Potential conflict of interest: One of the authors, J.M., has a significant financial interest in IL28B genotyping. To manage this potential conflict of interest, J.M. (and the entire Duke team) was blinded to all posttransplant outcomes at the time of genotyping. Only after all genotyping was performed and the results disclosed to Mayo investigators were outcomes metrics disclosed to the Duke collaborators, including J.M. No one at Mayo has any financial interest in IL28B testing. Dr. McHutchison is employed by and receives grants from Merck/Schering-Plough. Dr. McHutchison receives grants from Roche and holds intellectual property rights IL28B patent with Labcorp. Dr. Goldstein receives grants from GSK Research and holds intellectual property rights IL28B patent with Labcorp.
Polymorphism in the interleukin-28B (IL28B) gene region, encoding interferon (IFN)-λ3, is strongly predictive of response to antiviral treatment in the nontransplant setting. We sought to determine the prevalence and impact on clinical outcomes of donor and recipient IL28B genotypes among liver transplant recipients. The cohort study included 189 consecutive patients infected with hepatitis C virus (HCV) who underwent liver transplantation between January 1, 1995, and January 1, 2005, at the Mayo Clinic, Rochester, MN. Genotyping of the polymorphism rs12979860 was performed on DNA collected from all donors and recipients in the cohort. Sixty-five patients received IFN-based antiviral therapy. The CC IL28B variant was less common in the chronic HCV-infected recipients than in non-HCV donor livers (33% versus 47%, P = 0.03). IL28B recipient genotype was significantly predictive of fibrosis stage, with TT genotype being associated with more rapid fibrosis (Pearson chi-square P = 0.024 for the comparison G versus A). Donor and recipient IL28B genotype were independently associated with sustained virologic response (P < 0.005). The presence of IL28B CC variant in either the recipient (R) or donor (D) liver was associated with increased rate of sustained virologic response (D-non-CC/R-non-CC = 3/19 [16%] versus D-CC/R-non-CC = 11/22 [50%] versus D-non-CC/R-CC = 5/12 [42%] versus R-CC/D-CC = 6/7 [86%], P = 0.0095). IL28B genotype was not significantly associated with survival (overall/liver-related). Conclusion: Recipient IL28B TT genotype is associated with more severe histological recurrence of HCV. Recipient and donor liver IL28B genotype are strongly and independently associated with IFN-based treatment response in patients after orthotopic liver transplantation. The data suggest that CC donor livers might be preferentially allocated to patients with HCV infection. (Hepatology 2011;)
Chronic hepatitis C virus (HCV) infection is the most common indication for liver transplantation in the Western world. Recurrence of HCV infection is the most common cause of death and graft loss following liver transplantation.1 Although treatment with interferon (IFN)/peginterferon (pegIFN) and ribavirin (RBV) is feasible in the posttransplant setting, sustained virologic response (SVR) rates are lower than in nontransplant patients2 and many patients do not tolerate therapy.
Genetic variation in the region of the IL28B gene on chromosome 19, coding for IFN-λ3, has recently been demonstrated to be strongly associated with SVR in patients with genotype 1 chronic HCV infection who are treated with pegIFN plus RBV in the nontransplant setting.3-5IL28B polymorphism has also been associated with spontaneous HCV clearance.6, 7 To date, the mechanism underlying the association between IL28B variants and natural/treatment-induced clearance remains unclear. The role and relevance of IL28B polymorphism (recipient/donor) to HCV recurrence after transplantation and the HCV treatment outcome after transplantation have not previously been investigated.
We therefore performed a retrospective analysis of a well-characterized HCV transplant cohort in which both recipient and donor liver DNA was tested for IL28B genotype.8 We tested for an association between recipient/donor IL28B variants and treatment outcome, as well as natural history after transplantation, including time to HCV recurrence and patient survival.
ALT, alanine aminotransferase; CI, confidence interval; HCV, hepatitis C virus; HR, hazard ratio; IFN, interferon; IL, interleukin; IQR, interquartile range; MELD, Model for End-Stage Liver Disease; OLT, orthotopic liver transplantation; RBV, ribavirin; SVR, sustained virologic response.
Patients and Methods
The study population consisted of consecutive chronic hepatitis C patients who underwent orthotopic liver transplantation (OLT) between January 1, 1995, and January 1, 2005, at the Mayo Clinic, Rochester, MN, with follow-up in all patients. DNA was collected prospectively from all donors and recipients. Data were obtained according to a standard protocol. A subset of the patients was treated with IFN-based regimens. The study protocol was approved by the Institutional Review Board of the Mayo Clinic and was carried out in accordance with institutional guidelines.
Virological (genotype, viral load), biochemical, and hematological data were measured in the certified Mayo Clinic laboratories. A sensitive qualitative assay was used to detect serum HCV RNA (COBAS Amplicor HCV Test, version 2.0 assay; Roche Molecular Systems) with a sensitivity of 100 IU/mL. Genotypes were assigned using nucleotide primers specific for a 401–base pair target sequence within the NS5 (nonstructural protein 5) region with sequence information compared with published HCV type reference sequences using the FASTA algorithm (Wisconsin Genetics Computer Group, Madison, WI).
DNA Extraction and IL28B Genotyping
DNA was extracted from stored paraffin-fixed liver tissue blocks using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA) assay. Tissue was used because whole blood was not available. Donor and recipient DNA was tested for the polymorphism rs12979860 using the ABI TaqMan allelic discrimination kit and the ABI7900HT Sequence Detection System (Applied Biosystems, Carlsbad, CA). The possible genotypes for this biallelic polymorphism are: CC, CT, and TT, where the CC variant has previously been associated with good response to pegIFN plus RBV therapy in patients infected with genotype 1 HCV.3, 9
Hepatitis C Recurrence and Allograft Histology.
Recurrent hepatitis C was defined histologically. Liver biopsies were performed routinely at 1, 3, and 5 years after OLT, and also when clinically indicated. Hepatitis C recurrence was defined as the presence of HCV RNA in the serum, as well as allograft histology showing lymphocytic infiltrates suggestive of recurrent HCV infection in the absence of other Banff criteria for acute cellular rejection. Biopsies were read by experienced hepatopathologists who used a standard system for biopsy interpretation.
Sustained Virologic Response.
SVR was defined as undetectable serum HCV RNA at 24 weeks after the end of treatment.
For follow-up study, retransplantation and mortality were recorded.
Baseline characteristics were compared using Mann-Whitney/Kruskal-Wallis Tests or χ2 tests for continuous and categorical variables, respectively. Univariate Cox regression analyses and backward and forward stepwise analyses were applied to build a multivariate proportional hazards model assessing risk factors for recurrent HCV infection, and for 5-year graft survival. The following factors were considered: IL28B genotype, sex, age, donor age, diabetes mellitus, body mass index corrected for the amount of ascites removed during the transplant procedure, Model for End-Stage Liver Disease (MELD) score, HCV genotype, platelet count, prothrombin time, aspartate aminotransferase, alanine aminotransferase (ALT), bilirubin, gamma glutamyl transferase, creatinine, iothalamate clearance, albumin, sodium, glucose, cholesterol, triglycerides, HCV RNA, warm ischemia, cold ischemia, hepatocellular carcinoma as an indication for transplantation, alcohol use as a contributing factor to liver disease, race ± time to recurrent hepatitis C, and antiviral therapy, in the analysis of graft survival. Univariate and multivariate logistic regression analyses were performed to assess which factors at the time of recurrence were predictive of SVR after peginterferon treatment. Factors that were statistically significant in univariate analysis with a P < 0.10 were entered in the multivariate Cox regression model. Several models were fit, and the final model included the covariates with the best fit to the data, according to chi-square test. All statistical analyses were performed with SAS version 9.1 (SAS Institute, Cary, NC).
A total of 220 consecutive patients infected with chronic hepatitis C underwent OLT between January 1, 1995, and January 1, 2005. Only patients alive and not requiring retransplantation in the first 90 postoperative days were included in this analysis. Donor and recipient liver tissue was available for IL28B genotyping in 189 patients. Genotype at the polymorphic site rs12979860 on chromosome 19 was suitable for analysis in 171 recipients (90%) and in 172 donors (91%). In 155 patients, both donor and recipient IL28B genotype were successfully characterized. Table 1 shows pre-OLT characteristics.
