National Institutes of Health Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Dr. Derek LeRoith, National Institutes of Health, Clinical Endocrinology Branch, Room 8D12, Building 10, Bethesda, Maryland 20892. E-mail: derek{at}helix.nih.gov.
In a landmark study, van den Berghe et al. (1) showed, quite convincingly, that patients in an intensive care setting, when treated with intensive insulin therapy to normalize blood glucose levels, experienced significantly reduced mortality and morbidity. In that prospective, randomized, and controlled trial, over 1500 patients in a surgical intensive care unit were divided into two groups, conventionally and intensively treated. In the conventionally treated group, insulin infusion was used only when the blood glucose exceeded 215 mg/dl and the blood glucose level was kept between 180 and 200 mg/dl. In the intensively treated group, blood glucose levels were maintained between 80 and 110 mg/dl. Mortality fell in the intensively treated patients by almost half while they were in the intensive care unit; in-hospital mortality fell to a similar degree. In addition, morbidity was significantly reduced. There were fewer episodes of septicemia, organ failure, and polyneuropathy.
This study has elicited enormous interest primarily among endocrinologists and diabetologists who feel that the paradigm for treating in-hospital diabetics and even nondiabetics (only 13% of patients in the van den Berghe study had a previous diagnosis of diabetes, whereas 75% were hyperglycemic on admission) needs radical rethinking. Indeed, both the American Diabetes Association and the American Association for Clinical Endocrinologists have issued consensus statements regarding in-hospital treatment of diabetic patients, basing their statements quite appropriately on this and other related studies (2).
Two recent observational studies confirmed the improved outcomes of critically ill patients using intensive insulin therapy (3) and the use of glucose-insulin infusion to treat acute myocardial infarction (4). However, the work of van den Berghe et al. (1) remains the only randomized, controlled, interventional study from which conclusions can be drawn.
One of the most intriguing questions that arises from these studies is this: what are the mechanisms involved in the improved outcomes following intensive insulin therapy and glycemic control in these patients admitted to the intensive care unit? In their most recent study published in this issue of the JCEM (5), the investigators present results of their findings with regard to the GH/IGF axis. Previous studies on this axis have revealed some interesting findings that are worth reviewing. In patients with chronic obstructive pulmonary disease, GH administration for 3 wk demonstrated an improvement in nitrogen (N2) balance as well as increased maximal inspiratory pressure, suggesting a degree of reversal of ventilatory muscle function (6). GH administration has also been shown to improve N2 balance in patients with severe burns, patients recovering from trauma on parenteral nutrition, patients with severe sepsis, and other critically ill patients. GH administration can lead to hyperglycemia by causing insulin resistance, and these investigators counterbalanced this effect by simultaneous administration of IGF-1, which by itself may cause hypoglycemia. The combination of GH and IGF-1 also results in an additive effect on N2 balance (7).
In an earlier study, GH administration to patients in an intensive care setting following cardiac or abdominal surgery, severe trauma, or acute respiratory failure resulted in a surprising increase in morbidity and mortality (8). The effects led the investigators to speculate that the poor outcome may be the result of secondary metabolic disturbances such as hyperglycemia and/or increased lipolysis (9). However, they also noted that GH is known under certain circumstances to augment the production of reactive oxygen species and inflammatory cytokines that may play a role in the poor outcome of very ill patients.
In the present study, van den Berghe et al. (5) studied in detail the GH-IGF-IGF-binding protein axis, hoping to understand the mechanism(s) that improved intensive insulin therapy outcomes in their trial. Their hypothesis was that insulin therapy would enhance the IGF-1 axis, which is clearly anabolic. To their surprise, the effect of therapy was to raise GH levels and reduce circulating IGF-1, IGF binding protein-3 (IGFBP-3), and the acid-labile subunit (ALS)a picture of GH resistance (10). Although this effect ran counter to previously held beliefs (and data) that insulin is required for GH action in liver, as shown by the GH resistance of type 1 diabetes with underinsulinized liver tissue, this unusual result may be explained teleologically. GH counteracts the action of insulin at various peripheral target tissues such as muscle and fat; therefore, a state of GH resistance at these tissues may improve insulin sensitivity. This is seen experimentally in the GH antagonist transgenic mouse model (11). The reduction in circulating IGF-1, IGFBP-3, ALS, and GH-binding protein in the subjects of the present (1) study suggests GH resistance at the liver. The data suggesting GH resistance at the liver are strongly supported by the reduction in levels of IGF-1 mRNA, as it is extremely sensitive to the biochemical effect of GH. On the other hand, the possibility of GH resistance at other target tissues remains to be determined. Interestingly, all these changes in IGF-1, IGFBP-3, and ALS did not correlate with patient survival and leads one to question whether the effects are an epiphenomenon of GH resistance with a more important role being played by the improved insulin sensitivity that GH resistance allows. The mechanisms for GH resistance need to be explored in future studies. Some intriguing data are already available from studies involving sepsis in rodents, in which GH resistance was seen in both liver and muscle at a postreceptor level involving signal transducer and activator of transcription 5, and TNF- has been invoked as one of the causative agents (12).
The question raised by these and other investigators is whether the improved outcomes in morbidity and mortality are due to the tight metabolic control or the insulin itself via a nonmetabolic effect. The Diabetes and Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI) study would suggest that improved outcomes may be achieved in seriously ill patients by affecting the metabolism of various tissues (13). However, the exact mechanisms remain to be uncovered, and investigations will continue to uncover new and intersecting pathways. For example, during acute stress, catecholamine excess inhibits pancreatic ß-cell insulin release and increased lipolysis secondarily affects insulin action; both lead to hyperglycemia whether the patient is a previously known diabetic or not. Under these conditions, improved metabolic control using insulin is an obvious course of action. In the case of acute myocardial dysfunction (DIGAMI study), the improved glucose metabolism by cardiac muscle may replace fatty acid metabolism, which is deleterious to myocardial function in the postinfarct state.
More recently, interest in the nonmetabolic effects of insulin has grown in an attempt to understand many of the complications of diabetes, particularly those involving the cardiovascular system (14). However, these studies may also help to explain results of the van den Berghe trials. Insulin affects the vasomotor responses of the vasculature via nitric oxide production in vascular endothelial and smooth muscle cells and also plays a significant role in the production of inflammatory cytokines. Is it therefore possible that the improved outcomes in patients receiving intensive insulin therapy are partly the result of improved metabolic control and also secondary to nonmetabolic effects of insulin such as reductions in inflammatory cytokines? Indeed, these investigators have measured representative examples of acute phase proteins C-reactive protein and mannose-binding lectin. High C-reactive protein levels upon admission and lack of reduction in the less intensively treated group were associated with a poorer prognosis, suggesting that intensive insulin therapy may have strong antiinflammatory effects (15). On the other hand, they found, using multivariate logistic regression analysis, that the lowered blood glucose level was related to reduced mortality and not to the insulin dose administered (16). Further exploration of these types of questions in these clinical trials and animal models will surely increase our understanding and help design appropriate algorithms for in-hospital patient care.
In summary, the studies by van den Berghe et al. are driving a whole new area of research and clinical applications. Among the many questions that remain to be answered are the following: 1) how applicable are their findings to other situations of in-hospital diabetic and nondiabetic patients? 2) how readily will the medical community accept these proposed new algorithms and paradigms? 3) what areas of basic and clinical research should be investigated to further our understanding of the mechanisms involved in the marked improvements seen by these interventions?
Presumably, both the research funding agencies and the healthcare management organizations will appreciate the advantages of further research and implementation of this exciting new paradigm of management.
Footnotes
Abbreviations: ALS, Acid-labile subunit; IGFBP-3, IGF binding protein 3.
Received April 20, 2004.
Accepted May 17, 2004.
References
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |