Immunogenicity and safety of ebolavirus vaccines in healthy adults: a systematic review and meta-analysis of randomized controlled trials

ABSTRACT

Background

This review aimed to systematically evaluate the immunogenicity and safety of the candidate Ebola virus vaccine (EVV).

Methods

We searched five databases for randomized controlled trials (RCTs) evaluating the effects of EVV on healthy adults. The primary outcomes were relative risk (RR) of sero-conversion or sero-response of EVV in healthy adults between the groups that received EVV and the controls.

Results

Twenty-nine RCTs (n = 23573) were included. There was a significant difference in RR of sero-conversion of EVV (RR 13.18; 95% CI 11.28–15.41; I² = 33%; P < 0.01) between the two groups. There was a significant difference in RR of adverse events (AEs) of EVV (RR 1.49; 95% CI 1.27–1.74; I² = 88%; P < 0.01), although no difference in RR of serious AE (SAE) between the two groups. Subgroup analysis showed that there was no significant difference in RR of AEs for DNAEBO, EBOV-GP, MVA, and rVSVN4CT1 vaccines, compared with controls.

Conclusions

The DNAEBO, EBOV-GP, MVA, and rVSVN4CT1 vaccines are likely to be safe and immunogenic, tending to support the vaccination against Ebola disease. These findings should provide much-needed evidence for public health policy makers to develop preventive measures based on disease prevalence features and socio-economic conditions.

KEYWORDS:

• Ebolavirus
• vaccine
• Ebola disease
• immunogenicity
• safety
• systematic review
• meta-analysis

1. Introduction

Ebola disease due to the Ebola virus has been responsible for several major outbreaks in Africa since first being identified in 1976 [1]. Outbreaks of Ebola disease have high morbidity and mortality, which bring enormous financial and logistic burden to public health systems of affected countries and can lead to chaos worldwide [2–4].

The lack of effective treatments and the lethality of Ebola disease make safe and effective vaccination a major medical need, which has promoted several vaccine development programs [5]. So far, the recombinant replication competent vesicular stomatitis virus-based vaccine expressing the glycoprotein of a Zaire Ebolavirus licensed as ERVEBO® (rVSVΔG-ZEBOV-GP), Zabdeno®（Ad26.ZEBOV）and Mvabea® (MVA-BN-Filo) have been proven to be with high efficacy to prevent Ebola disease in high-risk population or contacts of recently confirmed cases in West African countries [6–8]. However, some safety concerns associated with these vaccines have been reported [6,8]. The need for safe and effective vaccines is even more crucial, as vaccination not only prevents infection but also limits the severity of these diseases [9,10].

Many other vaccines have been tested in healthy volunteers and have promising results on both safety and immunogenicity, although the protocols are very different (design, dose administered or time of follow-up) [5,7,11]. With the diversity of clinical trials against Ebola disease, only data on effective vaccines can control future outbreaks, such as the one that happened in West Africa or more recently in Uganda.

To the best of our knowledge, except for a systematic review and network meta-analysis [12], no other systematic review and meta-analysis have been performed to summarize published data on safety and immunogenicity of candidate vaccines against Ebola virus in healthy adults. Therefore, this meta-analysis aimed to assess the safety and immunogenicity of candidate vaccines against Ebola disease in healthy adults.

2. Methods

The study was registered with the open-access PROSPERO International prospective register of systematic reviews (CRD42022385148). This study was carried out strictly following the recommendations from the Cochrane Handbook [13] and the latest Preferred Reporting Items for Systematic Reviews of Interventions (PRISMA) guidelines [14].

2.1. Search strategy and eligibility criteria

A comprehensive literature search of PubMed/Medline, EMBASE, the Cochrane Library databases, the WHO Ebola vaccine landscape, and bioRxiv/medRxiv was conducted (from inception to 2 August 2023) for randomized controlled trials (RCTs) that investigated the safety or immunogenicity of ebolavirus vaccines (EVV) in healthy adults, with the search updated on 5 September 2023. There were no restrictions on language, region, date, participant demographics, or publication status. Descriptors were identified in Medical Subject Headings (MeSH), Descritores em Ciências da Saúde (Decs), and Embase Subject Headings (Emtree). The Cochrane-validated filter for RCTs was applied [15]. The search strategy was adapted according to descriptors in each database. Additionally, references of all included studies were also hand-searched to identify any potentially eligible studies.

