Estrogen and progesterone receptors in breast cancer: comparison between enzyme immunoassay and computer-assisted image analysis of immunocytochemical assay

Automated measurement of estrogen receptor in breast cancer: a comparison of fluorescent and chromogenic methods of measurement

Whereas FDA-approved methods of assessment of estrogen receptor (ER) are 'fit for purpose', they represent a 30-year-old technology. New quantitative methods, both chromogenic and fluorescent, have been developed and studies have shown that these methods increase the accuracy of assessment of ER. Here, we compare three methods of ER detection and assessment on two retrospective tissue microarray (TMA) cohorts of breast cancer patients: estimates of percent nuclei positive by pathologists and by Aperio's nuclear algorithm (standard chromogenic immunostaining), and immunofluorescence as quantified with the automated quantitative analysis (AQUA) method of quantitative immunofluorescence (QIF). Reproducibility was excellent (R(2)>0.95) between users for both automated analysis methods, and the Aperio and QIF scoring results were also highly correlated, despite the different detection systems. The subjective readings show lower levels of reproducibility and a discontinuous, bimodal distribution of scores not seen by either mechanized method. Kaplan-Meier analysis of 10-year disease-free survival was significant for each method (Pathologist, P=0.0019; Aperio, P=0.0053, AQUA, P=0.0026); however, there were discrepancies in patient classification in 19 out of 233 cases analyzed. Out of these, 11 were visually positive by both chromogenic and fluorescent detection. In 10 cases, the Aperio nuclear algorithm labeled the nuclei as negative; in 1 case, the AQUA score was just under the cutoff for positivity (determined by an Index TMA). In contrast, 8 out of 19 discrepant cases had clear nuclear positivity by fluorescence that was unable to be visualized by chromogenic detection, perhaps because of low positivity masked by the hematoxylin counterstain. These results demonstrate that automated systems enable objective, precise quantification of ER. Furthermore, immunofluorescence detection offers the additional advantage of a signal that cannot be masked by a counterstaining agent. These data support the usage of automated methods for measurement of this and other biomarkers that may be used in companion diagnostic tests.

Ductal carcinoma in situ with microinvasion: Clinicopathologic study and biopathologic profile

Data regarding the biologic behavior and surgical management, in particular the axillary lymph node excision, of ductal carcinoma in situ with microinvasion (DCIS-MI) are controversial. Therefore, we decided to study the histopathologic characteristics, the biopathologic profile, as well as the follow-up of a group of patients with DCIS-MI. Thirty-one cases of DCIS-MI, 21 of whom were treated with axillary lymph node dissection, were studied. All cases were classified according to the Van Nuys classification, and the extension of DCIS was quantified. The biopathologic profile (ER, PR, MIB 1, p53, c-erbB-2) as well as the follow-up was also investigated. The results did not reveal any statistically significant differences between the two groups, and there was no statistically significant relationship between the extension of DCIS and the number of microinvasion (MI) foci or maximum MI diameter, or between Van Nuys classification of DCIS and again the number of MI foci or maximum MI diameter. DCIS-MI seems associated with good prognosis. None of the patients had relapses or metastases. Our data seem to suggest that the natural history of DCIS-MI resembles DCIS, and we, therefore, suggest that all the surgically removed area should be examined histologically to avoid missing foci of infiltrating breast cancer larger than 1mm.

Biopathologic profile of breast cancer core biopsy: is it always a valid method?

For breast cancer management biopathologic profile and particularly the expression of estrogen receptor (ER) and progesterone receptor (PR) is considered essential. In advanced cases, core biopsy results are the only data available. To evaluate reliability of data, results of ER, PR, MIB1, p53 and c-erbB2 on core biopsy were compared with those on surgical specimens. Results showed a statistically significant concordance for ER and PR in pT1 but not in pT2 tumors, possibly due to breast cancer heterogeneity. MIB1 results were worse with no significant concordance even for pT1 group. There was statistically significant concordance in pT1 and pT2 groups for p53 and c-erbB 2, probably due to the high number of negative cases for these markers. We recommend more core biopsies for larger tumors since core biopsy has a high probability for giving unreliable data in these cases. In conclusion, this study showed that core biopsy has a high probability for not very reliable data in bigger tumors where the results obtained might be the only data available. A higher number of core biopsy is recommended in those cases.

