Summary
Atrial fibrillation (AF) confers a substantial risk of stroke. Recent trials comparing vitamin K antagonists (VKAs) with non-vitamin K antagonist oral anticoagulants (NOACs) in AF were performed among patients with so-called “non-valvular” AF. The distinction between “valvular” and “non-valvular” AF remains a matter of debate. Currently, “valvular AF” refers to patients with mitral stenosis or artificial heart valves (and valve repair in North American guidelines only), and should be treated with VKAs. Valvular heart diseases, such as mitral regurgitation, aortic stenosis (AS) and aortic insufficiency, do not result in conditions of low flow in the left atrium, and do not apparently increase the risk of thromboembolism brought by AF . Post-hoc analyses suggest that these conditions probably do not make the thromboembolic risk less responsive to NOACs compared with most forms of “non-valvular” AF. The pathogenesis of thrombosis is probably different for blood coming into contact with a mechanical prosthetic valve compared with what occurs in most other forms of AF. This may explain the results of the only trial performed with a NOAC in patients with a mechanical prosthetic valve (only a few of whom had AF), where warfarin was more effective and safer than dabigatran. By contrast, AF in the presence of a bioprosthetic heart valve or after valve repair appears to have a risk of thromboembolism that is not markedly different from other forms of “non-valvular” AF. Obviously, we should no longer consider the classification of AF as “valvular” (or not) for the purpose of defining the aetiology of the arrhythmia, but for the determination of a different risk of thromboembolic events and the need for a specific antithrombotic strategy. As long as there is no better new term or widely accepted definition, “valvular AF” refers to patients with mitral stenosis or artificial heart valves. Patients with “non-valvular AF” may have other types of valvular heart disease. One should emphasize that “non-valvular AF” does not exclude patients with some types of valvular heart disease from therapy with NOACs.
Résumé
La fibrillation atriale (FA) est associée à un risque d’accident vasculaire cérébral. Les essais récents comparant antivitamines K (AVK) et anticoagulants oraux directs non-AVK (NACO) dans la FA ont été réalisés chez des patients avec FA, dite « non valvulaire ». La différence entre FA « valvulaire » et « non valvulaire » reste néanmoins sujet à controverses. Actuellement, la FA « valvulaire » fait référence aux patients avec rétrécissement mitral ou prothèse valvulaire cardiaque (ainsi qu’aux plasties mitrales dans les recommandations nord américaine) et nécessite un traitement par AVK. Les valvulopathies, telles que l’insuffisance mitrale, le rétrécissement aortique ou l’insuffisance aortique ne sont pas associées à un bas débit dans l’oreillette gauche et semblent ne pas augmenter le risque thromboembolique lié à la FA . Des analyses post-hoc suggèrent que le bénéfice des NACO dans ces pathologies n’est pas différent de celui constaté pour les autres patients avec FA « non valvulaire ». Le processus de thrombogénicité est probablement différent des autres formes de FA lorsque le sang entre en contact avec du matériel de prothèse valvulaire mécanique. Ceci pourrait expliquer les résultats négatifs du seul essai réalisé à ce jour avec un NACO chez des patients avec prothèses mécaniques (dont seulement une minorité avaient de la FA) et où la warfarine à été plus efficace et plus sûre que le dabigatran. En revanche, la FA chez des patients avec une bioprothèse ou une plastie valvulaire semble associée à un risque thromboembolique assez similaire à celui attendu pour une FA « non valvulaire ». Manifestement, il ne faut plus envisager la classification d’une FA comme « valvulaire » (ou « non valvulaire ») pour définir l’étiologie de l’arythmie, mais dans le but de déterminer un risque différent d’évènements thromboemboliques et pour établir un traitement antithrombotique spécifique. Tant qu’il n’y aura pas de meilleur terme ou de nouvelle définition largement acceptée, la FA « valvulaire » fait référence aux patients avec rétrécissement mitral ou avec prothèse valvulaire (mécanique ou biologique). Les patients avec FA « non valvulaire » peuvent avoir d’autres types de valvulopathies. Il faut donc insister sur le fait que la définition de FA « non valvulaire » n’exclut pas certains patients avec valvulopathies des possibilités de traitement par NACO.
