Randomized controlled trial of molnupiravir SARS-CoV-2 viral and antibody response in at-risk adult outpatients

Abstract

Viral clearance, antibody response and the mutagenic effect of molnupiravir has not been elucidated in at-risk populations. Non-hospitalised participants within 5 days of SARS-CoV-2 symptoms randomised to receive molnupiravir (n = 253) or Usual Care (n = 324) were recruited to study viral and antibody dynamics and the effect of molnupiravir on viral whole genome sequence from 1437 viral genomes. Molnupiravir accelerates viral load decline, but virus is detectable by Day 5 in most cases. At Day 14 (9 days post-treatment), molnupiravir is associated with significantly higher viral persistence and significantly lower anti-SARS-CoV-2 spike antibody titres compared to Usual Care. Serial sequencing reveals increased mutagenesis with molnupiravir treatment. Persistence of detectable viral RNA at Day 14 in the molnupiravir group is associated with higher transition mutations following treatment cessation. Viral viability at Day 14 is similar in both groups with post-molnupiravir treated samples cultured up to 9 days post cessation of treatment. The current 5-day molnupiravir course is too short. Longer courses should be tested to reduce the risk of potentially transmissible molnupiravir-mutated variants being generated. Trial registration: ISRCTN30448031.

Fig. 1. Disposition of PANORAMIC participants approached and provided virology samples.

< LLOQ, below lower limit of quantification.

Fig. 2. Baseline viral load and spike antibody relationships with each other, CRP, demographics and co-morbidities.

a Baseline viral load increase with age. b Baseline viral load decrease with time since symptom onset. c Baseline viral load decrease with increased baseline spike antibody (S-Ab). d Baseline viral load increase with increase in CRP measured in capillary blood. e Geometric mean viral load ±95%CI with time compared with a t test showing baseline viral load higher in males and in patients with kidney disease. f Geometric mean S-Ab ±95%CI with time compared with a t test showing baseline S-Ab lower in males and participants who were not fully vaccinated. LLOQ is lower limit of quantification which for viral load was 109 cp/mL. For (a–d) the slope was compared with zero using the f statistic, for figures (e) and (f) and t test was used to compare geometric mean values. *P < 0.05, **P < 0.01, ***P < 0.001. Source data are provided as a Source Data file.

Fig. 3. Viral and spike antibody dynamics in molnupiravir (purple) and Usual Care (grey) arms.

a Median viral load ±95%CI with time, below limit of quantification (BLOQ) measures substituted with LOQ/2 for molnupiravir (n = 253) and Usual Care (n = 324) compared with a Mann–Whitney test. b Viral load with time showing multivariable linear viral dynamic model predictions for a typical individual (female, median age, S-Ab and time since symptom onset) receiving molnupiravir (purple line) or Usual Care (black line). For molnupiravir a piecewise increase then decrease in elimination rate was estimated with a change point at the end of treatment compared with a likelihood ratio test. c Simulation from the viral dynamics model for 1000 subject demographic combinations sampled from the data. Line shows the 50th percentile of the probability of viral load being <109 cp/mL on the last day of treatment for 5 to 14 days treatment. Error bars are 95% prediction intervals. d Geometric mean spike antibody ±95%CI with time compared with a t test. e Spike antibody with time showing multivariable linear antibody dynamic model predictions for a typical individual (female, median age) receiving molnupiravir (purple line) or Usual Care (black line) compared with a likelihood ratio test. f Model-derived individual predictions of spike antibody doubling time versus viral load area under the curve (AUC) from day 0 to 14. *P < 0.05, **P < 0.01, ***P < 0.001. Source data are provided as a Source Data file.

Fig. 4. Mutations occurring over time in viral sequences.

a Number of transition and transversion mutations in molnupiravir-treated (purple) and Usual Care (grey) participants. Transition mutations (G to A, A to G, C to T and T to C) are shown as solid lines, transversions (G to C, C to G, A to T, C to A, G to T, T to G and T to A) are shown as dotted lines; standard deviation denoted by error bars. b Shannon entropy over time in participants receiving molnupiravir or Usual Care. Participants in each group with viral loads above the LLOQ by day 14 (Molnupiravir Persistent; purple and Usual Care Persistent; orange) and participants with <LLOQ viral loads by Day 14 (Molnupiravir Resolved; grey or Usual Care Resolved; yellow) are shown; standard deviation denoted by error bars. b Shannon entropy over time in participants receiving molnupiravir or Usual Care. Participants in each group with viral loads above the LLOQ by day 14 (Molnupiravir Persistent; purple and Usual Care Persistent; orange) and participants with <LLOQ viral loads by Day 14 (Molnupiravir Resolved; grey or Usual Care Resolved; yellow) are shown. c Numbers of sequenced samples at each time point. Source data are provided as a Source Data file.

Fig. 5. Maximum likelihood phylogenetic trees of consensus sequences from sequenced samples.

a sequences (n = 627) from molnupiravir treated participants. b sequences (i = 809) from participants receiving Usual Care. Baseline samples (Day 1) are shown in purple, samples collected between days two and five (corresponding to the duration of molnupiravir treatment) are shown in grey, samples collected between Days 6 and 14 (after treatment has finished in the molnupiravir arm) are shown in red. The scale bar is shown.

Fig. 6. Viral mutations in molnupiravir and Usual Care groups.

a The number of mutations (mean= 553 and standard deviation=391), in samples taken at treatment-end (Day 5) is significantly higher (Mann–Whitney test) in molnupiravir-treated participants with detectable virus at Day 14 (orange, n = 100 participants) compared with those who cleared virus between Days 5–14 (red, n = 87 participants). Box-plot midline is the median, the upper and lower limits the 75th and 25th percentile, the whiskers 1.5 times the interquartile range. Data beyond that are represented as points. There is no difference between the numbers of mutations in the comparable Usual Care groups (mean = 45 and standard deviation=31.2, n orange = 94 participants, n red = 158 participants). b Viral load trajectories (mean and standard deviation) for molnupiravir-treated and Usual Care participants with any know spike neutralisation-escape missense mutation occurring at any time point post baseline at consensus level (allele frequency (AF) > 50%) in red, below consensus level (1–50%) in purple or with no mutations (grey). Participants (441) were required to have viral load data for baseline, Day 5 and Day 14 and sequence data for all positive viral loads. The table shows the total number of participants with consensus, below consensus and zero known neutralisation escape mutations for molnupiravir and Usual Care groups, c Consensus level (>50%) amino acid substitutions occurring in Day 14 sequences from molnupiravir (purple) and Usual Care (grey) participants. Mutations occurring in more than one participant’s sample are shown above the rest. Schematic of the SARS-CoV-2 genome is shown along the bottom. ****P < 0.0001. Source data are provided as a Source Data file.

Fig. 7. Culture rate inferred from microscopy observation of cytopathogenic effects combined with a positive result by lateral flow immunochromatography of culture media, from nasopharyngeal swabs inoculated on to Calu-3 cells and incubated for 7 days.

a Overall culture rate split by baseline, on treatment and post treatment for each group compared with a two-sided Fisher’s Exact Test. b Distribution of culturable samples by day of sampling versus viral load. Overall totals were 169 culture positive samples out of 1018 screened. No significant associations between treatment group or any of the tested covariates were found with probability of culture positivity in samples taken after Day 6. ns: P > 0.05. Source data are provided as a Source Data file.