Efficacy of amiodarone for the prevention of atrial fibrillation recurrence after cardioversion

Summary

The restoration and maintenance of sinus rhythm is a desirable strategy for many patients with atrial fibrillation (AF) since it has been associated with improvement in symptoms and a better quality of life. Sinus rhythm can be achieved by pharmacological or electrical cardioversion or after catheter ablation of AF. Despite high rates of successful cardioversion, AF recurrence remains a major challenge. Anti-arrhythmic drug therapy currently plays a significant role in maintaining sinus rhythm after cardioversion. Amiodarone is the most commonly prescribed anti-arrhythmic drug for patients with AF. This is due to its particular electrophysiological properties and superior anti-arrhythmic effects in comparison with other anti-arrhythmic drugs. Understanding the cardiac electrophysiology and arrhythmogenesis mechanisms may result in identification of new targets for anti-arrhythmic therapy. The aim of this article was to review amiodarone’s clinical pharmacology and evaluate evidence supporting amiodarone for treatment and prevention of AF recurrence after cardioversion.

Atrial fibrillation (AF), the most common sustained cardiac arrhythmia, is a major cause of cardiovascular morbidity and mortality due to stroke, heart failure and sudden death.¹ A recent study estimates that in 2016, 7.6 million people over 65 years in the European Union had AF, and by 2060, this will increase by 89% to 14.4 million. This corresponds to a 22% increase in AF prevalence, from 7.8 to 9.5%.²

Several randomised trials investigated the two treatment strategies for patients with AF, rhythm control with restoration and maintenance of sinus rhythm (SR) versus ventricular rate control by lowering the heart rate: AFFIRM (Atrial Fibrillation Follow-up of Rhythm Management), HOT-CAFE (How to Treat Chronic Atrial Fibrillation), PIAF (Pharmacological Intervention in Atrial Fibrillation), RACE (RAte Control versus Electrical conversion) and STAF (Strategies of Treatment of Atrial Fibrillation).^3-7

Results from these studies showed that the rhythm-control strategy was no more efficient than rate control in terms of stroke and mortality. However, restoration and maintenance of normal SR has been associated with an improved tolerance to exercise and a better quality of life. Approaches to restore SR include anti-arrhythmic drugs (AADs), electrical cardioversion and catheter ablation. High recurrence rate of AF has been reported after cardioversion, highlighting the need for AADs. Multiple AADs are available for AF rhythm control, the choice of drug being driven by the AAD’s reported efficacy,

patient co-morbidities and risks of side effects. Amiodarone was first used in cardiology in 1962 as an anti-anginal agent. Five years later it was withdrawn due to its side effects. Afterwards, in 1974 it was re-introduced for its anti-arrhythmic effects both for atrial and ventricular arrhythmias. Amiodarone has been largely used for pharmacological cardioversion of AF and for maintenance of SR after direct current cardioversion (DCC) or catheter ablation. Amiodarone is also particularly used in patients with concomitant AF and impaired left ventricular function or underlying ischaemic heart disease.

This review aimed to evaluate the accumulated data on amiodarone’s efficacy and safety for the treatment and prevention of AF recurrence after cardioversion. It also examined the particular pharmacological characteristics of amiodarone that are thought to play a role in the superior anti-arrhythmic effects in comparison with other anti-arrhythmic drugs.

We conducted a review of the medical literature, which included all relevant articles published between 1978 and October 2019 using the MEDLINE, EMBASE databases and the COCHRANE library. To achieve different aims, we used several MeSH terms for our search strategy. For evaluating the efficacy of amiodarone after cardioversion we used: ‘atrial fibrillation’, ‘atrial flutter’, ‘amiodarone’, and ‘conversion’ or ‘cardioversion’. In order to compare amiodarone with other anti-arrhythmic drugs we added in our search strategy the following MeSH terms: ‘flecainide’, ‘propafenone’, ‘quinidine’, ‘disopyramide’, ‘sotalol’, ‘donedarone’, ‘betablockers’, ‘carvedilol’ ‘metoprolol’. ‘nebivolol’, ‘atenolol’, ‘bisoprolol’. To study the efficacy of amiodarone after catheter ablation, we used the same terms and we added: ‘catheter ablation’ or ‘pulmonary vein isolation’.

We also reviewed the guidelines for AF management published in 2010, 2012 and 2016 for additional references. The literature search resulted in 8 230 publications and we selected only articles published in English or if they had a summary available in English.

To meet our inclusion criteria, studies had to have two groups: a study group and a control group. The study group must have received anti-arrhythmic treatment with amiodarone at an appropriate dosage to prevent episodes of AF and maintain SR. For the control group, either placebo or rate-control drugs (digoxin, calcium channel blockers, beta-blockers) could have been used or no treatment. For the secondary objective of assessing differences between two anti-arrhythmic drugs, the control group could have received any of the anti-arrhythmic drugs other than amiodarone that have shown efficacy on AF.

Pharmacology of amiodarone Pharmacokinetics and pharmacodynamics of amiodarone

Amiodarone is an iodine-rich benzofuran derivative with a structure similar to the thyroid hormones.⁸ Pharmacokinetic characteristics are summarised in Table 1. After oral administration, absorption is limited, the bioavailability ranging from 35 to 65% with extensive individual differences.⁹ Due to its lipophilic nature, it has a large volume of distribution and high concentrations of amiodarone are found especially in adipose tissue but also in highly perfused organs such as the liver, lung, spleen or skin.^10,11

Table 1. Pharmacokinetics of amiodarone.

Pharmacokinetic parameters
Bioavailability (%)	35-65
Binding to plasma proteins (%)	96
Vd (l/kg)	60
Onset of action after oral administration (weeks)	1-3
Onset of action after iv administration (hours)	1-2
Normal therapeutic plasma concentration (g/ml)	1-2.5
Half-life after chronic therapy, mean (days)	56
Half-life after iv administration (hours)	4.2-34.5
Metabolism	CYP3A4, CYP2C8
Excretion	Bile, renal elimination is
	minimal

After oral administration, amiodarone onset of action is delayed, from two to three days, often one to three weeks and even longer. Its elimination half-life is multiphasic as follows: in the first three to 10 days after drug withdrawal, amiodarone plasma concentration initially decreases by 50%, followed by a terminal half-life of 26 to 107 days, with a mean of around 53 days.⁹ Following repeated intake of amiodarone, administered more frequently than its half-life, initial accumulation in tissues is normal until steady state.¹² Without a loading dose, a steadystate concentration is achieved between 130 and 535 days. After long-term administration, amiodarone might accumulate in excess in tissues, which explains why some of the amiodarone adverse effects increase in frequency over time.¹³

Amiodarone undergoes hepatic metabolism through the cytochrome p450 enzyme and is eliminated by biliary excretion, while < 1% is excreted unchanged via the renal route.¹⁴ Amiodarone’s most prominent metabolite is desethylamiodarone (DEA). DEA is a pharmacologically active compound with properties similar to the parent drug, although with a much longer elimination half-life.¹⁵ The normal therapeutic plasma amiodarone and DEA levels usually range between 1.0 and 2.5 μg/ml.¹⁶ However, studies have shown that serum concentrations do not correlate well with amiodarone efficacy or side effects.^17-19

If administered intravenously, amiodarone has distinct pharmacokinetics compared to the oral formulation. Bioavailability becomes 100%, while the peak plasma concentration occurs 15 minutes after injection.¹⁶ The distribution to tissues is rapid, within 30 to 45 minutes the decrease in concentration reaching 10% of the peak value. The metabolism is also different in terms of less DEA being produced.¹⁶

Amiodarone has equally complex pharmacodynamics that are summarised in Table 2.⁹ It has numerous effects and possesses class I, II, III and IV action.¹⁴ The electrocardiographic and electrophysiological effects presented in Table 2 might vary according to the route of amiodarone administration, but these differences are further discussed below.

Table 2. Pharmacodynamics of amiodarone.

	In vitro electrophysiological characteristics		In vivo electrophysiological measurements
		In vivo ECG
Amiodarone mechanism of action		measurements
Class III effect - blocks IK	↑ APD	↑QT	↑ ERP AVN
	↑ ERP		↑ ERP HPS
			↑ TERPA TERPV
Class I effect - blocks INa	↓ SN auto- maticity	↓ sinus rate	↑ AH
Class II effect - anti-adrenergic	↓ CV	↑ PR	↑ HV
Class IV effect - blocks ICa,L		↑ QRS

↑= increase; ↓ = decrease; ECG = electrocardiographic; APD = action potential duration; ERP = effective refractory period; SN = sinus node; CV = conduction velocity; AVN = AV node; HPS = His–Purkinje system; A = atrium; V = ventricle.

Amiodarone effect on atrial remodelling

Atrial remodelling has been proven to play a major role in the occurrence and maintenance of AF.^20,21 The main pathophysiological mechanisms contributing to AF are electrical remodelling, structural remodelling, autonomic nervous system changes, and Ca²⁺ handling abnormalities.²² Recognition of the mechanisms of remodelling in AF is important since they may lead to more specific therapies that target the underlying substrates.

Electrical remodelling is the first mechanism that occurs at the onset of AF and promotes AF through a re-entry-prone substrate. Changes in atrial frequency are determined by changes in the physiology of ion channel activation.

Studies have shown that the inward-rectifying K1 current (IK1) is a potential anti-arrhythmic target with an increased level in permanent AF, compared to normal SR human atrial myocytes.²³ IK1 provides the repolarising current during the terminal repolarisation (phase 3 of the cardiac action potential). It is one of the most important currents in the perpetuation of re-entrant mechanisms, as was documented by both experiments^24,25 and computer simulations.^26,27 Therefore blockage of IK1 results in prolongation of the atrial potential duration and effective refractory period (ERP).

Amiodarone blocks the channels that are upregulated by remodelling such as IK1 or by stimulation of the sympathetic nervous system. Therefore amiodarone’s pharmacological properties are thought to play a key role in treatment of longstanding persistent tachycardia.²⁸ Furthermore, the inhibitory effect of amiodarone on the inward Na+ Ca²⁺ currents determines a decrease in excitability and conductivity of cardiac tissues, especially when stimulated at higher frequencies.²⁹

Structural remodelling involves atrial enlargement and interstitial fibrosis. Left atrium dilation has been proven to be a precursor of AF as well as a prognostic factor for AF recurrence after cardioversion. A study showed that following the AF cardioversion, the left atrial fractional shortening was higher, while the left atrial stunning was lower in patients treated with amiodarone than in the subjects who received propafenone.³⁰

A key component of atrial remodelling is altered intracellular Ca²⁺ signalling.^31,32 In normal hearts, Ca²⁺ enters the cells with each action potential through L-type calcium channels (ICa,L), triggering Ca²⁺ release. An elevated heart rate in AF leads to intracellular Ca²⁺ overload that further engages several defence mechanisms. ICa,L reduction contributes to a decrease in Ca²⁺ inward current, shortening the AF-related action potential duration (APD), which promotes re-entry.^33,34 Moreover it has been shown that patients at risk for AF have lower levels of ICa,L than patients with a low risk for AF, due to the downregulation mechanism.³⁵

Shinagawa et al. used an experimental model with the aim to compare the anti-arrhythmic effects of amiodarone with flecainide and dofetilide in atrial tachycardia-associated AF, as well as to analyse these AAD’s influence on the impact of atrial tachycardia on L-type calcium channel expression.³⁶ Mongrel dogs were subjected to atrial tachycardia at 400 beats per min for seven days while the AADs were started three days prior to the onset of tachypacing.

They found that amiodarone had superior efficacy in the prevention of AF associated with atrial tachycardia remodelling, compared to flecainide or dofetilide. Furthermore, only amiodarone prevented both the changes in atrial ERP as well as the L-type Ca²⁺ channel expression induced by tachycardia. Amiodarone also reversed already established remodelling after four days of atrial tachycardia. Therefore the amiodarone effect in preventing atrial remodelling may contribute significantly in the superior efficacy of this AAD.

Electrophysiological properties of amiodarone: effect on atrial-His, His-ventricular, QT intervals

It has been demonstrated that amiodarone’s electrophysiological effects are significantly different following intravenous and oral administration.^37-39 These differences are certainly relevant to understand intravenous amiodarone’s efficacy in the acute conversion of cardiac arrhythmias compared to the prophylactic effects of oral amiodarone in preventing arrythmia recurrences.

When administered intravenously, amiodarone has a major effect on the atrioventricular (AV) node, which results in prolongation of the nodal conduction time, represented by the atrial-His (AH) interval.^40,41 Moreover, amiodarone significantly lengthens the functional refractory period (FRP) and ERP of the AV node.^40,42 This effect would lead to the termination of paroxysmal re-entrant supraventricular tachycardias and would account for reducing the ventricular response in AF and atrial flutter, which occasionally results in conversion. The same studies have shown that intravenous amiodarone administration has no significant effect on the atrial or ventricular ERP or on the QT interval.⁴⁰

By contrast, if amiodarone is administered orally, it may significantly lengthen the QT interval and prolong the APD in the atria, AV node and ventricles.^40,43 This explains the wide spectrum of amiodarone’s anti-arrhythmic effects as it increases the ERP in all myocardial tissues. Wellens et al. also found that following oral amiodarone, the ERP of the atrium and ventricle lengthens and the His-ventricular (HV) interval increases while intravenous amiodarone did not change these values. The AV node ERP and the AH interval were prolonged both after intravenous and oral amiodarone.⁴⁴

Efficacy of amiodarone versus other antiarrhythmics for AF

Class IC and class III AADs have been widely used for AF pharmacological conversion and their use could reduce AF recurrence rate by about 20 to 50% versus placebo.^45-47 The most important studies on the use of amiodarone versus other anti-arrhythmic drugs for the prevention of recurrent AF after conversion of AF are listed in Table 3.

Table 3. The most important randomised and non-randomised trials on the use of amiodarone vs other anti-arrhythmic drugs for the prevention of AF recurrences after cardioversion.

Trial	AADs evaluated	Number	Mean age (years)	Main results
Fragakis 201248	Amiodarone VS amiodarone + ranolazine for conversion of recent-onset AF	51	63	65 VS 88% conversion to SR
Santas 201249	Amiodarone + irbesartan VS irbesartan after cardioversion, for maintenance of SR	94	64	26.7 VS 55.1% AF recurrence at 18 months
Freemantle 201147	Amiodarone VS dronedarone VS sotalol vs flecainide VS propafenone	6629	61	41.9 VS 63.4% dronaderone 37.7 VS 51.9% sotalol 31 VS 32% flecainide 40 vs 47% propafenone AF recurrence
DYONISOS 201050	Amiodarone VS dronedarone for persistent AF	504	64	42 VS 63.5% AF recurrence at 12 months
Pitagora trial 200851	Amiodarone VS flecainide/propafenone VS sotalol for AF in patients with pacemakers for sinus node disease	176	72	40 vs 28 VS 24% SR at 1 year
Vijayalaskshmi 200652	Amiodarone VS sotalol for patients with persistent AF planned for DCC	94	64	81 VS 92% conversion to SR 63 vs 39% SR at 6 months
Niu 200653	Amiodarone VS sotalol for AF	102	56	78.4 VS 70.6% conversion to SR 67.5 VS 41.7% SR at 12 months 44.4 VS 26.7% SR at 24 months
SAFE-T 200554	Amiodarone VS sotalol for AF	665	67	27.1 VS 24.2% spontaneous conversion 27.7 VS 26.5% failed DCC 487 VS 74 days to AF recurrence
Kochiadakis 200455	Amiodarone VS propafenone for maintenance of SR	146	63	34.7 VS 44.59% AF recurrence at 9.8 and 3.8 months, respectively
Kanoupakis 200456	Amiodarone VS carvedilol for conversion and maintenance of SR	145	65	93.3 vs 91.5% conversion to SR 16.6 VS 27.9% AF recurrence at 4 weeks
Manios 200357	Amiodarone + diltiazem for conversion and maintenance of SR for persistent AF	111	64	100 VS 85% conversion to SR 26.4 VS 53.3% AF recurrence at 6 weeks
AFFIRM sub-study 200358	Amiodarone VS sotalol VS flecainide/propafenone after cardioversion in patients with persistent AF	410	69	62 VS 38 VS 23% in SR at 1 year
Kochiadakis 200059	Amiodarone VS sotalol for recurrent, symptomatic AF	186	63	47 VS 77% AF recurrence at 6 and 8 months, respectively
CTAF 200060	Amiodarone VS sotalol or propafenone to prevent recurrences of AF	403	65	35 VS 63% AF recurrence at 16 months
Martinez-Marcos 200061	Amiodarone VS propafenone VS flecainide for paroxysmal AF	150	60	64 VS 72 VS 90% in SR at 12 hours
Villani 199262	Amiodarone VS disopyramide for prevention of AF recurrences	76	65	32 VS 57% AF recurrence at 13 months
Zehender 199263	Amiodarone VS quinidine + verapamil for persistent AF	40	58	60 VS 55% in SR at 2 years
Martin 198664	Amiodarone VS disopyramide for paroxysmal AF	70	69	79 VS 55% in SR at 16 months
Vitolo 198165	Amiodarone VS quinidine in the prophylaxis of AF after DCC	54	53	78.5 VS 46.1% in SR at 6 months

AAD, anti-arrhythmic drug; AF, atrial fibrillation; SR, sinus rhythm; DCC, direct-current conversion.

In the critical appraisal section, we focus on the most important studies among those listed in the tables. We highlight the studies with the largest number of patients, such as CTAF, SAFE-T, EAST-AF, DIONYSOS. Thereafter we also mention studies on fewer patients that confirm the above findings: AMIO-CAT confirmed the results from EAST-AF. Moreover, these studies on a large number of patients were the basis of the recommendations listed in the new guidelines for the management of patients with persistent AF.

Critical appraisal

The literature search resulted in 4 308 publications on amiodarone compared to other anti-arrhythmic drugs used in patients with AF undergoing pharmacological or electrical cardioversion. Amiodarone was tested against flecainide, propafenone, dronedarone, sotalol, carvedilol, diltiazem, disopyramide, quinidine and ranolazine. Thirty-one studies were selected, out of which 11 were randomised, controlled trials: four studies compared amiodarone with class I anti-arrhythmic drugs,^51,55,62,65 while six compared amiodarone with sotalol,^51-54,58,59 and one study compared amiodarone with dronedarone.⁵⁰ There were 452 patients in the studies comparing amiodarone versus class I AADs, 504 patients when comparing amiodarone and dronedarone, and 1 633 patients in the amiodarone-versus-sotalol studies.

Altogether the results show that class I and III anti-arrhythmic drugs reduced AF recurrence after catheter ablation by 20 to 50%. Direct comparison between amiodarone and class IA, IC or other class III anti-arrhythmic drugs showed better results for amiodarone in preventing recurrences of AF.

In the Canadian trial of AF (CTAF), 403 patients with paroxysmal and persistent AF with at least one episode in the preceding six months were randomly assigned to receive amiodarone, sotalol or propafenone.⁶⁰ For the first 14 days, the amiodarone dose was 10 mg/kg body weight daily, followed by 300 mg per day for four weeks, after which a maintenance dose of 200 mg each day was prescribed. During a mean follow up of 16 months, 35% of patients taking amiodarone had AF recurrence, compared with 63% of the group treated with propafenone or sotalol (p < 0.001).⁶⁰

In SAFE-T (Sotalol Amiodarone Atrial Fibrillation Efficacy Trial) 665 patients with persistent AF were randomised to amiodarone, sotalol or placebo.⁵⁴ Amiodarone was given at a dose of 800 mg daily for 14 days, followed by 600 mg for the next 14 days, and 300 mg per day for the first year, after which a dose of 200 mg per day was prescribed.

After 28 days of treatment, amiodarone and sotalol were equally effective in AF conversion to SR (27.1 vs 24.1%, respectively, p = 0.45). Amiodarone was however superior for SR maintenance with a longer time to first AF recurrence compared with sotalol (median 487 vs 74 days, p = 0.002) (p = 0.002) or placebo (median 487 vs 6 days, p = 0.001).⁵⁴ Similarly, a sub-study of the AFFIRM trial found that amiodarone was more successful in maintaining SR after a follow up of one year compared with sotalol (60 vs 38%, p = 0.002) or with class I agents (62 vs 23%, p < 0.001).⁵⁸

Amiodarone was also compared against dronaderone in the DIONYSOS (Double-Blind Trial to Evaluate the Efficacy and Safety of Dronaderone) ramdomised, controlled trial that enrolled 504 patients with persistent AF.⁵⁰ Amiodarone was given to 255 patients at a dose of 600 mg once daily for the first 28 days and 200 mg once daily thereafter, while dronaderone’s dose was 400 mg twice a day. The primary composite endpoint included AF recurrence or premature study discontinuation and the result was 75.1 vs 58.8% for dronaderone and amiodarone, respectively [hazard ratio (HR) 1.59, p < 0.001].

At 12 months of treatment, amiodarone was associated with lower risk of AF recurrence (42 vs 63.5%). On the other hand amiodarone had a worse safety profile than dronaderone (13.3 vs 10.4%) with less premature drug discontinuation for patients with dronaderone versus amiodarone (HR 0.76, p = 0.227). A recent study showed that dronaderone might be a good therapeutic option to replace amiodarone in selected patients with amiodarone-induced hyperthyroidism for shortterm discontinuation.⁶⁶

Regarding the combination of amiodarone with other anti-arrhythmics, it seems that the addition of ranolazine would produce conversion much faster than amiodarone in monotherapy. Tsanaxidis et al. studied a population of 173 patients, mean age of 68 years with recent-onset AF, enrolled for pharmacological cardioversion of AF.⁶⁷ The primary endpoint was the time to AF conversion in the amiodarone-versusamiodarone + ranolazine group and the secondary endpoint was the conversion rate within 24 hours in the two groups.

The authors demonstrated that the combination of amiodarone + ranolazine led to conversion to SR 10.8 hours faster than with amiodarone alone (p < 0.0001). Furthermore, the 24-hour conversion rate with the combination of amiodarone + ranolazine was higher than with amiodarone in monotherapy (98 vs 58%, p < 0.001). Their results were confirmed by Koskinas et al., who found in a group of 121 patients with recent-onset AF, a conversion time 3.1 hours shorter for the combination amiodarone + ranolazine compared to amiodarone in monotherapy (p = 0.001), and a higher conversion rate of 81 vs 54% (p = 0.02).⁶⁸

Efficacy of amiodarone in AF recurrence after direct-current conversion

The success rate of electrical cardioversion was reported to be 75 to 87%, while the AF recurrence rate after conversion may be up to 50%, especially within four weeks.^69-71 Strong evidence supports that pre-treatment with amiodarone is effective in increasing not only the acute restoration of SR but also the longterm maintenance of SR.^72-74

The European Society of Cardiology (ESC) recommends anti-arrhythmic pre-treatment prior to conversion in order to enhance success and to prevent further AF recurrences.¹ If AAD is planned for SR maintenance after cardioversion, the guideline recommendation is to start amiodarone therapy a few weeks before cardioversion in order to achieve effective drug levels. However the guidelines assign this use of amiodarone to a class IIa recommendation with evidence level B (derived from a single randomised, controlled trial or large non-randomised studies).

The most important studies on the use of amiodarone for the prevention of recurrent AF after conversion are listed in Table 4.

Table 4. The most important randomised and non-randomised trials on the use of amiodarone for the prevention of AF recurrences after cardioversion.

Trial	Amiodarone administration	Number	Mean age (years)	Main results
Galperin 201475	Amiodarone for maintenance of SR after cardioversion	51	68	81.5 VS 54.2% SR at 18 months
Jong 200616	Amiodarone for maintenance of SR after cardioversion	76	66	38 VS 36% AF recurrence at 5 years between 200 and 100 mg amiodarone
Vijayalakshmi 200652	Amiodarone in patients with AF with DCC	94	64	63 VS 16 in SR at 6 months between amiodarone and no drug
Channer 200473	Amiodarone VS placebo before and after electrical cardio- version for persistent AF	161	67	51 vs 16% in SR at 8 weeks between amiodarone and placebo
Boos 200477	Short-duration oral amiodarone after cardioversion of AF	35	61	47.1 VS 16.7% in SR at 16 months between amiodarone and placebo
Manios 200357	Amiodarone and recurrence rate after cardioversion	111	64	26.4 VS 50% at 6 weeks between amiodarone and no drug
GEFACA 200178	Amiodarone VS placebo for termination of persistent AF and maintenance of SR	95	63	79.54 VS 38.46% conversion to SR 37.14 VS 80% AF recurrence at 2.7 months
CHF-STAT 199879	Amiodarone for maintenance of SR after in patients with AF + heart failure	103	67	31 VS 7.6% SR at 4 years amiodarone VS placebo
Opolski 199780	Amiodarone for conversion and maintenance of SR after failed electrical cardioversion	49	62	65% SR after first ineffective DC cardioversion and followed by pre-treatment with amiodarone and repeat DCC 52% SR at 12 months
Chun 199581	Amiodarone for maintenance of SR in patients with refractory AF	110	60	87% SR at 1 year 70% SR at 3 years 55% at 5 years
Gosselink 199282	Amiodarone for maintenance of SR after cardioversion	89	63	53% SR at 3 years
Brodsky 198783	Amiodarone for maintenance of SR after DCC in patients with dilated LA	28	61	39% SR at 6 months 35% SR at 1 year
Gold 198684	Amiodarone for refractory AF	68	59	79% SR at 21 months
Horowitz85	Amiodarone for paroxysmal and persistent AF resistant to quinidine	38	60	55% in SR (of paroxysmal AF) and 45% SR (of persistent AF) at 15 months

AF, atrial fibrillation; SR, sinus rhythm; DCC, direct-current conversion; LA, left atrium.

Critical appraisal

The literature search resulted in 1 035 publications on amiodarone use in adult patients with AF undergoing pharmacological or electrical cardioversion. Amiodarone was prescribed in the study group, while in the control group placebo or no treatment was used. Six studies including 528 patients were randomised, controlled trials and were published between 1995 and 2014.^{52,57,73,75-77} Patients had persistent AF with a duration between one month and one year and received amiodarone 200 to 800 mg for one to 52 weeks. The left atrial diameter was 50 mm in all six studies and the length of follow up was one to 16 months. Overall, the long-term maintenance of SR was 34 to 50% for patients treated with amiodarone and five to 17% for patients treated with placebo or not receiving anti-arrhythmic drugs.

The study of Channer et al. enrolled 161 patients with persistent AF, who were randomised to three groups: two groups of patients were treated before electrical conversion for two weeks with 400 mg amiodarone, and it was continued at a dose of 200 mg for eight weeks in one group and 52 weeks in the other, while the placebo group received placebo throughout the study.⁷³

Of the patients pre-treated with amiodarone, 21% (26/123) converted to SR before direct-current conversion (DCC), while all 38 patients on placebo were still in AF at the moment of DCC (absolute difference 21%, CI: 10–29%, p = 0.002). At eight weeks following successful DCC, amiodarone was more effective in the maintenance of SR compared to placebo [63/123 (51%) vs 6/38 (16%), p < 0.001]. After one year, more patients on longterm amiodarone remained in SR versus patients on short-term amiodarone [30/61 (49%) vs 21/62 (33%), p = 0.085]. However a higher rate of adverse effects requiring discontinuation was noted for long-term amiodarone (18%) compared to short-term treatment (8%) or placebo (3%).73 Overall, the six random, controlled trials supported the use of amiodarone for long-term maintenance of SR with minor side effects.

Efficacy of amiodarone in AF recurrence after catheter ablation

Catheter ablation has gained a significant role in restoring SR in patients with symptomatic AF. Although a high success rate with catheter ablation is achievable, it may vary between 60 and 80% for paroxysmal AF and between 50 and 60% for persistent AF.^86-88 One of the most frequent causes of early recurrence of arrythmias post ablation is inflammation due to tissue damage inflicted by the radiofrequency energy.⁸⁹ While early recurrence is not considered a sign of failure in catheter ablation treatment, it is strongly indicative of potential late arrythmia recurrence. As a potential solution to this issue, the short-term use of AADs has been suggested.^90,91 This would serve both as a preventative mechanism for early arrythmia recurrence, as well as facilitating left atrium reverse remodelling in the long term.

The most important studies on the use of amiodarone for the prevention of recurrent AF after conversion of AF are listed in Table 5.

Table 5. The most important randomised and non-randomised trials on the use of amiodarone for the prevention of AF recurrences after catheter ablation.

Trial	AAD administration	Number	Mean age (years)	Main results
POWDER AF 201892	Anti-arrhythmic drugs VS no drug after catheter ablation of recurrent AF	173	62	2.7 VS 21.9% AF recurrence at 1 year
Kettering 201793	Short-term amiodarone after catheter ablation of persistent AF	230	61	16.5 VS 29.6% within first 3 months 18.3 VS 27% at 1 year arrhythmia recurrence between amiodarone and no AAD
EAST AF 201694	Short-term (90 days) amiodarone or class I VS placebo after catheter ablation of AF	2 038	63	59.0 VS 52.1% at 90 days 69.5 VS 67.8% at 1 year free rate from recurrent atrial tachyarrhythmias
AMIO-CAT 201695	Short-term amiodarone after catheter ablation of AF	212	61	39 VS 48% atrial tachyarrhythmia at 6 months between amiodarone and placebo
Lodzinski 201496	Amiodarone or sotalol VS no AAD VS last ineffective AAD used before PVI, after catheter ablation of paroxysmal AF	180	50	47.5 VS 54.3 VS 45.5% SR at 2 months
Mohanty 201497	Amiodarone discontinuation 4 months before ablation VS ablation performed without amiodarone discontinuation	112	61	57 vs 79% AF termination during ablation 66 VS 48% SR at 32 months
Wu 200898	AAD VS placebo after catheter ablation of AF	74	62	13.5 VS 37.8% AF recurrence at 3 months 29.7 VS 24.3% at 12 months 8.1 VS 8.1% at more than 12 months
Turco 200799	Amiodarone VS no drug after catheter ablation of AF	107	57	30 VS 34% AF recurrence at 12 months

AAD, anti-arrhythmic drug; AF, atrial fibrillation; SR, sinus rhythm; PVI, pulmonary vein isolation.

Critical appraisal

The literature search resulted in 2 887 publications on the efficacy of amiodarone in AF recurrence after catheter ablation, out of which we included eight studies (Table 5). All of them were randomised, controlled trials and were published between 2007 and 2018. Eight were open label and one was double blind.⁹⁵ Four studies included patients from a single centre, one study was from two centres,⁹⁵ and three studies were multicentre trials.^92,94,97 In two studies pulmonary vein isolation (PVI) was used,^92,95 while in the other six PVI + other ablation techniques, such as mitral isthmus line, cavo-tricuspid line or complex fractionated atrial electrograms (CFAE) were used.

In all eight trials amiodarone was given after catheter ablation and compared to placebo or other anti-arrhythmic drug. One study included only patients with paroxysmal AF,⁹² one study only patients with persistent AF,⁹⁷ and the other six studies both patients with paroxysmal and persistent AF. Patients were followed for two to 17 months after catheter ablation.

Amiodarone did reduce AF recurrence during short-term follow up (six to 12 weeks) but did not improve freedom from AF during the mean follow-up period of eight months. While effective in reducing recurrence of arrythmias in the early stages post ablation, amiodarone did not appear to be an effective long-term solution. The potential negative side effects in patients presenting with structural heart diseases or heart failure, as well as elderly patients, strike a delicate balance between advantages and drawbacks.

In the EAST-AF (Efficacy of Antiarrhythmic Drugs Short- Term Use After Catheter Ablation for Atrial Fibrillation) trial with over a 90-day use of AADs, a significant reduction in early arrythmias was confirmed, however, this did not translate into improvements in the later stages.⁹⁴ The AMIOdarone After CATheter Ablation for Atrial Fibrillation (AMIO-CAT) trial reported similar results with no difference in the AF-free rate at six months after six or eight weeks, respectively, of AAD use.⁹⁵

Mohanty et al. investigated the impact of peri-procedural amiodarone on the outcome of AF ablation for 112 patients with long-standing persistent AF.⁹⁷ The patients were randomised to two groups: one that discontinued amiodarone four months before ablation, and a control group that underwent AF ablation without amiodarone discontinuation.

For the patients in the off-amiodarone group, a higher number of non-pulmonary vein triggers were revealed compared to the other group (75 vs 43%, p < 0.001). Patients in the on-amiodarone group had however lower fluoroscopy, radiofrequency and procedural times than those who discontinued amiodarone therapy. At six months’ follow up, both groups had similar rates of freedom from AF (71% in the off-amiodarone group vs 75% in the on-amiodarone group, p = 0.702). At 32 ± 8 months, a higher success rate without AAD was demonstrated for patients in the off-amiodarone group (66 vs 34%, p < 0.002). The study by Zhang et al. suggests that a second ablation procedure renders better results than AAD use in cases of persistent AF.¹⁰⁰

Safety profile of amiodarone

Amiodarone has a wide spectrum of adverse effects, with a prevalence of 15% in the first year of administration, which may increase up to 50% in the case of prolonged administration.¹⁶ Potential adverse effects include toxicity involving the lungs, heart, thyroid gland, liver, skin, eyes and nerves (Table 6).

Table 6. Adverse effects of amiodarone.

Adverse effect	Incidence (%)	Reference
Pulmonary toxicity	2	102
Bradycardia	5	103
Torsade de pointes	<1	104
Hyperthyroidism	3	105
Hypothyroidism	4-22	8
Hepatitis and cirrhosis	<3	102,106
Nausea, anorexia, constipation	30	107
Photosensitivity	25-75	105
Grey-blue discoloration	< 10	105
Corneal microdepositions	> 90	101
Optic neuropathy	< 1	108,109
Neurologic: ataxia, tremor, peripheral neuropathy	3-30	110,111

While the majority of adverse effects are reversible after amiodarone dose reduction or discontinuation, fatal complications have also been reported. These include pulmonary fibrosis, liver cirrhosis or bradycardia that may lead to cardiac arrest. At risk for pulmonary fibrosis are patients with an underlying lung disease or with a dosage of amiodarone greater than 400 mg/day. Therefore amiodarone should not be used in patients with significant pulmonary disease, significant liver disease or with symptomatic conduction system disease.

Follow up of patients on amiodarone treatment is mandatory to detect and limit its adverse effects. Routine screening is recommended every six months, with lung function assessment including chest radiography, liver and thyroid function tests, ECG, and an annual eye examination.¹⁰¹

Moreover, amiodarone acts as an inhibitor of several cytochrome P450 pathways, such as CYP3A4 or CYP2C9, therefore resulting in alterations in plasma concentrations of many common drugs. The main amiodarone drug interactions are listed in Table 7.^16,112–114 An important interaction is potentiation of the warfarin anticoagulant effect, with the recommendation to monitor prothrombin times closely and to reduce the dose of anticoagulant if needed.^115,116 Amiodarone can also increase digoxin levels, therefore dose reduction and close monitoring of serum digoxin levels is recommended.¹¹⁴ To avoid adverse drug– drug interactions, patients under treatment with amiodarone should therefore consult their cardiologist whenever a new drug is prescribed.

Table 7. Main interactions of amiodarone with other drugs.

Amiodarone plus	Interactions
Coumadin derivates	Potentiates the anticoagulant effect and increases phrotrombin time
Digoxin	Increases digoxin level
Beta-blockers, calcium channel blockers	May cause bradycardia, sinus arrest or AV block
Class IA anti-arrhythmics drugs, macrolide antibiotics, tricyclic antidepressants	May prolong QTc
Hypokalaemia-causing drugs (diuretics)	May increase the risk of Torsades de points
DOACs	May increase its bioavailability, but does not affect efficacy or safety outcomes
Statins (lovastatin, simvastatin)	May increase the risk of rhabdombyolisis or myopathy
Flecainide	Increases flecainide levels

The most important studies on the side effects of amiodarone versus other AADs are listed in Table 8. We evaluated the withdrawal rate due to AAD side effects as well as pro-arrythmic events. Following the comparison studies between amiodarone and class I or III anti-arrhythmics, we obtained the following results: amiodarone had a lower withdrawal rate than class I drugs combined and no statistically significant difference when compared to dronaderone or sotalol. As for pro-arrhythmia, amiodarone had a lower side-effect rate than class I AADs and no clear difference was observed between amiodarone and sotalol or dronedarone.

Table 8. The most important trials that evaluated amiodarone vs other AAD for withdrawals due to side effects and for pro-arrhythmic events.

Trial	AADs compared	Number	Mean age (years)	Greater withdrawals due to side effects, RR (95% CI)	Greater pro-arrhythmia, RR (95% CI)
DYONISOS 201050	Amiodarone VS dronaderone	504	64	Amiodarone 1.37 (0.9-2.09)	Amiodarone 1.95
Pitagora 200851	Amiodarone vs flecainide/propafenone	176	72	Amiodarone 1.11 (0.23-5.4)	Not studied
Pitagora 20085	Amiodarone VS sotalol	176	72	Amiodarone 5.86 (0.34-100.89)	Not studied
Vijayalaskshmi 200652	Amiodarone vs sotalol	78	64	Sotalol 0.38 (0.04-3.14)	Not studied
Niu 200653	Amiodarone vs sotalol	102	56	Sotalol 0.71 (0.24-2.1)	Not studied
Kochiadakis 200455	Amiodarone VS propafenone	254	63	Amiodarone 8.74 (2.09-36.4)	Similar 1.03 (0.15-7.10)
AFFIRM sub-study 200358	Amiodarone VS class I drug	410	69	Class I drug 0.33 (-0.21-0.53)	Class I drug 0.20 (0.08-0.51)
AFFIRM sub-study 200358	Amiodarone VS sotalol	410	69	Sotalol 0.83 (0.47-1.47)	Sotalol 0.49 (0.17-1.42)
Kochiadakis 200059	Amiodarone VS sotalol	186	63	Amiodarone 3.44 (1.01-11.75)	Similar 0.94 (0.14-6.46)
Villani 199262	Amiodarone vs disopyramide	76	65	Disopyramide 0.35 (0.1-1.18)	Not studied
Vitolo 198165	Amiodarone vs quinidine	54	53	Similar 0.93 (0.06-14.09)	Similar 0.93 (0.06-14.09)

AAD, anti-arrhythmic drug.

Another large cohort study on amiodarone and mortality retrospectively enrolled 141 500 patients with AF. They did not find a significant increase in risk of death with amiodarone use.^117,118 The TREAT-AF study included 122 465 patients with newly diagnosed AF and found that amiodarone use was not associated with increased risk of death.¹¹⁹ This study also included the largest cohort with concomitant AF and heart failure on amiodarone therapy and reported no association between amiodarone and increased mortality rates in patients with a diagnosis of heart failure at baseline.

Based on studies that compared amiodarone with other AADs, we propose the clinical approach outlined in Fig. 1 for SR maintenance after cardioversion or catheter ablation.

Fig. 1.

Clinical approach for maintenance of sinus rhythm after cardioversion or catheter ablation.

Conclusion

AF is a time bomb. The high and continuously increasing prevalence of AF has led to a real pandemic of heart problems, which has a negative impact on patients’ quality of life. Rapid treatment with cardioversion and prevention of other episodes is essential, given the vicious circle known as ‘fibrillation induces fibrillation’.

Over the last 45 years, amiodarone has been used extensively in all countries of the world. The experience and knowledge gained with this anti-arrhythmic unequivocally recommends it for conversion to SR and maintaining normal rhythm in patients with AF. Moreover, knowledge of the pharmacokinetics and pharmacodynamics of this anti-arrhythmic, along with an understanding of the pathophysiological mechanisms behind AF, provides an additional rationale for early use of amiodarone to prevent long-term complications of AF.