What do we need to know about RAVs clinicallyregist2.virology-education.com/2016/14EU/27_Zeuzem.pdf · 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30. LDV RAVs Class RAVs Y93H. GT1a. Sensitivity

What do we need to knowabout RAVs clinically ?

Stefan Zeuzem, MDUniversity of Frankfurt

Germany

14th European HIV & Hepatitis WorkshopRome, 25-27 May, 2016

Background• Resistance associated variants (RAVs) or resistance associated

polymorphisms (RAPs) with reduced drug susceptibility are observed in patients with chronic HCV

– Due to the high replication rate of HCV (1012 / day)– Error-prone RNA-dependent RNA polymerase

(10-4 to 10-5 per copied nucleotide)

• All possible single and double mutants are predicted to be generatedmultiple times each day

• Not all RAVs are fit enough to persist (but potentially under selectionpressure of a respective DAA). Most NS3A RAVs disappear after EoT(HCV1b > HCV1a), while many NS5A RAVs persist

• The level of resistance in vivo may differ between a RAV in previouslyuntreated patient and the same RAV selected by a DAA (= TEV) (e.g. due to compensatory mutations)

Rong et al., Sci Transl Med 2010

Zeuzem et al., Ann Intern Med 2015Jacobson et al., AASLD 2016

Group GTPosition

K24 M28 (L28)†

Q30 (R30)* L31 P32 S38 H58

(P58)‡ A92 Y93

EBR RAVs1a -¶ A, G, T D, E, H,

G K, L, R F, M, V -¶ -¶ D -¶ C, H, N, S

1b -¶ Any Any Any L -¶ D K AnyNS5A Class RAVs 1a or 1b Any Any Any Any Any Any Any Any Any†Wild type amino acid at position 28 is M in GT1a and L in GT1b*Wild-type amino acid at position 30 is Q in GT1a and R in GT1b.‡Wild type amino acid at position 58 is H in GT1a and P in GT1b¶No EBR RAVs identified at this position

Note that EBR RAVs are a subset of the NS5A class RAVs

How to define a RAV ?

• RAV per specific drug or drug class(e.g. resistance against elbasvir or any NS5A inhibitor)

• Genotypic backbone in vitro and in vivo(e.g. genotype 1a or 1b replicons, other genotypes)


• RAV per specific drug or drug class(e.g. resistance against ledipasvir or any NS5A inhibitor)


• Level of reduced susceptibility in vitro(e.g. any aa change vs >2.5-fold vs >100-fold)

NS5A RAVs - Overview

GT 1a GT1bM28T Q30R L31M/V Y93H/N L31V Y93H/N

Ledipasvir >20x >100x >100x >1,000x >1,000x/-

Ombitasvir >1,000x >100x<3x

>10,000x <10x >50x>100x

Daclatasvir >100x >1,000x >100x >1000x

>1,000x >10,000x >20x >20x

Elbasvir >10x >10x>10x

>100x <10x >10x/->100x

Velpatasvir <10x <3x >20x >100x>1,000x <3x/-

Odalasvir >20x <10x <3x >1,000x <3x<3x

<10x

ABT-530 <3x <3x <3x <10x <3x <3x

MK-8408 <10x <10x <10x <10x <10x <10x

Cheng et al., EASL 2012; Wong et al., AAC 2013; Krishnan et al., AAC 2015; Fridell et al., Hepatology 2011; Liu et al., AAC 2015; Cheng et al., EASL 2013; Patel et al., EASL 2015; Ng et al., CROI 2014; Asante-Appiah et al., AASLD 2014

EC50 fold-change compared to WT replicon





• Methodology of RAV detection (sensitivity threshold)(e.g. deep sequencing with a cut-off at 1% orpopulations sequencing with a cut-off at 15-25%)

G T 1 b

S e n s it iv ity T h re s h o ld (% )

Pre

va

len

ce

(%

)

0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 00

1 0

2 0

3 0

L D V R A V sC la s s R A V s

Y 9 3H

G T 1 a

S e n s it iv ity T h re s h o ld (% )

Pre

va

len

ce

(%

)

