- Email: [email protected]
Sa1570 Test Performance of EUS for Chronic Pancreatitis: Novel Meta-Analysis Using Bayesian Techniques Designed for Imperfect Reference Standards
Abstracts
18.5 mg/L), respectively.Amoxicillin cystic levels range from 0 to 3.23 mg/L (mean: 0.59), null in 10/19 patients. By comparison with blood levels, tissue penetration was 2.7 % for amoxicillin and 2,6% for clavulanate. Conclusion: Antibiotic cystic diffusion after 1g amoxicillin/200mg clavulanate infusion is very low, leading to inefficiency to decrease bacterial proliferation. These results suggests that an increase in antibiotic dose or an antibiotic class switch should be evaluated in this setting.
Sa1567 60 Consecutive Incidental Pancreatic Cysts Referred to Endoscopic Ultrasound: Cost of Evaluation and Outcome Noel M. Lee*, Rinjal Brahmbhatt, Deepak V. Gopal, Mark E. Benson, Ryan De Lee, Patrick Pfau Department of Medicine, UW School of Medicine and Public Health, University of Wisconsin, Madison, WI Background and Aims: Pancreatic cysts are often found incidentally on imaging for other purposes at increasing rates. After the original cyst is noted, many patients undergo a series of imaging and tests to evaluate the pancreatic cyst. The cost of this evaluation and outcome of these patients is not known. Methods: EUS exams performed from January 2010 to January 2011 were reviewed. 60 consecutive patients were evaluated with indication for EUS as “cyst found on imaging”. We examined how the cyst was originally found, indication for original imaging, and total number of imaging before EUS. The size and location of the cyst, EUS obtained cytology, and cyst fluid CEA were recorded. Outcome of patients after EUS was evaluated as well as the cost of workup for these incidental cysts. Cost was determined by hospital facility plus professional charges. Results: 60 patients were evaluated. 53 cysts (88.3%) were found incidentally on abdominal CT; 1.7% on chest CT; 5% MRI; 5% transabdominal ultrasound. 36 (60%) of patients were initially evaluated/imaged for pain (5 epigastric pain, 4 RUQ, 3 RLQ, 2 LUQ, 4 LLQ, and 18 with generalized pain). 8 (13.3%) were imaged for non-pancreatic cancer workup, 4 (6.7%) had nausea, vomiting, or diarrhea, and 4 (6.7%) had abnormal laboratory findings. Average cyst size was 2.36 cm. 25 cysts were found in the head, 22 body, 9 tail, and 3 peripancreatic area. 71.7% of patients had greater than one imaging test before EUS, with a mean average of 2.8 tests per patient (range 19). EUS-FNA cytology was obtained in 71.7% of patients; 8 nondiagnostic, 23 negative, 5 suspicious for mucinous neoplasm, 3 neuroendocrine tumor, 1 adenocarcinoma. 8 patients had a cyst fluid CEA ⬎192. Of the 60 patients, 11 patients eventually underwent surgery, 40 patients were followed with surveillance imaging, 6 were recommended no followup, 3 were lost to followup, and 2 had infections with antibiotic treatment. Only 16 patients (26.7%) had a definitive diagnosis by cytology or surgical pathology. The mean cost of evaluation per person for an incidental pancreatic cyst was $17,606, range $11,669- $32,457. Conclusions: 1)The evaluation of incidental pancreatic cysts leads to extensive imaging with high associated costs 2) Despite the high cost and multiple tests, only a minority of patients will have a definitive diagnosis made.
