ORIGINAL RESEARCH ARTICLE
`
`Clin Phormocokinei 2(IJ4: 43 (8): 529538
`03125963/04/0(D8{J529/$31.00/0
`© 2004 Adis Data Information BV. All rights reserved.
`
`Pharmacokinetic Profile of
`
`Intramuscular Fulvestrant in
`
`Advanced Breast Cancer
`
`Iohn F.R. Robertson} Bjorn Erikstez'n,2 Kent C. Osborne} ]ohn Pippen/1
`Steven E. Come,5 Leroy M. Parker,6 Stan Gertler,7 Mike P. Harrison3 and
`David A. Clarke8
`
`1 Nottingham City Hospital, Nottingham, UK
`Radium Hospital, Oslo, Norway
`Baylor College of Medicine, Houston, Texas, USA
`Texas Oncology, Dallas, Texas, USA
`Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA
`Dana-Farber Cancer Institute, Boston, Massachusetts, USA
`Ottawa Regional Cancer Center, Ottawa, Ontario, Canada
`AstraZeneca Pharmaceuticals, Macclesfield, UK
`
`O0\lG\U‘I>-l>DJI\.)
`
`Abstract
`
`Objective: To characterise the pharmacokinetics of a long—acting formulation of
`fulvestrant following intramuscular administration of single and multiple doses.
`Study design: Pharmacokinetic investigations of single and multiple doses of
`fulvestrant were conducted within two global phase III efficacy studies that
`compared intramuscular fulvestrant with oral anastrozole in postmenopausal
`women with hormone—sensitive advanced breast cancer (study 0020, conducted in
`Europe, Australia and South Africa, and study 0021, conducted in North
`America).
`Methods: Patients received once—monthly intramuscular injections of fulvestrant
`250mg (1 X 5mL for S21 months in study 0020; 2 X 2.5mL for S30 months in
`study 0021). Serial blood samples were collected for the first 28 days after the
`initial dose and immediately prior to all subsequent monthly doses. Plasma
`fulvestrant concentrations were determined by high—performance liquid chro-
`matography—tandem mass spectrometry.
`Patients: Twenty—six (study 0020) and 193 (study 0021) postmenopausal women,
`comprising the pharmacokinetic subgroups of the phase III efficacy trials, were
`studied. Patients had shown disease progression or recurrence following previous
`hormonal therapy for advanced disease or had relapsed after adjuvant endocrine
`therapy with a nonsteroidal antiestrogen.
`Outcome measures and results: For single—dose fulvestrant 250mg, area under
`the concentration-time curve from time zero to 28 days (AUC2g), maximum
`observed plasma concentration (Cmax), minimum observed plasma concentration
`at 28 days (Cmin) and time to maximum plasma concentration (tmax) were
`determined. For multiple—dose fulvestrant 250mg once monthly, steady—state
`trough concentrations (Ctrough) were determined. Plasma fulvestrant concentra-
`
`AstraZeneca Ex. 2060 p. l
`Mylan Pharms. Inc. V. AstraZeneca AB IPR20l6-013 16
`
`

`
`530
`
`Robertson et all.
`
`tions reached a peak at a median of 7 days (range 2-8 days) postdose, and declined
`biexponentially with a slower phase commencing approximately 2-3 weeks
`postdose. Intersubject variability in Cm); and AUC2g was approximately 6-fold
`and 4-fold, respectively. Mean parameters for single—dose fulvestrant were:
`AUC2s, 148 pg 0 day/L; Cmax, 8.2 ttg/L; Cmm, 2.6 ttg/L; tmax, 7.0 days. Geomet-
`ric mean Crrough increased from 2.57 to 6.15 ug/L (study 0020) and from 2.38 to
`6.52 ug/L (study 0021) over the first 6 months, reaching steady—state concentra-
`tions of approximately 6-7 ug/L (study 0020) or 9 ug/L (study 0021). Preliminary
`pharmacokinetic analysis, using a naive pooled data approach, suggests that
`observed single— and multiple—dose plasma profiles can be adequately described
`with a two—compartment kinetic model. Model—generated steady—state AUC25;
`values were approximately 300 pg 0 day/L.
`Conclusions: The intramuscular formulation of fulvestrant displays predictable
`kinetics and approximately 2-fold accumulation on administration once monthly.
`At the proposed therapeutic dosage (250mg once monthly), plasma fulvestrant
`concentrations are maintained within a narrow range throughout the administra-
`tion interval,
`thus ensuring stable systemic drug exposure during long—term
`treatment.
`
`Fulvestrant (Faslodex®)1, a new estrogen recep-
`tor (ER) antagonist, is now available for the treat-
`ment of honnone receptor—positive metastatic breast
`cancer
`in
`postmenopausal women
`that
`has
`progressed during prior
`antiestrogen
`therapy.
`Fulvestrant acts by blocking ER transcriptional ac-
`tivities and downregulating cellular levels of the
`ER.[1'3] Unlike the nonsteroidal antiestrogens (e.g.
`tamoxifen, toremifene and raloxifene), which dis-
`play mixed estrogen antagonist and agonist proper-
`ties,[4v5l fulvestrant causes complete abrogation of
`estrogen—sensitive gene transcription and has no
`known estrogen agonist effect.[6] Fulvestrant has a
`unique mode of action that offers the potential for
`continued hormonal treatment in patients with ad-
`vanced, honnone—sensitive breast cancer that has
`
`progressed with tamoxifen treatment. The efficacy
`of fulvestrant
`in this treatment setting has been
`confirmed in a phase II clinical study, where women
`treated with fulvestrant achieved a clinical benefit
`
`rate of 69% after disease progression with tamoxifen
`treatment.[7] Fulvestrant also offers potential thera-
`peutic advantages over aromatase inhibitors in the
`postmenopausal setting as, unlike aromatase inhibi-
`
`tors, it is able to block the trophic effects of exoge-
`nous estrogens. In two global phase III studies car-
`ried out in postmenopausal women with advanced,
`tamoxifen—resistant breast cancer (trial 0020, con-
`
`ducted in Europe, South Africa and Australia, and
`trial 0021, conducted in North America), fulvestrant
`was similar to anastrozole for the primary efficacy
`endpoint, time to progression (median 5.5 vs 5.1
`months [trial 0020] and 5.4 vs 3.4 months [trial
`
`0021] for fulvestrant and anastrozole, respectively).
`Secondary endpoints included objective response
`rate, clinical benefit rate and median duration of
`
`response, which were not significantly different be-
`tween the two treatment groups.
`In both trials,
`fulvestrant and anastrozole were equally well toler-
`ated.[8v9]
`
`Fulvestrant has low aqueous solubility and has
`been developed as a long—acting, oil—based formula-
`tion for use as a once—monthly intramuscular injec-
`tion. This parenteral depot fonnulation provides ad-
`equate bioavailability and offers potential com-
`pliance advantages over existing breast cancer
`treatments. Intramuscular administration can offer
`
`sustained plasma drug concentrations, and will also
`
`1 The use of trade names is for product identification purposes only and does not imply endorsement.
`
`© 2004 Adis Data Information BV. All rights reserved.
`
`Clin Phormocokinet 2004; 43 (8)
`
`Astrazeneca Ex. 2060 p. 2
`
`

`
`Fulvestrant in Breast Cancer
`
`531
`
`be less affected by vomiting and subsequent tablet
`loss than oral agents. Preliminary findings in post-
`menopausal women with primary breast cancer indi-
`cate that slow—release intramuscular fulvestrant ex-
`
`hibits linear kinetics over the dose range 50—250mg
`following single—dose administration.[10]
`The aim of the present series of investigations,
`conducted within the confines of the phase III clin-
`ical trial programme, was to characterise the phar-
`macokinetic profile of fulvestrant after single and
`multiple doses.
`
`suppressants, low—dose progestins or ketoconazole
`was not permitted during the studies, but continua-
`tion of existing treatment with these agents was
`allowed at the investigator’s discretion.
`The studies were approved by the relevant local
`ethics committees and were conducted in accor-
`
`dance with the principles of Good Clinical Practice
`and the Declaration of Helsinki. Written informed
`
`consent was obtained from each patient prior to
`study entry. Efficacy and tolerability findings from
`these studies are reported elsewhere.[8v9]
`
`Materials and Methods
`
`Pharmacokinetic Study Design
`
`Patients
`
`Both studies recruited postmenopausal women
`with hor1none—sensitive advanced breast cancer who
`
`had experienced either disease recurrence or pro-
`gression following previous hormonal therapy for
`advanced disease or whose disease had relapsed
`following adjuvant endocrine therapy with a non-
`steroidal antiestrogen. Patients were defined as hor-
`mone sensitive if they: (i) had received 212 months
`of adjuvant hormonal therapy before relapse; (ii)
`had shown 23 months of tumour remission or
`
`stabilisation following hormonal therapy; or (iii)
`were ER positive or progesterone receptor (PgR)
`positive.
