Relationship between the interleukin-28b gene polymorphism and the histological severity of hepatitis C virus–induced graft inflammation and the response to antiviral therapy after liver transplantation
Hepatitis C is one of the most common liver diseases worldwide; it is often complicated by the development of cirrhosis and hepatocellular carcinoma. In Europe, the prevalence of hepatitis C virus (HCV) infection is approximately 1.5%. If an HCV infection is not treated or is ineffectively treated, hepatocellular function may deteriorate progressively.1, 2 For patients with end-stage chronic liver disease, liver transplantation (LT) is currently the treatment of choice.3 Recurrent HCV infection is one of the most important graft diseases3, 4 that can occur after LT. The course of hepatitis C in a graft is usually more severe than the course of hepatitis C in nontransplant patients.3–6
HCV-induced fibrosis in a liver graft is an important determinant of morbidity and mortality in patients after LT; the mode of presentation varies from minor symptoms in a clinically stable patient to a rapid loss of graft function leading to graft failure. In some patients, retransplantation may be indicated despite persistent antiviral treatment.5, 7, 8 Currently, standard antiviral therapy for HCV infection involves the administration of both pegylated interferon-α2a (PEG-IFN-α2a) and ribavirin (RBV) for up to 1.5 years. A sustained virological response (SVR; ie, the patient is serum-negative for HCV by polymerase chain reaction 6 months after the cessation of therapy) occurs in approximately 40% to 50% of patients with genotype 1 and in 80% of patients with genotype 2 or 3.9-12 After LT, such therapy is associated with frequent adverse events, and response rates, varying between 30% and 40%, are substantially lower in all HCV-positive patients.13-15
Reinfection with HCV may trigger excessive synthesis and deposition of extracellular matrix (ECM), which is usually associated with the activation of cytokine release.16, 17 Activated macrophages, lymphocytes, bile duct epithelia, endothelia, and myofibroblasts are sources of inflammatory and fibrogenic cytokines and growth factors, which can stimulate hepatic stellate cells to synthesize ECM molecules that initiate fibrosis.17 Relevant risk factors for the development of graft fibrosis have been identified; they include high levels of viremia in the early posttransplant period, HCV genotype 1b, multiple episodes of rejection, the immunosuppression regimen, and the donor's age.18-24 However, appreciable variations in graft inflammation and fibrosis, susceptibility to HCV infection, spontaneous viral clearance, and antiviral therapy responses suggest the existence of endogenous risk factors that influence the evolution of graft disease after LT. In several trials, the maximum capacity for producing different cytokines in response to appropriate stimulation has been shown to vary between individuals.25, 26 Genetic polymorphisms of enzyme systems, cytokines, and growth factors, which are involved in the mediation of immunomodulation, inflammation, ECM turnover, and antioxidative stress, may explain the substantial differences in the severity of graft damage between individuals. Previously, studies of genetic associations were undertaken in which the number of patients with recurrence of HCV infection was small, the disease severity was not specified, and/or the controls were inadequate.27, 28 Recently, in several trials, notable associations between interleukin-28b (IL-28b) gene single-nucleotide polymorphisms (SNPs) and (1) spontaneous clearance of HCV and (2) favorable responses to standard antiviral therapy (PEG-IFN–RBV) for HCV infection have been demonstrated, and they suggest the influence of multiple genes.29-33
The IL-28b gene encodes the antiviral protein interferon-λ (IFN-λ), which exhibits antiviral properties in response to IFN-α; it is up-regulated by peripheral blood mononuclear cells and hepatocytes during infection with HCV.2, 34-37 Published data suggest that its nucleotide substitution (T→G, rs8099917) lowers the cytokine expression of IL-28b and thereby reduces its antiviral potential in comparison with the T allele.29
Data on the role of IL-28b polymorphisms (rs8099917) are currently not available for HCV-positive patients after LT. However, considerations about the role of IL-28b-variants in histological and biochemical severity of HCV-recurrence, occurrence of acute cellular rejection (ACR) and antiviral treatment success are promising. Because there are similarities between the reinfection of a graft with HCV after LT and the pretransplant course of infection with HCV, we postulated that genetic variants of IL-28b (T→G, rs8099917) might play a role in the development of HCV-related graft disease and its response to treatment.
