A randomized, multicenter study comparing steroid-free immunosuppression and standard immunosuppression for liver transplant recipients with chronic hepatitis C


  • This study was supported by a grant from Roche.


This randomized, prospective, multicenter trial compared the safety and efficacy of steroid-free immunosuppression (IS) to the safety and efficacy of 2 standard IS regimens in patients undergoing transplantation for hepatitis C virus (HCV) infection. The outcome measures were acute cellular rejection (ACR), severe HCV recurrence, and survival. The patients were randomized (1:1:2) to tacrolimus (TAC) and corticosteroids (arm 1; n = 77), mycophenolate mofetil (MMF), TAC, and corticosteroids (arm 2; n = 72), or MMF, TAC, and daclizumab induction with no corticosteroids (arm 3; n = 146). In all, 295 HCV RNA–positive subjects were enrolled. At 2 years, there were no differences in ACR, HCV recurrence (biochemical evidence), patient survival, or graft survival rates. The side effects of IS did not differ, although there was a trend toward less diabetes in the steroid-free group. Liver biopsy samples revealed no significant differences in the proportions of patients in arms 1, 2, and 3 with advanced HCV recurrence (ie, an inflammation grade ≥ 3 and/or a fibrosis stage ≥ 2) in years 1 (48.2%, 50.4%, and 43.0%, respectively) and 2 (69.5%, 75.9%, and 68.1%, respectively). Although we have found that steroid-free IS is safe and effective for liver transplant recipients with chronic HCV, steroid sparing has no clear advantage in comparison with traditional IS. Liver Transpl, 2011. © 2011 AASLD.

It has been estimated that 3 to 4 million people in the United States are chronically infected with hepatitis C virus (HCV). Complications of this infection, including decompensated cirrhosis and hepatocellular carcinoma, are the main indications for liver transplantation in this country and account for approximately 40% of cases. In comparison with other patients, those undergoing transplantation with chronic HCV infections have worse long-term survival.1, 2 Although successful treatments with antivirals (pegylated interferons and ribavirin) prevent liver failure and reduce the risk of hepatocellular carcinoma,3 only half of treated patients respond to therapy.4 Furthermore, even fewer patients respond if the treatment is delayed until decompensation has ensued or transplantation has been performed.5 Thus, HCV recurrence remains a significant risk for patients who undergo liver transplantation, and the disease appears to progress more rapidly in this setting.6

It has been suggested that more intensive immunosuppression (IS)—particularly high-dose corticosteroids or lymphocyte-depleting therapy for the treatment of acute cellular rejection (ACR)—accelerates the progression of recurrent HCV. It has not, however, been determined whether the use of daily maintenance corticosteroid therapy (primarily in the first few months after transplantation) presents the same risk of aggressive recurrent HCV.7, 8 Because considerable controversy remains about the risks and benefits of maintenance corticosteroids after liver transplantation for HCV,9-11 we conducted a randomized, prospective, multicenter trial to compare the safety and efficacy of 3 IS regimens in HCV-positive liver transplant recipients and specifically to determine whether a steroid-free regimen is safe and effective.


ACR, acute cellular rejection; CI, confidence interval; CMV, cytomegalovirus; DAC, daclizumab; HCV, hepatitis C virus; IS, immunosuppression; MMF, mycophenolate mofetil; NS, not significant; OR, odds ratio; PRED, prednisone; RAI, rejection activity index; TAC, tacrolimus.


Study Design

This open-label, randomized, prospective, multicenter trial compared the safety and efficacy of 3 IS treatment regimens in HCV-positive liver transplant recipients. Because the main objective was to examine the feasibility of steroid avoidance, subjects were randomized with a weighted 1:1:2 ratio to 1 of 3 arms: arm 1 (n = 77), which received tacrolimus (TAC; Prograf, Astellas Pharma US, Inc., Deerfield, IL) and corticosteroids; arm 2 (n = 72), which received TAC, corticosteroids, and mycophenolate mofetil (MMF; CellCept, Hoffmann-La Roche, Inc., Nutley, NJ); or arm 3 (n = 146), which received daclizumab (DAC) induction (Zenapax, Hoffmann-La Roche), TAC, and MMF (a steroid-free regimen). The treatment regimens are summarized in Fig. 1. Specific doses are detailed in the legend for Fig. 1. Notably, subjects received DAC on days 0 (2 mg/kg), 3 (2 mg/kg), and 8 (1 mg/kg); this differed from the standard 2-dose regimen typically used at that time.12 The study endpoints included patient and graft survival, ACR, advanced HCV recurrence, and adverse events related to IS treatment.13

Figure 1.

Three IS protocols for liver transplant recipients with chronic HCV. TAC was administered at 0.08 to 0.12 mg/kg/day in 2 equal doses; the first dose was 20% of that amount within 72 hours, and the target trough levels were 10 to 15 ng/mL through week 6 and 5 to 12 ng/mL thereafter. As for corticosteroids, 500 to 1000 mg of methylprednisolone (or the equivalent of intravenous hydrocortisone or dexamethasone) was administered intraoperatively; the dosage was orally tapered to 10 mg/day by day 30 and to 5 mg/day by day 90. MMF was administered at 2 to 3 g/day in 2 divided doses, with the first dose given within 12 hours; the dosage was adjusted for adverse events. The dosage of DAC was 2 mg/kg on days 0 (within 12 hours) and 3 and 1 mg/kg on day 8.

The study group initially consisted of 312 adult patients (18 years old or older) with end-stage liver disease due to an HCV infection who were going to receive their first liver transplant. However, 17 patients who were positive for anti-HCV antibodies were found to be HCV RNA–negative at the time of transplantation and on subsequent retesting. They were excluded from further analysis; this left 295 evaluable subjects. Organ sources included both deceased and living donors. The exclusion criteria included the following: previous transplantation; hepatitis B core antibodies, hepatitis B surface antigen, or HCV antibodies in the donor; an ABO-incompatible donor; and anti–human immunodeficiency virus antibodies in the donor or the recipient. Patients who had contraindications for the use of calcineurin inhibitors or who were hospitalized in the intensive care unit at the time of transplantation also were not eligible. The protocol was approved by the institutional review boards of each of the 18 participating US centers. All subjects agreed to participate and signed the consent form.

Liver biopsy was performed 90 days, 1 year, and 2 years after transplantation. HCV recurrence was evaluated according to the Batts and Ludwig system for chronic hepatitis,14 and advanced recurrent HCV was defined for the purposes of this study as an inflammation grade ≥ 3 and/or a fibrosis stage ≥ 2. ACR was defined by the 1997 Banff criteria.15 Significant rejection was defined as a global Banff grade ≥ 2 and a rejection activity index (RAI) score ≥ 4. By necessity, day-to-day clinical decisions were based on liver biopsy interpretations by each study center's local pathologist. However, histological data presented in this report were derived from a review of biopsy slides by a single central hepatopathologist (G.J.N.). As previously reported,16 significant interobserver agreement was demonstrated between the center pathologists and the central pathologist (κ = 0.76 for recurrent HCV and κ = 0.62 for ACR, P < 0.001).

Treatment of ACR

Liver biopsy for cause was performed when consecutive test results demonstrated a 1.5-fold elevation in the serum aspartate aminotransferase or alanine aminotransferase level above the upper limit of the normal range or a serum bilirubin increase > 0.3 mg/dL. Because clear pathological differentiation between mild ACR and early recurrent HCV was difficult to achieve and because it was possible that the rejection treatment could trigger the activation of HCV, mild ACR (Banff grade 1 or an RAI score of 1-3) was treated with an increase in the TAC level to 12 to 15 ng/mL without a corticosteroid bolus and recycle. Moderate to severe ACR (a Banff grade ≥ 2 and an RAI score ≥ 4) was treated with a 1.0-g bolus of methylprednisolone, which was followed by a 6-day steroid taper of intravenous methylprednisolone (200, 160, 120, 80, 40, and 20 mg and then none) or oral prednisone. Although the 6-day steroid taper duration was suggested by the protocol, this was left to the investigators' discretion with the goal of returning patients to their original IS regimen as soon as possible. Corticosteroid-resistant ACR, which was defined as the histological persistence of ACR (unchanged or worse according to repeat liver biopsy) after steroid therapy, could be treated with anti-lymphocyte antibody therapy at the discretion of the study site.

