PDF: OAC-ALONE
(2019)Objective
To compare oral anticoagulation alone versus combined oral anticoagulation and single antiplatelet therapy in patients with atrial fibrillation and stable coronary artery disease beyond 1 year after coronary stent implantation
Study Summary
• Primary endpoint: OAC alone 15.7% vs combination 13.6% (HR 1.16, P=0.20 for noninferiority)
• Major bleeding higher with combination therapy (10.4% vs 7.8%), though not statistically significant
Intervention
OAC alone (warfarin or DOACs) versus OAC plus single antiplatelet therapy (aspirin or clopidogrel) for patients >1 year post-coronary stenting
Inclusion Criteria
Age ≥20 years, AF beyond 1 year after coronary stenting, currently on combination of OAC and single antiplatelet agent
Study Design
Arms: OAC alone vs OAC plus single antiplatelet therapy (aspirin or clopidogrel)
Patients per Arm: 344 OAC alone, 346 OAC plus antiplatelet
Outcome
• Major secondary outcome (primary + major bleeding): 19.5% vs 19.4% (HR 0.99, P=0.016 for noninferiority, met)
• ISTH major bleeding: 7.8% vs 10.4% (HR 0.73, P=0.22)
Bottom Line
This randomized trial failed to establish noninferiority of OAC alone compared to combined OAC and antiplatelet therapy for the primary composite endpoint due to premature termination and inadequate statistical power, making the study inconclusive. The trial is the first randomized comparison in this population but warrants future larger studies to determine the optimal antithrombotic regimen.
Major Points
- First randomized trial comparing OAC alone vs OAC plus antiplatelet in AF patients with stable CAD >1 year post-stenting
- Trial prematurely terminated after enrolling only 696 of planned 2000 patients over 38 months due to slow enrollment
- Noninferiority not established for primary endpoint (death/MI/stroke/embolism): 15.7% OAC alone vs 13.6% combination (HR 1.16, 95% CI 0.79-1.72, P=0.20 for noninferiority)
- Noninferiority met for major secondary endpoint (primary + major bleeding): 19.5% vs 19.4% (HR 0.99, 95% CI 0.71-1.39, P=0.016 for noninferiority)
- MI occurred in 8 (2.3%) OAC alone vs 4 (1.2%) combination patients; all MIs in warfarin-treated patients
- Stroke/systemic embolism: 13 (3.8%) OAC alone vs 19 (5.5%) combination patients
- ISTH major bleeding: 27 (7.8%) OAC alone vs 36 (10.4%) combination (HR 0.73, P=0.22)
- Stent thrombosis very low: 2 (0.58%) in OAC alone group only
- 75.2% received warfarin, 24.8% DOACs; 85.9% aspirin, 14.5% clopidogrel as antiplatelet
- Substantial crossover: 12.2% OAC alone to combination (mainly for PCI), 9.0% combination to OAC alone (mainly for bleeding)
- Mean age 75 years, mean CHADS2 score 2.5, median 4.5 years from last PCI
Study Design
- Study Type
- Multicenter, prospective, open-label, randomized, noninferiority trial
- Randomization
- Yes
- Blinding
- Open-label for treatment allocation. Blinded independent clinical event committee for outcome adjudication. Study group assignments blinded to statistician, clinical event committee, steering committee, and sponsor.
- Sample Size
- 690
- Follow-up
- Median 2.5 years (IQR 1.8-3.4 years)
- Centers
- 111
- Countries
- Japan
Primary Outcome
Definition: Composite of all-cause death, myocardial infarction, stroke, or systemic embolism
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 47 of 346 (13.6%, annualized rate 5.5%) | 54 of 344 (15.7%, annualized rate 6.4%) | 1.16 (0.79-1.72) | P=0.20 for noninferiority, P=0.45 for superiority |
Limitations & Criticisms
- Trial prematurely terminated due to slow enrollment, enrolling only 696 of planned 2000 patients, resulting in severely underpowered sample size
- Noninferiority not established for primary endpoint, making trial inconclusive
- Open-label design may have introduced bias in patient adherence, outcome reporting, and clinical decision-making, though event adjudication was blinded
- Large noninferiority margin (HR 1.5) may not be clinically acceptable
- Noninferiority margin was redefined during the study (from 4% absolute difference to HR 1.5) to adjust for extended follow-up
- Substantial crossover between treatment groups (12.2% OAC alone to combination, 9.0% combination to OAC alone) may have diluted treatment effects
- Follow-up data not prospectively collected in standardized manner for all patients (9.3% from referring physicians, 2.9% by contacting patients)
- Japanese-specific INR targets (1.6-2.6 for age ≥70 years) differ from global standard (2.0-3.0), limiting generalizability
- Japanese rivaroxaban dose (15 mg) lower than global standard (20 mg), though blood concentrations comparable due to lower body weight
- Only 24.8% received DOACs; trial conducted when warfarin was still predominant
- Heterogeneity in antiplatelet choice (aspirin vs clopidogrel) adds variability
- All myocardial infarctions occurred in warfarin-treated patients; no MIs in DOAC patients
- Small number of stent thrombosis events (n=2, both in OAC alone group) limits conclusions about this critical outcome
- Study conducted only in Japan; generalizability to other populations uncertain
- No data on whether patients had optimal medical therapy for coronary disease
- 14 protocol violations identified but not excluded from analysis
- Time in therapeutic range significantly different between groups using post-hoc INR range of 2.0-3.0 (54.9% vs 47.9%, P=0.004), suggesting possible differences in anticoagulation intensity
Citation
Circulation. 2019;139:604–616