Efficacy and Safety of Rivaroxaban in Patients with Heart Failure and Non-Valvular Atrial Fibrillation: Insights from ROCKET AF


Rivaroxaban is a direct factor Xa inhibitor that is more consistent and predictable than warfarin.1 ROCKET-AF was an international, multicenter, randomized, double-blind, study designed to assess the noninferiority of rivaroxaban to warfarin for the prevention of stroke and systemic embolic events in patients with non-valvular AF.2 The trial showed noninferiority of rivaroxaban with a significant reduction in intracranial hemorrhage (ICH).2

Heart failure (HF) is an additional risk factor for thromboembolic events in patients with AF3 and those with HF, as compared to those without HF, on vitamin K antagonists spend even less time on average in the therapeutic range and thus may have increased risk of bleeding and/or thromboembolism.4 The authors of "Efficacy and Safety of Rivaroxaban in Patients with Heart Failure and Non-Valvular Atrial Fibrillation: Insights from ROCKET-AF" postulated that the predictability of rivaroxaban made it a useful alternative to warfarin in patients with AF and HF. Since 62.5% of patients in the ROCKET AF trial had a history of HF or a low ejection fraction (EF), this large subgroup (n=9033) allowed for exploration of the efficacy and safety of rivaroxaban in AF patients with HF.


ROCKET-AF was a non-inferiority trial that compared rivaroxaban 20 mg daily (or 15 mg daily in patients with creatinine clearance 30-49 mL/min) with adjusted-dose warfarin (target international normalized ratio [INR] 2.5, range 2.0-3.0) in patients with non-valvular AF. Patients had to have electrocardiographic evidence of AF and a moderate to high risk of AF (defined as history of stroke, transient ischemic attack, or systemic embolism, or CHADS2 risk score of at least 2). The patients of interest in this subgroup analysis were those with a history of HF or a left ventricular EF <40% on any imaging modality. The efficacy of rivaroxaban versus warfarin was also assessed in patients classified by EF (≥ or <40%), New York Heart Association (NYHA) class, HF with preserved and reduced EF, CHADS2 score, and device (implantable cardioverter defibrillator (ICD) or biventricular-ICD [BiV-ICD]).

The primary efficacy outcome was all stroke or systemic embolism. Secondary efficacy outcomes included death from all-causes, myocardial infarction, and the composite (and individual components) of stroke, systemic embolism, or vascular death. Intention-to-treat (ITT) analysis was used for all efficacy outcomes. Endpoints were measured until the time of site notification of study termination. The primary safety endpoint was major or non-major clinically relevant bleeding with secondary safety endpoints of ICH and hemorrhagic stroke.

Cox proportional hazards models were used to assess the association with risk of outcomes for (1) patients with vs. patients without HF; (2) rivaroxaban vs. warfarin within subgroups defined by HF; and (3) rivaroxaban vs. warfarin within subgroups of HF patients defined by EF, NYHA class, and CHADS2 score.


Patients with HF were higher risk (Table 1), although on average two years younger, more likely to have persistent AF, and had higher mean CHADS2 scores. Other characteristics are summarized in Table 1.

Table 1: Summary of Patient Characteristics.

Heart Failure

Heart Failure

(N =4530)


(N =2551)







Persistent AF, n (%)

3789 (83.6%)

3708 (82.3%)

1965 (77.0%)

2023 (78.2%)

CHADS2 Score, mean (SD)

3.7 (0.9)

3.7 (0.9)

3.2 (0.9)

3.1 (0.8)

Ejection Fraction <40%, n (%)

1043 (33.3%)

1102 (34.5%)

0 (0.0%)

0 (0.0%)

Concurrent aspirin use, n (%)

1373 (30.3%)

1428 (31.7%)

657 (25.8%)

640 (24.7%)

Overall, patients with HF had a higher risk of the composite of stroke, systemic embolization, or vascular death, all-cause death and vascular death compared to patients without HF (Table 2). The rates of stroke or systemic embolization per 100 patient-years with rivaroxaban vs warfarin were similar among patients with HF (1.90 vs. 2.09; adjusted HR 0.91: 95% CI, 0.74-1.13) and without HF (2.10 vs. 2.54; adjusted HR 0.84: 95% CI, 0.65-1.09, p=0.62). Safety endpoints also were similar with rivaroxaban vs. warfarin among patients with HF and among patients without HF.

Table 2: Summary of Safety and Efficacy Outcomes, Heart Failure vs. No Heart Failure (events per 100 patient-years, (total events))


Heart Failure

No Heart Failure

Heart Failure vs. No Heart Failure HR (95% CI)

p value

Efficacy Outcomes

Stroke or Systemic Embolization

1.99 (343)

2.32 (232)

0.94 (0.78, 1.13)


Stroke, Systemic Embolization, or Vascular Death

5.00 (835)

3.50 (346)

1.28 (1.11, 1.47)


All-Cause Death

5.26 (879)

3.37 (335)

1.34 (1.17, 1.55)


Vascular death

3.53 (600)

1.75 (176)

1.65 (1.37, 1.98)


Safety Outcomes

Major or non-major clinically relevant bleeding

14.12 (1766)

15.73 (1158)

1.00 (0.92, 1.08)


Hemorrhagic Stroke

0.29 (41)

0.45 (38)

0.73 (0.45, 1.2)


Intracranial Hemorrhage

0.53 (74)

0.77 (65)

0.84 (0.58, 1.22)


The reduction in hemorrhagic stroke with rivaroxaban noted in the overall trial was also seen in patients with HF (adjusted HR 0.38: 95% CI 0.19-0.76, p=0.067). There was little difference in efficacy or safety amongst other subgroups, except for an increased risk of bleeding in those treated with rivaroxaban with ICDs or BiV-ICDs (Table 3).

