Most of the scientific trials looking for drugs to use against Covid-19 were crap.
They either weren’t randomized—they didn’t compare people who get a drug to similar people who don’t—or they were statistically underpowered, with too few subjects to draw any real conclusions.
Now, don’t go all “#NotAllTrials” on me. For sure, the last year has seen notable, lifesaving research that changed the course of the pandemic—like the British Randomised Evaluation of Covid-19 Therapy Trial (or “Recovery”) result showing that the cheap steroid dexamethasone helps seriously ill people survive. Boom. Good stuff. But broadly, even the good trials found more drugs that didn’t work than drugs that did—important to know, but it doesn’t tell you which drugs get people home from the hospital sooner.
In the last couple weeks, researchers in Europe and the United States have spun up new efforts to change all that—large-scale, randomized trials of multiple drugs designed to get some real answers. Mass vaccination has crushed Covid-19 case counts and deaths around the world, but countries like India and Brazil are still in crisis. Even in places where Covid-19 seems to be getting under control, it might never disappear completely. It would be, I’m saying, very good to have drugs that made people with Covid survive more often, or even just feel better sooner. The fact that they’re not here already—and that studies with the best chance of finding them are only spinning up now—illustrates everything that’s wrong with the way drugs currently get identified and evaluated.
Some numbers: Between the emergence of Covid-19 and last November, ClinicalTrials.gov and the World Health Organization International Clinical Trial Registry Platform marked more than 2,000 Covid drug trials looking at more than half a million patients and 2,895 different drugs or combinations of drugs—the “arms” of a trial. According to two leaders of the US Food and Drug Administration—acting commissioner Janet Woodcock and Kevin Bugin of the Center for Drug Evaluation and Research, writing in the journal Nature Reviews Drug Discovery—just 5 percent of those trial arms were randomized or statistically sound. “They were just too small, and most of them didn’t even enroll the number they aspired to, so we ended up with no information,” Woodcock tells me. “All these patients sucked into hundreds of trials, and they didn’t get any results.”
Woodcock and Bugin confirmed what Covid researchers already knew. I wrote about this problem almost a year ago, when a study of 1,551 Covid drug trials registered in the database ClinicalTrials.gov between March and May 2020 showed that barely a third could yield high-quality results. At about the same time, an article in Stat looking at Covid trials over the first half of 2020 found that nearly half planned to enroll fewer than 100 subjects, likely too small to show anything—and that one in six studies were on the drugs chloroquine and hydroxychloroquine. (Those continued even after evidence emerged that the drugs were useless against Covid-19.) None of these researchers used this kind of language, but I’ll say it again: crap.
Now, Woodcock—along with colleagues centered at the National Institutes of Health, and other sprawling networks of researchers around the world—are working to get some new momentum. The World Health Organization is relaunching its multi-arm Solidarity trial: It started in March of 2020 to look at antiviral medications, but none of those panned out. Now, since Covid can induce a dangerous overreaction by a person’s immune system, the trial will look at a set of repurposed immune-suppressing drugs. (Nature reports that the trial will be looking at an immune suppressor called infliximab, the anti-inflammatory imatinib, and an antimalarial called artesunate. WHO spokesperson Tarik Jašarević wouldn’t confirm those drugs but said that “intense preparatory work is underway in more than 45 countries to prepare to rapidly assess these new treatment options.”)
In the US, a new iteration of a series of trials called Accelerating Covid-19 Therapies and Vaccines (sigh, “Activ”) will also take on repurposed drugs—the researchers haven’t said which ones yet—to treat people with milder or earlier Covid. Earlier Activ trials found that, for example, monoclonal antibodies helped fight the virus, but they are expensive and difficult to use—they have to be given intravenously, which requires access to a hospital. Another Activ trial helped work out protocols for using drugs that treat blood clots—another bad Covid symptom. This new one is Activ-6, which “plans to look at people who are very early in their disease, and look at oral, repurposed agents,” Woodcock says. “It’s probably something we should have started earlier. I wish we had done that.”
There are other European studies too. One already underway will also be going after useful repurposed drugs. It’s called—deep breath, here we go—Randomised, Embedded, Multi-Factorial, Adaptive Platform Trial for Community-Acquired Pneumonia. (Yes, that’s Remap-Cap.) It has already chalked up a rare win by getting good results with the drugs tocilizumab and sarilumab, inhibitors of an inflammatory molecule called interleukin-6; block that and you get an anti-inflammatory, immunosuppressive effect.
Remap-Cap was successful because—like the folks running Recovery and those now on Solidarity and Activ-6—it is a “platform trial,” built for scale and speed. “We had designed it to be ready for a pandemic. The ‘A’ and the ‘P’—it was designed to be an adaptive platform,” says Anthony Gordon, chair in anesthesia and critical care at Imperial College and the UK chief investigator on the trial. “That’s why we were studying community-acquired pneumonia. We knew you could acquire it, and we’d have the general design features we could adapt should a pandemic come along.” Covid-19 is, after all, a community-acquired pneumonia itself. Gordon’s team put together a large network designed to run rapid, parallel tests. They started with antibiotics but could easily shift to antivirals or other drugs on the same master protocol.
