Continued Influence of Preoperative Renal Function on Outcome of Orthotopic Liver Transplant (OLTX) in the US: Where Will MELD Lead Us?


*Corresponding author: Thomas A. Gonwa,


Renal function is a component of the Model for End Stage Liver Disease (MELD), We queried the 1999–2004 OPTN/UNOS database to determine whether preoperative renal function remained an important determinant of survival in primary deceased donor liver transplant alone patients (DDLTA) or primary combined kidney liver transplant patients (KLTX). We examined preoperative creatinine, renal replacement therapy (RRT), incidence of KLTX, and patient survival in the 34 months before and after introduction of MELD and performed a multivariate Cox regression analysis of time to death.

Preoperative renal function is an independent predictor of survival in DDLTA but not in KLTX. When compared to DDLTA with a preoperative serum creatinine of 0–0.99 mg/dL, patients with serum creatinine from 1–1.99 mg/dL, >2.0 mg/dL, those requiring RRT, and those receiving KLTX had a relative risk of death following transplant of 1.11, 1.58, 1.77, and 1.44 respectively. KLTX requiring RRT had better survival than DDLTA requiring RRT. Since introduction of MELD, KLTX, preoperative creatinine, and number of patients requiring preoperative RRT have increased. Despite this, patient survival following orthotopic liver transplant (OLTX) in the 34 months after introduction of MELD is not different than prior to introduction of MELD.


Preoperative renal function has been previously shown to influence postoperative morbidity and mortality in patients undergoing orthotopic liver transplant (OLTX 1–8). It can be used to predict not only the need for postoperative renal replacement therapy (RRT) but also the risk of postoperative infection (7–10). Previous studies analyzed data from transplants performed prior to 1998. Nair et al. demonstrated a stepwise decrease in survival following OLTX depending on preoperative renal function (2). Patients with an average preoperative serum creatinine of 0.8 mg/dL had a 5-year patient survival of 62% compared to a 5-year survival of only 42% in patients with a preoperative serum creatinine of 2.7 mg/dL. Organ Procurement and Transplantation Network/United Network for Organ Sharing (OPTN/UNOS) data from 1988 to 1995 demonstrated that patients with a preoperative serum creatinine >2 mg/dL had a 5-year survival of only 50% (10). Furthermore, patients requiring preoperative RRT had worse outcomes compared to those not requiring RRT (5,11).

Combined kidney liver transplant (KLTX) has been utilized for patients with failure of both organs. Utilizing OPTN/UNOS data, it has been previously shown that patients receiving KLTX did better than patients with preoperative serum creatinine > 2 mg/dL who received a liver alone (10). This has put pressure on many centers to perform KLTX in patients with elevated preoperative serum creatinine rather than OLTX alone. Coincident with these findings, the Model for End Stage Liver Disease (MELD) system was introduced to quantify the risk of death due to liver disease (12,13). The main components of the MELD score are bilirubin, International normalized ratio (INR), and serum creatinine. This score was adopted by the OPTN/UNOS to allocate organs for those awaiting OLTX based on medical necessity. There were several fears voiced during the debate over MELD, including concerns that patients with worse renal function would be transplanted resulting in lower overall survival or possibly an increase the number of KLTX. Initial analyses from two separate large transplant centers have shown an apparent increase in postoperative mortality that correlates with pretransplant MELD score (14,15). However, three studies have been published examining the impact of the introduction of MELD on post transplant survival that utilized the OPTN/UNOS database (16–18). None of them shows an increase in post OLTX mortality following the introduction of MELD although follow-up is short.

Over the last few years there have been great advances in RRT and the appreciation that early intervention in patients with renal dysfunction awaiting OLTX may improve survival. This may allow transplant centers to better manage patients with poor renal function who present for OLTX. We undertook the present study to address several issues. First, does pretransplant renal function continue to influence post OLTX survival? Second, has the introduction of the MELD system increased the number of patients with poor renal function receiving OLTX or the number of KLTX being performed? Finally, has the introduction of the MELD system influenced overall survival following liver transplantation?

Materials and Methods

The OPTN/UNOS database was queried for all primary deceased donor OLTX performed between 4/27/1999 and 12/26/2004. Patients undergoing deceased donor liver transplant alone (DDLTA) or combined simultaneous KLTX were included in the study. Data were extracted in regards to pretransplant renal function reported in these patients. Preoperative serum creatinine and the need for pre operative RRT was utilized to stratify patients. Patients were divided into four groups, DDLTA with a serum creatinine 0–0.99 mg/dL, DDLTA with a serum creatinine 1–1.99 mg/dL, DDLTA with a serum creatinine >2.0 mg/dL and DDLTA requiring preoperative RRT. During this time period, pretransplant RRT was collected two ways in the OPTN/UNOS database and we utilized both definitions for this study. The first is any form of RRT prior to removal from the waiting list for OLTX (the primary pre-MELD definition), and the second definition is any form of RRT twice in the week preceding OLTX (the primary post-MELD definition). Patients who met either definition were included in the RRT group in order to make comparisons between the pre MELD and the post MELD groups. Patients who lacked data on pretransplant renal function were excluded from the analysis. Routine diagnostic and demographic data were also extracted on the patients. For the patients receiving KLTX, renal diagnoses were examined by extraction from the renal transplant database. For the analysis of the effect of MELD, two time periods were utilized. The post-MELD period covered the 34-month time period from the institution of the MELD system for distribution of organs to the end of the study period (February 27, 2002 to December 31, 2004). The pre-MELD period comprised the 34 months preceding the institution of the MELD system (April 27, 1999 to February 26, 2002). MELD scores are only reported for those patients within the MELD era.

A multivariate Cox proportional hazards analysis was also performed to determine whether renal function had an impact on patient survival when controlling for other important donor and recipient characteristics. This was limited to patients transplanted in the post-MELD era. Variables included in the model were: transplant group, recipient age, gender, ethnicity, diagnosis, MELD score, medical condition, ventilated support status, donor age, ethnicity and cold ischemia time. The results of the Cox regression are presented as relative risks (RR), their 95% confidence limits, and p-values.

Patient and graft survival rates as of February 17, 2006 were calculated utilizing the Kaplan-Meier method, and differences in survival were assessed using the logrank test. Demographics were compared using chi square tests. All statistical analyses were performed using SAS, version 9.1. (SAS Institute, Cary, NC)


In this time period, 25 023 patients were identified for the study. Of these, 24 133 were DDLTA and 890 were KLTX. The pre-MELD period consists of 11.298 patients (11 010 DDLTA and 288 KLTX) and the post-MELD period has 13 765 patients (13 163 DDLTA and 602 KLTX). Analysis of patient survival rates according to preoperative serum creatinine (DDLTA no RRT) versus DDLTA patients requiring RRT preoperatively versus patients receiving KLTX is presented in Figure 1. As can be seen, patient survival rates were worse for every level of increasing renal dysfunction in patients who received a DDLTA. Sixty month survival was 79.1%, 72.2%, 63.1%, 63.9% and 69.6 % for DDLTA serum creatinine 0–0.99 mg/dL, 1.0–1.99 mg/dL, > 2.0 mg/dL, patients requiring RRT, and patients receiving a KLTX respectively. Overall, DDLTA patients did better than KLTX patients (all DDLTA vs. KLTX, p = 0.0022). However, KLTX patients had increased survival compared to DDLTA patients requiring RRT and DDLTA patients with creatinine >2.0 mg/dL (p < 0.0001). In data not shown, there was no effect of pretransplant creatinine or dialysis status on survival following KLTX. In data not shown, the same effect of pretransplant serum creatinine was seen in patients undergoing repeat transplant. At every renal function level, patients undergoing repeat transplant had worse patient survival. Furthermore, at every pretransplant renal function level, repeat transplant patients did worse than primary transplant patients. The difference between those with best renal function and those with worst renal function was similar in the repeat transplant recipients.

Figure 1.

Patient survival in months post transplant following primary deceased donor liver transplant alone (DDLTA) stratified according to pretransplant serum creatinine (Crt) 0–0.99 mg/dL, 1.0–1.99 mg/dL, >2.0 mg/dL, or need for renal replacement therapy (RRT) or patients receiving a combined kidney liver transplant (KLTX), 4/27/1999–12/26-2004.

