Su1398 Accuracy of EUS-Guided Fine-Needle Aspiration for the Diagnosis of Unexplained Bile Duct Strictures in a Hispanic Population

Abstract

Accuracy of EUS-Guided Fine-Needle Aspiration for the Diagnosis of Unexplained Bile Duct Strictures in a Hispanic Population Jorge L. Weber, Keyur Patel, Kenneth R. McQuaid, M. Hossein Saboorian, Guy Lindberg, Peter B. Marcus, Carlos G. Micames Gastroenterology, Hospital Bella Vista, Mayaguez, Puerto Rico; Caris Diagnostics, Dallas, TX; Gastroenterology, Duke University Medical Center, Durham, NC Introduction: Despite advances in imaging and sampling methods, identifying the etiology of biliary strictures remains a diagnostic challenge. Brush cytology obtained during ERCP has limited sensitivity, ranging from 30 to 50%. Until now, only a few studies with limited number of patients have looked at the accuracy of EUS for diagnosing the etiology of biliary strictures following non-diagnostic brush cytology. We report the largest series to date and the first specifically evaluating the role of EUS-FNA in a Hispanic population. Methods: Consecutive patients who underwent EUS for evaluation of indeterminate biliary strictures documented by ERCP following negative intraductal tissue sampling from Feb 2008 to Nov 2010 were included. Diagnostic accuracy of EUS was calculated using histopathology, malignant cells on cytology or 6 month follow up as reference standard. Results: A total of 59 cases of indeterminate biliary strictures were identified. EUS was performed by a single endosonographer using a curvilinear echoendoscope. Seven cases were excluded because of insufficient follow up. Fifty-two cases (mean ( SD) age 67.9 12.9 yrs.) (33 males; 19 females) were analyzed. FNA was performed in 49 cases where a pancreatic or thickened bile duct wall ( 3 mm) was present on EUS (mean size 29.1 12.26 mm). A median of 5 (range 2-7) passes were made. Biliary strictures were caused by malignancy in 35 cases (67%): adenocarcinoma 31, pancreatic lymphoma 2, and metastatic small cell carcinoma 2. Among the benign diagnoses, 9 had chronic pancreatitis, 1 pseudocyst, 3 choledocholithiasis, 1 autoimmune pancreatitis, and 3 idiopathic. Overall sensitivity for EUS-FNA was 91.4%, specificity 100%, NPV 85%, PPV 100%, and accuracy 94.2%. Thirty cases (58%) had no mass seen on CT or MRI. EUS-FNA had a sensitivity of 93.8%, NPV 93.3%, and accuracy 96.7%. Twenty-two cases (42%) had a mass seen on imaging. Sensitivity, NPV and accuracy for EUS-FNA was 89.5%, 60%, and 91%, respectively. Forty-five cases (87%) had strictures located in the distal third of the bile duct. EUS-FNA had a sensitivity of 89.7%, NPV 84.2%, and accuracy 93.3%. Seven cases (13%) had strictures located in the proximal mid and upper third of the bile duct. EUS-FNA had a sensitivity, NPV, and accuracy of 100%. Nominal logistic model indicated no abdominal pain, female gender and mass on CT/MRI imaging as independent predictive factors for malignancy. Conclusions: EUS-FNA is a highly specific and sensitive technique for the diagnosis of indeterminate biliary strictures following negative intraductal tissue sampling. EUS can diagnose malignant strictures in the majority of cases that are undetectable by CT or MRI. EUS-FNA not only can assist management by obtaining a tissue diagnosis preoperatively, but it can also determine resectability and identify malignancies that do not require surgery.