Safety and immunogenicity of a SARS-CoV-2 recombinant protein vaccine with AS03 adjuvant in healthy adults: interim findings from a phase 2, randomised, dose-finding, multi-centre study

Study design and participants

This is an ongoing Phase 2, randomised, modified double-blind, parallel group, dose-ranging study (clinicaltrials.gov NCT04762680). The study is being conducted in 20 clinical research centres in the USA and Honduras, with a planned duration of approximately 13 months. Here, we present interim safety and reactogenicity data up to study day (D) 43, three weeks after the second vaccination, and immunogenicity data up to D36.

Adults, aged 18 years and older, were eligible. Women who were pregnant or lactating, or, for those of childbearing potential, not using an effective method of contraception or abstinence, were excluded. To allow evaluation of vaccine performance in high-risk groups, individuals with pre-existing medical conditions, those who were immunocompromised (except those with recent organ transplant or chemotherapy) and those with a potentially increased risk for severe COVID-19,⁹ were eligible for participation in the study. The inclusion/exclusion criteria are described in detail in the Appendix (Appendix – Supplementary Methods 1).

The study was undertaken in compliance with the International Conference on Harmonisation (ICH) guidelines for Good Clinical Practice and the principles of the Declaration of Helsinki. The protocol and amendments were approved by applicable Independent Ethics Committees/Institutional Review Boards, including the Advarra Institutional Review Board for the USA, and the regulatory agencies as per local regulations. Written informed consent was obtained from the participants before any study procedures were performed.

Randomisation and masking

Participant randomisation was stratified by age group (18–59 years and ≥60 years), by positivity for antibodies to SARS-CoV-2 S antigen using a rapid serodiagnostic test (positive/negative; COVID-19 IgG/IgM Rapid Test Cassette; Healgen Scientific, Houston, TX, USA) and by high-risk medical conditions (yes/no). Participants were randomly assigned, using an interactive response technology system, in a 1:1:1 ratio to receive two intramuscular injections, 21 days apart (D1 and D22), of one of the three candidate vaccine formulations containing 5 μg (low-dose), 10 μg (medium-dose) or 15 μg (high-dose) of CoV2 preS dTM antigen with a fixed dose of AS03 adjuvant. A subset of 120 participants who tested negative by the SARS-CoV-2 rapid serodiagnostic test was additionally stratified by age and study group and randomly assigned to provide samples for assessment of cell-mediated immunity (CMI) and mucosal antibodies.

This study had a modified double-blind design, such that all participants and outcome assessors were masked to group assignments and unblinded study staff involved in vaccine preparation were not involved in safety outcome assessments.

Procedures

The recombinant protein antigen CoV2 preS dTM was produced using a Sanofi Pasteur proprietary insect cell baculovirus expression vector system, based on S protein sequence from the Wuhan-Hu-1 reference strain, as previously described.⁷ The AS03 Adjuvant System (GlaxoSmithKline Vaccines, Rixensart, Belgium) is an oil-in-water emulsion containing 11·86 mg α-tocopherol and 10·69 mg squalene per 0·5 mL vaccine dose, described previously.^7,10 The AS03-adjuvanted CoV2 preS dTM vaccine formulations were presented in two separate vials: a multi-dose vial containing AS03 and a single-dose vial containing one of the three antigen dose solutions. An equal volume of the adjuvant emulsion was added to the vial containing the antigen and mixed prior to injection. Vaccinations (0.5 mL per dose) were administered by qualified and trained study personnel by intramuscular injection into the deltoid region of the upper arm.

Blood samples and nasopharyngeal swabs were collected before each vaccination for determination of SARS-CoV-2 naïve/non-naïve status. SARS-CoV-2 naïve or non-naïve status at D1 and/or D22 was determined for all participants on serological samples using ELECSYS electrochemiluminescence immunoassays for anti-SARS-CoV-2 anti-S (S-ELECSYS; Elecsys Anti-SARS-CoV-2 S, Roche, Indianapolis, IN, USA EUA202698) on D1 only and anti-nucleocapsid (N) (N-ELECSYS; Elecsys Anti-SARS-CoV-2, Roche, EUA200514) on D1 and D22, and on nasopharyngeal swabs using nucleic acid amplification tests (NAAT; Abbott RealTime SARS-CoV-2 assay, EUA200023) on D1 and D22, following the manufacturer’s instructions. SARS-CoV-2 naïve participants at D1 and D22 (naïve D1+D22) were those who tested negative for S-ELECSYS at D1 and negative for both N-ELECSYS and NAAT at both timepoints; the non-naïve subset of participants tested positive for at least one of the three tests at D1 and/or D22 (non-naïve D1/D22) (Appendix – Methods 2).

