22G versus 25G biopsy needles for EUS-guided tissue sampling of solid pancreatic masses: a randomized controlled...

No comparative study of 22-gauge biopsy needles (PC22) and 25-gauge biopsy needles (PC25) has been conducted. We prospectively compared the diagnostic accuracy of PC22 and PC25 in patients with pancreatic and peripancreatic solid masses.

We conducted a randomized noninferiority clinical study from January 2013 to May 2014 at Samsung Medical Center. A cytological and histological specimen of each pass was analyzed separately by an experienced pathologist. The primary outcome was to assess the diagnostic accuracy using the PC22 or PC25. Secondary outcomes included the optimal number of passes for adequate diagnosis, core specimen yield, sample adequacy, and complication rates.

Diagnostic accuracy of combining cytology with histology in three cumulative passes was 97.1% (100/103) for the PC22 and 91.3% (94/103) for the PC25 group. Thus, noninferiority of PC25 to PC22 was not shown with a 10% noninferiority margin (difference, -5.8%; 95% CI, -12.1 to -0.5%). In a pairwise comparison with each needle type, two passes was non-inferior to three passes in the PC22 (96.1% vs. 97.1%; difference, -0.97%; 95% CI -6.63 to 4.69%) but noninferiority of two passes to three passes was not shown in the PC25 group (87.4% vs. 91.3%; difference, -3.88%; 95% CI, -13.5 to 5.7%).

Non-inferiority of PC25 to PC22 diagnostic accuracy was not observed for solid pancreatic or peripancreatic masses without on-site cytology. PC22 may be a more ideal device because only two PC22 needle passes was sufficient to establish an adequate diagnosis, whereas PC25 required three or more needle passes.