reviews
`
`Annals of Oncology 23: 1378-1386, 2012
`doi:10.1093/annonc/mdr593
`Published online 8 February 2012
`
`The sequential use of endocrine treatment for advanced
`breast cancer: where are we?
`
`C. Barrios‘, J. F. Forbes2, W. Jonat3, P. Conte4, W. Gradishar5, A. Buzdars, K. Ge|mon7,
`M. Gnant8, J. Bonneterreg, M. Toilo, c. Hudis” & J. F. R. Robertsonl2*
`7lnternal Medicine Department, PUCBS School of Medicine, Porto Alegre, Brazil; 2School of Medicine & Public Health, University of Newcasfle, Newcastle, Australia;
`3Department of Obstetrics and Gynaecology, University of Kiel, Kiel, Germany; “Deparvnent of Oncology and Hematology, University of Modena and Reggio Emilia,
`Modena, Italy; 5F1’obert H. Lurie Comprehensive Cancer Center, Northwestern University Feinberg School of Medicine, Chicago; 5Deparm7ent of Breast Medical
`Oncology, University of Texas MD Anderson Cancer Center, Houston, USA; 7Department of Medical Oncology, University of British Columbia, Vancouver, Canada;
`8Department of Surgery, Comprehensive Cancer Centre Vienna, Medical University of Vienna, Vienna, Austria; glntegrated Clinical Research Unit, Centre Oscar
`Lambret, Lille, France; 70Breast Surgery Department, Kyoto University, Kyoto, Japan; "Breast Cancer Medicine Service, Memorial Sloan—Kettering Cancer Center,
`New York, USA; 12/-"acuity of Medicine and Health Sciences, Nottingham University, Derby, UK
`
`Received 16 November 2011; accepted 23 November 2011
`
`U)
`E
`.9>
`
`asC
`
`Background: Hormone receptor—positive advanced breast cancer is an increasing health burden. Although
`endocrine therapies are recognised as the most beneficial treatments for patients with hormone receptor—positive
`advanced breast cancer, the optimal sequence of these agents is currently undetermined.
`Methods: We reviewed the available data on randomised controlled trials (RCTS) of endocrine therapies in this
`treatment setting with particular focus on RCTS reported over the last 15 years that were designed based on power
`calculations on primary end points.
`Results: In this paper, data are reviewed in postmenopausal patients for the use of tamoxifen, aromatase inhibitors
`and fulvestrant. We also consider the available data on endocrine crossover studies and endocrine therapy in
`combination with chemotherapy or growth factor therapies. Treatment options for premenopausal patients and those
`with estrogen receptor—/human epidermal growth factor receptor 2—positive tumours are also evaluated.
`Conclusion: We present the level of evidence available for each endocrine agent based on its efficacy in advanced
`breast cancer and a diagram of possible treatment pathways.
`Key words: advanced breast cancer, algorithm, endocrine, hormone receptor positive, postmenopausal, treatment
`pathway
`
`introduction
`
`Despite early diagnosis and improving treatment options for
`primary breast cancer, there continues to be a substantial
`number of women who relapse with advanced disease.
`Approximately 80% of breast cancer cases in Western countries
`are estrogen receptor positive (ER+) [1] and for the majority of
`these patients, endocrine therapy is an appropriate option in
`both the adjuvant and advanced setting. This manuscript
`reviews the available data on randomised controlled trials
`
`(RCTs) of endocrine therapies in the treatment of hormone
`receptor—positive (HR+) advanced breast cancer.
`
`HR+ postmenopausal patients with
`advanced breast cancer
`
`Before contemporary phase III trials involving third generation
`aromatase inhibitors (Als), RCTs were much smaller in size and
`
`*Correspondence to: Prof. J. F. R. Robertson, University of Nottingham, Department
`of Medicine and Health Sciences, Royal Derby Hospital, Utfoxeter Road,
`Derby DE22 SDT, UK. Tel: +44—i332—72488i; Fax: +44—i3S2—724880;
`E—rnail: lCJi‘ll'I robei‘tscin@nottingi'iarn.ac.uk
`
`were seldom prospectively powered to test for either superiority
`or equivalence between the two arms. Indeed, assumptions
`were made that different endocrine agents such as tamoxifen,
`megestrol acetate (MA) and aminoglutethimide had equivalent
`efficacy (but different side—effect profiles) based on small
`datasets where type 2 errors were a distinct possibility.
`
`tamoxifen versus high—dose estrogens
`
`Tamoxifen is a selective ER modulator (SERM), which
`antagonises estrogen signalling in the treatment of HR+
`advanced breast cancer. The high—dose estrogens were known
`to be effective in breast cancer treatment, possibly by increasing
`p53 levels [2]. A review of RCTs comparing tamoxifen with
`high—dose estrogens reported that overall response rates
`(ORRs) were comparable (33% versus 31%) [3]. In the initial
`report of diethylstilbestrol (DES) compared with tamoxifen
`(n = 143), there were also no differences observed in time to
`treatment failure, duration of response or overall survival (05)
`[4]. Although, a subsequent update reported that survival was
`significantly longer with DES [5], tamoxifen became the initial
`
`© The Author 2012. Published by Oxford University Press on behalf of the European Society for Medical Oncology.
`All rights reserved. For permissions, please email: journals.permissions@oup.com
`
`
`
`
`
`
`
`
`
`glcgz/QIEHIIEIIIO/{IISIQAIHHSHI')IdOHSIILIOI.IQ[l(\0I.{LI9SI’E[IE/310'S1'EI.I.II1D!.p.IU‘_IXO'DHU[l'[l'2//Id1fjI{IIIOIJPQPEOIHMOG
`
`
`
`
`
`
`
`
`
`Astrazeneca Ex. 2072 p. 1
`Mylan Pharms. Inc. v. Astrazeneca AB IPR2016-013 16
`
`

`
`Annals of Oncology
`
`endocrine therapy of choice due to its improved side—effect
`profile.
`
`tamoxifen versus MA
`
`MA is thought to treat breast cancer by inhibiting pituitary
`function and thus suppressing luteinising hormone and the
`subsequent production of estrogen. In at least five RCTs [6—11],
`tamoxifen was shown to have comparable efficacy with MA in
`terms of ORR and OS and a better side—effect profile.
`
`tamoxifen versus SERMs
`
`Tamoxifen has also been tested against several other SERMs.
`Analyses have shown that tamoxifen was comparable to
`toremifene (n = 1421) [12] or idoxifene (n = 220) [13] and was
`superior to droloxifene [ORR (P = 0.02) and time to
`progression (TPP) (P < 0.001)] [14] and to arzoxifene
`[progression—free survival (PFS; P = 0.01)]
`[15].
`Overall, tamoxifen was therefore deemed to be as good as, or
`better than, all alternative SERMs with phase II crossover
`studies showing cross—resistance between tamoxifen and other
`SERMS.
`
`tamoxifen versus first- and
`
`second-generation Als
`
`AIs are thought to work by inhibiting aromatase signalling,
`which ultimately blocks the estrogen receptor. The first-
`generation AI aminoglutethimide was shown to be comparable
`with tamoxifen alone [16, 17] or with aminoglutethimide plus
`tamoxifen [18, 19]. The latter trials are among the first to study
`an AI in combination with an antiestrogen and no
`improvement was observed over the antiestrogen alone.
`
`tamoxifen versus other endocrine
`
`agents (meta-analysis)
`
`Fossati et al. [20] reviewed 35 RCTs comparing tamoxifen with
`a range of other endocrine therapies, including ovarectomy,
`MA, AIs, medroxyprogesterone acetate, SERMs, goserelin and
`fluoxymesterone. They reported an ORR of 30% with
`tamoxifen versus 29% with the other agents and an OS hazard
`ratio of 1.02 [confidence interval (CI) 0.94-1.10].
`Tamoxifen became the standard therapy for advanced breast
`cancer, having demonstrated first—line efficacy when compared
`with a range of other endocrine agents in advanced breast
`cancer.
`
`third-generation Als: anastrozole and
`Ietrozole (competitive, non-steroidal)
`and exemestane (non-competitive,
`steroidal) versus MA as second-line
`endocrine therapy
`
`These trials were the first endocrine therapy RCTs prospectively
`powered to demonstrate significant differences in clinical
`ouLcome(s). Anastrozole showed no significant difference in
`
`TTP from MA on an initial analysis [21, 22]. However,
`a planned subsequent analysis found anastrozole 1 mg o.d. to
`be associated with significantly increased OS versus MA
`(median 26.7 versus 22.5 months, respectively; P < 0.025) [23].
`Two studies of Ietrozole 2.5 mg o.d. versus MA showed no
`significant difference in TTP or OS [24, 25]. Exemestane
`resulted in an increased TTP (4.7 versus 3.8 months; P = 0.037)
`and a significantly longer OS (median OS not reached for
`exemestane at time of publication versus 28.5 months for MA;
`P = 0.039) compared with MA [26]. AIs were initially
`introduced based on the improved side—effect profile but
`similar TTP versus MA. Subsequently, this decision was
`supported by the OS data with anastrozole and the increased
`efficacy seen with exemestane.
`
`third-generation Als: anastrozole and
`Ietrozole (competitive, non-steroidal)
`and exemestane (non-competitive,
`steroidal) versus tamoxifen as first-line
`endocrine therapy
`
`Anastrozole was shown to be superior to tamoxifen in terms of
`TTP in a North American—based trial where almost 90% of
`
`patients were known to be HR+ [27]. No significant difference
`in TTP was reported in TARGET, a ‘Rest of the World’ study.
`However, only 45% of patients in TARGET were known to
`have an HR+ tumour [28]. In a pooled retrospective analysis of
`the two trials including patients with known HR+ tumours,
`anastrozole was shown to be superior to tamoxifen for TTP but
`not for OS [29]. Letrozole significantly prolonged TTP
`compared with tamoxifen but, again, no significant difference
`in OS was observed [30]. Exemestane had similar PFS and OS
`compared with tamoxifen using the log—rank test; when PFS
`was assessed using the Wilcoxon test, it was significantly longer
`with exemestane than tamoxifen [31].
`Overall, the third— generation AIs were deemed more effective
`in terms of disease control than MA and tamoxifen and were
`
`well tolerated and so have become the preferred first—line
`endocrine therapy. This finding is similar to the adjuvant
`settings, where third—generation AIs have been compared with
`tamoxifen in large trials [32—35].
`
`Fulvestrant: 250 mg dose
`
`Fulvestrant is a selective ER down regulator (SERD) that binds,
`blocks and increases degradation of ER, resulting in inhibition
`of estrogen signalling [36]. It was initially approved at a dose of
`250 mg/month after studies showed that it was as effective as
`anastrozole 1 mg/day in the treatment of HR+ advanced breast
`cancer in the second—line setting, after tamoxifen [37].
`
`Fulvestrant: 500 mg dose
`
`Fulvestrant 500 mg was compared with fulvestrant 250 mg
`in a phase III RCT in the second—line setting in women with
`advanced breast cancer in the CONFIRM study. The primary
`end point TTP was significantly longer for patients
`receiving fulvestrant 500 mg versus fulvestrant 250 mg
`
`Volume 23 | No. 6 | June 2012
`
`o|oi:10.1093/annonc/mdr593 | 1379
`
`
`
`
`
`
`
`
`
`‘gz/§.I‘l2'('lI.I'2]‘U0/§1IS..I3A}'LIf1SLl'L[0[..I3AAOI.[LI9SiE[II01[i]/\[112fl.I0'S[‘BU.I'f'l()_l‘p.I0‘IXO'3IIOULIQ//Iduq1110.1}PQPBOIUAAOG
`
`
`
`
`
`
`
`
`
`Astrazeneca Ex. 2072 p. 2
`
`

`
`(hazard ratio = 0.80; P = 0.006). The difference in 05 did not
`reach statistical significance (P = 0.091) at the initial analysis
`[38]. This finding is fully consistent with the increased biological
`eifects seen with the 500 mg dose compared with 250 mg [39].
`Fulvestrant 500 mg was also compared with anastrozole
`1 mg/day in the metastatic setting in the phase II FIRST trial
`(11 : 205). TTP was significantly prolonged with fulvestrant
`500 mg (hazard ratio : 0.626; P = 0.0496) [40]. Adverse events
`were comparable between treatment arms. Data from the
`FIRST study showed that the significant difference in TTP had
`persisted with longer follow—up (23.4 months with fulvestrant
`versus 13.1 months with anastrozole; hazard ratio = 0.66; P =
`0.01) [41].
`In summary, fulvestrant 500 mg has a biologically greater effect
`and provides a clinically meaningful benefit over fulvestrant 250
`mg. The standard dosing schedule of fiilvestrant should now be
`500 mg and, based on its increased efficacy, should be considered
`earlier in the treatment of advanced disease.
`
`endocrine crossover studies
`
`In contrast with the comparative wealth of data from head—to—
`head studies, there are only two RCTs assessing the impact of
`treatment sequence. In a first—line study, letrozole 2.5 mg was
`associated with longer initial TTP than tamoxifen 20 mg (9.4
`versus 6.0 months; P = 0.0001); yet, there was no significant
`difference in survival [30].
`In a 60—patient subgroup of the TARGET study, time to first
`progression was 11.3 months with anastrozole and 8.3 months
`with tamoxifen [42]. The time from randomisation to second
`progression was 28.2 months for patients who started on
`anastrozole and crossed over to tamoxifen and 19.5 months for
`
`the opposite regimen. However, the study is not sufficiently
`powered to draw conclusions.
`Although there is a scarcity of data from robust RCTs, the
`available non—randomised data regarding the effects of
`endocrine sequence demonstrate that response to first—line
`therapy predicts for response to subsequent endocrine therapy
`[43, 44]. However, there are no data showing that one
`treatment sequence is preferable to another.
`
`endocrine therapy plus chemotherapy
`
`According to accepted convention, the concomitant use of
`chemotherapy and hormonal therapy is not recommended in
`the treatment of breast cancer, as the two mechanisms are
`considered theoretically to be antagonistic.
`The Australian and New Zealand Breast Cancer Trials Group
`evaluated doxorubicin 50 mg/m2 plus cyclophosphamide
`750 mg/ml in sequence, and in combination, with tamoxifen
`20 mg b.i.d. (n = 339). As patients were not selected based on
`HR status, it is not surprising that the response rates were
`variable between groups. However, TTP was not significantly
`different and OS was almost identical, irrespective of treatment
`sequence [45].
`Tominaga et al. [46] reported that MA in combination with
`cyclophosphamide, doxorubicin and fluorouracil (CAF)
`chemotherapy was better than CAF alone. However, this design
`is chemotherapy with or without endocrine therapy (in an HR
`
`Annals of Oncology
`
`unknown population) rather than endocrine therapy with or
`without chemotherapy in HR+ advanced breast cancer.
`Overall, clinical trials in this area are lacking, particularly
`with combinations of the newer classes of endocrine agents
`such as AIs or SERD (fulvestrant) with or without
`chemotherapy.
`
`the treatment of ER+/human epidermal
`growth factor receptor 2-positive
`postmenopausal patients with
`advanced breast cancer
`
`Although a notable proportion of patients with breast cancer
`have human epidermal growth factor receptor 2-positive
`(HER2+) tumours, there are currently no definitive data on
`when to use anti—HER2 agents and hormone therapy in the
`advanced breast cancer setting. However, some studies have
`been conducted in these patients.
`The addition of trastuzumab or lapatinib to AI therapy,
`anastrozole or letrozole, respectively, has shown clinical benefit
`for patients with tumours that were HR+/HER2+. In both
`studies, addition of the growth factor inhibitor improved
`clinical benefit rate (CBR) and PFS but there was no significant
`difference in 05 (P = 0.325 for trastuzumab [47, 48]; not
`reported for lapatinib [49]).
`To date, there are no studies comparing endocrine therapy
`with chemotherapy in this setting [50].
`Fulvestrant with or without lapatinib was evaluated in
`a phase III study in patients with HR+ advanced breast cancer.
`At the third interim analysis, no improvements were observed
`in PFS or 05 with the addition of lapatinib to fulvestrant.
`However, in patients with HER2+ tumours, a trend towards
`improved PFS was observed (5.9 versus 2.8 months for
`fulvestrant + lapatinib versus fulvestrant alone; P = 0.29).
`Treatment was generally well tolerated [51].
`In all of these studies, the data suggest that the addition of
`a HER2—targeted therapy increased the efficacy of the endocrine
`agent by almost doubling both CBR and TTP. There are no RCT
`comparisons of the combination versus the anti—HER2 therapy
`alone. Whether an individual patient receives combination of
`endocrine and HER2—targeted therapies or an endocrine therapy
`alone is a decision for each patient and their physician [52].
`
`combined endocrine and growth
`factor therapies
`
`Cristofanilli et al. [53] reported prolonged PFS with anastrozole
`plus gefitinib (n = 43) versus anastrozole plus placebo (11 = 50;
`HR = 0.55; 95% CI 0.32-0.94) in postmenopausal women with
`HR+ metastatic breast cancer. This, however, was not reflected
`in the neoadjuvant RCT of anastrozole versus anastrozole plus
`gefitinib [54]. In a study of tamoxifen with or without gefitinib,
`no PFS benefit was reported with the addition of gefitinib to
`tamoxifen [55].
`In a phase III trial, letrozole plus temsirolimus offered no
`PFS advantage over letrozole alone in ER+ metastatic breast
`cancer [56]. However, in a randomised phase II trial of 111
`patients with HR+ and HER2—negative tumours and with prior
`exposure to AIs, tamoxifen with everolimus was superior to
`
`1380 | Barrios et al.
`
`Volume 23 | No. 6 | June 2012
`
`
`
`
`
`
`
`
`
`‘gz/§.I‘l2't'lI.I'2]‘U0/§1IS..I3A}'LIf1SLl'L[0[..I3AAOI.[LI9SIE[II01[I]/\[112fl.I0'S[‘BLI.I't'l()_l‘p.I0‘IXO'3IIOULIQ//Iduq1110.1}PQPBOIUAAOG
`
`
`
`
`
`
`
`
`
`AstraZeneca Ex. 2072 p. 3
`
`

`
`Annals of Oncology
`
`tamoxifen alone in terms of CBR (61.1% versus 42.1%) and
`median TTP (8.5 versus 4.5 months; P = 0.008, exploratory
`analysis) [57]. The mTOR inhibitors warrant fiirther clinical
`evaluation in combination with endocrine therapies,
`particularly SERMs.
`In a randomised phase II study (n = 156), the addition of the
`monoclonal IGF—1 receptor antibody antagonist AMG 479 to
`exemestane or fulvestrant provided no additional PFS benefit
`for patients with HR+ metastatic or locally advanced breast
`cancer [58].
`It has been suggested that therapies targeted at growth factor
`signalling may help to overcome acquired resistance to
`endocrine therapy. However, current data are lacking and
`further, robust clinical investigations are required.
`
`the treatment of HR+ premenopausal
`patients with advanced breast cancer
`
`In premenopausal patients, ovarian ablation has been the
`standard treatment for over 100 years [59, 60]. Ovarian
`ablation [oophorectomy, radiotherapy, luteinising hormone-
`releasing hormone agonists (LHRHa)] with or without
`tamoxifen perhaps remains the most common initial
`therapeutic endocrine choice in premenopausal women.
`However, tamoxifen has been shown to be an effective
`monotherapy agent too.
`Tamoxifen is approved for the treatment of premenopausal
`patients with ER+ advanced breast cancer, and two small RCTs
`have shown that it has comparable efficacy (in terms of
`response rates and OS) to oophorectomy [61, 62]. Goserelin, an
`LHRHa, is recognised as an effective alternative to
`oophorectomy in pre/perimenopausal women following phase
`III evaluation [61, 63, 64]. In a meta—analysis of four studies by
`Klijn et al. [65] (11 = 506), the combination of LHRHa plus
`tamoxifen resulted in significantly prolonged PFS (P < 0.001)
`and 05 (P = 0.02) relative to either agent alone.
`Of note, in the largest of these studies, combination therapy
`was compared with sequential therapy. Although the TTP was
`longer for the combination, there was no difference between the
`two arms in terms of ‘time to total failure’ (Unpublished data;
`AstraZeneca on file).
`AIs are not suitable for use alone in premenopausal women due
`to the high oestradiol levels in these patients; AIs must therefore
`be used in combination with ovarian suppression. Fulvestrant
`250 mg has not been evaluated as a sole therapy in premenopausal
`women with advanced breast cancer but fulvestrant 250 mg in
`combination with goserelin has been reported to have a CBR rate
`of 45% in premenopausal patients (n = 20) [50].
`In summary, there are no firm data to suggest that ablation
`of ovarian function in premenopausal women renders them
`equivalent to postmenopausal patients, but until any other data
`become available, this appears the most logical therapeutic
`approach and current trials have shown some degree of success.
`
`discussion
`
`In many of the key studies reported to date, observed
`improvements in TTP did not translate into OS improvements.
`Therefore, how do we weight end poinLs—i.e. CBR compared
`
`with TTP compared with OS? (Figure 1). Furthermore, the
`relevance of an end point depends on the mechanism of action
`of the treatment. For example, ORRs may not be appropriate
`for agents that slow or delay disease progression. Therefore, it is
`particularly important to select end points appropriately for
`studies in advanced breast cancer.
