Cervical cytology in the detection of uterine clear cell carcinoma: Diagnostic predictors from a case-control study.

Abstract

Uterine clear cell adenocarcinoma (CCC) is a rare, aggressive malignancy with poor prognosis. The present study aimed to identify and describe its characteristic morphological features in cervical cytology. This was a 3-year retrospective case-control study. Cases included cervical samples of histopathologically proven endometrial and cervical CCC. Controls included cervical samples of histopathologically proven endometrial serous carcinoma (n=15), endometrioid adenocarcinoma (n=20), and endocervical adenocarcinoma (n=15). Twenty-eight cytomorphological features were evaluated; the strength of association was determined by odds ratio (OR) and Cramer's V, and the diagnostic accuracy of statistically significant features was assessed. Cases consisted of histopathologically proven 25 (34.7%) endometrial and 13 (18.0%) cervical CCC. Corresponding cervical samples were available for a total of 14 (36.8%) patients, of which 13 (92.8%) were positive for epithelial cell abnormality. On univariate analysis, three cytomorphological variables were significant predictors of uterine CCC: presence of dense cytoplasm (OR=88; V=0.72), deep nuclear membrane irregularities (OR=17.5; V=0.55), and coarse chromatin (OR=21.3; V=0.46). Dense cytoplasm had the highest positive predictive value (92%) and high specificity (97.8%), whereas coarse chromatin had the highest sensitivity (92.3%) and negative predictive value (96.7%). The presence of dense cytoplasm and deep nuclear membrane irregularities in the tumour cells were strong predictors, and coarse chromatin a moderate predictor, of uterine CCC compared to its close cytological mimics.