Effect of IL28B Genotype on Early Viral Kinetics During Interferon-Free
Treatment of Patients With Chronic Hepatitis C
TOM W. CHU,* ROHIT KULKARNI,‡EDWARD J. GANE,§STUART K. ROBERTS,?CATHERINE STEDMAN,¶
PETER W. ANGUS,#BRETT RITCHIE,** XIAO-YU LU,‡DAVID IPE,‡URI LOPATIN,* SOREN GERMER,‡‡
VICTOR A. IGLESIAS,§§ROBERT ELSTON,??PATRICK F. SMITH,* and NANCY S. SHULMAN‡
*Roche, Nutley, New Jersey;‡Genentech, South San Francisco, California;§Auckland Clinical Studies, Auckland, New Zealand;?The Alfred Hospital, Melbourne,
Victoria, Australia;¶Christchurch Clinical Studies, Christchurch, New Zealand;#The Austin Hospital, Heidelberg, Victoria, Australia; **The Royal Adelaide, Adelaide,
South Australia, Australia;‡‡Roche, TRS, Nutley, New Jersey;§§Roche, Basel, Switzerland; and? ?Roche, Welwyn, United Kingdom
BACKGROUND & AIMS: Although interleukin 28B (in-
terferon, lambda 3) (IL28B) genotype affects the response
of patients with chronic hepatitis C to peginterferon and
ribavirin, little is known regarding its effect on response to
direct-acting antivirals in interferon-free combinations. We
analyzed the effects of IL28B genotype on the viral kinetic
(VK) response to an interferon-free combination of the
nucleoside polymerase inhibitor mericitabine (RG7128)
and the hepatitis C virus (HCV) protease inhibitor dano-
previr. METHODS: We performed a double-blind, dose-
escalation study of patients with chronic HCV genotype 1
infection who were interferon treatment naive or had not
responded to previous therapy with peginterferon and
ribavirin. Patients were sequentially assigned to 1 of 7
cohorts then randomly assigned to groups that received
up to 13 days of treatment with mericitabine (500 or 1000
mg, twice daily) plus danoprevir (100 or 200 mg, every 8
hours, or 600 or 900 mg, twice daily) or placebo. Eighty-
three of 87 patients were genotyped for the IL28B single-
nucleotide polymorphism rs12979860. VKs were analyzed
only in patients who received 13 days of treatment, at
optimal doses, using a biphasic model to describe first-
and second-phase slopes of viral decay during therapy.
RESULTS: At day 14 (the end of interferon-free treat-
ment), the mean reduction in the serum level of HCV
RNA was slightly greater in patients with the CC poly-
morphism (5.01 log10IU/mL) than those without (4.59
log10IU/mL). Modeling revealed that patients with the CC
polymorphism had slightly better early VKs, most appar-
ent in the ?-phase of viral decay. A mixed effect on the
?-phase was observed, which was reduced in magnitude
but prolonged in patients with CC, who also had better
on-treatment response to peginterferon and ribavirin dur-
ing follow up. CONCLUSIONS: IL28B genotype ap-
pears to affect early VKs in patients with chronic
hepatitis C receiving interferon-free treatment.
Keywords: INFORM-1; Genetics; Direct-Acting Antiviral
of sustained virologic response (SVR) after treatment with
peginterferon plus ribavirin in patients with chronic hep-
ost interleukin 28B (interferon, lambda 3) (IL28B)
genotype is the most important baseline predictor
atitis C virus genotype 1 infection.1,2In a seminal genome-
wide association study, patients with a CC genotype at the
rs12979860 locus on chromosome 19 had a 2- to 3-fold
greater rate of SVR when treated with peginterferon plus
ribavirin than patients with the less favorable TT geno-
type.1Several classes of direct-acting antiviral agents
(DAAs) are under development, and it is expected that,
when used in combination with peginterferon plus riba-
virin, these new drugs will increase SVR rates and decrease
the required duration of therapy for many patients with
chronic hepatitis C.3–8The inclusion of interferon in these
regimens makes it likely that host IL28B genotype will
significantly influence disease management decisions as
well as treatment outcomes, as a preliminary study has
Unfortunately, treatment with interferon-containing
regimens is associated with significant adverse effects in a
large proportion of patients. Consequently, interferon-
free combinations of DAAs have become an area of con-
siderable clinical interest. To this end, we recently showed
in the INterferon-Free regimen fOR the Management of
HCV (INFORM-1) study that a dual oral combination
regimen composed of the nucleoside polymerase inhibitor
mericitabine and the HCV protease inhibitor danoprevir
produced a rapid and substantial viral load decline that
was maintained throughout 2 weeks of treatment.10A
host IL28B polymorphism has also been shown to affect
the spontaneous clearance of HCV infection11; thus, it is
conceivable that genetic variation at this locus may also
influence the outcome of treatment with interferon-free
regimens. For this reason, we conducted an exploratory
analysis to determine the effect of host IL28B genotype on
the early viral kinetic response to mericitabine and dano-
previr in patients enrolled in the INFORM-1 study.
