Rationale: Rapid on-Site Evaluation (ROSE) of cytologic specimens acquired with EUS-guided fine needle aspiration (EUS-FNA) represents the most accurate available technique to reach a definitive diagnosis in patients with pancreatic solid masses. Cytologic interpretation, however, requires a high degree of expertise rarely found outside high volume centers and ROSE is not available in many countries. This has created a barrier to the widespread dissemination of EUS in the community and throughout the world, because the lack of cytologic expertise has resulted in a low diagnostic accuracy and, therefore, in a limited perceived utility of EUS. A device that is able to: (i) acquire histologic core biopsy samples usually easier to be interpreted; (ii) be used by most of the endosonographers and not only by the experts; (iii) have a performance at least not inferior to ROSE, will represent a major breakthrough in the field of EUS tissue acquisition. The availability of such needles will determine a shift from cytology to histology that will overcome some of the limitations of cytology and ROSE, thus strongly contributing to the diffusion of EUS throughout the world and in the community.

Objectives: To compare the performance and the diagnostic accuracy of EUS-guided fine needle biopsy (EUS-FNB) coupled with ROSE with that of EUS-FNB alone using an FNB needle.

Study design: International randomized multicenter trial. Study population: Patients ≥18 years old, referred for EUS-guided tissue sampling of a solid pancreatic mass.

Intervention: EUS-guided tissue acquisition by means of either EUS-FNB with ROSE or EUS-FNB alone, using one of the following FNB needles: Procore 20-gauge, SharkCore 22-gauge or Acquire 22-gauge.

Main study parameters/endpoints: The main endpoint is the diagnostic accuracy, measured against the gold standard diagnosis that will be surgical resection specimen or in non-operated patients the results of other diagnostic work-up (other tissue sampling techniques and imaging studies) or the clinical course of the disease. Secondary endpoints include: i) safety; ii) presence of tissue core; iii) feasibility to perform additional immunohistochemical/molecular biology analyses; iv) time of the sampling procedure.

Since its initial report in 1992, endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) has become an increasingly important tool for the evaluation of lesions of the gastro-intestinal tract and of adjacent organs. The diagnostic accuracy of EUS-FNA ranges from 60% to 90% according to the site of investigation and it is especially low for neoplasms such as stromal tumors, lymphomas, and well-differentiated adenocarcinomas that are difficult to be diagnosed by cytology alone. Moreover, the accuracy of EUS-FNA strongly relies on rapid on-site evaluation (ROSE) of the adequacy of the collected specimens by a cytopathologist or a cytotechnician, who can also help in establishing the need for additional samples to perform ancillary studies that are required in some cases to reach an effective diagnosis.

However, cytology requires a high degree of expertise rarely found outside high volume tertiary care centers and ROSE is not available in many countries Both these needs have created a barrier to the dissemination of EUS in the community and in many countries, because the lack of cytological expertise has resulted in a low diagnostic accuracy and, therefore, in a limited perceived utility of EUS.

Therefore, it would be of vital importance to have needles able to provide at the same time material for ROSE and histological core biopsy specimens to allow for further analyses, i.e. immunohistochemistry and molecular analysis. The availability of such needles would determine the centers with an established ROSE service to continue to use it and would also increase the chances that the patient will leave the service with a diagnosis and will have available additional material, so much needed in difficult cases or, in the near future, necessary to perform molecular studies in order to drive treatments. On the other hand in centers with no ROSE availability, needles with an accuracy not inferior to the one obtainable with ROSE will help overcome the limitations of cytology and ROSE and will facilitate the widespread utilization of EUS in the community and throughout the world.

To answer this important question, the investigators propose to perform an international multicenter randomized study with the aim of comparing EUS-FNB with ROSE versus EUS-FNB without ROSE using three novel needles (the 20-gauge ProCoreTM, the 22-gauge SharkCoreTM and the 22-gauge AcquireTM needle) in patients with solid pancreatic masses. These needles have become recently available and preliminary results for both pancreatic and non-pancreatic lesions are extremely encouraging Indeed, all these needles demonstrated a very high accuracy rate (>92%). Each center involved in the present study must have at least 2 of the 3 needles available. The non-inferiority design of the study will test the investigators hypothesis that EUS-FNB, by providing adequate samples for histologic examination, will perform at least as good as EUS-FNB with ROSE. The choice of the 20G ProCore ™, the 22G SharkCore™ or 22G Acquire™, instead of the 25G or the 19G, balances the need to use a needle that acquires enough tissue to perform all the studies needed to reach the definitive diagnosis, with its usability, i.e. a needle that can be used by most, if not all the endosonographers and not only by the experts. In this regards, the 25-gauge seems too small to gather enough tissue in a consistent number of patients while the 19-gauge is less maneuverable and more difficult to use thus preventing its utilization by all endosonographers.

This is an international randomized multicenter trial with two parallel arms in a (1:1) ratio. Consecutive patients with solid pancreatic masses and an indication to perform EUS-guided tissue acquisition will be evaluated and, if eligible, will be enrolled into the study.Patients will be randomized in a 1:1 ratio, using random 10 patients block sizes for allocation concealment between groups. An online randomization module will be made available to the participating centers. Randomization will take place after the lesion will have been visualized with EUS and the patient will be found suitable for inclusion. The choice of the needle to be used will be at the discretion of each endosonographer and will be done before randomization so that the choice of the needle does not create bias in the results. Nor the endoscopist, neither the pathologist will be blinded to which needle will be used.

The sample size has been calculated in order to demonstrate the non-inferiority of EUS-FNB without ROSE compared to EUS-FNB with ROSE in terms of diagnostic accuracy, having established a clinically acceptable margin of non-inferiority of 5%. The reported diagnostic accuracy of EUS-FNA with ROSE is 92%. With a type I error α of 5% and a power 1 - β of 80%, the total required sample size amounts to 730 patients (one-sided hypothesis testing of categorical data, comparing two binomial proportions of independent samples. Calculations executed with PASS, version 14.0.3). Considering, than, a 9.5% of patients to add in order to counteract the estimated rate of drop-out and lost to follow-up, 800 patients will be needed on the whole that is 400 per group.

Inclusion Criteria:

• Solid pancreatic mass referred for EUS-guided tissue acquisition
• Lesion can be visualized with EUS and needle puncturing can be technically feasible
• Written informed consent.

Exclusion Criteria:

• Known bleeding disorder that cannot be sufficiently corrected with co-fact or fresh frozen plasma (FFP)
• Use of anticoagulants that cannot be discontinued
• International Normalized Ratio (INR) >1.5 or platelet count <50.000
• Cystic lesions even with solid component
• Previous inclusion in other or present study
• Pregnancy