Author:
Yu Shuang-Ni,Li Joshua,Wong Sio-In,Tsang Julia Y S,Ni Yun-Bi,Chen Jie,Tse Gary M
Abstract
AimsThe probabilistic approach is widely adopted for breast fine needle aspiration cytology. However, a definite cytological diagnosis is not always possible for C3 (atypia) cases, which poses a management dilemma as this represents a mixed category of benign and malignant cases. It would be beneficial to be able to predict malignancy based on specific cytological features in C3 aspirates.MethodsA comprehensive panel of cytological features (including quantitative, cytomorphological and background features) in a large cohort of C3 breast aspirates with subsequent histological excisions was evaluated to identify relevant morphological criteria predicting the risk of subsequent malignancy.ResultsA total of 229 C3 specimens with histological follow-up were included. Malignant outcome was found in 30.1% of specimens and the majority were invasive cancers. Features that showed a significant association with malignant outcome included older age (p=0.001), lower percentage of epithelial cell clusters and high percentage of single cells (p=0.002), cribriform architecture in cell clusters (p=0.034), presence of intracellular mucin (p=0.027), increased cell clusters without myoepithelial cells (p=0.048), diminished fibromyxoid stromal fragments (p=0.001), reduced bipolar nuclei (p=0.021) and the presence of necrosis (p=0.023). Except for the percentages of single cells and cell clusters without myoepithelial cells, all other features were shown to be independent risk predictors in multivariate analysis.ConclusionsC3 aspirates were associated with a significant probability of histological malignancy. Certain quantitative, cytomorphological and background features were potentially helpful in predicting the risk of a malignant outcome. The prediction could be clinically useful in the management of C3 cases.
Subject
General Medicine,Pathology and Forensic Medicine
Cited by
19 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献