COMMENTARY

Calling Out the FDA's 'We Don't Regulate Medicine' Mantra

Opinion: The FDA Can Do a Better Job of Regulating Drug Approvals

Vinay Prasad, MD, MPH

Disclosures

May 24, 2018

The US Food and Drug Administration (FDA) likes to say, "We don't regulate the practice of medicine." As an example, in a recent journal article, authors from the FDA wrote: "[T]he practice of medicine is outside the FDA's purview."[1] And regarding the FDA's decision to rescind the approval for Avastin in breast cancer, an FDA spokeswoman said, "The drug will still be available and the FDA doesn't regulate the practice of medicine."[2]

"We don't regulate medicine" is something of an FDA catch phrase, and while it is technically true, it is also naive. It's like a cop who sees a car swerving and shrugs it off with, "I don't tell people how to drive."

Of course, the FDA does not tell you how to practice medicine (and no one wants them to), and the cop does not tell you how fast to accelerate and where to turn. But they are each responsible for enforcing standards—in the cop's case, for how we drive; and in the FDA's case, the standard for drug approval.

The problem with the FDA is that they have set the bar too low.

The problem with the FDA is that they have set the bar too low and, at times, failed to meet even their own stated threshold for drug approval. Consider this: Nearly two thirds of cancer drugs are being approved solely on the basis of surrogate markers of benefit—like tumors shrinking on scans—as opposed to improvements in quantity or quality of life.[3] This would be okay as long as the FDA later asked drug makers to show that these drugs improve survival or quality of life. But research by Chul Kim, of Georgetown University, and I has found that only 14% (5/36) of drugs approved on the basis of a surrogate later showed a survival benefit with a follow-up of 4.4 years on the US market.[3]

Some approvals that are based on a surrogate endpoint are through the FDA's accelerated approval program, meaning there is a postmarketing study to verify efficacy, and having this program makes sense and is a good idea.

However, other approvals that are based on a surrogate are regular approvals, meaning the FDA doesn't require a postmarketing study of efficacy. Here you have to be cautious. You really need to know that the surrogate has a proven track record of predicting survival because you may not get more information later.

The FDA agrees and says regular approvals must be based on "established" surrogates.[4] Here is a tweet of a slide from an employee of the FDA:

The FDA says clearly, "For regular approval...validation is necessary."

But the FDA is not demanding that validated surrogates be used. Kim and I found that for 11 of 30 (37%) regular approvals, there is no validation study in the entire medical literature, period.[4] They just don't exist. In only 3 of 30 (10%) approvals is there a strong, proven correlation between the surrogate and survival.[4]

The cop who says, "I don't tell people how to drive" is right. But when someone blows past at 92 miles an hour and the cop doesn't act, you might respond, "But you enforce the speed limit, right?" Similarly, the FDA doesn't tell you how to practice medicine. But they do say that regular approval requires validated surrogate endpoints, and they don't enforce that rule.

Growing Concern

Approving many new drugs give the impression of advancement and innovation, but it is only advancement and innovation if patients are better off. There is a growing cadre of experts who are concerned that this may not always be the case.

Ajay Aggarwal in Nature recently wrote, "[A] drug that shrinks tumors might not help to extend people's lives." He added, "Approvals that let drugs stay in the marketplace on the basis only of quick, easy surrogate end-points are unlikely to produce highly effective treatments; we will simply get more drugs providing marginal value."[5]

This is precisely my concern.

When the FDA Forgets Their 'We Don't Regulate Medicine' Mantra

But in some cases, the FDA is doing the opposite—giving approvals where they are not necessary, where doctors can already use those drugs for that purpose as part of the practice of medicine.

For example, the FDA seems to forget their "we don't regulate medicine" slogan when it comes to certain approvals—specifically, if a cancer drug is already on the US market and the FDA approves it for a second or third purpose on the basis of bad data.

Doctors already have access to these drugs; they can already use them for the alternative purpose. It's called off-label prescribing, and it is ubiquitous in cancer medicine. It will even be reimbursed, if recommended by one of several compendia or guidelines. If the FDA doesn't regulate medicine, and the data behind these expanded indications are bad, there is no need for the FDA to weigh in. They don't have to approve it.

Just look at sunitinib for adjuvant kidney cancer, an approval that came out this year. We've used sunitinib for metastatic kidney cancer for over a decade, but this year, the FDA said it is authorized for use in the adjuvant setting, for high-risk patients after the tumor is resected.[6]

Make no mistake—this was a bad approval.

Make no mistake—this was a bad approval. The drug didn't improve survival in two adjuvant studies.[7] It didn't even delay recurrence in one.[7] And it lowers quality of life.[8,9] The doctors who discussed it at the drug advisory meeting had a misunderstanding of what had been shown.[10] They didn't seem to know that overall survival results were in the supplement of the paper. Researchers have gone as far as calling this approval "regulatory capture," meaning it was so bad that it is as if the FDA is working for the drug companies, not the people.[11]

To the FDA: If you don't regulate the practice of medicine, you don't need to give second or third approvals based on unproven surrogates, like sunitinib in the adjuvant setting for kidney cancer, or pertuzumab for neoadjuvant breast cancer.[12]

After all, doctors could still use it if they wanted. If they felt strongly, they could even put it in the guidelines. Instead, you can say, "We don't regulate the practice of medicine. Do what you want, but you will only have our blessing when you show that the drug truly benefits patients."

Constructive Criticism

Unquestionably, the world is a better place with the FDA than without them. The FDA does an okay job, but they can be better. Here is my constructive criticism:

  • Only use regular approval when it meets your own standard—the surrogate endpoint used in the clinical trial has been validated.

  • As long as you continue to say, "We don't regulate the practice of medicine," let that cut both ways. You don't have to bend over backwards to give second or third approvals based on bad data. You can say no.

Cops don't tell us how to drive, but we want them to pull over reckless drivers. The FDA is no different. No one wants them in the exam room, but they have a mandate to approve cancer drugs that are safe and effective. "Effective" means that they make people live longer or better, and not merely change the results of CT scans. It is time that the FDA remembered that.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....