Table 1. Baseline Patient Characteristics Before Transplantation, According to Recipient IL28B Genotype
Continuous variables are expressed as medians (interquartile range).
Donor age, years
Male sex (%)
HCV Genotype (%)
Prothrombin time, seconds
Iothalamate clearance, mL/minute
During a median follow-up of 4.6 years (interquartile range [IQR], 2.4-6.9), 148 patients (80%) had evidence of recurrent hepatitis C, the median time from OLT to diagnosis of recurrence being 1.0 year (IQR, 0.3-2.2). Twenty-one patients were diagnosed with recurrent hepatitis C after protocol biopsy at 1 year after OLT, nine patients after protocol biopsy at 3 years after OLT, and one patient after protocol biopsy at 5 years after OLT. The remaining patients were not diagnosed using protocol liver biopsies but by biopsies taken because it was deemed necessary by the treating physician, usually because of a rise in aminotransferase levels.
IL28B Genotype Distribution
DNA from 171 liver transplant recipients and 172 donor livers was successfully typed for the polymorphism rs12979860 (Table 2). Both genotype frequencies were in Hardy-Weinberg equilibrium. The CC IL28B variant was significantly less common among the transplant recipients who were chronically infected with HCV than the non–HCV-infected donor livers (33% versus 47%, P = 0.03). A total of 155 patients could be defined according to donor:recipient IL28B genotype pairs (CC versus non-CC). The respective frequencies were: donor non-CC:recipient non-CC = 34%, donor CC:recipient non-CC = 32%, donor non-CC:recipient CC = 19%, donor CC:recipient CC = 15% (Table 2).
Table 2. IL28B Genotype Frequency
IL28B Genotype Frequency
IL28B genotype was not determined from 18 recipients and 17 donor livers.
All patients had virological recurrence of HCV infection following liver transplantation. A total of 110/171 (64%) of recipients in whom the IL28B single-nucleotide polymorphism was successfully genotyped were diagnosed with recurrent hepatitis C by the fifth postoperative year. Time to recurrence was delayed in recipients with the CC IL28B genotype compared to those with CT and TT genotypes (5-year recurrence: 78% versus 87% versus 100%, respectively; P = 0.0173).
Multivariate Cox regression analysis showed that the recipient IL28B C allele was an independent predictor of delayed recurrence of hepatitis C at 2 years: hazard ratio (HR), 0.619; 95% confidence interval (CI), 0.434-0.883; P = 0.0081 (Table 3). Pretransplant MELD score (HR, 1.05; 95% CI, 1.017-1.085; P = 0.0029) and pretransplant ALT level (HR, 1.004; 95% CI, 1.001-1.007; P = 0.0042) were associated with shorter time to recurrence. The recipient IL28B C allele remained an independent predictor of delayed recurrence of hepatitis C at 5 years: HR, 0.632; 95% CI, 0.466-0.856; P = 0.0031 (Table 3). Pretransplant MELD score and pretransplant ALT level were both also associated with shorter time to recurrence at 5 years. The relationship between recipient IL28B genotype and time to recurrence of hepatitis C was independent of donor IL28B genotype (recipient IL28B genotype, P = 0.030 and P = 0.015 when donor IL28B genotype was forced into the 2-year and 5-year models).
Table 3. Multivariate Cox Regression Survival Analysis Assessing Factors Predictive of Recurrent Chronic Hepatitis C at 2 and 5 Years
HCV Recurrence, 2 Years
Recipient IL28B genotype
Overall (additive genetic model)
CC vs TT
CT vs TT
Pretransplant MELD score
Pretransplant ALT level (IU/mL)
HCV Recurrence, 5 Years
Recipient IL28B genotype
Overall (additive genetic model)
CC vs TT
CT vs TT
Pretransplant MELD score
Pretransplant ALT level (IU/mL)
Among patients for whom donor liver IL28B genotype was available, recurrent hepatitis C was diagnosed in 85/172 (49%) at 2 years post-OLT, and in 114/172 (66%) at 5 years post-LT. Donor IL28B genotype was not associated with time to recurrence of hepatitis C (log-rank P = 0.5566 and 0.3369, for 2-year and 5-year survival analyses, respectively).