2.2. Study selection

Inclusion criteria were determined a priori and organized according to the PICOS acronym: (P) Population. Healthy adults ≥18 years; (I) Interventions. Any brand or type of ebolavirus vaccine; (C) Comparisons. Receiving no intervention (including placebo); (O) Outcomes of interest: Number of participants with sero-conversion or sero-response.

For studies involving prime/boost vaccination, only data from the prime vaccination of 28 days was evaluated. Studies of nonrandomly assigned participants receiving EVV, using combination of vaccines, with participants ≤18 years, or lacking of information on outcomes, were excluded.

The primary outcome was immunogenicity, which was assessed by either relative risk (RR) of seroconversion rate (proportion of participants with at least a four-fold increase in antibody titer for enzyme-linked immunosorbent assay (ELISA) 28 days after vaccination) or sero-response rate (proportion of participants with a seropositivity measured by ELISA titer above a prespecified threshold 28 days after vaccination). Given the variability of the method used to assess immunogenicity of EVV, only seroconversion or sero-response rates based on ELISA titers at 28 days were considered. Only the seroconversion rate was used if both seroconversion and sero-response results were available in a study, as seroconversion result is less sensitive to baseline status than sero-response. Secondary outcomes were SAEs and AEs occurring within the first 14 days after vaccination.

Two investigators (JTY and LZ) screened independently titles and abstracts, and then assessed full texts of selected abstracts. Discrepancies were resolved by discussion or by a third investigator (CYW).

2.3. Data extraction

We developed a Microsoft Excel template to systematically abstract data. Extracted data included: first author name and year of publication, country/region, study design, age, follow-up, proportion of male participants, vaccine type, and dosing information, outcomes, and study sponsorship (sponsored by Government and/or Industry). If a study has more than one arm, only data from arms relevant to the review were extracted. We contacted authors of (potentially) eligible studies for additional information.

2.4. Statistical analysis

We opted for a frequentist approach to compare safety and immunogenicity between candidate vaccines using meta-analysis for binary endpoint. The Mantel–Haenszel method was used to calculate the pooled hazard ratio (RR) and 95% confidence interval (CI) using a fixed- or random-effects model according to the heterogeneity [16,17]. Heterogeneity between studies was quantified using I² statistics. I² values were considered to represent low, moderate, and high degrees of heterogeneity when values were less than 25%, 25–75%, and more than 75%, respectively.

Subgroup analysis was performed for the types of vaccines or the participants coming from different Continents/regions, and sensitivity analysis was performed for incidence of AEs when studies sponsored by industrial companies were excluded. Publication bias was assessed using egger’s test [18]. If publication bias was detected, the ‘trim and fill’ method was used to investigate the effect of publication bias. This method conservatively incorporates hypothetical negative unpublished studies to mirror the positive studies which lead to the asymmetry of funnel plot. All analyses were carried out using the RevMan 5.4 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2020).

2.5. Risk of bias assessment

Two reviewers (JTY and CYW) independently assessed the risk of bias for included trials using the Cochrane risk-of-bias tool for randomized trials (RoB 2.0) [19]. The overall certainty of the body of evidence was rated by using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach, taking into account overall risk of bias, consistency of effect, imprecision, indirectness, and publication bias [20,21]. In case there were serious concerns in any of these domains, the quality of evidence was rated down. We incorporated the overall RoB 2.0 judgments into our GRADE assessment.

3. Results

3.1. Search results and characteristics of included studies

We identified 1706 records through a systematic database search and 1289 additional records from gray literature (WHO Ebola vaccine landscape, and bioRxiv/medRxiv). Following the removal of duplicates and title and abstract review stages, 39 full-text articles of the remaining records were retrieved. Of these, four were excluded for no control group one due to the study design (not RCT) [22–25], four due to related to children [5,26–28], two for data duplication [29,30]. No additional articles were retrieved from the reference lists of included studies. Therefore, 29 studies with 23,573 participants were eligible for inclusion in this systematic review (Figure 1) [7,11,31–55].

Figure 1. PRISMA flow diagram of the article screening process.

The main characteristics of the included studies are summarized (Table 1). Four studies were conducted in more than one Continent [35,39,40,52], and other 24 studies were done in 5 Continents: Africa (n = 10 studies) [7,32,33,43,44,48,51,53–55], Asia (n = 2) [56,57], Europe (n = 5) [31,36,42,47,49], North America (n = 7) [11,34,37,41,45,46,50], and Oceania (n = 1) [38].

Table 1. The characteristics of included studies.

Source	Country/Region	Design	Age (years)†	Follow-up	Gender‡	Vaccine type	Dosing Information	Main outcomes	Sponsorship
Agnandji et al, 2016	Switzerland (Geneva)	RCT, phase 1, double-blind	18–65	6 months	61.0%	rVSV-ZEBOV-GP	rVSVΔG -ZEBOV-GP (1 × 10^7 PFU), rVSVΔG -ZEBOV-GP (5 × 10^7 PFU), placebo	Seropositivity rate measured by ELISA of 50 or more, erythema, swelling/induration, pain, fever, chills, myalgia, headache, fatigue, gastrointestinal symptoms, oral vesicle, arthralgia, arthritis	Government
Anywaine et al, 2019	Uganda (Masaka), Tanzania (Mwanza)	RCT, phase 1, observer-blind	18–50	8 months	79.0%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10^8 TCID 50), placebo	ELISA positive response rate to GP [Kikwit], injection site pain, injection site pruritus, injection site warmth, headache, fatigue, fever, myalgia, arthralgia, chills, nausea, vomiting, rash, pyrexia, pruritus (generalized)	Government and Industry
Barry et al, 2021	Burkina Faso, Cote d’Ivoire, Kenya, Uganda	RCT, phase 2, observer-blind, placebo controlled	18–50	24 months	68.6%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10^8 TCID 50), placebo	Proportion of positive ELISA GP [Z] response or responses of more than reference value, injection-site pain, and fatigue, headache, myalgia	Government and Industry
Bockstal et al, 2022	USA	RCT, phase 1, Double blind, placebo controlled	18–50	12 months	44.4%	Ad26.ZEBOV, Ad26.SUDV, Ad26.MARV, MVA-BN-Filo	Ad26.Filo (9 × 10^10 VP), MVA-BN-Filo (5 × 10^8 infectious units)	ELISA GP-specific neutralizing antibody responses and cellular immune responses, injection site pain/tenderness, headache, fatigue, chills, myalgia, decrease in hemoglobin	Industry
Clarke et al, 2020	Australia, USA	RCT, phase 1, Double blind, placebo controlled	27–60	6 months	46.0%	rVSVN4CT1-EBOVGP1	rVSVN4CT1-EBOVGP1 2 doses (d0, d28) low dose 2.5 × 10^4 intermediate dose 2.0 × 10^5 High dose 1.8 × 10^8	ELISA positive response rate to GP, fever, malaise or fatigue, myalgia, headache, nausea, vomiting, chills, arthralgia, diarrhea	Government
De Santis et al, 2016	Switzerland (Lauzanne)	RCT, phase 1-2a, double-blind	18–65	6 months	49.0%	ChAd3-EBO-Z	ChAd3-EBO-Z (2.5 × 10^10 VP), ChAd3-EBO-Z (5 × 10^10 VP), placebo	Proportion of positive ELISA GP [Z] response, pain, swelling, erythema, axillary lymphatic-node, chills, fatigue or malaise, headache, fever, nausea, musculo-articular pain, abdominal pain, conjunctivitis, rhinitis, sweating	Government
ElSherif et al, 2017	Canada	RCT, phase 1, observer-blind	18–65	6 months	42.0%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP (1 × 10^5 PFU), rVSVΔG-ZEBOV-GP (5 × 10^5 PFU), rVSVΔG-ZEBOV-GP (3 × 10^6 PFU), placebo	Seroconversion as at least 4-fold increase from ELISA baseline, pain, abdominal pain, chills, arthralgia, fever, diaphoresis, diarrhea, fatigue, headache, myalgia, nausea	Government
Fries et al, 2020	Australia	RCT, phase 1, double-blind	18–50	12 months	47.0%	EBOV-GP	EBOV GP nanoparticle vaccine (6.5, 13, 25, or 50 µg) 0.5 mL, placebo	ELISA anti-EBOV GP antibody concentration, injection site pain, redness, swelling, and bruising, fatigue, headache, muscle pain, diarrhea, nausea, joint pain, chills, vomiting, fever	Government and Industry
