Correlation of final pathology with the ROSE used during the EBUS-TBNA

Abstract

Introduction: When using EBUS-TBNA, the use of rapid onsite evaluation(ROSE) reduces the overall cost by reducing the duration of operation and the amount of unnecessary biopsy. The aim of this study was to determine correlation of ROSE used during EBUS-TBNA with the final pathologic diagnosis and the cellular subgroup in malignant group. Methods: This prospective study involved 449 lymphadenomegaly (LAM)/mass lesion belonging to 219 consecutive patients who experienced EBUS-TBNA under ROSE. Patient demographic data, morphologic features of mediastinal/hilar lesions (lymphadenopathy/mass), diameter, number of samples, ROSE results, and final pathology results were recorded. Comparisons of ROSE and final pathology results were compared. Results: In total of 219 patients, 133 were male(60.7%). The mean age was 62±14 years. The mean diameter of the sampled LAM/masses was 14.8±1.0mm. The mean number of needle pass in a LAM was 3.1±1.3 and the mean number of pass leading to a ROSE diagnosis was 1.5±0.9. According to ROSE results, 54.3% of the lesions were lymphoid material/reactive, 4.5% were inadequate, 5.8% were granulomatous inflammation, and the remaining 35.4% were malignant pathology. As a result of the definite pathology, 2.4% of lesions were inadequate, 58.1% were benign pathologies, and 39.4% were malignant. The diagnostic success rate for ROSE in malignancies was 93.5% and false negativity was 4%. Regarding whether ROSE detected the malignant cell type/subtype assignment correctly in 145 lesions reported as lung cancer in final pathology, ROSE was correct 64.1%. Conclusion: Despite the high correlation of ROSE with the final pathology, ROSE could moderately predict tm subtype/cell type in lung ca.