Pii: s0002-9149(99)00168-x

Comparison of Sotalol Versus Quinidine
for Maintenance of Normal Sinus Rhythm
in Patients With Chronic
Atrial Fibrillation
Mary Ross Southworth, PharmD, Dawn Zarembski, PharmD, Marlos Viana, PhD, Many clinicians choose sotalol for the prevention of
(range 42% to 58%), quinidine 53% (range 48% to 59%),
recurrences of atrial fibrillation (AF) as an alternative to
and control 32% (range 26% to 39%). When combining
quinidine, which has been associated with an increase
and comparing mortality effects, the following studies
in long-term mortality. Using meta-analytic techniques,
met the same inclusion criteria: 4 sotalol studies, 9 quin-
we compared the effects on maintenance of sinus
idine studies, and 7 control studies. The point estimates
rhythm and mortality of combined groups of patients
and corresponding credibility intervals for mortality in
with chronic AF treated with sotalol, quinidine, or a
the 3 groups were sotalol 2.2% (range 0.6% to 4.8%),
control drug. Rates of conversion at 6 months and mor-
quinidine 3.0% (range 1.7% to 4.7%), and control 1.1%
tality were combined for each group after performing
(range 0.3% to 2.4%). Sotalol and quinidine are compa-
sensitivity analysis to test for homogeneity. Bayesian
rable in their ability to maintain sinus rhythm at 6
estimates and corresponding 95% credibility intervals
months (about 50%) and both agents are superior to
were constructed to compare the probabilities of achiev-
control. There is a trend for both agents to increase
ing sinus rhythm and mortality among groups. A litera-
ture search revealed 4 sotalol studies, 6 quinidine stud-
mortality with long-term therapy. These data do not
ies, and 5 control studies that met inclusion criteria
support choosing sotalol over quinidine as a safer alter-
established a priori. The point estimates for maintaining
native for preventing recurrences of chronic AF.
normal sinus rhythm (at 6 months) and corresponding
ᮊ1999 by Excerpta Medica, Inc.
credibility intervals for the 3 groups were sotalol 50%
(Am J Cardiol 1999;83:1629 –1632)
Although controversial, the use of antiarrhythmic have conducted a meta-analysis to possibly shed light
agents for maintenance of sinus rhythm (SR) in on this controversy. Our objective was to compare the patients with atrial fibrillation (AF) is a common clin- efficacy and safety of quinidine, sotalol, and control in ical practice. However, the results of several studies1,2 maintaining SR in patients with chronic AF.
have questioned the wisdom of using Vaughn-Wil-liams type I antiarrhythmic drugs for this indication.
In a previously published meta-analysis by Coplen et al,1 quinidine was compared with control for main- MEDLINE databases of the National Library of Med- taining normal SR. Whereas quinidine proved to be icine. From 1985 to the present, the search term “so- more efficacious, it was at the cost of a significantly talol” was crossed with “atrial fibrillation” and “atrial higher mortality rate. Further, analysis of data from flutter”; results of this search were crossed with the the Stroke Prevention in Atrial Fibrillation Study2 broad search strategy “clinical trial.” A similar search revealed an increased risk for cardiac and arrhythmic was performed from 1990 to the present using the mortality in patients with congestive heart failure re- search term “quinidine” These data were combined ceiving antiarrhythmic drugs (mostly quinidine and with data extracted from the previous meta-analysis1 procainamide). Consequently, sotalol has become to form a comprehensive collection of trials. In addi- more frequently prescribed as an alternative to agents tion, reference lists associated with extracted articles such as quinidine,3 although superiority over other were searched for pertinent studies.
antiarrhythmic drugs has not been conclusively dem- The criteria for study inclusion were determined a onstrated. In the absence of a robust, placebo-con- priori. Studies that (1) involved the therapy of chronic trolled trial comparing the 2 antiarrhythmic drugs, we (Ͼ72 hours1) AF or flutter, (2) used a randomized trialdesign, and (3) had data available at 3, 6, or 12 monthsregarding mortality and/or the proportion of patients From the Departments of Pharmacy Practice and Medicine, Section of remaining in normal SR were included. Criteria for Cardiology and the Division of Biostatistics, School of Public Health,University of Illinois at Chicago, Chicago, Illinois; and the Department exclusion from the study were studies that (1) in- of Pharmacy Practice, Chicago College of Pharmacy, Downers volved therapy of postoperative AF or paroxysmal Grove, Illinois. Manuscript received October 13, 1998; revised AF, (2) used nonracemic or intravenous sotalol (or manuscript received February 11, 1999, and accepted February 14.
quinidine), or (3) used crossover design.
