`
`Comparison of the Short-Term Biological Effects of 7a-[9-(4,4,5,5,5-
`
`pentafluoropentylsulfinyl)-nonyl] estra-1,3,5, (10)-triene-3,173-diol
`
`(Faslodex) versus Tamoxifen in Postmenopausal Women with
`Primary Breast Cancer1
`
`John F. Robertson,” Robert I. Nicholson, Nigel J. Bundred, Elizabeth Anderson, Zenon Rayter, Mitchell Dowsett,
`John N. Fox, Julia M. W. Gee, Alan Webster, Alan E. Wakeling, Charles Morris, and Michael Dixon
`Departrrent of Surgery, Nottingham City Hospital, Nottingham, United Kingdom [J. F. R.]; Tenovus Centre for Cancer Research, Welsh School of Pharnacy, Cardiff, \Na|es
`CF10 3XF [R.|.N., J.M.WG.]; Departrrent of Surgery, South Mancheéer University Hospital, Mancheéer M20 8LR, United Kingdom [N.J.B.]; Clinical Research
`Departnent, Christie Hospital National Health Service Trust, Mancheéer M20 4BX, United Kingdom [E.A.]; Departnent of Biochenistry, Royal Marsden Hospital, London
`$V\B 6JJ, United Kingdom [M. Do.]; Bristol Royal
`lnfirrrary, Bristol DS2 8HV\I, United Kingdom [Z. R.]; Castle Hill Hospital Cottingham, East Yorkshire HU16 5JQ, United
`Kingdom [J. N. F.]; A§raZeneca, Maoclaifield SK10 4TF, United Kingdom [A.W, A. E.W, C. M.]; and Departrrent of Surgery, Western General Hospital, Edinburgh EH4 2XU,
`Sootland [M. Di.]
`
`ABSTRACT
`
`INTRODUCTION
`
`701-[9—(4,4,5,5,5—Pentafluoropentylsulfinyl)—nonyl]estra—1,3,5, (10)—triene—
`3,17B—diol (ICI 182,780; Faslodex) is a novel steroidal antiestrogen. This
`partially blind, randomized, multicenter study compared the effects of
`single doses of long—acting ICI 182,780 with tamoxifen or placebo on
`estrogen receptor (ERoz) and progesterone receptor (PgR) content, Ki67
`proliferation—associated antigen labeling index (Ki67LI), and the apo-
`ptotic index in the primary breast tumors of postmenopausal women.
`Previously untreated patients (stages T1—T3; ER—positive or —unknown)
`were randomized and received a single i.m. dose of ICI 182,780 50 mg
`(n = 39), ICI 182,780 125 mg (n = 38), or ICI 182,780 250 mg (n = 44) or
`oral tamoxifen 20 mg daily (n = 36) or matching tamoxifen placebo
`(n = 43) for 14-21 days before tumor resection surgery with curative
`intent. The ER and PgR H—scores, together with the Ki67LI were deter-
`mined immunohistochemically in the matched pretreatment biopsy and
`the posttreatment surgical specimens. The apoptotic index was deter-
`mined by terminal deoxynucleotidyltransferase—mediated dUTP—biotin
`nick end labeling on the same samples. The effects of treatment on each of
`these parameters were compared using analysis of covariance. ICI 182,780
`produced dose—dependent reductions in ER and PgR H—scores and in the
`Ki67LI. The reductions in ER expression were statistically significant at
`all doses of ICI 182,780 compared with placebo (ICI 182,780 50 mg,
`P = 0.026; 125 mg, P = 0.006; 250 mg, P = 0.0001), and for ICI 182,780
`250 mg compared with tamoxifen (P = 0.024). For PgR H—score, there
`were statistically significant reductions after treatment with ICI 182,780
`125 mg (P = 0.003) and 250 mg (P = 0.0002) compared with placebo. In
`contrast, tamoxifen produced a significant increase in the PgR H—score
`relative to placebo, and consequently, all doses of ICI 182,780 produced
`PgR values that were significantly lower than those in the tamoxifen-
`treated group. All doses of ICI 182,780 significantly reduced Ki67LI
`values compared with placebo (ICI 182,780 50 mg, P = 0.046; 125 mg,
`P = 0.001; 250 mg, P = 0.0002), but there were no significant differences
`between any doses of ICI 182,780 and tamoxifen. ICI 182,780 did not alter
`the apoptotic index when compared with either placebo or tamoxifen.
`Short—term exposure to ICI 182,780 reduces the ERoz in breast tumor cells
`in a dose—dependent manner by down—regulating ER protein concentra-
`tion. The reductions in tumor PgR content by ICI 182,780 demonstrate
`that ICI 182,780, unlike tamoxifen, is devoid of estrogen—agonist activity.
