FULL MANUSCRIPT
JOURNAL OF INTERFERON & CYTOKINE RESEARCH
Volume 00, Number 00, 2014
ª Mary Ann Liebert, Inc.
DOI: 10.1089/jir.2014.0088
Homozygosity for HLA Group 2 Alleles Predicts
Treatment Failure with Interferon-a and Ribavirin
in Chronic Hepatitis C Virus Genotype 1 Infection
Meadhbh Collison,1 Jun Liong Chin,2 Ahmed Abu Shanab,2 Ross Mac Nicholas,2
Ricardo Segurado,3 Suzie Coughlan,1 Jeff Connell,1 Michael J. Carr,1
Raphael B. Merriman,2 P. Aiden McCormick,2 and William W. Hall1
Host genetic factors influence treatment responses to antiviral therapy in chronic hepatitis C virus (HCV)
infection. We retrospectively investigated associations between host genetic markers and treatment-induced
virologic responses to dual therapy with interferon-a and ribavirin in chronically infected HCV genotype 1
(g1)– and genotype 3 (g3)–infected individuals. A total of 171 patients (89 HCV g1 and 82 HCV g3 infected)
were investigated for genetic markers influencing treatment-induced sustained virologic response (SVR).
Overall, SVR was observed for 46/89 (52%) HCV g1- and 57/82 (70%) HCV g3-infected patients. Of the 4
interleukin 28B (IL28B) single-nucleotide polymorphisms (SNPs), rs12979860 was the host genetic marker
most significantly associated with failure to achieve an SVR in HCV g1-infected individuals [P = 3.83 · 10 - 4;
odds ratio (OR) = 5.61; confidence interval (CI) = 2.07–15.18] and gave a positive predictive value for treatment
failure of 81.3% for minor homozygotes (TT). Using additive (P = 3.54 · 10 - 4) and dominant models
(P = 3.83 · 10 - 4), a dosage effect of the T allele was observed, with the dominance term not significant for this
SNP. Logistic regression showed an association between HLA-C1/C1 and rapid virologic response in HCV g1
infections with an OR relative to the heterozygote of 10.0 (95% CI: 1.6–62.5, P = 0.014). HLA-C2 homozygosity was a significant predictor of nonresponse to treatment in HCV g1–infected individuals (P = 0.023).
and its transport within the cells is mediated primarily by the
equilibrative nucleoside transporter (ENT), ENT1 (Fukuchi
and others 2010). Morello and others (2010) identified an
SNP, termed rs760370, within the ENT1 gene to be associated with rapid virologic response (RVR) in HIV/HCV-coinfected patients who had received peg-IFN-a with ribavirin.
Cysteine-cysteine chemokine receptor 5 (CCR5) is expressed
on the surface of various cell types, including monocytes,
macrophages, memory T cells, Th1 cells, and immature
dendritic cells (He and others 1997), and studies that investigated the role of CCR5D32 in HCV infection have reported
conflicting results regarding spontaneous clearance of HCV
and susceptibility to infection (Mascheretti and others 2004;
Goulding and others 2005; Nattermann and others 2011).
Significant fibrosis (Ishak stage > 2) has been associated with
a reduced chance of achieving SVR, and has also been shown
to have a negative impact on early viral kinetics during dual
therapy (Guedj and others 2012). An association with the
rs738409 SNP in the PNPLA3 gene and presence of cirrhosis
and accelerated fibrosis progression has been reported,
Introduction
D
ual therapy for chronic hepatitis C virus (HCV) infection has consisted of pegylated interferon-a in
combination with ribavirin; however, this is only effective
in 40%–60% of HCV genotype 1 (g1) infections (Chou and
others 2013). Several host genetic factors are known to influence treatment response. The interleukin 28B (IL28B)
single-nucleotide polymorphism (SNP) rs12979860 is currently employed, in conjunction with other laboratory and
clinical parameters, as a pretreatment predictor of sustained
virologic response (SVR). Evidence exists to suggest that
typing for a number of different IL28B SNPs and identifying
an IL28B haplotype associated with treatment response may
more accurately predict treatment outcome than typing for
rs12979860 alone (Smith and others 2011).
Additional host markers predicting spontaneous HCV
clearance and treatment response in chronic HCV (CHC)infected individuals have also been described. The uptake of
the guanosine nucleotide analogue ribavirin by hepatocytes
1
2
3
National Virus Reference Laboratory, University College Dublin, Dublin, Ireland.
Liver Unit, St Vincent’s University Hospital, Dublin, Ireland.
Centre for Support and Training in Analysis and Research (CSTAR), University College Dublin, Dublin, Ireland.
1
2
suggesting this SNP as a marker for treatment response
(Trepo and others 2011; Valenti and others 2011).
The natural killer (NK) cell receptor family, killer
immunoglobulin-like receptor (KIR), and its ligand, HLAC, are known to play an important role in the immune response to many viral infections, including HCV [reviewed
in Rajagopalan and Long (2005)]. While some combinations
of KIRs and HLA-C appear to have a protective effect
against HCV infection or are associated with an SVR, other
combinations have been shown to have a detrimental effect
and contribute to chronic infection, disease progression, or
nonresponse to therapy (Khakoo and others 2004; Knapp
and others 2010). A relationship between KIRs, HLA-C, and
IL28B polymorphisms has also been identified (Dring and
others 2011; Suppiah and others 2011). Based upon these
data, we have investigated whether a combination of innate
immune genes and other loci may better predict the response
to therapy and improve the selection of patients for treatment, rather than 1 genetic biomarker alone.
In this retrospective, single-center, observational study,
we analyzed a suite of genetic markers in a cohort of HCVmonoinfected patients to evaluate the association of host
genetic markers, viral genotype, and early viral load kinetics
on treatment response in patients receiving interferon-based
therapy with ribavirin.
Materials and Methods
Study population
One hundred seventy-one individuals chronically infected
with HCV were included in the present study. All were monoinfected with either HCV g1 or g3. The HCV g1-infected cohort
comprised 89 individuals (46 responders and 43 nonresponders)
and the HCV g3-infected cohort consisted of 82 individuals (57
responders and 25 nonresponders). HCV viral load (IU/mL) was
determined for individuals at baseline (pretreatment) and at 4
and 12 weeks after starting treatment. Treatment responders
were defined as those with HCV RNA decline ‡ 2-log10 IU/mL
from pretreatment HCV viral load at week 12 and who had
undetectable levels of HCV RNA 24 weeks after treatment
cessation. Treatment nonresponders were defined as those individuals who did not achieve an on-treatment RNA decline
‡ 2-log10 IU/mL from baseline levels at 12 weeks or who had
detectable HCV RNA levels 24 weeks after treatment was
discontinued. Plasma and serum samples were obtained from
the National Virus Reference Laboratory (NVRL), University
College Dublin (UCD). Samples had been collected between
2002 and 2012 and stored at - 20C. Ethical approval for this
study was obtained from the ethics and medical research committee of St. Vincent’s University Hospital.
Treatment schedule
Peginterferon alpha 2a (Pegasys) was administered subcutaneously at a dose of 180 mg per week. Ribavirin was also
given at a dose of 1,000 mg/day for < 75 kg and 1,200 mg/day
for > 75 kg in the HCV g1-infected individuals and at
800 mg/day for HCV g3-infected individuals. Alternatively,
peginterferon alpha 2b (pegIntron) was administered subcutaneously at a dose of 1.5 mg per week and ribavirin was
given at a dose of 800–1,400 mg/day for HCV g1- and HCV
g3-infected individuals. Length of treatment was 48 weeks
for HCV g1 and 24 weeks for g3, with treatment cessation
COLLISON ET AL.
after 12 weeks if there was not a > 2 log10 decline in HCV
viral RNA for HCV g1 and g3.
Virological testing
Measurements of HCV viral load were made using either
the Cobas TaqMan HCV 1.0 assay (Roche Diagnostics,
Basel, Switzerland) or the VERSANT HCV 3.0 RNA (bDNA)
assay (Siemens Healthcare, Erlangen, Germany). Data were
analyzed in IU/mL using previously determined conversion
factors of 2.56 and 5.2 for the Roche and Siemens assays,
respectively. HCV genotyping data were generated using the
Innogenetics VERSANT HCV genotype 2.0 assay (Siemnes
Healthcare).
DNA isolation and genotyping
Genomic DNA was extracted manually from serum and
plasma using the viral RNA minikit as described by the manufacturer (Qiagen, Crawley, United Kingdom). Each gene of
interest was amplified by polymerase chain reaction sequencespecific primers (PCR-SSP) using HotStarTaq Master Mix
(Qiagen). Oligonucleotide primers (Metabion, Martinsried,
Germany) used to type the HLA-C genes (HLA-C1 and HLAC2) and the 4 KIR genes (KIR2DL3, KIR2DL2, KIR2DS3, and
KIR2DS2) were as previously described (Tajik and others
2010). For HLA typing, the final optimized concentration of
HLA-C1 and -C2 primers in each reaction was 0.4 mM, and for
the human growth hormone (HGH) endogenous control primers was 0.2 mM. For KIR typing, the final optimized concentration of KIR primers was 1 mM, and the HGH primers was
0.4 mM. End-point PCR for KIR and HLA typing was carried
out on a Veriti thermocycler (Applied Biosystems, Foster City,
CA) under the following conditions: 95C for 15 min; 40 cycles
of 94C for 1 min, 65C for 1 min, and 72C for 1 min; and
72C for 10 min. External quality control panels from the International Histocompatability Working Group (IHWG) in the
Fred Hutchinson Cancer Research Center (Seattle, WA) were
used to validate the genotyping methodologies and optimize the
specificity and sensitivity of the assays prior to screening. For
CCR5 genotyping, oligonucleotide primer sequences used for
end-point PCR across the 32-bp deletion within CCR5 (rs333)
were designed using Primer3Plus software (Primer3Plus.com)
and produced amplicons of 202 and 170 bp for wild-type and
mutant alleles, respectively. Primers were used at a final concentration of 1 mM (F: 5¢GTCTTCATTACACCTGCAGCTC
TC3¢, R: 5¢ATGGTGAAGATAAGCCTCACAGC3¢).
Assay-on-demand SNP genotyping assays (Applied Biosystems) were used for allelic discrimination of IL28B
rs12979860, rs8099917, rs7248668, ENT1 rs760370, and
PNPLA3 rs738409. The assays-by-design SNP genotyping
assay (Applied Biosystems) was employed for allelic discrimination of rs4803221. A detection mix consisting of a 40 ·
mix of unlabelled PCR primers and TaqMan minor-grovebinding probes labeled with a 5¢ VIC (2¢-chloro-7¢-phenyl-1, 4dichloro-6-carboxyfluorescein) for the major alleles and FAM
(6-carboxyfluorescein) for the minor alleles for each SNP was
diluted to 25 · in nuclease-free water. Allelic discrimination
assays were performed on the ABI7300 SDS Real-Time PCR
System (Applied Biosystems) with fluorescence acquisition
occurring during the annealing/extension stage. Pre- and postassay plate reads were performed for 1 min at 60C to determine the change in fluorescence during the real-time PCR.
HLA-C INFLUENCE ON TREATMENT RESPONSE IN CHRONIC HCV
Statistical analysis
Hardy–Weinberg equilibrium tests were performed for
SNPs investigated in the present study for association with
treatment response with a 1 degree of freedom (d.f.) chisquare goodness-of-fit test in the R statistical package,
version 2.15.0. Genetic association tests were performed in
PASW Statistics version 18. Single markers were tested for
association with nonresponse using a 2 d.f. chi-square test
on the 3 genotype categories, or a 1 d.f. chi-square test
assuming a dominant mode of inheritance on a reference
homozygote versus collapsed heterozygous and ‘‘risk’’ homozygous genotypes. Presence or absence of the KIR genes
was tested for association with nonresponse with a 1 d.f. chisquare test. Linear regression was applied to investigate the
additive (dosage) effect of alleles on baseline log10-transformed
HCV viral load. The dynamics of change in viral load from
baseline through to week-4 and -12 follow-up time-points
were examined using repeated-measures analysis of variance
(ANOVA). Genotypes of each genetic marker were examined
as a categorical between-subject factor, time as a withinsubject factor, and a genotype · time interaction term, to model
genotype-mediated differences in the pattern of change in
viral load over time.
