`
`Winrich Rauschning’ and KathleenI. Pritchard?
`' formerly Clinical Research Oncology, Klinge Pharma GmbH, WeihenstephanerStrasse 32, D-81673 Munich,
`Germany, and * Division of Medical Oncology, Toronto-Bayview Regional Cancer Centre and Sunnybrook
`Health Science Centre, University of Toronto, 2075 Bayview Avenue, Toronto, Ontario, Canada M4N 3M5
`
`Key words: antiestrogens, breast cancer, droloxifene,clinicaltrial,
`
`Abstract
`
`Droloxifene, a new antiestrogen, has theoretical advantages over tamoxifen based on preclinical data. These
`include higheraffinity to the estrogen receptor, higher antiestrogenic to estrogenic ratio, and moreeffective
`inhibition of cell growth and division in ER positive cell lines, as well as less toxicity, including reduced carci-
`nogenicity in animal models. Droloxifene also exhibits more rapid pharmacokinetics, reaching peak concen-
`trations and being eliminated much more rapidly than tamoxifen. A phase II study compared droloxifene in
`dosages of 20, 40, and 100 mg daily in postmenopausal womenwith metastatic, or inoperable recurrent, or
`primary locoregional breast cancer who had not received prior hormonal therapy. Of 369 patients rando-
`mized, 292 wereeligible and 268 evaluable for response. Response rates (CR + PR) were 30%in the 20 mg
`group, 47% in the 40 mg group, and 44% in the 100 mg group (40 mg vs 20 mg, p = 0.02; 100 mg vs 20 mg, p =
`0.04; pooled 40 + 100 mg vs 20 mg, p = 0.01). Median response durationalso favouredthe higher dosages (20mg
`group = 12 months; 40 mg group = 15 months; 100 mg group = 18 months). When adjusted for prognostic
`factors, time to progression wassignificantly better for the 100 mg (p = 0.01) and the 40 mg (p = 0.02) group
`compared to the 20 mg group. Droloxifene increased SHBG and suppressed FSHat all dosages and sup-
`pressed LH at the 40 and 100 mg dosages. These hormonaleffects increased with increasing dosage. Short-
`term toxicity was generally mild, and similar to that seen with other antiestrogens. Droloxifene appears active
`and tolerable. It may have a particular role in situations in which rapid pharmacokinetics, or an increased
`antiestrogenic to estrogenic ratio, are required.
`
`Introduction
`
`Endocrine therapy remains a mainstayin the treat-
`ment of breast cancer, both in the adjuvantsetting
`and for metastatic disease. The antiestrogen tamox-
`ifen has become standard first-line therapy for
`metastatic disease in postmenopausal and fre-
`quently in premenopausal women. Tamoxifenis al-
`so increasingly used as standard adjuvant therapy
`for node-positive and node-negative postmenopau-
`
`sal women,althoughits role in the adjuvant therapy
`of premenopausal women remains less clear. Ta-
`moxifen has becomeso widely used largely because
`of its very acceptable short-term toxicity profile in
`comparison to earlier antiestrogens and to histor-
`ically standard hormonessuch as estrogen. Now
`that tamoxifen has been used for longer periods in
`adjuvant therapy and is being studied as a preven-
`tive in womenat high risk of developing breast can-
`cer, its short and long-term toxicity profiles are un-
`
`Address for offprints: W. Rauschning, Clinical Research Oncology, Klinge Pharma GmbH, WeihenstephanerStrasse 32, 81673 Munich,
`Germany
`
`AstraZeneca Exhibit 2047 p. |
`InnoPharma Licensing LLC v. AstraZeneca AB IPR2017-00905
`Fresenius-Kabi USA LLC v. AstraZeneca AB IPR2017-01912
`
`
`
`84
`
`W Rauschning and KI Pritchard
`
`dergoing increasing scrutiny. Several recent reports
`suggest that tamoxifen may exert beneficial influ-
`ences on both bone mineralization and lipid metab-
`olism, perhaps because of its mild estrogenic ef-
`fects. Some of tamoxifen’s more serious, although
`infrequent, short-term complications such as hy-
`percalcemia, deep venous thrombosis, and pulmo-
`nary embolism,as well as an increased incidence of
`endometrial cancer, however, are probably also re-
`lated to this mild estrogenic effect. In addition, in
`animal models, tamoxifen has shown a probably
`non-hormonal hepatic carcinogenicity which has
`not, as yet, been demonstrated to occur in humans.
`Thus, the developmentoverthe last 5—10 years of a
`number of newer antiestrogens has raised consid-
`erable interest in the oncology community. Many of
`these antiestrogens are, however, very similarto ta-
`moxifen, both in their estrogenic to antiestrogenic
`profiles and in their pharmacokinetic properties.
`Droloxifene (3-hydroxy-tamoxifen) is a new non-
`steroidal antiestrogen of some particular interest,
`since preclinical in vitro and in vivo testing have
`shown a numberof differences which represent po-
`tential advantages in comparison to tamoxifen.
