`Copyright 0 1995 by Annual Review; Inc. All rights reserved
`
`‘D QENVNIEVCE Further
`Qutck links to onllne content
`
`Annu.Rev.Pharrnacol.Toxicol.1995.35:195-211.DownloadedfromwwwannualreviewsorgbyJohnsHopkins
`
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`TAMOXIFEN: Toxicities and
`
`Drug Resistance During the
`Treatment and Prevention of
`
`Breast Cancer
`
`V. Craig Jordan
`Robert H. Lurie Cancer Center, Northwestern University Medical School, Chicago,
`Illinois 60611
`
`KEY WORDS:
`
`endometrial cancer. rat liver cancer. antiestrogens. raloxifene, ICI 182,780
`
`ABSTRACT
`
`Tamoxifen, a nonsteroidal antiestrogen, is the endocrine therapy of choice for
`all stages of breast cancer. There are six million women-years of experience
`with tamoxifen, and the drug has produced survival advantages in node-posi-
`tive and node-negative patients who have had 2—5 years of adjuvant tamoxifen
`therapy. A low incidence of side effects has been reported with tamoxifen,
`resulting in the proposal to use the antiestrogen as a preventive agent for breast
`cancer. Three separate clinical trials are currently under way—in the United
`States, Italy, and the United Kingdom. Current concerns about tamoxifen are
`the development of rat liver tumors during long-term treatment and an in-
`creased incidence of endometrial carcinomas observed in patients. Another
`concern is the development of drug resistance to long-term tamoxifen therapy.
`There is increased interest in both determining the mechanism of drug resis-
`tance and evaluating new antiestrogens that may be more beneficial as a
`preventive, as an adjuvant therapy, or for the treatment of advanced breast
`cancer.
`
`INTRODUCTION
`
`Tamoxifen is a nonsteroidal antiestrogen (1—4) that exhibits antitumor prop-
`erties in laboratory animals (5—9). Although many compounds were investi-
`
`195
`
`0362-1642/95/0415-0195$05.00
`
`AstraZeneca Exhibit 2017 p. 1
`InnoPharma Licensing LLC V. AstraZeneca AB
`IPR2017-00904
`
`
`
`196
`
`JORDAN
`
`gated in the 1960s and 70s (10), only tamoxifen was successfully developed
`in the laboratory for the treatment of breast cancer (10-12). The initial focus
`of anti-breast cancer drug development in the 1970s was as a palliative therapy
`for advanced breast cancer. However, adjuvant therapy and the exploration of
`long-term tamoxifen therapy (13, 14) has revolutionized breast cancer treat-
`ment. An overview analysis of randomized clinical trials of adjuvant tamoxifen
`therapy demonstrates that
`long-term (i.e. more than one year) adjuvant
`tamoxifen is an appropriate strategy to control the recurrence of both node-
`positive and node-negative breast cancer (15). Tamoxifen confers a survival
`advantage to those women with breast cancer who are treated for more than
`two years. Tamoxifen is the hormonal treatment of choice for all stages of
`breast cancer.
`The use of tamoxifen as a treatment for breast cancer has over six million
`
`women-years of experience. The drug has a low reported incidence of serious
`side effects (16), and five years of therapy is commonplace (17).
`The success of tamoxifen as a treatment for breast cancer has fueled interest
`
`in the drug as a preventive agent. Estrogen is known to promote the develop-
`ment of breast cancer, so it is only natural that an antiestrogenic drug that has
`been extensively clinically tested would be the leading candidate for evalua-
`tion. The pharmacology of tamoxifen is complex (18, 19), but the scientific
`rationale for testing tamoxifen has merit. An important component of the drug’s
`pharmacology is the target-site specificity of tamoxifen; the drug can act as
`an antitumor agent (probably as an antiestrogen) in the breast but can also be
`estrogenic at physiological sites to prevent bone loss and decrease circulating
`cholesterol.
