Employment of meld score for the prediction of survival after liver transplantation

Abstracts

OBJETIVO: Analisar a acurácia geral do escore MELD pré-operatório para a predição da sobrevivência pós-transplante hepático (TH) e explorar fatores preditivos da sobrevivência de médio prazo (24 meses). MÉTODOS: Estudo de corte transversal incluindo pacientes transplantados pelo Serviço de Cirurgia Geral e Transplante Hepático do Hospital Universitário Oswaldo Cruz, Universidade de Pernambuco, entre 15 de julho de 2003 e 14 de julho de 2009. Utilizou-se análise da área sob curva ROC (receiver operating characteristic) como medida-resumo do desempenho do escore MELD e se exploraram fatores preditivos da sobrevivência de médio prazo utilizando análise uni e multivariada. RESULTADOS: A sobrevivência cumulativa de três, seis, 12 e 24 meses dos 208 pacientes estudados foi 85,1%, 79,3%, 74,5% e 71,1%, respectivamente. O escore MELD pré-operatório apresentou baixo poder discriminatório para a predição da sobrevivência pós-TH. Por análise univariada, identificaram-se a transfusão intraoperatória de hemácias (p<0,001) e plaquetas (p=0,004) e o tipo de anastomose venosa hepatocaval (p=0,008) como significativamente relacionados à sobrevivência de médio prazo dos pacientes estudados. No entanto, por análise multivariada, observou-se que apenas a transfusão de hemácias foi um fator preditivo independente deste desfecho. CONCLUSÃO: O escore MELD apresentou baixa acurácia geral para a predição da sobrevivência pós-transplante dos pacientes estudados, entre os quais, apenas a transfusão intraoperatória de hemácias foi identificada como fator preditivo independente da sobrevivência de médio prazo após o TH.

Análise de sobrevida; Pacientes; Transplante de órgãos; Transplante de fígado; Mortalidade

ORIGINAL ARTICLE

^IMaster's Degree, Health Sciences, Faculty of Medical Sciences and Institute of Biological Sciences, University of Pernambuco - FCM/ICB/UPE

^IIAssistant Professor, Department of Surgical Medicine, FCM/UPE

^IIIPhD in Surgery, Federal University of Pernambuco -UFPE

^IVAssociate Professor, Department of Surgical Medicine, FCM/UPE

^VPhD Graduate, Surgery UFPE

^VIAssistant Professor, Department of Surgical Medicine, FCM/UPE

^VIIProfessor, Department of Surgical Medicine, FCM / UPE

Address correspondence to

ABSTRACT

OBJECTIVE: To assess the overall accuracy of the preoperative MELD score for predicting survival after liver transplantation (LT) and appraise medium-term (24 months) predictors of survival.

METHODS: We conducted a cross-sectional study including patients transplanted by the Department of General Surgery and Liver Transplantation of the Oswaldo Cruz University Hospital, University of Pernambuco, between July 15^th, 2003 and July 14^th, 2009. We used analysis of area under ROC (receiver operating characteristic) as a summary measure of the performance of the MELD score and assessed predictors of medium-term survival using univariate and multivariate analysis.

RESULTS: The cumulative survival of three, six, 12 and 24 months of the 208 patients studied was 85.1%, 79.3%, 74.5% and 71.1%, respectively. The preoperative MELD score showed a low discriminatory power for predicting survival after TH. By univariate analysis, we identified intraoperative transfusion of red blood cells (p <0.001) and platelets (p = 0.004) and type of venous hepatocaval anastomosis (p = 0.008) as significantly related to medium-term survival of the patients studied. However, by multivariate analysis only red blood cell transfusion was a significant independent predictor of outcome.

CONCLUSION: The MELD score showed low overall accuracy for predicting post-transplant survival of patients studied, among which only intraoperative transfusion of red blood cells was identified as an independent predictor of survival in the medium term after TH.

Key words: Survival analysis. Patients. Organ transplantation. Liver transplantation. Mortality.

