Using fl ow cytometry for the diagnosis of lung neuroendocrine carcinoma with samples from fi ne-needle aspiration and pleural fl uid

Abstract

Mark E. Costaldi, Adam Lall, Richard Burack, Zhongren D. Zhou

Objective: Diagnosis of primary and metastatic lung neuroendocrine carcinoma (LNC) is challenging with fine-needle aspiration (FNA) since the morphologic features of LNC and lymphoid lesions overlap. Flow cytometry (FC) is commonly utilized for lymphoid lesions; however, FC with cluster differentiation (CD) 56 and cytokeratin antibodies can also be used for the diagnosis of neuroendocrine neoplasia. Our studies compared the role of FC to immunohistochemistry (IHC) for the diagnosis of LNC from cytological specimens. Materials and Methods: Cases of LNC with both cytology and FC were identified in our pathology databases from 2008 until June 2012. IHC and FC results were compared in these cases. Results: Fifteen of 17 cases had immunohistochemical studies from cell block including cytokeratin (11/11, 100%), thyroid transcription factor-1 (6/8, 75%), synaptophysin (12/14, 86%), chromogranin (4/13, 31%) and CD 56 (10/11, 91%). Of all 17 cases, 13 cases had FC performed with antibodies for CD 56, cytokeratin (Cam5.2) and CD 45. Twelve of the 13 cases were positive for CD 56 (12/13, 92%), which was similar to the IHC results (10/11, 91%). Eleven cases were positive for cytokeratin by IHC and 10 were positive (10/12, 83%) for cytokeratin by FC. One case in our study which was CD 56+/Cam5.2+/CD 45− was diagnosed as carcinoid tumor by final morphology. Conclusions: Our results suggest that FC utilizing antibodies for CD 56/cytokeratin/CD 45 is a reliable alternative method for detecting LNC from FNA specimens when IHC fails or becomes unavailable. However, the morphology is still necessary for the diagnosis since the FC panel is not wholly specific for the subtypes of LNC.