The management of patients awaiting transplantation is a growing concern. This retrospective cohort study examined outcomes of hepatitis C virus (HCV)-infected kidney candidates who remain waitlisted. Records from 315 HCV+ kidney candidates evaluated between 1992 and 2002, were reviewed. A total of 300 (95.1%) patients were receiving renal replacement therapy at evaluation, median duration 48.2 + 4.3 months. The diabetes prevalence was 42.9% in HCV+ candidates, compared to 35.9% among 602 currently listed HCV− patients (p = 0.023). Liver disease, defined by abnormal hepatic biochemistry or histology, was observed in 59% patients. Median post-evaluation follow-up was 1440 ± 75 days; 138 candidates were transplanted. Kaplan–Meier survival was higher among transplanted than non-transplanted patients (p = 0.003). Of 177 patients not transplanted, 76 were delisted, mostly due to death (45%) and non-compliance (28%), infrequently because of liver disease (8.8%). A Cox regression model was fit to examine risk factors for waitlist death; only diabetes was associated (HR: 2.17, 95% CI: 1.1–4.1, p = 0.02), while liver disease was not. This study demonstrates that, in waitlisted HCV+ kidney patients, diabetes occurs with increased prevalence and is a major mortality determinant. Diabetic HCV+ kidney candidates are therefore a patient subgroup that requires frequent and careful reevaluation to ensure ongoing transplantability.
Approximately 6% of candidates die while awaiting kidney transplantation (1). Consequently, the management of patients on the kidney transplant waiting list is evolving into an increasingly important area of concern (2). Transplantation is the treatment of choice for hepatitis C virus (HCV)+ patients with end-stage renal disease (ESRD) (3,4). HCV infection, which has a prevalence of 7–40% in the transplant population, is associated with a greater risk of both allograft failure as well as mortality (5–11). Studies have suggested that the increased death rates may be attributable to comorbid complications of decompensated liver disease and a greater susceptibility to infections (5,8,12). However, some of these retrospective studies examined outcomes in patients in whom a diagnosis of hepatitis C infection was made only after transplantation. This may have resulted in under-recognition of more advanced liver disease at the time of transplantation, accounting for the high rates of decompensated liver disease in their populations.
On the other hand, an increased prevalence of diabetes mellitus has been observed in HCV+ individuals in the general population, as well as after transplantation in both kidney and liver recipients (13–18). Diabetes mellitus is also the leading cause of ESRD in the United States and a major predictor of mortality. The impact of either diabetes or the above HCV-related comorbidities on survival among waitlisted HCV+ kidney candidates is not well characterized. Thus, there were two purposes of this single-center study; first, to determine the prevalence of pre-existing diabetes in HCV+ patients undergoing evaluation for kidney transplantation in our transplant program and second, to examine factors affecting the transplant candidacy of HCV+ patients who remain on the waiting list.
Materials and Methods
We have retrospectively examined outpatient medical records from 315 HCV+ patients evaluated for kidney transplantation at the University of Pennsylvania Medical Center between July 1, 1992 and June 30, 2002. Hepatitis C infection had been diagnosed before transplantation in all patients following implementation in our center in July 1992 of a screening program for HCV infection for all potential kidney transplant candidates. We excluded from analysis (i) HCV+ candidates listed for combined liver–kidney transplantation and (ii) patients lost to follow-up whose outcomes were unknown or for whom information was missing.
All patients underwent standard outpatient pre-transplant evaluation. This involved evaluation by a transplant physician, transplant coordinator, financial coordinator and social worker. All patients underwent the customary battery of blood testing including hepatic biochemical tests of AST, ALT, albumin, PT and PTT. After the initial evaluation, all patients were discussed at a transplant evaluation meeting attended by all members of the transplant team. The transplant team then recommended additional studies required by candidates to complete their evaluation. All patients were placed on the active regional waitlist for kidney transplantation pending completion of their evaluation requirements. Of the 315 HCV+ patient cohort, 304 patients were listed for kidney alone, while 11 patients were placed on both the kidney and kidney–pancreas waitlists. Once listed, patients could be delisted in the event of transfer to another transplant program, death, interim development of preclusive comorbidities (such as advanced liver disease, severe coronary artery disease) or documentation of either patient non-compliance or return to active substance abuse.
Diabetic patients and all other candidates with a history suggestive of coronary artery disease underwent non-invasive cardiac stress testing. Coronary angiography was done on patients with a positive stress test and revascularization procedures performed as indicated. Other selective consultants were consulted as clinically indicated.
Diagnosis and screening of HCV and evaluation of liver disease
Commencing July 1992, all patients were screened for HCV. Initially, samples from all evaluated patients were screened for the presence of HCV antibodies using a second-generation enzyme linked immunoassay (EIA-2; Abbott Laboratories, Chicago, IL); positive samples by the EIA-2 assay were confirmed by RIBA until 1998, when the EIA-2 assay was replaced by a third-generation assay. Patients with positive tests were considered to have HCV infection. Since 1999, qualitative HCV RNA screening by reverse-transcriptase polymerase reaction (RT-PCR) has routinely been performed in all potential kidney candidates regardless of their HCV antibody status. During the era of routine HCV RNA measurement, the presence of either a positive HCV antibody with a positive HCV RNA or a positive HCV RNA in the absence of a positive anti-HCV has been considered to reflect HCV infection. An evaluation by a hepatologist and subsequent liver biopsies were performed at the discretion of the kidney transplant team between 1992 and 1997, generally reserved for patients with abnormal hepatic biochemistries. Since 1998, a liver biopsy has been incorporated as part of the customary pre-transplant evaluation of patients infected with hepatitis C. Following the implementation of liver biopsies as standard protocol, liver histology has consistently been graded by one hepatopathologist according to the METAVIR scoring system (19). Patients with evidence of significant bridging fibrosis or cirrhosis were deemed to be unsuitable for kidney transplantation alone. The use of interferon-α for treatment of HCV-related liver disease has been at the discretion of the consulting hepatologist. No patients received ribavirin.
Additionally, serological testing for hepatitis B surface antigen, anti-hepatitis B surface antibody (anti-HBsAb) and anti-hepatitis B core antibody (anti-HBcAb) was performed in every patient. Other routine virologic screening included CMV IgG, EBV titers, HIV Ab by ELISA.
Additional data collected
Candidate age and weight at time of evaluation, gender, race, the presence or absence of pre-existing diabetes mellitus and a family history of diabetes, pre-transplant weight, cause of ESRD, dialysis duration and history and previous renal transplant history, were recorded. Additional risk factors for HCV such as intravenous drug abuse and blood transfusions were documented, as well as ALT levels at the time of evaluation and liver biopsy results when performed. A liver biopsy, when performed, was considered to be histologically positive if there was any degree of inflammation or fibrosis in the tissue specimen. Liver disease was defined by the presence of any histologically positive liver biopsy and/or any ALT level > 35 U/L. Dates of evaluation, listing and death, where applicable, were documented. Cause of death of waitlisted patients, if known, was also ascertained.
All analyses were performed using SPSS software (SPSS, Inc., Chicago, IL). Means of continuous data were compared by ANOVA methods and categorical variables were compared using either the χ2 test or Fisher's exact test when data were sparse. Kaplan-Meier analysis was used to compare patient survival of transplanted with non-transplanted HCV-infected candidates. Since the waitlist comprised the entire study population at one time or another, the non-transplant cohort comprised all patients initially. Survival for patients never transplanted was calculated as the time from the date of listing until death or until the date of the current analysis; for patients who were transplanted, the time from listing until transplant was credited to the non-transplant group. These patients were then censored at the time of transplantation at which point they entered the transplant cohort and their survival time credited to the transplant. A second Kaplan-Meier analysis was performed on HCV+ kidney candidates not transplanted, where survival was compared according to the presence or absence of diabetes mellitus. A Cox regression model was fit to investigate risk factors for mortality among waitlisted patients not transplanted. This model included all baseline demographic variables (age, gender, race) as well as other covariates that have previously been associated with mortality in this population (diabetes, duration of ESRD, alcohol use, prior transplant, presence of liver disease). Finally, to better assess the influence of underlying liver disease on candidate mortality, a second model was fit that was limited to only listed patients who had undergone pre-transplant liver biopsy. A p-value ≤ 0.05 was considered to be significant. For this analysis, the liver histology covariate was dichotomized into the absence or presence of any liver injury.
