Atrial fibrillation (AF), characterised by rapid and irregular activation of the atrium, is the most common arrhythmia in clinical practice, and 2.3 million people in North American and 4.5 million people in the EU are estimated to have paroxysmal or persistent AF.1 An estimated 0.4–1% of the general population are affected with AF,1 with prevalence increasing to nearly 10% in the octogenarian population.2–4 The ageing population has led to an increase in AF-related hospital admissions by 66%,5 presenting a significant impact for healthcare providers with a total cost burden of approximately €13.5 billion in the EU.6 AF can significantly affect morbidity and mortality if left untreated, particularly in women. Patients with non-valvular AF face a two- to seven-fold increased risk of ischaemic stroke relative to patients without AF,7 and studies have shown that other factors such as intercurrent illnesses and age can further increase the risk of stroke.7–9 In addition to an increased long-term risk of stroke,10 AF is associated with tachycardia-induced cardiomyopathy, congestive heart failure and increased mortality; the all-cause mortality rate in patients with AF is nearly double that of age-matched individuals without AF.11 Mortality rate is also closely linked to the severity of underlying heart disease,2,11–14 and as heart failure can both promote and be exacerbated by AF, patients who present with both conditions have a poor prognosis.6,15 Managing patients with AF therefore presents a major challenge.
Pathophysiology – Atrial Fibrillation (AF) Begets AF
AF is the result of a complex interaction between initiating triggers, such as ectopic foci, perpetuating tissue substrates and electrophysiology.16–18 Atrial remodelling, stimulated by rapid atrial rate, denotes the process by which AF modifies electrical, functional and structural properties, subsequently leading to the continued initiation and maintenance of arrhythmia.19,20 These AF-induced atrial changes inevitably lead to disease progression in a vicious self-perpetuating cycle, endorsing the notion that ‘AF begets AF’ (see Figure 1).
Electrical remodelling is characterised by changes in atrial refractoriness, atrial conduction and sinus node function. The calcium channel levels present the most clinically significant ionic change in AF. Under normal conditions, calcium enters muscle cells during each action potential and contributes to cellular repolarisation, but the rapid atrial firing and increased number of action potentials in AF lead to an increased influx of calcium ions. Short-term inactivation and long-term downregulation of L-type calcium current (ICaL) are then introduced to compensate for potential calcium overload. Significantly, the reduction in ICaL over the long term shortens the duration of the action potential and atrial effective refractory period (ERP) and reduces the ERP adaptation to rate.20,21 Abnormal expression of other ionic currents has also been described in AF, including the rapid sodium current and several potassium currents.21–24 Although electrical remodelling starts rapidly within the first minutes of AF, the electrical changes are often rapidly reversible upon restoration of normal sinus rhythm.
Contractile remodelling of atrial muscle can occur rapidly or over a period of weeks to months owing to tachycardia.20,25 Myolysis and reduced release of systolic calcium caused by downregulation of ICaL have been implicated in the observed loss of contractility and reduced atrial transport function.20,26,27 Contractile remodelling presents significant clinical consequences, including thrombus formation, atrial dilatation, and potential disease progression due to the coexistence of multiple wavelets. Even after restoration of sinus rhythm, contractility is only recovered over time, likely because atrial damage caused by myolysis needs to be repaired. Over this period of time, patients are still at high risk of atrial blood stasis and thromboembolitic events.28 Structural modifications, many of which are irreversible, result from both sustained AF and underlying cardiovascular diseases.26 Atrial fibrosis and loss of myocardial mass are frequently seen in AF, but such structural changes are difficult to distinguish from those of associated heart disease. Atrial fibrosis is associated with disturbances in intra-atrial conduction and susceptibility for multiple re-entry wavelets and AF. It is not yet clear how structural changes in the atria are involved in the development of and progression to persistent AF, but some studies have suggested a role for structural remodelling in reducing atrial contractility and enhancing atrial dilatation and altering electrophysical refractoriness and conduction.20,23
Rate Control versus Sinus Rhythm Control
The strategic management of patients with AF is a heavily debated topic, where the initial decision of how to manage patients is a choice between rate control and sinus rhythm control. Control of ventricular rate can be achieved pharmacologically or by ablation or modification of the atrioventricular (AV) node and implantation of a permanent pacemaker. Restoration and maintenance of sinus rhythm aims to correct the disturbance in rhythm by using electrical cardioversion or antiarrhythmic drugs.
