`Breast Cancer: Double-Blind Randomized Trial Showing
`a Dose Efi‘ect and Improved Efficacy and Tolerability
`Compared With Megestrol Acetate
`
`By P. Dombernowsky, |. Smith, G. Falkson, R. Leonard, L. Panasci, J. Bellmunl, W. Bezwoda, G. Gardin, A. Gudgeon,
`M. Morgan, A. Fornasiero, W. Hollmann, J. Michel, T. Hatschek, T. Tiabbes, H.A. Chaudri, U. Hornberger, and RF. Trunel
`
`Paras : To compare two doses of letrozole and
`megestrol acetate (MA) as second-line therapy in post-
`menopausal women with advanced breast cancer previ-
`ously treated with antiestrogens.
`Patients and Methods: Five hundred fifty-one pa-
`tients with locally advanced, locoregionally recurrent or
`metastatic breast cancer were randomly assigned to
`receive letrozole 2.5 mg (n = 174), letrozole 0.5 mg
`(n = 188), or MA 160 mg (n = 189) once daily in a
`double-blind, multicenter trial. Data were analyzed for
`tumor response and safety variables up to 33 months of
`follow-up evaluation and for survival up to 45 months.
`Results: Letrozole 2.5 mg produced a significantly
`higher overall obiective response rate (24%) compared
`with MA (16%; logistic regression, P = .04) or Ietroxole
`0.5 mg (13%; P = .004). Duration of obiective response
`was significantly longer for letrozole 2.5 mg compared
`with MA (Cox regression, P = .02). Letrozole 2.5 mg
`
`was significantly superior to MA and letrozole 0.5 mg in
`time to treatment failure (P = .04 and P = .002, respec-
`tively). For time to progression, letrozole 2.5 mg was
`superior to letrozole 0.5 mg (P = .02), but not to MA
`(P = .07). There was a significant dose effect in overall
`survival in favor of letrozole 2.5 mg (P = .03) compared
`with letrozole 0.5 mg. letrozole was significantly better
`tolerated than MA with respect to serious adverse expe-
`riences, discontinuation due to poor tolerability, cardio-
`vascular side effects, and weight gain.
`Conclusion: The data show letrozole 2.5 mg once
`daily to be more effective and better tolerated than MA
`in the treatment of postmenopausal women with ad-
`vanced breast cancer previously treated with antiestro-
`gens.
`J Clin Oncol 16:453-461. © 1998 by American Society
`of Clinical Oncology.
`
`NDOCRINE THERAPY is an important option in the
`treatment of advanced breast cancer, with tamoxifen
`the most widely used first—line drug. Approximately 40% of
`patients who relapse after tamoxifen may achieve further
`clinically useful tumor control with second-line therapy},2
`but the optimum choice of second-line treatment has not yet
`been defined. Progestins and aromatase inhibitors (AIs) are
`both commonly used; progestins (megestrol acetate [MA]
`and medroxyprogesterone acetate) can cause edema, weight
`gain, vaginal bleeding, hypertension, and thromboembolic
`problems,2 while aminoglutethimide, until recently the most
`commonly used AI, can cause rash, drowsiness, and leth-
`argy.2
`Lately, more selective AIs3'8 have been developed, includ-
`ing letrozole (CGS 20267), a new orally active, potent,
`highly selective, nonsteroidal competitive inhibitor of the
`aromatase enzyme.9 It has been reported to be approxi-
`mately 10,000 times as potent as aminoglutethimide in vivo,
`with no evidence of inhibition of progesterone or corticoste-
`rone synthesis at doses required to inhibit estrogen synthe—
`sis.10 In animal models, it has been shown to lead to almost
`complete regression of estrogen—dependent dimethyl—
`benzanthracene (DMBA)-induced mammary tumors.10 Phase
`I studies have shown letrozole to be effective in suppressing
`estrone, estradiol, and estrone sulfate by more than 75% to
`80% at doses of 0.1 mg to 5 mg/d, with no clinically relevant
`
`effects on other hormones of the endocrine system (includ-
`ing glucocorticoids, mineralocorticoids, and thyroid hor-
`mones).11'14 Objective tumor response rates of approxi—
`mately 25% have been obtained in postmenopausal patients
`with advanced breast cancer after failure of previous therapy.
`Tolerability was excellent, with minimal side effects.”'16
`This study reports the results of a multicenter, interna-
`tional, double—blind, randomized trial to compare the effi-
`cacy and safety of two doses of letrozole versus MA as
`second-line therapy in the treatment of women with ad-
`vanced breast cancer previously treated with an antiestro-
`gen. The dose of letrozole 0.5 mg was selected because it
`was the lowest dose to achieve maximal estrogen suppres-
`sion. The dose of 2.5 mg was chosen because it was still
`selective and well tolerated and could perhaps achieve a
`higher degree of aromatase inhibition at the level of the
`tumor. The two doses differ by a factor of five to ensure that
`a dose effect, if present, could be detected.17
`
`From the Letrozole International Trial Group (AR/3C2).
