Role of the Watchman Device in Managing Non-Valvular Atrial Fibrillation
The Watchman device should be considered primarily for patients with non-valvular atrial fibrillation who have absolute contraindications to long-term oral anticoagulation but remain at high risk for ischemic stroke. 1, 2
Mechanism and Efficacy
- The left atrial appendage (LAA) is the source of approximately 90% of thrombi in patients with non-valvular atrial fibrillation, making it a logical target for stroke prevention strategies. 1
- The Watchman device provides percutaneous LAA occlusion (LAAO) to prevent thrombus formation and subsequent thromboembolism. 3
- Two major randomized controlled trials have evaluated the Watchman device against warfarin:
- A meta-analysis of these trials demonstrated that patients receiving the Watchman had significantly fewer hemorrhagic strokes than those on warfarin, though there was an initial increase in ischemic strokes in the device group. 1
- When periprocedural events were excluded, the difference in ischemic stroke rates was not statistically significant. 1
Patient Selection Criteria
- Primary candidates: Patients with non-valvular AF at high risk of stroke (CHA₂DS₂-VASc score ≥2 or ≥3) who have absolute contraindications to long-term oral anticoagulation. 1, 2
- Secondary consideration: Patients with poor tolerance or adherence to oral anticoagulation therapy. 1
- The FDA approval specifies that patients should be deemed suitable for short-term warfarin therapy during the periprocedural period. 1
- Centers for Medicare & Medicaid Services (CMS) states the device is an option for patients suitable for short-term warfarin but deemed unable to take long-term oral anticoagulation. 1
Procedural Considerations and Risks
- Serious periprocedural complications occur in approximately 6-7% of cases, including: 1, 2
- Pericardial effusion requiring drainage
- Device embolization
- Procedure-related ischemic stroke
- A learning curve exists, with higher complication rates observed in the first few implantations at each center. 1
- Post-implantation, dual antiplatelet therapy (DAPT) is typically required for at least 6 weeks, which may expose patients to increased bleeding risk. 1, 3
- Thrombus formation on the device is not uncommon (up to 7.2% per year) and is associated with risk of ischemic stroke during follow-up. 1, 2
Comparison with Oral Anticoagulants
- Oral anticoagulation remains the preferred first-line therapy for stroke prevention in most patients with AF and elevated stroke risk. 1, 2
- No randomized controlled trials have directly compared the Watchman device with direct oral anticoagulants (DOACs), which have better safety profiles than warfarin. 1, 2
- A recent retrospective matched cohort study suggested that in high-risk patients with prior major bleeding events, the Watchman device was associated with lower rates of major bleeding events, TIAs, and ischemic strokes compared to DOACs over 5 years. 4
- Post-implantation antithrombotic regimens vary, with options including:
Current Guideline Recommendations
- American College of Cardiology/American Heart Association/Heart Rhythm Society (2019): Percutaneous LAA occlusion may be considered in patients with AF at increased risk of stroke who have contraindications to long-term anticoagulation (Class IIb, Level of Evidence B-NR). 1
- CHEST Guideline (2018): LAA occlusion is suggested in patients with AF at high risk of ischemic stroke who have absolute contraindications for oral anticoagulation (Weak recommendation, low quality evidence). 1
Limitations and Caveats
- Device-related thrombosis appears more common with antiplatelet therapy than with oral anticoagulation post-implantation. 5
- Transesophageal echocardiography (TEE) should be performed at 45 days and 12 months post-implantation to assess for device-related thrombus and residual peridevice flow. 3
- For surgical LAA occlusion during cardiac surgery in patients with AF, the need for long-term oral anticoagulation remains unchanged despite the procedure. 1
- The Lariat device (an alternative LAA closure approach) has shown incomplete occlusion in approximately 20% of cases with inexperienced operators. 1