The pharmaceutical pipeline for atrial fibrillation

Abstract

Atrial fibrillation (AF) is associated with a significant burden of morbidity and increased risk of mortality. Beyond outstanding advances in catheter ablation procedures, antiarrhythmic drug therapy remains a corner-stone to restore and maintain sinus rhythm. However, potentially life-threatening hazards (proarrhythmia) and significant non-cardiac organ toxicity have made new drug development of prominent relevance. Multichannel blocking, atrial selectivity, and the reduction of the risk of adverse events have all constituted the main theme of modern antifibrillatory drug development. Dronedarone, an analog of amiodarone, has the unique characteristic of being the first antiarrhythmic drug demonstrated to reduce hospitalizations in AF. Dronedarone is associated with less systemic toxicity than amiodarone, although being less effective for sinus rhythm maintenance. Atrial selective agents have been developed to target ion channels expressed selectively in the atria. Among the most promising drugs of this class is vernakalant, which has been shown effective for the acute conversion of AF with small risk of proarrhythmia. Finally, increasing evidences support antiarrhythmic effectiveness of traditional non-antiarrhythmic drugs, such as renin-angiotensin system blockers, statins, and omega-3 fatty acids.

In this article, we will focus on recent advances in antiarrhythmic therapy for AF, reviewing the possible clinical utility of novel antifibrillatory agents.

Key words::

• Antiarrhythmic therapy
• atrial fibrillation
• atrial-selective drugs
• dronedarone

Abbreviations
AF	=	atrial fibrillation
SR	=	sinus rhythm
RRR	=	relative risk reduction
ARR	=	absolute risk reduction
NNT	=	number needed to treat

Key messages

• Multichannel blocking, atrial selectivity, and the reduction of the risk of adverse events have all constituted the main theme of modern antifibrillatory drug development.

• Dronedarone, a multichannel blocker analog of amiodarone, has been recently approved for sinus rhythm maintenance in AF, although its benefit across all the spectrum of AF patients is not definitely demonstrated.

• Vernakalant, a promising atrial-selective agent that has recently been approved in Europe, appears an effective and safe drug for rapid termination of AF, although the real incremental value in this setting compared to other available antiarrhythmic agents merits further investigation.

Introduction

Atrial fibrillation (AF) is the most common sustained arrhythmia in Western countries, with an estimated 30 million patients affected by 2050 across the United States and Europe alone (1). Atrial fibrillation has a significant impact on morbidity, mainly related to symptoms, heart failure and thromboembolic events, and is the most frequent arrhythmic cause of hospital admission in the US (2,3). In addition, AF is associated with excess mortality independently of thromboembolic complications (4,5).

The primary therapeutic goal in the management of AF is the restoration and maintenance of sinus rhythm. Recently, international societies have reconsidered the problems of trials of AF treatment and have suggested that additional end-points may be important in the assessment of different AF treatments (6,7). These include reduction of important clinical outcomes in AF, such as mortality, hospitalization, and stroke, as well as relevant economic analyses such as the cost of treatment (6,7).

To date, the most effective treatment for AF is catheter ablation (8). On the other hand, antiarrhythmic drug therapy still remains the most widely adopted rhythmcontrol strategy and also constitutes an important component of treatment for many patients undergoing catheter ablation. Currently available antiarrhythmic drugs, however, have significant limitations. In particular, the use of class I antiarrhythmic agents is limited by increased risk of proarrhythmia, with a demonstrated increase in cardiovascular death in patients with structural heart disease (9,10). Among class III agents, sotalol has been demonstrated to increase mortality in high-risk patients (11), and dofetilide is rarely used due to the substantial risk of torsade de pointes and the need for in-hospital initiation (12). Ultimately, amiodarone, the most widely used antiarrhythmic agent, has multiple mechanisms of action including class III properties and has minimal proarrhythmic potential (13). On the other hand, amiodarone has never been proven to reduce hospitalization or mortality in AF patients (14,15) and is flawed by high risk of non-cardiac toxicity, such as thyroid and hepatic toxicities, pulmonary fibrosis, neuropathy, corneal pigmentation, and skin discoloration (13). As a result, the development of new antiarrhythmic drugs for AF is crucial. The main theme of modern antifibrillatory drug development has been the minimization of the risk of cardiac and systemic toxicities, while maintaining and, possibly, intensifying antiarrhythmic effectiveness (16). To this aim, significant effort has been directed toward the development of drugs that act through multiple ion channels blockade, resulting in a net beneficial antiarrhythmic effect with minimal toxicity (17–19). Moreover, the identification of ion channels selectively or preferentially located in the atria has led to the development of atrial selective antiarrhythmic agents, in order to minimize simultaneous unwanted effects on the ventricles (20–25). Finally, there is increasing evidence supporting antiarrhythmic properties of traditionally non-antiarrhythmic agents (i.e. upstream therapies), such as inhibitors of the renin-angiotensin-aldosterone system (25–28), statins (29,30), and omega-3 fatty acids (31–33).

A large number of preclinical studies have evaluated these compounds, elucidating their pharmacokinetic and pharmacodynamic properties, which can be found summarized in recent elegant reviews (16). However, it is not unusual in cardiology that what may appear a promise in preclinical studies actually reveals serious pitfalls in the clinical setting. For example, suppression of ventricular ectopy after myocardial infarction by antiarrhythmic drugs demonstrated very effective in preclinical studies led to an unexpected increase in mortality in clinical studies (9).

Therefore, the purpose of this article is to review the current advances in antiarrhythmic drug therapy for AF, discussing the pharmacologic properties with a critical appraisal on potential clinical applications of antifibrillatory agents in the pipeline.

Antiarrhythmic drug analogs of amiodarone

Amiodarone is the most effective available drug for restoring and maintaining sinus rhythm (14,15,34,35) and is the most widely used antiarrhythmic drug for AF. Amiodarone is an action potential-prolonging agent, with little effects on depolarization and conduction velocity. From an electrophysiological standpoint, amiodarone blocks multiple ion channels, including Na⁺, K⁺, and Ca²⁺ channels (36). Moreover, it has significant vasodilatatory and antiadrenergic properties (36,37). Vasodilatatory effects are mediated by the major metabolite of amiodarone, N-desethylamiodarone (38), and may partially contribute to the clinical efficacy of amiodarone.

The major drawback of amiodarone consists in its high degree of lipophilicity, which, together with its long half-life, causes accumulation and toxicity in several organ systems (13). In recent years, intense research has focused on the development of analogs of amiodarone, with the intent of replicating its antiarrhythmic effects while minimizing the risk of toxicity.

