Search

Objective The study objectives were to clarify the clinical findings and the causes of intractability and mortality of upper gastrointestinal (UGI) bleeding in inpatients. Methods The patients were divided into Inpatient (Ip) and Outpatient (Op) onset groups, and their characteristics, clinical and bleeding data, and outcomes were compared. Patients or Materials Our study included 375 patients who developed UGI bleeding during hospitalization or were admitted after being diagnosed with UGI bleeding in an outpatient setting from January 1, 2015, to June 30, 2020. Results The Ip group had worse general condition; increased percentages of comorbidities; and more common use of proton pump inhibitor, anti-coagulant, and steroid than the Op group. Compared with the Op group, the Ip group had lower serum albumin levels and platelet counts at the onset of bleeding, whereas rebleeding, mortality, and bleeding-related death rates were higher. Multivariate analysis of the Ip group revealed that the risks of rebleeding included endoscopic high-risk stigmata, maintenance dialysis, and duodenal bleeding, whereas the risks of mortality were gastric ulcer and a Charlson Comorbidity Index update score of ≥3. Conclusion UGI bleeding in the Ip group was associated with higher rebleeding and mortality rates. Because of their poor general health condition, the pathology of UGI bleeding in these patients may differ from that of patients with common UGI bleeding. A different approach for the care and prevention of UGI bleeding in inpatients is required.

Background Jackhammer esophagus (JE) is a hypercontractile esophageal motility disorder diagnosed using high-resolution manometry (HRM). We sought to determine the clinical presentation and therapeutic data of patients with JE in Japan. Methods The study included patients with JE, diagnosed through HRM performed for suspicious esophageal motility disorders. Demographics, esophagogastroduodenoscopy, radiology, and therapy data were collected from patient charts. Results Among the 4,412 HRM tests performed, 89 patients (61.6 ± 13.4 years; 64 males, 25 females) were diagnosed with JE (2.0%). Dysphagia was the most frequent symptom (80%), followed by chest pain (40%) and heartburn (25%). Esophagogastroduodenoscopy showed abnormal findings in 32% of patients: corkscrew/rosary beads appearance in 26%, narrowing in 11%. Eosinophilic infiltration (> 15 eosinophils/high power field) was diagnosed in 21%. Esophagography showed abnormal findings in 9% of the patients. For the initial therapy, 47 patients received medical treatment followed by peroral endoscopic myotomy (21 patients) and laparoscopic myotomy (two patients). Thirteen patients did not receive any treatment and 10 of those (77%) reported spontaneous resolution of symptoms. Patients who required invasive treatment experienced severe disability in their quality of life and greater maximal distal contractile integral than those who did not. Conclusions HRM showed that the prevalence of JE was very low (2%). Esophagogastroduodenoscopy revealed some characteristic features of JE in patients. Some patients showed improvement of symptoms without invasive treatments. Follow-up with/without medical treatment should be considered before performing invasive treatment in patients whose distal contractile integral is relatively low and the quality of life is not impaired.

Significant atrophic gastritis in the fundic gland region is a well-known endoscopic finding observed in autoimmune gastritis (AIG). The endoscopic features of early AIG have not been reported. Iron deficiency, vitamin B12 deficiency, anemia, or neurological symptoms may not be observed in the early stages of AIG, and it may thus be difficult to diagnose early AIG based on clinical findings. We treated a 50-year-old Japanese female whose condition was suspected to be early AIG. The endoscopic findings showed normal gastric pyloric gland mucosa, and diffuse reddened and edematous gastric fundic gland mucosa. Pathologically, local infiltration of lymphocytes and decrease of parietal cells was present in a deep part of the gastric fundic gland mucosa. Blood tests showed that the titer of parietal cell antibody (PCA) was 1:320 and the gastrin level was 820 pg/ml. We determined that the patient had AIG because she also had Hashimoto’s disease, the PCA titer was high, the serum gastrin level was slightly increased, and inflammation was observed only in the gastric body on the endoscopic images. To the best of our knowledge, this is the first case report of endoscopic findings that suggest early AIG, before atrophic changes were observed.

A 47-year-old Japanese woman presented with epigastric discomfort and anorexia. Upper endoscopy showed type 4 advanced gastric cancer in the gastric antrum with stenosis of the pyloric portion. Abdominopelvic CT revealed peritoneal dissemination, and stage IV advanced gastric cancer was diagnosed. Laparoscopic gastrojejunostomy and enterostomy were performed. Trousseau syndrome occurred 7 days post-surgery. Despite appropriate treatment, the patient developed a pulmonary embolism 13 days later and died 18 days post-surgery. Even in young cancer patients without lifestyle disease complications, it is important to pay attention to Trousseau syndrome as part of cancer management.

Autoimmune gastritis (AIG) is chronic atrophic gastritis caused by an autoimmune mechanism of unknown etiology and presents with various pathological conditions by causing an achlorhydria state through parietal cell damage. The most characteristic endoscopic finding in AIG is advanced corpus-dominant mucosal atrophy. A recent study that examined several cases in Japan revealed the presence of endoscopic features other than corpus-dominant advanced atrophy. Remnants of oxyntic mucosa and sticky adherent dense mucus were found in ≥30% of cases, and hyperplastic polyps were found in ≥20% of cases. In image-enhanced endoscopy (IEE), white globe appearance (WGA) was observed in 32% of AIG cases. Additionally, some reports have stated that the findings in AIG cases using IEE showed cast-off skin appearance (CSA) and foveola type mucosa; however, a consensus is yet to be achieved. These endoscopic results were found in cases of advanced-stage AIG. There have been few reports concerning early-stage AIG cases. In these few reports, all of the cases were pathologically diagnosed as early AIG. In all of the cases, the pathological findings almost always showed neither parietal cell destruction nor atrophy. Endoscopic findings such as “mosaic pattern with slight swelling of the areae gastricae”, “diffuse reddened and edematous gastric fundic gland mucosa”, and “pseudopolyp-like nodules” may be common characteristics of early images. In such early cases, high antibody titers, no atrophic changes, and few clinical abnormal findings were shown. Endoscopists are expected to update their knowledge regarding AIG diagnosis with the evolution of imaging equipment.

Objective The incidence of osteoporosis is increasing with the rapid aging of the Japanese population. Bisphosphonates are first-line agents used for the treatment of osteoporosis, but they can cause upper gastrointestinal mucosal injury. This study investigated symptoms and upper gastrointestinal mucosal injury associated with oral bisphosphonates. Methods Symptoms were evaluated using the F-scale questionnaire, and esophageal mucosal injury and gastroduodenal ulceration were assessed by endoscopy. Patients were stratified by the type of bisphosphonate (alendronate, risedronate, or minodronate), treatment schedule (once weekly or every four weeks), and the concomitant use of other medications [antithrombotic agents, nonsteroidal anti-inflammatory drugs (NSAIDs), or acid suppressants]. Patients The subjects included 221 patients treated with oral bisphosphonates for at least one month. Results The median F-scale total score was 4 (0-34), reflux score was 2 (0-20), and the mean dyspepsia score was 2 (0-16). Endoscopy showed esophageal mucosal injury of Grade A or worse (Los Angeles classification) in 22/221 patients (10.0%) and gastroduodenal ulcers in 9 patients (4.1%). The dyspepsia score in patients who took minodronate every four weeks was significantly lower (p

The accidental ingestion of a press-through package (PTP) sheet is associated with the risk of gastrointestinal injury and puncture. When pain occurs in the laryngeal pharynx, the PTP may already be stuck in the upper esophageal region, and urgent endoscopic treatment is necessary. A plain chest X-ray image should be performed first to identify the ingested PTP, but this was not successful in the present two patients. As a next step, a CT scan (which involves high radiation exposure) is commonly used. In our patients (a 76-year-old woman and a 59-year-old man), tomosynthesis was used to confirm the presence and location of a PTP sheet as a pre-endoscopic diagnosis. With tomosynthesis, the level of radiation exposure was reduced to approx. 1/10 of that of simple CT; the imaging time is also shorter at ~ 5 s. Tomosynthesis can be performed at approx. 1/3–1/4 the cost of simple CT. The usefulness of tomosynthesis as a pre-endoscopic diagnostic tool was demonstrated in our patients, and its further utilization is expected.

Functional dyspepsia (FD) is a gastroduodenal disorder that presents as postprandial fullness, early satiation, or epigastric burning despite no evidence of a structural disease. Proton pump inhibitors (PPIs) are often the first choice for treating FD. However, some patients need additional medication because of residual symptoms despite a certain level of benefit from the PPI. For these patients, a combination of PPI and other agents has a possibly more beneficial effect than changing their medication. This study aimed to evaluate the efficacy of an initial PPI followed by combination therapy with PPI and acotiamide in FD patients with residual symptoms after an initial PPI. We enrolled 105 patients who started an initial PPI (20 mg of esomeprazole once a day). Twenty-three patients with residual symptoms received 100 mg of acotiamide, a cholinesterase inhibitor, three times a day with esomeprazole as a combination therapy for 2 weeks. The symptoms were evaluated using the modified Frequency Scale for the Symptoms of Gastroesophageal Reflux Disease (mFSSG). Eighteen of 23 patients (78%) achieved an overall improvement in symptoms. Almost all FD-related symptoms statistically improved after the combination therapy, with an improvement in the mFSSG score relevant to the postprandial distress syndrome and epigastric pain syndrome. The symptoms improved regardless of age, sex, and the pre-combination therapy score of the mFSSG. Our findings suggest that the combination therapy of acotiamide and PPI may be effective in selected FD patients with insufficient improvement with an initial PPI. However, well-designed trials are required to confirm the efficacy.

We investigated the effects of rikkunshito, in combination with a proton pump inhibitor, on symptoms and quality of life in patients with proton pump inhibitor-refractory gastroesophageal reflux disease. The subjects were 47 patients with gastroesophageal reflux disease with residual symptoms such as heartburn following 8 weeks of proton pump inhibitor therapy. We administered these subjects rikkunshito in combination with a proton pump inhibitor for 6-8 weeks. We scored their symptoms of heartburn, fullness, abdominal discomfort, and abdominal pain, and surveyed their quality of life using the Reflux Esophagitis Symptom Questionnaire, comprising questions concerning daily activities, meals (changes in amount and favorite foods), and sleep (getting to sleep and early morning waking). Improvement was seen in all symptoms, and quality of life scores for meals and sleep also improved. These results indicate that combination therapy with rikkunshito and a proton pump inhibitor improves quality of life related to eating and sleep in patients with patients with proton pump inhibitor-refractory gastroesophageal reflux disease.

Objective Differentiated gastric cancer generally develops in the atrophic gastric mucosa, although undifferentiated cancer is sometimes encountered in patients with severe atrophic gastritis. We characterized the endoscopic features of undifferentiated gastric cancer in patients with severe atrophic gastritis. Methods Stage IA early gastric cancer was diagnosed in 501 patients who were admitted to our hospital between April 2003 and March 2012. The endoscopic and pathological findings were compared among 29 patients with undifferentiated cancer and severe atrophic gastritis, 104 patients with undifferentiated cancer and mild/moderate atrophic gastritis and 223 patients with well-differentiated cancer and severe atrophic gastritis. Endoscopic atrophic gastritis was classified according to the Kimura-Takemoto classification as no gastritis, C-1 and C-2 (mild), C-3 and O-1 (moderate) or O-2 and O-3 (severe). Results The tumors were larger and showed deeper mural invasion in the patients with undifferentiated cancer and severe atrophic gastritis than in those with well-differentiated cancer and severe gastritis or undifferentiated cancer and mild/moderate gastritis. On endoscopy, undifferentiated cancer associated with severe gastritis was often red in color. Conclusion It is often difficult to diagnose early undifferentiated gastric cancer, especially in patients with severe atrophic gastritis. The present study characterized the important endoscopic features of such tumors.

A 53-year-old woman was referred to our hospital for management of gastric varices that ran transversely across the greater curvature of the gastric body, detected during routine upper gastrointestinal endoscopy. CT identified a low-density calcified mass near the tail of the pancreas and the splenic hilum. Based on the results of radiographic and pathological investigations, the tumor was diagnosed as solid pseudopapillary neoplasm (SPN), and the gastric varices were considered to have developed secondary to occlusion of the splenic vein by the tumor mass. This is a rare case of SPN associated with splenic vein occlusion and left-sided extrahepatic portal hypertension.

Aim: In non-alcoholic steatohepatitis (NASH), fibrosis begins around the central veins, as also happens with alcoholic liver disease, so the symptoms of portal hypertension may be due to central vein occlusion. The aim of this study was to define the prevalence of esophagogastric varices and the clinical outcome after endoscopic treatment in NASH patients with severe fibrosis. Methods: The subjects were 72 patients with clinicopathologically confirmed NASH who had bridging fibrosis (F3) or cirrhosis (F4) determined by the examination of liver biopsy specimens, and who underwent upper gastrointestinal endoscopy. The prevalence and pattern of endoscopically detected varices at the time of liver biopsy were evaluated. The results of NASH patients (n = 11) with endoscopically treated esophageal varices were compared to those with alcoholic (n = 67) and hepatitis C virus-associated cirrhosis (n = 152). Results: Esophagogastric varices were detected in 34 out of the 72 (47.2%) patients; esophageal varices in 25 (34.7%) and gastric varices in nine (12.5%), while six of these patients had variceal bleeding. In NASH patients, the cumulative recurrence-free probability at 24 months after endoscopic treatment was 63.6%, the bleeding-free probability was 90.9%, and the 5-year survival was 100%. Only one out 11 patients died of liver failure at 70 months after treatment. Conclusion: About half of NASH patients with severe fibrosis had esophagogastric varices. The clinical status and course of the varices do not necessarily improve after endoscopic treatment. NASH patients with esophagogastric varices need to be followed up carefully, like patients with other chronic liver diseases.

Background: Nodular gastritis (NG) was considered a physiological change with little pathological significance, mostly in young women. In recent years, however, it has been often reported in patients with Helicobacter pylori (H. pylori) infection, or in patients with gastroduodenal ulcer/gastric cancer, suggesting possible clinical significance. Methods: From July 2003 to July 2006, 59 patients were diagnosed with NG among 32 404 patients examined endoscopically. The incidence of NG was evaluated in relation to age, sex, H. pylori infection status, symptoms leading to endoscopy, associated lesions in the upper digestive tract at the time of NG diagnosis, and existence of other systemic conditions. Results: The NG patients consisted of 13 out of 18 152 (0.07%) male patients and 46 out of 14 252 (0.32%) female patients, with a mean age of 45.3 ± 17.7 years. All 28 patients who were examined for H. pylori infection were positive. Endoscopic examination was performed for precordial pain and upper abdominal pain in 24 (40.7%) patients, symptoms of gastroesophageal reflux disease in eight (13.6%) patients, and symptoms of functional dyspepsia in six (10.2%) patients. NG was associated with duodenal ulcer in eight (13.6%) patients, hyperplastic gastric polyps in five (8.5%), gastric ulcer in one (1.7%), and gastric cancer in one (1.7%) patient. Conclusion: NG is a specific gastritis resulting from H. pylori infection that may be strongly associated with H. pylori-related lesions.

Objective The aim of this study was to analyze the effects of anti-thrombotic therapy and nonsteroidal anti-inflammatory drugs (NSAIDs) on gastroduodenal bleeding. Patients and Methods The study subjects were 544 patients (421 males and 123 females, mean age, 64.2 years) who were treated endoscopically for bleeding gastroduodenal ulcers from January 1995 to August 2008. Of the 544 patients, 276 (50.7%) had a history of treatment for ≥1 month with an antithrombotic agent or NSAIDs, including low-dose aspirin (n=94), other NSAIDs (n=91), warfarin (n=43), or any combination of the three (combination treatment group; n=48). On the other hand, 268 patients had not previously received any of these drugs (control group). Clinical features and endoscopic therapeutic results were assessed and compared. Results Helicobacter pylori infection was detected in 187 of the 241 (77.6%) patients examined. Of the 544 patients, 199 (36.6%) attended the Department of Cardiology or Cardiovascular Surgery, and 170 (31.3%) patients were already being treated with antiulcer medication, including proton pump inhibitors (n=18 [3.3%]). Forty (7.4%) patients suffered from rebleeding after intervention; the incidence of ulcer rebleeding was not significantly different between patients being treated with any such drugs (4.7% [13/276]) and those that had not previously received any antithrombotic agent or NSAIDs (10.1% [27/268]). Conclusion Antithrombotic therapy and NSAIDs use contributed to bleeding in 50.7% of patients with gastroduodenal ulcers. These drugs are a major cause of ulcer bleeding, but are not necessarily considered a risk factor for rebleeding after endoscopic hemostasis.

