UPMC Physician Resources
TAVR Update- 2016 Current Indications and Future Directions
Dr. John Schindler discusses transcatheter aortic valve replacement and the most common heart valve disease encountered in western countries.
Upon completion of this activity, participants should be able to:
- Inform health care providers of the changing indications regarding patients who may benefit from trans-catheter heart valve replacements.
- Improve understanding and interpretation of echocardiography results.
- Improve the care of elderly patients with severe aortic stenosis by educating HCP’s about the mortality benefits of transcatheter heart valve implantation when compared to high risk surgery.
- 5-year outcomes of transcatheter aortic valve replacement or surgical aortic valve replacement for high surgical risk patients with aortic stenosis (PARTNER 1): a randomised controlled trial The Lancet , Volume 385 , Issue 9986 , 2477 – 2484
- Outcomes in Patients Undergoing Surgical or Self-Expanding Transcatheter Aortic Valve Replacement. J Am Coll Cardiol. 2015 Jul 14;66(2):113-21
- Transcatheter or Surgical Aortic-Valve Replacement in Intermediate-Risk Patients. N Engl J Med 2016 Apr 2
- Evolution of Transcatheter Aortic Valve Replacement. Circulation Research. 2014; 114: 1037-1051
- Clinical Outcomes at 1 Year Following Transcatheter Aortic Valve Replacement. JAMA. 2015;313(10):1019-1028. doi:10.1001/jama.2015.1474
Dr. Schindler has financial interests with the following proprietary entity or entities producing health care goods or services as indicated below:
- Investigation in clinical trials, involving Medtronic, Edwards, and Boston Scientific
The University of Pittsburgh School of Medicine is accredited by the
Accreditation Council for Continuing Medical Education (ACCME) to
provide continuing medical education for physicians.
The University of Pittsburgh School of Medicine designates this enduring material for a maximum of .75 AMA PRA Category 1 Credits™. Each physician should only claim credit commensurate with the extent of their participation in the activity. Other health care professionals are awarded (.075) continuing education units (CEU) which are equivalent to .75 contact hour.
For your credit transcript, please access our website 4 weeks post-completion at http://ccehs.upmc.edu and follow the link to the Credit Transcript page. If you do not provide the last 5 digits of your SSN on the next page you will not be able to access a CME credit transcript. Providing your SSN is voluntary.
Release Date: 11/8/2016 | Last Modified On: 11/8/2016 | Expires: 11/8/2017
I do have several disclosures to present. I am an investigator on multiple clinical trials, I do not take any financial remuneration other than just travel and meals for the required events that I have to attend.
So just you know the TAVR is dominating the medical literature, and in addition to our journals it also is dominating the lay press. And you can see this was in the Wall Street Journal just 2 weeks ago talking about how Medicare may be bad for your heart because of the recent data suggests that the transcatheter valve therapies have been so successful and at times Medicare kind of lags with respect to the clinical data. So you know I personally think that they've done a good job of keeping up with this, it's such a rapidly evolving field and hopefully you'll understand where we are today and not be in that lag phase. And this was also in the New York Times talking about building these better valves. Again with technology the companies have done such a great job at finding the problems and then finding solutions so that we can get better outcomes.
So just the obligatory slide with respect to you know a little bit of prevalence in pathophysiology. We all know that aortic stenosis is the most common heart valve disease encountered in Western countries with a male predominance. The prevalence is actually up to 10% in octogenarians and nonagenarians. There is no medical therapy to prevent the progression of aortic stenosis, we know that there have been trials using HMG-CoA reductase inhibitors as well as bisphosphonates but they have all been negative trials. So unfortunately in 2016 we really have no medical therapy for these patients. And when symptoms are present it has an extremely high mortality rate with up to 50% mortality within 1 to 2 years.
And just to review with respect to the classification, this is from the most recently published guidelines in 2014. We've changed to actually referring to this as stages of disease, A, B, C and D. So A is just those patients who are at risk for AS; B are those with mild to moderate disease. We really want to focus today on those who have severe disease that may be asymptomatic, stage C, with normal LV systolic function and C2 with reduced LV systolic function because there is a fine distinction there. And then those patients with severe symptomatic disease, certainly the stage D1 patients are the most straightforward patients and we all know that if someone has a high gradient, a low aortic valve area and symptoms that are suggestive of valvular heart disease, specifically aortic stenosis, the treatment pathway is very straightforward. It gets a little more complex in some of the patients in the stage D2 and certainly in stage D3 and we'll touch on that a little bit as we go through this.
