Stuck on the concretes

Would you agree that the following is a proper goal in medicine: Making sure that medicines are prescribed only to those who would benefit from them, so that money is not spent needlessly on drugs that, at best, do no good (and may in fact do harm) to the patients that take them?

Would you also agree that the government has a role in making sure that that goal is met?

I believe the answer to the first question is "yes" -- but evidence shows that the answer to the second must be "no."

The FDA, in its approvals process, is trying ever harder to make sure that drugs are only sold to patients who need them, and for whom there is solid evidence that the drug will do what it is supposed to do. Sounds good, right? Unfortunately, allowing the government to meddle in the sale of drugs this way is stifling innovation in medicine.

When the FDA approves a drug for sale, the approval is not granted unconditionally. A manufacturer of women's clothing does not have to place any restrictions on who can buy its wares; if that means a man buys the clothes for the purposes of cross-dressing or an overweight woman buys an outfit she doesn't really fit into, the government doesn't care. Not so for a drug. FDA approval is always granted for a particular "indication." What is an "indication" in this context? "Indication" used to mean "for a disease" -- that is, if a drug was shown to be effective against breast cancer, the FDA would approve it for the treatment of breast cancer. But these days, indications are far more specific than that. The FDA generally will only approve a drug in the exact subset of people that was studied in a clinical trial: If a birth control pill was tested in women between the ages of 18 and 34, the drug will be approved only for women in that age group. If a prostate cancer drug was tested in men who had already been treated with X, Y, and Z chemotherapy drugs, it will only be approved for men who have already gotten X, Y, and Z.

The problem with such concrete-bound definitions of when a drug can be used is that there is often no reason to use those concretes in the definition. There may be no reason why that prostate cancer drug wouldn't work in men being treated for the disease for the first time. There may even be evidence, in the form of another clinical trial, that it does work in that setting. Too bad: Because the FDA bureaucrats have only reviewed the information from the "already got X, Y, and Z" trial, that is all they will consider. Thus, the judgment of bureaucrats considering a limited set of information is expected to supersede a patient's own judgment as well as the judgment of his or her doctor, who may know much more data than the bureaucrats about the drug in general and certainly knows more about the patient he or she is treating.

Take the following examples:

• Gardasil. Gardasil is a vaccine against human papillomavirus (HPV), one of the most common sexually transmitted infections. Its effectiveness was tested in girls and young women aged 9 to 26. Consequently, the FDA approved it only for women in that age group. But women older than 26 were not excluded from Gardasil's clinical trial because the vaccine was less likely to work in them; they were only excluded because, if a woman is going to contract the virus, she is likely to do so before the age of 26. There is no reason to suspect that a 27-year-old -- or a 37-year-old -- who has not been infected would not benefit from receiving the vaccine. Yet the FDA insisted on the age restriction in Gardasil's indication.
• Ixempra. Ixempra is a chemotherapy drug recently approved to treat breast cancer -- but only breast cancer that has spread to other organs in the body, and only after patients have already had treatment with three other specific types of chemotherapy. Again, there is no reason to believe that Ixempra would not work for a woman whose cancer had spread but who had not yet received all three of those other types of chemotherapy.
• Lybrel. Last year, I wrote about how Lybrel, the only birth control pill approved to be taken 365 days a year, is no different from taking another brand of pill and simply skipping the placebo pills, thus sparing a woman from menstruation. No other birth control pill can advertise this -- although there is no reason why any other birth control pill should not work.
• Avastin. Avastin is a cancer drug used in combination with chemotherapy to treat several types of cancer -- but not only does the FDA recognize only two of those types (colorectal cancer and lung cancer, and only after those cancers have spread to other parts of the body), it also dictates what chemotherapy drugs are considered acceptable to use in combination with Avastin. Despite ample evidence that Avastin works in breast cancer, the FDA has yet to approve it for that use. Furthermore, when Avastin is used to treat lung cancer, the FDA specifies that it must be used in combination with two specific drugs, paclitaxel and carboplatin. If patients want to receive Avastin in combination with the two related (and very similar) drugs docetaxel and cisplatin, that's just too bad -- again, there is no reason it shouldn't work.

Doctors and researchers come up with new ways to use drugs far faster than the FDA can possibly hope to approve such uses -- and that's if the pharmaceutical company chooses to go through the lengthy and costly bureaucratic process of getting a new indication approved at all. As a result, the FDA's recommendations for how to treat patients lag far behind actual, current medical practice. How could they not? A bureaucracy cannot possibly match the speed of individuals and their doctors making the decisions that are, individually, right for them.

Perhaps some patients might not be comfortable trying a drug in a setting in which it has not been previously tested, such as 30-year-olds getting vaccinated with Gardasil. But should all patients be denied the ability to do so? Certainly not: That is a violation of the patient's right to purchase whatever drugs he or she wishes, and of the drug manufacturer's right to sell the drug to whomever it can.

Fortunately, at the moment FDA approvals do not necessarily limit the treatment patients actually get: Once a drug has been approved for sale, a doctor can prescribe it for uses other than those for which it was approved (so-called "off-label" usage). But the company cannot promote the drug for such uses -- for example, an Avastin sales rep is currently not permitted to say one word about breast cancer, even if a doctor asks him or her about it. And, because some insurance companies use FDA approvals to determine what drugs they will cover, patients may have to pay for so-called "experimental" use out-of-pocket, even though they have already paid for prescription drug coverage.

How much money has been wasted on obtaining additional FDA approvals because the scope of a first approval was unnecessarily narrow? How many patients have suffered or died because they weren't allowed to try a therapy that had every reason to work, and no reason not to? The FDA's clinging to concretes, rather than allowing patients and physicians to exercise their own judgment, is expensive indeed.