DREAM: Rosiglitazone reduced progression to diabetes by 62%

Both rosiglitazone and ramipril aided regression to normoglycemia in participants with IGT or IFG.

COPENHAGEN — Rosiglitazone maleate stopped 60% of people with impaired fasting glucose or impaired glucose tolerance from developing diabetes, and rosiglitazone and ramipril respectively helped 71% and 16% of people with these conditions to return to normoglycemia.

The Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication trial, or DREAM, was a two-by-two factorial, double blind, randomized controlled trial of rosiglitazone (Avandia, GlaxoSmithKline) and ramipril (Altace, King) undertaken by researchers at the Population Health Research Institute, McMaster University and Hamilton Health Sciences, Hamilton, Ontario, Canada. It was presented at the European Association for the Study of Diabetes Annual Meeting.

“DREAM is the trial that provides the best estimate of the effect of ACE inhibitors in preventing diabetes in people with IFG and IGT and no previous cardiovascular disease,” said Salim Yusuf, MD, a cardiologist and epidemiologist, and director of the Population Health Research Institute and Hamilton Health Sciences.

“Rosiglitazone’s effect on diabetes and regression and ramipril’s effect on regression, as well as both drugs’ effect on ALT [alanine aminotransferase], suggest that both drugs may be favorably affecting or remodeling organs involved in glucose physiology,” said Hertzel C. Gerstein, MD, professor in the department of medicine and the department of clinical epidemiology and biostatistics at McMaster University.

Study design and participants

Participants (n=5,269) were aged 30 years or older and had either IGT or IFG. They were recruited from 191 centers in 21 countries and assigned to either placebo or 15 mg daily ramipril, or placebo or 8 mg daily rosiglitazone. They were followed for three years.

The primary outcome of the study was a composite of incident diabetes or death. The secondary endpoint was regression to normoglycemia in participants with IGT or IFG.

The average age of participants was 54.7 years, and more were women (59.2%). Isolated IGT was present in 57.5%, isolated IFG in 14%; both conditions were present in 28.5% of participants.

Additional characteristics of the cohort included hypertension (43.5%; mean blood pressure 136 mm Hg/83 mm Hg), smoking (44.6%) and sedentary lifestyle (26.8%). Participants’ average BMI was 30.9 and their average weight was 187 lb. The waist-hip ratio was 0.87 for women and 0.96 for men.

Participants’ fasting plasma glucose levels were 104.4 mg/dL and their two-hour plasma glucose levels were 156.6 mg/dL.

Ramipril arm results

Adherence to ramipril at three years was 75.4% vs. 80.9% of patients given placebo. Cough was cited by 9.7% of participants as a reason for discontinuation of the drug.

BP was reduced in patients given ramipril to a mean of 128.3 mm Hg systolic and 78 mm Hg diastolic (P<.0001 for both). Participants in both groups gained weight.

Ramipril had a “small, favorable effect” on liver function, according to Yusuf. The researchers measured ALT in both arms for the first year of the study, and found that the drug reduced ALT over placebo (P=.04). “This may imply an effect on the liver and perhaps fatty infiltration of the liver,” Yusuf said.

Four hundred seventy-five participants (18.1%) either developed diabetes (n=449; 17.1%) or died (n=31; 1.2%) vs. comparable numbers in those given placebo: 517 total (19.5%), with 489 (18.5%) developing diabetes and 32 (1.2%) dying. This pointed to a 9% reduction in diabetes with the drug, but this was not a significant difference.

There was no significant heterogeneity found for any of the primary outcomes in participant subgroups, including IFG or IGT, combined IFG and IGT, age, BMI, waist-hip ratio and systolic BP.

Regression to normoglycemia was examined at two years, and the researchers found a greater enhancement of regression to normoglycemia with ramipril (P=.001). Yusuf said that “this is a robust finding and not dependent on where you put the cutoff,” in reference to the recent American Diabetes Association change in abnormal fasting plasma glucose cutoff levels.

“These data would suggest that if you treat 1,000 individuals with ramipril for three years, you will enhance regression in about 104 people,” Yusuf said.

There was a difference in average two-hour plasma glucose levels with ramipril compared with placebo, 135 mg/dL vs. 140.4 mg/dL, respectively (P=.01).

Rosiglitazone arm results

Adherence to rosiglitazone at three years was 79.5% vs. 84% of patients given placebo. Edema was cited by 4.8% of participants as a reason for discontinuation of the drug.

The researchers were concerned that hepatotoxicity would be a problem with rosiglitazone, similar to the effects seen with troglitazone, a drug that was pulled from the market due to hepatotoxicity in 2000. Rosiglitazone, however, reduced ALT levels (P<.0001).

Rosiglitazone also lowered BP. Average final systolic BP in the rosiglitazone arm was lower (129.4 mm Hg; P=.0001), as was average final diastolic BP (78.4 mm Hg; P<.0001).

Weight gain in patients given rosiglitazone was not favorable compared with placebo. “Participants were counseled to use dietary and physical activity to reduce the risk of diabetes, and that counseling was provided to 93% of participants at every visit,” said Gerstein.

Regardless, there was significant weight gain in patients given rosiglitazone compared with a slight, insignificant reduction in weight in patients given placebo. “By the end of the trial, participants allocated to rosiglitazone gained about 2.2% of their body weight,” Gerstein said.

Interestingly, in patients given placebo “there was a progressive increase in the waist-hip ratio, which may reflect the natural history of IFG/IGT in this population. In contrast, participants allocated to rosiglitazone had no change in waist-hip ratio over time,” Gerstein said. An increase was seen, however, in hip circumference in patients given rosiglitazone (P<.0001), “suggesting that the weight was gained in the hip and not gained in the abdomen to the same extent,” he said.

Three hundred six participants (11.6%) either developed diabetes (n=280; 10.6%) or died (n=30; 1.1%). This was a 60% reduction in the primary outcome compared with those given placebo (n=686; HR=0.40; 95% CI, 0.35-0.46). There was a 62% reduction in the rate of development of diabetes alone (HR=0.38; 95% CI, 0.33-0.44), but rosiglitazone had no significant effect on death alone.

In terms of regression to normoglycemia, 50.5% of those given rosiglitazone compared with 30.3% of those given placebo regressed to normoglycemia (P<.0001) as defined by the previous ADA cutoff (HR=1.71). If adjusted to the new ADA cutoff level, the greater rate of regression was still seen (38.6% vs. 20.5%; HR=1.83; P<.0001).

By the end of the study, rosiglitazone reduced average fasting plasma glucose by 9 mg/dL and two-hour plasma glucose by 29 mg/dL (P<.0001 for both).

In prespecified subgroups including sex, age, geographical region, isolated IFG, isolated IGT, combined IFG and IGT, as well as weight subgroups (total weight, BMI, waist-hip ratio, waist circumference and hip circumference), the results consistently favored rosiglitazone, according to Gerstein.

“This drug was effective regardless of the baseline risk for diabetes, so those with IFG or IGT developed diabetes at a rate of about 6% per year. Those with both IFG and IGT developed diabetes at a rate of about 14% per year, and despite the differences in rate, the effects were similar,” he said.

“To think of this from a clinical perspective, for every 1,000 people treated with rosiglitazone for about three years, 144 cases of diabetes will be prevented and about 200 people will progress to normoglycemia, with an excess of about four cases of heart failure,” Gerstein said, adding that the durability of the glycemic effects of both rosiglitazone and ramipril are currently being assessed in a washout phase and will be reported at a future meeting.

Cardiovascular events

There was an 8% higher cumulative hazard ratio for participants given ramipril compared with those given placebo, but this difference was not significant. Yusuf said that DREAM was not designed to examine cardiovascular events, and the event rates (at less than 1% per year) were low.

“This is a different population compared with the HOPE, PEACE and EUROPA trials, which showed a clear and definite impact on cardiovascular events,” Yusuf said. “DREAM deliberately excluded individuals at high risk for cardiovascular events, because it was already clearly proven that ramipril saved lives, prevented heart attacks and prevented strokes. So that was not the question that this trial tried to address,” he said.

“However, people at lower risk may have less activation of the reninangiotensin system, so the effects of an ACE inhibitor may be more modest. Because this excluded such people at high risk, fewer people were on drugs that raise glucose, like a diuretic or a beta-blocker,” Yusuf said. Because of the study design, low-risk population and short duration, emphasis should not be placed on the DREAM results pertaining to vascular events.

Yusuf said that the researchers did not recommend ramipril for diabetes prevention. “However, many people with IFG or IGT have other indications for the use of ACE inhibitors, such as hypertension, heart failure or vascular disease. In these individuals, the favorable effect on glucose is one added reason to use ramipril,” Yusuf said.

With rosiglitazone, there was no effect on the overall composite of cardiovascular events. Heart failure rates, however, were higher in the rosiglitazone group and 14 participants (0.5%) experienced heart failure vs. two (0.1%) in the placebo group (P=.01). None of the cases were fatal congestive heart failure. – by Evan Young

For more information:

• Bosch J, Yusuf S, Gerstein HC, Wareham NJ. Results of the DREAM trial (Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication). Presented at: European Association for the Study of Diabetes Annual Meeting; Sept. 14-17, 2006; Copenhagen.