Anti-Influenza Serum and Mucosal Antibody Responses after Administration of Live Attenuated or Inactivated Influenza Vaccines to HIV-Infected Children

Title of Contribution: Anti-Influenza Serum and Mucosal Antibody Responses after Administration of Live Attenuated or Inactivated Influenza Vaccines to HIV-Infected Children
Author(s): Adriana Weinberg1, Lin-Ye Song2, Robert Walker3, Maria Allende3, Terence Fenton2, Julie Patterson-Bartlett1, Sharon Nachman4, George Kemble3, Ting-Ting Yi3, Patricia Defechereux5, Diane Wara5, and Myron Levin1*
Affiliation(s): 1U CO Denver, Aurora, CO; 2Harvard Sch Publ Health Ctr for Biostat in AIDS Res, Boston, MA; 3MedImmune, Inc, Gaithersburg, MD; 4Stony Brook NY State U, Stony Brook, NY; 5UCSF, San Francisco, CA.

Background: Cold adapted influenza vaccine (CAIV) significantly reduces cases of influenza in immune competent children compared to trivalent inactivated influenza vaccine (TIV) during seasons when the vaccine and circulating viral strains are matched and not well matched. We studied both vaccines in HIV-infected children by measuring the magnitude, persistence, and hetero-subtypic profile of serum and mucosal anti-influenza antibodies from HIV-infected children immunized with CAIV or TIV. The CAIV used in this study has not been registered and is not available outside of the US. Methods: 243 HIV-infected children (≥5 to < 18 years of age) on stable antiretroviral therapy were randomized to receive either CAIV or TIV. Blood and saliva were obtained at enrollment and at 4 and 24 weeks post-immunization (wpi), and assayed for anti-influenza HAI and neutralizing antibodies in serum and anti-influenza IgA and IgG antibodies in saliva. Results: HAI titers were measured against vaccine strains (A H1N1 New Caledonia, A H3N2 Wyoming and B Jilin); and against mismatched strains (A H3N2 Sydney and B Yamanashi). At baseline, HAI geometric mean titers (GMTs) varied from 13 (B Yamanashi) to 90 (A H3N2 Wyoming). Both CAIV and TIV significantly increased the HAI titers at 4 and 24 wpi against vaccine and mismatched strains. HAI titers were higher in TIV compared with CAIV recipients at all time points and for all strains except A H1N1 New Caledonia at 24 wpi. At baseline, neutralizing antibody (neut) GMTs varied from 62 for influenza B Jilin to 1751 for influenza A H3N2 Wyoming. At 4 wpi, 4-fold increases in neut titers against each of the 3 strains in the vaccines were observed in 12% to 56% of subjects. Neut GMTs were higher in TIV vs. CAIV recipients both at 4 and 24 wpi. Anti-influenza salivary IgG (all 3 vaccine strains combined), adjusted for total salivary IgG, significantly increased both in CAIV and in TIV recipients from medians of 3 and 2 EU/µg/ml, respectively, at baseline, to 4 and 7 EU/µg/ml, respectively, at 4 wpi; and 3 and 5 EU/µg/ml, respectively, at 24 wpi. At 4 wpi, but not at 24 wpi, influenza-specific salivary IgG levels were significantly higher in TIV vs. CAIV recipients. Anti-influenza salivary IgA was detected in a minority of subjects at baseline and did not appreciably increase at 4 wpi in either group of vaccinees. HAI titers significantly correlated with neut and anti-influenza-salivary IgG at all time points, but with anti-influenza-IgA only at baseline.  On day 3 pi, 22 CAIV recipients shed the vaccine strain A H1N1 and 11 shed the vaccine strain B. Compared with non-shedders, H1N1 shedders had significantly lower baseline H1N1-specific HAI and neut titers. An inverse relationship between serum antibodies and influenza B shedding could not be demonstrated, possibly because of the lower number of shedding events. Baseline anti-influenza-salivary IgG and IgA were significantly inversely associated with vaccine virus shedding. Conclusions: CAIV and TIV increased anti-influenza serum and salivary IgG antibodies of HIV-infected children against the vaccine and mismatched strains. Baseline serum and salivary IgG and IgA antibodies were inversely associated with vaccine virus shedding in CAIV recipients.