`
`TREATMENT AND PREVENTION OF
`
`BREAST CANCER
`
`ABSTRACT.—Tamoxifen is a nonsteroidal antiestro-
`
`gen that has found successful applications for each
`stage of breast cancer in the treatment of selected pa-
`tients. Tamoxifen was originally introduced for the
`treatment of advanced disease in postmenopausal
`women; however, the drug is now also available for
`the palliative treatment of premenopausal women
`with estrogen receptor (ER) positive disease. The
`proven efficacy of tamoxifen and the low incidence of
`side effects made the drug an ideal agent to test as an
`adjuvant therapy for women with node-positive breast
`cancer. Laboratory studies indicate that long-term
`treatment schedules may provide maximal benefit in
`preventing recurrence, and recent analysis of clinical
`trials demonstrates that between 2 and 5 years of ad-
`juvant tamoxifen therapy provides a survival advan-
`tage for postmenopausal women with node-positive
`disease. Similarly, adjuvant studies in node-negative
`breast cancer have demonstrated an increase in the
`
`disease-free survival of both pre- and postmenopausal
`patients with ER-positive tumors. However, the ex-
`tended use of tamoxifen has raised questions about
`the long-term safety of antiestrogen therapy. 0f spe-
`cial concern is the impact of tamoxifen on ovarian
`function in premenopausal women and the potential
`risks to the fetus if pregnancy occurs. Fortunately,
`there are no reports about the teratogenicity of ta-
`moxifen in the human, but it is important that phy-
`sicians counsel women about the risk of pregnancy.
`Tamoxifen should not be used if a patient is preg-'
`nant.
`
`Initial concerns that the long-term administration
`
`134
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 1
`InnoPharma Licensing LLC V. AstraZeneca AB
`IPR2017-00904
`
`
`
`of an antiestrogen would increase bone loss and in-
`crease the risks of coronary heart disease appear to
`be unwarranted. Tamoxifen has some estrogen-like
`activities in postmenopausal women and causes a
`preservation of bone in the lumbar spine and a de-
`crease in circulating cholesterol. Indeed, a reduction
`in fatal myocardial infarction (MI) has been noted dur-
`ing 5 years of tamoxifen therapy, possibly the direct
`result of a prolonged reduction in circulating choles-
`terol.
`
`However, the estrogen-like qualities of tamoxifen
`that could be valuable as a hormone replacement
`therapy for all postmenopausal women following a di-
`agnosis of breast cancer may also increase the risk
`for developing endometrial carcinoma. To date, there
`are only a few reports of endometrial carcinoma be-
`ing diagnosed during adjuvant therapy with tamox-
`ifen; however, any instances of uterine bleeding or
`spotting should be followed up with an endometrial
`biopsy.
`There are some concerns about large doses of ta-
`moxifen promoting liver cancer in rats. These results
`are of particular concern if tamoxifen is to be used as
`a preventive in normal women. However, tamoxifen is
`acting as a weak estrogen in the rat model, and other
`estrogens (eg, conjugated estrogens and estrogens
`used in oral contraceptives) are much more potent
`promoters of liver cancer. There are no reports of an
`increased incidence of hepatocellular carcinoma dur-
`ing long-term tamoxifen therapy; in fact, there is a re-
`port of the use of tamoxifen to treat the disease suc-
`cessfully.
`The estrogen-like properties of tamoxifen may en-
`courage clones of breast cancer cells to grow in re-
`sponse to the drug. This form of drug resistance has
`been described in the laboratory, and a new pure an-
`tiestrogen is being evaluated for use as a second-line
`agent following the failure of long-term adjuvant ta-
`moxifen therapy.
`The extensive clinical experience with tamoxifen
`has encouraged its evaluation as a preventive in
`women at risk of developing breast cancer. Several
`clinical studies are being positioned in different coun-
`tries to address this question. All the laboratory stud-
`ies demonstrate that tamoxifen will prevent or sup-
`press mammary cancer, and clinical experience has
`
`Curr Probl Cancer, May/June 1992
`
`135
`
`AstraZeneca Exhibit 2021 p. 2
`
`
`
`shown a 40% reduction in the incidence of second
`
`primary breast cancers.
`The completion of randomized prevention trials
`with tamoxifen will place a valuable medical option in
`the hands of the physician who treats women at risk
`for breast cancer.