`Patients who had received more than one prior
`endocrine treatment for advanced breast cancer
`
`were excluded from the studies, as were those with
`
`severe/uncontrolled systemic disease or hepatic im-
`pairment (total bilirubin >l.5 times the upper limit
`of normal [ULN]; alanine aminotransferase [ALT]
`
`or aspartate aminotransferase [AST] >2.5 times the
`ULN in the absence of hepatic metastasis; or ALT or
`AST >5 times the ULN in the presence of hepatic
`metastasis). Patients with a history of treatment with
`fulvestrant or an aromatase inhibitor, or recent expo-
`sure to extensive radiotherapy, cytotoxic therapy,
`estrogen replacement
`therapy, any non—approved
`drug (within 4 weeks) or a luteinising hormone-
`releasing hormone (LHRH) analogue (within 3
`months) were also excluded from study participa-
`tion.
`Initiation of corticosteroids, adrenocortical
`
`European Study (0020)
`
`A randomised, open—label, parallel—group study,
`conducted primarily in Europe, Australia and South
`Africa, investigated the plasma kinetics of fulves-
`trant following single monthly (28 i 3 days) intra-
`muscular injections of fulvestrant 250mg, adminis-
`tered in a total volume of 5mL into the gluteus
`maximus muscle. Serial blood samples were collect-
`ed for the first 28 days of treatment and single blood
`samples were obtained before each subsequent dose,
`for a period of 321 months.
`
`North American Study (0021)
`
`parallel—group
`double—blind,
`A randomised,
`study conducted in North America investigated the
`plasma kinetics of
`fulvestrant
`following single
`monthly (28 i 3 days) intramuscular injections of
`fulvestrant 250mg, administered in two 2.5mL in-
`jections (one in each buttock). To preserve study
`blindness, matching anastrozole placebo tablets
`were supplied for once—daily oral administration.
`Serial blood samples were collected for 28 days
`after the first dose and single blood samples were
`obtained before each subsequent dose, for a period
`of £30 months.
`
`Pharmacokinetic Assessments
`
`Serial blood samples (l0mL) for pharmacokine—
`tic analyses were collected predose and at 3 hours
`and (at any time) 1, 2, 7, 10, 14, 21 and 28 days after
`the first dose of fulvestrant. A further blood sample
`was collected 21 days after the third dose of fulves-
`
`© 2004 Adis Data Information BV. All rights reserved.
`
`Clin Pharrnacokinet 2004; 43 (8)
`
`AstraZeneca Ex. 2060 p. 3
`
`

`
`532
`
`Robertson et all.
`
`In addition, predose
`(study 0021 only).
`trant
`(trough) blood samples (l0mL) were taken immed-
`iately before each monthly fulvestrant injection for
`the duration of each study. Samples were collected
`in lithium—heparin tubes and centrifuged for 10 min-
`utes at 1000g within 10-15 minutes of blood sam-
`pling. The plasma was immediately separated, fro-
`zen and stored at —20°C until required for analysis.
`
`Assay Method
`Plasma fulvestrant concentrations were deter-
`
`mined by means of a validated highly sensitive and
`specific high—performance liquid chromatography
`(HPLC)—tandem mass spectrometry method, using
`deuterated fulvestrant as
`the internal
`standard.
`
`Fulvestrant was extracted from a 0.5mL heparinised
`plasma sample using 21nL of 10% (v/v) hexanel
`2—propanol
`and,
`following
`centrifugation,
`the
`organic layer was evaporated to dryness under a
`stream of nitrogen at 40°C. The dry residue was
`reconstituted in 200},LL of methanol/water (5 : 1 v/v)
`and a 20ttL aliquot was injected onto a 5cm x 4mm
`lnertsil® C18 HPLC column (GL Sciences, Inc.,
`Tokyo, Japan) and eluted at a flow rate of 0.7 mL/
`min. Detection was performed with a Sciex API
`lII+® triple quadrupole mass spectrometer (Perkin-
`Elmer, Boston, MA, US A) in positive multiple reac-
`tion mode (MRM). The following ions were select-
`ed for measurement (precursor/product transitions):
`m/z 607—>589, the product ion of fulvestrant, was
`measured against that of the m/z 613+595 product
`ion of the [2H6]fulvestrant internal standard.
`Concentrations of fulvestrant were calculated
`
`from peak area ratios (fulvestrant/intemal standard)
`by reference to calibration series constructed by
`adding known amounts of fulvestrant (0.25—50 ttgl
`L) to control plasma and extracting these standards
`in parallel with the test samples. Quantification was
`performed using a weighted (1/concentration?) least
`squares linear regression line generated from the
`standard samples. No samples with concentrations
`higher than the top standard were found, and dilu-
`tion into the working range with blank plasma was
`therefore not necessary. The performance of the
`assay was monitored throughout use by inclusion of
`quality control samples prepared at low, medium
`
`and high concentrations in blank human plasma, and
`analysed in duplicate with each batch of sample
`analysis. The acceptance/rejection criteria were
`based upon a coefficient of variation (CV) of 20%.
