Efficacy, safety, and predictive factors for a positive yield of EUS-guided Trucut biopsy: a large tertiary referral center experience.

OBJECTIVES: Endoscopic ultrasound-guided trucut biopsy (EUS-TCB) technique has the advantage of obtaining tissue for histological examination rather than for cytology alone. However, the diagnostic yield may depend on factors related to both technical aspects and the lesions sampled. Safety of EUS-TCB is yet to be established in a large number of procedures. The aim of the study was to determine factors predicting a positive diagnostic yield, and safety for EUS-TCB in a large tertiary referral center-based service.

METHODS: All patients were referred for EUS-guided tissue sampling as a part of their diagnostic workup. Linear-array echoendoscope (GF-2000-OL5, KeyMed) with a 19-gauge trucut needle (Quick-Core, Wilson-Cook) was used by two operators to obtain tissue samples. Clinical data, details of the EUS-TCB, post-procedure complications, and histology were prospectively collected between May 2002 and February 2008.

RESULTS: In total, 247 patients (143 men) aged 57-73 (median 66) had EUS-TCB performed. Lesions sampled were in the pancreas (113), esophagogastric wall (34), and extra-pancreatic areas (100) (lymph nodes: 52). The maximum diameter of the lesion/wall thickness ranged from 0.6 to 5.4 cm (median 3). One to five passes were made (median 3) to obtain tissue cores 2-18 mm (median 10) in length. The procedure failed in 6% of cases. The overall diagnostic accuracy was 75%. The overall complication rate was 2% (bronchopneumonia, minor hemoptysis, minor hematemesis, mucosal tear, retropharyngeal abscess) with no procedure-related deaths. Site of lesion (pancreatic vs. extra-pancreatic, P<0.032), site of biopsy (stomach vs. duodenum vs. esophagus, P<0.001), and number of passes (< or =2 vs.>2, P<0.013) were predictors of a positive diagnostic yield in univariate analysis. However, only the site of biopsy (P<0.001, 95% CI: 0.58-2.32) and number of passes (P=0.05) were independent predictors in multinominal logistic regression.

CONCLUSIONS: Diagnostic yield of EUS-TCB is higher when lesion is approached through the stomach and better when more than two passes were made. In this large series, the complication rate of 2% associated with EUS-TCB was similar to that reported with EUS-fine needle aspiration technique.