Protocol design for randomized clinical trial to compare underwater cold snare polypectomy to conventional cold snare polypectomy for non-pedunculated colon polyps of size 5-10 mm (COLDWATER study).

BACKGROUND: Colorectal cancer is internationally the third leading cause of death from a malignant disease. The aim of screening colonoscopy in adults > 45 years of age is early diagnosis and treatment of precancerous polyps. Endoscopic polyp removal (polypectomy) can be achieved with various techniques depending on the size, morphology, and location of the polyp. According to current guidelines, small non-pedunculated polyps should be removed with a cold snare after the colorectal lumen has been insufflated with air (conventional cold snare polypectomy).In recent years, several studies have described the benefits of water aided colonoscopy, as well as the safety and efficacy of underwater cold snare polypectomy for large colon polyps. However, there are insufficient data on conventional and underwater techniques for small polyps, the most commonly diagnosed colorectal polyps. METHODS: We have designed a prospective randomized double-blind clinical trial to compare the safety and efficacy of conventional and underwater cold snare polypectomy for non-pedunculated polyps 5-10 mm in size. A total of 398 polyps will be randomized. Randomization will be carried out using the random numbers method of Microsoft Excel 2016. The primary endpoint is the muscularis mucosa resection rate. Secondary endpoints are the depth and percentage of R0 excisions, complications, and the recurrence rate at follow-up endoscopy 6-12 months after polypectomy. DISCUSSION: We hypothesize underwater polypectomy will result in a higher muscularis mucosa resection rate. The results of our study will provide useful data for the development of guidelines in polypectomy techniques for non-pedunculated polyps 5-10 mm in size. CLINICAL TRIAL REGISTRATION: Clinicaltrials.gov, NCT05273697.

Gastric tube volvulus following an Ivor-Lewis esophagectomy.

Gastric tube conduit is the method of choice for restoring continuity of the digestive track after a partial or total esophagectomy. Redundant gastric conduit (i.e. an elongated, floppy conduit) is a rare cause of dysphagia in patients with long survival. Gastric tube volvulus is exceedingly rare with only three cases described in the literature. We present the diagnostic and therapeutic course of a 57-year-old man who presented to our department with gastric tube volvulus 32 months after an Ivor-Lewis esophagectomy. Diagnosis was made with computed tomography and volvulus was reduced endoscopically. To the best of our knowledge, this is only the fourth case of gastric tube volvulus described in the English literature. This rare situation might be a consequence of a redundant gastric tube. Endoscopic volvulus decompression was successful in our case.

The incidence of HPV infection in anal cancer patients in Greece.

BACKGROUND AND AIM OF THE STUDY: Although anal cancer represents a relatively uncommon malignancy, its incidence over the last five decades, has been reported as increased for both sexes, worldwide. Human papillomavirus (HPV) infection has been shown to be a major cause for its development. The aim of the present study is to report on clinical, epidemiological and virological data of squamous anal cancer in Greek patients. PATIENTS AND METHOD: Between January 2002 and December 2010, 11 Greek patients (6 females) who were diagnosed as suffering from squamous cell anal or perianal cancer, were treated in our Hospital. Formalin fixed paraffin embedded tissue samples, obtained at the time of the anal biopsy or surgery, were analyzed by PCR in order to identify the presence as well as the type of HPV infection. RESULTS: Overall, the presence of HPV DNA was detected in 6 out of the 11 patients (54.5%). The "high risk" HPV DNA was detected in 3 of them (2 women and 1 man), while the "low risk" HPV DNA was detected in the remaining three (2 women and 1 man). CONCLUSION: The incidence of HPV infection in squamous cell anal cancer Greek patients, is lower than other Western countries, probably reflecting differences in sexual habits in the Greek population.

Palliative gastrectomy and other factors affecting overall survival in stage IV gastric adenocarcinoma patients receiving chemotherapy: a retrospective analysis.

OBJECTIVE: Most patients with gastric cancer present with locally advanced or metastatic disease and usually receive palliative therapy. We sought to identify factors influencing overall survival in patients with stage IV gastric cancer receiving palliative chemotherapy. PATIENTS AND METHODS: The records of 311 patients with histological diagnosis of gastric adenocarcinoma were retrospectively reviewed and 17 clinicopathological and therapeutic parameters were evaluated for their influence on overall survival. RESULTS: In multivariate analysis nine factors were found to independently influence survival: no previous palliative gastrectomy [Hazard ratio (HR, 12; CI 7.969-18.099)], single agent chemotherapy instead of combination chemotherapy (HR, 1.35; CI 1.068-1.721), histological grade III (HR, 1.39; 95% CI 1.098-1.782), the presence of hepatic (HR, 1.6; 95% CI 1.246-2.073) and abdominal metastasis (HR, 1.33; 95% CI 1.039-1.715), CA 72-4 > 7 U/L (HR, 1.39; 95% CI 1.026-1.887), LDH > 225 U/L (HR, 1.72; 95% CI 1.336-2.236], need for blood transfusions (HR, 1.58; 95% CI 1.213-2.082), and weight loss > 5% (HR, 1.96; 95% CI 1.352-2.853) at the time of initial diagnosis. Patients were stratified as low (0-2 factors), intermediate (3-6 factors) and high (7-9 factors) risk and the median survival was 76, 40 and 11 weeks, respectively. CONCLUSION: Nine clinical and laboratory factors that adversely affect survival in patients with stage IV gastric cancer who receive chemotherapy were identified. Their concurrent presence seems to have an additive effect as patients with seven to nine factors have the worse prognosis. Palliative gastrectomy and combination chemotherapy appear to be associated with improved survival.

Does Helicobacter pylori identification in the mucosa of the gallbladder correlate with cholesterol gallstone formation? / ¿Existe una correlación de la identificación de Helicobacter pylori en la mucosa de la vesícula biliar con la formación de cálculos biliares de colesterol?

OBJECTIVE: Helicobacter pylori (H pylori) represents a potential initiator of cholesterol crystallization and it has been proposed that it is related to gallstone formation. In this study, any possible association between the H pylori identification in the mucosa of gallbladder and cholesterol gallstone formation was evaluated. METHODS: Gallbladders containing pure or mixed cholesterol gallstones (cholelithiasis group, n = 89) and gallbladders without gallstones (control group, n = 42) were submitted to standard histopathological examination for H pylori detection, as well as to nested polymerase chain reaction amplification for H pylori DNA detection. RESULTS: Helicobacter pylori was identified in the gallbladder's epithelium in four patients with cholelithiasis and in two patients in the control group by histology. In all the cases which were found to be H pylori positive by histological examination, H pylori DNA were also detected. No correlation between gallstone formation and H pylori detection in the biliary epithelium was found. A higher incidence of acute inflammation in the cholelithiasis (22.5% vs 9.5%, p = not significant [ns]) and in the H pylori positive groups (33% vs 17.6%, p = ns) were histologically detected. A higher incidence (10% vs 0%), p = ns) of H pylori in gallbladders with gallstones and acute inflammation, compared to gallbladders with acute inflammation but without gallstones, was noticed. CONCLUSION: Helicobacter pylori is detectable in low frequency in the mucosa of the gallbladder and it does not seem to act as a lithogenic component for cholesterol gallstone formation. Its higher incidence in gallbladders with gallstones and acute inflammation, suggests a possible accessory role in a subset of patients with cholelithiasis.

OBJETIVO: Helicobacter pylori (H pylori) representa un iniciador potencial de la cristalización del colesterol, y se ha propuesto que guarda relación con la formación del cálculo biliar. En este estudio, se evaluó cualquier posible asociación entre la identificación de H pylori en la mucosa de la vesícula y la formación del cálculo biliar de colesterol. MÉTODOS: Las vesículas que contienen cálculos biliares de colesterol puros o mixtos (grupo de colelitiasis, n = 89) y vesículas sin cálculos biliares (grupo control, n = 42) fueron sometidos a un examen histopatológico estándar con el fin de detectar el H pylori descubrimiento, así como a la amplificación de la reacción en cadena de polimerasa para la detección de ADN H pilori. RESULTOS: El Helicobacter pylori fue identificado mediante histología en el epitelio de la vesícula en cuatro pacientes con el colelitiasis y en dos pacientes en el grupo de control. En todos los casos que resultaron ser H pylori positivo por el examen histológico, se halló también DNA H pylori. No se halló correlación ninguna entre la formación del cálculo biliar y la detección de H pylori en el epitelio biliar. Se detectó histológicamente una incidencia más alta de inflamación aguda en la colelitiasis (22.5% contra 9.5%, p = no significativo [ns]) y en los grupos H pylori positivos (33% contra 17.6%, p = ns). Se observó una incidencia más alta (10% contra 0%), p = ns) de H pylori en las vesículas con los cálculos biliares e inflamación aguda, en comparación con las vesículas con la inflamación aguda pero sin cálculos biliares. CONCLUSIÓN: Helicobacter pylori es detectable en baja frecuencia en la mucosa de la vesícula y no parece actuar como un componente litogénico en la formación del cálculo biliar de colesterol. Su mayor incidencia en las vesículas con cálculo biliar e inflamación aguda, hace pensar en un posible papel auxiliar en un subconjunto de pacientes con colelitiasis.

Does Helicobacter pylori identification in the mucosa of the gallbladder correlate with cholesterol gallstone formation?

OBJECTIVE: Helicobacter pylori (H pylori) represents a potential initiator of cholesterol crystallization and it has been proposed that it is related to gallstone formation. In this study, any possible association between the H pylori identification in the mucosa of gallbladder and cholesterol gallstone formation was evaluated METHODS: Gallbladders containing pure or mixed cholesterol gallstones (cholelithiasis group, n = 89) and gallbladders without gallstones (control group, n = 42) were submitted to standard histopathological examination for H pylori detection, as well as to nested polymerase chain reaction amplification for H pylori DNA detection. RESULTS: Helicobacter pylori was identified in the gallbladder's epithelium in four patients with cholelithiasis and in two patients in the control group by histology. In all the cases which were found to be H pylori positive by histological examination, H pylori DNA were also detected. No correlation between gallstone formation and H pylori detection in the biliary epithelium was found. A higher incidence of acute inflammation in the cholelithiasis (22.5% vs 9.5%, p = not significant [ns]) and in the H pylori positive groups (33% vs 17.6%, p = ns) were histologically detected. A higher incidence (10% vs 0%), p = ns) of H pylori in gallbladders with gallstones and acute inflammation, compared to gallbladders with acute inflammation but without gallstones, was noticed CONCLUSION: Helicobacter pylori is detectable in low frequency in the mucosa of the gallbladder and it does not seem to act as a lithogenic component for cholesterol gallstone formation. Its higher incidence in gallbladders with gallstones and acute inflammation, suggests a possible accessory role in a subset of patients with cholelithiasis.

Alcoholic hepatitis: from pathogenesis to treatment.

UNLABELLED: Alcoholic hepatitis is a serious complication of alcohol abuse due to its high mortality rates particularly at short term. It may complicate pre-existing alcoholic fatty liver or cirrhosis and is mainly diagnosed on clinical and laboratory grounds although liver biopsy is occasionally needed to exclude other pathology and confirm the diagnosis. Accumulating evidence suggests that cytokines and immunity are actively involved in its pathogenesis. Management includes abstinence and supportive care. Treatment with corticosteroids has been studied in several clinical trials with conflicting results. However, recent evidence supporting the beneficial effect of TNF-alpha inhibition provides an encouraging alternative. Here we summarise the current state in diagnosis and management of alcoholic hepatitis and briefly review the latest advances in pathophysiology that may lead to new therapeutic strategies for this difficult clinical condition. DATA SOURCES: Medline 1966-2005, EMBASE/Excerpta Medica 1980-2005, The Cochrane Library (2005 Issue 2) and contact with authors of published reports.

Raltitrexed (tomudex) administration in patients failing multiple prior chemotherapy regimens in advanced colorectal cancer: a pilot study.

PURPOSE: To evaluate efficacy of Raltitrexed, a specific thymidilate synthase inhibitor, in patients with advanced colorectal cancer (ACC) failing multiple prior chemotherapy regimens (e.g. 5-FU+LV, CPT-11, etc). METHODS: 20 patients with ACC; 13 males/7 females, median age 64 (range: 53-69), median Karnovsky PS: 80 (70-90), and sites of metastases; liver: 16, lung: 6, lymph nodes: 9, peritoneal: 8 and a life expectancy of at least 3 months, were entered in the present pilot study of Raltitrexed administration. All patients had progressed after prior chemotherapy with 5-FU+LV and subsequently CPT-11, and some had received further infusional 5-FU, Raltitrexed was administered at a dose of 3 mg/m2 i.v. every 21 days. RESULTS: 3 patients obtained stable disease (SID), 15%, with tumor marker decline (CEA, CA-19.9). Time-to-progression was 4.8 months (2.2-7) and survival 7.4 months (6.0-7.8). Toxicity was in general not severe and consisted mainly of myelosuppression; neutropenia (WHO) grade 2: 45% and grade 3: 22%, and anemia grade 1-2: 40%. CONCLUSION: Response to treatment with Raltitrexed is limited in patients with ACC failing multiple prior chemotherapy regimens, however, a limited percentage of patients with SD derived clinical benefit.

cag+ Helicobacter pylori induce transactivation of the epidermal growth factor receptor in AGS gastric epithelial cells.

The gastric pathogen Helicobacter pylori is known to activate epithelial cell signaling pathways that regulate numerous inflammatory response genes. The aim of this study was to elucidate the pathway leading to extracellular signal-regulated kinase (ERK) 1/2 phosphorylation in H. pylori-infected AGS gastric epithelial cells. We find that H. pylori, via activation of the epidermal growth factor (EGF) receptor activates the small GTP-binding protein Ras, which in turn, mediates ERK1/2 phosphorylation. cag+ strains of H. pylori are able to induce greater EGF receptor activation than cag- strains, and studies with isogenic mutants indicate that an intact type IV bacterial secretion system is required for this effect. Blockade of EGF receptor activation using tyrphostin AG1478 prevents H. pylori-mediated Ras activation, inhibits ERK1/2 phosphorylation, and substantially decreases interleukin-8 gene expression and protein production. Investigations into the mechanism of EGF receptor activation, using heparin, a metalloproteinase inhibitor and neutralizing antibodies reveal that H. pylori transactivates the EGF receptor via activation of the endogenous ligand heparin-binding EGF-like growth factor. Transactivation of gastric epithelial cell EGF receptors may be instrumental in regulating both proliferative and inflammatory responses induced by cag+ H. pylori infection.

Jejunogastric intussusception presented with hematemesis: a case presentation and review of the literature.

BACKGROUND: Jejunogastric intussusception (JGI) is a rare but potentially very serious complication of gastrectomy or gastrojejunostomy. To avoid mortality early diagnosis and prompt surgical intervention is mandatory. CASE PRESENTATION: A young man presented with epigastric pain and bilous vomiting followed by hematemesis,10 years after vagotomy and gastrojejunostomy for a bleeding duodenal ulcer. Emergency endoscopy showed JGI and the CT scan of the abdomen was compatible with this diagnosis. At laparotomy a retrograde type II, JGI was confirmed and managed by reduction of JGI without intestinal resection. Postoperative recovery was uneventful. CONCLUSIONS: JGI is a rare condition and less than 200 cases have been published since its first description in 1914. The clinical picture is almost diagnostic. Endoscopy performed by someone familiar with this rare entity is certainly diagnostic and CT-Scan of the abdomen could also help. There is no medical treatment for acute JGI and the correct treatment is surgical intervention as soon as possible.

Comparative study of tropisetron with the addition of dexamethasone or alprazolam in breast cancer patients receiving adjuvant chemotherapy with CEF (cyclophosphamide, epirubicin and 5-fluorouracil).

We studied tropisetron (T) in patients with breast cancer receiving standard adjuvant chemotherapy with CEF (cyclophosphamide, epirubicin and 5-fluorouracil) over 3 consecutive cycles; T was administered alone or in combination with dexamethasone (D) or alprazolam (A). 50 women entered and during the 1st cycle patients received T i.v. before chemotherapy and the same dose orally on each of the following 3 days. In the 2nd cycle, T was administered together with D and during the 3rd cycle, T was combined with A and continued with T over the ensuing 3 days post-chemotherapy. Stress was present in 23 women and was evaluated for its impact on antiemetic response. Differences in the emetogenic response were found for nausea and vomiting mainly with the addition of A. The combination of T+A was superior to T and T+D in acute emesis (P<0.001). Concerning delayed emesis, differences were detected with both T+D and T+A (being equally effective) and superior to T alone (P<0.001). The emetogenic potential was decreased by the addition of A in comparison to T alone (P=0.001). Patients without stress had no difference, while patients with stress had a significantly better antiemetic result with the addition of D or A to T. In conclusion, T provides a satisfactory result in controlling nausea and emesis caused by moderately emetogenic CT regimens. Addition of D or A improves the antiemetic effect, and A provides better coverage in women with stress, a finding worth exploration in larger confirmatory studies.

Rapid urease test is less sensitive than histology in diagnosing Helicobacter pylori infection in patients with non-variceal upper gastrointestinal bleeding.

BACKGROUND AND AIMS: The validity of the rapid urease (CLO) test to diagnose Helicobacter pylori infection in patients with bleeding ulcers has been questioned. The aim of this paper is to evaluate the validity of the CLO test in comparison with histology in diagnosing H. pylori infection in patients with acute upper gastrointestinal bleeding (UGB), irrespective of non-steroidal anti-inflammatory drug (NSAID) use. METHODS: Upper gastrointestinal endoscopy was performed within 24 h of admission for all patients with UGB admitted to the Department of Pathophysiology, Medical School, Athens, for a period of 12 months. Patients with variceal bleeding, previous gastric operation, recent treatment with proton pump inhibitors (< 2 months) and those with a history of H. pylori eradication therapy were excluded from the study. At least four biopsies (two from the antrum and two from the body) were obtained for the CLO test and histology (modified Giemsa). RESULTS: Seventy-two consecutive patients (aged 18-90 years, 51 men, 21 women) were included. Forty-six patients (64%) used NSAID. Thirty-two patients (44%) were found to be positive for H. pylori infection by the CLO test, while 44 patients (61%) were found to be positive on histology (P<0.045, 95% CI, 0.004-0.331). The sensitivity and specificity of the CLO test were 68 and 93% respectively; positive and negative predictive values were 94 and 65%, respectively. The age of the patient and visible blood in the stomach did not influence results of either the CLO or histology. CONCLUSIONS: The CLO test, performed within 24 h of hospital admission in patients with UGB, irrespective of NSAID use, is unreliable for the detection of H. pylori infection. The age of the patient and the presence of blood in the stomach do not seem to influence these results.

The CLO test is unreliable in diagnosing H. pylori infection in post-surgical stomach; is there any role of H. pylori in peptic ulcer recurrence?

AIM: To evaluate the validity of the CLO test in detecting Helicobacter pylori in patients with gastric operation and to investigate the relationship of H. pylori with peptic ulcer recurrence in these patients. METHODS: In this prospective study, 110 consecutive patients, the majority of whom had undergone gastric operation for benign disease (n = 102), were included. Eighty patients (62 males), aged 38-87 years, had had a gastrectomy (10 Billroth I, 70 Billroth II), and 30 patients (27 males), aged 36-73 years, had had a vagotomy (13 vagotomy plus gastroenterostomy, 17 vagotomy plus pyloroplasty). H. pylori was sought on multiple biopsy specimens, using CLO test and histology (modified Giemsa stain). The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the CLO test were estimated using histology as 'gold standard'. RESULTS: Overall, 21 gastrectomy patients (26%) were H. pylori-positive by CLO and 25 (31 %) were H. pylori-positive by histology. The estimated sensitivity, specificity, PPV and NPV of the CLO test, using histology as 'gold standard', were 68%, 91%, 77% and 86%, respectively. The CLO test was positive in 67% of vagotomy patients (20 of 30), while 50% (15 of 30) were H. pylori-positive by histology. The estimated sensitivity, specificity, PPV and NPV of the CLO test were 87%, 53%, 65% and 80%, respectively. H. pylori prevalence by histology was 50% in patients with vagotomy and 31% in those with gastrectomy (P = 0.0787). Recurrent ulcers were observed in 8/30 patients (27%) after vagotomy and in 10/72 patients (14%) after gastrectomy. Recurrent ulcer was documented in 6/15 H. pylori-positive patients with vagotomy (40%), and in one of 25 H. pylori-positive patients with gastrectomy (4%). This difference was significant (Fisher's exact test, P = 0.007, relative risk 5.091, 95% CI 0.819-31.64). CONCLUSION: The CLO test seems to be unreliable in diagnosing H. pylori in post-surgical stomach. The H. pylori prevalence is higher, although not significantly, in vagotomized patients compared with gastrectomized patients, and in this group is closely related to the presence of recurrent ulcer. So, at least in this group of patients, it is strongly recommended to look for and eradicate H. pylori.

Expression of HLA-DR, costimulatory molecules B7-1, B7-2, intercellular adhesion molecule-1 (ICAM-1) and Fas ligand (FasL) on gastric epithelial cells in Helicobacter pylori gastritis; influence of H. pylori eradication.

There is evidence that Helicobacter pylori infection up-regulates the expression of HLA class II molecules by gastric epithelial cells (GEC). In this study we evaluated whether GEC are capable of expression of costimulatory molecules in H. pylori gastritis. The expression of FasL by GEC, before and after eradication of H. pylori, was also studied. Thirty patients (23 men) aged 27-81 years (53.67 +/- 13.99 years (mean +/- s.d.)) with dyspepsia were studied. Upper gastrointestinal endoscopy was performed and six biopsies were obtained (antrum, n = 3; corpus, n = 3) for Campylobacter-Like Organisms (CLO) test and histology; 23 (16 men) were H. pylori+ and seven (all men) were H. pylori- by both methods and served as controls. Helicobacter pylori eradication therapy was given to H. pylori+ patients and all patients were re-endoscoped after 116 +/- 9 days. Formalin-fixed paraffin-embedded tissue sections were stained by the ABC immunoalkaline phosphatase method. In H. pylori gastritis HLA-DR was expressed and correlated with disease activity (P < 0.01). No HLA-DR was observed in controls. In H. pylori-eradicated patients significant decrease of HLA-DR was found (antrum, P < 0. 001). ICAM-1 was expressed by GEC in 80% of H. pylori+ patients; ICAM-1 expression did not correlate with gastritis parameters and decreased significantly after eradication (antrum, P < 0.01). B7-1 and B7-2 were expressed on H. pylori+ samples and their expression decreased after eradication, albeit not significantly. Weak epithelial expression of both B7 molecules was observed in all the controls. FasL was steadily expressed by GEC in both H. pylori+ and H. pylori- patients and remained almost unchanged after eradication. These findings suggest that GEC may acquire antigen-presenting cell properties in H. pylori infection through de novo expression of HLA-DR and costimulatory molecules. This seems to be attenuated after eradication and resolution of mucosal inflammation. The same cells exhibit the capacity to control the inflammatory process, probably by inducing apoptotic cell death to Fas-bearing infiltrating lymphocytes.

Ranitidine versus ranitidine plus octreotide in the treatment of acute non-variceal upper gastrointestinal bleeding: a prospective randomised study.

AIM: To compare the efficacy of ranitidine with that of ranitidine plus octreotide in the treatment of non-variceal upper gastrointestinal (UGI) bleeding. DESIGN: Prospective, randomised, open study. PATIENTS AND METHODS: Upper GI endoscopy was carried out during the first 24 hours in all patients with UGI bleeding who had been admitted within a period of 18 months. Patients with variceal bleeding, and those who had undergone any type of gastric operation, were excluded. Eighty-four patients (58 men and 26 women) aged 21-92 years (mean age: 61.2 +/- 15.0 SD) were included. Patients were randomised to receive ranitidine 50 mg tid intravenously alone (Group A: 44 patients, 29 men), or in combination with octreotide 100 micrograms tid subcutaneously, the second drug given for three days only (Group B: 40 patients, 29 men). The study end-points were discharge without operation, emergency surgical intervention or death. The number of blood units given and the days of hospitalization were also recorded. RESULTS: Aspirin and non-aspirin NSAID use before bleeding was reported by 16/44 (36%) patients in Group A and by 19/40 (47.5%) patients in Group B (p = 0.38, OR = 0.63, 95% CI = 0.26-1.51). The endoscopically detected pathology and bleeding stigmata did not differ between the groups (p = 0.86, p = 0.64, OR = 0.78, 95% CI = 0.3-1.99, respectively). Mean use of blood units (p = 0.16) and days of hospitalization (p = 0.25) did not differ. Three patients in Group A (6.8%) and three in Group B (7.5%) required surgical intervention (p = 1.0, OR = 1.1, 95% CI = 0.21-5.84). CONCLUSION: Ranitidine plus subcutaneous octreotide is not superior to ranitidine alone in the management of patients with acute non-variceal UGI bleeding.