- Top of page
Postoperative atrial fibrillation (POAF) is 1 of the most common complications after cardiac surgery, predisposing patients to a significantly increased risk of mortality and morbidities including stroke, hemodynamic compromise, and ventricular arrhythmia.1–5 POAF has been reported to occur in 10% to 40% of patients undergoing off-pump coronary artery bypass graft (OPCAB), which has been routinely performed in >50 major cardiovascular centers across China.6 OPCAB has not been verified to decrease the incidence of POAF compared to conventional on-pump coronary artery bypass graft (CABG).7,8
The renin-angiotensin system and angiotensin II type 1 receptor have been implicated in the atrial structural and electrical remodeling, which could lead to the development of atrial fibrillation (AF).9–12 The blockade of the renin-angiotensin system using an angiotensin-converting enzyme inhibitor (ACEI) or an angiotensin receptor blocker (ARB) has been shown to favorably affect atrial remodeling in animal and human models.12–15 These findings suggest that preoperative ACEI or ARB use might reduce the incidence of POAF by impacting atrial remodeling. However, a few studies examining the clinical effect of ACEI or ARB on POAF have yielded conflicting results.16–20 In addition, little attention has been paid to the drug class and treatment duration (TD) regarding ACEI or ARB administration. The effects of ACEI or ARB on POAF were not previously studied separately, despite different mechanisms of ACEI and ARB. ACEI prevents the degradation of bradykinin, whereas ARB does not.21 ARB has been shown to have a stronger systemic anti-inflammatory effect compared with ACEI.22 It has been reported that ACEI has a significant protective effect on preventing new onset AF, but the effect of ARB was not found to be significant in a meta-analysis including nonsurgical patients.23
Therefore, we aimed to investigate whether the preoperative use of ACEI or ARB, ACEI alone, and ARB alone exerted an effect on the incidence of POAF in patients undergoing OPCAB. In addition, we examined whether the TD of ACEI or ARB had influence on the incidence of POAF.
- Top of page
The study examined 1050 consecutive patients who underwent elective isolated OPCAB at Asan Medical Center, Seoul, South Korea, from January 2006 to December 2009. Patients with AF, atrial flutter, paced rhythm on preoperative electrocardiogram, history of AF, hyperthyroidism, patients undergoing emergency surgery, and patients enrolled in other clinical trials were excluded. The research protocol was approved and the requirement of written informed consent was waived by our institutional review board.
Patient data were obtained through the review of electronic medical records by 2 trained investigators independently. In the case of disagreement, they reviewed the records together. Clinical data included age, gender, preoperative electrocardiogram, left main coronary artery disease (≥70% stenosis on preoperative coronary angiography), cerebrovascular disease (history of ischemic or hemorrhagic stroke, carotid artery stent insertion, or ≥80% occlusion in the carotid artery), peripheral vascular disease (history of claudication, amputation for arterial insufficiency, aortoiliac occlusive disease, vascular reconstruction surgery, stent insertion at a peripheral artery, or aortic aneurysm), chronic kidney disease (serum creatinine >1.5 mg/dL or estimated glomerular filtration rate <60 mL/min/1.73 m2), renal failure (requirement of dialysis), diabetes mellitus (fasting blood glucose ≥126 mg/dL or requirement of antidiabetic medication), hypertension (systolic blood pressure ≥140 mm Hg or diastolic pressure ≥90 mm Hg in patients without diabetes mellitus and chronic kidney disease, systolic blood pressure ≥130 mm Hg or diastolic pressure ≥80 mm Hg in patients with diabetes mellitus or chronic kidney disease, or requirement of antihypertensive medication), dyslipidemia (low-density lipoprotein cholesterol >100 mg/dL, prediagnosed dyslipidemia, or requirement of lipid-lowering medications including statin and nonstatin lipid-lowering drugs), unstable angina, previous myocardial infarction, congestive heart failure, previous percutaneous coronary intervention, previous CABG, chronic obstructive pulmonary disease, liver disease (liver cirrhosis, hepatitis, or hepatocellular carcinoma), left atrial diameter, left ventricular ejection fraction, EuroSCORE (logistic), smoking status, duration of surgery, infused volume during surgery (crystalloid, colloid, packed red blood cell, and cell saver blood), the number of anastomosed coronary vessel, and weight gain at postoperative day 1. Laboratory data included preoperative hematocrit, serum creatinine, total bilirubin, albumin, and C-reactive protein.
