Decreased Survival in Liver Transplant Patients Requiring Chronic Dialysis: A Canadian Experience

Abstract

Background. 

Chronic kidney disease is associated with increased mortality among nonrenal organ transplant recipients. End-stage renal disease (ESRD) is a serious complication after orthotopic liver transplantation (OLT). It is unclear if the outcomes of these individuals are different from nontransplant patients requiring dialysis or a kidney transplant.

Methods. 

We report the incidence of ESRD in OLT recipients and compare their outcomes to matched dialysis controls. We analyzed 4186 patients who received an OLT in Canada between January 1981 and December 2002 and 228 matched, nontransplant, chronic dialysis controls.

Results. 

The incidence of ESRD after OLT was 2.9% (n=120). The unadjusted mortality rate for those who required chronic dialysis was 49.2% compared with 26.8% in those who did not develop kidney failure (P<0.0001). The survival of OLT recipients on dialysis was lower than the matched chronic dialysis cohort (log-rank test, P=0.01). A kidney transplant was performed in 24% of the OLT recipients and 21% of the matched dialysis cohort, and their overall survival was similar. The OLT patients who remained on dialysis had a significantly lower survival when compared with matched dialysis patients who did not receive a kidney transplant (log-rank test, P=0.0002).

Conclusions. 

Mortality was greater for OLT recipients on dialysis than would be expected from a matched, nontransplant, dialysis cohort. Kidney transplantation may abrogate some of this increased mortality risk.

With the overall improvement in orthotopic liver transplantation (OLT), chronic kidney disease (CKD) has evolved as a major complication that negatively influences long-term patient survival. Chronic kidney disease after OLT is multifactorial, but calcineurin inhibitor (CNI) associated nephrotoxicity is well recognized (^1–4). Other risk factors associated with CKD in OLT are advanced age, lower pretransplant kidney function, postoperative acute kidney injury, pre- or postoperative diabetes mellitus, hepatitis C, and hepatorenal syndrome at the time of transplantation (^{1, 5–9}).

End-stage renal disease (ESRD) has been reported in 2% to 10% of OLT recipients by 10 years after transplantation (^{7, 10}). The risk of ESRD in this population was best described by Ojo et al. (¹¹) using the data from the Scientific Registry of Transplant Recipients. They reported a cumulative incidence of kidney failure in OLT of 8% at 1-year and 18% at 5-years, which was higher than heart, lung, or heart-lung transplant recipients. Despite the high rate of kidney failure associated with OLT, little is known regarding the outcomes of these patients after they begin renal replacement therapy.

We conducted a population-based, cohort study of all OLT recipients in Canada performed over a 22-year period. Our first aim was to report the incidence of ESRD requiring chronic dialysis in OLT recipients and examine their outcome with those who did not reach ESRD. Our second aim was to compare the survival of OLT recipients on dialysis and after kidney transplantation with that of a matched, nontransplant, chronic dialysis cohort.

METHODS

Sources of Data

Data were extracted from the Canadian Organ Replacement Register (CORR), Canada’s national database on treatment for end-stage organ failure. This database, managed by the Canadian Institute for Health Information, captures patient demographic, clinical and outcome variables on patients who have received solid organ transplants and chronic renal dialysis in the country.

Study Subjects

This study included patients with a first registered treatment in CORR of a liver transplant performed at all Canadian transplant programs (seven adult and two pediatric) between January 1, 1981 and December 31, 2002. Patients were excluded from the study if their first registered treatment in CORR was a combined transplant involving a kidney, they had chronic renal dialysis treatment before the liver transplant, they had a preemptive kidney transplant after the liver transplant, or if they went on to have dialysis but became lost to follow-up or recovered renal function. The definition of chronic dialysis was clinically determined by each center. The outcome of mortality was also obtained from CORR. Dialysis data is well captured because a standardized form is submitted to CORR for each patient when they initiate dialysis and again when there is a death. The number of transplant patients in our registry was compared with the aggregate numbers received from each transplant facility on an annual basis to verify complete capture of subjects.

For our second aim, OLT recipients requiring chronic dialysis were matched to a cohort of patients, also from CORR, with ESRD on chronic dialysis, but who had not received a solid organ transplant (n=228). Patients were well-matched on gender, age ±10 years, initial dialysis modality, initial dialysis treatment facility, and year of initial dialysis treatment. An attempt was made to match two dialysis patients to every OLT recipient. For one OLT recipient, no match could be made. For 10 recipients, only one match could be made. Information on diabetic status was available for 110 of the 120 OLT patients and 213 of 228 of the matched dialysis cohort.

Statistical Analysis

We used means and standard deviations to summarize continuous data. Categorical data were compared using chi-square test, whereas Tukey pooled t test was used for continuous data. Actuarial survival was reported using life tables to compute estimates of patient survival. Patient survival time (patient-years) was calculated from the time of OLT to death or the end of the observation period (December 31, 2002). Patients were censored at the time of kidney transplant and if they remained alive at the end of the observation period. We used SAS Enterprise Guide Version 2.05 (SAS Institute, Cary, NC) for all the statistical analyses.

RESULTS

End-Stage Renal Disease After Liver Transplant

Of the 4186 OLT performed during the study period, 120 patients (2.9%) experienced ESRD requiring chronic dialysis and 4066 patients did not develop ESRD. The median time elapsed between OLT and the initiation of chronic dialysis was 5 years, with a mean of 5.1±3.7 years. Of patients who went on to require dialysis, 97.5% (n=117) were initiated on hemodialysis, and the remaining 2.5% (n=3) started with peritoneal dialysis.

Survival According to Chronic Dialysis Status

The baseline characteristics of the OLT recipients according to dialysis status are summarized in Table 1. The OLT recipients who received chronic dialysis included significantly more men and were older than the group that did not require chronic dialysis. Age was also explored as a categorical variable. Among the 570 pediatric (age <18 years) individuals in this cohort, 0.7% (n=4) required chronic dialysis. Chronic dialysis was started in those 18 to 44 years old in 3.1% (n=34/1,114), 45 to 59 years in 3.0% (n=51/1,726), and for more than or equal to 60 years old in 4.0% (n=31/776) of individuals. Of the identifiable primary causes of liver failure, chronic hepatitis C (n=754) accounted for the highest proportion, 3.8% (n=29), of those who required chronic dialysis.

TABLE 1: Patient characteristics by dialysis status

Table 2 shows outcomes according to dialysis status. In total, 49.2% (59 of 120) of the OLT recipients requiring dialysis died during the observation period, compared with 26.8% (1,091 of 4,066) of the nondialysis group. The median time between initiation of chronic dialysis and death was 1.5 years with a mean 1.6±1.3.

TABLE 2: Patient outcomes by dialysis status

Survival on Dialysis After Kidney Transplantation

Table 3 shows the results of the matching process of chronic dialysis patients to the OLT dialysis cohort. No significant mismatches were present on any of the key demographic variables. The 5-year patient survival was 17% in the OLT dialysis group and 43% in the matched chronic dialysis controls (Fig. 1). There was a significantly decreased survival for the OLT recipients compared with the matched dialysis cohort (log-rank χ²=12.83, df=1, P=0.01, Fig. 1).

TABLE 3: Comparison of OLT recipients on dialysis and a matched dialysis cohort

FIGURE 1.: Life table analysis of OLT recipients after initiation of dialysis compared with a matched chronic dialysis cohort. The 5-year patient survival was 17% for those on dialysis after OLT and 43% in the matched dialysis cohort.

A similar proportion of kidney transplants were performed in each cohort (24.2% vs. 20.6%). The survival of OLT recipients who received a kidney transplant was comparable with the matched dialysis cohort who received a kidney transplant (80% vs. 90%, log-rank χ²=1.269, df=1, P=0.26). However, the OLT recipients who remained on dialysis had a decreased survival rate compared with similar dialysis patients (log-rank 11.5% vs. 36.5% χ²=13.82, df=1, P=0.0002) (Fig. 2).

FIGURE 2.: Life table analysis of OLT recipients and a matched chronic dialysis cohort, with or without kidney transplantation. The 5-year patient survival after kidney transplantation was 80% in the OLT recipients and 90% in the chronic dialysis cohort. The 5-year patient survival in those who remained on chronic dialysis was 11.5% in the OLT dialysis group and 36.5% in the matched chronic dialysis controls.

DISCUSSION

We show that individuals post-OLT who initiate chronic dialysis have an increased risk of death compared with matched, nontransplant, chronic dialysis controls. Although it has been shown that CKD after nonrenal organ transplantation is associated with increased mortality, we report the poor outcome in this cohort after initiating dialysis when compared with a matched dialysis cohort (^{11, 12}). The OLT patients who received a kidney transplant exhibited a similar survival to those of matched chronic dialysis patients who underwent kidney transplantation. Although the number of patients analyzed was relatively small and could represent a type II error, this suggests the OLT patients may have a proportionally greater survival benefit with kidney transplantation than other dialysis patients.

The incidence of ESRD after OLT was 2.9% in our study population. This is lower than previously reported in single center studies, ranging from 4% to 10% with follow up of 10 to 13 years (^{3, 4, 7, 13–16}). The OLT group which required chronic dialysis exhibited a longer mean survival from time of OLT than those who did not develop ESRD. This highlights the issue that one must survive long enough to eventually develop ESRD and require chronic dialysis. This phenomenon of competing risks between ESRD and death has been shown in a previous study of OLT (⁵). Our findings are in keeping with other observational studies in U.S. populations. One study found increased mortality in 45 patients with an OLT as compared with OLT recipients without kidney failure (⁷). Another study found equally poor survival in 43 patients who reached dialysis compared with those with preserved kidney function. This latter study did show, however, that 10-year survival was significantly improved in those who received a kidney transplant, versus remaining on dialysis (¹). Because ESRD is well associated with increased cardiovascular mortality, (¹⁷) we thought a more valid approach was to adopt a matched dialysis population as our control group, rather than making comparisons with a nonkidney failure cohort. As a result, our findings suggest that factors associated with the development of ESRD in OLT provide additional risk for mortality than dialysis alone.

We could not determine the risk factors for developing ESRD from the information provided by the registry data, and this was not the purpose of the study. Other limitations of this study were that variables that may have influenced survival, such as immunosuppressive medication use, cardiac status, laboratory or dialysis treatment parameters, and race. For the period under consideration for this study (1981–2002), race was not always a mandatory variable and consequently was not captured consistently, making it unsuitable for analysis in this study. Despite our control population being well-matched on the selected variables, we cannot exclude residual confounding from unmeasured factors. A potential strength of our study is that data was compiled from all organ transplant centers across Canada, rather than from a single center, enhancing the generalizability of our findings.

Implications for this research relate to the recognition that mortality for conventional dialysis patients is further increased in patients with OLT. Factors that identify progression to ESRD in this cohort would potentially allow for interventions to slow the progression of CKD. As CNI-associated toxicity is of particular concern, different strategies aimed at renal protection may include lowering the dose, early withdrawal, or even avoidance of CNI. Further studies should be done to prospectively explore these questions.

In conclusion, we report the results from a national Canadian registry of OLT patients who developed kidney failure requiring dialysis therapy. Although the incidence of dialysis dependent chronic renal failure remains relatively low (2.9%), the survival in this cohort is significantly worse when compared with those who do not develop ESRD or to a matched, chronic dialysis cohort. This increased mortality risk seems to be abrogated in those who receive a kidney transplant.