Assessment and Treatment of HCV Recurrence

Protocol liver biopsy was recommended on days 0 (transplant date), 90, 365, and 730. A window of ±30 days was allowed for the 90-day biopsy, a window of ±60 days was allowed for the 1-year biopsy, and a window of ±90 days was allowed for the 2-year biopsy. Protocol biopsy samples were obtained for 83.4%, 75.4%, and 62.0% of the living subjects within these 3 windows. However, later protocol liver biopsy samples remained relevant only if the subjects had not already reached the study's histological endpoint of advanced hepatitis. Year 2 protocol biopsy samples were available for 82% of these patients. Biopsy compliance did not differ between the groups. Event-driven biopsy was performed as clinically indicated, and the results of these additional biopsy procedures were included in the analysis of the HCV recurrence rate. Advanced HCV recurrence was defined as the presence of histological evidence of grade 3 (moderate) or higher necrosis/inflammatory activity and/or the presence of stage 2 (periportal/septal) or higher fibrosis. Advanced HCV recurrence was required for the consideration of antiviral therapy, which was left to the discretion of the investigators at each study site. Viral replication, which was reported as the log HCV RNA level, was measured with a branched DNA assay (Chiron Corp., Emeryville, CA) because this assay offered the largest dynamic range for viral quantification that was available at the time of the study's initiation.17 HCV RNA was measured at the baseline (before transplantation) and 90, 365, and 730 days after transplantation. HCV RNA levels did not affect patient management decisions.

Safety Outcomes

New-onset diabetes mellitus was defined as the use of insulin for 30 or more days within the first posttransplant year among patients who did not have active diabetes before transplantation. New-onset hypertension was defined as the presentation of the condition at the 1-year follow-up visit in patients who did not have the condition at the baseline; new-onset hyperlipidemia was defined in the same way. The rates of infections and malignancies were also reported.

Infection Prophylaxis

Intravenous ganciclovir or oral valganciclovir was administered for a minimum of 7 days after transplantation as prophylaxis against cytomegalovirus (CMV) infections. After 7 days, all patients were managed according to the institutional protocol. No CMV prophylaxis was given to CMV-positive patients who received an organ from a seronegative donor. Prophylaxis for Pneumocystis carinii, fungal, and other bacterial infections was administered to patients according to the institutional protocol.

Statistical Methods

A sample size of 312 subjects was deemed sufficient to discover a difference in the primary endpoint between the groups (arms 1 and 2 versus arm 3) after 1 year. Assumptions were based on a 2-sided χ2 test with continuity correction, a type I error rate of 0.05, a type II error rate of 0.20 (power = 80%), a significance level of 0.013 (based on the Dunn-Sidak adjustment for multiple comparisons), a 10% dropout rate, and efficacies of 25%, 35%, and 50% in arms 1, 2, and 3, respectively.

With an intent-to-treat approach, all data were analyzed with SAS software (version 9.1.3, SAS Institute, Inc., Cary, NC) with a 5% significance level. Follow-up pairwise comparisons between groups were performed with the Bonferroni multiple-comparison adjustment. Continuous data were compared with the Kruskal-Wallis test and the Wilcoxon 2-sample test. Categorical data were compared with Fisher's exact test (2-sided) for 2 × 2 tables and with the likelihood ratio χ2 test for larger tables. A per-protocol analysis was also performed but is not included here because the number of assessable subjects at 2 years was reduced by 39%, and the outcomes were not different from those of the intent-to-treat analysis.

Patient and graft survival, the time to the first episode of ACR, clinically significant HCV recurrence, and other adverse events were estimated with the Kaplan-Meier product-limit method18 and were compared with the log-rank statistic.19 Outcomes were censored at the last biopsy for recurrence and at the last follow-up or death for other events. Death and retransplantation were events for graft survival. If the last biopsy occurred within a protocol biopsy visit window, the patient was censored on the last day of that window.

A Cox regression analysis was performed for advanced HCV recurrence. We looked at both the study endpoint (at least grade 3 or at least stage 2) and advanced fibrosis alone (Metavir stage 3 or 4). Because of the significant potential for confounding, we included all variables in the univariate analysis in the Cox analysis. This analysis looked at all subjects, regardless of the IS regimen, and at the treatment arms.


Study Groups

Two hundred ninety-five subjects at 18 liver transplant centers in the United States were randomized. Seventy-seven were randomized to arm 1 (TAC and prednisone), 72 were randomized to arm 2 (TAC, prednisone, and MMF), and 146 were randomized to arm 3 (TAC, DAC, and MMF). The recipient and donor characteristics for the 3 study groups are summarized in Table 1. Deceased donors accounted for most donor organs (88.1%). In comparison with recipients in arms 1 and 2, recipients in arm 3 received fewer organs from male donors (P < 0.05). The groups were comparable in all other categories.

Table 1. Patient and Transplant Characteristics
CategoryArm 1: TAC/PRED (n = 77)Arm 2: TAC/PRED/MMF (n = 72)Arm 3: TAC/DAC/MMF (n = 143)
  • *

    The data are presented as means and standard deviations.

  • P < 0.05 for arm 3 versus arms 1 and 2.

  • The data are presented as medians and ranges.

 Sex, male [n (%)]55 (71.4)54 (75.0)105 (71.9)
 Age (years)*51.3 ± 6.751.6 ± 7.851.3 ± 7.4
 Age > 55 years [n (%)]18 (23.4)16 (22.2)35 (24.0)
 Race [n (%)]   
  Caucasian56 (72.7)54 (75.0)101 (69.2)
  African American4 (5.2)2 (2.8)16 (11.0)
  Other17 (22.1)16 (22.2)29 (19.9)
 Weight (kg)*82.3 ± 16.586.8 ± 17.286.0 ± 18.5
 HCV genotype 1 [n (%)]53 (68.8)59 (81.9)105 (71.9)
 Model for End-Stage Liver Disease score*20.5 ± 6.621.6 ± 6.321.1 ± 7.4
 Source [n (%)]   
  Deceased donor70 (90.9)62 (86.1)128 (87.7)
  Living donor7 (9.1)10 (13.9)18 (12.3)
 Sex, male [n (%)]52 (67.5)50 (69.4)78 (53.4)
 Sex mismatch, yes [n (%)]29 (37.7)32 (44.4)57 (39.0)
 Cold ischemia time (hours)   
  Deceased donor7.2 (0.5-12.6)7.3 (0.9-18.8)7.1 (0.5-14.4)
  Living donor2.4 (1.5-3.5)2.3 (0.7-5.2)2.3 (0.5-4.5)
 Age (years)*39.8 ± 19.038.2 ± 15.341.5 ± 16.2
 Age > 55 years [n (%)]18 (23.4)9 (12.5)26 (18.2)

The mean follow-up was 20.9 ± 7.2 months (range = 0-32 months). Reasons that patients were withdrawn over the 2-year study period (other than death or graft loss) included withdrawal for safety and/or the need to change to another IS regimen (cyclosporine or rapamycin; 15 in arm 1, 10 in arm 2, and 27 in arm 3) and noncompliance or loss to follow-up (2 in arm 1, 4 in arm 2, and 9 in arm 3).


The IS doses are summarized in Table 2. Although the dose of TAC was lower during the first 90 days in arm 3 (P < 0.05), the mean trough levels were equivalent in the 3 arms. MMF was added as rescue therapy after ACR at some point for 12 subjects in arm 1 (see Table 2 for the time windows), mostly during the first posttransplant year. Similarly, 18 subjects in arm 3 received steroids after an episode of ACR. No differences were observed in the use of concurrent medications for antibacterial, antifungal, antiviral, lipid-lowering, and antihypertensive treatments (data not shown). There was a tendency for subjects in arms 1 and 2 to require insulin more often within the first postoperative month than subjects in arm 3 (16% and 14% versus 6% with 1 month and 23% and 20% versus 10% within 3 months), but this difference did not reach statistical significance (P = 0.09).

Table 2. IS Agents
 DaysArm 1: TAC/PREDArm 2: TAC/PRED/MMFArm 3: TAC/DAC/MMFP Value
  • NOTE: The data are presented as means and standard deviations.

  • *

    P < 0.05 for arms 1 and 2 versus arm 3.

  • MMF was used in 11, 12, and 8 subjects in arm 1 for 1 to 90, 91 to 365, and 366 to 730 days, respectively (see the text).

  • P < .001 for arm 1 versus arm 2 and/or arm 3.

  • §

    P < 0.05 for arm 2 versus arm 3.

  • PRED was used in 11, 18, and 10 subjects in arm 3 for 1 to 90, 91 to 365, and 366 to 730 days, respectively (see the text).

  • P < .001 for arm 3 versus arm 1 and/or arm 2.

Oral TAC dose (mg/day)1-907.7 ± 3.87.7 ± 3.76.7 ± 3.9*
91-3655.2 ± 2.85.7 ± 4.65.1 ± 3.1NS
366-7304.0 ± 2.44.2 ± 3.24.2 ± 2.7NS
TAC trough level (ng/mL)810.0 ± 3.711.1 ± 6.610.8 ± 5.0NS
3012.2 ± 5.811.1 ± 4.611.1 ± 4.7NS
909.5 ± 4.611.0 ± 5.59.6 ± 4.2NS
3658.4 ± 4.68.9 ± 3.29.0 ± 4.1NS
7307.3 ± 3.76.9 ± 3.27.5 ± 3.1NS
Oral MMF dose (mg/day)1-90218 ± 5631920 ± 3461811 ± 453§
91-365245 ± 6001603 ± 4971381 ± 570§
366-730156 ± 4801368 ± 5711301 ± 553
Oral steroid dose (mg/day)1-9012.1 ± 5.511.0 ± 3.22.0 ± 9.7
91-3656.0 ± 4.74.9 ± 3.52.1 ± 8.6
366-7306.2 ± 6.05.0 ± 4.50.8 ± 3.9


