Breast Cancer Vol. 14 No. 1 January 2007 Original Article Immunohistochemical Expression of Hormone Receptors and The Histological Characteristics of Biochemically Hormone Receptor Negative Breast Cancers Rieko Nishimura 1, Toshiaki Saeki, Shozo Ohsumi 2, Yoichi Tani 4, and Shigemitsu Takashima 2 Departments of 1 Clinical Laboratory and 2 Surgery, National Hospital Organization Shikoku Cancer Center; Department of Breast Oncology, Saitama Medical School; and 4 Department of Medical Science, Dako Japan Inc., Japan Background: Most of the discordant cases between biochemical and immunohistochemical (IHC) assays for hormone receptor (HR) status in breast cancers are due to negative findings from the biochemical assay but positive IHC findings. However determining HR status based on IHC only in biochemically HR negative breast cancers has never been studied. The aim of this study is to examine the histological characteristics in immunohistochemically HR positive but biochemically HR negative breast cancers. Methods: IHC staining for HRs in 45 biochemically HR-negative breast cancers was done. The relationship between HR status by IHC and the histological characteristics was assessed. Results: In 45 cancers, 105 (0.4%) were estrogen receptor- (ER) or progesterone receptor- (PR) positive by IHC. The enzyme-immunoassay (EIA) HR titer was higher in immunohistochemically HR-positive tumors (ER: 2.7 fmol/mg protein; PR: 0.8 fmol/mg protein) than in negative tumors (0.6 fmol/mg protein in both HRs). IHC-assessed ER positivity on histological sections was high in some tumor types, such as mucinous carcinoma (77.8%), invasive micropapillary carcinoma (66.7%), and infiltrating ductal carcinoma of no special type with abundant stroma (60.2%). Among infiltrating ductal carcinomas of no special type, low nuclear grade tumors were all ER positive and high nuclear grade tumors showed low ER positivity by IHC, even in biochemically HR negative cancers. Conclusion: The IHC-assessed HR status may reflect tumor cell behavior, such as overall and disease-free survival and endocrine response, better than HR status as assessed by the enzyme-immunoassay method. Immunohistochemically HR-positive but biochemically HR-negative breast cancers include infiltrating ductal carcinomas of no special type with low nuclear grade and some tumor types with high stromal content. We can assess the true HR status by IHC especially these tumors. Breast Cancer 14:100-104, 2007. Key words: Breast cancer, Hormone receptor, Biochemistry, Immunohistochemistry Introduction Assessment of hormone receptors (HRs) in breast cancers with the immunohistochemical Reprint requests to Rieko Nishimura, Department of Clinical Laboratory, National Hospital Organization Shikoku Cancer Center, 160 Kou, Minamiumemoto-machi, Matsuyama, Ehime 791-0288, Japan E-mail: rnishimu@shikoku-cc.go.jp Abbreviations: HR, Hormone receptor; ER, Estrogen receptor; PR, Progesterone receptor; IHC, Immunohistochemical; EIA, Enzyme-immunoassay; HE, Haematoxylin and eosin Received June 15, 2006; accepted September 29, 2006 (IHC) method on paraffin sections has recently become wide spread 1, 2). Most of the cases with discordant biochemical assay and IHC results are due to negative findings from the biochemical assay but positive IHC findings -5). The main reason for this discordance is thought to be the histology of the tumors; carcinoma with sparse cellularity 5, 6), tumor heterogeneity 4), and low levels of cytosolic protein close to the cut offs 4). An altered hormone binding site but intact epitope for the monoclonal antibody can also be a cause for the discrepancy 6). Basing HR status on IHC only among biochemically HR negative breast cancers has never been studied. 100