`Reductions in tumor cell proliferative activity (as indicated by Ki67LI)
`show that ICI 182,780 is likely to have antitumor activity in the clinical
`setting.
`
`Received 11/9/00; accepted 7/25/01.
`The costs of publication of this article were defrayed in part by the payment of page
`charges. This article must therefore be hereby marked advertiserrent in accordance with
`18 U.S.C. Section 1734 solely to indicate this fact.
`1 This trial was sponsored and funded by AstraZeneca Pharmaceuticals (Macclesfield,
`United Kingdom).
`2To whom requests for reprints should be addressed, at Department of Surgery,
`Nottingham City Hospital, Hucknall Road, Nottingham, NG5 1PB, United Kingdom.
`
`Estrogens act as endocrine growth factors for at least one-third of
`breast cancers (1), and their effects are mediated via the ER3 pathway.
`Several approaches have been adopted to treat hormone-sensitive
`breast cancer. In premenopausal women these include reducing cir-
`culating estrogen by ovarian ablation or by inhibiting ovarian estrogen
`production. In postmenopausal women, the mainstays of therapy are
`the prevention of estrogen binding to its receptor using an antiestrogen
`or lowering estrogen levels with aromatase inhibitors. The antiestro-
`gen tamoxifen is the most widely used hormonal treatment for all
`stages of breast cancer (2). However, tamoxifen possesses partial
`agonist activity which has positive effects on bone (3, 4) and blood
`lipids (5), but which also has unwanted side effects, including in-
`creased endometrial proliferation (6), a small increase in the risk of
`endometrial cancer (7-9), tumor flare at the start of treatment (10),
`and tamoxifen-mediated tumor stimulation upon progression (11).
`Currently,
`there are two other clinically available nonsteroidal,
`mixed agonist/antagonist antiestrogens, toremifene, which is used in
`the treatment of breast cancer (12), and raloxifene, which is being
`used in the management of osteoporosis (13). These two agents,
`together with tamoxifen, comprise a group of compounds that are
`described as SERMS (14). No new SERM has yet provided significant
`advantages over tamoxifen in the treatment of breast cancer in terms
`of either efficacy or tolerability, and all SERMS discovered to date
`show some degree of partial agonist activity. Furthermore, cross-
`resistance between the new SERMS and tamoxifen may limit their
`application in advanced disease after adjuvant tamoxifen treatment
`(15). Despite the potential advantages of the partial agonist properties
`of the SERMS, a drug that acts as a nonagonist (pure) antiestrogen
`may be an important step toward improving breast cancer treatment
`(16).
`Fulvestrant (Faslodex), formerly known as ICI 182,780, is a novel
`estrogen antagonist that, unlike tamoxifen, has no estrogen—agonist
`activity (Fig. 1). Preclinical and early clinical studies (17-40) suggest
`that ICI 182,780 has biological effects indicative of improved clinical
`efficacy in the treatment of breast cancer. The main features are ER
`down-regulation, antiproliferative activity,
`induction of apoptosis,
`lack of cross-resistance with tamoxifen, and the absence of ER-
`agonist activity.
`ICI 182,780 has a binding affinity for the ER that is ~100 times
`
`3The abbreviations used are: ER, estrogen receptor(s); SERM, selective estrogen
`receptor modulator; ICI 182,780, 7a-[9-(4,4,5,5,5-pentafluoropentylsulfinyl)-nonyl]estra-
`1,3,5, (10)-triene-3,17[3-diol; PgR, progesterone receptor(s); Ki67LI, Ki67 proliferation-
`associated antigen labeling index; AI, apoptotic index; DAB, diaminobenzidine tetrahy-
`drochloride; ANCOVA, analysis of covariance; ICA, immunocytochemical assay; NRS,
`normal rabbit serum.
`6739
`
`Downloaded from cancerres.aacrjournalsorg on July 29, 2014. © 2001 American Association for Cancer Research.
`
`AstraZeneca Ex. 2030 p. 1
`Mylan Pharms. Inc. V. AstraZeneca AB IPR20l6-01325
`
`
`
`SHORT-TERM EFFECTS OF ICI 182,780 IN PRIMARY BREAST CANCER
`
`Tamoxifen
`NME2
`
`ICI 182,780
`
`on
`
`'(‘cH,)9so(cH,)3ci=2ci=s
`
`H0
`
`Q. 1
`
`Fig. 1. Chemical structures of the nonsteroidal SERM, tamoxifen, and of the novel
`nonagonist (pure) antiestrogen, ICI 182,780.
`
`greater than that of tamoxifen (17), and in animal models, ICI 182,780
`markedly attenuates the ability of the ER to activate or inhibit gene
`transcription (20 -22). Several different mechanisms may underlie this
`effect, including impaired dimerization, increased ER turnover, and
`disrupted nuclear localization (23-25). ICI 182,780 treatment blocks
`the uterotrophic effects of ER agonists (estrogens) and of partial
`agonists such as tamoxifen (26 -28) and raloxifene (29) and reduces
`ER levels in the tumors of women with primary breast cancer (30).