Sa1568 Diagnostic Performance of EUS-FNA for Cystic and Non Cystic Pancreatic Neuro-Endocrine Tumors: Single Center Experience From the Moffitt Cancer Center Akshay K. Gupta*, Jonathan R. Strosberg, Pamela Hodul, Larry Kvols, Shivakumar Vignesh GI Oncology, Moffitt Cancer Center, Tampa, FL Aim: Pancreatic neuroendocrine tumors (pNETs) are rare tumors with malignant potential. EUS and EUS-guided FNA (EUS-FNA) have been shown to be superior to other imaging methods in the preoperative localization and diagnosis of pNETs. The objective if this study is to describe the EUS features of nonmetastatic, cystic and non-cystic pNETs seen at our center over the last 6 years and to evaluate the performance of EUS-FNA in diagnosis of PNETs at our center. Methods: The study was approved by the IRB at the Moffitt cancer center. We sought to identify all cases of histologically confirmed, non-metastatic pNETs, which underwent EUS prior to surgical resection at our institution between Jan 2005 and Dec 2010. A prospectively maintained database of all the pNETs seen at our institution since 1999 was used as the primary source. All the clinical, endoscopic and pathologic information was abstracted from our electronic medical records. Results: Total 34 patients were included in the study, with 10 cystic and 24 solid pNETs. Mean age of the patients was 59.1 yrs (SD 12.8), with a majority female (61.8%). All of them were discovered incidentally, or due to nonspecific symptoms. All the EUS exams were performed by one of three experienced endoscopists, and the lesion was imaged in all cases. Among the cystic tumors, 50% were part cystic and part solid, and 50% were fully cystic. The cystic tumors were more commonly seen at the head/uncinate, and the solid tumors were more commonly seen in the body/tail region (p ⫽ 0.02). There was no significant difference in the size between cystic and solid tumors (mean diameter 23.4 mm Vs 21.3mm; P ⫽ 0.06). The main pancreatic duct and rest of
the pancreatic parenchyma were normal appearing in a majority of the cases (91.7% and 94.1% respectively). Fluid could be aspirated from 50% of the cystic tumors, all with a CEA level ⬍ 192 ng/ml. With surgical pathology as the gold standard, the overall sensitivity of EUS-FNA in diagnosing cystic tumors was 57.1%, and for solid tumors, 94.1% (p⫽ 0.03), there was no significant difference in the number of needle passes required, between cystic and solid tumors. Conclusions: EUS is a sensitive test for imaging PNETs. EUS FNA is much more sensitive in diagnosing solid PNETs, vs. cystic PNETs. Our results indicate that EUS-FNA may have a higher sensitivity for diagnosis of cystic PNETs than the reported sensitivity of EUS-FNA for all pancreatic cystic tumors.
Sa1569 The Usefulness of Endoscopic Ultrasonography in Acute Cholecystitis Without Choledocholithiasis on Multidetector Computed Tomography Yong Woon Shin*, Byoung Wook Bang, Jung IL Lee, Jin-Woo Lee, Kye Sook Kwon, Seok Jeong, Don Haeng Lee, Hyung Gil Kim, Young Soo Kim Inha University School of Medicine, Incheon, Republic of Korea Background: Choledocholithiasis commonly occurs in patients with symptomatic cholelithiasis. Although recent technical innovation of multidetector computed tomography (MDCT) scan enhances diagnostic yield of choledocholithiasis, it is considered to have some limitation in evaluation of common bile duct (CBD). Aims: The purpose of this study was to evaluate the usefulness of endoscopic ultrasonography (EUS) in detection of choledocholithiasis who was diagnosed as acute cholecystitis without choledocholithiasis on MDCT. Methods: 334 patients with acute cholecystitis and no evidence of CBD stone on MDCT underwent EUS between March 2006 and April 2011. If CBD stone was suspected on the basis of EUS results or clinical symptoms, final diagnosis was obtained by endoscopic retrograde cholangiopancreatography (ERCP). Patients’ medical records were retrospectively analyzed for clinical symptoms, biochemical data, and results of imaging studies. Results: MDCT did not detect CBD stone in 42 (12.6%) patients among 334 patients with acute cholecystitis. The causes for these discrepancies could be contemplated as their small size (n⫽20, 47.6%), isodensity (n⫽18, 42.8%), impacted stone (n⫽1, 2.4%), and misdiagnosis (n⫽3, 7.1%). 37 (11.1%) patients were detected CBD stones by EUS while MDCT failed to notice the existence. With EUS used as a triage tool, unnecessary diagnostic ERCP and its complications could be spared in 255 (76.3%) patients. Intraoperative cholangiography (IOC) was performed in 157 patients. However, IOC did not detect any CBD stone. Conclusions: It seemed that MDCT might not take a major role in detecting CBD stone and EUS should be performed as an add-on test in patients with acute cholecystitis for CBD evaluation. Especially, EUS is routinely recommended in patients with abnormal liver enzymes, pancreatitis, and dilated CBD. IOC is not necessary if CBD evaluation was performed by EUS preoperatively.
Sa1570 Test Performance of EUS for Chronic Pancreatitis: Novel MetaAnalysis Using Bayesian Techniques Designed for Imperfect Reference Standards Joseph Romagnuolo*1, Nandini Dendukuri2, Ian Schiller2, Lawrence Joseph2 1 Medical University of South Carolina, Charleston, SC; 2McGill University, Montreal, QC, Canada Background: EUS is one of the most accurate tests for diagnosing “minimal change” chronic pancreatitis (MCCP) based on multiple studies comparing to reference standards (RS). However, a lingering impediment in providing summary estimates of EUS test performance using traditional meta-analytic techniques has been that the RS varies from study to study, and RS’s for MCCP are imperfect themselves. We aimed to use Bayesian methods to correct for imperfections in the RS and summarize the studies. Methods: Abstracts from a PubMed search (1966-2011, “EUS” and “pancreatitis”) were manually reviewed for comparisons of EUS accuracy to RS. Eligible manuscripts with extractable 2x2 data of EUS in suspected MCCP were then retrieved and abstracted. Quality and risk of bias were graded with published (QUADAS) scores. Cutoffs of ⱖ3 and ⱖ4 (of 9) criteria were both considered, given results of prior receiver operator characteristics (ROC) curve studies. Prior information on known/perceived accuracy of the existing RS were elicited from experts. A bivariate meta-analysis model was used to pool the sensitivity and specificity estimates across studies, adjusting for the imperfect nature of the model and RS. We modeled “prior” information on RS accuracy as well as modeled with “non-informative” priors, and compared the models with the conventional approach (treating RS as “perfect”). Pooled sensitivities/specificities (95% Bayesian credible intervals (CI)) and summary ROC curves were obtained from each model. In lay terms, the model considers that some discrepancies between EUS and RS could be due to RS being wrong -i.e. “false-positive” EUS could really be “false-negative” RS. Stratified analyses explored impact of prevalence, study era/year of study, and
AB204 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4S : 2012
www.giejournal.org
Abstracts
radial/linear on between-study heterogeneity. Results: 730 abstracts were manually reviewed; 14 met criteria and 2x2 data were abstracted. 5 RS were used alone or in combination: ERCP, function tests, radiologic/clinical follow-up, and histology. As seen in Fig 1, adjusting for the imperfect RS yielded higher test performance characteristics for EUS (solid lines) than if perfection in RS was assumed (dashed lines). Similar results were obtained with “non-informative prior”. Median pooled sensitivity & specificity were 87.1% (95%CI: 67-97%) & 92.9% (95%CI: 71-99%) for the ⱖ3 cutoff. Adjusted RS sensitivities’ & specificities’ ranges were 74-95% & 85-95%, for MCCP; histology was most accurate (95%, 93%), and function tests least sensitive but specific (74%, 93%). Conclusions: EUS diagnostic performance in MCCP can be summarized from existing literature using novel Bayesian methods to correct for imperfections in the array of RS used to date. EUS accuracy is high, especially in comparison to estimated “true” test performance of the RS to which it is being compared.
Sa1572 Association of EUS Features With PanIN Lesions in Chronic Pancreatitis (CP): A Single Center Experience Julia K. Leblanc*, Jey-Hsin Chen, Leticia P. Luz, Mohammad A. AL-Haddad, Stuart Sherman, Lee Mchenry, John M. Dewitt Indiana University, Indianapolis, IN Introduction: Pancreatic intraepithelial neoplasia (PanIN) lesions and chronic pancreatitis (CP) are both considered histologic precursors to pancreatic adenocarcinoma. The frequency of PanIN lesions in chronic pancreatitis unknown. Objective: To evaluate the association between PanIN lesions and EUS abnormalities in patients with CP. Hypothesis: We hypothesized that increasing PanIN grade would be associated with an increased number of EUS criteria for CP. Methods: EUS and pathology databases at our hospital from 2000 to 2009 were queried for patients with a diagnosis of chronic pancreatitis who underwent EUS and any pancreatic resection within one year. EUS parenchymal (hyperechoic foci, hyperechoic stranding, lobularity ⫾ honeycombing, cysts, calcification/stones) and ductal criteria (irregularity of the main pancreatic duct, dilated side branches, hyperechoic duct margins, PD dilation, PD stones) from the body and tail of the pancreas were abstracted. The relationship between PanIN grade (1-4) and EUS criteria (total, parenchymal, and ductal) for CP was evaluated using chi-square test. Results: 127 of 150 (85%) patients who had EUS for CP and a pancreatic resection had a PanIN lesion on surgical pathology. 91 of 127 (71%) patients (48 men, mean age 54) had surgery a mean of 91.7 (⫾89.3) days after EUS. Pancreatic resection involved the head in 63 (69%) and distal pancreas in 28 (31%). PanIN 1, 2, 3, and 4 lesions were seen in: 76 (84%), 13 (14%), 2 (2%), and 0% respectively. PanIN lesions were seen in 63 (69%) of pancreas head and 28 (31%) of distal pancreatic resections. PanIN 1 lesions were more often associated with the pancreas head resections (p⫽0.026). PanIN 2 and 3 lesions were not associated with site of pancreas resected (p⫽0.76). Overall, the mean (SD) number of total EUS criteria for CP, parenchymal criteria, and ductal criteria was: 4.9 (⫾2.7), 3.7 (⫾1.8), and 1.4 (⫾1.4) respectively. The most frequent parenchymal and ductal criteria were: hyperechoic stranding in 72 (79%), any lobularity in 71 (78%), dilation in 41 (45%), and hyperechoic duct margins in 22 (24%). Number of EUS criteria by PanIN grade is shown in Table 1. There was no association between the degree of PanIN lesion and number of: ductal (p⫽0.73), parenchymal (p⫽0.73), or total criteria for CP (p⫽0.82). Parenchymal lobularity with honeycombing was associated with higher PanIN grades (p⫽.02) Conclusion: PanIN lesions were frequently found in resected pancreas specimens from patients with EUS features of CP. While the PanIN grade was not predicted by number of EUS criteria for CP, the number of advanced PanIN lesions is small. Therefore, there is a potential association between the degree of PanIN lesions and number of EUS criteria for CP. Further study is needed.
Figure 1.
Sa1571 Diagnosis of Focal Pancreatic Masses by Quantitative Low Mechanical Index Contrast-Enhanced Endoscopic Ultrasound (EUS) Dan Ionut Gheonea*1,2, Costin T. Streba1, Ana Maria Ioncica1, Tudorel Ciurea1, Adrian Saftoiu1 1 Gastroenterology, University of Medicine and Pharmacy, Craiova, Romania; 2University of Medicine and Pharmacy Carol Davila, Bucharest, Romania
Number of EUS criteria by PanIN grade
Introduction: Second-generation intravenous blood-pool ultrasound contrast agents are increasingly used in endoscopic ultrasound (EUS) for characterization of microvascularization, differential diagnosis of benign and malignant focal lesions, as well as improved staging and guidance of therapeutic procedures. Aims And Method: The aim of our study was to prospectively compare the vascularisation patterns in chronic pseudotumoral pancreatitis and pancreatic cancer using quantitative low mechanical index (MI) contrast-enhanced EUS. We included 51 patients with chronic pseudotumoral pancreatitis (n⫽19) and pancreatic cancer (n⫽32). Perfusion imaging started with a bolus injection of Sonovue (2.4 ml), followed by analysis in the early arterial (wash-in) and late venous (wash-out) phase. Perfusion analysis was performed by post-processing of the raw data (time intensity curve [TIC] analysis). TIC analysis was performed inside the tumor and the pancreatic parenchyma, with depiction of the dynamic vascular pattern generated by specific software. Statistical analysis was performed on raw data extracted from the TIC analysis. Final diagnosis was based on a combination of EUS-FNA, surgery and follow-up of minimum 6 months in negative cases. Results: Pseudotumoral chronic pancreatitis showed in the majority of cases a hypervascular appearance in the early arterial phase of contrast-enhancement, with a dynamic enhancement pattern similar with the rest of the parenchyma. Statistical analysis of the resulting series of individual intensities revealed no statistically relevant differences (p⫽0.78). Pancreatic adenocarcinoma was usually a hypovascular lesion, showing low contrast-
www.giejournal.org
enhancement during the early arterial and also during the late venous phase of contrast-enhancement, also lower than the normal surrounding parenchyma. We found statistically significant differences in values during TIC analysis (p⬍0.001). Conclusion: Low MI contrast enhanced EUS technique is expected to improve the differential diagnosis of focal pancreatic lesions. However, further multicentric randomized studies will confirm the exact role of the technique and its place in imaging assessment of focal pancreatic lesions.
Mean (SD) EUS criteria for CP
PanIN (nⴝ91)
Total
Parenchymal
Ductal
1 (n⫽76) 2 (n⫽13) 3 (n⫽2)
4.8 (2.9) 5.5 (1.6) 7.5 (2.1)
3.5 (1.9) 4.3 (1.1) 5.5 (0.7)
1.4 (1.4) 1.5 (1.3) 2.0 (1.4)
Sa1573 Yield of Repeat Endoscopic Ultrasound With Fine Needle Aspiration in Suspicious Pancreatic Lesions Brian R. Boulay*, Umayer Ali, Brett Sleesman, Allan Halline Digestive Diseases & Nutrition, University of Illinois at Chicago, Chicago, IL Background: The management of patients with pancreatic lesions and atypical cells or hypocellular specimens following endoscopic ultrasound (EUS) guided fine needle aspiration (FNA) is controversial. There is little information regarding the yield of repeat FNA in these lesions. Aim: Determine the yield of positive cytology in patients undergoing repeat EUS-guided FNA for solid or cystic pancreatic lesions, as well as possible factors contributing to positive cytology. Methods: Retrospective electronic chart review of patients referred to a single tertiary care center for endoscopic ultrasound over a 4.5 year period. Patient charts were abstracted for baseline patient characteristics, findings on EUS exam,
Volume 75, No. 4S : 2012
GASTROINTESTINAL ENDOSCOPY
AB205
18.5 mg/L), respectively.Amoxicillin cystic levels range from 0 to 3.23 mg/L (mean: 0.59), null in 10/19 patients. By comparison with blood levels, tissue penetration was 2.7 % for amoxicillin and 2,6% for clavulanate. Conclusion: Antibiotic cystic diffusion after 1g amoxicillin/200mg clavulanate infusion is very low, leading to inefficiency to decrease bacterial proliferation. These results suggests that an increase in antibiotic dose or an antibiotic class switch should be evaluated in this setting.
