Dr. Martha Sola-Visner - Red Cell and Platelet Transfusions in the NICU: Update 2022

you Good morning everyone. Welcome to our surgery grand round this morning. I want to thank you all for being here. I have the privilege of introducing our speaker for today, Dr. Martha Salove-Isner. As many of you know, Dr. Salove-Isner is originally from Ecuador and did her medical school training at Catholic University of Santiago de Guayecuil. After medical school, she moved to Florida where she completed her residency in pediatrics and then a fellowship in neonatal and perinatal medicine at Chan's teaching hospital in Guayecuil. She joined the faculty in Florida after her training where she remained there for five years. And then she made her way up the east coast first to Drexel in Philadelphia and now here at Boston Children's Hospital where she has been on faculty since 2007. She is an associate professor of Harvard Medical School and I already knew this but having now looked through her CV I can say with even more confidence is a true clinician and scientist. She is a very accomplished researcher as well as clinician. She is a frequent attending in the neonatal intensive care unit here where we get the opportunity to care for patients along with her team and to learn from her. In addition to that, she is a history of accomplished RO1 funding and is the current PI for two project grants centered around neurodevelopmental consequences of neonatal, anemia, thrombocytopenia and the treatments for such. Her overlapping clinical and research career endeavors have true impact on optimizing the care of our sick neonates. For those reasons we are honored that she agreed to join us today to update us on the current standards for red cell and platelet transfusions and the NICU and we are greatly looking forward to her presentation. So without further delay I'll let her go ahead and get started. Thank you so much. Hey Jamie, thank you so much for the kind introduction and thank you very much for the invitation to present today. It's nice to see so many familiar faces. I'm on call tonight so I'll be there in rounds tomorrow with whoever is on service and hopefully we won't have too much interaction in the middle of the night. But it's a pleasure to be here and talk about transfusions. I'm going to share my screen. Can everybody see my screen? This is actually a very good time to talk about red cell and platelet transfusions in the NICU because the last two years have seen the publication of probably the most impactful trials in this field since you know the last decade. So there is a lot of new information to discuss and I'm going to dedicate the first half of this talk to talk about red cell transfusions and then the second half to talk about platelet transfusions. Most of this will be clinical trials or we'll talk a little bit also about mechanistic implications. So red cell transfusions are among the most common interventions in the NICU. This year is a table from a very recently published paper where we looked at transfusions in newborn infants in seven birth hospitals, hospitals that deliver babies in the United States in the years 2013 to 2016. We looked at the percentage of babies at different gestational ages that receive different blood products. And as you can see here in the smallest gestational age group less than 27 weeks gestation, 70% of them received at least one red cell transfusion, most of them actually more than two or more. In the slightly larger, more mature babies 27 to 28% actually decreased to 44 and then it became increasingly smaller as the gestational age and the maturity of the babies increased. So particularly among preterm infants, this is a very common intervention. So what are we thinking about when we give a red cell transfusion? What is it that we want to accomplish? Well, we give one as you all know because we want to improve tissue oxygen delivery to these babies and we all know that hemoglobin is an important part of the equation that determines the oxygen delivery to the organ centiages together with cardiac output and the oxygen saturation. The problem is that it is not completely clear or it wasn't completely clear how much hemoglobin a baby needs to assure adequate tissue oxygen delivery. It is also important to recognize that this question is different in adults and in neonates. It is different because in adults we need to assure adequate tissue oxygen delivery for our organs to function normally, to simply to function. In babies, we not only need to assure the function of the organs but we need to support the growth of this organs and this is particularly important when it comes to the brain. So this is a figure that shows the growth, the increase in brain weight over time and as you can see here, this growth is extraordinarily fast in the first month of life in the last three months of pregnancy. Then in the first year of life, this is how much it grows. Then in the second year and then it really increases and weight fairly little all the way into adulthood. So the brain grows in this first year of life is extraordinarily rapid and as a consequence, 60% of the total energy consumption of the baby is by the brain and this is compared to 20% of the total energy need of adults. And this is including adults that mostly sit in front of the computer and think all day. So the amount of energy and oxygen that is required by the neonatal brain is substantially higher than that comparatively speaking than that required by the adult brain. So up until 2005, there had been no trials whatsoever comparing different transfusion thresholds in babies. And then in 2005 and 2006, two major trials were published almost back to back. The first one was the so-called Iowa trial by Dr. Ed Bell in 2005. This was a single institution randomized control trial of red cell transfusions in babies that enrolled the total of 100 low birth peri-low birth weight babies. And then a year later, we had the pine study led by Hiroshima, Polani and a chop in 2006. And this was a multi center study in the US and Canada that enrolled 451 patients. And both trials were very similar in that they compared liberal versus restrictive red cell transfusion strategies, both enrolled very low birth with babies. In both of them, the threshold levels varied based on post-little age and based on respiratory support. And both of them reported short term, meaning at the time of naked discharge and long term outcomes. And I won't talk too much about the short term outcomes because there was nothing too excited. But the long term outcomes were very different between the two trials. So University of Iowa reported that long term at a median age of 12 years. So 12 years after these babies had left the NICU, they looked at the long term outcomes. And what they found was that the babies transfused liberally, had worse neurocognitive scores, had smaller brain volumes. And this was particularly driven by the girls. So the liberally transfused girls had the worst outcomes. And therefore, they concluded that a restrictive transfusion approach was better. And then the pine published their long term outcomes. Long term for them was 18 to 21 months. So these babies underwent Bayley scores evaluations at 18 to 21 months. And they found that the restrictive transfusion group had more cognitive delays at correct at this corrected age. So in that a liberal transfusion strategy was probably more appropriate. I will not discuss today for the sake of time what the potential reasons for this is. But obviously it left the neonatologist with a big question mark. What do you do with this? So what we did is designed to additional trials, larger and more conclusive trials to ask the same questions again. So the first was the top, the transfusion of premature trials. This was a very large trial sponsored by the NICHD, the neonatal network. And it was in the United States. And the second one was the ethno. This was a European trial that was mostly run in NICU in Germany. And both of them were published during the pandemic in the year 2020. So the first results to be published came from the ethno, the European trial. And this came out in JAMA in August of 2020. As you can see here, this was a randomized trial that compared liberal and restrictive transfusion trials in extremely low birth weight babies. It involved 36 European NICU mostly in Germany. And as you can see here, there were liberal results, these are hematocrystals. And they were restrictive thresholds, these are hematocrystals too, which both varied, depending on the time after birth, they were higher in the first week, lower in the next two weeks, and lower after 21 days. And they also varied depending on the clinical condition of the baby. But they called being critical, which required higher crits or being non critical. So both of these trials gave a variety of thresholds, depending on the clinical condition, the allocation group and the postnatal age. In total, the ethno study randomized a little over a thousand babies. And approximately 450 were included in the primary outcome analysis in the liberal group. And 478 in the restrictive group. You can see here the composition of the study population. Those were small babies. The median interstational age at birth was 26 weeks. The weight, the birth weight was about 750 grams. And they actually stratified based on birth weight. So they had an even number of babies up off and below 750 grams at birth. All around the biggest difference was that the babies assigned to the liberal group, 81% of them got transfused. And they received the total volume of approximately 40 milliliters. And the babies that were assigned to the restrictive group, only 60% of them got transfused. So 40% avoided transfusion altogether. And they received the total volume through the NICU hospitalization of 19 MLs. And this graph shows that over time the weekly mean hematocrets became different between the restrictive group and the liberal thresholds. This was statistically significant, although I will say clinically the difference was not huge. In terms of the primary outcome, it was a composite outcome of either death or neurodevelopmental impairment measured at 24 months. And when they looked at this results, there was absolutely no difference, not even a hint of a difference between the restrictive and the liberal groups in terms of the primary outcome or in terms of the individual components, death or cognitive deficit or cerebral palsy. All of the outcomes were the same between restrictive and liberal group. They also looked at the outcomes at the time of NICU discharge, all the usual suspects, IVH, PBL, BPD, Neck, ROP, intestinal perforation. And there was absolutely no statistically significant difference in any of these outcomes. So this study concluded that using a restrictive threshold was safe in terms of the outcomes that they had studied. And then in December of 2020, just five months later, the North American study was published. This was a much larger study. It was similar, but significantly larger with 1800 babies, almost twice the size than the previous one and 850 roughly in each group completed the primary outcome data. This was the top transfusion criteria. This is given in hemoglobin. I made a table with hematocrystallate. But it was very similar design in that it the actual hemoglobin or hematocryst recommended for transfusion, very depending on whether the baby was assigned to the liberal arm or the restrictive arm. They also decreased with advancing postlato H, so highest in the first week of life. And they also vary depending on respiratory support. The top study did not have critical state versus non-critical state. It just limited itself to the respiratory support of the baby. The population was very similar. The main gestational age was a little under 26 weeks compared to 26 weeks in the European trial. The birth weight was around 750 grams for all the groups. And they were evenly distributed between males and females. So quite quite similar. And as you can see here, there was also since starting in the first week of life, a progressive difference between the lower and the higher threshold groups in the mean weekly hemoglobin level. So they achieved a statistically significant separation between the groups. And the primary outcome similar to it was also death for neurodevelopmental impairment at 24 months of age, a two years of age. So when they looked at this primary outcome, once again, there was no difference. In fact, the relative risk was one between the restrictive and the liberal and the liberal group in the composite outcome. And looking at the different components of the primary outcome, which was death, which were death or neurodevelopmental impairment defined this cognitive delay, moderate or severe cerebral palsy, severe vision impairment or severe hearing impairment. Again, there were absolutely no differences between these two groups. They also looked at the outcomes at the time of NICU discharge. And once again, there was no difference in any of the usual outcomes survival to discharge PPBR or PIBH, PBL or the letter soup that we're using in NICU or NIC. And the only difference was the number of transfusions per infants. So the higher the babies assigned to the higher hemoglobin thresholds had more transfusions than the