CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

age-standardized mortality ratio was 2.74 (95%CI: 2.54-2.93) and among men and 3.15 (95%CI: 2.89-3.46) among women. During years 2-3, 4-5, 6-7 and 8-9 after ART start, age-standardized mortality ratio was 2.20 (95%CI: 2.02-2.39), 1.90 (95%CI: 1.72-2.10), 2.66 (95%CI: 2.42-2.92), and 4.52 (95%CI: 4.12-4.95). Conclusion: Even after starting treatment, HIV-infected persons remain at elevated risk of death compared to a HIV-uninfected population of the same age in Zambia. Even though HIV infected men experience a higher rate of death than HIV infected women after starting ART, treated HIV women experience a greater risk of death compared to age-standardized uninfected Zambian women. Enhanced engagement and widespread use of TB prophylaxis are needed for HIV infected persons to reach mortality rates of HIV uninfected persons. 902 PREDICTORS OF LOSS OF VIRAL LOAD SUPPRESSION AMONG MSM IN ATLANTA Jeb Jones , Jennifer Taussig, Jodie L. Guest, Colleen F. Kelley, Patrick S. Sullivan Emory University, Atlanta, GA, USA Background: Inequalities in the HIV care continuum between Black and White MSM living with HIV, including maintenance of viral suppression with effective ART, contribute to disparities in morbidity, mortality, and HIV transmission rates between these groups. We conducted an interim analysis to assess predictors of incident loss of VL suppression among MSM in Atlanta, GA to gain a better understanding of individual factors that contribute to maintenance and loss of viral suppression and to inform effective interventions to reduce these disparities. Methods: The EngageMENt study is an ongoing longitudinal cohort of HIV positive Black (n=207) and White (n=193) MSM in Atlanta, GA designed to examine racial disparities in the HIV care continuum. VL measurements were obtained at 0 and 12 months. Additional VL measures were available by self-report at 3 and 6 months. Among men who were virally suppressed at baseline, we compared the rate of loss of viral suppression (VL>40 copies/mL) between Black and White MSM. Potential predictors of incident loss of VL were measured at baseline and at each follow-up visit. Unadjusted and adjusted Cox proportional hazards models were used to assess predictors of incident loss of VL suppression. Results: The rate of loss of viral suppression was 20.2/100 person-years (95%CI: 13.2, 29.6) among Black MSM and 11.9/100 PY (95%CI: 7.0, 18.9) among White MSM [unadjusted hazard ratio (HR) = 1.7, 95%CI: 0.9, 3.2]. Anxiety (HR = 2.3, 95%CI: 1.2, 4.3) and ARV non-adherence (HR = 2.4, 95%CI: 1.2, 4.9) were associated with incident loss of viral suppression in unadjusted models. Anxiety (HR = 2.0, 95%CI: 0.9, 4.5), ARV non-adherence (HR = 1.7, 95%CI: 0.8, 3.6), lack of health insurance (HR = 1.4, 95%CI: 0.7, 3.3), and not being in care (HR = 3.3, 95%CI: 0.9, 10.0) were associated with higher hazard of loss of viral suppression, though none was statistically significant in the adjusted model. Conclusion: Approximately 1 in 5 Black MSM and 1 in 10 White MSM experienced incident loss of viral suppression per year in our cohort. Individual- level factors such as mental health issues and insurance status may be contributing to incident loss of viral suppression and need further exploration. Results of this interim analysis might change in terms of magnitude or statistical significance. This study will assist in the design of tailored interventions for Black and White MSM to prevent loss of and minimize differences in maintenance of HIV viral suppression. 903 IMPACT OF HIV TEST-AND-TREAT INITIATIVE IN MIAMI-DADE COUNTY, FLORIDA Karalee Poschman 1 , Emma C. Spencer 2 , David Goldberg 2 , Kira Villamizar 2 , Tiffany Adams 2 , Jeffrey Beal 2 1 CDC, Atlanta, GA, USA, 2 Florida Department of Health, Tallahassee, FL, USA Background: Rapid access to antiretroviral therapy (ART) immediately following HIV testing is upheld as a prevention tool to reduce HIV transmission and improve outcomes along the HIV care continuum. In 2016, the Miami-Dade County Health Department launched a test and treat (T&T) initiative to offer same-day or next-day access to ART following initial HIV diagnosis. This study aims to evaluate HIV care outcomes, including viral load (VL) suppression (<200 copies/mL) and retention in HIV care (two or more HIV-related labs, medical visits or prescriptions at least three months apart), for persons whose HIV was diagnosed in Miami-Dade County in 2017. Methods: Clinical and epidemiological data reported to the Florida Department of Health HIV/AIDS surveillance systemwere matched to lab, medical visit and prescription records in Ryan White Program databases, county health