`Submitted April 14, 1997; accepted September 25, 1997.
`Supported by a grantfrom Novartis Pharma AG.
`Address reprint requests to RF Trunet, MD, Novartis Pharma SA,
`92506 Rueil—Malmaison Cedex, France.
`© 1998 by American Society of Clinical Oncology.
`0732-183X/98/1602-0008$3. 00/0
`
`Journal of Clinical Oncology, Vol 16, No 2 (February), 1998: pp 453-461
`
`453
`
`Downloaded from ascopubs.0rg 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 2139 p. 1
`InnoPharma Licensing LLC V. AstraZeneca AB IPR2017-00900
`
`
`
`454
`
`Patients
`
`PATIENTS AND METHODS
`
`Postmenopausal women with histologically or cytologically con-
`firmed breast cancer and with measurable or assessable locally ad-
`vanced or locoregionally recurrent or metastatic disease were eligible
`for the study. Postmenopausal status was defined by no spontaneous
`menses for at least 5 years; spontaneous menses within the past 5 years,
`but amenorrheic (eg, spontaneous or secondary to chemotherapy or
`hysterectomy) for at least 12 months, and luteinizing hormone (LH) and
`follicle-stimulating hormone (FSH) levels greater than 40 IU/L (or
`according to the definition of the postmenopausal normal range of the
`laboratory involved); bilateral oophorectomy; or radiation castration
`and amenorrheic for at least 3 months. Patients had to have tumors with
`positive or unknown estrogen or progesterone receptor status. Patients
`were regarded as estrogen or progesterone receptor-positive if any assay
`(cytochemical, immunochernical, irnmunohistochemical, or radioimmu-
`noassay) of primary or secondary tumor tissue was positive. Patients
`were regarded as receptor unknown if no assay was known to be
`positive or negative. Patients with estrogen receptor-negative, but
`progesterone receptor-positive status, or vice versa, were considered as
`receptor-positive and were included in the trial. Patients were required
`to have failed to respond to previous antiestrogen therapy either by
`relapsing on adjuvant therapy (given for 26 months) or within 12
`months of stopping treatment, or by progressing on first-line antiestro-
`gen treatment for metastatic disease. Previous treatment with chemother-
`apy as neoadjuvant or adjuvant treatment or for advanced disease (no
`more than one regimen) was allowed. Other inclusion criteria were a
`World Health Organization (WHO) performance status 5 2,
`total
`bilirubin level less than 1.5 and/or transaminases less than 2.6 times
`upper limit of normal, creatinine concentration less than 1.5 times upper
`limit of normal, WBC count 2 3.0 X 109/L, neutrophil count 2 1.5 X
`109/L, hemoglobin Z 10 g/dL, platelets 2 75 X 109/L, and total serum
`calcium level less than 2.75 mmol/L.
`Exclusion criteria were rapidly progressive disease (CNS involve-
`ment, or diffuse lymphangitis carcinomatosa of the lung); inflammatory
`breast cancer; disease limited to hilar enlargement, pleural effusion, or
`ascites; hepatic metastases that involved more than one third of the
`liver; concurrent or previous malignant disease (other than contralateral
`breast carcinoma, in situ carcinoma of the cervix treated by cone biopsy,
`or adequately treated basal or squamous cell skin carcinoma); history of
`deep vein thrombosis or pulmonary embolism; uncontrolled cardiac
`disease or diabetes mellitus; adrenal insufficiency; or Cushing’s syn-
`drome.
`
`Adjuvant endocrine therapy other than antiestrogens, oophorectomy,
`or radiation castration, and previous first-line endocrine therapy other
`than antiestrogens for advanced disease, were not allowed. Systemic
`corticosteroids for more than 15 days,
`investigational drugs, and
`concomitant anticancer treatment, except for radiotherapy to areas not
`being evaluated, were not permitted during the trial. Bisphosphonates
`started in the 6 months before or during the trial were not allowed if
`bone metastases were the sole manifestation of disease.
`All patients gave written consent to participate in the study, which
`was approved by the relevant
`local ethical review board and the
`Freiburger Ethik-Kommission International. The study was conducted
`according to Good Clinical Practice requirements.
`
`Study Design
`This randomized, double-blind, comparative trial was conducted in
`91 centers in 10 countries. Letrozole 2.5 mg or 0.5 mg once daily, or MA
`
`DOMBERNOWSKY ET AL
`
`160 mg once daily, were assigned equally, using the double-dummy
`technique, to patients who met the inclusion criteria.
