EUS-Guided Fine Needle Tissue Acquisition Using a Nitinol Ultra Flex 19-Gauge Needle for Transduodenal Lesions: a Multicenter Prospective Study

Background and Aim: EUS-guided fine needle biopsy (EUS-FNB) using 19G needles to obtain core biopsy tissue samples may be difficult when performed transduodenally. Recent data on a limited number of patients have suggested that a new nitinol 19-gauge needle (ExpectTM 19 Flex needle, Boston Scientific Corp, Natick, MA, USA) can improve the performance of EUS-FNB from the duodenum. Aim of this study was to determine the feasibility, diagnostic yield and accuracy of EUS-FNB with the nitinol ultra flex 19-gauge needle for lesions approached from the duodenum. Methods: This is a multicenter prospective study performed in 6 tertiary referral centers. Patients with solid lesions, which needed transduodenal EUS-FNB were considered eligible. The EUS-fine needle tissue acquisition technique, in which the stylet is removed before inserting the needle in the working channel of the echoendoscope to increase needle flexibility was used in all cases. Three needle passes were attempted in all patients with the Expect 19G Flex needle. In case of failure the use of another needle was at the discretion of the endoscopist. Technical success rate, diagnostic yield and accuracy were the primary endpoints. Safety and assessment of difficulty of needle use were the secondary endpoints. Results: Up to now, 172 cases have been included (mean age 66 13 years, 38% females): 143 were pancreatic or peripancreatic mass lesions and 29 were lymph nodes (mean size 33.112.4mm), 59.9% of which were approached from the duodenal bulb and the remaining from the second (37.2%) or third (2.9%) duodenal portion. FNA was technically feasible in 156 lesions (90.7%) and an adequate histological sample was obtained in 132 cases (76.7%). Considering malignant vs. non-malignant diseases, sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio were 65.8% (95%CI 57.7-73.1), 100% (95%CI 75.9-100), infinity and 0.34 (95% CI 0.27- 0.43), respectively, yielding an overall diagnostic accuracy of 68.6%. Five major complications were reported (3 cases of bleeding, 1 perforation and 1 acute pancreatitis). The needle use was judged easy in 63.1% and 62.2% of cases for lesions approached from the bulb and from the second/third duodenal portion, respectively (pZ0.767). Conclusion: Our multicenter prospective study demonstrated that EUS-FNTA using the new nitinol 19G Flex for sampling mass lesions through the duodenum is technically successful in most cases, but diagnostic yield and accuracy are suboptima