Patterns of Resistance-Associated Substitutions in … of patterns of resistance-associated substitutions (RASs) selected after treatment failure to the different DAA-based regimens

Gastroenterology 2018;154:976–988

CLINICALLIVER

Patterns of Resistance-Associated Substitutions in PatientsWith Chronic HCV Infection Following Treatment WithDirect-Acting Antivirals
Julia Dietz,1 Simone Susser,1 Johannes Vermehren,1 Kai-Henrik Peiffer,1
Georgios Grammatikos,1 Annemarie Berger,2 Peter Ferenci,3 Maria Buti,4 Beat Müllhaupt,5

Bela Hunyady,6 Holger Hinrichsen,7 Stefan Mauss,8 Jörg Petersen,9 Peter Buggisch,9

Gisela Felten,10 Dietrich Hüppe,10 Gaby Knecht,11 Thomas Lutz,11 Eckart Schott,12

Christoph Berg,13 Ulrich Spengler,14 Thomas von Hahn,15 Thomas Berg,16 Stefan Zeuzem,1

and Christoph Sarrazin1,17, for the European HCV Resistance Study Group

1Department of Internal Medicine 1, University Hospital Frankfurt, Frankfurt, Germany, German Center for Infection Research(DZIF), External Partner Site Frankfurt, Germany; 2Institute for Medical Virology, University Hospital Frankfurt, Frankfurt,Germany; 3Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria; 4Hospital Universitario ValleHebron and Ciberehd del Instituto Carlos III, Barcelona, Spain; 5Swiss Hepato-Pancreato-Biliary Center and Department ofGastroenterology and Hepatology, University Hospital Zürich, Zürich, Switzerland; 6Somogy County Kaposi Mór TeachingHospital, Kaposvár, Hungary; 7Practice of Gastroenterology, Kiel, Germany; 8Practice of Gastroenterology, Düsseldorf,Germany; 9Institute for Interdisciplinary Medicine IFI, Hamburg, Germany; 10Practice of Hepatology, Herne, Germany;11Infektiologikum, Frankfurt, Germany; 12Department of Hepatology and Gastroenterology, Charité Universitätsmedizin Berlin,Berlin, Germany; 13Department of Internal Medicine I, University of Tübingen, Tübingen, Germany; 14Department of InternalMedicine I, University of Bonn, Bonn, Germany; 15Department of Gastroenterology, Hepatology and Endocrinology,Medizinische Hochschule Hannover, Hannover, Germany; German Center for Infection Research (DZIF), Hannover-Braunschweig Site, Germany; 16Department of Gastroenterology and Rheumatology, University Hospital Leipzig, Leipzig,Germany; and 17Medizinische Klinik II, St. Josefs-Hospital, Wiesbaden, Germany

Abbreviations used in this paper: DAA, direct-acting antiviral; DCV,daclatasvir; EOT, end of treatment; GT, genotype; HCV, hepatitis C virus;LDV, ledipasvir; NS, nonstructural protein; PCR, polymerase chain reac-tion; PEG IFN, pegylated interferon; PI, protease inhibitor; RASs, resis-tance-associated substitutions; RBV, ribavirin; SMV, simeprevir; SOF,sofosbuvir; SVR, sustained virologic response; 3D, 3 DAA regimen(paritaprevirDombitasvirDdasabuvir); 2D, 2 DAA regimen (paritaprevirDombitasvir).

Most current article

© 2018 by the AGA Institute0016-5085/$36.00

https://doi.org/10.1053/j.gastro.2017.11.007

BACKGROUND & AIMS: Little is known about substitutions thatmediate resistance of hepatitis C virus (HCV) to direct-acting an-tivirals (DAAs), due to the small number of patients with treat-ment failure in approval studies. It is important to identifyresistance patterns to select effective salvage treatments.METHODS: We performed a comprehensive analysis forresistance-associated substitutions (RASs) in HCV genes(nonstructural protein [NS]3, NS5A, NS5B) targeted by DAAs. Wecompared NS3, NS5A, and NS5B sequences from 626 patients inEurope with DAA failure with sequences from 2322 DAA-naïvepatients, infectedwith HCV genotypes 1 to 4.We considered RASsto be relevant if they were associated with DAA failure in patientsor conferred a greater than twofold change in susceptibilitycompared with a reference strain in in vitro replicon assays. Datawere collected on pretreatment status, DAA regimen, the treat-ment initiation date and duration, and virologic response. Patientswho received at least 4weeks of antiviral treatmentwere includedin the analysis. RESULTS: RASs in NS3 associated with sime-previr or paritaprevir failure include R155K and D168E/V. Inaddition, several RASs were specifically associated with failure ofsimeprevir (Q80K/R in patients with genotype 1a or 4) or par-itaprevir (Y56H in combination with D168V in patients withgenotype 1b). Y93H in NS5A was the RAS most frequentlyassociated with failure of daclatasvir, ledipasvir, or ombitasvir inpatients with genotype 1b infection, and L31M was associatedwith failure of daclatasvir or ledipasvir, but not ombitasvir. RASsin NS5A were heterogeneous among patients with HCV genotype1a or genotype 4 infections. In patients with HCV genotype 3,Y93H was associated with resistance to daclatasvir, but no RASswere associated with ledipasvir failure, pointing to a limitedefficacy of ledipasvir in patients with genotype 3. Amongpatients failed by sofosbuvir-containing regimens, L159F was

enriched in patients with genotype 1b (together with C316N) orgenotype 3 infection, whereas the RAS S282T was rarelyobserved. CONCLUSIONS:We compared RASs in NS3, NS5A, andNS5B among patients failed by DAA therapy. Theses varied withthe HCV genotype and subtype, and the different drug classes.These findings might be used to select salvage therapies.

Keywords: HCV; RASs; DAA; Virologic Treatment Failure.

everal combination regimens based on direct-acting
Santivirals (DAAs) that target proteins with crucialfunctions in the hepatitis C virus (HCV) replication cycle areapproved for the treatment of chronic hepatitis C.1 Althoughrates of sustained virologic response (SVR) are high,because of the large number of infected patients, a sub-stantial number will require a rescue treatment.2 Virologicfailure to DAA-based therapies is associated with theselection of resistant viral isolates and retreatment withthe same regimen has limited efficacy.1,3,4 Therefore, the
http://crossmark.crossref.org/dialog/?doi=10.1053/j.gastro.2017.11.007&domain=pdf


EDITOR’S NOTES

BACKGROUND AND CONTEXT

Little is known about substitutions that mediate resistanceof HCV to direct-acting antivirals (DAAs). However, it isimportant to identify resistance patterns to selecteffective salvage treatments.

NEW FINDINGS

Sequencing of HCV NS3, NS5A and NS5B in 626 patientswith a DAA failure in Europe revealed characteristicresistance patterns which varied with the HCV genotypeand subtype, and the different drug classes.

LIMITATIONS

This retrospective study was non-controlled and mayfavor the analysis of difficult-to-treat patients.

IMPACT

The identified typical treatment-selected resistancepatterns for broadly currently used DAA regimensenable the selection of specific retreatment optionsbased on the result of resistance analyses.

March 2018 HCV RASs in European DAA-experienced Patients 977

CLINICAL

LIVE

R

knowledge of patterns of resistance-associated substitutions(RASs) selected after treatment failure to the different DAA-based regimens may be useful for selection of individual,effective salvage treatment options and resistance testing isrecommended by international guidelines in this setting.5

Due to the high antiviral efficacy, limited data from theapproval studies are available on RASs selected in patientswith treatment failure. The major sofosbuvir (SOF) RASS282T was observed in a few patients only.6–8 In phase 2/3studies for simeprevir (SMV) plus sofosbuvir, only 34 pa-tients had a virologic treatment failure.9–11 For regimenswith sofosbuvir and the nonstructural protein (NS)5A in-hibitors ledipasvir (LDV) and daclatasvir (DCV), 51 and 36treatment failures were detected in the phase 2/3 studies,respectively.7,12–15 Finally, for the large phase 3 develop-ment program of the combination of paritaprevir/ritonavir,ombitasvir, and dasabuvir (3D regimen) overall 40 patientswith virologic failure were reported.16–18

Because of the lack of the availability of a commercialassay and the limited number of patients with virologicfailure at each study site, a centralized HCV resistanceanalysis was performed for the European resistance studygroup. Based on the data from a very large number ofDAA-naïve patients (n ¼ 2323) and patients with a failure tosofosbuvir/pegylated interferon/ribavirin (SOF/RBV þ PEGIFN) (n ¼ 63), sofosbuvir/ribavirin (SOF/RBV) (n ¼ 132),simeprevir/sofosbuvir (n ¼ 55), ledipasvir/sofosbuvir (n ¼232), daclatasvir/sofosbuvir (n ¼ 89), and 3D/2 DAAregimen (paritaprevirþ ombitasvir) (2D) (n ¼ 55), for thefirst time a comprehensive and comparative analysis of HCVRASs patterns was possible in the present study.

