Hi, my name is Phil Spinella. I'm going to give you my presentation on Massive Bleeding in Critically Ill Children. So there's a difference between Life-Threatening Bleeding and Massive Bleeding. Life-Threatening Bleeding can be both Massive Bleeding, but also a small amount of bleeding in enclosed compartments such as the brain or even within the eye. So it's important when we discuss Severe Bleeding, whether we're talking about Massive Bleeding, which is a large quantity of bleeding versus Life-Threatening Bleeding. Now ATLS does categorize the amount of Massive Bleeding into four different stages according to the percent of blood volume lost. And in the adult population, Massive Transfusion has been defined many ways. Classically, it has been 10 or more units of red cells or whole blood within the first 24 hours. But more recently, the definition has been changing to include either four units of any product over a short period of time or four units of red cells, recognizing that the time period at which the patient is at highest risk of bleeding to death is within hours, not over 24 hours. Now the definition in children have also been multiple definitions in the past, but the first data-driven definition was published by our group, by NEF, primarily. And here we noticed that when more than 40 mLs per kilo of all blood products are transfused within the first 24 hours, that's when the risk of death started to increase. And this has become the most commonly used definition of Massive Transfusion in children. There are limitations to the data, as you can see in the second bullet, but it is the best that we have at the moment. Now Massive Transfusion definitions have only been defined or data-driven in patients, children with traumatic injury. They have not been replicated in other cohorts that can have severe bleeding, such as those requiring cardiac surgery on ECMO, for example. Now when it comes to massive bleeding in children, we performed a survey now about six years ago that indicated that while Massive Transfusion protocol activation is uncommon, it does happen often enough that it's important to study and to improve our practices. And most interest, well, one of the interesting points of the survey was that the vast majority of the indications to activate a Massive Transfusion protocol was physician discretion. And that's basically because we don't have hemodynamic parameters or lab values that can currently accurately predict who's severely bleeding or not. Now the slides that are in a white background are new slides that were added by myself with new literature that wasn't available in time for SCCM to put into their own slides. I'm reading their slides. But most recently, Julie Leonard and I, with a group of others, published a prospective observational study in over 449 patients where there was a Massive Transfusion protocol activation and getting at least one blood product or had more than 40 cc's per kilo of any blood product within the first six hours. So this was the first large multi-center study characterizing the epidemiology practice patterns and outcomes for children with life threatening bleeding due to any cause. And in this cohort, 55% were male, the median age was seven years of age. And then the etiology of bleeding was most common in trauma, 46% with 34% being operative and about 20% being medical bleeding. In the operative group, they were most commonly cardiac surgery patients. In the medical group, they were most commonly GI bleeding patients. Interestingly, the amount of time it took to the first red cell transfusion for these children with life threatening hemorrhage was eight minutes. But you can see a much more delayed response with plasma and platelets, 34 minutes on average for plasma, 42 minutes for platelets, you know, indicating a pretty significant delay in getting plasma or platelet products to children with life threatening bleeding. The incidence of morbidities such as ARDS and acute kidney injury were high at 20 and 18.5%. 28 day mortality in the entire cohort was 37.5%. But there were definitely differences according to etiology, you can see 28 day mortality was highest for those with medical bleeding, compared to 36% for those with traumatic bleeding and 23% for those with operative bleeding. When it came to death due to hemorrhage in this cohort, 65% of the deaths from bleeding occurred within six hours and 86% by 24 hours, showing that, as you might expect, when children die of severe bleeding, they die very quickly, almost completely within the first 24 hours. Now when it comes to the mechanisms associated with massive bleeding, historically, the emphasis has been on reduced oxygen delivery from reduced preload, which is why in the past, ATLS and other resuscitation algorithms have focused on giving red cell transfusions first, and then basic and delaying the use of plasma and platelets. But in the past 10 to 15 years or so, there's been recognition that there's a primary coagulopathy that occurs as well as an endotheliopathy. And the combination of shock, coagulopathy, and endotheliopathy has been coined as a blood failure. The etiology primarily investigated in adults have shown that with hypoperfusion in a massively bleeding patient leads to increased thermodulin and protein C activation, which leads to a systemic anticoagulation and hyperfibrinlytic profile in some patients, not all, that in addition to that hypothermia and acidosis, you know, further increases the shock, coagulopathy, and endothelial injury. And in the past, this has been called the lethal triad, but actually more recently, another group has rephrased this as the death diamond. The fourth part of the diamond is hypocalcemia. So between acidosis, hypothermia, coagulopathy, and hypocalcemia, those four elements are associated with increased death from hemorrhage. Related to the recent