Recurrent hepatitis C virus (HCV) infection is a major cause of morbidity and mortality after liver transplantation for HCV-related end stage liver disease. Although previous studies have shown a short-term effect of interferon-based treatment on fibrosis progression, it is unclear whether this translates to improved graft survival. We evaluated whether treatment of recurrent HCV leads to an improved graft survival. Cohort study included consecutive HCV patients who underwent liver transplantation between 1 January 1995 and 1 January 2005 in the Mayo Clinic, Rochester, MN. Two hundred and fifteen patients were included in the study. During a median follow-up of 4.4 years (interquartile range 2.2–6.6), 165 patients (77%) had biopsy-proven recurrent HCV infection confirmed by serum HCV RNA testing. Seventy-eight patients were treated. There were no differences in MELD-score, fibrosis stage or time towards HCV recurrence between treated and untreated patients at time of recurrence. There was a trend for greater frequency of acute cellular rejection among untreated patients. The incidence of graft failure was lower for patients treated within 6 months of recurrence compared to patients not treated within this time-period (log rank p = 0.002). Time-dependent multivariate Cox regression analysis showed that treatment of recurrent HCV infection was statistically significantly associated with a decreased risk of overall graft failure (hazard ratio 0.34; CI 0.15–0.77, p = 0.009) and a decreased risk of graft failure due to recurrent HCV (hazard ratio 0.24; CI 0.08–0.69, p = 0.008). In conclusion, although a cause and effect relationship cannot be established, treatment of recurrent HCV infection after liver transplantation is associated with a reduced risk of graft failure.
Chronic hepatitis C virus (HCV) infection is the most common indication for liver transplantation in the United States and Europe (1). Recurrence of HCV infection after transplantation is a serious cause of morbidity and mortality, resulting in death, graft loss or cirrhosis in one-third of HCV-infected recipients by the fifth postoperative year (2). Although treatment with interferon or peginterferon and ribavirin is feasible in the posttransplant setting, sustained virological response (SVR) rates are lower than in nontransplant patients (3).
Two randomized controlled trials have reported the histological effects of antiviral therapy on treated and untreated recipients with biopsy proven recurrence of HCV infection (4,5). Both studies noted improvement in fibrosis score among treated patients. In addition, others found that sustained virological responders treated for posttransplant recurrent HCV infection experienced a lower incidence of cirrhosis during long-term follow-up than the nonresponders (6, 7). It is, however, unknown whether antiviral therapy for recurrence of HCV infection confers improved graft survival.
As the relative contributions of direct and immune mediated graft injury to the course of posttransplant graft survival are not known, until now antiviral treatment is typically offered to patients with clear evidence of HCV-mediated graft injury. However, with the current lack of scientific evidence on the optimal timing of antiviral treatment after liver transplantation for chronic hepatitis C, some physicians use a lower threshold for treatment than others. Moreover, patient related factors, such as recovery from posttransplant complications and psychological status, might play a role in the decision to start treatment.
The aim of this study was to assess whether antiviral treatment improves graft survival in patients with recurrent hepatitis C after transplantation, using a minimal definition of HCV recurrence.
Single center, retrospective cohort study.
All consecutive chronic hepatitis C patients who underwent liver transplantation between 1 January 1995 and 1 January 2005 in the Mayo Clinic in Rochester, MN. Both patients receiving their first graft and patients who underwent retransplantation could enter the study.
The study protocol was approved by the Institutional Review Board of the Mayo Clinic and was carried out in accordance with institutional guidelines. Patients gave informed consent for their data to be used in the study.
Data were obtained on patient demographics (gender, ethnicity and age at transplantation), anthropomorphics (height and weight), donor demographics (donor age) and transplant procedure (cold and warm ischemic time). The body mass index (BMI) was calculated as weight divided by height squared.
Treatment characteristics (peginterferon/standard interferon, ribavirin, dose reductions, use of erythropoietin, treatment duration and response to treatment) were documented.
SVR is defined as undetectable serum HCV ribonucleic acid (RNA) at 24 weeks after the end-of-treatment.
Analytical procedures: Virological data (genotype and viral load), biochemical data (creatinine, bilirubin, glucose, sodium, AST, ALT, gamma GT, cholesterol and triglycerides) and hematological data (hemoglobin, platelet count and INR) were measured in the certified Mayo Clinic laboratories. The model for end-stage liver disease (MELD)-score was calculated as previously described (3.8 × log(bilirubin) + 11.2 × log(INR) + 9.57 × log(creatinine) + 6.43) (9).
