engleski

Logistic regression analysis for identifying type 2 diabetics with poor response to sulfonylurea therapy

Background: Sulfonylureas are used as first- line therapy drugs in type 2 diabetics ; however, therapy failure occurs in approximately 30% of patients. We hypothesize that at least one part of the cause lies in genetic factors. Sulfonylureas are insulin secretagogues that act by binding to pancreatic beta cells sulfonylurea receptor (SUR-1) encoded by two genes: ABCC8 and KCNJ11. The aim of this study was create a logistic regression model to identify patients with poor response to sulfonylurea therapy using demographic (age, gender), anthropometric (body mass index, BMI), biochemical (fasting and postprandial glucose and lipid concentration) and genetic (ABCC8 polymorphisms: SUR1 exon 16 (-3C/T), SUR-1 exon 31 (Arg1273Arg) and SUR-1 exon 33 (S1369A) and KCNJ11 polymorphism E23K) parameters. Methods: The study included 251 unrelated type 2 diabetics on sulfonylurea therapy. Polymorphisms were detected using polymerase chain reaction - restriction fragment length polymorphism. Biochemical parameters were determined on automatic analyzer Olympus AU 2700 (Olympus, Hamburg, Germany). Based on the HbA1c concentration, patients were divided into two groups: patients with HbA1c concentration over 53 mmol/mol Hb were classified as patients with poor outcome. All variables are analysed using univatiate regression analysis. Statistically significant variables were included into multivariate regression model. Results: Univariate regression analysis identified following variables as statistically significant: fasting glucose (OR (95% CI) = 1.60 (1.37-1.88)) ; postprandial glucose (1.28 (1.17-1.40)) ; SUR1 exon 16 polymorphism (1.58 (1.06-2.36)) and SUR-1 exon 31 polymorphism (0.51 (0.33-0.77)), while age, gender, BMI and lipid concentration didn't differ significantly across subgroups. After inclusion into stepwise multivariate regression model, fasting and postprandial glucose and SUR-1 exon 31 polymorphism remained significant (1.45 (1.22- 1.74) ; 1.14 (1.03-1.27) and 0.48 (0.30-0.80), respectively). Percent of correctly classified cases was 74.10%. Conclusion: Type 2 diabetics with high fasting and postprandial glucose and wild type allele of the SUR-1 exon 31 (Arg1273Arg) polymorphism are more likely to have poor glycemic control expressed as high HbA1c concentration.

diabetes melitus ; sulfonylurea ; polymorphism ; pharmacogenetics

nije evidentirano