Angiographic and clinical outcomes of rosiglitazone in patients with type 2 diabetes mellitus after percutaneous coronary interventions: a single center experience.

A beneficial effect of thiazolidinediones includes the reduction of intermediate markers, suggesting a potential for reducing atherosclerosis and restenosis. The objective of this study was to determine if rosiglitazone (RSG) reduced the odds of restenosis and if RSG improved the odds of clinical outcomes after percutaneous coronary intervention (PCI) in type 2 diabetes mellitus (DM) patients. A total of 609 patients with 734 lesions were selected from the period between January 1, 2001 and January 31, 2004. These patients were divided into 2 groups: a "control" group representing patients seen between January 1, 2001 and September 2002 when RSG was not available in our hospital and a "RSG treatment" group representing patients seen between September 2002 and January 31, 2004 when RSG was available in our hospital. Thus, 213 patients with 253 lesions (1.19 L/P) were placed in the RSG group and 396 patients with 481 lesions (1.21 L/P) were placed in the control group. Subgroup analysis based on the PCI received had 88 patients in the RSG arm receiving balloon angioplasty and 125 patients receiving coronary stenting; the control group had 187 and 209 patients, respectively, in the subgroups. Primary endpoint was angiographic restenosis at 6 months, and secondary endpoints were death, myocardial infarction, and target lesion revascularization. More patients in the control group were insulin-requiring, had poorer left ventricular function, but had a larger preprocedural minimal lumen diameter (pre-MLD). At 6 months, restenosis and reocclusion rates were lower in the RSG group (P = .014 and P = .006, respectively). Twenty-nine patients died in the control group versus 1 in the RSG group (P <= .001). RSG (P = .019), stenting (P = .005), preprocedural reference vessel diameter (P = .017), metformin (P= .022), pre-MLD (P < .001), hyperlipidemia (P = .016), and combined RSG and metformin (P = .020) were predictors of restenosis, while RSG (P = .016) and metformin (P = .029) were predictors of survival. In conclusion, RSG was found safe and well tolerated and was associated with reduced odds of restenosis, reocclusion, and mortality rates in type 2 DM patients independent of glycemic control and PO performed.