Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

All right. So I'm very excited for our next speaker. Um, Doctor Lisa Moos was kind enough to uh accept the invitation when I first sent it out to her. I wasn't sure. Um And I was ecstatic when we got the yes. Um an introduction for Doctor Moore would take um probably the entire time that uh you're gonna be speaking, but just to give a few highlights here, um she um went to medical school at the Uniformed Services University of Health Sciences in Bethesda. Um and then did her Im and PMRI fellowship at Walter Reed. She's now a professor of medicine um at the Edward Herbert School of Medicine of the Uniformed Services, University of Health Sciences. She serves as the Associate Dean for Assessment and Professional development and she is a retired colonel of the US Army Corps. Um I first came across Doctor Moore as a medical student when I was handed the print version of the uh chest VT E guideline guidelines by a hematologist who told me this is the Bible for what to do. Um And I try to attend as many talks as I can by her whenever she's taking a chest. Um She has held more positions at chest than I knew existed. I was looking at your CV the other day and it goes on for about two pages. Um So the dedication to service is really incredible. Um And um more than 80 peer reviewed publications. So it's with a great pleasure. Um That Doctor Moos is gonna come speak to us. Thank you again. Thank you. II think he set me up twice here now with that introduction and trying to follow a who has all those cool videos. I can't win. So, um, I guess I can only go up from there, right? All right. Uh I also have no conflicts of interest. Um, I'm from the government, I'm here to help. There are my objectives. I won't read those for you. So we're gonna start with a case that should be familiar to most of you. This is a 52 year old woman who comes in, uh, complaining of six hours of acute onset of left-sided chest pain. She has a past medical history that's significant only for hypertension that's been well controlled on medications. Uh, nothing recent, but she did say about two weeks ago, she took a long haul flight that was greater than uh 12 hours on exam. She's normotensive 1 35/85. But in asking her, that sounds a little bit low for what her baseline has been with her chronic hypertension and she's tachycardic to 100 and 10, her respiratory rate is 22 and her oxygen saturation on room air is a little bit low at 94%. They do a chest X ray which is unremarkable. Uh D dimer is markedly elevated. And so they go on to CT which does confirm uh multiple segmental pulmonary emboli on the left side and you can see an elevated RV to LV ratio. Uh in addition to some other findings, if I were to show you some additional cuts, a little bit of reflux into the IVC. So the question is how to best manage this patient. And this is the one that um I guess has kept me in business, right? Uh This has evolved, I thought it was so simple when I was a resident and then a fellow and getting called for patients, I'm like just read the guideline. Uh and, and then you realize that none of these patients read the guideline, right? So they don't follow and it really has evolved into trying to understand how to take this information at presentation and risk stratify the patients a little bit further understanding that they're not all the same. And in fact, these patients in this, what we would call intermediate risk category that you see in the center of the slide are the ones that are most difficult for us to uh determine their clinical course when they walk through the door. And that's because we understand there is just such this wide spectrum of disease in, in pulmonary embolism. And luckily for us, about 70% of them will do absolutely fine. They have normal hemodynamics, their right ventricle works well and they're gonna do fine with just anticoagulation. And in fact, a proportion of those in recent studies, anywhere up to about 25% of them can be treated entirely as outpatients. If they do need to come into the hospital, it's usually a short 1 to 2 day stay before we can send them out. There. Is that 5% that comes in, that's very easy to identify as well. They're very high risk. They come in and they've had syncope or sudden cardiac death. They're significantly hypotensive. I don't know that we figured them out yet either. To be honest, it's not the focus of today's talk, but I don't think all high risk patients are the same either. Uh and we need to do a better job of understanding how to approach them. But what we do know is we have to be aggressive. We need to go straight to reperfusion. It's this 25% of patients that uh are, are giving us a little bit of pause, right? And similar to the one that I presented to you and these are patients that look ok. Their hemodynamics appear not to be significantly altered. Um Their oxygenation may be a little off but not, they're not significantly hypoxemic. They seem to have some RV, dy dysfunction. But, but again, overall, they, they look, ok, but some of them are gonna rapidly deteriorate and some of them are actually gonna do fine. They're gonna look like those 70% and it's trying to understand how we can better sort those out and why are we? Let's see. All right. Seems to be stuck here on the arm. I wanna go. It's advancing here. You're trying to go here, right? Try that. Maybe I'll talk to her. Let's see the way that works. OK. Thank you. Um So we're focusing on the right ventricle because we know that patients with any degree of RV, dysfunction are going to have a worse outcome than those who do not. And the physiology makes sense. This is a graphic that uh is from the NT 2019 European Society guidelines. Um I hate to show guidelines other than chest, but uh I will say that these are, these are well done as well. Uh And I really do like this graphic and it points out the fact that when you have that acute increase in RV, afterload, the first thing you're going to see is RV dilatation. And we measure