Large analysis looks at Medicare beneficiaries with atrial fibrillation

Treatment with rivaroxaban versus apixaban was associated with a significant increased risk of major ischemic and hemorrhagic events versus apixaban in Medicare beneficiaries ages 65 or older with atrial fibrillation, a retrospective cohort study found.

The adjusted incidence of a composite primary outcome of major ischemic or hemorrhagic events was 16.1 per 1,000 person-years for rivaroxaban, compared with 13.4 per 1,000 person-years for apixaban (HR 1.18, 95% CI 1.12-1.24), reported Wayne Ray, PhD, of Vanderbilt University in Nashville, and co-authors in JAMA.

Increased risk with rivaroxaban versus apixaban also was seen for major ischemic events (8.6 versus 7.6 per 1,000 person years respectively; HR 1.12, 95% CI 1.04-1.20), and major hemorrhagic events (7.5 versus 5.9. per 1,000 person years respectively; HR 1.26, 95% CI 1.16-1.36).

“Although previous retrospective cohort studies have examined the comparative effectiveness of rivaroxaban and apixaban in atrial fibrillation, the present study makes three distinct contributions to the evidence needed to guide clinical practice,” Ray and colleagues wrote. The study has a larger sample size, includes patients taking reduced doses of both medications (23% of both the apixaban and rivaroxaban groups), and has a primary study outcome that was “an integrated measure of the benefits and harms of anticoagulation,” the researchers noted.

“Previous studies reported both an efficacy outcome—ischemic stroke or systemic embolism—and a separate safety outcome that consisted of major bleeding, predominantly gastrointestinal events,” Ray and co-authors continued. “However, the relative clinical importance of non-fatal gastrointestinal or other extracranial bleeding versus strokes or other intracranial bleeding has been controversial.”

“In contrast, the hemorrhagic component of the primary outcome for the present study included only the most severe extracranial bleeding, that which was fatal, thus permitting a single measure of the clinical impact of anticoagulant choice.”

The researchers identified 581,451 Medicare patients with atrial fibrillation who began rivaroxaban (n=227,572) or apixaban (n=353,879) treatment between January 2013 and November 2018, and followed them through November 2018 (median follow-up 174 days).

Mean age was 77 and 50.2% were women, and the mean number of medications taken was 12. All participants had Medicare part D coverage for prescription medications and a diagnosis of atrial fibrillation/flutter in the 90 days prior to enrollment.

Prior to inverse probability of treatment weighting, people in the rivaroxaban group were younger (76.3 versus 77.4 years), less likely to be women (48.3% versus 51.4%) and had a lower prevalence of risk factors than the apixaban group. “After weighting, differences in covariate prevalence were minimal, with all standardized differences no more than 0.01,” the authors noted.

For individual components of the primary outcome, rivaroxaban versus apixaban treatment conferred higher respective risks for:

  • Ischemic stroke: adjusted rate 8.3 versus 7.2 per 1,000 person-years; HR 1.12.
  • Hemorrhagic stroke: adjusted rate 2.5 versus 1.7 per 1,000 person-years; HR 1.48.
  • Fatal extracranial bleeding: adjusted rate 1.4 versus 1.0 per 1,000 person-years; HR 1.41.

Rivaroxaban versus apixaban treatment also conferred increased risk for secondary outcomes, including:

  • Nonfatal extracranial bleeding: 39.7 versus 18.5 per 1,000 person-years; HR 2.07, including gastrointestinal sites (35.2 versus 16.3 per 1,000 person-years; HR 2.09).
  • Mortality: the unadjusted rate for total mortality for rivaroxaban was less than for apixaban but was increased after adjustment (44.2 versus 41.0 per 1,000 person-years; HR 1.06).
  • Fatal ischemic or hemorrhagic events: (4.5 versus 3.3 per 1,000 person-years; HR 1.34).

The majority of patients took standard doses of apixaban (5 mg twice daily) and rivaroxaban (20 mg once daily), but 23% of each group took reduced doses (apixaban 2.5 mg twice daily or rivaroxaban 15 mg once daily). Compared with patients receiving standard doses, those taking reduced doses were older (82.8 versus 75.2 years), more likely women (62.8% vs 46.4%), and had greater prevalence of other risk factors for stroke and bleeding.

The rivaroxaban group on reduced doses had increased risk for the primary outcome (adjusted rate 27.4 vs 21.0 per 1, 000 person-years; HR 1.28) as well as the combined ischemic components, the combined hemorrhagic components, and for the secondary outcome of nonfatal extracranial bleeding. Total mortality did not differ significantly from those receiving apixaban.

For patients receiving standard doses, rivaroxaban had increased risk for the primary outcome and the combined component major hemorrhagic events compared to apixaban, but risk for major ischemic events was not significantly different.

“These findings add to a body of literature suggesting that apixaban is associated with lower bleeding risk and possibly greater thromboembolic protection compared with rivaroxaban,” noted Peter Zimetbaum, MD, of the Beth Israel Deaconess Medical Center in Boston, and co-authors in an accompanying editorial.

“In the absence of randomized trials directly comparing direct oral anticoagulants, the choice of direct oral anticoagulant in clinical practice settings is determined by clinician preference, local practices, and insurance coverage,” they added. The present study, the wrote, “represents the largest and most contemporary evidence from clinical settings on the differential effectiveness and safety associated with apixaban and rivaroxaban.”

“Although rivaroxaban and apixaban have similar half-lives of approximately 12 hours, rivaroxaban is administered once daily while apixaban is administered twice daily,” Zimetbaum and co-authors pointed out.

“This difference in dosing schedules results in significantly higher variability in peak-to-through fluctuations with rivaroxaban,” they wrote. “Patients treated with rivaroxaban may therefore experience higher bleeding risk during high peak levels, and greater stroke risk during low trough levels.”

The researchers acknowledged that study data did not permit investigation of mechanisms to account for differential risk between medications. A substantial proportion of patients discontinued treatment, they added. Findings apply only to Medicare patients age 65 and older.

  1. Treatment with rivaroxaban versus apixaban was associated with a significant increased risk of major ischemic and hemorrhagic events versus apixaban in Medicare beneficiaries ages 65 or older with atrial fibrillation.

  2. The adjusted incidence of a composite primary outcome of major ischemic or hemorrhagic events was 16.1 per 1,000 person-years for rivaroxaban and 13.4 per 1,000 person-years for apixaban (hazard ratio [HR] 1.18, 95% CI 1.12-1.24).

Paul Smyth, MD, Contributing Writer, BreakingMED™

This study was supported by the National Heart, Lung, and Blood Institute (NHLBI).

Ray reported receiving grants from NHLBI.

Zimetbaum reported no conflicts of interest.

Cat ID: 913

Topic ID: 74,913,282,464,494,730,913,192,255,925