M1448 Endoscopic Ultrasound (EUS) Is Highly Effective in Establishing An Etiology in Idiopathic Pancreatitis Sarba Kundu, Jason Conway, John A. Evans, Girish Mishra Background: Despite thorough evaluations, a precise etiology for acute pancreatitis can remain elusive which unfortunately necessitates unnecessary and sometime invasive testing with untoward outcomes. EUS provides superb pancreaticobiliary imaging with an additional opportunity to determine a cause in idiopathic pancreatitis. Aim: To determine the effectiveness of EUS in establishing an etiology for idiopathic pancreatitis. Methods: 222 patients were documented to have idiopathic pancreatitis after screening 3210 endoscopic ultrasound procedure reports performed between 9/23/03 and 4/26/08. Patients with either an initial episode or recurrent episodes of pancreatitis with either a normal prior ERCP/CT/ MRCP or transabdominal ultrasound were included into the study. Exclusion criteria were as follows: history of or suspicion of chronic pancreatitis, history of any alcohol abuse, suspected pancreatic mass, suspected gallstones, history of pancreatic surgery, Billroth anatomy, post cholecystectomy syndrome and triglyceride levels greater than 500. Pathological diagnosis was confirmed when fine needle aspirates were available. Results: In ninety six (43%) patients evaluated for idiopathic pancreatitis, EUS identified a discrete etiology with some patients having multiple findings as shown in Table 1. Nearly one-half were found to have obvious gallstones, CBD stones, PD stones, or microlithiasis. Unsuspected neoplasia (cystic lesions, IPMN, neuroendocrine lesions and adenocarcinoma) was identified in approximately 12%. Conclusions: EUS identified an abnormality in nearly two-thirds of patients with normal prior imaging with the vast majority exhibiting benign biliary pathology; causality however, cannot be determined. We unearthed a surprisingly high number of neoplastic lesions which further supports the benefit of EUS in patients with idiopathic pancreatitis. Table 1. Findings
Findings Specific for cause of Pancreatitis Findings Consistent with Chronic Pancreatitis Large Stones in: Common Bile Duct (CBD) Gall Bladder (GB) Pancreatic duct Total Microlithiasis in: CBD GB Total Suspected Anatomic anomalies Suspected Pancreatic tumors
16 15 2 33
7.2 6.8 0.9 14.9
20 43 63 6 27
9 19.4 28.4 2.7 12.2
Note: Panc Anomalies & tumors are not included in specific cause of pancreatitis
M1449 Endoscopic Ultrasound Restaging of Gastric Cancer Following Neoadjuvant Therapy James J. Farrell, Fritz C. Eilber Background: There is a resurgent role for EUS in the staging of gastric cancer due to its value in selecting patients with locally advanced gastric cancer for neoadjuvant therapy (MAGIC Trial, N Engl J Med 2006 Jul 6;355 (1): 11-20). There is also an emerging interest in EUS for assessing response to gastric cancer therapy to decide upon continued treatment, change in treatment or surgery. However, the accuracy of EUS for gastric cancer restaging after neoadjuvant therapy is unknown. The aim of this study was to assess the accuracy of T and N staging with EUS of gastric cancer after neoadjuvant therapy. Methods: A contemporary cohort analysis of all patients undergoing EUS for gastric cancer at UCLA from 2004-2008 was performed. Only patients with definitive surgical T and N staging were included in the final analysis. We compared the accuracy of EUS T and N staging between patients who went straight to surgery (control group) to those who underwent neoadjuvant therapy followed by EUS restaging and then surgery (neoadjuvant group). EUS nodal staging was based solely on morphology. Fisher’s Exact test was used to compare the accuracies between both groups. Results: 73 patients with gastric cancer underwent EUS evaluation with the same Radial Echoendoscope system at our institution between 2004-2008. 34 were excluded from analysis (no surgical pathology available, endoscopic therapy). 39 patients had definitive surgical T and N staging: 27 (69%) patients went straight to surgery (control group) and 12 (31%) underwent neoadjuvant therapy (neoadjuvant group). Of these 12 patients, 9 had a post treatment EUS. When comparing the neoadjuvant group with the control group, there was no significant difference in EUS accuracy for all T stages (78% vs. 74%, pZn.s.), or just T3 or 4 stages (85% vs. 67%, pZn.s.) The accuracy EUS for N stage in the neoadjuvant group was 67% compared to 63% in the control group (pZn.s.) (Table 1). Conclusions: In this contemporary cohort analysis, the accuracy of EUS T and N staging of gastric cancer is comparable in the presence or absence of treatment. Unlike the limited accuracy for EUS restaging in esophageal cancer after neoadjuvant treatment, there may be a role for relook EUS
in patients undergoing treatment for gastric cancer to assess response to therapy and to guide further treatment. Table 1. EUS T and N Stage Accuracy based on Surgical Staging Tumor Stage T0,1,2 Neoadjuvant Group(9) Control Group(27) p value
T Lymph N Accuracy Node Stage Accuarcy T3,4 T0-4 T3,4 N0 N1 N0,1
M1450 EUS Surveillance of Suspected Lower Risk Mucinous Cystic Neoplasms of the Pancreas Tamas A. Gonda, Peter S. Francisco, Amrita Sethi, Shashin Shah, Vasudha Dhar, Charles J. Lightdale, Stavros N. Stavropoulos, Peter D. Stevens Introduction: Evolving guidelines for the management of small (!3 cm) asymptomatic cystic lesions of the pancreas suggestive of IPMN or MCN recommend periodic surveillance. The natural history of these lesions is just beginning to be described. The goal of our study was to examine change in morphology and fluid analysis (CEA, DNA quantity and quality, Kras mutations, LOHs) and determine if these parameters can predict progression. Methods: We have searched our endoscopic, radiologic and laboratory database for all patients who have undergone pancreatic cyst fluid analysis in the last 10 years and identified those that were surveyed by EUS at least 6 months after initial diagnosis. We excluded lesions for which management is non-surgical (finding of central scar/ microcystic morphology suggestive of SCAs) and high risk mucinous lesions such as main duct IPMNs, suspected MCNs R3 cm or lesions with mural nodule. Fluid molecular analysis was performed by RedPathIP (Pittsburgh, PA). Results: We identified 49 patients with suspected lower risk mucinous cysts for which surveillance is generally recommended; 31 underwent repeated cyst fluid FNA. The average surveillance interval was 21þ/-11 months. The majority of cysts did not show a notable change in size (O5 mm increase was seen in 12%), and only 2 patients developed new mural nodules. 25% and 20% of patients have shown a O50% increase or decrease in fluid CEA (however only 5/31 showed a greater then 1 log change). In linear regression model there was no association between initial size, CEA or size and CEA change. 6/9 patients with rising CEA underwent surgery and 5/6 had low grade lesions. Both patients with a mural nodule had invasive cancer. Surveillance with molecular markers was also performed in 15/28 patients. Molecular features were stable in 13/15 cases; in 5 of these a rise in CEA was seen but low grade lesions were found suggesting a superiority of molecular analysis to CEA in risk-stratifying lesions during surveillance. In 2/15 cases new high amplitude LOH or Kras mutations were found and these correlated with development of a mural nodule. Conclusions: At nearly two years follow up, the majority (88%) of suspected mucinous cystic lesions for which surveillance is recommended are stable in size. However, fluid CEA level fluctuations are frequent and in the absence of other morphologic or molecular findings suggesting progression, a rise in CEA level does not appear to reliably predict neoplastic progression. Unlike CEA, molecular markers are more stable between measurements and progression correlates with development of high risk morphologic features.
M1451 Prevalence of Features of Chronic Pancreatitis in a Dutch Cohort of Individuals At High-Risk for Developing Pancreatic Cancer Femme Harinck, Jan-Werner Poley, Irma Kluijt, Marcel G. Dijkgraaf, Hendrik M. Van Dullemen, Robin Timmer, Marco Bruno, Paul Fockens Introduction: Individuals at high-risk for pancreatic cancer (PC) are (1) mutation carriers of PC prone hereditary syndromes and (2) first-degree relatives of patients with PC who have familial PC. Previous studies revealed a high prevalence of chronic pancreatitis (CP)-like features in these high-risk individuals undergoing screening endosonography (EUS). The aim of this study was to determine the prevalence of CP features in a Dutch cohort of individuals at high-risk for developing PC entering a yearly surveillance program. Methods: Asymptomatic high-risk individuals prospectively underwent EUS. The operating endosonographer scored presence of nine validated features of CP; echogenic foci, strands, lobularity, cysts, stones, duct dilatation, duct irregularity, hyperechoic duct margins, and visible side branches. Since interpretation of EUS-images is known to be highly observer dependent, the videotaped examinations were re-evaluated and re-scored by four expert endosonographers. The interobserver agreement among the experts was assessed using interclass correlations (ICC) statistics. ICCs !0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8 and 0.8-1 were considered as poor, fair, moderate, good and very good respectively. Results: Eighteen individuals (M/F 7/11), age 41-63 years, median 55 years who had underwent baseline screening with EUS were included.
Volume 69, No. 5 : 2009 GASTROINTESTINAL ENDOSCOPY AB245
Seven individuals were from familial PC kindreds and eleven were mutation carriers of PC prone hereditary syndromes. At initial screening, the number of features of CP scored per individual ranged from 0 to 4. The operating endosonographer scored three or more features in five individuals (28%) and four features were detected in four (22%) individuals. When re-evaluated by experts, three or more features were suggested in 19.5% of cases (range 11-28%). The maximum number of items scored per individual was four and found in 5,5% of cases (range 0-11%). The ICC for the number of features scored per individual was fair (0.3). The agreement for specific features of CP varied from poor to moderate. Conclusion: Based on these preliminary results, features of CP were less commonly seen at baseline screening in this Dutch cohort of individuals at high-risk for developing PC compared to previous studies. Previous studies found three or more CP-like features in 78% of cases. In our series three or more features of CP were found at initial examination in 28% of cases and by re-evaluation of the examinations by experts in 19.5%. Further surveillance of these individuals will reveal if features of CP progress over time and may serve as a marker for early neoplastic changes.
M1452 Usefulness of Endoscopic Ultrasonography (Plain Combined with Contrast-Enhanced) in the Differentiation Between Autoimmune Pancreatitis and Pancreatic Cancer Takuya Ishikawa, Yoshiki Hirooka, Akihiro Itoh, Hiroki Kawashima, Toshifumi Kasugai, Eizaburo Ohno, Hiroshi Matsubara, Ryoji Miyahara, Yoshiaki Katano, Naoki Ohmiya, Yasumasa Niwa, Hidemi Goto Objectives: Autoimmune pancreatitis (AIP) is a unique form of chronic pancreatitis, and we need to differentiate it from pancreatic cancer (PCA) in the process of diagnosis. Endoscopic ultrasonography (EUS) is known as a useful examination in pancreato-biliary field, but reports on EUS findings with AIP are limited. The aim of this study is to investigate the usefulness of EUS (plain combined with contrastenhanced) in the differentiation between AIP and PCA. Methods: We reviewed 36 patients with AIP (male: femaleZ33: 3, mean age 62.9) based on the criteria established by Japan Pancreas Society in 2006 between July 2003 and October 2008 and 36 patients with PCA (male: femaleZ24: 12, mean age 61.6) based on pathological diagnosis between April 2005 and October 2008. All patients underwent EUS, and we investigated 6 findings as follows; location, internal echo, border with surroundings, upstream dilation of main pancreatic duct (MPD), wall thickness of bile ducts, and lymph nodes enlargement. Contrast-enhanced EUS (CE-EUS) using Perflubutane (SonazoidÔ, Daiichi-Sankyo, Japan) was performed for 16 patients with AIP and 20 patients with PCA, and enhancement effects and diagnostic capability of CE-EUS were evaluated. Enhancement effects compared with surrounding pancreatic parenchyma were observed for 1 minute continuously, and 3 minutes, 5 minutes after injection of contrast medium, and were classified into 3 patterns as follows using time intensity curve; pattern A (pA): Equal enhancement effect continued for 5 minutes, pattern B (pB): Equal enhancement effect continued for 1 minute, and decreased after 3minutes, pattern C (pC): Enhancement effect decreased within 1 minute. (Results) The locations of the lesions of AIP and PCA were, head: body and tail: whole body Z 18 (50%): 5 (13.9%): 13 (36.1%) and 25 (69.4%): 11 (30.6%): 0 (0%). All lesions were hypoechoic, and stripe or mesh hyperechoic areas were seen only in AIP (16.6% (6/ 36), 6% (2/36)). Anechoic areas were seen only in PCA (25% (9/36)). The incidence of clear borders was higher in PCA (19.4% (7/36): 83.3% (30/36)) (p!0.001). The diameter of upstream MPD was more increased in PCA (2.89: 5.06 (mm)) (pZ0.007), and the incidence of wall thicknesses of bile ducts was higher in AIP (63.6% (7/11): 11.1% (2/18)) (pZ0.01). There was no significant difference in lymph nodes enlargement. Enhancement patterns of AIP and PCA were, pA: pB: pCZ 12 (75%): 2 (12.5%): 2 (12.5%) and 0 (0%): 5 (25%): 15 (75%). Judging the pB or pC as PCA, the sensitivity, specificity, and accuracy rate of CE-EUS were 75%, 100%, and 88.9%. Conclusions: EUS combined with CE-EUS is useful for differentiating AIP and PCA.
M1453 The Combined Endoscopic Ultrasound and Secretin Endoscopic Pancreatic Function Test (EUS/PFT) Influences Diagnosis and Management of Chronic Pancreatitis Tyler Stevens, John J. Vargo, Gregory Zuccaro, John A. Dumot, Mansour A. Parsi Introduction: The combined EUS/PFT provides a sensitive evaluation of pancreatic structure and function. Aim: Determine how EUS/PFT results affect diagnosis and treatment of patients with suspected CP. Methods: Retrospective cohort study of pts who underwent EUS/PFT in the past 3 years. Radial EUS was performed (normal !4 criteria) with endoscopic collection of duodenal samples at 15, 30, and 45 min after intravenous secretin (normal peak bicarb R80 mmol/L). The medical record was reviewed to determine whether a final clinical diagnosis of CP was made, and what treatments were offered. Results: A total of 247 pts had the EUS/PFT [Figure].
AB246 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 5 : 2009
207 pts were evaluated for suspected minimal-change CP (abdominal pain, negative CT). The final diagnosis was based on clinical presentation and EUS/PFT results. 98% of those with concordant abnormal results (EUSþ/PFTþ) were given a final diagnosis of CP (ruled in). 93% of those with concordant normal results (EUS-PFT-) were told they did not have CP (ruled out). 79% of those with discordant results (EUSþ/PFT- or EUS-/PFTþ) were ruled in. A total of 71 patients (34%) were ruled in: 53 were prescribed enzymes, 12 had nerve blockade or spinal cord stimulation, and 7 had surgical resection. 136 (66%) pts were deemed not to have CP: Of these, 42 were found to have another organic cause of pain and 94 were diagnosed with a functional pain syndrome. An additional 40 pts underwent EUS/PFT for staging of established CP (CT positive): Based on EUS results, 8 were referred for endoscopic, and 7 for surgical management. 7 had a normal PFT, and enzymes were stopped. EUS/FNA revealed adenocarcinoma in 3 pts. Conclusion: The combined EUS/PFT results present complimentary information: EUS detects early parenchymal abnormalities, screens for neoplasm, and identifies duct obstruction. PFT assess exocrine function which helps in diagnosis and in the decision to give enzymes.
Figure. EUS/PFT results and clinical diagnosis.
M1454 Clinical Factors Influence Accurate Endosonographic Assessment with Miniprobe for Early Gastric Cancer Shunsuke Yamamoto, Tsutomu Nishida, Shusaku Tsutsui, Motohiko Kato, Yoshito Hayashi, Takuya Yamada, Hideharu Ogiyama, Toshiyuki Yoshio, Hideki Iijima, Masahiko Tsujii, Norio Hayashi Background and Aims: With the advent of endoscopic submucosal dissection (ESD) techniques, pretreatment diagnosis for invasive depth of early gastric cancer (EGC) has become increasingly important. Endoscopic ultrasonography (EUS) is a useful tool to determine the vertical cancer invasion depth. The aim of this study was to evaluate the clinical factors affecting the diagnostic accuracy of pretreatment EUS in EGC. Patients and Methods: A total of 56 patients (mean age 68 years; 45 male, 11 female) pretreatment diagnosed as early gastric cancer was enrolled in this study. All patients were underwent EUS examination with a 20 MHz endosonography miniprobe and treated at Osaka university hospital, Osaka, Japan, from April 2007 to July 2008. We reviewed the medical records of 56 patients and compared pretreatment EUS staging with histopathologic staging according to the clinical parameters such as 1) the diameter, 2) the location and 3) the surface pattern (elevated or depressed). Results: 46 patients were confirmed as T1 m and 10 patients as T1 sm cancer through histopathologic examination of the resected specimen. The overall diagnostic accuracy of EUS for predicting tumor invasion depth was 82% (46/56, T1 mZ43/46, T1 smZ3/10). 1) Diameter of carcinoma was significantly larger in the misdiagnosed lesions (nZ10) than in accurately diagnosed lesions (mean 31 vs 15mm, pZ0.003). 2) Misdiagnosed lesions included significantly higher rate of corpus lesions than accurately diagnosed lesions (90 vs 50%, pZ0.03). 3) There was no statistically significant difference between elevated and depressed types in the diagnostic accuracy of tumor invasion. Conclusion: EGC with large tumor size or located in the upper part of the stomach is more frequently associated with an incorrect miniprobe EUS diagnosis of tumor invasion depth.
M1455 Is EUS Necessary in the Management of Patients with High Grade Dysplasia in Flat Barrett’s Mucosa? Mihir Bakhru, Vanessa M. Shami, Vani J. Konda, Irving Waxman, Samer El-Dika Introduction: The use of Endoscopic Ultrasound (EUS) in patients with high-grade dysplasia (HGD) in the setting of Barrett’s esophagus is debatable. Even with high frequency probes, it is difficult to distinguish the subtleties of high-grade dysplasia (HGD) from intramucosal cancer (ICA) or even cancer invading the submucosa. Some patients with nodular Barrett’s and HGD are found to have malignant lymph nodes by EUS-fine needle aspiration (FNA). However, there is very little reported on the use of EUS FNA in the setting of flat Barrett’s with HGD. The aim of this study was to investigate the impact of EUSFNA in this subset of patients. Methods: A total of 22 consecutive patients with flat HGD who were referred to the University of Virginia and the University of Chicago for EUS between Jan 2005 - June 2008 were