Endocrine management of breast cancer—biology and current practice
- First Online:
- Cite this article as:
- Schneider, H.P.G. & Jackisch, C. Eur Clinics Obstet Gynaecol (2005) 1: 102. doi:10.1007/s11296-005-0016-3
- 13 Views
Breast cancer frequency is related to age. There are impressive advances in the diagnostic armament and surgical techniques of breast cancer, and yet, it has continued its deadly impact. In women with operable breast cancer, the histologic status of the axillary lymphnodes remains the most useful prognostic information. Today, breast cancer is viewed as a systemic disease with spreads to local and distant sites at the same time. There is a high rate of occult disease. Factors of major influence on breast cancer include reproductive experience, ovarian activity, benign breast disease, familial tendency, genetic differences, dietary considerations and specific endocrine factors. Tamoxifen still is the standard endocrine treatment for hormone-receptor-positive breast cancer both in the adjuvant and metastatic settings. Because of its weak oestrogenicity, tamoxifen may not be optimally effective and can increase the risk of endometrial cancer and stroke. Patients may also be refractory or become resistant to tamoxifen treatment. Aromatase inhibitors block the synthesis of oestrogen, have low intrinsic oestrogenic activity and, thereby, offer the potential of being more effective than tamoxifen. The challenge of the coming years is to identify women who are best treated with an aromatase inhibitor upfront rather than a tamoxifen-to-aromatase inhibitor sequence. Third-generation aromatase inhibitors offer greater benefits in oestrogen receptor (ER)-positive, progesterone-negative and HER-2 overexpressing tumours. Overall, no more than 2–3% of breast cancer occurrences can be attributed to the inherited mutation of either maternal or paternal origin. Autocrine and paracrine regulation of local oestrogen biosynthesis in normal and tumour breast tissue is via growth factors acting upon aromatase activity, which is preferably expressed in the tumour-bearing quadrant. Aromatase regulation operates against a concentration gradient of oestrogens, comparing peripheral plasma to local tissue levels in both premenopausal and postmenopausal women. Clinical observations suggest to routinely determine both ER subtypes and its variants, by which, the prediction of the response of the breast tumours to endocrine therapy may improve. Polymorphisms of cytochrome CYP-17, CYP-1A1 and COMT are found to be associated with an increased risk of breast cancer. To identify high-risk genotypes in women may delineate the individual at increased risk of breast cancer. Arguments in favour of an impact of postmenopausal oestrogen–progestin therapy on pre-existing tumours derive from observations in epidemiologic studies showing no increase in non-invasive breast cancer with hormone therapy; on the other hand, several studies point to a lower grade and earlier stage of disease in hormone users, with subsequently better survival rates. The older a postmenopausal woman, the greater her risk of developing an increase in breast density with hormone therapy. This is considered a good reason to recommend discontinuation of hormone therapy for at least 2 weeks prior to mammography in women older than age 65 who have dense breasts. As yet, there is no endocrine or biological evidence as to a sequential use of non-cross-resistant endocrine therapies. After progression on adjuvant and first-line tamoxifen, ovarian ablation is an appropriate second-line therapy. For premenopausal women with ovarian ablation, the use of a third-line therapy with an aromatase inhibitor appears feasible. In postmenopausal women, a wide choice of endocrine treatment options is available and an optimal sequence has yet to be determined. The widening range of adjuvant endocrine options represents an opportunity to prolong patient benefits in the treatment of hormone-receptor-positive breast cancer. Further refinement is currently investigated. For example, tamoxifen therapy followed by an aromatase inhibitor may lead to a reduction in endometrial pathologies associated with tamoxifen. New adjuvant treatments with chemotherapy or endocrine agents is being used increasingly to down-stage locally advanced and large operable breast cancers, which often become fully resectable; in addition, initially operable tumours requiring mastectomy may be successfully removed by breast-conserving surgery. Patient selection is important to optimise new adjuvant endocrine therapy; only patients with oestrogen-receptor-rich breast cancer are candidates, and postmenopausal women are likely to benefit the most.
KeywordsReproductive experienceEndocrinologyClinical managementNew developmentsFuture perspectives
George Beatson, a Scottish surgeon, reported on his experience of a remission of breast cancer following bilateral oophorectomy in premenopausal women more than a century ago . Ever since, the possible relationship of ovarian function and mammary tumourigenesis never escaped our clinical conscience. The major influence on breast growth at puberty is oestrogen. Increasing levels of oestrogen first result in an increase in the size and pigmentation of the areola and in the formation of a mass of breast tissue just underneath. Similar to the uterus and the vagina, breast tissue expresses both oestrogen receptors (ER), ERα and ERβ . The expression of ERs will not occur in the absence of prolactin. While oestrogen, in some primate mammals, stimulates growth of the ductal portion of the glandular system, progestin influences the growth of the alveolar components of the lobule. Neither oestradiol nor progesterone alone, or in combination, could provide optimal breast growth and development. Further differentiation of the glandular breast requires insulin, cortisol, thyroxine, prolactin and growth hormone . A large variety of growth factors is also involved, but their molecular mechanisms have yet to be determined. Experimental evidence in rodents proves that progesterone is the key hormone required for mammary growth and differentiation, while oestrogen is necessary because progesterone receptors will only be synthesised in the critical presence of oestrogen .
