Histochemistry of Steroid Receptors in Breast Cancer: An Overview*

Transcription

1 ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 9, No. 3 Copyright 1979, Institute for Clinical Science, Inc. Histochemistry of Steroid Receptors in Breast Cancer: An Overview* LOUIS P. PERTSCHUK, D.O.* ERIC GAETJENS, Ph.D. ANNE C. CARTER, M.D., DAVID J. BRIGATI, M.D. DONG S. KIM, M.D., and ELLIS H. TOBIN, B.S Departments o f Pathology and Medicine, Division o f Endocrinology, State University o f New York, Downstate Medical Center, Brooklyn, NY and Kings County Hospital Center, Brooklyn, NY ABSTRACT Immunofluorescence and a new histochemical technique were employed to assay 226 breast cancer specimens for estrogen receptor. Results showed an overall correlation of 91 percent w hen com pared to those of biochemical assays. The histochemical technique is rapid, easy to perform and reveals the same param eters as does immunofluorescence without the need for antiserum. Tumor cell receptor heterogeneity and location of receptor in cytoplasm or nucleus is readily defined by both methods. These histologic tests should prove to be useful in extending the availability of estrogen receptor analysis to all patients w ith breast cancer. Several methods have been described for the detection and study of estrogen receptor (ER) in morphologically intact tissue. F lu o rescein ated estradiol was used to localize ER in rabbit breast and uterine cells2,4 and later to identify ER in human mammary cancer and normal endom etrium.1 Im m unofluorescence (IF) techniques have b een quite successful. * Supported in part by U.S.P.H.S. Research grants CA 23623, CA and contract NOl-CN from the National Cancer Institute and the Jack R. Aron Research Foundation. t Reprint requests to Dr. L. P. Pertschuk, Box 25, Downstate Medical Center, 450 Clarkson Avenue, Brooklyn, NY For IF, both the polym erized5,12,13,15,16 and non-polym erized form of estradiol (E2) have been em ployed as the binding ligand, and im m unoperoxidase m ethodology has been utilized together with the latter.9,10,11 Labeling studies involving the use of antisera prepared in different anim al species have certain inherent disadvantages and their em ploym ent for detection of ER is no exception. V arious batches of antiserum not uncommonly contain e n tirely different am ounts of specific antibody. Additonally, unknown and unw anted antibodies may unexpec /79/ $00.90 Institute for Clinical Science, Inc.

2 220 PERTSCHUK, GAETJENS, CARTER, BRIGATI, KIM AND TOBIN tedly cross-react w ith tissue components or may block activity of the desired reaction. The problem is com pounded by w ide variations in the m olar ratio of fluorescein to protein, or peroxidase to protein, in the labeled immunoglobulin required to make the receptor visible. Variables of this nature make attempts at standardization o f such m ethods im possible. A different approach for the detection of ER was recently reported3 whereby a conjugate of E 2, bovine serum albumin (BSA) and fluorescein isothiocyanate (FITC), was used as the binding hormone. This molecule was composed of 11 moles of FITC and 24 moles of E 2 per mole of BSA, with E 2 linked at position 6. However, the observed results did not correlate w ith those of biochem ical assay. A d iffe re n t co n ju g ate has b een em ployed by us composed of five moles of FITC and four moles of E 2 per mole of BSA in which E 2 was linked at position 17. Results of ER assay using this ligand were compared to those obtained by the dextran-coated charcoal assay (DCC).14 In this paper, the IF and histochemical m ethods for d etection of ER are com pared. Materials and M ethods * Ayerst Laboratories, Inc., New York, NY. Biopsy specim ens were obtained from 226 women with breast cancer at the time of surgery and frozen in liquid nitrogen as soon as possible. Specim ens w ere divided into two portions, one half for biochemical assay and the other for IF or histochem ical analysis. For IF, the binding hormone was used in a polym erized, w ater soluble state (polyestradiol phosphate [PEP]).* Frozen tissue sections, four microns thick were incubated in PEP for one hour, fixed in paraform aldehyde and then exposed to sheep anti-estradiol serum followed by FIT C-labeled anti-sheep im m unoglobulins.f M ultiple controls at each step of the procedure w ere necessary.12,15 For the histochem ical technique, frozen sections of sim ilar thickness w ere incubated w ith 47 to 94 pm oles FITC- BSA-E2 in phosphate buffered saline containing 10 percent ethanol. After two hours, the tissue was fixed in acetoneethanol and triple w ashed in buffer.14 T he histochem ical m ethod required few controls. Parallel sections w ere exp o sed to th e sam e co n cen tratio n s of FITC-BSA-E2 plus 50 to 500 fold excess m olar concentrations of diethylstilbesterol (DES), the antiestrogen CI-628J: and unlabeled E 2 for competitive binding studies. A nother section was incubated w ith FITC-BSA w hich had not been conjugated to E 2. Processed slides were m ounted in buffered gly cero l and exam ined by u l traviolet (UV) microscopy. Hematoxylin and eosin stained sections w ere examined concurrently. The presence of endogeneous, in vivo bound E 2 was determ ined by indirect IF on pre-fixed tissue sections for both assay systems. Specim ens w ere analyzed for the following features: (1) estim ated proportion of epithelium to stroma; (2) presence of viable tum or cells; (3) estim ated percentages of positive and negative malignant cells; (4) fluorescent intensity of positive cells; (5) cytoplasmic or nuclear location of fluorescence; and (6) degree of successful com petition of DES, CI-628 and unlabeled E 2. Specim ens were recorded as positive for ER w hen 10 percent or more of viable tum or cells exhibited fluorescence yet showed no significant fluorescence after in cu b a tio n w ith FITC-BSA, b u t exh ib ite d a significant dim inution of intensity after exposure to com petitors. Known positive ER and negative ER I Calbiochem-Behring, Somerville, NJ. } Warner-Lambert/Parke-Davis, Ann Arbor, MI.

