Principal Findings:

A total of 543 patients were randomized, 270 to the pioglitazone arm, and 273 to the glimepiride arm. The baseline clinical characteristics were fairly comparable between the two arms, except hypertension and former smokers, which were more frequent in the glimepiride arm. The mean baseline glycosylated hemoglobin (HbA_1C) was 7.4% in both arms. The levels of fasting blood glucose, insulin, low-density lipoprotein (LDL), triglycerides, and C-reactive protein were similar between the two arms; high-density lipoprotein (HDL) cholesterol was slightly higher in the glimepiride arm (43.4) compared with the pioglitazone arm (40.8).

The primary efficacy endpoint, least square mean change in percent atheroma volume from baseline, increased significantly in the glimepiride arm (0.73%, 95% confidence interval [CI] 0.33-1.12%, p < 0.001), whereas it decreased slightly in the pioglitazone arm (0.16%, 95% CI -0.57 to 0.25%, p = 0.44) (p = 0.002 between the two arms). Similarly, the change in maximum atheroma thickness increased by 0.011 mm (95% CI -0.0002 to 0.022 mm) in the glimeperide arm, and decreased by 0.011 mm (95% CI -0.002 to 0.0004 mm) in the pioglitazone arm (p = 0.006 between the two arms). Another secondary outcome, the change in 10 mm of the most diseased segment, showed no difference between the two groups.

There was a 0.19% greater reduction in mean HbA_1c levels in the pioglitazone arm compared with glimepiride (-0.55% vs. -0.36%, p = 0.03), although this difference was not apparent until 2 years of follow-up. Compared with baseline, there was a greater reduction in fasting insulin levels (-5.0 µU/ml vs. 1.33 µU/ml, p < 0.001), a greater increase in HDL (5.7 vs. 0.9, p < 0.001), a greater reduction in triglycerides (-16.3 vs. 3.3, p < 0.001), a greater reduction in hs-CRP (-1.0 vs. -0.4, p < 0.001), a smaller increase in systolic blood pressure (0.1 vs. 2.3, p = 0.03), a greater decrease in diastolic blood pressure (-0.9 vs. 0.9, p = 0.003), but no significant increase in LDL cholesterol (2.1 vs. 1.1, p = 0.69) in the pioglitazone arm compared with the glimepiride arm, respectively.

The composite endpoint of cardiovascular death, nonfatal myocardial infarction, or stroke was similar between the glimepiride and pioglitazone arms (2.2% vs. 1.9%, respectively; p = 0.78). Similarly, the incidence of hospitalization for congestive heart failure and coronary revascularization was 1.8% and 1.5% (p = 1.0), and 11.0% and 10.7% (p = 0.93), for glimepiride and pioglitazone, respectively. The incidence of side effects between pioglitazone and glimepiride, respectively, was: hypoglycemia (15.2% vs. 37.0%, p < 0.001), bone fractures (3.0% vs. 0%, p = 0.004), peripheral edema (17.8% vs. 11.0%, p = 0.02), and angina (7.0% vs. 12.1%, p = 0.05). Serum urea nitrogen >30 mg/dl was significantly higher in the pioglitazone arm (10.7%) compared with the glimepiride arm (4.8%) (p = 0.009), although serum creatinine >2.0 mg/dl was similar between the two arms (1.1% vs. 0.7%, p = 0.12). Hemoglobin decrease >3 g/dl was also significantly higher in the pioglitazone arm (4.1%) compared with the glimepiride arm (0.7%) (p = 0.01).