Strain development of A/H1N1pdm09 candidate vaccine viruses for the 2021-22 northern hemisphere influenza season.

The vaccine effectiveness (VE) of the A/H1N1pdm09 component of the 2017-18 quadrivalent live attenuated influenza vaccine (QLAIV) was improved by performing rational haemagglutinin (HA) mutagenesis. Introducing N125D, D127E, D222G and R223Q substitutions into the HA protein of A/Slovenia/2903/2015 (A/SLOV15) enhanced replicative fitness in primary human nasal epithelial cells (hNECs). This allowed A/SLOV15 to overcome inter-strain competition in QLAIV, resulting in improved VE.During strain development for the 2021-22 QLAIV formulation, A/H1N1pdm09 LAIV viruses containing wild-type (WT) HA and neuraminidase (NA) sequences were found to replicate poorly in embryonated eggs and hNECs. We aimed to enhance replicative fitness via the HA mutagenesis approach that was performed previously for A/SLOV15. Therefore, combinations of these four mutations were introduced into the HA protein of representative 6B.1A-5a.2 viruses, A/Victoria/2570/2019 and A/Victoria/1/2020 (A/VIC1). Replicative fitness of A/VIC1 V7 was improved ~30-fold in eggs and ~300-fold in hNECs relative to its parent, without compromising other critical LAIV characteristics.

Replication of live attenuated influenza vaccine viruses in human nasal epithelial cells is associated with H1N1 vaccine effectiveness.

In the 2013-2014 and 2015-2016 influenza seasons, live attenuated influenza vaccine (LAIV) generated reduced vaccine effectiveness (VE) against circulating H1N1 strains. This reduced VE coincided with the introduction of pandemic 2009 H1N1 (A/H1N1pdm09) vaccine virus reassortants, in place of pre-2009 seasonal H1N1 strains. Here, we explored one specific hypothesis for reduced VE; decreased replicative fitness of A/H1N1pdm09 strains in humans. Two A/H1N1pdm09 strains with reduced VE, A/California/07/2009 (A/CA09) and A/Bolivia/559/2013 (A/BOL13), were compared to pre-2009 seasonal H1N1 strains, A/New Caledonia/20/1999 (A/NC99) and A/South Dakota/6/2007 (A/SD07). Initial results showed that A/H1N1pdm09 strains had reduced multi-cycle infectivity in Madin-Darby Canine Kidney (MDCK) cells, compared to their pre-2009 counterparts. The A/BOL13 viral titre was found to be 2.65 log10/mL lower when measured by multi-cycle 50% tissue culture infectious dose (TCID50) assay compared to single-cycle fluorescent focus assay (FFA). By contrast, clinically effective A/NC99 titres differed by only 0.54 log10/mL. In human alveolar (A549) cells, A/H1N1pdm09 strains replicated less than pre-2009 strains, with A/CA09 and A/BOL13 generating lower peak viral titres over 5 days. This phenotype was corroborated in physiologically relevant, primary human nasal epithelial cells (hNECs). Here, peak titres for pre-2009 strains A/NC99 and A/SD07 were 8.43 log10 TCID50/mL and 8.52 log10 TCID50/mL, respectively, versus 6.89 log10 TCID50/mL and 6.06 log10 TCID50/mL for A/H1N1pdm09 strains A/CA09 and A/BOL13. This confirmed a reduced ability of A/H1N1pdm09 strains to sustain replication in human respiratory cells. Using this information, H1N1 candidate A/Slovenia/2903/2015 (A/SLOV15) was characterised for replacement of A/BOL13 in the 2017/18 LAIV. A/SLOV15 produced comparable single and multi-cycle infectivity titres (Δ 0.16 log10/mL) and reached a peak titre 1.23 log10 TCID50/mL higher than that of A/BOL13 in hNEC cultures. Taken together, these data suggest a reduction in sustained multi-cycle replication in human cells as a plausible root cause for reduced A/H1N1pdm09 VE.

Haemagglutinin stability was not the primary cause of the reduced effectiveness of live attenuated influenza vaccine against A/H1N1pdm09 viruses in the 2013-2014 and 2015-2016 seasons.

During the 2013-2014 influenza season, the quadrivalent live attenuated influenza vaccine (QLAIV), had lower than expected vaccine effectiveness (VE) against circulating A/H1N1pdm09 viruses in the USA. The underlying reason proposed for this was that the A/H1N1pdm09 vaccine strain, A/California/07/2009 (A/CA09), had a thermally unstable haemagglutinin (HA) protein. Consequently, a new A/H1N1pdm09 candidate strain, A/Bolivia/559/2013 (A/BOL13), was developed for inclusion in the 2015-2016 QLAIV. A key parameter for selection of A/BOL13 was its more thermostable HA phenotype compared with A/CA09. During the 2015-2016 season, QLAIV containing A/BOL13 was found in some studies to have improved, but still with suboptimal, VE against circulating A/H1N1pdm09 viruses and was not recommended for use by the CDC in the US market in the 2016-2017 influenza season. This suggested that improved HA thermostability had not entirely resolved the reduced VE observed. One hypothesis for this was that, by improving thermostability, the A/BOL13 HA protein had been over-stabilised, compromising its activation at the low endosomal pH required for successful viral entry. Here we demonstrate that, while the A/BOL13 HA protein is more stable than that of A/CA09, its thermal and pH stability were comparable with historically efficacious LAIV strains, suggesting that the HA had not been over-stabilised. Furthermore, studies simulating potential heat exposure during distribution by exposing QLAIV nasal sprayers to 33 °C for 4 h showed that, while remaining within product specification, A/CA09 viral potency was statistically decreased after 12 weeks at 2-8 °C. These data suggest that although unfavourable HA protein stability may have contributed to the reduced VE of A/CA09 in 2013-2014, it was unlikely to have affected A/BOL13 in 2015-2016. We conclude that HA stability was not the primary cause of the reduced effectiveness of LAIV against A/H1N1pdm09 viruses in the 2013-2014 and 2015-2016 seasons.