Dr John Mandrola


March 09, 2013

Everybody knows the most important aspect of treating atrial fibrillation is stroke prevention. So far, all we have for this important treatment goal is drugs.

Warfarin and now the novel anticoagulants prevent stroke but expose the patient to the risk of bleeding. It's a tough trade-off. It pits an immediate and ongoing risk (bleeding) for future benefit (stroke prevention). Let's call these systemic solutions.

Electrophysiologists like targeted strategies. It's what we do: we like to find the focus and take it out. This is the thinking underlying elimination of the left atrial appendage (LAA), perhaps our most lethal attachment. Surgeons can ligate or staple the appendage at the time of surgery, but this isn't going to work in the legions of people with AF. It's too invasive.

Enter the Watchman device. Delivered percutaneously in an electrophysiology lab, the device holds the promise of occluding the left atrial appendage and preventing embolic complications without the need for systemic anticoagulation. It's a great theory, and all involved (doctors and patients) would love the idea of not using drugs that increase bleeding risk. Bleeding causes fear; that's a truth.

The early studies with Watchman were modestly promising. In 2009, the PROTECT-AF trial, published in the Lancet, showed that the device was noninferior to warfarin for stroke prevention, but the net clinical benefit was limited by higher procedural complication rates. And the complications were serious: stroke due to air embolism, cardiac perforation, tamponade, and device embolization. A follow-up registry study, Continued Access (CAP), showed that experience matters. When researchers kicked the lower-enrolling centers out of the trial, acute complication rates got better. The technology looked better.

In 2010, an FDA advisory panel narrowly recommended approval (7 to 5), but the final decision awaited more studies.

The PREVAIL trial, which was (supposed) to be presented here at the American College of Cardiology 2013 Scientific Sessions , was designed to give more data and offer additional safety data. The slides were made available, but due to an embargo break by Boston Scientific, the maker of Watchman, the trial has been pulled from the program. Investigators stand by the data, and the company says it was an honest breach of protocol.

Sue Hughes from heartwire has the details of the slide presentation. Similar in design to PROTECT-AF, 407 patients were randomized to either the device or warfarin. The trial met two of three primary end points. On safety, Watchman met the seven-day safety end point, with only six of 269 patients suffering a major complication (no deaths). Noninferiority criteria were also established in a comparison of stroke or embolism after seven days. At 18 months, however, there was no difference in the composite end point of stroke, embolism, or death from cardiac causes. Dr David Holmes, the principal investigator of PREVAIL, noted that 18-month data are limited by the small number of patients that have reached that level of follow-up.

My take:

My decidedly negative view of LAA occlusion as a stroke-prevention strategy has not been changed. Given the hoopla surrounding PREVAIL and the small numbers of patients that have completed follow-up, it's hard to say the technology is dead. But it's on life support.

Think about it. Here are eight reasons why I am doubtful about Watchman.

  1. The entire notion of targeting LAA occlusion for stroke prevention in patients with AF is flawed. Stroke is a systemic disease, with both physiologic and anatomic considerations other than just the left atrial appendage.

  2. The human left atrial appendage is a complex anatomical structure that does not lend itself well to closure. Even experienced Watchman operators report persistent leaks in more than a third of their patients. This led the Texas researchers to conclude:

    Given the extent of comorbidities in some patients with AF and the possible presence of a systemic coagulation and platelet dysfunction, an approach that targets only LAA occlusion may not be warranted.

  3. The device is supposed to prevent ischemic stroke, yet the main benefit in PROTECT-AF was driven by lower hemorrhagic strokes. Perhaps there is something to the flawed-strategy thesis?

  4. The procedure is invasive and highly dependent on the operator's skill. Acute complications can be severe. Although the data show the complication rate has come down, 2.2% is still significant.

  5. The most appealing patients for this technology, those who are at high risk for anticoagulation, were excluded from the trial. In PROTECT-AF, nearly 30% of the patients with a Watchman had a CHADS2 score of 1 and therefore would have been eligible for aspirin alone.

  6. The efficacy statistics are not strong. Confidence intervals are wide. We just don't know. When noninferiority measures are used in highly selected patients and patients are operated on in highly selected centers, it is tough to know what will happen in the real world.

  7. Good alternatives for stroke prevention are available. If you are going to cite science, you have to point to the more than 50 000 patients studied in the trials of the novel anticoagulants. Dabigatran, rivaroxaban, and apixaban have all compared favorably to warfarin in a higher-risk cohort of patients with AF.

  8. The device is a permanent indwelling device lodged in the human heart. That's significant. Once endothelialized, it isn't coming out easily. As tools go, placing a device in the left atrial appendage is a big hammer.

I am keeping an open mind on this technology. But for now, the Watchman clearly does not pass the most important test: I would not have it placed in me. Once your atria fibrillates, you think differently about treatments.

I hate to say this, but it looks like pills are the best we have to prevent stroke in susceptible patients with AF.



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.
Post as: