Importance of Host Genetic Factors HLA and IL28B as Predictors of Response to Pegylated Interferon...
Transcript of Importance of Host Genetic Factors HLA and IL28B as Predictors of Response to Pegylated Interferon...
nature publishing group ORIGINAL CONTRIBUTIONS LI
VE
R
1246
The American Journal of GASTROENTEROLOGY VOLUME 106 | JULY 2011 www.amjgastro.com
INTRODUCTION Current therapy for chronic hepatitis C (CHC), a combination of
pegylated- α interferon (pegIFN- α ) and ribavirin (RBV), achieves a
response rate of 48 – 88 % ( 1 ). Viral factors, such as genotype, base-
line viral load, rapid virological response, early virological response,
and amino-acid pattern in the viral genome ( 2 – 4 ), and host factors,
including age, sex, race, liver fi brosis, and obesity, have been reported
to be associated with pegIFN- α / RBV therapy outcome ( 5 ).
Because pegIFN has direct antiviral eff ects and RBV promotes
a type 1 cytokine-mediated immune response that can enhance
antiviral immune responses, there might be an association
between immunogenetic characteristics and response to antiviral
therapy ( 6 ).
Strong, multispecifi c T-cell responses are thought to be mainly
responsible for viral elimination and disease pathogenesis during
hepatic C virus (HCV) infection ( 7 ). Th e eff ective presentation
of viral antigens to CD4 + and CD8 + T cells by HLA class I and
class II molecules, respectively, is the key regulator of optimum
immune response against viral infection. Two alleles are consis-
tently associated with viral clearance and the decreased disease
Importance of Host Genetic Factors HLA and IL28B as Predictors of Response to Pegylated Interferon and Ribavirin Paloma Mu ñ oz de Rueda , PhD 1 , 2 , 8 , Miguel- Á ngel L ó pez-Nevot , MD, PhD 3 , 8 , Pablo S á enz-L ó pez , PhD 3 , Jorge Casado , BS 1 , Antonia Mart í n-Casares 3 , Pablo Palomares , MD, PhD 1 , Rosa Quiles , PhD 1 , 2 , Ana Gila , MD, PhD 1 , 2 , Manuel Romero-G ó mez , MD, PhD 2 , 4 , Esther-Jos é Pav ó n , PhD 1 , Jos é -Antonio Mu ñ oz , PhD 1 , Á ngel Carazo , PhD 1 , Paloma Sanz-Cameno , PhD 2 , 5 , Ricardo Moreno-Otero , MD, PhD 2 , 5 , Mois é s Diago , MD, PhD 6 , Josefa Le ó n , PhD 1 , 2 , Á ngeles Ruiz-Extremera , MD, PhD 2 , 7 and Javier Salmer ó n , MD, PhD 1 , 2
OBJECTIVES: Viral factors are considered the best predictors of response to treatment for chronic hepatitis C (CHC), but genetic factors are known to have an important role in this respect. This paper investigates the relationships among the host genetic factors HLA and IL28B, viral factors, and the outcome of combination therapy.
METHODS: A multicenter retrospective cohort of 428 previously untreated CHC patients was treated with pegylated interferon / ribavirin (pegIFN / RBV) for 48 weeks. In all, 378 (88 % ) of these patients were genotype 1 or 4, and 50 (12 % ) were genotype 2 or 3.
RESULTS: Multivariate logistic regression showed the rs12979860 CC genotype (adjusted odds ratio (aOR) = 4.3, 95 % confi dence interval (95 % CI): 2.6 – 7), the HLA-DQB1 * 0301 allele (aOR = 2.08, 95 % CI: 1.2 – 3.5) and age, viral genotype, and viral load levels to be signifi cantly associated with sustained virological response (SVR). When the variable rs12979860 was eliminated, the area under the receiver operating characteristic (ROC) curve (AUC) decreased signifi cantly (0.76 vs. 0.69; P = 0.03). AUC values derived from viral factors were lower than those corresponding to host genetic factors (0.67 vs. 0.72, respectively; P = 0.04). The HLA-DQB1 * 0301 and A * 0201 alleles were associated with rs12979860 CC genotype and SVR ( P < 0.0001).
CONCLUSIONS: The HLA-DQB1 * 0301 allele and IL28B genotype are factors that are associated independently with SVR. There is a synergism between the HLA-DQB1 * 0301 and HLA-A * 0201 alleles with polymorphism rs12979860 CC, which increases the SVR rate. IL28B genotype is the best predictor of SVR.
SUPPLEMENTARY MATERIAL is linked to the online version of the paper at http://www.nature.com/ajg
Am J Gastroenterol 2011; 106:1246–1254; doi: 10.1038/ajg.2011.82; published online 14 June 2011
1 Unidad de Aparato Digestivo, Hospital Universitario San Cecilio , Granada , Spain ; 2 CIBEREHD (Instituto de Salud Carlos IIII) , Granada, Spain ; 3 Servicio de Inmunolog í a, Hospital Universitario Virgen de las Nieves , Granada , Spain ; 4 Unidad de Hepatolog í a, Hospital Ntra. Sra. de Valme , Sevilla , Spain ; 5 Unidad de Hepatology, Hospital Universitario de la Princesa , Madrid , Spain ; 6 Unidad de Hepatolog í a, Hospital General de Valencia , Valencia , Spain ; 7 Unidad de Pediatr í a, Hospital Universitario San Cecilio , Granada , Spain ; 8 These two authors contributed equally to this work . Correspondence: Paloma Mu ñ oz de Rueda, PhD, Unidad de Aparato Digestivo, Hospital Universitario San Cecilio , c / Dr Oloriz no16, 18012 Granada , Spain . E-mail: [email protected] Received 19 August 2010; accepted 11 February 2011
© 2011 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
1247
LIV
ER
HLA and IL28B Predict SVR in Patients With HCV
severity of hepatitis C: HLA-DQB1 * 0301 and HLA-DRB1 * 11, as
well as DRB1 * 11-DQB1 * 0301 haplotype ( 8 – 10 ). HLA associa-
tions with respect to IFN treatment response have been intensively
investigated across global populations. To date, a clear and repro-
ducible association has not been found between HLA alleles and
therapeutic outcome ( 11 – 13 ).
