Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Performance Assessment of a Fully Automated Deep Sequencing Platform for HCV Resistance Testing

Performance Assessment of a Fully Automated Deep Sequencing Platform for HCV Resistance Testing Antiviral Therapy 2019; 24:417–423 (doi: 10.3851/IMP3318) Original article Performance assessment of a fully automated deep sequencing platform for HCV resistance testing 1,2 1,2 1,2 1,2 1,2 Christophe Rodriguez , Mélanie Mercier-Darty , Alexandre Soulier , Lila Poiteau , Mélanie Wlassow , 1,2 1,3 1,2† 1,2† Slim Fourati , Christophe Hézode , Jean-Michel Pawlotsky , Stéphane Chevaliez * Department of Virology, National Reference Center for Viral Hepatitis B, C and Delta, Hôpital Henri Mondor, Université Paris-Est, Créteil, France INSERM U955, Créteil, France Department of Hepatology, Hôpital Henri Mondor, Université Paris-Est, Créteil, France These authors jointly supervised this work *Corresponding author e-mail: stephane.chevaliez@aphp.fr Background: International liver society guidelines recom- treatment regimen were extracted and sequenced by two mended to perform HCV resistance testing at baseline of methods including population sequencing of the NS3, first-line therapy with certain combination regimens or NS5A and NS5B coding region reference method and the prior to retreatment in patients previously exposed to a deep sequencing-based Sentosa SQ HCV Genotyping Assay. direct-acting antiviral (DAA) containing regimen. Cur- Results: A high concordance rate with Sanger sequencing, rently, no standardized assays have been developed as the reference method, was found for the NS3, NS5A and purchasable kits for HCV resistance testing. The aim of this NS5 coding regions, regardless of the genotype-1 subtypes. study was to evaluate the performance of the Sentosa SQ The deep sequencing-based assay was more sensitive than HCV Genotyping Assay, a novel deep sequencing-based population sequencing to detect minority variants, repre- assay, to identify resistance-associated substitutions senting less than 10% of the viral populations, but also (RASs) in the NS3 protease, NS5A protein domain I and some variants representing up to 30% of the viral qua- NS5B polymerase regions for patients infected with HCV sispecies, as expected. genotypes-1a and 1b. Conclusions: The Sentosa SQ HCV Genotyping Assay can be Methods: Serum samples collected from patients with confidently used in clinical practice in the indications of HCV chronic hepatitis C infection who failed to achieve a sus- resistance testing for these subtypes. Technical improve- tained virological response after receiving a DAA-containing ments are now required to allow for pangenotypic coverage. Introduction HCV infects approximately 71 million people world- study, the incidence of HCV DAA treatment failures wide. Patients with chronic HCV infection are at a high was of the order of 6% globally [6]. Most DAA-based risk of developing cirrhosis and hepatocellular car- treatment failures are relapses, and they occur more cinoma [1]. An estimated 700,000 annual deaths are frequently when one or several of the following param- attributable to chronic HCV infection [2,3]. The devel- eters are present: treatment-experienced patients, cir- opment and approval of new combinations of direct- rhosis, infection with HCV genotype-1a or 3a, presence acting antiviral (DAA) drugs yielded HCV infection of the Q80K polymorphism in the NS3 protease region, cure rates over 95% globally in both large-scale clinical presence of resistance-associated substitutions  (RASs) trials and the real-world setting when these combina- in the NS5A region at baseline, and/or short treatment tions were used according to international liver society duration [7]. guidelines [4,5]. Despite the high rates of virological The viral populations that constitute the HCV qua- cure achieved with these treatments, HCV infection is sispecies differ by amino acid polymorphisms that not eliminated upon first-line therapy in up to 10% of emerge by mutation during replication and are sub- cases, depending on the patient’s profile and the admin- sequently selected based on their effects on viral fit- istered regimen. In the French HEPATHER cohort ness [8]. Natural polymorphisms may confer reduced ©2019 International Medical Press 1359-6535 (print) 2040-2058 (online) 417 AVT-19-OA-4532_Rodriguez.indd 417 AVT-19-OA-4532_Rodriguez.indd 417 22/11/2019 11:37:13 22/11/2019 11:37:13 C Rodriguez et al. susceptibility to a given DAA or, more frequently, to capacity to identify RASs in the NS3 protease (nucleo- a DAA class. As a result, DAA administration selects tide position 1–798 according to the H77-1a prototype viral strains carrying RASs on their genome which strain), NS5A (nucleotide position 39–600) and NS5B bear reduced susceptibility to the drugs administered, polymerase (nucleotide position 990–1677) regions for thereby contributing to treatment failure. Some interna- patients infected with HCV genotypes-1a and 1b. tional liver society guidelines recommended to perform The aim of the present study was to evaluate the per- HCV resistance testing at baseline of first-line therapy formance of the Sentosa SQ HCV Genotyping Assay with certain combination regimens, in order to tailor for the determination of HCV resistance profiles in treatment in patients harbouring RASs in the NS5A patients infected with HCV genotype-1a or 1b who region [9,10]. However, the most recent version of the failed to achieve a sustained virological response (SVR) European Association for the Study of the Liver (EASL) after receiving a DAA-containing treatment regimen. guidelines does not recommend using HCV resistance testing prior to first-line therapy, because of the high Methods barrier to resistance and global efficacy of new pangen- Clinical specimens otypic regimens [11]. Liver society guidelines all rec- ommend performing HCV resistance testing covering Nineteen serum samples collected from patients with the NS3 protease, NS5A and NS5B polymerase regions chronic hepatitis C infected with HCV genotype-1a prior to retreatment in patients previously exposed to a (n=14) or 1b (n=5), followed in the Department of DAA-containing regimen [10,11]. Hepatology of the Henri Mondor University Hospital, DNA sequence analysis has been based for many were studied. All patients had been previously treated years on so-called ‘population’ sequencing (that is, by a DAA combination and failed to achieve SVR. The direct sequence analysis of a PCR product) by means samples were taken at retreatment baseline for HCV of the Sanger method [12]. Population sequencing resistance testing, frozen and stored at -80°C until test- detects 20% to 30% of the viral populations present ing. Serum HCV RNA levels were measured by means in the quasispecies [13]. Thus, minor populations are of a real-time PCR assay (Abbott RealTime HCV Assay, not detected by this method. They can be identified Abbott Molecular, Des Plaines, IL, USA) [15]. The HCV by clonal sequence analysis, but this approach is time- genotype and subtype were determined by means of our consuming and cumbersome and cannot be applied to in-house population sequencing technique targeting the clinical practice. Thus, more sensitive sequencing meth- NS5B gene followed by phylogenetic analysis, the ref- ods were required. Deep sequencing methods now offer erence method for HCV genotype determination [16]. new opportunities for the detection of minor viral pop- ulations present at frequencies as low as 1%. Study design Currently, no standardized assays have been devel- The presence of HCV RASs in the three regions of oped as purchasable kits for HCV resistance testing. interest was assessed in all samples at baseline of Resistance testing therefore relies on homemade tech- retreatment by two methods. The reference method niques based on population or deep sequencing. A was population sequencing of the NS3 protease, NS5A limited number of laboratories has made these tests protein domain I and NS5B polymerase coding regions available in Europe and in other regions, and the perfor- by means of our in-house techniques [17]. The results mance of their assays has not been externally validated. were compared with those generated by the deep The Sentosa SQ HCV Genotyping Assay (Vela Diag- sequencing-based Sentosa SQ HCV Genotyping Assay. nostics GmbH, Hamburg, Germany) is a novel, fully The study was conducted in accordance with the Inter- automated