OFomarkers in Cancer Cfriemoprevntion Muer, A.B., Bartsch, H., Boffetta, P., Drssted, L. and Vaitlo, H. eds ARC Scentific PJbljcations No. 154 International Agency ter Research on Cancer, Lyon, 2001

There is overwhelming evidence that alterations in endogenous hormone metabolism - as a form of endogenous exposure—may bean important metabolic riskfactorfor the development of . This chapter reviews current theories and major epidemiological findings that link endogenous hormones (sex and their metabolites, but also insulin, and insulin-like growth factor-I (IGF-l)) to breast cancer risk. Knowledge about these metabolic risk factors can be used to identify women at increased risk of breast cancer, who might benefit most from chemoprevention. In addition, modification of high-risk endocrine profiles may itself become a central target of chemoprevention. Possible intervention strategies include improvement of insulin sensitivity, reducing concentrations of IGF-1 in blood and breast tissue, reducing ovarian overproduction of androgens, inhibiting the activity of aromatase and other enzymes involved in formation within the breast, and modifying estrogen metabolism within the breast (e.g., decreasing 16(x- and 4a-hydroxylation, and increasing O-methylation of catecholestrogens). Several of these possible strategies are illustrated with examples of chemopreventive agents currently in use or proposed for use to prevent breast cancer,

Introduction hormones, the use of exogenous hormones for Breast cancer is predominantly a disease of west- contraception or for treatment of menopausal ern, industrialized countries. There is over- symptoms has also been suggested to be related to whelming evidence from studies on migrants and breast cancer risk. from time trends in incidence rates that the high The identification of specific alterations in incidence rates of breast cancer in western coun- endogenous hormone metabolism as risk factors tries are due mainly to lifestyle factors, including for breast cancer risk is a topic of extensive nutrition. Nutritional factors most strongly research. Knowledge of such hormonal risk factors implicated include low levels of physical activity may help understand why more traditional risk and an energy-dense diet rich in fats, refined car- factors (e.g., obesity, lack of physical activity, late bohydrates and animal protein. These factors also , use of certain exogenous hormones) appear related to an earlier average age at menar- are related to breast cancer. In addition, this che, increasing average body height and an knowledge can be used to design and test strategies increasing prevalence of obesity—three well estab- for breast cancer prevention, through changes in lished risk factors for breast cancer that are also lifestyle (nutrition, hormone use) or by use of spe- known to be associated with changes in hormone cific chemopreventive agents. Favourable changes metabolism. An additional risk factor, which prob- in endogenous hormone profiles can then be taken ably also affects breast cancer risk through changes as intermediate evidence for a possible preventive in endogenous hormone metabolism, is a child- effect. bearing pattern characterized by relatively late first The first section of this chapter reviews the asso- pregnancy and low parity. Besides endogenous ciations of breast cancer risk with circulating levels

149 Biomarkers in Cancer Chemoprevenlion of sex steroids and with modifications in sex (Ehrmann et al., 1995), but also in postmenopausal metabolism locally within the breast. In the women with upper body type (android) obesity next section, basic aspects of insulin-like growth (Kaaks, 1996). A4- and testos- factor-I (IGF-1) metabolism are discussed, and the terone are the immediate precursors for estrogen associations of breast cancer risk with circulating synthesis. Especially after the menopause, when levels of IGiLI, its binding proteins (IGFBPs) and the ovaries no longer produce the enzyme insulin are reviewed. In the third section, a num- aromatase needed for estrogen synthesis, the levels ber of possible strategies for breast cancer preven- of bioavailable androgens unbound to sex-hor- tion through modification of endogenous hor- mone binding globulin in plasma are a major mone metabolism are presented. The final section determinant of formed within adipose presents concluding remarks about the possible tissue, including the breast. use of markers of endogenous hormonal exposures The ovarian androgen excess hypothesis and in prevention studies. estrogen excess hypothesis are both supported by results from several prospective cohort studies, Sex steroids which have shown increased breast cancer risk in Circulating sex steroids as determinants of breast postmenopausal women having elevated plasma cancer risk levels of and A4-androstenedione, Ovarian sex steroids clearly play an important role reduced levels of sex-hormone binding globulin in promoting the development of breast cancer. and increased levels of total and bioavailable estra- Risk is consistently increased in women with early diol (Thomas et al., 1997a). The increases in breast menarche and late menopause (Key & Pike, 1988) cancer risk associated with this high-androgen and hence is associated with a longer lifetime pro- steroid profile may be due particularly to the rise in duction of and progesterone by the bioavailable plasma estrogen levels, plus increased ovaries. Furthermore, breast cancer incidence rates estrogen formation within breast tissue itself. rise more steeply with age before menopause than Estradiol is a strong mitogen for breast epithelial after, when the ovarian synthesis of estrogens and cells, and is therefore thought to have a key role in progesterone practically ceases, and when the pro- promoting tumour development (Henderson & duction of androgens gradually reduces to about Feigelson, 2000). However, risk may also be influ- half the premenopausal levels. Ovariectomy at an enced directly by androgens, through androgen early age reduces breast cancer incidence and receptors within breast tissue (Adams, 1998; Birrell ovariectomy at the time of cancer diagnosis dimin- etal., 1998). ishes the risk of tumour recurrence (Secreto & There is considerable evidence that the combi- Zumoff, 1994). nation of estrogens and progestogens increases A popular theory is that breast cancer risk is breast cancer risk further than exposure to excess increased in women who have increased plasma estrogens alone (Bernstein & Ross, 1993; Key & levels of bioavailable estradiol unbound to sex- Pike, 1988). Major observations supporting this hormone binding globulin, and elevated estradiol "estrogen-plus-progestogen' hypothesis are that: (1) concentrations within breast tissue (estrogen breast epithelial cells have the highest mitotic excess hypothesis) (Bernstein & Ross, 1993; Key & activity in the luteal phase of the menstrual cycle, Pike, 1988). A second theOry is that breast cancer when progesterone production peaks; (2) pre- risk is increased in women with elevated levels of menopausal women who are obese or who have androgenic steroid hormones, in particular testos- severe forms of ovarian androgen excess (as in terone and A4-androstenedione (ovarian androgen polycystic ovary syndrome) on average experience excess hypothesis) (Bernstein & Ross, 1993; Kaaks, some reduction in breast cancer risk, and this may 1996; Secreto & Zumoff, 1994). These two theories be explained by frequent anovulatory menstrual are complementary as ovarian over-production of cycles and an impaired luteal-phase progesterone androgens and increased plasma levels of total or production; (3) in postmenopausal women, the bioavailable estrogens often occur together, most use of combined estrogen-plus-progestogen hor- strikingly in patients with hyperandrogenic syn- mone replacement therapy increases breast cancer dromes such as the polycystic ovary syndrome risk to a greater extent than replacement therapy