Very recently, another host genetic factor, a single-nucleotide
polymorphism on chromosome 19q13, rs12979860, has been
strongly associated with sustained virological response (SVR) to
treatment with pegIFN / RBV in a cohort of > 1,600 individuals
chronically infected with genotype 1 of the HCV ( 14 ). Th is vari-
ant is near the IL28B locus, which encodes type III IFN- λ 3. Th ese
authors state that the CC genotype in rs12979860 doubles the
chance of successful HCV treatment. Another study ( 15 ) found
that the CC genotype strongly enhances resolution of HCV infec-
tion among individuals of European or African ancestry. Accord-
ing to these results, rs12979860 polymorphism is the fi rst genetic
marker that predicts response to treatment for HCV genotype 1
in a genome-wide association study. IFN- λ 3, which belongs to
the type III family of IFNs, plays a role in innate immunity ( 16 ).
Type III IFNs are related to IL-10 and are similar to type I IFNs
with respect to their antiviral action ( 17,18 ).
Th e aim of this study was to explore the infl uence of the
host genetic factors (HLA class I A, B, and Cw loci and
HLA class II DRB1 and DQB1 loci and rs12979860 variations)
as predictive factors of SVR to pegIFN- α / RBV therapy for CHC
infection and to compare them with viral factors. A large cohort
of 428 Spanish Caucasian individuals was categorized through
molecular genotyping techniques for HLA and IL28B.
METHODS Patients A total of 428 previously untreated CHC patients were enrolled
in this multicenter retrospective cohort study. Th is cohort
included 378 patients (88 % ) with genotype 1 or 4 and 50 (12 % )
with genotype 2 or 3. Only 271 subjects had biopsy-proven
CHC. Th e histological study was carried out in accordance with
Scheuer ’ s grading of necroinfl ammatory activity (I0 – I8) and
fi brosis staging (F0 – F4), with modifi cations ( 19 ). Th e diagnosis
of CHC was based on the permanent detection of HCV-RNA
serum. Th e patients showed no evidence of hepatitis B virus,
HIV, alcoholism or autoimmune or drug-induced liver disease.
In all, 167 patients were treated at San Cecilio University Hospital
(Granada), 120 at Nuestra Se ñ ora de Valme Hospital (Sevilla), 101
at La Princesa Hospital (Madrid), and 40 at University General
Hospital (Valencia), all in Spain. All of these patients were treated
between 2001 and 2004.
Treatment and response to therapy Th e treatment consisted of 180 μ g / week pegIFN- α -2a (Pegasys,
Roche Diagnostics, PEGASYS ® , Roche, Basel, Switzerland) in combi-
nation with RBV dose-adjusted to body weight (800 – 1,200 mg / day;
Copegus, Roche Diagnostic) for 48 weeks. SVR was defi ned as
undetectable serum HCV-RNA 24 weeks aft er stopping treatment;
the patients whose HCV-RNA levels were detectable aft er the
completed period of treatment, together with relapsed responder
patients (undetectable HCV-RNA during treatment but detect-
able aft er discontinuation), were considered to have a non-SVR.
Subjects with insuffi cient viral response at 12 or 24 weeks discon-
tinued therapy per protocol as treatment failures and were consid-
ered to have a non-SVR.
Ethical considerations All subjects provided informed consent to participate and for
the collection and storage of serum and peripheral blood for
DNA extraction. Th e study protocol conformed to the ethical
guidelines of the 1975 Declaration of Helsinki. Prior approval for
the study was obtained from the applicable ethics committees.
Virological assays HCV genotyping was determined by reverse hybridization
(Inno-LIPA II HCV Innogenetics SA, Ghent, Belgium). Quanti-
tative measurement of viral load (cutoff < 15 IU / ml, HCV
Ampliprep TaqMan, Roche Molecular Systems, Pleasanton, CA)
was carried out at baseline and at weeks 12, 24, 48 (end of treat-
ment), and 72 (end of follow-up) with samples that were stored
at − 80 ° C.
HLA class I and HLA class II genotyping Class I HLA-A, B, and Cw loci and class II HLA-DRB1 and DQB1
loci genotyping was performed in 428 subjects. Geno typing was car-
ried out by SSO LABType (One Lambda, Canoga Park, CA). Target
DNA was amplifi ed by PCR using sequence-specifi c primers, fol-
lowed by hybridization with allele-specifi c oligo deoxynucleotides
coupled with fl uorescent phycoerythrin-labeled microspheres. Th e
fl uorescence intensity was determined on a Bio-Plex 200 system
(Luminex xMAP, Austin, TX). HLA allele assignment was per-
formed with HLA-Tools soft ware (Los Angeles, CA).
IL28B genotyping Rs12979860 genotyping was performed in 428 subjects using a
Taqman 5 ′ allelic discrimination assay (Custom Assay Service;
Ref AHI05OJ). Th e primers used were forward GCCTGTCGTGT
ACTGAACCA and backward GCGCGGAGTGCAATTCAAC.
Th e Taqman probes for the reverse strand were TGGTTC G CGC
CTTC labeled with VIC and CTGGTTC A CGCCTTC labeled
with FAM. Single-nucleotide polymorphism amplifi cation assays
were used according to the manufacturer ’ s instructions. PCR reac-
tions were carried out in a total volume of 4 μ l with the following
amplifi cation protocol: pre-incubation at 50 ° C for 2 min and at
95 ° C for 10 min, followed by 40 cycles of 95 ° C, 15 s; 60 ° C, 1 min.
Th e PCR results and genotype of each sample were automatically
attributed by the SDS 2.2.1 soft ware for allelic discrimination
(Applied Biosystems, Foster City, CA).
Statistical analysis Th e dependent variable was the viral response (non-SVR vs. SVR).
Th e degree of association between the response and independent
variables was determined by calculating the crude odds ratio (cOR)
The American Journal of GASTROENTEROLOGY VOLUME 106 | JULY 2011 www.amjgastro.com
1248 LI
VE
R
de Rueda et al.
and its 95 % confi dence interval (CI). In the analysis of frequency of
the HLA class I and class II alleles according to the response, signi-
fi cant P values were corrected (Pc; Bonferroni ’ s correction) for the
number of alleles detected at each locus. When the dependent vari-
able was carrier vs. noncarrier of HLA-DQB1 * 0301 or rs12979860
CC vs. rs12979860 CT / TT genotype, the baseline variables were
compared between groups using the χ 2 test. To determine the
independent eff ect of each factor, the OR was adjusted (aOR) using
a binomial logistic regression model. Th e covariates included in
the multivariable model were age, sex, alanine aminotransferase,
γ -glutamyltransferase, viral genotype, viral load, HLA-DQB1 * 0301,
and IL28B. A histological study was not included because we only
had 271 biopsy-proven cases of CHC. Our model shows that:
P =+ − − + + + + +
1
1 1 04 1 17 0 88 0 8 0 73 1 47e xG xA xVL xDQ xIL( . . . . . . )
where P is the probability of achieving SVR, G is the viral geno-
type (genotype 1 or 4 = 0 and genotype 2 or 3 = 1), A is the patient ’ s
age ( > 40 years old = 0 and ≤ 40 years old = 1), VL is the viral
load ( > 600,000 IU / ml = 0 and ≤ 600,000 IU / ml = 1), DQ is HLA-
DQB1 * 0301 (HLA-DQB1 * 0301( − ) = 0 and HLA-DQB1 * 0301( + ) = 1
and IL is rs12979860 (rs12979860 CT / TT = 0 and rs12979860
CC = 1). Th e criterion for statistical signifi cance was P ≤ 0.05.
