Frequency of visualization of presumed celiac ganglia by endoscopic ultrasound.

BACKGROUND AND STUDY AIMS: Celiac ganglia can be visualized by endoscopic ultrasound (EUS). It is unknown how often ganglia are visualized during EUS, and what clinical factors are associated with ganglion visualization. The aim of this study was to prospectively evaluate the frequency of visualization of presumed celiac ganglia by EUS and to identify factors that predict their visualization. PATIENTS AND METHODS: Clinical, demographic, EUS, and cytologic data were collected prospectively from 200 unselected patients who were undergoing EUS in a tertiary referral centre. When presumed celiac ganglia were visualized, their size, number, location, and echo features were noted. When presumed ganglia were aspirated, the results of cytology were recorded. RESULTS: The most common indication for EUS was investigation of a pancreatic mass or cyst (25 %). Presumed celiac ganglia were identified in 81 % of patients overall. Logistic regression analysis determined that female sex and having no prior history of gastrointestinal surgery were independently associated with ganglion visualization. Among patients whose ganglia were visualized, more ganglia were seen per patient with linear echo endoscopes (2, range 0 - 5) than with radial echo endoscopes (1, range 0 - 4) ( P = 0.001). Presumed celiac ganglia were aspirated in 10 patients; and cytologic examination revealed neural ganglia in all of these. CONCLUSIONS: Celiac ganglia can be visualized by EUS in most patients who undergo upper gastrointestinal EUS examinations, and are best seen with linear-array echo endoscopes. Ganglia can usually be differentiated from lymph nodes on the basis of their endosonographic appearance.

A comparison of endoscopic ultrasound, magnetic resonance imaging, and exam under anesthesia for evaluation of Crohn's perianal fistulas.

BACKGROUND & AIMS: To determine accuracy of endoscopic ultrasound (EUS) and magnetic resonance imaging (MRI) for evaluation of Crohn's disease perianal fistulas. METHODS: Thirty-four patients with suspected Crohn's disease perianal fistulas were prospectively enrolled in a blinded study comparing EUS, MRI, and examination under anesthesia (EUA). Fistulas were classified according to Parks' criteria, and a consensus gold standard was determined for each patient. Acceptable accuracy was defined as agreement with the consensus gold standard for > or =85% of patients. RESULTS: Three patients did not undergo MRI; 1 did not undergo EUS or EUA; and consensus could not be reached for 1. Thirty-two patients had 39 fistulas (20 trans-sphincteric, 5 extra-sphincteric, 6 recto-vaginal, 8 others) and 13 abscesses. The accuracy of all 3 modalities was > or =85%: EUS 29 of 32 (91%, confidence interval [CI] 75%-98%), MRI 26 of 30 (87%, CI 69%-96%), and EUA 29 of 32 (91%, CI 75%-98%). Accuracy was 100% when any 2 tests were combined. CONCLUSIONS: EUS, MRI, and EUA are accurate tests for determining fistula anatomy in patients with perianal Crohn's disease. The optimal approach may be combining any 2 of the 3 methods.

Neural network analysis of EUS images to differentiate between pancreatic malignancy and pancreatitis.

BACKGROUND: The differentiation of focal pancreatitis and pancreatic adenocarcinoma is problematic and often resolved only by pancreaticoduodenectomy. EUS is the most sensitive imaging modality for both conditions, yet ultrasonic criteria for distinguishing the two have not been described and differentiation remains difficult. The aims of this study were to develop a self-learning computer program that can analyze EUS images and differentiate malignancy from pancreatitis, and to compare results obtained with this system with EUS interpretation by experienced endosonographers. METHODS: Twenty-one patients with pancreatic cancer and 14 with focal pancreatitis were included. The diagnosis was confirmed histologically in all cases and each patient had undergone EUS. A single EUS image from each procedure was used for computer analysis. The results were compared with the EUS diagnosis reported at the actual procedure as well that of an endosonographer who reviewed videotapes of the procedures. RESULTS: The software program differentiated focal pancreatitis from malignancy with a maximal 89% accuracy. With sensitivity set at 100% for malignancy, the program was 50% specific and accuracy was 80%. Sensitivity and accuracy of the endosonographer's impression at the time of EUS were, respectively, 89% and 85%. A sensitivity of 73% and accuracy of 83% were achieved with blinded interpretation of EUS videotapes. CONCLUSIONS: Analysis of EUS images with computer software programs is feasible and compares favorably with human interpretation. The application of this technology to EUS and other imaging scenarios could be a useful adjunct to diagnostic endoscopy and warrants further investigation.

Utility of endoscopic ultrasonography in endoscopic drainage of pancreatic pseudocysts in selected patients.

OBJECTIVES: To determine the effect of endoscopic ultrasonography (EUS) on endoscopic drainage of pancreatic pseudocysts and to determine patency with fistula dilation and placement of multiple stents. PATIENTS AND METHODS: Between September 1995 and January 1999, 19 patients underwent endoscopic drainage of pancreatic pseudocysts, 17 of whom were assessed by EUS before drainage. Radial EUS scanning was used to detect an optimal site of apposition of pseudocyst and gut wall, free of intervening vessels. A fistula was created with a fistulatome, followed by balloon dilation of the fistula tract. Patency was maintained with multiple double pigtail stents. The primary goal of this retrospective study was to determine whether EUS affected the practice of endoscopic drainage of pancreatic pseudocysts. RESULTS: In 3 patients, drainage was not attempted based on EUS findings. In the other 13 patients (14 pseudocysts), creation of a fistula was successful on 13 occasions, and no immediate complications occurred. However, 1 patient subsequently developed sepsis that required surgery. All other patients were treated with balloon dilation, multiple stents, and antibiotics, with no septic complications. Of 14 pseudocysts (in 13 patients), 13 (93%) resolved. CONCLUSIONS: Results of EUS may alter management of patients considered for endoscopic drainage of pancreatic pseudocysts. Endoscopic ultrasonography was useful for selecting an optimal and safe drainage site. The combination of balloon dilation, multiple stents, and antibiotics appears to resolve pancreatic pseudocysts without septic complications.

Impact of EUS-guided fine-needle aspiration on lymph node staging in patients with esophageal carcinoma.

BACKGROUND: Preoperative identification of lymph node metastases associated with esophageal carcinoma may influence treatment. EUS is the most accurate method for locoregional staging of these tumors. The impact of EUS-guided fine-needle aspiration (EUS-FNA) on lymph node staging in esophageal carcinoma is unclear. METHODS: From May 1996 to May 1999, 74 patients with esophageal carcinoma underwent preoperative EUS. After October 1998 EUS-guided FNA was performed on nonperitumoral lymph nodes greater than 5 mm in width. The results of EUS with and without FNA were retrospectively reviewed and compared. Final diagnosis was based on surgical results or EUS-guided FNA malignant cytology. Ten of the 74 patients had to be excluded for lack of lymph node stage confirmation. Final diagnosis was obtained in the remaining 64 patients (33 from the EUS only group and 31 from the EUS-FNA group). RESULTS: The results of EUS versus EUS-FNA for lymph node staging were sensitivity 63% versus 93% (p = 0.01), specificity 81% versus 100% (not significant), and accuracy 70% versus 93% (p = 0.02), respectively. Complications comprised 1 patient who developed self-limited bleeding after dilation that did not preclude completion of the EUS (1%, 95% CI [0%, 7%]). CONCLUSIONS: EUS-FNA is more sensitive and accurate than EUS alone for preoperative staging of locoregional and celiac lymph nodes associated with esophageal carcinoma. EUS-FNA of nonperitumoral lymph nodes in patients with esophageal carcinoma is safe and should be routinely performed when treatment decisions will be affected by nodal stage.

EUS-guided fine-needle aspiration combined with flow cytometry and immunocytochemistry in the diagnosis of lymphoma.

BACKGROUND: Limited information is available regarding the use of EUS-guided fine-needle aspiration (EUS-FNA) in the diagnosis of lymphoproliferative disorders. The aim of this study was to evaluate the yield of this technique in the primary diagnosis of lymphoma. METHODS: The records were reviewed of 38 consecutive patients with GI lesions and/or enlarged lymph nodes identified on imaging studies that raised a suspicion of lymphoma who underwent EUS-FNA of lymph nodes or the gut wall. Final diagnosis was based on clinical follow-up, imaging studies, or surgical findings. RESULTS: Twenty-three patients with lymphoma and 15 patients with benign disease or reactive lymphadenopathy were identified. The overall sensitivity, specificity, and accuracy of EUS-FNA cytology with flow cytometry/immunocytochemistry (FC/IC) for the diagnosis of lymphoma were, respectively, 74%, 93%, and 81%. When comparing patients who had EUS-FNA with FC/IC versus those who had EUS-FNA without FC/IC, sensitivity was 86% versus 44% (p = 0.04), specificity was 100% versus 90% (not significant), and accuracy was 89% versus 68% (not significant). CONCLUSION: EUS-FNA can provide cytology specimens diagnostic for lymphoma. Selective use of FC/IC in patients with suspected lymphoma improves the yield of EUS-FNA and may guide diagnostic evaluation and treatment decisions.

Acute extraluminal hemorrhage associated with EUS-guided fine needle aspiration: frequency and clinical significance.

BACKGROUND: Complications with EUS-guided fine needle aspiration cytology (EUS-guided FNA) are rare and include perforation, infection, pancreatitis, and intraluminal bleeding. To date, the ultrasound appearance and clinical significance of perilesional bleeding during EUS-guided FNA have not been described. The aim of this study was to analyze the frequency of acute extraluminal hemorrhage associated with EUS-guided FNA. METHODS: From September 1998 to October 1999 EUS-guided FNA was performed during 227 of 1104 EUS procedures. Patient follow-up and complications were recorded and retrospectively analyzed. OBSERVATIONS: Three patients were identified with acute extraluminal hemorrhage at the site of the aspiration during EUS (frequency 1.3%: 95% CI [0%, 2.8%]). The bleeding manifested as an expanding echopoor region adjacent to the sampled lesion. No clinically recognizable sequela arose from the bleeding. All patients were treated with a short course of antibiotics and outpatient observation. Preprocedure coagulation and platelet assessment did not predict which patients were at risk for this complication. CONCLUSION: Acute extraluminal hemorrhage occurring during EUS-guided FNA is a rare complication with a characteristic ultrasound appearance. Recognition of this event might be important to allow the endoscopist to terminate the procedure and thereby minimize the potential for more serious bleeding.

Diagnosis of chronic pancreatitis. Is a gold standard necessary?

The question that serves as this article's title is rhetorical. Clinicians have diagnosed and managed chronic pancreatitis without a gold standard for decades and must continue to do so in the foreseeable future. Although clinicians have a much wider array of diagnostic tools available for the diagnosis of chronic pancreatitis, a single readily applied gold standard remains elusive. Diagnostic studies are rarely compared with a true gold standard--histopathology. Furthermore, even if a safe biopsy technique were available, it might fall short of a gold standard, given the patchy nature of early-stage chronic pancreatitis. Indeed, different stages of chronic pancreatitis require not only recognition of the different clinical presentations but also different levels of intensity of diagnostic testing to establish the diagnosis confidently. The diagnosis in most patients with chronic pancreatitis can be made confidently with a good clinical history and a limited number of currently available structural and functional tests. No single diagnostic study, functional or structural, suffices for all patients. It is also axiomatic that patients with intractable abdominal pain in whom early-stage chronic pancreatitis is suspected represent a challenge for clinicians partly because of this lack of a single, dependable gold standard. Perhaps we have reached the point at which further refinements of current tests of structure or function are not beneficial because increased sensitivity is countered by loss of specificity. We suggest that a new approach to developing a gold standard for the diagnosis of chronic pancreatitis is necessary. With advances in the understanding of the mediators of the inflammatory process, it may be possible to devise a test to assess the earliest events in this disease.

The different courses of early- and late-onset idiopathic and alcoholic chronic pancreatitis.

