The byzantine world of pharmaceutical regulation has recently broken into the public consciousness, causing a bit of a panic. Aducanumab—the first new Alzheimer’s treatment in nearly two decades—was approved by the Food and Drug Administration on June 7 despite scant evidence of benefit, and against the nearly unanimous advice of the agency’s expert advisers. Op-eds called the decision, which could trigger billions of dollars in new government spending, a “false hope,” “bad medicine,” and “a new low.” (FDA officials have said that their decision was based on “rigorous science,” and that it reflects the willingness of people with Alzheimer’s and their families’ to accept a treatment that might help, despite “some degree of uncertainty.”) On Thursday, the FDA tried to clarify that the drug should be used only for patients with mild dementia; the next day, amid concerns about inappropriate interactions between the drugmaker and FDA officials, Acting Commissioner Janet Woodcock called for her own agency to be investigated.
This isn’t the first—or fifth—run of bad press the agency has received, but one gets the sense from recent coverage that some crucial threshold has now been crossed, that the rising floodwaters of ineptitude have finally yielded a catastrophe. But even if this blunder’s inner workings are more public than those of the past, there’s little reason to believe we’ll see systemic change at the FDA. The long-standing and gradual erosion of the agency’s scientific standards makes me think instead of the eroding coastlines and thawing icebergs associated with climate change. For decades now, alarmed scientists have been crying out for action as things have gotten worse and worse. And for decades now, little has been done.
The FDA’s standards began to slide in the late 1980s and early ’90s. AIDS activists, desperate to slow a devastating and mysterious illness, had pushed for the creation of new pathways for approving treatments more quickly—and, effectively, on thinner evidence. A new program in 1992 allowed for “accelerated approval” on the basis of surrogate markers, which are indirect measures of a drug’s benefit, assessed via laboratory or imaging tests, that stand in for more meaningful outcomes such as life expectancy. But the implementation of these accelerated processes was criticized by some scientists and patients, even at the time. In 1994, for example, The New York Times cited skeptics who worried that “no one can tell if the drugs work.” Eight months later, the AIDS activist organization ACT UP San Francisco called Anthony Fauci a “pill-pushing pimp” for supporting CD4 immune-cell counts and viral loads as surrogate markers. They were completely invalid, the activists wrote, and nothing more than “a marketing exec’s wet dream.”
That early change in standards worked out for the best. We now know that CD4 counts and viral load are excellent markers of HIV/AIDS status, and the types of HIV drugs that were being questioned at the time—reverse-transcriptase inhibitors and protease inhibitors—did turn out to be effective. Newer versions still form the backbone of lifesaving, multidrug regimens. In fact, their success has been used to wave off criticism of the FDA’s declining standards for years.
But that level of success is not at all the norm. Most treatments in medicine will prove only modestly effective and come with real risks. The use of regulatory shortcuts has grown more common anyway. In 1992, about 40 percent of all drugs qualified for at least one of the FDA’s expedited programs for approval. By 2018, that rate had doubled. Use of surrogate endpoints has now become routine—60 percent of approvals in recent years have relied on them. The repeated protests of agency critics along the way have done nothing to slow the trend. In May 2000, for example, the government approved the use of saline breast implants despite safety concerns. “This decision really sets the FDA standard of safety at a new low,” the health-policy expert Diana Zuckerman said at the time. More than 20 years later, she continues to decry the agency’s still-declining standards: The aducanumab approval “sets a very dangerous precedent that could harm patients’ health,” she said recently. (Acting Commissioner Woodcock has disputed the idea that evidence standards are declining; in her view, “we know far more now about a drug when it is approved than in the past.”)
Indeed, public controversies over drug approvals have only proliferated. In 2015, the FDA approved flibanserin, a treatment meant to improve female libido, after having rejected it twice. “This is a product that is neither very effective nor particularly safe,” one former FDA official told reporters at the time, and it sets “a precedent that a drug for women’s sexual health has to be treated in a very special way.” The agency reached another low the following year, when its leaders overruled their own staff to approve eteplirsen, a treatment for a rare genetic condition called Duchenne muscular dystrophy. The drug’s poor quality of evidence led a dissenting official to declare that it was nothing more than a “scientifically elegant placebo.” The 21st Century Cures Act easily cleared Congress a few months after that, though two former ACT UP members and a former FDA commissioner had declared in a Times op-ed that the new law’s encouragement of even-lower low standards was “potentially placing patients at unnecessary risk of injury or death.” In 2019, the bioethicist Ezekiel Emanuel warned about the approval of a slew of new cancer therapies with uncertain or very limited benefits, noting that “drugs with unproven effectiveness sell false hope to desperate patients.”
The FDA says that it must balance the importance of meeting rigorous standards with the need to provide timely access to potentially lifesaving medicines. “Commentators have noted FDA’s increasing use of these programs over the last decade, often with a view that the increase is driven by a loosening of our approval standards,” Woodcock and the FDA official Peter Marks wrote in 2019. “In reality, the FDA’s standards have not changed. Instead, the increased use of expedited approval pathways is directly related to the increasing numbers and scope of these programs provided by Congress, as well as the kinds of medicines that are being developed, and the types of diseases that are being studied.”
