CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

405 Early CSF Viremia and CNS T-Cell Infiltrate in a Nonaccelerated SHIV Infection Model Sandhya Vasan 1 ; Piyanate Sunyakumthorn 2 ; Matthew D.Wegner 2 ; Eric Lombardini 2 ; Joanna M. Hellmuth 3 ; Jintanat Ananworanich 4 ;VictorValcour 3 ; Robert O’Connell 2 ; Serena S. Spudich 5 ; Nelson L. Michael 6 1 US Military HIV Rsr Prog, APO, AP, USA; 2 Armed Forces Rsr Inst of Med Scis, Bangkok, Thailand; 3 Univ of California San Francisco, San Francisco, CA, USA; 4 Military HIV Rsr Prog, Bethesda, MD, USA; 5 Yale Univ Sch of Med, New Haven, CT, USA; 6 US Military HIV Rsr Prog, Bethesda, MD, USA Background: Recent studies describe CNS inflammation and viremia during the earliest stages of HIV infection. We sought to develop a non-accelerated model to characterize these effects in rhesus macaques. Methods: Macaques were infected with a single SHIV1157ipd3N4 challenge intrarectally (9 males) or intravaginally (3 females). Weekly plasma SHIV RNA and CD4+ T cell counts were quantified by RT-PCR and flow cytometry. Immune activation markers (IP-10, MCP-1 and IL-15) were measured by Luminex at weeks 2 and 12 post infection (W2 and W12) in plasma and W12 in CSF. At W12 necropsy, brain sections from 6 SHIV+ and 6 uninfected control animals were stained by immunohistochemistry (IHC) with CD3, CD4, and CD68 antibodies and quantified as immunoreactive cells per 40 HPF. Results: Plasma SHIV RNA mimicked early human HIV infection with mean peak and W12 set-point viremia at 6.4 and 5.1 log 10 copies/mL, respectively. CSF viremia was detectable in the 4 animals with highest W12 plasma viral load. Plasma IL-15, MCP-1 and IP-10 at W2 were significantly elevated over baseline (19.7 vs 9.8 pg/mL, p=0.0002; 436 v 215 ng/mL, p<0.0001; 384 vs 70 pg/mL, p=0.0001, respectively), and normalized by W12. However, while there was no elevation of IL-15 in the W12 CSF, MCP-1 and IP-10 W12 CSF levels were significantly elevated over W12 plasma (284 vs 136 ng/mL, p<0.0001; 282 vs 112 pg/mL, p=0.0005) and over normal control CSF (284 vs 126 pg/mL, p=0.03; 282 vs 117 pg/mL, p=0.004). The W12 CSF/serum albumin ratio was <5 x 10 -3 in all 12/12 animals, consistent with an intact blood brain barrier. IHC revealed no evidence of CD68+ or CD4+ infiltrate in midbrain, frontal cortex, or basal ganglia. A subset of SHIV-infected animals had evidence of a mild CD3+ T cell infiltrate with qualitative pavementing along the vascular epithelium, and clustering in the brain parenchyma. Interestingly, CD4+ T cells were increased in the meninges of a subset of SHIV-infected animals vs controls, in the absence of CD68+ cells. Conclusions: Early CSF viremia and IP-10 and MCP-1 elevations reflect a discrete neurovirologic process consistent with published human studies. This is accompanied by a mild, non-specific CD3+, CD4- infiltrate in the in brain parenchyma and CD4+ T cells in the meninges in some SHIV+ animals in the absence of macrophages. The SHIV1157ipd3N4 non- accelerated challenge model reflects pathophysiologic changes characteristic of acute/early HIV infection in humans, and can serve as a model for future interventional studies. 406 Relative Frequency of Drug Resistance Mutations on Individual HIV-1 Genomes in HAND Teresa Evering ; Leslie St. Bernard; Jeremy Abolade; Hiroshi Mohri; Martin Markowitz Aaron Diamond AIDS Rsr Cntr, New York, NY, USA Background: The ability of the CNS to act as sanctuary may result in discordant drug-resistance mutations (DRMs) in CSF and plasma. Compartmentalization and independent evolution of DRMs in CSF HIV-1 variants has been described. We hypothesized that in those with HAND differences in the relative frequency of DRMs in the CSF and plasma would be observed. Methods: We used SGA to study paired CSF and plasma samples of 12 HIV+ subjects from the CHARTER study with no neurocognitive impairment (NCN) (N=5), asymptomatic neurocognitive impairment (ANI) (N=4), mild neurocognitive disease (MND) (N=2) and HIV-associated dementia (HAD) (N=1). Subjects were viremic and on treatment at the time of sampling. The Stanford HIVdb program and phylogenetic analyses were used to identify HIV-1 DRMs on each SGA-derived genome and determine viral compartmentalization respectively. Results: On average, 25 SGS were analyzed per compartment per subject (N=621). We found statistical evidence of compartmentalization between CSF- and plasma-derived viruses (p<0.05) in individuals with NCN (2/5=40%) and HAND (5/7=71%). HIV-1 DRMs were commonly found, with 11/12 (92%) subjects with at least 1 DRM in the CSF or plasma (median 7, range 1-18). Multiple DRMs were commonly identified on SGA-derived genomes (range 0-6). For all subjects with NCN, the relative frequency of each DRM was statistically similar between compartments. In contrast, in subjects with HAND, 57% (4/7) of individuals demonstrated statistically significant differences in the relative frequency of at least one DRM in the CSF and plasma (Bonferroni corrected p -value thresholds from p<0.0028 to p<0.0125). This represents a trend toward a statistical difference in the relative frequency of discordant DRMs in the CSF and plasma in those with NCN and HAND (p=0.08 by Fisher’s exact test). When identified, the relative difference in DRMs reflects a higher percentage of DRMs in the plasma as opposed to the CSF (21 events vs. 1, p<0.0001 by Fisher’s exact test). 3 subjects with HAND (3/7=43%) demonstrated one DRM in the CSF that was not identified in paired plasma. Longitudinal analyses of 2 subjects with HAND reveal the development of significant differences in the relative frequency of DRMs in the CSF and plasma over time - both in the presence (N=1) and absence (N=1) of changes in the cART regimen. Conclusions: Using SGA, statistically significant differences in the frequency of DRMs in the CSF and plasma are readily found in those with HAND. 407 Dual Role of Activated and HIV-Specific CD8 T Cells in CSF During Acute HIV Infection Lydie Trautmann 1 ; Cari Kessing 2 ; Eugène Kroon 3 ; Donn Colby 3 ; Shelly Krebs 4 ; Napapon Sailasuta 5 ; Jintana Intana 3 ;VictorValcour 6 ; Jintanat Ananworanich 4 ; Serena S. Spudich 7 ; for the SEARCH 010/011/013 Study Groups 1 Military HIV Rsr Prog, Silver Spring, MD, USA; 2 The Scripps Rsr Inst, Jupiter, FL, USA; 3 SEARCH, Bangkok, Thailand; 4 Military HIV Rsr Prog, Bethesda, MD, USA; 5 Univ of Hawaii, Honolulu, HI, USA; 6 Univ of California San Francisco, San Francisco, CA, USA; 7 Yale Univ Sch of Med, New Haven, CT, USA Background: HIV enters the CNS compartment as early 8 days post estimated exposure in humans. Infiltration of CD8 T cells into the CNS is a recognized feature of many neurodegenerative diseases and is seen in HIV-associated dementia. The number of CD8 T cells present in the cerebrospinal fluid (CSF) during HIV infection is elevated compared to other CNS diseases; however, the role of these CD8 T cells and whether they are serving a detrimental or protective effect during acute HIV infection is unknown. Methods: We analyzed CD8 T cells from CSF in a unique cohort enrolled during the earliest stages of acute HIV infection (RV254/SEARCH010 cohort, n=26) compared to chronic HIV patients (SEARCH011, n=9, cART naïve) and uninfected controls (RV304/SEARCH013, n=8). We analyzed the absolute number and phenotype of CD8 T cells from the CSF and blood by flow cytometry. Phenotypic analyses were completed; we then correlated these with CSF HIV RNA and inflammatory markers from CSF. After in vitro expansion of the CSF CD8 T cells, we examined Vbeta repertoire usage by flow cytometry; HIV-specificity was determined by intracellular cytokine staining using peptide pools covering Env, Pol, Gag and Nef. Results: The total numbers and frequency of activated CSF CD8 T cells were elevated in chronic and acute HIV-infected participants compared to controls (P=0.04 and P<0.0001, respectively). Activated CD8 T cells correlated with CSF HIV RNA (P=0.001), and with markers of CNS inflammation [neopterine (P=0.0005), IP-10 (P=0.004), CD163 (P<0.0001), and sCD14 (P=0.02)]. CD8 T cells in the CSF harbored a more restricted Vbeta repertoire compared to total and activated CD8 T cells from peripheral blood. During acute infection, different Vbeta families were expanded in the CSF and peripheral blood, suggesting local specific expansions. Finally, HIV-specific CD8 T cells were detected during acute infection in CSF and activated CD8 T cells in blood. These CD8 T cells were found to target different HIV proteins in the CSF compared to peripheral blood. Conclusions: These results highlight the dual role of activated CD8 T cells in CSF during HIV infection. Their increase in the CSF early in HIV infection and their correlation to neuro- inflammatory markers suggest that they could play a role in the development of HIV-associated neuroinflammatory disorders. Their unique T cell repertoire and HIV-specificity in the CSF in acute infection suggest that they could also play a role in controlling viral replication in the CNS compartment.

Poster Abstracts

154

CROI 2016