Review
`
`@ CrossMa.rk
`A Good Drug Made Better: The Fulvestrant
`Dose—Response Story
`
`John F.R. Robertson,1 Justin Lindernann,2 Sally Garnett,2 Elizabeth Anderson,3
`Robert I. Nicholson,4 Irene Kuter,5 Julia M.W. Gee4
`
`Abstract
`Sequential use of endocrine therapies remains the cornerstone of treatment for hormone receptor—positive advanced
`breast cancer, before the use of cytotoxic chemotherapy for unresponsive disease. Fulvestrant is an estrogen receptor
`(ER) antagonist approved for the treatment of postmenopausal women with ER+ advanced breast cancer after failure
`of prior antiestrogen therapy. Initially approved at a monthly dose of 250 mg, the recommended fulvestrant dose was
`revised to 500 mg (500 mg/mo plus 500 mg on day 14 of month 1) after demonstration of improved progression-free
`survival versus fulvestrant 250 mg. We have reviewed the dose-dependent effects of fulvestrant, both from a retro-
`spective combined analysis of dose-dependent reduction of tumor biomarkers in the presurgical setting (3 previously
`reported studies: Study 18, Neoadjuvant Endocrine Therapy for Women with Estrogen-Sensitive Tumors, and Trial 57)
`and from a review of clinical studies for advanced breast cancer in postmenopausal women. Analysis of presurgical
`data revealed a consistent dose-dependent effect for fulvestrant on tumor biomarkers, with increasing fulvestrant
`dose resulting in greater reductions in ER, progesterone receptor, and Ki67 labeling index. The dose-dependent
`biological effect corresponds with the dose-dependent clinical efficacy observed in the treatment of advanced
`breast cancer after failure of prior antiestrogen therapy. Although it remains to be determined in a phase III trial, cross-
`trial comparisons suggest a dose-dependent relationship for fulvestrant as first-line treatment for advanced breast
`cancer. Overall, biological and clinical data demonstrate a strong dose-dependent relationship for fulvestrant, sup-
`porting the efficacy benefit seen with fulvestrant 500 mg over the 250 mg dose.
`
`Clinical Breast Cancer, Vol. 14, No. 6, 381-9 © 2014 Elsevier Inc. All rights reserved.
`Keywords: Advanced breast cancer, Endocrine therapy, Estrogen receptor, Postmenopausal, Tumor biomarkers
`
`Introduction
`
`Endocrine therapies provide effective and well—tolerated treat-
`ments for postmenopausal women with hormone receptor—positive
`breast cancer (estrogen receptor—positive [ER+] and/or progester-
`one receptor—positive [PgR+]), both in the adjuvant settingl and for
`the treatment of advanced disease?
`
`Aromatase inhibitors (AIS), which block production of estrogen
`through their
`interaction with the estrogen—producing enzyme
`
`‘Graduate Entry Medicine and Health School (GEMS), University of Nottingham,
`Derby, UK
`2AstraZeneca, Alderley Park, Macclesfield, UK
`5Forn1erly Astiazeneca, Alderley Park, Macclesfield, UK
`4Breast Cancer Molecular Pharmacology Group, School of Pharmacy and
`Pharmaceutical Sciences, Cardiff University, Cardiff, UK
`5Massachusetts General Hospital, Boston, MA
`Submitted: Mar 28, 2014; Revised: Jun 10, 2014; Accepted: Jun 17, 2014; Epub:
`Jun 24, 2014
`
`Address for correspondence: John F. R Robertson, MD, Division of Medical
`Sciences Sc Graduate Entry Medicine, School of Medicine, Faculty of Medicine Sc
`Health Sciences, University of Nottingham, Royal Derby Hospital Centre, Derby
`DE22 SDT, UK
`Fax: +44 (0)1332 724880; e-mail contact: john.robertson@nottingharn.ac.uk
`
`aromatase, have demonstrated increased efficacy compared with the
`ER antagonist tamoxifen in postmenopausal women as first-line
`endocrine treatment for ER+ advanced breast cancer
`345 and as
`
`therapy for postmenopausal women with early breast
`adjuvant
`cancer.7’9 As such, AIS are now considered the standard of care as
`adjuvant endocrine therapy for postmenopausal women with hor-
`mone receptor—positive breast cancer.
`that
`is an ER antagonist
`Fulvestrant, a 17B—estradiol analog,
`competes with endogenous estrogen for binding to the ER”)
`However, unlike
`tamoxifen, which exhibits partial
`estrogen
`agonist activity, fulvestrant has no recognized estrogenic effect. It is
`thought that this is due to the fact that on binding to the ER,
`fulvestrant induces a conformational change, leading to degradation
`of the ER and complete inhibition of ER signaling in animal
`models.“
`
`Unfortunately, resistance to endocrine therapy will eventually
`develop. Although optimal sequencing of appropriate hormone
`therapies is the ideal approach, few randomized controlled trials
`have directly compared the effects of changing the order in which
`2 different agents are given.2 Furthermore, the paucity of data led
`
`i526-8209/S - see lronlmaller © 2014 Elsevier Inc. All righls reserved.
