Therapy with Metformin—Rosiglitazone vs. Metformin Alone

Source: Fonseca V, et al. JAMA 2000;283:1695-1702.

Most antidiabetic agents attack only one of several causes of diabetes. Metformin promotes glucose lowering by reducing hepatic glucose production and gluconeogenesis by enhancing peripheral glucose uptake. Rosiglitazone promotes glucose transporters and activating adipocyte differentiation.1 Fonseca and colleagues evaluated the efficacy of metformin-rosiglitazone therapy in patients whose type 2 diabetes was poorly controlled with metformin alone.

This was a randomized, double-blind, placebo-controlled trial in 36 outpatient centers in the United States. A total of 348 patients aged 40-80 years with a mean fasting glucose of 216 mg/dL, a mean glycosylated hemoglobin of 8.8%, and a mean body mass index of 30.1 kg/m2 were randomized.

Patients received 2.5 g/d of metformin plus placebo, 2.5 g/d of metformin plus 4 mg/d of rosiglitazone, 2.5 g/d metformin, and 8 mg/d rosiglitazone.

Glycosylated hemoglobin levels, fasting plasma glucose levels, insulin sensitivity, and beta-cell function improved significantly with the metformin-rosiglitazone therapy in a dose-dependent manner. The mean glycosylated hemoglobin decreased by 1% in the 4 mg/d rosiglitazone group and 1.2% in the 8 mg/d rosiglitazone group. Of patients receiving the 8 mg/d dose of rosiglitazone, 28.1% achieved a glycosylated hemoglobin of less than 7%. Dose-dependent increases in body weight and total and low-density lipoprotein cholesterol levels were observed in both rosiglitazone groups vs. placebo.

The data suggested that combination treatment with rosiglitazone and metformin improved glycemic control, insulin sensitivity, and beta cell function more effectively than treatment with metformin alone.

Comment by Ralph R. Hall, MD, FACP

The conclusion printed in the original article in JAMA stated that this benefit occurred with once-daily metformin-rosiglitazone therapy. I could not believe that the metformin could be given in this dose once daily without substantial side effects. A telephone call to Dr. Fonseca confirmed that this was an error in the editing. She indicated that the metformin had been given in multiple doses as it should have been.

This is a carefully conducted study that points to the usefulness of this combination of antidiabetic drugs. The study was carried out over a 26-week period for the drug combination. HbA1C levels in the rosiglitazone groups decreased after four weeks and plateaued by week 18. It was noted that the HDL cholesterol decreased to an extent that the ratio of HDL to LDL cholesterol did not change. Of particular interest is that no one in the rosiglitazone group experienced elevations in the alanine amino transferase. There is still concern, however, that if rosiglitazone is used in enough patients that liver problems will be identified.

It is also of note that in a smaller study of patients (less well conceived from a statistical point of view) treated with metformin and repaglinide with slightly lower baseline levels of blood glucose, that the HbA1C levels fell 1.4%.2 Repaglinide acts by stimulating insulin release by slightly different mechanisms than the sulfonylureas. We therefore have a number of combinations available for patient treatment.

However, in order to reach the desired goal of a HbA1C of less than 7%, a third oral drug or insulin would have been needed in more than two-thirds of the patients in this study.

The big question today is can we prevent beta cell failure by more intensive treatment earlier in the course of the disease? Does glucose toxicity cause permanent damage to the beta cell if present for only a few months? Should we start combination therapy without a trial of exercise and diet in order to prevent permanent damage? A number of investigators believe this to be a distinct possibility.


1. Lehmann JM. J Biol Chem 1995;270:1661-1669.

2. Moses R, et al. Diabetes Care 1999;22:119-124.