`
`Animal studies demonstrate that tamoxifen prevents mammary carcinogen-
`esis (5, 7, 20—22), and clinical studies show that adjuvant tamoxifen therapy
`decreases the incidence of second primary breast cancers by 40% (15). Post-
`menopausal bone density is maintained by tamoxifen treatment (23-25), which
`could ultimately lead to the prevention of osteoporosis. Tamoxifen also de-
`creases low-density lipoprotein cholesterol levels in postmenopausal women
`(26-28). This positive property of tamoxifen may be responsible for the de-
`crease in hospital visits for the treatment of any cardiac condition (29) and the
`significant decrease in fatal myocardial infarction for women treated for five
`years with tamoxifen (30, 31). These data provide a scientific basis for pla-
`cebo-controlled clinical trials to test tamoxifen’s ability to prevent breast
`cancer.
`
`PREVENTION STUDIES
`
`Unlike the laboratory models of mammary tumorigenesis, where all animals
`develop tumors and the efficacy of tamoxifen is readily demonstrated, targeting
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`AstraZeneca Exhibit 2017 p. 2
`
`
`
`TAMOXIFEN AND BREAST CANCER
`
`197
`
`the appropriate population at risk for breast cancer is difficult. Numerous risk
`factors have been identified, and these have been reviewed elsewhere (32).
`However, because the incidence of breast cancer is small in the general pop-
`ulation, women volunteers with a high-risk profile must be recruited. It is
`essential to design a double-blind, placebo-controlled trial, but the large num-
`bers of volunteers required and the long time period necessary to obtain a
`statistically significant result mandates data management, compliance moni-
`toring, and an enormous clinical trials effort.
`Three clinical trials are currently recruiting and following volunteers to test
`tamoxifen’s ability to prevent breast cancer. The first trial was begun at the
`Royal Marsden Hospital in 1986 (33, 34). The initial goal was to recruit 2000
`women as a Vanguard study and monitor the progress of the volunteers.
`Healthy women aged 30—70 are eligible provided they have a family history
`of breast cancer on the maternal side, with at least one first-degree relative
`(sister, mother, daughter) under the age of 45 years who has developed breast
`cancer or bilateral breast cancer or with a first-degree relative and at least one
`other maternal relative affected. The women are randomly assigned to receive
`20 mg of tamoxifen or a placebo daily for at least eight years. At five years
`(by June 1993) compliance for the 2012 women assigned to the tamoxifen arm
`of the study was greater than 70% (35).
`There is a significant increase in hot flashes (34% vs 20%), vaginal discharge
`(16% vs 4%), and menstrual irregularities (14% vs 9%) for tamoxifen- vs
`placebo-treated women. Safety monitoring shows no obvious effects on radial
`bone mineral density, but fibrogen, antithrombin III, and cholesterol levels
`decrease out to five years in the tamoxifen-treated women.
`Most importantly, the Marsden study demonstrates an increased incidence
`of uterine fibroids and benign ovarian cysts as a result of tamoxifen treatment.
`An in-depth study (36) of the postmenopausal women demonstrated that ta-
`moxifen causes potentially malignant changes in the endometrium, but trans-
`vaginal ultrasonography can be used to identify those women who should be
`monitored. These findings resulted in approval by the Department of Health
`(July 1993) to recruit an additional 15,000 women volunteers from sites around
`the United Kingdom. Recruitment of additional volunteers has also been con-
`ducted for more than a year in Australia.
`The second prevention trial began recruiting volunteers from throughout
`North America in May 1992. The study, funded by the National Cancer
`Institute, will recruit 16,000 women who will be randomly assigned to be
`treated with tamoxifen (20 mg daily) or a placebo for five years. Those eligible
`for entry into the study include any woman over the age of 60 or women
`between the ages of 35 and 59 years whose five-year risk of developing breast
`cancer, as predicted by the Gail model (37), equals that of a 60-year-old
`woman. Any woman over 35 years of age, with a diagnosis of lobular carci-
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`AstraZeneca Exhibit 2017 p. 3
`
`
`
`198
`
`JORDAN
`
`noma in situ (LCIS) treated with biopsy alone, is eligible for entry. In the
`absence of LCIS, the risk factors for entry vary with age, so a 35-year-old
`woman must have a relative risk of at least 5, while a 45-year-old woman’s
`relative risk must be 1.8 to be eligible for entry.
`Seven thousand women were recruited in the first year. Pretrial concerns
`that younger women at risk would not volunteer for the trial are not substan-
`tiated by the population distribution. About one third of the volunteers are
`35—50 years old, with relative risks ranging on average between 10 for the
`youngest participants to 4 for the 50-year-olds.
`By December 1993, 11,000 women had been recruited to the prevention
`trial. Administrative concerns about the development of uterine carcinoma
`during tamoxifen treatment halted recruitment to National Surgical Adjuvant
`Breast Project (NSABP) trials for several months during 1994, but recruitment
`was reinitiated in June 1994 after the Food and Drug Administration had
`reviewed the concerns.
`
`The final tamoxifen prevention trial was initiated in Italy (38) by the Euro-
`pean Institute of Oncology and the Milan Cancer Institute. Up to 20,000
`volunteers who are over the age of 45 but who have already had a hysterectomy
`will be recruited. The aim of these restrictions is to avoid the complications
`of both pregnancy and endometrial cancer. Volunteers are being randomly
`assigned to tamoxifen (20 mg daily) or a placebo for five years. There were
`more than 3000 volunteers recruited by July 1993.
`For the first time, the clinical trials community is evaluating a therapy to
`prevent breast cancer. Although the majority of women recruited to the trials
`will not develop breast cancer, they will experience symptoms and side effects
`from tamoxifen. The evaluation of the toxicity of tamoxifen in the trials is
`extremely important, not only to determine the therapeutic value of the inter-
`vention but also to assess whether compliance can be maintained by the study
`population. Extra attention is being paid to chronic toxicities.
`
`TOXICITIES OF TAMOXIFEN
`
`Considerable concern has been expressed about the potential toxic effects of
`tamoxifen that could become critical in any evaluation of the drug given to
`women without breast cancer. These concerns are listed in Table 1 and have
`
`been the subject of a recent commentary (39). Ocular problems and the small
`increase in thromboembolic disease has been adequately reviewed (16), but
`the potential of tamoxifen to be carcinogenic is a serious risk.
`In the laboratory, tamoxifen can stimulate the growth of human endometrial
`carcinoma but can block the growth of a breast tumor transplanted in the same
`immune-deficient mouse (40). This possibility was demonstrated in patients
`when a 40% decrease in second primary breast cancers but a threefold increase
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`AstraZeneca Exhibit 2017 p. 4
`
`
`
`TAMOXIFEN AND BREAST CANCER
`
`199
`
`Table 1 Potential long-term toxicities from tamoxifen therapy
`
`Ocular problems
`Thromboembolic disorders
`Endometrial cancers
`Liver cancer
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`in endometrial cancer was observed in an adjuvant clinical trial of tamoxifen
`(41). Seventeen endometrial cancers were reported for the 1000 patients taking
`tamoxifen after eight years of evaluation. The rate of detection is two per
`thousand per year (42). A similar rate is reported by the NSABP (43). However,
`the NSABP study reports 6 deaths out of the 23 patients who developed
`endometrial carcinoma over the 6 years of evaluation for the 2639 women
`(43). The causes of death and the association with the duration of tamoxifen
`treatment are shown in Table 2. It is clear that women must be monitored for
`
`the development of endometrial carcinoma, but perhaps most importantly,
`patients must be screened before therapy to ensure that preexisting endometrial
`carcinoma is not present.
`Investigators are currently interested in determining whether tamoxifen pro-
`duces a higher-grade, more aggressive endometrial carcinoma. Initial reports
`from the Yale—New Haven tumor registry suggested that women were “at risk
`for high-grade endometrial cancers that have a poor prognosis" (44). Current
`comparisons of histological grade and stage in patients treated with tamoxifen
`(42, 43, 45) demonstrate the same proportions noted for the general population.
`
`Table 2 Patients randomly assigned to the tamoxifen arms of NSABP
`BM who died of EC“
`
`Patient
`
`Age
`
`Time on
`tamoxifen
`(months)
`
`Diagnosis of EC
`after tamoxifen
`(months)
`
`Cause of
`death
`
`1
`
`2
`3
`4
`5
`6
`
`66
`
`68
`63
`58
`54
`68
`
`—
`
`EC
`
`Never took
`tamoxifen
`CV diseaseVD
`0
`5
`EC
`0
`9
`EC
`73
`22
`EC
`23
`42
`Pulmonary
`0
`65
`embolism
`
`“Endometrial cancer
`“Cardiovascular disease
`
`AstraZeneca Exhibit 2017 p. 5
`
`
`
`200
`
`JORDAN
`
`Nevertheless, tamoxifen does not retard the development of endometrial car-
`cinoma, and clonal selection may result in premature changes in pathology.
