 Thank you so much for that kind introduction, I'm really flattered by it and I'm really flattered and really honored to be giving this Lewis lecture. I wish I could be there in person and I look forward to the time when we can all see each other in person again and put this all behind us. I appreciate everyone dialing in today. I'd like to talk to you a little bit about things that we've been involved with over the last decade or so and I've titled this talk improving the safety of ACS treatment from reducing bleeding to defining transfusion thresholds because I think the two issues are really interrelated and you may wonder, why is an interventional cardiologist interested in transfusion medicine? I'll take you through a little bit of the history of how I got involved but it's really exciting work that we and others around the country are doing to try and define exactly what the role of blood transfusion is in patients who are undergoing coronary intervention as well as those with acute coronary syndromes. I don't have any non-research relationships with industry. I have some funding to the institution for my role on a randomized clinical trial. Of course, I get a salary from the American Heart Association for being the editor-in-chief and I do want to thank Jeff Carson for letting me borrow some of his slides. This is Grand Rounds so I think it's important for us to start talking about cases because this is how we used to learn in medicine and I hope this is how we continue to learn. I'm going to present to you a 67-year-old farmer who presents to you with dyspnea and congestion. He has a history of various infections that he survived but his symptoms have worsened to the point of the fact that he's having stridor. His physician treats him with bloodletting and bleeds him about 11 units of blood thinking that the more he bleeds him, the better the patient will get and unfortunately the patient died soon after. Now, if you're a student of history like I am, you will recognize this clinical presentation as being George Washington and his personal physician was Benjamin Rush shown in this mural leaning over him. Benjamin Rush was one of the signers of the Declaration of Independence and a very big proponent of bloodletting as a treatment strategy. Now, if you train an interventional cardiology when I did in 2003-2004, which was the heyday of combination antithrombotic therapy and large bore access in the femoral artery, you would think that perhaps we were proponents of bloodletting. But I think we've gotten a little bit more refined about what bleeding means in this patient population and how we can potentially even change that. For those of you who are also students of history of medicine, you may recognize this person on the right hand side. This is James Blundell. James Blundell was an obstetrician gynecologist who in 1820 performed the very first successful human-to-human transfusion shown in the painting or the drawing to the left. He had been experimenting with some animal-to-animal transfusion, which were summarily unsuccessful, but when faced with a woman who was suffering from peripartum hemorrhage, he hooked up his own arm to her and transfused her with blood and saved her life. Now, remember at the time that he did this, the discovery of blood groups was several decades away. Landsteiner hadn't actually developed that classification scheme until the early 1900s. 1901, if I remember correctly. So this is a pretty risky thing to do. And we've had a tremendous amount of information about the role of blood transfusion and improving the safety of blood transfusion, except in one area. And that is really what is the role of blood transfusion and what is the optimal transfusion threshold in patients who have ischemic heart disease. And finally, let's talk about a little bit more of a contemporary patient. Here's a 75-year-old female with a history of hypertension type 2 diabetes, AFib, who presents with ACS. In fact, there may be patients in your emergency department right now who fit this profile. She's 59 kilograms. I don't know if that actually exists in Texas, but maybe it does. She's anemic with a hemoglobin of 8.5 and her creatinine is 1.6. Normal ejection fraction. And I think consistent with the guidelines, she's scheduled for invasive risk stratification. And as part of that decision-making process, the questions that we face as clinicians is, her hemoglobin is low. Should she be transfused? When she goes to the cath lab, what access site should be used? And if she undergoes PCI, what combination and duration of dual anti-clinotherapy should she receive? All of these decisions take into account the balance between worsening the ischemic risk and trying to balance that against reducing the bleeding risk. So I think this is not an unusual situation. As we all know, the ACS process begins with plaque rupture, which is shown for you on the right-hand side of the screen. And the patient in the cath lab, they undergo iatrogenic plaque rupture with deployment of the balloon. That leads to generation of thrombin, which leads to platelet activation and platelet aggregation. And it is our treatment for ACS that is directed at these various different elements. We have anti-thrombin therapy. We have inhibitors of platelet activation, both intravenous and oral. Certainly, we don't use a lot of platelet aggregation inhibitors anymore, like 2B3A inhibitors. And then, of course, the injured vessel wall needs to be treated either with stenting or with ultimately bypass surgery. When you couple those two things together, which is invasive procedures with potent anti-thrombotic therapy, that sets up the sort of perfect situation for a patient to have a bleeding event. And I think that what's happened over time is that the bleeding risk is really front-loaded in patients with acute coronary syndrome or those undergoing PCI. It's really the highest at the time the patient undergoes