https://orcid.org/0000-0002-6498-9212
To the Editor—We read with interest the findings from Matrajt et al [1], who propose a novel approach to live attenuated influenza vaccine (LAIV) strain selection based on results from mathematical models. They suggest that vaccine strains should be antigenically distant enough from preexisting immunity to allow vaccine replication, while remaining antigenically close enough to circulating strains to protect from infection. We fully agree that this approach is worth exploring, not only for strain selection, but also for decisions about the optimal age and frequency of childhood immunization with LAIV. However, there are a number of complexities to LAIV immunogenicity worth highlighting that cannot be ignored if such strategies were to be used in practice.
The proposed model assumes that preexisting serum antibodies from prior infection or vaccination are the main driver of LAIV “take” and immunogenicity. Our own data using Russian-backbone LAIV in Gambian children aged 24–59 months partially support this, with preimmunization serum hemagglutination inhibition (HAI) titer (but not T-cell response or mucosal immunoglobulin A [IgA]) being the key determinant of LAIV shedding [2].
We have recently stratified this cohort based on seropositivity to the H3N2 strain included in the vaccine (A/Hong Kong/4801/2014 [A/HK]) and 2 older H3N2 strains antigenically similar to those potentially encountered during the children’s lifetime (A/Switzerland/9715293/2013 [A/Sw] and A/Texas/50/2012 [A/Tex]). Children who were A/HK seropositive (HAI titer ≥1:10) at baseline (n = 169) were less likely to shed H3N2 than those seropositive to A/Sw or A/Tex (but A/HK seronegative; n = 28) or those seronegative (HAI titer <1:10) to all 3 H3N2 strains (n = 45) (Figure 1A and 1B)). Shedding did not differ significantly between seronegative children and those seropositive to A/Sw or A/Tex only.
![Shedding and immunogenicity to Russian-backbone live attenuated influenza vaccine (LAIV) in Gambian children aged 24–59 months, stratified by preexisting serum antibodies to H3N2 strains. A, Percentage of children shedding H3N2 at day 2 after LAIV. B, Cycle threshold (Ct) from H3 hemagglutinin (HA)–specific reverse-transcriptase polymerase chain reaction assays on nasopharyngeal swab samples as a marker of viral load on day 2 after LAIV. C, Immune responses to LAIV determined by responses on day 21 after immunization (definitions provided elsewhere [2]). T-cell response denotes H3 HA–specific CD4+ T cells producing interferon γ and/or interleukin 2. Seronegativity was defined as a hemagglutination inhibition titer <1:10 to A/Hong Kong/4801/2014 (A/HK), A/Switzerland/9715293/2013 (A/Sw), and A/Texas/50/2012 (A/Tex). Proportions were compared using the Fisher exact test, and Ct values were compared using the Kruskal-Wallis test with Dunn posttest. *P < .05; †P < .001. Abbreviations: A/HK+, A/HK seropositive; A/HK−, A/HK seronegative; A/Sw+, A/Sw seropositive; A/Tex+, A/Tex seropositive; IgA, immunoglobulin A; NS, not significant at P < .05.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/cid/70/10/10.1093_cid_ciz773/1/m_ciz773f0001.jpeg?Expires=1750868384&Signature=O0MFQmgfg84kXwORbm-uD~rl88v0vpBYLIPo9kmQPDGjBPeVuvUmCUJUNNut~tCZ4SvbRDU5aZTOtBlIeRTr0zf6MJ4tJKomTZyHKHSYDP~6Nrn-dNBxFl-5j0hpmpbPFa0Fhj5nc77QEni6F7T9Gg0iovZmMyLz1GJh0ySP97ccaldYf-TZLNUI3q58yjqaZrdxF7fyKsQVuZ60vCBNKTTXtGfDJE8sK~i2rTqt6hpThMvzGNRwCbZ0JLKMkjFuRUC4BT3eLaj8TiEPoLF2HbC5C0OL3qwIq0g8saLFmrLbyHq3x94EbXMluxlCWH3wE6TW20gQhAiHLmqDvdDtfQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Shedding and immunogenicity to Russian-backbone live attenuated influenza vaccine (LAIV) in Gambian children aged 24–59 months, stratified by preexisting serum antibodies to H3N2 strains. A, Percentage of children shedding H3N2 at day 2 after LAIV. B, Cycle threshold (Ct) from H3 hemagglutinin (HA)–specific reverse-transcriptase polymerase chain reaction assays on nasopharyngeal swab samples as a marker of viral load on day 2 after LAIV. C, Immune responses to LAIV determined by responses on day 21 after immunization (definitions provided elsewhere [2]). T-cell response denotes H3 HA–specific CD4+ T cells producing interferon γ and/or interleukin 2. Seronegativity was defined as a hemagglutination inhibition titer <1:10 to A/Hong Kong/4801/2014 (A/HK), A/Switzerland/9715293/2013 (A/Sw), and A/Texas/50/2012 (A/Tex). Proportions were compared using the Fisher exact test, and Ct values were compared using the Kruskal-Wallis test with Dunn posttest. *P < .05; †P < .001. Abbreviations: A/HK+, A/HK seropositive; A/HK−, A/HK seronegative; A/Sw+, A/Sw seropositive; A/Tex+, A/Tex seropositive; IgA, immunoglobulin A; NS, not significant at P < .05.
However, the immunogenicity data reveal a more complex pattern. Although HAI seroconversion is significantly lower in A/HK-seropositive children and reflects patterns of shedding, the T-cell and mucosal IgA responses follow different trends (Figure 1C). A ≥2-fold CD4+ T-cell response is generated most often in children seronegative to all 3 strains, but at similar levels in A/HK-seropositive or A/Sw-seropositive/A/Tex-seropositive (A/HK-seronegative) children. In addition, prior serostatus has no impact on the likelihood of a ≥2-fold mucosal IgA response. This is in keeping with our findings that shedding increases the odds of seroconversion and a T-cell response, but not mucosal IgA responses [2].
Furthermore, approximately 50% of children generated a T-cell response to A/HK hemagglutinin despite being A/HK seropositive, challenging the view that giving LAIV may not be worthwhile in the face of preexisting serum antibody immunity to vaccine strains. Similar T-cell data are observed for matrix and nucleoprotein (data not shown). The interpretation of any such data is of course limited by the lack of a correlate of protection for LAIV, although both T cells and IgA have been shown to be protective in the absence of seroconversion and are likely to be cocorrelates of protection, along with serum antibodies [3–7].
The approach taken by Matrajt et al [1] to improve LAIV effectiveness is promising. Well-designed immunogenicity studies in children to both inform and test further models could help refine our approach to LAIV use in the future.
Notes
Financial support. This work was supported by the Wellcome Trust (Intermediate Clinical Fellowship 110058/Z/15/Z to T. I. d. S.).
Potential conflicts of interest. All authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
