CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

THURSDAY, FEBRUARY 26, 2015 Session P-N10 Poster Session

Poster Hall

2:30 pm– 4:00 pm HCV: Resistance to Antiviral Agents 696 Characterization of Naturally Occurring Resistance to HCV NS5A Inhibitors Jennifer Cook; Owen Solberg; Alicia Newton; Suqin Cai; Arne Frantzell; Jacqueline Reeves; Christos J Petropoulos; JonathanToma; Wei Huang Monogram Biosciences, South San Francisco, CA, US

Background: Clinical trials with some NS5A inhibitor-containing regimens have resulted in less favorable responses among individuals with GT1a compared to GT1b HCV. To assess whether quantitative or qualitative differences in naturally occurring resistant variants may explain the discordant response rates, we used deep sequencing and recombinant replicons to determine the prevalence and resistance profiles of NS5A inhibitor resistant variants in HCV from DAA treatment-naïve individuals. Methods: NS5A regions were amplified from 109 plasma samples (1a=71, 1b=38) and incorporated into a Con1 luciferase reporter replicon. NS5A sequencing was performed using the Illumina MiSeq platform. NS5A inhibitor resistance-associated variants (RAVs) located at positions 28, 30, 31, 32, and 93 were recorded (lower threshold=0.5%). RAVs were introduced into reference replicons by site-directed mutagenesis (SDM). Replicons containing plasma-derived NS5A sequences or SDMs were evaluated for NS5A inhibitor susceptibility Results: NS5A inhibitor RAVs were detected in 11/71 (15.5%) GT1a viruses: M28T=3, Q30H=4, L31M=3 and Y93C/H/N=6. Among the 11 viruses with GT1a RAVs, six harbored a single RAV while five had more than one RAV. RAVs were present at >10% (range: 13.6-99.2%) of the quasispecies in 6/11 viruses and at <10% (range: 0.92-3.1%) for the remaining five viruses. Similarly, major NS5A RAVs were detected in 7/38 (18.4%) GT1b viruses: L31M=1, Y93H=7. RAVs were present at >10% (range: 20.7-98.8%) of the quasispecies in 2/7 viruses and at <10% (range: 0.54-1.1%) for the remaining five viruses. Codons for all observed RAVs differed by only one nucleotide from the predominant “wild type” codon, except L31M in GT1b. Replicons containing patient NS5A sequences with RAVs exhibited large reductions in susceptibility to NS5A inhibitors. Reductions in NS5A inhibitor susceptibility were greater when RAVs were tested in the context of GT1a NS5A sequences (FC >150) compared to GT1b NS5A sequences (FC<20). Conclusions: The prevalence of resistant variants was similar between GT1a and GT1b isolates. Naturally occurring NS5A resistance variants were more diverse among GT1a viruses, compared to GT1b. Slightly higher proportions of resistant variants were present within GT1a virus populations compared to GT1b. The combination of multiple resistance pathways, lower genetic and resistance barriers may provide advantages for GT1a viruses to escape NS5A inhibition. 697 Hepatitis C Q80K Prevalence in BC, Canada, Determined by a Public Domain Assay Background: HCV genotype 1a (GT1a) infections harbouring a baseline Q80K polymorphism in the NS3 gene have reduced virologic response to IFN-based HCV treatments containing simeprevir. We aimed to develop, validate, and freely disseminate a NS3 clinical sequencing assay to detect Q80K and other NS3 mutations, and establish their prevalence in British Columbia (BC), Canada. Methods: RNA was extracted using a NucliSens easyMag, amplified by nested RT-PCR and sequenced by Sanger or Illumina sequencing. Sanger chromatogram interpretations were performed automatically using in-house analysis software (ReCall). HCV genotypes were identified using INNO-LiPA HCV 2.0 (5’UTR); sequence-based genotype assignment was verified phylogenetically in comparison with reference sequences. Our assay was validated via comparison with 70 sequences from an external laboratory and relative to consensus sequences from next-generation sequencing (MiSeq). To establish the prevalence of Q80K in BC, samples from 376 GT1 LiPA HCV+ individuals diagnosed in 