Insulin vs. strict blood glucose control to achieve a survival benefit after AMI?

Patients with diabetes mellitus have a 1.5–2 times higher risk of death after myocardial infarction than do nondiabetic patients. In diabetic myocardium, the consumption of fatty acid as metabolic fuel is thought to be increased, whereas glycolysis is impaired both in ischaemic and non-ischaemic areas. As the consumption of fatty acids instead of glucose requires more oxygen, such a shift may be deleterious particularly when oxygen supply is limited. More than four decades ago, the concept of infusing glucose together with insulin and potassium (GIK) to protect the ischaemic myocardium was introduced. Early clinical studies on the effect of such a ‘metabolic cocktail’ yielded promising results suggesting that GIK might be a way to reduce morbidity and early mortality in patients with an acute myocardial infarction (AMI). As the possible mechanism behind a cardioprotective effect of GIK, Opie proposed the promotion of glycolysis in cardiomyocytes and the diversion of fatty acids to adipocytes thereby reducing oxygen consumption in the heart. With the advent of aggressive and acute revascularization strategies (thrombolysis and percutaneous transluminal coronary angioplasty), the potential of metabolic supportive therapy with GIK was not further pursued. In 1995, the Diabetes and Insulin-Glucose infusion in AMI (DIGAMI) study was the first to randomize diabetic patients with an AMI to either ‘intensive insulin therapy’ or standard treatment. Intensive insulin therapy comprised intravenous infusion of GIK as soon as possible after the AMI and continued for 48 h. Thereafter, patients in the intensive insulin therapy group were submitted to a ‘stricter’ blood glucose control regimen with subcutaneous insulin continued for 3 months after discharge. This intervention resulted in a mean blood glucose level after 24 h of 9.6 mmol/L when compared with a significantly higher 11.7 mmol/L in the control group. At hospital discharge (8.2 vs. 9.0 mmol/L) and 3 months follow-up (8.5 and 9.0 mmol/L), small differences in blood glucose control between the two study groups were maintained. In the intensive treatment arm, mortality risk at 1 year was reduced by a relative 29%. In addition, there was a significant decrease in re-infarction and new heart failure. It remained unclear, however, how much of the benefit was due to an acute effect of GIK and how much was mediated by strict blood glucose control with insulin in the days and months after the AMI. Therefore, Malmberg et al. embarked on a subsequent study, DIGAMI 2. In this study, patients with diabetes mellitus and AMI were randomized into three groups. The first group received GIK infusion for 24 h followed by an insulin-based long-term glucose control aiming for the current target for fasting blood glucose levels in diabetes mellitus of 5–7 mmol/L. The second group received GIK infusion for 24 h followed by ‘standard’ glucose control. The third group received ‘routine management’ according to the local practice. This was indeed the correct study design to answer the question that was raised and the authors ought to be congratulated for their intention to scrutinize their initial findings. Unfortunately, however, several unanticipated problems arose which substantially limited possible conclusions from this study. The first obstacle was that the set target for fasting blood glucose of 5–7 mmol/L for group 1 was never reached. More dramatically, the level of blood glucose control ended up being identical in all three study groups, and this despite the fact that