Bioprosthetic Valve Thrombosis: A Perhaps Not So Rare Complication

This transcript has been edited for clarity.

Christopher V. DeSimone, MD, PhD: Hello, and welcome back to the Mayo Clinic–Medscape video series. I'm Dr Christopher DeSimone, a cardiologist and medical director of CV marketing at the Mayo Clinic in Rochester, Minnesota.

Today we will be discussing bioprosthetic valve thrombosis (BPVT). I'm joined by my colleague, Dr Sorin Pislaru, consultant, professor of medicine, and an expert in this area. Welcome, Dr Pislaru.

Sorin V. Pislaru, MD, PhD: Thanks for having me, Chris. It's a great pleasure to be here to talk about BPVT.

DeSimone: I know this is an area of passion for you. To start off, what is BPVT and when do you find it occurring?

Pislaru: This is a long story. Carpentier implanted the first xenograft in a patient in the 1960s. At that time, neither echocardiography nor cardiac CT was available for routine evaluation, so it took a little while for the cases of BPVT to trickle into the literature.

For a long, long time the disease was considered to be a rare event. There are case reports of one, or at most two, patients who had BPVT through the 1970s and 1980s. Then people started to realize that this may be more common than thought.

We have a group of several cardiologists and cardiac surgeons here at the Mayo Clinic who became interested in the phenomenon early on. We started looking deeper into our reports.

BPVT is, simply put, thrombosis forming over a bioprosthesis. Patients are not supposed to have that event, given that we routinely provide anticoagulation after implantation. It was felt that once you are past those 3 months in which the valve re-endothelializes, you're out of the woods.

What we've seen is that you do actually have thrombosis that occurs much later than that, meaning 2-4 years post-implantation, and even as late as 10 years after the initial event. However, the bulk of BPVT occurs somewhere between years 1 and 3, which represents the peak period for risk of that occurring.

Determining the Incidence Rate of BPVT

DeSimone: This seems like a very serious topic that could have a lot of ramifications for our patients. I know it's very difficult to determine the incidence of BPVT depending on the population that you study, but what has the incidence been in your population?

Pislaru: Just as you pointed out, this largely depends on which population you look at and the methodology you're using. We started retrospectively reviewing this in early 2011 and 2012 and noticed that the incidence was not very high. One of the first papers published by our group looked at patients who needed a reoperation for a bioprosthetic valve. In that series, about 11% of valve failures and subsequent reoperations were due to the presence of BPVT.

Of course, that doesn't represent the incidence among the total number of patients. If you retrospectively look at how many bioprosthetic valves were implanted at the Mayo Clinic during that study interval and derive a percentage from that total number, the incidence would be about 1%.

Dr Raj Makkar and colleagues had a huge impact on the literature with their 2015 paper in The New England Journal of Medicine, in which they looked systematically with four-dimensional CT (4DCT) at patients who had a transcatheter aortic valve replacement (TAVR). Following that study, the incidence becomes much higher. They quoted an incidence rate of around 10%-12% at 1 month. And there are several registries from various groups that show a BPVT incidence rate of about 10%-15% at 1 month post-implantation in TAVR.

What accounts for the different incidence rates between these two? We never looked with CTs in a systematic fashion in our population, so we could have missed some patients. At the same time, remember that not all TAVR patients in the clinical trials required anticoagulation, with some placed on dual antiplatelet therapy. Was that the problem and the reason why they have more thrombosis? It's very hard to tell. My guess is that the incidence rate is probably somewhere in between 1% and 10%.

Transcatheter vs Surgical Valves

DeSimone: Given that we're now doing lots of transcatheter valves, is there something inherently different about that valve or is it just the antithrombotic/anticoagulant regimen?

Pislaru: It's a good question and, fortunately, we now have an answer. For a long time, people said that transcatheter valves and surgical valves are different, meaning that there is neo-sinus because you now have the valve stand that's implanted there, and you have the native valve that remains in place, whereas the surgical valve will require removal of the native aortic valve.

This led to a lot of discussion about why the incidence is so low in surgical valves and so high in transcatheter valves. I always felt that they should not be all that different and that perhaps we do not find them in surgical valves because we do not look.

In the past couple of years, both the PARTNER 3 and the CoreValve Group published their results in the randomized trial for low-risk patients. Both studies showed that patients in surgical groups and transcatheter groups had a significant incidence of BPVT. It was actually in the order of 10%-20% for surgical valves, which was not that different from transcatheter valves at 1-year when the incidence was 20%-30%. So, I think the incidence rates are approximately the same with these valves, and if you start looking for it then you will start finding it.

Long-term Concerns in Patients With

DeSimone: Very interesting. Let's say you have an episode of BPVT, and you treat it however you want to treat it, perhaps with anticoagulation. You reimage the patient and they do fine. In the long term, can you tell us if there are any adverse outcomes, that these patients are at higher risk of? Should we be imaging them more rigorously? What do you think?

