This transcript has been edited for clarity.
John M. Mandrola, MD: Hi, everyone. This is John Mandrola from theheart.org | Medscape Cardiology. I'm joined today by cardiac surgeon Professor Richard Whitlock, from McMaster University in Hamilton, Ontario, to discuss his late-breaking presentation of the LAAOS III trial at the 2021 meeting of the American College of Cardiology. LAAOS III was a clinical trial of add-on surgical left atrial appendage closure in patients with atrial fibrillation (AF) who were having heart surgery for other reasons, such as coronary disease or valvular heart disease. This was a Canadian-led global trial, measuring stroke and systemic embolism as a primary outcome. Welcome, Richard.
Richard Whitlock, MD, PhD: Thank you, John. LAAOS III answers a persistent question within surgery and cardiology: Does left atrial appendage occlusion prevent ischemic stroke? And we showed that surgical left atrial appendage occlusion does prevent stroke; it reduced the risk for stroke by 33% from the time of surgery. If you exclude the perioperative period when patients have strokes for other reasons, such as clamping the aorta and cutting out calcified valves, the effect is actually 42% from day 30 on. So this was a definitive trial.
LAAOS Plus (Not Sans) Anticoagulation
Mandrola: You had a specific question in mind. Tell us more about these 4800 patients. What kind of background therapy were they on?
Whitlock: LAAOS III was a multinational effort involving 27 countries and 105 sites. The patients were undergoing cardiac surgery for other indications besides AF. They had to have AF and a CHA2DS2VASc stroke risk score of at least 2. Now, this wasn't a how-to-do-a-procedure trial or a standalone procedure like percutaneous devices. Patients had already been exposed to some risk by having the cardiac surgery; these patients all should be on baseline and ongoing oral anticoagulation. We wanted to explore whether left atrial appendage occlusion provided an additive benefit on top of usual care that includes oral anticoagulation. And indeed, we demonstrated that this is the case.
Mandrola: Rich, tell us about some of the trial characteristics. There was blinding and adjudication of outcomes. There was intention to treat and crossovers. Give us a rundown.
Whitlock: That's correct. Because this was an additive procedure, we came up with a fairly unique way of maintaining blinding. An email indicating the allocation was sent only to the surgeon participating in the surgery. It said, "Please occlude the appendage" or "Please do not occlude the appendage." The surgeons followed that 93% of the time, so there were about 7% crossovers. It was slightly different between the groups because of issues such as adhesions, thrombus that was found in the left atrial appendage — issues like that. The surgeons who performed the procedures were asked not to indicate in the operative report whether they occluded the appendage or not, but just to state that the patient was participating in this trial and the appendage may or may not have been occluded.
The ongoing antithrombotic management was continued, independent of the knowledge of allocation. Surgeons follow our patients for the short term. The ongoing management of these medications is usually carried out by cardiologists or primary care physicians, and they were not aware, nor was the patient or any other healthcare professional involved outside the operating room, aware of the patient's surgical treatment. I believe it was a rigorous study. Also, stroke was a hard endpoint. This is not a surrogate; these strokes were substantial.
Mandrola: A lot of clinical trials enroll relatively ideal patients. It's a pretty special environment, and it makes it tough to generalize, but my take on LAAOS III was that you enrolled a fairly typical cardiac surgery population.
Whitlock: Yes. The mean age of this group was 71 years, the mean CHA2DS2VASc score was 4.2. So indeed they were quite high-risk patients. Two thirds of these patients underwent some type of valvular procedure, which makes LAAOS III unique in that a lot of the trials in the past have excluded the valvular patients with AF. About one third were female. It would have been great to have had more female representation, but looking at our local screening logs, the number was consistent with what we were seeing in patterns of practice. I would say that the LAAOS III population was well representative of the population undergoing cardiac surgery in contemporary practice.
Mandrola: The P value is quite low and the confidence interval was wide, but the low end was a 15% reduction. Naysayers may say, look, it was only a 2.2% absolute risk reduction at 3.8 years. How would you answer that?
Whitlock: Again, this is a 33% risk reduction at 3.8 years. Or should it be hazard reduction? Stats aren't my strong suit. These AF strokes are large. That's the first point. The second point is, if you look at the Kaplan-Meier curves, they continue to diverge. The number needed to treat at 5 years is around 37. If you're looking out to 10 years, it will probably be down into the teens, so there's ongoing benefit. This is a one-time therapy; there aren't issues of compliance or adherence. It's a gift that keeps on giving.
Surgical Technique and TEE Critical
Mandrola: I completely agree with you. I want to pivot to the technique. We'll talk about percutaneous closure later, but I want you to talk to other surgeons now. What technique did you use for closing the appendage in LAAOS III and why is it important?
