`
`European journal of Cancer Vol. 32A, No. 4, pp. 576-588, 1996
`Copyright © 1996 Published by Elsevier Science Ltd. All rights reserved
`Printed in Great Britain
`0959-8049/96 8l5.00+0.00
`
`S0959-8049(96)00032-9
`
`Special Paper
`
`New Endocrine Therapies for Breast Cancer
`
`A. Howell,‘ 8. Downey2 and E. Anderson3
`
`‘CRC Department of Medical Oncology and 3Tumour Biochemistry Laboratory, Christie Hospital NHS Trust,
`Manchester, U.K.; and 2Department of Surgery, University Hospital of South Manchester, Manchester, U.K.
`
`INTRODUCTION
`
`IT IS 100 YEARS since Beatson removed the ovaries of a 33-year
`old patient with advanced breast cancer in May 1895 [1]. She
`responded to treatment for 42 months which led later to the
`general acceptance of this form of therapy. The mechanism of
`the response was unknown at the time as the term ‘hormone’
`was not coined until later [2].
`In the 100 years since Beatson made his observation, a
`series of endocrine therapies have been introduced, many of
`which have later been discarded [3]. Since nearly all endocrine
`therapies give equivalent response rates and response dur-
`ations, the reason for change has been the need to reduce
`the toxicity of treatments and to make them more widely
`applicable. Thus, the relatively toxic high—dose oestrogens
`have been replaced by the relatively non-toxic tamoxifen,
`adrenalectomy by aromatase inhibitors and androgens by pro-
`testogens.
`
`THE NEED FOR NEW ENDOCRINE TI-IERAPIES
`
`The active and continuing search for new agents is fuelled
`by the considerations outlined in Table 1. We need agents
`with increased efficacy in advanced breast cancer, as adjuvants
`
`Table 1. Some aims of modem endocrine therapy for breast cancer
`
`Increased efficacy
`
`0 Advanced disease — response rate
`— response duration
`0 Adjuvant therapy — cure rate
`— delay in relapse
`— any activity if found
`
`- Prevention
`
`Decreased toxicity
`
`0 General
`
`0 Endocrine
`
`—— gastrointestinal
`symptoms, asthaenia
`— sweats/flushes
`—- weight gain
`
`Improved general
`health
`
`0 Cardiovascular — less events
`I Skeleton
`— less events
`0 Uterus
`— no proliferation
`
`Correspondence to A. Howell.
`Revised and accepted 12 Jan. 1996.
`
`after surgery and for prevention. We also need to aim for
`even greater decreases in toxicity and to design agents which
`increase women’s general health as well as having an antitu-
`mour effect. Improvements are required with respect to gen-
`eral health since breast cancer often arises during the period
`of a woman’s life when the activity of the ovaries is declining,
`or has been artificially interrupted leading to menopausal
`symptoms and long-term increase in the risk of cardiovascular
`and skeletal problems. These problems can be reversed by
`oestrogen and progesterone replacement
`therapy. We are
`reluctant to use these hormones in the clinic because of fear of
`
`recurrence. It would be of great value if we could design an
`antitumour endocrine agent which also eliminated meno-
`pausal problems.
`Although we may consider that current endocrine therapies
`are an improvement over older ones, there remains room for
`considerable improvement. The aim of this review is to con-
`sider the efficacy, toxicity and general health aspects of newer
`endocrine therapies which are in clinical trial, but which are
`not yet (or may never be) commercially available. The three
`most active areas of clinical trial are the new anti—oestrogens
`together with the new aromatase inhibitors and antiprogestins.
`Other areas which are regarded as endocrine therapies and
`which include the use of vitamin D analogues, retinoids and
`somatostatin analogues will not be covered in view of the
`paucity of clinical results to date. Nor will we mention LHRH
`analogues or progestins since there are few new data concern-
`ing these agents.
`
`ANTI-OESTROGENS
`
`The activities of agents which resemble the triphenylethy-
`lene anti-oestrogens such as tamoxifen were first assessed in
`the 1940s [4, 5] but it was not until the demonstration that
`tamoxifen was as active but less toxic than oestrogens and
`androgens that this type of therapy gained wide acceptance,
`first as treatment for advanced disease [6], then as an adjuvant,
`and more recently for the prevention of breast cancer [7] . Two
`main avenues have been taken in order to attempt to improve
`on tamoxifen. One is to chemically alter the non—steroidal
`triphenylethylene ring structure of tamoxifen or to produce
`new non—steroidal ring structures, e.g. the benzothiaphenes
`
`576
`
`|nnoPharma Exhibit 10510001
`
`
`
`New Endocrine Therapies for Breast Cancer
`
`577
`
`Tamoxifen
`
`Toremifene
`
`Droloxifene
`
`
`
`Raloxifene
`
`0\/\N
`
`O
`
`Figure 1. Structures of non-steroidal anti-oestrogens clinically available or in clinical trial.
`
`(Figure 1). The second is to produce steroidal analogues of
`oestrogen with growth inhibiting activity (Figure 2).
`The two types of antioestrogen, steroidal and non—steroidal,
`appear to have different mechanisms of action which may
`account for differences in their activity and side-effect profiles.
`The triphenylethylene anti-oestrogens bind to the oestrogen
`binding sites of oestrogen receptor (ER) monomers which
`then combine to form dimers. This process of dimerisation
`facilitates binding of the ER to specific oestrogen response
`elements (ERE) in the vicinity of oestrogen-regulated genes.
`The ER protein contains two trans-activating functions
`(TAFs) both of which are active when oestrogen binds to the
`molecule, resulting in the range of gene transcription and gene
`repression associated with the effect of oestrogen. Tamoxifen
`binding to ER results in activation of TAF1 in a manner
`similar to oestrogen, but activation of TAF2 is abrogated by
`tamoxifen. Thus, tamoxifen is a partial agonist because it
`activates TAF1 and an antagonist because it inhibits TAF 2
`[8, 9].
`
`OH
`
`HO
`
`/"(CH2)9S0(CH2)3CF2CF3
`
`ICI 182,780
`
`Figure 2. Structure of the steroidal specific
`oestrogen ICI 182,780.
`
`‘pure’ anti-
`
`Thc activity of the steroidal anti-oestrogens appears to be
`different. For example, the specific antioestrogen ICI 182780
`binds to ER, but because of the long side chain on the 7
`alpha position of the molecule it appears to stearically hinder
`receptor dimerisation [10]. There is evidence that ER turnover
`is increased with an associated reduction of detectable ER
`
`molecules in the cell [11, 12]. In the absence of receptor
`dimerisation, binding of ER to ERES may be abolished or
`attenuated. In vitro, virtually no transcriptional activity of ER
`has been detected in cells treated with specific anti-oestrogens.
`
`Non-steroidal anti-oestrogen: (NSAEs)
`Five NSAES have completed their preclinical testing pro-
`gramme and are in clinical trial. The clinical trial programmes
`of TAT-59, raloxifene and idoxifene, are still in their early
`stages whereas for toremifene and droloxifene, phase III trials
`comparing each agent with best standard therapy (e.g.
`tamoxifen) are in progress.
`The rationale for deciding to embark upon a clinical trial
`programme for a new NSAE depends upon laboratory studies
`indicating some measure of superiority over tamoxifen. Given
`preclinical evidence of superiority, the trial programme then
`needs to demonstrate that the new NSAE has some measure
`
`of superiority over standard therapy (i.e. tamoxifen) in the
`clinic. Areas which may need to be addressed in preclinical
`studies and clinical studies are shown in Table 2.
`
`In preclinical studies, evidence of superiority over tamoxifen
`may be sought in one or more of the following areas; receptor
`binding, antitumour activity,
`the balance between tumour
`antagonism and peripheral agonism, whether there are poten-
`tially useful alternative mechanisms of action, and activity
`against tamoxifen-resistant cells. We will examine how each
`of the new agents fare in these areas compared with tamoxifen.
`The structures of each of the NSAEs are shown in Figure l.
`
`|nnoPharma Exhibit 10510002
`
`
`
`578
`
`A. Howell et al.
`
`Table 2. Assessment of non-steroidal anti-oestrogens (NS/IE5)
`
`(A)
`
`(B)
`
`In the laboratory
`Oestrogen receptor binding
`Tumour antagonism
`Peripheral antagonist/agonist ratio
`Alternative mechanisms of action
`Activity against tamoxifen-resistant cell lines
`
`§":“‘*’!"!"
`
`In the clinic
`Activity as first-line therapy
`Activity in tamoxifen-resistant tumours
`Side-effect profile
`Utility of peripheral antagonist/agonist ratio
`Pharmacokinetics
`
`U1>hU)k\):-‘
`
`All are based on the triphenylethylene structure of tamoxifen
`with the exception of raloxifene which is a benzothiaphene.
`
`Oestrogen receptor binding. All the NSAEs have greater affin-
`ity for the ER relative to tamoxifen and a greater binding to
`ER relative to oestrogen compared with tamoxifen with the
`exception of toremifene (Table 3) [13—l7].
`
`Tumour antagonism. Preclinical antitumour activity is most
`often tested against human mammary tumour cell
`lines
`(usually MCF—7 cells) both in culture and when transplanted
`into athymic nude mice. In addition, activity is usually tested
`against carcinogen-induced mammary tumours in rodents.
`Some of the available data are summarised in Table 4 [13, 14,
`16-25]. In some studies no direct comparison with tamoxifen
`was made and these are not cited. The antitumour activities
`of toremifene and raloxifene in vitro and in vivo were less or
`
`equally active as tamoxifen, whereas those of droloxifene
`and TAT-59 were apparently more active. Idoxifene was
`apparently more active than tamoxifen but not to the same
`extent as droloxifene and TAT-59.
`
`Peripheral agonist/antagonist ratio. All of the NSAEs are par-
`tial agonists, and it is customary to look for compounds which
`have low or reduced agonist activity since it is thought that
`high activity may result in reduced antitumour effects. How-
`ever, increased agonist activity may be beneficial with respect
`to the skeletal and cardiovascular systems. The relative
`agonist/antagonist effects of the five non-steriodal anti-oestro-
`gens in the immature rat uterus assay are shown in Table 5
`[14, 16, 17, 23, 26-28]. Antagonist activity is assessed by
`
`Table 3. Binding to oestrogen receptors (ER) relative to tamoxifen
`and to oestrogen ofNSAE: in clinical trial
`
`Binding to ER % binding to ER
`relative to
`relative to
`tamoxifen
`oestrogen
`
`Tamoxifen
`Toremifene
`Droloxifene
`Raloxifene
`TAT-59
`Idoxifene
`
`—
`X 1
`X 10
`?
`x 10
`x 2.5
`
`5
`5
`7 .5
`>100
`10
`12.5
`
`[Ref.]
`
`[13]
`[13]
`[14]
`[15]
`[16]
`[17]
`
`percentage reduction in weight of the oestrogen primed uterus
`caused by the antioestrogen, and agonist activity is assessed
`by growth stimulation of the uterus by the antioestrogen
`in the absence of oestrogen priming. In general, the new
`compounds have less agonist and more antagonist activity
`than tamoxifen; raloxifene and idoxifene appear to be the
`most antagonistic and least agonistic.
`
`Alternative mechanisms of action. It is known that high con-
`centrations of non-steroidal anti-oestrogens cause cell death
`in ER-negative as well as ER-positive cell lines in vitro. The
`mechanism of non-receptor mediated cell death is not well
`understood, but may be related to calmodulin antagonism
`[13], inhibition of protein kinase C activity or to antioestrogen
`binding sites, which seem to be present in cells irrespective of
`their ER status. Whether these activities are associated with
`
`response to NSAEs in vivo is not known, although if they
`were, responses would be expected irrespective of receptor
`status, which is rarely found clinically. Nevertheless, interest
`in increasing the degree of calmodulin antagonism is being
`investigated. For example, idoxifene is four times more active
`as a calmodulin antagonist than tamoxifen [29]. It remains to
`be seen whether more potent antagonists can be synthesised
`which, if active, should increase tumour response rates.
`
`Activity against tamoxifen-resistant cell lines. In vitro evidence
`that a new NSAE is effective against tamoxifen-resistant cell
`lines would indicate a use for such an agent in tamoxifen-
`resistant tumours in the clinic. Few data are available for
`
`lines, but Jarman (personal
`NSAEs in tamoxifen-resistant
`communication) has shown that idoxifene is 10 times more
`active against
`the tamoxifen-resistant cell
`line RL-3 than
`tamoxifen itself, and this observation has led to a clinical trial
`of this agent in patients who have failed tamoxifen. Clinical
`trials assessing cross-resistance of NSAEs are outlined in the
`section on ‘Activity of newer NSAEs in tamoxifen-resistant
`tumours’.
`
`Clinical data
`
`As judged by receptor binding, antitumour activity and
`agonist activity,
`toremifene appears to be very similar to
`tamoxifen in preclinical studies. Droloxifene looks much more
`active than tamoxifen in all three assays, although there are no
`data on whether it has an alternative mechanism of action or
`whether it is active in tamoxifen failures. We have been unable
`
`to find full reports on raloxifene, but it looks highly active with
`respect to ER binding and oestrogen antagonism in the rat
`uterus assay, although disappointing in animal model systems.
`Both TAT-59 and idoxifene appear to have excellent antioes-
`trogen profiles. The question now arises of whether these
`preclinical data are reflected in the clinical experience with
`each drug. Data on the first— and second-line activity, side-
`effect profile, peripheral antagonist/agonist ratios and the
`pharrnacokinetics of the NSAEs are summarised below as far
`as they have been tested.
`
`New non-steroidal anti-oestrogens as first-line therapy. The
`current results of the trials of NSAEs as first-line therapy for
`advanced disease are summarised in Table 6. All the studies
`
`were performed in patients with ER-positive or unknown
`tumours, and all were previously untreated with endocrine
`therapy for advanced disease. There were no significant differ-
`ences in response rates between tamoxifen and toremifene in
`
`|nnoPharma Exhibit 10510003
`
`
`
`New Endocrine Therapies for Breast Cancer
`
`579
`
`Table 4. Preclinical antitumaur activity of non-steroidal anti-oestrogen: compared to tamox-
`ifen (TAM)
`
`Human cell lines in
`vitro
`
`Human cell lines in
`the nude mouse
`
`Carcinogen-induced rat
`tumours
`
`Toremifene
`Droloxifene
`Raloxifene
`
`TAT-59
`Idoxifene
`
`< X 1 [18,19]
`X 10 [20]
`P
`
`~>< 10 [16]
`><l.5 [17]
`
`2 NT
`> TAM [21]
`NT
`
`> TAM [24]
`> TAM [25]
`
`DMBA = to TAM [13]
`R3230AC > TAM [14,21]
`NMU < TAM [22]
`DMBA < TAM [23]
`DMBA > TAM [16]
`NMU > TAM [17]
`
`NT, not tested against TAM. DMBA, dimethylbenz(a)anthracene. NMU, nitrosomethylurea.
`
`anti-
`non—steroidal
`of
`activity
`Table 5.Agonist/antagonist
`oestrogens in the immature rat uterus assay
`
`% Agonism
`
`% Antagonism
`
`[Ref.]
`
`Tamoxifen
`Toremifene
`Droloxifene
`
`50
`43
`35
`
`[2 6]
`[27]
`[14]
`[28]
`[23]
`83
`5
`Raloxifene
`[16]
`>TAM
`?
`TAT-59
`
`
`
`15 8 5Idoxifene [17]
`
`50
`?
`~20
`
`TAM, tamoxifen.
`
`Table 6. Results of studies using newer non-steroidal anti-
`oestrogens as first-line endocrine therapy for advanced breast cancer
`
`Drug [Ref.]
`
`Dose
`(mg/day)
`
`No. of patients
`
`Response
`CR/PR (%)
`
`Toremifene [32]
`(Phase II trials
`summarised)
`
`Toremifene [30]
`Tamoxifen
`(Phase III trial)
`
`Toremifene [31]
`
`Tamoxifen
`(Phase III trial)
`
`Droloxifene [33]
`(Randomised phase
`II trial)
`
`TAT-59 [59]
`(Randomised Phase
`II trial)
`
`20
`60
`240
`
`240
`40
`
`60
`200
`20
`
`20
`40
`100
`
`10
`20
`40
`
`14
`93
`38
`
`31
`31
`
`221
`212
`215
`
`84
`88
`96
`
`15
`11
`13
`
`21
`52
`68
`
`29
`44
`
`21
`22
`19
`
`30
`47
`44
`
`15
`55
`31
`
`receive 20, 40 or 100 mg of droloxifene per day. This study
`comprised a large number of patients and reported signifi-
`cantly higher and impressive response rates at the two higher
`doses compared to the lower one (Table 6). Time to pro-
`gression in higher dose groups was also significantly longer.
`Preliminary data on TAT 59 look promising. Phase II studies
`with raloxifene are in progress and it is too early for idoxifene
`to have been assessed as a first-line agent.
`
`Activity of newer NSAES in tamoxifen-resistant tumours. New
`endocrine therapies are usually tested after tamoxifen failure.