Epidemiology of breast cancer
The lifetime probability of developing breast cancer varies from 12.5% in the US to less than 5% in some Eastern European and Asian countries. Thus, the top-ranking United States has a probability of 1 in 8, which is double the risk in 1940 . Typical Western European figures, such as those for Germany, are 1 in 10 . Since the beginning of the 1990s, breast cancer incidence has plateaued; increases were limited to women over the age of 50 and a rate of about 0.4% per annum, and were also limited to localised disease. Mortality rates remained constant for a long time, until they began to decline in the 1990s. Localised breast cancer constitutes about 60% of all breast cancers; its 5-year survival rate has risen from 72% in the 1940s to 97% . The major reason for this is earlier diagnosis because of greater utilisation of screening mammography and increased use of chemotherapy, and also, a continuing decline in mortality. The 5-year survival rate for breast cancer with regional spread is 79%; with distant metastases, the rate is 22%. The breast is the leading site of cancer in Western countries (about one-third of all cancers), and has only recently been exceeded by lung and bronchial cancer as the leading cause of death from cancer in women .
Breast cancer frequency is related to age. While 94% of all breast cancers occur in women over the age of 40, around 6% of all cases are apparent below the age of 40 and 15% under the age of 50 . The overall 5-year survival rate for women who develop breast cancer under the age of 45 is 83%, compared to 88% for women aged 65 and older.
Despite impressive advances in surgical and diagnostic techniques, breast cancer has continued its deadly impact. With operable breast cancer, the histologic status of the axillary lymph nodes remained the most useful prognostic information [8, 9]. Negative axillary lymph nodes are predicative of higher survival rates. Therefore, breast cancer was viewed as a disease of stepwise progression, a concept on which surgeons based their procedures. This concept has been challenged by viewing breast cancer as a systemic disease, which spreads to local and distant sites at the same time. There is occult metastatic disease at the time of presentation. In many cases, the dissemination of tumour cells has occurred before surgery. Surgery is curative in cancers prior to invasion, which are not systemic at the time of diagnosis.
Risk factors for breast cancer
Relative risk >4.0
Older than 65 years of age
Two or more first-degree relatives with early disease
Dense postmenopausal breast
Relative risk 2.1–4.0
One first-degree relative
Atypical hyperplasia (histol.)
High-dose thorax radiation
High postmenopausal bone density
Relative risk <2.1
First-term pregnancy after 30 years of age
Menarche before 12 years of age
Menopause after 55years of age
No experience of breast-feeding
History of carcinoma (endometrium, ovary, colon)
Daily alcohol intake
Reproductive experience and breast cancer
As the great majority of women (85%) who develop breast cancer do not have an identifiable risk factor other than age, thus, every woman might be considered at risk. The most evident additional impact is individual reproductive experience.
Impact of reproduction
The risk of breast cancer is augmented with the increase in the age at which a woman delivers her first full-term child. A woman bearing her first child before the age of 18 has about one-third the risk of one who first delivers after the age of 35. To be protective, pregnancy must occur before the age of 30. Indeed, women over the age of 30 years at the time of their first delivery have a greater risk than women who never become pregnant . While we observe an increasing risk with increasing age , there is, nevertheless, a significant protective effect with increasing parity, even when adjusted for age at first birth and other risk factors . Considering the increased incidence of breast cancer over the last decades, delayed child-bearing and fewer children may have contributed significantly to this phenomenon.
A large case-control study reported that pregnancy transiently increases the risk of breast cancer for up to 3 years after a woman’s first childbirth, but this will then be followed by a lifetime reduction in risk . It has also been reported that a concurrent or recent pregnancy (3–4 years previously) adversely affects survival, even when adjusting for the size of tumour and the number of nodes . This may reflect accelerated growth of an already present malignancy; breast cells that have already begun malignant transformation are adversely affected by the hormones of pregnancy, while normal stem cells become more differentiated and resistant. These effects are very likely mediated by oestrogen and progesterone. There is also, however, evidence for the presence of luteinising hormone (LH) receptors in breast tissue and the potential of human chorionic gonadotropin (HCG) to contribute to the protective differentiation of breast cells .
When breast cancer is diagnosed during pregnancy, tumour prognosis is independent of the time at which the pregnancy was terminated . In theory, full-term pregnancy may protect against breast cancer by invoking complete differentiation of breast cells, while abortion increases the risk by allowing breast cell proliferation in the first trimester of pregnancy, but not the full differentiation that occurs in later pregnancy. Therefore, the question whether the risk of breast cancer is associated to the number of abortions experienced by individual patients is hampered by a problem of recall bias; women who develop breast cancer are more likely to reveal their history of induced abortion than healthy women. If this recall bias is avoided by looking at data from national registries rather than personal interviews, the risk of breast cancer was identical in women with and without induced abortions [17, 18].
The breast cancer protection afforded only by the first full-term pregnancy suggests a trigger effect, producing a permanent change in the breast tissue and making it less susceptible to malignant transformation. The first pregnancy also has a lasting impact on a woman’s hormonal milieu. A small but significant elevation of oestriol, a decrease in dehydroepiandrosterone and dehydroepiandrosterone sulphate, and lower prolactin levels persist for many years after delivery [19, 20].
Lactation may offer a moderate protective effect (20% reduction) on the risk of breast cancer . However, both the Nurses‘ Health Study and a Norwegian prospective study with a long duration of breast-feeding did not find a benefit on either premenopausal or postmenopausal breast cancer incidence [22, 23]. If any, the impact of lactation must be small. Worldwide-available data suggest a reduction of the risk of breast cancer by 4.3% per year of breast-feeding, together with a potential reduction in accumulative incidence by the age of 70 of more than 50% . A French meta-analysis carried out in 2000 indicated that breast-feeding reduces the risk of breast cancer by about 10–20%; however, this effect was limited to premenopausal women .
Familial tendency—family history
Most cases of breast cancer arise in individuals without a specific family history. In contrast to these sporadic breast cancers, affected female relatives have about twice the rate of the general population. While there is an excess of bilateral disease in patients with a family history, their relatives have about a 45% lifetime chance of developing breast cancer. Clinically speaking, one should emphasise that most women with an affected relative will never have breast cancer. The figures show that only 1 in 9 women with breast cancer has an affected first-degree relative.
The average woman of 50 years of age has a risk of only 1–2% of being diagnosed with breast cancer in each subsequent 5-year period of her life. Breast cancer will be the cause of death for less than 5% of women who survived to 50 years of age without a history of breast cancer. Approximately 75% of women diagnosed with breast cancer today will survive the disease. The breast and ovarian cancer gene (BRCA1) associated with familial cancer is localised to 17q12-q21. Mutations in BRCA1 are believed to be responsible for approximately 20% of familial breast cancer and 80% of families with both early-onset breast and ovarian cancer. Other genetic alterations have also been observed in breast tumours. Overall, no more than 2–3% of breast cancers can be attributed to inherited mutation ; this inheritance can be either maternal or paternal; male carriers experience an increased risk for colon and prostate cancers . Another important locus, BRCA2, on chromosome 13q12-q13, accounts for up to 35% of families with early-onset breast cancer, but with a lower rate of ovarian cancer . BRCA1 and BRCA2 together account for 80% of families with multiple cases of early-onset breast cancer .
Mutations in a dominant breast-cancer-susceptibility gene account for less than 5% of breast cancer cases in the general population. The combined presence of ovarian cancer and three or more cases of breast cancer within a family are strong predictors of BRCA1 mutations. Genetic screening should be reserved for patients from high-risk families. Family history characteristics associated with the presence of BRCA1 are early-age-onset of breast cancer within a family, relatives with ovarian cancer, three or more relatives with breast cancer or Ashkenazi ancestry. Identification and counselling for families who have the appropriate history but fail to demonstrate BRCA1 or BRCA2 mutations should be exactly the same as when the mutations are found.
So, what to do in a high-risk family situation? If high-risk women undergo prophylactic mastectomy, they will experience a major reduction but not total prevention of breast cancer, since the mutation is present in every cell and prophylactic mastectomy does not remove all tissue . Similarly, following prophylactic oophorectomy, a carcinoma may arise from peritoneal cells. Clinical breast examination and mammography are recommended every 6–12 months, beginning between the ages of 25 and 35. An evaluation every 6 months is appropriate, as BRCA1-related tumours have been demonstrated to be faster-growing tumours. Proper counselling and support should be offered to those women who might be candidates for prophylactic mastectomy. Pelvic examination, serum CA-125 levels and transvaginal ultrasonography with colour Doppler imaging are recommended every 6–12 months for women under the age of 40. Prophylactic oophorectomy and hysterectomy are recommended after the completion of childbearing, preferably before the age of 35. Following surgery, oestrogen-only therapy is appropriate and acceptable. However, the effect of oral contraceptives is unsettled; there are contradictory reports as to whether the risk of ovarian cancer is reduced or remains unaffected [31, 32]. A larger case-control study concluded that BRCA1 (but not BRCA2) mutation carriers had small increases in the risk of breast cancer in oral contraceptive users for at least 5 years with an odds ratio of 1.33 and a confidence interval of 1.11–1.60 .
Specific endocrine factors
There have been previous studies indicating that subnormal levels of etiocholanolone were found from 5 months to 9 years before the diagnosis of breast cancer in women living on the island of Guernsey . Low levels of androsterone and etiocholanolone were observed to correlate with an increase in breast cancer only in women under the age of 50. Over the age of 50, the reverse was true . Despite these rare observations, specific endocrine factors predominantly relate to endogenous oestrogen and progesterone.
Many lines of evidence suggest some oestrogen-related promoter function for the development of breast cancer. The response of an organ to the proliferative effects of a hormone may be a progression from normal growth to hyperplasia to neoplasia. Accordingly, the risk of breast cancer could be determined by the cumulative exposure of breast tissue to oestrogen . This contention is supported by: (1) the condition is 100 times more common in women than in men; (2) breast cancer invariably occurs after puberty; (3) untreated gonadal dysgenesis and breast cancer are mutually exclusive; (4) a 65% excess rate of breast cancer has been observed among women who have had an endometrial cancer; (5) breast tumours contain ERs, which are biologically active, as indicated by the presence of progesterone receptors in tumour tissue . While such observations suggest an element of oestrogen dependence, a higher oestriol level is hypothesised to actually protect against the more potent effects of oestrone and oestradiol. This may explain the protective effect of early pregnancy. Consistent with this observation are also the higher rates of urinary oestriol excretion in healthy premenopausal Asiatic women with their lower risk of breast cancer compared to Caucasians .
The predictive value of reproductive risk factors of breast cancer is increased by combining them. As an example, individual age and age at first full-term birth would not only reflect the total exposure to oestrogen, but also the effect of sex steroids on the final differentiation of the glandular breast induced by pregnancy and lactation as major determinants of susceptibility to cancer . Other contributing factors to individual variation in exposure to oestrogen are weight changes, differences in exercise and dietary intake of certain nutrients. In overweight premenopausal women, the risk of breast cancer is lower compared to normal-weight individuals; however, in postmenopausal women, excess weight is associated with either an unchanged or slightly increased risk . This may be related to a more marked increase in total and free oestrogen levels in overweight postmenopausal women compared to lower levels with increasing weight in premenopausal women. The epidemiologic literature provides little support for a major contribution of dietary fat to the risk of breast cancer. Nevertheless, there is a correlation between intra-abdominal fat and the risk of breast cancer as a consequence of excessive caloric consumption but, however, not a specific dietary component . Altogether, studies of intakes of alcohol, fat, antioxidant vitamins and fibre have produced conflicting results. Phyto-oestrogens, with their structural similarity to physiologic oestrogens, when ingested, have both oestrogen-agonist and oestrogen-antagonist effects in humans. Flaxseed, as a source of mammalian lignanes and α-linoleic acid, has been shown to exert anti-oestrogenic effects by binding to ERs and inhibiting the synthesis of oestrogen. The incidence of breast cancer is the lowest in regions where the intake of soy, an abundant source of phyto-oestrogens, or of flaxseed is high. Whether or not this inverse relation is direct or only indicative of other influencing factors is a matter of debate .
Breast tissue metabolism of oestrogen
Breast cancer tissue contains all the enzymes necessary for the formation of oestradiol from circulating precursors, including aromatase, sulfatase and 17β-hydroxy steroid dehydrogenase (17β-HSD) [43–45]. Oestradiol formation in normal breast and breast cancer tissues occurs via two main pathways: the “aromatase pathway,” which transforms androgens into oestrogens, and the “sulfatase pathway,” which converts oestrone sulfate (E1S) into oestrone (E1), which is then transformed into E2 by reductive 17β-HSD activity.
Autocrine and paracrine regulation of local oestrogen biosynthesis in normal and tumour breast tissue is via growth factors acting upon aromatase activity; this enzyme is preferably expressed in the tumour-bearing quadrant of the breast compared to distant areas of the same quadrant or other quadrants. Apparently, aromatase regulation operates against a concentration gradient of oestrogens, comparing peripheral plasma to local tissue levels both in premenopausal and postmenopausal women [46, 47].
Oestradiol levels in breast tumours of postmenopausal women remain as high as in the premenopause, while the plasma levels decrease. This clearly points to the discrepancy between the two compartments and would implicate the necessity of mechanisms that require local factors. Several other studies had confirmed such observations and underline the hypothesis that local production of oestradiol is the source of the steroid at breast level . Androstenedione was found at lower concentrations in the tumour compared to fatty tissues of all quadrants; such differences were not seen with testosterone. Finally, E1S is highly concentrated in the tumour. These observations are reconciled with the fact that androstenedione is the major precursor for local oestrogen synthesis. It also points to the importance of local aromatase activity, which was comparable in all tissues, in contrast to 17-OH HSD, of which differences were observed. We measured this enzyme by substrate-to-product conversion; no specific type has been distinguished. One should, however, bear in mind that these laboratory estimates cannot be easily extrapolated to real activity in the tissue because stimulatory and inhibitory factors do play additional roles. Furthermore, the promoter for aromatase in tumours is different from that in fatty tissues. Quantitative evaluation indicates that, in human breast tumours, E1S via sulfatase is 100–500 times higher a precursor for E2 than androgens via aromatase [48, 49].
Biosynthesis of oestrogen
According to current concepts, both cytochrome CYP-17 (encoding P-450 17α-hydroxylase) and cytochrome CYP-19 (encoding P-450-aromatase) are involved in oestrogen biosynthesis; polymorphisms of both genes have been identified in the general population. Women who are heterozygous or homozygous for a cytochrome CYP-17 polymorphism have been shown to produce high serum oestradiol concentrations; however, this polymorphism is not unequivocally associated with increased risk of breast cancer . Ongoing studies demonstrate a link between polymorphisms of the P-450-aromatase gene with an increased risk of breast cancer . Oestrogen production may also be influenced by a variation in tissue-specific promoters of aromatase gene expression . A detailed investigation on the expression of aromatase in human breast tumours  has demonstrated a change in promoter I.4 in normal breast tissue to promoter PII and PI.3 in breast cancer, which are more active and may result in increased synthesis of aromatase messenger ribonucleic acid (mRNA).