Breast Cancer Vol. 14 No. 1 January 2007 We studied the IHC expression of HRs in biochemically HR negative breast cancers and examined the histological characteristics of these tumors. Materials and Methods Study Population A total of 45 primary breast cancers, which were negative for both estrogen receptor (ER) and progesterone receptor (PR) by enzymeimmunoassay (EIA) between April 1, 1990, and December 1, 2000, were studied. During that period, 1850 primary breast cancers were resected in our institution. Of 1850 resected tumors, 527 tumors were HR negative and 849 tumors were ER- and/or PR- positive by EIA. The HR status of 474 tumors was unknown. HR negative breast cancer by EIA methods accounted for 527 of 176 primary breast cancers (8.%). In 527 HR-negative breast cancers, appropriateparaffin blocks were obtained in 45 tumors. Of these tumors, 14 tumors were from breast conserving surgery specimens and 202 tumors were from total mastectomy specimens. EIA for Hormone Receptors Partial tumor specimens were frozen and sent to a commercial laboratory (Otsuka Assay Co. Ltd., Tokushima, Japan) at the time of operation. The titers of HRs were analyzed by routine EIA methods in the laboratory. The cut-off value for HR status positivity was determined to be 1 fmol/mg protein for ER and 10 fmol/mg protein for PR, according to the manufacturer s recommendation. Tumors with ER and PR titers less than 5 fmol/mg protein by EIA were analyzed as HR-negative in laboratory reports. Histological Sections for Immunohistochemical Staining and Histological Characteristics Haematoxylin and eosin (HE)-stained slides of all cases were reviewed, and a representative section was chosen for the study. Serial sections were cut for IHC analysis and HE-staining for histological comparison. The histological diagnoses used in this study were performed on re-cut HE sections instead of the original diagnostic slide. The stromal content of the tumor was visually estimated in infiltrating carcinoma of no special type. In infiltrating carcinoma of no special type, nuclear grade was evaluated by the Modified Scarff- Bloom-Richardson Grading System. Immunohistochemical Analysis for Hormone Receptors IHC analysis for HRs was performed on paraffin sections by the standardized staining methods using an automated instrument (Autostainer Plus, DakoCytomation, Carpinteria, CA, USA). An anti- ER monoclonal antibody (clone 1D5, dilution 1:50, DakoCytomation) and anti-pr antibody (clone PR66, dilution 1:800, DakoCytomation) were used for primary antibodies, and then detected by an EnVision Plus kit (DakoCytomation). The heating method was used for antigen retrieval according to the manufacturer s recommendation. Evaluation of Staining Results For each immunohistochemically stained slide, we visually estimated the percentage of tumor cells showing nuclear reactivity. The slides in which more than 10% of tumor cells were stained were evaluated as positive. Staining intensity was not evaluated. Normal ducts in each slide were used for positive controls. In infiltrating ductal carcinoma of no special type, the relationship between IHC status of the ER and estimated stromal content or nuclear grade was compared. Results The IHC status of HRs is given in Table 1. Among 45 cases of biochemically HR-negative breast cancers, 105 tumors (0.4%) were ER or PR positive. There were 104 (0.1%) ER positive tumors and 60 (17.4%) PR positive tumors. Only 1 tumor was ER negative and PR positive. Among the 241 ER negative tumors, 1 had less than 10% ER-positive cells, and 12 tumors had less than 1% ER positive cells. Table 1 Immunohistochemical Status of ER and PR in 45 Cases of Biochemically Hormone Receptor Negative Breast Cancers Receptor No. (%) ER+ PR+ ER+ or PR+ 104 (0.1%) 60 (17.4%) 105 (0.4%) ER, estrogen receptor; PR, progesterone receptor; +, positive; -, negative. 101
Nishimura R, et al Biochemically Hormone Receptor Negative Breast Cancers Fig 1 Immunohistochemical status of hormone receptors and biochemical results. (a)estrogen receptor (ER) and (b)progesterone receptor (PR). EIA titer in hormone receptors is higher in immunohistochemically hormone receptor-positive tumors (average 2.7 fmol/mg protein for ER and 0.8 fmol/mg protein for PR) than negative tumors (average 0.6 fmol/mg protein for ER and PR). (Student s t-test) Table 2 Relationship of Immunohistochemical Status of Estrogen Receptor and Histological Subtypes of Biochemically Hormone Receptor Negative Breast Cancers Histological subtypes ER+ ER- Total Infiltrating ductal carcinoma of no special type Mucinous Metaplastic Invasive micropapillary Infiltrating lobular Microinvasive ductal carcinoma Non-invasive ductal carcinoma 85 (29.2%) 7 (77.8%) 0 ( 0%) 2 (66.7%) 1 (.%) ( 20%) 6 (28.6%) 206 (70.8%) 2 (22.2%) ( 100%) 1 (.