Dronedarone

Dronedarone has a structure similar to amiodarone, but with two important differences: dronedarone does not have iodine moieties, and has a methylsulfonamide group incorporated in the benzofuran ring. These two features result in a decreased potential of thyroid dysfunction and in decreased lipophilicity and half-life, thus reducing the tissue accumulation of the drug (39). Dronedarone preserves the multichannel blockade features of amiodarone, together with its vasodilatatory and antiadrenergic effects (40). In cardiomyocytes, dronedarone blocks inward currents, including rapid Na⁺, and L- and T-type Ca²⁺ currents, and also outward K⁺ currents, and the pacemaker current I_f (40,41). Translating such pharmacodynamic properties into clinically visible effects, dronedarone decreases heart rate and blood pressure and lengthens the PR and QT intervals (40,42). The pharmacokinetics of dronedarone differs significantly from that of amiodarone (40,41). After oral administration with meals, the bioavailability is about 15% due to extensive first-pass metabolism, which is mediated by the hepatic cytochrome CYP3A4. Peak plasma concentration is reached within 4 to 6 hours after oral ingestion, and steady-state levels are reached within 4 to 8 days. The elimination half-life of dronedarone is 13 to 24 hours and is significantly shorter than that of amiodarone, which requires up to several months for complete removal from adipose tissue stores. The extensive metabolism by the hepatic CYP3A4 results in significant interactions with many agents that either inhibit or induce the CYP3A4. Among the most notable cardiovascular drugs that act as inhibitors of the CYP3A4, thus increasing the plasma levels of dronedarone, are non-dihydropyridine calcium channel blockers (i.e. verapamil and diltiazem) and several statins (e.g. simvastatin, atorvastatin). Of note, at variance with amiodarone, dronedarone is not metabolized by CYP2C9 and CYP2C19, which are involved in the metabolism of warfarin. Furthermore, dronedarone is a direct inhibitor of the renal organic cation transport, resulting in an increase in serum creatinine that occurs early after treatment initiation and persists up to 1 week after treatment discontinuation (42,43). This increase in serum creatinine levels, however, does not reflect kidney damage, since the rate of glomerular filtration remains unaffected, when measured with the inulin clearance method (44). Dronedarone is contra-indicated in patients with severe hepatic impairment, while no dosage adjustment is required for patients with renal insufficiency.

The antiarrhythmic effects of dronedarone, as well as its effects on morbidity and mortality in different patient populations, have been tested in multiple randomized clinical trials (Table I) (42, 43,45–49). The Dronedarone Atrial Fibrillation Study After Electrical Cardioversion (DAFNE) trial (42) compared three doses of dronedarone (400, 600, and 800 mg twice daily) with placebo on the end-point of AF recurrence in 270 patients with persistent AF scheduled for elective cardioversion. The median time to first AF recurrence was 60 days in the dronedarone (400 mg twice daily) group, as compared to 5.3 days in the placebo groups (relative risk reduction (RRR) = 55%; 95% confidence interval (CI) 72% to 28%; P = 0.001; absolute risk reduction (ARR) at 6 months = 25%; number needed to treat (NNT) = 4). Surprisingly, no significant reduction in AF recurrence was reported with higher doses of dronedarone (i.e. 600 and 800 mg twice daily), and based on these findings the authors suggest the absence of a dose-response effect. In secondary analyses, dronedarone was shown to reduce ventricular rate in case of AF recurrence (P = 0.0001) as compared to placebo, a result further confirmed in an appropriately designed randomized trial (Table I) (45).

Table I. Randomized clinical trials of dronedarone.

Characteristics	DAFNE (42)	EURIDIS-ADONIS (43)	ERATO (45)	ANDROMEDA (49)	ATHENA (47)	DIONYSOS (48)
Year	2003	2007	2008	2008	2009	2010
Patients enrolled, n	270	1,244	174	627	4,628	504
Design of the trial	Dronedarone versus placebo	Dronedarone versus placebo	Dronedarone versus placebo	Dronedarone versus placebo	Dronedarone versus placebo	Dronedarone versus amiodarone
Clinical scenario	Persistent AF undergoing CV	SR with at least 1 AF/AFL in the preceding 3 months	Permanent AF	Recent decompensated HF with NYHA class III/IV, LVEF<35%	SR with AF/AFL in the previous 6 months and ≥1 additional risk factor^a	Persistent AF with indication to CV
Primary end-point	Time to first AF	Time to first AF	Ventricular rate during exercise	Death or hospitalization for HF	Death or cardiovascular hospitalizations	Time to first AF or drug discontinuation
Treatment regimen	400–800 mg b.i.d.	400 mg b.i.d.	400 mg b.i.d.	400 mg b.i.d.	400 mg b.i.d.	400 mg b.i.d.(D) versus 600 mg daily for 28 days and 200 mg daily thereafter (A)
Exclusion criteria	Permanent AF, HF with NYHA class III–IV, LVEF<35%	Permanent AF, HF with NYHA class III–IV, renal failure	HF with NYHA class III–IV	Recent MI, acute pulmonary edema	Permanent AF, HF with NYHA class IV, decompensated HF	HF with NYHA class III–IV; unstable angina or recent MI
Follow-up, months	6	12	6	2	21	7
Results	60 days (D) versus 5.3 days (P)	116 days (D) versus 53 days (P)	Reduction of ventricular rate by 25.6 bpm (D) versus 2.2 bpm (P)	Death: 8.1% (D) versus 3.8% (P); Hospitalizations: 22.9% (D) versus 15.7% (P)	Hospitalizations: 36.9% (D) versus 29.3% (P); Death: 5% (D) versus 6% (P)	AF recurrence: 63.5% (D) versus 42% (A); Drug discontinuation: 10.4% (D) versus 13.3% (A)

^aHypertension, diabetes, prior cerebrovascular accident or systemic embolism, left atrium diameter ≥ 50 mm, LVEF < 40%.

AF = atrial fibrillation; AFL = atrial flutter; CV = cardioversion; SR = sinus rhythm; HF = heart failure; NYHA = New York Heart Association; LVEF = left ventricular ejection fraction; MI = myocardial infarction; D = dronedarone; P = placebo; A = amiodarone; b.i.d.= twice daily.

With regard to safety, a gradual increase in the rate of premature drug discontinuation was observed with increasing doses of dronedarone, with 3.9%, 7.6%, and 22.6% discontinuation rates in the 400 mg b.i.d., 600 mg b.i.d., and 800 mg b.i.d.groups. The high rate of drug discontinuation associated with the 600 mg b.i.d.and 800 mg b.i.d.doses of dronedarone may also provide a more plausible explanation to the apparent lack of antiarrhythmic superiority of higher doses of dronedarone, since analyses were carried out on the intention-to-treat basis. The most frequent causes of premature drug discontinuation were gastrointestinal side-effects (diarrhea, nausea, and vomiting). Interestingly, although a progressive lengthening of the QT interval was observed with increasing doses of dronedarone, no torsade de pointes was reported. Again, this is an important similarity with amiodarone, which is known to cause prolongation of the QT interval with very small torsadogenic potential (41). After the results of DAFNE, a dose of dronedarone of 400 mg twice daily was selected for subsequent studies.