`Therefore, ICI 182,780 seems to act as an ER down-regulator, because
`it functionally blocks the ER and reduces cellular ER levels such that
`the receptor is rendered unavailable or unresponsive to estrogen or
`estrogen agonists.
`The PgR gene is an estrogen-regulated gene (34), so drugs with
`estrogenic activity will increase its expression. Accordingly, tamox-
`ifen has been shown to increase PgR levels (35), whereas initial work
`on primary breast tumors found that a short-acting formulation of ICI
`182,780 reduced PgR levels (30), suggesting that it is devoid of
`estrogen-agonist activity and may have a different mechanism of
`action to that of tamoxifen. Additional evidence that ICI 182,780 and
`tamoxifen have different underlying modes of action comes from
`studies showing that tamoxifen-resistant tumors remain sensitive to
`ICI 182,780 treatment in \/itro (18, 19), in vivo (36, 37), and in the
`clinic (38-40).
`ICI 182,780 has antiproliferative effects, as assessed by immuno-
`histochemical detection of the Ki67 proliferation-associated antigen
`(30-32). Previous small clinical studies have suggested that both
`tamoxifen and ICI 182,780 increase apoptosis in primary human
`breast cancer (33).
`The study reported here represents the first direct randomized
`comparison of the short-tenn biological effects of ICI 182,780 (50
`mg, 125 mg, or 250 mg as a single i.m. injection) with tamoxifen (20
`mg/day p.o. for 14-22 days) and tamoxifen placebo in women with
`primary breast cancer. It is also the first investigation of any dose-
`response effect of ICI 182,780 and the first time that the biological
`effects of the clinical trials formulation (250 mg) have been assessed.
`The end points of the trial were ERa (referred to as ER for the
`remainder of this paper) and PgR H-scores, Ki67LI, and the AI.
`
`itant therapy, demography, current medical conditions, hematology, and bio-
`chemistry screening.
`Patients were included if they were postmenopausal (>12 months since the
`last menstrual period and/or had castrate levels of follicle-stimulating hormone
`>40 IU/liter) and had a clinically staged, histologically confirmed T1, T2 or T3
`primary breast cancer. They had to be fit for surgery within 1 month and have
`a tumor large enough to provide sufficient biopsy samples. Patients were
`ER-positive or -unknown at entry to the trial. The study was approved by the
`Ethics Committees of all centers.
`Patients were not eligible for the study if they had evidence of metastatic
`disease or had received any prior treatment for their primary tumor. Other
`exclusion criteria were:
`(a) treatment with hormone replacement therapy
`within 4 weeks of starting the trial; (b) baseline hematology or clinical
`chemistry outside the normal range; (0) risk of human immunodeficiency virus,
`hepatitis B, or hepatitis C transmission; (d) history of disease affecting steroid
`metabolism;
`(e)
`bleeding
`diathesis
`or
`thrombocytopenia
`(platelets
`<l00 X 109/liter); or any other reason that could jeopardize the protocol.
`Treatment with drugs known to affect sex hormone status could not be
`commenced during the trial (e.g., ketoconazole or prednisolone), although the
`patient could continue to receive such drugs if they were being taken before the
`study and the patient’s hormone status was stable.
`Patients were randomized to one of the following treatments: single i.m.
`dose of ICI 182,780 50 mg (n = 40), 125 mg (n = 40), and 250 mg (n = 41);
`tamoxifen, 20 mg, once daily p.o. for 14-21 days (n = 37); or tamoxifen
`placebo, once daily p.o. for 14-21 days (n = 43). Patients were scheduled for
`tumor resection surgery with curative intent between day 15 and day 22 after
`the start of treatment. On the day of surgery, patients were reassessed for
`concomitant therapy, concomitant conditions, hematology, and biochemistry.
`All patients returned for postsurgical assessment on day 57.
`
`Tumor Sampling
`
`The Tru-cut/core biopsy taken at the first clinic attendance for diagnostic
`purposes was used as the prerandomization tumor sample. Where possible, a
`minimum of three cores was taken, sufficient to provide material for the three
`laboratories. The posttreatment specimen was obtained at definitive surgical
`resection. All of the tissue samples were fixed in 3.7% formalin immediately
`after removal, then embedded in paraffin wax for sectioning and subsequent
`analysis of biological markers.
