Rosiglitazone delays developing DM (DREAM)

Clinical Question

Does rosiglitazone delay the development of diabetes in patients with impaired glucose tolerance or impaired fasting glucose levels?

Bottom Line

Patients at increased risk of developing diabetes were less likely to develop diabetes if taking rosiglitazone (Avandia) than if given a placebo. We don't know how well rosiglitazone compares with other interventions also known to delay diabetes: diet and exercise, metformin, or acarbose. We also don't know if clinically relevant outcomes are improved. (LOE = 1b)

Reference

DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medication) Trial Investigators; Gerstein HC, Yusuf S, Bosch J, et al. Effect of rosiglitazone on the frequency of diabetes in patients with impaired glucose tolerance or impaired fasting glucose: a randomised controlled trial. Lancet 2006;368:1096-1105.  [PMID:16997664]

Study Design

Randomized controlled trial (double-blinded)

Funding

Industry + govt

Allocation

Concealed

Setting

Outpatient (any)

Synopsis

To be eligible for this study, patients had to have a fasting blood glucose level between 6.1 mmol/L and 7 mmol/L (110 mg/dL to 126 mg/dL); impaired glucose tolerance, defined as a glucose level between 7.8 mmol/L and 11.1 mmol/L (140 mg/dL to 200 mg/dL) measured 2 hours after a 75-g glucose load; or both. Additionally, they had to demonstrate at least 80% compliance with taking rosiglitazone during a 17-day run-in period. The authors excluded patients with diabetes and cardiovascular disease. All patients received diet and exercise advice, which is very different than a formal diet and exercise intervention. More than 5200 patients received rosiglitazone 4 mg daily for the first 2 months, then 8 mg daily for the remainder of the study, or placebo. The researchers evaluated the patients sequentially for a median of 3 years. The main outcome -- diabetes or death -- was evaluated by intention-to-treat analysis. At the end of the study, 11.6% of patients taking rosiglitazone developed this composite end point compared with 26% of the control patients (number needed to treat [NNT] = 7; 95% CI, 6-8). Since there was no difference in deaths, this NNT was entirely due to delayed development of diabetes. Only 10.6% of the patients taking rosiglitazone developed diabetes compared with 25% of the control patients (NNT= 7; 6-8). Approximately 25% of the patients taking rosiglitazone stopped taking the medication, compared with 21.5% of those taking placebo (number needed to treat to harm [NNTH] = 30; 18-95). Additionally, 0.5% of the patients taking rosiglitazone developed congestive heart failure compared with 0.1% of the control patients (NNTH = 220; 123-619). A word of precaution: Knowler and colleagues (NEJM 2002;346:393-403) compared lifestyle changes with metformin and placebo, and found that lifestyle changes were more effective than metformin. Since the DREAM study omitted the most effective intervention, we really don't know how rosiglitazone stacks up.