Epidemiology and Prognostic Significance
Atrial fibrillation (AF) and heart failure (HF) are two major cardiovascular problems that predispose to each other and are associated with considerable morbidity and mortality. AF is the most common sustained arrhythmia in clinical practice, affecting 0.4–1% of the general population, with its prevalence increasing with age.1,2 In addition, the estimated prevalence of HF in the general population is 2–3%, rising similarly with age.3,4 The prevalence of AF in patients with HF ranges from 10 to 30%,5 and has been observed to increase in proportion to the severity of HF from <10% in those with New York Heart Association (NYHA) functional class I HF to approximately 50% in those with NYHA functional class IV HF.6 The annual incidence of AF in patients with HF who are in sinus rhythm (SR) is approximately 5%.7 The independent prognostic significance of AF in the setting of HF remains controversial;8,9 however, most evidence suggests that patients with both AF and HF have a worse prognosis than HF patients who are in SR.10–12
The Synergism Between Atrial Fibrillation and Heart Failure
The increased propensity for AF in HF can be explained by structural and electrophysiological atrial remodelling that creates an environment favourable to the development and maintenance of AF.13 Experimental HF promotes the induction of sustained AF by causing atrial interstitial fibrosis that impairs local conduction.14 Patients with HF and no history of atrial arrhythmias exhibit atrial enlargement, loss of functioning atrial myocardium and impaired atrial conduction, which lead to increased inducibility and sustainability of AF.15 Atrial tachycardia-induced ionic remodelling is altered in the presence of HF.16,17 A recent study showed that left ventricular (LV) systolic dysfunction in patients in SR is associated with a shortening of the atrial cellular effective refractory period, which may contribute to a predisposition to AF.18
On the other hand, AF has significant haemodynamic consequences that compromise ventricular function and may exacerbate HF. The loss of atrial enhancement of ventricular filling may considerably decrease cardiac output. Rapid ventricular rate compromises ventricular filling, increases myocardial oxygen demand and decreases coronary perfusion. Furthermore, variation in R-R intervals impairs cardiac contraction because of abnormalities of the force–frequency relationship.19 In addition, tachycardia-related cardiomyopathy can be the result of a sustained, rapid ventricular response during AF.20
Rationale of Treatment
Prevention of thromboembolism, rate control and cardioversion and maintenance of SR are the main objectives of AF treatment.
Prevention of Thromboembolism
HF is a risk factor for thromboembolism in patients with AF. Thus, antithrombotic therapy with either aspirin (81–325mg daily) or a vitamin K antagonist to maintain an international normalised ratio (INR) between 2 and 3 is mandatory. The therapeutic choice should depend on the presence of additional risk factors for thromboembolism, the risk of bleeding complications and patient preferences.21
Rhythm Control or Rate Control?
The matter of rhythm control versus rate control as a long-term therapeutic strategy for the management of AF in patients with HF has been addressed by a recently published multicentre, prospective, randomised trial. The Atrial Fibrillation and Congestive Heart Failure (AF-CHF) trial compared the maintenance of SR (rhythm control) with control of the ventricular rate (rate control) in patients with an LV ejection fraction (LVEF) of ≤35%, symptoms of congestive HF and a recent episode of AF. The primary outcome was the time to death from cardiovascular causes; 1,376 patients were enrolled and were followed for a mean of 37 months. There was no significant difference in the primary outcome (27% in the rhythm-control group and 25% in the rate-control group; p=0.59). In addition, there was no significant difference in any of the secondary outcomes, including death from any cause, worsening HF or stroke. Moreover, the rate-control strategy eliminated the need for repeated cardioversion and reduced rates of hospitalisation. On the basis of these findings, rate control can be considered as a primary therapeutic approach in patients with AF and HF.22
The lack of superiority of a rhythm-control strategy may be explained by the low efficacy and potential toxicity of currently available antiarrhythmic therapies and also by the fact that AF in patients with HF may be a simple marker of disease severity due to poor ventricular function, neurohormonal activation or inflammation, with no independent prognostic value.23
Class I antiarrhythmic drugs are contraindicated in HF patients because of their proarrhythmic and negative inotropic effects, which may exacerbate symptoms and worsen prognosis.24,25 Therefore, restoration of SR can be accomplished by electrical cardioversion, IV ibutilide or oral amiodarone. Ibutilide is associated with a high incidence of torsades de pointes in patients with an LVEF <35% and should be used cautiously in this clinical setting.26 A recent study demonstrated that electrical cardioversion is a safe and effective method for the restoration of SR in patients with AF, reduced LVEF and NYHA HF ranging from class II to IV. Interestingly, optimised HF treatment with β-blockers, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers and mineralocorticoid receptor blockers is associated with higher cardioversion success rates.27
According to current guidelines, amiodarone and dofetilide are the antiarrhythmic agents of choice for maintenance of SR in HF patients.21 A substudy of the Congestive Heart Failure: Survival Trial of Antiarrhythmic Therapy (CHF STAT) demonstrated that patients with AF in the setting of HF have a significantly greater tendency to spontaneously convert to SR during chronic amiodarone therapy, while patients who convert have a lower mortality rate than those who do not. Moreover, the drug prevented the development of newonset AF and significantly reduced the ventricular response in those with persistent AF.28 However, the use of amiodarone in this clinical setting is associated with several risks. Development of bradyarrhythmia may lead to discontinuation of digoxin or permanent pacemaker implantation.29 Furthermore, as shown in the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT), amiodarone therapy may be associated with a higher mortality in NYHA class III patients.30 Finally, extracardiac toxicity may hinder the long-term use of the drug. The Danish Investigations of Arrhythmia and Mortality on Dofetilide in Congestive Heart Failure (DIAMOND-CHF) study showed that in patients with HF and reduced LVEF, dofetilide, a class III agent, is effective in converting AF, preventing its recurrence and reducing the risk of hospitalisation for worsening HF, with no effect on mortality.31
A retrospective analysis of patients with AF or atrial flutter at baseline in the DIAMOND studies demonstrated that dofetilide is effective for restoration and maintenance of SR in patients with LV dysfunction. The study also suggested that restoration of SR is associated with improved survival.32
Dronedarone is a multichannel blocker with electrophysiological properties similar to those of amiodarone, but without the iodine substitute. Thus, it does not cause iodine-related adverse reactions. ATHENA is a recently published, randomised, double-blind, placebo-controlled trial that showed that dronedarone reduces the incidence of hospitalisation due to cardiovascular events or death in patients with paroxysmal or persistent AF. A reduction in the number of hospitalisations for AF was mainly responsible for the reduction in the rate of hospitalisation due to cardiovascular events.33 Twenty-one per cent of patients enrolled in the trial had a history of HF with NYHA class II or III symptoms and 12% had LVEF <45%. A subgroup analysis revealed that the benefit from dronedarone use in HF patients is similar to that of the entire group. However, a previous randomised, double-blind, placebo-controlled trial, the Anti-arrhythmic Trial with Dronedarone in Moderate to Severe CHF Evaluating Morbidity Decrease (ANDROMEDA), which enrolled patients with symptomatic HF, an LVEF of <35% and recent hospitalisation with new or worsening HF, was terminated prematurely because treatment with dronedarone was associated with increased early mortality related to the worsening of HF.34 The difference in outcome between the two trials may be a result of the different study populations. ANDROMEDA enrolled patients with advanced decompensated HF. In contrast, ATHENA excluded patients with decompensated HF within the previous four weeks or NYHA class IV HF. Given the results of ANDROMEDA, dronedarone is contraindicated in patients with severe HF and LV dysfunction. Azimilide is a class III antiarrhythmic agent possessing both IKr- and IKs-channel-blocking properties. A subgroup analysis of the Azimilide Postinfarct Survival Evaluation (ALIVE) trial showed that azimilide is safe and effective in preventing AF reccurence in post-myocardial infarction patients with depressed LV function.35
Despite the lack of randomised controlled data, there is emerging evidence suggesting that non-antiarrhythmic drug therapy may decrease the incidence of AF in patients with CHF. Angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers may prevent or delay the development of AF in patients with HF by unloading the left atrium and inhibiting atrial fibrosis;36–45 however, 3-hydroxy-3-methyl-glutaryl-CoA (HMG CoA)-reductase inhibitors may prevent the development of AF in HF patients.46 Finally, experimental data suggest that selective aldosterone blockade may contribute to AF prevention in the setting of AF.47
Antiarrhythmic therapy is suboptimal because it is associated with significant toxicities and a high recurrence rate. Catheter ablation of AF has grown rapidly and is emerging as an effective alternative therapeutic option for AF that is resistant to pharmacological rhythm or rate control. Although randomised trial data are lacking, observational studies have demonstrated that ablation of AF in patients with HF and a reduced LVEF is associated with a significant improvement in cardiac function48–51 as well as symptoms, exercise capacity and quality of life.48 These studies originate from highly experienced centres, are based on a limited number of patients, have a non-randomised design and a short follow-up and are not powered to assess mortality. However, catheter ablation of AF, when performed by experienced ablationists, seems to be a reasonable alternative in selected HF patients when pharmacological rhythm or rate control fails. A recently published randomised prospective, multicentre clinical trial compared pulmonary vein isolation with atrioventricular (AV) node ablation with biventricular pacing in patients with symptomatic drug-resistant AF, an LVEF of ≤40% and NYHA II or III HF. Pulmonary vein isolation was found to be superior to AV node ablation with biventricular pacing in terms of quality of life, six-minute walk distance and LVEF. In addition, the pulmonary-vein-isolation strategy resulted in high rates of freedom from both AF and antiarrhythmic medications.52
According to current American College of Cardiology/American Heart Association/European Society of Cardiology (ACC/AHA/ESC) guidelines, ventricular rate is considered controlled when the ventricular response ranges between 60 and 80 beats per minute at rest and 90 and 115 beats per minute during moderate exercise.21
In the acute clinical setting, digoxin is the recommended initial treatment option, especially in haemodynamically unstable patients. However, digoxin is relatively slow and ineffective in controlling heart rate. Esmolol, which is an intravenous β1-selective adrenergic receptor blocker with a very short duration of action, may be a useful therapeutic option due to its ability to be titrated. Intravenous (IV) amiodarone is a suitable alternative agent when both rapid rate control and cardioversion are considered appropriate.53
Digoxin, β-blockers or their combination are recommended for longterm rate control.54 Digoxin has reduced efficacy in states of increased sympathetic tone, as occurs in worsening HF and during exercise.55,56 According to recently published ESC guidelines, the use of a β-blocker in patients with HF and reduced LVEF is imperative.54 The four recommended β-blockers are metoprolol CR/XL, carvedilol, bisoprolol and nebivolol. However, there are no randomised trials on the role of β-blockers in patients with AF and HF. A retrospective analysis demonstrated that the use of metoprolol or carvedilol resulted in a significant improvement in LVEF in patients with AF and HF.57 A retrospective analysis of the US Carvedilol Heart Failure Trials Program showed a significant improvement in LVEF and a trend towards reduced mortality and HF hospitalisation.58 A retrospective analysis of the Cardiac Insufficiency Bisoprolol Study (CIBIS) II showed that bisoprolol had no effect on mortality in patients with AF.59 A retrospective analysis of the Carvedilol Or Metoprolol European Trial (COMET) demonstrated that treatment with carvedilol compared with metoprolol offers additional benefits on mortality and morbidity among patients with AF.60 A post hoc analysis of the Metoprolol CR/XL Randomized Intervention Trial (MERIT)-HF study showed that metoprolol had no effect on total or cardiovascular mortality in patients with AF.61
A recent observational study showed that in unselected patients with both AF and HF, β-blocker alone and the combination of β-blocker and digoxin were associated with a similar decrease in the risk of death, whereas treatment with digoxin alone was associated with no survival benefit, similar to patients without any rate-control treatment.62
Non-dihydropyridine calcium-channel antagonists (verapamil, diltiazem) should be avoided as they may worsen congestive heart failure (CHF) due to their negative inotropic effect.
Atrioventricular Node Ablation and Ventricular Pacing
When pharmacological rate control is unsuccessful or poorly tolerated, radiofrequency catheter ablation of the AV node with permanent pacemaker implantation, the so-called ‘ablate and pace’ strategy, should be considered. To date, there is only one randomised controlled study showing that in patients with HF and chronic AF, the ablate and pace strategy is effective and superior to drug therapy in controlling symptoms, with no effect on cardiac performance and disease progression.63 A subgroup analysis of a study examining the effect of this procedure on long-term survival demonstrated that there is no difference in survival between the ablate and pace strategy and drug therapy among patients with AF and HF.64 In a study that examined the effect of AV node ablation on long-term survival in patients with drug-refractory AF and LV dysfunction, LVEF nearly normalised in 29% of patients after ablation, in whom observed survival was comparable to that of normal subjects, suggesting that AF-induced LVEF reduction is reversible in many patients and highlighting potential survival benefits of rate control.65 However, chronic pacing from the right ventricular apex may induce asynchrony, which may cause adverse haemodynamic effects and worsening of LV dysfunction and mitral regurgitation.66 Thus, the choice of pacing site appears crucial in the context of HF. A randomised trial that compared chronic biventricular pacing with right ventricular pacing in patients undergoing ablation of the AV node for management of AF with rapid ventricular rates showed that biventricular pacing is associated with a significant improvement in the six-minute walk test and LVEF. The beneficial effect of cardiac resynchronisation was greater in patients with impaired systolic function or with symptomatic HF.67
Other Device Therapies
Cardiac Resynchronisation Therapy
Multiple randomised controlled clinical trials have confirmed the significant benefit of cardiac resynchronisation therapy (CRT) in NYHA III–IV HF patients who are symptomatic despite optimal medical therapy and who have a reduced LVEF (<35%) and QRS prolongation (QRS width >120ms). With few exceptions, these trials have enrolled patients in normal SR. However, about one-third of patients with severe HF and left bundle branch block have AF.68,69
To date, only one prospective randomised controlled trial designed to assess the clinical efficacy and safety of CRT in patients with chronic AF and NYHA III HF has been published.70 It showed an improvement in exercise tolerance with effective biventricular pacing compared with conventional VVIR pacing. It was later demonstrated that the benefit of CRT was maintained over a 12- month follow-up period.71 Several observational studies have reported a similar benefit of CRT in patients with chronic AF or SR.72–75 A recent meta-analysis of prospective cohort studies comparing the impact of CRT on patients in AF versus SR demonstrated that patients in AF show significant improvement after CRT, with similar or improved LVEF compared with SR patients, but smaller benefits in terms of functional outcomes.76 The presence of AF may not allow optimal delivery of biventricular pacing due to competition by the interference of conducted beats. AV junction ablation may maximise CRT delivery by avoiding inhibition by the rapid intrinsic AV nodal conduction.
Recent prospective, observational studies have highlighted the potential role of AV junction ablation in improving the outcome of HF patients with AF undergoing CRT.77–79 The Atrio-VEntricular Junction Ablation Followed by Resynchronization Therapy (AVERT)-AF trial is an ongoing prospective, randomised, double-blinded, multicentre trial that will test the hypothesis that AV junction ablation followed by CRT significantly improves exercise capacity and functional status compared with pharmacological rate control in patients with chronic AF and depressed LVEF.80
AF is particularly common in implantable cardioverter–defibrillator (ICD) recipients, the vast majority of whom have structural heart disease and LV dysfunction. Persistent AF is associated with appropriate ICD shocks and deterioration of CHF in patients with LV dysfunction.81 Moreover, there may be an association between ventricular tachyarrhythmias and paroxysmal atrial tachyarrhythmias among ICD recipients.82 The Inhibition of Unnecessary RV Pacing with AV Search Hysteresis in ICDs (INTRINSIC RV) trial showed that in patients with no history of AF, the risk of death was dramatically increased from 3.2 to 8.9% when they developed AF shortly after receiving an ICD.83 However, data regarding the efficacy of ICD therapy in the primary prevention of sudden cardiac death in AF patients are scarce and contradictory. Despite the fact that a retrospective analysis of the Multicenter Automatic Defibrillator Implantation Trial (MADIT) II showed that AF patients benefit from ICD therapy,84 a retrospective analysis of the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) demonstrated that compared with placebo or amiodarone, ICD may not improve survival in patients with AF.85
AF and CHF are the two global ‘epidemics’ of cardiovascular disease. A causal reciprocal relation and a complex interplay exist between them. They often constitute a vicious circle and management of both of these conditions remains a medical challenge. In the aftermath of the AF-CHF trial, it seems reasonable to suggest that rate control can be considered as a primary therapeutic approach in patients with AF and HF. New treatment approaches such as catheter ablation and CRT may alter the management of these patients. However, large randomised controlled trials are needed to assess their role.