150 Endogenous hormone metabolism as an exposure marker in breast cancer chemoprevention studies with estrogens alone (Ross et al., 2000; Schairer et The principal enzyme involved in producing al., 2000). Nevertheless, several case—control estrogens from androgenic precursors is aromatase studies (Secreto & Zumoff, 1994), though not one (CYP19), which converts 4-androstenedione and (small) prospective cohort study (Thomas et al., testosterone into and estradiol, respec- 1997b), have shown increases in breast cancer risk tively. Aromatase activity is higher in adipose tis- in premenopausal women who have elevated sue of breast cancer patients than in tissue from androgen levels in plasma or urine. The latter sug- women with benign breast disease (Miller & gests that only mildly hyperandrogenic women Mullen, 1993). Furthermore, in breasts containing who maintain regular ovulatory cycles, but who a tumour, aromatase expression and activity are may have somewhat elevated bioavailable estro- higher in quadrants bearing tumours than in the gens, may be at increased risk, whereas the more other quadrants. A second enzyme in breast tissue severely hyperaridrogenic women may experience that may contribute to estrogen formation is 3- a relative protection because of chronic anovula- hydroxysteroid dehydrogenase, which converts tion and impaired ovarian progesterone produc- (DHEA) and DHEA sul- tion. Direct evaluations of associations of breast fate (DHEAS) into M-androstenedione, the princi- cancer risk with circulating levels of estradiol and pal androgenic precursor for aromatization into progesterone are complicated by the wide varia- estrogens (estrone). tions of these two hormones during the menstrual A second important regulatory mechanism for cycle, and so far have not led to any definitive modulation of estrogenic activity within breast tis- conclusions. sue is the interconversion of the less potent estro- There is extensive evidence that reductions in gen estrone into the more potent estradiol by 1 7J3- sex-hormone binding globulin synthesis and hydroxysteroid dehydrogenase type 1, and vice increases in ovarian sex steroid synthesis are both versa, by 1 73-hydroxysteroid dehydrogenase type related to chronic hyperinsullnaemia, which in 2. In normal breast tissue and breast cancer cells, turn is often a consequence of obesity, lack of the balance in activity between the two isozymes physical activity, and hence insulin resistance generally appears to favour the conversion of (Kaaks, 1996) (Figure 1). The effects of insulin on estrone into estradiol (Thu & Conney, 1998). hepatic sex-hormone binding globulin production A third mechanism by which hormonally active and steroidogenesis may be mediated at least par- estrogens (i.e., with receptor binding activity) can tially by an increase in IGF-I bioactivity (see below). be formed is the hydrolysis of sulfate or glucuronyl groups from, respectively, and Estrogen synthesis and metabolism within breast estrone glucuronate or estradiol glucuronate. The tissue sulfated or glucuronated compounds are formed in Only sex steroids unbound to sex-hormone bind- the liver and other tissues by steroid sulfotrans- ing globulin can diffuse from the circulation into ferases and glucuronosyl transferases, respectively, target tissues, where they may bind to estrogen to form more water-soluble compounds that can and androgen receptors and exert their biological be excreted in bile and urine. These compounds effects. Circulating levels of bioavailable sex have little or no -binding affin- steroids are thus a key determinant of local estro- ity, but their large amounts in the circulation form genic or androgenic activity within the breast. a reservoir from which active estrogens can be Another important determinant of estrogenic formed by hydrolysis (Zhu & Conney, 1998). activity within breast tissue, however, is the local A fourth mechanism by which active estrogens metabolism of sex steroids (Figure 2). Indeed, stud- are formed is hydrolysis of estrogen fatty acid ies have shown substantially higher concentra- esters. These esters are formed by estrogen acyl- tions of estrogens in breast tissue than in the cir- transferase and, because of their high lipophulicity, culation, especially in postmenopausal women. accumulate in relatively high concentrations in Furthermore, several steroid-metabolizing enzymes fatty tissues. While estradiol fatty acid esters have are present, and active, in breast tissue (Zhu & little or no estrogen receptor-binding affinity, they Conney, 1998). Similar enzymatic activity is found form an important local reservoir for estradiol for- in adipose tissue elsewhere in the body. mation by the action of specific esterases.

151 Biomarkers in Cancer Chemoprevention

Insulin resistance I

Chronic hyperinsu linaemia

Ovarian lGFBP1 t- in plasma and 10 androgen synthesis 1' [FBP-2 tissues

Growth Total IGF-1 activity 1' Plasma hormone IGF-I 1 androgens I

Plasma sex- Plasma hormone binding bioavailable globulin L androgens

Ir

Plasma bioavailable Peripheral estrogens estrogen synthesis

Figure 1. Role of insulin and IGF-I in regulating levels of total and bioavailable sex steroids IGF-I, Insulin-like growth factor-I; IGFBP-1, insulin-like growth factor binding protein-i; IGFBP-2,

Besides enzymes involved in regulating levels of CYP2C family, CYP3A family, 450cf/F) that form active versus less active or bound forms of estrogen estrogen hydroxylation products. A second type of within breast tissue, several other enzymes form enzyme - catechol-O-methyltransferase —further estrogen metabolites that may have their own bio- metabolizes hydroxy-estrogen metabolites to logical activities and physiological roles (Figure 2). methoxy-estrogens. The first major classes are cytochrome P450 The hydroxylating P450 enzymes all require enzymes (CYP1AI, CYP1A2, CYPZB1, CYP2B2, NADPH, and are located in mitochondria. Estrogen

152 Sulfates Glucuronides Fatty acid 0-Methylated esters catechols sulfo- A glucuronyl estrogen \ransferase transferase / / acvl COMT sulfata se glucuronidase esterase//" tra;-isreras

Estrogens 2-OH I 4-01-1 (Catechol) estrogens Androgens + M-Androstenethone Estrone IT 1713-HSDI IT 173-HSD2 Other hydroxylated estrogens Testosterone a romatase Estradiol P450 oxvlases

Figure 2. Synthesis and metabolism of estrogens within breast tissue (After Zhu & Conney, 1999)

