| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;82:1927-1937
© 2006 The Society of Thoracic Surgeons
a Department of Critical Care Medicine, University of Calgary, Calgary, Alberta, Canada
b Department of Cardiac Sciences, University of Calgary, Calgary, Alberta, Canada
c Department of Medicine, University of Calgary, Calgary, Alberta, Canada
d Department of Community Health Sciences, University of Calgary, Calgary, Alberta, Canada
e Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada
f Centre for Health and Policy Studies, University of Calgary, Calgary, Alberta, Canada
g Department of Intensive Care, Austin Hospital, Melbourne, Victoria, Australia
Accepted for publication June 12, 2006.
* Address correspondence to Dr Ghali, Faculty of Medicine, University of Calgary, 3330 Hospital Dr NW, Calgary, AB T2N 4N1, Canada. (Email: wghali{at}ucalgary.ca).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Perioperative amiodarone has been proposed as an intervention to decrease the incidence of AF after cardiac surgery. Several studies have evaluated the efficacy of amiodarone in this setting. However, their results are difficult to generalize owing to differences in study populations, inconsistent use of perioperative ß-blockade, and considerable diversity in the protocols for amiodarone administration. While prior systematic reviews have assessed the use of perioperative amiodarone, results have often been limited by coincident evaluation of multiple interventions for prevention of postoperative AF, by inclusion of few or duplicate studies, by omission of more recent publications, or by failure to perform a pooled analysis [7–13]. Recently, the Prophylactic Oral Amiodarone for the Prevention of Arrhythmias that Begin Early After Revascularization, valve Replacement or Repair (PAPABEAR) trial, the largest study of perioperative amiodarone in cardiac surgery, has been reported and significantly expands the existing body of evidence on the efficacy of amiodarone for preventing postoperative AF [14]. Therefore, our primary objective was to provide an up-to-date evaluation of the effect of amiodarone on postoperative AF. Our secondary objectives were to assess the impact of prophylactic amiodarone on major perioperative cardiovascular complications including ventricular tachyarrhythmias, neurologic events (transient ischemic attack or stroke), and mortality, as well as to assess the impact on duration of hospitalization and total health care costs. In addition, we conducted meta-regression and sensitivity analyses to determine whether particular study quality or clinical factors, in particular perioperative ß-blocker use, influence the apparent effect of amiodarone on the incidence of AF.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The Medline (through April 2006), Embase (through April 2006), and Cochrane Controlled Clinical Trials Register (through April 2006) databases were searched using an approach recommended for systematic reviews of RCTs [15]. We derived four comprehensive search themes that were then combined using the Boolean operator "and." The first theme used a highly sensitive RCT filter method [16]. The second theme, amiodarone, was created by a search using an exploded medical subject headings and textword search for: "amiodarone." The third theme, atrial fibrillation, was created by using the Boolean search term "or" to search for the following terms appearing as both exploded medical subject headings and text words: "atrial fibrillation" or "atrial flutter" or "supraventricular tachycardia." The fourth theme, cardiac surgery, was created by a search using an exploded medical subject heading and textword search for "cardiac surgery" or "coronary artery bypass" or "valvular."
Study Selection Criteria
An initial screen of identified abstracts was conducted independently by two persons (S.M.B. and P.D.G.) to confirm the report of original data. Relevant reports were then independently reviewed in full by the same two persons for eligibility based on four inclusion criteria: (1) study design (RCTs), (2) target population (patients undergoing cardiac surgery), (3) intervention (trials of amiodarone versus control), and (4) outcome (documentation of postoperative AF).
Data Extraction
The reviewers then independently extracted data from all studies fulfilling eligibility criteria. Data extracted included details of study protocol, details of amiodarone administration, and baseline clinical characteristics. The primary outcome was the incidence of postoperative AF. Secondary outcome measures included differences in heart rate at onset of AF, time to onset and duration of AF, incidence of ventricular tachyarrhythmias and neurological events, mortality, duration of hospitalization, and health care costs. In RCTs with multiple interventions or treatment arms or in studies with factorial designs, data were extracted, when possible, for those groups comparing amiodarone and placebo only. Authors of the studies were contacted for additional information when applicable.