`The lower limit of quantification of the assay was
`0.25 ttgfL, while assay precision, expressed as the
`intra— and inter—assay coefficients of variation
`(CV%), at this level were 19.5% and 16.2%, respec-
`tively, and both generally 37% at the higher concen-
`trations (as assessed in study 0021). Accuracies
`were 97.3—109% of the theoretical concentrations
`
`for the three quality control standards.
`
`Pharmacokinetic Analysis
`
`The maximum observed plasma concentration
`(Cmax), time to reach the maximum plasma concen-
`tration (tmax) and minimum plasma concentration
`(Cmm) during the 28-day period after the first dose
`were determined from the individual plasma con-
`centration-time profiles. In the multiple-dose stud-
`
`ies, trough plasma concentrations (ctfough) were de-
`termined directly from blood samples collected im-
`mediately prior to each monthly injection. The
`plasma concentration-time data for fulvestrant were
`analysed by non—compartmental methods using vali-
`dated software (WinNonlin version 1.5; Scientific
`
`Consulting, Cary, NC, USA). The area under the
`concentration-time curve from time zero to 28 days
`(AUC23) [i2 days] after the first dose was calculat-
`ed using the linear trapezoidal method. Estimates of
`accumulation, time to steady state and the plasma
`
`elimination half-life (t1/25) were obtained using phar-
`macokinetic models, with first-order input and dis-
`tribution, fitted to the single—dose plasma concentra-
`
`tion-time data and also to the multiple-dose Ctfough
`data from each trial by using a naive pooled data
`approach (i.e.
`it was assumed that all the plasma
`concentrations came from the same patient). The
`data were analysed using WinNonlin with 1/y
`weighting. Model-generated pharmacokinetic para-
`meter estimates were compared for the single and
`multiple doses.
`
`Statistical Analysis
`
`Data were analysed descriptively and, for para-
`meters with a log-normal distribution, Cmax, Cmm,
`
`Ctfough and AUC23 values were summarised as the
`
`© 2004 Adis Data Information BV. All rights reserved.
`
`Clin Pharmacokinet 2004; 43 (8)
`
`AstraZeneca Ex. 2060 p. 4
`
`

`
`Fulvestrant in Breast Cancer
`
`533
`
`Table I. Baseline demographic and clinical characteristics of trial
`0020 and 0021 study populations
`
`geometric mean, CV, range and standard deviation.
`The tmax values were summarised as the median and
`
`the range. The geometric mean was calculated as
`exp (u) and the CV was calculated as (equation 1):
`1
`
`CV% = 100 X|:€Xp(62)-1:|A
`
`(Eq. 1)
`
`where u and 63 are the mean and variance of the
`log—transfor1ned data.
`In those instances where
`350% of plasma fulvestrant concentrations were
`non—quantifiable at a given timepoint, the geometric
`mean and CV were calculated by substituting the
`limit of quantification (0.25 ug/L) for the missing
`values. If >50% of plasma concentrations were non-
`quantifiable,
`the geometric mean and CV values
`were reported as unknown.
`
`Results
`
`Patient Demographics
`
`The baseline demographic and clinical character-
`istics of the two study populations are summarised
`in table 1. Plasma samples for determination of the
`single—dose kinetic profile of fulvestrant were ob-
`tained from 16 of 222 patients randomised to receive
`the drug in study 0020 and from four of 206 patients
`who received the drug in study 0021. However, in
`view of the limited size of this latter patient cohort,
`single—dose data are given for indicative purposes
`only. Trough plasma samples for determination of
`the multiple—dose kinetic profile were obtained from
`26 and 193 patients in studies 0020 and 0021, re-
`spectively. Both study cohorts had a mean age of 63
`years and mean bodyweights were similar: 69kg and
`72kg for studies 0020 and 0021, respectively. The
`majority of patients were Caucasian, although the
`North American cohort included 10% Black and 4%
`
`Hispanic patients.
