View Abstract

A7326 - Negative Predictive Value of Endobronchial Ultrasound Guided Needle Aspiration of Mediastinal and Hilar Lymph Nodes
Author Block: M. Frohlich1, H. Wang2, C. Robitaille1, L. Sakr1; 1Medicine, Division of Pulmonary Diseases, Jewish General Hospital, Montreal, QC, Canada, 2Pathology, Jewish General Hospital, Montreal, QC, Canada.
Background: Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) of mediastinal and hilar lymph nodes (LNs) is performed for diagnosis and staging of lung cancer and malignancies which metastasize to the mediastinum and hilar areas. Finding malignant cells on LN aspiration alters staging, management and prognosis. On the other hand, the absence of malignant cells on LN aspiration allows down staging of a primary malignancy, or excluding evidence of recurrent or progressive disease. The negative predictive value (NPV) of EBUS-TBNA and its predictors is poorly understood. Objective: The purpose of this study is to assess the NPV of EBUS-TBNA of mediastinal and hilar LNs and to define potential clinical, radiological and pathological determinants. Methods: Chart review of all mediastinal and hilar LN specimens obtained by EBUS- TBNA between October 2012 and August 2017, was undertaken. Specimens showing absence of malignancy were included. Subsequent diagnosis of cancer was ascertained on pathology obtained from other tissue biopsy, mediastinoscopy, or surgical resection specimens, or from clinical follow-up. Results: 119 patients were included (58% male). Samples were reported as benign (69), non-necrotizing granulomas or histiocytic clusters suggestive of sarcoid (34), necrotizing granulomas (12) or non-diagnostic (4). Lymph node stations sampled by EBUS-TBNA were: 2R (5), 4R (56), 4L (18), 7 (78), 10R (4), 10L (2), 11R (52), and 11L (36). A cancer diagnosis was made in 32 patients, 31 of whom had EBUS-TBNA specimens reported as benign, and 1 reported non-diagnostic. Among those 32 patients, 8 had previous history of malignancy, 31 had a lung mass on CT (28 had a single mass, 3 had multiple masses), 10 had early stage disease (N0), 8 had confirmation of malignancy in LN stations not reached or assessed by EBUS (1 EUS, 5 surgery, 2 Mediastinoscopy), 3 had true false negative EBUS-TBNA specimens as similar LN stations were found negative on EBUS but positive on mediastinoscopy (2) or repeat EBUS-TBNA (1). Interestingly, none of the patients with sarcoid-like features on EBUS-TBNA developed cancer on follow-up. Patients without a cancer diagnosis had a median follow-up of 303 days. Conclusion: The NPV of EBUS-TBNA in our patient cohort is 97% when taking into account only LN stations reached and/or sampled by EBUS-TBNA, and 73% when taking into account all LN stations whether reached/sampled or not by EBUS-TBNA. Interestingly, none of the patients with sarcoid features on EBUS-TBNA developed cancer. Further analysis to determine potential determinants of NPV of EBUS-TBNA is undergoing.