REVIEW ARTICLE
`A r o m a t a s e I n h i b i t o r s i n t h e T r e a t m e n t a n d P r e v e n t i o n o f
`B r e a s t C a n c e r
`
`By Paul E. Goss and Kathrin Strasser
`
`Purpose: The purpose of this article is to provide an
`overview of the current clinical status and possible
`future applications of aromatase inhibitors in breast
`cancer.
`Methods: A review of the literature on the third-
`generation aromatase inhibitors was conducted. Some
`data that have been presented but not published are
`included. In addition, the designs of ongoing trials with
`aromatase inhibitors are outlined and the implications
`of possible results discussed.
`Results: All of the third-generation oral aromatase
`inhibitors—letrozole, anastrozole, and vorozole (non-
`steroidal, type II) and exemestane (steroidal, type I)—
`have now been tested in phase III trials as second-line
`treatment of postmenopausal hormone-dependent
`breast cancer. They have shown clear superiority com-
`pared with the conventional therapies and are there-
`
`SEVERAL CLASSES OF endocrine agents that antago-
`
`nize the effects of estrogen are useful in the treatment
`of estrogen receptor (ER)–positive breast cancer.1 For
`example, selective estrogen receptor modulators (SERMs)
`and pure antiestrogens antagonize ER function by binding
`competitively to the receptor. Steroidal antiestrogens addi-
`tionally reduce ER concentration by inducing estrogen
`receptor degradation.2 Surgical, medical, and radiation-
`induced ovarian ablation and aromatase inhibitors antago-
`nize the action of estrogen by reducing its levels both in the
`circulation and in normal and malignant breast tissue.
`Aromatase (estrogen synthetase) inhibitors have become
`the established second-line treatment for ER-positive met-
`astatic breast cancer after the SERM tamoxifen. The third-
`generation aromatase inhibitors are currently being com-
`pared with tamoxifen in first-line metastatic, adjuvant, and
`neoadjuvant settings. Should they prove superior to tamox-
`ifen in terms of disease response,
`toxicity, and, most
`importantly, patient survival, they might replace tamoxifen
`as first-line endocrine therapy. Based primarily on a supe-
`rior side effect profile, anastrozole has recently been ap-
`proved as first-line therapy of metastatic breast cancer in
`several countries. The efficacy and excellent tolerability of the
`newer aromatase inhibitors in the treatment of breast cancer
`might lead to their use as chemopreventives in healthy women
`considered at significant risk of developing breast cancer. To
`this end, studies are underway to investigate their ability to
`alter surrogate markers of breast cancer risk.
`
`fore considered established second-line hormonal
`agents. Currently, they are being tested as first-line
`therapy in the metastatic, adjuvant, and neoadjuvant
`settings. Preliminary results suggest that the inhibitors
`might displace tamoxifen as first-line treatment, but
`further studies are needed to determine this.
`Conclusion: The role of aromatase inhibitors in pre-
`menopausal breast cancer and in combination with
`chemotherapy and other anticancer treatments are ar-
`eas of future exploration. The ongoing adjuvant trials
`will provide important data on the long-term safety of
`aromatase inhibitors, which will help to determine their
`suitability for use as chemopreventives in healthy
`women at risk of developing breast cancer.
`J Clin Oncol 19:881-894. © 2001 by American
`SocietyofClinicalOncology.
`
`In this article, the rationale for the use of aromatase
`inhibitors in breast cancer treatment, their mechanism of
`action, and preclinical test systems used in their evaluation
`are briefly reviewed. The current clinical status of third-
`generation aromatase inhibitors is discussed and ongoing
`clinical trials of these agents are described. Possible future
`applications of aromatase inhibitors in the treatment and
`prevention of breast cancer are also outlined.
`There may be specific biologic and pharmacologic rea-
`sons for giving aromatase inhibitors after tamoxifen. On the
`
`From the Division of Hematology/Oncology, Princess Margaret
`Hospital, Toronto, Ontario, Canada.
`Submitted April 27, 2000; accepted September 18, 2000.