By means of diff erent models of logistic regression and receiver
operating characteristic (ROC) analysis, we studied the discrimina-
tory power of the diff erent predictive response factors to distinguish
individuals with SVR from those with non-SVR. Th us, the area under
the curve (AUC) ROC was independently obtained for the variables
age, viral load, viral genotype, HLA-DQB1 * 0301 (positive or nega-
tive), and rs12979860; in addition, four logistic models were studied:
model 1 used age, viral genotype, pre-treatment viral load, carrier of
HLA-DQB1 * 0301, and rs12979860 genotype; model 2 used all cov-
ariates except rs12979860 genotype; model 3 used genetic host fac-
tors (age + HLA-DQB1 * 0301 + rs12979860 genotype), and model 4
used viral factors (age + viral load + viral genotype). When the AUC
was ≥ 0.7, the model was considered to have predicted the response.
All statistical calculations were performed using SPSS soft ware
version 15.0 for Windows (SPSS Inc., Chicago, IL).
RESULTS Baseline clinical, virological, and histological characteristics and predictive factors of virological response Th e main clinical, virological, and histological features of all
428 patients are summarized in Table 1 . In bivariate analysis, the
factors associated with SVR were age ≤ 40 years old, viral geno-
type 2 or 3, γ -glutamyltransferase ≤ 37 U / l, and baseline viral load
≤ 600,000 IU / ml.
Study of HLA class I and class II and SVR association In the class I region, there were 35, 58, and 33 alleles at the
HLA-A, B and Cw loci, respectively. In the class II region, there
were 35 alleles at the HLA-DRB1 locus and 16 alleles at the
HLA-DQB1 locus. HLA-A * 0201, B * 4001, B * 4402, Cw * 0304,
DRB1 * 1101, and DQB * 0301 were associated with SVR, and
HLA-A * 0301, DRB1 * 0701, and DQB1 * 0202 were associated with
non-SVR, but when the P value was corrected (Bonferroni ’ s cor-
rection), only the P value for HLA-DQB1 * 0301 was statistically
signifi cant ( Table 2 ). In all, 100 of 428 patients had the HLA-
DQB1 * 0301 allele: 31 (31 % ) had a non-SVR and 69 (69 % ) had a
SVR; but in the HLA-DQB1 * 0301-negative group ( n = 328), 156
(48 % ) had a non-SVR and 172 (52 % ) had a SVR ( Table 2 ). Th e
characteristics of the HLA-DQB1 * 0301-positive and -negative
patients are shown in Supplementary Table I online, and those
of the HLA-DQB1 * 0301-positive patients ( n = 100) with SVR and
non-SVR are shown in Supplementary Table II .
Th e haplotype DRB1 * 1101-DQB1 * 0301 was studied on the
basis of the linkage disequilibrium between the HLA-DQB1 * 0301
and DRB1 * 1101 alleles. In all, 45 patients had the DRB1 * 1101-
DQB1 * 0301 haplotype, of whom 35 (78 % ) had SVR, whereas of
the 383 without this haplotype, 222 (58 % ) had SVR ( P = 0.03).
When the Bonferroni ’ s correction was applied, this result was no
longer statistically signifi cant.
When the patients were stratifi ed according to their viral geno-
type, the SVR rates in the HLA-DQB1 * 0301-positive and -negative
patients infected by viral genotype 1 or 4 were 68 % (60 / 88) and
48 % (140 / 290), respectively ( P = 0.001; Pc = 0.016); whereas in the
HLA-DQB1 * 0301-positive and -negative patients infected with
viral genotype 2 or 3 the corresponding values were 75 % (9 / 12)
and 84 % (32 / 38), respectively ( P = 0.4) (Mantel – Haenszel test,
P = 0.005). For an interaction term between viral genotype and
DQB1 * 0301 in logistic regression, P = 0.041.
Study of IL28B and association with SVR Rs12979860 genotypes were unequivocally assigned in all cases
except in fi ve individuals ( n = 423). Th e rs12979860 CC geno-
type frequency was 33 % ( n = 142), CT was 51 % ( n = 220), and TT
was 14 % ( n = 61). We found an association of the CC genotype
with SVR ( Table 2 ). When the patients were stratifi ed according
to viral genotype, the frequency of the CC genotype was higher
in viral genotype 2- or 3-infected patients (25 / 50; 50 % ) than in
viral genotype 1- or 4-infected patients (117 / 373; 31 % ; P = 0.009).
Th ere was a signifi cant association between genetic variation in
IL28B and SVR among individuals infected with viral genotype
1 or 4 ( P < 0.0001), but among those infected with viral genotype
2 or 3 no such association was found ( P = 0.7) (Mantel – Haenszel
test, P = 0.0001). Th e charac teristics of the rs12979860 CC and
CT / TT patients are shown in Supplementary Table III . Th e
rs12979860 CC genotype patients ( n = 142) with SVR vs. non-SVR
diff ered only with respect to viral load and baseline alanine amino-
transferase ( Table 3 ). However, the patients with CT / TT geno-
type and SVR were younger, had a lower γ -glutamyltransferase,
a lower viral load and a higher proportion of viral genotype 2 or 3
( Table 3 ). Accordingly, the patients with CT / TT genotype achieved
SVR only if they also presented other predictive factors.
