r 1990 S. Karger AG. Basel 0030 2414 90 0474 0313 $ 2.75 0
Oncology 1990:47:313-317
Immunocytochemical Determination of Epidermal Growth Factor Receptor with Monoclonal EGFR1 Antibody in Primary Breast Cancer Patients1 Pierantonio Bevilacqua, Giampietro Gasparini, Sandro Dal Fior, Giuseppe Corradi St. Bortolo Hospital. Vicenza. Italy
Key Words. Epidermal growth factor receptor • Steroid hormone receptors • Histopathological factors • Breast cancer
Introduction There is currently an increasing interest in the role of autocrine factors in the control of breast cancer cell growth. Epidermal growth factor (EGF') is a polypep tide that stimulates the growth of particular human breast cancer cell lines in vitro [ 1,2]. A prerequisite for its action seems to be the presence of the specific receptor situated in plasma membrane of the target cells [3,4). The EGF receptor (EGFr) has been detect ed in 40-50% of human breast carcinomas by radio ligand biochemical analysis (5, 6] or with imrnunocytochemistry [7, 8], Several reports demonstrate the
1 The authors thank G. Pelizzari for the statistical analysis. G. Pietribiasi lor technical assistance and D. Mazzocco for assistance with the manuscript.
presence of EGFr also in other neoplasias, i.e. lung, hepatoma and brain tumors [9 11]. The EGFr has an external domain which binds EGF and transforming growth factor a, a transmem brane section, and a cytoplasmatic portion that is similar to that of the protein encoded by the oncogene V-erb-B-2 [12], Sainsbury et al. [13] have demonstrated that the determination of EGFr in breast cancer may be rele vant in vivo, in fact its expression was found to be the most important variable in predicting relapse-free and overall survival in stage I-II disease. Slamon et al. [14] have found the prognostic relevance also of the am plification of the FlER-2/neu oncogene, which is close ly related to the EGFr gene. In the present study we assess, by an immunocytochemical assay, the frequency of EGFr positivity in human mammary tumors. We compared the EGFr
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
Abstract. Epidermal growth factor (EGF) has been shown to be important in regulating the growth of breast cancer cells in vivo because of its mitogenic action on some breast cancer cell lines in vitro. Immunocytochemi cal analysis of EGF receptor (EGFr) was carried out on frozen sections in 134 primary breast cancer patients. Overall 68 of 134 (51%) of the tumors were EGFr positive. There was no correlation between EGFr positivity and menopausal status. Regarding the histopathological features, no significant correlations were observed between EGFr expression and tumor size, grading and lymph nodes status. Estrogen (F.R) and progesterone (PgR) receptors were detected by an immunocytochemical assay and an equal distribution of EGFr was found regarding steroid hormonal receptors expression. Finally, there was only a positive trend between the prolifera tive activity of the tumors, as measured by Ki-67 antibody, and the amount of EGFr. Our results suggest the presence of a subclass of breast tumors, characterized by the absence of ER and or PgR and the presence of EGFr, whose growth appears to be mediated by autocrine growth factors rather than by steroid hormones. The overall picture is that of an independent relationship between EGFr expression and the known prognostic factors in breast cancer.
expression with steroid hormonal receptors, certain histopathological characteristics and the proliferative activity of the tumors to evaluate if EGFr is an in dependent parameter in primary breast cancer.
Patients, Materials and Methods One hundred and thirty-four consecutive cases of stage I—11 breast carcinoma ( T i-T ja; N 0- 2 ; Mo) were analyzed. The patients were treated with mastectomy or quadrantectomy plus radio therapy alone in histologically node-negative cases and plus chemo therapy or hormonotherapy (tamoxifen) in histologically node positive cases. The median age of the patients was 55 years (from 31 to 70 years). Menopausal status was defined according to the following criteria: patients were considered in premcnopause when at the time of surgery they were actively menstruating or when less than I year had elapsed from spontaneous menopause; patients were con sidered in perimenopause when 1-5 years had elapsed from spon taneous menopause or oophorectomy and patients were considered in postmenopause when more than 5 years had elapsed from spon taneous menopause or oophorectomy. For histologic evaluation surgically obtained breast cancer spec imens were stripped of blood, fatty tissue and necrotic areas and were immersed for 2 min in a beaker of isopentane, suspended in liquid nitrogen and then removed from the isopentane and im mersed directly in the liquid nitrogen. The specimens were evaluated first in cryostat examination and then fixed in formalin and paraffin sections and stained with hematoxylin and eosin for light micro scopic analysis. Tumors were classified by the histologic type according to the criteria of the National Surgical Adjuvant Breast Project [15]. Grading was according to the criteria of Bloom and Richardson [16] on the basis of nuclear grade and presence or absence of tubular formation. All identifiable lymph nodes in the axillary specimens were examined by light microscopy. EGFr was analyzed using the monoclonal EGFR1 antibody, isolated by Waterfield et al. [17] using cells of the epidermoid carcinoma line A43I as immunogen (clone EGFR1 Amcrsham International Lab.. UK). Immunocytochemical staining was carried out with an avidin-biotin complex immunoperoxidase technique [18]. Frozen samples were cut into 4- to 6-pm sections at —20 C and fixed in cold acetone at —10 C for 10 min. Sections were washed briefly in 0.01 M PBS. pH 7.2 and then incubated at room temperature for 20 min with normal blocking serum diluted 1:10, with primary F.GFR1 antibody diluted 1:10. biotinylated with sec ondary antiserum 1:200 (horse antimouse Ig) and then incubated in avidin-biotinylated horseradish peroxidase complex for 30 min (Vector Lab. Burlingame. Calif.. USA). Each incubation was fol lowed by washes in PBS for 5 min. except for the first step. Labeling was developed with a diaminobenzidine (1 mg/ml in Tris-HCl) as substrate for 10 min and followed by counterstaining with hema toxylin. Controls w'crc obtained by omission of the primary reagent. For hormonal receptors expression specimens were analyzed using the monoclonal antireceptor antibodies (ER-ICA and PgR-ICA kits) developed by Abbott Diagnostics (Abbott Laboratory, North Chicago. 