`
`136
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 3
`
`
`
`THE ROLE OF TAMOXIFEN IN THE
`
`TREATMENT AND PREVENTION OF
`
`BREAST CANCER
`
`INTRODUCTION
`
`It has been known since the turn of the centuryl‘ 2 that about one-
`third of patients with advanced breast cancer will respond to some
`form of additive or ablative endocrine therapy. However, the reason
`for the apparently arbitrary responses remained unclear until the es-
`trogen receptor was describede‘_5 and utilized in selecting patients
`with a high probability of responding to endocrine therapy.6 Ap-
`proximately 60% of ER-positive patients will have an objective re-
`sponse to endocrine therapy.6
`Tamoxifen (ICI 46,474; Nolvadex®) is the trans isomer of a substi-
`
`tuted triphenylethylene that has antiestrogenic activity in laboratory
`animals7 and causes the regression of carcinogen-induced rat mam-
`mary tumors?—11 Tamoxifen is a competitive inhibitor of estrogen
`action and blocks the binding of [3H]estradiol to the E111?"13 Al-
`though a whole range of different nonsteroidal antiestrogens were
`described in the 19605,14 only tamoxifen was developed successfully
`for the treatment of breast cancer.”’ 16 The first preliminary clinical
`evaluation of tamoxifen to treat advanced breast cancer was con-
`
`in
`the Christie Hospital
`ducted by Cole and colleagues17 at
`Manchester. The efficacy of tamoxifen was equivalent to either an-
`drogen or high-dose estrogen therapy in postmenopausal patients,
`but the side effects noted with tamoxifen were modest. Several re-
`
`ports”"19 subsequently confirmed the efficacy of tamoxifen in pro—
`ducing palliation in postmenopausal women with advanced breast
`cancer.
`
`In 1973, Nolvadex®, the ICI brand of tamoxifen (as its citrate salt),
`
`was approved in the United Kingdom for the treatment of advanced
`breast cancer by the Committee on the Safety of Medicines. Similar
`approval was given in the United States by the Food and Drug Ad-
`ministration (FDA) on December 30, 1977. Nolvadex® is now avail-
`
`able in more than 110 countries as first-line endocrine therapy for
`
`Curr Probl Cancer, May/June 1992
`
`137
`
`AstraZeneca Exhibit 2021 p. 4
`
`
`
`the treatment of advanced breast cancer in postmenopausal women.
`Tamoxifen was initially used in premenopausal women to treat
`menometrorrhagia20 and to induce ovulation in infertile women.“ 22
`Although tamoxifen causes an increase in circulating estrogen levels
`during the treatment of advanced disease in premenopausal
`women,”24 it causes an objective response in about one-third of
`patients. Evidence has been accrued to demonstrate that tamoxifen
`produces a response rate similar to oophorectomy;25‘ 26 however, the
`reported studies are small and do not possess the statistical power
`to detect significant differences. Nevertheless, the FDA has approved
`the use of tamoxifen to treat ER-positive advanced breast cancer in
`premenopausal women.
`During the past 20 years, a dramatic change has occurred in the
`strategic application of antiestrogen to treat breast cancer. Approxi-
`mately 170,000 new cases of breast cancer will be diagnosed annu-
`ally in the United States, but less than 2.0% will be advanced disease.
`The rest will be node-positive and node-negative breast cancers with
`no detectable distant metastatic spread. Twenty years ago, patients
`would have waited for disease recurrence before treatment. The use
`
`of adjuvant therapy (ie, treatment following mastectomy) is now ac-
`cepted as a useful strateg for destroying micrometastases (Fig 1).
`Tamoxifen has found a valuable place in the medical armamentar—
`ium because it has been shown to have proven efficacy as an adju-
`vant therapy” and has a low incidence of side effects. Indeed ta-
`moxifen has been so successful that clinical trials are planned to
`study the use of tamoxifen in healthy women for preventing the de-
`velopment of breast cancer.”29
`During the next decade, up to 2 million women in the United
`States will have a diagnosis of breast cancer. It is hoped that tamox—
`ifen therapy will produce a beneficial effect in a significant propor-
`tion of these patients. The ubiquitous use of tamoxifen now man-
`dates a serious evaluation of both the strategic use of antiestrogens
`in the clinic and the potential consequences of long-term antihor—
`monal therapy.
