Observational Study on the Beneficial Effect of Preoperative Statins in Reducing Atrial Fibrillation After Coronary Surgery

CARDIOVASCULAR

Observational Study on the Beneficial Effect of Preoperative Statins in Reducing Atrial Fibrillation After Coronary Surgery Giovanni Mariscalco, MD, Roberto Lorusso, MD, PhD, Catherine Klersy, MD, MS, Sandro Ferrarese, MD, Matteo Tozzi, MD, Davide Vanoli, MD, Bruno Vito Domenico, MD, and Andrea Sala, MD Department of Surgical Sciences, Cardiac Surgery Unit, Varese University Hospital, University of Insubria, Varese; Cardiac Surgery Unit, Civic Hospital, Brescia; Service of Biometry and Clinical Epidemiology, Scientific Direction, IRCCS Fondazione Policlinico San Matteo, Pavia; Department of Surgical Sciences, Vascular Surgery Unit, Varese University Hospital, University of Insubria, Varese, Italy

Background. Recent evidence supports the important role of inflammation in atrial fibrillation (AF) after coronary artery bypass grafting (CABG) and there is growing evidence that statin has cardiac antiarrhythmic effects. The aim of this study was to assess the efficacy of preoperative statins in preventing AF after CABG in a longitudinal observational study. Methods. Over a two-year period, 405 consecutive patients underwent isolated CABG procedures. Univariate analysis was performed exploring the relationship regarding statin use and AF development. A propensity score for treatment with statins was obtained from core patient characteristics. The role of statin therapy on postoperative AF was assessed by means of a conditional logistic model, while stratifying on the quintiles of the propensity score. All analysis was performed retrospectively.

Results. Postoperative AF occurred in 29.5% of the patients with preoperative statin therapy compared with 40.9% of those patients without it (p ⴝ 0.021). No statistical differences among development of AF and type, dose, or duration of preoperative statin therapy were observed. Preoperative statins were associated with a 42% reduction in risk of AF development after CABG surgery (odds ratio [OR] 0.58, 95% confidence interval [CI] 0.37 to 0.91, p ⴝ 0.017, while stratifying on the propensity score). No different effect of statins on AF was observed with respect to age groups (< 70 and >70 years) (interaction p ⴝ 0.711). Conclusions. Preoperative statins may reduce postoperative AF after CABG. Patients undergoing elective revascularization may benefit from a preventive statin approach. (Ann Thorac Surg 2007;84:1158 – 65) © 2007 by The Society of Thoracic Surgeons

A

operative determinants of AF was evaluated as a secondary endpoint.

trial fibrillation (AF) is the most common complication after cardiac surgery, with important clinical and economic implications [1, 2]. Although several studies have identified various baseline predictors of postoperative AF, the pathophysiologic mechanisms remain unclear [1]. Observational evidences seemed to testify an inflammatory component of this postoperative arrhythmia [3, 4]. Statin drugs, which have both antioxidant and antiinflammatory properties, may attenuate postoperative AF development and constitute a potential preventive approach [5]. Patients with preoperative statins undergoing cardiac surgery demonstrated a lower incidence of postoperative AF [6 – 8]. Therefore, the purpose of the present study was to evaluate the influence of preoperative statin therapy on AF development after coronary artery bypass grafting (CABG). The identification of further operative and postAccepted for publication May 7, 2007. Address correspondence to Dr Mariscalco, Department of Surgical Sciences, Cardiac Surgery Division, Varese University Hospital, Varese, I-21100, Italy; e-mail: [email protected].

© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc

Patients and Methods Patients and Study Design Over a two-year period (2005 to 2006), all consecutive patients scheduled for primary isolated CABG (n ⫽ 422) were considered for the study. Patients with permanent pacemaker (n ⫽ 5), chronic AF (n ⫽ 7), or inflammatory diseases that required therapy with steroids or nonsteroidal antiinflammatory drugs (n ⫽ 5), were subsequently excluded. Thus, 405 patients represented the study population; 320 (79%) were operated on cardiopulmonary bypass (CPB) and 85 (21%) underwent off-pump CABG. The study cohort had an average age of 66.1 ⫾ 9.4 years (range, 34 to 85 years) and contained 81% men. Demographic data are listed in Table 1. Although all data were prospectively recorded into the hospital computerized database registry that remained consistent over the study period, this is an observational retrospective study. The database registry captures 100% 0003-4975/07/$32.00 doi:10.1016/j.athoracsur.2007.05.021

Ann Thorac Surg 2007;84:1158 – 65

MARISCALCO ET AL STATINS AND POSTOPERATIVE ATRIAL FIBRILLATION

1159

Unit

Statin (n ⫽ 218)

p Value

No Statin (n ⫽ 187)

AF (n ⫽ 141)

p Value

No AF (n ⫽ 264)

[y] [%] [Kg/m2]

65.5 ⫾ 9.1 16.6 26.8 ⫾ 3.8

0.274 0.305 0.600

65.5 ⫾ 9.7 21.0 26.6 ⫾ 3.7

70.5 ⫾ 8.0 22.7 26.3 ⫾ 3.9

⬍0.001 0.139 0.165

63.7 ⫾ 9.2 16.3 26.9 ⫾ 3.6

[%] [1–4] [%] [%] [%] [%] [mm] [%] [%]

22.1 3.1 ⫾ 0.9 5.5 64.1 11.8 53.6 ⫾ 11.0 38.4 ⫾ 5.3 29.8 65.6

0.556 0.930 0.999 0.025 0.073 0.898 0.784 0.665 0.834

24.7 3.1 ⫾ 1.0 5.4 52.7 18.4 53.7 ⫾ 11.6 38.1 ⫾ 4.7 32.3 64.5

21.3 3.1 ⫾ 1.0 10.6 62.1 12.9 51.8 ⫾ 11.4 37.6 ⫾ 6.2 30.2 68.3

[%] [%] [%] [%] [%] [%] [%] [score]

70.5 31.5 50.5 8.8 13.0 5.6 5.6 4.0 ⫾ 2.3

0.657 0.512 0.023 0.859 0.058 0.425 0.825 0.443

73.0 28.1 38.7 8.1 20.5 7.6 4.9 4.2 ⫾ 2.8

72.1 33.1 39.6 12.2 19.4 6.5 5.3 4.9 ⫾ 2.5

66.2 25.8 55.4 n.a.

0.151 0.730 0.476

73.1 27.8 51.7 n.a.

70.6 28.9 57.8 45.7

Variables Demographics Age Females BMI Cardiac Unstable CCS History of AF Prior AMI Prior PCI LVEF LA AP dimension LMCA stenosis RCA stenosis Comorbidities Hypertension Diabetes Dyslipidemia COPD PVD CVA Renal failure Euroscore Preoperative medications ␤-blockers Ca-antagonists ACE-Inhibitors Statins

[%] [%] [%] [%]

0.464 0.678 0.002 0.340 0.384 0.025 0.450 0.821 0.324

24.6 3.1 ⫾ 1.0 2.7 56.8 16.7 54.6 ⫾ 11.1 38.5 ⫾ 4.2 31.6 63.1

0.818 0.305 0.107 0.059 0.325 0.999 0.999 ⬍0.001

70.8 28.1 48.4 6.5 15.3 6.5 5.0 3.6 ⫾ 2.4

0.420 0.473 0.242 0.021

66.4 25.5 51.4 58.2

Data are mean ⫾ SD or pecentages. ACE ⫽ angiotensin-converting enzyme; AF ⫽ atrial fibrillation; AMI ⫽ acute myocardial infarction; BMI ⫽ body mass index; CCS ⫽ Canadian Cardiovascular Society angina classification; Ca-antagonists ⫽ calcium channel blockers; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebrovascular accident; LA ⫽ left atrial diameter; LMCA ⫽ left main coronary artery; LVEF ⫽ left ventricular ejection fraction; PVD ⫽ peripheral vascular disease; RCA ⫽ right coronary artery.

of cardiac surgeries performed at Varese University Hospital and is validated by random audits of patient charts. Data entry is managed by surgeons, anesthetists, and perfusionists involved in the care of the patients. Coredata input had a “check-in– check-out” character at times of admission, operation, intensive care unit, and at discharge or subsequent clinical visits, whereas other observations (such as atrial fibrillation) are entered on a daily basis and confirmed at “check-out.”

Data Definitions All variables analyzed in the study are listed in the Appendix. Preoperative patient characteristics were evaluated after current guidelines. In particular, hypertension was considered according to the Joint National Committee VI, diabetes mellitus according to the American Diabetes Association, and dyslipidemia according to the National Cholesterol Education Program–Adult Treatment Panel II criteria [9 –11]. Angina was defined by

the Canadian Cardiovascular Society Classification [12]. Chronic obstructive pulmonary disease (COPD), peripheral vascular disease, and renal failure were defined according to the European system for cardiac operative risk evaluation [13]. The latest was also considered to quantify the preoperative surgical risk [13]. The study was approved by our Institutional Review Board (protocol nr.117) and, because of its retrospective nature, individual patient consent was waived. The study was also not supported by any external source of funding.

Clinical Management and AF Definition All patients were subjected to preoperative coronary angiography, transthoracic echocardiography, and colorDoppler ultrasound of epiaortic vessels. Preoperative medications, including ␤-blockers, diuretics, antihypertensives, statins, and calcium-channel blockers, were routinely omitted on the day of the operation and re-

CARDIOVASCULAR

Table 1. Clinical Characteristics of the Population Study and Relationship With Statin Use and Postoperative Atrial Fibrillation

1160

MARISCALCO ET AL STATINS AND POSTOPERATIVE ATRIAL FIBRILLATION

Ann Thorac Surg 2007;84:1158 – 65

Table 2. Operative and Postoperative Variables With Relationship With Statin Use and Postoperative Atrial Fibrillation Variables CARDIOVASCULAR

Operative factors On-pump procedures Grafts CPB-time ACC-time IABP Postoperative data Inotropic drugs Respiratory failure Perioperative AMI Reoperation for bleeding Ventilation time ICU stay LOS Hospital mortality

Unit

Statin (n ⫽ 218)

p Value

No Statin (n ⫽ 187)

AF (n ⫽ 141)

p Value

no AF (n ⫽ 264)

[%] [n] [min] [min] [%]

76.0 2.8 ⫾ 0.8 84.7 ⫾ 37.8 53.1 ⫾ 21.7 2.3

0.110 0.567 0.311 0.838 0.075

82.8 2.9 ⫾ 0.8 88.6 ⫾ 29.9 53.5 ⫾ 18.8 5.9

78.7 2.9 ⫾ 0.7 88.2 ⫾ 31.0 52.5 ⫾ 15.6 7.1

0.999 0.316 0.522 0.621 0.029

79.2 2.8 ⫾ 0.9 85.6 ⫾ 35.8 53.6 ⫾ 22.3 2.3

[%] [%] [%] [%] [h] [h] [days] [%]

68.7 12.4 5.9 3.8 24.1 ⫾ 60.4 42.4 ⫾ 69.0 7.8 ⫾ 4.0 0.0

0.322 0.334 0.827 0.791 0.283 0.473 0.134 0.999

73.7 9.3 5.1 3.2 30.1 ⫾ 46.5 47.1 ⫾ 53.3 8.4 ⫾ 4.7 0.5

73.0 15.0 3.6 3.6 33.0 ⫾ 75.9 52.9 ⫾ 86.5 8.2 ⫾ 3.2 0.7

0.492 0.065 0.356 0.999 0.105 0.112 0.575 0.348

69.3 8.7 6.5 3.4 23.5 ⫾ 36.7 40.1 ⫾ 42.4 8.0 ⫾ 4.8 0.0

Data are mean ⫾ SD or pecentages. ACC ⫽ aortic cross clamp; AF ⫽ atrial fibrillation; AMI ⫽ acute myocardial infarction; balloon pump; ICU ⫽ intensive care unit; LOS ⫽ lenght of hospital stay.

started after the operation, unless clinically contraindicated. Commercially available statins screened for in this study were atorvastatin, pravastatin, simvastatin, lovastatin, rosuvastatin, and fluvastatin. All patients were daily monitored until discharge with continuous electrocardiographic telemetry, as well as standard 12-lead electrocardiography. Additional recordings were collected at clinical suspicion of AF. Only AF episodes lasting longer than 15 minutes were considered [14]. Atrial flutter or tachycardia were disregarded because of their different mechanisms [15]. Amiodarone, either oral or intravenously administered, constituted the standard pharmacologic treatment of AF. Patients without successful rhythm cardioversion and with persistent AF were given warfarin and were discharged to the referring hospital, with a planned electric cardioversion within thirty days.

Surgical Techniques and Operative Management All procedures were performed through a median sternotomy approach, using either an on-pump or off-pump technique. The anesthetic management followed identical routine methods. Cardiopulmonary bypass was of standardized fashion, with ascending aortic cannulation and two-stage venous cannulation of the right atrium. The CPB was conducted using a membrane oxygenator with hemodilution and mild systemic hypothermia. Myocardial protection was afforded by intermittent cardioplegia (crystalloid or blood, antegrade or retrograde routes). Off-pump procedures were conducted using the CTS stabilization system (Guidant, Santa Clara, CA) at near normothermia. Intracoronary shunts were used whenever possible. To reduce blood loss, blood was recollected with a suction cardiotomy reservoir in the CPB group, whereas

CPB ⫽ cardiopulmonary bypass;

IABP ⫽ intraaortic

a cell saver was used in the off-pump group. All procedures had the left internal mammary artery supplying the left anterior descending artery. Other conduits included the right internal mammary artery, radial artery, and saphenous vein, as found appropriate. Proximal anastomoses were performed using a side-biting clamp. After surgery, patients were transferred to a dedicated cardiovascular intensive care unit (ICU). Heart rate, electrocardiogram, central venous and arterial pressures, and acid-base and blood gases were continuously monitored during the ICU stay. Inotropic support was provided if the ventricular contractility was considered frankly impaired in order to achieve stable hemodynamic conditions. Perioperative need for blood products was determined on an individual, patient-by-patient basis; in general, blood transfusions were administered when hemoglobin was less than 9 g/dL. Fluid intake and output were monitored hourly throughout the hospital stay.

Statistics Clinical data were prospectively recorded and tabulated with Microsoft Excel (Microsoft Corp, Redmond, WA). Table 3. Postoperative Atrial Fibrillation Features and Statin Use Event AF occurrence AF, days after surgery AF recurrence AF at discharge

p Value

No Statins

[%] 34.8 29.5 [days] 2.1 ⫾ 0.9 2.1 ⫾ 0.8

0.021 0.683

40.9 2.1 ⫾ 1.0

[%] [%]

0.202 0.723

26.5 7.1

Unit

All Patients

21.5 6.1

AF ⫽ postoperative atrial fibrillation.

Statins

16.4 4.9

Ann Thorac Surg 2007;84:1158 – 65

MARISCALCO ET AL STATINS AND POSTOPERATIVE ATRIAL FIBRILLATION

1161

Table 4. Statin Characteristics and Postoperative Atrial Fibrillation

Type Sinvastatin Atorvastatin Othersa Dose Duration

Unit

All Patients

AF

[%] [%] [%] [mg/die] [days]

44.2 37.7 18.2 21.7 ⫾ 10.0 45.5 ⫾ 22.0

43 38.4 18.6 21.6 ⫾ 10.3 47.2 ⫾ 24.7

p Value

No AF

0.905

0.904 0.596

46.9 35.9 17.2 21.7 ⫾ 9.9 44.8 ⫾ 20.9

a

Pravastatin, lovastatin, rosuvastatin, and fluvastatin (grouped).

AF ⫽ postoperative atrial fibrillation.

Continuous variables were tested for normal distribution by the Kolmogorov-Smirnov test and compared between groups with the unpaired Student ttest for normally distributed values; otherwise the Mann-Whitney Utest was employed. In case of dichotomous variables, group differences were examined by ␹2 or Fisher exact tests as appropriate. The likelihood of receiving a preoperative treatment with statins was taken into account by computing a propensity score (by means of a logistic model) based on the preoperative characteristics listed in Table 1. The role of statin therapy on postoperative AF occurrence (primary endpoint) was assessed by means of a conditional logistic model, while stratifying on the quintiles of the propensity score [16]. Further, the modifying effect of age (ⱕ70 and ⬎70 years) was assessed by including an interaction term in the model. As a secondary endpoint, potential operative and postoperative determinants of AF were also fitted in the model to assess their prognostic role. These included noncollinear clinically relevant variables, with a pvalue less than 0.15 at the univariate analysis. Kaplan-Meier cumulative AF-free survival was computed according to statin treatment and plotted. Results are expressed as mean ⫾ SD for continuous variables and frequencies for the categoric ones. A two-sided pvalue less than 0.05 was considered statistically significant. Statistical analysis was performed with Stata 9.2 (StataCorp, College Station, TX) and with SPSS, release 13.0 for Windows (SPSS Inc, Chicago, IL).

ated with a new-onset of AF (45.1% vs 31.6%, p ⫽ 0.027). No surgical variables, with the exception of postoperative intraaortic balloon pump (IABP) requirement, were found associated with postoperative AF (Table 2). Also, no differences were evident concerning the modality of revascularization. Postoperative AF occurred in 111 (34.7%) of 320 patients operated on CPB versus 30 (35.3%) of 85 patients subjected to off-pump revascularization (p ⫽ 0.999). No differences were observed between patients with preoperative statins and patients without them, with the exception of patients affected by dyslipidemia and previous myocardial infarction (p ⫽ 0.023 and p ⫽ 0.025, respectively). In particular, the time of AF occurrence, AF recurrence, and the persistence of the arrhythmia at discharge were not influenced by statin use (Table 3). Withdrawal from preoperative statin therapy occurred in 51 patients (21.6%) essentially because of transitory transaminase increase (62.7%). However, statin withdrawal was not associated with postoperative AF development, AF recurrence, or persistent AF at discharge (p ⫽ 0.840, p ⫽ 0.573, and p ⫽ 0.874, respectively; data not shown).There were also no statistical differences between

Results Univariate Analysis Univariate analyses regarding statin use and AF occurrence are represented in Tables 1 and 2. The overall incidence of AF in the study population was 34.8%. Postoperative AF occurred in 29.5% of the patients with preoperative statin therapy compared with 40.9% of those without it (p ⫽ 0.021). The use of calcium-channel blockers, of angiotensinconverting enzyme-inhibitors, and of ␤-blockers did not influence the AF development (Table 1). In detail, patients with preoperative ␤-blocker therapy compared with patients without it demonstrated similar AF rates (33.1% vs 37.5%, p ⫽ 0.420). However, withdrawal from preoperative ␤-blocker therapy was significantly associ-

Fig 1. The Kaplan-Meier cumulative estimates of postoperative atrial fibrillation (AF) free survival in patients with (---) preoperative statin therapy and in patients without (—) it (log-rank, p ⫽ 0.018). Data under the curves represent patients at risk.

CARDIOVASCULAR

Variable

1162

MARISCALCO ET AL STATINS AND POSTOPERATIVE ATRIAL FIBRILLATION

CARDIOVASCULAR

development of AF and type, dose, or duration of preoperative statin therapy (Table 4). Preoperative statin use was associated with less AF in the on-pump group (p ⬍ 0.001) compared with the off-pump group (p ⫽ 0.060). Patients with preoperative statins had a shorter length of hospital stay (LOS) compared with patients without them, although not significant (7.8 ⫾ 4.0 vs 8.4 ⫾ 4.7 days, p ⫽ 0.134). When considering the entire study population, no differences were observed in LOS in patients affected by AF compared with patients without it (p ⫽ 0.575). Also, no differences were found in patients with preoperative statins affected by AF or not (7.7 ⫾ 4.4 vs 7.9 ⫾ 3.0 days, p ⫽ 0.816). There was no AF-specific difference in hospital mortality (p ⫽ 0.348) and in oneyear mortality (p ⫽ 0.300). The Kaplan-Meier actuarial estimates of occurrence of postoperative AF in both groups showed a significantly better AF-free survival in the statin group (Fig 1).

Multivariable Analysis (Stratified on the Propensity Score) We then evaluated the role of statins while stratifying on the propensity score to adjust for the likelihood of receiving the treatment, given the presenting characteristics. Indeed, preoperative treatment with statins was associated with a 42% reduction in risk of AF development after CABG surgery (odds ratio [OR] 0.58, 95% confidence interval [CI] 0.37 to 0.91, p ⫽ 0.017). No different effect of statins on AF was observed in the two age groups (ⱕ70 and ⬎70 years) (interaction p ⫽ 0.711). In addition, at the analysis of the secondary endpoints, patients with withdrawal from preoperative ␤-blocker therapy and patients with postoperative need of IABP both appeared to be at higher risk of AF development, although statistical significance was not reached (OR 1.64, 95% CI 0.97 to 2.72, p ⫽ 0.062 and OR 2.29, 95% CI 0.65 to 8.06, p ⫽ 0.198, respectively).

Comment In our prospective observational study, preoperative statin therapy conferred a 42% reduction in risk of AF after CABG (OR 0.58, 95% CI 0.37 to 0.91). This result accounted for the likelihood of the patients being treated with statins, given the presenting characteristics. Other potential predictors of AF were postoperative withdrawal of ␤-blockers and postoperative IABP requirement. On one hand, requirement for IABP in the perioperative period can be assumed to mirror the status of myocardial dysfunction with the need of postoperative inotropic drugs. Catecholamine enhancement triggered activity and automaticity, both key points in the development of postoperative AF [15, 17]. On the other hand, ␤-blockers antagonize the effects of catecholamines on the myocardium and have been studied extensively as prophylactic agents in this setting. Withdrawal from preoperative ␤-blocker therapy has largely been postulated as causing postoperative AF [1, 2, 15, 18]. The beneficial statin effect on AF has been described in different reports. Statins have been demonstrated protec-

Ann Thorac Surg 2007;84:1158 – 65

tive against AF in patients with stable coronary disease [19]. They also decreased the AF recurrence after external cardioversion and the incidences of death, unstable angina, and cardiac arrhythmias [20, 21]. Two recent studies observed a significant reduction in AF risk after cardiac surgery in patients with preoperative statin use [6, 7]. Similar results with a threefold decrease in the odds of AF have been confirmed after noncardiac thoracic surgery [22]. The atorvastatin for reduction of myocardial dysrhythmia after cardiac surgery study (ARMYDA-3) is the only randomized, controlled trial of pretreatment with statins in elective cardiac surgery patients [8]. In this study, Patti and colleagues [8] randomized 200 patients to preoperative atorvastatin (40 mg/day) or placebo starting seven days before operation, observing a significant decrease in postoperative AF in the atorvastatin-treated group (35% vs 57%, p ⫽ 0.003), with a 61% reduction in AF risk at the multivariate level. However, the protective mechanisms exerted by statins remain unclear. The benefit of preoperative statins may reflect their antiinflammatory effects and the role of inflammation in the pathophysiology of postoperative AF has been widely demonstrated [1, 3, 4]. As a matter of fact, inflammatory markers as cytokine or circulating adhesion molecules are attenuated in patients receiving statins after cardiac operations [5, 23]. Antioxidant effect, direct antiarrhythmic effects by cell membrane ion channel stabilization or direct protection of the ischemic myocardium, and modulation of extracellular matrix are also reported properties [24, 25]. Moreover, the beneficial effects of statin therapy are not limited to patients with dyslipidemia, but the incidence of adverse cardiovascular outcomes and the need for coronary revascularization are reduced even in patients with normal cholesterol levels [26]. The protective effects of statins on postoperative AF are valuable because this arrhythmia is the most frequent complication after cardiac surgery and it is associated with increased mortality and expenditure [1, 2]. In light of the importance of postoperative AF to patient outcome, there has been a great deal of interest in preventing this complication pharmacologically. Amiodarone, ␤-blocker agents, and sotalol have been found to reduce but not abolish this postoperative arrhythmia [18]. Several patients still experience AF despite prophylactic therapy, and a large number of the remaining patients are unnecessarily exposed to such drugs with potential adverse events [18]. Moreover, patients with deteriorated left ventricular function (LVF), COPD, and renal failure are frequently excluded from a prophylactic approach with part of the drugs; nevertheless, these patients are the ones most affected by AF, as demonstrated in several studies, and would benefit the most from a preventive treatment [18]. Statin therapy has a low cost, is a low risk preventive AF strategy, and all cardiac surgery patients may benefit from such prophylactic regimen. In our study, however, no AF effects were observed in LOS in the entire population study nor in patients with preoperative statins. This discrepancy, compared with the ARMYDA-3 study could be justified by the inclusion criteria [8]. In our study, emergency operations were

included. As a matter of fact, without considering unstable cases (23.5%), patients affected by postoperative AF experienced a significantly longer hospital stay (7.9 ⫾ 2.7 vs 7.2 ⫾ 2.1 days, p ⫽ 0.025). Patients with statin therapy also revealed a shorter, but not significant LOS compared with patients without it (7.3 ⫾ 2.2 vs 7.6 ⫾ 2.6 days). Our study was an observational investigation based on prospectively collected data, and a selection bias may have been presented because there was no homogenous indication for the use, type of lipid-lowering agents, and duration of preoperative statin therapy. Indication and type of statins were decided by the referring physicians. However, our analysis accounted for this by assessing the role of statins while stratifying on the propensity score. Moreover, the two groups of patients, statin and no statin use, were comparable in preoperative and operative characteristics. The two groups were also reasonably balanced in the absolute number of cases. Furthermore, our aim was to explore only the statin effects on postoperative AF, without modify at all preoperative drug regimens. Our peripheral cardiac surgery department does not permit a preventive elective AF prophylactic strategy as reported by Patti and colleagues [8], with no possibilities to hospitalize patients several days before the operation (seven days as in the ARMYDA-3 study; our hospitalization before the operation was 2.4 ⫾ 1.2 days). But more importantly, emergent cases (23.5% of our patients) are inevitably excluded from such therapy. The latest are most often older patients, suffering from deteriorated LVF and requiring inotropic drugs with prolonged hospital stay; both AF risk factors as demonstrated in several studies [1, 2, 14, 27]. Moreover, in all studies regarding the statin effect on postoperative AF, patients with a preoperative history of AF were excluded [6, 8]. As largely demonstrated, these patients most often are at higher risk of developing postoperative AF [1, 2, 4]. In our study we did not introduce any exclusion criteria as we aimed to observe the highest possible proportion of patients admitted for CABG. As a matter of fact, only 9% of the subjects were not included in the final analysis, reflecting the complete CABG patient population. On the contrary, in the ARMYDA study about 40% (123 of 323) of hospitalized patients were excluded [8]. Finally, we believe that observational studies like ours are still needed after the publication of controlled clinical trials. Indeed, results of controlled clinical trials may differ from results obtained in clinical practice. Generally, randomized trials involve selected centers only and frequently include younger, selected elective patients, with a lower incidence of comorbidities. Discrepancies between controlled trials and clinical practice may be more pronounced when invasive procedures are involved. An operator’s technical skill and careful patient follow-up may enhance the benefits of the procedure in the context of the clinical trial. Therefore, confirmation of the trial results by observational studies is eagerly needed. This study, conducted in an unselected population, aimed at that purpose. A further limitation of this study is the lack of the baseline assessment of inflammatory mark-

MARISCALCO ET AL STATINS AND POSTOPERATIVE ATRIAL FIBRILLATION

1163

ers, which have been associated with postoperative AF [1, 3, 23]. In conclusion, preoperative statins seem to reduce postoperative AF development after CABG and constitute a useful preventive pharmacologic approach in patients undergoing elective revascularization.

References 1. Hogue CW Jr, Creswell LL, Gutterman DD, Fleisher LA. Epidemiology, mechanisms, and risks: American College of Chest Physicians guidelines for the prevention and management of postoperative atrial fibrillation after cardiac surgery. Chest 2005;128(2 suppl):9S–16. 2. Mathew JP, Fontes ML, Tudor IC, et al. A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA 2004;291:1720 –29. 3. Tselentakis EV, Woodford E, Chandy J, Gaudette GR, Saltman AE. Inflammation effects on the electrical properties of atrial tissue and inducibility of postoperative atrial fibrillation. J Surg Res 2006;135:68 –75. 4. Koch CG, Li L, Van Wagoner DR, Duncan AI, Gillinov AM, Blackstone EH. Red cell transfusion is associated with an increased risk for postoperative atrial fibrillation. Ann Thorac Surg 2006;82:1747–56. 5. Chello M, Patti G, Candura D, et al. Effects of atorvastatin on systemic inflammatory response after coronary bypass surgery. Crit Care Med 2006;34:660 –7. 6. Marin F, Pascual DA, Roldan V, et al. Statins and postoperative risk of atrial fibrillation following coronary artery bypass grafting. Am J Cardiol 2006;97:55– 60. 7. Ozaydin M, Dogan A, Varol E, et al. Statin use before by-pass surgery decreases the incidence and shortens the duration of postoperative atrial fibrillation. Cardiology 2006; 107:117–21. 8. Patti G, Chello M, Candura D, et al. Randomized trial of atorvastatin for reduction of postoperative atrial fibrillation in patients undergoing cardiac surgery: results of the ARMYDA-3 (atorvastatin for reduction of myocardial dysrhythmia after cardiac surgery) study. Circulation 2006;114: 1455– 61. 9. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med 1997;157:2413– 46. 10. Davidson MH, Maki KC, Pearson TA, et al. Results of the National Cholesterol Education (NCEP) Program Evaluation ProjecT Utilizing Novel E-Technology (NEPTUNE) II survey and implications for treatment under the recent NCEP Writing Group recommendations. Am J Cardiol 2005;96:556 – 63. 11. Aguilar-Salinas CA, Garcia-Garcia E, Lerman-Garber I, Gomez-Perez FJ, Rull JA. Making things easier is not so easy. The 1997 American Diabetes Association criteria and glucose intolerance. Diabetes Care 1998;21:1027– 8. 12. Campeau L. Grading of angina pectoris (Letter). Circulation 1976;54:522–3. 13. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16: 9 –13. 14. Mariscalco G, Engstrom KG, Ferrarese S, et al. Relationship between atrial histopathology and atrial fibrillation after coronary bypass surgery. J Thorac Cardiovasc Surg 2006;131: 1364 –72. 15. Kalman JM, Munawar M, Howes LG, et al. Atrial fibrillation after coronary artery bypass grafting is associated with sympathetic activation. Ann Thorac Surg 1995;60:1709 –15. 16. Austin PC, Mamdani MM. A comparison of propensity score methods: a case study estimating the effectiveness of postAMI statin use. Stat Med 2006;30:2084 –106.

CARDIOVASCULAR

Ann Thorac Surg 2007;84:1158 – 65

1164

MARISCALCO ET AL STATINS AND POSTOPERATIVE ATRIAL FIBRILLATION

CARDIOVASCULAR

17. Waldo AL. Mechanisms of atrial fibrillation, atrial flutter, and ectopic atrial tachycardia--a brief review. Circulation 1987;75(pt 2):III37– 40. 18. Young-Xu Y, Jabbour S, Goldberg R, et al. Usefulness of statin drugs in protecting against atrial fibrillation in patients with coronary artery disease. Am J Cardiol 2003;92: 1379 – 83. 19. Siu CW, Lau CP, Tse HF. Prevention of atrial fibrillation recurrence by statin therapy in patients with lone atrial fibrillation after successful cardioversion. Am J Cardiol 2003; 92:1343–5. 20. Dotani MI, Elnicki DM, Jain AC, Gibson CM. Effect of preoperative statin therapy and cardiac outcomes after coronary artery bypass grafting. Am J Cardiol 2000;86:1128 –30; A1126. 21. Amar D, Zhang H, Heerdt PM, Park B, Fleisher M, Thaler HT. Statin use is associated with a reduction in atrial fibrillation after noncardiac thoracic surgery independent of C-reactive protein. Chest 2005;128:3421–7. 22. Chello M, Carassiti M, Agro F, et al. Simvastatin blunts the increase of circulating adhesion molecules after coronary artery bypass surgery with cardiopulmonary bypass. J Cardiothorac Vasc Anesth 2004;18:605–9. 23. Wright DG, Lefer DJ. Statin mediated protection of the ischemic myocardium. Vascul Pharmacol 2005;42:265–70. 24. Bonetti PO, Lerman LO, Napoli C, Lerman A. Statin effects beyond lipid lowering--are they clinically relevant? Eur Heart J 2003;24:225– 48. 25. Collard CD, Body SC, Shernan SK, Wang S, Mangano DT. Preoperative statin therapy is associated with reduced cardiac mortality after coronary artery bypass graft surgery. J Thorac Cardiovasc Surg 2006;132:392– 400. 26. Burgess DC, Kilborn MJ, Keech AC. Interventions for prevention of post-operative atrial fibrillation and its complications after cardiac surgery: a meta-analysis. Eur Heart J 2006;27:2846 –57. 27. Delaney AP, Lee RP, Kay S, Hansen P. Early invasive revascularisation for patients critically ill after acute myo-

