CROI 2017 Abstract e-Book

Abstract eBook

Oral Abstracts

Results: 295 HCC cases diagnosed in HIV/HCV-coinfected patients have been included in the GEHEP-002 cohort. The median (Q1-Q3) age was 49 (46-52) years and 265 (90%) were male. HCV genotype distribution was: Gt 1, 114 (48%); Gt 2: 3 (1%); Gt 3, 85 (36%) and Gt 4, 35 (15%). Since 1999, when the first HCC case was recorded, 34 (11.5%) cases have occurred in patients with previous SVR. The proportion of HCC cases in patients with previous SVR was 16.7% (1 out of 6), 10.9% (15 out of 137) and 7.9% (10 out of 126) in period 1, 2 and 3, respectively (Figure 1). By contrast, this proportion increased to 30.8% (8 out of 26) in the DAA IFN-free period (p<0.01 for the comparison between period 1-3 vs 4) (Figure 1). Twenty-one patients with previously treated HCC received subsequent therapy with DAA IFN-free regimens. HCC recurred in 1 (4.7%) of them. Conclusion: The proportion of HCC cases diagnosed in HIV/HCV-coinfected patients with previous SVR has significantly increased parallel to the arrival of DAA IFN-free strategies. This finding may be, at least partially, explained by the fact that DAA have allowed treating patients at advanced stages of liver disease in which the protective effect of SVR on the risk of HCC could be less marked.

Oral Abstracts

140 IDENTIFICATION OF HCV INTER-SUBTYPE RECOMBINANT STRAINS WITH BREAKPOINTS IN NS5A Suqin Cai , Alicia Newton, Wei Huang, Jeannette Whitcomb Monogram BioScis, South San Francisco, CA, USA Background: Hepatitis C virus (HCV) is classified into seven major genotypes (GT) and numerous subtypes. HCV GT and subtype are associated with differences in clinical outcomes, response to treatment and epidemiology. A better understanding of recombination in HCV is of practical importance and evolutionary interest. Methods: Recombinant HCV strains were identified during routine drug resistance testing of clinical specimens. HCV RNA was isolated from plasma and NS5A amplicons were generated by RT-PCR and sequenced on a next-generation sequencing platform. Sequencing reads were analyzed using a proprietary analysis pipeline with a subtyping routine that aligns a subset of reads to a library of reference HCV strains. Samples with a large proportion of reads that aligned to two different references were evaluated as potential recombinant viruses. Consensus sequences were compared to subtype reference sequences (Bootscan, Simplot) to define recombination breakpoints. Phylogenetic analysis was performed (Clustal) to determine relatedness of independently collected HCV specimens. Results: Ten 1a/1b recombinant viruses were identified from diverse geographic collection centers. All 10 had 5’ NS5A 1a sequences and 3’ NS5A 1b sequences. The recombination breakpoint for 7 of these viruses was consistently at amino acid (aa) 386 and for 2 of the viruses at aa 366 of NS5A relative to the H77 reference. All 9 of these 1a/1b recombinants had the same recombination locus in 1b at aa 348 relative to the Con1 reference. A single 1a/1b recombinant had a 1a breakpoint at aa 277 (H77) and a 1b breakpoint at aa 214 (Con1). The recombinant NS5As were longer than H77 and Con1 NS5A due to duplication of sequences at the site of recombination. Phylogenetic analysis demonstrated that the 9 1a/1b recombinant viruses with breakpoints at 366/386 in 1a and 348 in 1b were all closely related suggesting that this may represent a circulating recombinant form of HCV. We also identified one inter-genotype recombinant (1a/4), with a breakpoint in NS5A at aa 349 of H77 and at aa 339 of the GT4 ED43 reference. Conclusion: HCV recombination is thought to be rare. However, the prevalence of recombinant forms may have been underestimated due to the use of genotyping methods, such as single-locus sequencing, that are unlikely to detect recombination. The independent identification of a similar recombinant strain frommultiple samples and diverse geographic collection areas suggests that the strain exists as a circulating recombinant form. 141 SYNERGY BETWEEN COMBINATIONS OF ANTI-HCV BROADLY NEUTRALIZING MONOCLONAL ANTIBODIES Madeleine C. Mankowski 1 , Andrew I. Flyak 2 , Stuart C. Ray 3 , James E. Crowe 2 , Justin R. Bailey 3 1 Johns Hopkins Univ, Severna Park, MD, USA, 2 Vanderbilt Univ, Nashville, TN, USA, 3 Johns Hopkins Univ, Baltimore, MD, USA Background: No single broadly-neutralizing monoclonal antibody (bNAb) effectively neutralizes all strains of hepatitis C virus (HCV). Therefore, induction of antibody responses that target multiple distinct epitopes may be needed for an effective vaccine. This vaccination strategy would be most effective if bNAbs interact synergistically, and less effective if bNAbs interact antagonistically. We hypothesized that bNAbs binding to distinct epitopes would have increased neutralizing breadth and synergistic potency when used in combination. We tested 35 bNAb combinations to define effects on neutralizing breadth and potency. Methods: Twelve bNAbs binding to distinct epitopes on HCV E2 were screened individually and in 35 2-bNAb combinations for neutralization of a panel of 11 genetically diverse HCV pseudoparticles (HCVpp). Neutralization of HCVpp by these combinations was compared to expected neutralization predicted by the Loewe additivity model, defining the interactions of the component bNAbs as either synergistic, additive, or antagonistic. Results: BNAb combinations showed significantly greater neutralizing breadth than individual bNAbs at the same total antibody concentration, neutralizing a median of 63% of HCVpp, while individual bNAbs neutralized a median of 55% (p<0.05, unpaired T-test). None of the 35 bNAb combinations tested showed evidence of antagonism. Most showed additive potency, and one combination, HepC74 + HepC98, showed synergy (neutralization in excess of additive effects) at 8 of 9 antibody concentrations tested (p<0.05, paired T-test). Neutralization of HCV by this combination was 3-fold greater than neutralization by either bNAb individually at the same total antibody concentration. Conclusion: By screening 35 bNAb combinations, we identified a combination that displays both increased neutralizing breadth and synergistic potency, many bNAb combinations with additive neutralization, and no combinations displaying antagonism. These results demonstrate that induction of multiple bNAbs targeting distinct epitopes could be an effective vaccine strategy for HCV. Specific induction of HepC74- and HepC98-type antibodies could be particularly effective. 142 NEUTRALIZING ANTIBODY DEVELOPMENT DURING HIV-1 INFECTION Penelope Moore, Univ of the Witwatersrand, Johannesburg, South Africa A preventative HIV vaccine is likely to require broadly neutralizing antibodies (bNAbs) able to recognize diverse circulating viruses from across the world. Such antibodies have not yet been elicited by vaccination. However some HIV-1-infected individuals naturally mount bNAb responses during chronic infection, suggesting a need for prolonged maturation towards breadth. Longitudinal studies of how bNAbs develop may provide a template for immunogen design, by defining the interplay between antibodies and HIV envelope evolution. Understanding how certain members of an antibody lineage mature towards breadth, whereas others become evolutionary dead-ends, provides insights for vaccine strategies. Similarly, studies of donors who fail to develop bNAbs despite equivalent levels of antigenic stimulation will highlight potential roadblocks. Overall, this presentation will highlight the dynamic relationship between virus and antibody in shaping breadth, with implications for HIV immunogen design.

58

CROI 2017