Abstract
Without clinical efficacy data, vaccine protective effect may be extrapolated from animals to humans using an immunologic marker that correlates with protection in animals. This immunobridging approach was used for the two-dose Ebola vaccine regimen Ad26.ZEBOV, MVA-BN-Filo. Ebola virus (EBOV) glycoprotein binding antibody data obtained from 764 vaccinated healthy adults in five clinical studies (NCT02416453, NCT02564523, NCT02509494, NCT02543567, NCT02543268) were used to calculate mean predicted survival probability (with preplanned 95% confidence interval [CI]). We used a logistic regression model based on EBOV glycoprotein binding antibody responses in vaccinated non-human primates (NHPs) and NHP survival after EBOV challenge. While the protective effect of the vaccine regimen in humans can be inferred in this fashion, the extrapolated survival probability cannot be directly translated into vaccine efficacy. The primary immunobridging analysis evaluated the lower limit of the CI against predefined success criterion of 20% and passed with mean predicted survival probability of 53.4% (95% CI: 36.7-67.4).