And then also just so you are aware of the echo parameters and how we sort of classify things with respect to mild, moderate or severe aortic stenosis really as I said when the mean gradient is greater than 40 and the peak velocity across the aortic valve is greater than 4 m/s with a valve area of less than 1 cm sq, again that is the severe category and those patients are straightforward. We get a lot of patients where you'll see the echo reports have discordant data where the mean gradient may be 25, the aortic valve area may be .8, what do you do in those patients? So again it takes a real concerted effort to look at the echocardiogram and make sure that we are really getting good data from these studies.
And one other thing I want to point out is that in larger patients with very high BMIs, high body surface areas, they you know a valve area of 1 cm sq or 1.1 cm sq for that fact in a patient who might weigh 300 pounds, a larger male, there is a problem because if you index the aortic valve, which is listed there on the bottom on the far right if you index that area is less than .6 cm sq per meter sq, and that also is a parameter that we look at. So that also is a parameter that we look at, so it's not just the straight up aortic valve area because we see a lot of these patients who are 85, 90 years old, very frail elderly women who you know their weight is only 110 pounds. So they might be okay with an aortic valve area of .9 cm sq when you index their valve.
Okay, so let's jump into the cases because I know that's what I was asked to do, to present some real cases here so we can open up the discussion. So the first patient that I have is an 82 year old male with diabetes, hypertension, atrial fibrillation and severe asymptomatic aortic stenosis. He is satisfied with his health related quality of life and remains active, he gardens and he takes care of his one acre property in North Hills. He wants to discuss TAVR because his older brother came to us about 9 months ago, his brother presented with decompensated heart failure and NSTEMI and he was found to have a severe lesion in his proximal LAD so we treated both processes at the same time with a PCI of the LAD and a TAVR with a 31 mm core valve, and he's done exceptionally well. So now his brother is saying you know I don't want to follow the course of my brother, I don't want to be admitted with heart failure and a heart attack, he almost died. Can I just get the TAVR now?
His medications are appropriate, he's seeing the appropriate physicians who are taking care of his risk factors. So with respect to his blood pressure, he's 160/82, his pulse is in the 70s, he's controlled with respect to his AF and actually he's very well compensated on examination, so he's not in heart failure. And when we look at the spectral Doppler tracings of his echo there is some variability here because there is variable stroke volumes, he has atrial fibrillation so there is variations in his R-R intervals, so we can see the peak velocities here are somewhat variable, but if you look at this one where there is a longer R-R there is a mean gradient of 50 mm of mercury and a peak velocity of 4.2 m/s. And the valve area is calculating out anywhere between .8 and 1.1.
So he goes on to have a heart catheterization prior to us seeing him. He had single vessel CAD with just an isolated distal LAD stenosis and given his lack of angina pectoris it was just decided to manage that medically. He has normal filling pressures and mild pulmonary hypertension. His PA saturation was 65%. So when you calculate out his risk for an aortic valve replacement using the Society of Thoracic Surgeons Predicted Risk of Mortality, it's just an online program that you can put in all the clinical predictors, it comes out for this patient to have surgical risk of AVR of 3.4%. so anyone under 3% is considered low risk for an open operation, 3 to 8% is considered intermediate risk and above 8% is considered high risk.
So the question is how should this patient be managed? So I'm going to open it up to my esteemed panel here. What do you guys want to do, any thoughts about this? Should we just continue to observe him? Should we do exercise stress test with imaging? Should we just take him to TAVR? Or should we get him to the operating room? I think that's perfectly reasonable but I think what that gets to the point now is is the patient asymptomatic, okay?
Now in this patient I didn't quite, you know he's active, he's doing the things that he wants to do so we had a pretty good sense that he's satisfied. And so that's why it's important I think it's very nice that when I evaluate these patients the family members are typically there because sometimes this generation of A, downplay their symptoms, B, they just want to you know give the doctor the information they want to hear. So sometimes there can be discordant data from what the patient is telling you and from what's really happening at home; but his wife was with him. She said I watch him, he's out there, he does you know mow the grass still, you know doesn’t seem to be limited from that standpoint. So I think that's a very valid point, but it gets to the question of you know is the patient really asymptomatic? And what would we do if we weren't sure about that? So if I ever hear that someone maybe you know some family member is saying he is a little bit limited, the patient says they are not. And since this is such a slow insidious process sometimes patients just get used to living with the limitations that have been forced on them from their disease process.
