ABSTRACT
INTRODUCTION: Morphology is routinely used for detecting malignant cells in body fluids, but it has limitations. Recently, flow cytometry (FCM) is used as an effective tool for studying non-haematological malignancies. The main objective of this study is to standardize a simple and rapid FCM test for the detection of malignant epithelial cells in body fluids. MATERIALS AND METHODS: Body fluids that had been processed for cytology/cytology and FCM were enrolled in this prospective study. We developed a fluorescent-labelled, monoclonal antibody panel composed of cell surface markers for this FCM assay. We compared the results of cytology/cell block and FCM. RESULTS: A total of 121 fluid samples were studied. Comparing the diagnostic performance of cytology/cell block and FCM, 52 (43%) cases were positive and 60 (49.5%) cases were negative for carcinoma cells by both techniques. Nine cases showed discordant results between the two techniques. Six cases were cytology+ but FCM- and three cases were FCM+ cytology-. Clustered Epithelial Cell Adhesion Molecule (EpCAM)-positive events with high scatter properties were definitive for positive diagnosis by FCM. We studied PD-L1 expression in 13 cases by FCM. Six cases were reported as false negative by this FCM assay due to hypocellularity and lack of EpCAM expression in malignant cells. CONCLUSIONS: This FCM assay is simple, easier and cost-effective yielding sensitive results with no inter-observer variability. FCM would become a valuable tool to complement routine diagnostic cytology and reduces misdiagnosis.
ABSTRACT
Immunophenotyping plays a major role and is essential for establishing the diagnosis of chronic lymphocytic leukemia (CLL). Though CLL has a characteristic phenotype, diagnosis may be challenging due to immunophenotypic overlap with other B cell non-Hodgkin's lymphomas (B-NHL). Markers like CD200, CD43, CD20 and CD45 were found valuable in CLL and we investigated their diagnostic efficiency and accuracy in 174 patients with leukemic B-NHL. On the integration of four markers by a scoring system, 96% (49/51) of CLL cases showed a score of 3 or 4 and 90% (36/40) of non-CLL cases had a score of 0 or 1. This scoring system for CLL diagnosis showed a sensitivity of 98.2% and 96% in the analytical cohort and validation cohort respectively, which was significantly higher than the classical Matutes score. Hence we strongly suggest considering the expression of CD200, CD20, CD43 and CD45 in the diagnosis of B-NHL cases.
