Background and objective: EUS-guided FNA is a sensitive method to obtain cytologic specimens from solid lesions in close proximity to the GI tract. Although FNA provides cells for analysis, large-caliber Tru-cut biopsy (EUS-TCB) needles obtain samples that can be used for additional histopathologic analysis. We assessed the additional diagnostic yield of EUS-TCB in patients with solid mediastinal lesions in whom EUS-FNA was performed.
Patients and design: In the period from July 2003 to July 2007, all patients with mediastinal lesions accessible to EUS-FNA and EUS-TCB were evaluated. In all patients, a mean of 3 passes of EUS-FNA was followed by EUS-TCB. Cytologic and histologic specimens were evaluated by 2 pathologists blinded for patient condition. A final diagnosis was obtained by combining all information present (EUS-FNA and EUS-TCB results, mediastinoscopy, bronchoscopy [if performed], and other investigations).
Results: The diagnostic accuracy of EUS-FNA, EUS-TCB, and the combination of both techniques was 93%, 90%, and 98%, respectively (not significant). In EUS-FNA-negative patients, EUS-TCB provided a final diagnosis in an additional 3 patients (5%). Malignant disease found by EUS-FNA could be specified by EUS-TCB in 15 patients (25% of patients). The granulomatous disease established by cytologic samples of clinically suspected tuberculosis could be specified by EUS-TCB in 2 patients (3%). In 1 patient (2%), both FNA and TCB were inconclusive.
Limitations: Retrospective study.
Conclusions: The diagnostic yield of EUS-FNA and EUS-TCB is comparable. We recommend limiting the use of EUS-TCB to specific cases in which EUS-FNA is not conclusive.