babies that were assigned to the lower thresholds. And this was statistically different. So the conclusions from those two studies published in 2020 are clearly that the use of restrictive red cell transfusion thresholds in preterm infants, restrictive within the limits study, not below the limits study. But within the limit study does not increase mortality and is not associated with major neurodevelopmental impairment that at the time of discharge or at 24 months of age. And therefore, it is likely that the recommendation currently should be to use restrictive transfusion thresholds within the limits study in top and in Edna. And I put together tables with a two with a hematocrypt that were studied by both above groups in this slide. So with this, it would seem that all the questions that we had regarding red cell transfusions in babies would be answered. We can use restrictive transfusion thresholds. However, before we say that there are no more questions, let me go for a minute and talk about the physiopathy of the anemia of prematurity, which is why we ended up transfusing these babies. So if we look at the factors that contribute to the anemia of prematurity, the first one is the low-iron resource. It turns out that 80% of the iron that is present in a full-term baby is accreted during the third trimester of pregnancy. This is the time where babies store an iron and are born with iron storages. And this is lost for the premature babies, especially in the most extremely premature. The second reason is to the bottom of the losses, the hemorrhage into the laboratory, that if we draw one CC or blood, this actually represents a substantial percentage in a 500 gram. The next one is the inadequate elevation of aristropoietin in response to anemia, which is a well-known issue in the units. And then finally, we have the rapid growth rate, which expands the blood volume. So you need a fair bit of aristropoietis just to maintain the hemoglobin and to keep up with this rapid expansion of the blood volume in a growing brini. So all these factors contribute to the anemia of prematurity. And how does this anemia of prematurity potentially affect the brain? Well, I already told you that hemoglobin is part of the equation for oxygen delivery. So one mechanism is, of course, decreased oxygen delivery to this growing brain. But another mechanism that is potentially more important is the iron deficiency that accompanies anemia of prematurity in most cases. And it turns out we know a lot about the effects of early life iron deficiency on brain energy metabolism from multiple animal studies, basic science studies and clinical studies. And in a nutshell, this is from a recent review in 2020. If you have early life infancy iron deficiency, most of these studies don't come from pre-tronerionates. They're from full-term babies in the first year of life who had anemia during infancy. So we know is that if you have early life iron deficiency, this leads to mitochondrial dysfunction. This might occur in this function and SAP with less production of ATP, the energy substrate in the brain. And in the presence of less ATP, you develop a simplified neuronal structure. Mostly these are neurons that have less dendrites, less complex dendrites and therefore build less connections. Even if the iron deficiency is then corrected, this remains because it's a structural issue that leads to persistent structural abnormalities at the microscopic level in the brain. And this persistent structural abnormalities manifest themselves as abnormal adult brain function and behaviors. And in fact, again, going back to the iron deficiency literature, the consequences of iron deficiency in early human development and infancy are not severe CP or hearing loss or vision loss or death. They are actually more subtle deficits that present later in life. So studies have shown babies who, people who had iron deficiency in infancy up to 19 years of age or early adulthood and have shown that those people who had iron deficiency in infancy had long-term deficits mostly in executive function, ADHD, difficulties in school and recognition memory. As well as adolescent behavioral problems. So the manifestations of early life iron deficiency are probably not detectable in a Bayley scale at two years of age, but rather much later in adolescents with executive function issues, memory issues and then social and behavioral problems. So the question is, it is very clear that the use of restrictive thresholds is safe in terms of the outcomes that have been measured. Death, severe neurocognitive deficit, cerebral palsy, neuro vision impairment or hearing impairment. That is very clear. However, the clinical studies in the field of iron deficiency and a number of preclinical mouse models where we can make the mice have lobotomy induced anemia in neonatal life and then we can follow them for consequences in adulthood clearly show that anemia in neonatal life might be associated with behavioral dysregulation that becomes more evident with age. I didn't show the mouse studies because of time, but they are emerging now. So there is a school age follow-up of the top infants, the infants that were enrolled in the top North American study that is currently ongoing and that is going to follow this baby up to at five years of age. So I do think that it will be very important to look out for this to make sure that the more severe anemia doesn't go with deficits that are not recognizable this early. So in the meantime, the recommendations are to await the results of tip top that's a school age follow-up of the top infants that I just mentioned. I think in the meantime, we should apply transfusion strategies that are based on the available clinical evidence, which would support the restrictive approach. Again, within the limits studied in top and end not below the limits studied in top and end not that we have no data for the safety. And then I think it's really important to apply whatever strategies we can to maintain as high as possible a hematopryt without increasing transfusions. The late core clamping in the hospital where the babies are born is known to increase your iron storages limit as much as possible the hemorrhage into the laboratory and importantly provide adequate iron supplementation because this might be the most helpful thing we can do for babies. And perhaps in a different setting, this is something to discuss in our babies with short gut syndrome who are not fed and may not be getting iron supplementation. So with this, I'm going to switch gears and change to talk for the remaining of the talk about playlet transfusions in neonates. I think I'm going to leave