111.. USA). The assay uses a peroxidase-antiperoxidase (PAP) technique for the visualization of both receptors in frozen
Bcvilacqua/Gasparini/Dal Fior/Corradi
tissue sections. Frozen samples were cut into 4- to 6-pm sections at —20 C, thaw-mounted onto glass slides and fixed in 4% formaldehyde-phosphate buffer saline solution (PBS) for 10 min. The slides were transferred to cold methanol at —10 C for 5 min and then to cold acetone at —10 C for up to I month before assay. Stored sections were washed in PBS prior to the ER-ICA and PgR-ICA and then incubated tor 15 min with normal goat serum (blocking agent). The primary antibodies (antihuman estrogen, ER. and progesterone, PgR. rat monoclonal antibodies) were added dropwise to one section of each specimen, control antibody (rat IgG) was added to the second section from the same specimen and sections were incubated with the bridging antibody (goat-antirat IgG) for 30 min Slides were washed again in PBS for 5 min before and after a further 30 min incubation with rat PAP complex. Sections were immersed in distilled water before counterstaining with Mayer's modified hematoxylin for 5 min. Sections were rinsed in water for 5 min and mounted in Kaiser's gelatine. Controls consisted of ER- and PgR-positive cells provided with the kits and treated with primary or control antibody. In addition, a series of tumor specimens were treated with primary or control antibody. For each tumor a total of 30 fields were counted and a mean percentage staining figure was calculated. Tumors were clas sified as ER-ICA or PgR-ICA positive if more than 5% of nuclei were stained positively. For the cell kinetics determination, sections were first washed briefly in 0.01 M PBS, pH 7.2, and then incubated at room tem perature for 20 min in normal blocking serum and successively without washing, with Ki-67 monoclonal antibody (Dakopatts Ltd. UK) diluted 1:50 for 60 min, biotinylated horse antimouse Ig for 30 min (Vector Lab.. Burlingame. Calif.. USA) and avidin-biotinyl ated horseradish peroxidase complex for 30 min (Vector Lab.. Burlingame. Calif., USA). Between incubations the slides were washed with PBS. Labeling was developed with a diaminobenzidonchydrogen peroxide substrate until nuclear staining was easily detectable and sections were then lightly counterstained with hema toxylin. Control sections were stained with an unrelated mono clonal antibody (EGFr) and also omitting primary antibody. The number of cells with nuclear staining was determined counting the number of positive nuclei and the total number of nuclei in 12 high-power fields (h.p.f.; x 40 objective. Zeiss photomicroscope) ensuring that the whole section was scanned. An average of 1,000 nuclei per section were counted and if heterogeneity was a feature, the number of h.p.f. assessed was increased. Areas of the section with the highest labeling rates ’were used by counting. The mean of two percentages of Ki-67 labeled nuclei was calculated and used in the statistical analysis. For all the immunochemical staining the same tumor was evaluated independently by 2 observers. The degree of positivity was graded by a modified semiquantitative approach as described by McCarty et al. [19] which incor porates both the intensity and the distribution of staining in a single numerical score. The score was calculated with the formula = I(i+ 1 ) x Pi and consists of the sum of the staining intensities i ( —(1); +(2); + + (3)) multiplied by the percentage of cells stained (Pi) in each category. Heterogeneity was defined as the proportion of positively stained tumor cells. The x 2 lest for contingency tables was employed to test for association between EGFr with receptor expression, menopausal status, cell kinetics and histopathological features.
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
314
315
EGFr in Primary Breast Cancer
Table 1. Correlations between EGFr and histopathological characteristics
Positive Negative
Grading, n
Tumor size, n
Lymph nodes, n
n
%
1
2
3
1
2
3
-
+
68 66
51 49
40 40
25 26
3 0
4 3
29 31
35 32
40 32
28 35
Results EGFR I antibody produced immunostaining label ing in the cytoplasmic membranes of tumor cells, the reaction tended to be granular cytoplasmic with some peripheral accentuation. The percentage of stained cells varied from case to case and there were patchy positive and negative areas (focal positivity) and areas in which the neoplastic cells showed different staining intensities. Three histochemical classes were defined ( —; + ; + + ) on the basis of the estimated proportion of the entire tumor cell population showing staining. Using the above semiquantitative method there were: 66 (49%) negative tumors ( —) and 68 (51%) positive tumors, 25 of which had moderate levels ( + ) and 43 had high levels ( + + ) of EGFr expression. There were 47 women in pre or perimenopause and 87 in postmenopause. No correlation was observed between EGFR1 positivity and menopausal status (X2 = 0 .001).