0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 00

1 0

2 0

3 0

L D V R A V s

C la s s R A V s

Y 9 3H

N=3509 N=1887

GT1a Position LDV RAVs NS5A Class RAVs

24 G/N/R G/N/R26 E28 A/G/T A/G/T/V30 E/G/H/L/K/R/T C/E/G/H/I/L/K/R/S/T/Y31 I/F/M/V I/F/M/V32 L L38 F F58 D D/L92 K/T K/T93 C/F/H/N/S C/F/H/L/N/R/S/T/W

GT1b Position LDV RAVs NS5A Class RAVs

28 M31 I/F/M/V I/F/M/V32 L L58 D D92 K K93 C/H/N/S C/H/N/S

Methodology of RAV detection

Date on file, GileadHongmei Mo, personal communication

Frequency and characteristics of RAVs in treatment-naive and DAA-experienced patients - European RAVs database: Serum samples of 3549 European HCV-infected patients Population-based sequencing for NS3, NS5A, and NS5B Considered relevant if associated with treatment failure or shown to confer

>2-fold changed drug susceptibility in comparison to reference strain

No RAVs detected in G1a or G2

Dietz J, et al. EASL 2016, Barcelona. #PS007

RAVs in SOF/PR and SOF/RBV failures

16%

40%32%

55%

0%

20%

40%

60%

80%

L159

F

C316

H/N

Freq

uenc

y of

RAV

s*

DAA-naive (n=706)SOF/PR; SOF/RBV (n=55)

0%11%

0%

20%

40%

60%

80%

L159

F

DAA-naive (n=313)SOF/PR; SOF/RBV (n=41)

EC50 fold change: 1.6 (L159F together with C316N) EC50 fold change: 1.3

RAVs in DAA-naive pts (53%) RAVs in SOF/PR and SOF/RBV pts (61%)

RAVs in DAA-naive pts (0%)RAVs in SOF/PR and SOF/RBV pts (11%)

G1b G3

NS5B NS5B

*Double variants were calculated as single RAVs. Thus, sum of single RAVs frequencies is not identical with rate of patients with RAVs.


RAVs in SMV/SOF failures NS3

NS3

<1% 1% <1% <1% 1% 1%3% 3%

44%

0%

20%

40%

60%

80%

F43V

Q80

K/R

S122

R

R155

K

A156

S/TV

D168

E/N

/V

Freq

uenc

yof

RAVs

*

DAA-naive (n=706) SMV/SOF (n=33)

34%

1% <1% <1%

59%

41%

24%

0%

20%

40%

60%

80%Q

80K

R155

G/K/

T

A156

T/V

D168

E/VFr

eque

ncy

of R

AVs*

DAA-naive (n=766)SMV/SOF (n=18)

RAVs in DAA-naive pts (36%) RAVs in SMV/SOF pts (82%)

RAVs in DAA-naive pts (3%)RAVs in SMV/SOF pts (44%)

EC50 fold change: 10 20–100 20–100

EC50 fold change: 10->100

4%

25%

50%

0%20%40%60%80%

Q80

R

D168

E/VFr

eque

ncy

of R

AVs DAA-naive (n=99) SMV/SOF (n=4)

RAVs in DAA-naive pts (4%)RAVs in SMV/SOF pts (75%)

EC50 fold change: no data in G4

G1a

G1b

G4NS3


Frequency and characteristics of RAVs in treatment-naive and DAA-experienced patients - European RAVs database:


RAVs in DCV/SOF versus LDV/SOF failures (G1)


NS5A – DCV/SOF NS5A – LDV/SOF

1% 6% 4% 3% <1% <1% 1%8% 8%

67%

25%17%

8%

0%

20%

40%

60%

80%

100%

K24R

M28

A/T/

V

Q30

E/H/

L/R

L31M

H58D

A92T

Y93

C/F/

N

Freq

uenc

y of

RAV

s* DAA-naive (n=766)

DCV/SOF (n=14)

1% 6% 4% 3% <1% <1% 1%2% 7%

59%

9% 2% 2%

34%

0%20%40%60%80%

100%

K24R

M28

T/V

Q30

E/H/

L/R

L31M P32L

H58D

A92T

Y93

C/F/

N

DAA-naive (n=766)LDV/SOF (n=47)

RAVs in DAA-naive pts (13%)RAVs in DCV/SOF pts (83%)

RAVs in DAA-naive pts (13%)RAVs in LDV/SOF pts (82%)

