CROI 2017 Abstract e-Book

Abstract eBook

Oral Abstracts

found at high frequencies in the contemporaneous CSF sample (42-50%). 12/13 participants did not show significant compartmentalization between CSF and plasma. None of the participants carried major drug-resistance mutations in either compartment. Conclusion: This study provides evidence, for the first time, that HIV-1 compartmentalization in the CNS can occur as early as AHI, within days of exposure, though this early compartmentalization onset is uncommon. Further studies are needed to understand the forces determining initial sequestration or enrichment of T/F variants in the CNS compartment, and potential long-term implications of these findings for persistence of HIV-1 reservoirs and neurological impairment in HIV. 72 CD4/CD8 RATIO ASSOCIATES WITH NEUROPSYCHOLOGIC PERFORMANCE DURING EARLY HIV INFECTION Leah T. Le 1 , Kevin Robertson 2 , Brinda Emu 1 , Richard W. Price 3 , Magnus Gisslén 4 , Henrik Zetterberg 4 , Roxane Fabre 5 , Christian Pradier 5 , Serena Spudich 1 , Matteo Vassallo 6 1 Yale Univ, New Haven, CT, USA, 2 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 3 Univ of California San Francisco, San Fransisco, CA, USA, 4 Sahlgenska Academy at the Univ of Gothenburg, Gothenburg, Sweden, 5 Univ de Nice-Sophia Antipolis, Nice, France, 6 Cntr Hosp de Cannes, Cannes, France Background: In chronic HIV, inversion of the CD4/CD8 ratio and its negative trajectory over time are associated with increased morbidity and mortality, systemic immune activation/immunosenescence, and risk of neurocognitive impairment, regardless of antiretroviral therapy (ART) treatment status. We examined whether the trajectory of the CD4/CD8 ratio predicted neuropsychologic performance even during primary HIV infection (PHI). Methods: We longitudinally assessed blood and cerebrospinal fluid (CSF) markers of inflammation/immune activation and neuropsychological testing performance (NPZ4, an average of 3 motor and one processing speed test, and cognitive domains: motor, executive function, processing speed, and memory), in ART-naïve participants enrolled during PHI (<12 months after HIV infection). The majority of participants commenced ART during follow-up independent of the study. Spearman correlation and linear mixed models assessed the relationships between the trajectory of blood CD4/CD8 ratio over time and neurocognitive performance as well as blood and CSF markers of immune activation and injury. Results: 109 PHI participants enrolled (median duration of infection 3.3 months, median plasma HIV RNA 4.63 log 10 copies/mL, 525 total observations, and 4 median observations per participant). At baseline, median CD4/CD8 ratio and NPZ4 scores were 0.53 (IQR: 0.40, 0.86) and -0.20 (IQR: -0.56, 0.34) respectively. Mean CD4/CD8 ratio decreased with longer delayed time from infection to starting treatment (p= 0.01). Greater change in NPZ4 scores correlated with greater change in CD4/CD8 ratio over the course of follow-up (p = 0.02). Multivariate analysis adjusting for time since each visit and baseline, age, viral load, alcohol abuse and treatment status revealed that every unit increase in the CD4/CD8 ratio was associated with a 0.15 increase in NPZ4 score (95% CI: 0.002, 0.29; p = 0.047). Among cognitive domains, change in processing speed correlated with the trajectory of CD4/ CD8 ratio over time (p = 0.03). The trajectory of the CD4/CD8 ratio negatively correlated with change in CSF neurofilament light chain (NFL), a marker of active neuronal injury, independently from cART (p = 0.04). Conclusion: Though early ART had a beneficial effect on the CD4/CD8 ratio, this marker was associated with neuropsychological testing performance and NFL, a marker of brain injury, in PHI, independently from treatment. Its trajectory may thus predict neuropsychologic performance even during early infection. 73 ASYMPTOMATIC HIV-1 CSF ESCAPE IS UNCOMMON AND IS NOT ASSOCIATED WITH NEURONAL DAMAGE Sarah Beth Joseph 1 , Laura P. Kincer 1 , Natalie M. Bowman 1 , Henrik Zetterberg 2 , Magnus Gisslén 2 , Kevin Robertson 1 , Serena Spudich 3 , Joseph J. Eron 1 , Richard W. Price 4 , Ronald Swanstrom 1 1 Univ of North Carolina Chapel Hill, Chapel Hill, NC, USA, 2 Sahlgrenska Academy at the Univ of Gothenburg, Gothenburg, Sweden, 3 Yale Univ, New Haven, CT, USA, 4 Univ of California San Francisco, San Fransisco, CA, USA Background: HIV-1 can be detected in the cerebrospinal fluid (CSF) of some subjects on antiretroviral therapy (ART). The persistence of HIV-1 in the CNS during ART is both a potential barrier to curing HIV-1 and may damage the CNS. Methods: Single copy viral load (VL) testing was done on CSF and plasma samples from 96 neurologically asymptomatic subjects infected with HIV-1 and on stable ART. We examined the relationship between CSF VL and the presence of pleocytosis (a marker of CNS inflammation) and neurofilament light chain protein (NFL; a marker of neuronal injury) in CSF samples. Single genome amplification (SGA) and/or Illumina MiSeq deep sequencing were used to genetically characterize env genes in the CSF of 5 subjects with CSF VLs high enough for analysis. For 2 subjects, we also explored the origins of CSF escape populations by comparing viral populations in the CSF during ART to viral populations in the blood and/or CSF prior to ART (Figure 1). Finally, we examined whether env genes from subjects with CSF escape are macrophage-tropic based on their ability to enter cells expressing low levels of CD4. Results: 17% (16) of subjects had a CSF VL above the limit of detection (LOD) of the single copy assay, and 76% (73) had a plasma VL above the LOD. Further, 6% of subjects (6 of 96) had asymptomatic CSF escape, with plasma VL ≤ 40 cp/ml and a CSF VL > 40 cp/ml, with a median CSF VL of 163 cp/ml. Overall, we did not observe a relationship between CSF VL and CSF NFL (r = 0.06, p = 0.54), nor did subjects with CSF escape have higher levels of CSF NFL in comparison to other subjects (p = 0.1). Further, in longitudinal analyses of two subjects with CSF escape (one persistent and one transient), we did not observe abnormal or increasing levels of CSF NFL. In contrast, CSF escape was associated with greater CSF WBC (median = 2, range = 1-134 cells/μl; p ≤ 0.01). Finally, most CSF escape viruses were CCR5-using and adapted to entering T cells (R5 T cell-tropic); however, one persistent escape population was dominated by moderately macrophage-tropic variants. Conclusion: These cross-sectional results suggest that in individuals without neurologic symptoms, CSF escape is uncommon, and escape viruses are typically R5 T cell-tropic and not associated with detectable neuronal injury. While our findings are limited by the paucity of CSF escape in this well-suppressed cohort and the cross-sectional design, we found no association between asymptomatic CSF escape during ART and a CSF marker of neuronal damage.

Oral Abstracts

28

CROI 2017