Sedation during upper GI endoscopy in cirrhotic outpatients: a randomized, controlled trial comparing propofol and fentanyl with midazolam and fentanyl.

BACKGROUND: Patients with liver cirrhosis frequently undergo diagnostic or therapeutic upper GI endoscopy (UGIE), and the liver disease might impair the metabolism of drugs usually administered for sedation. OBJECTIVE AND SETTING: To compare sedation with a combination of propofol plus fentanyl and midazolam plus fentanyl in cirrhotic outpatients undergoing UGIE. DESIGN: A prospective, randomized, controlled trial was conducted between February 2008 and February 2009. MAIN OUTCOMES MEASUREMENTS: Efficacy (proportion of complete procedures using the initial proposed sedation scheme), safety (occurrence of sedation-related complications), and recovery time were measured. RESULTS: Two hundred ten cirrhotic patients referred for UGIE were randomized to 2 groups: midazolam group (0.05 mg/kg plus fentanyl 50 µg intravenously) or propofol group (0.25 mg/kg plus fentanyl 50 µg intravenously). There were no differences between groups regarding age, sex, weight, etiology of cirrhosis, and Child-Pugh or American Society of Anesthesiologists classification. Sedation with propofol was more efficacious (100% vs 88.2%; P < .001) and had a shorter recovery time than sedation with midazolam (16.23 ± 6.84 minutes and 27.40 ± 17.19 minutes, respectively; P < .001). Complication rates were similar in both groups (14% vs 7.3%; P = .172). LIMITATIONS: Single-blind study; sample size. CONCLUSION: Both sedation schemes were safe in this setting. Sedation with propofol plus fentanyl was more efficacious with a shorter recovery time compared with midazolam plus fentanyl. Therefore, the former scheme is an alternative when sedating cirrhotic patients undergoing UGIE.

Efficacy and safety of endoscopic ultrasound-guided drainage of pancreatic pseudocysts using double-pigtail plastic stents: A single tertiary center experience.

OBJECTIVES: Pancreatic pseudocysts (PPC) are fluid collections with a well-defined wall that persist for more than 4 weeks inside or around the pancreas as a result of pancreatic inflammation and/or a ductal lesion. PPC have been successfully treated with endoscopic ultrasound (EUS)-guided drainage using different stents. This study aimed to evaluate the safety and efficacy of EUS-guided drainage of PPC using double-pigtail plastic stents in a tertiary hospital. METHODS: Patients with PPC referred for EUS-guided drainage between May 2015 and December 2019 were included in this case series. The primary endpoint was to evaluate the efficacy (clinical success) and safety (adverse events and mortality) of EUS-guided drainage of PPC. Secondary endpoints included technical success and pseudocyst recurrence. RESULTS: Eleven patients (mean age, 44.5±18.98 years) were included in this study. The etiologies for PPC were acute biliary pancreatitis, chronic alcoholic pancreatitis, and blunt abdominal trauma. The mean pseudocyst size was 9.4±2.69 cm. The clinical success rate was 91% (10/11). Adverse events occurred in three of 11 patients (27%). There were no cases of mortality. The technical success rate was 100%. Pseudocyst recurrence was identified in one of 11 patients (9%) at 12 weeks after successful clinical drainage and complete pseudocyst resolution. CONCLUSION: EUS-guided transmural drainage of PPC using double-pigtail plastic stents is safe and effective with high technical and clinical success rates.

Instant messenger smartphone application for endosonographer/cytopathologist real-time interaction at a distance in EUS-FNA for solid pancreatic lesions.

Background and study aims Telecytopathology (TCP) may allow proper and timely evaluation of endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) specimens. The aim of this study was to evaluate the feasibility of TC through a multiplatform instant messenger smartphone application to evaluate specimens of EUS-FNA of pancreatic solid lesions. Patients and methods Twenty-three patients (14 male/9 female; median age: 56 yr.; age range: 33 - 86) with a solid pancreatic lesion were included. Exclusion criteria were as follows: age < 18 yr and predominantly cystic lesions. During each EUS-FNA, after each pass, the aspirated material was spread over a glass slide and was stained by the endoscopist. The glass slide was then reviewed on a microscope with a smartphone fitted in, and the most representative fields were captured and sent to the cytopathologist using WhatsApp Messenger. Results In initial evaluation using TCP rapid on-site evaluation (ROSE), adequate cellularity of the glass slide was detected in 16 of 23 patients (69.6 %). An initial diagnosis of malignancy (positive or suspicious) was possible in 14 of 23 patients (60.8 %). Conclusion The current study demonstrated the feasibility of a low-cost, Internet-based, telecytopathology system using WhatsApp Messenger to provide ROSE of EUS-FNA slides in patients with solid pancreatic lesions.

Endoscopic resolution of pseudocyst infection and necrosis as a complication of endoscopic pseudocyst drainage. A case report.

CONTEXT: Endoscopic therapy has recently gained importance as an option for the treatment of pancreatic necrosis. We report a case of transgastric endoscopic resolution of pseudocyst infection and necrosis as a complication of endoscopic pseudocyst drainage. CASE REPORT: The patient underwent several sessions of endoscopic removal of debris and necrotic material, intercalated with daily lavage and aspiration through a nasocystic tube. CONCLUSIONS: At the present time, fourteen months after pancreatic pseudocyst secondary infection, the patient remains asymptomatic with no radiographic signs of pseudocyst recurrence. Transmural endoscopic therapy followed by continuous drainage and necrosectomy is a reasonable option for treating the pancreatic pseudocyst secondary infection.

Endoscopic ultrasound guided fine-needle aspiration core biopsy: comparison between an automatic biopsy device and two conventional needle systems.

BACKGROUND: endoscopic ultrasound guided fine-needle aspiration (EUS-FNA) allows cytologic and/or histologic diagnosis of lesions within or adjacent to the gastrointestinal tract. However, the amount of tissue obtained with a regular 22 gauge needle is not always satisfactory. With the development of a needle XNA-10J-KB (Shot-Gun) that resembles the automatic liver biopsy needle, it is expected that significant samples be obtained more frequently (core biopsy), optimizing histological analysis. OBJECTIVE: to compare samples obtained with EUS-FNA using 3 different needle systems: GIP, NA-10J-1 and Shot-Gun. METHODS: 19 patients underwent EUS-FNA for diagnosis (5) or tumor staging (14). Mean age was 58.9 years (range 27-82), being 50% men. All patients were submitted to EUS-FNA with the 3 needle models. The Shot-Gun model was "shot" when its tip was near the target inside the lesion, followed by aspiration. Samples were submitted for cytologic and histologic examination. RESULTS: mean lesion size was 3.0 cm (range 0.8-5.5 cm). Final diagnoses were made after surgery or intra-operative biopsy: 13 pancreatic tumors (12 adenocarcinomas and 1 neuroendocrine tumor), 4 chronic pancreatitis, 1 acute pancreatitis, and 1 cholangiocarcinoma. Specimens adequate for cytologic diagnosis were obtained in 13/19 (68. 4%) patients using GIP model, in 14/19 (73.7%) with NA10J-1 model, and in 17/19 (89.5%) with ShotGun, model (p=0.039). Histologic analysis was possible in 10/19 (52.6%) patients using the GIP model, in 14/19 (73.7%) with NA10J-1, and in 17/19 (89.5%) with Shot-Gun, model (p=0.005). Adequate samples for cytologic or histologic assessment in 16/19 (84.2%) patients using the GIP model, in 17/19 (89.5%) with NA10J-1, and in 18/19 (94.7%) with Shot-Gun, model (p=0.223). In two cases biopsies were negative due to very hard tumors. CONCLUSION: the Shot-Gun needle obtained better samples for histological diagnosis than NA10J-1 needle and GIP.

[I Brazilian consensus of endoscopic ultrasonography]. / I Consenso Brasileiro de Ecoendoscopia.

BACKGROUND: In the last 20 years, several papers have focused on demonstrating the impact of endoscopic ultrasonography findings on the management of different clinical scenarios in digestive disease. This fact is an indirect evidence of the difficulty of popularization of the method. On other hand, the limited availability of endoscopic ultrasonography in Brazil is a direct evidence of this limitation. This was the rationale for the organization of a consensus meeting on endoscopic ultrasonography. It was aimed to identify the best evidence that support the use of endoscopic ultrasonography in gastroenterology. METHODS: A panel of experts on endoscopic ultrasonography was selected based on the files of the Gastroenterology and Endoscopy Societies and on the registries of endoscope manufacturers. Two members of the meeting selected the relevant topics that were transformed into questions. The topics and the questions were debated among the experts five months before the consensus meeting. The experts were asked to perform systematic reviews in order to answer the questions so it could be possible to grade the answers based on the strength of the evidence. During the two days of the meeting the answers were presented, debated and voted. Consensus was reached when a minimum of 70% of the voters were in agreement. The final consensus report was submitted to the experts' evaluation and approval. RESULTS: Seventy nine questions were debated by the experts at the pre-Consensus meeting. As the result of this debate 85 questions came out and were assigned to the members of the panel. During the Consensus meeting 22 experts debated and voted 85 answers. Consensus was reached for several clinical scenarios for which the impact of endoscopic ultrasonography findings were supported by level 1 evidences: differential diagnosis of subepithelial lesions and thickening of gastric folds, staging and diagnosis of unresectable esophageal cancer, indirect signs of peritoneal involvement of gastric cancer, MALT gastric lymphoma and rectal cancer staging, diagnosis of common bile duct and gallbladder stones, diagnosis of chronic pancreatitis and differential diagnosis of a solid mass in chronic pancreatitis, differential diagnosis of the pancreatic cyst, prediction of the results of the endoscopic treatment of esophageal varices and diagnosis and staging of non-small cell lung cancer. CONCLUSIONS: There are the highest levels of evidences that support the indication of endoscopic ultrasonography for several digestive diseases and even for non-small cell lung cancer.

II Brazilian consensus statement on endoscopic ultrasonography.

BACKGROUND AND OBJECTIVES: At the time of its introduction in the early 80s, endoscopic ultrasonography (EUS) was indicated for diagnostic purposes. Recently, EUS has been employed to assist or to be the main platform of complex therapeutic interventions. METHODS: From a series of relevant new topics in the literature and based on the need to complement the I Brazilian consensus on EUS, twenty experienced endosonographers identified and reviewed the pertinent literature in databases. The quality of evidence, strength of recommendations, and level of consensus were graded and voted on. RESULTS: Consensus was reached for eight relevant topics: treatment of gastric varices, staging of nonsmall cell lung cancer, biliary drainage, tissue sampling of subepithelial lesions (SELs), treatment of pancreatic fluid collections, tissue sampling of pancreatic solid lesions, celiac neurolysis, and evaluation of the incidental pancreatic cysts. CONCLUSIONS: There is a high level of evidence for staging of nonsmall cell lung cancer; biopsy of SELs as the safest method; unilateral and bilateral injection techniques are equivalent for EUS-guided celiac neurolysis, and in patients with visible ganglia, celiac ganglia neurolysis appears to lead to better results. There is a moderate level of evidence for: yield of tissue sampling of pancreatic solid lesions is not influenced by the needle shape, gauge, or employed aspiration technique; EUS-guided and percutaneous biliary drainage present similar clinical success and adverse event rates; plastic and metallic stents are equivalent in the EUS-guided treatment of pancreatic pseudocyst. There is a low level of evidence in the routine use of EUS-guided treatment of gastric varices.

Comparison between endoscopic brush cytology performed before and after biliary stricture dilation for cancer detection.

BACKGROUND: Confirmation of malignancy within biliary strictures is endoscopically challenging. Dilation of strictures has been reported to enhance cytological diagnosis. AIM: To compare brush cytology results before and after biliary stricture dilation. PATIENTS AND METHODS: Patients with extra-hepatic biliary stricture at endoscopic retrograde cholangiopancreatography were included in the study. Brushing was performed before and immediately after dilation using a 10 Fr dilating catheter. Cytology samples were classified as: negative for malignancy, presence of atypical cells, insufficient material, suspicious for malignancy or positive for malignancy. Final diagnosis was established by surgery, biopsy or follow-up. RESULTS: Biliary brush cytology was performed in 50 patients, with an overall sensitivity of 40% and 27.5%, before and after dilation, respectively. The combination of results increased cancer detection rate to 45%. There were 5/50 (10%) minor complications and one death related to the procedure. CONCLUSIONS: Brush cytology performed before biliary stricture dilation has a similar cancer detection rate to that following dilation, although the combination of results enhances sensitivity.

Chronic radiation proctopathy: A practical review of endoscopic treatment.

Chronic radiation proctopathy (CRP) is a troublesome complication of pelvic radiotherapy. The most common presentation is rectal bleeding. CRP symptoms interfere with daily activities and decrease quality of life. Rectal bleeding management in patients with CRP represents a conundrum for practitioners. Medical therapy is ineffective in general and surgical approach has a high morbid-mortality. Endoscopy has a role in the diagnosis, staging and treatment of this disease. Currently available endoscopic modalities are formalin, potassium titanyl phosphate laser, neodymium:yttrium-aluminum-garnet laser, argon laser, bipolar electrocoagulation (BiCAP), heater probe, band ligation, cryotherapy, radiofrequency ablation and argon plasma coagulation (APC). Among these options, APC is the most promising.

Efficacy and safety of endoscopic ultrasound-guided drainage of pancreatic pseudocysts using double-pigtail plastic stents: A single tertiary center experience

OBJECTIVES: Pancreatic pseudocysts (PPC) are fluid collections with a well-defined wall that persist for more than 4 weeks inside or around the pancreas as a result of pancreatic inflammation and/or a ductal lesion. PPC have been successfully treated with endoscopic ultrasound (EUS)-guided drainage using different stents. This study aimed to evaluate the safety and efficacy of EUS-guided drainage of PPC using double-pigtail plastic stents in a tertiary hospital. METHODS: Patients with PPC referred for EUS-guided drainage between May 2015 and December 2019 were included in this case series. The primary endpoint was to evaluate the efficacy (clinical success) and safety (adverse events and mortality) of EUS-guided drainage of PPC. Secondary endpoints included technical success and pseudocyst recurrence. RESULTS: Eleven patients (mean age, 44.5±18.98 years) were included in this study. The etiologies for PPC were acute biliary pancreatitis, chronic alcoholic pancreatitis, and blunt abdominal trauma. The mean pseudocyst size was 9.4±2.69 cm. The clinical success rate was 91% (10/11). Adverse events occurred in three of 11 patients (27%). There were no cases of mortality. The technical success rate was 100%. Pseudocyst recurrence was identified in one of 11 patients (9%) at 12 weeks after successful clinical drainage and complete pseudocyst resolution. CONCLUSION: EUS-guided transmural drainage of PPC using double-pigtail plastic stents is safe and effective with high technical and clinical success rates.

Endoscopic ultrasonography and alcoholic patients: can one predict early pancreatic tissue abnormalities?

CONTEXT: The diagnosis of chronic pancreatitis in early stages remains challenging and agreement between various methods is poor. Alcohol consumption is the most important cause of chronic pancreatitis. OBJECTIVE: The aim of this study was to evaluate the endosonographic evidence of pancreatic disease in patients with alcohol abuse. SETTING: Consecutive patients referred for endoscopic ultrasonography for various indications between May 2001 and December 2003. PATIENTS: Two hundred twenty-eight patients were enrolled in the study. One hundred 89 patients were studied; 39 (17.1%) were excluded because the pancreas could not be completely assessed (n=23) or because a previous diagnosis of chronic pancreatitis and/or pancreatic neoplasm was suspected (n=16). DESIGN: Prospective study. The patients were divided into 2 groups: alcoholics (n=17) and non-alcoholics (n=172). MAIN OUTCOME MEASURE: The assessment of pancreatic disease according to parenchymal and ductal changes as previously described by Catalano and Sahai (scores 1 and 2, respectively). RESULTS: The male:female ratio was 7.5 and 0.9 in the alcoholic and the non-alcoholic groups, respectively (P=0.002). Mean ages were 53.0 and 47.8 years, respectively (P=0.156). Alcoholic patients had significantly (P<0.001) more endosonographic abnormalities than non-alcoholics for both scores. A ROC curve comparison between patients and controls showed that the best specificity and sensitivity combination was obtained for the number of EUS features greater than 1 and 2 for the scores 1 and 2, respectively. CONCLUSION: Alcoholic patients had significantly more endosonographic abnormalities reflecting possible early changes of chronic pancreatitis. By using reported scores, values up to 1 or 2 can be seen in patients with no risk for pancreatitis, suggesting that chronic pancreatitis should be considered only for scores equal to or greater than 2 or 3 for scores 1 and 2, respectively.

PROPOSAL OF A CLINICAL CARE PATHWAY FOR THE MANAGEMENT OF ACUTE UPPER GASTROINTESTINAL BLEEDING.

