Genetic and Antigenic Characterization of Influenza Viruses
The 2017–18 influenza vaccine virus components were selected in March 2017, during one of two biannual WHO-sponsored vaccine consultation meetings to review influenza data generated by GISRS laboratories. The recommended Northern Hemisphere 2017–18 trivalent influenza vaccine composition consists of an A/Michigan/45/2015 (H1N1)pdm09-like virus, an A/Hong Kong/4801/2014 (H3N2)-like virus, and a B/Brisbane/60/2008-like (B/Victoria lineage) virus. An additional influenza B virus (B/Phuket/3073/2013-like [B/Yamagata lineage]) was recommended for quadrivalent vaccines.**** These recommendations reflect an update to the A(H1N1)pdm09 virus component to a more contemporary influenza A(H1N1)pdm09 virus (an A/California/7/2009 (H1N1)pdm09-like virus was replaced with an A/Michigan/45/2015 (H1N1)pdm09-like virus), compared with the recommendation for the Northern Hemisphere 2016–2017 influenza season and are the same as the vaccine virus recommendations made for the 2017 Southern Hemisphere influenza vaccine.
Most influenza vaccines licensed in the United States, with the exception of cell culture–based inactivated influenza vaccine (ccIIV4) and recombinant influenza vaccines (RIV3 and RIV4) are produced through propagation of candidate vaccine viruses (CVVs) in eggs. Historically, CVVs provided to manufacturers have been egg-derived. Egg propagation of influenza viruses, particularly influenza A(H3N2) viruses, often leads to genetic changes that might have antigenic implications. The vaccine viruses selected for the Northern Hemisphere 2017–18 vaccine were representative of most, but not all, circulating influenza viruses at that time, and had the fewest and least substantial egg-adapted changes. In August 2016, the Food and Drug Administration approved the use of cell-derived CVVs for inclusion in ccIIV4.†††† For the 2017–18 season, the influenza A(H3N2) component of this vaccine is manufactured using a cell-derived CVV. The other components of this vaccine are manufactured using egg-derived CVVs. Production of influenza vaccines using cell-grown CVVs and cell-based technology can circumvent antigenic changes that might be associated with egg propagation, particularly for influenza A(H3N2) viruses.§§§§
Data obtained from antigenic characterization are important in the assessment of the similarity between reference vaccine viruses and circulating viruses. In vitro antigenic characterization data acquired through hemagglutination inhibition (HI) assays or virus neutralization assays are used to assess whether genetic changes in circulating viruses affect antigenicity, which could affect vaccine effectiveness. Since the 2014–15 season, many influenza A(H3N2) viruses lack sufficient hemagglutination titers for antigenic characterization using hemagglutination inhibition assays. Therefore, representative influenza A(H3N2) viruses are selected for antigenic characterization using the virus neutralization focus reduction assay to assess the ability of various antisera to neutralize infectivity of the test viruses. For nearly all influenza-positive surveillance samples received by CDC, next generation sequencing (NGS), which employs genomic enrichment practices (5–7), adapted by CDC, Nextera library preparation (Illumina, San Diego, California) and NGS using MiSeq (Illumina, San Diego, California), is performed to determine the genetic identity of circulating viruses. The genomic data are analyzed and submitted to public databases (GenBank or GISAID EpiFlu). CDC has antigenically or genetically characterized 877 influenza viruses collected and submitted by U.S. and international laboratories since May 21, 2017, including 117 influenza A(H1N1)pdm09 viruses, 495 influenza A(H3N2) viruses, and 265 influenza B viruses.
Phylogenetic analysis of the HA genes from the A(H1N1)pdm09 viruses collected since May 21, 2017, showed that all but one were in subclade 6B.1, and one virus belonged to clade 6B (Figure 3). All A(H1N1)pdm09 viruses were antigenically similar (analyzed using HI with ferret antisera) to the 6B.1 virus A/Michigan/45/2015, the recommended influenza A(H1N1)pdm09 reference virus for the 2017 Southern Hemisphere and 2017–18 Northern Hemisphere influenza vaccines.
Four hundred ninety-five influenza A(H3N2) viruses collected globally since May 21, 2017, were sequenced, and phylogenetic analysis of the HA genes illustrated that multiple clades/subclades were cocirculating (Figure 3). The HA genes showed extensive diversity and belonged to clades 3C.2a or 3C.3a, with 3C.2a predominating (Figure 3). The 3C.2a and the 3C.2a1 subclade circulated in approximately equal proportions. A representative set of 153 influenza A(H3N2) viruses (51 international and 102 United States) were antigenically characterized, and most (97%) A(H3N2) viruses were well-inhibited (reacting at titers of less than or equal to fourfold of the homologous virus titer) by ferret antisera raised against A/Michigan/15/2014 (3C.2a), a cell propagated A/Hong Kong/4801/2014-like reference virus representing the A(H3N2) component of the 2017 Southern Hemisphere and 2017–18 Northern Hemisphere influenza vaccines. A smaller proportion (33%) of influenza A(H3N2) viruses were well-inhibited by antiserum raised against egg-propagated A/Hong Kong/4801/2014 reference virus representing the A(H3N2) vaccine component, which is likely because of egg-adaptive amino acid changes in the HA of the egg-propagated virus.
A total of 85 influenza B/Victoria-lineage viruses were phylogenetically analyzed, and all HA genes belonged to genetic clade V1A, the same genetic clade as the vaccine reference virus, B/Brisbane/60/2008. However, two deletion subclades were detected in 2017. One subclade has a 6-nucleotide deletion (encoding amino acids 162 and 163) and the other subclade has a 9-nucleotide deletion (encoding amino acids 162, 163 and 164). The 162–163 double deletion in the HA was detected in viruses circulating in multiple countries, with the majority identified in the United States, although the three viruses with 162–164 triple deletion were only detected in Hong Kong SAR, China. Thirty-nine (72%) B/Victoria lineage viruses were well-inhibited by ferret antisera raised against MDCK-propagated B/Brisbane/60/2008 reference virus, representing the B/Victoria lineage component of the 2017 Southern Hemisphere and 2017–2018 Northern Hemisphere influenza vaccines. However, 28% of B/Victoria lineage viruses reacted poorly with ferret antisera raised against MDCK-propagated B/Brisbane/60/2008, which correlated with the 162–163 double deletion and the 162–164 triple deletion in the HA.
Phylogenetic analysis of 180 influenza B/Yamagata-lineage viruses indicate that the HA genes belonged to clade Y3 (Figure 3). A total of 99 representative influenza B/Yamagata-lineage viruses (59 international and 40 United States) were antigenically characterized, and all were antigenically similar to B/Phuket/3073/2013, the reference vaccine virus representing the influenza B/Yamagata-lineage component of the 2017 Southern Hemisphere and 2017–18 Northern Hemisphere quadrivalent vaccines.