Goldstein et al, 2022	USA	RCT, phase 1, observer-blind, placebo controlled	18–50	24 months	42.9%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10^8 VP), placebo	ELISA positive response rate to GP, headache, fatigue, myalgia, arthralgia, chills, fatigue, headache, myalgia, nausea, pyrexia, vomiting	Government and Industry
Halperin et al, 2017	USA, Spain, Canada	RCT, phase 3, double blind, placebo controlled	18–65	24 months	45.9%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP 2 × 10^7 PFU, High dose 1 × 10^8 PFU, placebo	ELISA positive response rate to GP, PRNT seroresponse rate, injection site adverse events, fever, headache, arthralgia, pain, chills, fatigue	Government and Industry
Halperin et al, 2019	USA, Spain, Canada	RCT, phase 3, double blind, placebo controlled	18–65	24 months	45.9%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP 2 × 10^7 PFU, High dose 1 × 10^8 PFU, placebo	ELISA positive response rate to GP, PRNT seroresponse rate, injection site adverse events, fever, headache, arthralgia, pain, chills, fatigue	Government and Industry
Heppner et al,2017	USA	RCT, phase 1, double-blind	18–61	12 months	53.0%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP (3 × 10^3 PFU), rVSV-ZEBOV-GP (3 × 10^4 PFU), rVSVΔG-ZEBOV-GP (3 × 10^5 PFU), rVSVΔG -ZEBOV-GP (3 × 10^6 PFU), rVSV-ZEBOV-GP (9 × 10^6 PFU), rVSVΔG-ZEBOV-GP (2 × 10^7 PFU), rVSVΔG -ZEBOV-GP (1 × 10^8 PFU), placebo	Seroconversion for IgG ELISA as four times or more than baseline titer, abdominal pain, diarrhea, fatigue, shivering or chills, headache, joint aches and pain, joint swelling or tenderness, vomiting, myalgia, nausea, fever, sweats arm pain, local tenderness	Government
Huttner et al, 2015	Switzerland (Geneva)	RCT, phase 1–2, double-blind	18–65	12 weeks	54.0%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP (3 × 10^5 PFU), rVSVΔG -ZEBOV-GP (1 × 10^7 PFU), rVSVΔG -ZEBOV-GP (5 × 10^7 PFU), placebo	ELISA seropositivity rate, local pain, fever, fatigue, headache, myalgia, chills	Wellcome Trust through WHO
Ishola et al, 2022	Sierra Leone	RCT, phase 2, double-blind	≥18 year	24 months	82.0%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10^8 VP), MenACWY(0·5 mL) followed by a saline (0·5 mL, 0·9%)	ELISA seropositivity rate, injection-site pain, headache, myalgia, fatigue, and arthralgia	Industry
Kennedy et al, 2017	Liberia	RCT, phase 2, double-blind	≥18 year	12 months	63.4%	ChAd3-EBO-Z, rVSV-ZEBOV-GP	ChAd3-EBO-Z (1 × 10^11 VP), rVSV-ZEBOV-GP (2 × 10^7 VP), placebo	Positive antibody response as IgG antibody titer against EBOV-GP of more than 607 EU, headache, muscle pain, feverishness, fatigue, joint pain, rash, nausea, abnormal sweating, mouth ulcer	Government
Kibuuka et al, 2015	Uganda (Kampala)	RCT, phase 1b, double-blind	18–50	24 months	82.0%	DNA	EBO vaccine (4 mg), placebo	Proportion of positive response by ELISA titer of 30 or higher, pain, tenderness, swelling, redness, chills, aching joints, fatigue, headache, fever, nausea, muscle aches	Government
Kieh et al. 2022	Guinea, Liberia, Sierra Leone, and Mali	RCT, double-blind	≥18 year	12 months	55%	Ad26.ZEBOV, rVSVΔG‑ZEBOV‑GP	Ad26.ZEBOV (0.5 ml; 5 × 10^10 VP), rVSVΔG-ZEBOV-GP(1.0 ml, 9.4 × 10^7 PFU), placebo	injection-site reactions and symptoms (e.g. feverishness and headache) events	Government and Industry
Ledgerwood et al, 2010	USA	RCT, phase 1, double-blind	18–50	48 weeks	53.0%	rAd5.ZEBOV-GP	rAd5.ZEBOV.GP (2 × 10^9 VP), rAd5.SBOV.GP (2 × 10^10 VP), placebo	ELISA positive response rate to GP [Z], pain, myalgia, headache, fever, myalgia, malaise, chills, nausea, redness, swelling	NR
Li et al, 2017	China	RCT, phase 1, double-blind	18–60	6 months	48.0%	rAd5.ZEBOV-GP	rAd5.ZEBOV-GP (1.6 × 10^11 VP), rAd5.ZEBOV (4 × 10^10 VP), placebo	Number of responders assessed by neutralizing antibody titer against Ad5 > 1:200, induration, swelling, itch, fever, headache, fatigue, cough, pain, joint pain, muscle pain, throat pain, diarrhea, vomiting	Government