Associations among HLA alleles, rs12979860 genotype, and SVR Th e associations between the HLA alleles found in our population
(126 HLA class I and 51 HLA class II alleles) and the rs12979860
© 2011 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
1249
LIV
ER
HLA and IL28B Predict SVR in Patients With HCV
genotype were evaluated in SVR and non-SVR patients. Th e
HLA-DQB1 * 0301 and A * 0201 alleles ( Figure 1 ) were associ-
ated with the CC genotype and SVR. To establish whether any of
these alleles had a synergic eff ect with rs12979860 CC on SVR,
a simple logistic regression analysis with a polytomic variable
was performed for four groups of patients ( Figure 1 ). Th e
reference group was formed by patients without either of the two
alleles (HLA-DQB1 * 0301-negative or HLA-A * 0201-negative,
and rs12979860 CT / TT genotype). Patients who were HLA-DQB1 *
0301-positive and had the rs12979860 CC genotype ( n = 29) had
a 5.8 times greater probability of SVR than did those with neither
of the two. Note that HLA-DQB1 * 0301-positive and rs12979860
CT / TT genotype patients had a 2.9 times greater probability of
SVR than did those with neither of the two ( P < 0.0001). In all, 62
patients were HLA-A * 0201-positive and had the rs12979869 CC
genotype; 89 % of these achieved SVR, as opposed to 48 % of those
who had neither of these alleles ( P < 0.0001). Moreover, the HLA-
A * 0201-positive and rs12979869 CT / TT genotype patients did
not diff er signifi cantly from the patients without either of these
alleles ( Figure 1 ).
Table 1 . Baseline clinical and virological characteristics and histological features and predictive factors of sustained virological response
All patients, n =428 (100) Non-SVR, n =187 (44) SVR, n =241 (56) cOR a (95 % CI, P value)
Age (years)
≤ 40 164 (38) 51 (31) 113 (69) 2.35 (1.56 – 3.55, < 0.0001)
> 40 264 (62) 136 (51) 128 (49) 1
Sex
Female 180 (42) 85 (47) 95 (53) 1.28 (0.87 – 1.89, 0.21)
Male 248 (58) 102 (41) 146 (59) 1
Viral genotype
1 50 (12) 9 (18) 41 (82) 4.05 (1.92 – 8.58, < 0.0001)
1 or 4 378 (88) 178 (47) 200 (53) 1
Fibrosis stage ( n =271)
F0, F1 or F2 209 (77) 82 (39) 127 (61) 1.55 (0.88 – 2.74, 0.13)
F3 – F4 62 (23) 31 (50) 31 (50) 1
Necroinfl ammatory grade ( n =263 )
I0 – I4 164 (62) 65 (40) 99 (60) 1.27 (0.77 – 2.10, 0.35)
I5 – I8 99 (38) 45 (45) 54 (55) 1
ALT, U / l
≤ 40 44 (10) 20 (46) 24 (55) 0.91 (0.49 – 1.70, 0.76)
> 40 384 (90) 165 (43) 219 (57) 1
AST, U / l
≤ 40 133 (31) 55 (41) 78 (59) 1.14 (0.75 – 1.73, 0.53)
> 40 295 (69) 133 (45) 162 (55) 1
GGT, U / l
≤ 37 186 (44) 71 (38) 115 (62) 1.73 (1.16 – 2.58, 0.007)
> 37 242 (56) 126 (52) 116 (48) 1
Baseline viral load, IU / ml
≤ 600,000 141 (33) 41 (29) 100 (71) 2.35 (1.52 – 3.62, < 0.0001 )
> 600,000 287 (67) 146 (51) 141 (49) 1
ALT, alanine aminotransferase; AST, aspartate aminotransferase; CI, confi dence interval; cOR, crude odds ratio; GGT, γ glutamyltransferase; SVR, sustained virological response. Values are absolute with percentages in parentheses. a Logistic regression: reference groups, > 40 years old, male, viral genotype 1 or 4, fi brosis staging F3 – F4, necroinfl ammatory grade I5 – I8, ALT > 40 U / l, AST > 40 U / l, GGT > 37 U / l, viral load > 600,000 IU / ml.
The American Journal of GASTROENTEROLOGY VOLUME 106 | JULY 2011 www.amjgastro.com
1250 LI
VE
R
de Rueda et al.
DISCUSSION A large number of studies have identifi ed viral- and patient-
related factors for pre-treatment prediction of the probability
of SVR ( 5 ). To date, the most important of these factors have
been the viral ones. Nevertheless, genetic host factors are seen
as becoming more important, with the development of high-
throughput technologies such as genome-wide association study.
Our study examines the eff ects of genetic diversity in the human
major histocompatibility complex (HLA-A, B, and C class I loci
and DRB1 and DQB1 class II loci) and the rs12979860 genotype
upstream of IL-28B on the outcome of pegIFN / RBV treatment.
Th e data used were obtained from a large ( n = 428) well-charac-
terized cohort of Spanish patients. We present evidence that there
is a signifi cant association of one class II allele HLA-DQB1 * 0301
and the rs12979860 CC genotype with SVR, the latter being
a better predictive factor of SVR than are viral factors.
Th e multivariate logistic regression model shows that age ≤ 40
years old, viral genotype 2 or 3, baseline viral load ≤ 600,000 IU / ml,
carrier of HLA-DQB1 * 0301 and rs12979860 CC genotype were
the main independent factors associated with SVR to pegIFN / RBV
therapy. All these variables, except HLA-DQB1 * 0301, were previ-
ously associated independently with SVR ( 14,20 – 23 ). As pegIFN
has direct antiviral eff ects and RBV promotes a type 1 cytokine-
mediated immune response that can enhance antiviral immune
responses, there might be an association of immunogenetic
characteristics with response to antiviral therapy. Accordingly, the
genetic variations at HLA loci with respect to antigen presentation
might be related to the response to pegIFN-based therapy.
In our review of relevant publications, we identifi ed only three
reports that examined whether specifi c HLA alleles were associ-
ated with SVR to pegIFN / RBV therapy for CHC infection ( 11 – 13 ).
Th e A * 02, B * 58, and DPB1 * 1701 HLA alleles were independently
associated with SVR by Rhodes et al. ( 11 ). Th e number of patients
was higher ( n = 343) ( 11 ), but they were separated into two races,
Caucasian Americans and African Americans, which diminished
the number of patients per arm, and allele frequencies were very
low in certain subgroups. However, what is most important is that
the P values were not corrected for the number of alleles found
in each locus ( 11 ). In fact, in our study, the HLA-A * 0201 allele
was statistically associated with SVR, but according to Bonferroni ’ s
correction this value was not signifi cant.
