E v o l v i n g Tren d s i n L i v e r Tra n s p l a n t a t i o n Listing and Liver Donor Allocation Russell H. Wiesner,
MD
KEYWORDS Liver transplantation MELD Liver donor allocation Chronic liver disease KEY POINTS The MELD-based allocation policy is excellent at prioritizing patients with chronic liver disease on the waiting list based on survival. Exception criteria are needed because of such conditions as hepatocellular cancer, with tumor progression being used as a surrogate for patient survival. MELD can be tweaked by adding serum sodium and other variables; but overall the impact would be minimal and reprograming is expensive and cumbersome.
INTRODUCTION
The success of liver transplantation in the past three decades as a life-saving procedure for patients with end-stage liver disease has led to the ever-increasing disparity between the demands for liver transplantation and the supply of donor liver organs. This demand has been fueled by an increasing prevalence of cirrhosis and hepatocellular cancer (HCC) related to the hepatitis C epidemic, and to the increase in obesityrelated liver disease.1 Today, nearly 1 in 10 patients waiting for a donor liver organ die on the waiting list.2 Therefore, donor allocation and distribution remains a challenge and a moral issue as to how these organs can be equitably distributed. This article reviews the evolution of the liver allocation policy and discusses in detail the challenges we face today. HISTORY
Donor liver allocation dates back to the Transplantation Act of 1983 when there were only a few liver transplant centers in the United States. However, with the increasing success of liver transplantation and expanded indications, the number of patients seeking this life-saving procedure continued to grow. At the same time, the number
Disclosure: None. Mayo College of Medicine, Rochester, MN 55905, USA E-mail address: [email protected] Clin Liver Dis 18 (2014) 519–527 http://dx.doi.org/10.1016/j.cld.2014.05.014 1089-3261/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
liver.theclinics.com
520
Wiesner
of institutions offering liver transplantation also increased. By the mid-1990s, livers were allocated based on blood type; time on the liver transplant waiting list; and whether a patient was in the intensive care unit (ICU), hospital, or an outpatient. There was no criterion to define which patients should be in the ICU or which patients should be hospitalized. Indeed, many centers admitted patients to the ICU solely to facilitate their placement at the top of the liver waiting list. In retrospect, it was noted in the model for end-stage liver disease (MELD) era that the number of patients proceeding to liver transplantation from the ICU dropped dramatically from 24% in the pre-MELD era to 13% in the post-MELD era.3 There was also noted to be a strong relationship between the number of centers in an organ procurement unit, and the number of patients undergoing liver transplant directly from the ICU. An attempt to rectify this problem was made in 1998 by implementation of a new system incorporating the Child-Turcotte-Pugh score (CTP) as an index for liver disease severity and prognosis. This scoring system was originally used to predict the outcome of portal-cava shunt surgery in patients with cirrhosis.4 Adoption of the CTP-based allocation system was intended primarily to reflect waiting list mortality and severity of liver disease. However, the use of the CTP score for the prioritization of liver transplant candidates had several major drawbacks. First, ascites and encephalopathy were subjectively assessed and were influenced by such therapies as diuretics, albumin administration, and lactulose therapy. In addition, a score of three points was allocated for any serum bilirubin value higher than 3.0 mg/dL and any serum albumin level value of 2.8 mg/dL or less. This “ceiling and floor affect” for bilirubin and albumin, respectively, meant that many patients ended up with the same overall CTP score, even though they may have had vastly different values of these two variables. Indeed, a patient with a serum bilirubin of 25 mg/dL received the same CTP score as the patient who had a serum bilirubin of 3 mg/dL. As a result, time spent on the waiting list became the major selection factor of liver candidates having identical CTP scores. Finally, there was no parameter in the CTP score that reflected renal function, a key prognostic marker in patients with end-stage cirrhosis. Two studies found time spent on the waiting list was not associated with risk of death on the waiting list.5,6 In the end, the revised allocation policy based on the CTP score ultimately proved unworkable and unfair to patients with the most severe liver disease. Because of considerable disagreement among transplant centers on how livers should be allocated, the United Network for Organ Sharing (UNOS) and the Department of Health and Human Services intervened in 1999 and challenged the transplant community with “The Final Rule.”7 Among the conditions of the Final Rule were that allocation policies should be based on objective and measurable medical criteria of patients or categories of patients who are medically suitable candidates for liver transplantation. In addition, it noted that patients should be rank-ordered according to severity of disease and predicted mortality on the liver waiting list. The Final Rule stipulated that waiting time should be deemphasized, that allocation should be designed to achieve equitable allocation of organs among patients, and that organs should be distributed over as broad a geographic area as feasible in order of decreasing medical urgency. Finally, the Final Rule noted that neither place of residence nor place of listing should be a major determinant of access to a liver transplant. This challenge was ultimately met by the adoption of the MELD score by UNOS in February 2002. MELD
The MELD score was first used and published in 2000 by Malinchoc and colleagues8 to predict survival in patients undergoing elective transjugular intrahepatic portal-cava
Evolving Trends in Liver Transplantation
systemic shunts. In 2001, the Mayo Clinic group modified the score to predict mortality in a wide range of liver disease etiologies and severities.9 In the final model, cause of liver disease was eliminated from the model and three biochemical variables were used: (1) serum bilirubin, (2) international normalized ratio (INR), and (3) serum creatinine level. Retrospective studies demonstrated that the MELD score had a high degree of discriminative ability in prediction of 3-month survival in patients with cirrhosis, regardless of the cause of liver disease, and was independent of complications of liver disease, such as ascites, encephalopathy, variceal bleeding, and spontaneous bacterial peritonitis.9 The final assessment of the MELD score was made using 3437 patients on the UNOS liver transplant waiting list. In this study the MELD score was found to be superior to the CTP score for predicting 3-month survival.10 The C statistic for predicting 3-month survival on the waiting list was 0.83. This study led to the adoption of the MELD score for liver donor allocation in the United States in February 2002. Since that time, the MELD score has been validated around the world and is presently used in many countries today to allocate donor liver organs.11–13 Indeed, there are more than 1000 papers published in the last decade on the MELD score. The MELD score has been thoroughly scrutinized and a decade later remains the standard by which to predict mortality in patients with end-stage liver disease. ADVANTAGES OF THE MELD SCORE
Advantages of the MELD score are its statistical validation and the use of objective widely available laboratory tests (serum bilirubin, serum creatinine, and INR of prothrombin). Today several online calculators are available for calculating the MELD score. It initially was thought that the MELD score might prioritize sicker patients who were more likely to die following liver transplantation. The impact of the MELD allocation policy resulted in a reduction in waiting list registration by 12%, a reduction in death on the waiting list by 5%, decreased median waiting times from 676 to 416 days, and patients transplanted within 30 days of listing increased from 23% in the pre-MELD era to 37% in the post-MELD era.14 In the MELD era although sicker patients were transplanted posttransplant survival actually slightly improved, thus dispelling concerns that many had about reducing death on the waiting list, but increasing death posttransplant.15,16 However, transplantation of sicker patients with higher MELD scores has led to an increase in the cost of liver transplantation and resource use.17,18 Finally, the MELD-based allocation system has allowed increasing transparency and allows evidence-based decision making to refine allocation and distribution policy to maximize outcomes using this scarce donor resource. DISADVANTAGES OF THE MELD SCORE
The MELD allocation policy is an urgency-based system, which means that the risk of death on the waiting list is paramount, irrespective of the degree of survival benefit after liver transplantation. In reality, the MELD era has seen a small increase in overall patient survival despite increases in MELD score at the time of transplant. However, this increase in survival may be related to the increased number of patients with HCC with low MELD scores who are given exception points. In the final analysis, patients with the highest MELD score have reaped the most benefit.19 The second potential disadvantage is that there is considerable variation in MELD score known to occur based on laboratory methodology.20–22 This is particularly true of INR and serum creatinine measurements. Differences in measurement can mean up to a seven MELD point variation depending on how these biochemical
521
522
Wiesner