Martin et al, 2006	USA	RCT, phase 1, double-blind	18–44	12 weeks	67.0%	DNA	EBODNA012–00-VP(2 mg), EBODNA012–00-VP (4 mg), EBODNA012–00-VP (8 mg), placebo	ELISA positive response rate to GP [Z], pain, myalgia, headache, fever, myalgia, malaise, nausea, induration, skin discoloration	NR
Milligan et al, 2016	UK	RCT, phase 1, observer-blind	18–50	8 months	36.0%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10 8 TCID 50), placebo	Percentages of vaccine responders assessed by ELISA, pain, erythema, warmth, pruritus, swelling, induration, fatigue, headache, fever, nausea, myalgia, chills, rash, nausea, pyrexia, arthralgia, vomiting	Government and Industry
Mutua et al, 2019	Kenya (Nairobi)	RCT, phase 1, observer-blind	18–50	12 months	68.0%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10^8 TCID 50)	ELISA positive response rate to GP [Kikwit], injection site pain, injection site pruritus, injection site warmth, headache, fatigue, fever, myalgia, arthralgia, chills, nausea, vomiting, rash, pyrexia, pruritus (generalized)	Government and Industry
Pollard et al, 2021	France, UK	RCT, phase 2, double-blind, placebo-controlled	18–65	12 months	48.4%	Ad26.ZEBOV, MVA-BN-Filo	Ad26.ZEBOV (5 × 10^10 VP), MVA-BN-Filo (1 × 10^8 VP), placebo	ELISA positive response rate to GP [Z], pain, swelling, pruritus, erythema, fatigue, headache, myalgia, chills, arthralgia, nausea, fever, infections	Government and Industry
Regules et al, 2017	USA	RCT, phase 1, observer-blind	18–50	8.5 months	71.0%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP (3 × 10^6 PFU), rVSVΔG -ZEBOV-GP (2 × 10^7 PFU), rVSVΔG-ZEBOV-GP (1 × 10^8 PFU), placebo	Positive response for ELISA was a titer of 50 or more, Seroconversion was a quadrupling of the titer over the baseline value, abdominal pain, site pain, headache, myalgia, nausea, fever, fatigue, chills, arthralgia, diaphoresis, diarrhea	Government
Samai et al, 2018	Sierra Leone	RCT, phase 2/3 unblinded	≥18 year	6 months	60.6%	rVSVΔG-ZEBOV-GP	rVSVΔG-ZEBOV-GP (single dose 2 × 10^7 PFU)	Vaccine efficacy ≥ 50% or there was a 2-fold infection risk increase; safety outcome	Government
Tapia et al, 2016	USA, Mali (groups 3B and 3 C)	RCT, phase 1, double-blind	18–65	5 months	45% (U.S.A.), 80% (Mali)	ChAd3-EBO-Z	ChAd3-EBO-Z (1 × 10^10 VP), ChAd3-EBO-Z (1 × 10^11 VP), ChAd3-EBO-Z (2.5 × 10^10 VP), ChAd3-EBO-Z (5 × 10^10 VP), placebo	Proportion of positive ELISA GP [Z] response or responses of more than reference value, pain and tenderness, fever, fatigue, myalgia, arthralgia, headache, chills, nausea	Government
Tapia et al, 2020	Cameroon, Mali, Nigeria, and Senegal	RCT, phase 2, observer-blind	≥18 year	6 months	54.0%	ChAd3-EBO-Z	ChAd3-EBO-Z vaccine (1 × 10^11 particle units) placebo	Proportion of positive ELISA GP [Z] response or responses of more than reference value, pain redness, swelling, fatigue, fever, gastrointestinal, headache	Government, Wellcome Trust through WHO, and industry
Zhu et al, 2015	China	RCT, phase 1, double-blind	18–60	4 weeks	49.0%	rAd5.ZEBOV-GP	rAd5.ZEBOV.GP high dose (1.6 × 10^11 PV), rAd5.ZEBOV.GP low dose (4 × 10^10 PV), placebo	ELISA EC90 positive response, pain, myalgia, headache, fever, myalgia, fatigue, vomiting, cough, diarrhea, induration, redness, swelling, joint pain	Government and Industry
Zhu et al, 2017	Sierra Leone	RCT, phase 2, double-blind	18–50	6 ^‡months	54.0%	rAd5.ZEBOV-GP	rAd5.ZEBOV-GP (1.6 × 10^11 VP), rAd5.ZEBOV-GP (8 × 10^10 VP), placebo	Number of responders assessed by neutralizing antibody titer against Ad5 > 1:200, induration, swelling, rash, fever, headache, fatigue, cough, pain, joint pain, muscle pain, throat pain, diarrhea, itch, vomiting	Government and Industry