Forty-four episodes of moderate to severe ACR occurred during the study period. Most episodes (68.2%) occurred in the first 90 days after transplantation, and all but 1 episode (97.7%) occurred within the first postoperative year. Clinically significant ACR occurred during the first 90 days in 14.3%, 6.9%, and 9.6% of the subjects in arms 1, 2, and 3, respectively [not significant (NS)]. These episodes of ACR were moderate (Banff grade 2) in 61.5%, 80.0%, and 75.0% of the subjects in arms 1, 2, and 3, respectively and were severe (Banff grade 3) in the rest. By day 730, clinically significant ACR had occurred in 14.3%, 12.5%, and 13.7% of the patients in the 3 arms. None of these differences between the groups were significant. Ductopenic rejection occurred in 4 subjects (2, 1, and 1 in arms 1, 2, and 3, respectively).

HCV Recurrence

HCV RNA was detectable at the time of study entry in all subjects (17 anti-HCV–positive cases in whom the virus was undetectable were excluded). HCV RNA spontaneously cleared in a single patient (0.3%); for this subject, HCV RNA remained undetectable without treatment 90, 365, and 730 days after transplantation. The mean HCV RNA levels were similar in the 3 IS arms on days 0, 90, 365, and 730. However, most viral levels exceeded the upper limitation of quantitation for the assay, so it is possible that differences might not have been detected. Biochemical hepatitis was defined as the elevation of at least 1 of the 3 following tests: alanine aminotransferase (>100 IU/mL), aspartate aminotransferase (>100 IU/mL), and total bilirubin (>1.5 mg/dL). It was observed by day 90 in 43.8%, 39.1%, and 27.7% of the subjects in arms 1, 2, and 3, respectively (P < 0.05 for arm 3 versus arms 1 and 2). Biochemical abnormalities were present in 49.2%, 42.4%, and 38.0% of the subjects on day 365 (NS) and in 33.9%, 34.0%, and 29.9% of the subjects on day 730 (NS).

Clinically significant HCV (defined by protocol as an inflammation grade ≥ 3 and/or a fibrosis stage ≥ 2) occurred in 48.2%, 50.4%, and 43.0% of the subjects in arms 1, 2, and 3, respectively, within 1 year and in 69.5%, 75.9%, and 68.1% within 2 years (Fig. 2A). Grade 3 or higher inflammation was present in the first-year liver biopsy samples of 18.8%, 16.6%, and 23.2% of the patients in arms 1, 2, and 3, respectively. By year 2, 35.4%, 24.3%, and 33.9% of the patients had a similar inflammation grade (Fig. 2B). Stage 2 or higher fibrosis was found in 46.8%, 50.4%, and 43.1% of the patients in arms 1, 2, and 3, respectively, by year 1 and in 66.3%, 75.9%, and 68.2% by year 2 (Fig. 2C). None of these differences were statistically significant. Fibrosing cholestatic hepatitis occurred in 7 subjects (2, 2, and 3 in arms 1, 2, and 3, respectively).

Figure 2.

Histological progression of recurrent HCV according to the treatment arm: (A) the probability of histological progression according to the study endpoint (grade ≥ 3 and/or stage ≥ 2; P = 0.60), (B) the probability of grade 3 or higher inflammation alone (P = 0.31), and (C) the probability of stage 2 or higher fibrosis alone (P = 0.43).

Ninety-two patients in the 3 arms (31.2%) received antiviral therapy [11 received pegylated interferon alone, and 81 received pegylated interferon and ribavirin (51.1% in year 1 and 32.6% in year 2)]. Treatment was more common in patients with stage 3 or 4 fibrosis versus patients with lower stages of fibrosis (P < 0.001 at years 1 and 2). However, because the treatment was not part of this protocol, we do not have details for the treatment doses, duration, or outcomes. Nonetheless, the use of the treatment was similar in the 3 arms and did not affect patient or graft survival during the 2-year study period (data not shown).

