By Michael Crawford, MD, Editor
SOURCES: Philippart R, et al. Prognostic value of CHA2DS2-VASc score in patients with ‘non-valvular atrial fibrillation’ and valvular heart disease: The Loire Valley Atrial Fibrillation Project. Eur Heart J 2015;36:1822-1830.
Breithardt G, Baumgartner H. Valvular heart disease among non-valvular atrial fibrillation: A misnomer, in search of a new term. Eur Heart J 2015;36:1794-1797.
The CHA2DS2-VASc score (CVS) for the prediction of stroke and other thromboembolism risk in patients with atrial fibrillation (AF) has been validated in patients with non-valvular AF (see Table 1). AF patients with mitral stenosis or a prosthetic left heart valve are known to have a high risk of thromboembolism, and vitamin K antagonists are recommended for them regardless of their CVS.
However, little data are available on how to manage AF patients with native, non-rheumatic valve disease. Thus, investigators from France and the United Kingdom tested the hypothesis that the CVS would work well in such patients. The hypothesis was tested in the echocardiography database of a large hospital in Tours, France, to identify 8053 AF patients without valve disease (n = 6851) and those with either aortic stenosis (AS) or regurgitation (AR) and mitral regurgitation (MR) (n = 1202) between 2000 and 2010.
Thromboembolic events were identified after a mean follow-up of 868 days in 627 patients. The AF patients with valve disease (61% MR, 24% AR, 32% AS) had a higher risk of events (hazard ratio [HR], 1.39; 95% confidence interval, 1.14-1.69; P = 0.001), even after adjustment for anticoagulant and antiplatelet use. The severity of valve disease was not associated with more events, but patients with aortic valve disease had higher event rates than those with MR. The event rate per year increased with increasing CVS in those with and without valve disease, and the predictive value of the CVS was the same in both groups.
Comparing a CVS of 0-1 to 2-3, the event rate/year not on anticoagulants increased from 1.62% to 6.19% in AF patients without valve disease and from 1.90% to 5.98% in those with valve disease. The authors concluded that in “non-valvular AF” patients with (no mitral stenosis or valve prostheses) the presence of left ventricular valve disease increased the risk of thromboembolic events and that this result correlated with higher CVS scores.
COMMENTARY
Almost all the research on anticoagulants and AF has been in so-called “non-valvular AF.” It is now clear that this terminology was imprecise. What was really meant was “AF patients with no other high-risk condition for thromboemboli,” such as rheumatic mitral stenosis and prosthetic valves. The latter patients had clear indications for anticoagulation if they developed AF. Patients with non-rheumatic mitral valve disease and patients with aortic valve disease fell into a gray zone in which there wasn’t much data. Thus, this retrospective observational study focusing on AF patients with and without gray zone valve disease is of interest. In their traditional non-valvular AF patients, 22% had left heart valve disease (LHVD), 60% of which was non-rheumatic MR. This LHVD subgroup of the “non-valvular” AF patients had a higher incidence of thromboembolic events. However, they were older and had more comorbidities than the group with absolutely no valve disease, which was reflected in higher CVSs. On multivariate analysis, only age, female sex, and the CVS were predictive of events. Of course, the first two are included in the CVS. In fact, in this study the CVS had the same predictive value in both groups, which suggests that it can be used for all AF patients without rheumatic mitral stenosis or a prosthetic valve.
These distinctions are important for choosing an oral anticoagulant. We know that the new oral anticoagulants (NOAC) are at least as good as vitamin K antagonists (VKA) in “non-valvular AF,” but not in prosthetic valves. There are no comparative data in mitral stenosis, but most would favor VKA in these patients. Can we use NOACs in the AF patients with non-rheumatic, non-prosthetic LHVD? This study suggests that if we use the CVS we can. Other studies comparing the NOACs to VKAs have included some LHVD patients (e.g., ROCKET AF), and the results were similar. Thus, the available data suggest NOACs are adequate in LHVD patients with AF.
Clearly, the term “non-valvular AF” is confusing. The American College of Cardiology/American Heart Association/European Society of Cardiology guidelines define “non-valvular AF” as the absence of rheumatic disease, mitral stenosis, prosthetic valves, or valve repair. That leaves a lot of LHVD on the table. Perhaps we need new terminology in this area as the editorial accompanying this article suggests. The immediate practical conclusion of this study is that the CVS score can be used to determine who needs oral anticoagulant therapy in “non-valvular AF” patients with LHVD with confidence.
