A Meta-Analysis Comparing Endoscopic Ultrasound-guided Fine-needle Aspiration With Endoscopic Ultrasound-guided Fine-needle Biopsy

Background: The superiority between endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) and endoscopic ultrasound-guided fine-needle biopsy (EUS-FNB) is still a debate. The purpose of this meta-analysis was to compare the diagnostic value and safety of these 2 needles in sampling of all solid lesions and separately pancreatic lesions. Materials and Methods: PubMed, EMBASE, and the Cochrane library were searched for randomized controlled studies that reported the comparison of FNA and FNB in solid mass. Outcome measures included diagnostic accuracy, number of needle passes, diagnostic adequacy, presence of tissue cores, and adverse events. Standard error, upper and lower confidence intervals at 95% confidence interval for the risk were obtained using Review Manager Version 5.3 which was also used to generate forest plots for pooled analysis. The random or fixed effect model was applied depending on the heterogeneity (I (2)). Results: Eighteen randomized control trial studies with a total of 2718 patients (1141 patients with EUS-FNA, 1108 with EUS-FNB, and remaining 469 patients were sampled with both needles alternatively) were included in the meta-analysis. FNB group has relatively good diagnostic accuracy relative risk (RR): 0.94, 0.92-0.97; P=0.0002), diagnostic adequacy (RR: 0.95, 0.9-1.0; P=0.04) and high quality histologic yield compared (RR: 0.77, 0.64-0.93; P=0.007) with the FNA group in solid gastrointestinal lesions, and the number of needle passes to obtain sufficient tissue (mean difference: 0.54, 0.45-0.64; P<0.00001) was lower in the FNB group. For solid pancreatic disease only, there was no difference in diagnostic accuracy (RR: 0.97, 0.93-1.01, P=0.13) or quality histologic yield (RR: 0.60, 0.29-1.23; P=0.16). The rate of adverse events (RR: 1.04, 0.48-2.29; P=0.92) did not significantly differ between FNA and FNB groups. Conclusions: In solid gastrointestinal lesions, FNB is associated with a relatively better diagnostic adequacy, diagnostic adequacy and tissue cores rates, and less number of needle passes. For solid pancreatic disease only, there is no difference in diagnostic accuracy or tissue cores rates.