Biopathological profile of multiple synchronous homolateral and bilateral breast cancers

It still needs to be verified whether multiple syncronous homolateral and bilateral breast cancers represent intramammary spread of a single tumor or two or more separate neoplastic events. To clarify this problem, we studied the biopathological profile of 46 homolateral and 20 bilateral cases. The cancers were always surgically removed and processed at the same time. The expression of estrogen receptors (ER), progesterone receptors (PR), MIB 1, p53, and c-erbB-2 was determined. Computer-assisted image analysis (CAS 200) was used to evaluate ER, PR, MIB 1, and p53. The histological concordance was 95.6% in homolateral and 50% in bilateral cases. The immunophenotype profile of multiple homolateral neoplasms showed a concordance between 93.47% for ER and 78.26% for p53. The results were statistically significant for all parameters except for p53. In bilateral cancers, there was a significant statistical concordance for ER. These data strongly suggest that both mechanisms may exert an influence and, in particular, that in the majority of homolateral carcinomas, there may be intramammary spread of tumor cells. In multiple bilateral tumors, however, the great diversity of the histological aspects and the differences in the immunophenotype pattern suggest that the vast majority of these may constitute independent multiple events.

Prognostic indexes in breast cancer: comparison of the Nottingham and Adelaide indexes

The search for single independent prognostic factors in breast cancer has often produced conflicting results. Therefore, prognostic indexes have been compiled by combining several parameters. In this study we compare the Nottingham Prognostic Index (NPI), which is based on traditional prognostic factors (diameter of the neoplasm, lymph node status and histological grade) with the Adelaide Prognostic Index (API), which is based on the tumour diameter and two biological parameters: oestrogen receptors and cell kinetics. We considered 82 cases of breast cancer observed over the period 1987-1990 with a minimum follow-up of 60 months. The NPI gives a better definition of the prognostic profile for each patient. Our results indicate three prognostic groups (good, moderate, unfavourable), which differ with respect to disease-free survival (DFS; P=0.0024) and overall survival (OS; P=0.0033). In contrast, the API scores showed no significant correlation with OS or DFS. The use of prognostic indexes, especially when compiled using traditional parameters, is a useful aid to the clinician, since they can provide a reliable indication of how individual tumours will evolve.

Breast cancer in young women: clinicopathological features and biological specificity

Literature data suggest that breast cancers occurring in young patients may be different from those arising in older women. In this study the clinicopathologic characteristics of 50 patients under 40 years of age were compared with those of patients aged over 60. Patients under 40 years old more frequently had a family history of breast cancer than did older patients (24% vs 17%) and had more often used oral contraceptives (29% vs 13%); on average they had experienced menarche 1 year earlier. For early onset breast carcinomas there was a higher frequency of grade 3 tumours (38% vs 17%) and oestrogen receptor negativity (46% vs 20%). In addition, in younger patients the carcinomas were mostly DNA aneuploid (78% vs 58%), with a higher proliferation rate (48% vs 26%) and more frequent c-erbB-2 overexpression (48% vs 26%) and p53 alteration (30% vs 8%). Our data demonstrate that breast cancers arising in young women have a significantly different biopathological profile from those in older patients, with a predominance of unfavourable prognostic parameters.