Since 1988, about 2600 patients, including more than 400-positive patients, underwent LT for end-stage liver disease. Hepatic sonography and liver biopsy were also undertaken when they were indicated (eg, there was evidence suggesting rejection).
Primary immunosuppression consisted predominantly of calcineurin inhibitors (cyclosporine or tacrolimus) with or without mycophenolate mofetil; the corticosteroid dose that was administered initially was tapered, and the corticosteroid was discontinued over a period of 8 weeks. Satisfactory allograft function was achieved in all patients by the time of discharge from the hospital. Patients remained clinically stable for at least 1 year after LT. All the patients had hepatitis C viremia and exhibited laboratory and hepatic histological features consistent with hepatitis C.
After consent was obtained for genotyping and retrospective data analysis, venous blood was collected from 212 patients reinfected with HCV between April 2007 and April 2009. For the selection of a homogeneous cohort of patients reinfected with HCV, patients who had a major vascular complication (eg, hepatic artery thrombosis; n = 4), cholestatic hepatitis (n = 2), serum positive for hepatitis B surface antigen (n = 1), appreciable alcohol intake after LT (male, >40 g/day; female, >20 g/day; n = 2), recurrent autoimmune disease (n = 1), or ischemic biliary lesions (n = 5) or underwent retransplantation less than 1 year after initial LT (n = 4) were not subjected to further evaluation. Patients in whom the evolution of hepatic fibrosis was uncertain (<2 graft biopsy samples and <1 year of follow-up after LT, n = 10) were excluded. After these exclusions, 183 individuals remained for analysis. Furthermore, histological data on patients who responded to antiviral therapy were included in the analysis of inflammation and fibrosis only when HCV replication was occurring before the onset of SVR. The median duration of retrospective observations after LT was 7.0 years (range = 1-17 years). Between April 2007 and April 2009, 1 patient died of hepatocellular carcinoma recurrence 7 years after LT, 2 patients underwent retransplantation 7 and 12 years after initial LT for graft cirrhosis, and 1 patient died of HCV-related graft failure 1.5 years after LT. Previously deceased patients were not considered suitable for analysis because no material for genotyping was available.
Demographic data for the analyzed patients are listed in Table 1.
Table 1. Demographic Data for the Patients
Total number of patients reinfected with HCV
Gender (male/female), n (%)
115 (62.8)/68 (37.2)
94 (51.6)/88 (48.4)
Age (years), median (range)
Period of observation (years), median (range)
Duration of exposure to HCV (years), median (range)
Initial immunosuppression, n (%)
Cyclosporine A and steroids or cyclosporine A, mycophenolate mofetil, and steroids
Tacrolimus and steroids or tacrolimus, mycophenolate mofetil, and steroids
ACR, n (%)
62 (33.9)/121 (66.1)
HCV genotype (1b/1a/2/3/4), n
PEG-IFN-α2a, PEG-IFN-α2b, IFN-α, and IFN-β
130, 26, 12, 4
RBV, amantadine, and silibinin
152, 7, 2
Antiviral treatment completed (n = 159, 86.9%), n (%)
Antiviral treatment incomplete (n = 24, 13.1%), n
Evaluated biopsy samples, n
Antiviral Therapy and Occurrence of ACR
Antiviral treatment was undertaken in 159 patients (86.9%) at least once. Treatment was based predominantly on an IFN-RBV combination according to standard recommendations [PEG-IFN-α2a (n = 130), PEG-IFN-α2b (n = 26), and RBV (n = 152)]. IFN-α (n = 12), IFN-β (n = 4), amantadine (n = 7), and silibinin (n = 2) were administered in different combinations as rescue therapy. Twenty-four patients who did not complete antiviral therapy were not considered for further fibrosis analysis. The decision to administer antiviral treatment was based on an absence of contraindications and on an assessment of the severity of HCV-induced disease based on serum aminotransferase levels, the presence of serum HCV RNA, and hepatic histological findings (after the exclusion of other relevant pathologies such as graft rejection). IFN-based treatment was undertaken for at least 12 months. No detailed quantitative analysis of changes in serum HCV RNA during treatment was undertaken, but qualitative data on the HCV RNA status are provided in the documentation of responses to treatment, posttreatment relapses of disease, and nonresponses (NRs) to treatment. End-of-treatment virological response (ETR; undetectable HCV on the date on which therapy was completed) and SVR (undetectable HCV at least 6 months after the completion of treatment) were used in the definitions of the outcomes of treatment. NR to primary therapy (detectable HCV RNA at the completion of treatment) and relapse after the completion of therapy (ETR-SVR difference) were regarded as indicators of treatment failure (TF).
