The New England Journal of Medicine April 18, 2019

Antithrombotic Therapy after Acute Coronary Syndrome or PCI in Atrial Fibrillation (AUGUSTUS)

Renato D. Lopes, Gretchen Heizer, Ronald Aronson et al.

Bottom Line

In patients with atrial fibrillation and recent ACS or PCI, an antithrombotic regimen of apixaban and a P2Y12 inhibitor without aspirin resulted in significantly less bleeding and fewer hospitalizations compared to regimens that included a vitamin K antagonist, aspirin, or both.

Key Findings

1. Major or clinically relevant nonmajor bleeding occurred in 10.5% of the apixaban group vs. 14.7% of the vitamin K antagonist (VKA) group (HR 0.69; 95% CI, 0.58 to 0.81; P<0.001) [1.3].

2. Major or clinically relevant nonmajor bleeding occurred in 16.1% of the aspirin group compared to 9.0% of the placebo group (HR 1.89; 95% CI, 1.59 to 2.24; P<0.001).

3. Patients receiving apixaban had a significantly lower incidence of death or hospitalization than those receiving a VKA (23.5% vs. 27.4%; HR 0.83; 95% CI, 0.74 to 0.93; P=0.002).

4. The incidence of ischemic events (composite of stroke, myocardial infarction, definite or probable stent thrombosis, or urgent revascularization) was similar between the apixaban and VKA groups (6.7% vs. 7.1%).

5. Patients in the aspirin group had a similar incidence of death, hospitalization, and ischemic events compared to those in the placebo group, demonstrating no significant efficacy benefit to offsetting the increased bleeding risk.

Study Design

Design

RCT

Open-Label / Double-Blind

Sample

4,614

Patients

Duration

6 mo

Median

Setting

33 countries

Population Patients ≥18 years of age with atrial fibrillation (paroxysmal, persistent, or permanent) who had an acute coronary syndrome or underwent PCI, and planned to take a P2Y12 inhibitor for at least 6 months.

Intervention Apixaban (5 mg or 2.5 mg twice daily) vs. Vitamin K antagonist, AND Aspirin (81 mg daily) vs. Placebo, in a 2x2 factorial design. All patients received a concurrent P2Y12 inhibitor (primarily clopidogrel).

Comparator Vitamin K antagonist (target INR 2.0-3.0) and/or Placebo (depending on 2x2 randomization allocation), alongside a concurrent P2Y12 inhibitor.

Outcome Major or clinically relevant nonmajor (CRNM) bleeding according to International Society on Thrombosis and Haemostasis (ISTH) criteria.

Study Limitations

• The comparison between apixaban and VKA was open-label, which could introduce bias, although outcomes were evaluated by a blinded central adjudication committee.

• The trial was not adequately powered to detect small but clinically meaningful differences in individual rare ischemic endpoints, such as stent thrombosis.

• The time in therapeutic range (TTR) for patients in the VKA group was only 59%, which, while typical for global trials, may have slightly inflated the relative bleeding rates compared to a perfectly managed VKA cohort.

• The follow-up duration was limited to 6 months, though this covers the highest-risk period for both recurrent ischemic events and bleeding after ACS or PCI.

Clinical Significance

The AUGUSTUS trial provided definitive evidence that 'double therapy' (a direct oral anticoagulant plus a P2Y12 inhibitor) is superior to traditional 'triple therapy' (VKA plus dual antiplatelet therapy) in patients with atrial fibrillation who require anticoagulation and undergo PCI or experience ACS. By using a 2x2 factorial design, the trial clearly isolated the harms of early aspirin use, transforming international guidelines to routinely recommend dropping aspirin 1 to 4 weeks after an ACS or PCI event to minimize bleeding without compromising ischemic protection.

Historical Context

Historically, patients with atrial fibrillation requiring oral anticoagulants who also presented with ACS or underwent PCI posed a major therapeutic dilemma. To prevent stroke and stent thrombosis, guidelines traditionally recommended 'triple therapy' (VKA + Aspirin + P2Y12 inhibitor). However, triple therapy was associated with unacceptably high rates of major bleeding. The WOEST trial (2012) first suggested that dropping aspirin reduced bleeding without increasing ischemic events. Subsequent DOAC trials—PIONEER AF-PCI (rivaroxaban) and RE-DUAL PCI (dabigatran)—reinforced the safety of DOAC-based double therapy. The AUGUSTUS trial, with its unique 2x2 factorial design, definitively isolated the independent effects of the anticoagulant choice (apixaban vs. VKA) and the antiplatelet choice (aspirin vs. placebo), providing the most robust evidence to date against the routine use of prolonged triple therapy.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

In the context of the AUGUSTUS trial, patients with atrial fibrillation and a recent PCI required both anticoagulation and antiplatelet therapy. What are the distinct pharmacological targets of apixaban, clopidogrel, and aspirin, and how does understanding these mechanisms explain the synergistic increase in bleeding risk when all three are combined?

Key Response

Apixaban reversibly inhibits Factor Xa in the coagulation cascade, clopidogrel blocks the P2Y12 ADP receptor to prevent platelet activation, and aspirin irreversibly inhibits COX-1 to reduce thromboxane A2 production. Understanding these distinct but overlapping hemostatic mechanisms helps students grasp why 'triple therapy' profoundly disrupts normal hemostasis and exponentially increases bleeding risk compared to dual therapy.

Resident

Based on the findings of the AUGUSTUS trial, how should you manage the discharge antithrombotic regimen of a standard-risk patient with atrial fibrillation who just underwent a percutaneous coronary intervention (PCI) with a drug-eluting stent?