We found that the HLA class II allele DQB1 * 0301 is associated
with SVR, aft er adjusting and controlling for the most important
virological confounding factors. Previous studies have shown
that patients with higher levels of CD4 + T-cell responses are
more likely to develop SVR ( 24,25 ). It is assumed that the HLA-
DQB1 * 0301 allele may more eff ectively present HCV immuno-
dominant epitopes to CD4 + Th cells than do other alleles; indeed,
the immune-modulating epitope of HCV for the HLA-DQB1 * 0301
allele has been described ( http://www.hcv.lanl.gov ) ( 26 ), and this
allele was strongly associated with SVR in the current study. We
examined whether the presence of the DRB1 * 1101-DQB1 * 0301
haplotype might infl uence SVR, and found signifi cant diff erences;
however, following Bonferroni ’ s correction, this signifi cance disap-
peared. Th is fi nding corroborates the role of the HLA-DQB1 * 0301
Multivariate analysis model Figure 2 presents the aOR estimates for a logistic regression
model (model 1) including factors that were independently asso-
ciated with SVR. Table 4 shows the interindividual diff erences in
SVR obtained by the diff erent predictors in our model (Methods
section). Th e rs12979860 CC genotype was associated with a
more substantial diff erence in the rate of SVR than other baseline
predictors included in the model.
When the multivariate analysis was performed on the data for
267 patients for whom we were aware of the fi brosis stage, it was
found that fi brosis stage was not associated with SVR independ-
ently of the other factors (F0 – F2, aOR = 1.5, 95 % CI: 0.78 – 2.3,
P = 0.2).
Using ROC curve analysis, we sought to identify pre-treatment
variables that were predictive of SVR, and evaluated the power of
diff erent multivariate logistic models with diff erent predictors in
order to distinguish between SVR and non-SVR based on AUC
ROC ( Figure 3 ). Th e AUC for rs12979860 was greater than that
for HLA-DQB1 * 0301 and other covariates ( Figure 3a ). Model 1
(sensitivity and specifi city were 73 and 67 % , respectively) had
the largest AUC, as expected, because it included all the predic-
tive covariates ( Figure 3b ). When we extracted the variable
rs12979860 (model 2), the sensitivity and specifi city were 72 and
50 % , respectively, and the AUC was lower than in model 1 with
signifi cant diff erences ( P = 0.03). Comparison of the host genetic
factors model (model 3) and the viral factors model (model 4)
revealed statistically signifi cant diff erences (higher AUC in
model 3, P = 0.04) ( Figure 3b ).
Table 2 . HLA class I and II alleles with P value ≤ 0.05 before Bonferroni’s adjustment and rs12979860 genotype association with SVR
cOR a (95 % CI) P value Pc
HLA class I and II alleles ( n =428)
A*0201 ( + ) 1.738 (1.10 – 2.74) 0.02 0.7
A*0301 ( + ) 0.533 (0.31 – 0.91) 0.02 0.7
B*4001 ( + ) 5.345 (1.20 – 23.78) 0.02 1.1
B*4402 ( + ) 2.472 (1.03 – 5.918) 0.03 1.7
Cw*0304 ( + ) 4.315 (1.25 – 14.96) 0.02 0.6
DRB1*0701 ( + ) 0.529 (0.33 – 0.84) 0.007 0.2
DRB1*1101 ( + ) 2.2 (1.02 – 4.9) 0.04 1.4
DQB1*0202 ( + ) 0.589 (0.37 – 0.94) 0.03 0.48
DQB1*0301 ( + ) 2.02 (1.25 – 3.25) 0.003 0.04
IL28B ( n =423)
rs12979860 CC 3.84 (2.44 – 6.05) < 0.0001
CI, confi dence interval; cOR, crude odds ratio; Pc, Bonferroni’s correction; SVR, sustained virological response. a Logistic regression: the cOR is the probability of response of the patients who are carriers of the allele, with regard to non-carriers (reference groups).
© 2011 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
1251
LIV
ER
HLA and IL28B Predict SVR in Patients With HCV
allele, which acts independently of HLA-DRB1 * 11 and could
also be associated with other alleles of HLA-DRB1 * such as 0401,
0407, 1303, and 1402. HLA-DRB1 * 1101 and DQB1 * 0301 have
been reproducibly shown to be involved in the spontaneous res-
olution of HCV infection in European populations ( 8 ). Diverse
studies have identifi ed signifi cant genetic associations with diff er-
ent HLA-DQB1 * 0301-bearing haplotypes, suggesting that HLA-
DQB1 * 0301 might be one of the most prominent factors in HCV
clearance ( 8 ).
At the end of 2009, three independent genome-wide association
studies reported that single-nucleotide polymorphisms near the
IL28B (IFN- λ 3) region were associated with treatment response
( 14,27,28 ). We found that the patients with the rs12979860 CC
genotype had a four times greater probability of SVR than did
those with the CT or TT genotype. But aft er stratifying for geno-
type, there was no signifi cant association between rs12979860 and
SVR among individuals infected with HCV genotype 1, indicating
that the prognostic value of the risk allele for treatment response
might be limited to individuals with diffi cult-to-treat HCV geno-
types. Th e same results were obtained by McCarthy et al. ( 29 )
and Rauch et al. ( 30 ), but in this case with genomic variants in
the rs8099917 locus.
Table 3 . Baseline characteristics of patients who are rs12979860 CC and rs12979860 CT / TT and relationship between non-SVR and SVR
rs12979860 CC ( n =142) rs12979860 CT / TT ( n =281)
Non-SVR, n =33
(23 % ) SVR, n =109
(77 % ) P a Non-SVR, n =151
(54 % ) SVR, n =130
(46 % ) P a
Age (years) b
≤ 40 9 (16) 46 (84) 0.1 40 (38) 66 (62) < 0.0001
> 40 24 (28) 63 (72) 111 (63) 64 (37)
Sex
Female 15 (28) 38 (72) 0.3 68 (55) 56 (45) 0.8
Male 18 (20) 71 (80) 83 (23) 74 (47)
Fibrosis stage ( n =267)
1 – 2 11(18) 50 (82) 0.2 69 (48) 75 (52) 0.2
3 – 4 7 (34) 15 (68) 24 (60) 16 (40)
Necroinfl ammatory grade ( n =259)
1 – 4 7 (16) 36 (84) 0.06 57 (48) 62 (52) 0.5
4 – 6 14 (33) 28 (67) 30 (54) 25 (45)
ALT (U / l)
≤ 40 8 (50) 8 (50) 0.02 12 (41) 17 (59) 0.2
> 40 26 (21) 100 (79) 137 (54) 115 (46)
AST (U / l)
≤ 40 8 (22) 29 (78) 0.8 46 (49) 48 (51) 0.3
> 40 26 (24) 79 (76) 105 (56) 82 (44)
GGT (U / l) b
≤ 37 21 (27) 57 (73) 0.8 54 (45) 67 (55) 0.006
> 37 15 (24) 49 (76) 99 (62) 61 (38)
Viral genotype b
2 or 3 4 (16) 21 (84) 0.4 5 (20) 20 (80) < 0.0001
1 or 4 29 (25) 88 (75) 146 (57) 110 (43)
Baseline viral load b (IU / ml)
≤ 600,000 5 (11) 41 (89) 0.01 34 (37) 59 (63) < 0.0001
> 600,000 28 (29) 68 (71) 117 (62) 71 (38)
ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, γ -glutamyltransferase; SVR, sustained virological response. Values are absolute with percentages in parentheses. a Bivariate analysis: P value calculated by χ 2 test or Fisher’s test. b Mantel – Haenszel test: P ≤ 0.05.