deep sequencing-based assay comprising a national Conference on Harmonization guidelines, customized version of the epMotion 5075 robotic liq- applicable regulations, and the principles of the Decla- uid handling system for RNA extraction and sequence ration of Helsinki. All patients gave written informed library preparation (Sentosa SX101), a customized ver- consent for the use of leftover specimens. sion of an Ion One Touch device for template prepa- ration including emulsion PCR (Sentosa ST401), the HCV resistance testing by means of in-house Ion Torrent technology for deep sequencing (Sentosa population sequencing methods SQ301), and software for data analysis and reporting Briefly, total HCV RNA was extracted from 400 µl of (Sentosa Link and Sentosa Reporter, respectively). This serum by means of QIAsymphony DSP Virus/Pathogen assay has received CE-IVD mark in Europe for HCV kit (Qiagen GmbH, Hilden, Germany), according to the genotype determination and resistance testing in clinical manufacturer’s instructions. The RNA pellet was eluted practice. We recently reported on the ability of this assay with 60 µl of RNAse-free water with 0.04% NaN . to correctly identify the HCV genotype and subtype Complementary DNA synthesis was performed with [14]. The current generation of the assay also has the the OneStep RT-PCR kit (Qiagen GmbH) with sets of 418 ©2019 International Medical Press AVT-19-OA-4532_Rodriguez.indd 418 AVT-19-OA-4532_Rodriguez.indd 418 22/11/2019 11:37:13 22/11/2019 11:37:13 Deep sequencing platform for HCV resistance testing primers adapted to the viral regions targeted (Table 1). A means of the commercial web-based software NGS nested PCR was used to amplify NS3, NS5A and NS5B HCV Module (SmartGene, Lausanne, Switzerland). coding DNA fragments. PCR products were purified by means of NucleoFast 96 PCR plate kit (Macherey- Sequence data Nagel GmbH & Co. KG, Düren, Germany) and directly The GenBank/ENA/DDBJ accession numbers of the sequenced by means of the BigDye Terminator Cycle sequences reported in this paper are PRJNA397404. v3.1 sequencing kit (ThermoFisher Scientific, Courta- boeuf, France) on an ABI 3100 sequencer (Applied Results Biosystems, Foster City, CA, USA), according to the Characteristics of the study population manufacturer’s instructions. Nucleotide sequences were manually corrected and aligned to genotype-1a Nineteen patients previously exposed to different DAA- or 1b reference sequences, respectively, and amino acid containing regimens were included. Most of them changes were deduced from the nucleotide sequences. (78.9%) were males, with a mean age of 57.8 years (Table 2). The patients were mainly infected with geno- HCV resistance testing by means of Sentosa SQ HCV type-1a (73.7%) and 12 of them (63.2%) had a baseline Genotyping Assay HCV RNA level >800,000 IU/ml. Of the 19 patients, Briefly, nucleic acid extraction was performed from 13 (68.4%) had compensated cirrhosis and 4 had 530 µl of serum on the Sentosa SX101 robotic instru- advanced fibrosis (F3 in the METAVIR scoring system). ment using Sentosa Virus Total Nucleic Acid Plus II kit Prior DAA-containing regimens included: pegylated (Vela Diagnostics). The NS3, NS5A and NS5B coding interferon (IFN) with ribavirin plus daclatasvir (n=8); regions were RT-PCR-amplified by means of Veriti Dx pegylated IFN plus daclatasvir and asunaprevir (n=3); 96-Well Thermal Cycler (Applied Biosystems). After sofosbuvir and ribavirin (n=1); sofosbuvir and sime- purification of PCR products using magnetic beads, a previr (n=2); sofosbuvir and ledipasvir with or with- 200-nucleotide fragment library was prepared on Sen- out ribavirin (n=2); daclatasvir and asunaprevir (n=1); tosa SX101. The samples were barcoded by ligation, daclatasvir and simeprevir (n=1); and ritonavir-boosted pooled into a single tube and amplified by emulsion PCR paritaprevir, ombitasvir and dasabuvir with ribavirin on Sentosa ST401i. Deep sequencing was performed by (n=1). Fourteen patients were retreated, according to means of the Sentosa SQ Sequencing Kit on the Sentosa the European recommendations available at the time, SQ301 Sequencer, based on Ion Torrent technology. with the combination of sofosbuvir and simeprevir Primary data analysis was automatically performed for 12 weeks (n=11), sofosbuvir plus ledipasvir with using Sentosa SQ Reporter software. Assembled NS3 ribavirin for 24 weeks (n=1) or sofosbuvir, simeprevir contigs (944-base pair fragment), NS5A contigs (604- and daclatasvir with ribavirin for 24 weeks (n=2). An base pair fragment) and NS5B contigs (685-base pair SVR was achieved in 10 patients (71.4%). The remain- fragment) were aligned with genotype-1a and 1b refer- ing four patients failed, including three patients who ence sequences, respectively. The raw sequencing data received the combination of sofosbuvir and sime- generated by the Sentosa assay were also analysed by previr for 12 weeks and one patient treated with the Table 1. Primers used for nested PCR amplification prior to population sequencing of the HCV NS3 protease, NS5A domain I and NS5B polymerase genes, according to the HCV genotype (1a or 1b) HCV genotype Primer 5′-3′ sequence Gene Nucleotide position References 1 NS3-G1F1 ATG GAR AAG AAR RTY ATY RTN TGG G NS3 3278–3300 Personal 1 NS3-G1F2 ATG GAR AYY AAG CTY ATY ACN TGG G NS3 3278–3300 Personal 1 NS3-G1R CTY TTN CCR CTN CCN GTN GGN GCR TG NS3 4025–4048 Personal 1a NS3-1a-2s CCG ATG GAA TGG TCT CCA AGG NS3 3384–3405 [23] 1a NS3-1a-2a GAG AGG AGT TGT CCG TGA ACA C NS3 3965–3984 [23] 1b NS3-1b-1s GGC GTG TGG GGA CAT CAT C NS3 3316–3332 [24] 1b NS3-1b-1a GGA GAT GAG TTG TCT GTG AA NS3 3968–3985 Personal 1 NS5A-G1xF GGA TGA ACCGGC TSA TAG C NS5A 6084–6103 Personal 1 NS5A-G1xF1 ACC CAG CTS CTG ARR AGG C NS5A 6200–6219 Personal 1 NS5A-G1xR ACG TAR TGG AAR TCC CCC AC NS5A 6626–6646 Personal 1 Sn755 TAT GAY ACC CGC TGY TTT GAC TC NS5B 8255–8278 [25] 1 5B-SI766 CTG YTT TGA CTC CAC NGT RAC NS5B 8266–8287 [26] 1 GEN1A.R1 CCG GGC AYG AGA CAC GCT GTG ATA AAT G NS5B 9277–9305 [27] 1 GEN1B.R1 TGC GGC ACG AGA CAV GCT GTG ATA TG NS5B 9279–9305 [27] Antiviral Therapy 24.6 419 AVT-19-OA-4532_Rodriguez.indd 419 AVT-19-OA-4532_Rodriguez.indd 419 22/11/2019 11:37:13 22/11/2019 11:37:13 C Rodriguez et al. Table 2. Demographic and virological features of the 19 NS5A RASs were found at retreatment baseline in patients included 17 of the 19 patients (89.5%). All of the NS5A RASs found (at positions 24, 28, 30, 31, 58 and 93) had been Parameter Value (n=19) previously described to be associated with NS5A inhib- itor-containing regimen failures in vivo. The RASs were Mean age, years ±ds (range) 57.8 ±9.6 present in single (n=6), double (n=8) or triple (n=3) (46–81) mutants. The most frequent amino acid substitutions Male, n (%) 15 (78.9) were at positions 31 (L31M in six patients and L31V Mean HCV RNA level, log IU/ml ±ds (range) 6.0 ±0.5 (5.2–6.7) in two patients), 30 (Q30R in two patients, Q30K in HCV RNA level >800,000 UI/ml, n (%) 12 (63.2) two patients, Q30E in two patients and Q30H in one HCV genotype patient) and 93 (Y93H in five patients and Y93C in Genotype 1a, n (%) 14 (73.7) two patients; Table 4). All of these RASs confer vari- Genotype 1b, n (%) 5 (26.3) ous levels of reduced susceptibility to first- and second- Fibrosis stage, n (%) 19 (100) generation NS5A inhibitors [18]. Mild fibrosis (Fibroscan ≤ 7.0), n (%) 2 (10.5) NS5B RASs, including L159F and C316N, were Moderate fibrosis (Fibroscan > 7.0 and ≤9.5 kPa), n (%) 0 (0) detected in only one patient infected with genotype-1b Advanced fibrosis (Fibroscan > 9.5 and ≤12.5 kPa), n (%) 4 (21.1) who had been previously exposed to a combination of Compensated cirrhosis (Fibroscan >12.5), n (%) 13 (68.4) sofosbuvir and ledipasvir (Table 5). Both substitutions Treatment-experienced, n (%) 19 (100) have been reported to be associated with reduced sus- Prior DAA-containing regimen Pegylated IFN/ribavirin + daclatasvir, n 8 ceptibility to sofosbuvir [19]. Pegylated IFN/ribavirin + daclatasvir + asunaprevir, n 3 Sofosbuvir with ribavirin, n 1 NS3 protease RASs found by Sentosa SQ HCV Sofosbuvir + simeprevir, n 2 Genotyping Assay Sofosbuvir/ledipasvir ± ribavirin, n 2 NS3 RASs were found at baseline of retreatment in 12 Daclatasvir + asunaprevir, n 1 of the 19 patients (63.2%) with the Sentosa assay, using Daclatasvir + simeprevir, n 1 the internal assay software for interpretation. All of the Paritaprevir/ritonavir/ombitasvir + dasabuvir with 1 RASs found