A couple of weeks ago, I tested positive for COVID. I went on the antiviral drug Paxlovid right away. I felt significantly better after being on it for three or four days. When I took a home test, I was still positive, but only had a very faint line.
Then, after finishing the five-day course of the drug, I started feeling sicker again. My fever returned. I started testing strongly positive.
What happened to me has happened to lots of other people—including Anthony Fauci—and has been dubbed “Paxlovid rebound.” The rebounds don’t seem to be landing people in the hospital or leading to death, but they’re not fun. And more importantly, they have public health implications, because people who thought they weren’t contagious anymore actually might be. What we don’t really know yet is what’s driving these rebounds and how common they truly are.
Lizzie O’Leary: In December of 2021, the Food and Drug Administration granted Pfizer an emergency use authorization for Paxlovid. The US government placed an order for millions of courses of treatment. What was the public reaction to the release of the drug?
Rachel Gutman-Wei: It was called a monster breakthrough, miraculous, the biggest advance in the pandemic since the vaccines.
And what garnered the headlines and praise were some pretty amazing numbers from Pfizer’s clinical trial, about how well the drug worked to keep people out of the hospital and keep them from dying. What did the data say about Paxlovid?
There were two main sections of the clinical trial. Number one was in unvaccinated people who had some kind of risk factor for severe COVID—that’s the high risk group. The high risk portion in unvaccinated people found an 89 percent reduction in risk of hospitalization and death, which is great. It’s really, really important that that population can be protected because they’re at the highest risk of going to the hospital, having severe complications, possibly dying from COVID. And there are still 300 people or so dying a day from COVID and presumably some of them are not vaccinated and have these sort of risk factors. So protecting them is super important.
But there was another part of the trial, what’s called the standard risk trial. What happened in that group?
It was a different story in the standard risk trial. So that’s people who are vaccinated, but it excluded anybody who had received a vaccine in the last 12 months. It also included those people who were unvaccinated, but don’t have risk factors. They didn’t really find any significant effects of the drug on preventing hospitalization, ICU admissions, or death. They found trends pointing in the right way, but they weren’t statistically significant. They also didn’t find any benefit in terms of having symptoms for a shorter amount of time, second alleviation, any of that. So it’s not as clear from the standard risk data, what exactly Paxlovid has to offer for vaccinated people.
How much data exists on a vaccinated person or a vaccinated and boosted person taking Paxlovid and what kind of efficacy the drug has for them?
There was an Israeli study about a month and a half ago that looked at both vaccinated and unvaccinated people, and it was, crucially, done during the omicron era. The Pfzier trial was conducted pretty much completely in the delta era. So some of the difference, the daylight between the amazing results in the high risk trial and the not as great results in the standard risk trial could also have something to do with the variant. But this Israeli study basically found that people over 65, whether they’re vaccinated or not, got a ton of benefit from Paxlovid and people under 65 got virtually none, regardless of their vaccination status.
Now that we are in this scenario where people can get it—and get it for free, because the US government has purchased a lot—who is supposed to be prescribed Paxlovid right now? I think that is a little bit confusing to people.
Yeah, and it’s confusing to doctors too. The emergency use authorization from the FDA is fairly broad. It’s basically: If you’re over 12 and you have COVID and it’s been five days or fewer since your symptoms started and you have a risk factor, then you can have it. And, the risk factor-worth language has been pretty broad since the beginning of COVID for lots of good reasons. But at the same time, you could have a perfectly healthy 15-year-old with mild asthma, be eligible for Paxlovid.
How widespread is its use?
Since it was authorized in December and as of June 26th, which is the most recent HHS data that’s available, more than 1.8 million Americans have taken it. So it’s a lot.
While Paxlovid clearly has benefits, there are still a lot of unknowns about it, including why it can cause a weird side effect known as Paxlovid mouth. What is that?
People have described it as bitter, sour, metallic. One person described it to me like if you mixed grapefruit juice and soap together, somebody said it tastes like burnt grapefruit. It doesn’t sound good. Pfizer said that that was happening in five or 6 percent of people, but almost everybody I spoke to for the article about Paxlovid mouth said no way, it’s happening to way, way, way more people. I spoke to a couple of dozen people just in my reporting on Paxlovid and every single one had Paxlovid mouth.
Then there is the issue of rebound. I think it might be helpful to define what we’re talking about when we say rebound—how would you describe it?
I would say that rebound is when you have taken Paxlovid, you’ve finished your course, your symptoms have abated, you test negative—and then within a few days you get a reversal. You test positive again, your symptoms come back.
Rebound makes me wonder whether there’s any data on if the treatment is extending the course of the infectious period. It feels like a very thorny question, not just for individuals, but for public health, right?
