Marios Georgakis: Why Did Two Major Clinical Trials for Heart Disease Fail?

Okay, I'm Eric Topo, and this is Ground Truth, a drill that you've joining us today. I have the real delight to have this discussion with Marius, your doctor, the University of Munich, who is a physician scientist, actually a neurologist who studies astrosyrosis, genomics, particularly the cerebral ventular, but he certainly is tuned into the cranberry and brosis anywhere in the body, and he wrote a drill in his own subject called the codon about the Zeus trial, and we're going to discuss these two trials, big trials, pivotal trials that failed. So welcome, Marius. Thank you, Eric. Nice to be here. Thank you again for the invite. Yeah, no, going to be great. So this is kind of unprecedented in my decades as a cardiologist to see two pivotal trials, one with over 6,000 patients, one with over 8,000 patients, and they both failed. And maybe there's common threads here, maybe not, but let's first, before we get into that, talk about astrosis and inflammation. What's your sense about, you know, people are talking, oh, well, this proves the inflammation hypothesis is wrong, what's your sense about the process of astrosyrosis? Yes, thank you, Eric, so I think the history of inflammation and astrosyrosis actually goes a little bit, right, so over 100 years in principle. It was the beginning of the 20th century when Nikolaiyanitskiv, a pathologist who fed rabbits with high levels of cholesterol, and he observed under his microscope, astrosyrosyroidic lesions in their arthin, they are ortes at the time. He was able to see this, what we call today, foam cells, right, so macrophages filled with cholesterol and lipids, and he already assumed at the time that these are in principle phagocytes, macrophages, so this was the first connection to any immune response, if you wish, and that's already early 20th century, so over 100 years ago, and of course, since then we've done massive progress, right, so after that, for decades, immunostatic studies confirmed in principle that those are immune cells, macrophages, t-cells, through quantification of cytokines within human plaques, we were able to see that there are high levels of inflammatory mediators, and then even more recently, with high throughput technologies like single-cell RNA, using special transcriptomics, we just came close to recognize the very complex immune microenvironment, I mean, they're talking about multiple different types of macrophages, t-cells, neutrophils, mast cells, dendritic cells, and so it's a really complex and mature immune microenvironment, and so I don't believe that anybody doubt that inflammation and inflammation, once is actually taking place within the vessels, within arthrosyroidic plaques, and I think the big question was always whether targeting this process would actually lead to any kind of therapeutic benefit, and so yeah, and so yeah, that's good, I think no one will question, although some did, and like for example, in the New York Times coverage of this loose trial, he said, oh, well, the inflammation hypothesis is wrong, it's not that there's lacking inflammation in arteries with arthrosyrosis, it's just that we haven't found at least so far a target of inflammation that is going to be safe and highly effective. So also, I want to mention all those that are joined, if you have any questions for marios, don't hesitate to put them in the message, I'll try to get to them as we move along. All right, so we have, we established that inflammation is a big part of arthrosyrosis, and in fact, even a paper this week was about the autoimmune aspects of this whole condition, which is interesting, but let's get into the idea that we have these markers, okay. High-stensity CRP has been out there for a couple of decades, we got Paul Richter that's all he talks about is high sensitivity CRP. We have interleukin 6, it can be measured, but isn't usually measured clinically so much as HSR, CRP, and of course, we have the lipids. We have LDL, HDL, triglyceride, APOB, and LP, these are what we have from blood tests. What did these blood tests tell us? Yes, so it's been, as you said, it's been now, I think, over two decades that the first large scale epidemiological studies actually assessed the relationship of high-sensitivity seriac rotting levels in the blood with a risk of cardiovascular disease, and actually so the pretty robust associations, right, so both case control studies later on also prospective cohort studies, suggesting that people who have high levels of seriac rotting, if you follow the map for several years, people with higher levels have high risk of developing myocardial infarction stroke or cardiovascular death, and in principle later on, more or less, we came to similar conclusions for interleuking six levels, so also interleuking six measuring in blood, both in primary prevention, so people who don't have cardiovascular disease, but also in secondary prevention settings, people who do have cardiovascular disease, they seem to be predictable with your cardiovascular events, and this seems to hold through, even after we statistically agenda other cardiovascular risk factors, including the markers you mentioned LDL cholesterol, HDL cholesterol, triglyceride levels, but also other comorbidities that we know cardiovascular risk like type 2 diabetes, blood pressure, BMI, so these associations have been out there for several decades, and of course it's the question, what do they tell us, I think, in two different dimensions, one, can we interpret them in a way that they represent markers of vascular inflammation, if you wish, and on the one hand and on the other hand, can it mean that these very same markers, these molecules could be potentially cause all mediators of cardiovascular risk, and so potentially promising therapeutic talk. You just said the magic word causal, right, that's the big deal, so LDL cholesterol has pan out to be an extraordinary surgery, perhaps one of the best surgeons that we know in all of medicine, because everything that has lowered LDL, whether it's statins, PCSK9, Xetomy, you know, all these different things, have correlated with improved outcomes, but we saw HDL, which has also been associated with low HDL, with atherosclerosis, and trials were done like niacin and others to raise HDL, they fail, similarly with triglyceride, it's high triglyceride, so what is causal here? Now, you use a technique as other genomesis do, and researchers, Mendelian randomization, with Mendelian randomization, could you explain for the uninitiated what that is, and whether or not, things like LP, L, H, S, C, R, P, and