When Berkeley resident Myriam Misrach tested positive for the coronavirus last month, she started taking the antiviral pill Paxlovid the same day. During the five-day course of treatment, her cough and shortness of breath usually subsided, but a few days after she took the last pill, her symptoms returned.
48 hours later, she also had a fever, headache, nausea, a runny nose and lost her sense of taste, she said. And she tested positive for the virus again — despite testing negative and feeling much better a few days earlier.
“I had everything in the book,” said Misrach, 66, who has been vaccinated and boosted. “It was not a mild case at all.”
Misrach continued to test positive for two weeks after that and is still coughing today, although the other symptoms have disappeared. What was even more puzzling, she said, was that her husband had just taken Paxlovid and it “worked beautifully” for him — he started feeling better almost immediately and stayed that way, though he, too, tested positive after initially testing negative.
“I don’t blame Paxlovid, but I think they should study it more,” she said.
With the number of Americans taking the Pfizer drug soaring, many people report a similar “rebound” after taking the drug — including some vaccine scientists and doctors who have documented their experiences on Twitter. In addition to a recurrence of symptoms, rebound also means that someone who thought they had recovered may still be contagious and will have to isolate for extra days.
All known cases of Paxlovid viral rebound appear to have resolved without patients requiring hospitalization, say doctors prescribing the drug and researchers studying the problem. They overwhelmingly agree that this gives them no break from prescribing the antiviral medication, which in clinical trials reduced the risk of hospitalization and death from COVID by nearly 90%. They say that if someone qualifies for Paxlovid, the patient should still get it, despite the potential to experience rebound symptoms, because it delivers on its promise by indeed keeping people out of the hospital.
Rebound, also called relapse, is not uncommon in infectious diseases. Doctors often see it in patients who have taken antibiotics or antivirals, where the infection returns after treatment is completed because the virus or pathogen has not been completely cleared, said Dr. Prasanna Jagannathan, a Stanford immunologist and infectious disease physician.
Paxlovid’s rebound phenomenon is an example of what happens when a new drug — probably the most controlled drug in recent history, second only to COVID vaccines — begins to be widely used in the real world and deliver results that may not have been observed in clinical trials at such high levels. This doesn’t mean the drug has failed, scientists and doctors noted, but rather, it needs further study and may require adjustment of dosage or duration of use.
It’s not clear why rebound occurs, or how often it occurs in the real world. In Pfizer clinical trials, it occurred in 2% of people taking Paxlovid. Many doctors prescribing the antiviral say they hear about rebound from patients anecdotally, and that it’s more common than in studies. But that may be partly due to reporting bias, with the people who experienced a rebound reporting it earlier than those who didn’t have the problem.
“We’ve all heard anecdotes from patients we’ve cared for who experience this, so it’s clearly a phenomenon,” said Jagannathan, who has prescribed Paxlovid to 25 to 30 patients and observed rebound in two of them. “Nobody knows yet what that real number is.”
Pfizer and the US Food and Drug Administration are monitoring rebound cases for further investigation. Providers and patients can report cases to the respective Pfizer and FDA adverse event reporting systems.
Researchers are looking at a few possible explanations for viral rebound.
A small study, not yet peer-reviewed, suggests the problem is probably not drug resistance due to viral mutation or a problem with a patient’s immune response. Instead, patients may not have been adequately exposed to Paxlovid. This could mean that instead of the five-day course currently approved by the FDA, people may need to take the drug for longer or at a different dosage. The study, posted to a pre-print site this week, looked at three vaccinated and boosted adults who took Paxlovid, including one who experienced rebound. That person was infected with the BA.2 ommicron subvariant.
Patients and health care providers can report cases of suspected Paxlovid rebound to Pfizer and the U.S. Food and Drug Administration, who are monitoring the phenomenon for further investigation.
To report it to Pfizer, visit Pfizer’s COVID-19 Treatment Adverse Event Reporting website and submit a form online.
To report it to the FDA, go to FDA MedWatch and submit a form online or by fax at 1-800-332-0178. Call 1-800-332-1088 for questions.
“Our hypothesis or best guess at this point is that we think there is not enough drug exposure to get rid of the virus,” said the study’s lead author, Dr. Aaron Carlin of UC San Diego, who studies emerging and recurrent viral infections and how they interact with the immune system. “There will probably be studies to see if people need 10 days instead of five days to try and prevent that rebound.”
Another small study, originally published in late April and updated last week by the VA Boston Health System, also suggests that the reason for relapses is not that the virus mutated after patients took Paxlovid. The authors said further research is needed to determine the cause of relapse.
Pfizer’s Paxlovid studies were conducted in vaccinated and unvaccinated people when delta and earlier variants were circulating. Now it is people infected with omicron and omicron subvariants who are taking the drug, including many who have been vaccinated. So it’s possible that their immune system may respond a little differently to the drug, which may help explain the rebound. It is also possible that omicron and its sub-variants lead to a longer period of viral shedding than delta, so that people now need a course longer than five days.
The UC San Diego study analyzed several coronavirus variants and their susceptibility to Paxlovid, and found no significant differences in how they responded to the drug. But there is some evidence to suggest that the neutralizing antibody response in vaccinated people is lower to omicron than to delta, “so there may be something about omicron and the immune system contributing to this (rebound), but we don’t understand that yet,” said Carlin.
If it turns out that the virus does become resistant to Paxlovid in the future, it may help to combine it with other antiviral drugs. Treating HIV with a single drug leads to drug resistance almost immediately, but treating HIV with three drugs does not, Carlin said.
“It’s a warning, but I don’t think it’s a reason people shouldn’t take the drug,” Carlin said. “It is still very effective. We just need to understand if it can be used in a way to make it even better to avoid the rebound.”
Catherine Ho (she/her) is a staff writer for the San Francisco Chronicle. Email: cho@sfchronicle.com Twitter: @Cat_Ho