`Randomization was stratified for each participating country. Within
`each country, treatments were assigned and packed 1:1:1 according to
`computer-generated permuted blocks of size 6 or size 3. Those countries
`that would include small centers had a block size of 3, and all countries
`had a block size of 3 for “reserve” supplies (ie, after enrollment of the
`planned number of patients had been completed in the center, additional
`patients could be enrolled following the reserve supplies). Each patient
`enrolled was assigned the treatment pack with the lowest available
`randomization number at a center.
`Patients were seen for tumor evaluation before the start of trial
`treatment and every 3 months during the trial. All patients were enrolled
`over 18 months from March 1993 to September 1994. Patients were
`monitored for tumor response and safety variables for up to 33 months
`(median, 25.5 months) and for up to 45 months for survival (median,
`18 to 20).
`At each visit, superficial or palpable lesions were measured, chest
`x-ray was performed, and severity of pain, performance status, and
`adverse experiences were recorded. In addition, pulse rate, blood
`pressure, weight, and routine hematology and biochemistry parameters
`were measured. ECG was performed at baseline and at 3 and 12 months
`(or when the patient discontinued the study).
`Bone scans and/or skeletal surveys, liver ultrasonograms, or com-
`puted tomographic (CT) scans were performed at baseline and at 6
`months and repeated every 3 months if positive at baseline. A
`quality-of—life questionnaire, the QLQ-C30 of the European Organiza-
`tion for Research and Treatment of Cancer (EORTC),18 was completed
`by patients at baseline and at each visit while on trial treatment.
`The primary efficacy end point was overall objective tumor response
`(complete response [CR] and partial response [PR]), assessed using
`International Union Against Cancer (UICC) criteria.19 Measurable
`disease was defined as the presence of metastatic lesions measurable in
`one or two dimensions using radiographic methods (x-ray, CT scan,
`ultrasonogram, or nuclear magnetic resonance) or physical methods (ie,
`palpation and measurement by calipers). Osteolytic bone lesions were
`considered as assessable, but nonmeasurable. Osteoblastic lesions or
`mixed blastic/lytic lesions at the same site and with no other site of
`metastatic disease, pleural effusion, hilar enlargement, and ascites were
`regarded as nonassessable, nonmeasurable disease. Hilar enlargement
`evaluated by CT scan was accepted as assessable, but nonmeasurable
`disease. Nonassessable, nonmeasurable lesions could not be monitored
`for response, but their continued presence would downgrade a CR to a
`PR in measurable lesions, and any increase in extent of such lesions or
`the appearance of new nonassessable, nonmeasurable lesions consti-
`tuted disease progression (PD). The overall tumor response of each
`patient was verified by independent blinded external peer review (two
`medical oncologists and one radiologist). Peer review assigned an
`assessment of overall tumor response: CR, PR, stabilization of disease
`(NC), PD, or not assessable (NA). CR, PR, and NC had to be confirmed
`on two occasions at least 4 weeks apart (generally, at the next scheduled
`3-month assessment). Other end points included time to progression
`('I'I‘P) of disease (interval between start of trial
`therapy [date of
`randomization] and the earliest diagnosis of PD validated by peer
`review whereby patients who died before any on—treatment tumor
`staging could be performed were counted as having progressed), time to
`treatment failure (TTF; interval between start of trial therapy [date of
`randomization] and the earliest event of PD, withdrawal of trial therapy
`for any reason, death for any cause, withdrawal of consent, or loss to
`follow-up), and time to death, as well as duration of objective response
`(CR5 and PRs). Duration of clinical benefit (CR, PR, and NC 2 6
`
`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 2139 p. 2
`
`
`
`Table 1. Patient Demographics and Baseline Data
`% of Patients
`Letrozole
`2.5 mg
`(n = 174)
`
`Letrozole
`0.5 mg
`[n = 188)
`
`Variable
`
`455
`
`MA
`[n = 189)
`
`64.6 t 10.5 63.6 i 9.1
`
`64 I 9.5
`
`20.2
`44.7
`35.1
`
`30.3
`30.9
`37.2
`
`62.8
`29.3
`6.4
`
`11.2
`25.5
`63.3
`
`36.7
`18.6
`44.7
`
`50.0
`38.3
`l l .7
`
`60.6
`21.8
`17.6
`
`34.6
`65.4
`
`21.3
`8.0
`38.8
`
`4.3
`
`19.0
`54.6
`26.4
`
`25.3
`29.9
`42.5
`
`58.6
`30.5
`8.6
`
`7.5
`31.6
`60.9
`
`32.8
`24.7
`42.5
`
`51 .1
`34.5
`14.4
`
`69.0
`20.7
`10.3
`
`32.8
`67.2
`
`19.0
`12.1
`33.9
`
`2.3
`
`16.4
`55.0
`28.6
`
`25.4
`31.7
`40.7
`
`51 .3
`37.0
`9.5
`
`11.6
`30.2
`58.2
`
`37.0
`21.7
`41.3
`
`46.0
`45.0
`9.0
`
`59.8
`21.7
`18.5
`
`32.3
`67.7
`
`21.2
`13.8
`34.9
`
`3.7
`
`LETROZOLE AND ADVANCED BREAST CANCER
`
`months) was subsequently added in accordance with current standards
`of reporting.