Key Response

The trial demonstrated that an antithrombotic regimen of apixaban plus a P2Y12 inhibitor (like clopidogrel) without aspirin resulted in significantly less bleeding and fewer hospitalizations compared to regimens including a vitamin K antagonist or aspirin. Residents should recognize that dropping aspirin early (often at discharge) and maintaining 'dual antithrombotic therapy' is now the standard of care for most of these patients.

Fellow

While AUGUSTUS showed a clear bleeding benefit for omitting aspirin, there was a numerical, non-significant increase in stent thrombosis in the placebo arm. Given that the trial was powered for safety rather than rare ischemic events, how should this influence your decision to drop aspirin in a patient with highly complex coronary anatomy, such as a left main bifurcation stent?

Key Response

Because the trial was not sufficiently powered to definitively rule out a small increase in rare but catastrophic ischemic events like stent thrombosis, fellows must learn to individualize care. In patients with exceptionally high thrombotic risk (e.g., complex bifurcation or left main stenting), a short, tailored course of triple therapy (1 to 4 weeks) may still be justified before stepping down to dual therapy.

Attending

The AUGUSTUS trial shifted the paradigm by demonstrating that reducing bleeding through the omission of aspirin also significantly reduced hospitalizations, even without a mortality benefit. In your clinical practice and teaching, how do you frame the trade-off between the highly visible, common risk of bleeding and the rare, statistically underpowered, yet clinically devastating risk of stent thrombosis when counseling patients and junior staff?

Key Response

Attendings must navigate the nuances of shared decision-making and risk stratification. The rationale highlights the teaching point that while the default is now to drop aspirin to prevent hospital readmissions for bleeding, the 'art of medicine' involves identifying the outlier patient whose ischemic risk warrants deviating from the trial's broad population average, emphasizing customized, patient-centered risk calculation.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The AUGUSTUS trial employed a 2x2 factorial design to evaluate apixaban versus VKA and aspirin versus placebo simultaneously. What are the key statistical assumptions regarding interaction terms in this design, and how would a significant synergistic interaction between the anticoagulant and antiplatelet therapies complicate the interpretation of the main effects?

Key Response

A 2x2 factorial design assumes there is no significant interaction between the interventions, allowing for independent analysis of the margins (main effects). If apixaban's safety benefit heavily depended on the presence or absence of aspirin, interpreting the main effect of apixaban across both aspirin strata combined would be statistically misleading. Researchers must evaluate the interaction P-value to validate the independent conclusions.

Journal Editor

In the AUGUSTUS trial, the comparison between aspirin and placebo was double-blinded, but the comparison between apixaban and a vitamin K antagonist was open-label. As a peer reviewer evaluating this methodology, what specific biases does this introduce for the primary endpoint of 'clinically relevant nonmajor bleeding,' and what safeguards would you demand to ensure the validity of these results?

Key Response

Open-label designs are inherently susceptible to ascertainment and reporting biases, especially for subjective endpoints like 'clinically relevant nonmajor bleeding.' An editor would look for robust methodological safeguards, such as a blinded central endpoint adjudication committee (a PROBE design) and strict, objective, laboratory-driven criteria for bleeding events (e.g., ISTH criteria involving specific hemoglobin drops) to mitigate this threat to validity.

Guideline Committee

How do the results of the AUGUSTUS trial, alongside PIONEER AF-PCI and RE-DUAL PCI, provide the evidentiary basis for the current ACC/AHA and ESC guideline recommendations regarding the transition from triple to dual antithrombotic therapy in patients with atrial fibrillation undergoing PCI?

Key Response

AUGUSTUS provided definitive evidence that a NOAC-based dual therapy regimen (apixaban + P2Y12 inhibitor) is superior in safety to VKA-based triple therapy. This directly informs the Class I recommendation in modern guidelines (e.g., 2020 ESC guidelines on AFib or 2021 ACC/AHA/SCAI guidelines on coronary revascularization) to discontinue aspirin early (typically up to 1 week post-PCI) and continue a NOAC plus a P2Y12 inhibitor to minimize bleeding without an unacceptable increase in ischemic events.

Clinical Landscape

Noteworthy Related Trials

2013

WOEST Trial

n = 573 · Lancet

Tested

Dual therapy (Warfarin plus Clopidogrel)

Population

Patients receiving oral anticoagulants who undergo PCI

Comparator

Triple therapy (Warfarin plus Clopidogrel plus Aspirin)

Endpoint

Any bleeding episode within 1 year

Key result: Omitting aspirin and using dual therapy resulted in significantly less bleeding without increasing the rate of thrombotic events.

2016

PIONEER AF-PCI

n = 2,124 · NEJM

Tested

Rivaroxaban plus P2Y12 inhibitor

Population

Patients with nonvalvular atrial fibrillation undergoing PCI with stenting

Comparator

Standard triple therapy (VKA plus DAPT)

Endpoint

Clinically significant bleeding

Key result: Rivaroxaban-based dual therapy strategies significantly reduced the rates of clinically significant bleeding compared to standard triple therapy.

2017

RE-DUAL PCI

n = 2,725 · NEJM

Tested

Dabigatran plus P2Y12 inhibitor

Population

Patients with atrial fibrillation undergoing PCI

Comparator

Standard triple therapy (Warfarin plus Aspirin plus P2Y12 inhibitor)

Endpoint

Major or clinically relevant nonmajor bleeding

Key result: Dual therapy with dabigatran significantly reduced the risk of bleeding compared to triple therapy with warfarin, with noninferiority for thromboembolic events.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis