Randomized trial comparing 15 vs 5 actuations per pass during endoscopic ultrasound-guided fine-needle biopsy for specimen acquisition of solid pancreatic lesions

Abstract

Background and aim: There is currently no established number of actuations (to-and-fro movements) per pass during endoscopic ultrasound-guided fine-needle biopsy (EUS-FNB). This study aimed to compare 15 vs 5 actuations in terms of adequate specimen acquisition of solid pancreatic lesions.

Methods: In this prospective, randomized, crossover, noninferiority, single-center study, eligible patients underwent EUS-FNB using a 22-G Franseen needle with both 15 and 5 actuations per pass, performed in a randomized order, from October 2020 to December 2021. The acquired specimens from each pass were separately evaluated. The primary outcome was the accuracy of the histological diagnosis per pass. The noninferiority margin was set as 15%.

Results: Data from 85 patients were analyzed, revealing pancreatic cancer in 73 patients. The accuracy of the histological diagnosis in the 15 and 5 actuations groups was 83.5% (71/85) and 77.7% (66/85), respectively. The difference was -5.8% (95% confidence interval -15.6-3.4), which does not indicate noninferiority of the five actuations group. Among the secondary outcomes, the 15 actuations group was significantly superior to the five actuations group in terms of the obtained core tissues (1.88 [interquartile range 0.89-3.64] mm² vs 1.66 [0.83-2.71] mm² [P = 0.031]) and subjective evaluation of cytology specimens for pancreatic cancer (69.0% vs. 31.0%, P = 0.005).

Conclusions: The noninferiority of five actuations in the accuracy of the histological diagnosis was not confirmed, and 15 actuations are preferred during EUS-FNB for solid pancreatic lesions.

Keywords: Franseen needle; endoscopic ultrasound-guided fine-needle biopsy; histological diagnosis; solid pancreatic lesion; specimen acquisition.