3 ER IN BREAST CANCER BY HISTOCHEMISTRY 221 specim ens w ere processed in parallel w ith each group of tumors assayed. Results U sing e ith e r IF or h isto ch em ical technique, tum or cell heterogeneity of ER was a prom inent feature of the majority of specimens. This was evidenced by cellular gradations of staining intensity varying from weak to brilliant. Brightness of fluorescence was noticeably greater w ith the histochem ical m ethod, preparations were cleaner and easier to interpret and nuclear staining, where present, was more pronounced than with IF. ER was occasionally apparent in norm al structures w ith both tech n iq u es. These included non-neoplastic mammary tissue, epiderm is and hair follicles. Visualization of tum or cells w ithin the dermis was superior with the histochemical procedure as binding to collagen was m inim al. N ecrotic cells often show ed non-specific staining by IF but not with FITC-BSA-E2. Non-specific binding of FITC-BSA was not encountered. H istologic diagnoses of the tum ors studied are shown in table I. Comparison of results of biochemical and histologic analyses are summarized in tables II and III. Correlation of results was slighty but not significantly better by the histochem ical m ethod (92 percent) than with IF (89 percent). The overall correlation rate was 91 percent. Comparison of results of histologic and biochemical ER assay was possible for 207 breast cancers. O f these, 129 were positive for ER and 59 were negative for ER by all methods. However, included in this group were two specim ens with ER only in normal anatomical structures, and 17 which were either entirely necrotic or lacked id entifiable tum or cells. Such specim ens may be sources o f false positive and false negative biochem ical assay results but may be identified as being of no value for assay by either histologic m ethod. TABLE I P a t h o lo g ic D ia g n o ses o f Specim ens A ssayed fo r E strogen R eceptors in B reast Cancers Primary breast cancers Intraductal carcinoma 11 Infiltrating duct cell carcinoma 117 Medullary carcinoma 7 Scirrhous carcinoma 10 Colloid carcinoma 2 I n s i t u and infiltrating lobular carcinoma 15 Mixed types: 3 Lobular and duct cell carcinoma 2 Medullary and duct cell carcinoma Recurrent carcinoma 30 Metastases 22 No tumor identified 226 Com parison was not possible in 19 cases. Four specim ens did not contain sufficient solubilized cytosol protein for DCC. All were histologically positive for ER. Two others, also positive for ER by histology, thaw ed in transit for biochem i cal assay. E leven tum ors show ed a predominance of nuclear ER histologically but were assayed biochemically by DCC which will not detect nuclear ER. In two specim ens, the tissue subm itted for histology was necrotic while the specimen assayed by DCC was positive. TABLE I I Comparison o f R e s u lts o f E strogen R ecep tor A ssay in B rea st Cancer by H is t o lo g ic and B io ch em ica l T ech n iq u es Positive, all methods 129* Negative, all methods 59t Positive by histologic assay Insufficient solubilized protein for DCC/SGA 4 Positive by histology Specimen for DCC thawed 2 Nuclear estrogen receptor predominant histologically; no SGA 11 Tissue for histochemistry necrotic Positive DCC 2 Failure to correlate 19 *Includes two specimens with estrogen receptors in normal structures only. Tumor is negative by histochemistry, tlncludes nine necrotic specimens and seven without tumor. Dextran-coated charcoal assay/sucrose gradient assay.