Retransplantation and mortality were recorded. Graft failure was defined as occurring at the time of retransplantation or death.
HCV recurrence and allograft histology
Liver biopsies were routinely performed at 1, 3 and 5 years after transplantation and when clinically indicated. A minimal definition of recurrence of HCV infection was applied. HCV recurrence required allograft histology showing lymphocytic infiltrates suggestive of recurrent HCV infection as determined by an experienced pathologist (absence of other Banff criteria for acute cellular rejection, i.e. absence of endotheliitis and cholangitis), confirmed by the presence of HCV RNA in the serum (positive qualitative HCV RNA test or quantitative HCV RNA test showing >103 IU/mL).
Baseline characteristics were compared using Mann–Whitney and χ2 tests.
The Kaplan–Meier method was used to assess overall graft survival and to estimate the effect of treatment on graft survival after recurrence. In the latter analysis, but not in the Cox regression analysis, patients who had severe recurrence resulting in graft failure or death within 6 months after recurrence were excluded by using 6 months after recurrence as time 0. Patients were classified according to history of treatment at 6 months to avoid survivor-treatment bias (10). Groups were compared using log-rank tests.
Univariate Cox regression analyses and backward and forward stepwise analyses were applied to build a multivariate proportional hazards model assessing risk factors for graft failure. In this analysis, which was performed on an intention-to-treat basis, treatment was modeled as a time-dependent covariate to represent the ability of patients undergoing treatment during follow-up to change their status from ‘untreated’ to ‘treated’. Several models were fit, the final model including the covariates with the best fit to the data, according to Chi-square test (11). The following factors were considered: gender, age, donor age, treatment, diabetes mellitus, BMI, MELD-score, fibrosis stage, genotype, platelet count, INR, AST, bilirubin, creatinine, albumin, ALT, cholesterol, triglycerides, HCV RNA, warm ischemia and cold ischemia. Importantly, in all models, treatment was consistently significantly associated with decreased graft failure.
The final model for overall graft failure that provided the best fit to the data included the following covariates: albumin, MELD-score, BMI, platelet count and treatment. The final model for graft failure due to hepatitis C included the following covariates: albumin, MELD-score and treatment. The reported hazard ratios are the relative increases in hazard associated with increases of 10 g/dL for the covariate albumin and 10 years for donor age. Since the definition of SVR is undetectable serum HCV-RNA by sensitive molecular tests at 24 weeks after the end-of-treatment, we used this time point as time 0 for classifying response versus nonresponse. The results are reported as hazard ratios with 95% confidence intervals.
Between 1 January 1995 and 1 January 2005, 220 liver transplantations were performed for HCV-related liver disease. Five patients did not give informed consent and the study thus describes 215 liver transplantations. One hundred seventy-eight patients (83%) had a complete follow-up until 1 December 2006, 6 months prior to closure of the database.
During a median follow-up of 4.4 years (interquartile range (IQR) 2.2–6.6) 165 patients (77%) had evidence of recurrent HCV infection, the median time from OLT to diagnosis of recurrence being 1 year (IQR 0.3–2.2). Twenty-four patients were diagnosed with recurrent HCV after protocol biopsy at 1 year after OLT, 12 after protocol biopsy at 3 years after OLT and 2 patients after protocol biopsy at 5 years after OLT. The remaining 127 patients were diagnosed after biopsy for clinical indications such as elevated liver enzymes. There were no differences in occurrence of graft failure among patients diagnosed with recurrent HCV after protocol biopsy or after biopsy for elevated liver enzymes. Of the 50 patients without proven recurrence, 10 were treated before transplantation and were HCV RNA negative, 21 had no signs of recurrence on either biopsy or by HCV RNA testing and 19 had signs of recurrence on either biopsy (n = 8) or by HCV RNA testing (n = 11), but not on both tests.
Treatment was initiated in 78 patients; 17 patients were treated with standard interferon and 61 patients were treated with pegylated interferon. Sixty-nine patients (88%) received combination therapy with ribavirin. Dose reductions were needed in 57 patients (73%) and 39 patients (50%) needed erythropoietin treatment to maintain hemoglobin levels. The median duration of treatment was 46 weeks (IQR 27–57). Twenty-six patients (34%) achieved SVR and 45 patients (58%) did not achieve SVR: 40 patients were nonresponders and 5 patients had an end-of-treatment response but relapsed. The remaining 7 patients could not be evaluated for SVR: 2 completed their therapy in another center, 4 died before they reached the end-of-follow-up and 1 was treated recently and had not reached end-of-follow-up yet. The average time (mean ± SEM) from transplantation for treated and untreated patients was 78.7 ± 3 months versus 87.9 ± 2.4 months, respectively (p = 0.04). The distribution of CsA:TAC at enrollment for treated and untreated patients was 9:69 and 17:70, respectively (p = 0.159).