Background
Atrial fibrillation (AF) is the most common arrhythmia, and confers a substantial risk of stroke. In the absence of anticoagulation, thromboembolic risk ranges from < 1% per year – similar to the background risk of the age-matched population – to > 20% per year. The risk of stroke and systemic embolism in AF may be assessed by simple clinical risk factors and scoring systems . This has led to the wide use of oral anticoagulation as a preventive strategy for most patients with AF, unless clearly at very low risk . The recent availability of non-vitamin K antagonist oral anticoagulants (NOACs) is likely to increase the number of AF patients efficiently treated for stroke prevention. Recent trials comparing vitamin K antagonists (VKAs) with NOACs in AF were performed among patients with so-called “non-valvular” AF, and excluded patients at high risk of thromboembolism, such as those with AF accompanying mitral stenosis or with mechanical prosthetic valves. Beyond the higher risk of stroke and ethical issues in the clinical development of NOACs, a reason for excluding these patients in trials testing NOACs was the possibility that the pathogenesis of thromboembolism may be substantially different from that in other AF patients. The distinction between “valvular” AF and “non-valvular” AF still remains a matter of debate, however, with different designations adopted in the literature.
We discuss the definitions of the terms “valvular” and “non-valvular” AF in different trials with NOACs and in current guidelines. We also review the thromboembolic risk associated with AF in the presence of the various valve diseases, and the qualitative type of possible thrombus in such conditions. All of these factors may have implications for clinical practice and future investigations.
Differing definitions of “valvular atrial fibrillation”
Trials of thromboprophylaxis in atrial fibrillation
The issue of “valvular AF” definition is relevant because most of these patients were excluded from recent trials testing NOACs in patients with AF . Consequently, NOACs have been registered and are currently indicated only for patients with so-called “non-valvular AF”. The reasons for excluding patients with “valvular AF” included uncertainties about whether the mechanism of thrombogenesis in such patients is similar to that occurring in the more common forms of “non-valvular” AF and, consequently, whether a similar anticoagulation strategy is appropriate. The criteria for excluding such patients were, however, variable ( Table 1 ). In the RE-LY trial testing dabigatran versus warfarin, “history of heart valve disorders” was an exclusion criterion, and the disorders were defined as prosthetic valve or haemodynamically relevant valve disease, resulting in the exclusion of patients with AF and severe mitral or aortic insufficiency or severe AS .
| Study drug | Study acronym/name | Year of publication | Atrial fibrillation exclusion criteria related to valve disease |
|---|---|---|---|
| Apixaban | AVERROES | 2011 | Valvular disease requiring surgery, prosthetic mechanical heart valve |
| Apixaban | ARISTOTLE | 2011 | Clinically significant (moderate or severe) mitral stenosis, prosthetic mechanical heart valve |
| Apixaban | ARISTOTLE-J | 2011 | Valvular heart disease |
| Betrixaban | EXPLORE-Xa | 2013 | Prosthetic mechanical heart valve |
| Dabigatran | PETRO | 2007 | Mitral stenosis, prosthetic valves |
| Dabigatran | RE-LY | 2009 | History of heart valve disorder (including haemodynamically relevant valve disease and prosthetic valve) |
| Edoxaban | Edoxaban phase II study | 2012 | Comorbid rheumatic valvular disease, history of valvular surgery, infective endocarditis |
| Edoxaban | ENGAGE-AF-TIMI 48 | 2013 | Moderate or severe mitral stenosis, unresected atrial myxoma, mechanical heart valve |
| Rivaroxaban | ROCKET-AF | 2011 | Haemodynamically significant mitral valve stenosis, prosthetic heart valve |
| Rivaroxaban | J-ROCKET-AF | 2012 | Haemodynamically significant mitral valve stenosis, prosthetic heart valve |
| Ximelagatran | SPORTIF III | 2003 | Mitral stenosis, previous valvular heart surgery, active infective endocarditis |
| Ximelagatran | SPORTIF V | 2005 | Mitral stenosis, previous valvular heart surgery, active infective endocarditis |
The ROCKET-AF trial, evaluating rivaroxaban against warfarin, excluded only haemodynamically significant mitral valve stenosis and prosthetic heart valves, but permitted the inclusion of patients with other diseases in native valves, as well as patients treated with annuloplasty, commisurotomy or valvuloplasty .
In ARISTOTLE, evaluating apixaban versus warfarin, patients with “clinically significant (moderate or severe) mitral stenosis”, as well as “onditions other than AF that require chronic anticoagulation (e.g. prosthetic mechanical heart valve)” were excluded, therefore allowing the inclusion of patients with native valvular heart disease other than mitral stenosis and bioprosthetic heart valves .