Episodes of rejection were confirmed by the application of the Banff criteria to the histology of graft biopsy samples.38 Sixty-two of 183 patients developed at least 1 episode of ACR. These episodes were successfully treated by the administration of a corticosteroid bolus or by the modification of the immunosuppression regimen. Four patients had corticosteroid-resistant cellular rejection that responded to antibody therapy.
Diagnosis of Reinfection With HCV
The diagnosis of reinfection with HCV was made by the detection of HCV RNA in serum. All patients had antibodies to HCV and a detectable viral load after LT. An early method for assessing the viral load was applied; specifically, a standardized quantitative test with a lower limit of detection of 615 IU/mL (Versant bDNA 3.0 quantitative assay, Bayer Diagnostics) was used. A more recent quantitative assessment of the viral load was also applied (Cobas AmpliPrep/TaqMan, Roche, Mannheim, Germany); this assay has a lower threshold of detection (15 IU/mL).
Analysis of Laboratory and Histological Data
Six hundred five liver biopsy samples (at least 2 per patient) were evaluated. The severity of inflammation was graded with the classification proposed by Desmet and Scheuer (none, 0; minimal, 1; mild, 2; moderate, 3; and severe, 4).39 Fibrosis was staged on a scale of 0 to 4 (0, absent; 1, mild without septa; 2, moderate with few septa; 3, numerous septa without cirrhosis; and 4, cirrhosis). Subsequently, with respect to the degree of fibrosis, the patients were divided into a mild group and an advanced and progressive group (Table 2 and 3). The Desmet-Scheuer score was selected because of its superior reproducibility in comparison with other semiquantitative systems used for assessing fibrosis.39, 40 For each patient, the median grade of inflammation and the median and maximal stages of fibrosis that occurred during infection with HCV were determined before statistical analyses were undertaken. Mean levels of AST, ALT, glutamate dehydrogenase (GLDH), gamma-glutamyltransferase (GGT), and bilirubin were measured during routine follow-up evaluations and at other times when indicated. Data obtained after successful antiviral treatment were not included in assessments of the progression of hepatic inflammation and fibrosis because of the absence of the etiological agent.
Table 2. Hepatic Histological Data
NOTE: Grades of inflammation and stages of fibrosis are expressed as proportions of n.
Years after LT
Biopsy samples (n = 605), n
Portal fields, mean (standard deviation)
Inflammation grade, n (%)
Grade, median (range)
Fibrosis stage, n (%)
Stage, median (range)
Genomic DNA was extracted from peripheral blood leukocytes of the recipients with the NucleoSpin Blood L DNA isolation kit (Macherey-Nagel GmbH, Germany). IL-28b genotyping was undertaken with the TaqMan SNP genotyping assay (Applied Biosystems) as previously described.29-31 Polymerase chain reaction was carried out with 96-well plates with a reaction volume of 10 μL. The thermal cycle conditions were 60°C for 30 seconds, 95°C for 10 minutes, 40 cycles of 95°C for 15 seconds, and 60°C for 1 minute in accordance with the manufacturer's instructions.
Data were collected from a prospective database. SPSS 18.0 statistical software (SPSS, Chicago, IL) was used for statistical calculations involving the following data: IL-28B genotypes, HCV genotypes, donor and recipient ages, duration of infection, response to antiviral treatment, episodes of rejection, donor and recipient genders, grades of inflammation, stages of fibrosis, and serum levels of aminotransferases (ALT and AST). Differences between continuous and categorical variables were calculated with the Mann-Whitney U test, Kruskal-Wallis test, chi-square test, and Fisher's exact text as appropriate. A 2-tailed P value < 0.05 was accepted as significant. The power calculation was undertaken at the Institute for Medical Statistics (http://statistics.msi.meduniwien.ac.at); the genotype distribution was tested for the deviation from the Hardy-Weinberg equilibrium at the Institute for Human Genetics (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl).
