Abstract Body

While RV144 vaccine-induced IgG3 to HIV-1 Env variable loops 1 and 2 (V1V2) correlated with decreased risk of HIV-1 acquisition, the responses were not durable. We studied IgG subclasses (IgG1-IgG4) in HIV-uninfected RV144 volunteers who completed their primary vaccination series 6-8 years earlier and were randomized to receive booster vaccinations with either ALVAC-HIV/AIDSVAX B/E (Gr1), AIDSVAX B/E (Gr2) or ALVAC-HIV (Gr3) at 0 and 6 months (RV305).

ELISA IgG1-IgG4 to HIV gp120 A244gD, and gp70V1V2 scaffold proteins 92TH023 (subtype AE) and CaseA2 (subtype B) were assessed in plasma and cervico-vaginal mucus (CVM) at study entry, 2 weeks post 1st and 2nd boosts, and 6 and 12 months post 2nd boost.

At study entry, geometric mean titers (GMT) of IgG to gp120 A244gD in plasma ranged from 50-67 in all groups. We did not detect a boosting effect of any IgG subclasses in plasma and CVM in Gr3. IgG1 GMT increased 2 weeks post 1st boost in Gr1 and Gr2, but significantly declined (p<0.01) 6 months post 2nd boost (Gr1=4850 and 255; Gr2=5279 and 252, respectively). They were significantly higher than those in RV144 at peak immunogenicity and 6 months post last injection (492 and 88, p<0.01). IgG3 GMT increased 2 weeks post 1st boost (Gr1=168, Gr2=159), but not after the 2nd boost, and were significantly lower than RV144 at peak immunogenicity  (532, p<0.01). While IgG4 was undetected in RV144, it increased to 951 and 611 and to 800 and 588 2 weeks post 1st and 2nd boosts in Gr1 and Gr2, respectively. IgG1-IgG4 to gp70V1V2 scaffolds was not detected at study entry. IgG1 GMT to AE and B V1V2 were similar in Gr1 and Gr2 after the 1st boost (Gr1-Gr2/AE=528-524 and Gr1-Gr2/B=50-43) but lower after the 2nd boost (Gr1-Gr2/AE=146-147 and Gr1-Gr2/B=28-24). IgG4 to AE and B V1V2 were detected in Gr1 and Gr2 but not detected in RV144. RV144 IgG3 to V1V2 scaffolds were not boosted in Gr1 and Gr2, and IgG2 remained weak for all antigens tested. In CVM, most responses to gp120 A244gD were IgG1 and IgG4, while to scaffolds were mostly IgG1.

Late boosts with AIDSVAX B/E, with or without ALVAC-HIV, induced plasma HIV-specific IgG responses that were predominantly IgG1 and IgG4, were significantly higher than RV144 responses, but did not improve magnitude nor durability of IgG3, and were similar to subtypes in CVM. These studies highlight the need for new strategies to durably boost HIV-specific, protective antibody responses observed in RV144.