Thank you very much. So a confession at the beginning, which is that once I had a title for my talk, I moved away from the nutrition part and just stuck with the machines part. So sorry to disappoint anybody who was expecting otherwise, but I hope that this is nonetheless an interesting and exciting talk. So my disclosures, I have no financial disclosures related to this talk. I do have more of a clinical perspective that I bring to this talk, something that I think is much more interesting and hopefully for those of you who work with ECMO, something that you'll actually be able to take away and use in the future. So the layout for the talk is going to be first discussing some of the risks for immunosuppression in ECMO patients that do not have to do with ECMO, both direct and indirect. Then we'll talk a little bit about mechanical circulation and immunity, specifically looking at cardiopulmonary bypass as sort of a control population to study the effects specifically of the machine as it interacts with the human. And then we'll talk about ECMO and the population we use ECMO with. And then lastly, we'll talk about some opportunities that we can use this information. So once again, we've talked a lot about, I mean, we've talked already a little bit this morning about immunosuppression and you may have seen some other talks this Congress about immunosuppression. There's lots of things that can do it, but I'm going to say that it was the machines. Machines did it, but did they do it? How can you really know that given the complexity of this population? So we'll pop a red pill and we'll go down a bit of a rabbit hole. So immunosuppression is very common in the ICU as Dr. Riley talked at length about, and that was just in one specific population. There are, as you guys all know, there's a very heterogeneous population of patients who end up in ICUs and they can have immunosuppression for any number of reasons. It couldn't possibly have been this friendly looking robot. So there's the underlying disease process, both the chronic diseases that they brought in with them to the ICU, the acute process that brought them to the ICU. There's what we do to them in the ICU with various pharmacologic interventions, steroids, other immunosuppressants for transplant patients, et cetera. There is a large degree of malnourishment. I won't invoke vitamin C for fear of never talking about anything else again. The patients often receive a lot of blood products, especially at the ECMO population, which itself is an immunomodulatory agent. And then many of them are on hemodialysis, which is another form of extracorporeal life support that has its own immunomodulatory effects. Similarly, infectious risk is very high in this population. So if you're in the ICU, you are at risk for exposure to infectious agents, independent of what you brought in with you to the ICU. So you are subjected to the presence of invasive devices, endotracheal, intravascular, urinary, potentially rectal. You're exposed to people coming in and out of your room all the time. And this can be especially true in ECMO patients due to the duration of illness and support. So how do we know it was the machines? Again, immunosuppression is common in the ICU. Infectious risk is high in the ICU. You've got these multiple confounders of the diseases. You've got the treatments. And again, with ECMO, you've got an especially high acuity population because ECMO is typically used as a salvage or last line therapy for only the most ill patients for whom there is no other option. So this is kind of an apples to oranges comparison. So how are we going to look at this and say it was the machines? And again, it's going to be looking at other forms of extracorporeal life support. So hemodialysis has been studied and it's been shown to have some immunomodulatory effects, but so has cardiopulmonary bypass, which being very similar to ECMO, is what we're going to talk about and try to do more of an oranges to lemons comparison. It's not apples to apples, but at least these are both citrus. So I don't know how many people are familiar with cardiopulmonary bypass or ECMO, but they're both very similar. So they are both continuous flow extracorporeal life support systems. They both have relative stasis of blood in various sections of the machine, and they are both composed of tubes and pumps and oxygenators. So a lot of similarities, but there are also a lot of differences. To me, and as it pertains to this talk, the biggest difference is going to be in the baseline population characteristics. So a patient who's undergoing cardiopulmonary bypass is somebody whom a cardiothoracic surgeon has determined to be an eligible candidate for surgery, and typically a very high-risk surgery. They are undergoing a relatively short procedure where a long cardiopulmonary bypass time is measured in hours, and a short ECMO run is measured in days. There's also an element of air-blood interface with cardiopulmonary bypass because the venous blood is maintained in a reservoir, and so it just kind of sits there floating and kind of percolating around. And then there's often hypothermia induced in patients undergoing cardiopulmonary bypass, which is not always the case with ECMO. There are some other differences with the types of pumps. The ECMO machines are typically high-speed centrifugal pumps, magnetically levitated, whereas cardiopulmonary bypass is typically a low-speed roller pump. There are venous, it's usually a passive venous drainage in cardiopulmonary bypass. But you know, for what we're looking at, it's just about as close as you can get for someone who's not critically ill. And here's some schematics of the two for comparison. So what are then the immunologic effects of cardiopulmonary bypass? This is something that's been studied relatively extensively, and it has been shown to induce a SERS-like response. So primarily through the contact system, so as blood enters the extracorporeal circuit of the cardiopulmonary bypass machine, the contact of the blood with a foreign surface activates the complement system, it starts to upregulate cytokines, and these things all feed into the complex response that I'm not going to go into detail with. But essentially, it's a very SERS-like response. This is exacerbated by the hypothermia that we use in many cardiac cases, by the stress of surgery itself, by potentially ischemia reperfusion insults, if that is a component of the surgery. And there is some evidence that perhaps it's decreased by steroids and by the use of more biocompatible components of the cardiopulmonary bypass system. The effects that this has on the immune system do affect both the innate and the adaptive immune response. So just a couple of quick sample slides. So this is a time-dependent reduction in HLA expression. So T1 is pre-cardiopulmonary bypass, T2 is immediately following, T3 is post-op day one, and T4 is discharged from ICU. And as you can see, there's a time-dependent decrease in HLA expression. There's also a time-dependent change in the relative expression of PDL1 and of the ratio of PD1 over time. And again, this is by days with the T1 through 4, and also in minutes. So the duration of exposure to cardiopulmonary bypass in minutes can affect this. And there's some evidence as well that it's mediated by myeloid-derived suppressor cells. This is a relatively new frontier on this. But as you can see in the bottom graph, a population of myeloid-derived suppressor cells is invoked at this time 1, or time 2, sorry, right after cardiopulmonary bypass, and then it fades over time. And I don't know if any of you attended the immunology talk on sepsis yesterday, but there's a lot of investigation into the role of myeloid-derived suppressor cells in the immune response to sepsis. And so maybe there's a lot of similarities here. So that's cardiopulmonary bypass, but what about ECMO? The same thing. It's very repetitive here. ECMO also induces a SERS-like response. It also does it through primarily the contact system, complement, cytokines, cellular activation. And in this somewhat busy big circle here, you can see on the right side of the circle is the endothelium, and that's to represent the patient circulation itself. And on the left side is the ECMO tubing, so the biomaterial which over time develops a coating of adsorbed proteins, but nonetheless will lead to activation of all these inflammatory cascades. In addition with ECMO, again, due to the high-speed centrifugal pump, there's going to be some hemolytic effect as well, and some uncontrolled release of intracellular cytokines from blood products being spun through a tiny cone at somewhere between 3,000 and 9,000 RPM. So do we have evidence of this? Of course we do, and I have to show you the slides to show it. So we have here, again, time-dependent increases in all of these pro- and anti-inflammatory cytokines. In this study, it was venoarterial ECMO, which is the exception to the entire blood circulation being pumped through the lungs, compared to patients in cardiogenic shock who were being supported medically, and that is the control group. And as you can see, there is, again, an increase and a time-dependent increase, both over the course of days and over the course of hours, in these pro- and anti-inflammatory cytokines. And similarly, there's an upregulation of myeloid-derived suppressor cells. So the orange circles on the right side of the bottom-most graph are an increase in the population of myeloid-derived suppressor cells. So perhaps there's some similarities with these inflammatory processes, and sepsis, and biotrauma, and hypoxia, and all those other things. So those are the direct effects of ECMO, but again, there's a lot of indirect effects in the specifically ECMO-supported population. They typically receive more blood transfusions due to ECMO, related to bleeding and hemolysis. We often have higher hemoglobin targets for patients who are suffering from hypoxic respiratory failure. We use immunomodulatory therapies for the underlying diseases that are being supported by ECMO. They have these unique, indwelling, large-bore cannulas, which are frequently in the femoral position, which also represent an inflammatory insult. And these can be present for long durations. At UCLA, I think our record, I think it's eight or nine months that a patient was supported on ECMO with the same cannulas. There's also the, potentially, depending on the type of ECMO cannulation, you have surgical site infection risk with cut-downs, staples, sutures, things like that that are sitting there for the same length as the cannulas. So here's a fun little graph representing the kind of second-hit components of ECMO. So you have this primary disease burden, this underlying acute and chronic disease that the patient has brought in with them, and is probably the reason that they're being placed onto ECMO. Then you have ECMO. You have the tubing. You have the pump. You have the oxygenator, and you squeeze all this blood through there, and then it comes out back into the circulation. And of course, it has no idea what's just happened, and it's very inflamed, and there's a lot of cytokine storm that's happening. But so what? Who cares? If I tell you that ECMO induces an immunosuppressed state, you're going to not put somebody on ECMO because they are going to have a little bit of a, you know, reduction in their ability to respond to infection, or because they're going to get a little bit of inflammation. You know, these are, again, salvage therapies. When patients need it, they need it. So we use it. So what are we going to do with this information? So the first thing is to remember that patients who are supported by ECMO, regardless of the underlying reason for it, or the cannulation strategy, are at very high risk for infectious complications. So you must have a high index of suspicion for it, especially because the usual indicators of infection are going to be extremely affected by the presence of the ECMO circulation. So fever is not going to be present in somebody who's got a heater-cooler in their ECMO circuit. Tachycardia may not be present in somebody who's on veno-arterial ECMO, or it may be present all the time because of why they're on veno-arterial ECMO. Tachypnea, same for veno-venous or veno-arterial. The lungs are supported, in this case, by an extracorporeal circuit, so you may not see these things. So you need to have a very high index of suspicion. We often use low... We have a low threshold for antibiotics, and we often use them for long durations. Because again, you have a very difficult time with source control. These cannulas cannot be easily removed. If someone gets a CLABSI, step one, take out the line. Someone gets CLABSI, but they're on ECMO, you can't just take out their cannulas. Not if you needed to put them on ECMO in the first place. Paying attention to meticulous wound care, avoiding and minimizing non-ECMO causes, so mobilizing, feeding them, making sure that they're replete with their micronutrients, and then of course, getting them off ECMO if you can. Personally, I use this a lot to emphasize with the families, when I talk to the families or with the patients, if they're in a position to do so, as an example of the inevitable complications that are related to support by extracorporeal means. That though there are no hard expiration dates for these machines, they are practically limited by complications. Bleeding, clotting, and then, as pertains to what we're talking about today, infection. We have seen many cases of progressive multi-organ dysfunction syndrome that ultimately leads to death despite ECMO in our population. We presume that it's sepsis in many cases, and that sepsis is probably related to their underlying immunomodulatory and immunosuppressed state, inability to fight off infections, and perhaps there's some degree of culture-negative sepsis that is purely inflammatory going on. So I use this as an approach to explain to families the limits of this. And then, as I've mentioned a couple times, there's a lot of similarities to sepsis. So maybe this is something that we will start to be approaching much more as we approach the inflammatory process that is ongoing with sepsis and systemic infection. Some other exciting future directions that we aren't really able to take full advantage of yet, but our frontiers are improved biocompatibility. So most ECMO tubing is heparin-coated at the moment, but there are other biocompatible surface coatings that are essentially designed to endothelialize the tubing. And data is kind of less than exciting so far, but that is hopefully something that we can use. The other thing is a different analysis of anti-inflammatories as far as the risks-benefits for people who support it with ECMO. Arginine supplementation is something that's been explored in cardiopulmonary bypass and shown to, due to its mechanism with myeloid-derived suppressor cells, increase the number and viability of the mononuclear cell lines, all of them. But this is something that's been tried in other fields and there have been inconclusive effects there, so we'll have to see. Inflammatory adsorbers, so this is the schematic for cytosol. This was something that was very popular during COVID. Very easy to splice into the membrane and, you know, inflammatory mediators just stick to these beads and then they're gone, hooray. But the case reports on this have ranged from banal to miraculous. So whether this is something that needs to be incorporated into every ECMO circuit is something that we're going to need to demonstrate. And then again, with the similarities to sepsis and the inflammatory response to other infectious diseases or syndromes, as these are developed, we'll have to see if they can be applied as well to ECMO. So as with Roy Batty and Blade Runner, these machines, they are complex. They are neither pure good nor pure evil, but we, you know, hopefully we can get some good out of them. So thank you very much. Thank you.

ECMO

immunosuppression

inflammatory response

infection risk

critical care