Evaluation of the clinical utility of EBUS-TBNA sampling with ROSE (Rapid On-Site Evaluation) in the diagnosis of mediastinal and hilar metastatic non-lung extrathoracic primary carcinoma

Background: Endobronchial Ultrasound (EBUS) has an established role in lung cancer diagnosis. However, a proportion of patients have non-lung primary carcinoma diagnosed when sampling the mediastinum and hila, requiring alternative treatment. We analysed the incidence of these cases at our institute. Methods: Retrospective analysis of prospectively maintained EBUS database to identify all cases of non-lung extrathoracic primary carcinoma at our unit between 1/1/2013 and 31/6/2020. Cases had medical notes, radiology, EBUS report including ROSE and final cytology report reviewed. Benign cases, haematological malignancies, malignant mesothelioma and thymic malignancies were excluded. Results: A total of 79 cases were identified of 2806 patients having EBUS in the study period. 65% male, 35% female. Mean age 69 (range 31-93). A total of 92 nodes were sampled, mean size 1.87cm, average of 4 passes per node. The majority (84%) achieved the diagnosis with sampling of just one nodal station. The nodes most commonly sampled were station 7 (28%), 4R (25%), 11L (10%), 11R (9%), 10R (7%), 2R (5%), 4L (5%) and 3P (2%). Final pathology showed the following malignancies: 23% Gastrointestinal, 21% renal, 12% breast, 12% urological, 11% sarcoma, 5% melanoma, 2% endometrial, 1% olfactory neuroblastoma, 1% squamous cell tongue and 12% unknown but certain non-lung primary carcinoma based on radiology and pathology. Conclusion: EBUS is useful diagnostic test in non-lung primary carcinoma diagnosis, potentially avoiding further invasive investigations and ensuring correctly targeted systemic oncological treatment.