One hundred eighty-three patients who were reinfected with HCV after LT underwent IL-28b genotyping for an assessment of the rs8099917 polymorphism (GG, 6.6%; GT, 45.9%; TT, 47.5%). The findings were correlated with laboratory and histological variables of HCV-induced liver disease. The frequencies of the G and T alleles for the entire cohort were 29.5% and 70.5%, respectively. Data on the IL-28b genotypes with respect to the grades of hepatic inflammation, the stages of hepatic fibrosis, the groups of patients classified according to hepatic inflammation and fibrosis, the serum levels of aminotransferases, the level of viremia early after LT, the outcomes of antiviral therapy, and the occurrence of ACR are summarized in Tables 3 and 4.
Table 3. IL-28b Genotypes, Serum Biochemistry, and Hepatic Histology
NOTE: ALT, AST, GGT, GLDH, total bilirubin, and viremia levels (10 log) are presented as means and standard deviations. ACR is expressed as a proportion of n.
Inflammation grade, median (range)
Low grade (0-2, n = 101)
High grade (3-4, n = 82)
Fibrosis stage, median (range)
Fibrosis group, median (maximum)
Low stage (0-2, n = 100)
Advanced stage (3-4, n = 83)
Viremia level 1 year post-LT (U/mL)
ACR, n (%)
Table 4. IL-28b Genotypes and Outcomes of Antiviral Therapy (n = 159)
IL-28 Genotype, n (%)
HCV was undetectable at the completion of treatment.
Primary therapy failure.
HCV was undetectable 6 months after the completion of treatment.
HCV was detectable 6 months after the completion of treatment.
Six hundred five graft biopsy samples were evaluated with respect to the grades of inflammation and the stages of fibrosis. The distribution of IL-28b genotypes (GG, GT, and TT) differed significantly between groups with high and low grades of inflammation (grades 0-2, n = 101; grades 3-4, n = 82; P< 0.001). The T allele was more abundant than the G allele among patients with low grades of inflammation. The median histological grade of inflammation differed significantly among IL-28b genotypes (3.0 for GG, 2.5 for GT, and 2.0 for TT, P< 0.001, Figs. 1 and 2). However, IL-28b genotypes were similar in groups with different median stages of fibrosis (P = 0.402) and different maximal stages of fibrosis during the 5-year median period of reinfection with HCV after LT (n = 100 for stages 0-2 and n = 83 for stages 3-4, P = 0.586; Table 3 and Figs. 1 and 3). During follow-up (evaluations 0.5, 1, 3, 5, 7, and 10 years after LT), significant associations were found between the G allele and the median grade of inflammation at 1 (P = 0.046), 3 (P = 0.004), and 5 (P = 0.002) years after LT (Fig. 2). However, no associations were found between the IL-28b genotypes and the mean stages of fibrosis (0.5, 1, 3, 5, 7, and 10 years after LT, P = 0.414, 0.489, 0.263, 0.910, 0.832, and 0.836, respectively; Fig. 4).
The GG, GT, and TT genotypes correlated significantly with the mean serum levels of aminotransferases measured on routine evaluation dates (ALT, 87.0 U/L for GG, 57.0 U/L for GT, and 47.3 U/L for TT, P = 0.001; AST, 72.3 U/L for GG, 54.7 U/L for GT, and 43.7 U/L for TT, P = 0.003; Table 3 and Fig. 5). However, a relationship between these genotypes and the serum levels of GGT, GLDH, and bilirubin was not apparent (Table 3). The mean viremia levels (10 log), which were measured in 143 patients within the first year after LT before the initiation of antiviral treatment, differed among the IL-28b genotypes [6.4 (0.6) U/mL for GG, 6.3 (0.7) U/mL for GT, and 5.8 (0.8) U/mL for TT, P < 0.001; Table 3 and Fig. 6].
