Introduction
Introduction
Atrial fibrillation (AF) is one of the most common complications after coronary artery
bypass grafting (CABG). However atrial fibrillation is not merely a simple rhythm
disorder. AF reduces left ventricular function due to the lack of active diastolic
left ventricular filling and impaired contractility caused by tachyarrhythmia phases.
Additionally, postoperative AF prolongs the patient's stay in the critical care unit,
which is associated with clinical complications and increased costs.
Many studies have been undertaken to develop pharmacological or other antiarrhythmic
therapies or algorithms for the prevention of postoperative AF. One of the most effective
drugs in the prevention of postoperative AF is amiodarone®. However this therapy is
associated with a high rate of drug-related complications such as hyperthyroidism,
pulmonary fibrosis, and a significant reduction in left ventricular inotropy. Thus,
the prophylactic administration of amiodarone® has to be considered as controversial.
Several clinical and experimental studies showed n-3 polyunsaturated fatty acids (PUFA)
to be effective for the prevention of atrial fibrillation. PUFA have shown significant
antiarrhythmic effects on left atrial tissue in rat experiments. The consumption of
fish, resulting in higher plasma PUFA concentrations, has been associated with a lower
incidence of AF over a 12-year follow-up. A recent study showed the beneficial effect
of preoperative oral PUFA on the postoperative occurrence of atrial fibrillation after
coronary arterial bypass surgery. Experimental studies using atrial rat cardiomyocytes
showed that PUFA prevent the induction of fibrillation. In a dog model, PUFA suppressed
the occurrence of ventricular tachycardia (VT) after infarction.
There are several theories explaining these effects. PUFA provide effective membrane
stabilization in the myocardial cell by prolonged inactivation of the fast sodium
outward channel, resulting in a longer refractory time. Another mechanism is the modulation
of calcium release, probably through interaction with the sarcoplasmic reticulum.
This is hypothesized to be caused by a direct interaction of PUFA with arachidonic
acid metabolism.
These interactions seem to be responsible for the reduction of major cardiac events,
including sudden cardiac death, ventricular fibrillation, and AF, after the use of
PUFA.
Aim of the study
The aim of this study was to assess the efficacy and safety of intravenously administered
n-3 PUFA in preventing the occurrence of atrial fibrillation after CABG.
Methods
Methods
Study design
The study was designed as a prospective randomized double-blinded trial. The protocol
was approved by the ethics committee of our university (file reference 74/04). This
study was not supported by a grant, nor was it funded by any pharmaceutical company.
Patients were randomized to the control group and to the interventional group. In
addition to usual peri- and postoperative therapy, i. v. saturated free fatty acids
were given to the control group and PUFA to the interventional group. The infusion
pump was started at least 12 hours before CABG surgery and continued until transfer
from the ICU to a normal ward.
Patients
The study cohort consisted of 102 patients (70 men, 32 women, mean age 67 ± 9.3 years).
The consecutive patients were admitted to the department of cardiovascular surgery
from December 2005 to October 2006 for elective CABG surgery. Inclusion criteria were
age of more than 18 years, normal sinus rhythm, stable hemodynamic conditions and
freedom from angina at rest. Exclusion criteria were concomitant valve surgery, prior
history of supraventricular arrhythmias, and current antiarrhythmic therapy other
than beta-blockers and calcium channel antagonists. Informed consent had to be given
by all patients.
PUFA therapy
52 patients were randomized to the interventional group, 50 served as the control
group. In the control group, free fatty acids (100 mg soya oil/kg body weight/day)
were infused via a perfusion pump from the time of admission to hospital until transfer
to a normal ward. In the interventional group, PUFA were given at a dosage of 100 mg
fish oil/kg body weight/day. Free fatty acids and PUFA were purchased from Fresenius
(Bad Homburg, Germany). Lipovenös® 10 % served as a free fatty acid in the control
group. 100 ml Lipovenös® contains 10 g soya oil. The dosage was 1 ml/kg body weight/day;
this is equivalent to 100 mg soya oil/kg body weight/day. Omegaven® served as PUFA
in the interventional group. 100 ml Omegaven® contains 10 mg fish oil. The dosage
was 1 ml/kg body weight/day, which is equivalent to 100 mg fish oil/kg body weight/day.
