Home Search Abstracts Browse Sessions Program Committee E-mail Abstract Author View Session


Session 75 Poster Abstracts
Neuropathogenesis: Clinical Correlates and Observational Studies
Tuesday, 1:30 - 3:30 pm
Poster Hall


505    
Cerebrospinal Fluid HIV Concentration is Highly Associated with Blood CD8+ T Cell Activation and Antiretroviral Treatment Responses
E Sinclair*1, R Ronquillo2, M Bigos2, A Nilsson1, S Deeks1, C Csajka1, D Verotta1, and R Price1
1Univ. of California, San Francisco, USA and 2Gladstone Inst. Virology and Immunology, San Francisco, CA, USA

Background:  HIV RNA is frequently detected in the CSF, but its relationship to the concentration of HIV in plasma is highly variable. To examine factors contributing to CSF HIV levels, we performed a multivariate analysis of salient laboratory and clinical variables (including markers of T-cell activation in blood and CSF and treatment status) in a cohort sample.

Methods:  This is a cross-sectional analysis of baseline findings in the Sentinel Neurological Cohort of HIV-infected subjects recruited for longitudinal CSF analysis. Subjects were segregated into 3 groups based on treatment status and plasma (and CSF) HIV concentrations at study entry: Group S (successes), 26 subjects on antiretroviral therapy (ART) with plasma or CSF viral loads <500 copies/mL; F (failures), 21 subjects on ART with a viral load >500 copies; and N (no ART), 27 off therapy. Six-color flow cytometry measured the percentage of CD3+CD4+ and CD3+CD8+ cells expressing the activation markers HLA-DR (DR+) and CD38 (38+) in blood and CSF. A generalized additive model (GAM, as implemented by Splus) was used to determine the order of influence on the response marker CSF HIV viral load of candidate explanatory variables including plasma HIV viral load, subject group, blood CD4 and CD8 T cell counts, blood and CSF CD4+ and CD8+ activation, CSF white cells, CSF protein and CSF:blood albumin ratio.

Results:  Plasma viral load was not significantly different in groups N and F but both differed from group S. By contrast, treatment had a greater effect in the CSF, with the F group differing from N and nearly as low as S (see the table). The order of CD8+ T-cell activation in blood was like that of the CSF viral load:  group N>F>S. The GAM analysis showed that the order of influence on the CSF viral load was: subject treatment group > blood CD8+ cell activation > plasma HIV, with little additional effect of the other variables.

 

 

Plasma Viral Load

CSF Viral Load

Blood

CD8+DR+38+

Group

Median log10

Viral Load

Median %

N

4.24

3.42

63.1

F

3.86

1.40

39.1

S

1.24

1.24

25.6

 

Conclusions:  These results show that ART generally has a greater effect on CSF than plasma HIV concentrations, an effect evident even in the presence of drug resistance with limited systemic treatment response and continued viremia. The superior benefit of partially effective treatment in CSF compared to plasma may be mediated by reduction of peripheral immune cell activation. Because CD8+ T cells are not likely responsible for transport or enhancement of infection in the CSF compartment, their activation may serve as an indicator of parallel events involving other cells or factors that are less readily measured. 

Keywords: CSF; CD8 T cells; treatment