OUP user menu

Glycaemic control in acute coronary syndromes: prognostic value and therapeutic options

Raffaele De Caterina, Rosalinda Madonna, Harald Sourij, Thomas Wascher
DOI: http://dx.doi.org/10.1093/eurheartj/ehq162 1557-1564 First published online: 2 June 2010


Type 2 diabetes and acute coronary syndromes (ACS) are widely interconnected. Individuals with type 2 diabetes are more likely than non-diabetic subjects to experience silent or manifest episodes of myocardial ischaemia as the first presentation of coronary artery disease. Insulin resistance, inflammation, microvascular disease, and a tendency to thrombosis are common in these patients. Intensive blood glucose control with intravenous insulin infusion has been demonstrated to significantly reduce morbidity and mortality in critically ill hyperglycaemic patients admitted to an intensive care unit (ICU). Direct glucose toxicity likely plays a crucial role in explaining the clinical benefits of intensive insulin therapy in such critical patients. However, the difficult implementation of nurse-driven protocols for insulin infusion able to lead to rapid and effective blood glucose control without significant episodes of hypoglycaemia has led to poor implementations of insulin infusion protocols in coronary care units, and cardiologists now to consider alternative drugs for this purpose. New intravenous or oral agents include the incretin glucagon-like peptide 1 (GLP1), its analogues, and dipeptidyl peptidase-4 inhibitors, which potentiate the activity of GLP1 and thus enhance glucose-dependent insulin secretion. Improved glycaemic control with protective effects on myocardial and vascular tissues, with lesser side effects and a better therapeutic compliance, may represent an important therapeutic potential for this class of drugs in acutely ill patients in general and patients with ACS in particular. Such drugs should be known by practicing cardiologists for their possible use in ICUs in the years to come.

  • Diabetes
  • Acute coronary syndromes
  • Glucose control
  • Insulin
  • Incretins
  • Glucagon-like peptide 1


Diabetes mellitus (diabetes) and cardiovascular disease (CVD) are two widely interconnected entities. The Euro Heart Survey on Diabetes and the Heart8 and our own data1 indicate not only a high prevalence of diabetes, but also the high rates of undiagnosed diabetes or prediabetic states such as impaired glucose tolerance (IGT) or impaired fasting glucose (IFG) in patients with stable or unstable coronary heart disease (CHD). With CHD ranking as the number 1 cause of death worldwide,2 with diabetes increasing by two to three times the risk of CHD,3 and with diabetes and the often preceding metabolic syndrome dramatically increasing their prevalence in Europe4 over the past 20 years, diabetologists and cardiologists have started to join their forces to improve the management of the millions of patients suffering from both diseases.4

The pathophysiological links between diabetes and its preceding states on the one hand, and CVD on the other are complex and involve hyperglycaemia, insulin resistance, and β-cell dysfunction, and a clustering of the risk factors for atherosclerosis.5 We will here focus on hyperglycaemia and ways of correcting it. We will here (i) review the evidence linking hyperglycaemia in patients with acute coronary syndromes (ACS) with an adverse prognosis, (ii) review strategies adopted so far to control hyperglycaemia in the acute cardiac care setting, and (iii) hypothesize a role for recently developed drugs—incretin analogues and mimetics—in the treatment of hyperglycaemia.

Prognostic role of hyperglycaemia, impaired glucose tolerance, and diabetes in patients with acute coronary syndromes

