View Abstract

A7299 - Diagnostic Utility of EBUS-TBNA for Evaluating Mediastinal Adenopathy in Patients with Extrathoracic Malignancy and Prognostic Implication of Granulomatous Inflammation
Author Block: M. Lane1, Q. M. Fasihuddin2, C. Ohri3, J. S. Wang Memoli1; 1Division of Pulmonary, Critical Care and Respiratory Services, MedStar Washington Hospital Center, Washington, DC, United States, 2Pulmonary Medicine, Mid-Atlantic Permanente Medical Group, Baltimore, MD, United States, 3Department of Internal Medicine, MedStar Washington Hospital Center, Washington, DC, United States.
Rationale: Intrathoracic lymphadenopathy in patients with history of extrathoracic malignancy is common and raises concern for metastases. Data on the diagnostic performance of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in this population are limited. Furthermore, the significance of finding granulomatous inflammation is unclear.
Objectives: We sought to evaluate the diagnostic performance of EBUS-TBNA in patients with mediastinal lymphadenopathy and extrathoracic malignancy, as well as the frequency and outcomes associated with granulomatous inflammation.
Methods: We performed a retrospective review of 670 patients having undergone EBUS-TBNA from August 2010 through June 2016 at our center. Eighty-nine patients had a history of extrathoracic malignancy and suspected mediastinal metastases based on radiography. Patients with non-malignant findings underwent further investigation or were followed radiographically for at least six months. Longer-term outcomes based on chart review were then analyzed.
Results: Malignant intrathoracic lymphadenopathy was found in 31 patients (35%) (six (7%) of these were new or separate lung cancers). Findings were non-diagnostic or benign in 58 (65%). Sixteen (18%) were false negatives. The sensitivity, negative predictive value, and accuracy for diagnosing malignancy in this population were 66%, 69%, and 81%, respectively. The most common extrathoracic malignancies with mediastinal metastases were head-and-neck (27%), breast (20%), and colorectal (12%). Sixteen patients (18%) in our sample had granulomatous inflammation; one of these was diagnosed with mediastinal metastases based on clinical and radiographic follow-up. Ten (62.5%) patients with granulomatous inflammation were true negatives; of those, three (30%) had malignancy progression outside the mediastinum with median follow-up of 18 months. There was no difference in the rate of malignancy progression between those with granulomatous inflammation and those with other benign findings (36% vs 23%, p=0.2). Patients with granulomatous inflammation were more likely to have a history of breast cancer (50% vs 16%, p=0.002).
Conclusion: When used to evaluate suspected mediastinal metastases in patients with history of extrathoracic malignancy, EBUS-TBNA has lower sensitivity and negative predictive value compared to its use in staging lung cancer (reported diagnostic yield of 90%). Although EBUS-TBNA may be a first diagnostic modality, benign results may warrant further investigation because of the high false negative rate in this population. Granulomatous inflammation did not have prognostic implications regarding disease progression or recurrence in our study. Larger studies are needed to confirm these findings and establish further recommendations in this population.