COMMENTARY

TAVR vs SAVR in Low-Risk Patients: What We Do and Don't Know

Wayne B. Batchelor, MD, MHS; Tom C. Nguyen, MD

Disclosures

March 16, 2021

Wayne B. Batchelor, MD, MHS: Hello, everyone. I'm Wayne Batchelor. I'm the director of interventional cardiology and interventional cardiology research and education at the Inova Heart and Vascular Institute in Falls Church, Virginia. I'm joined by Dr Tom Nguyen, who is a cardiothoracic surgeon and chief of cardiothoracic surgery at the University of California, San Francisco. He's also co-director of the Heart and Vascular Center. Welcome, Tom.

Tom C. Nguyen, MD: Thank you, Wayne.

Batchelor: Since 2019, we've been seeing more transcatheter aortic valve replacements (TAVRs) performed in the United States than surgical aortic valve replacement (SAVR). This is much to, perhaps, the chagrin of some surgeons, but I think most surgeons are really on board and see the value of both therapies.

Nguyen: As a surgeon, I recognize that there are many ways to treat patients with structural heart diseases, both at the aortic and the mitral valve. Because of that, I think it's really important that surgeons and cardiologists be involved with these newer technologies because I think there's a big role for them in treating high-risk patients and intermediate-risk patients as well as the low-risk patients.

Today's conversation will focus on patients who are low risk.

Batchelor: Maybe you can talk a little bit about some of the recent trials.

Nguyen: The seminal trial that we should focus on when we talk about low-risk patients was in 2019, called the PARTNER 3 trial. There are a couple of really important things to know about that trial. The average age of TAVR patients was 73 years and the average age of SAVR patients was 73 years. The average STS score was 1.9% and the primary endpoints were mortality, stroke, or rehospitalization.

For TAVR at 1 year, the primary endpoint was 8.5%; for SAVR it was 15.1%. The sub-analysis, interestingly, for that low-risk trial showed no difference in death between TAVR vs SAVR. Really, that study kind of put TAVR for low-risk patients on the map and really eventually opened the door for the FDA approval of TAVR for the treatment of patients who are low risk.

Later, the 2-year results came out, and what it showed was that TAVR is still favorable for patients who are low risk, but interestingly, not as significantly as before. What also was interesting was that TAVR is still noninferior to SAVR at 2 years, but the main difference was due to a significant difference in rehospitalization.

This is something that isn't really discussed much, but at 2 years there was actually increased incidence of death, stroke, and valve thrombosis in TAVR vs SAVR. If you look carefully at the curves that have been published, you start seeing the curves between TAVR and SAVR starting to converge a little bit at 2 years.

There's still a lot to know. I think TAVR is a potentially viable option for this patient risk cohort, but I do think it's important for us as clinicians to really know the data.

Batchelor: There's another important trial that came out around the same time — the Evolut Low Risk. Of course, this was randomizing patients to receive the Medtronic self-expanding prosthesis vs SAVR. Similar endpoints, slightly different; this one included all-cause death or disabling stroke at 24 months. It carried out the follow-up a little longer, perhaps to address some of those concerns that you had in regard to the 2-year PARTNER 3 data.

Interestingly, within that study there was no real increased mortality seen with TAVR. In fact, TAVR fared extremely well. Of course, there's meta-analytic data with systematic overviews of all of these studies, including PARTNER, Evolut Low Risk, SURTAVI, and NOTION, that suggest that the point estimate still favors TAVR.

As an interventional cardiologist, I might be sticking my chest out a little bit here, thinking that I've got some advantage. I think your cautionary notes can't be underemphasized. For example, these trials systematically excluded the bicuspid cohort. What are those anatomic subsets that you as a surgeon would really be careful about when choosing between TAVR vs SAVR?

Nguyen: It's important to emphasize that the results of the trial only apply to the study population. You can imagine in a randomized controlled trial that you have a list of inclusion and exclusion criteria, and you essentially have a study population and the results from it. The results only apply to that study population. The reason why that's important, and you hinted at it, is because there are strict exclusion criteria. For the PARTNER 3 study, there are 31 exclusion criteria. For Evolut/CoreValve, 17% of patients were excluded; 34% of patients were screened out. That's really, really important.

We don't have any published data at least beyond 2 years. That's really important as well. If a patient comes in who's low risk and says, "Hey, doctor, what can we expect at 5 years, 6 years, or 7 years?" We don't have any published randomized data in that respect. It's not saying that TAVR or SAVR is better. We have to be honest about the stuff that we don't have any data behind.

Another thing I think is important is that the average age of the patients was 73 years. If a patient who's 65 and low risk says, “Hey, what should I expect?” We have to remind ourselves that the average age of these patients in the trial was 73, so we really don't know.

In the trials, the concomitant surgery rate was 7.9% for TAVR vs 26.4% for SAVR. That's important because the primary endpoints were death, stroke, or rehospitalization. You can imagine that if a patient had a SAVR and a CABG vs a TAVR PCI, then potentially the SAVR and CABG patient would have a higher likelihood of being rehospitalized. That may explain why the primary endpoint was a little bit higher for SAVR. We don't know.

