Liver retransplantation of patients with hepatitis C infection is associated with acceptable patient and graft survival



Infection with hepatitis C virus (HCV) is the leading cause of liver transplantation (LT), while liver retransplantation (RT) for HCV is controversial as a result of concerns over poor outcomes. We sought to compare patient and graft survival after RT in patients with and without HCV. We performed a retrospective chart review of all patients undergoing RT at our center between February 1998 and April 2004. Indications for RT, HCV status, patient, and donor characteristics, laboratory values, and hospitalization status at RT were collected. A total of 108 patients (48 HCV and 60 non-HCV) underwent RT during the study period, with mean post-RT follow-up of 1,096 days (range, 0-2,888 days). Grafts from donors aged >60 years were used less frequently in HCV patients at RT (6%) compared with LT (47%), P < 0.001. There was no difference between HCV vs. non-HCV patients in 1- and 3-year patient survival (respectively, 79% vs. 63%, and 71% vs. 63%) and graft survival (respectively, 67% vs. 66%, and 59% vs. 56%). Post-RT mortality and graft failure in HCV patients occurred within the first year in 89% of patients, and 83% were unrelated to HCV recurrence. We conclude that patients should not be excluded from consideration for retransplantation solely on the basis of a diagnosis of HCV. Liver Transpl 13:1717–1727, 2007. © 2007 AASLD.

Infection for hepatitis C virus (HCV) is currently the leading indication for liver transplantation (LT) in the United States, and the number of HCV patients with end-stage liver disease is expected to double by 2020.1 Recurrence of HCV infection after LT is universal and may have an accelerated course leading to cirrhosis in 10%-30% of patients at 5 years after LT.2, 3 The treatment options to prevent progression to allograft cirrhosis are limited. Antiviral therapy for HCV after LT appears to be less effective than in immunocompetent patients.4 In addition, a severe cholestatic from of HCV associated with high virus loads may lead to early allograft failure.5–7 The onset of graft cirrhosis is associated with poor outcomes in the absence of RT, with a risk of clinical decompensation of 42% at 1 year and mortality rates of 60% and 90% at 1 and 3 years, respectively.8

RT for HCV recurrence is controversial because of the poor outcomes reported in several series,1–4, 9–14 and the decision whether to retransplant a patient with HCV has become a major issue at most LT centers. On the basis of epidemiologic trends, most LT centers will confront an increasing burden of HCV patients with graft failure in the next decade.1–3 At present, there are no universally accepted criteria for RT in patients with allograft failure due to HCV recurrence.15

The aims of the present study were to compare the outcomes of RT in HCV vs. non-HCV patients, and in HCV patients undergoing RT for HCV recurrence vs. non–HCV-related graft failure.


HCV, hepatitis C virus; LT, liver transplantation; RT, liver retransplantation; UNOS, United Network for Organ Sharing; MELD, Model for End-Stage Liver Disease; ACR, acute cellular rejection; ICU, intensive care unit.


Study Design

This was a retrospective cohort-based study that included all patients who underwent a second LT at Mayo Clinic Transplant Center, Jacksonville, Florida, between March 1998 and April 2004. Patient and graft survival after RT were determined through February 2006. The study was approved by the Mayo Clinic Institutional Review Board.

Patients were grouped as either HCV or non-HCV, according to their indication for initial LT. HCV patients undergoing RT for recurrent HCV (group 1) were compared with HCV patients undergoing RT for non-HCV indications (group 2). Group 1 patients were also compared with non-HCV patients undergoing late RT (>90 days after initial LT) (group 3). Graft failure was defined by graft dysfunction leading to RT or patient death. HCV-related liver disease was defined by the presence of serum hepatitis C antibodies (anti-HCV) by second-generation enzyme-linked immunosorbent assay, in addition to the presence of circulating HCV ribonucleic acid (HCV-RNA) in serum by reverse transcriptase–polymerase chain reaction or reactivity for anti-HCV by recombinant immunoblot assay before transplantation.

HCV Recurrence and HCV-Related Graft Failure

Liver biopsies were performed after LT and RT at 1 week, 4 months, and annually per protocol and as clinically indicated for liver enzyme abnormalities. Significant HCV recurrence was defined by an increase in liver enzymes (≥2-3 times baseline or upper limits of normal), detectable HCV by reverse transcriptase–polymerase chain reaction, and histologic features of HCV recurrence (Batts Ludwig grade ≥2 or progressive fibrosis).16

Graft failure was attributed to HCV recurrence in those patients with progressive graft injury and fibrosis that was histologically consistent with marked HCV recurrence. Patients with biliary or vascular complications directly contributing to graft loss and without progression to advanced fibrosis (bridging fibrosis or cirrhosis) on examination of liver biopsy samples were designated as having a non-HCV indication for RT. Histologic review of all explants was also performed to look for other causes of graft failure.

Criteria for RT

All patients presenting with graft failure, regardless of HCV status or recurrence as the cause of graft failure, were considered for RT. Exclusion criteria for RT were applied uniformly to all patients. RT was denied or withheld in the following circumstances: (1) recurrent hepatocellular carcinoma or known metastatic malignancy; (2) presence sepsis or active infection; or (3) recidivism or noncompliance.

Clinical Parameters

Patient variables that were analyzed included age, sex, body mass index, total bilirubin (mg/dL), creatinine (mg/dL), international normalized ratio, serum albumin (g/dL), HCV genotype, HCV load (both at initial LT and at RT), hospitalization status, and the presence and severity of ascites and hepatic encephalopathy at the time of RT. Child-Pugh-Turcotte score and Model for End-Stage Liver Disease (MELD) score were calculated on the basis of laboratory values at the time of LT and RT. MELD was also calculated at the time of listing for RT, and the interval from listing to RT was determined. Patients who received dialysis on 2 occasions during the week before transplantation were assigned a creatinine value of 4 mg/dL for statistical analysis, based on the United Network for Organ Sharing (UNOS) MELD calculator.17 Donor and operative factors that were analyzed included age, and cold and warm ischemia time.

HCV Testing

Anti-HCV antibodies were detected by a commercially purchased assay (EIA-2.0, Abbott, Abbott Park, IL) and a 5-antigen recombinant immunoblot assay (RIBA 3.0 SIA, Chiron, Emeryville, CA). Viral RNA was extracted from aliquots of 100 μL of sera by using a chaotropic lysis protocol supplied in the Amplicor Hepatitis C Virus v2.0 Test kit (Roche Diagnostics, Indianapolis, IN), as previously described.9 HCV RNA quantitation was expressed in IU/mL. Before May 2001, HCV RNA quantitation was performed by using a branched chain DNA assay (Quantiplex Version 2.0, Chiron) and expressed in viral equivalents/mL (vEq/mL). The lower limit of sensitivity of this assay was 0.2 × 106 vEq/mL. Since May 2001, HCV-RNA quantitation was determined by the Versant HCV RNA 3.0 branched DNA Assay, with a lower limit of detection of 615 IU/mL (Bayer, Norwood, MA). A conversion factor of 158,730 was used to transform HCV RNA quantitation from vEq/mL to IU/mL for purposes of standardization.

Donor Procurement and Surgical Technique

Organ procurement was carried out as described elsewhere with aortic and portal perfusion with University of Wisconsin solution.18 All transplantation procedures were performed by using the piggyback technique, without venovenous bypass, as previously described.19 Patients with concomitant hepatocellular carcinoma received pretransplantation chemoembolization as per protocol.20 LT was declined for hepatocellular carcinoma patients with extrahepatic involvement, documented tumor vascular invasion, or tumor size >8 cm.

Immunosuppression and Allograft Rejection

Immunosuppression for initial LT involved 3 drugs: tacrolimus (Prograf, Astellas Pharma US, Deerfield, IL); prednisone; and mycophenolate mofetil (Cellcept, Nutley, NJ) or azathioprine. Target trough levels of tacrolimus were 8-12 ng/mL during the first 4 months and 5-8 ng/mL thereafter. Prednisone was tapered and discontinued during the first 4 months after LT. Mycophenolate mofetil or azathioprine were discontinued 4 months after LT. Immunosuppression after RT involved the same 3-drug regimen, but prednisone was rapidly tapered and discontinued over the first 2 weeks in HCV patients.

Deviation from these regimens was at the discretion of the treating hepatologist in the following circumstances: (1) Patients having neuropsychiatric side effects or unable to achieve therapeutic levels with tacrolimus were converted to cyclosporine (Neoral). (2) Patients with renal insufficiency on tacrolimus or cyclosporine were converted (>30 days after transplantation) to sirolimus (Rapamune, used as of 2000) or combined with it while reducing calcineurin inhibitor target levels. (3) Mycophenolate mofetil was discontinued earlier than 4 months after transplantation in cases of a diagnosis of malignancy, infections, bone marrow suppression, or gastrointestinal side effects. (4) Corticosteroids were not rapidly tapered in patients with graft rejection, concerns about inadequate primary immunosuppression, or conversion to sirolimus therapy. (5) Basiliximab (Simulect, used as of 2002) induction was used in patients with renal insufficiency at transplantation.

Acute cellular rejection (ACR) was defined by histologic assessment of liver tissue. ACR was graded as mild, moderate, or severe according to the criteria for global assessment proposed by Demetris et al.21 Patients with biopsy-documented moderate to severe ACR were treated with intravenous methylprednisolone (total dose, 1-3 g) divided into 3 alternate-day doses. Prednisone tapering followed a standard protocol, which was not modified in the presence of ACR. Muromonab (OKT3, Ortho Biotech, Raritan, NJ) antibody or antithymocyte globulin (Thymoglobulin, Genzyme, Cambridge, MA) infusion was reserved for patients with ACR resistant to intravenous corticosteroids. Chronic rejection was histologically defined by the updated international Banff schema for liver allograft rejection.22

Statistical Analysis

Statistical analysis was performed by SPSS software, version 14.0 (SPSS, Chicago, IL). Patient characteristics among different groups were compared with the Mann-Whitney test or the Fisher exact test, and within the same group at LT and RT by paired t test or Wilcoxon signed rank test. Patient and graft survival in different patient groups was compared by Kaplan-Meier analysis by log rank testing. The impact of clinical and laboratory criteria was analyzed by univariate and multivariate Cox proportional hazard regression. Multiple factors achieving a P value of <0.1 on univariate analysis were entered into a multivariate stepwise regression analysis. All tests were 2 tailed, and significance was defined at a P value of < 0.05.



A total of 761 patients underwent initial LT at our center during the study period. Of these, 294 (39%) were in HCV patients, and 467 (61%) were in non-HCV patients. During the same time period, 108 patients had a second LT: 48 (44%) in HCV patients, including 3 patients undergoing initial LT elsewhere, and 60 (56%) in non-HCV patients, including 7 patients undergoing initial LT elsewhere. The retransplant rate in HCV and non-HCV patients undergoing in-center primary LT during the study period was 15.3% (45 of 294) and 11.3% (53 of 467), respectively (P = NS). RT was performed >90 days after LT (late RT) in 44 patients who underwent primary LT at our center: 29 (63%) HCV infected and 15 (28%) non-HCV infected (P < 0.001).

Sixteen HCV patients (mean patient age, 48 ± 8 years and initial LT donor age of 59 ± 17 years) with significant HCV recurrence and graft dysfunction did not undergo RT during the inclusion period. Of these, 1 was on the waiting list and later retransplanted, 5 were denied RT as a result of recidivism and 1 for noncompliance, 3 died within a month of decompensation before RT evaluation could be performed, 1 patient opted for hospice care, and 5 patients had transferred their care elsewhere and did not return to the study center for RT evaluation. Eleven of these patients died by the end of follow-up. No HCV patients with initial graft failure were denied RT at the study center for reasons other than the outlined exclusion criteria.

The causes of graft failure in HCV and non-HCV patients leading to RT are outlined in Table 1. There was no difference in primary nonfunction as an indication for RT in HCV and non-HCV patients. Twenty-five HCV patients underwent RT for recurrent HCV (group 1); all were late RT. Twenty-three HCV patients underwent RT for non-HCV indications (group 2). Twenty-two non-HCV patients underwent late RT (group 3). The clinical characteristics of all HCV patients, of all non-HCV patients, and of the 3 groups are outlined in Table 2. There was a smaller proportion of donors aged >60 years at RT (6%) compared with LT (47%) in HCV patients (P < 0.001).

Table 1. Causes of Initial Graft Failure Leading to Liver Retransplantation in Patients With and Without HCV
Cause of graft failureNo. of HCV patients (LT-RT interval, days)No. of non-HCV patients (LT-RT interval, days)
  • Abbreviations: HCV, hepatitis C virus; LT, liver transplantation; RT, liver retransplantation; PNF, primary nonfunction.

  • *

    Denotes the subset of HCV-infected patients with additional findings of chronic rejection on explant.

PNF6 (2, 2, 3, 3, 3, 3)5 (2, 2, 3, 12, 13)
Poor initial graft function and graft failure (non-PNF)2 (65, 71)4 (14, 40, 44, 63)
Biliary or arterial complications13 (161 ± 276) * 1 (287)30 (100 ± 171)
Venous (portal or hepatic) complications2 (129, 837)6 (6, 18, 30, 131, 217)
Recurrent disease25 HCV (724 ± 487) * 3 (646, 891, 883)4 primary sclerosing cholangitis (2,882, 2,920, 3,062, 3,599)
Other7 chronic rejection (1,030 ± 1,837)
 3 venoocclusive disease (331, 1,513, 1,680)
 1 graft rupture (1)
Table 2. Patient and Donor Characteristics at RT in All HCV and Non-HCV Patients*
CharacteristicAll HCV (n = 48)All non-HCV (n = 60)P value (HCV vs. non-HCV)Group 1 (HCV patients/HCV RT indication) (n = 23)Group 2 (HCV patients/non-HCV RT indication) (n = 23)P value (group1 vs. group 2)Group 3 (non-HCV patients/late RT) (n = 22)P value (group 1 vs. group 3)
  • Abbreviations: HCV, hepatitis C virus; RT, liver retransplantation; LT, liver transplantation; PNF, primary nonfunction; INR, international normalized ratio of prothrombin time; MELD, Model for End-Stage Liver Disease; CPT, Child-Pugh-Turcotte; ICU, intensive care unit; NA, not applicable; n, number of patients (denominator shown if data are missing for some cases). Values are provided as mean ± 1 standard deviation, and as n (%) unless otherwise specified.

  • *

    Characteristics also include HCV patients undergoing RT for recurrent HCV (group 1), HCV patients undergoing RT for non–HCV-related indications (group 2), and non-HCV patients undergoing late RT (>90 days after LT) (group 3).

  • Virus loads were not titrated to end point in all cases.

Age (yr)52 ± 850 ± 140.953 ± 652 ± 100.350 ± 150.8
 >608 (17%)15 (25%)0.33 (12%)5 (22%)0.53 (14%)1
Male gender (%)35 (73%)32 (53%)0.04716 (64%)19 (83%)0.214 (64%)1
Body mass index28 ± 627 ± 70.627 ± 628 ± 50.527 ± 60.8
 LT-RT interval (days)444 ± 490447 ± 1,0190.02724 ± 487140 ± 264<0.0011,175 ± 1,4290.7
 Median24640 68225 353 
>1 yr22 (46%)11 (18%)0.00319 (76%)3 (13%)<0.00111 (50%)0.08
Creatinine at RT (mg/dL)1.5 ± 0.81.6 ± 10.91.6 ± 0.61.4 ± ± 0.90.2
 >29 (19%)15 (25%)0.54 (16%)5 (22%)0.718%1
Bilirubin at RT (mg/dL)14 ± 1615 ± 170.815 ± 1913 ± 120.516 ± 210.9
 >10 (%)20 (42%)27 (46%)0.710 (40%)10 (43%)141%1
INR at RT2.1 ± 1.31.8 ± 1.50.092 ± 12.1 ± ± 0.30.005
Listing-RT interval (days)57 ± 10528 ± 490.1580 ± 12531 ± 720.00866 ± 650.7
MELD at listing for RT22 ± 823 ± 110.623 ± 820 ± 80.1820 ± 90.18
 >25 (%)17 (35%)22 (37%)111 (44%)6 (26%)0.25 (23%)0.2
MELD at RT24 ± 1023 ± 110.524 ± 1024 ± 100.821 ± 90.2
 >2521 (44%)22 (37%)0.610 (40%)11 (48%)0.85 (23%)0.2
Change in MELD from listing to RT2.3 ± 5.3−0.7 ± ± 3.63.6 ± ± 4.20.5
Albumin (mg/dL)3 ± 0.53.2 ± 0.60.133 ± 0.53.1 ± ± 0.50.02
CPT9.8 ± 28.9 ± 20.04810.3 ± 2.29.3 ± ± 1.90.002
Child class C (%)25 (52%)21/49 (43%)0.415 (60%)10 (44%)0.46 (27%)0.04
Hospitalization (%)25 (52%)43 (72%)0.04613 (52%)12 (52%)18 (49%)0.4
ICU (%)11 (23%)23 (38%)0.0994 (16%)7 (30%)0.33 (14%)1
HCV load at RT (×1,000 IU/mL), median (range)325 (0 to >7,692)NANA454 (0 to >7,692)83 (0 to 3,552)NANANA
  MEq/mL/158,730 IU (%)25/44 (57%)17 (68%)8/19 (42%)0.13
Genotype 1 (%)27/31 (87%)NANA18/20 (90%)9/11 (82%)0.6NANA
Donor age at LT (yr)56 ± 2154 ± 230.660 ± 2053 ± 210.261 ± 160.9
 >60 (%)21/45 (47%)25/53 (47%)112/22 (55%)9 (39%)0.48/15 (53%)1
Donor age at RT (yr)36 ± 1741 ± 190.0838 ± 1933 ± 160.341 ± 200.4
 >60 (%)3 (6%)10 (17%)0.142 (8%)1 (4%)14 (18%)0.4
Warm ischemia time (min)39 ± 2036 ± 130.1342 ± 2637 ± 110.935 ± 120.4
 >60 (%)3 (6%)3 (5%)12 (8%)1 (4%)11 (4%)1
Cold ischemia time (h)6.9 ± 1.87 ± ± 1.66.7 ± 1.90.168 ± 2.10.15
 >12 (%)1 (2%)1 (2%)10 (0%)1 (4%)0.51 (4%)0.5
Length of follow-up (days)1,083 ± 6821,107 ± 8840.9917 ± 5951,264 ± 7360.071,159 ± 8330.3


Outcomes were less favorable after RT vs. LT, with patient and graft survival rates of 74% vs. 88% and 66% vs. 80%, respectively, at 1 year, and 70% vs. 82%, and 62% vs. 72%, respectively, at 3 years (P = 0.007 for patient survival; P = 0.03 for graft survival). However, there was no statistical difference between HCV and non-HCV patients in post-RT patient and graft survival (Fig. 1) after a mean follow-up of 1,095 ± 797 days (range, 0-2,888 days). All surviving patients had at least 1 year of follow-up. The actuarial patient and graft survival in HCV patients after RT were 79% & 71% at 1 year (n = 48), and 67% & 59% at 3 years (n = 34). Actuarial patient and graft survival in non-HCV patients after RT were 70% and 63% at 1 year (n = 60), and 66% and 56% at 3 years (n = 44). In addition, there was no difference in post-RT patient and graft survival between HCV patients undergoing RT for HCV recurrence (group 1) vs. non-HCV indications (group 2), or vs. non-HCV patients undergoing late RT (group 3) (Fig. 2). Patient and graft survival at 1 year after RT were 76% and 68% in group 1, 83% and 74% in group 2, and 77% and 68% in group 3. Twenty-one HCV patients had marked HCV recurrence after RT, including 12 patients in group 1 who underwent antiviral treatment, but those patients had 90% patient and graft survival at 1 year. Antiviral therapy was only provided for histologically defined HCV recurrence after LT and RT, with no preemptive treatment in any patients.

Figure 1.

(A) Patient survival after RT in HCV (solid line) and non-HCV (dashed line) patients. (B) Graft survival after RT in HCV (solid line) and non-HCV (dashed line) patients.

Figure 2.

(A) Patient survival after RT in HCV patients with HCV-related graft failure (group 1, solid line) and non–HCV-related graft failure (group 2, dashed line), and in non-HCV patients undergoing late RT (group 3, dotted line). (B) Graft survival after RT in HCV patients with HCV-related graft failure (group 1, solid line) and non–HCV-related graft failure (group 2, dashed line), and in non-HCV patients undergoing late RT (group 3, dotted line).

One-year post-RT patient and graft survival according to individual clinical criteria with respective univariate and multivariate Cox regression analysis in HCV and non-HCV patients, and in group 1 patients are outlined in Tables 3–6. The studied criteria were: patient age >60 years, male patient gender, MELD score at RT >25, creatinine >2 mg/dL, bilirubin >10 mg/dL, Child class C at RT, LT-RT interval >1 year, intensive care unit (ICU) status at RT (vs. non-ICU), donor age >60 years, warm ischemia time >60 minutes, and virus load at RT >1 mEq (or equivalent to 158,730 IU/mL) in the case of HCV patients. Only one patient, or none, per group in our series had a cold ischemia time >12 hours, and we were thus unable to analyze this factor.

Table 3. Univariate Cox Regression Analysis of Post-RT Patient and Graft Survival by Clinical Criteria in all HCV and Non-HCV Patients
Clinical criteria at RTAll HCV (n = 48)All non-HCV (n = 60)
1-year patient survivalHazard ratio [95% CI] P value1-year graft survivalHazard ratio [95% CI] P value1-year patient survivalHazard ratio [95% CI] P value1-year graft survivalHazard ratio [95% CI] P value
  1. Abbreviations: RT, liver retransplantation; HCV, hepatitis C virus; 95% CI, 95% confidence interval; MELD, Model for End-Stage Liver Disease; LT, liver transplantation; ICU, intensive care unit.

Patient age >60 yr63% vs. 83%2.8 [0.8–9.3] 0.09950% vs. 75%27 [0.9–7.9] 0.0740% vs. 80%4 [1.7–9.5] 0.00240% vs. 71%2.5 [1.1–5.5] 0.03
Male patient sex86% vs. 62%0.5 [0.2–1.3] 0.277% vs. 44%0.6 [0.2–1.6] 0.369% vs. 71%1.2 [0.5–2.7] 0.766% vs. 61%0.9 [0.4–1.9] 0.8
MELD at RT >2576% vs. 82%1.4 [0.5–4.5] 0.562% vs. 78%2.0 [0.8–5.5] 0.1646% vs. 84%3.2 [1.4–7.6] 0.00741% vs. 76%2.5 [1.1–5.3] 0.02
Creatinine at RT >2 mg/dL78% vs. 80%0.9 [0.2–3.9] 0.867% vs. 72%1.1 [0.3–3.8] 0.953% vs. 76%1.7 [0.7–4.1] 0.347% vs. 69%1.5 [1.1–5.3] 0.3
Bilirubin at RT >10 mg/dL70%vs. 86%1.9 [0.6–5.6] 0.360% vs. 79%1.8 [0.7–4.7] 0.356% vs. 84%3.0 [1.2–7.4] 0.0248% vs. 78%1.6 [1.2–6.0] 0.02
Child class C at RT76% vs. 83%1.5 [0.5–4.7] 0.568% vs. 74%1.3 [0.5–3.4] 0.757% vs. 82%2.2 [0.8–5.7] 0.1252% vs. 71%1.5 [0.6–3.5] 0.4
LT-RT interval >1 year86% vs. 73%0.6 [0.2–2.0] 0.477% vs. 65%0.7 [0.3–2.0] 0.582% vs. 67%0.9 [0.3–2.6] 0.873% vs. 61%0.9 [0.4–1.9] 0.9
ICU at RT82% vs. 78%1.1 [0.3–4.1] 0.982% vs. 67%0.7 [0.2–2.4] 0.652% vs. 81%2.7 [1.2–6.3] 0.0244% vs. 77%2.5 [1.2–5.5] 0.02
Donor age >60 yr33% vs. 82%7.6 [2–28.9] 0.00333% vs. 73%5.6 [1.5–20.1] 0.00960%vs. 72%1.2 [0.4–3.6] 0.750% vs. 66%1.3 [0.5–3.5] 0.6
Warm ischemia time >60 min33% vs. 82%3.2 [0.7–15.2] 0.1433% vs. 73%2.6 [0.6–11.9] 0.2100% vs. 68%0.04 [0–62.7] 0.467% vs.63%0.7 [0.09–5.2] 0.7
HCV load at RT >1 MEq or equivalent76% vs. 84%1.3 [0.4–4.6] 0.672% vs. 68%0.8 [0.3–2.2] 0.7
Table 4. Multivariate Cox Regression Analysis in HCV and Non-HCV Patients
SurvivalHazard ratio95% confidence intervalP value
  1. Abbreviations: HCV, hepatitis C virus; MELD, Model for End-Stage Liver Disease.