EC50 fold change: >1000 >100 >100 >100 >100

G1a

G1b

<1% <1% 9% 4% 7%

73% 82%

0%20%40%60%80%

100%

L28V

R30H

L31

F/I/M

/V

A92T

Y93HFr

eque

ncy

of R

AVs*

DAA-naive (n=706) DCV/SOF (n=14)

<1% <1% 9% 4% 7%3%

40%

3%

83%

0%20%40%60%80%

100%

L28V

R30H

L31

F/M

/V

A92T

Y93H

DAA-naive (n=706) LDV/SOF (n=41)



EC50 fold change: 2–10 20–100 2–100 >100



RAVs in DCV/SOF versus LDV/SOF failures (G3 & 4)NS5A – DCV/SOF NS5A – LDV/SOFG3

G4

7% 2%11%

84%

0%

20%

40%

60%

80%

100%

A30K

Y93H

Freq

uenc

y of

RAV

s* DAA-naive (n=313)DCV/SOF (n=21)

7% 2%7%0%

20%40%60%80%

100%

A30K

Y93H

DAA-naive (n=313)LDV/SOF (n=15)



EC50 fold change: >1000 no data

2%

84%100%

0%20%40%60%80%

100%

L28M

L30R

/S

Freq

uenc

y of

RAV

s*

DAA-naive (n=99)DCV/SOF (n=2)

2%

64% 57%

29%43%

0%20%40%60%80%

100%

L28M

Q30

R

M31

V

Y93C

/H

DAA-naive (n=99) LDV/SOF (n=9)



2–10 >100 EC50 fold change:



RAVs in 3D failures






• Methodology of RAV detection (sensitivity threshold)(e.g. deep sequencing with a cut-off at 1% or populationssequencing with a cut-off at 15-25%)

• Impact of RAVs on SVR rates(e.g. combination partner, treatment duration, stage of disease)

Simeprevir + Sofosbuvir in GT1a-infected patients (TN+TE): Impact of the NS3 RAV Q80K

97 96

0

20

40

60

80

100

SMV + SOF

SVR

(%)

SVR w/o Q80K SVR with Q80K

Kwo et al., OPTIMIST-1, EASL 2015; Lawitz et al. , OPTIMIST-2, Hepatology 2015

Tx duration 12 weeks

Cirrhosis No

Q80K frequency 40%


84

97

73

96

0

20

40

60

80

100

SMV + SOF SMV + SOF

SVR

(%)


Tx duration 8 weeks 12 weeks

Cirrhosis No No

Q80K frequency 42% 40%Kwo et al., OPTIMIST-1, EASL 2015; Lawitz et al. , OPTIMIST-2, Hepatology 2015


84

9792

73

96

74

0

20

40

60

80

100

SMV + SOF SMV + SOF SMV + SOF

SVR

(%)


Tx duration 8 weeks 12 weeks 12 weeks

Cirrhosis No No Yes

Q80K frequency 42% 40% 47%Kwo et al., OPTIMIST-1, EASL 2015; Lawitz et al. , OPTIMIST-2, Hepatology 2015

SVR12 Rates by Treatment Regimen (LDV/SOF)and Duration: Patients without

Cirrhosis

98 9990

99 99 99

0

20

40

60

80

100

LDV/SOF 8 Weeks LDV/SOF 12 Weeks LDV/SOF 12 Weeks

With NS5A RAVs No NS5A RAVs

30/32 107/108 187/ 189 504/509 79/88 298/300

SVR1

2 (%

)

TN < 6M TN TE Studies included for analysis: LDV/SOF 8 weeks: GS-US-337-0118 (LONESTAR 1), GS-US-337-0108 (ION-3); LDV/SOF 12 Wks TN: GS-US-GS-US-334-1274 (Bleeding Disorder), GS-US-337-0102 (ION-1), GS-US-337-0108 (ION-3), GS-US-337-0113 (Japan 1), GS-US-337-0115 (ION-4), GS-US-337-0122 (Electron 2), GS-US-337-0131(China), GS-US-337-0118 (LONESTAR 1), GS-US-337-1406, GS-US-337-1468 (LEPTON); LDV/SOF 12 Wks TE: GS-US-337-0109 (ION-2), GS-US-337-0113 (Japan 1), GS-US-337-0115 (ION-4), GS-US-337-0124 (SOLAR-2), GS-US-334-1274 (Bleeding Disorder), GS-US-337-0118 (LONESTAR 1) , GS-US-337-0131(China), GS-US-337-1406, GS-US-337-1468 (LEPTON)

