Follow-Up and Comparative Assessment of SARS-CoV-2 IgA, IgG, Neutralizing, and Total Antibody Responses After BNT162b2 or mRNA-1273 Heterologous Booster Vaccination

Abstract

Background: Priming with ChAdOx1 followed by heterologous boosting is considered in several countries. Nevertheless, analyses comparing the immunogenicity of heterologous booster to homologous primary vaccination regimens and natural infection are lacking. In this study, we aimed to conduct a comparative assessment of the immunogenicity between homologous primary vaccination regimens and heterologous prime-boost vaccination using BNT162b2 or mRNA-1273.

Methods: We matched vaccinated naïve (VN) individuals (n = 673) with partial vaccination (n = 64), primary vaccination (n = 590), and primary series plus mRNA vaccine heterologous booster (n = 19) with unvaccinated naturally infected (NI) individuals with a documented primary SARS-CoV-2 infection (n = 206). We measured the levels of neutralizing total antibodies (NTAbs), total antibodies (TAbs), anti-S-RBD IgG, and anti-S1 IgA titers.

Results: Homologous primary vaccination with ChAdOx1 not only showed less potent NTAb, TAb, anti-S-RBD IgG, and anti-S1 IgA immune responses compared to primary BNT162b2 or mRNA-1273 vaccination regimens (p < 0.05) but also showed ~3-fold less anti-S1 IgA response compared to infection-induced immunity (p < 0.001). Nevertheless, a heterologous booster led to an increase of ~12 times in the immune response when compared to two consecutive homologous ChAdOx1 immunizations. Furthermore, correlation analyses revealed that both anti-S-RBD IgG and anti-S1 IgA significantly contributed to virus neutralization among NI individuals, particularly in symptomatic and pauci-symptomatic individuals, whereas among VN individuals, anti-S-RBD IgG was the main contributor to virus neutralization.

Conclusion: The results emphasize the potential benefit of using heterologous mRNA boosters to increase antibody levels and neutralizing capacity particularly in patients who received primary vaccination with ChAdOx1.

Keywords: COVID‐19; antibodies; anti‐S1‐IgA; anti‐S‐RBD IgG; booster; immune response; neutralizing antibody; vaccination.

FIGURE 1

Summary of the study cohort and timeline of sampling. The study included a total of 879 samples. We classified study subjects into two mutually exclusive groups: (1) vaccinated naïve (VN; n = 673) and (2) unvaccinated naturally infected (NI; n = 206). The VN group was further classified to three subgroups: (1) Partially vaccinated group included samples collected post‐one dose of either ChAdOx1, mRNA‐1273, or BNT162b2. (2) The primary series group included samples collected post–two homologous doses of either ChAdOx1, mRNA‐1273, or BNT162b2. (3) The primary series plus one booster dose group included samples collected post–two doses of ChAdOx1, followed by a heterologous booster shot of either mRNA‐1273 or BNT162b2. * denotes non–mutually exclusive groups, comprising a total of 98 samples collected from 20 VN subjects at five time points (T1–T5). T1 and T2, collected post–first dose, T3 and T4, collected post–homologous second dose, and T5, collected post–heterologous booster (third) dose. Initially, 20 samples from 20 study subjects were collected at each time point; however, two samples at T4 and T5 were excluded due to SARS‐CoV‐2 infection during the follow‐up period. The NI group was further classified to symptomatic, pauci‐symptomatic, and asymptomatic.

FIGURE 2

Assessment of vaccine‐induced immunity after heterologous booster with mRNA vaccines. (A) NTAb neutralizing total antibody levels measured by CL‐900i® (IU/mL). (B) TAb total antibody levels measured by CL‐900i. (C) Anti‐S‐RBD IgG antibody levels (BAU/mL) measured by CL‐900i®. (D) Anti‐S1 IgA ratios measured by Euroimmun. Each circle represents a single sample. Black bars indicate interquartile range (IQR). Statistical significance was determined using Kruskal–Wallis test. p value asterisk denotes to *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001.

FIGURE 3

Longitudinal antibody response in VN individuals who received heterologous booster doses. (A) NTAb neutralizing total antibody levels measured by CL‐900i® (IU/mL). (B) TAb total antibody levels measured by CL‐900i. (C) Anti‐S‐RBD IgG antibody levels (BAU/mL) measured by CL‐900i®. (D) Anti‐S1 IgA ratios measured by Euroimmun. A total of 98 samples collected from 20 VN subjects at five time points (T1–T5). T1 and T2, collected post–first dose (~36 and ~75 days from first dose, respectively), T3 and T4, collected post–homologous second dose (~104 and ~205 days from first dose, respectively), and T5, collected post–heterologous booster (third) dose (~296 days from first dose). Initially, 20 samples from 20 study subjects were collected at each time point, however, two samples at T4 and T5 were excluded due to SARS‐CoV‐2 infection during the follow‐up period. The NI group was further classified to symptomatic, pauci‐symptomatic, and asymptomatic. Statistical significance of antibody levels among paired samples was assessed using Friedman test. Mann–Whitney test was conducted for comparisons between T5: ChAdOx1^1,2 + BNT162b2³ and ChAdOx1^1,2 + mRNA‐1273³. p value asterisk indicates *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001. Only significant correlations are shown.

FIGURE 4

Pairwise correlation of neutralizing total antibody (NTAb) titers with anti‐S‐RBD IgG and anti‐S1 IgA levels among VN individuals. Scatter plots (left) and Spearman's r and p values' correlation matrices (right) for (A) NTAbs/anti‐S‐RBD IgG and (B) NTAbs/anti‐S1 IgA were generated. Correlation coefficients in the range 0–0.39, 0.40–0.59, 0.6–0.79, and 0.8–1 suggest weak, moderate, high, and very strong correlations, respectively. Scatterplots were used to depict the direction, form, and strength of correlations. All p values were two sided at a significance level of 0.05. p values < 0.001 is represented as 0.001.

FIGURE 5

Pairwise correlation of neutralizing total antibody (NTAb) titers with anti‐S‐RBD IgG and anti‐S1 IgA levels among NI individuals. Scatter plots (left) and Spearman's r and p values' correlation matrices (right) for (A) NTAbs/anti‐S‐RBD IgG and (B) NTAbs/anti‐S1 IgA were generated. Correlation coefficients of 0–0.39 indicate a weak, 0.40–0.59 a moderate, 0.6–0.79 a strong, and 0.8–1 a very strong correlation. Scatterplots were used to depict the direction, form, and strength of correlations. All p values were two‐sided at a significance level of 0.05. p values < 0.001 is represented as 0.001.