`
`ADJUVANT TAMOXIFEN THERAPY
`
`CONCEPTS
`
`A variety of in viva laboratory models have been used to support
`the development of clinical trials that assess the efficacy of long-
`term or indefinite adjuvant therapy with tamoxifen in stages I and II
`disease. Unfortunately, these models do not precisely replicate the
`clinical situation, but their results clearly demonstrate the advan-
`tages of extended therapy in suppressing the appearance of palpa-
`
`138
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 5
`
`
`
`
`
`INITIATION
`
`CPCARCINOGENESIS
`
`PROMOTION
`
`I
`:
`o '0 0
`0° 0
`00000
`O 0000
`
`10-15 years
`
`Detection
`
`\ i-10years
`
`FIG 1.
`Development of breast cancer. At the point of detection, node-positive or negative disease
`can be classified as ER positive or negative. Adjuvant therapy with an antiestrogen can
`delay recurrence of the disease. but ER-negative disease that is refractory to treatment
`ultimately develops.
`
`ble tumors. Tamoxifen can therefore be considered a chemosup-
`pressive therapy.30
`The laboratory evidence demonstrating the tumoristatic proper-
`ties of tamoxifen is reviewed below, with examples from carcinogen-
`induced rat mammary tumor models and human breast tumors
`transplanted into athymic mice.
`
`Dimethylbenzanthracene-Induced Rat Mammary Carcinoma
`A single oral administration of dimethylbenzanthracene (DMBA, 20
`mg in 2 mL peanut oil) to 50-day-old female Sprague Dawley rats
`results in the appearance of palpable mammary tumors after 50 to
`150 days.31 The biology and endocrinology of this animal model
`have been studied extensively.32 Established tumors are hormone-
`dependent (predominantly prolactin) and respond to oophorectomy
`indirectly, by reducing estrogen-stimulated prolactin release (anties-
`trogens inhibit estrogen-stimulated prolactin release both in vivo33
`and in vitro“). Unfortunately, metastasis does not occur in this
`model, and a hormone-dependent model that precisely replicates
`adjuvant
`therapy is not available. Tamoxifen inhibits growth of
`
`Curr Probl Cancer, May/June 1992
`
`139
`
`AstraZeneca Exhibit 2021 p. 6
`
`
`
`DMBA-induced rat mammary tumorsf"11 However, this design does
`not provide a model for the treatment of minimal disease. To create
`this situation, DMBA can produce microfoci of transformed cells in
`the mammary chains after several weeks of promotion by circulating
`hormones. Tamoxifen therapy (in different doses or different dura-
`tions) can then be applied to determine whether the appearance of
`tumors can be preventedas‘39
`Tamoxifen (12.5 to 800 rig/day), administered for a month (starting
`1 month after DMBA), causes a dose-related decrease in the number
`
`of mammary tumors and a dose-related delay in the appearance of
`tumors.38 However, all the animals eventually develop at least one
`tumor. In contrast, continuous treatment with tamoxifen prevents
`tumorigenesist‘39 although termination of therapy results in the
`appearance of tumors.40 Therefore, tamoxifen exhibits the properties
`of a tumoristatic agent in the DMBA-induced rat mammary carci-
`noma model.
`
`N-nitrosomethylurea-Induced Hat Mammary Carcinoma
`The N—nitrosomethylurea (NMU) model was originally believed to
`be metastatic,41 but these reports have not been confirmed. Never-
`theless, NMU-induced mammary tumors have been shown to be
`more directly dependent on estrogen for growth,42 although a com—
`bination of pituitary and ovarian hormones is probably necessary to
`provide an optimal growth environment.
`Administration of NMU to female rats, followed by the daily ad-
`ministration of tamoxifen for several weeks, results in a delay in the
`appearance of palpable mammary tumors.”44 In contrast, the con-
`tinuous administration of tamoxifen completely suppresses the ap-
`pearance of palpable tumors.“‘45 Cessation of therapy (even of large
`doses of tamoxifen) does, however, result in the appearance of mam—
`mary tumors in up to 50% of animals.45
`
`Heterotransplantation of Breast Cancer Cell Lines
`Into Athymr'c Mice
`tumors can be grown successfully in athymic
`Primary breast
`(immune-deficient) mice.46 However, the take rate is often modest,
`and hormone-dependent growth is difficult to demonstrate rou-
`tinely.
