CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Methods: We selected 268 HIV+ adults who were assessed in the CHARTER study and who were taking 3-drug ART; underwent lumbar puncture; and had HIV RNA in plasma and CSF ≤ 50 copies/mL. CXCL10, TNF-α, and IL-6 were quantified in CSF by bead suspension array immunoassay in 2 independent samples: the Training Set (TS, n=144) and the Validation Set (VS, n=124). Data were analyzed using conventional parametric, non-parametric, and multivariable regression methods. Results: The two sets were similar in age, self-reported race/ethnicity, sex, body mass index, AIDS status, current and nadir CD4 + T-cell count, self-reported ART duration, and CPE values. The most common ART regimen was efavirenz-tenofovir-emtricitabine (n=49). All HIV RNA levels were ≤ 50 copies/mL. In both groups, higher CPE values correlated with lower levels of CXCL10 (TS: r=-0.30, p<0.001; VS: r=-0.30, p<0.001) and TNF-α (TS: r=-0.22, p=0.01; VS: r=-0.19, p=0.03) but not IL-6 (TS: r=-0.11, p=0.23; VS: r=0.07, p=0.48). Multivariable models combined data from both groups and adjusted for other statistically significant correlates of each biomarker as well as correlates of CPE. These models confirmed the associations between higher CPE values and either lower CXCL10 (model R 2 =0.18, p<0.001) or TNF-α levels (model R 2 =0.05, p=0.04). Conclusions: During suppressive ART, regimens that are estimated to have better distribution into the CNS are associated with less inflammation in CSF, as indicated by levels of CXCL10 and TNF-α. This may reflect better suppression of HIV RNA in the CNS. Estimated ART drug distribution into the CNS did not correlate with IL-6 levels in CSF, which is consistent with findings that IL-6 can both remain elevated during suppressive ART and be associated with HAND (e.g., JAIDS 2015 68:281-8). While cross-validation is a strong approach, longitudinal analyses are needed to further strengthen confidence in these findings. 413 Role of HCV Coinfection on CSF Biomarkers in HIV Patients Andrea Antinori 1 ;Valentina Fedele 1 ; Carmela Pinnetti 1 ; Patrizia Lorenzini 1 ; Stefania Carta 1 ;Veronica Bordoni 1 ; Federico Martini 1 ; Francesca Ceccherini-Silberstein 2 ; Adriana Ammassari 1 ; Carlo Federico Perno 2 1 Inst Nazionale Malattie Infettive “L. Spallanzani”, Rome, Italy; 2 Univ of Rome Tor Vergata, Rome, Italy Background: The effect of HCV co-infection on neurological injury has not been explored. Aimwas to evaluate changes in CSF biomarkers of neurodegeneration and inflammation and to analyse contribution of HCV to neuroinjury. Methods: Retrospective analysis of neopterin, neurofilament (NFL), sCD14 (ELISA assays), MCP-1 ( Luminex assay) concentrations in CSF/plasma paired samples from HIV-infected patients (pts). Lumbar puncture (LP) achieved for neurological signs/symptoms, CNS staging of lymphoma, or neurological disease. By fitting a multivariate linear regression model, factors independently associated with CSF biomarkers were identified. Results: 238 CSF/plasma pairs from 234 pts were included: 79%male, median age 47 (IQR, 39-51), HIV transmission: heterosex 34%, MSM 21%, IVDU 27%; 79% in CDC C. Neurological signs/symptoms in 37.8%. At LP, median CD4 was 184/mm3 (IQR 93-408), nadir 75/mm3 (IQR 21-156); 173 pts (72.7%) were on ARV, with median log 10 HIV-RNA of 2.2 (IQR 1.6-4.4) in plasma and <1.7 (IQR 1.6-3.2) in CSF. In 51% plasma HIV-RNA was <50 cp/mL. Median log 10 HCV-RNA in plasma and CSF was 5.8 and 1.0. Detectable HIV-RNA in both plasma and CSF was independently associated with higher levels of neopterin (p=0.001) and NFL (p=0.002), while CD4>350/mmc with lower levels of MCP-1 (p=0.022) and of sCD14 (p=0.009), when compared with CD4<200. Pts with neurological disease showed a higher concentration of sCD14 (p=0.013). 67 pts were HIV+/HCV+, with detectable plasma HCV-RNA in 52 (29%) (median log 10 HCV-RNA in plasma: 5.7); 13 (25%) of them had detectable CSF HCV-RNA (median log 10 1.04). At multivariable analysis, higher plasma HCV-RNA was associated to a lower concentration of CSF neopterin (p<0.001), NFL (p<0.001), MCP-1 (p= 0.001) and sCD14 (p= 0.001). Differences in biomarkers are shown in Fig. 1. In a multivariate analysis restricted to HCV-RNA+ pts, higher CSF HCV-RNA was associated with increased CSF neopterin (p=0.010), MCP-1 (p<0.001), sCD14 (p=0.005). CSF values of neopterin, MCP-1, sCD14 and NFL in HIV+ remained significantly higher than those observed in HIV+/HCV+. Conclusions: HIV-RNA contemporary detectable in plasma and CSF is associated to increased values of neuroinflammatory and neuroinjury biomarkers. Interstingly, HCV-RNA in CSF independently correlates with increased neuroinflammation and/or neuroinjury, while this effect seems to be reversed in case of HCV-RNA present only in plasma. The clinical implications of these findings are warranted.

Poster Abstracts

157

CROI 2016