`hllp://dx.doi.org/l 0.l 0l 6/i.cbc.20l4.06.005
`
`Clinical Breast Cancer December 2014
`
`331
`
`AstraZeneca Ex. 2071 p. 1
`Mylan Pharrns. Inc. v. AstraZeneca AB IPR2016-013 16
`
`

`
`Fulvestrant Dose—Response Story
`the authors of a recent review to conclude that no definitive rec-
`
`ommendations could be made regarding the sequencing of endo—
`crine therapies in patients with advanced breast cancer, and that
`patients should receive the most efficacious treatment
`in that
`setting, while also considering specific side effect
`issues for that
`patiCI1t.Z Early preclinical data demonstrated a lack of cross—
`reactivity between fulvestrant and tamoxifen, with fulvestrant
`inhibiting the growth of tamoxifen—resistant tumors.” Similarly in
`the clinical setting, many postmenopausal women with advanced
`breast cancer
`that
`responded to first-line fulvestrant
`remained
`, 4
`13 1
`Furthermore,
`responsive
`to further
`endocrine
`treatment.
`tumors that have responded to prior treatment with an anties—
`trogen15’16 or an AIIUS may retain sensitivity to subsequent
`treatment with fulvestrant.
`
`Presurgical studies provide the opportunity to perform a detailed
`analysis and comparison of biomarker expression and biomarker
`response with various experimental drug treatments. As an example,
`the selective ER modulator tamoxifen was reported to increase PgR
`levels as a result of its partial estrogen agonist activity.” However,
`downregulation of ER with fulvestrant leads to reduction in PgR
`protein levels through disruption of ER—dependent
`transcription
`of the PgR gene, as shovsm in a randomized comparison with
`tamoxifen, highlighting the distinct mechanisms of action of these
`2 agents.2O Reduction in Ki67 expression, a nuclear antigen and
`marker of cell proliferation, is reported to correlate with treatment
`response to endocrine therapy in ER+ breast cancer,“ and Ki67 in
`short—term neoadjuvant studies has been shown to predict outcome
`in long—term adjuvant trials.2Z
`Clinical efficacy of fulvestrant was demonstrated in post—
`menopausal women with advanced breast cancer that had pro—
`6,
`,4
`1232 andwas
`gressed or recurred on prior antiestrogen therapy
`originally approved at a monthly dose of 250 mg. However, a dose—
`dependent
`effect was
`subsequently shown, with improved
`progression—free survival (PFS) for fulvestrant 500 mg (500 mg/mo
`intramuscular [IM] injection plus 500 mg on day 14 of month 1)
`versus the 250 mg dose. This led to approval of the 500 mg dose for
`the treatment of postmenopausal women with ER+ advanced breast
`cancer after failure of prior antiestrogen therapy.25
`This review investigates the dose—dependent effects of fulvestrant
`more broadly, in terms of both the reduction of tumor biomarkers
`in the presurgical setting and the clinical efficacy for the treatment
`of breast cancer.
`
`range of 50 to 500 mg administered using the commercially avail-
`able long—acting formulation. Data from Study 18,20 Neoadjuvant
`Endocrine Therapy for Women with Estrogen—Sensitive Tumors
`(NEWEST),27 and Trial 5728 were combined in this analysis.
`
`Study Designs
`Study 18
`Study 18 was a randomized, multicenter, partially blinded study
`that compared placebo, tamoxifen, fulvestrant 50 mg, fulvestrant
`125 mg, and fulvestrant 250 mg before surgery in postmenopausal
`women with previously untreated primary breast cancer.20 Patients
`received a single IM dose of fulvestrant 50 mg, 125 mg, 250 mg, or
`tamoxifen 20 mg daily, or tamoxifen placebo daily for 14 to 21 days
`before surgery. Only data from patients whose tumors were ER+ or
`PgR+ have been included in the current analysis. When patients
`had more than 1 tumor, baseline data from only the primary tumor
`were included.
`
`NEVVEST
`
`NEWEST (ClinicalTrials.gov identifier NCT0093002) was a
`randomized, multicenter, open—label, phase 11 study comparing
`fulvestrant 500 mg (500 mg/mo plus 500 mg on day 14 of
`month 1) with fulvestrant 250 mg/mo for 16 weeks before surgery
`in postmenopausal women with ER+ locally advanced breast
`cancer.27 Tumor biomarker levels at week 4 have been used in the
`
`present analysis for the closest consistency with data from Study 18
`and Trial 57.
`
`Trial 57
`Trial 57 (ClinicalTrials.gov identifier NCT00259090) was a
`randomized, multicenter, double—blind, phase II trial comparing
`fulvestrant 500 mg (single 1M dose) plus anastrozole (1 mg orally
`once daily for 14-21 days), fulvestrant 500 mg plus anastrozole
`placebo, or anastrozole plus fulvestrant placebo before surgery in
`postmenopausal women with ER+ primary breast cancer.28 Before
`protocol amendment, Trial 57 included a treatment phase in which
`patients were randomized to receive fulvestrant 250 mg plus anas-
`trozole (n : 6),
`fulvestrant 250 mg plus anastrozole placebo
`(n : 6), or anastrozole 1 mg plus fulvestrant placebo (n = 6).
`Although patient numbers are small and should be interpreted with
`caution, data for this initial treatment phase have been included
`for completeness in this analysis.
`
`Biological Rationale for a
`Dose-Response Relationship
`for Fulvestrant
`
`Dose—dependent reduction of tumor biomarkers after fulvestrant
`treatment was first demonstrated in a short—term presurgical study
`in postmenopausal women with primary breast cancer.Z6 After daily
`injections of a short—acting formulation of fulvestrant, reductions in
`ER expression and Ki67 labeling index were greater in patients with
`ER+ breast cancer who received a fulvestrant 18 mg daily injection
`compared with those who received a fulvestrant 6 mg daily
`injection.
`We now extend the study of dose dependency by presenting
`a retrospective analysis of tumor biomarker data extracted from
`3 previously reported presurgical studies over a fulvestrant dose
`
`Tumor Biomarker Expression and
`Statistical Analyses
`ER, PgR, and Ki67 expression were determined in each study by
`immunochemistry on sections of formalin—fixed, paraffin—embedded
`tissue. Study 18 used the following antibodies: ER, H222 (Abbott
`Laboratories, Abbott Park, IL); PgR, KD68 (Abbott); Ki67, MIB—1
`(Coulter Electronics, Luton, UK). In NEWEST,
`the antibodies
`used were the following: ER, 1D5 (Dako Ltd, Carpinteria, CA);
`PgR, 636 (Dako Ltd); Ki67, MIB—1 (Coulter Electronics). The
`antibodies used in Trial 57 were as follows: ER, 6F11 (Novocastra,
`Newcastle, UK); PgR, 636 (Dako Ltd); Ki67, Clone MIB—1 (Dako
`Ltd). Antigen retrieval methods and secondary detection methods
`varied between the studies and have been described.20’27‘28 ER, PgR,
`and Ki67 expression levels at pre— and post—treatment (14-21 days
`
`382
`
`Clinical Breast Cancer December 2014
`
`Astrazeneca Ex. 2071 p. 2
`
`

`
`Table 1
`
`change From Baseline in ER H—score
`
`]0/m F.R. Robertson et fll
`
`IBack-Transformed Least Squares Mean change From Baseline (%) (95% Cl)
`Trial 57~lnitial Phase
`Trial 57 Main Phase
`
`Study 18
`(_
`69.5 to 28.9)
`82.5 to —15.9
`(_
`—67.6 (—83.7 10 —35.5
`(_
`—75.2
`87.0 to —52.4)
`—84.0 (—91.7 to —69.1
`)
`
`) )
`
`—37.3
`—61.7
`
`Treatment
`Placebo
`Tamoxifen
`
`Fulvesiranl 50 mg
`Fulvestrani 125 mg
`Fulvestrani 250 mg
`Fulvestrani 500 mg
`Fu|vestran1250 mg plus anastrozole
`Fulvestrani 500 mg plus anastrozole
`Anastrozole
`
`—52.9 (433.0 to —4o.1)
`
`40.7 (—3o.3 to 14.4)
`
`—21.o (—5e.2 to 42.4)
`
`—43.2 (—68.4 to 2.1)
`
`5.8 (—41.0l0 89.7)
`
`—44.6 (—53.9 to —33.4)
`
`48.9 (—58.1 to —37.rs)
`—14.7 [~29] to 3.5)
`
`Abbreviations: Cl : confidence Interval; ER : estrogen; NBNEST : Neoadjuvant Endocrine Therapy for Women with Estrogen-Sensitive Tumors.
`
`post-treatment in Study 18 and Trial 57 and at week 4 in NEWEST)
`were determined by manual counting under light microscopy. ER
`and PgR expression were determined as the H—score, calculated
`as(0.5><%__)
`l (l><% 1)
`l
`(2><% l
`l)
`l
`(3><% l
`l 1):
`where % ::, % +, % ++, and % +++ represent the overall per—
`centage positivity of very Weak, weak, moderate, and strong staining,
`respectively. Ki67 expression was determined as the labeling index,
`derived from the number of positively stained epithelial cells,
`expressed as a percentage of the total number of cells counted.
`Tumor biomarker expression data were analyzed by study using
`an analysis of covariance (ANCOVA) model (log—transf0rmecl ratio
`of post— to pretreatment) With the log—transf0rmed baseline value
`and treatment included as factors. The least squares mean and
`confidence interval
`(CI) values were bacl<—transformed to the
`original scale. To assess the impact of fulvestrant dose While
`allowing for between—study variability, a second AN COVA model
`was produced including log—transformed baseline, dose (as a
`continuous variable), and study as factors. The first ANCOVA
`included all
`treatment groups within each trial;
`the second
`ANCOVA included only placebo and the fulvestrant 50 mg,
`125 mg, 250 mg, and 500 mg treatment groups. For the placebo
`data to be log—transformed, a dose of 0.5 mg rather than 0 mg was
`used for the purpose of this analysis.
`
`ER H-Score
`In Study 18, NEWEST, and Trial 57, a dose—dcpendent effect
`was seen over the dose ranges investigated for reduction in ER
`expression. In each study, the greatest reduction in ER expression
`was seen with the highest fulvestrant dose. In Study 18, greater
`reduction in ER was observed for
`fulvestrant 250 mg versus
`tamoxifen, and in Trial 57, greater reduction in ER expression was
`observed for fulvestrant 500 mg versus anastrozole. In Trial 57, no
`additional reduction in ER expression was observed for fulvestrant
`500 mg plus anastrozole compared with fulvestrant 500 mg alone
`(Table I; Figure l).
`
`PgR H-Score
`of fulvestrant was also
`€1T€Ct
`A consistent dose—dcpendent
`observed in Study 18, NEWEST, and Trial 57 for reduction in PgR
`expression. The greatest reduction in PgR expression was seen with
`the highest fulvestrant dose within each study. An increase in PgR
`
`expression was seen in the tamoxifen treatment group in Study 18.
`In Trial 57, no additional reduction in PgR expression was observed
`for
`the combination of fulvestrant 500 mg plus anastrozole
`compared with fulvestrant 500 mg alone or anastrozole alone.
`Similar reductions in PgR expression were observed for fulvestrant
`500 mg alone and anastrozole alone (Table 2; Figure 2).
`
`Ki67 Labeling Index
`Ki67 labeling index was reduced after treatment in each fulves-
`trant treatment group in each study. In Study 18 and NEW/EST,
`the greatest reduction in Ki67 labeling index was seen with the
`highest fulvestrant dose. In Trial 57, which also included the small
`initial cohort of patients treated with fulvestrant 250 mg (n = 6),
`there were no meaningful differences
`in K167 labeling index
`reduction between the fulvestrant
`treatment groups (Table 3;
`Figure 3).
`Overall results from the ANCOVA model show a consistent
`
`dose—dcpendent effect for fulvestrant over the dose ranges analyzed
`for ER and PgR H—score and Ki67 labeling index. Results for the
`
`Figure 1
`
`change From Baseline in ER Expression
`
`Change
`f'°'",
`"a5e""°
`(%)
`
`100
`2 Trial 57(r)
`3°
`60 ATria|57
`44,
`200
`6°
`
`-40
`«W
`400
`
`I
`P
`
`4o
`
`$tudy1B(II)
`NEWF5T(n)
`Trial 57(i) In)
`Trial 57 in)
`
`l
`
`I
`T
`
`31
`
`l
`
`I
`F50
`
`as
`
`I
`
`I
`F125
`
`35
`
`‘
`a
`I
`1
`I
`I
`F250
`
`4o
`92
`5
`
`4
`
`I
`
`a
`§
`I
`I
`I
`F500 F250 F500
`+A +A
`
`9;
`35
`
`5
`
`:1
`
`l
`'
`I
`A
`
`6
`:7
`
`Least squares mean and 9596 confidence Interval;output from an analysis Licovariance model
`of ERchange from baseline (natural log transformed) with treatment as a factor.
`Trial 57(D:datafrom initial patients inTrial 57, treated with F250, priorto protocol amendment.
`A. anastrozole; ER. estrogen receptor; F50/125/250/500, fulvestrant 50/1 15/350/500 mg; E placebo;
`T. lamoxifert
`
`Abbreviation: NEWEST : Neoadjuvant Endocrine Therapy for Women With Estrogen-Sensitive
`Tumors.
`
`Clinical Breasl Cancer December 2014
`
`383
`
`Astrazeneca Ex. 2071 p. 3
`
`

`
`Fulvestrant Dose—Response Story
`
`Table 2
`
`change From Baseline in PgR H-Score
`
`Treatment
`Placebo
`Tamoxiten
`Fulvestrant 50 mg
`Fulvestrant 125 mg
`Fulvestrant 250 mg
`Fulvestrant 500 mg
`Fulvestrant 250 mg plus anastrozole
`Fulvestrant 500 mg plus anastrozole
`Anastrozole
`
`IBack-Transformed Least Squares Mean change From Baseline (%) (95% Cl)
`Trial 57~lnitial Phase
`Trial 57 Main Phase
`
`Study 18
`40.3 (—25.8 to 165.4)
`160.1 (27.7 to 429.8)
`—62.7
`80.6 to —28.6)
`(_
`—7B.8 (—88.4 to —61.3)
`86.4( 92.8 to
`74.2)
`
`67.3( 81.0to
`
`43.7)
`
`—91.4 (—95.o to —35.o) 47.5 ( 82.8 to 60.3)
`
`—49.2 (—32.9 to 50.9)
`
`—65.9 (—88.8 to 4.0)
`
`—e3.2 r—77.2 to —4o.e)
`
`—5e.3 (—75.5 to —29.o)
`—59.2 (—75.0 to —33.6)
`
`Abbreviations: Cl : confidence Interval; NEWEST : Neoadjuvant Endocnne Therapy for Women with Estrogen—Sensitive Tumors; PgR : progesterone.