The American Journal of GASTROENTEROLOGY VOLUME 106 | JULY 2011 www.amjgastro.com
1252 LI
VE
R
de Rueda et al.
with the rs12979860 CC genotype are related to some of the
routes involved in the recognition of diff erent HLA specifi cities.
Th us, an IL28B genotype with higher levels of rs12979860 during
treatment may induce the maturation and activation of antigen-
presenting cells. Th ese antigen-presenting cells will present an
immunodominant peptide via HLA-DQB1 * 0301 to the CD4 +
Th 0 cells, which diff erentiates to CD4 + Th 1 anti-HCV. Further-
more, IL28B can induce HLA class I expression on hepatocytes
infected with HCV, and individuals with HLA-A * 0201 will more
effi ciently control HCV infection because they present a viral
peptide with a low rate of mutations that prevents escape from
anti-HCV cytotoxic CD8 + lymphocytes.
Th ese new genetic predictive factors will have to compete
with other viral predictors of response such as viral load or viral
Because of the growing importance of genetic host factors in
treatment outcome, the associations between the rs12979860
genotype and all the HLA alleles found in our population were
evaluated in SVR and non-SVR patients. It was found that only
two alleles were associated with the CC genotype and SVR:
HLA-DQB1 * 0301 and A * 0201. Th e HLA-DQB1 * 0301 allele
is independently associated with SVR, whereas patients with
HLA-A * 0201 require the presence of the CC polymorphism in
rs12979860 to increase the SVR rate. Although the genotypic rela-
tion between the HLA class II molecules and the IFN- λ ones is
well known and has recently been described ( 31 ), the phenotypic
relationship between the genotypes DQB1 * 0301 and rs12979860
CC with SVR are described in the present paper. We suggest that
at least some of the benefi cial eff ects of the response in patients
DQB1*0301 (+)rs12979860 CC
vs. DQB1*0301 (–)rs12979860 CT/TT
DQB1*0301 (–)rs12979860 CC
vs. DQB1*0301 (–)rs12979860 CT/TT
DQB1*0301 (+)rs12979860 CT/TT
vs. DQB1*0301 (–)rs12979860 CT/TT
A*0201 (+)rs12979860 CC
vs. A*0201 (–)rs12979860 CT/TT
A*0201 (–)rs12979860 CC
vs. A*0201 (–)rs12979860 CT/TT
A*0201 (+)rs12979860 CT/TT
vs. A*0201 (–)rs12979860 CT/TT
OR0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 2423
Figure 1 . Associations between rs12979860 polymorphism and HLA alleles evaluated in sustained viral response (SVR) and non-SVR patients. DQB1 * 0301( + ) and rs12979860 CC patients had a crude odds ratio (cOR) of 5.8 (95 % confi dence interval (CI): 2.3 – 14.8, P < 0.0001); DQB1 * 0301( − ) and rs12979860 CC patients had a cOR of 4.8 (95 % CI: 2.8 – 8, P < 0.0001); DQB1 * 0301( + ) and rs12979860 CT / TT patients had a cOR of 2.9 (95 % CI: 1.6 – 5.1, P < 0.0001); A * 0201( + ) and rs12979860 CC patients had a cOR of 8.6 (95 % CI: 3.2 – 23.4, P < 0.0001); A * 0201( − ) and rs12979860 CC patients had a cOR of 2.3 (95 % CI: 1.2 – 4.3, P = 0.013); A * 0201( + ) and rs12979860 CT / TT patients had a cOR of 1.5 (95 % CI: 0.8 – 2.5, P = 0.16). Each group of patients was compared with patients who were DQB1 * 0301( − ) or A * 0201( − ) and who had the rs12979860 CT / TT genotype (reference groups). cOR is the probability of SVR for each group of patients with respect to the reference group (patients with neither of the two alleles).
Genotype 2/3 vs. 1/4
Viral load ≤600,000 vs. > 600,000 IU/ml
Age ≤40 vs . >40 years old
DQB1*0301 (+) vs. DQB1*0301 (–)
rs12979860 CC vs. rs12979860 CT/TT
0 1 2 3 4 5 6 7 8
aOR
Figure 2 . Multivariate logistic regression: predictors of sustained viral response (SVR) to pegylated interferon / ribavirin therapy. Covariates of model: sex, age, alanine aminotransferase, γ -glutamyltransferase, viral genotype, viral load, carrier of the HLA-DQB1 * 0301 allele, and rs12979860 polymorphism. Final model: viral genotype (adjusted odds ratio (aOR) = 3.2, 95 % confi dence interval (CI): 1.4 – 7.2), age (aOR = 2.4, 95 % CI: 1.5 – 3.8), viral load (aOR = 2.2, 95 % CI: 1.4 – 3.6), carrier of DQB1 * 0301 (aOR = 2.08, 95 % CI: 1.2 – 3.5), and rs12979860 polymorphism (aOR = 4.3, 95 % CI: 2.6 – 7). aOR is the probability of SVR for each predictor with respect to the reference group to controlling for other variables. Hosmer and Lemeshow test: P = 0.8.
© 2011 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY
1253
LIV
ER
HLA and IL28B Predict SVR in Patients With HCV
genotype. Using ROC curve analysis, the most powerful predictive
pre-treatment variable of SVR was found to be rs12979860.
In conclusion, the HLA-DQB1 * 0301 allele and the rs12979860
genotype are signifi cant independent predictors of response to
pegIFN / RBV treatment in patients with chronic HCV infection.
Nevertheless, there is synergism between the HLA-DQB1 * 0301
and HLA-A * 0201 alleles with genotype rs12979860 CC, which
increases SVR valuation. Th ese results confi rm that studying
patients ’ immunogenetic factors can help in planning new, more
eff ective therapeutic strategies. Finally, the rs12979860 genotype
is the best predictor of response in patients with CHC treated
with pegIFN / RBV.
CONFLICT OF INTEREST Guarantor of the article : Javier Salmeron, MD.
