As previously mentioned, Papanicolaou was the first to report the presence of breast epithelial cells in NAF, and found malignant cells in 1 of 438 asymptomatic women (Papanicolaou et al., 1958). NAF was found to contain not only epithelial cells, but also foam cells, a term used to describe the "foamy" appearance of the cytoplasm. He speculated, "It thus appears possible that under the term foam cell we are dealing with a variety of cell types that, although morphologically indistinguishable..., may vary in origin." Almost 50 years later, after numerous studies using panels of epithelial and macrophage markers, the origin of foam cells remains an area of debate (King et al., 1984; Krishnamurthy et al., 2002; Mitchell et al., 2001). In the report, Papanicolaou also evaluated breast cyst fluid collected from 100 subjects and contrasted cytologic findings in NAF with those in breast cyst fluid. He noted a relative scarcity of foam cells in breast cyst fluid, which are generally the most frequent cellular component of NAF. Leukocytes and macrophages were also scarce in cyst fluid but relatively common in NAF.
The number of epithelial and foam cells and ratio of epithelial to foam cells have been assessed in different breast cancer risk populations (King et al., 1984; Papanicolaou et al., 1958; Sauter et al., 1997). It was found that as breast cancer risk increased, the number of epithelial cells, as well as the ratio of epithelial to foam cells, increased.
Increased breast density suggests more proliferative activity. Increased breast density as seen on mammography has been linked to increased breast cancer risk (Wolfe, 1976). Among a population of women in whom NAF cytology was collected, those with the greatest mammographic density were found to have a fourfold increased risk of atypical hyperplasia (Lee et al., 1992a).
Longitudinal studies have demonstrated the usefulness of abnormal NAF cytology in predicting future breast cancer risk. A prospective study which enrolled 2071 Caucasian women found that, after an average of 12.7 years of follow-up, the relative risk (RR) for women who yielded various cytologic categories of NAF vs. women who yielded no NAF (RR = 1) were as follows: unsatisfactory specimen, 1.4; normal cytology, 1.8; epithelial hyperplasia, 2.5; and atypical hyperplasia, 4.9 (Wrensch et al., 2001). A follow-up study involving 4046 women who were followed for a median of 21 years found that, compared with women from whom no fluid was obtained, whose incidence of breast cancer was 4.7%, the adjusted RRs for women with various NAF cytologic findings were 1.4 for those with unsatisfactory aspirate specimens, 1.6 for those with normal cytology in the aspirates, 2.4 for epithelial hyperplasia, and 2.8 for atypical hyperplasia. Thus, longer follow-up demonstrated a consistent, albeit somewhat lower, increased risk related to worsening NAF cytology, and is consistent with the implications of a fine needle aspiration or excisional biopsy demonstrating atypical hyperplasia (Wrensch et al., 2001).
Multiple aspiration visits have been demonstrated to increase the detection of abnormal epithelial cells in NAF (King et al., in press). Two hundred seventy-six women without known breast cancer underwent nipple aspiration. Among women in whom NAF was collected, hyperplastic cells were found in 34/178 (19.1%) at visit 1, which increased to 73/209 (34.9%) by visit 5. Atypical cells were found in 6.7% at the initial visit, and in 18.2% of NAF specimens in at least one of five visits.
The presence of tumor at the margin of a surgical biopsy presents a treatment dilemma, since approximately half of the time re-excision fails to find residual tumor. On the other hand, tumor recurrence rates are significantly higher if margins are not resected until they are tumor free (Sauter et al., 1999). NAF cytology has been used to evaluate the presence of residual breast cancer. Atypical and malignant cytology observed in NAF samples collected after excisional breast biopsy but before or concurrent with definitive surgery (Sauter et al., 1999) were significantly associated with residual ductal carcinoma in situ (DCIS) or invasive cancer. It was felt that pathologic factors such as tumor distance from the biopsy margin, multifocal/multicentric disease, subtype and grade of DCIS or invasive cancer (IC), tumor and specimen size, tumor and biopsy cavity location, presence or absence of extensive DCIS, and biopsy scar distance from the nipple would optimize a model to predict the presence of residual breast cancer among women with a biopsy with an involved or close tumor margin. The model (Sauter et al., 2001), which included both NAF cytology and pathologic parameters, was superior in predicting residual breast cancer (94%) to models using NAF cytology (36%) or pathologic parameters (75%) alone. NAF cytology also was useful in predicting which patients had one or more lymph nodes involved with tumor, which could prove useful in determining which subjects should receive chemotherapy.
While numerous studies point to the high specificity of NAF cytology in breast cancer diagnosis (King et al., 1975; Papanicolaou et al., 1958; Sauter et al., 1997), cytologic findings are occasionally difficult to interpret. Perhaps the chief difficulty is in the differentiation of benign from malignant papillary growths. This dilemma is found primarily in the cytologic evaluation of SND, which is often the result of a benign papilloma on histopathologic review which can appear suspicious for carcinoma to the cytopathologist not highly familiar with NAF and SND cytologic evaluation (Papanicolaou et al., 1958; Sauter et al., in press-b).
Was this article helpful?