interlooking six have ever been demonstrated to be causal through Mendelianization? Yeah, so, as you mentioned, the question of causality, I think, is an interesting one with regards to therapeutic development, right? So, is that, if a molecule, if a mechanism, it's causal for cardiovascular disease, progression, atherosclerosis development, then it means that if we target this mechanism, we should expect benefits in cardiovascular history, the actual clinical benefits, and the problem we have is that with the epidemiological studies I mentioned before, these are classical observational studies, so the way the problem of this studies is what we call confounding, so if we take, for example, let's say, interlooking six levels in the blood, we measure it in a big population, and then try to associate it with outcomes, we know that the groups of people who have high IL-6 versus low IL-6 differ not only in terms of IL-6 levels, but also in terms of other multiple confounding factors, right? So, and these confounding factors might also be associated with cardiovascular risks, so it makes it very, very hard to disentangle it. The association receives in principle reflecting a really causal association or not No matter how much we adjust with statistical methods because they're always will be remaining a so-called unmeasured confounding like a variation we cannot measure with the methods we have and The idea of Mendelian randomization is that it's core a very very simple one and it It dictates that if a mechanism is causal for the pathogenesis of a disease then Genetic variation that interferes with this mechanism should also Change the trajectory of the disease. We're interested in right and So if we can find a genetic variant or ideally more than one genetic variance that influence the mechanism we are interested in let's say I'll signalling reliably Then we can go take this variant and go and test if it is associate risk of cardiovascular disease and If it does then under specific assumptions we can say that this is probably cause why do we say that because this approach is less Less biased as compared to classical epidemiology Because people carry their genetic the genetic variance from birth it cannot be that Our diet has influenced our genetic variance cannot be that our LDL levels have influenced our genetic variance and So theoretically the groups of faults that differ in terms of this genetic variance should be otherwise Kind of kind of equal. So in other words in the same way that People in the population might differ because of their genetic Architecture in terms of heights and people are shorter some people are taller in the same way Some people have higher or lower activity of our six signaling activity As encoded within their genome and we can then go and Leverage that using modern big data sets and go and explore if this is associated with the risk of cardiovascular disease and The reason why this is an interesting in this particular case particularly in cardiovascular disease this has Served as so far or had served as so far Pretty well in terms of finding as you said causal drivers that can be translated into If you case use therapies So we we had when we do this exercise for LDL cholesterol no matter what Target we select that Actually we use our medications target to lower LDL cholesterol if we find variation there any variant that lowers LDL cholesterol in these genes also lowers risk of cardiovascular disease and The same is true for outlet. Lay as you mentioned and the same act has had also been done for Interlooking six on the other hand If we ran the same exercise for C-reactive protein we find genetic variants within the gene coding CRP that influence CRP levels those genetic variants to not so any Association with the risk of coronary artery disease or other cardiovascular disease and points and the same is true for HDL If we find variants that raise HDL but have no influence on LDL cholesterol They do not influence the risk of cardiovascular disease That's what that just said is really important, okay? Because even though we're going to talk about how HSC or P was used in these trials as an entry criteria particularly in zoos And it isn't been established that HSC or P is a causal factor No less the fact that anything can raise HSC or P you have a cold You had a bump on your knee. I mean, you know anything can raise HSC No less chronic illness of many kinds so this is a real problem of using a non-specific marker And of course, you know, you brought up the HDL story We we saw how very high HL can actually be associated with Adverse outcomes. Yeah, you're you're your explanation that was really lucid Which is great But we're going to get further into that by the way Carsten Schmidt is asked about oxidized social lipid ox PL we'll get into that because that may be especially Important with respect to L.P. L.A. The refined the marker of what's actually the the the inflammatory trigger Okay, so we have this background where you know L.P. L.P. L.A. is causal L.D. L. is causal Maybe and even in looking six has been shown causal the others may be not so much now Enter zoos We get this trial 6300 plus people and I was shocked because I didn't know what the entry criteria you had to have some chronic kidney disease and Some prior heart disease I risk from heart disease and an HSC or P of two or greater two milligrams per Leader, can you comment about the entry criteria there? Yes, and We we kind of course go into the Mendelian randomization results, but actually I believe we're very informative of designing The interlocking six or taking the decision to bring the interlocking six hypothesis forward because I think that's important but in terms of zoos so the inclusion criteria for zoos were The patients needed to have evidence of arthritis cardiac cardiovascular disease Chronic kidney disease as defined by a GFR below 60 or high sensitivity serious progeny of both two and The this was I assume a controversial Controversial decision boy boy novel However, there were good reasons for that so number one was That we know patients with chronic kidney disease to have higher Seriactive protein levels. This is this is a fact and We assume this to mean that they have higher levels of systemic inflammation Number two, we know that patients with chronic kidney disease have a higher cardiovascular risk and This is This is important because this is a risk that is potentially unaddressed by modern Pharmacotherapies, but at the same time means that a trial can be completed probably Faster because those patients accumulate events in a shorter period of time and third I think at the time that