with population sequencing except three ribavirin, n were detected by Sentosa HCV SQ Genotyping Assay in high proportions (>80%). The three RASs not identified DAA, direct-acting antiviral; IFN, interferon. by the assay were V55I, Y56F and S122T, which confer reduced susceptibility to asunaprevir, grazoprevir and combination of sofosbuvir, simeprevir and daclatasvir simeprevir, respectively. plus ribavirin for 24 weeks who discontinued treat- Some RASs were detected by the deep sequencing- ment prematurely owing to a severe pulmonary arterial based Sentosa assay and not with our in-house popu- hypertension. lation sequencing method, including V36A (n=1) and V36M (n=1), Q80R (n=1), D168E (n=1), D168V HCV resistance testing by means of in-house (n=1) and D168Y (n=1), and V170A (n=1). In all but population sequencing methods three cases, they were present as minor viral popula- Our in-house population sequencing techniques were tions, representing 1.1% to 7.1% of the viral qua- used for NS3 protease, NS5A domain I and NS5B resist- sispecies (Table 3). The remaining RASs (D168E/V/Y) ance testing. NS3 RASs were found at retreatment base- represented between 15.9% to 33% of the viral qua- line in 11 of the 19 patients (57.9%), including 5 who sispecies, as reported by the Sentosa assay software had never been exposed to any protease inhibitor. All (Table 3). of the amino acid substitutions found (at positions 55, 56, 80, 122, 155 and 168) had been previously reported NS5A RASs found by Sentosa SQ HCV Genotyping to be associated with protease inhibitor-containing Assay regimen failures in vivo [18]. The RASs were present NS5A RASs were found at retreatment baseline in in single (n=9), double (n=1) or triple (n=1) mutants. 18 of the 19 patients (94.7%). All of the NS5A RASs The most frequent amino acid substitutions were at detected by population sequencing except two were positions 80 (Q80K in three patients and Q80L in one also identified by the Sentosa HCV SQ Genotyping patient) and 155 (R155K in four patients; Table  3). Assay in high proportions (>50%). The two RASs RASs at both positions are known to confer resistance not detected by the Sentosa assay included L28M and to simeprevir in vitro and in vivo, and reduced suscepti- P58S, which both confer reduced susceptibility to first- bility to second-generation protease inhibitors, includ- generation NS5A inhibitors in genotype-1b-infected ing grazoprevir and voxilaprevir [18]. patients. 420 ©2019 International Medical Press AVT-19-OA-4532_Rodriguez.indd 420 AVT-19-OA-4532_Rodriguez.indd 420 22/11/2019 11:37:13 22/11/2019 11:37:13 Deep sequencing platform for HCV resistance testing Table 3. RASs in the NS3 region found by in-house population Table 4. RASs in the NS5A domain I region found by in-house sequencing and Sentosa SQ HCV Genotyping Assay, population sequencing and Sentosa SQ HCV Genotyping Assay, respectively respectively In-house In-house Deep sequencing-based population Deep sequencing-based Sentosa population Sentosa SQ HCV Genotyping sequencing SQ HCV Genotyping Assay (n=19; sequencing (n=19; frequency NS5A RAS (n=19) frequency in the HCV quasispecies) NS3 protease RAS (n=19) in the HCV quasispecies) K24 V36 Q 2 – A – 1 (1.6%) M/L28 M – 1 (7.1%) T 2 – L 1 1 (99.8%) V – 3 (2%; 6.1%; 1.5%) V55 Q/R30 A 1 1 (97.2%) E 3 4 (21.2%; 72%; 42.3%; 72.9%) Y56 H 1 2 (2.8%; 98.6%) F 1 – R 2 5 (99.8%; 20.2%; 13.4%; 37.1%; 84.4%) Q80 L31 K 3 3 (98.7%; 97.1%; 98.3%) M 6 8 (99.6%; 99.7%; 98.8%; 99.5%; 19.3%; L 1 1 (99.9%) 4.5%; 99.5%; 99.3%) R – 1 (1.1%) V 2 2 (99.4%; 99.6%) S122 H/P58 T 1 – P 2 – R155 S 1 – K 4 4 (98.4%; 78.8%; 79.6%; 95.5%) Y93 D168 C 2 3 (99.6%; 3.6%; 54.2%) E 1 2 (15.9%; 94.9%) H 5 5 (98.7%; 99.7%; 93.1%; 99.5%; 99.5%) Y – 1 (33%) V 1 2 (16.7%; 99.7%) NS5A resistance-associated substitutions (RASs) reported to be associated with V170 resistance to NS3 protease inhibitors in patients infected with genotypes-1a or A – 1 (4.3%) 1b failing a direct-acting antiviral containing regimen [18]. NS3 resistance-associated substitutions (RASs) reported to be associated with resistance to NS3 protease inhibitors in patients infected with genotypes-1a or web-based NGS HCV Module software developed by 1b failing a direct-acting antiviral containing regimen [18]. SmartGene. Using this method, three additional RASs were identified as compared with using the Sentosa SQ Several RASs were detected by the deep sequencing- Reporter internal assay software, including a RAS at based Sentosa assay, but not by our in-house popula- position 122 in the NS3-coding region and RASs at tion sequencing method. They included M28V (n=3), positions 28 and 58 in the NS5A-coding region. Q30E (n=1), Q30R (n=3) and Q30H (n=1), L31M (n=2) and Y93C (n=1), which were generally present Discussion as minor viral populations, representing 1.5% to 6.1% of the viral quasipecies. Five NS5A RASs (L31M and Treatment failure with DAA-based regimens is often Q30E/R) present as more than 10% of the viral qua- associated with the selection of viral variants with sispecies in the Sentosa assay were not detected by reduced susceptibility to the administered drug(s) and, Sanger sequencing (Table 4). most often, to other drugs from the same class(es). HCV resistance testing may be useful to guide treat- NS5B polymerase RASs found by Sentosa SQ HCV ment decisions with specific regimens in particular Genotyping Assay patient populations, when the presence of pre-existing NS5B RASs at retreatment baseline were not detected RASs in certain proportions has an impact on treat- in any patient using the Sentosa HCV SQ Genotyping ment outcomes. A number of factors should be taken Assay (Table 5). into consideration, such as the availability of resist- ance testing in clinical practice, the availability of HCV resistance testing by means of Sentosa SQ HCV alternative treatment options, and the consequences of Genotyping Assay plus sequence analysis by means of treatment failure associated with resistance. Because the SmartGene NGS HCV Module treatment options that do not need to be tailored to Raw sequencing data from the Sentosa HCV SQ the baseline RAS profile are available, it is not justi- Genotyping Assay were reanalysed by means of the fied to recommend systematic HCV resistance testing Antiviral Therapy 24.6 421 AVT-19-OA-4532_Rodriguez.indd 421 AVT-19-OA-4532_Rodriguez.indd 421 22/11/2019 11:37:13 22/11/2019 11:37:13 C Rodriguez et al. Table 5. RASs in the NS5B polymerase region found by with the estimated sensitivity of both methods (1% for in-house population sequencing and Sentosa SQ HCV deep sequencing, 20–30% for population sequencing). Genotyping Assay, respectively Discrepant results could be explained for at least NS3 RASs by the time between failure and HCV resistance Deep sequencing-based Sentosa testing. As suggested NS3 RASs are replaced by the wild In-house population SQ HCV Genotyping Assay (n=19; NS5B RAS sequencing (n=19) frequency in the HCV quasispecies) type within 24 to 48 weeks [20,21] Despite the relatively small number of patients stud- L159 ied, our study suggests that the deep sequencing-based F 1 – Sentosa SQ HCV Genotyping Assay can be confidently C316 used to characterize the HCV resistance profile prior N 1 – to retreating patients who failed a DAA-based regimen. The study included only genotype-1a- and 1b-infected NS5B resistance-associated substitutions (RASs) reported to be associated with patients, the most prevalent subtypes in industrialized resistance to NS3 protease inhibitors in patients infected with genotypes-1a or 1b failing a direct-acting antiviral containing regimen [18]. countries, because the Sentosa SQ HCV Genotyping Assay has been designed so far only for these subtypes. However, the assay has the capacity to identify RASs prior to first-line therapy, as stated in the EASL Rec- for all genotypes and a second version of the assay ommendations on Treatment of Hepatitis C 2018 [11]. with improved genotype coverage and with the ability In contrast, HCV resistance testing is useful to guide to detect RASs that are associated with resistance to retreatment decisions in patients who have previously sofosbuvir is in development. failed an IFN-free DAA-based regimen. Nevertheless, Our study has limitations. First, it was performed no standardized commercial assay is yet available for with a relatively small number of patients. Secondly, HCV resistance testing, which still relies on homemade the performance of this assay in non 