6 July 2010 


By Nor Azmi Kamaruddin*
National University of Malaysia
Kuala Lumpur

The introduction of thiazolidinedione (TZD) in the late 1990s heralded a new landmark in the management of diabetes mellitus. It was the first among the anti-diabetic agents that specifically addressed the issue of insulin resistance; the major underlying pathophysiology of type 2 diabetes mellitus (T2DM).

However in 2007 serious concerns were raised regarding the possible association between rosiglitazone, one of the two major TZDs and myocardial infarction. It came about as a result of the infamous meta-analysis conducted by Steve Nielsen that appeared in the New England Journal of Medicine (NEJM). Historically the issue first appeared when the Federal Drug Authority (FDA) approved the drug and noticed the possible link among the data that were presented by GlaxoSmithKline (GSK) for the purpose of the drug registration. It requested GSK to conduct an randomised controlled trial (RCT) to address the issue. However what came out in 2006 was the “A Diabetes Outcome Progression Trial“ (ADOPT) study that dealt with the subject of durability of rosiglitazone when compared to glyburide and metformin. Prior to the publication of ADOPT, a second request was made by FDA and as a result of which the “Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes” (RECORD) study was initiated.

Soon after the publication of the Nielsen’s meta-analysis, several other papers appeared on both sides of the Atlantic that either supported or opposed his original findings. These included several different forms of meta-analysis as there were concerned about the methodology employed by the Nielsen’s paper. Similarly several databases were reviewed and published some from huge insurance companies especially from North America. At the same time NEJM allowed the premature publication of the RECORD study which was inconclusive as the study had yet to be completed. Meanwhile FDA requested GSK to initiate another study to look into the issue and the “Thiazolidinedione Intervention with Vitamin D Evaluation” (TIDE) study was set in motion although the study involved the use of vitamin D presumably to answer the concern of osteoporotic fractures as well. Unfortunately, a definitive answer isn’t forthcoming. The TIDE trial which compared rosiglitazone against its counterpart pioglitazone isn’t scheduled to end until October 2015!

The completed RECORD study was finally published in the Lancet in June of 2009. By which time the damage had already been done and the study suffered from the lack of power as a result of patients’ withdrawal. The other criticism was the use of composite end-points where myocardial infarction, the complication in question was grouped together with other major adverse cardiovascular events thus diluting the contribution of myocardial infarction in the final analysis. The authors of the study concluded that “our evidence is insufficient to rule out a small increased risk of myocardial infarction caused by rosiglitazone when compared with other glucose-lowering agents”. However they did warned against the use of Rosiglitazone in patients with a history of heart failure or with previous conditions that might have led to myocardial dysfunction. Similarly TZDs should also be used with caution in women at high risk of osteoporotic fractures.

The issue of the cardiovascular safety of rosiglitazone will be addressed by the FDA in the middle of July 2010. In the meantime Reuters in its 1st of June 2010 issue reported that GSK had already paid USD 60 million in the form of out of court settlement to a group of patients in the United States while the Drug Control Authority in Saudi Arabia had suspended rosiglitazone for a period of 6 months pending further review of the drug.

Whatever the fate of rosiglitazone, one important issue had been brought to our attention. There is serious need to translate glycaemic improvement in any clinical drug trials into actual significant clinical end-points especially in terms of cardiovascular morbidity and mortality. As a result of the controversy surrounding rosiglitazone the need to evaluate any diabetes intervention in terms of cardiovascular end-points has now become an important if not necessary reality.

* The author was on the insulin resistance advisory board of GSK Malaysia. In the past he had received research and travel grants from GSK(M) as well as delivered lectures on behalf of the company.