Data collection
After surgery, all patients were transferred to the intensive care unit (ICU). Continuous
rhythm monitoring was performed until transfer to a normal ward. Standard 12-lead
ECG was performed daily from admission to hospital until transfer. After admission,
all patients were examined by transthoracic echocardiography. This method yielded
parameters for left ventricular ejection fraction, left ventricular diastolic diameter,
and left atrial diameter.
Cardiopulmonary bypass- and cross-clamping times were documented for each patient.
Primary end point of this study was the development of postoperative atrial fibrillation
as detected by monitoring or 12-lead ECG during the ICU period. AF was defined as
any confirmed episode of AF for longer than 15 minutes. After occurrence of postoperative
AF, participation in the study ended for these patients.
The secondary end point was the length of stay in the ICU and in hospital.
Statistical analysis
The efficacy of intravenously administered PUFA for the prevention of postoperative
atrial fibrillation after CABG was investigated. Sample size calculation was based
on a 30 % occurrence of postoperative AF in the control group and a 20 % occurrence
in the PUFA group. A sequential method of testing was used. This method tests the
result of two different interventions on the probability of two alternative clinical
events. For the incidence of atrial fibrillation a probability of 25–30 % was assumed,
based on the literature. The expected probability of an effective therapy should not
be lower than 0.5. The above-described sequential test investigates the possibility
of a result after each step. The precondition is the independence of patient characteristics.
Preceding statistical analysis, the patients had to be randomized pairwise.
Differences in the length of stay in the ICU and in hospital were analyzed using unifactorial
variance analysis. Normal distribution of all investigated factors was proved after
transformation by log-normal transformation ([Fig. 1
Fig. 1 Total stay in hospital: total length of stay in the hospital shows a log-normal distribution.
Results
Results
The demographic, clinical and surgical characteristics of the patients in the two
groups were similar and showed no significant differences. There was no statistical
significant difference between the two groups “atrial fibrillation” and “sinus rhythm”
with regard to the parameters “body weight” and “age” ([Table 1
Table 1 Demographic data.
Sex
Body weight (kg)
Age (yrs)
p-value
Control group
male
32 = 64 %
88.75 ± 11.05
66.65 ± 10.65
n. s.
female
18 = 36 %
71.06 ± 8.84
70.72 ± 8.18
n. s.
Intervention group
male
38 = 73 %
83.47 ± 12.04
61.21 ± 14.12
n. s.
female
14 = 27 %
75.21 ± 16.92
74.42 ± 9.23
n. s.
Parallel to these findings, a comparison of the parameters “left atrial diameter”
and “left ventricular ejection fraction” did not show a significant difference between
the two groups ([Table 2
Table 2 Echocardiography data.
LA diameter (mm)
Ejection fraction ( %)
p value
Control group
40.57 ± 5.09
52.31 ± 15.68
n. s.
Intervention group
40.01 ± 5.14
52.00 ± 15.00
n. s.
Finally, there was no statistically significant difference between the groups with
regard to the parameters “cardiopulmonary bypass” and “cross-clamping”. In the group
with atrial fibrillation, the cross-clamping times and cardiopulmonary bypass times
tended to be longer.
Efficacy of PUFA in preventing postoperative atrial fibrillation: sequential testing
method
For the calculation of the acceptance and rejection line, we used the formula depicted
in the appendix. The two graphs ([Fig. 2 Fig. 3 Fig. 2 Fig. 3 Table 3
Fig. 2 This graph shows the results of the sequential testing method on the original data
of the interventional group. After the 18th test, the procedure can be stopped with
a probability of error of 0.01, since the acceptance line has been crossed.
