INTRAOPERATIVE management of blood glucose concentration in patients with diabetes mellitus has traditionally focused on avoidance of profound hypoglycemia or hyperglycemia. In contrast, the relative importance of strict perioperative regulation of blood glucose within the normal range has received little emphasis, in large part because the benefits and risks of such a strategy have yet to be convincingly demonstrated in controlled clinical trials. In this issue of Anesthesiology, Ouattara et al. 1provide compelling evidence indicating that tight control of intraoperative blood glucose positively affects patient outcome in diabetic patients after coronary artery bypass graft surgery. Using a multivariate analysis, the authors demonstrated that the risk of sustaining cardiovascular morbidity was increased more than sevenfold in patients with refractory hyperglycemia (defined as four consecutive determinations of blood glucose concentration exceeding 200 mg/dl despite insulin treatment) as compared with those in whom blood glucose concentration was more tightly controlled (< 150 mg/dl). The authors used a continuous infusion of insulin administered using a modified Portland protocol to treat blood glucose concentrations greater than 180 mg/dl. The objective of treatment was to maintain the blood glucose concentration between 150 and 200 mg/dl intraoperatively and less than 140 mg/dl postoperatively in the intensive care unit (ICU). Thirty-six percent of the patients enrolled in the study required insulin treatment according to the authors’ criteria. Fifty percent of these patients had inadequate control of intraoperative blood glucose concentration. These findings were associated with an increased in-hospital mortality rate (11.4 vs. 2.4%) and a prolonged duration of stay (> 96 h) in the ICU (46 vs. 19%) as compared with patients in whom better control of blood glucose was achieved.
The results of Ouattara et al. are quite provocative. A question that remains unanswered is whether hyperglycemia is truly a cause of increased cardiovascular morbidity and mortality or whether the correlation of hyperglycemia and cardiac complications is merely an epiphenomenon. Results from animal studies indicate that acute hyperglycemia alone, independent of chronicity, underlying diabetes, or alterations in plasma insulin concentration, adversely modulates endogenous and pharmacologically induced cardioprotective signal transduction pathways.2Hyperglycemia increases myocardial infarct size, impairs endothelial function, adversely affects coronary microcirculatory regulation, and attenuates coronary collateral development in part by increasing the production of deleterious quantities of reactive oxygen species and blunting nitric oxide–dependent protective mechanisms.3Several recent clinical trials also strongly suggest that aggressive control of blood glucose decreases overall and cardiac-related mortality in a variety of patient subpopulations. In an important prospective, randomized trial, Van den Berghe et al. 4demonstrated that intensive insulin therapy (target blood glucose concentration of 80–110 mg/dl) decreased in-hospital mortality by more than 30% in patients admitted to the ICU as compared with those who received insulin only if the blood glucose concentration exceeded 210 mg/dl. Sixty percent of the patients enrolled in this study had undergone cardiac surgery. Finney et al. 5demonstrated that hyperglycemia in excess of 145 mg/dl predicted an increase in mortality in ICU patients (primarily cardiac and thoracic surgical patients). Krinsley et al. 6reported similar findings in patients admitted to the ICU for a variety of medical conditions involving all organ systems. Interestingly, a multivariate statistical analysis identified administration of insulin as an independent predictor of death,5whereas control of blood glucose concentration rather than the dose of insulin correlated with improvements in outcome.5,7Taken together, these experimental and clinical findings suggested that insulin may activate prosurvival pathways in myocardium3and, further, that control of blood glucose is likely to play an important role in protecting against ischemic injury. Interestingly, Quattara et al. 1showed that preoperative use of insulin was associated with increased postoperative risk, but preoperative insulin treatment did not demonstrate a statistically significant interaction with poor intraoperative glycemic control as an independent predictor of morbidity. These results may indicate that preoperative use of insulin identifies a higher risk group of patients, but acute, sustained intraoperative hyperglycemia independently predicts an increased risk of morbidity. The findings also suggest that it may not be possible to identify relative “insulin resistance” solely on the basis of preoperative insulin requirements, especially in cardiac surgical patients subjected to cardiopulmonary bypass (CPB). It is interesting to speculate that thiazolidinedione insulin-sensitizing agents may improve patient outcome by enhancing the degree to which tight control of blood glucose concentrations may be achieved with exogenous insulin. Large-scale clinical trials will be required to confirm this hypothesis, however.
