A novel approach for predicting the efficacy of updated vaccines when the correlate of protection is known

Objectives: Traditional vaccine efficacy trials enroll tens of thousands of patients and may take months to complete. For vaccines that are regularly updated to address antigenic change due to evolution – such as influenza and SARS-CoV-2 vaccines – these resource- and time-intensive efficacy trials become infeasible. As a result, little is known about the efficacy of these vaccines at their time of deployment. In this poster, we propose a novel, efficient trial design for assessing vaccine efficacy using a population PK/PD modeling approach. Methods: In our proposed trial design, we fit a population PK/PD model to longitudinal immunological correlate data in a relatively small number of patients. Rather than assessing infection frequency, we use the PK/PD model to project vaccine efficacy based on the relationship between the immunological correlate and protection from disease. To test our trial design, we performed trial simulations using virtual patients generated from a population PK model fitted in Phoenix NLME to neutralizing antibody kinetics data after administration of the Spikevax SARS-CoV-2 vaccine1. Neutralizing antibody titers have been demonstrated to predict protection from symptomatic SARS-CoV-2 infection2. Thus, it is possible to bridge from these antibody titers to estimates of vaccine efficacy in terms of protection from symptomatic COVID-19. In simulated trials enrolling 9 to 50 patients, we projected population-level protection from COVID-19 while accounting for uncertainty in the population PK model and the PK/PD relationship. Results: In our trial simulations, we demonstrate that Spikevax efficacy against symptomatic COVID-19 could have been projected to exceed 50% with 95% certainty with a PK/PD-model-based trial of only 24 patients. Conclusions: This efficient trial design could be used to determine the efficacy of updated vaccines as well as support optimization of vaccine dose, schedule, and adjuvants.