• Li-Han Hsu, Jen-Sheng Ko, Dong-Ling You, Chia-Chuan Liu, and Nei-Min Chu.
• Division of Pulmonary and Critical Care Medicine, Koo Foundation Sun Yat-Sen Cancer Center, Taipei, Taiwan. lhhsu@mail.kfcc.org.tw
• Respirology. 2007 Nov 1; 12 (6): 848-55.

Objective And BackgroundIntegrated PET and CT (PET/CT) is accurate in detecting hilar-mediastinal metastases. However, it has a moderate positive predictive value, necessitating pathological verification, especially in situations in which the result would make a difference to treatment. This study aimed to evaluate the performance of transbronchial needle aspiration (TBNA) for hilar-mediastinal lesions suspicious on PET/CT.MethodologyA retrospective study was conducted on 19 patients with a total of 25 positive hilar-mediastinal lymph nodes localized on PET/CT. Standard TBNA technique with rapid on-site cytopathology was performed.ResultsThe mean short-axis diameter of the positive lymph nodes identified on PET/CT was 9.9 +/- 3.0 mm. The sensitivity, specificity and diagnostic accuracy of PET/CT-guided TBNA were 81.8%, 100% and 84%, respectively. The number of needle passes to successful lymph node aspiration or a diagnosis of cancer was 2.36 +/- 0.49. Nine of the 25 positive lymph nodes (36%) on PET/CT were smaller than 1.0 cm. The accuracy and sensitivity of TBNA for these subcentimetre nodes was 88.9% and 87.5%, respectively. TBNA replaced surgical sampling in 15 patients (78.9%) with positive lymph nodes on PET/CT. In seven non-small cell lung cancer patients, diagnosis and staging were possible in the one procedure. No complications were encountered.ConclusionPET/CT can identify small malignant lymph nodes that can then be successfully biopsied by TBNA with on-site cytopathology.

23 keywords...

hide…