Here’s What Scientists Know About Paxlovid Rebound

Dr. David Ho has spent his entire career studying HIV and other viruses, so he thought he knew what to expect when he recently became infected with SARS-CoV-2. His symptoms weren’t very severe, but after discussing it with his doctor, Ho decided to take Paxlovid, the antiviral COVID-19 therapy made by Pfizer, for the five-day course. At 69, he fit the description of someone who should. The U.S. Food and Drug Administration (FDA) and the Centers for Disease Control and Prevention (CDC) recommend Paxlovid for people at higher risk of severe COVID-19, including seniors.

Ho carefully tracked his infection. He had access to PCR testing and genetic sequencing in his lab at Columbia University, as well as rapid antigen kits for self-testing at home. He started taking Paxlovid a day after he first experienced cold symptoms and an antigen test turned positive. A PCR test on day 2 confirmed the infection, and Ho continued taking the daily therapy of two different pills for five days, as indicated. Every day from day 4 to day 9, his antigen tests were negative, and two PCR tests performed on day 5 and day 7 were also negative.

But on day 10, he had a headache, runny nose, and mild cough. Ho decided to test himself at home, and to his surprise, he was again positive, which a PCR test confirmed. His team sequenced the virus to compare it to the genetic sequence from his earlier infection and found that it was identical to the virus that he had been infected with days before—suggesting that the virus hadn’t mutated to become resistant to Paxlovid, nor had he become infected again with a different virus. The same infection had come roaring back after being quelled for a few days.

“I was surprised to see the rebound,” he says. “I was expecting Paxlovid to take care of things.”

A rebound like Ho’s was always a known potential side effect of the drug. In studies, Pfizer reported that 1% to 2% of people taking the medication experienced rebound. The company says the rates of rebound in the treated group in its study and among those receiving placebo were similar, indicating that “elevated nasal viral RNA is uncommon and not uniquely associated with treatment.” But a growing number of people are reporting infection rebounds after taking Paxlovid—so many that “there’s no way it’s occurring at 1% or 2%,” says Ho, who has also shared his data and discussed it with scientists at Pfizer. “It’s happening quite a bit.” Anecdotes from people who test positive, take a course of Paxlovid, test negative, and then days later test positive again abound on social media. It’s been scientifically documented; researchers at the University of California, San Diego recently described one case of rebound in a May 18 paper published on the pre-print server Research Square (which means it has not yet been peer reviewed). And on May 24, the CDC issued a health advisory instructing doctors caring for patients who take the drug about how to manage rebound cases.

On May 31, CDC director Dr. Rochelle Walensky warned people taking the drug that they “might get symptoms again,” but told CBS News that those symptoms are generally mild and that so far, have not been serious enough to send people to the hospital for more intensive care. She stressed that the benefits of the drug therapy continue to outweigh the potential risk of rebound infections.

Ho decided to study the phenomenon further when he learned that a colleague, Dr. Michael Charness at the VA Boston Healthcare System, had a similar experience. The scientists (who are both vaccinated and boosted) teamed up and conducted what is probably the most comprehensive analysis of the phenomenon to date, because they were able to conduct daily testing and track the flip from negative to positive and also perform genetic sequencing of the virus to confirm that the infection was caused by the same virus rebounding back, rather than a new infection. Their paper, published on Research Square, describes their own cases, as well as eight additional ones in which people reported rebounding virus after taking Paxlovid.

Here’s what we know about Paxlovid rebound so far.

Why do people even take Paxlovid if their COVID-19 could rebound?

Paxlovid is a combination of two drugs: a protease inhibitor used to treat HIV infections that blocks the virus from making critical proteins it needs to replicate, and another that prevents the liver from breaking down the drug too quickly in the body.

It’s not currently FDA approved, but doctors can prescribe it under an emergency use authorization. The authorization was based on studies conducted by Pfizer that showed the drug was up to 88% effective in protecting unvaccinated people from needing hospitalization for COVID-19, or from dying of the infection. The company is continuing to study Paxlovid in those who are vaccinated and boosted to see if their immune responses somehow react differently to the drug.

In a statement to TIME, a Pfizer spokesperson said “We are continuing to monitor the data, but we have not seen any resistance emerge to date in patients treated with Paxlovid. We believe the return of elevated detected nasal viral RNA is uncommon and not uniquely associated with treatment. We remain very confident in Paxlovid’s clinical effectiveness at preventing severe outcomes from COVID-19 in high-risk patients.”

Why does Paxlovid rebound happen?

Scientists are still trying to determine why people with COVID-19 are taking Paxlovid, testing negative for several days, and then testing positive again. But they are exploring several potential reasons. One hypothesis is that people are simply getting infected again in quick succession with different variants of SARS-CoV-2, since cases are mounting in the country and because the virus is mutating so quickly. Another explanation might be that because the virus mutates so rapidly, it has developed a way to escape Paxlovid and become resistant to its effects.

But Ho believes something else is going on. His small study showed that three of the infections in his report—his, Charness’, and that of one other person—were rebounds of the first, not new strains or a new Paxlovid-resistant variant. Based on the genetic sequencing data, “we showed the rebounding virus is not resistant to the protease inhibitor [drug in Paxlovid], because there is no change in protease sequence,” he says. “It’s not reinfection with another virus. The sequences are identical.” Pfizer says its studies similarly showed that the virus is not developing resistance to the protease inhibitor drug.

Instead, Ho believes the virus is just doing what viruses do in response to antiviral medications. No antiviral drugs—whether against SARS-CoV-2 or HIV—actually kill any virus present in an infected person; they work by blocking certain steps in the virus’ replication cycle, freezing them in time so the virus cannot continue to churn out more copies of itself. Paxlovid specifically interrupts the step involving the protease enzyme, which splices the long string of polypeptides that the virus makes once it infects a cell. That lengthy protein needs to be cut into its component proteins, which then play roles in making a new virus. By inhibiting that step, Paxlovid leaves the virus in a suspended state—and depending on when the drug was started, an infected person could have thousands of these long polypeptide strings already made and floating around in the body, creating a latent reservoir of viral product that could become active again. The drug prevents these intermediary forms from getting cut and infecting cells, but once the drug stops after day 5, the enzyme is no longer inhibited and can go back to splicing and making viral proteins. That could lead to a rebound, because more virus is being made that can infect cells again.

“We know the drug has a very short half life, which means it doesn’t linger for long in the body,” says Ho. “The virus hasn’t been killed by the drug, but only blocks its replication. There is still a reservoir of viral material, so for a virologist, that’s the first thing we think of. Once the drug is washed out, that reservoir can become active again.”

Ho is currently doing experiments in his lab to figure out how long the intermediary polypeptide of SARS-CoV-2 can remain in the body, in order to determine whether the 5-day treatment window is long enough. It’s possible that one way to minimize rebound infections is to extend the number of days people take Paxlovid, but Pfizer would have to conduct new studies to determine how many additional days would be needed, and how safe and effective the drug would be if that were the case. Pfizer is considering whether to conduct these studies in order to better understand if specific populations might need longer treatment with the medication.

Should you take Paxlovid again if you test positive after finishing the 5-day course?

According to the CDC’s recent health advisory on Paxlovid rebound, the agency does not recommend that people who have finished the five-day course restart taking the pills. Instead, the CDC advises people who test positive again to restart a five-day period of isolation and to wear masks for 10 days after rebound symptoms begin.

Ho says the CDC reached out to discuss best strategies for managing rebound prior to issuing the new guidance. The CDC does not specifically recommend that people on Paxlovid test themselves more regularly after finishing their treatment, but Ho advises that people follow current advice and run a rapid at-home test if any symptoms, however mild, reappear, regardless of whether they have taken Paxlovid or not.

How sick will I get if I test positive again after taking Paxlovid?

The CDC notes that most people who rebound experience mild symptoms the second time, and most will recover without the need for additional treatment.

If I test positive after taking Paxlovid, am I still contagious?

Yes. Anyone who tests positive on a rapid test—even after finishing a full course of the drug treatment—is again contagious and can pass the virus on to others. It’s important for people taking Paxlovid to be on the lookout for the possibility of a rebound infection and use rapid at-home antigen tests as often as they can, and confirm the diagnosis with a PCR test if possible, so they will know if they turn positive again.

In the 10 cases in Ho’s report, two people transmitted the virus to others in their household after they had relapsed. “The idea of somebody who goes virus-negative and then could become virus-positive again—that’s of great concern,” he says.

Given the possibility of a rebound, is it worth taking Paxlovid?

The emergency use authorization currently restricts doctors to prescribing the drug only to people at higher risk of developing serious disease, which includes older people and those with compromised immune systems. Otherwise healthy people, especially those who are vaccinated and boosted, may not benefit from the therapy, especially since people need to start taking it as close to when they first experience symptoms as possible in order for it to be effective. The rebound might just extend quarantine periods for them as well, since they might recover without the drug and not experience rebound infections in a shorter period of time.