MethodsPatients

In this noninterventional epidemiological study, serum of4240 patients with a chronic hepatitis C infectionwith genotypes

1 to 6 was sent from study sites of different European countriesto our laboratory for routine diagnostic HCV resistance testing.Information of the pretreatment status was provided on theresistance analysis submission form. We excluded individualswith insufficient data regarding the treatment status and pa-tients who received investigational DAAs within a clinical study.Moreover, special geno- and subtypes (GT1c, 1e, 1l, GT2k/1bchimeras that were published separatly,19 GT5 and 6) andboceprevir/telaprevir-treated patients were not included,leaving 2322 DAA-naïve and 626 DAA-experienced patients forall further analyses (Supplementary Figure 1). After completionof resistance testing, data were collected from individuals with aDAA treatment failure in a retrospective manner, including thedocumentation of limited parameters, such as cirrhosis, thepretreatment status, the administered DAA regimen, the treat-ment initiation date and duration, as well as the virologicresponse. Patients who received at least 4 weeks of antiviraltreatment were included. Investigations were performed accord-ing to the Declaration of Helsinki, and approval of the usage ofpatient blood samples and retrospective collection of data forresearch purpose was obtained from the local ethics committee(Ethikkommission der J. W. Goethe-Universität Frankfurt).

NS3, NS5A, and NS5B Amplificationand Sequencing

TheHCVRNAextraction and cDNAsynthesiswere performedas described previously.20 The NS3, NS5A, and NS5B DAA targetregions were amplified by nested polymerase chain reactions(PCRs) using 1/10 of cDNA and outer PCR product, respectively,and the Fast Cycling PCR Kit (Qiagen, Hilden, Germany). All PCRamplifications started with an initial denaturation step for5 minutes at 95�C. Each cycle consisted of a denaturation step at96�C for 5 seconds, and an annealing at the temperaturedescribed in the supplementary information for 5 seconds fol-lowed by an extension step at 68�C for 1 minute. Finally, PCRreactions were completed by an extension step at 72�C for 1minute. The number of PCR cycles and the primer sequences arelisted in the supplementary information. Resulting PCR productswere gel-purified using the QIAquick Gel Extraction Kit (Qiagen).Population sequencing was performed with primers as denotedin the supplementary information and the Big Dye Terminatorv3.1 Cycle Sequencing Kit on an ABI Prism 3130xl GeneticAnalyzer (Applied Biosystems, Foster City, CA).

Analysis of RASsAll sequences were proofread and aligned using BioEdit

version 7.2.5 (T. Hall, Ibis Therapeutics, Carlsbad, CA). Weconsidered RASs as relevant if they were described to beassociated with treatment failure in vivo and/or if theyconferred a greater than twofold changed drug susceptibility incomparison with a wild-type reference strain in in vitro repli-con assays that were mainly conducted by the industry andpublished in several studies.1,3 All analyzed RASs are listed inthe supplementary information. Fold change resistance levelsare an indication, but they depend on the type of the assay usedand are not directly comparable between different studies. Asdescribed in previous reports,20,21 also minority RASs wereregarded as relevant that were detectable as mixed peaks in theelectropherogram, whereby a sensitivity of approximately 15%to 20% can be assumed for population sequencing.20 Areevaluation of HCV geno- and subtypes was performed on the

978 Dietz et al Gastroenterology Vol. 154, No. 4

CLINICALLIVER

basis of the obtained nucleotide sequences and sequences weredesignated as wild-type when no RASs were detected.

We compared RASs before and after a DAA-based treat-ment, and when combined RASs were detected in one viralvariant they were counted as single RAS. To calculate the fre-quency of RASs, the number of each RAS was divided by theoverall number of target gene sequences. As in the vast ma-jority of cases, matched samples of the same patient before andafter a DAA treatment failure were not available, we performedoverall comparisons of DAA-naïve versus DAA-experiencedpatients for each genotype and treatment regimen.

ResultsVirological Characteristics of PatientsWith a DAA Failure

The vast majority of patients had a relapse (84%),whereas a nonresponse/breakthrough was detected in 6%of individuals only (the remaining 10% of patients dis-continued treatment or the type of failure was not speci-fied). Interestingly, patients with a nonresponse/breakthrough exhibited higher prevalences of DAA-specificRASs compared with individuals with a relapse (86% vs66%, respectively). Moreover, we detected in 1.7% of pa-tients another genotype after DAA failure compared withthe one determined before treatment initiation by com-mercial assays (data not shown).

NS5B RASs After Sofosbuvir/Ribavirin �Pegylated Interferon

We investigated the prevalence of NS5B nucleotideinhibitor RASs in patients after treatment with

Figure 1. NS5B nucleotide inhibitor RASs following sofosbuvirDAA-naïve patients. Frequencies of RASs in patients infected winvestigated patients is displayed at the top of the figure as wellwith the range stated in parentheses. Typical patterns of RASsindicated in red letters. At the bottom, resistance levels are spsequences analyzed is indicated.

sofosbuvir/ribavirin with or without pegylated interferon incomparison with their frequencies before treatment initia-tion. Patients with GT1a (n ¼ 35) or GT2 (n ¼ 26) did notdevelop NS5B RASs after treatment failure (data not shown).Interestingly, in individualswith GT1b, L159Fwas selected incombination with C316N (32% and 55%, respectively) aftertreatment failure (Figure 1A). In GT3, slight increases ofL159F (5%) and V321A (2%) were observed, whereas bothvariants were not found in DAA-naïve patients (Figure 1B).Themedian sampling time after end of treatment (EOT)was 5to 6 months for both regimens. C316N is frequently associ-ated with L159F in GT1b and was also observed after failureto other sofosbuvir-containing regimens (data not shown).However, L159FandV321Aarenot associatedwith a reducedsofosbuvir susceptibility. The only known variant conferringsofosbuvir resistance in vitro is S282T, which was notdetected in any of these patients.

NS3 RASs After Simeprevir/SofosbuvirTreatment Failure

After failure to a simeprevir/sofosbuvir treatment,characteristic RASs developed within NS3. Patients withGT1a showed an increase of Q80K (56%) which confersonly low-level resistance to simeprevir in vitro but isassociated with a reduced treatment response in vivo.Moreover, medium- to high-level resistant R155 and D168variants were frequently detected (44% and 25%, respec-tively); R155K and D168E were most prominent(Figure 2A). In GT1b, the resistance profile was dominatedsolely by D168 variants (41%) and D168V was character-istic (Figure 2B). Patients infected with GT4 who did notachieve an SVR were rare (n ¼ 4) and here Q80R and D168E

/ribavirin±pegylated-interferon treatment in comparison withith GT1b (A) and in individuals with GT3 (B). The number ofas the median time of the RAS analysis in relation to the EOTare highlighted by red frames and major increasing RASs areecified according to Svarovskaia et al,52 and the number of

Figure 2. NS3 RASs detected in patients after simeprevir/sofosbuvir treatment in GT1a (A), GT1b (B), and GT4 (C) in com-parison with their frequencies in DAA-naïve patients. At the top of the figure the number of patients is included, the mediantime of the RAS analysis in relation to the end to treatment with the range stated in parentheses is indicated. Patterns of RASsare highlighted using red frames and major increasing RASs are designated in red letters. The resistance levels according toReferences 29, 30, and 53, as well as the number of sequences investigated are listed at the bottom.


CLINICAL

LIVE

R

were selected (Figure 2C). The median sampling time wasbetween 4 and 7 months after EOT for all GT.

NS5A RASs in Patients With Daclatasvir/Sofosbuvir Versus Ledipasvir/Sofosbuvir Failure

To investigate differences in the development of RASs,we compared the frequencies of NS5A RASs after failure todaclatasvir or ledipasvir in combination with sofosbuvir inpatients with GT1–4. In GT1a-infected patients, high-levelresistant RASs appeared at position Q30 after failure todaclatasvir/sofosbuvir and ledipasvir/sofosbuvir (71% and

52%, respectively) and Q30H/R variants were most com-mon. Moreover, L31 variants were detectable to a lesserextent (24% and 14%, respectively). However, differenceswere observed for both regimens, as moderate frequenciesof Y93 variants (30%) were characteristic for a ledipasvir/sofosbuvir failure and were infrequent after daclatasvir/sofosbuvir (Figure 3A).

The resistance profile in GT1b was comparable for bothregimens and L31 and Y93 variants increased (58% and83% after daclatasvir/sofosbuvir; 50% and 80% afterledipasvir/sofosbuvir), with L31M and Y93H being mostprevalent (Figure 3B). In GT3, high-level resistant Y93H was


CLINICALLIVER


CLINICAL

LIVE

R

also characteristic after daclatasvir/sofosbuvir administra-tion (66%), whereas it was completely lacking in ledipasvir/sofosbuvir-treated patients (Figure 3C). In GT4, infectionmainly occurred with subtypes 4a and 4d and only a limitednumber of patients could be investigated after daclatasvir/sofosbuvir failure (n ¼ 3). Interestingly, low-level resistantL30R was already detectable at relatively high frequenciesin DAA-naïve patients, especially in those with subtype 4d.After DAA failure, RASs at position L28 were prominent(66% daclatasvir/sofosbuvir and 39% ledipasvir/sofosbuvir), which mainly was caused by an increase of low-level resistant L28M. Moreover, Y93 variants appeared (33%daclatasvir/sofosbuvir and 39% ledipasvir/sofosbuvir) withY93C/H variants mainly detected (Figure 3D).