Table 3: Efficacy and Safety Endpoints Amongst Selected Heart Failure Subgroups (reported as events per 100 patient-years (total events)).

Stroke or non-CNS embolism

Major or Non-Major Clinically Relevant Bleeding




Rivaroxaban vs. Warfarin HR (95% CI)

p value



Rivaroxaban vs. Warfarin HR (95% CI)

p value

NYHA Class


I or II

1.90 (111)

2.02 (118)

0.94 (0.73, 1.22)


14.83 (631)

14.15 (610)

1.08 (0.97, 1.21)



1.88 (49)

2.10 (54)

0.90 (0.61, 1.32)

12.45 (232)

13.54 (249)

0.96 (0.80, 1.15)

HF Classification


Preserved EF

2.00 (93)

2.06 (95)

0.98 (0.74, 1.31)


14.18 (487)

14.81 (504)

1.00 (0.88, 1.13)


Reduced EF

1.27 (29)

1.70 (39)

0.75 (0.46, 1.22)

15.05 (241)

13.89 (227)

1.14 (0.95, 1.37)

CHADS2 Score



1.30 (10)

1.16 (9)

1.09 (0.44, 2.69)


15.96 (83)

10.02 (56)

1.54 (1.10, 2.16)



1.96 (154)

2.18 (170)

0.90 (0.72, 1.12)

14.06 (805)

14.42 (822)

1.02 (0.92, 1.12)

Device Therapy


No Device

1.96 (147)

2.08 (156)

0.94 (0.75, 1.18)


13.08 (722)

13.72 (753)

0.99 (0.89, 1.09)



0.33 (1)

1.96 (6)

0.17 (0.02, 1.39)

32.43 (56)

16.37 (35)

2.00 (1.31, 3.05)


The findings comparing rivaroxaban with warfarin among patient with HF were similar to the overall trial findings. Thus, the authors concluded that rivaroxaban should be considered as an alternative to warfarin in patients with AF and HF.


AF occurs in 12-41% of those with HF and correlates with HF severity.5 Management of these patients can be difficult as patients with HF have more variability in time spent in the targeted therapeutic range of warfarin treatment.4 Thus, use of a predictable anti-coagulation agent in this population might be of even greater benefit than the average patient with AF. ROCKET-AF included a large cohort with HF at the time of randomization, unlike prior studies with direct factor Xa inhibitor apixaban, ARISTOTLE and RE-LY.2,6,7 This allowed for a unique opportunity to assess the efficacy and safety of rivaroxaban in these patients.

This study of HF patients in the ROCKET-AF trial was the first major analysis of direct factor Xa inhibitors in heart failure patients. The results showed little difference between the efficacy and safety of rivaroxaban compared to warfarin in patients with HF as compared to patients without HF. Thus, the authors concluded that rivaroxaban is an appropriate alternative to warfarin in patients with both AF and HF.

This study raises several important questions, however. While the number of patients with HF in the ROCKET-AF trial was indeed quite high (n=9033), the analysis was still that of an underpowered subgroup, and thus the neutral results should be cautiously interpreted. There were no statistically significant differences between treatments either in patients with or without HF. These results could indeed imply that rivaroxaban performed similarly in those with and without HF compared to warfarin, but without an initial trial design structured to identify differences in these populations, it is difficult to declare the lack of separation between the outcomes as a true representation of the efficacy and safety of rivaroxaban in this population. Thus, these findings should be considered hypothesis-generating.

Since patients enrolled in ROCKET-AF had a moderate to high risk of embolism, with the vast majority having a CHADS2 score ≥ 3, this raises the question of how efficacy and safety is balanced in those with a low to moderate risk of embolization and in particular, those with AF and HF who otherwise have low risk of embolization.

The results of this analysis suggest that rivaroxaban is indeed a possible alternative to warfarin in patients with both AF and HF. The authors have laid crucial groundwork for further investigation of the safety and efficacy of rivaroxaban in patients with both AF and HF.


  1. Kubitza, D., Becka, M., Wensing, G., Voith, B. & Zuehlsdorf, M. Safety, pharmacodynamics, and pharmacokinetics of BAY 59-7939—an oral, direct Factor Xa inhibitor—after multiple dosing in healthy male subjects. Eur J Clin Pharmacol 2005; 61:873–80.
  2. Patel, M. R. et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011; 365:883–91.
  3. Predictors of thromboembolism in atrial fibrillation: I. Clinical features of patients at risk. The Stroke Prevention in Atrial Fibrillation Investigators. Ann Intern Med1992; 116: 1–5.
  4. Rose, a J. et al. Patient characteristics associated with oral anticoagulation control: results of the Veterans AffaiRs Study to Improve Anticoagulation (VARIA).J Thromb Haemost 2010; 8: 2182–91.
  5. Maisel, W. H. & Stevenson, L. W. Atrial fibrillation in heart failure: epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol 2003; 91: 2D–8D.
  6. Connolly, S. J. et al. Apixaban in patients with atrial fibrillation. N Engl J Med 2011; 364:806–17.
  7. Granger, C. B. et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2011;365: 981–92.

< Back to Listings