interest in what is called hemostatic resuscitation, or damage control resuscitation, is the recognition that giving crystalloids to a patient with severe bleeding can actually be harmful, even though ATLS still suggests that for a massive amount of time, the patient can still be at risk for hemorrhagic resuscitation. Even though ATLS still suggests that for a massively bleeding patient to start with crystalloids, many practitioners have avoided starting to resuscitate patients with obvious or clear, you know, exsanguination and are not using crystalloids or going straight to blood products and using blood products in a balanced manner. And we'll talk a little bit about more of that. We'll talk a little bit more about that in the next few slides. There isn't much data in children or even adults, you know, confirming that giving more crystalloids is independently associated with worse outcomes. But there was this one study from children treated at combat support hospitals in Iraq and Afghanistan suggesting more crystalloids may have led to more mortality and more morbidity. As I just mentioned, since the wars in Iraq and Afghanistan, there's been a rethinking on how we should treat patients with severe bleeding. The overall bundle of care is called damage control resuscitation, and that is made up of many different elements, as you can see here on the slides. But the central tenant of damage control resuscitation is hemostatic resuscitation, and that's giving blood products in a balanced manner. And while we've been thinking about balancing transfusion as ratios, whether a one-to-one ratio of plasma to red cells or platelets to red cells, you can also look at balancing blood product administration and thinking about it as a deficit or not. The amount of red cells minus the amount of plasma, or the amount of red cells minus the amount of platelets, and that includes the magnitude of the imbalance between those products. There is good adult data indicating that giving a balanced transfusion is associated with improvement in shock and coagulopathy and potentially improved outcomes as well. There really isn't as strong data in other patient populations and adults. The best data evaluating ratios in adults is from the proper trial. This is the randomized controlled trial of high versus low plasma and platelet ratios relative to red cells. And in this seminal study, sorry, let me go back to this slide, okay, well, let me go back. In that seminal study, while it did not show any difference in mortality, there was reduced death from bleeding in the adult trauma patients where there was, within 24 hours, when a higher ratio of plasma to red cells were used. There is much less data regarding ratios in children. This study published in 2015 evaluated ratios and transfusing lower ratios, you know, was associated with increased mortality. This was ratios of plasma to red cells, actually. Now when it comes to more recent data evaluating both ratios and deficits in children, from that 449 patient cohort that I described to you before, when we only evaluated children with traumatic life-threatening hemorrhage, there was 191 patients in this cohort. And after adjusting for many potential confounders, as you see on the slide, a high plasma to red cell ratio was associated with improved six-hour survival. And higher plasma deficits, as well as higher platelet deficits, were associated with increased mortality. And if you think about a deficit, a lower deficit is more equal to a balanced resuscitation. So you would expect lower deficits, higher survival. So if there's a hot bigger deficit, there's a risk of increased death. And that's what our data in this study revealed, which to us is evidence that deficits may be more accurate assessment of balanced resuscitation compared to ratios. This is online in peds critical care medicine as of last month. Now I've been only talking about blood components so far. The most efficient way of giving a balanced resuscitation is to actually use whole blood. And before we talk about whole blood, it's important to describe the differences. Whole blood can be either warm and fresh, given immediately within eight hours. And most of the military data published on warm, fresh whole blood is military data. That's compared to cold stored whole blood, which is stored at two to six degrees Celsius. It can be stored actually up to 35 days in certain storage solutions. But most centers right now are storing cold stored whole blood for 14 to 21 days. And the vast majority of the civilian data, and especially recently, is with low titer group O whole blood. So low titer group O whole blood is low titer for both anti A and anti B, less than 256, according to the military, and less than 200 for the Red Cross. Most of the civilian data published in the past few years is with low titer group O whole blood. That's different than ABO specific whole blood that has been published in some of the military data from about 10 years ago, when their studies were evaluating warm, fresh whole blood. I think it's important for the audience to realize that group O whole blood stored cold, since it includes all of the required transfusion transmitted disease testing, is FDA licensed, as I mentioned, potentially up to 35 days of storage. And it is an American Association of Blood Banks standard product. Now going back into the literature for whole blood, interestingly, at the Children's Hospital of Philadelphia in 1991, Mano and colleagues did a randomized controlled trial comparing whole blood to individual components in a one to one to one unit ratio. In this, there was a three arm randomized controlled trial with fresh whole blood, cold stored whole blood, less than 48 hours, and then reconstituted whole blood. And in this RCT, that was in children requiring cardiopulmonary bypass for cardiac surgery, those children that received either fresh whole blood or cold stored whole blood, that's not on the slide, there was a statistical reduction in blood loss, according to chest tube output, in the cohorts receiving whole blood compared to components. So this is an RCT in children requiring cardiac surgery showing both fresh whole blood and cold stored whole blood caused less bleeding. Now you compare that to this 2004 paper by Mu and colleagues and in this RCT, they compared cold whole blood to just plasma and platelets. And it was only transfused during the cardiopulmonary bypass. Post-op in the ICU, they all got components, which was very different than the Mano study that continued the intervention of post-op in the ICU. In this study that didn't do that, there was no difference in their primary outcome, which was a composite outcome, but there was an association with increased ICUs, length of stay, and cumulative balance for those that got whole blood, although they didn't adjust, even though it was an RCT, it was the groups were in balance, there were many more patients on ECMO in the whole blood group. And if they, it would have been interesting to see if they had adjusted for that difference between the groups, if the results would have been different or not. Now there was this one adult study that looked at modified whole blood. This was whole blood that was filtered, that with the leukoreduction filtered, that removes platelets. So it really isn't whole blood at all, it was basically red cells and plasma. And in this small study in adults, there were no difference in mortality rates between the two groups. But when you start to think about now low titer group O whole blood compared to individual components, you can see that there are many potential and some data-proven advantages of low titer O whole blood compared to the potential risks. And all of these statements are backed up by data in this manuscript published by Mark and I now a few years ago. But when you compare whole blood to individual components, whole blood is a more potent product because there's less anticoagulants and additives when you put a one-to-one-to-one back together again. Whole blood has coleslaw platelets in it, there's good RCT data showing that a coleslaw platelet is more hemostatically active, associated with improved measures of coagulopathy in a few RCTs. There's increased storage duration of a platelet-containing product with whole blood, as I mentioned, can at least take it out to 14, 21 days, potentially even 35. This allows you to give a whole blood containing, a platelet-containing product at places that can't afford to keep a five-day shelf life product with room temperature stored platelets, which is what you get when you use components. When you use group O whole blood, there's basically no risk of a fatal hemolytic reaction because you're only using group O whole blood. When you use individual components, as soon as you know the ABO type of the patient, the goal is to change to ABO-compatible products. When you do that, due to human error risk, you can, there's a one in 80,000 risk of a fatal hemolytic reaction, so it's actually safer to go with a group O-based product alone. Again, since it's stored cold, there's less of a bacterial contamination risk with whole blood compared to components. And then a very big advantage are the logistic advantages of giving one product, whole blood, compared to three different components. The fact that you can get the product in faster and maintain a balance for resuscitation is at least, you know, theoretically a potential advantage. And then there's now a few papers in adults. I'm about to show you one pediatric paper indicating an independent association with improved survival when whole blood is, group O whole blood is used compared to components. What are the potential risks of whole blood? Well, when you give group O whole blood to a non-O recipient, there is a theoretical risk that that incompatible plasma may increase immune complex development and lead to endothelial injury. But while this is a theoretical risk, there really is no data confirming it. If anything, there's adequate data showing that the use of incompatible plasma with group A plasma in adults with severe bleeding does not have any adverse consequences at all. There is a potential for increased waste if you're going to stock whole blood now in your hospital blood bank. If you don't use it before it expires, that may lead to more waste. But this can be mitigated by using group O whole blood in all patients with massive bleeding. And this has been done in places such as in South Texas where they have a less than one percent waste rate for group O whole blood in their system. There's, of course, then the ease of over-resuscitation since it is a lot easier to give. But that, you know, should be able to be controlled with good QI processes and monitoring of the incorporation of whole blood into your program. Now I don't have time to go through all of the the bullet points on that past slide, but I do want to show you that when you add the individual components back together again to include cryo relative to a unit of whole blood, because again of the additives that are in each of these units, you end up getting a much more dilute product when you use these in combination. The red cell concentration is down to 29 percent relative to 38 to 50 with whole blood. The platelet count is in the 80,000 range compared to, you know, above 150. And even your coagulation factors are much less, again with reconstituted whole blood in a one-to-one-to-one ratios. And this is very likely one of the potential reasons why we're starting to see in some studies that whole blood is independently associated with improved survival. There's still no RCTs testing this hypothesis yet, but they are actually about to start. Now Christine Leeper