CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

329 PGT121 ANTIBODY ENGINEERING: ENHANCED INFECTED CELL KILLING AND DRUG-LIKE PROPERTIES Mini Balakrishnan, Nathan D. Thomsen , Craig S. Pace, Xue Zhang, Magdeleine Hung, Mark R. Nagel, Brian A. Carr, Y. E. Hu, Helen Yu, John A. Corbin Gilead Scis, Inc, Foster City, CA, USA Background: A therapeutic agent that mediates the selective destruction of cells harboring latent HIV may sustainably suppress viremia in the absence of antiretroviral therapy and offers the potential for a sterilizing cure. Anti-HIV-1 envelope (Env) antibodies from elite neutralizers can neutralize a high percentage of viral strains across clades. Some of these broadly neutralizing antibodies (bNAbs) can recognize Env expressed on the surface of infected cells and mediate killing via effector function. Methods: A set of patient-derived bNAbs targeting different Env epitopes were tested for natural killer (NK) cell mediated killing of HIV-infected CD4+ T cells to identify a bNAb for further optimization. Enhancement of effector function was achieved through glycoengineering of the fragment crystallizable (Fc). Fragment antigen-binding (Fab) variants were generated via genetic engineering and screened using a panel of biochemical and functional assays for improvements in drug-like properties. Results: PGT121 was selected as the lead bNAb for engineering on the basis of killing activity and previously demonstrated efficacy in SHIV infected monkeys (Barouch et al. Nature 2013). Fc glycoengineering increased binding affinity to Fc gamma receptor 3A ~10-fold and similarly enhanced NK cell-mediated antibody dependent cellular cytotoxicity of HIV-infected CD4+ T cells (mean EC50 = 0.82 μg/mL, mean Emax = 40%) compared to PGT121 (mean EC50 = 3.8 μg/mL, mean Emax = 10%). Genetic engineering eliminated several liabilities within the variable region of PGT121 Fab. Removal of T-cell epitopes reduced ex vivo T-cell activation rate from 32% to 12% of donors, indicating a reduction in the potential for clinical immunogenicity. Removal of three glycosylation motifs improved manufacturability by eliminating the need to monitor and control Fab glycosylation. Fab engineering did not impair Env recognition as assessed by neutralization activity (median IC50 = 0.0106 μg/ml, 67% breadth) compared to PGT121 (IC50 = 0.0163 μg/ml, 66% breadth) against a panel of 142 clade B clinical isolates. Conclusion: Engineering of PGT121 enhanced its ability to kill HIV-infected cells and yielded a Fab variant with superior drug-like properties, while maintaining the breath of HIV Env recognition. Further validation is required to determine if this approach will be generally applicable to optimizing bNAbs for the selective elimination of cells harboring latent HIV reservoir. 330LB VRC01 INFUSION HAS NO EFFECT ON HIV-1 PERSISTANCE IN ART-SUPPRESSED CHRONIC INFECTION Sharon Riddler 1 , Christine Durand 2 , Lu Zheng 3 , Justin Ritz 3 , Richard A. Koup 4 , Julie Ledgerwood 4 , Bernard Macatangay 1 , Joshua C. Cyktor 1 , John W. Mellors 1 , for the ACTG A5342 ProtocolTeam 1 Univ of Pittsburgh, Pittsburgh, PA, USA, 2 Johns Hopkins Univ, Baltimore, MD, USA, 3 Harvard Univ, Boston, MA, USA, 4 NIAID, Bethesda, MD, USA Background: ART blocks infection of new cells but has no impact on cells already infected with latent or active proviruses. Broadly neutralizing monoclonal antibodies (bnMAb) may promote clearance of viremia and virus-expressing cells through antibody-dependent mechanisms. We evaluated whether the CD4-binding site bnMAb VRC01 affects HIV persistence in chronically-infected individuals on ART. Methods: A5342 was a phase 1, randomized, double-blind, placebo-controlled, parallel arm study. Participants with ART-suppressed viremia (<40 copies/ml) were randomized to Arm A: 2 infusions of VRC01 (40 mg/kg) at entry and week 3 and 2 infusions of placebo (saline) at weeks 6 and 9; or Arm B: 2 infusions of placebo at entry and week 3 and 2 infusions of VRC01 at weeks 6 and 9. Primary outcomes were safety and change in cell-associated HIV RNA/DNA ratio (CAR/CAD) from baseline (BL) to week 6. Plasma viremia (single copy HIV RNA assay [SCA]) and PMA/ionomycin-stimulated virus production (HIV RNA copies/ml) from CD4+T-cells were also assessed. Changes from pre- to post-VRC01 (Arm A: entry to week 6; Arm B: week 6 to week 12) time points across both arms were evaluated. Results: 40 participants were randomized; 20 per arm. Median age was 52 y; median CD4+ T-cell count was 696 /mm3. No treatment-related adverse events ≥grade 3 were reported during study follow up. No significant difference between VRC01 and placebo was observed for change in CAR/CAD ratio from BL to week 6 (median fold change: 1.12 vs. 0.83, p=0.16, 95% CI (0.75, 2.42)), or from pre-to post-VRC01 time points with both arms combined (1.24, 95% CI (0.83, 1.69), p=0.29; Table). At entry, 22/40 (55%) participants had SCA ≥1 copy/ml. At week 6, there was no difference in the proportion with SCA ≥1 copy/ml between the arms (42% vs. 37%, p=1.0). There were also no significant differences between arms from BL to week 6, or from pre- to post-VRC01 time points with both arms combined in PMA/ionomycin stimulated virus production (all p>0.05). Conclusion: In individuals with chronic ART-suppressed HIV infection, VRC01 infusions were safe and well tolerated but did not affect plasma viremia, cellular HIV RNA/DNA levels, or stimulated virus production from CD4+T-cells. Potential mechanisms being evaluated to explain the lack of response include viral resistance to VRC01, poor penetration of VRC01 to sites of virus expression, or inherent inability of VRC01 to clear virus particles or virus-expressing cells.

Poster and Themed Discussion Abstracts

331 ENHANCEMENT OF THE CD4 CHIMERIC ANTIGEN RECEPTOR AGAINST HIV INFECTION Anjie Zhen, Mayra A. Carrillo , Valerie Rezek, Scott G. Kitchen Univ of California Los Angeles, Los Angeles, CA, USA

Background: Cell-based gene therapy has quickly become the new frontier for treatment of HIV infection. Clinical trials have shown CD4 chimeric antigen receptor (CD4CAR) modified T cells to be moderately effective against HIV and safe when introduced through adoptive T cell transfer. This CAR expresses CD4 attached to an intracellular CD3 zeta- domain. When bound to the HIV envelope glycoprotein 120 (gp120), CD4 CAR expressing CD8 cells elicit an immune response against infected cells. Despite its success in vitro, T cells modified with the CD4 CAR showed limited efficacy against HIV infection in clinical trials. Studies suggest that CD4 CAR modified CD8 cells may have lost their efficiency due to their susceptibility to HIV infection via CD4 expression, excessive ex vivo activation and cell processing, and immune exhaustion/premorbidity. Thus, we aim to overcome

CROI 2017 130