`Safety and tolerability were assessed using the National Cancer
`Institute common toxicity criteria.
`
`Statistical Methodology
`
`Assuming a significance level of 5% (two-tailed), the total sample
`size was calculated on the basis of overall objective tumor response as
`540 patients, with 146 patients per arm, increased by approximately
`23% to allow for each pair of treatments to be compared, and for a small
`proportion of patients being lost
`to follow-up without having an
`assessment of tumor response. The sample size had approximately 80%
`power to detect a 13% absolute superiority of one of the letrozole arms
`over MA, assuming an overall objective tumor response rate of 15% for
`MA.
`
`No interim analysis was planned. The core trial data were analyzed 9
`months after the end of enrollment. Six months later, the extension data
`for all variables were analyzed. Overall survival is updated every 6
`months until 90% of all enrolled patients have died. These subsequent
`analyses were performed mainly to confirm the core trial results and to
`obtain more mature estimates of time events. The extension data and
`most recent survival data are presented here.
`The primary end point (overall objective tumor response) was
`analyzed by logistic regression, presenting odds ratios with 95%
`confidence intervals (Cl). Time to progression, time to treatment failure,
`and overall survival were analyzed using Cox proportional hazards
`regression models presenting risk ratios with 95% CIs. Treatment
`comparisons were adjusted for the following baseline covariates: age,
`body-mass index, receptor status, dominant site of disease, extent of
`disease (anatomical locations), disease-free interval, prior antiestrogen
`therapy, response to prior antiestrogen therapy, previous chemotherapy,
`previous or concomitant bisphosphonates, and performance status.
`Significance levels were not adjusted for multiple comparisons or
`multiple end points. The main comparison of interest was letrozole 2.5
`mg versus MA.
`Duration of overall objective response (CR3 and PRs), duration of
`clinical benefit, 'ITP, TI‘F, and time to death were estimated by the
`Kaplan-Meier product-limit method. Treatment comparisons of dura-
`tion of response and duration of clinical benefit were not adjusted for
`baseline covariates, because of confounding with response.
`Analyses were performed on the intent-to-treat data set, which
`included all patients who received study medication. The analysis of
`safety data was descriptive, presenting 95 % CIs of differences between
`treatments. Where appropriate, Fisher’s exact test or a x2 test was used.
`
`Patients
`
`RESULTS
`
`A total of 551 patients took study medication and were
`included in the intent-to-treat analyses (letrozole 2.5 mg,
`n = 174', letrozole 0.5 mg, n = 188; MA, n = 189). Table 1
`shows that
`the three treatment arms were similar with
`
`respect to demographic and baseline characteristics.
`
`Efi‘icacy
`
`shows the Kaplan—Meier
`Figure 1
`Overall survival.
`estimates for time to death. The median time to death was
`
`Age, years (mean : SD)
`Age class, years
`5 55
`56-69
`2 70
`Dominant site‘
`Soft tissue
`Bone
`Viscera
`Anatomic sites‘
`l
`2
`3
`Disease-tree interval
`0 (stage IV)
`< 24 months
`2 24 months
`Overall receptor status
`ER+PgR+
`ER+ or PgR‘
`ER? PgR?
`Performance status
`0
`l
`2
`Prior chemotherapy
`None
`Adiuvant only
`Therapeutic : adiuvant
`Prior antiestrogen
`Adiuvant only
`Therapeutic 1' adiuvant
`Response to antiestrogen therapy
`CR + PR
`PD + unknown < 6 months'l'
`Adiuvant antiestrogens
`Biphosphonates before trial entry or
`concomitantly
`Radiotherapy to areas not being
`
`10.38.0evaluated lot tumor response 9.0
`Abbreviations: ER, estrogen receptor; PgR, progesterone receptor.
`‘Dominant site and number of anatomic sites do not sum to 100% as peer
`review considered that some patients did not have evidence of malignant
`disease at baseline.
`
`
`
`
`
`'l'PD or response is unknown, but antiestrogen was given For <6 months.
`
`654 days (21.5 months) for letrozole 0.5 mg, 655 days (21.5
`months) for MA, and 770 days (25.3 months) for letrozole
`2.5 mg. There was a significant dose effect in favor of
`letrozole 2.5 mg compared with letrozole 0.5 mg (P = .03).
`There was no significant difference between letrozole 0.5
`mg and MA (P = .4) or between letrozole 2.5 mg and MA
`(P = .15). These results are listed in Table 2.
`
`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 2139 p. 3
`
`
`
`456
`
`DOMBERNOWSKY ET AL
`
`
`
`r 1.0
`
`
`
`r
`
`:1
`
`1
`153
`150
`165
`e
`
`y
`
`a
`
`1
`131
`135
`136
`12
`
`15
`
`I
`105
`111
`103
`21
`1e
`Months
`
`l
`so
`as
`33
`24
`
`‘T
`
`27
`
`I
`as
`45
`42
`so
`
`y
`
`33
`
`|
`11
`17
`15
`as
`
`1
`0.5m
`2.5m
`MA
`:19
`
`.3
`.
`1’!
`g
`a
`~6
`,§
`:
`E" 0.3
`a"
`0-2
`
`0.1 -
`
`0.0 ~1
`n=188
`n=174
`was
`sum
`
`0.9
`
`0.8
`
`fl
`0.7 "
`:3.
`0.6 §
`g
`0.53
`
`0.4 ;
`a
`
`0.3
`0-2
`
`0.1
`
`0.0
`
`Fig 1. Kaplan-Meier estimates
`for time to death: letrozole 0.5 mg
`(- -l, letrozole 2.5 mg i-l, MA 1'
`' ‘l
`(Plot curtailed when <5 patients/
`treatmentarm). Deaths: 126 (67.0%)
`on letrozole 0.5 mg, 103 (59.2%) on
`letrozole 2.5 mg,
`on
`MA. Arrows indicate censored obser-
`vations.
`
`,,
`
`w
`Lifting“
`ialfiwiltihtJth:
`Lulu “‘3
`
`,,
`
`TTP. The median TTP was 5.6 months for letrozole 2.5
`
`mg, 5.5 months for MA, and 5.1 months for letrozole 0.5 mg
`(Fig 2). There was no significant difference between letro-
`zole 2.5 mg and MA (P = .07), although the strong trend
`
`Table 2. Summary of Analysis of Main End Points
`Treatment Comparison
`
`End Point
`Overall survival
`Risk ratio
`95% Cl
`P
`1TP
`Risk ratio
`95% Cl
`P
`111:
`Risk ratio
`95%Cl
`P
`
`Obiedive response (CR + PR)
`Odds ratio
`95% Cl
`P
`Duration of CR + PR‘
`Risk ratio
`95% Cl
`P
`Duration of CR + PR + NC 2 6‘
`Risk ratio
`95% Cl
`P
`
`0.5 mg:2.5 mg
`
`0.5 mgzMA
`
`2.5 mgtMA
`
`0.82
`1.12
`1.34
`1.02 to 1.76 087101.44 0.63 '1: 1.08
`.03
`.38
`.15
`
`0.80
`1 .04
`1 .35
`1.04 to 1.75 0.81 to 1.32 0.62 to 1.02
`.02
`.78
`.07
`
`0.77
`1.08
`1.47
`1.15t01.89 0.86 to 1.36 0.61 t00.99
`.002
`.52
`.04
`
`1 .82
`0.60
`0.42
`0.23 to 0.76 0.32 to 1.12 1.02 to 3.25
`.004
`.1 1
`.04
`
`0.42
`0.58
`1 .32
`0.54 to 3.23 0.25 to 1.34 0.20 to 0.86
`.54
`.19
`.02
`
`0.44
`0.52
`1.22
`0.68 to 2.19 0.30 to 0.89 0.26 to 0.73
`.50
`.01
`.001
`
`Abbreviations: 0.5 mg, 0.5 mg letrozole; 2.5 mg, 2.5 mg letrozole; MA, 160
`mg MA.
`'Unadiusted For baseline cavariates; all other comparisons adjusted accord-
`ing to the protocol and analysis plan.
`
`favored letrozole 2.5 mg (95% CI suggested as great as a
`38% reduction in the risk of progressing with letrozole 2.5
`mg to a 2% increase in the risk compared with MA), or
`between letrozole 0.5 mg and MA acetate (P = .8) in overall
`TTP. The treatment difference between letrozole 2.5 mg and
`letrozole 0.5 mg was significant (P = .02). These results are
`listed in Table 2.
