CROI 2019 Abstract eBook

Abstract eBook

Oral Abstracts

Methods: sCD30 levels were measured in banked cerebrospinal fluid (CSF) samples and matching plasma from healthy HIV-uninfected controls (n=18), HIV-infected viremic individuals (n=52), individuals on suppressive ART (n=40), HIV-controllers (n=10), and participants with CSF escape (plasma RNA <50 copies/mL, detectable CSF RNA; n=10). sCD30 levels (median, IQR) were compared across groups and correlated with CSF HIV RNA and markers of axonal injury and myeloid cell activation using nonparametric tests. Results: Compared with uninfected controls (30 ng/mL, 23-50), plasma sCD30 levels were elevated in viremic participants (75 ng/mL, 53-116; p<0.001), but not in those on suppressive ART (35 ng/mL, 31-39). In contrast, CSF sCD30 levels remained elevated in ART-suppressed individuals (34 ng/mL, 19-46; p=0.002) and in those with CSF escape (33 ng/mL, 27-40; p=0.004) compared with controls (18 ng/mL, 11-23). Interestingly, individuals with very low level CSF HIV RNA (detectable but <40 copies/mL) had higher CSF sCD30 than those with higher RNA levels (quantifiable above the limit of detection) and to participants with undetectable CSF RNA (median 33 vs 24 vs 19 ng/mL, p=0.005). No association was observed between CSF sCD30 and plasma HIV RNA, concurrent or nadir CD4 T cell count, duration of infection, plasma sCD30, or CSF total protein. CSF sCD30 correlated with CSF neurofilament-light chain, a marker of axonal injury (r=0.36, p<0.001), but not with neopterin, a marker of myeloid cell activation. Conclusion: Soluble CD30 levels remain elevated in the CSF but not plasma of HIV-infected individuals on ART. In addition, CSF sCD30 is correlated with neuronal injury markers and low-level residual CNS viremia, but not with markers of myeloid cell activation or general CNS inflammation. CSF sCD30 appears to be produced in the setting of very low, but not higher, levels of CSF HIV RNA, which may reflect virus release compared to the high burst size characteristic of productive viral infection.

neopterin) were measured in participants with plasma HIV <20 cps/mL. We compared these markers between participants with CSF HIV RNA detected below the LOQ and those with CSF HIV not detected. Mann-Whitney and Fisher’s exact tests were used to determine statistical significance. Results: Forty-one participants with plasma viral suppression had mean age 54 yrs, mean CD4 T cells 612 cells/uL, and median 20 yrs on ART. 21 of 41 participants had HIV RNA detected in CSF, with 13 having CSF HIV RNA detected below the LOQ (i.e., <20 cps/mL), and 8 with quantifiable CSF HIV (median = 47 cps/mL). Participants with CSF HIV RNA detected below the LOQ did not differ from those with CSF HIV not detected when compared for age, years on ART, nadir or current CD4 T cells, race, sex, substance use history, or current use of an integrase inhibitor. When compared to participants with CSF HIV RNA not detected, participants with CSF HIV detected below the LOQ had increased CSF albumin (median 26 vs. 20 mg/dL; p<0.04), CSF protein (median 39 vs. 30 mg/ dL; p<0.05), and CSF:blood albumin ratio (median 6.5 vs. 4.5; p<0.05) (Figure 1). There were no significant differences between the two groups for CSF leukocytes (median 1-2 cells/uL) or for plasma or CSF neopterin. Conclusion: Detection of HIV in CSF below a current commercial assay’s LOQ was associated with higher CSF albumin, protein, and albumin ratio when compared to no detection of HIV in CSF, but did not associate with standard HIV metrics or treatment history. Detection at unquantifiable levels by this clinical assay may accurately differentiate low level HIV associated with changes in the blood-brain barrier, even during plasma viral suppression.

Oral Abstracts

127 EFFECT ON PLASMA NFL, A MARKER OF NEURONAL INJURY, AFTER SWITCHING FROM TDF TO TAF Linn Hermansson 1 , Richard W. Price 2 , Aylin Yilmaz 1 , Staffan Nilsson 3 , Scott McCallister 4 , Tariro Makadzange 4 , Moupali Das 4 , Henrik Zetterberg 1 , Kaj Blennow 1 , Magnus Gisslén 1 1 Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden, 2 University of California San Francisco, San Francisco, CA, USA, 3 Chalmers University of Technology, Gothenburg, Sweden, 4 Gilead Sciences, Inc, Foster City, CA, USA Background: Tenofovir alafenamide (TAF) is associated with significantly lower plasma tenofovir concentrations than tenofovir disoproxil fumarate (TDF), thereby decreasing bone and renal side effects. Limited data are available on TAF pharmacokinetics and effect in the central nervous system (CNS). One concern that has been raised is that because TAF is a stronger substrate for P-glycoprotein (P-gp) than TDF, it could decrease its CNS exposure, since substrates for P-gp are subject to active blood-brain barrier efflux. Plasma neurofilament light protein (NFL) is a sensitive marker of neuronal injury in a variety of neurodegenerative conditions, including neuronal injury in HIV

126 CSF HIV RNA DETECTED AT <20 COPIES/ML ASSOCIATES WITH BLOOD- BRAIN BARRIER MEASURES Shelli Farhadian , Jennifer Chiarella, Rachela Calvi, Hetal Mistry, Tobias Kirchwey, Payal Patel, Michelle Chintanaphol, Marie L. Landry, Serena S. Spudich Yale University, New Haven, CT, USA Background: A subset of people living with HIV (PLWH) has quantifiable HIV RNA in cerebrospinal fluid (CSF) despite plasma HIV suppression; this ‘CSF escape’ can associate with inflammation and injury in the central nervous system. However, it is unknown whether detection of CSF HIV RNA below the limit of quantitation (LOQ) on commercially available clinical assays associates with clinical parameters or biomarkers of neuropathogenesis during antiretroviral therapy (ART). Methods: PLWH on ART for >1 year consented to a research blood draw and lumbar puncture. HIV-1 RNA testing was performed on paired CSF and plasma samples using the Roche Ampliprep Taqman_v2.0 assay (lower LOQ 20 cps/ mL). Markers of blood brain barrier dysfunction and neuroinflammation (CSF albumin, protein, and WBC) and monocyte activation (plasma and CSF

47

CROI 2019