Spontaneous regression of oesophageal varices after long-term conservative treatment. Retrospective study in 20 patients with alcoholic liver cirrhosis, posthepatitic cirrhosis and haemochromatosis with cirrhosis

Spontaneous regression of oesophageal varices in liver cirrhotics without sclerotherapy or shunt operation has only been known in alcoholic cirrhosis after alcohol abstinence. Therefore, 20 liver cirrhotics of different aetiologies were controlled over 13 years (six alcohol, nine hepatitis, five haemochromatosis). Under strict alcohol abstinence, all underwent treatment with lactulose and ammonia-reducing amino acids to improve the urea synthesis in the liver. Since gastrointestinal bleeding was not observed, neither sclerotherapy nor shunt operation were performed. Initially, all patients had oesophageal varices (nine stage III, three stage II-III, eight stage II). Following conservative therapy, eight cirrhotics showed total regression and twelve showed stage I-II. Their Child-Pugh index, and urea synthesis rate improved significantly. Possible causes for the spontaneous regression of oesophageal varices are strict abstinence from alcohol, spontaneous seroconversion in six posthepatic B-cirrhoses and consequent phlebotomy in haemochromatosis.     

Stimulation of insulin secretion by medium-chain triglycerides in patients with cirrhosis

Oral medium-chain triglycerides were given to 10 normal volunteers, 12 cirrhotics (group I) without and 28 cirrhotics (group II) with abnormal portal systemic communications (ascites, splenomegaly, oesophageal varices, or surgically-created portacaval shunts). After 30 ml of medium-chain triglyceride oil there was no appreciable change in serum glucose levels in any of the three groups nor in serum insulin levels in the normals and in cirrhotics in group I. However, there was a significant increase in serum insulin levels in the cirrhotic patients in group II. It is suggested that the rise in serum insulin levels after medium-chain triglycerides noted in the cirrhotics with shunts is due to shunting of insulin-containing portal blood around the liver (anatomical shunts) and to a diminished hepatic cell mass capable of extracting insulin (functional shunt). This differential response of serum insulin levels to medium-chain triglycerides may prove to be of value in detecting the presence of abnormal portal systemic communications in cirrhotic patients.     

Alkaline gastro-oesophageal reflux: dual probe pH monitoring.

Although the aetiology of Barrett's oesophagus or columnar line oesophagus (CLO), remains unknown, bile reflux has been implicated as a factor in its pathogenesis. This study aimed to detect alkaline reflux in gastro-oesophageal reflux patients using dual probe pH monitoring. Thirty patients with histologically diagnosed CLO, 15 age and sex matched patients with oesophagitis (grade 1-3), and 15 healthy volunteers were studied by dual probe, 18 hour pH monitoring and analysis of the bile acid content of oesophageal refluxate. Total acid exposure and acid exposure in the upright and supine postures were greater in CLO subjects than in oesophagitis patients and controls. Furthermore, the number of reflux episodes lasting more than five minutes and the duration of the longest reflux episode were significantly greater in the CLO subjects than the oesophagitis and control subjects. Nine subjects with CLO and oesophagitis, however, were not identified as refluxers, although six had a bile acid concentration in their oesophageal aspirate higher than the 95th centile value of the controls. There was no correlation between the oesophageal pH and the bile acid contents of refluxate. It is concluded that dual probe pH monitoring is not useful in detecting alkaline refluxers. pH monitoring, although the only subjective test available to identify acid refluxers, is not a sufficiently sensitive test with which to define alkaline reflux.     

Effects of endoscopic variceal treatment on oesophageal function: a prospective, randomized study

AIM : Endoscopic methods are currently the most widely used techniques for the treatment of bleeding oesophageal varices (BOV). However, a number of complications may limit their usefulness. We conducted a prospective, randomized comparison of variceal ligation versus sclerotherapy in cirrhotics after the control of variceal haemorrhage to study the relative short-term risks of these two procedures with respect to oesophageal motility and gastro-oesophageal reflux. METHODS : Seventy-three patients with established cirrhosis and an episode of variceal bleeding controlled by one session of endoscopic therapy were randomized to treatment with sclerotherapy or ligation until variceal eradication. In 60 of these patients, oesophageal manometry and 24-h intra-oesophageal pH monitoring were performed at inclusion and 1 month after variceal eradication. RESULTS : After variceal eradication with sclerotherapy, peristaltic wave amplitude decreased from 76.2 +/- 14.7 mmHg to 61.6 +/- 17.7 mmHg (P = 0.0001), simultaneous contractions increased from 0% to 37.9% (P = 0.0008), and the percentage of time with pH < 4 increased from 1.60 +/- 0.25 to 4.91 +/- 1.16% in channel 1 (P = 0.0002) and from 1.82 +/- 0.27 to 5.69 +/- 1.37% in channel 2 (P = 0.0006). In contrast, the above parameters were not disturbed with ligation. CONCLUSION : Our data define the advantages of ligation over sclerotherapy with respect to post-treatment oesophageal dysmotility and associated gastro-oesophageal reflux.     

Sleep and nocturnal acid reflux in normal subjects and patients with reflux oesophagitis.

Nocturnal gastro-oesophageal reflux may be important in the pathogenesis of reflux oesophagitis. This study aimed to determine whether: (1) gastro-oesophageal reflux occurs during sleep in patients with reflux oesophagitis and, if so, to explore the mechanism, and (2) the sleep pattern of patients with oesophagitis is different from that of control subjects. After a standard evening meal, simultaneous manometric, oesophageal pH, and polysomnographic recordings were obtained in 11 patients with endoscopic oesophagitis and 11 control subjects. Patients with gastrooesophageal reflux disease had significantly more total reflux episodes throughout the nocturnal monitoring period than control subjects (105 v 6). Ninety two of 105 episodes of gastro-oesophageal reflux in patients occurred during the awake state and 10 during sleep stage II. A number of reflux episodes occurred during brief periods of arousal from the various sleep stages. Of the 105 reflux events recorded in patients, 42 were induced by transient lower oesophageal sphincter relaxation, 20 by stress reflux, 22 by free reflux mechanisms, and in 21 the mechanism was unclear. The sleep pattern and the time spent in each sleep stage was not different between the two groups. It is concluded that the awake state is crucial for the occurrence of nocturnal reflux episodes in normal subjects as well as in patients with reflux oesophagitis and that the difference between the frequency of gastro-oesophageal reflux between normal subjects and patients cannot be explained by different sleep patterns.     

Hypogonadism is not related to the etiology of liver cirrhosis

We investigated the clinical and laboratory findings of hypogonadism and feminization in male patients with viral or alcoholic cirrhosis to determine whether chronic liver disease plays a primary role in the development of sexual dysfunction and hormonal changes. Two groups of male patients with liver cirrhosis (23 alcoholic, 33 viral) age-and Child's gradematched, and 20 age-matched healthy men, as a control group, were included in this study. Clinical signs of hypogonadism and feminization were examined in the cirrhotic patients. Follicle-stimulating hormone, luteinizing hormone, prolactin, testosterone, free testosterone, estradiol, androstenedione, dehydroepiandrosterone sulfate, and sex hormone-binding globulin were estimated in all groups. Seminal fluid was also analyzed in 7 alcoholic and 15 viral cirrhotics. Serum levels of estradiol, androstenedione, and sex hormone-binding globulin were significantly higher, and free testosterone and dehydroepiandrosterone sulfate levels were significantly lower in both groups of cirrhotics compared with the control group. Child's C patients in both groups of cirrhotics were found to have higher estradiol and lower free testosterone levels than child's A and B patients. Alcoholic and viral cirrhotics had markedly reduced sperm motility and density. The differences between alcoholic and viral cirrhotic patients in the clinical signs of hypogonadism, serum levels of sex steroids, and the results of seminal fluid analysis were not statistically significant. These findings suggest that liver cirrhosis per se, independent of etiology, causes hypogonadism and feminization, and that the degree of hypogonadism and feminization correlates well with the severity of liver failure.     

Gonadal dysfunction and changes in sex hormones in postnecrotic cirrhotic men: a matched study with alcoholic cirrhotic men.

To investigate the gonadal dysfunction and changes in sex hormones in male patients with postnecrotic cirrhosis, and to compare them with those in alcoholic cirrhotic men, three age-matched groups of men (hepatitis B virus-related postnecrotic cirrhosis 27, alcoholic cirrhosis 21, normal controls 30) were studied. Twelve of the 21 (57%) alcoholic cirrhotics and 16 of the 27 (59%) postnecrotic cirrhotics had a history of impotence. Both alcoholic and postnecrotic cirrhotic patients had significantly lower basal testosterone, but higher estradiol and prolactin levels than the control group (p less than 0.05). However, no differences were noted between the two cirrhotic groups. The degree of reduced testosterone and increased prolactin levels correlated with the severity of the cirrhosis. Despite the low testosterone concentration, basal levels of follicle stimulating hormone (FSH) and luteinizing hormone (LH) were not increased in the cirrhotic patients. All the three groups studied had normal FSH and LH responses to the stimulation of exogenous gonadotropin releasing hormone. On the basis of these results, we conclude that: (1) impotence and low testosterone level are not infrequent findings in men with hepatitis B virus-related postnecrotic cirrhosis, especially in those with decompensated liver function. (2) The liver disease per se is important for the development of male sexual dysfunction. (3) The derangement of hypothalamic-pituitary function may play a role in the sexual dysfunction and changes in sex hormones in male patients with cirrhosis.     

Characteristics, prognosis and outcome of patients with oesophageal varices in a university hospital in Sweden 1994-1999

OBJECTIVE: Patients with liver cirrhosis, portal hypertension and oesophageal varices are known to have high morbidity and mortality. The knowledge of incidence, aetiology and outcome in Sweden in recent years is limited. MATERIAL AND METHODS: All patients with oesophageal varices diagnosed for the first time at Sahlgrenska University Hospital during the 6-year period 1994-1999 were retrospectively studied. Information about the aetiology of liver cirrhosis and oesophageal varices, as well as about the proportion of bleeding and non-bleeding varices, endoscopic and pharmacological treatment and outcome, was analyszed. RESULTS: 312 patients were retrieved, 297 with liver cirrhosis (197 diagnosed before first bleeding (P), 92 after bleeding (B) and 8 at autopsy) and 15 with portal vein thrombosis without cirrhosis. Fifty-four percent had alcoholic liver disease. Fifty-five percent in group B and 13% in group P had at least one bleeding episode during follow-up (p<0.001). There was no significant difference in survival between groups B and P. Twenty-six percent of the cirrhotics died of liver failure and 19% from variceal bleeding. In a multivariate analysis, variables predicting mortality were: Child-Pugh class, group B, age and bilirubin levels. CONCLUSIONS: Variceal bleeding is still a strong risk factor for recurrent bleeding, but few die from their first bleeding. This concurs with studies indicating declining mortality from variceal bleeding. However, this patient group still has a high mortality from other causes.     

Hepatic venous pressure gradient determination in patients with hepatitis C virus-related and alcoholic cirrhosis

OBJECTIVES: Few data exist regarding the degree of portal hypertension in hepatitis C virus (HCV)-related cirrhosis, as the majority of studies have included mainly patients with alcoholic cirrhosis. This study was aimed at comparing the severity of portal hypertension in patients with HCV-related or alcoholic cirrhosis. METHODS: In total, 59 cirrhotic patients with portal hypertension (HCV-related in 34 cases and alcoholic in 25) underwent main right hepatic vein catheterization, with determination of the wedged and free hepatic venous pressures, and of hepatic venous pressure gradient (HVPG). RESULTS: HVPG values did not differ between the two groups of patients (19.4 +/- 6.0 mmHg vs 18.5 +/- 3.5 mmHg; P = 0.51). The prevalence and degree of oesophageal and gastric varices and portal hypertensive gastropathy did not correlate with the aetiology. Patients with viral cirrhosis had a lower prevalence of previous bleeding than those with alcoholic cirrhosis, despite a similar proportion of large varices in the two groups and similar HVPG levels. In both groups of patients, HVPG did not differ between patients with previous bleeds and those without. CONCLUSIONS: The degree of portal hypertension in cirrhotic patients does not correlate with the cause of the disease. Thus, current statements on the management of portal hypertension, although based upon studies including mainly patients with alcoholic cirrhosis, can be applied also to patients with viral-related cirrhosis.     

Development of risky varices in alcoholic cirrhosis with a well-maintained nutritional status

AIM: To compare the nutritional status between alcoholic compensated cirrhotic patients and hepatitis C virus(HCV)-related cirrhotic patients with portal hypertension.METHODS: A total of 21 patients with compensated cirrhosis(14 with HCV-related cirrhosis and seven with alcoholic cirrhosis) who had risky esophageal varices were investigated. In addition to physical variables, including the body mass index, triceps skinfold thickness, and arm-muscle circumference, the nutritional status was also assessed using the levels of pre-albumin(pre-ALB), retinol-binding protein(RBP) and non-protein respiratory quotient(NPRQ) measured with an indirect calorimeter.RESULTS: A general assessment for the nutritional status with physical examinations did not show a significant difference between HCV-related cirrhosis and alcoholic cirrhosis. However, the levels of pre-ALB and RBP in alcoholic compensated cirrhotic patients were significantly higher than those in HCV-related compensated cirrhotic patients. In addition, the frequency of having a normal nutritional status(NPRQ ≥ 0.85 and ALB value 3.5 g/d L) in alcoholic compensated cirrhotic patients was significantly higher than that in HCV-related compensated cirrhotic patients.CONCLUSION: According to our small scale study, alcoholic compensated cirrhotic patients can develop severe portal hypertension even with a relatively well-maintained liver function and nutritional status compared with HCV-related cirrhosis.     

Is scintigraphy of value in the diagnosis of gastrooesophageal reflux disease?

One hundred and ten patients with suspected oesophageal symptoms were investigated by means of oesophageal endoscopy (OE), 24-h pH-metry, and oesophageal scintigraphy (ES). When 24-h pH-metry formed the basis for diagnosis of gastrooesophageal reflux disease (GERD), the sensitivity for ES at abdominal compression was 64%, but no statistically significant differences were found among erect refluxers (ER), supine refluxers (SR), and combined refluxers (CR). Only 4% of the GERD patients had pathologic oesophageal clearing at ES. The more severe the macroscopic oesophagitis found by OE, the more pronounced were the abnormal findings at 24-h pH-metry and at ES with abdominal compression. Increased postprandial reflux was associated with gastro-oesophageal reflux and hiatal hernia at ES with abdominal compression and the most severe form of oesophagitis, respectively. It was concluded that ES had too low sensitivity to be recommended as a screening test for GERD. Nevertheless, the specificity of 76% can to some extent help us to rule out GERD in patients.     

Latent portasystemic encephalopathy

Forty patients with chronic liver disease and portal hypertension but without clinical signs of portasystemic encephalopathy (15 patients with nonalcoholic cirrhosis, 15 patients with alcoholic cirrhosis, and 10 patients with minimal EEG changes) and a control group of 12 patients with chronic alcoholic pancreatitis were studied using an extensive psychometric program, which, in the same form, is used for expert reports on driving capacity. Of the cirrhotic patients, 60% were considered unfit to drive; in 25% driving capacity was questionable, 15% (only nonalcoholic cirrhotics) were considered fit to drive. In contrast 75% of the patients with alcoholic pancreatitis were considered fit to drive. Major defects were found only in three heavy alcoholics. Patients with alcoholic cirrhosis scored lower than patients with nonalcoholic cirrhosis. This was due, to differences in liver function rather than to the effect of alcohol consumption. Patients with minimal EEG changes were practically all considered unfit to drive.     

Somatostatin plus isosorbide 5-mononitrate versus somatostatin in the control of acute gastro-oesophageal variceal bleeding: a double blind, randomised, placebo controlled clinical trial

BACKGROUND: Variceal bleeding is a severe complication of portal hypertension. Somatostatin reduces portal pressure by decreasing splanchnic blood flow, and nitrates by diminishing intrahepatic resistance. Experimental studies have shown that the combination of somatostatin and nitrates has an additive effect in decreasing portal pressure. AIM: To compare the therapeutic efficacy of either intravenous infusion of somatostatin plus oral isosorbide 5-mononitrate or somatostatin alone in gastro-oesophageal variceal bleeding associated with liver cirrhosis. METHODS: A unicentre, double blind, placebo controlled, clinical trial was conducted. Sixty patients bleeding from oesophageal or gastric varices were randomised to receive intravenous infusion of somatostatin (250 microg/hour) plus oral isosorbide 5-mononitrate (40 mg/12 hours) (group I) or somatostatin infusion plus placebo (group II) for 72 hours. RESULTS: The two groups of patients had similar clinical, endoscopic, and haematological characteristics. Control of bleeding was achieved in 18 out of 30 patients (60%) in group I and 26 out of 30 patients (87%) in group II (p<0.05). There was no significant difference in mean transfusion requirements between the two groups: 2.6 (2.2) v 1.8 (1.6) respectively; means (SD). Mortality and side effects were similar in the two groups, but development of ascites was higher in group I (30%) than in group II (7%) (p<0.05). CONCLUSION: In cirrhotic patients with acute gastro-oesophageal variceal bleeding, addition of isosorbide 5-mononitrate to somatostatin does not improve therapeutic efficacy, induces more adverse effects, and should not be used.     

Gastroesophageal reflux--cause of hemorrhaging esophageal varices?

The results of the examinations do not depend on the peptic theory of the haemorrhage of the oesophageal varices in patients with liver cirrhosis, since the relative frequency of reflux troubles and of gastrooesophageal reflux in patients with liver cirrhosis and haemorrhage of the oesophageal varices was not found greater than in patients with liver cirrhosis and oesophageal varices without haemorrhage as well as the combination of reflux oesophagitis and oesophageal varices was rarely to be observed in the endoscopic material.     

Endoscopic ligation of oesophageal varices compared with injection sclerotherapy in primary biliary cirrhosis

BACKGROUND AND AIMS: Oesophageal varices are an important complication in primary biliary cirrhosis (PBC). However, there have yet to be any studies made on treatment of oesophageal varices in PBC. We therefore studied the efficacy and related complications of endoscopic variceal ligation (EVL) and endoscopic injection sclerotherapy (EIS) as an initial treatment in primary biliary cirrhotic patients. METHODS: From December 1985 to March 1999, 29 biliary cirrhotic Japanese patients with portal hypertension and oesophageal varices were treated in our clinics. Eleven patients were treated with EVL and EIS, and 18 patients underwent EIS only. The liver function, renal function and respiratory function were studied before and after endoscopic treatment and any complications were also examined. RESULTS: In stages III and IV, significant differences were observed in the serum levels for total bilirubin and gamma-glutamic pyruvic transaminase only in the EIS group. Significant differences were observed in the rate of appearance of pyrexia, retrosternal pain and pleural effusion between the EIS and EVL groups. CONCLUSION: EVL significantly reduced the adverse effects associated with EIS at the initial session in primary biliary cirrhotic patients.     

Characteristics,prognosis and outcome of patients with oesophageal varices in a university hospital in Sweden 1994–1999

Objective. Patients with liver cirrhosis, portal hypertension and oesophageal varices are known to have high morbidity and mortality. The knowledge of incidence, aetiology and outcome in Sweden in recent years is limited. Material and methods. All patients with oesophageal varices diagnosed for the first time at Sahlgrenska University Hospital during the 6-year period 1994–1999 were retrospectively studied. Information about the aetiology of liver cirrhosis and oesophageal varices, as well as about the proportion of bleeding and non-bleeding varices, endoscopic and pharmacological treatment and outcome, was analyszed. Results. 312 patients were retrieved, 297 with liver cirrhosis (197 diagnosed before first bleeding (P), 92 after bleeding (B) and 8 at autopsy) and 15 with portal vein thrombosis without cirrhosis. Fifty-four percent had alcoholic liver disease. Fifty-five percent in group B and 13% in group P had at least one bleeding episode during follow-up (p<0.001). There was no significant difference in survival between groups B and P. Twenty-six percent of the cirrhotics died of liver failure and 19% from variceal bleeding. In a multivariate analysis, variables predicting mortality were: Child-Pugh class, group B, age and bilirubin levels. Conclusions. Variceal bleeding is still a strong risk factor for recurrent bleeding, but few die from their first bleeding. This concurs with studies indicating declining mortality from variceal bleeding. However, this patient group still has a high mortality from other causes.     

Nonhereditary colonic angiodysplasias: Histomorphometric approach to their pathogenesis

The pathogenesis of colonic angiodysplasias, more accurately termed vascular ectasias (VE) has not been definitely established. The aim of this study was to assess that the VE of noncirrhotic patients are not associated with diffuse abnormalities of the colonic mucosal microvasculature unlike the VE of cirrhotic patients. Three groups of nine consecutive patients were studied: group I, control patients with an irritable bowel syndrome; group II, noncirrhotic patients with VE; and group III, alcoholic cirrhotics with VE. A histomorphometric analysis of normal-appearing colonic mucosa was achieved from biopsies taken at six predetermined sites. Noncirrhotics with VE had a significantly lower mean number of mucosal capillaries and a significantly lower mean cross-sectional area of mucosal capillaries than alcoholic cirrhotics with VE. Alcoholic cirrhotics with VE had a significant increase of all the vascular parameters compared to the control group. There was no difference between the control patients and the noncirrhotic patients with VE. These results suggest that the VE of noncirrhotic and cirrhotic patients are entities of distinct pathogenesis.     