Table 1 reports the relationship between some his topathological features and EGFr. x2 analysis shows only a positive trend between tumor size and the distribution of EGFr (%2 = 3.00). All pTj carcinomas were immunoreactive. No correlation was observed between EGFr distribution with grading (x 2 = 0.31) or lymph nodes status (x 2 = 1.66). EGFr expression was correlated with steroid hor mone receptor status determined with an itnmunocytochemical method using the monoclonals anti estrogen antibody and antiprogesterone antibody from Abbott Labs. Regarding ER expression there were 98/134 (73%) ER-ICA positive tumors and re garding PgR expression there were 69/134 (51%) PgR-ICA positive tumors. For both receptors the antibodies produced staining to the nucleus of breast cancer cells. The percentage of stained cells and the intensity varied from case to case, with a substantial heterogeneity in receptor expression in the majority of
Table 2. Correlation between EGFr and immunocytochemical steroid hormonal receptors determination EGFr
Positive Negative
ER-ICA
PgR-ICA
n
%
+
-
+
-
68 66
51 49
50 48
18 18
38 31
30 35
Table .3. EGFr and proliferative activity Ki-67 rate %
EGFr positive
negative
0 5-9 10-19 20-39 ^4 0
n
%
n
%
24 17 7 10 8
36 26 11 15 12
16 19 7 11 15
24 28 10 16 22
sections examined. As reported in table 2 no corre lation was found between EGFr distribution with ER-ICA (x 2 = 0.011) or PgR-ICA (x 2 = 1.06) posi tivity. Ki-67 antibody produced immunostaining labeling in nuclei of a proportion of cells that varied from case to case. There were 94 (70%) immunoreactive tumors with a range of Ki-67 positive cells from 5 to 65%. There was a positive trend (x 2 = 3.86) between the proliferative activity of the tumors by monoclonal Ki-67 antibody and the EGFr expression. In very high proliferating tumors (Ki-67 rate ^ 40%) the percent age of EGFR1 positivity was higher than that of negative tumors (22 vs, 12%). this difference did not reach statistical significance (table 3).
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
EGFr
316
It is widely held that breast cancer is a tumor characterized by a substantial biological heterogeneity being composed of various cell populations, with dif ferent metastatic potential. The immunohistochemistry done in our study confirms this intratumoral heterogeneity. In fact, breast cancers examined con tain clones of cells expressing EGFr, ER/PgR positive and negative cells and fast and low proliferating tu mors. In the present study we demonstrate the presence of EGFr in 51 % of primary breast tumors. This result is comparable with the percentage of EGFr positivity reported by other authors [1.5. 7], The correlation between EGFr expression and some histopathological characteristics of the carcinomas was investigated. Only a positive trend was found between EGFr and tumor size, however the comparison was mainly done within pTi and pT2 cases. All pT3 tumors were EGFR 1-positive. Contrary to the data of Battaglia e! al. [7] and Sainsbury et al. [13], which show a signifi cantly higher percentage of EGFr-positive tumors in stage II (lymph nodes positive) disease, our results found no difference in EGFr distribution regarding lymph nodes status. No correlation was observed be tween EGFr and histologic grading. A similar finding has been found by Walker et al. [8] and Skoog et al. [20], whilst Sainsbury el al. [5] and Delarue et al. [6] show that EGFr-positive carcinomas were generally poorly differentiated. Such differences are probably due to the subjective nature of grading criteria. Furthermore, the correlation between EGFr and steroid hormone receptors was investigated adopting immunohistochemical assays. In our series EGFr was not significantly correlated with either ER or PgR. A significant inverse correlation between EGFr and ER has been demonstrated by some [5, 21] but not all authors [1. 22], Delarue et al. [6] and Battaglia et al. [23] also show an inverse relationship between EGFr and PgR expression. The reasons for these discrepan cies are not clear. The independence of EGFr from steroid hormone receptors suggests that the subclass of tumors which are EGF + and ER/PgR— may represent breast cancer cells in which growth is pri marily regulated by autocrine factors and that are unresponsive to hormonal therapy. As reported in table 2, about one-half of ER-ICA and PgR-ICA negative tumors express EGFr. We found 18 EGFr + / ER —and 30 EGFr-F/PgR — tumors. As demonstrat
ed by Sainsbury et al. [13] such a subgroup of cases may be with poor prognosis. Finally, we have determined the cell kinetics with the monoclonal Ki-67 antibody using a rapid, practi cal and easily performed immunohistochemical meth od. Only a positive trend was observed between high proliferating tumors and EGFr expression. Thirtyeight percent of EGFr-positive tumors were fast pro liferating (Ki-67 >20%). Such a relationship was also demonstrated by Skoog et al. [20], as measured by cellular content of thymidine kinase isoenzyme (which increases in S-phase). On the contrary, a significant positive cor relation between EGFr and high S-phase was ob served by Walker and Camplejohn [8], Because defini tive data regarding EGFr expression and its relation ship with cell kinetics is poorly analyzed in the litera ture, this area of research needs to be studied further. In conclusion, our results revealed that EGFr pos itivity is confined only to certain subgroups of mam mary carcinomas. This finding is of interest when considering growth regulation of breast carcinomas as well as the possible use of EGFr as prognostic indica tor. The observation that EGFr. histopathologic fea tures, steroid hormone receptors and cell kinetics can be independent variables, suggests that these factors might be a powerful combination of prognostic in dicators that should be used for a better stratification of patients in adjuvant therapy clinical trials.