Address for reprints: Jerry Bauman, PharmD, University of Illinois at Two investigators screened all extracted articles Chicago, Department of Pharmacy Practice (M/C 886), 833 SouthWood Street, Chicago, Illinois 60612-7230.
independently for entry into the study. If the investi- gators disagreed or interpreted results differently, a 65 years of age. Within each included study, the third investigator resolved the discrepancy. In the gender distribution was similar. For studies that pro- event that a study did not provide data concerning vided data, combined prevalence of valvular disease mortality or efficacy rates, attempts were made to for quinidine, sotalol, and control drug were 55%, 9%, contact the author. A reject log was kept of all ex- and 65%, respectively; prevalence of coronary artery cluded studies. One investigator extracted data regard- disease was 23%, 15%, and 19%. Despite the dispar- ing efficacy and mortality; a second investigator con- ate prevalence of these concomitant diseases between firmed the accuracy of the extracted data.
the studies that provided those data, within each indi- Statistical analysis: Methods used to analyze data
vidual study, these rates were similar in the compared were similar to those used in a previously published groups.6–16 The mean duration of AF before restora- study.4 The hypothesis H of parametric homogeneity tion of SR ranged from 1.3 to 24 months. Information (combinability) among individual studies was as- regarding the prevalence of congestive heart failure sessed using the observed proportions (safety and efficacy) and corresponding posterior probability for With regard to cardioversion, a variety of tech- the hypothesis H. The posterior probability P(H) is niques were used. Some studies initiated antiarrhyth- derived from the Bayes factor, which is the ratio of averaged likelihoods for the homogeneity hypothesis sion,6,7,9,10,12,15,16 while others started drug therapy and its alternative, and its value quantifies its credi- only after successful electrical cardioversion.8,11,14 A bility according to the observed data.5 If P(H) is few studies only used electrical cardioversion in some calculated to be close to 0, then there is strong evi- patients,10,15,16 while one did not use electrical cardio- dence against the hypothesis. Conversely, if P(H) is version at all.13 It is difficult to assess the success rate close to 1, then there is strong evidence for the hy- of the acute cardioversion techniques, because many pothesis. For each group (quinidine, sotalol, and con- studies did not include information on patients who trol) P(H) was calculated based on all (k) studies and, did not achieve SR.7,10 For the purposes of our anal- additionally, based on k-1 studies (removing 1 study ysis we only included long-term (i.e., 3-, 6-, and at a time). Groups were considered homogenous with 12-month) efficacy rates in patients who achieved SR regard to efficacy and safety data if P(H) was Ͼ0.75.
Point estimates were provided with (Bayesian) 95% For evaluation of efficacy (the percentage of pa- tients remaining in SR at certain time points), the The observed number of patients remaining in nor- following time points were considered: 3, 6, and 12 mal SR at each predetermined time point versus the months after initiation of antiarrhythmic therapy. Ef- total number of patients achieving normal SR was ficacy rates in the quinidine group at 3 months were used to devise point estimates and 95% credibility reported in 8 studies and ranged from 55% to 91%. At limits (binomial parameters), representing the proba- 6 months, 6 studies provided efficacy data, which bility of remaining in normal SR for each time point.
ranged from 42% to 73%. Twelve-month data were Likewise, the probability of death in each drug group reported in 3 studies and ranged from 29% to 52%.
was also estimated for each included study.
With regard to the sotalol-treated group, efficacy dataat 3 months was available in 1 study only, 63%.
