Endoscopic ultrasound-fine needle aspiration (EUS-FNA) for pancreatic lesions: effectiveness in clinical practice.

BACKGROUND: Diagnosis of pancreatic masses is often difficult. Endoscopic ultrasound-fine needle aspiration has been proposed as the best single-step strategy. AIMS: To prospectively evaluate feasibility, effectiveness and safety of endoscopic ultrasound-fine needle aspiration of pancreatic masses in a consecutive study of unselected patients. METHODS: Two hundred ninety-three patients were enrolled in two referral Hospitals in Northern Italy. All patients were referred either due to the presence of imaging test abnormalities (suspected or evident masses, or features indirectly suggesting the presence of a mass) or due to clinical or biochemical findings suggesting pancreatic cancer in the absence of positive imaging. All patients underwent linear array endoscopic ultrasound and, when indicated, fine needle aspiration. All procedures were recorded prospectively. The final diagnosis was established at the end of follow-up or when the patients underwent surgery or died. RESULTS: Fine needle aspiration was indicated in 246 of 293 cases (84%), considered technically feasible in 232 of 246 cases (94%) and gave adequate samples for histopathological diagnosis in 204 of 232 cases (88%). Endoscopic ultrasound sensitivity, specificity and accuracy were 79, 60 and 72%, respectively; the corresponding figures for endoscopic ultrasound-fine needle aspiration were 80, 86 and 82%. There was good agreement with final diagnosis for endoscopic ultrasound-fine needle aspiration (kappa 0.673, 95%CI 0.592-0.753), greater than that for endoscopic ultrasound alone (kappa 0.515, 95%CI 0.425-0.605). There was one case of intracystic haemorrhage and one case of transient hyperthermia (0.3%). CONCLUSIONS: Endoscopic ultrasound-fine needle aspiration of pancreatic masses seems to be feasible, effective and safe in this consecutive study of patients.