The E23K variant of KCNJ11 encoding the pancreatic beta-cell adenosine 5'-triphosphate-sensitive potassium channel subunit Kir6.2 is associated with an increased risk of secondary failure to sulfonylurea in patients with type 2 diabetes.

CONTEXT: Several studies suggest that genetic factors may play a role in the different responses to antidiabetic therapy; however, conclusive evidence is still lacking. OBJECTIVE: The objective of the study was to investigate whether diabetic patients carrying the E23K variant in KCNJ11 are at increased risk for secondary sulfonylurea failure. DESIGN: Secondary sulfonylurea failure was defined as fasting plasma glucose greater than 300 mg/dl despite sulfonylurea-metformin combined therapy and appropriate diet, in the absence of other conditions causing hyperglycemia. SETTING: The study was conducted in an ambulatory care facility. PATIENTS: A total of 525 Caucasian type 2 diabetic patients were enrolled in the study. INTERVENTION: Sulfonylurea treatment was followed by sulfonylurea-metformin combined therapy and then insulin treatment. MAIN OUTCOME MEASURE: Secondary failure was the main outcome measure. RESULTS: Of the diabetic patients enrolled in the study, 38.5% were E23E homozygous, 51.4% were E23K heterozygous, and 10.1% were K23K homozygous. The frequency of carriers of the K allele was 58 and 66.8% among patients treated with oral therapy or secondary sulfonylurea failure, respectively (odds ratio, 1.45; 95% confidence interval, 1.01-2.09; P = 0.04). Adjustment for age, gender, fasting glycemia, glycosylated hemoglobin, age at diagnosis, and duration of diabetes in a logistic regression analysis did not change this association (odds ratio, 1.69; 95% confidence interval, 1.02-2.78; P = 0.04). Islets isolated from carriers of the K allele showed no differences in glucose-stimulated insulin secretion and a tendency toward reduced response upon glibenclamide stimulation (P = 0.09). After 24-h exposure to high (16.7 mmol/liter) glucose concentration, impairment of glibenclamide-induced insulin release was significantly (P = 0.01) worse with the E23K variant. CONCLUSIONS: These data suggest that the E23K variant in KCNJ11 may influence the variability in the response of patients to sulfonylureas, thus representing an example of pharmacogenetics in type 2 diabetes.