Relative efficacy and safety of NOACs over warfarin in AF not affected by valvular heart disease

Non–Vitamin K Antagonist Oral Anticoagulants in Patients With Atrial Fibrillation and Valvular Heart Disease

Literature - Renda G, Ricci F, Giugliano RP and Caterina R - J Am Coll Cardiol 2017: 69, March 2017 DOI: 10.1016/j.jacc.2016.12.038


Valvular heart disease (VHD) and atrial fibrillation (AF) often coexist and both increase the risk of stroke and systemic arterial embolic events (SSEE) [1]. Presence of moderate/severe mitral stenosis or a mechanical prosthetic heart valve confers a particularly high thromboembolic risk [2]. Both conditions have consistently been exclusion criteria in phase III trials comparing non-vitamin K antagonist oral antiocoagulants (NOACs) with warfarin. Current definitions of ‘valvular’ and ‘nonvalvular’ vary among studies and may be misleading.

Data of use of NOACs in patients with AF and mitral stenosis are therefore lacking. The RE-LY (dabigatran [3]), ROCKET AF (rivaroxaban [4]), ARISTOTLE (apixaban [5]) and ENGAGE AF-TIMI 48 (edoxaban [6]) did, however, include some VHD patients. Individually, these trials did not provide evidence of a differential effect of NOACs over warfarin in patients with and without VHD regarding the main efficacy and safety outcomes. These conclusions may however be limited by the limited numbers of patients enrolled per trial with various forms of VHD.

This systematic review and meta-analysis is therefore an aggregate evaluation of the relative performance of NOACs and warfarin specifically in VHD patients, based on data of all four phase III NOAC trials. 3950 patients had VHD in RE-LY, 2003 in ROCKET-AF, 4808 in ARISTOTLE and 2824 in ENGAGE AF-TIMI, amounting to a total of 13585 patients with VHD for this analysis. A total of 58098 patients were classified as without VHD in those four trials.

Main results

  • A similar rate of SSEE (RR: 1.13; 95%CI: 0.99 to 1.28) was seen in patients with VHD, as compared with those without VHD.
  • Significantly higher rates of major bleeding (RR: 1.30; 95%CI: 1.13 to 1.49) and all-cause death (RR: 1.34; 95% CI: 1.13 to 1.59) were seen in patients with VHD as compared with patients without VHD.
  • The rate of SSEE was lower in those treated with higher-dose NOACs as compared with warfarin, both in patients with VHD (RR: 0.70; 95%CI: 0.58 to 0.86) and without VHD (RR: 0.84; 95%CI: 0.75 to 0.95; test for subgroup interaction P=0.13; I2=57%).
  • The rate of major bleeding was similar with higher-dose NOACs as compared with warfarin, both in those with VHD (RR: 0.93; 95%CI: 0.68 to 1.27), or without VHD (RR: 0.85; 95%CI: 0.70 to 1.02) (test for subgroup interaction P=0.63; I2=0%); a significant statistical heterogeneity across studies was seen(Cochran’s Q P= 0.0001; I2= 78%).
  • Higher-dose NOACs reduced ICH compared with warfarin to a similar degree among AF patients with and without VHD (RR: 0.47; 95%CI: 0.24 to 0.93, and RR: 0.49; 95%CI: 0.41 to 059, respectively; test for subgroup interaction P=0.91; I2=0%).
  • When including all NOAC dose arms, largely similar results were seen to the analyses restricted to the higher NOAC doses. However, inclusion of lower doses of edoxaban and dabigatran decreased the magnitude of the risk reduction for SSEE seen with NOACs and resulted in fewer major bleeding and ICH seen with NOACs as compared with warfarin, in patients with or without VHD.


This meta-analysis shows that while patients with AF and VHD are on average at higher risk than AF patients without VHD, and had higher rates of major bleeding and all-cause death, the efficacy and safety of NOACs vs warfarin are similar between AF patients included in the phase III trials with or without VHD. Since the coexistence of VHD does not affect the overall relative efficacy and safety of NOACs in terms of SSEE and major bleeding, NOACs may be an attractive alternative to VKAs in patients with VHD, other than severe mitral stenosis or mechanical heart valves.

The newly proposed term ‘MARM-AF’, referring to Mechanical and Rheumatic Mitral valvular AF’ [7] may be useful to identify truly high-risk AF patients for whom VKAs are the anticoagulant of choice.

Editorial comment

Valvular Heart Disease Patients on Edoxaban or Warfarin in the ENGAGE AF-TIMI 48 Trial

De Caterina R, Renda G, Carnicelli AP, et al.,

J Am Coll Cardiol 2017: 69, March 2017 DOI: 10.1016/j.jacc.2016.12.031

This study was published simultaneously in J Am Coll Cardiol. It is a subanalysis of rates of SSE, major bleeding and net clinical outcomes in patients with and without VHD enrolled in the ENGAGE AF-TIMI 48 trial that evaluated edoxaban in comparison with warfarin. Subanalyses of patients with VHD in the other phase III NOAC trials had already been performed.