Results
Correlation of sustained virologic response
with IL28B SNPs
In HCV g1-infected individuals, a strong association was
found between all 4 IL28B SNPs (rs12979860, rs8099917,
rs4803221, and rs7248668) and SVRs (Fig. 1). No signifi-
FIG. 1. Percentage frequency of interleukin 28B (IL28B)
single-nucleotide polymorphisms (SNPs) in hepatitis C virus (HCV) genotype 1 (g1)–infected treatment responders
versus nonresponders. The frequency of the 3 allelic groups—
minor homozygous, heterozygous, and major homozygous—
is shown for the 4 IL28B SNPs: rs12979860, rs8099917,
rs4803221, and rs7248668. For each SNP, allele frequencies are compared between HCV g1-infected responders and
nonresponders.
3
cant association between the 4 IL28B polymorphisms and
HCV g3-infected patients was observed (Fig. 2). Both
dominant and genotypic models were used to identify the
most appropriate model to predict treatment failure for
the HCV g1-infected cohort. Assuming a dominant model,
the minor alleles of all 4 IL28B SNPs were found to be
significantly associated with treatment nonresponse (P =
3.83 · 10 - 4, P = 0.004, P = 0.0019, and P = 0.014), respectively. Rs12979860 was the most significant, with the T risk
allele the best independent genetic predictor of nonresponse
to treatment using the dominant model. With the genotypic
‘‘additive’’ model, a dosage effect of the minor T allele was
observed, with heterozygotes less likely to respond to
treatment than major homozygotes (CC), and minor homozygotes (TT) less likely to respond to treatment than heterozygotes (CT) with only 1 copy of the risk allele. The
rs12979860 minor homozygote (TT) gave a positive predictive value for treatment failure of 81.3%, with the presence of the major allele homozygote (CC) giving a negative
predictive value of 77.4% for treatment failure (Table 1).
In our cohort, while the IL28B haplotype described by
Smith and others (2011) was a significant predictor of treatment outcome in HCV g1-infected individuals (P = 0.0053),
the association with treatment nonresponse was not as strong
as rs12979860 alone. As rs12979860 was found to be the
most significantly associated with virologic response, this
SNP was used in all subsequent analyses.
While the IL28B SNPs were tested for association with
nonresponse with a 2 d.f. genotypic model, and with a 1 d.f.
dominant model for the minor allele, the pattern of the
genotypic odds ratios (ORs) did not indicate obvious dominance. A logistic regression model was fitted with an additive ( - 1, 0, 1) term and a dominance ( - 0.5, 0.5, - 0.5)
genotype parameterization to determine whether any significant
dominance variance was present. The dominance term was
not significant (P = 0.881; OR = 1.077) in the presence of the
additive term (P = 4.67 · 10 - 4; OR = 3.85), for rs12979860
(in HCV g1). Therefore, only the additive term was used
in subsequent analyses. In HCV g3-infected individuals, no
FIG. 2. Percentage frequency of IL28B SNPs in HCV
genotype 3 (g3)–infected treatment responders versus nonresponders.
4
COLLISON ET AL.
Table 1. Predictive Values of the IL28B
Genetic Variants in HCV g1- and HCV g3-Infected
Individuals Receiving Interferon/Ribavirin Therapy
HCV
genotype
g1
Host
marker
Reference
variant
Risk
variant
PPV
%
NPV
%
rs12979860
Ca
CC
Ca
CC
Ta
TT
Ga
GG
Ca
CC
Ca
CC
Ta
TT
Ga
GG
TT
Ta
GG
Ga
GG
Ga
AA
Aa
TT
Ta
GG
Ga
GG
Ga
AA
Aa
81.3
62.1
75.0
63.8
75.0
63.0
75.0
60.9
16.7
30.2
0.0
33.3
0.0
33.3
50.0
34.4
58.9
77.4
54.3
69.0
54.3
67.4
54.3
65.1
68.4
69.2
69.1
71.4
69.1
71.4
70.0
72.0
rs4803221
rs8099917
rs7248668
g3
rs12979860
rs4803221
rs8099917
rs7248668
Prediction measures were calculated using nonresponse as the
outcome of interest, and each genetic variant as the ‘‘test’’ for
nonresponse.
a
Refers to all genotypes containing that allele.
Bold values refer to the alleles with the highest positive predictive
and negative predictive values.
g1, Genotype 1; g3, genotype 3; HCV, hepatitis C virus; IL28B,
interleukin 28B; PPV, positive predictive value: the proportion of
test positive (risk variant carriers) who were nonresponders; NPV,
negative predictive value: the proportion of test negative (reference
variant carriers) who were responders.
FIG. 3. Frequency of HLA-C1 and -C2 alleles in HCV g1
and g3 treatment responders versus nonresponders.
in HCV g1 (n = 89) and g3 (n = 82) responder and nonresponder groups. No significant difference was observed
between either HCV g1-infected group (P = 0.575) or HCV
g3-infected group (P = 0.557). The frequency of the PNPLA3
rs738409 SNP was also examined between groups and no
significant difference was found between responders and
nonresponders in either HCV genotypes.
Influence of host genetics on early viral kinetics
association was observed between any of the IL28B SNPs
and virologic response.
