New through-the-needle brush for pancreatic cyst assessment: a randomized controlled trial

Current EUS technologies are suboptimal in the assessment of pancreatic cystic lesions (PCLs). We developed a new through-the-needle brush, the "loop brush," to improve the cellular yield, and thereby sensitivity, of EUS-guided FNA (EUS-FNA) of pancreatic cysts. In this study, we evaluated its safety and efficacy.

We performed an in vivo randomized controlled trial in pigs using artificial cysts. In 1 group, the loop brush was deployed through a 22-gauge needle using EUS-FNA into the cysts. In the control group, cystic puncture was performed with a standard needle. Loop brushes were visually inspected after the procedure. Cytologic assessment, cell counting, and hemoglobin analysis were performed in samples from both groups.

One hundred fourteen artificial cysts were punctured in 6 pigs, 57 in each group. Neither adverse events nor significant device malfunction occurred during loop brushing. Samples collected with the brush had nondetectable concentrations of hemoglobin in 72% of cases (41/57), and 26% (16/57) had <.6 g/dL, with no significant differences with the controls (P = .32). Brushing cell counts were associated with significantly increased cell counts (11.7 times the median difference, P < .0001). Cytologic smears were diagnostic in 77% of cases in the brushing group and 54% in the control group (P = .01 [Fisher exact test]; P = .006 [χ² test]).

The procedure using the new loop brush appears to be safe, causing neither significant bleeding nor device malfunction. Samples obtained with the loop brush were suitable for cytologic analysis and showed significantly higher cell yield than controls. Further clinical studies are warranted.