The Effect of Preexisting Immunity on Virus Detection and Immune Responses in a Phase II, Randomized Trial of a Russian-Backbone, Live, Attenuated Influenza Vaccine in Bangladeshi Children

Abstract

Background: In a 2012 Phase II clinical trial, 300 Bangladeshi children aged 24 to 59 months with no prior influenza vaccine exposure were randomized to receive a single intranasally-administered dose of either trivalent, Russian-backbone, live, attenuated influenza vaccine (LAIV) or placebo. Protocol-defined analyses, presented in the companion manuscript, demonstrate decreased viral detection and immunogenicity for A/H1N1pdm09, relative to the A/H3N2 and B strains. This post hoc analysis of the trial data aims to investigate the LAIV strain differences by testing the hypothesis that preexisting humoral and mucosal immunity may influence viral recovery and immune responses after LAIV receipt.

Methods: We used logistic regressions to evaluate the relations between markers of preexisting immunity (ie, hemagglutination inhibition [HAI], microneutralization, and immunoglobulin G and immunoglobulin A (both serum and mucosal antibodies) and LAIV viral recovery in the week post-vaccination. We then tested for potential effect modification by baseline HAI titers (ie, <10 versus ≥10) and week 1 viral recovery on the LAIV-induced serum and mucosal immune responses, measured between days 0 and 21 post-vaccination.

Results: Higher levels of preexisting immunity to influenza A/H3N2 and B were strongly associated with strain-specific prevention of viral shedding upon LAIV receipt. While evidence of LAIV immunogenicity was observed for all 3 strains, the magnitudes of immune responses were most pronounced in children with no evidence of preexisting HAI and in those with detectable virus.

Conclusions: The results provide evidence for a bidirectional association between viral replication and immunity, and underscore the importance of accounting for preexisting immunity when evaluating virologic and immunologic responses to LAIVs.

Clinical trials registration: NCT01625689.

Keywords: humoral immunity; immunogenicity; influenza vaccine; mucosal immunity.

Figure 1.

The association between preexisting immunity (ie, indicated by a tertile of the strain-specific immune marker on trial day 0) and subsequent protection from strain-specific viral shedding upon LAIV receipt (ie, indicated by no viral recovery from NPW specimens on trial days 2, 4, and 7; n = 145). Odds ratios for being shedding-negative (y-axis) are plotted versus the mean for each tertile of the immune marker at baseline (x-axis). The horizontal reference line indicates an odds ratio of 1. Abbreviations: LAIV, live, attenuated influenza vaccine; NPW, nasopharyngeal wash.

Figure 2.

A/H1N1pdm09-specific immune responses to placebo (n = 145) and LAIV receipt (n = 145). Of LAIV recipients, 100% were categorized as A/H1N1pdm09 shedding-negative (ie, indicated by no viral recovery on trial days 2, 4, and 7). The intensity of the color indicates the number of individuals at a given coordinate. P values are from paired t-tests comparing the log₂ titers on trial days 0 and 21. The diagonal reference line indicates equivalence between the time points. Abbreviation: LAIV, live, attenuated influenza vaccine.

Figure 3.

A/H3N2-specific immune responses to placebo (n = 145) and LAIV receipt (n = 145). Of LAIV recipients, 46% were categorized as A/H3N2 shedding-negative (ie, indicated by no viral recovery on trial days 2, 4, and 7). The intensity of the color indicates the number of individuals at a given coordinate. P values are from paired t-tests comparing the log₂ titers on trial days 0 and 21. The diagonal reference line indicates equivalence between the time points. Abbreviation: LAIV, live, attenuated influenza vaccine.

Figure 4.

B-specific immune responses to placebo (n = 145) and LAIV receipt (n = 145). Of LAIV recipients, 59% were categorized as A/H3N2 shedding-negative (ie, indicated by no viral recovery on trial days 2, 4, and 7). The intensity of the color indicates the number of individuals at a given coordinate. P values are from paired t-tests comparing the log₂ titers on trial days 0 and 21. The diagonal reference line indicates equivalence between the time points. Abbreviation: LAIV, live, attenuated influenza vaccine.

Figure 5.

Density plots illustrating the distributions of differences in the log₂ serum and mucosal antibody titers between trial days 0 and 21 by immunoassay, influenza strain, and treatment/shedding status (N = 290). P values are from likelihood ratio tests and indicate effect modification of the immune responses by shedding status (ie, shedding-negative versus shedding-positive). The vertical reference lines indicate 4-fold rises (ie, a log₂ difference ≥2) between the time points.