Combination therapy with pioglitazone plus metformin or sulfonylurea in patients with Type 2 diabetes Influence of prior antidiabetic drug regimen

Purpose: To evaluate two trials of combination therapy, pioglitazone together with metformin or sulfonylurea, for Type 2 diabetes mellitus and to examine how pretrial antidiabetic therapies may have influenced the results. Subjects and Methods: The results of two published trials that examine combination therapy, pioglitazone plus metformin or pioglitazone plus sulfonylurea were analyzed. A post hoe analysis was performed in which patients from both of these trials were subdivided on the basis of their oral antidiabetic therapy before enrollment into the trials. Those subsets receiving pretrial therapy with a sulfonylurea plus metformin -and discontinuing one of the agents before randomization -were compared to those receiving only one of the agents before enrollment. Results: Patients in the combination pioglitazone (30 mg/day) plus metformin therapy ann of one trial, who entered with stable metformin monotherapy, experienced a significant decrease in glycosylated hemoglobin (HbA(1C)) levels during the 16-week course of the trial (-1.0+/-0.1 [mean+/-S.E.] percentage points; P<.05). In contrast, those in the same arm of the trial whose pretrial therapy included metformin plus a sulfonylurea-and who discontinued the sulfonylurea before enrollment -experienced no significant change in HbA(1C) levels (0.2 +/- 0.2 percentage points; P>.05). The difference between groups was significant (P<.001). Patients in the combination pioglitazone (15 or 30 mg/day) plus sulfonylurea therapy arm of the other trial, who entered the trial on stable sulfonylurea monotherapy, experienced significant decreases in HbA(1C) levels (-1.0+/-0.1 and -1.4+/-0.1 percentage points, respectively, for the 15- and 30-mg pioglitazone arms; P<.05). Those in the same arm whose pretrial therapy included sulfonylurea plus metformin-and who discontinued the metformin before enrollment- experienced no significant change in HbA(1C) levels. Differences between the groups separated on the basis of pretrial antidiabetic regimen were significant (P < .001). Conclusions: The efficacy of pioglitazone as an add-on in combination therapy (with metformin or a sulfonylurea) is likely to be greater than in previous reports, because those trial results were influenced by treatment regimens used by a portion of the patients before enrollment. Trials of oral hypoglycemic agents must be carefully constructed to consider the effects of agents given prior to the trial. (C) 2003 Elsevier Inc. All rights reserved.