at Pittsburgh has published pretty much all of the data evaluating whole blood in children with traumatic injury. And in this 80 patient single center prospective observational study where she enrolled or included children that had either, that had, I'm sorry, more than 40 mls per kilo of all blood products in 24 hours as definition she used for massive transfusion. Then they adjusted for all the potential confounders. And in this adjusted model children who received whole blood as part of their resuscitation, not even the entire amount of their resuscitation, but just a part of it, had significant decreased mortality at both 72 hours and 28 days post-injury. In addition to that data suggesting improved survival in a separate paper published in 2021, she also showed that whole blood compared to components had a faster reduction in base deficit as a measure of shock and improved reduction in INR compared to components. So here you have now in children a small single center study indicating though not only does whole blood compared to components resolve shock quicker and has more efficacious at improving coagulopathy, also an independent association with improved survival. She has also published multiple papers indicating safety of group O whole blood. And when she has compared children who received whole blood compared to components, no increase in transfusion reactions, no increase in organ failure scores, or measures of homolysis in non-group O recipients of whole blood. Now moving on to antifibrinolytics, such as tranexamic acid in adult populations, there have been very large randomized controlled trials indicating improved survival. But there's been more limited data in children. There was this one study on PEDS trauma patients in a combat setting, suggesting that the use of the use of tranexamic acid was associated with reduced mortality. But this is, you know, just one study and it wasn't didn't adjust for all of the potential confounders that you would expect to see or would like to see in a retrospective study like this. It also didn't report on the amount of blood loss or the blood used in this cohort. Now, Christine Leeper and I have just published, again, from the large data set that I mentioned up front, the MATIC data set of over 449 patients. And in this cohort, we studied or compared children that received any antifibrinolytic, whether it be tranexamic acid or Amicar. And we compared those children to those that did not receive any antifibrinolytic. And in this entire cohort, no matter what the etiology of bleeding, the use of an antifibrinolytic was in unadjusted analyses associated with reduced six-hour mortality, reduced six-hour mortality due to bleeding. And then when we adjusted for potential confounders, there was an independent association with reduced mortality at both six hours and 24 hours, again, in those kids that received an antifibrinolytic. And when we did the forest plot, it didn't matter. Both TXA and Amicar both had similar reduced odds ratios. Just when we looked at them individually, the sample size got smaller, and they individually were not statistically significant. In this study, though, there was no difference with 30-day mortality. Now, when it comes to other intravenous hemostatic adjuncts, such as some of the coagulation factor concentrates that are on the market, whether it be recombinant factor 7A, fibrinogen concentrates, or the PCCs, there are multiple PCCs that are available internationally. The one in the States has factors 2, 7, 9, and 10, as well as some protein C and protein S. And the bottom line is, there's no data on any of these factor concentrates to suggest they may be helpful in children with traumatic bleeding. There are some small case series, but none of them that really are worth mentioning, to be fair. I will say there are fibrinogen concentrate versus cryoprecipitate trials that are ongoing. I think they've actually, they're finished. They're analyzing their data now and should be published soon. That will be interesting to see if the additional factors in cryoprecipitate, if they are beneficial compared to a fibrinogen-based concentrate. Now, when it comes to monitoring hemostatic treatment, you know, a big question is should we continue to use empiric ratios or should we use goal-directed therapy with either Rotem or Teg? There are no studies in children evaluating this concept, but there is a single-center RCT out of Denver. It's not on the slide. Again, these are not my slides, but the single-center RCT out of Denver did show in adults when Teg was used to direct resuscitation, there was improved mortality compared to classic coagulation tests such as platelet count, INR, PTT, fibrinogen. So, you know, take-home messages. Improvement in, you know, clinically meaningful definition of critical bleeding is needed. The indicators for initiating massive transfusion are not well described. It's not known if goal-directed protocols are superior to empiric ratios. The benefits of all of the factor concentrates and antifibrinolytics still need prospective evaluation. Hopefully there will be a randomized RCT of tranexamic acid versus placebo called the TIC-TOC trial, starting hopefully within the next year. And while the slide does say a case-by-case careful approach to the management of bleeding patients is recommended, data is starting to gather in both adult and pediatric populations, suggesting that it's reasonable to consider using group O whole blood compared to components, while we also pursue developing randomized controlled trials to study it definitively. With that, that ends the presentation and I will be taking questions during the live session for this part of the program. Thank you.

massive bleeding

life-threatening bleeding

massive transfusion

hemostatic resuscitation

tranexamic acid

blood failure

critically ill children