FDA Advisory Panel Votes 7 to 5 to Recommend Approval of Watchman LAA Closure Device

April 23, 2009

(UPDATED April 24, 2009) April 23, 2009 (Gaithersburg, Maryland) — A US Food and Drug Administration advisory panel voted 7 to 5 in favor of approving a device for closure of the left atrial appendage (LAA) that they say is comparable to long-term warfarin therapy for the prevention of stroke in warfarin-eligible patients with nonvalvular atrial fibrillation (AF).

The vote to recommend approval came with conditions, including that implantation be performed in centers with surgical backup and the creation of a physician certification program. The panel also recommended the creation of a registry and extended follow-up of current clinical trials.

"There certainly is some doubt about the data because it is a small study, but it's larger than what we see with other devices," said Circulatory Systems Devices Panel member Dr John Somberg (Rush University Medical Center, Chicago, IL). "We're in as strong a position as I've seen with many other devices that have even broader applications."

I voted yes, a tepid yes.

"I voted yes, a tepid yes," said neurologist Dr Gary Abrams (University of California, San Francisco). "I'm not convinced of the noninferiority [of the device], but I sit on the fence with that. I was swayed by the safety issues. . . . I think that once the early morbidity from this gets worked out as people get experience with it, I think it offers an option for people who need to stay on warfarin for many, many years."

FDA Watching the Watchman

Throughout the day, the panel debated and discussed the various analyses of the Embolic Protection in Patients with Atrial Fibrillation (PROTECT-AF) trial, the pivotal premarket approval study, and focused specifically on the safety and effectiveness of the device, known commercially as the Watchman (Atritech, Plymouth, MN). The Watchman is a percutaneously implantable fabric-covered expandable nitinol cage that is placed distally to the ostium of the LAA to occlude flow and prevent the migration of thrombus forming in the appendage.

As previously reported by heartwire --results of PROTECT-AF study were presented last month at the i2 Summit of the American College of Cardiology 2009 Scientific Sessions--the Watchman device was associated with a reduction in hemorrhagic stroke risk vs warfarin, and all-cause stroke and all-cause mortality outcomes were noninferior to warfarin.

It muddies up the answer . . . and then it's just a guess by almost every panel member.

During a presentation to the panel, Dr Julie Swain, who analyzed the data for the FDA, called PROTECT-AF a "complex study" and pointed out that using a noninferiority hypothesis when comparing a drug with a device is unusual. Given, however, that warfarin is the drug therapy, "it would be wonderful to have a device where you have an upfront cost, and then everything was fine, and you didn't have to give warfarin again," she said. "As physicians we all know the difficulties of giving warfarin."

Dr Norman Kato (Cardiac Care Medical Group, Encino, CA), who voted for approval, was particularly critical of the trial design, saying that interpreting noninferiority analyses makes him "nervous," and questioned whether the design was appropriate in this setting, particularly given the difficulties in adjudicating hemorrhagic stroke events.

"By not having a rigorous trial, we spend a lot of time trying to figure out if the statistics are right and how many different ways we can slice and dice this thing," he said. "It bothers me that we had to compromise on the trial, that it had to be noninferiority, because the times where we've had to deal with noninferiority vs a randomized, prospective trial we always get into this issue. It muddies up the answer we have to provide, and then it's just a guess by almost every panel member."

In her presentation, Swain noted that patients included in the PROTECT-AF study were at low risk for stroke. Roughly one-third of patients had a CHADS2 score of 1, while 67% had a score of 1 or 2. The American College of Cardiology/American Heart Association recommendations state that patients with a CHADS score of 1 can be treated with aspirin rather than warfarin. Moreover, the exclusion/inclusion criteria were extensive, noted Swain, and patients at higher risk of poor outcomes, including those with advanced heart failure, recent stroke or MI, and carotid disease were not studied.

Swain told the advisory panel that the results of PROTECT-AF are confounded by the use of other medications. All patients were taking warfarin to day 45, and, if there was no flow around the Watchman device, were then prescribed clopidogrel (Plavix, Bristol-Myers Squibb) for six months. All patients continued aspirin therapy for the remainder of the trial. Yet, as noted by the FDA, one-third of patients in both arms didn't receive medications according to the study protocol--controlled warfarin in the therapeutic range vs the device and short-term warfarin use. Some patients in the device arm were restarted with warfarin for various reasons, which confounds the intention-to-treat analysis results, said Swain.

Dr Thomas Vassiliades (Emory University School of Medicine, Atlanta, GA), who voted against recommending approval, said that while the statistical analysis showed the device to be equivalent to warfarin therapy, "there were too many confounding variables, such as antiplatelet therapy, to approve." Also, with two-thirds of patients having a CHADS score of 1 or 2, "it gives me concern about using the device in the general population," he said. "Finally, I thought that the data at two or three years were so miniscule and that implanting a device in the left atrium needed longer follow-up for me to feel comfortable approving it."

"I'm Comfortable This Is the Right Thing to Do"

Overall, most panel members felt the sponsor showed the device to be effective, although there were caveats. Many were uncomfortable with the size of the 800-patient study and the duration of follow-up. Others were uncomfortable making a decision about effectiveness with end points such as hemorrhagic stroke, which was included in the primary efficacy end point. They felt a decision on effectiveness was difficult when there were so few hemorrhagic strokes--five in control arm and one with the Watchman device--while others thought ischemic stroke, which occurred more frequently in the device arm, would have been a more reliable end point.

I would think it would be a mistake for the FDA to approve this device.

Regarding safety, there was also a divergence of opinion. Again, panel members were concerned about assessing the long-term safety of the data given the short-term PROTECT-AF study. Dr Michael Domanski (National Heart, Lung, and Blood Institute, Bethesda, MD) said it was hard to look at the data and say the device is safe. As observed in the PROTECT-AF trial, the risk of ischemic stroke was significantly higher in the device arm than in the control arm, and Domanski said that the high rate of hemorrhagic stroke, which was higher in the warfarin control arm, could have been avoided with more attention to patient follow-up rather than by implanting the invasive device.

"I think there were lots of complications," said Domanski. "I don't think they were minor at all. I would think it would be a mistake for the FDA to approve this device."

Still, opinions like this were the minority, and most felt the "devastating" effects of warfarin over time need to be balanced with the increased risks with the device, such as pericardial effusion, device embolization, and thrombus formation on the device. There was agreement among panel members that avoiding warfarin in certain patients was a significantly attractive enough reason to vote for approval.

"The risk of Coumadin is high, especially in an older population who fall or who are more fragile," said Dr Jeffery Brinker (Johns Hopkins Hospital, Baltimore, MD). "I'm comfortable this is the right thing to do, and I anticipate that the information we're asking from the sponsor with postapproval studies will be helpful in providing further encouragement."

In PROTECT-AF, safety events, particularly pericardial effusion, were more common in the device group, but these decreased over time with procedural modifications and enhanced training. Dr Fredric Resnic (Brigham and Women's Hospital, Boston, MA), who also voted for approval, stressed the need for training and certification for implanters, because the results achieved in the clinical trial would not be replicated if adopted by less experienced centers.

The FDA does not have to follow the recommendations of the advisory panel, although it usually does.

Atritech, a privately held company, funded the PROTECT-AF study.


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: