Arrhythmias and conduction abnormalities after coronary artery bypass graft (CABG) surgery are a common clinical problem. Atrial fibrillation is the most common arrhythmia with a reported incidence ranging from 10% to 50%(1–3). Postoperative atrial fibrillation is an important cause of morbidity and increased hospital length of stay (4–6), which in turn, results in increased costs of medical care and resource utilization.
The pathophysiology of postoperative atrial fibrillation is multifactorial. Previous studies of preoperative risk factors have yielded conflicting results, with advanced age being the only risk factor consistently identified (7). Intraoperative ischemic insult to the atria is thought to be an important predisposing factor for postoperative atrial arrhythmias. The atria are frequently electromechanically active during aortic cross-clamping and cardioplegic delivery, implying that atrial preservation is suboptimal (8,9). Such atrial activity during ischemic periods leads to anaerobic glycolysis, with subsequent lactate and acid accumulation. Perioperative ischemia and lactic acidosis can cause electrical instability, leading to postoperative arrhythmias (8,9). Other possible etiologic factors for post-CABG atrial fibrillation include pericarditis, excessive production of catecholamines, and intravascular volume changes (7).
Despite progressive improvements in CABG surgery results, patients with unstable angina and recent infarction continue to have increased morbidity and mortality (10). The increased risk may be a result of increased myocardial injury, which contributes to decreased ventricular function and an increased incidence of postoperative arrhythmias and low cardiac output syndrome (11,12). High-risk patients may undergo periods of ischemic metabolism while waiting for surgery and may therefore benefit from a metabolic intervention.
Several experimental studies have demonstrated the beneficial effect of glucose and insulin infusion on myocardial performance during CABG surgery (13–15), presumably via restoration of myocardial energy supplies. However, these studies have not focused on the effect of insulin on perioperative arrhythmias. In addition, the glucose-insulin infusion used in these studies has not been given as a part of the cardioplegic solution during the ischemic period.
We have previously demonstrated that insulin improves myocardial aerobic metabolism and increases post-CABG ventricular function (16,17). We hypothesized that improved aerobic metabolism will mitigate the deleterious effects of intraoperative atrial ischemia and thus decrease the risk of postoperative atrial arrhythmias. We therefore performed a randomized, double-blinded, placebo-controlled study to determine whether insulin-enhanced cardioplegia would decrease the rate of postoperative atrial fibrillation in a high-risk patient population.
Patients were part of a continuing trial assessing the effect of insulin cardioplegia on postoperative myocardial infarction and low cardiac output syndrome, and therefore, these endpoints are not presented in this article. A sample size of 500 patients was chosen in order to obtain a high power (> 99%) of detecting a 30% relative risk reduction in atrial fibrillation, based on an expected incidence of 30% and an α-level of 0.05.
All patients who were admitted to a coronary care unit after an ischemic episode (angina or infarction) and required urgent CABG during the same hospitalization were eligible to participate in this study. Eligible patients included those with postinfarct angina, Canadian Cardiovascular Society class IV angina, angina with rapidly increasing intensity, and severe left main disease. A total of 501 patients operated on in one university-affiliated center between December 1995 and October 1997 were included in the current study. Patients who required simultaneous valvular or extracardiac procedures, left ventricle (LV) aneurysm resection, or redo sternotomy were excluded. The study protocol was approved by our IRB, and all patients gave written, informed consent to participate.
All cardiac medications including β-adrenergic blockers were continued until the morning of surgery. Patients were premedicated with either sublingual lorazepam (1–3 mg) or IM morphine (5–10 mg) and perphenazine (2.5–5 mg) according to the anesthesiologist’s preference. Pulmonary artery catheters and arterial cannulas were used for perioperative monitoring. Patients were anesthetized by using a small-dose fentanyl regimen (10–15 μg/kg), aiming to early tracheal extubation. Isoflurane (0.4%–1.5%) and midazolam (up to 0.1 mg/kg) were used for maintenance of anesthesia. Pancuronium was used to facilitate tracheal intubation. Propofol infusion (2–4 mg · kg−1 · h−1) was started at the beginning of the cardiopulmonary bypass (CPB) and was continued 2–4 h postoperatively to keep patients sedated until they were warm and hemodynamically stable.
Surgery and Myocardial Protection
All patients underwent median sternotomy and were placed on CPB. A single, two-staged right atrial cannula was used for venous drainage. Left atrial venting was not performed. Hematocrit was maintained between 20% and 25% during CPB, pump flow rates between 2.0 and 2.5 L · min−1 · m−2, and mean arterial pressure between 60 and 80 mm Hg. Systemic body temperature was allowed to drift to 34°C, with active rewarming at the end of CPB. All anastomoses were completed during a single aortic cross-clamp period. Cold (10°C) or tepid (29°C) antegrade or retrograde blood cardioplegia was used for myocardial protection according to the surgeon’s preference. Our cardioplegic solution has been previously described (18), consisting of oxygenated blood mixed with crystalloid in an 8:1 ratio to achieve a final concentration of 6 mEq/L of magnesium sulfate, 50 mmol/L of glucose, and either a small (8 mEq/L) or large (27 mEq/L) concentration of potassium chloride. All patients received an initial infusion of the large potassium solution followed by maintenance with the small potassium formulation.
Patients were randomized to receive either standard blood cardioplegia (Control group) or standard blood cardioplegia enhanced with insulin (Insulin group). Patients who were randomized to the Placebo (Control) group had an equivalent volume of inactive diluent added to each bag of crystalloid cardioplegia. Patients randomized to the Insulin group had human insulin (Humulin R; Eli Lilly, Mississauga, Canada) added to the crystalloid solution to achieve a final concentration of 10 IU/L in the blood cardioplegic mixture.
Randomization was via an opaque sealed envelope opened at the time of anesthetic induction and was stratified by surgeon to account for differences in myoprotective strategies. The surgeon, anesthesiologist, and perfusionist were blinded to patient group assignment.
All patients were transferred postoperatively to the intensive care unit (ICU). Propofol infusions were discontinued in the ICU when patients were hemodynamically stable, were not bleeding excessively (>100 mL/h), and had reached a body temperature of 36.5°C. Patients’ tracheas were routinely extubated when they were awake, were following commands, were able to create a negative inspiratory force of -20 cm H2O, and had a vital capacity of >10 mL/kg (19). In hemodynamically unstable patients, sedation was continued as a morphine infusion supplemented with midazolam when needed.
Dopamine was used as a first line inotrope if the cardiac index was <2.0 L · min−1 · m−2 or if the systolic blood pressure was <90 mm Hg after optimization of filling pressures. More severe hemodynamic compromise led to the initiation of epinephrine, norepinephrine, dobutamine, or milrinone as indicated. Patients requiring sustained inotropic support were considered candidates for intraaortic balloon pump insertion (11). Stable patients were transferred to the surgical floor on the first postoperative day. Patients were started on β-adrenergic blockers 1 day postoperatively only if they were receiving these medications preoperatively.
Patients were continuously monitored by telemetry to a central nursing station monitor (Hewlett Packard M2350A; Andover, ME) while in the ICU and nursing ward. Monitors were continuously observed by a trained nurse and were set to alarm for all common arrhythmias. Telemetry was discontinued in patients who remained in sinus rhythm 48 h after transfer to the surgical ward, and only symptomatic arrhythmias were recorded thereafter. A 12-lead electrocardiogram was collected immediately after surgery and on the first, third, and fifth postoperative days. A single physician (MH) who was blinded to patient group assignment reviewed all electrocardiographic data.
Atrial fibrillation was defined as an irregular rhythm with no organized atrial activity persisting >30 s. Atrial flutter always occurred in association with atrial fibrillation and, therefore, was recorded as atrial fibrillation. Ventricular tachycardia was defined as six or more consecutive, regular, wide-complex beats at a rate of >120/min. Intracardiac conduction defects were defined as wide-complex beats with appropriate corresponding morphology. Patients were considered to have an arrhythmia if it occurred any time in the postoperative period, regardless of whether treatment was instituted.
Data were expressed as mean ± sd for continuous variables and as percentages for categorical variables. Univariate analyses were performed by using the χ2 test (or Fisher’s exact test where appropriate) for categorical variables and Student’s t-test (or Wilcoxon ranked sum test where appropriate) for continuous variables. Stepwise logistic regression models included all variables suggested by the univariate analyses (P < 0.25) or those judged to be clinically important. The model with the best Hosmer-Lemeshow goodness-of-fit statistic and receiver operator characteristic curve was chosen, as previously described (11). All analyses were performed by using the SAS program (SAS Institute, Cary, NC).
Of the 501 patients, 243 were randomized to the Insulin Treatment group and 258 patients to the Control group. Preoperative clinical characteristics of these patients are shown in Table 1. The two groups of patients were similar for all preoperative variables, with the exception of an increased prevalence of LV dysfunction in the Insulin group (Table 1).
Intraoperative variables for the two groups are displayed in Table 2. There were no statistically significant differences between the two groups of patients for any of the intraoperative characteristics listed, with the exception of peak serum glucose and potassium levels.
Postoperative variables were also similar for the two groups of patients including time to tracheal extubation (8.8 ± 6.7 vs 9.4 ± 7.1 h for Insulin versus Control, respectively), ICU length of stay (1.8 ± 3.2 vs 1.9 ± 3.7 days), and length of postoperative hospital stay (8.7 ± 14.6 vs 7.8 ± 10 days, all P values nonsignificant).
Atrial fibrillation was the most common postoperative arrhythmia, with a peak incidence on the second day after CABG (Table 3). The cumulative incidence of atrial fibrillation was 31.1% in the Insulin group and 30.1% in the Control group (P = 0.80). Stepwise logistic regression revealed the following predictors for postoperative atrial fibrillation (with odds ratios [OR] and 95% confidence intervals [CI] in parentheses): 1) age > 70 yr (OR 1.07, CI 1.05–1.09), 2) preoperative atrial fibrillation (OR 9.87, CI 2.69–36.31), and 3) preoperative creatinine level > 150 μmol/l (OR 10.23, CI 1.10–94.82).
The incidence of ventricular tachycardia was infrequent at all time points (Table 4). The cumulative incidence of ventricular tachycardia was 9.4% in the Insulin group versus 8.5% in the Control group (P = 0.71). Most episodes of ventricular tachycardia were nonsustained and did not require treatment.
Postoperative Conduction Defects
Right bundle branch block (RBBB) was the most common conduction abnormality postoperatively, with a peak incidence immediately after arrival in the ICU (Table 5). There was no difference in the incidence of RBBB between the two groups of patients. Stepwise logistic regression analysis revealed the following predictors for RBBB: 1) age >70 yr (OR 1.03, CI 1.00– 1.06), 2) female sex (OR 1.84, CI 1.06–3.20), and 3) diseased circumflex vessel (OR 2.68, CI 1.34–5.34). Left bundle branch block (LBBB) was uncommon, occurring in only 8.9% of all patients. There was no difference in the incidence of LBBB between the two treatment groups. The cumulative incidence of any intracardiac conduction abnormality (RBBB or LBBB) was 17.7% in the Insulin group versus 14.7% in the Control group (P = 0.37).
Temporary postoperative atrial or ventricular pacing was required in 39.3% of patients in the Insulin group and 42.9% in the Control group (P = 0.42). Pacing was used in the majority of patients for bradycardia, not for atrioventricular conduction abnormalities. However, on the fifth postoperative day, there were two patients in the Control group who still required pacing with epicardial wires versus none in the Insulin group.
On arrival to the ICU, 48.1% of patients in the Insulin group and 52.3% of patients in the Control group (P = 0.38) were free of any arrhythmia or conduction defect and were not paced. On the fifth postoperative day, the respective values were 70.8% and 75.2% (P = 0.32).
Postcardiac surgery arrhythmias are associated with significant morbidity and are an important cause of increased resource utilization (1–6,20). Many investigators have studied predictors of postoperative arrhythmias, particularly atrial fibrillation, and have examined various methods of prevention and treatment. In the current study, we studied the occurrence of arrhythmias and conduction defects in a high-risk patient group undergoing urgent CABG surgery using cardioplegic arrest and CPB. We focused on patients undergoing urgent CABG surgery (i.e., those requiring CABG during the same hospitalization as their admission for an acute ischemic event). We chose this patient population because of their increased incidence of cardiac morbidity (11,12), thereby improving the chance of detecting a statistically significant difference between treatment groups.
We were unable to detect a beneficial effect of glucose-insulin cardioplegia on the incidence of postoperative atrial fibrillation in the current double-blinded, randomized, controlled clinical trial. Independent predictors of atrial fibrillation were elderly age, previous atrial fibrillation, and renal insufficiency. Although several different risk factors have been previously identified (1–3), advanced age is the only consistent predictor of postoperative atrial fibrillation (7). It should be noted that all patients who were taking preoperative β-adrenergic blockers in the current trial continued to receive β-adrenergic blocker after CABG. Continuation of β-adrenergic blocker therapy after cardiac surgery results in a significant reduction in the incidence of atrial fibrillation (21,22).
Our randomized trial also failed to detect a beneficial effect of insulin-enhanced cardioplegia on postoperative conduction abnormalities. Previous studies using crystalloid cardioplegia have demonstrated conduction abnormalities in 20%–58% of patients after CABG, persisting in 3%–22% of patients at discharge (23,24). Studies using cold-blood cardioplegia have produced similar values (25,26). Flack et al. (27) compared warm- and cold-blood cardioplegia and found that the incidence of conduction abnormalities was significantly reduced in the normothermic cardioplegia group. The incidence of RBBB among the 501 patients in this study was small (12%–14% on arrival to the ICU and 4%–6% on the fifth postoperative day), which may indicate good myocardial protection in our patient population, even though normothermic cardioplegia was not used.
Exogenous glucose may be a superior substrate for the myocardium during periods of ischemia (17). Insulin stimulates the pyruvate dehydrogenase enzyme and improves Krebs cycle metabolism after ischemia (17,28). Additional possible beneficial effects of glucose-insulin include a reduction in circulating free fatty acids, which have deleterious effects on ischemic myocardium (29). In experimental studies with regional myocardial ischemia, glucose-insulin infusion decreased infarct size, increased ATP and creatinine phosphate levels, and improved ventricular function (30–32). Several clinical studies have shown beneficial effects of glucose and insulin infusions during cardiac operations (13,33,34), while others have not proven efficient (35).
The current large, randomized, controlled study of urgent CABG surgery patients reveals that cardioplegia, enhanced with glucose and insulin, has no effect on the incidence of postoperative arrhythmias or conduction abnormalities. Any possible beneficial effect of insulin infusion on myocardial metabolism was not reflected in postoperative electrical disturbances. Additionally, patient length of ICU and hospital stay was not affected by insulin-enhanced cardioplegia.
Our findings are in contrast to previously published studies. In a small, nonblinded, prospective randomized study, Lazar et al. (36) demonstrated a significant effect of perioperative glucose-insulin solutions on postoperative clinical outcomes in patients undergoing urgent CABG for unstable angina. These investigators used an insulin infusion delivered via a central vein at 0.05 IU · kg−1 · h−1. In the group receiving insulin, the authors found a significant reduction in inotrope scores, overall weight gain, ventilation times, and ICU and hospital lengths of stay. Insulin treatment also resulted in a decreased incidence of postoperative atrial fibrillation (13% vs 53% in the control group, P = 0.02).
There are several possible explanations for the discrepancies. Our study delivered insulin directly to the heart as part of the cardioplegic formulation. A previous clinical study from our institution demonstrated that this technique resulted in a significant improvement in load-independent indices of LV function (16). However, our treatment strategy did not include any insulin therapy after aortic cross-clamp removal. In contrast, Lazar et al. (36) continued insulin therapy for 12 hours after myocardial reperfusion. In addition, the smaller postoperative weight gain observed by Lazar et al. (36) in the insulin group may have contributed to the decreased incidence of atrial arrhythmias. A combination approach, involving both cardioplegic insulin delivery and postoperative IV therapy, may be optimal in preventing arrhythmias and improving LV functional recovery.
A limitation of this study is that telemetry was discontinued 48 hours after transfer to the surgical ward (3 to 4 days postoperatively for the majority of patients), in patients who remained in sinus rhythm. It is therefore possible that short episodes of asymptomatic atrial fibrillation may have been undetected. However, previous studies have demonstrated that most arrhythmias occur in the first three days postcardiac surgery (7). In addition, it is unlikely that more of these episodes would have occurred in one treatment group than the other. Given the randomized design of our trial and our sufficient sample size, we feel that our conclusions are justified.
In conclusion, we were unable to detect a beneficial effect of insulin-enhanced cardioplegia on atrial fibrillation after high-risk CABG surgery. Further studies are required to determine the optimal use of insulin during cardiac surgery.
1. Rubin DA, Nieminski KE, Reed GE, Herman MV. Predictors, prevention, and long term prognosis of atrial fibrillation after coronary artery bypass graft operations. J Thorac Cardiovasc Surg 1987; 94: 331–5.
2. Caretta Q, Mercanti CA, De Nardo D, et al. Ventricular conduction defects and atrial fibrillation after coronary bypass grafting: multivariate analysis of preoperative, intraoperative and postoperative variables. Eur Heart J 1991; 12: 1107–11.
3. Mathew JP, Parks R, Savino JS, et al., for the MultiCenter Study of Perioperative Ischemia Group. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes and resource utilization. JAMA 1996; 276: 300–6.
4. Wong DT, Cheng DC, Kustra R, et al. Risk factors of delayed extubation, prolonged length of stay in the intensive care unit, and mortality in patients undergoing coronary artery bypass graft with fast-track anesthesia: a new cardiac risk score. Anesthesiology 1995; 91: 911–5.
5. Lazar HL, Fitzgerald C, Gross S, et al. Determinants of length of stay after coronary artery bypass graft surgery. Circulation 1995; 92 (Suppl): II20–4.
6. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993; 56: 539–49.
7. Ommen SR, Odell JA, Stanton MS. Atrial arrhythmias after cardiothoracic surgery. N Engl J Med 1997; 336: 1429–34.
8. Smith PK, Buhrman WR, Levett JM, et al. Supraventricular conduction abnormalities following cardiac operations: a complication of inadequate atrial preservation. J Thorac Cardiovasc Surg 1983; 85: 105–15.
9. Mullen JC, Khan N, Weisel RD, et al. Atrial activity during cardioplegia and postoperative arrhythmias. J Thorac Cardiovasc Surg 1987; 94: 558–65.
10. Hammermeister KE, Morrison D. Coronary bypass surgery for stable angina and unstable angina pectoris. Cardiol Clin 1991; 135–55.
11. Rao V, Ivanov J, Weisel RD, et al. Predictors of low cardiac output syndrome after coronary artery bypass. J Thorac Cardiovasc Surg 1996; 112: 38–51.
12. Levy S. Factors predisposing to the development of atrial fibrillation. Pacing Clin Electrophysiol 1997; 20: 2670–4.
13. Oldfield GS, Commerford PJ, Opie LH. Effects of preoperative glucose-potassium on myocardial glygogen levels and on complications of mitral valve replacement. J Thorac Cardiovasc Surg 1986; 9: 874–8.
14. Haider W, Eckersberger F, Wolner E. Preventive insulin administration for myocardial protection in cardiac surgery. Anesthesiology 1984; 60: 422–9.
15. Haider W, Benzer H, Schutz W, Wolner E. Improvement of cardiac preservation by preoperative high insulin supply. J Thorac Cardiovasc Surg 1984; 83: 294–300.
16. Rao V, Borger MA, Weisel RD, et al., for the Insulin Cardioplegia Trial (ICT) Investigators. Insulin cardioplegia for elective coronary bypass surgery. J Thorac Cardiovasc Surg 2000; 119: 1176–84.
17. Rao V, Merante F, Weisel RD, et al. Insulin stimulates pyruvate dehydrogenase and protects human ventricular cardiomyocytes from simulated ischemia. J Thorac Cardiovasc Surg 1998; 116: 485–94.
