BACKGROUND AND AIMS: EUS-guided liver biopsy sampling using FNA and, more recently, fine-needle biopsy (FNB) needles has been reported with discrepant diagnostic accuracy, in part due to differences in methodology. We aimed to compare liver histologic yields of 4 EUS-based needles and 2 percutaneous needles to identify optimal number of needle passes and suction.

METHODS: Six needle types were tested on human cadaveric tissue: one 19G FNA needle, one existing 19G FNB needle, one novel 19G FNB needle, one 22G FNB needle, and two 18G percutaneous needles (18G1 and 18G2). Two needle excursion patterns (1 vs 3 fanning passes) were performed on all EUS needles. Primary outcome was number of portal tracts. Secondary outcomes were degree of fragmentation and specimen adequacy. Pairwise comparisons were performed using t tests, with a 2-sided P < .05 considered to be significant. Multivariable regression analysis was performed.

RESULTS: In total, 288 liver biopsy samplings (48 per needle type) were performed. The novel 19G FNB needle had significantly increased mean portal tracts compared with all needle types. The 22G FNB needle had significantly increased portal tracts compared with the 18G1 needle (3.8 vs 2.5, P < .001) and was not statistically different from the 18G2 needle (3.8 vs 3.5, P = .68). FNB needles (P < .001) and 3 fanning passes (P ≤ .001) were independent predictors of the number of portal tracts.

CONCLUSIONS: A novel 19G EUS-guided liver biopsy needle provides superior histologic yield compared with 18G percutaneous needles and existing 19G FNA and core needles. Moreover, the 22G FNB needle may be adequate for liver biopsy sampling. Investigations are underway to determine whether these results can be replicated in a clinical setting.