When Anticoagulation Isn’t an Option: What to Do?

For a large number of patients with nonvalvular atrial fibrillation (AF) and high risk of stroke, oral anticoagulation may be relatively contraindicated because of previous bleeding episodes, risk of falls, patient-related issues (ie, renal impairment or drug-drug interactions), or narrow therapeutic windows. There are also many who, for a wide variety of reasons, are nonadherent to their anticoagulant medication regimens. For these patients, a procedural option with a left atrial appendage closure device may be an alternative.

At the American College of Cardiology 2013 session in San Francisco, authors of the PREVAIL trial¹ reported the results of a multicenter study of left atrial appendage closure with the Watchman device vs warfarin for stroke/thromboembolic prevention in patients with nonvalvular AF. Although the oral presentation was canceled because of an embargo violation by the sponsor Boston Scientific, the results were presented and discussed extensively.

PREVAIL was primarily a safety trial with a non-inferiority design. There were 407 patients with AF requiring anticoagulation (mean CHADS₂ was 2.6) from 41 US centers randomized to receive either warfarin or undergo surgical closure with the Watchman device. The device had previously been evaluated in the PROTECT study² against warfarin and was rejected by the FDA because of concerns regarding increased bleeding and procedure-related complications. In this study, patients in the device group were given 45 days of warfarin therapy following implantation; there was a 95% implant success rate with a 4.4% rate of major complications.

The main safety end point in PREVAIL, which analyzed end points at 7 days and included death, ischemic stroke, systemic embolism, and procedure- or device-related complications requiring endovascular or cardiovascular intervention, occurred at a much lower rate (2.2%) than in PROTECT. Since the prespecified noninferiority criteria were established at an event rate of 2.6%, the device was deemed noninferior to warfarin on the basis of these results.

The second safety end point was a composite of the rate of cardiac perforation and risk for pericardial effusion with tamponade, which was 4.4% in Watchman group, much lower than 8.7% reported for the same end point in the PROTECT study.

With respect to efficacy end points, the Watchman device group had a similar rate of the composite end point of stroke or systemic embolism at 7 days following implantation (0.025%), compared with the warfarin group (0.020%). The rate for the triple composite endpoint of stroke, systemic embolism, and cardiovascular or unexplained death at 18 months was also the same for both groups (0.064%). 

Although the trial established its goal of noninferiority of the Watchman device, the outcomes with respect to safety and efficacy for the closure device are overshadowed by concerns about this trial’s design and conduct. Notably, the confidence intervals for the 18-month efficacy follow-up were very wide, crossing the border of noninferiority, because of small sample sizes of patients with long-term follow-up. Furthermore, the rate of stroke in the warfarin group was much lower than expected (0.7%), and there is little information provided on important measures, such as the time in therapeutic range for patients in the warfarin arm. Most important, there may still be a residual stroke risk even after implantation of the device that may necessitate oral anticoagulation, which may make the device less useful in patients for whom the indication is inability to tolerate oral anticoagulation. There also may be a steep learning curve for placement of the Watchman device, a procedure that requires a transseptal puncture of the interatrial appendage.

These limitations make it challenging to interpret and apply this noninferiority trial to clinical practice. Whether the FDA will approve this device on the basis of the results of the PREVAIL remains to be seen.   However, this promising technology may offer an attractive option to those patients who have a high CHADS₂ score, but cannot take oral anticoagulation medication. More news from the FDA to follow.


References:
1. Holmes DR Jr, Doshi S, Kar S, et al. Results of Randomized Trial of LAA Closure vs Warfarin for Stroke/Thromboembolic Prevention in Patients with Non-valvular Atrial Fibrillation (PREVAIL). ACC 2013.

2. Reddy VY, Holmes D, Doshi SK, et al. Safety of percutaneous left atrial appendage closure: results from the Watchman Left Atrial Appendage System for Embolic Protection in Patients with AF (PROTECT-AF) Clinical Trial and the Continued Access Registry. Circulation. 2011;123:417-424.