For atrial fibrillation with ACS or PCI, apixaban plus clopidogrel (without aspirin) is preferred (AUGUSTUS)

Reference

Lopes RD, Heizer G, Aronson R, et al, for the AUGUSTUS Investigators. Antithrombotic therapy for acute coronary syndrome or PCI in atrial fibrillation. N Engl J Med 2019;380(16):1509-1524.

Synopsis

These researchers identified adults with atrial fibrillation of any type who required long-term oral anticoagulation, had experienced either an acute coronary syndrome or PCI in the previous 14 days, and were planning to take a P2Y12 inhibitor (usually clopidogrel) for at least 6 months. The median age of the included patients was 71 years, the mean CHA2DS2-VASc score was 3.9 (adjusted stroke risk ~4% per year), the mean HAS-BLED score was 2.9 (increased risk for major bleeding); 29% were women. This was a Latin square design, with 4 study arms making 2 simultaneous comparisons (apixaban vs a vitamin K antagonist [VKA] such as warfarin, and aspirin vs placebo). Thus, patients were randomized into 1 of 4 groups: (1) apixaban plus aspirin, apixaban plus placebo, VKA plus aspirin, and VKA plus placebo. Recall that all patients were also taking a P2Y12 inhibitor such as clopidogrel or ticagrelor. Patients assigned to apixaban received a 5-mg dose twice daily, unless they had at least 2 of the following: 80 years or older, weight less than 60 kg, or a serum creatinine level at least 1.5 mg/dL, in which case they were given 2.5 mg twice daily. The target INR for the VKA was 2.0 to 3.0, and aspirin was given in an 81-mg daily dose. Patients were treated for 6 months, at which time they returned to usual care by their personal physician, and final outcomes were adjudicated at 7 months. Although the aspirin was double-blinded, warfarin and apixaban use were open label. Groups were balanced at the beginning of the study, and analysis was by modified intention to treat (each patient who received at least one dose of study drug was included). A total of 4683 patients entered the trial, and 88% were still enrolled at the end of the trial. Most losses to follow-up were due to death, and the number of deaths was similar between groups. INR control was good, with only 26% of measurements outside the range of 2.0 to 3.0, with 3% above and 23% below. The primary outcome was major bleeding (in a critical organ, associated with a 2g decrease in hemoglobin, or causing death) or nonmajor but clinically significant bleeding (requiring hospitalization, medical or surgical intervention, a clinic visit, or a change in therapy). For the comparison of apixaban with VKA, the risk of major bleeding (6.7 vs 10.5 events/100 person-years) and clinically relevant nonmajor bleeding (18.2 vs 26.1 events/100 person-years) were both significantly lower. This corresponds to a number needed to treat [NNT] to avoid one major bleed of 26, and an NNT to prevent one nonmajor clinical bleed of 17, both favoring apixaban. Major bleeding (11.1 vs 6.5 events per 100 person-years) and nonmajor clinical bleeding (30.0 vs 15.2 per 100 person-years) were both significantly more common with aspirin than with placebo (number needed to treat to harm = 22 and 7, respectively). Overall, the lowest bleeding rates were for the apixaban plus placebo group. Regarding vascular outcomes, there was a small reduction in stroke risk with apixaban vs VKA (1.2 vs 2.4 events per 100 person-years; NNT = 83 per year), but no difference between groups with regard to all-cause mortality, cardiovascular mortality, stent thrombosis, myocardial infarction, or the need for urgent revascularization for either of the comparisons (apixaban vs VKA or aspirin vs placebo). The only other difference between groups was a lower risk of hospitalization for the apixaban group (54.8 vs 66.5 events per 100 person-years; NNT = 8).