CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

is less understood, particularly among men who have sex with men (MSM). We examined the impact of age on the magnitude of racial/ethnic disparities in longitudinal HIV care indicators among MSM newly linked to HIV care. Methods: Adult MSM who successfully linked to care (i.e., ≥2 HIV care visits in ≤12 months) for the first time between 2004-14 in 11 US clinical cohorts in the NA-ACCORD were followed from HIV care linkage until 5 years after linkage, 31 December 2014, or death, whichever occurred first. We added and subtracted cumulative incidence curves for ART initiation, disengagement from care (i.e., not having ≥1 HIV care visit, CD4 count, or HIV RNA measure in ≤12 months), re-engagement in HIV care, VS (HIV RNA ≤200 c/mL), and loss of VS. We then integrated the area between the curves to estimate the mean percentage of person-time spent 1) in care, 2) on ART, and 3) with VS in the first 5 years after linkage, by race/ethnicity and age group. Analyses were adjusted for age (to reduce confounding within age groups), history of injection drug use, site, CD4 count, and HIV RNA at linkage. Results: A total of 11,003 MSM were included. MSM of most racial/ethnic and age groups spent on average >70% of person-time engaged in care and >50% of person-time on ART during the first 5 years after linkage to care (Table). Black MSM≥40 years of age and Hispanic MSM≥50 years of age spent less time in care, on ART, and with VS than white MSM in the same age range. Hispanic MSM <50 years of age spent more time in these stages than white MSM in the same age range, although most differences were not statistically significant. The magnitude of the black-white and Hispanic-white disparity generally increased with increasing age for these outcomes. Conclusion: The magnitude of racial/ethnic disparities in the HIV care continuum after inital HIV care linkage increases with increasing age. Clinical initiatives designed to reduce racial/ethnic disparities in engagement in care, early ART initiation, and VS among MSM with HIV should especially focus on minority MSM from older age groups.

log-binomial models with each outcome to estimate prevalence ratios (PRs) and 95% confidence intervals (CIs) according to key demographic variables. Results: Overall, 13,628 persons were eligible for analysis. Most were male (75.6%), men who report sex with men (MSM) (62.6% among males), African American (61.8%), and had ≥1 VL reported during 2016 (76.6%). The median age at diagnosis was 35.4 years (IQR 26.0-45.8). A total of 8859 (65.1%) PLWHA exhibited SVS and 6,004 (44.1%) exhibited DVS during 2016. Differences in both SVS and DVS were observed by age (>35 vs ≤35 years: SVS PR 1.23, 95% CI 1.20-1.26 and DVS PR 1.41, 95% CI 1.35-1.46) and race/ethnicity (African American vs White: SVS PR 0.88, 95% CI 0.85-0.90 and DVS PR 0.83, 95% CI 0.79-0.86; Hispanic vs White: SVS PR 0.83, 95% CI 0.79-0.88 and DVS PR 0.89, 95% CI 0.83-0.96). Conclusion: Nearly two-thirds of all PLWHA in NC achieved SVS in 2016, but fewer than half achieved DVS. Differences by age and race were apparent with both measures, suggesting that either approach may be useful in identifying disparities to be addressed; however, the SVS measure, which does not reflect the durable suppression needed for optimal clinical and prevention benefits, may provide an overly optimistic view of viral suppression in the population.

Poster Abstracts

941 UNDISCLOSED ANTIRETROVIRAL DRUG USE IN BOTSWANA – IMPLICATIONS FOR NATIONAL ESTIMATES

Sikhulile Moyo 1 , Simani Gaseitsiwe 1 , Kathleen M. Powis 2 , Terence Mohammed 1 , Comfort Maphorisa 1 , William Abrams 3 , Liziwe Chebani 4 , Kutlo Manyake 1 , Molly Pretorius Holme 5 , Tendani Gaolathe 1 , Joseph Makhema 1 , Shahin Lockman 5 , William Clarke 6 , Max Essex 5 , Vlad Novitsky 5 1 Botswana Harvard AIDS Institute Partnership, Gabarone, Botswana, 2 Brigham and Women’s Hospital, Boston, MA, USA, 3 CDC Botswana, Gaborone, Botswana, 4 Ministry of Health, Gaborone, Botswana, 5 Harvard University, Cambridge, MA, USA, 6 Johns Hopkins Hospital, Baltimore, MD, USA Background: Undisclosed ART use may affect national estimates and confound results of clinical trials. We assessed ARV traces among virologically suppressed individuals participating in the Botswana Combination Prevention Project (BCPP) who self-reported no prior ARV use. Methods: Plasma from 134 BCPP participants who reported no ART use and had undetectable HIV-1 RNA (£400 copies/mL) was screened for ARVs by high- throughput liquid chromatography coupled with Q-Exactive high-resolution mass spectrometry using data-dependent fragmentation and selected reaction monitoring at resolution of 17,500 (Marzinke et al. 2014 Clinica Chimica Acta 433:157-68). To obtain qualitative results, each specimen was compared to positive and negative controls for each drug (Abacavir, Amprenavir, Atazanavir, Darunavir, Efavirenz, Emtricitabine, Indinavir, Lamuvidine, Lopinavir, Maraviroc, Nelfinavir, Nevirapine, Raltegravir, Rilpivirine, Ritonavir, Saquinavir, Stavudine, Tenofovir, Tipranavir, and Zidovudine). Results: Among 3,596 HIV-positive participants enrolled in a household survey in BCPP, 953 (27%; 95% CI 25-28%) self-reported no prior use of ART, 135 (14%, 95% CI 12-17%) of whom had HIV-1 RNA£400 copies/mL. Plasma ARV traces were tested in 134 of these 135 individuals, 52 (39%, 95% CI 31-48%) of whom had detectable ARVs. Traces of 3 ARV drugs were found in 42 participants, 2 drugs in 9 participants, and one participant had traces to a single drug (EFV). The most commonly identified ARVs (EFV/NVP, FTC, TDF) represented regimens in Botswana’s national ART program. The overall proportion of HIV-infected

940 HIV VIRAL SUPPRESSION ASSESSED BY TWO METRICS IN NORTH CAROLINA DURING 2016 Manuela Bullo 1 , Erika Samoff 2 , Nicole Dzialowy Adams 2 , Brad Wheeler 2 , Kimberly A. Powers 1 , Anna Cope 3 1 University of North Carolina Chapel Hill, Chapel Hill, NC, USA, 2 North Carolina Division of Public Health, Raleigh, NC, USA, 3 CDC, Atlanta, GA, USA Background: To maximize HIV clinical outcomes and reduce onward transmission, durable viral suppression (DVS) is needed among people living with HIV/AIDS (PLWHA). However, most viral suppression analyses are based on measures that consider only a single viral load (VL) and thus cannot indicate durability. We sought to compare viral suppression prevalence using a single viral suppression (SVS) and DVS measure, and to assess disparities in both measures by age and race, among PLWHA in North Carolina (NC) during 2016. Methods: The NC Division of Public Health (DPH) maintains a registry of all PLWHA in an electronic surveillance system. Since 2013, all HIV care labs (VL and CD4 counts) are reportable to the NC DPH by law. We considered all PLWHA who were diagnosed between 2006 and 2015, were alive and resided in NC in 2016, and were ≥18 years old at diagnosis. We measured: a) prevalence of SVS, defined as VL ≤200 copies/ml at most recent measure in 2016; and b) DVS, defined as at least two separate VLs ≤200 copies/ml and with no VL >200 copies/ml during 2016. Eligible PLWHA with no VL recorded in 2016 were included and assigned to not having either outcome. We used univariable

CROI 2018 359