The skeptical cardiologist first mentioned the Watchman left atrial appendage occlusion (LAAO) device in a cynical post on advertising for medical devices and drugs that I experienced at the American College of Cardiology meeting in 2018.
I thought it would be ironic to get my picture taken inside the heart grasping the catheter and plugging up that pesky appendage.
At that time I wrote, “The Watchman device I was grasping has been approved for preventing stroke in patients with atrial fibrillation who are at high risk and can’t, or won’t, take blood thinners. Boston Scientific has been flying cardiologists to various cities for the last year to wine and dine them and fill them full of reasons to send their patients for the device. “
I also agreed with John Mandrola’s early skeptical take on Watchman here. John updated his arguments against LAAO in a Substack post (subtitled “percutaneous closure of the left atrial appendage may become one of cardiology’s biggest errors”) in 2021 (here) and provided a link to his excellent lecture on the topic.
Beware The Leaky Watchman
Unfortunately, the device is being implanted willy-nilly in patients with atrial fibrillation and patients need to understand it is not the panacea (eliminate stroke risk without having to take a blood thinner!) they are being promised.
On his most recent This Week in Cardiology, Dr. Mandrola does a great job of summarizing some of the issues surrounding LAAO and discussing a new paper that should raise further concerns. If you are interested in critical analysis of current cardiology studies and practice I can’t recommend TWIC highly enough. I tell every trainee who gets within a few feet of me on hospital rounds that listening if they can spend 20 minutes listening to TWIC that time will be the most densely educational part of their week.
I’m going to alternate text from the Medscape transcript of John’s podcast with my own observations.
Last week, authors from Athens Greece, first author, Athanasios Samaras, published meta-analysis of 48 studies (including randomized controlled trials [RCTs] and observational studies) looking at the incidence and prognosis of residual leaks following percutaneous LAAO.
Some background: The idea of LAAO is that if you completely occlude the appendage, clots cannot form there, and you reduce a substantial nidus for cerebral embolism and stroke. The added benefit, proponents say, is that after occlusion, the device endothelizes and you can stop oral anticoagulation, thereby reducing bleeding risk.
I have to admit, it’s a great story, and I am sure it’s easy to explain to patients.
It is remarkably easy to convince patients to get the device if you don’t share with them inconvenient truths about the lack of efficacy and the possibility of device-related thrombus. But typically patients don’t get all the information they need to make an informed decision before signing up for having the foreign body inserted inside their heart.
Dr M goes on to describe the problem of inadequate closure in the LAAO world:
One major problem is that our LAAs come in lots of different shapes and sizes, and it’s hard to get complete closure. What’s more, and this is a whopper of a barrier, if the device leaves a little leak, blood can more easily pool and clot. Surgical data has shown that incomplete LAA closure carries a serious risk of stroke.
Many studies have documented peri-device leaks (PDL) after percutaneous closure. The nice thing about the current paper is that it brings together and sums up the incidence and prognosis of these leaks.
It seems like weekly that I come across an image on X of a device-related thrombus case like this one:
All but four of the 48 included studies were observational. Total patients studied was nearly 62,000.
The main finding was the most sobering: PDL by transesophageal echocardiography (TEE) were found in 26% of patients. One in four. Wow.
Two notes on this: a) This incidence comes from centers that actually report things; you might expect that these centers are better at implantation than lower volume centers that don’t report data. b) This incidence is by the standard TEE.
The authors also report incidence of PDL by CT scanning — a more sensitive technique. Here, as I have reported recently, PDL were noted in 57% of patients. I find this quite sobering, but proponents will argue that prognostic data is based on TEE, and we don’t know what finding leaks by CT means.
The PDL ranged from 1 mm to > 5 mm and the larger the leak the greater the chance of experiencing thromboembolism. A PDL of 3 mm was found in 10% of the patients and was associated with a fourfold increase in stroke or systemic embolism.
Comparison to Surgical Obliteration of the LAA
“the word occlusion is in the name of LAAO. It’s not LAApO for partial occlusion. I can’t help but think this discovery, nearly 10 years after the seminal trials, is one of the explanations for higher ischemic stroke rates in the seminal trials, PROTECT and PREVAIL.
Of course, the other reason was the decision to stop oral anticoagulation (OAC) after occlusion. Recall that one of the best studies in the entire cardiology space, LAAOS 3, involved surgical LAAO at the time of cardiac surgery. Surgical LAAO reduced stroke significantly compared with no occlusion, but OAC was not discontinued.
Why did LAAOS 3 authors not stop OAC? Because atrial fibrillation (AF) is a systemic disease, and the appendage is only one source of stroke. Also, the entire evidence base of stroke and AF have shown the necessity of OAC to reduce AF-related stroke.
I wrote about LAAOS3 here. That post includes a TEE video of an LAA clot. Back then I bemoaned the lack of evidence preceding LAAOS3: “as one who visualizes the LAA routinely by transesophageal echocardiography I see how important the pump function of this appendage is to overall atrial function and how its pump function is essentially eliminated in postoperative patients. What damage was removing or obliterating the LAA doing to overall cardiac function? I even created a website with the URL “I am Joe’s Left Atrial Appendage”, attempting to drum up sympathy for this abused portion of the heart.”
Even Biased Authors Acknowledge Many Challenges with LAAO
I remain concerned about Joe’s left atrial appendage and, like Dr. M, wish we had good RCT data that deploying a nitinol frame structure with fixation anchors and a permeable polyester (PET) fabric cover is a good long-term solution to afib-related stroke:
I remain totally pessimistic about this preventive approach to stroke reduction, and I am truly saddened by the fact that our field embraced this technique without proper studies. Hundreds of thousands of procedures have been done, and if only a fraction of these patients had been enrolled in RCTs, we would know if this procedure worked.
A recent extensive review of LAAO notes all the issues John outlined in 2017: lack of efficacy data (“A key issue with these data is the lack of convincing evidence that supports the mechanism of action of LAAO; reducing cardiac thromboembolism due to exclusion of the LAA cavity from systemic circulation. Indeed, ischemic events were higher in the device arm in PROTECT AF and PREVAIL”), the high prevalence of device-related thrombus and device leaks. Despite these huge issues (“remaining challenges”) there was no call to put a halt on implantation.
The FDA approved LAAO with the Watchman 2.5 (Boston Scientific) device in 2015. In 2021, a second LAAO device (Amulet, Abbott) was approved by the FDA based on its noninferiority to the Watchman 2.5 device. There are 7 authors on the JACC review on LAAO and 6 of them are financially connected to Boston Scientific and/or Abbott, the LAAO makers.
The medical approach to stroke prevention in atrial fibrillation with newer oral anticoagulants (Eliquis) has been proven superior to warfarin. Before patients sign up for LAAO, they need to understand all the potential downsides of device implantation and that it does not reduce ischemic strokes compared to warfarin. They should strongly consider a second opinion from a cardiologist with no skin in the implantation game.