`In contrast, human breast cancer cell lines will grow into solid tu-
`mors following inoculation into the mammary fat in the axillary re-
`gion of athymic mice. Eli-negative lines grow without further treat-
`ment, whereas ER-positive lines grow only if the animals receive
`supplemental estrogen therapy.”48 Tamoxifen and its metabélites
`inhibit estrogen—stimulated growth of MCF-7 breast
`tumors.”50
`However, tamoxifen does not appear to be tumoricidal; long-term
`
`140
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 7
`
`
`
`treatment with tamoxifen alone does not destroy MCF-7 cells inocu-
`lated into athymic mice,” 51 and therapy for up to 6 months does
`not prevent subsequent estrogen therapy from reactivating quies-
`cent tumor cells. Therefore, tamoxifen demonstrates the properties
`of a tumoristatic agent in this model.
`
`EVOLUTION OF CURRENT PRACTICE
`
`When the strategic decision was made to employ therapeutic
`agents following mastectomy, one important step was to select the
`duration of adjuvant tamoxifen therapy. The data from the DMBA
`model indicated that a long-term or indefinite treatment schedule
`might be the optimal
`treatment
`strategy.39 But how long is
`long when an extrapolation is required from the rat model to the
`patient?52
`The majority of investigators selected a conservative course of 1 or
`2 years of adjuvant tamoxifen. This decision, however, was based on
`a number of reasonable concerns. Patients with advanced disease
`usually respond to tamoxifen for perhaps 1 0r 2 years;53 it was ex-
`pected, therefore, that ER-negative disease would be encouraged to
`grow prematurely during adjuvant therapy. If this growth occurred,
`then the physician would have already used a valuable palliative
`drug and would have only combination chemotherapy to slow the
`relentless growth of recurrent disease. A related argument involved
`changing the treatment strategy for the application of combination
`chemotherapy. Recurrent treatment cycles (for 2 years) were found
`to be of no long-term benefit for the patient. An aggressive course of
`short-term treatment
`(6 months) with the most active cytotoxic
`drugs would have the best chance of killing tumor cells before the
`premature development of drug resistance. The same argument was
`used to substantiate an intuitive reluctance to use long—term tamox-
`ifen therapy. Longer might not be better because it would lead to
`premature drug resistance. Finally, sincere concerns were expressed
`about the side effects of adjuvant therapy and the ethical issue of
`treating patients who might not recur. Although this argument fo-
`cused primarily on chemotherapy and node-negative patients, it is
`fair to say that few patients in the mid-19705 had received extended
`therapy with tamoxifen, so long-term side effects were largely un-
`known. The majority of tamoxifen-treated patients had only received
`about 2 years of treatment for advanced disease before drug resis-
`tance occurred. The potential side effects of thrombosis, osteoporo-
`sis, etc, were only of secondary importance. The use of tamoxifen in
`the disease—free patient and its proposed use as a preventive obvi—
`ously would change that perspective. The evolving strategy for the
`adjuvant treatment of breast cancer is shown in Figure 2. However,
`
`Curr Probl Cancer, May/June 1992
`
`141
`
`AstraZeneca Exhibit 2021 p. 8
`
`
`
`Combination E Short course (6 mo.)
`Chemotherapy
`to avoid resistance
`
`Tamoxifen m Evolution of long-term
`1970's
`therapy to exploit
`tumoristatic action
`
`
`
`199% —-
`
`FIG 2.
`Adjuvant treatment with combination chemotherapy evolved to therapy with a short (6—
`month) course of noncross-resistant drugs to avoid the early development of drug resis-
`tance. In contrast. early use of 1 year of tamoxifen therapy to produce control of disease
`recurrence was extended up to 5 years to suppress early recurrence of the disease. Cur—
`rent trials are evaluating indefinite tamoxifen therapy as maintenance therapy.
`
`the situation is perhaps more complex because tamoxifen is either
`combined with adjuvant chemotherapy or given alone.
