• Graft failure;
  • liver transplant;
  • outcome;
  • primary nonfunction


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

PNF following liver transplantation (LT) is an infrequent but life-threatening complication. Liver allocation under MELD is based upon recipient severity of illness, a known risk factor for the occurrence of PNF. The incidence of PNF since the application of MELD has not previously been reported. The SRTR database was studied since inception of MELD until September 2004 for all adult recipients of deceased donor LT. PNF was defined as graft loss or death within 14 days of LT secondary to PNF or without defined cause.

A total of 10545 transplants met inclusion criteria and PNF occurred in 613 (5.81%) of recipients. Univariate analysis demonstrated donor age, serum creatinine >1.5 mg/mL, hypertension and CVA as risk factors for PNF. Recipient factors included life support, mechanical ventilation, use of inotropes, hemodialysis, initial status 1 and use of a shared transplant. In the multivariate model only donor age and recipient serum creatinine, bilirubin, on life support and status 1 at transplant were significant risk factors for PNF.

In this analysis of PNF in the MELD era the incidence of PNF does not appear to have increased from prior reports. Risk factors for PNF are related to donor age and severity of recipient illness.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

Over the past 10 years the number of individuals awaiting orthotopic liver transplantation has grown from 3955 in 1994 to 17171 in 2003. During this same time, the number of deceased donors has only increased from 5099 in 1994 to 6455 in 2003 (1,2). This disparity has lead to an increase in the number of deaths on the waiting list and as a consequence the transplant community has greatly expanded the use of nonideal donors to improve the rate of transplant (1). A number of characteristics have been used to describe the expanded liver donor pool and include older age, steatotic grafts, seropositive for hepatitis B and or C, donors after cardiac death (DCD) and split livers (3,4). Only recently a definition of expanded criteria donors (ECD) has come into acceptance characterized by a risk index score that predicts allograft failure at 1 year. Donor factors associated with failure include age >40, African American race, CVA as a cause of death, DCD and split livers (5). The use of these ECDs while providing organ replacement for individuals in dire need of an allograft can result in initial poor graft function (IPF) and a higher incidence of primary graft nonfunction (PNF). Known risk factors for the occurrence of PNF have been reported to include older-age donors, steatotic livers, DCD donors, prolonged ischemic times and high-risk recipients (6–8).

The final rule on liver allocation was approved and instituted in February 2002 in an effort to reduce the number of deaths on the waiting list and to make allocation of scarce resources more equitable. This rule established the model for end-stage liver disease (MELD) as the primary determinant for deceased donor liver allocation. Initial review of the results of OLTx after inception of MELD have demonstrated decreased mortality on the waitlist, a decrease in the number of new registrations and transplantation of more acutely ill recipients (9).

The combination of an increased utilization of ECD organs and transplantation of more acutely ill individuals should be expected to yield suboptimal results compared to prior periods. Despite this seemingly contradictory allocation schema, results reported to date have not demonstrated inferior outcomes in the MELD era compared to prior periods (10). We have hypothesized that the increased use of nonideal organs and allocation based entirely on severity of recipient disease would increase the likelihood of PNF. In an effort to define risk factors for PNF in the MELD era we conducted a review of all deceased donor OLTx performed since the inception of MELD for the occurrence of PNF and a multivariate analysis of risk factors for PNF.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

To study the incidence and risk factors for PNF, we used the SRTR/UNOS liver and donor datasets. Specifically, SRTR datasets for liver transplant candidate registration, follow-up and transplant were linked with corresponding deceased donor datasets to produce the final data for analysis.


Subjects for inclusion into the study were selected from among all adult first liver transplant recipients from the inception of MELD allocation until September 2004. From the linked dataset, subjects meeting the following criteria were considered to have PNF: graft loss, retransplantation or patient death within 14 days of initial transplantation not secondary to hepatic artery thrombosis (HAT), biliary complication, recurrent disease or acute rejection.

Missing data

Variables with greater than 25% of data missing or unknown were excluded from further analysis. Missing data from specific variables were addressed as follows; for continuous variables missing data were replaced with the series mean value. Missing categorical variables were replaced with the result of least interest. Missing donor and recipient-related categorical data are detailed in Table 1. The following continuous data were replaced with series mean values; recipient serum creatinine (n = 409), recipient total bilirubin (n = 453), recipient warm ischemia (n = 2337), recipient cold ischemia (n = 1463), donor BUN (n = 81), donor creatinine (n = 38), donor total bilirubin (n = 307), donor SGOT (n = 279), donor SGPT (n = 287), donor and recipient age (n = 1).

Table 1.  Missing categorical data
 Number missing
Donor factors
Prerecovery Meds
 >3 inotropes38
 ETOH dependency126
 CVA as COD1
Recipient factors
 Variceal hemorrhage976
 Portal vein thrombosis665
 Prior abdominal surgery830
 Mechanical ventilation969
 Use of inotropes1203

Implausible data were replaced with the mean of the series. Implausible data were considered BMI ≥ 60 or ≤ 10, age ≥ 90, serum creatinine ≥ 25 mg/dL, total bilirubin ≥ 70, warm ischemia ≥ 150 min or ≤ 20 min, cold ischemia ≥ 30 h or ≤ 1 h. Donor race was recoded into the following categories Caucasian (n = 7144), African American (n = 1149), Asian (n = 160), American Indian (n = 37) and other (n = 98).


A univariate analysis was performed to determine potential risk factors for inclusion in the multivariate model. Categorical variables were compared using chi-square test with Fisher's exact test where appropriate and continuous data were compared using Student's t-test. Continuous data with greater than two groups was compared using a one-way analysis of variance (ANOVA). All factors with p ≤ 0.10 were included in the binary multivariate logistic regression model with the occurrence of PNF considered the endpoint. In the multivariate model donor age was converted to categorical data using the following age ranges; less than 40 years, 41–50 years, 51–60 years and greater than 60 years of age. Recipient total bilirubin was categorized into the following groups (0–3.99 mg/dL, 4–7.99 mg/dL, 8–11.99 mg/dL and >12.0 mg/dL. Recipient serum creatinine was also categorized using the following groups; 0–1.99 mg/dL, 2–3.99 mg/dL and >4.0 mg/dL. Recipient covariates for the multivariate model included gender, need for dialysis, life support, mechanical ventilation, inotropes, status 1 at listing, received a shared transplant, serum bilirubin and creatinine. Donor covariates included age, gender, race, serum creatinine >1.5 mg/dL, CVA as cause of death, history of hypertension, meet kidney ECD criteria and DCD donor. Because ECD criteria include CVA as cause of death, hypertension and serum creatinine >1.5 mg/dL, only meeting kidney ECD criteria was included in the multivariate model. The multivariate model was constructed in a forward stepwise fashion with p ≤ 0.05 considered significant. All statistics were computed using SPSS 12.0 (Chicago, IL). Data are presented as mean ± standard error of the mean (SEM).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References


From the SRTR SAF files a total of 10545 first liver transplants were available for further analysis. Using the definition of PNF described above a total of 613 (5.81%) subjects met the criteria for PNF. Recipient characteristics of the PNF group and non-PNF group are noted in Table 2. In general recipients experiencing PNF were well matched with those candidates not experiencing PNF. There was no difference in age, sex, race or BMI. The vast majority of liver transplant recipients were Caucasians with African Americans and Asians accounting for virtually all of the remaining recipients. Only a small proportion of recipients were of Arabic, Indian, American Indian or other descent. There was no contribution of recipient race to the development of PNF. In addition, previously defined risk factors for PNF, such as need for dialysis, mechanical ventilation and on life support or inotropes, were found more frequently in the PNF than non-PNF groups (Table 2). In a similar fashion, reflecting overall severity of illness, individuals with status 1 standing at listing had a much higher incidence of PNF. Laboratory studies demonstrated a difference between groups in that recipient total bilirubin and serum creatinine were significantly higher in the PNF group. Despite this finding, first and last MELD scores were not different between groups. Ischemia times, both warm and cold, have long been considered a significant risk factor for both initial poor function (IPF) as well as PNF, were only different for cold ischemia. Warm ischemic times were nearly identical while cold ischemia times differed by 42 min between groups.

Table 2.  Recipient characteristics
Age51.8 ± 0.1051.5 ± 0.4351.9 ± 0.100.369
Sex (M (%))7126 (67.6%)395 (64.4%)6731 (67.8%)0.087
BMI28.3 ± 0.0528.4 ± 0.2428.2 ± 0.050.448
 Caucasian8949 (84.8%)509 (83.0%)8440 (85.0%) 
 African American893 (8.47%)63 (10.3%)830 (8.36%) 
 Asian441 (4.18%)27 (4.40%)414 (4.17%) 
 American Indian52 (0.49%)2 (0.33%)50 (0.50%) 
 Pac-Islander20 (0.19%)1 (0.16%)19 (0.19%) 
 Other190 (1.80%)11 (1.79%)179 (1.80%) 
Risk factors
 Life support616 (5.84%)85 (13.9%)531 (5.34%)<0.0001
 PVT275 (2.61%)19 (3.1%)256 (2.6%)0.431
 Variceal hemorrhage468 (4.43%)32 (5.22%)436 (4.39%)0.333
 Ascites7145 (67.8%)408 (66.6%)6737 (67.8%)0.513
 Ventilator463 (4.39%)57 (9.30%)406 (4.09%)<0.0001
 Inotropes333 (3.16%)46 (7.50%)287 (2.89%)<0.0001
 Dialysis650 (6.16%)52 (8.48%)598 (6.02%)0.014
 Shared Txp2788 (26.4%)205 (33.4%)2583 (26.0%)<0.0001
 Initial status 1384 (3.64%)56 (9.14%)328 (3.30%)<0.0001
Laboratory values
 First MELD17.8 ± 0.0818.3 ± 0.2317.8 ± 0.070.050
 Last MELD23.5 ± 0.0623.3 ± 0.3323.5 ± 0.080.675
 Total Bili (mg/dL)3.01 ± 0.057.79 ± 0.372.72 ± 0.040.0001
 Serum Cr (mg/dL)1.27 ± 0.011.82 ± 0.061.23 ± 0.0080.0001
 Warm ischemia (min)40.8 ± 0.1741.99 ± 0.7640.73 ± 0.170.077
 Cold ischemia (h)7.76 ± 0.038.40 ± 0.147.71 ± 0.030.0001

Donor characteristics are presented in Table 3. Donor age was significantly older in those experiencing PNF, and there was a higher proportion of female donors in the PNF group. A higher number of African American (16.2% vs. 13.3%) and Asian donors (2.94% vs. 1.73%) developed PNF but donor race did not achieve statistical significance (p = 0.075). Donor factors for increased risk of renal graft loss including history of hypertension, serum creatinine >1.5 mg/dL, stoke as cause of death and the composite risk factor of meeting kidney criteria for ECD donor were significantly associated with PNF (Table 3). Other risk factors such as anti-HCV+, HBsAg+ and proteinuria were not identified as associated with PNF (data not shown). Pre-recovery medications such as vasopressors, steroids, inotropes, diuretics, T3, or use of three or more inotropes were not significantly associated with PNF. A reduction in the incidence of PNF was exhibited by T4 use. None of the laboratory studies tested, donor total bilirubin, SGOT and SGPT, were different between the PNF and non-PNF groups.