Patients and Methods
INFORM-1 was a randomized, double-blind, placebo-
controlled, dose-escalation trial in patients with chronic hepati-
Abbreviations used in this paper: DAA, direct-acting antiviral agents;
IL28B, interleukin 28B [interferon, lambda 3]; INFORM, INterferon-Free
regimen fOR the Management of HCV; SVR, sustained virologic response.
© 2012 by the AGA Institute
tis C including both interferon treatment-naive individuals and
previous nonresponders to peginterferon plus ribavirin.10The
study design is described in Figure 1, and the primary results are
published elsewhere.10Briefly, patients from 6 centers in Aus-
tralia and New Zealand were enrolled into 1 of 7 cohorts.
Patients in cohort A1/A2 received monotherapy with 1 of the 2
active drugs for 4 days followed by the combination of both
active drugs for a further 3 days. Patients enrolled in the 6
remaining cohorts (B–G) were randomized to 13 days of active
treatment with mericitabine (500 or 1000 mg twice daily) plus
danoprevir (100 or 200 mg every 8 hours, or 600 or 900 mg twice
daily) or placebo. After this time, all patients were allowed to
initiate treatment with peginterferon plus ribavirin. Patients
who received 13 days of combination therapy in cohorts C–G are
the subject of this analysis. Patients randomized to cohort B
received lower doses of both mericitabine and danoprevir and
had lower viral declines over the 13 days than the subsequent
cohorts and are excluded from the analyses. One subject in
cohort C with a baseline D168E mutation conferring a 27-fold
decrease in susceptibility to danoprevir is also excluded.
Serum HCV RNA levels were determined by COBAS TaqMan
HCV Test v2.0 assay (Roche Diagnostics, Burgess Hill, UK: lower
limit of quantification, 43 IU/mL; lower limit of detection, 15
IU/mL) at baseline; at days 1, 2, 3, 4, 5, 6, 7, 10, and 13 of
treatment with mericitabine plus danoprevir; study day 14
(washout); and at weeks 4 (study day 42) and 12 (study day 98)
of treatment with peginterferon plus ribavirin.
All patients enrolled in INFORM-1 were considered for inclu-
sion in this genetic analysis. Participation was voluntary, and the
patients’ data were anonymized. All participating patients signed
an appropriate informed consent form before donating blood
samples for DNA analysis. A total of 83 out of 87 patients in the
intention-to-treat population consented to donate DNA into the
repository for post hoc analyses. These samples were subse-
quently genotyped at the rs12979860 locus by direct sequencing
of regions in and upstream of the IL28B gene.
Blood samples (6 mL) were collected in EDTA-containing
tubes. DNA was extracted from whole blood using a silica
adsorption-based extraction method (MagNA Pure LC DNA
Isolation Kit I; Roche Applied Science, Indianapolis, IN). Geno-
typing for rs12979860 was performed by double-stranded DNA
sequencing using an ABI capillary sequencing apparatus and
“BigDye” chemistry (Applied Biosystems, Carlsbad, CA). Cycle
sequencing was carried out on an MJ Tetrad PCR machine
(Waltham, MA) using ABI BigDye Terminator (Applied Bio-
systems) chemistry according to the manufacturer’s instruc-
tions. After sequencing, the polymorphism analyses were per-
formed using PolyPhred software (licensed from University of
Viral load profiles were constructed for each patient over 13
days of combination therapy with mericitabine plus danoprevir.
Viral load differences at a given time point were compared using
a 2-sided Mann–Whitney test with type I error fixed at 10%.
Fisher exact test was used to compare the proportions that were
HCV RNA negative at a given time point.
To assess the effect of treatment and IL28B genotype on the
viral kinetics, viral decay profiles were fit to a nonlinear mixed
The population characteristics were modeled as fixed effects,
individual variations were modeled as random effects, and with-
in-individual variations were taken into account by an error
process. Yi(t) denotes the log10viral load of the ithsubject at time
t, V0idenotes the baseline (pretreatment) viral load on a log10-
scale of the ithsubject and Ai, ?1i, ?2irepresent viral kinetics
Figure 1. Study design of the INFORM-1 trial. Patients with chronic
HCV infection, including treatment naive, interferon treatment failure
(partial responder and relapse), and interferon null responders, were
randomized to either active study drugs or matching placebo by cohort.
previr at increasing doses. The patients were allowed to initiate pegy-
lated interferon/ribavirin treatment after completion of active study drug
Figure 2. Mean change in serum HCV RNA level between study days 1
and 14 by rs12979860 genotype in all patients in cohorts C–G (n ? 45)
who received 13 days of dual oral combination therapy with mericitabine
IL28B VIRAL KINETICS IN IFN-FREE TREATMENT 791
model parameters of the ithsubject. ? Represents the random
effect terms, which capture the interindividual variability (ie, an
individual’s deviation from the typical behavior of the treatment
group) and ?(.)and ?(.)capture the effect of treatment and IL28B
genotype on the viral kinetic parameters. The decision to include
or exclude these effects in the model was made by comparing
nested models using a likelihood ratio test. The Akaike infor-
mation criterion was also considered to strike a balance between
goodness-of-fit and model complexity.
?1And ?2capture the slope of the first and the second phase,
respectively. To make the model identifiable, we imposed the
restriction that ?1? ?2. The viral decay model as described
above does not fit viral rebound data. Thus, if a viral load curve
deviated from a viral decay trajectory, the data were excluded
from the model fitting.15To deal with data below the limit of
quantification, all HCV RNA measurements ?40 IU/mL were
imputed as 40 IU/mL (LLoQ of the assay). If more than 1 below
the limit of quantification value in a viral load trajectory was
imputed using this rule, a level-off effect was created, which the
viral decay model is unable to capture. It is however somewhat
crucial to capture this characteristic because in clinical terms it
is interpreted as a virologic response. To account for this effect,
the fitted model is modified as follows:
In INFORM-I, a total of 741 viral load data points from 45
eligible patients across 19 time points on 13 days of dosing were
available. HCV RNA measurements made at 12 and 16 hours
postdose on day 3 and day 4 were ignored because they create an
artificial saw-tooth pattern, leaving 607 data points on 45 sub-
jects across 15 time points for analysis. Seven data points qual-
ified as rebound (a data point qualified as a rebound if it
occurred after the nadir of the viral decay trajectory and was ?
0.5 log10higher than the HCV RNA measurement at the nadir)
and were excluded. Thus, a total of 600 data points on 45
subjects across 15 time points on 13 days of dosing were con-
sidered for model fitting and parameter estimation. Seventy-two
of these data points were below the limit of quantification and
imputed as 40 IU/mL.
A forward-stepwise approach was used to arrive at the final
model. First a model with no covariates was fitted, and then
covariates were added 1 at a time as fixed effects. The signifi-
cance of a fixed effect associated with a covariate for inclusion in
the model was assessed using a likelihood ratio test. The final
fitted model includes treatment and IL28B genotype fixed effect
on all viral kinetic parameters in the model.
All statistical analyses were done in R (version 2.10.0), and for
model fitting the nlme library was used.16Approximate length of
the phase-I duration (tPh-1) was computed as in Wolters et al.17
Clinical trial information is located at www.clinicaltrials.gov
Host IL28B genotype was determined by direct
sequencing for 83 out of 87 patients who received at least
1 dose of study medication. One-third of treatment-naive
patients had a CC genotype (21 of 63). The remaining
two-thirds of treatment naive patients carried the T allele
(34 [54%] had the CT genotype, and 8 [13%] had the TT
genotype). Only 2 (20%) of the 10 previous partial re-
sponders and none (0%) of the 10 previous null respond-
ers to peginterferon plus ribavirin had a CC genotype.