`
`Chemotherapy Plus Tamoxifen: Extending Tamoxifen Therapy
`In 1974, Hubay and coworkers54 initiated a clinical trial of chemo-
`therapy with cyclophosphamide, methotrexate, and 5-fluorouracil
`(CMF, low dose) versus chemotherapy plus tamoxifen (CMFT) versus
`CMFT plus bacillus Calmette-Guerin (BCG) vaccinations. No control
`group could be justified in light of the encouraging data for the suc-
`cessful adjuvant therapy with CMF, so 312 stage II pre— and post-
`menopausal patients were randomized to receive 1 year of treat-
`ment. Analysis of all the patients showed no benefit with the addi-
`tion of tamoxifen. Nevertheless, subanalysis of the ER-positive pa-
`tients approached significance (p = 0.08)
`for tamoxifen-treated
`groups and was significant in postmenopausal patients with four or
`more positive nodes (p = 0.05).
`The Ludwig Breast Cancer Group55 trial of 1 year of combination
`chemotherapy iCMF + prednisone) and tamoxifen in postmeno-
`pausal women showed a strong eflect (p < 0.0001) for the adjuvant
`therapy on disease-free survival when compared with observation
`only.
`The largest and most comprehensive analysis of adjuvant tamox-
`ifen and chemotherapy has been conducted by the National Surgical
`Adjuvant Breast Project (NSABP), in which 1,890 patients with stage
`II breast cancer (pre- and postmenopausal) were randomized to
`
`142
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 9
`
`
`
`combination chemotherapy with PF (L-phenylalanine mustard and
`5-fluorouracil), with or without tamoxifen?“58 The addition of ta-
`moxifen to chemotherapy in premenopausal patients provides no
`additional benefit.58 This may not be too surprising because chemo-
`therapy is known to cause ovarian failure,”60 so a maximal endo-
`crine response may have already occurred. However, some patients
`appear to show an adverse effect with the addition of tamoxifen. Pa-
`tients with progesterone receptor (PgR)—negative tumors do less well
`(p = 0.007) when tamoxifen is added to chemotherapy. The reason
`for the ability of tamoxifen to negate the benefits of PF chemother-
`apy is unknown. Laboratory studies61 have, however, demonstrated
`that tamoxifen can impair the action of some chemotherapeutic
`agents. A similar clinical effect has not been observed with the CMF
`chemotherapy regimen.
`An overall analysis of the NSABP trial (all entered patients) shows
`an increase in disease-free survival for the tamoxifen~containing reg-
`imen (p = 0,002).58 The benefit was observed entirely in the post-
`menopausal category. The response to tamoxifen is related to the
`nodal and steroid receptor status of the patients. Benefit with
`tamoxifen is noted in women over 50 years of age with steroid—
`receptor-positive tumors and four or more positive axillary nodes
`(p < 0.001).
`The early laboratory data39 indicate that long—term adjuvant tamox—
`ifen therapy may provide additional benefit. There are indications
`from both the Nolvadex Adjuvant Trial Organization (NATO)6?‘_64
`and NSABP studies of 2-year tamoxifen therapy that benefits accrue
`only as long as tamoxifen is given. This raises the clinical question of
`how long tamoxifen should be given.
`The first pilot clinical study of long-term tamoxifen therapy was
`initiated at the University of Wisconsin in 1977 by Tormey and Jor-
`dan.65 At the time this strategy was being considered, many patients
`with stage II disease had already completed 1 year of adjuvant ther-
`apy with combination chemotherapy and tamoxifen. Because the
`data from the DMBA model showed tamoxifen to be tumoristatic,
`
`therapy with tamoxifen was continued initially for 5 years; subse-
`quently, the decision was made to continue the drug indefinitely or
`until relapse.66 Patients are being maintained on indefinite tamox-
`ifen therapy, and some are now into their second decade of treat-
`ment. The patients have proved to be a valuable resource for studying
`the endocrinology,67‘69 pharmacology/,70 and safety of tamoxifen.71
`The potential efficacy of this treatment strateg/ was encouraging,
`and national randomized clinical trials were established through the
`Eastern Cooperative Oncology Group (ECOG). These trials are cur-
`rently being conducted (EST 5181 and EST 4181) to test the efficacy
`of chemotherapy and tamoxifen adjuvant therapy, or the combina-
`tion followed by another 4 years of tamoxifen versus indefinite treat-
`
`Curr Probl Cancer, May/June 1992
`
`143
`
`AstraZeneca Exhibit 2021 p. 10
`
`
`
`ment (Fig 3). An early analysis demonstrates that long-term tamox-
`ifen therapy confers an increase in disease-free survival.”
