Chapters Transcript Calcific Aortic Stenosis: Key Concepts and the Road Ahead – Cardiology Grand Rounds Philippe Pibarot, DVM, PhD, FACC, FAHA, FESC, FASE, discusses calcific aortic stenosis, looking at key concepts and towards the future. PHILIPPE PIBAROT: So, aortic stenosis. And I have some disclosure that I would like to mention before getting started. Some are relevant to this presentation. And I will discuss the the SAPIEN 3 device, which is not yet approved in the US. So as Brian mentioned, my background is in veterinary medicine. And so let me tell you this little story. He already told you the story of the tiger. Once I was trying to sedate this tiger. This was a-- I mean, I was young and I obviously did a judgment mistake. But this is another case. Actually, my first publication in cardiology was in Equine Vet Journal, which is actually a high-impact journal in veterinary medicine. And it took a long time to put together this case report/review. And this was a donkey, but a very special donkey. It was actually the mascot of a daycare system for young children. The problem is that as you see on the ECG, it developed third-degree AV block, and so was doing syncope three to five times a day. And so this donkey, like 600 pounds, was kind of falling down and becoming a serious risk factor for the kids. So two options-- euthanasia or putting a pacemaker. Of course the kids voted, and it was a pacemaker that was decided. And so we implanted a pacemaker, and the donkey went back to business and had a happy life. So calcific AS, as you know, calcific aortic valve disease is a progressive disease starting with sclerosis, calcification, thickening of the valve, and progressing to obstruction of blood flow. And it is a disease that's associated with a very important burden. It affects about one million people in North America. The prevalence, as you know, increases dramatically with age. In the population over 65 years old, 3% do have AS. And unfortunately, up to now, there is no medical therapy, no drug to reduce the progression of the disease or the outcomes. And so when a patient has a symptomatic [INAUDIBLE], the only option is to do an aortic valve replacement. And we have now two options, surgical or transcatheter, and we will discuss this. And this disease is directly responsible for about 20,000 deaths and 80,000 valve replacements per year in North America. And the average cost for an aortic valve replacement is, in Canada, $75,000. So any hope for pharmacotherapy in AS? Well, there are some new things I would like to share with you. First, AS is a very complex disease. And I won't go into detail here, but you have infiltration of lipids and oxidation and inflammation, activation of renin-angiotensin system, and fibrosis. And also, there are some mechanisms leading to calcification, to mineralization of the valve. And there are actually two main pathways. One, it is very well-known, which is osteoblastic transition. So there is an osteoblastic switch of the phenotype of the myofibroblast that becomes kind of osteoblast-like cells. And the second mechanism that is kind of underestimated that we underline in our group is apoptosis, that we know very well leads also to mineralization. So what about the drugs? Of course, the focus was initially put on statins, because inflammation and lipid-mediated inflammation, this was a good target. And actually, you know that three randomized trials later, statins failed. There is absolutely no difference between statin and placebo in those three trials. So now there are some interesting recent data, and especially this GWAS from George Thanassoulis where they actually found a strong association between the polymorphisms SNP in the locus of the LP agent-- so the gene encoding for lipoprotein(a)-- and the presence of aortic valve calcification, and also of incident clinical aortic stenosis. And so there is now more and more evidence, from this paper and other papers, that Lp(a) may be involved in the progression of AS. Actually, lipoprotein(a) is the main transporter of oxidized phospholipid. And oxidized phospholipid, once it is in the valve, is transformed in oxidized free fatty acid and LPC. LPC is the lysophosphatidylcholine, and it's a bad guy. It's a pro-inflammatory, pro-calcifying agent. And so patients who have a high Lp(a) may have a faster progression of the mineralization, and therefore of the stenosis. And well, the GWAS was essentially an association between the Lp(a) SNP and in the presence of arterial calcification. But when you see the patient, they already have the disease. So the next step was to demonstrate that the Lp(a) is associated with the progression of the disease once the disease has developed. And this is what we found, and this is unpublished data that I share with you today. It's a substudy of the ASTRONOMER trial. So this was one of the trials that failed in demonstrating any effect of statins. And actually, we looked at the association between Lp(a)'s circulating level and the progression of aortic stenosis measured by echo. And you see that the patients who are in the upper tertile, which is more than 58-- so this is close to the 50 that we use in clinic to say that the patient has elevated Lp(a)-- have a faster progression of aortic stenosis compared to the patient in the two other tertiles. And there was also a strong interaction with age. You see that the association between Lp(a) and progression is much more pronounced in the patients who were actually younger and probably had less calcified valves. And this was also associated with a high risk of cardiac events, which was essentially aortic valve replacement. 90% of this composite was aortic valve replacement. So I think there is now more and more evidence that there is an association between Lp(a) and oxidized phospholipid and the progression of AS. And this has important implication, because about 20% of the population do have elevated Lp(a). So this is 65 million people in North America. But as you know on the other end, Lp(a) is essentially determined genetically. So lifestyle modification will have only a minimal effect on Lp(a). And they currently available drugs, such as niacin, only achieve a modest reduction in Lp(a), and they have significant side effects. There is some hope. There is a Phase I and II trial that are currently in press in Lancet where they show that oligonucleotide antisense that is directed to Apo(a) reduces Lp(a) levels by 75%-- so very potent effect-- with minimal side effects. So we are working now on a trial using this compound, and see if it is able to slow the progression of AS. Now, there's another potentially interesting target that is already available, which is targeting renin-angiotensin system. This is a study that we did a few years ago. This is not randomized, of course. But this is a retrospective case-matched study. And actually, we found that patients who were on ARBs had slower progression of the stenosis compared to patients who had hypertension, or even patients who were on ACE inhibitors-- and even slower compared to patients who had no hypertension and had no medication targeting renin-angiotensin system. And when we look at survival, actually, interestingly, patients who were on ARBs had similar survival compared to the control patients, and better compared to patients who had hypertension or ACE inhibitors, although they had, as expected, a worse baseline risk profile. So I think this is a potentially other interesting target. And AS is really-- and this will be my last slide on pharmacotherapy-- is really at the intersection of several diseases-- atherosclerosis with the lipid infiltration. And we know that statins failed but maybe Lp(a) lowering will eventually work, especially in younger patients with mild disease. Hypertension, and activation of renin-angiotensin system may play a role, also, not only in the progression of aortic stenosis but also the progression of associated myocardial fibrosis. And so these medications, particularly ARBs, may be helpful. And finally, AS is also an osteoblastic disease. And so there are now some ongoing trials in UK assessing the effect of bisphosphonates and RANK antibodies. The name of the trial is SALTIRE II. So there are some things ongoing. Of course we'll have to wait several years before knowing if we are able to alter the progression of AS with pharmacotherapy. And for sure, I don't think that it will be "one drug fits all." It's a complex disease, and depending on the baseline characteristic of the patient-- the age, the stage of the disease-- we may need different medications. So pending the development of this pharmacotherapy, the only option that we have to treat the disease is aortic valve replacement, right? And there are two options, surgical or TAVR. And the challenge is to know when to intervene and how. Should we do a SAVR or a TAVR? So this is what we have in the guidelines, published last year. If the patient has symptoms and there is a high gradient, I should think this is a no-brainer. This is what they call a D1 stage. There is this new concept of disease staging in the Valvular Guidelines, as you know. And this is a class I indication for valve placement, OK? Not much debate about this patient. Now, if the patient has symptoms but has a low gradient-- so main gradient is lower than 40, peak jet velocity is lower than 4-- and the patient has a reduced LV ejection fraction, what is recommended to do is a dobutamine stress echo. It's a low-dose dobutamine, to 20 mic. And if the gradients increase up to 40 or more, then this is a class IIa recommendation for valve replacement, and this is classified as D2 stage in the guidelines. On the other hand, if the patient has a low gradient but preserved EF, then you need to look at the flow, if there is any low flow state, which is defined in the guideline as a stroke volume index lower than 25 ml per meter squared. So if there is low flow and the valve is heavily calcified-- and we eventually may use multi-slice CT to corroborate this finding, as we'll discuss later-- this is a D3 stage. These are the patients with a paradoxical low-flow, low-gradient that we heard described in 2007. And this is a new indication in the guidelines, a IIa recommendation. Now, if the patient has no symptoms-- well, if you have a very severe stenosis, which is defined in the American guidelines as a peak jet velocity more than 5, mean gradient at higher than 60, with low surgical risk, this is IIa recommendation for valve placement. So kind of a prophylactic, if you will, AVR. If there is no symptoms, it's just severe AS-- not very severe-- the LV ejection fraction is low, then it's a C2, and there is a class I indication for aortic valve replacement. There are relatively few of these patients. I've not seen many patients who have a high gradient, low EF, and who are really, truly asymptomatic. But you don't want to miss these patients, so if they have low EF, you need to operate on them. Now in the C1 stage, which is the patients with severe high-gradient AS, asymptomatic, preserved EF what the guidelines