The Role of Percutaneous LAA Closure in Stroke Risk Management for Patients With AF

Ashish A. Bhimani, MD, FHRS; Sandeep K. Goyal, MD, FHRS


November 15, 2022

This transcript has been edited for clarity.

Ashish A. Bhimani, MD, FHRS: Hello. My name is Ashish Bhimani. I'm one of the electrophysiologists at the Piedmont Heart Institute in Atlanta. I'm joined here today by one of my partners, Dr Sandeep Goyal. I'll let him introduce himself.

Sandeep K. Goyal, MD, FHRS: Hello. I'm Sandeep Goyal. I'm a cardiac electrophysiologist and medical director of the electrophysiology labs at Piedmont Atlanta Hospital. Dr Bhimani and I will be discussing percutaneous left atrial appendage (LAA) closure.

Bhimani: Thank you for joining me, Sandeep. I thought we could start off by talking a little bit about the patients whom we send for LAA closure.

Everyone's pretty well aware that when the Watchman device first came onto the scene, there were a couple of big studies, PROTECT-AF and PREVAIL. The outcomes weren't quite as clear as one way or the other. I think some people took it as optimistic that there was some efficacy, but others felt that it didn't really meet the technical noninferiority that they started with.

Initially, many of us thought it was certainly a good option, particularly for patients who can't be on a blood thinner — maybe not quite as good as a blood thinner, but better than nothing. That's really where we started — which patients should we send to get evaluated for this. What is your approach in thinking about who's a good candidate for LAA closure?

Goyal: I think you're absolutely right that the initial two big studies, PROTECT-AF and PREVAIL, gave us good data, but they left a cloud of uncertainty in terms of proving a definite noninferiority over warfarin.

However, given that those studies included patients who were able to take oral anticoagulation and showed that there is a good amount of stroke risk reduction in patients with LAA closure, and some of the difficulties were related to early procedural complications, it definitely gave us confidence early on that this would be a good alternative to consider for patients who are not good candidates for long-term oral anticoagulation.

Both the US Food and Drug Administration (FDA) and the Centers for Medicare & Medicaid Services (CMS) had similar thoughts, and that's why it was approved in the US for that patient population. We really looked for very good reasons that the patients are not able to tolerate long-term oral anticoagulation. However, I think that had changed quite a bit over time in terms of improved safety profile of the devices and the implant. We are not at a point where we would consider this to be a primary therapy replacing oral anticoagulation.

Certainly, I consider it much more in my practice for patients who are having any significant problems tolerating oral anticoagulation, where I expect them to have either recurrent bleeding or recurrent falls — things that are not going to be one time, but there is a chance that those problems will continue to recur — that will prevent them from being on long-term oral anticoagulation and are established good candidates for this therapy.

Bhimani: You bring up a good point. The studies initially were comparing warfarin against a first-generation initial device. Obviously, neither one of those is being used primarily in stroke risk reduction for our atrial fibrillation patients, so we're going to need more clinical trial data to really highlight who the best patients are.

One of the things that I've seen is that as the technology has improved, we've got newer devices and second-generation devices. Can you tell us a little bit about how they improve the safety and how you incorporate that into your decision-making?

Goyal: Even in my own practice, the cadence of LAA closure had significantly increased since the second-generation Watchman, also known as Watchman FLX, came out. There were, I think, several design improvements from the first-generation Watchman device to the Watchman FLX device.

A couple of the most important things, from a safety standpoint, were that the Watchman FLX device is closed from the front end. The original Watchman device was open, so you had sharp pieces that you led with inside the appendage. You had to place your delivery sheath deep inside the appendage, which was part of the reason I think there was higher risk for pericardial effusion.

In the Watchman FLX device, it's basically a closed device at the front. You could lead with a soft delivery method — what they call a ball — and implant the device very safely. It's essentially eliminated direct device-related pericardial effusion. The risk is not zero, but it's pretty close.

The second big change they made was making the device shorter. It was able to be accommodated by more LAA than the initial device was.

The third big thing would be the Watchman device has a screw at the back, which is how it's attached to its delivery system. They were able to cover much of that screw with the fabric. The reason that is important is because that is a potential site for thrombus formation. If you are able to reduce the surface area of a metallic component, then I think it certainly helps. I think that was a big advance that really made the procedure significantly safer.

It was proven in the PINNACLE FLX trial, which was a nonrandomized trial, but it had over 400 patients. The overall major complication rate was less than 0.5%, and the success rate for the device was over 98%, which speaks in terms of the development of the technology itself.

In the past year, we now have the Amulet device available, which is an alternative that has been available in Europe for several years but we haven't had access to. That has really added another layer of precision because now you are able to pick and choose devices on the basis of the individual patient characteristics and anatomies.

The Amulet device was compared with the first-generation Watchman device because when the trial started, that was the available device that we had. It's shown a very comparable efficacy and safety to the original device, but we don't have head-to-head data between Amulet and the Watchman FLX device.

Bhimani: I know that because of the points you made about the initial Watchman and which type of anatomy it would fit well, there was a question about what kind of preprocedural imaging [was needed] or how to assess the LAA before we move forward. These newer devices sound like they give us a little more flexibility to cover most appendages.

