New Watchman Data Put Spotlight on Ischemic Stroke in AF

Shelley Wood

September 23, 2014

WASHINGTON, DC — New, longer-term data on the Watchman left atrial appendage occluder (Boston Scientific) studied in the PREVAIL trial points to an increased risk of ischemic stroke in atrial-fibrillation patients treated with this novel device, a finding that appears to be countered, in part, by a lower risk of hemorrhagic stroke than that seen among patients taking warfarin.

But the higher rate of ischemic strokes in the Watchman-treated patients means that the second of two co–primary end points—and the only one that was positive in favor of the Watchman at 18 months—is now no longer statistically significant.

As previously reported by heartwire , PREVAIL had a primary safety end point (which was met) and two primary efficacy end points, only one of which was met. The trial was scheduled to be presented at the 2013 ACC meeting but was yanked last minute after an embargo break by the study sponsor. The results were subsequently published in the Journal of the American College of Cardiology.

In PREVAIL, the Watchman device failed to demonstrate noninferiority on the first co–primary end point at 18 months: a composite of stroke, systemic embolism, and cardiovascular/unexplained death among those who received the device compared with historical rates based on data from the PROTECT AF study. The second efficacy end point—a comparison of ischemic stroke or systemic embolism occurring more than seven days postrandomization—did meet the criterion for noninferiority at 18 months.

But as Dr Vivek Reddy (Mount Sinai Hospital, New York, NY) showed at TCT 2014 last week, eight new ischemic strokes have occurred in the Watchman arm since 2013, while no ischemic strokes have occurred in the comparator warfarin group. One additional hemorrhagic stroke occurred in the Watchman group in this period and two in the warfarin group, one of which was fatal. Overall, however, deaths were numerically higher among the warfarin-treated patients.

In a separate study, also released last week at TCT, Dr David Holmes presented a patient-level meta-analysis combining patients from PREVAIL and the earlier PROTECT AF study. When the two trials were combined, there was no difference in the primary efficacy outcome (stroke, systemic embolism, cardiovascular or unexplained death) between the Watchman and warfarin groups. But here again, the analysis also showed a statistically significant increase in the rate of ischemic strokes among Watchman-treated patients.

FDA Panel to Mull Stroke Data

The increase in ischemic strokes is widely believe to be the focus of a third FDA advisory committee review of the Watchman, which has twice received majority votes in favor of approval from FDA advisors.

Dr Gregg Stone (Columbia University, New York, NY), who moderated the TCT session where the new data were presented, said he had no direct knowledge of what the FDA requested or not. "However, these talks do show new and important data that [it's] fair to assume the FDA advisory board (and FDA) will be considering," Stone said in an email to heartwire . "The long-term results of PROTECT AF and PREVAIL are highly different, the former quite positive, the latter negative. It is undeniable that the control arm in PREVAIL did unexpectedly well, but 'the data are the data.' "

Likewise, also speaking with heartwire , Dr Matthew Price (Scripps Clinic, La Jolla, CA), who was one of the discussants at the TCT session where the new data were presented, said the FDA advisory panel members have their work cut out for them.

The statistics and methodology are "challenging" in themselves, he noted, predicting discussion around the ability to pool PROTECT AF and PREVAIL. There's also the issue of much lower-than-expected stroke rates in the comparator arm, as Reddy elaborated in his presentation.

PREVAIL, despite enrolling patients with a mean CHADS2 score of 2.6, produced a rate of stroke/embolism in the warfarin arm that was just one per 100 patient-years—much lower than that seen in other recent studies in this population, including studies of the new oral anticoagulants.

The rate of ischemic strokes in the warfarin-treated patients was even lower, at 0.3 per 100 patient-years.

"You can't just blame your control group as the reason that a study was negative," Price told heartwire , but he predicted that the FDA will be looking to its advisors for guidance on this point.

"I think there will be a healthy discussion. Here you have a trial which over continuing follow-up in a post hoc way is drifting in the wrong direction; however, the absolute event rates are sort of what one would expect and are better than the imputed placebo."

While Price predicts the upcoming panel won't be a "slam dunk" for the Watchman, his personal view, after seeing the new data, is that there is a role for the Watchman in a subset of patients.

"There are a large proportion of patients out there who are still not treated with oral anticoagulants, based on real or perceived complications. So even though we have these great new drugs, people at risk aren't always being treated with them, they don't want to take oral anticoagulants, they are noncompliant, they don't want to take a medication every day, or they are afraid of having a bleeding event."

He continued: "My [sense] would be that the Watchman, based on these data, is not a replacement for warfarin or oral anticoagulants in all patients, but for a narrow group of patients in whom long-term oral anticoagulation would not be the best thing for them, this would be a reasonable option."

That's a view also echoed by Stone, who believes the totality of data support acceptable safety for the Watchman, "with a notable reduction in periprocedural complications over time with operator experience and adequate training; compared with warfarin, there is a substantial safety benefit in terms of reduction in major bleeding and intracranial hemorrhage; there is somewhat less protection from ischemic stroke and systemic embolism with Watchman compared with warfarin, but this is offset by an observed reduction in CV/unexplained mortality with Watchman."

In other words, Stone continued, a "positive net clinical benefit."

"All in all," he said, "considering the totality of the data, my opinion is that the Watchman has a role to play for select patients at increased bleeding risk or in whom warfarin—and the [new oral anticoagulants]—are not appropriate chronic therapy."

Price disclosed grant support/research contracts with SentreHeart and consultant fees/honoraria/speaker's-bureau payments from Boston Scientific, Daiichi-Sankyo/Eli Lilly, Medtronic, St Jude Medical, Terumo Medical, the Medicines Company, WL Gore and Associates, and AstraZeneca. Reddy disclosed receiving grant support/research contracts and consultant fees/honoraria/speaker's-bureau payments from Boston Scientific, Coherex Medical, Biosense Webster, Cardiofocus, St Jude Medical, and VytronUS in addition to consultant fees/honoraria/speaker's-bureau payments from Medtronic. Holmes disclosed, "Both Mayo Clinic and I have a financial interest in the technology related to this research. That technology has been licenses to Boston Scientific." Stone had no relevant conflicts of interest.


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.