Table 1 shows the baseline characteristics at the time of recurrence for treated and untreated patients. Although treated patients had higher serum albumin levels at baseline, there were no differences in severity of recurrence as determined by MELD-score or fibrosis stage. Also, the time between liver transplantation and diagnosis of recurrence was similar for treated compared to untreated patients (p = 0.61, Table 1). Eleven out of 87 untreated patients and 10/78 treated patients developed CMV infection post OLT (p = 0.99). Twelve patients who received treatment for recurrent HCV infection and 24 untreated patients had received high dose prednisone treatment for rejection (800 mg on three alternating days) in the first month after OLT (p = 0.058).
Table 1. Baseline characteristics at time of recurrent hepatitis C virus (HCV) infection1
1Continuous variables are expressed as medians (inter quartile range).
2MELD = Mayo model for end-stage liver disease.
3HCC = Hepatocellular carcinoma.
Number of patients
Male gender (%)
Donor age, years
Diabetes mellitus (%)
Body mass index, kg/m2
Fibrosis stage (%)
Platelet count, ×109 cells/L
International normalized ratio
Aspartate aminotransferase, U/L
Alanine aminotransferase, U/L
HCV ribonucleic acid (×106 IU/mL)
Warm ischemia, min
Cold ischemia, min
Interval between liver transplantation and HCV recurrence, years
Pretransplant HCC3 (%)
Pretransplant alcohol abuse (%)
The majority (n = 57, 65.5%) of the patients who did not receive antiviral treatment did not have advanced fibrosis on most recent liver biopsy, as indicated by ≥stage 2 fibrosis. A total of 30 patients (34%) in the untreated group did have progressive fibrosis (≥stage 2 fibrosis stage). Reasons for not initiating antiviral treatment in these patients were examined in detail. Results are presented in Table 2. A history of acute cellular rejection was an important factor in five patients (5.7%), including one who had a history of severe acute cellular rejection. The others had Banff A1 (mild) acute cellular rejection. Further, in two patients, there were some features of rejection at the time of histological progression, leading to uncertainty regarding the diagnosis. Rejection was thus a factor in deciding not to treat in 8% of untreated patients. Other reasons for not initiating antiviral therapy despite fibrosis stage ≥2 included uncertain diagnosis (n = 4), patient refusal (n = 4), patient debility (1–3) and concerns regarding compliance (active alcoholism, n = 1). Treatment was contraindicated due to psychiatric (n = 2) hematologic (n = 7) and or poor renal function (n = 4) in a further 13 untreated patients.
Table 2. Reasons for not treating recurrent HCV Infection in untreated patients with recurrent hepatitis C virus (HCV) infection (n = 87)
1Uncertain diagnosis resulted from presence of ischemic cholangiopathy (n = 2) and some features of rejection (n = 2).
2Hematologic contra-indications included anemia, thrombocytopenia and leucopenia.
HCV recurrence mild
History of rejection
In order to test the hypothesis that treated patients could have been selected based upon their medical compliance, we compared the mean serum levels of tacrolimus and cyclosporine, which were monitored closely during the first month after transplantation, between treated and untreated patients as a measure for previous treatment compliance. No statistically significant differences were found in mean serum levels of tacrolimus and cyclosporine and their standard deviations (SD). The mean tacrolimus levels were 9.9 ng/mL (SD 2.8) in treated versus 9.6 ng/mL (SD 2.4) in untreated patients (p = 0.59). The mean cyclosporine levels were 240 ng/mL (SD 103) in treated patients versus 196 ng/mL (SD 87) in untreated patients (p = 0.29).
Effect of treatment on survival
Figure 1 shows the overall graft survival of all patients included in the study. Seventy-eight patients were treated and 87 patients remained untreated, despite of HCV recurrence. The median time between recurrence of HCV infection and start of interferon-based treatment was 0.23 years (IQR 0.02–1.5).
Retransplantation or death due to recurrent HCV infection occurred in 5 treated patients (6.4%) and in 22 untreated patients (25%), while the occurrence of graft failure due to other causes was similar in these two groups (Table 3).
Table 3. Causes of graft failure among patients with recurrent hepatitis C virus (HCV) infection1 (n = 165)
Treated (n = 78)
Untreated (n = 87)
1As proven by liver biopsy and confirmed by HCV ribonucleic acid testing.
