Diagnostic difficulties and pitfalls in rapid on-site evaluation of endobronchial ultrasound guided fine needle aspiration

BACKGROUND: One of the novel techniques utilizing fine needle aspiration (FNA) in the diagnosis of mediastinal and lung lesions is the endobronchial ultrasound (EBUS)-guided FNA. In this study, we describe five cases which had a discrepancy between on-site evaluation and final diagnosis, or a diagnostic dilemma when rendering the preliminary diagnosis, in order to illustrate some of the diagnostic difficulties and pitfalls that can occur in EBUS FNA. METHODS: A total of five EBUS FNA cases from five patients were identified in our records with a discrepancy between the rapid on-site evaluation (ROSE) and final diagnosis, or that addressed a diagnostic dilemma. All of the cases had histological confirmation or follow-up. The cytomorphology in the direct smears, cell block, and immunohistochemical stains were reviewed, along with the clinical history and other available information. RESULTS: Two cases were identified with a nondefinitive diagnosis at ROSE that were later diagnosed as malignant (metastatic signet-ring cell adenocarcinoma and metastatic renal cell carcinoma (RCC)) on the final cytological diagnosis. Three additional cases were identified with a ROSE and final diagnosis of malignant (large cell neuroendocrine carcinoma (LCNEC) and two squamous cell carcinomas), but raised important diagnostic dilemmas. These cases highlight the importance of recognizing discohesive malignant cells and bland neoplasms on EBUS FNA, which may lead to a negative or a nondefinitive preliminary diagnosis. Neuroendocrine tumors can also be difficult due to the wide range of entities in the differential diagnosis, including benign lymphocytes, lymphomas, small and nonsmall cell carcinomas, and the lack of immunohistochemical stains at the time of ROSE. Finally, the background material in EBUS FNAs may be misleading and unrelated to the cells of interest. CONCLUSIONS: This study illustrates the cytomorphology of five EBUS FNA cases that address some of the diagnostic challenges witnessed while examining these specimens during ROSE. Many of the difficulties faced can be attributed to the baseline cellularity of the aspirates, the bronchial contamination, the difficulty identifying neoplasms with bland cytology, the wide spectrum of diseases that can occur in the mediastinum with overlapping cytomorphologic features, the mismatch between the background material and the cell populations present, and the overall unfamiliarity with these types of specimens.