Troglitazone reduces hyperglycaemia and selectively acute-phase serum proteins in patients with Type II diabetes

Aims/hypothesis. Inflammation could play a part in insulin resistance. Thiazolidinediones, new antidiabetic drugs, possess anti-inflammatory effects in vitro. We investigated if acute-phase serum proteins are increased in patients with Type II (non-insulindependent) diabetes mellitus who had been treated with insulin and whether troglitazone has anti-inflammatory effects in vivo. Methods. A total of 27 patients (age 63.0 ‐ 1.7 years, HbA 1c 8.8 ‐ 0.3 %, BMI 32.7 ‐ 0.8 kg/m 2 , duration 15.2 ‐ 1.4 years, insulin dose 73.3 ‐ 7.0 U/day) participated in the study. The patients received daily either 400 mg troglitazone or placebo for 16 weeks. Blood samples were taken at baseline, at the end of therapy and after a follow-up time of 23 ‐ 4 days. Results. The concentrations of serum amyloid A (6.2 ‐ 1.1 mg/l) and C-reactive protein (6.1 ‐ 1.1 mg/ l) were increased (p < 0.001) and complement protein C3 (1.69 ‐ 0.05 g/l) was also above the reference range for healthy subjects. Placebo treatment had no effect on glucose or inflammation, whereas troglitazone reduced fasting glucose (from 10.4 ‐ 0.6 mmol/l to 8.1 ‐ 0.5 mmol/l, p < 0.01), HbA 1c (from 8.7 ‐ 0.3 % to 7.5 ‐ 0.3 %, p < 0.01), insulin requirements (from 75 ‐ 10 U/day to 63 ‐ 10 U/day, p < 0.05), serum amyloid A (from 6.3 ‐ 1.5 mg/l to 4.0 ‐ 1.3 mg/l, p = 0.001), a-1-acid glycoprotein (from 906 ‐ 51 mg/l to 729 ‐ 52 mg/l, p = 0.001) and C3 (from 1.72 ‐ 0.07 g/l to 1.66 ‐ 0.06 g/l, p < 0.05) but not a-1-antitrypsin, ceruloplasmin, C-reactive protein or haptoglobin significantly. Concentrations of glucose and acute-phase reactants had returned to those before treatment at the follow-up visit. Conclusion/interpretation. In Type II diabetic patients serum amyloid A and complement protein C3 are raised. Troglitazone exerts a selective reversible action on some acute-phase proteins and C3 but not on others in conjunction with the improvement in glucose metabolism. [Diabetologia (1999) 42: 1433‐ 1438]