CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

contribution of V3-specific antibodies to tier 1 virus (SF162) neutralization in TZM-bl cells. A regression model was used to investigate the association between neutralization titers and MTCT risk. Results: The maternal plasma V3-specific IgG response against the clade B peptide was high in magnitude (median MFI 23,177), but, binding was abolished (median MFI < 100) against the CRF1 V3 peptide, which differed from clade B peptide at 3 amino acid residues (K305T, I307T and A317R). Moderate plasma binding was observed against the clade A peptide (median MFI 8,739), which differed from the clade B peptide at positions I307V and H308R. Based on the shared amino acid differences between the V3 peptides, positions 305, 307, 308, and 317 flanking the V3 loop tip appear to be critical for maternal plasma V3 binding. Maternal plasma tier 1 virus neutralization significantly decreased in the presence of SF162 V3 as compared to the scrambled peptide and no peptide (median ID 50 671, 1399, 2930, for SF162 V3 peptide, scrambled peptide, and no peptide, respectively; p<0.0001 for both). The median plasma tier 1 virus neutralization inhibition with the V3 peptide was 50% (IQR 9-75%), yet the proportion of the maternal plasma neutralization inhibited by the V3 peptide did not predict MTCT risk. Conclusions: Our results suggest that in clade B HIV-1-infected women, V3 is an important tier 1 virus neutralization epitope that is dependent on residues that flank the V3-loop tip for IgG binding. Yet, additional epitope specificities may also contribute to the maternal neutralization responses associated with decreased MTCT risk. 905 Broad, Highly Avid Vaccine-Elicited Anti-V1V2 IgG Responses in HIV-Exposed Infants Genevieve Fouda 1 ; Coleen Cunningham 1 ; Elizabeth McFarland 2 ; Bill Borkowsky 3 ; NicoleYates 1 ; Erin McGuire 1 ; Hua-Xin Liao 1 ; Barton Haynes 1 ; GeorgiaTomaras 1 ; Sallie Permar 1 1 Duke University, Durham, NC, US; 2 University of Colorado, Aurora, CO, US; 3 New York University, New York, NY, US Background: We have previously reported that infant vaccination can induce high magnitude and durable HIV-1 envelope (Env)-specific IgG responses, including against the Env V1V2 loops. Importantly, in the trial of the moderately effective RV144 vaccine, anti-V1V2 IgG responses were associated with reduced risk of HIV acquisition. To further characterize this potentially-protective response in infants, we measured the breadth, IgG subclass distribution, and avidity of vaccine-elicited anti-V1V2 IgG. Methods: In PACTG 230, HIV-exposed infants were immunized with 4 doses of SF-2 rgp120+MF-59 (Chiron vaccine) or placebo between 0 and 20 weeks of age; 49 Chiron and 18 placebo recipients were included in the present study. Binding antibody multiplex assay (BAMA) was used to measure IgG responses against A244 V1V2 (clade AE), gp70 1086C V1V2 (clade C), gp70 V1V2 B case A2 (clade B), gp70 B case A2 B case A2 V169K, gp70 B case A2 mut3 (V169K/E172V/E173V mutations), and gp70 V1V2 (A) (clade A). IgG1 and IgG3 responses were measured against gp70 B case A2 V1V2 by BAMA. Avidity was measured by ELISA and the avidity index (AI) calculated by dividing the optical density in presence of urea by that without urea. Results: At birth, vaccine and placebo recipients had similar maternally-acquired anti-V1V2 IgG levels. At peak immunogenicity (week 24), vaccinees had higher frequency and magnitude multiclade anti-V1V2 IgG responses than placebo recipients (p <0.0001 for all antigens). Moreover, the magnitude of IgG responses significantly increased between weeks 0 and 24 in vaccine but not placebo recipients for all antigens except gp70 V1V2 (A), indicating that the vaccine induced antibodies that recognize the V1V2 region across clades. IgG responses against all V1V2 constructs except g70 V1V2 (A) remained higher at week 52 in vaccinees than in placebo recipients (p values from<0.0001 to 0.0083). The majority of infants had detectable levels of anti-V1V2 IgG1 antibodies (week 24: 89%), but 47% also had anti-V1V2 IgG3 at week 24. However the IgG3 response was short-lived. There was no difference in the avidity of the anti-V1V2 IgG that was maternally-acquired (birth) and vaccine-elicited (week 24 in vaccine recipients), with median AI: 0.89 and 0.78 at week 0, and 24, respectively (p=0.15). Conclusions: Adjuvanted Env vaccination of HIV-exposed infants can induce cross-clade specific, durable anti-V1V2 IgG that have avidity comparable to that of anti-V1V2 IgG acquired from their chronically-infected mothers. 906 Maternal Neutralization Escape Virus Variants Do Not Predict Infant HIV Infection Risk Caitlin Milligan 1 ; Maxwel Omenda 1 ;Vrasha Chohan 1 ; Katherine Odem-Davis 1 ; Barbra A. Richardson 1 ; RuthW. Nduati 2 ; Julie M. Overbaugh 1 1 Fred Hutchinson Cancer Research Center, University of Washington, Seattle, WA, US; 2 University of Nairobi, Nairobi, Kenya Background: Mother-to-child transmission (MTCT) of HIV provides a setting in which to study the immune correlates of protection from infection. The role of maternal neutralizing antibodies (NAbs) in preventing MTCT, however, is unclear. Variable findings reported in the literature may be due to challenges in sampling viruses and antibodies near transmission, differences in viruses tested, and availability of key clinical correlates including viral load. We sought to understand whether mothers who transmitted the virus to their infants via breastfeeding differed in their ability to neutralize their autologous virus variants at the time of transmission frommothers who did not transmit. Methods: HIV-infected mothers from Kenya were included based on the following criteria: 1)high plasma viral load (>4.6log 10 copies/mL); 2)breastfed ≥ 3 months; 3)infant was HIV-DNA negative at birth; 4)maternal sample available prior to, but near the time of transmission. Ten non-transmitting mothers (NTM) and 10 transmitting mothers (TM) were chosen based on these criteria and full-length HIV-1 envelope genes were cloned from PBMCs from each woman. These envelopes were used to generate pseudoviruses that were tested against maternal autologous/contemporaneous plasma in a TZM-bl neutralization assay. We estimated the association between risk of transmission and the log 2 IC50 for multiple virus variants per mother using logistic regression with clustered standard errors to account for intra-woman correlation. A t-test was used to compare the proportion of neutralization-resistant viruses in TM and NTM. Results: In total, 100 envelope genes were cloned from the 20 women (5 per woman) and displayed a range of neutralization sensitivities when tested as pseudovirus against maternal plasma. TM had a median log 2 IC50 of 8.20 (IQR 7.58, 9.01) and NTM had a median of 7.82 (IQR 6.56, 9.02). In a logistic regression, there was not a significant difference in autologous NAb responses between TM and NTM (OR 1.25, p=0.3). Neutralization-resistant viruses were detected in both groups at similar proportions when tested against autologous plasma at a 1:50 dilution (NTM 6% vs. TM 2%, p=0.3). Conclusions: HIV-infected mothers exhibit a mixture of mostly neutralization-sensitive viruses, however, both TM and NTM also harbor neutralization resistant variants around the time of transmission. These results suggest that MTCT during the breastfeeding period is not simply due to the presence of maternal neutralization escape variants.

Poster Abstracts

544

CROI 2015