Journal of Hepatology
Volume 44, Issue 1 , Pages 217-231, January 2006

Natural history and prognostic indicators of survival in cirrhosis: A systematic review of 118 studies

  • Gennaro D'Amico

      Affiliations

    • Gastroenterology Unit, Ospedale Cervello and University of Palermo, Italy
    • Corresponding Author InformationCorresponding author. Tel.: +39 091 6802780; fax: +39 091 6885111.
    • ,
  • Guadalupe Garcia-Tsao

      Affiliations

    • Section of Digestive Diseases, Yale University and CT-VA Healthcare System, Connecticut, USA
    • ,
  • Luigi Pagliaro

      Affiliations

    • Gastroenterology Unit, Ospedale Cervello and University of Palermo, Italy

published online 10 November 2005.

Article Outline

 

Back to Article Outline

1. Background 

Liver transplantation significantly improves the survival and quality of life of patients with end-stage cirrhosis. However, a large proportion of cirrhotic patients still die while on the transplant list because of an insufficient number of donors and because of the lack of an accurate prediction of life expectancy. Although many prognostic models have been proposed in the last two decades to predict mortality in cirrhosis, the Child–Pugh score [1] is by far the most largely used both in clinical practice and in clinical research. Recently, the model for end stage liver disease (MELD) has replaced the Child–Pugh score in the United States for prioritizing liver donor allocation [2]. MELD is considered more reproducible than the Child–Pugh score because it does not include subjective variables such as ascites and encephalopathy. However, obtaining the score requires computing and it is, therefore, much less practical than the Child–Pugh score for individual estimates at the bedside. Moreover, the MELD score has not been proven to be superior to the Child–Pugh score in terms of predictive accuracy [3], and both may be unsatisfactory when applied separately to compensated and decompensated patients [4]. It is in fact well known that life expectancies in compensated and decompensated cirrhosis are strikingly different and it is, therefore, conceivable that prognostic indicators may be different or may have a different weight according to the disease stage [4].

In this article we outline the natural history of cirrhosis and report on a systematic review of the literature regarding predictors of mortality in cirrhosis in order to evaluate whether variables included in the Child–Pugh and MELD scores and other variables have different validity throughout the different stages in the course of cirrhosis.

Back to Article Outline

2. Natural history of cirrhosis 

Cirrhosis is the end-stage of every chronic liver disease. Its natural history is characterized by an asymptomatic phase, termed ‘compensated’ cirrhosis followed by a rapidly progressive phase marked by the development of complications of portal hypertension and/or liver dysfunction, termed ‘decompensated cirrhosis’ (Fig. 1). In the compensated phase, portal pressure may be normal or below the threshold level identified for the development of varices or ascites (‘clinically significant portal hypertension’) [5]. As the disease progresses, portal pressure increases and liver function decreases, resulting in the development of ascites, portal hypertensive gastrointestinal (GI) bleeding, encephalopathy and jaundice. The development of any of these complications marks the transition from a compensated to a decompensated phase. Progression may be accelerated by the development of other complications such as (re)bleeding, renal impairment (refractory ascites, hepatorenal syndrome), hepatopulmonary syndrome and sepsis (spontaneous bacterial peritonitis). The development of hepatocellular carcinoma (HCC) may accelerate the course of the disease at any stage.

Classification of cirrhosis into a compensated and a decompensated stage is simple and reproducible and identifies patients at a similar rate of disease progression and survival. Decompensated cirrhosis is defined by the presence of ascites, variceal bleeding, encephalopathy and/or jaundice [6], [7]. Moreover, since ascites is most frequently the first of these signs to appear [8], it is usually considered a landmark sign of decompensated cirrhosis. Transition from a compensated to a decompensated stage occurs at a rate of ∼5–7% per year (Fig. 2) [9].

Survival of patients with compensated cirrhosis is significantly longer than that of decompensated patients with median survival times of >12 years and ∼2 years, respectively (Fig. 3A). Patients with compensated cirrhosis die mostly after transition to a decompensated stage. Survival while the patient remains in the compensated stage is calculated by censoring data at the first manifestation of decompensation and, as shown in Fig. 3B, is markedly higher than the probability of survival of a given clinical stage at a given point along the course of the disease, e.g. at diagnosis (Fig. 3A).

  • View full-size image.
  • Fig. 3. 

    Survival according to decompensation at diagnosis (A) and while remaining in the compensated or decompensated stages (B). Individual patient data from two prospective studies of the natural history of cirrhosis [8], [10]. [This figure appears in colour on the web.]

By combining data from two large natural history studies including 1649 patients [8], [10], four clinical stages or status of cirrhosis can be identified, each with distinct clinical features and a markedly different prognosis (Fig. 4). Each stage is defined by the presence or absence of complications of cirrhosis and was agreed upon in the recent Baveno IV consensus conference [11].

Stage 1 is characterized by the absence of esophageal varices and of ascites. While patients remain in this status, the mortality rate is as low as 1% per year. Patients exit this status at a cumulative rate of 11.4% per year: 7% because of the development of varices and 4.4% because of the development of ascites (with or without varices).

Stage 2 is characterized by the presence of esophageal varices without ascites and without bleeding. While patients remain in this status, the mortality rate is 3.4% per year. Patients leave this status by developing ascites (6.6% per year) or by developing variceal bleeding before or at the time of development of ascites (rate 4% per year).

Stage 3 is characterized by ascites with or without esophageal varices in a patient that has never bled. While patients remain in this status, the mortality rate is 20% per year, significantly higher than in the two former states. Patients exit this stage by bleeding (7.6% per year).

Stage 4 is characterized by GI bleeding with or without ascites. In this stage the one-year mortality rate is 57% (nearly half of these deaths occur within 6 weeks from the initial episode of bleeding).

Stages 1 and 2 correspond to patients with compensated cirrhosis while stages 3 and 4 refer to decompensated cirrhosis. HCC develops at a fairly constant rate of 3% per year and is associated with a worse outcome at whatever status it develops.

In the aforementioned consensus conference on portal hypertension it was suggested that this status classification should be adopted in future prognostic studies of cirrhosis and in randomized controlled trials [11].

2.1. Predictors of mortality: a systematic review 

2.1.1. (a) Methods 

All prognostic studies of cirrhosis were considered eligible for this review if they fulfilled the following criteria: (a) publication in the English language; (b) inclusion of adult patients with cirrhosis; (c) survival analysis was reported; d) follow-up ≥6 months, and (e) multivariable analysis of prognostic indicators of death risk was performed. Exclusion criteria were: (a) studies that included both cirrhotic and non cirrhotic patients; (b) studies assessing only short term mortality (≤6 months); (c) studies evaluating specific therapies (randomized controlled trials, prospective studies of post-therapeutic course, etc.) or prognosis after procedures (e.g. surgical shunts, transjugular intrahepatic portosystemic shunt), (d) studies including only patients with primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), metabolic or genetic liver diseases.

Studies were identified by a MEDLINE search using the following terms: (survival [ALL] OR mortality [ALL] OR predictor [ALL] OR prognosis [ALL] OR prognostic [ALL]) AND (multivariate OR Cox OR Cox's OR adjusted OR adjustment OR logistic [ALL]) AND cirrhosis [MESH]. Additional studies were located by manual search using references from retrieved articles.

The following data were extracted by two independent observers, based on pre-defined criteria: (a) aim of the study, explanatory (i.e. assessing the prognostic value of a specific variable according to a biologically plausible hypothesis) or predictive (assessing many variables without a primary concern for understanding the primary mechanism involved in their prognostic value); (b) design (prospective or retrospective); (c) inception cohort, i.e. whether patients entered the study at a well-defined stage; (d) characteristics of included patients; (d) sample size and follow-up; (e) survival analysis and causes of death; (f) analysis of prognostic factors; (g) study validation. The quality of the included studies was assessed according to widely accepted quality criteria for prognostic studies (marked with an asterisk in Table 1) [12], [13], [14].

Table 1. Summary of characteristics of 118 studies of predictors of mortality in cirrhosis
Characteristic of studiesN (%)
Aim
Explanatory76 (64)
Predictive42 (36)
Design
Prospective65 (58)
Retrospective48 (42)
*Inception cohort17 (15)
*Patients were included consecutively54 (46)
*Inclusion and exclusion criteria defined74 (63)
*Number of excluded patients specified34 (30)
*Cirrhosis well-defined95 (81)
*Candidate variables were identified a priori80 (68)
*Candidate variables included previously identified predictors of survivala96 (81)
*Relevant baseline data shownb74 (63)
Analysis specific to disease stage41 (35)
Compensated cirrhosis18 (15)
Decompensated cirrhosis23 (19)
Patients not previously included in other similar studies83 (86)
*Length of followup reported111 (94)
*Patients lost to followup reported63 (53)
*Number of deaths reported114 (97)
Causes of death reported81 (69)
*Ratio of number of deaths/number of variables >10 (i.e. no overfitting)44 (37)
*Missing data reported20 (17)
*Results validated internally or externally17 (15)

*Quality variables.

aFor studies prior to MELD, predictors should have included at a minimum Child (or Child–Pugh) score, Child (or Child–Pugh) class or its components. For studies after MELD, predictors must include Child (or Child–Pugh) or components or MELD or MELD components.

bAt a minimum the studies should have reported results on age, sex, etiology of cirrhosis and Child or Child–Pugh or all components, or MELD score or all its components.

The following five major quality criteria were used to select ‘good’ quality studies: (a) inclusion of consecutive patients; (b) relevant baseline patient characteristics were reported (age, gender, etiology of cirrhosis and Child class or components or MELD score); (c) length of follow-up was reported; (d) important variables with a previously established prognostic influence on survival were included in the analysis; (e) number of deaths were reported. Inclusion of patients as an inception cohort was dismissed as a quality criterion because it was reported in only 17 studies.

In order to achieve a more homogeneous and clinically sound assessment, the studies were grouped according to whether they included patients with compensated or decompensated cirrhosis. Compensated cirrhosis was defined by the absence of ascites, jaundice, portal-systemic encephalopathy or variceal bleeding. Correspondingly, decompensated cirrhosis was defined by the presence of any of these complications. Studies including both compensated and decompensated patients were considered in a separate group of ‘unclassifiable studies’.

The cumulative proportion of surviving patients after the longest observation period was recorded per each study (observed rates were used if cumulative proportions were not reported). However, since the length of follow-up was very different across studies, 1 and 2-year survival rates were also recorded to obtain more homogeneous estimates.

To identify the most accurate prognostic indicators, only variables in the first five levels of statistical significance (according to the P value pertinent to each single variable in the final multivariable analysis) were selected from each study as having a prognostic value [15], [16]. To provide a rough measure of the reproducibility of each prognostic variable, we calculated the ratio between the number of studies in which each variable was significant and the number of studies in which it was assessed. To explore the reliability of the prognostic indicators identified in all studies, a sensitivity analysis was performed by analysing only the good quality studies.

2.1.2. (b) Results 

A total of 804 references were obtained through a MEDLINE search performed on October 10, 2005; 11 more were obtained outside the search, for a total of 815 references. Of these, 685 were excluded because they were irrelevant or did not meet inclusion criteria by reading title and abstract. Five papers were unobtainable [17], [18], [19], [20], [21] and eight more were excluded after reading the paper because they did not meet inclusion criteria [22], [23], [24], [25], [26], [27], [28], [29]. Therefore, 117 studies met inclusion and exclusion criteria and constitute the basis of this investigation. Of note, one study separately described and analysed two patient populations (decompensated and compensated cirrhosis) [8] and therefore, this study is considered separately for each of these populations, bringing the total to 118 evaluated studies. Six were published between 1983–1985 [30], [31], [32], [33], [34], [35], 19 between 1985–1990 [8], [12], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51]; 27 between 1990–1995 [52], [53], [54], [55], [56], [57], [58], [59], [60], [61], [62], [63], [64], [65], [66], [67], [68], [69], [70], [71], [72], [73], [74], [75], [76], [77], [78]; 26 between 1995–2000 [79], [80], [81], [82], [83], [84], [85], [86], [87], [88], [89], [90], [91], [92], [93], [94], [95], [96], [97], [98], [99], [100], [101], [102], [103], [104]; and 40 between 2000 and October of 2005 [105], [106], [107], [108], [109], [110], [111], [112], [113], [114], [115], [116], [117], [118], [119], [120], [121], [122], [123], [124], [125], [126], [127], [128], [129], [130], [131], [132], [133], [134], [135], [136], [137], [138], [139], [140], [141], [142], [143], [144].

2.2. Description of studies 

The characteristics of the 118 studies evaluating predictors of long-term survival (>6 months) in patients with cirrhosis are summarized in Table 1. Almost two-thirds of the studies were explanatory, only 58% were prospective and 51% included consecutive patients. The most commonly met quality criteria included the reporting of inclusion and exclusion criteria, diagnosis of cirrhosis, number of deaths, length of follow-up, and inclusion of relevant prognostic variables. However, only a minority of studies fulfilled several other important quality criteria such as inclusion of an inception cohort (14%), reporting of missing data (17%) and validation (internal or external) (15%). Also, the adequate proportion of deaths/variables (>10) was only fulfilled in 37% while the rest demonstrated overfitting (Fig. 5). Only one study fulfilled all quality criteria [39]. Thirty-one studies met our criteria for a ‘good’ quality study [32], [37], [39], [47], [55], [57], [60], [62], [63], [66], [76], [81], [85], [89], [92], [95], [96], [97], [100], [103], [104], [112], [116], [117], [123], [124], [128], [130], [138], [142], [143].

  • View full-size image.
  • Fig. 5. 

    Number of variables included in the multivariable analysis as related to the number of observed deaths in 114 studies reporting this information. Each study is represented by a dot. The diagonal line represents the points where the ratio of deaths:variables is equal to 10. In the studies above the diagonal line, the ratio is <10 and there is a considerable risk of overfitting the data, and therefore of false-positive results. In studies below the diagonal line, the risk of overfitting is acceptably low.

2.3. Description of patients included in prognostic studies 

Overall, 23,797 patients were included in 118 studies. Their characteristics are reported in Table 2. The median number of patients per study was 129 but the range was wide (from 23 to 1217). The median age was 54 and two-thirds were male. Almost half of the patients had an alcoholic etiology of cirrhosis and, although not reported in all studies, only a minority (29%) corresponded to Child(–Pugh) class A. MELD score was only reported in nine studies with a median score of 12.

Table 2. Characteristics of cirrhotic patients included in 118 studies evaluating predictors of mortality
VariableN of studies with available informationMedianRange (IQRa)
Sample size (patients included/study)11812923–1217
Age1115444–67
% Male10867%23–100
Etiology of cirrhosis (All)115
Alcohol10946%0–100 (22–71)
Hepatitis C4735%0–100 (22–68)
Hepatitis B6611%0–100 (1–23)
Otherb7612%0–100 (0–21)
Cirrhosis stage41
Compensated1844%
Decompensated2356%
Child(–Pugh) score3385.3–10
Child(–Pugh) class
A7329%0–100 (18–50)
B6838%0–68 (30–46)
C7128%0–79 (14–39)
MELD score9127–18
Varices present2978%0–100 (46–100)
Followup period (months)11131.06–168 (15–60)
Mortality11436%11–100 (23–54)
Causes of death
HCC585%0–25 (1–9)
Variceal hemorrhage6810%0–40 (5–13)
Liver failurec7417%4–48 (11–25)
OLT1218%5–36 (8–27)
Median survival time (months)32331–186 (21–63)
One-year cumulative survival6478%24–100 (62–92)
Child(-Pugh) class A2395%75–100 (90–98)
Child(-Pugh) class B2080%63–99 (73–88)
Child(-Pugh) class C2245%15–95 (33–68)
Compensated2495%62–100 (91–98)
Decompensated2661%26–85 (56–70)
Two-year cumulative survival4275%44–100 (53–90)
Child(-Pugh) class A2090%70–100 (85–96)
Child(-Pugh) class B1670%50–96 (58–82)
Child(-Pugh) class C1838%8–90 (21–52)
Compensated1990%70–100 (87–97)
Decompensated1654%45–77 (48–72)
Final cumulative survival11461%0–89 (40–75)
N variables considered†100152–162 (10–21)
N deaths114496–486 (26–82)
N variables entered on multivariable analysis10492–46 (6–12)
N variables independently predictive of death11330–12 (2–4)

aInterquartile range.

bIncludes cryptogenic, non-alcoholic steatohepatitis, autoimmune, PBC, PSC.

cIncludes encephalopathy, hepatorenal syndrome, sepsis.

†Total number of assessed variables

2.4. Survival in prognostic studies 

As shown in Table 1, median followup time was 31 months with a median mortality rate of 36%. Median survival time, reported in only 32 studies, was 33 months. One-year, 2-year and final cumulative survivals were 78, 75 and 61%, respectively. One- and two-year survival times by Child class and by stage (compensated vs. decompensated) are depicted in Fig. 6, Fig. 7, respectively. The most commonly reported cause of death was liver failure (including hepatorenal syndrome and sepsis), followed by variceal hemorrhage and HCC.

  • View full-size image.
  • Fig. 6. 

    Box plots of one and two-year survival rates in Child–Pugh (or variants) class A, B and C in the studies reporting this information. The number of studies providing the relevant information is reported in Table 2.

2.5. Prognostic variables 

A total of 174 different variables were evaluated in these studies. Table 3 shows the grouping of these variables according to their purported clinical and pathophysiological role in cirrhosis. As shown in Table 4 the variable that was found to be the most common independent predictor of death was the Child(–Pugh) score, having been introduced in a multivariable analysis in 67 studies and having been among the first five significant predictors in 42 (63%) of them. This was followed by all components of the Child–Pugh score (albumin, bilirubin, ascites, encephalopathy, prothrombin time). Age is the only variable found to be predictive of survival in more than 10 studies that is not part of the Child–Pugh score. Variables found to be independently predictive of survival in at least one study are shown in Table 5 and of these, HVPG, MELD and the presence of HCC are remarkable because they were found to be predictive of death in over two-thirds of the 8–9 studies in which they were evaluated. Almost half of the variables evaluated were not significant in any study (Table 6) and remarkably ALT has been non-predictive of death in 31 studies in which it was evaluated. When restricting the analysis to the 31 studies that met criteria for ‘good’ quality (Table 7), the same most common prognostic variables were confirmed, that is, Child–Pugh score or its components and age.

