Diagnostic efficacy of endoscopic ultrasound-guided needle sampling for upper gastrointestinal subepithelial lesions: a meta-analysis

Background: An increasing number of studies have been conducted on the use of endoscopic ultrasound (EUS)-guided needle sampling for upper gastrointestinal subepithelial lesions (SEL). However, reported diagnostic efficacy varies greatly. Objective: To summarize up current evidences on the diagnostic efficacy of EUS-guided needle sampling for upper GI SEL. Method: A reproducible strategy was used to search four databases. Search results were evaluated for eligibility, and the quality of eligible studies was assessed by QUADAS-2. Pooled efficacy of EUS-guided needle sampling in upper GI SEL was calculated. Procedure-related complications, diagnostic errors, and independent factors related to a higher success rate were also recorded and analyzed. Results: Seventeen studies, comprising 978 attempts of EUS-guided needle sampling, were included in a meta-analysis. Pooled diagnostic rate of EUS-guided needle sampling was 59.9 %, with a heterogeneity I² of 55.2 %. Subgroup analysis showed no difference in diagnostic rate among fine needle aspiration (FNA), trucut needle biopsy (TCB), and fine needle biopsy (FNB), or among 19-, 22-, and 25-G needles. Subgroup analysis and meta-regression suggested that the cell block method might be correlated with a higher diagnostic rate. Few severe complications were reported. Diagnosis errors were rare. Conclusion: EUS-guided needle sampling is a safe, but only moderately effective method for pathology diagnosis of upper GI SEL. Choice of FNA/TCB/FNB, or 19 G/22 G/25 G does not seem to alter the overall diagnostic rate.