`These findings contrast with the effects of estrogens that only promote growth.
`In contrast, the concern about hepatocellular carcinoma is based upon lab-
`oratory studies alone. Numerous reports show that large doses of tamoxifen
`produce liver tumors in rats (46-48). Tamoxifen produces DNA adducts in rat
`liver (48—50) and protein adducts in vitro (51). It is hypothesized that tamoxifen
`can become metabolically activated through selective hydroxylation to form
`an unstable alkylating species (52). Nevertheless, even though DNA adducts
`can be formed by human liver microsomes in vitro (53), no practical demon-
`stration of DNA adduct formation has occured in humans. Indeed, there have
`
`been no reports of hepatocellularcarcinoma in women taking 20 mg tamoxifen
`daily and only two reports of hepatocellular carcinoma in women who took
`40 mg daily (41). Part of the problem is that the tumor is so rare—5 per 100,000
`of the population—that even a tenfold rise would be difficult to detect.
`The striking differences observed for the toxicology for tamoxifen in the rat
`may be species and dose related. Tamoxifen is used therapeutically at a dose
`of 250 ug/kg in both humans and rats (14). In contrast, rats are given 12 mg/kg
`of tamoxifen for half their lifetime to produce livertumors. This regimen (48)
`is approximately 40 times the human therapeutic dose given for about 8 times
`as long as the relatiVe human duration, which is five years or 6% of a woman’s
`life. The excessive doses of tamoxifen given to the rat could be overwhelming
`the capacity of the liver to such a degree that the rat must invoke unique
`metabolic routes to cope with the overdosing schedule. The equivalent exper-
`iment in humans would be a woman taking 800 mg tamoxifen daily for 40
`years.
`
`The relationship of liver carcinogenesis and tamoxifen is undoubtedly an
`important area of toxicology, not only to protect the treatment population but
`also to determine the relevance of certain animal models to human disease
`processes.
`
`Overall, the broad use of tamoxifen, both as a treatment and as a preventive
`agent in clinical trials has necessitated increased vigilance by the clinical
`community to identify untoward side effects as rapidly as possible. Tamoxifen
`is an effective therapy for breast cancer, and it has become an important
`cornerstone of treatment strategies. However, drug resistance is the inevitable
`result of any long-term therapy. The discovery of the processes of growth
`deregulation will provide additional therapeutic opportunities for the future.
`
`DRUG RESISTANCE
`
`The potential molecular mechanisms of resistance to antiestrogen therapy have
`recently been surveyed (54), but several new developments deserve comment.
`
`Annu.ReV.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`AstraZeneca Exhibit 2017 p. 6
`
`
`
`TAMOXIFEN AND BREAST CANCER
`
`201
`
`LOSS OF
`
`!_ RECEPTOR
`
`
`
`
`H UTATED
`"Begum;
`
`AmET'
`Ehi'hglbUCTION
`'
`H"
`
`
`lie—AL.—
`METABOLISM
`'—'
`
`
`
`
`
`TAMOXIFEN
`
`METABOLITES
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`INCREASED
`ESTRADI 0L
`
`Figure I The potential mechanisms of drug resistance against tamoxifen in human breast cancers.
`AEBP = antiestrogen binding protein. ER = estrogen receptor.
`
`Three major areas of concern are increased estrogen levels observed in pre-
`menopausal women during tamoxifen therapy, local tumor metabolism to
`unstable compounds that stimulate tumor growth, and the isolation of mutant
`estrogen receptors from tumors stimulated to grow by tamoxifen. These pos-
`sibilities are illustrated in Figure 1.
