Resistance analysis of hepatitis C virus genotype 1 prior treatment null responders receiving daclatasvir and asunaprevir


  • Potential conflict of interest: The authors are employees and shareholders of Bristol-Myers Squibb.

  • Supported by Bristol-Myers Squibb.

  • This study is registered with, number NCT01012895, and is also known as Study AI447-011.

Address reprint requests to: Fiona McPhee, 5 Research Parkway, Wallingford, CT 06492. E-mail:; fax: 203-677-6088.


In a sentinel cohort, hepatitis C virus (HCV) patients (primarily genotype [GT] 1a) were treated with daclatasvir (NS5A inhibitor) and asunaprevir (NS3 protease inhibitor). Preexistence, emergence, and persistence of resistance variants in patients who failed this treatment are described. HCV-infected null responders received daclatasvir (60 mg once daily) and asunaprevir (600 mg twice daily) alone (Group A, 11 patients) or with peginterferon alfa-2a and ribavirin (Group B, 10 patients) for 24 weeks. Resistance testing was performed on baseline samples and samples with HCV RNA ≥1,000 IU/mL at Week 1 through posttreatment Week 48. Resistance substitution susceptibility to inhibition by asunaprevir and daclatasvir was assessed using HCV replicon assays. In Group A, six GT1a patients experiencing viral breakthrough and one GT1a patient who relapsed had detectable NS5A (Q30E/R, L31V/M, Y93C/N) and NS3 (R155K, D168A/E/V/Y) resistance-associated variants at failure. Two of six viral breakthrough patients achieved SVR48 after treatment intensification with peginterferon alfa-2a and ribavirin. For 2/4 viral breakthrough patients not responding to treatment intensification, NS3 resistance variants changed (D168Y to D168T; R155K to V36M-R155K). At posttreatment Week 48, daclatasvir-resistant variants persisted while asunaprevir-resistant variants were generally replaced by wild-type sequences. The NS3 sequence remained unchanged in the one patient with NS3-R155K at baseline, relapse, and posttreatment Week 48. In Group B, no viral breakthrough was observed. Conclusion: The treatment failure of daclatasvir and asunaprevir in HCV GT1a patients was associated with both NS5A and NS3 resistance variants in prior null responders. NS5A resistance variants persisted while NS3 resistance variants generally decayed, suggesting a higher relative fitness of NS5A variants. (Hepatology 2013;53:902–911)


hepatitis C virus




lower limit of quantitation


polymerase chain reaction


sustained viral response.

The investigational direct-acting antivirals, daclatasvir and asunaprevir, are currently in clinical development for treating hepatitis C virus (HCV). Daclatasvir is a first-in-class, highly selective NS5A replication complex inhibitor with picomolar potency and broad HCV genotypic coverage.[1] Asunaprevir is a selective NS3 protease inhibitor with antiviral activity in vitro against HCV genotype (GT) 1 and GT4.[2] These direct-acting antivirals have demonstrated efficacy when individually combined with peginterferon alfa-2a and ribavirin to treatment-naive GT1 patients.[3-6] These regimens were well tolerated.

When peginterferon alfa-2a and ribavirin were added to the dual combination of daclatasvir and asunaprevir, all patients experienced sustained virologic response (SVR) at 48 weeks posttreatment.[7] The combination of daclatasvir and asunaprevir alone resulted in rapid suppression of HCV RNA levels in GT1 null responder patients.[7] This proof-of-concept study was the first to show that null responder HCV-infected patients could be cured with 24 weeks of an interferon-free regimen. Patients (n = 11; nine GT1a and two GT1b) were randomized to receive 60 mg daclatasvir once daily and 600 mg asunaprevir twice daily for 24 weeks. Thirty-six percent (n = 4; two GT1a and two GT1b) of patients achieved SVR24, six GT1a patients experienced viral breakthrough, and one patient relapsed 4 weeks posttreatment (Fig. 1). No resistance variants were detected at baseline for patients experiencing virologic breakthrough. Resistance variants to both daclatasvir and asunaprevir were detected, however, in all cases at or close to viral breakthrough. The rigorous analysis presented here characterizes virologic escape in patients who failed treatment with asunaprevir and daclatasvir, its association with baseline HCV polymorphisms, and decay of emergent drug-resistant variants to posttreatment Week 48.

Figure 1.

