Hepatitis C virus (HCV)–related cirrhosis is the most common indication for liver transplantation. However, the recurrence of HCV infections is almost universal, and the rate of fibrosis progression is accelerated so that 20% to 40% of recipients progress to allograft cirrhosis within 5 years.[1-4] The severity of HCV recurrence may be altered by the amount and type of immunosuppression. The treatment of acute cellular rejection (ACR) with repeated bolus injections of a corticosteroid (ST) is generally considered to have negative effects on recurrent HCV.[6, 7] However, no convincing data have shown that the use of daily maintenance ST therapy is associated with aggressive recurrent HCV. Although several clinical trials have examined the safety and efficacy of steroid avoidance,[8-14] whether steroid-free regimens would prove less deleterious for recurrent HCV remains controversial.
In Japan, too, HCV-related cirrhosis and hepatocellular carcinoma (HCC) are the most prevalent liver diseases, and living donor liver transplantation (LDLT) has become a treatment option for patients with these diseases. We have already reported the clinical outcomes of LDLT for 91 HCV-positive recipients. In this early experience, the survival rates and rates of progression to severe disease due to HCV recurrence were comparable to those for recipients of deceased donor liver transplantation described in the literature. Our standard immunosuppression protocol has consisted of tacrolimus (TAC) and a low-dose ST, with the administration of steroids gradually tapered and terminated at the end of the third month. Against this background, we conducted a prospective, randomized, multicenter trial to evaluate the effects of a steroid-avoidance regimen on HCV recurrence after LDLT. Because TAC monotherapy was considered weak as an induction immunosuppressive regimen, we adopted mycophenolate mofetil (MMF) as a substitute for steroids in the steroid-avoidance study arm. The aim of this prospective, randomized study was to compare the efficacy and safety of TAC/MMF and conventional TAC/ST regimens in HCV-positive LDLT recipients.
PATIENTS AND METHODS
In January 2004, this open-label, randomized, prospective, multicenter trial was started at 6 LDLT centers in Japan (ie, the hospitals of Kyoto University, Kumamoto University, Mie University, Nagoya University, Osaka City University, and Ehime University). This randomized clinical trial was registered at ClinicalTrials.gov (NCT00469131). Patients > 18 years old who were scheduled to undergo LDLT and suffered from HCV-related liver cirrhosis with positive serum results for HCV antibodies were eligible for inclusion in the study. The exclusion criteria were (1) a history of any organ transplantation or retransplantation, (2) ABO blood type incompatibility, (3) positive findings for hepatitis B surface antigen, (4) a small-for-size graft (graft/recipient body weight ratio < 0.8%), (5) renal dysfunction (serum creatinine level > 2 mg/dL), and (6) a platelet count < 3 × 104/mm3 or a white blood cell count < 1000/mm3. Patients with HCC were not excluded. In the selection criteria for HCC patients, no restrictions were placed on the number or size of tumors, but patients with extrahepatic metastasis or macroscopic venous invasion on preoperative imaging were excluded. HCC was classified according to the tumor-node-metastasis staging criteria.
All subjects provided written, informed consent to participate. Eligible patients were centrally registered and randomized with a Web-based enrollment system at the Translational Research Informatics Center. Randomization was stratified by institutions and HCC stage (stage III or other) and was performed in a 1:1 ratio to an ST group, which received TAC (Prograf, Astellas Pharma, Tokyo, Japan) and an ST, and an MMF group, which received TAC and MMF (CellCept, Hoffmann-La Roche, Basel, Switzerland). The protocol of this study was approved by the medical ethics committee of each participating university.
Histological recurrence of HCV (fibrosis stage ≥ F1 according to the METAVIR score), biopsy-proven acute rejection (BPAR) resistant to 2 sets of steroid pulse therapy, HCC recurrence, retransplantation, and patient death were defined as events, and the primary endpoint of the study was event-free survival. Secondary endpoints included patient survival, HCV viral loads, histological HCV recurrence, BPAR requiring steroid pulse therapy, and chronic rejection.
In both groups, the administration of TAC was started through a nasogastric tube within the first 12 hours after transplantation. The target whole-blood trough level for TAC was 10 to 15 ng/mL during the first 2 weeks, approximately 10 ng/mL thereafter, and 5 to 8 ng/mL from the second month. In the ST group, intravenous methylprednisolone was initiated at 10 mg/kg before graft reperfusion and was then tapered from 1 mg/kg/day on days 1 to 3 to 0.5 mg/kg/day on days 4 to 6 and to 0.3 mg/kg/day on day 7. Subsequently, oral prednisone was continued at 0.3 mg/kg/day until the end of the first month, and this was followed by 0.1 mg/kg/day until the end of the third month. After this time, steroid administration was terminated. In the MMF group (steroid avoidance), a steroid bolus injection immediately before reperfusion was also avoided. MMF was initiated through the nasogastric tube at a starting dose of 10 to 15 mg/kg on day 1, which was gradually increased to a target dose of 30 mg/kg, and this was continued for 6 months.
Assessment and Treatment of HCV Recurrence
The serum HCV RNA load was evaluated with polymerase chain reaction and an Amplicor HCV assay (Cobas Amplicor HCV Monitor, Roche Molecular Systems, Pleasanton, CA). HCV core protein was measured in serum as another index for the viral load with a previously reported enzyme immunoassay method. These viral loads were determined before LDLT and 1, 7, 14, and 28 days and 3, 6, and 12 months after LDLT.
Protocol liver biopsy was recommended 3, 6, and 12 months after LDLT and annually thereafter. Event-driven biopsy was performed as clinically indicated. All biopsy samples were evaluated by liver histopathologists at each center. For the diagnosis of HCV recurrence, the necroinflammatory activity (A0-A3) and the fibrosis stage (F0-F4) were assessed with the METAVIR score. HCV recurrence with a necroinflammatory activity classification ≥ A2 or a fibrosis stage ≥ F1 was considered for antiviral treatment with interferon (IFN) and ribavirin. From 2005 onward, splenectomy was performed during the recipient's transplant operation to increase the platelets and white blood cell counts suppressed by hypersplenism and to enhance the tolerability of IFN therapy.
Assessment and Treatment of Rejection
ACR was confirmed by liver biopsy and assessed according to the Banff criteria. In principle, mild ACR (Banff grade I or rejection activity index score of 1-3) was treated with an increase in the TAC level. Mild to severe ACR (Banff grade II/III or rejection activity index score ≥ 4) was treated with steroid pulse therapy (10 mg/kg bolus of methylprednisolone for 3 days) followed by steroid tapering, if necessary, in both groups.
The 1-year event-free survival rate for the TAC/steroid arm (the ST group) was estimated to be 20%; this was determined by an examination of historical data (n = 30) from Kyoto University Hospital. The 1-year event-free survival rate for the TAC/MMF arm (the MMF group) was expected to be 40%. Under the assumption of a 1-sided significance level of 5% and a power of 80%, the required sample size was calculated to be 60 patients per arm on the basis of the log-rank test.
Baseline and laboratory test data are summarized as medians and ranges. Categorical variables were compared with the χ2 test or Fisher's exact test. Continuous variables were compared with the Wilcoxon rank-sum test. Cumulative probability curves of events were calculated with the Kaplan-Meier method. When we evaluated the time to reach fibrosis stage 1 or 2 according to liver biopsy, outcomes were censored at the time of last biopsy or death for patients who did not reach each fibrosis stage. Differences between these curves were compared with the log-rank test. All data were analyzed with an intent-to-treat approach, and P < 0.05 was considered statistically significant.
Although the planned sample size was 120 patients, the enrollment period was ended in August 2010 after the enrollment of 79 patients because of shrinkage of the study population. Thirty-seven subjects were randomized to the TAC/ST arm (the ST group), and 42 subjects were randomized to the TAC/MMF arm (MMF group). Two patients in the ST group and 2 in the MMF group were excluded from the study after randomization because of protocol violations, so 35 patients in the ST group and 40 in the MMF group formed the cohort of the present analysis.
Recipient and donor characteristics and operative data are summarized in Table 1. No significant differences in these profiles were evident between the ST and MMF groups. One subject in the ST group withdrew consent to participate in the study on postoperative day 37. Patient outcomes were finally collected at the end of August 2011. The median follow-up was 55 months (range = 1-89 months).
Table 1. Patient, Transplant, and Immunosuppression Profiles
ST Group (n = 35)
MMF Group (n = 40)
NOTE: No significant differences in any variables (except for immunosuppression) were seen between the ST and MMF groups.
The data are presented as medians and ranges.
Total number of days of administration during the first year.
The rates of event-free survival (the primary endpoint) 1, 3, and 5 years after LDLT were 38.2%, 11.8%, and 5.9%, respectively, for the ST group and 25.0%, 17.5%, and 14.6%, respectively, for the MMF group (Fig. 1). No significant differences in event-free survival were evident between the groups (1-sided P = 0.45).
Early mortality (within 4 months after LDLT) occurred for 2 recipients in the ST group and for 3 recipients in the MMF group. The causes of death were graft dysfunction (n = 1) and cerebral bleeding (n = 1) in the ST group and cerebral bleeding (n = 1), bacterial pneumonia (n = 1), and sepsis (n = 1) in the MMF group. In the later period, 3 patients from the ST group died of HCC recurrence (n = 1), malignant lymphoma (n = 1), and graft failure due to a biliary stricture (n = 1), whereas 7 patients from the MMF group died of chronic rejection (n = 1), HCC recurrence (n = 1), HCV recurrence (n = 1), and other causes (n = 4). No recipients in either group underwent retransplantation. The 1-, 3-, and 5-year overall survival rates were 94.1%, 87.6%, and 82.7%, respectively, for the ST group and 92.5%, 84.5%, and 81.0%, respectively, for the MMF group (P = 0.28; Fig. 2).
HCV Histological Recurrence and Antiviral Therapy
During follow-up, histological HCV recurrence with a fibrosis stage ≥ F1 was confirmed for 27 patients in the ST group and for 35 patients in the MMF group. Cumulative recurrence rates with a fibrosis stage ≥ F1 (deaths censored) at 1, 2, and 3 years were 59.4%, 78.9%, and 85.9%, respectively, for the ST group and 74.2%, 81.9%, and 81.9%, respectively, for the MMF group (P = 0.57; Fig. 3). Recurrence with a fibrosis stage ≥ F2 was confirmed for 9 patients in the ST group and for 13 patients in the MMF group. Cumulative recurrence rates with a fibrosis stage ≥ F2 at 1, 3, and 5 years were 6.3%, 19.0%, and 24.4%, respectively, for the ST group and 16.7%, 28.4%, and 31.6%, respectively, for the MMF group (P = 0.46; Fig. 4).
Within 1 year after LDLT, 26 patients in the ST group and 34 patients in the MMF group who had developed histological recurrence of HCV were considered for antiviral therapy. Among these, 18 patients (69.2%) in the ST group and 21 patients (61.8%) in the MMF group started IFN/ribavirin treatment within 1 year (P = 0.60). As a result, a sustained virological response was achieved for 8 of 18 patients in the ST group (44.4%) and for 14 of 21 patients (66.7%) in the MMF group (P = 0.16).
HCV Viral Load
Changes in HCV RNA and HCV core protein loads are shown in Fig. 5. On day 28, HCV RNA levels were significantly lower for the MMF group (median = 250 KIU/mL, range = 0-17,660 KIU/mL) versus the ST group (median = 1700 KIU/mL, range = 0.02-13,000 KIU/mL, P < 0.05; Fig. 5A). HCV core protein levels were also significantly lower for the MMF group (median = 719 fmol/L, range = 20-26,200 fmol/L) versus the ST group (median = 7200 fmol/L, range = 20-32,000 fmol/L; P < 0.01; Fig. 5B). Thereafter, the viral loads were similar for the groups.
Immunosuppression and Rejection
The median total durations of the administration of steroids (in the ST group) and MMF (in the MMF group) were 92 and 175 days, respectively (Table 1). TAC whole-blood trough levels at 3, 6, and 12 months did not differ significantly between the ST and MMF groups. The addition of MMF or an ST was decided on the basis of the judgment of the investigators for 1 patient in the ST group and for 3 patients in the MMF group, respectively. BPAR requiring treatment with an ST bolus injection occurred in 4 patients from the ST group and in 13 patients from the MMF group (P = 0.051). According to the Banff criteria, ACR was mild for 2, moderate for 1, and severe for 1 in the ST group and mild for 3 and moderate for 10 in the MMF group. All episodes were observed within the first 3 months after LDLT in both groups. BPAR resistant to 2 sets of ST bolus treatment occurred in only 1 patient from the MMF group. Chronic rejection was diagnosed in only 1 patient, again from the MMF group.
