Comparative diagnostic performance of end-cutting fine-needle biopsy needles for EUS tissue sampling of solid pancreatic masses: a network meta-analysis.

ABSTRACT

BACKGROUND AND AIMS: Evidence is limited on the comparative diagnostic performance of newer end-cutting fine-needle biopsy (FNB) needles for tissue sampling of pancreatic masses. We performed a systematic review with network meta-analysis to compare the diagnostic accuracy of available FNB needles for sampling of solid pancreatic lesions. METHODS: A systematic literature review (Medline and Cochrane Database) was conducted for studies evaluating the accuracy of newer FNB needles in adults undergoing EUS-guided sampling of solid pancreatic masses. The primary outcome was diagnostic accuracy. Secondary outcomes were sample adequacy, diagnostic sensitivity, specificity, and adverse event rate. We performed pairwise and network meta-analyses and appraised the quality of evidence using Grading of Recommendations Assessment, Development and Evaluation methodology. RESULTS: Overall, 16 RCTs (1934 patients) were identified. On network meta-analysis, Franseen needles (Acquire; Boston Scientific, Marlborough, Mass, USA) significantly outperformed reverse-bevel needles (risk ratio [RR], 1.21 [95% confidence interval {CI}, 1.05-1.40] for accuracy and 1.31 [95% CI, 1.05-1.22] for adequacy) and FNA needles (RR, 1.21 [95% CI, 1.01-1.25] for accuracy and 1.07 [95% CI, 1.02-1.13] for adequacy). Likewise, the Fork-tip needle (SharkCore; Medtronic, Dublin, Ireland) was significantly superior to the reverse-bevel needle (RR, 1.17 [95% CI, 1.03-1.33] for accuracy and 1.09 [95% CI, 1.02-1.16] for adequacy) and to the FNA needle (RR, 1.09 [95% CI, 1.01-1.19] for accuracy and 1.03 [95% CI, 1.01-1.07] for adequacy). Other comparisons did not achieve statistical significance. As a consequence, Franseen (surface under the cumulative ranking score, .89 for accuracy and .94 for adequacy) and Fork-tip needles (surface under the cumulative ranking score, .76 for accuracy and .73 for adequacy) ranked as the 2 highest-performing FNB needles. When considering different needle sizes, 25-gauge Franseen and 25-gauge Fork-tip needles were not superior to 22-gauge reverse-bevel needles (RR, 1.18 [95% CI, .96-1.46] and 1.04 [95% CI, .62-1.52]). None of the tested needles was significantly superior to the other FNB devices or to FNA needles when rapid onsite cytologic evaluation was available. CONCLUSIONS: Franseen and Fork-tip needles, particularly 22-gauge size, showed the highest performance for tissue sampling of pancreatic masses, with low confidence in estimates.