CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

N=1274) was correlated with prespecified pooled virologic outcomes from the 2 studies. Virologic failures with HIV-1 RNA ≥200 c/mL had resistance analyses at confirmed failure or last visit. Results: High levels of virologic suppression of HIV-1 RNA to <50 c/mL at W48 was achieved: 91% (576/634) in the pooled B/F/TAF groups and 93% of subjects in both the DTG/ABC/3TC (293/315) and DTG+F/TAF (302/325) groups, respectively. There was rapid suppression to HIV-1 RNA <50 c/mL at W4: 76% (477/625) in the B/F/TAF group, 76% (236/311) in the DTG/ABC/3TC group, and 80% (258/324) in the DTG+F/TAF group by the Missing=Excluded approach. HIV-1 subtype B was present in 89% of patients. 92% (1048/1138) B subtype and 90% (123/136) non-B subtype had HIV-1 RNA <50 c/mL at W48. Pre-existing primary resistance mutations (-R) to any drug class were found in 18% (224/1274) of patients overall and consisted of NRTI-R in 2%, NNRTI-R in 13.2%, PI-R in 2.9%, and INSTI-R in 1.3%. One subject in the B/F/TAF group had Q148H+G140S at baseline and was suppressed with HIV-1 RNA <50 c/mL at W48. HIV-1 RNA <50 c/mL at W48 was similar with or without pre-existing resistance mutations (92%; 205/224 vs 92%; 966/1050, respectively). Through W48, 17 patients qualified for resistance testing (1.3% (8/634) B/F/TAF; 1.3% (4/315) ABC/DTG/3TC; 1.5% (5/325) DTG+F/TAF); none had emergent resistance to study drugs. Conclusion: Treatment with B/F/TAF, ABC/DTG/3TC, or DTG+F/TAF rapidly achieved and maintained high rates of virologic suppression in HIV-1 treatment- naïve subjects. The presence of pre-existing resistance mutations did not affect treatment outcomes. Development of primary drug resistance mutations to study drug was not observed through W48. 533 HIV DRUG RESISTANCE WITH EARLY VS DELAYED ANTIRETROVIRAL TREATMENT (HPTN 052) Philip J. Palumbo 1 , Jessica M. Fogel 1 , Ethan A. Wilson 2 , Sarah E. Hudelson 1 , Stephen Hart 3 , Laura Hovind 3 , Estelle Piwowar-Manning 1 , Maria Papathanasopoulos 4 , Mariza Morgado 5 , Shanmugam Saravanan 6 , Srikanth Tripathy 7 , Ying Q. Chen 2 , Myron S. Cohen 8 , Susan H. Eshleman 1 1 The Johns Hopkins University, Baltimore, MD, USA, 2 Fred Hutchinson Cancer Research Center, Seattle, WA, USA, 3 Frontier Science & Technology Research Foundation, Inc, Amherst, NY, USA, 4 University of the Witwatersrand, Johannesburg, South Africa, 5 Oswaldo Cruz Foundation - Fiocruz, Rio de Janeiro, Brazil, 6 YR Gaitonde Center for AIDS Research and Education, Chennai, India, 7 National Institute for Research in TB, Chennai, India, 8 University of North Carolina Chapel Hill, Chapel Hill, NC, USA Background: HIV Prevention Trials Network (HPTN) 052 was a multi-national, clinical trial that demonstrated reduced HIV transmission and health benefits of early antiretroviral treatment (ART). The study enrolled 1,763 serodiscordant couples and followed them for up to 10 years (2005-2015). At enrollment, HIV- infected index participants were randomized to the early ART arm (CD4: 350-550 cells/mm 3 at ART initiation) or the delayed ART arm (CD4: ≤250 cells/mm 3 at ART initiation). All index participants were offered ART at any CD4 cell count after the release of interim study results in May 2011. The rate of virologic failure was similar in the two study arms. We evaluated factors associated with HIV drug resistance in HPTN 052 participants with virologic failure. Methods: Virologic failure was defined as two consecutive viral load measures >1,000 copies/mL after 24 weeks on ART. HIV drug resistance was evaluated at ART initiation (baseline) and virologic failure using the ViroSeq HIV-1 Genotyping System and the Resistance Calculator Program at Frontier Science Foundation (Stanford v7.0 algorithm). Factors associated with HIV drug resistance were analyzed using Chi-square, t-tests, and logistical regression models using SAS software. Results: HIV genotyping results were obtained at the time of ART initiation (baseline) and at virologic failure for 211 (84.7%) of 249 participants (128 early arm; 83 delayed arm [22 started ART before May 2011]). Overall, 4.7% of participants had resistance at baseline and 35.5% had new resistance at failure. The frequency of new resistance at failure was lower among participants in the early arm compared to all participants in the delayed arm (30.5% vs. 43.4%, p=0.06), and compared to the subset of participants in the delayed armwho started ART before May 2011 (54.5%, p=0.032). The frequency of new resistance at failure was lower among participants with higher baseline CD4 cell counts (p=0.047) and lower baseline viral loads (p=0.0001), and was higher among those receiving a regimen of efavirenz, lamivudine, and zidovudine compared to other ART regimens (p=0.0074). In a multivariate model, new resistance at

failure was associated with baseline viral load (p=0.0008) and drug regimen (p=0.024). Conclusion: The frequency of new drug resistance at virologic failure was lower in adults with early ART initiation. The main factor associated with reduced drug resistance with early ART was lower baseline (pre-ART) viral load. 534 DEFINING CLINICALLY RELEVANT THRESHOLD FOR ULTRASENSITIVE HIV RESISTANCE TESTING Seth C. Inzaule 1 , Raph L. Hamers 1 , Marc Noguera-Julian 2 , Maria Casadella 2 , Mariona Parera 2 , Bonaventura Clotet 2 , Tobias F. Rinke de Wit 1 , Roger Paredes 2 1 Amsterdam Institute for Global Health and Development, Amsterdam, Netherlands, 2 IrsiCaixa Institute for AIDS Research, Badalona, Spain Background: Current HIV drug resistance tests, mainly based on Sanger- sequencing, have 20% detection thresholds that fail to detect minority variants (MVs). Ultrasensitive next-generation sequencing (NGS) offers potential as alternative lower-cost tests, but the clinical relevance of MVs is uncertain. Using data from a multi-country cohort study, we assessed which resistance threshold, including MVs, optimally predict an increased risk of virological failure (VF) on first-line NNRTI-based antiretroviral treatment (ART). Methods: Case-control study in the PanAfrican Studies to Evaluate Resistance cohort (PASER-M) in 12 sites in five countries. Cases were defined as those who had VF (viral-load (VL) ≥400 cps/ml) at 12 months or switched to second-line ART, and controls as those with VL<400 cps/ml on first-line ART at month 12, matched by country, age, pre-ART VL and CD4 cell counts. PDR was assessed by Illumina MiSeq next-generation sequencing of HIV-1 from pre-ART plasma samples, using the International Antiviral Society-USA 2017 drug resistance mutation list. We used logistic regression to assess the odds of VF at different PDR mutant detection thresholds (1% to 20%), compared with the cut-off used for major circulating variants in Sanger sequencing (≥20%). We calculated the relevant measures of diagnostic accuracy, summarized in Table. Results: We included 403 samples from 158 cases and 249 controls. Lowering the thresholds improved the sensitivity to identify cases, i.e. 12%, 13%, 16% and 18% at the 20%, 10%, 5% and 1% thresholds respectively. This came with a limited reduction in specificity (i.e. correctly identifying controls) of 98%, 96%, 96% at 20%, 10%, 5%, respectively, but a larger reduction to 92% at 1% threshold. DOR was 5.4, 4.0, 4.0 and 2.4 at the 20%, 10%, 5% and 1% thresholds, respectively. The OR of VF for the presence of PDR was 12.9, 8.3, 10.0 and 3.6 at 20%, 10%, 5% and 1%, respectively. NNT was 3, 4, 3 and 8 at 20%, 10%, 5% and 1% thresholds, respectively. A sensitivity analysis that excluded major variants (≥20%) showed that MVs at the 5% threshold were still significantly associated with VF (OR 7.6, 95%CI 1.1-53.4). Conclusion: Ultrasensitive resistance testing improves prediction of VF relative to the conventional 20% threshold. A 5%MV detection threshold could be an adequate cutoff for operationalization of ultrasensitive assays, although the marginal gain over 20% is small.

Poster Abstracts

535 PROVIRAL DRUG RESISTANCE ACQUIRED DURING EARLY THERAPY PREDICTS VIROLOGICAL OUTCOME Bindu P. Gopalan 1 , Reena Dsouza 1 , Karthika Arumugam 1 , Niharika Rajnala 1 , Udaykumar Ranga 2 , Anita Shet 3 1 St John’s Research Institute, Bangalore, India, 2 Jawaharlal Nehru Centre for Advanced Scientific Research, Bangalore, India, 3 Johns Hopkins University, Baltimore, MD, USA

CROI 2018 195