CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

479 PREDICTORS OF HIV GENOTYPE TESTING AMONG NEW DIAGNOSES LINKED TO CARE IN PHILADELPHIA Melissa Miller , Tanner Nassau, Makeda Carroll, Coleman Terrell, Kathleen Brady Philadelphia Dept of Pub Hlth, Philadelphia, PA, USA

Background: Genotypic testing (GT) is recommended by Public Health Service treatment guidelines and used as part of routine HIV surveillance to assess trends in HIV drug resistance and transmission patterns. Despite increasing rates of linkage to care (LTC), GT ≤90 days post diagnosis remains stable around 50%. This analysis aims to identify and address barriers to timely GT among persons who have successfully LTC. Methods: Data were extracted from Philadelphia’s Enhanced HIV/AIDS Reporting System on individuals aged >18+ newly diagnosed with HIV from 2013-2015 and LTC within 90 days of diagnosis. LTC was defined as having a viral load (VL) or CD4 < 90 days after diagnosis. Multivariable logistic regression models predicted timely GT, defined as GT < 90 days of diagnosis. Models were adjusted for race, sex at birth, age at diagnosis, exposure category, type of diagnosing facility (testing/linkage site, outpatient medical facility, inpatient facility, correctional facility, other), year of diagnosis, insurance status, concurrent HIV/AIDS diagnosis, LTC within 30 days, and viral suppression (VS, defined as a VL <200 copies/ ml at the last measure in the year after diagnosis). Results: 1,383 Philadelphia residents that LTC < 90 days of diagnosis were included in the analysis. Of these, 56% received timely GT. New diagnoses aged 45+ were less likely to have evidence of timely GT compared to diagnoses aged 18-24 (AOR, 0.6; 95% CI:0.4-0.8). Those diagnosed at correctional facilities (AOR,0.4; 95% CI:0.2-0.7), inpatient facilities (AOR, 0.5; 95% CI:0.3-0.8), and outpatient medical facilities (AOR, 0.6; 95% CI:0.4-0.9) were all less likely to have timely GT compared to those diagnosed at testing and linkage sites. Individuals without insurance at diagnosis were 45% less likely to have timely GT compared to privately insured individuals (AOR, 0.5; 95%CI:0.4-0.8) Individuals who achieved VS within 1 year post diagnosis were 1.6 times as likely to have received timely GT as those who did not (95%CI:1.3-2.0). Conclusion: Our data show poor adherence to recommendations for baseline GT in persons newly diagnosed with HIV. Timely GT is critical in improving long term care outcomes and in creating local transmission networks that provide entry points for intervention to reduce new HIV infections, both in keeping with National HIV/AIDS Strategy efforts. Age, insurance status, diagnosing facility, and VS are all factors that should be considered when targeting efforts to increase GT at the patient and provider level. 480 PRETREATMENT DRUG-RESISTANCE INCREASE IN 3 FOCAL POINTS OF MEXICO, 2012–2015 Claudia García-Morales 1 , Santiago Avila-Rios 1 , Daniela Tapia-Trejo 1 , Verónica S. Quiroz-Morales 1 , Samuel Navarro 2 , Carlos A. Barrera-Arellano 3 , Jesús Casillas-Rodríguez 4 , Akio Murakami-Ogasawara 1 , Gustavo Reyes-Terán 1 , for the HIV MexNet Group 1 Natl Inst of Respiratory Diseases, Mexico City, Mexico, 2 Hosp General Tijuana, Tijuana, Baja California, USA, 3 Capasits Cancún, Cancún, Quintana Roo, Mexico, 4 Clínica Especializada Condesa, Mexico City, Mexico Background: Pre-treatment drug resistance (PDR) is a concern for the management of HIV infection, with several low- and middle-income countries recently reporting levels of HIVDR above 10% amongst antiretroviral treatment (ART) naïve individuals. In Mexico, a nationally representative survey estimated a 16% PDR level in 2015. Following up on this study, we longitudinally assessed HIV PDR in three focal points of the Mexican HIV epidemic. Methods: HIV-infected ART-naïve individuals were recruited from 2008 to 2015: 589 from Tijuana, 1,142 from the central Metropolitan Zone (MZ), including Mexico City and the State of Mexico, and 547 from Cancun. Plasma HIV pol sequences were obtained. PDR was estimated using the Stanford HIVdb tool (v.7.0). Results: From 2012 to 2015, the overall PDR was 13.5% (95%, CI 10.7%-16.7%) in Tijuana, 14.9% (95%, CI 12.0%-18.2%) in Cancun and 16% (95%, CI 12.9%-19.4%) in the MZ. PDR to NNRTI was higher than to NRTI and PI in the MZ (10.6%) and Cancun (7.9%) (p<0.05), but not in Tijuana (5.7%) (p>0.05). No differences in overall or drug class-specific PDR were observed between the