Medical Examiner

The FDA Should Approve Drugs Based on Evidence, Not Emotions

A new bill and Donald Trump could come together to undermine the FDA’s ability to safeguard the public from dangerous new drugs.  

A man reaches for a medicine bottle as he takes his prescription pills on February 25, 2009 in Miami, Florida.

We should be strengthening the FDA’s regulatory oversight. Instead, we are weakening it at our own peril.

Joe Raedle/Getty Images

On Dec. 7, the New York Times ran an inspiring, dangerous story about Celine Ryan, a 50-year-old whose advanced colon cancer was successfully treated with experimental immunotherapy. Inspiring, because researchers repeatedly rejected Ryan’s requests to participate in a clinical trial, but her heroic refusal to accept their verdict led to her eventual admission—and remission. Dangerous, because although the article emphasizes that success in one patient proves very little, Ryan’s story served as powerful anecdotal reinforcement of a widespread and mistaken view that groundbreaking cures are stuck in trial stages, and that patients need to take it upon themselves to fight bureaucratic gatekeepers for access. This type of narrative undermines the necessity of the Food and Drug Administration, whose scientific standards are crucial to keeping medical costs down, preventing health crises, and ensuring objective evaluations of new drugs.

The consequences of buying into this mistaken picture are illustrated perfectly by the recently approved 21st Century Cures Act, an enormous piece of health care legislation that includes everything from increased funding for the National Institutes of Health to programs that will fight opioid addiction. Packaged with these uncontroversial goodies, however, is a provision that asks the FDA to relax evidentiary standards for drug approval by granting weight to “real-world evidence” and “patient-experience data.” Supporters argue that the provision will allow patients quicker access to cures. If passed, it could also incentivize competitive development of new drugs by lowering their cost to market—the current gold-standard for evidence of efficacy is randomized placebo-controlled trials, which are time-consuming and expensive. Health policy experts, on the other hand, are rightfully nervous about the unintended results of laxer standards: A market flooded with potentially unsafe drugs that are no more effective than placebo.

From the perspective of a cancer sufferer who has just read Ryan’s story—gutsy individual triumphs over hidebound bureaucracy—it’s hard to take calls for caution seriously. How many others have suffered and died, the logic goes, because they didn’t stand up to the system? Patient advocacy groups—often funded by industryroutinely take this position, criticizing the FDA for excessive conservatism that forecloses on Americans’ right to decide what constitutes a promising treatment and whether or not to take it. (As one Wall Street Journal op-ed put it in 2002: “FDA to Patients: Drop Dead.”) The push for less stringent FDA standards makes willing but unlikely bedfellows of pharmaceutical companies and everyday people, united against the perceived inefficiency of callous government elites who prefer exercising power to saving lives.

But this picture of the FDA couldn’t be further from the truth. In reality, regulations are in place to protect everyday people against the interests of pharmaceutical companies and our own impatience for new medicine. Left unregulated, the market for medicine can generate catastrophic results. Just ask Louise Medus, one of thousands of severely deformed “thalidomide babies” born in the late 1950s and early ’60s. (Medus, who is British, was born with limb deformities.) Thalidomide, developed to treat morning sickness, was made available in 49 countries for two years—despite early evidence that the drug might cause rare birth defects including phocomelia, the underdevelopment or absence of limbs. “I can’t imagine what my dad thought or felt when a grim-faced doctor led him to a delivery room an hour after my birth,” Medus told the Guardian. “All I know is that he almost fainted with shock when I was fully revealed and blurted out: ‘Surely you’re not going to allow a child in this state to live.’ ”

Unlike many other countries, the United States managed to avoid the thalidomide crisis, thanks to a heroic FDA regulator named Frances Oldham Kelsey. Despite intense pressure from the drug’s manufacturer, Oldham demanded further testing, pointing to inconclusive results about thalidomide’s efficacy in humans and initial evidence that the drug caused nervous system side effects. Six applications for approval were denied, and the subsequent birth of countless deformed babies like Medus vindicated the wisdom of Kelsey’s tenacity. For her efforts, she received the President’s Award for Distinguished Federal Civilian Service from John F. Kennedy in 1962. That same year Congress unanimously passed the landmark Kefauver Harris Amendment, which raised the evidentiary bar for new drug approval.