Table 3.  Donor characteristics
Age40.7 ± 0.1745.3 ± 0.7340.7 ± 0.170.0001
 Sex (M (%))6289 (59.6%)340 (55.5%)5949 (59.9%)0.030
 BMI26.1 ± 0.0526.53 ± 0.2326.16 ± 0.070.193
 Caucasian8771 (83.2%)486 (79.3%)8285 (83.4%) 
 African American1317 (12.5%)99 (16.2%)1317 (13.3%) 
 Asian190 (1.80%)18 (2.94%)172 (1.73%) 
 American Indian48 (0.46%)2 (0.33%)46 (0.46%) 
 Pac-Islander38 (0.36%)2 (0.33%)36 (0.37%) 
 Other82 (0.78%)6 (0.98%)76 (0.77%) 
Laboratory values
 Creat1.38 ± 0.021.43 ± 0.061.37 ± 0.020.376
 Total Bili1.00 ± 0.021.02 ± 0.111.00 ± 0.020.763
 SGPT61.8 ± 1.3362.7 ± 6.861.7 ± 1.350.863
 SGOT71.9 ± 1.6778.9 ± 10.978.9 ± 1.650.999
Risk factors N (%)
 Hypertension2888 (27.4%)205 (33.4%)2683 (27.0%)0.027
 Serum Cr > 1.5 mg/dL2012 (19.1%)136 (22.2%)1876 (18.9%)0.004
 Death by stroke4695 (44.5%)328 (53.5%)4367 (44.0%)0.001
 Meet kidney ECD2546 (24.1%)208 (33.9%)2338 (23.5%)0.0001
Prerecovery medications given
 Steroids6882 (65.3%)380 (62.0%)6502 (65.5%)0.079
 Diuretics4545 (43.1%)279 (45.5%)4266 (43.0%)0.214
 T43335 (31.6%)161 (26.2%)3174 (32.0%)0.003
 DDAVP4331 (41.1%)237 (38.7%)4094 (41.2%)0.212
 Dopamine2951 (27.9%)172 (28.1%)2779 (28.0%)0.967
 Dobutamine161 (1.53%)15 (2.45%)146 (1.47%)0.056
 Inotropic support5990 (56.8%)353 (57.6%)5637 (56.8%)0.687
 Three or more inotropes331 (3.14%)13 (2.12%)318 (3.20%)0.136
 DCD donor212 (2.015)18 (2.94%)194 (1.95%)0.092

From this analysis significant variables (p < 0.10) were entered into a univariate logistic regression and data are shown in Table 4. Both donor and recipient variables weighed heavily on the development of PNF. Donor age >60 was the most significant donor related factor associated with PNF, increasing the odds of its development by 2.01 (1.62–2.49 95% CI). Donor criteria for classification as an ECD including hypertension, CVA and serum Cr >1.5 mg/dL also consistently demonstrated increased odds of PNF development, as did the composite of a liver donor meeting criteria for a renal ECD status. The latter increased the OR of PNF by 1.67 (1.40–1.98 95% CI). Recipient variables contributing to an increased OR of PNF consisted almost exclusively of those associated with severity of illness. The requirement for life support including mechanical ventilation, inotropes and dialysis as well as status 1 listing increased the OR of PNF ranging from 1.45 to 3.14. While final MELD was not associated with PNF, total bilirubin and serum creatinine heavily influenced the development of PNF. Total bilirubin >12 mg/dL increased the OR of PNF development to 10.96 (8.64–13.91 95% CI) and serum creatinine >4 similarly increased to OR of PNF to 4.41 (3.05–6.37 95% CI). The only variables not associated with severity of recipient illness to contribute to the development of PNF were use of a shared transplant and CIT >10 h, which had OR of PNF of 1.43 (1.20–1.70 95% CI) and 1.36 (1.10–1.68 95% CI).

Table 4.  Univariate analysis
VariablesOdds ratios95% CIp
 <40 years
 41–50 years1.240.99–1.570.064
 51–60 years1.571.25–1.96<0.0001
 >60 years2.011.62–2.49<0.0001
 Gender (female)0.830.71–0.980.0301
 Serum Cr >–1.490.0441
 ECD criteria1.671.40–1.98<0.0001
 DCD donor1.520.89–2.410.0947
Prerecovery Meds
 Sex (female)0.860.73-1.020.0873
 Life support2.852.22–3.62<0.0001
 Shared Txp1.431.20–1.70<0.0001
 Status 13.142.38–4.13<0.0001
Total Bili
 0–3.99 mg/dLReference 
Total creatinine
 0–1.99 mg/dLReference 
Cold ischemia time
 0–10 hReference 
 >10 h1.361.10–1.680.004

Recipient severity of illness factors predominated in the multivariate analysis (Table 5). Renal function and serum bilirubin showed a higher odds ratio of developing PNF, with progressively increasing values. Recipient status 1 or on life support also increased the likelihood of developing PNF with odds ratios of 1.82 and 2.17, respectively. Of the donor variables included in the multivariate model, only donor age greater than 60 achieved significance.

