P-LUM-06: Diagnostic accuracy of combined needle use of the new 20-gauge procore fine needle biopsy and the 25-gauge fine needle aspiration needle in solid gastrointestinal lesions

BACKGROUND AND OBJECTIVES: Endoscopic ultrasound (EUS) fine needle biopsy (FNB) needles were designed to improve histologic yield, while maintaining the flexibility to permit ease of use. However, the diagnostic benefit of EUS-FNB needles as compared to the conventional fine needle aspiration (FNA) needles remains unclear. This study aimed to identify the diagnostic accuracy of combined needle use of a new 20-gauge ProCore FNB needle and a 25-gauge FNA needle. METHODS: For this study, cases were selected in which both the 20-gauge ProCore FNB needle and the 25-gauge FNA needle were used during a single EUS procedure for sampling of a solid pancreatic mass, lymph node, or submucosal mass. Tissue acquisition was performed in the course of the ASPRO trial, a multicenter randomized study comparing the diagnostic value of 20-gauge ProCore FNB and 25-gauge FNA needles (ClinicalTrials.gov: NCT02167074). The protocol allowed additional sampling with the nonassigned needle, at the discretion of an endosonographer. RESULTS: Of all 615 patients who were randomized for the ASPRO study, 74 were sampled with both needle types. In these combined needle cases, FNA was used first in 24 and ProCore FNB in 50 cases. Target lesions encompassed 39 solid pancreatic lesions, 18 submucosal masses, and 17 lymph nodes. Most pancreatic lesions were located in the head (25/39), lymph nodes were mainly located in the abdomen (15/18), and submucosal lesions were located in the stomach (8/20), esophagus (4/20), small intestines (3/20), and rectum (3/20). The main reason to use ProCore in addition to FNA was to collect more tissue for ancillary testing (79%). FNA was generally used in addition to ProCore to allow for on-site pathological assessment (76%). The regimen of FNA followed by ProCore resulted in 100% accuracy, while for ProCore followed by FNA, 88% was reached (P = 0.086). The type of target lesion, either pancreatic or nonpancreatic, did not affect the diagnostic accuracy of the two sampling regimens (odds ratio 1.2, 95% confidence interval 0.22–6.61, P = 0.834). CONCLUSION: FNA followed by ProCore FNB provides for a 100% diagnostic accuracy and tends to outperform ProCore FNB followed by FNA. Accuracy rates were independent of the type of target lesion.