`This clinical situation is particularly interesting with respect
`to NSAEs, since we discover whether drugs which are thought
`to have similar mechanisms of action show cross-resistance or
`
`cross—sensitivity.
`With the exception of one small study [34], the response
`rate (CR + PR) to toremifene after tamoxifen failure is 5% or
`lower (Table 7) [30, 34-40]. However, a number of patients
`had prolonged stabilisations of disease in these studies
`although it is difficult to assess numbers of these because short
`and long durations were combined in most. If we define at
`least 6 months as a ‘no change’ (NC) response [41], we
`estimate that NC may be 20-25%.
`There were no responses in a small study using raloxifene
`as a second-line treatment [39]. However, responses and NC
`were seen with droloxifene and idoxifene after tamoxifen
`
`failure, suggesting that TAT 59 ‘phenyl’ ring subsitutions
`of the tamoxifen molecule may produce non—cross-resistant
`NSAES. However, responses were lower than expected for
`standard second-line therapy, such as megestrol acetate, sug-
`
`Table 7. Response to NSAEs in tamoxifen-resistant breast cancer
`
`Toremifene
`
`Dose No. of CR+PR NC
`(mg) patients
`%
`% (Duration of NC) [Ref.]
`
`200
`240
`240
`200
`240
`100
`
`10-10
`20
`
`9
`34
`50
`102
`23
`26
`14
`33
`14
`
`33
`0
`4
`5
`0
`15
`0
`27
`14
`
`35
`26
`44
`23
`22
`19
`NK
`4
`21
`
`[34]
`(NK)
`[35]
`(5-27 months)
`[36]
`(> 2 months)
`(med 7.8 months) [37]
`(med 6.0 months) [30]
`(> 6.0 months)
`[38]
`[39]
`[5 9]
`[40]
`
`(‘Long’)
`(> 5-12 months)
`
`the randomised trials that have been published to date [30,
`31]. These data conflict with several phase II studies where
`toremifene showed higher response rates than tamoxifen at
`doses of 60 and 240 mg/day [32]. These high response rates
`have not been seen in the phase III studies using comparable
`doses.
`
`Droloxifene
`Raloxifene
`TAT-59
`Idoxifene*
`
`The results of a major international phase II trial of droloxi-
`fene were reported recently [33]. Patients were randomised to
`EJC 32:4-B
`
`NC, no change. NK, not known.
`*Given mainly as third-line endocrine therapy.
`
`|nnoPharma Exhibit 1051.0004
`
`
`
`580
`
`A. Howell et al.
`
`gesting some cross-resistance despite these chemical modifi-
`cations.
`
`Side—eflect profiles ofNSAE. If new NSAEs have comparable
`activity to tamoxifen but have reduced side-effects, then these
`are of clinical interest and could replace tamoxifen. Some of
`the more common side-effects are shown in Table 8 [32, 33,
`39, 42]. Although no precise data are available for idoxifene
`and few for TAT-59, there appears to be little difference
`with respect to tamoxifen and the new NSAEs. Differences
`between drugs may be true differences, but could also be
`explained by the assiduousness with which toxicity was
`sought, or the size of the study (e.g. the raloxifene data was
`based on only 14 patients).
`
`Utility of the peripheral agonist/antagonist ratio. A major aim,
`stated in studies of new NSAEs, is for less agonistic and more
`antagonistic molecules. The aim is laudable with respect to
`the tumour and the endometrium but, in terms of women’s
`general health, more agonistic activity towards bone and the
`cardiovascular system would probably be beneficial.
`All five new NSAEs have agonist activity in viva since
`they reduce gonadotrophin levels and increase SHBG (sex
`hormone binding globulin) (with the possible exception of
`idoxifene). With the exception of raloxifene, there appears to
`be no data available on the effects of new NSAEs on bone
`
`density, lipids and the endometrium. Raloxifene had equival-
`ent activity to Premarin on bone and reduces low density lipid
`cholesterol significantly, but has no significant effect on high
`density lipid cholesterol. Raloxifene also showed no stimu-
`latory effect on the endometrium and if this compound has
`good antitumour activity, it may be an attractive choice in the
`clinic [43].
`
`NS/IEpharmacokz'netz'cs. Toremifene has similar pharmaco-
`kinetics to those of tamoxifen [44]. It takes longer for idoxifene
`to reach steady-state concentrations (6-12 weeks) than tamox-
`ifen. The terminal half-life of idoxifene in patients on pro-
`longed therapy is 23 days, which is longer than tamoxifen
`[40]. Therapeutic levels of droloxifene are reached within the
`first day of therapy, in contrast to tamoxifen, where thera-
`peutic concentrations are reached after 11 days, but the ter-
`minal half-life of droloxifene is short at 25 h [45, 46]. This
`means that droloxifene may be more suitable than tamoxifen
`when considering the approaches to sequencing with other
`endocrine therapies or chemotherapies. The pharmacokineties
`of raloxifene do not appear to have been published, making
`the available pharmacokinetic data for the new NSAEs incom-
`plete.
`
`Table 8. Incidence of common side-effects with new NS/IE5
`
`I-Iot flushes Lassitude Nausea/vomiting
`(%)
`(°/o)
`(%)
`
`[Ref]
`
`30
`Tamoxifen
`19
`Toremifenc
`29
`Droloxifene
`43
`Raloxifene
`10
`TAT~59
`“similar to
`Idoxifene
`TAM”
`
`10
`8
`29
`14
`3
`
`[42]
`[32]
`[33]
`[39]
`[59]
`
`10
`10
`26
`36
`3
`
`TAM, tamoxifen.
`
`Summary. We can say, from very large and extensive clini-
`cal phase III studies, that the preclinical data have been borne
`out with regard to the similarity in effectiveness of toremifene
`compared to that of tamoxifen. The preclinical studies pro-
`duced no data on cross-resistance, but the clinical studies
`appear to show cross-resistance between the two molecules
`with respect to their antitumour effects. The small number of
`responses that were seen could well have been related to the
`withdrawal of tamoxifen [47]. We have been unable to find
`data on toremifene with respect to bone, lipids and the endo-
`metrium. One possible advantage of toremifene is that it does
`not appear to produce DNA adducts in the standard assays.
`Droloxifene appeared active in the preclinical screens and
`also appears to be so in the clinic, although there are, as yet,
`no clinical phase III data available. It has some cross-sensi-
`tivity with tamoxifen and its short half-life may make it parti-
`cularly useful in alternating schedules. In preclinical tests, it
`looks less agonistic than tamoxifen, but does show agonist
`activity in the areas reported to date (reduced LH and FSH,
`increased SHBG), but more data are required.
`In rat uterus assays, raloxifene was shown to be only weakly
`agonistic, but paradoxically has proved to be a clinically useful
`agent for the treatment of osteoporosis and has a favourable
`effect on lipids. The animal data may predict the clinical effect
`on the uterus since raloxifene (in preliminary studies) is said
`to have little agonist effect on this organ [43].
`Chemical data available on TAT-59, a Japanese NSAE,
`suggest it may be very active. Idoxifene has been specifically
`designed to be active in tamoxifen failures and is now in
`trial in this clinical situation. Its preclinical and early clinical
`activity look promising.
`
`STEROIDAL (‘PURE’) ANTIOESTROGENS
`Substitutions of the oestrogen molecule at various positions
`can produce compounds with antioestrogenic activity [48-
`52]. The oestrogen molecule was chosen as a basis for further
`development of anti-oestrogens because it was felt by Wake-
`ling and Bowler [53] that further alteration of the triphenyle-
`thylene molecule was unlikely to lead to strikingly better
`anti-oestrogens.
`This new generation of anti-oestrogens have been described
`as ‘pure’ or specific anti-oestrogens since they have little or no
`agonist activity in preclinical studies. Clinical data, in addition
`to preclinical data, are only available for the steroidal antioes-
`trogen ICI 182,780 and these are summarised in Tables 9 and
`10 [26, 54-63].
`ICI 182,780 binds to ER with the same affinity as oestradiol
`[26]. It is superior to tamoxifen when tested against human
`mammary tumour cell lines in vitro [26] and cell lines trans-
`planted into nude mice [54]. It is also active in tamoxifen-
`resistant cell lines in vitro [56—58] and when transplanted into
`nude mice [63] (Table 9).