Table 3. Variables (n=174) evaluated in 118 studies
Patient demographics (n=2)
Age, gender
Hepatic insufficiency (n=28)
Child–Pugh score/class, albumin, bilirubin, encephalopathy, prothrombin timea, MELD, MELD change, aminopyrine breath test, pseudocholinesterase, factor V, factor VII, ICG clearance, caffeine clearance, pre-kallikrein, GEC, fibronectin, arachidonic acid, complement, cholesterol, nutrition, midarm circumference, phase angle
Total serum protein, fibrinogen, OGC, bilirubin/GGT ratio, gynecomastia, skinfold thickness
Portal hypertension (n=26)
Ascites, varices, UGI hemorrhage, platelets, HVPG, HVPG change, γ-globulins, spleen size, hypersplenism, PV diameter, PV flow velocity, liver perfusion
Abdominal collaterals, portosystemic shunting (scintigraphy), portosystemic shunting (angiography), hepatic blood flow, portal blood flow, splenic vein diameter, azygos blood flow, post-sinusoidal resistance, hepatic vein O2 saturation, liver O2 extraction, portal vein resistive index, hepatic artery caliber, mesenteric vein caliber, FHVP
Hyperdynamic circulation/volume abnormalities (n=32)
BUN/azotemia, creatinine, serum sodium/hyponatremia, blood pressure, TNF, ICAM, bile acids, vascular spiders, norepinephrine, ADH, plasma renin activity, urinary sodium
Edema, aldosterone, urinary potassium, diuretic use, heart rate, cardiac output, systemic vascular resistance, glomerular filtration rate, palmar erythema, urine volume, urine osmolarity, endotoxin, diastolic pressure, circulation time, plasma volume, renal resistance, plasma osmolarity, renal blood flow, creatinine clearance, right atrial pressure
Liver inflammation/necrosis/fibrosis/cholestasis/histology (n=17)
AST, hyaluronic acid, liver iron, histological score, lactic dehydrogenase, liver size, GGT, nodule size,lymphatic cyst, efferent vessels, calprotectin, alkaline phosphatase, flat Doppler waveform HV
ALT, procollagen III, AST/ALT ratio, laparoscopic abnormalities
Complications related to cirrhosis and their markers (n=32)
HCC, decompensated cirrhosis, SBP/infection, WBC, AFP, PIVKA-II, PV tumoral thrombosis, alcoholic hepatitis, hemoglobin, UGI hemorrhage+PSE, time since UGI bleed, autonomic neuropathy, HPS, Apache score, time in ICU, use of pressors in ICU
Ascites protein, QTc interval, pulmonary pressure, O2 saturation, AaDO2, ascites WBC count, ascites PMN count, number of decompensating events, IgG, IgM, IgA, fibrin split products, hemoccult, opsonins, temperature, constipation
Etiologic factors (n=16)
Alcohol abstinence, alcoholic etiology, viral load, time since diagnosis, MCV, HBsAg, etiology
Hepatitis C etiology, route of viral exposure, HCV genotype, active viral infection, intravenous drug use, CD4 count, duration of HCV infection, anti-HDV, HBeAg
Co-morbidities and other factors (n=18)
T4, smoking, diabetes, serum potassium, HIV
Cholelithiasis, cardiac disease, glycemia, hypertension, any co-morbidity, serum chloride, serum iron, triglycerides, incapacity index, diet, symptoms, asthenia, anorexia
Treatment (n=3)
IFN, endoscopic treatment
Non-specified treatment/prednisolone

Variables in italics are those that were significant on multivariable analysis in at least one study. ICG, indocyanine green; GEC, galactose elimination capacity; OGC, oral glutamine challenge; HVPG, hepatic venous pressure gradient; PV, portal vein; FHVP, free hepatic venous pressure; HV, hepatic vein; SBP, spontaneous bacterial peritonitis; WBC, white blood cell count; PIVKA-II, serum protein induced by vitamin K absence or antagonist II; AFP, alfa-fetoprotein; ICU, intensive care unit; HIV, human immunodeficiency virus; IFN, interferon.

aIncluding prothrombin activity, INR.

Table 4. Variables that were most commonly found to be significant predictors of death in cirrhosis in 118 studies (n=7)
VariableN of studies in which variable was among the first 5 significant variablesN of studies evaluating the variable% of studies in which variable was among first 5/total of studies
CPS/class426763
Albumin317939
Bilirubin298136
Age278034
Ascites246239
Encephalopathy247034
Prothrombin time197525

In more than 10 studies.

Table 5. Variables that were found to be significant in 1–10 studies (n=84)
Variables significant among the first 5 in 2–10 studies divided by the total studies in which the variable was tested (%)Variables significant among the first 5 in only one study divided by the total studies in which the variable was tested (%)Variables significant among the first 5 and tested in only one study
Varices9/36 (25%)Factor V1/2 (50%)Complementa
Gender9/68 (13%)Midarm circumference1/2 (50%)MELD change
UGI hemorrhage8/32 (25%)PIVKA-II1/2 (50%)Caffeine clearance
Platelets7/33 (21%)Portal vein tumoral thrombosis1/2 (50%)Arachidonic acid
MELD6/8 (75%)Viral load1/2 (50%)Fibronectin
HVPG6/9 (67%)Smoking1/2 (50%)Phase angle
HCC6/9 (67%)Histological score1/3 (33%)PV flow velocity
BUN/Azotemia5/13 (38%)Alcoholic hepatitis1/3 (33%)HVPG change
γ-globulins5/23 (22%)Time since diagnosis1/3 (33%)Hypersplenism
Creatinine5/32 (16%)MCV1/3 (33%)Liver perfusion
Alkaline phosphatase5/33 (15%)Portal vein diameter1/4 (25%)ICAM
Aminopyrine BT4/8 (50%)Norepinephrine1/4 (25%)Calprotectin
Pseudocholinesterase4/9 (44%)Lactic dehydrogenase1/4 (25%)Flat Doppler wave HV
Sodium/hyponatremia4/19 (21%)Etiology1/4 (25%)Efferent vessels
AST4/40 (10%)Antidiuretic hormone1/5 (25%)Nodule size
Factor VII3/5 (60%)Diabetes1/5 (20%)Lymphatic cyst
ICG clearance3/11 (27%)Urinary sodium1/6 (17%)UGIH+PSE
Nutrition3/13 (23%)Renin1/7 (14%)Time in ICU
Blood pressure3/13 (23%)AFP1/9 (11%)HIV
Pre-kallikrein2/2 (100%)Potassium1/9 (11%)Apache score
TNF2/2 (100%)Hemoglobin1/11 (9%)Autonomic neuropathy
Hyaluronic acid2/2 (100%)Liver size1/16 (6%)Time since bleed
Liver iron2/2 (100%)HbsAg1/20 (5%)HPS
Decompensated cirrh2/2 (100%)GGT1/25 (4%)Pressors in ICU
T42/2 (100%)Alcoholic etiology1/48 (2%)Endoscopic treatment
Bile acids2/4 (50%)
IFN treatment2/6 (33%)
SBP/infection2/7 (29%)
Galactose elimination2/8 (25%)
WBC2/11 (18%)
Vascular spiders2/12 (17%)
Alcohol abstinence2/12 (17%)
Cholesterol2/13 (15%)
Spleen size2/17 (12%)

aC3, C4, CH50, alternative and classical pathways, PIVK, protein induced by vitamin K or antagonist II, PIVK, serum protein induced by vitamin K absence or antagonist II (PIVKA-II).

Table 6. Non-significant variables (n=83)
VariableN of studies in which the variable was testedVariableN studies in which the variable was testedVariableN studies in which the variable was tested
ALT31Splenic vein diameter2Renal blood flow1
HCV etiology9Azygos blood flow2Renal resistance1
Treatmenta9Plasma volume2Creatinine clearance1
Edema8Urinary osmolarity2R atrial pressure1
Aldosterone5Diastolic pressure2Laparoscopic abn1
Urine potassium5AST/ALT ratio2Endotoxin1
Diuretic use5Procollagen III2Opsonins1
Cholelithiasis5Ascites protein2Ascites WBC1
Abdo collaterals4Pulmonary pressure2Ascites PMN1
Heart rate4O2 saturation2FSP1
Total protein3QTc interval2IgG1
Skinfold thickness3Constipation2IgA1
PSS (scintigraphy)b3HCV genotype2IgM1
Cardiac output3Viral exposure route2CD41
SVR3Serum chloride2Co-morbidity1
GFR3Anorexia2N decompensating events1
Palmar erythema3OGC1AaDO21
Urine volume3Hepatic artery caliber1Temperature1
Cardiac disease3Mesenteric vein caliber1Hemoccult1
Active infection3PSS (angiography)1IVDU1
Anti-HDV3Portal blood flow1Duration of HCV inf1
HbeAg3PV resistive index1Incapacity index1
Asthenia3Post-sinusoidal resistance1Diet1
Glycemia3Hepatic vein O2 sat1Triglycerides1
Fibrinogen2Liver O2 extraction1Serum iron1
Bilirubin/GGT rat2FHVP1Symptoms1
Gynecomastia2Circulation time1Hypertension1
Hepatic blood flow2Plasma osmolarity1

OGC, oral glutamine clearance.

aNon-specified or prednisolone.

bBy scintigraphy.

Table 7. Variables that were most commonly found to be significant predictors of death assessed in 31 ‘good’ studies
VariableN of good studies in which variable was among first 5 significant onesN of good studies evaluating variable% of studies in which variable was among first 5/total of studies
CPS/class132065
Bilirubin112348
Albumin112348
Age112839
Prothrombin time82138
Encephalopathy71450
Ascites41429
Gender52322
BUN3475
Platelets31030

When the analysis is performed separately for studies that included only compensated or only decompensated cirrhotic patients, the most common prognostic variables in each group are different (Table 8), with variables related to portal hypertension (platelet count, varices, spleen size) appearing in the compensated group, and variables related to renal insufficiency or HCC appearing in the decompensated group.

Table 8. Variables significant in studies including or analyzing compensated (n=18) and decompensated cirrhosis (n=23) separately
Compensated cirrhosis (n=18 studies)Decompensated cirrhosis (n=23 studies)
VariableN of sign studiesN of studies eval%VariableN of sign studiesN of studies eval%
Age101471Child–Pugh91656
Albumin101759Encephalopathy61250
Bilirubin81553HCC4666
Platelet count61346Bleeding41136
Gender41429Creatinine41527
Prothrombin time31030Prothrombin time41625
Child–Pugh22100Albumin41625
Vascular spiders22100BUN/Azotemia3933
Spleen size2633Ascites31127
Varices2729Bilirubin31718
γ-globulins2729Age31817

Back to Article Outline

3. Discussion 

Prognosis is an essential part of the baseline assessment of any disease. It is not only the basis for the information that a physician provides to the patient, but is also the basis for any decision-making process. However, the applicability of the available prognostic information to the individual patient is frequently unsatisfactory because individual patients are different from the average patient population on which a prognostic estimation is based. Apart from the individual biologic variability, a very important source of this diversity is the specific disease stage of the individual patient. In order to provide prognostic information, a physician needs to recognize in his/her patient a specific pattern of clinical characteristics from patients included in predictive studies. The bulk of the methodology for prognostic studies has been developed from these concepts and its application to cirrhosis has been discussed in detail elsewhere [13], [145] and recommendations on how to perform and how to interpret prognostic studies have been developed [14], [146], [147], [148]. We have based this systematic review on these recommendations as indicated by the evaluation of quality items reported in Table 1.

A remarkable finding of this systematic review is the very large number of prognostic studies of cirrhosis performed in 23 years, at least 118 studies including a total of 23,797 patients. The fact that the number of studies has been increasing over time (40 in the past 5 years) indicates uncertainty and dissatisfaction with available prognostic models. Methodological problems in these studies leading poor reproducibility may partly explain this dissatisfaction. In fact, only 1 of the 117 reviewed studies fulfilled all the considered quality criteria and only 17 provided a validation of the results, of which only five based the validation on an independent patient sample. The major problems, however, lie in the inclusion of non-consecutive patients, incomplete reporting of inclusion/exclusion criteria, incompleteness of follow-up and the inclusion of patients at differing disease stages without a separate analysis. These deficiencies, particularly the inclusion in the same study of patients at different disease stages, lead to irreparable and yet non-measurable biases and render the results non-applicable to clinical practice.

In addition to the clinical aspects of methodology, the statistical methodology plays a key role in the assessment of predictors of survival. In the majority of the studies the number of variables included in the final analysis did not allow for controlling the risk of overfitting the data, i.e. the ratio of the number of deaths to the number of variables was <10:1 [147], [149], [150] (Fig. 5). Overfitting of data translates in a high risk of false positive results, which may well be a further explanation of the low reproducibility of the results for most studies.

A second major consideration from this systematic review is the wide range of reported survival rates (Table 2). While the variability of the final survival was obviously expected on the basis of the wide range of observation times across the studies, the variability of survival at fixed follow-up times (1 and 2 years, Table 2) should be interpreted, at least in part, as an expression of the heterogeneity of the included patients. In fact, this variability was appreciably reduced when the survival rates were assessed by Child(–Pugh) class or by stage (decompensated vs. decompensated) (Fig. 6, Fig. 7). It is notable that, while there is obvious overlap among survival times across the different Child–Pugh classes (Fig. 6), this overlap is almost negligible when analyzed by compensation stage (Fig. 7). However, even considering the compensated (or decompensated) stage of cirrhosis, there is still an appreciable variability of the 1 or 2-years survival rate across studies. This residual variability may be further explained, at least in part, by the inclusion of patients at different times along the course of their compensated or decompensated disease stage. Although somewhat intriguing, this concept may be easily understood if one considers the differences between a patient with compensated cirrhosis and different degrees of portal hypertension (e.g. with and without varices) and if one considers a patient with decompensated cirrhosis with ascites at different time intervals since the development of ascites (e.g. newly developed vs. ascites developed 2 years before). The prognosis of the compensated patient without varices will be better than that of the patient with varices and the prognosis of the decompensated patient with newly diagnosed ascites will be better than the one from ascites that developed two years ago. This stresses the importance of including inception cohorts in prognostic studies. Unfortunately this inclusion criterion was met in only 17 of the 117 reviewed studies.

However, despite methodological problems in the evaluated studies, this systematic review allowed for the identification of ‘robust’ predictors of death in cirrhosis. Not only was this evident in the total number of studies in which these variables were significant, but also by calculating a ratio between the number of studies in which each variable was significant and the number of studies in which it was assessed. With larger number of studies, a large ratio is an indirect measure of validity as each study that confirms the predictive value of a variable, provides indirect proof of its validity. This robustness is, therefore, independent of the quality of the studies. Of note, we only report on predictors that were statistically significant at the first five levels of multivariable analysis, as this yields more reproducible and powerful results by omitting variables that may be statistically significant at a lower level [15], [16].

By far, the most consistent and ‘robust’ predictor of death in cirrhosis is the Child–Pugh score and/or its components (albumin, bilirubin, ascites, encephalopathy and prothrombin time). This was the variable most commonly introduced in multivariable analyses in the 118 studies (57%) and the one most commonly found to be an independent predictor of survival (63% of the time): importantly, exactly the same parameters (Child–Pugh score, bilirubin, albumin, prothrombin time, encephalopathy and ascites) were the most frequent significant variables when only the ‘good’ quality studies were analyzed. A first important conclusion from this review is that the Child–Pugh score and/or its parameters as well as another important and powerful predictor of death, the patient's age, must be included in any future prognostic studies in cirrhosis, independent of disease stage. This includes studies in which parameters, such as the MELD score and the HVPG, that have been less frequently studied but that nevertheless were found to be predictive of death in over two-thirds of a reasonable number of studies, should be evaluated. On the other hand, our results also demonstrate that other parameters, such as the ALT, should no longer be incorporated into prognostic models as it has been consistently and repeatedly found to be non-predictive of death.

Another important finding of this large systematic review is that, as hypothesized and as shown earlier [151], prognostic markers differ depending on the stage of cirrhosis. In studies that included only patients with compensated cirrhosis, the Child(–Pugh) score was still among the most frequent significant predictors of death despite the absence of ascites, encephalopathy and jaundice, because its laboratory components, bilirubin, albumin and prothrombin time continued to be among the most frequent predictors, indicating that even subtle abnormalities in these laboratory parameters are predictive of death. In addition to these markers of liver insufficiency, in the compensated stage, significant predictors that come to light are those related to portal hypertension, such as the presence of varices, splenomegaly and platelet count as well as γ-globulin levels (as hyper γ-globulinemia is an indirect marker of portosystemic shunting). This probably indicates that, in a compensated stage, measurements of portal pressure will be of important prognostic value. This is strengthened by a recent study that showed that in patients with no varices and no ascites (stage 1 of our classification system) the most important predictor of the development of varices was an HVPG of >10mmHg [152]. Conversely, the set of significant prognostic variables in the group of patients with decompensated cirrhosis reflect a more advanced stage, as bleeding and HCC become predictive of death. It is in this group that the Child–Pugh score (and its components) has the most important prognostic value. In addition to the Child–Pugh score, parameters that reflect a further deterioration of the circulatory status of the cirrhotic patient, such as parameters of renal dysfunction (creatinine and blood urea nitrogen/azotemia) arise as powerful prognostic indicators in this setting and, therefore, it is not surprising that the MELD score (which incorporates creatinine in addition to markers of liver dysfunction) has become a valuable method to allocate organs. On the contrary, it is predictable that the MELD score would not be useful to predict survival in patients with compensated cirrhosis.

Lessons learned from this large systematic review should allow us to make suggestions for future studies on prognostic indicators of cirrhosis so that the information can be used in clinical practice. Regarding methodological issues, it will be of major importance to (a) include patients at a well-defined stage in the course of cirrhosis (inception cohort) thereby homogenizing the study population and (b) to avoid the error of overfitting, observed in a significant proportion of studies. From the clinical point of view, it will be important to assess prognostic variables separately for the different stages of cirrhosis, at a minimum, separating those with compensated and those with decompensated cirrhosis. Better still, and as recently concluded in the Baveno IV consensus conference, prognostic indicators should be targeted at the four specific subgroups of patients with cirrhosis that have a different risk of dying (Fig. 4). In patients with decompensated cirrhosis, any study of predictors of death should include important variables identified by the majority of studies, such as the Child–Pugh score (or its components) and age. In patients with compensated cirrhosis (or status 1 and 2), particularly in those who remain at a compensated stage, the risk of dying is low and in this group of patients it would be more useful to look at predictors of decompensation rather than at predictors of mortality.

In clinical practice, our results strengthen the current use of the Child–Pugh score in cirrhotic patients at large and the use of the MELD score in decompensated cirrhotic patients.

Back to Article Outline

References 

  1. Child CG, Turcotte JG. Surgery and portal hypertension. Major Probl Clin Surg. 1964;1:1–85
  2. Kamath PS, Wiesner RH, Malinchoc M, Kremers W, Therneau TM, Kosberg CL, et al. A model to predict survival in patients with end-stage liver disease. Hepatology. 2001;33:464–470
  3. Durand F, Valla D. Assessment of the prognosis of cirrhosis: Child-Pugh versus MELD. J Hepatol. 2005;42:S100–S107
  4. D'Amico G, Garcia–Tsao G, Abraldes J, Ginès P, Kim WR, Schepis F, et al. Prognostic indicators in cirrhosis. In: de Franchis R, editor. Portal Hypertension IV. Proceedings of the Fourth Baveno International Consensus Workshop on Definitions, Methodology and Therapeutic Strategies. Blackwell, 2005.
  5. D'Amico G, Garcia-Tsao G, Cales P, Escorsell A, Nevens F, Cestari R, et al. Diagnosis of portal hypertension: how and when. In:  DeFranchis R editors. Portal hypertension III. Proceedings of the third baveno international consensus workshop on definitions, methodology and therapeutic strategies. Oxford: Blackwell Science; 2001;p. 36–64
  6. Saunders JB, Walters JRF, Davies P, Paton A. A 20-year prospective study of cirrhosis. Br Med J. 1981;282:263–266
  7. Gines P, Quintero E, Arroyo V. Compensated cirrhosis: natural history and prognosis. Hepatology. 1987;7:122–128
  8. D'Amico G, Morabito A, Pagliaro L, Marubini E. Survival and prognostic indicators in compensated and decompensated cirrhosis. Dig Dis Sci. 1986;31:468–475
  9. D'Amico G. Natural history of compensated cirrhosis and varices. In:  Boyer TD,  Groszmann RJ editor. Complications of cirrhosis: pathogenesis, consequences and therapy. American Association for the Study of Liver Diseases; 2001;p. 118–123[Course Directors]
  10. D'Amico G, Pasta L, Madonia S, Tarantino G, Mancuso A, Malizia G, et al. The incidence of esophageal varices in cirrhosis. Gastroenterology. 2001;120:A2
  11. de Franchis R. Evolving consensus in portal hypertension report of the baveno IV consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol. 2005;43:167–176
  12. Infante-Rivard C, Esnaola S, Villeneuve JP. Clinical and statistical validity of conventional prognostic factors in predicting short-term survival among cirrhotics. Hepatology. 1987;7:660–664
  13. Chianciano consensus conference on prognostic studies in hepatology. Ital J Gastroenterol Hepatol 1998; 30: 580-3.
  14. Oxford Centre for Evidence Based Medicine. Levels of Evidence and Grades of Recommendation; 2001 [Ref Type: Internet Communication].
  15. Wells CK, Feinstein AR, Walter SD. A comparison of multivariable mathematical methods for predicting survival–III. Accuracy of predictions in generating and challenge sets. J Clin Epidemiol. 1990;43:361–372
  16. Heckerling PS, Conant RC, Tape TG, Wigton RS. Reproducibility of predictor variables from a validated clinical rule. Med Decis Making. 1992;12:280–285
  17. Buligescu L, Voiculescu M, Costin A, Draghici I. Study of the prognosis factors in liver cirrhosis. Med Interne. 1987;25:157–165
  18. Casaril M, Micciolo R, Gabrielli GB, Bellisola G, Corrocher R. Prognostic score in liver cirrhosis developed using the Cox's proportional hazard regression model. Ric Clin Lab. 1987;17:67–76
  19. Adler M, Bourgeois N, Van de SJ, Gelin M. A Pugh score of 8 adequately selects patients with parenchymal cirrhosis for liver transplantation. Transpl Int. 1992;5:S175–S178
  20. Altman C, Grange JD, Amiot X, Pelletier G, Lacaine F, Bodin F, et al. Survival after a first episode of spontaneous bacterial peritonitis. Prognosis of potential candidates for orthotopic liver transplantation. J Gastroenterol Hepatol. 1995;10:47–50
  21. Habib A, Mihas AA, Abou-Assi SG, Williams LM, Gavis E, Pandak WM, et al. High-density lipoprotein cholesterol as an indicator of liver function and prognosis in noncholestatic cirrhotics. Clin Gastroenterol Hepatol. 2005;3:286–291
  22. Ekindjian OG, Devanlay M, Duchassaing D, Leluan G, Kammerer J, Fouet P, et al. Multivariate analysis of clinical and biological data in cirrhotic patients: application to prognosis. Eur J Clin Invest. 1981;11:213–220
  23. Rossi L, Milani A, Marra L, Siciliano M. Grading scores and survivorship functions in liver cirrhosis: a comparative statistical analysis of various predictive models. Hepatogastroenterology. 1986;33:240–243
  24. Pomier-Layrargues G, Huet PM, Infante-Rivard C, Villeneuve JP, Marleau D, Duguay L, et al. Prognostic value of indocyanine green and lidocaine kinetics for survival and chronic hepatic encephalopathy in cirrhotic patients following elective end-to-side portacaval shunt. Hepatology. 1988;8:1506–1510
  25. Thomsen BL, Moller S, Sorensen TI. Optimized analysis of recurrent bleeding and death in patients with cirrhosis and esophageal varices. Copenhagen esophageal varices sclerotherapy project. J Hepatol. 1994;21:367–375
  26. Shiomi S, Kuroki T, Ueda T, Takeda T, Ikeoka N, Nishiguchi S, et al. Clinical usefulness of evaluation of portal circulation by per rectal portal scintigraphy with technetium-99m pertechnetate. Am J Gastroenterol. 1995;90:460–465
  27. Borroni G, Maggi A, Sangiovanni A, Cazzaniga M, Salerno F. Clinical relevance of hyponatraemia for the hospital outcome of cirrhotic patients. Dig Liver Dis. 2000;32:605–610
  28. Zhao C, Chen SB, Zhou JP, Xiao W, Fan HG, Wu XW, et al. Prognosis of hepatic cirrhosis patients with esophageal or gastric variceal hemorrhage: multivariate analysis. Hepatobiliary Pancreat Dis Int. 2002;1:416–419
  29. Degre D, Bourgeois N, Boon N, Le Moine O, Louis H, Donckier V, et al. Aminopyrine breath test compared to the MELD and Child-Pugh scores for predicting mortality among cirrhotic patients awaiting liver transplantation. Transpl Int. 2004;17:31–38
  30. Orrego H, Israel Y, Blake JE, Medline A. Assessment of prognostic factors in alcoholic liver disease: toward a global quantitative expression of severity. Hepatology. 1983;3:896–905
  31. Schlichting P, Christensen E, Andersen PK, Fauerholdt L, Juhl E, Poulsen H, et al. Prognostic factors in cirrhosis identified by Cox's regression model. Hepatology. 1983;3:889–895
  32. Christensen E, Schlichting P, Fauerholdt L, Gluud C, Andersen PK, Juhl E, et al. Prognostic value of Child-Turcotte criteria in medically treated cirrhosis. Hepatology. 1984;4:430–435
  33. Poynard T, Zourabichvili O, Hilpert G, Naveau S, Poitrine A, Benatar C, et al. Prognostic value of total serum bilirubin/gamma-glutamyl transpeptidase ratio in cirrhotic patients. Hepatology. 1984;4:324–327
  34. Milani A, Marra L, Siciliano M, Rossi L. Prognostic significance of clinical and laboratory parameters in liver cirrhosis. A multivariate statistical approach. Hepatogastroenterology. 1985;32:270–272
  35. Naveau S, Poynard T, Abella A, Pignon JP, Poitrine A, Agostini H, et al. Prognostic value of serum fibronectin concentration in alcoholic cirrhotic patients. Hepatology. 1985;5:819–823
  36. Christensen E, Schlichting P, Andersen PK, Fauerholdt L, Schou G, Pedersen BV, et al. Updating prognosis and therapeutic effect evaluation in cirrhosis with Cox's multiple regression model for time-dependent variables. Scand J Gastroenterol. 1986;21:163–174
  37. Mannes GA, Thieme C, Stellaard F, Wang T, Sauerbruch T, Paumgartner G. Prognostic significance of serum bile acids in cirrhosis. Hepatology. 1986;6:50–53
  38. Villeneuve JP, Infante-Rivard C, Ampelas M, Pomier-Layrargues G, Huet PM, Marleau D. Prognostic value of the aminopyrine breath test in cirrhotic patients. Hepatology. 1986;6:928–931
  39. Gines P, Quintero E, Arroyo V, Teres J, Bruguera M, Rimola A, et al. Compensated cirrhosis: natural history and prognostic factors. Hepatology. 1987;7:122–128
  40. Tanaka R, Itoshima T, Nagashima H. Follow-up study of 582 liver cirrhosis patients for 26 years in Japan. Liver. 1987;7:316–324
