Immunization with trivalent inactivated influenza vaccine in partially immunized toddlers

OBJECTIVE. Children >= 6 months of age who have previously received 1 dose of trivalent inactivated influenza vaccine are recommended to be given an additional single trivalent inactivated influenza vaccine dose the following fall. Limited data exist documenting the immunogenicity of 2 doses of influenza vaccine given in separate years to young children, and it is not known if the antigen content of each of the 2 doses of vaccine must be identical or similar to optimally immunize children in this age group. In 2004, the A/H3N2 and B antigens contained in trivalent inactivated influenza vaccine were changed from those in the 2003-2004 influenza vaccine, providing the opportunity to assess the effect of such a change on the single-dose recommendation in trivalent inactivated influenza vaccine-experienced toddlers. PATIENTS AND METHODS. We conducted an observational, nonrandomized, open-label study comparing immunogenicity and reactogenicity of 2 doses of trivalent inactivated influenza vaccine in 2 groups of healthy children aged 6 to 23 months. Children who had received 1 dose of 2003 trivalent inactivated influenza vaccine the previous season received 1 dose of 2004 trivalent inactivated influenza vaccine according to current guidelines (group 1). Trivalent inactivated influenza vaccine-naive toddlers received the standard 2 doses of 2004 trivalent inactivated influenza vaccine 1 month apart (group 2). Blood was obtained 4 weeks after the second dose of trivalent inactivated influenza vaccine. The primary outcome measure was antibody response to the 3 vaccine antigens in the 2004 trivalent inactivated influenza vaccine after 2 doses of vaccine, as determined by hemagglutination-inhibition antibody titers. Noninferiority of the antibody response was based on the proportion of subjects in each group achieving a titer of >= 1:32 postvaccination to antigens (H1N1, H3N2, and B) contained in the 2004-2005 vaccine. For each antigen, the antibody response was proposed to be noninferior if the upper bound of the 95% confidence interval of the difference between the proportion of children in the 2 groups with postvaccination titers >= 1:32 was < 15%. Reactogenicity was a secondary outcome and was assessed by parental diaries or telephone follow-up. RESULTS. Fifty six of 58 previously immunized children (group 1) and 63 of 64 vaccine-naive children (group 2) completed the study. The groups were similar, except group 1 was older at receipt of the second trivalent inactivated influenza vaccine. Reactogenicity did not differ by age or time between doses. Antibody responses to the unchanged influenza A/H1N1 antigen at 4 weeks after the second trivalent inactivated influenza vaccine dose were similar in both groups, with good responses as measured by geometric mean titer (75.2 vs 69.1) and percentage with antibody titers >= 1:32 (82.1% group 1 vs 85.7% group 2). For the A/H3N2 antigen, which changed between 2003 and 2004, there was a significantly higher geometric mean titer in group 1 compared with group 2 (156 vs 53.7), but both groups had very high rates of seroconversion that were not statistically different (91% vs 84%). The antibody response to influenza B was significantly lower in group 1 recipients, who received only a single dose of 2005 vaccine, as measured by both geometric mean titer and percentage with antibody >= 1:32. The group 1 geometric mean titer was 13.8, and the group 2 geometric mean titer was 49.1. Only 27% of children in group 1 achieved antibody levels >= 1:32 to influenza B compared with 86% in group 2. Using logistic regression, we also determined that older children had less potentially seroprotective levels to influenza B. Overall, noninferiority of the antibody response for group 1 compared with group 2 was confirmed for influenza A/H3N2, was marginally significant for A/H1N1, and was not confirmed for influenza B. CONCLUSIONS. The assessment of immune responses in children after changes in vaccine composition is important, because influenza vaccines change frequently, affecting not only antibody responses in partially immunized toddlers, but potentially immune responses in more fully immunized individuals. In this study, a change in 2 different vaccine antigens enabled us to assess and compare the impact of the original priming antigens after relatively minor changes in 1 antigen (A/H3N2) or after considerable antigenic changes in another vaccine antigen (B). Our subjects demonstrated relatively good responses to the vaccine antigen change characterized by relatively minor changes (A/H3N2). Circulating virus may have primed infants in both groups to antigen more closely related to the 2004 influenza A/H3N2 strain. The high A/H3N2 antibody response to the second dose of trivalent inactivated influenza vaccine in children who were immunized the previous fall with a different vaccine is consistent with the fact that more children in group 1 were alive during this epidemic and, therefore, were more likely to have experienced priming with natural infection. In contrast, a decreased antibody response to the influenza B antigen was seen in children primed with the earlier 2003 vaccine, suggesting that the major change in B virus lineage in the 2004 vaccine reduced the priming benefit of previous vaccination. Our findings are reminiscent of antibody responses in children seen after immunization with different but novel influenza antigens, such as swine flu vaccine (influenza A/swine/1976/37-like virus). Our results should be taken into account when evaluating new vaccines in young children for novel viruses, such as new pandemic strains of influenza. The need for multiple doses of vaccine to produce potentially protective antibody levels in children needs to be considered, even when vaccine is in short supply.