Comparison of clinical/histological outcomes according to puncture sites in endoscopic ultrasound-guided fine needle biopsy for large pancreatic masses: Multicenter randomized prospective pilot study.

ABSTRACT

OBJECTIVES: There are no recommendations regarding the optimal puncture site in endoscopic ultrasound-guided fine needle biopsy (EUS-FNB). This multicenter randomized prospective study compared the diagnostic accuracy and histological findings according to the sampling site for pancreatic masses larger than 3 cm. METHODS: Consecutive patients with pancreatic masses larger than 3 cm indicated for EUS-FNB were included in the study. Patients were randomly assigned to two groups for the initial puncture site (central vs. peripheral sampling of the masses). A minimum of four passes were performed, alternating between the center and the periphery. The primary outcome was diagnostic accuracy. RESULTS: A total of 100 patients were equally divided into the central group and the peripheral group. The final diagnosis revealed malignancy in 95 patients (pancreatic cancer [n = 89], neuroendocrine tumor [n = 4], lymphoma [n = 1], metastatic carcinoma [n = 1]), and benign conditions in five patients (chronic pancreatitis [n = 4], autoimmune pancreatitis [n = 1]). There was no significant difference in diagnostic accuracy between the puncture sites. However, combining samples from both areas resulted in higher diagnostic accuracy (97.0%) compared to either area alone, with corresponding values of 88.0% for the center (P = 0.02) and 85.0% for the periphery (P = 0.006). CONCLUSIONS: Both central sampling and peripheral sampling showed equivalent diagnostic accuracy in detecting malignancy. However, combining samples from both areas generated superior diagnostic yield compared to using either sampling site alone. For pancreatic masses larger than 3 cm, it is advisable to consider sampling from various areas of the masses to maximize the diagnostic yield.