that by that RV to lb ratio on CT, we can measure it with echocardiogram. Uh And we can look for hormonal evidence or biochemical evidence by the uh N terminal PRO BNP. If this goes on that stretch, of course, leads to increased RV, wall tension ultimately to RV ischemia. And so we start to look for measurements such as the elevated troponin suggesting that that is happening. And as this continues, of course, you see increa or decrease in the RV contractility decrease in the RV output, which then with the interdependence and the RV overload, pushing the interventricular septum into the LV. We start to see decreased LV, preload, decreased LV output. Um and then of course cardiogenic shock and death. So, trying to disrupt this cycle and identify that RV function early is what we've focused on and further risk stratifying. So if you look at the current classification systems, one is the American heart association chest doesn't actually have its own classification system. Uh But we would describe these patients similarly to the American Heart association. And that's looking at massive submassive low risk pe other terms, it can be high risk, intermediate risk, low risk. And you can see that they are similar in their approach. In that we look at hemodynamic status, we look at prediction scores in these classification systems. We're using the pulmonary embolism severity index or the simplified version, the S Pezzi. And then whether or not we have evidence of RD RV dysfunction. And you can see that these two classification systems are in agreement. Although with the European Society of cardiology, you see they are now subdividing that intermediate or submassive group into intermediate low and intermediate high. The intermediate low group is one that has stable hemodynamics, some evidence of RV dysfunction. And they classify that as either RV enlargement or um elevated BNP or troponins. The high risk has both. So this is a, a population that is further along that circular decompensation curve and therefore more worrisome. Now they use the Pezzi and the, the simplified Pezzi. I was part of developing those. I'm a big fan as well, but I would tell you that it's not that they're the only one out there and there are other risk scores. You can use. The important thing is to use some type of baseline score to ask yourself. Is this someone that I need to worry more about? Ironically, these scores are very good at telling you that patients are gonna do. OK. So if they have a low Pez, a low simplified Pezzi, a low Geneva score, they have none of the hestia criteria, they're going to do fine. In fact, these are the patients that can probably be discharged. We know that patients that then have some of these markers in these scores are probably not gonna do as well, but they're not their specificity just isn't good enough. We can't look at them and say with any certainty that this is a patient with intermediate risk that may deteriorate and may require more aggressive therapy. Um So right now, what you'll see, this is a nice summary and a recent review by Greg Piazza that just shows you general approach, right. Patient with a pe can they be anticoagulated? That's obviously what we'd like to do upfront regardless if they can't. Then we need to think about putting an IVC filter in if they can we start that. But then we have to say, do we need to do more than that? And for the low risk, they're fine. No, we don't need to do anything. And probably for many of those intermediate and whether or not um those need to be in the hospital again, varies a little bit by society. You look all the way over on the right. And these high risk, we need to consider some type of reperfusion right up front. And you'll notice with that higher intermediate risk, you're starting to see the words, well, consider, consider escalating therapy. If you look at the European Society of guidelines, they have a very similar approach. So if they have evidence of RV dysfunction, they then look for evidence by biomarkers. So whether there's an elevated troponin or BNP, their guidelines emphasize the troponin again, just because that's a little further along. And you can see if they have both RV dysfunction and uh evidence of elevated troponins. They again are saying, consider being a lot more aggressive, consider treating them like a high-risk patient. And when you look at our most recent chest guidelines, uh not really that different. We do say low risk can probably be treated at home, high risk systemic thrombolysis upfront. And in these intermediate risk, then we say start with anticoagulation. But in the select patients who deteriorate and think about escalating to reperfusion. And our recommendation in the current set of guidelines is to do that with systemic thrombolysis as opposed to catheter directed therapies. Now, that's based on the literature up through 2021 I'm gonna walk through whether that's changing yet. But this is the current recommendation and the reason for this is that there really are no randomized controlled trials that compare CDT to systemic thrombolysis significantly or CDT to anticoagulation in large patient populations. Um So we did not feel comfortable recommending that in, in 2021. But you probably were, are asking yourself, all right. So why are they saying consider in this group of patients? And it's really based on a single trial? Um uh Most of you are probably aware of the pyot trial. This was the second trial that looked at systemic thrombolysis versus anticoagulation in patients with acute pe. The first was about 10 years prior to this one, maybe 15 and that looked at higher risk patients. The pho trial was designed specifically to look at that intermediate high risk group. So evidence of RV dysfunction, elevated troponin and then they were randomized to, to neck to place in this case uh versus uh placebo which was standard anticoagulation. The primary outcome was similar to the older study which has been a criticism for many of us that the primary outcome was a combination of all cause mortality and the need to escalate care and their secondary outcomes then looked at mainly safety. So serious adverse events, they didn't really look at recurrent uh VT E they fo focused on major bleeding and intracranial hemorrhage and the need for escalation on its own. And in terms of the primary outcome, systemic thrombolysis was superior, this was similar to their prior trial. So now they're saying, well, now in this intermediate high risk, systemic thrombolysis gives you a mortality benefit. But again, remember this was a combined mortality and escalation of care. And when you break it down, this was entirely driven by escalation of care. There was no difference in mortality between systemic thrombolysis and anticoagulation. And you can see that the biggest thing that led to escalation in care was uh the need for CPR developing hypotension. So this clinical deterioration, it got people nervous and they went on to secondary thrombolysis. And when you looked at those safety outcomes, you have nonintracranial major bleeding, which was higher in the patients that got systemic thrombolysis. And of course, the rate of intracranial hemorrhage and stroke was also higher. So similar to other trials, you might have an outcome benefit. Although mortality was not significant and you're paying for it with increased intracranial hemorrhage and major bleeding, which is why none of the guidelines say start with systemic thrombolysis in patients with intermediate uh pulmonary embolism it makes you wonder even whether we're doing any benefit in these high risk. But when you're watching them come in and they've already had sudden cardiac death, we need to do something. I think again, that that area is, is evolving. What's driving all those guidelines to say, consider this reperfusion in those patients is that the patients that did deteriorate in the PA trial and then got secondary thrombolysis, almost 90% of them did very well. So it worked even if you waited a day or two to give the lytic therapy and those patients were discharged home, saying to uh you know, again, translating into the guidelines of, well, don't, don't, don't be too aggressive upfront but watch them closely. And if they deteriorate, then consider secondary lytics. And the problem with this is who are those patients? Right? Um We don't know really how to identify who might be more likely to deteriorate because that's the patient group that I would put in the ICU or at least an intermediate ICU. I would be watching them very closely as opposed to they're in that intermediate risk group. But I don't think they're gonna deteriorate. I think they're gonna do OK, I can probably put them on the ward. II wouldn't send them home. But how do I know which patients to be most concerned about? And we really, we don't have a universal definition of what clinical deterioration looks like. Is it a certain drop in their BP or their mean arterial pressure? Is it an increase in their heart rate, their respiratory rate? Is it decreasing oxygenation? Does it have to be a combination of all of the above? And as I mentioned, the current tools we have that we think stratify outcomes like the pulmonary embolism severity index are just not specific enough. They don't tell us which of these patients is necessarily gonna deteriorate. And so some of the more emerging research is trying to figure out who are those patients. This is a II think a nice summary in that paper I referenced earlier from, from Doctor Piazza that shows that real spectrum in there in the middle of, I've got this patient with acute pe that isn't clearly low or clearly high risk. Now, what types of things do I need to look at? So we can list them fairly easily, we can look at their heart rate, we can look at their BP, right? Either systolic or mean or both. We definitely wanna focus on the RV. So do we have enough information from the diagnostic CT or should I get a well done echocardiogram to look at that? The TSI certainly is something that can be helpful if you can get that, I'm gonna add in these biomarkers of the troponin and the BNP. And then I also need to think about bleeding, right? If I'm going to be more aggressive, how high a risk is there bleeding. Because if I can't give that anyway, does that change that whole algorithm? And where I've got the patient and a couple of other things to think about that he doesn't have in that diagram. But uh the, the heart type fatty acid binding protein is something that has shown some promise in detecting patients that are at higher risk. And we know that patients that come in with concomitant DVT in addition to this acute pe are also at higher risk. So another marker, but how do I put all these together? You know, and II really liked when he added this portion in of these unmet needs and and I would agree entirely, there are several unmet needs in this heterogeneous group of patients. So identifying those predictors of early deterioration. How long do you monitor them? So you say, OK, not gonna be aggressive upfront. Do I reassess them in six hours? 12 hours? 24? Can I let them go? 36. Is there some point where secondary reperfusion is not as effective or may be associated with higher adverse outcomes? I don't think we really know that identifying those patients again. Um that may need that more aggressive. Still doesn't answer a question of in today's age. What is that more aggressive? Do I give them lytics? Do I go to CDT? Do I do Surgical Thrombectomy if I have that expertise? Um And then how do I know that what I'm doing? Is working. So to me, the holy grail is right there, how can I take the information and accurately assess who needs more aggressive therapy upfront? And I think it's gonna morph a little bit as we look in the next five, maybe 10 years of not only in that intermediate group but the high risk group who's going to need what therapy right upfront, we have a couple scores. So my current practice, if I got someone that I'm bringing in and I'm concerned that they might