Among previous partial responders, 8 (80%) had the CT
Figure 3. Mean change in serum HCV RNA level between study days 1
and 14 by rs12979860 genotype in all patients in the 2 highest dose
cohorts (F and G; n ? 15) in which patients received 13 days of dual oral
combination therapy with mericitabine 1000 mg twice daily plus dano-
previr 900 mg twice daily.
Table 1. Point-Wise Comparison of Viral Load Curves Using Mann–Whitney Test
Mean change in HCV
RNA in CC (n ? 12)
Mean change in HCV RNA
in non-CC (n ? 33)
Mann–Whitney test for
equality of means (P value)
NOTE. The mean reduction in serum HCV RNA level is significantly different in patients with CC genotype when compared with non-CC genotypes
at days 7, 10, 13, and 14.
aP ? .05.
bP ? .1.
792 CHU ET ALGASTROENTEROLOGY Vol. 142, No. 4
genotype, and none had the TT genotype. For the previ-
ous null responders, 6 (60%) had the CT genotype, and 4
(40%) had the TT genotype. Mean baseline serum HCV
RNA levels were similar in patients with the CC genotype
(649 ? 0.48 log10IU/mL) and in those with non-CC
genotypes (6.31 ? 0.61 log10IU/mL).
Robust viral load reductions were seen in patients with
CC, CT, and TT genotypes in the 5 cohorts (C, D, E, F, and
G; n ? 45) that received 13 days of active treatment with
both drugs (Figure 2) and in the subset of patients in the
2 highest dose cohorts (F and G, n ? 15) (Figure 3).
The mean reduction in serum HCV RNA level began
to differentiate at day 7 in patients with CC versus the
non-CC genotypes and continued through the end of
treatment (Table 1). A similar differential HCV decline
was observed whether cohorts were considered individ-
ually or collectively: a 0.37–0.66 log10(IU/mL) greater
mean reduction was observed for patients with the CC
genotype within each dose group, except for cohort E
(danoprevir 600 mg twice daily, mericitabine 1000 mg
twice daily), which included only a single CC genotype
patient. Therefore, dosing did not appear to be associ-
ated with the differential viral load decline observed
between patients with CC versus non-CC genotypes at
Modelling of viral kinetic profiles revealed that both ?
and ? slopes of viral decay differed by IL28B genotype. The
slope of the ? phase was greater in patients who carried
the CC genotype (P ? .0187, Figure 4, Table 2). Whereas
the slope of the ? phase was lower in patients with the CC
genotype (P ? .0086), this was partially offset by a longer
duration of the ? phase (P ? .0109).
A higher proportion of patients with the CC genotype
had undetectable HCV RNA levels at the end of interfer-
on-free treatment overall (50% vs 27%, respectively, P ?
.174) and in the highest dose cohorts (100% vs 33%,
respectively, P ? .077) (Table 3). Viral clearance rates
between patients were higher overall in CC versus non-CC
genotypes on treatment with peginterferon plus ribavirin
during the follow-up phase of the trial (week 4: 83% vs
27%, respectively, P ? .001; and week 12: 92% vs 64%,
respectively, P ? .134).
This small post hoc analysis suggests that IL28B
genotype has an influence on early viral kinetics during
treatment with an interferon-free DAA regimens. To
more precisely describe the effect of host genotype on
early viral kinetics, the data were fit according to a
biphasic model of viral decay, which takes into account
the effects of treatment and IL28B genotype. We se-
lected this approach because it requires data only from
the early segments of the viral load trajectory and is
suitable for use with small data sets. Moreover, this
method avoids potential bias in selecting time points
that define the duration of the ? and ? phases. Inter-
estingly, our analyses suggest that patients with the CC
genotype have decreased ? and increased ? phase slopes
Figure 4. Early viral kinetics during combination treatment (ie, day 1 to
day 14) by IL28B genotype for all subjects in active arm from cohorts C
to G who received 13 days of dual oral combination therapy with meri-
citabine plus danoprevir.
Table 2. Summary of Early Viral Kinetics During Combination Treatment, ie, Day 1 to Day 14, by IL28B Genotype for All
Subjects in Active Arm From Cohorts C to G
? Phase slope (-HCV RNA
log10/day) Length of ? phase (h)
? Phase slope (-HCV RNA
CC Non-CCCC Non-CCCC Non-CC
Mann–Whitney test (P value)
CV, coefficient of variation.
