Beneficial effects of intensive insulin therapy in critically ill patients

Abstract

Critically ill patients in intensive care units (ICU) for more than a few days have a mortality of approximately 20% world-wide. These critically ill patients, in the absence of a previous diagnosis of diabetes, commonly exhibit stress hyperglycemia and insulin resistance [1]. Many of these critically ill patients die of multiorgan dysfunction (MOD) and sepsis. Since stress hyperglycemia has been shown to associate with impaired polymorphonuclear neutrophil function and bactericidal activity [2], the question arose whether lowering of blood glucose in critically ill patients would decrease morbidity and mortality. To address this question the Leuven group of Van den Berghe and associates performed a prospective, randomized control study in 1548 patients admitted to their surgical ICU who were receiving mechanical ventilation [3]. The patients were randomly assigned to receive intensive insulin therapy (maintenance of blood glucose between 80 and 110 mg/dl) or conventional treatment (insulin treatment only if blood glucose exceeded 215 mg/dl and then to maintain blood glucose between 180 and 200 mg/dl). At baseline, the two groups were comparable relative to age, gender, body mass index, APACHE II score and therapeutic intervention scoring system (TISS) and blood glucose. Two-thirds of the admissions to the surgical ICU were due to cardiac surgery in both groups. The primary endpoint of the Leuven study was ICU mortality. The intensive insulin treatment group had a significantly lower ICU mortality (4.6 vs 8%, P ¼ 0.04). This difference in mortality was primarily observed in those 451 patients who were in the ICU longer than 5 days (10.6 vs 20.2%, P ¼ 0.005). Multiple organ dysfunction (MOD) with a proven septic focus was also decreased (8 vs 33 cases, P ¼ 0.02) in the intensive insulin group as was the percentage of patients requiring greater than 14 days of ventilatory support (7.5 vs 11.9%, P ¼ 0.003). Other beneficial effects of intensive insulin therapy as compared to conventional treatment are shown in Table 1. In the total Leuven study population, the mean morning blood glucose value was 103±19 mg/dl in the intensive treatment group and 153±33 mg/dl in the conventional treatment group. Hypoglycemia (blood glucose less than 40 mg/dl) occurred in approximately 5% of the intensive treatment group without serious complications. Blood glucose measurements were performed every 1–4 h and the insulin dose adjusted according to a strict algorithm by intensive care nurses supervised by a physician not involved in the study. Although the Leuven study was performed in a surgical ICU, similar results have been recently reported from a medical-surgical ICU [4]. In a posthoc analysis of the Leuven results, a linear correlation between hyperglycemia and risk for death was observed, thus suggesting a direct toxic effect of glucose [5]. Moreover, in the conventional treatment group those patients with moderate hyperglycemia (110–150 mg/dl) had a lower risk of death than those patients with a blood glucose between 150 and 200 mg/dl. The moderate hyperglycemia group in the conventional treatment arm of the study, however, had a higher risk of death than the intensive insulin treatment group who had a blood glucose less than 110 mg/dl. In the presence of insulin resistance, glucose toxicity could be mediated by glucose overload in tissue sites in which glucose uptake is independent of insulin, including endothelial, epithelial and immune cells as well as central and peripheral nervous system and hepatocytes [6]. In that regard, of note are the observations in the Leuven study that the incidence of acute renal failure, the need for red blood cell transfusions and critical illness polyneuropathy were significantly less in the intensive insulin treatment group [3]. The beneficial effect of the decreased incidence of polyneuropathy may have been a factor in the diminished duration of mechanical ventilation in the intensive insulin treatment group. In contrast to the mitochondrial abnormalities observed in hepatocytes of the conventional treatment group, there was no morphological or functional evidence of