recommend is to do an exercise testing to make sure that they're really asymptomatic. And if they have poor exercise tolerance, falling blood pressure, then you should consider valve replacement with a IIa. On the other hand, if they are truly asymptomatic, then it is generally recommended to do a watchful waiting strategy, unless you have rapid stenosis progression. And we'll come back on some factor that you can use to eventually consider surgery. So let's-- because this is a very hotly debated topic, this C1 stage, you know, so the patient with asymptomatic severe AS. And you have here a patient-- I mean, I'm sure that you often see this type of patient. You know, 65-year-old woman, asymptomatic-- she claimed to be asymptomatic. LV ejection fraction is 60%. AS severity on echo, you have a peak jet velocity of 5.1. One year ago, it was 4.8. So this is a progression of 0.3 within the past year. The mean gradient is 64. Valve area is 0.65, index 0.35. So clearly, hemodynamically severe stenosis. I think the first thing in this patient, as recommended in the guidelines, is to make sure that she is really asymptomatic. And well, to be sure of that, you can question the family. This is always important. I remember one patient claiming to be asymptomatic until we asked to his wife, and his wife saying, yeah, but Tony, you remember the other day, you collapsed in the swimming pool while swimming. You lost consciousness. This is somewhat symptomatic, so. But if you don't-- you know? So indeed, if you do exercise testing in patients who are a priori asymptomatic, about 1/3, in fact, develop exercise-limiting symptoms, falling blood pressure. And these patients have worse outcomes. What about the patient we discussed today? Well, she did a good test with a 10 METS. No falling blood pressure. So now we can say, well, yeah, OK, she is really a patient with truly asymptomatic severe AS, a true C1 stage. And the dilemma here is should we do a watchful waiting, a "wait for symptoms" strategy, which is generally supported by the guidelines? Or should we consider an early prophylactic surgery? And there is more and more studies suggesting that you should intervene earlier in this patient. Because otherwise, some potentially irreversible myocardial dysfunction and symptoms may develop. So it's a big debate. And I think we should individualize the risk stratification and the decision-making with regard to timing on the AVR. And to do this, you need to look at the valve and valve hemodynamics, but you also need to look downstream of the valve to the aorta and arterial system, and to the left ventricle as well, to the chambers that are upstream, actually. So let's start with the valve first. Well, we very well know from this pioneering paper from Catherine Otto that the baseline severity of the stenosis is a powerful predictor of outcome. So the more severe the stenosis, the closer you are to aortic valve replacement or death, you know? And when the peak jet velocity at baseline is more than 4, there is a very high risk that you're going to get an event in the short term. And also, if the stenosis progressed rapidly over time, which is 0.3 or more meters per second per year, this is a powerful risk factor for adverse events. So again, if we come back to the patient I presented, peak jet velocity is 5.1. Fast progression, 0.3. So these are kind of two risk factors. And if you look in the guidelines, well, very severe AS and rapid stenosis progression are IIa recommendations in the guidelines. The very severe stenosis is not defined exactly the same way in the European guidelines-- peak jet velocity, 5.5. In the American guidelines, it's more 5. And I think 5 is more relevant. And so this patient has these two risk factors. So that's good. But up to now, we only look at the hemodynamics. And as very well emphasized in an editorial by Catherine Otto, we also need to look at the valve itself. And this valve from this patient looks bad. I mean, it is calcified. And as shown in this paper published by Raphael Rosenhek in the New England, when you have moderate/severe calcification, it's a powerful independent predictor, even if you adjust for baseline peak jet velocity of adverse events. And indeed, if you look at the guidelines, European guidelines, the severe valve calcification is here, and it's a IIa. In the American guidelines, no mention about valve calcification. Interesting. Why? I think, having talked to Blase Carabello and Catherine Otto, they decided not to include this because valve calcification, as you know, is very difficult to assess by echo. It is semi-quantitative. It is highly dependent on the settings, on the observer. If I want to make a severe calcification, I just have to increase the gain. But I think nowadays, we have new tools, and I think calcium scoring by multi-slice CT is accurate and reproducible in providing quantitative assessment of calcification. And I think it may eventually get into the guidelines. And you can use it to corroborate the hemodynamic severity, especially in patients with low-flow, low-gradient AS, and also to predict outcomes, as I will show in the next slide. The only thing that we found is that you need to use different cut points of aortic valve calcium score in women versus men. So in women, it's a score higher than 1,200, versus 2,000 in men. So, big difference. And then we said when we found this result, we say, of course. Women have smaller aortic annulus, so you need less calcium to reach hemodynamically severe stenosis. So we decided to index the calcium score for the cross-sectional area of the aortic annulus. And then we were expecting that we would reconcile the cut point. But actually, not at all. You still have a difference. 300 in women, in terms of what we call the density-- so unit per centimeter squared of annulus-- versus 500. So still, a difference. What about the case? Well, she had a 1,640, in terms of aortic valve calcium score, and density close to 500, which is consistent with hemodynamically severe stenosis. And you don't really need that to confirm the stenosis severity, but this was just as an information. The other thing that is interesting is that the calcium score, using the same cut points, also predicted very well mortality-- so [INAUDIBLE] points-- both in the whole cohort and in patients treated medically-- again, same collaboration that we did with Mayo and [INAUDIBLE] in Paris. So a very powerful independent predictor-- as previously shown by Raphael Rosenhek, but here you have a quantitative measurement. And again, you need to use different cut points in women and men. Why? We don't know. Why do we need to use different cut points for the different-- and as you know, it's very difficult to unveil the mystery of a woman's heart, so. But I think one of the hypothesis is that maybe women have relatively more valvular fibrosis besides the valve calcification. And the CT will only capture the calcification, not the fibrosis. So this is an hypothesis, and we have some recent data from explanted valves that would tend to support this hypothesis. Now, let's go a little bit further, in terms of potentially new future perspective and molecular imaging. So what we measured so far is the valve calcification, but there is also the possibility to assess the active mineralization within the valve using sodium fluoride PET-CT. So here you have the uptake of sodium fluoride, and this really gives you an idea of what is the active uptake of minerals now, you know? And as nicely shown by the group of Dweck and Newby, you have an increase in the sodium fluoride uptake with the progressing stage of the disease, which is not necessarily the case with FDG, which is more a marker of valvular inflammation that goes a bit up and down. And in this interesting work that was shared by Mark Dweck, it is interesting to see that the uptake of sodium fluoride actually very well predicts the region of the valve that has progressed, in terms of mineralization, at the two-year follow-up. And they also showed that the sodium fluoride update provides incremental value, predictive value, to predict the mineralization. So I mean, this is still a work in progress. I think it is not ready yet for prime-time in routine practice. Calcium scoring, probably, yes. But sodium fluoride PET-CT, we need more data. So we until now, we look at the valve. OK, fine. But we also need to look what happened downstream to the arterial system. Because about 30% to 80% of the patients with calcific AS do have concomitant systemic arterial hypertension. So hypertension is really overrepresented in AS, partly because as we discussed earlier, there is some common ground in terms of etiology and mechanism. And hypertension actually doubles the risk of mortality in AS. This has been shown in the SEAS trial, for example. So we need to systematically measure blood pressure at the time of echo and calculate the arterial hemodynamics, the systemic arterial compliance, and systemic vascular resistance-- this is easy to do in the echo lab-- in order to assess the presence of abnormal arterial hemodynamics. And given that, you know, these patients with calcific AS often have concomitant hypertension, their ventricle is facing a double hemodynamic load, a valvular load because of the stenosis, but also an arterial load because of the reduced arterial compliance and hypertension. So somewhere in the assessment of the patient, you need to assess this total, this global hemodynamic load that the ventricle is feeling. And this is what we attempted to do by proposing this new index, which is the valvulo-arterial impedance. This is actually the estimated [INAUDIBLE] of systolic pressure that we obtain. This estimation is obtained by simply adding the mean gradient across the aortic valve plus the systolic blood pressure that you measure at the time of echo. And this is divided by the stroke volume index. And it's severely increased when it's higher than 4.5. And the patient today, as a said, the valvulo-arterial impedance of 5.1, so clearly high. And we've shown that these patients, they have reduced survival compared to the patients with lower and arterial impdance-- and also compared to patients in the general population. So I think this is a useful parameter that we should incorporate in the assessment. The last but not the least, you know, is the repercussion of the valvular stenosis and the arterial load and the total hemodynamic load on the cardiac chamber and, of course, the left ventricle. And also the left atrium and right ventricle-- I won't have time to discuss those two, but I will focus on the left ventricle. And when you look at the LV ejection fracture, you may think that you have a wonderful day. But when you look at other parameters, it might be a different story. And you have to remember that in AS, what your eyes see-- so the LV ejection fraction-- is not necessarily what your heart feels. And the LV ejection fraction can grossly underestimate the extent of myocardial dysfunction in patients with concentric remodeling, with concentric hypertrophy-- which is generally the case in patients with AS. So you need to keep in mind that an LV ejection fraction of 55, 50, or even 60 in a patient with pronounced concentric remodeling is not normal function, OK? So to overcome this pitfall, I think we should look to parameters of longitudinal function. Longitudinal function is much more sensitive to the impairment of myocardial systolic function because it reflects the [INAUDIBLE] cardiac function which is primarily affected. And we have now speckle tracking to easily measure the global longitudinal strain. And as I emphasize in this editorial, I think we are ready for prime time. There are some issues, but this maybe should be considered in the next guidelines edition. Because about one-third of patients who are asymptomatic, preserved EF, actually have markedly reduced global longitudinal strain, and these patients have worse outcomes in the short term. So now BNP is another way to assess the repercussions, the consequence, of the hemodynamic load on the left ventricular function. And, well, in this case, we measured the BNP. It was 190. And you say, OK, does it predict outcomes? Well, if I go into the literature, well, the cut point to predict outcomes in AS varies from 70 to 250. So in some studies, it would be considered no risk. In another, it would be considered high-risk. Well, I think what is important to keep in mind is that the BNP, the normal reference value, varies depending on age and gender. So for these patients, well, a value around 90, 95 is normal, OK? So yes, now I can say, well, 190 is about twofold higher than the normal. And this is a very interesting paper from the Mayo, where the first author is actually one of my former PhD students, and she did this work when she was a PDF, a post-doctoral fellow. And actually, the concept that they introduce is rather than using the absolute value of BNP measured in the patient, use the ratio. So the ratio of the BNP measured in the patient to the normal reference value for the age and sex of the patient. And when it's higher than 2, then you start seeing the risk of mortality increasing. And when you are higher than 3, clearly you have a poor outcome. So this patient has a BNP of 2, so it kind of is on the green curve. So intermediary, but has some increased risk. So now I have this patient who was a C1 stage, true asymptomatic severe AS. And actually, now you have a portfolio of risk factors. You know, she has very severe AS, just on the edge. She has severe valve calcification. She had rapid stenosis progression, high Zva, reduced longitudinal strain, elevated BNP. I think now you have enough to individualize the risk of this patient and potentially recommend an earlier surgery, which could be a IIa or IIb, if you look at the guidelines. So I think my message here regarding this part of my presentation is I think we should convert the watchful waiting concept into an active surveillance process, where we actually measure additional risk markers and better evaluate the risk of having worse outcomes in the short term. Just a few words on this difficult issue of low-flow, low-gradient. I know that we already discussed these things with Brian and others. But actually, it's a challenging subset of patients. There are two subtypes, as I mentioned in the introduction-- the classical low-flow with the low EF and the paradoxical low-flow with the preserved EF. Actually, these two patients have markedly different LVEF. But they both have reduced [INAUDIBLE]. They are both in low-flow state for different reasons, for different mechanisms. Here you have the heart failure with the systolic dysfunction. Here it's more the HPEF type of patient with preserved EF but restrictive physiology. And the problem is that on the top of this low-flow state, you have an aortic stenosis. And these two patients have quite a modest gradient, despite in both cases, they had a severe stenosis and a one valve replacement. But the issue when the patient is in low-flow state it is that it is very difficult to differentiate a true severe stenosis that would benefit from valve replacement versus a pseudo-severe stenosis where, in fact, the flow is too low to completely open a valve that is only mildly or moderately stenotic. So you need to do dobutamine stress echo and a manipulation of flow to differentiate these two situations. Because at low flow, you're going to have a small valve area, but it may be pseudo-severe, because not open completely. And you have a low gradient because it is highly flow-dependent. So the gradient may be pseudo-normalized. So what do we have in the guidelines for the classical low flow with the low EF? So this is, again, the D2 stage. IIa recommendation if the patient presents with a high-gradient, small valve area at any stage of dobutamine stress echo. I won't spend too much time on the classical low flow, but more on the preserved EF, the paradoxical low flow, because it's a more debated subset of patient. As you know from in practice, you see this type of low-flow, low-gradient more frequently in elderly women with concomitant hypertension, because they have this pronounced concentric remodeling. And this is an example here. 82-year-old woman. She has an hypertension, relatively well-controlled with ACE. No evidence of CAD. She has heart failure symptoms. Preserved EF, but we measured the global longitudinal strength by speckle tracking. It is reduced. She has severe diastolic dysfunction with increased E over E prime. You see this very pronounced concentric remodeling. And well, AS severity