%) 2 (66.7%) 12 ( 80%) 15 (71.4%) 291 9 15 21 Total 104 241 45 The IHC status of HRs and biochemical results are shown in Fig 1. For ER, the average EIA titer was 2.7 fmol/mg protein in immunohistochemically HR-positive tumors and 0.6 fmol/mg protein in immunohistochemically HR-negative tumors. For PR, the average EIA titer was 0.8 fmol/mg protein in immunohistochemically HR-positive tumors and 0.6 fmol/mg protein in immunohistochemically HR-negative tumors. Student s t-test demonstrated that the EIA titer of ER was statistically significantly higher in immunohistochemically HR-positive tumors than negative tumors (p<0.0001), and that the EIA titer of PR was higher in immunohistochemically HR positive tumors than negative tumors, but without statistical significance (p=0.40). The relationship between histology and ER positivity on histological sections is shown in Table 2. In infiltrating ductal carcinoma of no special type, 85 of 291 (29.2%) were ER positive, 7 of 9 (77.8%) mucinous carcinomas were ER positive, 2 of (66.7%) invasive micropapillary carcinomas were ER positive, and 1 of (.%) infiltrating lobular was ER positive. None of the cases of metaplastic carcinoma were ER positive. In microinvasive or non-invasive ductal carcinoma, 9 of 6 (25%) were ER positive. The relationship between the IHC status of ER and estimated stromal content among infiltrating ductal carcinoma of no special type is presented in Table. Seventy percent was used for the cut-off, because it was statistically more significant than 0% cut-off. The ER positive rate was statistically higher in less cellular tumors by the Chi-square test (p<0.0001). In addition, nearly 80% of mucinous carcinomas were ER positive on IHC stain- 102
Breast Cancer Vol. 14 No. 1 January 2007 Table Relationship of Immunohistochemical Status of Estrogen Receptor and Stromal Content of Biochemically Hormone Receptor Negative Infiltrating Ductal Carcinoma of No Special Type Stromal Content in Tumor ER+ ER- Total More than 70% (estimated) Less than 70% (estimated) 50 (60.2%) 5 (16.8%) ing. About one-fourth of non-invasive or microinvasive carcinomas were ER positive on histological sections. The relationship between ER IHC status and nuclear grade in infiltrating ductal carcinoma of no special type is presented in Table 4. Among biochemically HR negative infiltrating ductal carcinomas of no special type, all nuclear grade 1 tumors were ER positive, whereas only 12.5% of nuclear grade carcinomas showed ER positivity on immunostaining. Discussion (9.8%) 17 (8.2%) 8 208 Total 85 206 291 2 =54.1, p<0.0001, Chi-square test. Table 4 Relationship of Immunohistochemical Status of Estrogen Receptor and Nuclear Grade in Biochemically Hormone Receptor Negative Infiltrating Ductal Carcinoma of No Special Type Nuclear grade ER+ ER- Total 1 2 1 ( 100%) 60 (.0%) 12 (12.5%) 0 ( 0%) 122 (67.0%) 84 (87.5%) 1 182 96 Total 85 206 291 Nuclear grade is based on Modified Scarff-Bloom-Richardson Grading. 2 =45.7, p<0.0001, Chi-square test. Originally, the HR status of breast cancers was assessed on cytosols prepared from frozen tissue by biochemical assays 1, 2). The method of choice in the 1970s was a dextran-coated charcoal ligandbinding assay (DCC), and in the mid 1980s a cytosole EIA was developed 1, 2). At the same time, the IHC method was introduced, but the IHC was not widely adopted because of the need for frozen tissue sections and low sensitivity 1, 2). After more practical methods of IHC using paraffin sections were introduced in the late 1980s, the assessment of HRs in breast cancers by IHC has been widely used 1, 2). Assessment of HR status by IHC has been shown to have higher discriminating power than biochemical assays for predicting overall and disease-free survival 7) and the highest sensitivity and predictive value for endocrine response 8, 9). The concordance rate between biochemical assays and IHC on paraffin sections has been reported to be from 68% to 89.9% -5). Most of the discordant cases are negative on the biochemical assay but positive on the IHC -5). The discordance has been thought to be caused by the histology of the tumors as it occurs most often in carcinomas with sparse cellularity 4-6) such as tubular or mucinous carcinomas, tumors with abundant fibrous stroma, and tumors with a large intraductal component. Tumor heterogeneity 4) and low levels of cytosolic protein close to the cut-offs 