IL28B VIRAL KINETICS IN IFN-FREE TREATMENT793
compared with patients with other IL28B genotypes.
However, the net effect on ? phase decay is partially
offset by the prolonged duration of the ? phase in CC
genotype patients. Therefore, differences in early viral
kinetics because of host IL28B genotype in patients
receiving interferon-free regimens may be more readily
observed during the ? phase of viral decay, as is sug-
gested from visual inspection of the mean viral decay
curves (Figures 2 and 3). Our data also suggest that the
relative reduction in viral load between CC and non-CC
patients is not associated with dosing, although the
numbers of patients in these 2 subgroups are too small
to make a definitive statement. Future studies of inter-
feron-free regimens should determine whether there is
an interaction between the doses of the DAAs and
Recent reports describing the effect of IL28B polymor-
phism on early viral kinetics in HCV genotype 1 patients
treated with pegylated interferon and ribavirin have em-
phasized that the rs12979860 CC genotype is associated
with an increased ? phase decline compared with non-CC
genotypes.18,19In contrast, our observations suggest that
an increase in the ? phase decline is the most prominent
feature of early viral decay during treatment with an
interferon-free DAA regimen in patients with the
rs12979860 CC genotype. A potential explanation for this
difference is that inhibition of viral production is more
sensitive to the effects of IL28B genotype during treat-
ment with pegylated interferon and ribavirin than during
treatment with an interferon-free DAA combination.
Mechanistically, the ? effect may reflect host differences
in the contribution of the innate immune response (en-
dogenous interferon responses) to the loss of infected
Based on these data, determination of IL28B polymor-
phism may be important for analysis of treatment out-
comes in clinical trials of interferon-free regimens. These
results, however, need to be confirmed in larger studies
and with different combinations of direct-acting antivirals
and in longer studies to determine the ultimate impact of
these kinetic differences on SVR.
As expected and seen in other studies, patients with the
CC genotype had more robust antiviral responses in com-
parison with the non-CC genotype after stopping meri-
citabine plus danoprevir and continuing subsequent
treatment with peginterferon plus ribavirin.
IL28B genotyping will continue to play an important
role in determining the likelihood of response to treat-
ment with peginterferon plus ribavirin and to emerging
triple (single DAA plus peginterferon/ribavirin) and qua-
druple (2 DAAs plus peginterferon/ribavirin) antiviral reg-
imens. However, the importance of IL28B genotype on
response to future interferon-free combination DAA reg-
imens remains to be determined with analyses of larger
and longer duration studies with interferon-free therapy
1. Ge D, Fellay J, Thompson AJ, et al. Genetic variation in IL28B
predicts hepatitis C treatment-induced viral clearance. Nature
2. Thompson AJ, Muir AJ, Sulkowski MS, et al. Interleukin-28B poly-
morphism improves viral kinetics and is the strongest pretreat-
ment predictor of sustained virologic response in genotype 1
hepatitis C virus. Gastroenterology 2010;139:120–129.
3. Hézode C, Forestier N, Dusheiko G, et al. Telaprevir and peginter-
feron with or without ribavirin for chronic HCV infection. N Engl
J Med 2009;360:1839–1850.
4. McHutchison JG, Everson GT, Gordon SC, et al. Telaprevir with
peginterferon and ribavirin for chronic HCV genotype 1 infection.
N Engl J Med 2009;360:1827–1838.
5. Kwo PY, Lawitz EJ, McCone J, et al. Efficacy of boceprevir, an NS3
protease inhibitor, in combination with peginterferon alfa-2b and
ribavirin in treatment-naive patients with genotype 1 hepatitis C
infection (SPRINT-1): an open-label, randomised, multicentre
phase 2 trial. Lancet 2010;376:705–716.
