Natural endogenous ligands for benzodiazepine receptors in hepatic encephalopathy.

Benzodiazepines of natural origin (NBZDs) have been found in human blood and brains as well as in medicinal plants and foods. In plasma and brain tissue there are i.e. diazepam and nordiazepam equal to commercial drugs but there are also other benzodiazepine-like compounds termed "endozepines", which act as agonists at the benzodiazepine receptors of central type (CBR). A synthetic pathway for the production of NBZDs has not yet been found, but it has been suggested that micro-organisms may synthesize molecules with benzodiazepine-like structures. Hence NBZDs could be of both endogenous and exogenous source and be considered as natural anxyolitic and sedative. Interestingly there are also natural compounds, such as the polypeptide Diazepam Binding Inhibitor (DBI) acting as an "inversive agonist" implicated in fair and panic disorders. It has been suggested that NBZDs may play a role in the pathogenesis of hepatic encephalopathy (HE). Multidirectional studies evaluated NBZDs levels (1) in the blood of normal subjects, of cirrhotic with or without HE and in commercial benzodiazepine consumers; (2) in the blood of cirrhotic treated or not with a non-absorbable antibiotic; (3) in several constituents of our diet. In conclusion, NBZDs increase sometime in cirrhotics with or without HE but they reach concentrations not higher than those found in commercial benzodiazepines consumers. Hence NBZDs must be considered as occasional precipitating factor of HE and benzodiazepine antagonists only symptomatic drugs. The finding that NBZDs may be in part synthesized by intestinal bacterial flora and in part constituent of our diet underlines the importance to feed cirrhotic patients with selected food.

CagA and VacA Helicobacter pylori antibodies in gastric cancer.

BACKGROUND: Infection with different genotypes of virulent Helicobacter pylori strains (cytotoxin-associated gene A [CagA]- and/or vacuolating cytotoxin A [VacA]-positive) can play a role in the development of atrophic gastritis, duodenal ulcer (DU) and gastric cancer (GC). OBJECTIVE: To determine whether patients with GC and H pylori-negative histological staining had previously been infected with H pylori CagA- and/or VacA-positive virulent strains. METHODS: Twenty-three GC patients with a mean (+/- SD) age of 68.14+/-9.8 years who tested H pylori-negative on histological staining took part in the study. Three control groups were included. The first group comprised 19 patients with past H pylori infection and DUs eradicated 10 years earlier, with a mean age of 58+/-18.2 years. H pylori-negative status for this group was determined every year with Giemsa staining, and follow-up testing occured 120+/-32 months (mean +/- SD) after therapy. The subsequent control groups included 20 asymptomatic children, with a mean age of 7+/-4.47 years, and with H pylori-negative fecal tests; the final group contained 30 patients without clinical symptoms of H pylori infection, with a mean age of 68+/-11.6 years, who tested H pylori-negative by histological staining. RESULTS: Prevalence of CagA and VacA seropositivity, respectively was 82.6% and 73.91% in GC patients; 84.2% and 84.2% in H pylori-negative DU patients; 25% and 5% in H pylori-negative children; and 36.6% and 16.6% in the patients without clinical symptoms on histological staining. CagA and VacA antibody positivity was not significantly different between GC patients and patients with DUs that had been eradicated 10 years earlier. Significant positivity was found between the children's group and the H pylori-negative (with past DUs) group (P<0.001). A statistically significant difference was found in age between groups (P<0.03). CONCLUSIONS: Patients with GC, even when H pylori-negative at the time of the present study, may have been infected by H pylori before the onset of the disease, as confirmed by CagA and VacA seropositivity. These data reinforce the hypothesis that H pylori may be a direct carcinogenic agent of GC.

Evaluation of rifaximin, placebo and lactulose in reducing the levels of benzodiazepine-like compounds in patients with liver cirrhosis: a pilot study.

Benzodiazepine-like compounds (BZDs), either taken with the diet or synthesized by intestinal bacterial flora, may represent a precipitating factor for hepatic encephalopathy (HE) in cirrhotic patients. We evaluated whether a diet and/or treatment with rifaximin or lactulose can reduce serum concentrations of BZDs in 18 cirrhotic patients without HE. Patients were given a standard diet for 7 days to keep the dietary intake of BZDs constant and were then randomized to a 7-day treatment with rifaximin 1,200 mg/day, lactulose 10-20 g three times daily, or placebo. Blood samples were collected at enrollment, at the end of the diet and drug treatment periods, and 7 days after the drug was stopped (follow-up). Serum concentrations of BZDs were measured by a radioligand binding technique after high-performance liquid chromatography extraction and purification and were expressed as diazepam equivalents (DE). No change in serum BZD concentrations was observed during the diet, while a statistically significant decrease from 105.6 +/- 66.5 to 63.5 +/- 49.5 pmol DE/ml was achieved in rifaximin-treated patients (p < 0.05) but not in patients treated with lactulose or placebo. During the followup, serum BZD concentrations returned to 104.5 +/- 74.0 pmol DE/ml in rifaximin-treated patients (p < 0.05 vs. end-treatment values), while no significant change was observed in the lactulose- and placebo-treated patients. These data indicate that control of bacterial flora with cyclic administration of rifaximin plays a pivotal role in avoiding increased plasma concentrations of BZDs, which represent a precipitating factor for HE inpatients with severe liver disease.

Helicobacter pylori antibodies (CagA and VacA) detection. The link between cancer and infection.

BACKGROUND: Infection with different genotypes of virulent Helicobacter pylori (Hp) strain, CagA and/or VacA positive, can play a role in the development of atrophic gastritis, duodenal ulcer (DU), and gastric carcinomas (GC). This study was undertaken to investigate if patients with GC with Hp negative histological Giemsa staining had a past infection by virulent strains of Hp CagA and/or VacA positive. METHODS: Twenty GC, (average age +/- SD) 68.14+/-9.8 years old, Hp negative to histological take part to the study. Two-control group were included: 19 Hp infected patients with DU eradicated 10 years before, 58+/-18.2 yrs. Hp negative status was determined every year with histology and follow-up after therapy was 120+/-32 months; range 96-144 months. Twenty asymptomatic children, 7+/-4.47 yrs, with Hp negative faecal test. The immunoblot assay was used to detect serum antibodies against CagA and VacA. RESULTS: Prevalence of CagA and VacA seropositivity was 90 and 95% in GC, 84 and 84% in DU Hp negative, 25 and 5% in children Hp negative, respectively. CagA and VacA antibody positivity was not significantly different between GC and patients with DU eradicated 10 years before. A true significant positivity was found against children (''t''-Student test; p<0.0001). Statistical difference was found in age between groups p<0.03. CONCLUSIONS: Patients with GC, although Hp negative at present, could be infected by Hp before the appearance of the disease as confirmed by CagA and VacA seropositivity. These data may reinforce the idea to consider Hp as a direct carcinogenetic agent of GC.

Gastric carcinoma risk index, type III intestinal metaplasia and Helicobacter pylori status on antrum and body biopsies in a prospective general population study.

BACKGROUND: To evaluate the utility of 2 biopsies of antrum and gastric body on routinary endoscopy for the assessment of type III intestinal metaplasia (IM-3) and Helicobacter pylori (Hp) status, 1750 patients (pts) (895 males and 855 females, mean age 60.2) were considered from June 1998 to June 2000. METHODS: Specimens were graded 0 to 3 for atrophy, IM-3 and Hp status. 620 pts treated previously with antibiotics or not eligible for biopsy were excluded from initial 2360 pts. RESULTS: IM-3 (score >0), was found in 118 pts (6.7%), 86 pts (4.9%) only in the antrum. Ten of 355 pts (2.8%) with normal endoscopy findings and 47 of 702 (6.6%) with non erosive endoscopic gastritis resulted IM-3 positive in the antrum. 709 pts (40.5%) were found positive for Hp in antrum or/and corpus. The presence of Hp and IM-3 in the antrum was not correlated (p=0.99; spearman test). A positive correlation (p=0.000) between duodenal ulcer and Hp was found when antral Hp positivity was taken into account. Gastric carcinoma risk index (GCRI) was found in 358 pts (20.4%); in this group 131 pts (36.6%) were Hp positive, 82 pts (23%) have IM-3, 184 pts (51.4%) have atrophy. CONCLUSIONS: The incidence of IM-3 is low (6.7%) in routinary endoscopy. Normal endoscopy does not exclude the presence of IM-3. The biopsy is necessary to discover IM-3 in the antrum in 5.3% of pts with normal or aspecific endoscopic gastritis. Application of the GCRI might be useful to identify a group of patients carrying a higher risk for gastric carcinoma.

Gastric atrophy and intestinal metaplasia changes 8 years after Helicobacter pylori eradication. A blind, randomised study.

BACKGROUND: Chronic atrophic gastritis and intestinal metaplasia are regarded as predisposing factors for gastric cancer associated with Helicobacter pylori infection, and their severity appears to influence gastric cancer risk. Our purpose was to determine the outcome of chronic gastritis after H. pylori eradication in a long-term follow-up. METHODS: Fifty-four consecutive patients with duodenal ulcer and H. pylori infection were enrolled in the study. Endoscopic examination with antral and corporal biopsy was done at baseline and yearly after conventional eradication therapy (omeprazole 40 mg b.i.d., amoxocyllin 1 g b.i.d and clarithromycin 500 mg b.i.d.). Gastritis, atrophy, and metaplasia were graded according to the updated Sydney System. RESULTS: Twenty-four patients were successfully treated; infection persisted in 14 and 16 dropped out (during the first 5 years of follow-up). Inflammation and mean neutrophil activity significantly decreased in patients in whom H. pylori was eradicated. Glandular atrophy improved in 2 and disappeared in 5/17 patients, whereas intestinal metaplasia improved in 3 and disappeared in 2/12. In the patients in whom H. pylori persisted, inflammatory infiltrate, atrophy and intestinal metaplasia had not significantly decreased during follow-up. In contrast, glandular atrophy worsened in 2 and developed in 5/7 patients. Similarly, intestinal metaplasia did not improve when present and developed in 5/13 cases. CONCLUSIONS: In a long-term follow-up, H. pylori eradication does not affect glandular atrophy, but it seems to prevent the development of precancerous lesions such as intestinal metaplasia.

Diagnostic use of endoscopic mucosal resection in ''borderline'' gastric lesions unresolved with standard biopsies.

BACKGROUND: Endoscopic mucosal resection (EMR) of gastric superficial malignancies less than 20 mm in size and flat or slightly elevated without ulceration can be a definitive treatment, but its role in lesions of uncertain etiology or in which standard biopsies specimens fail to determine diagnosis is uncertain. EMR was performed in 7 patients previously diagnosed as having low grade dysplasia (Category 3 of Vienna classification) by standard biopsies on polypoid or flat gastric lesions. METHODS: After day spraying with 0.2% indigo carmine and injection of 20 ml saline with adrenaline 1/20000, EMR of flat or sessile polyps (size between 5 to 15 mm) was performed by the Cap and Suction technique (Inoue). RESULTS: In 3 patients a previous diagnosis of low grade dysplasia was changed into high grade dysplasia, in 1 patient adenocarcinoma was found at EMR histology. In 3 patients EMR confirmed diagnosis made with routine endoscopy biopsies and finally in 2 patients dysplasia was down-graded into intestinal metaplasia. CONCLUSIONS: EMR may be considered in diagnostic gastric lesions with low grade dysplasia at standard biopsies (Category 3 of Vienna Classification of gastrointestinal neoplasia).

Ammonia and endogenous benzodiazepine-like compounds in the pathogenesis of hepatic encephalopathy.

BACKGROUND: Ammonia and endogenous benzodiazepines (BDZs) are two of the most important agents among those taken into consideration in the pathogenesis of hepatic encephalopathy (HE). METHODS: Venous ammonia and endogenous BDZs sera levels were assayed in 58 liver cirrhosis patients (34 male, 24 female) free of commercial BDZs. Endogenous BDZs were measured by binding assay after high-performance liquid chromatography purification. Ammonia was assessed by colorimetric test. RESULTS: Endogenous BDZs and ammonia were significantly higher in Child-Pugh class C than in class B and class A (P < 0.05), correlating to the severity of the liver dysfunction but not with the degree of HE. A significant difference, in fact, was noted between degree 0 (no HE) versus III-IV of HE (P < 0.05), but not between degrees I-II versus III-IV. Regression analysis performed to find a correlation between the ammonia and BDZ levels in HE resulted negative. CONCLUSION: Clinical evidence is provided in cirrhotic patients that ammonia and endogenous BDZ levels do not correlate with each other in the outcome of HE.

Endogenous benzodiazepines.

The existence of endogenous benzodiazepines such as diazepam and nordiazepam has been provided in human blood and brains as well as in medicinal plants and foods. It must be stressed, however, that in plasma and brain tissue there are also other benzodiazepine-like compounds termed 'endozepines' which are not halogenated. A synthetic pathway for the production of benzodiazepine-like compounds and endozepines has not yet been found, hence it may be surmised that these compounds could be of exogenous source. Changes in the level of endogenous circulating benzodiazepines due to food or drug ingestion could be responsible for pathological conditions. Clinical experiments were designed in order to study the levels of the endogenous benzodiazepines in vegetables and in the blood of control subjects and of cirrhotic patients. These patients accumulate benzodiazepines because of decreased liver metabolization capacity and impaired renal secretion, reaching plasma concentrations similar to those recorded in commercial benzodiazepine consumers.

Increased expression of peripheral benzodiazepine receptors and diazepam binding inhibitor in human tumors sited in the liver.

The peripheral benzodiazepine receptor system triggers intracellular metabolic events and has been associated with cell proliferation. Its endogenous ligand, the diazepam binding inhibitor, contributes to steroidogenesis by promoting cholesterol delivery to the inner mitochondrial membrane. The present study was undertaken to verify whether this system is altered in tumors sited in the liver. Peripheral benzodiazepine receptors and diazepam binding inhibitor were studied using immunocytochemistry and in situ hybridization in 9 human tumors sited in the liver, in liver hyperplasia, cirrhotic nodular regeneration, intestinal adenocarcinoma and in surrounding non-tumoral tissue. Immunocytochemical staining and in situ hybridization demonstrated that peripheral benzodiazepine receptors and diazepam binding inhibitor were more prominently expressed in neoplastic cells than in non-tumoral tissue. They were present in the same cells, suggesting that diazepam binding inhibitor may act in an intracrine manner in these cells. Higher peripheral benzodiazepine receptors and diazepam binding inhibitor expression in tumor cells suggest an implication of this system in the metabolism of neoplastic cells. Furthermore the evaluation of peripheral benzodiazepine receptor and diazepam binding inhibitor expression might be useful in evaluating malignancy and in diagnostic approaches of tumors in liver tissue.

May plasma cholesterol level be considered a neoplastic marker in liver disease from cirrhosis to hepatocellular carcinoma?

BACKGROUND: Although the role of cholesterol in tumourigenesis is unclear, it is used by the tumoural cells for biosynthetic processes and for steroid synthesis. AIM: To accertain whether plasma cholesterol levels might be a reliable neoplastic marker of a developing hepatocellular carcinoma in patients with liver cirrhosis. PATIENTS: Plasma cholesterol has been studied in 287 liver cirrhosis patients without hepatocellular carcinoma and in 132 patients with hepatocellular carcinoma. RESULTS: Cholesterol (mean +/- SEM) was higher in hepatocellular carcinoma patients when compared with age-, sex- and Child-Pugh class matched cirrhotic controls. In Child-Pugh class A, B and C with uncomplicated liver cirrhosis these values were, respectively, 142.0 +/- 2.5, 117.3 +/- 2.5, 97.4 +/- 2.9 vs 172.5 +/- 4.7, 163.8 +/- 7.9, 153.5 +/- 8.0 +/- mg/dl in patients with hepatocellular carcinoma (p < 0.001). A significant increase of cholesterol (p < 0.001) has been reported in the patients with liver cirrhosis when complicated by hepatocellular carcinoma and it was not related to cholestasis. CONCLUSIONS: This observation seems to suggest that the enhanced cholesterol biosynthesis by tumoural cells leads to a rise in plasma cholesterol of patients with cancer, and, moreover, that, this increase may be used as a neoplastic marker indicating the development of a tumour in patients with liver cirrhosis.

Tryptophan metabolism and hepatic encephalopathy. Studies on the sedative properties of oxindole.

Oxindole administration (1-100 mg/kg i.p.) to mammals decreases locomotor activity, reduces muscular tone and blood pressure and at larger doses causes coma and death. Utilizing both HPLC and GC/MS, we showed that oxindole is present in the blood, brain and other organs of several animal species, including humans. We demonstrated that oxindole is a tryptophan metabolite able to significantly decrease neuronal excitability by modifying the function of voltage-operated sodium channels. Its synthesis requires the availability of indole, which is formed in the gut. When liver function is impaired, a sufficient amount of indole reaches systemic circulation and is oxidized into oxindole, which seems to be one of the responsible agents for the neurological symptoms found in the course of liver impairment.

Up-regulation of peripheral benzodiazepine receptor system in hepatocellular carcinoma.

Increased number of peripheral benzodiazepine receptors (PBRs) have been found in some tumors outside the liver. The present study was to verify whether the PBR system is altered in hepatocellular carcinoma (HCC). The levels of endogenous benzodiazepine-like compounds (BZDs), measured by radioreceptor binding technique after HPLC purification and the endogenous ligand for PBRs, termed diazepam binding inhibitor (DBI), measured by radioimmunoassay utilizing a specific antibody for human DBI, were studied in the blood of 15 normal subjects, 12 liver cirrhosis and 10 patients with HCC. The levels of BZDs in serum were increased hundred fold in liver cirrhosis patients and slightly elevated in HCC patients. DBI was found to be increased in HCC patients. The binding recognition sites for PBRs (Bmax) were increased 4 to 7 fold in HCC tissue in comparison with that found in non-tumoral liver tissue (NTLT). On the contrary the concentrations of DBI were found to be significantly decreased in HCC tissue in comparison with the respective NTLT. These results seem to suggest an implication of PBRs and of their putative endogenous ligands in the metabolism of these neoplastic cells and possibly in their proliferation. The up-regulation of PBRs found in HCC tissue seems to indicate an increased functional activity of these receptors and opens up the possibility of new pharmacological and diagnostic approaches while the changes in the circulating endogenous ligands for the above receptors might be envisaged as early markers of tumorigenesis in liver cirrhosis.

Endogenous benzodiazepine-like compounds and diazepam binding inhibitor in serum of patients with liver cirrhosis with and without overt encephalopathy.

BACKGROUND/AIM: Despite some controversy, it has been suggested that endogenous benzodiazepine plays a role in the pathogenesis of hepatic encephalopathy. The aim of the present study was to evaluate the concentrations of endogenous benzodiazepines and the peptide, diazepam binding inhibitor, in the blood of patients with liver cirrhosis with and without overt encephalopathy, and to compare these levels with those of consumers of commercial benzodiazepines. SUBJECTS: Normal subjects (90), benzodiazepine consumers (14), and cirrhotic patients (113) were studied. METHODS: Endogenous benzodiazepines were measured by the radioligand binding technique after high performance liquid chromatography (HPLC) purification. The presence of diazepam and N-desmethyldiazepam was assayed by HPLC-electrospray tandem mass spectrometry. Diazepam binding inhibitor was studied in serum by radioimmunoassay. RESULTS: Endogenous benzodiazepines were below the limit of detection in 7% of patients with encephalopathy. When detectable, their levels were at least comparable with those of benzodiazepine consumers and correlated with the liver dysfunction but not the stage of encephalopathy. Serum levels of diazepam binding inhibitor tended to decrease when endogenous benzodiazepines levels increased. CONCLUSIONS: Endogenous benzodiazepines may accumulate in patients with liver cirrhosis during the course of the disease, and the phenomenon appears to be independent of the presence or absence of encephalopathy.

Diazepam binding inhibitor and total cholesterol plasma levels in cirrhosis and hepatocellular carcinoma.

Cholesterol is used by cells for biosynthetic processes and for steroid synthesis. Although the role of cholesterol in tumorigenesis is not clear it is known that steroids are important factors in human carcinogenesis. A polypeptide, diazepam binding inhibitor (DBI), which is an endogenous ligand for peripheral benzodiazepine receptors enhances steroidigenesis by promoting cholesterol delivery to the inner mitochondrial membrane which represents the rate-limiting step of steroid biosynthesis. We have assayed the total cholesterol (TC) and the DBI plasma concentrations in patients with liver cirrhosis complicated by hepatocellular carcinoma (HCC) in comparison with those of uncomplicated liver cirrhosis. TC and DBI levels have been studied in 73 cirrhotic patients and in 23 patients with HCC. Both TC and DBI levels were higher in HCC patients when compared with age, sex and Child-Pugh class matched cirrhotic controls. The values (mean+/-S.D.) in patients in Child-Pugh class B and C with and without HCC were respectively 128+/-30 mg/dl vs. 106+/-27 mg/dl (P < 0.01) and 2.05+/-0.78 pmol/ml vs. 0.78+/-0.84 pmol/ml (P < 0.0001). The data may be the result of the metabolic influence of tumors that enhances steroid biosynthesis during tumor proliferation.

Benzodiazepine-like compounds in the plasma of patients with fulminant hepatic failure.

BACKGROUND: Benzodiazepine-like compounds have been implicated in the pathogenesis of encephalopathy after fulminant hepatic failure. METHODS: The levels and the nature of benzodiazepine-like compounds were determined in six cases of fulminant hepatic failure during the course of the disease. Blood samples were collected on admission and a few days later, when the neurologic status had improved in five cases and immediately before death in one case. The compounds were measured in sera with a binding technique after high-performance liquid chromatography purification and analyzed with mass spectrometry. RESULTS: Their levels were highly variable in those with severe encephalopathy and were still increased on awakening in some cases. Diazepam and N-desmethyldiazepam were inconsistently present. CONCLUSIONS: The inconsistent presence of benzodiazepine-like compounds in encephalopathy after fulminant hepatic failure and their persistence, in some cases, at high levels on awakening from coma seem to indicate that the encephalopathy is not strictly dependent on the levels of these compounds.

Mesenteric vein thrombosis: a rare cause of abdominal pain in cirrhotic patients--two case reports.

Mesenteric vein thrombosis is a rare disorder which can develop rapidly with intestinal infarction or subacutely with abdominal pain due to intestinal ischemia. Despite the availability of modern diagnostic tools, which allow an early diagnosis in most cases, the mortality from this disease has not significantly diminished over the years. The problem is that the syndrome is rare and unusual and the clinical presentation is usually vague or confusing. Particularly in cirrhotic patients, this diagnosis requires the exclusion of several other complications of liver disease, like spontaneous bacterial peritonitis, tense ascites or portal thrombosis. Here, we report the occurrence of acute mesenteric vein thrombosis in two patients with liver cirrhosis. Severe subcontinuous abdominal pain out of proportion to the physical findings and abdominal distension were the major symptoms in both patients. Magnetic resonance imaging in one case and ultrasound scan with color Doppler followed by computed tomography in the other patient confirmed the diagnosis and enabled an appropriate early therapy to be undertaken.

Adult celiac disease and primary sclerosing cholangitis: two case reports.

The association of primary sclerosing cholangitis and celiac disease is uncommon. Herein, we report on 2 different cases which developed this association. Case 1 was a 59 year-old female who firstly complained of symptoms of cholestasis. The diagnosis of primary cholangitis was made on liver biopsy, and the endoscopic retrograde cholangiopancreatography (ERCP) showed narrowing and irregularity of the extra- and intrahepatic bile ducts. The results were positive for antiendomysial antibodies and the jejunal biopsy confirmed the coexistence of celiac disease, which was asymptomatic until that moment. The gluten-free diet ameliorated the index of cholestasis. Case 2 was an old man suffering from undiagnosed celiac disease for at least 5 years prior to admission at our Department. The diagnosis was based on the histological examination of a jejunal biopsy. The patient did not follow the gluten-free diet and was again admitted to our Department 6 years later with symptoms of cholestasis. The liver biopsy and ERCP confirmed the diagnosis of primary sclerosing cholangitis.

Antibacterial activity of rifaximin reduces the levels of benzodiazepine-like compounds in patients with liver cirrhosis.

Benzodiazepine-like compounds are present in trace amounts in the blood of normal subjects and increase in liver cirrhotic patients with or without encephalopathy. Their increased presence may, however, represent an occasional precipitating factor of hepatic encephalopathy. The source of these compounds is still unknown, but they are constituents of our diet since benzodiazepine receptor ligands have been described in plants, vegetables and in animals. They may also be synthesized, at least in part, by intestinal bacterial flora. In this article we report that the level of these compounds in the blood decreased by 40% after therapy with rifaximin, which reduces the aerobic and anaerobic intestinal bacterial flora. This observation indicates that intestinal bacterial flora is involved in the production of these compounds and that repeated short-term medications with this non-absorbable antibiotic may be useful in reducing the levels of benzodiazepine-like compounds in patients with liver cirrhosis.