Abstract
BACKGROUND: While endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA) is considered a preferred technique for tissue sampling for solid lesions, fine needle biopsy (FNB) has recently been developed. AIM: To compare the accuracy of FNB vs FNA in determining the diagnosis of solid lesions. METHODS: A retrospective, multi-center study of EUS-guided tissue sampling using FNA vs FNB needles. Measured outcomes included diagnostic test characteristics (i.e., sensitivity, specificity, accuracy), use of rapid on-site evaluation (ROSE), and adverse events. Subgroup analyses were performed by type of lesion and diagnostic yield with or without ROSE. A multivariable logistic regression was also performed. RESULTS: A total of 1168 patients with solid lesions (n = 468 FNA; n = 700 FNB) underwent EUS-guided sampling. Mean age was 65.02 ± 12.13 years. Overall, sensitivity, specificity and accuracy were superior for FNB vs FNA (84.70% vs 74.53%; 99.29% vs 96.62%; and 87.62% vs 81.55%, respectively; P < 0.001). On subgroup analyses, sensitivity, specificity, and accuracy of FNB alone were similar to FNA + ROSE [(81.66% vs 86.45%; P = 0.142), (100% vs 100%; P = 1.00) and (88.40% vs 85.43%; P = 0.320]. There were no difference in diagnostic yield of FNB alone vs FNB + ROSE (P > 0.05). Multivariate analysis showed no significant predictor for better accuracy. On subgroup analyses, FNB was superior to FNA for non-pancreatic lesions; however, there was no difference between the techniques among pancreatic lesions. One adverse event was reported in each group. CONCLUSION: FNB is superior to FNA with equivalent diagnostic test characteristics compared to FNA + ROSE in the diagnosis of non-pancreatic solid lesions. Our results suggest that EUS-FNB may eliminate the need of ROSE and should be employed as a first-line method in the diagnosis of solid lesions.