Rapid on-site evaluation and final cytologic diagnoses correlation during endobronchial ultrasonography

Abstract

Background: The cost-effectiveness and low rate of inadequate sampling with the use of rapid on-site evaluation (ROSE) along with endobronchial ultrasonography (EBUS) is well established. Our aim in this study was to evaluate the correlation of ROSE during EBUS and final cytologic diagnosis and also to see if ROSE might predict the subtype of lung cancer.

Patients and Methods: All consecutive subjects who attended our clinic between January 2016 and January 2019 for the evaluation of pathologic mediastinal and/or hilar lymph nodes (LNs)/mass using EBUS were enrolled into our prospective study. ROSE was performed in the same operating room with EBUS. ROSE results during EBUS were recorded. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of ROSE compared with the final cytologic diagnosis were evaluated.

Results: We enrolled 684 LN/mass lesions belonging to 328 patients into this study. When we compared ROSE results and final cytologic diagnosis, these procedures agreed on 91.6% of the LNs (P<0.001). The sensitivity of ROSE and final cytologic diagnosis for granulomatous inflammation was 72.5%, and for lung cancer, it was 89.2% (P<0.001). The sensitivity of ROSE for the adenocarcinoma subtype of lung cancer was 67.7%, and it was 70% for small cell lung cancer.

Conclusion: ROSE may help to recognize non–small cell lung cancer during EBUS, especially the adenocarcinoma subtype of lung cancer, which will help ensure having sufficient material for molecular analysis.

Endobronchial ultrasonography (EBUS) is a well-known minimally invasive bronchoscopy technique used for the evaluation of mediastinal/hilar lymph nodes (LNs) and mass lesions. It has comparable results with mediastinoscopy, which is known as the gold standard.1,2 Endobronchial ultrasonography-transbronchial needle aspiration (EBUS-TBNA) has a high diagnostic yield for lung cancer and granulomatous lung diseases and has limited complication rates.2 In addition, there are various studies showing the cost-effectiveness and low rate of inadequate sampling with the use of rapid on-site evaluation (ROSE) along with EBUS-TBNA.3–6 However, the opportunity of using ROSE is not widely available because of the need for a cytopathologist. Some recent studies also showed that telecytopathology was also possible and shortened the procedure time.7

Due to the improvement in science and technology, we live in a personalized treatment era for various diseases. Related to that, targeted therapies have had an increasing role in lung cancer treatment in recent years. Especially for non–small cell lung cancer (NSCLC), there have been new drug options based on molecular testing results. For the different targeted therapy options, epidermal growth factor receptor, programmed death ligand-1, anaplastic lymphoma kinase, and c-ROS oncogene 1 are the most common targets searched for in cytopathologic material via molecular tests.8,9 Therefore, it is important to be sure about obtaining adequate material during EBUS-TBNA for use in molecular testing in case of need. Our aim in this study was to evaluate the correlation of ROSE during EBUS-TBNA and final cytologic diagnosis (final diagnosis based on evaluation of smear and/or cell block with immunohistochemical staining), and also to see if ROSE might predict the subtype of the lung cancer.