`
`Sing|e—Dose Phormocokinetics
`
`Following the first dose of intramuscular fulves—
`trant, the time needed to achieve peak plasma con-
`centrations indicated prolonged release of fulves-
`trant from the injection site (tmax generally varying
`
`Characteristic
`
`Age at entry (years)a
`mean (SD)
`median
`
`range
`Weight at baseline (kg)a
`mean (SD)
`median
`
`range
`Race (%)
`White
`Black
`
`Hispanic
`other
`
`Study 0020
`(n = 26)
`
`Study 0021
`(n = 193)
`
`62.9 (10.0)
`61
`
`63.2 (11.2)
`64
`
`43-85
`
`33-89
`
`68.1 (9.8)
`69.0
`
`71.2 (14.2)
`72.1
`
`40.9-81.2
`
`39.5—126.8
`
`100
`0
`
`0
`0
`
`Hormone receptor status (%)
`84.6
`ER+ and/or PgR+
`0
`ER-and/or PgR-b
`15.4
`ER and PgR status unknown
`WHO performance status at baseline (%)
`0
`46.2
`1
`50.0
`2
`3.8
`
`Local regional radiotherapy (%)
`Radiotherapy for metastasis (%)
`Cytotoxic chemotherapy (%)
`Adjuvant hormonal treatment (%)
`yes (<12 months to relapse)
`yes (212 months to relapse)
`no
`
`73.1
`19.2
`38.5
`
`3.8
`42.3
`53.8
`
`Hormonal treatment for advanced breast cancer (%)
`yes (remission for 23 months)
`65.4
`yes (remission for <3 months)
`0
`no
`34.6
`
`a Data for n = 25 and n = 188, respectively.
`b
`ER negative and PgR negative, or ER negative and PgR
`unknown, or ER unknown and PgR negative.
`ER = estrogen receptor; PgR = progesterone receptor.
`
`between 2 and 8 days). Thereafter, the concentra-
`
`tions declined slowly to about a quarter of the maxi-
`
`mum by 28 days after administration. Individual
`
`plasma concentration—ti1ne profiles demonstrated
`
`some interpatient variability in trial 0020; Cmax and
`
`AUC23 values ranged from 3.8 to 23.8 pg/L and 73
`
`to 309 ug - day/L, respectively, in studies 0020 and
`
`85.5
`9.8
`
`4.1
`0.5
`
`86.5
`6.7
`6.7
`
`46.1
`43.5
`10.4
`
`47.7
`31.6
`63.2
`
`8.3
`50.3
`41.5
`
`51.8
`2.6
`45.6
`
`© 2004 Adis DC11‘C1 Information BV. All rights reserved.
`
`Clin Phorrncrcokinet 2004: 43 (8)
`
`AstraZeneca Ex. 2060 p. 5
`
`

`
`534
`
`Robertson et all.
`
`Table II. Single-dose pharmacokinetic parameters for fulvestrant
`after an initial intramuscular dose of fulvestrant 250mg in postmen-
`opausal women with advanced breast cancer. Data are expressed
`as geometric mean (CV%) [n], except where indicated
`
`Study 0020
`8.20 (63.8) [16]
`2.62 (33.4) [14]
`6.97 (1 .86—7.95) [15]
`143 (45.3) [13]
`
`Parameter and unit
`Cmax (ug/L)
`Gmin (HQ/L)
`tmax (days)a
`AUC28 (HQ 0 day/L)
`a Median (range) [n]
`AUC2s = area under plasma concentration-time curve from time
`zero to day 28; Cmax = maximum plasma concentration; C.-mn =
`minimum plasma concentration; CV% = coefficient of variation;
`tmax =
`t0 Cmax.
`
`Study 0021
`4.76 (68.1) [4]
`2.38 (47.7) [131]
`8.8 (6.97—12.0) [4]
`88.4 (47.3) [4]
`
`0021. Mean pharmacokinetic parameters, derived
`from individual patients, are summarised in table 11.
`Because the differences between the actual and
`
`nominal sampling times of individual patients, a
`mean plasma concentration-time profile was not
`
`constructed. However, the data were adequately de-
`scribed by fitting a two—compartment pharmacokin—
`etic model, with first—order input, to a naive pool of
`all individual points. The relationship between the
`observed concentrations and the model-predicted
`
`profile is illustrated in figure 1. Primary and secon-
`
`100-
`
`dary parameter estimates are given in table Ill. It
`was also possible to describe the single—dose data
`with a one—compartment model. However,
`it was
`found that this model under—predicted multiple—dose
`concentrations as a result of the slow terminal phase
`known to occur beyond 21 days.[10]
`
`Mu|’rip|e—Dose Phcnrmcncokinefics
`
`Repeated monthly intramuscular administration
`of fulvestrant 250mg initially resulted in progressive
`
`increases in Ctrough. In trials 0020 and 0021, respec-
`tively,
`the geometric mean Ctrough after the first
`injection was 2.6 and 2.4 ug/L, gradually increasing
`to 6.5 and 6.2 pg/L and approximating to steady
`state after the sixth dose. Although there was some
`interpatient variation (typical CV 30-40%), individ-
`
`ual patient and mean Ctrough Values subsequently
`exhibited minimal fluctuation during both trials, in-
`dicating an absence of any ti1ne—dependent changes
`during long—ter1n administration (figure 2).