the questions at the end. So any questions on red cell transfusions we can talk about then neonatal trombocytopene. And bleeding are very common among incubations. In fact, this is the population of patients that has the highest incidence of bleeding and then of any hospital population. Trombocytopene is quite common 18 to 35% of all babies that are admitted to the naked will develop low playlet count at some point during the state. And the incidence is inversely proportional to the gestational age of the baby. So in the smallest, most premature babies up to 70% will have a low playlet count at some point. So the smaller the baby, the more likely to have trombocytopene. The incidence of bleeding, as I said, is also high more than approximately 20% of all infants that have the birth weight less than 1500 grams will have entrepreneurial bleeding that's most commonly IVH. Although only about half of those will be severe or significant. So not surprisingly because of this association of high incidence of trombocytopene and high incidence of bleeding, neonatologist frequently give playlet transfusions to try to prevent bleeding in trombocytopene babies, particularly to those who are most immature. So this let again results from the same study that I mentioned in seven birth hospitals in the US. What we looked at transfusions for different blood products in 2013 to 2016. So this is relatively recent data. And as you can see here 34% of babies less than 27 weeks received at least one playlet transfusion and this percentage decreased rapidly as the babies became older and more mature. We also looked at the pre transfusion playlet counts for those babies and what playlet count were those transfusions ordered. And as you can see here, the median pre transfusion playlet count was somewhere around 70,000 for all babies independently of the gestational age. This was a little bit surprising because this is this is quite a high pre transfusion playlet count. So the question is if we are giving transfusions to increase the playlet count and reduce the risk of bleeding. What is the evidence that playlet transfusions decrease the incidence or the severity of bleeding in pre-term neonates? Up until 2019, the best evidence the only evidence came from one randomized trial, which was published 30 years ago in 1993. This randomized trial had enrolled 152 very low birth weight babies in the first week of life. They only looked in the first week of life. And they randomized them to get a playlet transfusion whenever the playlet count dropped below 150,000. The goal was to normalize, keep a normal playlet count. Or only if the playlet count dropped below 50,000. And what they found in this study was that there was no difference in the frequency or in the severity of intraventricular hemorrhage between the treated and the control group. So this study concluded already in 1993 that transfusing very low birth weight babies with playlet counts greater than 50,000 in the first week of life does not reduce the incidence of severity of IVH. Nevertheless, transfusions continued to be given at greater than 70,000. So the next question was, well, is there a correlation between the degree of thermoselopenia and the risk of bleeding? Is there a good correlation that says that if we increase the playlet count, the risk of bleeding comes down in a parallel way? And this study in 2009 looked at the playlet counts immediately before bleeding in the babies that had minor bleeding, just a little oozing from on to sides, little blood transfusions, the babies that had major bleeding, the babies that had no bleeding. And interestingly, as you can see, there was absolutely no difference in the playlet counts. And these were babies that had severe thermoselopenia that had less than 60,000. And in fact, the median, the mean playlet count was a little higher in the babies that had major bleeding compared to the ones that had minor or the babies that had no bleeding. So this study found no association between the playlet count and the risk of bleeding. This was a study that we published Kate's burger was one of our fellows in 2016. And we looked at over a thousand babies and evaluated the pre-trans, the the native playlet counts prior to the diagnosis of IVH in like I said about a thousand very low birth weight babies. What we found was that compared to babies who didn't have any thermoselopenia in the first week of life, having thermoselopenia was associated with a higher risk of IVH. But the degree of thermoselopenia didn't make any difference. The risk was the same whether you were between 90 and 100,000, whether you were between 70 and 80,000 or 40 and 50,000 or 10 to 30,000. So there is no correlation between the actual playlet count and the risk of bleeding. And the same thing has now been shown in those and in pediatric patients with chemotherapy and use thermoselopenia. So perhaps increasing the playlet count does not necessarily translate into decreasing the bleeding risk, even though it seems so obvious. So if transfusing playlets does not decrease the bleeding risk or may not decrease the bleeding risk. Do playletranspiration decrease neonatal morbidity and mortality in general? There were a number of observational studies that were published over the last 15, 20 years by As and by others that actually showed consistently that babies who received playlet transfusions had a higher relative risk of death than non-transpuse neonates. The number of playlet transfusions actually predicted the mortality rate in this very large study of thermoselopenic neonates. And specifically in neck increase numbers of playlet transfusions were associated with a higher incidence of short-gathering cholestases and with a higher mortality. So in all these observational studies playlet transfusions were correlated with higher morbidity, higher mortality. But the problem is that in any observational study you just cannot tell whether playlet transfusions are a marker of severity of illness. And in fact we look at this and yes the babies that go playlet transfusions were sicker than the ones that they not. Or whether playlet transfusions actually contribute to the morbidity and mortality. And this was the state of the science until 2019 when this landmark study was published. This was the largest most important randomized trial of playlet transfusion thresholds that has been published to death by an occurrence and one stand worth. So this study included neonates with gestational age less than 34 weeks, playlet counts less than 50,000. So they only looked at severe thermoselopenia and they had to have a head ultrasound within the six hours before the randomization