Atrial fibrillation (AF) occurs frequently soon after coronary artery bypass grafting (CABG) and often results in increased mortality and morbidity, particularly in patients with heart failure. New-onset AF is also a common event in the early period after discharge from a cardiac surgery clinic. Current guidelines recommend β blockers as first-line medication for the prevention of AF after CABG. In this prospective study, we investigated the effectiveness of the highly selective β1 receptor antagonist bisoprolol compared to the less selective β blocker carvedilol in preventing postdischarge AF after CABG in patients with decreased left ventricular function. Three hundred twenty patients (231 men, 89 women, mean age 66 ± 10 years) with ejection fraction <40% who underwent CABG and were then referred to an in-hospital cardiac rehabilitation program were randomized to receive bisoprolol (n = 160) or carvedilol (n = 160) starting 4 to 5 days after surgery. Bisoprolol was started at 1.25 mg 1 time/day and carvedilol was started 3.125 mg 2 times/day. All patients underwent continuous telemetric electrocardiographic monitoring for 5 days after entry in the study and thereafter 2 times/day routinely up to hospital discharge. During follow-up, 23 patients (14.6%) in the bisoprolol group and 37 patients (23%) in the carvedilol group developed AF (relative risk 0.6, confidence interval 0.4 to 0.9, p = 0.032). Twenty-six percent of all AF episodes were asymptomatic. At the 4-week outpatient visit, those in the bisoprolol group showed a significantly greater decrease in heart rate, being in sinus rhythm or AF (−15.6 ± 3 vs −9.4 ± 3 beats/min, p = 0.021), whereas changes in systolic and diastolic blood pressures did not differ significantly. In conclusion, bisoprolol is more effective than carvedilol in decreasing the incidence of postdischarge AF after CABG in patients with decreased left ventricular function.
To date, many pharmacologic approaches to preoperative and early postoperative prophylaxis of atrial fibrillation (AF) have been attempted. Current guidelines suggest β blockers as first-line medication for the prevention of AF after coronary artery bypass grafting (CABG). Although previous studies have employed nonselective β blockers, more recent studies have examined the use of more cardioselective β blockers for prevention of post-operative atrial fibrillation (POAF). Cardioselective β blockers carry the advantage of greater tolerability and compliance and are preferred in patients with left ventricular (LV) dysfunction. Carvedilol and bisoprolol are more widely used in clinical practice in patients with LV dysfunction and feature similar clinical efficacy. Recently, the 2 drugs were reported to be effective in preventing POAF in patients undergoing cardiac (i.e., valvular and CABG) surgery. No study has specifically compared the relative effectiveness of carvedilol and bisoprolol in preventing AF in patients with decreased LV function who had undergone CABG and are subsequently referred to postdischarge cardiac rehabilitation.
Methods
The study population included 320 patients with coronary artery disease, LV systolic dysfunction, and recent CABG who were referred consecutively to our cardiac rehabilitation division, Istituto di Ricovero e Cura a Carattere Scientifico, where they stayed for an expected period of 4 weeks, from January 2004 to December 2006.
Subject eligibility was determined at the initial screening visit. Patients were included in the study if they had a recent (≤5 days) CABG, LV ejection fraction <40%, sinus rhythm, and no AF episode after CABG. Patients were excluded if they had a history of asthma or severe chronic obstructive pulmonary disease, severe kidney diseases, contraindications to β blockers, sick sinus syndrome, heart rate <60 beats/min, systolic blood pressure <90 mm Hg, and concomitant treatment with class I or III antiarrhythmic drugs. Patients with documented episodes of AF occurring within 4 months before surgery were also ineligible for this study. All patients gave written informed consent to participate in the study, which was approved by the institutional review board.
After baseline screening on hospital admission (4 to 5 days after CABG), patients were randomly allocated on a 1:1 basis to receive oral bisoprolol (group I, n = 160) or carvedilol (group II, n = 160). Drugs were dispensed by the pharmacologic department that was also responsible for randomization codes. Drugs were administered by a nurse and investigators and patients were blinded.
At enrollment, patients received the first drug administration (bisoprolol 1.25 mg 1 time/day or carvedilol 3.125 mg 2 times/day). The dose of each drug was subsequently increased, according to standard practice and clinical judgment, until the maximum tolerated dose was achieved. Standard therapies were not restricted.
To detect POAF episodes, all patients underwent continuous electrocardiographic monitoring with a telemetric system for 5 days after study entry. Electrocardiographic data were stored for 24 hours and reviewed daily. For suspected AF, a 12-lead electrocardiogram was obtained to confirm occurrence of AF. Thereafter, an electrocardiogram was obtained 2 times/day routinely and anytime the patient developed new symptoms or if physical examination revealed any irregular rhythm up to hospital discharge. Two blinded cardiologists evaluated electrocardiograms. Only POAF episodes lasting >5 minutes were counted. For persistent POAF (≥15 minutes), amiodarone was administrated. When amiodarone was ineffective, electrical cardioversion was permitted. After discharge, each patient returned for a daily electrocardiographic recording for up to 4 weeks from first admission, independent of whether POAF was detected during in-hospital stay.
A 2-dimensional, M-mode, and Doppler echocardiogram was obtained at baseline in all patients according to recommendations of the American Society of Echocardiography. LV ejection fraction was calculated using the Simpson rule algorithm.
