Background and aim: This study was designed to compare the efficacy of rosiglitazone and glibenclamide in individuals with type 2 diabetes over a 12-month period.
Methods and results: A total of 598 patients were randomized to double-blind treatment for 52 weeks with rosiglitazone 4 mg/d (n=200), rosiglitazone 8 mg/d (n=191) or glibenclamide (n=207; dose adjusted up to 15 mg/d over the first 12 weeks according to clinical response). Changes in fasting plasma glucose (FPG), haemoglobin A1c (HbA1c), fasting insulin and its precursor peptides, and lipids were measured and safety was evaluated. Significant reductions in HbA1c levels at 52 weeks compared with baseline were seen in all treatment groups (rosiglitazone 4 mg/d=-0.3%, P=0.0003; rosiglitazone 8 mg/d=-0.5%, P<0.0001; glibenclamide=-0.7%, P<0.0001). Mean FPG levels were also significantly reduced in all treatment groups (rosiglitazone 4 mg/d=-1.4 mmol/l; rosiglitazone 8 mg/d=-2.3 mmol/l; glibenclamide=-1.7 mmol/l; P<0.0001 vs. baseline for all treatments). Rosiglitazone therapy reduced plasma insulin, proinsulin, split proinsulin and free fatty acid levels compared with glibenclamide. Rosiglitazone improved insulin resistance while a worsening was seen with glibenclamide. Total:high-density lipoprotein cholesterol ratios were reduced with glibenclamide and unchanged with rosiglitazone. All treatments were generally well tolerated.
Conclusions: The efficacy of rosiglitazone 8 mg/d in improving glycaemic control in patients with type 2 diabetes is comparable to that of glibenclamide. However, rosiglitazone reduced insulin resistance and proinsulin levels whereas glibenclamide use was associated with an increase in fasting insulin and proinsulin. This suggests that in the long term, rosiglitazone may protect the beta-cell whereas glibenclamide is likely to increase the burden.