Antibody kinetics following mRNA booster vaccination as correlates of protection against SARS-CoV-2 infection

Abstract

Introduction

Understanding the dynamics of SARS-CoV-2 antibody levels post-booster vaccination is crucial for optimizing vaccination strategies, as higher immunoglobulin responses are considered to correlate with reduced infection risk.

Material and Methods

Data were collected longitudinally from a cohort of healthcare workers at four time points: day of booster vaccination and 28, 180, and 360 days after. We utilized non-linear mixed effects models to map the trajectory of antibody kinetics for immunoglobulin A and G (IgA and IgG) binding (%) against the wild-type (WT) and Omicron BA.1 spike protein. Participants’ age, sex, type of mRNA vaccine, and timing of infection were considered as potential covariates, and best-fit models were chosen based on lowest statistical criteria. We then analyzed the association between each antibody isotype and risk of SARS-CoV-2 infection in this cohort through Cox proportional hazards models, and predicted protection against infection.

Results

The best-fit antibody kinetics models showed that decay rates post-booster and growth rates post-infection were higher with Pfizer boosters compared to Moderna. Binding antibody growth rates post-infection were higher in the elderly, females, and those with late infections. In our survival analyses, after adjustment for age, sex, vaccine type, and time-varying COVID-19 incidence, having high (above 75th percentile) antibody levels for WT IgG and BA.1 IgA at booster response (day 28) showed significant reductions in infection risk; hazard ratio (HR) = 0.47 (0.22-0.98) and 0.36 (0.17-0.78) compared with low antibody levels (below 25th percentile), respectively. To account for antibody kinetics, we used time-varying antibody levels and found similar results, with HRs of 0.22 (0.07-0.74) per percent increase in binding against WT IgG and 0.09 (0.01-0.66) against BA.1 IgA, but with higher prediction ability (AIC and BIC). On average, ≥80% booster vaccination-elicited protection against infection lasts 155 days.

Discussion

In conclusion, antibody kinetics of WT IgG and BA.1 IgA are correlated with reduced infection risk. To predict infection risk with high accuracy, follow-up measurements of antibody levels are essential. Duration of protection estimates may be used to inform frequency of booster vaccinations.