BACKGROUND/AIMS: Compared with alcoholic pancreatitis, little is known about the natural history of idiopathic pancreatitis. Two hundred forty-nine patients with alcoholic pancreatitis and 66 patients with idiopathic chronic pancreatitis seen at our institution between 1976 and 1982 were investigated. METHODS: Records were analyzed retrospectively from the onset of symptomatic disease, and patients were followed up prospectively until 1985. Patients with early-onset (n = 25) and late-onset (n = 41) idiopathic chronic pancreatitis had a median age at onset of symptoms of 19 and 56 years, respectively. RESULTS: The gender distribution was nearly equal in idiopathic chronic pancreatitis, but 72% of patients with alcoholic pancreatitis were men (P = 0.001 vs. idiopathic). In early-onset idiopathic pancreatitis, calcification and exocrine and endocrine insufficiency developed more slowly than in late-onset idiopathic and alcoholic pancreatitis (P = 0.03). However, in early idiopathic chronic pancreatitis, pain frequently occurred initially (P = 0.003 vs. late and alcoholic) and was more severe (P = 0.04 vs. late and alcoholic). In late-onset idiopathic pancreatitis, pain was absent in nearly 50% of patients. CONCLUSIONS: There are two distinct forms of idiopathic chronic pancreatitis. Patients with early-onset pancreatitis have initially and thereafter a long course of severe pain but slowly develop morphological and functional pancreatic damage, whereas patients with late-onset pancreatitis have a mild and often a painless course. Both forms differ from alcoholic pancreatitis in their equal gender distribution and a much slower rate of calcification.

Acute gastrointestinal bleeding. Experience of a specialized management team.

The initial experience of a specialized management team organized to provide expedient care for all acute major gastrointestinal bleeding in protocolized fashion at a large referral center is presented. Of the 417 patients, 56% developed bleeding while hospitalized. Upper gastrointestinal bleeding accounted for 82%. The five most common etiologies included gastric ulcers (83 patients), duodenal ulcers (67 patients), erosions (41 patients), varices (35 patients), and diverticulosis (29 patients). Nonsteroidal anti-inflammatory drugs were implicated in 53% of gastroduodenal ulcers. The incidence of nonbleeding visible vessels was 42% in gastric and 54% in duodenal ulcers. The rates of rebleeding were 24% (20 patients) in gastric ulcers and 28% (19 patients) in duodenal ulcers. Predictive factors for rebleeding included copious bright red blood, active arterial streaming, spurting, or a densely adherent clot. The rebleeding rate for esophagogastric varices was 57%. The mortality rate overall was 6% (27 patients), with rates varying from 3% (five patients) for gastroduodenal ulcers to 40% (14 patients) for esophagogastric varices. The morbidity rate for the entire patient population was 18% (77 patients), dominated by myocardial events (34 patients). The average length of hospitalization for gastroduodenal ulcers was 5 days, for diverticulosis 8 days, and for varices 10 days. The major efforts of a specialized Gastrointestinal Bleeding Team would be best directed at both reducing the morbidity associated with acute bleeding and reducing the overall cost of care.

Influence of hepatic reserve and cause of esophageal varices on survival and rebleeding before and after the introduction of sclerotherapy: a retrospective analysis.

Esophageal variceal sclerotherapy has been enthusiastically accepted as the procedure of choice for patients with variceal hemorrhage. Because the relationships among liver function, different causes of varices, survival, and rebleeding rates have not been well established in sclerotherapy trials, this enthusiasm may be unjustified. We studied these relationships in 80 patients with bleeding esophageal varices who were admitted to hospitals affiliated with our clinic between 1978 and 1980 and who did not receive sclerotherapy and in 162 patients admitted between 1980 and 1982 who received sclerotherapy with ethanolamine oleate. In both groups of patients, survival and bleeding-free intervals were significantly related (P less than 0.005 and P less than 0.01, respectively) to hepatic reserve (Child's class). In addition, patients with nonalcohol-related liver disease and poor hepatic reserve (Child's class C) had reduced survival and bleeding-free intervals compared with patients in class C with alcohol-related liver disease. Similar probabilities of survival and bleeding-free intervals were noted for Child's class subgroups and etiologic subgroups in the sclerotherapy and nonsclerotherapy groups, although a formal comparison was not made because of the retrospective nature of this study. Indications that sclerotherapy increases survival and reduces rebleeding may be due to different distributions of Child's classes and causes of varices within sclerotherapy and nonsclerotherapy groups in published control trials.

Ultrasonography and computed tomography in focal fatty liver. Report of two cases with special emphasis on changing appearances over time.

The features of focal fatty infiltration of the liver are described in 2 patients. In 1 patient, hepatic malignancy was initially suspected, and the true nature of the lesion was not appreciated until months later. The nondiagnostic features on ultrasonography and the characteristic features on computed tomography are described. Additional diagnostic information was obtained by guided liver biopsy and from repeat computed tomography months later, when partial or total resolution of the lesions was observed.

Human endoscopic ultrasonography.

Conventional transcutaneous ultrasound examinations are often compromised by intervening pulmonary or bowel gas and have limited resolution. Ultrasonic probes of frequencies greater than 5 MHz, which enhance resolution, cannot be used successfully on the skin surface, because they do not penetrate deeply enough to view intraabdominal organs in most adults. To overcome these problems, we tested an ultrasonic endoscope which had a 10-MHz, 64-element linear assay, generated real-time images at resolutions of less than 1 mm, and was an integral part of a 35-mm-long and 13-mm-wide endoscopic rigid tip. Thirty-two studies were performed in 15 healthy subjects, 4 patients with pancreatic cancer and 6 patients with chronic pancreatitis, and 1 patient each with a gastric ulcer and a suspected pancreatic abscess. We demonstrated that this procedure is safe, provides high resolution real-time ultrasound visualization of the heart, aorta, spleen, pancreas, liver, gallbladder, kidneys, and gastrointestinal mucosa and can detect moderate-sized pancreatic tumors and hepatic metastases less than 1 cm in diameter. Because endoscopic visualization of gastrointestinal mucosa and ultrasound examination of extraluminal organs can be obtained during a single procedure, rapid differentiation among mucosal and intramural disease of the hollow gut and disease of extraluminal organs should be possible with this diagnostic technique.

Is pancreatic duct obstruction or stricture a major cause of pain in calcific pancreatitis?

Forty-seven patients with calcific pancreatitis were submitted to ERCP, 24 of whom were designated 'painless' and 23 'painful', according to whether or not pain had occurred during the previous year. The pancreatic duct was cannulated in 39 of these patients. The incidence of pancreatic duct obstruction or stricture was of the same order in the painless (13/20, 65 per cent) and painful (15/19, 79 per cent) groups. Alcohol was the aetiological factor in 46 patients. Twenty-four (52 per cent) continued to drink, and were equally divided between the painless (12) and painful (12) groups. Four out of 13 in the painless and 7 out of 14 in the painful groups, with a stricture or obstruction, continued to drink. The difference was statistically not significant. These results suggest that ERCP findings of pancreatic duct obstruction or stricture need not necessarily be pertinent to the pain in patients with chronic pancreatitis.

Colonoscopy in lower gastro-intestinal haemorrhage.

The value of colonoscopy in the diagnosis and management of lower gastro-intestinal haemorrhage was assessed in 107 patients who were examined during a 2-year period. The main indications were to define uncertain radiological findings, to further investigate the cause of bleeding in patients with normal barium studies and to carry out polypectomy. Radiological lesions were confirmed in 27, defined in 12, and refuted in 14 patients. Twenty-eight lesions were demonstrated on colonoscopy which were not diagnosed by barium enema or sigmoidoscopy. A diagnosis was made by colonoscopy in more than two-thirds of the patients with frank blood loss and normal radiographs. In cases of occult bleeding it was farless helpful. Colonoscopy significantly improved diagnosis and management of gastro-intestinal bleeding when the cause was in doubt after standard investigations. In addition, 37 polyps were removed through the colonoscope from 30 patients.

The Mallory-Weiss syndrome. A prospective study in 130 patients.

The clinical features of the Mallory-Weiss syndrome were prospectively documented in 130 of 1 667 patients submitted to endoscopy for gastro-intestinal haemorrhage, an incidence of 7,8%. The important clinical features found in these patients were: alcohol abuse (21%); retching or vomiting (38%); salicylate ingestion (36%); dyspepsia (39%). The correct clinical diagnosis was therefore often difficult to make. Additional lesions were found in 40% of patients at endoscopy. Blood loss was relatively small, and surgery was not required in any patient. The 2 deaths that occurred were not attributable to haemorrhage. A high index of suspicion and early endoscopy are required to establish the diagnosis.