Why the FDA rejected fluvoxamine as a Covid-19 drug

The FDA made a reasonable decision — but one that still shows much of what’s wrong with our current system for emergency approvals.

Last year, researchers who were testing cheap generic drugs in the hope that one or more of them might prove to work as a Covid-19 treatment stumbled across a promising candidate: the antidepressant fluvoxamine.

In a massive randomized controlled trial, called Together, researchers at McMaster University compared eight different repurposed drugs, and found most of them — including ivermectin, the antiparasitic that many embraced as a Covid-19 miracle cure — failed to do much against the disease. But fluvoxamine appeared to reduce severe disease by about 30 percent. While fluvoxamine had already shown some promise in small-scale trials last year, small-scale trials can sometimes turn up spurious good results, so most people didn’t take fluvoxamine seriously until the impressive data from the Together trial.

“This already feels different from hydroxychloroquine and company given the high quality of the research,” Paul Sax argued in NEJM Journal Watch, which analyzes recent research. “We might finally be onto something.” Government regulators, though, remained more skeptical — in part because the regulatory system isn’t exactly designed for adding new indications for drugs that have already been approved by the FDA without a pharmaceutical company sponsoring them.

Another researcher who was convinced of the case for fluvoxamine, David Boulware, decided to take matters into his own hands. The FDA didn’t know how to deal with submissions for a drug to be approved for a new indication without someone responsible for the submission? Fine. He’d submit it himself. In December, he wrote and submitted an emergency use application for fluvoxamine as a treatment for Covid-19.

In a lot of ways, it was a heartwarming story about the power of citizen science. But that’s not how it turned out.

This week, the FDA rejected the application for an emergency use authorization of fluvoxamine. Regulators argued that the results from the Together trial were more ambiguous than they looked — most of the benefits came from a reduction in extended observation in the emergency room, an endpoint fairly specific to the study’s clinical setting in Brazil and not necessarily all that useful. They pointed out that since the Together trial, additional studies have attempted to find a record of fluvoxamine’s benefits, and mostly haven’t found results as large.

In an unusual step, the FDA released an explanation for the rejection, and for the most part it’s very reasonable. But the whole episode still showcases what’s broken about how we review and approve drugs.

The drug approval game

For eight months, the National Institutes of Health, which maintains an up-to-date database of research findings on treatments for Covid-19, didn’t update the fluvoxamine page with any information on the new, promising studies. (The NIH states on that page today, as it has for the last year, that “There is insufficient evidence for the COVID-19 Treatment Guidelines Panel (the Panel) to recommend either for or against the use of fluvoxamine for the treatment of COVID-19.”)

That frustrated researchers, especially this past winter as omicron cases started to grow and the best treatments for Covid-19, like Paxlovid, were not widely available. Many of them told me that with results like these, the FDA would approve a drug that had a pharmaceutical company backing it, and that what was working against fluvoxamine was what they considered its biggest upside: that it was cheap and well-known.

To be clear, fluvoxamine was already approved by the FDA — for obsessive-compulsive disorder (OCD). That means doctors can prescribe it in any context they think is appropriate, and it’s frequently prescribed off-label for anxiety and depression. It could also be prescribed off-label for Covid-19, but many doctors aren’t willing to do such off-label prescriptions. For that reason, the drug’s advocates wanted the FDA to determine that fluvoxamine is additionally indicated for Covid-19, so that treatment for the disease would be officially added to its listed uses alongside OCD. But many experts were skeptical.

“I don’t think the FDA ever will approve it for Covid,” Eric Lenze, the co-author of some early research on fluvoxamine, told me in December. “The reason the FDA will never approve it for Covid is exactly the reason it’s so useful for Covid; namely, it’s cheap and it’s widely available. No one can make any money off it, so no one is going to spend the money to appeal to the FDA to approve it.”

“The guidelines are overly conservative in that they have not yet endorsed fluvoxamine,” Ed Mills, one of the lead researchers of the Together trial, told me in November. Why was the FDA not giving fluvoxamine the same review it would give other drugs? “They don’t know how to deal with submissions where there isn’t someone to be responsible for it,” Mills said. The process of adding an indication is generally initiated by the drug developer, whose lobbyists work closely with the FDA to make sure they’re submitting the evidence the FDA wants to see for approval.