`
`TTF. The median TTF was longest for letrozole 2.5 mg
`(5.1 months) compared with MA (3.9 months) and letrozole
`0.5 mg (3.2 months). Letrozole 2.5 mg was significantly
`superior to MA in overall TTF (P = .04). The dose effect in
`favor of letrozole 2.5 mg compared with letrozole 0.5 mg
`was also significant (P = .002). There was no significant
`difference between letrozole 0.5 mg and MA (Table 2).
`Objective response. Table 3 lists the objective response
`rates and associated C1s for the three treatment arms. There
`
`was a significant d0se effect in favor of letrozole 2.5 mg
`compared with letrozole 0.5 mg (P = .004), and a Signifi-
`cant superiority of letrozole 2.5 mg over MA (P = .04).
`There was no significant difference between letrozole 0.5
`mg and MA (Table 2).
`Duration of objective response was significantly longer
`for letrozole 2.5 mg compared with MA (P = .02). The
`median duration of objective response was not reached for
`letrozole 2.5 mg, compared with 17.9 months for MA, and
`18.2 months for letrozole 0.5 mg. Covariates predictive for
`overall objective response were dominant site of disease and
`performance status. Patients with soft tissue as the predomi-
`nant site of disease and good performance status (0 or 1)
`were more likely to respond to one of the three treatments
`than patients with bone or Visceral involvement and perfor-
`
`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 2139 p. 4
`
`
`
`LETROZOLE AND ADVANCED BREAST CANCER
`
`457
`
`
`
`
`
`.§ 0.8
`
`0.7-
`
`3g
`
`‘15.
`
`5 0.6-
`r.
`5 0.51
`E
`‘z
`0-41
`3
`.§ 0-3-
`‘5
`1
`g 0-2
`0.1
`
`Fig 2. Kaplan-Meier estimates
`far TI'P: letrozole 0.5 mg (- -), letro-
`zale 2.5 m9 1‘), MA (- -.)
`(Plot
`curtailed when <5 patients/hear
`mentarm). Disease progression: 128
`patients (68.1%) on letrozole 0.5
`mg, 120 (69.0%) on letrozole 2.5
`mg, 145 (76.7%) an MA. Arrows
`indicate censored observations.
`
`0.0
`
`
`
`n=188
`n=174
`"-109
`sun
`
`57
`74
`71
`a
`
`3
`
`9
`
`40
`52
`37
`12
`
`13
`31
`13
`1a
`
`15
`
`21
`
`0.1
`
`0.0
`
`0.5 mg
`2.5 mg
`MA
`24
`
`mance status 2 (data not shown). Patients with predomi-
`nantly soft tissue disease achieved an objective response rate
`of 48% with letrozole 2.5 mg and 40% with MA. For
`patients with predominantly bone metastases, the objective
`response rate was 15% with letrozole 2.5 mg and 10% with
`MA. For patients with visceral involvement, an objective
`response rate of 16% was observed with letrozole 2.5 mg
`and of 8% with MA. In addition, letrozole 2.5 mg appeared
`to be more efficacious than letrozole 0.5 mg or MA in
`patients who did not respond to first-line antiestrogen
`treatment (response rates, 29%, 7%, and 15%, respectively).
`Response rates, including patients with stable disease for
`26 months, were 27% for letrozole 0.5 mg, 35% for
`letrozole 2.5 mg, and 32% for MA. There was no significant
`difference among the three treatment arms. There was
`overall a significant difference between treatments in dura-
`
`Table 3. Overall Tumor Response
`Letrozole
`Letrozole
`MA
`0.5 mg
`2.5 mg
`in =188)
`(n =1741
`(n = 189)
`Variable
`No.
`%
`No.
`%
`No.
`%
`
`
`24
`
`41
`
`16.4
`23.6
`12.8
`Objective response
`11.1 -21 .7
`17.2-29.9
`8.0-17.6
`95% Cl
`4.2
`8
`6.9
`l 2
`3.2
`6
`CR
`12.2
`23
`16.7
`29
`9.6
`18
`PR
`15.3
`29
`10.9
`19
`14.4
`27
`NC
`56.1
`106
`53.4
`93
`55.9
`105
`PD
`
`
`
`
`
`
`17.0 21 12.1 2332Not assessable' 12.2
`
`31
`
`'Patients with incomplete documentation of tumor lesions, patients iudged by
`peer review not to have evidence of malignant disease, or patients who had PRs
`or NC that was not confirmed at least 4 weeks later.
`
`including stable disease 26 months
`tion of response,
`(log-rank, P = .001), with both doses of letrozole being
`superior to MA (Cox regression, P = .001 for 2.5 mg and
`P = .01 for 0.5 mg; Table 2). The median duration of
`clinical benefit was 18.1 months for letrozole 0.5 mg, 23.5
`months for letrozole 2.5 mg, and 14.5 months for MA.
`Subjective assessments.
`Slightly fewer patients (41%)
`in the letrozole—2.5 mg arm had a worsening of pain during
`treatment compared with the low-dose letrozole (50%) and
`MA (49%) arms. Fewer patients experienced a deterioration
`in WHO performance status in the group that received
`letrozole 2.5 mg group (41%) compared with MA (55%) 08,
`P = .01). No major differences in quality of life were
`apparent between treatments over time. The higher inci-
`dence of dyspnea in the MA arm (adverse experiences) was
`reflected in consistently higher dyspnea scores in the MA
`arm in the quality-of—life scale. Similarly, patients in the MA
`arm scored worse in physical functioning than in the 2.5-mg
`letrozole arm, with intermediate scores for patients on 0.5
`mg letrozole, which reflects the higher levels of worsening
`of performance status in patients on MA. The early nausea
`that occurs in some patients with letrozole was noted in the
`nausea and vomiting subscale in the quality-of-life instru-
`ment, but the symptom disappeared with time.
`
`Tolerability and Safety
`
`Table 4 lists the type and frequency of adverse experi-
`ences, irrespective of relationship to trial medication, re-
`ported during the follow-up period of up to 33 months. The
`lowest frequency of adverse experiences occurred in the arm
`
`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 2139 p. 5
`
`0OL‘o-wgssemoidunnwstuenedunmodmd
`
`.6.o\lD
`
`9co
`
`
`
`.0on
`
`.0N
`
`
`
`
`
`
`
`458
`
`DOMBERNOWSKY ET AL
`
`that received letrozole 0.5 mg (78% of patients) compared
`with 85% of patients on letrozole 2.5 mg and 90% on MA. In
`each body system, the percentage of patients who reported
`adverse experiences was similar in all treatment arms, with
`the exception of adverse experiences affecting the cardiovas—
`cular system, which were significantly (Fisher’s exact test,
`P = .02) higher for MA (20% of patients) compared with
`2.5 mg or 0.5 mg letrozole (10% and 11% of patients,
`respectively). The most frequently reported adverse experi-
`ences considered by the investigator to be drug-related (2.5
`mg and 0.5 mg letrozole v MA) were nausea (6% and 11% v
`4%), headache (7% and 6% v 5%), peripheral edema (6%
`and 3% v 4%), hot flushes (5% and 5% v 4%), fatigue (5%
`and 4% v 6%), and weight gain (2% and 2% v 9%).
`Significantly more patients experienced serious adverse
`experiences (death,
`life—threatening, hospitalization) with
`MA (29%) compared with letrozole 2.5 mg (10%; 95% CI,
`11% to 27%) or letrozole 0.5 mg (15%; 95% CI, for 5.5% to
`22%). The majority of serious adverse experiences with MA
`were cardiovascular events, predominantly thromboembolic
`
`Table 4. Adverse Experiences by WHO-Preferred Terminology
`in Greater Than 5% of Patients Irrespective of Trial Drug Relationship
`Letrozole
`Letrazole
`MA
`0.5 mg
`2.5 mg
`160 mg
`(n =1881
`(n = 174)
`(n = 189)
`No.
`%
`No.
`%
`No.
`%
`
`WHO Terms
`
`Any adverse experience
`Musculoskeletal poin‘
`Nausea
`
`147
`48
`36
`
`Dyspnea
`Headache
`Arthralgia
`Fatigue
`Peripheral edema"
`Constipation
`Pain—body as a whole
`Weight increase'f
`Abdominal pain
`Vomiting
`Coughing
`Pain chest—body as a whole
`Dizziness
`Wral infection
`Diarrhea
`Insomnia
`Hot flushes
`Dyspepsia
`Rash’
`Pruritus
`Anorexia
`
`21
`25
`16
`11
`15
`16
`16
`4
`10
`15
`8
`14
`8
`11
`11
`11
`11
`8
`8
`2
`5
`
`78.2
`25.5
`19.1
`
`11.2
`13.3
`8.5
`5.9
`8.0
`8.5
`8.5
`2.1
`5.3
`8.0
`4.3
`7.4
`4.3
`5.9
`5.9
`5.9
`5.9
`4.3
`4.3
`1.1
`2.7
`
`148
`47
`19
`
`16
`22
`23
`19
`15
`13
`5
`4
`10
`13
`14
`12
`6
`12
`11
`6
`10
`9
`10
`3
`9
`
`85.1
`27.0
`10.9
`
`9.2
`12.6
`13.2
`10.9
`8.6
`7.5
`2.9
`2.3
`5.7
`7.5
`8.0
`6.9
`3.4
`6.9
`6.3
`3.4
`5.7
`5.2
`5.7
`1.7
`5.2
`
`1 70
`57
`17
`
`31
`17
`15
`21
`15
`16
`9
`16
`16
`10
`14
`14
`13
`12
`5
`6
`7
`1 1
`6
`10
`9
`
`89.9
`30.2
`9.0
`
`16.4
`9.0
`7.9
`11.1
`7.9
`8.5
`4.8
`8.5
`8.5
`5.3
`7.4
`7.4
`6.9
`6.3
`2.6
`3.2
`3.7
`5.8
`3.2
`5.3
`4.8
`
`‘Composite terms.
