Federal health regulators on Tuesday issued a warning that COVID-19 patients who have taken the antiviral treatment Paxlovid may experience a rebound and test positive again two to eight days after initial recovery.
The warning comes more than a month after droves of patients began swapping accounts on social media of COVID rebounds after taking Paxlovid.
The alert from the Centers for Disease Control and Prevention said it continues to recommend Paxlovid for patients at high risk for serious complications from infection. It also said that people with a recurrence of COVID-19 symptoms, or a new positive test after having tested negative, should isolate again for at least five days. It added that people should wear a mask for a total of 10 days after rebound symptoms start.
The CDC warning cited recent reports by researchers documenting some Paxlovid rebounds, including among patients who have been vaccinated and boosted.
The CDC warning comes as two small but provocative new studies from different teams of researchers in Boston and New York suggest such relapses may not be that unusual. The scientists also found that, in some rebound patients, viral levels were likely high enough to be contagious.
Additionally, one study found that two patients who relapsed inadvertently infected family members. In one instance, a 67-year-old man with no symptoms six days after completing Paxlovid infected a 6-month-old family member. The transmission took place after the window of time suggested by the CDC for isolating.
“This is not rare, otherwise why are we seeing these clusters?” said Dr. Michael Charness, chief of staff at the VA Boston Healthcare System and co-lead author of the study, which analyzed COVID infections in 10 patients, ages 31 to 71 who were fully vaccinated and had received at least one booster shot. The study was posted online Monday and has not been peer-reviewed.
Paxlovid is an at-home treatment prescribed at the first sign of infection to patients at high risk of serious COVID complications. The treatment consists of a total of 30 pills — three pills taken twice a day for five days.
When Paxlovid manufacturer Pfizer received emergency use authorization for the drug in December from federal regulators, its data indicated about 2 percent of patients in its trial experienced a rebound, as did roughly 1.5 percent of those who received a placebo.
The Charness study doesn’t directly refute the Pfizer data, but does suggests that rebounds are more common. Patients in that study experienced relapse symptoms beginning between three and eight days after completing Paxlovid, and symptoms lasted between three and ten days.
Antigen tests performed while patients were relapsing indicated they stayed positive for a median of six days, and until as long as day 18 after their initial, pre-Paxlovid positive test.
For a comparison, the Charness group looked at a separate group of COVID patients who had not taken Paxlovid: nearly 1,000 personnel in the National Basketball Association. None had a relapse of COVID. They have not published that data.
“We cite the NBA data to say this is clearly different from what had been observed” by Pfizer, said Dr. David Ho, director of the Aaron Diamond AIDS Research Center at Columbia University and a coauthor of the Charness and the NBA studies.
A Pfizer spokesman said the company continues to monitor data from its ongoing studies of Paxlovid, as well as reports from doctors and patients of rebounds. He said all the data are “consistent with our observations” from the company’s drug trial.
The new CDC warning notes that a COVID relapse with a brief return of symptoms “may be part of the natural history of SARS-CoV-2 (the virus that causes COVID-19) infection in some persons, independent of treatment with Paxlovid and regardless of vaccination status.”
Federal data show more than 668,000 courses of Paxlovid had been prescribed as of May 14. Doctors say that despite the rebounds, it is an effective medicine that has kept at-risk people out of the hospital.
A second study from a team of Boston researchers analyzed seven patients who had rebounded after taking Paxlovid and found levels of live virus in three of them for up to nine days. One of the patient’s test sample showed live virus for 11 days after finishing Paxlovid.
“This strongly increased the suspicion that they are contagious,” said Dr. Mark Siedner, an infectious disease clinician and researcher at Massachusetts General Hospital and coauthor of the study, which has not been peer-reviewed.
“It’s not just that they rebound. It’s that they rebound and it seems like they start from the beginning, their virus goes way up,” Siedner said. “It’s a really unique phenomenon.”
Siedner’s team found no evidence that the virus developed a resistance to Paxlovid.
Siedner and other researchers say the rebound phenomenon raises urgent questions about whether patients should be treated with a longer course of Paxlovid or perhaps another medication.
The CDC alert does not definitively answer that question.
“There is currently no evidence that additional treatment is needed with Paxlovid or other anti-SARS-CoV-2 therapies in cases where COVID-19 rebound is suspected,” it said.
The rebounds with such high levels of live virus has prompted Siedner to question if there is something about Paxlovid that might be contributing to the phenomenon.
“It makes us wonder, are we not using the drug properly or long enough, or is this something inherently about Paxlovid that doesn’t allow the immune system to kick in?” he said.
Siedner’s team, which includes researchers from Brigham and Women’s Hospital as well as the Broad Institute and Ragon Institute of MGH, MIT and Harvard, is launching a new study that hopes to answer some of these questions.
They will test the immune system of people who have rebounded to see if the immune response from those who received Paxlovid is different from those who did not.
Dr. Kathryn Stephenson, an assistant professor at Harvard Medical School and an infectious disease physician at Beth Israel Deaconess Medical Center, is also conducting a Paxlovid study, monitoring patients who just got started on the antiviral with COVID testing for two to three weeks to detect rebound information on symptoms.
She said that hers and other small studies of a few dozen people are helpful but that much larger, rigorous studies are urgently needed to understand and address rebounds.
“I think it is Pfizer’s responsibility to produce and share this data rapidly — it’s their drug [that received emergency use authorization,]” Stephenson said.
“It’s not fair that individual clinicians and researchers are now trying to catch up and collect this data ourselves.”