Safety and immunogenicity of heterologous boost immunization with an adenovirus type-5-vectored and protein-subunit-based COVID-19 vaccine (Convidecia/ZF2001): A randomized, observer-blinded, placebo-controlled trial

Background Heterologous boost vaccination has been proposed as an option to elicit stronger and broader, or longer-lasting immunity. We assessed the safety and immunogenicity of heterologous immunization with a recombinant adenovirus type-5-vectored Coronavirus Disease 2019 (COVID-19) vaccine (Convidecia, hereafter referred to as CV) and a protein-subunit-based COVID-19 vaccine (ZF2001, hereafter referred to as ZF). Methods and findings We conducted a randomized, observer-blinded, placebo-controlled trial, in which healthy adults aged 18 years or older, who have received 1 dose of Convidecia, with no history of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection, were recruited in Jiangsu, China. Sixty participants were randomly assigned (2:1) to receive either 1 dose of ZF2001 or placebo control (trivalent inactivated influenza vaccine (TIV)) administered at 28 days after priming, and received the third injection with ZF2001 at 5 months, referred to as CV/ZF/ZF (D0-D28-M5) and CV/ZF (D0-M5) regimen, respectively. Sixty participants were randomly assigned (2:1) to receive either 1 dose of ZF2001 or TIV administered at 56 days after priming, and received the third injection with ZF2001 at 6 months, referred to as CV/ZF/ZF (D0-D56-M6) and CV/ZF (D0-M6) regimen, respectively. Participants and investigators were masked to the vaccine received but not to the boosting interval. Primary endpoints were the geometric mean titer (GMT) of neutralizing antibodies against wild-type SARS-CoV-2 and 7-day solicited adverse reactions. The primary analysis was done in the intention-to-treat population. Between April 7, 2021 and May 6, 2021, 120 eligible participants were randomly assigned to receive ZF2001/ZF2001 (n = 40) or TIV/ZF2001 (n = 20) 28 days and 5 months post priming, and receive ZF2001/ZF2001 (n = 40) or TIV/ZF2001 (n = 20) 56 days and 6 months post priming. Of them, 7 participants did not receive the third injection with ZF2001. A total of 26 participants (21.7%) reported solicited adverse reactions within 7 days post boost vaccinations, and all the reported adverse reactions were mild, with 13 (32.5%) in CV/ZF/ZF (D0-D28-M5) regimen, 7 (35.0%) in CV/ZF (D0- M5) regimen, 4 (10.0%) in CV/ZF/ZF (D0-D56-M6) regimen, and 2 (10.0%) in CV/ZF (D0-M6) regimen, respectively. At 14 days post first boost, GMTs of neutralizing antibodies in recipients receiving ZF2001 at 28 days and 56 days post priming were 18.7 (95% CI 13.7 to 25.5) and 25.9 (17.0 to 39.3), respectively, with geometric mean ratios of 2.0 (1.2 to 3.5) and 3.4 (1.8 to 6.4) compared to TIV. GMTs at 14 days after second boost of neutralizing antibodies increased to 107.2 (73.7 to 155.8) in CV/ZF/ZF (D0-D28-M5) regimen and 141.2 (83.4 to 238.8) in CV/ZF/ZF (D0-D56-M6) regimen. Two-dose schedules of CV/ZF (D0-M5) and CV/ZF (D0-M6) induced antibody levels comparable with that elicited by 3-dose schedules, with GMTs of 90.5 (45.6, 179.8) and 94.1 (44.0, 200.9), respectively. Study limitations include the absence of vaccine effectiveness in a real-world setting and current lack of immune persistence data. Conclusions Heterologous boosting with ZF2001 following primary vaccination with Convidecia is more immunogenic than a single dose of Convidecia and is not associated with safety concerns. These results support flexibility in cooperating viral vectored and recombinant protein vaccines. Trial registration Study on Heterologous Prime-boost of Recombinant COVID-19 Vaccine (Ad5 Vector) and RBD-based Protein Subunit Vaccine; ClinicalTrial.gov NCT04833101.