Fluvoxamine research had been largely funded through Fast Grants, a private philanthropic effort to make Covid-19 research work happen, and as the drug is generic, no one would make money from its approval for Covid-19. “It’s very disappointing as a scientist to see that it’s actually not about clinical evidence, it’s about lobbying,” Mills told me.

The FDA’s rejection notice this week made their thinking clear, and it’s clearly not purely about lobbying.

It’s important to note that the crucial justification for fluvoxamine as a treatment is much weaker now than it was this winter when Boulware filed the application. At the time, there was a serious dearth of effective Covid-19 treatments that could be taken at home rather than in the hospital. Monoclonal antibodies, the first line of treatment in earlier waves of the pandemic, weren’t working well against omicron. Many other therapies were only recommended for hospitalized patients. There was no simple pill a person could take at home while their case was mild to prevent progression into severe disease.

Today, there is: antiviral drug Paxlovid. Even fluvoxamine’s strongest advocates agree that Paxlovid works a lot better — it appears to reduce severe disease by 80 to 90 percent. And while this winter Paxlovid was scarce, today there are plenty of doses in the US — though many sick Americans still have trouble accessing the drug because of a lack of primary care doctors they can talk to, while too many doctors remain misinformed about when to prescribe it.

But Paxlovid isn’t a panacea, and it’d still be good to have more options in our portfolio. “There are effective therapeutics that are available. But not everyone has access to them. Not everyone can tolerate them. Some people have contraindications,” Boulware argued in response to the FDA rejection. “And if you go elsewhere in the world, low- and middle-income countries, they have access to no therapeutics.” Still, that Paxlovid, which is a significantly better option, is now widely available weakens the case for fluvoxamine in the US, even though countries that don’t have Paxlovid access should likely make their own calculus.

On the whole, then, the FDA’s decision to decline the EUA for fluvoxamine seems reasonable — even to me, a person who has been enthusiastic about the research supporting fluvoxamine. However, the decision still highlights a lot that should be improved about how the FDA makes and communicates decisions about Covid-19 treatment.

Our Covid-19 treatment failures

For much of the pandemic, if you tested positive for Covid-19, the advice from public health authorities was to do nothing unless your symptoms worsened. Until recently, the official CDC page about what to do if sick with Covid-19 only advised you to wear a mask, wash your hands, and clean high-touch surfaces to avoid infecting those around you. If your breathing deteriorates or you show signs of severe illness like confusion or an inability to stay awake, the CDC advises you to go to the hospital.

Recently, the CDC added an info box highlighting that if you are at high risk of severe disease, treatment may be available. But for people who aren’t categorized as high-risk — which includes older adults or those with medical conditions — the recommendations still don’t include any treatment options.

At first, the lack of treatment recommendations was likely because the evidence for any treatment option was pretty weak. Early in the pandemic, treatments like hydroxychloroquine were hyped but turned out not to work. Later, ivermectin was embraced as a miracle cure. (It isn’t.)

But the lack of treatment options was also the product of a process that wasn’t very good at identifying them and communicating that information to a confused public. The fluvoxamine clinical trial — and many other clinical trials of prospective treatments — was funded by private philanthropy because government processes were too slow to rely on. NIH official recommendation pages meant to summarize the state of research for various treatments were often months out of date; I wrote in November 2021 that the fluvoxamine page had last seen an update in the previous April.

And instead of the FDA proactively working with researchers to set up clinical trials the agency would be willing to rely on to recommend or disrecommend drugs, researchers had to design and conduct trials themselves, and then some doctors had to fill out the EUA application to get the FDA to look at the work they’d done.

Right now, the need for fluvoxamine is limited, the evidence is mixed, and the FDA’s decision not to recommend the drug is pretty reasonable. But ideally, the FDA would have been actively involved in the research process as soon as fluvoxamine first showed promise, and the government would have participated in designing and funding more definitive trials instead of waiting for a submission from an interested group of citizens.

The fact that the evidence about fluvoxamine is still inconclusive at this point is a good reason not to issue an EUA — but it’s also a sign of a glaring failure in our system for investigating promising Covid-19 treatments. Covid-19 is going to be with us for a long time, and other pandemics might be on the horizon. The process for developing treatments — and communicating with the public about them — needs to get better.