References 1 Fitzpatrick, S.L.; La Chance. M.P.; Schultz, G.S.: Character ization of epidermal growth factor receptor and action on hu man breast cancer cells in culture. Cancer Res. 44: 3442-3447 (1984). 2 Roos, W.; Fabbro, D.; Rung. W.: Costa. S.D.: Eppenbergcr. V.: Correlation between hormone dependency and the regulation of epidermal growth factor receptor by tumor promoters in human mammary carcinoma cells. Proc. natn. Acad. Sci. USA 83: 991-995 (1986). 3 Downward, J.; Parker, P.; Waterfield. M.D.: Autophosphorylalion sites on the epidermal growth factor receptor. Nature. Lond. 311 483-485 (1984). 4 Hcldin. C.H.; Westermark. B.: Growth factors: Mechanisms of action and relation to oncogenes. Cell 37: 9-15 (1984). 5 Sainsbury, J.R.C.; Farndon, J. R.; Sherbet. V.G.; Harris. A.L.: Epidermal growth-factor receptors and oestrogen receptors in human breast cancer. Lancet /': 364 366 (1985). 6 Delarue. J.C.: Friedman, S.; Mouriesse. H.; May-Levin, F.; Sancho-Garnier. FL: Contesso. G.: Epidermal growth factor receptor in human breast cancers: Correlation with estrogen and
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
Discussion
Bevilacqua/Gasparini/Dal Fior/Corradi
EGFr in Primary Breast Cancer
8
9
10
11
12
13
14
15
16 Bloom. H.J.G.: Richardson. W.W.: Histological grading and prognosis. Br. J. Cancer II: 359-377 (1957). 17 Waterfield. M.D.: Mayers. E .L.V.; Stroobant. P.; Bcnnct. P.L.P.; Young. S.; Goodfellow, P.N.; Banting, G.S.; Ozanne. B.: A monoclonal antibody to the human epidermal growth factor receptor. J. cell. Biochem. 20: 149-161 (1982). 18 Hsu, S.M.: Raine. L.; Fanger. H.: A comparative study of the peroxidase antiperoxidase method and an avidin-biotincomplex method for studying polypeptide hormones with ra dioimmunoassay antibodies. Am. J. clin. Path. 75: 734 738 (1981). 19 McCarty. K.S. Jr: Miller. L.S.; Cox, E.B.; Konrath. J.: McCarty. K.S., Sr.: Estrogen receptor analyses: Correlation of biochemical and immunohistochemical methods using mono clonal antireceptor antibodies. Archs Path. Lab. Med. 109: 716-721 (1985). 20 Skoog. L.; Macias. A.; Azavedo, E.; Lombardero. J.: Klitcnberg. C.: Receptors for EGF and oestradiol and thymidine kinase activity in different histological subgroups of human mammary carcinomas. Br. J. Cancer 54: 271-276 (1986). 21 Perez. R.; Pascual. M.; Macias. A.; Loge. A.: Epidermal Growth Factor receptors in human breast cancer. Breast Cancer Res. Treat. 4: 189-193 (1984). 22 Peyrat. J.P.; Bonnettere. J.; Vandewalle. B.; Djiarre, J.: Lefebvre. J.: Récepteurs de l'EGF dans les cancers du sein humains. Relations avec les récepteurs hormonaux. Annls Endocr. 45: 412-413 (1984). 23 Battaglia. F.; Polizzi. G.; Scambia. G.; Rossi. S.; Benedetti Panici. P.; Iacobelli, S.: Crucitti. F.; Mancuso. S.: Receptors for epidermal growth factor and steroid hormones in human breast cancer. Oncology 45: 424-427 (1988).
Giampietro Gasparini. MD St. Bortolo Hospital USSL 8 1-36100 Vicenza (Italy)
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
7
progesterone receptors. Breast Cancer Res. Treat. 11: 173-178 (1988). Battaglia, F.; Scambia, G.: Rossi, S.; Bcncdctti Panici, P.; Bellantone, R.; Polizzi, G.; Querzoli, P.; Negrini. R.; Iacobclli, S.; Crucitti, F.: Mancuso, S.: Epidermal growth factor receptor in human breast cancer: Correlation with steroid hormone recep tors and axillary lymph node involvement. Eur. J. Cancer clin. Oncol. 24: 1685-1690 (1988). Walker, R.A.; Camplejohn. R.S.: DNA flow cytometry of hu man breast carcinomas and its relationship to transferrin and epidermal growth factor receptors. J. Path. 150: 37-42 (1986). Hwang, D. L..: Tay. Y.C.; Lin. S.S.: Lev-Ran, A.: Expression of epidermal growth factor receptors in human lung tumors. Can cer 58: 2260-2263 (1986). Costrini, N.V.; Beck. R.: Epidermal growth factor: Urogastrone receptors in normal human liver and primary hepatoma. Cancer 51: 2191-2196 (1983). Liberman. T.A.; Razon. N.; Bartal. A.D.; Yarden. Y.: Schlessinger. J.; Soreq, H.: Expression of epidermal growth factor receptors in human brain tumors. Cancer Res. 44: 753 760 (1984). Downward. J.: Yarden. Y.; Mayes, E.; Scrace. G.; Totty. N.; Stockwell, P.; Ullrich. A.; Schlessinger, J.; Waterfield. M.D.: Close similarity of epidermal growth factor receptor and v-erb-B oncogene protein sequences. Nature. Lond. 307: 521 526 (1984). Sainsbury, J.R.; Farndon, J.R.; Needham, G.K.; Malcolm. A.J.; Harris, A.L.: Epidermal-growth-factor receptor status as predictor of early recurrence of and death from breast cancer. Lancet 20: 1398-1402 (1987). Slamon. D.J.: Clark. G.M.: Wong. S.G.; Levin. W.J.: Ullrich. A.; McGuire. W. L.: Human breast cancer: correlation of re lapse and survival with amplification of HER-2/neu oncogene. Science 2351: 177-182 (1987). Fisher. E. R.; Gregorio. R.M.; Fisher. B.: The pathology of invasive breast cancer: A syllabus derived from findings of the National Surgical Adjuvant Breast Project (protocol No. 4). Cancer 36: 1-85 (1975).