Specifi c author contributions : Concept and study design:
Paloma Mu ñ oz de Rueda, Miguel Angel L ó pez-Nevot, and Javier
Salmeron; statistical analysis and interpretation; wrote and edited
the manuscript: Paloma Mu ñ oz de Rueda; consulted on manuscript
preparation: Miguel Angel L ó pez-Nevot and Javier Salmeron;
monitored the patients: Angeles Ruiz-Extremera, Ana Gila, and
Pablo Palomares; contributed patients to the study and monitored
these patients: Manuel Romero-G ó mez, Paloma Sanz-Cameno,
Ricardo Moreno-Otero, and Mois é s Diago; carried out laboratory
work: Jorge Casado, Antonia Mart í n-Casares, Rosa Quiles,
Esther-Jos é Pav ó n, and Jos é -Antonio Mu ñ oz; data acquisition
and creating the database: Á ngel Carazo and Josefa Le ó n. All
the authors reviewed the manuscript and approved the fi nal
version.
Financial support : Th is work was partially supported by the
Instituto de Salud Carlos III (Fondo de Investigaciones Sanitarias)
grant PI06 / 0939, Programa de estabilizaci ó n I3SNS, Conserjer í a
de Salud de la Junta de Andaluc í a PI387 / 05 and Ciberehd.
Potential competing interests: None.
Table 4 . Assessing probability of achieving SVR in patients with chronic hepatitis C
Age Viral load
Viral genotype DQ0301B1*
rs12979860 polymorphism
Probability of SVR ( % )
1 1 1 1 1 98
1 0 1 1 1 96
1 1 1 0 1 96
1 1 0 1 1 94
1 1 1 1 0 92
0 0 0 1 1 76
0 1 1 0 0 71
0 0 0 0 1 60
0 0 1 0 0 53
1 0 0 0 0 46
0 1 0 0 0 44
0 0 0 1 0 42
SVR, sustained virological response. Viral genotype 1 or 4=0, viral genotype 2 or 3=1; age > 40 years old=0, ≤ 40 years old=1; viral load > 600,000 IU / ml=0, ≤ 600,000 IU / ml=1; DQB1*0301( − )=0, DQB1*0301( + )=1; rs12979860 CT / TT=0, rs12979860 CC=1.
1.0
a b
0.8
IL28B
Mode 1: allcovariates
Model 2: withoutre12979860
Model 3:age+DQB1*0301+rs12979860
Model 4: age+viralload+viral genetic
Reference line
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.2 0.4
1-Specificity0.6 0.8 1.0
Viral genotype
Age
Viral load
DQB1*0301
Reference line
0.6
Sen
sitiv
ity
Sen
sitiv
ity
0.4
0.2
0.0
0.0 0.2 0.4 0.6
1-Specificity0.8 1.0
Figure 3 . Receiver operating characteristic (ROC) curve analysis of predictive factors of sustained viral response to pegylated interferon / ribavirin treatment in patients with chronic hepatitis C. ( a ) ROC curves for the variables independently: age (area under the curve (AUC) = 0.59), viral load (AUC = 0.59), viral genotype (AUC = 0.56), DQB1 * 0301 (AUC = 0.56), and rs12979860 polymorphism (AUC = 0.63). ( b ) ROC curves for four logistic models: model 1, age + viral load + viral genotype + carrier of DQB1 * 0301 + rs12979860 polymorphism (AUC = 0.76); model 2: age + viral load + viral genotype + carrier of DQB1 * 0301 (AUC = 0.69); model 3: age + carrier of DQB1 * 0301 + rs12979860 genotype (AUC = 0.72); model 4: age + viral genotype + viral load (AUC = 0.67). AUC model 1 vs. AUC model 2, P = 0.03; AUC model 3 vs. AUC model 4, P = 0.04.
The American Journal of GASTROENTEROLOGY VOLUME 106 | JULY 2011 www.amjgastro.com
1254 LI
VE
R
de Rueda et al.
7 . Folgori A , Spada E , Pezzanera M et al. Early impairment of hepatitis C virus specifi c T cell proliferation during acute infection leads to failure of viral clearance . Gut 2006 ; 55 : 1012 – 9 .
8 . Hong X , Yu RB , Sun NX et al. Human leukocyte antigen class II DQB1*0301, DRB1*1101 alleles and spontaneous clearance of hepatitis C virus infection: a meta-analysis . World J Gastroenterol 2005 ; 11 : 7302 – 7 .
9 . Singh R , Kaul R , Kaul A et al. A comparative review of HLA associations with hepatitis B and C viral infections across global populations . World J Gastroenterol 2007 ; 13 : 1770 – 87 .
10 . Yee LJ . Host genetic determinants in hepatitis C virus infection . Genes Immun 2004 ; 5 : 237 – 45 .
11 . Rhodes SL , Erlich H , Im KA et al. Associations between the human MHC and sustained virologic response in the treatment of chronic hepatitis C virus infection . Genes Immun 2008 ; 9 : 328 – 33 .
12 . Yee LJ , Im K , Wahed AS et al. Polymorphism in the human major histocompatibility complex and early viral decline during treatment of chronic hepatitis C . Antimicrob Agents Chemother 2009 ; 53 : 615 – 21 .
13 . Dai CY , Chuang WL , Hsieh MY et al. Human leukocyte antigen alleles and the response to pegylated interferon/ribavirin therapy in chronic hepatitis C patients . Antiviral Res 2010 ; 85 : 396 – 402 .
14 . Ge D , Fellay J , Th ompson AJ et al. Genetic variation in IL28B predicts hepatitis C treatment-induced viral clearance . Nature 2009 ; 46 : 399 – 401 .
15 . Th omas DL , Th io CL , Martin MP et al. Genetic variation in IL28B and spontaneous clearance of hepatitis C virus . Nature 2009 ; 461 : 798 – 801 .
16 . Sheppard P , Kindsvogel W , Xu W et al. IL-28, IL-29 and their class II cytokine receptor IL-28R . Nat Immunol 2003 ; 4 : 63 – 8 .
17 . Marcello T , Grakoui A , Barba-Spaeth G et al. Interferons α and λ inhibit hepatitis C virus replication with distinct signal transduction and gene regulation kinetics . Gastroenterology 2006 ; 131 : 1887 – 98 .
18 . Dellgren C , Gad HH , Hamming OJ et al. Human interferon-l3 is a potent member of the type III interferon family . Genes Immun 2009 ; 10 : 125 – 31 .