Within NS5B, S282T was found after daclatasvir/sofos-buvir or ledipasvir/sofosbuvir failure in a few patients only(for daclatasvir/sofosbuvir and GT3 n ¼ 2 [5%], for dacla-tasvir/sofosbuvir and GT4 n ¼ 1 [33%], for ledipasvir/sofosbuvir and GT1a n ¼ 1 [1%]; for ledipasvir/sofosbuvirand GT1b n ¼ 1 [1%], for ledipasvir/sofosbuvir and GT4n ¼ 2 [12%]).

The median sampling time point was 3 to 6 months afterEOT for all genotypes and regimens.

RASs in 3D/2D FailuresComponents of the 3D regimen are the protease inhibi-

tor (PI) paritaprevir/ritonavir, the NS5A inhibitor ombi-tasvir, and the non-nucleoside NS5B inhibitor dasabuvir,which is approved as triple combination (3D) for thetreatment of GT1 and as dual therapy (2D, consisting ofparitaprevir/ritonavir and ombitasvir) for GT4.

Within GT1a NS3, increasing RASs after 3D failure werelocated at position R155 and D168 (22% and 44%) andhere characteristic RASs were medium-level resistantR155K and D168V. In GT1b, high-level resistant D168V wasselected frequently in combination with Y56H. RegardingNS5A, in GT1a the frequencies of the most common RASswere 30% and 60% at positions 28 and 30. Here medium-to high-level resistant M28T/V and Q30H/R were mainlydetected. Of note, 41% of individuals with NS5A RASsharbored NS3 Q80K, which is a higher prevalence comparedwith 35% in DAA-naïve patients. In GT1b, medium-levelresistant Y93H was predominant (78%). Non-nucleosideNS5B RASs were infrequent within GT1a with slight in-creases of S556G and A553V observed, whereas the preva-lence of low-level resistant S556G was relatively high inGT1b (69%) (Figure 4A and B). The median sampling timewas 3 months after EOT for both subtypes.

So far, we could investigate only 4 patients infected withGT4 and a failure to the 2D regimen. In NS3, all patientscarried D168 variants and high-level resistant D168V wasmost frequent. The resistance pattern in NS5A was more

=Figure 3. Prevalence of NS5A RASs after daclatasvir/sofosbuvirnaïve patients in GT1a (A), GT1b (B), GT3 (C), and GT4 (D). The totime of the RAS analysis related to the EOT, and the range is indred frames with major increasing RASs labeled in red. Accordingat the bottom as well.

diverse with increasing RASs at position L28 (75%), andhere high-level resistant L28V was typically found(Figure 4C). For this limited number of patients, the sam-pling time points after EOT were varying.

Patterns of RASs and Retreatment OptionsWe identified typical patterns of RASs in individuals with

a failure to a DAA-based treatment. In PI-experienced pa-tients, RASs that were commonly selected after failure of asimeprevir- or a paritaprevir-based treatment were R155Kin GT1a and D168E/V in GT1b. In GT4, different RASs atposition Y56, Q80, and D168 appeared (Table 1). Moreover,more than 3 months after EOT, NS3 RASs were detectable inonly approximately one-half of PI-experienced patients withGT1b, whereas they were more frequent in GT1a (>80%including Q80 variants).

Regarding the NS5A inhibitor resistance profiles, weidentified Q30H/R variants in GT1a as typical treatment-selected RASs, which were commonly detected after adaclatasvir, ledipasvir, and ombitasvir-based treatment. Incontrast, the frequencies of Y93 variants were only mod-erate in GT1a and were variable according to the NS5A in-hibitor used (rare after daclatasvir and ombitasvir, morefrequent after ledipasvir). Altogether, the GT1a resistancepattern was diverse and included additional individual RASsfor each regimen. In GT1b, Y93H was selected using eachregimen. Interestingly, most patients had 1 NS5A RAS aftertreatment failure, except daclatasvir/sofosbuvir-experienced individuals with GT1b; here, 2 RASs werecommon, which can be attributed to the selection of L31M incombination with Y93H. Three RASs were not detected inGT1b-infected patients and were also rare in GT1a (Table 2).Within GT4, NS5A RASs mainly appeared at position L28and to a lesser extent at position Y93, but GT4 exhibits adiverse development of RASs according to the divergentsubtypes.

Finally, we evaluated the theoretical treatment andretreatment options in accordance with broadly approvedtreatment regimens available for the different genotypes. InGT1- and GT4-infected patients, RASs within NS3 plus NS5Amight impair a retreatment. For GT2 and 3, currently no PIsare available in many countries, and sofosbuvir/ribavirin isa suboptimal treatment for difficult-to-treat GT2- and allGT3-infected patients. Therefore, only NS5A RASs wouldrepresent a restriction for the retreatment of DAA-experienced patients. For all genotypes, S282T, which israrely selected in patients with sofosbuvir failure, mightlimit a sofosbuvir-based retreatment. We determined thatnearly all DAA-naïve individuals independent of the geno-type could be RASs-free treated with at least 1 regimen.

For DAA-experienced patients, multiple options areavailable for GT1. As NS3 plus NS5A RASs are infrequent in

or ledipasvir/sofosbuvir treatment failure compared with DAA-p of the figure shows the patient number included, the medianicated in parentheses. Patterns of RASs are highlighted usingto References 12 and 54 to 56, resistance levels are specified


CLINICALLIVER

Table 1.Summary of the Most Frequent Treatment-Selected RASs According to the HCV Genotype and Treatment RegimenThese characteristic RASs were defined to have a more than 10% increased prevalence after treatment failurecompared with DAA-naïve patients (exceptions are marked with asterisks).

GT1a GT1b GT2 GT3 GT4

NS3 NS5A NS5B NS3 NS5A NS5B NS5A NS5B NS5A NS5B NS3 NS5A NS5B

Simeprevir/sofosbuvir

R155Kc

D168Ebn.d. no RASs D168Vd n.d. L159Fb

C316Nbnot applicable not applicable Q80Rc

D168Ecn.d. no RASs

Daclatasvir/sofosbuvir

n.d. Q30H/Rd

L31Mdno RASs n.d. L31Mb

Y93HcL159FC316N

no patients Y93Hd S282Ta,b n.d. L28Mb S282Tb

Ledipasvir/sofosbuvir

n.d. Q30H/Rd

L31Md

Y93Hd

S282Ta,b n.d. L31Mc

Y93HdL159Fb

S282Ta,b

C316Nb

not applicable no RASs no RASs n.d. L28Mc

Y93C/HdS282Tb

3D/2D R155Kc

D168VdM28T/Vd

Q30RdS556Gb Y56Hb

D168VdY93Hc L159Fb

C316Nb

S556Gb

not applicable not applicable Y56Hc

D168VcL28Vc

Y93Hcn.d.

Sofosbuvir/ribavirin±pegylated-interferon

n.d. n.d. no RASs n.d. n.d. L159Fb

C316Nbn.d. no RASs n.d. L159Fa,b not applicable

n.d., not determined; not applicable, DAA regimen not approved for the respective genotype; no RASs, no RASs detected;no patients, DAA regimen is approved, but no samples were obtained.aIncrease in prevalence after treatment failure <10%, but not detected in DAA-naïve patients.b2-20-EC50 fold-change compared with wild-type replicon.c20- to 100-fold change.d>100-fold change.


CLINICAL

LIVE

R

patients who received sofosbuvir plus 1 additional DAA, aretreatment according to the presence of RASs is possiblefor more than 80% of patients, as RASs are usually lackingfor the alternative DAA target region. Major restrictions applyto 3D failures, inwhichmultipleRASs in all DAA target regionsenable a RASs-free retreatment only in 29% of patients. InGT3, Y93H is highly prevalent after daclatasvir/sofosbuvirfailure, and therefore only 16% of patients could be retreatedRASs-free, as alternative regimens not targeting NS5A arelacking so far. However, resistance development after ledi-pasvir/sofosbuvir is rare in GT3 because of the reducedantiviral activity, which enables a retreatment for 95% ofpatients. In GT4, RASs within NS3 plus NS5A are generallyuncommon. Only in the 2D-treated individuals analyzed sofar, NS3 plus NS5A RASs are frequent, precluding a RASs-freeretreatment for all patients (Figure 5).

DiscussionSo far, data on the prevalence of HCV RASs were mainly

available for treatment-naïve patients or for a small numberof patients with DAA failure who participated in approvalstudies. In the present analysis, we investigated RASs in thereal world in a large European resistance database

=Figure 4. Frequencies of NS3, NS5A, and NS5B non-nucleosidGT1a (A) and GT1b (B), as well as after 2D treatment failure in paThe overall number of patients is specified at the top of the figurto the EOT, and the range is indicated in parentheses. Patternmajor RASs were marked in red. The resistance levels accordsequences analyzed, are stated at the bottom.

comprising various patients with a DAA failure infected withHCV genotypes 1 to 4.

The development of RASs to sofosbuvir is rare. It seems,only S282T confers resistance in vitro and it was detected ina low number of patients after treatment failure only.22–24

We identified S282T in a few patients after daclatasvir/sofosbuvir or ledipasvir/sofosbuvir failure and it was rarein GT1 and GT3, but was more frequent in GT4. Moreover,L159F occurred in association with a sofosbuvir failure inGT1b and GT3 without conferring resistance in vitro, but thehalf maximal effective concentration was slightly increasedtogether with S282T.22,25,26 We observed slight increases ofL159F in GT3, but it was frequently selected in combinationwith C316N in GT1b-infected patients who failed to sofos-buvir/ribavirin or other sofosbuvir-containing regimens asreported in another study.27 Overall, the selection ofsofosbuvir RASs is rare, and the importance of selectedvariants remains unclear. However, it is important to notethat a retreatment with sofosbuvir seems to be associatedwith an increased likeliness for a treatment failure and se-lection of S282T.4,28

Similar to our study, after simeprevir/sofosbuvir failure,R155K and D168E were detected in GT1a, whereas in GT1bD168V was prominent and GT1b RASs vanished faster

e RASs in patients without a response to a 3D treatment intients with GT4 (C) in comparison with DAA-naïve individuals.e together with the median time of the RAS analysis in relations of RASs are highlighted using red frames and increases ofing to References 48 and 57 to 59, together with number of

Table 2.Comparison of the Number of NS5A RASs Selected in Patients With a Failure to an NS5A Inhibitor-ContainingRegimen

Number of NS5A RASs in DAA-failure patients

NS5A inhibitor

Daclatasvir Ledipasvir Ombitasvir

GT1a Patients without NS5A RASs 4 (24%) 17 (19%) 3 (10%)Patients with 1 NS5A RAS 7 (41%) 46 (51%) 20 (67%)Patients with 2 NS5A RASs 5 (29%) 25 (28%) 7 (23%)Patients with 3 NS5A RASs 1 (6%) 2 (2%) 0 (0%)NS5A sequences available 17 90 30

GT1b Patients without NS5A RASs 2 (17%) 6 (6%) 3 (17%)Patients with 1 NS5A RAS 2 (17%) 50 (53%) 14 (78%)Patients with 2 NS5A RASs 8 (66%) 38 (40%) 1 (6%)Patients with 3 NS5A RASs 0 (0%) 0 (0%) 0 (0%)NS5A sequences available 12 94 18


CLINICALLIVER
compared with GT1a.11,29 Four months after EOT, we did
not detect NS3 RASs anymore in more than one-half ofGT1b-infected patients. The few GT4 simeprevir/sofosbuvirfailure patients developed D168E, which was also identifiedin a pegylated-interferon-based study with simeprevir.30

Most likely, the relatively short half-life of RASs selectedby PIs explains high SVR rates in patients with PI-failurewho received retreatment regimens including a PI plussofosbuvir without an NS5A inhibitor.31

Most patients with DAA failure in this study weretreated with an NS5A inhibitor plus sofosbuvir and typical

and diverse NS5A RASs patterns were observed dependingon the NS5A inhibitor and the geno-/subtype. In GT1a,Q30H/R variants were selected after failure to both,daclatasvir or ledipasvir plus sofosbuvir. The selection ofY93 variants as major RASs was characteristic for ledi-pasvir/sofosbuvir, whereas only slight increases wereobserved after daclatasvir/sofosbuvir. Therefore, for aretreatment, the administration of second-generation NS5Ainhibitors like velpatasvir and pibrentasvir28,32–35 or thecombination of sofosbuvir, grazoprevir, elbasvir, andribavirin is feasible.36 However, in daclatasvir/sofosbuvir

Figure 5. (Re-)treatmentoptions with broadly avail-able DAAs according tothe presence of RASs, thepretreatment history, andthe genotype.


CLINICAL

LIVE

R

failures, the lack of Y93H may enable effective retreatmentoptions using ledipasvir or ombitasvir. Also, the combina-tion of a PI like simeprevir plus sofosbuvir with or withoutribavirin seems to be an effective retreatment option asnaturally occurring simeprevir high-level resistant RASsare rare.37–39 In GT1b, patterns of NS5A RASs were similarfor ledipasvir and daclatasvir, and, compared with GT1a,Y93H was highly prevalent. As Y93H variants seem topersist in the long-term follow-up and confer also ombi-tasvir resistance, for GT1b retreatment, the inclusion of aPI is probably mandatory and a rescue treatment withsimeprevir/sofosbuvir is also reasonable. Alternatively,regimens including a PI plus a second-generation NS5A-inhibitor could be used.28,32–36 Remarkably, we detectedhigher frequencies of NS5A RASs in GT1b compared with1a (80%–90% vs 70%–80%) and this was even morepronounced in the ledipasvir/sofosbuvir approval studies(90% vs 70%).7 Longer follow-up after virologic failure isrequired to determine potential differences in the half-lifeof NS5A RASs in GT1a vs 1b.

Failure to daclatasvir/sofosbuvir in GT3 was associatedwith a strong increase of Y93H, which was also shown in theapproval study.13 Remarkably, Y93H was not detected afterledipasvir/sofosbuvir, probably due to the reduced antiviralactivity of ledipasvir in GT3.40,41 Second-generation NS5Ainhibitors showed improved activities against GT3 isolatesharboring Y93H and recently also GT3-sensitive PIs wereapproved. SVR rates after a rescue treatment with sofosbuvirplus a second-generation PI like voxilaprevir or pibrentasvirare not available. However, although SVR rates after retreat-ment with sofosbuvir/velpatasvir/voxilaprevir or initialtreatmentwith glecaprevir/pibrentasvir were 95% and 97%,most failure patients were infected with GT3.35,42,43 Thus,retreatment of GT3 failure patients most likely will remain achallenge, and the addition of ribavirin may increase treat-ment efficacy in difficult-to-treat patients. In GT3-infectedpatients with decompensated cirrhosis, SVR rates to velpa-tasvir/sofosbuvir were substantially lower compared withthe group that additionally received ribavirin.44 Moreover, arecent study suggested that the addition of ribavirin toretreatment regimens, like daclatasvir or velpatasvir plussofosbuvir, could increase SVR rates.45 Thus, the addition ofribavirin may be an option for rescue treatments in difficult-to-treat GT3-infected patients as well as in DAA-naïvepatients in countries in which second-generation DAA regi-mens will not be available in the near future.

In patients with GT4, we identified NS5A L28M and to alesser extent Y93C/H/S as characteristic daclatasvir/sofos-buvir and ledipasvir/sofosbuvir failure RASs. Notably, themoderate frequency of Y93 variants enables a retreatmentwith velpatasvir/sofosbuvir, as these variants confer onlylow-level velpatasvir resistance.

The 3D regimen led to the selection of R155K in NS3 ofGT1a and D168V in GT1b, which was detected frequently incombination with Y56H, enhancing the resistance level. Thisresistance profile is similar to that of simeprevir, but Y56Hexclusively occurred after paritaprevir administration.Within GT1a NS5A, M28T/V and Q30H/R were mainlydetected and M28 variants are characteristic for an

ombitasvir treatment.46 As M28 variants were not enrichedafter daclatasvir or ledipasvir, this may enable ombitasvir asa rescue option and may explain relatively high SVR rates inpatients who received 3D as rescue treatment.37 In GT1b,Y93H was the key RAS also for ombitasvir, and thus for allfirst-generation NS5A inhibitors. NS5B non-nucleoside RASswere rare in GT1a, but in GT1b, S556G was selected incombination with C316N. In the pivotal trials, the preva-lence of GT1a RASs was comparable, but we detected higherfrequencies of RASs in GT1b, which is explainable by the lownumber of GT1b-infected patients investigated in thesestudies.46,47 In GT4, typical RASs detected after 2D failurewere D168V in NS3 and L28V in NS5A, whereas Y93 vari-ants were infrequent. In the approval study, the 3 patientswith a virologic failure developed similar RASs.48,49 As GT4is rare in Europe, more patients with virologic failure haveto be investigated to describe typical resistance patterns.

We considered (re)-treatment options with broadlyapproved DAAs based on RAS analyses. In DAA-naïve pa-tients, moderate RAS frequencies and several treatmentalternatives enable a RASs-free treatment for all GT. ForDAA-experienced patients, the retreatment options accord-ing to the RASs-free approach are decreasing. Interestingly,57% (GT1a), 93% (GT1b), and 100% (GT4) of patients withan NS5A inhibitor failure had no NS3 RAS, enabling aPI-based retreatment. Vice versa, 86% (GT1a) and 83%(GT1b) of PI-experienced individuals exhibited no NS5ARASs and might receive an NS5A inhibitor. Especially for 3Dfailures, alternatives are lacking, as RASs appear in multipleDAA targets and first-generation NS3 and NS5A inhibitorresistance profiles are overlapping. Another report alsodetected retreatment restrictions in patients with 3D failureand identified multiple RASs.50 However, a retreatmentusing 3D plus sofosbuvir is efficient, as shown in a smallstudy.51 Other possible or recently approved retreatmentoptions, like sofosbuvir/grazoprevir/elbasvir/ribavirin orsofosbuvir/velpatasvir/voxilaprevir, are highly effective inDAA-experienced patients and RASs had no apparent ef-fect.35,36 As these second-generation and multiple targetingregimens are not available in many countries, recently re-ported experiences with rescue treatments in small groupsof patients with DAA failure according to resistance profilesare of high importance. Data presented at recent interna-tional conferences showed high efficacies of this approachwith SVR rates of approximately 90%.37,39

In the present study, geno- and subtype-specific resis-tance profiles of a large number of patients with DAA failureare described enabling quantitative and comparative com-parisons. One limitation of this study is the noncontrolledanalysis of samples from patients with DAA failure in aretrospective study. This could favor the analysis of, forexample, difficult-to-treat patients with an urgent need ofretreatment and may lead to a sample selection bias. Thereported RAS frequencies therefore do not exactly representthe prevalence of RASs in European patientswith DAA failure.

In summary, the identified typical treatment-selectedresistance patterns for broadly currently used DAA regi-mens enable the selection of specific retreatment optionsbased on the result of resistance analyses. However,


CLINICALLIVER

resistance profiles for GT1b- and 3a-infected patientslargely overlapped, which restricts individualized RASs-freeretreatment options considerably.

Supplementary MaterialNote: To access the supplementary material accompanyingthis article, visit the online version of Gastroenterology atwww.gastrojournal.org, and at https://doi.org/10.1053/j.gastro.2017.11.007.

References
1. Sarrazin C. The importance of resistance to direct anti-
viral drugs in HCV infection in clinical practice. J Hepatol2016;64:486–504.

2. Gower E, Estes C, Blach S, et al. Global epidemiologyand genotype distribution of the hepatitis C virus infec-tion. J Hepatol 2014;61:S45–S57.

3. Pawlotsky JM. Hepatitis C virus resistance to direct-acting antiviral drugs in interferon-free regimens.Gastroenterology 2016;151:70–86.

4. Lawitz E, Flamm S, Yang JC, et al. Retreatment of pa-tients who failed 8 or 12 weeks of Ledipasvir/Sofosbuvir-based regimens with Ledipasvir/Sofosbuvir for 24 weeks.J Hepatol 2015;62:S192.

5. EASL. Recommendations on treatment of hepatitisC 2016. J Hepatol 2017;66:153–194.

6. Gane EJ, Stedman CA, Hyland RH, et al. Nucleotidepolymerase inhibitor sofosbuvir plus ribavirin for hepatitisC. N Engl J Med 2013;368:34–44.

7. Wyles D, Dvory-Sobol H, Svarovskaia ES, et al. Post-treatment resistance analysis of hepatitis C virus fromphase II and III clinical trials of ledipasvir/sofosbuvir.J Hepatol 2017;66:703–710.

8. Gane EJ, Hyland RH, An D, et al. Efficacy of ledipasvirand sofosbuvir, with or without ribavirin, for 12 weeks inpatients with HCV genotype 3 or 6 infection. Gastroen-terology 2015;149:1454–1461.e1.

9. Kwo P, Gitlin N, Nahass R, et al. Simeprevir plus sofos-buvir (12 and 8 weeks) in hepatitis C virus genotype1-infected patients without cirrhosis: OPTIMIST-1, aphase 3, randomized study. Hepatology 2016;64:370–380.

10. Lawitz E, Matusow G, DeJesus E, et al. Simeprevir plussofosbuvir in patients with chronic hepatitis C virus ge-notype 1 infection and cirrhosis: a phase 3 study(OPTIMIST-2). Hepatology 2016;64:360–369.

11. Lawitz E, Sulkowski MS, Ghalib R, et al. Simeprevir plussofosbuvir, with or without ribavirin, to treat chronicinfection with hepatitis C virus genotype 1 in non-responders to pegylated interferon and ribavirin andtreatment-naive patients: the COSMOS randomisedstudy. Lancet 2014;384:1756–1765.

12. Sarrazin C, Dvory-Sobol H, Svarovskaia ES, et al. Prev-alence of resistance-associated substitutions in HCVNS5A, NS5B, or NS3 and outcomes of treatment withledipasvir and sofosbuvir. Gastroenterology 2016;151:501–512.e1.

13. Nelson DR, Cooper JN, Lalezari JP, et al. All-oral 12-weektreatment with daclatasvir plus sofosbuvir in patients with

hepatitis C virus genotype 3 infection: ALLY-3 phase IIIstudy. Hepatology 2015;61:1127–1135.

14. Sulkowski MS, Gardiner DF, Rodriguez-Torres M, et al.Daclatasvir plus sofosbuvir for previously treated or un-treated chronic HCV infection. N Engl J Med 2014;370:211–221.

15. Poordad F, Schiff ER, Vierling JM, et al. Daclatasvir withsofosbuvir and ribavirin for hepatitis C virus infection withadvanced cirrhosis or post-liver transplantation recur-rence. Hepatology 2016;63:1493–1505.

16. Ferenci P, Bernstein D, Lalezari J, et al. ABT-450/r-ombitasvir and dasabuvir with or without ribavirin forHCV. N Engl J Med 2014;370:1983–1992.

17. Feld JJ, Kowdley KV, Coakley E, et al. Treatment of HCVwith ABT-450/r-ombitasvir and dasabuvir with ribavirin.N Engl J Med 2014;370:1594–1603.

18. Zeuzem S, Jacobson IM, Baykal T, et al. Retreatment ofHCV with ABT-450/r-ombitasvir and dasabuvir withribavirin. N Engl J Med 2014;370:1604–1614.

19. Susser S, Dietz J, Schlevogt B, et al. Origin, prevalenceand response to therapy of hepatitis C virus genotype 2k/1b chimeras. J Hepatol 2017;67:680–686.

20. Dietz J, Susser S, Berkowski C, et al. Consideration ofviral resistance for optimization of direct antiviral therapyof hepatitis C virus genotype 1-infected patients. PLoSOne 2015;10:e0134395.

21. Bartels DJ, Sullivan JC, Zhang EZ, et al. Hepatitis C virusvariants with decreased sensitivity to direct-acting anti-virals (DAAs) were rarely observed in DAA-naive patientsprior to treatment. J Virol 2013;87:1544–1553.

22. Tong X, Le Pogam S, Li L, et al. In vivo emergence of anovel mutant L159F/L320F in the NS5B polymeraseconfers low-level resistance to the HCV polymerase in-hibitors mericitabine and sofosbuvir. J Infect Dis 2014;209:668–675.

23. Svarovskaia ES, Dvory-Sobol H, Parkin N, et al. Infre-quent development of resistance in genotype 1–6 hep-atitis C virus-infected subjects treated with sofosbuvir inphase 2 and 3 clinical trials. Clin Infect Dis 2014;59:1666–1674.

24. Gane EJ, Abergel A, Metivier S, et al. The emergence ofNS5B resistant associated variant S282T aftersofosbuvir-based treatment. Hepatology 2015;62:322A.

25. Svarovskaia ES, Dvory-Sobol H, Doehle B, et al. L159Fand V321A sofosbuvir treatment-emergent HCV NS5Bsubstitutions. Hepatology 2014;60:218A.

26. Donaldson EF, Harrington PR, O’Rear JJ, et al. Clinicalevidence and bioinformatics characterization of potentialhepatitis C virus resistance pathways for sofosbuvir.Hepatology 2015;61:56–65.

27. Di Maio VC, Cento V, Lenci I, et al. Multiclass hep-atitis C virus resistance to direct acting antivirals inreal life interferon-free regimens failures advocatesfor tailored second-line therapies. J Hepatol 2017;66:S82–S83.

28. ZeuzemS,FlammS,TongM,etal.A randomized,controlled,phase 3 trial of sofosbuvir/velpatasvir/voxilaprevir orsofosbuvir/velpatasvir for 12 weeks in direct acting antiviral-experienced patients with genotype 1–6 HCV infection: thePOLARIS-4 study. Hepatology 2016;64:59A.

http://www.gastrojournal.org



http://refhub.elsevier.com/S0016-5085(17)36353-9/sref1









































































































































CLINICAL

LIVE

R

29. Lenz O, Verbinnen T, Fevery B, et al. Virology analyses ofHCV isolates from genotype 1-infected patients treatedwith simeprevir plus peginterferon/ribavirin in Phase IIb/IIIstudies. J Hepatol 2015;62:1008–1014.

30. Fevery B, Verbinnen T, Peeters M, et al. Virology ana-lyses of HCV genotype 4 isolates from patients treatedwith simeprevir and peginterferon/ribavirin in the Phase IIIRESTORE study. J Viral Hepat 2017;24:28–36.

31. Sulkowski MS, Vargas HE, Di Bisceglie AM, et al.Effectiveness of simeprevir plus sofosbuvir, with orwithout ribavirin, in real-world patients with HCV geno-type 1 infection. Gastroenterology 2016;150:419–429.

32. Bourlière M, Gordon SC, Ramji A, et al. Sofosbuvir/vel-patasvir/voxilaprevir for 12 weeks as a salvage regimenin NS5A inhibitor-experienced patients with genotype1–6 infection: the phase 3 POLARIS-1 study. Hepatology2016;64:102A.

33. Poordad F, Pol S, Asatryan A, et al. MAGELLAN-1, Part2: glecaprevir and pibrentasvir for 12 or 16 weeks inpatients with chronic hepatitis C virus genotype 1 or 4and prior direct-acting antiviral treatment failure.J Hepatol 2017;66:S83–S84.

34. Pilot-Matias T, Krishnan P, Schnell G, et al. Resistanceanalysis in the MAGELLAN-1 study (Part 2): glecaprevir/pibrentasvir therapy in HCV-infected patients who hadfailed prior DAA regimens containing NS3/4A proteaseand/or NS5A inhibitors. J Hepatol 2017;66:S708–S709.

35. Bourliere M, Gordon SC, Flamm SL, et al. Sofosbuvir,velpatasvir, and voxilaprevir for previously treated HCVinfection. N Engl J Med 2017;376:2134–2146.

36. de Ledinghen V, Laforest C, Hezode C, et al. Retreatmentwith sofosbuvir plus grazoprevir/elbasvir plus ribavirin ofpatients with hepatitis C virus genotype 1 or 4 whopreviously failed an NS5A- or NS3-containing regimen:the ANRS HC34 REVENGE study. Clin Infect Dis 2017;Epub ahead of print.

37. Vermehren J, Susser S, Dietz J, et al. Retreatment ofpatients who failed DAA-combination therapies: real-world experience from a large hepatitis C resistancedatabase. J Hepatol 2016;64:S188.

38. Hezode C, Chevaliez S, Scoazec G, et al. Retreatmentwith sofosbuvir and simeprevir of patients with hepatitisC virus genotype 1 or 4 who previously failed adaclatasvir-containing regimen. Hepatology 2016;63:1809–1816.

39. Chevaliez S, Trimoulet P, Dorival C, et al. Effect ofresistance-associated substitutions on retreatment ofdirect acting antiviral-exposed patients in the real-worldsetting (ANRS CO22 HEPATHER). J Hepatol 2017;66:S84.

40. Cheng G, Tian Y, Doehle B, et al. In vitro antiviral activityand resistance profile characterization of the hepatitis Cvirus NS5A inhibitor ledipasvir. Antimicrob Agents Che-mother 2016;60:1847–1853.

41. Gane EJ, Kowdley KV, Pound D, et al. Efficacy of sofos-buvir, velpatasvir, and GS-9857 in patients with hepatitisc virus genotype 2, 3, 4, or 6 infections in an open-label,phase 2 trial. Gastroenterology 2016;151:902–909.

42. Sarrazin C, Cooper C, Manns M, et al. No impact ofRASs on the high efficacy of SOF/VEL/VOX for 12 weeks

in DAA-experienced patients: an integrated resistanceanalysis of the POLARIS-1 and POLARIS-4 studies.J Hepatol 2017;66:S299.

43. Krishnan P, Schnell G, Tripathi R, et al. Pooled resistanceanalysis in HCV genotype 1–6-infected patients treatedwith glecaprevir/pibrentasvir in phase 2 and 3 clinicaltrials. J Hepatol 2017;66:S500.

44. Curry MP, O’Leary JG, Bzowej N, et al. Sofosbuvir andvelpatasvir for HCV in patients with decompensatedcirrhosis. N Engl J Med 2015;373:2618–2628.

45. Vermehren J, Dietz J, Peiffer KH, et al. High sustainedvirologic response rates in HCV genotype 3 patients withand without cirrhosis treated with daclatasvir/sofosbuviror velpatasvir/sofosbuvir ± ribavirin according to base-line resistance analysis. J Hepatol 2017;66:S83.

46. Krishnan P, Tripathi R, Schnell G, et al. Resistanceanalysis of baseline and treatment-emergent variants inhepatitis C virus genotype 1 in the AVIATOR study withparitaprevir-ritonavir, ombitasvir, and dasabuvir. Anti-microb Agents Chemother 2015;59:5445–5454.

47. Krishnan P, Tripathi R, Schnell G, et al. Pooled analysisof resistance in patients treated with ombitasvir/ABT-450/r and dasabuvir with or without ribavirin in Phase2 and Phase 3 clinical trials. Hepatology 2014;60:1134A–1135A.

48. Schnell G, Tripathi R, Beyer J, et al. Hepatitis C virusgenotype 4 resistance and subtype demographic char-acterization of patients treated with ombitasvir plusparitaprevir/ritonavir. Antimicrob Agents Chemother2015;59:6807–6815.

49. Hezode C, Asselah T, Reddy KR, et al. Ombitasvir plusparitaprevir plus ritonavir with or without ribavirin intreatment-naive and treatment-experienced patients withgenotype 4 chronic hepatitis C virus infection (PEARL-I): arandomised, open-label trial. Lancet 2015;385:2502–2509.

50. Di Maio VC, Cento V, Lenci I, et al. Multiclass HCVresistance to direct-acting antiviral failure in real-lifepatients advocates for tailored second-line therapies.Liver Int 2017;37:514–528.

51. Poordad F, Bennett M, Sepe TE, et al. Ombitasvir/paritaprevir/r, dasabuvir, and sofosbuvir treatment ofpatients with HCV genotype 1-infection who failed aprior course of DAA therapy: the Quartz-I study.J Hepatol 2016;64:S767–S768.

52. Svarovskaia ES, Gane E, Dvory-Sobol H, et al. L159Fand V321A sofosbuvir-associated hepatitis C virus NS5Bsubstitutions. J Infect Dis 2016;213:1240–1247.

53. Lenz O, Verbinnen T, Lin TI, et al. In vitro resistanceprofile of the hepatitis C virus NS3/4A proteaseinhibitor TMC435. Antimicrob Agents Chemother 2010;54:1878–1887.

54. Fridell RA, Wang C, Sun JH, et al. Genotypic andphenotypic analysis of variants resistant to hepatitis Cvirus nonstructural protein 5A replication complex in-hibitor BMS-790052 in humans: in vitro and in vivo cor-relations. Hepatology 2011;54:1924–1935.

55. Zhou N, Hernandez D, Ueland J, et al. NS5A sequenceheterogeneity and mechanisms of daclatasvir resistancein hepatitis C virus genotype 4 infection. J Infect Dis2016;213:206–215.




















































































































































CLINICALLIVER

56. Hernandez D, Zhou N, Ueland J, et al. Natural prevalenceofNS5Apolymorphisms in subjects infectedwith hepatitisC virus genotype 3 and their effects on the antiviral activityof NS5A inhibitors. J Clin Virol 2013;57:13–18.

57. Pilot-Matias T, Tripathi R, Cohen D, et al. In vitro andin vivo antiviral activity and resistance profile of thehepatitis C virus NS3/4A protease inhibitor ABT-450.Antimicrob Agents Chemother 2015;59:988–997.

58. Krishnan P, Beyer J, Mistry N, et al. In vitro and in vivoantiviral activity and resistance profile of ombitasvir, aninhibitor of hepatitis C virus NS5A. Antimicrob AgentsChemother 2015;59:979–987.

59. Kati W, Koev G, Irvin M, et al. In vitro activity and resis-tance profile of dasabuvir, a nonnucleoside hepatitis Cvirus polymerase inhibitor. Antimicrob Agents Chemother2015;59:1505–1511.

Received May 26, 2017. Accepted November 3, 2017.

Reprint requestsAddress requests for reprints to: Christoph Sarrazin, MD, Department ofInternal Medicine 1, Theodor-Stern-Kai 7, D-60590 Frankfurt, Germany.e-mail: [email protected]; fax: þ49–69–6301–84441.

Conflicts of interestThese authors disclose the following: Johannes Vermehren: Speaking and/orconsulting fees from Abbott, AbbVie, Gilead, Bristol-Myers Squibb,Medtronic, Merck/MSD, and Roche. Georgios Grammatikos: researchsupport: Gilead.Peter Ferenci: Speaking and/or consulting fees: Merck/MSD,Gilead, AbbVie, BMS. Research support: Gilead. Speaker: Gilead, AbbVie,BMS. Maria Buti: Speaking and/or consulting fees: Gilead, Merck/MSD,AbbVie, Janssen. Clinical Trials: Gilead, Merck/MSD, AbbVie, Janssen. BeatMüllhaupt: Speaking and/or consulting fees: Merck/MSD, Janssen, AbbVie,Boehringer Ingelheim, Intercept, Astra, Bayer, BMS, Gilead. Researchsupport: Gilead. Bela Hunyady: Speaking and/or consulting fees: AbbVie,Boehringer Ingelheim, BMS, Fresenius-Kabi, Gilead, Janssen, Merck/MSD,Roche. Stefan Mauss: Speaking and/or consulting fees: AbbVie, BMS,Gilead, Janssen, Merck/MSD, ViiV. Jörg Petersen: Clinical studies: AbbVie,Arrowhead, BMS, Eisai, Falk, Gilead, Hepatera, Hologic, Intercept, Janssen,Merck/MSD, Roche, Siemens, Vertex. Speaking and/or consulting fees:Abbott, AbbVie, Arrowhead, Assembly Pharma, BMS, Contravir, Gilead,GSK, Intercept, Kedrion, Janssen, Merck/MSD, Novira, Roche. Sponsoredlectures: Abbott, BMS, Boehringer, Gilead, Intercept, Kedrion, Janssen,Merck, Merz, MSD, Novartis, Roche. Research Support: BMS, Novartis,Roche. Peter Buggisch: Speaking and/or consulting fees: AbbVie, BMS, Falk,Gilead, Janssen, Merz Pharma, Merck/MSD. Gisela Felten: Speaking and/orconsulting fees: Merck/MSD, AbbVie Dietrich Hüppe: Speaking and/orconsulting fees: Merck/MSD, Janssen, Roche, BMS, Falk, BoehringerIngelheim, Novartis, Norgine, Abbott/AbbVie, Gilead, Ferring, MerckleRecordati, Echosens. Gaby Knecht: Grants: Gilead. Thomas Lutz: Grants:BMS, AbbVie, Gilead, ViiV, Janssen, Merck/MSD, Deutsche Leberstiftung.Eckart Schott: Speaking and/or consulting fees: AbbVie, BMS, Gilead,Merck/MSD Christoph Berg: Speaking and/or consulting fees: AbbVie, BMS,Gilead, Merck/MSD. Ulrich Spengler: Speaking and/or consulting fees:AbbVie, BMS, Falk, Gilead, Intercept, Merck/MSD. Thomas von Hahn:Speaking and/or consulting fees: AbbVie, BMS, Janssen. Thomas Berg:Speaking and/or consulting fees: AbbVie, Alexion, Bayer, BoehringerIngelheim, BMS, Gilead, GSK, Intercept, Janssen, MSD/Merck, Merz,Novartis, Sequana Medical and Roche. Research support: AbbVie, Roche,BMS, Gilead, Novartis, Merck/MSD, Intercept, Janssen, Novartis, SequanaMedical, and Pfizer. Stefan Zeuzem: Speaking and/or consulting fees:Abbvie, BMS, Gilead, Janssen, Merck/MSD. Christoph Sarrazin: Speakingand/or consulting fees: Abbott, AbbVie, BMS, Gilead, Intercept, Janssen,Merck/MSD, Roche. Research support: Abbott, Gilead, Janssen, Roche,Siemens. Speaker: Abbott, AbbVie, BMS, Gilead, Intercept, Janssen, Merck/MSD, Roche, Siemens. The remaining authors disclose no conflicts.