Zeuzem et al., AASLD 2015Sensitivity threshold at 1% (deep sequencing)

SVR12 Rates by Treatment Regimen (LDV/SOF) and Duration: TN Patients with Cirrhosis

96 10088

96 100 100

0

20

40

60

80

100

LDV/SOF LDV/SOF+RBV LDV/SOF


26/27 65/68

12 weeks 24 weeks

SVR1

2 (%

)

10/10 27/27 8/9 19/19

Studies included for analysis: LDV/SOF 12 Wks: GS-US-334-1274 (Bleeding Disorder), GS-US-337-0102 (ION-1), GS-US-337-0113 (Japan 1), GS-US-337-0115 (ION-4), GS-US-337-0122 (ELECTRON-2), GS-US-337-0131(China), GS-US-337-1406;LDV/SOF+RBV 12 Wks: GS-US-337-0102 (ION-1), GS-US-337-0113 (Japan 1), GS-US-337-0122 (ELECTRON-2); LDV/SOF 24 Wks: GS-US-337-0102 (ION-1), GS-US-334-1274 (Bleeding Disorder)


SVR12 Rates by Treatment Regimen and Duration: TE Patients with Cirrhosis

89 8796 100

0

20

40

60

80

100

LDV/SOF+RBV LDV/SOF


59/66 206/214

12 weeks 24 weeks

SVR1

2 (%

)

13/15 84/84

Studies included for analysis: LDV/SOF+RBV 12 Wks: GS-US-337-0109 (ION-2), GS-US-337-0113 (Japan 1), GS-US-337-0118 (LONESTAR-1), GS-US-337-0122 (ELECTRON-2), GS-US-337-0123 (SOLAR-1), GS-US-337-0124 (SOLAR-2), GS-US-337-1118 (Retreatment), P7977-0523 (ELECTRON); LDV/SOF 24 Wks: GS-US-337-0109 (ION-2), GS-US-337-0121 (SIRIUS), GS-US-334-1274 (Bleeding Disorder)






• Methodology of RAV detection (sensitivity threshold)(e.g. deep sequencing with a cut-off at 1% or populationssequencing with a cut-off at 15-25%)

• Impact of RAVs on SVR rates(e.g. combination partner, treatment duration, stage of disease)

Grazoprevir + Elbasvir for 12 wks in TN/relapse GT-1a patients: Impact of NS5A RAVs on SVR rates

98% 98% 98% 98%58%

86% 72% 91%

0%

20%

40%

60%

80%

100%

EBR RAVs NS5A class RAVs EBR RAVs NS5A class RAVs

Population sequencing Next generation sequencing

SVR1

2

Patients without RAVs Patients with RAVs

No RAVS: 414/438‡

(95%)

No RAVS: 352/438‡

(80%)

No RAVS: 396/439‡

(90%)

No RAVS: 289/439‡

(65%)

EBR RAVs EBR RAVsNS5A Class RAVs NS5A Class RAVs

1424

405414

7486

345352

3143

389396

136150

284289

5% 20% 10% 35%

PREV

ALEN

CE

†NGS with 1% ST supplemented by Population Sequencing when NGS not available. ‡ One GT1a was missing baseline population sequencing data but had baseline NGS dataEBR RAV List = For GT1a: M/L28T/A/G, Q/R30E/H/R/G/K/L/D, L31M/V/F, H58D, or Y93C/H/N/S NS5A Class RAV List = Any variant from reference strain at NS5A position 24, 28, 30, 31, 32, 38, 58, 92 or 93

Population Sequencing Deep Sequencing at 1% Sens. Threshold†

Jacobson et al., AASLD 2015

Effect of RAVs at Specific Baseline Positions on Likelihood to Achieve SVR12

RAV Position

SVR12Subjects with RAVs

(1% ST NGS)

SVR12Subjects With RAVs (Population

Sequencing)

24 15/18 (83.3%) 4/4 (100.0%)

28 61/68 (89.7%) 29/33 (87.9%)

30 14/23 (60.9%) 4/10 (40.0%)