Interestingly, despite the fact that bicuspid patients were excluded in the low-risk trial — and actually all of the randomized trials so far in TAVR, — the FDA has approved TAVR for bicuspid patients. But we don't have randomized data comparing bicuspid TAVR vs SAVR.

Batchelor: With some of the newer valve prostheses that we're implanting, I think we have pretty good long-term data from the surgical literature. We're getting there with TAVR. We certainly have up to 5-10 years and so far it looks pretty darn good. The nature and frequency of structural degeneration of a TAVR valve seem to be fairly similar to those of a bioprosthetic surgical valve.

In all honesty, we just don't have that 10- to 15-year data. When we're playing this out over the entire length or duration of someone's life, when we're making decisions in perhaps 50-, 60-, up to 70-year-old patients who are low risk, we have to recognize that they're probably, in some cases, going to need one, two, or even three more procedures.

Nguyen: Right. It is a little more difficult to do a coronary intervention if you have a prior TAVR because it's harder to cannulate the coronary ostia. There's a lot that we know we don't know. I'm a huge advocate of TAVR. I believe it's the future. I just think we, as clinicians, need to be very honest and transparent with what we know and what we don't know when we have shared decision-making conversations with our patients.

Batchelor: For the patient who you think is going to need two to three more procedures during their lifetime, do you think there's any advantage to doing surgery first, perhaps with root enlargement vs reserving surgery for the second or third operation? How does this play into your surgical decision-making?

Nguyen: There are a lot of creative options for these lower-risk patients who may want a tissue valve in the setting of not knowing the long-term durability of TAVR but wanting something done to minimize the risk of needing open-heart surgery. Essentially, what we're trying to come up with is an algorithm to minimize the patient needing open-heart surgery.

So we sit down and have an honest conversation with the patient. One potential option is that if a patient really wants to have something done to statistically minimize the need to have something additionally done, if they're young, we recommend a mechanical valve. We have some pretty good data for younger patients with mechanical valves and the likelihood of the need for re-intervention.

The drawback, obviously, is that you need to be on warfarin and it's lifelong. For patients who don't want to be on warfarin and don't want a mechanical valve, an algorithm was proposed by Michael Mack, where you potentially consider a TAVR at the beginning.

You can imagine an algorithm where you do a TAVR when they're 55 or 60 years old, but also explain to the patient that we don't have a whole lot of data on TAVR in 55- or 60-year-olds with bicuspid valve stenosis, but we think it'll last this long. When the patient becomes 60, 65, or so, then he or she will buy their first sternotomy to go in there and take that valve out, which is relatively easy but not inconsequential.

We actually published this at a recent meeting about explanted TAVR valves and the consequence of that. It's doable. From there, you can put in either a mechanical or tissue valve (most likely a tissue valve).

As you alluded to, we surgeons putting in that tissue valve need to make sure that we put in the largest valve possible because we need to think about the long-term plan for that patient. I would advocate that we try to put in at least a 23-mm valve. If we can't, we need to do a root enlargement, as you mentioned.

Hopefully, by then the patient is 60 or so, and that will potentially buy another 10-15 years. After that, if something happens, they could do a TAVR inside that bioprosthetic valve, called a valve-in-valve.

Essentially, that will get the patient 30-35 years or so, where he or she would have only had an open-heart procedure once. One thing we need to know is that there isn't a whole lot of data about valve-in-valve either, as far as durability data. Again, it's just important to be very honest and transparent about what we know.

Batchelor: We're starting to get some emerging data on TAVR within TAVR — so-called TAV-in-TAV vs TAV-in-SAV. In the Redo-TAVR registry, it's over 63,000 patients examined from about 2005 to 2019. They drilled it down to around a little over 600 patients who had other TAV-in-TAV or TAV-in-SAV, and found on a propensity-matched analysis lower gradients in the TAV-in-TAV patients but a little bit more aortic regurgitation.

In the TAV-in-SAV, higher gradients probably related to, again, surgical sutures and just the amount of hardware that's taken up by the surgical valve prosthesis, but less aortic regurgitation. I think TAVR works very well for a preexisting TAVR or SAVR, but there might be some minor nuances in terms of the differences between gradient and regurgitation between those two strategies. At least it looks like it's quite safe and appropriate to consider in either bioprosthesis.

Any comments on that?

Nguyen: It's a research topic of mine. I present on this fairly frequently and there's not a lot known. Here's what we know. I think in a high-risk patient who's had a prior bioprosthetic valve, TAVR valve-in-valve is not a bad option.

In a low-risk patient who has had a prior bioprosthetic valve and you have structural valve deterioration, I think most would agree that if you send them to a tertiary care center that is comfortable doing the operation, the risk of mortality for a reoperation is relatively low. This patient should undergo reoperation and then get the valve replaced.

