COVID-19 has become less of an urgent threat than it was in 2019 largely because of vaccines and growing immunity from natural infections, but antiviral treatments have also changed the course of the disease. The most popular of these is nirmatrelvir-ritonavir, sold under the brand name Paxlovid, which the U.S. Centers for Disease Control and Prevention (CDC) recommends for older people and anyone over age 12 who is at higher risk of COVID-19 complications. But people taking the drug have reported incomplete recovery, or testing positive again after testing negative once they finished the five-day course of the oral medication.
In a study published in the Annals of Internal Medicine, researchers led by Dr. Mark Siedner, at Harvard Medical School and Massachusetts General Hospital, investigated the phenomenon, and report that around 20% of people taking Paxlovid could experience rebound infections. The researchers also cultured the virus from these rebound patients, and confirmed live virus, which suggests that patients are still infectious and therefore can spread the virus to others.
Paxlovid rebound has been a highly debated topic in the COVID-19 medical community, since the drug’s maker, Pfizer, reported in its studies submitted to the U.S. Food and Drug Administration (FDA) that it occurred in about 2.3% of people. Since Paxlovid has been on the market, other studies have documented even higher rebound rates, around 14%. Unlike those studies, which mostly looked retrospectively at whether people developed symptoms again or tested positive again after testing negative, Siedner’s study set out to specifically investigate the rebound effect. The researchers took samples from 142 people who tested positive for COVID-19 and received prescriptions for Paxlovid and took the pills three times a week for two weeks and then weekly until their virus was detectable. The scientists not only tested for the virus, but if they found any, they also cultured it to determine if it could potentially cause infections.
They found that 20% of people rebounded after completing the five-day treatment course of Paxlovid, and that those who experienced rebound continued to harbor live virus in their noses for up to 14 days.
“Our study was small and needs to be verified, but based on our data, we need to balance the fact that Paxlovid is a very important drug and should be used in higher risk people, with the risk of rebound in 20% of people,” says Siedner.
Siedner also explored the discrepancy between Pfizer’s initial estimates of rebound, and the higher estimates from other studies. He found that the key to quantifying the rate of rebound is testing people more frequently, to capture changing levels of the virus. Pfizer’s study was not designed to study rebound and its scientists only tested participants three times: at days five, 10 and 15 after the first positive COVID-19 test. When Siedner and this team looked only at samples they collected on the same days that Pfizer’s researchers did, they found a similar 2.4% rebound rate as Pfizer had. Looking at samples from only those three days missed 80% of the rebound cases, he says. “When you don’t sample enough, you miss the rebound, and our study was able to fill in the gaps,” he says. “If you’re not studying it closely enough, you may not find it.”
Read more: Why Not Everyone Should Take Paxlovid
Understanding the rate, along with the mechanics of rebound, is important since rebounding virus remains contagious, as Siedner’s group found. They did not study whether people experiencing rebound actually spread the virus to others, but given that they found live virus in those nasal samples, it’s reasonable to assume that rebounding people are still contagious. Currently the CDC recommends that people isolate for five days after they first test positive—Siedner says that many of those who rebound after taking Paxlovid “are still shedding live virus for about three times that length of time. People who rebound in my opinion should have prolonged isolation periods”
Dr. David Ho, a professor of microbiology and immunology at Columbia University was among the first to report higher rates of rebound. Based on on-going studies from his lab, he says an eight-day Paxlovid treatment, instead of a five-day course, could significantly reduce the amount of rebound. Without any immune response, the virus’s half life in the body—the time it takes for the amount of virus to drop by half—is about 24 hours. Adding Paxlovid can drop the amount of virus by 32-fold, but the timing of the drug treatment needs to align with peak levels of the virus. Since that can be hard to achieve, taking the drug for a longer period of time could help. “We think that if you treat for eight days you can pretty much wipe out the rebound phenomenon,” says Ho.
More From TIME
Siedner’s study supports the longer course of treatment. People in the trial who began taking Paxlovid early, within a day or two of testing positive, were more likely to rebound than those starting the treatment a few days later. But he is reluctant to advise delaying treatment, since it’s challenging to find the right window when there is enough virus present. And extending the treatment course would require a change in the drug’s label, and the FDA would likely require an additional study documenting the benefits of the longer treatment. While Pfizer is currently studying such a longer course in people with weakened immune systems, it’s not clear if the company would undertake similar research in other groups.
“We are continuing to monitor the data, but believe the return of elevated, detected nasal viral RNA—also known as viral rebound or COVID-19 rebound—is not uniquely associated with any specific treatment,” a Pfizer spokesperson tells TIME. “We remain very confident in Paxlovid’s clinical effectiveness at preventing severe outcomes from COVID-19 in patients at increased risk.”
Given the current recommendations for Paxlovid, there isn’t an effective way to avoid rebound, but Siedner says it’s possible to manage it and reduce the risk of spreading COVID-19 because of it. People on the drug could test themselves again with an at-home antigen test five days after they test negative. If they remain negative, they likely have not rebounded, but if they are positive, they need to isolate again until they test negative.
“We really want to reinforce the fact that clinical trials established the fact that people at very high risk of COVID-19 complications can benefit from taking Paxlovid,” says Siedner. “Our data in no way counters that evidence. But people need to be aware that Paxlovid rebound is common, and understand that they need to isolate properly if they do rebound.”