`
`High Estrogen Levels
`
`Tamoxifen causes an elevation in circulating estrogen levels in premenopausal
`patients (55—58), and a high estrogen environment could reverse the antitumor
`actions of tamoxifen. Tamoxifen is most effective in vitro in a low—estrogen
`environment; however, in vivo, tamoxifen is converted to a range of anti-
`estrogens with high affinity for the receptor. The reversal of the actions of
`tamoxifen as an antitumor agent is complicated. Studies in athymic mice
`demonstrate that high levels of circulating estradiol (>1600 pg/ml) will par-
`tially reverse the growth inhibitory effects of low levels (40 ng/ml) of circu-
`lating tamoxifen (59). Clinical experience demonstrates that tamoxifen alone
`is effective in the control of Stage I and IV breast cancer in premenopausal
`patients (15, 60—62). Nevertheless, if patients with Stage IV disease initially
`respond and then fail treatment, there is a 30% probability of a second response
`to oophorectomy (63).
`
`AstraZeneca Exhibit 2017 p. 7
`
`
`
`202
`
`JORDAN
`
`Local Metabolism
`
`The pharmacokinetics and metabolism of tamoxifen have been extensively
`studied in patients (64—66), and there is no evidence that high levels of estro-
`
`genic metabolites are produced during long-term therapy. However, the tumor
`cells or the stromal component could locally metaboliZe tamoxifen to potent
`estrogens that could stimulate tumor growth. In the laboratory long-term tam-
`oxifen treatment will eventually cause the growth of MCF-7 breast cancer cells
`transplanted into athymic mice (67, 68). The tumors are estrogenoreceptor
`positive and grow in either athymic rats or athymic mice in response to either
`estradiol or tamoxifen (68, 69). Pure antiestrogens will block tamoxifen-stim-
`ulated growth; therefore, tamoxifen must be converted to.estrogens locally to
`stimulate growth through the estrogen receptor (70).
`Tamoxifen is metabolized to 4-hydroxytamoxifen in the mouse (71). This
`metabolite is a potent antiestrogen (72) that is known to have antitumor activity
`in the athymic mouse model (73). However, the potent antiestrogenic Z isomer
`is unstable (74, 75) and can be converted to the weakly antiestrogenic E isomer
`(76—78). If the isomerization occurs locally, the net antiestrogenicity of tam-
`oxifen will decrease, but this would not in itself account for increased tumor
`
`growth because an estrogenic stimulus is required. Minute amounts of Metab-
`olite E (tamoxifen without the dimethylaminoethane side chain) have been
`detected in human tumors during tamoxifen therapy (79). Fortunately, this
`metabolite of tamoxifen is too weakly estrogenic to promote tumor growth
`alone (64, 78). Nevertheless, the metabolite is unstable and can isomerize to
`a potent estrogen (78). If large quantities of this estrogenic metabolite can
`accumulate in the tumors, they could be the stimulus for tamoxifen-stimulated
`tumor growth. The hypothesis (80, 81) that tamoxifen-stimulated growth de-
`pends upon the simultaneous isomerization of metabolites to weak antiestro-
`gens and potent estrogens is illustrated in Figure 2.
`We recently addressed the question of metabolite isomerization as the mech-
`anism of tamoxifen-stimulated growth by determining the ability of tamoxifen
`derivatives, which cannot isomerize, to cause tumor growth. We found that a
`fixed-ring version of tamoxifen (Figure 2) can adequately support and develop
`ligand-stimulated tumor growth (82). Osborne and coworkers (83) have con-
`firmed our finding but also report that the related compound toremifene (see
`section on new agents) stimulates tumor growth, as does a tamoxifen derivative
`lacking the ether oxygen in the alkylaminoethoxy side chain. This latter com-
`pound cannot form Metabolite B, so the hypothesis is untenable. At present it
`is unclear how tamoxifen-stimulated tumor growth occurs within the cell, but
`one possibility is that the development of mutated estrogen receptors could
`alter the pharmacology of the antagonist to an agonist.
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`AstraZeneca Exhibit 2017 p. 8
`
`
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`TAMOXIFEN AND BREAST CANCER
`
`203
`
`/
`ON'K
`
`NC
`
`POTENT
`ANTIESTFIOGEN
`
`Isousmzs
`____,'
`
`wEAK
`ANTIESTROGEN
`
`H0
`4-HYDROXYTAMOXIFEN
`N/
`0N \
`
`umon
`METABOLIC
`
`H0
`
`u<
`
`
`
`TAMOXIFEN
`
`BLOCK
`
`BLOCK
`
`BREAST
`RESPONSE TUMOR
`I GROWTH
`
`FIXED RING
`TAMOXIFEN
`
`ROUTE
`
`vsnv umon
`METABOLIC
`ROUTE
`
`I
`'
`:V
`
`0"
`
`WEAK
`
`EsTROGEN
`
`\
`
`¢——-—-—
`__._..