`The two—compartment disposition model derived
`previously from single—dose data, with first—order
`input and distribution rate constants, was used to
`provide the initial estimates for a fit of the pooled
`
`Observed
`O
`— Predicted
`
`O
`O
`
`O
`O
`
`8
`
`O o
`
`A 10-
`_l
`
`
`
`0
`
`6
`
`1 _
`
`
`
`53C
`
`.2
`‘<6
`CD0C
`E
`8
`
`Q
`
`0-1
`
`0
`
`I
`7
`
`I
`14
`
`I
`21
`
`I
`28
`
`I
`35
`
`Time (days)
`Fig. 1. Comparison of observed plasma fulvestrant concentrations with the predicted profile derived from a fitted model of the pooled data
`following the first intramuscular dose of fulvestrant 250mg (study 0020).
`
`© 2004 Adis DCITCI Information BV. All rights reserved.
`
`Clin Phcfrmcfcokinet 2004; 43 (8)
`
`AstraZeneca Ex. 2060 p. 6
`
`

`
`Fulvestrant in Breast Cancer
`
`535
`
`Table III. Model parameters and derived pharmacokinetic data for the first dose of fulvestrant during once-monthly intramuscular adminis-
`tration of fulvestrant 250mg in postmenopausal women with advanced breast cancer. Values in parentheses are CV%
`Parameter and unit
`
`Study 0021 (n = 131)
`Study 0020 (n = 28)
`31 (3)
`31 (4.7)
`Apparent total clearance (L/h)
`13 000 (150)
`14 000 (38)
`Apparent volume of central compartment (L)
`0.28 (158)
`0.385 (48)
`Apparent absorption rate constant (day—1)
`39 000 (108)
`32 000 (22)
`Apparent volume of second compartment (L)
`73 (60)
`46 (40)
`Distributional clearance (L/h)
`57 (8.3)
`51 (26)
`Disposition half-life (days)
`333 (3.0)
`369 (4.1)
`AUG.” (ug 0 day/L)
`8.3 (8.8)
`11.8 (6.6)
`Cmax 019/L)
`4.6 (11.2)
`4.2 (8.3)
`tm ax (days)
`AUC... = area under the plasma concentration-time cur\/e from time zero to infinity; Cmax = maximum plasma concentration; CV% =
`coefficient of variation; tmax = time to Cmax.
`
`individual multiple-dose data. The resulting fitted
`profiles illustrate the relationship between the fitted
`profile and the actual plasma concentration-time
`data (figure 3). The use of a two—compartment
`model, based on parameter estimates taken from a
`previous single—dose trial of intramuscular fulves-
`trant 50—250mg,[10] provided an adequate fit for the
`data generated following the first injection, and was
`successfully extended to fit
`the multiple-dose
`
`Ctrough data. The two—compartment, model-generat-
`ed pharmacokinetic parameter estimates (transfer
`and elimination rate constants and volume of the
`
`central compartment) for the multiple-dose data
`were similar (within 20%) to those for the single-
`dose data. These were estimated with better preci-
`sion, as the trough samples provide additional infor-
`
`mation on the slower terminal phase (AUC5, repre-
`senting approximately 60% of the total AUC). Thus,
`the two—compartment pharmacokinetic model was
`considered suitable to describe exposure to fulves-
`trant following both single and multiple doses. The
`
`model appears to adequately describe the Ctrough
`values observed in patients from both studies for up
`to 2.5 years after the first dose and, thus, indicates
`linearity of fulvestrant pharmacokinetics during re-
`peated administration. Steady-state plasma fulves-
`trant exposures (AUCss), predicted according to the
`structural model, were very similar in both studies
`
`(294 and 336 pg - dayfL for trials 0020 and 0021,
`respectively). By comparison with the single—dose
`data, AUCSS represented an approximate 2-fold ac-
`cumulation.
`
`There was no evidence of a change in the phar-
`macokinetic behaviour of fulvestrant compared with
`that following the first injection, and it was possible
`to predict the multiple-dose pharmacokinetics of
`fulvestrant by using modelling techniques. Plasma
`concentrations of fulvestrant were maintained with-
`
`in a narrow range throughout the assessment period,
`consistent with
`the
`linearity
`of
`fulvestrant
`pharmacokinetics during repeated administration
`over a prolonged period.
`
`Discussion
`
`As shown in previous investigative studies in
`patients with breast cancer,[7] the present longer-
`term studies confirm that fulvestrant
`is released
`
`slowly into the systemic circulation after intramus-
`cular administration of the long-acting formulation
`and that plasma concentrations are maintained at
`detectable levels over the 4-week administration
`
`interval. The pharmacokinetics of fulvestrant, when
`administered as an intramuscular injection, are cha-
`racterised by a narrow plasma concentration range
`(approximately 3-fold) maintained throughout the
`
`administration interval and an apparently long t1/Z5
`(40-60 days). The latter is reflected in the relatively
`long time to reach steady state following repeated
`administration (3-6 months). However, it should be
`noted that although accumulation does occur (ap-
`proximately 2- to 3-fold), the majority of this is
`achieved after three doses.
`
`© 2004 Adis DC1‘l‘C1 Information BV. All rights reserved.
`
`Clin Phclrmclcokinet 20041: 43 (8)
`
`Astrazeneca Ex. 2060 p. 7
`
`

`
`536
`
`Robertson et all.
`
`100 —
`
`QU)
`3E
`2
`
`g
`
`§
`C
`8
`
`10 ‘
`
`:2
`
`"
`A
`W :1
`
`3-‘
`—
`
`IA
`
`__
`
`__
`
`" 5 ¥ %
`
`It‘
`
`__
`
`==
`
`__
`
`__
`
`::
`
`__
`
`__
`
`Study 0020
`Study 0021
`
`A
`
`__
`
`‘
`
`_
`
`--
`
`==
`
`‘_| W
`
`" I
`
`__
`
`1
`
`
`0
`2
`4
`6
`8
`10
`12
`14
`1 6
`1 8
`20
`22
`24
`
`Time (mo)
`Fig. 2. Geometric mean (: SD) trough plasma fulvestrant concentrations during once—monthly intramuscular administration of fulvestrant
`250mg in studies 0020 and 0021.
`
`the
`As with many slow-release preparations,
`pharmacokinetics of
`long—acting
`intramuscular
`fulvestrant are controlled by its slow systemic re-
`lease, and therefore its intrinsic elimination charac-
`
`teristics cannot be assessed directly. Single—dose
`data from volunteer studies using an intravenous
`formulation indicate that
`fulvestrant
`is
`rapidly
`cleared from the circulation with a well defined t1/25
`of approximately 14-18 hours.[”] For the long—act-
`ing intramuscular formulation, the apparent long t1/25
`indicates that drug release is much slower than elim-
`ination. The resulting plasma profile can therefore
`be considered to reflect the rate of drug absorption;
`the overall elimination rate constant ke approxi-
`mates to the absorption rate constant l<a.[12v13] Thus,
`although alterations in clearance rate (e.g. due to
`hepatic or renal impairment or drug interactions)
`might affect overall systemic exposure (AUC) on
`repeated administration, minimal immediate effects
`on the plasma profile would be expected.
`
`Earlier indications of modest drug accumulation
`occurring during once—monthly intramuscular ad-
`ministration of long—acting fulvestrant 250mg[7]
`were confirmed in the present studies, with Ctfough
`(predose) approximately doubling between the first
`and sixth doses as steady state was approached.
`
`lntrapatient variation was minimal in those patients
`where numbers of sequential trough samples were
`obtained, even though the interindividual variability
`was maintained.
`
`The aim of these pharmacokinetic analyses was
`to investigate the consistency of the concentration
`data from subsequent doses of fulvestrant with the
`more extensive data from the first administration, as
`
`a diagnostic of appreciable change in the underlying
`pharmacokinetics, and not
`the identification of
`covariates that could explain individual variation.
`The naive pooled data approach ignores individual
`differences and, consequently, the method does not
`differentiate interindividual variability from in-
`traindividual variability. Given the large number of
`potentially confounding variables
`(age, weight,
`pathophysiological changes, activity, etc.), the para-
`meter estimates using the naive pooled data ap-
`proach are usually considered poorer than those
`provided by rigorous two—stage or nonlinear mixed
`effects modelling approaches. However, by using
`the substantial body of single—dose profiles and con-
`tinuous multiple trough data from larger numbers, a
`satisfactory exploratory model has been provided,
`which appears to describe the observed data over a
`prolonged period of administration.
`
`© 2004 Adis D0110 Information BV. All rights reserved.