HCV RNA levels over time for patients treated with asunaprevir and daclatasvir who experienced viral breakthrough or relapse. All patients were infected with HCV genotype 1a. For Patients 1 through 6, peginterferon alfa-2a and ribavirin were added to the dual-asunaprevir and daclatasvir regimen at or close to the time of viral breakthrough. Data are modified from a previous study.[7] HCV RNA <LLOQ but detectable is represented by <25 IU/mL on the graph, whereas undetectable HCV RNA is represented by <10 IU/mL.

Materials and Methods

Study Design

A detailed description of this study was published[7] and is described briefly below. This open-label, phase 2a study recruited patients from seven centers in the United States and performed in accordance with the Declaration of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements. All patients provided written informed consent.

Patients had chronic HCV GT1 infection with RNA ≥105 IU/mL, no evidence of cirrhosis, and no response to previous HCV therapy. For a patient to be defined as a prior null responder, the patient had to have experienced at least 12 weeks of therapy with peginterferon alfa and had ≤2 log10 decline in HCV RNA. Eligible patients were randomly assigned to Group A or B. Group A patients received dual therapy consisting of daclatasvir (60 mg, orally, once daily) and asunaprevir (600 mg, orally, twice daily) for 24 weeks. Group A patients were eligible to have peginterferon alfa-2a and ribavirin added to their regimens for an additional 48 weeks (as indicated) if viral breakthrough occurred. Group B patients received quadruple therapy consisting of daclatasvir (60 mg, orally, once daily), asunaprevir (600 mg, orally, twice daily, dose adjustment was not permitted), peginterferon alfa-2a 180 μg per week subcutaneously, and ribavirin, oral twice daily, with doses determined according to body weight (1,000 mg daily in patients with body weight of <75 kg, and 1,200 mg daily in patients with body weight of ≥75 kg) for 24 weeks. The primary endpoint of the study was undetectable HCV RNA at posttreatment Week 12. SVR was defined as continued undetectable HCV RNA 12 weeks after cessation of treatment (SVR12).

Laboratory Assessment

HCV RNA was determined at a central laboratory using the Roche COBAS TaqMan v. 2 assay (Roche Molecular Diagnostics) with a lower limit of quantitation (LLOQ) of 25 IU/mL. HCV genotype was determined using VERSANT HCV Amplification 2.0 Kit (LiPA) (Siemens) and IL28B genotype (rs12979860 SNP) was determined by polymerase chain reaction (PCR) amplification and sequencing. Plasma samples for resistance testing were collected at baseline, Days 1 to 7, 14, and at Week 3, and every 2 weeks from weeks 4-12. After Week 12, for Group A, samples were collected every 2 weeks until Week 24 unless peginterferon alfa-2a and ribavirin was added, in which case samples were collected every 4 weeks. For Group B, samples were collected every 4 weeks from weeks 12-24. Posttreatment samples were collected at weeks 4, 12, 24, 36, and 48.

Genotypic and Phenotypic Analyses

Resistance testing was performed on available baseline samples and samples with HCV RNA ≥1,000 IU/mL at Week 1 through posttreatment Week 48. Population sequencing was performed using methods as described.[8-10] Baseline sequences have been deposited in GenBank under accession numbers KC591725-KC591768. PCR amplification was performed on ≥2 PCR reactions per sample where possible to assess for primer bias. For clonal analysis, PCR amplicons were cloned into the TOPO vector and transformed into TOP10 E. coli using a commercially available kit (TOPO TA cloning kit, Invitrogen, Carlsbad, CA) according to the manufacturer's instructions, with ≥30 individual colonies expanded and sequenced for each analysis. Resistance-associated NS5A and NS3 substitutions were introduced into HCV GT1a (H77c) replicon with adaptive variants P1495L and S2204I.[11] These replicons were monitored for phenotypic changes to asunaprevir and daclatasvir as described.[11, 12]