Although it would seem logical to choose rhythm control whenever possible given the underlying mechanisms promoting AF and disease progression, randomised trials have presented conflicting data on morbidity and mortality outcomes associated with rhythm control versus rate control. Both the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) and Rate Control vs. Electrical Cardioversion for Persistent Atrial Fibrillation (RACE) studies found no differences in survival between rate control and rhythm control in terms of mortality or morbidity.29,30 However, rhythm control was associated with higher rates of hospitalisation and more frequent adverse effects. Conversely, some studies have demonstrated reduced mortality with rhythm control relative to rate control.31,32 While the AFFIRM, RACE, Pharmacologic Intervention in Atrial Fibrillation (PIAF) and Strategies of Treatment of Atrial Fibrillation (STAF) studies found no difference between rate control and rhythm control in terms of quality of life,30,33–36 other studies suggest that the restoration and maintenance of sinus rhythm may impart other clinical benefits, such as improved symptomology, greater exercise tolerance, reductions in atrial remodelling and improved left ventricular function.37–41 However, inconsistencies in the available data have impeded the widespread use of the rhythm control strategy in early AF.
As AF is a progressive disease, early restoration of sinus rhythm may suppress symptoms and minimise atrial remodelling and, subsequently, AF progression. A recently published position paper suggests implementing a new strategy for managing early AF, including rhythm control as well as aggressive detection and management of related conditions, in order to prevent atrial remodelling and potentially allow for halted disease progression, and possibly even a reversion of arrhythmogenic changes in the atria.42
Under this new proposed paradigm, rhythm control should be initiated immediately following the first presentation of AF, thereby potentially delaying remodelling and preventing the onset of irreversible structural remodelling, while also allowing for the assessment and treatment of underlying diseases and potential underlying causative factors. Intervention with antiarrhythmic drugs can be very effective, but electrical cardioversion is available if antiarrhythmic drugs are unable to induce the return of sinus rhythm. After cardioversion, the use of antiarrhythmic drugs improves the chances of maintaining sinus remodelling and preventing electrical remodelling: a study in which electrical remodelling appeared to ‘reverse’ within the first few weeks after cardioversion suggested that antiarrhythmic drugs may be most indispensable during this period.43 Moreover, the rationale for early conversion of AF to sinus rhythm becomes clearer when considering that conversion becomes increasingly difficult with a prolonged duration of AF.44
Guidelines Guiding Practice
Thus far, studies comparing rate versus rhythm control have failed to identify any single strategy that can be applied optimally across all patients with AF. Patients enrolled in AFFIRM and RACE generally had serious cardiac disease or other co-existing illnesses: the enrolment criteria were ≥1 risk factor for stroke or death (including older age), and patients had to be able to tolerate AF if rate-controlled in AFFIRM and persistent AF refractory to electrical cardioversion in RACE.29,30 The results of these trials are therefore likely ungeneralisable to the wider patient population, as younger patients without or with minimal structural heart disease, accounting for as many as 30% of all AF patients,45 were excluded from these studies due to their low risk of fatal events. Furthermore, in the AFFIRM trial amiodarone was the most commonly used drug and its safety profile may have contributed to the high rate of adverse events seen in the rhythm control group.30
Current comprehensive practice guidelines for the management of patients with AF, developed jointly by the American College of Cardiology (ACC)/American Heart Association (AHA)/European Society of Cardiology (ESC) and sanctioned by the Heart Rhythm Society (HRS), were last updated in 2006. These guidelines state that rate control could be considered a reasonable initial therapy for older patients with persistent AF who also have hypertension or heart disease, depending on symptomology.6
Rhythm control may present a better initial therapy for younger patients, particularly those with paroxysmal ‘lone AF’, in which AF is not associated with any demonstrable structural heart disease.6 The selection of an antiarrhythmic drug should be based on the characteristics of the individual patient, accounting for arrhythmia burden, type of underlying heart disease and co-morbid conditions, the nature, intensity and frequency of symptoms, adverse effects and patient preference.6 Class Ic agents flecainide and propafenone, and beta-blocker sotalol are recommended as first-line therapies in patients with lone AF.
Patients with heart disease fall into three categories according to the ACC/AHA/ESC guidelines: those with hypertension, those with coronary heart disease and those heart failure, by order of increasing severity (see Figure 2). Hypertensive patients with left ventricular hypertrophy are at an increased risk of torsades de pointes owing to early ventricular after-depolarisations;46,47 drugs that do not prolong the QT interval are therefore the preferred first-line therapy. The class Ic agents propafenone and flecainide are suitable choices in the absence of ischaemia or marked left ventricular hypertrophy in these patients.6 In AF patients with cardiovascular disease without heart failure, the substantial beta-blocking activity of sotalol is the first-line recommendation due to its neutral effect on survival48 and shorter half-life and less long-term toxicity compared with amiodarone. However, the QT-prolonging effects of sotalol can increase the risk of torsades de pointes and proarrhythmia. Amiodarone is the preferred choice in patients with cardiovascular disease also presenting with heart failure, with dofetilide as a reasonable second-line alternative.49 Patients with heart failure are especially susceptible to the ventricular proarrhythmic potential of antiarrhythmic drugs owing to their myocardial vulnerability and electrolyte balance. Although amiodarone has a high incidence of side effects and organ toxicity and dofetilide presents a risk of proarrhythmia, these drugs still present the best option for these patients50,51 and are therefore recommended as first-line therapy for AF patients with heart failure. It should be noted that class Ic agents are contraindicated in the population of AF patients with structural heart disease; earlier studies such as the Cardiac Arrhythmic Suppression Trial (CAST) found that patients with ischaemia or prior myocardial infarction exposed to this class of drugs faced an increased mortality risk.52,53 The use of catheter ablation for sinus rhythm maintenance is an option in patients who are non-responsive to antiarrhythmic drug therapy.54
Despite the guidelines concerning the management of AF, available data suggest that the recommendations are often not followed in clinical practice. Market research data have suggested a severe underuse of the class Ic agents; where the guidelines suggest that class Ic agents are suitable for approximately half of all AF patients, their actual use may actually be only 19%.55 A study of antiarrhythmic drug use in the US from 1991 to 2000 showed that while the use of class Ic agents increased from 0.5 to 2.9% of visits, amiodarone use increased from 0.2 to 6.4% of visits – a 5.8-fold difference in amiodarone use relative to that of propafenone and flecainide.56 An analysis of antiarrhythmic drug use in the Fibrillation Registry Assessing Costs, Therapies, Adverse Events and Lifestyle (FRACTAL) study found a significant difference in drug use according to the speciality of the prescribing physician: electrophysiologists appeared to prefer amiodarone and cardiologists tended to use class Ic agents, while internists exhibited no clear preference.57 Moreover, the data also suggest that class Ia agents are prescribed more often and class Ic agents less often than recommended by the guidelines. A recent observational study of the management of AF patients and consistency with guidelines in France found that therapeutic strategies applied in practice were largely consistent with the available guidelines, although amiodarone was frequently used in patients with no apparent heart disease.58
The relative efficacies of the various antiarrhythmic agents available for the management of AF are as yet unknown, leaving drug selection to be driven by drug-safety issues instead – a factor that has long been established in the guidelines. In particular, the use of amiodarone, though effective in sinus rhythm maintenance, is limited by its long-term safety profile59–61 and potential for end-organ toxicity and severe extracardiac side effects, even at low doses.62 The use of amiodarone should therefore be restricted to second-line therapy, except in patients with concomitant heart failure or substantial left ventricular hypertrophy, in whom amiodarone appears to offer an overall advantage. Selection of an antiarrhythmic agent for the maintenance of sinus rhythm should therefore be undertaken with careful consideration of a patient’s concomitant cardiovascular conditions, as well as the safety and tolerability profile of the agent in question.
Atrial Fibrillation and Structural Heart Disease
Many patients with AF present with concurrent cardiovascular conditions such as left ventricular hypertrophy, coronary artery disease, or heart failure. Out of 17,949 adults with AF enrolled in the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study, 49% had hypertension, 29% were diagnosed with heart failure and 35% had a history of coronary heart disease.1 Similar findings of a high prevalence of comorbid conditions have also been reported in other trials. Of the 4,060 patients enrolled in AFFIRM, 51% were hypertensive, 26% had coronary artery disease, 5% had valvular disease and 23% had a history of congestive heart failure. Only 12% were found to be free of any apparent cardiac disease.29 In FRACTAL, where baseline criteria were analysed in 963 patients, 49% had hypertension, 25% had coronary artery disease, 17% had valvular disease and 18% had heart failure.63
Concomitant structural heart disease is a frequent accompanying and/or underlying factor of AF, and can present an increased risk of proarrhythmic events. Since the CAST trials in which class Ic drugs were found to have detrimental effects on the wellbeing and survival of patients with concurrent cardiovascular conditions,52 studies evaluating antiarrhythmic drugs in AF have greatly focused on cardiac safety. It follows that the identification of heart disease in patients with AF is extremely important, and that cardiac disease, if present, is an obvious point of consideration in selecting the appropriate antiarrhythmic agent. Although amiodarone is recommended for use in patients with heart failure or substantial left ventricular hypertrophy, amiodarone continues to be used excessively in the treatment of AF, even in cases in which other antiarrhythmic agents are better suited and indicated for as first-line therapy in the guidelines – highly contradictive considering the emphasis on cardiac safety. Amiodarone has been shown to induce a number of clinically significant extracardiac effects, particularly over a long period of use, including pulmonary and liver toxicity, thyroid imbalances, photosensitivity, neuropathy and blindness; all of these adverse events are absent in sotalol, dofetilide and the class Ic agents flecainide and propafenone. Indeed, a meta-analysis comparing amiodarone and class Ic drugs for termination of recent AF concluded that the efficacies were similar, and all were suitable for use as an alternative therapy for patients in whom class Ic and other more rapidly acting drugs cannot be used.64 It is important for clinicians to understand the inter-relationships between structural heart disease and antiarrhythmic drugs not only in terms of proarrhythmic risk, but also in terms of safety and toxicities.
