Atrial fibrillation (AF) is the most common cardiac arrhythmia encountered in clinical practice. The prevalence of AF increases with age;1 one and approximately 2.3 million US adults are currently afflicted with AF, with that number projected to increase to 5.6 million by the year 2050.2
AF typically results in an irregular rhythm with poorly controlled ventricular rate and loss of atrial mechanical systole; the combination increases the risk for congestive heart failure (CHF), cardiomyopathy and stroke. Maintenance of sinus rhythm can offset these adverse physiologic effects. However, current rhythm control strategies most commonly incorporate antiarrhythmic drugs, which are associated with significant side effects, including proarrhythmia, bradycardia, heart block as well as systemic toxicity. On the other hand, a rate control strategy allows AF to persist, potentially creating long-term, adverse atrial (and ventricular) remodeling.
Thus, uncertainty regarding the optimal treatment strategy for AF—namely, aggressive attempts at maintenance of sinus rhythm using antiarrhythmic drugs and electrical cardioversion versus ventricular rate control using atrioventricular (AV) nodal blocking agents and/or AV junction ablation—led to the initiation of several randomized trials comparing these approaches. The largest of these trials was the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial.3
AFFIRM was a randomized, multicenter clinical trial, sponsored by the National Heart, Lung, and Blood Institute, of 4,060 patients who were either ≥65 years of age or had at least one risk factor for stroke or death, which included the following: hypertension, diabetes mellitus, CHF, previous stroke, previous transient ischemic attack, systemic embolism, left atrial size ≥5cm, left ventricular ejection fraction (LVEF) ≤40%, or fractional shortening ≤25%. Patients were assigned to either a rate-control or rhythm-control strategy and subject to long-term follow-up.There was no statistical difference in total mortality between the two strategies; however, a rhythm-control strategy was associated with increased risk for hospitalization, adverse drug effects and thromboembolism. The latter occurred mainly in patients randomized to the rhythm-control strategy after anticoagulation had been discontinued because of apparent maintenance of sinus rhythm. Sub-study analysis of this trial has resulted in significant insights in the management of patients with AF.
In the AFFIRM trial, rate control was achieved using pharmacologic methods including β-adrenergic blockers, non-dihydropyridine calcium channel blockers, digoxin and combinations of these drugs to a target heart rate ≤80bpm at rest and either a heart rate not higher than 110bpm during a six-minute walk test (6MWT) or an average heart rate ≤100bpm during a 24-hour Holter (with no heart rate ≥110% of maximum age-adjusted predicted rate).
The most effective drug for rate control was a beta-blocker (with or without digoxin), which provided adequate rate control in 85% of patients.4 However, 37% of patients required a change or addition of another drug over long-term follow-up and 5% of patients required AV junctional ablation after failing maximal pharmacologic therapy. These data suggest that although ventricular rate control is a desired goal of therapy, practically this can be quite difficult to achieve pharmacologically.
An important question has been whether ventricular rate control to the desired targets favorably affects patient outcomes.5 To resolve this issue, patients in the AFFIRM trial with AF randomized to the rate-control arm were grouped according to quartiles of resting as well as exercise heart rate. Overall, target resting heart rate was achieved in 63% of patients and target exercise heart rate was achieved in 81% of patients. Surprisingly, the degree of heart rate control (as assessed either at rest or during exercise) did not influence rehospitalization, quality of life, New York Heart Association (NYHA) functional class, 6MWT distance, or overall survival.
In a pre-specified sub-study involving 410 AFFIRM patients randomized to a rhythm-control strategy who were eligible to receive two of three antiarrhythmic medications (amiodarone, sotalol, or a class I drug), a second randomization to one of these three drugs was performed.6 The primary end-point was the percentage of patients at one year who were alive, were taking the assigned antiarrhythmic medication, had not undergone cardioversion, and were in sinus rhythm. Compared separately with both sotalol and class I antiarrhythmic drugs (most commonly quinidine, procainamide, or propafenone), amiodarone proved to be the most efficacious, with no statistical difference in success between patients receiving sotalol versus class I therapy. Importantly, antiarrhythmic drugs were frequently discontinued during the first year due to medication-induced side effects.
At the same time, compared with an 'intention-to-treat' analysis of the primary AFFIRM report, an 'on-treatment analysis' found that the presence of sinus rhythm was independently associated with a survival benefit.7 This analysis adjusted for multiple variables and treated sinus rhythm and antiarrhythmic drugs as separate variables. In this sub-study, antiarrhythmic drug use was associated with increased mortality. These data suggest that a method of restoring sinus rhythm associated with fewer side effects could be highly desirable.
In patients receiving antiarrhythmic medications, the risk of proarrhythmia was quite low. In fact, the cumulative incidence of all arrhythmic events was only 5% at five years.8 In contrast, the overall mortality in this group of patients was 27%.Thus, most of the mortality was accounted for by an increased incidence of non-cardiac events in patients randomized to the rhythm control strategy.
Stroke and Atrial Fibrillation
Prior to AFFIRM, the American College of Cardiology (ACC)/American Heart Association (AHA)/European Society of Cardiology (ESC) practice guidelines allowed for the discontinuation of long-term anticoagulation once sinus rhythm was restored.9 Thus, the AFFIRM protocol allowed for stoppage of warfarin if sinus rhythm was maintained for a minimum of four weeks (and preferably 12 weeks) using an antiarrhythmic drug in the rhythm-control group.Overall, 8.2% of patients had a stroke event, including ischemic stroke (most commonly due to cardioembolism), primary intraparenchymal hemorrhage, and subdural or subarachnoid hemorrhage.The stroke risk was similar for patients in the rate-control and rhythm-control arms.However, the presence of AF was associated with a 60% increase in risk of stroke and use of warfarin was associated with a 69% decrease in risk of stroke.10
There are several possible explanations for this increase in stroke events in the rhythm-control group. First, the study highlights the overall inefficacy of antiarrhythmic drugs as only 56% of patients were in sinus rhythm at the time of stroke. Secondly, it is likely that these patients also had frequent paroxysms of asymptomatic AF. An under-appreciation of this phenomenon at the time of the AFFIRM study may have led to the premature discontinuation of anticoagulation therapy in many patients. Finally, it is possible that stroke events result from other cardiac sources, such as a left ventricular thrombus or a patent foramen ovale10 or other non-cardiac vascular origins.
Mortality in AFFIRM
During a mean follow-up of 3.5 years, more deaths were observed in the rhythm-control arm than the rate-control arm (356 versus 310 deaths, p=0.07).11 When the cause of death was investigated in greater detail, there was no difference in the rate of cardiac deaths (arrhythmic and non-arrhythmic) between the two management strategies. This was also true for vascular deaths. The increase in the number of deaths in the rhythm control group was attributed almost entirely to an approximate 50% increase in non-cardiovascular deaths in these patients. Specifically, there were increased deaths due to pulmonary causes (predominantly pneumonia) and cancer (principally lung cancer).11 The survival curves for non-cardiovascular deaths began to diverge at one year.
Most patients in the rhythm-control arm were treated with amiodarone; however, one can only speculate why this drug was associated with an increased risk of pneumonia and cancer. Interestingly, direct amiodarone-induced pulmonary toxicity contributed to only three pulmonary deaths in the rhythm-control arm. Similarly, since warfarin may have an antineoplastic effect, discontinuation of the drug may have contributed to some loss of oncologic protection and resulted in a slight increase in lung cancer.11 However, it is important to recognize that in AFFIRM, patients were randomized to a particular treatment strategy rather than a specific drug; thus, there can be potential bias in trying to link mortality to a specific drug.
While AFFIRM represents the largest trial of its kind to date, it remains to be determined whether its findings can be generalized to other patient populations. These include younger patients without risk factors for stroke as well as patients with CHF and underlying left ventricular dysfunction. The latter group is being evaluated in the on-going Atrial Fibrillation and Congestive Heart Failure (AF-CHF) trial.12
In addition, new treatment modalities in current practice and on the horizon hold promise. These strategies include 'atrial specific' antiarrhythmic agents that may promote maintenance of sinus rhythm while avoiding ventricular proarrhythmia and systemic toxicity, which hamper current antiarrhythmic therapy.13,14 Similarly, new anticoagulants such as direct thrombin and factor Xa inhibitors offer the hope of stroke prevention through more reliable anticoagulation, which can obviate the need for constant monitoring and reduce significantly the risk of bleeding.15,16 These advances may yet tip the balance back in favor of a rhythm control strategy. In addition, radiofrequency ablation of the pulmonary veins has been successful in long-term maintenance of sinus rhythm, representing a curative strategy that eliminated the need for pharmacotherapy for AF in previously drug-refractory patients.17