The studies enrolled similar populations — unvaccinated people with mild to moderate COVID-19 and at least one risk factor for severe disease — but they had slightly different measures of efficacy. Pfizer’s 89% figure comes from patients who started getting its pill, Paxlovid, within three days of their first COVID-19 symptoms. Merck’s 50% applies to patients who began treatment within five days. In the Paxlovid study, patients who started treatment within five days saw an 85% improvement in hospitalization or death versus placebo. Merck has not shared data on patients who got its drug within three days of symptom onset.
What the studies had in common was 100% efficacy against death, regardless of when patients started treatment. Merck’s study counted eight deaths among patients on placebo, and Pfizer’s observed 10.
On the safety side, the rate of side effects in both studies was similar between the treatment groups and placebo groups. In each study, fewer patients in the treatment group left the study due to side effects compared to those in the placebo group. Neither company has disclosed detailed data on the type and severity of side effects.
Each treatment is administered twice a day for five total days, amounting to 10 doses in total. Pfizer’s drug is co-administered with a common antiviral called ritonavir.
Will the antiviral be available for vaccinated patients with breakthrough infections?
Both companies conducted their studies entirely in patients who were at high risk of complications if they caught COVID and who also had not been vaccinated. That leads to a big question for policymakers: Should those who have been vaccinated, but who develop a breakthrough infection of SARS-CoV-2, be given the pills?
Right now that is a question without data. A third antiviral pill, from the biotech firm Atea and the large drug firm Roche, failed to prove it was effective in its own study, and Wall Street analysts suspect the reason is that the companies included vaccinated patients in the research. For those who have received the vaccine, hospitalization and death are much less likely. This means that it is harder for a drug to show efficacy, because there are fewer infections to prevent.
So regulators and public health officials will have to make a judgement on the risks and benefits of the COVID pills for people with breakthrough infections — without direct data in these populations.
Pfizer is running a clinical trial, with results due next year, that does include vaccinated patients, and the company’s executives have expressed confidence based on the results so far that the treatment should work. Both Merck and Pfizer are also running studies to show that the drugs can prevent people from developing symptoms if they take the antivirals after they are exposed to the virus.
Do the drugs work the same way?
No, not really. While both drugs interfere with the process the coronavirus uses to reproduce itself, each drug interferes at a very different point.
Merck’s drug throws a wrench into the works quite early. After someone takes molnupiravir, the drug is transformed into something uncannily similar to one of RNA’s chemical building blocks.
The modification is so subtle that not only will the coronavirus use molnupiravir in place of other building blocks when it replicates itself, but coronaviruses’ unusual proofreading mechanism can’t even pick up on the imposter compound. Over time, the drug will encourage the virus to introduce even more mistakes.
“Ultimately, this leads to what’s known as error catastrophe. It’s introducing so many different mutations that, eventually, nothing further can happen,” said Katherine Seley-Radtke, a medicinal chemist at the University of Maryland, Baltimore County. “You’ve got this completely mutated RNA.”
Pfizer’s drug, Paxlovid, acts at a completely different point in the virus’ reproductive process.
“It’s apples and oranges,” said Ronald Swanstrom, a biochemistry professor at the University of North Carolina School of Medicine.
Unlike molnupiravir, Paxlovid allows the strings of viral RNA to be assembled correctly. It even allows those strings to be used to create viral proteins, which are initially produced in one big chunk. Like a bolt of fabric before it’s cut to a clothing pattern, this protein needs to be chopped down to size before it can work.
That cutting is what Paxlovid prohibits. The drug is designed to bind to a particularly important point in an enzyme called a protease which slices up proteins. Without a functioning protease, the virus can’t create functional copies; no working virus, no problem.
Protease inhibitors have been used for decades to create more than a dozen drugs for HIV and hepatitis C; in some cases, they’ve also been used as cancer drugs.
“There’s a long history of medicinal chemistry targeting proteases,” said Bryan Dickinson, a chemical biologist at the University of Chicago.
Paxlovid is designed with a SARS-CoV-2-specific protease in mind, so it works more specifically on this coronavirus than molnupiravir.
But Paxlovid can’t work as well if it’s taken on its own. The body’s defense mechanisms will get rid of anything that it doesn’t recognize — including drugs, which can be digested by enzymes in a person’s liver. Another drug called ritonavir blocks the liver enzyme that would likely chew up Paxlovid, which gives the latter drug the space it needs to work.
