Response to Letter Regarding Article, “Differential Clinical Outcomes Associated With Hypoglycemia and Hyperglycemia in Acute Myocardial Infarction”
Dr Chaudhuri and colleagues propose that insulin therapy in the form of glucose-insulin-potassium (GIK) infusion may have lowered mortality in hyperglycemic patients with acute myocardial infarction in the Clinical Trial of Reviparin and Metabolic Modulation in Acute Myocardial Infarction Treatment and Evaluation–Estudios Clinicos Latino America (CREATE-ECLA) and Organization for the Assessment of Strategies for Ischemic Syndromes-6 (OASIS-6) trials. Their hypothesis is based on our Table 6,1 which shows that among patients with postadmission hyperglycemia (glucose ≥140 mg/dL), the unadjusted 30-day mortality rate was 8.3% in patients who received GIK compared with 9.7% in those who did not receive GIK. However, comparing outcomes between GIK-treated and untreated patients with postadmission hyperglycemia is biased and potentially misleading because GIK itself raises glucose levels, and this is a postrandomization comparison of a subgroup of patients. A more valid analysis compares outcomes between GIK-treated and untreated patients who had hyperglycemia on admission because admission glucose levels are not yet altered by GIK infusion, and this comparison preserves randomized groups and obviates the need for postrandomization statistical adjustments. In the combined Clinical Trial of Reviparin and Metabolic Modulation in Acute Myocardial Infarction Treatment and Evaluation–Estudios Clinicos Latino America (CREATE-ECLA) and Organization for the Assessment of Strategies for Ischemic Syndromes-6 (OASIS-6) trials,2 22 943 patients were randomized to GIK or control, of whom 10 729 had admission hyperglycemia (glucose ≥140 mg/dL). Mortality at 30 days occurred in 663 of 5401 (12.3%) of these patients randomized to GIK compared with 654 of 5328 (12.3%) control patients (P=1.0). Thus, GIK did not improve outcomes in patients with admission hyperglycemia. In addition, despite several mechanistic studies suggesting the cardioprotective effects of insulin, GIK may increase mortality at 3 days after acute myocardial infarction, with no overall benefit at 30 days or 6 months.2 Therefore, administration of insulin (as fixed-dose GIK infusion) cannot be recommended in acute myocardial infarction. Our posthoc analysis is also insufficient to support giving insulin to lower glucose to a range of 70 to 140 mg/dL in patients with acute myocardial infarction; such a strategy requires evaluation in adequately powered clinical trials.
Dr Gerstein has received honoraria from Sanofi-Aventis, GlaxoSmithKline, Eli Lilly, and Novo Nordisk; his institution has received research grants for studies led by Dr Gerstein from Sanofi-Aventis and GlaxoSmithKline; and he has served on advisory board for Sanofi-Aventis, GlaxoSmithKline, Eli Lilly, and Novo Nordisk. The other authors report no disclosures.