`The NSABP and the Stockholm trial”74 group have been per-
`suaded by the laboratory evidence and developing clinical data to
`embark on long—term tamoxifen studies. The NSABP has reported75
`on a successive registration trial of stage 11 breast cancer. Patients re-
`ceived chemotherapy (PF) and tamoxifen for 2 years, or the same
`regimen followed by an additional year of tamoxifen alone. The
`3-year tamoxifen therapy is superior to 2 years. However, the study
`is not randomized, and only patients who have successfully com—
`pleted 2 years of tamoxifen therapy are included for the third year.
`The Stockholm trial73'74 is much more complex; a flow chart is
`presented in Figure 4. The postmenopausal patient population is a
`mixture of node-positive and node-negative patients. Node-positive
`patients receive either radiotherapy or chemotherapy, whereas the
`majority of node-negative patients receive neither. All patients were
`subsequently randomized to 2 years of tamoxifen, 40 mg/day, or no
`endocrine therapy.
`The interim analysis (53 months) is encouraging.74 There is an in-
`crease in recurrence-free survival (p < 0.01), but not overall survival,
`
`A
`
`B
`
`STRATIFICATION
`Nodal Involvement
`
`1-3 posltlvo
`4-10 positlvo
`>10 pom“.
`
`Estrogen Receptor Result
`Palm.
`W"
`
`STRATIFICATION
`Nodal Involvement
`
`3mm
`>10 mm”
`
`59mg” mum” mw'"
`Posltlve
`Neg-five
`
`STEP ONE
`
`STEP TWO
`
`R “cum
`a
`'0' 12 We”!
`D
`O
`
`R
`a #Obummm
`D
`0
`
`"I".
`1' I AV b TH
`tor 12 total cycles
`
`III —-Cot°mug!“ TAM
`Z
`5 years
`E
`
`I"
`2
`E
`
`R
`
`a —-—cum x 12 cycles +
`D
`continuous TM! 10-! total of 5 years
`a ncurn tor 12 cycles + abumuon
`'2 —>CMFPT x 4 cycles + Oboervltlon
`E
`
`FIG 3.
`Schemas showing treatment in (A) ECOG study 5181 for premenopausal women and (B)
`ECOG study 4181 for postmenopausal women. After completion of 5 years of tamoxifen
`therapy, there is subsequent randomization to stop tamoxifen or continue it indefinitely. C
`= cyclophosphamide; M = methotrexate; F = 5 fluorouracil; P = prednisone; T = tamox-
`ifen; T8 = thiotepa; A = adriamycin; Vb = vinblastine; H = halotestin.
`
`144
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 11
`
`
`
`
`
`"High-risk" tumors
`(m 07 >33")
`- RT vs. Chemo
`- RT
`
`
`
`
`
`
`"Low-rlsk" tumors
`mom
`- mod. nu. mastectomy
`- breast conuwluon
`sum" ' m
`
`
`
`_ adjuvant
`1
`ndoerlno
`. my
`
`R
`A
`
`N '
`
`3
`,5
`I
`z
`E
`
`oa'lgoxmn
`‘0 mud."an
`
`2 yum
`
`alumina MRI—£002)!
`
`FIG 4.
`Design of Stockholm trial to determine effectiveness of long-term tamoxifen therapy.
`
`for the tamoxifen-treated arm. Tamoxifen therapy was ineffective in
`Eli-negative disease but very active in Ell-positive disease.
`An extension was initiated in 1983 to randomize patients in the
`tamoxifen-treated arm who were disease—free at 2 years to either
`stop tamoxifen or continue the drug for another 3 years. Information
`concerning the efficacy of this strategy is very encouraging"5 and has
`provided a wealth of valuable toxicological information and data
`about disease control.77“79
`
`Adjuvant Monotherapy With Tamoxifen: Postmenopausal Patients
`Several trials of tamoxifen monotherapy as an adjuvant to mastec—
`tomy were initiated toward the end of the 1970s. The major studies
`are summarized in Table 1. The study design compares tamoxifen
`therapy to a control arm,8° but only the ECOG trial was designed as
`a double-blind placebo study. Most of the clinical trial experience is
`with postmenopausal patients with stage 11 breast cancer. However,
`
`TABLE 1.