Six-month data were available in 4 studies and ranged The literature search identified 19 quinidine-related from 42% to 71%. No data were available for the studies and 14 sotalol-related studies. Of these, 10 12-month time point. In the control groups, the ranges sotalol and 10 quinidine studies were excluded for for 3, 6, and 12 months were 33% to 57% (6 studies), various reasons: study subjects were not being treated 0% to 38% (5 studies), and 5% to 28% (3 studies).
for chronic AF or were being treated for paroxysmal The most complete time point for data regarding AF (6 studies), data were not available for predeter- drug efficacy was 6 months, with 6 quinidine, 4 so- mined time points (9 studies), subjects were being talol, and 5 control studies providing data for analysis.
treated for postoperative AF (6 studies), inappropriate A total of 304 quinidine-treated, 155 sotalol-treated, drug form was used (IV quinidine, nonracemic so- and 185 control patients were compared with regard to talol) (3 studies), and the study used a crossover ability to maintain normal SR at 6 months.
design (2 studies). Of the remaining studies, 7 had Homogeneity testing revealed the posterior proba- control groups that were appropriate for inclusion in bility for remaining in normal SR at 6 months for all studies to be 0.775 for the quinidine group, 0.846 for Although studies of patients with paroxysmal AF the sotalol group, and 0.951 for the control group, were excluded a priori, 1 study6 included patients with implying that data from all of the groups were com- both chronic and paroxysmal AF. Because it was binable. The point estimates for remaining in SR with impossible to determine whether the deaths reported 95% credibility intervals for quinidine, sotalol, and in this study occurred in patients with chronic or control were 53% (48 to 59), 50% (42 to 58), and 32% paroxysmal AF, all patients were included in the mor- (26 to 39), respectively (Figure 1). Quinidine and sotalol were significantly more successful in maintain- All included studies6–16 were published between 1970 and 1995, and are indicated in Table I. Mean Mortality rates were computed for each group and ages reported in each study group ranged from 44 to were based on the occurrence of death throughout the 1630 THE AMERICAN JOURNAL OF CARDIOLOGYா
TABLE I Studies Included in Analysis
* Number of patients analyzed for mortality; see text for details.
† Graphed data.
— ϭ no information available.
FIGURE 2. Mortality during entire study duration shown as point
FIGURE 1. Percent of patients remaining in sinus rhythm at 6
estimates and 95% credibility intervals.
months shown as point estimates and 95% credibility intervals.
Q ؍ quinidine; S ؍ sotalol.
duration of each individual study; observations from tistical nonsignificance; however, there was a numer- 177 sotalol, 495 quinidine, and 365 control patients ical trend for both quinidine and sotalol to increase mortality with long-term therapy. Of the 15 patients Homogeneity testing revealed the posterior proba- who died in the quinidine group, 4 deaths were sudden bility for death throughout the study duration for all or attributed to ventricular arrhythmias. Three patients studies to be Ͼ0.999 for the control group, Ͼ0.999 for died in the sotalol group; 1 death was attributed to a the quinidine group, and 0.996 for the sotalol group, proarrhythmic event. None of the 3 deaths in the again implying that data from all of the groups were control group were arrhythmic or sudden in nature.
combinable. The point estimates and 95% credibilityintervals for probability of death for the quinidine, DISCUSSION
sotalol, and control groups were 3.0% (1.7 to 4.7), Since the early 1900s, quinidine has been consid- 2.2% (0.6 to 4.8), and 1.1% (0.3 to 2.4), respectively ered the drug of first choice for maintaining SR.
