Agent-based Simulation of Antibody Response Durability: Linking Antibody-secreting Cell Dynamics to Vaccine-Induced Serological Memory

We present an agent-based model to simulate the durability of antibody responses, focusing on the roles of plasmablasts and plasma cells in antibody production. The model explores how the probability of plasmablasts differentiating into plasma cells and the extended half-life of plasma cells shape the kinetics of immune responses. Previously, our simulations identified two distinct patterns of antibody waning: sustainers, who maintain plasma cells and stable antibody levels, and decayers, who lose antibody-secreting cells, resulting in a steep antibody decline. Here, we extend the model to quantify how changes in antibody-secreting cell dynamics—such as the likelihood of plasmablasts differentiating into plasma cells, the half-life of plasma cells, and the magnitude of the initial response—affect the frequency of decayers and the overall immune response. Additionally, we derive a mathematical framework to describe the trajectory of antibody levels, which accurately captures the dynamics of the antibody response against the SARS-CoV-2 Spike protein (Geometric Mean Titer) in a cohort of 5,834 individuals who received the adenoviral COVID-19 vaccine ChAdOx1 and 3,363 who received the mRNA vaccine BNT162b2. Our findings provide a quantitative basis for understanding variability in antibody response durability and its implications for vaccine design.