1a- or 1b-infected methods using population or deep sequencing. patients was not tested. Sentosa SQ HCV Genotyping Assay, the first avail- In conclusion, the present study assessing the clinical able kit for HCV genotyping and resistance testing, is performance of HCV resistance testing with the new deep a new automated deep sequencing-based assay that is sequencing-based assay Sentosa SQ HCV Genotyping standardized and easy-to-use without intensive training Assay showed excellent concordance with population or specialized skills. In the present study, we assessed sequencing in patients infected with genotypes-1a and the ability of this assay to characterize the HCV resist- 1b, suggesting that this assay can be confidently used ance profile in clinical samples from patients infected in clinical practice in the indications of HCV resistance with genotype-1a or 1b who failed a prior DAA-based testing for these subtypes. Technical improvements are treatment. Resistance profiles were in keeping with the now required to allow for pangenotypic coverage and exposure to different DAA-containing regimens, except enrichment of the database with reference prototype in some failing patients who had never been exposed to strains for determination of amino acid changes, as pre- one or more drug classes. viously reported for HIV resistance [22]. Our results show good concordance with popula- tion sequencing in most cases, especially when the web- Acknowledgements based SmartGene software NGS HCV Module was used to analyse the assay’s raw sequence data. Our data We would like to thank VELA Diagnostics GmbH, suggest that the Sentosa SQ Reporter software used for Hamburg, Germany for providing the Sentosa SQ HCV bioanalysis of raw sequence data needs to be improved, Genotyping Assay kits. in particular its database of prototype reference strain This work was supported by internal funding. sequences used for alignment. Interestingly, the design of the assay did not allow it to correctly identify RASs Disclosure statement in the NS5B-coding region, especially those able to con- fer reduced susceptibility to sofosbuvir (RASs at posi- JMP acted as advisors for Abbott and SC acted as advi- tions 159, 282 and 316) because these mutations are sors for Abbott and Cepheid. The remaining authors localized outside the target region, which is amplified declare no competing interests. by the assay. As expected, the deep sequencing-based assay was References more sensitive than population sequencing to detect 1. Polaris Observatory HCV Collaborators. Global prevalence minority variants, representing less than 10% of the and genotype distribution of hepatitis C virus infection in viral populations, but also some variants representing 2015: a modelling study. Lancet Gastroenterol Hepatol 2017; 2:161–176. up to 30% of the viral quasispecies. This is in keeping 422 ©2019 International Medical Press AVT-19-OA-4532_Rodriguez.indd 422 AVT-19-OA-4532_Rodriguez.indd 422 22/11/2019 11:37:13 22/11/2019 11:37:13 Deep sequencing platform for HCV resistance testing 2. Global Burden of Disease Cancer Collaboration, 16. Simmonds P, Smith DB, McOmish F, et al. Identification of Fitzmaurice C, Dicker D, et al. The global burden of cancer genotypes of hepatitis C virus by sequence comparisons in the 2013. JAMA Oncol 2015; 1:505–527. core, E1 and NS-5 regions. J Gen Virol 1994; 75:1053–1061. 3. GBD 2013 Mortality and Causes of Death Collaborators. 17. Hézode C, Chevaliez S, Scoazec G, et al. Retreatment with Global, regional, and national age-sex specific all-cause and sofosbuvir and simeprevir of patients with hepatitis C cause-specific mortality for 240 causes of death, 1990-2013: virus genotype 1 or 4 who previously failed a daclatasvir- a systematic analysis for the Global Burden of Disease Study containing regimen. Hepatology 2016; 63:1809–1816. 2013. Lancet 2015; 385:117–171. 18. Sorbo MC, Cento V, Di Maio VC, et al. Hepatitis C virus 4. Belperio PS, Chartier M, Ross DB, Alaigh P, Shulkin D. drug resistance associated substitutions and their clinical Curing hepatitis C virus infection: best practices from the relevance: update 2018. Drug Resist Updat 2018; 37:17–39. U.S. Department of Veterans Affairs. Ann Intern Med 2017; 19. Donaldson EF, Harrington PR, O’Rear JJ, Naeger LK. 167:499–504. Clinical evidence and bioinformatics characterization 5. Falade-Nwulia O, Suarez-Cuervo C, Nelson DR, Fried MW, of potential hepatitis C virus resistance pathways for Segal JB, Sulkowski MS. Oral direct-acting agent therapy sofosbuvir. Hepatology 2015; 61:56–65. for hepatitis C virus infection: a systematic review. Ann 20. Lenz O, Verbinnen T, Fevery B, et al. Virology analysis of Intern Med 2017; 166:637–648. HCV isolates from genotype 1-infected patients treated 6. Chevaliez S, Trimoulet P, Dorival C, et al. Effect of with simeprevir plus peginterferon/ribavirin in Phases IIb/III resistance-associated substitutions on retreatment of direct studies. J Hepatol 2015; 62:1008–1014. acting antiviral exposed patients in the real-world setting 21. McPhee F, Hernandez D, Zhou N, et al. Pooled analysis (ANRS CO22 HEPATHER). J Hepatol 2017; 66:S84. of HCV genotype 1 resistance-associated substitutions in 7. Buti M, Riveiro-Barciela M, Esteban R. Management NS5A, NS3 and NS5B pre-and post-treatment with 12 of direct-acting antiviral agent failures. J Hepatol 2015; weeks of daclatasvir, asunaprevir and beclabuvir. Antivir 63:1511–1522. Ther 2018; 23:53–66. 8. Pawlotsky JM, Hepatitis C. Drugs: is next generation the 22. Raymond S, Nicot F, Carcenac R, et al. HIV-1 genotypic last generation? Gastroenterology 2016; 151:587–590. resistance testing using the Vela automated next-generation sequencing platform. J Antimicrob Chemother 2018; 9. European Association for the Study of the Liver. EASL 73:1152–1157. Recommendations on Treatment of Hepatitis C 2016. J Hepatol 2017; 66:153–194. 23. Lawitz E, Sulkowski MS, Ghalib R, et al. Simeprevir plus sofosbuvir, with or without ribavirin, to treat chronic 10. AASLD/IDSA HCV Guidance Panel. Hepatitis C guidance: infection with hepatitis C virus genotype 1 in non- AASLD-IDSA recommendations for testing, managing, and responders to pegylated interferon and ribavirin and treating adults infected with hepatitis C virus. Hepatology treatment-naive patients: the COSMOS randomized study. 2015; 62:932–954. Lancet 2014; 384:1756–1765. 11. European Association for the Study of the Liver. EASL 24. Dieterich D, Bruce BR, Flamm SL, et al. Evaluation of Recommendations on Treatment of Hepatitis C 2018. sofosbuvir and simeprevir-based regimens in the TRIO J Hepatol 2018; 69:461–511. network: academic and community treatment of a real- 12. Sanger F, Air GM, Barrell BG, et al. Nucleotide sequence of world, heterogeneous population. Hepatology 2014; 60 bacteriophage phi X174 DNA. Nature 1977; 265:687–695. Suppl 1:220A. 13. Palmer S, Kearney M, Maldarelli F, et al. Multiple, linked 25. Lenz O, Fevery B, Vijgen L, et al. TMC-435 in patients human immunodeficiency virus type 1 drug resistance infected with HCV genotype 1 who failed previous mutations in treatment-experienced patients are missed pegylated interferon/ribavirin treatment: virologic analyses by standard genotype analysis. J Clin Microbiol 2005; of the ASPIRE trial. J Hepatol 2012; 56 Suppl 2:S5. 43:406–413. 26. Sarrazin C, Lathouwers E, Peeters M, et al. Prevalence of 14. Rodriguez C, Soulier A, Demontant V, et al. A novel the hepatitis C virus NS3 polymorphism Q80K in genotype standardized deep sequencing-based assay for hepatitis C 1 patients in the European region. Antiviral Res 2015; virus genotype determination. Sci Rep 2018; 8:4180. 116:10–16. 15. Chevaliez S, Bouvier-Alias M, Pawlotsky JM. Performance 27. Bourlière M, Sulkowski MS, Omata M, et al. An integrated of the Abbott real-time PCR assay using m2000sp and safety and efficacy analysis of >500 patients with m2000rt for hepatitis C virus RNA quantification. J Clin compensated cirrhosis treated with ledipasvir/sofosbuvir Microbiol 2009; 47:1726–1732. with or without ribavirin. Hepatology 2014; 60 Suppl 1:23. Accepted 16 May 2019; published online 21 May 2019 Antiviral Therapy 24.6 423 AVT-19-OA-4532_Rodriguez.indd 423 AVT-19-OA-4532_Rodriguez.indd 423 22/11/2019 11:37:13 22/11/2019 11:37:13 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Antiviral Therapy SAGE

Performance Assessment of a Fully Automated Deep Sequencing Platform for HCV Resistance Testing

Download PDF
7 pages

Loading next page...