So I just want to point out that exercise stress testing is indicated in these patients. I know you know a lot of times if someone comes through the stress lab and says they have aortic stenosis everybody says whoa, whoa, we don't stress people with aortic stenosis. But it actually is a class IIa indication to stress someone who is asymptomatic, again to see what happens to their blood pressure, see what happens to the gradient, see what happens to them as they are putting a little bit of excess stress on the body. So that's one thing.
And then the other thing in this gentleman, you know he had elevated blood pressure, 160/80 and that's another situation I think sometimes in primary care we are a little reluctant to give these patients additional medications because we are worried about the nature of the aortic stenosis and the fixed afterload problem that they have. But actually treating concomitant hypertension is a class I indication according to the guideline in asymptomatic patients, and you start with goal directed medical therapy. So that really comes from data from the SEAS trial, which was a trial that was looking at Simvastatin to halt the progression of aortic stenosis in asymptomatic patients. And as I said that was a negative trial. But when they went back and looked at the patient, these were all patients with hypertension, okay retrospective analysis. And the patients who were treated to normotensive levels did significantly better with respect to the time of cardiovascular death than those who were treated or left alone and sort of in that 160 to 180 range. And the treatment cutoff that they set here was 140 mm of mercury.
So I think the current guidelines do give us information, and say watchful waiting is appropriate as long as one of the following conditions is not present. And as I said, the asymptomatic patient stage CII who has LV dysfunction that's very different. So our patient has preserved LV systolic function so performing a procedure is not necessarily indicated unless the patient has very severe aortic stenosis. We've gotten away form using the term critical aortic stenosis, now we use severe and very severe. I don’t know that I agree with that but that's what the terminology is now, and so that's defined as a peak velocity greater than 5 m/s because we know that peak velocity alone is an independent predictor of mortality. And when we start to get to that level that's when sudden cardiac death really starts to increase.
If the patient had an abnormal stress test, so again if there is any concern I would put this patient on a treadmill and have imaging associated with it with an echo. And those patients are obviously candidates for cardiac surgery for another reason, if the patient had had the cath and they had sever triple vessel CAD with a high syntax score then obviously you are going to replace the valve at the time that they are getting their CABG.
And this is just a flow diagram of treating asymptomatic patients again defined as peak velocity of greater of 4 m/s. You can see the guidelines differ with respect to our American guidelines and the European guidelines. They throw in a few other indicators such as if there is the presence of severe left ventricular hypertrophy or if the patient has elevated natriuretic peptide levels and again peak velocities are for both a class I indication if it's in that very severe range.
Now you mentioned about the lack of data with respect to asymptomatic patients. So this is one data set that was recently published last fall and it's a Japanese registry. It's a retrospective analysis so it's not a randomized trial of treating patients with TAVR versus continued observation. But these patients did get some form of AVR, some of them were surgical and some were TAVR. And you can see that in the - it shows there were about 1800 patients, 300 patients received treatment even though they were classified as asymptomatic by - for whatever reason by the physicians who were caring for them. And when they went back and followed these patients you can see that all cause mortality significantly favors an early intervention as opposed to continued observation.
And what about hospitalization, because that's what this patient said to us, I don't want to be admitted to the hospital with heart failure. And you can see not surprisingly this one you know makes complete sense because a lot of our patients go from asymptomatic to winding up in some intensive care unit in decompensated heart failure in this - you know when they are transitioning phases. So hospitalizations for CHF were significantly improved in the early AVR group.
So some clinical pearls to remember when you are seeing these asymptomatic patients. A lot of the older literature says that you know if someone treated these asymptomatic, they've gone through an exercise stress test, you've really vetted those symptoms the annual risk of sudden cardiac death is about 1%. So it's actually quite low, and it was 1.5% in that Japanese registry. However 40% of those patients that were followed conservatively required AVR within 2 years of follow-up and a lot of those patients were admitted to the hospital as I said with decompensated heart failure. And the independent predictors that appeared to just come over and over again in all of these trials are increasing age, chronic kidney disease, inactivity and peak aortic valve velocity. And those are the patients that I see every Tuesday.
