CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

67% reduction) in the frequency of these activation markers on all 3 lineages after ART that was associated with the degree of decline in viremia. There was no significant decline in the percentage of HLA-DR+CD38- CD8+ T cells in the first two weeks of ART. Conclusion: While a prior study showed no effect of ART on the frequency of HLA-DR+CD38+ CD8+ T cells in HCs after 4 weeks of treatment, we show here that there is a substantial decline in these cells in HCs with low CD4+ T cells as early as 2 weeks after the initiation of ART. This is not associated with a decline in the percentage of HLA-DR+CD38- CD8+ T cells that are thought to contribute to the control of viral replication. Given the association of T cell activation with HIV associated morbidity, this study offers an immunologic rationale for initiating ART in HCs with low CD4+ T cell counts. 264 ELITE CONTROL OF HIV-1 INFECTION IS ASSOCIATED WITH REDUCED TRAILSHORT EXPRESSION Ana Catarina De Oliveira Virgens Paim 1 , Sekar Natesampillai 1 , Nathan Cummins 1 , Enrique Garcia-Rivera 1 , Nicole Kogan 1 , Ujjwal Neogi 2 , Gary D. Bren 1 , Background: Decline of CD4 T-cells in untreated HIV-1 infection is mainly due to apoptosis. TNF-related apoptosis inducing ligand (TRAIL) contributes to this CD4 T-cell decline but does not kill all infected cells. A novel protein, TRAILshort, which is expressed by HIV infected and uninfected cells, prevents the pro- apoptotic TRAIL from killing TRAIL receptor expressing cells and may promote HIV persistence. We hypothesized that HIV-1 elite controllers express less TRAILshort compared to viremic persons, leading to increased killing of HIV-1 infected cells, higher CD4 counts and lower HIV-1 reservoir size. Methods: Two independent cohorts were studied. Elite controllers (ECs) had undetectable HIV-1 RNA viral load for >1 year in the absence of ART (N=40 and 19 in discovery and validation cohorts). Viremic persons (VPs) had HIV-1 RNA viral loads >10,000 copies/ml off therapy (N=42 and 17). Expression of TRAILshort and full length TRAIL in PBMCs was assessed by RNAseq and flow cytometry. Plasma concentration of TRAILshort was assessed by antibody bead array and full length TRAIL by ELISA. Reservoir size was estimated by ddPCR for total HIV-1 DNA in PBMCs. Results: ECs were significantly older (51 yrs vs. 41 yrs, P<0.001) and had higher baseline CD4 T cell counts (991 cells/mm3 vs. 479 cells/mm3, P<0.001) compared to VPs. ECs had significantly lower total HIV-1 DNA content in PBMCs than VPs (82 copies/106 cells vs. 1572 copies/106 cells, P<0.001). In the discovery cohort, ECs had lower TRAILshort (P=0.002) and full length TRAIL (P=0.001) gene expression in PBMCs compared to VPs. TRAILshort surface expression on CD4 and CD8 T cells and monocytes was lower in ECs relative to VPs but not statistically significant. In the validation cohort, TRAILshort (P=0.06) and full length TRAIL (P=0.004) gene expression was lower in PBMCs of ECs vs. VPs. ECs had statistically significant lower plasma TRAILshort concentration (normalized to CD4 count) than patients with chronic HIV infection (P<0.001), primary HIV infection (P=0.002) and patients on long term ART (P=0.002). Conclusion: ECs have lower TRAILshort expression, higher CD4 T cell counts and lower HIV-1 reservoir size than VPs. Reduced TRAILshort expression may facilitate TRAIL-mediated killing of HIV-1 infected cells by the innate and adaptive immune system in ECs. TRAILshort may be an attractive novel target for immunomodulatory therapy to enhance immunologic control of HIV-1 infection. 265 HIV-DNA CONTENT IN PTFH CELLS IS ASSOCIATED WITH RESIDUAL VIREMIA IN ELITE CONTROLLERS Marcial García 1 , Vincent Morcilla 2 , María Ángeles Navarrete 1 , Katie Fisher 2 , Alfonso Cabello 3 , Juan Carlos López-Bernaldo 4 , Francisco Javier De La Hera 3 , Carlos Barros 5 , Manuel Fernández 3 , Vicente Estrada 6 , Miguel Górgolas 3 , Sarah Palmer 2 , José Miguel Benito 1 , Norma Rallón 1 1 Instituto de Investigación Sanitaria Fundación Jiménez Díaz, Madrid, Spain, 2 The Westmead Institute for Medical Research, Westmead, NSW, Australia, 3 Hospital Universitario Fundación Jiménez Díaz, Madrid, Spain, 4 Hospital General Universitario Gregorio Marañón, Madrid, Spain, 5 Hospital Universitario de Móstoles, Madrid, Spain, 6 Hospital Universitario Clínico San Carlos, Madrid, Spain Background: Low levels of HIV plasma viremia below the limits of detection of commercial assays (residual viremia) has been demonstrated in patients Stacey Rizza 1 , Steven G. Deeks 3 , Eric C. Polley 1 , Andrew Badley 1 1 Mayo Clinic, Rochester, MN, USA, 2 Karolinska Institute, Stockholm, Sweden, 3 University of California San Francisco, San Francisco, CA, USA

with cART-induced control as well as in those with spontaneous control of HIV replication. The source of residual viremia is highly debated and its potential relationship with levels of cell-associated HIV DNA has not been clarified to date. Herein, we have analyzed the HIV-DNA content in different CD4-T cell subsets and its potential association with residual viremia in elite controllers and in patients with cART-mediated suppression of HIV replication. Methods: Chronically HIV-infected patients maintaining undetectable pVL were included: 7 with spontaneous viral control (EC) and 9 with cART-mediated HIV replication control (cART-treated). Cell-associated HIV-DNA content was measured by ddPCR in purified resting T memory (rTm) and peripheral T follicular helper (pTfh) cell subsets as important compartments of HIV reservoir. Residual HIV viremia was quantified using a PCR single-copy assay (SCA) with a sensitivity of 0.3 copies/ml. Differences between groups were tested by non- parametric tests and associations by Spearman´s rho coefficient. Results: Lower levels of cell-associated HIV-DNA (median[IQR] Log copies/ million cells) was found in EC compared to cART patients in rTm (2.5[1.9–2.9] vs. 3.1[2.8–3.2]; p=0.059) and in pTfh (1.9[1.9–2.5] vs. 2.9[2.6–3.0]; p=0.025). In 3 of 7 EC (43%) and in 5 of 9 cART patients (56%) HIV could not be detected (<0.3 copies/ml). No significant differences (p=0.468) were found in median values of HIV-RNA between EC (9.5[1.5-16.8]) and cART patients (3[1.3-10.8]) in the subgroup of patients having detectable residual viremia (>0.3 copies/ mL). Interestingly, we found a significant and positive correlation between the HIV-DNA levels in pTfh cells and the residual HIV viremia (rho coefficient=0.928, p=0.008) in EC, and this was not observed in cART patients. Conclusion: Our results suggest that pTfh cells could be an important source of residual plasma viremia in EC patients. This could be the consequence of higher transcriptional activity of HIV in pTfh cells of EC compared to that in cART patients, what could explain the similar levels of residual viremia in both groups in spite of the lower HIV-DNA content in pTfh cells of EC compared to cART patients. Further studies are warranted to check if administration of cART to EC patients could help to reduce HIV-DNA in pTfh cell compartment and/or residual plasma viremia. 266 CXCR3+ FOLLICULAR HELPER TC ARE ASSOCIATED WITH HIV CONTROL DURING CHRONIC INFECTION Gonzalo Salgado-Montes de Oca , Perla Del Río-Estrada, Yuria Ablanedo Background: Follicular Helper CD4+ (TFh) are antigen experienced T cells found in secondary lymphoid organs such as lymph nodes (LN). Recently, different TFh subsets has been described during chronic SIV infection based on chemokine receptors expression including CXCR3-Tfh1, CCR4-Tfh2 and CCR6-Tfh17. We characterized TFh subsets proportions and activation patterns during chronic HIV infection and its association with disease progression and viral control. Methods: Cervical LN mononuclear cells (LNMC) from 22 chronic-untreated (CHR) and 7 treated-undetectable (ART) patients were characterized by flow cytometry including CXCR5hiPD-1hi (TFh), chemokine receptors (R5, R4, R6, X3, X5) and activation markers (CD38, HLA-DR, CD69). LNMC and PBMC HIV CA-DNA were measured by qPCR. T cell counts were assessed by BD True count kit. pVL was determined by m2000 system. Analysis were performed on Cytobank and Prism6 using non-parametric tests. Results: CHR participants had an average CD4+Tc count of 454 cells/ul. CD4+Tc represented 21.43% of LNMC. TFh were 3.11% of total LN CD4+Tc. Significant negative correlation was detected between TFh% and CD8+Tc count (Spearman r=-0.56, P=0.006). We found low beta chemokine receptors expression on TFh (Tfh2=2.6%, Tfh17=0.7%). CCR5 expression on TFh was 5.2% on CHR and 11.6% on ART, p=0.005. Of note, CXCR3+ Tfh1 are a prevalent population on both CHR (45.8%) and TAR (44.1%). Interestingly, Tfh1% from CHR negatively correlates with pVL (r=-0.55, P=0.007), PB CA-DNA (r=-0.58, P=0.004) and LN CA-DNA (r=-0.47, P=0.02). Tfh1 clustered as separated population on viSNE analysis. Furthermore, Tfh1 expressed significantly higher levels of CCR5, HLA-DR, CD38, CD69 when compared to CXCR3 negative TFh (p<0.0001 in all markers). Finally, CCR5 expressing Tfh1 were significantly higher on treated participants (9.5% CHR vs 26% ART; p<0.0001). Conclusion: Our study defined preferential chemokine receptors expression patterns on TFh during HIV infection including higher proportions of CXCR3+ Tfh1 cells. Tfh1 levels were associated with lower pVL and CA-DNA on LN and PB. Likewise, Tfh1 were found to be highly activated. These results suggest that Tfh1 Terrazas, Amaranta Y. Rivero-Arrieta, Gustavo Reyes-Terán National Institute of Respiratory Diseases, Mexico City, Mexico

Poster Abstracts

95

CROI 2019