Paxlovid Rebound in COVID-19: Where We Stand on Drug Regimens

This discussion was recorded on August 18, 2022. This transcript has been edited for clarity.

Robert D. Glatter, MD: Welcome. I'm Dr Robert Glatter, medical advisor for Medscape Emergency Medicine. Today, we'll be discussing the topic of Paxlovid rebound, which has been front and center in the news. Both Dr Fauci and President Biden experienced the phenomenon.

Joining us to discuss the topic is Dr Paul Auwaerter, professor of medicine and clinical director, Division of Infectious Diseases at Johns Hopkins University School of Medicine.

Welcome, Paul.

Paul G. Auwaerter, MD: I'm glad to be here. This is a topic many patients and physicians have questions about.

Glatter: Absolutely. I want to start off by defining for the audience what Paxlovid rebound is and how you would explain that.

Auwaerter: Right. I would divide it into two possible areas. One would be that people improve and then have some symptoms of recurrence, whether it's upper respiratory, lower respiratory, or perhaps fever. Does the syndrome recur?

The other would be, for those who especially have at-home antigen tests in hand, where we know that measures about 10⁵ or higher logs of the virus, does your test become negative and turn positive? Or does your test still stay positive (for those who haven't checked after taking a course of Paxlovid), usually 2-3 days after you stop, for up to a week later?

Glatter: For people who persistently test positive, there has been a question of whether they are infectious. If they're having this rebound and, let's say they're mildly symptomatic, do you advise that people do rapid antigen testing serially day to day in order to prevent exposing family members or coworkers?

Auwaerter: I don't know if everyone has tests available for daily assessment. I'll tell you, I had COVID-19 the same day as Anthony Fauci so I had many antigen tests. I was antigen positive up to day 11, did not even take Paxlovid, and then I ran out of tests, so I don't know if I was positive after that.

I did try to limit my exposure, wore a mask, and worked from home during that time. Of course, the recommendations are still the same for 10 days. I would say, certainly, for patients and myself, one of the reasons I didn't take Paxlovid was I had a commitment that I didn't want to miss in case I had Paxlovid rebound about 12 days after my illness.

Glatter: I'm glad you're doing better. That's good news.

Auwaerter: I'm doing fine. To answer the other part of your question, there have been reports of transmission within households, but not widespread.

Glatter: That speaks to the CDC's guidelines, which have been revised recently. People are concerned about exposing others during this period of rebound, and I think that's an important part to really dissect. That's where the controversy exists. Should we really be reentering society at that point?

Auwaerter: Considering vulnerable populations, you'd especially want to be very careful if people are living in your household who might be at risk. Much like the reason why the CDC practically reduced from 14 to 10 days and just downward, it's the practical aspect. How adherent are people? How many people are following through on that?

Paxlovid's Mechanism of Action

Glatter: Let me back up. Can you explain what Paxlovid is? What are the two drugs in it and how does it work? What would possibly be a mechanism that could be behind the rebound? I think that's the million-dollar question.

Auwaerter: To break it up, again, into two pieces, Paxlovid is nirmatrelvir combined with ritonavir. Nirmatrelvir is a specific protease inhibitor. Those of you who might have treated hepatitis C or had patients with HIV [who] are on these protease inhibitors. They inhibit a virally driven protein that cleaves a polyprotein into specific components so you can make new viruses and assemble.

Nirmatrelvir works, but its levels are not sufficient at the current dosing scheme, so it's boosted with ritonavir, which really has no specific activity against SARS-CoV-2 but raises the nirmatrelvir level. The unfortunate aspect is that ritonavir is a suicide inhibitor of CYP3A4, so it has really a tremendous list of drug interactions that you'd want to be very careful about in patients before considering a prescription.

Glatter: The drug-drug interactions. The people with chronic medical issues — diabetes, hypertension, obesity, people who are on blood thinners for atrial fibrillation or post stroke — are the people we're concerned about but need the drug the most. That's really our concern as clinicians.

Auwaerter: Yeah, that's right. In fact, the group that might be at highest risk are those with solid organ transplants, who may be on calcineurin inhibitors, such as cyclosporine or tacrolimus. That's where the drug interactions, even if you hold the drug, can lead to great toxicity. You can have syndromes like posterior reversible encephalopathy. There's a very strong recommendation not even to think about prescribing it in those patients.

Glatter: Do we know… Or what are your thoughts about the proposed mechanisms? There's been some early research in this, and we're talking about whether this is immune related. Is it circulating virus? Is it some other type of mechanism we're just not aware of at this point?

