Dr. Patricia Davenport - New Insights into Neonatal Platelet Transfusions: From the Bench to the Bedside

All right. Good morning, everyone. I have the distinct honor and pleasure today to introduce Dr. Patricia Davenport, who is one of our neonatologists here to come speak to us at Grand Rounds. Dr. Davenport was born in Cincinnati, Ohio, which is where she earned both her undergraduate and her medical degrees at the University of Cincinnati. Then she moved to New York City and did her residency at in Pediatrics at the New York University, and then she joined us here at Children's for her neonatology fellowship. We're very fortunate that she chose to stay here as a faculty neonatologist. I know that she personally, I know she spends a great deal of time teaching and mentoring younger trainees, including me and teaching me a lot of neonatology, which is super helpful. She also is leading a number of our institutions DEI efforts as well. Having worked with her a fair amount over the past year, I know that as a clinician, she's calm, thoughtful, and really highly collaborative. It's always a pleasure to see that she's on service or called by her at night. You know that it's going to be a good thoughtful and really great care for the patient. Her research efforts, though, have been particularly impressive. The main focus of her work is understanding the importance of platelets and the effects of platelet transfusions and critically ill newborn. She's earned an NIH Career Development Award for her work and has been invited to speak at a number of institutions around the country. She's used her expertise in this area to develop a number of clinical practice guidelines that have been helpful for us around platelet transfusion and unforectionated heparantetration and newborns. It's my pleasure and I'm very grateful that she accepted my invitation to come speak to us today about her work. Thank you, Dr. Devin for it. All right. Thank you, Brian. Very lovely introduction. It's such a joy to come talk to you guys today and I always just want to start off by saying all these amazing accolades Brian mentioned are really because of great mentorship and I know Dr. Solvissner is here somewhere today. But really, there's a thank you slide at the end, but for any trainees in the audience, your mentor is really the most important selection. You make maybe a side from your spouse, so it's a good one. All right. So like Brian said, I'm going to talk to you today about some of the work I've been doing from my fellowship and then now also as faculty about neonatal platelet transfusions and it's truly going to go from the bench to the bedside. So we'll start with some human clinical studies and then stand with me as we go through some cell culture and some mouse experiments and I promise I'll bring us back to babies at the end. So a brief background with some of you might be familiar with that come into the NICU, but thrombocytopenia is very common in the NICU. 18 to 35% of all neonates that we admit have thrombocytopenia at some time during their course. And the incidence of this actually goes up to about 70% in those babies who have a birth weight less than a thousand grams. And there's a lot of causes. There's inflammatory causes, which are some of the most common that we see. So we think about sepsis and the neckeratizing and our colitis. But there's also a lot of non-inflammatory causes. So genetic syndromes and interuterine growth restriction are also very common ideologies. Maybe not so seen in our NICU here, but if you go to a level three NICU, you'll see a little bit more of especially the growth restriction. And then unfortunately, these same babies have a very high incidence of bleeding, specifically intercranial hemorrhage, which makes all of us very nervous when we think about them having low platelet counts. And as a result, we've transfused these babies very liberally really over the past decades. So this was a study that one of our collaborators did recently looking at babies based on their gestational age, but then also in their week of life. And then I'm highlighting the 50th percentile here, but this is the platelet count that prompted transfusions. So you can see we're transfusing babies for platelets counts in the 50s, 60s, 70,000s on a regular basis. And it's really because we're hoping that by giving this platelet transfusion and increasing the platelet count, we're going to decrease their risk of bleeding. But unfortunately, that hasn't really panned out. We actually know that the platelet count is a poor predictor of bleeding, and it's actually even a specifically poor predictor of intercranial hemorrhage. And there's been many retrospective and observational studies that have come out that really show that platelet transfusions are associated with morbidity and mortality. But of course, the age-old question existed, which is well as really just from the saidapena and the need for transfusion actually just both markers of severity of illness. And can you really say that platelet transfusions cause harm? But in 2019, which is around the time that I started my fellowship, this trial finally was published. It's a Planet 2 trial. And this study randomized babies less than 34 weeks who had a platelet count less than 50,000. And they either were randomized to receive a transfusion at a high threshold of 50,000 or a low threshold of 25,000. And their primary outcome was death or major bleeding within 28 days after randomization. This is their table one. I'm just highlighting here. You can see these babies were very premature, so the average gestational age of birth was 26 weeks, and they all weighed around 700 to 750 grams. And also, of note here, I didn't highlight it, but about half over half were receiving treatment for infection at the time of enrollment. And strikingly, I think to a lot of people in the field who weren't following this literature, what they actually found was an increased incidence of death or major bleeding in the infants that were randomized to the high threshold group compared to the low threshold group. And this, you can see, is over the 28 days with a slow separation between the groups on this cap and my occurs. They also found on secondary analysis that there's a higher incidence of bronchopalmerid dysplasia or chronic lung disease of pre-matured in the babies, again, who were randomized to the high threshold group. And this obviously was shocking to a lot of people in the field and made us all really worried, are we causing harm? Are we doing the wrong thing? And we'll talk more about practice change later. But this was shocking to a lot of people. The same study actually followed these babies all the way out to their two-year outcomes, really looking at neurodevelopment and seeing if these liberal platelet transfusions actually continue to have deleterious effects as these babies get older. And one thing I will say, they had amazing fall up in this study, like 98 percent, with a caveat, though, of it was a very heterogeneous, heterogeneous, again, always a word, collection of assessments. So if anyone's familiar with two-year outcome fall up, we like to use the Bailey score. And this was a combination of Bailies and Griffith scores and surveys. And it was very eclectic, essentially. And so it was great and they got good outcomes and good numbers. But I think that's one caveat of this study. But nevertheless, what they did find is that their primary outcome of death, or they called it an unfavorable neurodevelopmental outcome, was fire, again, in the high threshold group compared to the low threshold group even at two years. And really striking to us is they also found that the babies in the high threshold group had a higher incidence of still needing oxygen two years after NICU discharge. So it's like a long time to still have some sort of oxygen requirement or chronic lung disease of prematurely. We had a really interesting