Improving influenza vaccine development with the pEpitope model: application to the 2018-19 season

The annual influenza vaccine has been shown to reduce flu-related hospitalizations and severe illness outcomes. Minimizing the antigenic differences between the vaccine strain and circulating strains ensures the vaccine will adequately prime the immune system against infection. We developed a theory of antibody response to infection following vaccination that produced a novel measure of antigenic distance. The model, called pEpitope, considers the modularity and hierarchy of antibody binding to the epitope regions of the viral hemagglutinin protein. The pEpitope model is able to explain over 90% of the variance in human vaccine epidemiological data from recent studies conducted by the US Centers for Disease Control and Prevention. Analysis of A(H3N2) strains circulating during the 2017-18 season identified the emergence of a new quasispecies cluster that is sufficiently distant from the 2018-19 vaccine and is therefore predicted to dominate. The pEpitope model is a valuable tool that predicts vaccine effectiveness to enhance vaccine strain selection and development.