During the COVID-19 pandemic, antivirals in development pipelines for influenza, Ebola, and other viral threats were rapidly repurposed to treat patients with COVID. One of these antivirals, molnupiravir, works in an unusual way—rather than blocking the virus outright, it disrupts the virus's copying machinery, forcing errors to accumulate in the virus’s RNA until it loses the ability to replicate. Molnupiravir was approved after trials showed it could reduce COVID-19-related hospitalizations by about half. But its results across different clinical trials were puzzling: in some studies it only lowered viral loads slightly, while in others it cleared the virus more rapidly. A new study published in the Journal of Clinical Investigation by Dr. Shadi Esmaeili and colleagues in the Schiffer Lab at Fred Hutch may finally explain why, revealing a hidden flaw in how we measure the efficacy of antiviral drugs.
While Paxlovid (nirmatrelvir/ritonavir) is currently the preferred treatment for COVID-19, a study in cancer patients suggests that molnupiravir may be as effective as paxlovid against more recent SARS-CoV-2 variants. Additionally, patients in this study receiving Paxlovid were significantly more prone to drug-drug interactions and adverse events.
The Fred Hutch team therefore decided to take a closer look at molnupiravir. They used mathematical modeling to simulate three major molnupiravir clinical trials: MOVe-OUT, conducted when earlier COVID-19 variants like Delta were circulating among unvaccinated patients; and PLATCOV and PANORAMIC, both conducted during the Omicron wave among vaccinated individuals. By modeling how the virus replicates in the body, how the drug moves through the bloodstream, and how effectively the drug suppresses the virus at different concentrations, the researchers were able to closely replicate the viral load patterns observed in each trial.
If trial results can be replicated with mathematical models, researchers can then ask additional questions of the data that were not tested in patients, such as what would happen with a longer treatment course, or an increased dose of the drugs?
Several components were factored in to model how effectively molnupiravir eliminated SARS-CoV-2 in the clinical trials: 1) The Viral Immune Dynamic (VID) model tracked virus replication in the body over time, accounting for a limited number of susceptible cells being infected and a lag phase between infection and virus production. The VID also accounted for two phases of immune response: an early innate response and a delayed acquired immune response. 2) The Pharmacokinetic (PK) model tracked molnupiravir after oral dosing as it moved from the gut to the bloodstream and respiratory tract before being cleared. 3) The Pharmacodynamic (PD) model mapped drug concentration to antiviral effect. The researchers used lab experiments to calibrate a standard S-shaped dose-response relationship (called a Hill equation) to estimate what fraction of newly produced viruses get converted into mutated, non-infectious particles at any given drug concentration. Finally, the model accounted for differences in drug potency observed in lab experiments versus patients.
Lead author Dr. Esmaeili shares one of their surprising key findings: “Modeling multiple trials simultaneously gave us a clinically important insight: molnupiravir appears to be more potent in vivo against Omicron than against earlier variants such as Delta.” In addition, while most drugs perform worse in patients than in controlled lab conditions, molnupiravir turned out to be six to seven times more potent in patients than in lab experiments.
The standard PCR test used to measure viral levels cannot distinguish between mutated, non-infectious viral particles and intact, infectious ones. Strikingly, the authors demonstrated that since molnupiravir works by mutating viral RNA, PCR-based viral measurements underestimated the drug's true antiviral impact.
"One of the most critical decisions in clinical trial design is the choice of endpoint. We found that commonly used virologic endpoints may substantially underestimate the drug’s true antiviral activity. These findings underscore the importance of aligning virologic endpoints in trials with the biological mechanisms of the antiviral agents and highlight how cross-trial modeling can generate novel biological insight and improve how we design and interpret antiviral studies." Esmaeili concludes.
The study also echoes a lesson from Paxlovid: extending molnupiravir treatment from five to ten days would likely reduce the viral rebound that some patients experience after finishing the course, a finding with practical implications for how the drug might be prescribed in the future.
Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium Member Dr. Steven Polyak contributed to this research.
The spotlighted research was funded by The National Institutes of Health National Institute of Allergies and Infectious Diseases.
Esmaeili S, Owens K, de Leon UA, Standing JF, Lowe DM, Zhang S, Watson JA, Schilling WH, Wagoner J, Polyak SJ, Schiffer JT. 2025. Molnupiravir clinical trial simulation suggests that polymerase chain reaction underestimates antiviral potency against SARS-CoV-2. J Clin Invest. doi.org/10.1172/JCI192052.
Science Spotlight writer Kelly Mitchell is a postdoctoral fellow in the Paddison Lab at Fred Hutch Cancer Center. She utilizes live cell reporters and CRISPR screening to study how glioblastoma cancer cells resist chemotherapy and radiation treatment. She obtained her PhD in cellular biology from Albert Einstein College of Medicine.
For the Media