`
`Clin Phcrrmcicokinet 2004; 43 (8)
`
`AstraZeneca Ex. 2060 p. 8
`
`

`
`Fulvestrant in Breast Cancer
`
`537
`
`Despite the decreasing numbers of patients at
`later timepoints, the good fit of the observed trough
`plasma data with the two—compartment model over
`multiple dose cycles indicates that there was no
`appreciable modification in the kinetics of fulves-
`trant during long—term administration (up to 2.5
`years). This consistency is further supported by the
`close similarity in model-generated estimates of
`steady—state plasma fulvestrant exposure in the two
`
`phase III studies. Therefore, the increase in plasma
`concentration that occurs with multiple doses seems
`predictable and appears to be solely due to the slow-
`release characteristics of this formulation rather than
`
`to any change in pharmacokinetic behaviour.
`
`The pharmacokinetics of the long—acting intra-
`muscular formulation of fulvestrant have previously
`been reported to be unaffected by the method of
`administration (either as 2 x 2.5mL or 1 x 5mL),
`
`0
`
`Observed
`
`— Predicted
`
`a
`
`100;
`
`3‘
`E2103
`3
`:
`5
`_
`.2
`—
`E
`_
`E
`_
`CD
`8O
`O
`
`1:
`:
`
`
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`b
`100:
`
`O‘
`3
`0‘
`.0
`0'
`_
`:§;-.§:’Eo‘5= 0:0 8 0000’
`1
`g‘,
`::e’a‘9:oa-s’ O O "
`3
`
`-
`o;n;u2».«o
`9
`v
`C
`- =: 0 V
`-.2
`9 - E3
`'
`-—
`.»-g
`C
`..._O,»~«
`8
`Q o0
`C
`O0
`81-0
`IO
`
`0
`
`I
`3
`
`I
`6
`
`I
`9
`
`I
`12
`
`I
`15
`
`I
`18
`
`I
`21
`
`I
`24
`
`I
`27
`
`I
`30
`
`I
`33
`
`Time(mo)
`Fig. 3. Comparison of the observed plasma fulvestrant concentrations with predicted mu|tip|e—dose profiles derived from a fitted model of
`the pooled data following once-monthly intramuscular administration ot fulvestrant 250mg in studies 0020 (a) and 0021 (b).
`
`© 2004 Adis DCITCI Information BV. All rights reserved.
`
`Clin Phcirmclcokinet 2004; 43 (8)
`
`AstraZeneca Ex. 2060 p. 9
`
`

`
`538
`
`Robertson et all.
`
`with the two regimens providing virtually superim-
`posable plasma profiles after the first dose.“‘‘] The
`present study shows that repeated administration of
`1 x 5mL and 2 x 2.5mL injections of fulvestrant
`250mg are equally effective in maintaining plasma
`concentrations in an empirically defined therapeutic
`range (that
`is, plasma fulvestrant concentrations
`generally reached in treatment schedules associated
`with efficacy) for at least 30 months. It therefore
`seems reasonable to conclude that these regimens
`may be used interchangeably, depending on which
`is considered more convenient in a particular clin-
`ical setting.
`
`Conclusion
`
`The intramuscular depot formulation of fulves-
`trant displays predictable kinetics and limited drug
`accumulation as a once—monthly treatment regimen.
`The formulation offers the assurance of stable drug
`exposure, with plasma fulvestrant concentrations
`maintained within a narrow range throughout the
`administration interval, and obviates patient com-
`pliance issues during long-terrn treatment.
`
`Acknowledgements
`
`This study was funded by AstraZeneca. The authors
`would like to thank Mr S.A. Rhead for provision of the
`pharmacokinetic data, and we acknowledge the contributions
`of Analytico Medinet BV, Breda, The Netherlands and Cedra
`Corp., Austin, Texas, USA for performing the plasma assays.
`The authors have provided no information on conflicts of
`interest directly relevant to the content of this study.
`
`ICI 182,780 (Fas-
`2. Howell A, Osborne CK, Morris C, et al.
`lodexTM): development of a novel ‘pure’ antiestrogen. Cancer
`2000; 89: 817-25
`
`3. Curran M, Wiseman L. Fulvestrant. Drugs 2001; 61: 807-13
`4. Jordan VC, Murphy CS. Endocrine pharmacology of anties-
`trogens as antitumour agents. Endocr Rev 1990; 11: 578-610
`5. MacGregor Jl, Jordan VC. Basic guide to the mechanisms of
`antiestrogen action. Pharmacol Rev 1998; 50: 151-96
`6. Wakeling AE, Dukes M, Bowler J. A potent specific pure
`antiestrogen with clinical potential. Cancer Res 1991; 51:
`3867-73
`
`7. Howell A, DeFriend DJ, Robertson JF, et al. Pharmacokinetics,
`pharmacological and anti—tumour effects of the specific anti-
`oestrogen ICI 182