Evaluation of fine needle biopsy (FNB) for endoscopic ultrasound (EUS) guided tissue acquisition of pancreatic masses to negate the need for rapid on-site evaluation: a multi-centre randomized control trial

Background: Endoscopic ultrasound guided fine needle aspiration (EUS FNA) is a key component for the investigation and diagnosis of solid pancreatic masse. It has a proven track record for being safe with complications and tumour seeding being a rare occurrence. The sensitivity, specificity and diagnostic accuracy for pancreatic lesions in many studies is high and exceeds 85%. Studies aiming to optimize EUS FNA have compared several factors including needle size, the number of needles passes and suction versus slow-pull. Initial studies showed that the use of rapid on-site evaluation (ROSE) had the biggest impact with increased diagnostic accuracy and reduced needle passes. However more recent reports show conflicting results on whether ROSE actually does influence outcome in EUS FNA. ROSE has historically been performed using FNA needles with cytological assessment. The act of smearing specimens (a necessary step for ROSE) to prepare cytology slides using cell aspirates obtained from FNA needles can be done without difficulty after necessary training and experience. The fine needle biopsy (FNB) needle (the ProCore needle) was introduced to further improve upon EUS guided tissue sampling. It is designed to collect a larger amount of core sample tissue by having a reverse bevel near the tip. Several studies favour the EUS FNB needle over the classic FNA needle for obtaining more adequate histological specimens with a high diagnostic yield. Data comparing the performance of EUS FNB without ROSE versus EUS FNA with ROSE is currently lacking. With this in mind, needles that provide histological specimens with an excellent (>95%) diagnostic yield (thus negating the need for ROSE) are the ideal devices for obtaining EUS guided diagnostic tissue. We therefore aim to compare the diagnostic yield of solid pancreatic masses using EUS FNA with ROSE versus EUS FNB without ROSE. Hypothesis: EUS FNB without ROSE will have equal diagnostic accuracy to EUS FNA with ROSE, hence negating the need for ROSE when EUS guided tissue acquisition is performed with FNB needle for solid pancreatic masses. Aim: To compare the outcomes of EUS FNA with ROSE versus EUS FNB without ROSE for solid pancreatic masses. Design: This will be a prospective, multi-centre, randomized control trial involving 600 patients with pancreatic solid mass of more than 1cm. There will be 10 centres participate in the trial, each recruiting 60 patients per hospital. Patients at each site will be randomized to either: Group A, (n=30, total = 300) which will undergo EUS FNA with the standard 22G EchoTip needle with ROSE. Group B (n=30; total=300), which will undergo EUS FNB with the 22G ProCore needle without ROSE.