Background: EUS-FNA often fails to make a definitive diagnosis in the evaluation of subepithelial lesions. The addition of jumbo biopsy forceps has the potential to improve diagnostic yield, but published series are limited.
Objective: To assess the likelihood of definitive diagnosis for subepithelial lesions by using jumbo biopsy forceps during EUS examination.
Design: Pooled retrospective analysis.
Setting: 6 tertiary referral centers.
Patients: All patients having undergone EUS examination for a subepithelial lesion in which jumbo biopsy forceps were used for tissue acquisition.
Main outcome measurements: Diagnostic yield of jumbo biopsy forceps use, complication rates, and comparison of diagnostic yield with that of EUS-FNA.
Results: A total of 129 patients underwent EUS with jumbo biopsy forceps; 31 patients (24%) had simultaneous EUS-FNA. The lesion locations were stomach (n = 98), esophagus (n = 14), duodenum (n = 11), colon (n = 5), and jejunum (n = 1). The average lesion size was 14.9 mm ± 9.3 mm. Overall, definitive diagnosis was obtained in 87 of 129 patients (67.4%) by using either method. A definitive diagnosis was provided by jumbo biopsy forceps use in 76 of 129 patients (58.9%) and by FNA in 14 of 31 patients (45.1%) (P = .175). The results in third-layer lesions were definitive with jumbo biopsy forceps in 56 of 86 lesions (65.1%) and with FNA in 6 of 16 lesions (37.5%) (P = .047). For fourth-layer lesions, the results with jumbo biopsy forceps were definitive in 10 of 25 (40.0%) and with FNA in 8 of 14 (57.1%) (P = .330). Forty-five of 129 patients (34.9%) experienced significant bleeding after biopsy with jumbo forceps and required some form of endoscopic hemostasis.
Limitations: Retrospective study.
Conclusions: Jumbo forceps are a useful tool for the definitive diagnosis of subepithelial lesions. The greatest benefit appears to be with third-layer (submucosal) lesions. The risk of bleeding is significant.
Copyright © 2012 American Society for Gastrointestinal Endoscopy. Published by Mosby, Inc. All rights reserved.