Response to Antiviral Treatment
One hundred fifty-nine patients (86.9%) underwent at least 1 full course of antiviral treatment (n = 118 for 1 course, n = 34 for 2 courses, n = 6 for 3 courses, and n = 1 for 4 courses). Eighty-six patients (54.1%) were HCV-negative at the time of completion of therapy (ETR). In 73 patients (45.9%), there was NR to treatment. Thirty patients in the ETR group experienced a relapse of HCV-related liver disease. Thus, after 6 months of follow-up, 56 fulfilled the criterion of SVR. None of the carriers of the homozygous G allele became consistently HCV-negative. Only 1 patient with the GG genotype became temporarily HCV-negative after the completion of the second course of treatment (ETR), but the disease in this patient subsequently relapsed. The distribution of IL-28b genotypes differed substantially with respect to the response to antiviral treatment (Table 4 and Fig. 7). In particular, a highly significant association was found between the IL-28b genotypes and SVR (GG, 0%; GT, 33.9%; TT, 66.1%) versus TF (GG, 10.7%; GT, 53.4%, TT 35.9%; P< 0.001) and ETR (GG, 1.2%; GT, 34.9%; TT, 63.9%) versus primary NR (GG, 13.7%; GT, 60.3%; TT, 26.0%; P< 0.001). The G and T allele frequencies were 17.0% and 83.0% in the SVR group, 37.4% and 62.6% in the TF group, 18.6% and 81.4% in the ETR group, and 43.8% and 56.2% in the primary NR group.
Occurrence of ACR
IL-28b genotypes were equally distributed among patients who developed at least 1 episode of ACR and patients who did not develop rejection (8.1% versus 5.8% for GG, 43.5% versus 47.1% for GT, and 48.4% versus 47.1% for TT, P = 0.798; Table 3).
In this study, we determined the prevalence of IL-28b genotypes (rs8099917) in a cohort of 183 patients with recurrence of HCV infection after LT. Hepatic histological grades of inflammation, stages of hepatic fibrosis, levels of viremia before antiviral treatment, serum levels of aminotransferases (AST and ALT), ACR occurrence, and responses to IFN-based antiviral treatment were assessed. The occurrence of ACR and the maximum stage of fibrosis that developed were not associated with an altered distribution of IL-28b genotypes. After a retrospective evaluation of 605 liver biopsy samples, the G allele was identified as a better marker of severe serum biochemical and histological indices of graft infection than the T allele. Furthermore, there was a strong association between the G allele and a lack of response to antiviral treatment in 103 of 159 patients. Thus, IL-28b polymorphisms may have important predictive value. Our findings constitute one of the first reports of the potential significance of IL-28b genotypes (rs8099917) in a cohort of patients with recurrent HCV infection after LT.
Genetic polymorphisms may influence a patient's susceptibility to liver damage.41-44 Recently, several independent studies of the natural history of HCV infection have found a significant association between 2 IL-28b SNPs (rs8099917 and rs12979860) and the response to antiviral treatment and spontaneous viral clearance.29-32 The IL-28b gene, located on chromosome 19, encodes IFN-λ3 (synonymous. IL-28b), which, like IFN-α, is involved in the suppression of viral replication. However, the tissue specificity of IFN-λ is higher because its receptors are expressed predominantly by epithelia (including hepatocytes).45, 46 IFN-λ3 seems to play an important role in spontaneous and IFN-α–mediated viral clearance.2, 34, 35 The G allele of the IL-28b polymorphism (rs8099917) affects antiviral clearance capacity by lowering IL-28b expression; it has been identified as a genetic risk factor for the failure of antiviral therapy to be substantially effective.29, 31, 32 Our findings are in accordance with observations made by Suppiah et al.,29 Rauch et al.,32 and Tanaka et al.,31 who demonstrated highly significant associations between IL-28b polymorphisms and spontaneous and treatment-induced viral clearance in the usual setting of HCV-related liver disease (unassociated with LT). Spontaneous elimination of HCV, which is rare after LT, was not observed in our cohort. However, the mean pretreatment levels of viremia, which were measured during the first year after LT, appeared to be significantly influenced by the presence of the G allele, and this supports the concept of endogenous, gene-regulated antiviral activity. Immunosuppression plays a key role in inhibiting spontaneous or treatment-induced viral elimination after LT.47,48 Unfortunately, precise data on antiviral treatment before LT were not available.
