Europe PMC

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

Antibody and T cell responses against wild-type and Omicron SARS-CoV-2 after the third dose of BNT162b2 in healthy adolescents

Show less

Author information

ORCIDs linked to this article

Preprint from Research Square, 25 Aug 2022
https://doi.org/10.21203/rs.3.rs-1961385/v1 PPR: PPR536221 

Preprint v1

This article is in the Europe PMC Open access subset. Refer to the copyright information in the article for licensing details.
Free full text in Europe PMC

This article is a preprint. A journal published article is available.

Preprint version history

  • Version 1 [25 Aug 2022]

Abstract 

High effectiveness of the third dose of BNT162b2 in healthy adolescents against Omicron BA.1 has been reported, but immune responses conferring this protection are not yet elucidated. In this analysis, our study (NCT04800133) aims to evaluate the humoral and cellular responses against wild-type and Omicron (BA.1, BA.2 and/or BA.5) SARS-CoV-2 before and after a third dose of BNT162b2 in healthy adolescents. At 6 months after 2 doses, S IgG, S IgG Fc receptor-binding, S-RBD IgG and neutralizing antibody responses waned significantly, yet neutralizing antibodies remained detectable in all tested adolescents and S IgG avidity increased from 1 month after 2 doses. The antibody responses and S-specific IFN-γ + and IL-2 + CD8 + T cell responses were significantly boosted in healthy adolescents after a homologous third dose of BNT162b2. Compared to adults, humoral responses for the third dose were non-inferior or superior in adolescents. The S-specific IFN-γ + and IL-2 + CD4 + and CD8 + T cell responses in adolescents and adults were comparable. Interestingly, after 3 doses, adolescents had preserved S IgG, S IgG avidity, S IgG FcγRIIIa-binding, and PRNT50 against Omicron BA.2, as well as preserved cellular responses against BA.1 S. Sera from 100% and 96% of adolescents tested at 1 and 6 months after 2 doses could also neutralize BA.1. Based on PRNT50, we predict 92%, 89% and 68% effectiveness against COVID-19 with WT, BA.2 and BA.5 1 month after 3 doses. Our study found high antibody and T cell responses, including potent cross-variant reactivity, after 3 doses of BNT162b2 vaccine in adolescents in its current formulation, suggesting that current vaccines can be protective against symptomatic Omicron disease.

Free full text 

Research Square

PPRID: PPR536221
EMSID: EMS153292Res Sq preprint, posted 2022 August 25
https://doi.org/10.21203/rs.3.rs-1961385/v1

Antibody and T cell responses against wild-type and Omicron SARS-CoV-2 after the third dose of BNT162b2 in healthy adolescents

Affiliations

  1. 1.The University of Hong Kong
  2. 2.Ishikawa Prefectural University
  3. 3.St Jude Children’s Research Hospital
  4. 4.University of Hong Kong
  5. 5.University of Melbourne
  6. 6.Centre for Immunology and Infection, Hong Kong

Copyright and license information

Copyright notice

This work is licensed under a CC BY 4.0 International license.

This article is a preprint. A journal published article is available.

Abstract

High effectiveness of the third dose of BNT162b2 in healthy adolescents against Omicron BA.1 has been reported, but immune responses conferring this protection are not yet elucidated. In this analysis, our study (NCT04800133) aims to evaluate the humoral and cellular responses against wild-type and Omicron (BA.1, BA.2 and/or BA.5) SARS-CoV-2 before and after a third dose of BNT162b2 in healthy adolescents. At 6 months after 2 doses, S IgG, S IgG Fc receptor-binding, S-RBD IgG and neutralizing antibody responses waned significantly, yet neutralizing antibodies remained detectable in all tested adolescents and S IgG avidity increased from 1 month after 2 doses. The antibody responses and S-specific IFN-γ+ and IL-2+ CD8+ T cell responses were significantly boosted in healthy adolescents after a homologous third dose of BNT162b2. Compared to adults, humoral responses for the third dose were non-inferior or superior in adolescents. The S-specific IFN-γ+ and IL-2+ CD4+ and CD8+ T cell responses in adolescents and adults were comparable. Interestingly, after 3 doses, adolescents had preserved S IgG, S IgG avidity, S IgG FcγRIIIa-binding, and PRNT50 against Omicron BA.2, as well as preserved cellular responses against BA.1 S. Sera from 100% and 96% of adolescents tested at 1 and 6 months after 2 doses could also neutralize BA.1. Based on PRNT50, we predict 92%, 89% and 68% effectiveness against COVID-19 with WT, BA.2 and BA.5 1 month after 3 doses. Our study found high antibody and T cell responses, including potent cross-variant reactivity, after 3 doses of BNT162b2 vaccine in adolescents in its current formulation, suggesting that current vaccines can be protective against symptomatic Omicron disease.

Introduction

Unvaccinated children and adolescents have a high risk of SARS-CoV-2 infection and it may be associated with hospitalizations, multi-system inflammatory syndrome and long COVID1–3. As one of the two most used vaccines worldwide, Pfizer-BioNTech-Fosun Pharma COVID-19 (BNT162b2) vaccine is a nucleoside-modified and lipid nanoparticle-formulated mRNA vaccine encoding the wild-type SARS-CoV-2 spike (S) glycoprotein4, which has demonstrated 95% efficacy in preventing COVID-19 after two doses in adults5. The Food and Drug Administration (FDA) issued an Emergency Use Authorization (EUA) for the use of BNT162b2 in adolescents aged 12–15 years on May 10, 20216. In a phase 3 study, efficacy of two-dose BNT162b2 was 100% in adolescents aged 12–15 years7. Our previous data also showed significantly higher humoral responses including total S IgG, virus neutralization, S IgG avidity and Fcγ receptor-binding antibody responses in adolescents aged 11–17 years after two doses of BNT162b2 than two doses of CoronaVac8.

Vaccine effectiveness (VE) has been found to decline at 6 months after the second dose of BNT162b2 vaccine in adults 9,10 and adolescents11. VE was also reduced during periods predominated by Omicron BA.1, which contains more than 30 mutations in its S protein, enabling dramatic neutralization escape12,13. Further Omicron sublineages have emereged, with a BA.2 epidemic wave affecting Hong Kong in January 2022, whilst BA.5 has become predominant worldwide since July 2022. In the UK, VE at 2 weeks after 2 doses of BNT162b2 vaccine declined to 65.5% in adults14, and 83.1% in adolescents aged 12–15 years15, respectively, for Omicron BA.1. Waning VE against variants of concern was enhanced by a booster dose in adults. In England, a real-world study showed 95% VE against symptomatic disease, and around 97%-99% against hospitalization or death at 14–34 days after a third dose of BNT162b2 in adults16. In Israel, a third dose of BNT162b2 had 95.3% VE, 93% against hospital admission, 92% against severe COVID-19 and 81% against death when compared with two doses in people aged 16 years or older17. A homologous third dose of BNT162b2 increased VE against symptomatic COVID to 67.2% in adults during BA.1 predominance14. It was hypothesised that a third dose of BNT162b2 in adolescents would further protect against Omicron BA.1 infection. In the US, Klein et al found 81% VE against emergency department and urgent care encounters in adolescents aged 16–17 years who received 3 doses of BNT162b2 during the BA.1 wave18. Yet, little is known about the humoral or cellular immune responses after 3 doses of BNT162b2 in healthy adolescents.

Antibody responses have been found to correlate with vaccine efficacy against symptomatic COVID-1919. Apart from antibody responses, CD8+ cytotoxic T lymphocytes (CTLs) can eliminate virus-infected cells directly and differentiated CD4+ T helper cells can coordinate a virus specific immune response20,21. Robust memory CD8+ and CD4+ T cells may provide long-lasting immunity against SARS-CoV-2 even in the absence of antibody responses and the neutralizing antibody escape by variants like Omicron22–24. Circulating effector T cells responses to the Omicron variants were preserved both in prior infected patients and vaccinated individuals12. However, binding and neutralising antibody and T cell responses against Omicron variants after the third dose of BNT162b2 vaccine in adolescents remain unknown.

Following our previous study, here we evaluated both humoral responses against the wild-type (WT) and Omicron BA.1, BA.2 and/or BA.5, including antibody binding and neutralizing functions, with ELISA-based assays and authentic plaque reduction neutralization test, and cellular responses against the WT and Omicron BA.1 by detection of intracellular IFN-γ+ and IL-2+ CD4+ and CD8+ T cells by flow cytometry, before and after the third dose of BNT162b2 in healthy adolescent aged 11–17 years compared to that in healthy adults.

Results

Enrolment of study participants

Fifty healthy adolescents aged 11–17 years and 80 healthy adults aged 18 years or older received a third dose of BNT162b2 by February 27, 2022 in our study (Fig. S1). Excluding participants who were infected during the study as determined by the presence of ORF8 antibodies25 or contributed no safety data and did not attend follow-up clinic, 28 adolescents aged 11–17 years (mean 13.7 years old) and 41 adults aged 18 years or above (mean 48.4 years old) were included in healthy safety analysis, with comparable sex and ethnicity distribution (Table S1). Primary immunogenicity was assessed in the evaluable analysis population which included participants with valid and timely immunogenicity results and no protocol deviations. Immunogenicity analyses were repeated in the expanded analysis population with relaxed vaccination and blood sampling intervals to further confirm the findings. Doses 1 and 2 were given 21–28 days apart. In evaluable analysis populations (adolescents N = 28, adults N = 33), bloods were collected 1 month after dose 2 (mean 28.5 days, post-dose 2), 6 months after dose 2 (mean 155 days, pre-dose 3), and 1 month after dose 3 (mean 22.7 days, post-dose 3). In expanded analysis populations (adolescents N = 28, adults N = 41), bloods were collected 31 days after dose 2 (post-dose 2), 160 days after dose 2 (pre-dose 3), and 25 days after dose 3 (post-dose 3). The protocol and statistical analysis plan are available in Supplementary materials.

Adolescent humoral immune responses are boosted and non-inferior to adults

For the primary humoral immunogenicity analysis, sera from evaluable adolescents and adults were collected, and antibody responses against the WT virus with SARS-CoV-2 Spike (S) IgG, S receptor-binding domain (S-RBD) IgG, S IgG avidity, and S Fcγ receptor IIIa (FcγRIIIa)-binding were tested by enzyme-linked immunosorbent assay (ELISA). ACE2-blocking antibody was estimated by surrogate virus neutralization test (sVNT). Plaque reduction neutralization test (PRNT) was also performed.

To investigate the durability of antibody responses in evaluable adolescents, the tests were performed at all timepoints including pre-dose 1, post-dose 2, pre-dose 3, and post-dose 3. An interim analysis of immunogenicity post-dose 2 has been previously performed8. The humoral responses moderately declined at pre-dose 3 when compared with that at post-dose 2 but significantly increased at post-dose 3 as measured by S IgG [geometric mean (GM)-optical density-450 (OD450) post-dose 2, 1.23 vs pre-dose 3, 0.98 vs post-dose 3, 1.41], S-RBD IgG (GM-OD450 2.57 vs 2.42 vs 2.88), sVNT (GM-% inhibition 97.1% vs 94.4% vs 97.2%), PRNT90 (GM-PRNT90 118 vs 58.8 vs 296), PRNT50 (GM-PRNT50 254 vs 137 vs 320), S IgG FcγRIIIa-binding (GM-OD450 2.10 vs 1.52 vs 1.92), but S IgG avidity index increased continually (GM avidity % 28.4% vs 52.0% vs 89.3%) (Fig. 1A).

Figure 1
Open in new tabFigure 1: Adolescents have boosted and non-inferior humoral immune responses to WT virus in comparison to adults after the third dose of BNT162b2 vaccine.

