The value of ThinPrep and cytospin preparation in pleural effusion cytological diagnosis of mesothelioma and adenocarcinoma

Abstract

The diagnosis of malignant mesothelioma requires an integration of the clinical presentation, radiological studies, and immunohistochemical stain of histological sections. Cytological diagnosis on pleural effusions of mesothelioma and pulmonary adenocarcinoma is highly desirable but debatable. A spectrum of cytological features has been found to be associated more commonly with malignant mesothelioma (e.g., peripheral cytoplasmic skirt, bubbly cytoplasm, cyanophilic cytoplasm, and scalloped border of cell balls) vs. adenocarcinoma (e.g., two‐cell population, inspissated cytoplasmic material, cytoplasmic vacuole, angulated and indented nuclei, and smooth border of cell balls) to only name a few. The current study is designed to assess whether the introduction of a liquid‐based technology such as ThinPrep (TP) can provide additional diagnostic value in addition to the conventional cytospin Diff‐Quik (DQ) preparations. Pleural effusion specimens were prepared with split samples for DQ‐stained cytospin and Papanicolaou‐stained liquid‐based TP. Fifteen pleural effusion samples with immunohistologically confirmed malignant mesothelioma and 13 pleural effusion samples of immunohistologically confirmed pulmonary adenocarcinomas were retrieved from our files. Both DQ cytospin‐ and Papanicolaou‐stained TP slides were evaluated for the known cytological features associated with malignant mesothelioma (25 cytological features) and adenocarcinoma (22 cytological features) without knowledge of the original cytological and histological diagnoses. The McNemar test was used to compare these two cytological preparations for both malignant mesothelioma and pulmonary adenocarcinoma. In the malignant mesothelioma group, 4 of 25 cytological features evaluated, bubbly cytoplasm (P = 0.002), vacuolated cytoplasm (P = 0.005), cell‐in‐cell arrangement (P = 0.007) and irregular nuclear contour (P = 0.083), were seen more frequently in the DQ cytospin preparation, as opposed to only one feature, nuclear size enlargement (P = 0.008), more readily seen using TP. In the pulmonary adenocarcinoma group, only 1 of 22 cytological features evaluated, presence of angulated or indented nuclei (P = 0.025), was seen more frequently in DQ as opposed to two features, presence of two‐ cell population (P = 0.04) and presence of micropapillary structures (P = 0.1), were seen more readily in TP. All other cytological features evaluated distinguishing mesotheliomas (20 features) and pleural adenocarcinomas (19 features) were seen equally readily in both types of specimen preparation techniques. This study suggests that the liquid‐based TP preparation of pleural effusions does not appear to provide additional diagnostic value when compared with the DQ cytospin preparation in the cytological distinction between mesothelioma and adenocarcinoma in pleural effusions. Most cytological features evaluated, 20 of 25 (mesothelioma) and 19 of 22 (adenocarcinoma), can be seen in both preparation techniques. Diagn. Cytopathol. 2005;32:137–144.