Characteristics of HCV+ transplant candidates
Since 1992, pre-transplant evaluation in our transplant program has identified HCV infection in 315 kidney candidates for whom outcomes are known. As shown in Table 1, the majority of HCV-infected patients were black and male gender. At the time of transplant evaluation, almost all patients had received renal replacement therapy, for a median duration of 48.2 ± 4.3 months. The median duration of follow-up since initial evaluation was 1440 ± 75 days and 138 (44%) of the cohort received transplants. Risk factors for HCV infection included a history of intravenous drug and/or prior transplantation in approximately one-fourth of patients. Clinical evidence of liver disease was observed in 59% patients. For patients who underwent pre-transplant liver biopsy, the median time interval between the biopsy and subsequent transplantation, death or most current follow up was 13.2 ± 3.1 months (range: 0.6–209 months). The overall prevalence of diabetes mellitus was 42.9% among the HCV+ candidates listed for transplantation. This compares to a rate of 35.9% among 602 currently listed HCV− patients (p = 0.023). Among the HCV+ patients, diabetes was present in 48% of candidates below the age of 30 years and 31% candidates between 30 and 40 years of age (Figure 1). Above this age range, the frequency of diabetes increased in an age-related manner; up to 63% in candidates 60–80 years old.
Table 1. Patient characteristics (n = 315)
*Histologically positive liver biopsy and/or any ALT level >35 U/L.
Male gender (%)
Black race (%)
Pre-evaluation diabetes (%)
Prior transplant (%)
Any liver disease (%)*
History of alcohol or IV drug abuse (%)
Pre-transplant liver biopsy (%)
Age at evaluation (yrs)
45.6 ± 0.6
Duration of ESRD at evaluation (mos)
48.2 ± 4.3
Time on waiting list (days)
695 ± 82
Survival of all HCV+ kidney transplant candidates
The impact of transplantation on HCV+ patient survival in this cohort is shown in Figure 2. Survival among the 138 patients who were transplanted was significantly higher than among their 177 counterparts who were not transplanted (p = 0.003).
Outcome of untransplanted HCV+ patients on the waitlist
With increasing emphasis being placed on management of the waiting list and prevention of mortality among patients awaiting transplantation, we next more closely examined the outcome of waitlisted HCV+ patients who were not transplanted (n = 177). Of these patients, 75 (42%) had diabetes, 108 (61%) were black, 48 (27%) had received a prior kidney transplant and 91 (55%) had evidence of liver disease. The mean age of this population was 47.4 ± 0.7 years and the median duration of dialysis at the time of evaluation was 58.7 ± 6.1 months. Seventy-six (43%) of the 177 patients not transplanted were removed from the active waitlist during the observation period. As shown in Figure 3, the major reasons for removal from the waitlist were death (45%) and non-compliance (28%). Fewer than 10% of delisted patients were removed from the list because of advanced liver disease (cirrhosis in six patients and hepatoma in one patient). Among the 34 patients delisted because of death, cardiovascular disease was the commonest cause of mortality (n = 16), although infection (n = 4) and other causes (n = 5, including drug overdose, operative death) were implicated as well. Mortality due to liver failure occurred in only three patients and in a further six patients, the cause of death was not established.
Cox regression models
A Cox regression model was fit to ascertain risk factors for mortality among the waitlisted patients awaiting transplantation (Table 2). Of the several factors included in the model, only diabetes was significantly associated with death (HR: 2.17, 95% CI: 1.13–4.13, p = 0.02), while pre-existing liver disease was not. A survival curve demonstrating the adverse impact of diabetes on patient survival is shown in Figure 4. Black race was inversely associated with pre-transplant waitlist mortality, although did not reach statistical significance (HR: 0.58, 95% CI: 0.33–1.03, p = 0.064).
Table 2. Cox regression of risk factors for mortality among 177 HCV+ patients waitlisted for kidney transplantation
*Histologically positive liver biopsy and/or any ALT level >35U/L.
ESRD duration >48 mos (vs. <48 mos)
Anti-HBsAb+ (vs. HbsAb-)
Any liver disease* (vs. none)
Prior transplant (vs. none)
Alcohol use (vs. none)
Male gender (vs. female)
Age at evaluation (years)
Black race (vs. non-black)
Diabetes (vs. no diabetes)
While hepatic biochemistries were measured in this entire above cohort, only 77 patients in the non-transplant cohort underwent pre-transplant liver biopsy. To adjust for the possible confounding effect that undocumented severe histological liver disease could have on candidate mortality, we limited a sub-analysis to these 77 biopsied patients. Three (4%) patients were delisted because of advanced liver histology, while the remaining patients were all considered to have biopsy findings acceptable for kidney transplantation. By Cox regression analysis, diabetes mellitus (HR: 4.4, 95% CI: 1.1–17.8, p = 0.038) and dialysis duration (HR: 1.1, 95% CI: 1.0–1.1, p = 0.036) were significantly associated with waitlist mortality in this population. Overall, 13 (17%) of the biopsied patient cohort died during the observation period.
Because only 17 (22%) of the non-transplanted cohort received interferon, this therapy covariate was not included in the above models. Eight interferon-treated patients had pre-existing diabetes, while nine patients were not known to have this metabolic disorder. Overall, just two patients, both non-diabetic, experienced a sustained virological response after being treated with interferon. Neither these two responders, nor the other seven non-diabetic recipients of interferon, developed diabetes after completing the therapy.
Current allocation policies dictate that waitlisted patients be medically cleared for transplantation on a continual basis until they are ultimately transplanted. This study confirms that, among the HCV+ kidney candidates, remaining on the waiting list is associated with a greater risk of mortality than being transplanted and this risk escalates with time. Additionally, we have demonstrated that the prevalence of diabetes mellitus is increased in HCV+ kidney transplant candidates and, together with dialysis duration, is strongly associated with death on the waiting list in this population. The leading cause of death in these patients is cardiovascular disease, while advanced liver disease is infrequently implicated in the mortality or delisting of waitlisted individuals.
A beneficial effect of transplantation for HCV+ patients with ESRD has been established previously (3,4). Neither study paid much attention to patients who remain on the waiting list. While the first of these did look at causes of death in their small cohort that was not transplanted, they did not examine risk factors for mortality, including comorbid conditions and histological evidence of pre-existing liver disease. In the second of these studies, HCV infection was retrospectively identified in patients by historic sample testing. Consequently, the contribution to patient mortality of pre-existing biochemical or histological liver disease, as well as any associated comorbid condition, was not clearly ascertained. Our observations extend these prior studies by corroborating the beneficial effect of transplantation on patient outcome in a large single center series of HCV+ transplant candidates. All patients in our series were documented to have HCV infection at the time of being listed for kidney transplantation. It is possible that the survival advantage could be a function of medical surveillance bias in that follow-up intensity may have been greater among transplanted patients. On the other hand, almost all non-transplant patients were on dialysis, a therapeutic modality that is likewise generally associated with frequent evaluation by a health care provider, in a dialysis unit. Our study has focused attention on patients remaining on the waitlist and has demonstrated an independent relationship between diabetes and pre-transplant mortality.
Among the HCV+ kidney transplant candidates, the effect of transplantation on liver disease remains uncharacterized, largely because of a paucity of studies that incorporate a thorough pre-transplant evaluation, including liver biopsy, genotyping and viral load. Published data suggest that up to 25% of HCV+ transplant candidates have stage 3 or 4 disease, but information regarding post-transplant histological progression is lacking (20–22). It is generally accepted that more advanced liver disease portends an increased risk of progressive hepatic injury and a higher risk of liver disease-related mortality after transplantation. Guidelines therefore advocate pre-transplant liver biopsy as a means of establishing the degree of baseline liver injury. Under these recommendations, the impact of liver disease on pre-transplant mortality has been hitherto unknown. Our study indicates that, by both clinical and histological criteria, waitlist mortality or delisting due to advanced liver disease is uncommon. We did not examine the impact of worsening liver disease occurring after the initial evaluation; however, the low frequency of mortality due to liver disease that occurred during the study period makes this phenomenon unlikely. The observation that cardiovascular disease was the commonest cause of death among our waitlisted HCV+ patients is consistent with a previous report in a dialysis population (23).
The association of diabetes with HCV infection has been increasingly recognized in the general population, as well as in patients after both liver and kidney transplantation (13–18). The present study demonstrates that pre-existing diabetes mellitus is significantly more common in HCV+ waitlisted kidney candidates than in uninfected counterparts. The 42.9% prevalence of diabetes among the HCV+ patients in our study is higher than the 18% observed by Stehman-Breen et al. in an HCV-infected dialysis cohort (23). We posit that the higher prevalence of diabetes in our patients is attributable to a more chronic exposure to HCV infection, as reflected by longer dialysis duration, as well as the fact that 26% of our patients had undergone prior transplantation. This latter exposure, in and of itself, is a well-recognized risk factor for diabetes in this population.
Although our program practice policy precluded distinguishing whether our diabetic HCV+ patients fit a type I or type II profile (unless they were potential pancreas candidates as well), several prior studies have consistently demonstrated that HCV is associated with insulin resistance and a type II diabetes pattern (24–26). In our study, we attribute the high frequency of diabetes in the cohort under 30 years of age to pre-existing type I diabetes mellitus acquired in childhood in patients who subsequently became infected with hepatitis C. The age-related increase in the prevalence of diabetes mellitus from the fourth decade onwards is in accord with the pattern seen in HCV+ patients in the general population. It remains to be determined whether sustained elimination of viremia with interferon diminishes the risk of subsequent new onset diabetes, either before or after kidney transplantation in potential kidney recipients.
The current national rate of patient death on the kidney waiting list is approximately 6% (1). Contributing factors include lengthening waiting times, as well as patient age and comorbidity. As a consequence, management of patients on the kidney transplant waiting list is an increasingly important area of concern. The consensus from the recent waitlist conference has proposed changes aimed at optimizing the transplantability of patients on the waiting list (2). Among the recommendations, it has been suggested that waitlisted patients be medically cleared for transplantation on a continual basis until they are transplanted. Under this proposal, our study has identified a subgroup of candidates at particularly high risk for pre-transplant mortality, namely HCV+ patients with pre-existing diabetes mellitus. Optimizing outcomes among listed candidates in this population may necessitate more stringent listing criteria in the first place, and thereafter, more frequent re-evaluation of those patients that are activated to accrue time.
We recognize several important limitations to our study. First, our data was retrospectively collected and we did not have complete information with regards to HCV viral load and genotype. Moreover, while elevated hepatic biochemistries are consistent with liver injury, they may underestimate liver disease severity in ESRD patients, a major reason underlying our program policy change in 1998 to incorporate pre-transplant liver biopsies into the standard evaluation of HCV+ kidney candidates. Consequently, the degree of histological damage was not available in all patients and could not be evaluated, either as an independent risk factor for pre-transplant mortality in the entire population, or as a way to identify those few patients that harbor HCV in the liver despite being HCV PCR negative in the serum (27). However, when we restricted our analysis to only patients who did undergo pre-transplantation liver biopsy, our findings were not different to the whole population and moreover, revealed that fewer than 4% were delisted because of histologically advanced liver disease. From a diabetes standpoint, it is also possible that our analysis did not capture the small percentage of patients who developed this metabolic complication after being listed, as has been recently observed (28). However, unrecognized diabetes would be expected to negatively impact on mortality in the cohort not known to have diabetes, and, if anything, mitigate the difference in outcome for patients with known diabetes. Lastly, we cannot exclude the contribution of residual confounding to our findings from unmeasured factors.
In conclusion, this large, single-center study of HCV+ kidney candidates, confirms that kidney transplantation results in superior outcomes compared to dialysis. Additionally, we have demonstrated that the prevalence of diabetes mellitus is higher among HCV+ kidney transplant candidates than their seronegative counterparts, with the frequency increasing in an age dependent manner. Diabetes was strongly associated with mortality among the HCV+ waitlisted transplant candidates while liver disease was not. The importance of this observation is underscored by the fact that among patients yet to be transplanted, death was the most frequent reason for removal from the waiting list. In an era in which the importance of ongoing medical clearance of waitlisted patients is being emphasized, this study identifies a common population subgroup that require frequent and careful re-evaluation to ensure that patients remain viable transplant candidates.