Long-term Analysis of Combined Liver and Kidney Transplantation at a Single Center FREE

End-stage renal disease once was considered a contraindication for liver transplantation (LT). Results of many studies show that varying degrees of renal failure in patients with chronic liver disease are associated with high mortality and morbidity after LT.^1- 4 However, since the first simultaneous combined liver and kidney transplantation (CLKT) was performed in 1983,⁵ several centers worldwide have reported successful long-term outcomes.^6- 12 The CLKT is now the procedure of choice for patients with concurrent end-stage liver and kidney disease.

When patients are already receiving long-term hemodialysis (HD), the decision to recommend a combined procedure is not complicated. In other patients with potentially reversible kidney dysfunction, it may be difficult to predict whether there will be sufficient return of native kidney function to recommend LT alone. Functional renal failure, or the hepatorenal syndrome (HRS), has not been accepted universally as an indication for a simultaneous kidney transplantation and has been reported as a reversible condition treated by means of LT alone.^13- 16 However, as the waiting time for liver organs increases, the incidence of HRS will increase, with the inevitable onset and prolonged duration of HD. In these circumstances, HRS can become a chronic, irreversible condition for which CLKT is the appropriate therapy.^17- 18

In February 2002, the model for end-stage liver disease (MELD) system replaced the United Network for Organ Sharing status classification for the allocation of liver organs.¹⁹ Because of the heavily weighted serum creatinine value in the calculation of the MELD score, candidates with renal failure have received organs more rapidly. As a result, there has been a considerable increase in nationwide volume of CLKT during the past 3 years, with more than 2000 CLKTs performed overall (Figure 1). This study was undertaken to review our experience with the combined procedures at the Dumont-UCLA Transplant Center.

We performed a retrospective review of data in all patients who underwent CLKT at the Dumont-UCLA Transplant Center from October 1, 1988, through August 31, 2004. Each patient initially was evaluated and accepted for LT by the liver transplant patient selection committee. Subsequently, the kidney transplant selection committee approved listing for a combined procedure. All transplantations were performed simultaneously except for one, which was completed within 48 hours.

Patient data were obtained by reviewing inpatient and outpatient medical records and our transplantation database. Data collected and analyzed included age, sex, cause of liver and kidney disease, HD requirements, United Network for Organ Sharing status or MELD score, preoperative and postoperative laboratory evaluation results (including levels of serum sodium, creatinine, prothrombin time/international normalized ratio, alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase, total bilirubin, and albumin), intraoperative data (including surgery time, liver graft cold and warm ischemia time, use of venovenous bypass, and transfusion of packed red blood cells), donor characteristics (including age, sex, race, cause of death, and peak serum sodium and creatinine values), posttransplantation immunosuppressive regimens, episodes of liver and kidney acute cellular rejection, need for retransplantation, and patient and graft survival.

In addition, data in all patients who underwent LT alone from 1998 through 2002 were reviewed for the presence of HRS before transplantation. This group was used for comparison with patients with HRS who underwent CLKT. Finally, data in a cohort of 743 adult patients who underwent cadaveric renal transplantation (CRT) were examined for incidence of acute cellular rejection at 1 year and for overall graft loss.

All LTs were performed with the liver in the orthotopic position by means of either caval interposition (n = 88) or piggyback technique (n = 11). The liver donor allografts were whole organs in 92 cases and segmental grafts in 7. The kidney allograft was implanted separately through a Gibson incision in 94 cases, through the existing LT chevron incision in 3 cases, and through a separate lower midline incision in 2 cases. All allografts were from ABO-compatible cadaveric donors with both grafts from the same donor. Cross-matches were not analyzed retrospectively.

Patients received methylprednisolone sodium succinate preoperatively and then tapered across 1 week, followed by oral prednisone when tolerated. Maintenance immunosuppression regimens varied during this period. A triple-drug regimen of cyclosporine, azathioprine, and prednisone was used from 1988 through 1994. Muromonab-CD3 (commonly referred to as OKT-3) induction or rescue therapies were used in 9 patients during this initial period. Routine use of tacrolimus was implemented in 1994 as part of a dual- or triple-drug regimen with prednisone and mycophenolate mofetil, the latter starting in 1996. The routine use of monoclonal or polyclonal antibodies as induction agents (daclizumab, basiliximab, or thymoglobulin in 1 pediatric case) was added to our regimen in 2002.

