Phase II, Randomized, Double-Blind Study of Two Dose Levels
`of Arzoxifene in Patients With Locally Advanced or Metastatic
`Breast Cancer
`
`By Aman Buzdar, Joyce A. O’Shaughnessy, Daniel J. Booser, John E. Pippen, Jr., Stephen E. Jones, Pamela N. Munster,
`Patrick Peterson, Allen S. Melemed, Eric Winer, and Clifford Huolis
`
`arzoxifene
`of
`daily dose
`a
`select
`Purpose: To
`(LY353381), a selective estrogen receptor modulator, for
`use in future studies in women with locally advanced or
`metastatic breast cancer who are either potentially tamox-
`ifen sensitive (T5) or tamoxifen refractory (TR).
`Patients and Methods: This trial was a randomized, dou-
`ble-blind, phase II study of arzoxifene 20 mg (n = 55) and
`50 mg (n = 57) in women with advanced or metastatic
`breast cancer. Patients were randomly assigned to balance
`for number of metastatic disease sites, prior tamoxifen
`therapy, and estrogen receptor status. The primary end
`point was tumor response rate (RR). Secondary end points
`included clinical benefit rate (CBR), time to progression (TTP),
`and toxicity.
`
`Results: Forty-nine patients were TS and 63 were TR.
`According to independent review, among TS patients, RR
`was higher in the 20-mg arm than the 50-mg arm (26.1% v
`
`8.0%), with a longer TTP (8.3 v3.2 months; P > .05). Among
`the TR patients, response rate was the same in the 20-mg
`and 50-mg arms (10.3%) with similar TTP (2.7 and 2.8
`months, respectively; P > .05). CBR was higher in the 20-mg
`arm than in the 50-mg arm among TS patients (39.1% v
`20.0%) and TR patients (13.8% v 10.3%). Arzoxifene was
`well tolerated. Dose-dependent toxicity was not demon-
`strated. There were no deaths during study.
`Conclusion: Arzoxifene is effective in the treatment of TS
`
`and TR patients with advanced or metastatic breast cancer
`at the 20-mg and 50-mg dose levels. Toxicities are minimal,
`and the therapy is tolerated. The 20-mg dose seems to be at
`least as effective as the 50-mg dose. Accordingly, arzox-
`ifene 20 mg/d was selected for further study in patients
`with breast cancer.
`
`J Clin Oncol 21:1007-1014. © 2003 by American
`Society of Clinical Oncology.
`
`REAST CANCER is the most common malignancy among
`women in the Western hemisphere and the second most
`common cause of cancer-related mortality. A substantial body of
`experimental, clinical, and epidemiologic evidence indicates that
`steroid hormones play a major role in the etiology of breast
`cancer.1 The effects of steroid hormones on breast epithelium are
`mediated through estrogen and progesterone receptors (ER and
`PgR, respectively).2 Tamoxifen has been the drug of choice for
`endocrine manipulation of both early and advanced stages of
`breast cancer.3 Its biologic effects are mediated primarily by
`inhibiting the actions of estrogen through its binding to the ER.
`Although tamoxifen is generally a well-tolerated drug, it does
`have significant side effects. These include hot flashes (20% to
`80%), thromboembolism (1% to 3%), and a variety of ocular
`toxicities and endometrial cancer. The risk of developing endo-
`metrial cancer with tamoxifen is estimated to increase two- to
`
`seven-fold in postmenopausal women receiving long-term treat-
`ment.4'9 Moreover, there are concerns that long-term use (> 5
`years) in the treatment of early-stage breast cancer is associated
`with the development of tamoxifen-dependent breast cancer.6
`Given the above concerns, considerable attention has been
`paid to developing more selective antiestrogens. The nonsteroi-
`dal benzothiophene selective ER modulator arzoxifene was
`designed to have potent ER antagonistic activity in the breast and
`endometrium while maintaining beneficial estrogen agonist ac-
`tivity on bone and lipids. ln preclinical studies, both arzoxifene
`and its desmethyl metabolite bound to the ER with high affinity
`and inhibited estrogen-dependent growth of MCF-7 cells.10
`Arzoxifene does not stimulate the uterine endometrium in
`
`ovariectomized rats; however, it does block estrogen-induced
`
`stimulation of the endometrium.11 It also demonstrated favorable
`
`effects on bone and lipids in preclinical studies.12
`A phase 1 study of four doses of arzoxifene (10, 20, 50, and
`100 mg) was conducted in 32 patients with previously treated
`breast cancer.13 The most common side effect was hot flashes
`
`(56%). Prospective evaluation of uterine safety, performed at
`baseline and after 12 weeks of treatment, showed no evidence of
`endometrial stimulation. Although responses were not seen, six
`patients had stable disease lasting for 6 months or longer. As no
`dose-dependent toxicity was identified in that study, this study
`was conducted to further evaluate the safety and efficacy of
`arzoxifene 20 mg and 50 mg in patients with advanced or
`metastatic breast cancer, and to determine the dose of arzoxifene
`to be used in future phase HI trials.
