Rationale Rosiglitazone and troglitazone both promote differentiation of pre-adipocytes into adipocytes in subcutaneous, but not omental fat, and reduce gamma glutamyl transferase, a surrogate marker for intra-abdominal and hepatic fat. Troglitazone has been shown by abdominal computed tomography (CT) and magnetic resonance imaging (MRI) scan to reduce the intra-abdominal adipose tissue area in type 2 diabetics. Similarly rosiglitazone has also been shown to increase subcutaneous but not intra-abdominal fat in patients with type 2 diabetes. In the same study it was also shown to cause a substantial reduction in hepatic fat. Central fat depots are believed to be associated with more cardiovascular risk than subcutaneous fat depots. By contrast, sulphonylurea-associated weight gain has been shown by abdominal CT scan to include increases in intra-abdominal adipose tissue.

The aim of this study is to compare the body fat distribution pattern of glibenclamide plus rosiglitazone versus glibenclamide and placebo (especially the intra-abdominal adipose tissue and abdominal subcutaneous adipose tissue areas) in patients with type 2 diabetes. It is hypothesised that rosiglitazone will lead to the accumulation of excess energy stores in the subcutaneous rather than the intra-abdominal adipose tissue depot. In addition, it is hoped that by having a positive effect on diastolic blood pressure, lipid levels, BMI, rosiglitazone will be shown to have a better cardiovascular risk profile when used in combination with glibenclamide rather than when glibenclamide is used alone.

Although insulin resistance has been shown to be a primary defect causing type 2 diabetes mellitus, insulin secretory defect has also been known to be an important factor in the development of type 2 diabetes mellitus. A previous study has shown that in Korean patients, early-phase insulin secretory defect may be the initial abnormality in the development of type 2 diabetes mellitus [56].

This study also aims to assess the efficacy and safety of glibenclamide plus rosiglitazone versus glibenclamide plus placebo therapy in Korean patients with type 2 diabetes. In addition, a previous study has shown that in Korean patients, early-phase insulin secretory defect may be the initial abnormality in the development of type 2 diabetes mellitus. This study aims to show that rosiglitazone treatment in Korean patients, regardless of their early phase insulin secretory ability, is just as efficacious and safe.

Objective(s) Primary To evaluate the effect of 12 months oral treatment with glibenclamide plus rosiglitazone versus oral glibenclamide plus placebo, on body fat distribution (as measured by the change in the ratio between the intra-abdominal adipose tissue and abdominal subcutaneous adipose tissue areas) in patients with type 2 diabetes.

Secondary

• To investigate the efficacy of glibenclamide plus rosiglitazone, compared to glibenclamide plus placebo on beta-cell function and insulin resistance as calculated by HOMA-B and HOMA-R.
• To investigate the efficacy of glibenclamide plus rosiglitazone, compared to glibenclamide plus placebo on fasting plasma glucose, insulin, fasting serum lipid profile (total cholesterol, triglycerides, HDL cholesterol, LDL cholesterol and total cholesterol to HDL cholesterol ratio).
• To investigate the efficacy of glibenclamide plus rosiglitazone, compared to glibenclamide plus placebo on early phase insulin secretion during an oral glucose tolerance test as measured by the insulinogenic index.
• To define further the clinical safety and tolerability of glibenclamide plus rosiglitazone through the assessment of physical examinations, vital signs, weight, routine laboratory tests, adverse experiences and electrocardiograms (ECGs).

Endpoint(s) Primary Change from baseline in the ratio (IAAT:SAT) between the intra-abdominal adipose tissue area (IAAT) and abdominal subcutaneous adipose tissue area [SAT] after 12 months treatment with oral glibenclamide plus rosiglitazone compared to oral glibenclamide plus placebo Secondary

Comparisons will be made between glibenclamide plus rosiglitazone and glibenclamide plus placebo treatment groups on Change from baseline after 6 and 12 months treatment with respect the following:

CT Scan

Derived from CT image at the lumbar IV level:

• abdominal subcutaneous adipose tissue area [SAT]
• intra-abdominal adipose tissue area [IAAT]

Derived from the CT image of the right leg at the thigh level (1cm below the gluteal fold):

• total subcutaneous adipose tissue area [TSAT] Derived from CT images at the lumbar IV and thigh level
• ratio between abdominal subcutaneous adipose tissue area [SAT] and total subcutaneous adipose tissue area of the thigh [TSAT]
• ratio between intra-abdominal adipose tissue area [IAAT] and total subcutaneous adipose tissue area of the thigh [TSAT] Derived from Oral Glucose Tolerance Test, glycaemic response to OGTT, difference

Subjects were selected by following considerations:

• Diagnosis of type 2 diabetes mellitus defined by the American Diabetic Association (ADA) criteria.
• Subject whose diabetes was managed by diet, exercise, and/or sulfonylurea, and/or metformin, who could be converted to treatment with glibenclamide 5mg om.
• 126mg/dL ≤ fasting plasma glucose (FPG) ≤ 270mg/dL and HbA1c level >7% at screening. In addition, the following criterion was applied prior to randomization: 126mg/dL ≤ fasting plasma glucose (FPG) ≤ 270mg/dL at baseline(after 2 months of run-in treatment with glibenclamide 5mg om only).