Analysis of host genetic markers and sustained
virologic response
HLA-C1 and -C2 typing was performed on 138 patients
(71 HCV g1 and 67 HCV g3) to evaluate the influence of
HLA-C on virological response and the frequency of HLA-C
in responders and nonresponders in HCV g1- and g3infected individuals (Fig. 3). Within the HCV g1-infected
group, HLA-C was found to be associated with nonresponse
to treatment (P = 0.023), and notably the frequency of HLAC2 homozygotes was higher in treatment nonresponders
compared with responders (7/37; 18.9% vs. 0/34; 0%). In
the HCV g3-infected group, no significant association was
observed between HLA-C type and nonresponse to treatment
(Fig. 3). There was a higher frequency of HLA-C1 homozygotes in nonresponders when compared with responders
(13/23; 56.5% and 13/44; 29.5%, respectively) and a lower
frequency of C1/C2 heterozygotes in nonresponders (9/23;
39.1%) when compared with responders (26/44; 59.1%) but
this was not significant (P = 0.09).
The frequencies of 4 KIR genes (KIR2DL3, KIR2DL2,
KIR2DS3, and KIR2DS2) in the study were compared between
responders and nonresponders in both HCV g1- (n = 87) and
g3-infected groups (n = 80; Fig. 4). No significant differences were found between frequencies of the genes in responders and nonresponders in either genotype groups. The
allele frequency of the CCR5D32 mutation was examined
Early viral load kinetics (log10 IU/mL) were examined for
association with the following genetic markers: IL28B
rs12979860, the IL28B haplotype, ENT1 rs760370, PNPLA3
rs738409, and CCR5D32. HCV RNA levels (IU/mL) were
obtained for patients at 3 different time-points: baseline
(pretreatment), 4, and 12 weeks after start of treatment.
All markers were examined to determine any difference
in baseline HCV viral load between allelic genotypes. The
markers ENT1 rs760370, PNPLA3 rs738409, CCR5D32, and
IL28B haplotype showed no association. IL28B rs12979860
FIG. 4. Comparison of the frequencies of the KIR genes in
HCV g1 and g3 treatment responders and nonresponders.
HLA-C INFLUENCE ON TREATMENT RESPONSE IN CHRONIC HCV
showed no association in the HCV g1-infected group
(P = 0.124), but was found to be associated with baseline
viral load in the HCV g3-infected group (P = 0.024). Baseline viral load was highest in major homozygotes (CC), and
lower with each additional copy of the T allele. The pattern of
allelic groups compared to changes in HCV RNA over time
was also examined. ENT1 rs760370 and CCR5D32 showed no
significant association in either HCV g1 or g3-infected
groups. However, the IL28B SNP rs12979860 was associated
with a difference in viral load over time between major homozygotes (CC), heterozygotes (CT), and minor homozygotes (TT) in both HCV g1-infected (n = 68; P = < 0.001) and
g3-infected (n = 70; P = 0.004) groups. The association of
rs12979860 was observed in both responders (n = 35;
P = 0.048) and nonresponders (n = 33; P = < 0.001) of HCV
g1-infected individuals, and in responders of HCV g3-infected
individuals (n = 49; P = 0.028) but not nonresponders (n = 21).
The effect of HLA-C1 and -C2 alleles on RVR (defined as
undetectable viral RNA at week 4 post-treatment) in individuals who achieved an SVR was investigated in HCV g1
cases. Due to the high RVR and SVR in the HCV g3infected cohort, too few individuals were present for statistical analysis. Logistic regression showed an association
between HLA-C1/C1 and RVR, with an OR relative to the
heterozygote of 10.0 [95% confidence interval (CI): 1.6–62.5,
P = 0.014].
Discussion
Since the identification of an association between polymorphisms near the IL28B gene and treatment-induced
clearance of HCV by independent genome-wide association
studies conducted in 2009 (Ge and others 2009; Suppiah and
others 2009; Tanaka and others 2009), IL28B SNPs and HCV
infection have been studied extensively. However, the best
genetic model to represent the association between IL28B
genotype and treatment-induced HCV clearance is still unclear. Some authors report a dominant effect of the minor risk
allele, with no significant differences observed between heterozygotes and minor homozygotes in response to treatment
(Tanaka and others 2009). This dominant model suggests that
the protective effect of the major allele of the IL28B SNPs is
primarily recessive and therefore only effective in individuals
homozygous for the major allele. Other studies suggest an
additive model, with a dosage effect of the risk allele such that
heterozygotes with only 1 copy of the risk allele respond
better to treatment than minor homozygotes with 2 copies of
the risk allele (Mangia and others 2010). In the present study,
all SNPs were examined using both dominant and additive
models to investigate the best model for predicting treatment
outcome. The pattern of the genotypic OR did not indicate
obvious dominance of the minor allele. The nonsignificance of
the dominance term in the IL28B SNP rs12979860 (P = 0.881)
in the presence of the additive term (P = 4.67 · 10 - 4) indicates
a dosage effect of the minor T allele in the cohort. A study by
McFarland and others (2014) identified a functional polymorphism (rs4803217) in the 3¢ untranslated region of the
IL28B (IFNL3) mRNA that influences the stability of the
transcript by targeting inhibition of HCV-induced microRNAs. The group found the IL28B SNP rs12979860 to be in
linkage disequilibrium with the functional rs4803217 SNP and
propose that the latter SNP is responsible for the association of
rs12979860 with clearance of HCV.
5
Prokunina-Olsson and others (2013) identified a novel
IFN-l family member, IFNL4, which is orthologous to but
distinct from known IFNs and other class 2 cytokines. The
IFNL4-creating polymorphism ss469415590[DG] allele is
perfectly correlated with the unfavorable rs12979860[T]
allele in Asians and well correlated in Europeans; however,
it is only moderately correlated in African populations. The
IL28B markers rs12979860 and rs8099917 are located 367bp downstream and 4-kb upstream of ss469415590, respectively. The ss469415590 variant and rs12979860 have
been shown to predict equivalently the virologic response to
dual therapy interferon-a and ribavirin in Caucasian HCV
g1/g4-infected patients (Real and others 2014) and in Asian
HCV g1-infected patients (Akkarathamrongsin and others
2014). However, typing of the former IFNL4 polymorphism
may be more informative for prediction of treatment response rates in chronic-HCV-infected African patients
(Prokunina-Olsson and others 2013). Pretreatment assessment of ss469415590 is also predictive in Asian CHC patients receiving triple therapy with the protease inhibitor
telaprevir in combination with IFN-a with ribavirin (Nagaoki
and others 2014).