`
`PATIENTS AND METHODS
`
`Study Design
`This was a randomized, double-blind, phase II study of arzoxifene. Each
`participating institutions’ independent review board gave approval to the
`
`
`
`From the AID. Anderson Cancer Center and US Oncology Research,
`Houston, and Baylor—Sammons Cancer Center and Texas Oncology. Dallas,
`TX; Memorial Sloan—Kettering Cancer Center, New York. NY: Eli Lilly and
`Company. Indianapolis, IN: and Dana—Farber Cancer Institute. Boston, MA.
`Submitted June 18, 2002; accepted December 9, 2002.
`This work was sponsored by Eli Lilly and Company.
`Address reprint requests to Aman Buzdar, MD, AID. Anderson Cancer
`Center, 1515 Holcombe Blvd, Box 424. Houston. TX 7703 0; email:
`abuzdar@mdanderson. org.
`© 2003 by American Society of Clinical Oncology.
`0732—183X/03/2106—1 007/$20. 00
`
`Journal of Clinical Oncology, Vol 21, No 6 (March 15), 2003: pp 1007-1014
`DOI: 10.1 200/JCO.2003.06.1 08
`
`1 007
`
`Downloaded from ascopubscrg by 151.194.33.114 on March 28, 2017 from 151.194.033.114
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`AstraZeneca Exhibit 2111 p. 1
`lnnoPharma Licensing LLC v. AstraZeneca AB lPR2017-00900
`Fresenius-Kabi USA LLC v. AstraZeneca AB IPR2017-01913
`
`

`

`1 008
`
`BUZDAR ET AL
`
`study design before enrolling patients. After providing informed consent,
`patients were randomly assigned to receive either 20 or 50 mg of arzoxifene
`(Eli Lilly and Company, Indianapolis, IN), taken as a single tablet daily with
`meals. Treatment was continued until disease progression or unacceptable
`toxicity occurred or informed consent was withdrawn. Patients experiencing
`disease progression at the 20-mg dose were eligible to receive further
`open-label arzoxifene treatment at a dose of 50 mg daily, at the investigator’s
`discretion. Treatment was discontinued for any study drugirelated grade 4
`toxicity. Randomization was performed using the Pocock and Simon
`method14 to maximize baseline treatment group balance according to three
`important prognostic factors: number of metastatic disease sites (< three
`or 2 three sites), prior tamoxifen therapy (yes or no), and degree of ER
`expression (high, low, or unknown). High ER expression was defined as 2
`50 fmol/mg of ER (biochemical) or Z 50% cells positive (immunohisto-
`chemistry), and low ER expression was defined as less than 50 fmol/mg of
`ER (biochemical) or less than 50% cells positive (immunohistochemistly).
`
`Eligibility Criteria
`
`The study population consisted of women who were potentially tamoxifen
`sensitive (TS), defined as no prior exposure to tamoxifen or patients who
`experienced relapse more than 12 months after cessation of adjuvant
`tamoxifen therapy, or tamoxifen refractory (TR). defined as patients who
`experienced relapse during adjuvant tamoxifen treatment (provided at least 1
`year had elapsed between initiation of tamoxifen and development of
`metastatic disease) or patients treated with tamoxifen as first-line therapy for
`metastatic disease whose disease was at least stable for 2 6 months 011
`tamoxifen and then progressed. All patients were women at least 18 years old
`with a pathologic diagnosis of locally advanced or metastatic breast cancer.
`Patients had an Eastern Cooperative Oncology Group performance status
`(PS) of 0 or 1. estimated life expectancy of 2 24 weeks, and tumors that were
`assessable or bidimensionally measurable and were ER- and/or PgR-positive.
`Patients with inoperable, locally advanced breast cancer were enrolled only
`if they were not good candidates, in the investigators judgment, for primary
`chemotherapy. Patients whose EIUPgR status was unknown were eligible if
`they were older than 50 years. Prior neoadjuvant or adjuvant chemotherapy
`was permitted if completed at least 6 months before diagnosis of metastatic
`disease. Patients who had received prior adjuvant hormonal therapy (includ-
`ing oophorectomy or ovarian irradiation) were enrolled, provided there was
`an interval of at
`least 12 months between completion of therapy and
`diagnosis of metastatic disease. Concomitant medications such as bisphos-
`phonates, hematopoietic growth factors, and palliative radiotherapy were
`permitted. Patients with child-bearing potential were required to use a barrier
`contraceptive method during and for 3 months after the trial.
`Patients were excluded from the study per investigator’s discretion if they
`had rapidly progressive disease, a serious concomitant systemic disorder, or
`predisposition to thromboembolic disorder; inadequate end-organ function
`(eg, serum creatinine 2 1.5‘
`times the upper limit of normal
`[ULN],
`bilirubin 2 1.5 times the ULN, and ALT or AST > 2.5 times the ULN);
`hypercalcemia', tumor known to be ER- and PgR—negative', or untreated brain
`metastases or were pregnant, breast-feeding, or had used any investigational
`agent within 4 weeks before study enrollment.
`The study was conducted in accordance with the principles of the
`Declaration of Helsinki. The protocol and consent process was approved by
`all relevant ethics boards, and all patients gave written consent before
`enrollment.
`
`Baseline Evaluations
`
`A complete history and physical examination including PS assessment,
`blood pressure, pulse, height, and weight were performed at baseline and at
`each subsequent physician visit. Assessment of tumor markers (eg, carcino-
`embryonic antigen, CA1573, and CA-125), urinalysis, and an ECG were also
`performed at baseline. Measurements of palpable lesions and tumor markers
`that were elevated at baseline were subsequently obtained monthly for 3
`months, then every 2 to 3 months. Additional testing included complete
`blood cell count, chemistry analysis, serum osteocalcin, sex hormones
`(luteinizing hormone [LH], follicle-stimulating hormone [FSH], estradiol,
`and sex-hormoneibinding globulin [SHBG]),
`and plasma levels of
`LY353381 and its desmethyl metabolite LY335562, which were obtained
`
`then every 2 to 3 months. Radiologic tumor
`monthly for 3 months,
`assessments were obtained at day 85 and then every 2 to 3 months while
`patients were em‘olled on the study. All patients who received at least 4
`weeks of treatment, had prestudy staging and tumor measurements, and had
`at least one tumor measurement while receiving treatment were considered
`assessable for the efficacy analysis. Toxicities were evaluated at each
`physician visit using the National Cancer Institute Common Toxicity Criteria
`grading system (version 1.0).15 All patients who received at least one dose of
`LY353381 were considered assessable for the safety analysis.
`
`Uterine Evaluations
`
`Uterine safety was prospectively evaluated in this trial for patients with an
`intact uterus. Transvaginal ultrasounds (TVUs) were performed no more than
`4 weeks before initiation of treatment, at day 85 (i 7 days), or sooner for
`those discontinuing from study for any reason. For patients who continued
`treatment beyond 12 weeks, a TVU was performed at least every 6 months
`in the first year and at least yearly thereafter. Endometrial thickness was
`considered significant, warranting further evaluation if it was more than 8
`mm at baseline or at subsequent evaluation or if the increase from baseline
`was 2 5 mm at a subsequent evaluation; follow-up occurred either with
`saline-infused sonohysterography, hysteroscopy/guided biopsy, or dilatation
`and curettage (D&C). If inadequate tissue was obtained by blind biopsy or
`D&C, further attempts to obtain endometrial tissue were required. Slides of
`endometrial tissue were reviewed centrally by Covance Inc (Princeton, NJ).
`
`Eflicacy Assessments
`
`Tumor response rate (RR) was assessed by the investigator and indepen-
`dent review panel using standard World Health Organization response
`criteria.16 Clinical benefit rate was prospectively defined as the sum of
`patients with complete response (CR), partial response (PR), or stable disease
`(SD) lasting 2 6 months during the study. An independent review panel
`consisting of three independent radiologists reviewed the data for all
`patients with a response or SD according to the investigator. There was
`no independent review for patients whose disease was assessed by
`physical examination alone. Tumor response data collected during the
`open-label dose-escalation phase were not
`included in the primary
`analysis of tumor response.
`Time to progression (TTP) was measured from the time of randomization
`until the time of documented progressive disease (PD), including death by
`any cause. The duration of response is identical to TTP but is only defined
`for patients who exhibit tumor response. Survival was defined as the time
`from randomization until death by any cause. Analyses of secondary end
`points were based on investigator-determined assessments.
`
`Pharmacokinetics
`
`Concentrations of LY353381 and its desmethyl metabolite LY335562 in
`plasma were evaluated monthly for the first 3 months, then every 2 to 3
`months while patients were enrolled on the study. The desmethyl metabolite,
`LY335562,
`is
`referred to in the protocol as unconjugated dihydroxy
`metabolite, LY335563. LY335563 is the hydrochloride salt of LY335562.
`Samples were collected at any time during each visit. Heparinized plasma
`collected from patients was analyzed for LY353381 and LY335562 using
`validated high-performance
`liquid chromatography/mass
`spectrometry
`(MS)/MS method.
`
`Statistical Methods
`
`The study was designed to enroll 37 patients per dose cohort, with the
`primary goal of selecting the better of two doses of arzoxifene.17 An early
`stopping rule was included in case one or both doses proved to be inactive.
