CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

362 PREVALENCE OF NEUROCOGNITIVE DISORDERS IN A WELL-TREATED AND AGING SWISS HIV-COHORT

Melanie Metral 1 , Isabella Locatelli 1 , Peter Brugger 2 , Klemens Gutbrod 3 , Ursula Kunze 4 , Isaure Nadin 5 , Marc Schwind 6 , Riccardo Pignatti 7 , Matthias Cavassini 1 , Renaud Du Pasquier 1 1 Univ of Lausanne, Lausanne, Switzerland, 2 Univ Hosp of Zürich, Zürich, Switzerland, 3 Univ Hosp of Bern, Bern, Switzerland, 4 Felix Platter Hosp, Basel, Switzerland, 5 Univ Hosps of Geneva, Geneva, Switzerland, 6 Klinik für Neuro, St. Gallen, Switzerland, 7 Servizio di Malattie Infettive, Lugano, Switzerland Background: The prevalence of HIV-infected associated neurocognitive disorders (HAND) remains high. In particular, the prevalence of HAND and the contribution of confounding factors in a well-treated aging population are unknown. This study aims to determine the prevalence and incidence of neurocognitive disorders (NCD) through longitudinal neurocognitive assessment over a 4 years period in a large well-treated and aging HIV cohort. In the present abstract, findings of the baseline milestone are presented. Methods: This is an ongoing, prospective, longitudinal, multicenter and multilingual (German, French and Italian) study within the Swiss HIV Cohort Study called the NAMACO study (Neurocognitive Assessment in the Metabolic and Aging COhort). So far, we have analysed 899 HIV-infected subjects older than 45 years who have undergone a thorough standardized neuropsychological assessment inspired by the tests used in the START study covering 7 cognitive domains (memory, attention, speed of information processing, executive functions, language, motor skills and sensory-perceptive abilities). To diagnose HAND, we referred to the Frascati criteria. Results: Our cohort (mean age 55±7.5 years, 80%men, 92% Caucasian) had a high proportion of undetectable viral load (96%, VL<50 copies/ml; median current CD4 632 cell/mm3 and median nadir CD4 180 cell/mm3). At baseline, the prevalence of NCD was 41%, with 27% of patients suffering from HAND (26% asymptomatic neurocognitive impairment (ANI), 0.7%mild neurocognitive disorder, and 0.7% HIV-associated dementia) and 14% due to”other”factors (mostly psychiatric disorders). Lower employment status (36%) was associated with a higher prevalence of ANI (31%) and”other”(25%). Motor skills (42% of our population), speed of processing (33%) and attention/working memory (33%) were the most frequently impaired domains. Conclusion: We were able to implement a standardized neuropsychological testing in three national languages, which provides a unique comprehensive view of the cognitive status of older HIV-infected patients in Switzerland. Despite the fact that HIV in this cohort is very well controlled (96% aviremic), the prevalence of HAND remains high (27%). However, the HAND population is composed mostly of ANI. Psychiatric disorders represent the most important confounding factor, accounting for one third of cognitive disorders. The most affected domains were motor skills, speed of processing and attention/working memory. 363 CNS MONITORING OF CART INTERRUPTION IN INDIVIDUALS TREATED DURING FIEBIG I ACUTE HIV Phillip Chan 1 , Donn Colby 1 , Victor Valcour 2 , Napapon Sailasuta 3 , Shelly J. Krebs 4 , Linda Jagodzinski 4 , Robert Paul 5 , Jintanat Ananworanich 4 , Serena Spudich 6 , for the RV411 and RV254/SEARCH 010 Study Groups 1 SEARCH, The Thai Red Cross AIDS Rsr Cntr, Bangkok, Thailand, 2 Univ of California San Francisco, San Francisco, USA, 3 Univ of Hawaii, Honolulu, HI, USA, 4 US Military HIV Rsr Prog, Silver Spring, MD, USA, 5 Univ of Missouri–St. Louis, St. Louis, MO, USA, 6 Yale Univ, New Haven, CT, USA Background: Monitored treatment interruption (TI) of combination antiretroviral therapy (cART) is a frequent strategy in studies of HIV remission. We assessed the central nervous system (CNS) effects of TI after sustained treatment started in early acute HIV infection (AHI) to monitor for adverse consequences of inflammation or dysfunction in the CNS and examine sources of viral rebound following withdrawal of cART. Methods: Eight participants with >24 months of viral suppression after a median of 2.8 (range 2.5-5.5) years of treatment initiated during Fiebig I (HIV RNA+, p24-, IgM-) AHI elected for TI followed by plasma HIV RNA monitoring every 3-7 days and cART resumption once plasma HIV RNA >1,000 cps/ml. Optional CNS measures during stable therapy prior to TI (pre-TI) and after post-TI rebound viremia (any plasma HIV RNA of >20 cps/ml) included cerebrospinal fluid (CSF) HIV RNA (LLOD 80 cps/ml), soluble CSF and plasma immune activation markers, and cerebral metabolites via 3T brain magnetic resonance spectroscopy (MRS). Participants completed a novel tablet-based measure of attention and inhibitory control from the NIH-Toolbox (Flanker test) during screening (wk -4), pre-TI wk 0, post-TI wk 1, after rebound viremia, and at 2, 4, and 12 weeks after resuming cART. Results: Plasma HIV rebound occurred at a median 26 (range 13-48) days after TI. Four participants underwent CSF sampling, revealing no detectable CSF HIV RNA (<80 cps/ml) pre-TI and during rebound viremia (corresponding plasma HIV RNA 30, 78, 739 and 13,462 cps/ml). Levels of CSF and plasma neopterin, MCP1, IP10, and sCD14 did not significantly change from pre-TI to rebound viremia. Brain MRS in basal ganglia and frontal white matter in 5 participants similarly revealed no significant changes in neuronal or inflammatory measures from pre-TI to rebound viremia. Flanker task performance remained similar across intervals of pre-TI wk 0 to plasma rebound (p=0.30). Flanker scores improved from the end of the TI to 12 weeks after resumption of cART (p=0.045) to levels exceeding performance at pre-TI. Conclusion: Longitudinal monitoring of the CNS did not reveal evidence of adverse CNS effects of TI after cART started during very early acute HIV infection. All of the four participants who had CSF sampling had no detected HIV RNA in the CSF during rebound plasma viremia. Examination of the CNS is feasible to integrate in systemic HIV studies and critical for a comprehensive understanding of the impact of interventions aimed at viral remission. 364 CSF HIV-1 COMPARTMENTALIZATION BY ENV DEEP SEQUENCING: RELATION TO NEURONAL INJURY Richard W. Price 1 , Magnus Gisslén 2 , Laura P. Kincer 3 , Ean Spielvogel 3 , Amy Lin 1 , Jaser Eusuff 4 , Serena Spudich 5 , Ronald Swanstrom 3 , Sarah Beth Joseph 3 , for theTHINC Study Group 1 Univ of California San Francisco, San Francisco, CA, USA, 2 Sahlgenska Academy at the Univ of Gothenburg, Gothenburg, Sweden, 3 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 4 Univ of California Berkeley, Berkeley, CA, USA, 5 Yale Univ, New Haven, CT, USA Background: HIV-1 infection of the central nervous system (CNS) is a constant feature of systemic infection and HIV RNA is detectable in cerebrospinal fluid (CSF) of virtually all untreated viremic individuals. CSF HIV can derive from circulating blood (equilibrated) but also from locally replicating populations (compartmentalized) due to either clonal expansion or sustained replication and evolution of diverse populations. We used partial HIV env deep sequencing to survey matched CSF and blood from untreated individuals grouped by blood CD4+ T cell counts, clinical status, and concentrations of CSF neurofilament light chain (NFL), a sensitive biomarker of neuronal injury. Methods: We cross-sectionally examined 43 archived CSF and blood sample pairs from untreated individuals from Gothenburg and San Francisco, and successfully generated sequences from 33 pairs divided into 4 groups: 1. neuroasymptomatic (NA) CD4 >200 cells/µL (N=8); 2. NA CD4 <200 with normal CSF NFL (NFL-) (N=8); 3. NA CD4 <200 with elevated NFL (NFL+) (N=10); and 4. clinically diagnosed HAD (N=7), all with elevated CSF NFL. Viral RNA was used to generate cDNA, and the V1/V3 region of env amplified and sequenced using Illumina MiSeq deep sequencing with PrimerID. Neighbor joining phylogenetic trees were constructed for each subject and visually segregated into 3 types: A. genetically diverse compartmentalized CSF populations comprising >1/3 of CSF env genes; B. CSF clonal expansions with limited diversity of CSF env genes; and C. fully equilibrated CSF and plasma populations. Results: Compartmentalization of CSF HIV populations was common through the course of infection, though its character changed (Table). Whereas clonal expansion predominated in early infection and in later infection without neuronal injury (Groups 1 and 2), diverse compartmentalized populations predominated in late infection with neuronal injury (Groups 3 and 4). However, all groups contained some individuals with diverse compartmentalization or equilibrated infection. Conclusion: Diverse compartmentalization in most individuals with elevated CSF NFL supports the importance of local CNS HIV replication to CNS injury. However, clonal expansion and equilibrated infection in some with elevated CSF NFL suggests that either it is not the sole or simple determinant of neuronal injury and that additional pathogenic factors are also involved, or that CSF sample and sequencing selection did not fully represent local replication in brain.

Poster and Themed Discussion Abstracts

CROI 2017 144