Biotech Values

[DewDiligence]

Lawmakers Blast FDA on Non-Inferiority Studies

http://www.reuters.com/article/idCNN276362720100827

›Fri Aug 27, 2010 5:55pm EDT
By Susan Heavey

WASHINGTON, Aug 27 (Reuters) - The U.S. Food and Drug Administration is cracking down on use of certain clinical trials that show a new drug is no worse than another already on the market, according to a government report released on Friday.

Such trials, known as non-inferiority trials, are used when drugmakers want to compare their experimental product to another one that is already FDA-approved. By showing that their new drug is no worse than another, it can also show some potential benefits such as fewer side effects.

But some critics question the benefit of approving a new drug if it is simply shown to be no worse than something already available to patients, especially if any additional benefits are slight.

The report, conducted by the Government Accountability Office and released by a bipartisan group of U.S. lawmakers, found the FDA is taking a tougher line in the wake of industry guidelines it issued in March.

"GAO's review of FDA's guidance showed that the agency has become more conservative in allowing evidence from non-inferiority trials to demonstrate a drug's effectiveness," the report said.

"First, FDA has limited the indications for which these trials may be used. Second, the agency has also become more rigorous in its review of evidence from non-inferiority trials," according to the GAO, which analyzed such trials looked at by the FDA between 2002 and 2009.

Additionally, the GAO said that repeatedly using non-inferiority trials for certain kinds of drugs can lead to "biocreep," in which successive approvals lead to drugs that are less and less effective and eventually "no more effective than a placebo." [Such an outcome can occur because each non-inferiority trial allows the experimental drug to be worse than the comparator by a prespecified “delta” and still be deemed non-inferior to the comparator.]

Drugmakers can work with the FDA to choose from a variety of clinical trial designs when developing a new therapy, but other kinds of studies are not without their weaknesses.

Placebo-controlled trials only prove a drug is better than a sugar pill -- that is, no treatment at all. Superiority trials can show a therapy is better than another one, but can take more time and be expensive, and are tougher to prove.

Proving a drug is actually no worse than another can also be challenging. According to the GAO, "non-inferiority trials are more complicated to design and their results are more difficult to interpret than other types of clinical trials." [Non-inferiority trials also have more room for cheating. To take an extreme case for the sake of illustration, if clinicians agreed in advance to give every patient the same efficacy outcome, the trial would be guaranteed to be a statistical success.]

In March, FDA officials gave pharmaceutical manufacturers advice on how to set up such trials. They can be used to prove a drug is no worse than another in people with certain infections, HIV, cancer and a few other conditions.

FDA objected to using non-inferiority trials in three cases but ultimately approved the drugs: Merck & Co Inc's antibiotic Noxafil, Bristol-Myers Squibb Co's HIV drug, Reyataz and Novartis' anemia drug Exjade. Fifteen other drugs were also approved using such trials during 2002 through 2009, the GAO said.

Democratic lawmakers and Senate Finance Committee ranking Republican Charles Grassley said in a statement the GAO's finding shows the agency needs to proceed with caution when evaluating such trials.

"The GAO report shows that these so-called non-inferiority trials have often proved to be an inferior means of reviewing the safety and efficacy of new drugs," said Representative Ed Markey, a Democrat. [Ed Markey has a strong opinion on many subjects, but I wonder how much he understands about biostatistics. I used to live in Markey’s congressional district.] "While the FDA has made strides to improve evaluations relying on non-inferiority trial data, the GAO report highlights important limitations of non-inferiority trials."

FDA officials had no immediate comment on the report.‹