Higher SARS-CoV-2 Spike Binding Antibody Levels and Neutralization Capacity 6 Months after Heterologous Vaccination with AZD1222 and BNT162b2

Abstract

Within a year after the emergence of SARS-CoV-2, several vaccines had been developed, clinically evaluated, proven to be efficacious in preventing symptomatic disease, and licensed for global use. The remaining questions about the vaccines concern the duration of protection offered by vaccination and its efficacy against variants of concern. Therefore, we set out to analyze the humoral and cellular immune responses 6 months into homologous and heterologous prime-boost vaccinations. We recruited 190 health care workers and measured their anti-spike IgG levels, their neutralizing capacities against the Wuhan-Hu-1 strain and the Delta variant using a surrogate viral neutralization test, and their IFNγ-responses towards SARS-CoV-2-derived spike peptides. We here show that IFNγ secretion in response to peptide stimulation was significantly enhanced in all three vaccination groups and comparable in magnitude. In contrast, the heterologous prime-boost regimen using AZD1222 and BNT162b2 yielded the highest anti-spike IgG levels, which were 3-4.5 times more than the levels resulting from homologous AZD1222 and BNT162b2 vaccination, respectively. Likewise, the neutralizing capacity against both the wild type as well as the Delta receptor binding domains was significantly higher following the heterologous prime-boost regimen. In conclusion, our results suggest that mixing different SARS-CoV-2 vaccines might lead to more efficacious and longer-lasting humoral protection against breakthrough infections.

Keywords: COVID-19; SARS-CoV-2; heterologous prime-boost vaccination; humoral and cellular immune response; variant of concern.

Figure 1

Heterologous prime-boost combining AZD1222 and BNT162b2 resulted in highest antibody levels at six months. Box plots compare anti-spike IgG antibody levels in BAU/mL six months into homologous or heterologous prime-boost vaccinations using AZD1222 and BNT162b2. Each dot represents one individual, while the boxplots show the medians, 25th and 75th percentiles, as well as the 95th percentile via the whiskers. The p-value resulting from the Kruskal–Wallis test was < 0.0001, and results from Dunn´s multiple comparisons test are indicated by asterisks and correspond to p < 0.05 (*), and p < 0.001 (***).

Figure 2

Heterologous prime-boost combining AZD1222 and BNT162b2 yielded highest neutralization capacities. Box plots indicate neutralizing capacities against the Wuhan-Hu-1 strain and the Delta Variant, respectively. Data are shown as percentages of neutralizing antibodies resulting from surrogate virus neutralization tests using RBD peptides. Asterisks indicate significant differences resulting from Kruskal-Wallis (p < 0.0001) with Dunn´s multiple comparisons test, with *** representing p < 0.001. The pound symbols (^#) indicate p-values resulting from Wilcoxon matched pairs signed rank tests.

Figure 3

Homologous prime-boost with BNT162b2 yielded the highest T cell responses. Dot plots show absolute numbers of IFNγ-producing ELISpots resulting from 5 × 10⁵ PBMCs stimulated with peptide pools representing the spike protein. Horizontal lines indicate means ± 2 SEM of IFNγ-producing ELISpots present in the absence of peptide stimulation. Each dot represents one individual and, in the case of more than mean + 2 SEM IFNγ-producing ELISpots, individuals were considered responders. p-values resulted from one-sample t-tests, comparing the numbers of ELISpots after peptide stimulation to the mean + 2 SEM numbers of ELISpots in the absence of stimulation. Pairwise comparisons via Wilcoxon matched pairs of unstimulated and stimulated numbers of ELISpots were 6.07 × 10⁻⁸ (AZD1222/AZD1222), 8.41 × 10⁻⁷ (BAZD1222/BNT162b2), and 4.49 × 10⁻⁹ (BNT162b2/BNT162b2). The table gives numbers of responders and non-responders resulting from each vaccination regimen. A chi-square test for independence confirms comparable proportions with a p-value of 0.9546.