Does intraoperative control of blood glucose concentration make a difference? Few studies have examined the impact of intraoperative management of blood glucose on patient outcome. Furnary et al. 8evaluated the relation between average blood glucose concentration on the day of surgery and during the first 2 postoperative days on cardiac- and non–cardiac-related mortality in 3,554 diabetic patients undergoing coronary artery bypass grafting using CPB. Hyperglycemia at any time during the study period predicted cardiovascular-related mortality. Lazar et al. 9prospectively compared the influence of aggressive control of blood glucose concentration with an infusion of glucose, insulin, and potassium (initiated before anesthetic induction and continued for 12 h postoperatively) and intermittent subcutaneous insulin on morbidity and mortality in patients undergoing coronary artery bypass grafting. Blood glucose concentrations averaged 170 and 140 mg/dl before CPB and in the ICU, respectively, in patients receiving glucose, insulin, and potassium as compared with 210 and 270 mg/dl, respectively, in those receiving conventional treatment. Glucose, insulin, and potassium therapy reduced the incidence of atrial fibrillation, sternal wound infections, use of inotropes, duration of ventilatory support, and mortality. The current results of Ouattara et al. 1extend these previous findings and suggest that less rigid intraoperative glucose control increases the risk of cardiovascular complications. Interestingly, preoperative and postoperative blood glucose concentrations were similar in patients with and without postoperative morbidity. These findings suggest that marginal control of intraoperative blood glucose contributed to the development of adverse outcome in cardiac surgical patients. Whether an even more aggressive glucose management strategy may have further reduced morbidity in this subset of insulin resistant patients remains to be determined. Interestingly, Carvalho et al. 10demonstrated that it was possible to achieve strict regulation of blood glucose during cardiac surgery when an insulin infusion was initiated before the onset of CPB when blood glucose concentration exceeded 145 mg/dl. These authors suggested that an earlier use of insulin, before the development of CPB-induced insulin resistance, may contribute to more efficient preservation of normoglycemia.
In conclusion, strong evidence exists to indicate that hyperglycemia alone, with or without diabetes, contributes to morbidity and mortality in patients at risk for myocardial ischemia and reperfusion injury, including those undergoing cardiac surgery. The preponderance of evidence also supports the contention that strict intraoperative control of blood glucose with continuous intravenous infusion of insulin favorably modifies the risk of major morbidity and mortality in patients during cardiac surgery. Whether these findings will also extend to high-risk patients undergoing noncardiac surgery has not been specifically evaluated. However, data obtained from experimental and clinical studies indicate that hyperglycemia produces a variety of deleterious effects that may adversely influence patient outcome. Therefore, we have proposed a strategy that emphasizes an intensive approach to treating and avoiding hyperglycemia in high-risk surgical patients with the objective of reducing overall morbidity and mortality in these most challenging patients.3,11
*Departments of Anesthesiology and Pharmacology and Toxicology, Medical College of Wisconsin, Milwaukee, Wisconsin. †Department of Anesthesiology, Medical College of Wisconsin and the Clement J. Zablocki Veterans Affairs Medical Center, Milwaukee, Wisconsin: Departments of Pharmacology and Toxicology and Medicine (Divsion of Cardiovasculr Diseases), Medical College of Wisconsin; Professor of Biomedical Engineering, Marquette University, Milwaukee, Wisconsin. ‡Professor and Director of Cardiac Anesthesia, Department of Anesthesiology, Medical College of Wisconsin and the Clement J. Zablocki Veterans Affairs Medical Center; Professor of Biomedical Engineering, Marquette University. email@example.com