The Potential Therapeutic Role of Insulin in Acute Myocardial Infarction in Patients Admitted to Intensive Care and in Those With Unspecified Hyperglycemia

While acute myocardial infarction and patients admitted to intensive care units carry a serious risk of mortality and morbidity, the concomitant occurrence of hyperglycemia enhances this risk. This has been established through various studies over the past decade (1–3). More recently, it has been shown that the treatment of these conditions with intravenous infusions of insulin results in marked improvements in clinical outcomes (3–5). While the normalization of blood glucose concentration may be one factor in the improvement of these outcomes, insulin may exert benefits of its own, independently of the plasma glucose concentrations.

Both septicemia and acute myocardial infarction are inflammatory states with markedly enhanced catabolism. There may be hyperglycemia and an increase in plasma free fatty acid (FFA) concentrations in both of these conditions. Increased FFA concentrations are known to be associated with a deterioration of clinical outcomes and may have toxic effects of their own on vascular reactivity and on the myocardium (6–9). FFAs may have effects on endothelial nitric oxide (NO) production (6), prostacyclin production (10), and the stability of prostacyclin in plasma (11). FFAs may also induce a state of insulin resistance through the induction of protein kinase C and inflammation through the suppression of inhibitor κB (IκB) in the skeletal muscle (12). Proinflammatory changes are also induced in the mononuclear cells by FFAs as reflected in an increase in nuclear factor-κB (NF-κB), the major proinflammatory transcription factor, an increase in reactive oxygen species (ROS) generation, and an increase in p47 ^phox (13). Thus, the normalization of FFA concentrations by insulin may produce a benefit. Similarly, it can be argued that insulin administration may facilitate the uptake of glucose by insulin-responsive organs, including the myocardium, thus providing for their metabolic needs. These mechanisms are based on our understanding of insulin as a metabolic hormone.

The last few years have witnessed the emergence of insulin in a new and a totally unexpected role. Recent data have shown that insulin has a profound, rapid, and potent anti-inflammatory effect at the cellular and molecular levels (14–17 similar to that observed with glucocorticoids both in vitro (18) and in vivo (19–21). Thus, insulin suppresses the proinflammatory transcription factor NF-κB and the expression of the adhesion molecule, intercellular adhesion molecule-1 (ICAM-1), and the chemokine, monocyte chemotactic protein-1 (MCP-1), in human aortic endothelial cells in vitro (14,15). Since one of the initial steps in inflammation is the adhesion of the circulating leukocytes to the endothelium and, thereafter, the arrival of more leukocytes to the site through chemokine action (22,23), the suppression of these two mediators of inflammation is cardinal in exerting an anti-inflammatory effect. Furthermore, the suppression of NF-κB may potentially lead to the inhibition of other potent mediators, like early growth response-1 (VEGF), and proinflammatory cytokines, like tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6), both of which are secreted in response to endotoxin challenge (24).

Recent observations in humans in vivo have also shown that a low-dose infusion of insulin into human subjects induces a significant acute anti-inflammatory effect within 2 h (16).

Thus, insulin causes a suppression of intranuclear NF-κB and an induction of the cytosolic IκB; it also causes a suppression of ROS generation and the suppression of p47^phox subunit of NADPH oxidase, the enzyme that converts molecular O₂ to the superoxide radical. Plasma concentrations of soluble ICAM-1 and MCP-1 as well as those of C-reactive protein (CRP) are reduced following insulin. These effects indicate a general anti-inflammatory action and are similar to those described for glucocorticoids in vitro and in vivo. In addition to NF-κB suppression, insulin has been shown to suppress two other proinflammatory transcription factors, activator protein-1 (AP-1) (25) and early growth response-1 (Egr-1) (17). These actions are of great interest since Egr-1 regulates the expression of tissue factor (TF) (26) and plasminogen activator inhibitor-1 (PAI-1) (27). TF is expressed on the surface of macrophages and foam cells in the atherosclerotic plaque. The exposure of the foam cells to blood following the rupture of the atherosclerotic plaque triggers the activation of the extrinsic coagulation cascade through TF-mediated activation of factor VII, which eventually results in the conversion of prothrombin to thrombin. Thrombin is a potent platelet proaggregator and also converts fibrinogen to fibrin. This sets the stage for thrombosis following plaque rupture during the evolution of acute myocardial infarction (22,23,28,29). The suppression of PAI-1 by insulin also has potential benefits since PAI-1 is an inhibitor of fibrinolysis (30,31). Insulin may thus facilitate clot dissolution in acute myocardial infarction. These actions of insulin may have further subtle effects on inflammation since procoagulatory processes may trigger proinflammatory processes. Indeed, activated protein C, an anticoagulant, has been shown to have beneficial effects in septicemia (32).

Insulin has recently also been shown to suppress AP-1 (25), a transcription factor that regulates the expression of matrix metalloproteinases (MMPs). MMPs are capable of lysing collagen and other matrix proteins and thereby allow the spread of inflammatory cells (33–36. They thus mediate the thinning and finally the rupture of the atherosclerotic plaque, which triggers thrombosis as described above. MMPs also mediate the initial intimal injury, which allows the monocytes to enter the subendothelial intima in between endothelial cells. The suppression of MMPs secondary to that of AP-1 would therefore be protective to the endothelium and to the fibrous cap of the plaque. This could be useful in the clinical setting of unstable angina, which is associated with the rupture of the atherosclerotic plaque. AP-1 and MMPs are also suppressed by glucocorticoids in vitro (37,38) and in vivo (19). Glucocorticoids also suppress Egr-1 and, more than likely therefore, TF and PAI-1 (39).

