A Viroimmunologic Model to Characterize the Antiviral Effect of Molnupiravir in Outpatients Infected With SARS-CoV-2: Implication for Treatment Duration

Abstract

Background: The antiviral efficacy of molnupiravir against SARS-CoV-2 is controversial. Here, we develop a model integrating viral and immune dynamics to characterize the mechanism of action of molnupiravir in vivo and its impact on viral dynamics during and after treatment.

Methods: We analyzed data from the PANORAMIC trial, where 577 outpatients were randomized shortly after symptom onset to receive usual care or molnupiravir for 5 days, with viral and immunologic data collected within 2 weeks. We developed a mathematical model that characterized virus-host interaction, accounting for the impact of molnupiravir on viral replication and mutagenesis. The model was used to explore the impact of longer treatment duration.

Results: Molnupiravir reduced RNA replication with an efficacy that reached 93% at the end of a 5-day treatment. This effect was mediated through 2 pathways: 1 that increased transition mutation frequency and 1 that directly inhibited viral production. Accordingly, 5-day treatment shortened the median time to clearance of RNA and infectious virus by approximately 2 days. Ten-day treatment could reduce the time to RNA clearance by 5 days and the occurrence of viral rebounds. Longer treatment durations might be needed for postexposure prophylaxis.

Conclusions: Our model suggests that molnupiravir acts primarily on viral replication, and not specifically on viral infectivity. Longer administration of molnupiravir may reduce the rebound rate, shortening the time to viral clearance.

Keywords: SARS-CoV-2; molnupiravir; mutagenesis; treatment duration; viral clearance.

Figure 1.

Nasopharyngeal viral load (left), spike antibodies (middle), and proportion of transition mutations (right) in the PANORAMIC trial. Upper, usual care alone (untreated, n = 324); lower, usual care with molnupiravir, 800 mg, twice a day (treated, n = 253); circles, observed data; empty circles, data below the limit of quantification; bold solid lines, mean; bold whiskers, 90% CI.

Figure 2.

Model for molnupiravir effect on SARS-CoV-2 replication. Abbreviation: NHC, N4-hydroxycytidine. Created in BioRender. Guedj, J. (2025) https://BioRender.com/k72z109

Figure 3.

Individual fits for nasopharyngeal viral load (left), spike antibodies (middle), and proportion of transition mutation (right) from 8 PANORAMIC participants. Gray circles, observed data; white circles, data below the limit of quantification; solid curves, model predictions (gray, untreated; blue, treated); black horizontal dashed lines, limit of quantification; red vertical dashed lines, symptom onset; shaded area, molnupiravir treatment period.

Figure 4.

Predicted impact of treatment duration on viral load, spike antibodies, and proportion of transition mutations. Horizontal dashed line, limit of quantification (2.0 log₁₀ copies/mL). *Time since treatment initiation. **Time after treatment completion. ***Calculated from treatment initiation to sustained negative result on polymerase chain reaction. Results are displayed as the median profile/value and its 90% CI for 100 replicates of 250 individuals. Abbreviation: AUC, area under the curve.

Figure 5.

Molnupiravir inhibits RNA viral production via 2 mechanisms of action. Left, viral production from infected cells predicted by the model; right, impact on viral load; solid gray, control; dotted cyan, molnupiravir direct viral inhibition (E_M$\to$V); dashed cyan, molnupiravir viral inhibition via transition mutation only (E_M$\to$P); solid cyan, combined effect; horizontal dashed line, limit of quantification (2.0 log₁₀ copies/mL). Results are displayed as the median profile/value and its 90% CI for 100 replicates of 250 individuals.

Figure 6.

Predicted probability of positive culture result during treatment with molnupiravir. Horizontal dashed line, 5% threshold. *Time since treatment initiation. **Time since treatment completion. Results are displayed as the median profile/value and its 90% CI for 100 replicates of 250 individuals, assuming once-daily sampling.