Hormone receptor status in breast cancer--a comparison between surgical specimens and fine needle aspiration biopsies

The present study was performed to evaluate the immunocytochemical analysis (ICA) of oestrogen (ER) and progesterone receptor (PR) in fine needle aspiration (FNA) biopsies from primary breast cancers as compared with the established enzyme immunoassays (ER-EIA and PR-EIA) based on cytosol homogenates from the corresponding resected tumour specimens. A total of 967 primary breast cancers were assessed for ER and PR content by both methods. Correlations between EIA and ICA expressed as percentage of tumour cells with a positive staining were highly significant (P < 0.001) for ER and PR. Staining intensity yielded only limited additional information. The concordance between the two techniques was about 80%. Evaluation of biological parameters by FNA may be useful to decide the optimal treatment for breast cancer patients.

The Prognostic Value of Hormone Receptor Detection by Enzyme Immunoassay and Immunohistochemistry: A Prospective Study in Patients with Early Breast Cancer

The main reason to determine estrogen (ER) and progesterone receptors (PR) in breast cancer is their predictive value for the response to endocrine therapy. In addition, ER and PR are often used as prognostic indicators. Enzyme immunoassay (EIA) and immunocytochemical assay (ICA) are two methods for determining ER and PR. These two methods have not been compared with each other in relation to clinical endpoints. In the present study we prospectively evaluated the prognostic value of ER and PR as determined by ICA and EIA in 223 and 207 patients, respectively, with early breast cancer. ER was positive in approximately 77% of patients, while PR was positive in approximately 65% of patients. The proportion of potential agreement beyond chance between EIA and ICA was 0.58 and 0.65 for ER and PR, respectively. The median follow-up was 86 months. Both ER and PR appeared to be weak prognostic factors. There were no differences in prognostic value according to the time point of analysis or cutoff value chosen, nor were there any differences in the prognostic value of hormone receptors detected by ICA or EIA. Both methods appear to be equivalent in terms of qualification and prognostic value.

Computer-assisted immunocytochemical determination of breast cancer steroid receptors. Frozen sections vs paraffin sections

Previously we have demonstrated that determination of oestrogen (ER) and progesterone (PR) receptors by immunocytochemical assay (ICA) on frozen sections (FS) and cytological smears with image analysis is effective for evaluating steroid receptors. The aim of this study was to determine concordance between ER and PR assessed by ICA on FS and paraffin sections (PS) both evaluated by image analysis. There were 115 breast carcinomas selected. For all cases, ER and PR determination was performed on FS and PS. Computer-assisted image analysis was performed using CAS 200. Results were expressed as percent positive area of neoplastic nuclei compared with total nuclear area of the examined neoplastic cells. Good correlation was demonstrated for both ER (r=0.759; concordance=83.4%) and PR (r=0.800; concordance=87.8%). The unexpected relatively low concordance for ER led to further investigations. We divided the 115 cases in two groups. The first group included specimens from our hospital; the second group specimens from suburban hospitals. In the first group there was better correlation for both ER (r=0.897) and PR (r=0.915) with a concordance of 91.5% and 93.6%, respectively. In the second group, correlation was worse for both ER (r=0.724) and PR (r=0.708), with a concordance of 77.9% and 83.9% respectively. From analysis of discordant cases we conclude that reduction in correlation and concordance with increased false negative cases in group 2 are probably due to delayed fixation. Our data suggest that ICA with automated image analysis is efficient in evaluating ER and PR on paraffin section only when the tumour samples are correctly fixed.

Effect of the aromatase inhibitor vorozole on estrogen and progesterone receptor content of rat mammary carcinomas induced by 1-methyl-1-nitrosourea

Vorozole, a nonsteroidal aromatase inhibitor, impedes the post-initiation stage of chemically induced mammary carcinogenesis. While various aspects of vorozole's effects on mammary carcinoma development have been investigated, little attention has been directed to determining the estrogen receptor (ER) and progesterone receptor (PR) content of mammary carcinomas that arise despite vorozole treatment. Female Sprague-Dawley rats were given an i.p. injection of 50mg MNU/kg body weight at 21 days of age and placed on diet supplemented with 0 or 3 mg vorozole/kg, which had no effect on mammary tumor development. Histologically confirmed carcinomas were evaluated for ER and PR by immunohistochemistry. In the control group, 78.8% of carcinomas were ER positive with an ER content ranging from 13.8 to 40.0%, similar to ER content of mammary ductal epithelial cells from non-carcinogen treated animals. PR content ranged from 4.4 to 45.2% and also was similar to levels of PR observed in ductal epithelial cells. ER was not correlated with PR in mammary carcinomas (r = 0.05, p > 0.80), whereas there was a significant correlation in ductal epithelium (r = 0.86, p = 0.006). In vorozole-treated rats, no ER negative carcinomas were observed and overall ER expression by vorozole was elevated (p < 0.03). All carcinomas from vorozole-treated rats expressed PR (2.5-60.2%) and correlation between ER and PR content was numerically greater in carcinomas from vorozole-treated animals (r = 0.42, p = 0.09). These data, which are considered hypothesis generating, provide evidence that low doses of vorozole in the diet select for mammary carcinomas with an increased ER positive phenotype.

Comparison of enzyme immunoassay and immunohistochemical measurements of estrogen and progesterone receptors in breast cancer patients

Before replacing enzyme immunoassay of estrogen and progesterone receptors by immunohistochemistry, results of both methods were compared on 437 samples obtained from breast cancer patients (342 primary breast carcinomas, 16 local recurrences, 49 biopsies, and 30 tumor specimens obtained after neoadjuvant treatment). Immunohistochemistry (IHC) results were first assessed semiquantitatively on the basis of the estimated proportion of positive tumor cells, and then quantitatively using the "quick score." Semiquantitative IHC hormone receptors results (positive > or = 10%) correlated well with enzyme immunoassay status (positive >15 fmol/mg protein) in 358 surgical samples (342 primary tumors and 16 recurrences), with overall concordance rates of 89.9% and 82.1%, respectively. Among the 100 discordant cases, a large intraductal carcinoma component was observed in 7 of 36 cases for estrogen receptor (ER) and 15 of 64 for progesterone receptor (PR). Thirty-five discordant cases also were observed near the cut-off values. Hormone receptor levels by enzyme immunoassay correlated strongly with the quantitative IHC "quick score." Whatever the method, hormone receptor status was associated with histologic grade (SBR) and tumor size, whereas age correlated strongly with ER positivity. Similar results were obtained for biopsy specimens and posttreatment samples. This comparison improved the reliability of the IHC technique, which is currently routinely used for ER and PR determination in the authors' institution.

Detection of progesterone receptor forms A and B by immunohistochemical analysis

AIM

The measurement of progesterone receptors (PR) is recommended as part of the clinical management of breast and endometrial cancers, and immunohistochemistry on formalin fixed tissue is now the method of choice. PR is expressed as two isoforms, PRA and PRB, and although both these proteins are expressed in hormone dependent cancers, there is evidence that a large proportion of tumours express a predominance of one isoform. Therefore, it is essential to document the individual detection of PRA and PRB by the presently available anti-PR antibodies. The aim of this study is to investigate the detection of PR isoforms A and B in formalin fixed, paraffin wax embedded cell lines and tissue sections by immunohistochemistry, using a panel of commercial and in house antibodies to human PR.

METHODS

PR negative cell lines stably transfected to express only PRA (MCF-7Mll/PRA) or PRB (MDA-MB-231/PRB), and tissue sections of human breast carcinoma and normal endometrium were stained using an immunoperoxidase method. A panel of primary PR specific antibodies was evaluated for ability to detect both PRA and PRB proteins, and for intensity and distribution of positive staining under optimal conditions.

RESULTS

Of the 11 antibodies assessed, only four recognised PRA and PRB similarly. Six recognised PRA proteins but were unable to detect PRB expression in the cell lines expressing only PRA or PRB. In tissues expressing high amounts of PRA and PRB, all antibodies tested demonstrated positive PR staining. However, in tissues expressing a predominance of PRB, differential staining patterns were observed, with variations in staining intensity and in the proportion of cells positive for PR.

CONCLUSIONS

Most PR specific antibodies tested failed to detect PRB in formalin fixed tissue by immunohistochemical techniques, despite their ability to do so by immunoblot analysis. These observations suggest that there are conformational differences between PRA and PRB that mask epitopes on the PRB protein recognised by most anti-PR antibodies. The selection of antibodies that recognise both PRB and PRA in formalin fixed tissue is essential for the accurate evaluation of PR positivity in clinical specimens.

Biopathological significance of single cell DNA aneuploidy measured by static cytometry in breast cancer

DNA ploidy image analysis in breast carcinomas has occasionally revealed cases with a diploid DNA content but with a single hypertetraploid element, the so-called single cell aneuploidy (SCA). To identify the biologic significance of this little known phenomenon, we selected 40 cases with SCA from a series of 599 consecutive operable breast carcinomas. The clinical, pathological and biological characteristics of SCA cases were compared with those of a control group of 40 pure diploid breast carcinomas. Hormonal receptor status, proliferative indexes (Ki-67) and p53 overexpression were determined immunohistochemically and quantitatively evaluated by image analysis. The overexpression of c-erbB-2 was determined semiquantitatively. SCA was observed in 6.6% of cases (40 of 599) and in 17% of otherwise diploid cases (40 of 236). Breast cancers with SCA occur in younger women (mean age 54.75 y vs 61.12 y, P<0.05), are smaller (mean diameter 20.00 vs 21.62 mm), less differentiated (percentage of G3 cases 13.2 vs 2.9), more frequently hormone independent (ER positivity 70.0 vs 77.5%; PgR positivity 57.5 vs 77.5%) and have a greater cell proliferative activity (mean S-phase: 4.6 vs 2.8, P<0.05), (percentage of Ki-67 immunostaining: 24.1 vs 19.7%). There is also more overexpression of c-erbB-2 and P53, particularly in pT1 cases in which the percentage of c-erbB-2 positive cells is 54 vs 32 (P value is not significant) and the percentage of p53 is 29 vs 4 (P<0.05). According to these results SCA may be a reliable marker of genetic instability and of greater biologic aggressiveness. Image cytometry of DNA content may be a cost effective means of identifying breast cancer patients with an increased risk of tumour recurrences despite otherwise favourable prognostic parameters.

Steroid receptor heterogeneity in relation to DNA index in breast cancer: a multiparameter flow cytometric approach on paraffin-embedded tumor samples

Steroid hormone (estrogen and progesterone) receptor (ER and PR) status at the time of breast carcinoma surgery is used as a marker for hormone dependency to guide adjuvant therapy. In a significant number of cases a discrepancy exists between the detected number of hormone receptors and the response to hormonal therapy. One of the explanations for this could be intratumoral heterogeneity. Our objective was to investigate the heterogeneity of steroid hormone receptor expression in breast cancer by using multiparameter flow cytometry (MP-FCM) on routinely processed formalin-fixed, paraffin-embedded tumors. A series of 232 routinely processed breast carcinomas were analyzed using a recently developed technique for the isolation of single cells from paraffin-embedded material. After dewaxing and rehydrating, 50-microm thick sections were heated for 2 hours at 80 degrees C in a citrate solution. Single-cell suspensions were prepared by a short pepsin digestion. The obtained single-cell suspensions were immunostained simultaneously for cytokeratin and ER or PR. Finally, DNA was stained using propidium iodide, after which the samples were analyzed on a flow cytometer. The fractions of ER- and PR-positive cells were determined in the total, as well as the G0 /G1 fraction of the diploid, and in case of nondiploid tumors, also in the G0 /G1 fraction of the aneuploid cell population. Of 232 cases, 88 (38%) were diploid, 38 (16%) were tetraploid, and 106 (46%) were aneuploid. In the diploid tumors the mean fraction of ER- and PR-positive cells was 81% and 76%, respectively. The ER- and PR-positive fractions in the total cytokeratin-positive fraction decreased significantly in the tetraploid (56% and 55%, respectively) and aneuploid tumors (both 47%, p < 0.0001). When analyzing the ER- and PR-positive fractions separately in the diploid and aneuploid cell populations of the nondiploid tumors, it became apparent that the ER and PR status in the diploid fraction of the tumor was significantly higher than in the aneuploid fraction (p < 0.0001). For the tetraploid tumors the mean ER- and PR-positive fractions were 79% and 76%, respectively, in the diploid fraction, and this decreased to 45% in the aneuploid cell subpopulation. In the aneuploid tumors this decrease was even more drastic: in the diploid cell population the ER- and PR-positive fractions were 66% and 62%, while this was 38% and 39% in the aneuploid population. These findings illustrate clearly the existence of a heterogeneous distribution of ER/PR expression in breast cancer, related to the loss of a diploid DNA index. Because of its objective quantification of subfractions within the same tumor, MP-FCM can be regarded as a superior method compared to more conventional techniques such as immunohistochemistry and biochemistry.

