OBJECTIVE: To evaluate and to compare the diagnostic yield of ERCP brush cytology (ERCP) and EUS-FNA in patients with biliary strictures and evaluates the agreement between general pathologists (GP) and expert GI pathologists (GIP) in the final diagnosis of biliary strictures. METHODS: Patients with biliary strictures documented by ERCP were included. Brush cytology was performed and during EUS, only visible mass lesions or localized bile duct wall thickening were aspirated. The gold standard method for diagnosis was surgical histology and/or follow-up. Tissue sampling results were: malignant, suspicious, atypical, insufficiently or benign. Specimens were interpreted by GP and GIP, blinded for prior tests results. RESULTS: 46 patients were included. Final diagnosis was malignancy in 37 (26 pancreatic--11 biliary) and benign in 9 (8 chronic pancreatitis--1 common bile duct inflammatory stricture). Sensitivity and accuracy for ERCP brush cytology were 43.2% and 52.2% for GP and 51.4% and 58.7% for GIP. Sensitivity and accuracy for EUS-FNA were 52.8% and 58.5%, respectively for GP and 69.4% e 73.2% for GIP. In comparison, the combination of brush cytology and EUS-FNA demonstrated higher sensitivity and accuracy for both GP (64.9% and 69.6%, respectively) and GIP (83.8% and 84.8%, respectively) and improved agreement with final diagnosis for both (mostly for GIP). CONCLUSION: Both, ERCP brush cytology and EUS-FNA has a similar yield for the diagnosis of biliary strictures. However, the combination of these methods results in an improved diagnostic accuracy. In addition, GIP might be expected to interpret specimens with greater accuracy than GP.
OBJECTIVE: To evaluate and to compare the diagnostic yield of ERCP brush cytology (ERCP) and EUS-FNA in patients with biliary strictures and evaluates the agreement between general pathologists (GP) and expert GI pathologists (GIP) in the final diagnosis of biliary strictures. METHODS:Patients with biliary strictures documented by ERCP were included. Brush cytology was performed and during EUS, only visible mass lesions or localized bile duct wall thickening were aspirated. The gold standard method for diagnosis was surgical histology and/or follow-up. Tissue sampling results were: malignant, suspicious, atypical, insufficiently or benign. Specimens were interpreted by GP and GIP, blinded for prior tests results. RESULTS: 46 patients were included. Final diagnosis was malignancy in 37 (26 pancreatic--11 biliary) and benign in 9 (8 chronic pancreatitis--1 common bile duct inflammatory stricture). Sensitivity and accuracy for ERCP brush cytology were 43.2% and 52.2% for GP and 51.4% and 58.7% for GIP. Sensitivity and accuracy for EUS-FNA were 52.8% and 58.5%, respectively for GP and 69.4% e 73.2% for GIP. In comparison, the combination of brush cytology and EUS-FNA demonstrated higher sensitivity and accuracy for both GP (64.9% and 69.6%, respectively) and GIP (83.8% and 84.8%, respectively) and improved agreement with final diagnosis for both (mostly for GIP). CONCLUSION: Both, ERCP brush cytology and EUS-FNA has a similar yield for the diagnosis of biliary strictures. However, the combination of these methods results in an improved diagnostic accuracy. In addition, GIP might be expected to interpret specimens with greater accuracy than GP.