317
Oncology 1990:47:313-317
Immunocytochemical Determination of Epidermal Growth Factor Receptor with Monoclonal EGFR1 Antibody in Primary Breast Cancer Patients1 Pierantonio Bevilacqua, Giampietro Gasparini, Sandro Dal Fior, Giuseppe Corradi St. Bortolo Hospital. Vicenza. Italy
Key Words. Epidermal growth factor receptor • Steroid hormone receptors • Histopathological factors • Breast cancer
Introduction There is currently an increasing interest in the role of autocrine factors in the control of breast cancer cell growth. Epidermal growth factor (EGF') is a polypep tide that stimulates the growth of particular human breast cancer cell lines in vitro [ 1,2]. A prerequisite for its action seems to be the presence of the specific receptor situated in plasma membrane of the target cells [3,4). The EGF receptor (EGFr) has been detect ed in 40-50% of human breast carcinomas by radio ligand biochemical analysis (5, 6] or with imrnunocytochemistry [7, 8], Several reports demonstrate the
1 The authors thank G. Pelizzari for the statistical analysis. G. Pietribiasi lor technical assistance and D. Mazzocco for assistance with the manuscript.
presence of EGFr also in other neoplasias, i.e. lung, hepatoma and brain tumors [9 11]. The EGFr has an external domain which binds EGF and transforming growth factor a, a transmem brane section, and a cytoplasmatic portion that is similar to that of the protein encoded by the oncogene V-erb-B-2 [12], Sainsbury et al. [13] have demonstrated that the determination of EGFr in breast cancer may be rele vant in vivo, in fact its expression was found to be the most important variable in predicting relapse-free and overall survival in stage I-II disease. Slamon et al. [14] have found the prognostic relevance also of the am plification of the FlER-2/neu oncogene, which is close ly related to the EGFr gene. In the present study we assess, by an immunocytochemical assay, the frequency of EGFr positivity in human mammary tumors. We compared the EGFr
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
Abstract. Epidermal growth factor (EGF) has been shown to be important in regulating the growth of breast cancer cells in vivo because of its mitogenic action on some breast cancer cell lines in vitro. Immunocytochemi cal analysis of EGF receptor (EGFr) was carried out on frozen sections in 134 primary breast cancer patients. Overall 68 of 134 (51%) of the tumors were EGFr positive. There was no correlation between EGFr positivity and menopausal status. Regarding the histopathological features, no significant correlations were observed between EGFr expression and tumor size, grading and lymph nodes status. Estrogen (F.R) and progesterone (PgR) receptors were detected by an immunocytochemical assay and an equal distribution of EGFr was found regarding steroid hormonal receptors expression. Finally, there was only a positive trend between the prolifera tive activity of the tumors, as measured by Ki-67 antibody, and the amount of EGFr. Our results suggest the presence of a subclass of breast tumors, characterized by the absence of ER and or PgR and the presence of EGFr, whose growth appears to be mediated by autocrine growth factors rather than by steroid hormones. The overall picture is that of an independent relationship between EGFr expression and the known prognostic factors in breast cancer.
expression with steroid hormonal receptors, certain histopathological characteristics and the proliferative activity of the tumors to evaluate if EGFr is an in dependent parameter in primary breast cancer.
Patients, Materials and Methods One hundred and thirty-four consecutive cases of stage I—11 breast carcinoma ( T i-T ja; N 0- 2 ; Mo) were analyzed. The patients were treated with mastectomy or quadrantectomy plus radio therapy alone in histologically node-negative cases and plus chemo therapy or hormonotherapy (tamoxifen) in histologically node positive cases. The median age of the patients was 55 years (from 31 to 70 years). Menopausal status was defined according to the following criteria: patients were considered in premcnopause when at the time of surgery they were actively menstruating or when less than I year had elapsed from spontaneous menopause; patients were con sidered in perimenopause when 1-5 years had elapsed from spon taneous menopause or oophorectomy and patients were considered in postmenopause when more than 5 years had elapsed from spon taneous menopause or oophorectomy. For histologic evaluation surgically obtained breast cancer spec imens were stripped of blood, fatty tissue and necrotic areas and were immersed for 2 min in a beaker of isopentane, suspended in liquid nitrogen and then removed from the isopentane and im mersed directly in the liquid nitrogen. The specimens were evaluated first in cryostat examination and then fixed in formalin and paraffin sections and stained with hematoxylin and eosin for light micro scopic analysis. Tumors were classified by the histologic type according to the criteria of the National Surgical Adjuvant Breast Project [15]. Grading was according to the criteria of Bloom and Richardson [16] on the basis of nuclear grade and presence or absence of tubular formation. All identifiable lymph nodes in the axillary specimens were examined by light microscopy. EGFr was analyzed using the monoclonal EGFR1 antibody, isolated by Waterfield et al. [17] using cells of the epidermoid carcinoma line A43I as immunogen (clone EGFR1 Amcrsham International Lab.. UK). Immunocytochemical staining was carried out with an avidin-biotin complex immunoperoxidase technique [18]. Frozen samples were cut into 4- to 6-pm sections at —20 C and fixed in cold acetone at —10 C for 10 min. Sections were washed briefly in 0.01 M PBS. pH 7.2 and then incubated at room temperature for 20 min with normal blocking serum diluted 1:10, with primary F.GFR1 antibody diluted 1:10. biotinylated with sec ondary antiserum 1:200 (horse antimouse Ig) and then incubated in avidin-biotinylated horseradish peroxidase complex for 30 min (Vector Lab. Burlingame. Calif.. USA). Each incubation was fol lowed by washes in PBS for 5 min. except for the first step. Labeling was developed with a diaminobenzidine (1 mg/ml in Tris-HCl) as substrate for 10 min and followed by counterstaining with hema toxylin. Controls w'crc obtained by omission of the primary reagent. For hormonal receptors expression specimens were analyzed using the monoclonal antireceptor antibodies (ER-ICA and PgR-ICA kits) developed by Abbott Diagnostics (Abbott Laboratory, North Chicago. 