`These include a 20-60 fold higheraffinity to the
`estrogen receptor [1], lower estrogenic and higher
`antiestrogenic effects on the rat uterus(higherther-
`apeutic index) [2], more effective inhibition of the
`growth of ER-positive human breast cancer cell
`lines [3-5], more effective reduction of S-phase pro-
`portion in a variety of cell lines [6], complete sup-
`pression of growth factor (IGF-1) stimulated prolif-
`eration of MCF-7 cells when a therapeutic concen-
`tration is achieved [3], more effective blockade of
`estrogen-activated c-myc-expression [5], higher
`production of TGF-B in MCF-7 cells [6, 7], and
`more effective growth reduction of various experi-
`mental and transplanted tumorsin animals (R 3230,
`DMBA,T61){3,5, 8, 9]. In addition, in comparative
`animal toxicity trials droloxifene is qualitatively
`and quantitatively better tolerated than tamoxifen
`[10-15]. Furthermore, droloxifene does not trans-
`form Syrian hamster embryofibroblasts as tamoxi-
`fen does, nor produce ovarian or Leydig cell tumors
`in mice [16], suggesting that it may not have the
`same potential to increase the incidence of endo-
`metrial carcinoma in women that tamoxifen does. It
`
`is also notable that droloxifene, in contrast to ta-
`moxifen, does not induce hepatic carcinomasin rats
`[17], suggesting a lack of hepato-carcinogenic po-
`tential in humans.
`In addition, several pharmacokinetic properties
`of droloxifene are of potentialclinical advantage in
`comparison to those of tamoxifen. First, droloxi-
`fene is itself the active anticancer compound [18,
`19], in comparison to tamoxifen, which must be me-
`tabolized in order to form the active product, 4-
`OH-tamoxifen [20]. As a result, the concentration
`of droloxifene necessary for antitumoractivity is
`reached within the first day of therapy [18, 19], in
`contrast to tamoxifen, for which effective concen-
`trations are attained only from the 11th day onwards
`[20]. Furthermore, extensive accumulation of ta-
`moxifen takes place, extending over 4 weeks[20],
`the wash-out time of tamoxifen being 7 times longer
`than that of droloxifene, which has a serum elim-
`ination half life of one to one and a half-days [19].
`Becauseof these encouraging preclinical charac-
`teristics, and following several phaseI-II studies in
`which droloxifene demonstrated safety, good toler-
`ability, and some efficacy in women who had previ-
`ously received tamoxifen and/or other endocrine
`therapy[21, 22], this new antiestrogen wasthentest-
`ed asfirst line therapy in the study described below,
`a multicentre, double-blind, randomized, dose find-
`ing study comparing 20, 40, and 100 mg of droloxi-
`fene given daily in postmenopausal women who
`had received no previous hormonal therapy for
`metastatic breast cancer. Early results from a pre-
`liminary analysis of some of the women entered on
`this study have been previously described [23].
`
`Patient selection and methods
`
`Study design
`
`The study described was carried out by investiga-
`tors from Austria, Belgium, Brazil, Canada, France,
`Germany,
`Spain,
`Sweden,
`Switzerland,
`the
`Netherlands, and the United Kingdom. Thetrial
`was designed to determine the optimal daily dose of
`droloxifene in postmenopausal women not previ-
`ously exposed to systemic anti-tumor hormonal
`
`AstraZeneca Exhibit 2047 p. 2
`
`
`
`therapy. Patients were randomizedto receive 20, 40,
`or 100 mg of droloxifene given in double-blind fash-
`ion. Each centre was supplied with numbered medi-
`cations in three doses. The randomization was done
`in groupsof six patients. There wasnostratification
`other than by centre.
`
`Patient selection
`
`Patients included were postmenopausal women
`with histologically proven breast cancer who,at the
`time of study entry, had advanced disease which
`could include distant metastatic cancer, inoperable
`recurrent loco-regional cancer, or inoperable pri-
`mary local or loco-regional cancer. All patients
`were required to have at least one lesion that was
`evaluable or measurable according to the World
`Health Organization (WHO)criteria [24]. Study
`candidates were required to have positive estrogen
`or progesterone receptor status in either the pri-
`mary tumor specimen or subsequently obtained
`specimens, or to have unknownreceptorstatus. Pa-
`tients with prior chemotherapy wereeligible only if
`such chemotherapy had been given as adjuvant
`treatment and was terminated at least a year prior
`to study. Patients were not permitted to have re-
`ceived any type of systemic endocrine therapy
`either with adjuvant intent or as treatment for
`metastatic disease.
`Exclusion criteria included a performancestatus
`of grade four, previous malignancy of other organs
`except adequately treated in situ carcinoma of the
`cervix uteri or basal or squamouscell carcinoma of
`the skin, a history of retinopathy, a history of severe
`liver disease, acute severe infectious disease, cur-
`rent thrombophlebitis or thrombosis, a history of
`leukocytopenia or thrombocytopenia notrelated to
`previous chemotherapy, and elevated calcium lev-
`els. Patients whose only lesions were malignantef-
`fusions, lymphangitis carcinomatosis, osteoblastic
`bonelesions, or lesion(s) that had been recently ir-
`radiated were not consideredeligible. Patients with
`inflammatory breast cancer or with brain metasta-
`ses were noteligible for study. Patients felt to be at
`very high risk because of rapid progression of dis-
`ease or extensive disease in liver or lung were not
`
`Droloxifene: a new antiestrogen
`
`85
`
`considered eligible for this trial of a hormonal
`agent.
