Pre-existing humoral immunity and CD4⁺ T cell response correlate with cross-reactivity against SARS-CoV-2 Omicron subvariants after heterologous prime-boost vaccination

Abstract

Background

Information regarding the heterologous prime-boost COVID vaccination has been fully elucidated. The study aimed to evaluate both humoral, cellular immunity and cross-reactivity against variants after heterologous vaccination.

Methods

We recruited healthcare workers previously primed with Oxford/AstraZeneca ChAdOx1-S vaccines and boosted with Moderna mRNA-1273 vaccine boost to evaluate the immunological response. Assay used: anti-spike RBD antibody, surrogate virus neutralizing antibody and interferon-γ release assay.

Results

All participants exhibited higher humoral and cellular immune response after the booster regardless of prior antibody level, but those with higher antibody level demonstrated stronger booster response, especially against omicron BA.1 and BA.2 variants. The pre-booster IFN-γ release by CD4⁺ T cells correlates with post-booster neutralizing antibody against BA.1 and BA.2 variant after adjustment with age and gender.

Conclusions

A heterologous mRNA boost is highly immunogenic. The pre-existing neutralizing antibody level and CD4⁺ T cells response correlates with post-booster neutralization reactivity against the Omicron variant.

1. Introduction

The spread of SARS-CoV-2 not only changed the lifestyle of most people, but also caused considerable social and economic damage, leading to significant loss of life. Currently, there are several types of vaccines being commonly administered to provide active immunization, including adenovirus vector vaccines such as the ChAdOx1 nCoV-19 COVID-19 vaccine (AstraZeneca, AZ), and mRNA vaccines such as the BNT162b2 COVID-19 vaccine (Pfizer-BioNTech, BNT) and mRNA-1273 COVID-19 vaccine (Moderna) [1,2]. Although these vaccines have demonstrated significant neutralizing antibody titers measured in blood 2–4 weeks post-inoculation and successfully suppressed the epidemic in some countries, a waning in antibody production 4–6 months after primary immunization has been noticed and was responsible for a substantial number of breakthrough infections caused by the emerging variants of concern (VOCs) [3,4]. In clinical studies, researchers found that two doses of the AZ or BNT vaccine worked well against COVID-19-related hospitalization and death, but there is a certain degree of waning in vaccine effectiveness, which subsequently leads to the universal recommendation for booster COVID vaccination [5]. Due to the limited availability of COVID-19 vaccines and to establish a better immune response to fight against emerging SARS-CoV-2 variants, heterologous prime-boost immunization has been applied in many countries [6,7]. Studies have shown that heterologous booster vaccination results in effective immune maturation and is able to overcome waning immunity issues, providing protection against Omicron BA.1 and BA.2 variants [8,9]. These evidences support the idea that this new heterologous strategy offers superior immunity compared with homologous booster vaccination [[10], [11], [12]].

Vaccination can trigger the body's immune response to protect against certain diseases. That response can be divided into two parts: humoral immunity and cellular immunity, which elicit strong antibody responses and robust CD4⁺ and CD8⁺ T cell responses respectively [[13], [14], [15]]. Past research primarily focused on antibody production, which has some shortcomings, as antibody-based immunity can be delayed or can decay relatively quickly [15,16]. On the other hand, T-cell response has been shown to be more enduring [17,18]. Moreover, studies have also shown that CD4⁺ and CD8⁺ T cell responses were not influenced when encountering different variants [19,20]. This evidence further confirms the importance of cell-mediated immunity measurements for evaluating vaccine efficacy.

In this study, we recruited healthcare workers (HCWs) previously primed with two doses of Oxford/AstraZeneca ChAdOx1-S vaccine to evaluate both humoral and cellular immunity before and after they received a heterologous booster dose of Moderna mRNA-1273 vaccine. Here, we applied a commercialized interferon-γ release assay (IGRA), QuantiFERON SARS-CoV-2 (Qiagen), to evaluate cellular immune response after booster dose administration. We also tested the cross-reactivity of neutralizing antibodies against Delta, Omicron BA.1 and BA.2 variants before and after receipt of the third dose of vaccine.

2. Materials and methods

2.1. Patients and samples

Healthcare workers in National Cheng Kung University Hospital who previously received two doses of Oxford/AstraZeneca ChAdOx1-S vaccine as primary immunization at least five months previously were recruited for this study. Participants provided written informed consent upon recruitment. The protocol of this study was reviewed and approved by the Institutional Review Board (IRB) of National Cheng Kung University Hospital (NCKU) (No. A-BR-110-051). One booster dose of Moderna mRNA-1273 vaccine was given to these participants with a minimum dose interval of five months. All participants had blood samples taken before booster dose and one month after the booster dose. Participants were followed up for up to 3 months after the booster and monitored for breakthrough infection, defined as acute respiratory illness with microbiological confirmation through rapid antigen test or PCR reaction.

2.2. Surrogate SARS-CoV-2 neutralization test –Wildtype, Delta, Omicron BA.1 and BA.2 variants

The levels of neutralizing antibodies (NAbs) were determined using a cPass SARS-CoV-2 Neutralization Antibody Detection Kit (GenScript, USA) and the process was performed according to the manufacturer's instruction [21]. Samples and controls were diluted with sample dilution buffer (1:10) and mixed with horseradish peroxidase conjugated recombinant SARS-CoV-2 receptor binding domain (RBD-HRP) solution at a ratio of 1:1 to create neutralization reaction mixtures. After a 30-min incubation at 37 °C, the mixtures were added to the corresponding well of an ACE2-coated 96-well plate. During incubation, the unbound RBD-HRP as well as non -neutralizing antibodies binding were captured on the plate and washed away during the washing step. Tetramethylbenzidine (TMB) solution was subsequently added to each well and the plate was incubated for 15 min in the dark. The reaction was quenched by adding Stop Solution, which turned the colour of the mixture from blue to yellow. The absorbance of the plate was immediately read using a microtiter plate reader at 450 nm. The percent signal inhibition was calculated as follows:

$$\mathit{Percent\; Signal\; Inhibition}=\left(1-\frac{\mathit{OD}\mathit{value\; of\; Sample}}{\mathit{OD}\mathit{value\; of\; Negative\; Control}}\right)\times 100\%$$

The results were interpreted as positive if the percent signal inhibition was >30%, which was the cutoff value. The detection of NAbs against wildtype, Delta, and Omicron BA.1 variants all followed a similar protocol. For wildtype NAbs detection, an additional conversion formula was used to transfer percent inhibition into concentration values in terms of IU/mL [22].

We used an Anti-SARS-CoV-2 (BA.2) Neutralizing Antibody Titer Serologic Assay Kit (ACROBiosystems, USA) [23] as the neutralizing antibody against Omicron BA.2 variant detection. The process was similar to that mentioned above, however the percent signal inhibition was calculated slightly differently as follows:

$$\mathit{Percent\; Inhibition}=\left(1-\frac{{\mathit{OD}}_{450\mathit{nm}}-{\mathit{OD}}_{630\mathit{nm}}\mathit{of\; Sample}}{{\mathit{OD}}_{450\mathit{nm}}-{\mathit{OD}}_{630\mathit{nm}}\mathit{of\; Negative\; Control}}\right)\times 100\%$$

The results were interpreted as positive if the percent signal inhibition was >20%, which was the cutoff value.

2.3. SARS-CoV-2 anti-spike protein antibody measurement (Roche Elecsys)

The anti-SARS-CoV-2 antibodies were determined using a Roche Elecsys anti-SARS-CoV-2 S test (Roche Diagnostics, Switzerland, an electrochemiluminescence immunoassay intended for qualitative detection of antibodies (including IgG) to the SARS-CoV-2 spike (S) protein receptor binding domain (RBD) in human serum and plasma. The assay was performed using a cobas e601 analyzer. The principle of the test is a double-antigen sandwich assay, and the duration of the total assay was 18 min. The procedure was followed as per the manufacturer's instructions. We incubated 20 μL of sample with biotinylated SARS-CoV-2 S-RBD-specific recombinant antigen and SARS-CoV-2 S-RBD-specific recombinant antigen labeled with a ruthenium complex to form a sandwich complex. After adding streptavidin-coated microparticles, the complex became bound to the solid phase. The mixture was then aspirated into the measuring cell where the microparticles were captured on the surface of the electrode. Unbound substances were then removed. Applying voltage induced chemiluminescent emission which could then be measured by a photomultiplier. The results were determined via a calibration curve.

2.4. QuantiFERON SARS-CoV-2

Cellular immunity was evaluated using a QuantiFERON SARS-CoV-2 test (QIAGEN, Germany) [[24], [25], [26]]. The protocol can be divided into two parts: 1) blood collection and T cell stimulation; and, 2) IFN-γ measurement. Blood samples were initially collected and placed in four different QFN SARS-CoV-2 blood collection tubes (BCTs) (negative control, positive control, antigen 1, antigen 2). Antigen 1 (Ag1) is an epitope of CD4⁺ T cells derived from the S1 subunit and antigen 2 (Ag2) is an epitope of CD4⁺ and CD8⁺ T cells derived from the S1 and S2 subunits. The positive control tube contains phytohemagglutinin and negative control tube contains no peptide [27]. BCTs were shaken thoroughly to dissolve the antigens on the wall and placed into a 37 °C incubator for 16–24 h. After incubation, BCTs were centrifuged for 15 min at 2000–3000 RCF(g) and the plasma was collected for IFN-γ detection. The second phase of the process was similar to ELISA. The working strength conjugate was added to each well followed by samples and standards. After a 2-h incubation, each well was aspirated and washed for 6 cycles. The Enzyme Substrate Solution was then added to each well. Following 30 min of incubation, Enzyme Stopping Solution was added to end the reaction. The optical density was measured by a microplate reader, and the final results of concentration (IU/mL) were obtained using a QuantiFERON R&D Analysis Software.

2.5. Statistical analysis

The statistical analysis was performed using GraphPad Prism and R program. Descriptive analyses of numerical variables were presented as median and interquartile range (IQR), and categorical variables were presented as frequency and percentage. Differences in continuous variables between groups were assessed with Mann-Whitney test or Wilcoxon signed-rank test. Even though the sVNT of 30% is arbitrarily set as the cut-off value for adequate neutralizing activity against SARS-CoV-2, a value <30% still represents neutralizing antibody level and correlates with anti-SARS-CoV-2 spike antibody even though it may not be enough to prevent most real life infection [28]. Therefore, we used the raw data of sVNT for further correlation analysis (Spearman correlation). Value differences were considered statistically significant if the p value was <0.05.

3. Results

A total of sixty-three participants (19 males; 44 females) aged from 24 to 61 years were recruited for this study. The baseline clinical characteristics and serological data before booster are shown in Table 1 . These participants were further categorized into high and low NAb (HNT, n = 20 vs. LNT, n = 43) based on pre-booster NAb with an inhibition value of 70%. There was no significant difference in the interval between blood sampling and the last COVID vaccination between these two groups. The pre-booster humoral immunity in LNT and HNT groups was significantly different, either in binding or neutralizing antibody (GMT of anti-SARS-CoV-2 spike antibody, 131.8 vs. 735.6 U/ml; GMT of NAb against wildtype, 47.2 vs. 762.2 IU/ml, all p value <0.0001). The NAb against wildtype, Delta, Omicron BA.1 and BA.2 before and after booster vaccination is shown in Fig. 1 (A, B, C & D). Before booster receipt, the HNT group presented higher neutralizing antibody against wildtype, and Delta variant than LNT group, but no similar difference was found against Omicron BA.1 and BA.2 variants. The participants, either in HNT or LNT, all demonstrated enhanced antibody response against all variants after booster. The neutralizing activity against wildtype and Delta variant was significantly boosted, and the difference between HNT and LNT diminished after booster vaccination. In contrast to relatively high neutralizing activity to wildtype and Delta variant before booster receipt, only 77.8% of the participants presented adequate neutralizing activity against Omicron BA.1, and 87.1% against BA.2 (inhibition>20%). Although the neutralizing activities against Omicron BA.1 and BA.2 were also enhanced after booster receipt in both groups, the HNT group still presented higher neutralizing activity against Omicron BA.1 and BA.2 than the LNT group.