Antiarrhythmic Drug Selection in the Absence of Structural Heart Disease
Patients with lone AF, or AF with minimal structural heart disease, are candidates for class Ic drugs or sotalol, although in this setting sotalol has a small risk of torsades de pointes. Flecainide and propafenone are highly effective in the termination of AF in patients without structural heart disease, with differing rates of success depending on whether the preparation is oral (50–80%) or intravenous (90%).65 Notably, these drugs are generally well tolerated and are essentially devoid of extracardiac organ toxicity. The guidelines state that the ability of this patient population to administer these agents orally in an outpatient setting may be beneficial,6 and this ‘pill-in-the-pocket’ approach has been shown to be safe in outpatients, with high patient compliance, a low incidence of adverse events and a marked reduction in emergency room visits and hospitalisations.66 Although flecainide has been shown to be one of the most effective drugs in cardioversion to sinus rhythm, with a success rate of 77–93%,67–70 efficacy is lost if the arrhythmia is present for more than 24 hours,68,69 a phenomenon proposed to result from electrical remodelling.71 However, intervention with intravenous flecainide early in AF recurrence (<10 hours) in patients who had previously failed pharmacological cardioversion and had undergone electrical cardioversion has been shown to successfully restore efficacy in 58% of cases of recurrent AF.72 This small window of opportunity necessitates strict rhythm monitoring, however. The pill-in-the- pocket approach has been proposed to reduce the interval between arrhythmia onset and treatment. Moreover, this approach has been proposed to be beneficial for patients with recurrent episodes of symptomatic AF, as such episodic therapy can reduce the risk of toxicity compared with sustained therapy.6 Early treatment, alongside the lower likelihood of structural heart disease in younger patients, could then potentially increase the long-term efficacy of therapy.
It is possible that the low use of class Ic drugs is an ongoing result of the fall-out from the CAST study data, but there remains a need to emphasise the important role of this drug class in patients with no or minimal structural heart disease, as the safety advantage of these drugs in this patient population cannot be denied. Clinical studies have found flecainide and propafenone to be similar in the number of patients who convert to sinus rhythm,73,74 although the studies have found the relative rates of conversion to be inconsistent.73,74
Rates and severity of adverse effects have also been similar between the two drugs.73–75 One study has stated that the adverse events experienced with propafenone were important enough to stop therapy, such that the probability of a patient’s compliance at one year tended to be higher with flecainide than with propafenone.76 While both drugs have demonstrated efficacy and safety, it is ultimately the clinician’s choice to use either flecainide or propafenone, based on the drug’s safety and tolerability profile and the amount of clinical experience, dosage and compliance considerations and clinical data available with each drug.
Conclusions
AF is a complicated condition with a significant impact on morbidity and mortality. With the potential to develop an increased risk of thromboembolism and irreversible structural changes to the heart, it appears that rhythm control with antiarrhythmic drugs early in the disease course could be beneficial. Although rate control may be better suited for elderly patients with structural heart disease, younger patients with healthier hearts deserve rhythm control at an early stage following diagnosis. The class Ic drugs have proved useful in this population of patients, although adherence to guideline recommendations is poor and other more dangerous drugs continue to be prescribed in lieu of the safer first-line agents. With the emphasis on cardiac safety, one would expect cardiologists to then select the best appropriate drug for their patients based on the type and severity of heart disease, concomitant disease, patient characteristics and long-term safety of the drug in question.