and you can reverse the coagulopathy without any fluid overload. Now if you have CRT going, then you can just give plasma infusion and then use your filter but just change it frequently. Do you have CRT? Yes, so you can do CRT. So start your steroid because so much of it is, just for 24 hours, so much of it is antibody related. At some point again, I wasn't talking about all the disease, but at some point, depending on your items 13, you have to decide if the patient's not getting better, if you need Eculizumab, if the ADAMS 13 is less than 10% because that's acquired TTP. There's a talk at the meeting tomorrow about that. And then the other thing is, if your ADAMS 13 is between 10% and 57% and you can't get them off the CRT, you send off complement levels. And if their complement levels are activated, then you're worried about lack of inhibitory complement and you can treat those with Eculizumab. I think I said Eculizumab the first time, I meant Rituximab. So if your ADAMS 13 is less than 10%, Rituximab to turn off the antibodies that are causing acquired TTP, if it's between 10 to 57%, then you make a decision about whether you want to use Eculizumab to inhibit the activated complement. So yes, you can do CRT plasma infusion. That lights off as us getting the hook. Okay. Appreciate that. Thank you again to our speakers. They'll be available outside the room. I appreciate you being here.

extracorporeal therapies

plasma exchange

hemoperfusion

TAMOF

sepsis

survival rates

toxins

liver failure

macrophage activation syndrome

research

efficacy