Baseline Analyses

IL28B genotype and baseline polymorphisms at amino acid positions associated with resistance are shown in Tables 1 and 2 for Groups A and B, respectively. In Group A, 10/11 patients carried the non-CC IL28B genotype, indicative of a null responder population. Baseline NS5A and NS3 polymorphisms were detected in patients with both CC and non-CC IL28B genotype. NS5A polymorphisms were observed at baseline for four patients (two patients [GT1b] had Q54Q/H and two patients [1-GT1b and 1-GT1a] had P/H58P/S/H polymorphisms). These substitutions alone are not associated with change in susceptibility to daclatasvir in vitro.[9, 11] Four GT1a patients had NS3 polymorphisms conferring <3-fold change in potency to asunaprevir in vitro (1-Q80L, 3-Q80K).[8] Three of six virologic breakthroughs on dual treatment had baseline NS3-80 polymorphisms. One patient (GT1a) had NS3-R155K which confers 27-fold resistance to asunaprevir. This patient subsequently relapsed. In Group B, 9/10 patients carried the non-CC IL28B genotype; baseline NS5A and NS3 polymorphisms were only detected in patients with the non-CC genotype. NS5A polymorphisms at amino acid positions associated with resistance were observed at baseline for four GT1a patients (Table 2; two patients Q/H54C/Y/H and two patients P/H58P/H). Variants H54C/Y and H/P58H/P confer no changes in susceptibility to daclatasvir.[9, 11] The same NS3 polymorphisms, Q80L and Q80K, observed in Group A were observed in Group B in two patients (1-GT1a and 1-GT1b); like 1a-Q80K, 1b-NS3-Q80L is not associated with a resistance phenotype to asunaprevir.[12]

Table 1. Properties of HCV GT1-Infected Patients Treated With Asunaprevir and Daclatasvir (Group A)
   Resistance-Associated Variants at Baseline Resistance Variants at Virologic FailurePlasma Concentrations Postdose on Day 14 (nM)
Patient IDGTIL28B GTNS3NS5AViral Response (Week of VBT)aNS3NS5A12 h24 h
  1. a


  2. b

    Numbers in parentheses correspond to the percentage of clones that encoded for the resistance variant(s).

  3. c

    Clonal analysis was not performed.

  4. d

    At posttreatment Week 48, HCV < LLOQ with confirmed undetectable HCV RNA at a subsequent visit. Accession numbers referring to NS3 Patient sequences are: KC591751, KC591756, KC591768, KC591750, KC591758, KC591757, KC591754, KC591748, KC591749, KC591767, KC591755. Accession numbers referring to NS5A Patient sequences are: KC591729, KC591742, KC591735, KC591734, KC591737, KC591728, KC591745, KC591726, KC591727, KC591731, KC591743.

  5. Abbreviations: BL, baseline; GT, genotype; n/a, not applicable; PT, posttreatment; SVR, sustained viral response; VBT, viral breakthrough.

   # of clonesVariant (%)b# of clonesVariant (%)b # of ClonesVariant (%)b# of ClonesVariant (%)bAsunaprevirDaclatasvir
11aCT3536VBT (W2)31D168Y(87); D168A(13)37Y93N(100)259249
21aCC32Q80K(100)32H58P(63)VBT (W4)30Q80K; R155K(100)31L31V—H58P(100)331204
31aCT3136VBT (W6)32D168Y(100)32Q30R—L31M(100)59198
41aCT37Q80K(100)32VBT (W8)36Q80K; D168E(100)42Q30R(76); L31V(100); Y93C(17)119267
51aCT40Q80L(100)34VBT (W10) then SVR24 after rescue37Q80L; R155K(100)31Q30R(45); L31V(77); H58P(39); Y93C(16); Y93N(13)32121
61aCT  VBT (W12) then SVR48 after rescue D168Vc L31V; H54Y; Y93Cc182617
71aCT35R155K(80)32Relapse (PT W4)47R155K(100)43Q30E(100)54284
81bTT  Q54H; P58P/SSVR36d n/a n/a343263
91bTT  Q54Q/HSVR48 n/a n/a66187
101aCT  SVR48 n/a n/a501241
111aTT Q80K SVR48 n/a n/a217195
Table 2. Baseline Properties of HCV GT1-Infected Patients Treated With Asunaprevir, Daclatasvir, Peginterferon Alfa-2a, and Ribavirin (Group B)
  Resistance-Associated Variants at Baseline
  1. Accession numbers referring to NS3 patient sequences are: KC591763, KC591752, KC591760, KC591761, KC591764, KC591766, KC591759, KC591747, KC591753, KC591762. Accession numbers referring to NS5A patient sequences are: KC591738, KC591746, KC591741, KC591736, KC591739, KC591732, KC591730, KC591725, KC591740, KC591733. GT, genotype.

