Chronic hepatitis C (CHC) is the leading indication for liver transplantation (LT) in the Western world. However, the recurrence of a hepatitis C virus (HCV) infection in the allograft is inevitable, and it is associated with progression to cirrhosis in up to 30% of patients 5 years after transplantation and with a higher rate of allograft failure and shorter survival in comparison with recipients without an HCV infection. Fibrosing cholestatic hepatitis (FCH) is an unusual but particularly severe manifestation of a recurrent HCV infection and carries an especially poor prognosis. Estimates of its frequency range from 2% to 14%.[2-4] It is typically characterized by early HCV recurrence with marked cholestasis, rapid progression to cirrhosis, and graft failure within the first 1 to 2 years after transplantation. Because of its relatively low frequency, only scarce data regarding this catastrophic presentation of an HCV infection are available, and there is a limited understanding of the factors favoring its development as well as the rate of response to antiviral therapy and recurrence after retransplantation. In recent years, polymorphisms near the interleukin-28B (IL-28B) gene have been shown to play a major role in HCV eradication (both spontaneous and after antiviral treatment) and in the severity of the clinical course of an acute HCV infection.[5, 6] The posttransplant course and liver histology have also been shown to depend on both recipient and donor IL-28B polymorphisms independently of viral eradication. The IL-28B gene codes for interferon λ. Thus, the mechanism for its effect on HCV-related disease is likely immune-mediated. Although FCH is considered to be a pauci-immune complication of recurrent HCV in the allograft, it is tempting to consider that IL-28B polymorphisms can affect its development and course as well. Two studies have examined this possibility and suggested that favorable recipient and donor genotypes are associated with an increased risk for FCH.[8, 9] In the present study, we sought to characterize patients with FCH from a large, single-center cohort of CHC transplant patients and to identify risk factors favoring its development; we included an analysis of IL-28B polymorphisms from both the recipient and the donor.
Chronic hepatitis C (CHC)–related cirrhosis is the leading indication for liver transplantation (LT). However, the recurrence of a hepatitis C virus (HCV) infection after transplantation is universal and is associated with worse outcomes. Fibrosing cholestatic hepatitis (FCH) is a particularly severe manifestation of a recurrent HCV infection and frequently results in graft failure and death. The identification of risk factors for FCH is important but has been limited by the low frequency of FCH. The interleukin-28B (IL-28B) genotype is important in an HCV infection: it is related to the clinical severity of an acute infection and may play a role in the development of FCH as well. Two hundred seventy-two consecutive LT cases for CHC were studied at a single institution. Consensus criteria were used to define an FCH cohort. The remainder of the study population served as a control group. The IL-28B genotype (at the rs12979860 locus) from both the donor and the recipient was determined, and other clinically relevant data were tabulated. A nonparametric statistical analysis was performed. Twelve cases of FCH were identified, and they were compared to a control group of 260 LT cases without FCH. A detailed analysis of clinical characteristics, including treatment responses and outcomes, was tabulated. FCH was associated with the earlier recurrence of HCV infections, higher HCV viral loads, and lower levels of immunosuppressive medications. There was a nonsignificant increase in recipient IL-28B non-CC genotypes in cases developing FCH. In conclusion, a high HCV viral load and earlier recurrence were identified as risk factors for FCH. It is still unclear what role immunosuppression plays in the pathogenesis of FCH and whether IL-28B polymorphisms constitute a risk factor. Collaborative studies with larger numbers of study subjects are needed in order to define these issues. Liver Transpl 19:1311-1317, 2013. © 2013 AASLD.
chronic hepatitis C
fibrosing cholestatic hepatitis
hepatitis C virus
PATIENTS AND METHODS
The study included 272 transplants performed for 255 patients with CHC between January 1995 and July 2010 at the Mayo Clinic (Rochester, MN). The characteristics of the population and the methodology have been previously described in detail. Patients were followed according to a standard protocol, which included mandatory biopsies 7 days and 4 and 12 months after LT and yearly thereafter. The diagnosis of FCH was based on the presence of compatible clinical and biochemical data along with liver biopsy findings supporting this condition. Specifically, the following criteria were used to define FCH: cholestasis and elevated bilirubin levels within the first year after LT, very high HCV RNA levels, and a liver biopsy or liver biopsy samples showing a rapid progression of fibrosis, a ductular reaction, a paucity of inflammation, and the ballooning of hepatocytes. The absence of alternative explanations for cholestasis/bilirubinostasis, such as biliary or hepatic artery complications and drug-induced liver injury, was also evaluated. In patients with no FCH, the recurrence of HCV was defined as a panlobular or portal lymphocytic infiltrate with the detection of quantifiable HCV RNA and no alternative cause. Donor and recipient DNA was extracted from stored paraffin-fixed liver tissue blocks with the QIAamp DNA mini kit assay (Qiagen, Valencia, CA). The polymorphism rs12979860 was tested in all cases (Applied Biosystems, Carlsbad, CA). The study protocol was approved by the institutional review board of the Mayo Clinic.
