Comparison of stepwise addition of prandial insulin to a basal-bolus regimen when basal insulin is insufficient for glycaemic control in type 2 diabetes: Results of the OSIRIS study

Aim. - The metabolic efficacy of adding prandial insulin in a stepwise manner to a straightforward basal-bolus regimen was compared in patients with type 2 diabetes mellitus (T2DM), suboptimally controlled by oral antidiabetic drugs (OADs) and once-daily basal insulin. Methods. - In this international randomized, parallel-group, non-inferiority study, 811 patients with poorly controlled type 2 diabetes using basal insulin were switched to insulin glargine (GLAR) for 6 months while continuing OADs. Patients with HbA(1c) > 7% and FPG < 120 mg/dL (n = 476) were then randomized to either group 1, GLAR + metformin (MET) + 3 x insulin glulisine (GLU), group 2, GLAR + MET + 1-3 x GLU, or group 3, GLAR + MET + insulin secretagogue (IS) + 1-3 x GLU, for 12 months. Objectives were to show the non-inferiority of efficacy of group 2 vs group 1 and vs group 3. Non-inferiority of group 2 vs group 1 was concluded if the upper limit of the 95% confidence interval (CI) for the HbA(1c) difference was <= to 0.4%. Results. - The adjusted HbA(1c) difference of group 2 vs 1 for the per-protocol population crossed the non-inferiority margin (0.228, 95% CI: -0.018-0.473). There was significantly less weight gain in group 2 compared with group 1, but adverse events were otherwise similar between the two groups. In patients with HbA(1c) < 8% at baseline, non-inferiority was achieved in group 2 vs group 1. Conclusion. - Although non-inferiority was not achieved, stepwise intensification of GLU added to GLAR showed efficacy close to that of the basal-bolus approach and with significantly less weight gain. (C) 2012 Elsevier Masson SAS. All rights reserved.