Quantitative SARS-CoV-2 anti-spike responses to Pfizer-BioNTech and Oxford-AstraZeneca vaccines by previous infection status

Abstract

Objectives: We investigated determinants of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) anti-spike IgG responses in healthcare workers (HCWs) following one or two doses of Pfizer-BioNTech or Oxford-AstraZeneca vaccines.

Methods: HCWs participating in regular SARS-CoV-2 PCR and antibody testing were invited for serological testing prior to first and second vaccination, and 4 weeks post-vaccination if receiving a 12-week dosing interval. Quantitative post-vaccination anti-spike antibody responses were measured using the Abbott SARS-CoV-2 IgG II Quant assay (detection threshold: ≥50 AU/mL). We used multivariable logistic regression to identify predictors of seropositivity and generalized additive models to track antibody responses over time.

Results: 3570/3610 HCWs (98.9%) were seropositive >14 days post first vaccination and prior to second vaccination: 2706/2720 (99.5%) were seropositive after the Pfizer-BioNTech and 864/890 (97.1%) following the Oxford-AstraZeneca vaccines. Previously infected and younger HCWs were more likely to test seropositive post first vaccination, with no evidence of differences by sex or ethnicity. All 470 HCWs tested >14 days after the second vaccination were seropositive. Quantitative antibody responses were higher after previous infection: median (IQR) >21 days post first Pfizer-BioNTech 14 604 (7644-22 291) AU/mL versus 1028 (564-1985) AU/mL without prior infection (p < 0.001). Oxford-AstraZeneca vaccine recipients had lower readings post first dose than Pfizer-BioNTech recipients, with and without previous infection, 10 095 (5354-17 096) and 435 (203-962) AU/mL respectively (both p < 0.001 versus Pfizer-BioNTech). Antibody responses >21 days post second Pfizer vaccination in those not previously infected, 10 058 (6408-15 582) AU/mL, were similar to those after prior infection followed by one vaccine dose.

Conclusions: SARS-CoV-2 vaccination leads to detectable anti-spike antibodies in nearly all adult HCWs. Whether differences in response impact vaccine efficacy needs further study.

Keywords: Antibody; Quantitative anti-spike antibody; SARS-CoV-2; Serology; Vaccine.

Fig. 1

Anti-spike IgG-positive results by days since first vaccination, by prior infection status and vaccine received. Tests performed after a second dose of vaccine are not included. The number of tests performed and positive and the resulting percentage is shown under each bar.

Fig. 2

The relationship between vaccine, age and probability of testing anti-spike IgG seropositive >14 days post first vaccination. Model predictions are shown using reference categories for sex and ethnicity (white, female, respectively) and in those without prior evidence of infection.

Fig. 3

Modelled quantitative anti-spike IgG responses following first vaccination by vaccine and previous infection status. Panels A and B show responses in previously infected healthcare workers (HCWs) and panels C and D HCWs without evidence of previous infection. Panels A and C show data for those receiving Pfizer–BioNTech vaccine and panels B and D Oxford–AstraZeneca vaccine. Model predictions are shown at three example ages: 30, 45, and 60 years. The shaded ribbon shows the 95% confidence interval. Values are plotted from 7 days prior to vaccination to illustrate baseline values (models are fitted using data from 28 days prior to vaccination onwards).

Fig. 4

Modelled quantitative anti-spike IgG titres following second Pfizer–BioNTech vaccination by previous infection status. Panel A shows those who were previous infected (including those previously infected at baseline or testing PCR-positive between vaccines) and panel B those who had no evidence of previous infection. Model predictions are shown at three example ages: 30, 45, and 60 years. The shaded ribbon shows the 95% confidence interval. Data were included in each model from 7 days before the second vaccination to allow pre-vaccination levels to be fitted correctly.