Details regarding preoperative use of ACEI or ARB were also obtained from electronic medical records, including the drug class and the TD if ACEI or ARB were prescribed. TD of ACEI or ARB was defined as the shortest duration that could be confirmed according to the medical records, and was divided into 4 groups: TD 1 = <1 week, TD 2 = 1 week to <1 month, TD 3 = 1 month to <1 year, and TD 4 = 1 year or longer. ACEI or ARB treatment was typically ceased on the day of surgery to prevent intraoperative refractory hypotension and restarted postoperatively at the intensive care unit at the attending physician's discretion. Data regarding preoperative use of β-blocker, calcium channel blocker, insulin, oral hypoglycemic agent, lipid-lowering drug, aspirin, clopidogrel, and diuretics were also acquired. β-blocker was used preoperatively, continued until the day of surgery, and resumed as soon as possible postoperatively.
The primary study end point was POAF prior to discharge. Based on the Society of Thoracic Surgeon National Adult Cardiac Database, POAF was defined as new-onset AF that required therapy including β-blocker, calcium channel blocker, amiodarone, anticoagulation, or electrical cardioversion irrespective of AF duration. Hemodynamically stable AF that did not require any of the listed intervention was not considered. POAF was reviewed during the postoperative period or from the arrival at the intensive care unit after surgery to before discharge. Rhythm was monitored continuously with lead II and V5 electrocardiogram in the intensive care unit. In the general ward, electrocardiogram was acquired once a day routinely and when the patients experienced newly developed symptoms, such as palpitation or dizziness, or when physical examination disclosed an irregular rhythm.
Continuous data are expressed as mean ± SD for normally distributed data or median (interquartile range) for nonparametric data, and categorical data as frequencies (percentages). Continuous variables were compared using the t-test or Mann-Whitney U test for parametric and nonparametric variables, respectively. Categorical variables were compared using the χ2 test or Fisher's exact test, as appropriate.
Crude and adjusted risks for POAF were compared using univariate and multivariate logistic regression analyses. Univariate analysis was performed independently with the covariates shown in Table 1. Variables with P values ≤0.20 in univariate analyses were candidates for inclusion in the multivariate logistic regression model with backward elimination. Model discrimination was assessed with c statistics (c index = 0.68, 0.67, 0.69 for ACEI or ARB, ACEI alone, and ARB alone, respectively) and model calibration was assessed with the Hosmer-Lemeshow goodness-of-fit test (P = 0.87, 0.84, 0.78 for ACEI or ARB, ACEI alone, and ARB alone, respectively).
Table 1. Perioperative Demographic and Clinical Characteristics
|Variables||Neither ACEI nor ARB, n = 643||ACEI or ARB, n = 407||ACEI Alone, n = 113||ARB Alone, n = 279 |
|Age, y||65.0 (57.0–70.0)||64.0 (58.0–70.0)||62.8 ± 8.8||65.0 (57.3–70.0)|
|Female sex||159 (24.7)||103 (25.3)||31 (27.4)||71 (25.4)|
|BMI, kg/m2||24.6 ± 2.6||24.6 (23.0–26.6)||24.6 ± 2.9||24.7 (23.0–26.7)|
|Diabetes mellitus||268 (41.7)||219 (53.8)a||53 (46.9)||154 (55.2)a|
|Hypertension||361 (56.1)||302 (74.2)a||77 (68.1)a||214 (76.7)a|
|Smoking||338 (52.6)||207 (50.9)||62 (54.9)||134 (48.0)|
|Dyslipidemia||486 (75.6)||322 (79.1)||91 (80.5)||220 (78.9)|
|Previous MI||102 (15.9)||85 (20.9)a||29 (25.7)a||54 (19.4)|
|Previous PCI||120 (18.7)||91 (22.4)||25 (22.1)||58 (20.8)|
|Previous CABG||9 (1.4)||1 (0.2)||1 (0.9)||0 (0.0)|
|Congestive heart failure||10 (1.6)||12 (2.9)||7 (6.2)a||5 (1.8)|
|Cerebrovascular disease||107 (16.6)||87 (21.4)||20 (17.7)||64 (22.9)a|
|PVD||18 (2.8)||13 (3.2)||3 (2.7)||10 (3.6)|
|COPD||41 (6.4)||27 (6.6)||4 (3.5)||22 (7.9)|
|Chronic kidney disease||39 (6.1)||58 (14.3)a||8 (7.1)||42 (15.1)a|
|Renal failure||13 (2.0)||20 (4.9)a||2 (1.8)||16 (5.7)a|
|Liver disease||13 (2.0)||6 (1.5)||2 (1.8)||4 (1.4)|
|Logistic EuroSCORE, %||1.8 (1.2–3.1)||2.3 (1.3–3.8)a||2.2 (1.3–3.5)||2.4 (1.4–3.9)a|
|LA size, mm||39.0 (36.0–42.0)||40.0 (37.0–44.0)a||40.0 (37.0–43.0)||40.0 (37.0–44.0)a|
|LVEF, %||60.0 (54.0–64.0)||59.0 (49.0–64.0)a||57.0 (43.8–63.0)a||59.0 (52.0–64.0)a|
|Laboratory findings|| || || || |
| Hematocrit, %||39.2 (35.9–42.0)||37.9 (34.0–41.1)a||38.2 ± 4.8||37.4 ± 5.3a|
| Creatinine, mg/dL||0.9 (0.8–1.1)||1.0 (0.8–1.2)a||0.9 (0.8–1.0)||1.0 (0.8–1.3)a|
| Total bilirubin, mg/dL||0.7 (0.6–0.9)||0.7 (0.5–0.8)a||0.7 (0.5–0.8)||1.0 (0.8–1.3)a|
| Albumin, g/dL||3.8 (3.5–4.0)||3.8 (3.4–4.0)a||3.8 (3.5–4.0)||3.7 (3.4–4.0)|
| CRP, mg/dL||0.2 (0.1–0.5)||0.2 (0.1–0.5)||0.2 (0.1–0.7)||0.2 (0.1–0.4)|
|Left main disease at CAG||134 (20.8)||88 (21.6)||18 (15.9)||64 (23.0)|
|Medications|| || || || |
| β-blocker||383 (59.6)||259 (63.6)||79 (70.0)a||172 (61.6)|
| CCB||471 (73.3)||278 (68.3)||71 (62.8)a||197 (70.6)|
| Insulin||171 (26.6)||160 (39.3)a||42 (37.2)a||110 (39.4)a|
| OHA||183 (28.5)||167 (41.0)a||45 (39.8)a||114 (40.9)a|
| Lipid lowering agent||397 (61.7)||280 (68.8)||84 (74.3)a||186 (66.7)|
| Statin||375 (58.3)||268 (65.8)a||79 (69.9)a||180 (64.5)|
| Nonstatin||11 (1.7)||4 (1.0)||1 (0.9)||3 (1.1)|
| Both statin and nonstatin||11 (1.7)||8 ((2.0)||4 (3.5)||3 (1.1)|
| Aspirin||446 (69.4)||282 (69.3)||75 (66.4)||195 (70.0)|
| Clopidogrel||324 (50.4)||223 (54.8)||66 (58.4)||146 (52.3)|
| Diuretics||54 (8.4)||119 (29.2)a||24 (21.2)a||88 (31.5)a|
|Intraoperative|| || || || |
| Number of anastomoses||3.0 (2.0–4.0)||3.0 (2.0–4.0)||3.0 (2.0–4.0)||3.0 (2.0–4.0)|
| Duration of surgery, min||225.0 (190.0–276.5)||220.0 (190.0–270.0)||220.0 (190.0–265.0)||221.0 (190.0–280.0)|
| Crystalloid, mL||1400.0 (1100.0–1800.0)||1400.0 (1000.0–1700.0)||1300.0 (1000.0–1700.0)a||1400.0 (1012.5–1700.0)|
| Colloid, mL||1100.0 (900.0–1500.0)||1050.0 (900.0–1500.0)||1200.0 (937.5–1500.0)||1000.0 (950.0–1500.0)|
| Packed RBC, units||0.5 (0.0–2.0)||1.0 (0.0–2.0)a||1.0 (0.0–2.0)||1.0 (0.0–2.0)a|
| Cell saver blood, mL||0.0 (0.0–200.0)||0.0 (0.0–200.0)||0.0 (0.0–200.0)||0.0 (0.0–200.0)|
|Weight gain at POD 1, kg||0.8 (0.0–1.7)||1.0 (0.1–1.8)||0.9 (0.1–1.7)||1.0 ± 1.4|
In addition, we performed propensity score matching analysis to adjust for potential confounders. The propensity score was the probability that a patient would receive medication and was computed using a logistic regression modeling including the covariates shown in Table 1. After calculating the propensity score, 302 patients who received ACEI or ARB, 104 patients who received ACEI alone, and 212 patients with ARB alone were successfully matched to the patients who did not in a 1:1 fashion, respectively.
Results are reported as odds ratios (OR) with a 95% confidence interval (CI). P values of <0.05 were considered to indicate statistical significance. All statistical analysis was performed using SAS software version 9.1 (SAS Institute, Cary, NC).
- Top of page
Patient demographics and perioperative variables are shown in Table 1. The overall incidence of POAF was 16.3% (171 of 1050 patients). POAF developed in 18.9% of patients receiving preoperative ACEI or ARB (77 of 407 patients) and in 14.6% of patients receiving neither ACEI nor ARB (94 of 643 patients). Of 407 patients receiving preoperative ACEI or ARB, POAF occurred in 18.6% of patients treated with ACEI alone (21 of 113 patients) and 18.6% of patients treated with ARB alone (52 of 279 patients).
Multivariate adjusted logistic regression analysis demonstrated that there were no associations between ACEI or ARB, ACEI alone, or ARB alone use and POAF (Table 2).
Table 2. Associations Between Preoperative Angiotensin-Converting Enzyme Inhibitor or Angiotensin Receptor Blocker Use and Postoperative Atrial Fibrillation
|Neither ACEI nor ARB|| ||Reference|| || |
|ACEI or ARB||Crude||1.36||0.98-1.90||0.07|
| ||PS matching||1.66||1.04-2.62||0.03|
| ||PS matching||1.30||0.57-2.97||0.53|
| ||PS matching||1.57||0.93-2.64||0.09|
|Treatment duration of ACEI or ARB||Multivariate|| || || |
| <1 week|| ||1.31||0.79-2.17||0.30|
| 1 week to <1 month|| ||1.01||0.60-1.72||0.97|
| 1 month to <1 year|| ||1.06||0.47-2.37||0.89|
| 1 year or longer|| ||2.12||0.97-4.64||0.06|
On the other hand, analysis of the propensity score matched cohort found that preoperative ACEI or ARB treatment was an independent risk factor for POAF (OR: 1.66, 95% CI: 1.04–2.62, P = 0.03), whereas ACEI alone was not (OR: 1.30, 95% CI: 0.57–2.97, P = 0.53). In the patients treated with ARB alone, increased incidence of POAF showed borderline significance (OR: 1.57, 95% CI: 0.93–2.64, P = 0.09) (Table 2).
We found that the incidence of POAF was 17.9% (27 of 151 patients) in the TD 1 group, 17.1% (27 of 158 patients) in the TD 2 group, 18.5% (10 of 54 patients) in the TD 3 group, and 29.5% (13 of 44 patients) in the TD 4 group. Multivariate adjusted logistic regression analysis indicated that there was no relationship between the duration of preoperative ACEI or ARB use and the occurrence of POAF in the TD 1, TD 2, and TD 3 groups (Table 2). In the TD 4 group, the increased risk of POAF was demonstrated to be of borderline significance (P = 0.06) (Table 2).
