CROI 2017 Abstract e-Book

Abstract eBook

Oral Abstracts

122 CTLs PARE DEFECTIVE HIV PROVIRUSES WITHOUT IMPACTING INFECTIOUS LATENT RESERVOIRS Szuhan Huang 1 , Allison S. Thomas 2 , Sara Karandish 1 , Dora Chan 1 , AdamWard 1 , Ya-Chi Ho 3 , Erika Benko 4 , Colin Kovacs 5 , R. Brad Jones 1 1 The George Washington Univ, Washington, DC, USA, 2 Boston Univ, Boston, MA, USA, 3 The Johns Hopkins Univ, Baltimore, MD, USA, 4 Maple Leaf Med Clinic, Toronto, Ontario, Canada, 5 Univ of Toronto, Toronto, Ontario, Canada Background: The majority of “shock-and-kill” studies have been performed in primary-cell models of HIV latency, which are imperfect representations of natural viral reservoirs. A need remains for a rigorous investigation of whether immune effectors in combination with latency-reversing agents (LRAs) can drive reductions in replication-competent proviruses from natural reservoirs. We treated ex vivo CD4+ T-cells from HIV+ individuals on long-term ARV therapy, with LRAs and autologous HIV-specific CTL clones (targeting non-escaped epitopes), and assessed the impact on total and intact-inducible proviruses. Methods: HIV-specific CTL clones targeting known HIV epitopes were isolated from ARV-treated subjects by limiting dilution, and killing activities were confirmed by flow cytometric assays. We developed an HIV eradication assay to test the abilities of these CTLs to reduce viral reservoirs in combination with romidepsin, vorinostat, bryostatin, ALT-803, or Pam3CSK4. Resting CD4+ T-cells from HIV+ leukapheresis samples were co-cultured with LRAs + CTLs for 5 days with ARVs, and activation/memory phenotypes were monitored. CD4+ T-cells were isolated after treatment and total/intact-inducible reservoirs were measured by cell-associated HIV DNA (ddPCR) and by quantitative viral outgrowth assay (QVOA). Results: Combinations of bryostatin and ALT-803+Pam3CSK4 with HIV-specific CTL generally led to significant decreases in cell-associated HIV DNA, with the greatest effects observed for bryostatin (up to 50% reductions, p < 0.01). Critically, these decreases in HIV DNA were not associated with measurable reductions in intact-inducible virus, regardless of the CTL clone or LRA combination used (powered to detect ~50% reductions with 95% confidence). Even when combined with PMA/ionomycin, CTLs were unable to drive reductions in intact-inducible virus. CTLs degranulated (CD107a) in response to autologous activated CD4+ T-cells that had been infected with virus from positive QVOA wells, ruling out a role for immune escape in our observation. Conclusion: Recently, it has been demonstrated that some defective proviruses can be expressed as antigens, enabling CTL recognition. Data from our ex vivo experiments are consistent with the preferential depletion of defective proviruses by CTLs, leading to reductions in HIV DNA without impacting intact proviruses. Understanding and overcoming the mechanisms limiting CTL against the intact-inducible reservoir may be key to successful CTL-based shock and kill interventions.

Oral Abstracts

123 DEVELOPMENT OF A PKC AGONIST DERIVED FROM INGENOL FOR HIV LATENCY DISRUPTION IN VIVO Jessica Brehm 1 , Vincent Tai 1 , David Irlbeck 1 , Robert Ferris 1 , Nancie Archin 2 , Matt Kanke 2 , Jean-Pierre Routy 3 , David M. Margolis 2 , Jun Tang 1 , David Favre 1 1 GlaxoSmithKline, Durham, NC, USA, 2 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 3 McGill Univ Hlth Cntre, Glen Site, Montréal, QC, Canada Background: Protein kinase C agonists (PKCa) are potent in vitro latency reversing agents (LRAs) that act synergistically with histone deacetylase inhibitors (HDACi) and bromodomain inhibitors (iBET), suggesting their potential use in HIV cure strategies. Here we confirm and extend prior testing of the PKCa Ingenol-B (IngB) in vitro. Because IngB is unstable at physiological pH conditions, a stabilized form of Ingenol (GSK445A) was developed and tested in vitro and in the non-human primate model. Methods: The ability of IngB to reverse latency in vitro was tested in reporter cell lines and in total or resting CD4+ T cells from stably-treated HIV-infected donors. Reactivation was measured by cell-associated RNA (caRNA), tat/rev induced limiting dilution assay (TILDA) and quantitative viral outgrowth assays (QVOA). IngB, GSK445A and iBET151 as single agents, fixed-dose combination and dose-response curves were measured. Cellular responses were profiled in vitro for cytotoxicity, cell signaling and transcriptomics. The pharmacology and biomarkers of GSK445A were analyzed in blood and tissues of healthy and chronically SIV-infected rhesus macaques (RM). Results: The effective concentrations 50 (EC50) of GSK445A and IngB were 1 and 13 nM, respectively, in HIV-Jurkat cell lines compared with ~200 nM for IngB in primary CD4+ T cells of HIV donors (caRNA), with latency reversal confirmed by TILDA and QVOA. Cell signaling and transcriptional profiling demonstrated rapid and potent activation of NFkB, Akt and MAPK pathways in CD4+ T cells. Combination activity of IngB with iBET151 showed a 5 to 10-fold increase in potency and maximal response by caRNA and was confirmed by TILDA, suggesting synergy in the reactivation of a large pool of latently infected cells. In vivo, a brief infusion of GSK445A at 5 to 20ug/kg was well tolerated in RM (n=20). Pharmacokinetics of GSK445A displayed a biphasic decline with exposure above the EC50 for 3-6 hours. Rapid increase in plasma IL-6, T cell trafficking and up-regulation of CD69 (10-60% in T cells) revealed dose-dependent responses in blood and lymphoid tissues, which returned to baseline after 6 to 48 hours. Conclusion: In vitro, GSK445A and IngB reversed HIV latency; activity was augmented by iBET151. GSK445A was developed as a stable Ingenol derivative that was well tolerated at an effective dose in RMmaking GSK445A a candidate for single and combination studies to test latency reversal and clearance in vivo. 124 HIV RNA REBOUND POSTINTERRUPTION IN PERSONS SUPPRESSED IN FIEBIG I ACUTE HIV Donn Colby 1 , Nicolas Chomont 2 , Eugene Kroon 1 , Louise Leyre 2 , Suteeraporn Pinyakorn 3 , Nelson L. Michael 3 , Merlin L. Robb 3 , Nittaya Phanuphak 1 , Jintanat Ananworanich 3 , for the RV411 and RV254/SEARCH010 Study Groups 1 SEARCH, The Thai Red Cross AIDS Rsr Cntr, Bangkok, Thailand, 2 Cntr de Rsr du Cntr Hosp de l’Univ de Montréal, Montréal, QC, Canada, 3 US Military HIV Rsr Prog, Bethesda, MD, USA Background: Initiation of antiretroviral therapy (ART) during acute HIV infection (AHI) minimizes establishment of a latent HIV reservoir. This could, in turn, delay viral rebound following treatment interruption (TI) and potentially induce a post-treatment controller state. We investigated time to viral load (VL) rebound after ART cessation in participants who initiated ART during AHI. Methods: Eight participants (7 male, 1 female) who initiated ART in Fiebig I (VL+, p24-, IgM-) acute infection and had VL<20 copies/ml on ART for a median of 2.8 (range 2.5-5.5) years before undergoing TI with VL monitoring every 3-7 days. VL, CD4/CD8, HIV DNA and inducible HIV RNA were examined. There was 85% power to reject the null hypothesis of 5% rate of VL < 50 copies/ml at 24 weeks post-TI if the true rate was 30% or greater. Results: The median (range) age was 29 (22-34) years. HIV subtypes were CRF01_AE (n=6) or CRF01_AE/B (n=2). Median (range) pre-ART values included HIV RNA 4.2 (3.3-4.9) log10copies/ml, HIV DNA 66 (0-490) log10copies/106CD4, and CD4 413 (227-565) cells/mm3. Prior to TI, median (range) CD4 was 561 (425-654) cells/mm3. Total HIV DNA (LOD 5 copies/106CD4) and inducible HIV RNA (LOD 1.4 tat/rev RNA+cells/106CD4) were undetectable for all participants. All participants experienced VL rebound post-TI at a median

49

CROI 2017