Fig. 3 This graph shows the results of the sequential testing method on the original data
of the control group. There is no effect in the placebo group.
Table 3 Postoperative development of atrial fibrillation.
Day postop.
Control group
Intervention group
1
5 Pat.
4.9 %
3 Pat
2.9 %
2
11 Pat.
11.58 %
7 Pat.
7.37 %
3
11 Pat.
11.58 %
7 Pat.
7.37 %
Compared to patients in sinus rhythm, there was a trend for patients who converted
to atrial fibrillation to remain longer in the ICU ([Fig. 4 Fig. 5
Fig. 4 Total time of stay in the ICU in hours: the gray bar represents the patients with
atrial fibrillation; the white bar represents the patients in sinus rhythm.
Fig. 5 Total time of stay in the hospital in days: the gray bar represents the patients
with atrial fibrillation; the white bar represents those in sinus rhythm.
Discussion
Discussion
Experimental studies
A large number of experimental studies have shown an effect of polyunsaturated fatty
acids (PUFA) on the myocardium, isolated myocardial cells [1 2 3 4 5
Clinical studies
There are not only positive study results describing the potential of PUFA in preventing
atrial fibrillation. A large randomized cohort study could not show this effect [6
Antiarrhythmic effects of PUFA
In a prospective randomized blinded clinical study, the benefit of a PUFA-rich diet
in patients after myocardial infarction was examined. The patients in the interventional
group showed a significantly lower rate of morbidity, mortality and ventricular arrhythmias
[7
Two clinical trials showed a beneficial effect of a PUFA-enriched diet on mortality
and sudden cardiac death [8 9 10
Several population studies have shown a significantly positive correlation between
fish consumption, plasma levels of PUFA, and the rate of serious coronary-related
events [11 12 13 14 15
Some clinical trials report a beneficial effect of PUFA on chronic or acute inflammation
[16 17 18 19 20
Clinical recommendations concerning PUFA
Scientific statements published by the American Heart Association [21 22 8 15 23
Studies in postoperative AF
Calo et al. [24
The results of Calo's study are essentially congruent with those found in our analysis.
However, there is an important difference between these two studies. In the study
of Calo, PUFA were administered orally. There were no tests on the bioavailability
or plasma levels. To exclude the problem of fluctuating plasma levels of PUFA, we
decided to administer PUFA intravenously. The PUFA dosage was adapted to the patient's
body weight as described above.
Discussion of statistical methods
Our study was planned as an intention-to-treat analysis. For simple statistical tests,
the inclusion of more than 400 patients would have been necessary. The statistical
calculations were more difficult due to the lower number of patients enrolled. Many
concomitant parameters were documented but were unusable for the calculation of clinical
predictors. The statistical methods of our data analysis have been validated. Thus,
our results are directly comparable to those derived from other trials. Additionally,
to a great extent, our results are congruent with the findings of substantially comparable
clinical trials.
Conclusions
Conclusions
The results yielded from our study data have a high clinical relevance. Perioperative
atrial fibrillation not only causes life-threatening complications but is also responsible
for excessive costs in the clinical treatment of CABG patients. Due to the proven
beneficial effect on perioperative atrial fibrillation, PUFA are recommended for the
perioperative therapy of patients undergoing CABG. Our results suggest that further
investigations should follow, which would include more patients. Thus, subgroup analysis
would be available, aiding calculations of clinical predictors. Intravenous administration
is advisable, to prevent the problem of fluctuating bioavailability.
Statement of responsibility
The authors had full access to and take responsibility for the integrity of the data.
All authors have read and agree to the manuscript as written.
Appendix
Acknowledgments
Acknowledgments
We thank Mrs. Mary Kay Steen-Müller M. D., Univ. Illinois, for her contributions to
the revision of this manuscript as a native speaker.
Conflict of Interest
None
Statement of responsibility
The authors had full access to and take responsibility for the integrity of the data.
All authors have read and agree to the manuscript as written.