2011 were examined. Sequences were also screened for mutations associated with boceprevir/telaprevir resistance. Results: Comparison of sequences generated by an external lab and with MiSeq consensus sequences revealed > 98% sequence concordance, and no discordant calls of Q80K. LiPA identified 231 (61%) individuals as GT1a, 91 (24%) as GT1b, while subtype could not be resolved in 54 cases (14%). The prevalence of Q80K was 61%, 11%, and 50% in LiPA GT1a, 1b and unresolved, respectively. Sequence analysis suggested that 52/54 individuals (96%) whose LiPA genotype was unresolved were actually GT1a, as were 28/91 (44%) cases that LiPA identified as 1b. Using a sequence-based assessment of HCV genotype, the prevalence of Q80K was 176/309 (57%) in GT1a compared to 1/66 (2%) in GT1b. Mutations associated with boceprevir or telaprevir resistance in this newly diagnosed population were also observed (V36M, N=2; T54S, N=5; and R155K, N=3). Conclusions: Our results suggest the overall prevalence of Q80K in GT1 in BC is 47%. Q80K was highly prevalent where LiPA results could not resolve GT1a vs 1b. Inconsistent HCV genotype 1a vs 1b assignment by LiPA UTR in comparison to sequence-based genotyping suggests that all individuals with GT1 HCV infection in BC should be screened for the Q80K polymorphism. Assay details and software are freely available to academic labs. 698 HCVNS3 Variants in HIV/HCV Coinfected Patients Before-After PegIFN/Ribavirin Enass Abdel-Hameed 1 ; Susan D. Rouster 1 ; Xiang Zhang 2 ; Jing Chen 2 ; Mario Medvedovic 2 ; Kenneth E. Sherman 1 1 University of Cincinnati, Cincinnati, OH, US; 2 University of Cincinnati, Cincinnati, OH, US Background: This study was designed to characterize HCV NS3 protease resistance associated variants (RAVs) by deep sequence analysis at baseline and after 12 weeks of PegIFN / ribavirin treatment in subjects who failed to achieve an early viral response. Overall genetic variability at NS3 protease drug resistant sites was also assessed. Methods: Thirty HCV/ HIV coinfected subjects were evaluated. All subjects were previously enrolled/consented in ACTG 5178 (SLAM-C), a study designed to evaluate the effects of prolonged PegIFN on hepatic fibrosis. An RT-PCR-amplified NS3 product was fragmented and primed into a sequencing library. The individually pooled libraries were sequenced with Illumina HiSeq next generation sequencing system set for single read. Results: The study cohort was predominantly male (83.3%), with a median age of 45years. 40%were white and 56.7% black. Baseline mean HCV log 10 viral load was 6.6 and HIV viral load was 3365.1. 56.7% had undetectable HIV due to effective cART. IL28B distribution for C/C, C/T, and T/T were 35, 45, and 20 % respectively. Sixteen patients were genotype 1a, 2 were 1a1b and 12 were type 1 no subtype. At baseline, protease RAVs was present in 73.3% of patients and expanded to 83.3 % of patients after 12 weeks. At baseline RAVs, V36L, T54S, T54A, and V55A were each detected in 3% of patients. V36M was detected in 4% of patients.I170V was present in 37% of patients. Q80K showed the highest prevalence at 44%. After 12 weeks of treatment, V55I and V36M expanded to 9% and 12 % of the population respectively, while Q80K was now detected in 50 % of the patients with a positive correlation to liver fibrosis stage (r= 0.5235, p =0.031). The proportion of subjects having detectable I170V decreased to 20% of patients. Over all amino acid positions studied, no statistical difference in relative abundance of RAVs within a patient before-after 12 weeks of treatment was seen. There was no relationship to IL28B, viral loads, age, gender or race. Jeffrey B. Joy 1 ; Celia K. Chui 1 ; Chanson J. Brumme 1 ; Mel Krajden 2 ; Andrea Olmstead 2 ;Winnie Dong 1 ;Wendy Zhang 1 ; Aram Karakas 1 ; Huong Hew 3 ; Richard Harrigan 1 1 BC Centre for Excellence in HIV/AIDS, Vancouver, Canada; 2 BC Centre for Disease Control, Vancouver, Canada; 3 Janssen Pharmaceuticals, Inc., Toronto, Canada

Poster Abstracts

432

CROI 2015