Table 1.

The clinical characteristics and serological data of participants in this study.

Characteristics	All participants (N = 63)	HNT group (N = 20)	LNT group (N = 43)
Anti-SARS-CoV-2 spike antibody before booster, GMT (95% CI), U/mL	227.52 (175.11, 295.60)	735.63⁎ (521.70, 1037.28)	131.82⁎ (108.60, 160.00)
Neutralizing antibody before booster, a GMT (95% CI), IU/mL	114.19 (77.76, 167.71)	762.24⁎⁎ (534.01, 1088.03)	47.23⁎⁎ (36.37, 61.31)
Age, median (IQR), years	39 (32, 45)	36.5 (29.75, 45)	42 (32, 45)
Age distribution, no. (%)
20–29 years	13 (20.6)	5 (25.0)	8 (18.6)
30–39 years	21 (33.3)	8 (40.0)	13 (30.2)
40–49 years	19 (30.2)	4 (20.0)	15 (34.9)
50–59 years	8 (12.7)	3 (15.0)	5 (11.6)
≧60 years	2 (3.2)	0 (0.0)	2 (4.7)
Sex, no. (%)
Male	19 (30.2)	5 (25.0)	14 (32.6)
Female	44 (69.8)	15 (75.0)	29 (67.4)
Underlying disease, no. (%)	0 (0.0)	0 (0.0)	0 (0.0)

^aSurrogate viral neutralizing antibody against SARS-CoV-2 wildtype.

^⁎The difference in anti-SARS-CoV-2 spike antibody levels before booster dose between HNT and LNT groups, p value <0.0001.

^⁎⁎The difference in neutralizing antibody levels against wildtype before booster dose between HNT and LNT group, p value <0.0001.

Fig. 1.

Neutralizing antibodies to wildtype SARS-CoV-2 or Delta, BA.1 and BA.2 variant before and after booster dose. The neutralizing antibodies against (A) wildtype (n = 63), (B) Delta (n = 63), (C) BA.1 (n = 63) and (D) BA.2 (n = 62) for participants in the HNT and LNT groups. The blue dots represent the HNT group, and the orange dots represent the LNT group. Lines and error bars indicate mean and standard deviation. The neutralizing antibodies for each group were measured before booster vaccination and one month after booster vaccination. *** p value <0.001, ** p value <0.01, * p value <0.05, ns indicates no significant difference (p value >0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The neutralizing activity against wildtype versus other variants before and after booster receipt in each blood sample is demonstrated in Fig. 2 . The neutralization activity against Delta variant was relatively comparable to wildtype activity either before or after booster receipt. The neutralization activity against Omicron BA.1 and BA.2 was much reduced compared to that against wildtype (Fig. 2 B, C, E, F) with various degrees of reduction. Even after booster receipt, the neuralization activity against Omicron BA.1 and BA.2 was still quite diverse among individuals, with inhibition ranging from 0% to 95%. The neutralizing activity against wildtype correlated well with the activity against Delta variant (r = 0.85, p value <0.0001), but not with Omicron BA.1 and BA.2. (Supplementary Fig. 1).

Fig. 2.