31 15/23 (65.2%) 5/13 (38.5%)

32 1/1 (100.0%) --

38 9/9 (100.0%) --

58 75/77 (97.4%) 48/49 (98.0%)

92 6/6 (100.0%) 3/3 (100.0%)

93 9/14 (64.3%) 5/8 (62.5%)

GT1a-Infected TN/TE Subjects given EBR/GZR 12 weeks (no RBV)

NGS using 1% ST supplemented by Population Sequencing when NGS not availableNS5A Class RAV List = Any variant from reference strain at NS5A position 24, 28, 30, 31, 32, 38, 58, 92 and 93

Grazoprevir + Elbasvir for 12 wks in TN/relapseGT-1b patients: Impact of NS5A RAVs on SVR rates

100% 100% >99% 99%98% 99% 98% 99%

0%

20%

40%

60%

80%

100%



SVR1

2



No RAVS: 220/265

(83%)

No RAVS: 184/265

(69%)

No RAVS: 211/265

(80%)

No RAVS: 172/265

(65%)

EBR RAVS EBR RAVSNS5A Class RAVS NS5A Class RAVS

4445

219220

8081

183184

5354

210211

9293

171172

17% 31% 20%35%

PREV

ALEN

CE

†NGS 1% ST supplemented by Population Sequencing when NGS not availableEBR RAV List = For GT1b: P32L, P58D or A92K or any variant at NS5A position 28, 30, 31 or 93NS5A Class RAV List = Any variant from reference strain at NS5A position 24, 28, 30, 31, 32, 38, 58, 92 or 93 Jacobson et al., AASLD 2015

Grazoprevir + Elbasvir for 12 wks in non-responderGT-1a patients: Impact of NS5A RAVs on SVR rates

97% 96% 97% 96%

29%64% 44%

76%

0%

20%

40%

60%

80%

100%



SVR1

2



No RAVS: 61/68(90%)

No RAVS: 54/68(79%)

No RAVS: 59/68(87%)

No RAVS: 47/68(69%)


27

5961

914

5254

49

5759

1621

4547

10% 21% 13%31%

PREV

ALEN

CE

†NGS 1% ST supplemented by Population Sequencing when NGS not available. ‡ One GT1a was missing baseline population sequencing data but had baseline NGS dataEBR RAV List = For GT1a: M/L28T/A/G, Q/R30E/H/R/G/K/L/D, L31M/V/F, H58D, or Y93C/H/N/S NS5A Class RAV List = Any variant from reference strain at NS5A position 24, 28, 30, 31, 32, 38, 58, 92 or 93 Jacobson et al., AASLD 2015

100% 100% 100% 100%100% 100% 100% 100%

0%

20%

40%

60%

80%

100%



SVR1

2



No RAVS: 51/52(98%)

No RAVS: 44/52(85%)

No RAVS: 48/52(92%)

No RAVS: 38/52(73%)


11

5151

88

4444

44

4848

1414

3838

2%15% 8% 27%

PREV

ALEN

CE

†NGS 1% ST supplemented by Population Sequencing when NGS not available. ‡ One GT1a was missing baseline population sequencing data but had baseline NGS dataEBR RAV List = For GT1a: M/L28T/A/G, Q/R30E/H/R/G/K/L/D, L31M/V/F, H58D, or Y93C/H/N/S NS5A Class RAV List = Any variant from reference strain at NS5A position 24, 28, 30, 31, 32, 38, 58, 92 or 93

Grazoprevir + Elbasvir + RBV for 16/18 wks in NR GT-1a patients: Impact of NS5A RAVs on SVR rates

Jacobson et al., AASLD 2015

How to define a RAV best clinically?

• RAVs (single and combination) per specific drug• Genotypic backbone in vitro and in vivo• Level of reduced susceptibility in vitro >100-fold• sensitivity threshold at 15-25% (population sequencing)

• Impact of RAVs on SVR rates in defined patientpopulation and a defined regimen (DAA combination, treatment duration)

• and there is even more complexity ........