For all the panicky commentary, the problem is a diffuse and slowly growing one, with specific harms that are tricky to assess. That’s why the approval of yet another costly, ineffective drug on the basis of a suspect surrogate marker ends up being like another deadly set of wildfires in the West: We’re told that it has to do with long-standing changes in the climate, but responding to the immediate crisis distracts from broader fixes. Even the agency’s “historic” and ill-advised sign-off on aducanumab might not move the needle of public opinion toward favoring greater scientific rigor over quicker access, according to Daniel Carpenter, a government professor at Harvard who has studied the FDA. It’s hard to feel invested in medical regulation without tangible, personal reminders of what happens when it fails. If aducanumab were to prove really dangerous for patients—if it ended up killing people—that would certainly grab attention. (There’s currently no evidence of major harm from aducanumab, though the FDA’s expert advisers did raise concerns about brain swelling in some clinical-trial patients.) The experience of children who were born with severe disabilities from thalidomide in the 1950s and ’60s, like that of the victims of sulfanilamide poisoning in the ’30s, prompted significant strengthening of government oversight. Thalidomide had that influence, Carpenter told me, “precisely because the nature of the deformities the victims had was so visceral, so evocative.” But the approval of a merely ineffective and expensive drug simply wouldn’t have the same effect. You need to see the health-care system’s version of a starving polar bear, he said, or else most people won’t care.
Even aducanumab’s staggering price tag of $56,000 a patient per year may not be conspicuous enough, because few will pay the full cost themselves. Most of the expense incurred for each new drug—which, for aducanumab, could be larger than NASA’s yearly budget—ends up being quietly subsumed into strained public-health budgets and rising health-insurance premiums. The actions of drug regulators, like those of industrial polluters, are often freighted with unacknowledged externalities. “The FDA specifically does not really worry about those larger societal issues and doesn’t really worry about cost,” Walid Gellad, a drug-policy researcher at the University of Pittsburgh, told me. Instead, the agency is judged by how many drugs it can approve. No regulator “wants to go before Congress or the American people and say, ‘This year we approved 50 percent fewer drugs than last year,’” Gellad said.
There are proposals for encouraging the FDA to pursue cost-effective therapies, just as there are proposals for encouraging reductions in carbon emissions. For example, health-policy researchers have suggested allowing the agency to shrink the window in which pharmaceutical companies can exclusively profit from their compounds, if those companies charge too much money for negligible benefits. Carpenter has recommended increasing the fees that companies pay to the FDA, and then having the agency use those resources to fund more-rigorous clinical studies. Medicare already has the authority to withhold payment for questionable treatments (and is now considering that option for aducanumab). But as with cap-and-trade policies for carbon emissions, aggressive approaches have failed in the face of powerful stakeholders. It’s not just the profit-driven pharmaceutical industry fighting these policies; patient groups are also effective advocates for quicker, easier access to new treatments.
In the absence of meaningful political progress or collective action, efforts to address the spread of bad or useless therapies—like efforts to address climate change—are left to individuals. The FDA may keep approving marginal treatments, but doctors don’t have to prescribe them. In theory, we could take a stand, and just say no to aducanumab. “There’s an assumption that physicians are good stewards and they’re going to take care of everybody and they’re not going to prescribe something that’s not effective,” Joseph Ross, a physician and health-policy expert at Yale, told me. “But you and I know how it really works in the real world.” Indeed, I do. For doctors, the allure of slowing the progression of a debilitating condition, satisfying patient requests, and even profiting from every prescription will make aducanumab hard to resist. Many of the passionate academic physicians I talk with seem willing to take up this fight in their own clinic, and recommend against the drug’s use, but they’re just idealists at the fringe, not the rank and file of the profession—91 percent of whom recently responded that they generally trust that the benefits of an FDA-approved therapy will outweigh its risks.
This is the tragedy of the concerned citizen: We are personally destined to fail. Just as switching to an electric car or turning your lights off won’t cool a warming planet, a minority of idealistic doctors won’t stop the flood of ineffective treatments. But it’s not impossible that the aducanumab fiasco will yield some systemic change. Vinay Prasad, an oncologist and drug-policy expert at UC San Francisco, suspects that the total cost of this or some future drug could end up being high enough to cause a fiscal crisis, prompting action at long last. “The American economy can handle a great deal of wasteful health-care spending,” he told me. “But it can’t tolerate an infinite number.” The FDA might have finally acknowledged these financial considerations when it rapidly reversed course to recommend that only a subset of patients with early Alzheimer’s disease receive aducanumab. That change could save insurers billions of dollars, and it makes them more likely to place explicit limits on which patients will have their expenses covered.
An interminable health-care-cost crisis will eventually spill over into other political issues, Prasad said. For example, the average worker experiences rising health-insurance premiums as stagnating wages. “It’s profoundly destabilizing in a society in ways we don’t see,” he told me. But whether politicians will do more than pursue empty hearings remains unclear. Before aducanumab, the exorbitant cost of new hepatitis C treatments, initially as much as $1,000 a pill, generated similar hand-wringing but no real policy change.
If an independent investigation leads to a reversal of the aducanumab decision, then perhaps such an unprecedented turn of events will prompt a real public reckoning. The most likely outcome, though, is that the bad headlines will recede and a new, lower bar for evidence will be set. With each alarming declaration of “the hottest day on record,” after all, it gets harder to remember what cooler temperatures felt like in the first place.