Blood samples were collected from participants at all study visits for immunogenicity assessments. SARS-CoV-2 neutralising antibody (NAb) titres against the D614G variant and the Beta (B.1.351) variant were measured using a pseudovirus neutralisation (PsVN) assay performed at Monogram Biosciences LabCorp (South San Francisco, CA, USA). The assay, based on previously described methodologies, measured NAb responses using HIV-1 pseudovirions expressing the full-length S protein of the parental D614G variant or the Beta variant.^11-13 Neutralisation by SARS-CoV-2 antibodies was measured by inhibition of luciferase activity in the target cells. Titres were calculated as the reciprocal of the serum dilution (dil) resulting in 50% neutralisation. Binding antibody (BAb) profiles were assessed by measuring SARS-CoV-2 anti-S protein IgG antibodies with an indirect enzyme-linked immunosorbent assay (ELISA) (Nexelis, Laval, Quebec, Canada), as described previously.⁷ The reference standard was human serum with known concentration of anti-S protein IgG antibodies; quantitative results were reported in EU/mL.

Whole blood samples were stimulated ex vivo with SARS-CoV-2 S Antigen (Spike-GCN4, Nexelis), using the TruCulture system (Rules-Based Medicine [RBM], Austin, TX, USA) as described previously.⁷ A microsphere-based multiplex immunoassay (TruCulture OptiMAP™ assay, RBM) was used to evaluate specific concentrations of interferon-γ [IFN γ], tumour necrosis factor-α [TNF-α], interleukin [IL]-2, IL-4, IL-5 and IL-13 on validated cytokine profiling panels. Reaction plates were analysed on a Luminex® platform and cytokine concentrations were calculated by adapted software (RBM) using a standard curve for specific cytokine production at each time point.

Participants were provided a diary card to capture solicited and unsolicited adverse events (AEs) for up to 21 days after the second injection; serious adverse events (SAEs), adverse events of special interest (AESIs) and medically attended adverse events (MAAEs) were collected over the duration of the study.

Outcomes

The primary safety objective was to describe the safety profile in all participants, for each candidate vaccine formulation. In this interim analysis, we describe primary safety endpoints up to D43 (21 days post-injection 2). Safety endpoints included the occurrence of immediate unsolicited systemic AEs occurring within 30 minutes of each injection; solicited injection site reactions (injection site pain, erythema and swelling) and solicited systemic reactions (fever, headache, malaise, myalgia, arthralgia and chills) up to seven days after each injection; unsolicited AEs reported up to 21 days after each injection; and MAAEs, serious adverse events (SAEs) and AESIs documented throughout the study. AESIs included anaphylactic reactions, generalized convulsions, thrombocytopenia and potential immune-mediated disorders.¹⁴

The primary immunogenicity objective was to describe the NAb response to the D614G variant in SARS-CoV-2-naïve participants (naïve D1+D22 subset) 14 days after the second vaccination (D36) for each study group. NAb and BAb responses were measured in all participants on D1, D22, and D36 to D614G. NAb responses to the Beta variant were measured at D36. Antibody responses were described based on geometric mean titres (GMTs) or geometric mean concentration (GMCs) for NAb and BAb antibodies respectively, the percentage of participants with ≥2-fold or ≥4-fold rise in antibody titres at each post-vaccination time point and the percentage of responders. Pre-vaccination titres below the assay lower limit of quantification (LLOQ; 1:40) were assigned a value of half of the LLOQ. Among participants who had NAb titres below the LLOQ at baseline, responders were defined as those with ≥2-fold increase in titres at a pre-defined post-vaccination time point. In participants with baseline titres above LLOQ, responders were those with a ≥4-fold increase in titres post-vaccination relative to D1. The geometric means of individual titre ratios (GMTRs), post-vaccination to pre-vaccination (D1), were described for each treatment group, following each injection. For cell-mediated immune responses, fold-rises for individual cytokines at D36/D1 and D22/D1 were calculated by dividing the post-vaccination (D22 or D36) by the pre-vaccination (D1) measurement; the ratios of fold-rises for cytokine pairs (e.g. IFN-γ to IL-4) and their 95% confidence intervals (CIs) were then computed.

NAb and BAb were also measured in a panel of human convalescent serum samples (Sanguine Biobank, iSpecimen, and PPD; n=79 samples) using the same assays as used on the participant serum samples, in the same laboratory and within the same timeframe. Convalescent samples were obtained from donors who had recovered from COVID-19 (with clinical severity ranging from mild to severe) and were asymptomatic at the time of sample collection, as described previously.⁷ Ratios of vaccine-induced antibody titres to convalescent serum titres were calculated for each antigen-dose group, by age group.

Statistics

All planned analyses were descriptive. It was estimated that a sample size of 130 evaluable SARS-CoV-2 naïve participants per group would enable a minimum observed GMT ratio between vaccine groups of 0.73, assuming a true GMT ratio of 1 and a standard deviation of 0.67 (estimated for the PsVN assay) with 95% probability. Assuming an attrition rate of 15% and capping the proportion of those testing positive by the SARS-CoV-2 rapid serodiagnostic test at 20% of the study population, a total study size of 720 participants (240 in each group) was planned.

The full analysis set (FAS) comprised all participants who received at least one study injection; those in the FAS were analysed according to the vaccine group to which they were randomised. The safety analysis set (SafAS) included all randomised participants who received at least one dose of the study vaccine and were analysed according to the vaccine actually received. Immunogenicity was assessed in participants from the per-protocol analysis set (PPAS), a subset of participants in the FAS who received both injections, provided blood samples at D1 and D36, did not have pre-specified protocol deviations for PPAS and did not receive an authorized COVID-19 vaccine prior to D36.

Primary immunogenicity endpoints were assessed in the SARS-CoV-2 naïve participant subset of the PPAS (PPAS naïve D1+D22); secondary immunogenicity endpoints were assessed in SARS-CoV-2 naïve and non-naïve participant subsets of the PPAS and FAS. CMI was analysed in a randomly assigned subset of the PPAS (CMIAS).

The 95% CIs for the GMTs/GMCs and GMTRs were calculated using normal approximation of log-transformed titres. The 95% CIs for the percentage of participants or responders were based on the Clopper-Pearson method.¹⁵ The 95% CIs for the differences in the percentage of participants with ≥2-fold or ≥4-fold rise and the percentages of responders were calculated using the Newcombe-Wilson score method without continuity correction.¹⁵ Statistical analyses were performed using SAS® Version 9.4 or later.

Role of the funding source

Funding was provided by Sanofi Pasteur and by federal funds from the Biomedical Advanced Research and Development Authority, part of the office of the Assistant Secretary for Preparedness and Response at the U.S. Department of Health and Human Services in collaboration with the U.S. Department of Defense Joint Program Executive Office for Chemical, Biological, Radiological and Nuclear Defense under Contract # W15QKN-16-9-1002. The funders were involved in the study design, data collection, data analysis, data interpretation, writing of the report, and the decision to submit the paper for publication. GlaxoSmithKline provided access to, and use of, the AS03 Adjuvant System.

Results

A total of 722 participants were enrolled and randomized in the study between 24 February 2021 and 8 March 2021; 721 received at least one injection (low-dose, n=240; medium-dose, n=239; high-dose, n=242). A total of 22 randomised participants discontinued the study by D43, none due to an AE (Appendix – Figure S1). Participant baseline demographic characteristics in the SafAS were balanced overall across treatment groups, and across age strata. Participants were 18 to 95 years old (18–59 years, n=360; ≥60 years, n=361) (Table 1); 33 were aged ≥75 years. Over 60% of participants had ≥1 high-risk medical condition (FAS; Appendix – Table S1).

The proportion of participants reporting at least one solicited adverse reaction (injection site or systemic) in the first seven days after any vaccination was similar between treatment groups (low- and high-dose groups, each 91·2%; medium-dose group, 89·9%), tended to be higher in younger adults (range across groups, 92·4–95·7%) than in older adults (86·9–89·1%), and after the second vaccination (82·3–86·3%) compared to the first vaccination (76·4–80·7%) (Figures 1A and 1B; Appendix – Table S2). The proportion of participants reporting at least one solicited reaction of grade 3 intensity after any injection was similar between treatment groups (range across groups [all age strata], 18·8–21·8%). The most frequently reported injection site reaction (all age strata) was injection site pain (range across groups, any intensity:81·9–83·7%; Grade 3: 5·0–6·3%). The most frequently reported solicited systemic reactions (any intensity, all age strata) were: malaise (range across groups, 58·0–61·1%), headache (56·1–57·1%) and myalgia (50·6–60·1%) and, for grade 3 intensity, malaise (8·4–13·0%) (Appendix – Table S2; for data by age group see Figure 1B). Grade 3 solicited injection site and systemic reactions were transient, with most occurring on the day of or the day after vaccination (Appendix – Figure S2) and resolving within two days without requiring medical attention.

