Computed tomography-guided transthoracic needle aspiration in diagnosing and staging hilar and mediastinal masses of lung cancer after negative bronchoscopy

This study targeted to assess the efficiency and safety of computed tomography (CT)-guided transthoracic needle aspiration (TTNA) in diagnosing and/or staging hilar and mediastinal masses and lymph nodes (LNs) suspected of lung cancer (LC). Over a period of 10 years, of 4460 diagnostic or staging flexible bronchoscopies (FBs) for suspected LC, 892 (20%) were negative despite bronchial washings, transbronchial needle aspiration, endobronchial, transbronchial lung, and/or brush biopsies. In 892, 22-gauge TTNA was attempted under contrast-enhanced CT guidance to diagnose lone peripheral pulmonary lesions (274), hilar or mediastinal masses (277) and hilar or mediastinal LNs (341) of 10 mm or greater short axis diameter. A total of 618 with hilar or mediastinal lesions were included in this prospective study. Overall, 500 (81%) TTNAs were diagnostic while the staging and diagnosing yield was 77% in 341 LNs, and the diagnostic yield 86% in 277 masses. Four hundred eighty-six cases of LC, 8 of lymphoma, 3 of tuberculosis, and 3 of metastatic extrathoracic malignancy were diagnosed by TTNA. Mediastinoscopy and/or thoracotomy corroborated LC in all of 158 cases TTNA-positive for LC. Of the 64 TTNA-negative cases, 30 were proven by surgical means to have LC, and 34 to have tuberculous lymphadenitis (14), sarcoidosis (10), lymphoma (7), mediastinal fibrosis (2), and Castleman disease (1). Surgical results used as a reference, sensitivity, accuracy, positive and negative predictive values of TTNA for LC were 84%, 86%, 100%, and 53% overall whereas they were 87%, 89%, 100%, and 53% for the masses, and 82%, 85%, 100%, and 53% for the LNs, respectively. During the follow-up of 396 nonsurgical cases, 229 cases TTNA-positive for LC and 48 TTNA-negative were proven to have LC whereas 14 TTNA-positive for benign disease and 6 TTNA-negative were proven to have non-LC disease by various biopsies. No false-positive results were obtained. In the remaining 99 positive for LC by TTNA, clinical follow-up suggested LC. Pneumothorax occurred in 74 (12%) cases (chest tubes required in 19), and minimal bleeding in 19 (3%). These results confirm that CT-guided TTNA is an efficient and safe method in diagnosing and staging FB-negative LC with hilar or mediastinal location and thus, is complementary to FB. © 2007 Lippincott Williams & Wilkins, Inc.