Stroke Prevention in Atrial Fibrillation and Valvular Heart Disease

Traditional Perspective of Valvular Heart Disease and AF

Rheumatic heart disease involving the mitral valve causes inflammation and fibrosis with disruption of the atrial architecture. There is increased left atrial pressure that contributes to left atrial dilation and increased wall stress. These acquired local conditions are thought to represent predisposing factors to the development of AF [4, 5].

While non valvular atrial fibrillation (NVAF) is known to increase the risk for ischemic stroke by 5-fold, the risk is even greater among patients with mitral stenosis, increasing up to 17-fold [6]. Approximately two-thirds of patients with valvular atrial fibrillation (VAF) (mitral stenosis) who survive a first ischemic stroke experience a second event in the subsequent decade. Half of the events occur in the first year. A retrospective study by Walker and colleagues reported that patients with rheumatic mitral stenosis in sinus rhythm had an incidence of 8% per year of an embolic event; this increased to 31.5% in patients with concomitant AF. Patients with rheumatic valvular disease characterized predominantly by mitral regurgitation were also at heightened risk for a cardioembolic event, even among those in sinus rhythm who had an incidence of 7.7% per year, increasing to 22% in those with AF [7].

While there have been no large randomized clinical trials to demonstrate the effectiveness of anticoagulation for thrombo-prophylaxis in patients with rheumatic valvular heart disease and atrial fibrillation, observational studies and retrospective analyses of patients with mitral stenosis do offer support for treatment. In a study by Fleming and colleagues, patients not anticoagulated had an embolic event incidence of 25% per year; in contrast, patients receiving anticoagulation, primarily a VKA, had 0.8% patient-year incidence of embolic events and a 0% recurrence rate [8]. A review of 254 patients with VAF reported that the incidence of an embolic event was 8 times greater in patients not receiving anticoagulation, versus those treated with an oral anticoagulant [9]. In 1977, the Joint Committee for Stroke published a statement that reviewed 11 studies evaluating thrombo-prophylaxis with anticoagulant therapy in patients with rheumatic mitral disease; they recommended anticoagulation in patients with rheumatic heart disease, particularly those with AF and a history of ischemic stroke [10].

Contemporary Perspective of Valvular Heart Disease and AF

AF currently affects over 2 million persons in the US, and this number is projected to increase to 10 million by the year 2050 [11]. The total number may actually be much higher, as nearly 50% of persons with AF do not experience signs or symptoms of the arrhythmia. AF is believed to account for at least 10-12% of all ischemic strokes [12, 13] and the risk increases steadily with age and other common conditions, including systemic hypertension.

The AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation has defined AF in the presence of rheumatic heart disease or either mechanical or bioprosthetic valve in the mitral position as VAF [14, 15]. The European Society of Cardiology Guidelines also determined VAF to be present with concomitant rheumatic heart disease (primarily mitral stenosis) or a prosthetic heart valve [16].

Clinical Trials of New Oral Anticoagulants

The development of new oral anticoagulants (NOACs) as an alternative to VKAs in patients with AF focused on those without valvular heart disease (NVAF) (Table 1). This decision was reflective of a changing landscape in Europe and North America, with a declining incidence and overall prevalence of rheumatic heart disease and rapidly increasing number of persons with AF at risk for ischemic stroke. It also acknowledged the wide phenotypic variability for cardioembolic stroke risk with mild, moderate and severe rheumatic mitral stenosis. Despite the clear decision to perform pivotal trials in patients with NVAF, patients with valvular heart disease were enrolled.

