The Nov. 9,2005 issue of JAMA has a free full text article by researchers at Cleveland Clinic who re-analyzed the data submitted by Bristol Myers Squibb and Merck for approval of muraglitazar.The "glitazar" are characterized as dual alpha/gamma PPAR activators which seem capable of lowering the blood sugar and lowering the triglyceride level and elevating HDL values.
A FDA committee has issued an approvable letter but the JAMA article analysis give a different conclusion than do the statistics submitted by the companies. There either is or is not an increase in cardiovascular mortality in the muraglitazar treated group depending on which analysis you believe.There is also a full text free editorial available that describes what has happened here and give a very interesting list of ways that drug companies( or any one) could manipulate the study and the data in a way to minimize any possible harm that the drug in question might cause. The other side of the coin is that the same could probably be done to maximize the likelihood of demonstrating possible harm. The two analyses differ mainly in how the disease categories are either lumped or split out. You do it one way and an increase in CV events seems significant while another classification leads to the opposite conclusion. Even so, the small number of CV events lead to rather wide confidence limits and neither conclusion could be said to be very robust.
So who is right? There may be no right or wrong here simply different valid ways of looking at the data.But, if approving drugs only after good safety data is submitted is what you desire then you would go with the Cleveland Clinic analysis. In light of the Vioxx situation, you would think the FDA might lean more toward that approach. The editorial says in part "Risk benefit assessment is a dynamic process with few absolutes." The data alone are often not enough but have to be considered in the context of the disease at hand. The disease prognosis and what if any other treatments are available and how effective and safe they are have to be thrown into the decision making process.Should the FDA require more pre marketing trials of sufficient power to more definitively answer the question of increased cv disease risk as suggested by the JAMA article authors or should the FDA accept the suggestion of the sponsors to do post marketing surveillance ? It is not easy to be in charge of approving drugs.
He said-he said clinical epidemiology disputes are not new.What seems new is the widening distrust of drug company data regard medication safety and efficacy to the point where academic researchers have taken it on themselves to "double check" the data.It also speaks to some degree of distrust of the FDA's ability.The era of trusting drugs because the government says they are safe is gone, if it was really ever here.Not only is there some double checking , it is having an impact.This particular drug was leading the pack of glitazars in the path of FDA approval-now its future is uncertain.