`The selection procedure was simply to choose the dose with the higher
`observed response rate between treatment groups. Assuming the true
`response rate is at least 15% higher on the better dose, this design has a 90%
`probability of selecting the better dose. Note that the selection design does
`not control for type I errors in the comparison of response rates between the
`dose cohorts, so the question of whether there was a statistically significant
`difference in response rate was not addressed. Exact 95% binomial confi-
`
`Downloaded from ascopubs.org by 151.194.33.114 on March 28, 2017 from 151.194.033.114
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`AstraZeneca Exhibit 2111 p. 2
`
`

`

`ARZOXIFENE IN ADVANCED/METASTATIC BREAST CANCER
`
`1 009
`
`
`Table 1. Baseline Patient and Disease Characteristics
`
`Arzoxitene 20 mg
`Arzoxitene 50 mg
`
`Characteristic
`No, of Patients
`%
`No, of Patients
`%
`
`Randomized patients
`Age] years
`Median
`Range
`Performance status“
`0
`1
`
`ER/PgR status
`ER+, regardless of PgR
`ERe/PgR+
`ER unknown/PgR unknown
`Sites of metastasis
`Bone
`
`36 to 8A
`
`55
`
`17
`38
`
`46
`5
`A
`
`40
`
`100
`
`31
`69
`
`8A
`9
`7
`
`73
`
`56
`33 to 84
`
`57
`
`24
`32
`
`52
`2
`3
`
`38
`
`100
`
`A2
`56
`
`91
`4
`5
`
`<57
`
`Lung
`Liver
`Skin
`At least three metastatic sites
`Tamoxifen sensitivity
`56
`32
`56
`31
`Refractory
`
`
`
`
`24 AA 25Sensitive 44
`
`32
`1 8
`16
`18
`
`16
`1 1
`1 1
`12
`
`29
`20
`20
`22
`
`18
`10
`9
`10
`
`Abbreviations: ER, estrogen receptor; PgR, progesterone receptor.
`*One patient on the 50-mg arm did not have baseline performance status assessed.
`
`dence intervals (CIs) were computed for response rates, but only for purposes
`of illustrating the precision of the point estimates.
`For end points other than response rate, standard statistical analysis
`methods were used to summarize and compare cohorts. Kaplan—Meier13
`estimation and the log-rank test were used to evaluate TTP. The Mantel-
`Haenszel X2 test was used to compare incidence of toxicities accounting for
`severity. Changes from baseline in various end points (eg, hormones.
`bone-related markers, and physical examinations) were assessed within study
`arms using the nonparametric sign test to allow for nonsymmetrical distri-
`butions, and between study arms using the Wilcoxon rank sum test. All
`significance tests for secondary end points were performed at the .05 level,
`whereas all CIs used the 95% level.
`
`RESULTS
`
`Patient Characteristics
`
`Between May 1998 and February 2001, 121 patients were
`entered onto the study, which was conducted at six study centers
`in the United States. Nine patients were not assigned to treatment
`because they either did not meet eligibility criteria (seven
`patients) or decided not to enter (two patients). One hundred
`twelve women were enrolled; 55 were randomly assigned to
`receive 20 mg of arzoxifene, and 57 patients were randomly
`assigned to receive 50 mg. Table 1 lists the baseline patient and
`disease characteristics of all randomized patients. Of note, there
`were more PS 0 patients in the SO-mg arm than in the 20-mg arm
`(42% V 31%). Also, there were more ER-positive patients in the
`SO-mg arm than in the ZO-mg arm (91% V 84%). Sixty-three
`patients were defined as TR and 49 were considered TS. Overall,
`the characteristics were well matched when comparing the 20-
`and SO-mg cohorts, regardless of tamoxifen sensitivity.
`The baseline characteristics of randomized patients are listed
`in Table 2 by dose and tamoxifen sensitivity. In the comparison
`of these cohorts, there were some differences in TS patients, with
`more ER—negative/PgR-positive and premenopausal patients in
`the 20-mg arm than the SO-mg arm, and more ER-positive/PgR-
`
`negative patients in the SO-mg arm than in the 20-mg arm. Also,
`although the average time from completion of adjuvant tamox-
`ifen to study enrollment was similar between TS patients in the
`20-mg and SO—mg arms (2.7 and 2.8 years, respectively), the
`average length of tamoxifen exposure was longer in the SO-mg
`arm than the 20-mg arm (5.6 V 3.2 years). Among the TR
`patients, there were more postmenopausal patients and patients
`with PS 0 in the SO-mg arm than in the ZO-mg arm.