on echo, 0.64. Index, 0.36. So if you stopped here, you would say, well, severe AS symptoms. This is a Class I indication for surgery. But the gradient is only 26. And of course, this raises uncertainty about the severity of the stenosis. And the patient is in low flow, so this might be a low flow, low gradient. Of course, you need to rule out measurement error. You need to look at the valve, which actually looks bad. And if you do a calcium scoring, this is what we tend to prefer to doing in this type of patient. In the classical low flow, low EF, low-dose dobutamine stress echo is great. In this type of patient with restrictive physiology, dobutamine stress echo may not be efficient in raising the flow. So we do calcium scoring, which was obviously very high in this patient, so you know that it is very likely to be hemodynamically severe, and it will progress very rapidly in the next six months. So she finally underwent TAVR and did well. And actually, this is one of the two subsets of patients-- and I will show another one later-- in PARTNER I where actually TAVR, in blue, in the [INAUDIBLE], the iris, was superior to SAVR in yellow, at least during the first year. So there might be an advantage in this patient with small, thick, non-compliant ventricle with small aortic annulus of doing a TAVR rather than a SAVR. Because these patients are at higher surgical risk, and also, they're at high risk of developing a prosthesis-patient mismatch because they have small annulus. It's really a challenge for the surgeon to go out from the OR with good hemodynamics. So what do we have in the guidelines? Well, this is a new entity, and so new recommendations. This is considered D3 stage in the American guidelines. And actually, it's defined as a small valve area, also indexed. Low gradient. Preserved EF, but reduced flow, so reduced [INAUDIBLE] index. Class IIa recommendation in both European and American guidelines for valve replacement. After confirmation of the stenosis severity-- so you need to confirm that the stenosis is severe and is the most likely cause of symptoms. The patients, of course, you only consider symptomatic patients. And there is another thing that is highlighted and has been in a nice editorial by Brian Lindman and Catherine Otto, actually before the guidelines, emphasizing the importance of identifying and treating hypertension in these patients. Because hypertension is frequent. It may contribute to the low flow. It may contribute to the low-flow, low-gradient, and to the symptoms. So there is hypertension, first we need to initiate or optimize antihypertensive therapy, and then reassess the flow, the gradient, and the symptoms. So the last part of my presentation is about-- so we discussed the indication and timing of AVR. Now the choice between TAVR and SAVR, because we have now these two options. If you look at the guidelines, there is a Class I indication for surgical AVR in the patients with low and intermediate surgical risk. So this is the standard of care for these patients. On the other hand, it's a Class I for TAVR in patients with prohibitive surgical risk and a predicted survival of more than one year post-TAVR. And there is also a Class IIa recommendation for TAVR in a patient with high surgical risk. So this is where we are. But let's kind of project into future perspective. Potentially new patient populations for TAVR, I won't have time to discuss all of them, but I think that TAVR may eventually be superior or at least be a very good alternative to SAVR in patients with COPD, chronic lung disease, especially if you transfemoral access with conscious sedation. Patients with diabetes-- I will show a slide that is very elegant work from Brian in this regard. Also from our center, we found that in PARTNER, TAVR is superior to SAVR in patients with a small aortic annulus. And this is actually the opposite in patients with a large annulus, and this is neutral in patients with midsize annulus. Potentially there are some recent data suggesting that maybe in patients with pulmonary hypertension and right ventricle dysfunction, TAVR may better protect the RV function and be associated with better outcomes. Low-flow, low-gradient, as we discussed, especially those with low EF, low-flow, low-gradient and no contractile reserve, and those with paradoxical low-flow, low-gradient. And finally, I want to raise this. It will be my last slides. And this is more provocative, but in patients with moderate AS, low LVEF, and heart failure symptoms. And we'll come back on this interesting subset. And of course-- and this is, I think, one of the most important steps. But TAVR may be non-inferior to SAVR in patients with intermittent surgical risk, and that would be the next very important step. And this is a very large subset. Very great study led by Brian Lindman. This is a substudy from the PARTNER Cohort A-- so the high risk. Actually, I had the pleasure to work on this with Brian on this paper. Actually, we found that in patients with diabetes, TAVR is superior to SAVR. And as I said, there are not many subsets where we see that. You know, there are the paradoxical low-flow in patients with diabetes. And we don't have this signal in patients with no diabetes. So we kind of speculated that maybe SAVR is associated with increased incidence of systemic inflammatory response syndrome and oxidative stress, and that in these patients with diabetes who are particularly vulnerable, this might be the reason for the superiority. So maybe in patients with diabetes-- obesity, as