This paper describes that almost all treatment efficacy comparisons were consistent in patients with and without VHD. The exception was all-cause death and the composite of death or disabling stroke, which appeared numerically to be better prevented by high-dose edoxaban than by warfarin in patients without VHD. Main efficacy results of ENGAGE AF-TIMI 48 were also consistent for the low-dose edoxaban regime, both in those with VHD and those without VHD (no significantly interactions).

Thus, in a large contemporary clinical trial in patients with AF and concomitant moderate-to-severe left-sided VHD, other than mechanical valves or moderate-to-severe mitral stenosis or prior valvular surgery, the relative efficacy of edoxaban vs. warfarin was mostly maintained for total SSEE, ischaemic stroke/ systemic embolic event and bleeding, as compared to AF patients without VHD.

Editorial comment

Breithardt [8] notes that the details of the subgroup of patients with valve disease in the four large NOAC trials were unknown until now, despite approval of all four NOACs in Europe and America. The fact that ‘nonvalvular’ AF also encompassed valve disease has led to uncertainties among physicians, and consequently many physicians have been hesitant in prescribing the NOACs to certain patients, even if they were included in the official drug labelling.

According to Breithardt, “the data offer a unique opportunity to make the distinction among the patients who had mechanical prostheses or mitral stenosis (who were excluded from these trials) and “nonvalvular” atrial fibrillation with valve disease (the “paradox”), as well as those who had no valve disease (“true nonvalvular AF”). This stratification will help to align the treatment decisions associated with these distinctions and provide reassurance in treating AF patients with the types of valvular heart disease studied here.”

It should be noted that the original RCTs had different designs and different inclusion/exclusion criteria, and there was no consistent definition of valve disease across the studies, nor was there any core-laboratory-based echocardiographic assessment. The significant statistical between-trial heterogeneity observed in the analysis of major bleeding is a further limitation. Because the trials do not allow any conclusion regarding the spectrum of severity of individual valve disorders, “it seems reasonable to be cautious in prescribing NOACs to more severely ill valve patients until more robust data show that NOACs can be safely given over the whole spectrum of valve disease, regardless of its underlying severity and related comorbidities.”

Acknowledging that a robust and intuitive classification is needed, Breithardt disagrees with the categorisation as proposed by De Caterina and Camm [7] of MARM-AF, which combines the presence of a mechanical valve and AF in associated with mitral stenosis. Breithardt prefers to keep these 2 categories separate, because the only data available on the efficacy of anticoagulation in the presence of mechanical valves, suggested a benefit of warfarin over dabigatran. A future trial of NOACs in patients with mitral stenosis would give information whether these two categories should be approached differently, or whether they may be combined under the flag of MARM-AF.

In his editorial comment, Breithardt proposes a simplified terminology, thereby skipping the term ‘nonvalvular’ altogether.


1. Nkomo VT, Gottdiener JS, Scott CG, Enriquez-Sarano M. Burden of valvular heart diseases: a population- based study. Lancet 2006;368:1005–11.

2. Halperin JL, Hart RG. Atrial fibrillation and stroke: new ideas, persisting dilemmas. Stroke 1988;19:937–41.

3. Ezekowitz MD, Nagarakanti R, Noack H, et al. Comparison of dabigatran and warfarin in patients with atrial fibrillation and valvular heart disease: the RE-LY trial (Randomized Evaluation of Long-Term Antico- agulant Therapy). Circulation 2016;134:589–98.

4. Breithardt G, Baumgartner H, Berkowitz SD, et al. Clinical characteristics and outcomes with rivaroxaban vs. warfarin in patients with non- valvular atrial fibrillation but underlying native mitral and aortic valve disease participating in the ROCKET AF trial. Eur Heart J 2014;35:3377–85.

5. Avezum A, Lopes RD, Schulte PJ, et al. Apixaban in comparison with warfarin in patients with atrial fibrillation and valvular heart disease: findings from the Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARIS- TOTLE) trial. Circulation 2015;132:624–32.

6. De Caterina R, Renda G, Carnicelli A, et al. Valvular heart disease patients on edoxaban or warfarin in the ENGAGE AF-TIMI 48 trial. J Am Coll Cardiol 2017;69:1372–82.

7. De Caterina R, Camm AJ. What is “valvular” atrial fibrillation? A reappraisal. Eur Heart J 2014; 35:3328–35.

8. Breithardt G. NOACs for Stroke Prevention in Atrial Fibrillation With Valve Disease - Filling the Gaps

J Am Coll Cardiol 2017. DOI: 10.1016/j.jacc.2017.01.012

Find Renda et al online at JACCFind De Caterina et al online at JACC

Facebook Comments


We’re glad to see you’re enjoying PACE-CME…
but how about a more personalized experience?

Register for free