Series: Landmark Trials in Endocrinology & Metabolism | Study #27
Category: Diabetes Prevention | Design: Multicentre, double-blind, randomised controlled trial | n: 5,269 | Follow-up: 3 years (median)
π Summary
Authors: DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medication) Trial Investigators (Gerstein HC et al.)
Journal: Lancet 2006;368:1096β1105 | DOI: 10.1016/S0140-6736(06)69420-8
The DREAM trial enrolled 5,269 adults aged 30 years or older with impaired fasting glucose or impaired glucose tolerance, or both, and no previous cardiovascular disease, from 191 sites in 21 countries. Participants were randomly assigned in a 2Γ2 factorial design to rosiglitazone 8 mg daily or placebo, and separately to ramipril or placebo. This summary focuses on the rosiglitazone arm. The primary composite outcome was incident type 2 diabetes or death. After a median follow-up of 3 years, the composite primary outcome occurred in 11.6% of participants assigned rosiglitazone versus 26.0% of those assigned placebo (HR 0.40; 95% CI 0.35 to 0.46; p<0.0001), a 60% relative risk reduction. The proportion who achieved normoglycaemia at end of study was 50.5% in the rosiglitazone group versus 30.3% in the placebo group (HR 1.71; 1.57 to 1.87; p<0.0001). Heart failure occurred in 14 participants (0.5%) in the rosiglitazone group versus 2 (0.1%) in the placebo group (p=0.01), a small absolute but statistically significant excess. There was no significant difference in cardiovascular event rates. The ramipril arm produced no significant reduction in diabetes incidence but did increase regression to normoglycaemia.
π Key Findings
| Outcome | Rosiglitazone | Placebo | Effect Size |
|---|---|---|---|
| Diabetes or death (primary composite) | 11.6% | 26.0% | HR 0.40 (0.35β0.46) Β· p<0.0001 Β· 60% RRR |
| Regression to normoglycaemia | 50.5% | 30.3% | HR 1.71 (1.57β1.87) Β· p<0.0001 |
| Heart failure | 14 patients (0.5%) | 2 patients (0.1%) | p=0.01 Β· Safety signal |
| Cardiovascular events overall | β | β | No significant difference |
| Ramipril arm β diabetes incidence | No significant reduction | ||
π¬ Expert Commentary
DREAM is a trial with a paradoxical legacy: it demonstrated the most potent pharmacological prevention of diabetes ever documented in a large RCT, and yet its active agent β rosiglitazone β was subsequently withdrawn from most markets owing to cardiovascular safety concerns identified in the RECORD trial and meta-analyses by Nissen and Wolski. The 60% reduction in diabetes incidence with rosiglitazone in DREAM is substantially larger than the 58% achieved with intensive lifestyle in the DPP, and the 50% normoglycaemia rate (versus 30% with placebo) was an extraordinary metabolic finding that reflected rosiglitazone’s mechanism of reversing insulin resistance by direct PPARΞ³ agonism. The mechanistic basis for this result β restoration of pancreatic beta-cell function and improved adipose tissue insulin sensitivity β is well established, and there is no pharmacological doubt that thiazolidinediones prevent or delay progression from impaired glucose tolerance to type 2 diabetes.
However, the heart failure signal in DREAM (0.5% vs 0.1%), though small in absolute terms over three years, was consistent with the known fluid-retention mechanism of PPARΞ³ agonism and foreshadowed the cardiovascular concerns that ultimately led to rosiglitazone’s market withdrawal in 2010 by the European Medicines Agency (with restricted use by the FDA). The consequence for diabetes prevention is that rosiglitazone’s remarkable efficacy in DREAM has not translated into clinical practice: the drug is not used for diabetes prevention, and the DREAM data remains largely an academic record of what PPARΞ³ agonism can achieve mechanistically. The DREAM trial is nonetheless important for understanding the biology of pre-diabetes, for the evidence it provides about thiazolidinedione mechanisms, and as a cautionary example of how surrogate endpoint efficacy can be dissociated from overall clinical benefit when safety signals emerge in longer-term follow-up.
Limitations: Rosiglitazone is no longer available for use in most countries due to cardiovascular safety concerns identified after DREAM was completed. The 3-year follow-up does not address whether prevention was sustained after drug discontinuation (durable remission vs glucose-lowering effect only). The heart failure excess, though small, reflects a clinically important mechanism-based adverse effect. The study was funded by GlaxoSmithKline and Sanofi-Aventis.
π BOTTOM LINE
DREAM demonstrated that rosiglitazone reduces the composite of incident diabetes or death by 60% and produces normoglycaemia in 50% of participants with impaired glucose tolerance, making it the most effective pharmacological diabetes prevention agent studied in a large RCT, but the subsequent identification of cardiovascular safety concerns with rosiglitazone has prevented clinical translation of these findings.
β Clinical Impact Rating: βββββ Limited by drug withdrawal (historically important)
