The safety and efficacy of mediastinoscopy when performed by general thoracic surgeons

Previous publications suggest that mediastinoscopy only obtains a biopsy of lymph node tissue in about 50% of patients; however, those data included results from nonthoracic surgeons.

A retrospective cohort study was performed using a database of a consecutive series of patients who underwent mediastinoscopy or video mediastinoscopy by general thoracic surgeons only.

Between January 1997 and September 2013, 1,970 patients underwent mediastinoscopy (video mediastinoscopy in the last 243). The indications were staging for known or suspected lung cancer in 68.5%. Morbidity occurred in 25 patients (1.3%). Significant bleeding occurred in 5 patients (0.25%): 2 patients required sternotomy, and bleeding in the other 3 was controlled with packing alone. No patients required transfusion. There were no 30-day operative deaths. Median operative time was 18 minutes, and 96.1% of operations were performed as outpatient procedures. Lymph node tissue was obtained from all patients, and biopsy specimens from at least two mediastinal stations were obtained for 98% who had non-small cell lung cancer. The false-negative rate for N2 lymph nodes that were accessible by mediastinoscopy was 8.2% when lymph nodes dissected at the time of pulmonary resection were used as the reference standard.

In the hands of general thoracic surgeons mediastinoscopy provides lymph node tissue from multiple stations essentially 100% of the time; has minimal morbidity and essentially no deaths; and is a short outpatient procedure. Specialty-specific data (and not national databases) should be used when the efficacy of mediastinoscopy is compared with endobronchial ultrasound.