Bristol-Myers Squibb Company and Merck & Co, Inc. announced results from both a phase III active-controlled study and a long-term phase II dose-ranging study for the investigational oral medicine muraglitazar (now known under the registered trade name Pargluva) during a late-breaking presentation of the 65th Annual Scientific Sessions of the American Diabetes Association (ADA).

The primary endpoint of each study was reduction in A1C levels (a measure of a person's average blood glucose over a two- to three-month time period) in patients with type 2 diabetes. Lipid effects, a secondary endpoint of both studies, were also measured.

The phase III study showed Pargluva 5 mg reduced A1C levels significantly more than pioglitazone1 30 mg in patients with type 2 diabetes. Both treatment groups were taking metformin (a drug used to treat type 2 diabetes). Significant effects were also seen on triglycerides and high-density lipoprotein cholesterol (HDL-C) levels. In the phase II study at 104 weeks, patients who received Pargluva 5 mg had a change of -1.52 per cent from baseline A1C (mean baseline of 7.92 per cent) with a final mean A1C level of 6.4 per cent.

Bristol-Myers Squibb and Merck are collaborators in the global development and commercialization of Pargluva. The New Drug Application (NDA) for Pargluva is currently under review by the US Food and Drug Administration (FDA). If approved, Pargluva would become the first marketed agent in a new class of investigational compounds called glitazars. Pargluva is a dual alpha/gamma PPAR (peroxisome proliferator-activated receptor) activator.

This phase III, double-blind, randomized, active-controlled trial was conducted in 1,159 patients with type 2 diabetes who had inadequate glycemic control on metformin alone. Patients were assigned to either a once-daily dose of Pargluva 5 mg (n=587) or pioglitazone 30 mg (n=572) for 24 weeks while continuing mean daily metformin doses of 1,854 mg and 1,851 mg in the Pargluva and pioglitazone groups, respectively. Starting at week 12, statin therapy could be initiated or adjusted as needed.

All analyses on mean change from baseline were adjusted for baseline level and used the last observation carried forward (LOCF) methodology. The mean baseline A1C value (8.1 per cent) and other disease characteristics were comparable between the two groups.

Pargluva Significantly Reduced A1C and Fasting Plasma Glucose
At week 24, the addition of Pargluva 5 mg reduced mean A1C values from baseline by 1.14 per cent compared with the addition of pioglitazone 30 mg, which resulted in a reduction of 0.85 per cent (p-value less than 0.0001 between treatment groups). Sixty per cent of patients taking Pargluva 5 mg attained the ADA recommended A1C goal (less than 7 per cent), compared to 45 per cent of patients taking pioglitazone 30 mg (p-value less than 0.0001 between treatment groups).

Pargluva 5 mg also reduced fasting plasma glucose, a secondary endpoint, more than pioglitazone 30 mg [-44 mg/dL from baseline versus -33 mg/dL from baseline, respectively (p-value less than 0.0001 between treatment groups)].

At week 12, triglyceride levels decreased 28 per cent and 14 per cent versus baseline in patients treated with Pargluva 5 mg and pioglitazone 30 mg, respectively (p-value equal to 0.0001 between treatment groups). In patients with a triglyceride level above 150 mg/dL at baseline, the reduction was 35 per cent and 19 per cent for Pargluva 5 mg and pioglitazone 30 mg, respectively (p-value less than 0.0001 between treatment groups). HDL-C levels increased 19 per cent in the Pargluva 5 mg group and 14 per cent in the pioglitazone 30 mg group (p-value equal to 0.0001 between treatment groups). In addition, non-HDL-C levels decreased 6 per cent, apoB levels decreased 12 per cent, and free fatty acids decreased 30 per cent in the Pargluva 5 mg group compared to decreases of 1 per cent, 6 per cent and 21 per cent, respectively, in the pioglitazone 30 mg group (all p-values less than or equal to 0.0001 between groups). There were no significant effects on LDL-C in either group.

At week 24, Pargluva 5 mg reduced fasting plasma insulin more than pioglitazone 30 mg [-5.0 microunits/mL from baseline versus -3.6 microunits/mL from baseline, respectively (p-value less than 0.0001 between treatment groups)]. Treatment with Pargluva 5 mg also increased insulin sensitivity more than pioglitazone 30 mg as measured by a decrease in homeostasis model assessment of insulin resistance (HOMA-IR) (p-value less than 0.0001 between treatment groups). Other secondary endpoints showed that CRP (C reactive protein) decreased 30 per cent and PAI-1 decreased 30 per cent in the Pargluva 5 mg group versus decreases of 24 per cent and 22 per cent, respectively, in the pioglitazone 30 mg group (p-value equals 0.04 for the CRP comparison and p-value equals 0.0002 for the PAI-1 comparison).

Discontinuation rates due to adverse events were 3 per cent for Pargluva 5 mg and 2 per cent for pioglitazone 30 mg. The incidence of serious adverse events was 4 per cent for Pargluva and 3 per cent for pioglitazone. Three deaths occurred during the 24 weeks of this study: two deaths in the Pargluva group (stroke and sudden cardiac death) and one death in the pioglitazone group (perforated ulcer). Investigators reported these cases not to be drug-related.

The rates of edema-related events for Pargluva versus pioglitazone were 9.2 per cent and 7.2 per cent, respectively, and weight gain was 1.4 kg vs. 0.6 kg, respectively. Investigator physicians reported that four patients developed events related to heart failure: three in the Pargluva group and one in the pioglitazone group, all of whom recovered with diuretic therapy and/or withdrawal of study drug. Confirmed hypoglycemia occurred in three patients taking Pargluva and one patient taking pioglitazone. There were no hypoglycemia-related serious adverse events or withdrawals.