By Michael H. Crawford, MD, Editor

SYNOPSIS: In three subgroups of coronary artery disease patients with atrial fibrillation, apixaban plus a P2Y12 inhibitor provided superior safety and similar efficacy outcomes as treatment with warfarin, aspirin, or both for six months.

SOURCE: Windecker S, Lopes RD, Massaro T, et al. Antithrombotic therapy in patients with atrial fibrillation and acute coronary syndrome treated medically or with percutaneous coronary intervention or undergoing elective percutaneous coronary intervention: Insights from the AUGUSTUS trial. Circulation 2019;140:1921-1932.

When patients with atrial fibrillation (AF) undergo percutaneous coronary intervention (PCI), the ideal antithrombic therapy would be an oral anticoagulant and aspirin plus a P2Y12 inhibitor. However, such regimens have increased bleeding risk compared to an oral anticoagulant plus a P2Y12 inhibitor without aspirin. This so-called dual therapy approach has not been shown to increase ischemic events. Still, the optimal regimen for different subgroups of patients with AF and coronary artery disease (CAD) is unclear.

The AUGUSTUS trial investigators specified three mutually exclusive subgroups to explore the safety and efficacy of antithrombic regimens. The patient subgroups were: acute coronary syndrome (ACS) treated medically, ACS undergoing PCI, and stable CAD undergoing PCI. The antithrombic regimen comparisons used a unique 2 × 2 factorial design wherein apixaban was compared to warfarin and aspirin was compared to placebo in patients taking a P2Y12 inhibitor. The primary outcome was major or clinically significant bleeding. Secondary outcomes were death or hospitalization and the composite of death, myocardial infarction (MI), stroke, stent thrombosis, or urgent revascularization. Patients were randomized for up to 14 days after ACS or PCI and treated for six months. From 492 sites in 33 countries, 4,614 patients were randomized, 2,811 with ACS and 1,784 with elective PCI. Among ACS patients, 1,714 were treated by PCI and the rest medically. For the entire cohort, the mean CHA2DS2-VASc score was 4, and the mean HAS-BLED score was 3, with no significant differences among the three groups.

The apixaban-treated cohort experienced less major bleeding compared to the warfarin cohort in the medically treated ACS patients (hazard ratio [HR], 0.44; 95% confidence interval [CI], 0.28-0.68), PCI-treated ACS patients (HR, 0.68; 95% CI, 0.52-0.89), and elective PCI patients (HR, 0.82; 95% CI, 0.64-1.04), but with similar death and ischemic events in all three groups (P = 0.71). The authors concluded that in the three subgroups of CAD patients with AF, apixaban plus a P2Y12 inhibitor provides superior safety and similar efficacy outcomes as treatment with warfarin, aspirin, or both for six months.


The authors of this prespecified subgroup analysis of the AUGUSTUS trial used a unique 2 × 2 factorial technique to compare antithrombic regimens in patients taking a P2Y12 inhibitor: apixaban vs. warfarin and aspirin vs. placebo. They demonstrated that the apixaban plus a P2Y12 inhibitor without aspirin was safer and equally efficacious as warfarin, aspirin, or both in ACS patients managed medically and PCI patients with ACS or stable CAD. In comparing apixaban vs. warfarin, the number needed to treat to prevent one significant bleed was 16 for the ACS medical group, 23 for the ACS-PCI group, and 39 for the elective PCI group. Also, the endpoint of death or hospitalization was reduced with apixaban vs. warfarin, with no difference in death or ischemic events across all three patient groups. In comparing aspirin to placebo, the number needed to harm (NNH) with aspirin was 33, 13, and 11 in the three groups, respectively. Further, the death or hospitalization and death or ischemic events endpoints were not different on aspirin vs. placebo. In the two PCI subgroups, the endpoint of stent thrombosis or MI was numerically lower in the aspirin group. However, this was offset by the higher bleeding rate, resulting in no net clinical benefit.

Other recent trials of DOACs vs. warfarin in PCI patients, such as PIONEER AF-PCI (rivaroxaban) and RE-DUAL PCI (dabigatran), also revealed lower bleeding rates with a DOAC plus a P2Y12 inhibitor vs. warfarin, P2Y12, and aspirin therapy, without an increase in ischemic events. However, it was unclear whether the results were driven by lower doses of DOAC (dabigatran) or the omission of aspirin. This AUGUSTUS trial substudy helps clarify these issues because the authors used the recommended stroke prevention doses of apixaban and employed a unique factorial technique that allowed for comparing aspirin to placebo. Interestingly, in the lower-dose dabigatran arm of RE-DUAL PCI, patients exhibited higher rates of stent thrombosis and MI. This AUGUSTUS data cannot establish the efficacy of not using aspirin in PCI patients, and there may be high-risk subgroups of post-PCI patients in whom triple therapy for some period is warranted.

There were limitations to this analysis. Although the safety of apixaban plus a P2Y12 inhibitor therapy in the total trial population was replicated, the subgroups were underpowered for bleeding risk. Also, since patients were randomized for up to 14 days after PCI, many received aspirin during their initial management of ACS. In addition, the study ended at six months after enrollment, so the ideal therapy after six months is unknown. The strongest feature was the inclusion of medically managed ACS patients, which can be up to one-third of all ACS patients. In this group, the NNH on aspirin was 11 vs. 33 in ACS PCI patients. Thus, there may be a stronger case for aspirin in PCI patients, especially those with ACS. In addition to other trials, AUGUSTUS suggests that dual antithrombic therapy without aspirin can be considered for six months after PCI or in medically treated ACS patients with AF to reduce bleeding risk. The challenge is identifying PCI patients who are at higher risk of ischemic events and also may need aspirin therapy for some period.