the procedure. They may get treated upstream with some anti-thrombotic therapy. But that arrow points to where the invasive procedure happens, which is where the bleeding risk is the highest. And then in the old days, in the bare metal stent era, that bleeding risk would dramatically fall once that 30-day point is reached because we were treating patients with dual anti-therapy for 30 days. So a lot of the focus has been on the inpatient setting and how we can reduce that risk for bleeding and potentially reduce the risk for transfusion in the inpatient setting. Only recently now have we taken it outside the hospital and I'll talk about that towards the end of the talk. If you look at the overall landscape of complications in patients undergoing PCI, and these are data from Fred Mosuti looking at the CATH PCI registry, the most common complication that happens in patients who undergo PCI, regardless of indication, and here I'm just showing you the STEMI and the non-STEMI, is bleeding. And in fact, if you look at what's happened with our current technologies, we've really been able to push the ischemic complications pretty low. The overall death rate from STEMI or non-STEMI is below 10% now. Shock, even though a lot of us are talking about shock, it's been one of the most popular things to think about, really shock is relatively rare when you compare it against other complications. Now, if those of you have been around for a while like I have, think about what these numbers mean in the context of the ISIS-2 trial and what 14-day mortality was in that step. And you can see that we've made a tremendous amount of progress with respect to ischemic complications, but bleeding is still something that we wrestle with in patients who come in. And we've also gotten a little bit more sophisticated about what that bleeding means. First, we've been able to show our group as well as others. These are data from the Washington Hospital Center that I'm showing you here, showing that it's the risk of bleeding or the incidence of bleeding is highly dependent on the patient's presentation. Patients who have stable angina or even unstable angina, really the vast majority of their bleeding complications can be related to the access side, but their overall risk of bleeding is relatively low. Think about the patient who comes in for a half with a positive stress test, for example. Now, if you take the patient and you increase their baseline level of risk for ischemic complications to the non-ST-segment elevation ACS, then their risk for bleeding goes up. If you take that to STEMI, it's even higher. And it's the patients who have shock who have the highest risk for bleeding. Now, that's probably because the patients with STS-segment elevation MI, for example, they're generally all undergoing invasive procedures and the patients with shock are undergoing invasive procedures plus large warframural access. So it may not necessarily be inherent to the ischemic risk of the patient as much as it is how we react to that ischemic risk and how we treat those patients. The other thing that we've learned about bleeding complications is that the proportion of bleeding that occurs related to the access side also changes with the clinical presentation of the patient. So while the patients with stable angina coming in for an outpatient cath and outpatient PCI have the lowest risk for bleeding, they have the highest proportion of that bleeding related to the vascular access site. When you go to the patient with the non-ST-segment elevation ACS, again, higher risk for bleeding compared with stable angina, but interestingly the proportion changes now. The vast majority of their bleeding is actually not related to the vascular access site, it's related to non-access site bleeding and mostly it's GI bleeding. And then if you change again to the STEMI population, the highest risk for bleeding, again, the vast majority of their bleeding risk is related to the vascular access site. Now the reason I bring this up is later on in the talk when we start talking about bleeding avoidance strategies, particularly as a related vascular access site, you're going to get the most bang for the buck in the patients who have vesti-segment elevation because they have on the highest risk for bleeding, so the most to gain. And number two, the vast majority of the bleeding events are related to the access site. So the access site strategy is going to have the most effect in these patients. And what I'll show you is that randomized trials have in fact shown that. Now we were among one of the first groups, certainly we were not the only ones to show that there was a strong association between bleeding complications that occur in the hospital and subsequent 30 day mortality as well as 30 day death or MI, as well as six month mortality. And even some groups have shown one year mortality and Steve Mnukin showed very nicely that there's an association between bleeding complications that occur in the hospital and subsequent stent thrombosis. We showed in this study looking at 26,000 patients that there's a gradation of risk, a dose response if you will. The worst, the bleeding complication, the higher the risk for bleeding or the higher the risk for death. Now, if you look at this red line, and I apologize for anyone who might be red, green color blind, but the red line, which is gusto severe bleeding, that's probably not surprising that it's going to have a very robust relationship with 30 day mortality. Remember, gusto severe bleeding isn't cranial hemorrhage or bleeding causing hypotension. What is surprising now is the fact that relatively lesser degrees of bleeding, so-called mild bleeding in this particular study also seem to be related to 30 day mortality. I want to think about what those mechanisms are. And you'll notice that I'm choosing my words very carefully here. I'm using the word association. Because again, outside of fatal bleeding or bleeding into your brain, this is probably an association and not directly causal. So as we start thinking about the future of antitrombotic therapy development, this is really the landscape that we're looking at. If you're worried just about the thrombosis and not about the bleeding, you're really just treating half of the patients. Now, as I mentioned, it's important to think about mechanisms of what this relationship is. Because if we understand the mechanisms, then we can start thinking about what the strategies are. So the first mechanism, of course, is confounding, right? None of these studies really randomized patients to bleeding versus no bleeding. Amazingly, there actually is a study that has done that. Many of you may know the feast trial. This is a trial of patients with symptomatic claudication where patients were randomized to serial phlebotomy over the course of six weeks compared with usual care. May not surprise you to learn that the serial phlebotomy strategy did not make a difference with respect to a mortality in that particular study. Now, I'm being a little bit facetious, of course. I mean, when patients bleed in the hospital, it's not a set amount of blood that they bleed. But confounding, I think, is something that we have to be aware of. Sick people bleed and sick people die. Severe bleeding we already talked about. If you bleed into your brain or you get hypotensive, that's probably directly a causal event. It's really the more mild or moderate bleeding that we're dealing with. And what is it that may underlie that mechanism? And I think it has to do with the fact that when a patient has a recurrent ischemic event, we intensify therapy. When a patient has a bleeding event, there's really not much we can do. We have to stop antithrombotic therapy. And that likely puts patients at risk for recurrent ischemic events and subsequent mortality. In fact, my colleague Tracy Wong has shown very nicely that patients who develop in-hospital bleeding, even if they survive that bleeding event and survive hospitalization at one year, they're less likely to be on evidence-based secondary prevention. So what can we do? Well, there are treatment strategies that we can use. Juditious use of blood transfusion. I'll talk a lot about that in this talk. Prevention is really important. Maybe we should just prevent bleeding to begin with. And the most potent way to do that, of course, is to make sure that these vascular access site strategies are evidence-based. So ultrasound, routine ultrasound for femoral access and of course an interest of mine, which is radial access, which probably is the most effective way to reduce bleeding risk in patients undergoing PCI. So this is really what I'm gonna focus on for the rest of the talk is blood transfusion and radial access. So let's talk about blood transfusion. I got interested in this a long time ago when I was a fellow. I was interested in, boy, it almost seems like a lifetime ago. Many of you may remember Medicare managed care and it was being rolled out. There was a lot of concern about the fact that it may cherry pick patients. So me and my research mentor at that time, Jamie Jollis, were trying to do a study of looking at Medicare managed care. And we looked at one year's worth of claims in community dwelling older patients who were Medicare beneficiaries. And what we found was that the number one procedure that they were being billed for was blood transfusion. It was more common than getting a head CT. So we were really amazed. And about that same time that we were looking at these data, the trick trial had come out. And those of you who are critical care trained may be familiar with the trick trial. That's the trial that showed that, you know, an aggressive transfusion strategy is no better than a conservative transfusion strategy and patients were critically ill. But if you look at sort of what transfusion triggers are across specialties or across different clinical presentations, you know, this is a great study by Paul Abert. It's a very old study by now, but I've not seen it repeated. Paul in Canada surveyed a series of different physicians, cardiologists, critical care specialists, surgeons and presented sort of contrived clinical scenarios. A patient who had trauma, patient with GI bleeding, a patient with sepsis and a patient with acute coronary syndrome and asked at what level of hemoglobin would you transfuse a patient presenting with these different clinical scenarios? And what he found was that the vast majority of respondents said that if a patient came in with ACS they would transfuse that patient if their hemoglobin was 10 or less. Whereas for other clinical situations they were willing to tolerate a much, much lower hemoglobin threshold. And the question is why is that? Well, in other clinical specialties and this is the trick trial that I mentioned previously this is Paul Abert's trial, 838 clinically critically ill patients not a lot of cardiology patients in this trial randomized to maintain a hemoglobin of seven versus nine with red blood cell transfusion. And what they found was that a liberal transfusion strategy that is maintaining a hemoglobin of seven was really not that much better than a restrictive transfusion strategy or maintaining a hemoglobin of seven. And so it is now basically part of clinical care in ICUs not to transfuse unless the patient is actively bleeding symptomatic from their anemia or if their hemoglobin really drops to seven or below. Now there are several other studies looking at transfusion triggers. If you look at, for example, patients who have acute GI bleeding a situation that probably all of us would say it's really important to make sure that the patient is typed and crossed and has several units available for transfusion to reduce their risk of dying. Well, in a randomized trial very difficult to do in the US this was done outside the US patients who were having an acute GI hemorrhage were randomized again to a restrictive or a liberal strategy maintaining a hemoglobin of seven versus nine. And what they found was amazingly something that would be counter to what all of us would think which is that the risk of re-bleeding was higher if they got a higher transfusion rate. So in other words, maintaining a hemoglobin of nine resulted in a significantly higher rate of re-bleeding. And the probability of survival was also lower. So even in a situation where a patient is actively bleeding from a GI source it appears that an aggressive use of transfusion is at the very least not better than a more conservative transfusion strategy than in fact maybe worse with respect to re-bleeding or the probability of survival. Now in cardiology though, it's a little bit different, right? Because our patients sometimes have known fixed coronary stenosis. And we know that hemoglobin is one of the variables in the oxygen delivery equation. And in this meta-analysis by Mark Sabatine looking across both STEMI as well as non-STEMI what he found was that there is a U-shaped relationship. Anemia clearly was associated with an increased risk of adverse outcomes. So if your hemoglobin was less than 10 you can see the point estimate for cardiovascular death by 30 days was significantly higher, two and a half times higher than if you had a normal hemoglobin. And if you adjusted for potential confounders you found the exact same thing. And in fact in this particular study they also looked at transfusion rates. And the transfusion rates interestingly were not that high, only about 5% STEMI about 3% non-STEMI and what they found again even for accounting for transfusion that this particular relationship seemed to hold. So perhaps there is a strong body of evidence to have a different transfusion threshold for patients who have fixed coronary stenosis or coronary artery disease particularly those who are coming in with acute ischemic heart disease like acute coronary syndrome. Now interestingly if you look at the literature on patients who have coronary artery disease and transfusion you can either look at that patient as being half full that is they don't need a lot more blood or half empty, they need a lot more blood. On the half empty side there was a study that looked at transfusion or actually anemia in patients who had a hemoglobin that dropped below 10 and found that there was a spike in cardiac output. It may interest you to know this is a study done at Duke by Jim Warren and Eugene Stead. I don't think you can do this study anymore but what they did was they took medical students put P.A. catheters in them and then gradually phlebotomized them to their hemoglobin drop below 10. And if their hemoglobin drop below 10 they found that the cardiac output spiked so they said that's a maladaptive response you gotta keep the hemoglobin at 10. Other studies have found the same thing there's a Harlan Krumholz's look at the co-operative cardiovascular project showing increased mortality of an inadequate was less than 33 that translates to about a hemoglobin of 10 or 11. On the other hand, we have plenty of data particularly in Jehovah's Witnesses for example where you can push the hemoglobin as low as six after bypass surgery and find no increase in mortality. So when you have this degree of evidence showing that you can either transfuse a patient or not transfuse a patient you have perfect clinical equipoise. And because of this uncertainty there are about 12 million units of blood transfused annually in the United States that's approximately three to six units for every single citizen in the city of Houston to get a unit of blood. So we really don't have any idea how we should be using blood transfusion in patients with ischemic heart disease. We have a lot of observational data so Matt Sherwood when he was a fellow with us he's now at Inova in Virginia looked at 1.3 million patients in the CAF PCI registry undergoing PCI and found that there was a tremendous amount of variability by sight and who gets transfused. Some sites weren't transfusing anyone undergoing PCI and some albeit a minority of sites were transfusing patients before PCI 8% of the patients or more of the patients undergoing PCI. So you can see there's this tremendous amount of variability in practice probably driven by the fact that there isn't a solid evidence base to guide us. When you put together all of the observational data or at least the largest observational studies being done a lot of these were done by our group again Matt Sherwood study we've looked at patients with acute coronary syndrome and found again that there's a strong relationship not causal perhaps, but an association between the aggressive use of blood transfusion and an increased risk for mortality as well as other ischemic events. And in our study in non-ST segmental elevation ACS patients what we found was that the relate that relationship existed down to a hemoglobin of eight. So in other words, if you transfused a patient and their hemoglobin was below eight there was no effect on mortality. If you transfuse the patient and their hemoglobin was above eight when you transfuse them that's where that manifestation of increased risk for mortality started being evident. So perhaps there are different triggers in the ACS population. Maybe it's not seven versus nine maybe it's eight versus 10 and that's something that we have to think about. The guidelines, these are the European guidelines previously said that you can transfuse an ACS patient if their hemoglobin is below eight. And in fact, I'll blow this up for you to see that recommendation. In patients with anemia and no evidence of active bleeding blood transfusion may be considered in the case of compromised hemodynamic status a hematocrit less than 25 or hemoglobin level less than eight. It's a 2B recommendation with a level of evidence C that level of evidence really does come from that observational study I just showed you that we did and published in JAMA in 2004. But the reality is these are inconsistent no adequately powered randomized trials. But despite that a restrictive policy may be considered and you can see how there's a lot of hedging in that language because there's no large adequately powered randomized trial. Now I had the privilege of participating in the clinical practice guidelines from the American Association of Blood Banks back in 2016, Jeff Carson led this and we actually chose not to recommend actually we chose not to make any recommendation at all. We couldn't recommend either for or against a liberal or restrictive transfusion threshold in patients with acute coronary syndrome because we just didn't have the data to support that recommendation. Kind of a side trivia note, we actually met here in Washington DC when we were developing this document and I remember very distinctly after that day all day of discussion we went to dinner and it just so happened that Anthony Fauci was sitting at the table next to us and I leaned over to one of my colleagues from Canada and I said, hey, that's Anthony Fauci and she said, who's that? And who would have thought that he would have such a huge role to play just a few years from then? So let's talk about this. Why would transfusion potentially be harmful to you? We know why it may be beneficial, right? Again, hemoglobin is part of the oxygen delivery equation to increase the hemoglobin theoretically you should be increasing oxygen delivery. Well, it turns out that stored red blood cells are not the same as the red cells that are floating around in our arteries and veins. Now I'm not a hematologist, I'm not even a basic scientist but Jonathan Stamler when he was at Duke spoke very slowly so that I would understand this and it turns out that stored red cells are low in two, three GPG so they don't actually deliver oxygen very well. They have a high oxygen affinity. In fact, they're depleted of nitric oxide almost immediately after they get stored and nitric oxide plays a fundamental role in oxygen exchange. I learned that respiratory exchange is about oxygen carbon dioxide but there's a third gas involved. So in the vein the hemoglobin molecule exists as a T isomer and in the lungs it takes up oxygen as well as nitric oxide to form something called S nitrosohemoglobin and it changes the confirmation to the R confirmation and at the level of the capillary and the deoxygenated organ the first thing it does is it releases the nitric oxide which causes capillary vasodilation and facilitates the delivery of oxygen into that tissue. If you don't have nitric oxide you impair oxygen delivery and maybe that's the lesion within stored red cells that may explain the fact that there's been it's been very challenging to show a benefit of blood transfusion in various patient populations. So packed red cells have something called the storage lesion which is characterized by depletion of nitric oxide which makes them function as nitric oxide sinks which can lead to vasoconstriction, platelet aggregation and ineffective oxygen delivery. It is also associated with increases in CRP and IL-6 and packed red cells may also exhibit procoagulant properties. The longer you store the red cells the greater the storage lesion. Now what's interesting is let's think about maybe we can give more fresh blood. So instead of blood that's been stored in the blood bank for a long time there's been a big push to say let's use fresher blood. Maybe we can overcome the storage lesion. Well, here's a randomized trial looking at critically ill adult patients randomized to fresh versus standard issue blood. Shockingly, there was no difference. So it sounds like once that storage lesion occurs there's probably no reversing it. And in fact, there was no benefit in this particular population of giving them fresh blood over standard blood. Similarly, this is one of my favorite randomized trial. This is Nancy Heddle's trial that she did in Canada. It's a very, very pragmatic study. These patients, only patients with type A or type O blood were randomized in this study. And the reason is that those are the blood types where blood sits on the shelf for a short amount of time or for a long period of time. So it's almost a natural randomization. And these patients were randomized in the blood bank in a one to two randomization to either receiving short-term stored red cells or long-term stored red cells. And what they found in fact, and I should also mention that the entire follow-up of this study was done through the electronic health record. So again, one of the first pragmatic trials to be done, you can see a very, very large sample size, almost 25,000 patients. And what she found was no benefit to new blood over older blood with respect to in-hospital mortality. So this whole idea of using fresh blood to overcome the potential adverse consequences of blood transfusion really has not panned out. And really it doesn't really matter if we make an effort to transfuse new blood, it makes no difference at all to the patient outcome. So now I showed you the slide previously, right? Because this is the trick trial. And again, to remind you, no difference in restrictive versus liberal transfusion strategies and critically ill patients. If you look at the subset of patients who had quote, cardiovascular disease. So only 257 patients in this trial had ischemic heart disease. Very few patients with acute coronary syndrome. In fact, a lot of these patients had a heart failure. You can see that the curves reverse themselves. So all of a sudden now, maybe the liberal transfusion strategy starts looking a little bit better. It's a subgroup analysis of a overall negative trial. So you wouldn't expect there to be a significant difference. But the change in that the curves is really intriguing, suggesting that maybe there is something to a different threshold in the patient with ischemic heart disease and that seven versus nine may not necessarily be what we need to think about in these particular patients. Well, you all are probably familiar with this trial, the reality trial, which is presented at ACC 2020. This was a trial led by Gabriel Steg and Gregory de Croc in France where they took 666 patients with acute myocardial infarction and randomized them to a restrictive versus liberal strategy. Their hemoglobin cutoffs were eight versus 10. Again, based on the observational data, eight versus 10, not versus seven versus nine. Now the primary endpoint of this trial was actually not a clinical endpoint. It was cost-effectiveness. I suspect that the journal editors made them put in the clinical outcomes and what they found at six months in short-term followup at 30 days was really no significant difference. So they were not able to show a benefit of a restrictive strategy or a liberal strategy in patients with acute myocardial infarction. They did follow the patients out to 12 months and look at this, the curves crossed. They crossed at about the four month time point. Now it looks like the liberal strategy is starting to look a little bit better. Again, not significantly better because again, this trial was not powered for clinical endpoints, but really an intriguing crossover of these curves. You can't really rely on this because again, it's not powered for clinical outcomes. The 12 month outcome wasn't the primary outcome. So we still don't know, but again, a lot of circumstantial evidence starting to really support equipoise in this particular patient population. And you can see here that the hazard ratios again, shown for you at less than five months versus five months. Now we did a pilot randomized trial at the Durham VA called CRIT pilot where we randomized about 60 patients with acute coronary syndrome to maintain a hemoglobin of eight versus 10. Jeff Carson did a smaller study called Mint where he randomized 110 patients with acute coronary syndrome to the same transfusion cutoffs. Our study showed potential harm with a liberal strategy. His study showed potential harm with a restrictive strategy. So again, perfect equipoise. And Jeff was able to use that to leverage NIH funding to get the Mint trial funded. Now he called this pilot study Mint. He calls this big study Mint. I thought he should call the pilot study Junior Mint, but we have different senses of humor, I guess. But Mint trial is ongoing now. It's 3,500 patients. This is going to be the definitive study. Looking at what the optimal transfusion threshold is in patients with acute coronary syndrome or randomizing 3,500 patients with STEMI or non-STEMI, type two MIs are allowed to a restrictive strategy, maintain a hemoglobin of eight versus maintaining a hemoglobin of 10. And the primary endpoint is at 30 days where over 2,600 patients into this trial, this is a very, very challenging trial to enroll in. It's not your standard acute coronary syndrome trial. These patients are very sick. If you think about the patients who have ACS in your particular hospital right now that are anemic, they are elderly, they have renal failure, they have cancer, these are very sick patients. And a completely blinded look at our event rates seem to show that we are at least on target with our event rates. And potentially, these are patients who are going to have the most of benefit from the results of the study. For our sites that are participating in this, I wanna publicly thank them. It's a very challenging trial to enroll in and we hope to have data within the next few years. So let's switch gears now because we don't know what to do about bleeding. We don't know what to do about transfusion. Maybe prevention is the best thing. And so let's talk about some prevention strategies. And again, the strategy I wanna focus on is really radial access. For those of you who know me, you know how I got started with radial access really pushed by the fellows who realized that if they go radial, they don't have to stay late to pull sheets. So one of our more enterprising fellows who ultimately went into EP said that, I came into work one day, the board was full and he said, we're just gonna do all these patients radial. And I said, okay. And that's how we got started. Now, we looked at close to 600,000 procedures in the CATH PCI registry to look and see, what is the overall prevalence of radial access in 2004 to 2007? And what were the associated outcomes? What we found was that there was no difference in procedure success between radial and federal, but there was about a 60% reduction in bleeding and vascular complications. And despite that potential benefit, radial access at that time between 2004 and 2007 in the United States only accounted for about 1.3% of all PCI procedures. The vast majority of hospitals were doing 0% radial at that time. And it was around 2008 or so where the rival trial was starting up. And I had the privilege of being the US National Coordinator for the rival study. For those of you who don't, who are not familiar with rival, rival was a 7,000 patient randomized trial of patients with acute coronary syndrome who were undergoing invasive risk stratification, randomized to undergo radial access versus femoral access for their diagnostic calf. This was a sub-randomization within the larger current Oasis 7 study. That study was looking at different doses of clopidogal versus aspirin. So really we had to use the end points that were part of that study, which was a composite of death MI stroke or non-cabbage related major bleeding in 30 days. We also were forced to use the definitions. So in this study, to qualify for a major bleeding event, you had to have, let's for example, if you bled into your eyes, but you didn't go blind, that didn't count. If you bled, but you didn't receive a transfusion, that didn't count. So these are major, major bleeding events that were