To many, progression of breast cancer and loss of estrogen receptor (ER) are synonymous. The perception is that normal breast tissue is ER*, is growth stimulated by estrogen, and is dependent upon estrogen for growth. Patients with ER+ invasive carcinoma have better prognosis than those with ER-negative cancer. Furthermore, loss of receptor makes the cancer more difficult to treat because hormonal manipulation is not useful. However, in the normal breast very few epithelial cells are ER+.
The number varies with the menstrual cycle but by immunohistochemical methods they are fewer than 10% of the normal epithelial cell population (Petersen et al., 1987; Jacquemier et al., 1990). Expression of ER increases in benign breast lesions.
Immunohistochemical analysis of ER expression in biopsy specimens from premenopausal patients found 6% of normal cells positive, 12% of epithelial cells in fibroadenomas, 13% in atypical lobular hyperplasia, and 31% in CIS (Jacquemier et al., 1990). However, in postmenopausal cases, 26% of normal cells, 25% of fibroadenoma cells, 73% of epithelial cells from proliferative lesions, and 30% from DCIS were positive.
In a case control study in which ER+ status of benign lesions was defined as the presence of any nuclear staining, 84% of cases versus 57% of controls were ER+ (Khan et al., 1994). Progesterone receptor status was similar-86% positive for both. The benign lesions were not further defined but the adjusted odds ratio for ER+ benign lesions in this study was 6.5 (Khan et al., 1994). In a study of 140 benign breast lesions of various types, including fibroadenomas, papillomas, and ductal hyperplasias, 70% showed some nuclear staining (Giri et al., 1989).
Many immunohistochemical studies use the H score system which multiplies the % of cells positive by the intensity of staining (1+, 2+, or 3+) for each positive cell.
Thus, the maximum score is 300 (all cells 3+). Using an H score of 20 as the criterion for classification as ER+, Schmitt (1995) found that the proportion of positive cases was similar for PBD and breast cancer, but proliferation indices, determined by MIB-1 antibody staining, were higher in ER+ than in ER-negative hyperplastic lesions. On the other hand, MIB-1 staining of DCIS and invasive cancer cases was greater in ER-negative than in ER+ cases, suggesting alteration in mitogenic response to estrogen with progression rather than alteration of receptor expression (Schmitt, 1995). However, due to the low H score used in this study, 80% of normal biopsies were also judged to be positive.
It has been reported that at least 25% of the epithelial nuclei stain strongly in 30% of benign breast lesions (Giri et al., 1989). This level of staining corresponds to biochemical assay values in excess of 10 fmol/mg protein (Giri et al., 1987), a level frequently used clinically as the cut off for classifying cancer as ER positive or negative.
This study (Giri et al., 1989) indicates that 30% of benign lesions would have H scores of 50 or more (at least 25% of cells at 2+ or more). Normal tissues, including areas co-existing with DCIS, were negative by this criterion whereas 12/39 fibroadenomas, 13/32 papillomas, 7/20 sclerosing adenoses, 12/49 ductal hyperplasias, and 21/48 DCIS were positive (Giri et al., 1989). Since an H score of 50 corresponds to 10 fmol/mg protein receptor protein, a level expressed by approximately half of breast cancers at diagnosis, it is clear that progression often involves increased ER in PBD and that loss of ER is a later event in progression.
The early expression of receptor most likely enables cells to respond to the mitogenic signal of estrogen thus promoting expansion of initiated epithehal cells. Loss of ER at later stages in breast cancer may reflect a loss of dependency on estrogen as a mitogenic signal. Amplification of erbB-2 is strongly associated with loss of ER expression in DCIS as well as in invasive carcinoma (Pavelic et al., 1992; Poller et al., 1993) and may reflect a role for an alternative mitogenic signal.
Overexpression of other growth factors may enhance hormone-independent growth in the absence of estrogen. Overexpression of heregulin-β2 in MCF-7 cells leads to estrogen independence and resistance to antiestrogens (Tang, C.K. et al., 1996). Some tumor suppressor genes may function by maintaining hormonal regulation, the deregulation of which has been observed to correlate with loss of heterozygosity(LOH). LOH at 17q21 has been significantly correlated with absence of both estrogen and progesterone receptors in breast cancer, and LOH at 17pl3.3 with loss of progesterone receptor (Ito et al., 1995). Hormonal regulation may be a central feature for many of the molecular alterations that occur in breast cancer.
Fred Raymond Miller
Advances in Oncobiology