PURPOSE: Endobronchial ultrasound (EBUS) has contributed to improved diagnostic yield over conventional, non-ultrasound-guided transbronchial needle aspiration(TBNA) of mediastinal lymph nodes. Processing and interpretation of samples may also have an effect on diagnostic yield. Our aim was to determine inter-observer variability, relative importance of cytological preparations, and other factors influencing the diagnostic yield of TBNA in suspected sarcoidosis.
METHODS: We analyzed mediastinal lymph node aspirates obtained during a randomized study of EBUS-TBNA versus conventional TBNA in patients with suspected sarcoidosis. Visible tissue fragments from the aspirates were processed as paraffin-embedded cell blocks and the remaining liquid samples were processed by the ThinPrep technique (Cytyc Corp., Boxborough, Massachusetts, U.S.A.). Each sample was reviewed by two cytopathologists, blinded to the procedure group. Lymph node location, size, number of aspirates, and diagnosis were recorded. A research cytopathologist recorded the presence of non-caseating granulomas on cytology and cell block preparations independently
RESULTS: 50 patients were enrolled and 152 lymph nodes available for analysis (95 EBUS-TBNA, 57 conventional TBNA). Overall diagnostic yield was 64.5% and there was no difference in yield per-lymph-node between EBUS-TBNA and conventional TBNA, although twice as many needle passes per node were performed in the conventional TBNA group. There was good diagnostic agreement between cytopathologists (k=0.677, p< 0.001). Both liquid-based cytology and cell blocks contributed to identifying granulomas, with each being solely positive in 22% of nodes. Sample review by the research cytopathologist significantly increased the diagnostic yield over the initial interpretation (64.5% vs 54.6%, p<0.001). The diagnostic yield was not affected by lymph node location, size, or number of aspirates per node.
CONCLUSIONS: Both liquid-based cytology and cell block processing are complimentary in maximizing diagnostic yield in EBUS-guided and conventional TBNA in suspected sarcoidosis. Good inter-observer agreement between cytopathologists was noted but a review by a lung cytopathologist further improved diagnostic yield. None of the clinical factors assessed impacted on the diagnostic yield of the procedure on a per node basis.
CLINICAL IMPLICATIONS: Tissue sample processing and interpretation are important factors in the evaluation of TBNA samples in sarcoidosis.
DISCLOSURE: The following authors have nothing to disclose: Alex Chee, Moosa Khalil, David Stather, Paul MacEachern, Stephen Field, Alain Tremblay
No Product/Research Disclosure Information