†Reported as median (IQR) unless otherwise stated.

‡Reported as percentage of males unless otherwise stated.

Ad5.ZEBOV: adenovirus type-5 vector-based Ebola vaccine; Ad26.ZEBOV: adenovirus type-26 vector-based Ebola vaccine; ChAd3-EBO-Z: replication-defective chimpanzee adenovirus 3 vector expressing Zaire Ebolavirus glycoprotein; EBODNA012–00-VP: Ebola virus glycoprotein DNA vaccine; ELISA: enzyme-linked immunosorbent assay; MVA-BN-Filo: modified vaccinia Ankara (MVA) expressing Zaire Ebolavirus glycoprotein and other filovirus antigens; NR: not reported; PFU: plaque-forming unit; RCT: randomized controlled trials; rVSVΔG-ZEBOV-GP: recombinant replication competent vesicular stomatitis virus-based vaccine expressing the glycoprotein of a Zaire Ebolavirus; TCID 50: 50% tissue culture infectious doses; VP: viral particles.

3.2. Quality assessment

Of all the 29 studies, 2 had an overall high risk of bias, 19 had some concerns of bias, and 8 had a low risk (Figure S1).

3.3. Primary outcomes

3.3.1. Seroconversion rate of participants

Twenty-eight of the 29 studies reported sero-conversion rate of participants (n = 8965 participants). There was a significant difference in RR of sero-conversion between the groups that received candidate vaccines against Ebola virus and controls (RR 13.18; 95% CI 11.28–15.41; I² = 33%; P < 0.01) (Figure 2).

Figure 2. RR of candidate vaccines against Ebola virus on seroconversion rate in healthy adults between the groups that received candidate vaccines against EVV and the controls.