Differences in baseline characteristics between the 2 treatment groups were evaluated by chi-square and unpaired t tests. Within-group changes in reported variables were evaluated by paired t test. Between-group comparisons were performed by unpaired t test or Mann-Whitney rank-sum test. Sample size calculation was based on an expected 20% occurrence of AF in the bisoprolol and carvedilol arms according to the rate of POAF reported in previous studies; we estimated the need of enrolling 320 patients with 80% power and significance 0.05 and assuming a dropout rate of 20%. Statistical analysis was carried out on an intention-to-treat basis.
The Kaplan-Meier method was used for outcome analysis using as an end point the occurrence of POAF. Rate of POAF occurrence between the 2 treatment groups was compared by log-rank test. Relative risks and 95% confidence intervals were calculated to compare outcome between groups by Cox proportional hazard model. Results are expressed as mean ± SD. A 2-tailed p value <0.05 was considered statistically significant.
Results
Of 388 patients screened, 320 patients meet the inclusion criteria and were included in the study. Twenty-nine patients were not included because of respiratory failure, 19 patients had renal failure in advanced stage, 10 patients had severe bradycardia, and 10 patients were on antiarrhythmic drugs.
Baseline patient characteristics are presented in Table 1 . There were no significant differences between the 2 groups with respect to all variables. Mean doses of bisoprolol and carvedilol used were 2.5 ± 0.2 and 12.5 ± 2 mg/day, respectively.
| Variables | Bisoprolol | Carvedilol |
|---|---|---|
| (n = 160) | (n = 160) | |
| Age (years) | 67.5 ± 8 | 64.6 ± 11 |
| Men/women | 120 (75%)/40 (25%) | 111 (69%)/49 (30%) |
| On-pump/off-pump coronary artery bypass grafting | 127 (79%)/33 (21%) | 135 (84%)/26 (16%) |
| New York Heart Association classes I/II/III | 55 (34%)/79 (49%)/27 (17%) | 51 (33%)/76 (47%)/32 (20%) |
| Diabetes mellitus | 63 (39%) | 69 (43%) |
| Dyslipidemia ⁎ | 88 (55%) | 96 (60%) |
| Hypertension (blood pressure ≥140/90 mm Hg) | 129 (81%) | 108 (68%) |
| Left ventricular ejection fraction (%) | 30 ± 2 | 30 ± 2 |
| Left atrial area (cm 2 ) | 19 ± 2 | 19 ± 1 |
| Left ventricular end-diastolic diameter (mm) | 72 ± 2 | 70 ± 2 |
| Distance walked at 6-minute walking test (meters) | 417 ± 120 | 423 ± 101 |
| Concomitant therapy | ||
| Angiotensin-converting enzyme inhibitor | 145 (91%) | 149 (93%) |
| Diuretic | 118 (74%) | 125 (78%) |
| Angiotensin receptors blocker | 51 (32%) | 46 (29%) |
| Aldosterone receptors blockers | 93 (58%) | 106 (66%) |
| Digoxin | 46 (29%) | 42 (26%) |
| Antiplatelet agent | 159 (99%) | 155 (97%) |
| Statin | 97 (61%) | 119 (74%) |
| Anticoagulant | 34 (21%) | 31 (19%) |
⁎ Included total cholesterol >200 mg/dl, low-density lipoprotein cholesterol >100 mg/dl, and triglycerides >150 mg/dl.
During follow-up, 58 patients (18.1%) developed POAF in the overall population. Rate of occurrence of POAF in the bisoprolol group was 14.6% (23 of 160) versus 23% (37 of 160) in the carvedilol group (relative risk 0.6, confidence interval 0.4 to 0.9, p = 0.032). Kaplan-Meier actuarial estimates of POAF occurrence in the study groups are shown in Figure 1 . Four of 23 patients (17%) on bisoprolol treatment had >1 episode of POAF (3.6 ± 0.4), and 5 of 37 patients on carvedilol had >1 episode of POAF (2.9 ± 0.6, p = 0.21). In 15 of 58 patients (26%) who had POAF, episodes were asymptomatic, without significant differences between treatment groups.
Electrical cardioversion was performed in 1 patient in the bisoprolol group and in 2 patients in the carvedilol group (p = 0.86).
Effects of bisoprolol and carvedilol on hemodynamic parameters are presented in Figure 2 and Table 2 . At the end of follow-up, in the bisoprolol and carvedilol groups a significant decrease in heart rate from baseline was detected, without significant effects on blood pressure ( Table 2 ). However, patients in the bisoprolol group had a significantly greater decrease in heart rate (−15.6 ± 3 vs −9.4 ± 3 beats/min, p = 0.021; Figure 2 ).
| Variable | Bisoprolol Group (n = 150) | Carvedilol Group (n = 148) | ||
|---|---|---|---|---|
| Baseline | 4 Weeks | Baseline | 4 Weeks | |
| Systolic blood pressure (mm Hg) | 127 ± 29 | 122 ± 21 | 129 ± 35 | 118 ± 27 |
| Diastolic blood pressure (mm Hg) | 83 ± 12 | 80 ± 13 | 86 ± 10 | 83 ± 12 |
| Heart rate (beats/min) | 74 ± 7 | 59 ± 6 ⁎ | 78 ± 10 | 68 ± 11 ⁎ |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
