Abstract
BACKGROUND: A number of reports of radiologically guided percutaneous biopsy of mediastinal masses have been described but techniques have varied, particularly the type of needle used. In this study mediastinal biopsies with fine aspiration needles and cutting needles have been compared, sometimes in the same patient. The results are reviewed with particular emphasis on the choice of biopsy needle and its influence on pathological diagnosis. METHODS: A retrospective review was undertaken of radiologically guided mediastinal biopsies performed between 1981 and 1991. RESULTS: Sixty fine needle aspiration biopsies (FNA) and 34 Tru-Cut biopsies of mediastinal masses were performed in 75 patients with fluoroscopic or computed tomographic guidance. Overall sensitivity and specificity in terms of diagnosis of malignant disease were 90% and 100% respectively for FNA biopsies, and 96% and 100% for Tru-Cut biopsies. Diagnostic accuracy in terms of precise diagnosis of the malignant or benign nature of a mass and its origin was 77% for FNA biopsies and 94% for Tru-Cut biopsies. For FNA biopsies sensitivity and accuracy were higher for carcinomatous lesions (96% and 88%) than for noncarcinomatous lesions (81% and 69%). The only significant complication encountered was a pneumothorax following a biopsy which required intercostal drainage. CONCLUSIONS: Radiologically guided percutaneous needle biopsy is a safe procedure which provides useful diagnostic information in the majority of cases. Fine needle aspiration techniques usually suffice for carcinomatous lesions but a cutting needle biopsy should be performed whenever possible when lymphoma, thymoma, or neural masses are suspected to obtain larger specimens for more accurate histological diagnosis.