CROI 2019 Abstract eBook

Abstract eBook

Oral Abstracts

plasma HIV RNA with median CD4 count 631 cells/mm³ (IQR 458-840) at study entry. At study entry, 16% had NCI, 29%were obese, and 40%were overweight. Over 3 years, 6% of participants developed NCI while 78% remained unimpaired. In multivariable models, increasing age (OR 1.04; 95% CI 1.00, 1.08; p=0.04), and having an obese (OR 2.45; 1.05, 5.70; p=0.04) or overweight BMI (OR 2.21; 1.00, 4.91; p=0.05) vs normal BMI were associated with increasing prevalence of NCI compared to those who remained unimpaired. Conclusion: Both greater age and obesity were independently associated with worsening cognitive function. These results extend previous work demonstrating a higher risk of NCI among obese PLWH by showing that obese individuals are also at greater risk of subsequently transitioning from unimpaired to impaired neurocognition. 130 HIGH HIV VIRAL BURDEN PERSISTS IN CXCR3+ GCTFH DESPITE VERY EARLY ART INITIATION Omolara Baiyegunhi 1 , Jaclyn Mann 1 , Adrienne Yanez 2 , Geetha H. Mylvaganam 2 , Krista Dong 2 , Thumbi Ndungú 1 , Bruce D. Walker 2 , Zaza Ndhlovu 2 1 University of KwaZulu-Natal, Durban, South Africa, 2 Massachusetts General Hospital, Boston, MA, USA Background: Early initiation of Combination Antiretroviral Therapy (cART) during acute HIV infection blunts peak viremia, reduces HIV viral reservoirs and preserves immune function, but treatment interruptions often results in rapid viral rebound. We studied persons identified and treated at the onset of plasma viremia, in many when viral load is less than 1000 RNA copies/ml to define the dynamics of HIV suppression in lymph node (LN) tissues. Additionally, we investigated the cell subset that remained persistently HIV infected. Methods: We studied 16 hyperacute HIV-infected subjects who initiated therapy in Fiebig stage I, subdivided into three groups based on when the LN sample was obtained. Group 1 was sampled within 3 months, group 2 was sampled at one year and group 3 was sampled after two years on therapy. Immunofluorescence (IF) microscopy and RNAscope in situ hybridization (ISH) techniques were used to quantify Gag p24 protein and Gagpol RNA respectively. The Cobas AmpliPrep HIV-1 test was used to quantify LN and plasma viral loads and viral Gag, Nef and Envelope sequencing were conducted using ABI 3130xl sequencing platform. Digital droplet PCR was used to quantify HIV RNA levels in FACs sorted LN cell subsets and follicular CD4+ T cells harboring HIV antigens were extensively phenotyped by flow cytometry. Results: Despite rapid plasma viral suppression at a median of 16.5 days, Gag p24 antigen and quantifiable RNA were readily detectable in the LN in 12 out of the 16 donors sampled in all three experimental groups. Moreover, sequencing analysis revealed viral evolution in Gag, Nef and/or Envelope sequences in 4 out of 6 LNs sampled >3 months after therapy compared to the transmitter found virus sequences obtained just before cART initiation. There was no significant reduction in Gagp24 antigen in LN samples obtained after a year on cART compared to the samples obtained within 3 months on cART (p=0.4). RNA quantification of FACs sorted TFH subsets showed significantly higher levels of Gag p24 mRNA copies in CXCR3+ follicular CD4+ T cells compared to other TFH subsets (p=0.01). Conclusion: Our results highlight the huge difference in viral load decay kinetics between peripheral blood and LN, despite very early cART initiation. Importantly, we identify that CXCR3+ TFH contribute significantly to viral persistence in the LN during therapy. These results underscore the need for future interventions directed at eliminating residual virus in tissue sanctuaries. 131 SIROLIMUS REDUCES T-CELL CYCLING AND IMMUNE CHECKPOINT MARKER EXPRESSION, ACTG A5337 Timothy J. Henrich 1 , Ronald Bosch 2 , Catherine Godfrey 3 , Hanna Mar 2 , Apsara Nair 4 , Elizabeth Hawkins 5 , Carl Fichtenbaum 6 , Michael Keefer 7 , Jonathan Z. Li 8 , Daniel R. Kuritzkes 8 , Michael M. Lederman 9 , Priscilla Hsue 1 , Steven G. Deeks 1 1 University of California San Francisco, San Francisco, CA, USA, 2 Harvard University, Boston, MA, USA, 3 National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA, 4 Frontier Science & Technology Research Foundation, Inc, Amherst, NY, USA, 5 Social & Scientific Systems, Silver Spring, MD, USA, 6 University of Cincinnati, Cincinnati, OH, USA, 7 University of Rochester, Rochester, NY, USA, 8 Brigham and Women’s Hospital, Boston, MA, USA, 9 Case Western Reserve University, Cleveland, OH, USA Background: Reversing immune dysfunction and inhibiting T-cell proliferation are critical to immune-based HIV cure strategies. A prior retrospective analysis of the use of sirolimus, an mTOR inhibitor, in HIV+ renal transplant recipients

