Glycemic control and treatment failure with pioglitazone versus glibenclamide in type 2 diabetes mellitus: a 42-month, open-label, observational, primary care study.

BACKGROUND

Insulin resistance and declining beta-cell function are the core defects in type 2 diabetes mellitus. It has been suggested that deteriorating glycemic control is related to baseline hemoglobin A(1c) (HbA(1c)) values and remaining beta-cell function.

PATIENTS AND METHODS

We report glycemic data from a 3.5-year, open-label, observational, primary care study comparing 30 mg/day pioglitazone with 3.5 mg/day glibenclamide add-on to stable metformin monotherapy in 500 patients with type 2 diabetes. Insulin commencement was considered for patients with HbA(1c) > or = 8.0% or when vascular complications occurred. The change in HbA(1c) compared with baseline and the difference in time to failure to maintain glycemic control were calculated.

RESULTS

At endpoint, HbA(1c) had decreased by 1.0% in the pioglitazone group (p < 0.005) and by 0.6% in the glibenclamide group (p < 0.05). Annual progression rates to insulin treatment were 6.6% (pioglitazone) and 16.4% (glibenclamide; p < 0.001 between-group difference). Mean weight increases of 3.5 +/- 0.42 kg in the pioglitazone group and 3.3 +/- 0.38 kg in the glibenclamide group were noted. Overall, both treatments were well tolerated.

CONCLUSIONS

Pioglitazone add-on to metformin revealed significant benefits in long-term glycemic control compared with glibenclamide. This difference may be explained by a large between-group difference in HOMA-S, which was shown to correlate significantly to the change in HbA(1c). This suggests that a strategy to reduce insulin resistance to lower the burden of the beta-cell is superior to treatment with glibenclamide.