Two large trials tested the efficacy of dronedarone for the maintenance of sinus rhythm in patients with history of AF, namely the European Trial in Atrial Fibrillation or Flutter Patients Receiving Dronedarone for the Maintenance of Sinus Rhythm (EURIDIS) and the American-Australian-African Trial with Dronedarone in Atrial Fibrillation or Flutter Patients for the Maintenance of Sinus Rhythm (ADONIS) (43). EURIDIS and ADONIS shared the same enrollment criteria and end-points, and their results were ultimately published in a single paper (43). Taken together, EURIDIS and ADONIS included a total of 1,239 patients that were in sinus rhythm for at least 1 hour at the time of randomization and had history of at least one documented episode of AF in the preceding 3 months. Patients were randomized to receive 400 mg of dronedarone twice daily or placebo for 1 year. The primary end-point of the study was time to first AF or atrial flutter recurrence, as assessed by transtelephonic ECG monitoring or 12-lead ECG. Dronedarone significantly prolonged the time to first AF/atrial flutter recurrence as compared to placebo both in the EURIDIS and in the ADONIS trials, with a median time to recurrence of 96 versus 41 days (P = 0.01) in the EURIDIS and of 158 versus 59 days (P = 0.002) in the ADONIS. These figures account for a statistically significant 25% RRR of AF/atrial flutter recurrence within 1 year associated with dronedarone use, although the ARR (11.1%) was quite small for a soft end-point such as AF/atrial flutter recurrence.

Despite such encouraging results, caution should be exercised when generalizing the results of randomized trials with dronedarone to the general population with AF. In fact, an analysis of treatment effects of dronedarone shows that the benefit of this compound is not the same across all the spectrum of patients with AF. Plotting the total number of AF/atrial flutter events avoided every 1,000 patients treated with dronedarone against the base-line risk of AF recurrence with placebo observed in each dronedarone trial, it appears that the antifibrillatory effectiveness of dronedarone is directly dependent on the base-line risk of AF recurrence (Figure 1). Moreover, it is predictable that, when base-line annual risk of AF recurrence is lower than 60%–65%, dronedarone would have minimal—if any—clinically meaningful antiarrhythmic effectiveness. On the other hand, dronedarone is not a completely harmless agent. The yearly risk of drug discontinuation because of a serious adverse effect was quite similar across all dronedarone trials, and approximates 12%. Therefore, taking into account also the risk of adverse events, dronedarone predictably loses clinically meaningful antifibrillatory effectiveness—and is potentially harmful—when the base-line annual risk of AF recurrence is lower than 65%–70% (Figure 1). The results of such analysis, notwithstanding the limitations inherent to statistical prediction models, call for a trial with dronedarone in a population of patients with annual risk of AF recurrence lower than that reported in published trials, before considering appropriate the use of dronedarone in the general population with AF.

Figure 1. Benefits and risks of dronedarone to prevent the occurrence of atrial fibrillation (AF) or atrial flutter (AFL). The number of AF/AFL events avoided (black squares) and the number dronedarone discontinuations for a serious drug-related adverse event (gray circles) every 1,000 patients treated are plotted against the annual risk of AF recurrence with placebo in each trial. Solid lines indicate linear regressions of treatment effects. Extrapolation of regression lines (dotted lines) toward their intersection suggests that the antiarrhythmic benefits of dronedarone may not outweigh the risks associated with its use when the patients' annual risk of AF recurrence is smaller than 65%–70% (intersection of dotted lines). AEs = adverse events; ADONIS (43) = American-Australian-African Trial with Dronedarone in Atrial Fibrillation or Flutter Patients for the Maintenance of Sinus Rhythm; EURIDIS (43) = European Trial in Atrial Fibrillation or Flutter Patients Receiving Dronedarone for the Maintenance of Sinus Rhythm; DAFNE (42) = Dronedarone Atrial Fibrillation Study After Electrical Cardioversion.

Of note, amiodarone displays a completely different antiarrhythmic behavior, with greater effectiveness than dronedarone, as demonstrated by a greater number of AF events avoided every 1,000 patients treated (Figure 2), and consistency of treatment effect across different base-line risks of AF recurrence. The antiarrhythmic superiority of amiodarone compared with dronedarone has been recently confirmed in a randomized trial in patients with persistent AF (Table I) (48). Moreover, the antifibrillatory effectiveness of amiodarone predictably outweighs the risks of serious adverse events across all AF patient populations (Figure 2).

Figure 2. Benefits and risks of amiodarone to prevent the occurrence of atrial fibrillation (AF) or atrial flutter (AFL). The number of AF/AFL events avoided (black squares) and the number of amiodarone discontinuations for a serious drug-related adverse event (gray circles) every 1,000 patients treated are plotted against the annual risk of AF recurrence with placebo in each trial. Solid lines indicate linear regressions of treatment effects. Extrapolation of regression lines (dotted lines) does not show intersection, suggesting that the antiarrhythmic benefits of amiodarone outweigh the risks associated with its use across all patient populations. AEs = adverse events; GEFACA (34) = Grupo de Estudio de Fibrilacion Auricular Con Amiodarona; SAFE-T (15) = Sotalol Amiodarone Atrial Fibrillation Efficacy Trial; Kochiadakis et al. (35); Channer et al. (14).

Dronedarone has also been reported to reduce the risk of death or hospitalization for cardiovascular causes in patients with AF. Subgroup analyses of EURIDIS and ADONIS suggested a 20% lower rate of death or hospitalization in the dronedarone arm, largely related to a reduction in the risk of cardiovascular hospitalization. To further evaluate the effect of dronedarone on death or hospitalizations, a large trial on AF/atrial flutter patients with moderate to high risk of cardiovascular death was designed, that is, the ATHENA (A Placebo-Controlled, Double-Blind, Parallel Arm Trial to Assess the Efficacy of Dronedarone 400 mg b.i.d.for the Prevention of Cardiovascular Hospitalization or Death from Any Cause in Patients with Atrial Fibrillation/Atrial Flutter) (47). The design of the ATHENA is the result of the modern ‘change of philosophy’ in AF treatment evaluation, which takes into account important clinical end-points beyond sinus rhythm restoration and maintenance, together with a peculiar focus on safety (6,7).