FundingThis study was supported by a DZIF German Center for Infection Researchgrant entitled “Geno- & phenotypic NS3, NS5A and NS5B inhibitorresistance analysis” to Christoph Sarrazin (TTU 05.902).

Other members of the European HCV Resistance Study group:

University Hospitals:C. Antoni, R. Vogelmann, M. Ebert (Mannheim, Germany); J. Backhus, T.Seufferlein (Ulm, Germany); J. Balavoine, E. Giostra (Genf, Switzerland); C.Berg (Tübingen, Germany); M. Cornberg, H. Wedemeyer, M. Manns(Hannover, Germany); A. De Gottardi (Bern, Switzerland); R. Esteban(Barcelona, Spain); T. Discher, E. Roeb (Gießen, Germany); M. Gress(Marburg, Germany); R. Günther, P. Wietzke-Braun (Kiel, Germany); A.Herrmann, A. Stallmach (Jena, Germany); D. Hoffmann (TU, München,Germany); H. Klinker (Würzburg, Germany); A. Kodal (Lübeck, Germany); F.Lammert (Homburg, Germany); M. Löbermann (Rostock, Germany); J.Schulze zur Wiesch, J. von Felden, F. Piecha, A. Lohse (Hamburg, Germany);T. Götze, P. Malfertheiner (Magdeburg, Germany); J. Mayerle (LMU,München, Germany); D. Moradpour (Lausanne, Switzerland); C. Moreno(Brüssel, Belgium); C Neumann-Haefelin, R. Thimme (Freiburg, Germany); L.Reinhardt, V. Ellenrieder (Göttingen, Germany); J. Schattenberg, M. Sprinzl,P. Galle (Mainz, Germany); J. Schmidt (Münster, Germany); E. Schott, H.-J.Epple (Berlin); J. Siebler (Erlangen); R. Stauber (Graz, Austria); M. Steckstor,W. Schmiegel (Bochum, Germany); W. Stremmel (Heidelberg); B. Strey(Duisburg), K. Tomasiewicz (Lublin, Poland); C. Trautwein (Aachen,Germany), R. Zachoval (LMU, München, Germany).

Academic Hospitals: W. Angeli (Kempten, Germany); S. Beckebaum(Essen, Germany); C. Doberauer (Köln, Germany); K. Ende (Erfurt,Germany); A. Erhardt (Wuppertal, Germany); A. Garrido-Lüneburg(Bad Oeynhausen, Germany); H. Gattringer (Steyr, Austria); D. Genné(Biel, Switzerland), M. Gschwantler (Vienna, Austria), F. Gundling(München, Germany); C. Hartmann (Esslingen, Germany); T. Heyer (Düren,Germany), C. Hirschi (Luzern, Switzerland); S. Kanzler (Schweinfurt,Germany); N. Kordecki (Lingen, Germany); M. Kraus (Altötting-Burghausen,Germany); U. Kullig (Dresden, Germany); L. Magenta, B. Terziroli Beretta-Piccoli (Locarno, Switzerland); M. Menges (Schwäbisch Hall, Germany); L.Mohr (Lahr, Germany); K. Muehlenberg (Regensburg, Germany); C.Niederau (Oberhausen, Germany); B. Paulweber (Salzburg, Austria); A.Petrides (Bochum, Germany); R. Piso (Olten, Switzerland); W. Rambach(Bad Kissingen, Germany); M. Reiser (Marl, Germany); B. Riecken(Ludwigsburg, Germany); J. Roth (Lauterbach, Germany); R. Schöfl, A.Maieron (Linz, Austria); A. Schneider (Korbach, Germany); M. Schuchmann(Konstanz, Germany); U. Schulten-Baumer (Eggenfelden, Germany); A.Seelhoff (Berlin, Germany); D. Semela (St. Gallen, Switzerland), A. Stich(Würzburg, Germany); C. Vollmer, J. Brückner (Augsburg, Germany); J.Ungemach (Frankfurt, Germany); E. Walter (Villingen-Schwenningen,Germany); A. Weber (Nürnberg, Germany); T. Winzer (Bad Hersfeld,Germany).

Local study sites (private practices), Germany:W. Abels (Nürnberg); M. Adler (Braunschweig); F. Audebert (Regenburg); C.Baermann (Freiburg); E. Bästlein (Köln); R. Barth (Nürnberg); K. Barthel(Leipzig); K. Baumgartl (Pfungstadt); W. Becker (München); J. Benninger(Regensburg); T. Beyer (Rosenheim); A. Bodtländer (Frankfurt); G. Böhm(Ludwigshafen); U. Bohr, A. Moll, U. Naumann (Berlin); N. Börner (Mainz);H.R. Bruch (Bonn); N. Busch (Würselen); O. Burkhard (Worms); C. Chirca(Bad Reichenhall); A. Dienethal (Frankfurt); P. Dietel (Leipzig); F. Dreher(Rottenburg); P. Efken (Minden); U. Ehrle (Pfungstadt); F. Emke (Osnabrück);J. Fischer (München); U. Fischer (Aalen); D. Frederking (Hamburg); B. Frick(Darmstadt); B. Gantke B. (Düsseldorf); P. Geyer (Fulda); T. Glaunsinger(Berlin);F. Goebel (München); U. Göbel (Cottbus); R. Graf (Leutkirch); M.Herder (Urbach); T. Heuchel (Chemnitz); S. Heuer (Bielefeld); R. Heyne(Berlin); K.H. Höffl (Traunstein); W. P. Hofmann (Berlin), F. Holst (Marburg);H. Hörster (Mönchengladbach); C. John (Berlin); M.C. Jung (München); B.Kallinowski (Schwetzingen); W. Kerzel (Forchheim); P. Khaykin (Frankfurt);M. Klarhof (Ettlingen); B. Knapp (Siegen); U. Knevels (Unna); A. Körfer(Peine); A. Köster (Friedrichshafen); B. Künzig (Waiblingen); A. Langenkamp,M. Kuhn (Kassel); M. Littman (Bad Saulgau); H. Löhr (Wiesbaden); L.Ludwig (Dornstadt); U. Lutz (Singen); P. Maurer (Bühl); C. Mayer (Marburg);V. Meister (Vechta); D. Moritzen (Bielefeld); M. Mroß (Berlin); M. Mundlos(Verden); O. Nehls (Stuttgart); K. Ningel (Nürtingen); A. Oelmann(Paderborn); H. Olejnik (Goch); E. Pascher (Herrieden); A. Philipp(Recklinghausen); M. Pichler (Stuttgart); F. Polzien (Braunschweig); R.Raddant (Geldern); M. Riedel (Köln); S. Rietzler (Albstadt Ebingen); A. Rump(Freiburg); W. Schmidt (Berlin), J. Schmidtler-von Fabris (Stadtbergen); L.Schneider (Fürth); A. Schober (Göttingen); J. Schwenzer (Berlin); T. Seidel(Weimar); G. Seitel (Karlsruhe); C. Sick (Bremen); K. Simon (Leverkusen); D.Stähler (Köln); H. Steffens (Berlin); B. Strey (Duisburg); K. Svensson(Bremen); W. Tacke (Königstein); K. Teubner (Suttgart); J. Thieringer(Frankfurt); U. Trappe (Hamm); J. Ullrich (Krefeld); S. Usadel (Freiburg); A.von Lucadou (Nürnberg); M. Werheid-Dobers (Bergisch-Gladbach); E.Zehnter (Dortmund); A. Zipf (Mannheim).





















mailto:[email protected]

* includes treatment-naive and PEG-IFN/RBV pre-treated patients** includes patients treated with other DAAs from clinical studies***published by Susser et al. J Hepatol. 2017

Overall n=4240 patientswith resistance analysis

n=2948 remainingindividuals

Insufficient data on treatment status/other**

(n=914)

Patients with BOC/ TVR+PEG/RBV failure

(n=325)

Patients with availabletreatment infomation (n=3326)

Patients with special GT:GT 1c, 1e, 1l (n=6)GT2k/1b (n=33)***

GT5 (n=8)GT6 (n= 6)

DAA-naive patients* n=2322

DAA-experienced patients (GT1-4)n=626

SOF/PEG/RBV failure (n=63)

SOF/RBV failure (n=132)

SMV/SOF ± RBV failure (n=55)

DCV/SOF ± RBV failure (n=89)

LDV/SOF ± RBV failure (n=232)

2D/3D ± RBV failure (n=55)

Supplementary Figure 1. Flow diagram of the overall patientcohort involved in the study.