`
`Drug Administration
`
`Long acting ICI 182,780 (AstraZeneca, Macclesfield, United Kingdom) was
`administered by i.m. injection into the gluteus maximus muscle. Patients were
`randomized to receive 50 mg of ICI 182,780 (1 ml), 125 mg of ICI 182,780
`(2.5 ml), or 250 mg of ICI 182,780 (5 ml). Tamoxifen was supplied as
`Nolvadex tablets containing 20 mg of tamoxifen (AstraZeneca) and adminis-
`tered at a dose of 20 mg/day. The tamoxifen placebo tablet (AstraZeneca)
`matched the 20 mg tamoxifen tablet. Both tamoxifen and tamoxifen placebo
`were administered p.o.
`
`Adverse Events Monitoring
`
`Adverse events (defined as the development of a new medical condition or
`the deterioration of a preexisting medical condition subsequent to or during
`exposure to the trial medications) were monitored throughout the study.
`Patients were followed up for adverse events for 57 days postdosing.
`
`Analysis of Tumor Marker Expression
`
`PATIENTS AND METHODS
`
`ER. ERa expression was assessed at the Tenovus Centre for Cancer
`Research, Cardiff Wales, on sections cut from the formalin-fixed, paraffin-
`embedded tissue specimens described above. All mounted sections were dried
`overnight at 60°C before being dewaxed and rehydrated to PBS (pH 7.2-7.4).
`Two hundred and one women with primary breast cancer participated in a
`Endogenous peroxidase activity was quenched by incubation in hydrogen
`multicenter, randomized, partially blinded study. The administration of tamox-
`peroxide (0.5% in methanol) for 10 min and then rinsing in running tap water
`ifen and tamoxifen placebo was double blind, and the administration of ICI
`for 5 min and in PBS for 5 min. Then sections were enzyme-digested in a bath
`182,780 (at one of three doses) was open. Postmenopausal women with
`of 0.02% Pronase E (Sigma Chemical Co., Poole, United Kingdom) in PBS at
`histologically proven primary breast cancer awaiting tumor resection were
`37°C before being rinsed as described previously. To block the nonspecific
`recruited to the study from June 1997 to May 1999. Each woman gave written
`staining, a blocking reagent, comprising 20% normal swine serum (Dako Ltd.,
`informed consent and underwent an initial eligibility screen in the week before
`Glostrup, Denmark) in PBS was applied to the sections and then “tapped off’
`randomization. Prestudy assessments included past medical history, concom-
`6740
`
`Downloaded from cancerres.aacrjourna|s.org on July 29, 2014. © 2001 American Association for Cancer Research.
`
`AstraZeneca Ex. 2030 p. 2
`
`
`
`Characteristic
`ER status
`n (%)
`
`PgR status
`n (%)
`
`Positive
`Negative
`Unknown
`Positive
`Negative
`Unknown
`
`SHORT-TERM EFFECTS OF ICI 182,780 IN PRIMARY BREAST CANCER
`
`Table 1 ER and PgR éatus of tun'ors—per-protocol patients
`ICI 182,780
`ICI 182,780
`50 mg
`125 mg
`33 (86.8)
`34 (89.5)
`4 (10.5)
`1 (2.6)
`1 (2.6)
`3 (7.9)
`29 (76.3)
`29 (76.3)
`7 (18.4)
`5 (13.2)
`2 (5.3)
`4 (10.5)
`
`Placebo
`29 (69.0)
`8 (19.0)
`5 (11.9)
`28 (66.7)
`10 (23.8)
`4 (9.5)
`
`ICI 182,780
`250 mg
`32 (74.4)
`6 (14.0)
`5 (11.6)
`29 (67.4)
`9 (20.9)
`5 (11.6)
`
`Tamoxifen
`27 (81.8)
`4 (12.1)
`2 (6.1)
`21 (63.6)
`9 (27.3)
`3 (9.1)
`
`before incubation overnight at room temperature with the primary antibody
`(diluted 1:2), which was the rat antihuman ERa antibody (Clone H222)
`supplied in the ER-ICA kit by Abbott Laboratories (North Chicago, IL).
`Sections were washed in PBS (5 X 4 min) and then a secondary biotinylated
`sheep antirat
`immunoglobulin (Amersham Life Science Ltd., Amersham,
`United Kingdom) diluted 1:500 in 20% normal swine serum was applied for 60
`min. Sections were washed again in PBS (5 X 4 min) before the avidin-biotin-
`horseradish peroxidase complex (Dako Ltd.) diluted 1:120 in PBS was added
`for 60 min with additional washing afterward in PBS (5 X 4 min). Then the
`DAB chromogen was applied (as supplied in the Abbott ER-ICA kit) to the
`sections and left for 10 min before rinsing in distilled water (2 X 3 min).
`Staining was enhanced by treating the sections with 0.5% copper sulfate in
`0.85% sodium chloride for 8 min and rinsing in distilled water (2 X 3 min).
`The sections were counterstained with 0.5% methyl green for 5 min, washed
`in distilled water (2 X 3 min), dehydrated, cleared, and mounted for exami-
`nation by light microscopy.