Risk Factors for Advanced Recurrent HCV

Advanced hepatitis (at least grade 3 and/or stage 2) is an ominous prognostic indicator for patients with HCV recurrence after liver transplantation and is typically considered to be the indication for the consideration of antiviral therapy (as it was in this study). A Cox regression analysis was used to assess factors independently associated with the development of advanced hepatitis after transplantation. A univariate analysis found only the cumulative steroid dose and early evidence of histological injury to be associated with the eventual achievement of this endpoint. On the other hand, a stepwise multivariate Cox regression showed that male donors, male recipients, living donors, cold ischemia times, ACR, and early histological recurrence (grade 2 inflammation on the 90-day liver biopsy sample or stage 1 fibrosis on the 1-year biopsy) were associated with the development of advanced hepatitis (Table 3).

Table 3. Factors Associated With the Histological Study Endpoint (Inflammation Grade ≥ 3 and/or Fibrosis Stage ≥ 2)
Observations (n)VariableUnivariate AnalysisMultivariate Analysis
P ValueOR (95% CI)P ValueOR (95% CI)
  1. NOTE: The analysis was performed with Cox regression. 10 patients are excluded because they had no biopsies. Donor age and sex were missing in 3 additional patients, and donor sex was missing in two additional patients.

280Donor, male0.690.94 (0.71-1.26)0.040.7 (0.49-0.98)
285Donor, black0.470.85 (0.55-1.32)  
283Donor age (years)0.171.01 (1-1.02)  
285Recipient, male0.961.01 (0.73-1.39)0.0012 (1.32-3.01)
285Recipient, black0.081.56 (0.96-2.54)  
285Recipient age (years)0.831 (0.98-1.02)  
285Living donor0.271.29 (0.82-2.04)0.0022.79 (1.48-5.25)
285CMV0.171.45 (0.86-2.46)  
243Cold ischemia time (hours)0.171.04 (0.98-1.09)0.031.07 (1.01-1.14)
285Arm 10.950.99 (0.71-1.38)  
285Arm 20.331.17 (0.85-1.62)  
285Arm 30.440.89 (0.67-1.19)  
285CellCept average dose until recurrence0.521 (1-1)  
285Steroid average dose until recurrence0.031.01 (1-1.02)  
285Grade 1 by day 90<.0013.18 (2.29-4.42)  
285Grade 2 by day 90<.0014.59 (3.16-6.68)<.0013.33 (2.11-5.26)
285Grade 1 by day 365<.0013.6 (2.56-5.05)  
285Grade 2 by day 365<.0014 (2.91-5.48)  
285Stage 1 by day 90<.0013.6 (2.55-5.07)  
285Stage 1 by day 365<.0014.46 (3.21-6.21)<.0015.19 (3.37-8)
285ACR during first 2 years0.421.18 (0.79-1.76)  
285ACR before recurrence0.161.35 (0.89-2.03)0.041.67 (1.03-2.7)
285OKT3/thymoglobulin for ACR before recurrence0.411.45 (0.59-3.53)  

Advanced fibrosis (stage 3 or 4) is an important posttransplant milestone because it typically indicates a high risk of subsequent hepatic decompensation. Thus, we also looked at advanced fibrosis alone (without any consideration of the grade) as an endpoint (Table 4). A univariate analysis showed that the donor age and early histological recurrence were associated with the development of advanced fibrosis. According to a multivariate Cox analysis, the donor age and grade 2 inflammation on the day 90 or year 1 liver biopsy sample were associated with the subsequent development of bridging fibrosis or cirrhosis; the administration of OKT3 or thymoglobulin approached significance.

Table 4. Factors Associated With Advanced Fibrosis (Metavir Stage 3 or 4) by Year 2
Observations (n)VariableUnivariate AnalysisMultivariate Analysis
P ValueOR (95% CI)P ValueOR (95% CI)
  1. NOTE: The analysis was performed with Cox regression. 10 patients are excluded because they had no biopsies. Donor age and sex were missing in 3 additional patients, and donor sex was missing in 2 additional patients.