19 . Caballero T , P é rez-Milena A , Masseroli M et al. Liver fi brosis assess-ment with semiquantitative indexes and image analysis quantifi cation in sustained-responder and non-responder interferon-treated patients with chronic hepatitis C . J Hepatol 2001 ; 34 : 740 – 7 .
20 . Shiff man ML , Suter F , Bacon BR et al. Peginterferon alfa-2a and ribavirin for 16 or 24 weeks in HCV genotype 2 or 3 . N Engl J Med 2007 ; 357 : 124 – 34 .
21 . Berg T , von Wagner M , Nasser S et al. Extended treatment duration for hepatitis C virus type 1: comparing 48 versus 72 weeks of peginterferon-alfa-2a plus ribavirin . Gastroenterology 2006 ; 130 : 1086 – 97 .
22 . Kamal SM , El Tawil AA , Nakano T et al. Peginterferon {alpha}-2b and ribavirin therapy in chronic hepatitis C genotype 4: impact of treatment duration and viral kinetics on sustained virological response . Gut 2005 ; 54 : 858 – 66 .
23 . Jacobson IM , Brown RS Jr , Freilich B et al. Peginterferon alfa-2b and weight-based or fl atdose ribavirin in chronic hepatitis C patients: a rand-omized trial . Hepatology 2007 ; 46 : 971 – 81 .
24 . Rosen HR , Weston SJ , Im K et al. Selective decrease in hepatitis C virus-specifi c immunity among African Americans and outcome of antiviral therapy . Hepatology 2007 ; 46 : 350 – 8 .
25 . Golden-Mason L , Klarquist J , Wahed AS et al. PD-1 expression is increased on immunocytes in chronic HCV and predicts failure of response to antiviral therapy: race-dependent diff erences . J Immunol 2008 ; 180 : 3637 – 41 .
26 . Lamonaca V , Missale G , Urbani S et al. Conserved hepatitis C virus sequences are highly immunogenic for CD4(+) T cells: implications for vaccine development . Hepatology 1999 ; 30 : 1088 – 98 .
27 . Suppiah V , Moldovan M , Ahlenstiel G et al. IL28B is associated with response to chronic hepatitis C interferon-alpha and ribavirin therapy . Nat Genet 2009 ; 41 : 1100 – 4 .
28 . Tanaka Y , Nishida N , Sugiyama M et al. Genome-wide association of IL28B with response to pegylated interferon-alpha and ribavirin therapy for chronic hepatitis C . Nat Genet 2009 ; 41 : 1105 – 9 .
29 . McCarthy JJ , Li JH , Th ompson A et al. Replicated association between an IL28B gene variant and a sustained response to pegylated interferon and ribavirin . Gastroenterology 2010 ; 138 : 2307 – 14 .
30 . Rauch A , Kutalik Z , Descombes P et al. Swiss Hepatitis C Cohort Study; Swiss HIV Cohort Study. Genetic variation in IL28B is associated with chronic hepatitis C and treatment failure: a genome-wide association study . Gastroenterology 2010 ; 138 : 1338 – 45 .
31 . Pietil ä TE , Latvala S , Osterlund P et al. Inhibition of dynamin-dependent endocytosis interferes with type III IFN expression in bacteria-infected human monocyte-derived DCs . J Leukoc Biol 2010 ; 88 : 665 – 74 .
Study Highlights
WHAT IS CURRENT KNOWLEDGE 3 Several viral factors, such as genotype, baseline viral load,
rapid virological response (4 weeks), early virological response (12 weeks), and amino-acid pattern in the viral genome, and several host factors, including age, sex, race, liver fi brosis stage, and obesity, have been associated with pegylated- α interferon / ribavirin therapy outcome.
3 Two HLA alleles are consistently associated with viral clearance and decreased disease severity of hepatitis C worldwide; these are HLA-DQB1 * 0301 and HLA-DRB1 * 11. The DRB1 * 11-DQB1 * 0301 haplotype is also associated with viral clearance and decreased disease severity of hepatitis C.
3 The HLA DR locus appears to be a prominent immuno-genetic factor infl uencing IFN treatment response.
3 The IL28B polymorphism is the fi rst genetic marker that predicts response to treatment for hepatic C virus genotype-1 with a genome-wide association study.
WHAT IS NEW HERE 3 A statistically signifi cant association of HLA class II allele
DQB1 * 0301 with sustained virological response (SVR) was found.
3 A statistically signifi cant association between the rs12979860 polymorphism and SVR was demonstrated in 428 patients of the same race (Caucasian) by PCR methods with single-nucleotide-polymorphism amplifi cation assays.
3 The HLA-DQB1 * 0301 allele and rs12979860 genotype are signifi cant independent predictors of response to pegylat-ed- α interferon / ribavirin (pegIFN / RBV) in patients with chronic HCV infection.
3 There is synergism between the HLA-DQB1 * 0301 and HLA-A * 0201 alleles with the rs12979860 CC genotype, which increases SVR valuation. The relationship between HLA alleles and IL28B has not previously been evaluated.
3 IL28B is the best predictor of SVR in patients with CHC treated with pegIFN / RBV, surpassing viral factors (viral genotype and viral load). No previous evaluation has been made of the viral and genetic factors, including IL28B, using receiver operating characteristic curves.
REFERENCES 1 . Poynard T , Yuen MF , Ratziu V et al. Viral hepatitis C . Lancet
2003 ; 362 : 2095 – 100 . 2 . Salmer ó n J , Mu ñ oz De Rueda P , Ruiz-Extremera A et al. Quasispecies as
predictive response factors for antiviral treatment in patients with chronic hepatitis C . Digest Dis Sci 2006 ; 51 : 960 – 7 .
3 . Salmer ó n J , Casado J , Mu ñ oz De Rueda P et al. Quasispecies as predic-tive factor of rapid, early and sustained virological responses in chronic hepatitis C, genotype-1, treated with peginterferon-ribavirin . J Clin Virol 2008 ; 41 : 264 – 9 .
4 . Mu ñ oz De Rueda P , Paton R , Casado J et al. Prospective studies of muta-tions in the E2-PePHD, NS5A-PKRBD, NS5A-ISDR and NS5A-V3 regions of hepatitis C virus genotype-1 and their relationship to peginterferon and ribavirin treatment response . J Virol 2008 ; 82 : 6644 – 53 .
5 . Kau A , Vermehren J , Sarrazin C . Treatment predictors of a sustained virologic response in hepatitis B and C . J Hepatol 2008 ; 49 : 634 – 51 .
6 . Feld JJ , Hoofnaglr JH . Mechanism of action of interferon and RBV in treatment of hepatitis C . Nature 2005 ; 436 : 967 – 72 .