Since 2000, the thiazolidinedione rosiglitazone (AVANDIA) is marketed for treatment of patients with type 2 diabetes (a-t 2000; 31: 66-7). Neither the clinical benefit nor the safety of long-term use of the glitazone have been proven in these patients. Long-term studies with clinical endpoints are lacking. On the other hand serious safety concerns arise from animal experiments, short-term clinical studies and spontaneous ADE reports: the drug may induce heart and liver failure and increases body weight.

The just published DREAM* study (1,2) suggests that the lincence-holder of rosiglitazone, GlaxoSmithKline, already aims at another target group of users:(3) otherwise healthy persons, only showing abnormal glucose tolerance or abnormal fasting glucose.** The study, which was co-financed by Glaxo, investigates whether 8 mg rosiglitazone daily reduces the incidence of diabetes diagnoses - fasting blood glucose above 125 mg/dl or two-hour level over 199 mg/dl in the oral glucose tolerance test (OGTT) - in these persons compared to placebo. 5,269 adults aged at least 30 years took part. In a 2 x 2 factorial design (a-t 2003; 34: 100) the prophylactic effect of the ACE inhibitor ramipril (DELIX etc.) was also investigated at the same time.

During three years the diagnosis of diabetes was made significantly less often in the rosiglitazone group than on placebo, at 10.6% vs. 25% (hazard ratio [HR] 0.38; 95% confidence interval [CI] 0.33-0.44). The median fasting blood glucose fell by 9 mg/dl and the two-hour OGTT level by 29 mg/dl. However, the effect on these surrogate parameters did not reflect any clinical benefit - on the contrary, rosiglitazone was harmful to its users: despite regular dietary advice, at the end of the study they had put on 2.2 kg more than the placebo group. At the final physical examination of the study, 6.8% reported peripheral oedema compared to 4.9% on placebo. In addition, significantly more study participants developed heart failure on rosiglitazone (0.5% vs. 0.1%; HR 7.03; 95% CI 1.60-30.9). A combined cardiovascular endpoint comprising myocardial infarction, stroke, cardiovascular death, heart failure, new angina pectoris or revascularisation, showed a trend to the detriment of rosiglitazone (2.9% vs. 2.1%; HR 1.37; 95% CI 0.97-1.94; p = 0.08). It is open to speculation how this trend would have developed if the study had not been prematurely stopped five months before the planned conclusion, due to knowledge of the data of the PROactive* study (a-t 2005; 36: 95-6). This study with the closely related pioglitazone also confirmed a significant rise in the risk of heart failure in diabetic patients (4). Transaminase levels were reported only for the first year of the study and only as less meaningful mean levels (GPT on rosiglitazone 4.2 U/l lower than on placebo). Whether there was any clinical sign of liver damage was not stated.

The ACE inhibitor ramipril has no effect either on the incidence of a diagnosis of diabetes or on the cardiovascular endpoints (2). The mortality does not differ with either of the drugs from mortality with placebo (1,2).

	Long-term intervention studies of the benefit and safety in patients are lacking for the glitazone rosiglitazone (AVANDIA), licensed for the treatment of type 2 diabetes.
	The just published three-year DREAM study confirms safety concerns: participants were otherwise healthy persons showing only abnormal glucose tolerance, but rosiglitazone increased the risk of heart failure and there was also a tendency to a greater risk of cardiovascular events.
	Heart failure has a particularly poor prognosis in diabetic patients. We, therefore, advise strongly against treating diabetics with rosiglitazone.
	We regard off-label-use for the prevention of diabetes as beyond any reasonableness.

(R = randomised study)

1

The DREAM Trial Investigators: Lancet 2006; 368: 1096-105

2

The DREAM Trial Investigators: N. Engl. J. Med. 2006, published online 15. Sept. 2006

3

Scrip 2006; Nr. 3193: 23

4

DORMANDY, J.A. et al.: Lancet 2005; 366: 1279-89

DREAM = Diabetes REduction Assessment with ramipril and rosiglitazone Medication; PROactive = PROspective pioglitAzone Clinical Trial In macroVascular Events

Abnormal glucose tolerance: defined in the study as fasting blood glucose level below 126 mg/dl and two-hour level in the oral glucose tolerance test [OGTT] between 140 and 200 mg/dl; abnormal fasting glucose: fasting blood glucose of at least 110 mg/dl but below 126 mg/dl.