CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

blood and CSF laboratories at each visit. VLs less than 40 were estimated as 20 if undetectable and as the exact copy number when detectable. Decay rates were calculated as change in log(VL) divided by number of days on ART. Multivariate linear regression was used to identify predictors of CSF VL decay rates. Results: 40 treatment-naïve patients (35 men, 5 women) underwent baseline LP and 30 returned for LP 2 weeks after initiation of ART. Median CD4 count was 259 (range 6-424). Median blood HIV VL was 42,269 copies/mm 3 (168-2,184,166) at baseline and declined to 205 (<40-26,417) at week 2. At baseline, 98% of subjects had detectable CSF HIV VL (median 1197, range 11-381,030); at 2 weeks 77% had detectable CSF HIV, and mean log(CSF VL) declined by 1.24. In bivariate analyses, log(CSF VL) at baseline and 2 weeks were correlated (p=0.0002) and log(CSF VL) and log(blood VL) at 2 weeks were correlated (p=0.003). Log-linear CSF VL decay rates were correlated with baseline CSF VL (p=0.0001) and with baseline CSF white blood cell count (p=0.03); there was a non-significant trend towards correlation with blood VL at baseline (p=0.1). Protease inhibitor use was associated with faster CSF viral decay (p=0.04) while integrase inhibitor use was associated with non-significantly slower decay, though this relationship did not hold for blood VL. Multivariate analyses (N=28) of predictors of CSF VL decay are shown in the table.

Conclusions: Baseline higher CSF VLs and faster blood VL decay rates predicted faster decay of HIV in CSF after initiation of ART. Though not significant in multivariate analyses, PIs were associated with faster CSF viral suppression while integrase inhibitors were associated with slower suppression, findings which may have implications for ART choice in patients with HIV-associated neurocognitive findings.

446 Rates of Nonconfounded HIV-Associated Neurocognitive Disorder After Early cART Teresa H. Evering 1 ; Allison Applebaum 2 ; Melissa La Mar 1 ; Donald Garmon 1 ; David Dorfman 3 ; Martin Markowitz 1

1 Aaron Diamond AIDS Research Center, an Affiliate of the Rockefeller University, New York, NY, US; 2 Memorial Sloan-Kettering Cancer Center, New York, NY, US; 3 Mount Sinai School of Medicine, New York, NY, US Background: HIV-Associated Neurocognitive Disorder (HAND) is an important complication of chronic HIV-infection with an estimated prevalence ranging from 19 - 45%. We investigated the prevalence of non-confounded HAND in an HIV positive cohort who previously initiated long-term cART with a median duration of infection of 1.6 months. Methods: Participants were randomly selected from the ADARC Primary HIV-1 Infection Cohort. Exclusion criteria included: treatment interruption since cART initiation, severe neuropsychiatric disorder, active major depressive disorder, DSM-IV diagnostic criteria for alcohol/substance abuse or dependence within 1 year prior to screen (excluding marijuana), untreated syphilis and positive hepatitis C serology. The Beck Depression Inventory (BDI) was used to determine severity of depressive symptoms. We assessed neurocognitive (NC) function comprehensively and those with global deficit scores (GDS) >0.50 were considered impaired. Associations between GDS scores and clinical parameters were evaluated using multiple linear regression. Results: Forty individuals screened and a high screen failure (SF) rate (14/40 = 35%) was observed. The most common reasons for SF were active methamphetamine/other substance dependence (5/14 = 36%) and active bipolar disorder (3/14 = 21%). Twenty-six, primarily non-Hispanic white (73%), male (100%) subjects were enrolled and underwent NC assessment. Mean age was 43 (28-71) years, with a median of 17 years of education (13-24). Median current and nadir plasma CD4+ T cell counts were 828 (506- 1411) and 359 (150-621) cells/ m L. Median duration of cART prior to enrollment was 5.7 years (2.2-9.9). Median 2010 CPE score at study entry was 7 (6-10) and all participants had plasma HIV-1 RNA <50 copies/ml at this time. Median BDI score was 1 (0-13). Median GDS was 0.17 (0.00-0.50). Only 1 (4%) participant was impaired. There was no association between GDS and any clinical/immunologic parameter ( p > 0.27, all variables). Individuals failing screening were demographically similar to those enrolled. Conclusions: Observed rates of HAND in this cohort of HIV-infected individuals without confounding comorbidities that initiated cART during acute/early infection are low. Our findings suggest substantial neuroprotective benefit of initiating cART during primary infection, but also highlight the pervasiveness of comorbid illness in those with HIV and suggest a significant contribution of comorbidities to observed HAND prevalence. Background: Highly active antiretroviral therapy (HAART) reduces morbidity and mortality due to human immunodeficiency virus (HIV), however cognitive impairment continues to persist. It remains uncertain whether HAART regimens with a high degree of central nervous system penetration effectiveness (CPE) exert beneficial neurological outcomes in HIV-infected (HIV+) individuals on stable treatment. The present cross sectional study examined the relationship between HAART CPE and brain integrity in HIV+ adults. Methods: Sixty-four HIV-infected individuals (18-82 years old) on stable HAART ( ≥ 3 months) were assigned a CPE score using a published ranking system and divided into high ( ≥ 7; n=35) and low (< 7; n=29) CPE groups. All participants completed laboratory and neuropsychological testing in addition to structural neuroimaging. Neuropsychological tests included measures known to be sensitive to HIV with values converted into standardized scores (NPZ-4) based on published normative scores. A semi-automated methodology was utilized to assess brain volumetrics within cortical (white and grey matter) and subcortical (caudate, putamen, and thalamus) regions of interest often affected by HIV. Primary analyses utilizing linear regression were computed to examine the relationship between total CPE score with NPZ-4 and brain volumetrics. Differences in demographic (age, education, race, gender) and viral factors (recent viral load, duration of infection, recent CD4, CD4 nadir) between high ( ≥ 7) and low (< 7) CPE groups were examined utilizing independent sample t tests and chi-squared analyses. Secondary analyses utilized analysis of variance to assess NPZ-4 and brain volumetric differences between HIV+ individuals with high and low CPE scores. Results: Primary analyses revealed no significant relationships between neuropsychological performance (F(1,63)=1.01; p > .05) or brain volumetrics (F(5,67)= .27, p > .05) and total CPE score utilizing linear regression. There were no differences in demographic or viral factors between the two CPE groups. Furthermore, no significant relationships were observed in the secondary analyses when individuals were categorically divided into high or low CPE groups ( p > .05). Conclusions: Long-term HAART regimens with a high degree of CPE were not associated with significantly improved neuropsychological or neuroimaging outcomes in HIV+ adults. Results suggest that alternate mechanisms may potentially contribute to better neurological outcomes in the era of HAART. 447 The Impact of HAART CNS Penetration Effectiveness on Brain Integrity in HIV+ Adults Laurie Baker 1 ; Robert H. Paul 1 ; Jodi M. Heaps 1 ; Mario Ortega 2 ; Christin Usher 1 ; JeeYoon Chang 2 ; Beau Ances 2 1 University of Missouri St Louis, St Louis, MO, US; 2 Washington University School of Medicine, St Louis, MO, US

Poster Abstracts

310

CROI 2015