that did not show a major IVH. The study was conducted largely in the United Kingdom and in neck is in Europe where they do point of care ultrasound. So the neonatologists could do that and show that there was no major IVH. They excluded infants with major conteneral malformations. Anybody who had a major bleeding in the previous 72 hours was excluded and then they could be enrolled later, including any IVH in this ultrasound in neonatromoselopenia or babies who were more advanced. Babies that were enrolled were randomized to receive playlet transfusions. Either when the playlet count dropped below 50,000 and this was the high threshold, notice that this is lower than what was being used in the United States in the observational study I showed. And the low threshold was less than 25,000 and the primary outcome that they looked at was death or major bleeding within the 28 days following the randomization. This shows the characteristics of the babies in total they had approximately 660 babies enrolled so fairly large study multi center. They had a median weight of 743 728 grams so very small babies in general and the median postnatal age that randomization was about seven or eight days. So they have the babies were randomized in the first week of life and the other half was mostly in the after the first week of life after the period of high risk. The primary outcome was death or major bleeding through trial day number 28 and what they found was that the babies randomized to the high threshold to be transfused more liberally at below 50,000. Had a statistically significantly higher incidence of bleeding or death that the babies that have been randomized to the more restrictive threshold. I'm going to repeat this the babies that God playlet transfusions at a higher limit limit more liberally had more bleeding in death than the babies that were allowed to drop to 25,000 before being transfused. Now this game as a little bit of a shock when this study was published in the New England Journal to the neonatal community. But if you think about it, it was consistent with all the previous observational studies that had shown no effect on bleeding and an association with mortality and morbidity in the neck you. So so this this was actually consistent with what we have been thinking even before the publication of a randomized trial. When they looked at the two separate outcomes individually the two components of this composite death or at least one major bleeding episode through 28 days. Both of them one in the same direction so both death as well as bleeding had a higher incidence in the liberal group compared to the to the restrictive group. And among all the secondary outcomes that they looked at they found that survival with bronchopulmonary dysplasia was also had a the presence of BPD was also higher in the liberal group. So they had more death more bleeding and a higher incidence of BPD among the survivors or all in the same direction or in the more liberal group. This year is a couple of my occurs where I'm showing you the survival without major bleeding and this is in the low threshold group and this isn't the high threshold group. I'm going to show something that is quite interesting to me and it is that the two lines don't diverge immediately after the babies were enrolled. This two lines begin to actually separate about seven to ten days later. So this separation didn't happen immediately after a play the transfusion this baby didn't drop that or bleed that they after or immediately after this this separation was was later and the lines continued to separate. It hasn't doesn't seem to have a collibrated even at 28 days. So I don't know what would have happened if they had continued to look importantly 90% of the babies that were as a sign 93% of the babies that were assigned to the high threshold group got a play the transfusion and only 50% of the ones assigned to the low threshold group got a play the transfusion because half of them avoided them all together by using the lower threshold. So in conclusion the planet to study that's the name of the study which is the highest level of evidence that we have today show higher bleeding and higher mortality among preterm neonates that were randomized to liberal player transfusions compared to those who were randomized to being more restrictive Lee transfused now every study has any patients and so did planet to so the first thing is that the planet to patients had a head ultrasound done within six hours prior to the random. So we don't know by design which threshold is better if you have already an IVH they did not address this question. The second problem was that 39% of the patients in planet to had received one or more play the transfusions prior to the randomization so almost 40% had received a play the transfusion prior to being randomized in the study and isn't clear why so this raised a lot of questions who were those babies why did they get those play the transfusions where they the highest risk babies in the first week of life and that's why we didn't know. We had less babies randomized in the first week of life where those the most critically ill babies that and you know this this generate some skepticism I would say and then the last question perhaps associated is are these restrictive thresholds only beneficial for stable or for low risk babies. So a couple of things since have attempted to answer this question the first thing was naturally a subgroup analysis by the planet to investigators and this was part of the same publication they did separately look at the babies that were less than 28 weeks which have the highest risk of leading highest risk of mortality compared to the babies greater than 28 weeks and as you can see here that the separation of the direction was the same the babies that were transfused with the higher play the count have more mortality or bleeding that those that were transfused at a more restrictive threshold even in this group of highest risk babies and this was statistically significant then they looked at the babies that had been enrolled in the first 72 hours of life that's that's the real highest risk period of leading the babies and those babies had a much smaller numbers of course you know but a very significant difference in their risk of bleeding and mortality. And again in the same direction so it looked like when you separately looked at the smallest babies and the babies that were enrolled in the highest risk period you know even though you may not have had statistical significance because of the numbers the directionality was the same and no matter in what group you looked. So these investigators were