`lSOIIERflE
`H0
`
`POTENT
`ESTROGEN
`
`METABOLITE E
`
`Figure 2 A proposed scheme for the metabolism of tamoxifen in breast tumors that could cause
`tamoxifen-stimulated growth. Tamoxifen could be converted to the potent antiestrogen
`4-hydroxytamoxifen and the weak estrogen referred to as Metabolite E. The key event in the
`hypothesis is the ability of the metabolites in the tumor cells to isomerize to a weak antiestrogen and
`a potent estrogen. The fixed-ring derivative of tamoxifen that cannot isomerize is biologically active
`at promoting growth of tamoxifen-dependent breast tumors. This observation makes the proposed
`scheme unlikely to be the major mechanism for tamoxifen-stimulated growth.
`
`Mutated Estrogen Receptors
`
`There is considerable interest in determining the biological relevance of mutant
`steroid hormone receptors. Laboratory models have demonstrated that specific
`mutations of the androgen (84) and progesterone receptors (85) can change
`the biological properties of antiandrogens and antiprogestins to full agonist
`molecules. Therefore, mutations in the estrogen receptor that change the phar-
`macology of an antiestrogen to an estrogen could explain tamoxifen-stimulated
`growth in tumors.
`The screening of clinical tumor material and cell lines has resulted in the
`identification of several mutations of the estrogen receptor (86-88), but the
`
`AstraZeneca Exhibit 2017 p. 9
`
`
`
`204
`
`JORDAN
`
`biological relevance of these findings is unclear. However, the impact of point
`mutations in the estrogen receptor on the pharmacology of antiestrogens can
`be examined under laboratory conditions. If MDA-MB-231 breast cancer cells
`(receptor negative) are transfected with either a wild-type estrogen receptor
`(ER) gene or an ER gene with a glycine-to-valine mutation at amino acid 400,
`the resulting transfected clones will respond to estrogen by decreasing growth
`(89). This laboratory model then becomes a test of the estrogenicity of any
`ligand receptor complex under controlled conditions. Pure antiestrogens pre-
`vent the inhibitory effect of estradiol in both wild-type and mutant cDNA
`(HEO) transfectants (89).
`In contrast, the antiestrogens that are normally partial agonists in assays
`involving wild-type receptors only express estrogenic activity in HEO trans-
`fected cells (90, 91). We have proposed a model that describes the changes in
`the folding of the mutant receptor around the antiestrogen that produces an
`estrogenic coupling (91). However, the HBO mutant is known to be a labora-
`tory cloning artifact, and until recently, no single~point mutations of the re-
`ceptor had been observed in nature.
`Tamoxifen-stimulated MCF—7 breast tumors that grow in athymic mice
`appear to have normal estrogen receptors. However, a screen of mRNAs for
`estrogen receptors in tamoxifen-stimulated tumors, using first reverse trans-
`criptase and polymerase chain reaction followed by single-stranded conforma-
`tional polymorphism, revealed a tumor line containing an estrogen receptor
`with a mutation (92). The error is a single-point mutation in the codon, which
`converts an aspartate to a tyrosine at AA351 in the steroid-binding domain
`(93). Preliminary studies with the cDNA for the novel mutant receptor
`demonstrate high activity for the protein when MDA-MB-231 cells are trans-
`fected with the gene and, most importantly, a conversion of the pharmacology
`of nonsteroidal antiestrogens to potent estrogenic activity (94). Although the
`development of drug resistance to tamoxifen is clearly not caused by mutation
`of the estrogen receptor alone, this mutation may be one of the many mecha’
`nisms that come into play during long-term tamoxifen treatment.
`A drug development program is well under way by the pharmaceutical
`industry to find either new first-line antiestrogens, which are less toxic than
`tamoxifen, or new second-line antiestrogenic agents to be used after tamoxifen
`treatment failure.