4 222 PERTSCHUK, GAETJENS, CARTER, BRIGATI, KIM AND TOBIN TABLE III Breast Cancers in Which Assay for Estrogen Receptors by Histologic Techniques Did Not Correlate with Results of Biochemical Analyses Estrogen receptor positive histologically, negative dextran-coated charcoal assay Scirrhous carcinoma 7 Few tumor cells in specimen 1 Micrometastasis 1 Tumor cell heterogeneous for estrogen receptor 5 Estrogen receptor negative histologically, positive dextran-coated charcoal assay All specimens heterogeneous for estrogen receptors 14 5 No correlation of results was o b tain ed in 19 s p e c im e n s (tab le III). F o u r t e e n w ere histologically positive for E R and negative by D C C. O f these, seven w e re scirrhous carcinom as, one was a m icro m etastasis a n d o n e was a skin b io p s y from a p a t i e n t w ith in flam m ato ry c a r cinoma. In these cases, it is possible that the co n ten t of E R o f the cytosol was too low for biochem ica l detection. T h e r e m ain ing s p e c im e n s in this group w e re eq u a lly d iv id ed into those w h ich w e re negative for ER by histology and positive FIG U RE 1. Infiltrating duct c e ll carcinoma. Columns of tumor cells exhibiting predominantly cytoplasmic ER. Fluorescence microscopy. Origi nal magnification x 500. b io c h e m ic a lly a n d th o se h istologically positive for E R yet neg ative on b io c h e m ical assay. T h e s e latter tum ors w e re all hetero g en o u s for E R a n d th ere is thus a possibility that th e diffe rent blocks of tis sue, separately assayed, may have con ta in e d entire ly different proportions of receptor. Approxim ately 12 p e rc e n t of the cases stu d ied e x h ib ite d in vivo b o u n d E 2 b u t in no case d id this interfere w ith the d e te c tion o f ER. Illustrations o f cytoplasm ic an d nuclea r E R are show n in figures 1 an d 2. Discussion Currently, biochem ical E R assays are li m i t e d in p e r fo rm a n c e to co m m e rc ia l laboratories and major m edical centers b ecause o f the sophisticated m eth o d o l ogy involved. Histologic tech n iq u e s, on the o th e r hand, are m uch less costly and r e q u ir e o nly t h e e q u i p m e n t n o rm a lly p r e s e n t in m o s t s u r g i c a l p a t h o l o g y laboratories. Results could b e in te rp re te d by hospital b ase d pathologists w ith very little extra training. C om pariso n of I F an d histochem ical m eth ods for assay of E R shows that there is little difference e ith e r in accuracy or in the param eters w h ich may b e analyzed. H ow ever, I F is subject to the vagaries d is c u s s e d previously, th e p ro c e d u re is m ore lengthy and, as a co n se q u en ce, tis sue sections m ay b ec o m e folded, torn or lost. T h e use o f P E P as the b in d in g hor m o n e adds the pro b lem o f p o ly m er d e g radation over a p erio d of weeks. T h e h isto c h e m ic a l t e c h n i q u e is less ex p e n siv e than IF, p ro c essin g is co m p lete in less than th ree hours and tissue sections are not traum atized. T h e c h e m i cals r e q u ir e d for reag ent preparatio n are com m ercially available and th e materials can be readily stan d ard iz ed an d could be m an ufactured in bulk. F e w e r controls are n e e d e d than w ith IF.