well-- TAVR may be superior. And as I said, the next step is really this, so the intermediate risk predicted between 4% and 8%. And there are some ongoing trials, you know, the Partner II, to which you participate, and the SURTAVI trial. And let me share some things regarding. I think the issue when we move to lower risk, the TAVR needs to be close to perfect, in terms of hemodynamics and clinical outcomes. And as you know, the main Achilles' heel is the paravalvular regurgitation. So to be able to move to the intermediate and, of course, the low-risk population, we need to fix that. There are some encouraging results. This has been presented at ACC. This, actually, the rate of moderate/severe paravalvular regurgitation in the different generations of balloon-expandable valves from Edwards. So you have the SAPIEN, SAPIEN XT, and then finally, SAPIEN 3 that has been presented at ACC. And you see that the rate of moderate/severe is 2.9% in the [INAUDIBLE] cohort and 4.2%, compared to 12% in PARTNER I. So it's not eliminated, but it is quite a significant improvement. The stroke also was an initial concern with TAVR, and of course even more a concern when you're discussing about intermediate or low-risk population. And with the SAPIEN 3, in the high-risk was 1.5% at 30 days and 2.6%-- which, I mean, becomes really interesting results. And even, I think, more impressive was the results of 30-day mortality, where you have now-- let's focus on the [INAUDIBLE] cohort. You have in the high-risk population-- so it's pretty high STS score-- a 2.2% 30-day all-cause mortality. And it's 1.1% in the [INAUDIBLE]. So it compares with the best surgical series. And it's going to be difficult to do much better than that. So I think this is encouraging. Of course, we have to wait for the one-year results. Especially for the intermediate risk, you cannot conclude on a 30-day result. So we have to be cautious. But I mean, this is encouraging. But the people [INAUDIBLE]-- and we had this discussion at dinner last night-- say, well, OK, I see you have a 30-day is good. And you have relatively low rate of moderate/severe-- you know, 3.8% when you put together high risk and intermediate risk. But you still have about 40% of mild PVR. And if we do a flashback in PARTNER I, you see that, well, moderate/severe, yes, it's associated with a twofold increase in two-year mortality. But mild is not benign at all. It was associated with a marked increase in mortality. And we were all surprised about this finding, because we had this, in memory, the data of native AR, and native AR, you know, mild is generally benign. But here it's a different context, I think. First, paravalvular regurgitation is maybe one of the most difficult things to assess by echo, or any imaging modality. So maybe here you have what was called mild, but maybe there is a substantial proportion of patients with moderate. And the other thing that is important is the acuteness of AR. You know, in chronic native AR, the AR developed progressively, and so mild AR is well tolerated. But here we have an acute AR. Because the patient, you know, has no AR before surgery, as a QAS. And then after TAVR, he has moderate/severe. So it's acute. And we did this paper with my colleague and friend Joseph [INAUDIBLE]. And actually, what you see here, the black curves are the patients who had no or trace AR before and after TAVR. And they do well. The red curve are the patients who had no or trace AR before but with moderate or severe after. And they do very poorly. OK? And interestingly, the blue curve are patients who had moderate, mild to moderate AR before, and they have moderate or severe after. And they do as well as those with none/trace AR. This is just to emphasize the concept that here, if the ventricle is preconditioned to AR and to volume overload, this is business as usual, and it will probably better tolerate AR. On the other hand, even in mild, acute PVR could be harmful in patients with pure AS. And this type of ventricle, which is small, thick, non-compliant ventricle, even at 10 mL, the AR could be a catastrophic, especially in patients with paradoxical low-flow, low-gradient. They may not tolerate a mild PVR. So mild PVR is not equivalent to our patient, OK? And having said that, if mild PVR was so bad, we won't have the survival results that we obtain in SAPIEN treatment. Again, it's only 30-day. But there are some studies where you have longer-term data. And this is the CoreValve, also presented as a late-breaking trial at ACC. What you see is you have around 7% to 10% moderate/severe, depending on the follow-up. And you have about 30% mild-- so much more than in SAVR, of course, as expected. And still, interestingly, the mortality was lower, not higher, with TAVR, and this was significant at [INAUDIBLE]. And the output of PVR, if you had A3, my PVR was killing a patient heavily. I mean, we would see the opposite signal. And it is interesting to see that this survival benefit is maintained to two years. And this was intriguing to me. How to explain that despite the rate of PVR I showed you, TAVR is significantly better than SAVR in this high-risk CoreValve? You could say, well, TAVR is less invasive than SAVR. OK. But then we would see a difference at 30-day mortality, right? And actually, there is no difference. Even two months is very similar. This is after, that the curves tend to diverge. So this is not the operative risk and the reduced operative risk. This is not the stroke. Stroke, you know, is somewhat similar and cannot explain the difference