deteriorate, I used the Bovis score. It was one of the first to come out. That's a European score. And what Doctor Bova suggested is rather than looking for that sustained hypotension below uh a a systolic of 90 looking at that in between of 90 to 100. Um Although I would say from a practical standpoint, I tend to look at how far down are they from their baseline BP um to, to help me any tachycardia above 100. Again, the evidence of RV dysfunction, which most of these patients have by their initial definition, you can see there are some other scores that have been suggested, they're using similar parameters. Although starting to bring in some measures of hypoxemia, which wasn't something we really focused on early and you don't see it in the, in a, in a lot of the risk scores, which is, you know, interesting. But I think early on, we realized that looking hemodynamically and how the RV was doing was a lot more prognostic than the degree of hypoxemia. And we could support the hypoxemia. But I think we're realizing that we were overlooking that as a, as an important marker uh and looking at measures of perfusion. So, bringing the lactate in is something that's a little bit newer and, and not included in the Bova score. So, in reality, I'll look at a boba score, but I'm also looking at the lactate. I'm also looking at the degree of hypoxemia. Um and, and I can't give you any systematic way to combine these. I think you have to look at all of them right now. There's another score that's being used in what's called the high P pho trial, which is looking at um the I get the high po and po three confused and they're both looking at the same using the news score, one's using half dose lytics, one's using CDT. Um But they're using the news score which you can see has similar markers to the other scores that I just showed you. Uh but they also bring in level of alertness. Um So another measure of perfusion and there are a couple of studies trying to look at early predictors or potential predictors of early clinical deterioration. You can see the first one here is interesting in that it's sort of the opposite of how do I know they're out of danger. So, the normalization of the heart rate within 48 hours. Um inpatients that were on anticoagulation had a decent predictive value. The next two studies there look are really looking at the opposite, what might identify patients that are at highest risk for that clinical deterioration. So they have some actual numbers here rather than just the presence or absence of RV, dysfunction. Looking at the actual right ventricular basal width, an actual cut off, that's a little bit higher than many studies used for the RV to LV ratio. Getting that TSI measurement by echocardiogram um as well as additional markers that can be very helpful from echo. And then you can see a final study there uh from Marco Zen say, looking that a initial mean arterial pressure less than or equal to 81 was an independent risk factor for 48 hour clinical deterioration. So we're starting to see a little more quantitative metrics on how we might identify patients most at risk. And then this was a study from uh Bush Romina and colleagues that was uh came out of the flash study for the registry which um has shown us quite a few things. So in the flash registry, these are patients that are being enrolled prospectively and followed, who are going to get catheter directed therapy. So this is not a randomized controlled trial. All of these patients are being treated with catheter directed throm uh thrombectomy or thrombolysis. And because they're doing this though, they have the opportunity to get hemodynamics. And they asked the question, how are these patients doing when we actually put a right heart cath in? And interestingly in these patients who by definition are intermediate risk in the flash registry. So they are normotensive. But up to a third of them actually have evidence of shock on invasive monitoring, which they define as a cardiac index less than or equal to 2.2. So they tried to determine, can we figure out who these patients are? Uh and they came up with a proposed composite shock score that looked at the troponin and BNP being elevated, moderate to severely reduced LV function, which they defined um based on either uh echo parameters or the uh RV to LD ratio presence of saddle embolus concomitant DVT, which we mentioned earlier and tachycardia and what they noticed when they did this was that these were I independent markers. But then, of course, as you added them up, they were predictive of the presence of normal tensive shock where almost 60% of the patients, if they had all six of those factors had normal tensive shock and certainly would be a group that we would be more concerned about and maybe should get more aggressive therapy right upfront. Other characteristics of these normal tens of shock patients is they had lower BMIs. Um not only was the heart rate above 100 but significantly higher. And as you looked at cos of the higher RV to LV ratio and reduced RV function. That of course also correlated with the evidence of uh normal tensive shock. And they made the interesting comment that 17% of those patients with normal tens of shock had a simplified pulmonary embolism severity index of zero, which would be low risk. So they come in with a normal BP if they had a pez A an sp of zero, some patients, some people might send them home and the guidelines would support that. And yet, here's a group that had subclinical shock. They don't define those 17% very well. The S Pez doesn't have a lot of comorbidities. So, or these patients that really have no reserve, right, have underlying cardiopulmonary disease or something else that put them at higher risk. They don't define those. Well, I did see another study um not as well done retrospective but did have almost 200 patients um from the Wellstar Health System who had undergone invasive therapy for their intermediate high risk pe and they found a higher number. So 44% were in cardiogenic shock, uh which was associated with increased risk of mortality. And they found a few other factors here associated with that um pre presentation with syncope. And again, in most guidelines, we would quantify or classify those as high risk right up front. Um But the elevated respiratory rate, a right bundle branch block, some similar markers. They did include the lactate and here they found an absence of deep vein thrombosis associated and I don't know if that's because there's been more clot that is already embolized and you don't have residual clot, um, hard to know. But interesting when you see the difference between the two and some of the more traditional markers, some of which were included in, in, um, Boucher's paper were not predictive. So, I don't, I don't know what to do with that yet. I II think again, looking at all of these and as I'm ticking more and more and more, I'm getting more and more and more concerned. So what do we do? What are our options for therapy in these intermediate patients and particularly these intermediate high risk? And thi this is what we have today. Of course, we can always go with systemic anticoagulation. We can use systemic thrombolysis, we can use reduced dose thrombolysis, we can do any number of catheter directed approaches and in many centers, Surgical Embolectomy. And I don't know um those of you that follow the pe literature might have seen an interesting statement from the American Heart Association, really pushing for primary surgical emb embolectomy in more patients. Um With the argument being that modern surgical techniques are so safe that looking at all of the data on surgical embolectomy is really not helpful because it's using very, very old studies where mortality and morbidity were higher. Um They don't have any data to say you should do that. But it's sort of a call to say, are we under treating some of these higher risk patients? And I certainly in the time that doctor Green, the only time has given me is I can't go through all these studies for you. Right. There are so many studies coming out looking at different, uh, groups, different types of catheters. They are almost all observational. They're not great. Um, so the best I can do for you today is to give you a network meta analysis because they really haven't been compared to each other in a head to head trial. So trying to make some sense of all of this data that we have to this point, I'm gonna show you two different network meta analysis. Um And they used slightly different techniques. I was talking to Doctor Dass last night. Um II had to go to chat GPT to take a lesson in statistics because II couldn't understand the difference between these two who used a similar name. They called it a frequent test network meta analysis, but they use very different ways of showing their results, different statistical analyses. Uh And so trying to sort through them. Um And I will try to balance that for you. But this is the first one and both of these were published about six months ago. And in this particular one, they did again, a very nice systematic review. These were done with uh Prisma diagrams, they were registered on prosparol, they were done, they were done well as best as I could tell in doing the literature search, they included randomized controlled trials and prospective observational studies. You're really not gonna have much in CDT if you it to just randomized controlled trials and this statistic you're seeing here is called the supra. Um It's, it's an interesting one but basically, it looks at the cumulative probability that this treatment is the best treatment compared to other treatments when it compares them across and does these triangulation and weights all the studies. And so you can see in terms of short term mortality CDT S appears superior to anticoagulation or systemic thrombolysis. And it really doesn't come close to the line of identity when you look at CDT versus systemic thrombolysis. Uh in terms of systemic thrombolysis uh being worse CDT versus anticoagulation, you can see also doesn't cross the line of identity suggesting that at least in terms of short term mortality CDT would be the superior treatment when you look at major bleeding, CDT does have higher bleeding than anticoagulation alone, but lower bleeding than with systemic thrombolysis. So, suggesting that sort of in between that a lot of us are, are thinking about um and really no difference in the major bleeding between systemic thrombolysis and CDT. And I think that's an interesting observation. It's something we're seeing. Most people sort of assume. I think people that are big fans of catheter directed therapy say you can do this and you don't have bleeding risk and that isn't true, but it may be safer than systemic thrombolysis. Uh When you look at intracranial hemorrhage, no difference with CDT versus anticoagulation. And really the biggest problem again is what, what we already know is systemic thrombolysis. So for all cause mortality, uh catheter directed was associated with a lower risk and they did a subset where they restricted this just to patients with intermediate high risk and found a similar uh result. It was however associated with a higher risk of major bleeding. There were numerically higher intracranial hemorrhage, but this was not statistically significant. And again, it was lower than with systemic thrombolysis um which again had no mortality benefit when, when compared to either of these. And so if you're trying to balance the best treatment or lowest event rate for efficacy was catheter directed. But for safety was anticoagulation. That doesn't help me much. That seems to be what I already know when I look at some of these. Um but it is taking um all of the data to date this meta analysis included over 80,000 patients. So, is it giving us a signal? And I think the authors say it does support the efficacy, but they do give the caveat of this is largely driven by observational studies and that we need randomized controlled trials, which I would agree with. Um it was associated with a higher risk of major bleeding um and a higher risk of intracranial hemorrhage. But that intracranial hemorrhage, as I said was um numerically higher, not statistically higher. Now, the second network meta analysis used a different statistical but it shows it in a way that I think most of us are more used to looking at and they looked at all cause mortality, comparing anticoagulation, systemic thrombolysis and catheter directed and each to each other. And you can see for all cause mortality, you're seeing a similar signal in that catheter directed versus either of the other two was superior. In terms of intracranial hemorrhage. You see that systemic thrombolysis certainly is um associated with a higher risk whether you compare it to CDT or anticoagulation. Both of those have lower risks of intracranial hemorrhage compared to the lytic therapy with no real difference. Again, a signal suggesting that there is a maybe a lower risk of intracranial hemorrhage with CDT in this study. But again, the confidence interval crosses one and for major bleeding, no difference here with catheter directed versus anticoagulation. But again, you can see that there's a little bit of a shift that says CDT may have a higher. So the the results of these are fairly consistent, a little bit different ways of doing it. And this study, although it used the exact same search approach only included 21,000 patients. And I had a hard time figuring out how one study using a search approach had 80,001 had 20. And when you really get down into the weeds, this second meta analysis decided to exclude cohorts where the groups that they were looking at were s really different at baseline. Um Because again, these aren't randomized trials. So if they had groups that really didn't fit certain criteria or look different at baseline, they excluded them. This study also in trying to make, uh, their conclusions used a approach that should be used if you're doing actually evidence based guidelines and that's looking at the certainty of the evidence um and the weight of the evidence and some additional judgments. So going through what we call grade, but what grade includes is an evidence to decision. A algorithm I guess is probably the best way to put it where the whole group goes through and you don't look just at the evidence, you start with the evidence, but then you bring in other factors, like most importantly, certainty. How certain are we, what is the confidence interval? How good is the data? Are these good studies? Are they? OK? Studies. What's the cost effectiveness? What's the equity across? What, what are access issues? So it brings a lot of other things into the decision. But in this second meta analysis, they're saying with moderate certainty, which is pretty good, the risk of death and major bleeding, including intracranial hemorrhage was lower with CDT than systemic lytic. So I think we can feel confident in that signal and what we're seeing across both both of these network meta analyses. And well, for time sake, I did not include another one, but there was a third one out last year that had similar results. When you compare CDT with anticoagulation, CDT does have a lower risk of death, may be associated with a similar risk of intracranial and major bleeding. And again, moderate certainty. They also do note that this was driven mainly by observational data. So as I said, II had a hard time putting these two studies together just because of the way they're done. But I think when you think about the real question here, right? Is CDT ready for prime time? These are sophisticated machine learning when we do network ana uh analyses and we're making a lot of assumptions when we do it and they, neither one of them actually showed what algorithms they chose to use and just like anything we do with machine learning, the output is only as good as the input. I don't know how many of you like to play around with some of the large language models, but you put a, a prompt in if you don't do it, well, you're not gonna get what you want coming out and um the more you do it, you can learn to tweak them. II don't know how they decided to do some of these inputs. Um One of the things that still limits this approach is that all catheter directed therapy is lumped into one group and these catheters are not all the same, they're very different and they can be be utilized just for local but lower dose lytics, they can be used from a mechanical standpoint. So whether that's a suction or some other way of doing that, they can be a combination of the two, they can use ultrasound, putting all of these into one, II think is problematic in in interpreting it. And similarly, when you look at systemic thrombolysis, they're lumping all those studies into one. And some of our newer studies are using lower doses of thrombolysis. So I'm not sure we can do that. The second network meta analysis did look at surgical embolectomy as well, although they didn't include that in any of their final uh plots. But um that also I think is limited by the fact that if you go too far back, I II, that's gonna bias the outcomes in surgical um patients. So it may be more efficacious it, but it may be at the cost of safety uh again, given all those limitations and really we need, we need good studies and they're coming. So I don't know whether it's gonna be a year or two. But I think, you know, when I give this talk again, it's going to change each time because all of these studies are ongoing and II think they're gonna help us quite a bit and these are the ones that are looking at catheter directed, but we also have ones looking at dosing of lytics. Could we find a safer dose and continue using peripheral catheters with systemic um lytic therapy? So a lot more to come, what do we do right now? I'm a big fan of pulmonary embolism response teams. I don't know how many of you have a per team, a couple. OK. I'm a big fan. Um and I am not part of the per consortium. I'm not promoting anything. But I think as, as I've tried to show you, there are so many decisions you're trying to make with this in between group and so many data points that you're looking at and we don't really have the optimal way to combine them. We're just sort of checking off and our are some should some be weighted more than others. I think the answer is probably yes. But by bringing together a team, we do a couple of things. One, we get multiple different lenses, which I think is helpful. You get the expertise of multiple specialties, it definitely helps the primary team and the patient cause you're getting one voice, not 10, like we used to when we would consult each of these specialties and it allows you to do a very individualized approach to each patient, including their, not only their risk of deterioration, but their risk of bleeding, their comorbidities and what they want you can bring the patient into your decision making here in a very systematic way. So the way these work is I inactivation patient in the ed D patient somewhere else in the hospital, uh many parts are actually using artificial intelligence in the radiology suite. So as soon as the imaging is interpreted, and they see pulmonary embolism and they see evidence of RV dysfunction, it will automatically alert the team. Um and most of these are done via teleconference. So recognizing that these are very busy people, they're all over the hospital. Some of them may not even be in the same hospital. At the same time, you can bring them all together to discuss the case. And the what the per consortion is arguing is that by doing this, we can be a lot more systematic, we can measure our outcomes. And I think early on this made sense to me, but the question I was always asking myself is, are we making a difference by utilizing all these resources? And there is some good data now that's coming out of both the perk consortium but other uh institutions and registries that patients that are managed by a a response team have lower rates of major or clinically relevant nonmajor bleeding. Importantly, they have a shorter time to reaching therapeutic anticoagulation, which we know is very important. And it didn't necessarily lead to the increased use of more um interventional based therapies or the use of IVC filters, which a lot of people I think who were critics or were hesitant, would say given the people or the specialties that comprise these teams, um the interventional radiologist and the interventional cardiologist, they're just gonna wanna all go to catheter directed. They're all gonna wanna put a filter in and that's not actually the case. And they're showing now that there are decreases in 30 day inpatient mortality. Um And importantly, in the patient population we're talking about today. And so based on that the European Society of uh Cardiology actually has a recommendation to, to set up a part wherever you have that capability. So where are we coming full circle again? II don't think I've been able to answer the questions for you today in this intermediate risk. I think the answer still today is anticoagulation, but there is a group that needs to be monitored. And my guess is that in a few years, we'll be able to better identify them. I also think we're going as we get more data with catheter directed that may end up being the way to go in these higher risk, maybe even high risk patients as opposed to the safety profile that we see with systemic thrombolysis. So I think this is this is evolving. I showed you this slide where Doctor Piazza has identified those problems. I kind of liked his final graphic, which is what are some potential solutions? What can we do? So in terms of diagnosis A I assisted. As I mentioned, some institutions are doing this, the earlier we intervene, the earlier we get them on anticoagulation and the earlier we make a decision about the best treatment, the better the patients are gonna do um using A I and machine learning to take all those factors I just reviewed for you and perhaps give us better decision making, right? Instead of me just checking off how many of these things we're seeing, letting large language models look at all that data, keep learning on all these patients and start alerting us that not only, hey, the, the CT shows significant pulmonary embolism, this degree of clot burden, this amount of RV dysfunction. But then mining the E Mr and looking at their hemodynamics, looking at their lactate level, looking at their creatinine, I'm just throwing all kinds of things out but pulling all that data and saying high risk should probably be in the ICU or high risk should go straight to reperfusion therapy. So I think there's this is an exciting time. I think we're gonna get a lot better at this. There may be other biomarkers coming out that can help us beyond just the BNP and their troponin using the per teams, I think in interpreting a lot of this data, I do like large language models, but they need human oversight, right? And so having that multidisciplinary team to look at some of this information and make those adjustments and then looking at how we can help the ourselves in the research community. So risk stratification, it is essential and patients with intermediate risk pe are the most heterogeneous. And um really do require clinical expertise, a lot of judgment to manage. So OK, keeping us in business here, uh we do need those studies. I can't wait to see them and we need to better identify patients that are gonna benefit. I don't think every patient needs a catheter directed approach, but there are definitely a group that will benefit in the long run. So I will stop there and I'm happy to take any questions or comments. Yes. Thank you so much. Yes, like we consulted a lot for these patients. It's either persistent or severe and, and you know, you mentioned that a little bit that uh you know, I wonder how you, how you approach those patients maybe are hemodynamically stable and has been for 48 hours but are stuck on, you know, high, you know, high flow requirements and, you