opportunity to contribute to this data because we had access to this very large data set from the peanut trial, which is a separate trial and unitology looking at the use of Epo for IVH. But what they have is an incredibly rich data set of many preterm babies, and they all had Bailey scores at two years. And they have very detailed transfusion information. So obviously it's retrospective, but what we were able to do is look back and then look at the incidence of even like receiving a transfusion, even one or no, and looking at the incidence of death or severe NDI in these babies. And we again found that even receiving one platelet transfusion during your hospital course was associated with an increased risk of death or severe NDI with a components of this also trending in the same direction. And for each additional platelet transfusion, we could show here that your risk of death or severe NDI increases. So if you look at this figure here, it's showing death or severe neurodevelopmental impairment, and then death or moderate to severe, because even moderate NDI is significant for families, especially. And because each of these babies had Bailey scores, we were able to actually have a little more granular data on what components of neurodevelopment were affected. And while you can see here platelets no versus yes, so this is just saying like, did you ever get transfused versus never? You can see that all components were decreased in those who did receive a platelet transfusion, but actually the motor component was the strongest affected. And obviously in this study, it's retrospective. It's definitely confounded by many things, which we tried to account for in our analysis. But this was our way of hopefully just trying to give a little bit more clarity on to what components of neurodevelopment seem to be affected. So just briefly from the clinical trials and clinical studies that have come out, I think we now know that liberal transfusion practices and pretermunates have worse short and long term outcomes. So higher mortality and major bleeding, higher incidence of BPD, higher incidence of poor neurodevelopment outcomes at two years, as well as increased need for oxygen support at two years. So of course, the question is why, like what's causing this? What are the mechanisms underlying all this farm? And this has really been the focus of the work that I've been doing over the past few years. So along the way from donor to baby, there's actually many places along the path that this harm could be kind of occurring, right? So there's donor, donor factors, there's the way we process things in the blood bank. We'll talk a little bit more about this yellow box later in the talk. There's many different ways to process platelets. But what I'm going to focus on right now is the fact that every single platelet transfusion that we give, of course, comes from an adult donor. So we're giving babies adult platelets and we'll talk about why that's important in a second. But really click on it a pause and just do a brief platelet biology 101 because we're going to talk about some things about platelets that I know when I joined Martha's lab, good old Wikipedia taught me a few things that I need to catch up on. So platelets circulate in a resting state. So they're discoid cells and they circulate happily and just freely in the bloodstream. And then they can get activated by a number of different agonists that either express from the end of the helium or express on different cells or from the presence of bacteria. And then they undergo this dramatic conformational change and they turn into these kind of more sticky looking cells. And platelets inside contain many different granules. There's alpha granules and dense granules and lysosomes. There's no nuclei if you remember that from med school. But nonetheless, they're very active and interactive cells. And when they get activated, it's incredibly complicated. Many receptors are expressed on the surface. The contents of their granules are released out into the surrounding serum or plasma. And then they're able to interact with many different things, the endoseleum, like I mentioned. But also they're able to interact with different cells. And here is a leukocyte, which we're going to talk a little bit more about. But I wanted to highlight here this interaction, which is Pselectin, which is expressed on the platelet surface after activation, is how platelets are able to interact specifically with neutrophils in monocytes. And it's through binding to the PSGL1 receptor on the cell surface. We're going to talk a little bit more about Pselectin as we go through this talk. So I just wanted to highlight that here. And with that biology in mind, we know that neonatal platelets are different from adult platelets. So they look the same. These are EMs of both neonatal and adult platelets. They have similar volume and structure, but they have very different surface receptor expression with activation. So when you activate a neonatal platelet, it expresses either less receptors or sometimes the downstream mechanism of something binding to that receptor is different in the neonatal platelet compared to the adult. And as a result, they have very different reactivity to different agonists. So you can isolate platelets and activate them in vitro and neonatal platelets are very hypo-reactive compared to adult platelets. So they react, they basically activate kind of less than an adult platelet would. And this was one study you just comparing the surface expression after activation of basically phybrinogen binding. So aggregation of platelets and Pselectin exposure. So again, when you activate a platelet, they're going to express Pselectin on the surface. And what you can see here on the bottom, these are different agents that we can use to activate platelets in vitro. And then white are preterm platelets, full term platelets in red and blue are adult platelets. And you can see for the majority of these activating agents, you can see that preterm and term babies activate or express Pselectin on the surface much less than adults do. And this is just a very important developmental difference between neonatal and adult platelets. And it does normalize over time, usually over the first kind of like seven to 12, 14 days. But actually, there's interesting studies now showing that some changes in development in the platelet actually can persist even as long as to like two years old. So it really takes a long time for a neonatal or a childhood platelet to truly become the platelets that circulate in adults. So for a long time, this thought that platelets were hyper reactive, people thought was a developmental deficiency in babies and actually led to us giving more and more transusions. Because we thought not only do they have low platelet counts, they also don't work, so we must give them more. But what we've realized now over the past few decades is that babies actually exist in this really unique balance called developmental hemostasis. And they do have hyper reactive platelets and they also have decreased levels of many of the pro-coagulation factors. But this is actually really nicely balanced by other things that are abnormal in baby's blood. So if you have ever looked at a baby's CBC, they always have very high hematocrit and high MCVs. They also have high Von Willough Grant factor levels. They have ultra long Von Willough Grant polymers. And these factors actually help these hyper reactive platelets still adhere and interact with the endothelium or the vascular wall. And in addition, they also have decreased anti-coagulation factors. So a combination of all of this actually leads babies to have intact hemostasis. And not only just intact, but maybe even a little bit more robust than adults. So this is a study looking at adult and neonatal blood and looking at closer times using the PFA 100. So it's basically an assessment of their