Background: At present only few studies directly compare the diagnostic yield of endobronchial ultrasound guided fine needle aspiration (EBUS-FNA) and transcervical video-assisted mediastinoscopy (TM) for mediastinal lymph node staging in patients with NSCLC. If and when EBUS-FNA may replace TM as Gold Standard remains controversial.
Methods: From April 2008 to December 2009, 36 patients with mediastinal lymphadenopathy underwent simultaneous EBUS-FNA/ TM at our institution. Among them were 26 patients with confirmed or suspected NSCLC.
Results: A total of 133 samples were obtained by EBUS-FNA and 157 samples by TM. EBUS-FNA achieved significantly less conclusive, but more indeterminate pathological results in comparison to TM (78.7% vs. 98.6%, p < 0.001; 14.9% vs. 1.4%, p = 0.007). Less paratracheal nodes were sampled by EBUS-FNA (right: 46.2% vs. 88.5%, p = 0.003; left: 23.1% vs. 65.4%, p = 0.005), while sampling rates in the subcarinal localisation were comparable (96.2% vs. 80.8%, p = NS). Among patients with confirmed NSCLC and conclusive EBUS-FNA/ TM findings (n = 18), the prevalence of N2/N3 disease was 66.7% (n = 12) according to TM findings. Diverging nodal stages were found in five patients (27.8%). Three patients who were N2 negative in EBUS-FNA were upstaged to N2 or N3 by TM, two patients with N2 status in EBUS-FNA were upstaged to N3 by TM.
Conclusions: Compared to TM, EBUS-FNA had a lower diagnostic yield and resulted in systematic mediastinal nodal understaging. At this point we suggest corroborating negative EBUS-FNA results by transcervical mediastinoscopy.