Seventeen of the 75 patients experienced BPAR treated with ST pulse treatment. The overall survival rate for the 17 patients with BPAR was significantly lower than the rate for the 58 patients without BPAR (68.2% versus 90.9% at 3 years and 51.9% versus 90.9% at 5 years, P < 0.001; Fig. 6). However, histological HCV recurrence rates did not differ between patients with BPAR and patients without BPAR. Rates of progression to a fibrosis stage ≥ F1 at 1 year were 78.5% and 64.9%, respectively (P = 0.66), and rates of progression to a fibrosis stage ≥ F2 at 3 years were 30.8% and 22.3%, respectively (P = 0.31).
As for infectious complications, bacterial and fungal infections such as sepsis and pneumonia were diagnosed for 1 patient in the ST group and for 4 patients in the MMF group. A cytomegalovirus infection was diagnosed for 2 patients in the ST group and for 1 patient in the MMF group. Clinical test results with respect to metabolic complications are summarized in Table 2. No significant differences in these data were apparent between the groups.
Table 2. Clinical Test Results
NOTE: The data are presented as medians and ranges. No significant differences in any variables were seen between the ST and MMF groups.
Systolic blood pressure (mm Hg)
Diastolic blood pressure (mm Hg)
Fasting blood glucose level (mg/dL)
Serum creatinine level (mg/dL)
Several risk factors have been implicated in the frequency and severity of recurrent HCV after liver transplantation.[6, 24] Many studies have found deleterious effects of potent immunosuppression (eg, high numbers of bolus injections of methylprednisolone, the use of anti-lymphocyte preparations, and high total cumulative doses of steroids). On the other hand, it had been not clarified as of 2004 (when the present study was started) whether the use of daily maintenance ST therapy was associated with aggressive recurrent HCV. Although there have been, to the best of our knowledge, 7 prospective, randomized studies reported as full articles since then,[8-14] the issue of steroid-free regimens remains controversial. All 7 studies included a calcineurin inhibitor as part of the immunosuppressive regimen (cyclosporin A in 2 studies and TAC in 5 studies). In 4 studies,[8, 9, 11, 12] the ST was simply removed from the regimen of the control arm, whereas the other 3 studies replaced the ST with daclizumab.[10, 13, 14] Although each steroid-free protocol was found to be safe and feasible in all studies, the results were divergent with respect to the effects on HCV recurrence. Lladó et al. reported favorable results for a steroid-free group in terms of short-term histological HCV recurrence. However, Manousou et al. reported contradictory findings. Moreover, the remaining 5 studies found no significant differences between steroid-free and control groups.
The present investigation was a randomized comparative study conducted among LDLT recipients to investigate the impact of a steroid-avoidance immunosuppression protocol on HCV recurrence. In the steroid-avoidance arm, steroids were completely removed; this included a methylprednisolone bolus injection immediately before reperfusion. The safety and efficacy of omitting steroids during and after liver transplantation have been demonstrated in clinical studies. To really evaluate the steroid-free effects, the regimen of the study arm should simply remove STs from the regimen of the control arm, as in the previous 4 studies.[8, 9, 11, 12] However, because we thought that TAC monotherapy would prove weak as an induction immunosuppressive regimen, we adopted MMF as a substitute for steroids in the steroid-avoidance arm of the study. Because MMF, an inositol monophosphate dehydrogenase inhibitor like ribavirin, has been proposed to exert antiviral effects, we also expected additional anti-HCV effects from the use of MMF in the study arm. Several studies have evaluated the anti-HCV effects of MMF in clinical liver transplantation settings.[27-30] Although antiviral properties have remained unconfirmed in most studies,[27-29] Fasola et al. reported a delay in the recurrence of severe hepatitis among recipients treated with MMF.
Consequently, the present study compared immunosuppressive regimens using an ST or MMF in combination with TAC. The primary endpoint was defined as event-free survival. For HCV-positive recipients, optimal immunosuppression is required to both prevent rejection and control the recurrence of HCV. Less potent immunosuppression is considered to reduce the severity of HCV recurrence. On the other hand, the risk of rejection will be increased, and this may lead to intractable rejection requiring repeated steroid boluses and thus adversely affect the risk of HCV recurrence. The risk of HCC recurrence and bacterial or viral infectious complications, which often cause fatal outcomes, are also affected by the strength of immunosuppression.