  41. Tsuji Y, Koga S, Ibayashi H, Nose Y, Akazawa K. Prediction of the prognosis of liver cirrhosis in Japanese using Cox's proportional hazard model. Gastroenterol Jpn. 1987;22:599–606
  42. Llach J, Gines P, Arroyo V, Rimola A, Tito L, Badalamenti S, et al. Prognostic value of arterial pressure, endogenous vasoactive systems, and renal function in cirrhotic patients admitted to the hospital for the treatment of ascites. Gastroenterology. 1988;94:482–487
  43. Gluud C, Henriksen JH, Nielsen G. Tthe copenhagen study group for liver disease. Prognostic indicators in alcoholic cirrhotic men. Hepatology. 1988;8:222–227
  44. Tage-Jensen U, Henriksen JH, Christensen E, Widding A, Ring-Larsen H, Christensen NJ. Plasma catecholamine level and portal venous pressure as guides to prognosis in patients with cirrhosis. J Hepatol. 1988;6(3):350–358
  45. Albers I, Hartmann H, Bircher J, Creutzfeldt W. Superiority of the Child-Pugh classification to quantitative liver function tests for assessing prognosis of liver cirrhosis. Scand J Gastroenterol. 1989;24:269–276
  46. Christensen E, Krintel JJ, Hansen SM, Johansen JK, Juhl E. Prognosis after the first episode of gastrointestinal bleeding or coma in cirrhosis. Survival and prognostic factors. Scand J Gastroenterol. 1989;24:999–1006
  47. Wang JY, Liu HY, Zhu WN, Han JQ, Dong CY, Shi YJ, et al. An analysis of prognostic factors in cirrhosis. Chin Med J (Engl). 1989;102:266–272
  48. Merkel C, Bolognesi M, Angeli P, Noventa F, Caregaro L, Sacerdoti D, et al. Prognostic indicators of survival in patients with cirrhosis and esophageal varices, without previous bleeding. Am J Gastroenterol. 1989;84:717–722
  49. Merkel C, Bolognesi M, Finucci GF, Angeli P, Caregaro L, Rondana M, et al. Indocyanine green intrinsic hepatic clearance as a prognostic index of survival in patients with cirrhosis. J Hepatol. 1989;9:16–22
  50. Adler M, Van Laethem J, Glibert A, Gelin M, Bourgeois N, Vereerstraeten P, et al. Factors influencing survival at one year in patients with nonbiliary hepatic parenchymal cirrhosis. Dig Dis Sci. 1990;35:1–5
  51. Tameda Y, Yoshizawa N, Takase K, Nakano T, Kosaka Y. Prognostic value of peritoneoscopic findings in cirrhosis of the liver. Gastrointest Endosc. 1990;36:34–38
  52. Bayerdorffer E, Lamerz R, Fliege R, Kopcke W, Mannes GA. Predictive value of serum procollagen-III-peptide for the survival of patients with cirrhosis. J Hepatol. 1991;13:298–304
  53. Finucci G, Bellon S, Merkel C, Mormino P, Tirelli M, Gatta A, et al. Evaluation of splanchnic angiography as a prognostic index of survival in patients with cirrhosis. Scand J Gastroenterol. 1991;26:951–960
  54. Kleber G, Sauerbruch T, Ansari H, Paumgartner G. Prediction of variceal hemorrhage in cirrhosis: a prospective follow-up study. Gastroenterology. 1991;100:1332–1337
  55. Merkel C, Gatta A, Zoli M, Bolognesi M, Angeli P, Iervese T, et al. Prognostic value of galactose elimination capacity, aminopyrine breath test, and ICG clearance in patients with cirrhosis. Comparison with the Pugh score. Dig Dis Sci. 1991;36:1197–1203
  56. Oellerich M, Burdelski M, Lautz HU, Binder L, Pichlmayr R. Predictors of one-year pretransplant survival in patients with cirrhosis. Hepatology. 1991;14:1029–1034
  57. Zoli M, Cordiani MR, Marchesini G, Iervese T, Labate AM, Bonazzi C, et al. Prognostic indicators in compensated cirrhosis. Am J Gastroenterol. 1991;86:1508–1513
  58. de Jongh FE, Janssen HL, de Man RA, Hop WC, Schalm SW, van Blankenstein M. Survival and prognostic indicators in hepatitis B surface antigen-positive cirrhosis of the liver. Gastroenterology. 1992;103:1630–1635
  59. Ferro D, Saliola M, Quintarelli C, Alessandri C, Basili S, Cordova C, et al. 1-year survey of patients with advanced liver cirrhosis. Prognostic value of clinical and laboratory indexes identified by the Cox regression model. Scand J Gastroenterol. 1992;27:852–856
  60. McCormick PA, Morgan MY, Phillips A, Yin TP, McIntyre N, Burroughs AK. The effects of alcohol use on rebleeding and mortality in patients with alcoholic cirrhosis following variceal haemorrhage. J Hepatol. 1992;14:99–103
  61. Merkel C, Bolognesi M, Bellon S, Zuin R, Noventa F, Finucci G, et al. Prognostic usefulness of hepatic vein catheterization in patients with cirrhosis and esophageal varices. Gastroenterology. 1992;102:973–979
  62. Merkel C, Bolognesi M, Bellon S, Bianco S, Honisch B, Lampe H, et al. Aminopyrine breath test in the prognostic evaluation of patients with cirrhosis. Gut. 1992;33:836–842
  63. Abad-Lacruz A, Cabre E, Gonzalez-Huix F, Fernandez-Banares F, Esteve M, Planas R, et al. Routine tests of renal function, alcoholism, and nutrition improve the prognostic accuracy of Child-Pugh score in nonbleeding advanced cirrhotics. Am J Gastroenterol. 1993;88:382–387
  64. Cabre E, Abad-Lacruz A, Nunez MC, Gonzalez-Huix F, Fernandez-Banares F, Gil A, et al. The relationship of plasma polyunsaturated fatty acid deficiency with survival in advanced liver cirrhosis: multivariate analysis. Am J Gastroenterol. 1993;88:718–722
  65. Oellerich M, Hartmann H, Ringe B, Burdelski M, Lautz HU, Pichlmayr R. Assessment of prognosis in transplant candidates by use of the Pugh-MEGX score. Transplant Proc. 1993;25:1116–1119
  66. Salerno F, Borroni G, Moser P, Badalamenti S, Cassara L, Maggi M, et al. Survival and prognostic factors of cirrhotic patients with ascites: a study of 134 outpatients. Am J Gastroenterol. 1993;88:514–519
  67. Zoli M, Iervese T, Merkel C, Bianchi G, Magalotti D, Marchesini G, et al. Prognostic significance of portal hemodynamics in patients with compensated cirrhosis. J Hepatol. 1993;17:56–61
  68. Arrigoni A, Gindro T, Aimo G, Cappello N, Meloni A, Benedetti P, et al. Monoethylglicinexylidide test: a prognostic indicator of survival in cirrhosis. Hepatology. 1994;20:383–387
  69. Bianchi G, Marchesini G, Zoli M, Bugianesi E, Fabbri A, Pisi E. Prognostic significance of diabetes in patients with cirrhosis. Hepatology. 1994;20:119–125
  70. Maroto A, Gines A, Salo J, Claria J, Gines P, Anibarro L, et al. Diagnosis of functional kidney failure of cirrhosis with Doppler sonography: prognostic value of resistive index. Hepatology. 1994;20:839–844
  71. Moller S, Bendtsen F, Christensen E, Henriksen JH. Prognostic variables in patients with cirrhosis and oesophageal varices without prior bleeding. J Hepatol. 1994;21:940–946
  72. Realdi G, Fattovich G, Hadziyannis S, Schalm SW, Almasio P, Sanchez-Tapias J, et al. Survival and prognostic factors in 366 patients with compensated cirrhosis type B: a multicenter study. The investigators of the European concerted action on Viral Hepatitis (EUROHEP). J Hepatol. 1994;21:656–666
  73. Urbain D, Muls V, Makhoul E, Ham HR. Prognostic value of thallium-201 per rectum scintigraphy in alcoholic cirrhosis. J Nucl Med. 1994;35:832–834
  74. Sugimura T, Tsuji Y, Sakamoto M, Kotoh K, Fukutomi T, Sakai H, et al. Long-term prognosis and prognostic factors of liver cirrhosis in the 1980s. J Gastroenterol Hepatol. 1994;9:154–161
  75. Homann C, Garred P, Graudal N, Hasselqvist P, Christiansen M, Fagerhol MK, et al. Plasma calprotectin: a new prognostic marker of survival in alcohol-induced cirrhosis. Hepatology. 1995;21:979–985
  76. Ohta M, Hashizume M, Kawanaka H, Akazawa K, Tomikawa M, Higashi H, et al. Prognostic significance of hepatic vein waveform by Doppler ultrasonography in cirrhotic patients with portal hypertension. Am J Gastroenterol. 1995;90:1853–1857
  77. Urbain D, Muls V, Thys O, Ham HR. Aminopyrine breath test improves long-term prognostic evaluation in patients with alcoholic cirrhosis in Child classes A and B. J Hepatol. 1995;22:179–183
  78. Violi F, Ferro D, Basili S, Cimminiello C, Saliola M, Vezza E, et al. Prognostic value of clotting and fibrinolytic systems in a follow-up of 165 liver cirrhotic patients. CALC Group. Hepatology. 1995;22:96–100
  79. Merkel C, Marchesini G, Fabbri A, Bianco S, Bianchi G, Enzo E, et al. The course of galactose elimination capacity in patients with alcoholic cirrhosis: possible use as a surrogate marker for death. Hepatology. 1996;24:820–823
  80. Merli M, Riggio O, Dally L. Does malnutrition affect survival in cirrhosis? PINC (Policentrica Italiana Nutrizione Cirrosi). Hepatology. 1996;23:1041–1046
  81. Salerno F, Borroni G, Moser P, Sangiovanni A, Almasio P, Budillon G, et al. Prognostic value of the galactose test in predicting survival of patients with cirrhosis evaluated for liver transplantation. A prospective multicenter Italian study. AISF Group for the study of liver transplantation. Associazione Italiana per lo Studio del Fegato. J Hepatol. 1996;25:474–480
  82. Vorobioff J, Groszmann RJ, Picabea E, Gamen M, Villavicencio R, Bordato J, et al. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis: a 10-year prospective study. Gastroenterology. 1996;111:701–709
  83. Adler M, Verset D, Bouhdid H, Bourgeois N, Gulbis B, Le Moine O, et al. Prognostic evaluation of patients with parenchymal cirrhosis. Proposal of a new simple score. J Hepatol. 1997;26:642–649
  84. Chan CC, Hwang SJ, Lee FY, Wang SS, Chang FY, Li CP, et al. Prognostic value of plasma endotoxin levels in patients with cirrhosis. Scand J Gastroenterol. 1997;32:942–946
  85. Fattovich G, Giustina G, Degos F, Tremolada F, Diodati G, Almasio P, et al. Morbidity and mortality in compensated cirrhosis type C: a retrospective follow-up study of 384 patients. Gastroenterology. 1997;112:463–472
  86. Gentilini P, Laffi G, La Villa G, Romanelli RG, Buzzelli G, Casini-Raggi V, et al. Long course and prognostic factors of virus-induced cirrhosis of the liver. Am J Gastroenterol. 1997;92:66–72
  87. Homann C, Varming K, Hogasen K, Mollnes TE, Graudal N, Thomsen AC, et al. Acquired C3 deficiency in patients with alcoholic cirrhosis predisposes to infection and increased mortality. Gut. 1997;40:544–549
  88. Jover R, Carnicer F, Sanchez-Paya J, Climent E, Sirvent M, Marco JL. Salivary caffeine clearance predicts survival in patients with liver cirrhosis. Am J Gastroenterol. 1997;92:1905–1908
  89. Magliocchetti N, Torchio P, Corrao G, Arico S, Favilli S. Prognostic factors for long-term survival in cirrhotic patients after the first episode of liver decompensation. Ital J Gastroenterol Hepatol. 1997;29:38–46
  90. Caregaro L, Alberino F, Amodio P, Merkel C, Angeli P, Plebani M, et al. Nutritional and prognostic significance of serum hypothyroxinemia in hospitalized patients with liver cirrhosis. J Hepatol. 1998;28:115–121
  91. Merkel C, Morabito A, Sacerdoti D, Bolognesi M, Angeli P, Gatta A. Updating prognosis of cirrhosis by Cox's regression model using Child-Pugh score and aminopyrine breath test as time-dependent covariates. Ital J Gastroenterol Hepatol. 1998;30:276–282
  92. Serfaty L, Aumaitre H, Chazouilleres O, Bonnand AM, Rosmorduc O, Poupon RE, et al. Determinants of outcome of compensated hepatitis C virus-related cirrhosis. Hepatology. 1998;27:1435–1440
  93. Stanley AJ, Robinson I, Forrest EH, Jones AL, Hayes PC. Haemodynamic parameters predicting variceal haemorrhage and survival in alcoholic cirrhosis. Quaterly J Med. 1998;91:19–25
  94. Amodio P, DelPiccolo F, Marchetti P, Angeli P, Iemmolo R, Caregaro L, et al. Clinical features and survival of cirrhotic patients with subclinical cognitive alterations detected by the number connection test and computerized psychometric tests. Hepatology. 1999;29:1662–1667
  95. Bustamante J, Rimola A, Ventura PJ, Navasa M, Cirera I, Reggiardo V, et al. Prognostic significance of hepatic encephalopathy in patients with cirrhosis. J Hepatol. 1999;30:890–895
  96. Hu KQ, Tong MJ. The long-term outcomes of patients with compensated hepatitis C virus-related cirrhosis and history of parenteral exposure in the United States. Hepatology. 1999;29:1311–1316
  97. Macias-Rodriguez MA, Rendon-Unceta P, Martinez-Sierra MC, Teyssiere-Blas I, Diaz-Garcia F, Martin-Herrera L. Prognostic usefulness of ultrasonographic signs of portal hypertension in patients with child-pugh stage a liver cirrhosis. Am J Gastroenterol. 1999;94:3595–3600
  98. Patch D, Armonis A, Sabin C, Christopoulou K, Greenslade L, McCormick A, et al. Single portal pressure measurement predicts survival in cirrhotic patients with recent bleeding. Gut. 1999;44:264–269
  99. Testa R, Valente U, Risso D, Caglieris S, Giannini E, Fasoli A, et al. Can the MEGX test and serum bile acids improve the prognostic ability of Child-Pugh's score in liver cirrhosis?. Eur J Gastroenterol Hepatol. 1999;11:559–563
  100. Degos F, Christidis C, Ganne-Carrie N, Farmachidi JP, Degott C, Guettier C, et al. Hepatitis C virus related cirrhosis: time to occurrence of hepatocellular carcinoma and death. Gut. 2000;47:131–136
  101. del Olmo JA, Pena A, Serra MA, Wassel AH, Benages A, Rodrigo JM. Predictors of morbidity and mortality after the first episode of upper gastrointestinal bleeding in liver cirrhosis. J Hepatol. 2000;32:19–24
  102. Fattovich G, Giustina G, Christensen E, Pantalena M, Zagni I, Realdi G, et al. Influence of hepatitis delta virus infection on morbidity and mortality in compensated cirrhosis type B. The European concerted action on Viral Hepatitis (Eurohep). Gut. 2000;46:420–426
  103. Ganne-Carrie N, Christidis C, Chastang C, Ziol M, Chapel F, Imbert-Bismut F, et al. Liver iron is predictive of death in alcoholic cirrhosis: a multivariate study of 229 consecutive patients with alcoholic and/or hepatitis C virus cirrhosis: a prospective follow up study. Gut. 2000;46:277–282
  104. Guechot J, Serfaty L, Bonnand AM, Chazouilleres O, Poupon RE, Poupon R. Prognostic value of serum hyaluronan in patients with compensated HCV cirrhosis. J Hepatol. 2000;32:447–452
  105. Alberino F, Gatta A, Amodio P, Merkel C, Di Pascoli L, Boffo G, et al. Nutrition and survival in patients with liver cirrhosis. Nutrition. 2001;17:445–450
  106. Amodio P, Del Piccolo F, Petteno E, Mapelli D, Angeli P, Iemmolo R, et al. Prevalence and prognostic value of quantified electroencephalogram (EEG) alterations in cirrhotic patients. J Hepatol. 2001;35:37–45
  107. Fernandez-Esparrach G, Sanchez-Fueyo A, Gines P, Uriz J, Quinto L, Ventura PJ, et al. A prognostic model for predicting survival in cirrhosis with ascites. J Hepatol. 2001;34:46–52
  108. Fattovich G, Ribero ML, Pantalena M, Diodati G, Almasio P, Nevens F, et al. Hepatitis C virus genotypes: distribution and clinical significance in patients with cirrhosis type C seen at tertiary referral centres in Europe. J Viral Hepat. 2001;8:206–216
  109. Franca AV, De Souza JB, Silva CM, Soares EC. Long-term prognosis of cirrhosis after spontaneous bacterial peritonitis treated with ceftriaxone. J Clin Gastroenterol. 2001;33:295–298
  110. Puthumana L, Chaudhry V, Thuluvath PJ. Prolonged QTc interval and its relationship to autonomic cardiovascular reflexes in patients with cirrhosis. J Hepatol. 2001;35:733–738
  111. Deltenre P, Rufat P, Hillaire S, Elman A, Moreau R, Valla D, et al. Lack of prognostic usefulness of hepatic venous pressures and hemodynamic values in a select group of patients with severe alcoholic cirrhosis. Am J Gastroenterol. 2002;97:1187–1190
  112. Fattovich G, Pantalena M, Zagni I, Realdi G, Schalm SW, Christensen E. Effect of hepatitis B and C virus infections on the natural history of compensated cirrhosis: a cohort study of 297 patients. Am J Gastroenterol. 2002;97:2886–2895
  113. Hui AY, Chan HL, Leung NW, Hung LC, Chan FK, Sung JJ. Survival and prognostic indicators in patients with hepatitis B virus-related cirrhosis after onset of hepatic decompensation. J Clin Gastroenterol. 2002;34:569–572
  114. Porcel A, Diaz F, Rendon P, Macias M, Martin-Herrera L, Giron-Gonzalez JA. Dilutional hyponatremia in patients with cirrhosis and ascites. Arch Intern Med. 2002;162:323–328
  115. Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in healthy human subjects, hospitalized patients, and patients with liver cirrhosis. Eur J Appl Physiol. 2002;86:509–516
  116. Abraldes JG, Tarantino I, Turnes J, Garcia-Pagan JC, Rodes J, Bosch J. Hemodynamic response to pharmacological treatment of portal hypertension and long-term prognosis of cirrhosis. Hepatology. 2003;37:902–908
  117. Bal JS, Thuluvath PJ. Prolongation of QTc interval: relationship with etiology and severity of liver disease, mortality and liver transplantation. Liver Int. 2003;23:243–248
  118. Botta F, Giannini E, Romagnoli P, Fasoli A, Malfatti F, Chiarbonello B, et al. MELD scoring system is useful for predicting prognosis in patients with liver cirrhosis and is correlated with residual liver function: a European study. Gut. 2003;52:134–139
  119. Hui JM, Kench JG, Chitturi S, Sud A, Farrell GC, Byth K, et al. Long-term outcomes of cirrhosis in nonalcoholic steatohepatitis compared with hepatitis C. Hepatology. 2003;38:420–427
  120. Jepsen P, Vilstrup H, Moller JK, Sorensen HT. Prognosis of patients with liver cirrhosis and spontaneous bacterial peritonitis. Hepatogastroenterology. 2003;50:2133–2136
  121. Korner T, Kropf J, Kosche B, Kristahl H, Jaspersen D, Gressner AM. Improvement of prognostic power of the Child-Pugh classification of liver cirrhosis by hyaluronan. J Hepatol. 2003;39:947–953
  122. Liangpunsakul S, Ulmer BJ, Chalasani N. Predictors and implications of severe hypersplenism in patients with cirrhosis. Am J Med Sci. 2003;326:111–116
  123. Longheval G, Vereerstraeten P, Thiry P, Delhaye M, Le Moine O, Deviere J, et al. Predictive models of short- and long-term survival in patients with nonbiliary cirrhosis. Liver Transpl. 2003;9:260–267
  124. Nidegger D, Ragot S, Berthelemy P, Masliah C, Pilette C, Martin T, et al. Cirrhosis and bleeding: the need for very early management. J Hepatol. 2003;39:509–514
  125. Pessione F, Ramond MJ, Peters L, Pham BN, Batel P, Rueff B, et al. Five-year survival predictive factors in patients with excessive alcohol intake and cirrhosis. Effect of alcoholic hepatitis, smoking and abstinence. Liver Int. 2003;23:45–53
  126. Schenk P, Schoniger-Hekele M, Fuhrmann V, Madl C, Silberhumer G, Muller C. Prognostic significance of the hepatopulmonary syndrome in patients with cirrhosis. Gastroenterology. 2003;125:1042–1052
  127. Serra MA, Escudero A, Rodriguez F, del Olmo JA, Rodrigo JM. Effect of hepatitis C virus infection and abstinence from alcohol on survival in patients with alcoholic cirrhosis. J Clin Gastroenterol. 2003;36:170–174
  128. Bell H, Jahnsen J, Kittang E, Raknerud N, Sandvik L. Long-term prognosis of patients with alcoholic liver cirrhosis: a 15-year follow-up study of 100 Norwegian patients admitted to one unit. Scand J Gastroenterol. 2004;39:858–863
  129. Gildea TR, Cook WC, Nelson DR, Aggarwal A, Carey W, Younossi ZM, et al. Predictors of long-term mortality in patients with cirrhosis of the liver admitted to a medical ICU. Chest. 2004;126:1598–1603
  130. Giron-Gonzalez JA, Martinez-Sierra C, Rodriguez-Ramos C, Macias MA, Rendon P, Diaz F, et al. Implication of inflammation-related cytokines in the natural history of liver cirrhosis. Liver Int. 2004;24:437–445
  131. Heuman DM, Abou-Assi SG, Habib A, Williams LM, Stravitz RT, Sanyal AJ, et al. Persistent ascites and low serum sodium identify patients with cirrhosis and low MELD scores who are at high risk for early death. Hepatology. 2004;40:802–810
  132. Moreau R, Delegue P, Pessione F, Hillaire S, Durand F, Lebrec D, et al. Clinical characteristics and outcome of patients with cirrhosis and refractory ascites. Liver Int. 2004;24:457–464
  133. Park DK, Um SH, Lee JW, Lee JB, Kim YS, Park CH, et al. Clinical significance of variceal hemorrhage in recent years in patients with liver cirrhosis and esophageal varices. J Gastroenterol Hepatol. 2004;19:1042–1051
  134. Romero-Gomez M, Grande L, Camacho I. Prognostic value of altered oral glutamine challenge in patients with minimal hepatic encephalopathy. Hepatology. 2004;39:939–943
  135. Yoneyama K, Nebashi Y, Kiuchi Y, Shibata M, Mitamura K. Prognostic index of cirrhotic patients with hepatic encephalopathy with and without hepatocellular carcinoma. Dig Dis Sci. 2004;49:1174–1180
  136. Yoneyama K, Taniguchi H, Kiuchi Y, Shibata M, Mitamura K. Prognostic index of liver cirrhosis with ascites with and without hepatocellular carcinoma. Scand J Gastroenterol. 2004;39:1272–1279
  137. Biggins SW, Rodriguez HJ, Bacchetti P, Bass NM, Roberts JP, Terrault NA. Serum sodium predicts mortality in patients listed for liver transplantation. Hepatology. 2005;41:32–39
  138. Giron-Gonzalez JA, Martinez-Sierra C, Rodriguez-Ramos C, Rendon P, Macias MA, Fernandez-Gutierrez C, et al. Adhesion molecules as a prognostic marker of liver cirrhosis. Scand J Gastroenterol. 2005;40:217–224
  139. Huo TI, Wu JC, Lin HC, Lee FY, Hou MC, Lee PC, et al. Evaluation of the increase in model for end-stage liver disease (DeltaMELD) score over time as a prognostic predictor in patients with advanced cirrhosis: risk factor analysis and comparison with initial MELD and Child-Turcotte-Pugh score. J Hepatol. 2005;42:826–832
  140. Kayali Z, Ranguelov R, Mitros F, Shufelt C, Elmi F, Rayhill SC, et al. Hemosiderosis is associated with accelerated decompensation and decreased survival in patients with cirrhosis. Liver Int. 2005;25:41–48
  141. Papatheodoridis GV, Cholongitas E, Dimitriadou E, Touloumi G, Sevastianos V, Archimandritis AJ. MELD vs Child-Pugh and creatinine-modified Child-Pugh score for predicting survival in patients with decompensated cirrhosis. World J Gastroenterol. 2005;11:3099–3104
  142. Pineda JA, Romero-Gomez M, Diaz-Garcia F, Giron-Gonzalez JA, Montero JL, Torre-Cisneros J, et al. HIV coinfection shortens the survival of patients with hepatitis C virus-related decompensated cirrhosis. Hepatology. 2005;41:779–789
  143. Ruf AE, Kremers WK, Chavez LL, Descalzi VI, Podesta LG, Villamil FG. Addition of serum sodium into the MELD score predicts waiting list mortality better than MELD alone. Liver Transpl. 2005;11:336–343
  144. Nahon P, Ganne-Carrie N, Degos F, Nahon K, Paries J, Grando V, et al. Serum albumin and platelet count but not portal pressure are predictive of death in patients with Child-Pugh A hepatitis C virus-related cirrhosis. Gastroenterol Clin Biol. 2005;29:347–352
  145. Infante-Rivard C, Villeneuve JP, Esnaola S. A framework for evaluating and conducting prognostic studies: an application to cirrhosis of the liver. J Clin Epidemiol. 1989;42:791–805
  146. Diamond GA. Future imperfect: the limitations of clinical prediction models and the limits of clinical prediction. J Am Coll Cardiol. 1989;14:12A–22A
  147. Laupacis A, Sekar N, Stiell IG. Clinical prediction rules. A review and suggested modifications of methodological standards. J Am Med Assoc. 1997;277:488–494
  148. Justice AC, Covinsky KE, Berlin JA. Assessing the generalizability of prognostic information. Ann Intern Med. 1999;130:515–524
  149. Harrell FE, Lee KL, Matchar DB, Reichert TA. Regression models for prognostic prediction: advantages, problems, and suggested solutions. Cancer Treat Rep. 1985;69:1071–1077
  150. Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Ann Intern Med. 1993;118:201–210
  151. Pagliaro L, Pasta L, D'Amico G, Pulco A, Politi F, Vizzini G. Natural history of cirrhosis: predictors of death risk; overview of published studies and validation in a new test set. In:  Dobrilla G,  dePretis M editor. Advances in hepatobiliary and pancreatic diseases. Dordrecht: Kluwer; 1995;p. 82–93
  152. Groszmann RJ, Garcia-Tsao G, Bosch J, Grace ND, Burroughs AK, Planas R, et al. Beta-blockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med, in press.

PII: S0168-8278(05)00684-7

doi:10.1016/j.jhep.2005.10.013

Journal of Hepatology
Volume 44, Issue 1 , Pages 217-231, January 2006

Article Tools

* Image Usage & Resolution