17P HSD1, 17-hydroxysteroid dehydrogenase type 1; 17P HSD2, 173-hydroxysteroid dehydrogenase type 2; COMT, catechol-O-methyl- -. transferase 01 (û Biomarkers in Cancer Chemoprevention metabolites with hydroxyl groups at positions 1, 2, a number of unique biological activities apparently 4, 6, 7, 14, 15, 16, 17 or 18 can be formed (Zhu & not mediated by the estrogen alpha receptor. In Conney, 1998). Some of these metabolites are particular 2-methoxyestradiol inhibits the prolifer- formed predominantly in liver, but others (e.g., the ation of several cancer cell lines in vitro, including 2-, 4- and 16-hydroxylation products) are also human breast cancer, and in vivo also has strong formed in relatively large amounts in breast tissue. anti-angiogenic effects (Zhu & Conney, 1998). 2-Hydroxyestradiol has markedly reduced estrogen The distribution of catechol-0-methyl trans- receptor-binding activity, but 4-hydroxyestradiol ferase activity appears to follow a bimodal pattern and 16-hydroxyestradiol retain potent hormonal due to a polymorphism in the COMT gene. About activity by binding estrogen receptors. Because of 25% of Caucasians are homozygous for the low- this and other lines of evidence, it has been activity allele. Some recent epidemiological studies hypothesized that an increase in 16-hydroxylation (Huang etal., 1999; Lavigne etal., 1997; Thompson (relative to 2-hydroxylation) may be a determinant et al., 1998) but not all (Millikan et ci!., 1998) have of breast cancer risk (Bradlow et aL, 1986). This shown an increase in breast cancer risk in women hypothesis was confirmed in some recent epi- with this gene variant, although the results varied demiological studies (Kabat etal., 1997), but not by in association with either pre- or postmenopausal others (Ursin et ai., 1999), where levels of 16- and risk, and between obese or non-obese women. 2-hydroxylated estrogens were measured in urine Taken together, there is some evidence that high (Meilahn et al., 1998). Further recent studies do catechol-0-methyl transferase activity may protect suggest, on the other hand, that increased forma- against breast cancer by increasing levels of 2- tion of 4-hydroxyestrogens in target tissues, methoxyestradiol, which has anti-estrogenic prop- including human breast, may play an important erties, and by faster deactivation of 4-hydroxy- role in estradiol-induced carcinogenesis. A key estradiol, which is carcinogenic. mechanism involved may be metabolic redox cycling of 4-hydroxyestradiol, catalysed by 1G F-1 and IGF-binding proteins cytochrome P450 enzymes (Liehr, 2000). The The hypothesis that risk of cancer (at various organ potential role of increased 4-hydroxylation in sites) may be related to circulating levels of total breast cancer development awaits confirmation IGF-1 and its binding proteins (IGFBPs), and to an from epidemiological studies. increase in IGF-I bioactivity at a tissue level, has The 2- and 4-hydroxyestrogens have a catechol recently received much attention from epidemiol- structure (i.e., with hydroxyl groups on two adja- ogists. Reasons are that IGF-I in general inhibits cent carbon atoms in the aromatic A ring), which apoptosis and stimulates cell proliferation (Jones means that these compounds are chemically reac- & Clemmons, 1995; Werner & Le Roith, 1997) and tive. Due to their reactivity, they may for instance thus stimulates tumour development. In addition, bind to DNA, and thus cause DNA damage or levels of IGF-I and several of its binding proteins mutations. A strong carcinogenic action has been (especially IGFBP-1, -2 and -3), as well as local IGF- observed only for 4-hydroxyestradiol, and not for I bioactivity within tissues, are intricately related the 2-hydroxy metabolite. The enzyme catechol- to nutritional status and energy metabolism O-methyltransferase (COMT) further metabolizes (Thissen et al., 1994). Furthermore, as mentioned these 'catechol estrogens' by formation of 2- and 4- above, the IGF-I/IGFBP system, in interaction with methoxyestradiol compounds. O-Methylated insulin, appears to be central in regulating the syn- estrogens are more lipophilic, have very long half- thesis and plasma levels of sex-hormone binding lives, and do not bind to the classical (alpha) estro- globulin and androgens. gen receptor. Earlier studies on the chemical reactivity and IGF-1, IGFBPs, and !GF-I bioactivity potential genotoxicity of catechol estrogens led to IGF-I bioactivity is the overall resultant of complex the suggestion that enzymatic O-methylation was interactions of endocrine, paracrine and autocrine primarily a detoxification pathway for these cate- sources of IGF-I and IGFBPs with cellular receptors. chol intermediates. More recent results, however, Although IGF-1 bioactivity is believed to increase suggest that the 0-methylated products may have generally when total IGF-I concentrations rise,

154 Endogenous hormone metabolism as an exposure marker in breast cancer chemoprevention studies

IGM bioactivity is modulated to a large extent by This relationship between risk and total plasma the IGFBPs. IGF-I was found exclusively for women who devel- A first level at which IGFBPs modulate IGF-I oped breast cancer before the average age at bioactivity is regulation of the efflux of circulating menopause. Besides increased levels of IGF-I, either IGF-1 through the capillary barrier. At least six dif- as total concentration or relative to IGFBP-3 levels, ferent IGFBPs have been identified in tissues and in two case—control studies have shown that elevated the circulation (Jones & Clemmons, 1995). plasma insulin levels also increase breast cancer Although IGF-I and IGFBPs are produced in practi- risk, in both premenopausal (Bruning et al., 1995; cally all human tissues, most IGF-I and IGFBPs in Del Giudice et al., 1998) and postmenopausal blood plasma are produced in the liver. Over 90% (Bruning et al., 1992) women. of circulating IGF-I is complexed with IGFBP-3 and with another glycoprotein called acid-labile sub- Possible strategies for chemoprevention unit (ALS), IGFBP-3 has a very high affinity for IGF- Different types of intervention aiming at reducing I, and the large IGF-I/IGFBP-3/ALS complex can- breast cancer risk by favourably changing endoge- not pass through the capillary barrier to target tis- nous hormone metabolism have been proposed, sues. There is recent evidence that IGFBP-5, which including energy restriction and body weight loss has even higher affinity for IGF-I than IGFBP-3, (in obese or overweight women), long-term may form a similar IGF-I/IGFBP-5/ALS complex. changes in diet and physical activity level, and use Practically all of the remaining IGF-1 is bound to of various chemopreventive agents. Strategies for IGFBP-1, -2, -4 and -6, which have lower binding prevention of breast cancer through alterations in affinities for IGM (compared with IGFBP-3 and hormone metabolism may target a number of IGFBP-5), do not form complexes with ALS, and mechanisms. These include reduction of circulat- are small enough to cross the endothelial barrier. ing IGF-1 levels and/or IGF-I bioactivity at a tissue Therefore, a decrease in plasma IGFBP-3 level, with level, reduction of ovarian (and/or adrenal) sex a transfer of IGF-I to smaller IGFBPs not complexed steroid production, increase of hepatic sex-hor- with ALS, is believed to increase IGF-I availability mone binding globulin production, reduction of to its tissue receptors. Reductions in plasma con- local estrogen synthesis and modification of local centrations of the smaller IGFBPs, and particularly estrogen metabolism. Each of these strategies in IGFBP-1 and -2, are also thought to increase the principle may lead to measurable alterations in bioavailability of circulating IGF-I (Jones & levels of hormones or hormone metabolites in Clemmons, 1995). blood or urine. Chemopreventive agents may also A second level at which IGFBPs modulate IGF-I be used to diminish the binding of hormones to bioactivity is the target tissue itself, where IGFBPs their receptors. regulate the binding of IGF-I to its receptors. Although binding proteins have been mostly pro- Reducing IGF-1 bioactivity posed to inhibit receptor binding by complexing A class of compounds that lower IGF-I levels are IGF-I, results from in-vitro studies suggest that, growth hormone-releasing hormone (GHRH) depending on the relative concentrations of IGM antagonists, including the natural compound and IGFBPs, certain IGFBPs (e.g., IGFBP-1, -2, -3 somatostatin, and a number of somatostatin ana- and -5) may actually enhance IGF-I binding to its logues (e.g., octreotide, vapreotide, lanreotide, receptors (Jones & Clemmons, 1995). octastatin, somatuline (BIM 23014)). These com- pounds decrease IGF-I levels by inhibiting the pitu- Epidemiological studies on insulin, IGF-1 and breast itary secretion of growth hormone (Kath & cancer risk Hoffken, 2000; Kineman, 2000). In addition to the Three case—control studies and one prospective suppression of the growth hormone/IGF-I axis, cohort study have shown an increase in breast can- GHRH antagonists appear to inhibit tumour cer in women with elevated plasma IGF-1 or with growth directly through a local mechanism, which elevated IGF-I for a given level of IGFBP-3 (Bohike may be related to reduced local synthesis of et al., 1998; Bruning et al., 1995; Del Giudice et al., insulin-like growth factor-Il (IGF-11). This direct 1998; Hankinson et al., 1998; Peyrat et al., 1993). effect might be mediated by blocking the

155 Biamarkers in Cancer Cliemoprevention paracrine/autocrine actions of locally produced plus progestogen replacement therapy was difficult GHRH, which may normally act through receptors to evaluate because of lack of detail in the infor- for the vasoactive intestinal protein or pituitary mation from published studies. Nevertheless, a adenylate cyclase-activating polypeptide - pep- large case-control study and a cohort study tides that are structurally similar to GHRH. recently showed significantly stronger increases in Somatostatin analogues have been tested as part of risk with combined estrogen plus progestogen hormonal adjuvant therapy in breast cancer therapy than with estrogens alone (Ross et al., patients, but have not been evaluated for chemo- 2000; Schairer et ai., 2000). prevention in healthy women. The use of estrogen analogues for treatment and A second class of agents that decrease especially possibly chemoprevention of breast cancer was ini- the hepatic production and circulating levels of tially motivated by the concept that such com- total IGM are exogenous estrogens (e.g., in oral pounds would diminish estrogenic effects in breast contraceptives, postmenopausal estrogen replace- tissue by competing with natural estradiol for ment therapy, or combined estrogen-progestogen estrogen receptor binding, without having a recep- replacement therapy) (Campagnoli et al., 1992, tor-activating capacity (Goldstein, 1999). A large 1995; Ho et al., 1996) and estrogen analogues meta-analysis of randomized intervention trials, (, , droloxifene) (Pollak et al., including more than 75 000 breast cancer patients, 1992). Especially weak estrogen analogues have confirmed the efficacy of such analogues, showing been regularly proposed as potential chemopre- dramatic improvements in ten-year recurrence-free ventive agents against breast cancer, and this was survival in the tamoxifen treatment group (Early initially motivated mainly by their anti-estrogenic Breast Cancer Trialists' Collaborative Group, 1992). effects, through competitive binding to estrogen Another large, double-blind and placebo-con- receptors (but without receptor activation). trolled randomized trial also showed a strong Estrogenic compounds decrease the liver synthesis reduction in breast cancer incidence among 13000 of IGF-1 only if taken orally but not when admin- initially cancer-free women (Fisher et al., 1998). istered transdermally. A likely explanation is that The latter finding was not confirmed, however, by oral administration leads to relatively high smaller trials in about 2500 women in the United estrogen concentrations in the portal vein and Kingdom (Powles et al., 1998) and in about 5400 hence in the liver, and that only such high con- women in Italy (Veronesi et al., 1998). A recent trial centrations can substantially decrease hepatic IGF- among 7700 postmenopausal women with osteo- I production. Oral administration of androgenic porosis showed a strong protective effect of ralox- compounds, including some progestagens ifene against breast cancer occurrence (Cummings produced from the androgenic precursor 19- etal., 1999). nortestosterone (e.g., , norgestrel, A third group of agents that have been tested for levonorgestrel) and present in some types of oral potential chemopreventive activity against breast contraceptive and hormone replacement therapy, cancer, and which also appear to decrease circulat- leads to an increase in synthesis and plasma levels ing IGF-I levels, are retinoic acid analogues, such as of IGF-I. So far, no clear effect of fenretinide (4-hydroxyphenylretinamide) and all- intake on circulating IGF-1 levels has been trans-retinoic acid. A randomized intervention trial reported. among about 3000 Italian women with surgically The epidemiological evidence for associations removed stage I breast cancer showed no signifi- of breast cancer risk with oral contraceptive use, cant effect of fenretinide on the occurrence of con- use of estrogen replacement therapy or hormone tralateral or ipsilateral tumors. However, an inter- replacement therapy has been reviewed (IARC, action was detected between fenretinide and 1999), as well as that for tamoxifen (IARC, 1996). menopausal status, with a possible beneficial effect There is a small increase in risk associated with use in premenopausal women (Veronesi et al., 1999). of combined oral contraceptives, as well as with IGF-1 bioactivity may be decreased by improv- use of estrogen replacement therapy. The associa- ing insulin sensitivity. An improvement of insulin tion of risk with oral contraceptives based on sensitivity and decreases in hepatic glucose output progestogens only, and with combined estrogen lead to lower endogenous insulin secretion, and

