This review describes key aspects of the development of the rVSVΔG-ZEBOV-GP Ebola vaccine and key activities which are continuing to further expand our knowledge of the product. Extensive partnerships and innovative approaches were used to address the various challenges encountered during this process. The rVSVΔG-ZEBOV-GP Ebola vaccine was initially approved by the European Medicines Agency and prequalified by the World Health Organization in November 2019. It was approved by the United States Food and Drug Administration in December 2019 and approved in five African countries within 90 days of prequalification. The development resulted in the first stockpile of a registered Ebola vaccine that is available to support outbreak response. This also provides insights into how the example of rVSVΔG-ZEBOV-GP can inform the development of vaccines for Sudan ebolavirus, Marburg virus, and other emerging epidemic diseases in terms of the types of approaches and data needed to support product registration, availability, and the use of a filovirus vaccine.

The rVSVΔG-ZEBOV-GP vaccine demonstrated efficacy in preventing Ebola virus (EBOV) disease in a ring vaccination clinical trial conducted during the 2014–2016 West Africa outbreak and is licensed by regulatory agencies, including the US FDA and the EMA. Here, we present two studies that evaluated the durability of immunogenicity and protection from an EBOV challenge up to ~12 months following vaccination with rVSVΔG-ZEBOV-GP in nonhuman primates (NHPs). Cynomolgus macaques were vaccinated with either one or two doses of rVSVΔG-ZEBOV-GP or a saline control and were challenged intramuscularly with EBOV at a target dose of 1000 pfu at ~4 months (Study 1) or ~8 or ~12 months (Study 2) after the last vaccination. All vaccinated animals developed robust ZEBOV-GP-specific IgG and neutralizing antibody titers, which were sustained until the last time point tested prior to the challenge. The majority of animals (88–93%) challenged with EBOV at ~4 or ~8 months post-vaccination survived, whereas the survival rate was lower (53%) in animals challenged ~12 months post-vaccination. These results demonstrate that both one-dose and two-dose regimens of the rVSVΔG-ZEBOV-GP vaccine induced durable ZEBOV-GP-specific antibody titers in NHPs and provided high levels of protection against a lethal EBOV challenge up to ~8 months post-vaccination. In this stringent challenge model, decreased protection was observed at ~12 months post-vaccination despite sustained antibody levels.

The recombinant vesicular stomatitis virus-Zaire Ebolavirus envelope glycoprotein vaccine (rVSVΔG-ZEBOV-GP) was highly effective against Ebola virus disease in a ring vaccination trial conducted during the 2014–2016 outbreak in Guinea and is licensed by regulatory agencies including US FDA, EMA, and prequalified by WHO. Vaccination studies in a nonhuman primate (NHP) model guided initial dose selection for clinical trial evaluation. We summarize two dose-ranging studies with the clinical-grade rVSVΔG-ZEBOV-GP vaccine candidate to assess the impact of dose level on immune responses and efficacy in an NHP Ebola virus (EBOV) challenge model. Forty-six cynomolgus macaques were vaccinated with a wide range of rVSVΔG-ZEBOV-GP doses and challenged 42 days later intramuscularly with 1000 pfu EBOV. Vaccination with rVSVΔG-ZEBOV-GP induced relatively high levels of EBOV-specific IgG and neutralizing antibodies, measured using the same validated assays as used in rVSVΔG-ZEBOV-GP clinical trials. Similar responses were observed across dose groups from 1 × 10⁸ to 1 × 10² pfu. A single vaccination conferred 98% protection from lethal intramuscular EBOV challenge across all dose groups. These results demonstrate that robust antibody titers are induced in NHPs across a wide range of rVSVΔG-ZEBOV-GP vaccine doses, correlating with high levels of protection against death from EBOV challenge.

Although adaptive trials have become popular, little research has been done to investigate the effect of missing data in adaptive trials. We consider three different types of missing data mechanisms—missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR)—and introduce the “mixture missing mechanism” (MMM) to an adaptive three-period crossover study that uses the “maximizing procedure.” These results are compared to a traditional nonadaptive equal allocation crossover study. Simulations suggest that certain missing data mechanisms can result in biased estimates. For equal allocation, the bias is uniform between treatments so treatment comparisons are unbiased. However, for the maximizing procedure, the bias is not uniform between treatments, so treatment comparisons are biased. For the MNAR and MMM mechanisms, unusually large bias occurs in the placebo group, leading to a substantial loss of power.