Is the FDA being guided by science or politics?
Experts worry that decisions made by the Trump administration will undermine COVID-19 therapies.
Teresa Carr is a Texas-based investigative journalist and the author of Undark’s Matters of Fact column. This story originally featured on Undark.
When President Donald Trump fell ill with COVID-19 two weeks ago, his doctors gave him an experimental therapy that contains synthetically-produced versions of the body’s disease-fighting proteins, known as antibodies. Regeneron Pharmaceuticals, the developer, is still testing the therapy for safety and efficacy, and it is not widely available.
But ordinary citizens can get a natural source of those antibodies: convalescent plasma, an age-old treatment that involves giving a person sick with an infectious disease like COVID-19 a transfusion of antibody-rich blood from people who have recovered from it. The Food and Drug Administration (FDA) announced the emergency use authorization of plasma for COVID-19 at a presidential news conference on Aug. 23. Trump heralded the move as a “historic breakthrough.” Health and Human Services Secretary Alex Azar called the treatment “a major advance.” And based on the experience of patients receiving plasma through the FDA’s expanded access program, the agency’s commissioner, Stephen Hahn, claimed that it reduced mortality by 35 percent.
Hahn said Americans could trust his judgement that those results are “pretty clinically significant” because he’s a cancer doctor. But, as explained in a Factcheck.org analysis, Hahn grossly misrepresented data from a Mayo Clinic study (something he himself eventually acknowledged on Twitter). But even beyond that, the study is far from definitive. It is still under peer review and, crucially, it does not include a control group—patients who didn’t get plasma. Without that point of comparison, it’s impossible to say for certain whether the therapy provided any benefit over the usual care.
The push to authorize an unproven treatment based on preliminary findings mirrors what happened with the drug hydroxychloroquine last spring, when the FDA issued an emergency use authorization for the Trump-favored treatment. Two and a half months later, the agency had to revoke it due to reports that the drug was harming patients rather than helping them.
After the plasma announcement, science Twitter lit up with outrage over the misrepresentation of evidence. The Center for Integration of Science and Industry at Bentley University, for example, called out Hahn for “shocking anti-science behavior.” Fred Ledley, the center’s director, told me that until the press conference, he had refrained from commenting on political issues. “But that was unconscionable,” he said. “The abject lies coming from the FDA just scared everybody.”
The pandemic has added urgency to the question of how best to speed therapies to patients. One approach is to accelerate research by streamlining controlled clinical trials — a process already underway in the United Kingdom—as well as through mining medical records for data on how patients who aren’t part of a trial respond to a treatment in the real world. Another complementary tack is for the FDA to greenlight therapies faster—in some cases based on less strict standards of evidence. However, ultimately, how fast a therapy gets to market is moot if the medical community and public don’t trust the FDA to make a sound determination of whether it works and is safe.
The FDA has spent decades establishing a science-based framework for rigorously reviewing products and must be allowed to apply it without political interference, said Ledley. “I’m not confident this will happen,” he added. Neither are seven former FDA commissioners who, in an op-ed for The Washington Post, expressed concern that the Trump administration’s fingerprints were all over FDA decision making—not just in encouraging the agency to promote unproven treatments, but also in curbing its rule-making authority and trying to rush approval of a vaccine. Those politically motivated actions, they argued, are eroding the public’s confidence in any vaccines that gain approval—and public health authorities in general.
Until now, the FDA has had an impressive track record, says Kenneth Kaitin, director of the Tufts Center for the Study of Drug Development in Boston. “All the work they have done to generate that level of trust is being undermined,” he says. “And that’s a scary proposition.”
“There has been a call to speed up the drug development process for decades and decades,” says Kaitin. Nonetheless, he points out that on average it takes about the same amount of time to go from discovery to market approval now as it did three decades ago—about 15 years. The vast majority of that timeline is fairly evenly split between development, where researchers identify promising candidates and test them in animals, and the three standard phases of human trials. In the final step, the FDA reviews the evidence and gives the therapy a thumbs up or down.
The push to go faster often focuses on the FDA. But “the approval part is lightning fast,” says Ledley. While standard reviews can take a year, the FDA has established two pathways to rule on drugs for serious conditions in weeks to months, especially if they are the first available treatment or have significant advantages over alternatives. In addition, the agency has three programs to help accelerate the testing phases—by consulting with drugmakers on how to compress clinical trials, for example. More than half of the 367 new drugs the FDA approved from 2011 through 2018 traveled on at least one of these fast lanes.
The urgency of the pandemic has prompted researchers to find new ways to speed up human studies. “COVID has really challenged the whole system to try to really push the boundaries and try new things, given the urgency, the necessity, and all the data that’s coming up,” says Brenda Huneycutt, director of FasterCures, part of the Milken Institute, a nonprofit, nonpartisan think tank based in California. Researchers have figured out how to work with participants remotely, for example, saving people from having to travel as often to sometimes distant trial sites. They are also simplifying trials—narrowing research questions, making sign-up easy, and collecting less extraneous data.