6. McHutchison JG, Manns MP, Muir AJ, et al. Telaprevir for previ-
ously treated chronic HCV infection. N Engl J Med 2010;362:
Table 3. Mean Change in Serum HCV RNA Levels During 13 Days of Treatment With Mericitabine Plus Danoprevir and After
4 and 12 Weeks of Treatment With Peginterferon Plus Ribavirin
Mean change in HCV RNA
on day 14 after 13 days
of treatment with
Patients with undetectable HCV RNA (?15 IU/mL)
14, n (%)
Week 4 of
Week 12 of
All patients (cohorts C–G; n ? 45)
Highest dose cohorts (cohorts F
and G; n ? 15)
aDifference in the response rate between CC and non-CC at week 4 is statistically significant (P ? .001).
794CHU ET ALGASTROENTEROLOGY Vol. 142, No. 4
7. Jacobson IM, McHutchison JG, Dusheiko GM, et al. Telaprevir in Download full-text
combination with peginterferon and ribavirin in genotype 1 HCV
treatment-naïve patients: final results of phase 3 ADVANCE study
(abstract 211). Hepatology 2010;52(Suppl 1).
8. Poordad F, McCone J, Bacon BR, et al. Boceprevir (BOC) combined
with peginterferon alfa-2b/ribavirin (P/R) for treatment-naïve pa-
tients with hepatitis C virus (HCV) genotype (G) 1: SPRINT-2 final
results (abstract LB-4). Hepatology 2010;52(Suppl 1).
9. Akuta N, Suzuki F, Hirakawa M, et al. Amino acid substitution in
hepatitis C virus core region and genetic variation near the inter-
leukin 28b gene predict viral response to telaprevir with peginter-
feron and ribavirin. Hepatology 2010;52:421–429.
10. Gane EJ, Roberts SK, Stedman CA, et al. Oral combination therapy
with a nucleoside polymerase inhibitor (RG7128) and danoprevir
for chronic hepatitis C genotype 1 infection (INFORM-1): a ran-
domised, double-blind, placebo-controlled, dose-escalation trial.
11. Thomas DL, Thio CL, Martin MP, et al. Genetic variation in IL28B
and spontaneous clearance of hepatitis C virus. Nature 2009;
12. Neumann AU, Lam NP, Dahari H, et al. Hepatitis C viral dynamics
in vivo and antiviral efficacy of interferon-? therapy. Science 1988;
13. Wu H, Ding AA, DeGruttola V. Estimation of HIV dynamic parame-
ters. Stat Med 1998;17:2463–2485.
14. Wu H, Ding AA. Population HIV-1 dynamics in vivo, applicable
models and inferential tools for virological data from AIDS clinical
trials. Biometrics 1999;55:410–418.
15. Wu H. Statistical methods for HIV dynamic studies in AIDS clinical
trials. Stat Med 2005;14:171–192.
16. Pinheiro JC, Bates DM. Mixed-effects models in S and S-plus. New
York: Springer, 2000.
17. Wolters LMM, Hansen BE, Niesters HGM, et al. Viral dynamics
during and after entecavir therapy in patients with chronic hepa-
titis B. J Hepatol 2002:37;137–144.
18. Lindh M, Lagging M, Arnholm B, et al. IL28B polymorphisms
determine early viral kinetics and treatment outcome in patients
receiving peginterferon/ribavirin for chronic hepatitis C genotype
1. J Viral Hepat 2011;18:e325–331.
19. Bochud PY, Bibert S, Negro F, et al. IL28B polymorphisms
predict reduction of HCV RNA from the first day of therapy in
chronic hepatitis C. J Hepatol 2011;55:980–989.
20. Guedj J, Rong L, Dahari H, et al. A perspective on modelling
hepatitis C virus infection. J Viral Hepat 2010;17:825–833.
Received February 7, 2011. Accepted December 26, 2011.
Address requests for reprints to: Tom Chu, MD, PhD, Virology-
Pharma Research Early Development, 340 Kingsland Street, Nutley,
New Jersey 07110. e-mail: email@example.com.
Conflicts of interest
The authors disclose the following: T.W.C., R.K., X.-Y.L., D.I., S.G.,
V.A.I., R.E., P.F.S., and N.S.S. are employees of F. Hoffmann-La
Roche Ltd; U.L. was an employee of F. Hoffman-La Roche Ltd at the
time the study was conducted; S.K.R. is a Roche Global Advisory
Board member. The remaining authors disclose no conflicts.
Supported by F. Hoffmann-La Roche Ltd.
IL28B VIRAL KINETICS IN IFN-FREE TREATMENT 795