`
`Patient assessment and follow up
`
`At entry, all patients were required to be staged
`with a chest radiograph, bone scan, radiographs of
`all suspicious bone scan lesions, an ultrasound or
`computer tomographic scan ofthe liver, liver func-
`tion tests including an SGOT, SGPT and y-GT,an
`LDH,alkaline phosphatase, serum glucose, biliru-
`bin, creatinine, BUN, and calcium as well as a he-
`moglobin, sodium, potassium level, white blood cell
`count with differential, and a platelet count. Target
`metastases were required to be determined and
`measured on entry, and remeasured every two
`months for the first six months, then every three
`months. The technical method for measuring the
`tumor was notto be changed duringthe trial. Com-
`plete restaging as described above was required ev-
`ery six months.
`
`Response assessment
`
`The quality of the data obtained wasassured by two
`methods. All data in the case record forms were
`checked against hospital records of the patients by
`monitors of the two sponsoring companies, Rhone-
`Poulenc and Klinge Pharma. Furthermore, in re-
`sponse evaluation meetings, the investigators joint-
`ly evaluated the sources used for tumor response
`for all patients. As a result of these response eval-
`uation meetings, the tumorstatus ofall patients was
`peer reviewed.
`The criteria used to determine objective re-
`sponse (according to WHO) were:
`
`all
`remission: disappearance of
`a) complete
`Known disease determined by two observations
`made notless than four weeks apart;
`b) partial remission: a decrease of 50% or more in
`total tumor load; no lesion should have pro-
`gressed (2 25%); no new lesion(s) should have
`occurred;
`c) no change: decrease of less than 50% or in-
`
`
`
`
`
`AstraZeneca Exhibit 2047 p. 3
`
`
`
`
`
`
`
`
`
`Patients enrolled 124=(33.6)112 (30.4) 133 (36.0) 369 (100)
`
`
`
`
`
`
`Ineligible: 21=(18.8) 27~—(21.8) 29 (21.8) 77 (20.9)
`
`
`
`no UICClesions —
`6
`(5.4)
`4 16=(4.3)(3.2) (4.5)
`
`
`
`2
`(1.6)
`~
`5
`(4)
`neg. hormonereceptor status
`3
`(2.6)
`1
`(0.8)
`-
`1
`(0.3)
`no proof ofbreast cancer
`-
`
`2=(1.6) 1 (0.8) 3 (0.8)
`
`
`inflammatory breast cancer
`-
`4
`(3.2)
`-
`4
`(1.1)
`high risk patients
`-
`
`
`- 2=(1.5) 3 (0.8)
`
`not postmenopausal
`1
`(0.9)
`prior palliative chemo
`~
`-
`3.
`(2.3)
`3
`(0.8)
`
`
`
`prior hormonal treatment 2~~(1.6)2 (1.8) 2 (1.5) 6 (1.6)
`
`
`
`adj chemonot terminated at least one year prior
`(0.3)
`1
`(0.8)
`1
`-
`to study enrollment
`wot
`(1.6)
`6
`(0.8)
`1
`2
`incomplete baseline staging
`(0.3)
`1
`(0.8)
`1
`-
`~
`history of retinopathy
`(0.5)
`2
`(1.5)
`2
`-
`~
`hypercalcemia on entry
`
`
`
`- 2=(0.5)2 (1.5)
`previous other cancer
`~
`10 24=(6.5)(8.1) 8 (6.0)
`
`
`
`
`primary operable
`6
`(54)
`9
`(73)
`7
`(5.3)
`23
`(6.2)
`Notevaluable
`7
`(6.3)
`3.
`(24)
`1
`(0.8)
`8
`(2.1)
`early withdrawals
`4
`(3.5)
`1
`(08)
`1
`(0.8)
`4
`(1.7)
`early deaths
`2
`(1.8)
`3.
`(2.4)
`-
`3
`(0.8)
`lost to follow-up
`~
`(3.0)
`2
`(1.6)
`4
`7
`(1.9)
`target lesions not followed according to protocol
`1
`-
`1
`(0.8)
`1
`(0.3)
`insufficient treatment compliance
`~
`Open
`~
`
` - 1 (0.8) 1 (0.3)
`
`
`
`
`
`
`
`(4.6)
`
`
`
`(2.6)
`
`(0.9)
`
`86
`
`W Rauschning and KI Pritchard
`
`crease of less than 25% in the tumorsize; no
`new lesions;
`d) progressive disease: increase of the tumorsize
`by 25% or more, or appearance of new lesions.
`
`Results
`
`Patient characteristics
`
`Between June of 1988 and March of 1991, 369 pa-
`tients were randomized from 29 centres in Europe,
`9 centres in Canada and4 centres in Brazil. Of 369
`patients randomized, 268 are evaluable for re-
`sponse. Seventy seven were ineligible for reasons
`including primary operable breast cancer (24), no
`WHOevaluable lesion (16), prior hormonal ther-
`apy (6), incomplete baseline staging (6), negative
`hormonereceptorstatus (5), high risk status (4), in-
`
`flammatory breast cancer (3), not post-menopausal
`(3), palliative chemotherapy :(3), hypercalcemia
`(2), previous other cancer (2), no proof of breast
`cancer(1), adjuvant chemotherapy not terminated
`at least oneyearpriorto study enrollment(1), histo-
`ry of retinopathy (1). Twenty-three patients werein-
`evaluable,
`largely because early withdrawals or
`deaths, loss to follow-up or failure to repeat the in-
`vestigations required to determine response. Table
`1 illustrates the reasonsfor ineligibility and ineva-
`luability by treatment group, andin total.
`The distribution of demographic data and risk
`factors forall evaluable patients is displayed in Ta-
`ble 2. The three groupsdid not differ with respect to
`age distribution (mean = 64), disease-free interval
`or hormonereceptor status. If anything, patients
`with unfavourable prognostic factors wereslightly
`overrepresented in the higher dose groups,in that
`32% of the 20 mg group had an unfavourable prog-
`
`Table 1. Patients enrolled and evaluated.
`
`
`Patients
`
`Treatment group
`20 mg
`No.
`
`(%)
`
`40 mg
`100 mg
`No.
`(%)
`No.
`(%)
`
`
`Total
`No.
`
`(%)
`
`
`88=(71.0) 268=(72.6) 96 (72.2)
`
`
`
`Evaluable 84=(75.0)
`
`
`
`AstraZeneca Exhibit 2047 p. 4
`
`
`
`nosis (Possinger score > 7), while 39% in the 40 mg
`and 43% in the 100 mg group scored unfavourably
`(see Table 3 for calculation of Possingerscore) [25].
`Similarly, 33% of the 20 mg group hadvisceral me-
`tastases, while 40% of the 40 mg group and 45% of
`the 100 mg group had this generally acknowledged
`poor prognostic factor. These differences in the dis-
`tribution of prognostic factors were notstatistically
`significant.
`
`Response to droloxifene
`
`Responseratesfor all patients evaluated are shown
`on Table 4. Response rates (CR + PR) were 30% in
`the 20 mg group, 47% in the 40 mg group, and 44%
`in the 100 mg group. Median duration of response
`was 12 months in the 20 mg group, 15 monthsin the
`40 mg group, and 18 monthsin the 100 mg group.
`
`Droloxifene: a new antiestrogen
`
`87
`
`Whenthe three groups were compared,adjusting
`for prognostic factors using logistic regression, both
`the 40 mg (p = 0.02) and 100 mg (p = 0.04) groups
`had significantly higher response rates than the 20
`mg group. The pooled 40 and 100 mg groups also
`had a higher response rate than the 20 mg group (p
`= 0.01).
`Time to response wasrelatively short. In all three
`groups, 50% of the remissions were obtainedin the
`first two months(see Fig. 1). Patients who respon-
`ded later were generally those with bone metasta-
`ses. The median pain score of patients on study de-
`creased by nearly half within the first two weeks. (20
`mg group 1.4 (+0.8) to 0.7 (+0.9); 40 mg group 1.2 (+
`0.8) to 0.7 (+ 0.9); 100 mg group 1.6 (+ 0.8) to 0.9 (+
`0.9).
`Time to progression was 5.6 monthsin the 20 mg
`group, 8.3 months in the 40 mg group, and 6.4
`monthsin the 100 mg group(see Fig. 2). When ana-
`
`Table 2. Demographic data of patients evaluated for efficacy (n = 268), (percentage of patients in relation to size of treatment group).
`
`Treatment group
`Variable
`20 mg (n = 84)
`40 mg (n = 88)
`100 mg (n = 96)
`No.
`(%)
`No.
`(%)
`No.
`(%)
`
`
`Age (yrs, mean + std.dev.)
`Performance status (WHO)onentry
`Grade 0
`Grade I
`GradeII
`Grade III
`no data
`Disease free interval
`
`64.2
`
`49.2
`
`64.6
`
`49.1
`
`64.2
`
`19.9
`
`39
`35
`10
`0
`0
`
`(46.4)
`(41.7)
`(11.9)
`(0.0)
`(0.0)
`
`34
`45
`5
`3
`1
`
`(38.6)
`(51.1)
`(5.7)
`(3.4)
`(1.1)
`
`36
`38
`14
`6
`2
`
`(37.5)
`(39.6)
`(14.6)
`(6.3)
`(2.1)
`
`43
`41
`
`19
`65
`
`57
`27
`
`(51.2)
`(48.8)
`
`(22.6)
`(77.4)
`
`(67.9)
`(32.1)
`
`51
`37
`
`16
`72
`
`54
`34
`
`(58.0)
`(42.0)
`
`(18.2)
`(81.8)
`
`(61.4)
`(38.6)
`
`49
`47
`
`19
`77
`
`54
`42
`
`(51.0)
`(49.0)
`
`(19.8)
`(80.2)
`
`(56.3)
`(43.8)
`
`
`
`
`
`
`
`
`DW2years 34 (40.5) 34 (38.6) 43 (44.8)
`
`>2 years
`43
`(51.2)
`43
`(48.9)
`46
`(47.9)
`no data
`7
`(8.3)
`11
`(12.5)
`7
`(7.3)
`Receptor status (primary)*
`positive
`unknown
`Prior adjuvant chemotherapy
`Yes
`No
`Risk factor (Possinger Score)**
`0-6
`27
`Dominantsite of disease
`
`(27.1)
`26
`(31.8)
`28
`(29.8)
`25
`soft tissue
`(27.1)
`26
`(28.4)
`25
`(36.9)
`31
`bone
`
`
`
`
`
`
`28 (33.3) 35 (39.8) 44visceral (45.8)
`
`* p = 0.573, Chi’ test.
`** p = 0.112, Mantel Haenszeltest.
`
`
`
`AstraZeneca Exhibit 2047 p. 5
`
`
`
`
`
`
`AstraZeneca Exhibit 2047 p. 6
`
`88
`
`W Rauschning and KI Pritchard
`
`Table 3. Possinger Score.
`
`Prognostic factors
`Points
`
`
`Site
`Skin, Soft Tissue
`Bone, Effusion
`Lung:
`nodular
`
`solitary
`multiple
`Lymphangiosis carcinomatosa
`Liver
`Bone marrow carcinosis
`
`Receptor Status
`positive
`unknown
`
`1
`1
`
`3
`5
`6
`6
`4
`
`1
`2
`
`negative
`Relapse-free interval
`<2 years
`2 2 years
`
`
`3
`
`1
`
`<7 Points
`Favourable prognosis
`
`Unfavourable prognosis 27 Points
`
`lyzed using the log ranktest, the difference in time
`to progression was notsignificant (p = 0.20), but
`when adjusted for prognostic factors using Cox’s
`proportional hazards model,
`it achieved signifi-
`cance, with the 40 mg (p = 0.02) and the 100 mg (p=
`0.01) groups being significantly improved in com-
`parison to the 20 mg group. The 100 mg group was
`not, however,significantly different in comparison
`to the 40 mg group (p = 0.75).
`Thus, the results of multivariate analysis confirm
`
`that efficacy, as measured either by responserate or
`time to progression, washigherin the 40 and 100 mg
`dosage groups in comparison with the 20 mg group.
`Nostatistically significant difference between the
`40 and 100 mg dosage group could be detected.
`
`Toxicity
`
`Toxicity is reported for 369 patients enrolled in the
`study by March 31, 1992. All patients enrolled were
`includedin this toxicity evaluation, including those
`whowereineligible or inevaluable.
`Table 5 summarizes all clinical and laboratory
`toxicity affecting the various body systems, as an
`absolute number and as a percentage, listed by
`treatment group. The following symptoms were ob-
`served by more than 20% ofpatients: hot flashes: 20
`mg group: 32.1% (36 patients); 40 mg group: 32.3%
`(40 patients); 100 mg group: 28.6% (38 patients);
`lassitude: 20 mg group: 27.7% (31 patients); 40 mg
`group: 23.4% (29 patients); 100 mg group: 26.3%
`(35 patients); and nausea: 20 mg group: 22.3% (25
`patients); 40 mg group: 24.2%(30 patients); 100 mg
`group: 29.3% (39 patients). No difference in these
`or other adverse eventsin relation to dosage could
`be observed.
`Mild clinical adverse events occurred in 106 pa-
`tients, moderate events in 149 patients, severe
`events in 40 patients, and life-threatening adverse
`events in 5 of 369 patients. Table 6 summarizesse-
`vere or life-threatening events. Two life-threaten-
`
`Table 4. Best response of patients evaluated for efficacy (n = 268).
`
`
`Best response
`
`Treatment group
`20 mg (n = 84)
`No.
`(%)
`
`
`40 mg (n = 88)
`No.
`
`100 mg (n = 96)
`(%)
`No.
`(%)
`
`
`(11.5)
`11
`(8.0)
`7
`(6.0)
`5
`Complete response (CR)*
`(23.8)
`20
`Partial response (PR)
`(32.3)
`31
`(38.6)
`34
`(41.7)
`35
`No change (NC)
`(27.0)
`26
`(35.2)
`31
`(28.6)
`24
`Progression of disease (PD)
`(29.2)
`28
`(18.2)
`16
`(29.8)
`25
`CR/PR**
`(43.8)
`42
`(46.6)
`44
`(71.4)
`60
`CR/PR/NC***
`
`
`
`(81.8) 6872 (70.8)
`*
`p= 0.407, two-sided chi’ test for 3 x 2 tables.
`** bp = 0.055, two-sided chi’ test for 3 x 2 tables.
`p = 0.012, one-sided exact Fisher’s test for 2 x 2 tables; 20 mg group vs 40 mg group combined with the 100 mg group.
`*** D = 0.166, two-sided chi’ test for 3 x 2 tables.
`
`
`
`
`
`
`AstraZeneca Exhibit 2047 p. 7
`
`Droloxifene: a new antiestrogen
`
`89
`
`Time to Response
`Evaluable Patients
`with CR or PR
`
`(n = 108) 40
`
`
`
`Percent(byTreatmentGroup)
`
`50
`
`30
`
`20
`
`10
`
`0
`
`Treatment Group 1(20mg) —— 2(40mg)----- 3(100mg) - - -
`
`
`
`5
`
`10
`
`15
`
`20
`
`25
`
`Time from Entry in Study to Response (months)
`
`Fig. 1.
`
`Table 5. Patients with drug-associated adverse events by body system (percentage according to numberof patients in treatment group;
`more than one observation for each patient possible).
`
`Treatment group
`Adverse events (AE)
`20 mg (n = 101)
`40 mg (n = 108)
`100 mg (n = 122
`No.
`%
`No.
`%
`No.
`%
`
`
`Patients
`with clinical AE*
`with laboratory AE**
`Bodysystem,clinical AE
`Skin
`Center & periph nerv syst
`Autonom nervous system
`Special senses
`Psychiatric
`Gastrointestinal
`Liver andbiliary
`Heart rhythm
`Vascular
`Respiratory system
`Platelet, bleeding and clot
`Reproductive, female
`Bodyas a whole
`Body system, laboratory AE
`(6.0)
`8
`(5.6)
`7
`(0.9)
`1
`Liver andbiliary system
`(8.3)
`11
`(4.0)
`5
`(1.8)
`2
`Metabolic
`(2.3)
`3
`(0.0)
`0
`(0.0)
`0
`Platelet, bleeding andclot
`(1.5)
`2
`(0.0)
`0
`(0.0)
`0
`White bloodcell
`
`
`
`
`
`
`1 (0.9) 0 (0.0) 0Red bloodcell (0.0)
`
`(56.5)
`(8.1)
`
`(5.6)
`(7.3)
`(1.6)
`(0.0)
`(2.4)
`(22.6)
`(0.8)
`(0.8)
`(0.8)
`(0.8)
`(0.0)
`(12.9)
`(53.2)
`
`72
`18
`
`8
`9
`0
`1
`3
`25
`0
`0
`4
`1
`2
`9
`62
`
`(54.1)
`(13.5)
`
`(6.0)
`(6.8)
`(0.0)
`(0.8)
`(2.3)
`(18.8)
`(0.0)
`(0.0)
`(3.0)
`(0.8)
`(1.5)
`(6.8)
`(46.6)
`
`65
`4
`
`2
`8
`1
`0
`2
`25
`0
`1
`1
`0
`0
`15
`66
`
`(58.0)
`(3.6)
`
`(1.8)
`(7.1)
`(0.9)
`(0.0)
`(1.8)
`(22.3)
`(0.0)
`(0.9)
`(0.9)
`(0.0)
`(0.0)
`(13.4)
`(58.9)
`
`70
`10
`
`7
`9
`2
`0
`3
`28
`1
`1
`1
`1
`0
`16
`66
`
`* p=0.771, Fisher’s exact test for comparison for the 20 mg group vs 100 mg group one-sided.
`** p = 0.006, Fisher’s exact test one-sided: 20 mg vs 100 mg; p = 0.118 for 20 mg vs 40 mg.
`
`
`
`
`
`
`
`AstraZeneca Exhibit 2047 p. 8
`
`90
`
`W Rauschning and KI Pritchard
`
`Time to Progression
`(n = 268)
`
`1.0};
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`
`
`
`
`CumulativeSurvivalProportion
`
` Treatment Group 1(20mg) —— 2(40mg): ---- -3(100mg) - ~ -
`
`10
`
`15
`
`20
`
`25
`
`40
`
`45
`
`Time from Entry in Study to Disease Progression (months)
`
`Fig. 2.
`
`ing events involved pulmonary emboli; one oc-
`curred after an orthopaedic procedure, but there
`
`was no apparentcausefor the other, exceptfor the
`presence of metastatic disease, and the use of dro-
`
`Table 6. Patients with ‘severe’, ‘life threatening’ adverse events or ‘death’ (percentage according to numberof patients in treatment
`group; more than one observation for each patient possible).
`
`‘Treatment group
`Adverse events (AE)
`20 mg (n = 101)
`40 mg (n = 108)
`100 mg (n = 122)
`No.
`(%)
`No.
`(%)
`No.
`(%)
`
`
`Patients
`
`11
`1
`
`0
`0
`3
`0
`1
`0
`0
`1
`0
`1
`9
`0
`0
`
`(9.8)
`(0.9)
`
`(0.0)
`(0.0)
`(2.7)
`(0.0)
`(1.0)
`(0.0)
`(0.0)
`(0.9)
`(0.0)
`(0.9)
`(8.0)
`(0.0)
`(0.0)
`
`13
`6
`
`0
`1
`1
`0
`3
`1
`1
`0
`0
`2
`7
`0
`0
`
`with clinical AE
`with laboratory AE
`Body system,clinical AE
`Skin
`Musculoskeletal
`Central & peripheral nervous system
`Psychiatric
`Gastrointestinal
`Liver andbiliary system
`Metabolic
`Heart rhythm
`Vascular
`Respiratory system
`Body as a whole
`Resistance mechanism
`Platelet, bleeding and clot
`Body system, laboratory AE
`(4.5)
`6
`(1.6)
`2
`(0.0)
`0
`Liver andbiliary system
`(3.0)
`4
`(2.4)
`3
`(0.9)
`1
`Metabolic
`(0.8)
`1
`(1.6)
`2
`(0.0)
`0
`Red bloodcell
`
`
`
`
`
`
`Platelet, bleeding and clot (1.5) 0 (0.0) 0 (0.0) 2
`
`(10.5)
`(4.8)
`
`(0.0)
`(0.8)
`(0.8)
`(0.0)
`(2.8)
`(0.8)
`(0.8)
`(0.0)
`(0.0)
`(1.6)
`(5.6)
`(0.0)
`(0.0)
`
`20
`11
`
`4
`1
`3
`2
`6
`0
`0
`0
`2
`2
`7
`1
`2
`
`(15.0)
`(8.3)
`
`(3.0)
`(0.8)
`(2.3)
`(1.5)
`(4.1)
`(0.0)
`(0.0)
`(0.0)
`(1.5)
`(1.5)
`(5.3)
`(0.8)
`(1.5)
`
`
`
`
`
`
`AstraZeneca Exhibit 2047 p. 9
`
`Droloxifene: a new antiestrogen
`
`91
`
`Table 7. Median values of hormonesat entry into study, after 2 weeks, and the median value of the differences between the values on
`entry and the first visit (after 2 weeks).
`
`Treatment group
`Parameter
`20 mg (n = 66)
`40 mg (n = 74)
`100 mg (n = 80)
`p-value
`median
`median
`median
`Kruskal Wallis test
`
`
`SHBG’(nmol/l)
`
`on entry
`week 2
`difference
`
`61
`88
`17
`
`70
`99
`28
`
`64
`103
`27
`
`0.5626
`
`0.1792
`
`FSH? (U/ml)
`
`LH? (U/ml)
`
`63
`58
`63
`on entry
`50
`55
`59
`week 2
`~16
`-8
`-2
`difference
`20
`19
`17
`on entry
`16
`17
`18.5
`week 2
`0.0001
`46
`-3
`1
`difference
`
`
`0.7083
`
`0.0001
`0.7625
`
`Results of the statistical tests of the differences for being different from zero (Wilcoxon sign ranktest):
`' SHBG:20 mg group: p = 0.0001, 40 mg group: p = 0.0001, 100 mg group: p = 0.0001.
`? FSH:20 mg group: p = 0.0136, 40 mg group: p = 0.0001, 100 mg group: p = 0.0001.
`> LH:20 mg group: p = 0.0005, 40 mg group: p = 0.0001, 100 mg group: p = 0.0001.
`
`loxifene. The otherthreelife-threatening eventsin-
`volved infection (1), acute respiratory insufficiency
`(1), and hypoglycemic coma(1). It is unclear that
`any of these latter three events were related to dro-
`loxifene intake, as there were otherclinical expla-
`nations in each of the threecases.
`
`Hormonal effects of droloxifene
`
`Table 7 reports the median values of various hor-
`mone measurements at entry into study and after
`two weeks, as well as the median difference be-
`tween these values in patients in whom both tests
`
`Table 8. Serious adverse events in 1200 patients who have received droloxifene
`
`Event
`100 mg (n = 432) > 100 mg (n = 161)
`40 mg (n = 296)
`20 mg (n = 239)
`
`
`
`hypercalcemia
`pulmonary embolism
`suspected pulmonary embolism
`thrombophlebitis/phlebothrombosis’
`myocardial infarction
`oculomotoric paralysis
`retinal detachment
`atrial fibrillation
`
`Q*
`2Q*
`
`1*
`
`3*
`
`1-
`
`2-
`
`4*
`4*
`1*
`gx
`
`1*
`
`1:
`
`1*
`
`1*
`
`transient myocardialinsufficiency
`gastric cancer
`apoplexy
`headache with paresthesia
`thrombocytopenia”
`nausea, vomiting
`pneumonia, septicaemia
`death during pleural drainage
`anemia, worsening
`
`
`1*
`
`1*
`
`* Definitely, probably, or possibly related to droloxifene treatment.
`- Relationship with droloxifene treatmentnotlikely or cannot be established.
`“ Four of these patients are listed in addition as pulmonary embolism (suspected or diagnosed).
`» Patientis also listed as pulmonary embolism.
`
`
`
`92
`
`W Rauschning and KI Pritchard
`
`wereavailable. This table shows that sex hormone
`binding globulin (SHBG)wasinducedbyall dosag-
`es, more so in the 100 mg group (20 mg group: 35%
`induction; 40 mg group: 41% induction; 100 mg
`group: 65% induction). FSH appearedto beslightly
`suppressedin all three groups, with suppressionin-
`creasing with dosage. LH was suppressed only in
`the 40 and 100 mg groups, while there was no obvi-
`ous change in the 20 mg group.