Subgroup analysis based on the types of candidate vaccines demonstrated that the RR of seroconversion rates were lower in MVA (9.18), Ad26 or MVA (10.98), ChAd3 (11.91), DNAEBO (12.65), and EBOV-GP (12.77) compared to Ad5 (19.23), rVSVN4CT1 (25.09), rVSVZGP (22.58), and Ad26 (50.68) (Figure S2). Subgroup analysis based on the Continents from which the participants came demonstrated that the RR of seroconversion rates were lower in Oceania (12.77), Africa (21.04), Europe (22.14), and North America (26.40) compared to more than one Continent (38.95) and Asia (58.31) (Figure S3).

3.4. Secondary outcomes

3.4.1. Serious adverse events

All of the 29 studies (n = 23573 participants) found that candidate vaccines against Ebola virus did not increase the risk of SAEs compared with controls (RR 1.27; 95% CI 0.92–1.74; I² = 0%; P = 0.15) (Figure 3).

Figure 3. RR of candidate vaccines against Ebola virus on SAEs in healthy adults between the groups that received candidate vaccines against EVV and the controls.

3.4.2. Adverse events

Twenty-seven of the 29 studies reported on AEs (n = 9751 participants), and there was a significant difference in RR of AEs between the two groups (RR 1.49; 95% CI 1.27–1.74; I² = 88%; P < 0.01) (Figure S4). Subgroup analysis was performed to assess the heterogeneity among the included RCTs in the incidence of AEs, and the results showed that the RR for AEs were 1.24 (1.05, 1.47) for Ad5, 1.40 (1.08, 1.81) for Ad26, 1.85 (1.02, 3.37) for Ad26 or MVA, 1.92 (1.42, 2.59) for ChAd3, 1.12 (0.87, 1.43) for DNAEBO, 1.00 (0.89, 1.13) for EBOV-GP, 1.06 (0.83, 1.34) for MVA, 2.40 (0.34, 16.71) for rVSVN4CT1 and 1.42 (1.12, 1.79) for rVSVZGP compared with controls, respectively (Figure 4).

Figure 4. RR of different types of candidate vaccines against Ebola virus on AEs in healthy adults between the groups that received candidate vaccines against EVV and the controls after excluding studies sponsored by industrial companies.

3.4.3. Withdrawal

All of the 29 studies (n = 19617) examined withdrawal. The pooled estimate showed that there was a significant difference in RR of withdrawal due to AEs between the two groups (RR 0.59; 95% CI 0.51–0.68; I² = 0%; P < 0.01) (Figure S5).

3.5. Sensitivity analysis

Sensitivity analysis was carried out according to the sponsorship to account for heterogeneity in the incidence of AEs. When studies sponsored by the Industrial companies were excluded, there was a significant difference in RR of AEs between the two groups (RR 1.54, 95% CI 1.15–2.06, I² = 87%; P < 0.01) (Figure S6), which remained consistent with the overall analysis.

3.6. Publication bias

The funnel plot for seroconversion rates was generally symmetrical by visual inspection. Therefore, no publication bias was found in our study (Figure S7).

4. Discussion

To the best of our knowledge, except for a network systematic review and meta-analysis which concluded that the best vaccine to be used to stop future outbreak of Ebola disease is the rVSVΔG-ZEBOV-GP vaccine at dose of 2 × 10⁷ PFU [12], no other meta-analysis has been conducted to summarize published data on candidate vaccines against Ebola virus in healthy adults to identify the most effective in terms of safety and immunogenicity. However, network meta-analysis focuses only on the reference to the relative exclusion of non-reference paired comparisons and lacks of adjustment for multiplicity. Our meta-analysis is very necessary because it can overcome these shortcomings. Our study, based on 29 RCTs involving 23,573 healthy volunteers, concluded that candidate vaccines against Ebola virus were immunogenic and well tolerated, did not increase SAEs, although it may increase the incidence of AEs. However, subgroup analysis showed that the DNAEBO, EBOV-GP, MVA, and rVSVN4CT1 vaccine were immunogenic, safe, and well tolerated.

