Search

Background N-butyl-2-cyanoacrylate has been successfully used for the treatment of bleeding from gastric fundal varices (FV). However, significant rebleeding rates and serious complications including embolism have been reported. Objective Our purpose was to analyze the safety and efficacy of N-butyl-2-cyanoacrylate for FV bleeding by using a standardized injection technique and regimen. Design Retrospective. Setting Two tertiary referral centers. Patients A total of 131 patients (91 men/40 women) with FV underwent obliteration with N-butyl-2-cyanoacrylate by a standardized technique and regimen. Interventions (1) Dilution of 0.5 mL of N-butyl-2-cyanoacrylate with 0.8 mL of Lipiodol, (2) limiting the volume of mixture to 1.0 mL per injection to minimize the risk of embolism, (3) repeating intravariceal injections of 1.0 mL each until hemostasis was achieved, (4) obliteration of all tributaries of the FV, (5) repeat endoscopy 4 days after the initial treatment to confirm complete obliteration of all visible varices and repeat N-butyl-2-cyanoacrylate injection if necessary to accomplish complete obliteration. Main Outcome Measurements Immediate hemostasis rate, early rebleeding rate, bleeding-related mortality rate, procedure-related complications, long-term cumulative rebleeding-free rate, and cumulative survival rate. Results Initial hemostasis and variceal obliteration were achieved in all patients. The mean number of sessions was 1 (range 1-3). The mean total volume of glue mixture used was 4.0 mL (range 1-13 mL). There was no occurrence of early FV rebleeding, procedure-related complications, or bleeding-related death. The cumulative rebleeding-free rate at 1, 3, and 5 years was 94.5%, 89.3%, and 82.9%, respectively. Conclusion Obliteration of bleeding FV with N-butyl-2-cyanoacrylate is safe and effective with use of a standardized injection technique and regimen.

Background EUS-guided pancreatic pseudocyst and abscess drainage ideally requires the insertion of both a transmural stent and nasocystic catheter to ensure continuous irrigation and effective drainage. Sequential stent and nasocystic catheter insertion may be time consuming because of the difficulty in recannulating the pseudocyst after the initial stent placement. Objective Our purpose was to describe a novel one-step simultaneous double-wire technique that facilitates effective pancreatic pseudocyst and abscess drainage. Design To solve the problem of recannulating the pseudocyst cavity after the initial transmural stent placement, we developed a prototype 3-layer puncture kit that allows the simultaneous insertion of 2 guidewires at the initial puncture in one step. This puncture kit consists of a 22-gauge needle used for FNA, a 6F inner polytetrafluoroethylene (Teflon) catheter, and an outer 8.5F Teflon catheter. Setting Tertiary referral center. Patients Eight consecutive patients were treated. Interventions With use of a therapeutic linear echoendoscope and the assembled kit with the needle protruding out distally, the pseudocyst cavity was punctured by using electrocautery. The needle and the inner catheter were then withdrawn, leaving the outer catheter within the cavity. The size of this outer catheter permitted the simultaneous insertion of two 0.035-inch guidewires. Sequential insertion of a transmural stent and nasocystic catheter was then performed without a need for recannulation of the pseudocyst or abscess cavity. Main Outcome Measurements Successful endoscopic drainage. Results All patients were successfully treated without complications. Mean procedural time was 32.5 minutes (range 25-45 minutes). Conclusions With this novel technique, establishing an irrigation system for the treatment of pancreatic pseudocysts and abscesses becomes easier, faster, and safer.

Background Various techniques are available for EMR in the upper- and lower-GI tract. For early cancers of the esophagus, the "suck and cut" technique, which uses a transparent cap or variceal band ligator, is the most commonly practiced method. To facilitate multiple or circumferential EMR, a modified multiband variceal ligator (MBL) is introduced, which allows sequential banding and snare resection without the need to withdraw the endoscope. Objective To study the feasibility of modified MBL device in facilitating circumferential EMR of Barrett's esophagus (BE) that contains high-grade intraepithelial neoplasia (HGIN) and/or intramucosal cancer (IMC). Design To enable band delivery with a snare inserted in the therapeutic endoscope, the threading channel of the cranking device is enlarged from 2 to 3.2 mm. The 6-shooter MBL was used. Patients Ten consecutive patients (all men; median age, 62 years; range 43-82 years) with BE were treated. IMC and HGIN were found in 8 and 2 patients, respectively. Interventions EMR was performed with pure coagulating current when using a 1.5 × 2.5-cm mini hexagonal polypectomy snare. No submucosal saline solution injection was performed before resection. Results In 5 of 10 patients with circumferential BE of 2 to 9 cm in length (median, 4 cm), complete circumferential EMR was performed in 1 session by using 3 to 18 bands (median, 6). Four patients with 3- to 10-cm (median, 4 cm) long segment BE required 2 to 5 sessions (median, 3) with a total of 5 to 42 bands (median, 12). Another patient with multifocal HGIN and/or IMC in 24 of a total of 49 specimens was finally recommended for surgery because of technical difficulties caused by mural thickening after 4 sessions. No serious procedure-related complications were observed, except for 2 minor bleedings, which were controlled endoscopically. Seven patients developed strictures after circumferential EMR. All patients except 1 were successfully managed by weekly bougienage after a median of 5 sessions (range 3-11). Deep-wall tears developed in 1 patient during the fourth bougienage session, for which limited distal esophageal resection was performed with an uneventful outcome. Conclusions The novel technique of MBL-EMR described here facilitated and simplified circumferential removal of BE that contained HGIN and/or IMC. However, the method is associated with a very high stricture rate if circumferential EMR is performed in a single session. Complete removal of BE should be achieved by repeated partial EMR. Long-term follow-up is needed to observe for late recurrence and to determine the clinical impact of this method.

Background: There is no study of circumferential EMR in patients with Barrett's esophagus containing early stage malignant lesions. This study investigated the effectiveness and safety of circumferential EMR by using a simple snare technique without cap. Method: Patients with Barrett's esophagus containing multifocal high-grade intraepithelial neoplasia or intramucosal cancer, and patients with endoscopically nonidentifiable early stage malignant mucosal changes incidentally detected in random biopsy specimens were included in the study. A 30 × 50-mm polypectomy snare made of monofilament 0.4-mm steel wire was used without any additional device or submucosal injection. Results: Twelve patients (10 men, 2 women; median age 63.5 years, range 43-88 years) underwent circumferential EMR; 5 had multifocal lesions, and 7 had no visible lesions. Segments of Barrett's epithelium were circumferential (median length 5 cm) and completely removed. The median number of EMR sessions was 2.5. The median number of snare resections per EMR session was 5. The medial total area of mucosa in resected specimens per session was 3.8 cm 2 . Two patients developed strictures that were successfully treated by bougienage. Minor bleeding occurred during 4 of 31 EMR sessions. During a median follow-up of 9 months, no recurrence of Barrett's esophagus or malignancy was observed. Conclusions: Circumferential EMR with a simple snare technique is feasible, safe, and effective for complete removal of Barrett's epithelium with early stage malignant changes. (Gastrointest Endosc 2003;57:854-9.)

Background: Improved methods of tissue diagnosis for obstruction at the hilum of the liver (porta hepatis) have contributed substantially to the preoperative diagnosis of cholangiocarcinoma. Endoscopic brushing during endoscopic retrograde cholangiopancreatography (ERCP), with sensitivity of 20% to 100%, is the preferred technique for obtaining accurate pathologic results. Extensive hepatic resection with curative intent as well as modern approaches to palliative treatment are based on definitive diagnosis. This is a study involving endosonography-guided, fine-needle aspiration (EUS-FNA) for cytodiagnosis of potentially operable hilar cholangiocarcinoma when brush cytology was negative or unavailable. Methods: Ten consecutive patients (7 men, 3 women; age 47 to 78 years, median 59 years) with bile duct strictures at the hepatic hilum, diagnosed by CT and/or ERCP, underwent EUS-FNA using linear echoendoscopes and 22-gauge needles. Results: Adequate material was obtained in nine patients. Cytology revealed cholangiocarcinoma in seven and hepatocellular carcinoma in one. One benign inflammatory lesion identified on cytology proved to be a false-negative finding by frozen section. Metastatic locoregional hilar lymph nodes were detected in two patients, and in one patient the celiac and para-aortic lymph nodes were aspirated to obtain tissue proof of distant metastasis. There were no complications. Conclusions: When standard methods of tissue diagnosis are inconclusive, EUS-guided FNA may have a potential role in the diagnosis of primary cholangiocarcinoma of the hepatic hilum. As a new, minimally invasive approach, it proved to be technically feasible without significant risks.

Electrohydraulic lithotripsy has been shown to be an effective non-surgical treatment for common bile duct and intrahepatic duct stones. This technique was applied via the endoscopic retrograde route in 14 patients (mean age, 70) with the Mirizzi syndrome, all of whom were at high risk for surgery. The procedure was performed under strict cholangioscopic guidance. Twelve patients had a single stone and two had multiple stones impacted in the cystic duct. Stone diameter was 1.5 cm in one patient, 2 to 3 cm in nine patients, and greater than 3 cm in four patients. Insertion of the babyscope and stone fragmentation were successful in all patients. Complete stone clearance required one session in 12 patients and two sessions in two patients (both with multiple stones). In one patient post-procedural leakage of contrast medium from the cystic duct into the peritoneal cavity was noted. This was attributed to pressure necrosis induced by the impacted stone. The patient had an uneventful course of recovery, and leakage resolved with conservative management. Mortality was zero. Endoscopic treatment of the Mirizzi syndrome with electrohydraulic lithotripsy seems to be an effective and relatively safe alternative to surgery.

Background Pancreatic necrosis and pancreatic abscess are severe complications of acute pancreatitis. Surgery is associated with significant morbidity and mortality in these critically ill patients. Endoscopic therapy has the potential to offer a safer and more effective alternative treatment modality. However, its role needs to be further investigated. Methods This is a retrospective study of the outcome of consecutive patients with pancreatic necrosis and pancreatic abscess, all unfit to undergo surgery, who underwent a new aggressive endoscopic approach. The treatment includes (1) synchronous EUS-guided multiple transmural and/or transpapillary drainage procedures followed by balloon dilation of the cystogastrostoma or cystoduodenostoma, (2) daily endoscopic necrosectomy and saline solution lavage, and (3) sealing of pancreatic fistula by N-butyl-2-cyanoacrylate. Results Pancreatic necrosis and pancreatic abscesses were successfully drained in 13 patients, thus avoiding emergency surgery as an initial treatment. Surgery was completely avoided in 9 patients over a median follow-up of 8.3 months (range 3-81 months). Surgery was combined with endoscopic therapy in one patient because of abscess extension into the right paracolic gutter, which was not manageable by endoscopic drainage. Because of the "disconnected-duct syndrome," two patients later developed recurrent pseudocysts and underwent elective surgery. Complications included minor bleeding after balloon dilation and necrosectomy in 4 cases, which were self limiting or controlled endoscopically. Conclusions This aggressive endoscopic approach shows promising results. It expands the potential for endoscopic treatment in patients with pancreatic necrosis and/or pancreatic abscess.

macroscopic type; as the polyp grows larger and becomes semipedunculated or pedunculated, adenomatous or dysplastic foci appear first, followed by the cancerous lesion. Most adenocarcinomas found within HPs are well differentiated. The occurrence of poorly differentiated adenocarcinoma in an HP is rare. The typical hyperplastic morphology of HP, characterized by ‘‘sawtooth’’ glands and no detectable dysplasia, usually is seen predominantly at the surface. Therefore, biopsy specimens from these polyps may not reveal underlying adenomatous tissue or cancer. The association of cancer and HP is not as clearly established as is the case with adenoma. Nevertheless, we believe and recommend, on the basis of available data that demonstrate that most cases of carcinoma in HP occur in larger polyps, that all gastric HPs larger than 1 cm be resected endoscopically in toto.

Endoscopic papillectomy of benign papillary tumor is still not widely practiced. Intraductal growth has been considered a contraindication for endoscopic therapy. This prospective study evaluates endoscopic papillectomy for treatment of benign papillary tumors without and with intraductal growth.Monofilament snare and monopolar electrocoagulation were used for papillectomy. A 7F stent was placed in the pancreatic duct. Patients with distal intraductal growth underwent sphincterotomy and endoscopic resection after exclusion of more proximal growth.Between February 1985 and April 2004, 106 patients (109 lesions), 68 women, 38 men, median age 68 years (range 29-88 years) were included. Median tumor size was 2 cm (range 0.5-6 cm) with one session (range 1-8) required for removal. Nine patients had invasive carcinoma (8%). Surgery for incomplete removal or recurrence was performed in 12% of 75 patients without and 37% of 31 patients with intraductal growth (p0.01), respectively. Fifteen patients had recurrence (15%); but, only 4 required surgery. Endoscopic resection was curative (median follow-up, 43 months) in 83% without and 46% with intraductal growth (p0.001).Endoscopic papillectomy is safe and effective, and may be feasible in cases of intraductal growth. Surveillance and, if required, re-treatment are mandatory because of the risk of recurrence.

Background The treatment of pancreatic fistula can be difficult. A novel endoscopic approach to sealing pancreatic fistulas by using N-butyl-2-cyanoacrylate is described. Methods Twelve patients with pancreatic fistulas underwent endoscopic injection of N-butyl-2-cyanoacrylate into the fistulous tract, in addition to endoscopic drainage. Results Fistulas were closed successfully in 8 of 12 patients. A single treatment session was successful in 7 patients; a second session was required in one patient. In two patients, closure was temporary, and, in one patient, the treatment failed. One patient died 24 hours after treatment. He developed a pulmonary thromboembolism from a left popliteal vein thrombosis and died from complications of surgical thromboembolectomy. At autopsy, a pulmonary embolus was found, but there was no evidence of N-butyl-2-cyanoacrylate in the lungs. No procedure-related complication occurred over a median follow-up of 20.7 months (range 9-51 months). Conclusions In this preliminary study, occlusion of pancreatic fistulas by using N-butyl-2-cyanoacrylate glue was safe and effective, and obviated the need for surgery in a substantial proportion of patients. Further studies of the use of N-butyl-2-cyanoacrylate for closure of pancreatic fistula are warranted.

Background: Biliary leakage is a problematic complication of hepatobiliary surgery. A novel alternative method is described that can obviate the need for reoperation for refractory biliary fistula. Methods: Nine patients with large biliary leaks unresponsive to endoscopic drainage underwent N-butyl-2-cyanoacrylate glue occlusion at ERCP. Results: In 7 patients, occlusion was successful with prompt control of the fistula in a single session, averting reoperation. In 1 patient there was a partial response and in another the treatment was unsuccessful. No procedure-related complication occurred over a median follow-up of 35 months (range: 1.6-160 months). Conclusion: N-butyl-2-cyanoacrylate glue occlusion is a safe and effective endoscopic method for control of refractory bile leaks that eliminates the need for surgical reintervention.

Background: Metastases to the pancreas are usually found incidentally. Tissue diagnosis is imperative because imaging alone is incapable of differentiating them from primary pancreatic tumors. This study tested whether it is possible to differentiate metastases from other focal pancreatic lesions by using EUS-guided fine-needle aspiration (EUS-FNA) for cytodiagnosis. Methods: One hundred fourteen consecutive patients (mean age 61 years) with focal pancreatic masses, detected on CT, underwent EUS-FNA by using a linear-array echoendoscope and 22-gauge needles. Results: Adequate specimens were obtained from 112 lesions. Carcinomas were identified in 68 cases (60.7%), 56 (50%) of pancreatic origin and 12 (10.7%) from distant primary tumors. The metastases were all located in the head and body of the pancreas and measured 1.8 to 4.0 cm. The echo-texture was heterogeneous or hypoechoic in all cases and resembled that of primary tumors. Six of the 12 patients with metastatic disease had a prior diagnosis of cancer (breast, 3; renal cell, 2; salivary gland, 1), 4 of them with a recurrence and 2 with a second carcinoma metastasizing to the pancreas. Six patients without a prior diagnosis of cancer had metastases from renal cell, colonic, ovarian, and esophageal carcinomas; one metastasis was from an unknown primary and another was from a malignant lymphoma. These findings influenced the therapeutic strategy in 8 patients who underwent nonsurgical palliation. There were no complications. Conclusions: Pancreatic metastasis is an important cause of focal pancreatic lesions, but the EUS features are not diagnostic. Simultaneous EUS-FNA allows cytodiagnosis and can have a decisive influence on the selection of appropriate therapeutic strategies. (Gastrointest Endosc 2001;53:65-70.)

Background: Endoscopic treatment of giant colorectal polyps remains controversial because of concerns regarding coexistent malignancy, incomplete resection, and safety. Methods: We reviewed the clinical course after removal of 176 benign-appearing large (>3 cm) colorectal polyps, which were removed by endoscopic snare resection in 170 patients. These were termed "giant" polyps. Sessile polyps (n = 129) were removed piecemeal and pedunculated polyps (n = 47) transected at the stalk. Results: Bleeding was the only complication in 24% of polypectomy procedures (procedural in 58, immediate in 3, delayed in 6 patients). Except for one conservatively treated delayed bleed, all bleeds were treated endoscopically. Histology of resected polyps showed coexistent malignancy in 12%. Eight patients had malignant polyps that met "unfavorable" criteria and underwent surgery. Following complete endoscopic resection, 16 patients were lost to follow-up and 124 patients had follow-up of at least 6 months (117 benign and 7 "favorable" malignant polyps). Nineteen patients with benign polyps developed recurrences (18 benign, 1 malignant); one patient with a favorable malignant polyp had a malignant recurrence and underwent surgery. Conclusion: Endoscopic resection of benign-appearing giant colorectal polyps is feasible and safe. Complete excision is possible in patients with benign and favorable malignant polyps, but recurrence rates are high. Close surveillance to detect and treat recurrence is required. (Gastrointest Endosc 1996;43:183-8.)

longer enteroscope. Gastrointest Endosc 1991;38:55-8. 3. Lewis BS, Waye JD. Chronic gastrointestinal bleeding of obscure origin: role of small bowel enteroscopy. Gastroenterology 1988;94:1117-20. 4. Gostout CJ, Schroeder KW, Burton DD. Small bowel enteroscopy: an early experience in GI bleeding of unknown origin. Gastrointest Endosc 1991;37:5-8. 5. Swinson CM, Levi AJ. Is coeliac disease underdiagnosed? BMJ 1980;281:1258-60. 6. Mann JG, Brown WR, Kern F Jr. The subtle and variable clinical expressions of gluten-induced enteropathy (adult celiac disease, non-tropical sprue). Am J Med 1970;48:357-66. 7. Ahlquist DA. Fecal blood testing: demystifying the occult. In: Levin B, ed. Proceedings from the National Gastrointestinal Oncology Meeting. GI cancer: current approaches to diagnosis

Background: In spite of improvement of diagnostic modalities tissue diagnosis of obstructions at the liver hilum remains a challenge. Extensive hepatic resections with intention to cure or modern palliative concepts require exact diagnosis, which is rarely achieved preoperatively. Attempts were made by brushings and biopsies during ERCP with variable sensitivity. This is the first report of endosonograpghy guided fine-needle aspiration (EUS-FNA) for cytodiagnosis of hilar cholangiocarcinoma. Methods: 10 patients (7 male, 3 female, age: 47-78, mean age: 62.5) with strictures at the liver hilum, diagnosed by computed tomography and/or ERCP prospectively and consecutively underwent EUS-FNA with longitudinal echoendoscope and 22 gauge needles. Results: In 9 patients adequate material was achieved. Cytology revealed chloangiocarcinoma in 7 patients and hepatocellular carcinoma in one. 8 out of ten patients underwent surgery, in the other two lymph node and liver metastases were detected during pre-surgery examinations. In these EUS-FNA cytology revealed adenocarcinoma. One benign inflammatory lesion on cytology proved to be false negative in frozen section. No complication occurred. Accuracy, sensitivity, and positive predictive value were 89%, 89% and 100%, respectively. Conclusion: Primary cholangiocarcinoma of the hilum can be difficult to discern from other malignancies or benign lesions. These results suggest that EUS-FNA is a new less invasive approach for tissue diagnosis of Klatskin tumors, which for the first time proved to be technically feasible without significant risks.

Background: Mirizzi`s syndrome is defined as extrinsic compression of the common bile duct by an impacted stone in the cystic duct or the neck of the gallbladder. Surgery is often difficult due to extensive inflammation. Aim: Evaluation of endoscopic treatment of Mirizzi`s syndrome in the largest series reported yet. Methods: Retrospective evaluation of patients (pts) presenting with Mirizzi`s syndrome between 1990 and 1999.For initial stabilisation or safe transport to our center, temporary stents or nasobiliary drainage (NBD) were placed. If the stone could be caught into the Dormia basket, mechanical lithotripsy (ML) was performed. Otherwise electrohydraulic lithotripsy (EHL) using the Mother-Baby scope system (Olympus Co.,Tokyo, Japan) and the Walz lithotriptor (Fa. Walz, Rohrdorf, Germany) was performed to fragment stones under cholangioscopic view. Results: 48 f and 16 m pts with a median age of 71years (14-94y) were included. Median duration of biliary symptoms was 22 days (1 d-5 y). Patients presented with pain in 64%, obvious jaundice in 66% and painless jaundice in 22%. 19% were in poor general condition. 9 pts had undergone cholecystectomy median 3 y before (0.1-23 y). An initial endoscopic stent or NBD was placed in 30 pts (47%). Median size of stones was 2cm (0.7- 5cm). In 8 pts ML was performed. 52 pts required EHL. Complete duct clearance was achieved in 59 pts (92%). A single lithotripsy session was sufficient in 54 pts. 4 pts were treated by stenting or nasobiliary drainage only. A complicated course was observed in 4 pts: 2 pts with cystic duct leak at the site of pressure necrosis recovered conservatively.A 71y male not sent for the 2nd session of lithotripsy was treated with long term stenting in another hospital and developed small bowel perforation by the dislodged stent. A 90 y female with septicemia treated by NBD, had myocardial infarction after 2 d and died after 3 d. Subsequently, 12 pts underwent cholecystectomy. 4 of them were symptomatic after endoscopy due to cholelithiasis (3 patients had cholecystitis after 7 d, 7 d and 56 d, respectively; 1 pt had biliary colics after 2 y). 39 pts did not undergo surgery and have remained asymptomatic over a median follow up of 28 months (2-82). None of the pts developed biliary malignancy over a median follow up of 24 (0-85) months. Conclusion: Mirizzi`s syndrome, being considered a clear indication for surgical management, can be safely and effectively treated by endoscopy even in elderly and severely ill pts. Coincidence of gallbladder malignancy with Mirizzi`s syndrome was not observed.

Background: Metastases to pancreas cannot be differentiated from primary pancreatic tumors based on imaging alone. Tissue diagnosis is imperative to plan appropriate treatment. Endosonography guided fine needle aspiration (EUS-FNA) is an alternative for cytodiagnosis of pancreatic lesions. Methods: Amongst 114 consecutive EUS-FNA from pancreatic lesions, metastases were diagnosed in 12 patients (9.5%). Longitudinal echoendoscopes and 22 gauge needles were used for EUS-FNA. Results: Mean age of patients was 61 years (36-74). Out of 6 patients with cancer in history (breast 3, renal cell 2, salivary gland 1) 4 had recurrence while two had a second carcinoma metastasizing into pancreas. In those without prior cancer, cytology revealed metastases from renal cell, colonic, ovarian and esophageal primary tumors in 4 patients, respectively. One had malignant lymphoma, while the other primary site could not be identified. The clinical outcome was poor (median survival 6 months). 8 patients received non-surgical palliation and 3 underwent surgery for their pancreatic metastases. EUS showed metastases in the head/corpus of the pancreas, measuring 1.8-4 cm with inhomogeneous hypoechoic echotexture indistinguishable from primary pancreatic cancer. Conclusion: Pancreatic metastases are one of the important causes of focal pancreatic lesions. EUS features are not characteristic for diagnosis. A simultaneous EUS-FNA enables cytodiagnosis and may avoid surgery.