156 Endogenous hormone metabolism as an exposure marker in breast cancer chemoprevention studies this in turn leads to higher levels of IGFBP-1 and A second approach, which has also proven IGFBP-2. Natural ways of improving insulin sensi- effective especially in women with polycystic tivity are to lose body weight (for overweight or ovary syndrome, is the use of combination-type obese women) (Bosello et al., 1997; Guzick et al., oral contraceptives, which reduce ovarian hyper- 1994) or to increase physical activity. In addition, androgenism by diminishing the pituitary secre- insulin-sensitizing and hypoglycaemic drugs such tion of . A third, more experi- as metformin (Pugeat & Ducluzeau, 1999) and mental approach is to use new types of oral con- troglitazone (Henry 1997; Scheen & Lefebvre, traception, which contain gonadotropin-releasing 1999), biguanides or sulfonylureas (e.g., tolbu- hormone agonists to inhibit secretion of pituitary tamide and tolazamide) could be used. luteinizing hormone, as well as small amounts of exogenous estrogens to compensate for the block Increasing plasma sex-hormone binding globulin of endogenous (ovarian) sex steroid production levels (Spicer & Pike, 1994). Circulating sex-hormone binding globulin is pro- duced in the liver and, as mentioned above, is Reducing estrogen formation within the breast under negative control of insulin and IGF-I, which Since steroid sulfatases may increase the local for- both reduce its hepatic synthesis. Approaches to mation within the breast of estrogenic steroids increase sex-hormone binding globulin synthesis with receptor-activating capacity, steroid sulfatase include, first of all, improvement of insulin sensi- inhibitors could have considerable therapeutic, tivity. Weight loss or an increase in physical activ- and possibly preventive, potential. Several such ity both increase circulating sex-hormone binding inhibitors have now been developed. The most globulin levels (Guzick etal., 1994), as does the use potent to date is estrone-3-0-sulfamate (Purohit et of insulin-sensitizing drugs. Oral intake of exoge- al., 1999). Other steroidal sulfatase inhibitors are nous estrogens and estrogen analogues (e.g., estrone-3-methylthiophosphonate (Duncan et al., tamoxifen), which decrease the hepatic production 1993) and 3-0-methylphosphonate derivatives of of IGF-I, also causes a rise in plasma sex-hormone dehydroepiandrosterone (DHA-3-MTP), preg- binding globulin levels (Campagnoli et al., 1992). nenolone or cholesterol (Purohit et al., 1994). Non- Conversely, oral intake of androgenic compounds steroidal estrone sulfatase inhibitors include 4- reduces hepatic output and plasma levels of sex- methylcoumarin 7-0-sulfamate and its derivatives hormone binding globulin (Campagnoli et al., (14, 16 and 18) (Woo et al., 1998). The extent and 1994). Some study results have suggested a mild duration of the inhibition of estrone sulfatase may increase in hepatic production (Loukovaara et al., be monitored by measuring the activity of this 1995) or plasma levels of sex-hormone binding enzyme in white blood cells or by measuring the globulin (Adlercreutz et al., 198 7) at high intakes of decrease in the plasma DHEA/DHEAS concentra- , but this requires confirmation in tion ratio (Purohit et al., 1997). larger studies. Besides steroid sulfatase inhibitors, inhibitors of the enzyme aromatase (e.g., anastrozole, letrozole, Reducing ovarian sex steroid synthesis vorozole, aminoglutethimide, 4-hydroxyandro- Total plasma levels of testosterone are positively stenedione (Harvey, 1998; Singh et al., 1998) and related to breast cancer risk. As most of the circu- other compounds (Kelloff et al., 1998)) have been lating testosterone comes from the ovaries, this tested extensively in animal models and in clinical suggests that a reduction in ovarian sex steroid trials for breast cancer treatment. Aromatase output may also be an approach to prevent cancer. inhibitors are being proposed for phase III trials for A first approach to diminishing ovarian (and breast cancer treatment. The use of aromatase possibly also adrenal) androgen production is to inhibitors can lead to profound suppression of reduce circulating insulin levels. In obese and plasma estrogen levels (Kelloff et al., 1998). hyperandrogenic women, energy restriction with Nevertheless, strategies may be developed to body weight loss and use of insulin-sensitizing obtain chemopreventive effects without total sup- drugs have both been shown to decrease circulat- pression of aromatase and plasma estrogen levels ing androgen levels (Pugeat & Ducluzeau, 1999). (Kelloff et al., 1998).

157 Biomarkers in Cancer Chernoprevention

Several fiavonoids and isoflavonoids (phyto- This implies that either the hormonal parame- estrogens; e.g., , ) have been ter aimed at must be itself a direct cause of cancer found to be potent inhibitors of estrone reduction or else it must be at least very closely associated by 1 73-hydroxysteroid dehydrogenase type 1 in with other metabolic factors that are on the causal breast cancer cells (Makela et al., 1995, 1998). pathway (Lippman et al., 1990). Measurements of There is little information about compounds that hormones or hormone metabolites may then be may be used to regulate the balance between activ- used to evaluate the efficacy of a given prevention ities of the 170-hydroxysteroid dehydrogenase strategy. There are number of important problems, type 1 and type 2 isozymes in favour of lower estra- however. diol synthesis. First, the evaluation of whether a given hor- mone is directly related to tumour development is Modification of estrogen hydroxylation generally not straightforward. One problem is that On the basis of the theory that a high ratio of 16- hormonal parameters can have quite strong phys- to 2-hydroxyestradiol would increase risk, the use iological interrelationships. For example, in post- of chemicals that can shift hydroxylation towards menopausal women, elevated plasma insulin levels 2-hydroxylation have been proposed as chemo- lead to reductions in sex-hormone binding globu- preventive agents. One such compound is indole- lin, increases in plasma androgens and increases in 3-carbinol, a compound which occurs naturally in bioavailable estrogens. Sorting out which of these certain vegetables and which competes with estra- interrelated hormonal factors (bioavailable plasma diol for 16-hydroxylation (Michnovicz, 1998). The estrogens, androgens, insulin?) are more directly, shift from 16-hydroxylation to 2-hydroxylation or strongly, related to breast tumour development. can be monitored by measuring the ratio of the may not be possible using epidemiological meth- two types of estrogen hydroxylation product in ods alone, and may require complementary mech- urine (Michnovicz et al., 1997). It is possible, how- anistic evidence from experimental studies. When ever, that inhibition of 16-hydroxylation leads to possible effects on breast tissue of physiologically increased formation of 4-hydroxyestradiol, which interrelated hormones cannot be disentangled, a is potentially carcinogenic. reasonable approach to prevention may be to take modification of a global endocrine profile as the Hormone measurements as markers of endoge- target for prevention. For example, by improving nous exposure in chemoprevention studies: insulin sensitivity, one may at the same time conclusions and problem areas obtain an increase in plasma sex-hormone bind- As shown above, alterations in endogenous hor- ing globulin levels and reductions in plasma levels mone metabolism - as a form of 'endogenous' of total and bioavailable androgens and estrogens. exposure - may be an important metabolic risk Second, it is not always clear whether circulat- factor for the development of breast cancer. ing levels of hormones are a main determinant, or Knowledge about such metabolic risk factors can at least a good indicator, of levels within breast tis- be used to identify women at increased risk of sue. For example, in postmenopausal women, who breast cancer who might benefit most from have very low plasma concentrations of total and chemoprevention. In addition, as discussed in non-sex-hormone binding globulin-bound the third section of this chapter, modification of estradiol, the main determinant of estradiol high-risk endocrine profiles may become the concentrations within the breast may be its local central target for chemoprevention. In the latter synthesis from androgen precursors or from case, it is assumed that chemopreventive modifi- estrone. Therefore, if estradiol is a key hormone in cation of endogenous hormone levels (e.g., promoting breast tumorigenesis, increased circu- reduction in IGF-I levels in blood and other tis- lating bioavailable androgens, or increased breast sues; inhibition of aromatase activity; shifts of tissue activities of aromatase or 1713-hydroxylase, estradiol hydroxylation pathways towards less may be more important determinants of risk than 16- or 4-hydroxylation) will lead to reductions in plasma concentrations of bioavailable estradiol. breast cancer risk. Another example is the control of IGF-1 bioactivity

