| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;83:1326-1331
© 2007 The Society of Thoracic Surgeons
a Department of Cardiothoracic and Vascular Surgery & Institute of Clinical Medicine, Skejby Sygehus, Aarhus University Hospital, Aarhus N, Denmark
c Department of Intensive Care, Skejby Sygehus, Aarhus University Hospital, Aarhus N, Denmark
b Department of Clinical Epidemiology & Institute of Clinical Medicine, Aarhus University Hospital, Aarhus C, Denmark
Accepted for publication September 19, 2006.
* Address correspondence to Dr Zebis, Department of Cardiothoracic and Vascular Surgery & Institute of Clinical Medicine, Skejby Sygehus, Aarhus University Hospital, Brendstrupgaardsvej 100, DK-8200 Aarhus N, Denmark (Email: lrz{at}post.tele.dk).
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentReferences |
|---|
Methods: This randomized, controlled, double-blinded trial included 250 eligible consecutively enrolled patients undergoing coronary artery bypass grafting (CABG). They received 300 mg of amiodarone/placebo administered intravenously over 20 minutes on the first postoperative day and an oral dose of 600 mg of amiodarone or placebo twice daily for the first 5 postoperative days.
Results: The patients in amiodarone prophylaxis experienced a reduction in risk of atrial fibrillation of 14% (95% confidence interval [CI], 5.0% to 24%), with the number needed to treat at 6.9 (95% CI, 4.2 to 20), and the results for symptomatic atrial fibrillation showed a risk reduction of 18% (95% CI, 9.4% to 26), with the number needed to treat at 5.7 (95% CI, 3.9 to 11). Of the patients who developed atrial fibrillation in the placebo group, 84% experienced a symptomatic attack versus only 43% in the amiodarone group.
Conclusions: Postoperative prophylaxis with a high dose of oral amiodarone after an intravenous bolus infusion is a safe, practical, feasible, and effective regimen for CABG patients. It significantly diminishes the occurrence of postoperative atrial fibrillation.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
Although postoperative atrial fibrillation is often regarded as a temporary, benign, operation-related problem, it is associated with a twofold to threefold increased risk of adverse events [3, 4], including transient or permanent stroke, acute myocardial infarction, and death. A twofold increase in length of stay in the intensive care unit and prolongation of total hospital stay—with increased costs—have also been reported [5, 6].
The observed incidence of atrial fibrillation after coronary artery bypass grafting (CABG) has increased during the last 20 years. More advanced patient age, increased complexity of cardiac surgery, and better detection of AF may explain the increase [4, 7].
When atrial fibrillation occurs, ß-blockers, calcium-channel blockers, digoxin, magnesium, sotalol, quinidine, and amiodarone are used to control heart rate and restore sinus rhythm. Most of these antiarrhythmic agents have adverse effects [8]. The effectiveness of amiodarone is well documented for preventing atrial fibrillation and reducing its incidence after CABG operations [9–13]. However, regimens for administering this drug—intravenously preoperatively or postoperatively, or both [7, 11, 14] or orally, or both [9, 13]—as described in published studies, are difficult to implement in most centers owing to the need for an extended hospital stay or long preoperative treatment.
In most clinical settings, most patients are hospitalized the day before surgery. Postoperatively, they follow a fast-track schedule, with a few hours in the intensive care unit, early mobilization, and discharge within 5 to 6 days. This schedule precludes prolonged preoperative treatment, and intravenous infusion after the first postoperative day interferes with the mobilization program.
A solution to these logistic problems is the use of high-dose amiodarone postoperatively for a few days after an intravenous loading bolus. It is unknown, however, whether this regimen can achieve a prophylactic effect with minimal adverse effects [11]. In our study, we compared the risk of postoperative atrial fibrillation after CABG in two groups of patients: one receiving this easily administered regimen of high-dose prophylactic amiodarone and the other, placebo. We hypothesized that the risk of postoperative atrial fibrillation is reduced for patients receiving this prophylactic regimen. We further hypothesized that it is as efficient as previously reported ones [9–12].
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
The protocol for the study complied with the Helsinki II Declaration and was approved by the regional scientific ethical committee. All patients consented to participate in the study. The study was conducted according to the Good Clinical Practice standards and was monitored and approved by the Good Clinical Practice unit at Aarhus University Hospital.
Methods
This study was performed as a randomized, controlled, double-blinded trial. In the beginning of the study, one patient was drawn randomly each day to be included in the study. After 6 months, we shifted the regimen to include everybody scheduled for CABG. Each enrolled patient was randomly assigned to receive amiodarone or placebo according to a computerized prospective randomization schedule. Randomization was done in blocks of four and was stratified by age (
65 years and >65 years) and by preoperative use of ß-blockers.
Each patient received a randomization number, which was recorded and sent by fax to the pharmacy at Aarhus University Hospital, along with the patient’s Civil Registration number. The pharmacy decoded the randomization number, prepared the appropriate infusion and pills, and forwarded them, together with a sealed opaque envelope containing the randomization assignment, to the Department of Cardiothoracic and Vascular Surgery.
Intervention
At 8 AM on the morning after surgery, study participants, who were in the intensive care unit, received a bolus infusion containing 300 mg of amiodarone or placebo (5% aqueous dextrose solution) over 20 minutes, together with their first 600-mg dose of oral amiodarone or placebo. The oral dose of 600 mg amiodarone or placebo was administered twice a day (8 AM and 8 PM) for 5 days, with three 200-mg amiodarone tablets given in one capsule. Patients in the active treatment group received a total of 6.3 g amiodarone.
For patients unable to swallow the capsules, a nurse from another department opened them and dissolved the tablets in water. The study drug was terminated if systolic blood pressure dropped below 80mm Hg or if the heart rate dropped below 40 beats/min.
If possible, patients discontinued anticoagulation treatment 5 days preoperatively and continued the rest of their regular medication, including ß-blockers, until the day of surgery. Anticoagulation therapy and additional medications were resumed in the intensive care unit on the first postoperative day. The use of ß-blockers was resumed in patients who were receiving them preoperatively.
During surgery and up until the discharge from the intensive care unit, patients remained under continuous scope monitoring. After transfer to the ward, the attending physician performed standard 12-lead electrocardiograms (ECGs) daily. This monitoring was continued for the first 5 postoperative days. Additional ECGs were obtained in the presence of clinical signs (eg, irregular pulse, fainting, confusion, shortness of breath) or subjective manifestations (eg, palpitations, dizziness, fatigue, light-headedness) of atrial fibrillation. If atrial fibrillation developed, the study drug was discontinued and appropriate treatment was initiated.
On postoperative day 5, patients were transferred to local hospitals, which subsequently provided discharge summaries for the study. On postoperative day 30, patients were contacted by a study physician and asked whether clinical atrial fibrillation had developed. During the same call, patients were asked to contact their physician to schedule an ECG. Study staff obtained the ECG results. The observation period ended earlier if the patient reached a study end point.
All operations were performed through a median sternotomy. A standard cardiopulmonary bypass technique was established by cannulating the ascending aorta and the right atrium with a two-stage catheter. The aorta was cross-clamped, and for myocardial protection, cold crystalloid cardioplegia (St. Thomas Solution) was delivered antegrade at 20-minute intervals throughout the procedure. Noncoated silicone tubes and an oxygenator (QUADROX, Jostra, Hirrlingen, Germany) with an albumin-based coating (Safeline, Jostra) were used. Conditions of normothermia or moderate hypothermia (34°C) were established.
Bypass conduits included the internal thoracic/mammary artery, the radial artery, and the saphenous vein. The left internal mammary artery was anastomosed to the left anterior descending in most patients. Distal anastomoses were performed during a single period of aortic cross-clamping, and proximal anastomoses were performed with partial aortic clamping during rewarming.
End Points
The primary composite end point in the protocol was time to symptomatic atrial fibrillation, a permanent or transient stroke, acute myocardial infarction (AMI), or death. The secondary end points were time to symptomatic or asymptomatic atrial fibrillation, stroke, AMI, or death; time to symptomatic or asymptomatic atrial fibrillation; and time to symptomatic atrial fibrillation.
After collecting the data, we found that transient stroke developed in 2 patients, AMI developed in 1, and no one died. We therefore decided to narrow our end points exclusively to atrial fibrillation and symptomatic atrial fibrillation. For these two end points, we calculated the risk difference and number needed to treat, with 95% confidence intervals. In addition, the frequency of possible drug-related side effects (ie, bradycardia or hypotension) before postoperative day 30 was examined. Patients were evaluated for occurrence of end points 7 days and 30 days postoperatively.
Atrial Fibrillation
Atrial fibrillation was defined as fast, irregular, eddy current activation of the atrium with neutralization of its contractions. The ECG shows totally irregular ventricle rhythm with a constant irregular fluctuation of the length between the QRS-complexes. The QRS-complexes often appear with a frequency of about 150 to 200 beats/min, displaying a line flickering with narrow QRS-complexes.
Symptomatic atrial fibrillation was defined as the patient’s subjective discomfort, including palpitations, angina pectoris, shortness of breath, or fainting, or the need for acute intervention owing to hypotension or heart failure. An ECG was used to verify the diagnosis. Asymptomatic atrial fibrillation was defined as atrial fibrillation detected on the daily monitoring ECG or the discovery of an asymptomatic patient with an irregular pulse during ward rounds. Only asymptomatic atrial fibrillation lasting more than 30 minutes was counted as an event.
Possible contributing factors to atrial fibrillation development, such as pleura effusion, electrolyte derangement, fever, and volume overload, were examined and corrected. Medical treatment followed the routine department’s guidelines:
After transfer to the local hospital and subsequent discharge, treatment adhered to the guidelines of the department in charge of the patient.
Statistical Analysis and Sample Size
Baseline data and perioperative characteristics were compared using the Wilcoxon-Mann-Whitney rank sum test for nonparametric variables. The Pearson’s 
2 test was used to compare proportions.
We conducted intention-to-treat analyses of primary and secondary end points in the two groups. Kaplan-Meier curves were used to display survival until the end points, and Cox’s proportional hazards regression was used to estimate relative risks and to calculate p values and 95% confidence intervals. The proportional-hazards assumption was evaluated using Schoenfeld residuals. The 7-day and 30-day incidence rates, risk differences and number needed to treat were calculated with corresponding 95% confidence intervals The data were analyzed using Stata 9 software (StataCorp, TX).
Type I error was set at 0.05 and type II at 0.20, allowing for 80% power. The a priori hypothesis—that atrial fibrillation risk would be lowered from 25% in the placebo group to 10% in the treatment group—called for 96 patients in each group. The estimated dropout rate was 20% (48 patients). Thus, enrollment of 250 consecutive patients was needed. Ten participants from an earlier pilot study were enrolled in the full study.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
|
Most variables were equally distributed in the two groups. Exceptions were lesion location in the right coronary artery, which was overrepresented in the amiodarone group (p = 0.024), and the median number of grafted vessels (3.2 in the amiodarone group versus 2.9 in the placebo group; p = 0.008). Patients not receiving ß-blocker treatment had severe chronic obstructive pulmonary disease or they had atrioventricular block.
|
|
|
|
Atrial fibrillation developed in 14 patients in the amiodarone group and in 32 patients in the placebo group (Fig 3; Table 4). Of the patients with atrial fibrillation, 43% had symptomatic atrial fibrillation in the amiodarone group versus 84% in the placebo group. There was no statistically significant difference between the two groups with respect to outcomes labeled "time of occurrence after surgery" or "atrial fibrillation duration."
|
|
Two patients died during the trial, one after an AMI and one owing to a complex cancer. Transient stroke developed in 2 patients. The distribution among the treatment arms was equal: one death and one transient stroke occurred in both groups.
