Prevention of Instent Renarrowing With Aggressive Glucose Lowering With Pioglitazone in Diabetic Patients

Background: Despite drug-eluting stents (DES), diabetic patients remain at high risk of restenosis and poor clinical outcomes after percutaneous coronary intervention (PCI). Studies have suggested poor glycemic control and insulin resistance may be predictors of poor outcomes after PCI. There are conflicting studies as to whether strategies to improve glycemic control can improve outcomes after PCI. Thiazolidinediones, such as pioglitazone (PIO), may have anti-restenotic benefits, independent of glycemic control.

Study design: This study was a single centre prospective, randomized, open-label, blinded-endpoint (PROBE) parallel design trial. Type 2 diabetic patients, treated with diet or oral antidiabetic medication (sulfonylurea vs. metformin or combination; but no thiazolidinedione or insulin), who are undergoing elective or urgent PCI with stenting were eligible. Fifty type 2 diabetic patients were randomly assigned to either: intensive glycemic control: pioglitazone (PIO; 30 mg/d x 6 months) in addition to titration of oral hypoglycemic agents (OHA) to get HbA1c<6% (PIO: n=25) vs. conservative glycemic control: titration of OHA to get HbA1c<7% (CONTROL: n=25). Intravascular ultrasound (IVUS) was performed immediately after PCI and repeated at 6 months to determine the effect on instent neointimal plaque volume and area. Coronary stenting was carried out in a standard fashion, with routine use of a glycoprotein 2b/3a inhibitor during the procedure. From August 2002 until June 2005, DES were not permitted in the protocol. After June 2005, we amended the protocol to allow DES, as they had become routinely used in diabetic patients in our institution, especially for vessel size <3mm and/or lesion length>15mm. DES were used in 7 PIO and 11 CONTROL subjects, and bare metal stents (BMS) in the rest. Patients were then followed with clinic visits at 1, 3 and 6 months. OHA, other than pioglitazone, were adjusted in a stepwise manner in order to attain the HbA1c targets. Other concomitant medications, including anti-anginals, lipid-lowering therapy, and antihypertensive medication were adjusted according to their clinical need and current Canadian guidelines. After 6 months treatment, or before if clinically indicated, all subjects were to return for repeat cardiac catheterization, including repeat coronary angiography and IVUS of the intervened vessel to assess the serial change in luminal dimensions. Fasting blood was collected for plasma glucose, HbA1c, insulin, lipid profile, hs-CRP, adiponectin, leptin, matrix metalloproteinase-9, and interleukin-6 at the time of PCI and at the follow-up IVUS. If the patient developed recurrent ischemic symptoms before 6 months, the final IVUS could be performed earlier, if they were found to have clinically-significant restenosis (diameter stenosis > 50%). Otherwise, patients were still encouraged to have their protocol 6 month IVUS follow-up. 41 patients (n=20 PIO, n=21 CONTROL) had analyzable pairs of IVUS.

Study hypothesis: We hypothesized that there would be significantly less instent neointimal proliferation on IVUS at 6 months in the group receiving aggressive glycemic control plus the thiazolidinedione pioglitazone. We also hypothesized that the reduction in neointimal hyperplasia will likely relate to improvements in glycemic control (HbA1c) and insulin resistance. Additionally, we wanted to explore the biochemical predictors (glucose parameters, lipids, inflammatory markers, adipokines) for neointimal proliferation.