parameters are determined. Specifically, variability has been seen in the measurement of INR. However, major issues regarding serum creatinine have been raised, which can be influenced by extrarenal factors, such as total muscle mass, ethnicity, and gender. It has been noted that women tend to have lower median MELD scores compared with men at the time of transplantation. In addition, females are more likely to die on the waiting list in the post-MELD era compared with the pre-MELD era.23,24 Women are not only disadvantaged because of smaller size, but are also disadvantaged in that their serum creatinine underestimates the degree of renal dysfunction. A third disadvantage of the MELD system is that it has been shown that the quality of life does not correlate well with severity of liver disease as measured by the MELD score.25 Challenges that remain with regard to MELD allocation include that 15% of patients are prioritized incorrectly; the MELD exception scores are nonstandardized (discussed next), and HCC patients are overprioritized compared with non HCC patients (discussed next). In addition, there remains marked geographic disparity with regard to mean MELD scores at the time of transplant and regarding overall waiting time.26 Finally, because serum creatinine is one of the three factors making up the MELD score, there has been a four-fold increase in the number of liver-kidney transplants since the start of the MELD allocation system.27 Ongoing attempts are being made to further define which liver transplant patients need simultaneous liver-kidney transplant and which do not. MELD EXCEPTIONS
One of the major requirements of fair organ allocation is equitable access and distribution of organs to candidates regardless of their underlying liver disease, race, or gender. However, there are several patients with diseases who are poorly served by using the pure MELD allocation system; patients with HCC constitutes such a group of patients. In the absence of MELD exception points, patients with HCC could have an increased risk of waiting list drop-out, generally because of tumor progression even though typically they have a relatively low risk of waiting list mortality caused by liver disease severity. Other conditions underserved by the MELD allocation policy are patients with cirrhosis and hepatopulmonary syndrome or portal pulmonary hypertension, familial amyloidosis, and primary oxaluria. In HCC patients, MELD exception points were originally granted to patients whose tumor burden fell within the so-called Milan Criteria (stage 1–2) disease. Accordingly, it was determined that the risk of tumor progression, rather than 3-month death on the waiting list, should be used to adjust the MELD score for these patients. The adoption of a risk-based allocation system using MELD exception points has lowered waiting list mortality for HCC patients in the post-MELD era.28,29 Originally in 2002, stage 1 tumors (2 cm) were given a 15% risk of dying within 3 months, which equated to a MELD score of 24. Patients with stage 2 tumors (single tumor >2 cm but
MD
KEYWORDS Liver transplantation MELD Liver donor allocation Chronic liver disease KEY POINTS The MELD-based allocation policy is excellent at prioritizing patients with chronic liver disease on the waiting list based on survival. Exception criteria are needed because of such conditions as hepatocellular cancer, with tumor progression being used as a surrogate for patient survival. MELD can be tweaked by adding serum sodium and other variables; but overall the impact would be minimal and reprograming is expensive and cumbersome.
INTRODUCTION
The success of liver transplantation in the past three decades as a life-saving procedure for patients with end-stage liver disease has led to the ever-increasing disparity between the demands for liver transplantation and the supply of donor liver organs. This demand has been fueled by an increasing prevalence of cirrhosis and hepatocellular cancer (HCC) related to the hepatitis C epidemic, and to the increase in obesityrelated liver disease.1 Today, nearly 1 in 10 patients waiting for a donor liver organ die on the waiting list.2 Therefore, donor allocation and distribution remains a challenge and a moral issue as to how these organs can be equitably distributed. This article reviews the evolution of the liver allocation policy and discusses in detail the challenges we face today. HISTORY
Donor liver allocation dates back to the Transplantation Act of 1983 when there were only a few liver transplant centers in the United States. However, with the increasing success of liver transplantation and expanded indications, the number of patients seeking this life-saving procedure continued to grow. At the same time, the number
Disclosure: None. Mayo College of Medicine, Rochester, MN 55905, USA E-mail address: [email protected] Clin Liver Dis 18 (2014) 519–527 http://dx.doi.org/10.1016/j.cld.2014.05.014 1089-3261/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
liver.theclinics.com
520
Wiesner
