CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

Methods: We analyzed paired peripheral blood mononuclear cells (PBMC) and CSF cell pellets from 29 chronically HIV infected subjects with or without ART. Genomic DNA was extracted from PBMC using silica-based columns, while CSF cell pellets were directly lysed. Levels of HIV DNA (pol gene) were measured by digital droplet PCR (ddPCR) and normalized to cell numbers measured with RPP30. Non-parametric tests of associations between levels of HIV DNA and RNA in CSF and blood were carried out. Results: We investigated 20 subjects with undetectable HIV RNA (<50 copies/ml in blood plasma and CSF) and 9 subjects with detectable HIV RNA (median of 4.22 log 10 HIV RNA in blood and 4.11 log 10 HIV RNA in CSF). The median CD4 count was 498 (interquartile range [IQR]: 362-637). Patients on ART were mostly on regimens including a protease-inhibitor (86%) had a median time of ART exposure of 6.2 years (IQR: 4.6-11.6) and a median CNS penetration effectiveness (CPE) value of 7 (IQR: 7-7.8). HIV DNA was detected in 19 (66%) CSF pellets, including 10 (50%) samples in which HIV RNA was undetectable in CSF. HIV DNA levels in CSF cells positively correlated with HIV DNA levels in PBMC (P=0.03) and with HIV RNA in CSF (P=0.05) but not with the number of CSF leukocytes. Levels of HIV DNA in PBMC positively correlated with HIV RNA levels in blood (p=0.001). Similarly, the levels of HIV RNA significantly correlated between both compartments (P<0.0001). Interestingly, while levels of HIV DNA in blood were significantly lower in subjects on suppressive ART compared to untreated participants (P=0.01), HIV DNA levels in CSF did not differ between treated and untreated participants. Conclusions: Levels of HIV DNA and RNA correlated within and between blood and CSF compartments. However, suppressive ART was associated with lower HIV DNA levels in blood but not in CSF cells, despite relatively high CPE values. The HIV DNA reservoir in the CNS may not be effectively targeted with highly potent ART. 436 Antiretroviral Concentrations in Brain Tissue Are Similar to or Exceed Those in CSF. Namandjé Bumpus 2 ; Qing Ma 4 ; David J. Moore 1 ; Brookie M. Best 1 ; Ronald J. Ellis 1 ; Cristian L. Achim 1 ; Melanie Crescini 1 ; Courtney Fletcher 3 ; Igor Grant 1 ; Scott Letendre 1 The CNTN Group 1 University of California San Diego, San Diego, CA, US; 2 Johns Hopkins University School of Medicine, Baltimore, MD, US; 3 University of Nebraska Medical Center, Omaha, NE, US; 4 University at Buffalo, Buffalo, NY, US Background: Limited distribution of many antiretroviral therapy (ART) drugs into the central nervous system (CNS) results in much lower drug concentrations in cerebrospinal fluid (CSF) than in blood. Estimates of CSF distribution have been linked to HIV RNA levels in CSF and neurocognitive performance but results can be inconsistent. One possible reason for this inconsistency is that ART drug concentrations in CSF may not accurately reflect those in brain tissue. The objective of this analysis was to measure ART drug concentrations in brain tissue collected from adults dying with HIV disease. Methods: 9 HIV+ adults were evaluated in the California NeuroAIDS Tissue Consortium (CNTN) within 6 months of death; reported taking ART at that antemortem visit; and had detectable concentrations of at least one ART drug in serum at autopsy. Autopsies were performed within 30 hours of death. Brain tissue was collected and stored at -80 o C. Concentrations of 6 ART drugs (see Table) were measured in 3 brain tissue regions, globus pallidus (GP), cortical gray matter (CGM), and white matter (WM), by LC/MS with a lower limit of quantitation of 25 ng/mL. Results: Subjects were mostly men (82%) with a mean age of 40.4 (SD 5.0). The most common cause of death was pneumonia. ART drug concentrations in brain tissue in ng/mL are summarized in the Table. Concentrations of ATV, EFV, FTC, and 3TC were similar to published concentrations of these drugs in CSF but concentrations of TDF were higher than reported values in CSF. LPV concentrations in brain tissue were also higher than reported in CSF but only in WM. Drug concentrations appeared to vary by brain region: Across all drugs, concentrations were lower in CGM than in the other two regions (p=0.01, paired signed rank test). Table. Summary of Antiretroviral Concentrations in Brain Tissue. Conclusions: This is the first analysis of ART drug concentrations in human brain tissue. Concentrations of most drugs in this small analysis were similar to reported concentrations in CSF but TDF had higher concentrations than expected based on CSF reports. Regional variation in ART drug concentrations may be important for antiviral efficacy and toxicity. 437 HIV DNA Peripheral Reservoirs Have a Nonlinear Impact on Brain Pathology William Hey-Cunningham 2 ; Nadene Dermody 3 ; Phillip Chan 4 ; Bruce Brew 5 ; Kersten Koelsch 2 ; Lucette A. Cysique 1 1 NeuRA, University of New South Wales, Randwick, Australia; 2 Kirby Institute, University of New South Wales, Sydney, Australia; 3 Macquarie University, Sydney, Australia; 4 Queen Elizabeth Hospital, HKSAR, Hong Kong, China; 5 St. Vincent’s Hospital, University of New South Wales, Sydney, Australia Background: The HIV reservoir in peripheral blood mononuclear cells (PBMC) contributes to HIV-associated dementia (HAD) pathogenesis. However it is unclear whether it contributes to non-demented HIV-associated neurocognitive disorders (HAND) that are more common in chronic HIV infection (CHI), especially in HIV+ persons on long-term combined antiretroviral therapy (cART). Methods: Eighty adults (mean age=55, 1 female) with CHI on cART (>97%with undetectable plasma and CSF HIV RNA) underwent assessments of neurocognition and pre- morbid cognitive ability (Australian standardized index of education and reading) at two visits 18 months apart. HIV DNA in PBMC was measured by real-time PCR at the same time-points. Historical HIV disease data included: duration of HIV infection (median=19 years); ART during 1 st year of HIV infection (20%); current (556) and nadir (194) median CD4 count; none developed AIDS during study. Results: At baseline, 46% of the patients had non-demented HAND; 7.5% had HAD. At follow-up, neurocognitive function declined in 14%, and this was more likely in those with baseline HAD ( p <.03). Low pre-morbid cognitive ability was uniquely associated with HAD ( p <.05). Log(n) HIV DNA copies were stable between study visits (5.2 vs. 5.1 per 10 6 PBMC; r=.73). Univariate analyses showed that baseline HIV DNA was higher in those with longer duration of infection (r=.28; p<.02, Fig.1B) as well as those with lower pre-morbid cognitive ability (r=-.24; p<.04, Fig.1A) and was higher in those with no treatment during HIV infection 1 st year (5.4 vs. 4.5; p=.03, Fig.1C). Multiple regression models showed that baseline HIV DNA was not associated with baseline overall neurocognitive performance in unadjusted and adjusted (above significant factors plus CD4 count, age, cART duration) models. Change in HIV DNA between study time-points was associated with a decline in motor-coordination, r=-26, p <.03 (Fig.1D); and semantic fluency, r=-25, p <.03; and this remained significant in adjusted models (p<.02).

Poster Abstracts

304

CROI 2015