Sa1567 60 Consecutive Incidental Pancreatic Cysts Referred to Endoscopic Ultrasound: Cost of Evaluation and Outcome Noel M. Lee*, Rinjal Brahmbhatt, Deepak V. Gopal, Mark E. Benson, Ryan De Lee, Patrick Pfau Department of Medicine, UW School of Medicine and Public Health, University of Wisconsin, Madison, WI Background and Aims: Pancreatic cysts are often found incidentally on imaging for other purposes at increasing rates. After the original cyst is noted, many patients undergo a series of imaging and tests to evaluate the pancreatic cyst. The cost of this evaluation and outcome of these patients is not known. Methods: EUS exams performed from January 2010 to January 2011 were reviewed. 60 consecutive patients were evaluated with indication for EUS as “cyst found on imaging”. We examined how the cyst was originally found, indication for original imaging, and total number of imaging before EUS. The size and location of the cyst, EUS obtained cytology, and cyst fluid CEA were recorded. Outcome of patients after EUS was evaluated as well as the cost of workup for these incidental cysts. Cost was determined by hospital facility plus professional charges. Results: 60 patients were evaluated. 53 cysts (88.3%) were found incidentally on abdominal CT; 1.7% on chest CT; 5% MRI; 5% transabdominal ultrasound. 36 (60%) of patients were initially evaluated/imaged for pain (5 epigastric pain, 4 RUQ, 3 RLQ, 2 LUQ, 4 LLQ, and 18 with generalized pain). 8 (13.3%) were imaged for non-pancreatic cancer workup, 4 (6.7%) had nausea, vomiting, or diarrhea, and 4 (6.7%) had abnormal laboratory findings. Average cyst size was 2.36 cm. 25 cysts were found in the head, 22 body, 9 tail, and 3 peripancreatic area. 71.7% of patients had greater than one imaging test before EUS, with a mean average of 2.8 tests per patient (range 19). EUS-FNA cytology was obtained in 71.7% of patients; 8 nondiagnostic, 23 negative, 5 suspicious for mucinous neoplasm, 3 neuroendocrine tumor, 1 adenocarcinoma. 8 patients had a cyst fluid CEA ⬎192. Of the 60 patients, 11 patients eventually underwent surgery, 40 patients were followed with surveillance imaging, 6 were recommended no followup, 3 were lost to followup, and 2 had infections with antibiotic treatment. Only 16 patients (26.7%) had a definitive diagnosis by cytology or surgical pathology. The mean cost of evaluation per person for an incidental pancreatic cyst was $17,606, range $11,669- $32,457. Conclusions: 1)The evaluation of incidental pancreatic cysts leads to extensive imaging with high associated costs 2) Despite the high cost and multiple tests, only a minority of patients will have a definitive diagnosis made.
Sa1568 Diagnostic Performance of EUS-FNA for Cystic and Non Cystic Pancreatic Neuro-Endocrine Tumors: Single Center Experience From the Moffitt Cancer Center Akshay K. Gupta*, Jonathan R. Strosberg, Pamela Hodul, Larry Kvols, Shivakumar Vignesh GI Oncology, Moffitt Cancer Center, Tampa, FL Aim: Pancreatic neuroendocrine tumors (pNETs) are rare tumors with malignant potential. EUS and EUS-guided FNA (EUS-FNA) have been shown to be superior to other imaging methods in the preoperative localization and diagnosis of pNETs. The objective if this study is to describe the EUS features of nonmetastatic, cystic and non-cystic pNETs seen at our center over the last 6 years and to evaluate the performance of EUS-FNA in diagnosis of PNETs at our center. Methods: The study was approved by the IRB at the Moffitt cancer center. We sought to identify all cases of histologically confirmed, non-metastatic pNETs, which underwent EUS prior to surgical resection at our institution between Jan 2005 and Dec 2010. A prospectively maintained database of all the pNETs seen at our institution since 1999 was used as the primary source. All the clinical, endoscopic and pathologic information was abstracted from our electronic medical records. Results: Total 34 patients were included in the study, with 10 cystic and 24 solid pNETs. Mean age of the patients was 59.1 yrs (SD 12.8), with a majority female (61.8%). All of them were discovered incidentally, or due to nonspecific symptoms. All the EUS exams were performed by one of three experienced endoscopists, and the lesion was imaged in all cases. Among the cystic tumors, 50% were part cystic and part solid, and 50% were fully cystic. The cystic tumors were more commonly seen at the head/uncinate, and the solid tumors were more commonly seen in the body/tail region (p ⫽ 0.02). There was no significant difference in the size between cystic and solid tumors (mean diameter 23.4 mm Vs 21.3mm; P ⫽ 0.06). The main pancreatic duct and rest of
the pancreatic parenchyma were normal appearing in a majority of the cases (91.7% and 94.1% respectively). Fluid could be aspirated from 50% of the cystic tumors, all with a CEA level ⬍ 192 ng/ml. With surgical pathology as the gold standard, the overall sensitivity of EUS-FNA in diagnosing cystic tumors was 57.1%, and for solid tumors, 94.1% (p⫽ 0.03), there was no significant difference in the number of needle passes required, between cystic and solid tumors. Conclusions: EUS is a sensitive test for imaging PNETs. EUS FNA is much more sensitive in diagnosing solid PNETs, vs. cystic PNETs. Our results indicate that EUS-FNA may have a higher sensitivity for diagnosis of cystic PNETs than the reported sensitivity of EUS-FNA for all pancreatic cystic tumors.
Sa1569 The Usefulness of Endoscopic Ultrasonography in Acute Cholecystitis Without Choledocholithiasis on Multidetector Computed Tomography Yong Woon Shin*, Byoung Wook Bang, Jung IL Lee, Jin-Woo Lee, Kye Sook Kwon, Seok Jeong, Don Haeng Lee, Hyung Gil Kim, Young Soo Kim Inha University School of Medicine, Incheon, Republic of Korea Background: Choledocholithiasis commonly occurs in patients with symptomatic cholelithiasis. Although recent technical innovation of multidetector computed tomography (MDCT) scan enhances diagnostic yield of choledocholithiasis, it is considered to have some limitation in evaluation of common bile duct (CBD). Aims: The purpose of this study was to evaluate the usefulness of endoscopic ultrasonography (EUS) in detection of choledocholithiasis who was diagnosed as acute cholecystitis without choledocholithiasis on MDCT. Methods: 334 patients with acute cholecystitis and no evidence of CBD stone on MDCT underwent EUS between March 2006 and April 2011. If CBD stone was suspected on the basis of EUS results or clinical symptoms, final diagnosis was obtained by endoscopic retrograde cholangiopancreatography (ERCP). Patients’ medical records were retrospectively analyzed for clinical symptoms, biochemical data, and results of imaging studies. Results: MDCT did not detect CBD stone in 42 (12.6%) patients among 334 patients with acute cholecystitis. The causes for these discrepancies could be contemplated as their small size (n⫽20, 47.6%), isodensity (n⫽18, 42.8%), impacted stone (n⫽1, 2.4%), and misdiagnosis (n⫽3, 7.1%). 37 (11.1%) patients were detected CBD stones by EUS while MDCT failed to notice the existence. With EUS used as a triage tool, unnecessary diagnostic ERCP and its complications could be spared in 255 (76.3%) patients. Intraoperative cholangiography (IOC) was performed in 157 patients. However, IOC did not detect any CBD stone. Conclusions: It seemed that MDCT might not take a major role in detecting CBD stone and EUS should be performed as an add-on test in patients with acute cholecystitis for CBD evaluation. Especially, EUS is routinely recommended in patients with abnormal liver enzymes, pancreatitis, and dilated CBD. IOC is not necessary if CBD evaluation was performed by EUS preoperatively.
Sa1570 Test Performance of EUS for Chronic Pancreatitis: Novel MetaAnalysis Using Bayesian Techniques Designed for Imperfect Reference Standards Joseph Romagnuolo*1, Nandini Dendukuri2, Ian Schiller2, Lawrence Joseph2 1 Medical University of South Carolina, Charleston, SC; 2McGill University, Montreal, QC, Canada Background: EUS is one of the most accurate tests for diagnosing “minimal change” chronic pancreatitis (MCCP) based on multiple studies comparing to reference standards (RS). However, a lingering impediment in providing summary estimates of EUS test performance using traditional meta-analytic techniques has been that the RS varies from study to study, and RS’s for MCCP are imperfect themselves. We aimed to use Bayesian methods to correct for imperfections in the RS and summarize the studies. Methods: Abstracts from a PubMed search (1966-2011, “EUS” and “pancreatitis”) were manually reviewed for comparisons of EUS accuracy to RS. Eligible manuscripts with extractable 2x2 data of EUS in suspected MCCP were then retrieved and abstracted. Quality and risk of bias were graded with published (QUADAS) scores. Cutoffs of ⱖ3 and ⱖ4 (of 9) criteria were both considered, given results of prior receiver operator characteristics (ROC) curve studies. Prior information on known/perceived accuracy of the existing RS were elicited from experts. A bivariate meta-analysis model was used to pool the sensitivity and specificity estimates across studies, adjusting for the imperfect nature of the model and RS. We modeled “prior” information on RS accuracy as well as modeled with “non-informative” priors, and compared the models with the conventional approach (treating RS as “perfect”). Pooled sensitivities/specificities (95% Bayesian credible intervals (CI)) and summary ROC curves were obtained from each model. In lay terms, the model considers that some discrepancies between EUS and RS could be due to RS being wrong -i.e. “false-positive” EUS could really be “false-negative” RS. Stratified analyses explored impact of prevalence, study era/year of study, and
AB204 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4S : 2012
www.giejournal.org
Abstracts
radial/linear on between-study heterogeneity. Results: 730 abstracts were manually reviewed; 14 met criteria and 2x2 data were abstracted. 5 RS were used alone or in combination: ERCP, function tests, radiologic/clinical follow-up, and histology. As seen in Fig 1, adjusting for the imperfect RS yielded higher test performance characteristics for EUS (solid lines) than if perfection in RS was assumed (dashed lines). Similar results were obtained with “non-informative prior”. Median pooled sensitivity & specificity were 87.1% (95%CI: 67-97%) & 92.9% (95%CI: 71-99%) for the ⱖ3 cutoff. Adjusted RS sensitivities’ & specificities’ ranges were 74-95% & 85-95%, for MCCP; histology was most accurate (95%, 93%), and function tests least sensitive but specific (74%, 93%). Conclusions: EUS diagnostic performance in MCCP can be summarized from existing literature using novel Bayesian methods to correct for imperfections in the array of RS used to date. EUS accuracy is high, especially in comparison to estimated “true” test performance of the RS to which it is being compared.
Sa1572 Association of EUS Features With PanIN Lesions in Chronic Pancreatitis (CP): A Single Center Experience Julia K. Leblanc*, Jey-Hsin Chen, Leticia P. Luz, Mohammad A. AL-Haddad, Stuart Sherman, Lee Mchenry, John M. Dewitt Indiana University, Indianapolis, IN Introduction: Pancreatic intraepithelial neoplasia (PanIN) lesions and chronic pancreatitis (CP) are both considered histologic precursors to pancreatic adenocarcinoma. The frequency of PanIN lesions in chronic pancreatitis unknown. Objective: To evaluate the association between PanIN lesions and EUS abnormalities in patients with CP. Hypothesis: We hypothesized that increasing PanIN grade would be associated with an increased number of EUS criteria for CP. Methods: EUS and pathology databases at our hospital from 2000 to 2009 were queried for patients with a diagnosis of chronic pancreatitis who underwent EUS and any pancreatic resection within one year. EUS parenchymal (hyperechoic foci, hyperechoic stranding, lobularity ⫾ honeycombing, cysts, calcification/stones) and ductal criteria (irregularity of the main pancreatic duct, dilated side branches, hyperechoic duct margins, PD dilation, PD stones) from the body and tail of the pancreas were abstracted. The relationship between PanIN grade (1-4) and EUS criteria (total, parenchymal, and ductal) for CP was evaluated using chi-square test. Results: 127 of 150 (85%) patients who had EUS for CP and a pancreatic resection had a PanIN lesion on surgical pathology. 91 of 127 (71%) patients (48 men, mean age 54) had surgery a mean of 91.7 (⫾89.3) days after EUS. Pancreatic resection involved the head in 63 (69%) and distal pancreas in 28 (31%). PanIN 1, 2, 3, and 4 lesions were seen in: 76 (84%), 13 (14%), 2 (2%), and 0% respectively. PanIN lesions were seen in 63 (69%) of pancreas head and 28 (31%) of distal pancreatic resections. PanIN 1 lesions were more often associated with the pancreas head resections (p⫽0.026). PanIN 2 and 3 lesions were not associated with site of pancreas resected (p⫽0.76). Overall, the mean (SD) number of total EUS criteria for CP, parenchymal criteria, and ductal criteria was: 4.9 (⫾2.7), 3.7 (⫾1.8), and 1.4 (⫾1.4) respectively. The most frequent parenchymal and ductal criteria were: hyperechoic stranding in 72 (79%), any lobularity in 71 (78%), dilation in 41 (45%), and hyperechoic duct margins in 22 (24%). Number of EUS criteria by PanIN grade is shown in Table 1. There was no association between the degree of PanIN lesion and number of: ductal (p⫽0.73), parenchymal (p⫽0.73), or total criteria for CP (p⫽0.82). Parenchymal lobularity with honeycombing was associated with higher PanIN grades (p⫽.02) Conclusion: PanIN lesions were frequently found in resected pancreas specimens from patients with EUS features of CP. While the PanIN grade was not predicted by number of EUS criteria for CP, the number of advanced PanIN lesions is small. Therefore, there is a potential association between the degree of PanIN lesions and number of EUS criteria for CP. Further study is needed.
Figure 1.
Sa1571 Diagnosis of Focal Pancreatic Masses by Quantitative Low Mechanical Index Contrast-Enhanced Endoscopic Ultrasound (EUS) Dan Ionut Gheonea*1,2, Costin T. Streba1, Ana Maria Ioncica1, Tudorel Ciurea1, Adrian Saftoiu1 1 Gastroenterology, University of Medicine and Pharmacy, Craiova, Romania; 2University of Medicine and Pharmacy Carol Davila, Bucharest, Romania
Number of EUS criteria by PanIN grade
Introduction: Second-generation intravenous blood-pool ultrasound contrast agents are increasingly used in endoscopic ultrasound (EUS) for characterization of microvascularization, differential diagnosis of benign and malignant focal lesions, as well as improved staging and guidance of therapeutic procedures. Aims And Method: The aim of our study was to prospectively compare the vascularisation patterns in chronic pseudotumoral pancreatitis and pancreatic cancer using quantitative low mechanical index (MI) contrast-enhanced EUS. We included 51 patients with chronic pseudotumoral pancreatitis (n⫽19) and pancreatic cancer (n⫽32). Perfusion imaging started with a bolus injection of Sonovue (2.4 ml), followed by analysis in the early arterial (wash-in) and late venous (wash-out) phase. Perfusion analysis was performed by post-processing of the raw data (time intensity curve [TIC] analysis). TIC analysis was performed inside the tumor and the pancreatic parenchyma, with depiction of the dynamic vascular pattern generated by specific software. Statistical analysis was performed on raw data extracted from the TIC analysis. Final diagnosis was based on a combination of EUS-FNA, surgery and follow-up of minimum 6 months in negative cases. Results: Pseudotumoral chronic pancreatitis showed in the majority of cases a hypervascular appearance in the early arterial phase of contrast-enhancement, with a dynamic enhancement pattern similar with the rest of the parenchyma. Statistical analysis of the resulting series of individual intensities revealed no statistically relevant differences (p⫽0.78). Pancreatic adenocarcinoma was usually a hypovascular lesion, showing low contrast-
www.giejournal.org
enhancement during the early arterial and also during the late venous phase of contrast-enhancement, also lower than the normal surrounding parenchyma. We found statistically significant differences in values during TIC analysis (p⬍0.001). Conclusion: Low MI contrast enhanced EUS technique is expected to improve the differential diagnosis of focal pancreatic lesions. However, further multicentric randomized studies will confirm the exact role of the technique and its place in imaging assessment of focal pancreatic lesions.
Mean (SD) EUS criteria for CP
PanIN (nⴝ91)
Total
Parenchymal
Ductal
1 (n⫽76) 2 (n⫽13) 3 (n⫽2)
4.8 (2.9) 5.5 (1.6) 7.5 (2.1)
3.5 (1.9) 4.3 (1.1) 5.5 (0.7)
1.4 (1.4) 1.5 (1.3) 2.0 (1.4)
Sa1573 Yield of Repeat Endoscopic Ultrasound With Fine Needle Aspiration in Suspicious Pancreatic Lesions Brian R. Boulay*, Umayer Ali, Brett Sleesman, Allan Halline Digestive Diseases & Nutrition, University of Illinois at Chicago, Chicago, IL Background: The management of patients with pancreatic lesions and atypical cells or hypocellular specimens following endoscopic ultrasound (EUS) guided fine needle aspiration (FNA) is controversial. There is little information regarding the yield of repeat FNA in these lesions. Aim: Determine the yield of positive cytology in patients undergoing repeat EUS-guided FNA for solid or cystic pancreatic lesions, as well as possible factors contributing to positive cytology. Methods: Retrospective electronic chart review of patients referred to a single tertiary care center for endoscopic ultrasound over a 4.5 year period. Patient charts were abstracted for baseline patient characteristics, findings on EUS exam,
Volume 75, No. 4S : 2012
GASTROINTESTINAL ENDOSCOPY
AB205