Comparative accuracy of needle sizes and designs for EUS tissue sampling of solid pancreatic masses: a network meta-analysis

Background and Aims: Variable diagnostic performance of sampling techniques during EUS-guided tissue acquisition of solid pancreatic masses based on needle type (FNA versus fine-needle biopsy [FNB]) and gauge (19-gauge vs 22-gauge vs 25-gauge) has been reported. We performed a systematic review with network meta-analysis to compare the diagnostic accuracy of EUS-guided techniques for sampling solid pancreatic masses. Methods: Through a systematic literature review to November 2018, we identified 27 randomized controlled trials (2711 patients) involving adults undergoing EUS-guided sampling of solid pancreatic masses that evaluated the diagnostic performance of FNA and FNB needles based on needle gauge. The primary outcome was diagnostic accuracy. Secondary outcomes were sample adequacy, histologic core procurement rate, and number of needle passes. We performed pairwise and network meta-analyses and appraised the quality of evidence using GRADE (Grading of Recommendations Assessment, Development and Evaluation) methodology. Results: In the network meta-analysis, no specific EUS-guided tissue sampling technique was superior, based on needle type (FNA vs FNB) or gauge (19-gauge vs 22-gauge vs 25-gauge) (low-quality evidence). Specifically, there was no difference between 25-gauge FNA versus 22-gauge FNA (relative risk [RR], 1.03; 95% confidence interval [CI], 0.91-1.17) and 22-gauge FNB versus 22-gauge FNA (RR, 1.03; 95% CI, 0.89-1.18) needles for diagnostic accuracy, sample adequacy, and histologic core procurement. Findings were confirmed in sensitivity analysis restricted to studies with no rapid on-site cytologic evaluation and no use of the fanning technique. Conclusion: In a network meta-analysis, no specific EUS-guided tissue sampling technique was superior with regard to diagnostic accuracy, sample adequacy, or histologic procurement rate for solid pancreatic masses, with low confidence in estimates.