4) also was contribute. An altered hormone binding site with an intact epitope for the monoclonal antibody can be a cause of discrepancy 6). HR status by IHC only among biochemically HR negative breast cancers has, however, never been studied. In our study, about one-third of biochemically HR negative breast cancers were ER or PR positive by IHC. In comparison with the average EIA titer of HRs in histologically HR-positive and -negative tumors, both ER and PR are higher in histologically HR-positive tumors than in histologically HR-negative tumors. This is likely because histologically HR-positive tumors contain more HRs in the cytosols than histologically HR-negative tumors, and tumors with low levels of cytosolic protein close to the cut-offs were assessed as HRnegative by EIA. IHC-assessed ER positivity on histological sections was high in some tumor types, even in biochemically HR negative breast cancers; such as mucinous carcinoma (77.8%), invasive micropapillary carcinoma (66.7%), and infiltrating ductal carcinoma of no special type with abundant stroma(60.2%). This explains why some of the real ERpositive tumors with high stromal content were assessed as ER negative by EIA. Among infiltrating ductal carcinomas of no special type, low nuclear grade tumors were all ER positive and high nuclear grade tumors showed low ER positivity by IHC, even in biochemically HR negative cancers. The IHC-assessed HR status 10
Nishimura R, et al Biochemically Hormone Receptor Negative Breast Cancers may reflect tumor cell behavior; such as overall and disease-free survival and endocrine response, more than the HR status as assessed by EIA. In conclusion, about one-third of biochemically HR-negative breast cancers were ER- or PR- positive by IHC. We can assess the true HR status by IHC, especially for tumors with low nuclear grade but high stromal content. Acknowledgements We thank Ms. Tamami Yamamoto, Ms. Sachiko Morita, and Ms. Emi Nakaya at the Department of Clinical Laboratory, National Hospital Organization Shikoku Cancer Center, for their expert technical assistance. This study was partially supported by Grants-in-Aid for the National Hospital Organization Working Group on Breast Screening Pathology. References 1) Pertschuk LP, Axiotis CA: Steroid hormone receptor immunohistochemistry in breast cancer: past, present, and future. Breast J 5(1):-12, 1999. 2) Barnes DM, Hanby AM: Oestrogen and progesterone receptors in breast cancer: past, present and future. Histopathol 8:271-274, 2001. ) Biesterfeld S, Schröder W, Steinhagen G, Koch R, Veuskens U, Schmitz F-J, Handt S, Böcking A: Simultaneous immunohistochemical and biochemical hormone receptor assessment in breast cancer provides complementary prognostic information. Anticancer Res 17:472-470, 1997. 4) Ferrero-Poüs M, Trassard M, Le Doussal V, Hacène K, Tubiana-Hulin M, Spyratos F: Comparison of enzyme immunoassay and immunohistochemical measurements of estrogen and progesterone receptors in breast cancer patients. Appl Immunohistochem Mol Morphol 9():267-275, 2001. 5) Umemura S, Itoh H, Ohta M, Suzuki Y, Kubota M, Tokuda Y, Tajima T, Osamura Y: Immunohistochemical evaluation of hormone receptor for routine practice of breast cancer: highly senstive procedures significantly contribute to correlation with biochemical assays. Appl Immunohistochem Mol Morphol 11(1): 62-72, 200. 6) Stierer M, Rosen H, Weber R, Hanak H, Auerbach L, Spona J, Tüchler H: Comparison of immunohistochemical and biochemical measurement of steroid receptors in primary breast cancer: evaluation of discordant findings. Breast Cancer Res Treat 50:125-14, 1998. 7) Esteban JM, Ahn C, Battifora H, Felder B: Predictive value of estrogen receptors evaluated by quantitative immunohistochemical analysis in breast cancer. Am J Clin Pathol 102 Suppl 1:S9-12, 1994. 8) Harvey JM, Clark GM, Osborne CK, Allred DC: Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer. J Clin Oncol 17:1474-1481, 1999. 9) Elledge RM, Green S, Pugh R, Allred DC, Clark GM, Hill J, Ravdin P, Martino S, Osborne CK: Estrogen receptor (ER) and progesterone receptor (PgR), by ligand-binding assay compared with ER, PgR and ps2, by immuno-histochemistry in predicting response to tamoxifen in metastatic breast cancer: a southwest oncology group study. Int J Cancer 89:111-117, 2000. 104