ATHENA randomized a total of 4,628 patients with AF/atrial flutter and at least one additional risk factor, including hypertension, diabetes, prior cerebrovascular accident, an enlarged left atrium (i.e. ≥50 mm), or a left ventricular ejection fraction <40%, to receive dronedarone or placebo. The trial was adequately powered to address the primary composite end-point of hospitalization for cardiovascular cause or death from any cause. After a mean follow-up of 21±5 months, primary end-point was reached in 31.9% of patients allocated to dronedarone and in 39.4% of patients receiving placebo (RRR 24.2%; 95% CI 16.5%–31.2%; P< 0.001; NNT = 13). Such reduction in primary end-point was largely driven by a reduction of cardiovascular hospitalizations, whereas no significant reduction in all-cause mortality was observed (HR 0.844; 95% CI 0.660–1.080; P = 0.17). Importantly, the upper bound of the 95% CI for the relative risk of all-cause mortality reasonably excluded any clinically significant increase in the risk of death associated with dronedarone use. A post-hoc analysis suggested also a protective effect of dronedarone on ischemic stroke, which appeared independent of concomitant use of antithrombotic therapy and was more pronounced in patients with higher base-line risk of thromboembolic complications, as assessed by the CHADS₂ score (50). As the latter would be the subset of patients with the worst prognosis for thromboembolic complications, the finding would be easy to account for by saying—as the ATHENA authors did—that dronedarone has its greatest advantage in patients at higher risk of thromboembolism (50). From a methodological standpoint, this approach to the comparison of subgroups is highly flawed. Indeed, these results are based on a post-hoc subgroup analysis, not pre-specified in the original trial, with resulting P values not adjusted for multiple comparisons (47,50). Therefore, any reported result is very difficult to interpret, because if we go on testing long enough we will inevitably find something which is ‘significant’, and such a ‘significant’ result may be the one in a mass which we expect by chance alone (51).

Moreover, the beneficial effect of dronedarone on death or hospitalization reported in the ATHENA is not applicable to all patient populations. In fact, another trial that aimed at assessing the effect of dronedarone on cardiovascular hospitalization and mortality in a high-risk population led to disappointing results (49). Such trial was the Antiarrhythmic Trial with Dronedarone in Moderate to Severe CHF Evaluating Morbidity Decrease (ANDROMEDA), which compared dronedarone to placebo in a population of patients with a recent episode (<1 month) of decompensated congestive heart failure on the primary end-point of death from any cause or hospitalization for worsening heart failure. The ANDROMEDA was designed with the purpose of formally testing the safety of dronedarone in a high-risk population of patients and, as such, was not carried out in patients likely to receive the drug in clinical practice (i.e. patients with AF/atrial flutter). ANDROMEDA had to enroll a total of 1,000 patients. However, 7 months after study initiation, the trial was stopped with only 627 patients randomized, due to an excess mortality in the dronedarone group (i.e. 25/310 deaths in the dronedarone group versus 12/317 in the placebo group; HR 2.13; 95% CI 1.07–4.25; P = 0.03). Although such excess in mortality is likely over-estimated by the small number of deaths occurring over a short period of time in an under-powered sample of patients, the results of the ANDROMEDA trial raised important safety concerns regarding dronedarone use.

Putting together the results of dronedarone trials on the end-point of death or hospitalizations, dronedarone benefit appears inversely proportional to the base-line risk of the patient population, with a potential harm in high-risk patients, such as those included in the ANDROMEDA (Figure 3). Accordingly, the American Food and Drug Administration (FDA) has contra-indicated dronedarone in high-risk patients, such as those enrolled in the ANDROMEDA (i.e. patients with class IV heart failure, or class II to III heart failure who had a recent decompensation) (40,49), and has recommended its use for sinus rhythm maintenance in patients with left ventricular ejection fraction greater than 35% (40).

Figure 3. Effects of dronedarone on death or hospitalization for cardiovascular causes in different groups of patients. The number of death or hospitalization events (black squares) avoided (upper panel) or provoked (lower panel) every 1,000 patients treated is plotted against the annual risk of death or hospitalization with placebo in each trial. Solid line indicates linear regression of treatment effect and suggests an inverse relationship between dronedarone-associated benefits and base-line risk of patients, with harm for patients with an annual risk of death or hospitalization greater than 40%–45% (intersection of the regression line with the xaxis). The greatest benefit is observed in patients with low-to-moderate (25%–35%) annual risk of events. EURIDIS (43) = European Trial in Atrial Fibrillation or Flutter Patients Receiving Dronedarone for the Maintenance of Sinus Rhythm; ADONIS (43) = American-Australian-African Trial with Dronedarone in Atrial Fibrillation or Flutter Patients for the Maintenance of Sinus Rhythm; ATHENA (47) = A Placebo-Controlled, Double-Blind, Parallel Arm Trial to Assess the Efficacy of Dronedarone 400 mg b.i.d.for the Prevention of Cardiovascular Hospitalization or Death from Any Cause in Patients with Atrial Fibrillation/Atrial Flutter; ANDROMEDA (49) = Antiarrhythmic Trial with Dronedarone in Moderate to Severe CHF Evaluating Morbidity Decrease.

On the basis of the available evidence, however, we believe it reasonable to restrict the use of dronedarone to patients with moderate to high risk of AF recurrence—in whom the antiarrhythmic effectiveness clearly outweighs the risk of drug discontinuation because of adverse events (Figure 1)—and low risk of death or hospitalization, in order to maximize the benefit on death and hospitalization (Figure 2). Further studies on populations with lower risk of AF recurrence (<60% at 1 year) and moderate to high risk of death or hospitalization (40%–50% at 1 year), are warranted before considering appropriate the use of dronedarone across all the spectrum of AF patient populations.

Budiodarone

Budiodarone is an amiodarone analog that maintains iodine moieties but contains an ester modification that allows extensive metabolism by tissue esterases rather than by hepatic cytochromes (19). As a result, the half-life of budiodarone is significantly shorter than that of amiodarone (7 hours), while electrophysiological properties are similar (19). Only limited clinical data on budiodarone are available. Arya and colleagues evaluated the effect of budiodarone on 6 patients with paroxysmal AF and permanent pacemaker (52). Patients underwent sequential 2-week periods of treatment with increasing doses of budiodarone (200 mg b.i.d., 400 mg b.i.d., 600 mg b.i.d., and 800 mg b.i.d.), preceded and followed by a week of drug wash-out. The primary end-point was reduction of AF burden, defined as percent of time in AF, as assessed by pacemaker interrogation. At base-line, mean AF burden was 20.3% ±14.6% and decreased significantly with all doses of budiodarone, with greater treatment effects with increasing budiodarone doses. Moreover, budiodarone decreased also the duration of AF episodes, again with a dose-effect response.

These encouraging results were confirmed in a recent phase II clinical trial, the Paroxysmal Atrial Fibrillation Study with Continuous Atrial Fibrillation Logging (PASCAL) (53). In this trial, a total of 72 patients with paroxysmal AF and dual chamber permanent pacemaker with advanced rhythm-monitoring capabilities were randomized to placebo or increasing doses of budiodarone (200, 400, and 600 mg twice daily). Again, the primary end-point was decrease of AF burden as assessed by pacemaker interrogation. Increasing doses of budiodarone were associated with progressively greater reductions of AF burden (RRR 10%, P = 0.16 for the 200 mg b.i.d.; RRR = 54%, P = 0.015 for the 400 mg b.i.d.; and RRR = 75%, P = 0.005 for the 600 mg b.i.d.). With regard to safety, patients allocated to budiodarone experienced an increase of the thyroid-stimulating hormone, reversible after drug discontinuation. Moreover, an increase in serum creatinine levels was reported, likely due to a dronedarone-like inhibitory effect on renal organic cation transport. Interestingly, no QT prolongation was observed during the study course. On the other hand, the selection of patients with a permanent pacemaker may have limited the evaluation of some of the possible side-effects of budiodarone, such as depression of atrioventricular conduction velocity (19).

In conclusion, budiodarone is a promising amiodarone analog. However, further studies on broader populations of patients and with longer follow-up durations are necessary to better evaluate the effectiveness and safety of this new drug in AF.

Celivarone

Celivarone is a non-iodinated amiodarone analog that is currently undergoing extensive evaluation in clinical trials as an antiarrhythmic agent for both supraventricular and ventricular arrhythmias (17,54–56). Experimental data have shown that celivarone exhibits electrophysiological and hemodynamic effects similar to amiodarone, with acute antifibrillatory efficacy after intravenous and oral loads (54). A phase II dose-ranging study, the Maintenance of Sinus Rhythm in Patients with Recent Atrial Fibrillation or Flutter (MAIA), randomized 673 patients with a previous episode of AF or atrial flutter to four doses of celivarone (50, 100, 200, and 300 mg once daily) or placebo. The greater reduction of AF/atrial flutter recurrence was observed with the 50 mg dose of celivarone, as compared to placebo (52.1% versus 67.1%; P = 0.055) (55). No enhanced efficacy was reported with higher doses of celivarone, an effect similar to that already observed with dronedarone (42). With regard to safety, no torsade de pointes was reported, with a quite low rate of celivarone discontinuation because of adverse events (6.2% to 8% across different celivarone daily doses) (55). Recently, another phase II trial, the Controlled Dose Ranging Study of the Efficacy and Safety (CORYFEE), tested the efficacy of 300 mg or 600 mg once daily doses of celivarone for the conversion to sinus rhythm in patients with AF/atrial flutter undergoing planned electrical cardioversion. The results of the CORYFEE are not yet available.

Atrial selective antiarrhythmic drugs

The development of agents with selective pharmacologic effect on ion channels expressed in the atria but not in the ventricles is a major focus of modern antifibrillatory drug development. Many different ionic currents contribute to the action potential of human atrial myocytes, and the majority are shared with ventricular myocytes (Figure 4). However, several ionic currents are expressed predominantly or selectively in the atria (Table II) and constitute attractive targets for the development of atrial selective antiarrhythmic agents. To date, research on atrial selective compounds has focused mainly on two repolarization currents: the ultra-rapid potassium outward current (I_Kur) and the inwardly rectifying acetylcholine-sensitive current (I_KACh) (Figure 4). The latter is constitutively activated in an agonist-independent manner in patients with chronic AF (57), and blockade of its activity has been suggested to contribute to the antiarrhythmic effectiveness of several older antiarrhythmic compounds, such as flecainide (58).

Figure 4. Ionic currents involved in the action potential of atrial myocytes. Atrial selective currents are highlighted in gray.

Table II. Summary of atrial selective agents.

Agent	Target	Route	Stage of development	Notes
Vernakalant	Blocker of IKur, Ito, INa, IKr	IV, oral^a	Completed phase III studies; recently approved by the EMEA (only IV route)	High conversion in recent onset AF (<72 h); ineffective in atrial flutter. No data on patients with heart failure
Ranolazine	Blocker of late INa, IKur	IV, oral	Completed phase III studies as anti-ischemic agent; approved as antianginal agent	Reduces the occurrence of AF in subgroup analyses; no randomized trial on AF available
Xention	Blocker of IKur	IV, oral	Phase II	Highly selective IKur blocker but insufficient data
AZD7009	Blocker of IKur, Ito, INa, IKr	IV	Abandoned during phase II	Potent IKr blocker; possible high risk of proarrhythmia
AVE0118	Blocker of IKur, Ito, IACh	IV	Abandoned during phase IIa	Undisclosed reasons for interruption
AVE1231	Blocker of IKur, Ito, IACh	IV, oral	Phase I	Unselective blockade of atrial-selective targets; also blocks IKr, IKs, IKATP, ICaL, INa
NIP-141	Blocker of IKur, Ito, IKACh, ICaL	IV	Phase I	Multichannel blocker with possible anti-remodeling properties

Vernakalant is manufactured by Cardiome; Ranolazine is manufactured by CV Therapeutics.

^aUnder development.

IV = intravenous; AF = atrial fibrillation; AZD = Astra Zeneca; AVE = Sanofi Aventis; NIP = Nissan Pharmaceuticals.

Other atrial-selective molecular targets under active investigation include atrial-selective Na⁺ channel (59), small-conductance K⁺ channels (60), and other repolarization channels, such as the two-pore K⁺ channel TASK (61).

To date, truly selective modulators of atrial ion channels have not yet been developed, and most of the currently available ‘atrial selective’ agents are actually multichannel blockers with higher affinity to atrial ion channels.

Vernakalant

Vernakalant is among the most promising and most investigated atrial selective agents for the treatment of AF (62–72). Vernakalant acts through multichannel blockade, including atrial selective repolarization currents (I_Kur and I_KACh), but also several other currents, such as I_to and I_Na with a frequency-dependent effect (Figure 4) (62). On the other hand, blockade of I_Kr and I_Ks repolarization currents, which are highly expressed in the ventricles, is minimal. As a result, vernakalant slows the conduction velocity and prolongs refractoriness within the atria, with small effects on the ventricles (64). The atrial-specific repolarization-delaying effect appears more prominent in patients with paroxysmal rather than chronic AF (73), which is possibly related to a higher degree of electrical remodeling in the latter group.

Due to the short half-life (2 h), vernakalant has been mostly investigated as an intravenous agent (62,66,67), although an oral slow-release preparation has been developed and is undergoing phase III clinical evaluation (67). After intravenous administration, vernakalant undergoes metabolism by the hepatic cytochrome CYP2D6 (66). The extent of such metabolism is unclear, since plasma concentrations of vernakalant have been reported unaffected by the administration of CYP2D6 inducers or inhibitors (66,67).

Six randomized studies have evaluated the effect of intravenous vernakalant for the acute cardioversion of persistent AF (65,68–72) (Table III). These studies led to the recent (September 2010) approval of vernakalant for the acute conversion of AF by the European Medicines Agency (EMEA). In a phase II multicenter placebo-controlled trial, the Randomized, Controlled Trial of RSD1235, a Novel Antiarrhythmic Agent, in the Treatment of Recent Onset Atrial Fibrillation (CRAFT) (69), 56 patients with persistent AF were randomly assigned to one of two vernakalant intravenous doses (i.e. 0.5 mg/kg or 2 mg/kg in 10 minutes) or to placebo. A repeat infusion of vernakalant (or placebo) at a rate of either 2 mg/kg (for patients initially treated with the dose of 0.5 mg/kg), or 3 mg/kg (for patients initially treated with the dose of 2 mg/kg), was permitted if cardioversion was not reached within 30 minutes from the first infusion. The primary end-point was cardioversion of AF during or within 30 minutes after drug infusion. Patients allocated to the higher intravenous dose of vernakalant (i.e. 2 mg/kg followed by 3 mg/kg) had a significantly higher rate of conversion to sinus rhythm than those receiving placebo (61% versus 5%; P< 0.0005; RRR 59%; ARR 56%; NNT 1.8), together with a shorter median time-to-conversion (14 versus 162 minutes; P = 0.016). The encouraging results of the CRAFT led to four other clinical trials, the Atrial Arrhythmia Conversion Trials (ACTs) (Table III) (68,71,72,74). The ACT 1 randomized a total of 336 patients with persistent AF of either short (≤7 days) or long duration (8–45 days), to vernakalant (3 mg/kg in 10 minutes, followed by another 2 mg/kg if cardioversion was not reached within 15 minutes), or placebo (68). The primary end-point was conversion to sinus rhythm within 90 minutes of drug initiation. The conversion rate with vernakalant in patients with short-duration persistent AF was 51.7%, compared to 4% with placebo (P< 0.001). On the other hand, only 7.9% of patients with long-duration persistent AF converted to sinus rhythm with vernakalant, compared to 0% in the placebo arm (P = 0.09). The drug was well tolerated, with only transient minor side-effects (mainly dysgeusia and sneezing), and no torsade de pointes was reported. Such results were confirmed in two other trials, the ACT 3, which included a cohort of patients similar to that of the ACT 1, and the ACT 2, that focused on patients after cardiac surgery (65,72). Of note, ACT 3 and 2 included also a small subgroup of patients with atrial flutter, in whom vernakalant consistently failed to show an antiarrhythmic efficacy (65,72) (Table III). Finally, the ACT 4 was designed to further evaluate the safety of intravenous vernakalant in recent-onset AF patients, with results quite similar to the previous ACT studies. Of note, patients with decompensated heart failure were excluded from all ACT trials.

Table III. Randomized clinical trials of vernakalant.

Characteristics	CRAFT (69)	ACT-I (68)	ACT-II (65)	ACT-III (72)	AVRO (70)
Year	2004	2008	2009	Not yet published	Not yet published
Patients enrolled, n	56	336	161	276	254
Design of the trial	Vernakalant versus placebo	Vernakalant versus placebo	Vernakalant versus placebo	Vernakalant versus placebo	Vernakalant versus amiodarone
Clinical scenario	Persistent AF undergoing CV	Persistent AF undergoing CV	Postoperative AF/AFL	Persistent AF/AFL undergoing CV	Persistent AF undergoing CV
Primary end-point	Conversion of AF	Conversion of AF within 90 min of drug initiation	Conversion of AF within 90 min of drug initiation	Conversion of AF within 90 min of drug initiation	Conversion of AF within 90 min of drug initiation
Treatment Regimen	0.5 mg/kg in 10 min followed by 2 mg/kg if no CV or 2 mg/kg in 10 min followed by 3 mg/kg if no CV	3 mg/kg in 10 min followed by 2 mg/kg if no CV	3 mg/kg in 10 min followed by 2 mg/kg if no CV	3 mg/kg in 10 min followed by 2 mg/kg if no CV	3 mg/kg in 10 min followed by 2 mg/kg if no CV
Exclusion criteria	HF	HF with NYHA class IV	HF with NYHA class IV	HF	Not retrievable
Results	At 1 h: 53% (V) versus 5% (P); at 24 h: 79% (V) versus 45% (P)	AF ≤7 days: 51.7% (V) versus 4% (P); AF 8–45 days: 7.9% (V) versus 0% (P)	45% (V) versus 15% (P); no significant CV of AFL to SR	AF ≤7 days: 51.2% (V) versus 3.6% (P); AF 8–45 days: 9.4% (V) versus 2.7% (P); no significant CV of AFL to SR	51.7% (V) versus 5.2% (A)

AF = atrial fibrillation; AFL = atrial flutter; CV = cardioversion; SR = sinus rhythm; HF = heart failure; NYHA = New York Heart Association; V = vernakalant; P = placebo; A = amiodarone.

The relative effectiveness of vernakalant compared to other available antiarrhythmic agents is unknown, and studies of direct comparison with agents currently recommended for AF cardioversion are warranted. However, since currently available antiarrhythmic drugs have been evaluated extensively against placebo, an indirect comparison analysis of effectiveness can be performed. On indirect comparison, vernakalant appears one of the most effective antiarrhythmic agents for the acute conversion of AF, with effectiveness comparable to intravenous flecainide (Figure 5). Amiodarone, on the other hand, is the least effective antiarrhythmic agent in this setting, with a clearly smaller effectiveness compared to vernakalant (Figure 5).

Figure 5. Indirect comparison of treatment effects of different intravenous antiarrhythmic agents for the acute conversion of atrial fibrillation (AF). SR = sinus rhythm.

Surprisingly, amiodarone has been the only antiarrhythmic agent directly compared with vernakalant in a randomized study. This study was the A phase III prospective, Randomized, double-blind, active-controlled, multicenter, superiority study of Vernakalant injection versus amiodarone in subjects with Recent Onset atrial fibrillation (AVRO) (75), and the results were presented at the last Heart Rhythm Society annual meeting (70). In total, 254 patients with AF were enrolled in the study, with 116 patients in each group receiving either a 10-minute infusion of 3 mg/kg of vernakalant, eventually followed by another 10-minute infusion of 2 mg/kg if conversion to sinus rhythm was not reached within 15 minutes after the first infusion, or a 60-minute infusion of 5 mg/kg of amiodarone followed by a maintenance infusion of 50 mg over another 60 minutes. At 90 minutes after the first infusion, the proportion of patients converted to sinus rhythm was higher in the vernakalant arm (51.7%) than in the amiodarone arm (5.2%; P< 0.0001). Moreover, treatment with vernakalant resulted in a more rapid conversion to sinus rhythm, with a median conversion time of 11 minutes, and was associated with a higher rate of symptom relief compared to amiodarone (53.4% versus 32.8%, respectively; P = 0.0012). Also in this study, there was no reported case of drug-induced ventricular arrhythmias.

As previously mentioned, the results of the AVRO study were largely predictable by an accurate analysis of the available evidence (Figure 5). Further direct comparison studies with other available and more effective antiarrhythmic agents, such as intravenous flecainide, are certainly warranted before considering vernakalant of clear incremental therapeutic value for the acute conversion of AF.

In conclusion, vernakalant appears a new effective and safe drug for rapid termination of AF, although the real incremental value in this setting compared to other available antiarrhythmic agents merits further investigation. On-going studies with oral vernakalant will also reveal whether this agent may be useful for sinus rhythm maintenance.