know, would that push you more towards like a pa or would you really just focus on the rest of the? So, so it's a great question. The question is about those patients that um are staying, you know, relatively hemodynamically stable but have persistent hypoxemia or requiring high flow nasal oxygen. Um And what do you do with them? Um And II would tell you 10 years ago, I would focus more on the BP and say they're fine. I could hurt them more than I could help them leave them alone. And I'm not so sure I it I do take notice. Um part of it comes down to looking again at all of that data. And is this a patient that's in subclinical shock? So, if I had some of those other markers, in addition to that hypoxemia, I think I would be starting to move more aggressively and, and, and think about some type of secondary reperfusion. There's some data coming out early. I uh David Jimenez had a paper just this past year looking just at how patients do when we give them oxygen right upfront, even when they're not significantly hypoxic and showing some better outcomes. So that doing that is important. If you need to go beyond the um high flow nasal oxygen, then the issue becomes how do you do that without impairing the RV? And so decisions about whether you use noninvasive or if you're invasive ventilation, how do you manage the ventilator? Those are, those are tough and that's um something we've been trying to look at in the guidelines. And unfortunately, don't have any data. If you look at it, it's all in patients with ph or right heart failure. And I guess we can extrapolate as best we can. But um I think they're, they're difficult to manage if you go beyond that. So, um you know, I think my approach is that try to maximize that. But if there are other markers that this patient is not doing well, I would be more aggressive images. Hi. Uh Thanks so much for sharing your insights and um depth and width of knowledge about this subject and apparently computer generated images. It's just fun. Um I guess I, I'm struck by the circular um progression of heart dysfunction and the assessment tools where this the RV dysfunction, uh overload dysfunction is really this progressive dynamic uh events. And yet we're using all these static measures. I know the um assessment tools because the big uh thing is to predict the future, right at 25%. Why don't we use a, a delta biomarker or a delta ultrasound binding? Why are you using a singular? Uh So also a very good question. So the question is why are we using static markers particularly when we think about this spiral of RV dysfunction to identify patients that either may deteriorate or may, you know, uh need more aggressive therapy? And that I is also what we're looking at. And, and I think um about 67 years ago, I'm losing track of time. It was probably longer. But um uh uh Doctor Jimenez and I looked at, could we repeat the Pezzi whether it's the, you know, the uh the full Pezzi or the simplified Pezzi and look at it at baseline at 12, at 48 hours. So some of those markers looking at progression over time. And although it, those patients did seem to do a little bit worse, we couldn't find a good cut off on that. But in talking with my colleagues in the area, I think we all agree with you. It needs to be repeated measures. And so the part of that um diagram here when Doctor Piazza is talking about this, you know, newer prediction tool, novel biomarkers, it's really, to me it's how do you, what is that initial period of time? And how often are you measuring things during that time frame? And is this something too where if you've got continuous monitoring if you, you know, so you're watching the BP the whole time, the respiratory rate, the saturation and then how often are you getting these biomarkers that you can add into that? But, but I do think that's the future. Yes, sir. Um uh impairing CT S versus uh um that can be a problem what I, you know, 1 to 1 and we know that part of the puncture, the size and the bio, do you find that the CT maybe over reading this or um or which one? Yeah. So the question is um which one's more accurate in determining RV, dysfunction CT versus echo? And II think it depends on the operator. How, how good is your CT? And are you looking beyond the RV to LD ratio? I think that's the first thing to look at, but there are other markers, particularly as I said, the reflux of the dye, you can look at clot burden. You can look at many other things to say. How concerned am I? Um, with the echo, I think it depends on if you're doing a quick bedside versus if you've got a dedicated expert, cardiologist looking at it, doing all of the measures echo can give you a lot more sensitive information about the RV function. No question. Um So it's really, what do you have available? Who's doing the echo? Uh I don't know that, uh I certainly haven't looked at the literature to say how good is the RV to LD ratio compared to an echo? Say a mcconnell sign or the T II, I'd have to go back and look and see if they've been compared. But my question would be really the quality of each study upfront. But if you're looking at it and you're the clinician and you don't see anything else on that. CT and your bedside echo looks. Ok. And the biomarkers are OK. I probably wouldn't be too worried but because of that conflicting data, I might watch them the first day or two in the hospital on anticoagulation rather than send them out. We're gonna have to transition. But, um, you're gonna be around for lunch. Yeah. Uh probably right up until, but be here a little bit longer. A little bit more. Yes. Yes. Thank you all OK. So I'm gonna ask you all for CM E purposes. We have to transition to the other room. We're gonna do the oral poster presentations. Just be careful when you go in, there's um a uncovered extension cord and a projector in the middle aisle. Please don't trip. Um And then for our international audience and our people that are streaming, we're gonna stop broadcasting but we'll be back at uh what time at 1045 for the next talk? OK.