Meta-analysis Shows Watchman Still Noninferior to Warfarin, With Good Safety Outcomes

In patients with nonvalvular A-fib, left atrial appendage (LAA) closure decreases the likelihood of hemorrhagic stroke, cardiovascular death, and nonprocedural bleeding at 2 to 3 years of follow-up compared with warfarin therapy, according to a meta-analysis published in the June 23, 2015, issue of the Journal of the American College of Cardiology. However, closure was linked with more ischemic stroke than warfarin.

Next Steps:Meta-analysis Shows Watchman Still Noninferior to Warfarin, With Good Safety Outcomes“Further studies are needed to define risk thresholds for thromboembolism and bleeding at which patients with [A-fib] benefit from [LAA] occlusion therapy for stroke prevention and to compare the safety and efficacy of this strategy with target-specific oral anticoagulant agents,” write David R. Holmes Jr, MD, of the Mayo Clinic (Rochester, MN), and colleagues.

For the meta-analysis, the researchers looked at 2,406 patients with nonvalvular A-fib who were enrolled in PROTECT AF, PREVAIL, or their respective registries between 2005 and 2014. A total of 1,877 patients were treated with the Watchman device (Boston Scientific)—including 1,145 registry patients—and 382 patients from the control arms in both RCTs received warfarin.

Watchman was approved by the FDA in March for use in high-risk, nonvalvular A-fib patients.

Mean patient age ranged from 72 years in PROTECT AF to 75.3 years in the Continued Access to PREVAIL (CAP2) registry. Patient risk increased over the course of the study period, with CHADS2 scores increasing from 2.2 in PROTECT AF to 2.7 in Continued Access to PROTECT AF registry. Similarly, the CHA2DS2-VASc score increased from 3.5 to 4.5 (P < .0001 for both). Between 93.2% and 98.7% of patients were able to discontinue warfarin at 1 year.

Fewer Hemorrhagic, More Ischemic Strokes With Watchman

With regard to the composite efficacy endpoint of stroke, systemic embolism, and cardiovascular death, closure with Watchman met noninferiority criteria vs warfarin among the RCT populations. Event rates per 100 patient-years tended to favor LAA closure, with the exception of ischemic stroke, which was higher in Watchman-treated patients (table 1).

Table 1. Efficacy Among RCT Patients: Watchman vs Warfarin

When procedure-related strokes were excluded from the analysis, the difference in ischemic stroke disappeared (P = .21).

When the registry patients were included, the composite efficacy event rates were consistent. Kaplan-Meier analysis showed similar outcomes between the closure arms of the RCTs and the registry patients, even though the latter had higher baseline risk.

Analysis Primarily Informs Watchman Safety

In a telephone interview with TCTMD, Dr. Holmes explained that the impetus for this meta-analysis was the small sizes of PREVAIL and PROTECT AF. He said that when his group met with the FDA last year, “there was a great deal of interest in having this totality of data.”

After PREVAIL was published, Dr. Holmes commented that there had been concern that the data looked different than they had in PROTECT AF “even though PREVAIL was never powered to look at efficacy.” So this latest iteration of data, he said, shows consistency and reveals that the device is used in higher-risk patients in the real world.

In a telephone interview with TCTMD, Robert J. Sommer, MD, of Columbia University Medical Center (New York, NY), said the meta-analysis is a “rehash” of what is already known. “When you combine the short-term data of PREVAIL with the longer-term data of PROTECT AF, it’s not nearly as compelling as it was with PROTECT AF before.”

Given that the new RCTs looking at novel oral anticoagulants compared with warfarin include tens of thousands of patients, the researchers had a good idea of “what to expect from a warfarin control group,” he said. “But for whatever reason, in the PREVAIL study, the actual stroke rate was less than a third of what was anticipated…. There was really no way that the device could compete.”

Primarily, Dr. Sommer suggested, this meta-analysis “gives us more information about the safety of the procedure…. These combined data show that the overall safety of this procedure is really quite good; that the high rate of complications, which is what kept the FDA officially from approving the device, is really much lower than we initially showed; and that we can continue to do this procedure safely even at new centers.”

Additionally, “what this focuses on is the fact that the Watchman device is a local, site-specific therapy,” Dr. Holmes said. “Local, site-specific therapy does not do anything if there is carotid disease or if there's aortic atheroma.... Given the fact that most of the patients have strokes in the setting of nonvalvular A-fib from embolic events from the LAA, that means this device is going to work.”

If periprocedural strokes can be eliminated, “which I think to a large extent [they] have been dramatically decreased,” he said, “then I think there will be fewer ischemic strokes. There will still be more hemorrhagic strokes with the warfarin group, and it's the hemorrhagic strokes that are the most lethal and have the most associated morbidity.”

Uncertainty Over Late Outcomes

In an accompanying editorial, Jonathan W. Waks, MD, and Warren J. Manning, MD, both of Beth Israel Deaconess Medical Center (Boston, MA), say that “[t]hese data support the use of the Watchman as an alternative to warfarin for thromboembolic prophylaxis in [A-fib]. However, uncertainty remains regarding the overall safety and late efficacy of the Watchman device.”

Dr. Holmes said, “The late efficacy in terms of the stroke rate brings up the fact that not all strokes are going to come from the LAA.” Another concern that has been raised, he continued, “is that the Watchman device doesn't necessarily completely occlude the LAA.… We still do not have a perfect device. All of the devices we have are round devices and it's an oval structure. That's an important consideration, so we need to have better devices going forward.”

Drs. Waks and Manning also point out that the RCTs, and thus this meta-analysis, only included patients who were eligible for warfarin. “The safety of the Watchman in patients who are poor candidates for even short-term anticoagulant or antiplatelet therapy is currently unknown,” they say.

“Additionally,” the editorialists write, “the meta-analysis’ unusually high rate of hemorrhagic strokes in the warfarin-treated group may limit the beneficial effect of [LAA closure] in the general population.”

Though they expect the Watchman device to improve over time, Drs. Waks and Manning point out that “pharmaceutical approaches to [A-fib] stroke prevention are not stagnant, with the increasing use of nonvitamin K oral anticoagulants demonstrating improved efficacy and lower bleeding rates vs warfarin.”

While Watchman-treated patients may “pay early (procedural complications) and may also pay later (late thromboembolism),” they conclude, “[t]he jury is still out as we await clarity from long-term efficacy data.”


1. Holmes DR Jr, Doshi SK, Kar S, et al. Left atrial appendage closure as an alternative to warfarin for stroke prevention in atrial fibrillation: a patient-level meta-analysis. J Am Coll Cardiol. 2015;65:2614-2623.
2. Waks JW, Manning WJ. Left atrial appendage closure to reduce the risk of thromboembolic complications in atrial fibrillation: pay now and possibly pay later [editorial]? J Am Coll Cardiol. 2015;65:2624-2627.

Related Stories:

  • Dr. Holmes reports holding financial interest in technology related to the study, which has been licensed to Boston Scientific.
  • Drs. Waks and Manning report no relevant conflicts of interest.
  • Dr. Sommer reports participating in the PREVAIL trial and the CAP2 registry and holding stock in Coherex, which makes an LAA closure device.

We Recommend