1aCTQ54C; P58H/P

Overall Genotypic and Phenotypic Analysis of Virologic Failure

Genotypic and phenotypic analyses (Table 1) showed that all six patients (GT1a) experiencing virologic breakthrough and the patient (GT1a) experiencing relapse had detectable NS5A and NS3 resistance substitutions at or close to time of virologic failure. All variants were enriched at five amino acid positions associated with daclatasvir resistance. Phenotypic data on these substitutions in transient HCV RNA replicon assays have been described[9] and demonstrated that most of the substitutions confer substantial resistance. Additional analyses found that linked substitutions L31V-H58P and Q30R-L31M conferred >2000-fold resistance to daclatasvir (EC90 values were 144 ± 29, and 564 ± 5.8 nM, respectively) while H54Y (EC90 <0.06 nM) did not confer resistance. The same patients with NS5A resistance variants had NS3 resistance variants. Phenotypic data on NS3 substitutions in transient replicon replication assays have been described.[12] The Q80L+R155K substitution conferred 48-fold resistance (EC90 = 485 ± 30.0 nM).

Loss in daclatasvir and asunaprevir potency as a result of emergent NS5A and NS3 resistance substitutions, respectively, offers an explanation for virologic failure. As anticipated, daclatasvir-resistant variants conferred minimal crossresistance towards asunaprevir and asunaprevir-resistant variants conferred minimal effects on the activity of daclatasvir (Table 3). In general, linking of mutations in NS3 and NS5A genes did not significantly impact relative potencies of daclatasvir and asunaprevir; observed changes in drug potency were <3-fold (Table 3). For two Group A GT1b-infected patients, no viral breakthrough occurred during 24 weeks of treatment and neither patient experienced relapse during the 48-week follow-up period.

Table 3. Effects of NS3 Protease and NS5A Amino Acid Substitutions on Asunaprevir and Daclatasvir Potency in Genotype 1a Replicon Assays
NS3NS5AEC90 ± SD (nM)Fold ResistanceEC90 ± SD (nM)Fold Resistance
  1. Predominant resistance variants, as determined by clonal analysis (≥10% of clones) are shown. In general, EC90 values represent the average of at least three independent experiments (standard deviation provided). For variants where a standard deviation is not provided, EC90 values represent the average of two independent experiments. WT refers to the parental HCV genotype 1a (H77c) replicon[11] used to compare the impact of NS3 and NS5A resistance substitutions on asunaprevir and daclatasvir antiviral activity.

  2. a

    Transient HCV replication assay.

  3. Abbreviations: ND, not determined because the variant did not replicate; SD, standard deviation; WT, wild type.

Cell linewtwt10.1 ± 5.510.060 ± 0.0031
Transientwtawta4.5 ± 1.710.020 ± 0.0061
Patient 1D168Ywt938930.030 ± 0.0061
D168Tawta924 ± 32050.020 ± 0.0011
wtY93N7.7 ± 0.811156 ± 5319,267
D168TY93N934 ± 7592708 ± 10411,800
Patient 2Q80K+R155Kawta292 ± 114650.020 ± 0.0031
Q80K+D168Ywt>1000>990.050 ± 0.0091
wtL31V+H58P5.3 ± 0.01144 ± 292400
Q80K+R155KL31V+H58P639 ± 13863272 ± 214533
Q80K+D168YL31V+H58P>1000>99312 ± 1065200
Patient 3D168Ywt938930.030 ± 0.0061
wtQ30R+L31M6.1 ± 0.41564 ± 69400
D168YaQ30R+L31Ma1883 ± 156418>2000>100,000
Patient 4Q80K+D168Eawta209 ± 8460.010 ± 0.0021
wtQ30R+L31V6.1 ± 0.61>2000>33,333
wtaL31VaNDND42 ± 122100
wtL31V+Y93C7.2 ± 0.51>2000>33,333
Q80K+D168EQ30R+L31V374 ± 2937>2000>33,333
Q80K+D168EL31V+Y93C446 ± 8844>2000>33,333
Q80K+D168EaL31Va223 ± 45070 ± 143500
Patient 5Q80L+R155Kwt485 ± 30480.050 ± 0.0071
wtY93N7.7 ± 0.811156 ± 5319,267
wtQ30R+L31V6.1 ± 0.61>2000>33,333
wtL31V+H58P5.3 ± 0.01144 ± 292400
Q80L+R155KY93N348 ± 19734637 ± 12910,617
Q80L+R155KaQ30R+L31Va97 ± 3722>1000>50,000
Q80L+R155KaL31V+H58Pa101 ± 3322161 ± 438050
Patient 7R155Kwt270 ± 7270.04 ± 0.0101
wtQ30E12 ± 21373 ± 46217
Q80L-R155KQ30E438 ± 643350 ± 175833

For patients in Group A, trough plasma concentrations of daclatasvir 24 hours postdose on Day 14 ranged from 187-617 nM in Group A. Range in plasma concentrations of asunaprevir 12 hours postdose on Day 14 ranged from 32-501 nM. No correlation was observed between these trough plasma concentrations of daclatasvir and asunaprevir and virologic breakthrough (Table 1). In vitro resistance phenotypes (EC90 values in transient and stable replicon cell line assays) of emergent predominant resistance variants, however, were higher than observed drug exposures in plasma for daclatasvir and asunaprevir. Review of manual pill counts and dosing diaries suggested excellent adherence to treatment except for Patient 1, who admitted to several missed asunaprevir doses within the first 2 weeks of treatment.

