Utility of endobronchial ultrasound–guided mediastinal lymph node biopsy in patients with non–small cell lung cancer

Abstract

Invasive mediastinal biopsy is often necessary in the evaluation of non-small cell lung cancer (NSCLC), and mediastinoscopy has long been considered the reference standard. However, the emergence of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has resulted in controversy regarding it represents a suitable replacement for mediastinoscopy. We chose to determine the utility of EBUS-TBNA in evaluating the mediastinum in patients with NSCLC. The present study was a retrospective review of a prospective database of consecutive patients with NSCLC who underwent EBUS-TBNA for mediastinal evaluation from 2009 to 2011. The sensitivity, specificity, negative predictive value, and accuracy of EBUS-TBNA are reported. Also reported are the size of the lymph nodes biopsied and the number of instances in which EBUS-TBNA obviated the need for cervical mediastinoscopy. A total of 73 patients had a total of 140 mediastinal stations biopsied using EBUS-TBNA. Of the 73 patients, 30 had benign findings and underwent surgical resection, 1 of whom was found to have stage N2 disease. Of the remaining patients, 42 had a positive result and 1 had nondiagnostic biopsy findings for which malignancy was confirmed by mediastinoscopy. Mediastinoscopy would have changed the tumor stage and treatment planning in only 2 (2.7%) of the 73 patients. Overall, EBUS-TBNA had a sensitivity of 95%, a specificity of 100%, a negative predictive value of 94%, and an accuracy of 97%. EBUS-TBNA might be a feasible option for most patients with NSCLC for whom histologic assessment of the mediastinum is necessary. The rates of nondiagnostic and false-negative biopsy findings using EBUS-TBNA were low, small subcentimeter nodes could be routinely biopsied, and most patients with a radiographically positive mediastinum had their disease pathologically confirmed.