Several studies that examined the association of IL28B
SNPs on baseline viral load have found an association between the major homozygotes (CC) and higher baseline
viral loads (Ge and others 2009). In our cohort, IL28B
rs12979860 CC homozygotes had a higher baseline viral
load than CT heterozygotes, whom in turn had a higher
baseline viral load than TT homozygotes. However, after
transforming HCV viral load using log10, this association
was only significant in the g3-infected group (P = 0.004).
Lower baseline viral load for TT homozygotes has previously been shown in an HCV g3-infected cohort by Stenkvist and others (2013) and it is possible that this lower
baseline level may play a role in the detrimental effect of the
allele in treatment response.
The IL28B SNP rs12979860 was found to be significantly
associated with early viral kinetics in both HCV g1- and g3infected groups. The average viral load at the 3 time-points
(baseline, 4, and 12 weeks after start of treatment combined)
was significantly different between major homozygotes
(CC), heterozygotes (CT), and minor homozygotes (TT;
P = < 0.001 and P = 0.048, respectively). In the HCV g1infected group, rs12979860 CC major homozygotes had a
significantly lower estimated measure of mean viral load
than either heterozygotes (CT) or minor homozygotes (TT)
at week 4. As rs12979860 CC major homozygotes also had a
significantly increased chance of achieving SVR, this would
suggest that an early decline of viral load is a predictor of
treatment response. Analyzing HCV g1-infected responders
and nonresponders separately, viral loads in all 3 genotypes
of rs12979860 were reduced at week 4 in responders when
compared with nonresponders. An association between early
decline in viral load and treatment response has been shown
in previous studies (Ferenci 2004) and this study and others
suggest its potential therapeutic use in predicting treatment
outcome in HCV g1-infected individuals. Interestingly, in
the HCV g1-infected responder group, the 1 individual who
was IL28B rs12979860 minor homozygote (TT) achieved
RVR, with no detectable HCV RNA at week 4.
A significant finding of the present study is the absence
of significant associations of the 4 IL28B SNPs influencing dual-therapy, treatment-induced response rates in HCV
6
g3-infected individual, in support of findings by Rauch and
others (2010) who have reported that rs8099917 had no
significant impact on SVR in a Swiss cohort infected with
HCV g2 or g3. In contrast, in an Italian cohort, Mangia who
analyzed the rs12979860 SNP identified an association with
SVR in patients infected with HCV g2 and g3 who did not
achieve an RVR (Mangia and others 2010). Sarazzin and
others (2011) in a German study of HCV g2- and g3-infected
individuals found that rs12979860, but not rs8099917, was
significantly associated with RVR and SVR.
Morello and others (2010) found the SNP rs760370 at the
ENT1 gene to be significantly associated with RVR in HIV/
HCV-coinfected individuals, and speculated that the A > G
polymorphism affects the expression or activity of the ENT1
protein, which in turn modulates uptake of ribavirin in hepatocytes. No significant association was observed in the
present study between the rs760370 SNP and baseline viral
load, average viral load between 0 and 12 weeks, or pattern
of change of viral load between 0 and 12 weeks, in either
HCV g1- or g3-infected groups.
Studies that investigated a role of the CCR5D32 allele in
HCV acquisition report conflicting data with regard to
spontaneous clearance of the virus and susceptibility to infection (Mascheretti and others 2004; Goulding and others
2005; Nattermann and others 2011). CCR5D32 heterozygosity has a frequency of 9% in the Irish population
(Martinson and others 1997), and consequently the statistical power is limited in our study due to the low frequency of
the allele. Numbers of individuals within a study would
need to be very large in order to find a significant association between CCR5D32 and treatment outcome. In this regard, a large study conducted by Suppiah and others (2013)
to clarify the role of CCR5D32 in treatment response to dual
therapy in CHC-infected individuals found that CCR5D32
did not influence treatment response, or improve the power
of the IL28B SNP rs12979860 in predicting treatment response.
The NK cell receptor family, KIR, and its ligand, HLA-C,
are known to play an important role in the immune response
to many viral infections, including HCV. In HCV g1infected individuals in this study, all HLA-C2 homozygotes
failed to respond to treatment. These data suggest an important role for HLA-C2 in determining HCV treatment
outcome, perhaps through a suppression of NK cytotoxicity,
resulting in a reduced killing of infected hepatocytes. Suppiah and others (2011) hypothesized that as HLA-C1 is
known to activate NK cells more rapidly and aggressively
than HLA-C2 in the natural course of infection, upregulation
of HLA-C by IFN-a during treatment would also cause a
stronger activation of NK cells by HLA-C1 than HLA-C2.
NK cells are known to influence adaptive immune responses
to infection. The production of cytokines, such as IFN-g, by
NK cells influences T cell responses in lymph nodes, and
NK-cell-mediated killing of target cells also impacts T cell
responses, possibly by decreasing the antigenic load of the T
cells or by promoting antigen cross-presentation to cytotoxic
T cells (Krebs and others 2009). HLA-C phenotype may
influence the interaction between NK cells and the adaptive
immune system, which could explain the absence of HLAC2 homozygotes in HCV g1-infected individuals responding
to treatment in our present study (Andrews and others 2010).