Aim Because the procedure of balloon-occluded retrograde transvenous obliteration (B-RTO) causes extensive thrombosis of the major shunt that connects the spleen and gastric/renal venous systems, an increase in portal pressure is unavoidable. The aim of the present study was to assess the long-term outcome of B-RTO, including changes in esophageal varices. Methods B-RTO was conducted in 22 patients with gastric varices, who were divided according to the severity of esophageal varices at baseline; there were no esophageal varices (n = 7), F(1) varices (n = 11), and F(2) varices (n = 4). The outcome measures included the development/worsening of esophageal varices after B-RTO and survival rates. Results The cumulative bleeding-free probability for all 22 patients at 3 years after B-RTO was 100%. The overall 3-year survival was 94.4%. Seven patients who had no esophageal varices prior to B-RTO did not develop any after the procedure. Seven (63.6%) of the 11 patients with stage F(1) esophageal varices prior to B-RTO showed no changes in the varices after B-RTO, while two patients progressed to F(2) varices and two developed F(3) varices. The cumulative treatment-free probability of the esophageal varices at 24 months after B-RTO was 100% for patients without esophageal varices at baseline, 80.8% for patients with pre-existing F(1) varices, and 75% for those with pre-existing F(2) varices. Conclusion Although the B-RTO procedure is considered useful for the treatment of gastric varices, changes in hemodynamics due to obliteration of this major shunt must be taken into account and observed closely.

Recently, 3D image in endoscopic ultrasonography has been developed, so that examination become easier than 2D image. However, is it offer better result than 2D image. We investigated if it is useful for determination of the depth of cancer invasion and what factors influence the result. Ultrasound imaging system (Olympus Co.) is used. The method: Anticholinergic agent or glucagon is injected in order to suppress the peristalsis andthe 2 channel scope was inserted to the lesion vicinity. After the lesion is observed, the probe was inserted from the working channel, and water was irrigated from the other channel. When the lesions was covered the water, the scan was carried out. To stop respiration was directed under scan. One of the judgment of the depth of cancer invasion by images from radial scan, the other judgment of the depth of cancer invasion by made liner image, simulation three-dimensional image or multiple images from the computer taken images, they were separately determined. In 50 cases of esophageal cancer and 47 cases of gastric cancer, each results were compared with histological findings. Result 1. In esophageal cancer, the probe passed through the carcinoma in 49 cases, and whole scanning was possible in 98%. In judgment from radial scan, Tis and cancer invading the lamina propria were correctly determined in 67%, the cancer invading the muscularis mucosa (including slightly invasion the submucosa) in 59%, the submucosa in 92% , T2 in 100%, T3 in 63% , T4 in 100% and total on 72%. On the other hand judgment from images by computer, Tis and cancer invading the lamina propria were accurately determined in 83%, cancer invading the muscularis mucosa (including slight invasion the submucosa) in 66%, cancer invading the submucosa in 92%, T2 in 100%, T3 in 73%, T4 in 100% and total in 80%. 2. In gastric cancer, the judgment from radial scan, mucosal cancer was correctly determined in 81%, cancer invading the submucosa in 92% , T2 in 33% and total 81%. On the other hand, the judgment from images by taken computer, mucosal cancer was accurately determined in 84%, cancer invading the submucosa in 100%, T2 in 67% and total in 85%. 3. In the image with the big tumor, the invasion depth became the inaccuracy by the attenuation. The most important point in the judgment from images by computer was that judgment was made two direction, consecutive images of lesions. Conclusion: EUS 3D images were offer more accurate judgments than 2D. Information from two-direction and consecutive images of the lesion let us precise judgments.

The Purpose of the study; It is essential to diagnose mucosal cancer for determination of indication of endoscopic mucosal resection (EMR). We investigated the layer structure of the gastric wall, and diagnosis of the depth of invasion of early gastric cancer by using 30MHz-ultrasonographic probe (OLYMPUS XUM-30-25R) Methods; 1) (Experiment) The EUS findings of injected gelatine into the submucosa in resected stomach were compared with histological findngs. 2) (Clinical cases) The EUS findings in 31 cases of gastric cancer are compared with the pathological results. Results; 1) The resected gastric wall was delineated to 9 layers.1st and 2nd layers were gastric epithelium,the lamina propria and gastric glands, the 3rd layer was the interface ,the 4th layer was the muscularis mucosae, the 5th was the submucosa ,the 6th to 8th were the muscularis propria ,and 9th was the serosa.If glands were atrophic ,the lamina propria and grands would be described as hyperechoic layer and merge into the 3rd layer. 2) The muscularis mucosae were visualized in all cases.Accuracy of determination of the depth of cancer invasion was 85.7% in mucosal cancer, 80.0% in cancer invasing the submucosa , and over all accuracy was 83.8%.Case of overdiagnosis was caused by lesion concomitant with ulcer.Case of underdiagnosis was caused by microscopic invasion. Conclusion; The muscularis mucosae were delineated clearly and the accuracy of deternination of the cancer was inproved by the use of 30MHz probe.And also the area of cancer occupied all mucosa might be defined by 30MHz probe.

