CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Background: The extent of viral evolution during early ART impacting virological outcome is unknown in HIV-1 infected children. We hypothesised that acquisition of treatment relevant drug resistance mutations (TRDRMs) in the proviral compartment of children at six months of ART predicts virological failure (VF). Methods: Children ≤ 16 years initiated on ART (2 NRTI + 1NNRTI) from 2012 to 2015 were included in a nested case-control study. Six monthly CD4 counts HIV-1 RNA load and adherence was assessed. ‘Cases’ were VFs (VL ≥200 copies/ ml at 2 years of ART) and ‘controls’ were viremia suppressors (VL<200 copies/ ml during the 2-year period). Genotyping of proviral DNA fromwhole blood was performed by next-generation sequencing (NGS) in the partial pol gene of HIV-1 RT (amino acid 40 to 240) prior to (baseline) and at six months of ART by a paired-end protocol (Illumina). Plasma virus was genotyped by Sanger method at baseline and at VF. Mutations were identified and categorized using HIVdb (v8.4) and IAS-USA 2017 list. Groups were compared by the Mann Whitney U test and predictors for VF were assessed by the conditional logistic regression model (CLR). Results: Among 67 children (30 cases and 37 age-matched controls; median 8 years; 51%male) infected with HIV-1 Subtype C, baseline viral load was 5.4 log copies/ml and CD4 18.5%. NNRTI started was nevirapine in 70% of cases and 81% of controls. No significant differences were observed at baseline between the two groups (p=>0.05). At VF 86.6% (26/30) had DRMs in plasma virus. In CLR univariate model adherence (OR=6.4, 95%CI= 1.3 to 32.5, p=0.02) detection of at least one DRM (OR=6.2, 95%CI= 2.1 to 18.4, p=0.001), TRDRM (OR=9.1, 95%CI =2.5 to 31.9, p=0.001) and NNRTI DRMs (OR=4.9, 95%CI=1.6 to 14.6, p=0.004) showed association with an increased risk of VF. TRDRMs or TRNNRTI DRMs absent at baseline and detected at month six increased the risk of early VF (OR=15.6, 95%CI=3.2 to 75.5, p=0.001, OR=7.1, 95%CI=1.8 to 27.9, p=0.005 respectively). In the multivariable model, adherence (OR=9.7, 95%CI 1.6-55.8, p=0.011) and TRDRMs acquired at month six of ART (OR=11.4, 95%CI 3.1-42.3 p=0.001) were independently associated with VF. Conclusion: These data highlight the extent of viral evolution in the cellular compartment under ART jeopardizing the therapeutic outcome. Paired sequencing of the proviral compartment for TRDRMs prior to and at six months of ART by NGS may be explored for assessing the virological outcome in children initiating 1st line NNRTI based ART. Valerie F. Boltz 1 , Wei Shao 2 , Elias K. Halvas 3 , James A. McIntyre 4 , Robert T. Schooley 5 , Shahin Lockman 6 , Judith S. Currier 7 , Fredrick Sawe 8 , Evelyn Hogg 9 , Michael D. Hughes 10 , Mary F. Kearney 1 , John M. Coffin 11 , John W. Mellors 3 1 National Cancer Institute, Frederick, MD, USA, 2 Leidos Biomedical Research, Inc, Frederick, MD, USA, 3 University of Pittsburgh, Pittsburgh, PA, USA, 4 Anova Health Institute, Johannesburg, South Africa, 5 University of California San Diego, La Jolla, CA, USA, 6 Brigham and Women’s Hospital, Boston, MA, USA, 7 University of California Los Angeles, Los Angeles, CA, USA, 8 Kenya Medical Research Institute, Kisumu, Kenya, 9 Social &Scientific Systems, Silver Spring, MD, USA, 10 Harvard University, Cambridge, MA, USA, 11 Tufts University, Boston, MA, USA Background: The clinical significance of low frequency HIV-1 drug-resistant variants is controversial. We hypothesized that variants encoding both NRTI and NNRTI resistance mutations, but not single class resistance mutations, would be associated with virological failure of ART. To test this hypothesis, we developed an ultrasensitive, NGS-based single viral genome sequencing assay (uSGS) (Boltz, Retrovirology 2016) that can detect rare variants with linked drug resistance mutations (DRMs). Methods: Pre-ART plasma samples were obtained from the NVP/FTC/TDF arm of the ACTG A5208/OCTANE trial (Lockman, NEJM 2010) for women who had received (Trial 1 [T1]) or had not received (Trial 2 [T2]) single dose nevirapine (sdNVP) within 6-24 months of entry. HIV-1 RNA-derived cDNA libraries were generated with primer IDs for paired-end next-generation uSGS of RT codons 58-217. We compared the frequency of linked, single- and dual-class (NRTI and NNRTI) DRMs conferring resistance to >1 component of NVP/FTC/TDF ART (Stanford db v.8.4) in pre-ART samples fromwomen with and without subsequent virological failure of ART. Results: A total of 1,344,024 single-genome sequences were obtained from pre-ART plasma samples from 122 women: 61 women from T1 (prior sdNVP), 27 with ART failure; and 61 from T2 (no prior sdNVP), 13 with ART failure. Overall, linked DRMs were found in more women in T1 than T2: 15/61 (25%) vs 3/61 (5%)

p=0.004. Importantly, linked dual-class DRMs were found in more women in T1 with ART failure than in T1 without failure: 8/27 (30%) vs 2/34 (6%); p=0.017. In addition, linked dual-class DRMs were found in more women in T1 with failure than in T2 with failure although the difference did not reach statistical significance: 8/27 (30%) vs. 1/13 (8%) p=0.23. Linked single-class DRMs were not associated with failure in either trial. Conclusion: This study shows for the first time that rare HIV variants with linked resistance mutations to 2 drug classes (NRTI and NNRTI) are significantly associated with ART failure, whereas variants with linked single-class resistance mutations are not. Transient exposure to a single ARV (e.g., single dose nevirapine) can increase the frequency of dual-class resistance and the risk of ART failure as a consequence of stochastic mutations occurring in a virus population that had already expanded through selection for resistance to the first ARV.

537 ACCUMULATION OF HIV DRUG RESISTANCE ON FAILING FIRST-LINE ART IN LUSAKA, ZAMBIA Parker Hudson 1 , Lloyd Mulenga 2 , AndrewWestfall 3 , Ranjit Warrier 4 , Aggrey Mweemba 2 , Michael Saag 3 , Jeffrey S. Stringer 5 , Joseph J. Eron 5 , Benjamin H. Chi 5 1 University of Texas at Austin, Austin, TX, USA, 2 University of Zambia, Lusaka, Zambia, 3 University of Alabama at Birmingham, Birmingham, AL, USA, 4 Center for Infectious Disease Research in Zambia, Lusaka, Zambia, 5 University of North Carolina Chapel Hill, Chapel Hill, NC, USA Background: In many high-burden settings, access to routine HIV viral load (VL) and resistance testing is limited; as a result, individuals with virologic failure (VF) to first-line anti-retroviral therapy (ART) may remain on their regimen for prolonged periods. Methods: We analyzed data from a cluster-randomized trial of VL monitoring in Lusaka, Zambia. From 2006 to 2011, 12 randomized sites provided either routine VL monitoring at 3mo, 6mo, and every 6mo after ART initiation (intervention) or immunologic and clinical monitoring with discretionary VL testing (control). Stored specimens in the controls were tested for VF at study completion. VF was defined as the first of two consecutive plasma VLs > 400 copies/mL at least 4 weeks apart. For those with VF, HIV genotyping was performed retrospectively at baseline (BL) and on all subsequent specimens on first-line ART until participants changed to second-line ART or were censored (study completion, withdrawal, death). We used the 2016 Stanford HIV Resistance Database to predict drug susceptibilities (v8.2). Results: Of 1,973 enrolled participants, 165 (8.4%) developed VF over 4,446 person years of follow up. Among those with VF, ART regimens included ZDV (51%), d4T (32%), TDF (16%), NVP (84%), and EFV (16%). 464 genotype results were available, including 132 (80%) at BL and 116 (70%) at VF. 125 (76%) had at least one result between VF and censoring, for a cumulative 191 person-years on failing first-line ART. At BL, 23% had a major NNRTI mutation, which did not affect median time to VF in a Kaplan-Meier survival analysis (p=0.4). At VF, 82% had a major NRTI/NNRTI mutation; this increased to a cumulative 89% at last genotype (LG). M184 mutations increased from 2% (BL) to 59% (VF) to 71% (LG). K65R mutations increased from 2% (BL) to 11% (VF) to 13% (LG); among those on a failing TDF-based regimen, TDF resistance increased from 42% (8/19, VF) to 58% (15/26, LG). 70% developed resistance to second generation NNRTIs. The majority of resistance occurred by the time VF was detected (figure). Conclusion: In this sub-study, over half who failed first line therapy with TDF developed TDF resistance. Incremental increases in HIV drug resistance accumulated after VF on failing first-line ART. These data highlight the urgent need for effective adherence strategies and earlier detection of VF if we are to reach the ambitious 90-90-90 goals set forth by international agencies and local governments.

Poster Abstracts

536 RARE HIV VARIANTS WITH LINKED DUAL-CLASS RESISTANCE ARE ASSOCIATED WITH ART FAILURE

CROI 2018 196