- Email: [email protected]
PELD System
ALLOCATION
Excellent Liver Transplant Survival Rates Under the MELD/PELD System R.B. Freeman, A. Harper, and E.B. Edwards ABSTRACT The MELD/PELD (M/P) system for liver allocation was implemented on February 27, 2002, in the United States. Since then sufficient time has elapsed to allow for assessment of posttransplant survival rates under this system. We analyzed 4163 deceased donor liver transplants performed between February 27, 2002, and December 31, 2003, for whom follow-up reporting was 95% and 67% complete at 6 and 12 months, respectively. Kaplan-Meier survival analysis revealed 1-year patient and graft survival rates for status 1 of 76.9% and 70.4%, respectively, and 87.3% and 82.9% for patients prioritized by M/P (P ⬍ .0001 for status 1 vs M/P). When adult candidates were stratified by MELD score quartile at transplant, 1-year survival rates were 89.5%, 88.3%, 86.6%, and 78.1% for lowest to highest quartile (P ⫽ .0002) and graft survival rates were similarly distributed (85.0%, 84.5%, 82.7%, 73.0%, P ⬍ .0001). Candidates with hepatocellular cancer (89.6%) and other MELD score exceptions (88.8%) had slightly higher 1-year survival rates compared with standard MELD recipients (86.0%), which did not reach statistical significance (P ⫽ .089). Pediatric recipients had slightly better patient (88.7%) and graft (86.5%) survival rates at 1 year than adults but there were no significant differences among the PELD strata due to small numbers of patients in each PELD quartile. We conclude that patient and graft survival have remained excellent since implementation of the MELD/PELD system. Although recipients with MELD scores in the highest quartile have reduced survival compared with other quartiles, their 1-year survival rate is acceptable when their extreme risk of dying without a transplant is taken into consideration. HE MELD/PELD SYSTEM for liver allocation1 was introduced in the United States on February 27, 2002. The purpose for changing the allocation policy was to more accurately and objectively prioritize patients in need of liver transplantation by medical urgency. Preliminary results indicate a reduction in waiting list deaths and an increase in transplantation rates for waiting candidates.2 However, moving to a system that is designed to select candidates with higher medical urgency may have detrimental impact on the outcome of liver transplantation. Sufficient time has elapsed since implementation of this system to allow for accrual of follow-up data to calculate patient and graft survival esti-
T
© 2005 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 37, 585–588 (2005)
mates using Kaplan-Meier techniques. Our aim is to report patient and graft survival rates for adults and children based on the first 18 months’ experience with MELD/PELD and compare these rates for various strata of MELD/PELD scores at the time of transplant. From the Division of Transplant Surgery, Tufts-New England Medical Center, Boston, Massachusetts, USA. Address reprint requests to Richard B. Freeman, Jr, MD, Division of Transplant Surgery, Tufts-New England Medical Center, Box 40, 750 Washington Street, Boston, MA 02111. E-mail: [email protected] 0041-1345/05/$–see front matter doi:10.1016/j.transproceed.2004.12.099 585
586
FREEMAN, HARPER, AND EDWARDS
Fig 1. Kaplan-Meier plot of patient survival for all recipients receiving deceased donor liver transplants at status 1 compared with MELD/PELD candidates (P ⬍ .0001, log-rank).
METHODS Utilizing the National Organ Procurement and Transplantation Network (OPTN) database maintained by the United Network for Organ Sharing (UNOS), we collected data from all deceased donor liver transplant performed between February 27, 2002, and December 31, 2003. Calculated MELD/PELD scores were obtained for all recipients and note was made of any exceptional upgrade in MELD or PELD score that was granted by the Regional Review Boards (RRB). Recipients were stratified by calculated MELD/PELD score at transplant or by age or exception upgrade status. Follow-up information for patient and graft events was obtained from transplant recipient registration forms filed at the time of discharge from the primary transplant procedure or at 6 months or 1 year after transplant as mandated by OPTN policy.3 Patient survival was calculated from the date of transplant until the date of death as reported by the transplant center. Graft survival was calculated as the time from transplant to retransplantation or death. Kaplan-Meier curves were constructed with comparisons between groups performed using the log-rank test. A two-tailed P value ⬍.05 was considered statistically significant.
