Personalized prediction of SARS-CoV-2 vaccine-induced immunity after boost: a longitudinal analysis using joint modeling.

INTRODUCTION: The SARS-CoV-2 pandemic has revealed substantial inter-individual variability in immune responses, particularly following widespread primary vaccination and booster campaigns. These differences affect the durability of protective immunity and the need for additional booster doses. To optimize the management of current and future epidemics, there is a critical need for predictive tools that personalize immune monitoring and guide targeted booster strategies for vulnerable populations.
METHODS: In this study, we conducted a 15-month longitudinal analysis of a cohort of 1,000 individuals to identify key determinants of serological response following the first SARS-CoV-2 vaccine booster. We investigated how these factors influenced the risk of subsequent infection, and we developed statistical models to predict individual trajectories of anti-spike (S) IgG and neutralizing antibody (NAb) levels.
RESULTS-DISCUSSION: Our findings show that joint models (JMs), which integrate longitudinal antibody measurements with infection outcomes, significantly outperform traditional modeling approaches in predicting immune trajectories. This work underscores the potential of joint modeling to enable personalized immune surveillance, supporting strategies to sustain protective immunity in high-risk populations. In the future, this approach may be adapted for monitoring long-term immunity against other infectious diseases.