- Email: [email protected]
A prospective, randomized controlled trial comparing the efficacy and safety of Sotalol, Amiodarone, and Digoxin for the reversion of new-onset atrial fibrillation
ORIGINAL CONTRIBUTION
A Prospective, Randomized Controlled Trial Comparing the Efficacy and Safety of Sotalol, Amiodarone, and Digoxin for the Reversion of New-Onset Atrial Fibrillation
From the Departments of Emergency Medicine* and Cardiology,‡ Royal North Shore Hospital, Sydney; and the Faculty of Medicine, University of Sydney, New South Wales, Australia.§
Anthony P. Joseph, MB BS*§ Michael R. Ward, FRACP, PhD‡§
Received for publication August 20, 1999. Revision received March 15, 2000. Accepted for publication March 27, 2000. Address for reprints: Anthony P. Joseph, MB BS, Department of Emergency Medicine, Royal North Shore Hospital, Pacific Highway, St. Leonards, New South Wales, 2065, Australia; 61-2-9926-7922, fax 61-2-9906-8123; E-mail [email protected]. Copyright © 2000 by the American College of Emergency Physicians. 0196-0644/2000/$12.00 + 0 47/1/107655 doi:10.1067/mem.2000.107655
Study objective: A prospective, randomized controlled trial of new-onset atrial fibrillation was conducted to compare the efficacy and safety of sotalol and amiodarone (active treatment) with rate control by digoxin alone for successful reversion to sinus rhythm at 48 hours. Methods: We prospectively randomly assigned 120 patients with atrial fibrillation of less than 24 hours’ duration to treatment with sotalol, amiodarone, or digoxin using a single intravenous dose followed by 48 hours of oral treatment. Patients had ECG monitoring for 48 hours, and time of reversion, adequacy of rate control, and numbers of adverse events were compared. After 48 hours, those still in atrial fibrillation underwent cardioversion according to a standardized protocol. After 48 hours of therapy and attempted cardioversion, the number of patients whose rhythms had successfully reverted were compared. Results: There was a significant reduction in the time to reversion with both sotalol (13.0±2.5 hours, P<.01) and amiodarone (18.1±2.9 hours, P<.05) treatment compared with digoxin only (26.9±3.4 hours). By 48 hours, the active treatment group was significantly more likely to have reverted to sinus rhythm than the rate control group (95% versus 78%, P<.05; risk ratio 5.4, 95% confidence interval [CI] 1.5 to 19.2 ). In those patients whose rhythms did not revert to sinus rhythm, there was superior ventricular rate control in the sotalol group at both 24 and 48 hours compared with those who received either amiodarone or digoxin. There were also fewer adverse events in the active treatment group compared with the rate control group. Conclusion: Immediate pharmacologic therapy for new-onset atrial fibrillation with class III antiarrhythmic drugs (sotalol or amiodarone) improves complication-free 48-hour reversion rates compared with rate control with digoxin.
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
1
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
[Joseph AP, Ward MR. A prospective, randomized controlled trial comparing the efficacy and safety of sotalol, amiodarone, and digoxin for the reversion of new-onset atrial fibrillation. Ann Emerg Med. July 2000;36:1-9.] INTRODUCTION
Atrial fibrillation (AF) is one of the most common tachyarrhythmias seen in clinical practice (overall prevalence 4%, with more than half of affected patients older than 75 years) and results in hospital stays in the United States that are longer than for any other arrhythmia.1,2 Although the incidence of AF is increased with structural abnormalities such as left atrial enlargement, left ventricular hypertrophy, and left ventricular dysfunction, lone AF (AF without structural heart disease) accounts for about a third of the disease prevalence.2 The ineffective atrial contractions in AF reduce effective cardiac output, especially during exercise, and substantially increase the risk of stroke as a result of the formation and embolization of atrial thrombus.3 Sustained AF with rapid ventricular rate may also result in reversible tachycardia-induced left ventricular dysfunction,4,5 further increasing left atrial stasis.6 AF increases the risk of ischemic stroke sixfold to approximately 5% per year,7 and embolic strokes related to AF are usually large with death or severe neurologic deficit in 50% to 70% of cases.8 Anticoagulation markedly reduces the occurrence of embolic stroke in AF, but also increases risk of major bleeding, particularly in the elderly in whom AF is most prevalent.9 Consequently, chronic AF is associated with a significant increase in morbidity and mortality independent of underlying cardiac pathology10; this has prompted strategies to promote reversion to and maintenance of sinus rhythm in acute AF. The repeated atrial depolarization in AF rapidly induces electrophysiologic changes in the atrial myocardium that promote recurrence of the arrhythmia immediately after reversion,11,12 such that the earlier reversion to sinus rhythm occurs, the greater the likelihood of sustained maintenance of sinus rhythm. However, enthusiasm for early pharmacologic treatment to promote reversion has been dampened by high spontaneous early reversion rates13 and concerns regarding adverse effects of antiarrhythmic drugs, particularly proarrhythmia. Proarrhythmias occur as a result of the tendency of some antiarrhythmic agents (class I and class III) to cause heterogeneous prolongation of the QT interval. This leads to increased inhomogeneity of ventricular repolarization, which may facilitate the production of a sustained ven-
2
tricular arrhythmia such as polymorphic ventricular tachycardia of the torsade de pointes type.14,15 Patients most at risk for this arrhythmia are those with ventricular dysfunction, hypokalemia, or baseline prolongation of the QT interval.3 Although the class III drugs, amiodarone and d,l-sotalol hydrochloride (sotalol), are the most frequently used to revert atrial fibrillation in Australia, the risk of proarrhythmic events in a low-risk population such as new-onset AF has not been well defined. Adverse effects of class III antiarrhythmic drugs, such as hypotension, are far more frequent in patients with impaired cardiac or renal function, whereas increased risk of proarrhythmia in patients with AF has only been convincingly demonstrated in trials with class I antiarrhythmic drugs, such as quinidine.15,16 The risks and benefits of commonly used regimens of sotalol and amiodarone have not been previously compared with ventricular rate control alone with digoxin in the treatment of recent-onset AF. Thus, we performed a prospective, randomized, controlled multicenter trial to assess the efficacy and safety of amiodarone and sotalol in conversion of new-onset AF to sinus rhythm at 48 hours compared with rate control alone with digoxin. M AT E R I A L S A N D M E T H O D S
From November 1992 until September 1997, 120 patients with new-onset (historically of less than 24 hours’ duration) AF or atrial flutter with ventricular rate more than 100 beats/min were enrolled. All patients presented to the emergency departments of 1 of 3 hospitals (Figure 1). The protocol was approved by the Medical Ethics Review Committees of the hospitals involved, and informed consent was obtained from all patients. Inclusion and exclusion criteria are noted in Table 1. Wide-complex tachycardia, such as AF with Wolff–Parkinson-White syndrome, was excluded because of the risk of development of ventricular fibrillation if digoxin is used in the presence of an accessory pathway. Significant left ventricular dysfunction was defined as left ventricular ejection fraction less than 30% or diuretic therapy with more than 40 mg of frusemide or equivalent per day. Patients were prospectively individually randomly assigned to 1 of the following 3 drug regimens for 48 hours: 1. Digoxin 500 µg intravenously over 30 minutes followed by 250 µg orally every 6 hours for 4 doses and then 250 µg daily (125 µg daily if any renal impairment as defined by serum creatinine concentration >120 µmol/L)
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
2. Amiodarone 5 mg/kg intravenously over 30 minutes, then 400 mg orally every 8 hours for 6 doses 3. Sotalol 1.5 mg/kg intravenously over 30 minutes, then 80 mg orally every 8 hours for 6 doses The randomization process was computer-generated and administered centrally. Initially the randomization
Figure 1.
process was open, with the study drug known to the treating physicians. However, after the enrollment of 85 patients, the investigators believed it was preferable to blind the treating physicians to the selected drug until inclusion and exclusion criteria were met, and consent obtained. Baseline serum potassium, serum thyroxine, and thyroid-
Acute atrial fibrillation/flutter (<24 hours) (n=120)
Flow chart from point of enrollment to completion of study.
