CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

1116 Acute HIV Infection Transmission Among PeopleWho Inject Drugs in an Established Epidemic Setting Daniel Escudero 4 ; CalebWeinreb 4 ; Mark Lurie 4 ; Kenneth Mayer 2 ; Sandro Galea 3 ; Samuel Friedman 5 ; Brandon Marshall 4 1 Brown University, Providence, RI, US; 2 Fenway Health, Boston, MA, US; 3 Columbia University, New York, NY, US; 4 Brown University, Providence, RI, US; 5 National Development and Research Institutes, Inc, New York, NY, US Background: Among people who inject drugs (PWID), little is known about the contribution of acute HIV infection (AHI) to incident infections. Understanding its role in overall transmission may be crucial to improving the effectiveness of prevention strategies. We constructed an agent-based model (ABM) to estimate the proportion of transmission events attributable to AHI within an established HIV epidemic among PWID. Methods: The ABM was previously calibrated to reproduce the sociodemographics, risk behavior, and HIV epidemic trajectory observed in the New York metropolitan statistical area (MSA) population. Agents interact in a mixed, and dynamic sexual and injecting transmission network, representing a 100,000 population. Each agent has a unique, time- updated probability of acquiring or transmitting HIV determined by their risk behavior, partnerships, engagement in simulated prevention interventions (i.e., needle and syringe programs, HAART), and HIV disease stage. Using stochastic microsimulations, we catalogued transmission events based on the disease stage of the index agent to determine the proportion of infections transmitted during AHI (defined as the three month period following infection). Results: The calibrated model was able to approximate the epidemic trajectory among PWID in the New York MSA observed between 1992 and 2012. PWID comprised 1.9% of the general population in 1992, which decreased to 1.4% by 2012. Average annual incidence over this period was 0.07% for the general population; among PWID, incidence peaked at 3.5% during 1993-94 with a low of 1.7% in 2006. By 2012, 50% of HIV-infected PWID had initiated HAART, of whom 60%were virologically suppressed. Over the entire period, AHI accounted for 19% of incident HIV cases among PWID, with the following period-specific estimates: 15% (1992-1996, pre-HAART), 16% (1997-2004), and 23% (2005-2012). Conclusions: This study is the first to produce an estimate for the proportion of incident HIV infections attributable to AHI among PWID. Our model (which accounted for sexual and parenteral transmission, heterogeneous risk behavior, assortative mixing, and the expansion of HIV treatment and prevention interventions over the last two decades) produced AHI transmission estimates at the lower end of those previously published for non-drug-using populations, which may be due to our modeling of an established, declining epidemic. Further research and sensitivity analyses are needed to confirm these preliminary results. 1117 Decreasing Number of Undiagnosed HIV Infections in the Netherlands Ard van Sighem 1 ; Fumiyo Nakagawa 3 ; Daniela De Angelis 4 ; Chantal Quinten 5 ; Daniela Bezemer 1 ; Eline Op de Coul 2 ; Matthias Egger 7 ; Frank deWolf 6 ; Christophe Fraser 6 ; Andrew N. Phillips 3 1 Stichting HIV Monitoring, Amsterdam, Netherlands; 2 National Institute for Public Health and the Environment, Bilthoven, Netherlands; 3 University College London, London, United Kingdom; 4 MRC Biostatistics Unit, Cambridge, United Kingdom; 5 Europen Centre for Disease Prevention and Control, Stockholm, Sweden; 6 Imperial College London, London, United Kingdom; 7 Institute for Social and Preventive Medicine, University of Bern, Bern, Switzerland Background: Accurate estimates of the size of the HIV-infected population, including those not yet diagnosed, are important to understand the HIV epidemic and to plan interventions. We sought to estimate the number living with HIV as well as trends in the undiagnosed population, HIV incidence, and rate of diagnosis in the past 10 years. Methods: We used a multi-state back-calculation model to estimate HIV incidence, time between infection and diagnosis, and the HIV-infected population by CD4 count strata. The model was fitted to national surveillance data on new HIV and AIDS diagnoses from the ATHENA observational HIV cohort in the Netherlands. Rates of progression between the different states (primary infection, CD4 cell count ≥ 500, 350-499, 200-349, or <200 cells/mm 3 , and AIDS) were based on historical cohort data on untreated HIV-infected patients. Bootstrap techniques were used to calculate 95% confidence intervals (CI). Results: By the end of 2013, 29200 (95% CI 28000-30400) individuals, of whom 23400 (22200-24600) were still alive, were estimated to have been infected with HIV since the start of the epidemic in the 1980s. Based on registered HIV cases in ATHENA, we estimated that 91% of these patients, approximately 21300 (20200-22400), were still living in the Netherlands; the remaining 9%were not in care anymore because they moved abroad or were lost to follow-up. According to the model, the number of undiagnosed HIV-infected individuals decreased from 5150 (4850-5500) in 2003 to 3400 (2500-4650) in 2013. Of the undiagnosed individuals in 2013, 23% (19-27) were estimated to have been infected for less than one year, 53% (49-56) for one to 5 years, and 24% (19-30) for more than 5 years; 53% (51-55) had CD4 counts <500 cells/mm 3 . The annual number of new infections remained almost unchanged: 1060 (940-1200) in 2003, 1020 (890-1151) in 2008, and 860 (590-1260) in 2013. At the time of diagnosis, the estimated proportion of patients infected less than 2 years before their HIV diagnosis increased from 21% (18-23) in 2003 to 26% (22-31) in 2013, while the proportion infected less than 5 years before increased from 60% (57-63) to 67% (62-73). Conclusions: The number of undiagnosed HIV infections in the Netherlands is decreasing, but still almost a quarter has been infected for more than five years. Much greater increases in diagnosis rates are likely to be needed for a more substantial decrease in the annual number of new infections. 2:30 pm– 4:00 pm Modeling the Impact of HIV Interventions 1118 Predicted Impact of Antiretroviral Treatment on Preventing New HIV Infections in 53 Low- and Middle-Income Countries With Large HIV Epidemics AndrewM. Hill 1 ; Anton Pozniak 1 ; Katherine Heath 2 ; Alice Raymond 2 ; Mary Mahy 3 ; Nathan Ford 4 1 Chelsea and Westminster Hospital, London, United Kingdom; 2 St Mary’s Hospital–Imperial College Healthcare NHS Trust, London, United Kingdom; 3 Joint United Nations Programme on HIV/AIDS (UNAIDS), Geneva, Switzerland; 4 World Health Organization, Geneva, Switzerland Background: Clinical trials and observational studies have demonstrated that antiretroviral treatment can reduce the risk of HIV transmission. However, many countries still have low rates of antiretroviral treatment uptake. Methods: Standardised epidemiological data were compiled from 36 African and 17 non-African low- and middle-income countries with least 40,000 HIV-infected people. Estimates of new HIV infections in 2013 by country were calculated using the AIDS Impact Module (AIM) in Spectrum. Each country entered HIV prevalence rates from pregnant women attending antenatal care or from key risk groups, the numbers receiving antiretroviral treatment, the numbers of pregnant women taking antiretroviral therapy (ART) to prevent vertical transmission, and the national antiretroviral eligibility criteria into the AIM. ART coverage rate was defined as the total number receiving ART divided by the epidemic size in each country. HIV transmission rate was defined as the number of new infections per year divided by the epidemic size in each country. Weighted least squares (WLS) regression was used to investigate the association between HIV transmission rates and antiretroviral treatment coverage across the 53 countries. THURSDAY, FEBRUARY 26, 2015 Session P-Z3 Poster Session Poster Hall

Poster Abstracts

648

CROI 2015