ability to clot. And you can see here whether it's activated with epinephrine or ADP, babies in red have actually faster closer times. So it's not only are they able to clot, they're actually able to do it even maybe slightly bit faster than adults. So really the message there is that all these kind of differences in the neonatal hemostatic system must be there for reason and they actually are able to clot normally. And Arbolab was one of the first labs to actually prove that a proof of concept that when you give this basically different platelet into a neonatal system, you can alter the neonatal physiology. So this is done by a postdoc in Martha's lab prior to my joining. And what they did is they took cord blood from placentas and through some centrifugation steps basically created a thrombocytopenic sample. So they were able to remove the platelets while leaving all the other blood binds the same. And then you either isolated cord blood platelets or healthy adult platelets from a donor and then transfuse them into this baby thrombocytopenic blood. And in this mixing study what they essentially did show is that if you give cord blood, a peripheral blood or an adult blood transfusion, you actually had a shorter closer time compared to the cord blood that received a baby transfusion. And again, it's just the first time showing that when you give this transfusion of a different cell, you actually might be able to change something in the baby or in this case, the test tube of blood. But the important thing was that this closer time that it shortened to is actually at a level that's associated with some thrombotic disease in adults at times. So it does seem to be, it could be clinically significant. Okay, so with all that being said, moving into what we've been researching, it's now become clear that platelets are more than hemostatic cells. And the majority of research looking at the differences between adult and neonatal platelets has really focused on the hemostatic functions. But what we're really interested actually is in the abundance of immune functions that platelets have. So they release cytokines and chemokines. They interact with immune cells, as I've already told you, through Pselectin surface expression. And they also respond to infection. I don't know if nor tumor biology and vascular growth, which are also very important. But what we're going to really focus on is this immune function of platelets and really how is the neonatal platelet in the adult platelet different in terms of their immune potential and their immune function. Because what we wonder is does the neonate also live in a unique developmental immune state? And is this hyperreactive platelet? Because we know that they are actually very active immune cells, also part of a balance that exists in a preterm baby. And potentially when we transfuse these more hyperreactive adult platelets into this system, are we altering a balance that the neonate actually exists in that might be important? And that's been this question that we study. So of course the first question is just, are there actually differences in immune functions between neonatal and adult platelets? And this question hadn't been asked before. And so to answer this, we partnered with Chris Tom, who is a basic scientist and neonatologist set shop. And what we did is we isolated platelets from healthy term cord blood, so we collected from the placenta. And we washed the platelets and we compared it to washed platelets from just healthy adult volunteers. And we had eight and six. And this is just the diagram showing our processing. And then this is our little table one here. Again, they're all term babies. And then we have a range of years in our adult volunteers. And then what we did is for the first time, we used quantitative label-free mass spectrometry to do the, to compare the proteome of these two cells to each other. And so we identified around 4,000 proteins. And there had been previous studies of adult platelet proteomes. And we did capture 95% of the proteins that were previously described. But then we also expanded the number identified by 46%. I'm not the person to ask that something about this method with the proteome is new and exciting and interesting. And if anyone wants to know more about it, I'm happy to kind of refer you to the people who actually know more about it. But that is why it was able to actually expand the number of proteins that we identified. But interesting to us on principle component analysis, we did see that neonatal and adult platelets do cluster very separately from each other. And we found about 172 proteins that were up regulated in adults and 162 that were up in neonates. But what was really interesting is on gene ontology analysis. We found in the adult platelet proteome, they're really enriched for almost all immune, immune proteins. So defense response, immune response, regulation of the immune response. While the neonatal platelets are really enriched primarily in proteins related to metabolism. So it really does seem like the adult platelet is a little more pro-inflammatory or potentially like immune active compared to the neonatal platelet. We pulled out a few select proteins that we are interested in, one of which is beta 2M, which we'll talk a little bit more about as the talk goes on. But beta 2M is a protein that is initially identified as like a molecular shaperone protein for many different receptors, including like the MHC class one receptor. But now it's actually been found to be contained in platelets and it is released and we'll talk more about it later, but it can actually combine to monocytes and macrophages and alter their inflammatory response. And we did find that this is in higher concentrations in the adult platelet compared to the neonatal platelet. And then we also found CXCl12, which is another pro-inflammatory protein up-regulated in adults. And then the last one we're going to highlight here is PDGF beta, which again will come up later in the talk, which is a vascular growth factor contained in very high quantities in platelets. But again here we see there's still more in the adult platelet compared to the neonatal platelet. And we confirmed these on western blocks, which should be exact same trends. So that's all really great, but what about function, not just content? It's nice to know that they contain more, but do they actually interact more with cells or change the way that the neonatal immune system functions. And so I was I was mentioning before, there's been a lot of data looking at how platelets interact with the immune system over the past few decades. And one of the cells that platelets do interact with a lot are monocytes. And there's been a lot of data showing that through the p-selected and surface expression on platelets and through their release of rantys, which is a chemokine that they contain. And again our friend beta2m that I was just mentioning that we found in our proteome. Through the release of these and the expression of this platelets are able to interact with and alter the monocyte inflammatory response. And what people have shown in the past is that they can alter the release of pro-inflammatory cytokines from monocytes, but they can also change the migratory potential of the cells or kind of cause them to have increased extravacation out of the vasculature. And what we are interested in in this is people have looked at this, but only in the context of adult cells. No one's ever looked at it and really identified if you give a neonatal cell versus an adult or a neonatal platelet instead of an adult platelet, does the monocyte respond differently? And that was the question that we wanted to ask. So to do this, we first decided to investigate this part of the diagram. Do adult and neonatal platelets express different amounts of p-selectin, release different amounts of rantys, or release different amounts of beta2m to see if they