Ranolazine

Ranolazine is predominantly a late Na⁺ current (I_Na) blocker, with effects also on the ultra-rapid delayed rectifier K⁺ current (I_Kur) (76,77). The effect on late I_Na accounts for significant anti-ischemic properties, which have been demonstrated both in preclinical and clinical studies (76,78). Late I_Na is up-regulated during myocardial ischemia, which leads to increased cytosolic calcium concentration due to enhanced activity of Na⁺/Ca²⁺ exchanger, and increased susceptibility to delayed after-depolarization (76,79). Recent evidences suggest that late I_Na is increased also in patients with AF (80), and ranolazine has been shown capable of reversing such late I_Na current up-regulation (80).

As an anti-ischemic agent, up to 1,000 mg twice daily of ranolazine is administered orally, with variable bioavailability due to extensive first-pass metabolism, mainly CYP3A4-dependent (23,76,79,81). At concentrations within the therapeutic range for anti-ischemic activity (2–6 μM) (23,79), ranolazine produces a much greater depression of atrial versus ventricular I_Na, which is of potential interest in the treatment of AF. Interestingly, in experimental models of isolated canine pulmonary vein sleeve, ranolazine has been shown to suppress excitability and triggered activity, to terminate AF induced by acetylcholine and isoproterenol, and to prevent AF reinduction (23,82).

In a post-hoc analysis of the Metabolic Efficiency With Ranolazine for Less Ischemia in Non-ST-Elevation Acute Coronary Syndrome (MERLIN)-Thrombolysis in Myocardial Infarction (TIMI) 36 trial (78), which randomized more than 6,000 patients with acute coronary syndrome to ranolazine or placebo, treatment with ranolazine was associated with lower rate of new-onset AF (3.1% versus 4.3%; P = 0.01) (81). With regard to safety, ranolazine resulted in up to 6 ms mean increase in the QTc interval, which is likely due to its effects on the I_Kur current, although no case of drug-induced torsade de pointes was reported (78,81).

The potential usefulness of ranolazine in AF has been recently confirmed in a consecutive series of seven patients with recurrent AF after at least one failed antiarrhythmic agent (dofetilide, sotalol, propafenone, flecainide, or amiodarone) (83). In this series, ranolazine controlled the arrhythmia and maintained sinus rhythm at 27±11 weeks of follow-up in 5/7 patients (83).

Based on these preliminary findings, ranolazine may represent a safe atrial selective agent to treat AF, although appropriate randomized trials are warranted to evaluate its potential therapeutic role against AF.

Other atrial selective agents

Xention (XEN-D0101) is an atrial repolarization delaying agent, with a selective blockade of I_Kur(84). Xention has been demonstrated to prolong atrial refractoriness and suppress AF in canine models (84), with similar electrophysiological effects also on isolated human atrial myocytes (85). The phase I clinical development of Xention was initiated in October 2006.

AZD7009 exhibits mixed potassium and sodium ion channel-blocking properties and exerts potent effects on atrial refractoriness (86–88), without affecting significantly the ventricular refractoriness and QT interval (24). In a randomized study on 122 patients with persistent AF/atrial flutter undergoing planned cardioversion, intravenous AZD7009 increased the proportion of patients who converted to sinus rhythm in a concentration-dependent manner (25). Only one case of asymptomatic non-sustained polymorphic ventricular tachycardia was reported after AZD7009 infusion (25). Research on AZD7009 was stopped in 2006 for undisclosed side-effects (89).

AVE0118 is an atrial-selective multichannel blocker, with effects on I_Kur, I_KACh, and I_to (90). In animal models of AF, AVE0118 prolongs atrial refractoriness and repolarization in a dose-dependent manner without affecting ventricular repolarization (91,92). Such electrophysiological effects translate into effective conversion of persistent AF (91) and prevention of AF induction (92). Of note, the atrial repolarization delaying effect of AVE0118 appears specific for AF, whereas it actually shortens the atrial action potential duration during sinus rhythm (93). Clinical studies on AVE0118 have stopped in the IIa phase for unspecified reasons (94).

AVE1231 is a molecular congener of AVE0118, with similar unselective blockade of I_Kur (95). AVE1231 is currently in phase I clinical development (94).

NIP-141 is a multiple channel blocker with atrial selective action potential duration prolonging profile (96). In animal models, NIP-141 has been reported to inhibit atrial action potential shortening induced by rapid atrial pacing, with a possible anti-remodeling effect mediated by inhibition of L-type Ca²⁺ channel (97) (Table II).

In conclusion, modulation of ionic currents selectively expressed in the atria represents a main focus of modern antifibrillatory drug development. The main target of this novel drug category is the atrial selective repolarization current I_Kur (98). The clinical efficacy and safety of these compounds warrant adequate investigation in phase III studies.

Selective adenosine receptor agonists

The clinical effects of adenosine are mediated by four purinergic receptors, all with different functions (99,100). The stimulation of A₁ receptors accounts for the negative dromotropic properties of adenosine (99), due to prolonged refractoriness of the atrioventricular nodal conduction system. Such effect has important therapeutic applications in many supraventricular arrhythmias (101). In non-specialized myocytes, however, A1 receptor stimulation determines a shortening of refractoriness, with possible induction of atrial or ventricular fibrillation (101).

In the setting of chronic AF, in which a rate-control strategy is often attempted (102), the negative dromotropic effects of A₁ receptor stimulation may be of interest. Several selective A₁ receptor agonists are in the active phase of development. Tecadenoson is the compound of this class in the most advanced phase of development. The selectivity of tecadenoson on A₁ receptor appeared very high in preclinical studies (103,104), with significant effect on atrioventricular nodal conduction and refractoriness without causing hypotension or negative inotropic effects. A dose-ranging study in 32 healthy humans confirmed a dose-dependent effect on atrioventricular nodal conduction, without affecting the infra-nodal conduction (105). Four patients had drug-induced second- or third-degree atrioventricular block, and AF was induced in three patients.

A phase II clinical study with tecadenoson for rate control of chronic AF was completed in 2009, but results have not yet been published (106).

Upstream therapies for atrial fibrillation

The value of upstream therapies for AF has become increasingly appreciated in recent years, as the pivotal role of different non-electrophysiological triggers and substrates for AF has been elucidated (107). In particular, the exploration of the left atrial substrate has suggested that AF is a self-perpetuating disease, characterized by progressive electrical and tissue structural remodeling, which ultimately leads to loss of atrial myocytes and fibrosis (108,109).

The prevention of atrial structural remodeling through inhibition of biologic pathways leading to tissue fibrosis constitutes the rationale for up-stream therapies for AF. Several traditionally non-antiarrhythmic agents have been suggested for this purpose (Figure 6), including inhibitors of the renin-angiotensin-aldosterone system (28,110,111) and anti-inflammatory agents (29,30,33,112).

