CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Methods: 40 PHIV (4-19yrs age) who started ART <2 years of life and had undetectable viremia (<50 HIV copies/ml) for the past 5 years, were enrolled in 7 European research centers. HIV DNA copies per million peripheral blood mononuclear cells (PBMC) were measured by real-time PCR. Flow cytometry was used to investigate CD4 T cells for 1) co-expression of PD1 with IA (ICOS, CD38, Ki67 and HLA-DR); 2) co-expression of PD1 with ICP (TIGIT, LAG3, TIM3 and CTLA4); 3) intracellular cytokine production (IL2, IFNg, TNFα, IL21) after stimulation with ENV peptides. Pearson correlations and 2 group comparisons were performed using the Mann-Whitney T Test. P value<0.05 was considered significant. Results: Total PD1+ CD4 T cells positively correlated with HIV-DNA (r=0.46) as did CD4 T cells co-expressing PD1 with other ICP or IA (table 1). We then divided our cohort based on HIV DNA distribution into those with high (4th quartile) and low (1st quartile) HIV DNA. We found that PD1+ CD4 T cells co-expressing IA or ICP were higher in participants with high HIV DNA compared to low HIV DNA (table 1). PD1+ CD4 T cells (unstimulated) also showed correlations with ENV antigen activated circulating T follicular helper cells (Tfh) expressing CD40L (r=-0.41, p<0.05) with selective induction of IL2 (r=0.47, p<0.05) suggesting that PD-1 expression on CD4 can be associated with dysfunctional T:B cells interaction in response to HIV antigens. Conclusion: This study confirms that vertically HIV infected children and young adults under long-term viral control maintain the association between expression of PD1 on CD4 T cells and size of viral reservoirs and also implicates the size of the viral reservoir in altered Tfh functionality.

deletions, allowing reconstruction of the ancestor that infected these cells and is likely similar to both the founder virus and to variants comprising the HIV-1 reservoir. Indeed, the reconstructed virus matched an intact provirus from the same sample, demonstrating the accuracy of the approach. Conclusion: Despite very different immune systems, the HIV-1 proviral landscapes on ART were not obviously different between children and adults, with most proviruses containing large 3’ deletions. The low numbers of infected cells in children and in early-treated adults makes it difficult to detect intact proviruses. Here, we demonstrate the utility of viral reconstruction to infer the genetics of possible transmitted founder viruses and of intact proviruses that may comprise the HIV-1 reservoir. Characterizing the genetics of the HIV-1 reservoir in early-treated individuals can help guide the design of therapeutic interventions towards HIV remission. 812 CELL-ASSOCIATED HIV-1 DNA/RNA IN CHILDREN: PERFORMANCE OF REAL-TIME AND DIGITAL PCR Kathleen Gärtner 1 , Triantafylia Gkouleli 1 , Judith Heaney 2 , Paul Grant 2 , Sara Dominguez Rodriguez 3 , Eloise Busby 4 , Denise O'Sullivan 4 , Moira J. Spyer 1 , Caroline Foster 5 , Pablo Rojo Conejo 3 , Deborah Persaud 6 , Anita De Rossi 7 , Jim Huggett 4 , Eleni Nastouli 2 , for the EPIICAL Consortium 1 University College London, London, UK, 2 University College London Hospitals NHS Trust, London, UK, 3 Hospital Universitario 12 de Octubre, Madrid, Spain, 4 National Measurement Laboratory, Teddington, United Kingdom, 5 Imperial College Healthcare NHS Trust, London, UK, 6 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 7 University of Padova, Padova, Italy Background: Children perinatally infected with HIV-1 (PaHIV) require life-long antiretroviral treatment (ART). Despite ART, HIV persists in a latent reservoir, the cause of viral rebound after treatment interruption (TI). Robust methods to quantify the reservoir in perinatal infections are required. To detect cell-associated HIV-1 DNA (CA-DNA) and cell-associated HIV-1 RNA (CA-RNA) in suppressed PaHIV we compared two methods: quantitative Real Time PCR (qPCR) and digital droplet PCR (dPCR). Methods: In the CARMA-EPIICAL study, 40 European PaHIV on suppressive ART for ≤5 years were recruited. Total CA-DNA, total CA-RNA and unspliced (US) CA-RNA were quantified using qPCR (C1000, Bio-Rad) and dPCR (QX100 Droplet Analyser, Bio-Rad). Nucleic acids were extracted from PBMCs using the DSP virus/pathogen mini kit (Qiagen) on the Qiasymphony. Quantitative qPCR and dPCR were performed using primers in the LTR region for total CA-DNA and total CA-RNA and the pol region for US CA-RNA. To normalise copy numbers of CA-DNA and CA-RNA per 10 6 PBMCs reference genes were included in multiplex reactions. For qPCR a standard curve with known copy numbers was used in a 10-fold dilution series. The concordance analysis of qPCR and dPCR was determined with the Bland-Altman test and significance with Wilcoxon rank test. Results: HIV-1 CA-DNA could be detected in 36 of 40 PaHIV (<10-410 c/10 6 PBMCs for qPCR; <10-1420 c/10 6 PBMCs for dPCR). In seven of the 36 PaHIV CA-DNA copy numbers were below 10 c/10 6 PBMCs. Total CA-RNA was detected in 31 (<11-5789 c/10 6 PBMCs for qPCR, 11-857 for dPCR) and US CA-RNA in 23 of 40 patients (<11-274 c/10 6 PBMCs for qPCR, 11-325 for dPCR). Copy numbers of CA-DNA were significantly higher than CA-RNA, total CA-RNA copy numbers were significantly higher than US CA-RNA (see figure). Concordance analysis showed 97.4% agreement between qPCR and dPCR for total and US CA-RNA. Conclusion: We have demonstrated the detection of very low HIV CA-DNA and CA-RNA levels using both qPCR and dPCR in well suppressed PaHIV. The high agreement of concordance analysis suggests comparability of qPCR and dPCR for detecting low copy numbers of CA-RNA and validates use of both methods for diagnostic applications. The very low levels of CA-RNA expression could contribute to chronic immune activation and/or lead to production of infectious viruses. Further work to determinate the sensitivity of both methods and validate lower thresholds for CA-DNA and CA-RNA will be done.

Poster Abstracts

811 PROVIRAL LANDSCAPE IN CHILDREN PARALLELS ADULTS AND ENABLES RESERVOIR RECONSTRUCTION Jenna M. Hasson 1 , Mary Grace K.Katusiime 2 , Samuel Smith 3 , Michael J. Bale 1 , Liliana Perez-Rodriguez 3 , Divya Kilam 3 , Wei Shao 4 , Mark Cotton 2 , Eli A. Boritz 3 , John M. Coffin 5 , John W. Mellors 6 , Sean Patro 1 , Gert U.van Zyl 2 , Mary F. Kearney 1 1 National Cancer Institute, Frederick, MD, USA, 2 Stellenbosch University, Cape Town, South Africa, 3 NIAID, Bethesda, MD, USA, 4 Leidos Biomedical Research, Inc, Frederick, MD, USA, 5 Tufts University, Boston, MA, USA, 6 University of Pittsburgh, Pittsburgh, PA, USA Background: Characterizing HIV-1 proviruses that lead to viral rebound upon ART interruption could inform design strategies towards a functional cure. Methods for measuring the HIV-1 reservoir, such as the quantitative viral outgrowth assay, require collecting large sample volumes that are difficult to obtain from children. Here, we profile the proviral landscape in children and demonstrate the utility of “viral reconstruction” to characterize the genetics of the HIV-1 reservoir when sample volumes or proviral copy numbers are low. Methods: We performed near-full length (NFL) single-genome sequencing on 210 amplicons from PBMC of two children treated relatively early (9.0 and 9.3 months) and on ART for 7 years. The proviral landscape was compared to that of adults on ART (1056 genomes in the Proviral Sequence Database). Because recovering intact proviruses is rare in children and in adults who initiate ART early, we used the population of defective proviruses to reconstruct NFL ancestors that may be similar to the founder virus and/or to intact proviruses that persist on ART. Results: Similar to adults, ~98% of the proviruses were defective including 60%with large 3’ deletions of env/tat/rev. Proviral diversity (0.3% and 0.7% in p6-PR-RT) and proviral copy number (47 and 182 copies/10 6 PBMC) were low. In the child with the lower HIV-1 diversity and fewer 3’ deletions, we identified defective proviruses with sequences identical except for non-overlapping

CROI 2020 301