one step further and then in December of 2019 they published a subsequent secondary analysis of the data that really wanted to address specifically this question of whether the results are only applicable to the low risk babies and what they did in this really elegant and fascinating analysis is they took all the factors that are thought or known to be associated with higher mortality and higher risk of bleeding. Top of the list of course justational age we know the smallest the baby the higher the risk of bleeding and mortality they looked at postnatal age antinatal cortex steroids you know growth retardation female sex is protective sex is neck previous major bleed and they they incorporated all of these factors in a multi-variant logistic regression model that they developed to predict the baseline risk of major bleeding and mortality of the baby. So every baby was included in this model to come up with the baseline risk of bleeding or mortality based on these factors. So 653 babies that had been in planet two were then ranked based on this predicted model predicted baseline risk and then they were divided into four quartiles the babies at very low risk the most mature older within half-sepsis whatever. Low risk moderate and high risk you know high risk obviously the most returned the earliest the sickest you know whatever based on the model. As you can see here they then show the correlation between the predicted baseline risk of major bleeding and or death and then the actual observed incidents of major bleeding and or death in every one of these four groups and there was a pretty reasonable correlation you know the babies predicted to die more really did die more. Compared to the ones that were predicted to bleed and and die less so then they took each one of those groups which contain about 160 babies and compare the actual incidents of bleeding and death in the babies within the quartile that happened randomized to the restrictive threshold or to the more liberal play lay transition threshold. And as you can see here so the liberals were in the darker bars and the restrictive in the in the lighter bars and as you can see here within each baseline risk quartile the babies that were randomized to the liberal transfusion threshold had a higher incidence of bleeding or death compared to the babies that were randomized to the more restrictive threshold and in fact this difference went from about 4% absolute risk difference. So the highest risk babies seem to benefit the most from the restrictive threshold although every single group seemed to experience this benefit so putting this all together the conclusion was that the neonates with high baseline risk of bleeding or mortality benefited from the low prophylactic play the transmission thresholds as much or more than the lower risk of bleeding. So this obviously led to a flurry of additional questions. First of all is what are the mechanisms mediating the increase in bleeding and or mortality why why to play let's increase bleeding or mortality. One possibility could be that they lead to increase micro vascular thrombosis we do know that adult playlets are hyper reactive if we compare them to neonatal playlets and we in our laboratory about a decade ago showed that if we make neonatal blood trombocytopenic and we reconstituted with adult playlets we generate a little bit of a protombotic phenotype based on primary chemostasis. So in vitro this is true you know it has not been yet demonstrated in vivo but it's certainly a possibility. The second question is is the rapid volume infusion of the playlets contributing to the to the bleeds the babies in planet to receive 15 cc's per kilo typically given over 30 to 60 minutes because that's the typical practice for playlet transfusions and this is the standard in unatology. But if you think about what older children and adults receive in a normal playlet transfusion that is 5 milliliters per kilo. So we currently have a guideline of 10 milliliters per kilo and we suggest when possible to do it over two hours I think this is more important in the more premature babies and this is based on at least one randomized trial that compare playlet recovery in babies transfused with playlets over 30 minutes or over two hours. To see if there was any effect on both transfusion playlet counts and very briefly as you can see here these babies were about 32 weeks at the time of transfusion but 1800 grams this was not a birthday this was the time of transfusion they had a number of diagnosis associated with trombocytopenia but the important thing was that the 30 minute change in playlet count was the same in the short group in the group transfused over 30 minutes and in the group transfused over two hours. So it doesn't seem like transfusing over two hours will decrease your playlet recovery. So the last potential mechanism is that we know now from studies in the last decade that playlets are central mediators of inflammation of in neural responses. And we also know that neonatal playlets are less activated and interact less with neutrophils compared to adult playlets. So is it possible that the infusion of adult playlets into a neonate either triggers or worsens and inflammatory response and that that mediates the problem. So at least in healthy wild type mice in our laboratory if we transfuse 10 day old newborn mice with playlets from an adult mouse from an adult wild type mouse we see two hours after the transfusion and increase in interleuking six levels that last for about four hours and four hours later a significant increase in GCS levels. Now this is this is in mice but at least in mice a neonatal transfusion in newborn mice triggers and an inflammatory response. What about if the mouse has pre existing inflammation so I'm only showing a little bit of data from our studies looking at mice with LPS induced inflammation and either transfuse. You know this is no inflammation PBS and thyroid or LPS and just given a volume infusion of tyroids or LPS and playlets this are washed playlets suspended in entire its buffer and 18 hours after the transfusion you can see that the babies who received LPS and no playlets have normalized almost at I.O.6 levels while the babies that received playlets still have significantly higher levels. We see the same thing for KC this is the same that CXCO1 is a neutral fill chemo attract and and this is really important because it seems to be a cytokine that promotes neutral fill infiltration of the tissues that's also elevated by by playlets even 18 hours after the transfusion and we also see the same elevation with an anti inflammatory I.O.10. Also when we look at I.O.6 levels relative