Genotypic and Phenotypic Analysis of Individual Patients With Virologic Failure

Patient 1 (GT1a) had early viral breakthrough with detectable drug-resistant variants as early as Week 2 (Fig. 2) and started treatment intensification with peginterferon alfa-2a and ribavirin at Week 4. Emergent NS5A-Y93N conferred 19,267-fold reduced susceptibility to daclatasvir in vitro, and persisted through posttreatment Week 48. NS3-D168Y and NS3-D168A also emerged at virologic breakthrough and conferred 93-fold and 29-fold reduced susceptibilities to asunaprevir (Table 3) with 0.23 and 0.01-fold relative replication capacities (Fig. 2), respectively, versus a GT1a (H77c) reference replicon. NS3-D168T emerged as a minor variant (10%; 4/40 clones), determined by clonal analysis, at Week 12 (8 weeks into treatment intensification) conferring 205-fold reduced susceptibility to asunaprevir (Table 3) and 1.6-fold relative replication capacity when compared to GT1a (H77c) (Fig. 2). This became the predominant variant from Week 24 (20 weeks after treatment intensification) through posttreatment Week 36. D168T may have emerged from D168A based on synonymous codon usage. The low relative replication capacity of D168A was improved by changing to D168T (see comparison of in vitro replication capacities, Fig. 2). D168T was no longer detected by clonal analysis at posttreatment Week 48 due to outgrowth of wild-type virus. It should be noted that D168G (13-fold reduced susceptibility to asunaprevir) was detected in one sequenced colony at this timepoint.

Figure 2.

Time course of HCV RNA levels and resistance profiles of NS3 and NS5A for Patient 1, who had early viral breakthrough. RNA levels are shown over time for Patient 1 infected with HCV genotype 1a who received daclatasvir and asunaprevir initially, then started treatment intensification with peginterferon alfa-2a and ribavirin at Week 4 and stopped them at Week 24. The percentage of emerging NS3 D168 and NS5A Y93 resistance variants, as determined by clonal analysis or population sequencing, and their codons are shown over time. Clonal analysis was performed for both NS3 and NS5A (×) or only for NS3 (open circles); other timepoints were determined by population sequencing. The relative replication capacities described in the table insert represent the replication capacities for GT1a (H77c) replicon variants containing the different NS3 protease substitutions compared to the reference GT1a (H77c) replicon. HCV RNA <LLOQ but detectable is represented by <25 IU/mL on the graph, whereas undetectable HCV RNA is represented by <10 IU/mL. PT, posttreatment.

The time course of HCV viral load for Patient 2 (GT1a) indicated early viral breakthrough at Week 4 followed by viral suppression during treatment intensification with peginterferon alfa-2a and ribavirin for approximately 41 weeks and viral relapse within 4 weeks of treatment discontinuation (Fig. 3). At Week 4, NS5A-L31V-H58P was detected which exhibits 2,400-fold reduced susceptibility to daclatasvir (Table 3) while the major NS3 variant (100%, 30/30 NS3 clones) was Q80K-R155K, which confers ∼29-fold loss in potency to asunaprevir (Table 3). By Week 6, clonal analysis revealed several variants, although Q80K-R155K still predominated (∼68%, 27/40 NS3 clones). The patient responded to treatment intensification with peginterferon alfa-2a and ribavirin but relapsed; NS5A-L31V-H58P was detected, the same as at viral breakthrough, while the NS3 variant had changed to V36M-Q80K-R155K. At posttreatment Week 48, NS5A-L31V-H58P still persisted; however, a minor NS3 variant at Week 6 of dual treatment (V36M-Q80K, 12.5% [5/40 NS3 clones]) now predominated (75% [36/48 NS3 clones]) while Q80K-R155K and V36M-Q80K-R155K were no longer detected (Fig. 3).

Figure 3.

Time course of HCV RNA levels and resistance profiles of NS3 and NS5A for Patient 2, who had early viral breakthrough. RNA levels are shown over time for Patient 2 infected with HCV genotype 1a who received daclatasvir and asunaprevir initially, then started treatment intensification with peginterferon alfa-2a and ribavirin at Week 6 and continued for ∼41 weeks. Clonal analysis was performed at the indicated timepoints, with n indicating the numbers of colonies sequenced. All combinations of variants are shown; the Q80K-R155K-D168Y combination was seen in 1/40 colonies and may have been detected due to recombination during PCR amplification.[24] HCV RNA <LLOQ but detectable is represented by <25 IU/mL on the graph, whereas undetectable HCV RNA is represented by <10 IU/mL. PT, posttreatment.

At Week 6, Patient 3 (GT1a) experienced viral breakthrough (HCV RNA = 46 IU/mL). Resistance variants NS5A-Q30R-L31M and NS3-D168Y were detected at Week 7 (HCV RNA = 66504 IU/mL), with the former variant conferring 9,400-fold reduced susceptibility to daclatasvir and the latter conferring 93-fold reduced susceptibility to asunaprevir (Table 3; Supporting Fig. S1). Patient 3 received treatment intensification with peginterferon alfa-2a and ribavirin for ∼47 weeks but experienced relapse when treatment was halted. Assessment of NS5A and NS3 sequences over time revealed detection of NS5A-Q30R-L31M out to posttreatment Week 48, while NS3-D168Y was no longer detected (0/66 NS3 clones) at this timepoint.

When Patient 4 (GT1a) experienced viral breakthrough at Week 8, the predominant NS5A variant (67%; 28/42 NS5A clones) was Q30R-L31V (>33,333-fold reduced susceptibility to daclatasvir, Table 3; Supporting Fig. S2). The only NS3 variant detected was Q80K-D168E, which confers 46-fold reduced susceptibility to asunaprevir (Table 3; Supporting Fig. S2). Patient 4 responded to ∼46 weeks of treatment intensification with peginterferon alfa-2a and ribavirin but subsequently relapsed. NS5A and NS3 resistance variants detected during posttreatment follow-up were NS5A-L31V-Y93C (a predominant species at Week 12, 2 weeks after the initiation of the intensification therapy) and NS3-Q80K-D168E.

Patients 5 (Supporting Fig. S3) and 6 (clonal analysis was not performed) responded to treatment intensification with peginterferon alfa-2a and ribavirin (26 weeks for Patient 5 and 46 weeks for Patient 6) even though at viral breakthrough signature NS5A and NS3 resistance variants were detected (Table 3; Patient 5 only).

Patient 7 (GT1a) responded rapidly to treatment (Supporting Fig. S4) despite the preexistence of 1a-NS3-R155K (27-fold reduced susceptibility to asunaprevir) at baseline. No resistance to daclatasvir was observed at baseline. No viral breakthrough was detected during 24 weeks of treatment; however, at Week 4 posttreatment relapse occurred. Clonal analysis showed emergence of NS5A-Q30E (Q30E confers 6,217-fold reduced susceptibility to daclatasvir, Table 3). The predominant variant changed during 48 weeks posttreatment (Q30E to K and G then to Q). At posttreatment Week 48, the resistance-associated variant NS5A-Y93N accounted for ∼30% (11/36 clones) of the population after only being detected as a minor variant (3/91 clones) at posttreatment Week 36 (Supporting Fig. S4). The NS3-R155K variant was enriched at failure when compared to baseline (Supporting Fig. S4). The NS3 sequence remained unchanged from posttreatment Week 4 to posttreatment Week 48.

Genotypic and Phenotypic Analysis of an Individual Patient Responding to Quadruple Treatment

One patient (GT1a) in Group B had HCV RNA levels >1,000 IU/mL at Week 1. Sequence analysis revealed the enrichment of NS5A-L31M as early as 8 hours into treatment with daclatasvir, asunaprevir, peginterferon alfa-2a, and ribavirin, a mutation which was still detected as the predominant species at Day 3 (data not shown). The NS5A-L31M substitution confers reduced susceptibility to daclatasvir (3,350-fold) in vitro versus a GT1a (H77c) reference replicon.[11]


There is an unmet medical need for HCV GT1 patients who are prior null responders to peginterferon alfa and ribavirin therapy. Cure rates in this population are low even with the addition of recently approved direct-acting antivirals to peginterferon alfa and ribavirin.[13, 14] Herein, we report on a proof-of-concept study that involved patients with chronic HCV GT1 infection who had not responded to prior peginterferon alfa-2a and ribavirin treatment. Quadruple therapy (asunaprevir, daclatasvir, and peginterferon alfa-2a and ribavirin) resulted in a very high SVR rate (≥90%, 9/10 GT1a and 1-GT1b at 48 weeks posttreatment) in patients who did not respond to prior therapy. Even with the rapid enrichment of an NS5A resistance variant, the quadruple therapy was potent enough to result in viral clearance. A high rate of resistance-associated failure was observed in HCV GT1a patients treated with dual therapy (daclatasvir and asunaprevir alone).

