Fig. 6: Forest plot of SUPERNOVA observed efficacy versus PROVENT ToP model predicted efficacy based on SUPERNOVA data.

*Observed efficacy was defined based on planned intervention as 100(1 – relative risk) (%) of sipavibart versus comparator, where relative risk and two-sided 95% CI were evaluated with Poisson regression with robust variance, which includes treatment and randomisation stratification factors as covariates and adjusted follow-up time. Randomisation stratification factors were COVID-19 vaccination status within 6 months before randomisation (Yes or No), SARS-CoV-2 infection within 6 months before randomisation (Yes or No), and tixagevimab–cilgavimab use within 12 months before randomisation (Yes or No). ^†When events are not present in at least one arm, an exact approach is followed and 97.5% two-sided CI presented¹⁰. ^‡The PROVENT ToP model estimates and covariance parameter estimates were used to predict efficacy (%) and two-sided 95% CI based on actual intervention through evaluating average nAb ID₅₀ titres over 90 days or 180 days post any dose according to the timepoint. For the overall endpoint, average prevalence-adjusted nAb ID₅₀ titres were considered after mapping GISAID prevalence data to at-risk participants for variants with ≥5% prevalence. For matched non-F456L and XBB+ subvariant endpoints a similar approach was followed with exclusion of F456X variants (which are otherwise assumed an IC₅₀ of 1000), and inclusion of only XBB subvariants based on hedgehog variants that included a XBB Pango lineage, respectively. Daily serum mAb concentrations estimated from population pharmacokinetics were used to predict nAb ID₅₀ titres in at-risk participants based on IC₅₀ of 3.8, 83.1, and 1000 ng/mL for BA.2.86, JN.1 subvariant analyses and variants with F456L mutations, respectively. The comparator group includes participants who received tixagevimab–cilgavimab followed by placebo and those who received two doses of placebo. CI confidence interval, COVID-19 coronavirus disease 2019, IC₅₀ 50% inhibitory concentration, ID₅₀ 50% inhibitory dilution, GISAID Global Initiative on Sharing All Influenza Data, mAb monoclonal antibody, nAb neutralising antibody, N is the number of participants in the full analysis set without a positive RT-PCR test for SARS-CoV-2 at baseline by planned intervention, n is the number of participants with SARS-CoV-2 events for each endpoint by planned intervention, RT-PCR, reverse transcription-polymerase chain reaction, SARS-CoV-2 severe acute respiratory syndrome coronavirus 2, ToP threshold of protection.