Seasonal influenza vaccine strain selection by quantifying viral fitness

Each year, seasonal influenza vaccines often fail to provide adequate protection, a problem linked to the virus's swift and ongoing evolution. Ideally, vaccine strain recommendations aim to predict the dominant strain predicted to have the highest fitness level during the forthcoming season. To address this issue, researchers have developed FutureFlu, a framework grounded in biological principles.

FutureFlu determines the fitness score of influenza variants for future seasons by combining three distinct metrics: molecular genetic divergence, individual immune escape, and population-scale transmission dynamics. The findings reveal that the fitness scores of influenza variants have a strong positive correlation with their actual observed frequencies in the subsequent seasons.

Researchers tested FutureFlu across 24 seasons involving three influenza subtypes, observing that the model recommends antigenically matched vaccine strains significantly more often than the traditional annual recommendations. Specifically, FutureFlu accurately predicted A/H3N2 strains in 83.3% of seasons, compared to 45.8% by conventional methods, and B/Victoria strains in 75.0% of seasons, compared to 33.3% by conventional methods.

FutureFlu also outperforms existing methods for vaccine strain selection, even when antigenic data from hemagglutination inhibition (HI) assays are unavailable. This valuable supplement can assist the World Health Organization (WHO) in making informed vaccine recommendations. To facilitate global public health implementation, researchers have established an open online platform at futureflu.com.cn, providing real-time viral fitness predictions and vaccine suggestions.