Influence of apolipoprotein E polymorphism in alcoholic cirrhosis

OBJECTIVE: To assess whether the polymorphism of apolipo-protein E was associated with the development of alcoholic cirrhosis and could influence the severity of liver injury evaluated by the Child-Pugh score. METHOD: We investigated 75 alcoholic patients with a histological diagnosis of cirrhosis, with negative HBV, HCV serology and a control group of 54 subjects. Polymorphism of apolipoprotein E was performed using PCR. RESULTS: There was no difference for the allele frequency and the genotype in the cirrhotic group and the control group. Cirrhotic patients with allele epsilon 2 had higher concentration of albumin (P = 0.01) and a higher level of apolipoprotein AII (P < 0.05) than those with allele epsilon 3. They also had a higher concentration of apolipoprotein AI than cirrhotic patients with allele epsilon 3 and epsilon 4 (P = 0.01). There was a statistical difference between the three genotype groups for prothrombin time (P = 0.01). There was no statistical difference between the three genotype groups for Child-Pugh score. CONCLUSIONS: Polymorphism of apolipoprotein E was not associated with the development of alcoholic cirrhosis. However patients with allele epsilon 2 had better hepatocellular function.     

Vasodilatory effects of propylthiouracil in patients with alcoholic cirrhosis

BACKGROUND/AIMS: An experimental study has shown that propylthiouracil increases portal blood flow in normal rats. Whether propylthiouracil has a similar effect in patients with alcoholic cirrhosis remains to be demonstrated. The aim of this study was to evaluate the effects of oral propylthiouracil (300 mg) on systemic and portal hemodynamics in patients with alcoholic cirrhosis. METHODS: Plasma propylthiouracil levels were also measured by high performance liquid chromatography in five patients with alcoholic cirrhosis. In eight patients with cirrhosis, mean arterial pressure, cardiac output and portal blood flow were evaluated before and after placebo and propylthiouracil administration. Hemodynamic measurements were performed by the Doppler technique. The plasma peak level of propylthiouracil was achieved at 1.4 +/- 0.1 h in patients with alcoholic cirrhosis. This time was chosen to express hemodynamic changes. RESULTS: Propylthiouracil administration caused a significant increase in portal blood flow (+16.5%, p < 0.05) in patients with alcoholic cirrhosis. This effect was associated with a mild and significant rise in cardiac output (from 5.8 +/- 0.2 to 6.1 +/- 0.3 l/min, p < 0.05) and a decrease in peripheral vascular resistance (from 1171 +/- 69 to 1070 +/- 67 dyn . s-1 . cm-5, p < 0.01). A significant correlation was observed between changes in portal blood flow and peripheral vascular resistance (r = 0.79, p < 0.05). No significant changes were observed after placebo. CONCLUSIONS: Our findings show that propylthiouracil has a vasodilatory effect in patients with alcoholic cirrhosis. We postulate that this effect could be the mechanism by which propylthiouracil decreases hypermetabolic state, and increases oxygen delivery to the liver, in patients with alcoholic liver diseases.     

Cytomegalovirus viremia: risk factor for allograft cirrhosis after liver transplantation for hepatitis C

BACKGROUND: Despite recent advances in diagnosis and treatment, cytomegalovirus (CMV) infection continues to be a common cause of morbidity in liver transplant (LT) recipients. Because CMV infection suppresses cell-mediated immunity, which seems to be important in neutralizing hepatitis C virus (HCV) infection, we assessed the impact of CMV infection on histopathological HCV recurrence after LT. METHODS: The study group was comprised of 43 consecutive LT recipients with at least 6 months of histologic follow-up. Group 1 consisted of the 8 patients who developed CMV viremia after LT; group 2 comprised the 35 patients without CMV viremia. There was no significant difference with regard to age, initial immunosuppression, incidence of rejection, distribution of HCV genotypes, or mean follow-up between the groups. Semiquantitative histopathologic assessment of allograft hepatitis was performed using the Knodell's score. RESULTS: The mean total Knodell score of the final allograft biopsy was significantly greater in group 1 patients (P=0.016), with most of the difference due to periportal/bridging necrosis (P=0.009) and lobular activity subitem (P=0.01) scores. Half of the CMV viremic patients eventually developed allograft cirrhosis as compared with 11% of the CMV-negative patients (P=0.027). Accordingly, the cirrhosis-free actuarial survival by Kaplan-Meier estimates was significantly diminished in the CMV viremic patients. Glycoprotein B genotype analysis of CMV isolates revealed no significant differences between patients who did and those who did not develop allograft cirrhosis. CONCLUSIONS: After LT for chronic HCV, patients who develop CMV viremia incur a significantly greater risk of severe HCV recurrence.     

Liver disease in alcoholic cardiomyopathy: evidence against cirrhosis

The authors examined in detail the clinical and laboratory data and pathologic findings for 12 patients with alcoholic cardiomyopathy who were autopsied in the preceding ten years to determine the types of liver disease prevalent in this population. Neither alcoholic hepatitis nor cirrhosis was present in any patient, and most of the hepatic changes could be related to the effects of acute and chronic congestive heart failure. The major hepatic lesions included centrilobular congestion and/or ischemic necrosis, cardiac sclerosis (fibrosis about central veins and in perisinusoidal spaces), mild canalicular cholestasis, portal fibrosis, and nodular regenerative hyperplasia. This last finding may account for macroscopic nodularity resembling cirrhosis as well as portal hypertension in patients with alcoholic cardiomyopathy. Although alcoholic cardiomyopathy and alcoholic hepatitis or cirrhosis were mutually exclusive in the patients studied, the factors responsible for this are at present uncertain.     

Comparison of immunogenicity of vaccination and serovaccination against hepatitis B virus in patients with alcoholic cirrhosis

OBJECTIVES: The association of anti-HBs immunoglobulins and anti-HBV vaccine could increase the immunogenicity of the latter. The aim of this prospective randomized trial was to compare the immunogenicity of anti-HBs vaccination and serovaccination in alcoholic patients with cirrhosis. METHODS: Alcoholic patients with cirrhosis were randomized in 2 groups: a) Vaccination group: 3 i.m. injections of GenHevac B followed by one booster at month 9; b) Serovaccination group: same vaccination schedule followed by one i.m. injection of anti-HBs immunoglobulins (500 IU). RESULTS: Twenty-five patients (17 males and 8 females, mean age 56 years) were included in the study: 13 received a vaccination and 12 received a serovaccination. After 12 months, the seroconversion rates were 69% and 67% in vaccination and in serovaccination groups, respectively. The predictive factors of non responsiveness were as following: Child B cirrhosis, low number of CD8, a high CD4/CD8 rate, the existence of HLA DR7 antigen, and the absence of HLA DR1 antigen. CONCLUSION: In alcoholic patients with cirrhosis, serovaccination does not increase the immunogenicity of anti-HBs vaccination and should not be recommended.     

Relation between the quantity, type and duration of alcohol drinking and the development of alcoholic liver cirrhosis

Patients with alcoholic liver cirrhosis in Dalmatian region are presented in this paper and their social and medical characteristics are reviewed. The alcohol drinking habits of 211 patients with established clinical-laboratory criteria of alcoholic liver cirrhosis were investigated. The obtained results have been compared with the results of the control group. The average age of the patients was 51.6 years. Males were more numerous (2/3 of all examinees), while the average duration of alcohol consumption was 25.4 years. Wine was the most frequently consumed beverage (about 90% of cases) and the average daily intake was 151 g of pure alcohol. The authors prove by mathematical model that the occurrence of liver cirrhosis increases exponentially with the increase of the amount of alcohol consumed. Relative risk of the development of alcoholic liver cirrhosis increases many fold with the increase of the amount of alcohol consumed.     

A controlled clinical trial of methylprednisolone in patients with the cholestatic form of alcoholic liver cirrhosis

Fourty patients with alcoholic liver cirrhosis and refractory hyperbilirubinaemia were included in a prospective, double blind, comparative trial. Twenty of them were randomized to methylprednisolone (1 mg/kg/day i.m. over 3 days), and 20 to placebo (saline) of identical shape. In the active group a significant decrease in the bilirubin level (from 248 to 191 mumol/l, a drop by 23%; p < 0.05) was observed, which was not the case in the control group (from 292 to 300 mumol/l, an increase by 2.7%; p > 0.05). A decrease in the alkaline phosphatase activity was observed in both groups (by 11% in the active and by 20% in the control group: p < 0.05) while the encephalopathic indices did not improve in either. It is concluded that a short course of corticosteroid could speed-up the hospital stay and possibly slow down the progression of alcoholic cirrhosis.     

Effect of the type of beverage and meat consumed by alcoholics with alcoholic liver disease

We analyzed meat products and alcoholic beverage preference in patients with the three stages of alcoholic liver disease (ALD) compared with controls using diet history data. Daily consumption of total alcohol, types of alcoholic beverages, and types of meat and meat products in grams was obtained by dietary history taken from patients with biopsy proven stage of ALD. A strong association was found between the ALD subjects and total alcohol and beer consumption. There was a significant increase in the consumption of total pig products, pork, and offal in the ALD groups compared with controls. There was a significant positive correlation between beer consumption and pork in alcoholic hepatitis, total pork products in alcoholic hepatitis, and cirrhosis and offal in alcoholic hepatitis and cirrhosis. There was no correlation with the fatty liver stage of ALD. The strongest correlation was between beer and total pig products in the alcoholic hepatitis group. Wine consumption was negatively correlated with the consumption of pig products and beer in the alcoholic cirrhosis group. In conclusion, the association of total pig product consumption with cirrhosis mortality in various countries was validated by personal diet history data obtained from ALD patients in a tested clinical microcosm. The results suggest that this association may be modified by the type of alcoholic beverage that is preferentially consumed.     

Fiber morphometry of the external intercostal muscle. Comparison of dominant and nondominant sides in patients with severe COPD]

OBJECTIVES: Contribution of cellular immunity to the onset and perpetuation of alcohol-induced liver damage remains controversial. The aim of this work was to know whether T-cells participate in the pathogenesis of alcoholic liver injury by measuring the serum levels of sIL-2R in alcoholic patients with different degree of hepatic damage. PATIENTS AND METHODS: Fifty two patients and eighteen healthy subjects (Control group) were included. All patients were active drinkers of at least 100 grams/day of ethanol over a ten-years period. Serum sIL-2R was determined by ELISA. Liver biopsy was performed in all patients and liver function tests, serum immunoglobulins and complement proteins C3 and C4 were measured in all participants. The relationship between the sIL-2R and the severity of liver disease was studied. RESULTS: Circulating sIL-2R was higher in the group of patients than in the control (2.388 +/- 275.7 U/ml vs. 795.7 +/- 48.7 IU/mL; p < 0.001). There were not increased circulating sIL2R in those patients with alcoholic hepatitis. However, patients with cirrhosis showed increased serum sIL-2R regardless of the presence of alcoholic hepatitis. Furthermore, serum levels of sIL-2R inversely correlated with hepatic function test (r = -0.69; p < 0.001 for serum albumin; and r = -0.73; p < 0.001 for the prothrombin time) and were highest in those patients of the Child-Turcotte's class C. CONCLUSIONS: Circulating sIL-2R increases in alcoholic cirrhosis. However, our data do not support a contributory role of the cellular immunity, as assessed by circulating sIL-2R levels to the alcoholic liver damage. The increased serum sIL-2R in cirrhosis may result from defective heptic clearance of this molecule.     

Effects of silymarin in alcoholic patients with cirrhosis of the liver: results of a controlled, double-blind, randomized and multicenter trial

BACKGROUND/AIMS: Silymarin has protective effects in different experimental conditions, but its efficacy in human liver cirrhosis has not been completely established. Therefore, this study was carried out to determine the effect of silymarin in alcoholics with liver cirrhosis with respect to survival and clinical and laboratory changes. METHODS: From February 1986 to June 1989, we enrolled 200 alcoholics with histologically or laparoscopically proven liver cirrhosis in a randomized, double-blind multicenter trial comparing 450 mg of silymarin (150 mg/ three times per day) with placebo. The primary outcome was time to death, and the secondary outcome was the progression of liver failure. Additional analyses were also performed in 75 patients in whom anti-hepatitis C virus antibodies were measured after completion of the trial. RESULTS: One hundred and three patients were assigned to receive silymarin and 97 to receive placebo. The two groups were well matched for demographic and baseline clinical and laboratory features. A 2-year study period was completed in 125 patients (57 receiving silymarin and 68 receiving placebo). Twenty-nine patients (15 receiving silymarin, and 14 receiving placebo) died during the trial. Survival was similar in patients receiving silymarin or placebo. The effect of silymarin on survival was not influenced by sex, the persistence of alcohol intake, the severity of liver dysfunction or by the presence of alcoholic hepatitis in the liver biopsy. Silymarin did not have any significant effect on the course of the disease. No relevant side-effects were observed in any group. CONCLUSIONS: The results of this study indicate that silymarin has no effect on survival and the clinical course in alcoholics with liver cirrhosis.     

Proportion of the GSTM1 0/0 genotype in some Slavic populations and its correlation with cystic fibrosis and some multifactorial diseases

A homozygous gene deletion at the glutathione S-transferase M1 (GSTM1) locus of genomic DNA from blood spots was studied by PCR in the group of Slavic populations from the north-western and central-eastern regions of European Russia and in patients with lung cancer (LC), other tumors (OT), endometriosis (E), alcoholic cirrhosis (AC), cystic fibrosis (CF) and chronic bronchitis (CB). The frequencies of the GSTM1 0/0 genotype were 38.8% and 67.5% for both population groups, respectively. The proportion of the GSTM1 gene deletion genotype was estimated as significantly increased in LC (81%), OT (65%), E (81%), AC (77.3%), and in CB (73.6%) patients with symptoms of CB confirmed by X-ray but not in CB patients without X-ray evidence of disease (40.9%). A definite preponderance of GSTM1-0 homozygotes (51.1%) has been registered in CF patients of the pancreatic sufficient group with clear-cut pulmonological manifestations but not in those of the pancreatic insufficient group with predominantly intestinal or mixed clinical symptoms (41.2% and 37.5%, respectively). Earlier clinical manifestations and death before the age of 5 years are typical for GSTM1-deleted CF patients. These data support the notion that GSTM1 deletion should be considered as a convenient genetic marker for the early detection of groups at higher risk of many diseases caused by environmental and genetic factors, where manifestation depends on the lack of detoxification. High levels of GSTM1 0/0 genotypes in E patients favor the substantial contribution of certain environmental toxins in the pathogenesis of this widespread disease.     

Antral gastritis in chronic alcoholism. Role of cirrhosis and Helicobacter pylori

OBJECTIVES: Antral gastritis is frequent in alcoholics. The role of H. pylori in the pathogenesis of gastritis in these patients is not well known. The aim of our study was to study the role of H. pylori and cirrhosis in the pathogenesis of antral gastritis in alcoholic patients. METHODS: Seventy-nine patients were included in the study. All underwent upper gastrointestinal tract endoscopy with antral biopsies, independently of the presence of abdominal pain, and had serological examination for H. pylori antibodies. RESULTS: Cirrhosis and gastritis were present in 50 and 40 patients respectively, H. pylori serological assay and histological identification of the bacterium were positive in 35 (44%) and 19 (24%) patients respectively. Discrepancy between the 2 tests were observed more frequently in cirrhotic patients. A positive serology with a negative histologic examination for H. pylori was present for 18 cirrhotic and 4 noncirrhotic patients (p < 0.05). A gastritis without evidence of H. pylori was more frequent in cirrhotic than in noncirrhotic patients. H. pylori was histologically present in 11 of 29 cirrhotic patients and in 8 of the 11 noncirrhotic patients with a gastritis (p < 0.05). CONCLUSIONS: Discrepancies between histological examination and H. pylori serology in patients with cirrhosis might be due to the inhospitable environment for H. pylori in case of portal hypertension; the positive serology could be in relation with a past infection.     

Gastric clearance of serum albumin in patients with alcoholic liver cirrhosis

Gastric albumin clearance studies were carried out in 26 patients with alcoholic cirrhosis and in 10 control subjects. The mean clearance was significantly higher in patients than in control subjects. An increased protein gastric loss in cirrhotic patients was observed only in presence of gastritis. The presence or absence of distended esophageal varices did not influence the protein loss.     

Evidence for a tyrosine kinase-dependent activation of the adenylyl Cyclase/PKA cascade downstream from the G-protein-linked endothelin ETA receptor in vascular smooth muscle

Body retinoids are stored in the lipid droplets of hepatic stellate (Ito) cells. In chronic liver disease, the stellate cells differentiate into myofibroblast-like cells, a process whereby they lose their retinoid-containing lipid droplets. We studied the relation between liver retinoid content, the number of lipid droplets per stellate cell, and the number of stellate cells per mm2 in human alcoholic liver disease. Semithin sections of liver biopsies from normal subjects and patients with early (steatosis, inflammation, and mild fibrosis) and late (cirrhosis and cirrhosis with acute alcoholic hepatitis) alcoholic liver disease were morphometrically evaluated. Liver retinoid content was determined by HPLC. In normal patients, liver retinoid content was 901 +/- 213 IU/g of liver (mean +/- SEM). There was a decrease in liver retinoid content in early alcoholic liver disease (409 +/- 50 IU/g) and a further reduction in cirrhosis (153 +/- 50 IU/g). In patients with acute alcoholic hepatitis, retinoid content was strikingly low (5.2 +/- 1.8 IU/g). There was a progressive decrease in the number of stellate cells per mm2 associated with progressive liver damage. We found a fair correlation between the number of stellate cells per mm2 and liver retinoid content in all patient groups (overall correlation: 0.71). In normal subjects, the mean number of lipid droplets per stellate cell was 7.4 +/- 0.7. In patients with early alcoholic liver disease and in patients with alcoholic cirrhosis, this value was increased to 13.6 +/- 0.8 and 10.4 +/- 2.0, respectively. In patients with acute alcoholic hepatitis, only a few lipid droplets were present (4.2 +/- 0.5). There was a good correlation between liver retinoid content and mean number of lipid droplets in normal patients (r = 0.58). In alcoholic cirrhosis, however, correlation was poor (r = 0.34). In early alcoholic liver disease, the correlation was absent (r = 0.004). In conclusion, the major finding of our study is that the correlation between the mean number of lipid droplets per stellate cell and liver retinoid content varies according to the hepatic pathology considered. Marked lipid droplet accumulation occurs in stellate cells in early alcoholic liver disease and, to a lesser extent, in alcoholic cirrhosis, but there is no correlation between the mean number of lipid droplets per stellate cell and liver retinoid content. Therefore, not retinoids but probably lipids are responsible for the accumulation of lipid droplets. We also find that there is a fair correlation between the number of stellate cells per mm2 and liver retinoid content in all patient groups. Finally, we confirm the decrease in hepatic retinoid content that occurs in alcoholic liver disease in humans, even at the early stages of the disease.     

Scintigraphic diagnosis of cirrhosis: a receiver operator characteristic analysis of the common interpretative criteria

Receiver operator characteristic (ROC) techniques were used to determine relative importance of heterogeneity, bone marrow uptake, hepatomegaly, splenomegaly, and focal lesions in the scintigraphic diagnosis of cirrhosis, as well as to determine whether each of the criteria should be used as an identifier (high sensitivity) or as a discriminator (high specificity). Heterogeneity, splenomegaly, and bone marrow uptake were found to be good identifiers of hepatic disease. However, heterogeneity was found to be a poor discriminator for cirrhosis, splenomegaly showed some value as a discriminator, and bone marrow uptake was found to be a reliable discriminator for alcoholic liver disease. The presence of focal lesions was a good discriminator for metastatic disease, and was not specific for cirrhosis. Hepatomegaly also was not specific for cirrhosis. This method of analysis provides a relatively simple method for assembling a hierarchical guide to diagnostic criteria for the interpreting physician.     

Morbidity after video-laparoscopic cholecystectomy in cholelithiasis associated with liver cirrhosis. A case-control study

At present laparoscopic cholecystectomy represents the treatment of choice for symptomatic cholelithiasis. Authors performed a retrospective case-control study to evaluate whether cirrhosis associated with cholelithiasis increases the risk for morbidity of laparoscopic cholecystectomy. Twenty-one patients with cholelithiasis and cirrhosis (Child-Pugh class A or B) (group A) and 21 controls with cholelithiasis without cirrhosis (group B) entered the study. Controls were paired with cases for age, sex, and indication for cholecystectomy (simple cholelithiasis, acute cholecystitis). The two groups were compared for rate of conversion to open cholecystectomy (19% group A vs 9.5% group B; p = 0.31), morbidity (29.5% group A vs 5.3% group B; p = 0.17), median length of surgery (80 m in the two groups), and median time of postoperative hospitalization (5 days group A vs 3 days group B; p = 0.21). No difference among variables resulted to be statistically significant. Besides, neither common bile duct injuries nor intra or postoperative hemorrhages occurred in patients with cirrhosis. Authors conclude that the laparoscopic cholecystectomy can be considered a safe and effective surgical procedure also for patients with cholelithiasis associated with cirrhosis with a good residual hepatic function.     

Diencephalic and cerebellar pathology in alcoholic and nonalcoholic patients with end-stage liver disease

Formalin-fixed sections from the brains of 36 patients (30 alcoholic and 6 nonalcoholic) with autopsy-proven cirrhosis who died while in a hepatic coma were stained with hematoxylin and eosin, and examined for the presence of diencephalic, cerebellar, pontine, and basal ganglia lesions. Significant neuropathology was identified in 23 of 36 cases consisting of mammillary body and thalamic lesions characteristic of Wernicke encephalopathy (WE) (9 cases, all alcoholic patients) and cerebellar degeneration (20 cases, 17 alcoholic and 3 nonalcoholic patients). Clinical diagnosis of WE had been entertained during life in only 2 of these patients. All cases, alcoholic and nonalcoholic, manifested mild to severe Alzheimer's type II astrocytosis. No cases of central pontine myelinolysis nor acquired (non-Wilsonian) hepatocerebral degeneration were found. These findings show that the brains of a high proportion of cirrhotic patients with end-stage liver disease manifest concomitant unsuspected diencephalic and cerebellar pathology. The high incidence of WE underscores the need for early sustained treatment of alcoholic cirrhotic patients with vitamin B1. Evaluation of the neurological sequelae of liver transplantation, particularly of alcoholic patients with end-stage liver disease, may require a careful neurological and radiological assessment both before and after surgery.     

Angiosarcoma of the bladder: a review

BACKGROUND/AIMS: Correlations between serum levels of soluble tumor necrosis factor receptors p55 (TNFsRp55) and Child Pugh index have previously been reported in alcoholic patients with cirrhosis. We have undertaken this study to improve understanding of the role of tumor necrosis factor soluble receptors (TNFsRs) in alcoholic liver disease. METHODS: One hundred and two patients with alcoholic liver disease of various severity (23 pure steatosis, 22 fibrosis, seven acute alcoholic hepatitis without cirrhosis, 12 cirrhosis without acute alcoholic hepatitis, 14 cirrhosis with mild acute alcoholic hepatitis and 24 cirrhosis with severe acute alcoholic hepatitis) were studied. Blood was collected on EDTA and plasma was tested for TNFsR concentrations using ELISA assays. RESULTS: Plasma levels of TNFsRp55 and p75 increased progressively with the severity of liver disease, reaching a maximum in cirrhotic patients with severe acute alcoholic hepatitis. Plasma levels of TNFsRp55 in patients with fibrosis and of TNFsRp75 in patients with acute alcoholic hepatitis without cirrhosis were already higher than in healthy controls. In cirrhotic patients with or without acute alcoholic hepatitis TNFsRp55 and p75 were significantly increased compared with controls. In cirrhotic patients, plasma levels of TNFsRp55 correlated positively with all parameters of liver injury, whereas the TNFsRp75/ TNFsRp55 ratio correlated negatively. In cirrhotic patients with severe acute alcoholic hepatitis, the TNFsRp75/TNFsRp55 ratio was significantly lower than in all other groups. In cirrhotic patients with severe acute alcoholic hepatitis treated by prednisolone, the decrease in TNFsRp55 plasma levels between day 1 and day 15 was significantly more important in patients still alive at 2 months than in patients who died within 2 months. CONCLUSIONS: These results show that the expression of TNF-soluble receptors (TNFsRs) participates in the early phases of the alcoholic liver disease and that the TNFsRp75/TNFsRp55 ratio and plasma levels of TNFsRp55 may help to determine the diagnosis and the prognosis of severe acute alcoholic hepatitis in cirrhotics.     

The course of galactose elimination capacity in patients with alcoholic cirrhosis: possible use as a surrogate marker for death

There is increasing interest for the use of surrogate end points in the evaluation of treatments in patients with liver disease, but adequate validation is seldom available. This study aimed to describe the different course of galactose elimination capacity in patients with alcoholic cirrhosis who continued to drink or abstained from alcohol consumption during follow-up, and to validate changes in galactose elimination as a surrogate end point for death from liver-related causes. Forty-five patients with alcoholic cirrhosis (22 who continued drinking throughout the study period, and 23 who stopped drinking and were abstinent throughout the study period) were retrospectively selected among patients who had galactose elimination capacity measured at 6-month intervals. During follow-up 10 drinkers and 3 abstainers died of liver-related causes (P = .025). Abstainers showed a transient improvement in galactose elimination capacity, followed by a decrease. Continuous drinkers showed a reduction from the beginning. According to Cox's regression analyses, persistent alcohol abuse and galactose elimination capacity were separately related to the risk of death, but, when a time-dependent model was fitted containing galactose elimination capacity and persistent alcohol abuse, only the former remained significant. This implies that variations in the risk of death occurring as a consequence of abstinence from alcohol consumption may be predicted from changes in galactose elimination capacity, and that the mechanisms through which abstinence influences survival are strictly linked to the mechanisms responsible for the changes in the test. Because of the strict association of decrease in galactose elimination capacity and short survival, as proved in several series, this observation represents adherence to the criteria requested for adequacy of a surrogate end point. In conclusion, in alcoholic cirrhosis the decrease in galactose elimination capacity is an adequate surrogate end point for death from liver-related causes, which is worth testing in other conditions and in response to other treatments.