Table 5.  Multivariate analysis
Significant variables Odds ratio95% CIp
Recipient factors
Status 1 1.821.27–2.600.001
Life support 2.171.31–3.590.003
Total bilirubin0–3.99 mg/dLReference 
Serum creatinine0–1.99 mg/dLReference 
Donor factors
Donor age
<40 yearsReference 
41–50 years 1.080.85–1.380.515
51–60 years 1.340.99–1.800.057
>60 years 1.571.01–2.450.047


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

In this SRTR analysis of PNF after deceased donor liver transplant in the MELD era the most striking finding is a relative stability in the occurrence of PNF under the new allocation system. Our hypothesis at the inception of this project was that as programs transplanted individuals with a high severity of illness (higher MELD), PNF would occur more frequently. Additionally, with a more liberal policy on acceptable organs including older donors, more significant steatosis, donor instability, graft dysfunction in the form of PNF would be more common. After analyzing more than 10 000 deceased donor, adult first transplants, 613 (5.8%) cases of PNF, based upon our definition, were identified. In preparing for this project we estimated that the frequency of PNF would be substantially greater than 5%. This was based on single-center publications and an SRTR analysis of donors after cardiac death (DCD), in which PNF occurred in 6.4% of brain dead donors vs. 11.8% of donors after cardiac death (7). In this publication the author's definition of PNF was similar to ours with the exception of not excluding those with graft loss secondary to hepatic artery thrombosis. The incidence of PNF from other large single-center series has varied widely from 6.0–9.2% (11,12). In smaller series, rates of PNF have been much lower, 2.9% from 276 liver transplants at a single center (13). The reasons for the disparity in rates between large- and smaller-center series can only be speculated upon. It is possible that large centers are inherently more aggressive in both utilization of deceased donor organs and in severity of recipient illness considered acceptable for transplantation. The SRTR database represents all transplant programs and thus the rate of PNF seen in small volume programs may greatly offset that seen in larger, more aggressive centers. Another plausible explanation is that our analysis has underestimated the occurrence of PNF. We believe the latter explanation to be unlikely. In capturing all patient deaths and retransplants within 14 days of initial transplantation, it is unlikely that we excluded significant numbers of PNF events. Lastly, by excluding those with known etiology of graft failure we are unlikely to overestimate the incidence of PNF.

A number of donor factors were identified in the univariate analysis as associated with PNF including donor age, gender and other donor risk factors (hypertension, serum Cr >1.5 mg/dL, CVA as cause of death and meet kidney ECD criteria. Donor age as predictive of PNF is not surprising, as a number of publications have reported on outcomes utilizing older deceased donor livers for transplantation. In an SRTR analysis, donor age greater than 40 was the strongest predictor of graft loss and death, for individuals transplanted for viral hepatitis. In this multivariate analysis of greater than 11 000 liver transplant recipients, donor age greater than 40 was associated with a 1.67 increased risk of graft loss in HCV-infected recipients that increased to 2.21 when donor age was greater than 60 years (14). In a national review of 5150 liver transplants performed in Spain from 1994 to 2001, using a Cox proportional hazards model, the authors identified donor age greater than 50 and greater than 70 years with an increased risk of graft loss (RR = 1.27 and 1.4, respectively) (15). Despite the significance of the data, many of the events studied are well beyond that expected for PNF.