I think in general, most folks would agree that a high-risk/prohibitive-risk patient should get a TAVR valve-in-valve. A lower-risk patient in experienced hands will probably do very well with reoperation, get a surgical valve in place with the future plan of potentially valve-in-valve. For the intermediate-risk patient, we really don't know.

Batchelor: The bicuspid group of patients is really an important group because, again, they were systematically excluded from the trials that we were just discussing in low risk. That's a major proportion of some of the younger patients considered for TAVR or SAVR.

Are you aware of any randomized prospective trials? I'm not, but I may be missing something because really that's where the money is, so to speak, in terms of understanding the real truth.

Nguyen: Yeah, you're exactly right. You hinted at it. The bicuspids are often the ones that are lower risk, that tend to develop aortic stenosis earlier. We don't have much data in the lower-risk, younger patients who have bicuspid stenosis.

When we talk about databases, I do think that something that we should try to move toward as specialties and as societies is closer integration and communication with our respective databases. As you know, all TAVRs are recorded and logged in the TVT Registry and all SAVRs are in the STS National Database. Those databases don't communicate with each other, so it's hard to do a study. It's doable because studies have been done, but there's a lot of work involved. We need more studies looking at data from both databases, I think there's a lot of potential if we do that.

Batchelor: Having worked in multiple centers in the past, the best decision for a patient largely rests on the local milieu and the local expertise. In some cases that might bias toward the TAVR or in some cases it might bias more toward a surgical option. Do you want to speak to that in terms of how that plays into the multidisciplinary heart team?

Nguyen: We need to work very closely with our cardiology colleagues in a multidisciplinary fashion, in a heart team approach, to do what's best for patients. I think that's the most important thing that we need to do as surgeons. Probably even more important than that, I think we need to step back a little bit and not look at it as "surgeons vs cardiologists." I think as clinicians and as physicians, we need to think of ourselves as specialists who treat patients with structural heart diseases.

Eventually, we need a change of paradigm where we come up with specialists who have a little bit of a hybrid approach between a cardiologist and surgeon that can really efficiently and effectively do both procedures very well, and that will allow us to push the field.

What you alluded to is fascinating because if you have a hammer, then everything looks like a nail. Of course, if a patient with aortic stenosis or mitral regurgitation comes in to a cardiologist's office, that doctor is going to prefer a TAVR or a MitraClip; if they come in to a surgeon's office, they would be recommended surgery.

I would argue — and I know it might be somewhat debated, but we're all intelligent; we all went through medical school, etc. — that it's our training process that has biased us toward one decision or the other. It's not our inherent level of intelligence, per se.

What I'd like to eventually see is a deconstruction of that, where our training process is not cardiology specific or surgery specific but disease specific, where someone is treating patients with structural heart disease. That's what I'd love for the future. That's likely a pie-in-the-sky idea that may or may not happen, but I think it's the right way to approach diseases.

Batchelor: I think you're seeing a number of programs adopt that exact philosophy whereby, for example, the cardiothoracic surgical fellows are training in the cath lab more, and there's cross-fertilization and rotations going back and forth in both directions. I think what you're saying is extremely germane, and I think that training programs are recognizing this. The future bodes better than certainly the training experiences in the past.

This has been a really phenomenal discussion. In summary, we now have data from some pretty decently performed clinical trials on several thousand patients, indicating that when you exclude bicuspid disease, concomitant severe coronary artery disease, and severe aortic regurgitation, there appears to be no incremental mortality risk in TAVR. In fact, the trend might be in the opposite direction.

I think the cautionary note is that we really have excluded a good number of patients that are low risk, probably the majority of patients that would be coming to us on a daily basis. There still needs to be room for further investigation of how to manage young bicuspid patients, especially those who have aortopathy.

I think there's still some equipoise as to how to manage the patient with concomitant coronary artery disease. What do we do with the low- to intermediate-risk patient who has stentable multivessel coronary disease — maybe not left main, but stentable triple vessel? How do we manage that patient?

I think the reality is that because there are no obvious clinical randomized trials addressing some of these groups that we specifically refer to, including the bicuspid cohort, we're going to unfortunately have to rely on cohort studies that are derived from the registries that we have, which have some limitations but have as their strength a large amount of volume moving forward.

I want to bounce it back to you for any last suggestions or comments on how we can make better sense of this, make better decisions for the patients who face this, and also arm our academic community with better data on which to make guideline changes and recommendations.

Nguyen: Thank you, Wayne. I enjoyed the conversation as well. You mentioned several times the word "patient," and I think that we always need to remind ourselves that whatever we do, it needs to be patient centric; we need to put patients in the middle. In doing so, we need to have very honest and collaborative conversations with everyone at the table.

Batchelor: I think we've gone full circle in this conversation. It was really great to talk to you, Tom.

Nguyen: Thank you, Wayne.

Follow theheart.org | Medscape Cardiology on Twitter

Follow Medscape on Facebook, Twitter, Instagram, and YouTube

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....