Su1391 COMPARISON OF THE DIAGNOSTIC YEILD AND SAFETY OF 1ST GENERATION EUS CORE BIOPSY NEEDLES WITH FINE NEEDLE ASPIRATION AND NEXT GENERATION EUS CORE BIOPSY NEEDLES- A SYSTEMATIC REVIEW

Abstract

EUS-guided liver biopsy for the diagnosis of parenchymal liver diseases is being considered as an alternative to percutaneous method of liver biopsy. Several studies have evaluated the diagnostic yield from EUSLB with varying results. Different needle types have been used in the studies published to date. It is unknown if the discrepancy in the diagnostic yield is due to the various needle designs. We performed this meta-analysis to evaluate the diagnostic yield from different needle designs of similar gauge (19G only). We searched several electronic databases for all the studies evaluating the diagnostic yield of EUSLB for parenchymal liver diseases using 19G EUS-fine needle aspiration (FNA) or fine needle biopsy (FNB) needle (Procore and Shark Core needles). Studies were excluded if EUS FNA or FNB needle other than 19G was used, animal or ex-vivo models, biopsy for focal liver lesions, small case-series (<10). Measured outcomes included number of complete portal triads (CPT) which were pooled as weighted means. These were analyzed using random effects model. Heterogeneity was explored with meta-regression based on total specimen length (TSL). Weighted mean difference (WMD) were calculated to compare all three needles. A total of 9 studies with 394 patients (FNA needle (Boston-Scientific, Natick, Mass) in 78 patients, Procore needle (Cook Medical Inc, Winston-Salem, NC) in 127 patients and Shark Core needle (Medtronic, Sunnyvale, CA) in 189 patients, were included in this systematic review and meta-analysis. Weighted pooled mean with 95% confidence interval (CI) for Procore needle was 6.38 (3.92, 8.83), Cochran Q test P < 0.001, I2= 90%. Meta-regression analysis was conducted to explore heterogeneity. TSL was found to be a significant predictor of heterogeneity (intercept -0.05; TSL coefficient 0.36; P < 0.001) and R2 was 0.95. Therefore, longer specimen length resulted in higher number of complete portal tracts and completely explained heterogeneity across studies. Weighted pooled mean for Shark Core needle and FNA needle were 27.71 (8.27, 47.15), Cochran Q test P<0.001, I2=92% and 13.77 (4.18, 22.78), Cochran Q test P < 0.001, I2 =90%, respectively. Due to paucity of number of studies, meta-regression was not conducted for these analyses. WMD comparing Shark Core and Procore needles was 21.33 (-2.32, 44.99) (P=0.06). For the comparison between FNA and Procore needles WMD was 7.41 (-0.26, 15.10), (P=0.06). Finally, WMD for comparison of Shark Core with FNA was 13.91 (-16.92, 44.75) (P=0.37). Both Shark Core needle and FNA needle showed a trend towards obtaining higher complete portal tracts as compared to Procore needles. Further prospective studies comparing these needles are required at this time to identify optimal needle for EUS guided liver biopsy.Fig A: Forest plot showing complete portal tracts obtained by Shark Core needleFig. B: Forest plot showing complete portal tracts obtained by FNA needleView Large Image Figure ViewerDownload Hi-res image Download (PPT)