Formerly, the principal characteristics of a patient that were useful in treating breast cancer were the patient's age, menopausal status and previous response to endocrine therapy if applicable. The pathologist also provided valuable information related to the pathologic category and size of the lesion and the number of nodes exhibiting invasive disease. With the advances in flow cytometry and thymidine labeling of biopsies, the ploidy of the tumor and the number of cells composing the S-phase fraction now appear to be useful as prognostic factors in breast and other tumors (Hedley et al., Cytometry 6:327, 1985; Meyer and Wittliff, Int. J. Cancer 47:213, 1991). During the past two decades, the principal prognostic factors measured biochemically and employed in breast cancer were the sex-steroid hormone receptors.
Our understanding of molecular endocrinology and the absolute requirement of a receptor protein for expression of a hormone's action opened a new area of application to clinical chemistry. Currently more than 20 proto-oncogenes have been identified, and their protein products may be classified as either growth factors, growth factor or hormone receptor proteins, tyrosine kinases, G-regulatory-like proteins or nuclear associated proteins. A number of these have been purported to be candidates for better defining the biology and natural history of breast cancer (cf Gelmann & Lippman, IN: The Breast: Comprehensive Management of Benign and Malignant Diseases, K. I. Bland and E. M. Copeland III, eds., Ch. 19, W. B. Saunders, Philadelphia, 1991.)
Steroid Hormone Receptors
It is known that in order for sex hormones to promote their characteristic response in normal target organs such as the breast and uterus, estrogen and progestin must first bind to intracellular receptor proteins and associate with hormone response elements near endocrine induced genes in chromatin (Shepel and Gorski, BioFactors 1:71, 1988). No biologic response is possible unless both the hormone and the receptor protein are present in required quantities. It has been known for decades that many breast cancers require female sex hormones for their continued growth, i.e., they are hormonally dependent cancers. Some breast cancers, while not dependent upon estrogen for growth, may be inhibited by giving the patient high doses of estrogen-like compounds such as diethylstilbestrol and Tamoxifen (e.g., Fisher et al., J. Clin. Oncol. 1:227, 1983). Others continue to grow in the absence of large quantities of female sex hormones such as is the case in postmenopausal patients. Therefore, estrogen receptors (ER) and progestin receptors (PR) may be considered cellular markers of a breast cancer which is potentially endocrine responsive (Anonymous, Cancer 46:2759, 1980). Breast cancer patients most likely to respond to additive (administration of antihormones) or ablative endocrine manipulation (surgical removal of hormone-producing organs such as the ovary, adrenal and pituitary gland) may be identified from analyses of sex hormones receptors (e.g., Wittliff et al., IN: The Breast: Comprehensive Management of Benign and Malignant Diseases, K. I. Bland and E. M. Copeland III, eds., Ch. 43, W. B. Saunders, Philaelphia, 1991). The presence of sex-hormone receptors in a biopsy of either breast, endometrial or ovarian carcinomas is an index of good prognosis (greater disease-free and overall survival).
Receptor Polymorphism and Therapeutic Resistance
As many as 30-40% of human breast cancers are resistant to endocrine therapy in spite of the fact that they contain steroid receptors (cf Estrogen Receptors in Human Breast Cancer, W. L. McGuire, P. O. Carbone and E. P. Vollmer, eds., Raven Press, New York, 1975). Earlier we demonstrated that both ER and PR exhibit polymorphism (presence of isoforms) and suggested that their distribution may be related to endocrine responsiveness (e.g., Wittliff, Cancer 53:630, 1984). Using HPLC in various separation modes, we characterized the molecular heterogeneity of estrogen and progestin receptors and identified several variant forms (e.g., Wittliff et al., IN: The Use of HPLC in Receptor Biochemistry, p. 155, A. R. Kerlavage, ed., A. R. Liss Inc., New York, 1989). Recent studies have now shown that there are variant estrogen receptor mRNA molecules in certain human breast cancers using Northern blot analysis (Murphy and Dotzlaw, Mol. Endocrinol. 3:687, 1989). Similar findings have been made for progesterone receptor mRNA species with the exciting finding that certain progestational agonists autoregulate the levels of their own receptor proteins by inhibiting the transcription of the PR gene (Wei et al., Mol. Endocrinol. 2:62, 1988). Evidence from Gronemeyer et al. (J. Steroid Biochem. Molec. Biol. 40:271, 1991) suggests that certain antihormones such as RU-486 express their action depending upon the receptor isoform recognizing it.
It is now possible to use a variety of biochemical and molecular biological approaches to detect variant receptors with aberrant function in human breast cancer (e.g., McGuire, et al., J. Steroid Biochem. Molec. Biol. 43:243, 1992). The field is now sufficiently equipped to investigate the clinical significance of the expression of receptor variants in cancer.
Other Members of the Steroid/Thyroid Hormone Superfamily
There is increasing interest in receptors for other members of the superfamily such as 1,25 dihydroxyvitamin D3 in gynecologic neoplasms (Saunders et al., Gynecol. Oncol. 44:131, 1992). The Ah receptor, which appears to exhibit polymorphism, has now been shown to bind to halogenated aromatic hydrocarbons in a variety of species and produce tissue-specific toxic and biologic responses (Dennison, J. Biochem. Toxicol. 7:249, 1992). Androgen receptors have long been considered as predictive indices for selecting endocrine therapy for patients with prostate cancer (Benson et al., Cancer 59:1599, 1987). As a result of cloning these genes, many of the protein products are candidates as predictive indicators of therapeutic response to their cognate ligands, e.g., thyroid antagonists and the T3-T4 receptor. Retinoic acid and its analogs and their interrelationships with their receptors also appear to be of interest because of early studies of chemo-prevention.
Epidermal Growth Factor Receptors
Epidermal growth factor (EGF) is a single polypeptide chain with Mr = 6000 that is highly heat stable, particularly because of three disulfide bridges. EGF is known to interact with a number of organs in the body, including reproductive tissues. EGF receptor is a complicated molecule consisting of a large extracellular domain responsible for the association with EGF (Carpenter, Ann. Rev. Biochem. 48:193, 1979 and 56:881, 1987). The conformation of this domain is maintained by a large number of disulfide bridges and glycosylated residues. The transmembrane portion secures the receptor in the cytoplasmic membrane. The internal domain contains an ATP-binding site and exhibits tyrosine kinase activity. Interestingly, the receptor molecule autophosphorylates itself on tyrosine residues located near the carboxy terminus as well as on tyrosine residues of certain intracellular proteins when EGF occupies the binding site.
EGF receptors are present in certain breast, endometrial and ovarian cancers as detected by binding with radio-labeled EGF (Fitzpatrick et al., Cancer Res. 44:3448, 1984; Fekete et al., J. Clin. Lab. Anal. 3:137, 1989; Srkalovic et al., Cancer Res. 50:1841, 1990). The overexpression (i.e., increased numbers) of EGF receptors in a breast tumor biopsy appears to correlate with a shorter disease-free interval and with a decreased overall survival (Sainsbury et al., Lancet 1:1398, 1987). In contrast to steroid hormone receptors, high levels of EGF receptors in a breast tumor biopsy appear to be indicators of poor prognosis (e.g., Nicholson et al., Diagn. Oncol. 1:43, 1991). One of the most interesting new applications of EGF and its receptor is in the field of ovarian carcinoma where HER-2/neu levels appear to be controlled by this growth factor (Berns et al., Int. J. Cancer 52:218, 1992; Marth et al., Int. J. Cancer 52:311, 1992). These data indicate that EGF treatment down-regulates the synthesis of HER-2/neu oncoprotein.
c-erb B-2 (HER-2/neu) Oncoprotein
The neu oncogene, originally isolated from rat neuroblastomas, encodes a 185-kDa surface glycoprotein termed "p185 neu," and exhibits tyrosine kinase activity with a structure similar to the EGF receptor. When various molecular properties and chromosomal localization studies were conducted, it was revealed that the EGF receptor was distinct from p185 neu (Schecter et al., Science 229:976, 1985). The identity for the native ligand for p185 neu is the focus of research in many laboratories (Lupu, J. Steroid Biochem. Molec. Biol. 43:229, 1992).
The human homolog of c-neu gene, which is termed "c-erb B-2," is reported to be amplified (e.g., increase in copy number) in human breast cancer (Slamon et al., Science 244:707, 1989). This gene amplification has been correlated with decreased disease-free survival and lower overall survival. The oncogene has also been called HER-2/neu. Finally, overexpression of the HER-2/neu protein product also is correlated with amplification of the oncogene and is related to decreased disease-free survival and overall survival (Tandon et al., J. Clin. Oncol. 7:1129, 1989; Paik et al., J. Clin. Oncol. 8:103, 1990). Thus it appears that elevated levels of the oncoprotein are associated with poor prognosis in breast cancer patients. HER-2/neu gene overexpression in breast cancer appears to be associated with polyploid DNA content suggesting that these factors are interrelated (Bacus et al., Proc. AACR abs. 1079, 1991). The establishment of standardized methods for c-erb B-2 oncoprotein and the related EGF receptor by the Quality Assurance program of the Hormone Receptor Laboratory make these attractive candidates for incorporating into clinical trials.