No protocol biopsies were performed. All documented episodes of rejection were biopsy based. Liver biopsies were performed for elevation of serum transaminases or when indicated during a subsequent operative exploration. Kidney biopsies were performed when clinically indicated or for elevations in serum creatinine levels.

Patient survival was defined as time from transplantation to death or last follow-up. Liver graft survival was defined as time from transplantation to death, last follow-up, or retransplantation. Kidney graft survival was defined as time from transplantation to death, last follow-up, or return to HD. Hemodialysis days were defined as total inclusive days during which HD (conventional or continuous) was required, either from the first day of HD to transplantation or from transplantation to the last day requiring HD.

Hepatorenal syndrome was defined according to the major criteria outlined by the International Ascites Club (Table 1).²⁰ All patients with HRS in the CLKT and LT-only groups met the criteria. With few exceptions, CLKT was offered to patients with HRS who were receiving HD for a minimum of 4 weeks before transplantation, as previously recommended by Wilkinson and Cohen.¹⁷ The MELD scores were calculated for 83 patients; scores for 15 early patients were incalculable because no international normalized ratio values were reported. Survival curves were computed by using Kaplan-Meier methods, and P values for comparing survival were computed by using the log-rank test. The P values for comparing proportions were computed by using the Fisher exact test. Medians are reported for continuous outcomes, and the P values for comparing medians were computed by using the Wilcoxon rank sum test.

During the 16 years of this study, 98 patients (64 males and 34 females) underwent 99 CLKTs. This group included 89 adult (>18 years old) and 9 pediatric recipients, with a mean age of 46 years (range, 19 months to 67 years). One of the initial patients required retransplantation of both organs for liver graft failure 8 days after the initial CLKT. Twenty-six patients (26%) had previous transplants (16 liver, 8 kidney, and 2 heart). Causes of liver and kidney failure are shown in Table 2.

At the completion of the analysis, 31 patients had died. Causes of death included sepsis or multiple organ failure (16 patients), gastrointestinal hemorrhage and cardiac complications (4 patients each), and pulmonary embolus and ruptured cerebral aneurysm (1 each). Cause of death was unknown in 5 patients. With a median follow-up of 36 months (range, 0-161 months), the overall 1-, 3-, and 5-year patient survival rates were 76%, 72%, and 70%, respectively. We analyzed risk factors for donors and recipients and found that none was associated significantly with early posttransplantation mortality (Table 3).

Thirty-six liver grafts failed for 1-, 3-, and 5-year graft survival rates of 70%, 65%, and 65%, respectively. Death was the cause of graft failures in 24 patients (67%). Hepatic artery thrombosis and primary graft nonfunction accounted for 6 and 3 failures, respectively. The remaining 3 liver grafts were lost to portal vein thrombosis, hepatic outflow obstruction, and recurrent hepatitis C. Retransplantation eventually was performed in 11 patients (11%).

The kidney graft survival rates at 1, 3, and 5 years were 76%, 72%, and 70%, respectively. Twenty-nine (78%) of 37 grafts were lost because of death. Eight patients returned to HD secondary to sepsis and chronic rejection in 3 cases each, recurrent membranoproliferative glomerulonephritis associated with hepatitis B virus in 1, and unknown reasons in 1. One patient underwent retransplantation.

A CLKT was performed in 22 adult patients with HRS. All required HD for a median of 41 days (range, 4-234 days) before transplantation. After transplantation, 10 patients required no further HD, whereas the remaining 12 required supportive HD for a median of 16 days (range, 1-89 days). With a median follow-up of 27.9 months for this subgroup, the 1- and 3-year patient survival rates were both 72%.

To compare this subgroup of patients with a similar cohort of patients not receiving a kidney transplant, we separately reviewed data in all adult patients undergoing LT only at our institution from January 1, 1998, through December 31, 2002. After excluding patients with acute liver failure, a pretransplantation serum creatinine level less than 2.0 mg/dL (176.8 μmol/L), or patients requiring immediate retransplantation, we identified 148 patients with HRS, representing roughly one fifth of all recipients. Eighty (54%) of these patients required pretransplantation HD for less than 30 days and could be compared with the HRS subgroup that underwent CLKT. No patient in the LT-only group required more than 30 days of HD. With a comparable follow-up, the 1-year survival rates for patients with HRS undergoing CLKT or LT only were 72% and 66%, respectively (P = .88) (Table 4). A higher percentage (89%) of patients in the LT-only group required, on average, 1 additional week of HD after transplantation. However, most of these patients stopped receiving HD within 1 month after transplantation (Figure 2). With a median follow-up of 260 days, the median creatinine was 1.55 mg/dL (137.0 μmol/L; mean ± SD, 1.87 ± 1.22 mg/dL [165.3 ± 107.8 μmol/L]). These results document the rapid return of acceptable renal function in patients with HRS after LT alone.

Unlike patients in the LT-only group, patients undergoing CLKT who had HRS had required pretransplantation HD across a wide range of time. Thus, we compared results of the LT-only group with those of the CLKT group stratified according to duration of HD before transplantation and found similar outcomes for patients who underwent CLKT who required less than 8 weeks of HD and those who underwent LT only (Table 5). Conversely, the patients who underwent CLKT and who received HD longer than 8 weeks demonstrated a benefit in patient survival and improved resource use, as measured in fewer postoperative HD days and decreased duration of posttransplantation hospital stay.

As determined by means of nonprotocol biopsies, 25 patients had 34 episodes of liver allograft rejection. At 1 year, the cumulative liver acute cellular rejection rate was 23%.

Eleven patients had 14 episodes of kidney allograft rejection. The 1-year acute cellular rejection rate was 10%. Acute cellular rejection rates for liver and kidney grafts during use of different immunosuppressive regimens during the study are shown in Table 6.

For adult patients who underwent CLKT, the 1-year kidney acute cellular rejection rate was 14%. In a cohort of adults who underwent CRT across 5 years, the 1-year rejection rate was 23% (P<.01). Only 2 kidney allografts were lost to chronic rejection in the CLKT group at 7 and 9 years after transplantation, compared with 10 allografts lost within 5 years in the CRT group.

In this analysis of 99 CLKTs at a single institution, we demonstrated excellent long-term patient and graft survival, comparable with that in patients who underwent LT only at our institution.²¹ Short-term survival is reduced because of a high 6-month posttransplantation mortality rate, with most of these deaths caused by sepsis or multiple organ failure. A similar outcome was reported by Fong and coworkers²² in their United Network for Organ Sharing database analysis of 800 patients undergoing CLKT from October 1987 through October 2001. In our series, only previous organ transplantation and longer liver graft cold ischemia time approached statistical significance as risk factors for early mortality. These findings underscore the inherent severity of illness in patients with dual organ failure and their associated risks immediately after transplantation. The outlook is excellent for patients who survive the first 6 months.

Transplant physicians have long debated whether to add a concomitant kidney graft in patients with liver failure and HRS. With more potential recipients receiving HD awaiting transplantation, we are faced with the decision to place both organs into a single patient or to offer the kidney graft to a patient who has been on the deceased donor waiting list for many years. We hoped to identify guidelines to maximize the success of a simultaneous kidney transplantation and arrived at 2 conclusions.

First, there is eventual return of native kidney function for patients receiving short-term (≤30 days) pretransplantation HD who undergo LT only. Second, for patients receiving HD longer than 8 weeks, CLKT confers advantages regarding survival and resource use. We are unable to speculate as to the effects of prolonged HD (>30 days) on native kidney function in patients without a kidney transplant because we had no control group. We therefore recommend CLKT for patients who have HRS and receive HD longer than 8 weeks. This policy should eliminate the use of scarce kidneys for a population of patients who tend to regain native renal function after transplantation and allow sufficient time for native kidney function to become frankly irreversible. Early posttransplantation mortality and kidney use are improved when these specific guidelines are followed. The guidelines apply to transplant centers where waiting times are prolonged and patients with dual organ dysfunction are common.

Quadruple drug immunosuppression with monoclonal antibody induction was instituted for patients undergoing CLKT in 2002, coincident with the beginning of the MELD era. This kidney-sparing protocol allows the new kidney allograft to be protected from the normally higher maintenance levels of calcineurin inhibitors used in patients undergoing LT only. Induction therapy with basiliximab and daclizumab reduces the incidence of acute cellular rejection in kidney transplantation.^23- 24 Results of 2 recent studies showed a similar reduction in acute cellular rejection episodes in patients undergoing LT only—without adverse effects on patient or graft survival. In these reports, the 6-month acute cellular rejection rate was 35% in the basiliximab group²⁵ and 25% in the daclizumab group.²⁶ Of 88 patients who received induction therapy, we report only 2 cases (7%) of liver allograft rejection, which raises the question of whether multiorgan transplantation enhances the T-lymphocyte–depleting effect of monoclonal antibody induction. This low incidence of rejection likely accounts for the increased success of the combined procedure since monoclonal antibody induction was implemented, with 1- and 2-year patient and liver graft survival rates of 85% and 79%, respectively.

