Algorithmic approach by endobronchial ultrasound-guided transbronchial needle aspiration for isolated intrathoracic lymphadenopathy: A study in a tuberculosis-endemic country

• Main Authors: Chih-Hsi Kuo, Shu-Min Lin, Kang-Yun Lee, Fu-Tsai Chung, Po-Hao Feng, Te-Chih Hsiung, Yu-Lun Lo, Chien-Ying Liu, Han-Pin Kuo
• Format: Article
• Language: English
• Published: Elsevier 2014-08-01
• Series: Journal of the Formosan Medical Association
• Subjects: EBUS-TBNA; intrathoracic lymphadenopathy; malignancy; tuberculosis
• Online Access: http://www.sciencedirect.com/science/article/pii/S0929664613002143

Isolated intrathoracic lymphadenopathy (IT-LAP) is clinically challenging because of the difficult anatomic location and wide range of associated diseases, including tuberculosis (TB). Although sampling via endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) for histopathology is a major development, there is still room for improvement. This study aimed to investigate an algorithmic approach driven by EBUS-TBNA and conventional bronchoscopy to streamline the management of IT-LAP. Methods: Eighty-three prospectively enrolled patients with IT-LAP were subjected to an EBUS-TBNA diagnostic panel test (histopathology, cytology, and microbiology) and underwent conventional bronchoscopy for bronchoalveolar lavage. The results were structured into an algorithmic approach to direct patient treatment, workup, or follow-up. Results: The diagnostic yields of EBUS-TBNA based on histopathology were similar for each disease entity: 77.8% for malignancy, 70.0% for TB, 75.0% for sarcoidosis, 80.0% for anthracosis, and 70.0% for lymphoid hyperplasia (p = 0.96). The incidence of malignancy was 10.8% for total IT-LAP patients, and 12.0% and 33.7% for patients with TB and sarcoidosis, respectively. Thirty-five (42.2%) patients were symptomatic. The leading diagnosis was sarcoidosis (60%), followed by TB (20%), malignancy (11.4%), lymphoid hyperplasia (5.7%), and anthracosis (2.9%). By logistic regression analysis, granulomatous disease (odds ratio: 13.45; 95% confidence interval: 4.45–40.67, p < 0.001) was an independent predictor of symptoms. Seven (8.4%) and three (3.6%) IT-LAP patients diagnosed active TB and suggestive of TB with household contact history, respectively, were all placed on anti-TB treatment. Conclusion: The algorithmic approach streamlines patient management. It enables early detection of malignancy, correctly places nonmalignant patients on an appropriate treatment regimen, and particularly identifies candidates at high risk of TB reactivation for anti-TB chemoprophylaxis.