BACKGROUND/AIMS: Optimal management of proximal biliary strictures depends on staging and adequate tissue diagnosis of suspected malignant lesions. Sensitivity of ERCP brush cytology is poor. EUS/EUS-FNA of these lesions is challenging with limited data on its diagnostic accuracy. We report our experience of EUS/ EUS guided FNA of proximal biliary strictures in a tertiary referral centre. METHODOLOGY: All patients who underwent EUS guided FNA of hilar lesions between October 2003 to July 2007 were identified. The final diagnosis was determined by surgical pathology, results of EUS-FNA or follow-up. RESULTS: 32 patients underwent 36 procedures for hilar lesions during the study period. EUS detected a mass in 9/14 patients who did not have a mass seen on imaging. The mean followup period was 39.43 months. The cytological specimen was adequate in 26 patients. The final diagnosis was adenocarcinoma (24) and benign (8). The overall accuracy, sensitivity, specificity, positive predictive value and negative predictive value of EUS-FNA were 68%, 52 %, 100%, 100% and 54%, respectively. If only adequate aspirates were included in the analysis, the values were 74%, 60%, 100%, 100% and 55% respectively. CONCLUSIONS: EUS guided FNA is a useful modality in the diagnosis and management of hilar lesions.
BACKGROUND/AIMS: Optimal management of proximal biliary strictures depends on staging and adequate tissue diagnosis of suspected malignant lesions. Sensitivity of ERCP brush cytology is poor. EUS/EUS-FNA of these lesions is challenging with limited data on its diagnostic accuracy. We report our experience of EUS/ EUS guided FNA of proximal biliary strictures in a tertiary referral centre. METHODOLOGY: All patients who underwent EUS guided FNA of hilar lesions between October 2003 to July 2007 were identified. The final diagnosis was determined by surgical pathology, results of EUS-FNA or follow-up. RESULTS: 32 patients underwent 36 procedures for hilar lesions during the study period. EUS detected a mass in 9/14 patients who did not have a mass seen on imaging. The mean followup period was 39.43 months. The cytological specimen was adequate in 26 patients. The final diagnosis was adenocarcinoma (24) and benign (8). The overall accuracy, sensitivity, specificity, positive predictive value and negative predictive value of EUS-FNA were 68%, 52 %, 100%, 100% and 54%, respectively. If only adequate aspirates were included in the analysis, the values were 74%, 60%, 100%, 100% and 55% respectively. CONCLUSIONS: EUS guided FNA is a useful modality in the diagnosis and management of hilar lesions.