To our knowledge, Calne et al²⁷ first described immunoprotection of the kidney allograft by a liver allograft in an animal model. This hypothesis also was proposed in humans a decade ago by Rasmussen et al²⁸ and challenged by Katznelson and Cecka.²⁹ We report a significant reduction in both kidney acute cellular rejection episodes at 1 year and graft failure due to chronic rejection in the adult patients who under go CLKT vs the CRT cohort in our study. Similar findings were reported by Fong et al.²² These data support the theory that the liver provides immunologic protection, both short and long term, to the kidney.

The shortcomings of this analysis include its retrospective nature and limited data for the early patients undergoing CLKT. Statistical significance could not be demonstrated in any of the HRS group comparisons given the low power in each subgroup. Because we used no protocol liver or kidney biopsies, it is possible that rejection of one organ went undocumented, where a positive biopsy result for rejection of the other organ resulted in the initiation of treatment. Finally, we recognized that the immunosuppressive regimens for the CRT group are different than those for patients undergoing CLKT at our institution. However, it is safe to assume that patients undergoing CRT are better matched with their grafts given the ardent efforts at pretransplantation tissue typing in kidney transplantation. If one takes this aspect into account, the lower rejection rates in the CLKT group appear more impressive.

Most patients with HRS who progress to renal failure with or without short-term HD have been treated with LT alone. Before this analysis, we recommended CLKT when patients receive HD for longer than 1 month before transplantation. However, on the basis of current findings, we find that the acuteness of renal failure subsides after 2 months of HD and that CLKT after this time will not only offer improvement in patient survival but also reduce hospital expenditures for patient care.

Combined liver and kidney transplantation offers the best option for patients with simultaneous chronic liver and kidney failure when it is performed at a high-volume academic transplant center. Analysis of the more recent patients who underwent CLKT in our study demonstrates improved patient and graft survival most likely because of better perioperative medical management, shorter graft ischemia time, and more effective immunosuppressive regimens.

Correspondence: Ronald W. Busuttil, MD, PhD, Dumont-UCLA Transplant Center, Division of Liver and Pancreas Transplantation, David Geffen School of Medicine at UCLA, 10833 LeConte Ave, Room 77-120 CHS, Los Angeles, CA 90095.

Accepted for Publication: March 23, 2006.

Funding/Support: None reported.

Previous Presentation: This paper was presented at the 77th Annual Meeting of the Pacific Coast Surgical Association; February 18, 2006; San Francisco, Calif; and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.

John P. Roberts, MD, San Francisco, Calif: The authors are to be congratulated on their excellent patient and graft survival rates in an elderly population with combined liver and kidney failure. Their success speaks to the expertise at University of California, Los Angeles. This population presents a transplant team with therapeutic, diagnostic, technical, and ethical challenges.

The shortage of organs today forces us to decide between different patients with different potential outcomes. Competing for the kidneys used in the patients with combined liver and kidney failure are thousands of patients waiting for kidney transplantation in the United States—the patients on the waiting list in Los Angeles and San Francisco wait longer than 5 years to get a kidney transplant. These patients receiving HD face increased risk of death before and after kidney transplantation related to their time receiving HD. These patients undergoing kidney transplantation have a better patient and graft survival by about 10% to 15% than the results presented here. On the other hand, kidney failure in the long term after LT is associated with a 4-fold increase in the risk of death in patients after LT so that the LT in the patient who then goes on later to kidney failure may be a poor use of these livers.

This tension between competing groups of patients has created great controversy over the best use of kidneys in these populations. One preventable waste of organs would be to have a patient who had liver failure and acute kidney failure who received a kidney transplant but then went on to recover native kidney function resulting in the patient with 3 functioning kidneys.

The authors focused their article on this controversial area in trying to decide which patients have recoverable renal function. The authors found that patients with acute renal failure who did not undergo kidney transplantation but had a diagnosis of HRS actually had relatively good creatinine levels of 1.8 mg/dL (159.1 μmol/L) at 1 year. Their data suggest that a fair number of these patients have significant renal dysfunction a year after transplantation, but this is because 1 SD would include patients with a creatinine level greater than 3.0 mg/dL (>265.2 μmol/L). The authors also found better outcome in patients who received HD for 8 weeks compared with patients who received HD for less than 8 weeks.

The authors examined the incidence of rejection in this population. The liver long has been thought to provide immunologic cover for the kidney by preventing attack by the immune system. The authors find that the incidence of kidney rejection was less common in the population undergoing CLKT than in the population undergoing kidney-only transplantation. One confounding issue is the fact that the liver may demonstrate rejection clinically prior to an increase in creatinine levels suggests kidney rejection. This fact, combined with the lower risk of a liver biopsy, compared with a kidney biopsy, may lead to the liver being used as the warning for kidney rejection, which may result in a decrease in the diagnosis of kidney rejection. The relatively high percentage of patients with liver transplants who had rejection diagnosed and treated may have prevented or alleviated the rejection of the kidney.

My questions for the authors are as follows: There appeared to be some selection bias in which patients with HRS underwent kidney transplantation despite their 4-week HD rule. Could they identify the factors that led to this bias? Furthermore, given the number of patients who had marked renal dysfunction at 1 year and would be predicted to have an increased risk of death in the long term, can they develop factors that would predict for this outcome?

The authors also suggested that the patients who received more than 8 weeks of HD had a better outcome. Was this just a natural selection process eliminating patients who were too weak to survive until 8 weeks? If they are going to go to 8 weeks as their cutoff for transplantation, are they considering studying this prospectively or even randomly assigning the patients to kidney transplantation or not?

Finally, can they tell us the number of patients who had renal dysfunction at the time of the diagnosis of liver rejection that would suggest that they might be having synchronous kidney rejection?

Nicolas Jabbour, MD, Los Angeles, Calif: Did you study the function of the native kidneys after transplantation for HRS to see if these kidneys can recover even after 8 weeks of renal failure?

William H. Marks, MD, Seattle, Wash: This article raises some interesting points, and Dr Roberts touched on a couple of them in his discussion. I guess the additional question I have is, if you brush aside the economic advantage provided by giving the patients with HRS a kidney transplant, does the advantage of giving them a kidney offset the disadvantage of patients with kidney failure on the list who are awaiting kidney-only transplantation? In other words, is there truly an advantage that offsets the use of this precious resource?

Linda L. Wong, MD, Honolulu, Hawaii: I was just wondering if you had any idea how many patients during this same period were listed for both liver and kidney transplants but never received a transplant because they were too sick, were removed from the list, or died for some other reason.

Dr Busuttil: Dr Roberts has brought up several important issues regarding CLKT. The crucial issue that remains unanswered is how to identify patients with HRS who should receive a kidney at the time of LT. Many clinical parameters have been proposed, although none has been rigorously tested. These include the level of creatinine, the degree of oliguria or anuria, the number of pressers, and the results of MAG-3 [Tc-99m-mercaptoacetyltriglycine] scanning. Unfortunately, none of these has provided a very high predictive value, which is the crux of the issue. Our data suggest that receiving HD for 8 or more weeks is a start.

The other issue is that we need to do whatever we can to prevent patients from going into kidney failure after liver replacement, as has been shown to occur in about 20% of patients 5 years after transplantation, with a 4-fold increase in mortality. These figures provide the obvious impetus to select patients appropriately for CLKT.

Regarding the questions, Dr Roberts asked how we know that we are not wasting a kidney. We have a number of patients with 3 functioning kidneys, and that fact underscores the need for specific preoperative, prognostic indicators for CLKT. That question will be answered ultimately by means of a prospective trial. It also might be useful to perform MAG-3 scanning in patients with HRS at intervals before and after transplantation to assess native kidney function.

Regarding the question of selection bias in patients with HRS, the 4-week rule became our policy during the experience we have presented, and we now adhere to it pretty strictly. I have spoken about the issue of predictive factors, and I agree that a prospective study would be of value.

Regarding the frequency and degree of renal dysfunction in patients with biopsy-proved liver rejection, this is a valuable area for analysis but not one we have looked at specifically.

Dr Jabbour asked if we measured function of the native kidneys after transplantation for HRS. We did not, but MAG-3 scanning might be an appropriate means to do so.

Dr Marks asked if the survival of patients with HRS justified taking away kidneys from patients on the cadaveric kidney list. I do not know. If we can avoid having patients with 3 kidneys, that is what we need to do, and give the kidney to a patient on the cadaveric list rather than to the patient with HRS who has recoverable kidney function. We have an editorial that we hope will be coming out in the next couple of months, examining the issue of whether we are giving too many patients on the liver transplant list a kidney on the basis of their MELD score. The data suggest that we are losing more kidneys with this policy.

Dr Wong asked how many patients listed for liver-kidney transplantation were removed from the list or died before transplantation. In our recent experience, about 10 to 20 patients died each year.