In addition, age (OR: 1.06, 95% CI: 1.04–1.09, P < 0.0001) and left atrial diameter (OR: 1.05, 95% CI: 1.02–1.09, P = 0.005), which were assessed as continuous variables, were documented as independent risk factors for POAF by multivariate regression analysis.
- Top of page
The present study found that preoperative use of ACEI or ARB, ACEI alone, and ARB alone did not exert any beneficial effect on the occurrence of POAF in patients undergoing isolated OPCAB, and preoperative ACEI or ARB use, especially ARB use, was instead associated with an increased incidence of POAF.
There are previous studies that have investigated the effect of ACEI or ARB on the development of POAF, yielding discordant results. In patients undergoing CABG or valve surgery, some studies showed that there was no association between the preoperative use of ACEI or ARB and the incidence of POAF.18,19 On the other hand, ACEI alone and ACEI plus candesartan were reported to have decreased the incidence of POAF compared with the control group,17 and ACEI or ARB were demonstrated to have decreased the incidence of POAF.20 These studies included valve surgery as well as CABG and excluded patients with electrolyte imbalance, prior cardiac surgery, reduced ejection fraction, and enlarged left atrium; therefore, the patients in these previous studies were different from the patient cohort in our study, which may substantially affect the results. On the contrary, the preoperative use of ACEI has been shown to impose additional risk on the development of POAF in patients undergoing isolated CABG.16 This study may be more comparable to our study, based on both concordant results and similar patient cohorts investigated, which included only isolated CABG and not valve surgery despite different definitions of POAF.
In the present study, preoperative use of ACEI or ARB was associated with an increased incidence of POAF in patients undergoing OPCAB. Additionally, treatment with ACEI or ARB for 1 year or longer showed a trend toward an increased incidence of POAF. In fact, preoperative use of ACEI or ARB has been known to be a predictor of refractory hypotension,24,25 resulting in excessive volume loading or use of inotropes, which have been reported to be associated with POAF in patients undergoing cardiac surgery.26,27 Abrupt volume loading is likely to cause an acute atrial stretch that has been known to lower the threshold for the development of AF.28 Thus, it is plausible that the hypotensive effect of ACEI or ARB might contribute to the development of POAF. Moreover, a trend toward increasing POAF in patients treated with ARB, not ACEI, might be related to the finding that hypotensive events have been known to occur more often in patients treated with ARB than in those treated with ACEI.29 Likewise, the report that patients under long-term treatment with ACEI or ARB experienced severe hypotensive episodes intraoperatively might support our result regarding the clinical association between the longer duration of medication and POAF.30 ACEI or ARB has been established to benefit structural and functional atrial remodeling that are attributable to chronic atrial stretch.12–15 The patients undergoing OPCAB, however, are less likely to experience chronic atrial stretch that primarily occurs in valvular heart disease. The anti-inflammatory properties of ACEI or ARB31,32 might not provide the advantage of reducing POAF, given the attenuated inflammation during OPCAB compared with on-pump CABG.33 Therefore, we speculated that the hypotensive effect of ACEI or ARB might outweigh the beneficial effects of these medications in patients undergoing OPCAB.
The current study had a couple of strengths. First, it separately determined the effect of ACEI or ARB on the occurrence of POAF. Second, it investigated whether TD of ACEI or ARB was a factor that affected the incidence of POAF.
Our study has the following limitations. First, the specific dose and the exact duration of preoperative ACEI or ARB use were not considered, because such preadmission data were not available in this retrospective cohort study design. TD was taken to be the shortest duration that could be determined from medical records. Accordingly, it was possible that actual TD was longer than that used in the analysis. Second, the present study could not investigate the effect of combined ACEI and ARB administration because too few patients underwent such treatment (n = 15). Third, the sample size of the 4 subgroups according to the duration of medication might be too small to assess the effect of TD on POAF, so we could not reach a conclusive result regarding the effect of TD on POAF. Further investigations with a larger number of patients will be needed to confirm the effect of TD on POAF.