Atrial fibrillation (AF) is a common arrhythmia after cardiac surgery that usually occurs in the first 2–4 postoperative days in 20%–50% of patients.1–7 Patients who develop postoperative AF have more frequent morbidity and mortality, increased length of hospitalization, and increased cost of care.8–13 It remains unclear whether AF is directly responsible for the magnitude of these adverse outcomes or if AF is merely associated with the underlying causes of these complications. Regardless, a universal finding is that patients developing postoperative AF require longer hospitalizations, and 1 estimate puts the yearly cost associated with postoperative AF in the United States at >$1 billion.14 Thus, measures that prevent this complication may lead to improved patient outcomes.
The pathophysiology of postoperative AF is uncertain and likely multifactorial. The finding of increased levels of components of the complement cascade, white blood cell count, and C-reactive protein concentrations in patients who develop AF compared with patients remaining in sinus rhythm suggests a role of inflammation in the development of this arrhythmia.15–17 Two publications have suggested that perioperative glucocorticoid therapy can decrease the incidence of postoperative AF, presumably by attenuation of inflammation that accompanies cardiopulmonary bypass (CPB) and surgery.2,6 However, this treatment may be associated with complications, and the beneficial effects for the prevention of AF have not been a universal finding.3,6
Hemofiltration during CPB has been shown to remove low-molecular-weight molecules from the plasma including inflammatory mediators.18–20 The authors have previously performed a randomized, double-blind, placebo-controlled trial to identify whether perioperative steroids or hemofiltration during CPB, by attenuating inflammation, would reduce the duration of mechanical ventilation after cardiac surgery.21 In that study, hemofiltration decreased the time to tracheal extubation versus no intervention, but perioperative steroid therapy had no effect on extubation times compared with controls.21
Given the reported association between inflammation and AF after cardiac surgery, the fact that hemofiltration during CPB removes at least some inflammatory mediators, as well as our previous positive findings of hemofiltration in reducing times to tracheal extubation, we hypothesized that hemofiltration during CPB may have benefit in preventing AF after cardiac surgery.
We performed a retrospective review of patients who had previously been enrolled in a prospective, randomized, double-blind, placebo-controlled trial assessing the clinical benefits of steroid treatment versus hemofiltration in patients undergoing cardiac surgery with CPB.
After IRB approval and written informed consent, 192 adult men and nonpregnant women undergoing elective primary coronary artery bypass grafting (CABG) and/or valvular replacement or repair requiring CPB were enrolled. Subsequently, we obtained IRB approval to reexamine the patients' written charts and electronic records to determine whether postoperative AF occurred. Patients were excluded from enrollment in the original study for presence of congenital heart disease, left ventricular ejection fraction ≤0.35, severe pulmonary disease, previous difficult endotracheal intubation, pulmonary hypertension, neurologic deficits or disease, serum creatinine ≥2.0 mg/dL, recent or long-term steroid usage, insulin-dependent diabetes mellitus, or age ≥85 yr. In the retrospective analysis, patients with chronic AF were excluded.
Patients were randomized in a double-blind manner to 1 of 3 groups: Group I received placebo (saline) and no hemofiltration; Group II received 1 g of methylprednisolone IV immediately before induction of anesthesia followed by 4 mg of dexamethasone IV every 6 h during the next 24 h and no hemofiltration; and Group III received placebo (saline) and underwent hemofiltration during CPB. Hemofiltration was conducted with a Pro-Tec™ ultrafiltration device (Pall Biomedical Products, New York, NY) until 27 mL/kg of ultrafiltrate was removed.22 If the patient's blood volume was inadequate to obtain 27 mL/kg of ultrafiltrate, lactated Ringer's solution and/or colloid solutions were added to the CPB to obtain adequate volume in the venous reservoir for hemofiltration. Perfusionists were aware of subjects assigned to Group III but no other study assignment. All remaining operating room and intensive care unit (ICU) personnel were strictly blinded to group identity.
All patients received an anesthetic consisting of 25 μg/kg of fentanyl, 0.1 mg/kg of midazolam, isoflurane, and muscle relaxant. Standard monitoring for cardiac anesthesia and surgery was used, including arterial and central venous catheterization. Lactated Ringer's solution and colloid solutions were administered to maintain adequate circulating blood volume and hemodynamic stability.
Nonpulsatile CPB was conducted with a Univox membrane oxygenator (Bentley, Irvine, CA) and a Sarns 9000 CPB machine using a roller head (Sarns, Ann Arbor, MI). The circuit was primed with 2 L of Plasmalyte® solution (Baxter, Deerfield, IL), 25 g mannitol 12.5%, and 50 mEq sodium bicarbonate. All patients were expected to undergo normothermic CPB (>35°C) monitored by nasopharyngeal temperature. If hypothermic CPB was instituted, the patient's participation in the study was terminated and the data were not included in the analysis. Perfusion flow rates were maintained at 2.0–2.4 L·min−1·m−2. Mean arterial blood pressure was maintained at 50–80 mm Hg during CPB. Hyperkalemic cold blood cardioplegia was given every 20–30 min for myocardial protection. Anticoagulation was initiated with 300 U/kg of heparin (intestine porcine mucosal) (Elkins-Sinn, Cherry Hill, NJ) and maintained according to an activated clotting time (Hemochron 801, International Technidyne, Edison, NJ) of >480 s.
The plasma glucose concentration was measured from the arterial blood immediately after induction of anesthesia, at the onset of CPB, and upon final separation from CPB. Additional samples were obtained as needed to maintain the plasma glucose between 100 and 200 mg/dL. A plasma glucose level ≥250 mg/dL was treated with 5 U of regular insulin IV. Additional 5-U doses of insulin were given until the plasma glucose was ≤200 mg/dL.
After separation from CPB, heparin neutralization with protamine, and attainment of stable hemodynamics, a propofol infusion (25–50 μg·kg−1·min−1) was begun and continued in the ICU for sedation and until tracheal extubation. On arrival in the ICU, patients were assessed for evidence of neuromuscular blockade with train-of-four monitoring. Residual neuromuscular blockade was reversed with 0.05 mg/kg of neostigmine and 0.01 mg/kg of glycopyrrolate. Morphine sulfate was given for pain based on clinical assessment by the ICU nurses.
For the purposes of this retrospective analysis, all written hospital charts and electronic records were examined by one of the authors (WJM) who was blinded to the patient's treatment group. Patients were considered to have had AF if there was any electrocardiogram (ECG) with an irregular ventricular rhythm and absence of P waves or if AF was noted in any of the daily progress notes or hospital discharge summary.
Patient and procedural characteristics were compared across treatment groups using analysis of variance for continuous variables and the χ2 test (or Fisher's exact test) for categorical variables. Logistic regression was used to assess whether the incidence of AF was associated with randomized treatment. In addition to an unadjusted analysis, multiple logistic regression analysis was performed to assess whether the risk for AF differed across treatment groups after adjusting for age (treated as a continuous variable), type of procedure (valve versus no valve), and chronic obstructive pulmonary disease (none versus mild or moderate). These variables were selected based on previously reported associations with postoperative AF.23 For these analyses, the effect of treatment group was assessed using a 2 degrees of freedom χ2 test. Odds ratios (ORs) and corresponding 95% confidence intervals (CIs) were calculated to estimate the increased (or decreased) likelihood of AF for the methylprednisolone and hemofiltration groups compared with the placebo group. Other patient and procedural characteristics were assessed for a potential association with AF using logistic regression. In all cases, 2-sided tests were used with P values ≤0.05 considered significant. The sample size for this trial was based on a power analysis for detecting a difference among groups for the primary end point of time to tracheal extubation.
One hundred ninety-two patients were enrolled in the original trial but 3 patients did not complete the study because of protocol violations. In the subsequent analysis of AF, 4 patients were excluded on the basis of preexisting AF. Thus, 185 patients were available for this retrospective analysis. Of these patients, 92 (49.7%) underwent CABG alone, 85 (45.9%) underwent valve repair/replacement alone, and 8 (4.3%) underwent combined CABG and valve surgery. There were no differences among groups in demographics or surgical characteristics (Table 1).
Sixty patients (32%) developed postoperative AF. The incidence of AF did not differ across groups (P = 0.057). In a supplemental analyses that did not adjust for multiple comparisons, the incidence of AF was higher in the steroid and hemofiltration groups compared with the control group (OR = 2.59, 95% CI 1.16–5.77, P = 0.020 for steroid group; OR = 2.11, 95% CI 0.95–4.69, P = 0.065 for hemofiltration group) (Table 2, Fig. 1). Similar findings were obtained from an analysis that adjusted for other characteristics associated with AF (treatment comparison P = 0.108) (Table 2). From this secondary analysis, the only risk factor associated with the development of AF was age (65.4 ± 10.1 yr versus 61.4 ± 11.5 yr for patients with and without AF, respectively; P = 0.024) (Table 3). Gender, procedure type, smoking status, body mass index, and the duration of aortic cross-clamping, CPB, surgery, or anesthesia were not associated with the incidence of postoperative AF (Table 3).
Blood glucose concentrations obtained before steroid or placebo administration, during CPB, and after separation from CPB were higher for patients receiving steroids versus placebo (127.5 ± 23.5 mg/dL vs 117.1 ± 21.2 mg/dL, respectively; P = 0.008) and after CPB (161.0 ± 32.2 mg/dL vs 135.7 ± 28.8 mg/dL, respectively; P < 0.001). Plasma glucose concentrations were not different for patients receiving steroids compared with patients undergoing hemofiltration at any of these time periods.21
Recently, there has been interest in the role of systemic inflammation in the development of AF after cardiac surgery. In the current study, neither hemofiltration during CPB nor perioperative steroid therapy decreased the incidence of AF after cardiac surgery compared with placebo.
A relationship between AF after cardiac surgery and increased C-reactive protein levels, white blood cell counts, and concentrations of inflammatory cytokines, including interleukin-6, and complement activation has been reported.15–17 Cardiac surgery using CPB results in systemic inflammatory response syndrome ending in complement activation, proinflammatory cytokine release, and neutrophil activation.24 Hemofiltration effluent has been shown to contain tumor necrosis factor-α, interleukin-6, C3a, and C5a complexes.18,20,25 In this study, we used an established protocol of hemofiltration as described by Andreasson et al.22 and found no benefit in reducing the incidence of postoperative AF. These findings are corroborated by a previously performed study. Lucas et al.26 retrospectively reviewed the records of 175 patients undergoing CABG and found that CPB hemofiltration had no influence on the incidence of postoperative AF. On univariate analysis, hemofiltration was more frequently performed in patients who developed AF (71%) than in those who did not (53%) (P = 0.0468). However, hemofiltration was performed more frequently and more aggressively in patients with renal insufficiency, which may have confounded their results.
It is not surprising that intraoperative hemofiltration did not alter the incidence of AF in this study. AF typically occurs 48–96 h after surgery, and a short period of attenuating the inflammatory cascade during the operation probably has little benefit in preventing more remote complications. In fact, it has been shown that C-reactive protein levels do not peak until postoperative day 2, and complement C3b/c levels undergo a secondary elevation between postoperative days 2 and 4.15,17 This late spike in the levels of inflammatory mediators likely further negates the effects of a therapy confined to the intraoperative period.
In our study, steroid treatment was associated with a higher incidence of postoperative AF but the CIs were wide (unadjusted OR = 2.59, 95% CI 1.16–5.77; adjusted OR = 2.36, 95% CI 1.04–5.37) (Table 2). A lack of a benefit of steroids for reducing the frequency of postoperative AF contrasts with the findings of 2 randomized trials and a meta-analysis. Halonen et al.2 showed that 100 mg of hydrocortisone administered the evening of the operative day and then every 8 h for 72 h in patients undergoing CABG, aortic valve, or combined CABG and aortic valve surgery reduced the frequency of postoperative AF from 48% to 30% (P = 0.004). In that study, all patients received oral metoprolol titrated to heart rate, which might have influenced their results.1,2 Prasongsukarn et al.6 randomized 88 patients undergoing CABG surgery to placebo or 1 g of methylprednisolone before CPB and dexamethasone 4 mg every 6 h for 1 day after surgery or placebo. Steroid treatment decreased the incidence of AF from 51% in the controls to 21% in the steroid group patients (P = 0.003). Minor complications (gastrointestinal disturbance, hyperglycemia, mental confusion, etc.) were more common in the group receiving steroids.6 A recent meta-analysis of 14 trials showed that perioperative steroid use reduced the frequency of postoperative AF compared with placebo (relative risk 0.71, 95% CI 0.59–0.87, P = 0.001).27 The use of other medications including β-blockers was not controlled for in this analysis.
