Hi, thanks for tuning in. I'm Mike Stentz, and I've been tasked with answering the question, ECMO, do we even need to anticoagulate, which is a very controversial question. I'm an anesthesia intensivist at the University of Michigan, where I practice primarily in the cardiovascular ICU. My contact information is down there if you have any questions after the talk. So I have no financial disclosures. I do want to disclose that I am not a hematologist. So if you have questions about the clot and cascade, I'm not going to be the guy to answer those. I'm going to approach this more from the perspective of an intensivist and an ECMOlogist. So when I ask the question, do we need to anticoagulate? I think our possible answers are yes, no, maybe, I don't know. And can you repeat the question? I'm hoping that by the end of this talk, we'll have a little bit more clarity on this topic. So I thought it made the most sense to break this down into two questions. The first is why do we anticoagulate? And then the second question is what will happen if we don't anticoagulate? So first, we're going to address why do we anticoagulate? And the best answer I can come up with is we always have. So ever since the initial on-bypass surgery in 1952 by Dr. Gibbon, we've been using unfractionated heparin to induce a state of infinite anticoagulation in these patients. And in fact, in Bartlett's early descriptions of ECMO, he even uses the term infinite anticoagulation. And this was thought to prevent coagulation in the bypass circuit itself. In fact, we anticoagulated these patients so much that no one even bothered to measure how much we were anticoagulating them until the 1970s. You would just keep throwing more and more heparin at them. So the reason we have to anticoagulate these patients is that the bypass circuit is uniquely inflammogenic, if that's a word. So it will bind high molecular weight kininogen here, HK, and it'll also bind factor 12. So as soon as you go on bypass, these are going to adhere to the circuit. What happens next is that there's a chain reaction. So you're going to activate pre-calocrine, which is then going to activate calocrine, which then forms the sort of positive feedback loop. And it activates factor 12 over there, which then goes to all its activated forms and activates the intrinsic coagulation cascade, which is really unique to bypass and to ECMO, because pretty much every other surgical insult is going to activate our extrinsic coagulation cascade. Here, we're really activating the intrinsic coagulation cascade. But calocrine doesn't stop there. It also goes ahead and it activates the complement system. It activates neutrophils. It even activates endothelial TPA production. So you are both prothrombotic and coagulopathic at the same time. So then you're going to get all the evil humors, C3, C4, thromboxane, nitrous oxide, serotonin, free oxygen radicals, interleukins, histamine, all those sort of evil humors are going to make you both pro-inflammatory and also prothrombotic. So the question is, what role does systemic heparin play here? Well, we know that the main activity of heparin is through antithrombin, which then goes, heparin also has a little bit of activity against 2A and 10A and 9A, 11A, all those. So it reduces the fibrin burst from the extrinsic coagulation system. But remember here, we're largely worried about the intrinsic coagulation system. Now the fibrin generation still continues at low levels in the background, but you don't get quite the explosion that you would without the heparin. But it's also important to recognize that the heparin has essentially no effect on the complement activation and all the evil humors. So in other words, if we're asking what role does systemic heparin play, I think the answer we come up with is not much. So a lot of very smart people starting in the 90s came up with these biocompatible circuit materials. So what you have here is instead of just having the plastic of the bypass circuit in contact with the blood, they found very clever ways to coat it with heparin or in some cases albumin. There are even some circuits looking at trying to release nitrous oxide at this point of contact. But what these biocompatible circuit materials do is it prevents that very first step where the HK and the factor 12 adsorb into the circuit. So by preventing those from adsorbing into the circuit, you then break all the downstream effects. So you don't get nearly as much complement activation, you don't get the intrinsic coagulation cascade, you don't get all the evil humor circulating that you would with an uncoated circuit. It's also important to remember that ECMO circuits are simplified bypass circuits. So a lot of these initial data and experiences came from the cardiopulmonary bypass world, but the ECMO circuit is sort of dumbed down. So there's less places for it to coagulate. There are no blood air interfaces because there are no reservoirs in an ECMO circuit. And there are fewer moving parts. You're not going to have the suction, you're not going to have an LV vent, there's no cardioplegia lines. So there are fewer places where the ECMO circuit can clot off. So maybe we don't need to anti-coagulate quite as much. On the other hand, it's important to recognize that patients stay on ECMO a lot longer. And over time, you see some inflammatory markers like platelet activation increase, while other inflammatory mediators like, for example, complement activation seems to really go wild at first, but then within a few hours to days, it actually comes back to normal. So this graph here is not meant to scale at all. It's a purely conceptual graph, but you'll see that some inflammatory mediators peak early on, whereas others tend to increase as the length of the bypass run continues. So all that begs the question, what if we don't anti-coagulate our patient? The biggest and most common complication, of course, is mechanical thrombosis. About 10% of ECMO runs are complicated by mechanical thrombosis requiring circuit component exchange. And this is not a theoretical risk. These are both pictures that I've taken from my clinical practice, one showing a venous outflow line that is completely clotted off and another showing an oxygenator that over the course of a few hours just became completely encased in thrombin. Specific to our VA ECMO population, we also have to worry about blood stasis, because if you have decreased cardiac contractility on VA ECMO, you can get stasis within the ventricles, you can get stasis at the aortic root, you can get stasis in the pulmonary arteries. And if you don't find a way to either vent those chambers or find a way to anti-coagulate the patient, that intracardiac congestion and smoke can eventually become thrombus, which is obviously potentially catastrophic for our patients. And then we can't ignore the fact that the Extracorporeal Life Support Organization, ELSO, which is sort of the big end-all be-all organization in ECMO, still recommends anti-coagulation. This is from their website. These guidelines are easily available. They suggest unfractionated heparin at 50 to 100 units per kilo at the time of cannulation. And then they recommend maintaining a fairly high ACT or a reasonably high anti-10A goal. But it's important to note that these guidelines were last updated in 2014. And obviously, a lot has happened in the world of ECMO in the past eight years. So is anyone actually crazy enough to skip the heparin? The big thing is there are lots of case reports here. There are very few prospective studies. But often, we will withhold heparin in patients who are having complications. If they're thrombocytopenic, if they have a brain bleed, if they're developing some sort of problem that makes us say, you know, I really don't feel comfortable anti-coagulating this patient. So it makes a really nice convenient sample for us to study and report on. I'm also really lucky in that shortly after I was assigned this talk, there was published a very nice and thorough systematic review. So there have been several systematic reviews of ECMO without anti-coagulation. This is the most recent one I could find published in the ASIO Journal. And basically, what they found was, you know, the balance, of course, is always between to bleed or not to bleed. So bleeding complications, we know, will complicate up to 29% of ECMO runs. 10% of those are going to be major bleeding, although that's not always well-defined when these cases are reported. And intracranial hemorrhage, which of course is devastating, is 4% to 10% of those bleeds. On the other hand, clotting complications, as we already discussed, oxygenate or failure will occur in about 10% of patients. CVA will occur in 2% to 4% of patients. And there are some series looking at DVTs after decannulation that show that as many as two-thirds of patients will have a DVT after decannulation. So this systematic review identified 32 case reports and two prospective trials. They identified 201 patients who were maintained without anti-coagulation. A slight majority were VV, and then 90 VA, venovenous and venoarterial ECMO patients. 56% of these patients, they reported using a biocompatible ECMO circuit. The problem is a lot of the smaller case reports didn't report whether they used it or not, so we just don't know. And then only a very small number of these patients were enrolled in prospective trials. Most of these were retrospective case reports. So I tried to summarize their findings here. And there's no statistical comparisons, obviously, because these were all small studies and there was too much heterogeneity. But we can see here, if we're looking purely at thrombotic complications, they were pretty similar between the two groups. In other words, again, without statistical inference here, anti-coagulation does not seem to help prevent these thrombotic complications. On the other hand, looking at bleeding patients, we can see that any bleeding was pretty similar between the two. There was actually more major bleeding reported in the group without anti-coagulation, which we can argue probably suggests some indication bias there. In other words, patients who were at risk of major bleeding were less likely to be anti-coagulated. But you can see in the far right column that the rate of intracranial hemorrhage was still much higher in the anti-coagulation group. In other words, withholding anti-coagulation seems to reduce your risk of intracranial hemorrhage. Of course, it's important to take all of this with a grain of salt because we're looking at just 200 and some non-anti-coagulated patients, whereas we have 2,800 coagulated patients. So still a very, very small minority of patients are undergoing ECMO without anti-coagulation. I just wanted to highlight one of those studies. This was the group out of University of Rochester, and they reported their experience with veno-arterial ECMO without systemic anti-coagulation. Since 2016, they have not been routinely anti-coagulating veno-arterial ECMO patients. They reported 131 patients of whom 72 were not anti-coagulated, and they used biocompatible circuits unless the patient had a known history of HIT, heparin-induced thrombocytopenia. What they found was that there was a lower complication rate in the group that was not anti-coagulated, and this was statistically significant. They also found that patients were less likely to require transfusion if they were not anti-coagulated. And although the data aren't here, they also received fewer units if they were transfused. The mortality was higher in the group that was not anti-coagulated, but so was the lactate. So it's important to recognize that this group where they forewent anti-coagulation was likely a sicker cohort of patients, and again, raises that same problem of indication bias that we discussed earlier. But if they did a multivariate adjustment here, and it actually showed that anti-coagulation did not have any effect on overall mortality from ECMO. So then the big question is, well, there seems to be a little bit of case reports here and there. What's everyone else doing? There were two large international surveys of ECMO centers that showed almost every single ECMO center in the world reports using heparin. But the problem is they didn't specifically ask, do all your patients get heparin? Only that at least some patients got heparin, and of the patients who are anti-coagulated, heparin was by far the preferred anti-coagulant. So what the heck do we do with all that information? I put up a big warning sign here, because from here on out, everything I'm going to say is purely editorial and my synthesis of the data, your mileage may vary. So when do I anti-coagulate? I anti-coagulate every patient who has a strong indication for anti-coagulation besides ECMO. In other words, if you're on ECMO, because you had a massive pulmonary embolus, you have a strong indication for anti-coagulation, and I will anti-coagulate that patient. Patients with minimal pulsatility on veno-arterial ECMO, again, I'm terrified of getting that LV thrombus that's then going to flick off to the brain. So I will anti-coagulate patients who are on VA ECMO with minimal pulsatility. Patients who have a known hypercoagulable state, I will tend to anti-coagulate. Someone asked me about COVID the other day. COVID does seem to increase your risk of needing an oxygenator exchange. So I'm more prone to anti-coagulate my COVID patients, but that's sort of a gray area for me. And then if it looks like your oxygenator or any part of your circuit is starting to develop a fiber and stranding or thrombosis, I'm more prone to anti-coagulate those patients, even if I wasn't before. So then when do I not anti-coagulate? I do not anti-coagulate patients who are actively bleeding. Even if they have another indication for anti-coagulation, major bleeds on ECMO can be life-threatening. So I will pause anti-coagulation on those patients. High-risk patients, maybe they had a recent stroke and I'm worried about hemorrhagic conversion, or they have known esophageal varices, something like that. I will not anti-coagulate those patients. And then our practice pattern lately has been getting away from anti-coagulation for most VV patients. If you are on veno venous ECMO, have good heart pulsatility, have no other reason to be on anti-coagulation. Lately, I've been skipping anti-coagulation and that seems to go just fine. So let's try to put it all together here. We know that ECMO results in activation of the intrinsic clotting system as well as complement activation. We know that systemic anti-coagulation has historically been used to reduce thrombotic complications, but biocompatible ECMO circuits seem to reduce the need for systemic anti-coagulation. We know that a no anti-coagulation strategy has been reported to be safe, but this is a very small number of patients relative to the overall ECMO volume. And like so much else, unfortunately, more data are needed. So if you want to download these slides from the website, I tried to put together a little bit of a selected bibliography here. And I'd like to thank you all for your time.

anticoagulation

extracorporeal membrane oxygenation

bypass circuit

thrombotic complications

intracranial hemorrhage