Immunogenicity of heterologous prime/booster-inactivated and adenoviral-vectored COVID-19 vaccine: real-world data

Limited data are available on the responses to heterologous vaccine regimens for SARS-CoV-2, especially among countries using inactivated and adenoviral-vectored vaccines. A total of 77 participants who received heterologous prime/booster-inactivated COVID-19 vaccine and adenoviral-vectored vaccine were enrolled in our study. There were two comparison groups vaccinated with the homologous CoronaVac (N = 79) and AZD1222 (N = 80) regimen. All sera samples were tested for SARS-CoV-2 spike receptor-binding-domain (RBD) IgG using a chemiluminescent microparticle immunoassay (CMIA). The neutralizing activity in a subset of serum samples was tested against the original Wuhan strain and variants of concern, B.1.1.7 and B.1.351, using an enzyme-linked immunosorbent assay (ELISA)-based surrogate virus neutralization test (sVNT). The CoronaVac followed by the AZD1222 vaccine induced higher levels of spike RBD-specic IgG than that of two-dose CoronaVac or AZD1222 vaccines (p < 0.001). Sera samples of the CoronaVac/AZD1222 vaccine recipients elicited higher neutralizing antibody activity against the original Wuhan and the B.1.351 strain than in the recipients of the two-dose CoronaVac or AZD1222. Following the inactivated CoronaVac/adenoviral-vectored (AZD1222) vaccination administered 14–72 days apart, participants receiving the heterologous vaccine regimen had higher spike RBD-specic IgG and neutralizing activities than the homologous CoronaVac vaccine recipients. two-dose regimens 1 . The CombiVacS trial found that the Pzer’s BNT162b2 mRNA vaccine given as a second dose in individuals prime vaccinated with AZD1222 induced a robust immune response, with an acceptable and manageable reactogenicity prole 2


Introduction
Limited data are available on the responses to heterologous vaccine regimens for SARS-CoV-2, especially among countries using inactivated and adenoviral-vectored vaccines. In the Com-COV trial, heterologous prime-booster combinations of the adenoviral-vectored (AZD1222) and mRNA vaccines induced higher serum anti-spike receptor-binding domain (RBD) and neutralizing antibody titers than the homologous e cacy-proven adenoviral-vectored two-dose regimens 1 . The CombiVacS trial found that the P zer's BNT162b2 mRNA vaccine given as a second dose in individuals prime vaccinated with AZD1222 induced a robust immune response, with an acceptable and manageable reactogenicity pro le 2 .
Some countries have recommended the heterologous vaccine regimen, offering the mRNA vaccines as a second dose to young people who have already received the AZD1222 vaccine because of concerns about vaccine-induced thrombotic thrombocytopenia (VITT) 3 . The preliminary results showed that the heterologous vaccine regimen (AZD1222/BNT162b2) induced higher humoral and cellular responses to SARS-CoV-2 compared to the homologous vaccine regimen (AZD1222/AZD1222) 4 . Further research is required to comprehensively elucidate the immunological implications following different vaccine types and administration schedules.
Thailand has imported CoronaVac, which was developed by Sinovac Life Sciences, Beijing, China, since February 2021. Several phase 3 studies have shown acceptable safety and e cacy against symptomatic COVID-19 following two-dose CoronaVac vaccination [5][6][7] . Health professionals were the rst prioritized group receiving two-dose CoronaVac in Thailand. Adenoviral-vectored vaccine, imported from Korea in March 2021 and later produced from the Siam Bioscience company (Nonthaburi, Thailand) was initially prioritized for the elderly above 60 years of age. Preliminary studies have found the two-dose  CoronaVac regimen induced a lower, but acceptable, immune response compared to the two-dose   AZD1222 regimen; however, there was a signi cantly shorter waiting period between CoronaVac doses.  Thailand started vaccinations with the inactivated CoronaVac on 28 February, 2021, and AZD1222  vaccine on 16 March, 2021, with a 3-to 4-week and 10-week waiting period between doses, respectively. A 10-week waiting period was established based on recommendations of the Thailand FDA and e cacy studies identifying that a waiting period of <6 weeks resulted in lower immune stimulation than a period of 10 weeks 8 . The CoronaVac vaccine was associated with a rare focal neurological syndrome characterized by numbness, or sometimes weakness, in the limbs 9 . Although this self-limited adverse event is rare, individuals experiencing this side effect sought another regimen for their second shot, which was AZD1222.
It is possible to mix and match vaccines in speci c situations such as a vaccine shortage or for adverse reactions following vaccine administration. This study aims to assess the immunogenicity of heterologous prime/booster inactivated COVID-19 vaccine and the adenoviral-vectored vaccine currently available in Thailand to provide preliminary data on their immunogenicity.