All HCV patient survival   
 Recipient age >60 yr4.21.1–15.70.04
 Donor age >60 yr11.22.6–490.001
All HCV graft survival3.61.2–10.90.03
 Donor age >60 yr7.82.0–300.003
All non-HCV patient survival   
Recipient age >60 yr3.61.5–8.70.005
 MELD >25 yr2.61.1–6.20.04
All non-HCV graft survival   
 Bilirubin >10 mg/dL2.61.2–60.02
Table 5. Univariate Cox Regression Analysis of Post-RT Patient and Graft Survival by Clinical Criteria in HCV Patients Undergoing RT for Recurrent HCV
Clinical criteria at RTHCV patients undergoing RT for recurrent HCV (n = 25)
1-year patient survivalHazard ratio [95% CI] P value1-year graft survivalHazard ratio [95% CI] P value
  1. Abbreviations: RT, liver retransplantation; HCV, hepatitis C virus; LT, liver transplantation; ICU, intensive care unit.

Patient age >60 yr33% vs. 82%5.9 [1.1–33.2] 0.0433% vs. 73%2.8 [0.6–13.8] 0.2
Male patient sex88% vs. 56%0.3 [0.05–1.4] 0.1281% vs. 44%0.3 [0.07–1.2] 0.08
MELD at RT >2570% vs. 80%1.7 [0.3–8.5] 0.550% vs. 80%3.2 [0.8–13.4] 0.11
Creatinine at RT >2 mg/dL75% vs. 76%1.2 [0.1–9.9] 0.975% vs. 67%0.7 [0.09–5.7] 0.7
Bilirubin at RT >10 mg/dL70% vs. 80%1.7 [0.3–8.5] 0.550% vs. 80%3.2 [0.8–13.4] 0.11
Child class C at RT67% vs. 90%3.9 [0.5–33.6] 0.253% vs. 90%5.5 [0.7–44.9] 0.11
LT–RT interval >1 year84% vs. 50%0.3 [0.05–1.3] 0.1074% vs. 50%0.5 [0.1–2.2] 0.4
ICU at RT50% vs. 81%3.1 [0.6–17] 0.1950% vs. 71%1.7 [0.3–8.3] 0.5
Donor age >60 years0% vs. 83%9.3 [1.5–56.8] 0.020% vs. 74%7.3 [1.3–40.5] 0.02
Warm ischemia time >60 minutes0% vs. 83%8.0 [1.3–48.7] 0.020% vs. 74%6.0 [1.1–31.3] 0.03
HCV load at RT >1 MEq or equivalent77% vs. 75%0.9 [0.2–4.8] 0.971% vs. 63%0.7 [0.2–3.0] 0.7
Table 6. Multivariate Cox Regression Analysis in HCV Patients Undergoing Liver Retransplantation for Recurrent HCV
SurvivalHazard ratio95% confidence intervalP value
  1. Abbreviation: HCV, hepatitis C virus.

Patient survival   
Recipient age >6022.512.2–2260.008
Donor age >6018.51.9–1770.01
Graft survival   
Donor age >607.31.3–40.50.02

Graft Failure Leading to a Third LT

Six HCV (3 in group 1 and 3 in group 2) and 7 non-HCV patients received a third LT (Table 7). Graft failure leading to a third transplant in group 1 patients occurred within 12 days of RT and was unrelated to HCV. HCV recurrence was the cause of graft failure in 2 patients in group 2. Four HCV and 4 non-HCV patients survived after receiving a third graft, with mean follow-up of 1,098 days (range, 83-2,239 days).

Table 7. Causes and Time* of Graft Failure and Death in HCV and Non-HCV Patients
CharacteristicNo. patients (days from RT)No. patients (days from RT)
  • Abbreviations: HCV, hepatitis C virus; PNF, primary nonfunction; RT, liver retransplantation.

  • *

    Time is from RT in days.

  • Died after repeat RT.

Cause of graft failure  
No. patients67
Portal vein thrombosis1 (5)
Hepatic artery thrombosis/biliary necrosis1(14)2(21, 25)
Recurrent HCV2 (150, 190)
Other1 delayed graft failure (12)1 graft rupture (2)
Cause of death  
No. patients1222
Sepsis3(20, 81, 111)7(18, 24, 43, 68, 120, 145, 345)
Cardiac complications2(13, 83)4(0, 0, 3, 52)
Graft complications1 biliary stricture (355)1 PNF (1)
 1 recurrent HCV (298)1 hepatic artery thrombosis (4)
  2 hepatic vein thrombosis/stenosis (36, 80)
  1 recurrent venoocclusive disease (438)
Malignancy3 recurrent hepatocellular carcinoma (240, 251, 1,924)1 recurrent epithelioid hemangioepithelioma (188)
 1 lymphoma (450)1 gastrointestinal stromal tumor (1,878)
Other1 intracranial bleed (86)1 intracranial bleed (223)
  1 graft-vs.-host disease (83)
  1 unknown “natural causes” (784)
  1 accidental (1,334)

Mortality After RT

At the end of the study, 34 (31%) of 108 patients undergoing RT had died, including 5 of 13 patients undergoing a third LT (Table 7). A total of 65% of deaths occurred within 6 months of RT, and 73% of these were due to sepsis and cardiac events. Malignancy (recurrent in 5 of 6 patients) accounted for 50% of deaths occurring after 6 months after RT. In all, recurrent HCV was the cause of 1 post-RT death, and 2 post-RT graft failures, i.e., 17% of deaths or graft failures in HCV patients within the follow-up period.


Immunosuppression in HCV patients after LT and RT is compared in Table 8. All HCV patients' initial immunosuppressant regimens included a calcineurin inhibitor after LT (initial LT at Mayo Clinic Transplant Center, Jacksonville) and RT. The only difference in post-RT immunosuppression between group 1 and 2 patients was in rapid corticosteroid tapering in 67% of group 1 patients vs. 25% of group 2 patients (P = 0.002). Significant HCV recurrence in group 1 patients with at least 90 day graft survival after RT was similar for those with primary immunosuppression changes vs. without (63% vs. 64% P = 1), and for those receiving sirolimus vs. not (57% vs. 67% P = 1).

Table 8. Immunosuppression in HCV Patients After LT and RT
CharacteristicProportion of HCV patientsP value
  1. Abbreviations: HCV, hepatitis C virus; LT, liver transplantation; RT, liver retransplantation.

Primary immunosuppressant changed after transplantation17%40%<0.001
Sirolimus therapy provided at any point after transplant11%34%0.002
Basiliximab induction2%10%0.046
Early discontinuation of mycophenolate mofetil24%42%0.02
Rapid tapering of corticosteroid045%<0.001

Allograft Rejection

Corticosteroid-treated ACR developed in 25% of HCV patients after RT, compared with 38% after LT (P = 0.2). No patients required the use of OKT3 or antithymocyte globulin after LT or RT. Corticosteroid-treated ACR did not significantly affect post-RT patient or graft survival in group 1 patients. One-year post-RT patient and graft survival were 71% vs. 78% (P = 1), and 57% vs. 72% (P = 0.6) in those with vs. those without corticosteroid-treated ACR.

Ductopenic chronic rejection was identified on explant wedge biopsy in 4 HCV patients undergoing RT. Percutaneous liver biopsy before RT identified only one of these cases. Of the other 3 HCV patients, 2 were suspected of having cholestatic HCV on the basis of pre-RT percutaneous liver biopsy findings, and 1 was diagnosed with ischemic intrahepatic biliary strictures by cholangiography before RT. All 4 patients had received antiviral treatment after initial LT without reduction in immunosuppression. Two did not respond to treatment, 1 experienced relapse after cessation of therapy, and 1 cleared HCV and stopped therapy 1 month before RT with a sustained viral response after RT.


This retrospective cohort-based study confirms a number of previously published findings. The rate of RT for HCV in this series of 15.3% was in keeping with reported rates of 9%-20%.13, 23–25 Patient and graft survival were significantly lower after RT than the initial transplant.11, 23, 26 Post-RT mortality and graft failure occurred predominantly in the first 6 months after RT.13, 23, 27 Early mortality was mainly related to sepsis (most frequently from or including an abdominal source) and cardiac events, irrespective of HCV status.13, 23, 27

The primary finding of our study was that HCV status, or recurrence as the indication for RT, did not negatively affect short-term RT outcomes. This is in agreement with a multicenter study by Rosen et al.,28 who studied 70 HCV RT patients; however, it conflicts with large database studies that report an increased risk of patient death and graft failure in HCV patients undergoing RT.10, 11, 14 UNOS data from 1990 to 2002 reported patient survival rates after RT of 50% to 61% in HCV patients compared with 65% to 70% in non-HCV patients at 1 year, and 50 to 55% vs. 60 to 63% at 3 years, respectively.10, 23 RT outcomes in HCV patients in single-center studies are more variable, with 3-year patient survival rates ranging from 33%9, 27 to 62%.29 This variability may reflect the influence of patient-, donor-, and center-specific factors in determining outcomes.