Nonparametric tests were used for statistical analyses because of a lack of a normal distribution for many of the variables. Continuous variables were summarized as medians and interquartile ranges, and comparisons between groups were performed with the Mann-Whitney rank-sum test. Categorical variables were compared with Fisher's exact test. IL-28B polymorphisms had to be grouped as favorable or unfavorable because the expected frequency was inappropriate for a χ test when the 3 genotypes were being used separately in the analyses. HCV RNA levels reported to be higher than the upper limit of quantification were changed to the upper limit of quantification for purposes of statistical calculation. All statistical tests were implemented with Stata 9.2 (StataCorp, College Station, TX).
Twelve cases of FCH in 11 unique LT patients with CHC were identified [12 of 272 cases (4.4%) and 11 of 255 patients (4.3%)]. All transplants were obtained from deceased donors, and there were no cases of donation after cardiac death. The demographics, IL-28B polymorphisms, times to HCV recurrence, and final outcomes for all FCH patients are shown in Table 1. HCV recurrence occurred within the first 5 months after transplantation in all cases. Half of the episodes were treated with peginterferon and ribavirin, with only 2 of 6 showing a virological response. Each of these 2 patients had 1 CC genotype of IL-28B, either from the donor or the recipient. The immunosuppression for treated patients was tacrolimus for all nonresponders and either tacrolimus or cyclosporine for responders. Three patients underwent retransplantation. Two of the retransplant patients were viremic, and FCH recurred in 1 of them, whereas the third patient had a virological response by the time of retransplantation, and HCV did not recur afterwards. Nine of the 11 patients died: 8 from graft failure secondary to FCH (2 of them also presented with sepsis, and another had hemoperitoneum) and the ninth (patient 5) from a disseminated fungal infection (Aspergillus and cryptococcosis) while experiencing a virological response with peginterferon and ribavirin. The 2 patients successfully undergoing retransplantation survived for at least 5.8 years (patient 7; viremic) and 8 months (patient 10; nonviremic).
|Patient||Sex||Age (Years)||IL-28B||Time to FCH (Months)||Genotype||HCV RNA (U/mL log10)||Treatment Outcome||Treatment Duration (Weeks)||Final Outcome||Time to Outcome (Months)|
|1||Male||64||CT||CC||0.6||4||Positive||Not applicable||—||Death, sepsis||9|
|1st LT||46||CT||—||1.3||—||>6.89||Not applicable||—||Re-LT||2.5|
|2nd LT||47||CT||CC||0.4||—||>6.89||Not applicable||—||Death||2.4|
|3||Male||55||CT||CC||0.7||1b||7.36||Not applicable||—||Death, hemoperitoneum||1.8|
|5||Male||61||CC||CT||0.2||2||>6.89||Virological response||7||Death, sepsis||4.2|
|6||Female||39||TT||—||1.9||1a||>7.45||No responsea||4||Death, sepsis||3.6|
The 260 LT cases without FCH were used as a control group for comparison against FCH episodes. Demographics, biochemical data at the time of recurrence, and factors possibly involved in the pathogenesis of FCH were compared between the 2 groups (Table 2). Donors for patients with FCH tended to be older, but there were no other differences in the baseline characteristics of either the recipients or the donors. The time to recurrence was shorter for FCH patients, and these patients presented with higher total HCV RNA and bilirubin levels and showed a trend of higher alanine aminotransferase levels. The great majority of the patients remained on tacrolimus during the first 2 months after transplantation, although the areas under the curve, calculated from trough levels taken on days 10, 21, and 60 after transplantation, were higher for non-FCH patients. Patients with FCH also showed a trend of fewer episodes of acute cellular rejection (either treated with methylprednisolone boluses or moderate/severe) at day 7, but there was no difference when we considered the first 3 months after transplantation (early rejection). Cytomegalovirus was not associated with FCH. Recipient and donor IL-28B genotyping was successful in 238 (88%) and 241 LT cases (89%), respectively. As shown in Fig. 1, there were no differences in the frequencies of IL-28B polymorphisms between patients with FCH and patients without FCH.
