(CNN) The pink line on Erin Blakeney's first positive Covid-19 test was so light that she almost didn't believe it. But there was no denying the fever and sore throat that developed overnight, just a few days after she and her husband attended a large memorial service in late March. The couple wore KN95 masks, but many others in attendance had not, even as the service stretched past 90 minutes.
Blakeney, a 43-year-old researcher at the University of Washington's School of Nursing, is a breast cancer survivor. The Seattle resident says she doesn't meet any strict definition of being immunocompromised, which can raise someone's Covid-19 risk. Both Blakeney and her husband are fully vaccinated and boosted. But she didn't want to take chances, because she's taking medications to prevent a cancer recurrence and she lost a family member to Covid-19 in November 2020.
So when she tested positive April 3, she reached out to a doctor through a telemedicine visit and got a prescription for a Covid-19 antiviral medication, Paxlovid.
Blakeney knew that for Paxlovid to be effective, she had to take it early, within a few days of getting symptoms. That's what she and her husband did. They didn't get very sick and continued to improve until their five days of treatment were over.
She finished treatment on a Thursday and tested negative the next day. She thought she was done with Covid-19, at least for a few months.
But by Monday, she was starting to get congested again. When she woke up on Tuesday, April 12, she realized that she was really sick. Her husband was, too.
"We both took rapid tests, and we were both very positive again, and we were like 'Oh, my gosh, what just happened? We've never heard of this,' " Blakeney says.
Dr. Michael Charness, a neurologist who practices at Boston's Brigham and Women's Hospital and VA Boston Healthcare System, hadn't heard of it either until he saw it with his own eyes.
His patient was a 71-year-old man with asthma who started taking Paxlovid the first day he had symptoms because of his age and underlying medical condition.
The medication worked fast. By the second day, his patient's symptoms were almost completely gone. His patient finished the pills after five days, and was testing negative on home tests by day seven. On the ninth day after his first positive test, the man's symptoms -- including a runny nose, sore throat and wheezing -- returned. He tested positive once again.
"At the moment, I would say it's just a little mysterious," Charness said. "There are a lot of potential explanations for what's going on. They're all speculation, and it has to be much better studied."
Paxlovid, made by Pfizer, is a combination of two antiviral medications. The US Food and Drug Administration authorized it for emergency use in December, after a clinical trial found that a five-day course cut the risk of being hospitalized or dying by 88% in unvaccinated people at high risk of severe outcomes from a Covid-19 infection.
Many in health care hailed Paxlovid as a game-changer because it was so effective and convenient and had to the potential to keep people out of the hospital.
Last week, in his very first tweet, Charness detailed the rebound of his patient's Covid-19 infection after Paxlovid. His colleagues urged him to join social media after seeing other reports about Covid relapse in patients who were taking the medication.
His patient, who was fully vaccinated and boosted and in good overall health, was tested for Covid-19 multiple times throughout his treatment and then shortly after his symptoms returned.
Charness and his colleagues sequenced the genome of the man's coronavirus. It was the same throughout the course of the infection, so it wasn't a reinfection. And they couldn't find any mutations that had suddenly developed, so the virus hadn't suddenly changed to develop resistance to the medication. They also looked for 21 other respiratory pathogens that might be making the man sick and didn't find anything.
Charness says he doesn't want his case to scare anyone away from taking Paxlovid, which has been shown to be extremely effective at keeping mild cases of Covid from becoming more severe or even deadly.
The FDA noted these rebound cases in its own analysis of the data from the clinical trials. The phenomenon was found in a small percentage of cases, roughly 10 to 14 days after starting Paxlovid. It's not clear how common these may be. A clinical trial of the medication is still ongoing.
In a statement to CNN on the potential for rebound after taking the medication, Pfizer said that in the clinical trial, a small number of patients in both the group taking placebo pills and the group taking Paxlovid experienced higher viral loads 10 to 14 days after starting treatment compared with their viral loads at day 5. Because investigators noted this phenomenon in both groups, the company doesn't believe it is tied to the medication. They also saw no link between patients' viral loads and subsequent severe disease.
"We remain confident in its clinical effectiveness at preventing severe outcomes from Covid-19 in high-risk patients," the company said in a statement.
Charness said the possibility that patients could rebound regardless of treatment is interesting, and he hoped that Pfizer would release the data from its clinical trial so doctors could better understand what's happening in these cases.
Charness and his colleague Dr. Paul Sax, an infectious disease physician at Brigham and Women's Hospital, say that understanding why and how often the infection is coming back is important.
"Now that we know what happens, we have to include it in our messaging to people that it could happen, and my own opinion is that people should stay away from others while they're testing positive," Sax said.
The US Centers for Disease Control and Prevention says people who've had Covid-19 can return to their regular activities after five days as long as they've been fever-free for 24 hours and their symptoms are getting better. The agency does not recommend that a person wait to resume activities until they're negative on a rapid test, though some experts say rapid home tests are helpful for knowing when you can still pass the virus to others.
Sax says that if people test positive after Paxlovid, doctors and patients need to know whether they're actually contagious, and that's not clear right now.
He thinks it would also be helpful to know if people who rebound can have a second course of treatment, since they're outside the original window of symptoms.
Researchers and doctors are still learning how Paxlovid works. For months, the main worry for most people has been how to find some. Patients and doctors scoured the internet for pharmacies that stocked the pills.
These problems prompted the US government to launch a test-to-treat program, setting up one-stop locations like pharmacies where people could be tested for Covid-19, get an antiviral prescription at an in-store urgent care clinic and get that prescription filled in the same store, making the process much more convenient.
This week, the Biden administration has announced that supplies are no longer constrained, and it is taking new steps to increase use of the medication.
Many people who have rebound infection say their symptoms tend to be mild the second time around.
But Blakeney was far sicker when her symptoms returned. She had a high fever, a rapid heart rate and a "really awful" cough. She says her breathing was very wheezy.
"I ended up going to the emergency room," she said. Her case stumped the ER doctors, who wondered how this could happen after Paxlovid.
"They were like, 'we've never heard of this,' " Blakeney said.
After running tests to make sure she didn't have blood clots or pneumonia, they prescribed an inhaled medication called albuterol to help her lungs and an antibiotic to treat what they thought must be a secondary bacterial infection.
Blakeney did get better after taking those medications. She's now mostly recovered, though her cough has lingered.
She says the experience has made her think about how we're navigating the pandemic.
"If we're living in a world where we don't have mask mandates, should we be having people test to exit isolation? I would think yes," she said. "Or should we be doing something kind of like the food safety guides they have in restaurants, where it's like 'this place has good airflow and is following best practices to create as safe of an environment as possible'? So I think it raises a lot of questions.
"It was a surprising experience. It was not how we expected it to go."