Determining Factors in Diagnosing Pulmonary Sarcoidosis by Endobronchial Ultrasound-Guided Transbronchial Needle Aspiration

Abstract

Although the role of endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA) in pulmonary sarcoidosis has previously been investigated, the determining factors in diagnosing sarcoidosis by EBUS-TBNA without rapid on-site evaluation (ROSE) are unclear. Patients with clinically and radiographically suspected sarcoidosis underwent EBUS-TBNA without ROSE in a prospective study. Presence of non-caseating epithelioid cell granulomas was pathologic evidence of sarcoidosis. The EBUS-TBNA was performed in 120 patients, 111 of whom had confirmed sarcoidosis. For the patients with sarcoidosis (62 stage I, 49 stage II) EBUS-TBNA provided sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 93.69%, 100%, 100%, 56.25%, and 94.17%, respectively, in the diagnosis of sarcoidosis. Diagnostic yield of EBUS-TBNA for sarcoidosis was associated with disease stage, but not associated with serum angiotensin converting enzyme level, number of lymph node stations sampled per patient, or total number of passes performed per patient. At EBUS-TBNA, 284 mediastinal and hilar lymph nodes were aspirated in 111 patients. Multivariate logistic regression revealed that short-axis diameter and more than 1 needle pass per lymph node were independent risk factors associated with positive pathology. No major procedure-related complications were observed. Endobronchial ultrasound-guided transbronchial needle aspiration is a safe procedure with high sensitivity for diagnosing sarcoidosis, having a higher diagnostic yield in stage I than stage II. To obtain a higher diagnostic yield of EBUS-TBNA in pulmonary sarcoidosis without ROSE, operators should select the largest mediastinal or hilar lymph node accessible and puncture with 3 to 5 passes.