PDF: PREVAIL
(2014)Objective
Left atrial appendage (LAA) occlusion with the Watchman device versus warfarin in patients with nonvalvular atrial fibrillation (NVAF).
Study Summary
Intervention
Randomized 2:1 to Watchman device implantation vs. warfarin therapy. Patients had NVAF and CHADS2 ≥2 (or 1 with additional risk factor). Device group discontinued warfarin after successful implantation per protocol. Follow-up: median 12 months.
Study Design
Arms: Array
Outcome
• Late ischemic stroke/SE (>7 days) — 2.5% vs. 2.0%; difference 0.53% (95% CrI: –1.9 to 2.7%); noninferiority met
• Early safety (procedure-related events) — 2.2% (95% CrI upper bound 2.652%); met safety performance goal
• Stroke rates: Ischemic 1.9% (device) vs. 0.7% (warfarin); hemorrhagic stroke rare
• Procedural complication rates improved vs. PROTECT AF (e.g., pericardial effusion requiring surgery 0.4%)
Bottom Line
In patients with NVAF, LAA occlusion with the Watchman device was noninferior to warfarin for preventing ischemic stroke or systemic embolism occurring more than 7 days after the procedure. It did not meet the noninferiority criteria for the composite endpoint of stroke, systemic embolism, and cardiovascular death, partly due to a very low event rate in the warfarin group. However, the procedural safety of the device improved significantly compared with the prior PROTECT AF trial, making it a reasonable alternative to warfarin for stroke prevention in appropriate patients.
Major Points
- PREVAIL was a randomized trial comparing the Watchman LAA closure device against chronic warfarin therapy in 407 patients with NVAF and a CHADS2 score ≥2 (or 1 with other risk factors).
- The trial had two co-primary efficacy endpoints analyzed at 18 months.
- First Efficacy Endpoint (stroke, systemic embolism, CV/unexplained death): Noninferiority was NOT achieved. The event rate was 6.4% in the device group vs. 6.3% in the warfarin group (Rate Ratio 1.07; 95% CrI: 0.57 to 1.89).
- Second Efficacy Endpoint (ischemic stroke or systemic embolism >7 days post-procedure): Noninferiority WAS achieved.
- The primary safety endpoint was met, with the rate of early adverse safety events in the device arm (2.2%) being significantly lower than in the previous PROTECT AF trial and meeting the pre-specified performance goal.
Study Design
- Study Type
- Prospective, multicenter, randomized clinical trial with a Bayesian design.
- Randomization
- Yes
- Blinding
- Unblinded. Participants and clinicians were not masked to treatment assignment.
- Sample Size
- 407
- Follow-up
- 18 months for the primary analysis.
- Centers
- Up to 50 sites.
- Countries
- United States
Primary Outcome
Definition: The first co-primary efficacy endpoint was a composite of stroke (ischemic or hemorrhagic), systemic embolism, and cardiovascular/unexplained death at 18 months.
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 0.063 (18-month event rate) | 0.064 (18-month event rate) | 1.07 (0.57 to 1.89 (Credible Interval)) | Noninferiority NOT achieved |
Limitations & Criticisms
- The trial population was required to be candidates for long-term warfarin, so results do not apply to patients with contraindications to anticoagulation.
- The Watchman device was not compared against new oral anticoagulants (NOACs).
- The clinical significance of the loss of LAA mechanical function after occlusion is unknown.
- The low overall event rate, particularly in the over-performing warfarin control arm, provided limited statistical power to establish noninferiority for the primary efficacy endpoint.
Citation
J Am Coll Cardiol 2014;64:1-12