Liver Transplantation in the Era of Model for End-Stage Liver Disease

[hanklive39]

From Liver International
Liver Transplantation in the Era of Model for End-Stage Liver Disease
Posted 03/18/2004
Victor S. Wang; Sammy Saab
Abstract and Introduction
Abstract

Liver transplantation is challenged by organ shortage and prolonged waiting list time. The goal of the ideal organ allocation system is to transplant individuals least likely to survive without a liver transplantation, and maintain appropriate rates of postoperative survival. Currently, liver allocation in the United States is based on the model for end-stage liver disease (MELD). Studies have shown MELD to be objective and accurate in predicting short-term survival in patients with cirrhosis.
Introduction

End-stage liver disease is a major public health dilemma. In 1997, liver disease was associated with over 23 000 deaths and was the 10th leading cause of mortality in the United States.[1] As of July, 2003, there were 17 582 patients awaiting a liver transplantation.[2] The number of transplants performed has not kept up with the growing number of individuals listed. This has lead to substantial mortality in patients awaiting liver transplantation.

In the United States, liver organ allocation has gone through many stages of evolution. From an individual patient's perspective, transplantation should occur before the occurrence of life threatening complications and the substantial impairment in quality of life. From a public health perspective, the goal of an allocation system should be to reduce waiting list mortality while maintaining post-transplantation survival. In 2002, the United Network of Organ Sharing (UNOS) adopted the model for end-stage liver disease (MELD) as the new model to assess disease severity and thus determine organ allocation, and a new era of organ procurement began. This paper reviews the past and the present in liver allocation and will discuss future implications and directions in the area of liver allocation.

Victor S. Wang1, Sammy Saab2,3

1VA Greater Los Angeles Health Care System, 2Division of Digestive Disease, and 3Dumont UCLA Liver Transplant Center, University of California, Los Angeles, CA, USA

Prior to the MELD Score

The Child-Turcotte classification system was developed in 1964 for assessing operative risk in patients with end-stage liver disease with variceal bleeding undergoing surgical shunting.[3] This system comprised of five empirically chosen parameters: serum albumin, serum bilirubin, ascites, encephalopathy and nutritional status. To make the classification less cumbersome to use in clinical practice, Pugh modified the scoring system in 1972 by exchanging nutritional status with prothrombin time.[4] The new founded Child-Turcotte-Pugh (CTP) score later became well established as a system that penetrated both clinical and research applications in liver disease. Through the years, many other prognostic models have been developed but none were able to replace the CTP score.[5,6] The durability of the CTP was partly due to its simplicity of use, its reliance on readily available information, and its ability to be used at the bedside.

In 1983, a NIH consensus development conference affirmed liver transplantation was no longer considered an experimental procedure and was deemed to be a therapeutic modality for end-stage liver disease that deserved broader application. However, soon after the institution of orthotopic liver transplantation as a common procedure, the demand of liver organs greatly surpassed the supply of available organs. This ultimately led to the policy of 'sickest first', which allocates livers to those with the highest priority. To account for the level of urgency, the UNOS allocation system was devised to stratify patients on the waiting list. (Table 1) Because of the limited number of statuses, waiting time often became the tiebreaker.

To prevent liver transplant centers from listing patients before there was true indication for liver transplantation thus allowing accrual of valuable wait time, a 'minimal listing criteria' was developed in 1997[7] (Table 2). While this system helped standardized the indication for liver transplantation, the criteria failed to stratify the urgency status of patients already on the wait list. Even after its institution there was still a wide variability of urgency within each UNOS category.

Later in 1998, two other changes had taken place. First UNOS redefined urgency by using CTP scoring rather than hospital admissions as an indicator of severity. Consequently, under the new revised rules, the CTP score was calculated on a regular basis for each patient on the waiting list. Second, the Institute of Medicine critiqued the lack of urgency stratification in the current liver allocation in the US.[8] Their report concluded that a system that de-emphasized waiting time would be preferable to the current system. The need to eliminate waitlist time as a major determinant of organ allocation was subsequently supported by Freeman et al.[9] The authors studied 16 414 patients on the UNOS waiting list and found that center-specific waiting time accounted for almost none of the center-specific 2-year mortality.

The following year, the Department of Health and Human Services issued an amended Final Rule under the National Organ Transplant Act. This Rule proposed that: (1) organs should be allocated to candidates in the order of medical urgency; (2) the role of waiting lists should be minimized; and, (3) efforts should be made to avoid futile transplantation and ensure efficient use of scarce organs. The Rule also proposed that allocation of organs be performed under strict medical criteria, founded by the transplantation community. The response to this call came from UNOS, which eventually lead to the development of the MELD score.

To be continued in next reply...

[hanklive39]

Development and Validation of the MELD Score

The MELD was developed originally to predict 3-month survival in patients undergoing transjugular intrahepatic portosystemic shunts (TIPS). Malinchoc et al.[10] studied 231 patients from four different medical centers in the United States undergoing TIPS. Cox proportional-hazards regression of this data found four predictors of survival: total serum bilirubin, serum creatinine, the International Normalized Ratio (INR), and the etiology of liver disease (Appendix[11]). Using these variables, a model was created and was found to be superior to the Child-Pugh score in predicting survival. Furthermore, the authors then validated the model with an independent group of 71 patients from Netherlands.

Kamath et al.[12] studied the ability of the MELD to predict patient survival in a variety of different settings: (1) hospitalized patients with hepatic decompensation, (2) ambulatory noncholestatic cirrhotic patients, (3) patients with primary biliary cirrhosis, and (4) historical patients from the 1980s with cirrhosis. The 3-month concordance(c)-statistic showed that the model performed well in predicting 3-month mortality for all four groups (Table 3). c-Statistics are commonly used to evaluate diagnostic tests, and are equivalent to the area under the receiver-operating-characteristic curve. A c-statistic between 0.8 and 0.9 indicates excellent diagnostic accuracy and greater than 0.7 is generally considered a useful test. Furthermore, the authors looked at the contribution of disease etiology and complications of portal hypertension (ascites, variceal bleeding, spontaneous bacterial peritonitis, and encephalopathy). Neither disease etiology (Table 4) nor any of the complications of portal hypertension substantially increased the c-statistics.

The use of continuous scales (such as the MELD) to allocate organ grafts was further supported by Freeman et al.[13] The authors studied the use of a continuous medical severity score based on the CTP score in status 2B patients, while implementing more strict criteria to status 2A patients. Waiting time was used only as a tiebreaker. Comparison of 6-month periods before (67 patients) and after (75 patients) these changes showed a significant drop in the waitlist mortality (62% vs. 94%, P = 0.005). There were significant increases in the transplants done for status 2B patients, and a significant decrease for those done for status 2A patients. Patient and graft survival following transplant was similar for both groups of patients. This trial affirmed that waitlist mortality could be decreased with stricter prioritization based on urgency rather than waitlist time.

MELD in the Pretransplantation Setting

Several studies have assessed the role of MELD in predicting survival in patients awaiting liver transplantation. Brown et al.[14] compared the MELD score to CTP score in their ability to predict pretransplantation disease severity in a group of 42 UNOS status 2A patients at a single medical center. Status 2A patients were studied because they were more likely to reflect a group of more higher urgency for transplantation. Multivariate analyses showed that the MELD score was better able to predict preorthotopic liver transplantation (OLT) disease severity than CTP in mechanical ventilation (MELD: odds ratio (OR), 1.13; 95% confidence interval (CI), 1.02-1.24; P=0.016 vs. CTP: OR, 1.12; 95% CI, 0.67-1.87; P=0.66) and dialysis (MELD: OR, 1.23; 95% CI, 1.06-1.43; P=0.006 vs. CTP: OR, 0.71; 95% CI, 0.36-1.39; P=0.31). Further analysis showed that each MELD score point increased the likelihood of intubation by 13% and need for dialysis by 23%.

