Conflicts of Interest: Dr. O'Leary advises and is on the speakers' bureau for Vertex. She also advises Bristol Meyers Squibb. This study was supported by the National Institutes of Health (grant T32 DK060414 to Jennifer C. Lai) and the University of California San Francisco Liver Center (grant P30 DK026743 to Jennifer C. Lai and Norah A. Terrault).
In an effort to expand the supply of donor livers, transplant centers have increasingly used expanded criteria donors, including older donors, donors after cardiac death, and donors at risk for the transmission of viral diseases. For hepatitis C virus (HCV)–infected candidates awaiting transplantation, a liver from a hepatitis C virus antibody–positive donor [HCV(+)D] is an option. Over the last 5 years, approximately 3% of all deceased donor liver grafts in the United States have been procured from HCV(+)Ds.1 Although these donors represent a small percentage of all donors, the number of liver transplants from HCV(+)Ds has tripled over the last decade,2 and this makes it imperative for us to understand the precise risk associated with these grafts (if there is any).
Several studies have evaluated the association between the donor HCV status and graft loss with data from the United Network for Organ Sharing (UNOS)/Organ Procurement and Transplantation Network (OPTN) registry.2-4 Although no effect was seen, these studies did not capture critical factors known to be associated with graft loss in HCV-infected recipients, including acute rejection and cytomegalovirus (CMV) infection. In addition, these studies were not able to evaluate HCV disease severity, which is an outcome of particular importance because of the morbidity and costs associated with the management of recurrent HCV. Among the single-center studies including histological follow-up, some3, 5 but not all6, 7 reported more advanced fibrosis in recipients of HCV(+)D liver grafts versus recipients of hepatitis C virus antibody–negative donor [HCV(−)D] grafts. One of these studies suggested that the risk of advanced fibrosis was further increased with older HCV(+)Ds (≥50 years) versus younger HCV(+)Ds (<50 years),5 but all these studies were limited by small sample sizes and relatively short follow-up.
Because of the limitations of these previous studies, we used the multicenter Consortium to Study Health Outcomes in HCV Liver Transplant Recipients (CRUSH-C) database to evaluate the association between the donor HCV antibody status, the donor age, and the development of advanced fibrosis after liver transplantation.
CI, confidence interval; CMV, cytomegalovirus; CRUSH-C, Consortium to Study Health Outcomes in HCV Liver Transplant Recipients; HCV, hepatitis C virus; HCV(−)D, hepatitis C virus antibody–negative donor; HCV(+)D, hepatitis C virus antibody–positive donor; HR, hazard ratio; MELD, Model for End-Stage Liver Disease; OPTN, Organ Procurement and Transplantation Network; UNOS, United Network for Organ Sharing.
PATIENTS AND METHODS
Patient Population and Data Definitions
The cohort included in CRUSH-C has been previously described.8 In brief, CRUSH-C is a multicenter study group that includes 5 high-volume and experienced transplant centers in the United States: (1) the University of California San Francisco, (2) Baylor University Medical Center, (3) NewYork-Presbyterian Hospital/Columbia University Medical Center, (4) Virginia Commonwealth University, and (5) the University of Colorado. Data from all adult patients infected only with HCV who received a primary liver transplant between March 1, 2002 and December 31, 2007 were retrospectively collected from electronic health records and manual chart reviews. Donor data were obtained from the UNOS/OPTN registry. The donor age was available for 97 of the HCV(+)D grafts (98%) and 1097 of the HCV(−)D grafts (99%). For this study, patients were excluded from the analyses if they experienced graft loss within 30 days of transplantation (n = 44), they had negative HCV RNA findings immediately after transplantation in the absence of posttransplant antiviral treatment (n = 61), or their donor HCV antibody status was missing or not reported in the UNOS/OPTN registry (n = 58).
Donor HCV antibody testing was performed at the donor's local hospital, and the method of HCV antibody testing was not recorded in the UNOS/OPTN registry. Posttransplant variables were obtained from manual chart reviews performed by each center, and they included acute cellular rejection treated with high-dose bolus corticosteroids or anti-lymphocyte therapy, treated CMV infection, posttransplant HCV treatment, and sustained virological response.
The standard protocol at each of the 5 centers included annual liver biopsies for HCV-infected liver transplant recipients. However, in the event that a patient required a liver biopsy for cause, the protocol liver biopsy may have been deferred beyond the patient's annual transplant anniversary date. Therefore, both protocol and for-cause biopsy data were included in the analyses. Four centers used the Batts-Ludwig staging system for fibrosis,9 and 1 center used the Ishak staging system.10
Although the biopsying of HCV(+)D grafts was routine, because of the retrospective nature of this study, these data were inconsistently recorded in the medical record and were, therefore, not accessible for the purposes of this study. However, the standard practice at each of the 5 centers during the study period was to accept livers only from HCV(+)Ds without any evidence of fibrosis on frozen sections. The acceptance of grafts from HCV(+)Ds with minimal fibrosis was considered if the donors did not have any additional characteristics that might confer a greater risk of poor posttransplant outcomes (eg, older age or donation after cardiac death).
Each center used a standard immunosuppression regimen; however, the immunosuppression regimens were not uniform among the sites. Data regarding the use of (1) a calcineurin inhibitor (tacrolimus or cyclosporine) at the last follow-up and (2) low-dose corticosteroids as maintenance immunosuppression (>3 months after transplantation) were recorded. None of the sites routinely used induction therapy.
Primary Predictor and Endpoints
The primary predictor was donor HCV antibody status, which was obtained from the UNOS/OPTN registry. The primary endpoint was advanced fibrosis in the recipient; this was defined as the first finding of bridging fibrosis or cirrhosis on a biopsy sample (Batts-Ludwig stages 3-4 or Ishak stages 4-6) after transplantation. Biopsy samples were read locally by each institution's pathologist. Patients without a single posttransplant liver biopsy sample (n = 108) were excluded from the analyses for only this outcome. Patient mortality and graft loss were also assessed.
Characteristics were reported as means and standard deviations (continuous variables) or as numbers and percentages (categorical variables). The Student t test and chi-square tests were used to compare baseline characteristics between HCV(+)D and HCV(−)D groups. Unadjusted survival rates were calculated with Kaplan-Meier methods and were compared with the log-rank test. A Cox regression analysis was used to evaluate the association between the donor HCV antibody status and the outcome of interest. Only patients who had at least 1 posttransplant biopsy sample were included in analyses evaluating the outcome of advanced fibrosis. Other variables that were associated with P < 0.2 in the univariate analysis were then evaluated in the multivariate model. To build the final multivariate model, we eliminated variables with the highest P values in a backward, stepwise fashion until all the remaining variables yielded P < 0.05.
Posttransplant HCV treatments and episodes of treated acute cellular rejection were evaluated in the Cox models as time-varying covariates. Immunosuppression medications (eg, cyclosporine, tacrolimus, and/or low-dose steroids for maintenance immunosuppression) were tested separately in the final model and were included if they were significant at P < 0.05. All final models were adjusted for the center effect to account for any potential unmeasured center-specific confounders.
Statistical analyses were performed with Stata 11 (Stata Corp., College Station, TX). This study was approved by the institutional review boards at each of the 5 centers.
At the 5 centers, 1206 patients met the inclusion criteria for this study, and 99 recipients (8%) received an HCV(+)D graft. The mean follow-up time was shorter for HCV(+)D graft recipients (2.6 ± 1.5 years) versus HCV(−)D graft recipients (3.3 ± 1.8 years, P < 0.001). The characteristics of the cohort members and their donors are displayed in Table 1 by the donor HCV antibody status. Compared to the HCV(−)D graft recipients, the recipients of HCV(+)D grafts were older, but they otherwise were similar with respect to sex, race, laboratory Model for End-Stage Liver Disease (MELD) score, and hepatocellular carcinoma. HCV(+)D graft recipients were less likely to be on a tacrolimus-based immunosuppression regimen at the last follow-up. HCV(+)Ds were older, more likely to be African American, and more likely to die of stroke or other causes, and they experienced longer cold ischemia times, which resulted in a donor risk index that was significantly higher for HCV(+)Ds versus HCV(−)Ds.
Table 1. Characteristics of HCV-Infected Liver Transplant Recipients and Their Donors
HCV(+)D Grafts (n = 99)
HCV(−)D Grafts (n = 1107)
The data are presented as means and standard deviations.
Treatment with either pulse-dosed steroids or anti-lymphocyte therapies.
Indicating some patients were not on a calcineurin inhibitor.
Eight HCV(+)D graft recipients (8%) and 100 HCV(−)D graft recipients (9%) did not have a single posttransplant biopsy sample (P = 0.75). Among those who had at least 1 posttransplant biopsy, the mean number of biopsies was 3 for both groups (P = 0.38). The mean times to the first biopsy for HCV(+)D graft recipients and HCV(−)D graft recipients were 5.1 ± 6 and 7.1 ± 8 months, respectively (P = 0.02).
Donor HCV Status and Advanced Fibrosis
Advanced fibrosis occurred in 32% of HCV(+)D graft recipients (29/91) and in 28% of HCV(−)D graft recipients (278/1007, P = 0.39). The unadjusted 1- and 3-year rates of advanced fibrosis were 14% and 48% for recipients of HCV(+)D grafts and 7% and 33% for recipients of HCV(−)D grafts (log-rank P = 0.01; Fig. 1A).
In the univariate analysis, transplantation with an HCV(+)D graft was associated with a 62% overall increased risk of advanced fibrosis [95% confidence interval (CI) = 1.10-2.38, P = 0.01; Table 2]. Other donor factors that were associated with advanced fibrosis in the univariate analysis (P < 0.2) were age, height, and stroke or other cause of death; African American race was associated with a decreased hazard (Table 2). In the multivariate model evaluating only donor factors, the donor HCV antibody status and age as well as African American race remained significantly associated with the primary outcome (P < 0.05; Table 2). Further adjustments for significant recipient characteristics again did not substantially change the increased hazard associated with an HCV(+)D status for advanced fibrosis [hazard ratio (HR) = 1.58, 95% CI = 1.05-2.36, P = 0.03; Table 2 and Fig. 1B].
Table 2. Univariate and Multivariate Models Evaluating the Association Between Donor Factors and Advanced Fibrosis
Model With Donor, Recipient, and Posttransplant Factors†
HR (95% CI)
HR (95% CI)
HR (95% CI)
Donor height, stroke or other cause of death, and split/partial livers were also associated with a prespecified P value <0.20 in the univariate analysis, but they were not statistically significantly associated with advanced fibrosis in the multivariate analyses. Recipient obesity, donation after cardiac death, cold ischemia time, and share region were also evaluated, but they were not significantly associated with advanced fibrosis in the univariate or multivariate analyses.
Adjusted for the center effect.
Donor age (per 10 years)
Donor race (African American)
Recipient sex (female)
Posttransplant HCV treatment
Treated CMV infection
Because of previous single-center data suggesting worse outcomes with older HCV(+)Ds,5 we next performed a stratified analysis evaluating the association between the donor age and the donor HCV antibody status. Using the mean donor age of 45 years, we found no significant difference in the cumulative rate of advanced fibrosis between HCV(+)D graft recipients and HCV(−)D graft recipients (HR = 1.34, 95% CI = 0.70-2.54, P = 0.38; Fig. 2A) with grafts from donors ≤45 years. Although adjustments for the other significant predictors of advanced fibrosis among donors ≤45 years old, including recipient African American race and donor cause of death, decreased the hazard associated with the donor HCV antibody status, this reduction was not statistically significant (HR = 0.94, 95% CI = 0.47-1.87, P = 0.85). In contrast, among recipients of grafts from donors >45 years old, there was a higher rate of advanced fibrosis (HR = 1.78, 95% CI = 1.10-2.88, P = 0.02; Fig. 2B). Multivariate adjustments for other significant predictors of advanced fibrosis among donors >45 years old, including donor African American race, posttransplant HCV treatment, and treated CMV infection, did not substantially change the increased hazard of advanced fibrosis associated with the donor HCV antibody status (HR = 1.76, 95% CI = 1.06-2.93, P = 0.03).
Donor HCV Antibody Status and Patient and Graft Survival
By the end of follow-up, 30 HCV(+)D graft recipients (30%) and 295 HCV(−)D graft recipients (27%) experienced graft loss (P = 0.43). There was no difference in the proportions of HCV(+)D graft recipients and HCV(−)D graft recipients who underwent retransplantation (3% versus 4%, P = 0.64). The unadjusted patient survival rates for HCV(+)D graft recipients and HCV(−)D graft recipients were 90% and 91%, respectively, at 1 year and 73% and 79%, respectively, at 3 years (P = 0.11). The unadjusted graft survival rates for HCV(+)D graft recipients and HCV(−)D graft recipients were 89% and 90%, respectively, at 1 year and 69% and 77%, respectively, at 3 years (P = 0.10). In the multivariate analysis, the donor HCV antibody status was not an overall independent predictor of either mortality (HR = 1.28, 95% CI = 0.85-1.94, P = 0.24) or graft loss (HR = 1.17, 95% CI = 0.79-1.72, P = 0.44).
Donor HCV Status and Response to HCV Treatment
In all, 374 patients (31%), including 34 HCV(+)D graft recipients (34%) and 340 HCV(−)D graft recipients (31%, P = 0.45), received HCV treatment after transplantation. Among the patients who were treated, the overall rates of sustained virological response were similar between HCV(+)D graft recipients and HCV(−)D graft recipients (18% versus 24%, P = 0.38). Among the treated patients, 26 of the 34 HCV(+)D graft recipients (77%) and 257 of the 340 HCV(−)D graft recipients (76%) received HCV treatment before the onset of advanced fibrosis (P = 0.91). Three of the 26 HCV(+)D graft recipients (12%) and 60 of the 257 HCV(−)D graft recipients (23%) who received HCV treatment before the onset of advanced fibrosis progressed to advanced fibrosis after treatment (P = 0.16). In the multivariate analysis, the receipt of HCV treatment was associated with a decreased hazard of advanced fibrosis (HR = 0.72, 95% CI = 0.53-0.97, P = 0.03), death (HR = 0.54, 95% CI = 0.40-0.73, P < 0.001), and graft loss (HR = 0.63, 95% CI = 0.48-0.82, P = 0.001).
Our multicenter cohort study demonstrates that HCV(+)D grafts are not risk-free. Although previous UNOS-based studies found no association between the donor HCV antibody status and survival,2-4, 6 these studies could not evaluate the clinically significant outcome of fibrosis progression or account for key posttransplant events (ie, acute rejection and HCV treatment). Specifically, we found that liver transplantation with HCV(+)D grafts was associated with a 58% overall increased adjusted risk of advanced fibrosis in comparison with HCV(−)D grafts.
Importantly, this increased risk of advanced fibrosis associated with HCV(+)D grafts was most apparent among recipients of grafts from donors more than 45 years old; this finding persisted even after adjustments for other independent predictors of advanced fibrosis. This substantiates findings from a previous single-center study that reported more severe fibrosis in recipients of older HCV(+)D grafts (≥50 years; n = 9) versus recipients of younger HCV(+)D grafts (<50 years; n = 20).5 Because the regenerative capacity of hepatocytes is age-dependent,11 hepatocytes that have previously been injured by chronic inflammation in an older donor may have a decreased ability to repair and proliferate in comparison with hepatocytes from a younger donor. Additionally, age may be a surrogate for the presence of hepatic fibrosis in HCV(+)Ds, which has recently been reported to be an important predictor of severe HCV recurrence.12 Lastly, older donor livers may have higher rates of baseline clinical fibrosis before implantation.
