Methods: Between January 2002 and July 2008, all patients who underwent endoscopic treatment of their pancreatic duct disruption were followed prospectively until complete resolution (nZ114 cases involving 113 patients). All patients underwent pancreatic sphincterotomy. In 72 cases a pancreatic duct stent was placed, of which 47 stents crossed the disruption. Success was defined as resolution of the leak at 12 months confirmed by ERCP, MRCP or CT post stent removal. Logistic regression analysis was performed on the following variables with regard to their ability to predict complete resolution: age, gender, etiology, chronicity, pancreatic stenting, endoscopic pseudocyst drainage, percutaneous drainage, type of disruption (partial vs. complete), location of disruption (pancreatic body vs. other), and enteral feeding. Results: 113 patients (78 male), mean age 51.25 years (range 11-86 years) were included. 114 total cases were evaluated. 108 cases presented with pancreatic fluid collection, 3 with pancreatic ascites and 4 with fistula. Resolution occurred in 80 cases (70.2%) at 12 months; while 20 cases (17.5%) did not resolve. 9 cases (7.9%) were lost to follow-up within 12 months of presentation and 5 (4.4%) mortalities occurred unrelated to the procedures. 72 cases (63.2%) underwent pancreatic duct stenting of which 51 resolved (71%). 68 (59.6%) cases required cystenterostomy, of which 51 resolved (75%), and 5 cases (4.4%) required percutaneous drainage. Enteral feeding was employed in 81 cases (71.7%). Logistic regression analysis failed to identify any factor predictive of resolution. In partial duct disruptions, pancreatic duct stenting with endoscopic drainage of fluid collections resulted in an increased rate of resolution at 12 months (76.6%). In cases of complete pancreatic duct leak, the use of pancreatic duct stenting with endoscopic drainage of fluid collections had no additional benefit (14/25, 56%) compared to cystenterostomy alone (24/38, 63.1%) (P valueZ 0.61). Conclusions: Stenting of pancreatic duct disruptions should only be performed in cases of partial disruption. Complete ductal disruption appears better managed with transenteric drainage of the associated fluid collection. Future randomized control trials are needed for validating these results. T1323 Real-Time Intraductal Confocal Microscopy During ERCP: Feasibility and Technical Considerations Peter D. Stevens, Yang K. Chen, Raj J. Shah, Ram Chuttani, Douglas K. Pleskow Background: Despite advancements in ERCP technology (digital fluoroscopy, choledochoscopy, intraductal ultrasound, and tissue acquisition devices), accurate diagnosis of lesions remains a fundamental problem. Confocal laser microscopy has been reported to be a useful tool for in vivo imaging of cellular structures and can be used during endoscopy using a miniprobe (Cellvizio, Mauna Kea Technologies, Paris, France) passed through the working channel of an endoscope. This system enables in vivo histopathologic imaging of the digestive epithelium. Recently a Cellvizio miniprobe has been developed for imaging the biliary and pancreatic duct. The aim of this study is to demonstrate the feasibility of intraductal confocal imaging and to describe the optimal imaging methods. Methods: Patients referred for ERCP to one of 5 experienced pancreaticobiliary endoscopists at one of 3 centers were eligible for enrollment. Study subjects were chosen based on expected finding of either a discrete lesion to be targeted for imaging or a diffuse field change to be examined. Following performance of standard ERCP, the confocal probe was placed in contact with the area of interest via a SpyGlass choledochoscope (Boston Scientific, Natick, Mass), a single lumen catheter (Cook Medical, Winston Salem, NC), or an H-BIN catheter (Cook Medical). Intravenous fluorescein was given to enable imaging. Real-time images were recorded. Results: Confocal imaging of the bile or pancreatic duct was attempted 23 times in 22 patients. In 11 cases, the SpyGlass DVS was used. The target (10 biliary,1 pancreatic) was reached in all cases (11/11) However, the operators noted that the images were unstable and unsatisfactory for interpretation in 7/11 attempts. In 8 cases (7 biliary, 1 pancreatic), the Cook H-Bin catheter was used to direct the probe against the wall of the duct to obtain images. The target was reached in all cases (7/7) including. Images were reported as easy to obtain and stable for interpretation. In 4 cases a single lumen catheter was used for the probe and stable images were obtained in 3. Conclusion. Initial experience with the CellVizio confocal system shows that images from within the bile and pancreatic ducts can be obtained for real time interpretation. Contrary to initial expectations, images were more easily obtained using catheters rather than choledochoscopy. Work is in continuing to determine the optimal imaging methodology. T1324 Transpapillary Pancreatic Duct Stenting Impacts Treatment Outcomes in Patients Undergoing Endoscopic Transmural Drainage of Pancreatic Fluid Collections (PFCS) Shyam Varadarajulu, Jessica Trevino, Ashutosh Tamhane, C. Mel Wilcox Background: The role of pancreatic duct (PD) stenting in patients undergoing endoscopic transmural drainage of PFCs remains unclear. Aim: Evaluate the effect of PD stenting on treatment outcomes in patients undergoing endoscopic transmural drainage of PFCs. Methods: This study was executed by analysing data collected prospectively on all patients who underwent endoscopic transmural drainage of PFC over a 4-yr period at a tertiary care referral center. Patients with a definitive luminal compression (LC) underwent drainage by EGD and those without LC by EUS. Double pigtail stents were deployed in all patients; in addition, nasocystic catheters were deployed in those with abscess/necrosis. An ERCP was attempted whenever feasible in all patients and transpapillary pancreatic stents were placed when the duct disruption could be bridged. Treatment success was defined as improvement in symptoms and resolution of PFC on follow-up CT at 10 weeks. Results: Of 110 study patients (median age 52 yrs, 68% men) who underwent PFC drainage, 40 (36%) underwent PD stenting. Treatment was successful in 95 of 110 patients (86%). PFC types were: pseudocyst (62%), necrosis (20%), and abscess (18%). The median duration of follow-up was 6.2 months. Patients who underwent PD stenting were significantly more likely to have treatment success than those without PD stent placement (97.5% vs. 80.0%; Risk Ratio crudeZ1.48; pZ0.01). In multivariable analysis, the association between PD stenting and treatment success remained significant (Risk Ratio adjustedZ1.14; 95% CI: 1.01-1.29; pZ0.036) even after adjusting for type of pancreatitis, type and location of PFC, luminal compression at endoscopy, enteral nutrition, WBC count, and number of endoscopic interventions. There was no significant difference in rates of complications between patients who underwent and those who did not undergo PD stenting (0% vs 7.1%; pZ0.16). Conclusions: Transpapillary pancreatic duct stenting improves treatment outcomes in patients undergoing endoscopic transmural drainage of pancreatic fluid collections. T1325 Conventional Contrast and Guide Wire Technique for the Selective Cannulation of the Common Bile Duct: Effects On Post- ERCP Pancreatitis Alberto Mariani, Antonella Giussani, Cristian Vailati, Milena Di Leo, Pier A. Testoni Background: guidewire cannulation has been shown to reduce the risk of post- ERCP pancreatitis (PEP) by avoiding the opacification of the main pancreatic duct. Aim: To compare the effects of the conventional contrast ERCP and guidewire cannulation of the common bile duct on the rate of post-procedural pancreatitis. Patients and Methods: A total of 1249 consecutive patients with suspected biliary disease with an intact papilla who were referred for ERCP in the same centre were evaluated by four expert endoscopists, each of them changing the cannulation thecnique from the standard catheter (SC) to hydrophilic guidewire (GW) every three months, for a total of four years. By SC technique no more than three pancreatic opacifications were performed. Patients with chronic pancreatitis or pancreatic cancer were excluded. The diagnosis of post-ERCP pancreatitis was made in presence of typical abdominal pain persisting at 24 hours after the procedure associated with a serum amylase and/or lipase greater than three times the upper limit of normal. Results: PEP occurred in overall 60/1249 patients (4.8%): 35/678 (5.2%) in the GW group and 25/571 (4.4%) in the SC group (pZNS). The rate of PEP was significantly lower (pZ0.0001) when GW or SC where inserted into the CBD (3.5%) than (unintended) pancreatic duct (11.9%). In patients with unintended pancreatic duct cannulation, HG was associated with an higher rate of PEP (15.2%) than SC (8.4%) but was three times lower if two and three SC opacifications only were considered. Multivariate analysis demonstrated precut (ORZ3.07, p!0.02) and suspected sphincter of Oddi dysfunction (ORZ3.8, p!0.05) to be significant risk factors independently associated with PEP. Conclusions: for a selective cannulation of CBD the standard catheter and guidewire technique have a similar risk for inducing PEP. However, the guidewire could be the preferable technique when by the standard catheter the pancreatic duct had already been opacified. T1326 Efficiency of Combined EUS-FNA and Biliary or Duodenal Stenting in Patients with Non Operable Pancreatic Cancer Isabelle Trouilloud, Federico Salom, Jerome Huppertz, Benoit Terris, Luigi Mangialavori, Jean-Christophe Duchmann, Marianne Gaudric, Gilles Roseau, Philippe Rougier, Stanislas Chaussade, Frederic Prat Background and Aims: Endoscopic ultrasound-guided fine-needle aspiration (EUS- FNA) is essential in the workup of locally advanced and metastatic pancreatic cancer and can be coupled with ERCP during the same sedation/anesthesia when biliary and/or duodenal stenting are needed. This ‘‘tandem’’ procedure may be cost-effective and can reduce the length of hospital stay, but the efficiency and potential risks of this approach have not been clearly defined. The aim of this study was to investigate the efficiency and complications of EUS-FNA combined with biliary or duodenal stent placement during the same session. Methods: The cases of 97 consecutive patients referred to our center for suspected pancreatic cancer with biliary and/or digestive obstruction over a 36 months period were retrospectively studied. All patients had a locally advanced or metastatic pancreatic mass of unknown nature, were not eligible for surgery and underwent an EUS-FNA coupled with an ERCP with biliary and/or duodenal stenting within the same session. We collected histological results, clinical outcome data, complications, length of stay and timespan between endoscopic procedures and first chemotherapy treatment. Abstracts www.giejournal.org Volume 69, No. 5 : 2009 GASTROINTESTINAL ENDOSCOPY AB267