Preliminary analysis of hormones and cells in lavage fluid from distinct ducts in healthy women.

A141 Background: Current techniques for accessing the breast ducts provide promise for identifying intermediate markers. However, establishing the hormonal and cellular composition within the breast ducts is critical in determining whether the unit of study should be a duct, a breast, or a woman. Pooled nipple aspirate fluid has previously been analyzed for hormonal composition. Analysis has also been performed using ductal lavage fluid from distinct ducts in high risk women (1). To our knowledge, no study has assessed ductal lavage fluid of healthy women in order to establish correlations between ducts, breasts, and women. Materials and Methods: This analysis combines initial results from two hypothesis-generating IRB approved studies. Volunteers over 18 years of age consented to ductal lavage under local anesthesia and nipple aspiration. Fluid samples were spun and the supernatant analyzed for protein, estrone sulfate, estradiol, and progesterone. Epithelial cells and macrophages were also evaluated. Mulitlevel modeling was used to estimate intraclass correlation coefficients (ICCs) and to account for correlated observations in analyses of variance. This is a preliminary analysis of cell, hormone, and protein yield from 53 women and 218 ducts. Results: Our findings indicate that breast intraclass correlation coefficients (ICCs) were zero or near zero for log-normally distributed analytes. Subject ICCs were zero or near zero for estradiol and progesterone but were relatively high (0.10-0.35) for all other analytes. Age, Gail index, and menopause status were unrelated to protein, estradiol, progesterone, and epithelial cells. Odds of macrophage presence increased with age (p = 0.03) but were decreased in postmenopausal women (p = 0.003). For estrone sulfate, a significant interaction existed between age and menopause status (p = 0.01). Among postmenopausal women, age and Gail index were unrelated to estrone sulfate. For premenopausal women, estrone sulfate significantly increased with Gail index (p = 0.03) and significantly decreased in younger and in nulliparous women (p = 0.003 and 0.004, respectively). Discussion: There was little or no evidence of within-breast correlation for all analytes. However, ducts from the same subject seemed more similar than ducts from different subjects except for estradiol and progesterone, which showed little evidence of within-subject correlation. This would suggest that certain analytes are correlated within subjects while others vary duct to duct within the same subject. Because multilevel modeling is sensitive to small sample sizes, these conclusions could change with additional subjects. Results from additional subjects are being analyzed. [1] Bhandare D et al. Endocrine Biomarkers in Ductal Lavage Samples from Women at High Risk for Breast Cancer. Cancer Epidemiol Biomarkers Prev 2005; 14(11):2620-7.