18. Borger MA, Weisel RD. Myocardial protection. In: Cheng DC, David TE, eds. Perioperative care in cardiac anesthesia and surgery. Austin, TX: Landes Bioscience, 1999: 106–10.
19. Cheng DC, Karski J, Peniston C, et al. Early tracheal extubation after coronary artery bypass graft surgery reduces costs and improves resource use: a prospective, randomized, controlled trial. Anesthesiology 1996; 43: 160–8.
20. Aranki SF, Shaw DP, Adams DH, et al. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation 1996; 94: 390–7.
21. Kowey PR, Taylor JE, Rials SJ, Marinchak RA. Meta-analysis of the effectiveness of prophylactic drug therapy in preventing supraventricular arrhythmia early after coronary artery bypass grafting. Am J Cardiol 1992; 69: 963–5.
22. Andrews TC, Reimold SC, Berlin JA, Antman EM. Prevention of supraventricular arrhythmias after coronary artery bypass surgery: a meta-analysis of randomized control trials. Circulation 1991; 8 (Suppl): III236–44.
23. Chu A, Califf RM, Pryor DB, et al. Prognostic effect of bundle branch block related to coronary artery bypass grafting. Am J Cardiol 1987; 59: 798–803.
24. O’Connell JB, Wallis D, Johnson SA, et al. Transient bundle branch block following use of hypothermic cardioplegia in coronary artery bypass surgery: high incidence without perioperative myocardial infarction. Am Heart J 1982; 103: 85–91.
25. Wexelman W, Lichstein E, Cunningham JN, et al. Etiology and clinical significance of new fascicular conduction defects following coronary bypass surgery. Am Heart J 1986; 111: 923–7.
26. Baerman JM, Kirsh MM, de Buitleir M, et al. Natural history and determinants of conduction defects following coronary artery surgery. Ann Thorac Surg 1987; 44: 150–3.
27. Flack JE III, Hafer J, Engelman RM, et al. Effect of normothermic blood cardioplegia on postoperative conduction abnormalities and supraventricular arrhythmias. Circulation 1992; 86 (Suppl): II385–92.
28. Svedjeholm R, Hallhagen S, Ekroth R, et al. Dopamine and high-dose insulin infusion (glucose-insulin-potassium) after a cardiac operation: effects on myocardial metabolism. Ann Thorac Surg 1991; 51: 262–70.
29. Henderson AH, Most AS, Parmeley WW. Depression of myocardial contractility in rats by free fatty acids during hypoxic periods. Circ Res 1970; 26: 439–44.
30. Maroko PR, Libby P, Sobel SE, et al. Effect of glucose-insulin-potassium infusion on myocardial infarction following experimental coronary artery occlusion. Circulation 1972; 45: 1160–75.
31. Heng MK, Norris RM, Peter T, et al. The effects of glucose-insulin-potassium on experimental myocardial infarction in the dog. Cardiovasc Res 1978; 12: 429–35.
32. Opie LH, Owen P. Effects of glucose-insulin-potassium infusion on arteriovenous differences of glucose and free fatty acids and on tissue metabolic changes in dogs with developing myocardial infarction. Am J Cardiol 1976; 38: 310–21.
33. Borman B, Scheld HH, Podzuweit T, et al. Enhancement of myocardial energy potentials in man by glucose-insulin treatment and after ischaemic heart arrest. J Cardiovasc Surg 1985; 26: 182–6.
34. Svedjeholm R, Huljebrandt I, Håkansson E, Vanhanen I. Glutamate and high-dose glucose-insulin-potassium (GIK) in the treatment of severe cardiac failure after cardiac operations. Ann Thorac Surg 1995; 59: S23–30.
35. Wistbacka JO, Kaukoranta PK, Nuutinen LS. Pre-bypass glucose-insulin-potassium infusion in elective nondiabetic coronary artery surgery patients. J Cardiothorac Vasc Anesth 1992; 6: 521–7.
36. Lazar HL, Philippides G, Fitzgerald C, et al. Glucose-insulin-potassium solutions enhance recovery after urgent coronary artery bypass grafting. J Thorac Cardiovasc Surg 1997; 113: 354–60.