 
/lp/sage/performance-assessment-of-a-fully-automated-deep-sequencing-platform-bRdkUBCHCq

References (26)

Publisher
SAGE
Copyright
© 2019 SAGE Publications
ISSN
1359-6535
eISSN
2040-2058
DOI
10.3851/imp3318
Publisher site
See Article on Publisher Site

Abstract

Antiviral Therapy 2019; 24:417–423 (doi: 10.3851/IMP3318) Original article Performance assessment of a fully automated deep sequencing platform for HCV resistance testing 1,2 1,2 1,2 1,2 1,2 Christophe Rodriguez , Mélanie Mercier-Darty , Alexandre Soulier , Lila Poiteau , Mélanie Wlassow , 1,2 1,3 1,2† 1,2† Slim Fourati , Christophe Hézode , Jean-Michel Pawlotsky , Stéphane Chevaliez * Department of Virology, National Reference Center for Viral Hepatitis B, C and Delta, Hôpital Henri Mondor, Université Paris-Est, Créteil, France INSERM U955, Créteil, France Department of Hepatology, Hôpital Henri Mondor, Université Paris-Est, Créteil, France These authors jointly supervised this work *Corresponding author e-mail: stephane.chevaliez@aphp.fr Background: International liver society guidelines recom- treatment regimen were extracted and sequenced by two mended to perform HCV resistance testing at baseline of methods including population sequencing of the NS3, first-line therapy with certain combination regimens or NS5A and NS5B coding region reference method and the prior to retreatment in patients previously exposed to a deep sequencing-based Sentosa SQ HCV Genotyping Assay. direct-acting antiviral (DAA) containing regimen. Cur- Results: A high concordance rate with Sanger sequencing, rently, no standardized assays have been developed as the reference method, was found for the NS3, NS5A and purchasable kits for HCV resistance testing. The aim of this NS5 coding regions, regardless of the genotype-1 subtypes. study was to evaluate the performance of the Sentosa SQ The deep sequencing-based assay was more sensitive than HCV Genotyping Assay, a novel deep sequencing-based population sequencing to detect minority variants, repre- assay, to identify resistance-associated substitutions senting less than 10% of the viral populations, but also (RASs) in the NS3 protease, NS5A protein domain I and some variants representing up to 30% of the viral qua- NS5B polymerase regions for patients infected with HCV sispecies, as expected. genotypes-1a and 1b. Conclusions: The Sentosa SQ HCV Genotyping Assay can be Methods: Serum samples collected from patients with confidently used in clinical practice in the indications of HCV chronic hepatitis C infection who failed to achieve a sus- resistance testing for these subtypes. Technical improve- tained virological response after receiving a DAA-containing ments are now required to allow for pangenotypic coverage. Introduction HCV infects approximately 71 million people world- study, the incidence of HCV DAA treatment failures wide. Patients with chronic HCV infection are at a high was of the order of 6% globally [6]. Most DAA-based risk of developing cirrhosis and hepatocellular car- treatment failures are relapses, and they occur more cinoma [1]. An estimated 700,000 annual deaths are frequently when one or several of the following param- attributable to chronic HCV infection [2,3]. The devel- eters are present: treatment-experienced patients, cir- opment and approval of new combinations of direct- rhosis, infection with HCV genotype-1a or 3a, presence acting antiviral (DAA) drugs yielded HCV infection of the Q80K polymorphism in the NS3 protease region, cure rates over 95% globally in both large-scale clinical presence of resistance-associated substitutions  (RASs) trials and the real-world setting when these combina- in the NS5A region at baseline, and/or short treatment tions were used according to international liver society duration [7]. guidelines [4,5]. Despite the high rates of virological The viral populations that constitute the HCV qua- cure achieved with these treatments, HCV infection is sispecies differ by amino acid polymorphisms that not eliminated upon first-line therapy in up to 10% of emerge by mutation during replication and are sub- cases, depending on the patient’s profile and the admin- sequently selected based on their effects on viral fit- istered regimen. In the French HEPATHER cohort ness [8]. Natural polymorphisms may confer reduced ©2019 International Medical Press 1359-6535 (print) 2040-2058 (online) 417 AVT-19-OA-4532_Rodriguez.indd 417 AVT-19-OA-4532_Rodriguez.indd 417 22/11/2019 11:37:13 22/11/2019 11:37:13 C Rodriguez et al. susceptibility to a given DAA or, more frequently, to capacity to identify RASs in the NS3 protease (nucleo- a DAA class. As a result, DAA administration selects tide position 1–798 according to the H77-1a prototype viral strains carrying RASs on their genome which strain), NS5A (nucleotide position 39–600) and NS5B bear reduced susceptibility to the drugs administered, polymerase (nucleotide position 990–1677) regions for thereby contributing to treatment failure. Some interna- patients infected with HCV genotypes-1a and 1b. tional liver society guidelines recommended to perform The aim of the present study was to evaluate the per- HCV resistance testing at baseline of first-line therapy formance of the Sentosa SQ HCV Genotyping Assay with certain combination regimens, in order to tailor for the determination of HCV resistance profiles in treatment in patients harbouring RASs in the NS5A patients infected with HCV genotype-1a or 1b who region [9,10]. However, the most recent version of the failed to achieve a sustained virological response (SVR) European Association for the Study of the Liver (EASL) after receiving a DAA-containing treatment regimen. guidelines does not recommend using HCV resistance testing prior to first-line therapy, because of the high Methods barrier to resistance and global efficacy of new pangen- Clinical specimens otypic regimens [11]. Liver society guidelines all rec- ommend performing HCV resistance testing covering Nineteen serum samples collected from patients with the NS3 protease, NS5A and NS5B polymerase regions chronic hepatitis C infected with HCV genotype-1a prior to retreatment in patients previously exposed to a (n=14) or 1b (n=5), followed in the Department of DAA-containing regimen [10,11]. Hepatology of the Henri Mondor University Hospital, DNA sequence analysis has been based for many were studied. All patients had been previously treated years on so-called ‘population’ sequencing (that is, by a DAA combination and failed to achieve SVR. The direct sequence analysis of a PCR product) by means samples were taken at retreatment baseline for HCV of the Sanger method [12]. Population sequencing resistance testing, frozen and stored at -80°C until test- detects 20% to 30% of the viral populations present ing. Serum HCV RNA levels were measured by means in the quasispecies [13]. Thus, minor populations are of a real-time PCR assay (Abbott RealTime HCV Assay, not detected by this method. They can be identified Abbott Molecular, Des Plaines, IL, USA) [15]. The HCV by clonal sequence analysis, but this approach is time- genotype and subtype were determined by means of our consuming and cumbersome and cannot be applied to in-house population sequencing technique targeting the clinical practice. Thus, more sensitive sequencing meth- NS5B gene followed by phylogenetic analysis, the ref- ods were required. Deep sequencing methods now offer erence method for HCV genotype determination [16]. new opportunities for the detection of minor viral pop- ulations present at frequencies as low as 1%. Study design Currently, no standardized assays have been devel- The presence of HCV RASs in the three regions of oped as purchasable kits for HCV resistance testing. interest was assessed in all samples at baseline of Resistance testing therefore relies on homemade tech- retreatment by two methods. The reference method niques based on population or deep sequencing. A was population sequencing of the NS3 protease, NS5A limited number of laboratories has made these tests protein domain I and NS5B polymerase coding regions available in Europe and in other regions, and the perfor- by means of our in-house techniques [17]. The results mance of their assays has not been externally validated. were compared with those generated by the deep The Sentosa SQ HCV Genotyping Assay (Vela Diag- sequencing-based Sentosa SQ HCV Genotyping Assay. nostics GmbH, Hamburg, Germany) is a novel, fully The study was