Comparison of neutralizing antibodies against wildtype and other variants of concern (VOCs). (A)The neutralizing antibodies against wildtype and Delta variant before booster dose (n = 63). (B) The neutralizing antibodies against wildtype and BA.1 variant before booster dose (n = 63). (C) The neutralizing antibodies against wildtype and BA.2 variant before booster dose (n = 63). (D) The neutralizing antibodies against wildtype and Delta variant after booster dose (n = 63). (E) The neutralizing antibodies against wildtype and BA.1 variant after booster dose (n = 63). (F) The neutralizing antibodies against wildtype and BA.2 variant after booster dose (n = 62). The blue symbols represent the HNT group, and the orange symbols represent the LNT group. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Thirty-seven participants (15 from the HNT and 22 from the LNT group) participated in a QuantiFERON test before and after booster vaccination (Fig. 3 ). The release of IFN-γ in response to Ag1 (CD4⁺ T cells response) and Ag2 (both CD4⁺ and CD8⁺ T cell response) was relatively low before booster receipt, but significantly increased after booster receipt (p value <0.001), indicating enhanced cellular immunity. The IFN-γ induced by Ag2 (both CD4⁺ and CD8⁺ T cell response) was markedly higher than that induced by Ag1 (CD4⁺ T cell response) either before or after booster vaccination (Fig. 3A). The increase in INF- γ was much more prominent in the HNT group than in the LNT group (Ag1: p value = 0.006; Ag2: p value = 0.045). This implies that the pre-booster humoral immunity was relatively parallel to the post-booster enhanced cellular immunity (Fig. 3B, C).

Fig. 3.

The IFN-γ releasing assay (QuantiFERON ELISA) in HNT and LNT group before and after heterologous booster. The IFN-γ measured by QuantiFERON ELISA in response to Ag1 and Ag 2 in (A) all participants, (B) all participants by groups, (C) increase of INF-γ release by group. The blue symbols represent Ag1, and the orange symbols represent Ag2. Lines and error bars indicate mean and standard deviation. The increase in INF-γ means that the level of INF-γ after booster receipt reduces the level before booster, expressed by median (IQR). *** p value <0.001, ** p value <0.01, * p value <0.05, and indicates no significant difference (p value >0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

To explore the association between humoral and cellular immunity, we analyzed the correlation between the level of IFN-γ released before booster and the post-booster NAb against wildtype, Delta, Omicron BA.1, and BA.2 variants (Fig. 4 , and Supplementary Fig. 2). The level of IFN-γ in response to Ag1 correlated with post-booster NAb against BA.1 (r = 0.408, p value = 0.012), and BA.2 variant (r = 0.482, p value = 0.003). The correlation remains significant even after age and gender adjustment (Supplementary Table 1). In contrast, no significant correlation was found between the Ag2-induced INF-γ release and post-booster NAb against the Omicron BA.1 or BA.2 variant. To measure the level of IFN-γ more precisely, we rechecked the INF-γ level in the supernatants using quantitative ELISA (Supplementary Fig. 3A, B). The INF-γ level determined by quantitative ELISA correlated quite well with QuantiFERON ELISA results. The level of IFN-γ using QuantiFERON assay above 0.15 IU/mL was considered as adequate cellular immunity against SARS-CoV-2 infection in the literature [29,30]. We measured the corresponding optimal IFN-γ level (pg/mL) under the ROC curve using this cut-off value (0.15 IU/mL) (Supplementary Fig. 3C, D). The optimal IFN-γ value from quantitative ELISA corresponds to 27.4 pg/mL in response to Ag1 with an area under the curve (AUC) of 0.92, and 27.3 pg/mL in response to Ag2 with AUC of 0.884.

Fig. 4.

The correlation of pre-booster cellular immunity with post-booster humoral response. The correlation of pre-booster cellular immunity by IFN-γ releasing assay (QuantiFERON ELISA) in response to Ag1 with post-booster neutralizing antibody against BA.1 (A) and BA.2 (C). The correlation of pre-booster cellular immunity by IFN-γ releasing assay (QuantiFERON ELISA) in response to Ag2 with post-booster neutralizing antibody against BA.1 (B) and BA.2 (D). r indicates the Spearman correlation coefficient, and the grey area indicates 95% confidence interval. The black line indicates fitted regression line.

In the five-month follow-up visit, 13 out of 63 participants had become infected with SARS-CoV-2, with a breakthrough infection rate of 20.6%. Most cases clustered in the first wave of the COVID-19 outbreak in Taiwan, which was predominantly caused by Omicron BA.2. The breakthrough infection rate was higher in the LNT than in the HNT group (25.6% vs. 10.0%), which was parallel to the serological response after the booster.

4. Discussion

In this study, we demonstrated enhanced humoral and cellular immunity to SARS-CoV-2 in healthcare workers who were ChAdOx1 nCoV-19 primed, and mRNA-1273 vaccine boosted. Regardless of the prior humoral immunity status, most participants exhibited higher binding and neutralizing antibody levels after the booster. The vaccine-induced neutralizing activity was quite attenuated toward Omicron BA.1 and BA.2 variants compared to wildtype and Delta variant. The difference between neutralizing activity against wildtype and Omicron BA.1 or BA.2 was diminished but not completely restored after the mRNA-1273 vaccine booster. In addition, enhancement of humoral and cellular immunity after booster receipt correlated with pre-booster NAb level and CD4⁺ T cell response, implying the possibility of using these two biomarkers for predicting booster response. Further, we exploited the IGRA's optimal INF- γ level for adequate cellular immunity using quantitative IFN-γ ELISA.

A booster after primary COVID vaccination was introduced early during the 2nd wave of the pandemic due to waning antibody levels over time, the emergence of variants of concern (VOCs), and attenuated neutralization toward VOCs. The rise and spread of Omicron variants further increased the need for a 2nd booster in vulnerable populations. Even though the “immune escape” character of the Omicron variants leads to increasing breakthrough infections, vaccination prevents severe disease and reduces hospitalization. Our data shows that the booster notably increased antibody presence and induced more broadly neutralizing antibodies with higher cross-reactivity. The B cell affinity maturation to the RBD is an ongoing process; thus, protection against new variants of the RBD will rise over time. A study conducted by Sokal et al., suggested that antibody affinity maturation of the B memory pool after completing the primary vaccination series continues for at least six months [31]. Moreover, the boosting facilitates antibody affinity maturation, resulting in superior immune kinetics, breadth, and durability of neutralizing antibody, which confers resilience to viral escape mutations and is crucial in the defense against rapidly spreading Omicron subvariants [32,33]. In addition, the booster could also enhance antibody-dependent neutrophil/monocyte phagocytosis, complement activation, and natural killer cell activation [34].

Previous trials including 1862 participants showed that heterologous administration of BNT162b2 (Pfizer-BioNTech) in ChAdOx1 nCoV-19-primed participants induced immunogenicity noninferior to homologous BNT administration (both prime and booster were BNT vaccines, i.e., BNT/BNT) with tolerable reactogenicity and higher T cell responses. Compared with homologous ChAdOx1 nCoV-19 vaccination (ChAd/ChAd), heterologous ChAd/BNT was found to elicit higher immunogenicity (ChAd/BNT vs. ChAd/ChAd) antibody titer ratio: 9.2 [35]. A clinical trial demonstrated the interchangeability of mRNA vaccines, BNT162b2 and Moderna 1273 in the primary vaccination regardless of vaccine sequence. Moreover, the Moderna mRNA1273 vaccine could produce even better immune response as a second dose [36]. The results of the current study again validate the applicability and immunogenicity of a heterologous, adenovirus vector vaccine prime, and mRNA vaccine boost regimen and proves that these two SARS-CoV-2 vaccines work together effectively. The mRNA booster not only elicits strong humoral and cellular immune response in the adenovirus vaccine primed vaccinees but also the inactive vaccine, CoronaVac (Sinovac) or protein subunit vaccine, MVC-COV1901 (MVC) primed vaccines [37,38]. These results support the value and use of mRNA vaccines as the booster to enhance post-primary immune response regardless of prior vaccine platforms. This might provide flexibility in vaccination strategy-making and more effectively maximize the gains of vaccination with reduced cost and workflow, especially in low-income or resource-poor countries.

The IGRA is widely used to measure cellular immunity post-COVID vaccination or infection. The Ag1 tube contained 13-mer epitopes derived from the S1 subunit of the spike protein, which is recognized mainly by CD4⁺ T cells, while the Ag2 tube contained 13- and 8-mer epitopes derived from the whole spike protein, which is recognized by both CD4⁺ and CD8⁺ T cells. The Ag2 evoked higher INF-γ release than Ag1, probably due to the mixed effect from both CD4⁺ and CD8⁺ T cells [39]. However, only Ag1-induced INF-γ release, mainly CD4⁺ T cells response, which correlates with the post-booster neutralizing antibody level – a result that is consistent with the findings by Sahin et al. [15] In a previous study evaluating the decay and maintenance of long-term SARS-CoV-2 neutralizing antibodies in infected people, Wang, Z. et al. discovered that virus-specific CD4⁺ T cells were associated with long-term persistence of neutralizing antibodies [40]. This was accompanied by a higher level of CXCR3 ligands such as CXCL9 and CXCL10, a higher frequency of cTfh1 (circulating T follicular helper), and lower levels of cTfh2 and cTfh17. Our findings agree with the previous report and highlights the importance of CD4⁺ T cells and their potential implication in evaluating both infection and vaccine-induced immunity.

This study was conducted between Dec. 2021 and Mar. 2022 before the SARS-CoV-2 spread out endemically in Taiwan, when most people in Taiwan were infection naive. The first wave of the COVID outbreak in Taiwan occurred between Apr. and Jul. 2022, and was predominantly caused by Omicron BA.2. Following this outbreak, the health authority rapidly authorized the use of 2nd boosters for healthcare workers and vulnerable populations, and some of the participants in this study got their 2nd boosters after Jun 2022. Due to the high heterogenicity of this cohort, it is challenging to explore the longevity of neutralizing and cellular response merely induced by the mRNA boost. Following the release of the bivalent formulations of the Moderna and Pfizer-BioNTech COVID-19 vaccines authorized by the FDA on Sep. 2022, we believe that the use of bivalent vaccines will become more common and might ultimately replace monovalent vaccine administration in defense against evolving variants.

5. Conclusion

To summarize, we demonstrated that heterologous mRNA booster induced adequate immunogenicity, including both humoral and cellular response in adenovirus vector vaccine primed vaccinees. Also, pre-existing neutralizing antibody level and CD4⁺ T cell response correlated with post-booster cross-reactivity against Omicron variants and cellular immunity. Although the neutralizing activity against Omicron subvariants was elevated after booster receipt, the vaccine-induced protection was sub-optimal for preventing breakthrough infections. These findings highlight the necessity for booster vaccination and identify the candidate immunological biomarkers for evaluating humoral and cellular immunity after COVID vaccination.

Author contributions

CFS, CMS, and CCS are the guarantor of the content of this manuscript, had full access to all the data in the study and take responsibility for the integrity of the data. CFS, CMS and CCS contributed to the study conception and design, collection of data. CFS, YCF, and CMS organized the database and collect the clinical information. BT, PT, and NL organized and performed the experiment. CFS and YF wrote and revised the draft. TH, PC, NL, WK, CL and CCS contributed to the data interpretation, critical review of the manuscript. All author contributed to manuscript revision, read and approved the submitted version.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.