Functional analysis of promoters will be essential for a clear understanding of the control of aromatase expression in breast tumours and its role in cancer development, and may involve transcription factors specific to breast cancer cells contributing to the growth of breast tumours in an autocrine or paracrine fashion. The aromatase gene may finally act as an oncogene that initiates tumour formation in breast tissue .
Breast tissue sensitivity to oestrogen
Oestrogen may diffuse passively through cellular and nuclear membranes. However, specific cells and tissues express ERs to which oestrogen would bind to and form a ligand-receptor complex in order to activate specific sequences in the regulatory region of genes responsive to oestrogen, known as oestrogen-response elements. These genes in turn regulate cell growth and differentiation.
New discoveries regarding the mechanism of oestrogen action represent one of the most important scientific advances of today. Not only do oestrogen cells act differently from tissue to tissue and from cell to cell, but there are also variations among individual women. Physiologically active doses in one individual may produce less of an effect in another. ER levels are low in normal breast tissue, and high levels have been directly correlated with an increased risk of breast cancer . Receptor levels increase with age in some ethnic groups and, apparently, are higher in white women compared to black or Japanese women. This phenomenon may be related to the function of a tumour-suppressor gene, the loss of which may result in failure to down-regulate ERs, with resultant defects of the cell cycle and, finally, driving breast carcinogenesis .
The human ER belongs to the nuclear receptor superfamily of ligand-inducible transcription factors. The identification of ERβ has indicated that the cellular responses to ER ligands are far more complex. ERα and ERβ interact with the same DNA response elements and exhibit similar, but not identical, ligand-binding characteristics. ERβ binds oestrogen cells with a similar affinity to ERα and activates the expression of reporter genes containing oestrogen response elements in an oestrogen-dependent manner. In vitro, the α and β receptors form heterodimers with each other, and the β receptor decreases the sensitivity of the α form to oestrogen, thereby, acting as a physiologic regulator of the proliferative effects of the α receptor .
In order to evaluate the role of different ERs in breast cancer, the expression of both ER isoforms in normal and malignant breast tissue has been investigated. In normal breast tissue, the expression of ERβ predominated, with 22% of samples exclusively expressing ERβ; this was not observed in any of the breast tumour samples . Most tumours expressed ERα, either alone or in combination with ERβ. These tumours that co-expressed ERα and ERβ tended to be node-positive and of higher grade, indicating a trend towards an association with more poorly differentiated tumours. In tissue samples with co-expression of both receptor subtypes, the expression of ERα was always greater than that of ERβ. Further studies are under way to find out whether high levels of ERβ are predominantly expressed in normal human breast and not so in breast cancer cell lines. A most recent interim report  investigated the expression of ERβcx, which has a dominant-negative effect over ERα. The ERβcx variant has been found to be expressed in the ovary, testis, prostate and thymus. It preferentially forms heterodimers with ERα and ERβ; whereas ERα is inhibited by ERβcx, ERß1 is unaffected, suggesting that ERβcx can act as a dominant-negative inhibitor of ERα. The often contradictory results published regarding the role of ERβ in breast cancer could be related to the presence or absence of a number of splice variants, which exert dramatically different biological effects, in particular, ERβcx. Therefore, the groups of Gustafsson and Coombes investigated the expression of ERβcx in 82 frozen breast samples of benign, ductal carcinoma in situ and invasive cancer origin . There was a statistically significant association between the presence of ERβcx and the response to endocrine therapy. There was also a relationship between the presence of ERβcx and survival, with patients whose tumours express ERβcx having a longer survival rate.
Such observations suggest to routinely determine both ER subtypes and its variants. This may be helpful in predicting the response of the breast tumours to endocrine therapy, particularly as selective ER modulators (SERMs) would depend on a possible agonistic effect on the ERβ-type expression.
Catabolism of oestrogen
Oestrogen cells are catabolised predominantly by hydroxylation, with a resultant formation of 2-hydroxy-oestrone and 2-hydroxy-oestradiol, 4-hydroxy-oestrone and 4-hydroxy-oestradiol, and 16α-hydroxy-oestrone and 16α-hydroxy-oestradiol . The 2-hydroxy and 4-hydroxy metabolites are converted to anticarcinogenic methoxylated metabolites (2-methoxy-oestrone and 2-methoxy-oestradiol, 2-hydroxy-oestrone and 2-hydroxy-oestradiol-3-methylether, 4-methoxy-oestrone and 4-methoxy-oestradiol, and 4-hydroxy-oestrone and 4-hydroxy-oestradiol-3-methylether) by catechol O-methyltransferase (COMT). Some of these catechol metabolites of oestrogen cells are implicated with potential carcinogenic and cytotoxic effects . They are further metabolised to electrophillic quinoids, such as o-quinones, which can isomerise to their tautomeric p-quinone metides; the roles of these quinoids in mediating the adverse effects of oestrogens have not been investigated in detail. It is possible for these electrophillic and redox active quinoids to cause damage within cells by a variety of pathways. Catechol oestrogen-mediated redox cycling can cause lipid peroxidation, consumption of reducing equivalents, oxidation of DNA and DNA single-strand breaks .