of institutions offering liver transplantation also increased. By the mid-1990s, livers were allocated based on blood type; time on the liver transplant waiting list; and whether a patient was in the intensive care unit (ICU), hospital, or an outpatient. There was no criterion to define which patients should be in the ICU or which patients should be hospitalized. Indeed, many centers admitted patients to the ICU solely to facilitate their placement at the top of the liver waiting list. In retrospect, it was noted in the model for end-stage liver disease (MELD) era that the number of patients proceeding to liver transplantation from the ICU dropped dramatically from 24% in the pre-MELD era to 13% in the post-MELD era.3 There was also noted to be a strong relationship between the number of centers in an organ procurement unit, and the number of patients undergoing liver transplant directly from the ICU. An attempt to rectify this problem was made in 1998 by implementation of a new system incorporating the Child-Turcotte-Pugh score (CTP) as an index for liver disease severity and prognosis. This scoring system was originally used to predict the outcome of portal-cava shunt surgery in patients with cirrhosis.4 Adoption of the CTP-based allocation system was intended primarily to reflect waiting list mortality and severity of liver disease. However, the use of the CTP score for the prioritization of liver transplant candidates had several major drawbacks. First, ascites and encephalopathy were subjectively assessed and were influenced by such therapies as diuretics, albumin administration, and lactulose therapy. In addition, a score of three points was allocated for any serum bilirubin value higher than 3.0 mg/dL and any serum albumin level value of 2.8 mg/dL or less. This “ceiling and floor affect” for bilirubin and albumin, respectively, meant that many patients ended up with the same overall CTP score, even though they may have had vastly different values of these two variables. Indeed, a patient with a serum bilirubin of 25 mg/dL received the same CTP score as the patient who had a serum bilirubin of 3 mg/dL. As a result, time spent on the waiting list became the major selection factor of liver candidates having identical CTP scores. Finally, there was no parameter in the CTP score that reflected renal function, a key prognostic marker in patients with end-stage cirrhosis. Two studies found time spent on the waiting list was not associated with risk of death on the waiting list.5,6 In the end, the revised allocation policy based on the CTP score ultimately proved unworkable and unfair to patients with the most severe liver disease. Because of considerable disagreement among transplant centers on how livers should be allocated, the United Network for Organ Sharing (UNOS) and the Department of Health and Human Services intervened in 1999 and challenged the transplant community with “The Final Rule.”7 Among the conditions of the Final Rule were that allocation policies should be based on objective and measurable medical criteria of patients or categories of patients who are medically suitable candidates for liver transplantation. In addition, it noted that patients should be rank-ordered according to severity of disease and predicted mortality on the liver waiting list. The Final Rule stipulated that waiting time should be deemphasized, that allocation should be designed to achieve equitable allocation of organs among patients, and that organs should be distributed over as broad a geographic area as feasible in order of decreasing medical urgency. Finally, the Final Rule noted that neither place of residence nor place of listing should be a major determinant of access to a liver transplant. This challenge was ultimately met by the adoption of the MELD score by UNOS in February 2002. MELD
The MELD score was first used and published in 2000 by Malinchoc and colleagues8 to predict survival in patients undergoing elective transjugular intrahepatic portal-cava
Evolving Trends in Liver Transplantation
systemic shunts. In 2001, the Mayo Clinic group modified the score to predict mortality in a wide range of liver disease etiologies and severities.9 In the final model, cause of liver disease was eliminated from the model and three biochemical variables were used: (1) serum bilirubin, (2) international normalized ratio (INR), and (3) serum creatinine level. Retrospective studies demonstrated that the MELD score had a high degree of discriminative ability in prediction of 3-month survival in patients with cirrhosis, regardless of the cause of liver disease, and was independent of complications of liver disease, such as ascites, encephalopathy, variceal bleeding, and spontaneous bacterial peritonitis.9 The final assessment of the MELD score was made using 3437 patients on the UNOS liver transplant waiting list. In this study the MELD score was found to be superior to the CTP score for predicting 3-month survival.10 The C statistic for predicting 3-month survival on the waiting list was 0.83. This study led to the adoption of the MELD score for liver donor allocation in the United States in February 2002. Since that time, the MELD score has been validated around the world and is presently used in many countries today to allocate donor liver organs.11–13 Indeed, there are more than 1000 papers published in the last decade on the MELD score. The MELD score has been thoroughly scrutinized and a decade later remains the standard by which to predict mortality in patients with end-stage liver disease. ADVANTAGES OF THE MELD SCORE
Advantages of the MELD score are its statistical validation and the use of objective widely available laboratory tests (serum bilirubin, serum creatinine, and INR of prothrombin). Today several online calculators are available for calculating the MELD score. It initially was thought that the MELD score might prioritize sicker patients who were more likely to die following liver transplantation. The impact of the MELD allocation policy resulted in a reduction in waiting list registration by 12%, a reduction in death on the waiting list by 5%, decreased median waiting times from 676 to 416 days, and patients transplanted within 30 days of listing increased from 23% in the pre-MELD era to 37% in the post-MELD era.14 In the MELD era although sicker patients were transplanted posttransplant survival actually slightly improved, thus dispelling concerns that many had about reducing death on the waiting list, but increasing death posttransplant.15,16 However, transplantation of sicker patients with higher MELD scores has led to an increase in the cost of liver transplantation and resource use.17,18 Finally, the MELD-based allocation system has allowed increasing transparency and allows evidence-based decision making to refine allocation and distribution policy to maximize outcomes using this scarce donor resource. DISADVANTAGES OF THE MELD SCORE
The MELD allocation policy is an urgency-based system, which means that the risk of death on the waiting list is paramount, irrespective of the degree of survival benefit after liver transplantation. In reality, the MELD era has seen a small increase in overall patient survival despite increases in MELD score at the time of transplant. However, this increase in survival may be related to the increased number of patients with HCC with low MELD scores who are given exception points. In the final analysis, patients with the highest MELD score have reaped the most benefit.19 The second potential disadvantage is that there is considerable variation in MELD score known to occur based on laboratory methodology.20–22 This is particularly true of INR and serum creatinine measurements. Differences in measurement can mean up to a seven MELD point variation depending on how these biochemical
521
522
Wiesner
parameters are determined. Specifically, variability has been seen in the measurement of INR. However, major issues regarding serum creatinine have been raised, which can be influenced by extrarenal factors, such as total muscle mass, ethnicity, and gender. It has been noted that women tend to have lower median MELD scores compared with men at the time of transplantation. In addition, females are more likely to die on the waiting list in the post-MELD era compared with the pre-MELD era.23,24 Women are not only disadvantaged because of smaller size, but are also disadvantaged in that their serum creatinine underestimates the degree of renal dysfunction. A third disadvantage of the MELD system is that it has been shown that the quality of life does not correlate well with severity of liver disease as measured by the MELD score.25 Challenges that remain with regard to MELD allocation include that 15% of patients are prioritized incorrectly; the MELD exception scores are nonstandardized (discussed next), and HCC patients are overprioritized compared with non HCC patients (discussed next). In addition, there remains marked geographic disparity with regard to mean MELD scores at the time of transplant and regarding overall waiting time.26 Finally, because serum creatinine is one of the three factors making up the MELD score, there has been a four-fold increase in the number of liver-kidney transplants since the start of the MELD allocation system.27 Ongoing attempts are being made to further define which liver transplant patients need simultaneous liver-kidney transplant and which do not. MELD EXCEPTIONS
One of the major requirements of fair organ allocation is equitable access and distribution of organs to candidates regardless of their underlying liver disease, race, or gender. However, there are several patients with diseases who are poorly served by using the pure MELD allocation system; patients with HCC constitutes such a group of patients. In the absence of MELD exception points, patients with HCC could have an increased risk of waiting list drop-out, generally because of tumor progression even though typically they have a relatively low risk of waiting list mortality caused by liver disease severity. Other conditions underserved by the MELD allocation policy are patients with cirrhosis and hepatopulmonary syndrome or portal pulmonary hypertension, familial amyloidosis, and primary oxaluria. In HCC patients, MELD exception points were originally granted to patients whose tumor burden fell within the so-called Milan Criteria (stage 1–2) disease. Accordingly, it was determined that the risk of tumor progression, rather than 3-month death on the waiting list, should be used to adjust the MELD score for these patients. The adoption of a risk-based allocation system using MELD exception points has lowered waiting list mortality for HCC patients in the post-MELD era.28,29 Originally in 2002, stage 1 tumors (2 cm) were given a 15% risk of dying within 3 months, which equated to a MELD score of 24. Patients with stage 2 tumors (single tumor >2 cm but