`
`Antiturnor Activity
`
`Of the 112 randomized patients, six were not qualified for
`analysis because of the following reasons: no measurable disease
`(three patients),
`treatment with an excluded medication (one
`patient), unspecified criteria not met (one patient), and wrong
`medication code (one patient). The overall R by dose irrespec-
`tive of tamoxifen sensitivity is shown in Table 3. The investi-
`gator-assessed RR and clinical benefit rate (CBR) of the 20-mg
`cohort were 19.2% and 28.8%, respectively. The investigator-
`assessed RR and CBR of the SO-mg cohort were 7.4% and
`20.4%, respectively. The peer-reviewed RR had one fewer PR on
`the 20-mg cohort and one more CR on the SO-mg cohort. There
`was no clear difference in R between the 20- and SO-mg doses,
`showing that both are in an effective dose range.
`According to the investigators’ assessment, in the TS cohort,
`there were seven responders (RR, 30.4%) and four patients with
`SD 2 6 months in the 20—mg arm (CBR, 47.8%) and two
`responders (RR, 8%) and six patients with SD 2 6 months
`(CBR, 32%)
`in the SO-mg arm (Table 4). The independent
`review process confirmed all but one PR and one SD in the
`20-mg treatment arm, yielding an R of 26% and a CBR of 39%.
`In the SO—mg arm,
`the independent review panel found one
`additional CR, one fewer PR, and three fewer patients with SD,
`yielding an R of 8% and a CBR of 20%. The Kaplan—Meier
`
`Downloaded from ascopubs.org by 151.194.33.114 on March 28, 2017 from 151.194.033.114
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`AstraZeneca Exhibit 2111 p. 3
`
`

`

`t 01 O
`
`BUZDAR ET AL
`
`
`Table 2. Baseline Characteristics of All Randomized Patients
`
`Tamoxifen-Sensitive
`Tamoxifen-Refractory
`
`20 mg (n : 24)
`50 mg (n : 25)
`20mg (n : 31)
`50 mg (n : 32)
`
`No. of Patients
`%
`No, of Patients
`’70
`No. of Patients
`%
`Characteristics
`No. :of Patients
`%
`
`Age, years
`Median
`Range
`ER/PgR status
`Positive/positive
`Positive/negative
`Positive/unknown
`Negative/positive
`Unknown/unknown
`Performance status
`0
`1
`
`56
`43 to 84
`
`56
`33 to 84
`
`59
`36 to 81
`
`57
`37 to 82
`
`17
`2
`0
`3
`2
`
`8
`16
`
`71
`8
`0
`1 3
`8
`
`33
`67
`
`15
`6
`1
`0
`3
`
`9
`16
`
`60
`24
`4
`0
`1 2
`
`36
`64
`
`19
`7
`1
`2
`2
`
`9
`22
`
`61
`23
`3
`6
`6
`
`29
`71
`
`22
`8
`0
`2
`0
`
`15
`16
`
`69
`25
`0
`6
`0
`
`47
`50
`
`Menopausal status
`Premenopausal
`Postmenopausal
`Prior adiuvant
`chemotherapy
`Prior tamoxifen
`No. of disease sites
`72
`23
`68
`21
`60
`15
`58
`14
`< Three sites
`
`2 Three sites 28 10 42 10 40 10 32 9
`
`
`
`
`
`
`
`
`6
`18
`9
`
`8
`
`25
`75
`38
`
`33
`
`2
`23
`10
`
`6
`
`8
`92
`40
`
`24
`
`8
`23
`19
`
`31
`
`26
`74
`61
`
`1 00
`
`3
`29
`17
`
`32
`
`9
`91
`53
`
`1 00
`
`Abbreviations: ER, estrogen receptor; PgR, progesterone receptor.
`
`estimates for TTP were 8.3 months (95% CI, 3.4 to 18.4 months)
`in the 20-mg arm and 3.2 months (95% CI, 2.8 to 6.2 months) in
`the 50-mg arm (Fig 1). Thus, regardless of source of efficacy
`assessment (investigator or independent panel), both dose levels
`were effective, with the 20-mg dose of arzoxifene showing more
`responses in TS patients.
`in the TR cohort,
`According to investigators” assessment,
`there were three responders (RR, 10.3%) and one patient with
`SD 2 6 months (CBR, 13.8%) in the 20-mg arm and two
`responders (RR, 6.9%) and one patient with SD (CBR, 10.3%) in
`the 50-mg arm (Table 4). According to the independent review
`panel, TR patients had the same RR on both the 20-1ng and
`50-mg dose (10.3%), whereas CBR was higher on the 20-mg
`dose (13.8% V 10.3%). The Kaplan—Meier estimates for TTP
`were 2.7 months (95% CI, 2.5 to 2.9 months) in the 20-mg arm
`and 2.8 months (95% CI, 2.5 to 2.9 months) in the 50-1ng arm
`(Fig 2). Table 5 lists investigator-assessed response durations
`among qualified patients.
`
`There were six patients (four TR and two TS patients) who
`crossed over to open-label 50-mg treatment after they experi-
`enced PD on the 20-mg arm. During the double-blind 20-mg
`treatment phase, all
`six patients experienced PD within 3
`months. None of these patients achieved a tumor response during
`the open—label phase.
`A survival analysis was not performed because more than
`80% of the enrolled patients were still alive at the time of the
`final analysis.
`
`Toxi city
`
`One patient with a prior diagnosis of cholelithiasis on the
`20-mg arm experienced grade 3 transaminase and grade 3
`bilirubin elevations. Overall, grade 2 laboratory aberrations were
`reported in less than 5% of patients, with only 7% of patients
`experiencing grade 1 aberrations. There were no investigator-
`determined grade 4 laboratory abnormalities.
`
`Table 3. Tumor Response and Clinical Benefit Rate by Dose Irrespective of Tamoxifen Sensitivity
`
`Investigator-Assessed Peer-Reviewed
`20 mg
`50 mg
`20 mg
`50 mg
`
`(n : 52)
`(n : 54)
`(n : 52)
`(n : 54)
`
`1 + 4
`3 + 6
`0 + 4
`3 + 7
`Objective tumor response, CR + PR, n
`9.3
`17.3
`7.4
`19.2
`Response rate, %
`3.1 to 20.3
`8.2 to 30.3
`2.1 to 17.9
`9.6 to 32.5
`95% CI within group, %*
`74.8 to 20.9
`71.0 to 24.6
`95% CI between group, 981'
`3 + 6 + 4
`1 + 4 + 3
`3 + 7 + 5
`0 + 4 + 7
`Clinical benefit response, CR + PR + SD26 months, n
`25.0
`14.8
`28.8
`20.4
`Clinical benefit rate, %
`14.0 to 38.9
`6.6 to 27.1
`17.1 to 43.1
`10.6 to 33.5
`95% CI within group, 96*
`
`95% CI between group, 961'
`77.9 to 24.8
`74.9 to 25.3
`
`Abbreviations: CR, complete response; PR, partial response; SD, stable disease; CI, confidence interval.
`”Within-group confidence interval is exact (based on binomial distribution).
`‘l'Between-group confidence interval is based on normal approximation.
`
`Downloaded from ascopubsorg by 151.194.33.114 on March 28, 2017 from 151.194.033.114
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`AstraZeneca Exhibit 2111 p. 4
`
`

`

`ARZOXIFENE IN ADVANCED/METASTATIC BREAST CANCER
`
`1011
`
`
`
`Table 4. Investigator-Assessed Tumor Response and Clinical Benefit Rate by Tamoxifen Sensitivity
`Ta moxifen-Sensitive
`Tamoxifen-Refractory
`20 mg
`50 mg
`20 mg
`50 mg
`(n : 23)
`(n : 25)
`(n : 29)
`(n : 29)
`
`
`2 + 1
`0 + 2
`1 + 6
`Objective tumor response, CR + PR, n
`10.3
`8.0
`30.4
`Response rate, %
`0.8 to 22.8
`2.2 to 27.4
`1.0 to 26.0
`13.2 to 52.9
`95% CI within group, 98*
`711.0 to 17.9
`0.8 to 44.0
`95% CI between group, %‘1'
`2 + 1 + 1
`0 + 2 + 1
`1 + 6 + 4
`0 + 2 + 6
`Clinical benefit response, CR + PR + SD 26 months, n
`13.8
`10.3
`47.8
`32.0
`Clinical benefit rate, %
`3.9 to 31.7
`2.2 to 27.4
`26.8 to 69.4
`14.9 to 53.5
`95% CI within group, %*
`
`95% CI between group, 961' 713.3 to 20.2 711.6 to 43.2
`
`Abbreviations: CR, complete response; PR, partial response; CI, confidence interval; SD, stable disease.
`*Within-group confidence interval is exact (based on binomial distribution).
`‘l'Between-group confidence interval is based on normal approximation.
`
`0 + 2
`
`Table 6 presents clinical toxicities reported in at least 2% of
`patients treated with arzoxifene. There were no statistically
`significant differences in the toxicities observed between the
`treatment arms. The most common grade 2 clinical toxicities
`were hot flashes and nausea. Seven patients (6%) reported with
`grade 3 toxicities, including nausea/vomiting, rash, neuromotor
`toxicity (defined as fatigue and asthenia', n = 2), neuromood
`toxicity (defined as depression), headache, neurocerebellar tox-
`icity (defined as dizziness), and puhnonary toxicity (defined as
`dyspnea). These events did not result in drug discontinuation.
`There were no grade 4 clinical toxicities.
`Five patients discontinued from the study because of adverse
`events. One patient experienced deep venous thrombosis, dys-
`
`pnea, and edema 3 weeks after femoral rod placement for a
`pathologic fracture. According to the investigator, this serious
`event was considered possibly related to surgery as well as study
`drug. One patient was hospitalized for confusion and dyspnea
`and was subsequently diagnosed with a blockage in her carotid
`arteries. Another patient was hospitalized and diagnosed with a
`new primary cancer of the colon. Both of these serious events
`were considered unrelated to study drug. Two other patients
`discontinued treatment because of nonserious events: severe
`
`temporomandibular joint pain in one patient in the 20-mg ami
`and severe hot flashes in one patient in the 50-mg arm.
`There were five deaths that all occurred within 5 weeks of
`
`study discontinuation. Four patients died as a result of PD and
`
`1.0
`
` 0.6
`ProportionProgression—free p U
`
`
`
`0.1—
`
`Log—Rank P—Vatue .0949
`Wilcoxon
`P—Vatue .0941
`
`Fig 1. Kaplan-Meierla estimate for time
`to progression (months) in tamoxifen-sensi-
`tive patients.
`
`o
`Patients at risk
`23
`25
`
`6
`
`1 1
`a
`
`12
`
`7
`3
`
`18
`
`5
`3
`
`24
`
`3
`3
`
`Months
`
`30
`
`0
`1
`
`.36
`
`Months
`
`o SRM 20mg
`0 SRM 50mg
`
`Therapy Treatment Code — SRM 20mg
`
`~~ SRM 50mg
`
`Downloaded from ascopubsorg by 151.194.33.114 on March 28, 2017 from 151.194.033.114
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`AstraZeneca Exhibit 2111 p. 5
`
`

`

`
`
`0.0—-‘ i
`i
`.
`i
`i
`i
`i
`r
`O
`3
`6
`9
`‘12
`1 5
`18
`2‘1
`Months
`Patients at risk
`29
`29
`
`4
`‘1
`
`4
`‘l
`
`3
`O
`
`‘l
`O
`
`OSRM 20mg
`OSRM 50mg
`
`1012
`
`
`
`
`
`0.8—
`
`0.5—
`
`'7
`Lag—Rank P—Vulue .7047
`Wilcoxcfi
`P—vmue .8556
`|—i
`
`BUZDAR ET AL
`
`Fig 2. Kapian-Meier18 estimate for
`time to progression (months) in tamox-
`ifen-refractory patients.
`
`
`Progressiomfree 0.2—-
`Proportion
`
`O.‘l -
`
`7
`9
`
`4
`O
`Moflths
`Therapy Treatmer‘it Code ‘ SRM 20mg
`
`one died as a result of unspecified natural causes that were not
`disease related.
`
`Effects on the Endometrium
`
`A11 enrolled patients with an intact uterus underwent a
`baseline TVU evaluation. Because of the natural and potentially
`complex variations in endometrial
`thickness in menstruating
`women, only postmenopausal women are discussed in this
`report. Ninety-three of the l 12 patients were postmenopausal; 33
`of the 93 patients had undergone hysterectomy before enroll-
`ment. Sixty patients underwent baseline TVU evaluations, 21 in
`the 20-mg arm and 39 in the 50-mg arm. Forty-six of the 60
`patients had at least one follow-up TVU. Although none of the
`postmenopausal patients in the ZO-mg arm experienced signifi-
`cant endometrial thickness increases (see Patients and Methods,
`under Uterine Evaluations), one patient did experience vaginal
`bleeding attributed to atrophic vaginitis. Five patients in the
`50-mg arm had a significant increase in endometrial thickness.
`Four of these five patients did not undergo the study-defined
`
`Table 5. Investigator-Assessed Response Duration (months)
`
`
`Tamoxifen-Sensitive
`Tamoxifen-Refractory
`20 mg
`50 mg
`20 mg
`50 mg
`
`(n : 23)
`(n : 25)
`(n : 29)
`(n : 29)
`PR-8.3
`PR-é.4
`PR-12.9*
`PR-2.9
`PR-10.9
`PR-30.3*
`CR- 15.4“
`PR-4.5
`F'R-17.3’K
`CR-19.4
`PR-18.4
`PR-27.2’t
`PR-28.6*
`CR-29.6”
`
`Abbreviations: CR, complete response; PR, partial response.
`*Censored duration.
`
`~ SRM 50mg
`
`follow-up evaluations; however, two of the four patients had
`normalization of their endometrial thickness without diagnostic
`or therapeutic intervention. For the one patient who underwent
`further evaluation of her increased endometrium, the obtained
`tissue was insufficient for diagnosis.
`
`
`Table 6. Clinical Toxicities Reported in at Least 2% of Patients
`Total Patients
`
`(N : 112)
`CTC Gracie (no. of patients)
`
`Toxicity
`No.