Successful LT leads to the generation of a unique population of quasinormal individuals who appear to depend on effective interactions between 2 different genetic backgrounds. A liver allograft is usually colonized by the recipient's cell populations, such as endothelia and lymphatic tissue, and this creates a functional chimerism of biochemical processes.49 It would be interesting to determine the role of donor genetics in the pathogenesis of graft disease and its manifestations.
HCV-induced changes in grafts were characterized in 605 liver biopsy samples. In particular, median grades of inflammation and median and maximal stages of fibrosis that occurred during the period of exposure to HCV after LT (median = 5.0 years, range = 1-17 years) were determined. The use of an ordinal scale system for arithmetic analyses is limited, and an appropriate division of fibrosis stages during exposure to HCV was not permitted. Nevertheless, it is frequently used in an attempt to achieve a metric quantification of fibrosis. Furthermore, the development of fibrosis is not linear; accordingly, the maximal stage of fibrosis achieved and the median stage of fibrosis seemed to be the most reliable variables for describing the development of fibrosis.7 Although the semiquantitative classification advocated by Desmet and Scheuer is often criticized as being error-prone when interobserver validation is considered, this method for assessing liver pathology appears to provide the most accurate and reproducible description of inflammation and fibrosis.40 The majority of published studies that have assessed risk factors for the severity of recurrence of HCV have lacked information on hepatic histology and especially the longitudinal quantification of hepatic damage. Moreover, the recurrence of HCV infection comprises or consists of the entire spectrum of graft diseases, including asymptomatic viremia, graft hepatitis, advanced fibrosis, and cirrhosis, to be well characterized with state-of-the-art diagnostic methods on several protocol dates after LT.
The identification of noninvasive markers of hepatic inflammation and fibrosis may help us to differentiate those patients who experience recurrence of HCV infection but have stable graft function without significant progression of inflammation or fibrosis from those patients at risk of sustaining appreciable short-term graft damage, and this may also help us to define indications for antiviral treatment. Approximately 50% of graft recipients with recurrent HCV infection survive 10 years without any significant progression of fibrosis; some patients achieve this outcome without receiving antiviral treatment. Despite rigorous and expensive IFN-based antiviral therapy for HCV infection in graft recipients, the overall results tend to be poor, with only 30% to 40% of patients achieving SVR.14 Some reasonable doubts regarding the appropriateness of antiviral treatment in this group of patients seem justified. As long as IFN-α remains the crucial component of antiviral therapy, the identification of predictors of the outcome of such therapy remains important.
As for the potential success of management, frequent severe adverse events associated with IFN-RBV combination therapy are commonly accepted. However, the unnecessary exposure of nonresponders to IFN should be avoided, if it is possible, by the application of improved predictors of the outcome of therapy.50, 51 On the basis of the present findings, after LT, the IL-28b gene appears to be involved in the pathogenesis of HCV-related disease in grafts, and its genetic variants seem to influence the severity of graft inflammation. Although the mechanism of the G allele–related impairment of IL-28b function is not yet well understood, its potential value as a marker of inflammation and a predictor of the outcome of therapy should, in combination with other predictive markers, receive further consideration.29
The precise mechanisms by which genetic factors influencing the immune response may have implications for the management of chronic hepatitis C in its normal setting and after LT remain uncertain. It would seem desirable for the role of the IL-28b polymorphism to be investigated further in larger prospective studies, which ideally would address donor genetics and would include evaluations of gene variants involved in tissue inflammation, antioxidative defense, responses to antiviral agents, and ECM metabolism.
In conclusion, we have demonstrated a significant association between IL-28b genotypes (rs8099917) and the outcomes of IFN-based antiviral treatment and biochemical and histological indices of graft inflammation in 183 patients experiencing a recurrence of HCV infection after LT. Genetic variants of IL-28b (rs8099917) seem to be involved in the pathogenesis of HCV-related graft inflammation and to influence the response to HCV. The G allele may be not only a marker of severe HCV-induced graft inflammation after LT but also a predictor of an unfavorable outcome to its treatment with antiviral agents. IL-28b polymorphisms may facilitate the identification of patients at risk of developing severe graft hepatitis. Furthermore, genetic variants may be useful in predicting the potential response to antiviral therapy and in suggesting how treatment might be modified.