Demographic data
All participants were Thai; however, the mean age of participants in the homologous AZD1222 vaccine cohort was higher than that in the homologous CoronaVac and heterologous CoronaVac/AZD1222 because of vaccine prioritization during initial implementation (Table 1). Unlike the homologous vaccination cohorts, there were variations in intervals between the rst and second dose vaccinations among the heterologous CoronaVac/AZD1222 vaccinees (median: 26 days, IQR: 21-32 days). We analyzed the immunogenicity data of the heterologous CoronaVac/AZD1222 vaccinees in two sets. The rst set included all available data. The second set included heterologous CoronaVac/AZD1222 vaccinees who received vaccines 14-35 days apart and had their blood collected between 14 and 35 days post-second dose vaccination (Supplementary Fig. 1  CoronaVac followed by AZD1222 vaccine induced higher levels of spike receptor-binding domain-speci c IgG than two-dose CoronaVac and AZD1222 vaccines (P<0.001) ( Fig 1A).
The sVNT was based on antibody-mediated blockage of ACE2-spike protein-protein interaction. The percent inhibition represented the ability of sera from vaccinated individuals to block the interaction between the ACE2 receptor protein and the SARS-CoV-2 RBD, which re ects the neutralizing activity. Our results showed that CoronaVac/AZD1222 vaccine recipients had higher neutralizing activities against the original Wuhan and the B.1.351 strain than the recipients of two-dose CoronaVac and AZD1222 (P<0.01) ( Fig. 1B and D, respectively). In addition, the CoronaVac/AZD1222 vaccine recipients had higher neutralizing activities against the B.1.1.7 strain than the recipients of two-dose CoronaVac but were comparable to those in the recipients of two-dose AZD1222 (Fig. 2C) (P<0.01).

Lower neutralizing activities against SARS-CoV-2 variants
As compared to the percentage of neutralizing activity against the original Wuhan strain, we found that neutralizing activities of sera against B.1.1.7 and B.1.351 were approximately 1.25-1.5 times and 1.5-1.97 times lower than that against the original strain (Fig. 2). The decline of neutralizing activity against the two variants (1.5 x for B.1.1.7 and 1.97 x for B.1.351) was highest among those who received the homologous CoronaVac/CoronaVac vaccine (P<0.01) (Fig 2A). Although there was a reduction in neutralizing activities of sera against B.1.1.7 (1.25x) and B.1.351 (1.5x) in heterologous CoronaVac/AZD1222 vaccinees (Fig. 2B), a higher level than was observed compared to the homologous CoronaVac/CoronaVac counterparts. A similar trend was found between the individuals who received the homologous AZD1222/AZD1222 and heterologous CoronaVac/AZD1222 vaccine (Fig. 2C).

Discussion
As of August 2021, the Department of Disease Control, Ministry of Public Health, Thailand, has implemented policies designating the rst vaccine dose to be the CoronaVac vaccine and the second dose to be the AZD1222 vaccine, with more than 200,000 individuals getting vaccinated under this policy. The reasons are the shortage of AZD1222 and the decreased effectiveness of the two-dose CoronaVac vaccine against the variants of concern that are circulating in Thailand.
Our study enrolled recipients who had received heterologous prime/booster inactivated COVID-19 vaccine and adenoviral-vectored vaccine and sought antibody testing following vaccination, and they were compared with homologous vaccine recipients. Our immunogenicity data has shown that RBD-speci c IgG was detected in 100% of heterologous CoronaVac/AZD1222 recipients after the second dose, with a higher GMT than those elicited by the two-dose CoronaVac and AZD1222 vaccines. In addition, the CoronaVac/AZD1222 vaccine recipients had higher amounts of neutralizing activities against the original Wuhan and the B.1.351 strain than did the recipients of two-dose CoronaVac and AZD1222. Although the extent of the e cacy of the heterologous regimen has not been studied, the comparatively high level of immunogenicity compared to the homologous AZD1222 regimen supports its use as an alternative schedule, with the added bene t of a shorter waiting period between doses.
Researchers are investigating to determine the immune correlates of protection to use as surrogate endpoints for vaccine e cacy. In a recent preprint, protection against SARS-CoV-2 challenge in vaccinated non-human primates strongly correlated with levels of anti-S antibody binding and neutralizing activity 10 . Because of its ability to elicit a high RBD-speci c IgG and neutralizing activity following two-dose vaccination, a heterologous prime/booster regimen with CoronaVac/AZD1222 may provide more protection than the homologous CoronaVac regimen. Regarding the reactogenicity of the heterologous regimen, 1,100 individuals received the heterologous CoronaVac/AZD1222 as reported by in the heterologous CoronaVac/AZD1222 regimen.
The high immunogenicity pro le of the heterologous prime/booster regimen in this study is congruent with conclusions of the Com-COV study, which also found an increase in anti-spike RBD-speci c IgG and neutralizing titers compared to the heterologous regimen 1 . In the Com-COV study, the researchers investigated combinations of the P zer mRNA vaccine (BNT162b2) and the AZD1222 vaccines available in the United Kingdom. Two-dose AZD1222 administered 6 weeks apart elicited a lower immune response than when given 10-12 weeks apart. The underlying mechanism is likely due to the host anti-adenoviral antibodies elicited by the rst vaccination preventing the virus in the second vaccine dose to enter the cells when the second dose is given sooner than 10-12 weeks. Nevertheless, the P zer mRNA vaccine (BNT162b2) given as a second dose in AZD1222-primed individuals has been shown to induce a higher response than that of the AZD1222 given as a second dose. Heterologous vaccination regimens have been previously examined with experimental vaccines for HIV 14 , malaria 15 , and Ebola 16 , a precedent for this regimen's e cacy; however, the mechanism for increased immunogenicity from mixing CoronaVac/AZD1222 has yet to be elucidated.
There are a few noteworthy limitations to the current study. Because our study participants were recruited in a real-world setting before the Ministry of Public Health launched the vaccine recommendation, the schedule of heterologous vaccination did not follow the recently released guideline stating that two doses should be given 28 days apart. In this study, approximately 80.5% of individuals received the rst and second dose of the heterologous regimen at an interval between 14 and 35 days. Secondly, the timing of collecting blood samples after the second dose in the heterologous vaccine group was not the same as for the homologous vaccine group. This caveat has not statistically affected the conclusion but nonetheless should be considered. Third, the age demographic disparity was present between different vaccination regimen groups. The inherent nature of Thailand's vaccination policy, which prioritizes vaccination with the AZD1222 vaccine in elderly people, consequently led to a higher average age for the homologous AZD1222 regimen cohort. The increased average age in the homologous AZD1222 regimen cohort can lead to a lower immune response as also demonstrated in a study of immunogenicity of an RNA vaccine 17 . The age-related lower immune response is likely due to the "immunosenescence" phenomenon as a result of increases in terminally differentiated memory cell populations, lymph node brosis, and altered cytokine production among the elderly 18 . Lastly, cell-mediated immunity was not explored in this study.
In conclusion, a heterologous prime/booster regimen demonstrated a strong immune response. Further studies are underway to determine the reactogenicity and immunogenicity of a heterologous prime/booster regimen in clinical trials against other SARS-CoV-2 variants. With an acceptable immunogenicity pro le, here is the rst report to show that a heterologous prime/booster regimen with CoronaVac/AZD1222 would provide greater exibility for countries experiencing supply di culties and individuals with adverse events following the rst dose CoronaVac vaccination.
Materials And Methods

Study design
We performed a cross-sectional study in which leftover sera samples from participants seeking antibody testing following vaccination at the Center of Excellence in Clinical Virology, Department of Pediatrics, Faculty of Medicine, Chulalongkorn University between April and July 2021 were further analyzed. Only samples from participants who received heterologous prime/booster inactivated COVID-19 vaccine and adenoviral-vectored vaccine were used. The study protocol was approved by the Research Ethics Committee of the Faculty of Medicine, Chulalongkorn University (IRB number 491/64) and this trial is registered with Thai Clinical Trials Registry (TCTR20210628005). The committee waived the requirement for consent because the samples used were de-identi ed and anonymous.
We also included two comparison groups vaccinated with the homologous CoronaVac and AZD1222 regimen (IRB no. 192/64, TCTR20210319003) in the analysis. Participants in the comparison groups received CoronaVac or AZD1222 vaccines at the Banphaeo General Hospital, Samutsakorn Province, Thailand, between March and June 2021. Informed consent was obtained during the second-dose vaccination visit. Participants who consented to blood sampling at 21-35 days after full vaccination were scheduled for an extra blood sampling visit at the Banphaeo General Hospital, Samutsakorn Province.

Study vaccine
CoronaVac is an inactivated virus vaccine created from African green monkey kidney cells (Vero cells) that have been inoculated with SARS-CoV-2 (CZ02 strain). The virus was inactivated with β-propiolactone and nally absorbed onto aluminum hydroxide. Each vial contains 0.5 mL with 600 SU (equal to 3 µg) of inactivated SARS-CoV-2 virus as antigen 19 . Chimpanzee adenovirus Oxford 1 (ChAdOx1)-vectored vaccine (AZD1222) from Oxford/AstraZeneca is a non-replicating viral vector vaccine that stimulates an immune response against the coronavirus spike protein. One dose (0.5 mL) contains no less than 2.5 × 10 8 infectious units of chimpanzee adenovirus encoding the SARS-CoV-2 spike glycoprotein (ChAdOx1-S) 20 . The CoronaVac vaccine was given as a two-dose regimen administered 21 days apart and prioritized for adults aged 18-59 years. AZD1222 was given as a two-dose regimen administered 10 weeks apart for adults ≥ 18 years and prioritized for the elderly above 60 years of age.

Blood samples
Venous blood samples (5 ml) samples were collected at various time points after second dose vaccination in participants who received heterologous prime/booster inactivated COVID-19 vaccine and adenoviral-vectored vaccine. For participants who received homologous CoronaVac and AZD1222 vaccines, venous blood samples (5 ml) were collected at 21-35 days after the second dose vaccination.

Laboratory testing
All sera samples were tested for SARS-CoV-2 spike RBD IgG by SARS-CoV-2 IgG II Quant assay (Abbott Diagnostics, Sligo, Ireland). This assay quanti es speci c IgG against the RBD of the spike protein by using a chemiluminescent microparticle immunoassay (CMIA). The result was expressed as arbitrary units per milliliter (AU/mL) and the positive cut-off level was ≥ 50 AU/mL (upper limit: 40,000 AU/mL).
Results of this assay had a good correlation with neutralizing antibodies 21 .
The neutralizing activity in a subset of serum samples was tested against the original Wuhan strain and variants of concern, B.  22 . Then 100 µL of the sample mixture was subsequently added to a capture plate with pre-coated h-angiotensin-converting enzyme 2 (ACE2) protein and incubated for 15 min at 37°C. After a washing step, 100 µL of 3,3′,5,5′-tetramethylbenzidine (TMB) solution was added and the plate incubated in the dark for 15 min at 20-25°C. Then 50 µL of a stop solution was added to quench the reaction and the sample was read immediately at 450 nm. The percent inhibition of a sample was calculated as (1average optical density (OD) of sample/average OD of negative control) × 100%. Greater than or equal to 20% inhibition was considered indicative of the presence of neutralizing antibodies.

Statistical analysis
Baseline characteristics are reported as means and standard deviations (SD). IgG S1/RBD antibodies titers are presented as geometric mean titer (GMT) and 95% con dence interval (CI). The sVNT antibodies titers are presented as median with interquartile range. The difference of spike RBD-speci c IgG and percentage of inhibition between groups was calculated using the Mann-Whitney U test. The Wilcoxon site rank test was used to compare the different percentages of inhibition between wild type and variants.
Statistical analysis was done using Prism 8.0 (GraphPad, San Diego, CA). A P-value < 0.05 was considered statistically signi cant.