Skip to main content

The FDA Is Broken

How the Food and Drug Administration messes up approval of new drugs including the new one, aducanumab, that supposedly helps Alzheimer’s disease patients.

Between 2010 and 2015, drug companies submitted data to the U.S. Food and Drug Administration (FDA) for three different drugs for Alzheimer’s disease, a devastating condition that afflicts six million Americans. The FDA rejected all three because the manufacturers failed to provide convincing evidence that their drugs actually improved symptoms of Alzheimer’s much less cured the disease. How times change. Last month, when biotech company Biogen came to the FDA with data on aducanumab, yet another Alzheimer’s drug, the evidence the company brought once again failed to show that the drug could slow or stop Alzheimer’s cognitive decline. This time, however, the FDA gave the drug a green light, prompting Michel VounatsosCEO of Biogen, to pronounce the decision “historic.”

It was historic all right, but not in the way Vounatsos probably meant it. The scientific data for aducanumab is so flimsy and contradictory, ten of 11 members of the FDA’s advisory committee of outside experts voted against approving the drug (the eleventh abstained). Even the FDA’s own statistician recommended the drug be rejected. When higher ups at the agency went ahead and approved it, three advisory committee members quit in protest. In a scathing resignation letter to the FDA, committee member Aaron Kesselheim called it “probably the worst drug approval decision in recent U.S. history.”

In addition to aducanumab’s failure to improve symptoms, the drug also packs some nasty side effects. Patients in Biogen’s clinical trials who were given the dose of the drug that was ultimately approved by the FDA were three times as likely to suffer brain swelling and hemorrhages as patients given placebo. They were also more likely to have painful headaches, vision loss, disorientation and falls—exactly the kinds of symptoms Alzheimer’s patients don’t need. The public safety advocacy group, Public Citizen, has called for an investigation into the process that led to aducanumab’s approval and for FDA officials overseeing it to be fired. That won’t be nearly enough to fix what ails the FDA.

Aducanumab, whose trade name is Aduhelm, was approved through what’s known as an accelerated pathway, an approval process originally intended for life-threatening diseases for which there are few remedies. The fast-track process requires a lower level of evidence to gain FDA approval than the normal pathway. Once the drug is approved, companies are supposed to conduct confirmatory trials, studies designed to tell for sure if the drug is, in fact, safe and effective.

Biogen got aducanumab on the accelerated pathway after it had already tested the drug in two large randomized controlled trials. Neither trial showed the drug had any effect on memory loss, personality changes, or cognitive decline. In fact, the studies proved futile and were stopped because patients who were on the drug did no better than those on the placebo.

On the accelerated pathway, the company was permitted to test the drug’s effect on a “surrogate endpoint.” These are things that can be measured in the body, like the presence of a protein in the blood or tumor size. A drug’s effect on a surrogate endpoint is much easier to demonstrate than its effect on a “clinical endpoint,” such as improved survival or quality of life.  That means drugs can get approved faster using fewer patients and at far lower cost.

Cheaper and faster might seem like a good thing, except that surrogate endpoints often prove to have little or no connection to a drug’s more important clinical effects. In the case of aducanumab, Biogen’s surrogate endpoint was the development of amyloid plaques, clumps of protein commonly found in the brains of patients with Alzheimer’s. However, more than two dozen prior studies of amyloid reducing drugs failed to show any connection between slowing amyloid protein and improved cognition. Some showed worse outcomes. And Biogen already had two failed trials with clinical endpoints, like cognitive decline, memory loss, and death.

The FDA has given Biogen until 2030 to conduct yet another trial to determine whether or not aducanumab actually works on Alzheimer’s symptoms that really matter. For the next nine years patients can take a drug with unknown (and unlikely) benefits and substantial risks of harm. At $56,000 a year per patient, aducanumab also threatens to bankrupt Medicare, which is compelled by law to cover any drug approved by the FDA.

This is not the first time the FDA has given a drug the go ahead on the basis of a shaky surrogate endpoint. Companies now apply for accelerated approval more often than not for all kinds of drugs, not just those treating life-threatening diseases—and the FDA is increasingly likely to let them on the market. In 2020, nearly three quarters (73 percent) of new drugs received fast-track approval compared to less than 40 percent a decade ago.

If you like this article, please sign up for Snapshot, Portside's daily summary.

(One summary e-mail a day, you can change anytime, and Portside is always free.)

Most drugs approved on the basis of a surrogate endpoint remain on the market because confirmatory studies are either delayed or not conducted–some for more than a decade and in one case for 24 years. Now, under a new category that Richard Pazdur, FDA’s head of the oncology (cancer) division, has dubbed “dangling” approvals, the agency is ignoring even completed confirmatory studies that fail to show benefit or that showed excessive harms.

Last May, for example, the FDA reviewed the confirmatory trials for a set of drugs used to treat six different kinds of cancer. All of the drugs were approved on the basis of surrogate endpoints, such as the degree to which the drug shrank tumors or the time it took after treatment for a cancer to return. Not one of the confirmatory trials showed an improvement in patients’ quality of life or survival.  In some cases, they made quality of life worse. Yet the FDA quietly left four on the market. Two others were withdrawn voluntarily by the manufacturers.