to I.O.10 they are higher in in the LPS in the playlet transfused than the non-playlet transfused. So it seems like at least in newborn mice playlets can induce an inflammatory response and can prolong the inflammatory response of an already inflamed mouse. So while we try to continue our studies and look deeper into what what happens we now are going to be start working with the model of BPD to look at playlet transfusions in BPD. But in the meantime clinically I think it is absolutely essential for NICUs to implement restrictive playlet transfusion guidelines within the limits studied in planet to reduce unnecessary expensive and harmful playlet transfusions. And this is doable we implemented those guidelines like in 2019 this is a paper published by Patty Davenport and Crystal Eman and I about a year ago what we looked at what happened with our playlet transfusions in our NICU after the implementation of guidelines and I will have to thank all of you because I think surgery has been so supportive of our more restrictive approach to playlet transfusions. We don't need to go below 25,000 but I think everybody has been so good about supporting that and we actually decrease our playlet transfusions to a third of what we were seeing before the implementation of guidelines. So our NICU has done a good job at achieving this reduction. With this I thank you so much for your attention I wanted to leave enough time for questions and thank you so much again for for the invitation. I'm going to stop sharing so we can see each other. Wow, Mark that was so much information there. No I'm sorry. That was wonderful. No it's terrific. Very stupidity as a former clouder myself I really enjoyed that. I want to thank you for speaking to us so informatively but also for your leadership. These initiatives both scientifically and in our clinical implementation and protocols as you just demonstrated it has taken fact it's very difficult and very large institution like ours to get everybody to sort of follow along with when we have evidence and under your leadership. We have a lot of questions and I would like to ask you a little bit off the beaten track to ask sort of a big picture of question but let's start with red cells. We have a lot of babies aren't small adults yet 100% of the red cells that we give to babies are adult red cells because you can't donate blood if you're an enemy. We all understand the difference between particularly premature red cells in terms of the hemoglobin distribution oxygen dissociation curve. Is it a good thing or a bad thing that we give anemic babies adult hemoglobin right fetal hemoglobin is designed teleologically for the fetal environment where there's oxygen exchange across the center and not across the alveolar membranes. But it's generally a slow transition over the first several months from fetal hemoglobin to adult hemoglobin. Are we helping our harming babies by giving them adult blood? I don't know that there is a straight answer because given neonatal blood is impossible except in the setting of delayed core clamping or core milking. So in unatology there has been a lot of interest and effort in trying to transfuse gift the baby their blood before their bond and literally wait to clamp the cord or actually milk the cord right now the standard of care is to wait 30 to 60 minutes to clamp the cord to allow blood from the placenta to get back into the baby. So that's an autologous transfusion and that's the only possibility. My personal take about this is that fetal hemoglobin allows to compensate for the very low oxygen concentrations that the baby seen uterus or the uterus is very hypoxic. Once you are born you're relatively hyper-oxic part of the reason why you shut off your epoproduction is because you're suddenly hyper-oxic. So I think the babies should tolerate better the hemoglobin A in that hypoxic. I think the problem in what you cannot have both ways is when you want to restrict the oxygen saturation and there were several trials that in unatology that to prevent ROP try to restrict oxygen saturation without taking into account your hemoglobin concentrations. And then if you try to be restrictive with your hemoglobin and you try to be restrictive with your oxygen I do think you have a recipe for disaster because the high the hemoglobin F and the high hematocryl that you have in uterus sort of allow you to compensate for this hypoclinia in terms of oxygen delivery. And you know we had first of all those studies showed a higher mortality and a higher incidence of neck and neck and we also started having this flurry of anemia driven neck cases or transfusion neck. I didn't even talk about that as a whole other field of transfusion associated neck which probably is more mediated by the degree of anemia. That has not been seen more recently and I think it was probably a consequence of you know not thinking about oxygen saturation and hemoglobin levels as part of the same equation same biology but just separate clinical problems and living in silos research wise. So I do think that we have no option done to transfuse with hemoglobin A is not ideal in terms of oxygen transfer but I think if the babies are equally oxygenated it's less of an issue. But there is no question also that red cells have in new modulatory effects that might be different in babies that is more and more data about that independently of hemoglobin A. So I don't know if if I answered your question but I don't think anybody can say whether it's good or bad it is what it is. Yeah you're actually still like another question so you point out that you know intentional hypoxemia to prevent RLP that doesn't contemplate what happens in the face of anemia. So the hemoglobin level but if we do increase hemoglobin level by transfusing them we're actually giving them adult hemoglobin that does more poorly than hemoglobin F in the hypoxemic environment. So it's better to have a higher total hemoglobin level or higher percentage of hemoglobin F in the intentional hypoxemic that would be interesting experiment to do it and an amuleting you could do that in humans. Those are very important and interesting questions. I don't know but there was certainly a flurry of higher incidence of neck associated with anemia at times where the neonatologist's pendulum had swung towards trying to be restrictive on both ends. I have lots of questions on the stop at the end. Yeah lots of people. Martha thanks so much. That was a great summary a lot of data and certainly stimulating about specifically our surgical patients. I think we discussed these thresholds almost every day on rounds. I have a question about the platelet side of