In order to better understand the spectrum of valvular heart disease captured in the NOAC registration trials, one must carefully review individual study exclusion criteria. The Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF) excluded patients with prosthetic heart valves and those with hemodynamically significant mitral stenosis [17]. In contrast, the Randomized Evaluation of Long-Term Anticoagulation (RE-LY) trial excluded patients with prosthetic heart valves and those with hemodynamically significant native valvular heart disease, but did not distinguish between the aortic and mitral valves [18, 19]. The Apixaban for Reduction in Stoke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE) trial excluded patients with moderate-to-severe mitral stenosis and those with prosthetic heart valve [20]. The Apixaban versus Acetylsalicylic Acid to Prevent Stroke in Atrial Fibrillation Patients Who Have Failed or Are Unsuitable for Vitamin K Antagonist Treatment (AVERROES) trial excluded patients with valvular heart disease that was severe enough to require surgery. Perhaps the most inclusionary of the phase 3 trials, the Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation-Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48) trial, did not exclude patients with either bioprosthetic valves or those who had previously undergone a valve repair [21, 22].

RE-LY Trial

In the RE-LY trial, a direct thrombin inhibitor, dabigatran etexilate, was compared at two fixed doses with warfarin in a blinded study. The primary outcome was defined as either cerebral or systemic embolism. At 110mg dose, dabigatran was found to be non-inferior to warfarin for the rate of primary outcome. At the higher dose of 150mg, dabigatran was superior to warfarin for primary outcomes risk reduction [23].

In this trial, patients that were excluded had either a history of “severe valvular disorders,” defined by prosthetic valve or hemodynamically relevant valve disease [19, 23]. Despite the exclusion criteria, there were patients with valvular disease that were included in the trial. In total, 21.8% percent of the patients enrolled in the RE-LY trial had some form of valvular heart disease that did not exclude them from participation; only 1.1% of them had mild mitral stenosis. Additionally, a subgroup analysis compared the overall effects of dabigatran with warfarin in patients with and without symptomatic heart failure (defined as the presence of NYHA class II or higher symptoms in the six months before screening with a prior admission for congestive heart failure) [24]. In this analysis, 26.2% of participants in the symptomatic heart failure group and 20.1% in the group without NYHA class II or greater heart failure, had “valvular heart disease” [24]. The study showed no significant outcome differences with either dabigatran or warfarin in this subset of patients.

While the RE-LY trial illustrates the benefits of dabigatran for thromboembolism protection, the Randomized Phase II Study to Evaluate the Safety and Pharmacokinetics of Oral Dabigatran Etexilate in Patients after Heart Valve Replacement (RE-ALIGN) is perhaps the most direct evaluation of a NOAC versus VKA for patients with valvular heart disease - in this particular case, mechanical prosthetic valves [25].

The study compared anticoagulation with dabigatran against warfarin in patients with either mechanical aortic or mitral valve replacement (AVR, MVR) [25]. Patients who had undergone either AVR or MVR within 7 days, or at least 3 months earlier, were then randomly assigned to receive warfarin (dosed to an INR of 2 to 3 or 2.5 to 3.5, on the basis of thromboembolic risk) or dabigatran (dosed to obtain a trough plasma level of at least 50 ng per milliliter). The study was discontinued due to the high thromboembolic and bleeding events among patients in the dabigatran group. The composite of stroke, transient ischemic attack, systemic embolism, myocardial infarction, or death occurred in 15 patients (9%) in the dabigatran group and 4 patients (5%) in the warfarin group (hazard ratio in the dabigatran group, 1.94; 95% confidence interval [CI], 0.64 to 5.86; P=0.24); thromboembolic events were higher in the group started on dabigatran within 7 days of MVR. Additionally, all cases of valve thrombosis detected in clinically asymptomatc patients were reported in the dabigatran group (3%) [25].

Based on the available data, dabigatran for patients with mechanical valves cannot not be supported, showing no benefit and an excess risk; however, the RE-ALIGN trial did generate a dialogue and series of questions to better understand the findings that included a lack of uniform aspirin use and reported mean time within the pre-defined target drug range of approximately 50% [26]. The latter suggests changes in dabigatran pharmacokinetics in the setting of cardiac surgery or the dynamics of blood flow and metabolism after heart valve surgery. Whether similar properties should be expected in the setting of trans-aortic valve implantation (TAVR) will require further investigation.