occurring in this particular study. Surprisingly, we were shocked. We were shocked because there was no difference. We thought for sure, radial was gonna make a difference with respect to the bleeding end point. But what we learned was something that I told you at the very beginning, which is that most of the patients enrolled in this trial were non-ST segment elevation MI patients and the vast majority of their bleeding events were not related to the vascular access site at all. So in fact, our intervention was not gonna actually show a difference. In the patients who had ST segment elevation MI, which comprised about a third of the patient's randomized, there was a significant reduction in the primary endpoint. And then when we looked at a pre-specified endpoint of major vascular access site complications, a very, very vascular access site sensitive outcome. That's where we started showing a benefit and over 60% reduction with radial access versus femoral access. In fact, the follow-up of this trial was 99.9% complete. Only one patient out of 7,000 was lost to follow-up. The US enrolled about 180 patients in this trial. We were the ones who lost someone to follow-up. And I remember hearing that statistic and I leaned over to one of my colleagues from France and I said, wow, it's really embarrassing. I mean, we put very few patients into this study and we're the ones who lost someone to follow-up. And without missing a beat, he said, that's the risk you take when you do trials in third-world countries. So here we have a pretty solid body of observational data, at least one trial that shows a benefit with respect to vascular access site complications. The question is, why are the rates so low? So with Christian Helfer to the Pewds Sound VA, we surveyed 200 interventional cardiologists in the United States. And what we found was that they identified two barriers to radial adoption in the US. The first was there was no US-based trial data. So a lot of people looked at rival and said, well, that's not relevant to the way I do business because I'm not an interventional cardiologist in Europe. I'm in a US interventionist. And the second was that there were no training opportunities for people post-fellowship. We held a think tank at the FDA. And you may wonder, why is the FDA interested in this? Well, they were interested in this because they thought that it was an approach to make antithraumatic therapy safer. So we held the think tank. It was open to the public. The NIH was there. Lots of patients were there. Lots of practitioners were there. And at that day-long meeting, the outcome was, you got to do our US-based randomized trial of radial versus femoral. Easy, right? Well, it's not so easy. The number one challenge is randomization. At that time, people who were doing radial access were doing it for a reason. Maybe they had a bad experience. They may be unwilling to randomize to femoral. Someone who was not very well trained in radial may in fact be unable to randomize to radial. The second challenge was funding. This is not a new gizmo. It's not a drug. Who's going to fund this trial? We needed a pragmatic way to do the trial. And the third challenge was finding those cardiologists who were proficient in radial access who were also willing to randomize. Now, I filled out site surveys. You filled out site surveys. We're all challenged by recall bias. At that time, if a cardiologist said that they were doing 50% radial approach, they were probably really doing 30%. So now it just so happened that my mentor at the time, Bob Harrington, had received funding from the NIH to build randomization into the CAF PCI registry. So a very pragmatic approach to randomized trials. As part of that grant funding mechanism, he had to start a randomized trial by the end of the grant period. And this was the perfect question to study. Now we had to find who the patients were. And what we found after talking to a lot of thought leaders in the space was that females were the ideal patient population to randomize. First of all, women are significantly underrepresented in all of our interventional trials. The same goes for radial trials. And they present a unique challenge. Almost every model of bleeding shows that females are at increased risk for bleeding, but they have smaller radial arteries, potentially higher procedural failure rate. So that's the first challenge. A radial operator would be willing to randomize a female patient to either radial versus thermal because we don't know what's better. Particularly if we have procedures success or access site crossover as part of the co-primary endpoint. The second challenge, we did have this platform for a cheaper trial. Now we could use this National Cardiovascular Research Infrastructure, so-called NCRI, which embedded randomization of the CATHPCI registry to actually do this study. And what this does is leverages the workflow of the registry participants into a very streamlined way of collecting data. We could also go to that registry and find out who's doing radial access in women. And I'm showing you the exact data. We said show us the sites that are actually doing radial approach in females. And this is the data output. When we took these first 60 sites and made them our sites for our randomized trial, some of these sites had never participated in clinical research previously. So let me show you how this worked. A site participating in our trial identified a patient, consented them, randomized them within the database. Now as part of the normal registry workflow, this site ended all the demographics, medical history, procedural data into the CATHPCI registry. They were doing that anyway. The back office software built through the grant funding would auto-populate a clinical trial case reportable. So when the study coordinator went to fill it out, 60 to 70% of it was already complete. In clinical trials, we pay for time. So this reduced the time