`TRefers to weight increase reported as adverse experience, and not to
`measured body weight.
`
`phenomena, reported in 15 patients on MA (six thrombophle—
`bitis, one thrombosis, three myocardial infarctions,
`three
`cerebrovascular accidents, and two pulmonary emboli)
`compared with none on letrozole 2.5 mg and two events on
`letrozole 0.5 mg (one thrombophlebitis and one cerebrovas-
`cular accident).
`No clinically significant changes in blood pressure, pulse
`rate, or ECG were seen in any treatment arm during the
`study. More patients treated with MA experienced weight
`gain of 210% from baseline (MA, 11.5%; letrozole 2.5 mg,
`6%;
`letrozole 0.5 mg, 6%, measured body weight). No
`specific treatment—related abnormalities in routine labora—
`tory parameters were observed apart from a higher fre—
`quency of increased gamma glutamyl
`transferase (yGT)
`with MA (12%) compared with letrozole 2.5 mg or 0.5 mg
`(4% and 5%, respectively).
`
`Discontinuation
`
`In total, 80% patients in the letrozole-2.5 mg arm, 87% in
`the letrozole-0.5 mg arm, and 92% in the MA arm discontin-
`ued treatment, mostly because of PD (68%, 69%, and 77%,
`respectively). The percentage of patients who discontinued
`trial
`treatment due to poor tolerability was significantly
`higher with MA (11%) than with letrozole 2.5 mg (3%; 95%
`CI for the difference, 2% to 12%), compared with 6% in the
`letrozole-0.5 mg arm (95% CI for the difference, —1 to 10).
`Other discontinuations were due to protocol violation,
`noncompliance, withdrawal of consent, loss to follow—up,
`and administrative problems.
`
`DISCUSSION
`
`This is the first reported trial to compare the effects of the
`new aromatase inhibitor,
`letrozole, with an established
`second-line treatment, MA, in the treatment of postmeno—
`pausal women with advanced breast cancer, previously
`treated with an antiestrogen.
`The data show letrozole 2.5 mg to be superior to MA 160
`mg once daily with respect to overall objective response
`rate, duration of response, duration of clinical benefit, TTF,
`performance status, and tolerability. The data also demon—
`strate letrozole 2.5 mg to be superior to letrozole 0.5 mg in
`terms of overall objective response rate, TTP, TTF, and
`overall survival. The dose effect has also been reported in
`another large randomized trial
`that compared the two
`letrozole doses with the reference AI aminoglutethimide
`showing that letrozole 2.5 mg was significantly superior to
`letrozole 0.5 mg in overall survival.20 This dose-response
`effect is surprising and in contrast to endocrine data on
`
`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 2139 p. 6
`
`
`
`LETROZOLE AND ADVANCED BREAST CANCER
`
`459
`
`letrozole. The degree of estrogen suppression and inhibition
`of in vivo aromatization of androstenedione to estrone were
`
`similar for letrozole 2.5 mg and 0.5 mg}7 It also contradicts
`experience with other AIs and with MA. Apart from a few
`reports,”23 most studies with MA,24 arninoglutethimide,25
`fadrozole,26 and anastrozole7’27 have not reported any dose
`effect. One possible explanation for the superiority of
`letrozole 2.5 mg compared with 0.5 mg could be a greater
`inhibition of intratumoral aromatase activity with the higher
`dose. This explanation would be in keeping with a recent
`report of a dose-dependent inhibition of tumor growth with
`letrozole in an in vivo model of intratumoral aromatase.28
`
`Randomized comparative trials2139‘33 have shown that
`tumor response rates could be as low as 5% and as high as
`43% with progestins and A15 in second—line therapy of
`advanced breast cancer after failure to respond to tamoxifen
`treatment. In two randomized, double-blind trials, an objec-
`tive response rate of 11% to 13% was obtained with 1 mg
`fadrozole twice daily compared with an objective response
`rate of 12% to 16% with 40 mg MA four times daily.8 Daily
`doses of 1 mg and 10 mg anastrozole tested in two large
`randomized trials were reported to produce an objective
`response rate of 5% to 13% in postmenopausal patients with
`advanced breast cancer compared with an objective response
`rate of 6% to 10% with 160 mg MA.