How do they compare with monoclonal antibodies?
Regeneron Pharmaceuticals and Eli Lilly have each won FDA authorization for antibody combination therapies that keep recently diagnosed COVID-19 patients from hospitalization and death. In a Phase 3 study enrolling recently diagnosed patients at high risk for severe disease, Regeneron’s treatment reduced the risk of hospitalization or death by 70% compared to placebo. In a similar study, Lilly’s therapy showed an 87% reduction.
The biggest difference is one of convenience. The antibody treatments are administered intravenously in a one-time, roughly hour-long process (Regeneron’s is authorized for subcutaneous injection when an IV procedure is not feasible). That could make the treatments from Pfizer and Merck, taken orally at home, preferable to patients unable to visit an infusion center.
There’s also a difference in cost. Regeneron and Lilly have signed deals with the federal government to sell their treatments at about $1,250 per dose. Merck’s agreement with the U.S. works out to about $700 for a five-day course of molnupiravir. Pfizer is still negotiating contracts but is expected to set a similar price for Paxlovid.
How easy will they be to get?
A pill is a huge leap in terms of logistical ease over infused therapies like monoclonal antibodies. For those treatments, not only did people have to make their way to clinics for their infusions, but hospitals and other facilities had to set up places where people who were actively infectious could come get treated without risking others’ health. (The other antiviral authorized to treat COVID-19, Gilead’s remdesivir, is an infusion and approved only for hospitalized patients, but some data indicate that if it were to be given to patients earlier in their infections, it could have a greater effect. If its approval ever covered outpatients, however, it would still run into the same logistical challenges of an infused therapy.)
Still, the COVID pills come with a key challenge of their own. They’re most effective when given early in the infection, so people need to be able to get tested and get their prescription rapidly. And the U.S. testing landscape is still limited. PCR tests can take days to return a result, and though the Biden administration has upped its effort to expand the availability of at-home rapid tests, finding one at a store is still hit or miss — success feels like scoring this holiday season’s hottest gift. Any delay in getting diagnosed undercuts the power of these pills; even a day or two has real implications for a treatment meant to clear out an acute infection like COVID-19.
Will it affect a patient’s DNA?
This is really a question only for Merck’s molnupiravir, since it works by sneaking subtly corrupted parts into the coronavirus’s RNA sequence.
Once the virus has mutated too much, it can’t work — mission accomplished. But there’s a theoretical chance that molnupiravir could also influence normal human DNA when it replicates, too. If mutations happen during that process, it could spell real trouble.
Merck did some tests during molnupiravir’s development to check this possibility out. In two different types of animal studies using higher and longer doses than are given to humans, Merck’s scientists didn’t see any increased risk of unwanted mutations.
“We are very confident in the safety profile of molnupiravir based on our preclinical and clinical data,” executive vice president Dean Li told investors in an October conference call, according to a transcript in the financial database Sentieo.
But UNC’s Swanstrom isn’t completely convinced that the tests Merck did were sensitive enough. In August, he and his colleagues published a paper in the Journal of Infectious Diseases showing that a key metabolite of molnupiravir could mutate DNA in animal cells.
Given these results, Swanstrom said he would be particularly interested in seeing a long-term study of people who took molnupiravir to continue to monitor this potential effect over the next 10 or 20 years.
“This thing is going to go into thousands of people. And are we just going to ignore the fact that there’s this potential risk?” he said. “The risk could be zero. It could be no worse than going to get a dental X-ray — or it could do something more. But unless we find out, you know, we’re going to learn this lesson the hard way, way later than we should.”
What might the new antivirals mean for cancer patients?
Because the new antiviral Pfizer is developing is a protease inhibitor, infectious disease specialists are familiar with how it works. So we already know: These drugs have the potential to interfere with many therapies used to treat cancer, Tobias Hohl, chief of the infectious diseases service at Memorial Sloan Kettering Cancer Center in New York, told STAT.
“They’re going to be very helpful [in our] armamentarium, but they will not substitute for prevention or vaccination efforts because these are not medicines that are completely benign and harmless in terms of their drug-drug interactions and toxicity,” he said. “So we’re going to have to be careful and thoughtful about how we use our protease inhibitors.”
This story was originally published by STAT, an online publication of Boston Globe Media that covers health, medicine, and scientific discovery.