5 ER IN BREAST CANCER BY HISTOCHEM ISTRY 223 F ig u r e 2. Intraductal and infiltrating duct cell carcinoma w ith ER pri marily in n u clei. F lu o rescen ce m icroscopy. O riginal m agn ification x 500. Both histologic m etho ds allow for d e tectio n o f tu m o r cell h etero g en eity o f E R an d p red ic tio n of the m olecular re ceptor form as 4 S nuclear or 8S cytoplasmic. T h e re is ev id en c e that tum ors w ith p r e do m in an tly nu clea r E R may not b e as re spon siv e to horm onal th e ra p y as those w ith cytoplasm ic receptor.6,7 Histologic te c h n iq u e s are applicable to the assay of ti n y s p e c i m e n s, n e e d l e b i o p s i e s a n d m align ant effusions w h ic h are gen erally un su itab le for biochem ical assay. T h e p re s e n c e of E R in hu m an breast c a n cer is of proven prognostic value in selection o f can didates for ablative or ad ditive e n d o c rin e therapy.8 It is therefore im p o r t a n t th a t assay o f E R s h o u ld b e available to all w o m en u n d e r trea tm e n t for b r e a s t c a n c e r. T h e h i s t o c h e m i c a l assay sho u ld prove to b e useful in attain ing this goal. A cknow ledgm ents We are grateful for the dextran-coated charcoal assays performed on many o f these cases by Dr. Wil liam L. McGuire and Staff o f the University oftexas Health Science Center at San Antonio, TX and for other biochemical assays of different specimens by Dr. George A. D egenshein and Staff, Maimonides H ospital and M edical C enter, B rooklyn, NY. Calbiochem -Behring Corporation donated sheep anti-estradiol serum and Warner-Lambert/ParkeDavis Co. donated CI-628. Evelyn A. Rainford and Ethel Jones gave invaluable technical assistance. References 1. B arr o w s, G. H., St r o u p e, S., and G ray, L. A.: In vitro uptake and nuclear transfer of fluoresceinated estradiol in human mammary cancer and normal endometrium. Amer. J. Clin. Path. 70:330, D a n d l i k e r, W. B., L e v i s o n, S. A., and BRAWN, R. J.: Hormone binding by cells and cell fragments as visualized by fluorescence m icroscopy. Res. Commun. Chem. Pathol. Pharmacol. 74: , Le e, S. H.: Cytochemical study of estrogen re ceptor in human mammary cancer. Amer. J. Clin. Path. 70: , L e v iso n, S. A., D a n d l ik e r, W. B., B r a w n, R. J., and V a n d e r l a a n, W. P.: F luorescence polarization m easurem ent of the horm oneb in d in g site in teraction. E n d o crin o lo g y 99: , M a l a n, L. B. and Ja n s s, D. H.: Demonstration o f estrogen-binding protein in human mam mary epithelial cultures by fluorescent anti body staining. In Vitro 14:339, M c G u ir e, W. L., P e a r s o n, O. H., and S e g a l o f f, A.: Predicting hormone respon siveness in human breast cancer. Estrogen Re ceptors in Human Breast Cancer. McGuire, W. L., Carbone, P. P., and Vollmer, E. P., eds. New York, Raven Press, 1975, pp M c G u ir e, W. L., H o r w it z, D. T., G a r o l a, R. E., a n d C h a m n e s s, G. C.: H o r m o n e s in b r e a st c a n c e r u p d a te M e ta b o lism 27: , M c G u ir e, W. L., H o r w it z, K. B., P e a r so n, O. H., and SEGALOFF, A.: Current status of es trogen and progesterone receptors in breast cancer. Cancer 39: , M e r c e r, W. D., C a r l so n, C. A., W a h l, T. B., and T e a g u e, P. O.: Identification of estrogen receptors in human breast cancer cells by im

6 224 PERTSCHUK, GAETJENS, CARTER, BRIGATI, KIM AND TOBIN m unofluorescence and im munoperoxidase techniques. Proc. Soc. Exp. Biol. Med. 2:13, Mer cer, W. D., Ca r l so n, C. A., Wa h l, T. B., and Tea g u e, P. O.: Identification of estrogen receptors in mammary cancer cells by immunofluorescence. Amer. J. Clin. Path. 70:330, Mer cer, W. D., Ca r lso n, C. A., Wa h l, T. B., and T ea g u e, P. O.: Identification o f estrogen receptors in human breast cancer cells by imm unological techniques. Fed. Proc. 37:1312, PERTSCHUK, L. P.: Detection of estrogen binding in human mammary carcinoma by immunofluorescence: A new technique utilizing the binding hormone in a polymerized state. Res. Commun. Chem. Pathol. Pharmacol. 14: , PERTSCHUK, L. P., BRIGATI, D. J., CARTER, A., and McGu ir e, W. L.: D etection o f estrogen receptor in human breast cancer by immunofluorescence. Amer. J. Clin. Path. 68:104, PERTSCHUK, L. P., GAETJENS, E., CARTER, A. C., Brigati, D. J., Kim, D. S., and Fea l e y, T. E.: An improved histochemical method for detection of estrogen receptors in breast cancer: comparison with biochemical assay. Amer. J. Clin. Path., in press. 15. Per tsc h u k, L. P., To bin, E. H., Brig ati, D. J., Kim, D. S., Bloom, N. D., Ga etjens, E., Ber m a n, P. J., Ca rter, A. C., and D eg en- SHEIN, G. A.: Immunofluorescent detection of estrogen receptors in breast cancer. Cancer 41: , To bin, E. H., B loom, N. D., Pertsc h u k, L. P., Berm an, P. J., and D e g e n s h e in, G. A.: Estrogen and progesterone receptor proteins in human mammary carcinoma. Multiple Forms of Steroid Hormone Receptors. Agarwal, M. K., ed. N ew York, E lsevier/n orth H olland Biomedical Press, 1977, pp