Three patients experienced side effects that led to discontinuation of study medication. Bradycardia requiring short-term pacemaker treatment developed in 2 patients in the intervention group, and hypotension developed in 1 patient in the placebo group.
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
Other regimens have shown different outcomes in preventing postoperative atrial fibrillation. The use of prophylactic amiodarone has been well documented in the literature, with postoperative atrial fibrillation occurring in as few as 5% [12, 15] and in as many as 35% [16] of patients. Several studies [17–21] yielded results ranging from no effect to effects comparable with those reported here. This variability stems from the heterogeneity of patient populations, the route of drug administration, the duration of drug therapy, the presence of ß-blocker therapy, the atrial fibrillation definition used, and other factors.
We had an empirical rationale for focusing on patients with clinically symptomatic atrial fibrillation, because the group of patients experiencing subjective discomfort is most prone to atrial fibrillation-related adverse events [22]. Postoperative atrial fibrillation occurred on the second to third postoperative day, which is the time when the blood concentration of amiodarone is maximal. Our regimen presented a clinically feasible way to significantly reduce the risk of atrial fibrillation after CABG, achieving a reduction similar to that of other regimens of longer duration and higher cost.
We found that when atrial fibrillation developed in patients despite amiodarone treatment, they were less likely to have symptomatic atrial fibrillation. This indicates that even if amiodarone failed to prevent atrial fibrillation at the dose provided in our trial, it lessened the symptoms. Thus, most affected patients in our study (57%) experienced only a mild attack. Postoperative time to atrial fibrillation and its duration were similar in both treatment arms. The duration of atrial fibrillation is probably more influenced by the attending physician’s urge to intervene than by the medication. One patient in the amiodarone group experienced atrial fibrillation 3.5 hours after surgery, which was before initiation of study treatment.
Serious treatment-related side effects were not encountered. However, our study of 250 patients was too small to permit rare side effects to be observed.
Limitations
The strengths of our study include its randomized controlled design, blinding, and the feasibility of drug administration within the routine of a normal ward setting. The regimen did not interfere with routine admission and discharge schedules and was thus easy to implement. It also did not require that changes to be made in preoperative medications.
This study was designed as a clinical study, focusing on the development of atrial fibrillation as seen in our ward. The real estimate of patients developing atrial fibrillation might be higher than the proportion found in this study. The establishing of the atrial fibrillation diagnosis after hospital discharge required that the patients had symptomatic atrial fibrillation, for which they visited their physician. This could undermine the validity of our atrial fibrillation data. However, our study focused on clinically important atrial fibrillation, ascertained from medical discharge records and from a telephone interview with the treating physician at day 30. At this time as well as at discharge from the hospital, an ECG was performed in all patients and read by a study physician who was blinded to the treatment assignment. Thus, although we cannot be completely certain that we have captured all episodes of atrial fibrillation occurring in the late postoperative period, the risk of overlooking clinically important attacks is small. Furthermore, randomization makes it unlikely that risk underestimation would differ by treatment group.
Our 21% dropout rate could have been reduced to 4.4% if randomization had been performed after surgery. Two potential confounders were the number of grafted vessels and the number of grafted right coronary arteries, both of which were more frequent in the amiodarone group. Both these factors increase the likelihood of developing atrial fibrillation; however, because this imbalance affected the intervention group, it is unlikely to undermine our study findings.
Ina Hoegh, the project nurse, participated in enrolling and randomizing patients and in conducting follow-up. The study was funded by the following private sources: "Statens sundhedsvidenskabelige forskningsfond," "Skejby fonden," "Elin Holms forskningsfond," "Lily Benthine Lunds Fond," "Lykfeldts Legat," "Snedkermester Sophus Jacobsen og hustru Astrid Jacobsens Fond" and "Kirsten Anthonius Mindelegat."
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