Hyperglycaemia at admission for an ACS is associated with a less favourable outcome in patients with or without known diabetes610 and has been considered an acute stress response.12 Recently, Sinnaeve et al.11 could further establish the prognostic value of admission vs. fasting hyperglycaemia for in-hospital and 6-month mortality in diabetic and non-diabetic subjects with ACS. Their results indicate that such associations are stronger for in-hospital mortality and steeper for fasting hyperglycaemia. Such findings indicate that long-term glucometabolic dysregulation, in addition to stress hyperglycaemia, increases risk in ACS patients even in the absence of manifest diabetes. The Glucose Tolerance in Patients with Acute Myocardial Infarction (GAMI) study assessed the prevalence of abnormal glucose regulation using an oral glucose tolerance test (OGTT) in patients without known diabetes admitted for an AMI.12,13 An OGTT was performed at discharge from an intensive care unit (ICU) and after 3 months. At these time points, 35 and 40% of the patients had IGT, and 31 and 25% were newly diagnosed with diabetes, respectively (Table 1). It seems unlikely therefore that hyperglycaemia after an ACS is a mere stress response, whereas the existence of a pancreatic β-cell dysfunction together with insulin resistance before the ischaemic event seems more probable.14 The follow-up of the GAMI population indicated that in subjects with no known diabetes, abnormal glucose metabolism is associated with a substantially increased risk for further CHD events.16 In the Euro Heart Survey, a worse outcome was observed for stable and unstable CHD patients with newly detected diabetes after 1 year, but not for a composite group of patients with IFG or IGT,17 a finding somewhat contradictory to that obtained in the DECODE study that established glucose tolerance as a clear risk factor for cardiovascular events.15 Finally, in patients with established diabetes, both fasting glucose levels during an ACS11 and long-term glycaemic control before and after an ACS16,17 predict the risk of further cardiovascular events.

View this table:
Table 1

Glucose metabolism after discharge from a coronary care unit and after 3 months in patients with acute myocardial infarction

Blood glucose after OGTT
<141 mg/dL141–200 mg/dL>200 mg/dL
<7.8 mmol/L7.8–11.1 mmol/L>11.1 mmol/L
After hospital discharge (n = 164)34% (n = 55)35% (n = 58)31% (n = 51)
Three months after discharge (n = 144)35% (n = 50)40% (n = 58)25% (n = 36)
  • IGT, impaired glucose tolerance; OGTT, oral glucose tolerance test (at 2 h). From Norhammar et al.77

Thus, disturbances of glucose metabolism are widely prevalent in ACS and relate to an adverse outcome, irrespectively of the presence or absence of previously diagnosed diabetes.

Biological plausibility of a causal relationship between hyperglycaemia and cardiovascular events after acute coronary syndromes

Possible mechanisms explaining the relationship between hyperglycaemia at admission and prognosis are summarized in Table 2. Some of these mechanisms postulate the existence of a non-causal relationship between hyperglycaemia and cardiovascular events after ACS, in that hyperglycaemia would represent just a bystander of myocardial injury and poor prognosis. In the Atherosclerosis Risk In Communities (ARIC) study, analysing the association of traditional and non-traditional risk factors with the incidence of CHD in diabetic adults, such incidence was related both to sex, age, and ethnicity and to the waist-to-hip ratio, levels of HDL3 cholesterol, apolipoproteins A-I and B, albumin, fibrinogen, von Willebrand factor and coagulation factor VIII activity, and leucocyte count.18 This may reflect the underlying inflammatory reaction or microvascular injury related to atherosclerosis in diabetes and a tendency to thrombosis. Alternatively, non-traditional risk factors may be common antecedents for both diabetes and coronary events, which might be the final outcome of a common underlying derangement such as insulin resistance. This is suggested by the association of insulin resistance with high levels of fibrinogen, tumour necrosis factor-α, and plasminogen activator inhibitor-1.

View this table:
Table 2

Potential mechanisms explaining the relationship between hyperglycaemia at admission and prognosis

Decreased insulin sensitivityImpaired glucose utilization25
Fatty acids-mediated inhibition of glucose oxidationMyocardial cell death7,78
Injury of cardiomyocyte plasma membrane
Calcium overload and arrhythmias
Increased levels of catecholaminesIncreased myocardial damage and infarct size7,79
Hyperglycaemia-induced osmotic diuresis and volume depletionDecreased end-diastolic volume7
Increased infarct size
Congestive heart failure
Cardiogenic shock
Enhanced platelet activationIncreased rate and severity of thrombotic events78
Inflammatory-immune reactions with increased markers of inflammationImpairment of functional recovery of the myocardium after ischaemia7,78,79

The opposing view, however, is that high glucose levels per se have a causal role. The relationship of high blood glucose with risk of death or poor outcome after AMI seems to hold both for diabetic and non-diabetic patients.7,19,20 The 1-year mortality is 22% in patients with glucose levels >200 mg/dL (11.1 mmol/L), but 6% in those with glucose levels <101 mg/dL (5.6 mmol/L) upon admission.8 More recently, it has been reported that fasting blood glucose during an ACS is more strongly associated with outcome than admission glucose.11

Biological plausibility and correlation studies cannot however prove causality. Only intervention studies with selective modifications of glucose levels after hospital admission may confirm a true causal relationship.

Attempts at proving causality

Several strategies have been followed over the years related to glucometabolic interventions in ACS. The first strategy investigated was infusion of insulin, glucose, and potassium (GIK). The rationale was that provision of potassium increases ischaemia-depleted myocyte potassium levels to prevent ventricular arrhythmias, with the simultaneous administration of insulin to facilitate potassium transport into the cells.21 Additionally, glucose enables a more efficient energy metabolism compared with free fatty acids or ketone bodies. These studies, however, were done in diabetic as well as non-diabetic subjects and did not aim at improving glucose levels. Recent large studies, such as the CREATE-ECLA22 or the OASIS-6 GIK23 trials, have convincingly shown that GIK infusion has no effects on mortality, cardiac arrest, or cardiogenic shock in ACS. Furthermore, a combined evaluation of both studies showed that such therapy might be even harmful in the early phase of ACS, due to hyperkalaemia or the fluid challenge. Another important issue regarding the GIK concept is that despite provision of insulin, mean glucose levels even increased during therapy [in the CREATE-ECLA study from 162 mg/dL (9 mmol/L) at baseline to 187 mg/dL (10.4 mmol/L) after 6 h of GIK administration]. Since glucose levels are, as stated before, independently associated with mortality, the ‘glucometabolic-intervention’, as a matter of fact increasing glucose levels and therefore going in the wrong direction, can be interpreted as positive outcome for the tested hypothesis.

Controlling glucose with the continuous intravenous (i.v.) administration of insulin in diabetic patients with ACS with the hope of impacting on prognosis was the aim of the ‘Diabetes Mellitus, Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI)’ trials.2428 An insulin–glucose infusion for 24 h after hospital admission, followed by intensified insulin therapy after discharge, was superior to conventional treatment in improving acute and long-term glycaemic control.24 More relevant, the 1-year mortality was 18.6% in the infusion group vs. 26.1% in the control group (29% RRR). The main benefit was seen during the early post-infarction period, without any effect on reinfarctions and with only a trend of reduced rates of heart failure.25,26 Subsequent reports27,29 confirmed these observations, showing an absolute 11% reduction in long-term mortality (3–4 years) in patients allocated to insulin–glucose infusion (Figure 1A). The most striking benefit was seen in patients with high admission blood glucose and with no previous insulin treatment and a low risk.

Figure 1

All-cause mortality in DIGAMI 1 (A) and DIGAMI 2 (B). The Kaplan–Meier curves show: (A) mortality outcome (P = 0.027) with insulin treatment vs. oral antidiabetic drugs; (B) the absence of statistically significant decrease in mortality in all three groups [Group 1: acute insulin infusion (24 h) and then s.c. insulin; Group 2: acute insulin infusion (24 h) and then standard long-term therapy; Group 3: conventional therapy]. Modified from Malmberg et al.17,27

The findings of DIGAMI 1, however, were not confirmed in the larger DIGAMI 2 trial,27 which reported on 1253 patients with type 2 diabetes and ACS, allocated to one of three treatments: (i) acute insulin–glucose infusion followed by insulin-based long-term treatment; (ii) insulin–glucose infusion followed by standard glucose control; and (iii) routine metabolic management according to local practice. Mortality did not differ between groups, but mortality in the standard care group was, although not significantly, the lowest in the trial (Figure 1B). The interpretation of the DIGAMI 2 results, however, is uncertain. The study had to be stopped prematurely due to slow recruitment rate, resulting in a drop of statistical power to below 50%. Substantial differences in baseline and on-trial glycaemic control were seen between the two studies (Table 3). In particular, in DIGAMI 2, the acute decrease in glycaemia during insulin infusion was modest and not different between groups, and no difference in HbA1c could be achieved. Furthermore, advances in the treatment of ACS might also have been relevant, as the overall 1-year mortality was 65% of the mortality observed in DIGAMI 1. The epidemiological analysis from DIGAMI 2, however, confirmed that glucose level at admission was a strong, independent predictor of long-term mortality in type 2 diabetic patients with ACS.17

View this table:
Table 3

Comparison of glucose control in DIGAMI 1 and DIGAMI 2

Control group (n = 314)Insulin infusion group (n = 306)P-valueInsulin infusion + insulin s.c. long-term (n = 474)Insulin infusion alone group (n = 473)Control group (n = 306)P-value
HbA1c at randomization (%)8.0 ± 2.08.2 ± 1.9NS7.2 ± 1.77.3 ± 1.77.3 ± 1.7NS
HbA1c decrease0.4 ± 1.51.1 ± 1.6<0.0001∼0.5%∼0.5%∼0.5%NS
Blood glucose (mg/dL)
 At randomization283 ± 76277 ± 74NS230 ± 81223 ± 79232 ± 83NS
 After 24 h211 ± 74173 ± 59<0.0001164 ± 54164 ± 50180 ± 650.0001
 Δ24 h 72104 66 59 52
  • HbA1c, glycated haemoglobin; s.c., subcutaneous; Δ24 h, difference in blood glucose at randomization—24 h; NS, not significant. Data from Malmberg et al.24,27

Two other studies support the concept that glucose control matters in diabetic patients with CVD. In the Munich registry,30 optimization of care in diabetic patients with AMI was investigated. Part of the intervention studied was an i.v. insulin infusion aiming at normalizing hyperglycaemia within 12 h. In-hospital mortality in diabetic patients with ACS dropped from 29 to 17% with optimized care. Lazar et al.31 investigated a modified glucose–insulin–potassium infusion to normalize glycaemia in 141 diabetic patients undergoing coronary artery bypass grafting (>70% urgent procedures) vs. standard therapy [target glucose <250 mg/dL (13.9 mmol/L)]. The infusion was started before induction of anaesthesia and continued for 12 h after the arrival in the ICU. The duration of stay in the ICU and in the hospital was significantly shorter in the infusion group. Furthermore, recurrent ischaemia, wound infections, and survival after 2 years significantly improved in the intervention group.

Further support to a role for the normalization of hyperglycaemia to improve outcome in ACS comes from a study in a mixed, diabetic, and non-diabetic population that indicated that in ACS, lowering of glucose by whatever means is beneficial.32 Conversely, in a subanalysis of the SYMPHONY studies, glucose-lowering therapy including only insulin and/or sulfonylureas (insulin-providing) was associated with higher 90-day death, ACS, and severe recurrent ischemia compared with biguanide only and/or thiazolidinedione therapy (12.0 vs. 5.0%).33

To determine the effect of insulin in the management of hyperglycaemia in non-diabetic patients presenting with ACS, the observational Myocardial Infarction National Audit Project (MINAP) was conducted in 201 hospitals in England and Wales. Of patients having a clear treatment strategy, 36% received diabetic treatment (31% with insulin). Mortality at 7 and 30 days was 11.6 and 15.8%, respectively, for those receiving insulin, and 16.5 and 22.1%, respectively, for those who did not. Compared with those who received insulin, after adjustment for age, gender, co-morbidities, and admission blood glucose, patients who were not treated with insulin had a relative increased risk of death of 56% at 7 days and 51% at 30 days.34

Similarly, in the CARDINAL trial database, using a multivariable 30-day mortality model, neither baseline glucose nor the 24 h change in glucose predicted mortality in diabetic patients. However, in non-diabetic patients, higher baseline glucose predicted higher mortality [hazards ratio (HR) 1.12, per 0.6 mmol/L (11 mg/dL) increase], and a greater 24 h change in glucose predicted lower mortality (HR 0.91, for every 0.6 mmol/L drop in glucose in the first 24 h) at 30 days. At multivariable analysis, baseline glucose and 24 h changes remained significant mortality predictors at 180 days in non-diabetic patients.35

Some support for intensive insulin treatment in ACS needs to be borrowed from experiences in ICU. Furnary36 had observed a significant reduction of mortality and the elimination of sternal wound infection in diabetic patients after heart surgery treated by a single i.v. infusion of insulin. Hyperglycaemic patients admitted to an ICU and receiving an insulin infusion targeting blood glucose to <145 mg/dL (8.0 mmol/L) had reduced mortality by 30% compared with controls.37 In-hospital mortality increased progressively with increasing blood glucose, consistent with an observational study enrolling patients newly admitted to an ICU.38

Two studies by the Leuven group are of great relevance in this context.39,40 A first prospective, randomized, controlled trial showed reduced mortality and morbidity in critically ill hyperglycaemic—but not necessarily diabetic patients—with blood glucose normalization through intensive insulin therapy.39 Here, 1548 patients admitted to a surgical ICU were randomly assigned to receive either a continuous insulin infusion with a glucose target <110 mg/dL (6.1 mmol/L) or conventional insulin therapy (glucose targets 180 and 200 mg/dL (10.0 and 11.1 mmol/L, respectively)). At 12 months, 4.6% had died in the intensive group compared with 8% in the conventional treatment group. Intensive insulin therapy also reduced surgical mortality, episodes of septicaemia, polyneuropathy, end-stage renal disease, and the need of blood transfusion.39,41 Concerns have been raised because of the high mortality in the control group; the unusually frequent administration of parenteral calories; a preponderance of patients who had cardiac surgery in the single centre where the study was performed; and the fact that in all such studies, blinding of the investigators is nearly impossible.42 The same group, more recently, extended their findings to patients admitted to a medical ICU for predominantly pulmonary or gastrointestinal reasons.40 In patients who stayed in the ICU for >3 days, the primary target population, in-hospital mortality was significantly lower with intensive therapy (43%) than in the others (52.5%) (P = 0.009). Independent of the length of stay in the ICU, a lower morbidity and an earlier discharge were observed in the intensive insulin group.42

Two large randomized trials that were planned to evaluate glycaemic control in ICU patients were stopped prematurely. The ‘Glucontrol Study’ was stopped after recruitment of one-third of the anticipated 3500 patients due to the failure of achieving target glucose values and to increased rates of hypoglycaemia.43 In the VISEP study, in severe sepsis, increased rates of hypoglycaemia resulted in stopping of the insulin therapy arm. In the truncated study population, no benefit of insulin treatment could be found.44

More recently, the publication of the Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation (NICE-SUGAR) trial,45 involving diabetic and non-diabetic subjects, has ignited further discussion. Within 24 h after admission to an ICU, patients who were expected to stay ≥3 days were randomly assigned to undergo either intensive glucose control, with a target blood glucose range of 81–108 mg/dL (4.5–6.0 mmol/L), or conventional glucose control, with a target ≤180 mg/dL (10.0 mmol/L). The primary endpoint was death within 90 days after randomization. A total of 829 patients (27.5%) in the intensive control group and 751 (24.9%) in the conventional control group died (OR for intensive control, 1.14; P = 0.02). The two treatment groups showed good glycaemic separation, with a mean difference of 29 mg/dL (1.6 mmol/L).39,40 The results of NICE-SUGAR contrasted starkly with those of preceding trials, with an increase in the rate of the primary endpoint, death at 90 days, and with intensive glucose control (27.5 vs. 24.9%). Not surprisingly, severe hypoglycaemia occurred in many more patients in the intensive control than in the conventional control group (6.8 vs. 0.5%, P < 0.001). The treatment effect did not differ significantly between surgical and medical patients.45 The differences between this large study and the Leuven studies have been analysed.46 Probably, the main difference is that the Leuven studies compared intensive glycaemic management with standard management at that time—reduction of glucose level only if the level is markedly elevated (>215 mg/dL) (>11.9 mmol/L). In contrast, the glucose level in the conventional control group of the NICE-SUGAR trial was targeted at only a mildly elevated range—144–180 mg/dL (8.0–10.0 mmol/L)—and more than two-thirds of these patients still received i.v. insulin to accomplish this goal. There are however other explanations. One is that insulin itself may have deleterious effects (sympathetic activation, sodium retention, or mitogenic actions, see below).47 This has been however contradicted by Kosiborod et al.48 indicating that glucose normalization during AMI is beneficial, irrespective of whether the patients receive insulin or not. It is also possible that the complexities of intensive management of glucose distracted from other, more important, management practices in the ICU. It is also possible that stress hyperglycaemia is the body's proper response to illness—an attempt to shunt energy from skeletal muscle to critical organs. Lastly, the increased mortality could be related to hypoglycaemia49 (Figure 2) and the resultant neuroglycopenia, which are difficult to detect in intubated and sedated patients.5052 The occurrence of hypoglycaemia does indeed relate to later mortality in DIGAMI 2, although this relationship disappeared when adjusted for other risk factors. Hypoglycaemia during hospitalization (as hyperglycaemia at admission) predicts increased mortality,53 possibly due to increased myocardial ischaemia,53 adrenergic stimulation,54 impaired metabolism, and increased apoptosis.55 Hypoglycaemia, however, occurring more often in patients with lower body weight or longer diabetes duration, may be simply a marker of increased risk,56 as suggested by studies indicating that uninduced hypoglycaemia, but not iatrogenic hypoglycaemia, predicts a poorer outcome.48,57

Figure 2

Incidence of hypoglycaemia in (A) DIGAMI 1,27 after conventional therapy (oral glucose-lowering drugs) and glucose–insulin infusion for 24 h followed by multidose s.c. insulin; (B) DIGAMI 2,17 after conventional therapy (oral glucose-lowering drugs), glucose–insulin infusion for 24 h followed by multidose s.c. insulin and glucose–insulin infusion for 24 h followed by oral glucose-lowering drugs; and (C) after conventional insulin therapy [continuous insulin infusion aimed at targeting blood glucose between 180 and 200 mg/dL (10.0–11.1 mmol/L)] and continuous insulin infusion aimed at targeting blood glucose between 80 and 110 mg/dL (4.4–6.1 mg/dL). Data from Van den Berghe et al.39

As a consequence of these new data, statements of International Societies now need to be revised or amended. The 2006 Consensus Statement of the American College of Endocrinology (ACE) and the American Diabetes Association (ADA) contained recommendations for the managements of diabetic patients in the ICU.58 The document underlined the advantages of intensive glucose control by insulin in decreasing morbidity, mortality, and costs related to duration of hospital stay. Recognizing that hypoglycaemia remains the major problem for intensive blood glucose control, the ACE/ADA suggested frequent blood glucose measurements during insulin treatment and nutritional plans focusing on consistent carbohydrate consumption. The European Society of Cardiology (ESC)/European Association for the Study of Diabetes (EASD) recommends blood glucose control by intensive insulin treatment (Class I recommendation) in patients after heart surgery (level of evidence B), in patients admitted to an ICU (level of evidence A), and in patients with AMI (Class II, level of evidence B).4 The more recent 2008 ESC guidelines on ST-elevation MI suggest ‘to keep glucose levels within normal ranges in diabetic patients’ with ‘suggested’ target glucose levels between 90 and 140 mg/dL (5–7.8 mmol/L) and avoiding blood glucose levels <80–90 mg/dL (4.4–5 mmol/L).59 The more detailed 2008 Statement of the AHA suggests to consider ‘intensive glucose control in patients with significant hyperglycaemia (>180 mg/dL, 10 mmol/L), regardless of prior diabetes history’, and sets similar goals. In patients hospitalized in non-ICU settings, where glucose control is likely less optimal, efforts should be directed at maintaining plasma glucose levels <180 mg/day (10 mmol/L) with subcutaneous (s.c.) insulin.60

A fair summary of the disparate trials and recommendations would be that diabetic or hyperglycaemic patients with ACS, as well as undergoing cardiac surgery, most likely benefit from strategies aiming at improved glucose control predominantly by the use of continuous insulin administration if in the ICU, through a variety of molecular mechanisms (Figure 3). It remains highly debated, however, how tight the control of glucose levels should be.46