Postmenopausal women with a variant allele that codes for a COMT with low activity have a higher risk of breast cancer than women with a wild-type allele . On the other hand, 17β-hydroxysteroid dehydrogenase activity is higher in breast tumours than in normal breast tissue .
Taking these tissue-specific variations of oestrogen production and catabolism into consideration, there is reason to believe that cumulative exposure to oestrogen and its metabolites may vary distinctly within individual women. Polymorphisms of cytochrome CYP-17, CYP-1A1 and COMT are found to be associated with increased risk of breast cancer . To identify high-risk genotypes in women may delineate the individual at increased risk of breast cancer.
Experimental evidence indicates that, with increasing duration of exposure, progesterone can limit breast epithelial growth, as it does with endometrial epithelium [61–63]. In vitro studies of normal breast epithelial cells reveal that progestins exhibit proliferation . Human breast tissue specimens removed by reduction mammoplasty after the patients were treated with oestradiol and progesterone indicate that progesterone inhibits in vivo oestradiol-induced proliferation [61, 63]. Women who ultimately develop breast cancer do not have different blood levels of progesterone . Most studies indicate that high levels of oestrogen and progesterone during pregnancy have no inverse effect on the course of breast cancer diagnosed during pregnancy or when pregnancy occurs subsequent to diagnosis and treatment. This would argue against progesterone as a key risk factor.
Therapeutic oestrogen and progestin
Numerous epidemiologic studies have indicated a small increase in the risk of breast cancer associated with postmenopausal oestrogen–progestin therapy. The key research issue is whether postmenopausal hormone therapy initiates the growth of new breast cancers or whether the epidemiologic data reflect an impact on pre-existing tumours. Arguments in favour of the impact on pre-existing tumours derive from observations that, in the epidemiologic studies, no increase of non-invasive breast cancers has been reported; several studies also point to a lower grade and stage of disease in hormone users with subsequently better survival rates. The increase in breast density seen with postmenopausal hormone therapy is transient, reversible and not consistent with a continuing effect on cellular proliferation. Following discontinuation of hormone therapy, breast density rapidly disappears [66, 67]. This regression of hormone-induced abnormalities was found to occur within 2 weeks of cessation of treatment .
The older a postmenopausal woman, the greater her risk of developing an increase in breast density with hormone therapy. This is a good reason to recommend discontinuation of hormone therapy for 2 weeks prior to mammography in women older than the age of 65 who have dense breasts.
Clinical endocrine management of breast cancer
The presence of both oestrogen and/or progesterone receptor is predictive of the response to endocrine therapy . Premenopausal and younger patients are more frequently diagnosed as receptor-negative. Women with receptor-positive tumours survive longer and, following mastectomy, have longer disease-free intervals when compared to those with receptor-negative tumours. This correlation of positive ERs with a disease-free interval is independent of the presence of positive axillary nodes or size and location of the tumours. Also, node-negative and oestradiol receptor-negative breast tumours have the same high rate of recurrence as do patients with axillary lymph node metastases.
Apparently, patients with ERs are those with the more slowly growing tumours. This is consistent with a correlation of the ER status to the degree of differentiation of the primary tumour. Commonly, highly differentiated grade-I carcinomas are receptor-positive in contrast to grade-III tumours.
The presence of progesterone receptors also correlates with disease-free survival and does so only second to the number of positive nodes . As progesterone receptors depend on oestrogen and are indicative of biologically active ERs, the absence of progesterone receptors is considered to be an ominous sign. Complete absence of oestrogen and progesterone receptors by immunohistochemic staining is indicative of less favourable tumours, which require chemotherapy as the only treatment option available for this subset of patients.
A meta-analysis of the Early Breast Cancer Trialists’ Cooperative Group (EBCTCG) in 1995  looked into 2,000 patients of less than 50 years of age with a 15-year overall survival. Among these, the absolute benefit of bilateral ovarectomy versus not was 6% (52% vs. 6%); the relative benefit was 13%. Most of these women did not receive any chemotherapy. The Zoladex Early Breast Cancer Research Association (ZEBRA) trial introduced a GnRH agonist goserelin and compared its effect to CMF chemotherapy in 1,640 node-positive premenopausal patients. There was no difference seen in the progression-free or overall survival. The question was raised as to whether the effect of chemotherapy is largely endocrine-based or not [70, 71].
The evidence for ovarian ablation to be an effective treatment option in premenopausal women with enocrine responsive breast cancer has recently been investigated by various authors [72, 73]. Based on the outcome of these trials, the standard of care for premenopausal patients with endocrine responsive breast cancer is the administration of GnRH analogues for at least 2–3 years in combination with tamoxifen (20 mg) for 5 years. Particularly, women aged 40 and younger benefit the most, regardless of nodal status. If chemotherapy is indicated, this treatment option should be administered in sequence to chemotherapy .