INTRODUCTION

Liver transplantation (LT) is currently the most effective treatment for patients with terminal chronic liver disease. However, the shortage of liver grafts persists as the main limiting factor for its development ¹. In order to reduce the number of deaths on the waiting list ^2,3 and eliminate possible confounding factors related to the chronological system, a new policy for alloca tion of liver grafts, based on the MELD (Model for End-Stage Liver Disease) system has been adopted in Brazil since July 4, 2006⁴.

Based on easily measurable laboratory data ^2,5,6, the MELD score has been considered more transparent and objective ^2,5,7 for the ordination of liver disease on the waiting list for TH. Nevertheless, although it has been validated as a predictor of mortality in the list, the role of this scoring system for predicting survival after HT remains controversial ^3,6,8, which could suggest the need to readjust the graft allocation criteria for a model that also took into consideration the results of TH ⁸.

The objective of this study was to evaluate the accuracy of preoperative MELD score to predict survival after TH and also analyze factors predictive of medium-term survival (24 months).

METHODS

We conducted a cross-sectional cohort study of adult and adolescent patients (e" 16 years) undergoing orthotopic liver transplantation (cadaver donor) in the Department of General Surgery and Liver Transplantation of the Oswaldo Cruz University Hospital, University of Pernambuco (HUOC / UPE) between July 15^th 2003 and July 14^th, 2009.

All transplant recipients were listed in the Transplant Center of Pernambuco and had their data collected from the database of the Service itself, considering the end date of preservation until June 15, 2011. Retransplanted patients were considered only in relation to the first procedure and we excluded patients transplanted for acute liver failure / fulminant liver failure and those undergoing transplantation of multiple organs, such as "split-liver" or "domino". Similarly, we also excluded patients with incomplete data in their hospital records. The variables studied included data related to the donor, recipient and transplant center. For the "MELD" score we considered "pure" value recorded in the medical records at admission for transplantation or calculated with examinations collected on this occasion, without regard to the proposed correction to prioritize candidates in "special situations" (corrected MELD) ⁴.

For the preoperative diagnosis of hepatocellular carcinoma (HCC) we considered the criteria of the Barcelona Consensus⁹, confirmed by pathology of the explant. For the selection of patients with HCC candidates for TH, we used the "Milan criteria". The diagnosis of viral hepatitis was based on the positivity of serological markers of each viral type. After TH, tacrolimus, mycophenolate (mofetil or sodium) and prednisone were used as immunosuppressive treatment, without changes in the protocols applied from 2003 to 2009.

Continuous variables were expressed as medians and interquartile range, while categorical variables were presented in absolute frequencies and percentages. The odds of post-HT survival were estimated by the Kaplan-Meier product-limit method.

The overall accuracy of the MELD score to predict survival after TH was evaluated using ROC (receiver operating characteristic) curve analysis. The area under the ROC curve (AUC) was used as a summary measure of the performance of the MELD score to predict this outcome. Additionally, stratified analyzes were conducted among patients with or without HCC.

Still, using univariate and multivariate analysis, we assessed the variables as factors predictive of medium-term survival (24 months) after TH. Initially, the association of each variable with the outcome mentioned in univariate models was evaluated using the logrank and Cox proportional hazards test. Then, the variables whose association with post-HT survival showed a statistically significance lower than 20% were analyzed in multivariate Cox proportional hazards to identify independent predictors. This analysis was also adjusted according to the criteria for allocation (chronological vs. MELD) and the time of transplantation (quartiles).

The collected data were tabulated in duplicate and the level of statistical significance in all analyzes was 1% (p <0.01).

The study was registered at SISNEP - Nationa Information System on Ethics in Research involving Human Subjects (CAAE - 0003.0.106.000/10) and approved by the Ethics in Human Research cCommittee of the Oswaldo Cruz University Hospital - Cardiac Emegncy Room of Pernambuco, under protocol number 12/2010.

RESULTS

From July 15^th, 2003 to July 14^th, 2009 298 HTs were performed in 288 patients by the Department of General Surgery and Liver Transplantation HUOC / UPE. Eighty patients were ineligible and were excluded from the study: 63 were younger than 16 years, ten were transplanted for acute liver failure / fulminant hepatitis, one underwent a domino transplant, one underwent a transplant of the "split-liver" and five cases had incomplete data in their hospital records. Descriptive statistics corresponding to the remaining cases included in the study (n = 208) is summarized in Table 1.

The proportional cumulative survival of patients at three, six, 12 and 24 months after HT was 85.1%, 79.3%, 74.5% and 71.1%, respectively (Figure 1). Throughout the 95.1 months of follow-up, 76 patients died (36.5%) and seven (3,4%) were subjected to a retransplant. The main causes of death were infection, tumor recurrence and liver failure / transplant rejection. The median of follow-up of the surviving patients was 42.7 months, ranging from 24.8 to 95.1 months.

Figure 2 shows the ROC curve of MELD score to predict the medium-term survival after HT. Its corresponding AUC is 0.504. AUCs of 0.476, 0.477 and 0.504 were observed for the prediction of survival of three, six and 12 months after HT, respectively. For the strata HCC and non-HCC, the corresponding values of AUC for the prediction of survival of 24 months were 0.594 and 0.514.

By univariate analysis, we identified transfusion of red blood cells (p <0.001) and platelets (p = 0.004) and type of hepatic venous reconstruction (p = 0.008) as significantly related to survival after HT. However, multivariate analysis showed that only red blood cell transfusion was an independent predictor of medium-term survival of transplant patients. In this study, each unit of RBCs transfused during surgery increased the relative risk of death within 24 months at 13.5%, which ranged from 12.1% to 13.5% after multivariate analysis (Table 2).

DISCUSSION

The MELD score was initially described by Malinchoc et al. ¹⁰ as a mathematical model capable of predicting survival in the first three months of patients who underwent percutaneous placement of transjugular intrahepatic portosystemic shunts (TIPS). Later, the model was rigorously validated as a predictor of mortality for cirrhotic patients on waiting lists for transplants ^6,11-13 and was eventually incorporated into a new criterion for the allocation of grafts for HT ^4,14,15.

Nevertheless, the role of the MELD score to predict survival after HT, which has been considered a matter of vital importance to the issue of liver transplantation^6, remains controversial ^3,6,8. In a systematic review on the performance of the MELD score in this scenario, Cholongitas et al. ⁶ concluded that this score was not a good predictor of short-term mortality after TH and that further studies were needed to evaluate its performance in the medium and long term.

In this study, analysis of the area under the ROC curve (AUC) was used as a summary measure of the MELD's performance for the prediction of survival after HT. Normally, AUC values e" 0.7 are considered clinically useful, whereas values d" 0.5 are related to scores with no discriminatory power ^3,16,17. Based on the results found, the preoperative MELD score showed low overall accuracy for predicting survival after HT, similar to that described in other Brazilian studies ^3,18.

Also, since the survival of transplanted patients with HCC is influenced by parameters related to the tumor ¹⁹, we performed analyzes stratified between subgroups HCC and non-HCC. However, no statistically significant difference was found between the medium-term survival of these subgroups, and the MELD score had also low overall accuracy for predicting survival after HT in both strata. Although the MELD criteria have been adopted to prioritize the most severe patients on the waiting list, the majority of transplanted patients is still low MELD scores ^6,20-22. Furthermore, it has been shown that the score does not display the same prognostic precision in milder forms of hepatic cirrhosis ^6,23. Thus, the effect of hepatic dysfunction may not be evident in patients with MELD scores below 30, suggesting that only higher values would affect post-HT survival ²⁴. In accordance with these reports, only 3.4% of transplant patients by the Service under review had MELD scores e" 30 (data not shown). This low proportion of patients with high scores likely contributed to minimize its overall accuracy for predicting survival after TH in this series. Intraoperative transfusion of red blood cells has often been related to lower survival after liver transplants ^18,25-28, which can also be seen in this series, which identified a strong influence of the transfusion of blood components in the medium-term survival of transplant patients. This negative effect of blood transfusions probably stems from immunomodulation related to transfusions and changes of its components from storage process ²⁹, which may increase the risk of nosocomial infections, acute lung injury and development of autoimmune diseases in the long term ³⁰. Moreover, as emphasized by Mitchell et al., the need for blood transfusion may act only as a marker of severity of the patients undergoing HT ³¹, for whom lower survival rates are expected.

Given the current recommendations for the presentation of observational studies ³², we used a significance level of only 1% for the statistical analyzes in our work. In addition, we avoided the categorization of continuous variables in order to minimize losses in statistical power and the occurrence of residual confounding related to the dichotomization method ^33,34. Similarly, stratified analyzes were performed according to the diagnosis of liver disease and multivariate analysis adjusted for temporal variables that could serve as confounding factors in interpreting the results, as the criterion for the allocation of grafts and the time of transplantation ³⁵.

In addition to the socio-economic peculiarities concerning the study population ³⁶, this series has special importance for the study of the predictive value of MELD score for post-HT survival in view of the divergent results found in Pernambuco after the adoption of the MELD criteria. Unlike the observed in other Brazilian states, in Pernambuco we observed a significant reduction in mortality on the waiting list ³⁷ and better post-HT survival after the adoption of this model, probably as a result of increasing governmental encouragements to transplantation of organs and tissues ³⁵. On the other hand, the inclusion of more donor-related variables could add relevant information to this analysis, especially due to the considerable number of marginal donors employed by the Service in the study ^1,38,39.

It therefore follows that the MELD score was not accurate for predicting survival after HT in the Department of General Surgery and Liver Transplantation HUOC / UPE where only intraoperative transfusion of red blood cells was identified as an independent predictive factor of medium-term survival after HT.

ACKNOWLEDGEMENTS

The authors thank Ms. Tania M. Xavier, Miss. Juliana S. Gomes and Dr. Isaac Esmeraldino for their invaluable help with data collection. We also thank Dr. Gedson A. Maia and Dr. Joseph A. de Morais for reviewing the English texts and Dr. Joseph N. Figueroa and Dr. Ulysses Montarroyos for their excellent technical assistance during the statistical analysis.

