Effects of boosted mRNA and adenoviral-vectored vaccines on immune responses to omicron BA.1 and BA.2 following the heterologous CoronaVac/AZD1222 vaccination

Abstract

The coronavirus 2019 omicron variant has surged rapidly and raises concerns about immune evasion even in individuals with complete vaccination, because it harbors mutations. Here we examine the capability of booster vaccination following CoronaVac/AZD1222 prime to induce neutralizing antibodies (NAbs) against omicron (BA.1 and BA.2) and T-cell responses. A total of 167 participants primed with heterologous CoronaVac/AZD1222 for 4-5 months were enrolled, to receive AZD1222, BNT162b2, or mRNA-1273 as a third dose. Reactogenicity was recorded. Immunogenicity analyses of severe acute respiratory syndrome coronavirus 2-binding antibodies were measured using enzyme-linked immunosorbent assay. The NAb titers against omicron BA.1 and BA.2 were determined using the focus reduction neutralization test (FRNT50) and total interferon-γ responses were measured to observe the T-cell activation. A substantial loss in neutralizing potency to omicron variant was found at 4-5 months after receiving the heterologous CoronaVac/AZD1222. Following booster vaccination, a significant increase in binding antibodies and neutralizing activities toward delta and omicron variants was observed. Neutralization to omicron BA.1 and BA.2 were comparable, showing the highest titers after boosted mRNA-1273 followed by BNT162b2 and AZD1222. In addition, individuals boosted with messenger RNA (mRNA) vaccines develop a T-cell response to spike protein, whereas those boosted with AZD1222 did not. Reactogenicity was mild to moderate without serious adverse events. Our findings demonstrated that mRNA booster vaccination is able to overcome waning immunity to provide antibodies that neutralize omicron BA.1 and BA.2, as well as a T-cell response.

Keywords: AZD1222; COVID-19; booster vaccine; heterologous; mRNA vaccine; omicron.

Figure 1

Study design and measurement of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2)‐specific receptor‐binding domain (RBD)‐binding antibody responses. Schematic depicting a total of 167 vaccinated individuals with heterologous CoronaVac/AZD1222 enrolled in cohort study. They were assigned to receive booster vaccines, either AZD1222 (n = 60), BNT162b2 (n = 55), or mRNA‐1273 (n = 52), and blood samples were collected on Days 0, 14, and 28 after booster vaccination (A). The anti‐RBD IgG (BAU/ml) in sera from boosted individuals with different vaccines, AZD1222 (purple), BNT162b2 (green), and mRNA‐1273 (yellow), were compared (B). Error bars in B indicate the geometric mean titers (GMTs).

Figure 2

Neutralizing activities measured using surrogate virus neutralization test against delta and omicron (BA.1). A subset of samples from boosted individuals with AZD1222 (purple), BNT162b2 (green), and mRNA‐1273 (yellow), which was randomly selected to test the surrogate virus neutralization test (sVNT) that included sera collected at baseline (n = 15/group) and sera collected at 28 days post‐boost (n = 30/group). Neutralizing activities against delta (A) and omicron (BA.1) (B) were compared between pre‐ and post‐booster vaccination. Numbers above the bar graph indicate the percentage of inhibition between human angiotensin converting enzyme 2 and receptor binding domain (ACE‐2 and RBD, respectively) proteins. A comparison of the neutralizing activity between delta and omicron variants at 28 days after booster vaccination is shown in (C). Median values are shown as horizontal bars. Dotted lines indicate cut‐off values (30%). The comparison was perform using Wilcoxon signed‐rank test (two‐tailed). ***p < 0.001.

Figure 3

Neutralizing antibody (NAb) titers against omicron BA.1 and BA.2 measured using focus reduction neutralization test (FRNT50). A subset of samples from boosted individuals with AZD1222 (purple), BNT162b2 (green), and mRNA‐1273 (yellow), which was randomly selected to test the FRNT50 included sera collected at baseline (n = 10/group) and sera collected at 28 days post‐boost (n = 30/group). NAb titers against omicron BA.1 (A) and omicron BA.2 (B) were compared between baseline (Day 0) and 28 days post‐booster vaccination with different vaccines. Numbers above the plot indicate the geometric mean titers (GMTs). Fold increases for each comparison are denoted. NAb titers against BA.1 and BA.2 at 28 days after booster vaccination were compared (C). Statistical analysis was done using Wilcoxon signed‐rank test (two‐tailed). The horizontal dotted line indicates the limit of detectable value of FRNT50. Values below the limit of detection (<20) were set at a titer of 10 before statistical analysis. ns, no significant difference.

Figure 4

Comparison of total interferon‐γ (IFN‐γ)‐releasing T‐cell responses to severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) antigens. Heparinized whole blood samples from vaccinated individuals receiving heterologous CoronaVac/AZD1222 followed by a third booster with AZD1222 (n = 30, purple), BNT162b2 (n = 30, green), or mRNA‐1273 (n = 30, yellow) were stimulated by Ag1, which is a CD4+ epitope derived from receptor‐binding domain (RBD), minus negative control (Nil) (A), and Ag2 which is CD4+ and CD8+ epitopes derived from S1 and S2 subunits, minus negative control (Nil) (B). Horizontal bars indicate the median. The cut‐off values were represented by horizontal dotted line. Levels of IFN‐γ above cut‐off values (0.15 IU/ml and ≥25% of Nil) indicate a reactive response. Statistical analysis was done using Wilcoxon signed‐rank test (two‐tailed). ns, no significant difference; **p < 0.01; ***p < 0.001.

Figure 5

Forest plot showed the risk difference with 95% confidence intervals of adverse events (AEs) after booster vaccination. The proportion of participants with any grade of solicited AEs after receipt the third dose were compared between AZD1222 versus BNT162b2 vaccine (A), AZD1222 versus mRNA‐1273 (B), and BNT162b2 versus mRNA‐1273 (C).