Adequacy criteria of rapid on-site evaluation for endobronchial ultrasound-guided transbronchial needle aspiration: a simple algorithm to assess the adequacy of ROSE

Abstract

Background. Rapid on-site evaluation (ROSE) for endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has been advocated to qualitatively diagnose biopsy samples. However, adequate ROSE criteria during EBUS-TBNA are unclear. The aim of this study was to determine adequacy criteria of ROSE in EBUS-TBNA samples and suggest an appropriate algorithm. Methods. Patients who underwent EBUS-TBNA for nodal evaluation between March and July 2013 at Seoul National University Hospital were included prospectively. The ROSE slides were reviewed independently by two pathologists, and the results were compared to the final pathologic results. Diagnostic yields, sensitivity, specificity, and accuracy were calculated in order to make nodal evaluations. Results. EBUS-TBNA was performed on 300 lymph nodes in 133 patients. Samples were nondiagnostic in 7.7%, 6.3%, and 1.7% of the cytologic, histologic, and overall pathologic results, respectively. On the ROSE slides, a large tissue core size (>= 2 cm), microscopic anthracotic pigment (MAP), and increased lymphocyte density (LD; >= 40 cells/field [403, mean of 10 fields]) were significantly associated with adequate final cytologic or histologic results. Malignant cells were not statistically associated with adequacy but were considered a parameter indicating an adequate diagnosis. Using four sequential criteria, tissue core size, the presence of malignant cell, MAP, and LD >= 40 cells/field, the sensitivity and accuracy rates of ROSE increased from 64.4% to 98.6% and from 64.7% to 97.3%, respectively. Conclusions. A high adequacy rate of ROSE in EBUS-TBNA can be achieved by sequentially applying four criteria: tissue core size, malignant cells, MAP, and increased LD. (C) 2016 by The Society of Thoracic Surgeons