would potentially affect the monocyte response. So to do this, we collected adult blood and cord blood. We washed the platelets. We activated them with a small low dose of thrombin. And then what we measured is their surface expression of p-selectin, the rantys content, the beta2m content. In TGF beta, we also measured, without going into too much detail, beta2m and TGF beta compete for the same receptor. So we decided to also measure that. And interesting what we found is p-selectin and resting platelets as low as you would expect. But when you activate them, we did see that adult platelets do express more p-selectin on the surface. And again, that's been published before. So this was nice just to confirm something that we knew to be true. And then for the first time though, we did show that resting platelets do express or release a little beta2m out into the culture. But when activated, adult platelets release more. And as we know from our proteome, they're releasing more because they actually contain more beta2m. And so this was the first time showing that there is this developmental difference upon activation between the beta2m release. And we saw no difference though in rantys, which was the other side of kind of importance or TGF beta between the unatal or adult platelets. So then of course we wanted to ask with those differences in release date and expression, do they actually change the monocyte response? And so again, we took the same cells, but this time we also isolated monocytes from the cord blood and activated the platelets from both adults and unates. And then we used a control monocyte and then we exposed it to either activated neonatal platelets, so kind of simulating a transfusion of neonatal platelets or adult platelets, which is actually what happens in the unit. And then what we looked at here to look at the cytokine response is we measured IL-8 and MCP-1 levels, which are two pro-inflammatory cytokines released from monocytes. And then we measured CCR-2 and CCR-5 on their surface, which when these receptors are expressed on the monocyte surface, they're more pro-migratory, or they're moving to areas of inflammation. So first, looking at the inflammatory cytokines, we saw that the addition of platelets in general does increase the monocytes release of IL-8 and MCP-1. And while there's like little trends to the adult platelets releasing more than the neonatal platelets, it's not significant. So it seems like the cytokine response is kind of just platelet dependent, maybe not so much based on developmental differences. But interesting what we found is the CCR-2 and CCR-5 expression was significantly higher when you exposed the monocytes to adult platelets instead of neonatal platelets, suggesting that when you give these more active adult platelets to the neonatal monocytes, that monocyte now is primed essentially to extravacate more readily than if it was exposed to a neonatal platelet. And importantly, CCR-2 is the receptor for MCP-1, and CCR-5 is the receptor for ranthes. So two of the cytokines we've been measuring. Okay, so of course then we want to ask, well what's responsible for these differences? And of course we think about p-selected and beta-2M because these are two platelet factors we found to be different between neonatal and adult platelets. So through a series of experiments, we blocked p-selected by blocking the receptor. And it didn't change anything with the cytokines, but we did see that the CCR-5 surface expression on the monocyte is lower when the PSG-L-1 block is there, suggesting that that upregulation of those surface receptors does seem to be p-selected mediated. Blocking beta-2M has been very difficult. We're still on that journey in the lab, but in lieu of being able to do that, what we decided to do is just take neonatal monocytes and expose them to different doses of recombinant beta-2M and see if we change either surface receptor expression or cytokine release. And what we found is if you give recombinant beta-2M, you don't change anything about the CCR-2 or CCR-5 expression, but you do alter the ILE in MCP-1 release. So suggesting that there's kind of two different pathways regulating these monocyte responses, one which seems to be p-selected independent, which results in increased surface expression of receptors, and the other which seems to be beta-2M dependent, which might result inside a kind release. So a quick summary from our in vitro studies from the proteomes to the cell culture really proteins that are related to inflammatory and immune responses are more abundant in adult compared to neonatal platelets. And we know that activated adult platelets express more p-selected on their surface, and they also release more beta-2M than neonatal platelets, and that exposure of neonatal monocytes to activated adult platelets results in increased CCR-2 and CCR-5 surface expression, and that this seems to potentially be p-selected mediated. And exposure to platelets also does induce increased MCP-1 in ILE-8 release, and this might be due to the beta-2M released by the platelets. And this is kind of, we'll follow this figure through the talk as we add some more pieces to it, but this is what we found after these in vitro studies. Okay, so now we're going to go to the mice, because of course, what we wanted to ask is, well, if you give a platelet transfusion actually into a living organism, you actually elicit an inflammatory response. And so to do this, we developed an experiment where we took adult donors, because we really want to mimic that developmental mismatch that happens in the unit, and these adult donors would either wild types of normal mice or their p-selected knockout mice, so they have no p-selected on the platelet surface, trying to mimic what a neonatal platelet might look like that has less p-selected. We wash the blood and create a washed platelet product. And of course, we have to make sure that what we're transfusing is high quality, so we make sure the platelets are resting, but they're also active at a bowl, because we don't want to transuse like an inert cell. We want it to still have biological activity. And then we also show that these platelets are transfused mouse normally over time, and they're not just rapidly cleared from the circulation. So once we establish this model, what we do is we transuse these platelets in this experiment into p-10 pups, and we're using a p-10 pup that has rhombuside epineus, so there's CMPL knockout mice, which means they lack the receptor that binds to TPO, which is released from the liver, and that's how our bodies make platelets, so they don't have that receptor, and so they don't make platelets. So they have platelet counts around about 10 to 20 percent of normal. So into these mice, we give a platelet transfusion, and then two or four hours later, we bleed the mice, and we obtain the plasma to measure cytokines, and we'll talk about nets in a second, and then by flow, we reconstitute the cellular fraction, and we look at platelet-neutral aggregates and platelet monocyte aggregates, which are exactly what they sound like, just how many platelets are interacting with these different immune cells. So first, looking at the monocytes, this is our percent of platelet monocyte aggregates, so tyros is a sham, because obviously the baby mice who don't get a platelet transfusion still need to undergo that procedure, so we do just give them a sham transfusion of a buffer. So you can see if you give a wild-type platelet transfusion to these mice, you do increase the amount of platelet monocyte aggregates in circulation, and when you give a p-selected knockout platelet transfusion, the levels are lower, so suggesting that as we know, p-selected is very important for this interaction between these cells. And then we looked at a panel of monocyte cytokines, which hopefully are a little familiar to you now. KC in the mouse is IL-8, they don't have IL-8, so they have an analog called KC, MCP1, and RANTE's, and what we did see is that the transfusion of platelets, the wild-type platelets, do increase the levels of all three of these cytokines in the in the circulation. So KC's elevated, MCP1's elevated, and RANTE's is elevated. But when you transfuse knockout platelets, they don't change at all. And that's actually not surprising, right? Given in the cell culture, we also showed that the IL-8 and MCP1 release had no difference when we blocked p-selected, so it really seems like that's not important in this pathway. Next we look at NETS, and I don't know if everyone's familiar with them, but they're pretty interesting. These are web-like filaments of DNA, that's basically a spulse from the neutrophil. The neutrophil doesn't always die actually in this process, which is interesting, but these web-like filaments of DNA are studied with different proteins, and what they basically do is they try to trap and kill pathogens in the circulation, which could be great, but as you can imagine, having sticky filaments of DNA in your circulation is also probably not beneficial, and so they can have harmful effects, increasing platelet aggregation and micro thrombosis and tissue damage, and this is just a study by a collaborative artist showing you can actually block NETS formation, improve survival, and some mouth models. And so we wanted to measure the use and see if platelet transfusions alter NETS formation and circulation. Before I get there though, I also just wanted to share one interesting piece of information, which is that right after birth babies actually cannot make NETS, and that's because the placenta creates insocrates and NET inhibitory factor, because as you can imagine, making NETS in utero is probably not going to be beneficial for the baby, but it's usually cleared in the first few days after birth, but it's a really interesting field of research if anyone's interested into this inhibitory factor, but we always make sure in our mouse models or in the human studies we do to think about timing of when we're going to investigate NETS, because we don't want that factor to be present. All right, so when we looked, so it's talking about neutrophils for the first time in NETS, so when we looked at the platelet neutrophil aggregates, you can see they also increased with wild type transfusions, and these are also decreased when we give a piece like the knockout transfusion, so again, just suggesting that that is so important for this interaction between the cells. And we did see that when you give a transfusion, you do increase the NETS in circulation, and when you give a knockout transfusion, it actually doesn't make a difference. So really, again, the first time just showing that platelet transfusion is increasing NET formation. This slide is just a quick blurb to say that we also have a lot of other ongoing studies in neonatal-murray models, if anyone's interested in knowing more or collaborating, we're always happy to talk, but as you guys know from the unit, most of our babies have some sort of inflammation when they're from beside a penis. So we are using a non-lethal LPS model of inflammation, and in red here you can see that those who receive a platelet transfusion, compared to those that just have LPS, have worsened weight loss, and really exists on a different weight trajectory for the remaining of their life, essentially here. We also have a lethal E. coli sepsis model, and in red here you can see that those who get E. coli and a platelet transfusion have increased mortality, and then we also are working on a hypooxyamodil of bronchoplomnery dysplasia to see how platelet transfusions alter neonatal chronic lung disease. And again, this is just a little blurb, and hopefully more to come on some of these exciting studies. So just to add to our graphic here from what we learned from the mouse, so we found that transfusion of adult platelets increased again, some of the same cytokine levels that we saw in our in vitro work, but this time we've added now rantes here as another cytokine that we found increased. We know it increases interactions between monocytes, and now we've also added the neutrophil, and through these interactions with the neutrophil, we also found that it does increase the formation of neutrophil extracellular traps in the systemic circulation. And when we transfuse these p-selected knockout platelets, which again are a little more characteristic of a baby platelet, it didn't seem to change anything with our cytokine response. It does decrease the interactions between platelets and leukocytes, but it didn't change anything with the nets, which is like an interesting thing that we're still thinking about. And then we have studies ongoing, the one thing we didn't talk about here is beta 2M, and we do have the beta 2M knockout mice, and we would like to transfuse those platelets and see if it has a different pathway effect. Okay, you made it through all the basic students, and now we're back to babies, because of course what we really care about is babies. So what's happening in the babies admitted to the NICU? And this was a really beautiful study that I had the opportunity to be part of with Martha during my fellowship. And what we did is we took babies that were thrombocytopenic, both in RNICU and across the street at Beth Israel. Obviously we collected detailed information about the babies. And what we did is any baby that was about to get a clinically ordered platelet transfusion, we collected a blood sample for plasma within an hour before, and then two or four hours after the transfusion. And in that plasma, we measured our friends, panel of inflammatory cytokines, and we also measured net formation to see if these are actually also increasing in babies. One part of the study that was really unique is that we also collected a sample of the platelet transfusion itself. And so we were able to get a lot of information about the unit, the donor, and different factors in the blood bank, which we'll talk about. But in the unit itself, we also measured the level of cytokines. We also measured extracellular vesicles. Not going to talk about it too much, but platelets are really one of the main contributors of EVs into the circulation. And we also measured mitochondrial DNA, really just trying to characterize what is the product that we're giving to these babies. So this is our table one. It's very, the majority of babies came from children's, which you can see because our average birth weight here is 2.5 kilos. So it really represents our larger babies that we admit here in the just average gestational age of 36 weeks. That being said, we did have some very preterm babies that came from Beth Israel. And you can see their age at transfusion. About half were being treated for culture positive sepsis at the time of enrollment in the study. And we'll talk a little bit more about these pre-transfusion cytokine levels. And this is the platelet increment, so showing how much the platelet count increased after transfusion. So the first thing we just looked at is this large panel of cytokines that we measured pre and post-transfusion within each baby. And the first thing we just looked at is the full change. So comparing the post-transfusion level to the pre-transfusion level within every single baby. And so if it goes to the right, that means the level increased with transfusion. And if it goes to the left, it means the level decreased with transfusion. And what you can see is that two hours after transfusion, we saw very high increases in PDGF alpha and PDGF beta. And if you remember, we talked about PDGF beta way earlier in the proteome. We know it's contained in higher amounts in adult platelets. So maybe it's not surprising that we see it pop up here. And again, this is just a growth factor that actually does seem to have some important implications into airway remodeling. And maybe this could have some downstream effects on chronic lung disease in these babies. And then at four hours post-transfusion, we still see increases in PDGF alpha and beta. But now we see this very big increase in rantes, which if you remember, was one of the platelet released factors that we know can alter monocyte responses. And so again, it's a chemokine that enhances monocyte and neutrophil activation and accumulation. And it's been shown in multiple models to do this in mice. But so we are interested, obviously with these factors that we found increased. But one really important question is how much of the cytokine did we actually transfuse into the baby in the unit? And so we have that data and we had sent it. And so what these crafts are showing you here is the level of the cytokine pre-transfusion in all the babies. And then what the concentration of the cytokine was in the unit. And you can see for all we included Vegef here too, it was almost statistically significant, we were interested in it. So you can see for all of these, the unit itself does carry a good amount of the cytokine itself. So what we wanted to ask is then did we simply just give these babies a bolus essentially of cytokines? Or is there actually something that biologically is happening when these adult cells are in circulation? So to answer this, we could actually calculate we know what was observed. We know the value of the post transfusion cytokine value in the baby. But we also could calculate what we would have expected it to be because we know their pre-transfusion value, we know their blood volume, we know how much we gave in the unit. So if there's math, you could figure out what the expected concentration should have been. And what you can see is for PTGF alpha beta and Vegef, the expected, the observed to expected was actually lower. So meaning what we would have expected it to be was higher than what it was, which makes sense. I think essentially we are giving them a bolus of these cytokines, which is kind of being reabsorbed by the body or taken up by different cells. But the only one that didn't is rantes. So the amount of rantes in the circulation after transfusion was higher than we would have expected, just based on what we gave them in the unit. So this is implicated to us that something biological is happening in the baby and some cell in the baby is making rantes in response to this transfusion. What cell that is, we actually still don't know, platelets make a lot of rantes, monocytes make a lot of rantes, and we're in the process of trying to figure out what cellular source we think it's coming from, but it does seem to be biologically being produced in response to transfusion. We did look at net formation as I mentioned previously. So we have pre-transusion levels and post-transusion levels in every single baby. And you can see here they do overall increase. There's a few babies who have decreases in net formation post-transusion, but overall there's an overall increase in netosis and net formation in the circulation of these babies. Again, either two or four hours after transfusion. So of course, we want to know, well, what factors are associated with this? And like we talked about before, there's many places along the pipeline of a baby receiving a transfusion, where things can alter the outcome. So the sex and the age of the donor, there's a lot of really beautiful data suggesting that these are really important factors in both red cell transfusions and in platelet transfusions. And then obviously there's a lot of factors in the baby, the sex, to gestational age, their birth weight, how old they are. And then we use these kind of pre-transusion cytokine levels to think about the level of inflammation in the baby. Like giving a platelet transfusion to a septic inflamed baby has to be a little different than giving a transfusion to a non-inflamed baby. So these were all things we were thinking about. And then finally, there's the blood-banking practices, which I think I just put this slide in because I've learned so much more about blood-banking than I ever knew before. And I do think there's a disconnect between the clinicians in the unit and what's happening in the blood bank. And we don't always realize the nuance of what's happening. And we just are so happy to get our product and transfuse it. But really, like every platelet transfusion in every unit is not created equal. So when we think about it, all of our platelets for babies are luka reduced to prevent CMV. And then you think about prevention of grapharous host disease and you think about storage. So for a long time to prevent GVHD, we always irradiated all of our platelet products. And that's still how things are done here at children. So all of our platelet transfusions are irradiated. But now there's a new method called pathogen reduction, which is really interesting where they put basically a component into the unit that when exposed to UV light is activated and cross links all the DNA of white cells, but also of bacteria. And so that's a new way that we're doing pathogen reduction now. And then you either store the platelet in PAS or in plasma. So you can imagine there can be many different combinations. You could have an irradiated product in plasma or you could have an irradiated product in PAS. And so the new answers of blood-banking is something we also want to do investigate to see if that is something potentially affecting the outcomes in our babies. So when we looked at all of these factors, recipient factors, pre-transusion factors, post-transusion factors, donor factors, all these different things, the only thing that came up slightly is that storage time. There was a slight association between the storage time of the product and the ranties level in the baby. And so this is just showing that here. So as the unit storage increases, we can see that the ranties level in the baby also increased. And this is interesting because platelets used to only be stored to five days. And we're now pushing it a little bit more to seven days if they have a negative culture on the product. And so potentially this raises the question of, is that something we should be doing or not doing? But again, these are small N numbers. And then this also is very small N numbers. And it's really just preliminary data, but talking about what we store the platelets in, PAS versus plasma, we found that the platelets that are stored in plasma were associated with much higher ranties levels and net levels in the babies compared to those stored in PAS. There's a lot of caveats here, low numbers. Also all the PAS platelets came from Beth Israel, all the plasma came from childrens. So we're also looking at two different institutions essentially here. So there's a lot of caveats to this, but really this is preliminary data for the next set of investigations that Martha and I are planning really trying to get products from the blood bank and interrogate what the inflammatory consequences of each product are. So I think this is important information. And we obviously want to give our babies the right product that's going to have the least inflammatory consequences for them. Okay, summary, the whole summary here that we've gone through. So I hope if you take anything away from this talk and you remember nothing else, just remember that platelets are immune cells in addition to hemostatic cells. And they really do a lot of interactions with the immune system. And we should be thinking about those also when we decide to give a transfusion. Adult platelets are more prone flammatory than new natal platelets and platelet transfusions induce immune responses. So we know they increase their interactions with monocytes and neutrophils. They cause the release of different cytokines from the monocyte. They also increase surface expression of receptors that potentially create monocytes that are more pro migratory or they might extravisee