15 nov 2022 Episode 294 Update on EBOLA vaccines

Tue, 11/15/2022 - 20:21

Episode 294:  EBOLA Vaccines


In this follow-up episode, I try to summarize the most convincing data on different EBOLA vaccine platforms. Afterwards, I present a personal summary/interpretation of progress in EBOLA (vaccine) research.   


  1. Overview


A recent overview of vaccines that are either approved or in human trial is given by Ashish Sharma in Curr Opinion Pharmacol (Ep 294-1)




Mayinga refers to DRC strain from 1976.  The Kikwit strain from DRC 1995 is closely related (0.03 % divergence)

The West African strain from Guinee 2002-2008 is 3 % divergent.

The Makona strain (Guinee 2014) is 3 % divergent both from Mayinga-Kikwit 1976-1996 and from West-Africa 2002-2008. 







More additional practical information by Oyewale Tomori in Curr Opinion Immunol 2020 (Ep 293-5)




The two vaccines approved in the West are

  • ERVEBO from MSD: single dose replication competent recombinant VSV (vesiculo-stomatitis virus) expressing the Zaire EBOV (ZEBOV) glycoprotein (GP). It has shown efficacy in the real world as well (see below)
  • Zabdeno and Mvabea from J&J and Nordic Bavaria: heterologous prime with replication incompetent human Adeno26 expressing Zaire EBOV GP (Ad26-ZEBOV)  + boost with replication competent Modified Vaccina Ankara, expressing GP from Zaire EBOV, Sudan EBOV, Marburg and Tai Forest virus. (MVA-BN-Filo)    


Approved in China a replication incompetent human Adeno 5, expressing EBOV Zaire (ZEBOV) GP

Approved in Russia GamEvav Combi: a prime boost with VSV-ZEBOV and Ad5-ZEBOV


As can be seen, several other vaccines are in early clinical trial, including

  • Additional VSV-based, Hu Adeno 5, but also Chimp Adeno 3-based
  • DNA vaccine by Inovio
  • GP protein based by Novavax  (the information in Table 2 is not correct, as it refers to SARS-CoV-2, but there is a genuine EBOLA vaccine see Ep 293-14)
  • Polypeptide based (EpiVacEBOLA) is produced and apparently approved in Russia: no publications   



Most vaccines focus on the Zaire strain (either Mayinga 1976 or Makona 2014), but

  • Bivalent ChAd3 Zaire and Sudan vaccine 
  • Trivalent sVSVN4CT1 EBOVGP1, containing glycoprotein (GP), from Zaire, Sudan and Marburg
  • The MVA booster from Nordic Bavaria with glycoprotein from Zaire EBOV, Sudan EBOV, Marburg Virus and nucleoprotein from Tai Forest virus


  1. Specific cases:  I will focus on real world data (only for ERBEVO), protection in non-human primates (NHP) or phase 2 studies in humans.   


2.1. ERBEVO (VSV from Merck) has published real world data


Ep 294-2: EBOLA Ca suffit study in Guinea (Lancet 2017) = ring vaccination, open-label, cluster-randomised trial:  immediate and delayed vaccination in contacts and contacts of contacts were compared: seemingly 100 % protection

No cases of Ebola virus disease occurred 10 days or more after randomisation among randomly assigned contacts

and contacts of contacts vaccinated in immediate clusters versus 16 cases among all eligible

individuals in delayed clusters.


Ep 294-3: Rupiani Emerg Infect Dis 2022:  Retrospective study during the 2018-2020 outbreak in DRC:

  •  25% of patients vaccinated before symptom onset died compared with 63% of unvaccinated patients.





  • Vaccinated patients re­ported fewer EVD-associated symptoms, had reduced time to clearance of viral load, and had reduced length of stay at the Ebola Treatment Center





Reduction in deaths was not affected by timing of vaccination before or after EVD exposure →  support use of preexposure and postexposure


Ep 294-4: Joseph Inungu AIMS Public Health 2019:  Description of the 2018 epidemic in DRC:


The WHO reported that the rVSV-ZEBOV-GP vaccine candidate (Merck vaccine) showed a 97.5% efficacy rate in the trial in the DRC. Of more than 90,000 people vaccinated, only 71 developed Ebola. Fifty-six of those people presented symptoms fewer than 10 days after being vaccinated, suggesting the vaccine had not yet had time to fully protect them. It takes about 10 days for the immune protection to develop after vaccination.


Ep 294-5:  Thomas Monath in NPJ Vaccine describes the entire history of this vaccine until 2020.


2.2. Data on Zabdeno-Mvabea: combo monovalent Ad26 ZEBOV and MVA-BN-Filo (= quadrivalent MVA with GP from EBOV Zaire, EBOV Sudan, Marburg and NP Tai Forest viruses)


Many phase 1 and 2 studies, showing safety and immunogenicity but no solid real world data


Ep 294-6: Iain D. Milligany JAMA 2017: Phase 1 in US adults


All vaccine recipients had specific IgG detectable 21 days postboost and at 8-month follow-up.





At 7 days postboost, at least 86%of vaccine recipients showed Ebola-specific T-cell responses (but those were not very high).


Ep 294-7: Zacchaeus Anywain PLoS Medicine 2022  Phase II trial in African children and adolescents

All vaccinees displayed anti-EBOV GP antibodies after the 2-dose regimen, with higher responses in the 56-day prime-boost interval groups




No serious adverse events, but pyrexia higher in children than adolescents. 


Ep 294-8: In view of the difficulty of true human protection studies, Ramon Roozendaal (NPJ Vaccine 2020) proposes “immunobridging”: The association between EBOV neutralizing antibodies, glycoprotein (GP)-binding antibodies, and GP-reactive T cells and survival in non-human primates (NHP) was assessed by logistic regression analysis:


Binding antibodies against the EBOV surface GP were identified as the immune parameter with the strongest correlation to survival post EBOV challenge, and used to infer the predicted protective effect of the vaccine in humans using published data from phase I studies.  


Study design in NHP


Analysis of vaccine-elicited parameters in relation to survival of NHP




Receiver Operator Curve (ROC) in d shows that T cell responses (ELISpot) have a lower area under the curve (AUC) than antibody responses (either ELISA or neutralization assays).  Because ELISA is more robust it is selected as a surrogate marker of survival. 

The result of ELISA in human phase 1 and 2 studies can now be used to predict probability of survival after EBOV infection, as shown in Fig 5 (below).



Nevertheless, this hypothesis will need to be confirmed in the real world, as EBOV evolves differently (more rapidly) in NHP and challenge in this model was standardized (4 weeks after the MVA boost), while infection in humans will happen a diverse time frame after vaccination.



2.3. GamEvac Combi approved in Russia:  The vaccine consists of live-attenuated recombinant vesicular stomatitis virus (VSV) and adenovirus serotype-5 (Ad5) expressing Ebola envelope glycoprotein ZEBOV Makona strain.

I found only 1 paper on human phase I/II.  No data on NHP. 


Ep 294-9: I. V. Dolzhikova Human Vaccines and Immunotherapeutics 2017


  • Antigen-specific IgG geometric mean titer at day 42 was 3,277 (95 % confidence interval 2,401–4,473) in volunteers immunized at full dose.
  • Neutralizing antibodies were detected in 93.1 % of volunteers at full dose, with geometric mean titer 20.
  • Antigen-specific response in peripheral blood mononuclear cells was also detected in 100 % of participants, as well as in CD4+  and CD8 + T cells in 82.8 % and 58.6 % of participants vaccinated at full dose, respectively



These titers seem rather low, but it is difficult to judge, since assays are not standardized.


2.4. Human Adeno-5 expressing ZEBOV GP Makona strain: approved in China.  Several phase 1 and 2 studies in China and 1 in Sierra Leone.  No NHP studies identified


Ep 294-10:  Feng-Cai Zhu  in Lancet 2017: Phase II in Sierra Leone, comparing a single low (8 X 1010 viral particles) and high (1.6 X 1011 particles): apparently very strong and lasting antibody response.




The Ad5 vector is a common human serotype.  Therefore many subjects have pre-existing antibodies to the vector, which weaken the immune response to the vaccine antigen.  This has clearly been shown in candidate HIV and SARS-CoV-2 vaccines, which therefore have been abandoned.



 2.5. Chimp Adeno-3 vaccines


Ep 294-11: Daphne Stanley Nat Med Oct 2014: Combination Chimp Ad 3 and MVA, expressing GP Mayinga strain


Human adenovirus type 5 vectors (rAd5) encoding ebolavirus glycoprotein (GP) generate protective immunity against acute lethal Zaire ebolavirus (EBOV) challenge in macaques, but fail to protect animals immune to Ad5, suggesting natural Ad5 exposure may limit vaccine efficacy in humans.

Here we show that a chimpanzee-derived replication-defective adenovirus (ChAd) vaccine also rapidly induced uniform protection against acute lethal EBOV challenge in macaques.

Because protection waned over several months, we boosted ChAd3 with modified vaccinia Ankara (MVA) and generated, for the first time, durable protection against lethal EBOV challenge




Ep 294-12: Julie E. Ledgerwood  NEJM 2017  Phase I in humans with Chimp-Adeno 3 encoding the glycoprotein from Zaire and Sudan species.


At the 2×1011 particle-unit dose, glycoprotein Zaire–specific antibody responses were in the range reported to be associated with vaccine-induced protective immunity in NHP challenge studies and sustained to week 48.





Phase 2 trial are announced, but I could not find results.


Ep 294-13: Tapia in Lancet Infect Dis 2016 shows human phase I of a heterologous prime with Chimp Adeno3-GP Zaire EBOV and a boost with MVA-BN-Filo





A clear-cut boost effect only in the heterologous combination, but also no follow-up, as far as I can see.


2.6 DNA vaccine from Inovio:  combination of 3 optimized plasmids with Mayinga, Makona and the 2002 West-African strain = TRIVALENT DNA vaccine


Ep 294-14:   Ami Patel J Infect Dis 2019 Protective Efficacy and Long-Term Immunogenicity in Cynomolgus Macaques by Ebola Virus Glycoprotein Synthetic trivalent DNA Vaccines.


  • Multiple-injection regimens of the EBOV-GP DNA vaccine, delivered by intramuscular administration followed by electroporation, were 100% protective against lethal EBOV-Makona challenge.
  • Two injections of a simple, more tolerable, and dose-sparing intradermal administration followed by electroporation generated strong immunogenicity and was 100% protective against lethal challenge.


EBOV-GP DNA vaccination induced long-term immune responses in macaques that were detectable for at least 1 year after final vaccination and generated a strong recall response after the final boost.


Intramuscular + electroporation




Intradermal + electroporation




2.7. Protein vaccine from Novavax


Ep 294-15: Louis Fries J Infect Dis 2019: Phase I/II with Mayinga Ebola Virus Glycoprotein Nanoparticle Vaccine With Matrix-M as adjuvant


  • Two doses of EBOV GP with adjuvant showed a rapid increase in anti-EBOV GP immunoglobulin G titers with peak titers observed on Day 35 representing 498- to 754-fold increases from baseline;




  • Serum EBOV-neutralizing and binding antibodies using wildtype Zaire EBOV (ZEBOV) or pseudovirion assays were 3- to 9-fold higher among recipients of 2-dose EBOV GP with adjuvant, compared with placebo on Day 35, which persisted through 1 year.



2.7. mRNA vaccine


Ep 294-16: Meyer JID 2018: Two mRNA encoding GP Mayinga, with different signal peptide (wild type and IgG kappa) were formulated in Lipo-Nano-Particles and induced 100 % protective immunity in guinea pigs







2.8. Trivalent VSV with ZEBOV, Sudan virus and Marburg virus= tri-val N4CT1 pan-FiloGP1


Ep 294-17: Demetrius Matassov J Virol 2017 : Full protection against ZEBOV; SUDV and Marburg in macaques





















Ep 294-18: David Clarke Lancet 2020: Immunogenicity of rVSVN4CT1-EBOV GP1 = monovalent








ELISA titers are high, but neutralizing Ab and T cell ELISPOT seem rather weak.


The risk of antibody-dependent enhancement (ADE)?


Ep 294-19: Kuzmina Cell Rep 2021: Antibody-Dependent Enhancement of Ebola Virus

Infection by Human Antibodies Isolated from Survivors




Filovirus-specific monoclonal antibodies from human survivors present at low concentrations are capable of enhancement of infection in monocytes or monocytic cell lines via the immunoglobulin Fc Receptor


The enhancement can be caused by antibodies of various epitope specificities, neutralizing capacities, and subclasses.


ADE counteracts antibody-mediated protection and facilitates infection, suggesting that low levels of antibodies in

humans may facilitate virus spread







Ep 294-20: Fuyurama PLoS Negl Dis 2020: A complement component C1q-mediated mechanism of antibody-dependent enhancement of Ebola virus infection


Besides the common Fc receptor (FcR)-mediated mechanism of antibody-dependent enhancement (ADE), Ebola virus (EBOV) is known to utilize the complement component C1q for ADE of infection


The paper suggests that C1q-mediated ADE of EBOV infection is not dependent on antibodies (Ab) and simply caused by increased attachment of virus particles to the cell surface, followed by endocytosis.


Ab-dependent FcR- and Ab-independent  C1q mediated enhancement of infection reinforce each other.







Obviously, pathogenesis of EBOLA is difficult to study and regular human phase 3 vaccine studies are impossible to organize.  Nevertheless, my understanding of EBOLA has increased

  • Just like SARS-CoV-2, EBOLA  inhibits the type 1 IFN system and then leads to hyper-inflammation (cyto/chemokine storm)
  • Whereas fatal cases are associated with hyper-inflammation, exhaustion of the coagulation system, lymphocyte exhaustion and apoptosis,  the non-fatal cases are clearly associated with early antibody and T cell responses and lower inflammation.  Nevertheless, relatively high T cell responses are evident in survivors with sequelae (maybe a sign of “inflammatory” T cell activation?). 
  • There is some evidence that subclinical infection may exist in EBOLA contacts and that latent infection in humans could be a source of new outbreaks, years later.


With regard to vaccines:

  • ERVEBO, the VSV vaccine by Merck is the only one that has shown effectiveness in at least 2 or 3 field studies
  • Zabdeno-Mvabea, the Janssen vaccine, has been extensively studies in NHP and in a number of phase 2 trials, but, as far as I can see, there is no published phase 3 or real-world data.  Most probably, this vaccine will work and it has the advantage of the MVA boost with GP from 3 dangerous human filoviruses. Unfortunately, the experience with Ad26 in COVID has uncovered the risk of very rare, but potentially fatal thromboses and bleedings.
  • The Chinese vaccine, based on Ad5 has the disadvantage that this serotype is very common and that pre-existing antibodies could weaken the immune response.  The development of a similar Chinese vaccine for COVID has been stopped for that reason.
  • The Russian Gam Evac Combi vaccine (prime with VSV and boost with Ad5) could work.  In fact, if Ad5 is used as a boost, the pre-existing antibodies are probably less of a problem.  See also rather convincing data with the COVID Sputnik vaccine, which uses Ad26 (a rare serotype) as a prime and Ad5 as a booster.  However, I could only find 1 paper on phase ½ Gam Evac  in humans, but no proof of protection in animal models.
  • All other platforms (Chimp Ad3, DNA, mRNA, protein) are interesting developments, but more data are needed.
  • In general, it is difficult to compare the various vaccine platforms, as the assays to measure antibodies, T cells and also the challenge experiments are not really standardized.  We also do not know for sure which are the correlates of protection in humans.
  • In any case, since the issues encountered with the hu Ad26 and the Chimp Adeno COVID vaccines in humans, it is likely that the safety requirements for viral vector vaccines of any type (including VSV and MVA)  will be more severely scrutinized in the future. Hence more emphasis will be on mRNA and protein vaccines, potentially also optimized DNA vaccines, although, at present, they are not yet equally advanced as the vector vaccines.
  • The principle of heterologous prime-boost, as exemplified by Janssen Ad26 ZEBOV and MVA-BN-Filo (= quadrivalent MVA with GP from EBOV Zaire, EBOV Sudan, Marburg and NP Tai Forest viruses) is certainly interesting, because it has the potential to strengthen and broaden the immunity more than repeated monovalent vaccines using a single platform.  Hopefully, this concept can efficiently be translated into platforms with even lower risks (such as mRNA and protein).
  • Antibody- and/or complement-dependent enhancement has been shown in vitro in monocytic cells.  The importance of this mechanism in vivo is unknown.


Interesting website: https://microbewiki.kenyon.edu/index.php/Infection_Mechanism_of_Genus_Ebolavirus