PROTECT AF

Published

Holmes DR, et al. "Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial". Lancet. 2009. 374(9689):534-42.
PubMed • Full text

Clinical Question

Among patients with non-valvular atrial fibrillation, is percutaneous left atrial appendage closure with the WATCHMAN device non-inferior to warfarin for the prevention of vascular events?

Bottom Line

Among patients with non-valvular atrial fibrillation, percutaneous left atrial appendage closure with the WATCHMAN device (without continued anticoagulation) is non-inferior to warfarin for the prevention of vascular events.

Major Points

It is estimated that only 50% of patients with non-valvular atrial fibrillation who qualify for warfarin are actually treated. This is a concern as the annual risk of stroke among those who are not anticoagulated range from 1.5% in individuals aged 50–59 years to 23.5% for those aged 80–89 years. Alternatives to anticoagulation have thus been pursued. One strategy involves obliteration of the left atrial appendage (LAA) since this site is a common location of clot formation in non-valvular AF.

PROTECT AF was a randomized controlled trial conducted in the US and Europe which randomized 707 patients with non-valvular AF and CHADS2 score of ≥1 to either LAA closure with the WATCHMAN device or to warfarin control in a 2:1 fashion. Those randomized to LAA closure discontinued their warfarin therapy 6 weeks after device placement. The study's objective was to determine noninferiority of LAA closure compared to warfarin. Patients were followed for the primary composite outcome of stroke, cardiovascular or unexplained death, or systemic embolism. With a median follow-up of 18 months, the primary outcome rate was 3.0 per 100 person-years in the LAA closure group compared to 4.9 per 100 person-years in the warfarin control (RR 0.62) and non-inferiority criteria were met. Adverse events were more common in the device group compared to warfarin controls (RR 1.69); in the device group these occurred early, whereas in the warfarin group these occurred late. Notably ischemic strokes were more common in the device group than in the warfarin controls, and most of the events in the device group were attributed to periprocedural air embolism. The authors conclude that LAA closure may be a reasonable option for patients in whom long-term anticoagulation is not desirable. The 2014 AHA/ASA guidelines contend that LAA closure with the WATCHMAN device may be most appropriate for patients with AF at high risk of stroke who are poor candidates for anticoagulation.

In 2013, further follow up data from the original patients were published providing more longitudinal safety information. The PREVAIL trial, a subsequent 2014 study involving the WATCHMAN device following the original PROTECT AF protocol, failed to demonstrate non-inferiority of LAA closure to warfarin using the same primary composite endpoint but had significantly lower adverse events when performed by more experienced operators. The subsequent follow up ASAP study (ASA Plavix Feasibility Study With Watchman Left Atrial Appendage Closure Technology) showed the safety of using dual antiplatelets only without the warfarin bridging period allowing the device to be used in patients with absolute contraindications to any length of warfarin use. A 2015 meta-analysis of PROTECT AF and PREVAIL demonstrated that LAA closure reduced the rates of hemorrhagic stroke, cardiovascular or unexplained death, and nonprocedural bleeding compared to warfarin.

Guidelines

AHA/ACC/HRS focused update of the 2014 guideline (2019, adapted)

• Percutaneous LAA occlusion may be considered in patients with AF at increased risk of stroke who have contraindications to long-term anticoagulation. (COR IIb, LOE B-NR)
• Surgical occlusion of the LAA may be considered in patients with AF undergoing cardiac surgery, as a component of an overall heart team approach to the management of AF. (COR IIb, LOE B-NR)

Design

• Multicenter, open label, randomized control trial.
• N=707 patients with paroxysmal, persistent, or permanent non-valvular AF
  • LAA closure (n=463)
  • Warfarin control (n=244)
• Setting: 59 centers in USA and Europe
• Enrollment: 2005-2008
• Mean follow-up: 18 months
• Analysis: Intention to treat
• Primary outcome: Composite of stroke (including ischemic and hemorrhagic), systemic embolism and cardiovascular or unexplained death

Population

Inclusion Criteria

• Age ≥18 years
• Paroxysmal, persistent, or permanent non-valvular AF
• CHADS2 score ≥1

Exclusion Criteria

• Contraindicated for warfarin
• Contraindicated for aspirin or clopidogrel (Plavix)
• Comorbidity other than AF requiring chronic warfarin use
• NYHA class 4 HF
• Implanted mechanical valve
• Atrial septal or PFO device
• Platelets <100K or hemoglobin <10 g/dL
• LAA thrombus
• A patent foramen ovale with atrial septal aneurysm and right-to-left shunt
• Mobile aortic atheroma
• Symptomatic carotid artery disease
• Left ventricular ejection fraction (LVEF) < 30%

Patients underwent transoesophageal echocardiography prior to enrollment to assess for exclusion criteria

Baseline Characteristics

Results are from the LAA closure group.