We next looked at specific subgroups of the main analysis. Table 1 presents the patient survival rates for these various subgroups. When one compares DDLTA patients not requiring RRT to KLTX patients not requiring RRT, there is no difference in survival (p = 0.81) although there appears to be a trend toward better survival in the DDLTA patients. The 3-year patient survival rate in the DDLTA patients not requiring preoperative RRT was 80.6% compared to 73.1% in the KLTX patients not requiring RRT. KLTX patients requiring preoperative RRT had superior patient survival rates compared to DDLTA patients requiring preoperative RRT (74.8% vs. 68.3% at 3 years, p < 0.0003). Finally, patients who received a KLTX not requiring preoperative RRT with a preoperative serum creatinine ≥2 mg/dL did not have better survival compared to DDLTA patients not requiring RRT with a preoperative serum creatinine ≥2.0 mg/dL (69.9% vs. 69.8% at 3 years, p = 0.177). Of note is that 28.9% of KLTX patients did not require RRT prior to transplant. Of these, 80% had a preoperative serum creatinine of >2.0 mg/dL.

Table 1.  Kaplan-Meier patient survival rates following OLTX in the United States, 4/27/1999–12/262004, by transplant type and pretransplant renal function status
GroupNumber of transplants12 (%)24 (%)36 (%)p-Value
  1. DDLTA = deceased donor liver transplant alone.

  2. KLTX = combined kidney liver transplant.

  3. RRT = renal replacement therapy.

  4. Creat = preoperative serum creatinine (mg/dL).

  5. a versus c, p < 0.0001.

  6. b versus d, p = 0.8015.

aDDLTA, no RRT23 07088.483.980.60.081
bKLTX, no RRT25888.081.173.1 
cDDLTA, with RRT110375.1671.268.30.0003
dKLTX, with RRT63283.679.074.8 
eDDLTA, no RRT, Creat ≥ 2219079.774.169.80.177
fKLTX, no RRT, Creat ≥ 220886.179.469.9 

Demographic data on the five patient groups are presented in Tables 2–5. DDLTA patients with a creatinine >2.0 mg/dL or those requiring RRT were more likely to be Status 1. Since a major component of the MELD score is renal function, there was an increase in the MELD score corresponding to renal function with the DDLTA patients requiring RRT having the highest (Table 2). Donor characteristics are presented in Table 3. The KLTX patients on the whole received younger donors with shorter ischemia times. This may explain some of the advantage that these patients have in survival. Liver diagnosis is presented in Table 4. There were a higher percentage of patients with metabolic disease in the DDLTA requiring RRT and in the KLTX patients.

Table 2.  Recipient demographics and MELD score by pretransplant renal function and transplant type 4/27/99–12/26/2004
 DDLTA 0–0.99 mg/dL (%)DDLTA 1–1.99 mg/dL (%)DDLTA >2.0 mg/dL (%)DDLTA RRT (%)KLTX %
  1. 1p < 0.0001 across all groups.

  2. 2MELD score reported only for patients transplanted after 2/27/2002.

  3. DDLTA = Deceased donor liver transplant alone.

  4. KLTX = Combined kidney liver transplant.

Age < 18115.
Sex male158.571.667.463.663.7
Staus 118.33.811.521.33.3
MELD < 151,221.413.
Table 3.  Donor demographic data and cold ischemia time by pretransplant renal function and transplant type 4/27/99–12/26/2004
 DDLTA 0–0.99 mg/dLDDLTA 1–1.99 mg/dLDDLTA >2.0 mg/dLDDLTA RRTKLTX
  1. 1p < 0.0001 across all groups.

  2. 2p = 0.25 across all groups.

  3. 3p < 0.0001 across all groups.

  4. 4p = 0.007 across all groups.

Age < 18120.99.610.213.216.2
Sex male258.560.059.859.958.4
Liver cold ischemia time4
 Not reported14.515.616.818.515.4
 <6 h21.721.620.419.226.0
 6–12 h54.553.153.953.552.0
Table 4.  Recipient diagnosis by pretransplant renal function and transplant type 4/27/99–12/26/2004
 DDLTA 0–0.99 mg/dLDDLTA 1–1.99 mg/dLDDLTA >2.0 mg/dLDDLTA RRTKLTX
  1. DDLTA = Deceased donor liver transplant alone.

  2. KLTX = Combined kidney liver transplant.

  3. RRT = Renal replacement therapy.

  4. p < 0.0001 across all groups.

Acute necrosis7.26.711.618.15.8
Nonchol cirrhosis56.870.469.061.569.1
Chol cirrhosis11.
Other liver dis5.
Biliary atresia6.
Benign neoplasms0.
Table 5.  Primary renal diagnosis of patient receiving combined kidney liver transplant in the United States; 4/27/1999–12/26/2004
DiagnosisN (%)
Acute renal failure or Hepatorenal Syndrome2.0
Previous renal transplant8.3
Tubulointerstitial disease2.8
Glomerular disease14.0
Polycystic kidney disease8.1
Other renal diagnosis not specified38.5
No renal diagnosis or unknown0.7

Allocation of livers for transplant began to utilize MELD score on February 27, 2002. We examined the impact of this change on the distribution of preoperative serum creatinine, preoperative RRT, KLTX, and survival following OLTX in the United States. Figure 2 demonstrates the percentage of patients receiving OLTX who received a KLTX in the United States for the years 1999–2004. The actual number of KLTX transplants for those years is 99, 135, 134, 210, 246 and 279. As can be seen, there has been an increase in KLTX as a percentage and in absolute numbers beginning in 2002, the year that MELD was introduced. The primary renal diagnosis of patients receiving KLTX is shown in Table 5. As can be seen, 2% had a diagnosis of acute renal failure or hepatorenal syndrome and 0.7% had no renal diagnosis entered. Interestingly, 8.3% had a previous renal transplant as indicated by the diagnosis of failed renal transplant. As shown in Table 6, the number of KLTX (absolute number and % of OLTX) increased in the 34 months following the introduction of MELD compared to the 34 months preceding MELD. Furthermore, the number of patients with renal dysfunction was greater after introduction of MELD compared to pre-MELD as noted by the higher preoperative creatinine levels and the greater use of preoperative RRT.

Figure 2.

Combined kidney liver transplants performed in the United States, total number and percentage of total OLTX (living donor OLTX excluded).

Table 6.  Liver transplants by creatinine, renal replacement therapy (RRT), and KLTX before and after institution of the MELD system for allocation of livers
 Pre-MELD 4/27/99–2/26/02 N (%)Post-MELD 2/27/02–12/26/04 N (%)p-Value
  1. DDLTA = Deceased donor liver transplant alone.

  2. KLTX = Combined kidney liver transplant.

DDLTA 11010 (97.4)13163 (95.6)<0.0001
KLTX 288 (2.6)602 (4.4) 
 Preoperative serum 
DDLTA0–0.99 mg/dL5706 (51.8)6067 (46.1)<0.0001
DDLTA1–1.99 mg/dL4033 (36.6)5074 (38.5) 
DDLTA≥2.0 mg/dL868 (7.9)1322 (10.0) 
DDLTARRT403 (3.7)700 (5.3) 

Table 7 lists demographic characteristics pre and post MELD. After the introduction of MELD, there has been an increase in patient age, more minorities transplanted and more males. As noted, there has been an increase in donor age and a lowering of cold ischemia time as evidenced by a drop in patients receiving livers with a cold ischemia time of >12 h.

Table 7.  Selected demographic data pre-MELD versus post-MELD
 Pre-MELD 4/27/1999–2/26/2002Post-MELD 2/27/2002–12/26/2004p-Value
Recipient age <0.0001
Donor age <0.0001
Male sex62.4%65.5%<0.0001
Recipient ethnicity 0.0002
Cold ischemia time <0.0001
 <6 h22.7%27.6% 
 6–<12 h64.5%62.8% 
 12+ h12.8%9.6% 

Despite the changes seen in Tables 6 and 7 (worse renal function, more patients on RRT, and more KLTX), there was no decrease in overall patient survival rates after OLTX following the introduction of MELD (Table 8, group g vs. group h, p = 0.200). Furthermore, there has been no difference pre and post MELD in patient survival in those patients receiving DDLTA with no RRT, DDLTA with RRT or KLTX.

Table 8.  Kaplan Meier patient survival rates by transplant type pre- and post-MELD
 Number of transplantsMonths post-transplantp-Value
  1. DDLTA = Deceased donor liver transplant alone.

