Files
Abstract
Introduction: Annual vaccines are the primary intervention for limiting the burden of influenza disease, but vaccine effectiveness (VE) depends on many factors, including how similar the strains used for vaccine formulation are to circulating influenza strains. Thus, ensuring that the vaccine induces a broadly-reactive immune response to many possible strains of influenza is paramount for improving VE. The heterologous immune response to strains other than those used in vaccine development is understudied and difficult to measure. Methods: We use data from a prospective, open annual influenza cohort across three study sites (Port St. Lucie, FL; Pittsburgh, PA; and Athens, GA) from Fall 2013 through Spring 2022. Each individual could repeat in subsequent years, but was not required to maintain consecutive enrollment. Every year, individuals contribute a pre-vaccination blood draw, receive a Fluzone vaccine (Sanofi) and contribute a post-vaccination blood draw at a followup visit targeted either 21 or 28 days post-vaccination. All serum samples were used for hemagglutination inhibition (HAI) assays to a wide panel of homologous and heterologous historical influenza strains. We fit Bayesian multilevel models to analyze the heterologous immune response and breadth of vaccine response. Results: First, we quantified the causal effect of vaccine dose (standard dose vs. high dose Fluzone) on the heterologous immune response. We found that higher dose improved the response to some strains, but not all, so higher dose can induce a less broad vaccine response. In a followup study, we were able to generalize our strain-specific predictions to predictions about antigenic distance using linked genomic data. Finally, we developed an overall summary metric for breadth of vaccine response which had substantially lower variation across data from different labs than current metrics. Conclusions: Many factors such as vaccine dose may modulate the heterologous immune response in unintuitive ways, so breadth must be carefully assessed. Breadth can be assessed using multiple measures of antigenic distance, but our novel summary metrics provide a robust measurement for understanding the breadth of response of vaccine candidates across different lab groups.