A Prospective Multicenter Single-Blind Randomized Controlled Trial Comparing FNA and FNB in the Diagnosis of Solid Masses

Qian Chen,Bin Cheng,Tianan Jiang,Lizhou Dou,Jinglin Wang,Yawen Li,Yueming Zhang,Bo Sun,Hongbo Shan,Zhuang Deng,Guoliang Xu,Xiujiang Yang,Guiqi Wang

DOI: https://doi.org/10.14309/00000434-201710001-00773

2017-01-01

Abstract:Introduction: Endoscopic ultrasound (EUS) guided fine needle biopsy (FNB) with side fenestration on the needle was developed aiming to obtain both cytological aspirates and histological core samples. To reveal the diagnostic yield of FNB sampling solid mass compared to standard fine needle aspiration (FNA), we conducted a large-scale multicenter study in patients with pancreatic and non-pancreatic masses. We also aimed to establish optimal techniques in EUS tissue acquisition. Methods: This was a prospective multicenter single-blind randomized controlled trial. Between December 2014 and January 2016, 408 patients were enrolled from five tertiary-care academic medical centers and randomized at a 1:1 ratio to receive either 22G FNA or 22G FNB. Two tissue acquisition (TA) techniques (1st and 2nd passes using slow pull vs. 3rd and 4th passes using suction methods) were performed on all EUS needles. The primary outcome was the diagnostic yield comparing FNA and FNB on all solid lesions, pancreatic and non-pancreatic masses, respectively. Secondary endpoints were the quality of the histologic specimen (quantity of cell, tissue integrity, and blood contamination). Results: Among all lesion, FNB group (n=190) compared to FNA group (n=187) gained superior histological diagnostic yield (91.44% vs. 80%; p = 0.0015), and higher cytological diagnostic yield (85.56% vs. 78.95%; p = 0.093). For pancreatic mass, FNB group (n=123) achieved better histological diagnostic yield compared to FNA group (n=126) (92.68% vs. 81.75%; p = 0.0099), and also higher cytological diagnostic yield (88.62% vs. 79.37%; p = 0.00468), respectively. For non-pancreatic mass, there was no difference between FNB and FNA on diagnostic accuracy. Furthermore for pancreatic mass, when slow pull technique was applied in the first and second passes, FNB was superior to FNA on specimen quality and diagnostic yield. Conclusion: FNB needle achieved superior overall diagnostic yield compared to FNA needle among all studied solid lesions. For pancreatic masses, the 22G FNB needle also gained superior histological diagnostic accuracy and higher cytological diagnostic yield compared to 22G FNA. We further suggest choosing FNB over FNA on pancreatic lesion for TA when slow pull technique was applied in the first two passes.Table: Table. Technical characteristics and outcomes of EUS-FNA/FNB