in mortality. So I think the potential explanation is prosthesis-patient mismatch. Because we worked, again, together with Brian on this paper about prosthesis-patient mismatch in PARTNER and found that TAVR was associated with more PVR compared to SAVR, but on the other hand, less prosthesis-patient mismatch. So these patients have better effective orifice areas and lower gradients. And this was associated with, also, reduced mortality. Maybe in PARTNER, the difference in paravalvular regurgitation kind of overcomes the benefit of having less mismatch. But if you look at the CoreValve, there is a very important difference in hemodynamics between the two arms. You know, there's about 0.3 to 0.4 centimeters squared difference between the airways. The airways of the TAVR are clearly better and the gradients are better. And there is, from what I've been told by the investigators, much less prosthesis-patient mismatch in the TAVR versus SAVR. So this might be an explanation for this better survival in the TAVR. And I know that the CoreValve investigator will present at AATS that is coming next week, a late-breaking trial presentation about the effect of prosthesis-patient report/review. mismatch on survival on these patients. Now the last concept I would like to share with you is the impact of moderate AS in patients with a low-EF, low-flow, low-gradient. This is the results from the TOPAS study, which is a multi-center study on low-flow, low-gradient. And actually, in the patients treated medically, we found that the aortic valve area measured during dobutamine-- so after normalization of flow-- actually, the impact on survival is not at 1.0, which is the cut point for severe AS, but it is higher. So it appears that in these patients with low-EF, depressed-ventricle heart failure, a moderate AS could have a significant impact on mortality. So again, I think it's raised a concept, same as we discussed with mild PVR. You know, mild PVR may not be equivalent in all patients. Same with moderate AS. I think moderate AS is going to be well tolerated by a ventricle with good systolic function, but the same moderate AS may be poorly tolerated by a failed ventricle. And I have another finding that may be in support of this concept. Here, this is prosthesis-patient mismatch, the impact on 30-day mortality. So this is post-AVR. But interestingly, we found if you have severe mismatch, this is bad, whatever is your pre-existing LV ejection fraction. But in terms of moderate mismatch, which is equivalent to moderate stenosis, this is moderate residual stenosis, if you will-- well, it is well tolerated in patients with a pre-operative LVEF of more than 40% but poorly tolerated in those with depressed LVEF. So again, moderate AS for a good ventricle may actually be severe for a failed ventricle. So I know it's provocative, but we are at this stage, and we're working on that. Should we consider TAVR-- not SAVR, because I mean, TAVR is less invasive-- in patients with moderate AS, low LVEF, and heart failure therapy, and of course, optimized heart failure therapy? So that's-- and I think we need a trial on this. So this is my last slide. I think the next milestone on the road of AS therapy, we probably need an RCT on Lp(a) lowering in younger patients with mild AS. And there will be, I think, an RCT starting in 2016, which the name will be probably RELEASE. Also, we may need an RCT of ARB in patients with mild to moderate AS. Because there might be actually a triple benefit at the level of the valve, the level of the ventricle, and potentially the level of the aorta, by preventing the [INAUDIBLE] and loss of energy associated with that. We need to better identify and treat hypertension in AS. I think we underestimate and undertreat hypertension in AS. And Brian has very well emphasized this pitfall in his editorial. And we need to better recognize this low-flow, low-gradient entity, especially a paradoxical low-flow that is insidious. Because we often have the perception that if we have preserved EF, the flow is normal, but actually, a large proportion of patients may have a low flow despite a preserved EF. And we know that from the [INAUDIBLE], but we have to transpose it to AS. We of course need to wait for the one-year data to confirm non-inferiority of TAVR in these intermediate risk patients. I mean, the results are encouraging so far. But you know, 30-day is not enough for intermediate risk. For low risk patients, it's probably even worse. We probably need to eliminate, to be perfect, to eliminate completely PVR. And we need-- one thing that we don't have so far is the long-term data about of durability of the data. Because we assume that it will be as durable as surgical valve, but actually, we don't know. And in the results that we have now, up to five, seven years. But it's really in the infancy when we're talking about prosthetic valves. And finally, and this is the last point I made, is we need an RCT of TAVR versus medical therapy in patients with moderate AS and heart failure and low EF. And there will be, actually, a trial that should start in September, if everything works well, if FDA approve it. And we clearly want to have you on-board on this trial, because it's really at the intersection of valvular disease and heart failure, which is obviously two topics that you are very interested in. Created by Presenters Philippe Pibarot, DVM, FACC, FACE, FAHA, FESC, PhD Professor, Faculty of medicine, Laval University Director, Canadian Research Chair in Valvular Heart Disease View full profile