conducted in accordance with the Inter- automated deep sequencing-based assay comprising a national Conference on Harmonization guidelines, customized version of the epMotion 5075 robotic liq- applicable regulations, and the principles of the Decla- uid handling system for RNA extraction and sequence ration of Helsinki. All patients gave written informed library preparation (Sentosa SX101), a customized ver- consent for the use of leftover specimens. sion of an Ion One Touch device for template prepa- ration including emulsion PCR (Sentosa ST401), the HCV resistance testing by means of in-house Ion Torrent technology for deep sequencing (Sentosa population sequencing methods SQ301), and software for data analysis and reporting Briefly, total HCV RNA was extracted from 400 µl of (Sentosa Link and Sentosa Reporter, respectively). This serum by means of QIAsymphony DSP Virus/Pathogen assay has received CE-IVD mark in Europe for HCV kit (Qiagen GmbH, Hilden, Germany), according to the genotype determination and resistance testing in clinical manufacturer’s instructions. The RNA pellet was eluted practice. We recently reported on the ability of this assay with 60 µl of RNAse-free water with 0.04% NaN . to correctly identify the HCV genotype and subtype Complementary DNA synthesis was performed with [14]. The current generation of the assay also has the the OneStep RT-PCR kit (Qiagen GmbH) with sets of 418 ©2019 International Medical Press AVT-19-OA-4532_Rodriguez.indd 418 AVT-19-OA-4532_Rodriguez.indd 418 22/11/2019 11:37:13 22/11/2019 11:37:13 Deep sequencing platform for HCV resistance testing primers adapted to the viral regions targeted (Table 1). A means of the commercial web-based software NGS nested PCR was used to amplify NS3, NS5A and NS5B HCV Module (SmartGene, Lausanne, Switzerland). coding DNA fragments. PCR products were purified by means of NucleoFast 96 PCR plate kit (Macherey- Sequence data Nagel GmbH & Co. KG, Düren, Germany) and directly The GenBank/ENA/DDBJ accession numbers of the sequenced by means of the BigDye Terminator Cycle sequences reported in this paper are PRJNA397404. v3.1 sequencing kit (ThermoFisher Scientific, Courta- boeuf, France) on an ABI 3100 sequencer (Applied Results Biosystems, Foster City, CA, USA), according to the Characteristics of the study population manufacturer’s instructions. Nucleotide sequences were manually corrected and aligned to genotype-1a Nineteen patients previously exposed to different DAA- or 1b reference sequences, respectively, and amino acid containing regimens were included. Most of them changes were deduced from the nucleotide sequences. (78.9%) were males, with a mean age of 57.8 years (Table 2). The patients were mainly infected with geno- HCV resistance testing by means of Sentosa SQ HCV type-1a (73.7%) and 12 of them (63.2%) had a baseline Genotyping Assay HCV RNA level >800,000 IU/ml. Of the 19 patients, Briefly, nucleic acid extraction was performed from 13 (68.4%) had compensated cirrhosis and 4 had 530 µl of serum on the Sentosa SX101 robotic instru- advanced fibrosis (F3 in the METAVIR scoring system). ment using Sentosa Virus Total Nucleic Acid Plus II kit Prior DAA-containing regimens included: pegylated (Vela Diagnostics). The NS3, NS5A and NS5B coding interferon (IFN) with ribavirin plus daclatasvir (n=8); regions were RT-PCR-amplified by means of Veriti Dx pegylated IFN plus daclatasvir and asunaprevir (n=3); 96-Well Thermal Cycler (Applied Biosystems). After sofosbuvir and ribavirin (n=1); sofosbuvir and sime- purification of PCR products using magnetic beads, a previr (n=2); sofosbuvir and ledipasvir with or with- 200-nucleotide fragment library was prepared on Sen- out ribavirin (n=2); daclatasvir and asunaprevir (n=1); tosa SX101. The samples were barcoded by ligation, daclatasvir and simeprevir (n=1); and ritonavir-boosted pooled into a single tube and amplified by emulsion PCR paritaprevir, ombitasvir and dasabuvir with ribavirin on Sentosa ST401i. Deep sequencing was performed by (n=1). Fourteen patients were retreated, according to means of the Sentosa SQ Sequencing Kit on the Sentosa the European recommendations available at the time, SQ301 Sequencer, based on Ion Torrent technology. with the combination of sofosbuvir and simeprevir Primary data analysis was automatically performed for 12 weeks (n=11), sofosbuvir plus ledipasvir with using Sentosa SQ Reporter software. Assembled NS3 ribavirin for 24 weeks (n=1) or sofosbuvir, simeprevir contigs (944-base pair fragment), NS5A contigs (604- and daclatasvir with ribavirin for 24 weeks (n=2). An base pair fragment) and NS5B contigs (685-base pair SVR was achieved in 10 patients (71.4%). The remain- fragment) were aligned with genotype-1a and 1b refer- ing four patients failed, including three patients who ence sequences, respectively. The raw sequencing data received the combination of sofosbuvir and sime- generated by the Sentosa assay were also analysed by previr for 12 weeks and one patient treated with the Table 1. Primers used for nested PCR amplification prior to population sequencing of the HCV NS3 protease, NS5A domain I and NS5B polymerase genes, according to the HCV genotype (1a or 1b) HCV genotype Primer 5′-3′ sequence Gene Nucleotide position References 1 NS3-G1F1 ATG GAR AAG AAR RTY ATY RTN TGG G NS3 3278–3300 Personal 1 NS3-G1F2 ATG GAR AYY AAG CTY ATY ACN TGG G NS3 3278–3300 Personal 1 NS3-G1R CTY TTN CCR CTN CCN GTN GGN GCR TG NS3 4025–4048 Personal 1a NS3-1a-2s CCG ATG GAA TGG TCT CCA AGG NS3 3384–3405 [23] 1a NS3-1a-2a GAG AGG AGT TGT CCG TGA ACA C NS3 3965–3984 [23] 1b NS3-1b-1s GGC GTG TGG GGA CAT CAT C NS3 3316–3332 [24] 1b NS3-1b-1a GGA GAT GAG TTG TCT GTG AA NS3 3968–3985 Personal 1 NS5A-G1xF GGA TGA ACCGGC TSA TAG C NS5A 6084–6103 Personal 1 NS5A-G1xF1 ACC CAG CTS CTG ARR AGG C NS5A 6200–6219 Personal 1 NS5A-G1xR ACG TAR TGG AAR TCC CCC AC NS5A 6626–6646 Personal 1 Sn755 TAT GAY ACC CGC TGY TTT GAC TC NS5B 8255–8278 [25] 1 5B-SI766 CTG YTT TGA CTC CAC NGT RAC NS5B 8266–8287 [26] 1 GEN1A.R1 CCG GGC AYG AGA CAC GCT GTG ATA AAT G NS5B 9277–9305 [27] 1 GEN1B.R1 TGC GGC ACG AGA CAV GCT GTG ATA TG NS5B 9279–9305 [27] Antiviral Therapy 24.6 419 AVT-19-OA-4532_Rodriguez.indd 419 AVT-19-OA-4532_Rodriguez.indd 419 22/11/2019 11:37:13 22/11/2019 11:37:13 C Rodriguez et al. Table 2. Demographic and virological features of the 19 NS5A RASs were found at retreatment baseline in patients included 17 of the 19 patients (89.5%). All of the NS5A RASs found (at positions 24, 28, 30, 31, 58 and 93) had been Parameter Value (n=19) previously described to be associated with NS5A inhib- itor-containing regimen failures in vivo. The RASs were Mean age, years ±ds (range) 57.8 ±9.6 present in single (n=6), double (n=8) or triple (n=3) (46–81) mutants. The most frequent amino acid substitutions Male, n (%) 15 (78.9) were at positions 31 (L31M in six patients and L31V Mean HCV RNA level, log IU/ml ±ds (range) 6.0 ±0.5 (5.2–6.7) in two patients), 30 (Q30R in two patients, Q30K in HCV RNA level >800,000 UI/ml, n (%) 12 (63.2) two patients, Q30E in two patients and Q30H in one HCV genotype patient) and 93 (Y93H in five patients and Y93C in Genotype 1a, n (%) 14 (73.7) two patients; Table 4). All of these RASs confer vari- Genotype 1b, n (%) 5 (26.3) ous levels of reduced susceptibility to first- and second- Fibrosis stage, n (%) 19 (100) generation NS5A inhibitors [18]. Mild fibrosis (Fibroscan ≤ 7.0), n (%) 2 (10.5) NS5B RASs, including L159F and C316N, were Moderate fibrosis (Fibroscan > 7.0 and ≤9.5 kPa), n (%) 0 (0) detected in only one patient infected with genotype-1b Advanced fibrosis (Fibroscan > 9.5 and ≤12.5 kPa), n (%) 4 (21.1) who had been previously exposed to a combination of Compensated cirrhosis (Fibroscan >12.5), n (%) 13 (68.4) sofosbuvir and ledipasvir (Table 5). Both substitutions Treatment-experienced, n (%) 19 (100) have been reported to be associated with reduced sus- Prior DAA-containing regimen Pegylated IFN/ribavirin + daclatasvir, n 8 ceptibility to sofosbuvir [19]. Pegylated IFN/ribavirin + daclatasvir + asunaprevir, n 3 Sofosbuvir with ribavirin, n 1 NS3 protease RASs found by Sentosa SQ HCV Sofosbuvir + simeprevir, n 2 Genotyping Assay Sofosbuvir/ledipasvir ± ribavirin, n 2 NS3 RASs were found at baseline of retreatment in 12 Daclatasvir + asunaprevir, n 1 of the 19 patients (63.2%) with the Sentosa assay, using Daclatasvir + simeprevir, n 1 the internal assay software for interpretation. All of the Paritaprevir/ritonavir/ombitasvir + dasabuvir with 1 RASs found with population sequencing except three ribavirin, n were detected by Sentosa HCV SQ Genotyping Assay in high proportions (>80%). The three RASs not identified DAA, direct-acting antiviral; IFN, interferon. by the assay were V55I, Y56F and S122T, which confer reduced susceptibility to asunaprevir, grazoprevir and combination of sofosbuvir, simeprevir and daclatasvir simeprevir, respectively. plus ribavirin for 24 weeks who discontinued treat- Some RASs were detected by the deep sequencing- ment prematurely owing to a severe pulmonary arterial based Sentosa assay and not with our in-house popu- hypertension. lation sequencing method, including V36A (n=1) and V36M (n=1), Q80R (n=1), D168E (n=1), D168V HCV resistance testing by means of in-house (n=1) and D168Y (n=1), and V170A (n=1). In all but population sequencing methods three cases, they were present as minor viral popula- Our in-house population sequencing techniques were tions, representing 1.1% to 7.1% of the viral qua- used for NS3 protease, NS5A domain I and NS5B resist- sispecies (Table 3). The remaining RASs (D168E/V/Y) ance testing. NS3 RASs were found at retreatment base- represented between 15.9% to 33% of the viral qua- line in 11 of the 19 patients (57.9%), including 5 who sispecies, as reported by the Sentosa assay software had never been exposed to any protease inhibitor. All (Table 3). of the amino acid substitutions found (at positions 55, 56, 80, 122, 155 and 168) had been previously reported NS5A RASs found by Sentosa SQ HCV Genotyping to be associated with protease inhibitor-containing Assay regimen failures in vivo [18]. The RASs were present NS5A RASs were found at retreatment baseline in in single (n=9), double (n=1) or triple (n=1) mutants. 18 of the 19 patients (94.7%). All of the NS5A RASs The most frequent amino acid substitutions were at detected by population sequencing except two were positions 80 (Q80K in three patients and Q80L in one also identified by the Sentosa HCV SQ Genotyping patient) and 155 (R155K in four patients; Table  3). Assay in high proportions (>50%). The two RASs RASs at both positions are known to confer resistance not detected by the Sentosa assay included L28M and to simeprevir in vitro and in vivo, and reduced suscepti- P58S, which both confer reduced susceptibility to first- bility to second-generation protease inhibitors, includ- generation NS5A inhibitors in genotype-1b-infected ing grazoprevir and voxilaprevir [18]. patients. 420 ©2019 International Medical Press AVT-19-OA-4532_Rodriguez.indd 420 AVT-19-OA-4532_Rodriguez.indd 420 22/11/2019 11:37:13 22/11/2019 11:37:13 Deep sequencing platform for HCV resistance testing Table 3. RASs in the NS3 region found by in-house population Table 4. RASs in the NS5A domain I region found by in-house sequencing and Sentosa SQ HCV Genotyping Assay, population sequencing and Sentosa SQ HCV Genotyping Assay, respectively respectively In-house In-house Deep sequencing-based population Deep sequencing-based Sentosa population Sentosa SQ HCV Genotyping sequencing SQ HCV Genotyping Assay (n=19; sequencing (n=19; frequency NS5A RAS (n=19) frequency in the HCV quasispecies) NS3 protease RAS (n=19) in the HCV quasispecies) K24 V36 Q 2 – A – 1 (1.6%) M/L28 M – 1 (7.1%) T 2 – L 1 1 (99.8%) V – 3 (2%; 6.1%; 1.5%) V55 Q/R30 A 1 1 (97.2%) E 3 4 (21.2%; 72%; 42.3%; 72.9%) Y56 H 1 2 (2.8%; 98.6%) F 1 – R 2 5 (99.8%; 20.2%; 13.4%; 37.1%; 84.4%) Q80 L31 K 3 3 (98.7%; 97.1%; 98.3%) M 6 8 (99.6%; 99.7%; 98.8%; 99.5%; 19.3%; L 1 1 (99.9%) 4.5%; 99.5%; 99.3%) R – 1 (1.1%) V 2 2 (99.4%; 99.6%) S122 H/P58 T 1 – P 2 – R155 S 1 – K 4 4 (98.4%; 78.8%; 79.6%; 95.5%) Y93 D168 C 2 3 (99.6%; 3.6%; 54.2%) E 1 2 (15.9%; 94.9%) H 5 5 (98.7%; 99.7%; 93.1%; 99.5%; 99.5%) Y – 1 (33%) V 1 2 (16.7%; 99.7%) NS5A resistance-associated substitutions (RASs) reported to be associated with V170 resistance to NS3 protease inhibitors in patients infected with genotypes-1a or A – 1 (4.3%) 1b failing a direct-acting antiviral containing regimen [18]. NS3 resistance-associated substitutions (RASs) reported to be associated with resistance to NS3 protease inhibitors in patients infected with genotypes-1a or web-based NGS HCV Module software developed by 1b failing a direct-acting antiviral containing regimen [18]. SmartGene. Using this method, three additional RASs were identified as compared with using the Sentosa SQ Several RASs were detected by the deep sequencing- Reporter internal assay software, including a RAS at based Sentosa assay, but not by our in-house popula- position 122 in the NS3-coding region and RASs at tion sequencing method. They included M28V (n=3), positions 28 and 58 in the NS5A-coding region. Q30E (n=1), Q30R (n=3) and Q30H (n=1), L31M (n=2) and Y93C (n=1), which were generally present Discussion as minor viral populations, representing 1.5% to 6.1% of the viral quasipecies. Five NS5A RASs (L31M and Treatment failure with DAA-based regimens is often Q30E/R) present as more than 10% of the viral qua- associated with the selection of viral variants with sispecies in the Sentosa assay were not detected by reduced susceptibility to the administered drug(s) and, Sanger sequencing (Table 4). most often, to other drugs from the same class(es). HCV resistance testing may be useful to guide treat- NS5B polymerase RASs found by Sentosa SQ HCV ment decisions with specific regimens in particular Genotyping Assay patient populations, when the presence of pre-existing NS5B RASs at retreatment baseline were not detected RASs in certain proportions has an impact on treat- in any patient using the Sentosa HCV SQ Genotyping ment outcomes. A number of factors should be taken Assay (Table 5). into consideration, such as the availability of resist- ance testing in clinical practice, the availability of HCV resistance testing by means of Sentosa SQ HCV alternative treatment options, and the consequences of Genotyping Assay plus sequence analysis by means of treatment failure associated with resistance. Because the SmartGene NGS HCV Module treatment options that do not need to be tailored to Raw sequencing data from the Sentosa HCV SQ the baseline RAS profile are available, it is not justi- Genotyping Assay were reanalysed by means of the fied to recommend systematic HCV resistance testing Antiviral Therapy 24.6 421 AVT-19-OA-4532_Rodriguez.indd 421 AVT-19-OA-4532_Rodriguez.indd 421 22/11/2019 11:37:13 22/11/2019 11:37:13 C Rodriguez et al. Table 5. RASs in the NS5B polymerase region found by with the estimated sensitivity of both methods (1% for in-house population sequencing and Sentosa SQ HCV deep sequencing, 20–30% for population sequencing). Genotyping Assay, respectively Discrepant results could be explained for at least NS3 RASs by the time between failure and HCV resistance Deep sequencing-based Sentosa testing. As suggested NS3 RASs are replaced by the wild In-house population SQ HCV Genotyping Assay (n=19; NS5B RAS sequencing (n=19) frequency in the HCV quasispecies) type within 24 to 48 weeks [20,21] Despite the relatively small number of patients stud- L159 ied, our study suggests that the deep sequencing-based F 1 – Sentosa SQ HCV Genotyping Assay can be confidently C316 used to characterize the HCV resistance profile prior N 1 – to retreating patients who failed a DAA-based regimen. The study included only genotype-1a- and 1b-infected NS5B resistance-associated substitutions (RASs) reported to be associated with patients, the most prevalent subtypes in industrialized resistance to NS3 protease inhibitors in patients infected with genotypes-1a or 1b failing a direct-acting antiviral containing regimen [18]. countries, because the Sentosa SQ HCV Genotyping Assay has been designed so far only for these subtypes. However, the assay has the capacity to identify RASs prior to first-line therapy, as stated in the EASL Rec- for all genotypes and a second version of the assay ommendations on Treatment of Hepatitis C 2018 [11]. with improved genotype coverage and with the ability In contrast, HCV resistance testing is useful to guide to detect RASs that are associated with resistance to retreatment decisions in patients who have previously sofosbuvir is in development. failed an IFN-free DAA-based regimen. Nevertheless, Our study has limitations. First, it was performed no standardized commercial assay is yet available for with a relatively small number of patients. Secondly, HCV