Whitlock: LAAOS III was one of three studies. It's called LAAOS III because there were LAAOS I and LAAOS II. In those trials, we refined what we should permit in terms of closure techniques as well as other design features. In LAAOS I, we learned that a purse string around the base of the left atrial appendage or a simple whipstitch on the outside did not work well. We had a 50% failure rate. In LAAOS II, we demonstrated that stapler occlusion and amputation enclosure were very effective. By our definition of stumps less than 1 cm and no leaks across the line, we had almost 100% success in LAAOS II. You're always concerned about leaks and you're concerned about stump. If you amputate this appendage down to smooth atrium and close it with a 3.0 or 4.0 Prolene double layer, you're almost guaranteed not to leave a stump. And if you get a leak, you will know about it because you're not leaving the operating room when the patient is bleeding.
Now, regarding the double-layer linear closure from within the left atrium. If you're taking full-thickness bites that go through endocardium out epicardium and back in, it may work well, but I don't think it's as good as this technique. About 55% of the patients had the amputation and closure, and the other patients had a combination of closure with a stapler or epicardial closure device, the atrial clip. These were split between 10% and 15% each.
Mandrola: You mentioned that transesophageal echocardiography (TEE) was encouraged. Was this necessary?
Whitlock: To do this safely, TEE is necessary. These patients come in with AF and you've got them off their antithrombotic. It takes hours or a day to form a clot. So if you're taking off the left atrial appendage, you need TEE for several reasons. One is to make sure there's no clot. That helps you decide which technique you're going to use. If you use an epicardial closure and there is a big thrombus, you may be in trouble; you'll need to open things up and take it out. Second, the TEE will confirm closure. Our definition of closure was a stump less than 1 cm and no need to cross any lines, if you happened to be using a stapler or clip. That definition is somewhat arbitrary, but the fact that we used that as the definition in LAAOS III and we've got a successful and positive trial supports it. So, TEE is used to rule out thrombus and confirm closure.
Percutaneous LAAO Devices
Mandrola: I want to close with discussion about percutaneous closure. I know that a positive trial of left atrial appendage occlusion will create even more tailwinds for the endocardial closure business, but I see significant differences between what you did and percutaneous closure. First, surgical closure was in addition to anticoagulation; that's a different strategy from percutaneous closure. Second, I think surgical closure, being epicardial, is different. Third, surgical closure doesn't leave a foreign body exposed to the blood pool in the left atrium. Can you speak to that?
Whitlock: I agree. These results pertain to surgical left atrial appendage occlusion that's concomitant to other cardiac surgery. It is good news. It confirms that the left atrial appendage is a substantial source of stroke in these patients and that managing it has benefits. But how it's managed is important. Moving forward, we need to conduct a trial with percutaneous devices that explores this as additive to oral anticoagulation in higher-risk patients.
If you have a population that still has a more than 2% per year risk of stroke on direct oral anticoagulation, according to the RE-LY data or 1.6% in the AVERROES trial, and you can give these patients the additive benefits of occlusion with a percutaneous device, then that's something that needs to be tested. But so far we do not have that data. That can't be extrapolated. It is hypothesis-generating for them. We need to do that trial.
Mandrola: What about patients with AF who take anticoagulants but don't always stay on them? Also, a fairly significant number of patients have bleeding issues with anticoagulants, and it would be wonderful if we could close the appendage and discontinue anticoagulation, like the strategy with percutaneous closure, which is dubious. Should we have a trial of removal of anticoagulants in patients who have undergone surgical closure of the left atrial appendage?
Whitlock: My view would be, let's first settle the question of whether there is additive benefit, because the principle of treating patients with AF is stroke prevention, stroke prevention, stroke prevention. And if we can have an additive effect, that's wonderful. At the same time, you could certainly evaluate a subset of patients who come off oral anticoagulation. What are the effects in that subset? Obviously, it's a weaker trial because you lose some of that randomization balance.
I still maintain that if a patient can be sustained on an oral anticoagulant and they do well, then that's what's best for the patient. I mean, major bleeds are a problem and are linked to morbidity and mortality, but there are certainly a large portion of patients who tolerate their long-term anticoagulants. If they have to come off anticoagulants because as they age they become frail and are prone to falls or other bleeding conditions, then they come off and it's highly likely that they will still maintain some benefit. But that needs to be proven.
Mandrola: I know that the LAAOS III trial was a huge undertaking, and I congratulate you and thank you for adding to the evidence base of this important issue.
Whitlock: Thank you. Obviously, I'm speaking on behalf of a huge number of co-investigators, so congratulations to all of them because I think we have shifted outcomes for patients with this condition.
Follow theheart.org | Medscape Cardiology on Twitter
© 2021 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: John M. Mandrola, Richard Whitlock. LAAOS III Lessons: Surgical Technique and Anticoagulation Key to Cut Stroke in AF - Medscape - May 28, 2021.