While these studies have demonstrated the impact of donor age on long-term graft survival, the impact of donor age on PNF has only been reported on a much smaller scale. In a single-center review of 400 liver transplants, donor age was not found to increase the incidence of PNF (16) and other studies have confirmed this apparent lack of association of donor age with PNF (15,17,18). A review of the NIDDK liver transplant database looking at donor age and outcome has suggested a correlation between advanced donor age and PNF. The authors divided donors into two groups aged 6–49 and aged 50–73. PNF was seen in 7 of 579 (1.2%) in the young donor group vs. 5 of 193 (2.6%) in the older group. While the paper did not specifically evaluate the rates of PNF or retransplant, the 3-month, 1- and 2-year patient survival rates were less in recipients of the older donors (19). In this SRTR analysis, donor age was found to be significantly and independently associated with PNF, when analyzed as both a continuous and categorized variable. In fact donor age greater than 50 increased the odds ratio of PNF by 1.57 (1.25–1.96 95% CI) and for age greater than 60 by 2.01 (1.62–2.49 95% CI) compared to the reference group consisting of donor age <40. This study conclusively demonstrates the role of donor age in the development of PNF.

In this SRTR analysis, female deceased donor organs were found to reduce the OR of PNF by 17% in the univariate analysis but when corrections were made for other predictive variables in the multivariate model this association did not achieve statistical significance. The role of donor gender in predicting outcomes has been studied in a number of transplanted organs. In renal transplant, female grafts have been found to portend worse long-term outcomes (20). Speculation from authors has centered primarily on a reduced nephron mass transplanted into the recipient but other confounding variables such as donor age may contribute to this identified outcome. In a large database review donor gender was not found to predict outcome in liver transplant (20) but smaller single-center studies have shown adverse outcomes with gender mismatch, specifically female donor to male recipient (21,22). While subgroup analysis, from this report, demonstrated that female donors were older and more likely to meet kidney ECD criteria (data not shown) than males, adjustment in the multivariate model showed gender to not be a significant predictor of PNF.

One of the more striking findings in this study was the identification of meeting kidney ECD criteria as a risk factor for PNF in liver transplantation. Kidney ECD criteria includes donor age greater than 50 and 2 of 3 risk factors among CVA as cause of death, terminal creatinine >1.5mg/dL or history of hypertension or age greater than 60 and no additional risk factors. This constellation of findings (CVA as cause of death, age greater than 50, elevated serum creatinine and history of hypertension) is a risk factor for adverse outcomes in renal transplantation and has been used to define expanded criteria donors (ECD) in renal transplantation. Kidneys transplanted from these donors are associated with an increased risk of delayed graft function (DGF), and relative risk of graft failure of 1.7 or greater than a reference group consisting of donors aged 10–39 years (23). It is likely that these factors may also have an adverse effect on liver function. We have previously seen the importance of age in determining outcomes after liver transplantation but have not evaluated the role of elevated serum creatinine, death by CVA and hypertension (15–17). In Feng's paper on graft failure after liver transplant, donor age >40 and CVA as a cause of death were noted to be significant risk factors for graft loss (5). It is possible that these findings, in conjunction with advanced age are evidence of end organ damage from atherosclerotic disease that adversely affects the hepatic function, hepatocyte regeneration or viability following procurement (24).

Recipient factors played a dominant role in influencing the occurrence of PNF. Factors noted to contribute to PNF centered on severity of recipient illness. In the univariate model life support, mechanical ventilation, use of inotropes, hemodialysis and status 1 were more frequently seen in those with PNF. The role of life support impacting outcomes following OLTx has been studied in a number of single-center studies. In a paper by Markmann, primary graft survival was reduced by recipient creatinine greater than 2.0 mg/dL and by pretransplant mechanical ventilation using a multivariate model. In this analysis the combination of these two factors dramatically reduced 1-year graft survival to 41% vs. 76% with many of the losses occurring very early (25). In a review of transplantation for fulminant hepatic failure the authors noted that 13.2% of grafts were lost secondary to PNF, a rate much higher than seen in any group thus far. Hemodialysis increased the rate of graft loss three-fold in this review (26). The role of life support was not found to be significant in the multivariate model. While recipient serum creatinine and serum bilirubin were independent predictors of PNF, MELD score was not found in any of the models to predict PNF. A number of plausible explanations can be hypothesized such as discordance between SRTR reported MELD scores and final MELD scores at transplantation. It is also possible that the mathematical transformation of the predictor variables in the MELD score alter the contribution to outcome of PT (INR), creatinine and bilirubin. Lastly, we did not adjust MELD scores for points granted by regional review boards and this may also have altered the predictive value of MELD in the model.