158 Endogenous hormone metabolism as an exposure marker in breast cancer chemoprevention studies

within breast tissue, which depends not only on Bradlow, HI., Hershcopf, R.E. & Fishman, J.F. (1986) circulating levels of IGF-I and IGFBPs, but also on Oestradiol 16 alpha-hydroxylase: a risk marker for breast cancer. Cancer Sun'., 5, 573-583 local synthesis. Thus, it is not entirely clear, for instance, whether factors such as estrogen replace- Bruning, P.F., Bonfrer, J.M., Van Noord, PA., Hart, A.A., ment therapy or tamoxifen that may decrease the Jong-Bakker, M. & Nooijen, W.J. (1992) Insulin resistance hepatic synthesis and output of IGF-I, therefore and breast-cancer risk. mt. J. Cancer, 52, 511-516 reducing its plasma concentrations, will also sub- Bruning, 9.1, Van Doom, J., Bonfrer, 3M., Van Noord, stantially reduce IGF-I bioactivity within the PA., Korse, C.M., Linders, T.C. & Hart, A.A. (1995) breast. A third example is formation of hydroxy- Insulin-like growth-factor-binding protein 3 is decreased and methoxy-metabolites of estrogens within the in early-stage operable pre-menopausal breast cancer. mt. J. Cancer, 62, 266-270 breast. Blood concentrations of these metabolites may not reflect very accurately their levels in Campagnoli, C., Biglia, N., Belforte, P., Botta, D., Pedrini, breast tissue, because they can also be formed in E. & Sismondi, P. (1992) Post-menopausal breast cancer adipose tissue elsewhere in the body. risk: oral estrogen treatment and abdominal obesity induce opposite changes in possibly important biological Besides the problem that circulating levels of a variables. Eur. J. Gynaecol. OncoL, 13, 139-154 given hormone may not always reflect breast tissue concentrations, there can be major technical or Campagnoli, C., Biglia, N., Lanza, M.G., Lesca, L., Pens, Sismondi, logistic problems in the measurement of hormones C. & P. (1994) Androgenic progestogens oppose the decrease of insulin-like growth factor I serum or their metabolites in blood, urine or saliva. Such level induced by conjugated oestrogens in post- problems are the lack of sensitivity and specificity menopausal women. Preliminary report. Maturitas, 19, of methods for the detection of specific sex steroids 25-31 and their metabolites (e.g., hydroxylation prod- Campagnoli, C., Biglia, N., Pens, C. & Sismondi, P. (1995) ucts) and problems due to cyclic or pulsatile varia- Potential impact on breast cancer risk of circulating tions in hormone secretion over time (Michaud et insulin-like growth factor 1 modifications induced by al., 1999; Xu etal., 1999). Oral HRT in menopause. Gynecol. EndocrinoL, 9, 67-74 Cummings, SR., Eckert, S., Krueger, K.A., Grady, D., References Powles, T.J., Cauley, J.A., Norton, L., Nickelsen, T., Adams, J.B. (1998) Adrenal androgens and human breast Bjarnason, N.H., Morrow, M., Lippman, ME, Black, D., cancer: a new appraisal. Breast Cancer Res. Treat., 51, Glusman, JE., Costa, A. & Jordan, V.C. (1999) The effect 183-188 of raloxifene on risk of breast cancer in postmenopausal Adlercreutz, H., Hockerstedt, K., Bannwart, C., Bloigu, S., women: results from the MORE randomized trial. 1-lamalainen, E. Fotsis, T. & 011us, A. (1987) Effect of Multiple Outcomes of Raloxifene Evaluation. J. Am. Med. dietary components, including and phytoestro- Ass., 281, 2189-2197 gens, on enterohepatic circulation and liver metabolism Del Giudice, M.E., Fantus, 1G., Ezzat, S., McKeown- of estrogens and on binding globulin Byssen, G., Page, D. & Goodwin, RJ. (1998) Insulin and (SHBG).J. Steroid Biochem., 27, 1135-1144 related factors in premenopausal breast cancer risk. Bernstein, L. & Ross, R.K. (1993) Endogenous hormones Breast CancerRes. Treat., 47, 111-120 and breast cancer risk. Epidemiol Rev., 15, 48-65 Duncan, L., Purohit, A., Howarth, N.M., Potter, B.V. & Birrell, SN., Hall, R.E. & Tilley, W.D. (1998) Role of the Reed, M.J. (1993) Inhibition of estrone sulfatase activity by androgen receptor in human breast cancer. J. Mammary estrone-3-methylthiophosphonate: a potential therapeu- Gland Biol. Neoplasia, 3, 95-103 tic agent in breast cancer. Cancer Res., 53, 298-303 Bohike, K., Cramer, D.W., Trichopoulos, D. & Mantzoros, Early Breast Cancer Trialists' Collaborative Group (1992) C.S. (1998) Insulin-like growth factor-I in relation to pre- Systemic treatment of early breast cancer by hormonal, menopausal ductal carcinoma in situ of the breast. cytotoxic, or immune therapy. 133 randomised trials Epidemiology, 9, 5 70-5 73 involving 31,000 recurrences and 24,000 deaths among Bosello, O., Armellini, F., Zamboni, M. & Fitchet, M. 75,000 women. Lancet, 339, 1-15 (1997) The benefits of modest weight loss in type II dia- Ehrmann, D.A., Barnes, R.B. & Rosenfield, R.L. (1995) betes, mt. j. Obes. ReJet. Metab. Disord., 21 Suppl. 1, Polycystic ovary syndrome as a form of functional ovar- S10-S13 ian hyperandrogenism due to dysregulation of androgen secretion. Endocrinol, Rev., 16, 322.-353