zoos was designed There was another anti-inflammatory drug that actually had succeeded in two phase three studies and an old school drug called Coltsison which Is used with caution. It's not really contraintigated, but it's used with caution in patients with chronic kidney disease and so this probably represented Was was attracted because it represented the potential market that With less competition and what would be in patients without chronic kidney disease for an anti-inflammatory drug, so I think this was this was the rationale for the criteria But but but of course there are there are Good reason somebody to think why why this might not have been the optimal population for example While we know that those people have higher inflammation and have higher cardiovascular risk It doesn't mean we have no evidence that their higher cardiovascular risk is actually because of higher inflammation, right? So there is no reason to believe that and On the other hand if We assume that the higher level of CRP above two milligrams per liter means higher inflammation in the general population or higher vascular inflammation in the general population in patients with chronic kidney disease that Who's baseline CRP levels are anyways higher. We don't really know if this higher CRP Valley really represents higher or inflammation Uh, all these are of course interesting interesting considerations and novel novel beyond Uh, it's rules designed to additional Uh, trial programs one was in patients with Heart failure which was in heart failure with preventive sexual fraction Uh, the Hermes and the Hina trials that were also All this continued because of futility just a couple of days ago Uh, and the third one was arteness which is in the population with Uh, good my cardiovascular function. So I assume with subgroup analysis from zoos but also with uh with the full results from from the next trials. We're gonna Know whether The reason why zoos failed was really a population issue or not. Uh, I have to say I'm I'm Um, I find it hard to believe that this is really a population issue. Although I I always found that choice Uh, controversial uh, because the final hazard ratio was Which is the only thing we know from the trial was 0.99 so it's a really Uh, a brutally null result and uh, very difficult to assume that there were subgroups within this population that might have Um, uh, so in any benefit Right now just to be uh clear on the background because you took There had been five trials of colt disease randomized and they showed overall a pretty substance reduction, not just in heart attack, but also stroke. And culturesine is not exactly an ideal inflammatory drug, but it's certainly has some in there. And then there was the trial of the interleukin-1 beta and antibody, which cantoes, which did lower cardiovascular adverse events, but it also increased serious infection and it never moved forward. But it was another example of suppressing an inflammatory process and some better outcomes. But we haven't yet seen, you know, the definitive anti-inflammatory drug, like here, an IL-6 blocker antibody, which did achieve its goal of suppressing IL-6 as it did suppress CRP, but no outcome improved. So that was, of course, not. Well, I should say, by the way, it was an anti-sense oligoneucleotide, not an antibody. And there are other preparations that are targeting ILs that are different operations, not the IL-6, but also IL-6 receptor. Can you comment about these other drugs, and whether or not they're substantively different from the one that is in zoos? Yes. So I think we can take a step back and discuss a little bit the evidence base for targeting IL-6 in the space of cardiovascular disease. We mentioned Mendelian randomization, but beyond Mendelian randomization. So the evidence base before just supporting IL-6 signaling was actually pretty broad, right? So we have animal studies suggesting that if we target interleukin-6 signaling pharmacologically, we see mouse, mouse models factors for serious reductions, we had studies suggesting that IL-6 levels within vulnerable, histologically unstable plaques are higher, human plaques. We had evidence from epidemiological studies as we suggested, interleukin-6 levels, higher levels were associated with higher cardiovascular risk. We had secondary analysis from the contostrial that you already mentioned. So the theory here is that IL-1 beta is upstream of interleukin-6, and then CRP is the is the readout of this inflammatory mechanism. So the theory says that we inhibit interleukin-1 beta with lower L-6, and by doing so we'd probably lower cardiovascular risk. This was the theory before Zeus, and in support of this theory, secondary post-hook analysis from contost showed that the benefit you mentioned for conachinema, which was a 15% reduction in cardiovascular risk, among patients with a history of my cardiovascular function, was absolutely restricted to people for whom the antibody achieved the suppression of IL-6. So there's a additional evidence, but I believe the evidence that actually potentially moved the needle for people to take the decision to invest into this, in this program, was probably the genetic evidence, right? And the genetic evidence originated from the receptor, not the ligand itself. Well, that's key, right? I mean, so protein ligand is indeed causal, but the receptor has better evidence. So this is a very, a very interesting area of genomic research. So the genetic evidence for IL-6 receptor goes back to 2012, where when two independent groups, the results were published by both groups in the Lancet, found two genetic variants within the IL-6 receptor gene that were associated with reductions in CRP, to also be associated with reductions in coronary artery disease risk. And those variants we know nowadays, they were both correlated because of this phenomenon of lung disease equilibrium. We now assume one of the two to be the causal one. And this variant is actually a variant within the coding region. It turns, you know, elitering the DNA, another nine to a cytosine, which then means that a nominal acid is being changed in the protein sequence from aspartate to alanine. And this is in the IL-6 receptor. And it causes the IL-6 receptor to, that is usually bound to the membrane of cells to be cleaved by, by proteins. So to be, to be subject to enhanced cleavage. And by that, we see reductions in downstream signaling as captured by reductions in seriactive protein. And these reductions were translated also to reductions in risk of coronary artery disease. So this was the original evidence. And over the years, my group and others were able to detect more