Skip to Main Content

When H. Eric Xu fell sick and tested positive with Covid this month, he took pills of his own invention. 

Paxlovid, the primary oral antiviral available in the U.S., is in short supply in China. But Xu, a prominent biochemist at the Chinese Academy of Sciences, had helped devise his own in 2021 and, as the drug’s inventor, could legally take it even as it went through clinical trials. 


His symptoms were alleviated, and he wasn’t alone. In an 822-patient clinical trial published Wednesday in the New England Journal of Medicine, Xu’s antiviral, VV116, appeared just as effective as Paxlovid at easing patient’s symptoms. Newly diagnosed patients who took VV116 recovered after a median of 4 days, compared with 5 days for Paxlovid, though experts caution that the study was not large enough to say one was better than the other.

The drug is being developed by Chinese drugmakers Junshi Biosciences and Vigonvita Life Science.

The results could lend Chinese authorities and physicians a new tool as they try to keep a burgeoning outbreak at bay, while eventually adding another antiviral to the world’s arsenal.


“I hope the publication of these results will get approval much quicker,” Xu said in an interview.

Quicker, however, may not be quick enough. The virus has spread rapidly in China since authorities lifted many of the country’s strict mitigation measures in response to protests over its zero-Covid policies.

“It will take time to manufacture,” he added. “I think the pandemic will pass.”

Although promising, the data may not in fact be enough to win over regulators and other researchers, particularly in countries where alternatives are available. 

Part of the problem is the study’s main metric. While the FDA has previously approved antivirals, such as the flu drugs Xofluza and Tamiflu, that are shown to alleviate symptoms faster than alternatives, the primary endpoint of U.S. trials of antivirals to treat Covid has been their effectiveness in preventing hospitalization or death. . 

It isn’t actually clear how quickly Paxlovid — the drug VV116 was compared to — alleviates symptoms in all patients. 

Pfizer presented data at ID Week, an annual infectious disease conference, this year showing high-risk unvaccinated patients who took Paxlovid recovered faster than patients on placebo, said Davey Smith, an infectious disease physician at University of California-San Diego. But the company acknowledged this year that vaccinated, standard-risk patients who received Paxlovid didn’t recover significantly faster than placebo in a separate trial. 

The VV116 study contained a mix of vaccinated and unvaccinated patients, randomized to receive VV116 or Paxlovid. The majority — around 75% — had received at least a primary vaccine course. It was also smaller than typical non-inferiority studies, in which  researchers try to show a new drug is just as good as an existing one, said Smith. Usually, these studies have thousands of patients. 

“Our FDA would probably not go for this being an approved drug, based on this endpoint,” Smith said.

David Perlin, who develops new anti-infectives as chief scientist at the Center of Discovery and Innovation in New Jersey, said VV116 appeared to be a “good addition” and little different than the approved antiviral molnupiravir, which has a similar mechanism of action.  But he agreed there wasn’t enough data for U.S. authorities. 

“With 800 participants, I’m not sure the FDA is going to extract enough of the adverse events you would want to look for,” he said. 

The drug itself has attracted broad interest. It is a derivative of Gilead’s remdesivir, the first antiviral authorized for Covid-19 in May, 2020. Xu said he and a team of researchers, including collaborators at the Wuhan Institute of Virology, took remdesivir’s “parent” molecule, called GS-441524, and spent six months modifying it, until they struck upon a version that could neutralize the virus throughout the body, without causing toxicity. 

In early 2020, a pair of researchers called on Gilead to develop GS-441524 itself as a pill for Covid. In public statements and emails to the researchers, Gilead claimed it wouldn’t be effective or safe. And it would take longer to develop, because, unlike remdesivir, it had never been tested in humans. But the researchers argued that Gilead preferred remdesivir because it had a longer patent life. 

Those scientists now see VV116’s results as vindication.

“As for whether an effective pill could have been available far earlier in the pandemic if Gilead did develop GS-441524: The answer is yes” said Victoria Yan, one of the researchers and a graduate student at MD Anderson Cancer Center, in an email. 

Xu argued it’s likely GS-441524 would have worked to some degree, but that changes his team made ensured it was potent in the body. Yan agreed that one of the changes did help. 

Gilead reiterated in an email that the company didn’t develop GS-441524 because of concerns about “technical feasibility.” The company announced in January it is working on an oral antiviral, based off of GS-441524, which is now in Phase 3.

It’s among a small handful of efforts, including VV116, to develop antiviral pills that are as effective as Paxlovid but don’t come with the same drug-to-drug interactions that prevent some of the highest risk patients from taking Pfizer’s drug. 

These efforts could also make antivirals more available across the globe. Although molnupiravir, a less effective drug, is now widely available, access to Paxlovid has been “pretty limited,” said Julien Potet, a policy advisor to Doctors Without Borders. 

That could be because Pfizer started striking agreements with generic manufacturers later than Merck, molnupiravir’s developer, and Paxlovid is more complex to make, Potet said.   

“If reasonably priced, [VV116] could make a major difference,” he added in an email.

The trials to make these new drugs available in the U.S. will have to look different than the one Junshi ran. Currently, U.S. researchers can’t easily run non-inferiority studies against Paxlovid. Because the drug is under emergency use authorization, researchers can only get it from Pfizer, but academics have so far struggled to access the drug for outside trials. As part of the team of researchers working on National Institutes of Health trials for Covid, Smith said he tried and failed to get Paxlovid for a non-inferiority study.

“You could not run this non-inferiority trial here,” he said. 

Companies developing new Covid antivirals have complained to federal officials that it has become impossible to run studies, said one government scientist involved in such talks, who spoke on condition of anonymity to discuss confidential deliberations. The FDA has asked them to compare their experimental drug to standard-of-care, but Pfizer refused to provide Paxlovid for the studies.

“I’ve heard multiple times from multiple drug companies,” the researcher said.

Drugmakers have had to go to other countries where regulations or the standard-of-care are different, the scientist said, and then hope they can convince the FDA to accept the data. Gilead, for example, currently lists three study sites for its Phase 3 trial. Two are in Canada. One is in Romania. 

Pfizer, in a statement, said it “is prioritizing”  Paxlovid for “patients at high risk for severe COVID-19 through governments and government-funded research organizations” and that, while it has received many research requests, “unfortunately, it is impossible to support them all.”

“We evaluate each proposal, selecting those with strongest scientific basis, the greatest likelihood of generating data important to patients and physicians, the intention to generate new data that is not already being collected through different means, and the ability of the requesting organization to protect the welfare of patients participating in research studies,” they said. 

The story has been updated to correct a name and to include a comment from Gilead. 

Create a display name to comment

This name will appear with your comment

There was an error saving your display name. Please check and try again.