ROCKET-AF: Rivaroxaban Noninferior to Warfarin in AF Patients

August 10, 2011

August 10, 2011 (Durham, North Carolina) — A major clinical trial showing that rivaroxaban (Xarelto, Bayer/Johnson & Johnson) is noninferior to dose-adjusted warfarin for the prevention of stroke or major embolism in patients with atrial fibrillation is now published online August 10, 2011 in the New England Journal of Medicine [1].

The Rivaroxaban Once Daily Oral Direct Factor Xa Inhibitor Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET-AF), first presented at the American Heart Association (AHA) 2010 Scientific Sessions, reported by heartwire at that time, met its primary end point, with investigators showing rivaroxaban was noninferior to warfarin in terms of stroke and non–central-nervous-system (CNS) embolism.

In an intention-to-treat superiority analysis, rivaroxaban was not shown to be superior to warfarin, but it fared better when investigators analyzed only patients treated with the drug in an on-treatment superiority comparison. In the on-treatment superiority analysis, rivaroxaban reduced the risk of stroke and non-CNS embolization 21% compared with warfarin, a statistically significant difference.

"When you look at all three of these analyses, I think you can say, without question, that rivaroxaban is noninferior and at least as good as warfarin," lead investigator Dr Manesh Patel (Duke Clinical Research Institute, Durham, NC) told heartwire . "For clinicians who care for patients, that's just one part of the answer, that being the prevention of stroke. When you look at bleeding risks, rivaroxaban is similar to warfarin, with an important distinction for intracranial hemorrhage and fatal bleeding, which seems to be less with rivaroxaban. And finally, for adverse events, it's a once-a-day therapeutic that doesn't require monitoring, so it might have some advantages for patients with atrial fibrillation who are thinking about taking this drug."

When ROCKET-AF was presented last November, there was some talk about how rivaroxaban stacked up against dabigatran etexilate (Pradaxa, Boehringer Ingelheim), an oral anticoagulant recently approved by the US Food and Drug Administration (FDA) for the prevention of stroke and systemic embolism in patients with atrial fibrillation.

In the 18 000-patient Randomized Evaluation of Long-Term Anticoagulant Therapy (RE-LY) study, the 150-mg dose of dabigatran was superior to warfarin on the primary end point and similar in terms of bleeding rates. Investigators, however, balked at making comparisons, saying there was no scientifically valid way to compare rivaroxaban against dabigatran. Others, including Dr Douglas Zipes (Krannert Institute of Cardiology, Indianapolis, IN), were impressed with the data, saying that even with rivaroxaban noninferior to warfarin, the agent is a step above conventional warfarin therapy.

In an editorial accompanying the published study [2], Dr Gregory del Zoppo (University of Washington School of Medicine, Seattle) and Dr Misha Eliasziw (University of Calgary, AB) write that since the intention-to-treat analysis was negative, the data do not confirm a superiority benefit of rivaroxaban over warfarin. Instead of clarifying the noninferiority/superiority question, the editorialists believe that "the multiple analyses have muddied the waters regarding rivaroxaban's efficacy and effectiveness over warfarin."

Still, the editorialists state that for the management of atrial-fibrillation patients, oral alternatives to warfarin have arrived, and their simplicity of use is attractive for patients and clinicians. The agents appear to have an efficacy similar to that of warfarin, "with the proviso that comparisons seem to depend on how easily the patient can be treated with warfarin."


The ROCKET-AF investigators enrolled 14 264 patients with nonvalvular atrial fibrillation who were at an increased risk of stroke, such as those with prior history of stroke and those with a history of hypertension. In addition, 90% of the patients had a CHADS2 score of 3 or higher, higher than scores of atrial-fibrillation patients enrolled in other antithrombotic trials. Patients were randomized to 20-mg rivaroxaban once daily (or 15 mg in patients with moderate renal impairment at screening) or to dose-adjusted warfarin (titrated to an INR of 2.5).