`Over the past 10 years, P.E.G. has received industry funding for
`investigator-initiated clinical and laboratory studies of aromatase
`inhibitors as well as honoraria for presenting papers or acting in a
`scientific advisory capacity. Support of this nature has been received
`from manufacturers of all of the third-generation inhibitors that have
`been tested and/or approved for use, including vorozole (Janssen Ortho
`Inc, North York, Toronto, Ontario), letrozole (Novartis Pharmaceuti-
`cals Canada Inc, Dorval, Quebec), exemestane (Pharmacia & Upjohn,
`Mississauga, Ontario), anastrozole (AstraZeneca, Mississauga, On-
`tario, Canada), and liarozole (Janssen Ortho). K.S. has not received
`any financial support from industry.
`Address reprint requests to Paul E. Goss, MD, PhD, Division of
`Hematology/Oncology, Princess Margaret Hospital, 610 University
`Ave, Toronto, ON M5G 2M9, Canada; email: pegoss@interlog.com.
`© 2001 by American Society of Clinical Oncology.
`0732-183X/01/1903-881
`
`Journal of Clinical Oncology, Vol 19, No 3 (February 1), 2001: pp 881-894
`
`881
`
`Downloaded from ascopubs.org by Univ of Chicago Library on October 30, 2016 from 205.208.122.242
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1074-0001
`
`

`
`882
`
`the inhibitors may be more effective than
`other hand,
`tamoxifen if given as first-line treatment. For these reasons
`and also because tamoxifen is the current standard of care as
`first-line hormonal therapy for metastatic disease, as adju-
`vant therapy and as an approved chemopreventive in the
`United States, we have structured this review as aromatase
`inhibitors after tamoxifen, as first-line therapy, and in
`combination with other agents.
`
`INHIBITING ESTROGEN SYNTHESIS AS A
`THERAPEUTIC TARGET
`Aromatase is the enzyme complex responsible for the
`final step in estrogen synthesis, viz the conversion of the
`androgens androstenedione and testosterone to the estrogens
`estrone (E1) and estradiol (E2). There are substantial data
`showing that estrogen promotes and probably initiates
`breast cancer.3 Inhibiting estrogen at
`the source of its
`synthesis is therefore a logical
`target of breast cancer
`treatment.
`The sites of estrogen synthesis include the ovaries of
`premenopausal women; extragonadal sites such as fat,
`muscle, and skin; normal breast stromal cells; and breast
`tumor tissue. After ovarian failure, estrogen is synthesized
`in peripheral tissues and circulates at low, relatively non-
`fluctuating levels.4,5 This peripheral aromatization in post-
`menopausal women is almost completely inhibited by
`single-agent administration of any of the third-generation
`inhibitors.6,7 In contrast, there is a barrier to using aromatase
`inhibitors as monotherapy in premenopausal women. First,
`high levels of androstenedione compete initially with the
`inhibitors as substrate for the enzyme complex and conse-
`quently estrogen synthesis is not completely blocked.8-10
`Second, suppression of estrogen results in a reflex increase
`in gonadotrophin levels, provoking an ovarian hyperstimu-
`lation syndrome, which causes a steep increase of aromatase
`in the ovary and in turn overcomes, at least in part, the
`initial blockade to estrogen synthesis by the inhibitor.11
`However, although both type I (steroidal) and type II
`(nonsteroidal) inhibitors compete initially with the androgen
`precursors for the enzyme, the type I inhibitors subsequently
`inactivate the enzyme irreversibly, thus being referred to as
`suicide inhibitors. Therefore, with ongoing exposure to type
`I inhibitors ovarian estrogen synthesis might in principle be
`more completely suppressed. However, in premenopausal
`women given the second-generation inhibitor formestane
`this was not the case and estradiol levels were not signifi-
`cantly suppressed by monotherapy.12 Thus to date, aro-
`matase inhibitors have been tested predominantly in com-
`bination with GnRH-analogs in premenopausal women.
`However, with the more potent
`third-generation type I
`suicide inhibitor exemestane,
`the possibility of mono-
`
`GOSS AND STRASSER
`
`therapy in premenopausal women merits further investiga-
`tion at standard and higher doses.
`Increasingly, the female breast has itself been recognized
`as another important site of estrogen production. Stromal
`cells in breast adipose tissue produce estrogen that
`is
`biologically active in both a paracrine and an autocrine
`manner.13 This is probably responsible for the observation
`that estrogen concentrations in the healthy breasts of post-
`menopausal women are unexpectedly higher (four- to six-
`fold) than in serum and similar to those in premenopausal
`women.14 In addition up to 70% of breast cancer cells have
`been shown to synthesize estrogen as a result of intracellular
`aromatase expression.15-18 This explains why aromatase
`expression and activity are higher in breast tumors than in
`peritumoral fat and in tumor-bearing quadrants of the breast
`compared with those without tumors.19-23 There is increas-
`ing evidence that this local estrogen production may play a
`major role in tumor proliferation.24-27 Intratumoral aro-
`matase has been linked to response to the aromatase
`inhibitor aminoglutethimide18,28 but surprisingly not
`to
`estrogen receptor expression.18,29 Despite similar depletion
`of serum estrogen levels with the current third-generation
`aromatase inhibitors, variability in patient outcome on these
`drugs could be attributable to differences in inhibition of
`local estrogen synthesis.
`
`MODELS FOR EVALUATING AROMATASE INHIBITORS
`
`Potency and Reversibility
`
`For in vitro assessment of aromatase inhibitory capabil-
`ity, microsomal preparations from rat ovaries or from
`human placenta are used.30,31 Inhibition of the enzyme and
`potency of the inhibitor are determined by the amount of
`tritiated water released in the assay. By washing the
`microsomal preparations and measuring residual inhibition
`of aromatase, the inhibitor can be classified as reversible or
`irreversible.
`Depletion of serum estrogen levels has been used as a
`measure of the potency of aromatase inhibitors in blocking
`estrogen synthesis in peripheral tissues. However, using
`traditional assays, suppression below the detection limit has
`been noted with all of the third-generation inhibitors. This
`has made differentiating them clinically from one another
`difficult. In part this has been overcome by using a highly
`sensitive isotopic kinetic assay that relies on infusing
`[73H]androstenedione and [414C]estrone and measuring the
`conversion of androstenedione to E1 and E2. This assay has
`been used in male rhesus monkeys and in both healthy male
`volunteers and female breast cancer patients.31,32 Recently,
`more sensitive antibodies have also been developed. These
`have allowed differences in serum estrogen suppression to
`
`Downloaded from ascopubs.org by Univ of Chicago Library on October 30, 2016 from 205.208.122.242
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1074-0002
`
`

`
`AROMATASE INHIBITORS IN BREAST CANCER
`
`883
`
`Nonsteroidal
`
`Steroidal
`
`Table 1. Classification of Aromatase Inhibitors
`
`First Generation
`
`Aminoglutethimide
`
`Second Generation
`
`Third Generation
`
`Rogletimide
`Fadrozole
`
`Formestane
`
`Anastrozole
`Letrozole
`Vorozole
`Exemestane
`
`be demonstrated in postmenopausal women given various
`third-generation inhibitors.33
`
`hypertrophy can therefore be used as a measure of an
`inhibitor’s effect on intratumoral aromatase activity.
`
`Selectivity
`
`By incubating adult hamster ovarian tissue with luteiniz-
`ing hormone, the production rates of estrogen, progesterone
`and testosterone can be determined. Differences in the
`concentration that inhibits 50% for these steroid hormones
`are correlated with selectivity of suppression, an important
`feature of third-generation aromatase inhibitors.34
`
`Antitumor Activity and Chemopreventive Potential
`
`The animal models that have been used to demonstrate
`antitumor efficacy have included the hormone-dependent
`carcinogen-induced MNU and DMBA rat mammary tu-
`mors35,36 and spontaneous tumors in Sprague-Dawley
`rats.37 Several scenarios analogous to the clinical status of
`patients can be evaluated in these models. For comparability
`to treatment of breast cancer, reduction of established
`tumors and inhibition of tumor multiplicity are used. To
`determine their chemopreventive effects, aromatase inhibi-
`tors have been given before or after carcinogen administra-
`tion. Inhibition of tumor formation in these animals is
`viewed as a surrogate model for prevention of tumor
`initiation or promotion in humans.36
`The recently developed aromatase-transgenic mouse
`model (int-5/aromatase) allows evaluation of the effects of
`aromatase inhibitors on aromatase-overexpressing breast
`tissue.25 In these ovariectomized mice, aromatase overex-
`pression leads to increased estrogenic activity specifically in
`the mammary glands, resulting in the initiation of various
`preneoplastic changes such as hyperplasia and dysplasia.
`The ability of inhibitors to block or reduce these effects has
`been tested.26
`A useful model for assessing the effects of inhibitors
`directly on intratumoral aromatase is the MCF-7CA cell line.
`This is an MCF-7 cell line transfected with the human
`placental aromatase gene (MCF-7CA), which results in a
`10-fold increase in the expression of aromatase. When
`xenografted in athymic nude mice, which have been ovari-
`ectomized, this cell line is able to act directly as an estrogen
`“pump.”38,39 Inhibition of tumor growth or of uterine
`
`CLASSIFICATION OF AROMATASE INHIBITORS
`Aromatase inhibitors have been classified in a number of
`different ways, including first-, second-, and third-genera-
`tion; steroidal and nonsteroidal; reversible (ionic binding),
`and irreversible (suicide inhibitor, covalent binding)40-42
`(Table 1). A figure of the structures of the most important
`aromatase inhibitors is presented in Fig 1.
`The clinical significance of classifying the third-genera-
`tion inhibitors is uncertain. In the presence of ongoing drug
`administration,
`it
`is arguable whether irreversibility of
`enzyme inhibition is relevant. On one hand, comparable
`depletion of circulating estrogen in postmenopausal women
`to below the level of sensitivity of traditional radio-immu-
`noassays has been reported with either reversible or irre-
`versible third-generation inhibitors. However, as mentioned
`previously, more sensitive assays recently developed have
`helped to distinguish the capability of the different inhibi-
`tors in suppressing estradiol levels. Furthermore, irrevers-
`ible inhibition of aromatase may be relevant in suppressing
`premenopausal ovarian estrogen synthesis as mentioned
`above, and enzyme-binding characteristics may also be
`
`Fig 1. Structures of aromatase inhibitors.
`
`Downloaded from ascopubs.org by Univ of Chicago Library on October 30, 2016 from 205.208.122.242
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1074-0003
`
`

`
`884
`
`in the development of clinical resistance to
`important
`different classes of aromatase inhibitors. Steroids (eg, ex-
`emestane) also impart to an inhibitor the potential to affect
`other steroid levels (eg, androgens), either directly by the
`parent compound or indirectly by its metabolites. This could
`be relevant to mechanisms of tumor resistance and also
`might influence the potential of steroidal inhibitors to act as
`chemopreventives and to exert effects on other systems such
`as bone and lipid metabolism. Thus dissimilarities between
`the two nonsteroidal third-generation reversible inhibitors
`letrozole and anastrozole and the recently approved steroi-
`dal third-generation irreversible inhibitor exemestane may
`afford different clinical applications and therapeutic indices
`for these compounds.
`
`AROMATASE INHIBITORS AS MONOTHERAPY
`
`After Tamoxifen
`
`There are at least two preclinical observations suggesting
`that aromatase inhibitors may be particularly suitable after
`initial treatment with tamoxifen. First, in vitro hormone-
`dependent MCF-7 cells develop estrogen hypersensitivity
`when passaged in estrogen-deprived media.43 This leads to
`growth response to estrogen in concentrations four orders of
`magnitude lower than usually required.43 In vivo experi-
`ments have also shown that MCF-7 cells in nude mice
`initially regress in response to tamoxifen but are later
`stimulated by its weak estrogen agonist properties.44 Sec-
`ond, estrogen-deprived MCF-7 cells develop upregulation
`of aromatase, which in turn may result in increased auto-
`crine stimulation by estrogen.43 In principle,
`tamoxifen
`might have the same effect.
`Thus, theoretically, cessation of tamoxifen in a patient
`with disease progression and initiation of an aromatase
`inhibitor might simultaneously withdraw tamoxifen’s estro-
`gen agonist effect and deplete both locally produced and
`circulating estrogen to which the disease may be exquisitely
`sensitive.43,45
`These principles have been tested in several trials of
`aromatase inhibitors as second-line hormonal therapy in
`patients who experience disease progression while receiving
`tamoxifen. In this context, first-line endocrine therapy with
`tamoxifen means both as adjuvant and as first-line treatment
`for metastatic disease and both types of patients were
`enrolled in the metastatic second-line trials discussed be-
`low. Studies of aromatase inhibitors as third-line therapy are
`included, because most patients in these trials were also
`treated with tamoxifen as first-line therapy.
`The same strategy of giving an aromatase inhibitor after
`tamoxifen is being extensively studied in the adjuvant
`setting, and these trials are also discussed in detail below.
`
`GOSS AND STRASSER
`
`Finally, although the potential of aromatase inhibitors as
`monotherapy and single-agent treatment in chemopreven-
`tion is discussed in the next section, it is conceivable that the
`strategy of tamoxifen followed by an aromatase inhibitor
`might also be applicable in this setting.
`After tamoxifen as second-line therapy of metastatic
`For many years the progestin megestrol acetate
`disease.
`and the first-generation aromatase inhibitor aminoglute-
`thimide were the standard of care as second-line hormonal
`treatment of postmenopausal metastatic breast cancer after
`tamoxifen. Because they showed comparable clinical effi-
`cacy despite their different mechanisms of action, it was
`believed that the maximum potential of endocrine therapy
`had been reached. The side-effect profiles of these drugs,
`however, are clearly troublesome and frequently lead to
`toxicity-related withdrawal of treatment.
`The third-generation nonsteroidal aromatase inhibitors
`anastrozole, letrozole, and vorozole and the steroidal inhib-
`itor exemestane have significantly superior toxicity profiles
`compared with those of these conventional therapies and, to
`some extent, greater clinical efficacy. They have now all
`been studied as second-line therapy after tamoxifen against
`megestrol acetate, and letrozole and vorozole have also been
`compared with aminoglutethimide.46-55 Table 2 lists the
`results of these trials, including those from the recently
`published exemestane versus megestrol acetate trial. Only
`the doses that were approved for use are presented. Data
`from the two trials of anastrozole versus megestrol acetate
`were combined because the trial designs were identical.
`Significant efficacy and/or toxicity advantages were dem-
`onstrated for all of the inhibitors. Furthermore, none of them
`were significantly inferior to the comparator in any end
`point of efficacy. Importantly,
`in all
`trials,
`the third-
`generation aromatase inhibitors showed a significant advan-
`tage over standard treatment in at least one end point of
`toxicity. In particular, they were all clearly superior to
`megestrol acetate in terms of weight gain. The toxicity
`profiles of the third-generation inhibitors are similar, with
`the most common adverse events being nausea, vomiting,
`hot flashes, fatigue, and headaches. Importantly, the toxicity
`profiles reported from these trials are influenced by the fact
`that the patients were coming off treatment with tamoxifen
`(with its long half-life), and more accurate assessment will
`be possible from the first-line metastatic and adjuvant trials.
`In the studies that evaluated and reported quality of life,
`significant improvements compared with the conventional
`therapies were seen. None of the third-generation aromatase
`inhibitors have been compared head-to-head, and because
`of clear differences in trial designs and patient populations,
`the present studies are not comparable, either in terms of
`toxicity or efficacy. This has been reviewed in detail by
`
`Downloaded from ascopubs.org by Univ of Chicago Library on October 30, 2016 from 205.208.122.242
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1074-0004
`
`

`
`AROMATASE INHIBITORS IN BREAST CANCER
`
`885
`
`Table 2. Second-Line Therapy With Aromatase Inhibitors
`
`ANA v MA47, 48
`(1 mg)
`
`LET v MA51
`(2.5 mg)
`
`VOR v MA54
`(2.5 mg)
`
`FAD v MA57
`(2 mg)
`
`FAD v MA57
`(2 mg)
`
`EXE v MA55
`(25 mg)
`
`FOR v MA58
`(250 mg IM)
`
`LET v AG50
`(2.5/500 mg)
`
`VOR v AG53
`(2.5/500 mg)
`
`263/253
`12.6/12.2
`
`174/189
`24/16
`
`225/227
`11/8
`
`196/184
`11.3/16.3
`
`152/151
`13.4/11.5
`
`336/403
`15/12.4
`
`91/86
`16.7/16.9
`
`185/178
`19.5/12.3
`
`277/279
`23/18
`
`42.2/40.3
`
`35/32
`
`35.9/35.9
`
`37.4/41.2
`
`37.4/34.6
`
`42.2/38.6
`
`36.3/29.3
`
`47/37
`
`5.6/5.5
`5.1/3.9
`25/22
`2/9
`
`0/8
`
`2.7/3.6
`
`3.9/3.8
`
`5.3/5.8
`
`26/29
`1.3/13.7
`
`27.1/23.1
`
`25.8/27.9
`
`4.7/3.8
`3.8/3.7
`NR/28.4
`2.8/5.8
`
`7/10/2011 12.2/21.2
`19.6/7
`11.7/9.2
`
`11.8/18.8
`14.5/11.4
`
`12.6/5.0
`
`4/3.7
`
`20/32§
`
`15/9†
`3/9
`20/32*
`
`14.2/25.1
`20.4/11
`
`7.7/23.4
`21.9/13
`9.2/4.9
`9.2/3.8
`
`14.5/28.2
`36.2/11.4
`18.4/7.4
`19.7/6
`
`0.3/3.0
`9.2/5.0
`2.8/0.8
`
`27/23
`3/13
`
`8/13
`14/11
`3/5
`2/6
`11/23
`18/23
`10/7
`7/5
`
`3.4/3.2
`3/3
`28/20
`
`7/6
`5.3/4.4
`25.7/21.7
`
`4.9/3.4
`
`10.3/9.6
`3.8/5.6
`
`LET . MA VOR . MA
`
`2.0/0
`EXE . MA‡
`
`VOR . AG
`
`No. of Patients
`Response rate (complete 1
`partial response), %
`Complete response 1 partial
`response 1 stable disease
`. 24 weeks, %
`Median TTP, months
`Median TTF, months
`Median OS, months
`Increased weight/appetite,
`%
`Edema, %
`Hot flashes, %
`Thromboembolic disease, %
`Sweating, %
`Dyspnea, %
`Nausea, %
`Vomiting, %
`Anorexia, %
`Skin rash, %
`Quality of life
`
`NOTE. The two FAD v MA trials were of similar design; significant results are printed bold.
`Abbreviations: ANA, anastrozole; MA, megestrol acetate; LET, letrozole; VOR, vorozole; FAD, fadrozole; EXE, exemestane; AG, aminoglutethimide; NR, not
`reached.
`*More than 3 kg.
`†Moderate and severe.
`‡In general, but not on all subscales.
`§More than 3 kg.
`
`Hamilton and Piccart56 for the trials with anastrozole,
`vorozole, and letrozole. Thus although letrozole and ex-
`emestane seem to have performed particularly well com-
`pared with the other inhibitors in terms of efficacy, further
`studies will be needed to confirm this. For example, a trial
`of letrozole versus anastrozole as second-line therapy after
`tamoxifen is ongoing.
`There are two second-generation inhibitors that although
`not widely used are on the market. Fadrozole, a nonsteroidal
`inhibitor, is currently marketed in Japan. It was also tested
`in second-line as treatment of postmenopausal metastatic
`breast cancer after tamoxifen and showed efficacy and
`toxicity comparable to that of megestrol acetate57 (Table 2).
`The steroidal inhibitor formestane (4-OH-androstenedione)
`showed advantages over megestrol acetate as second-line
`treatment of metastatic breast cancer in terms of efficacy
`and tolerability but is administered intramuscularly, which
`is associated with injection-site reactions58 (Table 2).
`Liarozole, a novel agent with a dual mechanism of action
`viz potent
`inhibition of aromatase and of retinoic acid
`catabolism (a retinoic acid metabolism– blocking agent),
`has been withdrawn from clinical development for reasons
`
`of predominantly retinomimetic toxicities. Nevertheless, in
`phase II studies in postmenopausal patients,
`liarozole
`showed promising activity in both ER-positive disease after
`tamoxifen and in ER-negative breast cancer.59,60
`In summary, the third-generation aromatase inhibitors
`have now become standard second-line treatment of ad-
`vanced breast cancer because of their better toxicity profile
`and improved clinical efficacy compared with conventional
`therapies. Ongoing and future trials will allow comparisons
`in terms of efficacy and tolerability between the different
`agents. In the near future they might also partially supplant
`tamoxifen as first-line treatment as outlined below.
`After tamoxifen as third-line therapy of metastatic dis-
`ease. Exemestane is the only aromatase inhibitor that has
`been tested in phase II trials as third-line therapy, after
`tamoxifen and then megestrol acetate had been given.61,62
`Thirty percent of patients experienced clinical benefit (ie,
`complete response plus partial response plus stable disease
`for $ 67 months) in this trial. Other studies have tested
`aromatase inhibitors as third-line hormonal therapy after
`another inhibitor had been given as second-line treatment
`(Table 3).63-66 Only phase II results are available to date,
`
`Downloaded from ascopubs.org by Univ of Chicago Library on October 30, 2016 from 205.208.122.242
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1074-0005
`
`

`
`886
`
`Reference
`No. of patients
`Previous TAM
`E1, %
`E2, %
`E1-sulphate, %
`Complete 1 partial
`response, %
`Complete response 1 partial
`response 1 stable
`disease, %
`Median TTP, months
`
`GOSS AND STRASSER
`
`Table 3. Aromatase Inhibitors as Third-Line Therapy
`
`VOR After FOR
`
`ANA After FOR
`
`FOR After AG
`
`EXE After AG
`
`EXE After MA
`
`EXE After MA
`
`EXE After Nonsteroidal
`
`64
`21
`100% resistant
`
`63
`9
`Not reported
`247
`230
`270
`
`0
`
`62
`
`65
`112
`98%
`
`21
`
`43
`
`66
`78
`96%
`
`289*/256†
`26
`
`39
`
`4.9
`
`61
`91
`100% resistant
`211
`222
`213
`13
`
`30
`
`2
`
`62
`85
`100% resistant
`
`9
`
`29
`
`16
`
`67
`241
`Not reported
`261*/127†
`251*/113†
`259*/117†
`6.6
`
`24.0
`
`14.7
`
`Abbreviations: TTP, time to progression; nonsteroidal, nonsteroidal aromatase inhibitor.
`*After AG.
`†After other nonsteroidal aromatase inhibitor.
`
`and apart from the trial of formestane given after aminoglu-
`tethimide, most trials tested a third-generation agent (ex-
`emestane, vorozole, and anastrozole) after a second-gener-
`ation inhibitor (either formestane or aminoglutethimide).
`Clinical responses in these trials may be explained by the
`fact that estrogen levels are lowered further by administra-
`tion of a third-generation after a second-generation inhibi-
`tor. For example, this was shown in the study of vorozole
`given for 2 months to patients whose disease was respond-
`ing or stabilized on formestane. Estrogen levels were further
`suppressed by vorozole and returned to pretreatment levels
`once the patients restarted formestane (Table 3).63
`This observation, that clinical remissions can be obtained
`by incremental suppression of estrogen, is being explored
`further in a trial in which exemestane is given at a very low
`dose and after initial response is subsequently increased at
`each point of disease progression.
`Recently, the results of a phase II trial testing exemestane
`at two dose levels (25 mg once daily and 100 mg once daily)
`after a nonsteroidal inhibitor (aminoglutethimide, anastro-
`zole, letrozole, or vorozole) have been published.67 Inter-
`estingly, exemestane showed an overall response rate (com-
`plete response plus partial response plus no change for $ 24
`weeks) of 20.4% in patients who had already received
`another third-generation aromatase inhibitor.
`A response to the androgen analog exemestane after the
`nonsteroidal inhibitors might be explained by the fact that
`exemestane exhibits androgenic effects. These effects, which
`have been seen at doses of 200 mg/d might also exist at a
`lower, and clinically not apparent, level at the 25-mg dose.
`Similar to the second-line trials discussed above, the
`relative benefits of the inhibitors as third-line treatment in
`the phase II trials (as listed in Table 3) cannot be compared,
`because there were significant differences in the trial de-
`
`signs and patient populations. For example, not all trials
`required clinical resistance to the agents given as first- and
`second-line hormonal
`therapy. With the emerging data
`indicating efficacy for the aromatase inhibitors in first-line
`metastatic disease, it is unlikely that randomized phase III
`trials will be conducted in this setting.
`Two strategies of
`As adjuvant therapy after tamoxifen.
`using aromatase inhibitors after tamoxifen are being evalu-
`ated in adjuvant postmenopausal breast cancer trials (Fig 2).
`In the first, the inhibitors are being given as an extension
`after the initial standard 5 years of tamoxifen. The MA.17
`international intergroup trial, initiated by the National Can-
`cer Institute of Canada-Clinical Trials Group in 1998, is
`randomizing patients who are disease-free after 5 years of
`adjuvant tamoxifen to an additional 5 years of letrozole or
`placebo. In a similar design, the National Surgical Adjuvant
`Breast and Bowel Project is currently commencing a trial
`(B-33) of 2 years of exemestane or placebo after a standard
`5 years of adjuvant tamoxifen.
`The second approach to using aromatase inhibitors after
`tamoxifen is the use of both agents in sequence within the
`first 5 postoperative years. In this regard, preliminary results
`of a phase III study comparing 5 years of tamoxifen with 3
`years of tamoxifen followed by 2 years of aminoglutethim-
`ide showed a statistically significant (P 5 .006) survival
`advantage for the sequential arm. However, this included a
`difference in deaths unrelated to cancer and no impact on
`disease recurrence was seen with the aromatase inhibitor.68
`Several
`large ongoing trials are also investigating this
`approach. For example,
`the International Collaboration
`Cancer Group trial is comparing 2 years of exemestane after
`3 years of tamoxifen to a standard 5-year course of
`tamoxifen. Similarly, the Austrian Breast Cancer Study
`Group and the German Adjuvant Breast Cancer Group are
`
`Downloaded from ascopubs.org by Univ of Chicago Library on October 30, 2016 from 205.208.122.242
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1074-0006
`
`

`
`AROMATASE INHIBITORS IN BREAST CANCER
`
`887
`
`in chemoprevention in healthy women. These issues are
`elaborated on below.
`
`As First-Line Endocrine Therapy
`
`Aromatase inhibitors may be superior to tamoxifen as
`first-line hormonal therapy in breast cancer patients and
`even as chemopreventives. Tamoxifen has several well-
`described disadvantages. Adverse symptoms and impair-
`ment of quality of life are not infrequent on treatment and
`serious, albeit rare, side effects including endometrial can-
`cer and thromboembolism occur. As discussed above,
`tamoxifen dependence and estrogen hypersensitivity may
`develop with prolonged therapy.
`Aromatase inhibitors are very well tolerated clinically.
`Although approximately 10% of patients receiving tamox-
`ifen discontinue treatment because of adverse events, less
`than 3% of patients did so in the phase III trials with the
`third-generation inhibitors.47-56 Furthermore, in the MCF-
`7CA nude mouse model,
`the third-generation aromatase
`inhibitors have superior antitumor activity compared with
`tamoxifen.69 To date, two trials in metastatic breast cancer
`comparing a third-generation inhibitor, anastrozole, have
`shown equal response rates and superior toxicity profiles as
`compared with tamoxifen.70 However, in addition to supe-
`rior response rates and improved toxicity profiles,
`two
`factors may influence whether the third-generation inhibi-
`tors will prove to be better first-line therapy than tamoxifen.
`First, superior survival rates among patients given the
`inhibitor compared with tamoxifen will be important. Sec-
`ond, it will be important to determine whether the sequence
`of tamoxifen to aromatase inhibitor is inferior to aromatase
`inhibitor to tamoxifen. In the past, several trials of earlier
`inhibitors such as aminoglutethimide after tamoxifen and
`vice versa showed the sequence of the inhibitor after
`tamoxifen to be a superior strategy.71-73 Additional data
`from recently completed studies of letrozole versus tamox-
`ifen as first-line therapy for metastatic disease and as
`neoadjuvant therapy in postmenopausal receptor-positive
`disease should be available soon. A trial of exemestane
`versus tamoxifen in the first-line treatment of metastatic
`breast cancer is being planned. It will be of interest to
`review the comparable responses and survival of patients in
`these trials. Furthermore, this important question will also
`be addressed in part by some of the ongoing adjuvant trials
`comparing aromatase inhibitors to tamoxifen.
`Although the optimal strategy of