Humoral responses were compared at pre-dose 1, post-dose 2 (1 month after dose 2), pre-dose 3 (6 months after dose 2) and post-dose 3 (1 month after dose 3) in evaluable adolescents. A. Longitudinal analysis of S IgG (pre-dose 1 N=12, post-dose 2 N=21, pre-dose 3 N=21, post-dose 3 N=21), S-RBD IgG (pre-dose 1 N=21, post-dose 2 N=21, pre-dose 3 N=21, post-dose 3N=21), sVNT inhibition (pre-dose 1 N=21, post-dose 2 N=21, pre-dose 3 N=21, post-dose 3 N=18), PRNT90 (pre-dose 1 N=9, post-dose 2 N=9, pre-dose 3 N=9, post-dose 3 N=9), PRNT50 (pre-dose 1 N=9, post-dose 2 N=9, pre-dose 3 N=9, post-dose 3 N=9), S IgG avidity (pre-dose 1 N=0, post-dose 2 N=21, pre-dose 3 N=20, post-dose 3 N=21) and S IgG FcγRIIIa-binding (pre-dose 1 N=12, post-dose 2 N=21, pre-dose 3 N=21, post-dose 3 N=21) in evaluable adolescents. A third dose booster increased humoral responses except for S IgG FcγRIIIa-binding. Importantly, S IgG, S-RBD IgG, PRNT50 and S IgG avidity were higher post-dose 3 compared to post-dose 2, while there was reduction in S IgG FcγRIIIa-binding. Longitudinal analysis was determined using paired t-test after natural logarithmic transformation with p-values denoted. Data labels and centre lines show geometric means (GM) estimates, with corresponding 95% confidence intervals shown by error bars. B. Non-inferiority testing of S IgG (adolescent N=28, adult N=28), S-RBD IgG (adolescent N=28, adult N=33), sVNT inhibition (adolescent N=25, adult N=33), PRNT90 (adolescent N=14, adult N=14), PRNT50 (adolescent N=14, adult N=14), S IgG avidity (adolescent N=28, adult N=28) and S IgG FcγRIIIa-binding (adolescent N=28, adult N=28) in evaluable adolescents in comparison to adults. These humoral responses were non-inferior, except for S-IgG avidity, which was non-inferior and superior. Geometric mean ratios and their associated 2-sided 95% confidence intervals (Cis) were plotted. There was no associated 95% CI for PRNT50 as all values in both groups were equal, at the upper LOD of 320. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. ns, not significant. LOD: Limits of detection; LOQ: limits of quantification; WT: wild-type; S-RBD: Spike-receptor binding domain; sVNT: surrogate virus neutralization test; PRNT: plaque reduction neutralization test; FcγRIIIa: Fcγ receptor IIIa.

We studied antibody and T cell responses to third dose in adolescents and adults, and high antibody responses were found in both evaluable and expanded adolescents and adults (Tables 1 and S2). We tested whether the third dose was non-inferior in adolescents compared to adults, by calculating their geometric mean ratios (GMRs) and 95% confidence intervals (CI) of various immunogenicity outcomes, and assessed whether the humoral responses to the WT virus in adolescents were non-inferior to those in adults by the same methods as our previous study8. Compared to adults, humoral responses including neutralizing and binding antibodies were all non-inferior, or even superior, in evaluable adolescents after the third dose as measured by S IgG (GM-OD450 1.44 vs 1.39, GMR 0.97, 95% CI 0.91–1.03), S-RBD IgG (GM-OD450 2.93 vs 2.89, GMR 0.99, 95% CI 0.96–1.02), sVNT (GM % inhibition 97.0% vs 97.1%, GMR 1.00, 95% CI 1.00–1.00), PRNT90 (GM-PRNT90 285 vs 296, GMR 1.22, 95% CI 0.90–1.65), PRNT50 (GM-PRNT50 320 vs 320, GMR 1.00, 95% CI not applicable as all individual values were 320), S IgG avidity (GM-% avidity 81.0% vs 88.8%, GMR 1.10, 95% CI 1.04–1.15), and S IgG FcγRIIIa-binding (GM-OD450 1.87 vs 1.90, GMR 1.01, 95% CI 0.99–1.04) (Fig. 1B and S2A). Antibody responses were further confirmed in the expanded analysis population and similar results were found (Fig. S2B). These results indicate that the third dose of BNT162b2 induces high levels of humoral responses in adolescents, which are comparable to that in adults.

Table 1. Humoral immunogenicity outcomes against wild-type SARS-CoV-2 after the third dose of BNT162b2 in evaluable analysis population
Adolescents 3 dosesAdults 3 doses
S IgG on ELISA
N2828
GM OD450 value (95% CI)1.39 (1.32–1.48)1.44 (1.40–1.48)
% positive (>/=LOD at 0.3)100%, P > 0.9999100%
S-RBD IgG on ELISA
N2833
GM OD450 value (95% CI)2.89 (2.85–2.93)2.93 (2.85–3.01)
% positive (>/=LOD at 0.5)100%, P > 0.9999100%
S-RBD ACE2-blocking antibody on sVNT
N2533
GM % inhibition (95% CI)97.1% (97.0-97.2%)97.0% (96.9–97.1%)
% positive (>/=LOQ at 30%)100%, P > 0.9999100%
Neutralizing antibody on PRNT
N1414
GM PRNT90 (95% CI)263 (218–317)215 (166–279)
% positive (>/=LOD at 10)100%, P >0.9999100%
GM PRNT50 (95% CI)320 (320–320)320 (320–320)
% positive (>/=LOD at 10)100%, P >0.9999100%
S IgG avidity on ELISA
N2828
GM avidity index (95% CI)88.8% (85.8.8–91.9)81.0% (77.7–84.4)
S IgG FcγRIIIa-binding on ELISA
N2828
GM OD450 value (95% CI)1.90 (1.86–1.93%)1.87 (1.85–1.89%)
% positive (>/=LOD at 0.28)100%, > 0.9999100%

S, spike protein; ELISA, enzyme-linked immunosorbent assay; GM, geometric mean; OD, optical density; LOD, limit of detection; LOQ, limit of quantification; CI, confidence interval; RBD, receptor-binding domain; ACE-2, angiotensin-converting enzyme-2; sVNT, surrogate virus neutralization test; PRNT, plaque reduction neutralization test; PRNT90, 90% plaque reduction neutralization titre; PRNT50, 50% plaque reduction neutralization titre; FcγRIIIa, Fc gamma receptor III-a. P-values compare the proportion of positive responses between adolescents and adults by Fisher’s exact test.

Adolecent CD8+ T cell respnses are boosted post dose 3 of BNT162b2

IFN-γ+ and IL-2+ CD4+ and CD8+ T cells responses to SARS-CoV-2 overlapping S peptide pools were analyzed by flow cytometry. Compared to post-dose 2, T cells responses, including S-specific IFN-γ+ and IL-2+ CD4+ and CD8+ T cells were not significantly different at pre-dose 3 (Fig. 2A). S-specific IFN-γ+, IL-2+ CD8+ T cells increased significantly, with a respective 12.4-fold and 5-fold increase at post-dose 3 when compared to that at pre-dose 3 (Fig. 2A). The increased S-specific IFN-γ+, IL-2+ CD8+ T cell responses at post-dose 3 could also be detected in adults (Fig. S3A).

Figure 2
Open in new tabFigure 2: Adolescents have boosted CD8+ T cell responses to WT virus after the third dose of BNT162b2 vaccine.

A. Longitudinal analysis of S-specific interferon-γ (IFN-γ)+ CD4+(pre-dose 1 N=7, post-dose 2 N=7, pre-dose 3 N=21, post-dose 3 N=19), interleukin-2 (IL-2)+ CD4+ (pre-dose 1 N=7, post-dose 2 N=7, pre-dose 3 N=21, post-dose 3 N=19), IFN-γ+ CD8+ (pre-dose 1 N=7, post-dose 2 N=7, pre-dose 3 N=21, post-dose 3 N=19), and IL-2+ CD8+ (pre-dose 1 N=7, post-dose 2 N=7, pre-dose 3 N=21, post-dose 3 N=19) T cells responses in evaluable adolescents. A third dose booster increased the S-specific IFN-γ+ CD8+ and IL-2+ CD8+ T cell responses, while IL-2+ CD8+ T cell responses were higher post-dose 3 compared to post-dose 2. Longitudinal analysis was determined using paired t-test after natural logarithmic transformation with p-values denoted. Data labels and centre lines show geometric means (GM) estimates, with corresponding 95% confidence intervals shown by error bars. B. Non-inferiority test of S-specific IFN-γ+ CD4+ (adolescent N=26, adult N=28), IL-2+ CD4+ (adolescent N=26, adult N=28), IFN-γ+ CD8+ (adolescent N=26, adult N=28), and IL-2+ CD8+ (adolescent N=26, adult N=28) T cell responses in evaluable adolescents in comparison to adults. Geometric mean ratios (GMR) and two-tailed 95% confidence intervals (CI) were plotted. S-specific IFN-γ+ CD8+ and S-specific IL-2+ CD8+ T cell responses were non-inferior, while S-specific IFN-γ+ CD4+ and S-specific IL-2+ CD4+ T cell responses were non-conclusive. Geometric mean ratios and their associated 2-sided 95% confidence intervals (CIs) were plotted. **p<0.01, ***p<0.001, ****p<0.0001. ns, not significant. Cut-offs were drawn as grey lines. WT: wild-type.

The similar proportion of positive participants for WT S-specific IFN-γ+ (88.5% vs 82.1%) and IL-2+ (80.8% vs 78.6%) CD4+ T cells responses at a cut-off of 0.005% were detected in adolescents and adults after the third dose (Table 2). Interestingly, increased proportion of positive participants for WT S-specific IFN-γ+ (84.6% vs 42.9%, p = 0.002) and IL-2+ (76.9% vs 50.0%, p = 0.052) CD8+ T cells responses were found in adolescents when compared to that in adults (Table 2). This result was further confirmed in expanded analysis population (Table S3). We also calculated the geometric mean ratio for T cell responses in adolescents versus adults (Fig. 2B). Comparisons of S-specific IFN-γ+ and IL-2+ CD4+ T cell responses to adults were inconclusive as the 95% CI limits were wide and beyond the non-inferiority margin of 0.60 and 1 (Fig. 2B). However, S-specific IFN-γ+ (GMR 2.90, 95% CI 0.96–8.78) and IL-2+ (GMR 2.59, 95% CI 0.95–7.05) CD8+ responses were non-inferior in adolescents compared to that in adults as the lower bounds of their two-sided 95% CI were above 0.60 (Fig. 2B and S3B). The inconclusive S-specific IFN-γ+ and IL-2+ CD4+ T cell responses, but non-inferiority of S-specific IFN-γ+ and IL-2+ CD8+ responses in evaluable adolescents were further confirmed with that in expanded analysis populations (Fig. S3C). These results show that a third dose of BNT162b2 induces potent cellular responses in adolescents, comparable to those in adults.

Table 2. Cellular immunogenicity outcomes against wild-type SARS-CoV-2 after the third dose of BNT162b2 in evaluable analysis population
 Adolescents 3 dosesAdults 3 doses
T cell responses
S-specifìc T cell responses on flow cytometry
N2628
GM % IFN-γ+CD4+T cells0.058%0.054%
(95% CI)(0.031–0.111%)(0.027–0.106%)
% positive (>/=cut-off at 0.01%)88.5%, P = 0.7182.1%
GM % IL-2+CD4+ T cells0.046%0.040%
(95% CI)(0.022–0.095%)(0.021–0.075%)
% positive (>/=cut-off at 0.01%)80.8%, P > 0.999978.6%
GM % IFN-γ+CD8+ T cells0.045%0.016%
(95% CI)(0.023–0.091%)(0.006–0.038%)
% positive (>/=cut-off at 0.01%)84.6%, P = 0.00242.9%
GM % IL-2+CD8+ T cells0.027%0.010%
(95% CI)(0.012–0.059%)(0.005–0.020%)
% positive (>/=cut-off at 0.01%)76.9%, P = 0.05250.0%

S, Spike; GM, geometric mean; CI, confidence interval; IFN-γ, interferon-gamma; IL-2, interleukin-2. P-values compare the proportion of positive responses between adolescents and adults by Fisher’s exact test.

Humoral and cellular immunity is maintained against Omicron after the third dose in adolescents and adults

We also sought to understand whether the third dose of BNT162b2 had increased immune responses against Omicron BA.1, BA.2 and/or BA.5 in adolescents. Omicron-specific binding antibody responses including Omicron BA.1 and BA.2-S-IgG binding, IgG avidity and FcγRIIIa-binding antibodies, and BA.1, BA.2 and/or BA.5-neutralizing antibody as measured by PRNT50 compared those to WT were estimated in adolescents and adults. Interestingly, both S IgG and S FcγRIIIa-binding were conserved against both BA.1 and BA.2 after the third dose in evaluable adolescents when compared to those against WT (Fig. 3A). However, when compared to S IgG avidity against WT, it dramatically declined against BA.1, but was comparable to that against BA.2 both in adolescents and adults (Fig. 3A). 50% PRNT against BA.2 detected after the third dose in adolescents (GM PRNT50 262) was comparable to that against WT (GM PRNT50 320) (Fig. 3A), while 50% PRNT against BA.5 (GM PRNT50 60.6) was significantly decreased compared to WT after 3 doses.

Figure 3
Open in new tabFigure 3: Humoral and cellular immunity is maintained against WT, Omicron BA.1 or BA.2, or BA.5 after the third dose of BNT162b2 (BBB) vaccine in healthy evaluable adolescents and adults.

A. WT, BA.1, BA.2 and BA.5 SARS-CoV-2 Spike (S) IgG (Adult BBB N=28, Adolescent BBB N=28), S IgG avidity (Adult BBB N=28, Adolescent BBB N=28), and S IgG FcγRIIIa-binding (Adult BBB N=28, Adolescent BBB N=28) and PRNT50 (Adolescent BB N=25, Adolescent BB + 6 months N=25, Adolescent BBB N=14 for WT, N=9 for BA.2 and N=5 for BA.5) in evaluable adolescents and adults. Although antibody levels were quantitatively higher for BA.1 or BA.2 than WT, neutralization was lower for BA.1, BA.2 or BA.5 than WT. B. Omicron S WT reference pool and BA.1 mutation pool-specific interferon-γ (IFN-γ)+ CD4+ (Adult BBB N=28, Adolescent BBB N=22), interleukin-2 (IL-2)+ CD4+ (Adult BBB N=28, Adolescent BBB N=22), IFN-γ+ CD8+ (Adult BBB N=28, Adolescent BBB N=22), and IL-2+ CD8+ (Adult BBB N=28, Adolescent BBB N=22) T cells in evaluable adolescents and adults. The cellular responses were similar between WT and BA.1. Data labels and centre lines show geometric means (GM) estimates, with corresponding 95% confidence intervals shown by error bars. Statistical analysis was determined using paired t-test after natural logarithmic transformation with p-values denoted. **p<0.01, ***p<0.001, ****p<0.0001. ns, not significant. LOD: Limit of detection; WT: wild-type; FcgRIIIa: Fcg receptor IIIa; PRNT: plaque reduction neutralization test; Cut-offs were drawn as grey lines.

To investigate whether Omicron BA.1 variant could escape T cell recognition after the third dose in adolescents, an S mutation pool which contained peptides covering 37 BA.1-associated mutations was used. T cell responses were compared to those from the WT reference peptide pool. As expected, there were no significant differences between WT and BA.1 S-specific IFN-γ+ and IL-2+ CD4+ and CD8+ T cells responses in both adolescents and adults (Fig. 3B). These results indicate that the third dose of BNT162b2 elicits potent protection against Omicron subvariants both in adolescents and adults.

Estimation of vaccine efficacies based on neutralization titres for the third dose of BNT162b2 in adolescents

We used PRNT50 to extrapolate the VE against symptomatic COVID-19 with WT, BA.2 and BA.5 in evaluable adolescents as described before8,19,26,27. The mean neutralization values against WT, BA.2 and BA.5 compared to convalescent sera were 2.31, 1.57 and 0.44, which extrapolated to 92%, 89% and 68% VE, respectively, in adolescents after the third dose (Fig. 4). According to these data, the third dose of BNT162b2 substainally exceeds the WHO’s recommended 50% VE threshold as effective for use against SARS-CoV-2 infection, including BA.2 and BA.5, in adolescents 1 month after dose 3.

Figure 4
Open in new tabFigure 4: Vaccine efficacy estimations of three doses of BNT162b2 for adolescents based on neutralization titres against SARS-CoV-2 WT, Omicron BA.2 and Omicron BA.5.

VEs against symptomatic COVID-19 were predicted by neutralizing antibodies. The mean neutralization level (fold of convalescent) were measured by dividing the geometric mean titres of PRNT50 in healthy evaluable adolescents who received three doses of vaccine with the WT PRNT50 of 102 convalescent sera collected on days 28-59 post-onset of illness in patients aged 18 years or above. A point estimate of VE was extrapolated from the best fit of the logistic model in Khoury et al8,19,26,27. VE: vaccine efficacy; BBB: three doses of BNT162b2; WT: wild-type.

Discussion

This study is the first to evaluate a wide range of humoral and cellular outcomes following a third dose of BNT162b2 in adolescents aged 11–17 years. A third dose can significantly boost antibody responses and CD8+ T cell responses in adolescents. These responses are similar compared to adults. Importantly, a third dose BNT162b2 can provide protective levels of humoral responses to Omicron subvariants, and cellular responses to BA.1 when compared to WT.

A high level of protection against SARS-CoV-2 infection and hospitalization after 2 doses of mRNA vaccine in adolescents was found both in clinical trials and the real-world data28,29. Our previous data also showed higher levels of humoral and cellular responses in adolescents after 2 doses than 1 dose only8. A third dose boosts the waning antibody response in adolescents. Similar to adult data, we showed significant reductions in neutralizing antibodies, S IgG and FcγR-binding antibodies at 6 months after two doses of vaccine9. However, in almost all parameters tested, a third dose BNT162b2 was able to reestablish and enhance antibody responses. IgG avidity is used to measure the strength of binding of the S IgG response and is indicative of formation of germinal centre reactions and high quality antibodies30,31. Increasing IgG avidity over time correlates with the establishment of long-lasting spike antibody responses after both two and three doses. The non-inferior and superior S IgG avidity in adolescents compared to adults suggests that there may be more long-lasting antibodies generated in adolescents. However, a longer term follow-up will be required to observe whether these boosted S IgG, neutralizing and high avidity responses will be maintained in adolescents. The FcγRIIIa-binding responses are the only serological response that remained static following a third dose in adolescents. FcγRIIIa-binding antibodies are associated with antibody effector functions, clearance of immune complexes and killing of infected cells, thereby contributing to protection from severity during breakthrough Omicron infections32,33. These FcγR antibodies increased significantly in adolescents and adults following two doses of BNT162b28 but had not been assessed following three doses in adults or adolescents. Here we found that this potential correlate of protection does not increase following a third dose in adolescents, and the response was non-inferior compared to adults.

Maintenance of S-specific CD4+ T cell responses at 6 months after two doses vaccination is promising for the longevity of T cell responses in adolescents. The cross-reactive nature of T cell responses34,35 and their maintained responses against S observed here suggests long-lasting protection against future related variants in adolescents and adults. The lack of boosting in CD4+ T cell responses after a third dose of BNT162b2 was consistent with previous studies in adults36 which may be due to an immune ceiling being reached by significantly boosted and maintained responses following two doses. However, we observed a significant boost and enhancement in CD8+ T cell responses post-dose 3 in both adolescents and adults, although it was not seen in adults in previous studies36. This lack of BNT162b2 significant boosting of T cell responses after two doses was seen in some adult cohorts37 but not others38 might be related to small sample sizes, assay sensitivity, or the differences in HLA in different geographic locations that led to variable epitope presentation, and therefore, variable responses.

Decreased vaccine efficacy of two doses alongside the surge of breakthrough infections with the Delta variant (B.1.617.2) and Omicron BA.1 of SARS-CoV-2 prompted the rapid rollout of the third dose of BNT162b2 vaccine globally14. Omicron BA.1 was first reported to the WHO by South Africa on November 24, 202139, and caused significant concern due to a large numbers of mutations, especially in the Spike protein. Many studies in healthy adults have reported the dramatic escape of neutralization antibodies by Omicron variants after two doses, but increased antibody-based immunity against Omicron after 3 doses of BNT162b239,40. In healthy adults from another cohort, using the same assay, we found 50% PRNT GMT of 67.3 and 95.1 against BA.1 and BA.2 1 month after 3 doses 40,41. In contrast, our data showed that adolescents had high GMT against BA.2 after three doses, which was comparable to PRNT50 against WT after 2 or 3 doses of BNT162b2 vaccine in adults. This result is also consistent to our findings between BA.2 and WT in terms of S IgG avidity. Although BA.5 neutralisation was significantly lower than for WT after 3 doses in adolescents, it still predicted a VE of 68% against symptomatic disease. Our data also showed that adolescents had preserved levels of binding antibodies as measured by S IgG and S IgG FcγRIIIa-binding antibodies against BA.1 and BA.2 after the third dose of BNT162b2 when compared with that against WT. Indeed, another study also showed that Omicron S-specific binding for IgG and FcγRIIIa persisted at a high level across the two doses of mRNA vaccine42. Besides neutralizing antibody, the non-neutralizing antibodies like S IgG FcγRIIIa binding can lead to continued viral clearance and the killing of infected cells, finally, may contribute to the less severe Omicron infection32,33. Therefore, our results demonstrate that the third dose of BNT162b2 can provide protection against Omicron subvariants by potent cross-reactive binding and neutralizing antibody and T cells responses.

There are limitations in this study. First, we compared humoral and cellular responses after the third dose of BNT162b2 vaccine on a limited subset of samples due to the surge of breakthrough infections with the Omicron BA.2 of the SARS-CoV-2 in Hong Kong during the study period, when some participants were infected or defaulted follow-up clinic to avoid potential Omicron BA.2 transmission. We also did not define the differentiation status of S-specific T cells like the naïve, central memory, effector memory, and terminally differentiated effector populations, which may elucidate whether vaccine induced a long-lasting or exhausted T cell response in adolescents43–45. We did not investigate every Omicron subvariant at each timepoint due to assay availability. Hybrid immunity, whether by vaccinating infected adolescents or breakthrough infections in vaccinated adolescents, was not investigated in our study, but more studies on this will be important for understanding the long-term immunological implications of vaccination in this population46–48.

Taken together, our data suggest cross-reactive potent antibody and T cell responses are elicited by a third dose of BNT162b2 in adolescents, explaining the high vaccine effectiveness observed in real-world studies. We will further track the durability of immunogenicity after three doses BNT162b2 vaccine and hybrid immunity after breakthrough infections.

Star Methods

Study Design

Coronavirus disease-19 (COVID-19) Vaccination in Adolescents and Children (COVAC; NCT04800133) aimed at evaluating the humoral and cellular immunogenicity in children8. This study was approved by the University of Hong Kong (HKU)/Hospital Authority Hong Kong West Cluster Institutional Review Board (UW21-157).

Participants

This study included the healthy adolescents aged 11–17 years and adults aged 18 years or older who received three doses of BNT162b2 intramuscularly. Potential participants with stably healthy conditions, known history of COVID-19, history of severe allergy, significant neuropsychiatric conditions, immunocompromised states were included. Transfusion of blood products within 60 days, haemophilia, pregnancy or breastfeeding were excluded from this study.

Procedures

Potential participants were recruited via school, media, or referral in Hong Kong. Study physicians contacted and obtained informed consent from participants aged 18 years or above, or for underage participants, from their parents and legally acceptable representatives.

S-RBD, surrogate virus neutralization assay (sVNT) and plaque reduction neutralization test (PRNT)

Peripheral clotted blood was drawn, and the serum was stored at -80°C after separation. The SARS-CoV-2 S receptor-binding domain (R-SBD) IgG enzyme-linked immunosorbent assay (ELISA) and PRNT were carried out as previously described and validated8. sVNT was conducted according to the manufacturer’s instructions (GenScript Inc, Piscataway, USA) and as described in our previous publication. All sera were heat-inactivated at 56°C for 30 mins before testing27,49. Details for the detections of S-RBD IgG, sVNT and PRNT were performed by the same methods showed in our previous study8. Briefly, S-RBD IgG ELISA plates were coated overnight with 100 ng/well of purified recombinant S-RBD in PBS buffer, followed by the incubation with 100 μL Chonblock Blocking/Sample Dilution (CBSD) ELISA buffer (Chondrex Inc, Redmond, USA) at room temperature (RT) for 2 hrs. Then added the 1:100 diluted serum in CBSD ELISA buffer to the wells and incubated at 37°C for another 2 hrs. After washing with 0.1% Tween 20 PBS (PBST), the plates were incubated with 1: 5000 diluted horseradish peroxidase (HRP)-conjugated goat anti-human IgG (Thermo Fisher Scientific) at 37°C for 1 h and washed with PBST for 5 times. Finally, 100 μL of HRP substrate (Ncm TMB one, New Cell & Molecular Biotech. Ltd. China) was added for 15 mins before stopped this reaction by 50 μL of 2 M H2SO4. The optical density (OD) was analyzed in a Sunrise absorbance microplate reader (Tecan, Männedorf, Switzerland) at 450 nm wavelength. Each OD reading was calculated by subtracting the background OD in PBS-coated control wells with the serum of participants. Values at or above an OD450 of 0.5 were considered positive, otherwise were imputed as 0.25.