So what about - let's get back to our patient. So what about when he does develop symptoms? Or let's just say that he did say when he was gardening he was having difficulty with completing that task. So the recently published or presented data, presented and published I should say, at ACC in Chicago dealing with intermediate risk patients, because we said gentleman's STS score was just over 3%, was the big you know headliner at ACC. And it showed that when you are comparing TAVR to surgical AVR there appeared to be a norm inferiority result, meaning that this you know the P values weren't statistically powered to say that there was a superiority to TAVR. But clearly when we are talking about death from any cause or a disabling stroke it appears to be a wash, such that TAVR and SAVR are the same.
And then that's in the intention to treat analysis. There is also an as treated analysis because a lot of the patients, not a lot, but some of the patients in the surgical arm dropped out and so there is another analysis looking at the as treated, that means which patients truly got surgery and which patients truly got TAVR. And again a non-inferior finding suggesting that TAVR and SAVR appear to be the same.
Now when you look at only the patients that were treated from a transfemoral aspect because in this trial with the TAVR approach it could have been transfemoral, it could have been transapical and it could have been transaortic. So transapical and transaortic, those are thoracotomies, that's invading the chest wall. That's a big operation for someone. And so transfemoral is very well tolerated, so if you only look at the patients who had transfemoral compared to open surgical AVR there did appear to be a statistical significance, suggesting that the transfemoral TAVR appeared superior to SAVR in this intermediate risk cohort.
So back to our patient. Since he truly was asymptomatic we added a diuretic to improve his blood pressure control, Chlorthalidone. We encouraged him to remain active and educated him about the symptoms. We advised ongoing clinical follow-up in 6 months and we told him that really the risk of sudden cardiac death without AVR at this point was on the order of about 1 to 1.5%, because that's obviously a question that we get quite frequently. And we quoted the risk of 1.1% of all cause mortality with TAVR in his age group, and 3% risk of mortality as we said, that's what the STS predicted risk of mortality tells us. So at the end of the day he is coming back to see us here in another month, so next year I'll give you his follow-up information. Okay, any comments about the case?
Yeah great question. So basically a drop in the blood pressure with exercise, okay. And so we do put these - you know there are modified (inaudible) protocols, so even some of them with bad joints can typically get on and at least show us some level of exercise. But we are looking at a drop in the blood pressure with exercise and there is one data set that says an increase in the mean gradient greater than 20 mm of mercury is also a poor prognostic finding. And then a patient who may drop their ST segments on the EKG that persist into recovery that's another patient we get very concerned about. And then obviously just looking at the patient as they are exercising. If they are obviously much more short of breath then you would anticipate for that level of exercise, that's a softer call.
Great, let's get onto the second case. You put out a lot there. I'm not sure I can touch on everything. So the second case is a little more straightforward in the sense that he's a 92 year old gentleman, he lives independently and believe it or not we see this quite frequently as well. He's accompanied by his Power of Attorney, he has noticed more difficulty with carrying his garbage can to the curb for the weekly pickup for the last several months and this has progressed now to the point where he's short of breath with minimal activity and he also is endorsing PND. He does have a history of aortic valve disease and he underwent an AVR in 1999 with a 27 tissue valve, a Perimount valve. He also has atrial fibrillation, hypertension and nonobstructive cerebral vascular disease. He is on a pretty tight regimen but everything seemed pretty appropriate.
On exam he is relatively normotensive, his pulse is in the 90s. He's a thin guy and he's clearly in heart failure with elevated JVP, crackles, reduced breath sounds at the bases and he has a systolic and a diastolic murmur when auscultating. And here is his short axis of his - so the short axis of the aortic valve, for those of you not familiar with looking at transthoracic echoes, here is the aortic valve right here and you can see that there is an abnormal flow pattern here, this blue colored jet in diastole. So that's indicative of some degree of aortic insufficiency. And it's very eccentric. Typically if we see aortic insufficiency it tends to be in the central part of the valve. So the question is does this patient have a paravalvular regurgitant leak and failure of the prosthetic valve from that mechanism? Or is this truly just central aortic regurgitation and failure of the valve in that manner?
