CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

SGS represents a useful tool for reservoir assessment and further sequencing and analysis with respective clinical data is ongoing in the CARMA cohort.

814 DIFFERENCES IN THE INDUCED LATENT HIV RESERVOIR IN PERINATAL AND ADULT INFECTIONS Adit Dhummakupt 1 , Jessica Rubens 1 , Thuy Anderson 1 , Laura Powell 1 , Bareng Nonyane 2 , Lilly V. Siems 3 , Aleisha Collinson-Streng 1 , Tricia Niles 2 , Richard Jones 4 , Vicki Tepper 1 , Allison Agwu 1 , Deborah Persaud 1 1 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 2 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA, 3 Abt Associates, Inc, Cambridge, MA, USA, 4 George Washington University, Washington, DC, USA Background: The HIV latent reservoir in resting memory CD4+ T cells prevents cure. Novel therapies to reactivate and eliminate the reservoir are in clinical trials in adults, but not yet in pediatric populations. Methods: HIV proviral reservoir size was determined in perinatal (N=11) and adult infections (N=10) by digital droplet PCR (ddPCR) and with the intact proviral DNA assay (IPDA) in perinatal samples. The inducibility of the latent reservoir was determined with the Tat/rev Induced Limiting Dilution Assay (TILDA) that uses single-round (12 hr) T cell stimulation of CD4+ T cells with PMA/ ionomycin to maximally activate cells to induce proviral expression, measured as multiply-spliced HIV RNA Units per 10 6 CD4 cells (msRUPM). Markers of immune activation (CD69, CD25 and HLA-DR) and exhaustion (PD-1, TIM-3 and TIGIT) were also assessed. An enhanced TILDA with addition of PHA and for 18 hours was performed to enhance proviral expression in perinatal infections. Non-parametric tests were used for differences between paired and unpaired measurements; correlations were quantified by Spearman rank coefficient. Results: The median age was 15.8 yrs with a median duration of suppression of 6.7 yrs for perinatal infections, and 40.5 yrs with a median duration of suppression of 7.3 yrs for adult infections. We found that despite a higher proviral reservoir size (median 132.1 vs. 66.7 c/10 6 PBMCs) and similar rates of T cell activation with PMA/ionomycin (median %CD69 = 96.7% and 93.0%) in perinatal and adult infections, respectively, the size of the induced reservoir was significantly lower in perinatal than in adult infections (median msRUPM of 2.99 vs 11.92, p=0.020). With the enhanced TILDA, the size of the induced reservoir increased significantly in perinatal infections (1.5-fold to a median of 4.5 msRUPM; p=0.034), but not in adult infections. The proportion of induced provirus was significantly lower in perinatal infections at 1.6% compared with 4.0% in adult infections (p=0.030). At baseline, the proportion of HLA-DR+ T cells was significantly lower in perinatal compared with adult infections (median HLA-DR+ cells = 4.56% vs 10.5%, p=0.012), but not correlated with the induced reservoir size. Conclusion: The inducibility of the latent reservoir is substantially lower in perinatal compared with adult infections, possibly due to differences in baseline states of immune activation, with implications for latency reversal strategies towards ART-free remission.

813 ASSESSMENT OF HIV-1 DNA BY SINGLE-GENOME SEQUENCING IN CHILDREN ON SUPPRESSIVE ART Kathleen Gärtner 1 , Triantafylia Gkouleli 1 , Matthew Byott 2 , Judith Heaney 2 , Moira J. Spyer 1 , Anita De Rossi 3 , Deborah Persaud 4 , Paolo Palma 5 , Carlo Giaquinto 3 , Pablo Rojo Conejo 6 , Caroline Foster 7 , Anne-Geneviève Marcelin 8 , Paolo Rossi 5 , Eleni Nastouli 2 , for the EPIICAL Consortium 1 University College London, London, UK, 2 University College London Hospitals NHS Trust, London, UK, 3 University of Padova, Padova, Italy, 4 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 5 Bambino Gesu Children's Hospital, Rome, Italy, 6 Hospital Universitario 12 de Octubre, Madrid, Spain, 7 Imperial College Healthcare NHS Trust, London, UK, 8 Pitie-Salpetriere Hospital, Paris, France Background: Perinatally HIV-1 infected children on early suppressive ART (PaHIV) represent an important population in the era of cure. Reservoir characteristics can determine criteria and cut-offs in future trials. We present unique HIV-1 DNA single genome sequencing (SGS) data in PaHIV from a multicentre cross-sectional study. We describe the presence of intact proviruses, defective genomes and ART resistance associated mutations (RAMs). Methods: In the CARMA-EPIICAL study, 40 PaHIV on ART since <2 years of life and suppressed for ≥5 years were recruited in 7 centers. HIV-1 DNA was measured by real-time PCR. Near full-length SGS was performed in positive samples: manual extraction and limiting dilution touchdown PCR generated amplicons across the genome using in house and published primers. Products were analysed by gel, libraries generated by Nextera®XT DNA Kit and sequenced on the Miseq (Illumina). De novo assembly of genomes was performed, using an in-house bioinformatics pipeline with open source software. SMALT and LASTZ for alignment and the HIVSeqinR pipeline were utilised to describe intact or otherwise defective viruses. The Stanford database was used for resistance/ subtype analysis and the Geno2pheno tool for tropism assignment. Results: The majority of patients, 34/40, had detectable HIV-1 DNA (median 115.1 c/10 6 PBMCs, range 49.1-260.7 c/10 6 PBMCs). Initial findings on 4 patients, where 10 near full-length sequences were generated, are included in the figure. 2 viruses were subtype C, 1 subtype G and 1 CRF02_AG. Tropism assignment was possible in 9/10 sequences, 4/9 were CCR5, 5/9 were CXCR4. Intact sequences were identified in 4/10, however all contained APOBEC 3G/F mutations. In 2 patients (3/10 sequences) RAMs in the pol/RT region were found. Defective genomes were frequent, 6/9 genomes contained deletions (4/6 were large), while 2/9 sequences had frameshift mutations. Inversions and stop codon mutations were not detected. Conclusion: Our preliminary findings in this cohort suggest that the HIV-1 DNA landscape in PaHIV can be complex. Defective genomes with large deletions can be frequent but intact genomes are also present suggestive of a pool of virus that can rebound post treatment interruption. However, host driven APOBEC related hypermutations are present in long standing treated infection as well as RAMs.

Poster Abstracts

CROI 2020 302