The aim of this paper is if higher frequency probe :30MHz provide better resolution and result. Equipment: 20,30MHz probes were used. The method: 1.Experiment: In order to analyze the difference between layer structure by 20MHz and one by 30MHz, 6 esophageal resected specimens were scanned by 20 or 30MHz in water tub. 2. Clinical use: Before examination, anticholinergic agent or glucagon is injected in order to suppress the peristalsis. The 2 channel scope was inserted to the lesion vicinity, the probe was inserted from the working channel, and water was irrigated from the other channel. The scan was carried out by 20 or 30MHz. Twentyone cases of superficial esophageal cancer were examined by 20 and 30MHz and each images were compared with histological findings. Result 1. Experiment: Normal esophageal wall was delineated 9 layers by 20 and 30MHz, however using 30MHz the 2nd and 4th layers became clear. 2. Clinical cases: Two epithelial cancer were delineated by 30MHz, but one of them was difficult to determine by 20MHz. The determination of cancer limited to the lamina propria was correctly determined in 78% by 30MHz, 67% by 20MHz. Cancer invading the muscularis mucosa or small invasion to the submucosa were correctly determined in 100% by 30MHz, in 67% by 20MHz. Cancer invading the submucosa was determined in 89% by 30 and 20MHz. Conclusion: 30MHz probe provide better image of the layers from the epithelium to the muscularis mucosa. Accordingly 30MHz is useful for determined indication of mucosal resection via endoscopy.

Background: Endoscopic variceal sclerotherapy (EVS) and endoscopic variceal ligation (EVL) are known to be equally effective for the treatment of esophageal varices. However, the indications for each type of therapy have not been decided. Endoscopic ultrasonography (EUS) is a well-accepted procedure for the diagnosis of gastrointestinal lesions, and three-dimensional (3D) imaging has recently been developed. Aims: The aim of this study was to observe the hemodynamics of esophageal varices using 3DEUS and to examine the selection of endoscopic treatment. Subjects and Methods: 3D-EUS was done with a 3D ultrasonic image processing system (EU-M30, EU-IP-2, Olympus Optical Co. Ltd.) using a small probe (12 MHz, 3.4 mm in diameter) by the mechanical spiral scanning method (length of 40 mm and pitch of 0.5 mm) with water irrigation. The area from the cardia to the distal esophagus was scanned before treatment, and 3D images were prepared that clearly displayed the architecture of the esophageal varices. The pattern of variceal blood flow detected on 3D images was classified into 4 types according to its architecture. Type 1 was cardial inflow without collateral vessels, Type 2 was cardial inflow with collateral vessels, Type 3 was an azygos-perforating pattern, and Type 4 was a complicated pattern. Thirty-four patients (100%) with esophageal varices underwent 3D-EUS observation. Subsequently, 23 patients were treated by EVS using 5%ethanolamine oleate and 11 patients by EVL, after which the results were assessed. Results: Based on 3D-EUS data, 16 patients (47.1%) were classified as Type 1, 4 patients (11.8%) as type 2, 4 patients (11.8%) as type 3, and 10 patients (29.4%) as Type 4. Three sessions of either EVS or EVL initially eradicated Type 1 esophageal varices, but recurrence was observed in 2 of 11 patients (18.2%) treated by EVS and in 3 of 5 patients (60.0%) treated by EVL. Type 3 varices were eradicated by 2.0 sessions of EVL, with none of the 3 patients showing recurrence. Conclusions: It was easy to assess the architecture of esophageal varices on 3D images. Based on the 3D-EUS findings, varices could be classified into 4 types and this classification was useful for the selection of endoscopic treatment. To treat the cardial inflow type without collateral vessels, obliteration of the feeding vessels is necessary, so EVS is selected. When collateral vessels are present and variceal blood flow can shift to the collateral pathways, EVL is more efficient should be selected.