2Includes hepatorenal syndrome and hepatic artery thrombosis.
Liver related (overall) (%)
Liver-related (specified) (%)
Recurrent HCV (%)
Non-liver-related (overall) (%)
Non-liver-related (specified) (%)
Non-liver malignancies (%)
Figure 2 shows that overall graft survival was longer for patients treated within 6 months of HCV recurrence compared to patients with recurrent HCV infection not treated within this time-period (log rank p = 0.002). Time-dependent multivariate Cox regression analysis showed that treatment of recurrent HCV was statistically significantly associated with a decreased risk of graft failure due to recurrent HCV (HR 0.24; CI 0.08–0.69, p = 0.008) (Table 4) as was serum albumin (HR 0.11; CI 0.04–0.29, p < 0.001). Higher MELD-scores at the time of recurrence were associated with a higher risk of graft failure due to recurrent HCV (HR 1.31; CI 1.14–1.50, p < 0.001).
Table 4. Graft survival by treatment status1
Unadjusted hazard ratio
95% confidence interval
Adjusted hazard ratio
95% confidence interval
Events/10,000 patient years
Events/10,000 patient years
1In all Cox regression analyses, treatment was modeled as a time-dependent covariate.
2Adjusted for albumin levels, Mayo model for end-stage liver disease (MELD)-score, body mass index and platelet count.
3Adjusted for albumin levels and MELD-score (see also Statistics).
Overall graft failure
Graft failure due to recurrent hepatitis C
Treatment of recurrent HCV infection (HR 0.34; CI 0.15–0.77, p = 0.009), serum albumin levels (HR 0.14; CI 0.07–0.26, p < 0.001) and BMI (HR 0.93; CI 0.89–0.97, p = 0.001) were statistically significantly associated with a decreased probability of overall graft failure. The following factors were associated with an increased probability of overall graft failure: MELD-score (HR 1.23; CI 1.12–1.36, p < 0.001) and platelet count (1.01; CI 1–1.01, p = 0.030).
We performed an additional analysis in which patients who were retransplanted were censored at the time of retransplantation (n = 17) instead of being reincluded in the analysis. The effects of treatment on overall graft failure (HR 0.39; CI 0.18–0.88, p = 0.023) and on graft failure due to recurrent HCV (HR 0.27; CI 0.09–0.86, p = 0.027) remained essentially unchanged.
The effectiveness of treatment for chronic HCV infection is usually evaluated by the number of patients achieving SVR, defined by HCV RNA, which is persistently undetectable for at least 6 months following completion of antiviral therapy (12). Although SVR has been shown to be associated with an improved clinical outcome in nontransplant patients, liver transplant recipients have more complex considerations, including the potential effects of chronic activation of the interferon stimulated response element through exogenous interferon administration. As the relative contributions of direct (cytopathic) and immune mediated graft injury to the course of posttransplant graft failure are not known, antiviral treatment is typically offered to patients with clear evidence of HCV-mediated graft injury. In order to demonstrate a beneficial effect of HCV treatment on survival, ideally, a cohort of treated patients would be compared to a similar but untreated control group.
Because of the limited success of interferon-based treatment in the posttransplant setting, the initiation of treatment for recurrent HCV infection at our center has depended on patient and physician preference over the past decade, more than on disease severity. A prospective protocol for obtaining liver biopsies together with a minimalist definition for recurrence of HCV, provided us with a unique opportunity to assess the effect of HCV treatment on graft survival, studying comparable groups of treated and untreated patients. Importantly, our study shows for the first time that treatment of recurrent HCV infection is independently associated with an improved graft survival.
Several smaller studies have previously reported a probability of improvement in fibrosis score after treatment varying between 0% and 31% (6,13–18). A recent study by Carrion et al. showed that regression of fibrosis after treatment for recurrent HCV infection goes along with an improvement in portal pressure, as measured by hepatic venous pressure gradient, which may contribute to the effect of treatment on graft survival (4).
A study by Bizolon et al., including 25 treated patients with paired liver biopsies and 21 untreated controls, suggests that improvement in fibrosis score does not necessarily depend on achievement of SVR (19). This means that there likely is some benefit of treatment, even if no SVR is achieved. However, SVR remains an important prognostic indicator. Picciotto et al. showed in univariate analysis that survival was better for sustained virologic responders than for nonresponders after treatment of recurrent HCV (8).