MELD and 3-month waiting list mortality were studied in a large prospective study on 3437 patients on the Organ Procurement and Transplantation Network (OPTN) waiting list from November 1999 and December 2001.[15] Overall mortality during the 3 months was 12% (n=412). Higher MELD scores at the time of listing were directly proportional to higher mortality. Patients with MELD scores <9 had a 1.9% mortality and patients with scores >40 had a mortality of 71.3%. CTP scores showed an upward trend also with mortality and higher scores. However, c-statistic using 3-month mortality as the endpoint was significantly higher in the area under the ROC curve for the MELD score vs. CTP score. (MELD: 0.83; 95% CI, 0.81-0.84 vs. CTP: 0.76; 95% CI, 0.74-0.79; P<0.001). It is important to note that this analysis overestimates CTP's predictive value because CTP score was treated as a continuous variable to facilitate comparison instead of as a categorical variable with waiting time as the final determinant (which is how it was used clinically). However, this study demonstrated the ability of MELD to accurately predict 3-month mortality in patients with chronic end-stage liver disease.

A recent European study looked at MELD and longer-term pretransplantation mortality in 129 cirrhotic patients.[16] Overall 1-year mortality for this cohort was 24% (31 patients). MELD scores were found to be significantly different between the deceased group and the surviving group at 6 months (11 (range 6-19) vs. 7 (range 5-17); P<0.0007) and at 1 year (9; (range 5-17) vs. 6; (range 5-17); P=0.003). The MELD score was found to perform about equally as well as CTP at both of these time points. This study further demonstrates the international applicability of a uniform scoring system, as it reproduced results found in US-based studies.[14,15] While the previous studies found that the MELD was most predictive at 3 months, this study suggested that the MELD is useful up to 1-year pretransplantion.

While most of these studies looked at MELD scores calculated once at a designated point, there has been interest in looking at the change in MELD scores over time. Merion et al.[17] looked at 760 waitlisted patients and compared mortality prediction with single MELD score calculation vs. the change in MELD score from serial calculations. A total of 760 patients were included in the study. MELD scores were calculated at baseline upon placement on the waitlist, and also whenever one or more laboratory components changed. The authors found that serial MELD scores predicted waitlist mortality significantly better than baseline MELD scores or medical urgency status. Furthermore, each point of the MELD score was associated with a 22% increased risk for waitlist death (P<0.001). Magnitude and direction of change in MELD score during the previous 30 d were found to be significant independent mortality predictors. A positive change of MELD score of more than five points during a 30-d period showed and increased risk for death (RR 3.16; P<0.0001). The authors had suggested using the change in MELD as a tiebreaker for those with identical MELD scores. While these results have strongly supported that, there is reservation. Changes in MELD score may not necessarily reflect worsening liver disease.[18] Reversible factors such as dehydration and infection may cause a transient rise in the lab values used to calculate the MELD. Without prospective data, the final determination on the value of change in MELD is yet to be made.

To be continued in next reply...

[hanklive39]

MELD in the Post-Transplantation Setting
OLT

An ideal liver allocation model would not only allocate livers to the most urgent patients, but also determine which candidates who have the best chance at survival post-transplantation. Initial studies on post-transplant outcome and MELD were limited by small patient sample size and high patient survival.[14] However, two recent studies have further analyzed pretransplantation MELD scores as predictors of postoperative survival. Onaca et al.[19] studied 669 adult patients at a single center retrospectively and looked at the association of MELD scores calculated immediately prior to OLT and the association with mortality at 1 and 2 years. The authors found progressively lower survival for higher strata of MELD scores at 1 year (MELD < 15, survival=89% vs. MELD >25, survival=76%; P=0.002). At 2 years of follow-up, survival was also significantly worse for higher MELD strata (MELD <15, survival=86% vs. MELD >25, survival=75%; P=0.031). Upon stratification by liver disease etiology, patients with cholestatic diseases had a survival advantage over patients with noncholestatic diseases. In the group with MELD scores between 15 and 24, survival was higher with cholestatic disease by 6.5% at 30 d, 13.2% at 1 year, and 17.5% at 2 years.

The association between higher MELD scores and worse patient survival was also found in a study of 403 patients who had undergone OLT at UCLA Medical Center.[20] A significant difference (P=0.0006) in higher MELD strata on 1-year post-transplantation survival analysis. Post-transplantation survival at 1 year for MELD strata of <10, 11-18, 19-24, 25-35, and >36 was 90%, 89%, 90%, 79%, and 69%, respectively. Additionally, there were significant survival differences between higher and lower scored groups upon dichotamization of the data at scores of 24 (P<0.0001). Significant differences in survival were also seen at dichotamization at scores of 30 and 36 (P<0.001 for both scores). The results of both of these studies suggest that preoperative MELD scores are strong predictors of post-transplantation outcome to as far as 2 years post-transplantation. However, as apparent from both studies, that the largest decline in survival seen within the first 6 months in the strata with the highest MELD scores. The survival curves mostly plateau off from 6 months onward. This observance suggests that the predictive power of MELD is strongest within the 6-month postoperative period. In concurrence, the immediate postoperative period is when complications can affect post-OLT survival the most such as acute graft rejection, sepsis, and multiple organ failure.
Living Donor Liver Transplants (LDLT)

Under the duress of organ shortage, LDLT is an important alternative to cadaveric transplantation. The use of the MELD score in this alternative setting has also been investigated. There has been recent data to suggest that the MELD cannot predict survival as accurately in patients receiving LDLT. A study of 62 patients looked at postoperative survival and preoperative MELD scores in adult patients who underwent right hepatic lobe transplants from adult donors.[21] At 1-year follow-up, 59 (95%) patients survived, and 52 (84%) grafts survived. There were no significant differences in the mean MELD scores between the survivors and the deceased, and the overall mean score was 15.2 (range 6-40). Patients who had higher MELD scores (>/=18) did have significantly longer hospital stays post-transplant than patients with lower scores (<18) (35.2 d vs. 19.8 d, P=0.01). Similarly, a preliminary report on 29 patients who underwent LDLT had a significant inverse correlation between MELD score and length of hospital stay (r2=0.21, P=0.02) and operative blood use (r2=0.17l, P=0.04).[22] The poor correlation between MELD score and survival may be explained that patients who receive LDLT are generally less urgent than those who undergo cadaveric transplantation, as evident by their lower MELD scores. Of note, the mean pretransplant MELD score of cadaveric recipients is 23.5 as compared to 14.8 for those receiving LDLT.[23] Based on the cadaveric transplantation survival data, the survival greatly worsens at preoperative scores higher than those of an average LDLT patient. Perhaps, MELD and post-LDLT survival can be better elucidated once data from larger studies are available. Nevertheless, careful selection of recipients for LDLT is critical. Assessments of functional reserve of the recipient, of donor risk of death, and of likelihood of survival while awaiting a cadaveric organ are among the things that need to be carefully examined in the candidates.

MELD and Retransplantation

Retransplantation generally confers worse survival than those being transplanted for the first time. A retrospective analysis assessed the impact of MELD predicting post-transplantation survival in patients undergoing retransplantation in a cohort of 76 patients who underwent 80 re-OLT.[24] Median time to re-OLT from the first transplantation was 2.65 years (range, 0.25-10.7 years), and the overall survival was 68.9% at 1 year and 61.7% at 5 years. The c-statistic was higher using CTP (0.817) than MELD (0.679). With dichotamization of the group by MELD score, patients with a score of 25 and lower had 1- and 5-year survival rates of 88.9% and 79.0%, respectively, and those with a score above 25 have rates of 52.6 and 47.4% (HR 4.7, P= 0.046). This survival difference was also seen in CTP score. Patients with CTP<10 had 1 and 5 year survival of 100% vs. 50.0% and 39.8%, respectively, for those with scores 10 and greater (likelihood ratio 14.3, P=0.0002). This preliminary data suggest the potential use of the MELD in risk stratification in graft dysfunction and in predicting long-term outcome after retransplantation, but it is apparently not superior to CTP in this setting.

To be continued in next reply...

[hanklive39]

Criticisms of the MELD Score

The weaknesses of the MELD must be viewed under the pretense that the score was not designed for its current present day use of liver allocation. It was designed primarily for the prediction of survival in patients undergoing TIPS procedures. Over time, the mathematical equation has been adapted for use for prediction of pretransplantation and post-transplantation mortality in end-stage liver disease, and additionally in a variety of other noncirrhotic liver disease. Theoretically, the advantage of a system that relies completely on objective data would be more fair allocation. But there has been criticism that the score does not serve the best interest of all patients, particularly those with hepatocellular carcinoma, familial amyloidosis, hepatopulmonary syndrome, and polycystic liver disease. In addition, there may be an unfair gain in other scenarios such as in renal failure. Such clinical situations are among others that will need special consideration in the fine-tuning of the current system to move towards fairness of allocation. Liver allocation in the future will need to address these issues.
Renal function

The use of creatinine in the MELD model has roused speculation of an unfair advantage in patients with a fluctuating creatinine level. While serum bilirubin and prothrombin time have less day-to-day fluctuation, serum creatinine levels can fluctuate based on several reasons such as dehydration, hemorrhage, or use of nephrotoxic agents. Some causes are reversible and others are not. As an example of the former, creatinine can fluctuate transiently with a changing volume status. Use of diuretics in ascites management decreases effective intravascular volume and subsequently decreases glomerular filtration rate leading to rises in serum creatinine. More significant rises in creatinine may be secondary to intrinsic renal disease or onset of hepatorenal syndrome. UNOS has required serial testing of MELD scores every 7 d for patients above the score of 25, and every 30 d for those between 19 and 24. While this is laborsome, the testing can help streamline the scores for those with transient fluctuations. The clinical importance of pretransplant creatinine in post-transplant survival was demonstrated by a recent study that showed that patients with moderate and severe renal failure had significantly lower 2-year survival when compared with patients with normal pretransplant renal function (64% and 55% vs. 76%; P<0.05).[25] Of note, subjects with dual liver and kidney transplants were removed from the study population in this paper. The data suggests that post-transplant mortality is highly influenced by pretransplant renal function and that allocating liver organs to patients with higher MELD scores secondary to renal failure may lead to decreased post-OLT survival. Undeniably, creatinine is a strong factor in the MELD model, and investigation of any decline in renal function is warranted when observed. Such investigation may include but not limited to a history and physical and evaluation of volume status, a history of exposures to nephrotoxic agents (i.e. contrast dyes, nonsteroidal anti-inflammatory drugs, antibiotics), urinanalysis with microscopic evaluation, urine electrolytes and eosinophils, evaluation for obstruction, and possibly renal biopsy.
Hepatocellular Carcinoma (HCC)

Patients with hepatocellular carcinoma may initially have preserved synthetic liver function that will not be prioritized well by MELD score calculation, thus underestimating their urgency. Prior to implementation of the MELD score as the allocation method, there have been some attempts to mathematically calculate risk of HCC progression to estimate how this factor would contribute to the new allocation schema.[26] Previously HCC-adjusted MELD scheme stratified patients with T1 HCC (single lesion</=1.9 cm) with a MELD score equivalent to a 15% (most recently adjusted to 8%) 3-month mortality, and T2 HCC (one nodule 2-5 cm, or two to three nodules all </=3 cm) with a score equivalent to a 30% (now adjusted to 15%) 3-month mortality. Additional points equivalent to a 10% increase in pretransplant mortality are also given every 3 months until the patient is transplanted or no longer suitable for transplant. T3 HCC (one nodule >5 cm or two to three nodules at least one >3 cm) and T4 HCC (four or more nodules of any size or gross vascular invasion) are not eligible for listing.[3] There is criticism that this schema was made without much prior data on the pattern and rate of dropouts, and that liver cancer patients may have been unfairly given an advantage. Efforts to verify the fairness of the scheme suggest that further refinement is still needed.[27,28] The implementation of the exceptions placed on HCC has resulted in more aggressive screening for HCC prior to transplantation. Careful prospective assessment of post-transplants survival for transplantation done for HCC under the MELD system is needed to support the continued use of the HCC exception rules in the MELD system for resource allocation.
MELD Score and Quality of Life

While the MELD score is thought to be advantageous because of its sheer reliance on objectivity, there has also been criticism about its lack of attention to factors such as quality of life, encephalopathy, and ascites. A recent study of 66 patients looked at the association of MELD score and clinical and subclinical hepatic encephalopathy as diagnosed by neuropsychometric tests and EEG.[29] Unsurprisingly, the authors found poor correlation between the MELD score and presence of encephalopathy or ascites, and stressed that these clinical findings are important debilitating factors in end-stage liver disease. As the MELD score cannot completely account for certain factors that can help assess levels of disease severity, there is some continued debate about how to best incorporate this clinical data into the allocation system.

To be continued in next reply...

[hanklive39]

Discussion

With continued shortage of available liver organs, the transplantation community maintains high interest in the efficient use of scarce resources. The MELD is a useful system that uses objective data, and it has been shown to accurately predict short-term survival in the pretransplantation and post-transplantation setting. The advantages of this system includes its ease of use at the bedside, the strong clinical data that shows predictive ability of the MELD score in 3-month mortality pretransplantation in a variety of clinical settings, and the presence of data to support its use to predict long-term post-transplantation survival. The disadvantages of this system include the lack of quality-of-life issues and no validation in patients with hepatocellular carcinoma.

The installment of MELD in liver allocation of the last year has affected resource allocation. The question now is that are we doing better in overall outcome than in the pre-MELD era? Preliminary data has shown decreased waiting list mortality during the first 6 months under the new system.[23] However, with higher medical urgency patients being offered transplantation, we may begin to see an increase in post-transplantation mortality.

In the evolving system, with cadaveric transplants going to the sickest, the patients with maintained synthetic function are those who may fall through the cracks. The use of LDLT and the possibility of suboptimal grafts will likely become shunts for those with lower MELD scores.

The liver transplantation community needs to re-examine the data now that we have close to 2 years of data under the MELD system. Prospective studies on MELD's impact on survival are needed to further elucidate its role in pretransplant and post-transplant settings.


Tables for:
Liver Transplantation in the Era of Model for End-Stage Liver Disease

[Liver Int 24(1):1-8, 2004. © 2004 Blackwell Publishing]

Table 1. UNOS Status

    Status   Variables
    1   Fulminant liver failure
    Primary graft nonfunction (<7 d)
    Hepatic artery thrombosis (<7 d)
    Acute Wilson's disease
    2A   In ICU with Pugh>10:
      (1) Active GI bleed (Pugh>10)
      (2) Stage 3 or 4 coma
      (3) Hepatorenal syndrome
      (4) Refractory ascites
    2B   Pugh score 7-10 with:
      (1) GI bleed
      (2) Hepatorenal syndrome
      (3) Spontaneous bacterial peritonitis or refractory ascites
      (4) HCC
    3   Continuous medical care OR hospitalized <5 d

    UNOS, United Network of Organ Sharing; HCC, hepatocellular carcionoma.