(Figure 2). The posterior probabilities associated with However, the visible and influential publication of the mortality rate inequalities were Ͻ10%, implying sta- Coplen et al meta-analysis after the CAST trial17 ARRHYTHMIAS AND CONDUCTION DISTURBANCES/SOTALOL VERSUS QUINIDINE FOR ATRIAL FIBRILLATION further questioned the long-term safety of antiarrhyth- One may logically ask which antiarrhythmic agent mic drugs. In the analysis of Coplen et al, randomized, is safe in patients with AF? One possibility is the use controlled trials investigating the efficacy and safety of amiodarone.20,21 Another option being investigated of quinidine in AF were combined. Quinidine was in the Atrial Fibrillation Follow-up Investigation of more effective than control (53% vs 32% remaining in Rhythm Management trial22 is the feasibility of using normal SR at 6 months), but the odds of dying were no antiarrhythmic therapy at all and simply using 2.7 times higher in quinidine-treated patients. Al- therapies aimed at rate control. Nevertheless, in the though not all deaths in the quinidine group could be absence of definitive, prospective trials, the data pre- clearly attributed to drug toxicity, 5 were from un- sented here do not support the use of sotalol as a safer known causes and therefore could have been the result alternative to quinidine in the maintenance of SR in of proarrhythmia. As a result of this study and others, clinicians turned to alternative agents such as sotalolin hope of finding an efficacious agent with a superiorsafety profile.
1. Coplen SE, Antman EM, Berlin JA, Hewitt P, Chalmers T. Efficacy and safety
Similar to the head-to-head comparison of sotalol of quinidine therapy for maintenance of sinus rhythm after cardioversion: a and quinidine by Juul-Moller et al,14 the 6-month meta-analysis of randomized control trials. Circulation 1990;82:1106 –1116.
2. Flaker GC, Blackshear JL, McBride R, Kronmal RA, Halperin JL, Hart RG.
efficacy rates in our analysis were very similar be- Antiarrhythmic drug therapy and cardiac mortality in atrial fibrillation. J Am Coll tween the 2 drugs and both were significantly greater Cardiol 1992;20:527–532.
3. Phillips BG, Bauman JL. Prescribing trends and pharmacoeconomic consid-
than that achieved in the control group. Although eration in the treatment of arrhythmias. Focus on atrial fibrillation and flutter.
maintenance of SR with sotalol or quinidine is signif- PharmacoEconomics 1995;7:521–533.
icantly more effective than no therapy, about 50% of 4. Roberts SA, Viana MA, Nazari J, Bauman JL. Invasive and noninvasive
methods to predict the long term efficacy of amiodarone: a compilation of clinical
patients will have recurrent AF during the first 6 observations using meta-analysis. PACE 1994;17:1590 –1602.
5. Viana MA. Bayesian joint estimation of binomial proportions. J Educat Stat
In the present analysis, we found that there was a 1991;16:331–343.
6. Reimold SC, Cantillon CO, Friedman PL, Antman EM. Propafenone versus
similar numerical trend toward increased mortality in sotalol for suppression recurrent symptomatic atrial fibrillation. Am J Cardiol both the quinidine and sotalol groups when compared with control. However, in the sotalol and quinidine 7. Byrne-Quinn E, Wing AJ. Maintenance of sinus rhythm after DC reversion of
atrial fibrillation: a double-blind controlled trial of quinidine bisulphate. Br
group, the posterior probability associated with mor- tality rate inequalities was Ͻ10%, implying statistical 8. Lloyd EA, Gersh BJ, Forman R. The efficacy of quinidine and disopyramide
in the maintenance of sinus rhythm after electroconversion from atrial fibrillation.
nonsignificance. A closer look at the data reveals a S Afr Med J 1984;65:367–369.
variety of causes of death in the study patients. Of the 9. Hillestad L, Bjerkelund C, Dale J, Maltau J, Storstein O. Quinidine in
15 patients who died in the quinidine studies, 4 (27%) maintenance of sinus rhythm after electroconversion of chronic atrial fibrillation:a controlled clinical study. Br Heart J 1971;33:518 –521.
were sudden deaths or deaths due to lethal ventricular 10. Sodermark T, Jonsson B, Olsson A, Oro L, Wallin H, Edhag O, Sjogren A,
arrhythmias. Similarly, of the 3 patients who died in Danielsson M, Rosenhamer G. Effect of quinidine on maintaining sinus rhythm the sotalol trials, 1 (33%) could be directly attributed after conversion of atrial fibrillation or flutter. A multicentre study from Stock-holm. Br Heart J 1975;37:486 – 492.
to the drug. Although not all deaths can be clearly 11. Boissel JP, Wolf E, Gillet J, Soubrane A, Cavallaro A, Mazoyer G, Delahaye JP.
attributed to the drugs, these results are still distress- Controlled trial of a long acting quinidine for maintenance of normal sinus rhythm ing, e.g., none of the deaths in the 365 patients in the after conversion of sustained atrial fibrillation. Eur Heart J 1981;2:49 –55.
12. Hartel G, Vouhiha A, Konttinen A, Halonen PI. Value of quinidine in
control group were from ventricular arrhythmias or were maintenance of sinus rhythm after electric conversion of atrial fibrillation. Br sudden in nature. The observation that both sotalol and Heart J 1970;32:57– 60.
13. Zehender M, Hohnloser S, Muller B, Meinertz T, Just H. Effects of amiodarone
quinidine caused similar trends in increasing overall versus quinidine and verapamil in patients with chronic atrial fibrillation: results of a mortality, and combined with the results of other analy- comparative study and a 2 year follow up. J Am Coll Cardiol 1992;19:1054 –1059.
ses,1,2 underscores the potential hazards in using sotalol 14. Juul-Moller S, Edvardsson N, Rehnqvist-Ahlberg N. Sotalol versus quinidine
for the maintenance of sinus rhythm after direct current conversion of atrial
as an alternative to quinidine in patients with AF.
fibrillation. Circulation 1990;82:1932–1939.
There are certain limitations to our analysis. Pa- 15. Hohnloser S, van de Loo A, Baedeker F. Efficacy and proarrhythmic hazards
tient-specific variables such as left ventricular func- of pharmacologic cardioversion of atrial fibrillation: prospective comparison ofsotalol versus quinidine. J Am Coll Cardiol 1995;26:852– 858.
tion, left atrial size, and valvular function influence 16. Singh S, Saini RK, Dimarco J, Kluger J, Gold R, Chen Y. Efficacy and safety
recurrence rate and overall mortality in patients with of sotalol in digitalized patients with chronic atrial fibrillation. Am J Cardiol AF. Clearly, it would be worthwhile to know the 1991; 68:1227–1230.
17. Echt DS, Liebson PR, Mitchell LB, Peters RW, Obias-Manno D, Barker AH,
prevalence of each of these variables in our study Arensberg D, Baker A, Friedman L, Greene HL, Huthen ML, Richardson DW.
groups. Unfortunately, most studies included in our Mortality and morbidity in patients receiving encainide, flecainide, or placebo.
The cardiac arrhythmia suppression trial. N Engl J Med 1991;324:781–788.
review did not provide data concerning left ventricular 18. Massell BF, Chute CG, Walker AM, Kurland GS. Penicillin and the marked
function7–14,16 or atrial size7–12 (Table I). Further, the decrease in morbidity and mortality from rheumatic fever in the United States.
incidence of valvular disease was markedly dissimilar N Engl J Med 1988;318:280 –286.
19. World Health Organization. Rheumatic fever and rheumatic heart disease.
in the quinidine, sotalol, and control groups. Many of WHO Technical Report Series 764. Geneva, World Health Organization, 1988.
the studies that provided data on quinidine and control 20. Gosselink AT, Crijns HJ, VanGelder IC, Hillige H, Wiesfeld ACP, Lie KI.
were published between 1970 and 1981; whereas the Low dose amiodarone for maintenance of sinus rhythm after cardioversion ofatrial fibrillation or flutter. JAMA 1991;267:3289 –3293.
sotalol studies were published more recently. The 21. Zarembski DG, Nolan PE, Slack MK, Caruso AC. Treatment of resistant
greater incidence of valvular disease in the earlier atrial fibrillation: a meta-analysis comparing amiodarone and flecainide. Arch studies likely reflects the declining prevalence, over Intern Med 1995;155:1885–1891.
22. AFFIRM Investigators. Atrial fibrillation follow-up investigation of rhythm
management—the AFFIRM study design. Am J Cardiol 1997;79:1198 –1202.
1632 THE AMERICAN JOURNAL OF CARDIOLOGYா

Source: http://tigger.cc.uic.edu/labs/crc/WebApplication1/education/spring2002/02_rotation_article_2.pdf