Endoscopic ultrasound-guided fine-needle aspiration for non-small cell lung cancer staging: A systematic review and metaanalysis.


Published February 2007


Endoscopic ultrasound (EUS)-guided fine-needle aspiration (FNA) is a minimally invasive alternative technique for mediastinal staging of non-small cell lung cancer. A metaanalysis was performed to estimate the diagnostic accuracy of EUS-FNA for staging mediastinal lymph nodes (N2/N3 disease) in patients with lung cancer.


Relevant studies were identified using Medline (1966 to November 2005), CINAHL, and citation indexing. Included studies used histology or adequate clinical follow-up (> 6 months) as the “gold standard,” and provided sufficient data for calculating sensitivity and specificity. Summary receiver operating characteristic curves metaanalysis was performed to estimate the pooled sensitivity and specificity.


In 18 eligible studies, EUS-FNA identified 83% of patients (95% confidence interval [CI], 78 to 87%) with positive mediastinal lymph nodes (pooled sensitivity) and 97% of patients (95% CI, 96 to 98%) with negative mediastinal lymph nodes (pooled specificity). In eight studies that were limited to patients who had abnormal mediastinal lymph nodes seen on CT scans, the sensitivity was 90% (95% CI, 84 to 94%) and the specificity was 97% (95% CI, 95 to 98%). In patients without abnormal mediastinal lymph nodes seen on CT scans (four studies), the pooled sensitivity was 58% (95% CI, 39 to 75%). Minor complications were reported in 10 cases (0.8%). There were no major complications.


EUS-FNA is a safe modality for the invasive staging of lung cancer that is highly sensitive when used to confirm metastasis to mediastinal lymph nodes seen on CT scans. In addition, among lung cancer patients with normal mediastinal adenopathy seen on CT scans, despite lower sensitivity, it has the potential to prevent unnecessary surgery in a large proportion of cases missed by CT scanning.


Micames CG, McCrory DC, Pavey DA, Jowell PS, Gress FG.

PMID: 17296659 [PubMed – indexed for MEDLINE]