suggests that it may lead to lower CD4+ T cell HIV DNA, but prospective studies are lacking. Therefore, we sought to evaluate the safety of sirolimus in ART- suppressed HIV-infected individuals and its effect on immune function and HIV-1 reservoir size. Methods: A5337 was an open-label, single-arm, pilot study of 20 weeks of oral sirolimus treatment for HIV-infected individuals on ART with HIV RNA <40 cps/ mL. Eligibility criteria included at least 24 months on ART, HIV RNA < assay limit and CD4+ cell count ≥ 350 cells/mm3. Measures of T-cell activation and cycling, immune exhuastion and CCR5 expression (secondary efficacy outcomes) were compared by paired t-tests prior to vs after continuous oral sirolimus. Results: 32 participants enrolled in the study. Participants had a median age of 52 years, 28%were female and 56%were black non-Hispanic. The median baseline CD4+ cell count was 813 cells/mm3. Two participants did not start study drug, 14 completed less than 20 weeks of sirolimus, and 16 completed 20 weeks of therapy. Twenty weeks after initiating sirolimus, CD4+ cell counts declined by a mean of 118 cells/mm3 (p=0.04; n=16). Three participants had a grade 3 adverse event (stomatitis and perturbations of fasting glucose in a known diabetic) or a decrease in CD4+ cell count to <300 cells/mm3. Two participants stopped sirolimus because of assympomatic transient Epstein Barr viremia. Other individuals discontinued because of lower grade toxicities or minor, clinically insignificant laboratory abnormalities. Twenty weeks of sirolimus was associated with significant decreases in the percentages of cycling Ki67+ CD4+ and CD8+ T cells (mean change -0.5%, p=0.031, and -0.5%, p=0.005, respectively), PD-1+ CD8+ T cells (-2.9%, p=0.008), and CCR5+ CD8+ T cells (-3.9%, p=0.001). Conclusion: Sirolimus use was associated with significant reductions in CCR5 expression and T-cell markers of cell cycling and immune exhaustion. There was a relatively high rate of treatment discontinuation, in part because of protocol- defined stopping criteria. 132 PD-1 AND CTLA-4 BLOCKADE IN MACAQUES INDUCES T-CELL EXPANSION AND SIV REACTIVATION Justin L. Harper 1 , Shari Gordon 2 , Cristin Galardi 2 , Hong Wang 1 , Colleen S. McGary 1 , Colin T. King 1 , James Schawalder 2 , Andrew Z. Sanderson 3 , Brian A. Johns 2 , Jeffrey D. Lifson 4 , Jake D. Estes 5 , David M. Margolis 6 , Guido Silvestri 1 , David Favre 2 , Mirko Paiardini 1 1 Yerkes National Primate Research Center, Atlanta, GA, USA, 2 ViiV Healthcare, Research Triangle Park, NC, USA, 3 GlaxoSmithKline, Uxbridge, UK, 4 Frederick National Laboratory for Cancer Research, Frederick, MD, USA, 5 Oregon Health and Sciences University, Portland, OR, USA, 6 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: The HIV reservoir is largely composed of resting memory CD4+ T-cells, a large fraction of which express the co-inhibitory receptors PD-1 and CTLA-4 that limit T-cell activation. Furthermore, PD-1 expression contributes to CD8+ T-cell dysfunction during HIV infection. We hypothesized that a dual PD-1 and CTLA-4 immune checkpoint blockade (ICB) will facilitate T-cell activation, revert latency, and restore SIV-specific T-cell responses, thus enhancing clearance of the viral reservoir. Methods: 33 rhesus macaques (RMs) were i.v. infected with SIVmac239 and initiated ART (TDF/FTC/DTG) at day 60, which was maintained for 1 year. RMs received 4, weekly ICB infusions: control antibody (n=6); anti-PD-1 mAb (n=6); anti-CTLA-4 mAb (n=6); anti-CTLA-4 plus anti-PD-1 mAb (n=7); and bi-specific anti-CTLA-4/PD-1 (n=8). All RMs underwent analytic ART interruption (ATI). Peripheral blood (PB), lymph node (LN), and rectal biopsy (RB) were collected longitudinally. Results: ICB treatment was well tolerated and demonstrated PD-1 receptor occupancy in PB and LN, including in T follicular helper cells. Dual PD-1/CTLA-4 ICB induced a significant increase in the number of cycling (Ki-67+) memory CD4+ and CD8+ T-cells in LN (1.96- and 1.96-fold, respectively), with significant expansion in the number of effector memory T-cells in PB (4.61- and 2.20-fold, respectively; p<0.01 for all these measures). Enhanced T-cell cycling was dependent on CTLA-4 inhibition, highlighting the synergy of the dual blockade. In addition, dual ICB increased the frequency of activated (CD38+HLA-DR+) and T-bet+ cells in LN. Notably, increased CD4+ T-cell cycling induced by dual ICB resulted in higher levels of plasma SIV RNA (indicative of viral reactivation) relative to monotherapy. Both the frequency (viremia >100 copies/mL measured in 4 out of 7 RMs and at multiple time points) and magnitude (average of 913 SIV-RNA copies/mL) of plasma SIV RNA was increased in dual vs monotherapy. Following ATI, only anti-PD-1 monotherapy produced a

Oral Abstracts

49

CROI 2019