Introduction
Coronavirus Disease 2019 , caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), has severely impacted the world in terms of health, society, and economy [1]. The mass vaccination campaigns are fundamental to reducing the burden of disease and the subsequent economic recovery. As of April 7, 2022, 11.4 billion doses have been administered globally, and 64.7% of the world population has received at least 1 dose of a COVID-19 vaccine, but only 14.8% of people in low-income countries have received at least 1 dose [2]. Currently, national regulatory authorities have granted authorizations for more than 15 COVID-19 vaccines, including 4 adenovirus-based vector vaccines: ChAdOx1 nCoV-19 (AstraZeneca), Ad26.COV 2-S (Janssen), rAd26+rAd5 (Gamaleya), and Ad5 nCoV (CanSino, Convidecia).
Heterologous regimens have been proposed as an option to elicit stronger and broader, or longer-lasting immunity, which is particularly important for COVID-19 vaccines with moderate vaccine efficacy. The results from clinical trials and real-world studies suggested that heterologous prime-boost vaccination of adenovirus-vectored vaccines (ChAdOx1 nCoV-19 or Ad26.COV 2-S) followed by mRNA vaccines (BNT162b2 or mRNA-1273) induced stronger immune responses, and provided higher effectiveness than homologous ChAdOx1 nCoV-19 vaccination [13,14]. Additionally, the results from Com-COV2 and COV-BOOST trials showed heterologous immunization with ChAdOx1 nCoV-19 and NVXCoV2373 induced both humoral and T-cell immune responses superior to that homologous ChAdOx1 nCoV-19 vaccination [15,16]. Robust data on the safety and immunogenicity of heterologous schedules with different COVID-19 vaccines will help enhance deployment flexibility and improve access to vaccines.
Here, we present the safety and immunogenicity of a heterologous prime-boost vaccination of a recombinant adenovirus type-5-vectored COVID-19 vaccine (Convidecia) followed by a recombinant protein-subunit-based COVID-19 vaccine composed of dimeric receptor-binding domain (RBD) (ZF2001) in healthy adults.

Study design and participants
This study was designed as a randomized, observer-blinded, placebo-controlled trial to access the safety and immunogenicity of a heterologous prime-boost immunization with Convidecia and ZF2001 in Guanyun County, Jiangsu Province, China. The trial protocol was reviewed and approved by the institutional review board of the Jiangsu Provincial Center of Disease Control and Prevention (approval number: JSJK2021-A005-02) and performed in accordance with the Declaration of Helsinki and Good Clinical Practice. This trial was registered with ClinicalTrials.gov NCT04833101. The study protocol, including the CONSORT checklist, can be found in S1 Study Protocol and S1 CONSORT Checklist.
In the original protocol, participants who have received 1 dose of Convidecia were randomly assigned (2:1) to receive either 1 intramuscular dose of ZF2001 or placebo control (trivalent inactivated influenza vaccine (TIV)), administered at either 28 days or 56 days after priming. We made a protocol change to add an additional boost vaccination with ZF2001 at 4 months after first boost dose to further boost the immune responses for all the participants. Four permutations of prime-boost schedule were investigated, including receiving ZF2001/ ZF2001 at 28 days and 5 months after priming with Convidecia (referred to as CV/ZF/ZF (D0-D28-M5)), receiving ZF2001 at 5 months after priming (referred to as CV/ZF (D0-M5)), receiving ZF2001/ZF2001 at 56 days and 6 months after priming (referred to as CV/ZF/ZF (D0-D56-M6)), and receiving ZF2001 at 6 months after priming (referred to as CV/ZF (D0-M6)) ( Fig 1A).

Vaccines
Convidecia and ZF2001 have been authorized for conditional license or emergency use against COVID-19 in China and other countries, and manufactured by CanSino Biologics and Anhui Zhifei Longcom Biopharmaceutical, respectively. ZF2001 encodes the SARS-CoV-2 receptorbinding domain (RBD) antigen (residues 319 to 537, accession number YP_009724390), with 2 copies in tandem repeat dimeric form, and was manufactured in the CHOZN CHO K1 cell line as a liquid formulation containing 25 μg per 0.5 ml in a vial, with aluminium hydroxide as the adjuvant [17]. The control influenza vaccine is produced by Dalian Aleph Biomedical. Administration is via 0.5 ml intramuscular injection into the upper arm for both ZF2001 and TIV.

Randomization and masking
The randomization lists were generated by an independent statistician using SAS (version 9�4). Participants were stratified by age (18 to 59 years and �60 years), and then randomly assigned (2:1) to vaccine group receiving ZF2001 and "placebo" group receiving a TIV either 28 days or 56 days apart. Randomization was done by random block size of 6. Influenza vaccine was used as a placebo comparator rather than a saline placebo to maintain masking of participants and ensure that all participants obtain some benefit.
We masked participants, investigators, laboratory staff, and outcome assessors to the allocation of treatment groups, but not the prime-boost intervals. Personnel who prepared and administered vaccination were aware of group allocation, but they were not otherwise involved in other trial procedures or data collection, and were instructed not to reveal the identity of the study vaccines to the participants and other investigators.

Procedures
We recruited participants from a population cohort receiving Convidecia, in which participants had no history of SARS-CoV-2 infection before priming with Convidecia, confirmed by SARS-CoV-2 IgM/immunoglobulin G (IgGAU : PleasenotethatIgGhasbeendefinedasimmunoglobulinG ) antibodies rapid test kit (Vazyme, Nanjing, China). Eligible participants were randomly assigned to receive 1 shot of ZF2001 or TIV in a ratio of 2:1 at 28 days or 56 days after the prime vaccination. Four months after receiving the first boost dose, all the participants were administrated with an additional dose of ZF2001. Participants were monitored for 30 minutes post each vaccination for any immediate adverse reactions and instructed to record solicited adverse events up to day 7 and unsolicited adverse events up to day 28 after each dose on paper diary cards. Serious adverse events self-reported by participants were documented throughout the study. Adverse events were graded as mild (grade 1), moderate (grade 2), severe (grade 3), or life-threatening (grade 4) according to the scale issued by the China State Food and Drug Administration (version 2019). Safety data within 28 days after priming with Convidecia were collected in the same manner as the boost vaccination. If a participant developed fever accompanied by respiratory symptoms, he/she was instructed to seek medical attention and notify study staff. For the suspected COVID-19, the participant had a nasal/throat swab taken for PCR test.

PLOS MEDICINE
Blood samples were taken on day 28 (referred to as baseline), day 42, day 56, day 164, after priming in CV/ZF/ZF (D0-D28-M5) and CV/ ZF (D0-M5) regimens, and on day 28 (referred to as baseline), day 56, day 70, day 84, day 194, after priming in CV/ZF/ZF (D0-D56-M6) and CV/ ZF (D0-M6) regimens ( Fig 1A). A microneutralization assay with 50% tissue culture infectious dose of 100 in each well (100 TCID 50 ) was used to measure live virus neutralizing antibodies. IgG binding antibody concentrations against RBD and spike protein were detected by an indirect enzyme-linked immunosorbent assay (ELISA) (Vazyme Biotech). The World Health Organization (WHO) international standard for anti-SARS-CoV-2 immunoglobulin (NIBSC code 20/136) was used side by side as reference with the serum samples measured in this study for calibration and harmonization of the serological assays [18]. The WHO reference serum (1,000 IU/ml) is equivalent to neutralizing antibody titer of 1:320 against wild-type SARS-CoV-2. Seroconversion was defined as at least a 4-fold increase in the antibody titers at different time points after priming compared to the baseline level (on day 28 post priming).
Peripheral blood mononuclear cells (PBMCs) were collected at 28 days post prime dose and 14 days post each boost dose, and for the determination of T cells producing interferon (IFN) γ. IFN-γ responses to spike glycoprotein were evaluated using enzyme-linked immunospot (ELISpot) (Mabtech, Stockholm, Sweden) and were expressed as spot-forming units (SFUs) per million PBMCs. An ad hoc analysis was conducted, comparing the wild-type virus neutralizing antibodies to a sample serum panel from 40 participants receiving homologous immunization of Convidecia following a 0 to 56 days regimen and 20 participants receiving homologous prime-boost vaccination of Convidecia with a 0 to 6 months regimen in previous trials [19,20]. Details of assays are included in the Supporting information (S2 Text).

Outcomes
The primary outcomes were the occurrence of solicited local or systemic adverse reactions within 7 days post vaccination and neutralizing antibody titers against wild-type SARS-CoV-2 at day 14 after each boost vaccination. Safety secondary outcomes include unsolicited adverse events for 28 days after immunization and serious adverse events collected throughout the study. Immunological secondary outcomes include the binding IgG concentration against SARS-CoV-2-specific RBD and spike protein at day 14 after each boost vaccination, and neutralization antibody titers against wild-type SARS-CoV-2 and binding IgG concentration at day 28 post prime dose, at day 28 post first boost dose and at month 6 post second boost dose.
The exploratory outcomes were neutralizing antibodies against the Delta variant B.1.617.2 at 28 days post prime dose and 14 days post second boost dose, and cellular responses measured by IFN γ ELISpot in peripheral blood at 28 days post prime dose and 14 days post each boost dose.

Statistical analyses
We assumed that the geometric mean titer (GMT) of neutralizing antibodies was about 1:20 at baseline (28 days after 1 dose of prime vaccination with Convidecia). After the second dose, GMT in the vaccine group was expected to reach 1:60 at day 14 post boost vaccination, while that in the control group remained unchanged. Assuming a standard deviation (SD) of 4, 40, and 20 participants receiving vaccine and placebo control in each regimen group, respectively, was estimated to provide 81.6% power for declaring the superiority.
Participants who received at least 1 dose of vaccine were included in the safety analysis. The primary immunogenicity analysis was done in the intention-to-treat population, including all participants who were injected and donated blood samples for antibody tests post boost vaccination. Subgroup analyses were performed according to the age for neutralizing antibodies against wild-type SARS-CoV-2. We assessed the number and proportion of participants with adverse reactions post vaccination. The antibodies against SARS-CoV-2 were presented as GMTs, geometric mean fold increases (GMFIs), and seroconversion with 95% confidence intervals (CIs), and the cellular responses were shown as the average number of positive cells per PBMCs. We used the χ 2 test or Fisher's exact test to analyze categorical data, and t test to analyze the log-transformed antibody titers. Normal distribution of the data was tested using Kolmogorov-Smirnov test. For nonnormal distributed data, Wilcoxon rank-sum test was used. The correlation between concentrations of log-transformed neutralizing antibodies and binding antibody was analyzed using Spearman's correlation with 95% CIs. Hypothesis testing was 2-sided with an α value of 0.05. Statistical analyses were done by a statistician using SAS (version 9�4) or GraphPad Prism 8.0.1.

Immunogenicity
Neutralizing antibody responses against wild-type SARS-CoV-2. The neutralization responses were detectable in 50.8% (61/120) of the participants at 28 days post prime vaccination with Convidecia, with GMT of 8.3 (95% CI 7.0 to 9.7). At 14 days post first boost, GMT of wild-type virus neutralizing antibodies in recipients receiving ZF2001 at 28 days and 56 days post priming was 18.7 (95% CI 13.7 to 25.5) and 25.9 (17. Table C in S1 Data). Homologous immunization with Convidecia at "0 to 56 days" regimen and "0 to 6 months" regimen induced neutralizing antibodies with the GMTs of 32.0 (95% CI 23.7 to 43.0) and 123.6 (82.8, 184.5) at 28 days post boost dose, respectively, which were equivalent to those induced by heterologous boost immunization with ZF2001 (S1 Fig). Neutralizing antibodies 14 days after boost dose were numerically higher in participants aged 18 to 59 years than in those over 60 years across 4 regimens, but no significant difference were found between age subgroups due to small sample size (Fig  3 and Table D in S1 Data).
Similar patterns of binding antibody responses were observed in age subgroups, and the younger adults had numerically higher humoral responses than did the older adults (S2 and

Neutralizing antibody responses against SARS-CoV-2 variants
Twenty-eight days after prime immunization with Convidecia, the GMT of neutralizing antibody titers to B.1.617.  Table C in S1 Data). Similar with that after prime immunization with Convidecia, the GMRs of neutralizing antibodies against wildtype relative to Delta variant elicited by boost vaccination ranged 2.9 and 3.4 across 4 heterologous regimens (Fig 5A and 5B). In addition, we tested the serum samples for neutralizing antibodies against the Omicron variant in CV/ZF/ZF (D0-D56-M6) regimen. The results showed

Discussion
Our findings show that heterologous immunization of ZF2001 in individuals primed with Convidecia is not associated with safety concerns in this study. The lower frequency of systemic adverse reactions was reported after boost dose with ZF2001, compared with that after prime dose with Convidecia. Four prime-boost regimens induced robust immune responses, with GMTs of wild-type virus neutralizing antibodies ranged 91 to 141, which is equivalent to range 283 to 441 IU/ml using the WHO international standard. This is comparable to 344 IU/ ml in participants primed with Janssen Ad26.COV2-S and then given BNT162b2 [21]. Heterologous immunization of ZF2001 with prime-boost intervals of 28 days or 56 days induced 2.5-and 3.3-fold higher neutralizing antibodies against wild-type SARS-CoV-2 than those induced by a single dose of Convidecia, respectively. In addition, we also found that heterologous vaccinations with Convidecia and ZF2001 at an interval of 5 months or 6 months are more efficient in eliciting neutralizing antibodies, with GMFIs of 8.9 and 10.7 compared to a single dose of Convidecia, respectively.
In the present study, we found that the impact of dose interval on the immune responses was greater than the number of doses, which may relate to memory B cell maturation undergoing during 4 to 6 months [22]. Notably, heterologous vaccination with Convidecia and ZF2001 at an interval of 5 months or 6 months induced higher antibody titers than that elicited by immunization at 28 days or 56 days apart. Additionally, 2-dose schedules with D0-M5 and D0-M6 induced antibody level comparable with that elicited by 3 doses of heterologous immunization with D0-D28-M5 and D0-D56-M6 schedules. Zhao and colleagues [23] also showed that prolongedinterval ZF2001 (receiving 3 doses of ZF2001 at interval of month 0, 1, and 4, M0-M1-M4) induced higher binding and neutralizing antibodies than the short-interval ZF2001 (M0-M1-M2), including against SARS-CoV-2 prototype strain and variants of concern such as Delta and Omicron. These findings support the use of a prolonged booster interval to elicit stronger immune responses in persons who had previously received prime immunization of COVID-19 vaccine.
As we know, preexisting anti-adenovirus immunity is the biggest obstacle for the adenovirus-vectored vaccines to overcome, especially for Ad5 eliciting widespread preexisting immunity in the human population. In the previous Phase IIb trial of Convidecia, the boosting effect of homologous prime-boost regime apart 56 days on immune responses was limited due to high anti-Ad5 antibodies [19]. In order to minimize the negative effect of preexisting anti-Ad5 antibody, heterologous prime-boost regimens and a wider prime-boost interval are necessary to provide enhance of immune responses. Our study indicated that homologous Convidecia vaccination apart 6 months induced comparable antibodies with that elicited by heterologous Convidecia-ZF2001 immunization at an interval of 5 months or 6 months. The impact of dose interval was also observed in other vectored COVID-19 vaccines. There is a better immunogenicity when a second dose of Ad26 is given at 6 months after the first dose of Ad26 compared with 2 months [24]. ChAdOx1 nCoV-19 has also shown that a longer prime-boost interval (�12 weeks) provided higher protective efficacy than a short interval (>6 weeks) [25].