The Solidarity and Recovery trials ran on similar principles—big groups of people, lots of different drugs. In the US, the Activ program did, too, in the spring of 2020, when hospitals were undersupplied and beyond capacity, filled with Covid patients.
But Activ’s first round of studies were most helpful in finding what didn’t work. That’s not necessarily bad—knowing which drugs don’t help is as important as knowing which ones do. In fact, the new one, Activ-6, will try to do both. “When we got to Activ-6, we realized we had to redesign the criteria yet again. We weren’t just trying to find the most efficacious drugs,” says Stacey Adam, scientific program manager for cancer at the Foundation for the National Institutes of Health—that’s the group that deals with partnerships between NIH and private companies like drugmakers. “What were people using anecdotally, maybe to the detriment of patients? That actually became a criteria—was this being used?” she recalls.
So the researchers set out with an eye toward testing drugs already in use that might not work, and drugs no one had tried yet that might. They ended up with 1,500 suggestions from physicians, drug companies, and researchers. A panel of advisers got that down to 30 probables, and now the four that they’ll eventually unveil as the arms of the trial. A trial with multiple arms and a shared control group can ultimately use fewer volunteers per arm and still get statistically significant results. “We think we can make a go/no-go decision with hundreds of subjects,” says Sarah Dunsmore, program director in the Division of Clinical Innovation at the National Center for Advancing Translational Sciences. If a drug looks good, it might then advance to a clinical arm with a few thousand people.
That feels like a good way to deal with the kinds of problems Woodcock and others identified with the drug-trial pipeline. In fact, Woodcock’s FDA this week issued a giant set of new guidelines for creating more of these master-protocol trials.
But none of that solves an equally urgent problem: time. It’ll be difficult for any of these studies to find Covid patients to enroll. “The nature of a pandemic is, it will have peaks and troughs at different times and places in the world. Right now our biggest recruiters, you won’t be surprised, are in India,” says Gordon of Remap-Cap. “Now, they are not as well set up for research as the UK and US are, so their numbers of recruits are far fewer than we recruited in the UK, even though their caseload is higher.”
Vaccines won’t ever eliminate Covid-19 completely. In countries where they’re available, some people refuse to take them; some countries can’t afford them. So good Covid drugs are still important. But even so, these trials feel a little late. “The concept is great. The concept a year ago would have been even better,” says David Boulware, an infectious disease physician and researcher at the University of Minnesota Medical School who’s involved with Activ-6 and some other Covid trials. Don’t chalk the delay up to science, he says, but politics.
Woodcock was at the FDA last year and worked on therapeutics as part of the Trump administration’s Operation Warp Speed. That program helped produce the vaccines that are now beating the pandemic, but the urgency for therapeutics was less … urgent. “Obviously the prior administration was not really interested in research, because this was all going to go away by last Easter,” Boulware says, tweaking then president Trump’s unfounded optimism in the early days of the pandemic. “So really this all came about after January 20, 2021.”
Wherever politics is an obstacle, so is money. (As Carl Zimmer wrote in The New York Times, by last January the US government had spent about $18 billion on vaccine research and development and about $8 billion on therapeutics.) As always, though, pharma money does a lot to clear a drug’s path. One of the early sort-of successes in the hunt for Covid drugs was remdesivir, a glossy antiviral made by the drug company Gilead; a US study conducted with the company’s aid and materials found in April of 2020 that it decreased the length of time people had symptoms; later studies including Solidarity found no impact on survival.
But off-patent drugs don’t make pharma companies the same money, so they don’t get the same corporate push. Studying cheap, repurposed medicines usually requires government funding. “A lot of these are generic drugs. So why have these not been pushed forward? Because there’s no patent, and therefore no profit motive for a pharmaceutical company. There’s no drug company saying, ‘We’ll give you $10 million to look at this,’” Boulware says. “So the government has to do it, and the government has to want to do it. The good news is, these are drugs available in low- and middle-income countries, or down the street at the Walgreens. It’s not ‘We’ve invented some newfangled medicine that we don’t actually have any of, but six months from now we’ll have enough to treat 10,000 people and it is going to cost $10,000 a dose.’”
Which leads to the other, grander insight that Woodcock hopes will prevent small-trial chaos in the next pandemic—more economical and more efficient master protocols for testing more than one drug at once. “What was interesting about this pandemic, if you look at all the results for therapeutics, especially for the immunomodulators, is there’s currently still a lot of back and forth about what regimen should be used, and we have conflicting trial results,” Woodcock says. “What that usually means is, there were small treatment effects and the trials weren’t powered adequately to give a definitive answer.” Big, multi-arm, master protocol studies are meant to bridge the gap between big drug companies running big, expensive trials on big, expensive drugs and the small, idiosyncratic ones that don’t produce enough new knowledge. During the pandemic, there haven’t been enough of those government-funded, ambitious, middle-ground studies—a flaw in the system that cost lives.
- 📩 The latest on tech, science, and more: Get our newsletters!
- The 60-year-old scientific screwup that helped Covid kill
- Covid forced the US to make more stuff. What happens now?
- How to find a vaccine appointment and what to expect
- The statistical secrets of Covid-19 vaccines
- The case for letting people work from home forever
- Read all of our coronavirus coverage here