Leaks After Percutaneous Left Atrial Appendage Occlusion and Stroke Risk

Christopher V. DeSimone, MD, PhD; Mohamad Alkhouli, MD


June 23, 2022

Editorial Collaboration

Medscape &

Christopher V. DeSimone, MD, PhD: Hello, and welcome back to the Mayo Clinic–Medscape video series. I'm Dr Christopher DeSimone, cardiologist and director of cardiovascular marketing here at Mayo Clinic.

Today we're going to be discussing leaks after left atrial appendage closure and how this can increase stroke risk. I'm joined by my colleague, Dr Mohamad Alkhouli, professor of interventional cardiology here at Mayo Clinic. Welcome, Dr Alkhouli.

Mohamad Alkhouli, MD: Thank you, Chris. It's a pleasure.

DeSimone: You recently presented a late-breaking study at ACC. Can you tell our viewers the major findings and takeaways from the study?

Small Leaks Matter

Alkhouli: Perfect. Thank you. The study aimed to look to see if leaks that are resulting after left atrial appendage closure are associated with strokes in the long term. We know from surgical literature that if you close the appendage and there is the residual leak, that results in multifold increase in strokes over a long time.

We didn't have enough data to see if that is the case with percutaneous left atrial appendage closure. The only data we had were from the randomized trial where the number of patients wasn't enough to be able to do that comparison at that study.

We leveraged the NCDR database. It's an ACC database that encompasses pretty much all of the atrial appendage closure done in the country, so it allows us that large population of patients to test the hypothesis, which is, do leaks impact long-term stroke risk?

We ended up somehow with a little over 50,000 patients in the study. The major findings were, one, that leaks were not uncommon. About 1 in 4 patients had some sort of a leak after the procedure, detected at the 45-day echo that we typically do.

The second major finding is although the majority of those leaks were small, so by current standards, are less than 5 mm — that's considered small. Those were associated with about 10%-15% increased risk for stroke or TIA at the 1-year mark. These results challenge the current notion that small leaks don't matter.

DeSimone: That's very interesting. You would think that some people are more scared of larger leaks. They might put them on anticoagulation and keep them on or look at ways of how they could plug. Yours is very interesting, and it makes sense because maybe the small leaks promote going in and not coming out and strokes themselves.

Alkhouli: There were very few large leaks in the registry, fewer than 400 patients. They were not associated with events, but those patients remained on blood thinners, so there is that treatment bias.

The small leaks were many, so we're able to discern that difference. Physiologically, the average size of the middle cerebral artery is 3 mm. The 5-mm mark is arbitrary that has been practiced for a while, so that sort of challenges that. In Europe, they already use 3 mm as a cutoff.

Preplanning for Leak Prevention

DeSimone: Very good. For our audience, you mentioned 45 days, but what's your practice and how would our practice maybe change about looking for these leaks more frequently and aggressively?

Alkhouli: Prevention is better than a thousand cures. The challenge would be to first find ways to prevent the leaks. These patients are coming to us to come off blood thinners. If we want to keep them on blood thinners or try to plug them, that becomes too cumbersome for a preventive procedure. We want to aim at preventing leaks to start, aiming at zero leaks.

We have a few tools to do that now. We're not quite there yet. I forgot to mention that this was with the first-generation devices, the Watchman 2.5. We now have a new-generation device that has a better seal and a better design altogether that allows us to seal the appendage better. With the newer devices, the leak rate in other studies has been less, at about 10%-15% at 1 year, so that is already encouraging.

We also have many planning tools that we didn't have before. Now we have pre-CT scans that we could model, and we could actually do a virtual implant and predict whether this patient is going to have enough compression on the device and whether it is going to have leaks or not. I think we have much more planning that we are able to leverage now.

Lastly, there are newer devices. At the time of the study, there was only one device on the market. Now we have two. If you do the pre-CT, you model it and you decide, well, this patient will be better served with this device vs that device. I don't think we are there yet in terms of eliminating the leak, but leveraging all those tools together, I think we could minimize it to a very few percent.

DeSimone: Excellent. It makes sense. You find a problem and then you look for ways to improve the situation or how to better do it, one of these being this revised device. Walk us through how, when you see those leaks, people speak about coiling. Give us a brief overview for our audience.

Alkhouli: It's off label, obviously. We don't have much data about the efficacy. We know that there are data. There are data from Dr Killu here at Mayo Multicenter Registry that looked at the safety of this practice, and it's fairly safe. We don't have long-term efficacy data. I think it's an option, but I don't think it's an option that should be promoted. I think we should focus more on the prevention.

When we see these leaks, the major challenge, in all honesty, is how do we actually classify the leak and how do we measure it? We will look at echo. It's a two-dimensional imaging modality most of the time. When you look at the leak, two people may call it large or small, depending on how they're measuring it. If it's a crescent shape or if it's small, like a focal area, it makes a big difference in measuring those leaks.

I want to emphasize for people who do this procedure and actually interpret the images to take a look at the mechanism of the leak. That is a very important aspect. A 2-mm, small leak above the device near the [warfarin] ridge is very different from a 2- to 3-mm leak that leads to a fully uncovered lobe posteriorly. I think we can move away in the future from sizing to mechanisms and locations, probably.

DeSimone: That make sense. Those probably have their own inherent thrombogenicities and risks of that.

Alkhouli: Absolutely.

DeSimone: It seems like the preplanning is key, and the devices are getting better. Is there anything else coming down the pipeline that you would say is the next big thing?

Alkhouli: Yes. Many of the leaks result from non-coaxiality with the appendage. There is a limited space where you can actually access the left atrium and gain access to the left atrial appendage. Non-coaxiality is a big problem.

The manufacturers of both devices that are available in the US came up with steerable sheets. Those would allow us to be more coaxial and prevent more leaks. There are also niche devices that are being developed to tackle certain anatomies. If you have a double lobe, you might be better served with this device vs that device. I think even the interventional aspect is improving, not only the preplanning.

Intracardiac Echo and TEE

DeSimone: Excellent. I think that will be critical. I do ablations. We're doing the pulmonary veins and the [warfarin] ridge, as you say, and the appendage is right there. The steerability, I think, is going to help you tremendously. One other thing, too, and I know you're a big fan of this, is we use intracardiac echo often.

Do you want to speak, since you're an expert on intracardiac echo and TEE? There are a couple of things. Does this give you a better coaxiality of where you're looking, where you're deploying, and then post-deployment as well as safety of the procedure? We have things coming out. You're a pioneer, and this is 4D echo.

Alkhouli: We already have it. We were fortunate to do the first in-human case last year with a 4D ICE probe and it changed our practice. We're able now to offer patients same-day discharge. We've been doing it with moderate sedation for a while, even with a 2D probe. I think the images are equal or potentially, in my opinion, a little bit superior because you’re right there.

DeSimone: I agree with you.

Alkhouli: You get it.

DeSimone: You're right there.

Alkhouli: EP docs send the ICE probe to the left atrium all the time. I don't know why interventional cardiologists are a little bit worried about it, but it's safe. If you have a 4D probe and you could do X-plane imaging, you get a pretty good picture of the left atrial appendage. I think that is going to be the future.

DeSimone: Thank you, Mohamad, for these very insightful thoughts and your view of the future of the field. Thank you for joining us on the Mayo Clinic–Medscape Cardiology series.

Follow | Medscape Cardiology on Twitter

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


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.