CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

444 DISTINCT HIV POPULATIONS IN CSF AND BLOOD DURING ACTIVE CRYPTOCOCCAL MENINGITIS Maura Manion 1 , Camille Lange 2 , Kathy Huppler Hullsiek 3 , Brandon Keele 4 , Helene Highbarger 5 , David Meya 6 , David R. Boulware 3 , Frank Maldarelli 2 , Irini Sereti 1 1 NIAID, Bethesda, MD, USA, 2 National Cancer Institute, Frederick, MD, USA, 3 University of Minnesota, Minneapolis, MN, USA, 4 AIDS and Cancer Virus Program, Frederick, MD, USA, 5 Leidos Biomedical Research, Inc, Frederick, MD, USA, 6 Makerere University, Kampala, Uganda Background: Central nervous system (CNS) infections like cryptoccocal meningitis (CM) are postulated to influence the formation and persistence of an HIV CNS reservoir through trafficking of infected and activated cells, but the mechanisms by which compartmentalization may occur are unknown. Viral compartmentalization present at antiretroviral therapy (ART) initiation may affect the establishment of that reservoir. We investigated the role of pre-ART CM co-infection on the HIV reservoir by characterizing HIV-1 populations in plasma, PBMC and cerebrospinal fluid (CSF). Methods: Patients with HIV and CM infections (N=73) were enrolled in clinical trials of ART initiation (COAT, ASTRO-CM). They had lumbar puncture and phlebotomy prior to anti-fungal therapy and subsequent ART. Plasma and CSF HIV RNA were quantified. A subset of subjects with CSF viral RNA levels > plasma HIV levels (N=9) and subjects with plasma RNA > CSF RNA (N=9) underwent single genome sequencing (SGS) of full length HIV env in plasma and CSF. Cell-associated HIV DNA was recovered for a subset (N=5) of subjects. A total of 786 full length env SGs were obtained from plasma and CSF, and 206 SGs were obtained from cell-associated HIV DNA. SGs were aligned and subjected to phylogenetic analyses (MEGA). Analyses for population shift and R5/X4 tropism predictions (GENO2PHENO) were done. Immunologic and virologic data were analyzed in the context of clinical and demographic data. Results: Subjects with CSF pleocytosis (WBC>5 cells/µL), had higher levels of HIV in CSF than in plasma (p<0.05). In general, the proportions of variants with R5 and X4 CD4 entry phenotype were similar in CSF and plasma, but strong discordance was detected in 4 subjects, reflecting compartmentalization of CD4 phenotypes. Sensitive compartmentalization analysis revealed distinct HIV populations in CSF and plasma in 5 subjects. Additional analysis of cell- associated HIV variants present in PBMC and CSF cells revealed HIV variants in CSF as distinct from HIV in CSF cells, but indistinguishable from PBMC-derived HIV. Conclusion: Pre-ART compartmentalization of HIV populations in CSF and peripheral blood was detectable in the majority of subjects with CM. Increases in CSF cell numbers were associated with elevated levels of HIV in CSF, but the HIV variants present in these cells were not typically present in the CSF HIV populations, indicating that CSF pleocytosis in CM does not substantially contribute to establishing HIV populations in CNS. 445 NEUROLOGIC STABILITY WITH BRIEF ANALYTIC TREATMENT INTERRUPTION AFTER EARLY ART Joanna Hellmuth 1 , Donn Colby 2 , Eugène Kroon 2 , Carlo Sacdalan 2 , Phillip Chan 2 , Jintana Intasan 2 , Victor Valcour 1 , Trevor Crowell 3 , Linda Jagodzinski 3 , Khunthalee Benjapornpong 2 , Nelson L. Michael 3 , Jintanat Ananworanich 3 , Robert Paul 4 , Serena S. Spudich 5 1 University of California San Francisco, San Francisco, CA, USA, 2 SEARCH, Bangkok, Thailand, 3 US Military HIV Research Program, Silver Spring, MD, USA, 4 University of Missouri St Louis, St Louis, MO, USA, 5 Yale University, New Haven, CT, USA Background: The central nervous system (CNS) is a likely reservoir of HIV and is vulnerable to viral rebound and increased inflammation upon cessation of antiretroviral therapy (ART). Thus, careful evaluations of CNS outcomes are critical for HIV remission studies employing analytic treatment interruption (ATI). This study investigated changes in neurologic measures across three small, closely monitored HIV remission trials involving ATI in participants who previously initiated ART during acute HIV infection (AHI). Methods: Pre-ATI, participants received either vorinostat/hydroxychloroquine/ maraviroc (n=7), or no added intervention (n=9). Eight other participants received the broadly neutralizing VRC01 antibody at and during ATI. Criteria for restarting ART included confirmed plasma HIV RNA >1,000 cps/mL. Pre- and post-ATI (on the day of, or after ART resumption) assessments included standard

measures of mood and anxiety; ACTG-derived macroneurological exam; Color Trails 1 and 2; Grooved Pegboard; Trail-making A; and the computerized Flanker Task. Elective tests included cerebrospinal fluid (CSF) sampling (pre-, during ATI at first plasma HIV RNA > 20 cps/mL, and post-ATI) and brain diffusion tensor imaging (DTI; pre- and during ATI). Analyses employed paired t-test and ANOVA. Results: At ART initiation, 54% of participants were in Fiebig I/II. ATI was preceded by a median of 3.4 years on cART (IQR 2.6-4.8). Median ATI duration was 30 days (IQR 19-37). Comparing pre- vs. post-ATI measures, there was no change in PHQ-9 depression score (3.9 vs. 4.8; p=0.510; n=14), HADS depression score (2.2 vs. 2.2; p>0.999; n=14), HADS anxiety score (3.6 vs. 4.2; p=0.677; n=14), Distress Thermometer rating (2.2 vs. 3.4; p=0.252; n=14), prevalence of neurologic findings (26% vs. 26%) or number of neurologic findings (0.7 vs. 0.6; p=0.516; n=23). The global neuropsychological test z-score modestly improved from pre-ATI to post-ATI (0.7 vs. 1.0; p=0.007 n=12). There were no differences in Flanker performance pre-, during and post-ATI. Two participants had detectable CSF HIV RNA during ATI before cART resumption, one at 29 days (25 cps/mL) and one at 34 days (42 cps/mL). There were no differences in DTI measures pre- vs. during ATI (n=12). Conclusion: In a small sample, we identified no adverse neurologic outcomes in AHI participants who underwent brief, closely monitored ATI. Further studies are needed to validate the CNS safety of ATI in HIV remission trials for longer ATI durations and in other populations. 446 SPECIFIC CSF VIRAL ESCAPE FINDINGS COMPARED TO PRE-cART HIV ENCEPHALITIS Valentina De Zan 1 , Francesca Ferretti 2 , Simonetta Gerevini 1 , Laura Passeri 1 , Adriano Lazzarin 1 , Cinque Paola 1 1 San Raffaele Scientific Institute, Milan, Italy, 2 Chelsea and Westminster NHS Foundation Trust, London, UK Background: Suppressive combined antiretroviral therapy (cART) has drastically reduced the incidence of HIV encephalitis, however CNS breakthrough, or ‘cerebrospinal fluid (CSF) viral escape’, is an emerging phenomenon resulting in significant neurological debilitation. Although both pre-cART HIV encephalitis (HIVE) and CSF escape encephalitis (esc-HIVE) result from HIV replication in the CNS, they seem to differ in several ways. Aim of this study is to characterize and compare clinical, radiological and CSF aspects of these two conditions. Methods: We retrospectively examined clinical, radiological and CSF data from patients with either esc-HIVE (defined by occurrence of incident neurological symptoms in patients on ART for >9 months and detectable CSF HIV-RNA while undetectable in plasma, or CSF HIV-RNA >2-fold than plasma levels) or HIVE. We also compared levels of immune activation markers (CCL2, CXCL10, suPAR) measured in cryopreserved CSF samples from 8 esc-HIVE and 17 HIVE patients. Comparisons were made with Mann-Whitney U test or Fisher exact test, as appropriate. Results: Laboratory and CSF findings at the time of diagnosis of esc-HIVE or HIVE are reported in the Table. Clinical symptoms and signs included memory and cognitive impairment (esc-HIVE=10, HIVE=10), cerebellar signs (esc- HIVE=11, HIVE=3), focal signs (esc-HIVE=8, HIVE=2), alteration of consciousness (esc-HIVE=4, HIVE=3), agitation/psychosis (esc-HIVE=0; HIVE=4). In esc-HIVE, MRI showed areas of white matter hyperintensity either involving the periventricular or other brain or cerebellum regions in 16/19 cases (84%), edema with sulcal effacement in 10/19 cases (53%), and no abnormalities in 3 cases. In patients with HIVE, MRI findings showed cortical and/or subcortical atrophy in 14/16 cases (88%) and diffuse periventricular white matter hyperintensity, most frequently symmetrical involving frontal lobes, in 11/16 cases (69%). Conclusion: Despite some similarities in clinical presentations between esc-HIVE and pre-cART HIVE, MRI frequently showed an inflammatory pattern in esc-HIVE, with no atrophy, which was in turn common in all pre-cART HIV cases. In addition, CSF cells and proteins were higher and levels of HIV replication and macrophage activation CSF markers were lower in esc-HIVE. These findings suggest different underlying mechanisms between the two entities, with esc-HIVE associated with lower extent of HIV replication and presence of inflammatory response in the CNS.

Poster Abstracts

CROI 2018 158