`
`CONCLUSIONS AND FUTURE DEVELOPMENTS
`
`The enormous success of tamoxifen as a first-line endocrine therapy for all
`stages of breast cancer has encouraged a search for alternative antiestrogens
`that might ultimately replace, or at least compliment, tamoxifen. Extensive
`clinical testing of a number of tamoxifen derivatives is under way. Toremifene
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`AstraZeneca Exhibit 2017 p. 10
`
`
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`TAMOXIFEN AND BREAST CANCER
`
`205
`
`ON NMag
`
`ONNMlz
`
`Q]
`Q] “ a]
`Tamoxifen
`
`Q
`Q] “ g]
`Toreiititene
`
`NNMng
`
`0
`
`0H
`
`"°
`
`a
`
`3] @395
`
`HO
`
`%(CH2)93°(CH21:0F25Fa
`
`Droloxifene
`
`ICI 182,780
`
`Figure 3 The formulae of new antiestrogens for breast cancer therapy.
`
`(95, 96) and droloxifene (97) are completing phase III trials against tamoxifen
`in postmenopausal women (Figure 3). Idoxifene (98) is entering phase II trials
`in the United Kingdom.
`The pure antiestrogen ICI 182.780 (Figure 3) is in clinical trials in the United
`Kingdom (99) and offers the advantage that it could be used as a second-line
`therapy if long-term adjuvant treatment results in tamoxifen-stimulated tumor
`growth. This principle has been demonstrated in the laboratory (70). The pure
`antiestrogens have a complete inhibitory effect on estrogen action in the
`primate uterus (100), but most importantly, the mode of action appears to be
`different than the nonsteroidal antiestrogens. Pure antiestrogens cause the loss
`of estrogen receptor from tumors and estrogen target tissues (101—103); thus
`the tissue becomes refractory to additional estrogenic stimulation.
`Finally, new antiestrogens could be targeted for novel applications. The
`nonsteroidal antiestrogen keoxifene (now renamed raloxifene) (Figure 4) pre-
`serves bone density in the ovariectomized rat (104, 105), and large doses will
`prevent the development of rat mammary tumors (21). The compound also
`decreases circulating cholesterol in the rat (105) and has only a weak agonist
`
`AstraZeneca Exhibit 2017 p. 11
`
`
`
`Annu.Rev.Pharmacol.Toxicol.l995.35:l95-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`206
`
`JORDAN
`
`
`
`4-IHYDROXYTAMOXIFEN
`HIGH AFFINITY FOR en
`
`\
`
`0H ® RALOXIFENE
`IMPROVED BONE DENSITY
`TARGETED
`
`ESTROGEN/ANTIESTROGEN
`
`//I4
`
`{C H2}B SO(('.‘1‘|2):I C|=2CFa
`
`ICI 182, 780
`PURE ANTIESTROGEN
`
`Figure 4 The new antiestrogen strategies to be developed based on the knowledge about the high
`binding affinity of 4-hydroxytamoxifen for the estrogen receptor. Each of the new agents has high
`binding to the estrogen receptor. but unlike the pure antiestrogen 1C1 182,780, the antiestrogen
`raloxifene has target site—specific effects and will be used to treat osteoporosis.
`
`effect in the rat and mouse uterus (105). One could also predict that there is
`a low probability of rat liver carcinogenesis.
`Raloxifene is being developed as a treatment for osteoporosis in post-
`menopausal women. The large numbers of postmenopausal women who would
`be treated with raloxifene to prevent osteoporosis might also be protected from
`coronary heart disease and breast cancer and have with a low probability of
`developing endometrial carcinoma and hepatocellular carcinoma. A future
`decrease in the incidence of breast cancer may occur as a positive side effect
`from the prevention of osteoporosis by an antiestrogen with targeted estrogenic
`properties.
`
`Any Annual Review chapter, as well as any article cited in an Annual Review chapter,
`may be purchased from the Annual Reviews Preprints and Reprints service.
`
`1-800-347-8007; 415-259-5017; email: arpr@class.org
`
`Literature Cited
`
`1. Harper MJK, Walpole AL. 1966. Con-
`trasting endocrine activities of cis and
`trans isomers in a series of substituted
`triphenylethylenes. Nature 212:8?