Figure 3

Molecular mechanisms of insulin: upon binding its tyrosine-kinase receptor, insulin induces receptor dimerization and activation of cascade phosphorylation events, producing the following effects of insulin in peripheral cells: (A) metabolic effects, promoting glucose transport, glycogen and protein synthesis, inhibition of lipolysis, protection from apoptosis, and the release of NO (described as ‘anti-inflammatory’); (B) growth- and differentiation-promoting effects, which lead to promotion of inflammation and atherogenesis (i.e. mitogenic, pro-inflammatory insulin signalling). In acute coronary syndromes, there would be a favourable balance, provided hypoglycaemia is prevented. Indicated are the points of action of drugs blocking targets in insulin signalling. IRS-1, insulin substrate receptor-1; PI3-kinase, phosphatidylinositol(PI)3-kinase; eNOS, endothelial nitric oxide synthase; JNK, c-Jun NH2-terminal kinase; MEK, MAPK (mitogen-activated protein kinase)/ERK (extracellular receptor kinase)-kinase; p38, p38 MAPK; Akt or protein kinase B (PKB); wortmannin, PI3-kinase inhibitor; PD (PD98059) and UO126, ERK 1/2 inhibitors; SB202190, p38 MAPK inhibitor.

Acute glucose lowering without inducing hypoglycaemia: a possible role for glucagon-like-peptide 1 analogues and mimetics

As normalization of hyperglycaemia is most likely beneficial in ACS, strategies that overcome the risk of hypoglycaemia might be more beneficial than insulin infusion in ACS.

Glucagon-like peptide 1 (GLP1) is an incretin hormone, secreted by the L-cells in the ileum and the colon upon oral ingestion of glucose.61 The term incretin refers to the hormonal mechanism by which insulin secretion after oral glucose ingestion is substantially higher compared with an i.v. administration.62 GLP1 is a peptide hormone with a high similarity to glucagon, which contributes to glucose regulation via several mechanisms. Specific receptors are located on pancreatic islet β- and α-cells, in the stomach, the heart, kidneys, the liver, the lung, skeletal muscle, adipose tissue, and the brain.62 At the level of pancreatic islets, in a glycaemia-dependent fashion, insulin secretion is stimulated in β-cells and glucagon secretion is suppressed in α-cells. Thus, GLP1 reduces hyperglycaemia, but does not induce hypoglycaemia.63 In addition, proliferation of β-cells is stimulated and apoptosis is reduced. Gastric emptying is also delayed by GLP1 via afferent vagal stimulation. In the liver, gluconeogenesis is reduced and glucose uptake is increased64 (Table 4). In the brain, GLP1 regulates satiety, limiting weight gain and favouring weight loss.65 In the heart, GLP1 activates anti-apoptotic signals via phosphatidyl inositol 3-kinase and cyclic adenosine monophosphate. It also increases glucose uptake and improves ventricular function, protecting the myocardium in acute ischaemic situations.66

View this table:
Table 4

Actions of incretins

Target organ/tissueActionsReferences
PancreasIncreased synthesis of insulin, suppression of glucagon secretion, increased β-cell mass or differentiation of islet precursor cells into β-cells, inhibition of β-cell apoptosis62
StomachDecreased glucose transit by slowing of gastric emptying64
LiverIncreased glucose uptake, decreased gluconeogenesis80
HeartInhibition of cardiomyocyte apoptosis, cardioprotection66
MuscleIncreased glucose uptake and glycogen synthesis81
Adipose tissueIncreased glucose uptake, lipolysis82
BrainSuppression of appetite/increased satiety, weight loss65

Although GLP1 secretion is lower in diabetic subjects, the incretin effect is intact, and infusions of GLP1 normalize glucose levels in type 2 diabetic patients without causing hypoglycaemia.67 A limitation of native GLP1 is its short half-life of 1–2 min, which only allows continuous i.v. administration. GLP1 is cleaved at the aminoterminal end by the ubiquitous dipeptidyl peptidase-4 (DPP-4) and excreted renally. Long-acting analogues of GLP1, called GLP1 receptor agonists or incretin mimetics, have been developed. Such are exenatide and liraglutide. Exenatide shows 53% homology to GLP1, lacks the alanine at position 8, and is therefore not cleaved by DPP-4.62 It is administered at doses of 5 and 10 µg twice daily for s.c. administration. Liraglutide has 97% homology with native GLP1, is conjugated with a C-16 fatty acid, and thus circulates bound to albumin and resists cleavage by DPP-4.62 It is administered once daily s.c. Notable side effects of GLP1 analogues are nausea and rarely vomiting (Figure 4).

