Gettysburg Medical News
The Clinical View
by P.E. Hoffsten, MD
30, May 2007

THE AVANDIA MESS – MORE MEDICAL TERRORISM

            Using the internet, there is available to the general public a medical journal called the New England Journal of Medicine, electronic addition.  On Monday, May 20^th, there appeared in this journal, an article implying that a drug called rosiglitazone (Avandia) caused heart attacks and death.  The article was slanted in such a direction as to spectacularize a very speculative observation.  As would certainly be expected of our irresponsible television and other media reporting, the article made a huge splash to terrorize the public.  I would like to give some perspective to what I consider another example of very irresponsible reporting by our TV and news media. 

            My very dear friend and esteemed colleague, Dr. Ken Bartholomew commented on this report, “Did you also know that virtually every single person who ate carrots in the 1800’s is now dead.  This doesn’t necessarily mean that carrots kill people.”  To give perspective to this article, it should be recognized that the FDA would never approve a drug or release it to the public based on the flimsy evidence in this article.  By the same token, they can not be faulted for not withdrawing a drug from the market based on the flimsy evidence in this article.  Thus contrary to what one might believe, the FDA is continuing to do its job very well.  Our pompous politicians, such as Senator Charles Grassly of Iowa, would do better to comment on topics that they know something about rather than to hop on spectacularized bandwagons that they know nothing about. 

            To look at the actual data contained in the article, the authors “cherry picked” 42 trials from 116 trials available in the literature.  These 42 trails were one that would support the point that they were trying to make.  Note, none of these 42 trials were specifically designed to evaluate the potential for heart attacks to occur in diabetes on or off of Avandia.  These 42 trials analyzed the results from 15,560 patients who were treated with Avandia versus 12,283 patients who were not treated with that drug.  Of the 42 trials, 4 had no heart attacks in either group and 22 had no deaths in either group.

            The actual number of heart attacks in the Avandia treated group was 86 in 15,560 patients.  The actual number of heart attacks in the non Avandia group was 72 heart attacks in 12,283 patients studied.  If one does the ratios on these numbers, it turns out that there were 55 heart attacks per 10,000 patients in the Avandia group and 58 heart attacks per 10,000 patients in the non Avandia treated group.  Thus on absolute numbers, there were actually fewer heart attacks in the Avandia group than there were in the non Avandia group.  Hardly, does this qualify as showing that Avandia causes heart attacks. 

            The authors also point out that there were 39 cardiovascular deaths in the Avandia treated group but only 22 deaths from cardiovascular causes in the control group.  But by their own calculation, these numbers do not reach statistical significance for medical certainty that Avandia would cause an increased number of deaths from cardiovascular causes. 

            But the gauntlet has been dropped.  For the accused to deny the crime is rarely credible.  Thus for GlaxoSmithKline Pharmaceutical Company to say that their drug is safe, is no more credible than Steve Nissen to say that the drug is not safe.  While there have not been trials completed to directly analyze the effect of Avandia on heart attack, those trials are in process and will be forth coming in the near future.  Until then, a position recommending Avandia is standing in quicksand.

            What are the alternatives?  Rezulin is a drug of the Avandia class.  Rezulin was withdrawn from the market several years ago because of rare but serious liver damage that occurred in some patients.  As Dr. Nissan points out, multiple other drugs in the Avandia class have been tested and found to be either ineffective or toxic.  They have never been brought to the market.  There is one drug left in this class called pioglitazone (Actos).  This drug has been directly tested to determine if it causes heart attacks or not.  The statistics for this drug were favorable indicating that it prevented heart attacks in addition to better controlling diabetes.  Thus it would seem that prudent medical advice at this point is to advise patients to use Actos if they are going to use a drug in this class or to use nothing at all until Avandia’s safety is certified. 

            One of my very preceptive patients asked why a drug in this class would be used at all.  To answer this question, one must understand the vascular disease causing process of diabetes.  Specifically, type II diabetes renders the person insensitive to their own insulin.  This results in high blood sugars which very clearly are disease-causing making the risk of heart attacks substantially higher than that of the general public.  In the past, credible studies have shown that better diabetic control leads to a decreased incidence of heart attacks.  It therefore follows, that medications that would better control diabetes, would decrease the rate of heart attack.  Indeed that has been shown for a drug called metformin and for pioglitazone (Actos).  As an eye opener, the lifesaving effect of insulin has never been shown.  It is specifically designed to lower blood sugars and yet studies showing a decreased incidence of heart attack in insulin using patients versus those that don’t use insulin have never been shown.  Be that as it may, the concept of lowering blood sugars in diabetic patients seems to have merit to prevent overall complications.  Thus, the merit of introducing the Avandia class of medications. 

            Avandia does have certain predictable side effects including weight gain, a slight increase in the serum cholesterol and some degree of fluid of retention.  Its competing product, Actos, also causes slight weight gain and fluid retention.  Actos does seem to have a more favorable lipid profile impact than Avandia.  Some day when the truth is known (it is not known right now), it is my personal belief that Avandia will be shown to be a safe and effective medication.  Unfortunately,  because of irresponsible reporting, the drug’s reputation will be tarnished and its usefulness compromised.  The lay newspaper and TV media should stay out of the drug evaluation business or do a better job than they are right now.