111.. USA). The assay uses a peroxidase-antiperoxidase (PAP) technique for the visualization of both receptors in frozen
Bcvilacqua/Gasparini/Dal Fior/Corradi
tissue sections. Frozen samples were cut into 4- to 6-pm sections at —20 C, thaw-mounted onto glass slides and fixed in 4% formaldehyde-phosphate buffer saline solution (PBS) for 10 min. The slides were transferred to cold methanol at —10 C for 5 min and then to cold acetone at —10 C for up to I month before assay. Stored sections were washed in PBS prior to the ER-ICA and PgR-ICA and then incubated tor 15 min with normal goat serum (blocking agent). The primary antibodies (antihuman estrogen, ER. and progesterone, PgR. rat monoclonal antibodies) were added dropwise to one section of each specimen, control antibody (rat IgG) was added to the second section from the same specimen and sections were incubated with the bridging antibody (goat-antirat IgG) for 30 min Slides were washed again in PBS for 5 min before and after a further 30 min incubation with rat PAP complex. Sections were immersed in distilled water before counterstaining with Mayer's modified hematoxylin for 5 min. Sections were rinsed in water for 5 min and mounted in Kaiser's gelatine. Controls consisted of ER- and PgR-positive cells provided with the kits and treated with primary or control antibody. In addition, a series of tumor specimens were treated with primary or control antibody. For each tumor a total of 30 fields were counted and a mean percentage staining figure was calculated. Tumors were clas sified as ER-ICA or PgR-ICA positive if more than 5% of nuclei were stained positively. For the cell kinetics determination, sections were first washed briefly in 0.01 M PBS, pH 7.2, and then incubated at room tem perature for 20 min in normal blocking serum and successively without washing, with Ki-67 monoclonal antibody (Dakopatts Ltd. UK) diluted 1:50 for 60 min, biotinylated horse antimouse Ig for 30 min (Vector Lab.. Burlingame. Calif.. USA) and avidin-biotinyl ated horseradish peroxidase complex for 30 min (Vector Lab.. Burlingame. Calif., USA). Between incubations the slides were washed with PBS. Labeling was developed with a diaminobenzidonchydrogen peroxide substrate until nuclear staining was easily detectable and sections were then lightly counterstained with hema toxylin. Control sections were stained with an unrelated mono clonal antibody (EGFr) and also omitting primary antibody. The number of cells with nuclear staining was determined counting the number of positive nuclei and the total number of nuclei in 12 high-power fields (h.p.f.; x 40 objective. Zeiss photomicroscope) ensuring that the whole section was scanned. An average of 1,000 nuclei per section were counted and if heterogeneity was a feature, the number of h.p.f. assessed was increased. Areas of the section with the highest labeling rates ’were used by counting. The mean of two percentages of Ki-67 labeled nuclei was calculated and used in the statistical analysis. For all the immunochemical staining the same tumor was evaluated independently by 2 observers. The degree of positivity was graded by a modified semiquantitative approach as described by McCarty et al. [19] which incor porates both the intensity and the distribution of staining in a single numerical score. The score was calculated with the formula = I(i+ 1 ) x Pi and consists of the sum of the staining intensities i ( —(1); +(2); + + (3)) multiplied by the percentage of cells stained (Pi) in each category. Heterogeneity was defined as the proportion of positively stained tumor cells. The x 2 lest for contingency tables was employed to test for association between EGFr with receptor expression, menopausal status, cell kinetics and histopathological features.
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
314
315
EGFr in Primary Breast Cancer
Table 1. Correlations between EGFr and histopathological characteristics
Positive Negative
Grading, n
Tumor size, n
Lymph nodes, n
n
%
1
2
3
1
2
3
-
+
68 66
51 49
40 40
25 26
3 0
4 3
29 31
35 32
40 32
28 35
Results EGFR I antibody produced immunostaining label ing in the cytoplasmic membranes of tumor cells, the reaction tended to be granular cytoplasmic with some peripheral accentuation. The percentage of stained cells varied from case to case and there were patchy positive and negative areas (focal positivity) and areas in which the neoplastic cells showed different staining intensities. Three histochemical classes were defined ( —; + ; + + ) on the basis of the estimated proportion of the entire tumor cell population showing staining. Using the above semiquantitative method there were: 66 (49%) negative tumors ( —) and 68 (51%) positive tumors, 25 of which had moderate levels ( + ) and 43 had high levels ( + + ) of EGFr expression. There were 47 women in pre or perimenopause and 87 in postmenopause. No correlation was observed between EGFR1 positivity and menopausal status (X2 = 0 .001).