`ERa immunopositivity appeared clearly as a brown nuclear signal in tumor
`epithelial cells against a background of green nuclear counterstain. Tumor
`epithelial cell ER content in the pre- and posttreatment specimens for each
`patient was assessed by the consensus oftwo people (J. M. W. G. and R. I. N.)
`using the dual viewing attachment of a light microscope. Overall staining was
`assessed at X10, and a representative area was viewed at X40 to assess the
`number of positive tumor cell nuclei and staining intensity. The percentages of
`positive tumor epithelial cells in each staining intensity category (i .e., negative
`—/—; very weak +/—; weak +; moderate ++; and strong +++) were
`recorded for each sample, and positive-control breast cancer samples of known
`ER positivity were included in every assay to monitor assay performance.
`Results were expressed as the ER H-score where: H-score = [(0.5 X %
`I/
`)
`I (1X% I) I (2X% I
`I) I (3X% I
`I
`I)].Ava1ueof>0imp1ies
`an ER-positive state with a range of 0—300.
`PgR Expression. Levels of PgR in sections from the same samples were
`also assessed by the Tenovus Centre for Cancer Research, Cardiff Wales. The
`assay procedure was similar to that used to detect ER, except that the primary
`anti-PgR antibody (Clone KD68) was that supplied by Abbott Laboratories in
`the PgR-ICA kit, as was the DAB chromogen.
`In this assay the primary
`antibody was diluted 1:4, and no enzyme retrieval was used. Results were
`expressed as the PgR H-score, using the same equation as that used to calculate
`the ER H-score.
`
`was quenched using hydrogen peroxide (0.2%) in methanol for 10 min. The
`sections were then rinsed in water and PBS and microwaved (800 W) in 10 mM
`citrate buffer (pH 6.0) at power 7 for 15 min after boiling point was reached.
`After cooling for 20 min, sections were washed in PBS and nonspecific
`binding was blocked with 10% NRS in 0.5% casein/PBS containing 4 drops/ml
`of the avidin block supplied by Vector Laboratories (Peterborough, United
`Kingdom) for 15 min. The primary antibody was then applied at a dilution of
`1:50 in 10% NRS/0.5% casein/PBS containing 4 drops/ml of biotin block
`(Vector Laboratories), and the sections were incubated for 80 min at room
`temperature. After washing in PBS (2 X 5 min), the secondary biotinylated
`rabbit antimouse antibody (DAKO E413; Dako Ltd., Ely, United Kingdom)
`was applied at a dilution of 1:300 in 10% NRS/0.5% casein/PBS for 40 min,
`and after washing in PBS (2 X 5 min), the avidin biotinylated enzyme complex
`reagent (Vectastain ABC Elite kit; Vector Laboratories) was applied for 40
`min. After the final PBS wash (2 X 5 min),
`incubation with the DAB
`chromogen (“SigmaFast” 3,3-diaminobenzidine tablet set; Sigma Chemical
`Co.-Aldrich Company, Poole, United Kingdom) was performed for 8 min at
`room temperature before a wash in distilled water. Samples were counter-
`stained with 20% hematoxylin for 3—5 min, dehydrated, cleared, and mounted
`for examination by light microscopy. Results were expressed as the Ki67LI
`(the percentage of positively stained nuclei calculated after counting at least
`1000 tumor cells).
`Al. The AI was measured using the terminal deoxynuc1eotidy1transferase-
`mediated dUTP-biotin nick end labeling assay at the Royal Marsden Hospital,
`London, United Kingdom. After dewaxing and rehydration to deionized water,
`endogenous peroxidases were quenched with hydrogen peroxide (1%) in PBS
`for 15 min and washing three times in deionized water. Then sections were
`digested in 0.5% pepsin (pH 2) for 30 min at 37°C in a humidified chamber.
