Gross visual inspection by endosonographers during endoscopic ultrasound-guided fine needle aspiration

Background/objectives: A clear criterion for terminating endoscopic ultrasound fine needle aspiration (EUS-FNA) without rapid on-site evaluation (ROSE) has not been established. However, a possible solution includes gross visual inspection (GVI) of the sample obtained with EUS-FNA. We performed a retrospective study to elucidate the efficacy of GVI for the diagnostic yield of EUS-FNA. Methods: Patients who underwent EUS-FNA of a pancreatic mass using a standard 22-G needle from January 2017 to December 2017 were included in the study. At least two punctures were performed for each patient, and GVI was performed for each pass by endoscopists. The correlation between GVI and pathological findings were investigated per needle pass for the first two passes. Regarding GVI, we evaluated the presence of a visible core (with or without) and the sample quantity (large or small). Results: We evaluated 126 EUS-FNA specimens and analyzed 252 needle passes. A final diagnosis of malignancy was made for 119 patients (94%). Accuracy rates were 92.5% with a visible core and 70.0% without a visible core (p <0.01), and 85.2% for large sample quantities and 70.2% for small sample quantities (p < 0.01). Univariate analysis indicated that the presence of a visible core and large sample quantity were associated with accuracy. Multivariate analysis indicated that only the presence of a visible core was significant. Conclusions: GVI can predict the correct diagnosis when ROSE is unavailable. Evaluating the presence of a visible core is more sensitive than assessing the quantity of the sample obtained. (C) 2018 IAP and EPC. Published by Elsevier B.V. All rights reserved.