BACKGROUND: Upper gastrointestinal bleeding implies significant clinical and economic repercussions. The correct establishment of the latest therapies for the upper gastrointestinal bleeding is associated with reduced in-hospital mortality. The use of clinical pathways for the upper gastrointestinal bleeding is associated with shorter hospital stay and lower hospital costs. OBJECTIVE: The primary objective is the development of a clinical care pathway for the management of patients with upper gastrointestinal bleeding, to be used in tertiary hospital. METHODS: It was conducted an extensive literature review on the management of upper gastrointestinal bleeding, contained in the primary and secondary information sources. RESULTS: The result is a clinical care pathway for the upper gastrointestinal bleeding in patients with evidence of recent bleeding, diagnosed by melena or hematemesis in the last 12 hours, who are admitted in the emergency rooms and intensive care units of tertiary hospitals. In this compact and understandable pathway, it is well demonstrated the management since the admission, with definition of the inclusion and exclusion criteria, passing through the initial clinical treatment, posterior guidance for endoscopic therapy, and referral to rescue therapies in cases of persistent or rebleeding. It was also included the care that must be taken before hospital discharge for all patients who recover from an episode of bleeding. CONCLUSION: The introduction of a clinical care pathway for patients with upper gastrointestinal bleeding may contribute to standardization of medical practices, decrease in waiting time for medications and services, length of hospital stay and costs.

[Colonic polyps diagnosis by conventional video colonoscopes and chromoscopy with indigo carmine dye solution]. / Videocolonoscopia convencional e cromoscopia com índigo carmim no diagnóstico de pólipos colônicos.

BACKGROUND: Magnification colonoscopy and contrast chromoscopy with indigo carmine dye solution have been used to differentiate neoplastic polyps (adenomas and adenocarcinomas) from non-neoplastic (hyperplastic, inflammatory, juvenile) in an attempt to obviate endoscopic polypectomy. On the other hand, little published information exists concerning conventional video colonoscopes and chromoscopy for polyp histology prediction. Aim - To assess usefullness of conventional video colonoscopes and contrast chromoscopy with indigo carmine solution for differential diagnosis of colon polyps. METHODS: In a routine colonoscopy series, we performed chromoscopy with conventional video colonoscopes before endoscopic excision of detected polyps. If a sulcus pattern was observed on the surface of the lesion, it was classified as neoplastic. Polyps were classified as non-neoplastic if no sulcus was detected on its surface. These observations were then compared with histology. RESULTS: In the study period (18 months), we detected 133 polyps in 53 patients. We were able to compare results of histology and chromoscopy in 126 lesions. The sensitivity, specificity, diagnostic accuracy, negative predictive value, and positive predictive value were 56,4%, 79,2%, 65,1%, 52,8%, and 81,5%, respectively. CONCLUSION: On the base of the presented data, we concluded that conventional video colonoscopes and contrast chromoscopy with indigo carmine solution is not a good technique for differential diagnosis of colon polyps.

Efficacy and safety of endoscopic prophylactic treatment with undiluted cyanoacrylate for gastric varices.

AIM: To evaluate the efficacy and safety of undiluted N-butyl-2 cyanoacrylate plus methacryloxysulfolane (NBCM) as a prophylactic treatment for gastric varices (GV) bleeding. METHODS: This prospective study was conducted at a single tertiary-care teaching hospital between October 2009 and March 2013. Patients with portal hypertension (PH) and GV, with no active gastrointestinal bleeding, were enrolled in primary prophylactic treatment with NBCM injection without lipiodol dilution. Initial diagnosis of GV was based on endoscopy and confirmed with endosonography (EUS); the same procedure was used after treatment to confirm eradication of GV. After puncturing the GV with a regular injection needle, 1 mL of undiluted NBCM was injected intranasally into GV. The injection was repeated as necessary to achieve eradication or until a maximum total volume of 3 mL of NBCM had been injected. Patients were followed clinically and evaluated with endoscopy at 3, 6 and 12 mo. Later follow-ups were performed yearly. The main outcome measures were efficacy (GV eradication), safety (adverse events related to cyanoacrylate injection), recurrence, bleeding from GV and mortality related to GV treatment. RESULTS: A total of 20 patients (15 male) with PH and GV were enrolled in the study and treated with undiluted NBCM injection. Only 2 (10%) patients had no esophageal varices (EV); 18 (90%) patients were treated with endoscopic band ligation to eradicate EV before inclusion in the study. The patients were followed clinically and endoscopically for a median of 31 mo (range: 6-40 mo). Eradication of GV was observed in all patients (13 patients were treated with 1 session and 7 patients with 2 sessions), with a maximum injected volume of 2 mL NBCM. One patient had GV recurrence, confirmed by EUS, at 6-mo follow-up, and another had late recurrence with GV bleeding after 35 mo of follow-up; overall, GV recurrence was observed in 2 patients (10%), after 6 and 35 mo of follow-up, and GV bleeding rate was 5% (1 patient). Mild epigastric pain was reported by 3 patients (15%). No mortality or major complications, including embolism, or damage to equipment were observed. CONCLUSION: Endoscopic injection with NBCM, without lipiodol, may be a safe and effective treatment for primary prophylaxis of gastric variceal bleeding.

