Poor Immunogenicity, Not Vaccine Strain Egg Adaptation, May Explain the Low H3N2 Influenza Vaccine Effectiveness in 2012-2013

Abstract

Background: Influenza vaccination aims to prevent infection by influenza virus and reduce associated morbidity and mortality; however, vaccine effectiveness (VE) can be modest, especially for subtype A(H3N2). Low VE has been attributed to mismatches between the vaccine and circulating influenza strains and to the vaccine's elicitation of protective immunity in only a subset of the population. The low H3N2 VE in the 2012-2013 season was attributed to egg-adaptive mutations that created antigenic mismatch between the actual vaccine strain (IVR-165) and both the intended vaccine strain (A/Victoria/361/2011) and the predominant circulating strains (clades 3C.2 and 3C.3).

Methods: We investigated the basis of low VE in 2012-2013 by determining whether vaccinated and unvaccinated individuals were infected by different viral strains and by assessing the serologic responses to IVR-165, A/Victoria/361/2011, and 3C.2 and 3C.3 strains in an adult cohort before and after vaccination.

Results: We found no significant genetic differences between the strains that infected vaccinated and unvaccinated individuals. Vaccination increased titers to A/Victoria/361/2011 and 3C.2 and 3C.3 representative strains as much as to IVR-165. These results are consistent with the hypothesis that vaccination boosted cross-reactive immune responses instead of specific responses against unique vaccine epitopes. Only approximately one-third of the cohort achieved a ≥4-fold increase in titer.

Conclusions: In contrast to analyses based on ferret studies, low H3N2 VE in 2012-2013 in adults does not appear to be due to egg adaptation of the vaccine strain. Instead, low VE might have been caused by low vaccine immunogenicity in a subset of the population.

Figure 1.

Maximum likelihood phylogeny of hemagglutinin (HA) sequences from 1339 influenza A(H3N2) isolates collected from North America during the 2012–2013 influenza season. Vaccination status of the individuals from whom the isolates were collected is noted (purple = vaccinated; orange = unvaccinated; blank = unknown vaccination status). Amino acid sites in which 20 or more of the 1339 specimens differed from the vaccine strain IVR-165 are noted, with the amino acids colored according to the key and annotated according to their location in HA1, HA2, and predicted epitope sites (A–E). The location from which the isolates were collected is color coded according to the key. Abbreviation: HA, hemagglutinin.

Figure 2.

Correlations between prevaccination titers (lower left panels, blue) and fold changes in titers (upper right panels, red). Fold changes are defined as the post-vaccination titer divided by the prevaccination titer. Points are semitranslucent; darker points represent multiple individuals. The dotted black line shows the diagonal (1:1) curve. On each plot, the x-axis corresponds to the strain for that column, and the y-axis corresponds to the strain for that row.