Auwaerter: I think the jury's out. Clearly, we know, for example, in myself, where I'm not taking any medication, there is some evidence of rebound. I didn't have any symptoms. I just had viral carriage. I had continued improvement during that time.

One theory is that people now are taking the medicine very early. We're doing antigen tests very early, not later. Perhaps we're suppressing the virus so early, there's enough transcripts and so on and so forth that something builds up, you withdraw the drug, and then boom, you get the rebound.

The other is that, perhaps, you're impairing some of the earlier immune responses, and it depends on what people's immunological history is. Do they have the ancestral virus? Did they have alpha? Have they been immunized?

Now, with BA.5, especially, varying enough, there's enough immune escape that maybe the body hasn't enough chance to really adapt so by the time you withdraw the drug, there's not enough to keep the virus at bay. Those are just some of the ideas, and it may be multifactorial components.

A substantial number of people do rebound. Careful studies suggest that a few percent rebound even in natural infection, but it might be higher. Some preprints have recently suggested it could be as high as 20%-25% with careful analysis.

Glatter: That preprint you're referring to came out in early August. It was very eye-opening, to be honest with you, to hear that up to one third of patients are rebounding. These are people without taking Paxlovid or any other antiviral, for that matter. Obviously, we'll bring molnupiravir to the conversation because that certainly doesn't have the efficacy that Paxlovid has, especially in the early studies, but it is available and something that we look to in those who can't take Paxlovid.

Auwaerter: Right. This may have been happening all through the pandemic. We always knew you could have persistent virus by molecular analysis and polymerase chain reaction (PCR). We knew early on that, for immunocompetent patients, really by day 7 or 8, it was very hard to find culturable virus. We're still grappling with quality studies to see what's happening with BA.5, for example. Is that different? There's a sense that it might persist a bit longer, and that may be a component here as well. 

Testing Positive After Testing Negative: Is It Paxlovid Rebound?

Glatter: The other issue is related to testing. Could you get a false negative early on after your initial infection then, thereafter, test again and become positive? That certainly is a limitation of the rapid antigen testing.

Another issue that I've seen is that maybe someone developed a secondary bacterial infection, such as sinusitis, bronchitis, or pneumonia, and this masqueraded as a rebound, and truly, it's not a rebound. It's a secondary infection. I think that's another caveat that we have to think about in this discussion.

Auwaerter: Absolutely. I think that may be giving a bit of a false impression that there's even higher rates of clinical rebound after they take this because, often, patients do feel substantially better quicker, taking the drug. Then there's this rebound, which may be that some antigenic drivers are just stimulating the immune response as the virus makes some proteins, which aren't being assembled into real virus, but stimulate a bit of a rebound clinical syndrome.

Glatter: From what I'm seeing, younger patients who don't really fit the category of high risk — age 50-65 years or have comorbidities — are getting prescribed Paxlovid in urgent cares and even in emergency departments (EDs). This type of prescribing is problematic, and it speaks to indications for the drug that never really were developed for. It's important to really stress that there is a specific group that should be the target for these antivirals in general. Would agree with that?

Auwaerter: I do, absolutely. The emergency use authorization (EUA), of course, says risks, and everyone looks at the CDC risks. That's become a little more liberal over time. I handle it by having discussions with the patient. Some will say, "Well, look, I'm 54. I should get Paxlovid." I say, "Well, gosh, you're doubly boosted. Chances of serious illness are really low. There's a rebound issue."

If someone is 75, they have major health problems, they're immunosuppressed, I say, "Look, even though you may be immunized, I would feel much better and I strongly recommend you take the Paxlovid. We do know that it can help prevent serious illness, even though we don't have the kinds of studies yet to really suggest how well it works in these populations. It's extra insurance that you're going to do fine, and we'll deal with the rebound if it occurs," and then give isolation recommendations at that point.

Glatter: In terms of isolation, you bring up a good point. What is your recommendation and how would that jive with the CDC's recommendation in terms of isolation with a case of rebound?

Auwaerter: The clock resets, in a way, to be cautious. It's as if you have a second illness. In fact, the President Biden and Dr Fauci both got second courses that you could say fall a little bit outside the EUA in those cases, unless you think they're a new infection.

I usually advise that people do mask and isolate during those times of rebound because I think there is infectious virus. If they have two serial negatives, I'm like, "Good, you can go about, wear a mask for that 10-day period at that stage." I tend to rely on the antigen on the other side, too, to try to get people out of jail sooner.

Glatter: In terms of any other aspects of viral load rebound, Dr Ho, who's an HIV researcher you know well, had documented a personal case, where he had been at a COVID-19 conference in Paris and he ended up testing positive. He became convinced that the rebound phenomenon was due to Paxlovid and it wasn't just inherent as part of the actual disease itself.

