The variable impact of specific risk factors on survival outcomes based on pre-transplantation diagnosis was analyzed in adult liver transplant recipients reported to the Scientific Registry of Transplant Recipients: 778 with hepatitis B (HBV), 3463 with hepatitis C (HCV) and 7429 without viral hepatitis. Graft and patient survival for the HBV and no viral hepatitis groups did not differ significantly. The HCV group had significantly lower graft (p = 0.0019) and patient survival (p < 0.0001) than the no viral hepatitis group. Patient survival was significantly lower (p = 0.0011) for HCV compared to HBV patients; differences in graft survival approached significance (p = 0.0561). Donor age, which was not a risk factor in patients with HBV, was the strongest predictor of graft loss and death in patients with HCV, starting with donors >40 years. Donor age >60 years was the strongest predictor of graft loss and death in patients without viral hepatitis. The risks of graft loss and death were reduced for patients on tacrolimus-based immunosuppression with mycophenolate mofetil, regardless of disease etiology. There are clear differences in risk factors for poor outcomes based on underlying liver disease, particularly with regard to the impact of donor age.
Cirrhosis caused by viral hepatitis remains the most common indication for liver transplantation in the United States and Europe. Advances in antiviral prophylaxis and treatment of post-transplantation hepatitis B (HBV) have significantly improved patient and graft survival in HBV-infected recipients (1–3). The inevitability of hepatitis C (HCV) reinfection and aggressiveness following transplantation have resulted in a continued focus on factors that contribute to poor survival outcomes in this population (4–8). Based on published data, the International Liver Transplantation Society Expert Panel Consensus Conference on Liver Transplantation and Hepatitis C identified risk factors for decreased survival or increased disease severity in patients infected with HCV, including pre-transplant viral load, female gender, recipient age, non-white race, underlying disease severity, donor age and cytomegalovirus (CMV) infection; more controversial or less well-established risk factors included human leukocyte antigen matching, cold ischemia time and living donor transplantation (9). In an era of maintenance immunosuppression that includes combinations of anti-proliferative agents, calcineurin inhibitors and corticosteroids, the effects of excess immunosuppression in patients infected with hepatitis B or C continue to be examined.
The impact of various risk factors on post-transplantation outcomes within the general transplant population has been studied extensively. However, specific differences based on disease etiology and their specific implications for patient management have not been addressed. To evaluate the variable impact of specific risk factors on graft and patient survival, this retrospective study compared liver transplant recipients infected with HBV, HCV and those with other underlying liver diseases.
Materials and Methods
Data reported to the Scientific Registry of Transplant Recipients (SRTR, October 2003 release) for patients between the ages of 18 and 80 years who underwent primary, single-organ, liver transplantation in the United States between January 1, 1995 and April 30, 2001, were analyzed. This population originally was sampled as a part of a study assessing long-term outcomes of maintenance immunosuppression with mycophenolate mofetil (MMF) (10); therefore, only patients who were discharged from the hospital on tacrolimus-based immunosuppression therapy (i.e. tacrolimus and corticosteroids) with or without MMF were included in the analysis. Patients who received cyclosporine or other non-tacrolimus immunosuppression at discharge were excluded from the analysis (4756 patients). Patients were divided into three groups, based on pre-transplantation diagnosis: HBV, HCV and no viral hepatitis (i.e. non-cholestatic cirrhosis, cholestatic liver disease, acute hepatic necrosis due to diseases other than hepatitis B or C, metabolic diseases, malignancies and congenital and vascular abnormalities). Follow-up data were obtained 6 and 12 months post-transplantation, and annually thereafter, up to August 1, 2003, or as long as data were available.
Graft survival and patient survival were determined for each diagnostic group using univariate Kaplan–Meier analysis; differences between groups were compared using the log-rank test. Cox proportional hazards modeling was used to investigate the independent effects of multiple covariates on graft and patient survival. Categorical covariates were transformed into sets of binary variables, with one variable from each set omitted to represent the reference group. Patients with missing data were included in the analysis, but missing values were imputed as follows: race (n = 17) was replaced with other race, pre-transplant serum creatinine (n = 276) was replaced with serum creatinine <2 mg/dL, pre-transplantation hospitalization status (n = 11) was replaced with not hospitalized and donor age (n = 2) was replaced with mean donor age. The October 2003 release of the SRTR did not contain all data required for calculation of Model for End-Stage Liver Disease or Child–Pugh scores; therefore, hospitalization status at the time of listing (i.e. in the intensive care unit [ICU], hospitalized or not hospitalized) was used as an estimate of pre-transplant liver disease severity. Demographic and baseline clinical characteristic data were summarized with descriptive statistics. All analyses were performed using SAS Release 8.2 (SAS Institute, Cary, NC; Windows XP).
A total of 11670 patients between the ages of 18 and 78 years (mean age 49.7 ± 10.8 years), were included in the study: 778 (6.7%) infected with HBV, 3463 (29.7%) with HCV and 7429 (63.7%) with diseases other than hepatitis B or C. Demographic and clinical characteristics for the population are presented in Table 1.
Table 1. Demographic and baseline clinical characteristics
Hepatitis B (N = 778)
Hepatitis C (N = 3463)
No viral hepatitis (N = 7429)
Hepatitis B vs. C
Hepatitis B vs. no viral
Hepatitis C vs. no viral
Mean ± SD
48.2 ± 10.8
50.3 ± 8.7
49.5 ± 11.6
Weight (kg) Mean ± SD
78.3 ± 19.1
83.0 ± 18.5
79.5 ± 18.6
Pre-transplant serum creatinine ≥ 2 mg/dL
Pre-transplant hospitalization status
Mean ± SD
38.4 (± 16.8)
37.9 (± 17.0)
37.3 (± 17.4)
Cold Ischemia Time
Mean recipient age was similar across diagnostic groups (from 48.2 ± 10.8 to 50.3 ± 8.7 years). There were fewer female recipients in the HBV group (21.5%) compared to the HCV and no viral hepatitis groups (34.4 and 42.7%, respectively). The HBV group had fewer Caucasian patients (61.2%) than the HCV and no viral hepatitis groups (87.7 and 88.8%, respectively); this was due to a higher percentage of patients with race classified as others (i.e. predominantly of Asian ethnicity) in the HBV group (30.6 vs. 5.7 and 3.7%, respectively). More patients infected with HBV were in the ICU prior to transplantation (28.5%) compared to the other diagnostic groups (18.6% of patients with HCV and 23.5% of patients without viral hepatitis). The proportion of patients in each group with pre-transplant renal dysfunction (i.e. serum creatinine ≥ 2 mg/dL) was similar. Donors for the three groups were similar in terms of age (mean age from 37.3 ± 17.4 to 38.4 ± 16.8 years) and gender distribution (predominantly male). The majority of patients overall received organs from deceased donors; 2.6–4.0% of patients across groups had undergone living donor transplants. The median follow-up period post-transplantation was 2.98 years; the maximum follow-up was 8.3 years.
Graft and patient survival
Graft survival rates were significantly lower in patients infected with HCV compared to patients without viral hepatitis (p = 0.0019; 4-year survival of 71.3% vs. 75.1%; Figure 1); differences in graft survival between patients with HBV (75.5% at 4 years) and HCV approached statistical significance (p = 0.0561). There were no significant differences in graft survival between HBV-infected patients and patients without viral hepatitis (p = 0.75).
Patient survival was significantly lower in recipients with HCV compared to HBV (p = 0.0011) and no viral hepatitis (p < 0.0001; 4-year survival of 75.7, 81.7, and 79.2%, respectively; Figure 2). Patient survival rates for recipients in the HBV and no viral hepatitis groups did not differ significantly (p = 0.209).
Factors affecting survival outcomes
Graft loss Factors associated with graft loss based on Cox regression analysis are presented for each diagnostic group in Table 2. Only covariates found to be significant for at least one diagnostic group are presented.
Table 2. Covariates for graft loss based on pre-transplant diagnosis
Other variables included in the multivariable model (and reference group) were recipient age (per year), CMV −D/−R (−D/+R; +D/+R; −D/+R), donor gender female (donor gender male), donor age 21–30 years (11–20 years).
None of the model covariates were found to be independently associated with an increased risk of graft loss in patients infected with HBV. MMF and other race were associated with significant reductions in the risk of graft loss.
In HCV-infected recipients, the strongest predictor of graft loss was the age of the donor. Transplantation with organs from donors between the ages of 41 and 50, 51 and 60 and >60 years was associated with a linear increase in the risk of graft loss (HR = 1.67, 1.86 and 2.21, respectively; p < 0.001). Serum creatinine ≥ 2 mg/dL, African American race and pre-transplant ICU stay also were significant risk factors for graft loss. MMF and other race were associated with reductions in the risk of graft loss in patients infected with HCV.
In patients without viral hepatitis, the most prominent risk factor for graft loss was donor age >60 years (HR = 1.89, p < 0.001). With the exception of donors between the ages of 21 and 30 years, each donor age cohort was independently associated with a significantly increased risk for graft loss, with a progressive increase in risk for each decade of donor age. Additional risk factors for graft loss included pre-transplant ICU stay, serum creatinine ≥ 2 mg/dL, African American race, pre-transplant hospitalization, cold ischemia time ≥12 h, and more recent transplant year. MMF was associated with a reduced risk of graft loss in patients without viral hepatitis.
Death Risk factors for death (Table 3) in patients with HBV included African American race, serum creatinine ≥ 2 mg/dL and recipient age. MMF was associated with a reduced risk of death.
Table 3. Covariates for death based on pre-transplant diagnosis
Other variables included in multivariable model (and reference group): CMV −D/−R (−D/+R; +D/+R; −D/+R), donor gender female (donor gender male), donor age ≤ 10 years, 21–30 years, and 31–40 years (11–20 years).
Donor age was the strongest predictor of death in recipients with HCV. As for graft loss, the risk of death increased with each decade of donor age, starting with donors 41–50 years (HR = 1.71, p < 0.001) and increasing (HR = 1.84 [p < 0.001] for donors 51–60 years), such that the risk of death among patients who received organs from donors over the age of 60 years was almost twice that of the reference group (i.e. donor age between 11 and 20 years; HR = 1.92, p < 0.001). Additional risk factors for death included serum creatinine ≥ 2 mg/dL, pre-transplant ICU stay, African American race, pre-transplant hospitalization and recipient age. Variables associated with a reduced risk of death included other race and MMF.
Having a donor over the age of 60 years was the strongest risk factor for death in patients without viral hepatitis (HR = 1.73, p < 0.001), followed by pre-transplant ICU stay, serum creatinine ≥ 2 mg/dL, African American race, pre-transplant hospitalization, donor age 51–60 years, donor age 41–50 years, cold ischemia time ≥ 12 h, more recent transplant year and recipient age. MMF was associated with a decreased risk of death.
Analysis of donor age
The association between the age of the donor and poor outcomes in patients with HCV and without viral hepatitis, but not in patients with HBV, was examined further using Kaplan–Meier analysis of graft survival, which confirmed distinct patterns among the three diagnostic groups.
In patients infected with HBV, no relationship between increasing donor age by decade and graft survival was evident (Figure 3A), although graft survival rates dropped to 77.1% within the first year and to 65.2% by 4 years following transplantation in patients who received organs from donors >60 years of age.
As illustrated in Figure 3B, there were distinct differences in graft survival, already apparent within the first year following transplantation, for patients with HCV who received organs from donors over the age of 40 years. Four-year graft survival was 67.5% for donor age 41–50 years, 61.5% for donor age 51–60 years and 57.4% for donors >60 years of age. By comparison, graft survival rates of 79.6, 78.4 and 74.9% were observed when donors were 11–20, 21–30 and 31–40 years of age, respectively.
In patients with diagnoses other than viral hepatitis (Figure 3C), there were small (up to 2%), progressive decreases in graft survival with increasing donor age up to 60 years (from 79.7% at 4 years for donor ages 11–20 years to 73.2% for donor ages 51–60 years) and a sharp decrease in graft survival when the donor was over the age of 60 years (62.8% at 4 years).
The shortage of organ donors requires the use of so called marginal organs. Marginal organs include those obtained from non-heart-beating donors, from donors with hepatotropic viral infection (e.g. hepatitis B or C), organs with varying degrees of steatosis and, most importantly, organs from elderly donors. In expanding the donor pool, specific factors about the recipient, the donor, peri-operative events and immunosuppressive management need to be considered relative to one another to identify paradigms that optimize long-term survival. For example, an aging population as a growing source for organ donation requires that we consider liver allografts from elderly donors not by age alone, but in terms of the donor's medical history, the status of the organ and by controlling peri-operative factors, such as limiting cold ischemia time (11).
Matching specific donor factors to specific recipient factors may be important in determining long-term outcomes and disease recurrence; however, the impact of factors affecting patient and graft survival may vary based on underlying liver disease. Thus, identifying patterns of risk factors based on the underlying disease of the transplant recipient may help to determine populations that may be more amenable to organs from marginal donors. In this registry database analysis, we found that factors affecting survival differed based on the underlying disease of transplant recipients, and patients with HBV distinguished themselves from the other populations in terms of an overall decreased relative risk for poor outcomes.
Previous studies have demonstrated worse survival outcomes in HCV-positive compared to HCV-negative liver transplant recipients (4,12). Our study confirms these findings, and also demonstrates the relative progress in management of patients undergoing transplant for HBV cirrhosis. Just over a decade ago, many centers considered HBV a contraindication for transplantation, but more recently, improvements in prevention and treatment of recurrent HBV post-transplantation have increased graft and patient survival to levels observed in patients undergoing transplantation for other indications [2,3]. Our study demonstrates survival rates for patients with HBV that were similar to patients without viral hepatitis, and superior to those for patients infected with HCV.
In multiple regression analyses, none of the risk factors for graft loss were common to all three diagnostic groups; however, African American race (13,14), pre-transplant renal dysfunction (i.e. serum creatinine ≥ 2 mg/dL) and older recipient age were associated with an increased risk of death, regardless of pre-transplantation diagnosis.
In patients with HBV, the age of the donor was not associated with an increased risk of graft loss or death, although there were trends for worse outcomes when donors were over the age of 60 years (graft loss: HR = 1.45, p = 0.17; death: HR = 1.47, p = 0.19). The absence of a statistically significant, negative effect of older allografts in this specific population may have been due to the relatively small number of patients (approximately 7% of the study population) who underwent transplant for HBV cirrhosis in this study. Univariate analysis examining graft survival based on the age of the donor confirmed decreased survival in HBV-infected patients who received organs from donors >60 years of age.
The association between organs from older donors and decreased survival in patients with HCV is well documented (4,6–8,12–16). Studies have demonstrated a relationship between accelerated post-transplantation fibrosis progression in HCV-positive recipients and allografts obtained from older donors (12,15). In our analysis, the age of the donor surpassed all other risk factors for poor graft and patient survival in patients with HCV; most striking was the impact donor age extended to what most would regard as relatively young donors (i.e. <50 years of age). Donor age between 41 and 50 years was associated with a 67% increase in the risk of graft loss; the risk increased to 86% when donors were between 51 and 60 years of age, and was more than 2-fold (HR = 2.21) when donors were >60 years of age. We also found that donor age significantly impacted survival outcomes in patients without viral hepatitis, especially when donors were >60 years of age (HR = 1.89 for graft loss). The relative risks observed in patients without viral hepatitis were smaller in the cohorts including donors 41–50 (HR = 1.22) and 51–60 years of age (HR = 1.34) compared to those observed for patients with HCV. Russo and colleagues also reported an association between donor age and graft survival in patients without HCV (16). In their analysis, donor age ≥60 years was associated with lower rates of graft survival at 1 year than donor age <40 years only in patients with HCV; however, 5-year graft survival was worse for patients who received organs from older donors, regardless of underlying disease (i.e. HCV, cholestatic liver disease or alcoholic liver disease).
Another risk factor common to patients with HCV and patients without viral hepatitis was pre-transplant liver disease severity, as patients in the ICU prior to transplantation had a higher risk of graft loss and death compared to those who were not hospitalized. This finding is consistent with the previously noted association between severity of illness, as measured by Child–Pugh score, and outcomes in HCV recipients . Although a greater percentage of patients with HBV in this study were in the ICU prior to transplantation, disease severity was not associated with a significant effect on graft or patient survival.
The current paradigm in liver transplantation in terms of the impact of immunosuppression on hepatitis C recurrence and outcomes is that more immunosuppression is ‘bad’. However, this impression is largely based on the negative impact of additional immunosuppression used to treat acute cellular rejection on patient and graft survival (17–20). All patients included in this study were discharged on immunosuppression including tacrolimus and corticosteroids, with or without MMF. For all three diagnostic groups, immunosuppression including MMF was associated with reduced risk of graft loss and death compared to tacrolimus and corticosteroids alone. Conclusions regarding the use of MMF in this study, however, are limited by the absence of information regarding dosing or duration of exposure to any of the immunosuppressants in either of the regimens included as covariates in the regression models. This shortcoming of our analysis is due to the fact that the SRTR database contains no information on drug dosing, and immunosuppression regimen as a covariate was based on an intent-to-treat analysis of discharge medications. Furthermore, because of the historical nature of this sample, potential long-term corticosteroid use in earlier years also may have contributed to the treatment benefit in patients with HCV (21,22). Nevertheless, the positive outcomes associated with the addition of MMF to tacrolimus and corticosteroids in this study support the position that the degree of immunosuppression in and of itself is not deleterious to patients infected with HCV (23).
Further potential limitations associated with analyses of registry data should be considered in interpreting these results. Variables that are important to outcomes, such as comorbid conditions, may not be reported for all patients. In this analysis, it was not possible to include hepatocellular carcinoma (HCC) as a covariate, as it tended to be captured consistently only as a primary diagnosis. As of 2002, information regarding the presence of HCC has been captured more uniformly in the database, and should be available for future analysis. It is possible that HBV-infected patients were seen predominantly by a small number of focused centers, where the standard of care and center-specific treatment effects may have played a larger role in the outcomes. Other potential limitations are the accuracy and completeness of registry data. A recent study demonstrated that aggregate patient survival data reported to the Organ Procurement and Transplantation Network (OPTN) by individual transplant centers was accurate compared to OPTN data supplemented by mortality data from the Social Security Death Master File (24,25). Completeness of data from OPTN is determined by comparing expected follow-up to reported follow-up reports and, where appropriate, the SRTR uses supplemental sources to try to provide more complete follow-up data (24,25). Finally, individual reporting bias by transplant centers may result in under-reporting or over-reporting of outcomes; however, it is reasonable to expect that this would occur to similar degrees for each of the groups compared and, therefore, would not affect the overall conclusions of this study.
In conclusion, there are clear differences in the extent and impact of risk factors for decreased patient and graft survival, depending on the underlying liver disease of the recipient. This is particularly striking for donor age, which was the most important predictor of poor outcomes in patients with HCV, but had little or no impact in patients with HBV. Unfortunately, changing allocation schemes to give patients with HCV cirrhosis priority for organs from younger donors would likely find little support from the transplant community at this time. Expanded criteria donors have been defined for renal transplantation, and potential renal transplant recipients have agreed prior to listing to be considered for such expanded donor kidneys (26). Perhaps in future, this same strategy will be applied to liver transplantation, with patients agreeing to accept expanded criteria donor organs during the informed consent process. Findings from this study also indicate that triple immunosuppression therapy including MMF plays a favorable role in survival outcomes for patients with HCV.
The improvement in outcomes for patients transplanted for HBV disease represents one of the great triumphs of the 90s. We can only hope that the next decade will bring similar improvements in outcomes of patients transplanted for HCV disease.
The data reported here have been supplied by the United Network for Organ Sharing and University Renal Research and Education Association under contract with the Department of Health and Human Services. The interpretation and reporting of these data are the responsibility of the authors and do not represent an official policy or interpretation of the United States Government or any of its representatives. The analyses reported here have been conducted in ProSanos Corporation facilities, and were funded by Roche Laboratories, Inc. Dr. Lake receives grant support from Roche Laboratories, Inc. Dr. Wiesner serves as an advisor to Roche Laboratories, Inc. Drs. Lake and Wiesner serve as consultants to ProSanos Corporation. This work was presented in part at the 10th Annual Congress of the International Liver Transplantation Society, Kyoto, Japan.