Biostatistics Seminar Series - Peter Gilbert, PhD

Title:  Assessing immune correlates of protection in COVID-19 vaccine efficacy trials Abstract:  An objective of phase 3 randomized, placebo-controlled preventive COVID-19 vaccine efficacy trials is to assess immunological biomarkers measured after vaccination as various types of "immune correlates of protection," with applications including to identify predictors of COVID-19, modifiers of vaccine efficacy, and mediators of vaccine efficacy, and ultimately to define surrogate endpoints that can constitute the basis for traditional or accelerated approval of vaccines, as well as for making bridging predictions of vaccine efficacy to new conditions different from those in the phase 3 trial. In this talk I will summarize the current statistical analysis plan of the COVID-19 Prevention Network for assessing immune correlates of protection. Topics addressed include immunological biomarker sampling design, methods for combining biomarkers to develop best predictors of study outcomes, and four statistical frameworks for estimating causal effects that link immunological biomarkers to vaccine efficacy: (1) principal stratification for assessing vaccine efficacy moderation; (2) controlled effects analysis for comparing outcome risk under assignment to vaccination and a biomarker value vs. under assignment to placebo; (3) Stochastic interventional analysis comparing outcome risk under assignment to vaccination and a stochastic shift in the biomarker distribution vs. placebo; and (4) Mediation/natural direct and indirect effects analysis of the proportion of vaccine efficacy mediated though an immunological biomarker.