|FCH Cases (n = 12)b||Controls (n = 260)||P Value|
|Sex: female [n (%)]||5 (42)||80 (31)||0.5|
|Age at LT (years)a||54 (43-63)||52 (46-57)||0.6|
|Body mass index at LT (kg/m2)a||27 (25-30)||28 (23-32)||0.8|
|Donor age (years)a||59 (25-64)||45 (28-54)||0.1|
|Explant steatosis [n (%)]||2 (17)||16 (6)||0.1|
|Allograft steatosis [n (%)]||2 (17)||46 (18)||1.0|
|Preservation time (minutes)a||418 (349-486)||437 (377-511)||0.5|
|Anhepatic phase (minutes)a||78 (66-89)||79 (66-91)||0.7|
|Time to recurrence (months)a||1.6 (0.8-2.5)||4.3 (3.1-10.3)||<0.001|
|HCV RNA at recurrence (U/mL log10)a||7.36 (6.89-7.66)||6.54 (5.86-6.88)||<0.001|
|Alanine aminotransferase at recurrence (U/mL)a||234 (86-314)||136 (69-254)||0.08|
|Total bilirubin at recurrence (mg/dL)a||3.1 (1.8-12.6)||0.9 (0.6-1.5)||<0.0001|
|Immunosuppression: tacrolimus [n (%)]||11 (92)||237 (91)||0.5|
|Tacrolimus area under the curve 60 days after LTa||437 (327-530)||542 (470-638)||0.01|
|Acute cellular rejection at day 7 [n (%)]||0||56 (22)||0.08|
|Early acute cellular rejection [n (%)]||2 (17)||71 (27)||0.5|
|Cytomegalovirus infection [n (%)]||1 (8)||38 (15)||0.7|
FCH is an aggressive form of HCV infection almost exclusively seen in the setting of severe immunosuppression. Histologically, it is characterized by hepatocyte ballooning, intracellular and canalicular cholestasis, a ductular reaction, periportal and pericellular fibrosis, and a moderate inflammatory infiltrate.[4, 12] Because of the particular setting in which FCH occurs and its histopathological features, it is considered a pauci-immune complication in which hepatocyte damage is related not only to inflammation but also to direct viral cytotoxicity. After LT, FCH constitutes the severest form of HCV recurrence with almost universal mortality.
Antiviral treatment with peginterferon and ribavirin is the current standard of care for HCV in LT recipients and leads to successful eradication of the virus in approximately 30% to 35% of patients.[14, 15] However, because of the low frequency of FCH, the likelihood of an antiviral response in this setting is still uncertain: most available data are derived from single-patient reports and small series. Moreover, when this condition is identified, many patients are sick enough not to be considered eligible for antiviral treatment or fail to tolerate the side effects and have to stop therapy early. The endpoint of antiviral therapy may differ from the endpoint for non-FCH patients: one report suggested that therapy should be indefinite, and others showed a resolution of FCH with antiviral therapy despite virological relapse after the treatment was stopped.[2, 16] A virological response was documented in 4 of 7 patients treated by Gopal et al., but 2 of them relapsed and died from FCH after the suspension of the treatment. Cimsit et al. treated 5 patients and observed a positive response in 4 of them, but it was not clear from their publication whether a virological response (sustained or not) was accomplished. Satapathy et al. did not find any complete virological response in 8 patients who attempted antiviral treatment. Ong et al. described fatal outcomes for 3 of 4 treated patients, none of whom had a virological response. Graziadei et al. reported 7 treated patients, and 1 achieved a sustained virological response. In our cohort, only 2 of 6 treated patients had a virological response. Unfortunately, 1 died from sepsis after HCV RNA became undetectable. In the future, the use of direct-acting antivirals and peginterferon-free regimens should have a positive impact on virological responses and improve the prognosis for patients with FCH.
A recent meta-analysis has shown that immunosuppression with cyclosporine increases the efficacy of antiviral therapy in comparison with tacrolimus. Although it has been suggested that cyclosporine could also increase the efficacy of antiviral therapy in the setting of FCH, no study has provided sufficient evidence, including a control group with tacrolimus.[2, 3, 19] Moreover, there are no data to support the idea that any of the calcineurin inhibitors can favor FCH development in comparison with others. Our analysis does not suggest the association of any of these medications with FCH, but only 1 of the 6 treated patients was on cyclosporine, and this makes it impossible to draw any conclusions.
Retransplantation in the setting of FCH has a particularly poor prognosis. In one multicenter study assessing survival in retransplant patients with HCV, the 1-year survival rate was 0% for 4 patients with FCH, and this highly contrasted with the 1-year survival rate of 67% (n = 39) for patients with recurrent HCV but no FCH. This study also highlighted the lack of acceptance for retransplantation for patients with FCH because this was the second leading reported contraindication. Satapathy et al. reported on 5 patients who underwent retransplantation with no recurrence of FCH, but only 1 achieved long-term survival. Berenguer et al. reported 1 patient with no repeat FCH after retransplantation. In our series, only 1 of the 2 cases with viremia who underwent retransplantation developed recurrent FCH, and the other had long-term survival despite persistent viremia. In contrast, none of the nontransplant patients survived. Thus, repeat FCH is not a rule after retransplantation, and this should be considered particularly for patients experiencing a virological response if their clinical condition continues to deteriorate.
