Inflammation is gaining widespread attention for its role in the initiation and progression of cardiovascular disease. Intensive insulin therapy with tight glucose control is reported to result in reduced morbidity and mortality in inflammation related critical illness. Experimental evidence revealed that insulin attenuates the systemic inflammatory response in rats subjected to thermal trauma or endotoxin. Clinical studies reported that insulin decreases plasma levels of high-sensitivity C-reactive protein (CRP) and serum amyloid A (SAA) in patients with acute myocardial infarction (AMI) or undergoing coronary artery bypass grafts (CABGs). These observations have implicated that insulin functions as an anti-inflammatory molecule although the underlying mechanisms remain elusive. Our study has provided in vivo direct evidence that insulin treatment reduces serum and cardiac tumor necrosis factor-α (TNF-α) induction， attenuates coronary endothelial expression of P-selectin and intercellular adhesion molecule (ICAM)-1， which leads to decreased polymorphonuclear leukocytes (PMNs) adhesion to coronary endothelium and myocardial PMNs accumulation in myocardial ischemia/reperfusion (MI/R). In the in vitro cell culture experiments， the anti-inflammatory effect of insulin was abolished by the specific Akt inhibitor and blocked by the NOS inhibitor L-NAME. Furthermore， the suppression on TNF-α by either insulin or TNF-α neutralizing antibody improved viability and reduced apoptosis of I/R cardiomyocytes. These data indicates that insulin can inhibit TNF-α induction through an Akt-activated and NO-dependent mechanism in MI/R. The anti-inflammatory property elicited by insulin may contribute to its cardioprotective and prosurvival effects in the critically ill.

    On the other hand， accumulating evidence underlines the importance of maintaining euglycemia when considering insulin infusion treatment. Because hyperglycemia is reported to increase generation of reactive oxygen species which promotes inflammation， it is possible that hyperglycemia masked the benefits of insulin therapy. In the DIGAMI 2 study， the primary target of a fasting blood glucose level of 90冾126 mg/dl in the insulin group was not achieved in this trial and insulin glucose showed no treatment benefits. In the CREATE-ECLA trial， relative hyperglycemia occurred in the glucose-insulin-potassium (GIK) treatment group and GIK did not improve the mortality in patients with AMI. Our study in an in vivo rat model with AMI showed that hyperglycemia during ischemia not only significantly increased myocardial apoptosis and infarct size， but almost completely blocked the cardioprotective effect afforded by GIK， which is partially due to hyperglycemia-induced decrease of myocardial Akt activation. Therefore we propose that normalizing blood glucose is important for insulin therapy to provide clinically relevant anti-inflammatory and cardioprotective effects in AMI.