Vaccines are great, and you should get yours. But for almost two years now, alongside the showy and ultimately successful hunt for vaccines that fight Covid-19, scientists have also been looking for drugs to treat people who were already sick. That hasn’t gone quite as well: a scientific quest tangled up in hundreds of studies too small to give real answers, then sidetracked by hype over the antimalarial hydroxychloroquine and then the antiparasitic ivermectin. Those didn’t work. But the research continued—and in the past few weeks, it seems to have paid off.
First, in early October, the transnational pharmaceutical company Merck and a biotech company called Ridgeback announced that molnupiravir, a decades-old antiviral drug invented at Emory University, reduced the risk of hospitalizations and deaths in people with Covid-19 by a respectable 50 percent. Then, in early November, the equally transnational pharmaceutical company Pfizer (you might remember it from its mRNA-based Covid vaccine) announced that its purpose-built antiviral Paxlovid likewise reduced the hospital-and-death risk in high-risk Covid patients by a whopping 89 percent.
That’s a win, right? Covid vaccines are finicky, hard to manufacture, and hard to distribute because of, in the case of the ones based on mRNA, a snowflake-like need for ultracold freezers. The steroid dexamethasone is for the severely ill. Monoclonal antibodies have to be administered early to work, and, like the antiviral drug remdesivir, they’re expensive, and both require hospital visits for intravenous infusion. But molnupiravir and Paxlovid are “small-molecule” drugs, easier to make, more stable in storage and distribution, and—this is the big one—they’re just pills. You pop ’em. On a planet ravaged by a pandemic virus, with effective vaccines unavailable in most places, easy-to-use, relatively cheap drugs could help beat the disease. “They’re game-changing in the sense that they’re both pills and therefore relatively easy to take,” says Charles Gore, executive director of the Medicines Patent Pool, a UN-supported organization that sets up international licensing agreements for drugs. “You don’t have to lie down to get an infusion, or walk 150 miles to get to a hospital.”
That’s convenient. And both Merck and Pfizer have also announced extensive licensing agreements with generic drugmakers and tiered pricing for lower-income countries, which is good for places that haven’t been able to afford the vaccines that North America and Europe have been hoarding. “I think the combination of these drugs and vaccines will really make a difference,” says Jayasree Iyer, CEO of the Access to Medicine Foundation, which works on getting critical drugs into lower-income countries. Vaccine makers know they have to do the same kind of technology transfer and licensing agreements for those, Iyer says, “but that’s going to take longer, and the immediate need now has an immediate drug.”
So that’s the good news, then. Merck, first out of the gate, was working with a drug that has a nifty mode of action. It’s a synthetic analog of one of the nucleoside “bases” from which the virus builds its genetic material. During replication, the virus uses the fake version instead, and boom! Virus, killed.
In parallel with its drug trials, Merck was building a global supply chain for manufacturing partners and putting together a deal with the Medicines Patent Pool for easier licensing. That group has been a good partner to pharmaceutical companies since the first deals for HIV and hepatitis-C drugs and will also help ensure quality and international approval and distribution.
Building the systems to get the drugs to people before they were approved by regulatory bodies like the US Food and Drug Administration is an unusual approach. And Merck was doing the same with two vaccine candidates and another therapeutic drug—none of which worked. This is called “at risk,” a kind of bet that the drug will succeed on spec, and it costs money but saves time. “As we were going through Phase 1 and Phase 2 studies, they were building this global supply chain to have us in the position where, by the end of this year, we will have produced 10 million courses of treatment. And we still haven’t reached our advisory committee meeting with the FDA yet,” says Paul Schaper, Merck’s executive director of global pharmaceutical policy. “At the same time, we started talking to governments and other stakeholders about doing advanced purchase agreements. This was with high-income governments, regional authorities, middle-income governments, and global public health agencies.”
Schaper says the company will have 3 million courses of treatment ready to deploy to low- and middle-income countries as soon as approval comes, and then international generics makers will be able to make medicine for more than 100 countries. Molnupiravir is already approved in the United Kingdom, and the US government has agreed to buy 1.4 million courses of treatment. “Pricing indications, from some academic work on the Merck drug, suggest that it could be around $20 for the five-day course, falling to $10 if they use a process that the Gates Foundation has funded, which optimized production,” Gore says. “It’s significantly lower than what Merck is going to charge in the US, which I believe is around $700.”
(Take a moment to be angry about this, because the initial research at Emory was funded by federal grants—which is to say, US taxpayer dollars. So you might ask why US citizens have to pay any more for this drug than the companies’ costs to bring it to market. But then if you did ask that, what are you, some kind of socialist?)