`Digestion was terminated, and sections were rinsed for 1 min and washed five
`times for 5 min each in deionized water. Then sections were washed twice in
`
`Tris-buffered saline (pH 7.6) for 5 min and incubated for 1 h at 37°C in 100
`pol/slide of a reaction mixture containing 0.75 pl of terminal deoxynuc1eoti-
`dyltransferase, 0.50 pl of biotinylated 16dUTP, 10 pl of 50 mM cobalt
`chloride, and 20 pl of reaction buffer (1 M sodium cacodylate + 125 mM
`Tris-HC1 + 1.25 mg/ml BSA in deionized water). After washing twice in
`deionized water and three times in PBS, sections were incubated with horse-
`radish peroxidase-conjugated streptavidin (Dako Ltd.) diluted 1:4000 in
`PBS + 1% BSA + 0.5% Tween 20. Another three washes in PBS/Tween 20
`
`Ki67 Pr0]iferati0n—associated Antigen Expression. Ki67 antigen was
`assessed on sections of the pre- and posttreatment tissue specimens at the
`Christie Hospital, Manchester, United Kingdom, using the MIB-1 anti-Ki67
`antibody supplied by Coulter Electronics (Luton, United Kingdom). Briefly,
`slides were dewaxed and rehydrated to PBS (pH 7.6). Endogenous peroxidase
`
`preceded development with 0.05% DAB and 0.07% imidazole for 30 s and
`then 10 min of incubation with 100 pl of 1% hydrogen peroxide. Sections were
`washed in running tap water for 5 min and then immersed in 0.5% copper
`sulfate plus 0.9% sodium chloride in deionized water for 1 min. DAB was then
`inactivated with chloros and the sections were washed in running tap water,
`
`Characteristic
`Age (yr)
`
`Clinical disease staging
`n (%)
`
`Tumor grade at surgeryb
`n (%)
`
`Table 2 Dermgraphic characteristics of ER-positive per-protocol patients
`ICI 182,780
`ICI 182,780
`50 mg
`125 mg
`33
`34
`69.2
`68.7
`8.4
`7.3
`2 (6.1)
`5 (14.7)
`11 (33.3)
`10 (29.4)
`3 (9.1)
`1 (2.9)
`17 (51.5)
`18 (52.9)
`6 (18.2)
`8 (23.5)
`16 (48.5)
`15 (44.1)
`9 (27.3)
`9 (26.5)
`2 (6.1)
`2 (5.9)
`
`Placebo
`29
`65.9
`9.2
`5 (17.2)
`10 (34.5)
`1 (3.4)
`13 (44.8)
`6 (20.7)
`14 (48.3)
`7 (24.1)
`2 (6.9)
`
`ICI 182,780
`250 mg
`32
`66.1
`8.3
`2 (6.3)
`12 (37.5)
`0 (0.0)
`18 (56.3)
`9 (28.1)
`16 (50.0)
`6 (18.8)
`1 (3.1)
`
`Tamoxifen
`27
`68.7
`8.4
`6 (22.2)
`9 (33.3)
`1 (3.7)
`11 (40.7)
`6 (22.2)
`13 (48.1)
`7 (25.9)
`1 (3.7)
`
`n
`Mean
`SD
`T1
`T2
`T3
`Not T45
`G1
`G2
`G3
`GX
`aUnable to categorize, but definitely not T4.
`b G1, well-differentiated‘, G2, moderately differentiated; G3, poorly differentiated; GX, unassessable.
`6741
`
`Downloaded from cancerres.aacrjourna|s.org on July 29, 2014. © 2001 American Association for Cancer Research.
`
`AstraZeneca Ex. 2030 p. 3
`
`
`
`SHORT-TERM EFFECTS OF ICI 182,780 IN PRIMARY BREAST CANCER
`
`Pretreatment
`
`Post-treatment
`
`
`
`
`
`ICI182,780(2S0mg)
`
`Fig. 2. Comparison ofER expression in a biopsy
`sample taken pretreatment with that from a sample
`taken from the same tumor after treatment with ICI
`182,780 (250 mg; A), tamoxifen (B), and tamoxifen
`placebo (C). ER immunopositivity appears as a
`brovim nuclear signal against a background of the
`green nuclear counterstain. Photographs supplied
`by‘ R. r. N.
`
`Tamoxifen
`
`Placebo
`
`counterstained with hematoxylin in blued tap water (30 s), dehydrated, cleared,
`and mounted for examination by light microscopy. Results were expressed as
`the percentage of apoptotic cells in 3000 tumor cells.
`
`Statistical Analysis
`
`addition, any patients in the per-protocol population with a missing
`value for a tumor tissue marker were also excluded from the analysis
`for that particular marker. The ANCOVA allowed an overall assess-
`ment of differences between each dose of ICI 182,780 and tamoxifen
`and each dose of ICI 182,780 and placebo. A test for overall treatment
`
`Overall treatment effect P = 0.0003
`P = 0.049
`P = 0.0001
`P = o.ooos
`
`‘
`
`‘
`
`120
`
`100
`
`This trial was an exploratory trial, so the minimum power required
`for statistical testing was set at 80%. The four end points (surrogate
`tumor tissue markers) were considered equally important, so all were
`classed as primary end points. The secondary end points were toler-
`ability and pharmacokinetic data (phamracokinetic data are not pre-
`sented in this paper). This “per protocol” analysis included only those
`patients who received the full course of treatment, completed the end
`of treatment assessment for the primary end point, and had no signif-
`icant protocol deviations or violations. All analyses were carried out
`by the Biometrics Group, AstraZeneca.
`It was calculated that ~30 patients/group were needed to detect the
`following differences between ICI 182,780 and the comparator with
`80% power using a two-sided significance level of 5%: 0.3 for ER
`H-score‘, 0.4 for PgR H-score‘, 4.5 for Ki67', and 0.2 for apoptosis. To
`allow for ER-/PgR-negative tumors, a total of 201 patients were
`recruited and ~40 were randomized to each treatment group.
`The primary end point data were assessed statistically using
`ANCOVA according to treatment received with tenns included in the
`250 mg
`125 mg
`50 mg
`model for treatment, center, and the baseline tumor marker value.
`ICI 182,780
`iCi 182,780
`ICI 182,780
`(n = 32)
`(n = 32)
`(n = 31)
`Patients in the per-protocol population who were ER-negative were
`excluded from the analysis of ER, Ki67, and AI, and patients who
`Fig. 3. Posttreatrnent mean ER H-scores after a single i.m. injection of 50 mg, 125 mg,
`or 250 mg of ICI 182,780 or Tamoxifen 20 mg once daily p.o. or tamoxifen placebo.
`were PgR-negative were excluded from the analysis of PgR. In
`6742
`
`5 80
`I.I.l
`fl)V‘
`+r
`
`E(
`
`BOE
`
`60
`40
`
`20
`
`&
`
`§
`
`E
`
`
`
`
`
`Placebo ‘
`(n = 29)
`
`Tamoxifen
`(in = 25)
`
`Downloaded from ‘cancerr?e$.a:ac.ije«umalsrarg on July 29, 2014. © 2001 American Association for Cancer Research.
`
`Astrazeneca Ex. 2030 p. 4
`
`
`
`n
`Pretreatment mean H-score
`Percentage change (posttreatment)
`Overall treatment effect
`Treatment difference vs placebo (95% CI)
`
`SHORT-TERM EFFECTS OF ICI 182,780 IN PRIMARY BREAST CANCER
`
`Table 3 Smrrary of reailts for ER H-score
`ICI 182,780
`ICI 182,780
`50 mg
`125 mg
`31
`32
`136
`124
`—39
`—50
`
`Placebo
`29
`125
`—13
`P : 0.0003
`NAE
`
`ICI 182,780
`250 mg
`32
`113
`—59
`
`Tamoxifen
`25
`123
`—36
`
`—30 (—57, —4)
`P : 0.0255
`—2 (—29, 26)
`P : 0.8955
`
`—47 (—74, —21)
`P : 0.0006
`—19 (—46, 9)
`P : 0.1833
`
`—60 (—86, —34)
`P : 0.0001
`—32 (—59, —4)
`P : 0.0239
`
`—29 (—57, —0.2)
`P : 0.0485
`NA
`
`Treatment difference vs tamoxifen (95% CI)
`
`29 (0.2, 57)
`P : 0.0485
`
`ENA, not applicable.
`
`Overall treatment effect P = 0.0001
`P = 0.0090
`P = 0.0001
`P = 0.0001
`P = 0.0001
`
`P = 0.0002
`
`100
`
`80
`
`Mean:1:
`
`
`
`1SEM
`
`0
`
`Placebo
`(n = 28)
`
`250 mg
`125 mg
`50 mg
`ICI 182,780
`ICI 182,780
`ICI 182.780
`(n = 29)
`(n = 29)
`(n = 29)
`Fig. 4. Posttreatment mean PgR H-scores after a single i.m. injection of 50 mg, 125 mg,
`or 250 mg of ICI 182,780 or tamoxifen 20 mg once daily p.o. or tamoxifen placebo.
`
`Tamoxifen
`(n = 21)
`
`effect was undertaken. If this was significant at the 5% level, then the
`following pairwise comparisons were made:
`ICI 182,780 50 mg
`versus placebo; ICI 182,780 125 mg versus placebo; ICI 182,780 250
`mg versus placebo and ICI 182,780 50 mg versus tamoxifen; ICI
`182,780 125 mg versus tamoxifen; and ICI 182,780 250 mg versus
`tamoxifen. A supplementary comparison of tamoxifen versus placebo
`was also undertaken. For ER and PgR; the comparisons are presented
`as treatment differences with 95% confidence intervals. The mean
`
`change from baseline was also calculated for each treatment group
`and expressed as a percentage of the baseline mean value. For both
`Ki67LI and AI; the data showed evidence of nonnormality, so all
`values were log- (base e) transformed for the ANCOVA analysis, and
`the comparisons are, therefore, presented as treatment ratios with 95%
`confidence intervals. In addition, the median change from baseline
`was calculated for each treatment group and expressed as a percentage
`of the baseline median value. Plots of means : 1 SE by treatment
`group for each end point are also presented.