I think there were 10 patients in what he published. He hinted that he really was suspicious that the drug itself was more the cause of the rebound. But again, the data are not necessarily pointing to that, and I think there is controversy.

Auwaerter: I agree. I think there is the impression and the focus that people will anchor to what they've last done, so Paxlovid ends up taking the brunt of that. I think there is an impression and maybe it's just the focus, and we need to do more study to see if this is really a fair attribution or not. Or is it part of the milieu that people have now with some degree of neutralizing antibodies from immunization, infection, and so on that offers part of these dynamics here? I think it's probably real to some degree. The question is, how much?

Glatter: We're concerned, and patients have expressed to me that they don't want an intense rebound effect, and if they take Paxlovid, if they're going to be at risk for that. They'd rather just tough it out and take anti-inflammatories, acetaminophen, fluids, and supportive care. Some of the literature is certainly reflecting that the intensity of the rebound could be an issue in patients, especially older patients. Should that dissuade people from considering Paxlovid or another antiviral?

Auwaerter: I still say no on that, especially for people at high risk, with the idea that we still have 50 or 60 patients in our hospital with COVID-19 at any given time, and it's exactly these patients. Although the intense rebound may be disturbing, perhaps because you feel like you're improved, and suddenly, a day or two later feeling much worse. Again, we have not hospitalized anybody that I'm aware of with Paxlovid rebound. I can't say it can't happen. Certainly, on the other side, we have people still coming to the hospital with this.

Glatter: I think the goal is to keep people out of the hospital, obviously and reduce complications. We know that the drugs are working. I think that's, really, the ultimate outcome.

Auwaerter: Absolutely.

Glatter: The studies are ongoing with Pfizer. I think 10 days of Paxlovid vs 5 is really the crux of the issue because, with persistent circulating virus, especially in high-risk people, is there a downside to having a 10-day course? Are we looking at resistance? Is that a concern with 10 days vs 5 days?

Auwaerter: Well, in certain populations, it could be. You have more opportunity and pressure to select for viral resistance. Ten days may be the right answer. It may be combination therapy, which we tend to use in many viral infections. This is, hopefully, a shorter and ultimately, self-limiting infection even without treatment. Ten days may be the right amount here to try to avoid some of these more infection control–oriented problems. I've never had someone with really terribly serious illness have to land back in the hospital with the rebound. Another aspect I do tell patients is that even if you do rebound not to worry — because many of them are worried.

Glatter: There's also, I wanted to point out, a study that Pfizer ran [on] postexposure prophylaxis and whether there was any value to giving it to patients who had an exposure to someone who was positive. The data, from what I saw, didn't really pan out and it didn't have any role. It was interesting that they ran the study and that, certainly, there was no indication for giving it. Can I get your thoughts on that?

Current Therapies for Pre- and Postexposure Prophylaxis

Auwaerter: We know for influenza that oseltamivir, zanamivir, and baloxivir all work if you give it early enough after exposure to a degree. I think it's also a numbers game, and especially in a heavily immunized population, perhaps the numbers weren't big enough to really show an impact. I think it makes biologic plausibility to use it that way. Of course, with the EUA, we're not supposed to prescribe it that way currently.

If I had someone that was a close contact, who may have chronic lymphocytic leukemia (CLL), they can't mount antibodies, they're on Evusheld, for example, for preexposure prophylaxis, I would have little compunction if the drug were available to give it for postexposure prophylaxis in a really high-risk patient. Some of the most difficult patients we've had in hospital are those who have B-cell deficiencies that have prolonged viral carriage and illness, often for weeks and months at a time.

Glatter: Some of the Pfizer studies and other studies have pointed to the fact that hospitalized patients with high-risk features, as you're describing, should be candidates for antivirals, especially Paxlovid.

Auwaerter: I think that's exactly right. Those are groups where I also use monoclonal antibodies if they're available. We're currently limiting them to severely immunodeficient patients, such as transplant patients, just given the limited availability of the monoclonal since federal funding is no longer present, and doses, at least, are waning in state supplies.

Glatter: Assuming monoclonals are available and you have a higher-risk patient, would you prescribe remdesivir outpatient for a 3-day course as an intravenous infusion, assuming there were other contraindications, especially to Paxlovid?

Auwaerter: Yes. In the state of Maryland, we have a tiered scarcity level for bebtelovimab to really target people who, for example, are on rituximab, might have CLL, or that cannot be immunized — maybe they had a severe reaction to the mRNA vaccine — they get bebtelovimab. For others who might be at high risk, like transplant patients, 3 days of remdesivir is what we're using in that group.

