Aim: To assess the efficacy and tolerability of vildagliptin compared with thiazolidinediones (TZDs) as an add on to metformin treatment in a primary care patient population with type 2 diabetes.
Methods: This was a randomized, 12-week, open-label study comparing vildagliptin (100 mg, n = 1653) and TZD (agent and dose at the investigators' discretion, n = 825) add-on therapy in patients inadequately controlled [haemoglobin A(1C) (HbA(1c)): 7-10%] on a stable dose of metformin (> or =1000 mg/day). The primary objective was to test non-inferiority of vildagliptin to TZDs for the difference in change in HbA(1c) from baseline [established if the upper limit of the two-sided 95% confidence intervals (CI) did not exceed 0.4%].
Results: Mean (+/- s.e.) change in HbA(1c) from baseline to study endpoint was -0.68 +/- 0.02% in the vildagliptin group and -0.57 +/- 0.03% in the TZD group. The difference between groups was -0.11% (95% CI: -0.17% and -0.04%), establishing the non-inferiority of vildagliptin (p = 0.001) after 3 months of treatment. Vildagliptin was non-inferior to TZDs for subgroups of race, age and body mass index. Body weight increased in the TZD group (0.33 +/- 0.11 kg) and decreased in the vildagliptin group (mean: -0.58 +/- 0.09 kg; p < 0.001 for difference). Adverse events occurred in similar proportions of patients in both groups (vildagliptin: 39.5% and TZD: 36.3%) Hypoglycaemia and abnormal changes in liver enzymes were uncommon.
Conclusions: This short-term study suggests that vildagliptin is as effective as TZDs after 3-month treatment as an add-on to metformin in a primary care population that included diverse patient subgroups.