BACKGROUND: The purpose of this study was to determine interobserver variability, the relative importance of cytologic preparations, and factors influencing the diagnostic yield of mediastinal lymph node aspirates in suspected sarcoidosis. METHODS: Analysis of mediastinal lymph node aspirates obtained during a randomized study of endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) and conventional TBNA in patients with suspected sarcoidosis. All aspirates were processed by the monolayer ThinPrep method and cell blocks were prepared when feasible. The slides were reviewed by 2 cytopathologists, blinded to the study group. Lymph node location, size, number of aspirates, and diagnosis were recorded. A research cytopathologist recorded the presence of noncaseating granulomas in the monolayer and cell block preparations separately. RESULTS:Fifty patients were enrolled in the study, with 152 lymph nodes available for analysis (95 EBUS-TBNA, 57 conventional TBNA). The overall diagnostic yield was 64.5% on a per-lymph node basis. There was good agreement between cytopathologists (κ=0.677, P<0.001). Both liquid-based cytology and cell blocks were important in identifying granulomatous inflammation, with each being solely positive in 22% of cases. The diagnostic yield was not altered by lymph node location, size, or number of aspirates per node. CONCLUSIONS: Liquid-based cytology and cell block specimens are equally important in maximizing the diagnostic yield in EBUS-guided and conventional TBNA in suspected sarcoidosis. Good interobserver agreement between cytopathologists was noted, with improved diagnostic yield after review by a pulmonary cytopathologist. None of the clinical factors assessed impacted on the diagnostic yield of the procedure on a per-node basis.
RCT Entities:
BACKGROUND: The purpose of this study was to determine interobserver variability, the relative importance of cytologic preparations, and factors influencing the diagnostic yield of mediastinal lymph node aspirates in suspected sarcoidosis. METHODS: Analysis of mediastinal lymph node aspirates obtained during a randomized study of endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) and conventional TBNA in patients with suspected sarcoidosis. All aspirates were processed by the monolayer ThinPrep method and cell blocks were prepared when feasible. The slides were reviewed by 2 cytopathologists, blinded to the study group. Lymph node location, size, number of aspirates, and diagnosis were recorded. A research cytopathologist recorded the presence of noncaseating granulomas in the monolayer and cell block preparations separately. RESULTS: Fifty patients were enrolled in the study, with 152 lymph nodes available for analysis (95 EBUS-TBNA, 57 conventional TBNA). The overall diagnostic yield was 64.5% on a per-lymph node basis. There was good agreement between cytopathologists (κ=0.677, P<0.001). Both liquid-based cytology and cell blocks were important in identifying granulomatous inflammation, with each being solely positive in 22% of cases. The diagnostic yield was not altered by lymph node location, size, or number of aspirates per node. CONCLUSIONS: Liquid-based cytology and cell block specimens are equally important in maximizing the diagnostic yield in EBUS-guided and conventional TBNA in suspected sarcoidosis. Good interobserver agreement between cytopathologists was noted, with improved diagnostic yield after review by a pulmonary cytopathologist. None of the clinical factors assessed impacted on the diagnostic yield of the procedure on a per-node basis.