Variability in donor age between centers may be a contributing factor to differences in outcomes. Advanced donor age has been recognized to be associated with accelerated fibrosis progression along with decreased patient and graft survival in HCV patients undergoing primary LT.26, 30–33 Donor age also appears to be an important factor in RT for HCV. Carmiel-Haggai et al.,27 who studied 47 HCV RT patients, found an increased post-RT HCV patient mortality risk ratio (3.58, P = 0.002) for donor age ≥60 years. Jain et al.29 attributed an acceptable actuarial 3-year patient survival of 61.7% in 19 HCV patients undergoing RT to the use of younger donors (mean age, 31.8 ± 13.6 years). Pelletier et al.14 analyzed 1,718 patients (with and without HCV) undergoing RT from 1997 to 2002 UNOS data and found an increased covariate-adjusted mortality hazard ratio (1.3, P = 0.02) for donor age ≥60 years compared with age 40-59, and a reduced hazard ratio for donor age <40 years (≤0.76; P < 0.001). In our series, donor age >60 years was also a risk factor for death and graft failure by multivariate regression analysis in HCV patients, but was not a clear risk factor in non-HCV patients. This suggests that the use of younger donors in HCV patients at RT may preferentially improve their outcomes relative to non-HCV patients. If confirmed in a larger patient sample, this finding may impact the allocation of older donor grafts at RT. In our current series, only 6% of grafts used in HCV patients at RT were from donors aged >60 years, which may contribute to our better-than-expected patient and graft survival.

MELD score at the time of RT is another potential predictor of poor outcomes in HCV patients. MELD scores of >25 have been associated with decreased survival after primary LT,34 and RT.23 Burton et al.35 identified maximal utility when HCV patients underwent RT with MELD scores of 21-24, with decreased utility with MELD scores >28. Although 44% of HCV patients in this series had MELD scores >25 at RT, this did not greatly influence outcomes after RT. A relatively short listing-to-retransplant interval in our series may have prevented development of clinically important complications in this group of patients known to have increased short-term morbidity and mortality.8 Thus, it is unclear whether our results can be extrapolated to centers with long waiting times.

Differences in the immunosuppression used in our HCV RT patients did not appear to have a marked effect on outcomes. Deviation from the standard regimen was more common after RT. This was primarily due to baseline renal insufficiency and intolerance of standard regimens of tacrolimus and mycophenolate mofetil, highlighting the severity of illness in these patients. Changes in the primary immunosuppressant did not greatly influence HCV patient or graft survival in this series. Corticosteroids were tapered and discontinued by 2 weeks, which could have been beneficial. However, corticosteroid requiring ACR was similar to what has been reported in the literature and was not significantly different from patients who underwent LT. High-dose corticosteroid boluses but not maintenance corticosteroid therapy have been shown to be associated with more aggressive HCV recurrence and disease progression after LT.3

HCV loads >1 MEq at transplantation have been associated with diminished post-LT patient and graft survival, with limited data in RT. HCV load >1 MEq was not associated with worse outcomes after RT in all HCV patients, or in HCV patients undergoing RT specifically for HCV recurrence. Graft loss and death due to recurrent HCV were uncommon in this series. Additionally, HCV patients undergoing RT for recurrent HCV who then had corticosteroid-treated ACR or who met a histologic definition of clinically significant disease recurrence did not have decreased early graft survival. However, the study period was short, and longer follow-up is needed to define long-term outcomes in this group.

Additional factors influencing outcome after LT or RT for HCV including serum creatinine >2 mg/dL, bilirubin >10 mg/dL, Child-Pugh-Turcotte score, short retransplant intervals, ICU status, and prolonged cold and warm ischemia times have been previously described predictors, but they did not have a statistically significant effect in this series.10, 14, 26, 37 This may be due to limitations of a small sample size.

Of note, in addition to the use of younger donors, a number of risk factors such as advanced recipient age, retransplantation within 1 year of LT, ICU care at RT, and lengthy cold and warm ischemia times were infrequent in our HCV patients, particularly in those undergoing RT for HCV recurrence. This points toward a select HCV cohort, despite the absence of a prospective selection bias. However, it also demonstrates that good early outcomes can be obtained with RT for HCV recurrence, under favorable patient and donor conditions.

Short-term graft and patient survival after RT for HCV patients, regardless of indication for RT, may be similar to results obtained for non-HCV patients. On the basis of these results, younger donors should be used when considering RT in HCV patients, but further study is needed to better define the optimal recipient and donor criteria for RT.