Lessons from Avandia


By guest blogger Daniel Hoffman:

Last week an FDA advisory committee voted to keep GlaxoSmithKline’s (GSK) diabetes drug, Avandia, on the market, although the evidence presented by the agency’s staff and others at the two-day meeting contained serious criticisms of the product’s safety.  Twenty-one out of 33 committee members voted that Avandia increases heart risk and 22 panelists said the agency should either remove the drug from the market or severely restrict its use.

Shortly before the meeting, Dr. George Griffing of St. Louis University told a reporter that any decision from the FDA is a “lose-lose situation... In the best case, it doesn’t work and worst case, it’s harmful.”

There are a few, larger points that deserve attention.

First, it’s important to mention that while many safety concerns were raised about Avandia, the other therapeutic classes used for treating diabetes are also disquieting. Griffing, for example, pointed out that “the second-most common...[class] used in the treatment of diabetes, sulfonylureas, increases cardiovascular mortality by two-and-a-half-fold.”

Other clinicians raise questions about metformin, another common diabetes medication.

Insulin has long been associated with weight gain and other side effects.  A newer class of diabetes medications, the GLP-1 analogs (e.g., Byetta, Victoza and others on the way), has been linked to pancreatitis, thyroid changes, nausea and other conditions. 

In short, diabetes is a serious, growing condition and one where the medications used for treatment represent decidedly mixed blessings.  Whatever problems one may attribute to Avandia, the drug isn’t alone in the sense that all diabetes medications require some difficult tradeoffs.

Next, some observers raised questions about GSK’s conduct with respect to Avandia in the 11 years since the drug was first approved. One blogger posted internal GSK emails, made public as a result of discovery that occurred in numerous lawsuits against Avandia. One admonished the recipient that certain critical findings must “never see the light of day.”

The material is hardly comforting, but GSK’s actions don’t seem outside pharma’s usual course of business.  The internal Merck memos that came out on Vioxx, and those from AstraZeneca and Eli Lilly on their respective antipsychotics also contained their share of hair-raising messages. In fact, it’s probably a fair bet that a generous sample of emails from any company, can alter a reader’s breathing. (The memos from Ford and Bridgestone/Firestone that appeared a few years ago in connection with exploding SUV tires were stunners.) 

Finally, some writers claim GSK violated an implicit understanding between drug companies and patients. Matthew Herper, a senior editor at Forbes, described this compact as one where, “drug companies can make billions of dollars if they invent useful drugs and prove they are safe and effective...[I]t is pretty clear GlaxoSmithKline did not abide by the spirit of that agreement.” 

Well if physicians, regulatory agencies and others no longer place as much faith in the results of clinical trials designed and sponsored by pharma companies, that’s a good thing. Twenty years ago, a larger proportion of biomedical research was sponsored by the federal government and conducted entirely by full-time, academic researchers. At least that represented an alternative source of information. 

Beyond that, a growing number of practicing physicians and insurers are starting to doubt how transferable clinical trial results are to the real world. Patients outside of trials don’t conduct themselves in the same way as closely watched trial participants. For example, some trials in categories such as diabetes and obesity may include behavioral modification programs rarely used in clinical practice.

Then too, only a few thousand volunteers participate in the registration trials that pharma companies submit to get FDA approval for a drug. Trial subjects take the test drug for periods of weeks or months. A substantially different clinical profile can appear when hundreds of thousand or millions of people take a drug for years. That is why some insurers and large medical practices are starting to base their drug evaluations on reviews of their own databases to reveal how their many thousands or millions of patients responded in real life.

So if Avandia provided a boost to other means of evaluating drugs, besides the trials commissioned by interested parties, we’ll all be a little better for it. Ironic how lose-lose can sometimes create win-win.

To check out more Check Up items go to www.philly.com/checkup.