As reported previously, the per-protocol as-treated analysis showed rivaroxaban was noninferior to warfarin, with 1.7 events/100 patient-years in the rivaroxaban arm compared with 2.2 events/100-years in the warfarin-treated patients (p<0.001 for noninferiority). In the intention-to-treat analysis, rivaroxaban was noninferior to warfarin but failed to meet statistical significance for superiority. In looking at the as-treated population, studying all patients who received at least one dose of the drug and who were followed for events while taking the factor Xa inhibitor, rivaroxaban reduced the primary end point 21% (1.7 events/100 patient-years vs 2.2 events/100 patient-years for those taking warfarin; p=0.02 for superiority).

"When the study was first designed, nothing had beaten warfarin and reached market," said Patel. "So the first goal of ROCKET-AF was to demonstrate noninferiority, and most people would agree that in a randomized trial, noninferiority is best established when patients are taking the randomized treatment."

Like the editorialists, Patel said that since the intention-to-treat superiority analysis was negative, the data do not support a superiority benefit when rivaroxaban is compared with warfarin.

Regarding the primary safety end point, rates of major and nonmajor clinically relevant bleeding were 14.9% annually in the rivaroxaban-treated patients and 14.5% annually in the warfarin-treated patients (p=0.44). Interestingly, there was a significant 33% relative reduction in the risk of intracranial hemorrhage and a significant 50% relative reduction in the risk of fatal bleeding with rivaroxaban.

In their editorial, del Zoppo and Eliasziw point out that the reduction in intracranial hemorrhage was also observed in RE-LY, but the reason is unclear. "One possibility is the effect on a single target in the hemostatic system by the new antithrombotic agents vs the multiple targets by warfarin," they suggest.

Time in Therapeutic Range and the FDA Review Next Month

One of the concerns with ROCKET-AF is that the mean time spent in therapeutic range for warfarin patients was just 55%, which is lower than some of the other atrial-fibrillation studies. Patel noted that this might be the result of the older age of patients enrolled in the trial--the mean age was 73 years--and the fact that patients had a large number of comorbidities. In addition, the study enrolled patients from all over the world, and some clinicians in these countries might target the INR lower than doctors in the US and other Western countries.

"All that being said, we did do a quartile analysis, as some of the other studies have done, looking at the effect of rivaroxaban by quartiles of centers with their time in therapeutic range," Patel told heartwire . "The effect seems to be pretty consistent when compared against [warfarin-treated patients] in the highest quartiles of time in therapeutic range."

Rivaroxaban is scheduled to go before the FDA Cardiovascular and Renal Drugs Advisory Committee September 8, 2011. The panel will review ROCKET-AF and make recommendations on whether or not it believes rivaroxaban should be approved for the prevention of stroke and systemic embolism in patients with nonvalvular atrial fibrillation. The anticoagulant has already been approved by the FDA for prevention of deep venous thrombosis (DVT) in the setting of knee- or hip-replacement surgery.

Asked about the upcoming the FDA hearing, Patel said he has learned not to hazard a guess on the outcomes of advisory panels but added the FDA will be provided with a large, transparent data set "with robust findings for a common condition" for its review.

Top-line results from the ARISTOTLE trial, a study comparing the factor Xa inhibitor apixaban (Eliquis, Pfizer/Bristol-Myers Squibb) with warfarin in subjects with atrial fibrillation and risk factors for stroke suggest that the oral direct factor Xa inhibitor is noninferior to the older standard for the prevention of stroke and systemic embolism. Full results of that trial will be presented August 28 at the European Society of Cardiology 2011 meeting in Paris, France.

The ROCKET-AF study was sponsored by Johnson & Johnson and Bayer. Patel reports consulting fees for Ortho McNeil Janssen and Bayer and serving on an advisory board for Genzyme. Disclosures for the coauthors are listed in the paper.The editorialists report no conflicts of interest.