• Download figure
• Open in new tab

Figure 1: Solicited injection site (A) and systemic adverse reactions (B) after each injection for low-, medium- and high-antig en dose, by age group (SafAS)

L = low-dose group (5 μg). M = medium-dose group (10 μg). H = high-dose group (15 μg). Post-inj = post-injection.

Overall, there were four immediate unsolicited systemic non-serious AEs reported in four participants (Appendix – Table S2): two assessed by the investigator to be vaccine-related (grade 1 lymphadenopathy [low-dose group] and grade 1 paraesthesia [medium-dose group] on the same side as the injection site) and two unrelated, grade 1 presyncope (low-dose group) and grade 3 hypertension (high-dose group). The case of lymphadenopathy resolved with medication within five days, while the others resolved spontaneously within one day. After any vaccination, the proportion of participants reporting at least one unsolicited AE was comparable across treatment groups (ranging between 32·9% and 36·9%; Appendix – Table S2) and following the first (range across groups, 16·3–22·0%) and second injections (range across groups, 22·8–23·6%); similar patterns were observed for unsolicited adverse reactions (ARs). The proportion of participants reporting unsolicited AEs after any injection was slightly higher among the younger than the older adults (range across groups, 36·7–41·3% versus 29·2–36·9%), while the proportion of participants reporting unsolicited ARs was similar between age strata (range across groups, 14·3–24·8% among younger adults and 13.3–14.8% among older adults). The proportion of participants with grade 3 unsolicited AEs after any injection tended to be higher in the high antigen dose group (7·9%) than in the low- (2·1%) or medium-dose (2·5%) groups; this increase with higher antigen dose was also observed with Grade 3 unsolicited ARs. The majority of unsolicited ARs were compatible with reactogenicity symptoms (Appendix -Table S3), were of grade 1 or grade 2 intensity, occurred within the first four days after injections and generally resolved within seven days. Six participants in the high-dose group and one participant in the low-dose group were reported to have unsolicited AEs with the MedDRA preferred terms (PTs) of elevated blood pressure, elevated systolic blood pressure, essential hypertension or hypertension. These occurred shortly after vaccination, self-resolved within 1–2 days and occurred without any other associated symptoms in all but one case; one participant with grade 3 hypertension had macular rash and headache, all assessed as related to the study vaccine, and anxiety (assessed as unrelated). COVID-19 illness was reported for four participants with the onset of illness prior to D36 (14 days after the second vaccination); one in each of the low- and medium-dose groups and two in the high-dose group. MAAEs were reported for 62/721 (8·6%) participants overall, with no clear pattern between treatment groups (Appendix – Table S2). Of these, 14 were grade 3 intensity, seven (reported by five participants) were assessed by the investigator as related to the vaccine (low-dose, n=2; medium-dose, n=1; high-dose, n=4) and three (not related) were assessed as serious (medium-dose, n=2; high-dose, n=1). Grade 3 MAAEs tended to be more frequent in the high-dose group (3·3%; 8/241) compared to the low- and medium-dose groups (1.7% [4/240] and 0.8% [2/240], respectively); these included two events of grade 3 hypertension assessed as related to the study vaccine, both with a duration of 2 days. No AEs led to study discontinuation, and no AESIs were reported. Four SAEs were reported (two in each of the medium-dose and high-dose groups), none of which were considered by the investigator or the sponsor to be related to the study vaccine. The safety and reactogenicity profiles were similar between participants who were SARS-CoV-2 naïve at baseline (D1) and those who were non-naïve at D1 (Appendix - Table S2).

Among 611 participants in the PPAS, 598 had sufficient information to determine per-protocol SARS-CoV-2 naïve status at D1 and D22: naïve, n=521 (85·2%); non-naïve, n=77 (12·6%), with similar proportions across treatment groups (naïve: n=168, n=177, n=176; non-naïve: n=28, n=26, n=23 for low-, medium- and high-dose groups, respectively).

Among SARS-CoV-2 naïve participants in the PPAS, NAb titres and BAb concentrations to the D614G variant 21 days after the first injection (D22) showed minimal increase from baseline, regardless of antigen dose group, for both age strata (Figure 2A and Figure 3A).After the second injection (D36), Nab titres tended to increase with antigen-dose level in the younger age group but not in the older adults; titres were higher for younger adults than for older adults within each dose group (Figure 2A). On D36, the proportion of responders (≥2-fold-rise in NAb titres from D1 to D36) ranged from 97·4% to 100% across treatment groups for younger adults (18–59 years) and 95·1% to 100% for older adults (≥60 years); the proportion of those with a ≥4-fold-rise in Nab titres ranged from 97·4% to 100% among younger adults and 93·9% to 96·7% among older adults (Table 2). Similar observations were made for D36 BAb responses (Appendix – Table S4). For both age strata combined, the magnitude of the NAb response after two injections (D36) was comparable across treatment groups (D36 GMTs [1/dil] ranging from 2189 [low-dose] to 2895 [high-dose group]), and comparable to titres observed in the convalescent sera panel (2140 [95% CI 1543, 2967]) (Table 2). The ratio of vaccine-induced NAb titres to titres in the convalescent panel was 1·38, 1·85, and 2·40 among younger adults, and 0·76, 0·65 and 0·81 among older adults for the low-, medium- and high-antigen dose groups, respectively.

• Download figure
• Open in new tab

Figure 2: Neutralising antibody response to D614G, after each injection of low-, medium- and high-dose CoV2 preS dTM-AS03 formulations, by SARS-CoV-2 naïve status (PPAS).

Footnote: Boxes indicate median and quartile ranges. Outliers are plotted as individuals points. D, study day; n, number of participants available for each endpoint are shown in the table. For the panel of convalescent sera, n=79

• Download figure
• Open in new tab

Figure 3: Binding antibody response to D614G, following each injection of low-, medium- and high-dose CoV2 preS dTM-AS03 formul ations, by SARS-CoV-2 naïve status.

Footnote: Boxes indicate median and quartile ranges. Outliers are plotted as individuals points. D, study day; n, number of participants available for each endpoint are shown in the table. For the panel of convalescent sera, n=78

Among SAR-CoV-2 naïve participants with at least one high risk medical condition, antibody titres to D614G variant were comparable across all antigen dose groups in both age strata. Among naïve participants without high-risk medical conditions, higher NAb titres with increasing antigen dose level was observed in younger adults (Appendix – Table S5). After two doses, NAb titres among those without high-risk conditions were higher than those with ≥1 high-risk medical condition in the medium- and high-dose groups, particularly in the younger age stratum, although this was not observed in the low-dose group.

In participants in the PPAS who were non-naïve on D1 and/or D22, NAb titres (Figure 2B) and BAb concentrations (Figure 3B) to D614G variant increased by more than 10-fold in both age strata 21 days after a single injection (D22) in all antigen-dose groups. Higher titres and greater fold-increases were observed in the younger age strata compared to the older age group. After the first injection, a ≥2-fold rise in NAb titres was observed for 74·1% (20/27), 59·1% (13/22), and 73·9% (17/23) participants (all ages), in the low-, medium- and high-dose groups, respectively, and a ≥4-fold rise for 66·7% (18/27), 54·5% (12/22) and 69·6% (16/23) participants, respectively. GMTs increased further after the second injection, albeit to a lesser extent, with nearly all participants in each group achieving ≥4-fold rise in NAb and BAb levels at D36 (Table 2; Appendix – Table S5). The D22 titres in non-naïve participants were higher across all antigen-dose groups than those achieved among naïve participants after two doses (D36) (Figures 2 and 3).

The NAb response to the Beta variant was assessed at D36 only. In the PPAS naïve D1+D22 participants, GMTs were comparable between the low- and medium-dose groups, and slightly higher for the high-dose group, with titres approximately 10-fold lower than for the D614G variant (Table 2; Appendix – Figure S3). The pattern of NAb responses to the Beta variant with age was similar to that observed with responses to the D614G variant, with higher titres in younger compared to older adults. In the non-naïve participants, the Beta variant GMTs were also comparable between the low- and medium-dose groups with higher titres for the high-dose group (Table 2).

Cell-mediated immunity was assessed in a subset of 104 participants: 36 participants in the low-dose group, 31 participants in the medium-dose group, and 37 in the high-dose group. Increases of IFN-γ, IL-2, and TNF-α cytokines from pre-vaccination to D22 and D36 were more robust than were the increases for IL-4, IL-5, and IL-13, suggesting no Th2 cell bias in the cell-mediated responses (Appendix – Figure S4).