`Clinical Trials of Adjuvant Tamoxifen Monotherapy
`Daily Dose Duration
`Increase Disease
`Increased Correlation With
`(mg)
`(Year)
`Free Interval
`Survival
`Receptor Status
`
`Group
`
`Ludwig55
`Christie”82
`Danish33
`NATO62_ 64
`ECOG30
`Toronto‘
`French35
`Scottish"4
`
`20
`20
`30
`20
`20
`30
`40
`20
`
`1
`1
`1
`2
`2
`2
`3
`25
`
`Yes
`Yes
`Yes
`Yes
`Yes
`Yes
`Yes
`Yes
`
`No
`No
`No
`Yes
`No
`No
`Yes
`Yes
`
`Yes
`Not done
`Yes
`No
`Not analyzed
`Yes
`Yes
`Yes
`
`"Pritchard KI. Meakin JW. Boyd NF, et al: A randomized trial of adjuvant tamoxifen in postmenopausal
`women with axillary node positive breast cancer, in Jones SE, Salmon SE (eds): Adjuvant Therapy of
`Breast Cancer IV. Newv York, Grune & Stratton, 1984. pp 339—348.
`
`Curr Probl Cancer, May/June 1992
`
`145
`
`AstraZeneca Exhibit 2021 p. 12
`
`
`
`trials include both premenopausal patients and patients
`several
`with stage I disease. This heterogeneity has the potential to dilute
`and confound analysis of the results.
`Recently, Richard Peto of the University of Oxford27 has completed
`an overall analysis of the world literature on adjuvant therapy; such
`a review will not be attempted here. TWO points, however, do merit
`mention in this review: duration of therapy, and the relationship of
`receptor status of the primary tumor to response to tamoxifen ther-
`apy.
`Early trials with tamoxifen therapy were conservative, with only
`one year of therapy.“ 81—83 (Tamoxifen was then thought to be a tu—
`moricidal agent.) The largest study of 2 years of tamoxifen ther-
`apy has been conducted by the NATO in Great Britain and New
`Zealand.62’64 Only 46% of the entered patients (a mixture of pre-
`and postmenopausal patients with stage I or II disease) had
`hormone-receptor determinations, and these were assayed by differ-
`ent laboratories. This study is particularly important because it
`showed a survival advantage (p = 0.0019) for all patients receiving ta-
`moxifen, regardless of receptor status. These data are substantially
`supported by the Scottish trial,84 which also concluded that there is a
`steady improvement with tamoxifen treatment related to the receptor
`concentration of the primary tumor. The best disease-free survival is
`noted for those patients with ER values of over 100 fmol/mg cytosol
`protein. A controversial aspect of the Danish group’s report83 of 2-year
`tamoxifen adjuvant therapy is that, although patients with ER levels
`>100 fmol/mg cytosol protein had an increased disease-free survival
`(p < 0.01), there was an apparent reversal of the effect when the assay
`showed levels of 10—99 fmol/mg cytosol protein.
`All of the reported studies demonstrate an increase in the disease-
`free interval for tamoxifen—treated women either overall or for some
`
`re-
`subgroup. Clearly, analyses of whether local versus distant
`currences are controlled by tamoxifen will mirror survival advan-
`tages. The available data are limited and contradictory. The Christie
`stud 1'82 reports (nonsignificantly) fewer distant recurrences, as
`does the ECOG study.80 In contrast, the Ludwig group55 found that
`the improvement was confined to a lower incidence of locoregional
`recurrences in the tamoxifen arm. The NATO” and Scottish84 stud—
`
`ies support both positions, showing reductions in local and distant
`recurrences.
`
`In conclusion, longer treatment regimens with adjuvant tamoxifen
`appear to provide more benefit than shorter (1-year) regimens. Not
`all trials support the position that Eli-positive patients benefit from
`tamoxifen therapy. However, the revised Oxford overview analysis
`(1990) demonstrates that
`tamoxifen is more likely to benefit
`the
`receptor-positive patient.35
`
`146
`
`Curr Probl Cancer, May/June 1992
`
`AstraZeneca Exhibit 2021 p. 13
`
`
`
`Extended Tamoxifen Therapy Alone
`In May 1978, Delozier and coworkers initiated a trial of tamoxifen
`(40 mg/day) for 3 years versus no further treatment.86 The study pop-
`ulation was 179 postmenopausal women. Overall survival at 5 years
`did not differ significantly, but in ER-positive patients, tamoxifen im-
`proved both disease-free and overall survival. Tamoxifen had no ef-
`fect in EB—negative patients.