Tell me about pre-procedural imaging. Is that something you still think is warranted for most patients? Do you think that we're at the point where we can close almost all appendages, and that we can just go straight forward with the procedure?

Goyal: I think we have significantly improved our ability to perform LAA closure in the majority of patients. We only have rare patients that would not be able to be closed with the percutaneous device now. When only first-generation Watchman was available, I used pre-procedure imaging to decide which patients I would bring to the lab for closure and which I would not.

Now, I still personally try to get as much pre-procedural imaging as possible and really try to get CT, because that provides you a three-dimensional view of the appendage.

The reason for that is not as much whether we are going to be able to close this particular anatomy. It's true that, occasionally, it does tell us that there is going to be no suitable device, but more importantly, it allows us to assess the characteristics of the LAA and makes the implant itself much easier. You are able to obtain accurate sizing so that you know what size device you are going to use.

There are programs that are available where you can do a virtual implant of either the Watchman or the Amulet device and can get a large amount of information that allows you to have a much better idea of how the device is going to interact with the LAA and makes the procedure much more efficient and safer.

I think that's really the predominant use of the pre-procedural imaging today. When we started, it was more to decide yes or no.

Bhimani: Let's keep on that theme of the imaging and go into the procedure room itself. I know initially, one of the challenges was just, from an operational standpoint, having the implanter for the Watchman, and also we typically would have a second physician helping with the transesophageal echocardiography (TEE) to help guide the placement.

I understand that intracardiac echocardiography (ICE) is being used a little more. Can you talk a little bit about using ICE vs TEE and some tradeoffs that we might face with that?

Goyal: That has been a big change, at least in my practice. Nationally, ICE is being used more and more often for all kinds of structural heart and electrophysiologic interventions. LAA closure is no different. My personal practice is probably 90%-plus now with ICE, where 2 years ago, it was probably 5% with ICE.

TEE has many unique advantages. The biggest of those is that all of the trials have been done with TEE, so all the criteria we use are developed off the TEE and it's something that we have had long experience with. It's relatively cheap and easily available in most labs, and it could provide 3D imaging, so it's a very well-proven commodity.

However, one of the big challenges is that many of the patients requiring LAA closure, especially in the recent past, are older patients who are relatively frailer. If we could avoid general anesthesia, which is needed every time you use TEE for a procedural intervention, then that is really a big help.

That's where the initial use of ICE started, because we are able to do this under moderate sedation. It really reduces the recovery time for patients, and overall patient experience is significantly better.

I think there are a number of patients who have contraindications to the TEE because of esophageal varices, strictures, or things like that. For those patients, ICE was also gaining more utility. Then we have patients who are at more risk related to TEE. Fully anticoagulated patients tend to have a few more bleeding events. Any time you intervene in a different part of the body, oral injuries and esophageal injury are other risks that remain present.

One thing I have learned and realize in doing many of the transseptal punctures, both under ICE and TEE guidance, is that the inferior part of the septum is much better visualized in ICE. That is usually our target area for these procedures, so it actually provides better imaging.

Then, as you alluded to at the start, the workflow is a huge issue. Normally, you need three physicians, including an anesthesiologist, an implanter, and an echocardiographer. Use of ICE reduces it to two, so that allows better scheduling. All of these reasons have allowed ICE imaging to gain prominence.

In terms of the data, we have a significant amount going as far back as 2007 with small series. Much of the data have been single- or dual-center, small series data, but it has been shown to have very comparable safety and efficacy to TEE.

There are two prospective studies that are looking at it — one is ICELAA and another one is ICE Watchman — that would also provide more robust data on the safety and efficacy of the ICE-guided approach. I think that is really the future of imaging for LAA closure.

The 4D ICE is also coming onto the scene, which is really going to help improve the imaging, provide biplanar imaging, and make it easier for one physician to operate both the ICE and the device itself. It is really going to come down to reimbursement and whether the cost of the 4D ICE would be affordable for routine use.

I personally feel that if the technology is proven, then the cost constraints get figured out eventually, the price of the device will go down, and reimbursement will keep up with it if the technology is right for the patient. I'm a big proponent of using ICE for both patient and operational reasons.

Bhimani: Obviously, most electrophysiologists and interventional cardiologists are familiar with using ICE. Was the learning curve to switch from TEE to ICE for the procedures a long one, or do you think most people can pick that up pretty quickly?

Goyal: I think if you use ICE for other things, then it does make it a little bit easier. As electrophysiologists, many of us will do the right atrial– and right ventricular–based imaging from the ICE. Some of us, routinely, also examine the left atrium, the pulmonary veins, and the LAA from the left atrium with the ICE as just part of better understanding the anatomy for atrial fibrillation ablations.

I think that really helped me quite a bit when I started doing them with ICE. I initially used, for several cases, 3D mapping to best understand what the views are that we are going to get. For most electrophysiologists, it will probably not take more than eight to 10 implants to become proficient. You could even do co-implants, where you could have both TEE and ICE together.

