Detection and distribution of estrogen receptor in human breast cancer by a combined biochemical/immunohistochemical multiple microsample assay
Date
1987
Authors
Thornton, Ian George
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The relationship between high estrogen receptor (ER) levels and the regression of tumor growth following estrogen ablative therapy is well-established in human breast cancer. Generally however, even with treatment, tumors may become ER-negative and unresponsive to further hormonal therapy. This has traditionally been explained as a clonal expansion of ER-negative cells that have a selective growth advantage over ER-positive ones in the absence of estrogen. According to this hypothesis, once such a clone is established the tumor's growth can proceed unimpeded, after which reversal to a hormone-controlled state is improbable.
Commonly, the measurement of ER for clinical screening of patients for hormone therapy is done by steroid-binding assay (SBA) which measures the uptake of radiolabelled estradiol by a single sample of tumor. However, this method measures only previously-unoccupied receptor sites. Furthermore, if the sample does not consist entirely of tumor cells and has a high proportion of connective tissue, fluid, or serum contaminants, the quantitation of ER may be invalidated and tumors having a small proportion of highly ER-positive cells susceptible to endocrine response could remain undetected.
In order to further understanding of ER distribution in breast cancer a comparative multiple microsample assay was performed on 21 tumors . Four 13 X 1.5 X 1.5 mm (approx. 40 mg) microsamples from each tumor were divided equally and longitudinally, one half used for SBA by cold-agar gel electrophoresis (CAGE), and the other half for an estrogen receptor immunochemical assay (ER-ICA) employing monoclonal antibodies against ER.
For SBA , cytosols of microsamples were prepared by sequential homogenization and centrifugation , then incubated with tritiated 17B- estradiol. CAGE was used to separate receptor-bound hormone from unbound hormone . Uptake of estradiol was quantitated by scintillometry . Albumin , which binds estradiol nonspecifically, and in high concentrations interferes with ER estimation , was measured by radial immunodiffusion assay .
For ER- ICA 6 ~m frozen sections were incubated with monoclonal anti -human ER antibody. Following this, a peroxidase: anti-peroxidase (PAP) immunohistochemical procedure was applied to all sections in order to localize ER- bound antibody . Each stained section was examined by four independent observers to determine per cent carcinoma (PCC) and per cent specific ER staining (PSS) . SBA-derived ER concentrations were then corrected for PCC.
The proportion of tumor cells stained specifically for ER in each microsample was estimated and adjustment made for differences in staining intensity (SI). The resulting histoscore is an arbitrary value that allows these data to be ranked and compared with SBA-derived data.
Qualitatively, 68 % of all microsamples were ER-positive , and 28 . 5 % were negative by both methods, with an overall correlation of 0 . 97 .
Quantitatively, there was a correlation of 0.75 between SBA-derived corrected ER and ER-ICA derived histoscores for corresponding microsamples. The high correlation between SBA and ER- ICA data demonstrates the potential value of combined biochemical and immunohistochemical methods for screening breast cancer patients most likely to benefit from endocrine therapy, and thus avoiding the complications associated with other forms of therapy, such as radiation and chemotherapy.
The distribution of ER within the 21 tumors studied revealed a variety of staining patterns, both homogeneous and heterogeneous, the latter demonstrating various degrees of intermixing of ER- positive and ER-negative tumor cells .
Within individual tumors heterogeneous, as opposed to homogeneous , ER distribution patterns were seen in 12 of the 15 (80 %) ER-positive tumors . The checkerboard staining pattern observed in 9 of these might represent a variant type of tumor in which ER-negative status is attained randomly throughout the tumor, rather than by means of a focal, clonal expansion of ER-negative cells. If so , it may be possible that appropriate manipulation of the cell environment could reverse the tendency for breast tumor cells to become hormone-independent.