`In April 1978, the Scottish Trials Office (M11084 initiated a pilot
`clinical trial of tamoxifen versus no therapy in stage I or 11 breast
`cancer. This successfully evolved to become a randomized trial with
`1,312 participants. A total of 1,070 postmenopausal women were
`evaluated; almost half of these were stage I (node-negative) patients.
`Tamoxifen therapy (median = 5 years) produced a survival advan-
`tage (hazard ratio = 0.61) for node-positive patients. This trial of ad-
`juvant tamoxifen (10 mg bid) versus control shows the benefit of
`early long-term adjuvant tamoxifen therapy alone. The fact that no
`chemotherapy was included brings into question the value of com-
`bination chemotherapy over tamoxifen alone in postmenopausal
`women. TWO recent reports provide some insight into this dilemma.
`A recent study in Italy87 has evaluated the additional therapeutic
`advantage of CMF combination chemotherapy over 5 years of tamox-
`ifen alone. The Italian study demonstrates that tamoxifen alone is
`not significantly worse than when combined with CMF. In contrast,
`a recent NSABP study‘38 demonstrates that Adriamycin® (doxorubi-
`cin hydrochloride) and cyclophosphamide can enhance the efficacy
`of tamoxifen alone. Even though the analysis was early (3 years), a
`survival advantage was observed. There may, however, be some
`merit to combining a drug such as Adriamycin (the most potent sin-
`gle agent for chemotherapy of breast cancer) with triphenylethylene—
`type antiestrogens. Antiestrogens can prevent the efflux of cancer
`chemotherapeutic agents and enhance cytotoxicity.89
`The efficacy and patient acceptability provided the incentive to
`use long—term adjuvant tamoxifen therapy in node-negative disease.
`The NSABP90 established a double-blind, randomized clinical trial of
`pre- and post-menopausal women with Eli-positive node-negative
`disease. An early analysis shows an overall disease-free advantage for
`women who receive tamoxifen; but perhaps most importantly, pre-
`menopausal women also benefit. The investigation was originally to
`study 5 years of tamoxifen, but the decision has been made to con—
`tinue tamoxifen for up to 10 years. Rerandomization will compare 5
`versus 10 years of tamoxifen monotherapy. Overall, this large, ran-
`domized clinical trial provides an invaluable data base on the effi-
`cacy, toxicology, and patient acceptability of tamoxifen.
`In summary, these studies provide valuable support for the con—
`cept that long-term tamoxifen therapy can provide a survival advan-
`
`Curr Probl Cancer, May/June 1992
`
`147
`
`AstraZeneca Exhibit 2021 p. 14
`
`
`
`tage for receptor-positive patients with node-positive and node-
`negative disease. However, one of the immediate concerns is to de—
`velop an understanding of drug resistance with tamoxifen so that
`new strategies or therapeutic agents can be developed to build on
`the initial success of tamoxifen.
`
`FAILURE OF TAMOXIFEN THERAPY
`
`Long-term adjuvant tamoxifen therapy (up to 5 years) is now re-
`garded as the treatment of choice for selected groups of patients
`with breast cancer. Potential mechanisms of resistance to tamoxifen
`
`therapy include:
`
`0 hormone-independent (ER-negative) disease,
`' metabolism to estrogens, and
`o tamoxifen—stimulated tumor growth.
`
`At present, several forms of tamoxifen resistance have been de-
`scribed in the laboratory, but only Ell-negative growth has been
`identified in the clinic. We will consider drug resistance to tamox-
`ifen in order to develop additional treatment strategies with new an-
`tiestrogens.
`
`METABOLISM TO ESTHOGENS
`
`The use of long-term adjuvant tamoxifen therapy could produce
`metabolic tolerance and the conversion of tamoxifen (a derivative of
`
`a known estrogen, triphenylethylene) to.estrogenic metabolites. If
`this should occur, then tamoxifen would start to stimulate tumor
`
`grth t