Figure 6. Putative molecular effects of upstream therapies for atrial fibrillation. RAA = renin-angiotensin-aldosterone.

Inhibitors of the renin-angiotensin-aldosterone system

Atrial tissue remodeling is greatly enhanced by angiotensin II. Angiotensin II promotes growth of cardiac myocytes, vascular smooth muscle cells, and fibroblasts (113), increases atrial oxidative stress (114), mediates unwanted increase in ventricular oxidative stress induced by rapid atrial rate (115), and modifies electrophysiological properties by indirect actions on ion channels (116). Angiotensin II is also a potent pressor agent and directly stimulates aldosterone secretion, which further promotes atrial fibrosis (117). Moreover, overstimulation of the renin-angiotensin-aldosterone system chronically increases the left ventricular afterload, which ultimately leads to diastolic dysfunction, increased left atrial dimension, and higher risk of developing AF (111,116–118).

With these premises, inhibition of the renin-angiotensin-aldosterone system appears a reasonable strategy to prevent the occurrence of AF. Retrospective analyses of studies with renin-angiotensin system blockers in different cohorts of patients have reported a lower incidence or recurrence of AF in patients treated with renin-angiotensin inhibitors (26,119–122). In a recent large meta-analysis, the greatest benefit of renin-angiotensin inhibitors has been reported for secondary prevention of AF, especially after electrical cardioversion (28). Among primary prevention studies, a benefit was observed among patients with heart failure and those with hypertension and left ventricular hypertrophy, thus suggesting a possible benefit in patients with structural heart disease (123). On the other hand, renin-angiotensin inhibitors did not appear to prevent the occurrence of AF in patients after myocardial infarction (28). Despite these encouraging findings, a trial specifically aimed at assessing the effect of renin-angiotensin system blockade on AF recurrences reported negative results (110). This trial was the Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico-Atrial Fibrillation (GISSI-AF) (110), which randomized 1,442 patients with history of AF to either valsartan (an angiotensin receptor blocker) or matching placebo. At 1-year follow-up, AF recurred in 51.4% of the patients in the valsartan group versus 52.1% in the placebo group (HR 0.98; 96% CI 0.85–1.14; P = 0.83) (110). The results of the GISSI-AF question the effectiveness of renin-angiotensin system blockers for the secondary prevention of AF, although the wide use of other ACE inhibitors in the control group and the short duration of the follow-up may have confounded the results (110). Of note, we also found no improvement in the success rate of radiofrequency catheter ablation of AF in patients treated with renin-angiotensin system inhibitors (124), further supporting a lack of benefit in secondary prevention of AF.

With regard to aldosterone receptor antagonists, preclinical data suggest a potential role of these agents in AF management (111,117), although adequate clinical evidences in this setting are lacking.

In conclusion, inhibitors of the renin-angiotensin-aldosterone system appear an emerging treatment option for AF, mainly due to their significant anti-remodeling properties. Further clinical studies are necessary to identify the population of AF patients who may earn the greatest benefit from these agents.

Anti-inflammatory and anti-oxidant agents

Inflammation and oxidative stress contribute to electrical and structural remodeling of the atria (125–127) and are intimately linked to initiation and perpetuation of AF (127). Several anti-inflammatory and anti-oxidant agents have been tested against AF (29,33,112,128).

Steroids, the paradigm of anti-inflammatory agents, have been associated with a lower incidence of AF in preclinical and clinical studies (112,129), although the high rate of side-effects limits their usefulness as antiarrhythmics.

Inhibitors of the 3-hydroxy-3-methylglutaryl coenzyme A reductase (i.e. statins) are among the most promising anti-inflammatory and anti-oxidant agents with demonstrated antifibrillatory effectiveness (29,31,128,130,131). Multiple clinical studies have suggested an antiarrhythmic effect of statins against AF, especially when this condition is associated with increased inflammation, such as after cardiac surgery or acute coronary syndromes (128). Furthermore, we recently reported that the antifibrillatory effectiveness of statins may extend also to patients with sinus node dysfunction and a permanent pacemaker, a condition characterized by a high incidence of AF (29). Also in this condition, inflammation and oxidative stress have been associated with both structural changes in the atria leading to atrial fibrosis and dilatation, as well as with biochemical and electrophysiological changes in the individual atrial myocytes (29).

A limited amount of evidence supports the usefulness of omega-3 fatty acids against AF (32,33,127,132–134). In a randomized trial on 160 patients undergoing coronary artery bypass surgery, treatment with omega-3 fatty acids reduced the occurrence of postoperative AF by 46% compared to placebo (P = 0.013) (33). Similar findings were confirmed in another study on postoperative cardiac surgery patients (132), while a recent randomized trial failed to show a beneficial effect of omega-3 fatty acids on postoperative AF (133). Pooling together these results, there is inconclusive evidence supporting the use of omega-3 fatty acids to prevent the occurrence of postoperative AF (Figure 7).

Figure 7. Forest plot showing the individual and pooled hazard ratios of postoperative atrial fibrillation comparing therapy with omega-3 fatty acids versus placebo. Square boxes denote hazard ratio; horizontal lines represent 95% confidence interval (CI).

The benefit of omega-3 fatty acids in other populations of patients with AF is undefined and is being actively investigated in adequate randomized trials (135,136). To this regard, we recently reported, in a retrospective analysis of 1,500 patients with AF referred to our institution for pulmonary vein antrum isolation, that treatment with omega-3 fatty acids was associated with lower rate of early AF recurrences (134), thus suggesting a benefit also in patients eligible for AF catheter ablation.

In conclusion, statins represent the most promising anti-inflammatory and anti-oxidant agents with antiarrhythmic effectiveness against AF. Omega-3 fatty acids supplementation has been suggested to lower the incidence of AF, particularly in the postoperative period, although further studies are necessary to verify this finding.

Conclusions

In the upcoming years, plenty of novel pharmacologic strategies will become available to treat AF. Beyond demonstrating their antiarrhythmic effectiveness, research should also focus on identifying the subset of patients who may earn the greatest benefit from these compounds. Moreover, the role of these therapies in relationship to established curative approaches to AF, such as catheter ablation, should also be investigated.

Dronedarone is the first of these novel compounds approved for AF treatment. Its antiarrhythmic effectiveness seems lower in comparison to amiodarone, and the benefit across all the spectrum of AF patients is not clearly demonstrated. Intense research for atrial selectivity has led to the development of vernakalant, which appears very effective for the acute conversion of AF. Finally, upstream therapies have the important advantage of targeting AF substrates. Among these, statins are certainly the most promising ones, although further randomized studies are warranted to verify their real benefit in AF.

Declaration of interest: Dr. Natale has received compensation for belonging to the speakers’ bureau for St Jude Medical, Boston Scientific, Medtronic, and Biosense Webster and has received a research grant from St Jude Medical. The other authors declare no conflicts of interest.