Diagnostic Performance of Endoscopic Ultrasound-Guided Fine Needle Biopsy with Histological Analysis Versus Combined Cytohistological Analysis in Solid Pancreatic Lesions: A Systematic Review and Meta-Analysis

Abstract

Background: Recent advancements in endoscopic ultrasound-guided fine needle biopsy (EUS-FNB) needles enable reliable histological evaluation, raising questions about the necessity of routinely adding cytological analysis. Aim: To compare the diagnostic performance of EUS-FNB using histology alone versus combined cytohistological analysis for detecting solid pancreatic malignancies. Methods: A comprehensive search of MEDLINE, Embase, and Cochrane databases identified studies reporting diagnostic performance, tissue adequacy, processing methods, and reference standards of surgical specimens, clinical, and imaging assessment. Pooled sensitivity, specificity, tissue adequacy, and heterogeneity were analyzed using meta-analytic techniques. Results: Twenty-nine studies were included. Histological and combined analyses were performed in 1657 and 1271 patients, respectively. For histological analysis, pooled sensitivity was 90.5% (95% CI: 85.8–94.2, I² = 85.2%) and specificity was 98.2% (95% CI: 95.7–99.5, I² = 0%), with an area under the receiver operating characteristic curve (AUROC) of 0.95 (95% CI: 0.92–0.97). Combined analysis achieved slightly higher sensitivity of 91.4% (95% CI: 87.7–94.5, I² = 75.0%) and similar specificity of 98.0% (95% CI: 93.9–99.6, I² = 0%), with AUROC of 0.96 (95% CI: 0.93–0.98). Tissue adequacy was comparable in both groups (90.4% vs. 91.7%). Subgroup analyses showed that the stylet slow-pull technique yielded the highest sensitivities in both analyses. Among needle types, Franseen needles provided the best sensitivity in histological analysis, whereas side-fenestrated needles yielded the highest sensitivities in combined analysis. Conclusions: EUS-FNB with histological analysis shows diagnostic performance and tissue adequacy comparable to combined analysis. While histological assessment may serve as a feasible standalone approach, these findings require validation through future randomized controlled trials.