In this study, no association was observed between any of
the 4 KIR genes studied—KIR2DL3, KIR2DL2, KIR2DS3,
and KIR2DS2—with treatment response. Strikingly, the
COLLISON ET AL.
frequencies of all 4 genes were almost identical between
responders and nonresponders in both HCV g1- and g3infected groups and suggest that they may not be of as great
importance as genetic predictors of treatment response in the
context of HCV as others have shown (Suppiah and others
2011); however, this could be due to the limited power of
our study. Overall, the results of the study suggest that in
HCV g1-infected individuals receiving dual therapy, HLA-C
alone determines treatment outcome, regardless of KIR
genotype.
In the present study, IL28B SNPs and HLA-C2/C2 were
found to independently predict treatment nonresponse in
HCV g1-infected individuals. Forty-two of the 89 HCV g1infected individuals were CT for IL28B rs12979860 (47%).
Of these heterozygotes, almost half (19/42; 44%) responded
to treatment. Of note, all HLA-C2/C2 individuals (7/71)
were rs12979860 heterozygous, and none achieved an SVR.
These data suggest that homozygosity for HLA-C2 predicts
nonresponse to treatment, regardless of the fact that the
IL28B heterozygous genotype would predict a response rate
of almost 50%. In the multivariate analysis, an additive
model was used for both IL28B rs12979860 and HLA-C, and
no interaction was observed between the 2 loci. As a detrimental effect of HLA-C2 has been observed in HLA-C2
homozygotes but not HLA-C1/C2 heterozygotes, this may
account for an absence of an interaction between these 2
innate immune genes.
Potential limitations of this study include the cohort size
and the investigation of genetic influence on treatment
outcome specifically in individuals receiving interferon-a
and ribavirin therapy. Larger studies are required to investigate potential associations of IL28B rs12979860, HLA-C2,
and other genetic markers with treatment outcome in antiHCV protease and polymerase-based therapies due to the
prohibitive cost of the newer therapeutic options and the
cost benefit of identifying individuals who may respond to
dual therapy.
The findings in the present study demonstrate that IL28B
rs12979860 is the most significant predictor of treatment
response in HCV g1- but not g3-infected individuals and
that the HLA-C2 locus is an important pretreatment genetic
predictor of treatment failure also in HCV g1- but not g3infected individuals receiving dual IFN/RBV therapy. The
inclusion of HLA-C1 and HLA-C2 typing in a clinical algorithm requires further investigation for determining
treatment response rates in HCV g1-infected individuals and
also for determining the influence of these loci on treatmentinduced clearance with direct-acting antiviral therapy. These
findings are of particular relevance in resource-limited settings where the prohibitive cost of the direct-acting antivirals for treatment of CHC-infected individuals may preclude
their introduction and dual-therapy, interferon-containing
regimens may remain the only treatment option available.
Author Disclosure Statement
Meadhbh Collison: No relationships/conditions/circumstances that present potential conflict of interest. Jun Liong
Chin: No relationships/conditions/circumstances that present potential conflict of interest. Ahmed Abu Shanab:
Meeting expenses—AASL meeting 2013 sponsored by
Janssen and Janssen. Ross Mac Nicholas: No relationships/
conditions/circumstances that present potential conflict of
HLA-C INFLUENCE ON TREATMENT RESPONSE IN CHRONIC HCV
interest. Ricardo Segurado: Payment of consulting fee for
article preparation as part of CSTAR, which is a not-forprofit Academic Research Centre in University College
Dublin, which provides statistical support for academic and
industry researchers for an hourly fee. Coauthorship is independent of the fee, and assessed on the basis of ICMJE
guidelines (see cstar.ie). Suzie Coughlan: No relationships/
conditions/circumstances that present potential conflict of
interest. Jeff Connell: No relationships/conditions/circumstances that present potential conflict of interest. Michael J.
Carr: No relationships/conditions/circumstances that present
potential conflict of interest. Raphael B. Merriman: No relationships/conditions/circumstances that present potential
conflict of interest. P. Aiden McCormick: Board membership of Janssen, Abbvie, and M.S.D.; Money for secretarial
support for hepatoma clinic from Bayer. William W. Hall:
No relationships/conditions/circumstances that present potential conflict of interest.
References
Akkarathamrongsin S, Thong VD, Payungporn S, Poovorawan
K, Prapunwattana P, Poovorawan Y, Tangkijvanich P. 2014.
IFNL3 (IL28B) and IFNL4 polymorphisms are associated
with treatment response in Thai patients infected with HCV
genotype 1, but not with genotypes 3 and 6. J Med Virol
86(9):1482–1490.
Andrews DM, Estcourt MJ, Andoniou CE, Wikstrom ME,
Khong A, Voigt V, Fleming P, Tabarias H, Hill GR, van der
Most RG, Scalzo AA, Smyth MJ, Degli-Esposti MA. 2010.
Innate immunity defines the capacity of antiviral T cells to
limit persistent infection. J Exp Med 207(6):1333–1343.
Chou R, Hartung D, Rahman B, Wasson N, Cottrell EB, Fu R.
2013. Comparative effectiveness of antiviral treatment for
hepatitis C virus infection in adults: a systematic review. Ann
Intern Med 158(2):114–123.
Dring MM, Morrison MH, McSharry BP, Guinan KJ, Hagan R,
O’Farrelly C, Gardiner CM. 2011. Innate immune genes synergize to predict increased risk of chronic disease in hepatitis C virus infection. Proc Natl Acad Sci U S A 108(14):
5736–5741.
Ferenci P. 2004. Predicting the therapeutic response in patients
with chronic hepatitis C: the role of viral kinetic studies.
J Antimicrob Chemother 53(1):15–18.
Fukuchi Y, Furihata T, Hashizume M, Iikura M, Chiba K. 2010.
Characterization of ribavirin uptake systems in human hepatocytes. J Hepatol 52(4):486–492.
Ge D, Fellay J, Thompson AJ, Simon JS, Shianna KV, Urban
TJ, Heinzen EL, Qiu P, Bertelsen AH, Muir AJ, Sulkowski
M, McHutchison JG, Goldstein DB. 2009. Genetic variation
in IL28B predicts hepatitis C treatment-induced viral clearance. Nature 461(7262):399–401.
Goulding C, McManus R, Murphy A, MacDonald G, Barrett S,
Crowe J, Hegarty J, McKiernan S, Kelleher D. 2005. The
CCR5-delta32 mutation: impact on disease outcome in individuals with hepatitis C infection from a single source. Gut
54(8):1157–1161.
Guedj H, Guedj J, Negro F, Lagging M, Westin J, Bochud PY,
Bibert S, Neumann AU. 2012. The impact of fibrosis and
steatosis on early viral kinetics in HCV genotype 1-infected
patients treated with Peg-IFN-alfa-2a and ribavirin. J Viral
Hepat 19(7):488–496.
He J, Chen Y, Farzan M, Choe H, Ohagen A, Gartner S, Busciglio
J, Yang X, Hofmann W, Newman W, Mackay CR, Sodroski J,
7
Gabuzda D. 1997. CCR3 and CCR5 are co-receptors for HIV-1
infection of microglia. Nature 385(6617):645–649.
Khakoo SI, Thio CL, Martin MP, Brooks CR, Gao X, Astemborski J, Cheng J, Goedert JJ, Vlahov D, Hilgartner M,
Cox S, Little AM, Alexander GJ, Cramp ME, O’Brien SJ,
Rosenberg WM, Thomas DL, Carrington M. 2004. HLA and
NK cell inhibitory receptor genes in resolving hepatitis C
virus infection. Science 305(5685):872–874.
Knapp S, Warshow U, Hegazy D, Brackenbury L, Guha IN,
Fowell A, Little AM, Alexander GJ, Rosenberg WM, Cramp
ME, Khakoo SI. 2010. Consistent beneficial effects of killer
cell immunoglobulin-like receptor 2DL3 and group 1 human
leukocyte antigen-C following exposure to hepatitis C virus.
Hepatology 51(4):1168–1175.
Krebs P, Barnes MJ, Lampe K, Whitley K, Bahjat KS, Beutler
B, Janssen E, Hoebe K. 2009. NK-cell-mediated killing of
target cells triggers robust antigen-specific T-cell-mediated
and humoral responses. Blood 113(26):6593–6602.
Mangia A, Thompson AJ, Santoro R, Piazzolla V, Tillmann HL,
Patel K, Shianna KV, Mottola L, Petruzzellis D, Bacca D,
Carretta V, Minerva N, Goldstein DB, McHutchison JG.
2010. An IL28B polymorphism determines treatment response of hepatitis C virus genotype 2 or 3 patients who do
not achieve a rapid virologic response. Gastroenterology
139(3):821–827.e1.
Martinson JJ, Chapman NH, Rees DC, Liu YT, Clegg JB. 1997.
Global distribution of the CCR5 gene 32-basepair deletion.
Nat Genet 16(1):100–103.
Mascheretti S, Hinrichsen H, Ross S, Buggisch P, Hampe J,
Foelsch UR, Schreiber S. 2004. Genetic variants in the CCR
gene cluster and spontaneous viral elimination in hepatitis
C-infected patients. Clin Exp Immunol 136(2):328–333.
McFarland AP, Horner SM, Jarret A, Joslyn RC, Bindewald E,
Shapiro BA, Delker DA, Hagedorn CH, Carrington M, Gale
M, Jr., Savan R. 2014. The favorable IFNL3 genotype escapes mRNA decay mediated by AU-rich elements and
hepatitis C virus-induced microRNAs. Nat Immunol 15(1):
72–79.
Morello J, Cuenca L, Soriano V, Medrano J, Madejon A, Vispo
E, Barreiro P, Labarga P, Jimenez-Nacher I, Rodriguez-Novoa S. 2010. Influence of a single nucleotide polymorphism
at the main ribavirin transporter gene on the rapid virological
response to pegylated interferon-ribavirin therapy in patients
with chronic hepatitis C virus infection. J Infect Dis
202(8):1185–1191.
Nagaoki Y, Imamura M, Kawakami Y, Kan H, Fujino H, Fukuhara T, Kobayashi T, Ono A, Nakahara T, Naeshiro N,
Urabe A, Yokoyama S, Miyaki D, Murakami E, Kawaoka T,
Tsuge M, Hiramatsu A, Aikata H, Takahashi S, Hayes CN,
Ochi H, Chayama K. 2014. Interferon lambda 4 polymorphism affects on outcome of telaprevir, pegylated interferon
and ribavirin combination therapy for chronic hepatitis C.
Hepatol Res DOI: 10.1111/hepr.12336.
Nattermann J, Timm J, Nischalke HD, Olbrich A, Michalk M,
Tillmann HL, Berg T, Wedemeyer H, Tenckhoff H, Wiese
M, Kullig U, Gobel U, Capka E, Schiefke I, Guthof W,
Grungreiff K, Konig I, Roggendorf M, Sauerbruch T, Spengler U. 2011. The predictive value of IL28B gene polymorphism for spontaneous clearance in a single source outbreak
cohort is limited in patients carrying the CCR5Delta32 mutation. J Hepatol 55(6):1201–1206.
Prokunina-Olsson L, Muchmore B, Tang W, Pfeiffer RM, Park
H, Dickensheets H, Hergott D, Porter-Gill P, Mumy A, Kohaar I, Chen S, Brand N, Tarway M, Liu L, Sheikh F, Astemborski J, Bonkovsky HL, Edlin BR, Howell CD, Morgan
8
TR, Thomas DL, Rehermann B, Donnelly RP, O’Brien TR.
2013. A variant upstream of IFNL3 (IL28B) creating a new
interferon gene IFNL4 is associated with impaired clearance
of hepatitis C virus. Nat Genet 45(2):164–171.
Rajagopalan S, Long EO. 2005. Understanding how combinations of HLA and KIR genes influence disease. J Exp Med
201(7):1025–1029.
