The new allocation policy of the United Network of Organ Sharing (UNOS) based on the model for end-stage liver disease (MELD) gives candidates with stage T1 or stage T2 hepatocellular carcinoma (HCC) a priority MELD score beyond their degree of hepatic decompensation. The aim of this study was to determine the impact of the new allocation policy on HCC candidates before and after the institution of MELD. The UNOS database was reviewed for all HCC candidates listed between July 1999 and July 2002. The candidates were grouped by two time periods, based on the date of implementation of new allocation policy of February 27, 2002. Pre-MELD candidates were listed for deceased donor liver transplantation (DDLT) before February 27,2002, and post-MELD candidates were listed after February 27, 2002. Candidates were compared by incidence of DDLT, time to DDLT, and dropout rate from the waiting list because of clinical deterioration or death, and survival while waiting and after DDLT. Incidence rates calculated for pre-MELD and post-MELD periods were expressed in person years. During the study, 2,074 HCC candidates were listed for DDLT in the UNOS database. The DDLT incidence rate was 0.439 transplant/person years pre-MELD and 1.454 transplant/person years post-MELD (P < 0.001). The time to DDLT was 2.28 years pre-MELD and 0.69 years post-MELD (P < 0.001). The 5-month dropout rate was 16.5% pre-MELD and 8.5% post-MELD (P < 0.001). The 5-month waiting-list survival was 90.3% pre-MELD and 95.7% post-MELD (P < 0.001). The 5-month survival after DDLT was similar for both time periods. The new allocation policy has led to an increased incidence rate of DDLT in HCC candidates. Furthermore, the 5-month dropout rate has decreased significantly. In addition, 5-month survival while waiting has increased in the post-MELD period. Thus, the new MELD-based allocation policy has benefited HCC candidates. (Liver Transpl 2004;10:36–41.)
Hepatocellular carcinoma (HCC) is the most common malignant tumor arising from the liver. It is the seventh most common cause of cancer-related deaths worldwide.1, 2 In more than 90% of patients, HCC develops in the presence of an underlying liver disease, usually hepatitis B or C. Between 1976 and 1995, the incidence of HCC in North America increased from 1.4 to 2.4 in 100,000, primarily because of an increase in the incidence of hepatitis C.1, 3–5 During the same period, HCC-related mortality increased by 41%.6 Liver transplantation (LT) offers HCC patients the best chance of cure and long-term survival. The Milan criteria has been widely used as the selection guidelines for LT.7
The increasing shortage of donor organs and the concomitant waiting-list mortality has focused attention on the need to improve the stratification of LT candidates. The model for end-stage liver disease (MELD) score has emerged as a useful tool for estimating mortality.8–11 It was developed originally to predict the outcomes of transjugular portosystemic shunt procedures in patients who had chronic liver disease.12 However, MELD was recently validated retrospectively on five independent chronic liver disease data sets representing diverse patient populations with a broad spectrum of liver diseases and etiologies.8 Moreover, the model has been shown to be of use in ranking candidates according to their probability of dying during a defined period. It also has been shown to be useful in predicting mortality independent of etiology and the complications of portal hypertension (e.g., esophageal variceal bleeding, spontaneous bacterial peritonitis, portosystemic encephalopathy).8, 10 Wiesner et al found the MELD score to be superior to the Child-Pugh score for ranking candidates by the severity of their liver disease and their risk of dying.13
The new MELD-based allocation policy was implemented on February 27, 2002, by the United Network of Organ Sharing (UNOS) as the criteria for organ allocation in patients who had chronic disease and were awaiting LT.14 With the new policy, the priority of HCC candidates has been maintained. Candidates with stage T1 (one lesion <2 cm) or stage T2 (one lesion ≥ 2 cm but < 5 cm, or as many as three lesions less than 3 cm each) HCC have priority beyond the degree of hepatic decompensation. Recently, the priority MELD score for stage T1 and stage T2 HCC was reduced from 24 to 20 and from 29 to 24, respectively, and it now corresponds to a pretransplantation death probability of 8% (15% previously) and, within 3 months, a pretransplantation death probability of 15% (30% previously), respectively.15, 23
The aim of our study was to evaluate the impact of the priority MELD scores for HCC candidates with respect to incidence rate of DDLT, time to DDLT, dropout rate from the waiting list because of clinical deterioration or death, candidate survival while waiting on the list, and survival after DDLT for the period before or after the new allocation policy was established.
DDLT, deceased donor liver transplantation; HCC, hepatocellular carcinoma; LT, liver transplantation; MELD, model for end-stage liver disease; OPTN, Organ Procurement and Transplantation Network; UNOS, United Network of Organ Sharing.
Patients and Methods
Candidate and Data Collection
The UNOS database between July 1999 and July 2002 was reviewed for all HCC candidates who had registered as transplantation candidates or as transplant recipients or who had a status 2B justification form (pre-MELD) or an exceptional case request (post-MELD). All DDLT candidates were included for analysis of outcome after LT. The candidates who had a living donor or a domino transplant were excluded. Demographic information was collected for age, gender, date of listing, date of transplantation, initial serum bilirubin, creatinine clearance, international normalized ratio, date of dropout, reason for dropout, survival on the waiting list, and survival after DDLT. The candidates were grouped by time periods as pre-MELD or post-MELD. The pre-MELD period included candidates who were listed before February 27, 2002. The post-MELD period included candidates who were listed and had DDLT after February 27, 2002.