`”/0
`1
`2
`3
`4
`Hot Flashes
`20 mg
`50 mg
`Nausea
`20 mg
`50 mg
`Cutaneous toxicity
`20 mg
`50 mg
`Neuromotor toxicity
`20 mg
`50 mg
`Weight changes
`20 mg
`50 mg
`Alopecia
`20 mg
`50 mg
`Neuroheadache
`20 mg
`50 mg
`Vomiting
`0
`1
`0
`O
`2
`1
`20 mg
`
`50 mg 0 2 A 2 0 0
`
`
`
`
`
`Abbreviation: CTC, National Cancer Institute common toxicity criteria (version 1).
`
`42
`50
`
`21
`25
`
`1 1
`2
`
`A
`7
`
`5
`A
`
`5
`0
`
`2
`A
`
`13
`1 A
`
`7
`12
`
`5
`0
`
`1
`2
`
`2
`2
`
`3
`O
`
`1
`1
`
`10
`1 A
`
`3
`2
`
`1
`0
`
`1
`0
`
`1
`0
`
`0
`0
`
`0
`0
`
`0
`O
`
`1
`0
`
`0
`1
`
`0
`2
`
`0
`0
`
`O
`0
`
`0
`1
`
`0
`O
`
`0
`0
`
`0
`0
`
`0
`0
`
`0
`0
`
`0
`0
`
`0
`0
`
`23
`28
`
`1 1
`14
`
`6
`1
`
`2
`4
`
`3
`2
`
`3
`0
`
`1
`2
`
`Downloaded from ascopubsorg by 151.194.33.114 on March 28, 2017 from 151.194.033.114
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`AstraZeneca Exhibit 2111 p. 6
`
`

`

`ARZOXIFENE IN ADVANCED/METASTATIC BREAST CANCER
`
`1013
`
`Table 7. Summary of Changes from Baseline to End Point: Hormonal and Bone Biomarker Evaluations
`Treatment
`Group
`Baseline
`End Point
`Median
`
`Lab Test
`(mg)
`N“
`Median
`Median
`Change
`P
`
`FSH, IU/L
`
`LH, U/L
`
`Estradiol, pmoI/L
`
`SHBG/ nmoI/L
`
`20
`50
`20
`50
`
`20
`50
`20
`50
`
`53
`55
`50
`53
`
`35
`33
`53
`53
`
`34
`44
`23
`26
`
`39.6
`23.1
`65
`56
`
`33
`30
`20
`21
`
`22.4
`13.6
`77
`79
`
`*3
`*6
`*1 .5
`*1
`
`*73
`*73
`+7
`+1 1
`
`> .500
`
`.401
`
`.483
`
`> .500
`
`> .500
`70.8
`13.6
`15.9
`52
`20
`Osteocalcin, tug/L
`
`50 73.31 53 18.4 14.4
`
`
`
`
`Abbreviations: FSH, IoIIicIe-stimulating hormone; LH, Iuteinizing hormone; SHBG, sex-hormone—bincling globulin.
`*Number of premenopausal and postmenopausal patients with both baseline and postbaseline values.
`1P < .05 within dose group change from baseline (VViIcoxon signed rank test).
`
`Hormone and Bone Biomarker Evaluations
`
`Pharmaooki neti cs
`
`There were no statistically significant differences within or
`between treatment arms in changes from baseline in serum LH,
`FSH, estradiol, or SHBG (Table 7). However, when the data are
`analyzed according to menopausal status, decreases in LH and
`F SH with a concomitant increase in SHBG in postmenopausal
`women, consistent with a weak estrogen agonist effect on the
`pituitary, become evident (data not shown). These findings have
`been previously reported in postmenopausal women receiving
`tamoxifen therapy.”
`In the premenopausal patients, interpretation of F SH, LH, and
`estradiol
`is hampered by lack of information on timing in
`relation to menstrual cycle. However,
`the noted increase in
`estradiol has also been previously reported in premenopausal
`women treated with tamoxifen and may represent estrogen
`antagonist effect on the premenopausal pituitary or direct ovar-
`ian stimulation by selective estrogen-receptor modulators.19
`The serum osteocalcin level was measured at baseline, every
`4 weeks for 12 weeks, and then every 2 to 3 months until study
`discontinuation. The osteocalcin data analysis consisted of a
`comparison in median change from baseline to last observed
`value for all randomized patients, as well as a comparison
`between the two treatment groups. A total of 105 patients had at
`least
`two osteocalcin measurements (88 postmenopausal pa-
`tients; 17 premenopausal patients). Although there was a statis-
`tically significant decline in osteocalcin among: the postmeno-
`pausal patients in the 50-mg arm (median change, —3.5 ,LLg/L),
`such a decline was not noted in the postmenopausal patients in