Original article
A combination of models for end-stage liver disease and cirrhosis-related complications to predict the prognosis of liver cirrhosis

https://doi.org/10.1016/j.clinre.2012.04.014Get rights and content

Summary

Background

The Child-Pugh score, the model for end-stage liver disease (MELD) score, and the occurrence of cirrhosis-related complications are independent prognostic predictors used in the assessment of chronic liver diseases.

Objectives

The objectives of this study were to determine the best prognostic scoring system, and to create a combined method to predict the prognosis of liver cirrhosis more accurately.

Methods

We retrospectively reviewed 435 cirrhotic patients from January 2009 to June 2010 and evaluated their short- and medium-term survival. Child-Pugh, MELD and its advanced scoring systems were computed for each patient. The sensitivity and specificity of these scoring systems were analyzed and their validity was assessed using concordance (c)-statistics in predicting the prognosis of cirrhotic patients.

Results

Overall, 107 patients died within 6 months and 150 patients died within 1 year. The clinical and biochemical characteristics, cirrhosis-related complications, and the scores were significantly different among the survivors and patients who died. The largest area under the receiver operating characteristic curve was 0.741 for the integrated MELD (iMELD) at 6 months and 0.713 for iMELD at 12 months, indicating that iMELD was the best scoring system tested. Given this result, we created a new scoring system that combined iMELD and an index of cirrhosis-related complications, called iMELD-C. This novel system had c indexes of 0.758 for the 6-month survival and 0.746 for the 1-year survival.

Conclusions

The iMELD-C score is a better predictor of both short- and medium-term survival in patients with cirrhosis.

Introduction

The high incidence of mortality due to liver cirrhosis is a global public health problem. With the rapid progress of medical science, orthotopic liver transplantation (OLT) has been used to reverse the poor prognosis of liver cirrhosis patients; however, there is an increasing shortfall between the number of cirrhotic patients on waiting lists for OLT and the number of available donor livers [1]. In this regard, the model for the end-stage liver disease (MELD) scoring system was introduced in February 2002 as a tool to determine organ allocation priorities [2], [3].

The MELD score was calculated according to the original formula proposed by the Mayo Clinic:

3.8 × loge (bilirubin (mg/dl)) + 11.2 × loge (INR) + 9.6 × loge (creatinine (mg/dl)) + 6.4 × (etiology: 0 if cholestatic or alcoholic, 1 otherwise). (To avoid negative scores, laboratory values less than 1 mg/dl were rounded off to 1. The maximal value of creatinine was 4 mg/dl).

The MELD scoring system has many advantages, including its objectivity. MELD score calculation is based on the etiology of cirrhosis and three simple and objective laboratory variables, serum bilirubin, serum creatinine and INR. The only parameter in the model that requires subjective interpretation is the etiology of liver disease. However, when the disease etiology is excluded from the score calculation, the accuracy of the model is not appreciably affected [4]. Compared to MELD, the Child-Pugh score is more subjective, because it includes two subjective variables (ascites and encephalopathy) with the inevitable interobserver variation. Indeed, the Child-Pugh score has not been validated as a predictor of survival in patients with liver cirrhosis [5]. Further advantages of the MELD scoring system include its simplicity, ease of use, repeatability, and sensitivity to the dynamic changes of liver cirrhosis. Finally, the inclusion of creatinine measures into the MELD score calculation helps to reflects the influence of hepatorenal syndrome (HRS) on the prognosis of cirrhosis patients. The ability of MELD to predict 3-month mortality of cirrhosis patients has been shown by several studies to be approximately 80% accurate, i.e., to have a concordance (c)-statistic of 0.83 compared with 0.76 for the Child-Pugh score [6].

Despite these advantages, the MELD score does have some limitations, such as variability in the score because of the different laboratories and conditions, without hemodynamic index and so on.

To overcome these limitations, some scholars successively introduced four new mathematical equations based on both MELD and sodium (Na) levels, known as the MELD to sodium (MESO) index [7], the MELD with the incorporation of serum sodium (MELD-Na) [8], the integrated MELD (iMELD) score [9] and the MELD-ascites-sodium (MELD-AS) score [10]. The inclusion of Na indexes in these revised scoring systems was important because hyponatremia, which is a common event in liver cirrhosis, develops as a result of free water retention and is positively correlated with the severity of portal hypertension [11], [12]. Indeed, serum Na level inversely correlates with the severity of cirrhosis, and its measurement has been used to assess portal hypertension. Furthermore, serum Na is an early and independent predictor of HRS.

These revisions to the original MELD scoring system have proven beneficial. For example, when the MESO index is more or equal to 0.8, it can predict 85.4% of esophageal varices (EV), which has sensitivity of 85.4% and specificity of 84.4%. This performance is better than two other independent predictors of EV, thrombocytopenia and splenomegaly, whose sensitivity and specificity did not reach 75% [13]. Also, the MELD-AS score is considered advantageous because it considers indexes of serum Na and ascites, which are independent predictors of early mortality [14].

In addition to a high MELD score, some studies have demonstrated that patients who present with cirrhosis-related complications (for example, hepatic encephalopathy [HE] and EV bleeding) have a decreased survival rate compared with those without complications [15]. However, these major complications of cirrhosis are not assessed in the MELD scoring system. Despite the fact that survival in these patients is expected to be rather limited and a timely transplantation is needed, there has been no clear designation for the priority of organ allocation for these patients since the MELD scoring system was introduced. Therefore, the goal of this study was to combine the best scoring system and cirrhosis-related complications to provide an improved assessment of chronic liver disease. As detailed below, our data indicates that optimal identification of patients at risk of cirrhosis-related death can be obtained by adding four points to the score of the best scoring system if cirrhosis-related complications are present. We hypothesize that this combined scoring system could be used to predict the prognosis of liver cirrhosis more accurately.

Section snippets

Study population

From January 2009 to June 2010, patients with liver cirrhosis from our institute were prospectively evaluated and their medical files were retrospectively reviewed. We retrospectively studied 435 cirrhotic patients (337 males, 98 females; median age: 56 years of age; age range: 20 to 87). Liver cirrhosis was diagnosed on the basis of clinical, biochemical, and instrumental results. Liver disease etiology was hepatitis viral infection, including hepatitis B and C, in 344 patients, alcohol abuse

Demographics of patients subdivided according to 6-month and 1-year survival

During the one year follow-up period, the overall mortality rate at 6 and 12 months was 24.6% (107/435) and 34.5% (150/435), respectively. Of the 150 patients who died within 12 months, 47 were Child-Pugh class A (31%), 51 were class B (35%), and 52 were class C (35%). The causes of death were all related to liver disease. Of the 285 patients who survived more than one year, 156 were Child-Pugh class A (55%), 107 were class B (38%), and 22 were class C (8%).

The clinical characteristics of the

Discussion

Liver cirrhosis is the end-stage of many kinds of liver diseases with different etiology. The prognostic evaluation of patients with liver cirrhosis remains a challenging clinical issue, this is of importance because the demand for donor livers currently exceeds the number of patients on the waiting list for OLT. As such, an increased waiting time for transplantation leads to higher morbidity and mortality for patients with end-stage of liver disease. In this regard, improved prognostic

Disclosure of interest

The authors declare that they have no conflicts of interest concerning this article.

References (30)

  • R. Wiesner et al.

    The United Network for organ sharing liver disease severity score Committee. Model for end-stage liver disease (MELD) and allocation of donor livers

    Gastroenterology

    (2003)
  • F. Salerno et al.

    MELD score is better than Child–Pugh score in predicting 3-month survival of patients undergoing transjugular intrahepatic portosystemic shunt

    J Hepatol

    (2002)
  • M.R. Lucey 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

    Transplantation

    (1998)
  • M. Malinchoc et al.

    A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts

    Hepatology

    (2000)
  • T.I. Huo et al.

    Model for end-stage liver disease score to serum sodium ratio index as a prognostic predictor and its correlation with portal pressure in patients with liver cirrhosis

    Liver Int

    (2007)
  • Cited by (9)

    • Plasma claudin-3 is associated with tumor necrosis factor-alpha-induced intestinal endotoxemia in liver disease

      2019, Clinics and Research in Hepatology and Gastroenterology
      Citation Excerpt :

      A diagnosis of CHB was based on a history of serum hepatitis B surface antigen positivity for > 6 months, serum alanine aminotransferase levels greater than the upper limit of the normal level, serum total bilirubin < 10 times the normal level (171 µmol/L), and prothrombin time activity > 40% [11]. Liver cirrhosis was diagnosed on the basis of clinical, biochemical, and instrument results [12]. Decompensated LC was marked by the development of overt clinical signs, the most frequent of which were ascites, bleeding, encephalopathy, and jaundice [13].

    • The performance of prognostic models as predictors of mortality in patients with acute decompensation of cirrhosis

      2015, Annals of Hepatology
      Citation Excerpt :

      In Southern Brazil, Marroni, et al. evaluated 558 individuals on the waiting list for liver transplantation, and they observed that SNa-modified MELD scores are able to more accurately predict 3- and 6-month mortality (P < 0.001).16 The majority of the studies that evaluate the ability of MELD and its derivatives to predict mortality in cirrhotic patients are retrospective, and aim to evaluate 1-, 3-, 6- or 12-month mortality (Table 4).34,37 –39 As far as we are concerned, this is the first study to evaluate in-hospital mortality.

    • A revised scope in different prognostic models in cirrhotic patients: Current and future perspectives, an Egyptian experience

      2013, Arab Journal of Gastroenterology
      Citation Excerpt :

      All of the prognostic scoring systems had better prognostic accuracy for 1-year survival where the MELD-Na score had the highest AUC (0.743, 95% confidence interval (CI) 0.732–0.825) followed by MESO (0.736), iMELD (0.733), MELDNa (0.731), MELD (0.725), CrCTP (0.710) and CP (0.658). Based on the larger AUC of MELD-Na, we calculated the MELD-Na-C score, which was a combination of the MELD-Na score and an index of cirrhosis-related complications (C) [15]. We found that the MELD-Na-C had a high AUC for 1-year survival (0.753, 95% CI (0.748–0.859)).

    View all citing articles on Scopus
    View full text