`
`With endocrine therapy, prior response predicts the
`likelihood of subsequent response to another endocrine agent,
`and this should be taken into account when assessing whether
`to prescribe a subsequent endocrine therapy. However, for
`individual patients the duration of control beyond 6 months on
`one endocrine therapy does not predict for the duration of
`control beyond 6 months on a subsequent endocrine therapy.
`This fact suggests that individual tumours respond differently
`to different endocrine agents and that being able to select which
`endocrine agent an individual patient’s tumour is most
`sensitive to is a realistic, as well as a clinically worthwhile, goal.
`Treatment is continued until patients experience clinical
`disease progression, assuming the absence of serious adverse
`events. Stopping endocrine therapy is not recommended in
`advanced breast cancer, although some specific occasions do
`arise where the physician and patient may agree to this approach.
`It would seem worth testing intermittent endocrine therapy in
`future trials. This could either be with a single agent or involve
`multiple agents used in rotation in a predefined or randomly
`assigned sequence, with the aim of stopping or delaying the
`development of tumour resistance. There are limited data
`suggesting a degree of further benefit in individuals re—exposed
`to the same endocrine agent. Most data in this setting are with
`tamoxifen but it is all non—randomised. V\/hile it is not poor
`practise to reintroduce a prior treatment in a patient who
`previously responded, it is often not the best therapeutic option
`unless all endocrine options have been exhausted.
`The paucity of data from RCTs of sequencing of endocrine
`therapies in patients with advanced breast cancer means that no
`
`Better
`
`Worse
`
`Endpoints
`
`Time Io
`progression
`
`Response rates
`
`DVEFGII Sl.|Ni‘«‘B|
`
`CDC9
`
`Drug A
`Drug E
`
`Figure 1. Comparison of properties of drug A versus drug B. Which drug
`would you choose? Schematic representation of different potential end points
`and the level of ‘weight’ assigned to them. X represents the score assigned for
`each drug for each end point. The circles represent the amount of ‘weight’ one
`might assign to each end point. Consider how your opinion of drug A versus
`drug B would change if the locations of the markers moved.
`
`Volume 23 | No. 6 | June 2012
`
`doi:10.1093/annonc/mdr593 | 1381
`
`
`
`
`
`
`
`
`
`‘gz/§.I‘l2'('lI.I'2]‘U0/§1IS..I3A}'LIf1SLl'L[0[..I3AAOI.[LI9SiE[II01[i]/\[112/€10's{12uJno[p1o_ixo':3uouue//:duq1110.1}PQPBOIUAAOG
`
`
`
`
`
`
`
`
`
`Astrazeneca Ex. 2072 p. 4
`
`

`
`definitive recommendations can be made. There are few RCTs
`
`that have compared the same sequence of two drugs given in
`the opposite order: the study with letrozole and tamoxifen
`provides the most robust data in this setting [30]. The most
`reliable evidence currently available for possible sequences is
`provided by head—to—head trials that have been conducted in
`second— or third—line settings where the patients’ prior therapies
`are known and the therapies are proven to be effective in
`patients in that treatment setting. No trials have been
`conducted that specifically compared different combinations of
`endocrine agents in sequence. We have, therefore, prioritised
`selection of endocrine agents based on their known efficacy in
`this particular setting of advanced breast cancer. We have
`highlighted the level of evidence (Figures 2 and 3).
`
`treatment selection
`
`postmenopausal patients
`
`first line. For first—line treatment in advanced breast cancer,
`a non—steroidal AI is the standard choice, although there seems
`little to differentiate between anastrozole and letrozole in this
`
`Adjuvant treatment
`
`De nova.-‘no prior
`adjuvant endocrine therapy
`
`>1 year di5ease—free interval
`pnst-adjuvant tamoxifen
`
`Recurrence an adjuvant tamoxifen
`
`>1 year disease-free interval
`post-adjuvant AI
`
`Recurrence or progression an
`adjuvant Al
`
`First line
`ANASOZOLE “M1
`Di’
`LETROZOLEW‘
`Of
`EXEMESTANIE '31‘
`or
`Fulvestrant 500 mg l4‘|,4:)
`
`ANASTROZOLE """"‘
`0|’
`
`LETROZOLE M
`0'”
`EXEMESTANE ‘"1
`or
`
`Fulvestrant 500 mg 9”‘ ‘=1
`(>250 mg) 1"“
`
`ANASTROZOLE M 73'
`°'
`LETROZOLE ml
`or
`FULVESTRANT 500 mg I” 59'
`
`FU LVESTRANT500 mg
`(>250mgl‘ "91
`or
`Exemesrane
`or
`Tamoxifen
`
`FULVESTRANT 500 mg
`
`Annals of Oncology
`
`setting. The phase II data of fiilvestrant 500 mg versus an AI
`(anastrozole) in the first—line setting showed a significant
`advantage for fulvestrant 500 mg. Considering the long—term
`follow—up, fillvestrant has become a therapeutic option in this
`setting, especially if there is a contraindication to AIs or
`a problem with compliance. While some clinicians in some
`countries may accept the phase II data as being sufficient for
`this treatment option, a phase III study of first—line fulvestrant
`500 mg versus AIs is recommended to fiilly understand the
`potential benefits.
