PARIS – Rivaroxaban monotherapy bested combination therapy with rivaroxaban and an antiplatelet agent for patients with atrial fibrillation and stable coronary artery disease, with significantly more deaths and bleeding events seen with combination therapy.
The pronounced imbalance in all-cause and cardiovascular mortality (the hazard ratio favoring rivaroxaban monotherapy was 9.72) came as a surprise, and led to early cessation of the multisite Japanese trial, lead investigator, said at the annual congress of the European Society of Cardiology.
Several previous clinical trials had studied a reduced antithrombotic regimen for patients with atrial fibrillation (AFib) after percutaneous coronary intervention (PCI), said Dr. Yasuda, professor of medicine at Tohoku University, Sendai, Japan. Current guidelines recommend triple therapy with an oral anticoagulant plus aspirin and a P2Y12 inhibitor for the shortest duration possible, with combination therapy of an anticoagulant plus a P2Y12 inhibitor for up to 12 months. Once the 1-year post-PCI mark is reached, current European and American guidelines or consensus documents recommend monotherapy with an oral anticoagulant if AFib persists and the patient has stable coronary artery disease (CAD), explained Dr. Yasuda. “However, this approach has yet to be supported by evidence from randomized, controlled trials,” he said, adding “substantial numbers of patients in this situation continue to be treated with combination therapy, which indicates a gap between guidelines and clinical practice.”
The Atrial Fibrillation and Ischemic events with Rivaroxaban in Patients With Stable Coronary Artery Disease Study (), he said, was designed to address this practice gap, randomizing 2,200 individuals to receive monotherapy with rivaroxaban or combination therapy. A total of 1,973 patients completed follow-up.
Patients were included in the randomized, open-label, parallel-group trial if they had AFib and stable CAD and were more than 1 year out from revascularization, or if they had angiographically confirmed CAD that did not need revascularization. All 294 AFIRE study sites were in Japan.
The study’s primary endpoint for efficacy was a composite of stroke, systemic embolism, myocardial infarction, unstable angina requiring revascularization, and all-cause death.
Most of the patients (79%) were male, and the mean age was 74 years. About 70% of patients in each treatment arm had received prior PCI, and 11% had undergone previous coronary artery bypass grafting (CABG).
The monotherapy arm received rivaroxaban 10 or 15 mg once daily depending on renal status. Patients in the combination therapy arm received rivaroxaban, plus a single antiplatelet drug. This could be 81 or 100 mg aspirin daily, clopidogrel at 50 or 75 mg/day, or prasugrel at 2.5 or 3.5 mg/day.
On the recommendation of the data and safety monitoring committee, the trial was terminated about 3 months early because significantly more all-cause deaths were being seen in the combination therapy group, said Dr. Yasuda. In the end, patients were treated under the study protocol for a median 23 months and followed up for a median 24.1 months.
Kaplan-Meier estimates for the first occurrence of the composite efficacy endpoint showed that monotherapy had a rate of 4.14% per patient-year, while combination therapy had a rate of occurrence for the efficacy endpoint of 5.75% per patient-year.
These figures yielded a statistically significant hazard ratio (HR) of 9.72 favoring monotherapy (P less than .001) for the prespecified noninferiority endpoint. In a post hoc analysis, rivaroxaban monotherapy achieved superiority over dual therapy (P = .02).
Breaking down the composite efficacy endpoint into its constituents, deaths by any cause and cardiovascular deaths primarily drove the difference in treatment arms. Seventy-three patients in the combo therapy arm and 41 in the rivaroxaban arm died of any cause, and the cause of death was cardiovascular for 43 combination therapy patients and 26 monotherapy patients. This yielded HRs favoring rivaroxaban of 0.55 for all-cause mortality and 0.59 for cardiovascular deaths.
Hazard ratios for individual cardiovascular events were not statistically significantly different between treatment arms, except for hemorrhagic stroke, which was seen in 13 patients receiving dual therapy and 4 receiving rivaroxaban alone, for a hazard ratio of 0.30.