“It’s well-known scandals and crises like thalidomide that led to the FDA having the power it has today,” says Rachel Sachs, a law professor at Washington University who specializes in health law, innovation, and patient access. “Perhaps the existence of the FDA is a slight barrier to innovation. But in my view, it’s far more important to prevent unsafe, ineffective drugs from coming to market.”

Even if eroded standards don’t lead to a devastating public health crisis, it’s almost certain that they’ll lead to a massive increase in costly drugs that don’t actually work. As Sachs pointed out to me, ineffective drugs are already a problem. Under enormous pressure from patient advocacy groups and biotech companies, the FDA recently approved Exondys 51, a drug for a rare disease called Duchene’s muscular dystrophy, despite only limited evidence of efficacy and the sky-high cost of $300,000 per year. The manufacturer of Addyi, aka “pink Viagra,” secured approval for a drug that many experts believe to be not only ineffective but also dangerous. And a recent report suggests that more than a dozen cancer drugs with prices of more than $100,000 per year remain on the market without good evidence they work, approved on the basis of shoddy evidence. If anything, we should be strengthening the FDA’s regulatory oversight. Instead, we are weakening it at our own peril.

The reduced evidence standard proposed by the 21st Century Cares Act is a bad solution aimed at the wrong problem. Evidence is not collected or paid for by the FDA. (It’s paid for by the drug companies.) Reducing the standards simply means that the FDA will not have the necessary evidence it needs to ensure efficacy. “FDA approval is not what shows that a drug works; clinical data do that,” Derek Rowe, an expert on pharmaceutical development, wrote in response to the 21st Century Cures Act. “You can instruct everyone to collect less data, but then you will approve—and ask people and their insurance companies to pay for—more things that don’t actually work.”

So are we destined for rocketing insurance costs and another thalidomide? It’s still hard to gauge the potential consequences of the provision. According to Patti Zettler, a professor at Georgia State who served as associate chief counsel for the FDA, the provision’s vague wording gives the FDA a significant amount of interpretive authority. “There’s room for them to minimize the impact of the law and construe it narrowly,” she told me. Staff members of the FDA seem inclined to that approach, and recently offered a conservative interpretation of the new evidentiary standards in the New England Journal of Medicine.

Unfortunately, President-elect Trump seems to be on the bad side of the law.* His selection for head of the FDA might embrace the act, and open it to a free-for-all. Silicon Valley venture capitalist Jim O’Neill, rumored to be one of Trump’s picks for FDA commissioner, made an extreme case against FDA regulation of drug efficacy in a 2014 speech. “Let’s prove efficacy after they’ve been legalized,” he suggested, an absurd statement that reflects O’Neill’s frightening lack of scientific or medical qualifications. (For one thing, safety and efficacy are not separable factors in drug approval; side effects that are acceptable in effective cancer medication, say, would be unacceptable in headache medicine.) Trump is also rumored to be considering venture capitalist Scott Gottlieb, who, as a former deputy commissioner at the FDA, would be a more traditional choice.

Even without an FDA head who is abjectly anti-evidence, the bill’s passage paired with increased leniency toward pharmaceutical companies could be devastating. Imagine if a drug were developed with a focus on addressing a condition with a specific end point, for example, the ability to concentrate. Without randomized controlled trials, the data on these drugs is virtually guaranteed to be infected by bias, especially if it comes from the company that developed them. Autism would be an ideal condition, since autism research is plagued by difficulties distinguishing real treatment effects from placebo—exactly the kind of problem that high-quality trials are designed to address.

Next, pharma companies could sponsor patient advocacy groups’ campaigns for approval, using testimony from mothers whose children participated in experimental, low-quality trials with miraculous results. Trump tweets something about how the FDA needs to get its act in gear so autistic children don’t suffer. So, they do, and a potentially unsafe, ineffective, and highly profitable medication is approved. Rinse and repeat.

This will be the model for medication approval driven by outrage and desperation, rather than evidence and reason. It is tragic that some people end up waiting for effective treatment, and it is inspiring when patients like Ryan take matters into their own hands and end up beating the odds. I cannot imagine anything more painful than watching a loved one suffer or die, only to find out that speedier approval might have saved them. But allowing such stories to inform our medical decisions is not kind; it is potentially dangerous. This is precisely the kind of mistake that randomized controlled trials are meant to guard against in medicine. Let’s not make the same mistake with public policy.

*Update, Dec. 13, 2016: The wording of this sentence has been updated for clarity. (Return.)