them to tissue more readily. And then we also know that they cause neutrophils to create nets and circulation. And then finally the platelet unit itself is giving these babies a bolus of cytokines. And while it might not have a biological consequence, still giving them a bolus of these inflammatory cytokines likely has consequences in and of itself. And we think that these unintended effects of transfusion, which again we're giving for the best reasons and hopefully to prevent poor outcomes might also be contributing to harm. And I think these are important things as we learn more and more about the differences between platelets and the implications of that for transfusion. Okay, in the last few minutes we'll go very clinical. So what have we done with all this information? In our lab we're really passionate about doing thoughtful basic science as long as it has a clinical correlate or we think we can change practice. So we did change practice in our unit and thanks to all of you who also rotate through our unit and attend in our unit and make decisions on patients. So during my fellowship we re-wrote our platelet transfusion guideline to really highlight and reflect the data that we know, especially from the Planet 2 trial. And these are run charts, which is not my area of expertise. I worked with Kristi Leeman who does quality improvement work. But what you can see here is we were really able to change our platelet transfusion practices actually very quickly within a year. We really dropped our transfusion levels and really just wanted to thank everyone for their work on that because as other institutions across the country are trying themselves to change platelet transfusion practices, people have been having a really hard time and they look to us actually as like the experts and how to implement change. And I think it's really just a testament to how our unit is so research oriented and based and we want to provide the best care to our babies. So this is amazing and we keep updating and changing this. And I just want to highlight on this slide though even though we talk so much about the detrimental effects of platelet transfusion, there are still babies who need platelet transfusions 100%. It's just about making sure the risk benefit ratio is really in the benefit, you know, more benefit than risk for those babies. But if you have a baby who's actively bleeding or if you look at our transfusion guidelines, there are still babies we keep at 50,000 if they're immediately post-op or you know, there's numerous different things like I put them in a higher category. All right. And then finally, what else can we do? And I just want to say we just can keep doing really hopefully great research on both clinical and at the bench. And this is a project that I've been so lucky to involve with Jill and Volinda and then with Nicole and Ashley two of the critical care fellows. And we're looking at the inflammatory effects of ECMO within also the inflammatory effect of the transfusion given on ECMO. And then also hoping to see if we can characterize the inflammatory consequences of the total transfusion burden while a patient is on ECMO. So this is ongoing. It's very preliminary. We've had four babies that have been on study and we have all the samples collected for them. And we're planning actually right now to start our first pilot run looking at cytokine levels, net formation, but then also including some markers of human stasis and looking at thrombinantitrombin complexes. And again, oh, I didn't mention but these are all babies with CDH who are on the circuit. All right. And with that, I want to say thank you really to so many people. I think anyone who does research knows no one person does it. It's really tons and tons of people. So Martha and I kind of now co-lead the developmental hematology and the neonatal transfusion medicine lab. As I said in the beginning, Martha is my mentor through all of this and I really can't done any of it. And I still can't do any of it without her. She's just phenomenal. And these are the amazing people who actually work in our lab and do a lot of this work. And I'm just so lucky to work with them every day. Henry Feldman is our statistician. Having a friend who the statistician is amazing as I've learned. And he's really truly the best. And then obviously all of our collaborators and the clinical research team at Children's in newborn medicine. Martha and I started our clinical study at the same time that we changed our transfusion guidelines. So we really shot ourselves from the foot to try to study platelet transfusions while we minimized the amount of transfusions we give. And we really wouldn't have been able to do even the study we did without these amazing people who consent for us and approach families. And then of course, our funding for the lab. Like that's it. I love any questions if anyone has any. Well thank you. That was spectacular. It is it is a real art to be able to give a talk that melds the basic science with the clinical. And I think you just gave a lesson how to do that. It was spectacular. Well, thank you for rattling the cobwebs in my brain. 35 years ago, I spent my research fellowship time studying platelets and forgot most of the things and a little bit came back once it was wasn't known. But you present, I'm sure there's lots of questions. You present an obvious clinical question as you're trying to figure out all these details. Like when I was studying platelets, it was before I appreciated that children were not just small adults. But to all of us, now that's obvious. It isn't at all surprising that you point out that neonatal platelets, which were fetal platelets a few days ago, are different than adult platelets, just like fetal red cells and we even use fetal hemoglobin now to treat adult diseases. You did some experimentation with cord blow, which I was interested to see. Is there any practical limitation to having blood banks or platelet banks from blood from from courts? Yeah, no, it's a really good question. And it's one that a lot of people are asking and actually some of our collaborators in Utah, that is something that they are like so, so interested in. And I think it's just in the early stages. I think there's potential that it could work because you can get a lot of blood from a placenta. That's cord blood is, we use it as a surrogate for neonatal blood. Is it a perfect surrogate for it? No, but obviously, we can't get enough blood from a baby to wash their platelets and do all the studies we want. But you can from a placenta. And so theoretically, you could bank platelets from it. But I think it's still just like so early in the phase. But people are trying to think of outside the box ways to answer this question. Because then you essentially could give a neonatal transfusion to a baby. So I think it's very early in the stage. I have like 15 questions, would I ask you? I'll open it up to you. I might have questions from it. Yeah, correct. Thank you for accent talk. As surgeons, we spend a lot of time thinking about necrotizing or colitis. And there's a general correlation that the greater the severity of neck, the lower the platelets, and the absence of bleeding. Is there any threshold where you would suggest as a best guess where one should transfuse neonates with a significant necrotizing intercollitis? Yeah, it's a great question, especially for our unit. So we have dropped on our guideline every baby that's not immediately post-op or immediately post-op from a neurosurgical intervention to 25,000. I think that's the lowest we can go right now because that's the lowest that's been tested in a randomized trial. That being said, Martha is about to conduct a new randomized trial in preterm babies comparing 10 in 25. And so we're hoping to be able to drop it lower. But it's interesting because I think that population, the critically ill patient, whether it's from neck or