things. It's always fascinating me that we treat a number rather than function. And so what it you know and I think this this concept comes up a lot but is there a better way to go about deciding when to give a child platelets in terms of looking at their actual play function rather than their play. Yeah that's a great question so there have been some small studies looking at platelet function and they have not shown a great deal of correlation with bleeding. I think that's because platelet function changes also over the first 24 days of neonatal life and move slowly towards more adult functionality. But the biggest issue is that platelets in neonates are very hyper reactive. I told you that they just don't respond to the same way that the adults do. They don't aggregate and they don't activate as well. However if you do a test of whole blood he must a says they're actually shorter in babies than they are in in in adults. So they you know the PFA 100 or the bleeding times are shorter in babies than in adults despite the high productivity of the platelets. And this is because babies have a higher hematocrate have a higher MCB but mostly because they have a higher from the liver and factor level and they have the ultra long polymers. So the normal baby plasma is almost a TTP plasma and and this together with the high productivity of the playlists gives you a very unique balance. So what we hypothesized a few years ago is that the way to go would be to have a test of all blood he must a says primary he must a says rather than just a play let on or just a play let function to look at actually the whole combination of this baby's blood. And Emochi Deshmann who was a fellow here and now is a kind of Linsga she published two studies you know with me on that and what we found was that actually this a PFA 100 a CT ADP a clotting test in vitro in response to collagen and ADP correlated very well with the risk of bleeding as well as with a degree of reading while the play. The play they can didn't at all there was no correlation between play the counter bleeding scores we scored them prospectively but this test correlated and if the test improves the bleeding decrease and if the stress worsened the bleeding worsens the problem is that it requires 800 micro liters of blood extra blood. So that's a problem other groups have taken a different approach and European group that came up with a model developed a model to to predict the bleeding risk in the week following. A certain time so entering factors like ventilator gestational age postnatal age or this clinical factors into a calculator and coming up with a risk of bleeding for for the baby so this this is something that we currently do for BKB we do for. You know different sexist calculators there are a lot of neonatal online calculators that people used to predict the risk of sex is determined whether you get anti-biotics or not determined whether you get steroids or not so the same thing applicable to to bleeding and that's the idea and I would say that those are the main approaches. I'm okay currently has a funded study where she's looking at the IPF as a predictor of bleeding because younger playlets are more active so maybe it's not the total player count is the actual mass of more active and functional playlets and that has been shown in children with IP and in adults to work. So she's currently running a study and we're part of it looking at IPF and bleeding and you know we enroll all our termoceratomy babies that the parents allow us and follow. Bleeding scores and IPFs so there you're absolutely right I think the playlet function by itself is probably as poor as the player count but I think when you begin to think about the totality either the he mustatic totality with the test of he mustaches or you look at the clinical totality with the predictor score then I think your your prediction ability is going to be better so I think this is where we're heading. Hey Martha I have a question some days. Yes. So you you suggest that transfusing playlets does have some potential detrimental effects. So do you think that there will ever be a study to compare transfusing less than 25 versus no transfusion because of that less than 25 versus no transfusion. So the study of no prophylactic transfusion was done in adults and the unfortunately the non transfused ever had a higher incidence of intracranial bleeding so I don't think that no prophet boxes. So I think that there is a level of playlets that you need to simply maintain your chemostasis and their normal circumstances you need a small amount of playlets to block whatever's continuous holes you have and this level might be a little higher in babies because babies have such active growth and angiogenesis that the blood vessels are less stable less covered by basement membrane so probably a little more. So probably a little more prone to bleeding that level is about 5000 in adults and we know that from studies and it's unclear what it is in babies but we know is it's not greater than 25,000 just based on the clinical studies you know so we have a study that we have brought up to the new late on network together with the rabbit catalytic emory where we want to look at 20,000 as a result in babies and because I do think that we could go there and see what we can do. So I think that we could go there especially after the first week of life where the risk is so this is this is one of the questions that we are that we're asking we also want to enroll prenatally and focus on the extreme premature that the 23 to 27 weeks and retest 25 or so 15 the first week of life and then drop them to 20,000 to ask the question of can we go lower. So the answer is yes I think you can go lower probably I don't know if in every patient but probably yes in in some patients and many neonates and the issue is right now we don't have to safety data for that and the full term babies were successful or you know the animal we just that is no studies everything I showed you was was premature babies so this is a little bit of a dark hole and we extrapolate but I do think that we can do that. But I do think that is a that is room for lower I don't think zero profit access is a good idea. Well I think we're going on this for a long time these questions this is fascinating important topic for all of us and fortunately we're past time I really want to thank you for not only for the lecture but for all you've done for our education and for leading this whole arena for us at Boston Children's for a very long time now we're very appreciative. Thank you so much Stephen thank you very much for inviting me I really appreciate that. I'll see you guys.