Rauch A, Kutalik Z, Descombes P, Cai T, Di Iulio J, Mueller T,
Bochud M, Battegay M, Bernasconi E, Borovicka J, Colombo
S, Cerny A, Dufour JF, Furrer H, Gunthard HF, Heim M,
Hirschel B, Malinverni R, Moradpour D, Mullhaupt B, Witteck A, Beckmann JS, Berg T, Bergmann S, Negro F, Telenti
A, Bochud PY. 2010. Genetic variation in IL28B is associated with chronic hepatitis C and treatment failure: a genomewide association study. Gastroenterology 138(4):1338–
1345.e1–e7.
Real LM, Neukam K, Herrero R, Guardiola JM, Reiberger T,
Rivero-Juarez A, Salazar J, Mandorfer M, Merino D, Soriano
V, Rivero A, Macias J, Pineda JA, Caruz A. 2014. IFNL4
ss469415590 variant shows similar performance to
rs12979860 as predictor of response to treatment against
hepatitis C virus genotype 1 or 4 in caucasians. PLoS One
9(4):e95515.
Sarrazin C, Susser S, Doehring A, Lange CM, Muller T,
Schlecker C, Herrmann E, Lotsch J, Berg T. 2011. Importance of IL28B gene polymorphisms in hepatitis C virus
genotype 2 and 3 infected patients. J Hepatol 54(3):415–421.
Smith KR, Suppiah V, O’Connor K, Berg T, Weltman M, Abate
ML, Spengler U, Bassendine M, Matthews G, Irving WL,
Powell E, Riordan S, Ahlenstiel G, Stewart GJ, Bahlo M,
George J, Booth DR. 2011. Identification of improved IL28B
SNPs and haplotypes for prediction of drug response in
treatment of hepatitis C using massively parallel sequencing
in a cross-sectional European cohort. Genome Med 3(8):57.
Stenkvist J, Sonnerborg A, Weiland O. 2013. HCV RNA decline in chronic HCV genotype 2 and 3 during standard of
care treatment according to IL28B polymorphism. J Viral
Hepat 20(3):193–199.
Suppiah V, Armstrong NJ, O’Connor KS, Berg T, Weltman M,
Abate ML, Spengler U, Bassendine M, Dore GJ, Irving WL,
Powell E, Nattermann J, Mueller T, Riordan S, Stewart GJ,
George J, Booth DR, Ahlenstiel G, Michalk M, Malik B,
McClure P, Smith S, Sheridan D, Snape E, Fragomeli V,
Norris R, How-Chow D, Jonsson JR, Barrie H, Stelzer-Braid
S, Fletcher S, Applegate T, Grebely J, Matthews G, Bharadwaj M, Smedile A. 2013. CCR5-Delta32 genotype does not
improve predictive value of IL28B polymorphisms for
treatment response in chronic HCV infection. Genes Immun
14(5):286–290.
Suppiah V, Gaudieri S, Armstrong NJ, O’Connor KS, Berg T,
Weltman M, Abate ML, Spengler U, Bassendine M, Dore GJ,
COLLISON ET AL.
Irving WL, Powell E, Hellard M, Riordan S, Matthews G,
Sheridan D, Nattermann J, Smedile A, Muller T, Hammond
E, Dunn D, Negro F, Bochud PY, Mallal S, Ahlenstiel G,
Stewart GJ, George J, Booth DR. 2011. IL28B, HLA-C, and
KIR variants additively predict response to therapy in chronic
hepatitis C virus infection in a European Cohort: a crosssectional study. PLoS Med 8(9):e1001092.
Suppiah V, Moldovan M, Ahlenstiel G, Berg T, Weltman M,
Abate ML, Bassendine M, Spengler U, Dore GJ, Powell E,
Riordan S, Sheridan D, Smedile A, Fragomeli V, Muller T,
Bahlo M, Stewart GJ, Booth DR, George J. 2009. IL28B is
associated with response to chronic hepatitis C interferonalpha and ribavirin therapy. Nat Genet 41(10):1100–1104.
Tajik N, Shahsavar F, Nasiri M, Radjabzadeh MF. 2010.
Compound KIR-HLA genotype analyses in the Iranian population by a novel PCR-SSP assay. Int J Immunogenet
37(3):159–168.
Tanaka Y, Nishida N, Sugiyama M, Kurosaki M, Matsuura K,
Sakamoto N, Nakagawa M, Korenaga M, Hino K, Hige S, Ito
Y, Mita E, Tanaka E, Mochida S, Murawaki Y, Honda M,
Sakai A, Hiasa Y, Nishiguchi S, Koike A, Sakaida I, Imamura M, Ito K, Yano K, Masaki N, Sugauchi F, Izumi N,
Tokunaga K, Mizokami M. 2009. Genome-wide association
of IL28B with response to pegylated interferon-alpha and
ribavirin therapy for chronic hepatitis C. Nat Genet
41(10):1105–1109.
Trepo E, Pradat P, Potthoff A, Momozawa Y, Quertinmont E,
Gustot T, Lemmers A, Berthillon P, Amininejad L, Chevallier M, Schlue J, Kreipe H, Deviere J, Manns M, Trepo C,
Sninsky J, Wedemeyer H, Franchimont D, Moreno C. 2011.
Impact of patatin-like phospholipase-3 (rs738409 C > G)
polymorphism on fibrosis progression and steatosis in chronic
hepatitis C. Hepatology 54(1):60–69.
Valenti L, Rumi M, Galmozzi E, Aghemo A, Del Menico B, De
Nicola S, Dongiovanni P, Maggioni M, Fracanzani AL, Rametta R, Colombo M, Fargion S. 2011. Patatin-like phospholipase domain-containing 3 I148M polymorphism,
steatosis, and liver damage in chronic hepatitis C. Hepatology
53(3):791–799.
Address correspondence to:
Prof. William W. Hall
National Virus Reference Laboratory
University College Dublin
Belfield
Dublin 4
Ireland
E-mail: william.hall@ucd.ie
Received 28 May 2014/Accepted 24 July 2014