A good example of fast, agile trial design, said Huneycutt, is the U.K.’s Recovery megatrial of COVID-19 treatments, which has enrolled more than 12,000 patients across nearly 200 hospitals. So far, Recovery has shown that the steroid dexamethasone helps people with severe respiratory complications, but that neither hydroxychloroquine nor the HIV antivirals lopinavir and ritonavir help hospitalized patients. Results from trials of plasma and Regeneron’s experimental antibody treatment are still forthcoming.
One advantage U.K. researchers enjoy is easy access to patient records through the centralized National Health Service. Meanwhile, a fragmented health care system in the US makes it hard to take full advantage of electronic health records. Record-keeping software often isn’t compatible so it’s nearly impossible to reconcile records across different institutions. And with marketers hoovering up personal data wherever possible, many experts are also concerned about patient privacy.
Electronic records are an untapped resource, says Kaitin. For example, researchers can call on medical centers to scan the records for trial recruits—typically, a time-consuming process, especially for rare diseases. And real-world evidence from electronic records could also be an efficient way to show that an approved drug works for other indications. The clinical trial to prove that the hypertension drug telmisartan (Micardis) worked as well as ramipril (Altace) took seven years and cost tens of millions of dollars, for example. Meanwhile, in a 2018 study in JAMA Internal Medicine, researchers demonstrated that they could replicate those findings using electronic records in 12 weeks at less than 1 percent of the cost.
However, randomized clinical trials, which divide participants by chance into groups to compare treatments, remain the gold standard. Everyone I spoke with saw real-world data as an adjunct to clinical trials, not a replacement. Electronic records are incredibly complex, with as many as a million data points on each patient, and researchers are still working to validate artificial intelligence tools to mine these vast datasets. “There’s great potential there,” says Ledley. But “I don’t think we are there yet.”
While the FDA only issued a handful of emergency use authorizations during outbreaks of diseases such as Ebola and Zika, the pandemic has necessitated hundreds of such authorizations to get vital tests, medical devices, and therapies to the public quickly. These are typically based on a lower standard of evidence. Still, as someone who’s covered drug approvals for years, I find it disconcerting that Regeneron and Eli Lily have applied for emergency use authorizations for experimental antibody treatments with zero published, peer-reviewed evidence on whether they work or are safe. The drugs are likely to sail through anyway. In a recent five-minute video released by the White House, Trump touted the Regeneron treatment as a “cure” and says, “we’ve authorized it: I’ve authorized it.” (The president does not have the power to authorize drugs, and the FDA has not greenlighted either of the new therapies as of this writing.)
One consequence of getting therapies into the world ahead of trial results is that researchers can lose the opportunity to study them in a systematic way. Once a drug is widely available—and especially when the administration pushes it as a miracle cure—people will be understandably reluctant to participate in a clinical trial where they might be randomly assigned to a control group that doesn’t receive the treatment. That’s what happened with convalescent plasma. Despite enrolling more than 100,000 patients in the expanded access program, the FDA still lacks solid evidence on the therapy. “It’s sad,” says Joseph Ross, a professor of medicine and public health at Yale University. “We could have done so much more to complement what we’ve figured out works and doesn’t work from the Recovery trial.”
And yet a bipartisan congressional group is convinced that easing standards is the best way to get therapies to patients with serious diseases faster—even beyond the pandemic. The group has introduced the Promising Pathway Act in both houses of the U.S. Congress to require the FDA to add another approval pathway that would rely more heavily on real-world data and other less stringent measures of effectiveness.
Not all researchers are fans. “Analysts are already concerned about (expensive) products being approved without evidence of significant efficacy. This would make things worse,” Ledley writes in an email. It’s telling, he says, that the National Organization for Rare Disorders issued a statement arguing that the legislation “would lower FDA’s approval standards, exposing patients to unsafe and ineffective medicines.”
The FDA is charged with moderating this ongoing scientific debate over the tradeoffs between speed and quality of data. But experts I talked to have grown increasingly alarmed that politicians are putting their thumbs on the scale. So has the public at large. A recent Pew Research Center poll suggests that the public’s trust in a forthcoming COVID-19 vaccine has plummeted. Nearly 80 percent of Americans think that the vaccine approval process will move too fast, without fully establishing safety and effectiveness. Just 21 percent would definitely get a coronavirus vaccine when it’s approved—half the percentage that said that four months ago.
“The FDA will not authorize or approve any COVID-19 vaccine before it has met the agency’s rigorous expectations for safety and effectiveness,” Commissioner Hahn testified before a Senate committee on Sept. 23. “Science will guide our decisions. FDA will not permit any pressure, from anyone, to change that.”
We’ll see. The White House reportedly tried to nix the FDA’s strict guidelines for a vaccine authorization, but the agency published them on its website anyway. In the tug of war between the FDA and the administration, it’s not clear who will pull the vaccine over the finish line.
The implications of the loss of trust in public health institutions are “horrendous,” says Kaitin. Even before the pandemic, some Americans were skeptical about vaccines, and now many are questioning basic precautions like wearing a mask and social distancing, he says. “If anything is going to keep us in this pandemic for a long period of time, it may be the refusal of certain parts of the population to take steps to get control of the disease.”