Incidence rates were calculated in two time periods, pre-MELD and post-MELD, and are expressed as DDLT per person year. Time spent on the waiting list before February 27, 2002, was included in the calculation of pre-MELD person years, whereas time spent on the waiting list after February 27, 2002, was included in the calculation of post-MELD person years. The time to transplantation for all patients pre-MELD and post-MELD was calculated as 1 per incidence rate of DDLT, and the results are expressed in years per DDLT. The MELD score was calculated using the following formula:
The serum creatinine levels were capped at 4 mg/dL. The laboratory values for concentrations of creatinine less than 1 mg/dL were reset to 1 mg/dL before the calculations were performed.
The candidates who died or became too sick to receive a transplant because of clinical deterioration were included in calculations of the dropout rate. The maximum follow-up after implementation of MELD was 5 months. Therefore, the 5-month survival of HCC candidates on the waiting list, the 5-month candidate dropout rate, and the 5-month survival after DDLT were calculated for pre-MELD and post-MELD periods.
One-way ANOVA was used to analyze descriptive continuous variables. Categorical variables were analyzed using the chi-square test. Kaplan-Meier curves were generated for the survival data, and incidence rates were also compared between the pre- and post-MELD groups.
According to the UNOS database, 2,074 HCC candidates in the United States were listed for DDLT during the study period (pre-MELD period n = 1,717, and post-MELD period n = 357). Of the 1,717 candidates listed before the new allocation policy, 912 had waiting time in the pre-MELD period only, and the remaining 805 and had waiting time in both time periods. In the pre-MELD period, the 2B justification form for a priority Child-Pugh status beyond the degree of hepatic decompensation was available in 67.5% (616/912) of the candidates. In the 805 HCC candidates from pre-MELD period who carried their waiting time to the post-MELD period, an exceptional case request form was used for them to get a priority MELD score. This form was available for 86.8% (699/805) of the candidates. Similarly, in the post-MELD period, an exceptional case request form was available for 88% (314/357). Table 1 summarizes patient demographics and MELD score at listing of the HCC candidates who underwent DDLT. Table 2 summarizes the DDLT incidence rate for listed HCC candidates, time to transplantation in years, 5-month HCC candidate survival while on the UNOS/OPTN (Organ Procurement and Transplantation Network) list, 5-month dropout rate because of clinical deterioration or death, and 5-month patient survival after DDLT during pre-MELD and post-MELD periods. The ages and MELD scores of candidates who had DDLT were similar in both groups. The number of DDLTs performed for HCC candidates in the pre-MELD period was 535. In the post-MELD period, 644 DDLTs were performed. Of these, 404 candidates had waiting times that overlapped in the two time periods. The remaining 240 candidates transplanted in the post-MELD period had all their waiting time in the post MELD period.
Table 1. UNOS Database: Demographic Characteristics and MELD Scores at Time of Listing for the HCC Candidate Who Had DDLT During Two Time Periods
Abbreviations: DDLT, decreased donor liver transplantation; HCC, hepatocellular carcinoma; MELD, model for end-stage liver disease; OPTN, Organ Procurement and Transplantation Network; UNOS, United Network of Organ Sharing.
Age, gender and MELD score are calculated for the patients who underwent DDLT during pre- and post-MELD periods.
Only 77 of 535 candidates who underwent DDLT in the pre-MELD period had information to calculate the MELD scores.
Table 2. UNOS Database: DDLT Incidence Rate, Time to DDLT, 5-Month HCC Candidate Survival on List, 5-Month Dropout From List Because of Deterioration or Death, and 5-Month Patient Survival After DDLT in Pre-MELD and Post-MELD Periods
Abbreviations: DDLT, decreased donor liver transplantation; HCC, hepatocellular carcinoma; MELD, model for end-stage liver disease; NS, not significant; OPTN, Organ Procurement and Transplantation Network; UNOS, United Network of Organ Sharing.
Only 77 of 535 candidates who underwent DDLT in the pre-MELD period had information to calculate the MELD scores. The percentage of HCC candidates who dropped off the waiting list because of clinical deterioration or death was 25.9% (236/912) pre-MELD, 8.07% (65/805) in the period of overlap, and 6.7% (24/357) post-MELD (P < 0.001). Similarly, the percentage of HCC candidates who died while waiting was 14.5% (132/912) pre-MELD, 3.2% (26/805) in the period of overlap, and 4.5% (16/357) post-MELD (P < 0.001).
The total number of person years pre-MELD was 1,218.9, and the total person years post-MELD was 442.8. As noted earlier, the number of HCC candidates who had DDLT was 535 pre-MELD and 644 post-MELD. The DDLT incidence rate was 0.439 transplant/person years pre-MELD and 1.454 transplant/person years post-MELD (P < 0.001). The time to DDLT for the listed HCC candidates was 2.28 years/DDLT pre-MELD and 0.69 years/DDLT post-MELD (P < 0.001). The 5-month HCC candidate dropout rate because of clinical deterioration or death was significantly improved post-MELD (8.5%) compared with dropout pre-MELD (16.5%; P < 0.001). Figure 1 illustrates the national patient survival after DDLT. The 5-month survival was similar both pre-MELD and post-MELD. Tumor recurrence cannot be calculated after the new allocation policy because of the short follow-up time. Figure 2 shows that the 5-month survival for HCC candidates on the UNOS/OPTN list was 90.3% pre-MELD and 95.7% post-MELD (P < 0.001).
Liver transplantation is the acknowledged treatment of choice for patients who have early, unresectable HCC.16–19 This standard was reflected in the assignment by the UNOS of priority status 2B to candidates who had early HCC, which was equivalent to the priority assigned to candidates with Child-Pugh class C cirrhosis.20 Despite this priority, these candidates had to wait an extended time for a deceased donor. During the long wait, the disease progressed in many candidates to the extent that they were excluded from transplantation. The new MELD-based allocation policy maintained this priority for HCC candidates. However, we found that after the MELD-based policy was implemented, significantly more of the HCC candidates received a DDLT. The national waiting time for DDLT was significantly shorter for those listed in the post-MELD period compared to those listed pre-MELD.
In this study, the subjects were accrued over widely different periods of time, whereas the waiting list varied in size. Thus, the incidence rate was calculated to correct for changes in the waiting-list size during different times. The DDLT incidence rate for the listed HCC candidates was significantly higher post-MELD. The time to DDLT for HCC candidates was significantly shorter post-MELD. The dropout rate from the list included only medical reasons (clinical deterioration or death). The 5-month dropout rate from the waiting list for HCC candidates was reduced post-MELD. The percentage of deaths on the waiting list was highest pre-MELD. The HCC candidates in the post-MELD period as well as those who had waiting time in both periods and got priority MELD after February 27, 2002, had earlier DDLT. This finding may account for the decreased dropout rate and the decreased number of deaths on the waiting list for candidates post-MELD. Otherwise, all these candidates had lower 3-month mortality as evidenced by their lower MELD score. Furthermore, their 5-month survival on the waiting list also improved in the post-MELD period.
According to national data, the mean MELD score in transplantations for end-stage liver disease with other etiologies is 22.1 post-MELD.21 The percentage of HCC candidates who had DDLT within 30 days was 27% for those with T1 tumors and 45% for those with T2 tumors.21 Within 3 months of being listed for transplantation, more than 87% of the patients with HCC had received a transplant. This finding raised concern that HCC candidates were being given excessive priority. The issue was addressed in the UNOS/OPTN consensus meeting in January 2003, when the priority MELD score for HCC was decreased from 15% to 8% (equivalent to a MELD score of 20) for patients with stage T1 tumors and from 30% to 15% (equivalent to a MELD score of 24) for patients with stage T2 tumors.23 This modification took effect February 27, 2003. In addition, recent data indicate that HCC candidates with a single nodule ≤ 2 cm have less than a 10% risk of dying or becoming nontransplantable within 1 year. However, candidates with T2 HCC have a 60% chance of dying or becoming nontransplantable within 1 year of follow-up.21, 24, 25
According to an OPTN news release of November 13, 2002, preliminary data indicate that the new MELD-based allocation policy is functioning well.22 Data for the past 6 months under the new policy (between February 27,2002, and August 27, 2002) were compared with data from the same period in 2001. Early results indicate that the new policy is meeting the key goal of reducing the average death rate on the waiting list. Further analysis of short-term results showed that about one-fifth of the liver transplant recipients who had transplantation under the new policy had some form of HCC.
The national rate of transplantation for HCC post-MELD is 21.7%, compared to 8.8% pre-MELD.21 The increase in the rate of transplantation for HCC candidates post-MELD reflects the prioritization of these candidates. Does this prioritization of HCC candidates adversely affect sicker cirrhotic candidates? What are the waiting time, delisting rate, and pretransplantation mortality for cirrhotic candidates with other etiologies? Preliminary indications are that the new allocation system based on MELD has a favorable effect by reducing waiting list mortality while maintaining excellent early posttransplantation survival for all diagnoses.21 However, the marked advantage given to HCC candidates will need to be addressed further. The effects of the recent amendment to the priority MELD score for HCC candidates have yet to be determined.23
In conclusion, significantly higher proportions of candidates listed with the diagnosis of HCC are receiving a DDLT in the new MELD-based allocation system. The time to DDLT and the 5-month dropout rate from the waiting list for HCC candidates because of medical reasons has decreased since adoption of the new MELD scoring system. The current MELD-based UNOS/OPTN policy has improved the 5-month survival for HCC candidates on the waiting list. The impact of this new policy on the waiting time for DDLT for cirrhotic candidates who have other etiologies is unknown and needs further investigation. Based on the review of the UNOS data, the prioritization for HCC is increased under the new allocation policy.