Table 1 reports the relationship between some his topathological features and EGFr. x2 analysis shows only a positive trend between tumor size and the distribution of EGFr (%2 = 3.00). All pTj carcinomas were immunoreactive. No correlation was observed between EGFr distribution with grading (x 2 = 0.31) or lymph nodes status (x 2 = 1.66). EGFr expression was correlated with steroid hor mone receptor status determined with an itnmunocytochemical method using the monoclonals anti estrogen antibody and antiprogesterone antibody from Abbott Labs. Regarding ER expression there were 98/134 (73%) ER-ICA positive tumors and re garding PgR expression there were 69/134 (51%) PgR-ICA positive tumors. For both receptors the antibodies produced staining to the nucleus of breast cancer cells. The percentage of stained cells and the intensity varied from case to case, with a substantial heterogeneity in receptor expression in the majority of
Table 2. Correlation between EGFr and immunocytochemical steroid hormonal receptors determination EGFr
Positive Negative
ER-ICA
PgR-ICA
n
%
+
-
+
-
68 66
51 49
50 48
18 18
38 31
30 35
Table .3. EGFr and proliferative activity Ki-67 rate %
EGFr positive
negative
0 5-9 10-19 20-39 ^4 0
n
%
n
%
24 17 7 10 8
36 26 11 15 12
16 19 7 11 15
24 28 10 16 22
sections examined. As reported in table 2 no corre lation was found between EGFr distribution with ER-ICA (x 2 = 0.011) or PgR-ICA (x 2 = 1.06) posi tivity. Ki-67 antibody produced immunostaining labeling in nuclei of a proportion of cells that varied from case to case. There were 94 (70%) immunoreactive tumors with a range of Ki-67 positive cells from 5 to 65%. There was a positive trend (x 2 = 3.86) between the proliferative activity of the tumors by monoclonal Ki-67 antibody and the EGFr expression. In very high proliferating tumors (Ki-67 rate ^ 40%) the percent age of EGFR1 positivity was higher than that of negative tumors (22 vs, 12%). this difference did not reach statistical significance (table 3).
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
EGFr
316
It is widely held that breast cancer is a tumor characterized by a substantial biological heterogeneity being composed of various cell populations, with dif ferent metastatic potential. The immunohistochemistry done in our study confirms this intratumoral heterogeneity. In fact, breast cancers examined con tain clones of cells expressing EGFr, ER/PgR positive and negative cells and fast and low proliferating tu mors. In the present study we demonstrate the presence of EGFr in 51 % of primary breast tumors. This result is comparable with the percentage of EGFr positivity reported by other authors [1.5. 7], The correlation between EGFr expression and some histopathological characteristics of the carcinomas was investigated. Only a positive trend was found between EGFr and tumor size, however the comparison was mainly done within pTi and pT2 cases. All pT3 tumors were EGFR 1-positive. Contrary to the data of Battaglia e! al. [7] and Sainsbury et al. [13], which show a signifi cantly higher percentage of EGFr-positive tumors in stage II (lymph nodes positive) disease, our results found no difference in EGFr distribution regarding lymph nodes status. No correlation was observed be tween EGFr and histologic grading. A similar finding has been found by Walker et al. [8] and Skoog et al. [20], whilst Sainsbury el al. [5] and Delarue et al. [6] show that EGFr-positive carcinomas were generally poorly differentiated. Such differences are probably due to the subjective nature of grading criteria. Furthermore, the correlation between EGFr and steroid hormone receptors was investigated adopting immunohistochemical assays. In our series EGFr was not significantly correlated with either ER or PgR. A significant inverse correlation between EGFr and ER has been demonstrated by some [5, 21] but not all authors [1. 22], Delarue et al. [6] and Battaglia et al. [23] also show an inverse relationship between EGFr and PgR expression. The reasons for these discrepan cies are not clear. The independence of EGFr from steroid hormone receptors suggests that the subclass of tumors which are EGF + and ER/PgR— may represent breast cancer cells in which growth is pri marily regulated by autocrine factors and that are unresponsive to hormonal therapy. As reported in table 2, about one-half of ER-ICA and PgR-ICA negative tumors express EGFr. We found 18 EGFr + / ER —and 30 EGFr-F/PgR — tumors. As demonstrat
ed by Sainsbury et al. [13] such a subgroup of cases may be with poor prognosis. Finally, we have determined the cell kinetics with the monoclonal Ki-67 antibody using a rapid, practi cal and easily performed immunohistochemical meth od. Only a positive trend was observed between high proliferating tumors and EGFr expression. Thirtyeight percent of EGFr-positive tumors were fast pro liferating (Ki-67 >20%). Such a relationship was also demonstrated by Skoog et al. [20], as measured by cellular content of thymidine kinase isoenzyme (which increases in S-phase). On the contrary, a significant positive cor relation between EGFr and high S-phase was ob served by Walker and Camplejohn [8], Because defini tive data regarding EGFr expression and its relation ship with cell kinetics is poorly analyzed in the litera ture, this area of research needs to be studied further. In conclusion, our results revealed that EGFr pos itivity is confined only to certain subgroups of mam mary carcinomas. This finding is of interest when considering growth regulation of breast carcinomas as well as the possible use of EGFr as prognostic indica tor. The observation that EGFr. histopathologic fea tures, steroid hormone receptors and cell kinetics can be independent variables, suggests that these factors might be a powerful combination of prognostic in dicators that should be used for a better stratification of patients in adjuvant therapy clinical trials.