159 Biomarkers in Cancer Chemoprevention

Fisher, B., Costantino, J.P., Wickerham, D.L., Redmond, Kath, R. & Hoffken, K. (2000) The significance of somato- CX, Kavanah, M., Cronin, WM., Vogel, V., Robidoux, statin analogues in the antiproliferative treatment of car- A., Dimitrov, N., Atkins, J., Daly, M., Wieand, S., Tan- cinomas. Recent Results Cancer Res., 153, 23-43 Chiu, E., Ford, L. & Wolmark, N. (1998) Tamoxifen for Kelloff, G.J., Lubet, RA., Lieberman, R., Eisenhauer, K., prevention of breast cancer: report of the National Steele, YE., Crowell, J.A., Hawk, E.T., Boone, C.W. & Surgical Adjuvant Breast and Bowel Project P-1 Study. J. Sigman, C.C. (1998) Aromatase inhibitors as potential Nat! Cancer Inst., 90, 1371-1388 cancer chemopreventives. Cancer Epidemiol. Biomarkers Goldstein, S.R. (1999) Effect of SERMs on breast tissue, J. Prev., 7, 65-78 Endocrinol. Invest., 22, 636-640 Key, TJ. & Pike, M.C. (1988) The role of oestrogens and Guzick, D.S., Wing, R., Smith, D., Berga S.L. & Winters, progestagens in the epidemiology and prevention of S.J. (1994) Endocrine consequences of weight loss in breast cancer. Eur. J. Cancer Clin. Oncol., 24, 29-43 obese, hyperandrogenic, anovulatory women. Fertil, Kineman, R.D. (2000) Antitumorigenic actions of growth SterU., 61, 598-604 hormone-releasing hormone antagonists. Proc. Nat! Hankinson, S.F.., Willett, W.C., Colditz, G.A., Hunter, Acad. Sci. USA, 97, 532-534 DJ., Michaud, D.S., Deroo, B., Rosner, B., Speizer, FE. & Lavigne, J.A., Heizisouer, K.J., Huang, t-I.Y., Strickland, Pollak, M. (1998) Circulating concentrations of insulin- PT., Bell, D.A., Selmin, O., Watson, MA., Hoffman, S., like growth factor-I and risk of breast cancer. Lancet, 351, Comstock, G.W. & Yager, J.D. (1997) An association 1393-1396 between the allele coding for a low activity variant of Harvey, H.A. (1998) Emerging role of aromatase catechol-O-methyltransferase and the risk for breast can- inhibitors in the treatment of breast cancer. Oncology cer. Cancer Res., 57, 5493-5497 Huntingt), 12,32-35 Liehr, J.G. (2000) Is estradiol a genotoxic mutagenic car- Henderson, B.E. & Feigelson, H.S. (2000) Hormonal car- cinogen? Endocrinol. Rev., 21, 40-S4 cinogenesis. Ccucinogenesis, 21, 427433 Lippman, S.M., Lee, J.S., Lotan, R., Hittelman, W., Henry, R.R. (1997) Thiazolidinediones. Endocrinol. Metab. Wargovich, M.J. & Hong, W.K. (1990) Biomarkers as Clin. North Am., 26, 553-5 73 intermediate end points in chemoprevention trials. J. Nat! Cancer Inst., 82, 555-560 Ho, KY., O'Sullivan, AJ., Weissberger, A.J. & Kelly, J.J. (1996) Sex steroid regulation of growth hormone secre- Loukovaara, M., Carson, M., Palotie, A. & Adlercreutz, H. tion and action. Hormone Res., 45, 67-73 (1995) Regulation of sex hormone-binding globulin pro- duction by isoflavonoids and patterns of isoflavonoid Huang, CS., Chern, H.D., Chang, Kj., Cheng, C.W., conjugation in HepG2 cell cultures. Steroids, 60, 656-661 Hsu, S.M. & Shen, C.Y. (1999) Breast cancer risk associ- ated with genotype polymorphism of the estrogen- Makela, S., Poutanen, M., Kostian, ML., Lehtimaki, N., metabolizing genes CYP1 7, CYP1A1, and COMT: a multi. Strauss, L., Santti, R, & Vihko, R. (1998) Inhibition of genie study on cancer susceptibility. Cancer Res., 59, 17beta-hydroxysteroid oxidoreductase by in 4870-4875 breast and prostate cancer cells, Proc. Soc. Exp. Bio!. Med., 217,310-316 IARC (1996) IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol. 66, Some Makela, S., Poutanen, M., Lehtimaki, J., Kostian, ML., Pharmaceutical Drugs, Lyon Santti, R. & Vihko, R. (1995) Estrogen-specific 17 beta- hydroxysteroid oxidoreductase type 1 (E.C. 1.1.1.62) as a IARC (1999) IARC Monographs on the Evaluation of possible target for the action of phytoestrogens. Proc. Soc. Carcinogenic Risks to Humans. Vol. 72, Hormonal Contra- Exp. Bio!. Med., 208, 51-59 ception and Post-menopausal Hormonal Therapy. Lyon Meilahn, EN., De Stavola, B., Allen, D.S., Fentiman, I., Jones, J.I. & Clemmons, D.R. (1995) Insulin-like growth FIL., Sepkovic, D.W. & Kuller, L.H. (1998) Do factors and their binding proteins: biological actions. Bradlow, urinary oestrogen metabolites predict breast cancer? Rev., 16, 3-34 Endocrinol. Guernsey III cohort follow-up. Br. J. Cancer, 78, Kaaks, R. (1996) Nutrition, hormones, and breast cancer: is 1250-1255 insulin the missing link? Cancer Causes Control, 7, 605-625 Michaud, D.S,, Manson, J.E., Spiegelman, D., Barbieri, Kabat, G.C., Chang, Cj., Sparano, J.A., Sepkovie, D.W., R.L., Sepkovic, D.W., Bradlow, H.L. & Hankinson, S.E. Hu, X.P., Khalil, A., Rosenblatt, R. & Bradlow, H.L. (1997) (1999) Reproducibility of plasma and urinary sex Urinary estrogen metabolites and breast cancer: a case- hormone levels in premenopausal women over a one- control study. Cancer Epidemiol. Biomarkers Prev., 6, year period. Cancer Epidemiol. Biomarkers Prev., 8, 505-509 1059-1064

