Breast cancer characteristics in yielders and nonyielders of nipple aspiration fluid.

Abstract

71 Background: Nipple aspirate fluid (NAF) is an attractive biosample for the investigation of breast cancer risk factors. The reported NAF yield rate varies from 30% to 90% in various studies, raising questions about its value as a risk assessment and about the generalizability of data generated in NAF-yielders to non-yielders. To date, there is no data regarding the characteristics of breast cancers that arise in NAF-yielders and non-yielders. Methods: We examined breast cancer characteristics in NAF yielders and the non NAF yielders in an on-going case control study assessing the hormone concentrations of NAF in breast cancer cases and healthy screening controls. 299 women with recently diagnosed breast cancer were recruited from the Lynn Sage Breast Center. NAF collection was performed from the non cancer breast in the clinic, either before surgery or more than one week post-operatively. NAF yielders produced at least 2 ul of NAF. Each patient completed a detailed study questionnaire. Breast cancer characteristics were recorded on each participant. Results: Among 299 recruited patients 130 (40%) were non NAF yielders (group A) and 169 (60%) were NAF yielders (group B). The association of breast cancer risk factors was compared between them. There were no significant differences in the tumor characteristics. The tumor size, grade, number of positive lymph nodes, and fraction of hormone receptor positive, HER2 positive, and triple negative tumors was similar in NAF-yielders and non-yielders (smallest p value 0.295, for histology, ductal versus lobular). Among the major breast cancer risk factors, the only significant differences were that NAF yielders were younger than non-yielders (50.8 vs. 53.6 years, p < 0.0001) and were more likely to have a history of post-menopausal hormone use (27% among NAF yielders and 7% among non-yielders, p < 0.0001). Conclusions: Tumor characteristics of NAF-yielders and non-yielders are similar, suggesting that there is no qualitative difference between these two groups. The risk predictors developed in yielders can apply to non-yielders.