In the AVERROES trial, evaluating apixaban versus aspirin in patients considered “unsuitable” for VKAs, “valvular disease requiring surgery” was among the exclusion criteria and “unsuitability” for VKAs thus excluded patients with mechanical prosthetic valves .
In the last trial, the ENGAGE-AF study, which tested two strategies of edoxaban versus warfarin, patients with “moderate or severe mitral stenosis or a mechanical heart valve” were excluded, while the inclusion of patients with bioprosthetic heart valves and/or valve repair was permitted .
After the publication of the main trial results, some subanalyses are now addressing the outcomes of patients included with some sort of valvular heart disease, but only a few specific subgroup analyses of patients with “valvular heart disease” have been reported.
In a subgroup analysis of RE-LY in patients with symptomatic heart failure, 1283 (26%) of the patients with heart failure and 2661 (20%) of the patients without heart failure had some sort of “valvular heart disease”, but no information on outcomes is currently available for these patients .
The most detailed information comes from a secondary analysis of ROCKET-AF, which included 14,171 AF patients, 14% of whom had “significant” valvular disease, some with combined lesions . Mitral regurgitation was by far the most frequent valve disease, seen in 90% of the patients, while only 11% had AS. The authors raised the point that many patients with “non-valvular AF” have significant valve lesions, and this is likely to be even more common in patients seen outside the context of a randomized trial. The risk of stroke in those with “significant” valvular disease was found to be similar to that in patients without significant valve disease after controlling for stroke risk factors. Combined efficacy endpoints in patients with and without valvular disease were similar in patients treated with warfarin or rivaroxaban. Bleeding outcomes were similar in those without valvular disease, but were more frequent with rivaroxaban than warfarin in valvular disease patients. Intracranial bleeding was significantly reduced by rivaroxaban in patients with no valvular disease, and was also reduced, albeit not significantly, in valvular disease patients (non-significant interaction). Whether this effect is real or simply the result of multiple post-hoc analyses of the data is debatable. Anyway, the authors concluded that AF patients with and without valve disease experience the same stroke preventive benefit from oral anticoagulants.
In a subanalysis of ARISTOTLE, which has only been reported in preliminary form, 4808 (26%) of the enrolled patients had “at least moderate” heart valve disease . The results of this subanalysis were consistent with those of the overall ARISTOTLE trial, with no significant interaction according to the presence or absence of valvular heart disease for both stroke and systemic embolism and major bleeding.
In summary, exclusion criteria for concomitant valve disease varied slightly in pivotal trials with NOACs for stroke prevention in AF, with exclusion of most valvular disease patients implemented in some studies, while others included some patients with non-rheumatic valvular disease, valve repair or bioprostheses. However, a general term of “non-valvular AF” was used for the labelling of NOACs, because a clinician cannot refer to the specific inclusion criteria of each trial before prescribing an NOAC, which were not widely available at the same time in all countries. In recent years, this has led the scientific societies to redefine more precisely which patients with AF may be considered to have “valvular” and “non-valvular” AF.
Guidelines
In addition to the lack of absolute consistency reported above, the definitions of “valvular” and “non-valvular” AF also differ slightly in the various guidelines. In 2008, the American College of Chest Physicians guidelines proposed recommendations for patients with valvular heart disease and AF, including mitral stenosis and prosthetic heart valves ; no specific change to the definition was made in the latest edition of these guidelines, published in 2012 . The 2012 focused update of the European Society of Cardiology (ESC) guidelines on AF indicated that it is “conventional” to divide AF into cases that are described as “valvular” or “non-valvular”. Although stating that no satisfactory or uniform definition of these terms exists, the term “valvular AF” used in this guideline implied that AF was “related to rheumatic valvular disease (predominantly mitral stenosis) or prosthetic heart valves” . The 2011 American Heart Association/American College of Cardiology/Heart Rhythm Society AF guidelines said that: “the historical term ‘non-valvular AF’ is restricted to cases in which the rhythm disturbance occurs in the absence of rheumatic mitral valve disease, a prosthetic heart valve or mitral valve repair” . This was confirmed in the 2014 update, where non-valvular AF was defined as AF in the absence of rheumatic mitral stenosis, a mechanical or bioprosthetic heart valve or mitral valve repair .
Overall, the main scientific societies agree that patients with mitral rheumatic valve disease or a prosthetic valve (whether mechanical or biological) have “valvular AF”, but there are disagreements, mainly regarding patients with valve repair and possibly those with AF and rheumatic valve disease not located at the mitral valve.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