One mechanism underlying the anti-inflammatory effect of insulin may be through the release of NO from the endothelium; NO is known to exert an anti-inflammatory effect (40). Insulin has also been shown to induce an increase in the expression of NO synthase (NOS), the enzyme that generates NO (41). The induction of NOS and the release of NO acutely impart to insulin a vasodilatory property through the action of NO on the vascular smooth muscle. A similar effect of insulin on platelets allows it to exert an anti-aggregatory effect. The platelet has NOS that is activated by insulin to induce the generation of NO. NO exerts its effects on vascular smooth muscle and platelets through the activation of guanylate cyclase, which generates cGMP from GTP (42–44). The vasodilatory and antiplatelet effects are clearly of importance in the treatment of acute myocardial infarction. It is also of interest that NO may mediate at least a part of the anti-inflammatory effect attributable to insulin since the coincubation of an NOS inhibitor, nitro-arginine, with insulin and human aortic endothelial cells neutralizes the suppressive effect of insulin on the expression of ICAM-1 by these cells (14).

It should be mentioned that insulin sensitizers of the thiazolidinedione (TZD) class have been shown to exert anti-inflammatory effects including the suppression of NF-κB, ROS generation, and p47^phox subunit in mononuclear cell (MNC) and plasma concentrations of ICAM-1, MCP-1, TNF-α, and CRP. More recently, rosiglitazone has been shown to suppress plasma MMP-9 concentrations in diabetes (45). While one expects their effects to be useful in chronic situations, it is relevant that their anti-inflammatory effects are observed in 3–7 days in humans in vivo. In an experimental model of acute myocardial infarction, even a single dose of rosiglitazone has been shown to reduce myocardial damage by 50% (46,47). Therefore, the potential anti-inflammatory effect of TZDs in acute clinical situations requires investigation.

While insulin exerts a comprehensive anti-inflammatory effect of relevance to the treatment of different inflammatory conditions, as described above, it is remarkable that glucose exerts a powerful proinflammatory effect. Thus, the administration of 75 g of glucose orally results in an increase in ROS generation by the circulating polymorphonuclear leukocytes (PMNs) and MNCs, an increase in p47^phox protein, and a fall in plasma α-tocopherol concentrations. In addition to causing an increase in lipid peroxidation as reflected in an increase in plasma concentration of thiobarbituric acid-reacting substances (TBARS) (48–52). Glucose has also been shown to exert other prothrombotic effects (53). Clearly, therefore, glucose induces acute oxidative stress. Glucose has also been shown recently to induce an increase in intranuclear NF-κB, a fall in cytosolic IκB, an increase in IκB kinase (IKK)-α and IKKβ (49). Glucose induces similar proinflammatory changes in endothelial cells in vitro (54). These changes are clearly proinflammatory. In addition, it has also been shown that a 75-g glucose challenge in normal subjects induces an increase in AP-1 activity in MNCs along with an increase in the expression of MMP-2 in MNCs and an increase in plasma concentrations of MMP-2 and MMP-9 (50). As described above, MMPs aid in the spread of inflammation and in mediating the rupture of the atherosclerotic plaque. Similarly, glucose induces an increase in Egr-1 and TF expression in MNCs and an increase in plasma concentration of TF. Thus, several of the effects of glucose are diametrically opposite to those of insulin in the context of inflammation. Glucose also causes abnormalities in vascular reactivity (55). This may affect vital organ perfusion adversely especially during situations of increased demand.

These proinflammatory effects of glucose have also been demonstrated in vitro. The incubation of endothelial cells in high glucose concentrations results in ROS generation and an increase in intranuclear NF-κB (54,56,57). Similarly, the incubation of monocytic cell lines in high glucose concentrations results in the elevation of intranuclear NF-κB and AP-1 with an increase in the expression of the proinflammatory cytokine, TNF-α (58). These effects of glucose are clearly of clinical relevance since hyperglycemia at the time of admission to the hospital determines the mortality and the duration of admission to the hospital, independently of the primary diagnosis or the previous history of diabetes. Because this effect of hyperglycemia was demonstrated in a retrospective study, it is important that these data be confirmed in a prospective study. However, it is legitimate to consider the aggressive control of blood glucose concentrations with insulin in patients admitted to the hospital to determine whether the normalization of blood glucose concentrations in hyperglycemic subjects improves the clinical outcomes. Such a study would have both clinical and pharmaco-economic implications (59).

These observations lead us into a new paradigm (60) in which glucose and possibly other macronutrients have a proinflammatory action, while the hormone secreted in response to their intake has an anti-inflammatory effect. This new paradigm facilitates and extends our understanding of normal physiology and pathophysiology and also provides us with novel potential therapeutic uses of insulin and the normalization of plasma glucose concentrations. On the basis of the above discussion, it should be clear that any elevation of plasma glucose concentration is likely to be proinflammatory and thus potentially harmful. Similarly, the administration of insulin is likely to be anti-inflammatory and, thus, may be clinically beneficial. Therefore, when we use insulin to control elevated blood glucose concentrations, we benefit from both the reduction in the toxic effect of glucose and the anti-inflammatory effect of insulin. The clinical applications of this new paradigm are just beginning to be discovered. The possibilities appear to be immense and promising.