Computer-assisted immunocytochemical determination of breast cancer steroid receptors on cytological smears of excised surgical specimens compared with frozen sections

BACKGROUND

Due to the widespread use of fine needle aspirate biopsy the practice of determining estrogen (ER) and progesterone (PR) receptors in breast carcinoma from cytological smears (CS) is becoming very common. The aim of this study was to determine concordance between ER and PR assessed by immunocytochemical assay (ICA) on CS and FS both evaluated by image analysis since we have found no data in literature on this.

METHODS

104 breast carcinoma cases were selected. For all cases ER and PR determination was performed on CS, obtained by light scraping of the freshly cut surface of the excised surgical tumors at the time of frozen section diagnosis, and FS using the same monoclonal antibodies. Computer-assisted image analysis was performed in all cases using CAS 200. Results were expressed as percent positive area of neoplastic nuclei compared with total nuclear area of the examined neoplastic cells.

RESULTS

Good correlation was demonstrated between percent positive nuclear neoplastic area by ER-ICA on CS and FS (r = 0.759; P < 0.0001). Concordance of results was 90.19% (P < 0.001). Good correlation was also demonstrated between percent positive nuclear neoplastic area by PR-ICA in CS and FS (r = 0.889; P < 0. 0001). Concordance of results was 97.02% (P < 0.0001).

CONCLUSIONS

Our data suggest that ICA on CS with automated image analysis is efficient in evaluating ER and PR content in human breast cancer, especially when CS is the only method pathologists have to evaluate receptor status e.g. in advanced breast cancer cases when neoadjuvant therapy is necessary before surgery or when surgery is impossible.

A simple index using video image analysis to predict disease outcome in primary breast cancer

Image analysis was used to investigate the prognostic significance of immunostaining for oestrogen receptor (ER), p53 tumour-suppressor protein and tumour cell proliferation (MIB-1) in a random cohort of 200 primary breast cancer patients with between 4 and 7 years of clinical follow-up. Image measurements of the percentage of immunopositive cancer cell nuclei (% positive nuclear area) were recorded for the above tumour features for each patient. Assessment of relative risk using Cox's univariate analysis indicated that tumour size, number of cancer-involved nodes, MIB-1 and ER % positive nuclear area were significantly associated with breast cancer disease outcome, i.e., relapse-free survival and overall survival. In multivariate analysis, tumour size, number of involved nodes, ER and MIB-1 % positive nuclear area were retained as independent predictors of prognosis, depending on the image measurement cut-point used. A prognostic model, which can be used without reference to nodal involvement, was constructed for tumour size, ER cut-point of 30% positive nuclear area and MIB-1 cut-point of 10% positive nuclear area. Kaplan-Meier analysis of this image-based prognostic index identified 4 risk groups with predicted 5-year overall survival rates of 93%, 83%, 76.7% and 61.5%. We conclude that image measurements of ER and proliferative rate can be combined with tumour size to construct a prognostic index which reliably predicts disease outcome in primary breast cancer without knowledge of the nodal status of the patient.