The agency defends its increasing use of surrogate endpoints by pointing to a handful of success stories, such as Gleevec to treat leukemia, which went through accelerated approval  has proved highly effective. However, it is the very hit-and-miss nature of surrogates that caused the agency in the past to insist on studies of clinical benefit before releasing drugs onto the market.

By ignoring negative clinical outcomes in favor of a surrogate endpoint, the FDA has turned the scientific process on its head, if not the point of having a regulatory agency entrusted with ensuring the drugs that doctors prescribe, and patients take, are both safe and effective. “Dying with a smaller cancer rather than a bigger cancer is no consolation because you’re still dead,” says Ezekiel Emanuel, a breast cancer specialist and professor at the University of Pennsylvania. Meanwhile, patients and taxpayers get to fork over tens of thousands of dollars a year for the privilege of taking drugs that don’t work.

How Did We Get Here?

There was a time in the not-so-distant past when the FDA was among the most effective and respected consumer protection agencies in the world. The almost laughable scientific standards it uses today have their roots in the 1980s and ‘90s, when ACT UP and other AIDS activists accused the agency of sitting on potential cures. Gregg Gonsalves, an AIDS activist who is now an epidemiologist at Yale, says he and his fellow activists eventually realized that some early HIV drugs, rushed to market on the basis of surrogate endpoints, not only didn’t work, but they also “had clear and nasty side effects.”

By that time, right-wing anti-regulation groups, such as the Goldwater Institute, and patient advocacy groups, which are largely funded by the pharmaceutical industry, had taken up the cause of speedier approvals. “We sort of opened Pandora’s box,” says Gonsalves.

Many critics also point to the revolving door between the agency and industry. FDA officials are often rewarded with cushy industry jobs after they leave government. Former FDA commissioner Scott Gottlieb took a paid board position at Pfizer after his stint in government. Senior pharmaceutical executives also leave their jobs to join the FDA—often just long enough to help enact policy changes favorable to corporate profits. Patrizia Cavazzoni, who oversaw the approval of aducanumab as the head of FDA’s Center for Drug Evaluation and Approval, came to the agency in 2018 after many years in senior positions at Pfizer, Sanofi, and Lilly. Acting commissioner Janet Woodcock, a longtime government employee, has been accused of favoring industry, a charge that gained credence when she worried publicly over the stock price of biotech company Serepta in 2016, when its drug was under review.

The belief that surrogate endpoints signal meaningful benefit has become such an article of faith it’s now codified into law. The 21st Century Cures Act, passed in 2016 after intense lobbying by pharmaceutical and biotech companies, explicitly promotes surrogate markers and other unreliable types of scientific evidence to boost “innovation.” But as the University of Pennsylvania’s Emanuel cautions, if a drug doesn’t make you live longer, have a better quality of life, or reduce side effects or costs, “you can’t call it innovation.”


Acting commissioner Woodcock, undoubtedly responding to the firestorm of protest from researchers, doctors, and the agency’s own advisory committee members, has now called for an investigation of aducanumab’s approval by the inspector general of the Department of Health and Human Services. But short of ordering the drug off the market, the FDA’s decision to allow amyloid plaques to trump clinical outcomes sets a dangerous precedent for Alzheimer’s patients. The floodgates are now open for a host of drugs that companies quietly shelved in the past after testing showed they did nothing to slow Alzheimer’s cruel progress. Now companies are lining up to get their drugs on the fast track at the FDA. Eli Lilly, one of Cavazonni’s former employers, is first in line. The company says it plans to file for approval of its drug, donanemab, by the end of the year.

The Biden administration has an opportunity to address this problem and remind FDA administrators of their primary responsibility — protecting the public from snake oil. The first step is to appoint an FDA commissioner who comes from outside the agency and has not drunk the Kool-Aid of faster approvals. Woodcock, the acting commissioner, sees pharmaceutical companies as the agency’s partners. Biden should appoint an expert who has shown a stronger commitment to the public’s health than industry’s bottom line. Similarly, Cavazzoni needs to be replaced by someone who doesn’t have one foot in the corporate world.

The second thing Biden could do is require all drugs approved through a fast-track pathway be dispensed to patients with a prominent statement to the effect that the drug is experimental, and its efficacy and safety have not been confirmed. Patients should also be provided a website and a toll-free hotline dedicated solely to their specific drug, so they can report any problems and receive updates as the FDA obtains new data.

Finally, since drug makers are shifting the burden of testing onto patients, who are serving as unwitting guinea pigs for unproven drugs, they should be required to provide the drugs at cost until confirmatory studies can prove the drug’s benefit. That would serve as a strong incentive to conduct confirmatory studies in a timely manner. Even better, companies might quit trying to get slipshod drugs into the FDA’s testing pipeline in the first place.

Shannon Brownlee is a lecturer at George Washington University School of Public Health. Jeanne Lenzer is the author of The Danger Within Us; America's Untested, Unregulated Medical Device Industry and One Man's Battle to Survive It.

Support Nonprofit Journalism. If you enjoyed this article, consider making a donation to help us produce more like it. The Washington Monthly was founded in 1969 to tell the stories of how government really works —and how to make it work better.