Non-variceal gastrointestinal (NVGI) bleeding in cirrhosis may be associated with life-threatening complications similar to variceal bleeding.

To review NVGI bleeding in cirrhosis.

MEDLINE, Scopus, and ISI Web of Knowledge were searched, using the textwords "portal hypertensive gastropathy," "gastric vascular ectasia," "peptic ulcer," "Dieulafoy's," "Mallory-Weiss syndrome," "portal hypertensive enteropathy," "portal hypertensive colopathy," "hemorrhoids," and "cirrhosis."

Portal hypertensive gastropathy (PHG) and gastric vascular ectasia (GVE) are gastric lesions that most commonly present as chronic anemia; acute upper GI (UGI) bleeding is a rare manifestation. Management of PHG-related bleeding is mainly pharmacological, whereas endoscopic intervention is favored in GVE-related bleeding. Shunt therapies or more invasive techniques are restricted in refractory cases. Despite its high incidence in cirrhotic patients, peptic ulcer accounts for a relatively small proportion of UGI bleeding in this patient population. However, in contrary to general population, the pathogenetic role of Helicobacter pylori infection remains questionable. Finally, other causes of UGI bleeding include Dieulafoy's lesion, Mallory-Weiss syndrome, and portal hypertensive enteropathy. The most common non-variceal endoscopic findings reported in patients with lower gastrointestinal bleeding are portal hypertensive colopathy and hemorrhoids. However, the vast majority of studies are case reports and, therefore, the incidence, diagnosis, and risk of bleeding remain undefined. Endoscopic interventions, shunting procedures, and surgical techniques have been described in this setting.

The data on NVGI bleeding in liver cirrhosis are surprisingly scanty. Large, multicenter epidemiological studies are needed to better assess prevalence and incidence and, most importantly, randomized studies should be performed to evaluate the success rates of therapeutic algorithms.

610 patients of upper gastrointestinal haemorrhage were endoscoped over a period of eleven years from July 1985 to June 1996. Average age of the patients was 39.2 years. 82.6% were males and 17.4% were females. Duodenal ulcer (31.5%), erosive mucosal disease (30.8%), oesophageal varices (31.5%) and gastric ulcer (6.2%) were the major causes. Other causes included Mallory Weiss syndrome (10 patients), gastric polyp (3 patients), stomal ulcer (5 patients) and self-induced bleeding (3 patients). Multiple lesions responsible for bleeding were detectable in 6.6% of patients. Endoscopy was non-contributory in 50 (11.2%) patients. Haemorrhage was the first presentation in 8.5% patients of duodenal ulcer. A known ulcerogenic agent in 21% of duodenal ulcer cases precipitated the bleeding. 77.4% of duodenal ulcer patients responded to conservative management. Erosive gastritis (57.5%) was the commonest finding in the erosive mucosal group. Alcohol and analgesics were the major precipitating factors in these patients. Majority of oesophageal varices were treated by sclerotherapy. Mortality (20%) were highest in the oesophageal varices group.

Upper gastrointestinal (GI) bleeding is one of the most frequent causes of morbidity and mortality in the course of liver cirrhosis. The aim of this study was to determine the independent predictors of morbidity, mortality, and survival after the first episode of GI bleeding in patients with liver cirrhosis.

In a retrospective study of 403 cirrhotic patients who were admitted in the period January 1982 to December 1994 because of a first episode of GI hemorrhage, epidemiological factors, bleeding-related variables and cirrhosis-related variables that may be associated with hepatic and extrahepatic complications, mortality at 48 h and 6 weeks, and survival up to 30 June 1996 were assessed.

Forty-five percent of patients developed hepatic and/or extrahepatic complications, with a mortality rate of 7.4% at 48 h and 24% at 6 weeks. Renal failure, rebleeding, hepatocellular carcinoma, and hepatic encephalopathy were independent predictors of mortality. The Kaplan-Meier method showed a median survival of 30.9+/-4.5 months (95% confidence interval 22 to 39.7 months). The cumulative percentage of survivors was 60.2% at 1 year, 33.6% at 5 years, and 14% at 10 years. In a Cox's multiple regression analysis, age, hepatic encephalopathy, hepatocellular carcinoma, Child-Pugh grade, and renal failure were independently associated with long-term survival.

The first episode of GI bleeding in patients with liver cirrhosis is associated with high morbidity and mortality. Renal failure, rebleeding, hepatocellular carcinoma, and hepatic encephalopathy were independent risk factors for early death.

In the present article we describe updated information concerning the clinical feature of portal-hypertensive gastropathy (PHG), which is characterized by mucosal and submucosal vascular dilatation without inflammation. Although this lesion represents non-variceal bleeding, there is a wide variation of its prevalence. Portal pressure and some humoral factors may play important roles in its pathogenesis. Gastric acid secretory activity is reduced, whereas the gastric mucosal barrier is impaired. With regard to gastric mucosal haemodynamics, whether 'overflow' (i.e. active congestion) or 'stasis' (i.e. passive congestion) cause gastric mucosal hyperaemia is not known. A severe lesion is a potential source of bleeding, while mild lesions are of little clinical significance and endoscopic variceal obliteration aggravates PHG in some patients. In the treatment of PHG, pharmacological (e.g. propranolol), surgical (e.g. portosystemic shunt) and radiological (e.g. transjugular intrahepatic portosystemic shunt) procedures may be useful in preventing bleeding from PHG.

During the last 4 years, 147 patients suffering from portal hypertension with acute upper gastrointestinal bleeding were subjected to emergency endoscopy soon after they were resuscitated. Seventeen (11.5%) patients were referred to us with a clinical diagnosis other than portal hypertension. The causes of bleeding as seen during endoscopy were: oesophageal varices (n = 130; 88%), gastric varices (n = 11), gastric ulcer (n = 2) portal hypertensive gastropathy (n = 2) and erosive gastritis and duodenal ulcer in one patient each. All patient bleeding from oesophageal varices except one underwent emergency endoscopic sclerotherapy. One hundred and twenty-one (94%) stopped bleeding immediately. Rebleeding was seen in 11% and was effectively controlled by a second session of sclerotherapy in all but one patient. Twenty (14%) patients died. It is concluded that emergency endoscopy has a definite role in the management of patients with portal hypertension complicated by gastrointestinal bleeding.

This study was designed to evaluate the concentration and the regional distribution of TGF-alpha and EGF in normal and portal hypertensive human gastric mucosa. To this end we measured by RIA the gastric and duodenal concentration of TGF-alpha and EGF in subjects with chronic hepatitis, who had normal gastric endoscopic appearance, and in patients with liver cirrhosis with and without congestive gastropathy. Our results show that TGF-alpha concentration is significantly higher than EGF concentration in both the stomach and duodenum. No significant regional differences in the distribution of the two peptides were found. Moreover, the gastroduodenal tissue levels of TGF-alpha were comparable in subjects with and without hypertensive gastropathy. EGF gastric concentration was not altered in patients with congestive gastropathy. However, EGF duodenal tissue levels were significantly lower in patients with liver cirrhosis than in noncirrhotic subjects. We speculate that the higher level of TGF-alpha in the gastroduodenal mucosa may support the hypothesis that TGF-alpha and not EGF is the major physiological ligand for TGF-alpha/EGF receptor in the intact gut. Furthermore, the lower duodenal concentration of EGF in cirrhotics might partially explain the increased susceptibility of cirrhotic patients to duodenal ulcer.

To compare octreotide with injection sclerotherapy in the treatment of acute variceal haemorrhage, patients admitted with gastrointestinal bleeding and oesophageal varices confirmed by endoscopy were randomised to receive either emergency sclerotherapy with 3% sodium tetradecyl sulphate or octreotide (50 micrograms intravenous bolus plus 50 micrograms per h intravenous infusion for 48 h). At the end of the study period (48 h), the octreotide group also had sclerotherapy to obliterate the varices. 100 patients were recruited. Demographic features including the aetiology of portal hypertension and the Child-Pugh's grading of the two groups were similar. Bleeding was initially controlled in 90% of patients by emergency sclerotherapy and in 84% by octreotide infusion (95% confidence interval 0-19.5, p = 0.55). There were no significant differences between the two groups in early (within 48 h of randomisation) rebleeding (16% vs 14%), blood transfusion (3 units vs 3.5), hospital stay (5 days vs 6 days), or hospital mortality (27% vs 20%). No notable side-effects were associated with octreotide. We conclude that octreotide infusion and emergency sclerotherapy are equally effective in controlling variceal haemorrhage.

One hundred and seventy hospitalized patients with cirrhosis were included in a prospective and sequential study, to verify the prevalence and most frequent causes of bacterial infection. The differences in clinical and laboratory data between the two groups were analyzed: group I--80 patients who developed bacterial infection and group II--90 patients without bacterial infection. The prevalence or cumulative frequency of the development of bacterial infection during one hospitalization was 47.06%. Among these, the most frequent types of infection were: spontaneous bacterial peritonitis (SBP): 31.07%, urinary tract infection (UTI): 25.24% and pneumonia: 21.37%. Community infections were more frequent (56.25%) than nosocomial infections (32.50%) and they occurred sequentially in 11.25% of the cases. The agents responsible were gram negative bacteria in 72.34% of the cases. Clinical and biochemical parameters in bacterial infection were generally correlated with the severity of liver disease. Child-Pugh classification showed a predominance of class C in infected cirrhotic patients compared to non-infected ones. During hospitalization, the mortality rate of group I was 30% whereas in group II it was 5.55% (P = 0.0001). SBP and pneumonia were the most severe types of infection, with high mortality rates, 31.25% and 40.91%, respectively. These results indicate that bacterial infection is a severe complication in the course of cirrhosis.

Portal hypertensive gastropathy is a recently recognized important complication of cirrhosis. In the present study, the clinical features, portohepatic hemodynamics, and hepatic function were investigated in a series of 47 patients with cirrhosis. Mild gastropathy was found in 15 patients (32%) and severe gastropathy in 17 patients (36%). The presence of gastropathy seemed to be independent of age, sex, cause of cirrhosis, or grade of gastroesophageal varices. However, severe gastropathy was associated with an increase in portal venous pressure gradient (vs. control, P less than 0.01; vs. mild gastropathy, P less than 0.01), an increase in hepatic sinusoidal resistance (vs. control, P less than 0.01; vs. mild gastropathy, NS), and a decrease in hepatic blood flow (vs. control, P less than 0.01; vs. mild gastropathy, NS). In addition, patients with severe gastropathy had impaired metabolic activity of the liver, which was assessed by intrinsic clearance of indocyanine green (vs. control, P less than 0.01; vs. mild gastropathy, NS). These observations may have important therapeutic implications in patients with cirrhosis and portal hypertensive gastropathy.

Sixty-one patients with different degrees of liver failure, 23 with Child-Pugh class B and 38 with Child-Pugh class C, were studied and observed for 3 yr. Coagulation index analysis showed significantly lower values of prothrombin activity, more prolonged activated partial thromboplastin time, higher bilirubin and fibrinogen degradation products values in class C patients. Among all patients, 28 had fibrinogen degradation products values greater than 10 micrograms/ml, and in these patients a hyperfibrinolytic state was confirmed by higher values of circulating plasminogen activator antigen (17.3 +/- 8.7 ng/ml vs. 5.41 +/- 1.9 ng/ml; p less than 0.0001) and activity (6.6 +/- 2.1 IU/ml vs. 1.92 +/- 1.12 IU/ml; p less than 0.0001) and significantly lower plasminogen activator inhibitor antigen (6.4 +/- 3.5 ng/ml vs. 15.8 +/- 5.6 ng/ml; p less than 0.0001) and activity (3.6 +/- 2.2 IU/ml vs. 8.5 +/- 3.9 IU/ml; p less than 0.0001). Patients with positive fibrinogen degradation products had higher serum bilirubin (6 +/- 4 mg/dl vs. 2 +/- 2 mg/dl; p less than 0.0001) and lower fibrinogen (156 +/- 52 mg/dl vs. 194 +/- 62 mg/dl; p less than 0.02) than patients without hyperfibrinolysis. During the follow-up period, 41 patients died, 22 from fatal gastrointestinal hemorrhage and 19 from liver failure. Thirty patients experienced fatal (22 patients) and nonfatal (8 patients) gastrointestinal hemorrhage. Patients with positive fibrinogen degradation products or class C had a higher risk of gastrointestinal bleeding than patients with negative fibrinogen degradation products (odds ratio = 8) or class B (odds ratio = 3.5), respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Twenty patients with portal hypertension related to a variety of causes prospectively underwent colonoscopy for hematochezia (n = 10), hemoccult positive stool and anemia (n = 9), or polyp found with screening flexible sigmoidoscopy (n = 2) (includes 1 patient with anemia/heme-positive stool). Twelve patients (60%) had previously undergone a course of sclerotherapy, and 10 (50%) had endoscopic evidence of congestive gastropathy. Colonoscopic findings included mucosal abnormalities resembling multiple vascular ectasias in 14 (70%), 4 of whom also had endoscopic features suggesting a mild, chronic colitis. Neither signs of chronic liver disease nor stigmata suggestive of more severe portal hypertension correlated with the colonoscopic findings. Two patients required heater probe therapy for actively oozing lesions resembling vascular ectasias and an additional two patients sclerotherapy for bleeding midrectal varices. Although likely an overestimate of the frequency, this study suggests that portal colopathy can occur in portal hypertension. Vascular ectasialike lesions in such settings may be associated with acute as well as chronic gastrointestinal bleeding and may require pharmacological, directed endoscopic, or portal decompressive therapy. Additional studies are required to determine not only the pathophysiology but also the true frequency of this entity.

Variceal bleeding has a high mortality, as the majority of patients have cirrhosis, with hepatic coma, renal failure, ascites and clotting deficiencies as complicating factors. Bleeding varices must therefore be treated as an emergency. Resuscitation, endoscopic diagnosis and haemostasis are the cornerstones of treatment. Once bleeding varices have been identified, attempts to stop the bleeding must be made at once as this will lessen the chances of hepatic failure developing. Endoscopic sclerotherapy at the time of diagnosis is the best available treatment at present, although profusely bleeding varices can be difficult to see and inject. In these circumstances the passage of a Sengstaken tube should stop the bleeding, allowing later sclerotherapy to be successful. If rebleeding recurs and cannot be controlled, oesophageal transection with a stapling gun may be life-saving, although the varices may later recur and long-term endoscopic follow-up will be necessary. Portacaval shunting and the distal splenorenal shunt involve arduous surgery and are followed by a significant incidence of hepatic encephalopathy; they should be reserved for those few cases when simpler measures have failed, although shunts do lead to permanent decompression of the portal system. The acute variceal bleed may also be dealt with pharmacologically. Vasopressin, used in combination with nitroglycerin to lessen the harmful side-effects, is cheaper and as effective as terlipressin or somatostatin and its synthetic analogue octreotide. Several courses of injection sclerotherapy will be required to eliminate oesophageal varices. Thereafter, long-term follow-up will be necessary to deal with any recurrence. The place of non-selective beta-blockers is still contentious, but they do reduce portal pressure and may lessen the chance of rebleeding. There is also a growing role for hepatic transplantation, which not only eliminates the varices but also restores liver function to normal and greatly reduces the risk of subsequent hepatoma development.

Gastric mucosal lesions are common in patients with cirrhosis. Among them, snake skin pattern gastropathy (SSPG) is the most distinguishing one. A prospective study was conducted to investigate the incidence of SSPG in cirrhotic patients, the relationship between the degree of portal pressure and SSPG, and the possible association of SSPG with serum levels of gastrin and pepsinogen I. SSPG was found to be significantly more common in 100 cirrhotic patients than in 100 age- and sex-matched healthy controls (41% vs 0%, P less than 0.0001). Hepatic venous pressure gradient and serum gastrin and pepsinogen I levels were measured in 21 cirrhotic patients with SSPG and 25 cirrhotics without SSPG. There was no significant difference in hepatic venous pressure gradient (16.1 +/- 4.4 mmHg vs 16.1 +/- 4.9 mmHg, P greater than 0.05), serum gastrin level (78.0 +/- 26.7 pg/mL vs 80.1 +/- 32.5 pg/mL, P greater than 0.05) and serum pepsinogen I level (69.5 +/- 26.6 ng/mL vs 65.2 +/- 26.1 ng/mL, P greater than 0.05) in cirrhotic patients with or without SSPG. In conclusion, SSPG is common in cirrhotic patients. Portal pressure per se may not be the only factor causing SSPG--other aggressive factors may be needed together to cause the gastropathy. There is no evidence of correlation between serum gastrin or pepsinogen I level and SSPG.

The present study was intended to assess the incidence and the nature of gastric red spots (GRS) in patients with liver cirrhosis. Endoscopically, GRS was more frequently observed in patients with cirrhosis (n = 146) than those without cirrhosis (n = 103) (43.2% vs. 4.8%; P less than 0.001). There was no relationship between the incidence of GRS and the severity of cirrhosis or the size of varices. Portal venous pressure was higher in cirrhotics with GRS (n = 21) than those without GRS (n = 25) (33.7 +/- 6.0 mmHg vs. 28.2 +/- 4.8 mmHg; P less than 0.001). Morphometric analysis using the biopsied specimens was made in 16 cirrhotics with GRS, 12 cirrhotics without GRS, and 15 non-cirrhotics. The capillary bed occupation ratio (vascular area/mucosal area) was higher in cirrhotics with GRS (9.3 +/- 4.0%) than those without GRS 84.1 +/- 1.1%) or the non-cirrhotics (3.4 +/- 9.8%) (P less than 0.005, P less than 0.005), while there was no significant differences in the number of capillaries per unit area. Infiltrating inflammatory cell count was similar among the three groups. Extravascular red blood cell count per unit area was higher in cirrhotics with GRS (29.7 +/- 31.4) than those without GRS (5.4 +/- 5.1) or non-cirrhotics (5.4 +/- 6.3) (P less than 0.05, P less than 0.01). Furthermore, extravasation of red blood cells through defective portion of the endothelium and interposition of red blood cells in interepithelial spaces were observed electron microscopically in cirrhotics with GRS.(ABSTRACT TRUNCATED AT 250 WORDS)

Upper gastrointestinal hemorrhage is one of the more important complications of cirrhosis and a major cause of death in such patients. The main sites of bleeding are esophageal varices, gastritis, and peptic ulcers. In order to determine the prevalence of either potential bleeding lesions or of other endoscopic findings in hemodynamically stable individuals with various etiologies of cirrhosis, 510 consecutive cirrhotic patients, evaluated for possible orthotopic liver transplantation (OLTx) underwent an upper gastrointestinal endoscopy for combined diagnostic and therapeutic purposes. The patients were divided into two main groups: 319 patients with parenchymal liver disease and 191 patients with cholestatic liver disease. Gastritis was found significantly more often in patients with parenchymal liver disease than in those with cholestatic liver disease (49.8% vs 30.9%; P less than 0.001). In contrast, the prevalence of esophagitis, esophageal and gastric varices, gastric ulcer, duodenal ulcer, and duodenitis was similar in both groups. Normal endoscopic findings were present in 5.0% of the parenchymal group and 11.5% of the cholestatic group (P less than 0.02). Ascites and encephalopathy were found significantly more often in subjects with parenchymal liver disease as compared to those with cholestatic liver disease. Portal hypertension and its degree, as assessed by the presence and size of esophageal varices, was similar in both groups, and in both groups there was a statistically significant qualitative trend of increasing prevalence of esophageal varices with increasing severity of disease as estimated using Pugh-Child's criteria.

The prognostic value of plasma prekallikrein activity, prothrombin time, and serum albumin with regard to survival in chronic liver insufficiency was evaluated in 21 consecutive patients. Twenty patients had liver cirrhosis, and one patient had malignant liver disease (hepatocellular carcinoma). Eight patients died between 4 and 43 days after the time of blood sampling. These patients had a prekallikrein value less than 0.42. There were no overlapping prekallikrein values between patients who died and patients who survived (overlap index 0; p less than 0.001). Overlap index for prothrombin time was 0.35 (p less than 0.02), and for serum albumin 0.34 (p less than 0.02). In conclusion, plasma prekallikrein seems to indicate whether death is imminent in patients with liver insufficiency due to cirrhosis. Longitudinal studies of prekallikrein activity in different subgroups of patients with chronic and acute liver disease are recommended.

Factor VII and prekallikrein activities were studied in 37 patients with liver cirrhosis who were in a decompensated state. Sixteen of them died 30-70 days after admission; 21 survived and were discharged after 30-80 days. Seven who died and six survivors had signs of hyperfibrinolysis: factor VII activity differentiated the two groups independently of the presence of hyperfibrinolysis. The presence of hyperfibrinolysis significantly reduced prekallikrein activity, which did not differentiate clearly survivors from non-survivors. Long term follow up of survivors showed a good correlation between factor VII and prekallikrein activities with long term survival. Hyperfibrinolysis seemed to influence the clinical course of patients: 87% of patients with hyperfibrinolysis who died had fatal haemorrhagic episodes. Low factor VII activity may be a precursor of terminal liver insufficiency.

To characterize bleeding from gastric red spots in patients with cirrhosis, three groups of patients were studied: (a) 11 cirrhotic patients bleeding from gastric red spots, (b) 18 nonbleeding cirrhotic patients without gastric red spots, and (c) 13 noncirrhotic patients with endoscopic normal mucosa (controls). Histologic examination of antral biopsy specimens revealed a diffuse capillary ectasia without inflammation in 8 of the 11 cirrhotic patients with gastric lesions. Morphometric analysis disclosed a significantly greater mean mucosal capillary cross-sectional area in cirrhotic patients with gastric lesions (mean +/- SE, 1371 +/- 320 microns2) than in those without gastric lesions (541 +/- 61 microns2) (p less than 0.005) or controls (353 +/- 20 microns2) (p less than 0.001). Hypergastrinemia was detected in 8 of the 11 cirrhotic patients with lesions, in 2 of the 18 cirrhotic patients without gastric lesions, and in none of the controls (p less than 0.001). Gastrin serum levels correlated significantly (r = 0.80) with mean mucosal capillary cross-sectional area in patients with cirrhosis. Pepsinogen I serum levels below 20 ng/ml were observed in 7 of the 11 cirrhotic patients with lesions, in 1 of the 18 cirrhotic patients without lesions, and in none of the controls. These data indicate that bleeding from gastric red spots in patients with cirrhosis is a distinct entity characterized by vascular ectasia of the gastric mucosa. This condition seems to be associated with hypergastrinemia and low serum levels of pepsinogen I.

The amount of prostaglandin E2 in endoscopic biopsy specimens from the mucosa of the gastric corpus and antrum of 10 healthy volunteers and 8 patients with alcoholic cirrhosis (7 of whom had portal hypertension) was measured by radioimmunoassay. The mean prostaglandin E2 level in both corpus and antral mucosa was significantly lower in the cirrhotic patients (p less than 0.025). Release of prostaglandin E2 into the gastric cavity was also significantly less in the cirrhotic patients in the basal state (p less than 0.025) and in that of tetragastrin stimulation (4 micrograms/kg, s.c.; p less than 0.01). There was no significant difference in acid secretion between the two groups of subjects. The histologic findings of biopsy specimens showed no difference in the extent of chronic gastritis between the cirrhotic patients and healthy subjects. Therefore, this deficiency of prostaglandin E2 in the gastric mucosa of cirrhotic patients was not related to acid secretion or chronic gastritis.

The activities of Hageman factor, high molecular weight kininogen (HMWK), and prekallikrein were studied in patients who had chronic active hepatitis and cirrhosis. Serum HMWK and prekallikrein activities were decreased in chronic active hepatitis and cirrhosis, but Hageman factor activity was low in cirrhosis only. The reduction of prekallikrein, HMWK, and Hageman factor was dependent on the degree of liver failure. Similar prekallikrein values were found in serum samples, activated or not, with an excess of Hageman factor and HMWK, which suggests that the decrease of prekallikrein in liver disease is not influenced by the simultaneous decrease of Hageman factor and HMWK.

The frequency of infection at the time of admission with upper gastrointestinal haemorrhage has been determined in 149 successive cirrhotic patients admitted to an intensive care unit. Infection status was investigated by clinical examination, chest X-ray, and blood, urine and ascitic fluid culture. At initial examination infection was present in 32 patients (22 per cent) and was often in the form of septicaemia or spontaneous peritonitis; the bacteria responsible were frequently digestive in origin. At endoscopy, acute lesions of gastroduodenal mucosa were more frequent among infected patients, whereas gastro-oesophageal varices and chronic gastroduodenal ulcers were more frequent among the non-infected patients. Acute mucosal lesions were observed in 70 per cent of infected patients and in 19 per cent of non-infected patients. The mortality rate was higher in infected patients. Infection and the frequency of acute mucosal lesions were related to the severity of the cirrhosis. It is suggested that these lesions could be due to stress secondary to infection.

It is controversial whether the occurrence of ascites and gastrointestinal bleeding in cirrhosis is related to the severity of portal hypertension. Portal pressure was examined in 124 unselected patients with portal hypertension due to chronic liver disease to evaluate this issue. Portal pressure was less in patients without complications of chronic liver disease (11.7 +/- 3.0 mmHg, n = 16) as compared to patients who had bled from varices or erosive gastritis (16.6 +/- 3.4 mmHg, p less than 0.001, n = 49), who had ascites (16.2 +/- 3.0 mmHg, p less than 0.001, n = 78) or both (16.5 +/- 3.0 mmHg, p less than 0.001, n = 19). Portal pressure was similar in patients bleeding from varices and erosive gastritis (16.7 +/- 3.4 mmHg, n = 43; vs 16.2 +/- 4.0 mmHg, n = 6, respectively) and in patients with refractory and nonrefractory ascites (16.2 +/- 3.5, n = 21; vs 16.2 +/- 3.5 mmHg, n = 57). The lowest portal pressure recorded in a patient with variceal bleeding was 9.0 mmHg. The lowest portal pressure recorded in a patient with ascites was 8.0 mmHg. Esophageal varices (graded 0-4 at endoscopy) were larger in patients with a history of bleeding from esophageal varices as compared to patients without such a history (3.2 +/- 0.7 vs 2.0 +/- 0.9, p less than 0.001). Serum albumin concentration was greater in patients without ascites as compared to patients with ascites (33 +/- 5 vs 26 +/- 5 g/l p less than 0.001) but was similar in patients with refractory and nonrefractory ascites (25 +/- 7 vs 26 +/- 5 g/l, respectively).(ABSTRACT TRUNCATED AT 250 WORDS)

To clarify the role of endotoxaemia and congestion of the stomach in the development of acute haemorrhagic gastritis in cirrhotic patients and to investigate the mechanisms of gastric mucosal haemorrhage, the present study was undertaken using rats. Congestion of the stomach was produced by the ligation of gastric veins. Congestion of the stomach or endotoxaemia could not produce gastric mucosal haemorrhage by itself. However, petechial haemorrhage was induced when endotoxin was given to the rats with congestion of the stomach, and the gastric mucosal haemorrhage was largely prevented by administration of gabexate mesilate, an anti-kallikrein drug. Administration of bromelain, which releases prekallikrein and high molecular weight kininogen, instead of endotoxin, also induced gastric mucosal haemorrhage. These findings suggest that the cause of acute haemorrhagic gastritis may be the coexistence of endotoxaemia and congestion of the stomach due to liver cirrhosis and portal hypertension. The mechanisms of the haemorrhage may be as follows: Endotoxin-induced bradykinin acts on the dilated capillaries and small veins in the mucosa and markedly increases their permeability.