280Donor, male0.221.33 (0.84-2.11)  
285Donor, black0.140.54 (0.23-1.23)  
283Donor age (years)<.0011.03 (1.02-1.04)<.0011.03 (1.01-1.04)
285Recipient, male0.391.25 (0.75-2.08)  
285Recipient, black0.281.46 (0.73-2.91)  
285Recipient age (years)0.250.98 (0.95-1.01)  
285Living donor0.210.59 (0.26-1.35)  
285CMV0.151.72 (0.83-3.56)  
243Cold ischemia time (hours)0.321.04 (0.96-1.12)  
285Arm 10.701.1 (0.68-1.77)  
285Arm 20.991 (0.61-1.63)  
285Arm 30.750.93 (0.61-1.43)  
285CellCept average dose until recurrence0.911 (1-1)  
285Steroid average dose until recurrence0.391.01 (0.99-1.02)  
285Grade 1 by day 90<.0012.62 (1.65-4.14)  
285Grade 2 by day 90<.0013.82 (2.36-6.19)<.0012.87 (1.63-5.07)
285Grade 1 by day 365<.0012.68 (1.57-4.56)  
285Grade 2 by day 365<.0013.92 (2.3-6.67)<.0013.6 (1.79-7.24)
285Stage 1 by day 90<.0012.86 (1.78-4.59)  
285Stage 1 by day 365<.0013.45 (2.02-5.87)  
285ACR during first 2 years0.581.19 (0.64-2.19)  
285ACR before recurrence0.510.81 (0.44-1.5)  
285OKT3/thymoglobulin for ACR before recurrence0.871.12 (0.28-4.57)0.083.72 (0.86-16.2)

Safety Data

Table 5 shows the rates of infection, new-onset diabetes, new-onset hypertension, new-onset hyperlipidemia, and de novo malignancies. Although infections, diabetes, hypertension, and hyperlipidemia were quite common, there were no statistically significant differences between the study arms. There was, however, a trend toward less diabetes in the steroid-free IS group (P = 0.09). De novo malignancies were uncommon and also did not differ between the groups. Most posttransplant malignancies during the 2-year follow-up period were related to the recurrence or metastasis of hepatocellular carcinoma. CMV infections occurred in 18 cases (4, 6, and 8 in arms 1, 2, and 3, respectively).

Table 5. Likelihood of Complications by Study Year 2
ComplicationArm 1: TAC/PRED (%)Arm 2: TAC/PRED/MMF (%)Arm 3: TAC/DAC/MMF (%)
  1. NOTE: None of the differences between groups were statistically significant.

Infection requiring intravenous antibiotics/hospitalization41.636.134.8
New-onset insulin-dependent diabetes29.832.818.6
New-onset hypertension79.070.566.5
New-onset hyperlipidemia56.141.551.6
De novo malignancy: not hepatocellular carcinoma2.64.21.4
De novo malignancy: hepatocellular carcinoma6.56.92.0


The 1- and 2-year patient and graft survival rates were similar in the 3 arms. At the end of the second year, the patient survival rates in arms 1, 2, and 3 were 83.8%, 81.0%, and 86.1%, respectively (Fig. 3A). The 2-year graft survival rates were 79.1%, 79.8%, and 85.1%, respectively (Fig. 3B). The main causes of death included sepsis [n = 14 (3 in arm 1, 3 in arm 2, and 8 in arm 3)], malignancies [n = 8 (2, 5, and 1)], HCV recurrence [n = 5 (4, 1, and 0)], multiorgan failure without sepsis [n = 4 (2, 2, and 0)], intracerebral hemorrhaging [n = 2 (1, 1, and 0)], cardiopulmonary events [n = 3 (0, 1, and 2)], and unknown or other causes [n = 8 (0, 0, and 8)]. No statistically significant differences were observed between the arms.

Figure 3.

Kaplan-Meier graphs of (A) patient survival and (B) graft survival.


The outcomes of liver transplantation for HCV-infected recipients are inferior to the outcomes for patients undergoing transplantation for other indications.1 HCV recurrence that leads to graft failure is a major cause of posttransplant graft loss and death in HCV patients. Because antiviral therapy is poorly tolerated in the posttransplant setting and we have not yet reached the point at which it can predictably alter the course of recurrence, attention has mostly been focused on identifying and avoiding donor and host factors that are associated with progression to advanced-stage HCV. Older donor age,20-22 high HCV RNA levels,23, 24 CMV infections,25 lymphocyte-depleting IS agents, corticosteroids, and other factors remain topics of controversy. We confirmed the results of other studies with respect to the detrimental effects of donor age20-22 and sex20 on HCV progression. We found that an organ from a male donor was a risk factor for advanced hepatitis, although it was not for fibrosis alone. This was consistent with our observation that fibrosis progression was similar in the 3 treatment arms, despite the presence of fewer male donors in the steroid-free arm. Although it is unlikely, it is conceivable that this may have masked an otherwise higher rate of progression in this group. Nevertheless, none of the other aforementioned factors can be controlled within the current organ allocation system. What stood out in our analysis was that early histological recurrence, even after just 3 months, was highly associated with subsequent progression. This needs to be considered when antiviral treatment is being timed.