variants in this gene that are actually independent variants of the main one that actually also were associated with risk of cardiovascular disease. And kind of a dose response pattern. And this was the evidence for IL-6 receptor. And until very recently, there was no evidence for, for, for, for interlekin six. Now, our own work later on in that, that's a paper published last year in Nature Cardiovascular Research. If we leverage modern genetic data, since we can actually also find genetic variants in the locus of the interlekin six ligand itself, that are associated with lower expression. Those are variants, not in the code reason, but in the non-code reason of the gene and the influence, its expression within, within immune cells, particularly myeloid cells. And those variants reduce L-6 expression, they reduce CRP as a result, and they are also associated with a lower risk of cardiovascular coronary artery disease. But also spectrum of arthroscopotic disease, and points from arthroscopotic stroke, peripheral artery disease, abdominal aneurysm, carotid arthroscoposis, all of that. So while I do believe that there might be differences in the biology of targeting IL-6 and IL-6 receptor, I think that the main argument people use to support the T-L-6 receptor would be the right target, was the genetic evidence, but I do believe that there is also genetic evidence in the IL-6. The IL-6 ligand, the protein puzzle, okay. But that's good, because you know, there are a bunch of other drugs out there, they're much longer acting like small interfering IRNA and other ways to get at this IL-6 story. That we'll see in the future. But also the patient population made me important. I would submit, I want to know what you think, because you know, I've heard a substack on this too, that if you measure systemic markers, circular lead markers, they don't necessarily correlate at all with what's going on in the coronary arteries. And indeed, at the Munich meetings of the ESC, there was the FAI, the Fat attenuation index measurement, a quantitative measure of current inflammation to the three coronaries. It had terrible correlates with HSC or P. I mean, our value was 0.2, you know, so blood doesn't necessarily mean what's going on in the arteries, that a fair conclusion. I think that it absolutely is. I think CRP has served its role in, you know, raising awareness about the idea that inflammation is something that could be potentially measured, that it is correlated with cardiovascular risk and future cardiovascular events. But I agree with you that it is, it is a very non-specific marker of inflammation and it is probably time to start moving towards more specific biomarkers of hospital inflammation. So similar to the data you mentioned about FAI, I think the largest study today that tries to assess vascular inflammation was the PESA study by Spanish investigators, where they did beta MRI on the arteries of, I think, over 700 people, Carotid's, Howard has femoral, FTG pet and we have some evidence from histopathological studies that FTG uptake actually correlates with local plaque inflammation as quantified by macrophage contents, right? And I think in their paper they showed that if you try to correlate FTG uptake in the arteries with CRP, you also see a correlation very similar to what you mentioned. I think it was 0.24 around these lines. And I believe also from the perspective of experience, I mean whoever has worked at the hospital knows how non-specific elevations in CRP can be. And so I agree with you that we could become wiser if we had Good biomarkers if local plaque inflammation. Yeah. Yeah. Well, there isn't an ongoing trial of an oxidized LDL antibody, it's called or to cumab and it was also presented some of the work at the recent ESC unit, and they use the scan. FAAI, the AI quantification for an intra criterion for their antibody. Okay, so they are really going to the source, the coronary inflammation to test their antibody. I don't know whether it'll work, but I have to admire that they're using a very more specific tool to identify patient who are included. That's the first trial I know of that's zeroing in on actual coronary inflammation rather than relying on a non specific blood marker, whether or not that marker identified as causal. In fact, it was funny. Just a few days before the horizon trial came out, I know a little I had no idea it was coming out, and I still don't know why I didn't come out in the days during the big ESC meeting instead was waited for Friday evening. But just a few days there was a paper showing the causality and a Mendelian renomization of LPL, one of other other studies that have done so. Even in people who had LDL lowering, right, so it isn't just that you can get all the bang just from LDL lowering it was even more with targeting LPL. So this is really fascinating, and now we want to shift into horizon, but I do want to mention Avada Nazari did also type the prior work in Coltazine and Cantos, so to give credit, as you mentioned, all those questions we will try to get to them. There was a question about other markers, low grade inflammation and cardiovascular risk that Christian, Jackie has raised, we'll get to that. But now let's just switch gears, unless you want to go, you're something you're going to say, Marius. No, I also wanted to add to what you mentioned there are about using such biomarkers for selecting the right patients, but I think such biomarkers can also be useful for assessing efficacy in principle, right, so we now know we have we have quite a few trials to know that reductions in CRP, although they might be correlated with reductions in cardiovascular risk. This is definitely doesn't seem to be the rule, so in principle, we see, we have some trials, we have the serial trial testing method, we have the Cantos trial test in Kanakinu. We have some of the Coltazine trials, like convinced in stroke that measured CRP as well, and now we have zoos that measured CRP, and in principle, we know that changes in CRP actually do not perfectly correlated changes in cardiovascular risk. So also because zoos that actually achieved the largest reduction today, didn't translate to cardiovascular risk reduction, and so I believe that such biomarkers could be useful also for assessing efficacy at an earlier time point before taking the decision to invest in a massive phase three clinical trial. Without having previously any evidence of efficacy at the primary pathology level, which is the plaque. You know, you're making a really good point because these trials cost hundreds of millions of dollars, maybe even a billion dollars, like, you know, Horizon 8300+ participants started in 2019, so it went on seven years, I mean, it may have cost, you know, upwards of a billion dollars to do this trial, so talking about big, big shots here, big investment. Alright, so now Horizon is a different test, it's testing the ability to lower LP little A, we know LP little A is causal, we know that it describe Mendelian randomization that we reviewed, but the real test was we didn't have a drug that was specific to LP little A, and this was the first one, it was a Novartis drug, it was an anti-sense oligonucleotide. Um, it's criteria for entry for LP little A was not particularly if I, you know, threshold for some of the other trials, but nonetheless, it did a big test, and it was a bust. We don't know the details, just like Zeus, you know, we don't have all we have a press releases, right? But, um, this was not expected. In fact, you know, I have many patients that in clinic, they have high LP, very high LP little A. And they're very worried that just because we lower the LDL aggressively is out enough, they've been anticipating the ability of this drug in 2027, and the other trials of the multiple programs that are out there, they're not even going to be out till 2028 or 2029, even though they're different types of drugs and different thresholds and different patient populations, whatever. So, this was not anticipated, unless you were to thinking that LP little A wasn't causal, or that there was something wrong here with the patient population, how do you get to what went wrong in this trial? Yes, so I think that's, that's of course a really half a billion dollar question, right? And very, very difficult, very easy, probably to make hypothesis retrospectively about what went wrong, but, um, of course, I do believe that we should, you know, congratulate the trial teams for running this massive Herculean effort. Oh, yeah, absolutely. And so I think here we're talking about a different, a different setting. So again, like we're protein, little A, there was very, very strong evidence from human genetics, supporting that it is associated that it might be potentially causal for arthroscopic development. As opposed to interlocking six signaling, which is only to a very small extent, genetically determined. Uh, LP little A levels are largely genetically determined, so according to some studies up to 90%. And, uh, most of this, um, variation is originating at the gene coding for, uh, the protein itself, upper lipoprotein alpha. And the, uh, there again, uh, we have very strong evidence that each genetic variant that influences LP little A levels, uh, influences risk of coronary artery disease in a, in a, in a proportional way. And so this, this, uh, raised confidence that this is probably, this is probably a causal pathway. I have to say that there are a lot of potential limitations in this, we didn't come into that in this medial randomization studies, right? And potentially a lot of assumptions that need to be taken into account. But, uh, both of those cases we're talking about very, uh, very, very, very strong and well reproducible evidence, right? And, uh, similar to LDL, uh, it was the assumption that, uh, if we would lower LP, we would probably achieve reductions in, uh, in cardiovascular risk. Now, there are, there are a number of, um, there are a number of important considerations here. So genetics, uh, at its core, uh, what it tells us is about the lifetime exposure to differences in LP little A levels, right? And, uh, this is usually, we usually study populations in, uh, middle to late, um, to middle age to, to, to late life. So, it might be over 50 years, uh, median age. And, uh, those folks have been exposed to different LDL levels throughout their lifetimes. And then we go and intervene with, uh, with a medication now in, uh, horizon, for example, at, uh, at midlife for a very short period of time. And, uh, what I want to say is that the genetic evidence might not translate to, yeah, yeah, this contract is already LP A versus a short term. And actually, it's a very, very interesting case here because, um, uh, uh, uh, uh, uh, Steve Burgess and Brian Ference, I think it's a 2018 paper in some cardiology, they tried, they took the LDL evidence. And they said, okay, for LDL, we know how LDL reduction in, with pharmacotherapies actually, uh, performs. So we, we, uh, we have studied it deeply across different baseline LDL levels across different settings. And we know that if we lower LDL by, uh, let's say around 40 milligrams per decilator for five years, we achieve reductions of 20 to 25% in cardiovascular risk. And, uh, we know that if we then take the genetic variance that lower LDL levels, uh, we see a reduction over an average of 50, 55 years, let's see over a lifetime. that corresponds to close 50% risk reduction or the same reduction in LDL, but over the lifetime. And they tried to use this ratio of, let's say, genetic to clinical reduction to translate what we would expect for the LDL trials, right? And of course, this car is a lot of assumptions because we don't really know how reducing LDL is going to pursue, but in their analysis, assuming that the two would perform equal, they found that we would need the 100 milligrams per deciliter reduction in LDL to achieve a risk reduction that would be comparable to what we see with 40 milligrams per deciliter reduction in LDL cholesterol, which would be 20% to 25% risk reduction. And so if some of this is now the inclusion criteria of horizon, we see that they selected the population of with alkylilay levels above 70 milligrams per deciliter. I think the median was around 100 and 108. And given that the drug lowers alkylilay by around 70 to 80% on the basis of the phase two study. So we are below this 100 milligrams per deciliter average reduction. So and the trial was powered to detect the hazard ratio for around 0.8. So what I want to say is that it was a trial at the, what I want to say is that we should wait to see the final hazard ratio. I think it's a different story than in telekin six because I believe the, it will be a completely different interpretation if we see a hazard ratio of 0.8, six that barely cross one. There's a hazard ratio of one that is entirely known, right? So a null effect versus a at least going in the right direction of, yeah, the magnitude has anticipated. But also what you're alluding to here, Marius, is that the drug, even though it's a front-runner drug, it's not as powerful as some of the other drugs. I mean, there's SIRNAs that are getting 99% reduction, not 70 to 80% and if you say the length of follow up, could be longer too. And some of these are just given every three months or every six months and potentially even every year, one shot, one sub-Q injection. So there are lots more legs out there in this story of L.P. Little A. I hope one of the lessons that you're getting at will be that they extend the follow up, which, you know, amend the protocol, extend the follow up, of course, await for the details to come out of this horizon L.P. Little A, first trial. But it may not be as big a miss as you're getting it than people are right now anticipating, right? - I believe we're gonna learn much more from when we have the results of horizon, but also when we have the results of future trials because I think there are three ways of actually getting to a potentially stronger effect if L.P. Little A ends up really being causal, right? So one is to go into populations with higher levels, higher baseline levels of L.P. Little A, and some of the coming trials do that. The other way is to lower L.P. Little A more aggressively, the S.I. RNAs versus the anti-sensual oligonucleidides do that, the, as you said, the lower L.P. Little A by up to 97, 98, 99%. And the third way is to, as you said, extend the follow up and expose those people for longer periods of time to lower I.P. Little A. Yes, so it's all, I think we're going to, we're going to learn a lot from the coming trials, and I think it's also going to be very interesting from a human genetics point of view to get a deeper understanding of how human genetic findings can be translated to. And, you know, like a subgroup analysis may not have been planned since the trial started seven years ago, is to look at those who were on their longest time versus shorter duration just to get a sense about that. No, you're bringing up a great point there. So now let's get to this issue, which is L.P. Little A could be very high, but not having much oxidized fossil lipid containing the cargo, that is the pro-inflammatory cargo. The levels don't necessarily, currently with the, the potency of, of pro-inflammatory action. Anything that, well, we could measure oxidized fossil lipid in people and see whether or not that is part of the story. Any thoughts about that? Well, I believe that all those explanations are interesting and potentially plausible. I'm not aware whether oxidized fossil lipids were quantified in earlier phase two studies of the Leckerson, I have to admit. Or, but they have been in other studies and there wasn't such a, like for example, in studies of aortic stenosis, which L.P.A. does increase the risk and they were not, you know, aligned. And so, you know, you could get more insight about what's really going on. So, as opposed to the Zeus trial, where the issue was coronary versus soup, but here we're talking about within the L.P.A. the molecule, how much of it was carrying the culprit. Some people could have a high level and may not really denote any substantial risk. You know, that's a theory. There's some data to back it up, but it needs more work, anyway. Just, that's something I'm thinking about, but you know, we'll see. But now that's the story here, we want people with coronary inflammation. We want active coronary inflammation. So, if you can get a CT scan for $100, low resolution, and you could quantify the inflammation in each artery. So, forget the calcium score, which can be very misleading. People with zero calcium score can have very high extent of coronary inflammation in all arteries. We've seen that. So, the question is, do we do the future trials when we're trying to change lipids and change inflammation mediators? Should we be sure that we're getting the right patients in for, you know, like that trial I mentioned that's doing that? Will that make the trial smarter in terms of really going after the target of coronary artery inflammation? What do you think? Well, I do believe that, especially when it comes to entirely novel hypothesis, where we don't have previous evidence of efficacy. Because if we talk about LDL cholesterol lowering, I think there is very substantial evidence that LDL cholesterol is a great target of cardiovascular risk. And even appreciated by the FDA that allows approval just on the basis of LDL reduction. But when it comes to entirely novel hypothesis, like novel inflammatory target or novel targets related to other mechanisms of plaque biology, I think that it will be inevitable to not move towards more sophisticated study designs, that include more sophisticated assessment of plaque biology, plaque morphology, and some kind of assessment of the primary pathology. I still believe that we have to be pragmatic. I don't believe that a phase three cardiovascular or outcomes trial can really incorporate such kind of imaging, imaging markers. Well, you know, I'm not talking about an angiography, but the phase two study to assess efficacy or to provide proof of concept so that we then move on in a phase three study with higher confidence. I think that this would be reasonable, yes. And I think that this would be the direction to go. Yeah, I think, well, we disagree a little bit. I think you're right about phase two, it'd be easy, but getting a CT, fast CT is a couple of minutes, it's very low cost, and it could be done anywhere. I mean, it's all over, it's being marketed under catheter, like it's sliced bread or something. All right, now, one of the theories is, look, you just maxed out in cardiovascular disease. In fact, years ago, the companies were all saying we're not even studied anymore, 'cause we got the control group with all their drugs to lower LDL, there's not much else we can do. And then they came back to it with these recent, interventions, IL-6 and LPLA, but they were a bus so far, and some people say, "Well, look, you just max out, you just can't do any better, your thoughts." Yes, actually, in the sub-stock post you mentioned, this was one of my potential explanations for why he just failed as well, right, that inflammation was the wrong target altogether that we should never go after inflammation, and I think the main argument there of the people supporting that is not that inflammation doesn't play a role in cardiovascular disease. I think this is appreciated by most of the people in the community, but that inflammation ensued downstream of systemic cardiovascular factors that we are anyways treating very well and very aggressively that the reason any more benefits to extracts from doing that, and I think it is a reason of argument, but the other cardiovascular disease remains the leading cause of death worldwide, and even if we see these two trials, right, I mean, in both trials, there was residual cardiovascular risk. I mean, the patients