The combination of daclatasvir and asunaprevir appeared to have a higher barrier to resistance in GT1b than GT1a because resistance-associated failure was only observed in GT1a patients.[7] The higher clinical resistance barrier in patients infected with GT1b is supported by the findings in Japanese studies assessing the efficacy of daclatasvir and asunaprevir in HCV GT1b prior null responders.[15, 16] These GT1b patients still achieved SVR24 even with the preexistence of a signature resistance variant (NS5A-Y93H). Treatment of GT1a patients with the two direct-acting antivirals was associated with enrichment of both NS5A and NS3 resistance variants in the prior null responder population. NS5A substitutions were similar to those previously reported.[3, 9] It should be noted, however, that clonal analysis results indicated the presence of different resistance pathways during resistance selection in the GT1a patients who failed treatment in this small study (Supporting Figs. S1-4). These different resistance pathways may be related to the heterogeneity of the NS5A baseline sequence.[17] The NS3 amino acid positions susceptible to substitution in the presence of asunaprevir were similar to those described in an in vitro GT1a replicon resistance study.[12] This is suggestive of an in vitro/in vivo correlation for resistance to asunaprevir in GT1a. A relationship between baseline resistance and virologic breakthrough was unclear. Patient 7, with a preexisting NS3-R155K, was expected to experience virologic breakthrough but did not. Instead, three of six virologic breakthroughs had preexisting NS3-Q80 polymorphisms. Given the polymorphic nature of NS3-80 in GT1a sequences, its correlation with virologic failure requires further investigation.

The dual combination of daclatasvir and asunaprevir was sufficient to suppress viral breakthrough in Patient 7, who had a preexisting 1a-NS3-R155K. Although relapse was observed at Week 4 posttreatment in this patient, preexistence of the asunaprevir-resistant variant did not result in a delayed decline of HCV RNA. It is unknown if a cure could have been achieved with the addition of an interferon or third direct-acting antiviral. NS3-R155K was detected as the major emergent variant in GT1a patients failing treatment with boceprevir or telaprevir, whereas the other emergent resistance-associated variants to asunaprevir NS3-D168Y/E/T have also been detected in patients treated with TMC435 and vaniprevir.[18, 19] Emergent daclatasvir-resistant variants NS5A-Q30E/R, L31V/M, and Y93C/N have also been detected by other first-generation NS5A inhibitors that are based on the structure of daclatasvir.[20, 21]

In contrast, treatment of GT1 prior null responders, the majority of whom were infected with GT1a, with 24 weeks of the quadruple therapy (daclatasvir, asunaprevir, peginterferon alfa-2a, and ribavirin) resulted in a durable response with no confirmed relapse through 48 weeks of follow-up.[7] Interestingly, the regimen was robust enough to result in cure even with the early transient emergence of daclatasvir-resistant variants.[7] This suggests that the drug combination was sufficient to ultimately suppress the emergence of virally fit high level drug-resistant variants.

Addition of peginterferon alfa-2a and ribavirin to daclatasvir and asunaprevir as rescue or intensification therapy resulted in a cure for 33% (2/6) of patients (Patients 5 and 6) who experienced viral breakthrough to daclatasvir and asunaprevir.[7] These two patients had rapid declines in viral load at Week 2 (<25 IU/mL) but experienced virologic breakthrough at weeks 10 and 12, respectively. The HCV RNA levels were low (<10,000 IU/mL) at the time of treatment intensification, although they had detectable signature resistance variants to both daclatasvir and asunaprevir. Retreatment of prior null responders with peginterferon alfa and ribavirin normally results in <10% SVR. However, in the cases presented here, patients were able to respond to the quadruple combination. It is possible that restoration of innate and/or adaptive immunity occurred after clearance of the virus from serum, thus enabling these patients to subsequently respond to quadruple therapy. For example, viral clearance may have reduced PD-1 expression, thus allowing for recovery of the immune system. In addition, in vitro studies by Foy et al.[22] have shown that NS3 protease inhibitors lead to inhibition of HCV NS3/4A-mediated repression of interferon response factor (IRF3) activation, thus leading to restoration of transcription of interferon-inducible genes. For 50% (2/4) of those patients who eventually failed rescue, higher-level resistance variants or more fit NS3 resistance variants emerged but the NS5A variants remained unchanged.