On the interventional side, it more depends on how much experience they have. I think people who have experience would probably take the similar number of cases. If you don't have much ICE experience, then it may take 20-25 cases to get really proficient at ICE imaging.

Sometimes we are comparing TEE and ICE as the same technologies. Other than both being echocardiography, they are different technologies and you are never going to get the same views. You have to learn that the views from the ICE are going to be slightly different from the TEE, and that takes a little bit of getting used to.

Bhimani: I think that makes sense. Obviously, with the new devices and the change in how we can use imaging, I think the procedure is becoming safer, more operationally efficient, and maybe lower risk on some of the avenues of how we do the procedure.

Let's get back to the clinical arena a little bit. One of the biggest challenges is that we have a device that's great for patients who can't be on anticoagulation; yet, when they get the device, we requested they be on anticoagulation for typically at least that 6-week period while we're waiting to verify that there's no leak. Are there new data starting to show that maybe we can use different paradigms for anticoagulation or antiplatelet therapies to minimize risk in that short post-op period?

Goyal: I think that's a very critical point. That's something we all struggle with regarding how to get these patients safely through that period where anticoagulation is currently indicated. It's also very timely because the Amulet device was studied with dual antiplatelet therapy. We really don't need any period of oral anticoagulation after the Amulet device.

It's actually discouraged to use oral anticoagulation because there was a higher rate of pericardial effusions if you did get oral anticoagulation after the Amulet device implant, especially in the earlier period. The device led to thrombus risk that was actually lower than that with the Watchman device, even with dual antiplatelet therapy.

Many of us have been using dual antiplatelet therapy, even with the Watchman device before Amulet was available, for patients whom we did not think could tolerate oral anticoagulation or had very early bleeding on oral anticoagulation. On the basis of those data, the FDA, I think only a couple of weeks ago, has approved the Watchman FLX device to be used with dual antiplatelet therapy.

Now, this is not based on randomized trial data, but it's certainly a much larger number of patients from the American College of Cardiology Foundation's National Cardiovascular Data Registry (NCDR) Left Atrial Appendage Occlusion (LAAO) Registry that are available. The outcomes have been quite good. I think we tend to use too much oral anticoagulation and antiplatelets in these patients because that's what was used in the clinical trials.

The NCDR Registry also clearly showed that if you had patients on both aspirin and oral anticoagulation in those first 6 weeks, they had a pretty high risk for bleeding. If you took away the aspirin and just did oral anticoagulation, the bleeding risk dropped substantially without any change in stroke or device-related thrombus risk.

I think the real-world data are telling us that we need to pull back on the amount of anticoagulation and antiplatelets we are using in the early post-op period.

Bhimani: Especially for the ones who are at the highest risk for bleeding, which tend to be the ones who have the most to gain from having this procedure done.

I think that, from pre-op to inter-op to post-op, some of the new changes that are evolving are going to potentially make this procedure more accessible or more appropriate for a wider variety of patients. That's where the clinical data and the clinical trials are going to help guide us.

In terms of the device space, are there things that are coming down the pipeline or things that industry can help us with to try to evolve the devices for the next generation and what we might be looking for?

Goyal: The technology has evolved quite a bit. There are some Achilles heels that we are still dealing with. One is that there are a number of patients who still have leaks post-implant. If you have a large enough leak, then potentially, the benefit of the device goes down.

Again, one of the NCDR registry studies has shown that the leaks are not good after the device implant. Working on both the device design and procedural techniques to reduce the risk for leaks is really critical.

I think there has been much improved understanding, especially based on the postoperative CT imaging, as to which anatomies are more prone to leak and which are less prone to leak, and how to avoid those leaks. Getting much better preprocedural planning and using artificial intelligence–based platforms to guide what size and which device to use would be really, really helpful.

I think the second most feared problem is device-related thrombus. That's why we use oral anticoagulation early on. I think that risk is significantly reduced after the early period, but it's not eliminated. There are reports of late device-related related thrombus also, and there has been a large amount of work done in the device design space as to how to reduce that.

There is a device, Conformal, which is in clinical trials. They are using a thread tether rather than an actual metallic tether, and foam is covering the screw much more. The idea behind that is that this will help eliminate or reduce the risk for device-related thrombus even further. Obviously, that is an unproven claim at this point. We will see if clinical trial data support that.

It's going to be critically important for us to work on those two things. The third one would be if we find a device that would eliminate the need for anything more than a single antiplatelet agent, then that would allow our highest-risk patients, who probably benefit most from the device, to be able to safely have this implant done.

I think that if we can solve those three problems, that could be close to the holy grail. That's my wish list for what should be happening in the next 5-10 years.

Bhimani: Thank you. We all know that reducing stroke risk for our patients with atrial fibrillation is definitely one of the biggest challenges we face. it's exciting to see this space evolve, and hopefully, we'll see continued evolution, which ultimately leads to improve patient safety and outcomes.

Dr Goyal, thank you so much for joining me and being part of this. Thank you, Medscape, for the opportunity to have this conversation with everybody.

Goyal: Thank you. I really enjoyed the discussion.

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.