were in modern pharmacotherapies, I think in horizon, average LDL was 65, if I'm not mistaken, at baseline, and still over the trial duration, median four or five years, more than 10% developed new cardiovascular events. So there is still some unmet need there, and so I believe there is still some space for improving our efforts. And on the other hand, I find it hard to come in terms with the idea that the only way to treat the number one pathology causing the number one pathology, the number one cause of death is by treating systemic cardiovascular risk factors and not going down to the primary pathology, right? There has been decades of research into plaque biology, local mechanisms, driving, plaque destabilization, and plaque erosion, plaque rapture, and at the moment, all the pharmacotherapies we have act rather systemically rather than at the plaque microenvironment. And so I believe there is a very large module for research there, that would be, that's maybe. Well, I've since you're big into genomics, polygenic risk score for pharmacies is the most validated of all the polygenic risk score. But there's a little problem, okay? The little problem is, okay, let's say you have a very high polygenic risk score in the top 1% for heart disease, just until you win. It just tells you it could be at age 99 or it could be age 50, okay? So what I'm getting at is we want to know the status of that person at a moment in time, whether their arteries are inflamed. And I don't think we're going to be able to really, you know, practice the ceiling here of max therapy. And by the way, 65 LDL, no, we can get it down to 10 or 20 if we really want to. But the point is, we're doing that now on people who don't need to have that done, all right? We've got all these people that, you know, have, you know, look at TikTok or Instagram or, you know, whatever stuff that isn't have good evidence. And they say, I want to get my LDL down to, you know, zero. And of course, they don't really need to do that. And they're taking a lot of drugs to do that. And we don't know whether it's benefiting them, okay? So the point here is, and maybe it is, maybe it is. But I want to know if I'm going to go after somebody who doesn't have clear-cut risk, especially in a, I'm going to invest $500 million or $1 billion, I want to know they actually are at risk. What better way do we have to know they're at risk? And, and you, you show I, I'm sure the, the recent study that was presented at the last week, but also published at the same time in New England Journal of Medicine, the React study from Denmark and Spain, right? Assessment of, uh, uh, assessment of arthroscaroses with symbol technology principle, uh, uh, showing that after the age of 60, almost everybody has some level of arthroscaroses, but until then, uh, there are real differences in, uh, in the, in the trajectories. And what I found very fascinating as somebody being, uh, more into, um, cerebral vascular disease was that, uh, the proportion of people actually, um, having isolated coronary arthroscaroses. So they assessed, this is a study where they assessed in 16,000 people with ultrasound, carotid arteries and femoral arteries and with, uh, coronary CTA, their coronary arteries and in the general population, 16,000 people from 18 all the way up to, I think, 80 years. And, uh, provided for the first time a picture of, uh, arthroscaroses burden across the lifetime. But what I found very fascinating is that the proportion of people who had isolated coronary arthroscaroses was actually very low. It was less than 10% of the population, uh, being, uh, beyond coronary CTA, we, we, we, we have actually tools, ultrasound is extremely safe, radii, radiation free, very easy to do at any outpatient, uh, uh, clinic setting and, um, uh, readily available in principle. And we can use that to actually screen people who develop arthroscaroses, monitor them over time to see how this, how this, uh, where development, uh, yeah, I agree. You can look at plaque. The problem is it's kind of like, you know, cancer. You don't die from your tumor. You die from metastasis in arthroscarotic disease. You don't die from plaque. You die from plaque ruptured and, and, and fissure and erosion to cause heart attacks and sudden deaths. So the point is plaque itself isn't enough. You need to know it's really high risk plaque, high risk person, right? So I don't know that plaque alone by ultrasound is going to, I do agree with you. That's a very important study with publishing the New England Journal that, you know, after, and by the way, other studies have shown very young people, even teens can have significant plaque. It's less, you know, 20s and 30s. But the question is, is that plaque going to cause trouble in any given individual? And I don't know that plaque burden enough. Is that enough? Right? Well, I, I mean, can we do better than these non-specific markers, lipoproteins that are not yet shown that you suppress them, it changes events, or just plaque itself. Yeah. Well, I do believe that at the early stages, it's going to be hard to assess inflammation. And on the other hand, the current status quo is actually assessing just three spotters, right? So we assess age, sex, and some basic risk factors with the different equations that exist at the primary prevention setting, I mean, score two here in Europe, prevent in the U.S. And those definitely are very bad at predicting who is having autosclerosis and who doesn't have autosclerosis. Yeah. Yeah. And so I believe there is a lot of space if we move towards primary prevention to start, to start to lower risk at a population level. I want to prevent the plaque from accumulating in the first place, don't misunderstand my assertion here, just saying that once it's there, you know, it may not be so vulnerable such a liability as if we could just know that that person has the real potential to have a plaque event, right? Okay, this has been great. We've gone just about an hour here. I do want to open it up before we close if there are any further questions. But I've only picked up a few along the way. So if you have any other questions before I say we're done with Mario's who is just a gem and it's fun to compare our views and, you know, we don't agree on everything, but you also, you know, have some really excellent insights here. And I think everyone can appreciate that. Any other questions, did I miss anything that you want to bring up, Mario? No worries. Thanks a lot. It was a real pleasure. I think if anything, what I would like to highlight is because this probably took a lot of attention in the last weeks with Horizon and Zeus. It's the