department electronic health records and Medicaid claims. HIV care outcomes among antiretroviral-naïve patients whose initial HIV diagnosis was in Miami- Dade County in 2017 and who engaged in HIV care (n=950), including patients in T&T (n=80), were evaluated to determine the impact of T&T. Results: T&T did not significantly impact the rate of HIV care initiation within 30 days of diagnosis (85.0% vs. 81.5%). However, patients in T&T were more likely to achieve VL suppression within six months of diagnosis (87.5% vs. 66.1%, p<0.01) and be retained in care (91.3% vs. 81.6%, p=0.03). For patients with a suppressed VL within six months of diagnosis, the average number of days from diagnosis to VL suppression was lower for T&T (71 vs. 87, p<0.01). When evaluating patients retained in care, higher rates of VL suppression (90.4% vs. 76.1%, p<0.01) and more rapid VL suppression (72 vs. 89 days, p<0.01) persisted for T&T. Furthermore, patients in T&T were more likely to receive HIV resistance testing within three months of diagnosis (80.0% vs. 57.8%, p<0.01). Conclusion: While T&T did not significantly impact the timing of HIV care initiation, patients in T&T were more likely to achieve VL suppression within six months of diagnosis and progress to VL suppression more rapidly. Patients in T&T were also more likely to receive a baseline HIV resistance test, indicating a complete initial HIV care assessment. Rapid access to ART following HIV diagnosis can help reduce HIV-related mortality, improve health outcomes of those living with HIV and reduce HIV transmission through VL suppression. 904 ANTIRETROVIRAL REGIMEN DURABILITY IS NOT DRIVEN BY VIRAL FAILURE IN AN AFRICAN COHORT Christina Polyak 1 , Allahna L. Esber 1 , Francis Kiweewa 2 , Jonah Jonah Maswai 3 , John Owuoth 3 , Lucas Maganga 4 , Emmanuel Bahemana 5 , Yakubu Adamu 6 , Patrick Hickey 7 , Peter Coakley 1 , Anne K. Monroe 8 , Amanda D. Castel 8 , Alan Greenberg 8 , Trevor A. Crowell 1 , Julie Ake 9 1 Henry M Jackson Foundation, Bethesda, MD, USA, 2 Makerere University Walter Reed Project, Kampala, Uganda, 3 Walter Reed Project–Kisumu, Kisumu, Kenya, 4 Mbeya Medical Research Programme, Mbeya, Tanzania, United Republic of, 5 Walter Reed Program–Tanzania, Mbeya, Tanzania, United Republic of, 6 Walter Reed Program– Nigeria, Abuja, Nigeria, 7 Uniformed Services University of the Health Sciences, Bethesda, MD, USA, 8 George Washington University, Washington, DC, USA, 9 US Military HIV Research Program, Bethesda, MD, USA Background: Data on durability of first-line regimens in resource-limited settings are limited. We reviewed data from a large ongoing multinational African Cohort study (AFRICOS) to describe reasons and assess time to switching or stopping first-line antiretroviral therapy (ART). Methods: AFRICOS prospectively enrolls HIV-infected and uninfected adults at 12 President’s Emergency Plan for AIDS Relief (PEPFAR) supported facilities across 5 programs in Kenya (Kisumu and the South Rift Valley), Tanzania, Uganda, and Nigeria. ART regimen history is obtained at entry from available records and updated prospectively every 6 months. Reasons for switching or stopping ART are recorded by study physicians. For these analyses, we included HIV-infected participants who had documented ART start and stop dates, either prior to cohort entry or once enrolled. Time to switching or stopping a regimen was the primary endpoint used to assess durability. We generated Kaplan-Meier curves stratified by variables of interest and used the log-rank test to evaluate for significant differences. Results: Between January 2013 and June 2018, we enrolled 2820 HIV-infected adults (58% female) with a median age of 36 (IQR 30-44) years. Of these, 2663 (94%) were ART experienced and have initial ART start dates available, including 1154 (43%) that began ART before the study initiation in 2013. The first regimen for the majority (1396; 52%) was efavirenz/lamivudine/tenofovir disoproxl fumarate. The median duration of this regimen was 2.25 (IQR 0.94-3.88) years. The initial regimen was switched or stopped for 1207 (45%) participants for reasons including change in country guidelines (344; 29%), toxicity (281; 23%), stock-out (262; 21%) and regimen failure (107; 9%). Regimen durability did not differ by gender (Figure, panel A), but was reduced in the youngest age group evaluated (18-29 years; panel B), varied substantially by site (panel C), and was reduced with initial regimens containing D4T (panel D). Conclusion: In this large African cohort, the durability of first-line ART regimens was driven largely by factors other than viral failure. Specific regimens associated with high toxicity and abbreviated durability are no longer PEPFAR standard of care, but persistent programmatic factors that contribute to drug stock-outs and other barriers to ART maintenance require further investigation and intervention, especially as PEPFAR implements the programwide first-line transition to tenofovir/lamivudine/dolutegravir (TLD).

Poster Abstracts

CROI 2019 352