So pertinent labs included creatinine of 1.9, he's a little anemic. And the final TTT interpretation is that it's a mildly dilated LV with mild LV dysfunction, the EV is 40 to 45%. His RV function is normal, he has moderate tricuspid regurgitation. And the official read on this was that he has a moderate paravalvular regurgitant leak of the aortic prosthesis and there is also severe bioprosthetic valve stenosis. Now the mean gradient was read at 25, but the valve area is low at .8, and the dimensionless index, which we didn't talk about, just trust me that that's low as well And the peak velocity is 3.1. So this is one of those echoes where we are getting discordant data. The valve area is low but the mean gradient and the peak velocity are low, but we also know that he has reduced LV systolic function, so there may just be reduced stroke volume across that valve.
So the question now is what is the most appropriate next step? Should we just increase his diuretics and see him back in 3 months? Should we get a TEE to further evaluate that prosthesis? Should we refer him to cardiothoracic surgery or should we call palliative care? Any thoughts? I didn't make easy questions.
Why was he with his Power of Attorney?
Why was he with his Power of Attorney?
He never married and you know he's 92 and the guy lives - no, this guy is fully functional, you know he was talking to me about the Pirates and he's very interested in sports and you know it's just that these guys that live in his neighborhood have sort of adopted him. And so the Power of Attorney is a lawyer and he was willing to come and make this trip with him. Okay?
So I think we did all of the above actually. So the nice thing about the multidisciplinary clinic is that we have cardiac surgery there, we have the capabilities to do TEEs, we have palliative care services as well, we can call them. So we actually did have CT surgery engage with him, they evaluated him and felt that as a reduce sternotomy at the age of 92 they felt very comfortable calling him high risk for surgical AVR. Palliative care went in, he clearly understands his disease process and made it clear that if there was a therapy that could improve his quality of life and keep him out of the hospital he was all for it. And we did tell him to increase the diuretics and we didn't say follow-up in 3 months though.
So here is his TEE and this is a deep gastric view. So basically again for those not familiar with looking at these, this is the left ventricle, this is the aortic valve here. And the same thing here, it's just that this is 3D color imaging. So what you see is that the blue flow is the flow coming out of the left ventricle across the bioprosthesis. There is a lot of turbulence coming across this prosthesis with this (inaudible) color. So there is a lot of turbulence across the prosthesis indicative of some degree of bioprosthetic stenosis. And then this blue jet or red jet coming back into the left ventricle is the aortic insufficiency. So it's really a mixed aortic valve failure.
And then with respect to whether this is all paravalvular or intervalvular the 3D color we felt that there was really 2 separate jets. So there is one central here and then there is a smaller, it looks like a paravalvular jet out here. And again we had multiple images and there were multiple views to assist in this diagnosis.
So he did end up going for a heart catheterization, he has nonobstructive coronary disease, no surprise, he has severely elevated filling pressures, the right atrial pressure was 10, normal being 5. The wedge pressure was elevated at 25. He has moderate pulmonary hypertension. The aortogram showed that he has a mildly dilated aorta and there is moderate to severe aortic regurgitation, again on aortography we can't really tell if it's central or paravalvular. The prosthetic valve was not crossed. The formal read on his transthoracic echo showed that his LV was dilated, 5.6 cm. He did have mild LV dysfunction and again it was a moderate to severe - they interpreted the transthoracic as paravalvular regurgitant as I said, the TEE we felt that there were 2 separate jets.
So here is the spectral Doppler on him showing that the aortic stenosis does not appear to be as bad but the aortic regurgitation with the pressure half time of 298 seconds puts him in the moderate to severe category. So I think this is valve failure more from AR than AS, but clearly a mixed picture.
So we got a TAVR CTA on him that showed that he had excellent common femoral arteries as well as the external iliacs and common iliacs so our plan was to perform a transcatheter aortic valve within a surgical aortic valve prosthesis using a 26 mm SAPIEN which is a balloon expandable transcatheter heart valve. We were going to approach him from the right common femoral artery. The temporary pacer was placed in the left common femoral vein, we do use temporary pacers on all of our patients. We also elected to give him monitored anesthesia, really conscious sedation in the operating room. We have not been using general anesthesia in these transfemoral TAVRs because it's just so much better tolerated especially for someone who is 90 years old. And then our backup plan was if the AR persisted post-TAVR then we could potentially also offer him a PVL, paravalvular leak closure device, which are these just nitinol discs or Hershey Kiss shaped devices that just fill that space to prevent that regurgitation from continuing.