sip or for a long time in our unit based on their critical illness, we kept them at a higher threshold. And that was actually one of the biggest changes in our guideline was dropping that population down to the lowest threshold possible because again, yeah, if they're not bleeding, there's no evidence that they should be transfused at a higher level. And I think people for a long time did because they assumed if you have this preterm baby who's critically ill, they must be at increased risk of IVH. But we also now know that's not quite true. So I think right now the safest we can do is 25,000. So if you look at our guidelines, that's the lowest we let any baby go because we don't really have safety data lower than that. But hopefully it'll change over the next few years as Martha does her multi-center trial. This is a quick follow up. In the first study you presented, was there any subgroup analysis looking at neck to see if any platelet transfusion may have a beneficial effect on survival or other primary? It's a good question. In the like the planet two trial. So they didn't really have that granular level of data to look at that. But it is a good question. But people have looked at it again, not in a randomized trial, but just in retrospective studies. And there's been no evidence that there's really any benefit, which is interesting because there are populations in children and adults where it does seem like there's certain populations where platelet transfusion might be beneficial. But in the neonate, we've never found any evidence of benefit. Actually the only really small, small population is based on two case reports. So you can take it with what you, one is an ROP. There's a question about platelets and retinopathy of prematurity, which I think still just definitely needs a little more investigation. But outside of that, there's really no population that has been found to have any benefit. Yeah. Thank you for that talk. Do you think with all the information you've presented that will change the threshold for trans, for platelet level before we do surgery? You know, we take the threshold for 50,000. Do you think with that, this will set a precedent to start looking at changing that threshold? Yeah. So we have changed ours a little bit in the unit to the point that if a baby's thrombocytopenic preop will more likely just put platelets on hold for the OR. And if they're given, then that's wonderful and great because obviously we want to make sure that they have adequate hemostasis during the surgery. And then we do keep them post-op at a higher threshold. So in that immediate, I think it's 28 to 48 hours post-op, we keep them at a threshold of 50,000. But I do think it might change. Yeah, just the thinking of like what does it look like to be, you know, stable preop for a surgery? But I think to me the way to do it is just if it's low, but not low enough to transuse, maybe just have them on hold in the OR. And then obviously, if there's any concerns, they should be given. And again, that's the most like Martha, I always tried to caveat that like, there are definitely babies who need platelet transfusions. We're not anti-platelet transfusions, just like finding the right time and to do it. So I think then having it be guided more clinically would be kind of the right, at most likely. More? Oh. I'll say you know, compromise. Well, platelet counts. Is there any sauce and collaboration with these groups? Look at platelet impacts on compromised groups. Yeah, that's a little beyond my expertise. We focus mostly on babies. And obviously, thankfully, the incidence of cancer and neonates is very small. But there is a lot of work from other groups working on platelet transfusion and cancer. And I do think that's one of the populations where there are subsets that do benefit and other populations that seem to have harm. And a lot of our friends that we do collaborate with also work in those fields. So we talk with a lot of them as well too. But I think, I do think that every group, probably whether it's adults or children or people with cancer, there is nuance, I think, with transfusion, whether it's red cells or platelets or cryo or SFP. We could talk about any of them. But I think with all of them, there's actually a lot more nuance than we initially thought. And probably less is usually more for the majority. So thank you for the question. I must thank you one more because that clock's fast. That's right. You're emphasizing mostly the potential downsides to transfusing premature babies. But as we think about the physiologic changes that happen in various body systems over the course of development in early post-natal life, could it be that the difference in neonatal platelets could be a detriment and could contribute to things like necktizing anticolytus or even intracranial bleeding you showed in one of your early slides, the very dramatic difference in platelet response and F2 to ADP. You didn't show collagen, but collagen is a prominent stimulator. Is it possible that the immature vascular in the bowel or in the terminal matrix might benefit from more of an adult level of responsiveness in the patient? No, it's a good question. I mean, I think the thing we haven't studied it in our lab, but I do think all the studies in babies looking at if transfusion improves the outcomes of IVH or neck or things like that, we just don't see any benefit from the transfusion of adult platelets. So especially even in IVH, which is the one that people have investigated the most, we see that severe IVH occurs regardless of the platelet count to babies platelet count, but also regardless of the transfusion. So it really seems like even babies who maybe receive a transfusion and there were there are older studies now, but there's even a study that showed, well, what if you transfuse very liberally and you give them adult platelets and you get them up above 100,000? You know, maybe we're down here, but really we just need to get them all really high and we'll prevent IVH should know difference. So I think the thought is really especially for IVH. The platelet definitely has to have some role because of course there's a role in bleeding, but I think the majority of the pathology behind IVH is just the immaturity of the vascular. And a lot of some mouse studies and preclinable studies have just confirmed that, that it really just seems that the germinal matrix is just such a different vascular niche than anywhere else. It's just so sensitive to foluses or any, I mean also just intrinsic things in the baby, but one thing I think is interesting is just the last clinical pearl to leave you with is that in the planet 2 trial, they interestingly found increased bleeding, which was so confusing a little bit, right? You give these transfusions more liberally, the babies actually had more bleeding. And we think and the people who ran the study think that it's because the transfusions were given rapidly. So it actually was the bolus and then caused potentially IVH. And so you'll see in our guidelines in the unit, we recommend running it much slower, at least over an hour or two hours. You could even run it over three hours in the most premature babies. You have a certain window of time after it comes up from the blood bank that it has to be used so you obviously can't go past that. And in the trial that Martha's doing, everyone's running them much slower. And so we're expecting to not see hopefully that increased risk of bleeding. Because we think that might have been more of this like folus with a vascular. This answers your question, I don't know. But so run them a little slower. That's always, always better. Well, thanks for this really spectacular talk and also for the great care that you provide and the mentorship that you provide. And as you become appreciate for recognizing your mentor and where we gather. Thank you. And I don't mentor Brian. He's very wonderful. Thank you.