European Journal of Cancer 41 (2005) 34fl356
`
`Review
`
`European
`
`Journal of
`Cancer
`
`www.ejconline.com
`
`Endocrine treatment options for advanced breast cancer — the role of
`fulvestrant
`
`J.F .R. Robertson a”: S.E. Come b, S.E. Jones C, L. Beex d, M. Kaufmann 6, A. Makris f,
`J.W.R. Nortier g, K. Possinger 11, L.-E. Rutqvist 1
`
`a Unit of Surgery, Nottingham City Hospital, Hucknall Road, Nottingham NG5 IPB, UK
`b Beth Israel Deaconess Medical Center, Boston, MA, USA
`0 Charles A. Sammons Cancer Center, Dallas, TX, USA
`d University Medical Centre Nijmegen, Nijmegen, The Netherlands
`e Goethe University, Frankfurt, Germany
`fAcademic Oncology Unit, Mount Vernon Hospital, Middlesex, UK
`g Leiden University Medical Center, Leiden, The Netherlands
`h Humboldt University Berlin, Berlin, Germany
`i Karolinska Institute, Stockholm, Sweden
`
`Received 28 April 2004; accepted 20 July 2004
`Available online 11 November 2004
`
`
`
`Abstract
`
`For many years, tamoxifen has been the ‘gold standard’ amongst anti-oestrogen therapies for breast cancer. However, the
`selective aromatase inhibitors (AIs), anastrozole, letrozole and exemestane, have demonstrated advantages over tamoxifen as
`first-line treatments for advanced disease. Anastrozole is also more effective as an adjuvant treatment in early, operable breast
`cancer and is being increasingly used in the adjuvant setting. Generally, the selective oestrogen receptor modulators (SERMs),
`such as toremifene, droloxifene, idoxifene, raloxifene, and arzoxifene, show minimal activity in tamoxifen-resistant disease and
`show no superiority over tamoxifen as first-line treatments. In addition to these agents, other treatment options for advanced
`disease include high-dose oestrogens and progestins. Response rates for high-dose oestrogens and tamoxifen are similar, but
`the use of oestrogens is limited by their toxicity profile. Consequently,
`there is a need for new endocrine treatment options
`for breast cancer, particularly for use in disease that is resistant to tamoxifen or AIs. Fulvestrant ('_‘Faslodex’) is a new type
`of steroidal oestrogen receptor (ER) antagonist that downregulates cellular levels of the ER and progesterone receptor and
`has no agonist activity. This paper reviews the key efficacy and tolerability data for fulvestrant in postmenopausal women in
`the context of other endocrine therapies and explores the potential role of fulvestrant within the sequencing of endocrine ther-
`apies for advanced breast cancer.
`© 2004 Published by Elsevier Ltd.
`
`Keywords: Advanced breast cancer; Treatment; Fulvestrant; Faslodex; Postmenopausal
`
`1. Introduction
`
`address:
`
`Fulvestrant (‘Faslodex’) is a new type of endocrine
`agent, an oestrogen receptor (ER) antagonist that has
`—t
`.
`,
`Correspondmg author. Tel.: +44 115 969 1169x46859; fax: +44
`115 840 7618
`E—mail
`Robertson).
`
`john.robertson@nottingham.ac.uk
`
`(J.F.R.
`
`no agonist effects. It downregulates cellular levels of
`the ER, resulting in the decreased expression of the pro-
`gesterone receptor (PgR). This paper reviews key effi-
`cacy and tolerability data for fulvestrant in the context
`of other endocrine therapies and explores the potential
`.
`.
`.
`role of fulvestrant Within the sequence of endocrine
`agents used for treating pOStmenopausal women With
`advanced breast cancer.
`
`0959-8049/$ - see front matter © 2004 Published by Elsevier Ltd.
`doi:10.1016/j.ejca.2004.07.035
`
`AstraZeneca Exhibit 2073 p. 1
`InnoPharma Licensing LLC v. AstraZeneca AB IPR2017-00905
`
`

`

`J.F.R Robertson 6! a]. /European Journal of Cancer 41 (2005) 3467356
`
`347
`
`2. SERMs, oestrogens and ER downregulators
`
`3. Activity versus tamoxifen and in tamoxifen-resistant
`disease
`
`triphenylethylene
`a non-steroidal
`is
`Tamoxifen
`agent that has been the ‘gold standard’ selective oes-
`trogen receptor modulator
`(SERM) amongst
`anti—
`oestrogen therapies since the 1970s. Tamoxifen is an
`oestrogen antagonist
`in breast
`tissue, but acts as
`an oestrogen agonist in the bones and endometrium.
`This spares bone mineral density and serum choles-
`terol levels from the full effects of oestrogen depriva-
`tion [1,2], but is also associated with undesirable side
`effects such as an increased risk of endometrial cancer
`
`and thromboembolic events [3]. The selective aroma-
`tase inhibitors (AIS), anastrozole,
`[4,5]
`letrozole [6,7]
`and exemestane [8], have since been shown to have
`advantages over tamoxifen as first-line treatments for
`advanced disease, but a review of these data is beyond
`the scope of this paper.
`Several new anti-oestrogens have been developed
`since tamoxifen, some with similar mechanisms of ac-
`tion to tamoxifen and some that are very different.
`First-generation SERMs, such as toremifene, droloxif-
`ene and idoxifene, are tamoxifen analogues based on
`the non-steroidal
`triphenylethylene
`structure. The
`structurally
`distinct
`second-
`and third-generation
`SERMs (raloxifene, arzoxifene, EM-800 and ERA-
`923) are also non-steroidal, but are ‘fixed-ring’ benzo-
`thiophene derivatives, yet appear to retain some oes-
`trogen
`agonist
`activity.
`In
`contrast,
`fulvestrant,
`which has a steroidal structure closely resembling oest-
`radiol,
`is a new ER antagonist that has no agonist
`activity. Fig.
`1 shows the chemical structures of oest—
`radiol, fulvestrant, tamoxifen and raloxifene. Anti—oes—
`trogens are generally compared on the basis of their
`activity in tamoxifen-resistant disease,
`their ability to
`delay the development of resistance and their side-ef-
`fect profiles.
`
`Oestradiol
`
`Fulvestrant
`
`OH
`
`0H
`
`HO
`
`HO
`
`._J
`‘(CHZ)QSO(CH2)3CFZCF3
`
`Tamoxifen
`
`NMe
`
`O/\/
`
`Raloxifana
`
`
`2
`
`Fig. 1. Chemical structures of oestradiol, fulvestrant, tamoxifen and
`raloxifene.
`
`3.1. SERMS
`
`Several studies have confirmed that
`
`toremifene is
`
`cross-resistant with tamoxifen in advanced disease [97
`11]. Both these agents display similar efficacy and toler-
`ability in the advanced [12,13] and adjuvant settings [14].
`Efficacy results for the structurally related, droloxifene
`[15717] and idoxifene [18,19], have both been disap-
`pointing when either compared directly with tamoxifen
`or when used in tamoxifen-resistant disease; the devel-
`opment of these agents has now ceased.
`The second-generation, ‘fixed ring’ SERM arzoxifene
`has also shown poor efficacy in tamoxifen-resistant dis-
`ease [20]. Furthermore, a randomised Phase III trial of
`arzoxifene versus tamoxifen was terminated early be-
`cause of a lack of efficacy (Buzdar A, data not shown).
`Raloxifene also shows low efficacy in tamoxifen-resist-
`ant disease [21], but has shown promise in breast cancer
`prevention [22,23] and is currently being tested in this
`setting in the STAR (Study of Tamoxifen and Raloxif—
`ene) trial [24]. ERA-923 is a second-generation SERM
`that is currently in development for the treatment of
`tamoxifen-refractory metastatic breast cancer, but no
`efficacy data for this agent are available to date. In sum-
`mary, the SERMs as a group have thus far shown no
`superiority over tamoxifen as first-line advanced breast
`cancer treatments and minimal activity in tamoxifen-re-
`sistant disease.
`
`3.2. High—dose oestrogens
`
`Prior to the introduction of tamoxifen, high-dose
`oestrogens i such as diethylstilboestrol (DES) or ethi—
`nyl oestradiol 7 were generally considered the endo-
`crine treatment of choice for postmenopausal women
`with breast cancer [25]. Subsequently, the use of oes-
`trogens declined, but
`recent
`trial data have shown
`these drugs to have similar efficacy to tamoxifen [26]
`and to produce responses, even in those who have re-
`ceived extensive prior endocrine therapy [27]. However,
`the use of these agents is limited by their toxicity
`profile.
`
`4. Fulvestrant — a novel oestrogen antagonist that
`downregulates cellular levels of ER
`
`4. I . Biological eflects
`
`Fulvestrant blocks the trophic actions of oestrogen
`without exerting any partial agonist effects. Fulvestrant
`entered clinical development after preclinical studies
`suggested it was active in tamoxifen-resistant breast
`
`AstraZeneca Exhibit 2073 p. 2
`
`

`

`348
`
`J.F.R. Robertson et a1. / European Journal of Cancer 41 (2005 3467356
`
`PgR levels with fulvestrant treatment. Reductions in ER
`levels were statistically significant for all fulvestrant
`doses compared with placebo and for the 250 mg dose
`compared with tamoxifen. There were statistically sig-
`nificant reductions in PgR levels with fulvestrant 125
`and 250 mg compared with placebo. In contrast, tam-
`oxifen treatment produced significant increases in PgR
`levels compared with placebo, probably as a result of
`its oestrogen agonist activity. All doses of fulvestrant
`significantly reduced the Ki67 labelling index, but there
`were no significant differences compared with tamoxifen
`[36].
`
`4.2. Clinical eflficacy
`
`4.2.1. Fulvestrant versus anastrozole
`
`Two large Phase III trials (Trial 0021: North Ameri-
`can; Trial 0020: Rest of World [Europe, South Africa,
`Australia]) have compared the efficacy and tolerability
`of fulvestrant with anastrozole, in postmenopausal wo-
`men with advanced breast cancer who had progressed
`on prior endocrine treatment (mainly tamoxifen). Pa-
`tients were randomised to receive either fulvestrant 250
`
`mg, by monthly i.m. injection or a daily oral dose of
`anastrozole 1 mg and continued treatment until disease
`progression or withdrawal.
`In the North American trial, 400 patients received
`double-blind, randomised treatment with either fulves-
`trant (as 2X25 ml
`i.m.
`injections) (n=206) or oral
`anastrozole (n = 194) and were followed for a median
`of 16.8 months. Fulvestrant was found to be as effective
`
`as anastrozole in terms of time to progression (TTP)
`(Hazard Ratio (HR): 0.92; 95% Confidence Interval
`(CI) 07471.14; P = 0.43); median TTP was 5.4 months
`with fulvestrant and 3.4 months with anastrozole.
`
`Objective response (OR) rates were 17.5% for both
`treatments and clinical benefit
`(complete response
`(CR) + partial
`response
`(PR) + stable
`disease
`(SD) 2 24 weeks; CB) rates were 42.2% and 36.1% for
`fulvestrant and anastrozole, respectively. Median dura-
`tion of response (DOR; from randomisation to progres-
`sion) was 19.0 months for fulvestrant compared with
`10.8 months for anastrozole. An analysis using all rand-
`omised patients, defined for responders as the time from
`onset of response to disease progression and for non-re-
`sponders as zero, showed that mean DOR was signifi-
`cantly
`greater
`for
`fulvestrant
`compared with
`anastrozole;
`the ratio of average response durations
`being 1.35 (95% CI 11071.67; P < 0.01) [38].
`In the Rest of World (open) trial, 451 patients were
`randomised to receive either fulvestrant as a single 5
`ml
`i.m.
`injection (n = 222) or
`anastrozole orally
`(n = 229) and were followed for a median of 14.4
`months. Again, fulvestrant was shown to be at least as
`effective as anastrozole in terms of TTP (HR: 0.98;
`95% CI 08071.21; P = 0.84); median TTP was 5.5
`
`AstraZeneca Exhibit 2073 p. 3
`
`cancer [28730]. Fulvestrant acts by competing with oest-
`radiol for binding to the ER and has a higher affinity for
`the ER than tamoxifen [31]. Fulvestrant downregulates
`expression of ER and decreased activity of the ER path-
`way results in reduced expression of the PgR and re-
`duced proliferative and cell turnover indices both in
`vitro [32735] and in the clinical setting [36,37].
`In the presence of fulvestrant, ER is rapidly depleted,
`producing a loss of functional response to oestrogens
`after relatively short periods of time in in vitro studies.
`This is in contrast with the increases in ER levels seen
`
`on either oestrogen withdrawal or tamoxifen treatment
`[32]. The effects of three different fulvestrant doses and
`one dose of tamoxifen on cellular ER, PgR and Ki67
`levels were investigated in a study of postmenopausal
`women with primary breast cancer. Patients received
`either a single intramuscular (i.m.) injection of fulves-
`trant 50, 125 or 250 mg, or oral tamoxifen 20 mg, daily
`for 1421 days, prior to surgery of curative intent. Fig. 2
`demonstrates the dose-dependent reductions in ER and
`
`(a)
`
`P=o.ooo1
`l—PEO'UIHJE—l
`
`Mean:t‘lSEM
`
`
`
`
`
`Placebo
`(n:29)
`
`250 mg
`125 mg
`50 mg
`Fulvestrant Fulvestrant Fulvestrant
`{n=31)
`(n=32)
`(n=32)
`Overall treatment effect P=0.0003
`
`Tamoxifen
`(11:25)
`
`P=0.000‘l
`P=o.uom
`P410001
`'—__l
`
`P=0.000Z
`
`l
`
`P=0.UO3
`
`NS
`
`‘
`
`
`
`[3
`__
`
`.—
`I
`
`
`W
`
`
`
`;
`
`(b)
`
`‘IOO
`
`50
`
`50
`
`4o
`
`20
`
`0
`
`2
`U)
`Lu
`+1
`
`:3
`
`m E
`
`Placebo
`(n:28)
`
`50 mg
`125 mg
`250 mg
`FulvestrantFulvestrant Fulvestrant
`(n=29)
`(n=29)
`(n=29)
`Overall treatment effect P=0.0001
`
`Tamoxifen
`(n:21]
`
`(a) oestrogen
`Fig. 2. Post-treatment mean H-scores for cellular:
`levels. Figure
`receptor (ER) and (b) progesterone receptor (PgR)
`reproduced with the permission of Cancer Research [36]. SEM,
`standard error of the mean; NS, non-significant.
`
`

`

`J.F.R Robertson er a]. /European Journal of Cancer 41 (2005) 3467356
`
`349
`
`Table 1
`Response to subsequent therapy in patients who derived CB from
`fulvestrant
`
`
`
`PD Total
`
`Number of patients
`CR PR SD 2 24 weeks
`Patients who derived CB from first-line fulvestrant
`Anastrozole
`l
`0
`8
`Letrozole
`0
`l
`0
`Fadrozole
`0
`0
`l
`Tamoxifen
`0
`l
`7
`
`Megestrol acetate
`Medroxyprogesterone
`acetate
`
`0
`0
`
`0
`0
`
`l
`0
`
`7
`4
`0
`2
`
`0
`2
`
`16
`5
`l
`10
`
`l
`2
`
`
`
`
`
`
`
`
`
`
`
`Patients who derived CB from second-line fulvestrant
`37
`23
`Anastrozole
`0
`l
`13
`8
`3
`Letrozole
`0
`2
`3
`l
`l
`Formestane
`0
`0
`0
`5l0Megestrol acetate 8 2
`Table adapted from [44] and [42], with the permission of Breast Cancer
`Research and Treatment.
`CR, complete response; PR, partial response; SD, stable disease; PD,
`disease progression; CB. Clinical benefit.
`
`months for fulvestrant and 5.1 months for anastrozole.
`OR rates were 20.7% for fulvestrant and 15.7% for anas-
`
`trozole (odds ratio: 1.38; 95% CI 0.844229; P = 0.20).
`CB rates were 44.6% for fulvestrant and 45.0% for anas-
`trozole and the median DOR was 15.0 months and 14.5
`
`months for fulvestrant and anastrozole, respectively. In
`addition, mean DOR using all randomised patients was
`significantly greater for fulvestrant compared with anas-
`trozole, the ratio of average response durations being
`1.27 (95% CI 1.0541.55;P = 0.01) [38].
`The two Phase III trials were prospectively designed
`to be evaluated both individually and using combined
`data. A combined analysis of all patients included in
`both second-line Phase III trials demonstrated a signifi-
`cant 30% increase in mean DOR in patients treated with
`fulvestrant (ratio of average response durations: 1.30;
`95% CI 1.134.50; P < 0.01; Fig. 3) [40]. In addition
`to confirming the efi‘lcacy of fulvestrant that was ob-
`served in the individual trials, retrospective analyses of
`these combined data also showed fulvestrant had similar
`
`efficacy (in terms of OR rate, CB rate and DOR) to
`anastrozole in the subgroup of patients with visceral
`metastases. The median DOR in patients with visceral
`metastases was 17.5 months in the fulvestrant group
`compared with 11.7 months in the anastrozole group.
`Notably, in the subgroup of patients with visceral meta-
`stases only, seven of 69 (10%) fulvestrant-treated pa-
`tients achieved a CR, compared with one of 86 (1%)
`anastrozole-treated patients [41].
`The data from these two trials reiterate that fulves-
`
`trant is a novel agent with levels of activity in tamoxi-
`fen-resistant disease that distinguish it from SERMs
`and other anti—oestrogens. Furthermore,
`fulvestrant
`was at least as effective as anastrozole in this setting.
`In addition, retrospective analysis evaluating combined
`questionnaire data from the two trials showed that pa-
`tients can retain sensitivity to other endocrine agents
`(anastrozole,
`letrozole, and megestrol acetate) after
`receiving second-line fulvestrant (Table 1) [42].
`
`0.35
`
`
`
`Fulvestrant 250 mg
`----- -- Anastr'ozole1 mg
`
`Duration of response (months)
`
`Fig. 3. KaplaniMeier estimates for duration of response from onset of
`response to disease progression (combined analysis of all randomised
`patients included in Phase III trials). Copyright © 2003 American
`Cancer Society. Reprinted by permission of Wiley - Liss, Inc., a
`subsidrary of John Wiley & Sons, Inc. [40].
`
`4.2.2. Fulvestrant versus tamoxifen
`Fulvestrant and tamoxifen have been compared as
`first-line treatments in a trial including postmenopausal
`women with advanced breast cancer. Approximately,
`20425% of patients in this trial had previously received
`adjuvant tamoxifen, but no patients received prior endo—
`crine therapy for advanced disease. In this study, the
`median TTP was 6.8 months in the fulvestrant group
`and 8.3 months in the tamoxifen group. The between-
`treatment difference was non-significant
`(HR: 1.18;
`95% CI 09841.44; P= 0.088), but the upper limit of
`the 95% CI (1.44) did not satisfy the pre-defined crite-
`rion for non-inferiority ($1.25) of fulvestrant compared
`with tamoxifen. OR rates for fulvestrant and tamoxifen
`
`were similar (31.6% versus 33.9%; P = 0.45), but more
`tamoxifen-treated patients overall achieved CB (62.0%
`versus 54.3%; P = 0.03) [43]. Median DOR (from rand-
`omisation to progression) was 17.3 and 19.8 months
`for fulvestrant and tamoxifen, respectively [43]. In the
`prospectively defined subgroup of patients with ER—pos-
`itive and/or PgR-positive tumours, median TTP was
`similar for the fulvestrant and tamoxifen treatment
`
`groups (8.2 months versus 8.3 months; HR: 1.10; 95%
`CI 0894.36; P = 0.39) [43].
`Subsequent exploratory analyses of response by hor-
`mone receptor status showed that in the subgroup of pa-
`tients with tumours expressing both ER and PgR 44.3%
`of fulvestrant-treated patients and 29.8% of patients
`treated with tamoxifen experienced an OR (P 20.02)
`[43]. However, the authors note that these data were ret-
`rospectively derived and therefore should be interpreted
`with caution in terms of their clinical significance. Fur-
`ther confirmatory data are required. In addition, pa-
`tients who responded to first-line treatment with
`
`AstraZeneca Exhibit 2073 p. 4
`
`0.30
`0.25
`0.20
`0.15
`0.10
`
`0.05
`0.00
`
`
`
`Proportionresponding
`
`

`

`350
`
`J.F.R. Robertson et al. / European Journal of Cancer 41 (2005) 3464356
`
`fulvestrant may retain sensitivity to subsequent endo-
`crine therapy with a variety of agents, including anas-
`trozole, letrozole, fadrozole, tamoxifen, and megestrol
`acetate (Table l) [44]. This is similar to the findings
`noted above where tumours appeared to retain sensitiv-
`ity to other endocrine agents following second-line treat-
`ment with fulvestrant [42]. It therefore appears that the
`use of fulvestrant does not per se lead to end-stage hor-
`mone insensitivity.
`
`4.3. Tolerability
`
`In the second-line trials, both fulvestrant and anas-
`trozole were well tolerated, with few patients in either
`group withdrawing due to treatment-related adverse
`events (0.9% and 1.2% of the fulvestrant- and anastroz-
`ole-treated patients, respectively). Overall, the incidence
`and severity of events (generally mild to moderate)
`were similar between groups; the most common events
`in both groups included hot flushes, nausea, asthenia,
`pain, headache and pharyngitis. The incidence of
`events considered important with endocrine therapy
`such as weight gain, thromboembolic events and vagi-
`nitis was low for both fulvestrant and anastrozole
`
`[38,39].
`The use of placebo injections in the North American
`trial indicated that fulvestrant was well tolerated locally
`and that injection-site reactions were related to the injec-
`tion itself, as 27% of patients receiving fulvestrant com-
`pared with 23% of those receiving placebo reported
`injection-site reactions
`[38]. Overall, 86 fulvestrant
`courses (4.6%) of the total 1879 and 71 placebo courses
`(4.4%) of the total 1624 resulted in an injection—site reac—
`tion. This is supported by the clinical experience of phy-
`sicians administering the 2 x 2.5 ml fulvestrant regimen
`in the US. Here, nursing guidelines have previously sug-
`gested that, for adults, i.m. injections into large muscles
`such as the gluteus medias, should not usually exceed 4
`ml [45], therefore in the trial it was decided to use 2 X 2.5
`ml i.m. injections, one into each buttock, rather than a
`single 5 ml injection [38]. The pharmacokinetics of these
`two regimens have previously been shown to be similar
`[46]. Moreover, since the 2.5 ml injections were so well
`tolerated, most US institutions now prefer to administer
`fulvestrant as a single 5 ml i.m. injection, thereby reduc-
`ing the number of injections given to the patient (Astra-
`Zeneca, data on file).
`tamoxifen comparative
`In the fulvestrant versus
`study, both treatments were well tolerated with no sta-
`tistically significant differences in the incidence of pro-
`spectively defined adverse events of gastrointestinal
`disturbances, hot flushes, vaginitis and thromboembolic
`disease. However, the incidence of hot flushes was lower
`in the fulvestrant group than in the tamoxifen group and
`the difference approached statistical significance (17.7%
`versus 24.7%; P = 0.0501) [43].
`
`5. Fulvestrant — the US experience
`
`In April 2002, the US Food and Drug Administra-
`tion (FDA) granted approval for fulvestrant to be used
`for the treatment of hormone receptor—positive meta-
`static breast cancer in postmenopausal women with dis-
`ease progression following anti-oestrogen therapy. The
`use of fulvestrant as second-line therapy is increasing,
`although the drug continues to be used in patients
`who have progressed on both tamoxifen and an AI.
`Data from the two Phase III studies of fulvestrant versus
`
`anastrozole suggest that fulvestrant is as effective as an
`AI in the second-line setting and that DOR may, in fact,
`be longer with fulvestrant treatment [38,39].
`In the US, there appears to be a general underutilisa-
`tion of hormonal therapy and a lack of differentiation
`between fulvestrant and other hormonal agents such as
`anastrozole. At several oncology meetings in the US,
`surveys of treatment practice amongst US oncologists
`suggest that US physicians often favour chemotherapy
`in situations where European clinicians prefer further
`endocrine treatment. Furthermore, US physicians may
`utilise endocrine treatment for a shorter duration and
`
`switch to chemotherapy earlier than their European
`counterparts (Jones SE, data not shown).
`The personal experiences of the US physicians in-
`volved in the US Phase III fulvestrant versus anastrozole
`
`trial are in agreement with the formal DOR analysis,
`which suggest that, on average, fulvestrant-treated pa-
`tients respond for approximately 30% longer than those
`treated with anastrozole [40]. One investigator had 27 pa-
`tients included in the Phase III US trial, of these, five
`have had responses of >3 years (two for >4 years), four
`of these patients have now relapsed and have been un-
`blinded and all had been receiving fulvestrant and one
`is currently continuing double-blind treatment (Jones
`SE, data not shown). A second investigator had 16 pa-
`tients entered in the same trial; four of these patients
`(25%) had CB for 20444 months three of whom were
`found to be receiving fulvestrant after unblinding (Come
`SE, Personal Communication) (Table 2). This empha-
`sises the fact that there appears to be a population of pa-
`tients who have an extremely long DOR with fulvestrant.
`Furthermore, personal experiences from these physicians
`have shown that parenteral dosing can also be beneficial
`over oral dosing for some patients, particularly those
`demonstrating poor compliance with oral therapies.
`
`6. Sequencing of endocrine treatments
`
`Because of the toxicity associated with chemotherapy,
`it would be advantageous in appropriate patients to ex-
`tend the endocrine treatment window,
`thus deferring
`the initiation of more toxic treatments that are associated
`with acute and more severe side effects. Endocrine treat-
`
`AstraZeneca Exhibit 2073 p. 5
`
`

`

`J.F.R Robertson ez al. /European Journal of Cancer 41 (2005) 3467356
`
`351
`
`Table 2
`Case studies from the US phase III fulvestrant versus anastrozole trial (Come SE, Personal Communication)
`Efficacy results
`Patient characteristics
`Case study I
`o 55 years old
`0 ER-positive infiltrating lobular carcinoma
`o Developed metastases to the colon during adjuvant tamoxifen treatment
`
`0 SD of 7 months duration on anastrozole (after code-break)
`
`0 SD of 20 months duration on fulvestrant
`
`Case study 2
`o 59 years old
`o ER-positive/PgR-positive tumour
`o Metastases to liver, bone, skin, and lymph nodes
`
`0 SD of 39 months duration on fulvestrant
`0 SD of 7 months duration on anastrozole (after code-break)
`
`Case study 3
`o 64 years old
`o ER-positive tumour
`
`o CR of 44 months on fulvestranta
`0 Still receiving fulvestrant outside of trial setting and still in CR
`(currently of 55 months duration)
`
`0 Skin metastases (following first-line tamoxifen treatment)
`
`ER, oestrogen receptor; PgR, progesterone receptor; SD, stable disease; CR, complete response.
`a Patient withdrew from the trial whilst experiencing a CR (at that time of 44 months duration) to receive off-study treatment with fulvestrant
`closer to home.
`
`ment should generally continue as long as the patient re—
`mains hormone sensitive (i.e., achieving CB with hor-
`mone
`treatment)
`and when the patient becomes
`hormone resistant, chemotherapy treatment should then
`be initiated. For the past 20 years, the most commonly
`used first-line or adjuvant endocrine treatment for ad-
`vanced breast cancer has been tamoxifen. However, the
`
`non-steroidal AIs are now used routinely in the advanced
`setting and are also starting to be used as adjuvant treat-
`ment following the recently reported results from the
`‘Arimidex’, Tamoxifen, Alone or
`in Combination
`(ATAC) trial [47].
`Two schema of treatment options following adjuvant
`or first-line tamoxifen treatment (Fig. 4(a)) or adjuvant
`
`Adjuvant or ii-Iine TAM
`
`(b)
`
`I
`AdJuvant ANA
`
`I
`>12 months
`disease-free
`
`I
`Disease
`recurrence
`on ANA
`
`I
`>12 months
`disease-free
`
`(a)
`after TAM I
`after ANA
`
`Disease
`recurrence
`on TAM
`
`I
`iANA/LETl
`
`I
`FUL
`I
`EXE ANA/LET
`I
`I
`FUL
`EXE
`
`I
`MA
`
`I
`MA
`
`FUL
`I
`EXE
`
`I
`MA
`
`TAM
`I
`FUL":
`
`=
`
`V
`ANA/LET
`
`TAM
`I
`FUL
`I
`
`MA
`
`I
`EXE
`
`I
`
`EXE
`
`I
`FUL
`
`l
`I
`EXE
`I
`I
`
`V
`MA
`
`I
`v
`I
`II
`r"TAM'-'-l
`ANA/LET
`v
`I
`.
`I
`:"ANA/LET“:
`FUL
`I
`I
`I v
`v
`EXE FUL
`EXE
`I
`I
`I
`I
`I v
`v
`MA EXE
`FUL
`I
`I
`I
`I
`
`I I
`
`I
`
`FUL
`
`=
`
`V
`ANA/LET
`
`l
`I
`EXE
`I
`I
`
`V
`MA
`
`V
`MA
`
`V
`MA
`
`TAM. tamoxifen; ANA, anastrozole; LET. letrozole; FUL, fulvestrant; EXE, exemestane;
`MA, megestrol acetate
`
`—> Treatment option (no re-challenge)
`n} Treatment option (with re-chailenge)
`
`Fig. 4. Treatment options flow charts for patients previously treated with: (a) adjuvant or first-line tamoxifen; (b) adjuvant anastrozole.
`
`AstraZeneca Exhibit 2073 p. 6
`
`

`

`352
`
`J.F.R. Robertson et al. / European Journal of Cancer 41 (2005) 3467356
`
`anastrozole treatment (Fig. 4(b)) are proposed. These
`suggestions have been based on data from randomised
`clinical
`trials where available, or otherwise, on non-
`randomised data or the authors’ personal experiences.
`
`6.]. Treatment options following adjuvant or first—line
`tamoxifen
`
`If a patient experiences disease recurrence on adju-
`vant or progressive disease on first-line tamoxifen treat-
`ment (Fig. 4(a)), there is sufficient evidence to show that
`fulvestrant is comparable to anastrozole in these pa-
`tients [38,39], and the longer DOR with fulvestrant
`treatment may suggest an advantage for fulvestrant in
`this setting [40]. At the current time, despite these data,
`the wealth of clinical experience with AIs may mean that
`these agents could be the most comfortable treatment
`choice for many physicians.
`There is evidence to suggest that patients receiving
`fulvestrant may retain sensitivity to tamoxifen and
`non-steroidal AIs [42,44]. Fulvestrant or exemestane
`(a steroidal AI) could be considered following sec-
`ond-line treatment with a non-steroidal AI. Preliminary
`data from a small Phase H study suggests that fulves-
`trant is effective in this setting [48], although more data
`are available to support the reverse sequence (fulves-
`trant followed by an AI). High-dose oestrogens could
`also be used following AI failure [27]. However,
`the
`toxicities of this type of treatment may limit the use
`of this option. Data from a large Phase III trial
`is
`needed to clarify the optimal choice of sequence at this
`point.
`Exemestane has shown activity in women with ad—
`vanced breast cancer previously treated with non-ste-
`roidal AIs
`such as aminoglutethimide, anastrozole
`and letrozole [49]. Exemestane has shown superior effi-
`cacy and tolerability over megestrol acetate, after tam-
`oxifen
`failure,
`in
`postmenopausal women with
`advanced breast cancer [50], and therefore this agent
`is now chosen ahead of megestrol acetate in most situ-
`ations. Further data regarding the sequencing of AIs
`was presented in 2002 at the annual meeting of the
`American Society of Clinical Oncology. Bertelli and
`colleagues [51] investigated whether patients who had
`previously received exemestane could still benefit from
`treatment with anastrozole or letrozole after exemes-
`
`tane failure and vice versa. One PR, two SD and one
`disease progression (PD) were observed amongst the
`first
`five patients receiving non-steroidal AIs after
`exemestane. Responses to exemestane after treatment
`with non-steroidal AIs were similar to those observed
`
`in previous studies, being 3 PR, 3 SD, and 4 PD in
`the first 17 evaluable patients. Although preliminary,
`these data suggest that there is some evidence for a
`lack of cross-resistance between steroidal and non-ste-
`
`roidal AIs [51]. The Evaluation of Faslodex (fulves-
`
`trant) versus Exemestane Clinical Trial (EFECT) will
`compare the efficacy of fulvestrant and exemestane in
`postmenopausal women with advanced breast cancer
`who have progressed after prior non-steroidal AI treat-
`ment. Data from this trial will help to address this part
`of the algorithm.
`If a patient has a disease-free interval > 12 months fol-
`lowing adjuvant or first-line tamoxifen, the choice of
`subsequent treatment is slightly different. With no evi-
`dence indicating superiority of fulvestrant over tamoxi-
`fen in the first-line
`setting,
`a more
`appropriate
`treatment choice may be a non-steroidal A1 [447]. A sec-
`ond possibility, after a disease-free interval of >12
`months would be to rechallenge with tamoxifen and fol-
`low a similar schedule to that shown previously. How-
`ever, a patient may be uncomfortable receiving the
`same treatment and may want to receive a new agent,
`in which case an AI would be recommended.
`
`6.2. Treatment options following adjuvant anastrozole
`
`Anastrozole is the only non-steroidal A1 to have pro-
`ven efficacy for adjuvant
`treatment and provides an
`alternative to tamoxifen in this setting. Treatment op-
`tions following the use of adjuvant anastrozole are pre-
`sented in Fig. 4(b). There is a lack of randomised
`controlled trial data to support an optimal sequence fol-
`lowing either failure on adjuvant anastrozole or a dis-
`ease-free interval of >12 months. However,
`it would
`seem sensible in both situations to try an agent with a
`different mechanism of action such as an anti-oestrogen,
`although following a disease—free
`interval of >12
`months, rechallenge with anastrozole may also be a pos—
`sibility. However, these authors suggest that fulvestrant
`or tamoxifen would both be the valid choices in this set-
`
`ting, but because of the wealth of clinical experience
`with tamoxifen over the last 25 years, they would rec-
`ommend tamoxifen after adjuvant anastrozole treat-
`ment. A further
`reason for
`the use of tamoxifen
`
`followed by fulvestrant rather than the reverse sequence
`is that data from the two large Phase III fulvestrant ver-
`sus anastrozole trials show that this particular sequence
`works well [38,39]. While the reverse sequence has also
`been shown to be effective the volume of data is less
`extensive.
`
`The sequence of treatment choices following fulves-
`trant are similar to those presented earlier; the only dif-
`ference is that following disease recurrence on adjuvant
`anastrozole, there is a rationale for the use of meges-
`trol acetate ahead of exemestane, due to its different
`mechanism of action. Another possibility is the use of
`high-dose oestrogens [27]. Overall however, there is a
`lack of randomised trial data to clearly define an opti-
`mum sequence of endocrine therapies and until this be-
`comes available, clinical experience should shape future
`use.
`
`AstraZeneca Exhibit 2073 p. 7
`
`

`

`J.F.R Robertson et al. /European Journal of Cancer 41 (2005) 3467356
`
`353
`
`loading-dose regimen may allow steady-state levels of
`fulvestrant to be achieved more rapidly. Such an ap-
`proach may not impact on the long-term efficacy of
`the drug, but may allow early responses to be identified.
`It is possible to model the effects of the addition of a
`loading regimen on the attainment of steady-state ful-
`vestrant levels (Fig. 5). Here, an initial dose of 500 mg
`fulvestrant is given on day 0, followed by 250 mg fulves-
`trant on day 14. This is followed 14 days later by the
`standard fulvestr