resistance testing, which still relies on homemade the performance of this assay in non 1a- or 1b-infected methods using population or deep sequencing. patients was not tested. Sentosa SQ HCV Genotyping Assay, the first avail- In conclusion, the present study assessing the clinical able kit for HCV genotyping and resistance testing, is performance of HCV resistance testing with the new deep a new automated deep sequencing-based assay that is sequencing-based assay Sentosa SQ HCV Genotyping standardized and easy-to-use without intensive training Assay showed excellent concordance with population or specialized skills. In the present study, we assessed sequencing in patients infected with genotypes-1a and the ability of this assay to characterize the HCV resist- 1b, suggesting that this assay can be confidently used ance profile in clinical samples from patients infected in clinical practice in the indications of HCV resistance with genotype-1a or 1b who failed a prior DAA-based testing for these subtypes. Technical improvements are treatment. Resistance profiles were in keeping with the now required to allow for pangenotypic coverage and exposure to different DAA-containing regimens, except enrichment of the database with reference prototype in some failing patients who had never been exposed to strains for determination of amino acid changes, as pre- one or more drug classes. viously reported for HIV resistance [22]. Our results show good concordance with popula- tion sequencing in most cases, especially when the web- Acknowledgements based SmartGene software NGS HCV Module was used to analyse the assay’s raw sequence data. Our data We would like to thank VELA Diagnostics GmbH, suggest that the Sentosa SQ Reporter software used for Hamburg, Germany for providing the Sentosa SQ HCV bioanalysis of raw sequence data needs to be improved, Genotyping Assay kits. in particular its database of prototype reference strain This work was supported by internal funding. sequences used for alignment. Interestingly, the design of the assay did not allow it to correctly identify RASs Disclosure statement in the NS5B-coding region, especially those able to con- fer reduced susceptibility to sofosbuvir (RASs at posi- JMP acted as advisors for Abbott and SC acted as advi- tions 159, 282 and 316) because these mutations are sors for Abbott and Cepheid. The remaining authors localized outside the target region, which is amplified declare no competing interests. by the assay. As expected, the deep sequencing-based assay was References more sensitive than population sequencing to detect 1. Polaris Observatory HCV Collaborators. Global prevalence minority variants, representing less than 10% of the and genotype distribution of hepatitis C virus infection in viral populations, but also some variants representing 2015: a modelling study. Lancet Gastroenterol Hepatol 2017; 2:161–176. up to 30% of the viral quasispecies. This is in keeping 422 ©2019 International Medical Press AVT-19-OA-4532_Rodriguez.indd 422 AVT-19-OA-4532_Rodriguez.indd 422 22/11/2019 11:37:13 22/11/2019 11:37:13 Deep sequencing platform for HCV resistance testing 2. Global Burden of Disease Cancer Collaboration, 16. Simmonds P, Smith DB, McOmish F, et al. Identification of Fitzmaurice C, Dicker D, et al. The global burden of cancer genotypes of hepatitis C virus by sequence comparisons in the 2013. JAMA Oncol 2015; 1:505–527. core, E1 and NS-5 regions. J Gen Virol 1994; 75:1053–1061. 3. GBD 2013 Mortality and Causes of Death Collaborators. 17. Hézode C, Chevaliez S, Scoazec G, et al. Retreatment with Global, regional, and national age-sex specific all-cause and sofosbuvir and simeprevir of patients with hepatitis C cause-specific mortality for 240 causes of death, 1990-2013: virus genotype 1 or 4 who previously failed a daclatasvir- a systematic analysis for the Global Burden of Disease Study containing regimen. Hepatology 2016; 63:1809–1816. 2013. Lancet 2015; 385:117–171. 18. Sorbo MC, Cento V, Di Maio VC, et al. Hepatitis C virus 4. Belperio PS, Chartier M, Ross DB, Alaigh P, Shulkin D. drug resistance associated substitutions and their clinical Curing hepatitis C virus infection: best practices from the relevance: update 2018. Drug Resist Updat 2018; 37:17–39. U.S. Department of Veterans Affairs. Ann Intern Med 2017; 19. Donaldson EF, Harrington PR, O’Rear JJ, Naeger LK. 167:499–504. Clinical evidence and bioinformatics characterization 5. Falade-Nwulia O, Suarez-Cuervo C, Nelson DR, Fried MW, of potential hepatitis C virus resistance pathways for Segal JB, Sulkowski MS. Oral direct-acting agent therapy sofosbuvir. Hepatology 2015; 61:56–65. for hepatitis C virus infection: a systematic review. Ann 20. Lenz O, Verbinnen T, Fevery B, et al. Virology analysis of Intern Med 2017; 166:637–648. HCV isolates from genotype 1-infected patients treated 6. Chevaliez S, Trimoulet P, Dorival C, et al. Effect of with simeprevir plus peginterferon/ribavirin in Phases IIb/III resistance-associated substitutions on retreatment of direct studies. J Hepatol 2015; 62:1008–1014. acting antiviral exposed patients in the real-world setting 21. McPhee F, Hernandez D, Zhou N, et al. Pooled analysis (ANRS CO22 HEPATHER). J Hepatol 2017; 66:S84. of HCV genotype 1 resistance-associated substitutions in 7. Buti M, Riveiro-Barciela M, Esteban R. Management NS5A, NS3 and NS5B pre-and post-treatment with 12 of direct-acting antiviral agent failures. J Hepatol 2015; weeks of daclatasvir, asunaprevir and beclabuvir. Antivir 63:1511–1522. Ther 2018; 23:53–66. 8. Pawlotsky JM, Hepatitis C. Drugs: is next generation the 22. Raymond S, Nicot F, Carcenac R, et al. HIV-1 genotypic last generation? Gastroenterology 2016; 151:587–590. resistance testing using the Vela automated next-generation sequencing platform. J Antimicrob Chemother 2018; 9. European Association for the Study of the Liver. EASL 73:1152–1157. Recommendations on Treatment of Hepatitis C 2016. J Hepatol 2017; 66:153–194. 23. Lawitz E, Sulkowski MS, Ghalib R, et al. Simeprevir plus sofosbuvir, with or without ribavirin, to treat chronic 10. AASLD/IDSA HCV Guidance Panel. Hepatitis C guidance: infection with hepatitis C virus genotype 1 in non- AASLD-IDSA recommendations for testing, managing, and responders to pegylated interferon and ribavirin and treating adults infected with hepatitis C virus. Hepatology treatment-naive patients: the COSMOS randomized study. 2015; 62:932–954. Lancet 2014; 384:1756–1765. 11. European Association for the Study of the Liver. EASL 24. Dieterich D, Bruce BR, Flamm SL, et al. Evaluation of Recommendations on Treatment of Hepatitis C 2018. sofosbuvir and simeprevir-based regimens in the TRIO J Hepatol 2018; 69:461–511. network: academic and community treatment of a real- 12. Sanger F, Air GM, Barrell BG, et al. Nucleotide sequence of world, heterogeneous population. Hepatology 2014; 60 bacteriophage phi X174 DNA. Nature 1977; 265:687–695. Suppl 1:220A. 13. Palmer S, Kearney M, Maldarelli F, et al. Multiple, linked 25. Lenz O, Fevery B, Vijgen L, et al. TMC-435 in patients human immunodeficiency virus type 1 drug resistance infected with HCV genotype 1 who failed previous mutations in treatment-experienced patients are missed pegylated interferon/ribavirin treatment: virologic analyses by standard genotype analysis. J Clin Microbiol 2005; of the ASPIRE trial. J Hepatol 2012; 56 Suppl 2:S5. 43:406–413. 26. Sarrazin C, Lathouwers E, Peeters M, et al. Prevalence of 14. Rodriguez C, Soulier A, Demontant V, et al. A novel the hepatitis C virus NS3 polymorphism Q80K in genotype standardized deep sequencing-based assay for hepatitis C 1 patients in the European region. Antiviral Res 2015; virus genotype determination. Sci Rep 2018; 8:4180. 116:10–16. 15. Chevaliez S, Bouvier-Alias M, Pawlotsky JM. Performance 27. Bourlière M, Sulkowski MS, Omata M, et al. An integrated of the Abbott real-time PCR assay using m2000sp and safety and efficacy analysis of >500 patients with m2000rt for hepatitis C virus RNA quantification. J Clin compensated cirrhosis treated with ledipasvir/sofosbuvir Microbiol 2009; 47:1726–1732. with or without ribavirin. Hepatology 2014; 60 Suppl 1:23. Accepted 16 May 2019; published online 21 May 2019 Antiviral Therapy 24.6 423 AVT-19-OA-4532_Rodriguez.indd 423 AVT-19-OA-4532_Rodriguez.indd 423 22/11/2019 11:37:13 22/11/2019 11:37:13

Journal

Antiviral Therapy – SAGE

Published: Aug 1, 2019

There are no references for this article.