REFERENCES

• 1. Macedo FI, Miranda LE, Pádua TC, Fernandes JL, Neto OL, Lacerda CM. Effects of donor age on patient survival in liver transplantation: short-and long-term analysis. Hepatogastroenterology. 2009;56(93):1133-6.
• 2. Freeman RB, Wiesner RH, Edwards E, Harper A, Merion R, Wolfe R, et al. Results of the first year of the new liver allocation plan. Liver Transpl. 2004;10(1):7-15.
• 3. Brandão A, Fuchs SC, Gleisner AL, Marroni C, Zanotelli ML, Cantisani G, et al. MELD and other predictors of survival after liver transplantation. Clin Transplant. 2009;23(2):220-7.
• ⁴
  Brasil. Ministério da Saúde, Departamento de Cuidados da Saúde. Portaria GM nº 1160, de 29 de maio de 2006. Modifica os critérios de distribuição de fígado de doadores cadáveres para transplante, implantando o critério de gravidade de estado clínico do paciente. Diário Oficial [da República Federativa do Brasil] [Internet]. Brasília, 31 de mai. 2006; v. 103. Acessado em 12 abr. 2009. Disponível em: http://dtr2001.saude.gov.br/sas/PORTARIAS/Port2006/GM/GM-1160.htm
• 5. Freeman RB. Overview of the MELD/PELD system of liver allocation indications for liver transplantation in the MELD era: evidence-based patient selection. Liver Transpl. 2004;10(10 Suppl 2):S2-3.
• 6. Cholongitas E, Marelli L, Shusang V, Senzolo M, Rolles K, Patch D, et al. A systematic review of the performance of the model for end-stage liver disease (MELD) in the setting of liver transplantation. Liver Transpl. 2006;12(7):1049-61.
• 7. Freeman RB, Edwards EB. Liver transplant waiting time does not correlate with waiting list mortality: implications for liver allocation policy. Liver Transpl. 2000;6(5):543-52.
• 8. Onaca NN, Levy MF, Sanchez EQ, Chinnakotla S, Fasola CG, Thomas MJ, et al. A correlation between the pretransplantation MELD score and mortality in the first two years after liver transplantation. Liver Transpl. 2003;9(2):117-23.
• 9. Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R, Burroughs AK, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol. 2001;35(3):421-30.
• 10. Malinchoc M, Kamath PS, Gordon FD, Peine CJ, RanK J, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology. 2000;31(4):864-71.
• 11. Wiesner RH, McDiarmid SV, Kamath PS, Edwards EB, Malinchoc M, Kremers WK, et al. MELD and PELD: application of survival models to liver allocation. Liver Transpl. 2001;7(7):567-80.
• 12. Merion RM, Wolfe RA, Dykstra DM, Leichtman AB, Gillespie B, Held PJ. Longitudinal assessment of mortality risk among candidates for liver transplantation. Liver Transpl. 2003;9(1):12-8.
• 13. Wiesner R, Edwards E, Freeman R, Harper A, Kim R, Kamath P, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology. 2003;124(1):91-6.
• 14. Teixeira AC, Souza FF, Mota Gde A, Martinelli Ade L, Sankarankutty AK, Silva Ode C. Liver transplantation: expectation with MELD score for liver allocation in Brazil. Acta Cir Bras. 2006;21 Suppl 1:S12-4.
• 15. Ravaioli M, Grazi GL, Dazzi A, Bertuzzo V, Ercolani G, Cescon M, et al. Survival benefit after liver transplantation: a single European center experience. Transplantation. 2009;88(6):826-34.
• 16. Xiao L, Fu ZR, Ding GS, Fu H, Ni ZJ, Wang ZX, et al. Prediction of survival after liver transplantation for chronic severe hepatitis B based on preoperative prognostic scores: a single center's experience in China. World J Surg. 2009;33(11):2420-6.
• 17. Sempere L, Palazón JM, Sánchez-Payá J, Pascual S, de Madaria E, Poveda MJ, et al. Assessing the short- and long-term prognosis of patients with cirrhosis and acute variceal bleeding. Rev Esp Enferm Dig. 2009;101(4):236-48.
• 18. Boin IFSF, Leonardi MI, Udo EY, Sevá-Pereira T, Stucchi RSB, Leonardi LS. Aplicação do escore MELD em pacientes submetidos a transplante de fígado: análise retrospective da sobrevida e dos fatores preditivos a curto e longo prazo. Arq Gastroenterol. 2008;45(4):275-83.
• 19. Hemming AW, Cattral MS, Reed AI, Van Der Werf WJ, Greig PD, Howard RJ. Liver transplantation for hepatocellular carcinoma. Ann Surg. 2001;233(5):652-9.
• 20. Thuluvath PJ, Maheshwari A, Thuluvath NP, Nguyen GC, Segev DL. Survival after liver transplantation for hepatocellular carcinoma in the model for end-stage liver disease and pre-model for end-stage liver disease eras and the independent impact of hepatitis C virus. Liver Transpl. 2009;15(7):754-62.
