Skip to Main Content

The Food and Drug Administration declined Monday to authorize a 30-year-old generic antidepressant as a treatment for Covid-19, dealing a major blow to a small group of doctors who have organized around the pill for months, arguing that it could provide a cheap and accessible way to prevent hospitalizations and death both in the U.S. and around the world.

In an unusual two-page summary — the FDA does not generally disclose the reasoning behind rejections — regulators said that the doctors failed to provide adequate evidence of effectiveness of the drug, called fluvoxamine.

The submission was primarily based on a roughly 1,500-patient randomized, controlled trial in Brazil that found patients who received fluvoxamine early in the course of their disease were 32% less likely than patients who received placebo to be hospitalized or need emergency care of at least six hours. There were also 17 deaths in the fluvoxamine group compared to 25 in the placebo group, although the difference was not statistically significant.


Fluvoxamine can be found at local pharmacies for around $4. When the clinical trial results came out last summer, they produced hopeful headlines about the prospect of using fluvoxamine as a cheap oral treatment for Covid-19 — at a time when other antivirals were not yet on the market.

The FDA, however, said Monday that it was uncertain that the six-hour cutoff the Brazil study used was “a clinically meaningful threshold.”


Meanwhile, the results on hospitalizations and deaths alone were “not persuasive.” Neither, regulators added, was a smaller randomized clinical trial conducted in the U.S. or a number of real-world studies, because they were small, non-randomized, or employed different endpoints.

Regulators also pointed to a pair of larger randomized controlled trials that failed to show the same benefit the Brazil study did.

In a detailed rebuttal submitted last week and shared with STAT, David Boulware, the University of Minnesota infectious disease physician who has led the push to get fluvoxamine authorized as a Covid-19 treatment, said the FDA’s logic was “inconsistent.”

For instance, he noted, the Merck and Pfizer trials for their oral antivirals for Covid — molnupiravir and Paxlovid — also didn’t rely on a conventional definition of hospitalization. Instead, they defined hospitalization as more than 24 hours of “acute care.”

In an interview, Boulware said fluvoxamine could still have applications, even if the Pfizer pill, Paxlovid, is vastly more effective and now filling pharmacy shelves. For example, many high-risk patients can’t take Paxlovid because it could interact with a drug they’re taking to manage another condition.

Fluvoxamine could also be useful in middle- and low-income countries where the Pfizer and Merck pills are not yet widely available, Boulware said. The Merck pill reduces the risk of hospitalization from Covid-19 by around 30%, according to company clinical trials.

“It’s not my first choice as a physician but I should have the option,” Boulware said.

Fluvoxamine was unusual among experimental Covid-19 medicines. Because it was generic, no drug company stood to profit from bringing it to regulators, which can be a costly and time-consuming process.

Instead, it was championed by Boulware and a group of physicians and researchers, who saw it as the easiest path to cheaply preventing Covid-19 at a time when the only drugs available to prevent hospitalization from Covid-19 were monoclonal antibodies, which have to be injected or infused over an extended period.

Their efforts offered a test for how easily physicians could get a drug through regulators in a crisis without the support of patents or a pharma sponsor.

With the new drugs now available, Boulware acknowledged fluvoxamine is less crucial. But he said it could have been hugely useful during the Omicron surge in December, when he initially filed the paperwork for authorization.

At the time, neither Paxlovid nor molnupiravir were widely available and the new variant had rendered most monoclonals useless.

“This was highly relevant back in December, when we had no Paxlovid, no antibodies,” he said. “Now it’s less relevant.”

Still, he said, more trials are ongoing. And if more evidence comes in, he and other researchers may resubmit. Trials for other repurposed medicines are also ongoing.

Create a display name to comment

This name will appear with your comment