A patient with small bowel capillary dilatation and cirrhosis is reported. This patient had persistent, unexplained gastrointestinal bleeding. Small bowel capillary dilatation appears to be unique to patients with portal hypertension. The possible role of small bowel capillary dilatation in causing gastrointestinal bleeding is discussed.

A series of clinical steps in the severity of chronic liver disease (CLD) has been arranged to represent the case history of chronic active hepatitis and cirrhosis retrospectively. The 'steps' in such series of clinical stages have been chosen on the basis of physiopathological considerations, the state of health of the patients, and their ability to work. For every 'step' the laboratory findings were studied in order to verify the hypothesis that a laboratory case history exists, which is parallel to the clinical history of CLD. With the exception of serum albumin and, to a lesser extent, sulfobromophthalein-ki, which seem to reflect the progressive deterioration in hepatocellular function, the results suggest that most conventional tests do not seem to be of any value in monitoring a CLD patient in cases where the diagnosis is already known, whereas a physical examination provides a physician with more, and cheaper, information than laboratory tests.

Effects of subcutaneous calcium-heparin and vitamin K administration were studied in 30 cirrhotic patients showing low values of prothrombin time, antithrombin III, fibrinogen, platelet count, plasminogen, alpha 2-antiplasmin, raised levels of fibrin(ogen) degradation products and prolonged activated partial thromboplastin time. A group of 10 patients was first treated with K vitamin for 15 d; after vitamin K therapy interruption, a treatment with 5000 IU (8000 IU in 1 patient) every 12 h of subcutaneous calcium-heparin was started. In another group of 20 patients a treatment with 5000 IU (8000 IU in 2 patients) every 12 h of subcutaneous calcium-heparin was started immediately. The heparin administration in both groups had been performed for at least 2 weeks. No significant changes of blood coagulation picture were observed after vitamin K administration, while calcium-heparin treatment showed an increase in prothrombin time, fibrinogen, platelet count, plasminogen, alpha 2-antiplasmin, a decrease in fibrin(ogen) degradation products and a shortened activated partial thromboplastin time. There was no significant change in antithrombin III values.

One hundred and thirty-one patients underwent clinical and biological investigation with the following determinations performed on the same day; presence or absence of ascites, icterus and/or encephalopathy, coagulation study, biochemical determinations including albumin, transferrin and immunoglobulins immunoassays. The principal component analysis of biological data showed two sets of highly representative and inversely correlated data; one included coagulation tests, albumin and transferrin, and the other included immunoglobulin A/transferrin ratio, immunoglobulin A and total bilirubin. Clinical and biological data were computed using discriminant analysis between dead and survivors. Six parameters were then selected (total bilirubin, encephalopathy, factor V, AST, antithrombin III and transferrin) giving a correct prognosis in 81.6% (31/38) of cases in a test sample. Neither ascites nor immunoglobulins were useful for the estimation of the prognosis.

Sixteen clinical and biological variables were recorded in 116 episodes of upper digestive tract hemorrhage of known cause in cirrhotic patients. One-dimensional analysis reveals a significant correlation between six variables and the rupture of esophagogastric varices, whereas multiple linear regression and partial correlation analysis reduces the significant variables to two: a history of digestive hemorrhage and the nonalcoholic etiology of the cirrhosis. A value of the discriminant function exists for which the specificity and, consequently, the positive predictive value are equal to 100%, but with a sensitivity of 39%. These results mean that, in an explanatory approach, partial correlation analysis seems to constitute an indispensable complement to analysis of clinical and biological variables, since it reduces the chances of unwarranted explanatory interpretation. However, in a pragmatic approach, the recording of 16 variables does not permit a clear discrimination between ruptured varices and nonruptured varices; this suggests that either other factor(s) remain to be discovered or else that those related to ruptured varices and to acute ulcerations are the same.