In a Japanese study assessing the efficacy of daclatasvir and asunaprevir in HCV GT1b patients who were peginterferon-alfa/ribavirin ineligible or intolerant, the combination of suboptimal trough concentrations of both of these agents in plasma, as well as the preexistence of the resistance polymorphism NS5A-Y93H at baseline, may have played a role in patients experiencing virologic breakthrough.[23] In the study described herein, no clear relationship was observed between trough concentration values of daclatasvir and asunaprevir in the dual regimen and the ability of a patient infected with GT1a to achieve SVR. All but one of the HCV GT1a patients who experienced viral breakthrough were adherent. No baseline NS3 or NS5A resistance variants were detected by clonal analysis, even in the patient (Patient 1, Fig. 2) who experienced rapid viral breakthrough within the first 2 weeks of treatment. Taken together, the influence of drug exposure on emergence of resistance was difficult to ascertain from this small study examining HCV GT1a failures.

As anticipated from a prior null responder population, the majority of patients in this study carried a non-CC IL28B genotype, thus making a correlation between IL28B status and efficacy hard to assess. Studies in treatment-naive patients would be better suited to examine this relationship.

Viral fitness of emergent or enriched NS5A and NS3 variants was dependent on the variant and target. For three of four patients who experienced viral breakthrough and later failed treatment intensification with peginterferon alfa-2a and ribavirin, asunaprevir-resistant variants D168V or D168Y were detected at the time of virologic failure to dual therapy. The D168 resistance variants were no longer detected by clonal analysis 48 weeks posttreatment in these three patients. However, emergent NS3-D168E persisted in one patient. This is not surprising, given that D168E has been detected in treatment-naive patients,[10] suggesting that it is relatively fit. In contrast, emergent daclatasvir-resistant variants persisted throughout 48-week posttreatment. Different NS5A variants (Y93N, L31V-H58P, Q30R-L31M, and L31V-Y93C) were detected in four patients who experienced virologic failure to dual treatment, and all were fit relative to wild-type virus. Persistence of linked NS5A resistance variants (L31M/V-Y93H) out to 48 weeks posttreatment has also been observed in HCV GT1b-infected patients who failed treatment with daclatasvir.[23] Long-term follow-up studies are currently ongoing to assess whether these variants will be replaced by wild-type sequence. Apparent decay of a daclatasvir-resistant variant was observed in Patient 7, who relapsed during dual treatment. Clonal analysis revealed that emergent NS5A-Q30E was no longer detected at posttreatment Week 48. Interestingly, another resistance variant (NS5A-Y93N) outgrew the original NS5A resistance variant even though it was only first detected at posttreatment Week 36. Since Q30 substitutions linked with Y93N were no longer detected at posttreatment Week 48, NS5A-Y93N may offer a fitness advantage.

In conclusion, treatment of prior null responder patients with quadruple therapy (daclatasvir, asunaprevir, and peginterferon alfa-2a and ribavirin) resulted in all HCV GT1 patients being cured in this sentinel study. Treatment with dual therapy (daclatasvir and asunaprevir) also resulted in all of the GT1b patients being cured, while response rates were significantly lower in GT1a patients. When viral breakthrough occurred in patients infected with HCV GT1a receiving dual therapy, daclatasvir-resistant and asunaprevir-resistant substitutions emerged together. These substitutions were similar to those reported previously. Treatment intensification in patients who experienced virologic breakthrough was capable of providing SVR in a minority of patients despite prior null response to peginterferon alfa-2a and ribavirin. Finally, signature NS5A resistance-associated variants persisted throughout the study, while NS3 resistance-associated variants decayed, suggestive of a lower relative fitness cost of NS5A variants. Additional studies will enhance understanding of HCV treatment with daclatasvir and asunaprevir.


We thank the patients for their participation and commitment during the study. We also thank the investigators and contributors from each study site. The authors thank Bing He for involvement in the pharmacokinetic analysis. Professional medical writing and editorial assistance was provided by Carolyn Carroll, PhD, an employee of Bristol-Myers Squibb. Parts of this study were presented at 46th EASL Congress, Berlin, Germany, March 30-April 3, 2011, Oral Abstract 63, and HepDART 2011, Koloa, Hawaii, December 4-8, 2011.