The neutralizing antibodies against the Delta variant elicited by heterologous immunization with Convidecia and ZF2001 decreased about 2.9-to 3.4-fold relative to wild type across the 4 different regimens, and which was similar with that after the prime immunization. Nevertheless, heterologous schedules maintained higher neutralizing antibodies against Delta variant than prime vaccination. In addition, boost vaccination of ZF2001 in individuals primed with Convidecia increased neutralizing antibodies against Omicron and partially restored neutralization of the Omicron variant, but responses were still up to 19-fold decrease compared to wild type. Our results are consistent with those from studies by Servellita and colleagues and GeurtsvanKessel and colleagues and indicate that booster vaccinations are needed to further restore Omicron cross-neutralization by antibodies [26,27]. The use of ZF2001 as a boost does not further increase the cellular immunity responses obtained after the initial dose of Convidecia, which was in line with that reported in a previous trial with ZF2001 booster at interval of 4 to 8 months following 2-dose inactivated vaccines [28]. Compared with protein-subunit-based vaccines containing aluminium adjuvants, those with novel adjuvants could induce stronger immune responses. The results of Com-COV2 study showed that heterologous immunization with ChAdOx1 nCoV-19 vaccine and NVXCoV2373 (a Matrix-M adjuvanted recombinant spike protein vaccine) at an interval of 8 to 12 weeks induced both humoral and T-cell immune responses superior to that elicited by homologous ChAdOx1 nCoV-19 vaccine [15]. The lack of further increase in T-cell responses in our study might be related to the test method, since T cells detected would not continue to increase after reaching a plateau in the ELISpot assay.
Data from the Phase III efficacy trial showed a single dose of Convidecia could provide 57.5% protective efficacy against symptomatic COVID-19 at 28 days or more postvaccination and 91.7% vaccine efficacy against severe disease [29]. The ZF2001 vaccine was shown to be safe and effective against symptomatic and severe-to-critical COVID1-9 for at least 6 months after full vaccination. Three doses (30 days apart) of ZF2001 could provide 75.7% efficacy against symptomatic COVID-19 and 87.6% efficacy against severe-to-critical disease. In addition, vaccine efficacy against Alpha and Delta variants was 88.3% and 76.1%, respectively [30]. Our findings indicate that the heterologous schedule of Convidecia followed by a boost dose of ZF2001 with 5 to 6 months interval increased neutralizing antibodies by 9-to 17-fold, compared with that after an initial dose of Convidecia. Given the established associations between neutralizing antibody titers and vaccine efficacy [3,31], heterologous immunization with Convidecia and ZF2001 5 to 6 months apart are also likely to be highly effective, and could be considered in some circumstances for national vaccine programs.
This study has several limitations. First, the absence of a randomized control group completing the homologous Convidecia scheme. Although we selected 2 extend controls receiving homologous immunization of Convidecia following a 0 to 56 days regimen and 0 to 6 months regimen, which are both comparable with the cohorts receiving heterologous immunization with Convidecia and ZF2001 in baseline characteristics (Table A in S1 Data), there may be some potential bias. As an immunogenicity and reactogenicity study, we do not know whether the immune responses observed in our study will result in better effectiveness, and it is needed to be confirmed in real-world studies. We are unable, at this point, to determine whether higher antibody titers measured at 14 days post boost immunization will result in a more sustained elevation of antibodies, and this will be assessed at 6 months post second boost vaccination. Additionally, we did not collect blood samples before the third dose to assess the true impact of the third dose, and it would explain why there was little difference between 2-dose and 3-dose regimens. Lastly, the sample size of each regimen was relatively small, which limited the statistical power for age subgroup analyses. Findings from this study need to be validated in a large sample size.
In conclusion, our study shows that heterologous schedules of ZF2001 following the primary vaccination of Convidecia are not associated with safety concerns and can induce significant humoral immunity, particularly with a 5 to 6 months prime-boost interval. These results support flexibility in cooperating viral vectored vaccines and recombinant protein vaccine, subject to supply and logistical considerations.