Ann Thorac Surg 2007;84:1158 – 65

cardial infarction: impact on outcome and ICU resource utilisation. Crit Care Resusc 2003;5:258 – 65.

Appendix Additional Analyzed Variables Preoperative (29): age, sex, BMI (body mass index), BSA (body surface area), preoperative cardiac rhythm, CCS angina classification, angina presentation (stable or unstable), IABP (intraaortic balloon pump), previous AMI (acute myocardial infarction), previous PCI (percutaneous coronary intervention), left main stem stenosis, right coronary artery stenosis; left ventricular function, left atrial antero-posterior diameter; hypertension, diabetes mellitus, hypercholesterolemia, BPCO (chronic obstructive pulmonary disease), peripheral vascular disease, cerebrovascular accident, renal failure, hemoglobin (g/dL), EuroSCORE; ␤-blocker agents; calcium-channel blockers, ACE (angiotensin-converting enzyme)-inhibitors; type of statins, dose of statins, duration of preoperative statin therapy. Operative (11): CPB (cardiopulmonary bypass) use, cardioplegia-route (antegrade, antegrade and retrograde), cardioplegiatype (crystalloid or blood), cardioplegia-temperature (warm, cold), CPB time, ACC (aortic cross-clamp) time, CPB temperature, number of grafts, postoperative IABP, cardiac rhythm at the end operation, inotropic support. Postoperative (19): time on ventilator, ICU (intensive care unit) time, mediastinal bleeding, fluid balance, MB peak; respiratory failure, surgical revision for bleeding, perioperative AMI, hemotransfusion; atrial fibrillation (AF) occurrence, AF occurrence day, cardiac rhythm at discharge; LOS (length of hospital stay), hospital mortality, one-year mortality; postoperative: ␤-blocker agents, calcium-channel blockers, ACE-inhibitors, statins.

INVITED COMMENTARY Mariscalco and colleagues [1] report on the beneficial effect of preoperative administration of statin medications in reducing atrial fibrillation (AF) after coronary artery bypass grafting (CABG). This effect was preserved in a multivariate analysis that considered other pertinent preoperative and intraoperative variables. The mechanisms responsible for postoperative AF are not completely understood, but may include ischemia, anatomic changes in the atrium, transient electrophysiologic and hemodynamic changes perioperatively, and inflammation, among others. These arrhythmias likely occur in patients at risk (eg, anatomical substrate) because of a perioperative trigger. The motivation for the current study is the possibility that perioperative inflammation is related to the development of postoperative AF. In cardiac surgery, statins have recently received attention with the observation that pretreatment with atorvastatin reduced cytokine release and neutrophil adhesion in patients undergoing CABG with cardiopulmonary bypass. Two previous nonrandomized studies of preoperative statin therapy in patients undergoing CABG have shown that the incidence and © 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc

duration of postoperative AF were reduced. Most recently, a randomized trial of preoperatively used atorvastatin showed a significant reduction in postoperative AF in the treated group. Interestingly, the benefit of atorvastatin was not enjoyed by patients with left atrial enlargement or those who were undergoing heart valve surgery. The current report involves a relatively small number of patients and it is impossible to know if the study and control groups are similar except for the use of statins [1]. Nonetheless, this study adds to the accumulating evidence that statins may be a useful strategy for helping to prevent postoperative AF. We must keep in mind the entire body of evidence related to the prevention of postoperative AF [2]. We know that postoperative AF is related to preoperative variables, including patient age and type of operation. We also know that mild rather than moderate degrees of hypothermia may be protective. Accumulating evidence suggests that off-pump CABG may be associated with a lower incidence of postoperative AF than conventional CABG. We also know that pharmacologic interventions such as beta-blockers or amiodarone 0003-4975/07/$32.00 doi:10.1016/j.athoracsur.2007.06.102