whole idea that we can use genetics to identify drug targets, right? And I think we are, we went through a phase where this was hyped a lot as a concept. We went through a phase where it matured into the concept that, yeah, it might work sometimes, it might not work sometimes. And now we are going through a phase where we have two. trials, largely based on human, like whose ideas originated in human genetics, failing. And I just want to highlight here that I still strongly believe that human genetics can't point to causal biology. I don't believe we have any better tool at the moment to identify mechanisms that are causally involved in the pathogenesis of our human disease. But on the other hand, we have to remember that the fact that we lower with the risk biology doesn't mean that the risk, clinical translation, clinical development, those are different arts and different, they have their own nuances, but we shouldn't forget the value of. I'm with you, I'm with you on that. Now there's a couple of questions before I wrap up, there's a focus on inflammation mean that turisto would be more valuable for screening than clearly, just to be clear, clearly elucid and heart flow are competitive technologies for assessing coroner artery, but they don't look at inflammation. Turisto has FDA for exclusive inflammation and occasion, the others are looking at plaque, burden, vulnerable plaque, abnormal flow of reserve. So that's a big question of course, now that you could do a caristo in the future without even having to do CT angio, I don't know your thoughts, I mean I wrote up. I believe those technologies, because they move as we discuss towards assessing plaque presence and morphology directly and different features of plaque biology, I think there is a lot of promise there, but I also do believe that we still need. I would like to see those technologies implemented in the context of clinical trials, so that to see their value in real decision making situations, and yes, so this would be my answer. I mean, only a couple have been validated with histology in arteries and a caristo was, you know, eminently validated, and I mean I think that's what a real dilemma around, but of course what I wrote about was I thought that that technology has a lot of potential. Now, Ryan O'Young wrote a question, what can we learn about the discrepancies between Mendelian renemization and the trials, or would there be additional things we have to consider from now on? You test on that, but maybe you can amplify it. Yes, I do believe that considering the clinical context is important, because genetics does not really inform us about that, I think there are two, one critical example probably here, I think, is factor 11, the target, so very recently, Ascindexian gained so the efficacy in lowering non-cardium wall extract, no recurring stroke in patients with a history of non-cardium wall extract, and Ascindexian is a small molecule inhibitor of factor 11. And there was strong genetic evidence for factor 11 being involved in factor 11 variation being associated with the risk of eschemic stroke. And it's a very interesting example, because actually variation was most strongly associated with cardium wall extract, although there was a strong signal also for non-cardium wall extract, but when Ascindexian was tested in atrial fibrillation patients against modern anticoagulants, a fixabon, ribaroxabon, and so on, it failed to show efficacy. The study stopped early, was discontinued, because actually the drug performed worse than modern anticoagulants. And so, of course, a genetic study could never give you such a comparative analysis against the current clinical context and the fact that atrial fibrillation patients are actually being already treated with a very good drug. But then moving on, the drug showed efficacy, non-cardium wall extract, where we didn't have, let's say, where the signal was interestingly less strong, but at the same time, probably the clinical context was cleaner. So what I want to say, I think it's always important to take the clinical context and the clinical scenario in which a drug might be used into account when trying to interpret the genetic findings. Right. Well, I mean, along those lines, you know, the question is now that we have 20% of people have significantly elevated LPA, little A levels, okay. And we don't have a specific drug. We may never have a specific drug, I hope we will. But what are we doing in the meantime, right? And, you know, I think, yes, you could just get everybody's people with LPA, little A, get their LL down to 20, right? You could do that. I mean, theoretically, as long as they don't have some FH or other reasons to, you know, that it's too hard to suppress, but should you, you know, should you because, you know, it could be expensive with these PCSK9, now we have an oral blocker of that, which looks pretty good too. But nonetheless, these are not necessarily covered by insurance and they could be a big issue. They have side effects, certainly statins have side effects. So that is a question now. And I don't know how to answer that, except, you know, yeah, get everybody LDL down to 20, or see if they have inflammation, coronary inflammation. That, to me, needs to be assessed, prospectively because maybe we don't have to push this, we're going to be waiting for years now for people who have a very high LPA, little A level and our hand waving, and, you know, a lot of physicians, cardiologists are going to just go to the max reduction, but maybe that's not necessary in all these folks. That's maybe we should measure oxidized phospholibus or other means to determine. And by the way, if you have, if you do get the LDL down to very low levels and it's over six months or a year, what if they're still persistent inflammation? You know, this could be a very high risk patient. So anyway, I just leave that dangling because we don't know, right? I agree with you, Eric. I think the main answer is we don't know, we have to wait for the future trials and hope will be wiser in a few years. Yeah, I hope so too. Well, you've made us wiser today, Mario, so thank you for joining. I also want to thank all of the 400 folks who have joined us today. It's been a fun discussion. I've learned a lot. Hopefully you have as well. We've gotten to a lot of issues about current disease, the number one killer of man. And hopefully we're going to do better in the future and that we're not maxed out. That's let's hope we're not because I think there's lots of room for improved. All right. Well, thank you. Take care. Thank you, Eric. Real pleasure. Same here. Thanks. Bye-bye now. Bye-bye. Bye-bye.