IL28B Genotype and Treatment Response for Recurrent Hepatitis C
Analysis of the relationship between IL28B genotype and SVR was limited to patients for whom both recipient and donor IL28B genotype was available. A total of 65 patients received antiviral therapy for recurrent hepatitis C, 50 patients were treated with pegIFN, 15 were treated with standard IFN (77%), and 57 patients (92%) received combination therapy with RBV. Ribavirin starting dose was titrated to renal function. Five patients could not be evaluated for SVR: one patient was recently treated and had not reached the end of follow-up, three died due to sepsis within 6 months of stopping treatment and before they reached the end of follow-up, and one patient completed their therapy in another center. The relationship between IL28B genotype and treatment outcome was therefore investigated in 60 patients.
The rate of SVR was strongly associated with the IL28B genotype of both the recipient and donor liver. The rate of SVR according to recipient IL28B genotype was 58% versus 47% versus 0% for CC versus CT versus TT, respectively, odds ratio (OR) = 3.43; 95% CI, 1.42-8.3; P = 0.0062 (Fig. 1A). The rate of SVR according to donor genotype was 59% versus 30% versus 0% for CC versus CT versus TT, respectively; OR = 4.00; 95% CI = 1.46-10.98; P = 0.0071 (Fig. 1B). Notably, no patient with either recipient or donor TT genotype attained SVR (n = 14). SVR was independently associated with both recipient and donor IL28B genotype in a regression model that included both as predictors (P = 0.0026 and P = 0.0030, respectively). We then considered treatment outcome according to combined recipient:donor IL28B genotype pairing (CC versus non-CC genotype, given the primary results and data from the nontransplant literature that shows that the major benefit of the IL28B polymorphism occurs in CC homozygotes).9, 10 SVR was lowest in non-CC recipients of a non-CC donor liver, increased if either the recipient or the donor was CC at the IL28B polymorphism and was maximal in the setting of CC recipients of a CC liver (SVR rates 3/19 [16%] versus 11/22 [50%]/5/12 [42%] versus 6/7 [86%], P = 0.0095; Fig. 1C). No other clinical/biochemical variable was associated with virological clearance in this cohort.
Recipient IL28B CC genotype was associated with significantly less fibrosis formation in the early posttransplant period than recipients with non-CC genotypes. A total of 172 patients underwent a biopsy at year 1. Twenty-six had genotype TT, 48 had genotype CC, 80 had the heterogeneous CT genotype, and the IL28B genotype could not be determined in 18.
At 1 year after transplantation, when biopsy data were most complete, 9/28 (32%) patients with recipient IL28B TT genotype had fibrosis stage 2 or higher, compared to 6/52 (12%) patients with genotype CC and 20/80 (25%) patients with the heterogeneous CT genotype (Pearson chi-square P = 0.024 for the comparison CC versus TT). Donor IL28B genotype was not significantly associated with fibrosis at 1 year.
Allograft inflammation at 1 year, measured by histologic activity index, did not vary significantly with donor or recipient genotype (Mann-Whitney TT versus CC P = 0.430). Histologic activity index was higher, but not significantly so, at the time of recurrence of HCV in recipients with a recipient IL28B genotype: CC, 5.0 (range, 3.0-7.0); CT, 4.0 (range, 3.0-6.0); and TT, 3.0 (range, 2.5-4.0), Kruskal-Wallis (overall) P = 0.165.
ALT levels were lower for recipient IL28B genotype CC than for TT at 3-6 months post-OLT (138 U/L for TT, 93 U/L for CT, and 60 U/L for CC; Mann-Whitney P = 0.030 for CC versus TT). Although numerically higher among patients with donor CC genotype, median ALT levels did vary significantly with donor IL28B genotype (median ALT at 3-6 months post-OLT: 126 U/L for CC, 94 U/L for TT, and 74 U/L for CT at 3 months post-OLT; Mann-Whitney P = 0.580 for CC versus TT).
Mortality and Graft Failure
There was no statistically significant difference in overall graft survival according to recipient IL28B polymorphism (overall 5-year graft survival [n = 118]: 91% versus 76% versus 84% for CC versus CT versus TT genotypes [P = 0.2168]). There was also no significant effect of donor IL28B genotype on overall graft survival (5-year graft survival [n = 124]: 79% versus 84% versus 81% for CC versus CT versus TT genotype [P = 0.6977]). Neither recipient nor donor liver IL28B genotype was found to be significantly associated with liver-related mortality at 5 years (P = 0.3956 and P = 0.2418, respectively) (Fig. 2).
An analysis was also performed of the association of IL28B genotype with the frequency of a composite endpoint consisting of: histological evidence of cirrhosis, liver-related death/retransplantation and fibrosis stage ≥2. The analysis was censored for antiviral therapy. This clinical composite endpoint was significantly associated with recipient and donor, IL28B genotype (P = 0.047 and 0.040 for recipient and donor CC versus TT genotypes, respectively) (Fig. 3).
This study reports the association between IL28B genotype and virological treatment response and clinical outcome in HCV-infected patients following OLT. This unique cohort allowed interrogation of the respective roles of the IL28B genotype of hepatocytes (donor) and nonparenchymal cells of extrahepatic origin (recipient). We identified important roles for both donor and recipient IL28B genotype in determining treatment outcome. The data also suggest that recipient IL28B genotype may determine the severity of histological recurrence of hepatitis C as indicated by progressive fibrosis. These findings have potentially important implications for the management of HCV following liver transplantation.
The frequency of the CC variant in the transplant recipients was significantly lower than that in the non–HCV-infected donor livers. This is consistent with a role for the CC variant in spontaneous clearance of HCV, with enrichment for the non-CC variants in the chronic hepatitis C population. Indeed, a role for the CC variant in promoting natural clearance has recently been established.6, 7 Patients with the non-CC genotypes are also more likely to be prior nonresponders to IFN-based therapies before proceeding to liver failure and transplantation.
IL28B polymorphism, previously associated with treatment response in the nontransplant setting,4, 5, 7, 9, 10 strongly predicted for increased rate of SVR in the current cohort. Recipient and donor IL28B genotype were both independently associated with higher rates of SVR. Compared to the patients with matched recipient:donor non-CC variants, SVR rate was higher in patients with either a donor or recipient CC variant, and highest in patients with matched donor and recipient CC variants. No patient in which either the recipient or donor IL28B genotype was TT attained an SVR. The low SVR rates associated with antiviral therapy in patients with IL28B genotype TT suggests that IFN- and RBV-based treatment might be avoided until more effective treatment protocols are developed. IL28B genotyping of donors and recipients is likely to be cost-effective in liver transplant patients. In the same vein, because SVR rates of up to 80% can be achieved, recipients with IL28B CC donor or recipient genotype should be considered for antiviral treatment if they have not already been treated.
The associations between IL28B genotype, antiviral treatment outcome, time to histological recurrence, and more advanced fibrosis did not translate to a statistically significant benefit for either liver-related or overall survival. Trends for an association of recipient IL28B genotype were seen, however, for both overall and liver-related survival (P = 0.11 and P = 0.18, respectively, for CC versus non-CC genotypes). Our sample size may have been too small to detect an effect of IL28B genotype on these important individual outcomes. However, the frequency of a composite endpoint consisting of histological evidence of cirrhosis, liver-related death/retransplantation, and fibrosis stage ≥2 was significantly associated with recipient and donor IL28B genotype (P = 0.047 and 0.040 for recipient and donor CC versus TT genotypes, respectively). As previously observed, antiviral therapy for HCV was strongly associated with survival benefit in this cohort. It remains possible that IL28B genotype may be relevant to survival, through its association with IFN responsiveness and sustained virological clearance, but this study was underpowered to detect such an effect in this cohort, of which only a minority received treatment. The very high SVR rates (>80%) seen in patients with IL28B CC donor and recipient genotype raises the possibility that the impact of recurrence of HCV might be reduced by preferentially allocating IL28B CC donors to IL28B non-CC genotype recipients. There are many practical and theoretical considerations that would need to be addressed before such a change to organ allocation policy could be considered, but the results of this study raise the theoretical possibility of improving the utility of liver transplantation among recipients with HCV infection.