Assessment of Methodological Quality
Two reviewers (S.M.B. and P.D.G.) independently assessed methodologic quality. Items used to assess study quality were methods of randomization, allocation concealment, any blinding, use of a placebo, reporting of losses to follow-up or missing outcome assessments, evidence of important baseline differences, and use of a sample size calculation [17]. An overall quality score was determined for each study [18].
Statistical Methods
The rates for AF, ventricular tachyarrhythmias, strokes, and death from available RCTs were treated as dichotomous variables and combined to estimate the pooled odds ratio (OR) with 95% confidence intervals (CIs) using weighted fixed-effects or random-effects models when appropriate [19]. The presence of heterogeneity across studies was evaluated using the 
2 test for homogeneity and the I
2 statistic, with a I
2 value greater than 50% indicating at least moderate statistical heterogeneity [20]. Because the 2 statistic has a low sensitivity for detecting heterogeneity, a p value of 0.1 or less was considered significant for the presence of statistical heterogeneity [21]. Ventricular response rate, time to onset and duration of AF, duration of hospitalization, and health care costs were treated as continuous variables. The summary effects of these variables were calculated as the weighted mean difference. A weighted meta-regression and sensitivity analysis were performed to assess the effects of selected study quality and clinical factors on treatment effects. These a priori selected subgroups included the following: sample size of RCT, evidence of baseline differences, loss to follow-up or missed outcome assessment, type of cardiac surgery, elective or emergency surgery, timing and route of amiodarone administration, use of perioperative ß-blockers, calcium-channel blockers, or digoxin, age, and selected comorbidities. Publication bias was assessed for by use of an Egger's test and a visual assessment of a funnel plot [22, 23]. Statistical analyses were performed using Stata version 8.2 (StataCorp, College Station, Texas).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
= 0.90 ± 0.16), with all disagreements resolved by discussion.
|
|
Atrial Fibrillation After Cardiac Surgery
The definitions and methods for detection of AF were variable across studies. The incidence of AF in the amiodarone and control groups ranged between 5% and 35% and 21% and 53%, respectively (Table 2). The pooled OR for development of AF after cardiac surgery was 0.50 (95% CI: 0.43 to 0.59, p < 0.0001), suggesting a significant reduction in AF with use of amiodarone (test for heterogeneity 2 = 21.5, p = 0.26; I
2 = 16%; Fig 2). A cumulative meta-analysis including all 19 RCTs is presented in Figure 3. This analysis reveals that the PAPABEAR trial, the largest RCT assessing amiodarone, produced results that are consistent with the pooled results of preceding studies.
|
|
|
2 = 4.1, p = 0.9; Fig 4A) [14, 26–28, 30–35, 38]. Atrial fibrillation occurred an average of 6.7 hours later (95% CI: 0.7 to 12.6, p = 0.03) and for a duration 2.9 hours (95% CI: 1.2 to 4.6, p = 0.001) shorter in those receiving amiodarone. However, both of these pooled estimates had evidence of statistical heterogeneity (2 = 105, p < 0.001, and 2 = 365, p < 0.001) and require cautious interpretation.
|
|
2 = 14.4, p = 0.2; Fig 4B) [14, 25, 27–30, 32, 34, 35, 37, 38, 40].
Ventricular tachyarrhythmias occurred in 4.1% of patients (n = 121 of 2,928) [14, 24–28, 30, 32, 34–36, 38–40, 42]. The pooled OR for development of ventricular tachyarrhythmias was 0.39 (95% CI: 0.26 to 0.58, p < 0.0001), suggesting a reduction with amiodarone compared with control (heterogeneity 2 = 8.6, p = 0.8; I
2 = 0%; Fig 4C).
Neurologic events (transient ischemic attack or stroke) in the postoperative period were uncommon and occurred in 2% of patients (n = 50 of 2,562) [14, 25, 26, 28, 30, 32, 34–38, 40]. The pooled OR for postoperative neurologic events was 0.53 (95% CI: 0.30 to 0.92, p = 0.02), suggesting a reduction in patients assigned to amiodarone (2 = 5.9, p = 0.8; I
2 = 0%; Fig 4D). The overall operative mortality rate was low (3.0%; n = 82 of 2,775) and not significantly different for patients receiving amiodarone compared with control (OR = 0.96; 95% CI: 0.6 to 1.5, p = 0.8; heterogeneity 2 = 45.5, p = 0.9, I
2 = 0%) [14, 25–32, 34, 37, 38, 42].
