← Back

PRAGUE-17

Left Atrial Appendage Closure vs. Novel Anticoagulation Agents in Atrial Fibrillation

Share Share Copied! Compare

Year of Publication: 2020

Authors: Pavel Osmancik, Dalibor Herman, Petr Neuzil, ..., Vivek Y. Reddy

Journal: Journal of the American College of Cardiology

Citation: J Am Coll Cardiol 2020;75:3122–35

Link: https://doi.org/10.1016/j.jacc.2020.04.067

Clinical Question

Is left atrial appendage closure noninferior to direct oral anticoagulants for preventing cardiovascular and bleeding events in high-risk atrial fibrillation patients?

Bottom Line

Among high-risk AF patients, LAAC was noninferior to DOACs for preventing the composite of stroke, systemic embolism, cardiovascular death, major bleeding, and procedure-related complications, though procedural safety remains a concern.

        Major Points
        First randomized trial comparing LAAC directly with DOACs (primarily apixaban)
High-risk population: mean CHA2DS2-VASc 4.7, 47.8% with prior bleeding, 35.3% with prior stroke
LAAC achieved 90% procedural success rate with 4.5% major complications
No significant differences in individual components: stroke/TIA, cardiovascular death, or bleeding
Noninferiority maintained across all pre-specified and post-hoc analyses

      

Design

Study Type: Multicenter, randomized, open-label, noninferiority trial

Randomization: 1

Blinding: Clinical endpoint committee blinded to treatment allocation; participants and investigators not blinded

Enrollment Period: October 2015 to January 2019

Follow-up Duration: Median 19.9 months (695.9 patient-years)

Centers: 10

Countries: Czech Republic

Sample Size: 402

Analysis: Modified intention-to-treat analysis using Fine-Gray competing risk regression models with subdistribution hazard ratios, SPSS version 25.0

Inclusion Criteria

Nonvalvular atrial fibrillation
Indication for oral anticoagulation
History of bleeding requiring intervention or hospitalization, OR
History of cardioembolic event while taking anticoagulation, OR
CHA2DS2-VASc ≥3 plus HAS-BLED >2

Exclusion Criteria

Mechanical valve prosthesis
Mitral stenosis
Comorbidities other than AF mandating anticoagulation
Patent foramen ovale with large atrial septal aneurysm
Mobile aortic plaque
Symptomatic carotid arterial atherosclerosis
Clinically significant bleeding within 30 days
Cardioembolic event within 30 days
Creatinine clearance <30 ml/min
Left atrial or LAA thrombus on TEE

Arms

Field	Control	LAAC
Intervention	Direct oral anticoagulants at manufacturer-recommended doses (apixaban 95.5%, dabigatran 4.0%, rivaroxaban 0.5%)	Left atrial appendage closure with Amulet (61.3%), Watchman (35.9%), or Watchman-FLX (2.8%) devices, followed by aspirin plus clopidogrel for 3 months, then aspirin indefinitely
Duration	Long-term	Single procedure with long-term antiplatelet therapy

Outcomes

Outcome	Type	Control	Intervention	HR / OR / RR	P-value
Composite of stroke, TIA, systemic embolism, cardiovascular death, major or nonmajor clinically relevant bleeding, or procedure/device-related complications	Primary	13.42% per year (41 patients)	10.99% per year (35 patients)	0.84	0.44 (p=0.004 for noninferiority)
All stroke/TIA	Secondary	2.57% per year	2.60% per year	1	sHR 95% CI: 0.40-2.51
Cardiovascular death	Secondary	4.28% per year	3.18% per year	0.75	sHR 95% CI: 0.34-1.62
ISTH major/NMCRB	Secondary	7.42% per year	5.50% per year	0.81	sHR 95% CI: 0.44-1.52
Non-procedural bleeding	Secondary	7.42% per year	3.76% per year	0.53	sHR 95% CI: 0.26-1.06
Procedure-related complications	Adverse	0%	4.5% (9 patients)
Pericardial effusion	Adverse	0%	1.1% (2 patients)
Device embolization	Adverse	0%	0.5% (1 patient)
Procedure/device-related death	Adverse	0%	1.0% (2 patients)

Subgroup Analysis

Primary endpoint results were consistent across all subgroups including age, sex, CHA2DS2-VASc score, bleeding history, and center experience with no statistically significant interactions

Criticisms

Underpowered to detect differences in individual components like stroke alone
Short median follow-up (19.9 months) may not capture long-term differences
Open-label design with potential for bias
High-risk population limits generalizability to lower-risk AF patients
Composite endpoint mixing efficacy and safety outcomes may obscure clinically important differences
Procedural complications including 2 deaths highlight ongoing safety concerns with LAAC
Four centers were de novo LAAC sites which may have affected complication rates

Funding

Ministry of Health, Czech Republic (AZV 15-29565A)

Based on: PRAGUE-17 (Journal of the American College of Cardiology, 2020)