Table 2. Minimal Listing Criteria for Patients with Cirrhosis[7]

    Child-Pugh score >/=7
        or any Child-Pugh score and history of variceal bleed
        or any Child-Pugh score and history of spontaneous bacterial peritonitis
    HCC without lymph node and vascular involvement, and spread to other organs.

    HCC, hepatocellular carcinoma.


Table 3. MELD validation in a variety of clinical settings[12]

        Hospitalized   Ambulatory Noncholestatic   Ambulatory PBC   Historical
    N   282   491   326   1179
    3-month mortality   59   34   5   220
    3-month mortality c-statistic*   0.87 (0.82-0.92)   0.80 (0.69-0.90)   0.87 (0.71-1.00)   0.78 (0.74-0.81)

    *Values reported are the concordance statistic (95% CI). PBC, primary biliary cirrhosis.


Table 4. MELD Model with and without Disease Etiology[12]

        Hospitalized (n=282)   Noncholestatic (n=491)   Ambulatory PBC (n=326)   Historical (n=1179)
    MELD   0.086
    (0.81-0.92)   0.082
    (0.73-0.91)   0.87
    (0.71-1.00)   0.78
    (0.74-0.81)
    MELD
    +etiology   0.87
    (0.82-0.92)   0.80
    (0.69-0.90)   0.87
    (0.71-1.00)   0.78
    (0.74-0.81)

    Values reported are the concordance statistic (and 95% confidence intervals). PBC, primary biliary cirrhosis; MELD, model for end-stage liver disease.

tobe continued in next reply...

[hanklive39]

References for:
Liver Transplantation in the Era of Model for End-Stage Liver Disease

[Liver Int 24(1):1-8, 2004. © 2004 Blackwell Publishing]

