Performance of Endobronchial Ultrasound-Guided Transbronchial Needle Aspiration for the Diagnosis of Isolated Mediastinal and Hilar Lymphadenopathy

Abstract

Background: Although many studies have assessed the diagnostic utility of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in the context of a specific disease, few studies have assessed the overall diagnostic yield, sensitivity, and negative predictive value in patients with isolated mediastinal and hilar lymphadenopathy (IMHL). Objective: We evaluated the performance of EBUS-TBNA for diagnosing IMHL in a population with a high prevalence of concurrent or preexisting non-pulmonary malignancy. Methods: A retrospective chart review of patients who underwent EBUS-TBNA from October 2008 to April 2014 was performed to identify patients with IMHL. Patients with known or suspected primary pulmonary malignancy were excluded. When available, EBUS-TBNA results were cross-referenced with further diagnostic investigation or clinical diagnosis based on follow-up. Results: EBUS-TBNA was used to sample 765 lymph nodes from 350 patients. One hundred and fourteen (33.3%) patients had a concurrent or preexisting non-pulmonary malignancy. The overall yield of EBUS-TBNA for specific diagnosis was 300/350 (86%). The diagnostic yield for sarcoidosis, lymphoproliferative disease, metastatic lymphadenopathy from extrathoracic malignancy, and necrotizing granuloma was 123/149 (83%), 27/33 (82%), 20/25 (80%), and 13/19 (68%), respectively. Amongst 50 patients with non-diagnostic EBUS-TBNA, 25 yielded an insufficient sample and another 25 yielded only benign lymphoid material which was not representative of the underlying pathology. Overall, EBUS-TBNA had a sensitivity of 89%, a diagnostic yield of 86%, and a negative predictive value of 79%. Conclusion: For patients with isolated hilar or mediastinal lymphadenopathy and a high background prevalence of concurrent and preexisting non-pulmonary malignancy, EBUS-TBNA is a reliable first-line diagnostic investigation.