Absolutely. And the scenario in which people think they’re in the clear and then go out and infect other people is one of the reasons why it’s really important to find out exactly how widespread Paxlovid rebound is. I mean, if it’s really only happening in a few percent of people, then maybe it is okay to tell people, “If you finish your course, it’s been 10 days since the start of your symptoms, you’re free.” But if 10 or 20 percent of people are getting rebound, we might want to issue different guidance.
Pfizer has said that it’s seen a rebound rate of about 2 percent of patients who take Paxlovid, but as the drug is more widely adopted, patient experiences seem like they might tell a different story. What’s causing that mismatch?
Part of the problem with the Pfizer data is that all we have is a number. All we have is 2 percent. We don’t really know where they got that number from. They haven’t released the data behind it, at least not that I have seen. The Mayo Clinic study that found that it was closer to 1 percent was in 483 people, which is maybe not enough to really sus out the incidents when this is being given to close to two million people at this point. People I spoke to said that this would be the kind of thing that it would be good to study within a hospital system. So to just track everybody that a particular system prescribes Paxlovid to, and then see what happens. And there are certainly people who are doing that, but there just haven’t been results yet.
There was something in your story that really stuck out to me. And I feel like it highlighted why this is so important because you wrote that an individual patient would never know really if Paxlovid worked for them, because you can’t know how sick you would’ve gotten, if you didn’t take it. But, there’s this idea that if the drug doesn’t do that much for vaccinated people, then spending the country’s money on it might be a waste of our dwindling resources. How do we know if we are spending our money and energy well, as a country, in the arsenal of our tools to fight COVID?
This is a question that goes way beyond Paxlovid. America’s tendency to focus on one thing and run with it. We’re in an era of “Paxlovid is the thing now.” We’ve tried vaccines. We’ve given up on masks. We’ve given up on vaccine mandates. We’ve vaccinated some people. Let’s just throw Paxlovid at everybody, put all our eggs in that basket.
And as far as exactly how much money we should be spending on it, or whether it’s a waste, I think that’s more of a public health slash political judgment call. It’s really an open question whether that funding could be better spent doing further vaccine outreach, which has still been shown to connect with some holdouts and get them vaccination, or whether that could be better spent upgrading ventilation in schools and public buildings.
What else don’t we know about Paxlovid? And what do you want to know? When I went on it, I was struck by the fact that there’s basically no data on breastfeeding, and I’m at the end of breastfeeding my child. And my doctor saying, well, you probably shouldn’t, but there’s no real data. I was shocked by how little information there is.
I would say that’s a huge, outstanding question. One of the biggest ones for me is whether Paxlovid has any effect on long COVID. So there are some folks who have long COVID who have been pushing for research about whether Paxlovid can actually alleviate long COVID if you already have it. The other question is, does taking Paxlovid while you have COVID prevent you from getting long COVID afterward? That’s a really important question. Especially as a lot of the country has sort of adopted this posture of “Everyone’s going to get it. Consequences are what they are.” The consequences are a little bit different if a significant portion of the country is going to have these long term effects that impede their life.
But of course, all of this takes a long time to study, especially when it comes to tracking long COVID, and what kind of treatments might be helpful.
These are effects that last months and apparently years, so you would really want to follow up with people over months and years. So we’re not going to have answers on that for a while. Of you get COVID today and you’re somebody who has maybe a risk factor, but you’re young, you’re not super worried, nobody can tell you whether Paxlovid would decrease your chances of getting long COVID.
What do you think happens now? Is there going to be a souped-up Paxlovid? Will there be some other wonder drug?
As far as the future of Paxlovid itself, one of the interesting things is that Pfizer and the NIH are collaborating on a study of whether extending the dose time would help prevent some of those rebounds. One of the theories behind why people might be rebounding, if it is in fact related to Paxlovid, is that they’re just not taking the drug for long enough. If it turns out that an eight day course means that virtually nobody is rebounding, then that changes the calculus again. That’s one possible future. Another possible future is nothing changes about the drug, doctors and patients continue having these conversations. And some doctors decide to do it. Some doctors decide not to, depending on various risk factors. And it becomes a tool that is available to individuals, some of whom use it and some of whom don’t.
As far as broader implications, Paxlovid is something you give to people after they have already gotten COVID. And there’s a whole world of things that we as a country, we as a global community, could be doing to prevent that very first step, which is preventing people from getting COVID. So things like smarter approaches to masking, smarter and perhaps more enforced approaches to ventilation upgrades, testing requirements, improved access to testing, improved speed of testing, improved accuracy of testing. All these things could help stop chains of transmission and prevent people from even needing Paxlovid. And it would still be great to have it available if somebody did in fact get infected. But certainly it seems like it could do more good as part of a wider ecosystem of care.