`
`second ANCOVA, which adjusted for between—study variability,
`show that increasing fulvestrant dose results in greater reduction in
`ER and PgR H—score and Ki67 labeling index (P < .0001 for the
`dose—response relationship for each biomarker).
`
`Clinical Evidence of a Dose-
`
`Response Relationship for
`Fulvestrant
`
`Fulvestrant D0se—Response in Sec0nd—Line Therapy for
`Advanced Breast Cancer
`
`The clinical efficacy of fulvestrant at a dose of 250 mg/mo was
`established in the registration trials 0020 and 0021, which
`compared fulvestrant 250 mg with anastrozole for the treatment of
`postmenopausal women with advanced breast cancer
`that had
`progressed or recurred on prior antiestrogen therapy.23’24 In a
`combined analysis of data from both studies (fulvestrant, n : 428;
`anastrozole, n = 423), fulvestrant 250 mg was shown to be at least
`as effective as anastrozole with respect to time to progression (TTP).
`Median TTP was 5.5 months for fulvestrant 250 mg compared
`with 4.1 months for anastrozole (hazard ratio [HR], 0.95; 95.14%
`
`CI, 0.82—1.10; P : .48).“ This led to the approval of fulvestrant
`250 mg for the treatment of postmenopausal women with advanced
`breast cancer that had progressed or recurred on prior antiestrogen
`therapy. However, evidence of dose—dependent clinical efficacy with
`fulvestrant had already been suggested in these studies, because an
`initial 125 mg dose was dropped after a planned interim assessment
`that found no evidence for clinical efficacy at the fulvestrant 125 mg
`dose. Given the favorable tolerability profile of fulvestrant 250 mg,
`alternative dosing regimens were investigated.
`The phase III COmparisoN of Faslodex In Recurrent or Meta—
`static breast cancer (CONFIRM) trial was designed to compare
`fulvestrant 500 mg with fulvestrant 250 mg in patients with hor—
`mone receptor—positive, pretreated, advanced breast cancer. Ful-
`vestrant 500 mg significantly prolonged PFS versus fulvestrant
`250 mg. Median PFS was 6.5 months in the fulvestrant 500 mg
`group compared with 5.5 months in the fulvestrant 250 mg group
`(HR, 0.80; 95% CI, 0.68—0.94; P : .006), demonstrating a clear
`dose—dependent
`relationship
`for
`fulvestrant
`in
`this
`setting
`(Table 4).” Of note, the dose—dependent clinical efficacy seen in
`CONFIRM was not associated with a dose—dependent increase in
`tOXiCity, with no substantial differences between the treatment
`
`groups in terms of incidence and severity of adverse events. This
`increase in therapeutic index led to fulvestrant 500 mg becoming
`the recommended dose. This benefit was further confirmed in a
`
`follow—up analysis performed when approximately 75% of patients
`had died. Median overall survival was 26.4 months for fulvestrant
`
`500 mg compared with 22.3 months for fulvestrant 250 mg,
`indicating a clinically relevant difference in overall survival between
`the treatment groups (HR, 0.81; 95% CI, 0.69-0.96; nominal
`P : .016)?"
`
`Fulvestrant Dose—Re@onse in First—Lz'ne Therapy for
`Advanced Breast Cancer
`
`Cross—trial comparisons also suggest a dose—response relationship
`for fulvestrant as first—line therapy for advanced breast cancer. In
`Trial 25, fulvestrant 250 mg failed to demonstrate noninferiority
`compared with tamoxifen, the standard of care at the time of the
`trial, in postmenopausal women with advanced breast cancer pre—
`viously untreated with endocrine therapy for advanced disease.”
`
`Figure 2
`
`change From Baseline in PgR Expression
`
`Change
`
`450
`
`150
`::)5°''"E 200
`100
`50
`
`'5mdy18
`
`x Trial 57
`
`'5°
`1 00
`
`_
`
`Study 15 (II)
`NEWEST (ll)
`Trial 5‘l(|) (n)
`Trial W (ll)
`
`l
`P
`
`40
`
`I
`T
`
`32
`
`fgilff
`Q
`I
`I
`I
`I
`I
`I
`I
`F50
`F1 25
`F250
`F500
`F750 F500
`A
`+ A
`+ A
`
`3!
`
`35
`
`40
`92
`6
`
`99
`35
`
`6
`
`31
`
`6
`37
`
`Least squares mean and 95% confidence Interval; outputfrom an analysis afcovarlance model
`ofPgll changefrom baseline (natural log transformed) with treatment as afactor.
`Trial 57(l): data from Initial patlents InTvlal 57,treated with F250, prbvtn pmtnml amendment.
`A, arastrcIzale;F5D/125/25oISO0,Fu|vestrant 50/125/250/500 mg; P, placebo; PgR. prugenerane reoepwr;
`1'. lamudlen
`
`Abbreviation: NEWEST : Neoadjuvant Endocrine Therapy for Women With Estrogen-Sensitive
`Tumors.
`
`384
`
`Clinical Breast Cancer December 2014
`
`Astrazeneca Ex. 2071 p. 4
`
`

`
`Table 3
`
`change From Baseline in Ki67 Labeling Index
`
`J0/J71 F.R. Robertson et all
`
`Back-Transformed Least Squares Mean change From Baseline (%) (95% Cl)
`Trial 57~lnitial Phase
`Trial 57 Main Phase
`
`Treatment
`Placebo
`Tamoxifen
`Fulvesiranl 50 mg
`Fulvestrant 125 mg
`Fulvestrant 250 mg
`Fulvestrant 500 mg
`Fulvestrant 250 mg plus anastrozole
`Fulvestrant 500 mg plus anastrozole
`Anasirozole
`
`Study 18
`(_
`18.0 to 31.1)
`3.7
`51.3 to —15.5)
`35.8
`(_
`—23.3 (—40.6 to ~09)
`—46.1
`58.6 to —29.7)
`(_
`58.1 to
`(
`46.5
`31.6)
`
`—81.2 (—85.8 to —75.0) 28.2)
`
`45.5 ( 58.5 to
`
`79.0( 90.4 to
`
`53.7)
`
`—91.1 (—96.0 to —80.2)
`
`84.4( 92.9 to
`
`65.6)
`
`—74.4 [—81.5io —04.5)
`
`—83.4 (—88.5 to —76.0)
`85.0[ 89.1 to
`79.4)
`
`Abbreviations: Cl : confidence Interval; NEWEST : Neoadjiivant Endocrine Therapy for Women wiln Eslrogen—Sensilive Tumors.
`
`to demonstrate
`shown
`previously
`anastrozole was
`Because
`improvements in efficacy over tamoxifen,3 this was considered a
`surprising outcome for fulvestrant 250 mg. However, with the
`almost immediate separation of the TTP curves in this trial, it was
`hypothesized that the 3 to 6 months to steady state for the fulves—
`trant 250 mg regimen could have led to the underperformance of
`this treatment group.
`In the phase II Fulvestrant flRst—line Study comparing endocrine
`Treatments (FIRST) study, fulvestrant 500 mg was compared with
`anastrozole in postmenopausal women with advanced breast cancer
`who had not received endocrine therapy for advanced disease. The
`fulvestrant 500 mg dose regimen, which includes a 500 mg dose at
`day 14, was shown to be at least as effective as anastrozole in terms
`of the primary endpoint of clinical benefit rate (fulvestrant, 72.5%;
`anastrozole, 67.0%), and the secondary endpoint of TTP was
`significantly longer for fulvestrant 500 mg compared with anas—
`trozole.“ Safety data indicated that fulvestrant 500 mg has a similar
`tolerability profile compared with anastrozole 1 mg and is well
`tolerated as first—line therapy for advanced breast cancer. In a follow—
`up analysis, which was performed when disease had progressed in
`approximately 75% of patients, median TTP was 23.4 months for
`fulvestrant 500 mg compared with 13.1 months for anastrozole
`(HR, 0.66; 95% CI, 0.47-0.92; P = .01).” This was the htst trial
`to indicate that an alternative endocrine therapy may be more
`effective than an A1 in the first—line setting for advanced breast
`cancer and indirectly suggests a dose—response relationship for ful—
`vestrant 500 mg over fulvestrant 250 mg as first—line therapy for
`advanced breast cancer. Given that fulvestrant 250 mg demon—
`strated noninferiority to anastrozole (in the second—line setting of
`the registration trials 0020 and 002116>23~Z“), the significantly longer
`TTP with fulvestrant 500 mg versus anastrozole in the first—line
`setting also was indirect evidence of a dose—response relationship
`for fulvestrant.