Consistent with UNOS-registry based studies,2-4, 6 the rates of mortality and graft loss did not differ significantly with the HCV antibody status in our cohort. One explanation is that recipients of HCV(+)D grafts live for several years after the development of advanced recurrent disease, so differences in patient and graft survival require a longer duration of follow-up to become apparent. Alternatively, increased attention on the part of the provider to the posttransplant management of HCV(+)D graft recipients with rapid recurrence, including changes in immunosuppression, the less aggressive treatment of acute rejection, and the early initiation or longer duration of antiviral treatment, may attenuate differences in graft survival between HCV(+)D graft recipients and HCV(−)D graft recipients. Additional studies with long-term follow-up and more granular posttransplant event data are needed to further explore these hypotheses.
In agreement with previous UNOS-based studies, we found that there were significant differences in the baseline characteristics of HCV(+)Ds and HCV(−)Ds. This perhaps reflects the demographics of HCV antibody–positive individuals in the United States because in comparison with individuals of other races and ethnicities, African Americans have disproportionately higher rates of HCV antibody positivity and tobacco use (ie, they may be at increased risk of cerebrovascular death).13, 14 In addition, a recent study has demonstrated that in comparison with HCV antibody–negative individuals, HCV antibody–positive individuals (with or without viremia) in all age groups whose members are 30 years old or older (eg, 30-44, 44-59, and ≥60 years) experience a higher mortality rate ratio.13 These factors, in addition to their high proportion of the national share distribution (which is likely due to their HCV antibody positivity), explain the higher donor risk index associated with grafts from HCV(+)Ds.
Unlike the donor risk index, however, grafts from African American donors are associated with a decreased risk of advanced fibrosis in our analyses. Although these grafts are associated with a higher risk of graft loss in the donor risk index15 (which includes recipients with all etiologies of liver disease), a growing body of evidence suggests that the donor risk index may not be predictive of graft loss in HCV-infected liver transplant recipients.16, 17 In the nontransplant setting, African Americans with chronic HCV infections experience slower rates of progression to cirrhosis,18, 19 and whether these same protective immunological and genetic factors play a role in the posttransplant setting warrants a more detailed investigation.
One limitation of our study was the lack of detailed donor data, including biopsy samples and HCV RNA from donors. Unfortunately, data on donor fibrosis were not reported in the UNOS registry, which was the source of our donor data; only macrovesicular steatosis and microvesicular steatosis were reported, and these data were available for <20% of our entire cohort. In a European-based multicenter study, 43% of the 63 HCV(+)Ds were aviremic, and more than 50% had stage 1 fibrosis.12 However, the mean donor age of that cohort was 57.3 years (ie, those donors were more than a decade older than the donors in our cohort), so these findings may not be applicable to a United States–based cohort. In addition, the standard practice at each of the 5 centers was to accept only grafts from HCV(+)Ds without any evidence of fibrosis. The acceptance of grafts from HCV(+)Ds with minimal fibrosis was considered only for donors without any additional risk factors for poor posttransplant outcomes, such as older donor age or donation after cardiac death. Therefore, we believe that it is unlikely that there was a substantial difference in the baseline fibrosis between younger and older HCV(+)Ds. A second limitation was our lack of knowledge of the HCV viremic status of the donors. Because some of the HCV(+)Ds in our study were not viremic and were not at risk for potentially affecting HCV disease progression in the recipients, we would expect the reclassification of these aviremic HCV(+)D graft recipients to further strengthen the association between the donor HCV antibody status and posttransplant fibrosis severity.
Our study has important implications for the use of HCV(+)D grafts. We advise caution when grafts from HCV(+)Ds (especially older ones) are being used. Recipients of HCV(+)D grafts should undergo close monitoring for the early development of advanced fibrosis, and this monitoring approach may be tailored if future studies demonstrate that donor fibrosis or donor HCV viremia further modifies recipient outcomes. In light of more effective HCV therapy with direct-acting antiviral agents, further studies are warranted to determine whether early HCV treatment modifies this risk and whether recipients of HCV(+)D grafts should be selected according to their genotype and ability to tolerate antiviral therapy. Most importantly, because of the negative impact of advanced recurrent disease on health-related quality of life after transplantation20 and the higher risk with the use of HCV(+)D grafts, our study reminds us of the need to carefully balance the risk of wait-list mortality with the risk of posttransplant complications when HCV(+)D organs are being used.