Our meta-analysis comprehensively reviewed the evidence on candidate vaccines against Ebola virus in healthy adults up to 5 September 2023. Considering immunogenicity, we found that candidate vaccines were more immunogenic in healthy adults compared with controls, and the meta-analysis for immunogenicity had a good overall consistency. In the sensitivity analysis for seroconversion rate, when excluded the studies sponsored by industrial companies, the effect was consistent with the overall analysis. Subgroup analysis according to the types of EVV demonstrated that the RR of seroconversion rates were all higher in healthy adults, among which the highest was in Ad26 group (50.68) (Figure S2), which is consistent with the previous studies and support the high protective effect of this treatment option on the prevention of Ebola disease [11,33,34,49]. Subgroup analysis based on the Continents from which the participants came showed that the RR of seroconversion rates in healthy adults were all higher compared with controls, with Asia being the highest (58.31) (Figure S3).

Although seroconversion rates of some candidate vaccines were satisfactory, the safety profile of other candidate vaccines would be questionable for mass vaccination under the condition of no immediate risk of exposure. Some candidate vaccines against Ebola virus might increase the incidence of AEs in healthy adults (Figure S4). Subgroup analysis was conducted to evaluate the heterogeneity of AEs incidence among the included RCTs, and the results showed that the DNAEBO, EBOV-GP, MVA, and rVSVN4CT1 vaccine did not increase the incidence of AEs. The withdrawal rates of participants in all of the included 29 RCTs were quite low. Additionally, our meta-analysis also showed that candidate vaccines against Ebola virus did not increase the risk of SAEs. Subgroup analysis showed no regional differences in the incidence of adverse reactions among vaccine candidates for Ebola virus. Therefore, our meta-analysis demonstrated that EVV regimens were safe and well tolerated in healthy adults worldwide.

Vaccination is often one of the countermeasures against the infectious disease outbreaks [58–61]. The numerous Ebola virus outbreaks in Africa highlight the necessity of a safe and effective prophylactic vaccine against this virulent disease [62], although there are currently several licensed vaccines (or treatments) for Ebola disease caused by the Ebola virus. Encouragingly, our study shows that several candidate vaccines against Ebola virus are safe and effective. Our study adds to the knowledge database on the strategy for Ebola vaccination, especially in healthy adults among high-risk regions. Additional studies further evaluate the Ad26.ZEBOV, MVA-BN-Filo vaccination regimen in Africa, including juveniles, younger children, and infants, pregnant women, and healthcare workers who may have to travel to regions recurrently affected by outbreaks [63,64].

This study also has several limitations. First, methodological limitations were mainly due to data integrity (withdrawal was high [41]), design of included studies (blinding of participants and personnel [51]), and potential conflicts of interests. In order to make up for this drawback, we attempted to minimize potential biases in the review process according to the methods recommended by the Cochrane Collaboration [65] and specified in our PROSPERO protocol. A second limitation to be considered is the different ELISA assays methods and the different thresholds defining seroconversion rates for immunogenicity. Standardized methods would be most appropriate in order to make the conclusions of future studies more reliable [35]. The third concern was that some trials were not pre-registered in advance before enrolling participants, while others had modified the protocol. However, as AEs were negative, these sources of bias might not have significantly affected these outcomes.

5. Conclusion

In summary, this meta-analysis shows that the DNAEBO, EBOV-GP, MVA, and rVSVN4CT1 vaccines are likely to be safe and immunogenic, tending to support the vaccination against Ebola disease caused by the Ebola virus. These findings provide much-needed evidence for public health policy makers to develop preventive measures based on disease prevalence features and socio-economic conditions.

Declaration of interests

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

J.T.Y., M.S.M. and C.J.Q. conceived the study. J.T.Y. and L.Z. collected the data. C.Y.W. analyzed and interpreted the data, and wrote the first draft of the paper. All authors edited and approved the final manuscript.

Acknowledgments

We thank Jianping Liu for the guidance and insightful comments on this manuscript.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14760584.2023.2296937.

Additional information

Funding

This research was funded by a grant from Science and Technology Department of Henan Province [grant number 112102310306].