• Demographics: Age 72 years, 70% males
  • Race/ethnicity: Asian 1%, black 1%, white 92%, Hispanic/Latin American 5%
• CHADS2 score:
  • 1: 34%
  • 2: 34%
  • 3: 19%
  • 4: 8%
  • 5: 4%
  • 6: 1%
• PMH: HF 27%, HTN 89%, DM 24%, TIA/stroke 18%
• Age 75 years or more: 41%
• Previous warfarin use:
  • <1 year: 55%
  • ≥1 year: 44%
  • No estimate: 1%
• AF pattern: Paroxysmal 43%, persistent 21%, permanent 35%, unknown 1%
• AF onset:
  • <1 year: 15%
  • ≥1 year: 78%
  • No estimate: 7%
• LVEF: 57%

Interventions

• Patients underwent transesophageal echocardiography during screening phase to evaluate for exclusion criteria.

• Randomization to:
  • LAA closure with the WATCHMAN device (Atritech, Plymouth, MN, USA). The device is a self-expanding nickel titanium (nitinol) frame structure with fixation barbs and a permeable polyester fabric cover (which diameters from 21 mm to 33 mm to fit patient anatomy). Device placed using trans-septal catheter based approach guided by fluoroscopy and TEE. TEE imaging was done at 45 days, 6 months, and 12 months. Patients discontinued warfarin therapy if the 45-day TEE showed either complete closure of the LAA or residual peri-device flow (jet <5 mm in width). Once daily clopidogrel (75 mg) and aspirin (81–325 mg) until 6 month post procedure, then aspirin indefinitely.
  • Warfarin control: Warfarin titrated to target INR of 2.0-3.0. INR checked at least every 2 weeks for 6 months and at least once a month thereafter.

Outcomes

Comparisons are LAA closure vs. warfarin control unless otherwise specified.

Primary Outcome

Composite end-point (event/patient year)

21/694.1 vs. 18/370.8, RR=0.62 (95% CI 0.35–1.25), non-inferiority probability >99.9%, superiority probability=90.0%

Secondary Outcomes

Successful LAA device implantation

88% (408/463) of patients assigned to this intervention

Therapeutic INR on warfarin

66% of the time in patients on this intervention

Ischaemic stroke (event/patient year)

15/694.6 vs. 6/372.3, RR= 1.34 (95% CI 0.60–4.29), non-inferiority probability=71.8%, superiority probability=20.1%

Cardiovascular/unexplained death (event/patient year)

5/708.4 vs. 10/374.9, RR= 0.26 (95% CI 0.08–0.77), non-inferiority probability >99.9%, superiority probability=99.3%

Haemorrhagic stroke (event/patient year)

1/708.4 vs. 6/373.4, RR=0.09 (95% CI 0.00–0.45), non-inferiority probability >99.9%, superiority probability=99.8%

Systemic embolism (event/patient year)

2/707.8 vs 0/374.9

All stroke (event/patient year)

16/694.6 vs. 12/370.8, RR= 0.71 (95% CI: 0.35–1.64), non-inferiority probability 99.3%, superiority probability=76.9%

All-cause mortality (event/patient year)

21/708.4 vs. 18/374.9, RR 0.62 (95% CI: 0.34–1.24), non-inferiority probability >99.9%, superiority probability 90.7%

Subgroup Analysis

HR with treatment in men was lower than that for women (p=0.03; all other interaction tests p>0.40).

For men, HR 0.32 (95% CI 0.13–0.77)

For women, HR 1.47 (95% CI 0.52–4.11)

No other significant subgroup effects were seen (including age, CHADS2 score, type of AF, LAA size, LVEF). This did not change with the publication of the follow up data in 2013.

Adverse Events

Serious pericardial effusion (need for percutaneous or surgical drainage)

22 (4.8%) vs. 0

Major bleeding (required at least 2 units of packed red blood cells or surgery to correct)

16 (3.5%) vs. 10 (4.1%)

Haemorrhagic stroke (all but on in the control group resulted in death)

1 (0.2%) vs. 6 (2.5%)

Procedure-related ischaemic stroke

5 (1.1%) vs. 0

Device embolisation

3 (0.6%) vs. 0

Other procedural complications

2 (0.4%) vs. 0 (esophageal tear and procedure-related arrhythmia)

Criticisms

• Non-inferiority was not proven for prevention of ischemic stroke, which is the primary goal of intervention for AF
• Intervention patients are left on ASA long term, which itself offers mild stroke prevention benefit for patients with atrial fibrillation

Funding

Atritech, maker of the WATCHMAN device.