A novel EUS-guided fine-needle biopsy sampling (EUS-FNB) needle enabled physicians to obtain sufficient pathologic samples with fewer to-and-fro movements (TAFs) within the lesion. We compared the diagnostic yields of EUS-FNB with 3 and 12 TAFs at each puncture pass.

The primary endpoint of this multicenter, noninferiority, crossover, randomized controlled trial involving 6 centers was diagnostic sensitivity. Secondary endpoints were diagnostic accuracy and quantity and quality evaluation of EUS-FNB specimens. Length of the macroscopically visible core (MVC) and microscopic histologic quantity were used for quantitative evaluation. Macroscopic visual and microscopic histologic evaluations were performed for qualitative evaluation.

Among 110 patients (220 punctures, 110 for 3 TAFs and 12 TAFs each), 105 (210 punctures) had malignant histology. Diagnostic sensitivity for malignancy of 3 TAFs (88.6%) was not inferior to that of 12 TAFs (89.5%; difference, –.9%; 95% confidence interval, –9.81 to 7.86). Diagnostic accuracy for malignancy was 92.7% for 3 TAFs and 94.6% for 12 TAFs. Overall median MVC length was 13.5 mm in both groups. The 3-TAF group had a significantly higher rate of score ≥3 on macroscopic visual quality evaluation than the 12-TAF group (71.8% vs 52.7%, P  = .009). No significant intergroup differences existed in microscopic histologic quantity and quality evaluations (quantity evaluation, 88.2% for 3 TAFs vs 83.6% for 12 TAFs; quality evaluation, 90.0% for 3 TAFs vs 89.1% for 12 TAFs).

Diagnostic sensitivity and accuracy of EUS-FNB with 3 TAFs were not inferior to those with 12 TAFs for solid pancreatic lesions. The 3-TAF group showed significantly less blood contamination in sampled tissues than the 12-TAF group. (Clinical trial registration number: UMIN000037309.)