August 02, 2007

The Journal (what can I say, the editors run interesting stories) has a fascinating article today on an emerging dispute between cancer drug companies and the FDA over what standards should be used to approve drugs based on cutting edge science. The argument is that some treatments may not fit easily into the FDA's existing regulatory structure. Read the whole thing and make up your own mind.

But the thing that struck me was the following, from the FDA's director of medical policy, Dr. Robert Temple:

FDA officials defend the current approval standards, saying it is better to delay a new drug until there is definitive proof it works. "I don't think you want the FDA to approve a drug saying there's a chance this works. I think you need to know the therapy works," says Robert Temple, director of the office of medical policy at the FDA.

"If somebody uses a vaccine that doesn't work, they will be forgoing something else," Dr. Temple says. While he recognizes that some patients have "no options left," he says the best route for them is to appeal to the FDA and to drug manufacturers for permission to use experimental medicines.

I understand Dr. Temple's point, but I would also reply that most drug treatment is "experimental" in the sense that physicians often don't know - especially in complex diseases like cancer - whether or not a therapy will work in a specific patient before they try it.

Many drugs work only in a relatively small subset of the population (with exceptions, of course) and so most of the time physicians are trying a "let's do this and see if it works" approach. (There's nothing wrong with this, it just reflects the state of our medical knowledge.) But this also means that every time you try a new treatment you're forgoing the use of some other treatment that might work better. For most of us, failure just means switching to a different painkiller or statin. But the stakes are much higher in cancer, when lives can be measured in months.

Now, Dr. Temple can argue that our degree of confidence that a given cancer drug works ought to be very high, but individual clinicians and patients are going to disagree with him because their individual tolerance for risk or uncertainty is going to be different.

And the drug discovery process is also iterative. Doctors expand our knowledge of effective treatments by using them in ways companies and the FDA would never predict; indeed, as I understand it, close to half of all cancer drugs are used "off-label", i.e. for some other indication than is listed on the FDA-approved label. So there is plenty of drug "R&D" going on outside the FDA's purview just in the normal course of physician practice.

From my perspective, the question is is there another way of approving medicines based on novel mechanisms (like the cancer vaccines discussed in this article) that would help lower the costs and time required to get promising new treatments to patients? Former FDA Commissioner David Kessler says that "we have to find creative ways of getting cancer drugs to patients even if we end up being wrong a few times." I agree wholeheartedly.

Posted by Paul Howard at August 2, 2007 02:14 PM

Comments

Post a comment