Costs were reported in six studies; however, data could only be pooled from four studies [26–28, 37]. The mean (± SD) total cost was $18,548 (± $2,624) and $21,637 (± $4,744) for patients in the amiodarone and control groups, respectively. Amiodarone resulted in cost savings when compared with control using a conservative random-effects model ($2,527; 95% CI: -$500 to $5,815, p = 0.1) of marginal statistical significance (heterogeneity 2 = 9.5, p = 0.02).
Meta-Regression and Sensitivity Analysis for Atrial Fibrillation After Cardiac Surgery
There was no influence on the pooled estimate by any study quality factors including assessment of baseline differences ([coefficient] -0.02; 95% CI: -0.44 to 0.39, p = 0.9) and loss to follow-up or missed outcome assessment (0.15; 95% CI: -0.3 to 0.6, p = 0.5). Furthermore, no influence was seen based on several clinical factors including age (0.02; 95% CI: -0.02 to 0.7, p = 0.3), diabetes mellitus (0.002; 95% CI: -0.003 to 0.007, p = 0.5), congestive heart failure (-0.002; 95% CI: -0.007 to 0.002, p = 0.3), prior AF (-0.006; 95% CI: -0.4 to 0.3, p = 0.7), or the use of ß-blockers (0.003; 95% CI: -0.01 to 0.002, p = 0.7), calcium-channel blockers (0.005; 95% CI: -0.003 to 0.1, p = 0.1), digoxin (-0.1; 95% CI: -0.5 to 0.3, p = 0.6), or myocardial infarction (0.004; 95% CI: 0.001 to 0.008, p = 0.1). Further, the pooled effect estimate for the reduction in AF was similar across several stratified variables in a sensitivity analysis (Table 4).
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Recently, Aasbo and associates [7] suggested that while amiodarone decreases AF as well as additional cardiovascular complications, routine therapy should not be considered as yet; these researchers recommended an additional RCT to precisely define the impact of concomitant use of ß-blockers. That conclusion would support current consensus guidelines recommending that use of amiodarone be reserved for patients with contraindications to ß-blockers or with risk factors for AF such as valvular disease, left atrial enlargement, or supraventricular tachyarrhythmias [43]. However, similar to the findings by Aasbo and colleagues [7], only 48% of patients in our study were prescribed ß-blockers. Further, in one study, the majority of the study population had no prior ß-blocker therapy prescribed [40].
Thus, it would appear, despite perioperative ß-blockers being associated with a reduced risk of AF after cardiac surgery, ß-blockers are inconsistently prescribed. That has propagated further questions regarding the necessity and independent effect of prophylactic amiodarone [7, 11, 13, 42]. However, it is increasingly clear that current evidence as presented in our meta-analysis provides a compelling argument to the contrary. First, our meta-regression analysis of concomitant ß-blocker therapy failed to demonstrate any influence on the overall effect estimate for prevention of AF. Second, the PAPABEAR study a priori stratified patients on the basis of preoperative ß-blocker therapy [14]. This stratified analysis confirmed that in those randomly assigned to amiodarone, there was no meaningful difference in the incidence of AF regardless of preoperative ß-blocker use (15% on ß-blockers versus 16% not on ß-blockers). However, amiodarone added significant benefit relative to ß-blocker therapy alone (15% for amiodarone plus ß-blocker versus 25% for ß-blockers alone, p = 0.03). Similarly, despite a high rate (greater than 80%) of ß-blocker use, two large RCTs reported significant benefit with addition of amiodarone [32, 37]. Finally, the study by Auer and colleagues [40] showed added benefit for reducing AF with amiodarone plus ß-blocker compared with ß-blocker therapy alone [40].
Our meta-analysis also incorporated several secondary observations that add value to and strengthen the findings of prior reviews. First, in those patients that ultimately developed an episode of AF, amiodarone was associated with a lower ventricular response rate. Second, amiodarone reduced the occurrence of postoperative ventricular tachyarrhythmias. New-onset ventricular tachyarrhythmias after cardiac surgery are uncommon; however, their occurrence is associated with considerable morbidity and mortality [44]. Third, amiodarone was associated with a decreased risk of transient ischemic attack or stroke. Although the reported incidence was low, stroke is a potentially devastating complication [45]. Several independent risks for stroke after cardiac surgery, including postoperative AF, have been suggested [45]. Our data would suggest that amiodarone reduces the incidence of transient ischemic attack or stroke in a large heterogeneous patient population undergoing cardiac surgery. Fourth, this meta-analysis has shown that prophylactic amiodarone reduces the length of hospitalization by 0.6 days. Although this reduction is small, it represents considerable health resource savings when considering the number of cardiac surgical procedures performed annually. Furthermore, this would be supported by a trend towards cost savings with amiodarone. Fifth, our study failed to reveal a significant impact on mortality with amiodarone. This finding contrasts with large epidemiologic studies suggesting that AF results in greater mortality [1, 46]. However, there may be limitations to consider before concluding no difference in mortality with use of amiodarone from our study. Specifically, the pooled analysis was based on data from only 15 studies; no study incorporated or was powered to assess long-term mortality, and several studies may have limited generalizability.
Finally, our sensitivity analysis has suggested that the pooled effect estimate for reduction in AF with amiodarone is robust. Despite variability in the timing and route of amiodarone administration, there did not appear to be any significant impact on the pooled effect estimate. This finding suggests that, within the dosage ranges used in these studies, any protocol for administration of amiodarone that minimizes the total dosage and duration of therapy may be efficacious. This observation is potentially important when considering the need for abridged protocols for those patients requiring emergent surgery. In addition, similar to one large RCT, our analysis confirmed that amiodarone was equally efficacious for reduction in AF regardless of whether patients were undergoing either bypass alone or bypass combined with valvular surgery [14].
Before the routine implementation of amiodarone for AF prevention, the potential adverse events of therapy must be considered. While there is accumulating evidence of relative safety and patient tolerance [47, 48], case series and cohort studies have suggested an increased risk of pulmonary toxicity and postoperative acute respiratory distress syndrome in patients receiving amiodarone [49–51]. However, two recent studies have failed to show an association between either short-term or low-dose preoperative amiodarone and the development of acute respiratory distress syndrome [41, 49].
Unfortunately, few trials included in our review provided outcome data for acute or chronic adverse effects attributable to amiodarone. Acute adverse effects were variably described in only 13 studies [14, 24–26, 29, 31, 32, 34, 35, 37, 38, 40, 41]. The rates of postoperative nausea, bradycardia, hypotension, QT-interval prolongation, and need for temporary pacing were most commonly reported. Butler and coworkers [25] reported a higher incidence of bradycardia with amiodarone (78% versus 48%, p < 0.005). Similarly, one study found that bradycardia (rate less than 60 beats per minute) with need for temporary pacing was more common with amiodarone (48% versus 28%, p < 0.05) [29]. In contrast, however, other studies failed to corroborate these findings [24, 32, 34, 35, 37, 38]. Importantly, the incidence of postoperative insertion of a permanent pacemaker was reported in only two studies, with a low overall rate (less than 2%) and no difference for those receiving amiodarone [26, 32]. Treggiari-Venzi and colleagues [31] suggested amiodarone resulted in greater cardiovascular instability, prompting increased need and duration of vasoactive support. This trial was terminated early, however, and similar findings from larger studies have not been described [14, 24, 32, 37]. A recent systematic review focusing on side effects of amiodarone suggested hemodynamic compromise (bradycardia and hypotension) were more common when amiodarone was loaded intravenously in the postoperative period and exceeding 1,000 mg daily [48].