So here is his aortogram in the OR, again since his creatinine was 1.9 we did not give a lot of contrast for this but you can see here is the surgical prosthesis, contrast in the ascending aorta, the left ventricle is down here and you see the contrast coming back down into the left ventricle. And because there were two approaches into the left ventricle we wanted to make sure that we had the wire through the central hole and not in the paravalvular hole, so we are using 3-D TEE and you can see the wire traversing right through the surgical prosthesis into the left ventricle here. So we felt good about our wire placement. And then here is the balloon expandable prosthesis being inflated within the surgical prosthesis. It's this SV prosthesis, the height is about 19 mm. When we place these we get very concerned about obstruction of the coronary artery, so that's another reason why we do that TAVR CTA ahead of time. We had scrutinized that closely and the left main and the right coronary artery they were coming off at about 18 to 20 mm above the surgical prosthesis so we knew that the risk of coronary obstruction was extremely low. If we do have a case where we are concerned about coronary obstruction we will put a guide catheter into the left main or into the right coronary artery and do the valve replacement with the guide catheter in place so that we have access to the coronary because that can be a catastrophic event should there be occlusion of the left main coronary artery acutely. But again since this patient had high takeoffs we felt very comfortable with just proceeding straight up.
So there is the balloon inflated and expanding the SAPIEN prosthesis. You can see the TEE probe was down so we were still just using conscious sedation but for the deployment and for the crossing of the valve we did put the TEE probe down to assist us in the delivery of the prosthesis. So you can see we were quite pleased with our final implant, we wanted to be right flush with the top portion of the surgical prosthesis and that's really where we ended up.
And then we have our aortogram post procedure. So now again a pigtail catheter in the aorta injecting die up here. There is really no regurgitation, so that paravalvular leak really I don't know if - it's unlikely that we changed the surgical prosthesis because of such a fixed frame but we really didn't see much on subsequent imaging either with aortography or with the transthoracic echo.
So that frame, there is a surgical sewing ring there and we just literally force that device with the balloon such that the metal is now on the metal. So there have been no cases of valve embolization in this type of procedure. Very rarely there can be valve embolization if the implant depth is not appropriate or also if someone just has pure native aortic insufficiency, no prior surgery, we are just treating the aortic valve with no calcium and if you try to deploy a valve in that situation the valve embolization rate does tend to be in the 30 to 40% range, so that's why we are still not treating native AR with transcatheter heart valves but there are devices in development in that situation.
It's a great question and I think it does in certain situations so if you know that the patient is going to get a large bioprosthesis surgically implanted then yes, then down the road there are future therapies. And I'm going to hold that thought because I have a slide here addressing that specific question.
Any more questions please speak up.
Okay, so if we look at the valve and valve timeline we've been doing these now for over a year and doing them in clinical trials even for several years. And I think this really just highlights the - you know these patients that have severe bioprosthetic valve failure from AR are our most grateful patients because they come in in class IV heart failure and they leave in class I heart failure, it's that dramatic. So no doubt there is a lot of data that has led to the FDA approval of both commercial valves for this indication. And this gets to your question, Saleem, so this is data that was published in JAMA several years ago looking - this is from an international registry of patients who have been treated with valve and valve therapy, what I just showed you, a transcatheter valve within a surgical valve. And these independent predictors came to bear out to be problematic and one was the surgically implanted bioprosthesis. So such that if someone had a 21 mm or smaller valve they had a worse outcome than patients who had a greater than or equal to 21 mm valve. And that just probably gets to the fact that there is some patient prosthesis mismatch and so you have a small orifice and now we are putting more material within that orifice so we can exacerbate that particular condition.
And then also the type of valve failure is stenosis patients did worse than the regurgitation patients, which is what I'm showing you. Our patient had a greater than 21 mm valve, had AR as the primary mode of failure and then the last clinical predictor is whether it's transapical access or transfemoral access and as I showed you our patient had transfemoral access. So you know so there are other variables that you have to consider. But if you do have a larger patient with a large anulus that's going to get a big valve I think that's a provocative question. Now a lot of the surgeons are saying but we really should be using more mechanical prostheses because they are durable and then you don't have to subject the patient to repeat procedures. And I get that too. So I think it really comes down to having a discussion with the patient and really weighing the pros and cons of both approaches.