The data raises the question of whether pretransplant determination of IL28B genotype has consequences for the further management of HCV-infected patients with end-stage liver disease. Although patients with TT genotype experienced earlier histological recurrence of hepatitis C, more advanced fibrosis, and reduced SVR rates to subsequent pegIFN treatment, recipient IL28B genotype was not an important predictor of graft survival in our cohort. Therefore, in our opinion, liver transplantation should not be withheld from patients with this genotype. The data also suggest that IL28B genotype may be clinically useful in selecting patients who have the best risk:benefit ratio for IFN-based treatment post-OLT. This association between recipient and donor IL28B genotype, treatment response and graft outcome will need to be confirmed in larger prospective studies.
The treatment response data also allow speculation about the biology of the association between IL28B polymorphism and pegIFN/RBV responsiveness. That both donor and recipient genotype were important suggests that the IL28B polymorphism may be associated with both hepatocyte (innate) and nonparenchymal cell (innate and/or adaptive) immune mechanisms of IFN effect. This would be consistent with recent data suggesting that phase 1 decline while on treatment, reflecting virion production/release, is dramatically increased in patients with the good response variant, but that an effect on phase 2 decline, reflecting infected cell loss due to adaptive immune mechanisms, is also present.11 The molecular biology underlying these observations will be an important area for future translational research.
The association between the recipient CC IL28B genotype and delayed time to diagnosis of HCV recurrence is interesting. HCV recurrence was defined in this cohort as the combination of virological recurrence and histological disease. The results suggest that recipient, but not donor, IL28B genotype influences the progression from virological recurrence to histological disease. Because the key histological marker of recurrence was a typical portal and/or lobular lymphocytic infiltrate (of extrahepatic [recipient] cell origin), we speculate that recipient IL28B genotype may play a role in regulating the HCV-specific, human leukocyte antigen–independent,12, 13 adaptive immune response, either at the pattern recognition/signal transduction (dendritic cell) step or the effector (T lymphocyte, plasma cell) step. IL28 is known to induce Toll-like receptor 7 and Toll-like receptor 8 expression and expression of human leukocyte antigen,14, 15 potential mechanisms for variation in immune response with IL28B genotype.
The protein product of IL28B is IFN-λ3, one of the three members of the recently described type 3 IFN family.16, 17 Type 3 IFN are secreted in response to stimuli that also trigger type 1 IFN, and activate the common JAK/STAT (Janus kinase/signal transducer and activator of transcription) signaling pathway. A key distinction lies in the restricted expression profile of the unique IFN-λ receptor, present on hepatocytes but low or absent on CD34-positive bone marrow progenitor cells.16 IFN-λ inhibits HCV in vitro,18 and antiviral activity of recombinant IFN-λ1 (IL29) has recently been confirmed in patients infected with HCV genotype 1.19 The molecular consequences of the IL28B polymorphism, and the mechanistic explanation for the relationship with IFN responsiveness as yet remain unclear.
There are a number of limitations to this study. First, decisions about treatment assignment were not randomized, but rather were made by individual clinicians on a case-by-case basis, and selection bias cannot be excluded in both the analyses of treatment outcome as well as survival. Second, treatment regimens were not uniform, although there were no obvious differences according to recipient:donor genotype pairs. Finally, although the cohort is larger than most studies of HCV after OLT, power to detect smaller effects on survival was low. The data should therefore be considered limited to hypothesis generation.
In conclusion, the data suggest that recipient IL28B TT genotype is associated with more rapid histological recurrence of HCV. Recipient and donor liver IL28B genotype are strongly and independently associated with IFN-based treatment response in patients after OLT. Treatment was generally safe and has previously been associated with improved graft survival in this cohort. The data therefore support the preferential allocation of CC donor livers to patients with HCV infection. Prospective validation in larger multicenter cohorts is warranted.