Pislaru: I think this too is a field in motion. Our most recent study looked exactly at this question in patients identified as having a clinical episode of BPVT, meaning that a clinician recognized the event and treated the patient with anticoagulation. We wanted to see what happened with those patients at follow-up after anticoagulation resolved the initial problem. It turns out that in terms of strokes and peripheral embolic events, these patients do not do any differently from patients who never had BPVT globally. Now, these are small numbers, and if you look at larger numbers you may pinpoint differences.

But from what we have seen, there are two important things to keep in mind. The first problem is that 1 in 4 patients who had a single episode of valve thrombosis will have a second and sometimes a third or a fourth episode. In other words, you need to be very careful even after you've stopped treating them with anticoagulation because you believed you fixed the valve and that they were home free. My personal feeling is that if the patient who had a valve thrombosis is at a very low risk for bleeding, it would not be unreasonable to consider treating them with long-term anticoagulation.

The second problem is, do you restore prosthesis longevity? If you diagnose the disease and treat effectively with anticoagulation, are you going to restore the lifespan of the valve? The answer to that is, probably not.

We took patients who had successfully treated BPVT and matched them to patients who had the same type of valve, same size, same gender, same year of implantation, and so on and so forth. By following them up long-term at 10 years, we found that patients who did have valve thrombosis are more likely to need another intervention on the valve, be that a new valve-in-valve or a surgical intervention. The chance of that was close to 70% at 10 years in those with BPVT vs about 20%-25% in the control group. So, we think that BPVT hastens prosthetic degeneration, which you could imagine is related to the thrombotic event in the first place.

We will have long-term data from the randomized PARTNER and CoreValve Evolut trials in those patients found to have valve thrombosis by 4DCT. They'll be followed up as part of the clinical trial so we can see what happens at 3, 5, and 10 years' time. That will be very interesting to see.

The Latest Treatment Guidelines

DeSimone: It's nice to have an expert like you giving us a glimpse into the future. But if I can try to pin you down for our viewers and ask you right now: If you had a patient that came in with BPVT, which anticoagulant, plus or minus antiplatelet, would you choose, and for how long do you need to treat them? I know you alluded to the duration for this being "as long as we can while being safe," but what would you say are guidelines for our practitioners?

Pislaru: In general, for every patient that has a bioprosthetic valve, or a mechanical valve, for that matter, we recommend that they are on aspirin. All of the latest editions of the valvular heart disease guidelines, including the most recent ones in 2020, support the use of aspirin for every type of prosthesis. Many times, aspirin has been replaced by clopidogrel in patients who have concomitant coronary disease and stents, and so forth. So, that's a must.

When you diagnose thrombosis on the valve, then the question is, do you go with warfarin or a direct-acting anticoagulant? Because most of our experience has been with warfarin anticoagulation, that's what we tend to go with. As you know, the current guidelines allow the use of direct-acting anticoagulants for nonvalvular indications, such as atrial fibrillation, in patients who had bioprosthetic valves, but there is no long-term evidence for that.

I was obviously very disappointed when the GALILEO trial was stopped prematurely because of excess bleeding and excess death in patients treated with rivaroxaban in bioprosthetic valves after TAVR. That was not a good signal. Now, whether this is an issue across the class is hard to tell.

We did use direct-acting oral anticoagulants in a small number of patients, and several groups have used successful direct oral anticoagulants with BPVT. I would say that right this minute, there is more evidence with warfarin than with DOACs in terms of treating BPVT.

Then the question is, for how long? First, I would advise patience for all clinicians treating those with BPVT. In data we haven't yet published but are currently under review and hopefully will come out eventually [Editor's note: It is now published.], we looked at how long it takes for the gradient to normalize if you put somebody on anticoagulation. It turns out that you need to wait for a little while. The mitral valve prosthesis responds relatively fast. If you look at patients who have mitral valve BPVT, probably within 3-6 months, almost 100% of them will resolve. If you look at the aortic prosthesis and tricuspid prosthesis, it will take 6, 9, 12 months, and sometimes up to 18 months to resolve an event of BPVT, meaning that the morphologic abnormalities on the valve assessed by CT or echocardiography have disappeared, and that the gradient has returned to normal. It takes quite a while to obtain your full restoration of valve function in the aortic and tricuspid position, so you must be patient.

DeSimone: Thank you, Dr Pislaru, for these very important insights. And thank you for joining us on theheart.org | Medscape Cardiology.

Pislaru: It was a pleasure participating. Thank you for the invitation, Chris.

Follow theheart.org | Medscape Cardiology on Twitter

Follow Medscape on Facebook, Twitter, Instagram, and YouTube