There are several limitations to our study. The primary outcome of AF was variably defined across studies which could contribute to the disparity in events rates and influence the overall pooled estimate. Several of our secondary outcomes were also secondary outcomes in the RCTs and were variably reported, thus potentially limiting the inferences from our analysis. In addition, death after cardiac surgery was uncommon and inconsistently reported. Despite significant reductions in ventricular tachyarrhythmias and stroke with the use of prophylactic amiodarone, the possibility that amiodarone therapy may reduce operative mortality remains unproven.
Based on this study, we believe there is a convincing body of evidence to suggest that amiodarone should be considered as a first-line therapy and as routine prophylaxis for AF prevention after cardiac surgery. Our findings support this statement regardless of the presence of high-risk features for AF or the concomitant use of ß-blockers. At present, the most efficacious protocol for amiodarone administration perioperatively remains unclear; however, our findings suggest that a total dose in the range of 5 to 10 g is likely sufficient, independent of timing of initiation (preoperative, intraoperative, or postoperative) or route of administration (oral, intravenous, or combined). Likewise, the scheduled duration of therapy should be short in the postoperative period (less than 2 weeks). Therefore, for patients with planned elective surgery, amiodarone loading can begin preoperatively whereas for those undergoing emergency surgery, amiodarone could be initiated in the postoperative period. Finally, while we recommend use of amiodarone in routine clinical practice for patients without contraindications, we suggest that further surveillance for adverse effects is warranted.
In summary, amiodarone prophylaxis is associated with a significant decrease in the incidence of AF after cardiac surgery. Moreover, prophylactic amiodarone therapy is associated with fewer perioperative ventricular tachyarrhythmias and strokes and with a significant but modest reduction in hospital length of stay. Further, if AF occurred, amiodarone-treated patients had a significantly lower ventricular response rate. There is now a substantial body of evidence suggesting that prophylactic amiodarone should be implemented as routine therapy for patients undergoing cardiac surgery.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Roshanali, M. H. Mandegar, M. A. Yousefnia, F. Alaeddini, and B. Saidi Prevention of atrial fibrillation after coronary artery bypass grafting via atrial electromechanical interval and use of amiodarone prophylaxis Interactive CardioVascular and Thoracic Surgery, April 1, 2009; 8(4): 421 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Alsaddique, A. G. Royse, C. F. Royse, and M. A. Fouda Management of diastolic heart failure following cardiac surgery Eur. J. Cardiothorac. Surg., February 1, 2009; 35(2): 241 - 249. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Shrivastava, B. Smith, D. Caskey, and P. Reddy Atrial Fibrillation After Cardiac Surgery: Does Prophylactic Therapy Decrease Adverse Outcomes Associated With Atrial Fibrillation J Intensive Care Med, January 1, 2009; 24(1): 18 - 25. [Abstract] [PDF]
|
|
|
|
 |
|
 K. Lertsburapa, C. M. White, J. Kluger, O. Faheem, J. Hammond, and C. I. Coleman Preoperative statins for the prevention of atrial fibrillation after cardiothoracic surgery. J. Thorac. Cardiovasc. Surg., February 1, 2008; 135(2): 405 - 411. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. R. Zebis, T. D. Christensen, I. S. Kristiansen, and V. E. Hjortdal Amiodarone Cost Effectiveness in Preventing Atrial Fibrillation After Coronary Artery Bypass Graft Surgery Ann. Thorac. Surg., January 1, 2008; 85(1): 28 - 32. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. P. Mason, D. H. Marsh, J. M. Alster, S. C. Murthy, A. M. McNeill, M. M. Budev, A. C. Mehta, G. B. Pettersson, and E. H. Blackstone Atrial Fibrillation After Lung Transplantation: Timing, Risk Factors, and Treatment Ann. Thorac. Surg., December 1, 2007; 84(6): 1878 - 1884. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. F. Immer, A. Ackermann, E. Gygax, M. Stalder, L. Englberger, F. S. Eckstein, H. T. Tevaearai, J. Schmidli, and T. P. Carrel Minimal Extracorporeal Circulation is a Promising Technique for Coronary Artery Bypass Grafting Ann. Thorac. Surg., November 1, 2007; 84(5): 1515 - 1521. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. V. Conte Reducing the Incidence of Atrial Fibrillation Arch Surg, September 1, 2007; 142(9): 821 - 821. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
