CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

1005 Alcohol and Substance Use and Timing of Presentation to HIV Care Across the United States Jesse Abbott Klafter 1 ; Daniel R. Drozd 3 ; Michael J. Mugavero 2 ; Katerina Christopoulos 4 ; ChristopherW. Mathews 7 ; Joseph J. Eron 8 ; Kenneth H. Mayer 6 ; Matthew Mimiaga 5 ; Mari Kitahata 3 ; Heidi M. Crane 3 Center for AIDS Research Network of Integrated Clinical Systems (CNICS) 1 University of Washington, Seattle, WA, US; 2 University of Alabama at Birmingham, Birmingham, AL, US; 3 University of Washington, Seattle, WA, US; 4 University of California San Francisco, San Francisco, CA, US; 5 Harvard Medical School, Boston, MA, US; 6 Fenway Health, Boston, MA, US; 7 University of California San Diego (UCSD), San Diego, CA, US; 8 University of North Carolina, Chapel Hill, NC, US Background: Alcohol and substance use are known to negatively impact multiple steps along the HIV care continuum including retention, adherence, and viral suppression. The association between alcohol and substance use and testing and linkage to care is less well characterized. CD4 count at care entry is a function of HIV testing timing and subsequent linkage to care. We examined the association between alcohol and substance use and CD4 count at HIV care entry. Methods: Data from 6 CNICS sites from 2006-2013 were included. Patients complete a touch-screen-based assessment including alcohol and substance use measures as part of clinical visits. Patients were eligible for this study if they had no prior HIV care, were antiretroviral therapy naïve at the time of care entry, and had an assessment within 24 months of care entry (sensitivity analyses included 12 and 18 months). Our main outcomes were CD4 count as continuous and binary (<200) measures at care entry. We used linear and logistic regression to examine the association between these outcomes and alcohol and substance, adjusted for age, sex, race/ethnicity, depression, year of care entry, and site. Results: 2025 patients were eligible. Any current drug use (a composite of current amphetamine, cocaine or opiate use) (20%) was associated with higher CD4 count at care entry (34 cells/mm 3 higher p<0.05) and lower likelihood of CD4<200 (OR 0.5 p<0.01) compared to those who never used drugs. In separate models, current intravenous (IV) drug use (5%) was associated with higher CD4 count (86 cells/mm 3 higher p<0.01) and lower likelihood of initial CD4<200 (OR = 0.4 p<0.05) compared to those who never used IV drugs. In a third set of models including separate amphetamine, cocaine and opiate use, current cocaine use was associated with lower likelihood of CD4<200 (OR 0.4 p<0.01). High risk alcohol use was not associated with CD4 count at care entry in any of our models, compared to non-drinkers. Findings were similar in sensitivity analyses. Conclusions: Current substance use and IV drug use were associated with more timely presentation to HIV care as measured by initial CD4 count. This may reflect increased availability of HIV testing and linkage services among these patients or more frequent interactions with the health and/or criminal justice systems. These results have implications both for a universal test and treat strategies as well as for efforts to improve outcomes of HIV care among substance users. 1006 Marijuana Use and Its Nuanced RelationshipWith HIV Treatment ContinuumMetrics John A. Schneider 1 ; Ethan Morgan 1 ; Stuart Michaels 2 ; Britt Skaathun 1 ; LindsayYoung 1 ; RobertW. Coombs 3 ; Phil Schumm 1 ; DexterVoisin 1 ; Sam Friedman 4 UConnect StudyTeam 1 University of Chicago, Chicago, IL, US; 2 NORC, Chicago, IL, US; 3 University of Washington, Seattle, WA, US; 4 National Development Research Institute, New York, NY, US Background: Studies of relationships between drug use and HIV treatment have primarily focused on methamphetamine, cocaine or heroin use. Few studies, however, have focused on younger Black men who have sex with men (YBMSM) who have less drug use in aggregate when compared to other MSM, but likely have higher rates of marijuana use. We examine associations between marijuana use and key treatment continuummetrics in a population based sample of YBMSM in Chicago. Improving treatment continuum outcomes for YBMSM is critical to controlling the HIV epidemic domestically. Methods: From 2013-2014 a representative sample of YBMSM 16-29 years old in Chicago (n=626) was generated using Respondent Driven Sampling (RDS). HIV antibody/Ag and RNA testing were performed from dried blood spots. RDS-weighted models examined associations between marijuana use (never, intermittent or daily), HIV testing, and downstream treatment continuummetrics. Models were adjusted for age, education, condomless sex, group sex, EtOH use, depression and other drug use. Results: YBMSM had a 28% seropositivity rate; 31% of positives were virally suppressed. 32% of YBMSM reported using marijuana daily or multiple times daily, 27% never used and 41% reported intermittent use (weekly or less). MJ use was mildly correlated with ecstasy use (r=0.15; p<0.001) and popper use (r= 0.11; p=0.008), but not methamphetamine use (r=0.07; p=0.09)). In adjusted regression models, YBMSMwho used marijuana were more likely to be HIV seropositive (aOR, 3.56; p<0.05) and HIV positive unaware (12.80, p<0.001). Among HIV seropositive individuals, compared to no use, intermittent but not daily marijuana use was associated with worse retention in HIV care (2 or more visits 3 months apart in previous year) (aOR, 6.10; p=0.021). Marijuana use was not associated with linkage to care, adherence to ARVs or viral suppression. Covariates in these models including alcohol and other drug use were also not associated with any of the continuummetrics. Conclusions: Critical HIV treatment continuum components such as knowing one’s status and retention in care are related to intermittent marijuana use. Specific marijuana use information should be collected from clients engaging in care that includes frequency of use which may help target HIV treatment interventions. A focus on drugs used by most affected populations such as YBMSM and their nuanced relationship with continuummetrics is warranted, particularly in the context of increasing social acceptability of marijuana. 1007 “Test-and-Treat”in the Netherlands Ard van Sighem 1 ; Luuk Gras 1 ; Eline Op de Coul 2 ; Daniela Bezemer 1 ; Michiel van Agtmael 3 ; Godelieve de Bree 4 ; Peter Reiss 1 On behalf of the ATHENA National Observational HIV Cohort 1 Stichting HIV Monitoring, Amsterdam, Netherlands; 2 National Institute for Public Health and the Environment, Bilthoven, Netherlands; 3 VU University Medical Centre, Amsterdam, Netherlands; 4 Amsterdam Institute for Global Health and Development, Amsterdam, Netherlands Background: Early diagnosis and treatment of HIV, or ‘test-and-treat’, benefits individual patients and helps in preventing new infections. Treatment guidelines in the Netherlands now recommend starting combination antiretroviral treatment (cART) immediately, regardless of CD4 cell counts. We studied changes over time in the proportion of patients diagnosed with a recent infection and to what extent immediate treatment is adopted in clinical practice. Methods: All HIV-1-infected patients diagnosed in 2006-2013, i.e., when information on recent infection was consistently collected for all patients, were selected from the ATHENA national observational HIV cohort. Patients were considered recently infected if in the 6 months before diagnosis they had a HIV-negative test, an indeterminate western blot, additional evidence of a known risk exposure, or symptoms of acute infection. CD4 count at diagnosis was the first pre-cART CD4 count ≤ 3 months after diagnosis. Results: Of 9057 diagnosed patients, 1756 (19%) had evidence of recent infection: 1533 (87%) men who have sex with men (MSM), 124 (7%) men with other transmission route, and 99 (7%) women. Between 2006 and 2013, the proportion of recent infections increased from 21% to 31% in MSM (p<0.001), from 5% to 7% in women (p=0.008), and did not change in other men (7%; p=0.4). Recent infection was based on a previous negative test for 1188 (68%) patients, of whom 42% also had symptoms and 19% a known risk exposure. Overall, 884 (50%) patients had symptoms of acute infection. Median CD4 count was 500 (interquartile range, 364-680) cells/mm 3 and did not differ by risk group. Of all 9057 patients, 86% had entered into care ≤ 6 weeks after diagnosis. Between 2006 and 2013, the proportion starting cART ≤ 6 months after diagnosis was >95% for those diagnosed with CD4 <200, increased from 52% to 87% for CD4 200-349, from 15% to 63% for CD4 350-499, and from 8% to 42% for CD4 ≥ 500 cells/mm 3 (see figure). The proportion on cART among patients diagnosed in 2013 did not differ by risk group. In patients with recent infection, the proportion on cART ≤ 6 months increased from 21% to 62%.

Poster Abstracts

592

CROI 2015