Y93H and IL28B CC prevalence by region

Prev

alen

ce o

f IL2

8B C

C (%

)

22 22

30

54

0

20

40

60

NorthAmerica

Europe Oceania Asia Pacific

7133262

203929

109360

324597

4 8 6

17

0

20

40

60

NorthAmerica

Europe Oceania Asia Pacific

Prev

alen

ce o

f Y93

H (%

)

1443262

73929

21360

104597

GT1a/b

Zeuzem et al., AASLD 2015

26/2538

Association of Y93H and IL28B genotype


Y93H

Pre

vale

nce

(%)

13

5

25

51

12

0

5

10

15

20

25

30

GT 1 GT 1a GT 1b

CC Non CC

179/3800 36/774 142/574178/1348 153/1262

p <0.0001 p <0.0001p <0.0001

♦ 5,148 GT1 patients in Gilead HCV clinical trials with both NS5A sequencing and IL28B genotype data

SVR12 rates by Y93H or any NS5A RAV and IL28B genotype

9789

100 9798 93100 98

0

20

40

60

80

100

IL28B CC IL28B non-CC

With Y93H No Y93H With RAVs No RAVs

SVR1

2 (%

)

Studies included for analysis: GS-US-337-0118 (LONESTAR 1), GS-US-GS-US-334-1274 (Bleeding Disorder), GS-US-337-0102 (ION-1), GS-US-337-0108 (ION-3), GS-US-337-0113 (Japan 1), GS-US-337-0115 (ION-4), GS-US-337-0122 (Electron 2), GS-US-337-0131(China), GS-US-337-1406, GS-US-337-1468 (LEPTON), GS-US-337-0109 (ION-2), GS-US-337-0124 (SOLAR-2), GS-US-334-1274 (Bleeding Disorder), GS-US-337-1406

7476

429431

5461

11641196

153156

259280

350351

959977

♦ SVR12 data pooled from all guideline recommended LDV/SOF regimens


SVR12 Rates by Treatment Regimen and Y93H/IL28B: Patients without Cirrhosis

30

100 10094

100 100 100100 100

79

97 98 97

0

20

40

60

80

100

LDV/SOF 8 Weeks LDV/SOF 12 Weeks LDV/SOF 12 Weeks

SVR1

2 (%

)

TN < 6M TN TE

33

3131

11

102105

5151

468479

2323

239239

1516

1114

8585

266273

With Y93H No Y93HIL28B CC

IL28B non-CC

Studies included for analysis: GS-US-337-0118 (LONESTAR 1), GS-US-GS-US-334-1274 (Bleeding Disorder), GS-US-337-0102 (ION-1), GS-US-337-0108 (ION-3), GS-US-337-0113 (Japan 1), GS-US-337-0115 (ION-4), GS-US-337-0122 (Electron 2), GS-US-337-0131(China), GS-US-337-1406, GS-US-337-1468 (LEPTON), GS-US-337-0109 (ION-2), GS-US-337-0124 (SOLAR-2), GS-US-334-1274 (Bleeding Disorder), GS-US-337-1406

It is useful to detect RAVs ?• There is no “YES” or “No” answer• First and most important: Methodology for detection of RAVs must standardized

(and automated)• Pharmaceutical industry must fully publish available RAV data in collaboration

with academia• Usefulness of RAV testing will be patient population and treatment regimen

dependent• RAV testing most likely not required in patient populations with SVR rates > 99%• RAV testing most likely be clinical useful and cost-effective in population with

suboptimal SVR rates (definition < 95% / < 90% ?) & if population large enough– Regimens w/o a very high barrier to resistance drug– Treatment-experienced patients (in particular when exposed to DAAs)– Patients with cirrhosis– When the shortest possible treatment duration is economically important

• IL28B CC genotype is significantly associated with a higher prevalence of Y93H. This association is functionally not yet understood

• The relevance of (in particular NS5A) RAVs, baseline HCV RNA levels, and IL28B genotype on treatment outcome requires further intensive clinical research

Treatment decisions in patients with RAVs

• Switch drug class(exception: nucleosidic polymerase inhibitors)

• Treat longer (up to 24 weeks)• Add ribavirin• Use triple therapies (NI + PI + NS5A-Inhibitor)• Second generation PIs and NS5A-Inhibitors

(e.g. Glecaprevir, Pibrentasvir)• Peginterferon (??)

Documents

What do we need to know about RAVs clinicallyregist2.virology-education.com/2016/14EU/27_Zeuzem.pdf · 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30. LDV RAVs Class RAVs Y93H. GT1a. Sensitivity