Effect of cisapride on postprandial gastro-oesophageal reflux.

We studied the effect of cisapride on oesophageal motor function and postprandial gastro-oesophageal reflux in a randomised, double blind, placebo controlled crossover study. In 16 patients with symptomatic gastro-oesophageal reflux, cisapride 10 mg orally and placebo were studied on separate days according to identical protocols. Cisapride and placebo were given 30 minutes before a standard meal. Each study day was preceded by corresponding three day oral loading of cisapride (10 mg tds) or placebo. Lower oesophageal sphincter pressure, oesophageal body motility and oesophageal pH were monitored for 30 minutes before and three hours after the meal. Plasma cisapride concentrations were measured before and after dosing on both study days. With cisapride treatment, the plasma cisapride levels ranged from 48.1 (5.0) to 75.9 (6.9) ng/ml. Plasma levels were undetectable during placebo treatment. Cisapride enhanced acid clearance but had no significant effect on the duration of acid exposure, the rate of reflux episodes, the pattern of lower oesophageal sphincter pressure associated with the reflux episodes, basal lower oesophageal sphincter pressure or oesophageal peristalsis. These findings do not suggest a major role for cisapride, at the dosage tested, for the control of troublesome postprandial gastro-oesophageal reflux.     

Hyperglucagonaemia in cirrhotic patients and its relationship to the severity of cirrhosis and haemodynamic values

Plasma glucagon concentrations were measured in 160 cirrhotic patients (Pugh's grade A in 52 patients, Pugh's grade B in 64 patients and Pugh's grade C in 44 patients). These values were compared with plasma glucagon concentrations in 57 age and sex-matched healthy subjects. Systemic and portal haemodynamic measurements, effective renal plasma flow and creatinine clearance were recorded for each patient. Plasma glucagon levels were significantly increased in cirrhotic patients compared with healthy subjects. In addition, plasma glucagon levels were higher in cirrhotic patients with ascites than in those without ascites and were increased in relation to the severity of cirrhosis as assessed by Pugh's score. Multiple linear regression found that only Child-Pugh's score was estimated to be an independent predictor of hyperglucagonaemia in cirrhotic patients. However, in patients with different degrees of oesophageal varices and in patients without oesophageal varices, plasma glucagon concentrations were no different among the different groups of patients, but were still higher than plasma glucagon concentrations in healthy subjects. In contrast, plasma glucagon levels were negatively correlated with mean arterial pressure and systemic vascular resistance. The results of the present study suggest that impairment of liver function plays, in part, a role in increased plasma glucagon levels observed in patients with cirrhosis. In addition, these data support the hypothesis that hyperglucagonaemia may contribute, at least in part, to the pathogenesis of peripheral arterial vasodilatation in cirrhosis with portal hypertension.     

Combined Upper and Lower Gastrointestinal Endoscopy: A Prospective Study in Alcoholic and Nonalcoholic Cirrhosis

Upper gastrointestinal hemorrhage is one of the more important complications of cirrhosis. Most of the available data regarding the prevalence of upper and lower gastrointestinal sites of bleeding in cirrhotic patients have been obtained in individuals with alcoholic cirrhosis evaluated in the course of an acute gastrointestinal bleeding episode. Few data exist, however, as to the prevalence of either potential bleeding sites or of normal endoscopic findings in hemodynamically stable individuals with cirrhosis of any etiology. Five hundred ten cirrhotic subjects, who were evaluated for possible liver transplantation (OLTx) between January 1985 and June 1987, were included in this study. Seventy-five had alcoholic cirrhosis and 435 had nonalcoholic cirrhosis of various etiologies. Of these 510 patients, 412 underwent combined upper and lower gastrointestinal endoscopy and 98 underwent upper gastrointestinal endoscopy alone. Gastritis, gastric and duodenal ulcer disease were found significantly (each at least p less than 0.025) more often in patients with alcoholic liver disease than in those with nonalcoholic liver disease. The prevalence of the various lower gastrointestinal lesions in both groups was similar. Of particular interest is the fact that in alcoholic cirrhotics, the prevalence of gastritis, gastric ulcer and duodenal ulcer disease was unrelated to the degree of portal hypertension, whereas in the nonalcoholic cirrhotics the prevalence of gastritis and duodenal ulcer disease but not gastric ulcer disease was associated significantly with the degree of portal hypertension as assessed by the presence or absence of large esophageal varices, ascites, and hepatic encephalopathy.