`
`RESULTS
`
`Patient Characteristics. A total of 201 postmenopausal women
`(mean age, 67.6 years; range: 48-86 years) were randomized into the
`trial, and 190 completed the trial. One patient did not take any trial
`treatment, and 10 patients withdrew from the trial. The withdrawal
`rates were similar for the ICI 182,780 groups (1/treatment group) but
`four patients withdrew from the tamoxifen treatment group and three
`from the tamoxifen placebo group. Of those patients in the per-
`protocol population, 155 were ER-positive. Groups were well bal-
`anced with respect to age, disease stage, and tumor grade at surgery.
`The ER and PgR status of the tumors at study entry are given in Table
`1. The demographic characteristics of the ER-positive per-protocol
`patients in the five treatment groups are summarized in Table 2.
`ER Expression. Treatment of ER-positive tumors with ICI
`182,780 resulted in a marked reduction of nuclear ER content that
`could easily be seen under the light microscope (Fig. 2). This was
`confirmed by statistical analysis of the ER H-score; which showed a
`significant overall treatment effect (P = 0.0003). The posttreatment
`mean ER H-scores are shown in Fig. 3, and the summary of results are
`shown in Table 3. ICI 182,780 produced a dose-dependent reduction
`in the ER H-scores, and all doses significantly reduced the ER H-score
`compared with placebo. The reduction in ER H-scores seen at the
`lower doses of ICI 182,780 (50 mg and 125 mg) were not statistically
`significantly different from those caused by tamoxifen, although the
`comparison between the 250-mg dose of ICI 182,780 and tamoxifen
`did reach significance (P = 0.0239).
`PgR Expression. Analysis of the PgR H-scores showed a signif-
`icant overall treatment effect (P = 0.0001). Posttreatment mean PgR
`H-scores are shown in Fig. 4, and the summary of results is shown in
`Table 4. There was a dose-dependent reduction in PgR H-score with
`ICI 182,780, with the 125 mg and 250 mg doses of ICI 182,780
`producing significantly greater reductions in PgR H-score than pla-
`cebo. Tamoxifen caused a significant increase in PgR H-score com-
`pared with placebo; consequently, each dose of ICI 182,780 resulted
`in a PgR H-score that was significantly lower than that of tamoxifen.
`Ki67LI. Analysis of the Ki67LI showed a significant overall treat-
`ment effect (P = 0.0029). The posttreatment mean Ki67LIs are shown
`in Fig. 5 and the summary of results are shown in Table 5. ICI 182,780
`
`Table 4 Smrrary of results for PgR H-score
`ICI 182,780
`50 mg
`29
`47
`—12
`
`ICI 182,780
`125 mg
`29
`28
`—52
`
`ICI 182,780
`250 mg
`29
`33
`—67
`
`n
`Pretreatment mean H-score
`Percentage change (posttreatment)
`Overall treatment effect
`Treatment difference VS. placebo (95% CI)
`
`Placebo
`28
`30
`+43
`P : 0.0001
`NAE
`
`Treatment difference VS. tamoxifen (95% CI)
`
`—27 (—47, —7)
`P : 0.0090
`
`ENA, not applicable.
`
`—14 (—32, 5)
`P : 0.1455
`—40 (—60, —21)
`P : 0.0001
`
`—28 (—46, —10)
`P : 0.0030
`—55 (—75, —34)
`P : 0.0001
`
`—35 (—53, —17)
`P : 0.0002
`—62 (—82, —42)
`P : 0.0001
`
`6743
`
`Tamoxifen
`21
`49
`+63
`
`27 (7, 47)
`P : 0.0090
`NA
`
`Downloaded from cancerres.aacrjourna|s.org on July 29, 2014. © 2001 American Association for Cancer Research.
`
`AstraZeneca Ex. 2030 p. 5
`
`
`
`SHORT-TERM EFFECTS OF ICI 182,780 IN PRIMARY BREAST CANCER
`
`produced dose-dependent reductions in the Ki67LI such that the
`Ki67LI at each dose of ICI 182,780 was significantly lower than that
`in the placebo group. There were no significant differences in Ki67
`labeling between tamoxifen and any dose of ICI 182,780.
`AI. There was no statistically significant overall treatment effect
`on the AI (P = 0.2382). There was no difference in AI between any
`dose of ICI 182,780 and tamoxifen compared with control. Posttreat-
`ment mean values for apoptosis are shown in Fig. 6, and the summary
`of results are shown in Table 6.
`
`Drug Tolerability. In general, all of the drugs were well tolerated.
`The most commonly reported adverse event were those relating to
`breast surgery.
`
`DISCUSSION
`
`This study provides the first direct comparison of ICI 182,780 with
`tamoxifen and placebo on the biological end points of breast tumor
`ERot and PgR content, Ki67 labeling, and AIs in primary breast
`cancer.
`
`Previous preclinical and early clinical studies have indicated that
`ICI 182,780 is an estrogen antagon