RESULTS
Four thousand one hundred sixty-three liver transplants were reported to the UNOS/OPTN database for the study
Fig 2. Kaplan-Meier plots of patient survival for adult recipients receiving deceased donor liver transplants stratified by calculated MELD score at transplant. Status 1, pediatric, and exceptional recipients excluded (P ⫽ .0002, log-rank).
period. For these, 6-month follow-up forms were complete in 95% and 12-month follow-up reporting was received for 67%. Of the 4163 total, 3499 (84%) were adults and 256 (16%) were children. For the adult recipients, 2454 (70.1%) received their organ under standard MELD scores, 856 (24.5%) received transplants at increased priority because of hepatocellular cancer (HCC), and 197 (5.6%) were transplanted at increased priority by virtue of other exceptions to MELD granted by the RRBs. Overall patient survival for recipients receiving transplants at the emergent status 1 category was lower than for those transplanted in the MELD/PELD category (Fig 1). One-year actuarial patient survival for status 1 recipients was 76.4% compared with 87.3% for all other recipients (P ⬍ .0001). Graft survival rates are similarly distributed for status 1 (70.4%) and M/P (82.9%, P ⬍ .0001) recipients. All of these results compare favorably with results from previous allocation systems. When adult recipients are stratified by calculated MELD score at transplant, we find a very slight but statistically significant difference in survival rates for the most ill recipients (Fig 2) with recipients with MELD score ⬎ 35 having 12-month survival estimates of 78.1% compared with 86.6% for MELD scores of 26 to 35, 88.3% for
SURVIVAL RATES UNDER THE MELD/PELD SYSTEM
587
Fig 3. Kaplan-Meier plots of patient survival for adult recipients receiving deceased donor liver transplants stratified by standard MELD, HCC exception or other, non-HCC exception score (P ⫽ .0893, log-rank).
MELD scores of 16 to 25, and 89.5% for MELD scores of 7 to 15 (P ⫽ .0002). Graft survival was similarly distributed for adults stratified by MELD score at transplant (MELD ⬎ 35, 73.0%; MELD 26 to 35, 82.8; MELD 16 to 25, 84.5; MELD 7 to 15, 84.3; P ⬍ .0001). Patients who were granted increased priority by the RRBs, either because of a diagnosis of HCC meeting UNOS/OPTN stage 1 or 2 criteria or because of some other exceptional diagnosis, had slightly improved survival rates compared with those transplanted at standard MELD/ PELD scores (Fig 3), although these differences did not reach statistical significance (P ⫽ .089). Pediatric patients also had excellent survival rates but, due to many fewer patients in the individual PELD strata, no statistically significant differences in patient (Fig 4) (P ⫽ .1527) or graft (P ⫽ .144) survival was seen. DISCUSSION
Several recent reports have documented an association between MELD score at liver transplant and patient and
graft survival. However, all of these reports are based on cohorts of patients who were prioritized by the previous allocation system where MELD/PELD played no role in determining which patients were given organs. In this report, we found a significant difference in survival rates among the quartiles of MELD scores for adults. The most apparent finding is the reduced 1-year survival rate for the recipients with MELD scores ⬎ 35. Nonetheless, it is the excellent 1-year survival rates in excess of 75% for this group of patients who have essentially 0% chance of surviving that is most clinically relevant. Clearly, the candidates with MELD scores ⬎ 35 who centers have selected for transplantation should not be excluded on the basis of utilitarian arguments against allocating organs to the most ill candidates. Importantly, though, the recipients with MELD scores ⬎ 35 are likely to be selected by transplant center for transplantation due to some other favorable factors, such as age or lack of comorbidities, that are not captured by their high MELD scores. These early result for candidates with HCC are also encour-
Fig 4. Kaplan-Meier plots of patient survival for pediatric recipients receiving deceased donor liver transplants stratified by calculated PELD score at transplant (P ⫽ .153, log-rank).
588
aging. However, much longer follow-up periods will be required to allow for capture of recurrent HCC cases that will better define the long-term success for these candidates. In conclusion, these early patient and graft survival rates for liver transplantation under the MELD/PELD have not deteriorated since institution of this new system. Longer follow-up, more in-depth analyses of subgroups, particularly those with high MELD/PELD scores at transplant, and more complete reporting will be necessary to draw firmer conclusions.
FREEMAN, HARPER, AND EDWARDS
REFERENCES 1. Freeman RB, Wiesner RH, Harper A, et al: The New Liver Allocation System: moving towards evidence-based transplantation policy. Liver Transplantation 8:851, 2002 2. Freeman RB, Wiesner RH, Harper A, et al: Results of the first year of the new liver allocation plan. Liver Transplantation 10:7, 2004 3. See www.optn.org policies and bylaws for data reporting requirements.
Excellent Liver Transplant Survival Rates Under the MELD/PELD System R.B. Freeman, A. Harper, and E.B. Edwards ABSTRACT The MELD/PELD (M/P) system for liver allocation was implemented on February 27, 2002, in the United States. Since then sufficient time has elapsed to allow for assessment of posttransplant survival rates under this system. We analyzed 4163 deceased donor liver transplants performed between February 27, 2002, and December 31, 2003, for whom follow-up reporting was 95% and 67% complete at 6 and 12 months, respectively. Kaplan-Meier survival analysis revealed 1-year patient and graft survival rates for status 1 of 76.9% and 70.4%, respectively, and 87.3% and 82.9% for patients prioritized by M/P (P ⬍ .0001 for status 1 vs M/P). When adult candidates were stratified by MELD score quartile at transplant, 1-year survival rates were 89.5%, 88.3%, 86.6%, and 78.1% for lowest to highest quartile (P ⫽ .0002) and graft survival rates were similarly distributed (85.0%, 84.5%, 82.7%, 73.0%, P ⬍ .0001). Candidates with hepatocellular cancer (89.6%) and other MELD score exceptions (88.8%) had slightly higher 1-year survival rates compared with standard MELD recipients (86.0%), which did not reach statistical significance (P ⫽ .089). Pediatric recipients had slightly better patient (88.7%) and graft (86.5%) survival rates at 1 year than adults but there were no significant differences among the PELD strata due to small numbers of patients in each PELD quartile. We conclude that patient and graft survival have remained excellent since implementation of the MELD/PELD system. Although recipients with MELD scores in the highest quartile have reduced survival compared with other quartiles, their 1-year survival rate is acceptable when their extreme risk of dying without a transplant is taken into consideration. HE MELD/PELD SYSTEM for liver allocation1 was introduced in the United States on February 27, 2002. The purpose for changing the allocation policy was to more accurately and objectively prioritize patients in need of liver transplantation by medical urgency. Preliminary results indicate a reduction in waiting list deaths and an increase in transplantation rates for waiting candidates.2 However, moving to a system that is designed to select candidates with higher medical urgency may have detrimental impact on the outcome of liver transplantation. Sufficient time has elapsed since implementation of this system to allow for accrual of follow-up data to calculate patient and graft survival esti-
T
© 2005 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 37, 585–588 (2005)
mates using Kaplan-Meier techniques. Our aim is to report patient and graft survival rates for adults and children based on the first 18 months’ experience with MELD/PELD and compare these rates for various strata of MELD/PELD scores at the time of transplant. From the Division of Transplant Surgery, Tufts-New England Medical Center, Boston, Massachusetts, USA. Address reprint requests to Richard B. Freeman, Jr, MD, Division of Transplant Surgery, Tufts-New England Medical Center, Box 40, 750 Washington Street, Boston, MA 02111. E-mail: [email protected] 0041-1345/05/$–see front matter doi:10.1016/j.transproceed.2004.12.099 585
586
FREEMAN, HARPER, AND EDWARDS
Fig 1. Kaplan-Meier plot of patient survival for all recipients receiving deceased donor liver transplants at status 1 compared with MELD/PELD candidates (P ⬍ .0001, log-rank).
METHODS Utilizing the National Organ Procurement and Transplantation Network (OPTN) database maintained by the United Network for Organ Sharing (UNOS), we collected data from all deceased donor liver transplant performed between February 27, 2002, and December 31, 2003. Calculated MELD/PELD scores were obtained for all recipients and note was made of any exceptional upgrade in MELD or PELD score that was granted by the Regional Review Boards (RRB). Recipients were stratified by calculated MELD/PELD score at transplant or by age or exception upgrade status. Follow-up information for patient and graft events was obtained from transplant recipient registration forms filed at the time of discharge from the primary transplant procedure or at 6 months or 1 year after transplant as mandated by OPTN policy.3 Patient survival was calculated from the date of transplant until the date of death as reported by the transplant center. Graft survival was calculated as the time from transplant to retransplantation or death. Kaplan-Meier curves were constructed with comparisons between groups performed using the log-rank test. A two-tailed P value ⬍.05 was considered statistically significant.
RESULTS
Four thousand one hundred sixty-three liver transplants were reported to the UNOS/OPTN database for the study
Fig 2. Kaplan-Meier plots of patient survival for adult recipients receiving deceased donor liver transplants stratified by calculated MELD score at transplant. Status 1, pediatric, and exceptional recipients excluded (P ⫽ .0002, log-rank).
period. For these, 6-month follow-up forms were complete in 95% and 12-month follow-up reporting was received for 67%. Of the 4163 total, 3499 (84%) were adults and 256 (16%) were children. For the adult recipients, 2454 (70.1%) received their organ under standard MELD scores, 856 (24.5%) received transplants at increased priority because of hepatocellular cancer (HCC), and 197 (5.6%) were transplanted at increased priority by virtue of other exceptions to MELD granted by the RRBs. Overall patient survival for recipients receiving transplants at the emergent status 1 category was lower than for those transplanted in the MELD/PELD category (Fig 1). One-year actuarial patient survival for status 1 recipients was 76.4% compared with 87.3% for all other recipients (P ⬍ .0001). Graft survival rates are similarly distributed for status 1 (70.4%) and M/P (82.9%, P ⬍ .0001) recipients. All of these results compare favorably with results from previous allocation systems. When adult recipients are stratified by calculated MELD score at transplant, we find a very slight but statistically significant difference in survival rates for the most ill recipients (Fig 2) with recipients with MELD score ⬎ 35 having 12-month survival estimates of 78.1% compared with 86.6% for MELD scores of 26 to 35, 88.3% for
SURVIVAL RATES UNDER THE MELD/PELD SYSTEM
587
Fig 3. Kaplan-Meier plots of patient survival for adult recipients receiving deceased donor liver transplants stratified by standard MELD, HCC exception or other, non-HCC exception score (P ⫽ .0893, log-rank).
MELD scores of 16 to 25, and 89.5% for MELD scores of 7 to 15 (P ⫽ .0002). Graft survival was similarly distributed for adults stratified by MELD score at transplant (MELD ⬎ 35, 73.0%; MELD 26 to 35, 82.8; MELD 16 to 25, 84.5; MELD 7 to 15, 84.3; P ⬍ .0001). Patients who were granted increased priority by the RRBs, either because of a diagnosis of HCC meeting UNOS/OPTN stage 1 or 2 criteria or because of some other exceptional diagnosis, had slightly improved survival rates compared with those transplanted at standard MELD/ PELD scores (Fig 3), although these differences did not reach statistical significance (P ⫽ .089). Pediatric patients also had excellent survival rates but, due to many fewer patients in the individual PELD strata, no statistically significant differences in patient (Fig 4) (P ⫽ .1527) or graft (P ⫽ .144) survival was seen. DISCUSSION
Several recent reports have documented an association between MELD score at liver transplant and patient and
graft survival. However, all of these reports are based on cohorts of patients who were prioritized by the previous allocation system where MELD/PELD played no role in determining which patients were given organs. In this report, we found a significant difference in survival rates among the quartiles of MELD scores for adults. The most apparent finding is the reduced 1-year survival rate for the recipients with MELD scores ⬎ 35. Nonetheless, it is the excellent 1-year survival rates in excess of 75% for this group of patients who have essentially 0% chance of surviving that is most clinically relevant. Clearly, the candidates with MELD scores ⬎ 35 who centers have selected for transplantation should not be excluded on the basis of utilitarian arguments against allocating organs to the most ill candidates. Importantly, though, the recipients with MELD scores ⬎ 35 are likely to be selected by transplant center for transplantation due to some other favorable factors, such as age or lack of comorbidities, that are not captured by their high MELD scores. These early result for candidates with HCC are also encour-
Fig 4. Kaplan-Meier plots of patient survival for pediatric recipients receiving deceased donor liver transplants stratified by calculated PELD score at transplant (P ⫽ .153, log-rank).
588
aging. However, much longer follow-up periods will be required to allow for capture of recurrent HCC cases that will better define the long-term success for these candidates. In conclusion, these early patient and graft survival rates for liver transplantation under the MELD/PELD have not deteriorated since institution of this new system. Longer follow-up, more in-depth analyses of subgroups, particularly those with high MELD/PELD scores at transplant, and more complete reporting will be necessary to draw firmer conclusions.
FREEMAN, HARPER, AND EDWARDS
REFERENCES 1. Freeman RB, Wiesner RH, Harper A, et al: The New Liver Allocation System: moving towards evidence-based transplantation policy. Liver Transplantation 8:851, 2002 2. Freeman RB, Wiesner RH, Harper A, et al: Results of the first year of the new liver allocation plan. Liver Transplantation 10:7, 2004 3. See www.optn.org policies and bylaws for data reporting requirements.