Cathepsin D
Cathepsin D belongs to a class of acidic lysosomal proteases which are found in all cells. The cathepsin D gene is located at the extremity of the short arm of chromosome 11, near the H-ras oncogene (Rochefort, Cancer & Metastasis Rev. 9:321, 1990). The digestive pattern of cathepsin-D appears to be similar in breast cancer and normal breast. This interesting protease is estrogen induced and is secreted as a 52 kDa precursor. Pro-Cath-D, a phosphoglycoprotein is cleaved to mature forms of 48 kDa, 34 kDa and 14 kDa. Monoclonal antibodies have been produced against these forms and have been assimilated into an IRMA kit format (Rochefort et al., Cancer Cells, 2:383, 1990). Recent evidence suggests that over-expression of Cath-D in node negative breast cancer correlates with shorter disease-free survival and with decreased overall survival (Tandon et al., NEJM 322:297, 1990; Duffy et al., Clin. Chem. 38:2114, 1992; Kute et al., Cancer Res. 52:5198, 1992). These reports clearly indicate that expression of cathepsin D is related to prognosis of breast cancer patients. Therefore, we have established quality assurance materials to be used in the standardization of this test for clinical trials by cooperative groups such as the National Surgical Adjuvant Breast and Bowel Project and the Southwest Oncology Group.
Potential New Markers
Experimental and clinical studies suggest that analogues of luteinizing hormone-releasing hormone (LH-RH), such as Zoladex., Buserelin., Leuprolide. and [D-Trp6]LH-RH and of somatostatin such as Sandostatin. are useful for treatment of certain breast and endometrial cancers. Microanalytic methods of measuring LH-RH and somatostatin receptors in these tumors (Fekete et al., J. Clin. Lab. Anal. 3:1 37, 1989; Srkalovic et al., Cancer Res. 50:1841, 1990) now allow correlation of the levels of these regulatory proteins with clinical parameters to better identify endocrine-responsive carcinomas. Clinical trials may employ these receptor proteins as predictive tests for selecting patients for treatment with LH-RH and somatostatin analogs.
Other markers being investigated for their clinical utility include urokinase-type plasminogen activator (uPA) (Duffy et al., Cancer Res. 50:6827, 1990), the protooncogene p53 (Barbareschi et al., Am. J. Clin. Pathol. 98:408, 1992), Ki-67 growth fraction (Kamel et al., Am. J. Pathol. 134:107, 1989), pS2 protein (Foekens et al., Cancer Res. 50:3832, 1990), IGF-1 receptor (Foekens et al., Cancer Res. 49:7002, 1989; Bonneterre et al., Cancer Res. 50:6131, 1990), c-myc expression (Pavelic et al., Cancer Res. 52:2597, 1992), NM-23 (Klijn et al., J. Steroid Biochem. Molec. Biol. 43:211, 1992), certain heat-shock proteins and topoisomerases.
This century is closing with a greater use of clinical chemistry to ascertain the biology of cancerous lesions in order to improve the therapeutic management of these unfortunate patients. Previously, endocrinology consisted of measurements of the hormones themselves with little appreciation of the role of their receptor proteins. Advances in biochemistry and molecular biology now indicate that the primary molecule expressing a hormone's activity is the receptor. Furthermore, since the protein products of many oncogenes are involved in signal transduction and hence the differentiation and growth of tumors, this field has expanded considerably. Our success in treating cancer will be dictated in many ways by our ability to understand and control these regulatory events.
Summary of the Use of Prognostic Factors in Breast Cancer
1. In addition to serving as predictive indicators of hormonally-responsive carcinomas, elevated estrogen and progestin receptors in a breast biopsy is correlated with a longer disease-free interval and overall survival (better prognosis) of breast cancer patients.
2. Elevated levels of EGF receptors and the protein product of the c-erb B-2 oncogene in a breast carcinoma biopsy are correlated with a shorter disease-free interval and overall survival of patients with breast carcinoma.
3. Elevated levels of the lysosome-associated protease, cathepsin D, in a carcinoma biopsy is correlated with decreased disease-free interval and overall survival compared to breast cancer patients whose tumors lack this protease.