`
`Information available from other studies
`
`information is available from
`Some additional
`otherstudies of droloxifene. Table 8 summarizesse-
`rious adverse events in more than 1200 women
`treated with droloxifene worldwide. Other report-
`ed symptomsin these 1200 women,in addition to
`those seen in this Phase IJtrial, included gastroin-
`testinal discomfort, fatigue, headache, vertigo, ano-
`rexia, diarrhea, diffuse pain (bone, gastrointestinal,
`general), depression, excema, and pruritus vulvae.
`All of these symptoms, however, were reported to
`be mild.
`In nine phase I/II studies whicharestill partially
`open, 86 patients have received 200 or more mg of
`droloxifene daily [26-29]. Single doses up to 500 mg
`and continuous therapy with up to 400 mg daily
`have been well tolerated. The maximum tolerated
`doseis, therefore, still undefined.
`To date, 10 patients have received cumulative
`doses of more than 75 grams of droloxifene. This
`would correspond to a duration of 40 mg of droloxi-
`fene daily for more than 5 years. One patient has
`received a total of 132.6 g of droloxifene, corre-
`sponding to 9 years of therapy at 40 mg daily. No
`new adverse events have been observedin these pa-
`tients and the incidence of droloxifene side effects
`did not appear different than at lower doses. There
`has been no hint of cumulative toxicity.
`Tumorflares have not been reported with drolox-
`ifene, with the exception of one case report from
`Japan, in which a patient receiving a single dose of
`320 mg of droloxifene suffered from tumor flare
`symptoms[30].
`
`Discussion
`
`Droloxifene has been shownto have substantialac-
`tivity in patients with advanced breast cancer. In
`this large phase Itrial, complete responses were
`seen in 9% of patients and partial responsesin 32%.
`It appeared that responses were more commonin
`patients receiving 40 or 100 mg daily, than in those
`receiving 20 mg. Similarly, time to progression ap-
`peared somewhatlongerin the 40 or 100 mg groups.
`Droloxifene appeared to act rapidly, with mostre-
`sponses occurring within two monthsof starting
`therapy. There was rapid pain reduction. Whether
`this is significantly different than that observed with
`tamoxifen or with other antiestrogens, remains to
`be shown in randomizedtrials.
`Clinical experiencein this large phase II trial, and
`from 800 other patients treated around the world,
`suggests that droloxifene is extremely well tolerat-
`ed. Thromboembolic complications occur, but do
`not appear more commonthanin patients receiving
`other antiestrogens. Once again their comparative
`frequency remainsto be established in randomized
`trials.
`The tumorflares described in patients receiving
`tamoxifen have generally not been observed with
`droloxifene, but this must be clarified in a rando-
`mized setting. Droloxifene has been shownto in-
`crease sex hormonebinding globulin (SHBG) and
`to reduce FSH and LH levels in these postmeno-
`pausal women,suggesting that it may have somees-
`trogenic effects in humans, similar to those seen
`with tamoxifen. Whether this will result in im-
`provedlipid profiles remains to be established.
`Some of droloxifene’s pre-clinical characteris-
`tics, such as its increased antiestrogenicity and im-
`proved performance against tumorcell lines, sug-
`gestthat it may havebetteractivity than someof the
`antiestrogens used today. Whether droloxifene
`could result in higher responserates or in more pro-
`longed response duration in women with breast
`canceris, however, uncertain. Certainly it has been
`very difficult to exceed the 30% responserate seen
`in unselected patients with metastatic breast cancer
`whenvirtually any hormonal approachis used. We
`will, therefore, have to await comparative studies
`with tamoxifen and other second line hormonal
`
`
`AstraZeneca Exhibit 2047 p. 10
`
`
`
`
`
`therapies to see exactly how the responserate, re-
`sponse duration, toxicity profile, and rapidity of ac-
`tion of droloxifene compare in a randomizedset-
`ting.
`It is intriguing to speculate, however, that wheth-
`er or not droloxifene shows improved performance
`in this setting, it may be usefulin situations in which
`reduced estrogenicity is desirable. For example,it
`appears not to transform cells in vitro or to produce
`any of the malignancies seen with tamoxifen or with
`other anti-estrogens in animal models.If it is, in-
`deed,less estrogenic in humans,it may haveless ca-
`pacity to produce thromboembolic complications.
`Theseissues, however, also remain to be establish-
`ed in randomizedtrials.
`Becauseof droloxifene’s rapid pharmacokinetics
`and short half-life, sequenced hormonal chemo-
`therapy using droloxifene also has considerable ap-
`peal. There have been attempts to arrest cell divi-
`sion with tamoxifen, stimulate the cells to divide
`with estrogen, then follow with cytotoxic chemo-
`therapy. It would seem more appropriate to attempt
`this type of manoeuverusing a more rapidly cleared
`agent such as droloxifene. Similarly, trials of alter-
`nating therapy with droloxifene and a progestation-
`al agent would seem more appropriate thantrials
`with tamoxifen and progestationals.
`In summary, droloxifene is an active and toler-
`able drug in the treatment of postmenopausal wom-
`en with metastatic breast cancer. Its response rate
`and short-term toxicity appear very similar to that
`of tamoxifen. Proposed studiesto clarify its degree
`of cross-resistance, response rate, rapidity of re-
`sponse, and response duration in relation to tamoxi-
`fen and to other standard endocrine therapies, as
`well