Methods: A total of 239 patients (153 males, 86 females; mean age 53.3±13.7 years; range 14 to 85 years) who were diagnosed with or suspected to have lung cancer or mediastinal lymphadenopathy between January 2004 and December 2012 were performed transbronchial needle aspiration rapid on-site cytologic evaluation. Diagnosis success, localization and size of lymph nodes, and number of sampling were analyzed.

Results: Diagnostic specimens were obtained from a total of 227 patients (93.03%) and 297 lymph nodes were sampled. A statistically significant difference was detected between lymph node stations and dimensions (p=0.018). Subcarinal lymph nodes had the highest diagnosis rate (91.5%) and 49.5% of these were benign. Mean lymph node diameters were 23.9±10.4 mm in malignant group (n=123), 19.4±7.4 mm in benign group (n=135), and 19.1±8.8 mm in undiagnosed group (n=39) and the diameters were statistically significantly different from each other (p=0.001). A repeat transbronchial needle aspiration was performed in five patients and no metastasis was detected in three of them. All negative results (n=12) were confirmed with histopathological examination after resection.

Conclusion: On-site evaluation decreases the risk of inadequate specimens and increases diagnostic accuracy rate of transbronchial needle aspiration. Transbronchial needle aspiration may be the initial procedure of choice while staging lung cancer or diagnosing mediastinal lymph node. Procedure may be performed by thoracic surgeons in the operating room in patients planned for mediastinoscopy.