• 21. Yoo HY, Thuluvath PJ. Short-term postliver transplant survival after the introduction of MELD scores for organ allocation in the United States. Liver Int. 2005;25(3):536-41.
• 22. Merion RM, Schaubel DE, Dykstra DM, Freeman RB, Port FK, Wolfe RA. The survival benefit of liver transplantation. Am J Transplant. 2005;5(2):307-13.
• 23. Huo TI, Lin HC, Wu JC, Lee FY, Hou MC, Lee PC, et al. Different model for end-stage liver disease score block distributions may have a variable ability for outcome prediction. Transplantation. 2005;80(10):1414-8.
• 24. Kanwal F, Dulai GS, Spiegel BM, Yee HF, Gralnek IM. A comparison of liver transplantation outcomes in the pre- vs. post-MELD eras. Aliment Pharmacol Ther. 2005;21(2):169-77.
• 25. Ramos E, Dalmau A, Sabate A, Lama C, Llado L, Figueras J, et al. Intraoperative red blood cell transfusion in liver transplantation: influence on patient outcome, prediction of requirements, and measures to reduce them. Liver Transpl. 2003;9(12):1320-7.
• 26. Hendriks HG, van der Meer J, de Wolf JT, Peeters PM, Porte RJ, de Jong K, et al. Intraoperative blood transfusion requirement is the main determinant of early surgical reintervention after orthotopic liver transplantation. Transpl Int. 2005;17(11):673-9.
• 27. Massicotte L, Sassine MP, Lenis S, Roy A. Transfusion predictors in liver transplant. Anesth Analg. 2004;98(5):1245-51.
• 28. de Boer MT, Christensen MC, Asmussen M, van der Hilst CS, Hendriks HG, Slooff MJ, et al. The impact of intraoperative transfusion of platelets and red blood cells on survival after liver transplantation. Anesth Analg. 2008;106(1):32-44.
• 29. Vamvakas EC, Blajchman MA. Transfusion-related immunomodulation (TRIM): an update. Blood Rev. 2007;21(16):327-48.
• 30. Raghavan M, Marik PE. Anemia, allogenic blood transfusion, and immunomodulation in the critically ill. Chest. 2005;127(1):295-307.
• 31. de Morais BS, Sanches MD, Ribeiro DD, Lima AS, de Abreu Ferrari TC, Duarte MM, et al. Association between the use of blood components and the five-year mortality after liver transplant. Rev Bras Anestesiol. 2011;61(3):286-92.
• 32. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, STROBE Initiative. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting of observational studies. BMJ. 2007;335(7624):806-8.
• 33. Royston P, Altman DG, Sauerbrei W. Dichotomizing continuous predictors in multiple regression: a bad idea. Stat Med. 2006;25(1):127-41.
• 34. Altman DG, Royston P. The cost of dichotomising continuous variables. BMJ. 2006;332(7549):1080.
• 35. Batista TP, Sabat BD, Melo PS, Miranda LE, Fonseca-Neto OC, Amorim AG, et al. Impact of MELD allocation policy on survival outcomes after liver transplantation: a single-center study in northeast Brazil. Clinics. 2011;66(1):57-64.
• 36. Lacerda CM, Melo PSV, Lucena O, Amorim A, Almeida HM, Pereira LB. Transplante de fígado em hospital universitário do Recife: resultados iniciais. Rev Col Bras Cir. 2003;30(1):29-35.
• 37. Tenório AL, Macedo FI, Miranda LE, Fernandes JL, da Silva CM, Neto OL, et al. Survival on waiting list for liver transplantation before and after introduction of the model for end-stage liver disease score. Transplant Proc. 2010;42(2):407-11.
• 38. Lucena O, Amorim AG, Sabat B, Adeodato LCL, Miranda LEC, Lacerda CM. Transplante de fígado a partir de doadores em assistolia: resultados iniciais do grupo de transplante de fígado do Hospital Oswaldo Cruz, Universidade de Pernambuco. Rev Col Bras Cir. 2005;32(5):270-2.
• 39. Fonseca-Neto OCL, Miranda LEC, Sabat BD, Amorim AG, Adeodato L, Melo PSV, et al. O doador marginal: experiência de um centro de transplante de fígado. ABCD arq bras cir dig. 2008;21(1):1-5.

Employment of meld score for the prediction of survival after liver transplantation

Thales Paulo Batista, TCBC-PE^I; Bernardo David Sabat^II; Paulo Sérgio Vieira de Melo, TCBC-PE^III; Luiz Eduardo Correia Miranda, TCBC-SP^IV; Olival Cirilo Lucena da Fonseca-Neto, TCBC-PE^V; Américo Gusmão Amorim,ACBC-PE^VI; Cláudio Moura Lacerda, TCBC-PE^VII

Publication Dates

• Publication in this collection
  31 May 2012
• Date of issue
  Apr 2012

History

• Received
  15 Aug 2011
• Accepted
  18 Oct 2011