BACKGROUND: Mediastinoscopy is the gold standard for preoperative mediastinal staging of patients with suspected or proven lung cancer. Since the development of endoscopic ultrasound-guided biopsy via the trachea (EBUS-TBNA), this status has been challenged. The purpose of the study was to examine whether mediastinoscopy is necessary, when EBUS-TBNA is performed in a center with (1) a high level of expertise, (2) "bed side" microscopy by a pathologist, (3) general anesthesia, and (4) achievement of representative tissue from station 4R, 7 and 4L, that is, the same mediastinal stations that mediastinoscopy gives access to.

METHODS: A total of 95 consecutive patients with known or suspected lung cancer were referred for staging by EBUS-TBNA, which was performed as described.

RESULTS: Benign and malignant disease was found in the mediastinum of 6 and 13 patients, respectively. The remaining 76 patients were operated, resulting in 9 benign and 67 malignant diagnoses; spread was found to station 4R, 5, and 5 and 6 in 4 patients. The negative predictive value (NPV) was 63/67=0.94. However, if you exclude station 5 and 6, as they cannot be reached by neither EBUS nor mediastinoscopy, NPV was 66/67=0.99. The sensitivity was 0.76, and the specificity was 1.0.

CONCLUSIONS: When EBUS-TBNA is performed under optimal conditions including general anesthesia and "bed side" microscopy performed by a pathologist resulting in representative biopsies from station 4R, 7, and 4L, the NPV is so high that mediastinoscopy seems unnecessary.