Journal of the American College of Cardiology (JACC) July 08, 2014

Prospective Randomized Evaluation of the Watchman Left Atrial Appendage Closure Device in Patients With Atrial Fibrillation Versus Long-Term Warfarin Therapy: The PREVAIL Trial

David R. Holmes Jr, Saibal Kar, Matthew J. Price, et al.

Bottom Line

The PREVAIL trial demonstrated that while left atrial appendage closure with the Watchman device did not meet the prespecified noninferiority criteria for overall efficacy against warfarin, it achieved noninferiority for late ischemic events and showed significantly improved procedural safety compared to earlier studies.

Key Findings

1. At 18 months, the first coprimary efficacy endpoint (composite of stroke, systemic embolism, and cardiovascular/unexplained death) occurred at a rate of 0.064 in the Watchman group versus 0.063 in the warfarin control group [2.1.1].
2. The rate ratio for the first coprimary endpoint was 1.07 (95% credible interval [CrI]: 0.57 to 1.89); because the upper boundary of 1.89 exceeded the prespecified noninferiority margin of 1.75, statistical noninferiority was not formally achieved.
3. The second coprimary efficacy endpoint (stroke or systemic embolism >7 days post-randomization) occurred at a rate of 0.0253 in the device group versus 0.0200 in the control group (risk difference 0.0053 [95% CrI: -0.0190 to 0.0273]), which successfully achieved noninferiority.
4. Early procedure-related safety events occurred in only 2.2% of the Watchman arm, satisfying the pre-specified safety performance goal and marking a significant improvement over the preceding PROTECT AF trial.
5. The rate of pericardial effusions requiring surgical repair was exceptionally low at 0.4%, representing a significant decrease from the 1.6% observed in PROTECT AF (p=0.027).

Study Design

Design
RCT
Open-Label
Sample
407
Patients
Duration
18 mo
Median
Setting
Multicenter, US
Population Patients with nonvalvular atrial fibrillation (NVAF) who were eligible for short-term warfarin therapy and had a CHADS2 score ≥2 (or a score of 1 with an additional risk factor).
Intervention Percutaneous left atrial appendage (LAA) occlusion with the Watchman device, followed by short-term anticoagulation and subsequent discontinuation of warfarin.
Comparator Chronic, long-term oral anticoagulation with warfarin.
Outcome Two efficacy and one safety coprimary endpoints. Efficacy 1: composite of stroke, systemic embolism, and cardiovascular/unexplained death. Efficacy 2: stroke or systemic embolism >7 days post-randomization. Safety: occurrence of predefined procedure-related complications.

Study Limitations

The trial failed to achieve formal statistical noninferiority for its first composite efficacy endpoint, largely due to an unexpectedly low event rate in the control arm.
The trial was limited by a relatively small sample size (407 patients) and short initial follow-up duration (18 months).
The comparator arm utilized warfarin rather than direct oral anticoagulants (DOACs), which are the contemporary standard of care for nonvalvular atrial fibrillation.
The open-label nature of the trial, though practically necessary for comparing a device to a medication, introduces inherent biases.

Clinical Significance

PREVAIL was instrumental in confirming that LAA occlusion could be performed safely, effectively neutralizing the major critique of the earlier PROTECT AF trial which had a high rate of procedural complications. Although the trial missed one of its composite noninferiority endpoints, the absolute rates of stroke and systemic embolism were numerically low and comparable to well-managed warfarin therapy. This established the Watchman device as a reasonable, safe alternative for stroke prevention in patients with nonvalvular atrial fibrillation who have an appropriate rationale to avoid chronic oral anticoagulation, ultimately securing FDA approval for the device in 2015.

Historical Context

The concept of LAA occlusion gained serious traction with the PROTECT AF trial (2009), which demonstrated noninferiority to warfarin but revealed an alarming periprocedural complication rate (mostly pericardial effusions and procedural strokes). This safety signal prompted the FDA to mandate a second, confirmatory trial—PREVAIL. Conducted entirely in the US and specifically designed to assess whether improved operator training and procedural techniques could mitigate the early hazard, PREVAIL's primary mission was proving procedural safety. By drastically reducing complications, PREVAIL validated the modern era of structural interventions for stroke prevention in AFib.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

Why is the left atrial appendage specifically targeted for stroke prevention in nonvalvular atrial fibrillation, and what is the anatomical rationale for using a device like Watchman instead of systemic anticoagulation?

Key Response

Tests basic understanding of LAA anatomy as the primary source of thrombus in nonvalvular AF (accounting for >90% of cases) due to blood stasis, contrasting with valvular AF where thrombi can form elsewhere in the atria. It highlights the mechanical vs. pharmacological approach to stroke prevention.

Resident
Resident

Given that the PREVAIL trial did not meet its primary overall efficacy non-inferiority endpoint against warfarin, in which specific patient populations would you clinically justify recommending a Watchman device today?

Key Response

Encourages applying trial data to clinical decision-making. Residents must recognize that despite missing the primary efficacy endpoint, the reduction in late ischemic events and improved procedural safety make LAAC appropriate for patients with high bleeding risk or strict contraindications to long-term oral anticoagulation.

Fellow
Fellow

How does the implant success rate and procedural complication rate in PREVAIL compare to the earlier PROTECT AF trial, and what does this difference illustrate about the learning curve and operator volume in structural heart interventions?

Key Response

Fellows need to evaluate procedural data contextually. PREVAIL showed a significant drop in pericardial effusions and procedure-related strokes compared to PROTECT AF, highlighting the importance of operator experience, device iterations, and refined techniques in the evolution of structural cardiology.

Attending
Attending

When counseling a patient on the PREVAIL data, how do you balance the failure to achieve formal non-inferiority for overall efficacy with the achieved non-inferiority for late ischemic events to facilitate shared decision-making?

Key Response

Attendings must navigate complex, mixed trial results with patients. The discussion should focus on translating the nuances of early periprocedural risk versus long-term bleeding reduction into patient-centered outcomes, establishing acceptable trade-offs based on individual patient values and bleeding history.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The PREVAIL trial utilized a Bayesian statistical design to incorporate prior data from the PROTECT AF trial. What are the methodological advantages and potential bias risks of using Bayesian adaptive designs with informative priors in regulatory device trials?

Key Response

Challenges the researcher to critique statistical modeling. Using informative priors can increase statistical power and reduce required sample sizes, but it risks carrying over biases or assumptions from earlier trials, which is particularly problematic when procedural safety dynamics or patient populations are shifting between the trials.

Journal Editor
Journal Editor

From an editorial perspective, how should a journal handle the publication of a non-inferiority trial where the primary efficacy endpoint is formally missed, but secondary or 'late' endpoints are highlighted as successes? What safeguards prevent outcome reporting bias or spin in the abstract?

Key Response

Focuses on peer-review stringency. A rigorous reviewer would flag the emphasis on secondary endpoints (late ischemia) when the primary endpoint failed, assessing whether the manuscript's tone and abstract conclusions objectively reflect the statistical reality without misleading clinical practitioners.

Guideline Committee
Guideline Committee

Considering PREVAIL's mixed efficacy results and its use of warfarin as a control, how should current ACC/AHA guidelines grade the recommendation for LAA closure in patients with nonvalvular AF in the modern era where Direct Oral Anticoagulants (DOACs) are the standard of care?

Key Response

Committees must weigh historical evidence against the current standard-of-care. Because PREVAIL compared Watchman to warfarin rather than DOACs, current guidelines (e.g., 2019 AHA/ACC/HRS) appropriately assign LAAC a Class IIb recommendation, restricting it primarily to patients with high stroke risk who have poor suitability for any long-term anticoagulation.

Clinical Landscape

Noteworthy Related Trials

2009

PROTECT AF Trial

n = 707 · Lancet

Tested

Watchman left atrial appendage closure device

Population

Patients with non-valvular atrial fibrillation eligible for warfarin

Comparator

Warfarin

Endpoint

Composite of stroke, systemic embolism, and cardiovascular or unexplained death

Key result: Watchman device was non-inferior to warfarin for preventing stroke, systemic embolism, and cardiovascular death, though associated with a higher rate of initial safety events.
2020

PRAGUE-17 Trial

n = 402 · JACC

Tested

Percutaneous left atrial appendage closure

Population

Patients with non-valvular atrial fibrillation at high risk for stroke and bleeding

Comparator

Direct oral anticoagulants (DOACs)

Endpoint

Composite of stroke, TIA, systemic embolism, cardiovascular death, clinically relevant bleeding, or procedure/device-related complications

Key result: Left atrial appendage closure was non-inferior to DOACs for the composite endpoint in high-risk AF patients.
2021

Amulet IDE Trial

n = 1,878 · Circulation

Tested

Amplatzer Amulet LAA occluder

Population

Patients with non-valvular atrial fibrillation at increased risk of stroke

Comparator

Watchman LAA occluder

Endpoint

Composite of ischemic stroke or systemic embolism

Key result: The Amulet device was non-inferior to the Watchman device for safety and effectiveness, with higher procedural success and less device-related leak.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis