The Avandia Story: Failures of Regulation and Professionalism

Rosiglitazone (Avandia) is back in the news. First I opened my newspaper this morning to read the following article:

http://hosted.ap.org/dynamic/stories/U/US_MED_DIABETES_DRUG?SITE=CTWAT&SECTION=HOME&TEMPLATE=DEFAULT

Next, faithful reader Marilyn Mann kindly alerted me to a new editorial on rosiglitazone by Steve Nissen of the Cleveland Clinic, author of the meta-analysis that revealed the serious cardiovascular risks associated with the drug (citation below). Finally, through that article's citations, I was alerted to a New York Times article I had originally missed:

http://www.nytimes.com/2008/01/31/business/31censure.html?_r=1

Let's do a quick recap of the story Nissen tells. Early on there were reasons to be concerned that rosiglitazone's blood-sugar-lowering effects might come at the cost of cardiovascular risk, especially increased risk of congestive heart failure. This was important news because there are many other drugs available for management of diabetes, and one of the main reasons doctors want to control Type 2 diabetes is to reduce, not increase, heart risks. (More on that at the end.)

The FDA, meanwhile, was mostly focused on the first drug in that class to be aproved for diabetes, troglitazone (which I discuss at length in HOOKED). That drug (Rezulin) was causing big concern because of recently revealed risks of liver damage, in some cases so bad as to require liver transplant. The FDA wanted to get troglitazone off the market but feared reaction of the diabetes treatment community if it removed the only drug in this promising new class. Problem solved-- in 1999 the FDA quickly approved rosiglitazone plus another cousin, pioglitazone, both known to be relatively free of the liver problem; then later took troglitazone off the market. Trouble was the FDA was in such a rush that the heart risks of rosiglitazone (not shared, it seems, by pio-, thank goodness) got short shrift.

One prominent endocrinologist, Dr. John Buse, started to talk about the increased heart risks associated with the new drug. A delegation from the manufacturer paid him a visit. The upshot was that he ended up signing a contract essentially gagging him from commenting unfavorably on rosiglitazone. (Earlier post on the ins and outs of the Buse case at: http://brodyhooked.blogspot.com/2007/11/intimidation-of-dr-john-buse-some.html)

By October, 2006, GlaxoSmithKline had pretty good evidence from a large trial that there was a significant increased heart-failure risk with its drug, then selling in excess of $3B annually. The company shared the evidence with the FDA. Neither notified the medical profession or the general public. (Hence, the recent report from the US Senate reported by the AP, of how long the company knew the risks before revealing them.)

Nissen and his team conducted their independent meta-analysis and submitted it for publication in May 2007. As soon as he received a copy of the manuscript from the New England Journal of Medicine for his confidential review, Dr. Steven M. Haffner of the University of Texas-San Antonio medical center faxed a copy of it to GlaxoSmithKline, with whom he had a financial relationship, in violation of all the rules of journal publishing and academic peer review. However, GSK had already (as noted) been alerted to the risks of Avandia and merely noted that the figures calculated in Nissen's meta-analysis matched closely its own internal figures. The advance leak of the NEJM paper did however give GSK time to rush the publication of another study as a partial response to the unfavorable publicity. Eventually Avandia sales plummeted. (The Times reported in 2008 that UT-San Antonio was dealing internally with the Haffner accusations, as was NEJM. If any action was taken against him, a quick Google search reveals no sign of same.)

Nissen in his editorial provides a handy time line of all these events and lists the various failures of both government regulation and professionalism. I would add to his list the continued miseducation of both the physicians and the general public that the "answer" to Type 2 diabetes is better glucose control. (If you don't think this is the general impression, watch all the ads on daytime TV for glucose monitors, paid for by Medicare despite repeated evidence that home monitoring of Type 2 diabetes not treated with insulin does nothing to improve patient outcomes.) Starting with the massive UKPDS trial many years ago, several large studies have shown that unfortunately, tighter control of blood glucose does nothing to prevent the major complications of Type 2 diabetes, which are heart attack, stroke, and vascular conditions such as those leading to amputation of limbs. By contrast, aggressive attention to other cardiovascular risk factors, such as controlling blood pressure, and at least one drug, an old generic, metformin, can reduce the cardiovascular risk significantly.

So, if the scientific evidence says forget about close monitoring and tight control of blood sugar in Type 2 diabetes, why are both physicians and patients still so fixated on blood sugar? Well, if GSK is busy marketing a drug that lowers blood sugar, but that just so happens actually to increase cardiovascular risk, you'd expect a marketing campaign that says a lot about the importance of lowering blood sugar and keeps mum about real strategies for reducing cardiovascular risk. The fact that rosiglitazone once sold $3.3B worth of drug is good evidence of the resulting miscommunication--the glitazone drugs were never supposed to be first-line drugs for diabetes even under the best of circumstances, and if that much was sold, it could only be because what truly works for diabetes was completely distorted in the minds of most physicians.
I don't mean to say that it's solely on GSK's head that docs are bonkers about what really works for Type 2 diabetes. Clearly we all want to believe that diabetes = sugar--isn't that what all of us learned on our mother's knees?--and we don't want to believe evidence to the contrary no matter how high it piles up. However, it's one thing when this bias works all by itself, and quite another when millions of marketing dollars are spent to further convince us that what is untrue is true, thereby reinforcing the bias in spades. Even if we aggressively started today we will be years reversing this massive misinformation campaign.

Nissen SE. The rise and fall of rosiglitazone [editorial]. European Heart Journal, published online Feb. 12, 2010; doi 10.1093/eurheartj/ehq016.

NOTE ADDED 3/1/10: The original version of this post contained a typo that distorted the meaning; that typo has now been corrected. Sorry!