geographic zones. An increasing trend in overall PDR was observed in the MZ from 2008 to 2015 (10.2% to 16.1%, p=0.04), particularly associated with increasing PDR to NNRTI (3.6% to 12.3%, p=0.0072). Aditionally, PDR to TDF+FTC+EFV, the most widely used ART regimen in Mexico, significantly increased in this region (2.2% to 10.7%, p=0.0022), with PDR to EFV increasing from 0.6% to 10.0% (p=0.0011). Cancun also showed an increasing trend in PDR to NNRTI (4.7% to 13.3%, p=0.0167), TDF+FTC+EFV (6.3% to 13.3%, p=0.0167), and EFV (3.2% to 11.1%, p=0.0048) from 2012 to 2015. Similarly, Tijuana showed an increasing trend in overall PDR (11.6% to 17.5%, p=0.0167), but not to NNRTI. K103N was the most frequent DR mutation in the MZ (4.4%) and Cancun (4.2%), but not in Tijuana (1.4%). Clusters of drug-resistant viruses mostly frommales were observed in the MZ. The proportion of females in clusters increased in Tijuana and to a lesser extent in Cancun. Conclusion: HIV PDR significantly increased in the three focal points, but PDR showed zone-specific characteristics. NNRTI PDR significantly increased in Cancun and the MZ, but not in Tijuana. Importantly, PDR to the most widely used ART regimen in Mexico reached levels over 10% in Cancun and the MZ warranting immediate programmatic actions. Our observations also underscore the importance of implementing sub-national PDR surveys. 481 HIV DRUG RESISTANCE: A UNIQUE PERSPECTIVE ACROSS 4 AFRICAN COUNTRIES Brook Danboise 1 , Christina Polyak 1 , Alex Kasembeli 2 , Samoel Khamadi 3 , Babajide Keshinro Keshinro 4 , Hannah Kibuuka 5 , John Owuoth 6 , Sheila A. Peel 1 , Julie Ake 1 , Trevor Crowell 1 1 US Military HIV Rsr Prog, Bethesda, MD, USA, 2 Walter Reed Proj–Kericho, Kericho, Kenya, 3 Walter Reed Prog–Tanzania, Mbeya, Tanzania, United Republic of, 4 Walter Reed Prog–Nigeria, Abuja, Nigeria, 5 Makerere Univ Walter Reed Proj, Kampala, Uganda, 6 Walter Reed Proj, Kisumu, Kenya Background: The World Health Organization (WHO) has identified HIV drug resistance (HIVDR) as a significant threat to ending the AIDS pandemic and has called for increased surveillance for drug resistance. We evaluated HIVDR patterns among participants with and without a history of antiretroviral therapy (ART) in the prospective, multinational African Cohort Study (AFRICOS). Methods: AFRICOS enrolls adults at 11 PEPFAR-supported facilities in Uganda, Kenya, Tanzania, and Nigeria. At enrollment, all HIV-infected participants with viral load ≥1000 copies/mL receive HIVDR testing with pol subtype. We tallied mutations classified as conferring high-level or low-level resistance to nucleoside reverse transcriptase inhibitors (NRTIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs) and protease inhibitors (PIs) according to the Stanford HIVDB 7.0.1 and SmartGene IDNS software. For participants with no history of ART use, WHO surveillance drug resistance mutations (SDRMs) were also noted. Results: From 21 January 2013 to 9 December 2015, 1915 HIV-infected participants enrolled in AFRICOS. Of these, 746 had a viral load ≥ 1000 copies/mL and 253 had baseline HIVDR testing results available for inclusion in this analysis. Ninety-four (37%) were male and the median age of participants was 35 (interquartile range 28-41) years. One hundred eighty-eight (74.3%) participants were ART-naïve and 65 (25.7%) were ART-experienced at enrollment. Table 1 summarizes HIVDR for ART-naïve and ART-experienced participants in each country and the overall cohort. SDRMs among ART-naïve participants were rare (10/188, 5.3%), with K103N as the most frequent major mutation (8/188, 4.3%). HIVDR among ART-experienced participants was mostly driven by mutations conferring resistance to NRTIs (29/65, 44.6%) and NNRTIs (42/65, 64.6%), including a 10.8% rate of K65R. PI resistance was noted but uncommon (4/65, 6.2%). Conclusion: The prevalence of SDRMs among ART-naïve participants was low in Uganda and Nigeria, whereas moderate rates in Kenya and Tanzania suggest the need for evaluation of clinics, educational efforts and screening programs. Continued surveillance will inform local guidelines about HIVDR testing prior to ART initiation. Among ART- experienced participants, NRTI and NNRTI mutations were common, particularly in Kenya, Tanzania and Nigeria. Although currently uncommon, emerging resistance to PIs in these countries may eventually limit the effectiveness of current empiric second-line therapy.

Poster and Themed Discussion Abstracts

CROI 2017 201