Perhaps the one factor that can be modified to reduce the speed and severity of recurrence is the form of IS therapy. It remains unclear whether the choice of the calcineurin inhibitor influences the rapidity of HCV recurrence after transplantation. Although cyclosporine and other cyclophilin inhibitors decrease HCV replication in vitro,26-28 cyclosporine does not confer a benefit in vivo. Indeed, TAC has been associated with better patient and graft survival.27 All patients in this study received TAC. MMF also does not appear to have a significant antiviral effect despite early reports to the contrary.29, 30 We could find no evidence that MMF influenced HCV progression in this study. Finally, it could be argued that it would have been more appropriate to use a lower dose of TAC when it was combined with MMF in arms 2 and 3. Although this has been debated for years, there is no consensus for lowering the TAC levels when 3 drugs are used. Thus, we made the target drug levels the same for all groups.

The use of corticosteroids after liver transplantation has been a topic of debate for many years. Several studies have suggested that steroid avoidance is associated with lower rates of bacterial, fungal, and CMV infections, diabetes, hypertension, and hyperlipidemia, but this is accomplished at the cost of increased rejection.31 The risk of rejection is reduced when steroids are replaced by another IS agent, such as an interleukin-2 receptor–inhibiting antibody.11 We found no advantage with steroid avoidance in our study with respect to the posttransplant occurrence of infections, new-onset diabetes, hypertension, or hyperlipidemia. Thus, the remaining unknown is the way in which steroid avoidance affects the course of recurrent HCV infections. Steroids might accelerate the progression by driving viral replication and facilitating CMV coinfections. Alternatively, they might reduce the progression by diminishing the risk for ACR. Although steroids do not increase HCV replication in vitro,32 it has been suggested that steroid maintenance and repeated steroid boluses for treating rejection may be associated with more rapid progression of HCV recurrence and fibrosis.33 Small studies comparing steroid-free IS to standard therapy have shown conflicting results.11, 34-36 Our study, which is the largest to look at the impact of steroid avoidance on HCV recurrence, found no differences in the rates of HCV progression with steroid avoidance and standard IS regimens using steroids. However, our study observed subjects for only 2 years after transplantation, and some patients refused liver biopsy in the second year. Thus, we cannot dismiss the possibility that differences might have been observed if all subjects had been biopsied or the follow-up had been longer. We also cannot rule out the possibility that rejection-driven changes in IS, such as steroid boluses or the addition of MMF, might have masked differences between the randomized treatment regimens. Because of the similarities in the recurrence severity and rejection rates in the 3 groups, however, we feel that neither of these 2 possibilities is likely.

In summary, although we have shown no distinct advantage of steroid avoidance, we have found that steroid-free IS is safe in HCV-positive liver transplant recipients and does not increase the risk of rejection or fibrosis progression. Thus, this might be the preferred option in certain patients, such as those who already have obesity, diabetes, hypertension, or hyperlipidemia. Obesity and diabetes are associated with a more rapid progression of liver injury from HCV infection,37, 38 so steroid avoidance might confer a long-term benefit in this group that we were unable to uncover in this 2-year study.

Definitions of HCV recurrence in the posttransplant setting have been inconsistent and confusing. Although nearly all HCV-infected patients experience viral recurrence, not all experience a rapid progression of liver disease. Therefore, it is more appropriate to focus our attention on biochemical and histological recurrence, as we have in this study. We observed no differences between the 3 treatment arms with respect to the posttransplant levels of viremia or biochemical markers of hepatitis activity. It is important to note, however, that biopsy evidence of progressive HCV often occurred in the absence of biochemical abnormalities. Indeed, although only 29.9% to 34.0% had aspartate aminotransferase, alanine aminotransferase, or bilirubin elevations at year 2, 68.1% to 75.9% met the histological criteria for advanced recurrent hepatitis. This emphasizes the importance of regular surveillance liver biopsy in patients who undergo liver transplantation for HCV.

Until we have safer and more effective antiviral therapy for recurrent HCV in liver transplant recipients, the only viable options for reducing the rate of fibrosis progression in HCV-positive recipients may be (1) avoiding the use of older donors and (2) controlling comorbid host factors that may accelerate progression (eg, obesity, diabetes, and alcohol intake). Recent publications reporting that interleukin-28B polymorphisms might influence the rate of HCV progression after transplantation require confirmation.39, 40 We certainly require a better understanding of what makes HCV progress more rapidly in the posttransplant setting, and we look forward to the availability of more effective antiviral agents, which now seem within reach.