There are several potential reasons why we failed to find a decreased incidence of postoperative AF in patients receiving steroids compared with patients in other studies. First, the study by Prasongsukarn et al.6 included only CABG patients and the study by Halonen et al.2 included only patients undergoing CABG aortic valve or combined surgery. Our patient group was more heterogeneous and included patients undergoing mitral valve surgery. Second, our dose and duration of steroid treatment was less than in the study by Halonen et al.2 Third, we may have underestimated the frequency of postoperative AF relative to the other trials because we relied on intermittent ECGs and clinically detected AF. In contrast, in the studies by Halonen et al.2 and Prasongukarn et al.,6 patients were monitored with continuous telemetry ECG that was examined daily for AF. In fact, the frequency of AF in our control group was less than that reported in the prospective studies by Halonen et al.2 and Prasongsukarn et al.6 (21% vs 48% and 51%, respectively) but similar to that found in other larger retrospective studies.23
In the report by Prasongsukarn et al.,6 although steroid treatment was associated with a lower incidence of AF, there was no difference in the duration of hospitalization between treatment and control groups. Duration of hospitalization was not reported in the study by Halonen et al.2 Prasonsukarn et al.6 reported a higher rate of complications including gastrointestinal disturbance, hyperglycemia, mental confusion, etc. in patients who received steroids compared with placebo. These studies have inadequate power to detect potential catastrophic complications such as deep sternal wound infections that occur in <2% of patients. The routine use of steroids for preventing postoperative AF would subject 50%–70% of patients who would not have developed AF to these potential complications without any chance of benefit.
Given the small sample size for this study, the statistical power for detecting differences in the rate of AF between treatment groups may be limited. However, given the direction of the observed effects (OR = 2.59, 95% CI 1.16–5.77 for steroid group; OR = 2.11, 95% CI 0.95–4.69 for hemofiltration group), we believe our findings are consistent with the fact that neither steroids nor hemofiltration alter the frequency of AF after cardiac surgery. In addition, our data set did not include information on the use of medications associated with the occurrence or prevention of AF. Specifically, perioperative withdrawal of β-blockade and angiotensin-converting enzyme inhibitors has been associated with AF, and the perioperative institution of β-blockade and statins has been associated with a reduction in the frequency of AF.1,23,28 However, given the randomized nature of our study, we believe it is unlikely that the use of these medications differed significantly among the study groups.
In this study, hemofiltration during CPB did not decrease the incidence of AF after cardiac surgery. In contrast to previous reports, we found no benefit of perioperative steroids in reducing postoperative AF. In the current study, the data suggest a possible increase in the incidence of AF when patients receive perioperative steroids but the CIs are wide, and this result should be viewed with caution.
1.Halonen J, Hakala T, Auvinen T, Karjalainen J, Turpeinen A, Uusaro A, Halonen P, Hartikainen J, Hippelainen M. Intravenous administration of metoprolol is more effective than oral administration in the prevention of atrial fibrillation after cardiac surgery. Circulation 2006;114:I1–4
2.Halonen J, Halonen P, Jarvinen O, Taskinen P, Auvinen T, Tarkka M, Hippelainen M, Juvonen T, Hartikainen J, Hakala T. Corticosteroids for the prevention of atrial fibrillation after cardiac surgery: a randomized controlled trial. JAMA 2007;297: 1562–7
3.Halvorsen P, Raeder J, White PF, Almdahl SM, Nordstrand K, Saatvedt K, Veel T. The effect of dexamethasone on side effects after coronary revascularization procedures. Anesth Analg 2003;96:1578–83
4.Mahoney EM, Thompson TD, Veledar E, Williams J, Weintraub WS. Cost-effectiveness of targeting patients undergoing cardiac surgery for therapy with intravenous amiodarone to prevent atrial fibrillation. J Am Coll Cardiol 2002;40:737–45
5.Majahalme S, Kim MH, Bruckman D, Tarkka M, Eagle KA. Atrial fibrillation after coronary surgery: comparison between different health care systems. Int J Cardiol 2002;82:209–18; discussion 218–9