And as I said you know these patients are dramatically improved. This is data from all TAVR patients and you can see that 90% of them come to us in class III or IV heart failure and only 13% remain in class III or IV heart failure after their procedure. And the overwhelming majority are you know in class I or II. So when we can offer people a therapy that keeps them out of the hospital and makes them feel better and makes them live longer it's obviously a very impressive technology.
And remaining questions are still arterial access, stroke rates. I know this is a common question in paravalvular regurgitation so I will just go through these fairly quickly for the sake of time. So with respect to access routes I told you our you know go-to approach is a transfemoral because of the data consistently in multiple trials has shown the transfemoral leads to better outcomes. Word of big - if the transfemoral site is not usable then we are one of the highest centers in the country at using a subclavian access. We do use both left and right subclavian. We really have not done a direct aortic, we've done one over the past 18 months because again that's a thoracotomy, it's a tougher recovery for the patient.
Other approaches, the transapical with the balloon extendable valve I think is really going to go away because we keep seeing over and over again that this is another thoracotomy that patients have a hard time recovering from. Some centers have advocated using a venous access into the right common femoral vein creating an AV fistula with the aorta and the IVC, performing the TAVR then through the arterial system and then closing the AV fistula at the end with an Amplatzer device. We have not done that yet because again this center does not do subclavian access so we find that our alternate access go-to site is right here. And then there was a big French registry using carotic access, so that is another possibility. We have yet to do that, we've looked at a few patients. The stroke rates with carotic access are probably double so you know you have to weigh that and factor that in as well.
And quickly about stroke rates since we are talking about carotic access. Now this is just stroke rates across all you know patients who get transcatheter valve therapies. Different trials had different results. This was the first landmark study that was published. If you recall the stroke rate was 7%, people were wiring in the journals this is crazy, we are giving patients strokes and they could have a very good therapy otherwise. And you can see with improvements in the technology, with improvements in physician education the stroke rate is now probably somewhere between 1 and 4%, and that's typically what we quote to patients in our center.
And then lastly, paravalvular leak after the valve is placed, originally you know there were some problems with the initial designs of the valves but each company recognized that and quickly turned it around specifically. The Medtronic Core Valve you can see their original skirt just followed the frame on the inflow portion of the valve and when they came out with their second generation valve they brought that skirt down and that significantly improved things. They also changed the shape, it's now much more barrel shaped to sort of fill that anular space.
Edwards also had design improvements, these are both cobalt chromium frames but what they did was they actually put the sealing skirt around the inflow portion of the valve and that has dramatically improved their PVL rates.
This is one of the newer valves that we are using in study patients, it's called the Boston Scientific Lotus Valve, this is a braided nitinol frame. This is the only valve that's truly repositionable so we really like it. It has these intricate locking mechanisms that you can see here that allows us to fully place the valve, see its function and then decide if we like what it's doing to the conduction system, what it's doing with respect to the degree of paravalvular regurgitation. And if you don't like it, if it's the wrong size you can completely recapture, take it out of the body and upsize or downsize. So very interesting technology.
And then this is the Mitroflow Valve, which really has no metal at all and just has these rings both on the ventricular and aortic side. We just sort of anchor it within the anulus and using these devices you can see now that the PVL rates initially were in - these are moderate to severe PVL which we know correlates with poorer outcomes after TAVR, so we don't want to see high numbers here, 14 and 9% were the first you know big studies that were published using a core valve and since the design improvements have been enacted you can see the PVL rates have significantly declined now that we are in single digits.
So couldn't do it without the whole team. We've got a large team and I probably am leaving a lot of people out and we have Julie here is one of our team members, and a special shout out to Liz, Lisa and Donna who really run the Aortic Valve Outpatient Center.
So in conclusion TAVR has now been demonstrated to provide mortality benefits for those patients who are considered to be intermediate, high or extreme risk for surgical AVR. Procedural complications including CVA have decreased considerably with the improvements in technology and physician experience. And current randomized trials are enrolling low risk patients, we being one of those centers. Really the only cutoffs are going to be they would have to be at least 65 years of age and they cannot have a bicuspid valve. And they will come in one to one randomization to surgical AVR versus transcatheter valve. And there are also ongoing discussions with the FDA about designing a trial to evaluate the performance of TAVR in asymptomatic patients as well.
Thank you very much.