Our study has several limitations. The nonrandomized nature of the study may have led to selection bias. Patients with severe recurrence may not have been eligible for treatment, because they were too sick, or in contrast, more severe recurrence may have been an indication for treatment. However, the MELD-score and the fibrosis stage at the time of recurrence were similar for both groups, suggesting that the severity of recurrence was similar for both treated and untreated patients. Furthermore, when we started the Kaplan–Meier survival analyses at 6 months after recurrence, thus excluding patients in whom severe recurrence led to graft failure or mortality before treatment could be initiated, the difference in survival between treated and untreated patients was essentially unchanged and remained statistically significant.
Although we corrected in multivariable analysis for baseline factors shown to influence survival, the result of nonrandomized studies such as this might still be influenced by unmeasured variables (20). Since most of the patients were referred from other centers, we did not have reliable data on previous interferon-based treatment before transplantation. Thus, patients in the untreated group may have been selected for previous nonresponse. Nonresponders to interferon-based treatment are usually characterized by a higher proportion of genotype 1, male gender, higher fibrosis stage and higher viral load. Notably, there was no difference between these variables for the treated compared to the untreated patients at the time of recurrence, except that male gender was more prevalent in the treated group. This suggests, but does not prove, that previous nonresponse was not more frequent in the untreated group. Another potential source of selection bias may be patient compliance. It is possible that those patients who were presumed to be most compliant were more likely to be treated. However, the serum levels of immunosuppressive medication over the first month and their standard deviations were similar for treated and untreated patients, suggesting that they were at least equally compliant in taking their immunosuppressive medication. Moreover, all liver transplant patients undergo a thorough pretransplant evaluation and are only eligible for transplantation if they are compliant and adherent at that time. One of the strengths of our study is that our study cohort is larger than most studies among liver transplant patients with hepatitis C, which likely contributes to the reliability of the results.
Most of the patients in the untreated group did not have important histologic evidence of progressive fibrosis. Lack of progressive fibrosis is, in the absence of data that preemptive treatment is beneficial, a reasonable basis for not treating with antiviral therapy. Reasons for not initiating antiviral treatment among patients with more severe/progressive fibrosis needed to be examined in considerable detail. The higher, albeit not statistically significantly so, frequency of rejection in the untreated patients suggests that rejection may have been a factor in withholding antiviral therapy in some patients in the untreated group. Interestingly, this was something of a practice style issue. Some members of our group felt that peginterferon alpha should not be used in the context of a history or acute cellular rejection. The Food and Drug Administration (FDA)-approved labels for both IFN 2a and 2b cite (auto) immune hepatitis as a contraindication to using these agents and includes a black box warning regarding their use in immune hepatitis. Although many hepatologists ignore this warning and labeled contraindication in patients with recurrence of HCV following liver transplantation, it was not unreasonable to adhere to the label. The findings of this paper, including a lack of increased complications from clinically important new episodes rejection among treated patients and the observation that antiviral treatment is associated with improved outcomes adds weight to the argument for offering antiviral treatment even to patients with a history of mild acute cellular rejection. None of the patients in the antiviral treated group had allograft failure due to rejection. The untreated patients felt to have laboratory value parameters that contraindicated antiviral therapy all had hemoglobin, platelet, white cell count or creatinine values that were outside of the ranges cited on the FDA-approved labels for PEG-IFN as acceptable baseline values for initiation of therapy. This is interesting as, similar to for rejection, some of our hepatologists decided to ignore the labeling of PEG-IFN and/or RBV. The influence of practice style on antiviral treatment decision-making is apparent in the fact that hematologic and renal parameters were similar between the treated and untreated patients. It is also apparent that untreated patients included patients that were different from the treated group in other ways, including occurrence of recurrent alcoholism and general debility. The untreated group also inevitably self selected patients who refused antiviral therapy despite histologically progressive disease. The observed association of antiviral treatment improved posttransplant patient and graft survival should be interpreted carefully in light of these differences between treated and untreated patients.
The most certain conclusion from these results is that patients selected for and thought to be able to tolerate antiviral treatment have better patient and graft survival than recipients who are not selected for antiviral treatment. This observation raises the possibility that antiviral therapy might be the basis for improved outcomes in treated patients.
Randomized studies are needed to fully determine the impact of antiviral therapy on posttransplant patient and graft survival.
BJV received a stipend from the Netherlands Organization for Health Research and Development (ZonMw, project number 1900120283) for his work as a clinical research trainee. During his stay at the Mayo Clinic in Rochester, MN, he was also supported by a travel grant from the Trustfund of the Erasmus University Rotterdam, the Netherlands (grant number 97030.10/07.0275).
The sponsors did not have any influence on the design and conduct of the study; nor on collection, management, analysis and interpretation of the data; nor on preparation, review or approval of the manuscript.