Pyloric Brunner's gland hamartoma with atypical hyperplasia.

Brunner's gland hamartoma (BGH) is an extremely rare benign digestive tumor, generally located in the duodenal bulb. We report the case of a 51-year-old asymptomatic man with a large pedunculated BGH arising from the pylorus. It was successfully removed en bloc by endoscopic resection.

Heterotopic pancreas complicated by pseudocyst in the gastric wall diagnosed by endoscopic ultrasound-guided fine needle aspiration.

Heterotopic pancreas is usually asymptomatic and does not change throughout the patient's life, but sometimes it can present symp-toms and complications, which are rarely discrete in the literature. We present here a case of heterotopic pancreas in the gastric wall complicated with pseudocyst, and suggest that heterotopic pancreatic pseudocyst should be included in the differential diagnosis of gastric wall cysts.

Endoscopic treatment of esophageal varices in advanced liver disease patients: band ligation versus cyanoacrylate injection.

BACKGROUND: The results of variceal band ligation (VBL) in patients with advanced liver disease are unknown. Cyanoacrylate injection (CI) might have a better outcome than VBL in the treatment of esophageal varices (EV) in these patients. AIM: To compare VBL and CI in the treatment of EV in patients with advanced liver disease. PATIENTS AND METHODS: Thirty-eight patients with medium or large EV and Child-Pugh index of at least eight were randomized into two groups: VBL (n=20) and CI (n=18). The patients were followed-up for at least 6 months after the end of treatment. Main outcomes were eradication, bleeding, mortality, complication, and recurrence rates. RESULTS: Variceal eradication rates were similar in the VBL and CI groups (90 vs. 72%, P=0.39). Mean number of sessions until eradication was 3.17 and 3, respectively. Bleeding episodes until eradication were equally observed in both groups (P=0.17). Mortality (55 vs. 56%, P=0.52) and major complication rates (5 and 17%, P=0.32) were similar. Chest pain with dysphagia was more frequent in the CI group (55.6 vs. 10%, P=0.004). A higher risk of variceal recurrence was observed in the CI group (33 vs. 57%, P=0.04). CONCLUSION: No significant differences between the VBL and CI groups were observed in the treatment of EV in patients with advanced liver disease regarding mortality, variceal eradication, and major complications rates. However, minor complications and variceal recurrence were significantly more common in the CI group. In addition, there was a clear trend toward more bleeding episodes in patients included in the CI group.

Prediction of esophageal varices in hepatic cirrhosis by noninvasive markers.

OBJECTIVE: To determine whether Model for End-stage Liver Disease (MELD) Child-Turcotte-Pugh (CTP) classification, AST to platelet ratio index (APRI), and laboratory tests could predict the presence of esophageal varices (EV) or varices which need prophylactic therapy (medium or large size EV). METHODS: Three hundred patients with cirrhosis (193 men; mean age 53.1 years; majority with chronic C hepatitis) were prospectively analyzed. The presence of EV (any size and medium or large EV) was correlated with patients' characteristics (MELD, CTP classification, APRI, platelets count, and liver tests). RESULTS: One hundred and seventy-one patients (57%) had EV, of whom 35% (105) had varices which need prophylactic therapy (VPT). The distribution of EV according to CTP classification was as follows: A, 49%; B, 75.3% and C, 80%. Independent predictors of EV were: MELD higher than 8 (P=0.02); APRI higher than 1.64 (P=0.01); platelet count lower than 93,000/mm³ (P<0.01); aspartate aminotransferase higher than 1.34 × UNL (P=0.01), and total bilirubin higher than 1 mg/dl (P=0.04). MELD higher than 8 had the highest discriminative value for presence of EV (sensitivity=80.1%; specificity=51.2%; area under receiver operating characteristics=0.68). Factors independently associated with VPT were: thrombocytopenia (<92,000/mm³; P<0.01) and aspartate aminotransferase higher than 1.47 × UNL (P=0.03). Platelet count lower than 92,000/mm³ had sensitivity of 65.7%, specificity of 57.9%, and an area under receiver operating characteristics of 0.62 for the presence of VPT. CONCLUSION: High values on MELD are associated with EV and thrombocytopenia, with varices which need prophylactic therapy. As a result of their low sensitivity and specificity, it is suggested to maintain the recommendation of upper gastrointestinal endoscopy for all patients with cirhosis.