Risk factors for FCH development remain unresolved. Older donors, corticosteroid treatment for acute cellular rejection, high levels of HCV RNA, and quasispecies variants have all been implicated in previous reports.[8, 13, 22] In the present study, a trend of older donors as well as higher HCV RNA levels was identified for patients with FCH versus non-FCH controls. Remarkably, acute cellular rejection and the use of corticosteroid boluses were not increased in patients with FCH versus controls. Although previous studies have shown an increased risk with corticosteroid boluses, another recent study has also failed to report such an association, and this is in agreement with our findings. Notably, an analysis of serial tacrolimus levels (area under the curve) showed that patients with FCH were exposed to less immunosuppression. This, coupled with the lack of an association between corticosteroid boluses and FCH, de-emphasizes the role of immunosuppressors as the main risk factor for FCH development and suggests that other factors may be more important in the pathogenesis of FCH.
Recent studies have confirmed the major role that IL-28B plays in many features of HCV biology defining the course of posttransplant HCV. Table 3 shows results from studies addressing IL-28B in patients with FCH. A study by Graziadei et al. showed that recipients with FCH had a decreased frequency of the favorable IL-28B polymorphism in comparison with noncirrhotic controls with HCV recurrence, and this constituted an independent risk factor for FCH (hazard ratio = 2.92, 95% confidence interval = 1.15-7.42). After genotyping 2 IL-28B loci, rs12979860 and rs8099917, Hanouneh et al. could identify only increased expression of the favorable genotype for the rs8099917 locus in donors with FCH versus 10 selected controls with no FCH. This was associated with a 10-fold increased risk for FCH (hazard ratio = 9.6, 95% confidence interval = 1.1-83.1). Our study's figures for IL-28B polymorphisms in FCH patients look similar to findings from previous studies because a lower frequency of the favorable genotype was observed in recipients, whereas a higher frequency was noted in donors. However, none of these differences reached statistical significance. It is worth noting that the lone CC genotype recipient who developed FCH in our study received a liver from a 70-year-old donor (whose age was much higher than the average donor age, even in the FCH cohort).
|Recipients (%)||Donors (%)|
|FCH||No FCH||FCH||No FCH|
|Graziadei et al. (2012)||5a||70||25||26||60||14||50||50||0||41||44||15|
|Hanouneh et al. (2012)||12.5||50||37.5||20||50||30||50||50||0||20||60||20|
|Hanouneh et al.||62.5||37.5||0||40||50||10||75a||25||0||40||60||0|
Three things need to be considered when we analyze results for IL-28B and FCH. First, all studies have been limited by the small number of patients with FCH, which is due to the low frequency of the disease. Furthermore, genotyping has not been possible for all of them. Thus, these studies are certainly underpowered. For example, we had to group the unfavorable genotypes in order to be able to perform reliable statistics. Second, the control groups have been heterogeneous, and it is very likely that the studies by Graziadei et al. and Hanouneh et al. suffered from both selection and spectrum biases. The use of pathology scales specifically designed to address FCH features and severity may be helpful in better discriminating how the spectrum of HCV recurrence (from mild recurrence to FCH) correlates with IL-28B. Finally, confounding factors such as age, viral loads, and degrees of immunosuppression need to be adjusted in multivariate analyses, and this has been possible in only 1 of the reported studies (the one with the largest number of FCH patients).
The low prevalence of FCH makes it difficult to determine what risk factors are associated with this almost universally fatal condition. The outcomes after antiviral therapy or retransplantation are also difficult to predict. Although the 3 available studies addressing a possible association between IL-28B and FCH do not allow a firm conclusion to be drawn, all reported a several-fold decrease in the frequency of the favorable genotype (rs12979869) among recipients with FCH versus patients without FCH. Because of the small numbers of patients with FCH analyzed in each study, it is possible that there has been a systematic type II error across analyses on the basis of the lack of statistical power. A protective role for the C allele of IL-28B rs12979860 cannot be excluded. The role of IL-28B in FCH is undefined. Collaborative multicenter studies that are better powered and have a homogeneous approach to case and control definitions are needed to more definitively address this issue. The accurate identification of risk factors for FCH is likely to decrease the impact of FCH in the future.