CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Results: The 2625 patients were mostly male (75.4%), black (49.9%), and MSM (53.3%). Median age was 39 years old (IQR 29-48). Most patients (89.1%) were ART naïve at study entry. The principal anchor drugs at virologic suppression were NNRTI (48.9%), PI (36.5%) and INSTI (17.4%). Overall, 208 (7.9%) patients experienced virologic failure. A total of 155 (5.9%) patients experienced LLV to 51-200 copies/mL and 117 (4.5%) patients experienced LLV to 201-500 copies/ mL. There was a higher risk of virologic failure for successively higher LLV categories (Figure 1a), while only the blip 200-500 category was associated with a higher failure rate (Figure 1b). Both LLV 51-200 (Adjusted Hazard Ratio (aHR) 2.06 [1.26,3.37]) and LLV 201-500 (aHR 4.05 [2.49,6.58]) were associated with virologic failure (VF) in univariate and multivariate analysis. In sensitivity analysis excluding ART experienced patients, the association between LLV51- 200 and VF was no longer statistically significant. Blip 201-500 was directly associated with VF (aHR 2.25 [1.29,3.93]). Compared with NNRTIs, PIs were directly associated with VF. Black race (compared to white race: aHR 1.78 [1.20,2.64]) and PWID (compared to MSM: aHR 2.63 [1.50,4.61]) were associated with an increased risk of VF. Conclusion: Low level viremia between 201-500 copies/mL was associated with virologic failure. LLV between 51-200 copies/mL was also associated with virologic failure, particularly among ART experienced patients. Our findings suggest that patients with LLV below the current DHHS threshold for virologic failure (persistent viremia ≥ 200 copies/mL) may also be at increased risk for failure.

person-years (95% CI 6.0, 8.8), and stable across calendar years, for example 5.0 (95% CI 2.0, 12.5) in 2008 and 8.2 (95% CI 4.8, 14.2) in 2016 (P for trend = 0.13). A genotypic resistance test was obtained for 61 patients at FVF: 18 of 28 (64%) patients on an NNRTI had NNRTI resistance, 1/19 (5%) on a PI had PI resistance, and 0/13 (0%) on an INSTI had INSTI resistance. After FVF, second-line ART regimens most commonly contained a PI (38%), INSTI (37%), PI/INSTI (12%), or NNRTI (4%). Of 19 patients on an INSTI at FVF, 9 initiated second-line ART with a PI, 3 with an INSTI, 3 with PI/INSTI, 2 with PI/NNRTI/INSTI. Conclusion: In 2008–2016 in this cohort, one-fifth of patients who had FVF and subsequently changed ART were on an INSTI at the time of FVF and one-half on an NNRTI, and many had NNRTI resistance at switch. Over half of second-line ART regimens after FVF contained an INSTI. Assessing clinical outcomes on second- line ART is important for managing first virologic failure.

Poster Abstracts

499 PREDICTORS OF VIROLOGIC OUTCOME WHILE CONTINUING A PI-BASED ART REGIMEN IN ACTG A5288 Robert A. Salata 1 , Beatriz Grinsztejn 2 , Justin Ritz 3 , Ann Collier 4 , Peter Mugyenyi 5 , Evelyn Hogg 6 , Linda Wieclaw 7 , Robert Gross 8 , Catherine Godfrey 9 , Nagalingeswaran Kumarasamy 10 , Cecilia Kanyama 11 , John W. Mellors 12 , Carole Wallis 13 , Michael D. Hughes 3 1 Case Western Reserve University, Cleveland, OH, USA, 2 Instituto Nacional de Infectologia Evandro Chagas, Rio de Janeiro. Brazil, 3 Harvard University, Boston, MA, USA, 4 University of Washington, Seattle, WA, USA, 5 Joint Clinical Research Centre, Kampala, Uganda, 6 Social & Scientific Systems, Silver Spring, MD, USA, 7 Frontier Science & Technology Research Foundation, Inc, Amherst, NY, USA, 8 University of Pennsylvania, Philadelphia, PA, USA, 9 DAIDS, NIAID, Bethesda, MD, USA, 10 YR Gaitonde Center for AIDS Research and Education, Chennai, India, 11 University of North Carolina Project–Malawi, Lilongwe, Malawi, 12 University of Pittsburgh, Pittsburgh, PA, USA, 13 Lancet Labs and BARC SA, Johannesburg, South Africa Background: Antiretroviral (ARV) choices are challenging in resource-limited settings (RLS) after failure of 2nd-line therapy because of accumulated resistance. Many failing 2nd-line therapy without resistance remain on their 2nd-line therapy. Our objective was to evaluate demographic and predictors of successful virologic suppression in this prospective study. Methods: A5288 was an open-label strategy study in RLS in HIV-1 infected persons with confirmed plasma HIV RNA (VL) ≥ 1000 c/mL after 24 weeks of PI-based 2nd-line ART. The study sought to use newer ARVs (darunavir/r, etravirine and raltegravir) along with genotyping (GT), a cellphone adherence intervention (CPI) or standard of care (SOC), and real-time HIV VL monitoring to achieve VL suppression at week 48. Participants were assigned to 1 of 4 cohorts based on GT at time of study entry, previous ART history and any other GT available. This analysis focuses on the 287 (53%) enrolled in Cohort A (no lopinavir resistance) from Feb-2013 to Dec-2015. These 287 participants remained on 2nd-line PI-based ART regimen, with flexibility to change their NRTIs. Logistic regression was used to evaluate sex, age, baseline HIV-1 RNA, CD4 count, presence of resistance to at least one NRTI, and adherence support (CPI+SOC vs SOC) as predictors of the study’s primary endpoint: suppression of HIV-1 RNA ≤200 c/mL at week 48 (ITT).

498 SECOND-LINE ANTIRETROVIRAL THERAPY (ART) AFTER FIRST VIRAL FAILURE ON ART, 2008-2016 Thibaut Davy-Mendez , Sonia Napravnik, Oksana Zakharova, Joseph J. Eron University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: Antiretroviral therapy (ART) regimens currently recommended for initial HIV treatment are efficacious and tolerable, with a high likelihood of achieving and maintaining virologic suppression. However, managing patients with virologic failure can be challenging, and data on second-line ART following first virologic failure with current regimens is very limited. Methods: We included UNC CFAR HIV Clinical Cohort patients in care 2008– 2016, who were receiving ART with no prior evidence of virologic failure. Our outcome was defined as an ART change (including at least a change in anchor agent within 6 months) following first virologic failure (FVF) on ART, defined as the first detectable viral load (>1000 copies/ml) >24 weeks after ART initiation. Alternate definitions of FVF were considered in sensitivity analyses. We characterized the ART history of these patients and estimated incidence rates (new cases of second-line ART after FVF, divided by number of person-years without the outcome), overall and by calendar year, using Poisson regression. We examined genotype tests performed at FVF and defined resistance by class as ≥1 major IAS-USA mutation. Results: Between 2008 and 2016, among 2671 patients who contributed 13 814 person-years, 100 changed ART after FVF. Of these 100 patients, 44% were currently on an NNRTI anchor agent at FVF, 36% on a PI, 15% on an INSTI, and 3% on PI/INSTI (Fig.). The median year of ART initiation was 2008 (IQR 2004–2010). The overall incidence rate of FVF with ART change was 7.2 per 1000

Results: 56% of Cohort A participants were female, median age was 40. Median HIV-1 RNA was 4.3 log10 c/mL and CD4 count was 171 cells/mm3; 113 (39%) had

CROI 2019 185