`With non—steroidal AIs being widely used in the adjuvant
`setting, the choice of a different endocrine agent for first—line
`advanced disease has to be considered. In the CONFIRM trial,
`all patients were receiving a second hormone therapy, and
`fulvestrant 500 mg was superior to fulvestrant 250 mg in terms
`of the primary end point, TTP. Approximately half the patients
`were treated after adjuvant endocrine therapy and half after
`endocrine therapy for advanced disease. Approximately half of
`the patients had received prior AI and half prior tamoxifen. In
`the absence of other RCT data, fulvestrant 500 mg would
`appear to have the most RCT data in the post—adjuvant AI
`setting.
`There are non—randomised data that show that turnours will
`
`respond to other endocrine agents in the post—AI setting (e.g.
`exemestane, tamoxifen, MA), but these are selected datasets and
`are not obtained from RCTs.
`
`second line. In studies of second—line endocrine therapy for
`advanced disease, the third—generation AIs were considered
`superior to progestins [23—26]. The main benefits, which
`led to initial regulatory approval, involved safety: absence of
`significant weight gain and reduction of dyspnoea observed
`with MA [23, 26]. Survival benefits for non—steroidal AIs
`were seen on long—term follow—up with anastrozole [23].
`However, as non—steroidal AIs are now used much earlier,
`other endocrine agents should be considered for second
`line [67]. Similar second—line data are available for
`exemestane.
`
`Fulvestrant 250 mg is equivalent to Ms in the second—line
`setting in terms of TTP and OS. Similar results were seen in two
`large phase III studies (studies 20 and 21) of parallel design [3,
`68], which were then combined in a prospectively planned
`overview analysis [37]. In this study, 99% of patients had
`received tamoxifen as their prior endocrine therapy. Recently,
`fulvestrant 500 mg has been shown to be superior to fulvestrant
`250 mg in the second—line setting after failure of antiestrogen
`therapy [38]. In this study, 57.5% of patients had received prior
`tamoxifen and 42.5% had received a prior non—steroidal AI.
`The hazard ratio for PFS was 0-8 (P = 0.006) with a trend
`towards OS (P = 0.09) on the first data analysis. These findings
`are consistent with data from the phase II RCT in the first—line
`setting, which showed that fulvestrant 500 mg had greater
`efficacy than anastrozole.
`In terms of post—AI in advanced disease, the EFECT study
`reported no difference between fulvestrant 250 mg and
`exemestane. Again, since CONFIRM subsequently reported
`that fulvestrant 500 mg was superior to 250 mg, this provided
`an indirect comparison between fulvestrant 500 mg and
`exemestane post—AI given in the advanced selling.
`
`BASED ON RANDOMISED PHASE Ill DATA
`Based on randomised phase II data
`A treatment option, but not based on randomised, controlled, data
`
`Figure 2. Recommended order of selection of first—line endocrine agents
`in various therapeutic settings, based on level of evidence available.
`
`1382 | Barrios et al.
`
`Volume 23 | No. 6 | June 2012
`
`Astrazeneca Ex. 2072 p. 5
`
`
`
`
`
`
`
`‘gz/§.I‘l2'('lI.I'2fU0/§1IS..I3A}'LIf1SLl'L[0[..I3AAOI.[LI9SIE[II01[I]/\[112/€10's{12umo[p1o_ixo':3uouue//:duq1110.1}PQPBOIUAAOG
`
`
`
`
`
`
`
`
`
`
`
`

`
`Annals of Oncology
`
`Prior treatment
`
`Second line
`
`Third line
`
`Prior tamoxifen therapy
`
`Prior non-steroidalfixltherapy
`
`AidASFRDZOLEm"”'
`on
`IEFROZDIL ‘'5' "'
`Gr
`EXENESTANE '”l
`or
`RJLVESTRANTSGU rng"’l
`
`FuLuss11mn5uo mg
`l>15fl mg) I”!