For sVNT detection, 10 μL of serum were diluted at 1:10 and incubated with an equal volume HRP conjugated to the WT SARS-CoV-2 S-RBD (6ng) at 37°C for 30 mins, followed by the addition of 100 μL of each sample to each well of microtitre plates coated with angiotensin-converting enzyme-2 (ACE-2) receptor at 37°C for 15 mins. After washing and drying, 100 μL of 3,3’,5,5’-tetramethylbenzidine (TMB) was added and incubated at RT far away from light for 15 mins. Finally, the reaction was terminated, and the absorbance was read at 450 nm in a microplate reader. After confirmation that the positive and negative controls provided the recommended OD450 values, the % inhibition of each serum was calculated as (1 - sample OD value/negative control OD value) ×100%. Inhibition (%) of at least 30%, the limit of quantification (LOQ), was regarded as positive, while values below 30% were imputed as 15%.

The PRNT assay was performed in duplicate under a facility with biosafety level 3 as described before8. In brief, serum was diluted from 1:10 to 1:320, and then incubated with BetaCoV/Hong Kong/VM20001061/2020 (WT strain), hCoV-19/Hong Kong/VM21044713_WHP5047-S5/2021 (Omicron BA.1), hCoV-19/Hong Kong/VM22000135_HKUVOC0588P2/2022 (Omicron BA.2), or SARS-CoV-2/human/USA/COR-22-063113/2022 (Omicron BA.5) at 30 plaque-forming units in a culture plate (Techno Plastic Products AG, Trasadingen. Switzerland) at 37°C for 1h. Then the virus-serum mixtures were added onto Vero E6 TMPRESS2 cell monolayers and further incubated at 37°C for 1h. The plates were overlaid with 1% agarose in cell culture medium and incubated for 3 days. After fixing and staining, antibody titres were defined as the reciprocal of the highest serum dilution that resulted in > = 90% (PRNT90, a more stringent cut-off) or > 50% (PRNT50) reduction in the number of plaques. Values below the lowest dilution tested of 10 were imputed as 5.

S IgG, avidity and FcγRIIIa-binding

Detections of S IgG, avidity and FcγRIIIa-binding were carried out as previously described8. Briefly, proteins were diluted in PBS for specific antibody detection. Firstly, Plates (Nunc MaxiSorp., Thermofisher Scientific) were coated with 250 ng/mL WT (AcroBiosystems) or Omicron BA.1 (AcroBiosystems) or Omicron BA.2 (AcroBiosystems) SARS-CoV-2 S protein for IgG and IgG avidity detections, or 500 ng/mL WT (Sinobiological) or Omicron BA.1 (AcroBiosystems) S for FcγRIIIa-binding detections, or 300 ng/mL ORF8 (Masashi Mori, Ishiwaka University, Japan) at 37°C for 2 hrs25,50. The detection of ORF8 specific IgG was used to exclude infected individuals.

For IgG detection, plates were blocked with 1% FBS in PBS for 1 h before incubated with heat-inactivated (HI) serum, which was 1:100 diluted in 0.05% Tween-20/0.1% FBS in PBS at RT for 2 hrs. For antibody avidity, plates were washed three times with 8M Urea before incubated with IgG-HRP (1:5000, G8-185, BD) for 2 hrs. HRP was revealed by stabilized hydrogen peroxide and tetramethylbenzidine (R&D systems) for 20 mins, then stopped with 2N H2SO4 and analyzed with an absorbance microplate reader at 450 nm wavelength (Tecan Life Sciences). For FcγRIIIa-binding measurement, plates were coated with 500 ng/mL S protein and incubated with 1:50 diluted HI serum at 37°C for 1 h before incubated with 100 ng/mL biotinylated FcγRIIIa-V158 at 37°C for 1 h, followed by the detection of S specific FcγRIIIa-V158-binding antibodies by using streptavidin-HRP (1:10000, Pierce).

T cell responses

Peripheral blood mononuclear cells (PBMCs) were isolated from the whole blood of participants by density gradient separation and stored in liquid nitrogen before use. Firstly, PBMCs were thawed in RPMI medium supplemented with 10% human AB serum, then rested in a 37°C incubator for 2 hrs. The cells were stimulated with 1 μg/mL overlapping peptide pools representing the WT SARS-CoV-S proteins (Miltenyi Biotec, Bergisch Gladbach, Germany), or B.1.1.529/BA.1 S mutation pool (Miltenyi Biotec, Bergisch Gladbach, Germany) and WT reference pool (Miltenyi Biotec, Bergisch Gladbach, Germany), supplemented with 1 μg/mL anti-CD28 and anti-CD49d costimulatory antibodies (Clones CD28.2 and 9F10, respectively, Biolegend, San Diego, USA) at 37°C for 16 hrs. An equal volume of sterile double-distilled water (ddH2O) was used as a negative control. This mixture was stimulated for 2 hrs, followed by the addition of Brefeldin A (BFA,10 μg/mL; Sigma, Kawasaki, Japan) 51. Secondly, the cells were washed and immunostained with a fixable viability dye (eBioscience, Santa Clara, USA, 1:60), and antibodies against-CD3 (HIT3a, 1:60), CD4 (OKT4, 1:60), CD8 (HIT8a, 1:60), followed by fixed, permeabilized and stained with antibodies against IFN-γ (B27, 1:15) and IL-2 (MQ1-17H12, 1:15). All of these antibodies were purchased from Biolegend. Finally, data acquisition was carried out using flow cytometry (LSR II, BD Biosciences, Franklin Lakes, USA) and analyzed by FlowJo v10 software (BD, Ashland, USA). Antigen-specific IFN-γ+ and IL-2+ T cell results were finalized after subtracting the background (ddH2O) data and presented as the percentage of CD4+ or CD8+ T cells44. T cell responses against a single peptide pool were considered positive when the frequency of cytokine-expressing cells was higher than 0.005% and the stimulation index was higher than 2, while negative values were imputed as 0.0025%.

Outcomes

Humoral immunogenicity (S IgG and S-RBD IgG levels, sVNT %inhibition, 90% and 50% PRNT titres, S IgG avidity and FcγRIIIa-binding) and cellular immunogenicity markers (S-, Omicron S mutation-and Omicron WT reference- specific IFN-γ+ and IL-2+ CD4+ and CD8+ T cell responses) assessed after the third dose of BNT162b2.

Statistical analyses

Sample size

As the study was conducted during the Omicron BA.2 wave in Hong Kong, participants who were infected were excluded and some participants defaulted vaccination or follow-up clinic. All evaluable samples were tested by S-RBD IgG and sVNT, and sample sizes for more demanding assays, e.g. PRNT and T cell testing, were reduced based on laboratory capacity.

Analysis sets

The primary analysis of humoral and cellular immunogenicity outcomes was performed in the healthy adolescents and adults in the evaluable analysis population who received intramuscular injection of BNT162b2 vaccine on a per-protocol basis as described before8. All of these evaluable population remained uninfected during study visits based on self-reporting, ORF8 IgG negativity and negative baseline S-RBD IgG, had no major protocol deviations. Each immunogenicity outcome was calculated by GM, GM ratios (GMRs) were reported with a two-sided 95% CI, corresponding to a one-sided 97.5% CI, to test the non-inferiority hypothesis at the margin of 0.60. Non-inferiority analyses were further confirmed in the expanded analysis population. When both non-inferiority and inferiority were not met, the results were inconclusive. Participants with valid results at consecutive timepoints were compared by GM fold rise (GMFR). Immunogenicity outcomes data below the cut-off were imputed with half the cut-off value. Immunogenicity outcomes were analysed by an unpaired or paired t test after natural logarithmic transformation. The proportion of participants with a positive results was reported in percent with 95% CI derived from Clopper-Pearson method. The comparisons of proportions between groups were performed with the Fisher exact test.

Vaccine efficacy estimates

Neutralizing antibody titres were used to estimate the vaccine efficacies as a secondary objective as described before8,19. Briefly, the mean neutralizing levels (fold of convalescent) were derived by dividing the GMTs of PRNT50 for WT, Omicron BA.2 or BA.5 SARS-CoV-2 in healthy evaluable adolescents with that of 102 convalescent sera collected on days 28–59 post-onset of illness in patients aged ≥ 18 years. Point estimates of VE were extrapolated by the best fit of the logistic model generated from the online plot digitizer tool (https://automeris.io/WebPlotDigitizer/, version 4.5).

Supplementary Material

Acknowledgements

We thank the personnel from the Community Vaccination Centres at Ap Lei Chau Sports Centre, Gleneagles Hospital HK, and Sun Yat Sen Memorial Park Sports Centre as well as particularly Dr Victoria WY Wong of HKU, Ms Cindy HS Man and Dr Hon-Kuan Tong who operated the territory-wide vaccination programme at the CVC and assisted in clinical monitoring of vaccine recipients alongside our group. The investigators are grateful to all clinical research team members and laboratory staff of Department of Paediatrics and Adolescent Medicine, including Mr KW Chan and Dr Davy CW Lee, for their research support. We are indebted to the participants who have placed their trust in the investigators and joined the study. This study was supported by the research grant COVID19F02 and COVID19F10 awarded to Y.L. Lau by the Food and Health Bureau of the Government of Hong Kong, which was not involved in the study design, data collection, laboratory assays, statistical computation, interpretation or final conclusions of this project.

Author Information

kh.ukh@gnulyual

Author contributions

Y.L.L. conceptualised the study. Y.L.L., W.T., S. V., M.P., X. M., C.A.C., and D. L. designed the study. Y.L.L. led the acquisition of funding. Y.L.L., W.T., S.V. and M.P. supervised the project. J.S.R.D., C.A.C., D. L., J.S.R.D and S.M.C. led the study administrative procedures. J.H.Y.L. and D.H.Y.L. provided software support. S.M.C. contributed to recruitment of participants. Y.L.L. and J.S.R.D. provided clinical assessments and follow-up. D.L., S.C., J.H.Y.L., J.S.R.D. and Y.L.L. collected safety data. S.M.S.C., S.C., K.K.H.K., K.C.K.C., J.K.C.L., L.L.H.L., L.C.H.T., N.C.C., T. J., A. K., R.J.W. and M.P. developed and performed S-RBD IgG, sVNT and neutralisation antibody assays. C.A.C. and S.V. developed and performed the S IgG, IgG avidity, S IgG Fcγ receptor IIIa-binding and ORF8 antibody assays. The specialised ORF8 protein was developed and provided by M.M., X.M., Y. C., H.H.W.W., A.M.T.L. W.Y.L. and W.T. developed and performed the T cell assays. D.L. and J.H.Y.L. curated the data. D.L., J.H.Y.L. and D.H.L.L analysed the data. D.L., S.M.S.C., Y.L.L. and M.P. performed the vaccine efficacy extrapolation. D.L. visualised the data. X. M., C.A.C., D.L., J.S.R.D. and S.M.S.C. validated the data. X. M. wrote the first draft as supervised by Y.L.L., with input from C.A.C., D. L., J.S.R.D., and S.M.S.C.. All authors reviewed and approved the final manuscript.

Competing interests

The authors declare no competing interests.

References

History

  • Posted August 25, 2022.

Full text links 

Read article at publisher's site: https://doi.org/10.21203/rs.3.rs-1961385/v1

Read article for free, from open access legal sources, via Unpaywall: https://www.nature.com/articles/s41392-022-01282-7.pdfUnpaywall

Citations & impact 

This article has not been cited yet.

Impact metrics

Alternative metrics

Altmetric item for https://www.altmetric.com/details/134973865
Altmetric
Discover the attention surrounding your research
https://www.altmetric.com/details/134973865