Inclusion criteria met No exclusion criteria
Consent
Randomized
Protocol violations (n=5)
Digoxin (n=36)
Amiodarone (n=39)
Sotalol (n=40)
48 hours
48 hours
48 hours
Sinus rhythm (n=21)
Sinus rhythm at 48 hours (n=28; 78%)
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
Atrial fibrillation (n=15)
Sinus rhythm (n=30)
Atrial fibrillation (n=9)
Sinus rhythm (n=35)
Atrial fibrillation (n=5)
Cardioversion (n=10)
Cardioversion (n=7)
Cardioversion (n=3)
Sinus rhythm (n=7)
Sinus rhythm (n=7)
Sinus rhythm (n=3)
Sinus rhythm at 48 hours (n=37; 94%)
Sinus rhythm at 48 hours (n=38; 95%)
3
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
stimulating hormone were measured. Hyperthyroidism and hypothyroidism were defined by the individual laboratories’ normal ranges. All patients had ECG monitoring throughout the 48-hour trial period, and time of rhythm reversion and ventricular rate on reversion to sinus rhythm were noted. The ventricular rate of those remaining in AF was recorded at 4, 24, and 48 hours. If the patient’s rhythm had not reverted within 24 hours, intravenous heparin was then started with a 5,000-IU bolus and infusion at 1,000 IU/h and adjusted to activated partial thromboplastin time (APTT) of 2.0 to 2.5 times control. If the rhythm had not reverted to sinus rhythm by 48 hours and APTT prolongation had been achieved satisfactorily, electrical cardioversion was attempted using sequential shocks of 50, 100, 200, 300, and 360 J until sinus rhythm was restored. Cardioversion was postponed in the presence of left ventricular failure, systemic emboli, or inadequate anticoagulation (when APTT >2.0 times control had not been achieved for at least 4 hours). Twelvelead ECGs were recorded at baseline and 24 hours and the QTc interval was measured in lead II using Bazett’s formula -R ) averaged over 3 cycles. Echocardiography (QT/R was performed after at least 48 hours of sinus rhythm. Left atrial size and left ventricular fractional shortening were measured in the parasternal long-axis view. Adverse events noted included hypotension (mean arterial pressure [MAP] <70 mm Hg), onset of left ventricTable 1.
Inclusion and exclusion criteria. Inclusion criteria AF onset within 24 h Consent obtained Serum K+ >3.5 mmol/L and <5.5 mmol/L Serum creatinine <0.2 mmol/L Exclusion criteria AF present within 7 d and >24 h No consent Serum K+ <3.5 mmol/L and >5.5 mmol/L Serum creatinine >0.2 mmol/L Current β-blocker treatment Digoxin or sotalol treatment in last week Amiodarone treatment within 3 months Hypotension (MAP <70 mm Hg) Previous adverse reaction to any of trial medications Known thyroid disease Asthma/bronchospasm with β-blocker Wide-complex tachycardia Contraindication to anticoagulation Age <18 y Left ventricular dysfunction Pregnancy
4
ular failure (new inspiratory crepitations with signs of pulmonary venous congestion on chest radiograph), significant bradyarrhythmias (ventricular rate <40 beats/ min either in AF or sinus rhythm, or ventricular rate <60 beats/min with MAP <70 mm Hg), ventricular tachyarrhythmias (torsade de pointes polymorphic ventricular tachycardia or other), thrombophlebitis, stroke, hypersensitivity reactions (including bronchospasm), and prolongation of QTc interval by more than 50 ms. Trial medication was discontinued and observance of the trial protocol ceased for hypotension (MAP <70 mm Hg), tachyarrhythmias or bradyarrhythmias (as defined previously), or stroke. Because other therapies and treatment modalities were frequently used after discontinuation of trial protocol, the endpoints (reversion or ventricular rate) after discontinuation were analyzed as if the patients had remained in the trial until the completion of 48 hours or discharge from hospital. Statistical analysis was performed using SPSS statistical software (version 10.0; SPSS Inc, Chicago, IL). The primary outcome measure was successful reversion to sinus rhythm within 48 hours with antiarrhythmic treatment or cardioversion. Sample size (40 patients per group) was calculated before commencement of the trial allowing for pairwise comparisons between each of the 3 drugs. Assuming a reversion rate with digoxin of 50%, the sample size was calculated to be able to detect an absolute difference in reversion at 48 hours of 25% with each active treatment (ie, a 75% reversion rate with either amiodarone or sotalol) with P value less than .05 and a power of 0.80. Secondary endpoints included time to reversion, ventricular rate if still in AF at 4, 24, and 48 hours and adverse events. For continuous variables, all 3 groups were compared using analysis of variance, and post hoc testing was performed using Tukey’s test. For categorical variables, analysis was performed to compare the groups by χ2 testing or Fisher’s exact test where appropriate. Time to reversion was compared between the groups by Cox proportional hazards. Rate control was compared at 4, 24, and 48 hours by multiple linear regression with all baseline variables and treatment received entered as independent variables. R E S U LT S
One hundred twenty patients were enrolled in the trial. Five patients had protocol violations and were excluded from further analysis (4 patients in the digoxin and 1 in the amiodarone group). Baseline characteristics for the remaining 115 were similar in the 3 groups (Table 2). In 5
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
patients, the trial medication was ceased because of stroke (1 patient at 24 hours in the digoxin group), hypotension (2 patients in the sotalol group), or bradyarrhythmia (1 patient with 10.5-second sinus arrest on reversion in the amiodarone group and 1 patient with 12.0-second sinus arrest on reversion in the digoxin group). There were no deaths within the 48-hour study period. The numbers of patients with rhythm reversion by 48 hours in each of the sotalol and amiodarone groups were significantly greater than in the digoxin group (primary outcome measure, Table 3). This was because of significant increases in both spontaneous reversion with active drug treatment (occurring gradually over the 48 hours, Figure 2), and cardioversion success rate of those still in AF at 48 hours. Treatment with either sotalol or amiodarone significantly reduced time to reversion compared with digoxin therapy (Figure 2). The larger reduction observed in the sotalol group was not significantly different from that in the amiodarone group (P=.23; Table 3). No baseline variable was significantly associated with longer time to reversion by Cox proportional hazards, although there was a trend for those patients with increased left atrial size and left ventricular dysfunction. Of the 29 patients who remained in AF at 48 hours, cardioversion was not undertaken in 9 (Table 3). Four patients had significant left ventricular failure (all treated with digoxin) and cardioversion was postponed until
control of heart failure was achieved. One patient had a stroke (digoxin), 2 patients had inadequate anticoagulation (1 amiodarone and 1 sotalol), 1 patient had been withdrawn because of hypotension (sotalol), and 1 patient withdrew consent (amiodarone). When comparing the success of cardioversion at 48 hours, we found that significantly greater numbers of patients in the active treatment group who were still in AF had electrical reversion compared with those in the digoxin-treated group (100% versus 70%). At the end of the study period, patients treated with active therapy (sotalol or amiodarone) were significantly more likely to have reverted to sinus rhythm (P<.05; risk ratio 5.4; 95% CIs 1.5 to 19.2; Table 3). There were no significant differences between the treatment endpoints in the 3 treatment groups before and after the blinding to the study drug was initiated. In the patients who remained in AF at 4, 24, and 48 hours (Figure 3), sotalol therapy was associated with markedly lower ventricular rate compared with both digoxin and amiodarone (P<.05 at 24 and 48 hours). Higher ventricular rate at entry into the study was subsequently associated with higher ventricular rate if the patient was still in AF. Of those remaining in AF at 24 hours, adequate rate control (rate <100 beats/min) was achieved in 4 of 7 patients treated with sotalol but only 3 of 12 and 5 of 18 patients treated with amiodarone and digoxin, respec-
Table 2.
Baseline characteristics. Characteristic Atrial flutter Ventricular rate at enrollment (beats/min; mean±SEM) Age (y; mean±SEM) Male Serum K+ (mmol/L; mean±SEM) Thyroid-stimulating hormone (U/mL; mean±SEM) Hyperthyroid/hypothyroid Left atrium size (cm; mean±SEM) Left ventricular dysfunction* Underlying disease (known at admission) Ischemic heart disease Valvular heart disease Cardiomyopathy Hypertension No structural heart disease
Digoxin (n=36)
Amiodarone (n=39)
Sotalol (n=40)
1 141.3±3.8 64.9±2.0 20 4.09±0.06 2.30±0.34 0/2 3.95±0.10 6
1 144.4±3.1 61.3±2.6 25 4.14±0.05 2.64±0.99 3/1 3.97±0.11 8
1 137.5±3.6 62.8±2.4 19 4.24±0.07 1.97±0.19 1/0 3.84±0.10 6
3 4 1 10 18 (50%)
8 3 5 5 18 (45%)
7 1 0 6 26 (65%)
Left ventricular ejection fraction ≤40% on echocardiography after reversion.
*
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
5
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
tively. Similarly, at 48 hours adequate rate control for those still in AF was achieved in 3 of 5 patients treated with sotalol, but only 2 of 9 and 5 of 15 patients treated with amiodarone and digoxin, respectively. The drug regimens were well tolerated. The most common major adverse event was the onset of left ventricular failure; this occurred in those patients who had persistent or recurrent AF with poor rate control (Table 4). Of the 8 patients (6 treated with digoxin and 2 treated with amiodarone) who developed clinically significant left ventricular failure, only 4 (2 treated with digoxin and 2 treated with amiodarone) had significant left ventricular dysfunction on echocardiogram, suggesting that heart failure in the other 4 digoxin-treated patients was likely related solely to poor rate control. Despite its profound effect on ventricular rate in AF, sotalol treatment did not result in any clinically important bradyarrhythmias. Significant bradycardia occurred in only 2 patients, both on reversion to sinus rhythm. Sinus rate on reversion was significantly lower in patients who received sotalol compared with those treated with either digoxin (mean difference 11.8 beats/min; 95% CIs 3.7 to 20.0; P<.01) or amiodarone (mean difference 9.4 beats/min; 95% CIs 2.0 to 16.7; P<.01). However, the 2 patients with postreversion symptomatic bradycardia were treated with digoxin and amiodarone, respectively. Both patients were elderly with no clinical or echocardiographic evidence of structural heart disease. Temporary, and subsequently, permanent pacemakers were inserted for a pre-
sumptive diagnosis of sick sinus syndrome in each patient. There were 2 episodes of transient hypotension in the sotalol group. The incidence of other side effects was minimal with no bronchospasm or hypersensitivity reactions and only one episode of thrombophlebitis (in the amiodarone group). The QTc interval was significantly prolonged in the sotalol group over the 48-hour period compared with the other treatment groups (suggesting that amiodarone has minimal class III effects within 48 hours using the trial dosing regimen). However, prolongation of the QTc interval was not associated with any episodes of torsade de pointes ventricular tachycardia in the sotalol group. Sinus rhythm was restored without clinically important complications by 48 hours in significantly more patients treated with sotalol and amiodarone than with digoxin. DISCUSSION
This study demonstrates that 48-hour reversion rates from recent-onset AF are significantly improved by active treatment (d,l-sotalol or amiodarone) compared with the more conservative rate control strategy with digoxin. Although it is known that spontaneous reversion occurs commonly in recent-onset AF, there is some evidence from small studies that digoxin is no better (or worse) than placebo in achieving reversion to sinus rhythm.17 We propose that in the population presenting with acute AF, despite high spontaneous reversion rates with rate control alone, the risk of adverse effects with these “active”
Table 3.
Successful reversion from AF. Outcome Time to reversion (h; mean±SEM) Rate at reversion (beats/min; mean±SEM) Reverted (%) 4h 24 h 48 h Cardioversion deferred Cardioversion successful Sinus rhythm at 48 h (%) Still in AF at 48 h (%)
Digoxin (n=36)
Amiodarone (n=39)
Sotalol (n=40)
Active Therapy* (n=79)
26.9±3.4 78.4±2.9
18.1±2.9† 76.0±2.7
13.0±2.5‡ 66.6±1.6†§
15.5±1.9‡
9 (25) 18 (50) 21 (58) 5 7/10 28 (78) 8 (22)
12 (31) 27 (69) 30 (77) 2 7/7 37 (94)† 2 (6)†
16 (40) 32 (80)† 35 (88)‡ 2 3/3 38 (95)† 2 (5)†
28 (36) 59 (76)† 65 (83)‡ 4 10/10† 75 (95)‡ 4 (5)‡
*
Combined sotalol and amiodarone treatment. versus digoxin. ‡P<.01 versus digoxin. §P<.05 versus amiodarone. †P<.05
6
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
drugs is small and is outweighed by the complications arising from persistent uncontrolled atrial fibrillation, namely chronic AF, stroke, and heart failure. In the United States, d,l-sotalol and amiodarone have not been commonly used to treat new-onset AF. Although both drugs are classified as class III antiarrhythmic agents (blocking potassium efflux and prolonging both action potential and refractory period), they also have important non–class III electrophysiologic effects. Commercially available d,l-sotalol is a racemic mix of the d-isomer (pure class III effect), and the l-isomer (potent β-adrenergic blocking effects). In the absence of any underlying myocardial dysfunction or elevated drug levels caused by renal failure,15 proarrhythmia is rare, probably because of the added protection from β blockade. Similarly, amiodarone has β-blocking properties, as well as use- and voltagedependent sodium and calcium channel blocking effects.18 The β-blocking and sodium/calcium channel blocking effects of amiodarone occur earlier than the class III effects, and this was reflected by the lack of QT prolongation in the first 48 hours. With long-term administration of high doses (>400 mg/d), amiodarone is associated with a significant incidence of pulmonary, thyroid, liver, dermatologic, and neurologic side effects.19 However, these adverse
effects are largely dose-related and are very uncommon at the doses used for AF.20 The success of the dosing regimens used, which were convenient and easily administered, is at odds with some previously reported studies for both amiodarone and sotalol. Some of these differences may relate to the prior duration and subsequent follow-up of AF/flutter, and different dosing regimens and routes. Amiodarone treatment has previously been reported to have low reversion rates or minimal improvement beyond placebo.21-24 In these studies, however, treatment was not commenced until up to 3,22 7,21,23 or 10 days24 after the onset of the arrhythmia. Our study and other recent studies25 would suggest better results when amiodarone is given earlier. Improved efficacy occurred despite predominantly oral administration rather than the more invasive central venous route used in the above studies. Some of these studies21,22 have used equivalent or higher loading doses of intravenous amiodarone. The studies by Galve et al21 and Donovan et al22 found no difference between amiodarone and placebo or amiodarone, flecainide, and placebo for time to reversion at 8 and 24 hours, respec-
Figure 3. Figure 2.
Kaplan-Meier distribution of likelihood of persistence of AF versus time. *P<.05 versus digoxin by Cox proportional hazards.
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
Mean (±SEM) ventricular rate in patients remaining in AF at 4, 24, and 48 hours. Number of patients in AF at 4, 24, and 48 hours, respectively, by group: amiodarone 27, 12, 9; sotalol 23, 7, 5; digoxin 27, 18, 15. *P<.05 versus amiodarone-treated and digoxin-treated patients.
7
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
tively. However, the latter study found amiodarone demonstrated significantly better rate control at 8 hours than flecainide or placebo. Similarly, other investigators26 have reported low success rates with sotalol compared with quinidine for reversion of chronic AF (duration 48 hours to 6 months) when treatment was started at a mean of 6 weeks after commencement of the arrhythmia. Although others have found low reversion rates with sotalol with a shorter duration of arrhythmia, they have had very short follow-up.27-29 For example, Sung et al29 concluded that sotalol doses up to 1.5 mg/kg could not terminate AF/flutter, but was able to reduce ventricular rate, although follow-up was for 1 hour only. They noted previous studies that showed that sotalol was effective in preventing AF once sinus rhythm was restored and postulated the reason for this was the phenomenon of “reverse use-dependence.” This effect describes the ability of sotalol to prolong atrial refractoriness more at low than high heart rates, and may explain the “late” efficacy of sotalol in our trial. In our study, early (within 4 hours) reversion rates after starting sotalol treatment were equally low, but had markedly improved by 48 hours when ventricular rate had slowed. Clearly, longer follow-up would be necessary to document similar favorable outcomes in the ongoing prevention of AF recurrence. We have also shown that the expectant (rate control) approach with digoxin is associated with more complications. Despite the short follow-up and relatively high rate of reversion in the digoxin-treated group, a signifi-
cant increase in complications occurred mainly because of heart failure from uncontrolled ventricular rate. Digoxin therapy has been advocated as the treatment of choice because spontaneous reversion to sinus rhythm within 24 hours has occurred in more than 50% of patients with AF of short duration. 13 Although digoxin does not improve reversion rates, it does provide some rate control, but this is least effective in those whose rhythms do not revert early.30 Left ventricular failure, the most frequent adverse outcome in our study, occurred only in patients who had sustained AF with poor ventricular rate control. Heart failure occurred in digoxin-treated patients with poor rate control even in the absence of echocardiographic left ventricular dysfunction. We note that sotalol therapy did not induce heart failure even in patients with significant undiagnosed left ventricular dysfunction, perhaps because of its excellent rate-controlling effects. This trial has some limitations. Since many patients were excluded because of the use of concomitant medication or intercurrent illness, there are many patients to whom these data cannot be extrapolated. The trial was prospective and randomized, but not double-blinded and, as a result, does introduce some risk for bias. The follow-up for this trial was only for 48 hours; thus, it is uncertain whether the increase in early reversion would result in a better long-term outcome such as avoidance of chronic AF. In addition, other drugs are frequently used for rate control as an alternative to digoxin, particularly the cal-
Table 4.
Adverse events. Adverse Event Major Hypotension Bradycardia Tachyarrhythmia Hypersensitivity reaction Left ventricular failure* Stroke Total major events Complication-free reversion Minor Mean change in QTc interval‡ (ms) QTc interval increase >50 ms Thrombophlebitis
Digoxin (n=36)
Amiodarone (n=39)
0 1 0 0 6 (2) 1 8 26
0 1 0 0 2 (2) 0 3 36†
–23.3±5.8 1 0
–10.8±6.2 2 1
Sotalol (n=40)
Active Therapy (n=79)
2 0 0 0 0† 0 2† 38† 3.1±6.1§ 5 0
2 1 0 0 2 0 5† 74† 7 1
*Clinical
signs of left ventricular failure (numbers in parentheses indicate those in whom echocardiogram showed a left ventricular ejection fraction <40%). versus digoxin. ‡QTc indicates QT interval corrected for R-R interval (QT/R -R ). §P<.05 versus digoxin, amiodarone. †P<.05
8
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
cium channel blockers, and better rate control with these agents may prevent heart failure. The study does confirm that active treatment is more beneficial than digoxin alone both for reversion and rate control. This may be better validated by a larger prospective, randomized, doubleblinded study that compares sotalol, amiodarone, and other drugs such as flecainide with placebo. We conclude that early initiation of therapy with sotalol or amiodarone to promote reversion of recent-onset AF results in significantly improved patient outcomes at 48 hours with minimal adverse events compared with rate control alone with digoxin. Despite high reversion rates in the digoxin group, a higher rate of uncontrolled, persistent AF resulted in more early complications, in particular, the development of heart failure. Further study is warranted to evaluate whether consistent application of this policy results in a long-term reduction in morbidity and mortality associated with AF. We thank Dr. Rohan Rajaratnam, Dr. John Vinen, Dr. John Morgan, Dr. Gregor Campbell-Hewson, Dr. Clyne Fernandes, and Dr. Masami Miyashita for help with randomization and adherence to protocol; Dr. John Gunning (Royal North Shore Hospital) and Dr. Charles Pawsey (Concord Repatriation General Hospital) for the cooperation and help of their respective cardiology departments in adherence to protocols; and Dr. David Whalley for helpful advice with the manuscript.
REFERENCES 1. Bialy D, Lehmann MH, Schumacher DN. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation [abstract]. J Am Coll Cardiol. 1992;19:41A. 2. Kannel WB, Abbott RD, Savage DD, et al. Epidemiologic features of chronic atrial fibrillation: the Framingham study. N Engl J Med. 1982;306:1018-1022. 3. Prystowsky EN, Benson DW Jr, Fuster V, et al. Management of patients with atrial fibrillation. A Statement for Healthcare Professionals. From the Subcommittee on Electrocardiography and Electrophysiology, American Heart Association. Circulation. 1996;93:1262-1277. 4. Packer DL, Bardy GH, Worley SJ, et al. Tachycardia-induced cardiomyopathy: a reversible form of left ventricular dysfunction. Am J Cardiol. 1986;57:563-570. 5. Alam M, Thorstrand C. Left ventricular function in patients with atrial fibrillation before and after cardioversion. Am J Cardiol. 1992;69:694-696.
13. Danias PG, Caulfield TA, Weigner MJ, et al. Likelihood of spontaneous conversion of atrial fibrillation to sinus rhythm. J Am Coll Cardiol. 1998;31:588-592. 14. Reimold SC. Avoiding drug problems. The safety of drugs for supraventricular tachycardia. Eur Heart J. 1997;18(Suppl C):C40-44. 15. Hohnloser SH. Proarrhythmia with class III antiarrhythmic drugs: types, risks, and management. Am J Cardiol. 1997;80:82G-89G. 16. Flaker GC, Blackshear JL, McBride R, et al. Antiarrhythmic drug therapy and cardiac mortality in atrial fibrillation. The Stroke Prevention in Atrial Fibrillation Investigators. J Am Coll Cardiol. 1992;20:527-532. 17. Falk RH, Knowlton AA, Bernard SA, et al. Digoxin for converting recent-onset atrial fibrillation to sinus rhythm. A randomised double-blinded trial. Ann Intern Med. 1987;106:503506. 18. Kodama I, Kamiya K, Toyama J. Cellular electropharmacology of amiodarone. Cardiovasc Res. 1997;35:13-29. 19. Connolly SJ. Evidence-based analysis of amiodarone efficacy and safety. Circulation. 1999;100:2025-2034. 20. Kochiadakis GE, Igoumenidis NE, Solomou MC, et al. Efficacy of amiodarone for the termination of persistent atrial fibrillation. Am J Cardiol. 1999;83:58-61. 21. Galve E, Rius T, Ballester R, et al. Intravenous amiodarone in treatment of recent-onset atrial fibrillation: results of a randomized, controlled study. J Am Coll Cardiol. 1996;27:10791082. 22. Donovan KD, Power BM, Hockings BE, et al. Intravenous flecainide versus amiodarone for recent-onset atrial fibrillation. Am J Cardiol. 1995;75:693-697. 23. Boriani G, Biffi M, Capucci A, et al. Conversion of recent-onset atrial fibrillation to sinus rhythm: effects of different drug protocols. Pacing Clin Electrophysiol. 1998;21:2470-2474. 24. Hou ZY, Chang MS, Chen CY, et al. Acute treatment of recent-onset atrial fibrillation and flutter with a tailored dosing regimen of intravenous amiodarone. A randomized, digoxincontrolled study. Eur Heart J. 1995;16:521-528. 25. Kochiadakis GE, Igoumenidis NE, Simantirakis EN, et al. Intravenous propafenone versus intravenous amiodarone in the management of atrial fibrillation of recent onset: a placebocontrolled study. Pacing Clin Electrophysiol. 1998;21:2475-2479. 26. Hohnloser SH, van de Loo A, Baedeker F. Efficacy and proarrhythmic hazards of pharmacologic cardioversion of atrial fibrillation: prospective comparison of sotalol versus quinidine. J Am Coll Cardiol. 1995;26:852-858. 27. Halinen MO, Huttunen M, Paakkinen S, et al. Comparison of sotalol with digoxin-quinidine for conversion of acute atrial fibrillation to sinus rhythm (the Sotalol-Digoxin-Quinidine Trial). Am J Cardiol. 1995;76:495-498. 28. Reisinger J, Gatterer E, Heinze G, et al. Prospective comparison of flecainide versus sotalol for immediate cardioversion of atrial fibrillation. Am J Cardiol. 1998;81:1450-1454. 29. Sung RJ, Hanno LT, Karagounis L, et al. Intravenous sotalol for the termination of supraventricular tachycardia and atrial fibrillation and flutter: a multicenter, randomized, double-blind, placebo-controlled study. Am Heart J. 1995;129:739-748. 30. Jordaens L, Trouerbach J, Calle P, et al. Conversion of atrial fibrillation to sinus rhythm and rate control by digoxin in comparison to placebo. Eur Heart J. 1997;18:643-648.
6. Akosah KO, Funai JT, Porter TR, et al. Left atrial appendage contractile function in atrial fibrillation. Influence of heart rate and cardioversion to sinus rhythm. Chest. 1995;107:690696. 7. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22:983-988. 8. Cairns JA, Connolly SJ. Nonrheumatic atrial fibrillation. Risk of stroke and role of antithrombotic therapy. Circulation. 1991;84:469-481. 9. Warfarin versus aspirin for prevention of thromboembolism in atrial fibrillation: Stroke Prevention in Atrial Fibrillation II Study [see comments]. Lancet. 1994;343:687-691. 10. Jouven X, Desnos M, Guerot C, et al. Idiopathic atrial fibrillation as a risk factor for mortality. The Paris Prospective Study I. Eur Heart J. 1999;20:896-899. 11. Wijffels MC, Kirchhof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation. 1995;92:1954-1968. 12. Goette A, Honeycutt C, Langberg JJ. Electrical remodeling in atrial fibrillation. Time course and mechanisms. Circulation. 1996;94:2968-2974.
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
9
A Prospective, Randomized Controlled Trial Comparing the Efficacy and Safety of Sotalol, Amiodarone, and Digoxin for the Reversion of New-Onset Atrial Fibrillation
From the Departments of Emergency Medicine* and Cardiology,‡ Royal North Shore Hospital, Sydney; and the Faculty of Medicine, University of Sydney, New South Wales, Australia.§
Anthony P. Joseph, MB BS*§ Michael R. Ward, FRACP, PhD‡§
Received for publication August 20, 1999. Revision received March 15, 2000. Accepted for publication March 27, 2000. Address for reprints: Anthony P. Joseph, MB BS, Department of Emergency Medicine, Royal North Shore Hospital, Pacific Highway, St. Leonards, New South Wales, 2065, Australia; 61-2-9926-7922, fax 61-2-9906-8123; E-mail [email protected]. Copyright © 2000 by the American College of Emergency Physicians. 0196-0644/2000/$12.00 + 0 47/1/107655 doi:10.1067/mem.2000.107655
Study objective: A prospective, randomized controlled trial of new-onset atrial fibrillation was conducted to compare the efficacy and safety of sotalol and amiodarone (active treatment) with rate control by digoxin alone for successful reversion to sinus rhythm at 48 hours. Methods: We prospectively randomly assigned 120 patients with atrial fibrillation of less than 24 hours’ duration to treatment with sotalol, amiodarone, or digoxin using a single intravenous dose followed by 48 hours of oral treatment. Patients had ECG monitoring for 48 hours, and time of reversion, adequacy of rate control, and numbers of adverse events were compared. After 48 hours, those still in atrial fibrillation underwent cardioversion according to a standardized protocol. After 48 hours of therapy and attempted cardioversion, the number of patients whose rhythms had successfully reverted were compared. Results: There was a significant reduction in the time to reversion with both sotalol (13.0±2.5 hours, P<.01) and amiodarone (18.1±2.9 hours, P<.05) treatment compared with digoxin only (26.9±3.4 hours). By 48 hours, the active treatment group was significantly more likely to have reverted to sinus rhythm than the rate control group (95% versus 78%, P<.05; risk ratio 5.4, 95% confidence interval [CI] 1.5 to 19.2 ). In those patients whose rhythms did not revert to sinus rhythm, there was superior ventricular rate control in the sotalol group at both 24 and 48 hours compared with those who received either amiodarone or digoxin. There were also fewer adverse events in the active treatment group compared with the rate control group. Conclusion: Immediate pharmacologic therapy for new-onset atrial fibrillation with class III antiarrhythmic drugs (sotalol or amiodarone) improves complication-free 48-hour reversion rates compared with rate control with digoxin.
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
1
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
[Joseph AP, Ward MR. A prospective, randomized controlled trial comparing the efficacy and safety of sotalol, amiodarone, and digoxin for the reversion of new-onset atrial fibrillation. Ann Emerg Med. July 2000;36:1-9.] INTRODUCTION
Atrial fibrillation (AF) is one of the most common tachyarrhythmias seen in clinical practice (overall prevalence 4%, with more than half of affected patients older than 75 years) and results in hospital stays in the United States that are longer than for any other arrhythmia.1,2 Although the incidence of AF is increased with structural abnormalities such as left atrial enlargement, left ventricular hypertrophy, and left ventricular dysfunction, lone AF (AF without structural heart disease) accounts for about a third of the disease prevalence.2 The ineffective atrial contractions in AF reduce effective cardiac output, especially during exercise, and substantially increase the risk of stroke as a result of the formation and embolization of atrial thrombus.3 Sustained AF with rapid ventricular rate may also result in reversible tachycardia-induced left ventricular dysfunction,4,5 further increasing left atrial stasis.6 AF increases the risk of ischemic stroke sixfold to approximately 5% per year,7 and embolic strokes related to AF are usually large with death or severe neurologic deficit in 50% to 70% of cases.8 Anticoagulation markedly reduces the occurrence of embolic stroke in AF, but also increases risk of major bleeding, particularly in the elderly in whom AF is most prevalent.9 Consequently, chronic AF is associated with a significant increase in morbidity and mortality independent of underlying cardiac pathology10; this has prompted strategies to promote reversion to and maintenance of sinus rhythm in acute AF. The repeated atrial depolarization in AF rapidly induces electrophysiologic changes in the atrial myocardium that promote recurrence of the arrhythmia immediately after reversion,11,12 such that the earlier reversion to sinus rhythm occurs, the greater the likelihood of sustained maintenance of sinus rhythm. However, enthusiasm for early pharmacologic treatment to promote reversion has been dampened by high spontaneous early reversion rates13 and concerns regarding adverse effects of antiarrhythmic drugs, particularly proarrhythmia. Proarrhythmias occur as a result of the tendency of some antiarrhythmic agents (class I and class III) to cause heterogeneous prolongation of the QT interval. This leads to increased inhomogeneity of ventricular repolarization, which may facilitate the production of a sustained ven-
2
tricular arrhythmia such as polymorphic ventricular tachycardia of the torsade de pointes type.14,15 Patients most at risk for this arrhythmia are those with ventricular dysfunction, hypokalemia, or baseline prolongation of the QT interval.3 Although the class III drugs, amiodarone and d,l-sotalol hydrochloride (sotalol), are the most frequently used to revert atrial fibrillation in Australia, the risk of proarrhythmic events in a low-risk population such as new-onset AF has not been well defined. Adverse effects of class III antiarrhythmic drugs, such as hypotension, are far more frequent in patients with impaired cardiac or renal function, whereas increased risk of proarrhythmia in patients with AF has only been convincingly demonstrated in trials with class I antiarrhythmic drugs, such as quinidine.15,16 The risks and benefits of commonly used regimens of sotalol and amiodarone have not been previously compared with ventricular rate control alone with digoxin in the treatment of recent-onset AF. Thus, we performed a prospective, randomized, controlled multicenter trial to assess the efficacy and safety of amiodarone and sotalol in conversion of new-onset AF to sinus rhythm at 48 hours compared with rate control alone with digoxin. M AT E R I A L S A N D M E T H O D S
From November 1992 until September 1997, 120 patients with new-onset (historically of less than 24 hours’ duration) AF or atrial flutter with ventricular rate more than 100 beats/min were enrolled. All patients presented to the emergency departments of 1 of 3 hospitals (Figure 1). The protocol was approved by the Medical Ethics Review Committees of the hospitals involved, and informed consent was obtained from all patients. Inclusion and exclusion criteria are noted in Table 1. Wide-complex tachycardia, such as AF with Wolff–Parkinson-White syndrome, was excluded because of the risk of development of ventricular fibrillation if digoxin is used in the presence of an accessory pathway. Significant left ventricular dysfunction was defined as left ventricular ejection fraction less than 30% or diuretic therapy with more than 40 mg of frusemide or equivalent per day. Patients were prospectively individually randomly assigned to 1 of the following 3 drug regimens for 48 hours: 1. Digoxin 500 µg intravenously over 30 minutes followed by 250 µg orally every 6 hours for 4 doses and then 250 µg daily (125 µg daily if any renal impairment as defined by serum creatinine concentration >120 µmol/L)
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
2. Amiodarone 5 mg/kg intravenously over 30 minutes, then 400 mg orally every 8 hours for 6 doses 3. Sotalol 1.5 mg/kg intravenously over 30 minutes, then 80 mg orally every 8 hours for 6 doses The randomization process was computer-generated and administered centrally. Initially the randomization
Figure 1.
process was open, with the study drug known to the treating physicians. However, after the enrollment of 85 patients, the investigators believed it was preferable to blind the treating physicians to the selected drug until inclusion and exclusion criteria were met, and consent obtained. Baseline serum potassium, serum thyroxine, and thyroid-
Acute atrial fibrillation/flutter (<24 hours) (n=120)
Flow chart from point of enrollment to completion of study.
Inclusion criteria met No exclusion criteria
Consent
Randomized
Protocol violations (n=5)
Digoxin (n=36)
Amiodarone (n=39)
Sotalol (n=40)
48 hours
48 hours
48 hours
Sinus rhythm (n=21)
Sinus rhythm at 48 hours (n=28; 78%)
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
Atrial fibrillation (n=15)
Sinus rhythm (n=30)
Atrial fibrillation (n=9)
Sinus rhythm (n=35)
Atrial fibrillation (n=5)
Cardioversion (n=10)
Cardioversion (n=7)
Cardioversion (n=3)
Sinus rhythm (n=7)
Sinus rhythm (n=7)
Sinus rhythm (n=3)
Sinus rhythm at 48 hours (n=37; 94%)
Sinus rhythm at 48 hours (n=38; 95%)
3
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
stimulating hormone were measured. Hyperthyroidism and hypothyroidism were defined by the individual laboratories’ normal ranges. All patients had ECG monitoring throughout the 48-hour trial period, and time of rhythm reversion and ventricular rate on reversion to sinus rhythm were noted. The ventricular rate of those remaining in AF was recorded at 4, 24, and 48 hours. If the patient’s rhythm had not reverted within 24 hours, intravenous heparin was then started with a 5,000-IU bolus and infusion at 1,000 IU/h and adjusted to activated partial thromboplastin time (APTT) of 2.0 to 2.5 times control. If the rhythm had not reverted to sinus rhythm by 48 hours and APTT prolongation had been achieved satisfactorily, electrical cardioversion was attempted using sequential shocks of 50, 100, 200, 300, and 360 J until sinus rhythm was restored. Cardioversion was postponed in the presence of left ventricular failure, systemic emboli, or inadequate anticoagulation (when APTT >2.0 times control had not been achieved for at least 4 hours). Twelvelead ECGs were recorded at baseline and 24 hours and the QTc interval was measured in lead II using Bazett’s formula -R ) averaged over 3 cycles. Echocardiography (QT/R was performed after at least 48 hours of sinus rhythm. Left atrial size and left ventricular fractional shortening were measured in the parasternal long-axis view. Adverse events noted included hypotension (mean arterial pressure [MAP] <70 mm Hg), onset of left ventricTable 1.
Inclusion and exclusion criteria. Inclusion criteria AF onset within 24 h Consent obtained Serum K+ >3.5 mmol/L and <5.5 mmol/L Serum creatinine <0.2 mmol/L Exclusion criteria AF present within 7 d and >24 h No consent Serum K+ <3.5 mmol/L and >5.5 mmol/L Serum creatinine >0.2 mmol/L Current β-blocker treatment Digoxin or sotalol treatment in last week Amiodarone treatment within 3 months Hypotension (MAP <70 mm Hg) Previous adverse reaction to any of trial medications Known thyroid disease Asthma/bronchospasm with β-blocker Wide-complex tachycardia Contraindication to anticoagulation Age <18 y Left ventricular dysfunction Pregnancy
4
ular failure (new inspiratory crepitations with signs of pulmonary venous congestion on chest radiograph), significant bradyarrhythmias (ventricular rate <40 beats/ min either in AF or sinus rhythm, or ventricular rate <60 beats/min with MAP <70 mm Hg), ventricular tachyarrhythmias (torsade de pointes polymorphic ventricular tachycardia or other), thrombophlebitis, stroke, hypersensitivity reactions (including bronchospasm), and prolongation of QTc interval by more than 50 ms. Trial medication was discontinued and observance of the trial protocol ceased for hypotension (MAP <70 mm Hg), tachyarrhythmias or bradyarrhythmias (as defined previously), or stroke. Because other therapies and treatment modalities were frequently used after discontinuation of trial protocol, the endpoints (reversion or ventricular rate) after discontinuation were analyzed as if the patients had remained in the trial until the completion of 48 hours or discharge from hospital. Statistical analysis was performed using SPSS statistical software (version 10.0; SPSS Inc, Chicago, IL). The primary outcome measure was successful reversion to sinus rhythm within 48 hours with antiarrhythmic treatment or cardioversion. Sample size (40 patients per group) was calculated before commencement of the trial allowing for pairwise comparisons between each of the 3 drugs. Assuming a reversion rate with digoxin of 50%, the sample size was calculated to be able to detect an absolute difference in reversion at 48 hours of 25% with each active treatment (ie, a 75% reversion rate with either amiodarone or sotalol) with P value less than .05 and a power of 0.80. Secondary endpoints included time to reversion, ventricular rate if still in AF at 4, 24, and 48 hours and adverse events. For continuous variables, all 3 groups were compared using analysis of variance, and post hoc testing was performed using Tukey’s test. For categorical variables, analysis was performed to compare the groups by χ2 testing or Fisher’s exact test where appropriate. Time to reversion was compared between the groups by Cox proportional hazards. Rate control was compared at 4, 24, and 48 hours by multiple linear regression with all baseline variables and treatment received entered as independent variables. R E S U LT S
One hundred twenty patients were enrolled in the trial. Five patients had protocol violations and were excluded from further analysis (4 patients in the digoxin and 1 in the amiodarone group). Baseline characteristics for the remaining 115 were similar in the 3 groups (Table 2). In 5
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
patients, the trial medication was ceased because of stroke (1 patient at 24 hours in the digoxin group), hypotension (2 patients in the sotalol group), or bradyarrhythmia (1 patient with 10.5-second sinus arrest on reversion in the amiodarone group and 1 patient with 12.0-second sinus arrest on reversion in the digoxin group). There were no deaths within the 48-hour study period. The numbers of patients with rhythm reversion by 48 hours in each of the sotalol and amiodarone groups were significantly greater than in the digoxin group (primary outcome measure, Table 3). This was because of significant increases in both spontaneous reversion with active drug treatment (occurring gradually over the 48 hours, Figure 2), and cardioversion success rate of those still in AF at 48 hours. Treatment with either sotalol or amiodarone significantly reduced time to reversion compared with digoxin therapy (Figure 2). The larger reduction observed in the sotalol group was not significantly different from that in the amiodarone group (P=.23; Table 3). No baseline variable was significantly associated with longer time to reversion by Cox proportional hazards, although there was a trend for those patients with increased left atrial size and left ventricular dysfunction. Of the 29 patients who remained in AF at 48 hours, cardioversion was not undertaken in 9 (Table 3). Four patients had significant left ventricular failure (all treated with digoxin) and cardioversion was postponed until
control of heart failure was achieved. One patient had a stroke (digoxin), 2 patients had inadequate anticoagulation (1 amiodarone and 1 sotalol), 1 patient had been withdrawn because of hypotension (sotalol), and 1 patient withdrew consent (amiodarone). When comparing the success of cardioversion at 48 hours, we found that significantly greater numbers of patients in the active treatment group who were still in AF had electrical reversion compared with those in the digoxin-treated group (100% versus 70%). At the end of the study period, patients treated with active therapy (sotalol or amiodarone) were significantly more likely to have reverted to sinus rhythm (P<.05; risk ratio 5.4; 95% CIs 1.5 to 19.2; Table 3). There were no significant differences between the treatment endpoints in the 3 treatment groups before and after the blinding to the study drug was initiated. In the patients who remained in AF at 4, 24, and 48 hours (Figure 3), sotalol therapy was associated with markedly lower ventricular rate compared with both digoxin and amiodarone (P<.05 at 24 and 48 hours). Higher ventricular rate at entry into the study was subsequently associated with higher ventricular rate if the patient was still in AF. Of those remaining in AF at 24 hours, adequate rate control (rate <100 beats/min) was achieved in 4 of 7 patients treated with sotalol but only 3 of 12 and 5 of 18 patients treated with amiodarone and digoxin, respec-
Table 2.
Baseline characteristics. Characteristic Atrial flutter Ventricular rate at enrollment (beats/min; mean±SEM) Age (y; mean±SEM) Male Serum K+ (mmol/L; mean±SEM) Thyroid-stimulating hormone (U/mL; mean±SEM) Hyperthyroid/hypothyroid Left atrium size (cm; mean±SEM) Left ventricular dysfunction* Underlying disease (known at admission) Ischemic heart disease Valvular heart disease Cardiomyopathy Hypertension No structural heart disease
Digoxin (n=36)
Amiodarone (n=39)
Sotalol (n=40)
1 141.3±3.8 64.9±2.0 20 4.09±0.06 2.30±0.34 0/2 3.95±0.10 6
1 144.4±3.1 61.3±2.6 25 4.14±0.05 2.64±0.99 3/1 3.97±0.11 8
1 137.5±3.6 62.8±2.4 19 4.24±0.07 1.97±0.19 1/0 3.84±0.10 6
3 4 1 10 18 (50%)
8 3 5 5 18 (45%)
7 1 0 6 26 (65%)
Left ventricular ejection fraction ≤40% on echocardiography after reversion.
*
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
5
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
tively. Similarly, at 48 hours adequate rate control for those still in AF was achieved in 3 of 5 patients treated with sotalol, but only 2 of 9 and 5 of 15 patients treated with amiodarone and digoxin, respectively. The drug regimens were well tolerated. The most common major adverse event was the onset of left ventricular failure; this occurred in those patients who had persistent or recurrent AF with poor rate control (Table 4). Of the 8 patients (6 treated with digoxin and 2 treated with amiodarone) who developed clinically significant left ventricular failure, only 4 (2 treated with digoxin and 2 treated with amiodarone) had significant left ventricular dysfunction on echocardiogram, suggesting that heart failure in the other 4 digoxin-treated patients was likely related solely to poor rate control. Despite its profound effect on ventricular rate in AF, sotalol treatment did not result in any clinically important bradyarrhythmias. Significant bradycardia occurred in only 2 patients, both on reversion to sinus rhythm. Sinus rate on reversion was significantly lower in patients who received sotalol compared with those treated with either digoxin (mean difference 11.8 beats/min; 95% CIs 3.7 to 20.0; P<.01) or amiodarone (mean difference 9.4 beats/min; 95% CIs 2.0 to 16.7; P<.01). However, the 2 patients with postreversion symptomatic bradycardia were treated with digoxin and amiodarone, respectively. Both patients were elderly with no clinical or echocardiographic evidence of structural heart disease. Temporary, and subsequently, permanent pacemakers were inserted for a pre-
sumptive diagnosis of sick sinus syndrome in each patient. There were 2 episodes of transient hypotension in the sotalol group. The incidence of other side effects was minimal with no bronchospasm or hypersensitivity reactions and only one episode of thrombophlebitis (in the amiodarone group). The QTc interval was significantly prolonged in the sotalol group over the 48-hour period compared with the other treatment groups (suggesting that amiodarone has minimal class III effects within 48 hours using the trial dosing regimen). However, prolongation of the QTc interval was not associated with any episodes of torsade de pointes ventricular tachycardia in the sotalol group. Sinus rhythm was restored without clinically important complications by 48 hours in significantly more patients treated with sotalol and amiodarone than with digoxin. DISCUSSION
This study demonstrates that 48-hour reversion rates from recent-onset AF are significantly improved by active treatment (d,l-sotalol or amiodarone) compared with the more conservative rate control strategy with digoxin. Although it is known that spontaneous reversion occurs commonly in recent-onset AF, there is some evidence from small studies that digoxin is no better (or worse) than placebo in achieving reversion to sinus rhythm.17 We propose that in the population presenting with acute AF, despite high spontaneous reversion rates with rate control alone, the risk of adverse effects with these “active”
Table 3.
Successful reversion from AF. Outcome Time to reversion (h; mean±SEM) Rate at reversion (beats/min; mean±SEM) Reverted (%) 4h 24 h 48 h Cardioversion deferred Cardioversion successful Sinus rhythm at 48 h (%) Still in AF at 48 h (%)
Digoxin (n=36)
Amiodarone (n=39)
Sotalol (n=40)
Active Therapy* (n=79)
26.9±3.4 78.4±2.9
18.1±2.9† 76.0±2.7
13.0±2.5‡ 66.6±1.6†§
15.5±1.9‡
9 (25) 18 (50) 21 (58) 5 7/10 28 (78) 8 (22)
12 (31) 27 (69) 30 (77) 2 7/7 37 (94)† 2 (6)†
16 (40) 32 (80)† 35 (88)‡ 2 3/3 38 (95)† 2 (5)†
28 (36) 59 (76)† 65 (83)‡ 4 10/10† 75 (95)‡ 4 (5)‡
*
Combined sotalol and amiodarone treatment. versus digoxin. ‡P<.01 versus digoxin. §P<.05 versus amiodarone. †P<.05
6
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
drugs is small and is outweighed by the complications arising from persistent uncontrolled atrial fibrillation, namely chronic AF, stroke, and heart failure. In the United States, d,l-sotalol and amiodarone have not been commonly used to treat new-onset AF. Although both drugs are classified as class III antiarrhythmic agents (blocking potassium efflux and prolonging both action potential and refractory period), they also have important non–class III electrophysiologic effects. Commercially available d,l-sotalol is a racemic mix of the d-isomer (pure class III effect), and the l-isomer (potent β-adrenergic blocking effects). In the absence of any underlying myocardial dysfunction or elevated drug levels caused by renal failure,15 proarrhythmia is rare, probably because of the added protection from β blockade. Similarly, amiodarone has β-blocking properties, as well as use- and voltagedependent sodium and calcium channel blocking effects.18 The β-blocking and sodium/calcium channel blocking effects of amiodarone occur earlier than the class III effects, and this was reflected by the lack of QT prolongation in the first 48 hours. With long-term administration of high doses (>400 mg/d), amiodarone is associated with a significant incidence of pulmonary, thyroid, liver, dermatologic, and neurologic side effects.19 However, these adverse
effects are largely dose-related and are very uncommon at the doses used for AF.20 The success of the dosing regimens used, which were convenient and easily administered, is at odds with some previously reported studies for both amiodarone and sotalol. Some of these differences may relate to the prior duration and subsequent follow-up of AF/flutter, and different dosing regimens and routes. Amiodarone treatment has previously been reported to have low reversion rates or minimal improvement beyond placebo.21-24 In these studies, however, treatment was not commenced until up to 3,22 7,21,23 or 10 days24 after the onset of the arrhythmia. Our study and other recent studies25 would suggest better results when amiodarone is given earlier. Improved efficacy occurred despite predominantly oral administration rather than the more invasive central venous route used in the above studies. Some of these studies21,22 have used equivalent or higher loading doses of intravenous amiodarone. The studies by Galve et al21 and Donovan et al22 found no difference between amiodarone and placebo or amiodarone, flecainide, and placebo for time to reversion at 8 and 24 hours, respec-
Figure 3. Figure 2.
Kaplan-Meier distribution of likelihood of persistence of AF versus time. *P<.05 versus digoxin by Cox proportional hazards.
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
Mean (±SEM) ventricular rate in patients remaining in AF at 4, 24, and 48 hours. Number of patients in AF at 4, 24, and 48 hours, respectively, by group: amiodarone 27, 12, 9; sotalol 23, 7, 5; digoxin 27, 18, 15. *P<.05 versus amiodarone-treated and digoxin-treated patients.
7
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
tively. However, the latter study found amiodarone demonstrated significantly better rate control at 8 hours than flecainide or placebo. Similarly, other investigators26 have reported low success rates with sotalol compared with quinidine for reversion of chronic AF (duration 48 hours to 6 months) when treatment was started at a mean of 6 weeks after commencement of the arrhythmia. Although others have found low reversion rates with sotalol with a shorter duration of arrhythmia, they have had very short follow-up.27-29 For example, Sung et al29 concluded that sotalol doses up to 1.5 mg/kg could not terminate AF/flutter, but was able to reduce ventricular rate, although follow-up was for 1 hour only. They noted previous studies that showed that sotalol was effective in preventing AF once sinus rhythm was restored and postulated the reason for this was the phenomenon of “reverse use-dependence.” This effect describes the ability of sotalol to prolong atrial refractoriness more at low than high heart rates, and may explain the “late” efficacy of sotalol in our trial. In our study, early (within 4 hours) reversion rates after starting sotalol treatment were equally low, but had markedly improved by 48 hours when ventricular rate had slowed. Clearly, longer follow-up would be necessary to document similar favorable outcomes in the ongoing prevention of AF recurrence. We have also shown that the expectant (rate control) approach with digoxin is associated with more complications. Despite the short follow-up and relatively high rate of reversion in the digoxin-treated group, a signifi-
cant increase in complications occurred mainly because of heart failure from uncontrolled ventricular rate. Digoxin therapy has been advocated as the treatment of choice because spontaneous reversion to sinus rhythm within 24 hours has occurred in more than 50% of patients with AF of short duration. 13 Although digoxin does not improve reversion rates, it does provide some rate control, but this is least effective in those whose rhythms do not revert early.30 Left ventricular failure, the most frequent adverse outcome in our study, occurred only in patients who had sustained AF with poor ventricular rate control. Heart failure occurred in digoxin-treated patients with poor rate control even in the absence of echocardiographic left ventricular dysfunction. We note that sotalol therapy did not induce heart failure even in patients with significant undiagnosed left ventricular dysfunction, perhaps because of its excellent rate-controlling effects. This trial has some limitations. Since many patients were excluded because of the use of concomitant medication or intercurrent illness, there are many patients to whom these data cannot be extrapolated. The trial was prospective and randomized, but not double-blinded and, as a result, does introduce some risk for bias. The follow-up for this trial was only for 48 hours; thus, it is uncertain whether the increase in early reversion would result in a better long-term outcome such as avoidance of chronic AF. In addition, other drugs are frequently used for rate control as an alternative to digoxin, particularly the cal-
Table 4.
Adverse events. Adverse Event Major Hypotension Bradycardia Tachyarrhythmia Hypersensitivity reaction Left ventricular failure* Stroke Total major events Complication-free reversion Minor Mean change in QTc interval‡ (ms) QTc interval increase >50 ms Thrombophlebitis
Digoxin (n=36)
Amiodarone (n=39)
0 1 0 0 6 (2) 1 8 26
0 1 0 0 2 (2) 0 3 36†
–23.3±5.8 1 0
–10.8±6.2 2 1
Sotalol (n=40)
Active Therapy (n=79)
2 0 0 0 0† 0 2† 38† 3.1±6.1§ 5 0
2 1 0 0 2 0 5† 74† 7 1
*Clinical
signs of left ventricular failure (numbers in parentheses indicate those in whom echocardiogram showed a left ventricular ejection fraction <40%). versus digoxin. ‡QTc indicates QT interval corrected for R-R interval (QT/R -R ). §P<.05 versus digoxin, amiodarone. †P<.05
8
ANNALS OF EMERGENCY MEDICINE
36:1 JULY 2000
REVERSAL OF NEW-ONSET ATRIAL FIBRILLATION Joseph & Ward
cium channel blockers, and better rate control with these agents may prevent heart failure. The study does confirm that active treatment is more beneficial than digoxin alone both for reversion and rate control. This may be better validated by a larger prospective, randomized, doubleblinded study that compares sotalol, amiodarone, and other drugs such as flecainide with placebo. We conclude that early initiation of therapy with sotalol or amiodarone to promote reversion of recent-onset AF results in significantly improved patient outcomes at 48 hours with minimal adverse events compared with rate control alone with digoxin. Despite high reversion rates in the digoxin group, a higher rate of uncontrolled, persistent AF resulted in more early complications, in particular, the development of heart failure. Further study is warranted to evaluate whether consistent application of this policy results in a long-term reduction in morbidity and mortality associated with AF. We thank Dr. Rohan Rajaratnam, Dr. John Vinen, Dr. John Morgan, Dr. Gregor Campbell-Hewson, Dr. Clyne Fernandes, and Dr. Masami Miyashita for help with randomization and adherence to protocol; Dr. John Gunning (Royal North Shore Hospital) and Dr. Charles Pawsey (Concord Repatriation General Hospital) for the cooperation and help of their respective cardiology departments in adherence to protocols; and Dr. David Whalley for helpful advice with the manuscript.
REFERENCES 1. Bialy D, Lehmann MH, Schumacher DN. Hospitalization for arrhythmias in the United States: importance of atrial fibrillation [abstract]. J Am Coll Cardiol. 1992;19:41A. 2. Kannel WB, Abbott RD, Savage DD, et al. Epidemiologic features of chronic atrial fibrillation: the Framingham study. N Engl J Med. 1982;306:1018-1022. 3. Prystowsky EN, Benson DW Jr, Fuster V, et al. Management of patients with atrial fibrillation. A Statement for Healthcare Professionals. From the Subcommittee on Electrocardiography and Electrophysiology, American Heart Association. Circulation. 1996;93:1262-1277. 4. Packer DL, Bardy GH, Worley SJ, et al. Tachycardia-induced cardiomyopathy: a reversible form of left ventricular dysfunction. Am J Cardiol. 1986;57:563-570. 5. Alam M, Thorstrand C. Left ventricular function in patients with atrial fibrillation before and after cardioversion. Am J Cardiol. 1992;69:694-696.
13. Danias PG, Caulfield TA, Weigner MJ, et al. Likelihood of spontaneous conversion of atrial fibrillation to sinus rhythm. J Am Coll Cardiol. 1998;31:588-592. 14. Reimold SC. Avoiding drug problems. The safety of drugs for supraventricular tachycardia. Eur Heart J. 1997;18(Suppl C):C40-44. 15. Hohnloser SH. Proarrhythmia with class III antiarrhythmic drugs: types, risks, and management. Am J Cardiol. 1997;80:82G-89G. 16. Flaker GC, Blackshear JL, McBride R, et al. Antiarrhythmic drug therapy and cardiac mortality in atrial fibrillation. The Stroke Prevention in Atrial Fibrillation Investigators. J Am Coll Cardiol. 1992;20:527-532. 17. Falk RH, Knowlton AA, Bernard SA, et al. Digoxin for converting recent-onset atrial fibrillation to sinus rhythm. A randomised double-blinded trial. Ann Intern Med. 1987;106:503506. 18. Kodama I, Kamiya K, Toyama J. Cellular electropharmacology of amiodarone. Cardiovasc Res. 1997;35:13-29. 19. Connolly SJ. Evidence-based analysis of amiodarone efficacy and safety. Circulation. 1999;100:2025-2034. 20. Kochiadakis GE, Igoumenidis NE, Solomou MC, et al. Efficacy of amiodarone for the termination of persistent atrial fibrillation. Am J Cardiol. 1999;83:58-61. 21. Galve E, Rius T, Ballester R, et al. Intravenous amiodarone in treatment of recent-onset atrial fibrillation: results of a randomized, controlled study. J Am Coll Cardiol. 1996;27:10791082. 22. Donovan KD, Power BM, Hockings BE, et al. Intravenous flecainide versus amiodarone for recent-onset atrial fibrillation. Am J Cardiol. 1995;75:693-697. 23. Boriani G, Biffi M, Capucci A, et al. Conversion of recent-onset atrial fibrillation to sinus rhythm: effects of different drug protocols. Pacing Clin Electrophysiol. 1998;21:2470-2474. 24. Hou ZY, Chang MS, Chen CY, et al. Acute treatment of recent-onset atrial fibrillation and flutter with a tailored dosing regimen of intravenous amiodarone. A randomized, digoxincontrolled study. Eur Heart J. 1995;16:521-528. 25. Kochiadakis GE, Igoumenidis NE, Simantirakis EN, et al. Intravenous propafenone versus intravenous amiodarone in the management of atrial fibrillation of recent onset: a placebocontrolled study. Pacing Clin Electrophysiol. 1998;21:2475-2479. 26. Hohnloser SH, van de Loo A, Baedeker F. Efficacy and proarrhythmic hazards of pharmacologic cardioversion of atrial fibrillation: prospective comparison of sotalol versus quinidine. J Am Coll Cardiol. 1995;26:852-858. 27. Halinen MO, Huttunen M, Paakkinen S, et al. Comparison of sotalol with digoxin-quinidine for conversion of acute atrial fibrillation to sinus rhythm (the Sotalol-Digoxin-Quinidine Trial). Am J Cardiol. 1995;76:495-498. 28. Reisinger J, Gatterer E, Heinze G, et al. Prospective comparison of flecainide versus sotalol for immediate cardioversion of atrial fibrillation. Am J Cardiol. 1998;81:1450-1454. 29. Sung RJ, Hanno LT, Karagounis L, et al. Intravenous sotalol for the termination of supraventricular tachycardia and atrial fibrillation and flutter: a multicenter, randomized, double-blind, placebo-controlled study. Am Heart J. 1995;129:739-748. 30. Jordaens L, Trouerbach J, Calle P, et al. Conversion of atrial fibrillation to sinus rhythm and rate control by digoxin in comparison to placebo. Eur Heart J. 1997;18:643-648.
6. Akosah KO, Funai JT, Porter TR, et al. Left atrial appendage contractile function in atrial fibrillation. Influence of heart rate and cardioversion to sinus rhythm. Chest. 1995;107:690696. 7. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22:983-988. 8. Cairns JA, Connolly SJ. Nonrheumatic atrial fibrillation. Risk of stroke and role of antithrombotic therapy. Circulation. 1991;84:469-481. 9. Warfarin versus aspirin for prevention of thromboembolism in atrial fibrillation: Stroke Prevention in Atrial Fibrillation II Study [see comments]. Lancet. 1994;343:687-691. 10. Jouven X, Desnos M, Guerot C, et al. Idiopathic atrial fibrillation as a risk factor for mortality. The Paris Prospective Study I. Eur Heart J. 1999;20:896-899. 11. Wijffels MC, Kirchhof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation. 1995;92:1954-1968. 12. Goette A, Honeycutt C, Langberg JJ. Electrical remodeling in atrial fibrillation. Time course and mechanisms. Circulation. 1996;94:2968-2974.
JULY 2000
36:1
ANNALS OF EMERGENCY MEDICINE
9