References 1 Fitzpatrick, S.L.; La Chance. M.P.; Schultz, G.S.: Character ization of epidermal growth factor receptor and action on hu man breast cancer cells in culture. Cancer Res. 44: 3442-3447 (1984). 2 Roos, W.; Fabbro, D.; Rung. W.: Costa. S.D.: Eppenbergcr. V.: Correlation between hormone dependency and the regulation of epidermal growth factor receptor by tumor promoters in human mammary carcinoma cells. Proc. natn. Acad. Sci. USA 83: 991-995 (1986). 3 Downward, J.; Parker, P.; Waterfield. M.D.: Autophosphorylalion sites on the epidermal growth factor receptor. Nature. Lond. 311 483-485 (1984). 4 Hcldin. C.H.; Westermark. B.: Growth factors: Mechanisms of action and relation to oncogenes. Cell 37: 9-15 (1984). 5 Sainsbury, J.R.C.; Farndon, J. R.; Sherbet. V.G.; Harris. A.L.: Epidermal growth-factor receptors and oestrogen receptors in human breast cancer. Lancet /': 364 366 (1985). 6 Delarue. J.C.: Friedman, S.; Mouriesse. H.; May-Levin, F.; Sancho-Garnier. FL: Contesso. G.: Epidermal growth factor receptor in human breast cancers: Correlation with estrogen and
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
Discussion
Bevilacqua/Gasparini/Dal Fior/Corradi
EGFr in Primary Breast Cancer
8
9
10
11
12
13
14
15
16 Bloom. H.J.G.: Richardson. W.W.: Histological grading and prognosis. Br. J. Cancer II: 359-377 (1957). 17 Waterfield. M.D.: Mayers. E .L.V.; Stroobant. P.; Bcnnct. P.L.P.; Young. S.; Goodfellow, P.N.; Banting, G.S.; Ozanne. B.: A monoclonal antibody to the human epidermal growth factor receptor. J. cell. Biochem. 20: 149-161 (1982). 18 Hsu, S.M.: Raine. L.; Fanger. H.: A comparative study of the peroxidase antiperoxidase method and an avidin-biotincomplex method for studying polypeptide hormones with ra dioimmunoassay antibodies. Am. J. clin. Path. 75: 734 738 (1981). 19 McCarty. K.S. Jr: Miller. L.S.; Cox, E.B.; Konrath. J.: McCarty. K.S., Sr.: Estrogen receptor analyses: Correlation of biochemical and immunohistochemical methods using mono clonal antireceptor antibodies. Archs Path. Lab. Med. 109: 716-721 (1985). 20 Skoog. L.; Macias. A.; Azavedo, E.; Lombardero. J.: Klitcnberg. C.: Receptors for EGF and oestradiol and thymidine kinase activity in different histological subgroups of human mammary carcinomas. Br. J. Cancer 54: 271-276 (1986). 21 Perez. R.; Pascual. M.; Macias. A.; Loge. A.: Epidermal Growth Factor receptors in human breast cancer. Breast Cancer Res. Treat. 4: 189-193 (1984). 22 Peyrat. J.P.; Bonnettere. J.; Vandewalle. B.; Djiarre, J.: Lefebvre. J.: Récepteurs de l'EGF dans les cancers du sein humains. Relations avec les récepteurs hormonaux. Annls Endocr. 45: 412-413 (1984). 23 Battaglia. F.; Polizzi. G.; Scambia. G.; Rossi. S.; Benedetti Panici. P.; Iacobelli, S.: Crucitti. F.; Mancuso. S.: Receptors for epidermal growth factor and steroid hormones in human breast cancer. Oncology 45: 424-427 (1988).
Giampietro Gasparini. MD St. Bortolo Hospital USSL 8 1-36100 Vicenza (Italy)
Downloaded by: Karolinska Institutet, University Library 130.237.122.245 - 1/17/2019 8:09:39 PM
7
progesterone receptors. Breast Cancer Res. Treat. 11: 173-178 (1988). Battaglia, F.; Scambia, G.: Rossi, S.; Bcncdctti Panici, P.; Bellantone, R.; Polizzi, G.; Querzoli, P.; Negrini. R.; Iacobclli, S.; Crucitti, F.: Mancuso, S.: Epidermal growth factor receptor in human breast cancer: Correlation with steroid hormone recep tors and axillary lymph node involvement. Eur. J. Cancer clin. Oncol. 24: 1685-1690 (1988). Walker, R.A.; Camplejohn. R.S.: DNA flow cytometry of hu man breast carcinomas and its relationship to transferrin and epidermal growth factor receptors. J. Path. 150: 37-42 (1986). Hwang, D. L..: Tay. Y.C.; Lin. S.S.: Lev-Ran, A.: Expression of epidermal growth factor receptors in human lung tumors. Can cer 58: 2260-2263 (1986). Costrini, N.V.; Beck. R.: Epidermal growth factor: Urogastrone receptors in normal human liver and primary hepatoma. Cancer 51: 2191-2196 (1983). Liberman. T.A.; Razon. N.; Bartal. A.D.; Yarden. Y.: Schlessinger. J.; Soreq, H.: Expression of epidermal growth factor receptors in human brain tumors. Cancer Res. 44: 753 760 (1984). Downward. J.: Yarden. Y.; Mayes, E.; Scrace. G.; Totty. N.; Stockwell, P.; Ullrich. A.; Schlessinger, J.; Waterfield. M.D.: Close similarity of epidermal growth factor receptor and v-erb-B oncogene protein sequences. Nature. Lond. 307: 521 526 (1984). Sainsbury, J.R.; Farndon, J.R.; Needham, G.K.; Malcolm. A.J.; Harris, A.L.: Epidermal-growth-factor receptor status as predictor of early recurrence of and death from breast cancer. Lancet 20: 1398-1402 (1987). Slamon. D.J.: Clark. G.M.: Wong. S.G.; Levin. W.J.: Ullrich. A.; McGuire. W. L.: Human breast cancer: correlation of re lapse and survival with amplification of HER-2/neu oncogene. Science 2351: 177-182 (1987). Fisher. E. R.; Gregorio. R.M.; Fisher. B.: The pathology of invasive breast cancer: A syllabus derived from findings of the National Surgical Adjuvant Breast Project (protocol No. 4). Cancer 36: 1-85 (1975).
317
