TAVR in the US: Matured or Still Evolving?

An Expert Interview With Dr Susheel Kodali

Tricia Ward


March 21, 2016

Transcatheter aortic-valve replacement (TAVR) was approved by the US Food and Drug Administration (FDA) in late 2011 for the treatment of patients with severe symptomatic aortic stenosis deemed inoperable. This was on the basis of data from cohort B of the first PARTNER trial.[1] Subsequently, the procedure was approved as an alternative to surgical AVR for patients at extreme surgical risk[2] (Society of Thoracic Surgeons operative risk score ≥8% or at a ≥15% risk of mortality at 30 days). In 2015, the indications were expanded to include valve-in-valve repair of failed surgical bioprosthetic heart valves.

The European TAVR experience is longer, with the first devices receiving CE mark in 2007 and a broader selection of valves available to operators. The devices in use in the United States include the balloon-expandable Sapien valve (Edwards Lifesciences) and the self-expanding Corevalve (Medtronic). Since their initial approval, newer versions and a wider range of sizes have become available.

theheart.org | Medscape Cardiology interviewed Susheel Kodali, MD, assistant professor of medicine and director, Heart Valve Center, New York Presbyterian/Columbia Medical Center in New York about the current status of TAVR in the US and what to look for at the upcoming American College of Cardiology Scientific Sessions (April 2-4, 2016).

Susheel Kodali, MD

theheart.org | Medscape Cardiology: From the first days of doing TAVR, how do the latest versions of the FDA-approved TAVR devices, the Sapien 3 (Edwards Lifesciences) and the Corevalve Evolut R (Medtronic), compare with the earlier versions?

Dr Kodali: The evolution has been on multiple fronts, it's not just the devices. It's how we screen, how we operate; more experience overall has led to improved outcomes. The devices themselves contribute greatly, and the Sapien 3 has some significant advantages over Sapien XT in that it's a lower-profile system. It's a more forgiving valve in terms of paravalvular leak because of the skirt and everything else. But in addition to that, we are using CT sizing much more now than we did during the Sapien-XT era, so that obviously allows for better sizing and contributes to the excellent results.

In terms of the Medtronic Evolut, there are several real advantages that make it a much more user-friendly system; it's a true 14-French–equivalent system with the inline sheath. Patients who were done with alternative access in the past are now done transfemorally. My personal bias, and there has been some analysis to support it, is that the transfemoral approach has significantly better outcomes vs a transapical or a transaortic approach. The alternative accesses are surgical procedures that have a longer and more difficult recovery for the patients. If you do TAVR transfemorally, the outcomes are better.

We did an analysis several years ago. Gene Blackstone[3] from the Cleveland Clinic led it for the PARTNER trialists, where he did a propensity-matched analysis looking at transfemoral and transapical access routes. The outcomes were better with transfemoral access compared with transapical in terms of hard end points like mortality. Getting to smaller devices has helped as well. The other big thing with the Evolut device, for newer sites, for lower-volume sites, is the ability to reposition the device. It's not a fully retrievable valve, but you can get to 80% deployed and if you don't like the position, you can recapture it and reposition. That gives a level of comfort, especially for lower-volume operators.

theheart.org | Medscape Cardiology: Isn't there a concern that repositioning or increased device maneuvering might increase the stroke risk?

Dr Kodali: Any additional manipulation of the valve poses a potential concern for more embolic risk. Ideally you are going to position this appropriately on the first time and avoid repositioning. However, if you are fixing a malposition or a pop-out of the valve I think it is worth the risk/benefit threshold. In addition, the way that the Evolut deploys (it centers in the valve) makes it easier to get the correct position the first time.

theheart.org | Medscape Cardiology: The procedure was originally approved for inoperable patients, and now it is indicated for high-surgical risk patients. There will be data coming out at the ACC meeting on intermediate-risk patients, and the Low Risk Trial was recently approved. Do you see TAVR inevitably moving into lower-risk patients?

Dr Kodali: There is a trend in the US toward treating lower-risk patients if you look at the [Transcatheter Valve Therapy] TVT registry.[4] We have evolved our risk threshold in terms of who is appropriate based on our own experience and based on data. When we look at the initial trials in inoperable[1] and high-risk patients,[2] we learned that some of those patients were just too sick for any treatment. The heart team concept with surgeons and interventional cardiologists evolved to determine what they think is an appropriate utilization.

For example, a 90-year-old man undergoing a reoperation is technically going to be at intermediate risk. However, a lot of the heart teams across the country have said that in their opinion, that patient is high risk even though his STS score may only be 5%. The mean STS score in a high-risk transfemoral arm in the latest publication from TVT is 5%.[5] That illustrates an evolution in terms of how people are applying the therapy and what they consider high risk for surgery.

And what is high risk at one center may be different at another center. The surgical outcomes are going to be better at a high-volume surgical center than a low-volume surgical center. What one surgeon might consider high risk another surgeon will consider intermediate risk. Regardless, there has been an evolution toward treating lower-risk patients. We don't have any randomized clinical trial data comparing surgery vs TAVR in intermediate-risk patients except the NOTION trial,[6] which was published last year. It's from three centers in Europe and is an underpowered study. It was in all-comers 70 years or older, and their mean STS score was around 2.9%. There were fewer than 300 patients in the study, so as I said it was underpowered, but it suggested that the procedures were equal.

We are waiting for ACC, where PARTNER IIA, a large randomized trial of over 2000 intermediate-risk patients will be presented. They were randomized to surgery or TAVR with the Sapien XT system.

theheart.org | Medscape Cardiology: With the availability of a wider range of valve sizes, are there still patients in whom you would like to do TAVR, but their anatomy is unsuitable for the devices on the market?

Dr Kodali: Occasionally, but that is rare at this point. We do see anatomies that are not as well suited for TAVR (some bicuspid valves). Everything is in the risk/benefit threshold. If we see severe left ventricular flow outflow tract (LVOT) calcium, where there is a lower likelihood of getting a good result with a transcatheter valve and the patient is in that gray zone—not terribly high risk—they may be better suited for surgery. That is the direction we are moving toward—we are going to see whose anatomy is suitable for which procedure.

theheart.org | Medscape Cardiology: There are reports of minimal TAVR done under local anesthesia without transesophageal echo [TEE]. Is that being done much?

Dr Kodali: Part of it is driven by resource utilization because you have to be realistic, and part of it is driven by patient care. Avoiding anesthesia and TEE in these older patients ideally will avoid some complications. However, we have to be cautious about that because we don't want to sacrifice the excellent outcomes we have seen. With TEE imaging we can really assess the results in terms of paravalvular leak. If there is hypotension we can assess what's going on relatively quickly. There is a benefit to TEE.

Obviously, general anesthesia has its own risk thresholds. I think there is an intermediate approach. At high-volume centers that have extensive experience, I think it is appropriate to consider conscious sedation and transthoracic echo and fluoroscopic imaging, but those same centers should have a threshold to use a TEE probe if there are any concerns. In the end, you want to make sure you don't sacrifice the outcomes. The way I look at it is if somebody has high-risk anatomy— for example, there is a lot of LVOT calcium and sizing is borderline or there is some risk of coronary occlusion—then that patient should be done with TEE because you want to make sure that you avoid those complications and get the best result possible.

If CT shows that the patient has excellent anatomy, minimal LVOT calcium, very clean sizing, that patient is appropriate for an awake procedure. Then there are the clinical characteristics; some patients can tolerate an awake procedure and others can't. All of those things need to be factored in. But that can only be done with experience. If a new center is starting a TAVR program, I don't think it is appropriate to do an awake procedure in their first 10 or 20 cases. To get past the learning curve, those centers benefit significantly from having additional imaging.

theheart.org | Medscape Cardiology: Are these decisions about sedation and imaging being made in the context of a heart team discussion of each case?

Dr Kodali: Yes, absolutely.

theheart.org | Medscape Cardiology: What is the current status of cerebral-protection devices?

Dr Kodali: There are two main cerebral-protection devices that are already approved in Europe and that are either in trials or starting trials in the US. The Sentinel cerebral protection device (Claret Medical) is a dual filter placed using the right radial approach; one filter protects the innominate (brachiocephalic) artery and the second protects the carotid artery. The CLEAN-TAVI trial was presented by Axel Linke from Europe at TCT.[9] It showed a more than 50% reduction in embolic events by MRI with the Sentinel cerebral-protection device. Clinical events such as stroke were too low frequency to show a difference. It's hard to show an effect on other clinical events except with the larger trial.

The US Sentinel trial was initiated. It's a randomized trial with an MRI end point as the primary end point similar to CLEAN-TAVI, but there are also important secondary end points such as neurocognitive function. The idea behind the secondary end points is that that even if there is not a reduction in stroke, there may be an improvement in neurocognitive function by having fewer embolic events on MRI.

When we look at stroke scales, they cover only a fraction of the brain. The whole frontal lobe, for example, is not really assessed by these stroke scales. Declining cognitive function may be due to embolic events that go to that frontal lobe, and that is why neurocognitive testing is going to be important to assess any clinical benefit for these filters.

The Sentinel trial is with the dual filter. There was also the DEFLECT trial[7,8] with the TriGuard HDH (Keystone Heart), which is a shield that is placed across the aortic arch that deflects debris from going to the brain by deflecting it downstream. They are also in the process of initiating the REFLECT trial looking at the benefit of embolic protection in the US.

theheart.org | Medscape Cardiology: These are in clinical trials only and are not routinely used, correct?


Dr Kodali: Yes. They are not approved for use in the United States.

theheart.org | Medscape Cardiology: Can you comment on the bioprosthetic thrombosis issue that arose last year? These thromboses were seen in bioprosthetic surgical valves and transcatheter valves.

Dr Kodali: It started with the PORTICO IDE trial, which used the St Jude TAVR device. When the clinical trial began they had mandated CT at 30 days. This was 4-D CT, which is basically a CT over time, and when the investigators looked at these CTs they noticed there were significant incidences of abnormal leaflet motion. In the PORTICO-IDE trial it was seen in almost 40% (22 of 55) of patients.[10] The finding led to a temporary hold on the trial until people understood the phenomenon better, whether it's called abnormal leaflet motion, leaflet thrombosis, or clinical leaflet thrombosis (a lot of different names have been used).

After PORTICO IDE, there was some interest in two registries in Europe and in the US (the SAVORY; and RESOLVE registries, respectively) Routine imaging was not mandated in these registries, but they imaged at different time points (the median interval was 87 days). And they showed a 13% incidence of what they called "possible subclinical leaflet thrombosis" with different devices, the Sapien XT, Core Valve, Portico, as well as surgical valves.[9] The question then was, is this an imaging phenomenon, is there variable incidence with different valve types, and are there certain patient scenarios that are at higher risk? It's unclear, and we don't want to overreact to an imaging finding.

There is a suggestion of a possible link to neurologic events. However, we have to be cautious about that. There is no real temporal and causal relationship between the CT scan and the embolic events. It is something that bears further investigating. It is a real event, not just an imaging finding, because in the New England Journal of Medicine paper[10] they showed that the leaflet motion improved in every patient on therapeutic warfarin, whereas it did not improve in those that were not on therapeutic warfarin, suggesting that it's a real phenomenon. But we need to understand the link to clinical events, the true incidence, who is at risk and how to follow these patients. We are going to need a lot more data. It's going to be collected, and I'm sure the FDA is going to mandate that as part of future trials.

For now, we don't know enough to say definitively who is at risk, what should be done, how the patients should be followed, or what type of anticoagulation regimen is best. You could say "why not put everyone on warfarin?" but that is not benign because these are elderly patients at high risk for bleeding. And in this patient population, bleeding leads to higher mortality. It's not benign to anticoagulate every TAVR patient. There are trials being started with new oral anticoagulants, with dual antiplatelet therapy as well as single antiplatelet, that are going to look at this issue further. And so we will see. I think we just need to collect more data.

theheart.org | Medscape Cardiology: Do the current recommendations[11] to use clopidogrel in the first couple of months after TAVR, and then just lifelong aspirin still stand?

Dr Kodali: Correct.

theheart.org | Medscape Cardiology: Is the thrombosis issue affecting patient discussions on valves in terms of staging procedures? For example, a patient might be considering choose a bioprosthetic surgical valve with the understanding that in 10 or 15 years they could have a valve-in-valve with a TAVR.

Dr Kodali: We haven't seen any clinical correlation [for the leaflet thrombosis], but we keep it in our minds. I'm not using this to make my decision at this point, but I will always raise it as a possible limitation of the therapy, especially if we potentially move into lower-risk patients. As I noted earlier, the mean STS score in the TVT registry has declined over time. The risk threshold has decreased and maybe this will impact that, but there are not enough data to my mind right now to suggest that.

theheart.org | Medscape Cardiology: Is there anything that you see in Europe that you wish you had in the US?

Dr Kodali: Yes. They have multiple devices approved. We are getting there. The REPRISE III trial with the Lotus device (Boston Scientific) and the SALUS trial with the Direct Flow device (Direct Flow Medical) are ongoing. Other devices will enter into trials in the US, and we will see how they compare with the existing devices.

theheart.org | Medscape Cardiology: At the upcoming American College of Cardiology Scientific Sessions, what should our audience look for on TAVR?

Dr Kodali: There will be the two large trials in intermediate-risk patients and there is an interesting late-breaking trial looking at the impact of procedural volume on outcomes. In addition, there will be numerous other analyses, and we are learning more each time. But the big things to look for are the intermediate-risk data that will be presented (from PARTNER II and SAPIEN 3). I am interested to see what impact procedural volume will have on mortality (TVT Registry data).

theheart.org | Medscape Cardiology: Finally, what is next for TAVR? Do you foresee another big leap, or are we just going to see small incremental developments?

Dr Kodali: We still need devices to address the issue of aortic insufficiency. Bicuspid valves are a new indication that is going to be evaluated. I think we will see iterative improvements not only in the devices, but also in the patient populations.

Disclosures: Susheel K. Kodali, MD, has disclosed the following relevant financial relationships:
Serve(d) as a director, officer, partner, employee, advisor, consultant, or trustee for: Edward Lifesciences; Thubrikar Aortic Valve; Claret Medical; Meril Lifesciences; VS MED Tech
Served as a speaker or a member of the speaker's bureau for: St Jude Medical
Received research grants from: Edwards Lifesciences


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.