`
`2. Harper MJK, Walpole AL. 1967. A new
`derivative of triphenylethylene: effect
`on implantation and mode of action in
`rats. J. Reprod. Fem'l. 131101—19
`
`AstraZeneca Exhibit 2017 p. 12
`
`
`
`Annu.Rev.Pharmacol.Toxicol.1995.35:195-211.Downloadedfromwww.annualreviews.orgbyJohnsHopkins
`
`
`
`
`
`
`
`Universityon06/12/14.Forpersonaluseonly.
`
`TAMOXIFEN AND BREAST CANCER
`
`207
`
`Labhsetwar AP. 1970. Role of estrogens
`in ovulation. A study using the estrogen
`antagonist ICI 46,474. Endocrinology
`87:542—51
`Jordan VC. 1975. Prolonged antiestroge-
`nic activity of 1C1 46,474 in the ovari-
`ectomized mouse. J. Reprod. Fem'l. 52:
`251-58
`Jordan VC. 1974. Antitumour activity
`of the antioestrogen ICI 46.474 (tam-
`oxifen) in the dimethylbenzanthracene
`(DMBA)-induced rat mammary carci-
`noma model. J. Steroid Biochem. 5:354
`Jordan VC. Koemer S. 1975. Tamoxifen
`(1C1 46.474) and the human carcinoma
`88 oestrogen receptor. Eur. J. Cancer
`11:205—6
`Jordan VC. 1976. Effect of tamoxifen
`(ICI 46,474) on initiation and growth of
`DMBA-induced rat mammary carcino-
`mata. Eur. J. Cancer 12:419—24
`Jordan VC, Koemer S. 1976. Tamoxifen
`as an antitumour agent: role of oestradiol
`and prolactin. J. Endocrinol. 68:305-10
`Jordan VC. Jaspan T. 1976. Tamoxifen
`as an antitumour agent: oestrogen bind-
`ing as a predictive test for tumour re-
`sponse. J. Endocrinol. 68:453—60
`Jordan VC. 1984. Biochemical pharma-
`cology of antiestrogen action. Pharma-
`col. Rev. 36:245—76
`Jordan VC. 1988. The development
`of tamoxifen for breast cancer ther-
`apy: a tribute to the late Arthur L.
`Walpole. Breast Cancer Res. Treat.
`112197—209
`Lerner LI, Jordan VC. 1990. Develop
`ment of antiestrogens and their use in
`breast cancer: Eight Cain Memorial
`.sAward Lecture. Cancer Res. 50:4177-
`9
`Jordan VC. 1978. Use of the DMBA-
`induced rat mammary carcinoma system
`for the evaluation of tamoxifen treat-
`ment as a potential adjuvant therapy.
`Rev. Endocr. Relat. Cancer Suppl. Oct:
`49—55
`Jordan VC. 1983. Laboratory studies to
`develop general principles for the adju—
`vant
`treatment of breast cancer with
`antiestrogens: problems and potential for
`future clinical applications. Breast Can-
`cer Res. Treat. 3(Suppl.):73-86
`Early Breast Cancer Trialists' Collabo-
`rative Group. 1992. Systemic treatment
`of early breast cancer by hormonal,
`cytotoxic or immune therapy. Lancet
`339:1-15
`Morrow M, Jordan VC. 1993. Risk fac-
`tors and the prevention of breast cancer
`with tamoxifen.
`In Cancer Surveys:
`Breast Cancer, ed. JT Papademitriou, I
`Fentiman. 18:211—29. Imperial Cancer
`
`10.
`
`11.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`
`23.
`
`24.
`
`25.
`
`26.
`
`27.
`
`28.
`
`Res. Fund. London: Cold Spring Harbor
`Lab. Press
`1994. Long Term
`Jordan VC. ed.
`Tamoxifen Treatmentfor Breast Cancer.
`Madison: Univ. Wis. Press
`Jordan VC, Murphy CS. 1990. Endo-
`crine pharmacology of antiestrogens as
`Entitumor agents. Endocr. Rev. 11:578-
`10
`Jordan VC. 1993. Fourteenth Gaddum
`Memorial Lecture: A current view of
`tamoxifen for the treatment and preven-
`tion of breast cancer. Br. J. Pharmacol.
`110:507-17
`Turcot-LeMay L. Kelley PA. 1980.
`Characterizatio