T1447: Mucus Nodule of the Endoscopic Ultrasonographic Finding Is One of the Most Important Diagnostic Factors of the Branch Duct Type Intraductal Papillary Mucinous Neoplasms of the Pancreas

Abstracts T1444 Assessing Predictive Factors of Disagreement Between Endoscopic Ultrasound (EUS) and Endoscopic Cholangiography (ERC) for the Diagnosis of Common Bile Duct Stones: Is an ERC Always Needed? Jose Larino-Noia, Rocio Ferreiro, Julio Iglesias-Garcia, Patricia Garcia-Gonzalez, Pablo Herrera-Martin, Alvaro Perez-Gonzalez, Enrique Dominguez-Munoz Endoscopic Ultrasound (EUS) is highly accurate for the diagnosis of common bile duct (CBD) stones. A normal ERC is however not an unusual finding in patients previously diagnosed with CDB stones at EUS. Taking into account that these patients do not benefit from an ERC, we performed the present study aiming at defining potential predictive factors for discordance between EUS and ERC in this setting. METHODS: We performed a retrospective analysis of a prospectively collected database. Patients diagnosed with CDB stones at EUS, who later underwent an ERC for stone extraction, over a two-year period were included. EUS was performed under conscious sedation by two endosonographers using both linear or radial Pentax EUS, and the Hitachi EUB 900. ERC was performed under conscious sedation by an expert endoscopist using the Pentax duodenoscope. Age, gender, indication for EUS, presence of cholestasis, previous cholecystectomy, dilation of the CBD at EUS, size and number of CBD stones at EUS, and time (days) between EUS and ERC were recorded as potential independent variables associated with a normal ERC in these patients. Analysis was performed by stepwise multivariate logistic regression. RESULTS: 106 patients (57 female, mean age 71 years, range 21-83) with CBD stones at EUS were included. Time between EUS and ERC was of 3 days as median (range 1-81 days). ERC found no CBD stone in 36 cases (34%). A stone size ⬍5mm at EUS was the only variable associated with a normal ERC in the multivariate analysis (RR 0.290; 95%CI: 0.124-0.672). CONCLUSION: CBD stones smaller than 5 mms as detected by EUS tend to be expelled spontaneously. An ERC is therefore not always required in these cases.

T1445 Safety and Feasibility of Direct Pancreatic Cystoscopy Using EUS-Guided Through-the-Needle Fiber Optic Probe S. Ian Gan Introduction: Pancreatic cystic lesions (PCLs) are an increasingly frequently encountered diagnostic dilemma that are known potential precursors of pancreatic cancer. Despite the use of EUS and cyst fluid analysis, there is currently no non-surgical means of reliably establishing the cyst subtype, the risk of malignant transformation and the need for surgical intervention. Direct visualization of the cyst lining may provide information diagnostic of the cyst subtype.Methods: All patients referred for evaluation of PCLs were considered eligible. PCLs under 1.5cm in size were excluded. All patients underwent EUS and the PCL was assessed. A 19 gauge needle was advanced into the PCL and fluid was suctioned for analysis. A small amount of saline was injected into the cyst and a fiber optic probe (Spyglass, Boston Scientific) was advanced through the needle into the cyst. The cyst wall was visualized where possible. Patients were monitored for complications for 1hr, 24hrs and 1 week after the procedure. Results: Ten patients were recruited. Average age was 62.9 years. Six of the patients (60%) were women. Seven of the PCLs were in the body of the pancreas (70%) while 2 were in the head (20%) and one was in the tail (10%). Mean size of the PCLs was 2.35cm (range 1.5 to 4.4cm). The fiber optic probe could be advanced into the cyst in 9 of the 10 patients. In one patient, scope position was lost after cyst decompression. Visualization of the cyst wall was achieved in 6 patients (60%). In 3 patients, visualization could not be achieved due to cloudy or bloody fluid in the cyst. In all 6 patients with cyst wall visualization, the lining had the appearance of normal mucosa. Whitish debris could be seen in three patients. Only one complication occurred, with one patient complaining of abdominal pain for 4-5days after the procedure despite unremarkable bloodwork and repeat imaging. Conclusions: Direct pancreatic cystoscopy with EUS-guided probes appears safe and feasible in this preliminary study. The utility of direct pancreatic cystoscopy in the differentiation and diagnosis of PCLs remains in question.

T1446 Pancreatic Endocrine Tumors: Preoperative Localization and Differentiation of Malignant Versus Benign by EUS Combined With Contrast-Enhancement Takuya Ishikawa, Yoshiki Hirooka, Akihiro Itoh, Hiroki Kawashima, Eizaburo Ohno, Hiroshi Matsubara, Yuya Itoh, Yosuke Nakamura, Masanao Nakamura, Ryoji Miyahara, Naoki Ohmiya, Hidemi Goto Background and Aims: Pancreatic endocrine tumors (PETs) develop in relatively few patients, but it is often difficult to diagnose them correctly because of their small size and various clinical symptoms. The aim of this study was to investigate the usefulness of endoscopic ultrasonography (EUS) combined with

www.giejournal.org

contrast-enhancement in the preoperative localization of PETs and differentiation between malignant and benign PETs. Subjects and Methods: We retrospectively reviewed 63 pathologically certified PETs of 42 patients who underwent EUS, multiphasic multidetector computed tomography (MDCT), and trans-abdominal ultrasonography (US) at our institute since January 2001. We compared EUS, MDCT and US in the preoperative identification of PETs, and the characteristic findings of EUS with malignancy were investigated. Forty-two lesions which underwent contrast-enhanced EUS (CE-EUS) were classified into three types (A: homogenous enhancement, B: filling defect, C: poor enhancement), and the difference between malignant and benign was examined. Lesions more severe than ‘Well differentiated endocrine tumor, Uncertain behaviour’ were considered as malignant according to the 2004 WHO classification. Results: EUS showed high sensitivity (60/63, 95.2 %) in identifying PETs compared with MDCT (50/63, 79.3%) and US (28/63, 44.4%). Univariate analysis of EUS findings to identify predictors of malignancy showed diameter (P ⬍ 0.001), ultrasonographic texture (P ⬍ 0.001), and anechoic area (cystic change) (P ⫽ 0.023) were significant. Multivariable logistic regression analysis showed that heterogenous ultrasonographic texture was the most significant factor for malignancy (OR ⫽ 40.0; 95% CI ⫽ 8.976-178.24). Forty-one out of 42 lesions which underwent CEEUS showed obvious enhancement and enhancement types were classified as follows; A : B : C ⫽ 18 (90%) : 2 (10%) : 0 (0%) of benign lesions, and A : B : C ⫽ 2 (9.1%) : 18 (81.8%) : 2 (9.1%) of malignant lesions. Most heterogenous hypoechoic areas and anechoic areas corresponded to hemorrhage or necrosis on pathologic examination. They were identified as filling defects in CE-EUS, and were more clearly recognized than conventional EUS. The sensitivity, specificity, and accuracy rate of conventional EUS for malignancy were 86.9%, 88.8%, and 88.1%, while those of CE-EUS were 90.9%, 90.0%, and 90.4%, respectively. Conclusions: EUS has higher sensitivity in preoperative localization of PETs compared with MDCT and US. The characteristics of EUS and CE-EUS findings in malignant PETs were clarified, and they will improve the diagnostic accuracy of PETs.

T1447 Mucus Nodule of the Endoscopic Ultrasonographic Finding Is One of the Most Important Diagnostic Factors of the Branch Duct Type Intraductal Papillary Mucinous Neoplasms of the Pancreas Noritoshi Kobayashi, Kensuke Kubota, Takeshi Shimamura, Seitaro Watanabe, Shingo Kato, Atsushi Nakajima Purpose: The preoperative diagnosis of intraductal papillary mucinous neoplasms(IPMN) of the pancreas must be as reliable as possible because large pancreatectomy may be necessary. But it was very difficult that especially, diagnosis of branch duct type IPMNs wasn’t satisfied to distinguish malignant from benign even with using cytological analysis of cystic fluid. Endoscopic ultrasonography(EUS) is one of the most reliable imaging modalities of branch duct type IPMNs. The purpose of this study was to evaluate the preoperative diagnostic value of EUS in differentiating benign(adenoma and borderline malignant) and malignant(carcinoma in situ and invasive carcinoma) IPMNs. Experimental Design: We retrospective identified 31 patients between 2001 and 2005 who underwent preoperative EUS for branch duct type IPMNs in our hospital. Pathological diagnosis after surgical resection was adenoma(n⫽23), borderline malignant(n⫽1), carcinoma in situ(n⫽3), and invasive carcinoma(n⫽4). We focused on the diameter of cystic dilated branch duct, main pancreatic duct, mucus nodule, and lateral spread by EUS findings. We analyzed EUS findings compared with final pathological diagnosis. Results: Cystic dilated branch duct diameter(31.2 mm vs 37.1 mm P⫽0.2667) and main pancreatic duct diameter(6.3 mm vs 6.6 mm P⫽0.2459) wasn’t significantly correlated with benign and malignant IPMNs. But mucus nodule diameter of papillary protrusions(6.0 mm vs 25.1 mm P⬍0.0001) and lateral spread diameter of mucus nodule(9.7 mm vs 25.2 mm P⬍0.0001) was significantly correlated with benign and malignant IPMNs. Conclusions: Mucus nodule diameter of EUS finding was most reliable preoperative diagnostic findings distinguished benign from malignant IPMNs.

T1448 Patients With Suspected Pancreatic Cancer: Importance of Adjunctive FNA of Pancreatic Duct Fluid, Ascites and Liver Lesions Marc F. Catalano, Nathan Landesman, Tal Hazan, Nalini Guda, Joseph E. Geenen Role of EUS in the diagnosis & staging of pts w/ suspected PanCA is well established. Accuracy of EUS FNA tissue diagnosis is 80-90%. False negative results may be operator dependent, inadequate tissue sampling & a variety of anatomical variables. Staging accuracy is of critical importance in determining treatment options (resection vs palliative). Pts w/ unsuspected liver mets & those w/ malignant ascites confirm advanced disease & poor outcome. The role of adjunctive FNA of pancreatic duct fluid, ascites & liver lesions has been

Volume 71, No. 5 : 2010 GASTROINTESTINAL ENDOSCOPY AB279