While the incidence of PNF appears to be stable in the MELD era, it is important to analyze the rate of PNF in the context of changes in recipient risk factors that have occurred over time. For example the rate of hospitalization at the time of OLTx has declined over the past decade from 26% in 1995 to just 16% in 2004. Similarly, the percentage of recipients transplanted on life support has also declined from 14% in 1995 to 7% in 2004 (27). While a reduction in this risk factor for PNF would be expected to reduce the PNF rate in the MELD era this was not seen and may be explained by the rising incidence of renal failure in the MELD era recipients. The rate of combined liver-kidney transplantation has increased from 3% in 2001 to 5% in 2004 and is reflective of the importance of renal dysfunction in determining liver allocation priority under MELD. Aging of the donor population may also be adversely impacting the PNF rate in the MELD era. The average age of deceased donors has increased from 34.2 years in 1995 to 39.8 years in 2004 but more importantly the percentage of older donors (age 50–64 years) has increased from 19% in 1995 to 25% in 2004 (28).

An important advantage of liver allocation under MELD may be that individuals are transplanted before they become seriously ill, as reflected by a reduction in individuals transplanted who are hospitalized, ICU bound or on life support. This apparent reduction in severity of illness would be expected to reduce the incidence of PNF under MELD, which was not seen in this analysis. It is probable that a rising incidence of renal dysfunction coupled with an advancing donor age have offset this and resulted in an apparent stable PNF rate.

In summary, a SRTR database analysis was conducted to determine the incidence and risk factors for PNF under the current allocation system. The analysis has not demonstrated an increase in the incidence of PNF as liver allocation has shifted to a severity of illness-based algorithm. In the final multivariate analysis of risk factors for PNF severity of recipient disease and donor age where the principal risk factors for the occurrence of PNF.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

Support for this project was provided by the Julie Henry Fund at Beth Israel Deaconess Medical Center.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  • 1
    Delmonico FL, Sheehy E, Marks WH, Baliga P, McGowan JJ, Magee JC. Organ donation and utilization in the United States, 2004. Am J Transplant 2005; 5(4 Pt 2): 862873.
  • 2
    Hanto DW, Fishbein TM, Pinson CW et al. Liver and intestine transplantation: Summary analysis, 1994–2003. Am J Transplant 2005; 5(4 Pt 2): 916.
  • 3
    Briceno J, Marchal T, Padillo J, Solorzano G, Pera C. Influence of marginal donors on liver preservation injury. Transplantation 2002; 74: 522526.
  • 4
    Busuttil RW, Tanaka K. The utility of marginal donors in liver transplantation. Liver Transpl 2003; 9: 651663.
  • 5
    Feng S, Goodrich NP, Bragg-Gresham JL et al. Characteristics associated with liver graft failure: The concept of a donor risk index. Am J Transplant 2006; 6: 783790.
  • 6
    Strasberg SM, Howard TK, Molmenti EP, Hertl M. Selecting the donor liver: Risk factors for poor function after orthotopic liver transplantation. Hepatology 1994; 20(4 Pt 1): 829.
  • 7
    Abt PL, Desai NM, Crawford MD et al. Survival following liver transplantation from non-heart-beating donors. Ann Surg 2004; 239: 87.
  • 8
    Furukawa H, Todo S, Imventarza O et al. Effect of cold ischemia time on the early outcome of human hepatic allografts preserved with UW solution. Transplantation 1991; 51: 1000.
  • 9
    Freeman RB, Jr., Wiesner RH, Roberts JP, McDiarmid S, Dykstra DM, Merion RM. Improving liver allocation: MELD and PELD. Am J Transplant 2004; 4 (Suppl 9): 114131.
  • 10
    Freeman RB, Harper A, Edwards EB. Excellent liver transplant survival rates under the MELD/PELD system. Transplant Proc 2005; 37: 585588.
  • 11
    Busuttil RW, Farmer DG, Yersiz H et al. Analysis of long-term outcomes of 3200 liver transplantations over two decades: A single-center experience. Ann Surg 2005; 241: 905.
  • 12
    Jain A, Reyes J, Kashyap R et al. Long-term survival after liver transplantation in 4,000 consecutive patients at a single center. Ann Surg 2000; 232: 490.
  • 13
    Colina F, Lopez-Carreira M, Moreno E et al. A clinopathologic review of 8 liver graft primary nonfunctions. Hepatogastroenterology 1995; 42: 212.
  • 14
    Lake JR, Shorr JS, Steffen BJ, Chu AH, Gordon RD, Wiesner RH. Differential effects of donor age in liver transplant recipients infected with hepatitis B, hepatitis C and without viral hepatitis. Am J Transplant 2005; 5: 549.
  • 15
    Cuende N, Grande L, Sanjuan F, Cuervas-Mons V. Liver transplant with organs from elderly donors: Spanish experience with more than 300 liver donors over 70 years of age. Transplantation 2002; 73: 1360.
  • 16
    Busquets J, Xiol X, Figueras J et al. The impact of donor age on liver transplantation: Influence of donor age on early liver function and on subsequent patient and graft survival. Transplantation 2001; 71: 1765.
  • 17
    Grande L, Matus D, Rimola A et al. Expanded liver donor age over 60 years for hepatic transplantation. Clin Transpl 1998: 297301.
  • 18
    Washburn WK, Johnson LB, Lewis WD, Jenkins RL. Graft function and outcome of older (> or = 60 years) donor livers. Transplantation 1996; 61: 10621066.
  • 19
    Hoofnagle JH, Lombardero M, Zetterman RK et al. Donor age and outcome of liver transplantation. Hepatology 1996; 24: 8996.
  • 20
    Zeier M, Dohler B, Opelz G, Ritz E. The effect of donor gender on graft survival. J Am Soc Nephrol 2002; 13: 25702576.
  • 21
    Francavilla R, Hadzic N, Heaton ND et al. Gender matching and outcome after pediatric liver transplantation. Transplantation 1998; 66: 602605.
  • 22
    Marino IR, Doyle HR, Aldrighetti L et al. Effect of donor age and sex on the outcome of liver transplantation. Hepatology 1995; 22: 17541762.
  • 23
    Port FK, Bragg-Gresham JL, Metzger RA et al. Donor characteristics associated with reduced graft survival: An approach to expanding the pool of kidney donors. Transplantation 2002; 74: 12811286.
  • 24
    Iakova P, Awad SS, Timchenko NA. Aging reduces proliferative capacities of liver by switching pathways of C/EBPalpha growth arrest. Cell 2003; 113: 495506.
  • 25
    Markmann JF, Markmann JW, Markmann DA et al. Preoperative factors associated with outcome and their impact on resource use in 1148 consecutive primary liver transplants. Transplantation 2001; 72: 11131122.
  • 26
    Farmer DG, Anselmo DM, Ghobrial RM et al. Liver transplantation for fulminant hepatic failure: Experience with more than 200 patients over a 17-year period. Ann Surg 2003; 237: 666676.
  • 27
    Shiffman ML, Saab S, Feng S et al. Liver and intestine transplantation in the United States, 1995–2004. Am J Transplant 2006; 6(5 Pt 2): 11701187.
  • 28
    Marks WH, Wagner D, Pearson TC et al. Organ donation and utilization, 1995–2004: Entering the collaborative era. Am J Transplant 2006; 6(5 Pt 2): 11011110.