Figure 4

Main side effects after intake of glucose-lowering drugs for 2 weeks in association with glucagon-like peptide 1 receptor agonists (incretin mimetics) or inhibitors of dipeptidyl peptidase-4 (DPP-4) activity (incretin enhancers) vs. placebo. Reference doses: exenatide: 5 µg/day; liraglutide: 2 mg/day. Inhibitors of DPP-4: 200 mg/day. From Drucker and Nauck.62

Inhibitors of DPP-4, i.e. gliptins or ‘incretin enhancers’, such as sitagliptin, vildagliptin, saxagliptin, and alogliptin, prevent the breakdown of endogenous GLP1. In contrast to GLP1 analogues, they are not associated with weight loss or delayed gastric emptying.

Glucagon-like peptide 1-based therapies for glucose control in acute coronary syndromes?

GLP1-based therapies could be useful for glucose control as they do not cause hypoglycaemia, thus avoiding the need of monitoring glucose levels or dose adjustment. Because of the still uncertain causal role of hypoglycaemia as an independent risk factor in ACS patients, a trial of such drugs would be a good tool to test the hyperglycaemia causality hypothesis. Of note, GLP1, besides being involved in glucose regulation, also exerts promising effects on the cardiovascular system. GLP1 indeed increases NO-dependent endothelial function in experimental models,68,69 as well as in type 2 diabetic patients70 and healthy subjects,71 independent of glucose levels. In an experimental heart failure model, GLP1 improved myocardial glucose uptake and left ventricular systolic function.72 Both GLP173 and its analogue exenatide74 reduce ischemia/reperfusion injury in experimental models. In diabetic patients with heart failure, left ventricular function was improved after 5 weeks of treatment.75

Data on GLP1 in ACS are still sparse. Nikolaidis et al.76 administered GLP1 for 72 h in 11 patients (5 with diabetes) with an ST-elevation AMI and left ventricular dysfunction after successful reperfusion. GLP1 improved left ventricular ejection fraction from 29 ± 2 to 39 ± 2%, whereas no changes were seen in the control group.

Cardiovascular outcomes are currently studied in diabetic patients with incretin mimetics or DPP-4 inhibitors. Some of these studies are being conducted in acute cardiac situations. Exenatide is being investigated with regard to its glucose-lowering efficacy in the CCU (ClinicalTrials.gov Identifier: NCT00736229), as well as with regard to limitation of infarct size in ST-elevation MI and primary percutaneous coronary interventions (POSTCON II, NCT00835848). Sitagliptin is studied in combination with granulocyte colony-stimulating factor with regard to global myocardial function in ACS (SITAGRAMI, NCT00650143), as well as, by itself, with regard to β-cell function (BEGAMI, NCT00627744). Alogliptin is investigated with administration started immediately after an ACS (EXAMINE, NCT00968708) in a placebo-controlled study in 5400 type 2 diabetic patients with regard to cardiovascular death, non-fatal MI, and non-fatal stroke. Further studies are however also envisaged to test GLP1-based strategies in diabetic or hyperglycaemic patients with ACS even limited to the acute setting, where hyperglycaemia remains a viable target for interventions.


Current evidence on the prognostic role of hyperglycaemia in ACS allows us to suggest its direct involvement in short-term complications. However, a causal relationship could not so far be established because of limitations with the use of insulin, as well as a lack of appropriate studies. GLP1, its derivatives, or also orally active inhibitors of DPP-4 are promising agents to be tested in this context in the near future.

Conflict of interest: none declared.


View Abstract