160 Endogenous hormone metabolism as an exposure marker in breast cancer chemoprevention studies

Michnovicz, Jj., Adlercreutz, H. & Eradlow, H.L. (1997) progestin replacement therapy and breast cancer risk. j. Changes in levels of urinary estrogen metabolites after Am. Med. Ass., 283, 485-491 oral indole-3-carbinol treatment in humans. J. Nat! Scheen, A.J. & Lefebvre, P.J. (1999) Troglitazone: antihy- 89, 718-723 Cancer Inst., perglycemic activity and potential role in the treatment Michnovicz, J.J. (1998) Increased estrogen 2-hydroxyla- of type 2 diabetes. Diabetes Care, 22, 1568-1577 indole -carbinol. tion in obese women using oral -3 mt. j. Secreto, G. & Zumoff, B. (1994) Abnormal production of Obes. Relat. Metab. Disord., 22, 227-229 androgens in women with breast cancer. Anticancer Res., Miller, W.R. & Mullen, P. (1993) Factors influencing aro- 14,2113-2117 matase activity in the breast. j. Steroid Biochem. Mol. Bio!., Singh, A., Purohit, A., Coidham, N.G., Ghilchik, M.W. & 44,597-604 Reed, MJ, (1998) Biochemical control of breast aro- Millikan, R.C., Pittman, G.S., Tse, C.K., Duell, E., matase. Breast Cancer Res, Treat., 49 Suppl. 1, 59-14 Newman, B., Savitz, D., Moorman, PG., Boissy, R.J. & Spicer, DV. & Pike, M.C. (1994) Sex steroids and breast Bell, D.A. (1998) Catechol-O-methyltransferase and cancer prevention. J. Nat! Cancer Inst. Monogr., 16, breast cancer risk. Carcinogenesis, 19, 1943-4947 139-447 Peyrat, J.P., Bonneterie, Hecquet, B., Vennin, P., J., Thissen,J.P., Ketelslegers,J.M. & Underwood, L.E. (1994) Louchez, MM., Fournier, C., Lefebvre, J. & Demaille, A. Nutritional regulation of the insulin-like growth factors. (1993) Plasma insulin-like growth factor-1 (IGF-1) con- Endocrino!, Rev., 15, 80-101 centrations in human breast cancer. Eur. J, Cancer, 29A, 492-497 Thomas, H.V., Reeves, G.K. & Key, T.J. (1997a) Endogenous estrogen and postmenopausal breast cancer: Pollak, M.N., Huynh, H.T. & Lefebvre, S.P. (1992) a quantitative review. Cancer Causes Control, 8, 922-928 Tamoxifen reduces serum insulin-like growth factor I (IGF-I). Breast Cancer Res. Treat, 22, 91-100 Thomas, H.V., Key, T.J., Allen, D.S., Moore, J.W., Dowsett, M., Fentiman, I.S. & Wang, D.Y. (1997b) A prospective Powles, T., Feles, R., Ashley, S., Easton, D., Chang, J., study of endogenous serum hormone concentrations Dowsett, M., Tidy, A., Viggers, J. & Davey, J. (1998) and breast cancer risk in premenopausal women on the Interim analysis of the incidence of breast cancer in the island of Guernsey. Br. J. Cancer, 75, 1075-1079 Royal Marsden Hospital tamoxifen randomised chemo- prevention trial. Lancet, 352, 98-101 Thompson, PA., Shields, P.G., Freudenheim, J.L., Stone, A., Vena, JE., Marshall, JR., Graham, S., Laughlin, R., Pugeat, M. & Ducluzeau, P.H. (1999) Insulin resistance, Nemoto, T., Kadlubar, F.F. & Ambrosone, C.B. (1998) polycystic ovary syndrome and metformin. Drugs, 58 Genetic polymorphisms in catechoi-O-methyltrans- Suppl. 1, 41-46 ferase, menopausal status, and breast cancer risk. Cancer Purohit, A., Howarth, N.M., Potter, B.V. & Reed, M.J. Res., 58, 2107-2110 (1994) Inhibition of steroid suiphatase activity by Ursin, G., London, S., Stanczyk, EZ., Gentzschein, E., steroidal potential therapeu- methyithiophosphonates: Paganini-Hill, A., Ross, R.K. & Pike, MC. (1999) Urinary tic agents in breast cancer. J. Steroid Biochem. Mo!. Bio!., 2-hydroxyestrone/16alpha-hydroxyesttone ratio and risk 48,523-527 of breast cancer in postmenopausal women. j. Nat! Purohit, A., Froome, V.A., Wang, DY,, Potter, B.V. & Cancer Inst., 91, 1067-1072 Reed, MJ. (1997) Measurement of oestrone sulphatase Veronesi, U., Maisonneuve, P., Costa, A., Sacchini, V., activity in white blood cells to monitor in vivo Maltoni, C., Robertson, C., Rotmensz, N. & Boyle, P. inhibition of steroid suiphatase activity by oestrone- (1998) Prevention of breast cancer with tamoxifen: pre- 3-0-suiphamate. J. Steroid Biochem, Mol. Bio!,, 62, 45-51 liminary findings from the Italian randomised trial Purohit, A., Hejaz, HA., Woo, L.W., van Strien, A.E., among hysterectomised women. Italian Tamoxifen Potter, B.V. & Reed, M.J. (1999) Recent advances in the Prevention Study. Lancet, 352, 93-9 7 development of steroid suiphatase inhibitors. J. Steroid Veronesi, U., De Palo, G., Marubini, E., Costa, A., Biochem. Mo!. Bio!., 69, 227.-238 Formelli, F., Mariani, L., Decensi, A., Camerini, T., Del Ross, RIZ., Paganini-Hill, A., Wan, P.C. & Pike, M.C. Turco, M.R., Di Mauro, MG,, Muraca, MG., Del Vecchio, (2000) Effect of hormone replacement therapy on breast M., Pinto, C., D'Aiuto, G., Boni, C., Campa, T., Magni, A., cancer risk: estrogen versus estrogen plus progestin. J. Miceli, R., Perloff, M., Malone, W.F. & Sporn, M.B, (1999) Nat! Cancer Inst., 92, 328-332 Randomized trial of fenretinide to prevent second breast Schairer, C., Lubin, J., Troisi, R., Sturgeon, S., Brinton, L. malignancy in women with early breast cancer. j. Nat! Cancer Inst., & Hoover, R. (2000) Menopausal estrogen and estrogen- 91, 1847-1856

161 Biomarkers in Cancer Chemoprevention

Werner, H. & Le Roith, D. (1997) The insulin-like growth Xu X., Duncan, A.M., Merz-Demlow, BE., Phipps, W.R. factor-1 receptor signaling pathways are important for & Kurzer, M.S. (1999) Menstrual cycle effects on urinary tumorigenesis and inhibition of apoptosis. Cdt. Rev. estrogen metabolites. J. Clin. Endocrinol. Metcth,, 84, Oncog., 8, 71-92 3914-3918 Woo, LW., Howarth, N.M., Purohit, A., Hejaz, HA., Zhu, B.T. & Conney, A.H. (1998) Functional rote of estro- Reed, M.J. & Potter, B.V. (1998) Steroidal and non- gen metabolism in target cells: review and perspectives. steroidal suhfamates as potent inhibitors of steroid sulfa- Carcinogenesis, 19,1-27 tase. J. Med. Chem., 41, 1068-1083

Corresponding author ,

R Kaaks Unit of Nutrition and Cancer, International Agency for Research on Cancer, 150 Cours Albert Thomas 69372 Lyon cedex 08 France

162