  2. KLTX = Combined kidney liver transplant.

  3. RRT = Renal replacement therapy.

  4. 1Pre-MELD: 4/27/1999–2/26/2002.

  5. 2Post-MELD: 2/27/2002–12/26/2004.

 a. Pre-MELD10 60788.
 b. Post-MELD12 46388.583.579.8 
 c. Pre-MELD40374.269.867.50.658
 d. Post-MELD70075.772.168.2 
 e. Pre-MELD28884.579.974.70.895
 f. Post-MELD60285.179.373.1 
 g. Pre-MELD11 29887.683.580.20.200
 h. Post-MELD13 76487.782.878.9 

The results of the multivariate Cox regression analysis of time to death is presented in Table 9. As can be seen, renal function is an independent predictor of posttransplant survival while the Meld/PELD score is not. Other significant negative predictors include recipient in ICU, recipient on ventilated support, and cold ischemia time. Also recipient African American versus White had a relative risk (RR) of 1.39, p < 0.0001, recipient Hispanic versus White had a RR of 0.88, p = 0.06, and recipient Asian versus White had a RR of 0.72, p < 0.005. No diagnostic categories were significant. Other factors not reported here which were significant were age of recipient, age of donor, and donor ethnicity.

Table 9.  Multivariate Cox regression analysis of time to death within 9999 days of transplant for primary deceased donor liver transplants alone and liver-kidney transplants in the United States, 2/27/02–12/31/04
Variablep-ValueRR95% CI
  1. CI = confidence interval.

  2. CRT = creatinine in mg/dL.

  3. DDLTA = deceased donor liver transplant alone.

  4. KLTX = kidney liver transplant.

  5. M/P = MELD/PELD.

  6. RR = relative risk.

DDLTA-Crt 1–1.99 versus DDLTA Crt < 10.04611.11[1.00, 1.23]
DDLTA-Crt>2 versus DDLTA < 1<0.00011.58[1.36, 1.83]
DDLTA-RRT versus DDLTA Crt < 1<0.00011.77[1.47, 2.13]
KLTX versus DDLTA-Crt < 10.00041.44[1.18, 1.76]
Status 1 versus M/P < 150.8480.97[0.75, 1.27]
M/P 15–25 versus M/P < 150.1001.13[0.98, 1.31]
M/P >25 versus M/P < 150.05451.17[1.00, 1.37]
Recipient in ICU0.02981.22[1.02, 1.45]
Recipient on life support0.00011.47[1.21, 1.79]
Cold ischemia time 6–12 h versus <6 h0.02851.13[1.01, 1.26]
Cold ischemia time >12 h vs. < 6 h<0.00011.61[1.42, 1.83]


Preoperative renal function has long been known to be important in determining survival post-OLTX and studies throughout the 1990s emphasized this importance. Nair et al. examined the OPTN/UNOS database from 1988 to 1996 for the results in over 20 000 patients undergoing OLTX (2). They characterized patients according to calculated creatinine clearance with four groups. The four groups had average preoperative serum creatinine of 0.8, 1.4, 2.7 and 6.2 mg/dL. The average 5-year patient survival in these groups was 62%, 53%, 42% and 40%, respectively. They performed a multivariate analysis and found that preoperative serum creatinine was an independent predictor and had the highest Odds Ratio for predicting death post transplant. Interestingly, the patients in the latter two groups had an increased cause of death due to infection compared to the first group similar to an earlier paper which reported that increased infection rate correlated with renal function preoperatively (7). Another examination of the OPTN/UNOS database from 1988 to 1995 demonstrated that patients with a preoperative serum creatinine of ≥2.0 mg/dL had a 5-year patient survival of 50.4% compared to a 5-year patient survival of 62.2 % in patients undergoing KLTX (9). Markmann et al. examining single center data from 1992 to 1998 found a similar correlation with serum creatinine >2 mg/dL and patient survival (1). Furthermore, previous studies have demonstrated a deleterious effect of pretransplant renal dysfunction requiring RRT on posttransplant survival (5,11). Therefore, we wished to examine whether or not this had changed in recent years with the introduction of the MELD system for allocation, the use of newer immunosuppressive agents, better antimicrobial strategies, and an increased awareness of the deleterious effect of poor preoperative renal function on survival.

