Transbronchial Needle Aspiration Versus Transthoracic Needle Aspiration in the Diagnosis of Pulmonary Lesions

Abstract

Abstract In a prospective study (1989–1992) of pulmonary nodules/masses and/or mediastinal/hilar lesions, 329 patients underwent transbronchial needle aspiration (TBNA) using a cytology or a histology needle with or without fluoroscopy via flexible bronchoscopy as the initial procedure to sample mediastinal lesions (169) or lung lesions (160). TBNA established the diagnosis of malignant or benign disease in 68.1% (224 patients). The diagnostic yield was 89.3% (151/169) for mediastinal and 45.6% (73/160) for lung lesions without abnormal mediastinal/hilar areas. The undiagnosed patients (105) underwent transthoracic needle aspiration (TTNA) using a 19-gauge needle under fluoroscopy or computed tomography guidance. TTNA established the diagnosis in 69.5% (73). The diagnostic yield was 83.3% (15/18) for mediastinal lesions and 66.7% (58/87) for lung lesions. Benign disease was diagnosed by TBNA in 8.8% (20/224) and by TTNA in 23% (17/73). The benign disease most commonly diagnosed by TBNA was sarcoidosis, and TTNA was most useful in diagnosing various benign lung lesions. Neither TBNA nor TTNA established the diagnosis in 9.7% (32). Overall diagnostic rates for TBNA and TTNA were 68.1 and 22.2%, respectively. The combined diagnostic rate was 90.3%. There were no complications from TBNA but TTNA was complicated by pneumothorax in 20%. We suggest that TBNA should be the diagnostic procedure of choice in patients with mediastinal/hilar lesions. BNA and TTNA are complementary procedures.