Supplementary Table 1.Primers and PCR Conditions for the Amplification of the NS3 Protease

Primera GT Sequence (50 to 30) Annealing temperature and cycles

NS3-2F All GT TGTAAAACGACGGCCAGTATGGARAYYAAGVTYATYACNTGGG Outer: 53�C (45 cycles)NS3-1F All GT TGTAAAACGACGGCCAGTATGGARAAGAARRTYATYRTNTGGGNS3-1R All GT CAGGAAACAGCTATGACCCTYTTNCCRCTNCCNGTNGGNGCRTGNested NS3-1F All GT TGTAAAACGACGGCCAGTATCTTNCTNGGNCCNGCYGA Inner: 53�C (5 cycles), 67�C (35 cycles)Nested NS3-2F All GT TGTAAAACGACGGCCAGTATACTNCTNGGNCCNGCNGANested NS3-1R All GT CAGGAAACAGCTATGACCGCNACYTGRTANGTNTGNGG

aSequences adapted according to Besse et al., J Virol Methods, 2012;185:94-100.

March 2018 HCV RASs in European DAA-experienced Patients 988.e1

Supplementary Table 2.Primers and PCR Conditions for NS5A Amplification

Primer GT Sequence (50 to 30) Annealing temperature and cycles

NS5A_1b_5890_F 1b ACATTCTGGCGGGCTATGG Outer: 52�C (45 cycles)NS5A_1a_5890_F 1a ACATCCTTGCAGGGTATGGNS5A_1b_7378_R 1b TCCGCCAAGGCAGAAGACACNS5A_1a_7274_R 1a GACCACAGGTGGTTCGTAGNS5A_1a/b_6084_F 1a/1b TGGATGAACCGGCTGATAG Inner: 51�C (45 cycles)NS5A_1a_7063_R 1a ATGTTGCCGCCCATCTCNS5A_1b_7063_R 1b ATGTTCCCGCCCATCTCGT2ac_NS5A_out_f 2a/c GCGGTCCAATGGATG Outer: 48�C (45 cycles)GT2ac_NS5A_out_r 2a/c TGGCRTCSACCATGTCGT2ac_NS5A_inn_f 2a/c CCTTYGCKTCCAGAG Inner: 48�C (45 cycles)GT2ac_NS5A_inn_r 2a/c GCCACCCTCCAGATGGT2b_NS5A_out_f 2b CGGTCCAGTGGATGAAC Outer: 51�C (45 cycles)GT2b_NS5A_out_r 2b GCAAGGGAGCTGAGAGGT2b_NS5A_inn_f 2b GTCCAGTGGATGAACAG Inner: 51�C (5 cycles), 49�C (40 cycles)GT2b_NS5A_inn_r 2b CCGTCCACCCAAGAGNS5A-GT3-out-for 3 ACGCACTATGTTCCCGAGAG Outer: 54�C (45 cycles)NS5A-GT3-out-rev 3 TGATGGGCAGTTTCTCTTCCNS5A-GT3-inn-for 3 CGGTTACACCAGTGGATCAATG Inner: 52�C (45 cycles)NS5A-GT3-inn-rev 3 CGCCGGTCCAAGAATACGT4adr_NS5A_out_f 4 ATCATGAGCGGCGAG Outer: 46�C (45 cycles)GT4adr_NS5A_out_r 4 CTCAACCCTGGTAGCGT4adr_NS5A_inn_f 4 GTGACHTCCCTTCTCAG Inner: 46�C (5 cycles), 48�C (40 cycles)GT4adr_NS5A_inn_r 4 ACACCCTCCAAAGCG

Supplementary Table 3.Primers and PCR Conditions for NS5B Amplification (Nucleotide Inhibitor Amino Acid Positions159–321)


RZ_1b_NS5_out_fw GT 1, 2, 3, 4 CTCCGTGTGGRAGGACTTG Outer: 56�C (45 cycles)RZ_1a_NS5_out_fw CCGTGTGGAARGACCTTCTGP3_8616_reva GGCGGAATTCCTGGTCATAGCCTCCGTGAARZ_1b_NS5_out_fw GT 1, 2, 3, 4 CTCCGTGTGGRAGGACTTG Inner: 56�C (45 cycles)RZ_1a_NS5_out_fw CCGTGTGGAARGACCTTCTGP6_8611_reva AATTCCTGGTCATAGCCTCCGTGAAGACTC

aPrimer sequence adapted from Laperche et al., J Clin Microbiol, 2005;43:733-739.

988.e2 Dietz et al Gastroenterology Vol. 154, No. 4

Supplementary Table 4.Primers and PCR Conditions for NS5B Amplification (Non-Nucleoside Inhibitor Amino Acid Positions368–559)


P4_8253_fora,b GT1a/1b CCGTATGATACCCGCTGCTTTGACTCAAC Outer and inner: 65�C (45 cycles)P5_8253_fora,b GT1a/1b TCCTACGACACCAGGTGCTTTGATTCAACNS5B_1a_1702_Ra GT1a/1b GGGCATGAGACACGCTGTGNS5B_1b_1700_Ra GT1a/1b GCACGAGACAGGCTGTG

aPrimers used for both outer and inner PCRs.bSequence adapted from Laperche et al. J Clin Microbiol, 2005;43:733-739.

Supplementary Table 5.Sequencing Primers

Primer Gene GT Sequence (50 to 30)

NS3 M13 forward NS3 all GT GTAAAACGACGGCCAGNS5A_1a/b_6084_F NS5A GT1a/1b TGGATGAACCGGCTGATAGGT2ac_NS5A_inn_f NS5A GT2a/c CCTTYGCKTCCAGAGGT2b_NS5A_inn_f NS5A GT2b GTCCAGTGGATGAACAGNS5A-GT3-inn-for NS5A GT3 CGGTTACACCAGTGGATCAATGGT4adr_NS5A_inn_f NS5A GT4 GTGACHTCCCTTCTCAGP6-8611_rev NS5Ba GT 1, 2, 3, 4 AATTCCTGGTCATAGCCTCCGTGAAGACTCNS5B_1a/1b_seq1F NS5Bb GT1a/1b AGAACTGCGGYTATCGNS5B_1a/1b_seq2F NS5Bb GT1a/1b TACTACCTCACCCGTGACC

aFor sequencing of NS5B nucleotide inhibitor sites: amino acid positions 159–321.bFor sequencing of NS5B non-nucleoside inhibitor sites: amino acid positions 368–561.

March 2018 HCV RASs in European DAA-experienced Patients 988.e3

Supplementary Table 6.RASs With Greater Than Twofold Half Maximal Effective Concentration Analyzed Within NS3, NS5A,and NS5B of Different Genotypesa

Gene GT

NS3 1a F43L, Y56H, Q80K/R, R155G/K/T, D168A/E/H/N/YNS3 1b F43I/S/V, Y56H, Q80K/R, S122R, R155K/G/T, A156G/T/V, D168A/E/G/H/I/N/T/VYNS5A 1a K24G/N/R, M28A/G/T/V, Q30E/G/H/K/L/R/T, L31I/M/V, P32L, S38F, H58D, A92T, Y93C/F/H/L/N/R/S/T/WNS5A 1b L28T, R30H, L31I/F/M/V, A92K, Y93H/N/TNS5A 2 F28C/S, L31M, C92R, Y93HNS5A 3 A30E/K, L31F/M/V, Y93HNS5A 4 L28M/V, L30C/H/R/S/Q, M31I/L/V, P32L, Y93C/H/S/WNS5B (non-nucleoside) 1a C316Y, M414T, E446K/Q, Y448C/H, A553T, G554S, S556G/N/R, G558R, D559G/I/N/V, Y561HNS5B (non-nucleoside) 1b C316H/N/Y, S368T, M414I/T, Y448C/H, A553V, S556G/R, D559G/N,NS5B (nucleotide) 1,3 L159F, S282T, L320F, V321A

aFold change resistance levels are an indication but are dependent on the type of the assay used and are not directlycomparable between different studies.

988.e4 Dietz et al Gastroenterology Vol. 154, No. 4

Patterns of Resistance-Associated Substitutions in … of patterns of resistance-associated substitutions (RASs) selected after treatment failure to the different DAA-based regimens

Documents