As demonstrated in Figure 1, Tables 1 and 9, preoperative renal function appears to play a strong role in predicting postoperative survival. As seen previously in the analysis of Nair et al. (2), poor renal function is associated with a high RR of death with the highest RR (1.77) being in those patients requiring RRT. Thus despite advances in RRT and knowledge of the high RR of death associated with poor renal function, renal function remains an independent predictor of death following liver transplant. One potential limitation in this analysis is lack of further definition of the severity of renal disease. We lack precise diagnosis and no data are available on an important determinant of renal disease, proteinuria. This parameter might add more sensitivity to predicting posttransplant mortality or in predicting recovery of renal function following liver transplantation. In data not shown, we also examined the effect of renal function on survival in repeat transplants. The same pattern seen in primary transplants (worse patient survival with increasing renal dysfunction) was also seen in repeat transplant recipients. The difference between patients with ideal renal function (creatinine <1 mg/dL) and those with worse renal function was similar. However, in all categories of renal function the survival was worse compared to primary transplant recipients. In all categories of renal function, the repeat transplant recipients had a 10–15% decrease in 3-year patient survival compared to the primary transplant recipients. Thus, great care must be used when evaluating renal function in a patient being considered for repeat liver transplant.

There does not appear to be a difference (at least in 5-year survival) in primary liver transplant recipients with serum creatinine up to 2 mg/dL but patients with a serum creatinine ≥2 mg/dL and those requiring RRT clearly have inferior survival. However, in comparing these to the 4 groups in the paper by Nair et al., it appears that overall survival has improved remarkably. The first group in the paper by Nair corresponds to the first group in this paper (Creatinine < 1 mg/dL) and the 5-year survival has improved from 62% to 79.1%. Group 2 in the paper by Nair would correspond to the second group in this paper (Creatinine 1–1.99 mg/dL) and the 5-year survival has improved from 53% to 72.2%. Similar improvement for patients in groups 3 and 4 by Nair are apparent in this paper. Similarly, when we compared two groups we had previously reported (10, creatinine ≥2 mg/dL and KLTX) we found an increase in 5-year survival. Patients with a preoperative creatinine ≥2 mg/dL had an increase in 5-year survival from 50.4 % to around 63.1% and KLTX 5-year survival has increased from 62.2% to 69.7% even in patients on RRT. It is unclear what accounts for this dramatic increase in survival. We cannot determine this from the OPTN/UNOS database but can only speculate that better and more aggressive renal care, enhanced immunosuppressive strategies with more agents to choose from, improved antimicrobial strategies, and better overall care of the OLTX patients has led to the overall improvement in patient survival.

A trend noticed before is seen again in this study. Patients with increased serum creatinine preoperatively appear to have better survival with KLTX compared to DDLTA. However, this is only the case in those patients requiring RRT. (3-year survival 68.3% DDLTA patients requiring RRT vs. 74.8% with KLTX on RRT, p = 0.0004). DDLTA patients with a preoperative serum creatinine ≥2 mg/dL who are not on RRT had a 3-year survival of 69.8% compared to 69.9% in KLTX with a creatinine >2.0 mg/dL not requiring RRT (p = 0.177). We have no obvious explanation for the 5% difference in 3-year patient survival in patient receiving KLTX who did or did not require RRT. If one examines the data in Table 1, there is no difference between these groups at 12 and 24 months. Furthermore, at month 12 and 24, patients receiving KLTX who were not on dialysis with a pre operative creatinine >2 mg/dL have a better survival than similar DDLTA patients. Awareness of the improved survival of patients receiving KLTX compared to those on dialysis may be leading to an increase in KLTX as seen in Figure 2. However, the analysis presented here differs from a previous analysis (10). We confined the current analysis to those patients who were receiving a first liver transplant. In that population, the advantage of KLTX (in terms of 3-year patient survival) only exists in those patients who require RRT preoperatively. Patients with a serum creatinine >2.0 mg/dL who were not on RRT did not benefit (in terms of 3-year patient survival) compared to the similar group of patients receiving KLTX. We cannot comment on patients undergoing repeat transplant as the number of those patients receiving a KLTX was too low to make a meaningful interpretation.

Almost all of the increase in the number of KLTX is seen in the years 2002–2004 which corresponds with the introduction of MELD. With the introduction of MELD into the U.S. allocation system, there was speculation that there would be an increase in patients with poor renal function and those requiring RRT in those undergoing OLTX. This has indeed been the case as demonstrated in Table 6. Comparing the 34 months prior to and the 34 months after the introduction of MELD, there has been an increase in the number of patients receiving KLTX, patients on RRT prior to DDLTA, and a shift of more patients with elevated preoperative serum creatinine. This is due to the nature of MELD and is what it is designed to accomplish; the allocation of livers to the sicker patients who would have the most benefit of OLTX. It appears that this may be driving the increase in KLTX.

Another prediction made before the introduction of MELD was that there could be a decline in patient survival due to known deleterious effect of elevated creatinine on patient survival following OLTX. Despite the changes seen in Table 6, there does not appear to be any decrease in survival (at least during the first 3 years) since the introduction of MELD (Table 8). Despite more patients receiving KLTX and more patients on RRT, patient survival has remained steady since the introduction of MELD. As seen in Table 9, status 1 or MELD score is not a good predictor of posttransplant survival while preoperative renal function along with several other factors are. This is not surprising as MELD was not designed for this. It was designed to predict mortality in patients with cirrhosis. When one examines Table 9 and other data not shown it is hard determine why the overall survival is not changed. There are several factors with increased RR of death which have increased (poor renal function, RRT, KLTX, recipients African American or Hispanic, age) However, cold ischemia time had decreased and this may contribute positively. Further investigations are needed but perhaps improving care may be contributing to the trend.

The increase in KLTX does raise a dilemma. Who is an appropriate candidate for KLTX? Should it be the patient with a creatinine above 2 mg/dL or the patient on RRT pre transplant? Increased allocation of deceased donor kidneys to the KLTX population may have deleterious effects on the population of patients awaiting renal transplant alone. At the end of 2004, there were 56 822 patients on the U.S. national deceased donor kidney waiting list. The median time waiting was 37.7 months and climbing. According to a recent report, a patient age 40–59 who was waiting transplant while on dialysis had a death rate of 4.6% the first year, 8% the second year and 9.2% the third year (19). These rates increased for older patients. Thus, any diversion of kidneys from the deceased donor kidney pool to KLTX may have a deleterious effect on patients waiting for kidney transplant. The data presented here would suggest that in primary liver transplants, it should only be the patients requiring RRT who should receive a concomitant kidney transplant. In the post-MELD era, KLTX patients had superior survival at 12, 24 and 36 months posttransplant compared to DDLTA patients requiring RRT at the time of transplant (Table 8 85.1%, 79.3% and 73.1% vs. 75.7%, 72.1% and 68.2%, respectively). However, there may be patients with a decreased GFR not on RRT who would benefit from KLTX. These patients should be identified by renal biopsy which may provide assessment of the severity of renal disease and predict long term renal function. Suggested guidelines based on population studies are that patients with more than 30–35% glomerulosclerosis or 25–30% interstitial fibrosis regardless of cause should be considered for combined transplant (20). This must be balanced against the possible nephrotoxic effects of calcineurin inhibitors although it could be argued that sirolimus based immunosuppression might be utilized in these patients. This issue must be discussed (particularly in view of the superior graft survival in the deceased donor kidney patients compared to the KLTX) and new guidelines must be developed before firm recommendations can be made. Several recent publications have addressed this issue and set the table for the future debate (20–22). Future studies need to concentrate on the cause of renal failure in pre liver patients which precipitates the need for RRT, length of RRT prior to transplant, whether early and aggressive treatment of Hepatorenal syndrome with newer agents to improve renal function will improve post transplant survival and the effect of renal failure which persists posttransplant on the short and long-term survival of the liver transplant recipient. Furthermore, it has been suggested that prospective measurement of renal blood flow in patient with HRS or functional renal failure may help predict reversibility versus nonreversibility of renal dysfunction and improve the selection of patients for KLTX versus DDLTA (22) Finally, future studies are needed to determine if early aggressive pre operative, intra operative, and post operative RRT could improve the results in DDLTA patients with pre operative serum creatinine of >2.0 mg/dL.