Tamoxifen and other selective oestrogen receptor modulators
Tamoxifen was the first clinically available SERM. Tamoxifen’s structure is really related to clomiphene, a non-steroidal compound with similarity to diethylstilboestrol. Tamoxifen, in binding to the ER, competitively inhibits oestrogen binding. As tamoxifen has only 100–1,000 times less binding affinity of oestrogen in vitro, it must be more concentrated by that order to maintain the inhibition of breast cancer cells. When bound to the ER, tamoxifen prevents gene transcription by the TAF-2 pathway. In vitro, its actions are not cytocidal but, rather, cytostatic and, therefore, must be used long term. Tamoxifen failed to demonstrate increased activity with doses larger than 20 mg daily.
With more than 10 million women years of use, tamoxifen has established adjuvant therapy of breast cancer. The latter is defined to provide treatment in the absence of recognised active disease in order to reduce the risk of future recurrence or to minimise systemic recurrence in the presence of metastatic disease. Adjuvant treatment with tamoxifen achieves a highly significant improvement in recurrence, free survival and overall survival in postmenopausal women with breast cancer. The Early Breast Cancer Trialists’ Collaborative Group analysed 37,000 women from 55 randomised trials. After 5 years of therapy and a median follow-up of 10 years, a 27% reduction in tumour recurrence compared to placebo and a 26% reduction in mortality compared to placebo was observed. The beneficial effect of tamoxifen is evident no matter what the age of the patient, in both premenopausal and postmenopausal women, in node-positive and node-negative disease; in the latter, the effect of tamoxifen is rather small. The 10-year survival difference is greater than that at 5 years. If one summarises the adjuvant treatment effects of tamoxifen, chemotherapy or ovarian ablation, an extra 100,000 10-year survivors will be seen worldwide. The responses in advanced breast cancer are 30–35%, which is most marked in women with oestrogen-positive tumours; they may reach 75% in patients with tumours highly positive for ERs.
Randomised clinical trials have documented that a treatment duration of 5 years is superior to 2 years . There is no reason to extend tamoxifen treatment of breast cancer patients beyond 5 years .
Major disturbing cell effects of tamoxifen include an increase in hot flushing, cases of developed reversible retinopathy, more venous thrombi, endometrial cancer (6.3 per 1,000 patients after 5 years of treatment) and cataracts. As far as the oestrogenic activity of tamoxifen is concerned, 20 mg daily are nearly as potent as 2 mg oestradiol in lowering follicle-stimulating hormone (FSH) levels in premenopausal women . Those oestrogenic actions of tamoxifen include the stimulation of progesterone receptor synthesis, an oestrogen-like maintenance of bone and the cardiovascular system, and oestrogenic effects on the vaginal mucosa and the endometrium.
A large breast cancer prevention trial was initiated in the United States in 1992 . The study compared two groups of women; one treated with placebo and one with 20 mg tamoxifen daily for 5 years. Early in 1998 (after about 4 years of follow-up), the study was unblinded, since there were 49% (p<0.00001) fewer cases of invasive breast cancer; the incidence of non-invasive breast cancer was reduced by 50% (p=0.002). The risks observed were a 2.4-fold increase in postmenopausal endometrial cancer, a 2.8-fold increase in pulmonary embolism, a 1.6-fold increase in venous thrombosis and a 1.6-fold increase in cataracts. Another concern is that malignant breast tumours require a relatively long period of time to progress from the transformed cell to a clinically detectable tumour . The impact of tamoxifen may, thereby, reflect growth deceleration of a pre-existing tumour. A major argument against this assumption of an effect on pre-existing tumours is the observed persistent protection against recurrent disease over 10 years of tamoxifen adjuvant therapy follow-up . Interestingly, in the American preventive trial, tamoxifen reduced the incidence of breast cancer among BRCA2 carriers, but not in BRCA1 carriers; this may reflect the fact that most of the BRCA2 carriers had ER-positive tumours in contrast to the BRCA1 patients.
What is the balance of tamoxifen prevention and its associated risks? Apparently, the percentage of women achieving a net benefit is relatively small. Newer epidemiological studies from England, Italy and Australia reviewed the combined results of breast cancer prevention trials and added updated results .
Additionally, there are data available on the effect of raloxifene. Raloxifene is an anti-oestrogen that is registered for therapy for osteoporosis. In the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, investigating raloxifene prophylaxis against fractures in women with osteoporosis, the effect of raloxifene on the risk of invasive breast cancer was also investigated . Looking at ER-positive breast cancer, among a total of 41 cases, the annual rate per 1,000 was reduced from 3.71 to 0.59; this corresponds to a relative risk of 0.16 (95% CI=0.09–0.30). This impressive effect was not seen in patients with ER-negative invasive breast cancer (relative risk=1.13, CI =0.36–3.66) among 13 cases.
The experience with tamoxifen prophylaxis (20 mg daily for 5 years) has led to its recommendation for women with carcinoma in situ or atypical hyperplasia in a breast biopsy. In all other situations, the final answers are not conclusive. However, women at very high risk of breast cancer who choose tamoxifen treatment deserve support and the appropriate surveillance.
Aromatase P-450 is a member of the P-450 superfamily and a product of the CYP-19 gene. It converts C19 androgens to C18 oestrogens and is expressed in the ovary, testes, placenta, adipose mesenchymal cells, osteoblasts and chondrocytes of bone, vascular smooth muscle and endothelium, as well as numerous sites in the brain. The tissue-specific expression of aromatase is regulated by the use of tissue-specific promoters via alternative splicing. The aromatase inhibitors include anastrozole (Arimidex), letrozole (Femara) and exemestane (Aromasin). Randomised clinical trials indicate that treatment with an aromatase inhibitor yields a better clinical response compared to tamoxifen; there is slightly less hot flushing and no risk of abnormal endometrial change as indicators of smaller side effects [81, 82]. However, there are also major problems. There is an increase in fractures due to profoundly low oestrogen levels (almost a 99% decrease) and bone loss; this risk might be prevented by bisphosphonate treatment. In addition to hot flushing, other major side effects are arthritic complaints, reduced sexual function and myalgia . When tamoxifen therapy was followed by letrozole, an increase in disease-free survival could be observed . Another report pointed to better outcomes of exemestane after the standard therapy of 5 years of tamoxifen .
Neoadjuvant trials permit temporal sampling of breast tissue. Substudies in the phase-III trial with letrozole  have examined the impact of such biomarkers as ER, progesterone receptor and epidermal growth factor receptor family members, HER-1 and HER-2, on patient response. The potent and selective third-generation aromatase inhibitors, anastrozole, letrozole and exemestane are approved for clinical use in postmenopausal patients with advanced hormone-dependent breast cancer or in patients failing anti-oestrogen therapies. Definitive recommendations for both prophylactic and adjuvant treatment with aromatase inhibitors await to be defined for routine use outside clinical trials.
The tissue specificity of SERMs is based on the fact that agonist/antagonist interactions occur with the ERs and that different cohorts of co-activators and co-repressors are present in different tissue sites of action. A variety of new SERMS is currently being investigated; however, no relevant clinical results are at hand as yet.
The tissue specificity of selective aromatase modulators is based on the fact that one role of oestradiol is as a paracrine and intracrine factor in postmenopausal women and men. In addition, tissue-specific regulation of the aromatase gene is based on the use of tissue-specific promoters. These promoters employ different stimulatory and inhibitory factors in the various tissue-specific sites of expression. This concept has led to the definition of selective aromatase modulators (SAMs). It will be interesting to see whether SAMs may provide an improved relationship between the above-mentioned advantages and major problems seen with standard aromatase inhibition.
Despite the emergence of aromatase inhibitors as effective first-line treatment of advanced breast cancer (ABC), tamoxifen remains a widely prescribed therapy. It remains, however, clinically important that other endocrine therapies are effective in cases where the disease has become resistant to tamoxifen. A clinical benefit (complete or partial response or stable disease lasting 24 weeks and over) is what can be achieved in the chronic condition of ABC; otherwise, advanced disease is not curable with 5-year survival rates of less than 20%.
The validation of new drugs in clinical oncology requires huge cohorts of patients, which, nevertheless, form a group of tumours with heterogenous biological characteristics. In the adjuvant endocrine setting, every new drug or strategy provides a modest absolute benefit over the standard available treatment and the treated population taken as a whole. Tamoxifen is the standard endocrine treatment for hormone-receptor-positive breast cancer both in the adjuvant and metastatic settings. However, as tamoxifen is also weakly oestrogenic, it may not be optimally effective and increases the risk of endometrial cancer and stroke. In addition, patients may be refractory or may become resistant to tamoxifen treatment. Since aromatase inhibitors block the synthesis of oestrogen and have low intrinsic oestrogenic activity, they have the potential to be more effective than tamoxifen. Consequently, they have been introduced as an alternative to tamoxifen treatment. The challenge of the coming years will be to identify the women who are best treated with an aromatase inhibitor upfront rather than a tamoxifen-to-aromatase inhibitor sequence. Although only on the basis of retrospective analysis, the evidence to date suggests that the benefit of the third-generation aromatase inhibitors is greater in ER-positive, progesterone-negative and in HER-2-overexpressing tumours.
Breast oncologists look for strong predictive data from genomic and proteonic profiles to allow the best choice of endocrine treatments and their optimal sequence. The sequential use of non-cross-resistant endocrine therapies and the association of endocrine and biological therapy seem to be the way ahead, while much translational research needs to be carried out in parallel in order to understand which individual patients benefit from the incorporation of newer endocrine agents and more target-oriented approaches to their adjuvant therapy .
Neoadjuvant treatments with chemotherapy or endocrine agents is being used increasingly to downstage locally advanced and large operable breast cancers. Following these treatments, inoperable breast cancer often becomes fully resectable, and initially operable tumours requiring mastectomy may be successfully removed by breast-conserving surgery. Patient selection is important to optimise new adjuvant endocrine therapy: only patients with ER-rich breast cancer are candidates, and postmenopausal women are likely to benefit the most. There is some evidence to suggest that the nature of the tumour response is different for preoperative endocrine therapy compared with chemotherapy. This difference may result in a higher rate of complete tumour excision following breast-conserving surgery after neoadjuvant endocrine treatment . There appears to be a low rate of subsequent local recurrence in patients having breast-conserving therapy after neoadjuvant endocrine therapy.