   1. Martin J A, Smith B L, Mathews T J, Ventura S J. Births and deaths for 1998. National vital statistics report, Vol. 47, No. 25. National Center for Health Statistics, Hyattsville, MD, 1998.
   2. United Network for Organ Sharing. Data resources: available at http://www.unos.org. Accessed July 13, 2003.
   3. Child C G III, Turcotte J G. Surgery and portal hypertension. In: Child C G III, ed. The Liver and Portal Hypertension. Philadelphia: Saunders, 1964; 49-64.
   4. Pugh R N H, Murray-Lyon I M, Dawson J L, et al. Transection of the oesphagus for bleeding oesophageal varices. Br J Surg 1973; 60: 646-9.
   5. Schlicting P, Christensen E, Andersen P K, et al. Prognostic factors in cirrhosis identified by Cox's regression model. Hepatology 1983; 3: 889-95.
   6. Bircher J. Assessment of prognosis in advanced liver disease: to score or to measure, that's the question. Hepatology 1986; 6: 1036-7.
   7. Lucey M R R, Brown K A, Everson G T, et al. Minimal criteria for placement of adults on the liver transplant waiting list: a report of a national conference organized by the American Society of Transplant Physicians and the American Association for the Study of Liver Diseases. Liver Transpl Surg 1997; 3: 628-37.
   8. Institute of Medicine. Analysis of waiting times. In: Committee on Organ Procurement and Transplantation Policy, ed. Organ Procurement and Transplantation: Assessing Current Policies and the Potential Impact of the DHHS Final Rule. Washington, DC: National Academy Press, 1999; 57-78.
   9. Freeman R, Edwards E. Liver transplant waiting time does not correlate with waiting list mortality: implications for liver allocation policy. Liver Transpl 2000; 6: 543-52.
  10. Malinchoc M, Kamath P, Gordon F, et al. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31: 864-71.
  11. United Network for Organ Sharing. Policy 3.6. Available at: http://www.unos.org. Accessed July 18, 2003.
  12. Kamath P, Wiesner R, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001; 33: 464-75.
  13. Freeman R, Rohrer R, Katz E, et al. Preliminary results of a liver allocation plan using a continuous medical severity score that de-emphasizes waiting time. Liver Transpl 2001; 7: 173-8.
  14. Brown R, Kumar K, Russo M, et al. Model for end-stage liver disease and Child-Turcotte-Pugh Scores as predictors of pretransplantation disease severity, posttransplantation outcome, and resource utilization in the United Network for Organ Sharing status 2A patients. Liver Transpl 2002; 8: 278-84.
  15. Wiesner R, Edwards E, Freeman R, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003; 124: 91-6.
  16. Botta F, Giannini E, Romangnoli P, et al. MELD scoring system is useful for predicting prognosis in patients with liver cirrhosis and is correlated with residual liver function: a European study. Gut 2003; 52: 132-9.
  17. Merion R, Wolfe R, Dykstra D, et al. Longitudinal assessment of mortality risk among candidates for liver transplantation. Liver Transpl 2003; 9: 12-8.
  18. Kamath P, Kim R. Is the change in MELD score a better indicator of mortality than baseline MELD score? Liver Transpl 2003; 9: 19-21.
  19. Onaca N, Levy M, Edmund S, et al. A correlation between the pretransplantation MELD score and mortality in the first two years after liver transplantation. Liver Transpl 2003; 9: 117-23.
  20. Saab S, Wang V, Ibrahim A, et al. MELD score predicts 1-year patient survival post-orthotopic liver transplantation. Liver Transpl 2003; 9: 473-76.
  21. Hayashi P H, Forman L, Steinberg T, et al. Model for end-stage liver disease score does not predict patient or graft survival in living donor liver transplant recipients. Liver Transpl 2003; 9: 737-40.
  22. Ijperen M V, Brandhagen D J. MELD score predicts outcome after living donor liver transplantation. Hepatology 2002; A2030..
  23. Freeman R B, Harper A, Edwards E, et al. Results of the first six months of the MELD/PELD system: fewer waitlist deaths and more cadaveric transplants [abstract]. Am J Transplant 2003; 3 (Suppl. 5): S284.
  24. Yao F Y, Bass N M, Hirose R, et al. Prediction of survival after liver re-transplantation based on pre-operative Child-Turcotte-Pugh (STP) score and the model for end stage liver disease (MELD) score. Hepatology 2002; A247.
  25. Nair S, Verma S, Thuluvath P J. Pretransplant renal function predicts survival in pateints undergoing orthotopic liver transplantation. Hepatology 2002; 35: 1179-85.
  26. Cheng S, Freeman R, Wong J. Predicting the probability of progression-free survival in patients with small hepatocellular carcinoma. Liver Transpl 2002; 8: 323-8.
  27. Yao F Y, Bass N M, Nikolai B, et al. Liver transplantation for hepatocellular carcinoma: analysis of survival according to the intention-to-treat principle, and dropout from the waiting list. Liver Transpl 2002; 8: 873-83.
  28. Yao F Y, Bass N M, Nikolai B, et al. A follow-up analysis of the pattern and predictors of dropout from the waiting list for liver transplantation in patients with hepatocellular carcinoma: implications for the current organ allocation policy. Liver Transpl 2003; 9: 684-92.
  29. Yoo H Y, Edwin D, Thuluvath P J. Relationship of the model for end-stage liver disease (MELD) scale to hepatic encephalopathy, neuropsychometric testing, and ascites. Am J Gastroenterol 2003; 98: 1395-9.

Appendix: The Current MELD Score Calculation

MELD score=0.957xloge(creatinine mg/dL)+0. 378xloge(bilirubin mg/dl)+1.120xloge(INR)+0.643.

Minimum values for variables are set to 1.0 for calculation purposes. The maximum serum creatinine considered within the MELD score equation will be 4.0 mg/dl (i.e. for patients with a serum creatinine of greater than 4.0 mg/dl, the serum creatinine level will be set to 4.0 mg/dl). For patients on dialysis with two or more treatments within the prior week, the serum creatinine level will automatically be set to 4.0 mg/dl.

Respectfully,
Hanklive Lives for Today and Tomorrow!!!