William Mayoral

Nonmalignant obstruction is a common problem with metal stents in the treatment of esophageal cancer.

BACKGROUND The use of metal stents for the treatment of dysphagia due to esophageal malignancy is an important advance because of ease of delivery and their self-expandable property. Obstruction due to tumor overgrowth is a recognized complication, but nonmalignant obstruction in patients with metal stents is rarely reported. METHODS Database records of patients who had esophageal cancer and underwent metal stent insertion were reviewed. RESULTS A total of 116 patients were seen between October 1993 and October 1997. Four types of metal stents had been used (Ultraflex, Z Stent, Wallstent, and Esophacoil). Detailed follow-up information was available for 81 patients, who constitute the study sample. Forty-nine (60%) stent obstructions were reported, 26 of the 49 (53%) were due to tumor overgrowth and 23 (47%) were not associated with malignancy. Histologic analysis of the nonmalignant obstructing tissue showed granulation tissue (56%), reactive hyperplasia (22%) and fibrosis (22%). CONCLUSIONS Nonmalignant obstruction is a common although infrequently reported complication after placement of metal stents for esophageal cancer. The tissue response of the esophageal mucosa occurred with all 4 types of stents used. No specific characteristic of the stent or prior treatment seems to be related to obstruction of the stent in patients with either nonmalignant obstruction or tumor overgrowth.

Drug-induced liver disease.

Purpose of review To summarize the pertinent case reports, case series and clinical studies that described clinical, histological, epidemiological and mechanistic features of drug-induced liver disease in 2005. Recent findings Acetaminophen, highly active antiretroviral therapy and drugs for tuberculosis retained their preeminent position as the most commonly reported agents causing drug-induced liver disease, with acetaminophen continuing to be the leading cause of acute liver failure in the USA. While the frequency of drug-induced liver disease remains low, a large case-series of acute drug-induced liver disease from Spain and Sweden supported the observation that acute hepatocellular jaundice from a drug is associated with death or the need for transplant in at least 10% (known as Hys Law). With respect to using potentially hepatotoxic medications in patients with underlying liver disease, statins and second-generation thiazolidinediones were shown to be safe when used in patients with elevated baseline alanine aminotransferase or aspartate aminotransferase levels. Summary Drug-induced liver disease remains an important cause of acute liver failure, and research efforts by the National Institutes of Health and others are underway to better determine the risk factors and other host susceptibilities that will allow for the safer use of drugs in the future.

Cost-effective analysis of percutaneous liver biopsy (PLB) with or without ultrasound guidance (USG): results and outcomes

Cost-effective analysis of percutaneous liver biopsy (PLB) with or without ultrasound guidance (USG): results and outcomes

Cardiorespiratory complications during gastrointestinal endoscopy: prospective analysis of frequency and outcome

Cardiorespiratory complications during gastrointestinal endoscopy: prospective analysis of frequency and outcome

4664 Sphincter of oddi manometry and endoscopic sphincterotomy in patients with pancreatitis of unknown etiology: a long-term study and results in a tertiary referral center.

Sphincter of Oddi dysfunction (SOD) is responsible for 15% of PUE. Limited data is available about the benefit and long-term outcome of ES. The aim of this study was to evaluate the long-term outcome of ES in this group of patients. Methods: All SOM performed for PUE were reviewed. A total of 67 studies were identified. A basal sphincter of Oddi pressure (SOP) greater than 40mmHg was considered abnormal. Nineteen out of 67 patients (28%) with clinical pancreatitis had elevated SOP.All patients had biliary sphincter manometry (BSM). If the pressure was high, biliary sphincterotomy (BS) was performed. If BSM was normal, pancreatic sphincter manometry was performed (PSM), and if high, pancreatic sphincterotomy (PS) was performed. Initially, 18 patients had BS and 1 patient had PS. All patients were admitted overnight for observation. Further follow-up was obtained by telephone interview using a standard protocol, which included 21 variables addressing pain resolution, narcotic use, the need for hospital admissions or ER visits, and any further intervention. Results: Complete follow-up was obtained in all 19 patients. The mean follow up time was 54 months (Range 6 -126). Patients were categorized in 3 groups according to the presence or absence of pain with or without narcotics, and the recurrence or not of documented pancreatitis. Eleven patients (58%) were pain free. Four patients (23.5%) had mild pain that did not require narcotics, and four patients (21%) had pain that required narcotics on a daily basis. Four patients (21%) required a second procedure, which included a SOM. Two of them had a PS because of elevated pancreatic sphincter pressure. Extension of the original sphincterotomy was performed in the remaining two because of residual biliary sphincter pressure. Complications occurred in 4 patients (21%). Three had mild pancreatitis and one patient with prior sphincteroplasty had a duodenal perforation requiring surgery. Globally, fifteen patients (79%) felt better after the ES was performed, and 4 patients (21%) had no improvement. Conclusions: 1-SOD is frequently identified in patients with PUE. 2- Initial measurement of biliary sphincter pressure is adequate for identification and treatment of most patients with pancreatitis and SOD. 3- Isolated pancreatic sphincter dysfunction requiring PS is infrequent. 4- Further studies are required to confirm these results.

4669 Endoscopic biliary sphincterotomy in patients with post cholecystectomy pain and sphincter of oddi dysfunction: an effective and long lasting therapy.

Persistent abdominal pain is present in 15-20% of post cholecystectomy patients. After excluding other causes, patients can be defined using the Milwaukee classification and sphincter of Oddi manometry (SOM). Currently, limited data is available in reference to the long -term impact of biliary sphincterotomy on the clinical course of patients with elevated sphincter of Oddi pressure (SOP). The aim of this study was to evaluate the long-term outcome of endoscopic biliary sphincterotomy in patients with SOD. Methods: All SOM performed for post cholecystectomy pain and suspected SOD were reviewed. Patients with pancreatitis were excluded.A total of 92 manometries were identified. A basal SOP greater than 40 was considered to be abnormal(SOD). All measurements were from the biliary sphincter. No attempt was made to measure the pancreatic sphincter pressure. Thirty three patients (36%) had SOD. All patients with SOD underwent biliary sphincterotomy. Patients were admitted overnight for observation. Further follow-up was obtained by telephone interview using a standard protocol which included 21 variables addressing pain resolution, narcotic use, the need for hospital admissions or ER visits, and any further intervention. Results: Complete telephone interview and follow-up was obtained in all 33 patients. The mean follow-up was 45 months (range 6- 134). Patients were classified using Milwaukee classification into 3 groups: type I (1 patient); type II (7 patients) and type III (25 patients). 15(45%) patients were pain free (Type I= 1 patient; Type II=2 patients; Type 3= 12 patients); 5 (15%) patients had mild pain that did not require narcotics (Type II= 2 patients); (Type 3= 3 patients). 9 (27%) of patients did not improve (Type II=2 patients; Type III :7 patients). Six patients underwent a second ERCP with SOM. Four patients had elevated pancreatic sphincter pressure and underwent pancreatic sphincterotomy. In 2 patients, residual biliary pressure was present, and underwent extension of the biliary sphincterotomy. Complications occurred in 4 patients (12%) all of which consisted of mild pancreatitis. Overall, 25 patients (75%) felt better after biliary sphincterotomy.Conclusions:1- Biliary sphincterotomy is safe an effective in the majority of patients with SOD. 2-Benefits appear to be long term.3- Routine measurement of pancreatic sphincter pressure is not needed initially in this group of patients.

4565 Value of preoperative staging for pancreatic masses suspicious of malignancy.

Introduction: Imaging studies of pancreatic cancer have been reported to be accurate in the preoperative staging of pancreatic cancer. Systematic assessment of imaging studies regardless of a positive (malignancy) or negative (benign) Whipple or other surgical procedure has not been carefully assessed. We therefore retrospectively reviewed the results of preoperative imaging of pancreatic masses suspicious of cancer com of the histology of resection specimen and/or surgery. Methods: Between 10/96 and 10/99 CT, ERCP and EUS was performed approximately 2-4 weeks prior to surgery in 62 patients with suspected pancreatic cancer. The results of imaging were correlated with that of the histology of resection specimen or surgical findings. Results: Table summarizes the results of CT, ERCP, and EUS 38 patients underwent a Whipple procedure (36 adenocarcinomas; 1 pancreatic Crohn s disease, 1 pancreatitis with abscess); 9 patients underwent non Whipple surgery, and 15 were found to be unresectable at exploratorty laparotomy. Adenocarcinoma of the pancreas was found in 52 (84%) patients and benign pathology was found in 10 patients(16%). The accuracy for both CT , ERCP and EUS were 84%; 84% and 87% respectively. The positive predictive value (PPV) of CT, ERCP and EUS were of 84%,90.% and 87%, respectively. Of all patients deemed unresectable at surgery (n= 15), all had evidence of vascular invasion at the time of EUS. Conclusions: 1- CT, ERCP and EUS for preoperative staging do not show any significant difference in predicting malignancy. 2- In this series, 16% of cases had false positive imaging studies. 3- Surgery is still recommended in cases of positive imaging studies. 4- Pre-operative tissue diagnosis may be helpful to avoid unnecessary surgery.

3350 The complication rate of duodenal polypectomy.

Endoscopic polypectomy has generally been considered a safe procedure for gastric and colonic polyps. The risk of duodenal polypectomy has not been clearly established. Aim: To evaluate the complication rate subsequent to duodenal polypectomy in comparison to polypectomy of gastric or right colon polyps. Methods: Complications associated with duodenal, right colon or gastric polypectomy were identified by review of the CORI database, morbidity and mortality records, and chart review over the last three years. Right colonic polyps were classified as those identified in the cecum, ascending colon or hepatic flexure. Procedure reports were reviewed to identify polyps at least 10mm that were removed by polypectomy. Demographic data as well polyp characteristics (sessile vs pedunculated) were included in the analysis. Results: Overall, there were 250 right colonic polypectomies performed, 60 gastric polypectomies, and 26 duodenal polypectomies. Of these, 38 right colon polyps, 13 gastric polyps, and 15 duodenal polyps were 10mm or more in size. Complications associated with polypectomy performed in these polyps 10mm or greater in size are shown below. Patients undergoing duodenal polypectomy had a complication rate of 26% vs. 9.8% for gastric and right colon polypectomies combined (P=0.19; CI 0.83-8.7). One patient undergoing duodenal polypectomy had a perforation and underwent surgery. There was no procedure-related mortality. There was no difference in age or morphology of the polyps between groups. All complications occurred with removal of sessile polyps. Conclusion: In patients with polyps 10 mm or greater in size undergoing polypectomy at our institution, those that had duodenal polypectomy had a trend toward a higher complication rate than those having polypectomy in the right colon or stomach.

7143 Gastrointestinal manifestations in patients with classical and hypermobile ehlers-danlos syndrome.

Ehlers-Danlos syndrome (EDS) is a heterogeneous group of inheritable connective tissue disorders. Gastrointestinal (GI) manifestations including giant epiphrenic diverticula, mega-esophagus, GI bleeding and spontaneous bowel perforation have been reported without specification of EDS type. There has been minimal systematic evaluation of GI symptomatology in patients with EDS, and no evaluation of GI symptomatology specifically in the subgroup with classical or hypermobile EDS. Aim: To prospectively evaluate GI symptoms of patients with classical or hypermobile EDS referred to the GI service at the National Institute of Health (NIH). Methods: Overall, 23 patients with classical or hypermobile EDS from NIH EDS protocol were evaluated by the GI service between July 1998 and December 1999. Endoscopic evaluation was performed in consenting patients. Ten patients underwent EGD, 7 underwent colonoscopy, and 3 underwent catheter ultrasound (CUS) during endoscopy.Wall thickness of both bowel and descending aorta was assessed by CUS. GI symtomatology are as follows: (see table.)Results: Of patients undergoing EGD, 3 had a hiatal hernia (30%), 1 had antral erosions, and 2 had esophageal inflammation. On colonoscopy, 3 (43%) were reported to have a redundant colon, 1 had colonic inflammation, and 1 had submucosal vascular inflammation. Average wall thickness at CUS (total thickness/ muscularis propria in mm) was as follows: esophagus (2.9/2.2), gastric antrum (3.6), duodenum (2.1/1.1), descending colon (2.9/1.5), sigmoid (2.7/1.2), rectum (2.9/1.0). The average size of the descending aorta in the thorax at CUS was 18mm (largest one was 24 mm), and its average size in the abdomen was 12.8mm. Conclusion: Gastrointestinal symptoms are common in patients with classical or hypermobile EDS. The most common symptoms are probably related to GERD or IBS. Mucosal findings occur in this symptomatic population but are not frequent. Hiatal hernias, redundant colon, and thickening of the intestinal wall (especially the muscularis propria) may occur. This study is ongoing.

7178 Fna and mucosectomy in submucosal gi-lesions in adjunct to eus/cus.

EUS and catheter ultrasound (CUS) have been accurate in submucosal GIlesions. Fine needle aspiration (FNA) and endoscopic mucosal resection (EMR) have been reported to enhance the ability to correctly diagnose submucosal lesions. Aim: To characterize submucosal lesions of the upper GI tract diagnosed by EUS with either FNA or EMR. Methods: Patients referred for evaluation of submucosal lesions between January 1998 and November 1999 were retrospectively reviewed. All patients were evaluated with CUS and EUS (Radial, Linear array or both). The patients were analyzed in two groups. Group I consisted of 21 patients in whom FNA was performed. Group II consisted of 7 patients in whom EMR was performed. EMR was only performed in patients with lesions confined to the mucosa or submucosa. In one patient both FNA and EMR was performed. This patient was included in both groups. Results: Overall, 29 patients were included in this study. FNA was performed in 22 patients. The mean number of passes with the FNA needle was 3, and the range was 2-8. A diagnosis was rendered in 27% (6/22) of patients in whom FNA was performed. EMR was performed in 8 patients. A diagnosis was rendered in all 8 patients in whom EMR was performed (100%). Of all submucosal lesions evaluated 10% (3/29) were potentially malignant. All potentially malignant tumors were carcinoids. In the one case where both biopsy techniques were used, both the FNA sample and the EMR sample revealed carcinoid. No complications occurred. Specific diagnosis determined at FNA or EMR are below. Conclusions: 1- FNA was diagnostic in only 27 % of patients with submucosal lesions 2. EMR was accurate in obtaining the final diagnosis in all cases. 3. EMR was safely performed in lesions up to 26mm. 4. 10% of submucosal lesions were potentially malignant