6.Prasongsukarn K, Abel JG, Jamieson WR, Cheung A, Russell JA, Walley KR, Lichtenstein SV. The effects of steroids on the occurrence of postoperative atrial fibrillation after coronary artery bypass grafting surgery: a prospective randomized trial. J Thorac Cardiovasc Surg 2005;130:93–8
7.Zaman AG, Archbold RA, Helft G, Paul EA, Curzen NP, Mills PG. Atrial fibrillation after coronary artery bypass surgery: a model for preoperative risk stratification. Circulation 2000; 101:1403–8
8.Almassi GH, Schowalter T, Nicolosi AC, Aggarwal A, Moritz TE, Henderson WG, Tarazi R, Shroyer AL, Sethi GK, Grover FL, Hammermeister KE. Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg 1997;226:501–11; discussion 511–3
9.Aranki SF, Shaw DP, Adams DH, Rizzo RJ, Couper GS, VanderVliet M, Collins JJ Jr, Cohn LH, Burstin HR. Predictors of atrial fibrillation after coronary artery surgery. Current trends and impact on hospital resources. Circulation 1996;94:390–7
10.Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56: 539–49
11.Hakala T, Pitkanen O, Hippelainen M. Feasibility of predicting the risk of atrial fibrillation after coronary artery bypass surgery with logistic regression model. Scand J Surg 2002;91:339–44
12.Lahtinen J, Biancari F, Salmela E, Mosorin M, Satta J, Rainio P, Rimpilainen J, Lepojarvi M, Juvonen T. Postoperative atrial fibrillation is a major cause of stroke after on-pump coronary artery bypass surgery. Ann Thorac Surg 2004;77:1241–4
13.Tamis JE, Steinberg JS. Atrial fibrillation independently prolongs hospital stay after coronary artery bypass surgery. Clin Cardiol 2000;23:155–9
14.Steinberg JS. Postoperative atrial fibrillation: a billion-dollar problem. J Am Coll Cardiol 2004;43:1001–3
15.Bruins P, te Velthuis H, Yazdanbakhsh AP, Jansen PG, van Hardevelt FW, de Beaumont EM, Wildevuur CR, Eijsman L, Trouwborst A, Hack CE. Activation of the complement system during and after cardiopulmonary bypass surgery: postsurgery activation involves C-reactive protein and is associated with postoperative arrhythmia. Circulation 1997;96:3542–8
16.Lamm G, Auer J, Weber T, Berent R, Ng C, Eber B. Postoperative white blood cell count predicts atrial fibrillation after cardiac surgery. J Cardiothorac Vasc Anesth 2006;20:51–6
17.Ishida K, Kimura F, Imamaki M, Ishida A, Shimura H, Kohno H, Sakurai M, Miyazaki M. Relation of inflammatory cytokines to atrial fibrillation after off-pump coronary artery bypass grafting. Eur J Cardiothorac Surg 2006;29:501–5
18.Jois D, Israel-Biet D, Pouard P, Rolland B, Silvester W, Vouhe P, Safran D. High-volume, zero-balanced hemofiltration to reduce delayed inflammatory response to cardiopulmonary bypass in children. Anesthesiology 1996;85:965–76
19.Laffey JG, Boylan JF, Cheng DC. The systemic inflammatory response to cardiac surgery: implications for the anesthesiologist. Anesthesiology 2002;97:215–52
20.Tallman RD, Dumond M, Brown D. Inflammatory mediator removal by zero-balance ultrafiltration during cardiopulmonary bypass. Perfusion 2002;17:111–5
21.Oliver WC Jr, Nuttall GA, Orszulak TA, Bamlet WR, Abel MD, Ereth MH, Schaff HV. Hemofiltration but not steroids results in earlier tracheal extubation following cardiopulmonary bypass: a prospective, randomized double-blind trial. Anesthesiology 2004;101:327–39
22.Andreasson S, Gothberg S, Berggren H, Bengtsson A, Eriksson E, Risberg B. Hemofiltration modifies complement activation after extracorporeal circulation in infants. Ann Thorac Surg 1993;56:1515–7
23.Mathew JP, Fontes ML, Tudor IC, Ramsay J, Duke P, Mazer CD, Barash PG, Hsu PH, Mangano DT. A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA 2004;291:1720–9
24.Hall RI, Smith MS, Rocker G. The systemic inflammatory response to cardiopulmonary bypass: pathophysiological, therapeutic, and pharmacological considerations. Anesth Analg 1997;85:766–82
25.Millar AB, Armstrong L, van der Linden J, Moat N, Ekroth R, Westwick J, Scallan M, Lincoln C. Cytokine production and hemofiltration in children undergoing cardiopulmonary bypass. Ann Thorac Surg 1993;56:1499–502
26.Lucas MT, Limoli J, Schlut J, Holt DW. Does zero balance ultrafiltration decrease the incidence of post-cardiopulmonary bypass atrial fibrillation? J Extra Corpor Technol 2008;40:109–15
27.Whitlock RP, Chan S, Devereaux PJ, Sun J, Rubens FD, Thorlund K, Teoh KH. Clinical benefit of steroid use in patients undergoing cardiopulmonary bypass: a meta-analysis of randomized trials. Eur Heart J 2008;29:2592–600
28.Liakopoulos OJ, Choi YH, Kuhn EW, Wittwer T, Borys M, Madershahian N, Wassmer G, Wahlers T. Statins for prevention of atrial fibrillation after cardiac surgery: a systematic literature review. J Thorac Cardiovasc Surg 2009;138:678–86.e1