that the clinical study coordinator had to spend filling in data. They would fill in the only the unique pages for the trial and that created the analytic database. Two of the sites that weren't participating, participating in the CATHPCI registry filled out all the variables manually, hats off to them, it was a tremendous amount of data. And this is what led to the safe PCI for women trial. This is still unfortunately the only randomized trial performed solely in women in the interventional cardiology space. We randomized 1,800 women to radial versus femoral with a primary efficacy endpoint of bleeding or vascular complications and a primary feasibility endpoint of access site crossover. Here are the results. We found a 70% reduction in bleeding or vascular complications at the expense of about a 6% access site crossover, mostly for radial spasm. Now that's interesting, but what's more interesting for the clinical trialists in the audience is that we did this 1,800 patient study with clinical endpoint adjudication for only $5 million. And as all of you know, when you participated in trials, 80% of your patients come from 20% of your sites. In this study, we turned that on its head, 97% of our sites enrolled at least one patient and 71% of our sites enrolled at least 10 patients. This is just now a body of literature that has almost 19,000 randomized trials showing that radial access not only reduces major bleeding compared with femoral access, but also reduces mortality. Again, to take you back to the beginning of the talk, mostly in the STEMI population who have the most to gain from radial access. Even in the Safari trial, which was powered specifically for mortality comparing radial versus femoral, although stopped early because of lack of funding, they showed no difference. But if you look at the trial endpoint of bleeding, there's a very, very intriguing separation of curves. Again, suggesting a benefit of radial access and to Michelle LeMay's credit, he did do a meta analysis of his study Safari along with the other 13 studies, almost 11,000, almost 12,000 patients. Again, showing that there is a significant reduction in mortality favoring radial access, again, driven by the STEMI population. So how has radial access changed in the United States? We were 1% in 2007. While I'm pleased to say that in the second quarter of 2021, radial access accounted for 63% of all diagnostic cats and 54% of all PCIs in the United States. We've done this over the period of about a decade, a little shorter than the 200 years it took for the British Navy to put sauerkraut and limes on all their ships. And I think this proportion is gonna continue to increase as fellows who are trained in radial access go out into practice. We also now have a guideline recommendation. You all are probably familiar with the revascularization guidelines. We now have a class one recommendation with a level of evidence A for ACS patients and for stable ischemic heart disease patients to reduce the risk of death vascular complications or access site bleeding. Now we're of course in the DES era. So now we're gonna shift our bleeding risk to the outpatient setting because now patients we need to take dual anti-platelet therapy for a prolonged period of time. And so the bleeding issue has not gone away. It's just shifted from inside the hospital to outside the hospital. And I was pleased to be part of the author group that came up with this high bleeding risk criteria specifically related to dual anti-platelet therapy after drug-illuding stents. This is led by Phillip Urban. You can see major and minor criteria. Let's go back to the patient that we described at the beginning. This patient has the following high bleeding risk features. She's older, she has renal disease, she has anemia and she's on oral anticoagulation. So now we have to adapt our strategies not just for in the hospital for what happens after she leaves the hospital with respect to treatment approaches. So here's what we did with this patient. We went left radial. She had a significant LED lesion. Gender went intravascular ultrasound guided PCI in placement of a drug-illuding stent. She was treated with dual, actually not anti-platelet therapy with dual antithrombotic therapy with just the oral anticoagulant and clopidogrel, no aspirin as the trials have shown. Guideline directed medical therapy for secondary prevention. And amazingly her anemia workup showed that it was likely nutritional. About a third of anemia in the elderly is related to nutritional status. And I have a personal stake in this. My mom actually had nutritional anemia. And so it is something if you notice that in your patients, particularly if they're older, make sure you do a complete workup so you can identify the cause and get it treated. Post-PCI bleeding complications are common and they're associated with increased morbidity and mortality. I didn't show you the cost data but take my word for it. So I think as cardiologists, it's incumbent on us to try and reduce both the ischemic and bleeding risk. We don't actually know yet what the optimal transfusion threshold is in the patients with acute coronary syndrome. Lots of observational data, some pilot data, but the ongoing mint trial will really be the definitive answer helping us to try and guide transfusions in these patients. With respect to prevention, radial access is really the most effective strategy. And we really all should be using a radial first approach in our patients. And now the bleeding risk is shifting from procedural risk to long-term risk and antithrombotic therapy after PCI. But this is a constantly moving and shifting space, stent platforms, anti-platelet strategies are all evolving to try and reduce long-term bleeding risk without sacrificing efficacy. So again, it was a huge honor for me to present at the annual Lewis lecture. I wanna thank Honey for the invitation. I wanna thank all of you for dialing in.