Unfortunately, the present study failed to demonstrate any favorable results for the MMF group. Several factors should be considered in interpreting the present results. First of all, the final sample size (n = 75) was smaller than planned (n = 120). One reason was the recent trend toward cases using a small-for-size or ABO-incompatible graft, both of which were excluded from the present study. At Kyoto University, from the perspective of donor safety and with the development of strategies for small-for-size syndrome, the algorithm for graft type selection has been modified to primarily consider left lobe donation. Between 2006 and 2008, the selection of a small-for-size graft (graft/recipient body weight ratio < 0.8%) increased to 23.9% of all adult cases. Similarly, with improvements in ABO-incompatible protocols such as preoperative rituximab, ABO-incompatible cases have increased to more than 20%. The post hoc type II error based on the completed sample size of 75 patients and prespecified settings was elevated by 38%. However, the predictive power, representing the probability of obtaining a significant result after completion of the trial (n = 120) with the observed data (n = 75), was only 1% under the assumption of a uniform prior distribution. Therefore, although definitive conclusions cannot be reached with this decreased statistical power, we considered that significant differences regarding the primary endpoint may still not have been obtained even if all 120 subjects had been enrolled.
BPAR resistant to 2 sets of steroid pulse therapy was encountered in only 1 patient in this study, but BPAR treated with steroid pulse therapy tended to be more frequent in the MMF group versus the ST group (P = 0.051). The immunosuppressive power might plausibly have been weaker for the former group. In contrast to a steroid being intravenously injected during the first postoperative week, MMF was administered orally or via a nasogastric tube even immediately after the operation. Because the postoperative recovery of oral intake and gastrointestinal function is retarded for most LDLT recipients, the dosage or absorption of MMF might not have been sufficient to reach optimal blood levels in the immediate postoperative period. The addition of induction with interleukin-2 receptor antibody drugs, as in other studies,[8, 10, 11, 13, 14] was considered at the time of the protocol's creation, but the idea was abandoned because these agents were not covered by National Health Insurance in Japan. In any case, it could be argued that the beneficial effects of the steroid-avoidance regimen on HCV recurrence might be offset by the increased use of steroid bolus injections. Kato et al. reported the occurrence of acute rejection during the first year as the only factor associated with increased hepatic fibrosis at 1 year. However, this speculation is not supported by the finding that histological HCV recurrence rates did not differ between patients with BPAR treated with steroid pulse therapy and patients without it.
On the other hand, according to the analysis of the total study population, patients receiving steroid pulse treatment for BPAR showed significantly lower overall survival than those without steroid pulse treatment (Fig. 6). The causes of early mortality for patients receiving steroid pulse treatment were mostly related to infectious complications. Acute rejection is an established risk factor for infections within the first posttransplant year.[34, 35] Obviously, an optimal immunosuppression protocol for sufficiently decreasing the risk of acute rejection remains a prerequisite for HCV-positive recipients.
Notably, in the present study, the posttransplant HCV viral load was lower for the MMF group versus the ST group shortly after the operation. On day 28, the HCV RNA and HCV core protein levels were significantly lower for the MMF group. These findings suggest that HCV replication was suppressed in the MMF group, especially in the immediate posttransplant period. However, viral loads subsequently showed similar increases at 3 and 6 months, and they were decreased by 12 months in both groups. These changes were closely influenced by the induction of antiviral therapy. According to our criteria for starting antiviral therapy, 18 patients in the ST group and 21 patients in the MMF group started IFN/ribavirin treatment between 3 and 12 months. Antiviral therapy could also affect histological HCV recurrence and diminish the effects of differences in immunosuppression protocols.
In conclusion, this study failed to show any significant results. It is unknown whether this was due to the limited size of the sample (reducing the power to detect differences) or because our steroid-avoidance regimen had a minimal impact on LDLT outcomes for HCV-positive recipients. Recent meta-analyses[36, 37] have favored steroid-free protocols in terms of HCV recurrence, but the included studies have been heterogeneous and complicated to interpret. As previously mentioned, most individual trials have not shown results reaching the level of statistical significance. A larger multicenter trial with a more optimal protocol is required to better define the role of steroid-free regimens for HCV-positive recipients.
The authors acknowledge Ms. Naoko Kashiwagi and Ms. Emiko Uno for their support with the data management.