Rivaroxaban monotherapy also bested dual therapy in safety: The HR was 0.59 for the incidence of a major bleed on rivaroxaban versus combination therapy, using International Society on Thrombosis and Haemostasis–established criteria for major bleeding. In the dual therapy arm, 58 individuals experienced major bleeding – the study’s primary safety endpoint – compared with 35 in the monotherapy arm, for a hazard ratio of 0.59; nonmajor bleeding occurred in 198 dual therapy patients and 121 monotherapy patients, yielding a hazard ratio of 0.58.
The Kaplan-Meier estimate for major bleeding on monotherapy was 1.62% per patient-year, compared with 2.76% per patient-year for those on combination therapy. These findings, said Dr. Yasuda, were “generally” consistent across prespecified subgroups that included participant stratification by age, sex, and bleeding risk, among others.
Dr. Yasuda acknowledged the many limitations of the trial. First, early termination introduced the possibility of overestimating the benefit of rivaroxaban monotherapy. Indeed, said Dr. Yasuda, “the reductions in rate of ischemic events and death from any cause with rivaroxaban monotherapy were unanticipated and are difficult to explain.”
Furthermore, the open-label trial design could be a source of bias and the use of both aspirin and P2Y12 inhibitors for antiplatelet therapy “makes it uncertain whether the benefit of rivaroxaban monotherapy applies equally to the two combination regimens,” said Dr. Yasuda.
Rivaroxaban dosing in AFIRE was tailored to the Japanese study population, noted Dr. Yasuda. This means that the study is not immediately generalizable to non-Asian populations, needing replication before fully closing the knowledge gap about best long-term management of patients with AFib and stable CAD in the United States and Western Europe.
However, Dr. Yasuda pointed out, serum rivaroxaban levels in Japanese patients taking the 10- or 15-mg dose are generally similar to those seen in white patients taking a 20-mg rivaroxaban dose.
Freek Verheugt, MD, of Onze Lieve Vrouwe Gasthuis Hospital, Amsterdam, was the discussant for the presentation. He raised an additional concern: “East Asian patients are poor metabolizers of clopidogrel, which may have resulted also in underestimation of bleeding.” He cautioned that the AFIRE results may not be applicable to patients on a novel anticoagulant other than clopidogrel, or on vitamin K antagonists.
In his detailed critique of the AFIRE results, Dr. Verheugt cited the OAC ALONE trial, which used a similar study design and was also conducted in Japan. For, Dr. Verheugt pointed out that “You can see ... that it was not harmful in this 700-patient study to stop aspirin therapy 1 year after an intervention.” However, he said, “the net clinical benefit is not very different, either” between treatment arms in the OAC ALONE trial. “Given the low number of patients and the low number of events, this trial was not conclusive whatsoever” he added, so AFIRE’s findings were needed.
The safety data from AFIRE, with a study population triple that of OAC ALONE, makes the safety argument for monotherapy “a very easy winner,” said Dr. Verheugt.
Dr. Verheugt was not mystified by the lower all-cause and cardiovascular death rate in the monotherapy group. “What are the mechanisms that if you stop antiplatelet therapy you have a better ischemic outcome? How come?” asked Dr. Verheugt.
“Very likely, it is the bleeding ... that you prevent if you stop antiplatelet therapy,” he said, adding that it’s known from previous studies in individuals with acute coronary syndromes and AFib that “bleeding is correlated with mortality, and that’s also proven here.”
Though Dr. Verheugt joined Dr. Yasuda in calling for replication of the results in a non-Asian population, he concurred that the AFIRE results validate current practice for anticoagulation in AFib with stable CAD. “Stopping at 1 year is safer than continuation and, most of all, it saves lives,” he said.
Full results of AFIRE were published online at the time of Dr. Yasuda’s presentation (N Engl J Med. 2019 Sep 2.).
The study was funded by the Japanese Cardiovascular Research Foundation. Dr. Yasuda reported financial relationships with Abbott, Bristol-Myers, Daiichi-Sankyo, and Takeda. Dr. Verheugt reported financial relationships with BayerHealthcare, BMS/Pfizer, Boehringer-Ingelheim, and Daiichi-Sankyo.
SOURCE: Yasuda S. et al. ESC 2019, Hot Line Session 3, .