Good afternoon. I want to thank SCCM for asking me to present, and I would also like to thank Dr. Peter Alexander, a good friend of mine, for moderating this session. As always, it's a pleasure for me to speak about anticoagulation transfusion. I think many of you have heard me speak in the past. I'm very excited to be on this panel and hopefully have a fruitful discussion. I think soon, hopefully, I won't have to talk about anticoagulation transfusion because we'll have conquered the blood biomaterial barrier, and we can have artificial circuitry, and that success will allow us to use these type of advanced technologies such as extracorporeal circuitry with CRT, but more specifically with ECMO, to our greatest advantage in our highest risk patients. Again, I have no disclosures or conflicts of interest. I continue to enjoy the study of platelets and blood surface interactions. As I said, hopefully one day we'll conquer this. I think one of the most important things to remember when we're looking at pediatric patients versus adults is developmental hemostasis. We lost a guru in this field this past year in November, a battle against cancer. We lost Patty Massacott. You guys, many of you know her. She wrote many of the papers, and we collaborated on many chapters on ECMO together and papers together, and also the development of different protocols for monitoring and managing and troubleshooting patients on ECMO. I think one of the greatest things that she always was very influential on pushing on, and those of us who study this very closely feel very vetted to, is developmental hemostasis and how a human being from birth up to adulthood is an ever-changing combination of fibroanalytic system, procoagulant system, and anticoagulant system, and those levels are ever-changing until they get to adulthood where they have more stability and more appropriately sort of reasonable regulation of numbers that we feel are the normal. I think that we sometimes forget that and try to apply an infant or a six-year-old or a two-month-old to adult values, make them a little adult, and they most certainly are not. This kind of describes over time how our system changes when things become more relevant, and we're going to go through several more slides here that approach it using a paper by Toulon in 2016 that looked at this, and they did some excellent descriptions of it. What I've done is I've adapted that into graphs that give you a pictorial view of how we evolve from a neonate to an adult in our hemostatic system. In utero is probably where we first start, and any type of coagulation factors and regulation factors are produced at about five weeks of fetal development, and over the course of the fetal development to 40 weeks, we have a variability in how much of those factors are produced. I think one of the most important things, one of the most telling things that's really interesting is the cord blood is very, very coagulable, and it clots, and yet a baby has a very high risk of bleeding. You can see why at prematurity, when you're looking at 30 to 35 weeks, you have very low factors there, and of course, that makes sense that really premature neonates have a high risk of intraventricular hemorrhage and inability to stop the hemorrhage. At 40 weeks, even at that, we have a variability as to what we have for factors to stop clotting or to stop bleeding. Somewhere between 30 to 50% of our factors are in the adult range at that time. If we look at coagulation factor levels, and we look at the ones that we know from the cascade, the ones that are very influential in us forming a hemostatic clot, you can see that there's great variability amongst them all, and you see that range that goes from 0.4 to one of any one of those across the board, and there's not any kind of sense to it. It's just a mishmash of sort of tied up colored lines. But I think what this does tell us is that there is a reason or a rationale for it, because the babies don't just all bleed to death, and they don't just all not clot. They, in fact, sometimes clot more than we would like them to on the ECMO circuitry when their blood is stimulated and has an interaction with an artificial surface. So I think that the most important thing is that even from 11 to 17 years of age, we do not yet have the adult hemostatic system on board. And I think that when we look at our patients, what that does for us is it influences us, it pushes us to look at our anticoagulation guidelines, and to recognize they are guideline a starting point, but that we very likely will have to customize what we do and individualize what we do for each patient in how we anticoagulate them and how we support them. The end result, if you look at this graphic, is that until we get to adulthood, we have a decreased coagulation potential. That is what you might infer from this graph. In addition, our physiologic regulation is the other part of it. And when we're looking at physiologic regulation and fibrinolysis, so that we don't just clot off, that also doesn't come with normal values at birth. It evolves over the entirety of the time until adulthood. And so there's huge variability in how much protein S, protein C, plasminogen, and antithrombin we have. And you can see that we have a decreased regulation and an increase in fibrinolysis when we first start off in life, and it comes into a normal hemostatic, homeostatic range of balance by the time we reach adulthood. One of the ones that's far less studied, and part of that is because we use heparin, I think, in anticoagulation. Coagulation factors are interesting. They have a lot of names attached to them. The development of the cascade is far more interesting than thinking of a platelet sticking to things. But we are completely reliant upon primary hemostasis for the front end ability to prevent bleeding. And although it's far less studied, what we do know about it is that platelet counts are elevated frequently at birth in the neonate. There's also transient increases throughout that until they reach adult levels at about a year of age. And that although we take platelets out and they seem almost hypo-reactive when we test them within our in vitro studies, what we do know is that there is a shortened bleeding time in neonates and their transition to normal range for platelet counts and normal functionality is at about a month of life. The other thing that's really important and probably why there's a shortened bleeding time is that babies and neonates or neonates are born with a higher von Willebrand factor at birth. And so that may be why the platelets stick more quickly and transition to those sites of bleeding. We also know that between non-O and O blood groups, there is a difference in that reactivity and in the responsivity. So if I were to sum up what a neonate's hemostasis is, it is a combination of four things. A decreased coagulation potential, a decreased regulation potential, an increased fiber analytic potential, and an enhanced primary hemostasis. And those things comprise of normal hemostasis in the neonate. And then they get sick. And so that's where we get challenged. That's where we are working on protocols. That's where we're trying to identify what are the proper measures, managements, guidelines that we would use. This graph, this graphic that I created in 2010, was the first time I created it. And then I recreated it in 13 and 14. And now other people have used it in their presentations. But what this shows is ELSO data. And it shows all possible complications. And then it highlights hemorrhagic complications in the different groups, neonatal, pediatric, and adult. And if I were to apply 2010's data from ELSO and 2020's data, and I applied it a 10-year difference, and I applied the two over top of each other, there would be very little change. Because what we have done in ECMO is very good. We've designed great technology. We've designed excellent cannulas. We've designed a lot of excellent things. But what we have not yet conquered is that blood biomaterial interface, that interaction that occurs in the activation. And so these things continue to occur at a 40 or 30 or 20% level, depending on the age group. Clots also occur, but you can see the range of clots and where those clots occur is very different. So the majority of clots occur in our artificial circuitry, minimally within the patient. And when they do occur, the result is very severe. But it's rare in comparison to where we form other clots, which the oxygenator being the prime source of clot creation, and also clot harvesting, if we want to say that. So how do we monitor anticoagulation for UCLS? I've just told you we have this divergent population of humans. And we now have to kind of monitor these patients and we have to figure out how we're going to apply anticoagulation to them. And we've evolved a lot from where ACT was the sole method of monitoring to something that now gives us other opportunities like thrombolastography or thrombolastometry. Anti-TAN-A has come in. If a survey was done today, we would see that anti-TAN-A is the most frequently used method of monitoring our anticoagulation. APTT, we know, has a huge amount of downfalls, and for the reasons that I explained about developmental hemostasis. What we do know is we need to know whole hemostatic function, and we need to know anticoagulant effect. And so, you know, recommendation would be that you need to know both of those things. So you can't just solely use anti-TAN-A, and you can't solely just use ACT, or TEC, or Rotem, or APTT. You have to use a combination. So you need to use an anti-TAN-A. If you're using unfractionated heparin, that would be the best one to use, or APTT for a direct thrombin inhibitor, and that tells you anticoagulant effect. And then you would have to use one of the other methods above for whole blood assay. And the reason I continue to include ACT, although centers have gotten rid of it, is that not every center has the resources for TEC or Rotem. And, excuse me, ACT, TEC, or Rotem all provide in their respective measures, provide you with a real time identification or examination of how the blood is functioning, and what its hemostatic function is, and how long it takes to form a clot. And that is vital information we need when we manage patients on ECLS. This paper, I had the privilege of writing with Paul Monigal and Graham McLaren. John Snyderman, a former fellow of ours at SickKids, wrote the paper with Paul and Graham and myself when he was doing his fellowship in Australia. This shows the challenges that we have with our present monitoring tests. And I could go through all of them, but essentially, ACT has gotten a bad deal, but it depends how you use it, and it depends what you interpret from it. ACT is not correlatable with an anti-TENA or an APTT, because they measure very different things. Nor is a TEC and Rotem totally correlatable with an anti-TENA or an APTT, because again, we are measuring different things. We are measuring whole blood hemostatic capability versus anticoagulant effect. And you have to be aware that no one test is infallible. Every one of them has their challenges, and every one of them has incorrect or non-believable values at times. And you have to be careful. Therefore, it comes back to understanding your patient and understanding their condition and customization. I just show you this again from Toulon and the group, just to show you that there are, again, variability across age ranges, even when it comes to looking at APTT fibrinogens. So when we use blood products, of course, in critical care and in all advanced cares that we now provide in ICU management, we are trying to limit the amount of exposure to blood products when we can. And it isn't that because you go on ECMO, you should just keep plowing them full of blood products. On the other hand, you should also not prevent patients or restrict them from blood products when they are imperative to be necessary. What you have to understand is the blood products that you use. You must understand that platelets efficacy is most effective at one to two hours post-transfusion, not at the time of transfusion. You must understand that FFP and all of those factors that are within it have half lives. And so it is best used as a drug to be given fresh, not to be transfused over hours and hours. Cryoprecipitate is great, but it has essentially a concentration of four elements that help us to form hemostasis. Recombinant factor seven has very, very specific times where it has some utility in bleeding on ECMO. Prothrombin complex concentrate, used in trauma, used to not have to provide so much fluid and volume with FFP, provides you with five components. Not everything that FFP provides you, but those five components can help to provide hemostatic effect in specific instances. So again, customization, individualization. Antihemolytic factor is another compound that can be administered, rarely so in extracorporeal life support, but its factor in itself is a complex that could be used for acquired von Willebrand's disease that might be seen in a prolonged course of ECMO on centrifugal pumps. Antifibrinolytics, always being called procoagulants or not, they are antifibrinolytics. They prevent the breakdown of existing clot and they should be known as such. You have to be comfortable with them. They help you out in long-term ECMO cases. They help you out in challenging ECMO cases. And we do have other, and the other remains somewhat unknown. Some groups have used factor XIII concentrate. I have not. I've been asked, should I use it? You know, I think that if you can use an antifibrinolytic and prevent the, excuse me, if you use it, antifibrinolytic and prevent the breakdown of a clot, then that would be useful. Calcium, imperative, temperature, imperative, local control, hemostatic pressure, proline, quick clot, local control. These things are also important. You cannot forget the normal practices of how you stop bleeding. And of course, blood in itself is important and should be considered because it's part of the formation of the clot. So if we look at coagulation factors in plasma. I'm not gonna go through them all, but like I said, there is a half-life once we thaw it. And so it's much better used fresh than it is to be drawn out over a course of time of infusion. Cryoprecipitate, their constituents are four of them, as I said, factor VIII being one of those. We know factor VIII is actively consumed by the circuit. Factor XIII is another one, von Willebrand factor and fibrinogen. So they, it definitely helps in hemostasis. I show you this last slide because I think it's important. And it comes from one center. It comes from Junta Rulia's center in Houston in a paper that they wrote in 2018. I like how it's written. I like how it pushes you to consider other options. It pushes you to customize your therapies, but there are many other algorithms written and you can write your own algorithm using this as your original structure to help to decide, customize and individualize how you manage a patient who's bleeding on ECMO, how you manage a patient who requires adequate anticoagulation on ECMO. And with that, I'm gonna stop. And at the end, hopefully we will leave panel time for discussion. Thank you very much for your time. Hello, everyone. I'm Alaa Benbea, Pediatric Intensivist and Faculty at Johns Hopkins University in Baltimore, Maryland. Today, we will discuss survival and neurological and health related quality of life outcomes after ECMO. My disclosures are listed here all in form of research grants and I have no commercial disclosures. So today we will discuss briefly survival trends in ECMO patients, short and long-term outcomes in survivors and early and late approaches to improve outcomes. And to delve right in, the graph displayed here reflects data reported to the Extracorporeal Life Support Organization or the ELSO Registry from 1990 to 2019. The red bars represent number of patients and the blue line represents number of centers reporting ECMO data to the ELSO Registry. As seen, the volume of ECMO utilization has increased dramatically over time, especially in the last 10 years. There were 1,644 ECMO patients from 83 centers reported to the ELSO Registry in 1990 in contrast with over 15,000 ECMO patients from 420 centers reported to the registry in 2019. The inflection point in this graph is represented by the year 2009. A few events took place around that time. The first one was the H1N1 influenza pandemic with a surge of ECMO use in adults with respiratory failure and associated reports of improved outcomes with ECMO support versus conventional therapies. The second event was the publication of the results of the CSER trial, a randomized controlled trial of ECMO support in adults with severe acute respiratory failure conducted in the UK, which purported to show improved outcomes with the use of ECMO, though its findings were controversial at the time. And the third one was the advent of more compact and easy-to-use ECMO circuits and canvas that's allowing for less experienced centers to introduce or augment the use of this technology quite rapidly. Most ECMO growth is represented by adult cases. While neonatal respiratory failure represented more than 80% of total ECMO cases in 1990, neonates now comprise about 10% of all ECMO cases. Despite this tremendous growth in number of ECMO centers and number of ECMO cases, mortality for neonatal and pediatric ECMO represented in these graphs by the black line on top has remained relatively stable. The current thought is that over time and with improved technologies and expertise, patients canelated onto ECMO have had increasing medical complexity, thus explaining the lack of improvement in overall mortality rates. The top line in this graph represents patients without comorbidities and improved survival, while the bottom line represents patients with comorbidities with stable survival rates. The main primary indications for ECMO can be most easily summarized as pulmonary, cardiac, and eCPR or ECMO during active CPR for cardiac arrest. This table represents the latest publicly available international ELSO registry report. Overall survival to hospital discharge in critically ill patients supported on ECMO regardless of age and indications is 55%. There are some important differences in survival rates though. Survival to hospital discharge in selected neonatal populations, newborns with meconium aspiration syndrome, primary pulmonary hypertension of the newborn or neonatal sepsis can be as high as 75 to 87%, while survival in neonates with congenital diaphragmatic hernia is only about 50%. Overall, neonates with pulmonary indications for ECMO have 73% survival to hospital discharge. Survival in children and adults with primary pulmonary indications for ECMO is 60%. The neonates, children and adults with primary cardiac indications for ECMO have lower survival rates from 43 to 53%. And then lastly, patients of all ages with eCPR as the primary indication for ECMO have the lowest survival to hospital discharge ranging from 29% in adults to 42% in neonates and children. Long-term survival data for neonatal and pediatric ECMO are much more limited. A recent population-based corporate study from Ontario, Canada of patients less than 18 years of age receiving ECMO showed that overall survival to hospital discharge was 56.4% with a reassuring little decrease in long-term survival at one year, but with a drop-off to 42% up to five years of follow-up. Another study collecting data from 13 hospitals in California between 1999 and 2006 showed 46% survival to hospital discharge. 56 of 87 survivors were tracked for a median period of 1.2 years. Late deaths occurred in 5% of children and the readmission rate was 62%, a third of which was for respiratory infections. Importantly, 16% of long-term survivors were found to have neurologically debilitating conditions including epilepsy and developmental delay. With that, we'll segue into some of the main drivers for short and long-term outcomes among survivors. Acute neurologic injury is one of the most important drivers of survival and of unfavorable outcomes in pediatric ECMO. It occurs in 15 to 36% of ECMO cases in the form of hypoxic ischemic injury, intracranial hemorrhage, and ischemic stroke. Clinical and electrographic seizures have been described in 10 to 23% of ECMO patients, and brain death is frequently listed as a neurologic complication in ECMO populations, primarily because it is captured as such in the ELSO registry, where it's reported as occurring in 0.2 to 10% of patients, depending on age group and primary indication for ECMO. If neurologic injury occurs, it is by 89% and 10 to as many as 50% of survivors to hospital discharge have been found to have poor neurologic outcomes. The mechanisms of neurologic injury during ECMO are multifactorial. Profound cardiopulmonary failure or cardiac arrest leads to hypoxic ischemic injury pre-ECMO, rendering the brain vulnerable. The inflammatory state associated with critical illness is then compounded by exposure to the foreign surfaces of the extracorporeal circuit, which triggers a global innate immune response by activating the contact and complement systems, and provokes perturbations in pro-inflammatory and pro-thrombotic pathways implicated in the pathogenesis of neurologic injury. Added to that is the need for anticoagulation, rendering the patient at risk for both thromboembolic and hemorrhagic intracerebral complications. And lastly, the use of neck vessels, especially the right common carotid artery, is associated with increased odds of neurologic injury independent of age at cannulation. There are many risk factors for neurological complications and deficits that have been investigated, including ECPR, or extracorporeal cardiopulmonary resuscitation, and sepsis as ECMO indications, veno-arterial versus veno-venous mode, younger age, and pre-existing or newly developed organ dysfunctions. So how and when are neurologic, functional, and quality-of-life outcomes after ECMO assessed? There is no consensus on a set of outcome measures to be used in ECMO patients, which makes sense given the heterogeneity of ages and diagnosis. And these tables show findings from a systematic reviews of outcomes after pediatric ECMO conducted two years ago, in which we identified no fewer than 55 different outcomes measures used in pediatric ECMO. Overall, across 60 studies included in the systematic review, 10 to 50% of children scored more than two standard deviations below the population mean on cognitive testing. Behavior problems were identified in 16 to 46% of children tested, and severe motor impairment was reported in 12% of children. Quality-of-life of former ECMO patients evaluated at school age or adolescence ranged from similar to healthy peers to 31 to 53%, having scores more than one standard deviation below the population mean. We'll next break this down by age and ECMO indication. Data on neurologic outcomes of neonatal respiratory failure ECMO patients come primarily from a nationwide perspective follow-up program in the Netherlands. Serial longitudinal evaluations were conducted over 13 years in 178 neonatal ECMO survivors. Measures of cognition were stable over time. However, attention deficit and impaired verbal and visual-spatial memory were found at school age as well as in adolescence. This emphasizes the inability to predict future academic performance based on preschool cognitive assessments, and that neuropsychological assessments are needed to identify complex problems these children face, along with the growing-into-deficit phenomenon previously described in other neonatal critically ill populations. Data on pediatric respiratory ECMO outcomes are extremely limited. In a one-year follow-up study in the UK, neurodevelopmental problems were identified in 30 of 98 former mixed neonatal and pediatric respiratory ECMO patients, or 30%. The specific abnormalities noted included neurologic, such as seizures, motor or vision abnormalities in eight children, hearing with or without language delay in eight children, and behavioral problems in six children, with 27% having difficulty spanning these domains. An acute neurologic event on ECMO was found to be the only risk factors for neurodevelopmental concerns. Despite having neither a cardiac arrest nor an acute neurologic event documented, 24% still had neurodevelopmental concerns at one-year follow-up. Data from children who require cardiac ECMO support for acquired or congenital heart disease are much more robust. Studies using functional scales and neurofunctional scales, I should say, the Pediatric Cerebral and the Pediatric Overall Performance Category, PCPCR and POPC, to determine outcomes either at discharge or within three years of discharge, showed that 81 to 91% were favorable. And this was based on PCPC and POPC, less than three or less than four, or no change from baseline. PCPC and a POPC of three indicate moderate disability. Studies using more detailed evaluations that included cognitive testing revealed that at one to five years after discharge, 25 to 50% of those seen in follow-up tested in the severely disabled range, defined as more than two standard deviations below the mean for the population on a given test. And this included the Wechsler battery of tests as well as the Bailey scales for infant development. Studies that evaluated quality of life showed that 34 to 53% of cardiac ECMO survivors reported significantly diminished quality of life compared to age-matched healthy peers. And at least one study showed that quality of life was also significantly diminished compared with that of peers with cardiac disease, but no history of ECMO. And then lastly, the largest cohort reporting post-discharge outcomes in children who underwent eCPR stems from the THAPCA trials, the therapeutic hypothermia after pediatric cardiac arrest trials. Of 147 children who received eCPR, 30.5% survived to 12 months with favorable adaptive behavioral scores defined as Vineland Adaptive Behavioral Scale score of at least 70. A score of 70 is like two standard deviations below the population mean. In addition, 22% survived to 12 months with a VAPS2 score decreased by less than 15 points, which is less than one standard deviation. So what are some of the early and late strategies that we can use to improve the long-term outcomes of ECMO survivors? In terms of neurologic outcomes, using a nice diagram by Dr. LaRussa and colleagues, neurologic monitoring and assessment of neurologic complications is a requisite for ECMO patients, both early and late in the ECMO course and hospital stay. The best neuromonitoring by far is to keep the patient awake. And we do keep higher and higher risk patients awake for as long as possible. But not all neonates and not all children on ECMO are good candidates for decreasing sedation. And many patients remain comatose or are encephalopathic after ECMO cannulation. For these patients, there is active research surrounding best multimodal monitoring strategies, including sero head ultrasounds in babies, transcranial dopplers, EEG, cerebral oximetry, noninvasive cerebrovascular autoregulation monitoring, and plasma brain injury biomarker development. Neurologic morbidity, such as seizures, occur frequently in ECMO patients, as mentioned at the beginning of this talk. EEG monitoring for 24 to 72 hours after ECMO initiation has now become routine in many centers for that very reason. In this group of studies where EEG was conducted routinely, electrographic seizures were detected in 18 to 23% of children, 12 to 16% of children on ECMO suffered non-convulsive seizures, and 5 to 11% developed status epilepticus. EEG abnormalities were associated with mortality and unfavorable short-term outcomes. The good news is that seizures can be promptly treated if monitored for and detected in real time. And then abnormal neuroimaging. Any abnormal neuroimaging during and following ECMO is also independently associated with unfavorable outcomes after hospital discharge, and current recommendations call for advanced imaging whenever possible prior to hospital discharge. With that prompt recognition and treatment of any neurologic morbidity in the ICU is important, along with prevention of secondary injury, for example, utilizing a post-cardiac arrest bundle. Other general ICU care known to favorably impact long-term outcomes in critically ill children should be implemented in ECMO patients similar to other ICU populations, including minimizing sedation and managing pain, instituting early mobilization and rehabilitation, and delirium prevention and treatment strategies, assessing readiness to win from mechanical ventilation and careful glucose control. In the post-ICU period, also similar to other ICU patient populations, critically ill children who required ECMO support frequently suffer from chronic conditions, are burdened by rehospitalizations or repeated surgeries, are at risk for decline in functional status and independence, and for pediatric post-intensive care syndrome. Therefore, rehabilitation resources available to children post-ECMO need to be assessed and set in place prior to hospital discharge. A clear schedule for long-term follow-up and developmental services is crucial in ensuring that all survivors of critical illness requiring ECMO reach their full developmental potential. Follow-up is multidisciplinary, addressing all organ systems with potential sequelae after critical illness in terms of nutrition, hypertension in patients with chronic kidney disease, following severe AKI during critical illness, and other language and hearing testing, as well as functional respiratory testing. All follow-up needs to be disease-specific for children with chronic conditions and adjusted for children with neurologic morbidity who may benefit from long-term rehabilitation, developmental, and educational services. There are many professional society guidelines for specific diseases such as congenital heart disease and congenital diaphragmatic hernia, as well as by ELSO for ECMO survivors in general. These can be adjusted based on the norms of each institution, country, and the primary language of the patient and his or her family. With that, I conclude. Thank you for your time, and I look forward to the discussion at the end of this session. ♪♪

anticoagulation transfusion

neonates

developmental hemostasis

pediatric patients

ECMO management

neurologic outcomes

individualized care plans