`01
`EXEMESIINE '3“
`01
`PULVESTRANTESU rr1g'’‘l
`
`‘
`
`
`
`Anasrmzde
`0|’
`Lefrazde
`or
`Exernstane
`or
`Fulvestrantfiflu mg
`
`EXEMESTANEW1
`or
`FULVESTPANT250 mg i?"
`or
`Tamoxifen
`
`BASED ON RANDDMISEDPHASE H1 DATA
`A treatment option, but norbasedan mndomised, confroffed, data
`
`Figure 3. Recommended order of selection of second— and third—line agents in various therapeutic settings, based on level of evidence available.
`
`third line. There are currently few RCT data on which to base
`recommendations for third—line endocrine therapy in advanced
`disease except for EFECT where half the patients received
`second—line and half third—line endocrine therapy. There was no
`significant difference between exemestane and fulvestrant (250
`mg dose). However, 70% of the patients in this study were de
`novo hormone—resistant, which made it difficult to draw
`comparisons between agents. This RCT did, however, show
`a CBR of 30-35% in this particular setting. Therefore, it seems
`appropriate to start patients on third—line fulvestrant (but at the
`higher dose of 500 mg) or exemestane assuming they have had
`responses to prior endocrine therapy and have not received that
`particular agent previously.
`Based on the above considerations, we have compiled
`a diagram of possible treatment pathways (Figure 4).
`
`premenopausal patients
`
`There are few RCT data relating to endocrine therapy in
`premenopausal patients with advanced disease, which often
`involves visceral metastases. Endocrine therapy is appropriate
`in this setting, as long as there is no immediately life-
`threatening situation (e.g. lymphangitis carcinomatosis,
`extensive liver replacement). However, if premenopausal
`patients do not respond to an initial endocrine therapy, they
`are often switched to chemotherapy. Therefore, in the most
`serious cases, such as the presence of liver metastases, the most
`effective endocrine treatment (in terms of initial CBR rates)
`should be used initially, such as a combination of tamoxifen
`plus LHRHa [65]. Where the disease appears more indolent
`(e.g. soft tissue or bone metastases), LHRHa first, followed by
`further sequential endocrine therapy, appears appropriate.
`Als alone are not effective in premenopausal women but
`have shown efficacy in non—randomised phase II studies in
`combination with LHRHa [69]. Similarly, fulvestrant in
`combination with LHRHa has also shown efficacy benefits in
`premenopausal patients in non—randomised controlled studies
`[50, 70].
`
`HR+lHER2+ disease
`
`Study data consistently show that combinations of endocrine
`agents and HER2 inhibitors are additive to endocrine agents
`alone; monotherapies show approximately half the CBR and
`TTP of the combinations [48, 49, 51]. There are no randomised
`data comparing the combinations and anti—HER2 treatments
`alone. Furthermore, there are no RCT data on whether to
`continue anti—HER2 therapy with a subsequent endocrine
`agent.
`
`conclusions and recommendations
`
`Since few RCTs have reported significant differences in OS
`(although some have been observed in the second—line
`setting), a physician may consider selecting a cheaper product
`rather than the recommended treatment in a cost—sensitive
`
`environment. However, improvements in cancer care have
`often been by small steps, which have ultimately resulted in
`significant improvements in outcomes. Indeed, the median
`survival of patients with advanced breast cancer today is
`longer than two or three decades ago, even though no specific
`agent(s) can be determined as wholly responsible for this
`improvement.
`Despite the therapeutic advances in HR+ disease described
`above, new endocrine agents or treatment sequences are
`required that can significantly improve clinical outcomes,
`particularly survival. This is relevant for all ER+ tumours
`but particularly for specific subsets of patients such as
`those with HR+/HER+ tumours and for premenopausal
`patients.
`Non—steroidal AIs are currently considered to be the most
`effective first—line treatment of postmenopausal patients with
`advanced breast cancer, although recent fulvestrant 500 mg
`data have challenged this. A phase III trial comparing
`fulvestra