Glatter: There are other options besides Paxlovid, and I think we look at that as the only solution right now. Certainly, as an outpatient, it's attractive, but that's important to hit on.

Auwaerter: The remdesivir just presents a logistic challenge. No one wants to come back on 3 serial days for an IV infusion. Molnupiravir is a drug that we mentioned; it's the other oral antiviral. The MOVe-OUT trial had a 30% reduction in hospitalization or death at 28 days in an unimmunized population. Its impact, I think, in an immunized population is likely less. It's clearly second tier behind Paxlovid, monoclonals, and even remdesivir.

A recent secondary analysis in the Annals of Internal Medicine did suggest that there is less of a respiratory illness progression, and even people who did get into the hospital got out of the hospital earlier in the active molnupiravir arm. The actual trial had nine deaths in the placebo group and one in the molnupiravir group. If push comes to shove, for a high-risk patient, I do prescribe it because I think it does have some benefit. I don't completely discount it and say, "Oh, it's not worth it for patients." For the 50-year-old who is doubly boosted and so on, I don't think you're going to get much juice out of squeezing that one.

Glatter: Some people have talked about the mutation aspect of molnupiravir and that being a concern that patients have heard about. Is that something that, in your mind, should preclude physicians from prescribing it?

Auwaerter: We've used nucleoside analogs for quite a while. This is a 5-day course. I don't worry so much about the genotoxicity. Women are supposed to practice birth control for up to 4 days after they stop that drug. For men, it's 3 months. If there's family planning going on, I might talk about an alternative, or else that's the advice.

Other Therapeutics on the Horizon

Glatter: Are any new antivirals or COVID-19 therapeutics in the pipeline at this time that could even supplant Paxlovid as you see going forward?

Auwaerter: There has been an analog of colchicine, which has some very promising impact on the hyperinflammatory phase and ICU issues. That has an EUA in front of the FDA right now. There are other oral antivirals, and certainly, the most requested one is a non–ritonavir protease inhibitor, so we don't have to be as concerned with drug interactions.

Glatter: We do know, based on a recent article that Eric Topol put on Twitter, that medicines such as fluvoxamine, the antidepressant, ivermectin, and metformin really have no role. Metformin, possibly, in a secondary analysis, had some impact on primary endpoints.

Overall, we're learning that there are drugs that work and there are drugs that don't work. Looking at the evidence and practicing evidence-based medicine is very important.

Auwaerter: Yes, I don't know if it will persuade the ivermectin believers. Last winter, when sotrovimab was not available, there were no medications in the outpatient arena and we thought about fluvoxamine based on the Brazil TOGETHER trial. In a healthy population more like here in North America, it really didn't seem to have any impact.

The metformin is intriguing, but I don't think anyone is going to be running to prescribe that as an alternative here. Again, the repurposed drugs are great. We're learning more and more that that's a tough hill to climb to find success when you're trying to use an existing drug to impact a novel disease.

Final Thoughts

Glatter: Do you have one or two pearls for our audience to take away from this discussion?

Auwaerter: Although the rebound phenomenon, which may be due to infection itself or related to a drug like Paxlovid, is important, that if you have a high-risk patient, please still strongly consider antiviral therapy impacts. I wouldn't be dissuaded from that.

The second aspect is to know your drug interactions. We've seen people made ill by Paxlovid who land in the hospital. There are a number of drug interaction checkers. The one I like is from the University of Liverpool. They do a great job and have had a long history of HIV drug interaction checks with ritonavir. They have very clinically sound recommendations in their drug interaction checker.

Glatter: Thanks for decoding and deciphering what Paxlovid rebound is and the phenomenon of rebound, in general. I really appreciate your time, Dr Auwaerter. Thank you, again.

Auwaerter: Thank you for having me. We'll learn more because we're still trying to understand the mechanism and I think that will help inform more about what's going on and what to do.

Glatter: Absolutely. Thank you, again.

Robert D. Glatter, MD, is assistant professor of emergency medicine at Lenox Hill Hospital in New York City and at Zucker School of Medicine at Hofstra/Northwell in Hempstead, New York. He is an editorial advisor and hosts the Hot Topics in EM series on Medscape. He is also a medical contributor for Forbes.

Paul G. Auwaerter, MD, is a professor of medicine at Johns Hopkins University School of Medicine, and clinical director in the division of infectious diseases at Johns Hopkins Hospital in Baltimore, Maryland. He is an editorial advisor and hosts the Auwaerter on Infectious Diseases series on Medscape.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube