Selective Estrogen Receptor Modulators and Phytoestrogens

Abstract

Scientific achievements in the last two decades have revolutionized the treatment and prevention of breast cancer. This is mainly because of targeted therapies and a better understanding of the relationship between estrogen, its receptor, and breast cancer. One of these discoveries is the use of synthetic selective estrogen modulators (SERMs) such as tamoxifen in the treatment strategy for estrogen receptor (ER) positive breast cancer. Hundreds of thousands of lives have been saved because of this advance. Not only is tamoxifen used in the treatment strategy for patients who have breast cancer, but also for prevention in high risk premenopausal women. Another synthetic SERM, raloxifene, which was initially used to prevent osteoporosis, is also as effective as tamoxifen for prevention in high risk postmenopausal women. In certain regions of the world, particularly in Asia, a low incidence of breast cancer has been observed. These women have diets that are high in soy and low in fat, much unlike the western diet. Interest in the protective effects of soy derivatives has led to the research of phytoestrogens, metabolites of soy that are described by some as natural SERMs. As a result, many clinical questions have been raised as to whether phytoestrogens, which are also found in other natural foods, can protect against breast cancer. This article reviews the development and role of the more common SERMs, tamoxifen and raloxifene. In addition, this paper will also highlight the emerging studies on phytoestrogens and their similarity to SERMs.

Introduction

Great strides have been made in the last 25 years in the fight against breast cancer. One of the more notable developments has been the search for ways to prevent cancer. The development of Selective Estrogen Receptor Modulators (SERMs) has been a significant step towards achieving that goal. Tamoxifen, an antiestrogen in the breast and the pioneering SERM, has been the gold standard, and often the only choice in many countries for the treatment of breast cancer [1]. It also became the first drug ever to be approved by the United States (US) Food and Drug Administration (FDA) for the chemoprevention of breast cancer in high risk women [1]. This chapter will review the development of tamoxifen the prototypical SERM and its use and development as a chemopreventive agent. In addition this article will also highlight the emerging information regarding phytoestrogens that are being regarded by some as natural SERMs.

Background

By the turn of the 20^th century it was known that oophorectomy in pre-menopausal women with metastatic breast cancer could cause regression of the disease [2] [3]. This showed a link between products produced by the ovaries and the growth of some breast cancers. The product was found to be estrogen [4]. In 1936, Professor Antoine Lascassagne hypothesized that breast cancer was caused by a special hereditary sensitivity to estrogen and suggested that the development of an estrogen antagonist could prevent disease [5]. Over twenty five years later in 1962 Jensen and Jacobsen [6] described the estrogen receptor (ER) as the mediator of estrogen action, setting the stage for the manipulation of this receptor for multiple purposes [7].

Investigation of possible contraceptive agents led to the reinvention of ICI 46474 a failed contraceptive agent to become tamoxifen, the first targeted anti-cancer agent. The study of tamoxifen in the laboratory led to the finding that it inhibited the growth of ER positive breast cancer cells in vitro [8]. In addition, animal studies showed that tamoxifen prevented rat mammary carcinogenesis [9] [10] but had a stimulatory effect on rat uterine weight [11]. The actions of nonsteroidal antiestrogens were clearly not wholly explainable as estrogen agonists or antagonists and a model to describe their unique actions led to the development of the SERM concept [12] [13] [14].

What are SERMs?

SERMs are synthetic non-steroidal agents that bind to the ER and produce a change in the biologic activity of the receptor depending on the tissue type. The primary target site for SERMs, the ER, is a nuclear receptor. To fully understand the unique nature of SERMs the actions of estrogen on the body must be revisited. Estrogen in premenopausal women is primarily produced by the ovaries. There are multiple target sites for estrogen and it has various actions throughout the body. Estrogens decrease cholesterol levels by lowering the circulating low-density lipoproteins (LDL). Its actions also include maintenance of bone density in postmenopausal women, and hormonal regulation, and control of the menstrual cycle in premenopausal women. These actions are summarized in figure 1. In contrast the effect of SERMs depends on the target sites and is shown in figure 2.

Figure 1.

The sites of action for estrogen.

Figure 2.

The sites of action of tamoxifen.

A pure estrogen agonist would be one that stimulates the positive action of estrogen at all its targets. Conversely, a pure antagonist would inhibit all the actions of estrogen at all of its target sites. In contrast, SERMs have partial agonist and antagonist properties depending on the target site hence their uniqueness. Studies have shown that the partial agonist/antagonist properties depend on which associated coregulators are expressed when the receptor ligand interaction occurs [15]. The details of the receptor/ligand interaction help us understand the mechanism of action of SERMs.

Mechanism of Action

There are two aspects to the mechanism of action of SERMs: the pharmacokinectics or how the drug gets to the target site and the pharmacodynamics or what it does when it gets there. Tamoxifen, (figure 3), is a lipophilic prodrug that is easily absorbed by the gut without modification and 98% is bound to albumin after entering the circulation. It undergoes extensive metabolism in the gastrointestinal (GI) tract and in the liver into its less active form N-desmethyltamoxifen and two most active forms, 4-hydroxy tamoxifen and endoxifen [16] [17] [18] [19]. Each of the hydroxylated metabolites result from first pass metabolism in the liver. These compounds enter the bloodstream via the entero-hepatic circulation to reach their target sites [18] [20] [21]. The metabolites of tamoxifen are excreted via the fecal route as has been shown by animal studies using ¹⁴C radiolabeled tamoxifen [22]. These studies demonstrate 67 % of these metabolites enter the enterohepatic circulation and undergo further metabolism several times until excretion by the GI tract [23] [24]. 4-Hydroxy tamoxifen, and endoxifen have the same affinity for the ER as estrogen. Other metabolites of tamoxifen do not have as much effect or affinity for the ER as they lack the 4-hydroxy group [18]. Recent studies demonstrate that the potent tamoxifen metabolite endoxifen is produced by the product of the CY2D6 gene. In patients with mutations of the CYP2D6 gene or patients who take other medications that compete for the enzyme product, metabolism of tamoxifen to the potent metabolite endoxifen is affected and may therefore have less benefit [25] [26]. Raloxifene (figure 3), another SERM, is a polyphenol, which undergoes rapid conjugation in the GI tract and in the liver. In addition it also undergoes phase 3 metabolism by gut flora. The bacteria directly glucuronidate and sulfate this compound so that it is excreted [27] [28]. Since the drug does not reenter the enterohepatic circulation, it does not reach its targets as efficiently as tamoxifen. Also, a smaller percent enters the circulation as only 2% is bound to albumin and the half-life of raloxifene is 27 hours [29]. As a result of differences in metabolism and bioavailability, raloxifene is not as useful an agent in patients who already have breast cancer [30].

Figure 3.

A structural comparison of commonly studied phytoestrogens to SERMs

There are two isoforms of the ER, ERα and ERβ [6] [31] whose distribution and density varies depending on the target site. Both isoforms are found in the reproductive organs. Tamoxifen binds both receptors with equivalent affinity [32]. Endoxifen and tamoxifen have similar affinities for both isoforms [33]. Other ligands show preference for one isoform or the other, which may explain specific target tissue responses with various compounds. In many tissues, ERβ has anti-proliferative effects, whereas, ERα has proliferative effects [34]. Studies indicate that ER-β has an inhibitory effect on ER-α [35] [36] [37]. However, the biology is more complex than a simple agonist/antagonist interaction between the two receptors. The ratio of ERα to ERβ at a target site may be important in determining the overall action of a SERM on that tissue. A high ratio may correlate with high levels of cellular proliferation while a low ratio implies the opposite [38].

In the past, the interaction between SERMs and the ER was thought to be a simple case of a ligand switching its target receptor on or off. Through further research it is now known that this interaction is a more complex and dynamic process. The discovery of the steroid receptor co-activator protein (SRC1) helped further to elucidate this complex interaction [39]. The binding of a SERM to the ER results in a conformational change in the ER [40], which results in the exposure of different amino acids on the receptor. Studies using phage display created a fingerprint of exposed surfaces when tamoxifen or estrogen was bound to the ER. Different conformational changes occur in the ER depending on the ligand that binds to the ER. In addition, the fingerprint was different in ER-α vs. ER-β when they were bound to identical ligands [41].

This conformational change also promotes the binding of coregulators, which help determine the biologic activity of the SERM at the target site. Since the discovery of SRC1, dozens of other co-activator and co-repressor molecules have been discovered; all of which play some role in receptor modulation [15].

Finally, another dimension of signaling pathways can modulate the ER. Activation of the ER by other growth factor pathways can result in resistance to SERMs in a tumor.

This recruitment of specific co-regulators to the ligand receptor complex depends on the ligand that binds to the ER, the ER isoform, and “cross talk” with other growth factor pathways. SRC-3 is known to be important as a co-activator in breast cancer. In tumors and cancer cell lines that are HER2 positive and resistant to endocrine therapy with tamoxifen, studies demonstrate that SRC-3 is recruited to ER-α, but not ER-β in the presence of tamoxifen. In specimens from patients who were HER-2 negative and sensitive to endocrine therapy with tamoxifen, estrogen recruited SRC-3 to both ER isoforms, but tamoxifen did not. Finally, when SRC-3 was knocked down, there was reduced expression of the estrogen target gene, pS2 in MCF7 cells. After the SRC-3 knockdown in cells derived from HER2 positive tumors, there was a decrease in cell proliferation and the cells regressed in the presence of tamoxifen.

To summarize the molecular process thus far. Once a SERM binds to the ER it causes a change in the shape of the ER. This change of shape allows recruitment of co-activators, if it is destined to elicit an estrogenic response, or co-repressors if its response is anti-estrogenic. The binding of the coregulatory molecules leads to the activation of the promoter sequence of the estrogenic responsive gene [38]. This process is also controlled by the degradation and disassembly of complexes at the gene promoter site, which causes renewed activation of the signal to initiate RNA synthesis. In this way the SERM can specifically modulate the estrogen responsiveness of a target tissue.

Clinical Relevance

The full details of the mechanism of action of SERMs have yet to be precisely described however, their clinical importance as an advance in medicine is proven. Tamoxifen was initially tested in humans in the early 1970’s, before extensive anti-tumor testing in animals [42] [43]. Animal testing [1] [9] [10] refocused efforts and targeted the ER [44], thereby opening the door for chemoprevention. Through animal studies tamoxifen was found to have targeted anti-tumor activity and initially, anti-estrogenic activity correlated with anti-tumor activity. These findings led to extensive human trials that helped consolidate the actions of SERMs and refined their applications. In initial human studies tamoxifen, an “antiestrogen”, was found to lower bone density in pre-menopausal women [45]. However, the “estrogen-like” actions of tamoxifen, maintained bone density in post-menopausal women [46] [47]. In the uterus tamoxifen acts as an agonist and increases the risk of endometrial cancer in post-menopausal women [48]. The next sections review the large scale human chemoprevention trials of SERMs.

Chemoprevention

The first large human trial involving tamoxifen was the Royal Marsden study performed by Powles and colleagues [49] [50]. For this study approximately 3000 high-risk women were recruited and randomized to receive treatment with tamoxifen 20mg/day for 8 years or placebo. High risk status was determined by family history and a history of benign breast disease. The study found a decrease in LDL and loss of bone density in premenopausal women but, increased bone density in postmenopausal women and increased endometrial thickening on ultrasound study. Though this study initially showed no difference in the incidence of breast cancer, it was not powered to detect a difference in the development of breast cancer with either treatment group. Nevertheless, the twenty-year follow-up of this study does show a statistically significant reduction in the incidence of ER positive breast cancer in the tamoxifen treatment arm after the 8 years of treatment [51].

The National Surgical Adjunctive Breast and Bowel Project (NSABP) P-1 trial by Bernard Fisher and colleagues was the first major chemoprevention trial in the Unites Stated with tamoxifen [52]. Over 13,000 women were recruited for this study in multiple centers around the US. Once again high risk status was determined by family history, breast biopsy with pathologic findings of lobular carcinoma in situ or atypical ductal hyperplasia, no children, menarche by 12 and age at birth of first child of over 25. The initial results of the NSABP trial showed a 49% reduction in the risk of invasive breast cancer and a 50% reduction in the risk of non-invasive breast cancer. Tamoxifen also reduced the incidence of osteoporotic fractures. No difference was seen in the risk of myocardial infarction but there was an increased risk of deep venous thrombosis, endometrial cancer and cataracts in the tamoxifen group.

Based on these clinical trials in 1998, tamoxifen was approved by the US FDA for reduction of the risk of breast cancer in high-risk women. Despite the positive results of the NSABP P-1 trial the side effects noted in the tamoxifen group resurrected the interest in other SERMs that had similar chemopreventive profiles to tamoxifen but with a more desirable side effect profile. This has led to human trials with raloxifene an old compound, which had not been studied much since its discovery in the late 1970’s [53] [54].

Prevention of Osteoporosis

In laboratory studies raloxifene was shown to inhibit DMBA induced rat mammary carcinoma growth [55], and development [56] however, it was not as potent as tamoxifen. More importantly, raloxifene was as effective as tamoxifen in maintaining ovariectomized rat bone density but was less estrogen like than tamoxifen in the rodent uterus [13] or in stimulating mouse endometrial tumor growth [57]. The short half-life of raloxifene makes it a difficult drug to dose nonetheless; clinical trials with raloxifene have also helped define its pharmacology. The Multiple Outcomes for Raloxifene Evaluation (MORE) trial evaluated the effects of raloxifene in postmenopausal women [58] [59]. This study was extended to eight years as the Continuing Outcomes Relative to Evista (CORE) trial [52]. The results of the MORE/CORE trials demonstrated the effectiveness of raloxifene in preventing osteoporosis. In addition, raloxifene also inhibited the development of invasive breast cancer by 65% [60]. These clinical data justified the evaluation of raloxifene against tamoxifen to reduce the risk of breast cancer in high risk postmenopausal women. The Study of Tamoxifen and Raloxifene (STAR) trial, was a phase III double-blinded study that randomized eligible postmenopausal women at a high risk for breast cancer, to receive tamoxifen 20mg daily or raloxifene 60mg daily [61]. The STAR trial demonstrated the equivalence of raloxifene and tamoxifen in reducing the incidence of invasive breast cancer. Furthermore, raloxifene had a better side effect profile with a lower incidence of endometrial cancer and hyperplasia, deep venous thromboses and cataracts. A drawback of raloxifene however was its decreased effectiveness in preventing the development of non-invasive breast cancer after two years, when compared to tamoxifen. Currently raloxifene is FDA approved for the treatment and prevention of osteoporosis, and risk reduction for breast cancer in high risk postmenopausal women.

Extending Chemoprevention

With all the interest in therapeutic options it was reasonable to expect that the development of a chemopreventive agent such as tamoxifen would lead to interest in whether naturally occurring compounds have similar effects. Epidemiologic observations have made this question of even more interest and have led to research in the field of phytoestrogens. While the etiology may be unclear it has been well documented that Asian women have a lower incidence of breast and colorectal than Caucasian women [62]. Asian diets in particular are high in soy foods, which are felt to be responsible for this difference. When Asian women emigrate to western countries their incidence of breast cancer approaches that of the indigenous population [63]. This phenomenon has been observed in Japanese and Caucasian women who emigrate to the United States. It has also been observed that the risk of breast cancer in Asian Americans decreases in relation to increasing intake of soy derivatives [64]. Additionally, Chinese women who adopt a more westernized diet also appear to increase their incidence of breast cancer. All these findings have generated an interest in soy foods and its impact on hormone levels in the body

What are Phytoestrogens?

Phytoestrogens are plant derivatives that bear a structural similarity to 17-beta estradiol and act in a similar manner. Structures of common phytoestrogens, SERMs and 17- beta estradiol are shown in Figure 3. The principal phytoestrogen groups are flavonoids, lignans, coumestans and stilbenes [65] [66] [67]. Phytoestrogens are present in common foods such as soybeans, grains, fruits and vegetables. An in-depth review of the various types of phytoestrogens is beyond the scope of this article however, common properties of most phytoestrogens include their metabolism by gut flora to additional derivatives with varying estrogenic activity. Many studies have focused on isoflavones, which are a subgroup of the flavonoids, they include but are not limited to genistein, daidzein and biochanin A. These isoflavones have varying estrogenic activity [68] and isoflavones have been proposed as natural SERMs. Studies show that isoflavones act as antioxidants in vitro and exert antiproliferative activities [69] [70]. Equol (Figure 3), another phytoestrogen of the isoflavone family, is produced from daidzein by the action of intestinal flora. This metabolic conversion however occurs in only 30% of the population [71].

Lignans, the most prevalent phytoestrogens in the diet are found in whole wheat, fruits and vegetables. They are subdivided into enterolactones and enterodiol [65] with very weak estrogenic properties [72]. While there are many studies on isoflavones, there are significantly fewer studies on coumestans and stilbenes. Coumestans are potent activators of the ER signaling pathway but are not as prevalent in the diet. Resveratrol is the most common stilbene and its use as a chemopreventive agent against breast cancer is actively being studied in rodent models [65]. Phytoestrogens have varying degrees of estrogenic activity. In the next section we will consider the mechanism of action of phytoestrogens reveals that their interaction with the estrogen receptor which in some ways similar to the SERM/ER interaction.

Mechanism of Action of Phytoestrogens

As with SERMS phytoestrogens can bind to either ERα or ERβ however, phytoestrogens appear to have a higher affinity for ERβ [73]. This affinity may be dose dependent but overall phytoestrogens have a significantly lower affinity for the ER than estradiol [74] [75]. In addition the estrogenic potency of phytoestrogens varies within the particular phytoestrogen group. For example, within the flavonoid family genistein has greater potency than biochanin A, which has greater potency than daidzein [68]. Kuiper and colleagues [32] demonstrated that the stimulation of transcriptional activity by both subtypes of the ER vary depending on the estrogenic potency of the phytoestrogen. Further reporter gene assays have shown that synthetic estrogens and phytoestrogens have varying affinity for the ER and for each ER isoforms [73][75]. In broad terms, the phytoestrogens appear to bind to ERβ in preference to ERα. Hydroxylated SERMs in general have a higher binding affinity for both ERα and ERβ compared to phytoestrogens.

SERMs are nonsteroidal estrogens that become antiestrogenic by virtue of their correctly positioned side chain. The antiestrogen side chain is not present in phytoestrogens and may therefore limit their classifications as SERMs. Studies have found that isoflavones have both agonistic and antagonistic effects, though they are strong ERβ agonists and weak ERα agonists [76]. The presence of a correctly positioned phenolic ring and also the distance between the two opposing phenolic oxygens in the isoflavones structure is similar to that of 17 beta-estradiol (Figure 3). This similarity allows the isoflavones to bind to the ER, effectively displacing 17 beta-estradiol. This action may help explain how phytoestrogens help protect against breast cancer, because ERβ inhibits mammary cell growth as well as the stimulatory effects of ERα [77]. However, it is not certain whether isoflavones displace the estradiol by binding to a primary site on the ER, causing competitive binding between the isoflavones and the estradiol, or whether the isoflavones bind to a secondary site on the ER [78]. In contrast, genistein has been found to bind to the active site of ERβ [79]. Although the isoflavones have agonistic and antagonistic estrogenic effects, the phytoestrogens also induce differentiation as well as inhibit angiogenesis, cell proliferation, tyrosine kinase, and topoisomerase II; all of which will help prevent tumor growth.

As with the SERMs, studies have shown that the recruitment of co-regulatory molecules may be important in determining the function of phytoestrogens. In particular, isoflavones appear to selectively trigger ERβ transcriptional pathways, especially transcriptional repression. This affinity for the ERβ results in the exposure of activation weak function-2 (AF-2) on the surface of ERβ, which has greater affinity for certain coregulators compared to ERα [77]. Phytoestrogens also have differential activity on several ER associated signaling pathways. For example, Akt, which is normally phosphorylated secondary to activation of ERα, is upregulated by genistein and daidzein in ER positive breast cancer cell lines, while resveratrol has an inhibitory effect of the phosphorylation of Akt [80]. In contrast, in ER negative cell lines, resveratrol and daidzein activate Akt and genistein inhibits activation of Akt [80]. The effect of phytoestrogens on cell cycle regulators and transcription factors is significant as many novel synthetic agents are currently under development to inhibit pathways and factors that are up-regulated by the ER

It is important to stress at the outset that despite the fact that there have been numerous and extensive studies on the mechanisms of breast cancer chemoprevention with phytoestrogens, there is no clear certainty whether phytoestrogens are chemopreventive or may contribute to adverse outcomes in breast cancer [81].

Cell and Animal Studies of Phytoestrogens

Phytoestrogens have been classified as natural SERMs, and a review of cell and animal studies of phytoestrogens reveals some similarities to SERMs such as tamoxifen. The approach to these studies may be classified into three broad categories. The first are studies that focus primarily on the role of phytoestrogens as a chemopreventive agent. The second are those studies that focus on phytoestrogens as a treatment agent. The third are those studies that focus on the clinical effects when phytoestrogens are used continuously from neonates to adults.

The first category focuses on the chemopreventive effects of phytoestrogens in animal models that are subsequently treated with a chemical carcinogen. Animal studies have shown that when rats are treated with phytoestrogens and then exposed to a carcinogen they are less likely to develop breast cancer if exposure to phytoestrogens occurs at an early age [82] [83]. Lamartiniere and colleagues [82] demonstrated that the timing of exposure to phytoestrogens whether pre or post puberty, may influence their action on preventing mammary carcinogenesis. Lamartiniere [82] found that neonatal injections of genistein reduced the incidence of DMBA induced mammary tumors in rats. Further evaluation revealed that the overall effect of genistein on prepubertal rats appeared to be secondary to early differentiation in mammary tissues resulting in less active EGF signaling pathways in adulthood that may be protective against breast cancer. A recent meta-analyses by Warri et al [84] revealed pubertal exposure to phytoestrogens result in changes in the mammary gland morphology and signal pathways that mimic those induced by the estrogenic environment of early first pregnancy.

The second group of studies focus on the use of phytoestrogens in the realm of treatment in both tumor implanted athymic mice and breast cancer cell lines. Studies have shown that treating estrogen sensitive MCF-7 cell lines with genistein has an inhibitory effect on their growth [85]. However, not all studies have had such conclusive findings; in addition the action of phytoestrogens on breast cancer cells may be dose dependent. At low concentrations they may stimulate growth, and at high concentrations inhibit growth [72] [85] [86] [87] [88]. Indeed, the early study by Welshons et al [72] cautioned against the use of antihormonal therapies that did not block the ER for the treatment of breast cancer because high fiber or exclusively vegetarian diets with phytoestrogens containing food supplements could enhance the probability of tumor recurrence and growth. Nevertheless, certain phytoestrogens have also been noted to cause apoptosis of human breast cancer cells [89]. While phytoestrogens have been observed to cause these various actions in vitro, it is unclear that in vivo the concentrations needed to achieve these actions are attainable. In animal studies phytoestrogens have demonstrated a protective effect on the development of mammary cancer [90] [91]. Santell and colleagues [91] have shown that while genistein may inhibit breast cancer cells in vitro, treatment of tumor bearing athymic mice with genistein did not inhibit tumor growth. Furthermore the combination of phytoestrogens and tamoxifen to treat breast cancer may result in decreased efficacy of tamoxifen. In a study evaluating the development of tumor and the tumor latency period, tamoxifen treated mice fed a low dose isoflavone enriched diet had a higher tumor incidence and a shorter tumor latency period than placebo treated mice [92]. In addition tamoxifen associated mammary tumor prevention was also significantly reduced.

A third approach is the study of the effects from early exposure to phytoestrogens from the perinatal periods and onwards. This approach was recently adopted by Mardon and colleagues [93]. Rats perinatally or lifelong exposed to a rich isoflavone diet exhibited higher body weight and fat mass at 24 months of age. Perinatal exposure to phytoestrogens led to higher bone mineral density in later life [93]. The translation of these data to human epidemiology and pharmacology is the challenge.

Human Trials

Human trials on phytoestrogens differ from SERMs in that unlike the SERMs there are no major large-scale human trials. Human studies on phytoestrogens can be divided into two broad categories. The first are studies that evaluate the effect of phytoestrogens on estrogen biosynthesis and excretion, the second are those studies that evaluate the overall impact of dietary phytoestrogens on specific clinical endpoints such as menopausal symptoms and bone mineral density. Many studies have been done to examine the use of phytoestrogens as chemopreventive agents however, these studies are limited in that they are mainly retrospective. The possible mechanisms which may influence the relationship between phytoestrogens and breast cancer include the effect of phytoestrogens on estrogen biosynthesis and excretion, the interaction between phytoestrogens and the estrogen receptor, and the effect of phytoestrogens on cellular growth and proliferation in animal and cell studies [81].

Estrogen Biosynthesis and Excretion

Human studies on the effect of phytoestrogens on estrogen biosynthesis and excretion usually evaluate levels of circulating estrogen or steroid byproducts and metabolites in the urine. In addition in many of these studies the levels of phytoestrogens are also measured and factors that affect these levels are explored. Human studies have shown conflicting results regarding the overall effect of phytoestrogens. Lu and colleagues [94] treated 10 pre-menopausal women with a soy containing diet beginning on day two of the menstrual cycle to day two of the next cycle. Blood and urine samples were obtained before and during the initiation of the soy diet. Their results showed that the circulating levels of 17-beta estradiol decreased by 25%, however, cycle length did not change [94]. A dietary intervention study by Kumar and co-workers showed similar findings [95]. This study randomized women to receive 40mg of isoflavones day or placebo for a 12-week period. They found that serum free estradiol and estrone levels decreased. Serum hormone binding globulin increased and mean cycle length also increased. Conversely, a yearlong dietary intervention study by Maskarinec and co-workers [96] in premenopausal women did not find any difference in cycle length or hormone levels. These studies raise the question that while dietary intake of phytoestrogens is important, intake alone may not be the determinant of chemoprotective effect.

In addition to dietary intake, several other factors affect the levels of phytoestrogen derivatives in the body and their excretion in the urine, which may be of importance. As mentioned previously phytoestrogens are metabolized by gut flora to more active components. As a result substances that affect gut flora potentially affect the actions of phytoestrogens. Administration of antibiotics has been noted to decrease the serum concentration of enterolactone for a prolonged period [97]. Premenopausal women who are treated with long term antibiotics for urinary tract infections seem to be at higher risk for breast cancer, presumably because it alters the gut metabolism of phytoestrogens [98]. Smoking and obesity have been noted to decrease plasma enterolactone levels, however, tea, coffee, fiber and vegetables have the opposite effect [99]. In a study monitoring plasma enterolactone levels, women were noted to have a higher plasma concentration while on wheat bread 41.1nmol/L compared to 15.4nmol/L while on white bread [71]. A Finnish case control study [100] suggests that high enterolactone concentrations are associated with decreased breast cancer risk. Additionally, enterolactone levels are affected by both dietary intake and gut flora metabolism of ingested phytoestrogens.

In human studies, it is difficult to measure serum levels of phytoestrogens, because of a short half-life. Because most phytoestrogens are excreted in the urine, urine analysis of metabolites of phytoestrogens can be measured to give an indication of the phytoestrogens that predominate in the diet and even estimate the main dietary sources of the phytoestrogens [101]. Some metabolites of phytoestrogens such as enterolactone and equol are detectable in the urine. The excretion of equol in the urine has been proposed as a possible marker of the chemoprotective effect of phytoestrogens [102] [103]. Duncan and colleagues [104] studied the hormone profile of equol excretors versus equol non-excretors and found that regardless of the amount of phytoestrogens ingested in the diet, equol excretors had decreased levels of estrone, estrone-sulfate, testosterone, DHEA and higher levels of steroid hormone binding globulin. This steroid hormone profile has been found to be a protective profile for breast cancer. Of interest in this regard is the observation that the urinary excretion of phytoestrogens varies in different regions of the world [105]. Women in areas with a low incidence of breast cancer have higher urinary isoflavonoids than women living in areas with a high incidence of breast cancer. Vegetarians also have a higher concentration of isoflavonoids in their urine than omnivores [106]. However, it is unclear whether urinary enterolactone is protective against breast cancer or simply an indicator of a preferable (healthy) steroid profile. There are clearly other factors associated with whether or not phytoestrogens are protective than just the amount taken in the diet. The effect by which this is mediated remains unclear however, it may have to do with the levels of circulating sex steroid binding globulins. Phytoestrogens have been noted to stimulate the production of sex steroid binding globulin by liver cells [106]. In addition phytoestrogens have also been noted to be inhibitors enzymes involved in the synthesis of estrogen. This in turn leads to decreased circulating levels of free estrogen and less peripheral conversion of androgens to estrogens which are thought to be important in the development of cancer. Some phytoestrogens are known to decrease the conversion of androgens to estrogen by blocking the aromatase enzyme system. [107].

Phytoestrogens and Clinical Endpoints

The second group of human studies are those that focus on the effect of phytoestrogens on focal clinical endpoints. These endpoints vary and include alleviation of menopausal symptoms, maintenance of bone mineral density and development of breast cancer in some retrospective studies. Given recent concern regarding the possible adverse effects of hormone replacement therapy other alternatives for treatment of menopausal symptoms have been explored and phytoestrogens have played a significant role. A recent Cochrane review of the database revealed no clear evidence of the effectiveness of phytoestrogens in alleviating menopausal symptoms [108]. This not with standing there are some small trials which show a benefit to using phytoestrogens for treating menopausal symptoms. In a double blind prospective study sixty women were randomized to receive 60mg of isoflavones daily for 3 months or placebo [109]. The menopausal symptoms before and after treatment were recorded. Women receiving the phytoestrogens treatment noted a 57% and 43% decrease in the incidence of hot flashes and night sweats respectively. Similar results were seen in a small trial using a 6 week treatment of flaxseed for the treatment of menopausal symptoms [110]. Currently, phytoestrogens are being used clinically in the management of post menopausal symptoms in Europe [109]. Recently, prenylated flavonoids derived from hop are being used to treat menopausal symptoms. One such compound is 8-prenylnaringenin fig 3, that has strong estrogenic activity [111]. MenoHop an agent containing the phytoestrogen 8-prenylnaringenin, is currently being used to treat menopausal complaints in Belgium [112].

The relationship between phytoestrogens and bone health remains unclear, with some studies showing a benefit associated with phytoestrogen treatment and others showing none [113]. Supplementation of diet with isoflavones has been shown to help maintain lumbar spine bone density [114] [115]. A randomized double blind control trial was performed to compare with hormone replacement therapy (HRT), the effect of the phytoestrogen genistein on bone metabolism and bone mineral density [116]. Patients were randomized to receive either HRT daily (1mg of 17beta-estradiol and 0.5mg norethisterone) or genistein 30mg daily or placebo daily for a period of 1 year. On completion of this protocol women receiving the HRT and genistein had significantly increased bone mineral density in the femur compared to those in the placebo group. Another double blind randomized controlled trial comparing the effect of an isoflavone extract and placebo on bone mineral density had similar findings. In this study by Atkinson and colleagues [117] women receiving the isoflavones extract had a decreased loss of lumbar spine bone mineral content and bone mineral density

Direct studies on the efficacy of phytoestrogens in preventing breast cancer are difficult given the length of time required to perform such a study. However, surrogate endpoints such as the effect of phytoestrogens on breast cell proliferation and mammographic density have been studied. Short-term dietary supplementation with phytoestrogens stimulates breast epithelial proliferation [118]. This finding has also been noted in premenopausal women treated with prolonged phytoestrogen intake [119]. These histological findings are supported by the observation of increased high risk parenchymal sonographic patterns in women who report low dietary soy protein intake [120]. Other studies such as that by Maskarinec and colleagues [121] show a similar finding in mammographic density in women treated with prolonged phytoestrogen supplementation.

As seen in animal studies, the age at which a woman is exposed to phytoestrogens and length of exposure to phytoestrogens may be important in determining whether a protective benefit is obtained. A prospective study was performed by Key and colleagues [122] of over 30,000 women in Japan. Subjects completed a dietary questionnaire twice in a twelve-year span and were followed prospectively for the development of breast cancer. There was no relationship found between soy food consumption and the development of breast cancer, however this study was comprised of mostly non-adolescent women. Shu and colleagues [123] performed a retrospective case controlled study on Chinese women with breast cancer. Subjects completed a questionnaire regarding their dietary intake in adolescence. A high soy consumption as an adolescent was associated with a decreased incidence of breast cancer as an adult. This may also explain why when women emigrate to countries with a higher incidence of breast cancer than their native country, they are more likely to have a decreased incidence of breast cancer if they emigrated after puberty [124].

Conclusion

Since their discovery the use of SERMs in clinical practice continues to expand [125]. As our knowledge of phytoestrogens grows, so does our understanding of its interaction with the estrogen receptor and its ability to possibly act as a natural SERM. However, the molecular endocrinology of SERMs and phytoestrogens is very different so only a weak association exists between phytoestrogens and SERM action. Despite the advance-s in the treatment of breast cancer, prevention if possible is always superior to treatment. Currently tamoxifen and raloxifene are the first important steps in the quest to develop a complete preventative agent. In the future with more basic research perhaps phytoestrogens or their derivatives can play, a supportive role in chemoprevention as a substitute or as an agent to augment new and improved SERMs [126] [127].