Pfizer, on the other hand, has only just started making pills, but the company also has an agreement with the Medicines Patent Pool. Its Paxlovid blocks an enzyme that the SARS-CoV-2 virus uses to cut big chunks of protein into the sizes and shapes the virus needs to reproduce itself—a “protease inhibitor” like the drugs that were instrumental in the fight against HIV and AIDS. In Paxlovid, that protease inhibitor comes with another drug, ritonavir, that keeps a person’s liver from breaking down the protease inhibitor. A Pfizer spokesperson declined to make anyone from the company available for an interview, but the press release in which Pfizer announced its early trial results also quoted CEO Albert Bourla calling the drug “a real game changer,” for what that’s worth.
According to Pfizer’s spokesperson, the company expects to be able to make 120,000 full courses of the drug before the end of the year, and 50 million courses a year after that. The license deals the company has made via the MPP guarantee tiered pricing to 95 countries—more than half the humans on Earth. And Pfizer has said generic drugmakers in low-income countries need never pay them royalties—in fact, the company won’t collect royalties from anyone until the World Health Organization says the pandemic is over.
But the good international licensing deals only matter if the drugs work. And that’s still a bit of a question. The numbers from both trials look great, but they come from press releases instead of published journal articles, or even preprints. So independent researchers haven’t been able to scrutinize the details of the data—like demographics, for instance.
Both companies offered numbers from “interim analyses,” planned stopping points in a drug trial. Those are moments when an independent Data Safety and Monitoring Committee looks at the data collected thus far and says, basically, “hmm, keep going,” or “no, stop, this is terrible”—or, as happened here, that the results are so good it’s unethical to keep people on a placebo.
That’s tricky, though, because in these cases, the number of actual events in the studies—the number of times someone got sick or died—is relatively small. So it’s hard to extract useful conclusions. In the Merck trial, 28 out of the 385 people who got molnupiravir required hospitalization or died, versus 53 of the 377 people who’d received a placebo. That’s a roughly 50 percent reduction in risk, which sounds great. But tweak either of those numbers by, like, five people in either direction, and the risk reduction would change wildly.
The same goes for Paxlovid. The 89 percent risk-reduction number comes from 3 out of 389 people who got the drug requiring hospitalization—none died—versus 27 out of the 385 on the placebo arm going to the hospital. (Seven of them died.) That’s statistically significant, but it’s also teeny tiny. “They presented the findings based on early interim analyses, and then press-released the interim analysis. Most Data Safety and Monitoring Committees would not have recommended stopping the trials or making the announcements based on the small number of events that occurred,” says Edward Mills, a health research methods researcher at McMaster University and a leader of the Together Trial, which is testing a bunch of drugs that could be repurposed against Covid. “When you have small event sizes, you end up with dramatic-sounding effects numbers that are not believable.” Mills says that in his trial, the promising drug fluvoxamine had similar early results at an interim analysis, and his DSMC told him to keep going, that it was too early to tell if the numbers were real. (The fluvoxamine turned out to perform pretty well.)
Beyond that, both molnupiravir and Paxlovid have to be given early in the course of the disease, in what’s called the viremic stage, when the virus is still reproducing. One of the things that made SARS-CoV-2 a successful pandemic virus is its ability to transmit from person to person before they show any symptoms. People generally don’t get to the doctor with Covid until several days after they first get infected. By then, their career as a spreader of the disease is pretty much over. Pfizer’s big 89 percent number comes from data collected after just three days of infection. “It’s a completely specious way to present findings,” Mills says. “No one in their right mind would design a trial where the primary outcome was the first three days of infection.”
So why would these drug companies put such a big push behind drugs if their efficacy wasn’t all that they claimed? Reputation and market share—and maybe the ability to sell other drugs once Covid is under control. “The more drugs like these are available in India, China, and sub-Saharan Africa, then Pfizer can sell all the other drugs they want to sell,” Iyer says. “Hospitals can absorb patients with cancer, or with diseases for all the other drugs they make. And reputationally, it helps a lot.”
To be clear, none of this means that the drugs have no effect. It’s likely that, for example, Paxlovid’s 89 percent just shrinks over time. “If these medicines are just going to be used on hospital patients once they show up, 10 days into their illness, they’re not going to have much of an impact,” says David Boulware, an infectious disease physician at the University of Minnesota involved in multiple Covid drug trials. For one thing, people have to get tested for Covid before getting their prescription. “That’s complicated. It’s hard. Low-income countries, it’s easier to vaccinate people. If you can vaccinate the high-risk people, that’s a much better prospect than trying to deal with both diagnostics and treatment.”
Both drugs have complications in terms of actual use, too. Molnupiravir has been kicking around for 20 years, and it’s been aimed at lots of different diseases. Mills says most virologists had despaired of it ever finding a real use—in part because of mutagenicity issues that researchers have seen in analyses and trials over the years. It likely won’t be used in women who are or might become pregnant, for example. A recent article in Stat said that Merck’s animal testing of molnupiravir hadn’t found any problems, and Stat quoted a Merck executive VP telling investors that the company is “confident” in the drug’s safety profile.
Ritonavir, that second ingredient in the Pfizer package, doesn’t just inhibit liver activity; it also has well-known interactions with anticoagulant drugs and widely prescribed anticholesterol drugs. So people taking those may have to change their doses, or stop taking them for the five days of the full course. Not impossible, but not easy, either.
All that said, many researchers and public health experts believe that the new drugs will help. Even Mills says he thinks they’ll have an impact, though it’s hard to tell how big. And the total absence of deaths in the Paxlovid trial could be amazing, if the pattern holds over tens of thousands of people. “I am very optimistic here,” says Tulio de Oliveira, director of the KwaZulu-Natal Research and Innovation Sequencing Platform at the University of KwaZulu-Natal in South Africa. “The companies, Merck and Pfizer, released clear guidelines that the price of the drugs will be much lower to developing countries. Furthermore, they are already engaging with a number of local companies to produce the drugs in the developing world.”
The flip side of the big pharma companies having to justify their giant pre-orders to governments is that those governments may well have already seen more data than the companies have released publicly. “They have both trials. They have seen data we have not,” says Eric Topol, director of the Scripps Research Translational Institute. “They have seen beyond the press release and have bought, as you’ve seen, tens of millions of doses of the two pills.” To Topol, who has been tracking vaccine and drug development efforts throughout the pandemic, that suggests the unpublished data looks as good as the public numbers do.
In fact, lots of these questions could be obviated by a fuller release of the results from the companies’ trials. That especially the case for one that Merck has ongoing, on molnupiravir’s potential to keep people from getting sick after they’ve been exposed to the virus. That’s called post-exposure prophylaxis, and it would make a huge dent in Covid’s spread. “That’s much more exciting from a pandemic perspective, where you can do prophylaxis for people who are immunocompromised or have had high-risk exposures, and you can use these antivirals to stop transmission,” Boulware says. “If you’re exposed, you have a little virus, boom, it should knock it down to keep you from getting infected, and if you are already infected, it should reduce the viral load quicker.”
Another thing it’d be nice to know is whether the drugs work better in combination—either with each other or with something else. That’d be a classic move against a virus—HIV typically gets a three-drug combo, and hepatitis C gets two. Covid, as an illness, moves from the viremic stage to an inflammatory one, and then (when it gets serious) it acts like an autoimmune disorder. “The majority of patients, by the time they visit a doctor, are somewhere between the viremic stage and the inflammatory stage,” Mills says. “It only makes sense that you could combine an antiviral with an anti-inflammatory. The combination should have both, and probably at some point we’ll figure out how to add something that is immune modulatory, as well.”
But no one is running trials yet to figure out how that’d work. Here, as has been typical throughout the course of the pandemic, big drug companies with new candidates have conducted their own trials or funded and helped organize trials of the drugs by themselves. Large-scale, multi-arm, adaptive trials have worked with repurposed, older, and off-patent drugs. The researchers who run Activ-6, the trial that Boulware works on, have talked about incorporating the newer drugs, but “there’s nothing official,” he says. “It takes two to tango, so I’m not sure what will happen with Merck or others.”
Mills says he has asked Merck several times if the company would give him—or if he can buy—enough molnupiravir to trial it in combination with fluvoxamine, a psychiatric drug with anti-inflammatory properties that has done well in his trial so far. The drugmaker hasn’t responded (and a company spokesperson didn’t respond to me when I asked why). “I think they’re trying to hold onto their market share for as long as possible,” Mills says.
Obviously, prevention is still better (and cheaper) than treatment. But vaccines still aren’t available in most of the world. “Our hope was that the equity problem was going to be solved by companies planning for these things and governments not hoarding the vaccines a year ago,” Iyer says, adding that she’s more optimistic that a second generation of vaccines will be cheaper and more widespread. Meanwhile, the Medicines Patent Pool is helping establish mRNA manufacturing capabilities in South Africa. But until those efforts (and political pressure) start to show results, these drugs offer real hope—even if they don’t fulfill the lofty expectations set by the corporate press releases.
- 📩 The latest on tech, science, and more: Get our newsletters!
- Moderna’s CEO on the fight over the Covid vaccine’s future
- The race to develop a vaccine against every coronavirus
- How to put a vaccine card on your phone
- How to find a vaccine appointment and what to expect
- Need a face mask? Here are ones we like to wear
- Read all of our coronavirus coverage here