`
`Fulvestrant Dose Response in the Neoadjm/ant Setting
`NEWEST was the first study to compare the biological and
`clinical activity of the fulvestrant 500 mg dose regimen versus ful—
`vestrant 250 mg. Although the primary endpoint of NEWEST
`was biological (change in Ki67 labeling index from baseline to
`Week 4), the clinical data appeared to correspond with the dose—
`dependent reduction in tumor biomarkers seen at week 4. The
`
`tumor response rate at week 4 was 17.4% for the fulvestrant 500 mg
`group compared with 11.8% in the fulvestrant 2 50 mg group (odds
`ratio [OR], 1.68; 95% CI, 0.77-3.70; P: .19). At week 16, tumor
`
`response was 22.9% in the fulvestrant 500 mg group compared
`with 20.6% in the fulvestrant 250 mg group (OR, 1.30; 95% CI,
`0.64-2.64; P : .47) .27
`
`Fulvestrant in Combination Therapy
`Together with its distinct mechanism of action and reduced risk
`of cross—resistance with other endocrine treatments, the observa-
`tion of incomplete ER reduction with fulvestrant 250 mg,
`in
`the short,20 medium, and long term (Agrawal, in press),32 led to
`combination therapies being developed, aiming to further reduce
`ER activity and improve efficacy. The Fulvestrant and Anastrozole
`Combination Therapy (FACT) study compared the efficacy of a
`combination of anastrozole plus the fulvestrant 250 mg loading
`dose (LD) regimen (fulvestrant 250 mg 4» LD: 500 mg day 0, 250
`mg days 14 and 28, 250 mg/mo thereafter) versus anastrozole
`
`Figure 3 Change From Baseline in Ki67 Labeling Index
`
`f'°"‘.
`baseline
`(%)
`
`to
`20
`0
`
`-20
`
`‘°
`~80
`'10‘)
`
`studyiain)
`NEWET(n)
`Trlal 57(|) in)
`Trial 57 (n)
`
`ziireei
`ii Trial 570)
`A .ma|57
`
`}
`
`I
`P
`
`40
`
`I
`T
`
`32
`
`I
`F50
`
`I
`F125
`
`33
`
`as
`
`E
`I
`F250
`
`40
`92
`6
`
`1
`1
`a
`2
`I
`I
`F500 F250 F500
`+ A
`+ A
`
`99
`35
`
`5
`
`31
`
`i
`I
`A
`
`5
`:7
`
`Least squares mean and 9596 confidence intenralwutpul from an analysis cfcouarianne model
`of K167 labelling Index diange from baseline (natural log transformed)with treatment as I factor.
`Trill s7(otdmfrom initial patients inTri:| 57, treated with F250, pliarln pnmmoi amendment.
`A. nnnsirazole; F50/125/250/500,fu|vesm.int so/175/250/500 mg; P. p|lmd)0.'T. tlmoxifen.
`
`Abbreviation: NEWEST : Neoadjuvant Endocrine Therapy for Women With Estrogen-Sensitive
`Tumors.
`
`Clinical Breasl Cancer December 2014
`
`385
`
`Astrazeneca Ex. 2071 p. 5
`
`

`
`Fulvestrant Dose—Response Story
`
`Table 4
`
`PFS in Fulvestrant Monotherapy Trials for Advanced Breast Cancer
`
`Second—|ine
`studies
`
`Study 20*‘
`
`Study 21"‘
`
`EFECT
`
`FINDER1
`
`F|NDER2
`
`CONFIRM
`
`First-line
`studies
`
`Trial 25
`
`FIRST”
`
`Study Design
`
`Fulvestrant
`250 mg
`
`Fulvestrant
`250 mg + LD
`
`Fulvestrant
`500 mg
`
`Tamoxifen
`20 mg/d
`
`Anastrozole Exemestane
`25 mg/d
`
`3.7
`
`, randomized, open—label,
`Phase
`naralle group, mult'center study:
`tulvest ant 250 mg, n : 222;
`anastrozole, n : 22923
`nasel , andomized, doube— lind,
`parale—group, do ble—d mmy,
`ltioenter study: tulvestrant 250 mg,
`' anastrozo e, n : 19424
`andomized, doub
`lind,
`udy:
`
`'nestane, n
`andomized
`g oup, mu|t'
`ant 500 m
`250 mg —
`ant 250 mg, n :
`andomized, doub —
`g oup, mu|t'center
`ant 500 m
`250 mg —
`ant 250 mg, n : 4743
`randomized, double—blind,
`g oup, mu|t'center study:
`ant 500 mg, n : 362;
`ant 250 mg, n : 37425
`
`'
`
`Phase ll, randomized, doubleblind,
`parallelgroup, double—dummy,
`multicenter study: tulvestrant 250 mg,
`n : 313; tamoxifen, n : 2743”
`Phase I, randomized, open—label,
`paralle —group, multicenter study:
`tulvestrant 500 mg, n : 102;
`anastrozole, n : 10314
`
`Fulvestrant 250 mg: 250 mg days 0 and 28, 250 mg/mo thereafter; fulvestrant 250 mg + LD: 500 mg day 0, 250 mg days W4 and 28, 250 mg/mo thereafter; fulvestrant 500 mg: 500 mg days 0,
`14, and 28, 500 mg/mo thereafter.
`Abbreviations: CONFIRM — COmparisoN of Easlodex In Recurrent or Metastatic breast cancer; EEECT — Evaluation of Easlodex versus Exemestane Clinical Trial; EINDERW — Easlodex lnvestlgatioN of
`Dose evaluation in Estrogen Receptor—positive advanced breast cancer in Japan; FINDER2 : Faslodex lnvestigatiaN of Dose evaluation in Estrogen Receptor—positive advanced breast cancer in Europe;
`FlRST : Fulvestrant flRst-line Study comparing endocrine Treatments; LD : loading dose; PFS : progression-free survival.
`astudies 0020 and 0021 initially included afulvestrant 125 mg treatment group that was withdrawn because of lack of clinical activity.
`“Data from the FIRST follow—uo analysis.
`
`alone as therapy for postmenopausal women at first relapse after
`primary treatment for localized hormone receptor—positive breast
`cancer.“ More than 60% of the participants had received a prior
`endocrine therapy in the adjuvant setting. For
`the primary
`endpoint, median TTP was 10.8 months in the fulvestrant
`250 mg + LD plus anastrozole combination compared with
`10.2 months in the anastrozole alone treatment group (HR, 0.99;
`95% C1, 0.81—1.20; P = .91).55 The overall incidence ofAEs was
`similar between the 2 treatment groups. Secondary endpoints,
`including objective response rate, clinical benefit rate, and overall
`survival, were also similar between the 2 study arms, indicating no
`benefit
`for the anastrozole plus fulvestrant combination over
`anastrozole alone.
`
`The randomized phase 111 Southwest Oncology Group (SWOG)
`S0226 trial also compared the combination of anastrozole plus ful-
`vestrant 250 mg + LD with anastrozole alone as first—line therapy for
`postmenopausal women with metastatic breast cancer.“ The pri-
`mary endpoint of PFS was significant in favor of the combination
`group: 15.0 months compared with 13.5 months in the anastrozole
`alone group (HR, 0.80; 95% CI, 0.68-0.94; P : .007), and no
`safety concerns were raised with the fulvestrant plus anastrozole
`combination. In a retrospective analysis of those patients naive to
`prior tamoxifen therapy (414/694 patients: 59.7%), the median PFS
`Was 17.0 months
`in the combination group compared with
`12.6 months in the anastrozole alone group (HR, 0.74; 95% CI,
`0.59-0.92; P : .006). In those patients who had received previous
`
`386
`
`Clinical Breast Cancer December 2014
`
`Astrazeneca Ex. 2071 p. 6
`
`

`
`treatment with tamoxifen (280/694 patients: 40.3%), the median
`PFS was 13.5 months in the Combination group compared with 14.1
`months in the anastrozole alone group (HR, 0.89; 95% CI, 0.69—
`1.15; I’ = .37).“ In total, 166 of514 patients (32.3%) had received
`no prior adjuvant endocrine therapy in the FACT trial, whereas 414
`of 694 patients (59.7%) were naive to prior tamoxifen in SWOG
`S0226. In both trials, less than 2% of patients had received adjuvant
`therapy with an AI. Because the percentage of patients who were
`naive to prior adjuvant endocrine treatment was lower in the FACT
`trial, and any potential differences in clinical effectiveness may be
`more pronounced when comparing endocrine agents in hormone—
`naive patients, this could provide one potential explanation for the
`differences in efficacy between the FACT and SWOG S0226 trials.
`Furthermore, initial data from the Study of Faslodex, Exemestane
`and Arimidex (SoFEA) trial failed to demonstrate improved efficacy
`for the combination of fulvestrant with an AI over a monothcrapy
`treatment in the second—line setting. Similar PFS was reported for
`fulvestrant 250 mg + LD in combination with anastrozole compared
`with fulvestrant alone in postmenopausal patients with advanced
`breast cancer after progression on nonsteroidal AIs.55
`
`Discussion
`Our analysis of fulvestrant in 3 presurgical studies demonstrates
`a strong dose—dependent biological effect
`in the reduction of
`tumor biomarkers. Across each dataset analyzed, increasing fulves—
`trant dose leads to increased reduction in ER, PgR, and Ki67, and
`this dose—dependent reduction in tumor biomarkers corresponds to
`the dose—dependent clinical efficacy seen in postmenopausal women
`with advanced breast cancer in the second—line setting of trials
`0020 and 0021 (fulvestrant 250 mg vs. 125 mg) and CONFIRM
`(fulvestrant 500 mg vs. 250 mg). In the first—line setting, increased
`efficacy of fulvestrant 500 mg (vs. anastrozole in FIRST) versus
`250 mg (vs.
`tamoxifen in Trial 25) has been implied through
`indirect, cross—trial comparisons. Further cross—trial comparisons
`show that in the registration trials 0020 and 0021 (in the second—
`line setting), fulvestrant 250 mg was noninferior to anastrozole,
`but in the phase II FIRST trial (fulvestrant 500 mg vs. anastrozole
`in the first—line setting),
`fulvestrant demonstrated significantly
`longer TTP. The ongoing phase III Fulvestrant and AnastrozoLe
`COmpared in hormonal therapy Naive advanced breast cancer
`(FALCON) trial (ClinicalTrials.gov identifier: NCTOI602380) Will
`provide the additional