`Limited data are available from clinical studies (Table 10),
`but ICI 182,780 inhibits tumour proliferation and signifi-
`cantly reduces ER content when administered for 1 week
`before tumour resection [11]. It is active against tamoxifen-
`resistant metastatic human tumours in women and in vétro
`[(30-62]. As predicted from the nude mouse model,
`ICI 182,780 administration appears to result in a particularly
`long duration of tumour suppression. The median duration of
`response is, as yet, unknown since it has not been reached
`after 22 months of a phase II study, where tamoxifen-resistant
`tumours were treated with ICI 182,780 [61]. Preliminary data
`
`|nnoPharma Exhibit 1051.0005
`
`
`
`New Endocrine Therapies for Breast Cancer
`
`581
`
`The effect of ICI 182,780 on bone remains to be determined.
`From the limited data available from animal studies, there was
`no effect [65] or a negative effect [55] and clinical studies are
`required to resolve this issue. Thus, the specific antagonist
`ICI 182,780 may be an important new approach to antioes-
`trogen therapy. Overviews of the laboratory and clinical devel-
`opment of these compounds have been published recently
`[66, 67].
`
`AROMATASE INHIBITORS
`Aromatase inhibitors have been in the clinic for over 20
`
`years [68]. However, the toxicity profile of the major drug
`available (aminoglutethimide) led to it being used mainly as
`a third-line endocrine agent after second-line progestogen
`treatment. The major and most disturbing side-effect of
`weight gain with progestogen use led to a continued interest
`in aromatase inhibitors,
`the aim being to produce a non-
`toxic, easily administered compound with equivalent or better
`activity than aminoglutethimide, and which did not cause
`weight gain. The structure of aromatase inhibitors in the clinic
`and in clinical trials are shown in Figure 3.
`The aromatase inhibitors fall into two major groups: non-
`steroidal and steroidal compounds which appear to have dif-
`ferent mechanisms of action. Following the recognition that
`aminoglutethimide was a non-specific, reversible inhibitor of
`several cytochrome P4503 including aromatase,
`the quest
`began for more selective and more potent non-steroidal com-
`pounds. Rogletimide was an early example which showed
`improved enzyme (and pharmacological——no CNS effects)
`selectivity, but only comparable potency (Figure 3). Also,
`like aminoglutethimide, pyiidoglutethimide is a potent liver
`enzyme inducer and enhances its own metabolism. Fadrozole
`represented a major advance in potency (ca 500-fold) and
`selectivity, but the latter is not complete and 18-hydroxylase
`inhibition becomes apparent towards the upper end of the
`aromatase inhibitory dose response curve in women. Vorozole,
`letrozole and anastrozole (all
`triazole derivatives) combine
`potency and high selectivity for aromatase and have no dis-
`cernible effects on adrenal function at the maximally effective
`aromatase inhibiting doses. The latter three drugs have been
`shown to reduce circulating oestradiol levels in postmeno-
`pausal women to the limits of detection of the most sensitive
`assays currently available (Table 11) [69—75]. Although they
`are intrinsically reversible enzyme inhibitors, the long plasma
`half-lives of letrozole and anastrozole enable continuous
`
`enzyme inhibition to be achieved with simple once daily
`dosing. The question arises as to whether greater suppression
`of aromatase activity as is shown by such low hormone levels
`will result in greater response to treatment. An important issue
`for further research is the apparent lack of cross—resistance
`between various methods of aromatase inhibition (see
`Table 12) [76—80].
`The steroidal substrate analogue 4 hydroxyandrostenedione
`(Lentaron), was one of the first examples to be described and
`is the first to be developed. It has high enzyme selectivity, but
`poor oral bioavailability due to its high first-pass metabolism
`and,
`it
`is, therefore, provided as a parenteral formulation
`(twice monthly). Weak ‘hormonal’ (androgenic) effects are
`discernible in animals in the form of gonadotrophin sup-
`pression and in humans in the lowering of SHBG, although
`the latter was only seen with large twice daily oral dosing.
`Plomestane and exemestane represent second generation ster-
`oidal type of inhibitors and offer the potential for oral dosing.
`
`|nnoPharma Exhibit 1051.0006
`
`Table 9. Preclinical data for ICI 182,780 in comparison with
` tamoxifen
`
`Assessment
`
`Effect
`
`Oestrogen receptor binding
`
`Equal to oestrogen
`
`Tumour antagonism
`
`~ MCF-7 cells in vizro twice
`as active
`— MCF-7 cells in nude mice,
`growth suppressed twice as
`long
`— Animal tumours?
`
`Peripheral antagonist/agonist — Immature rat uterus
`ratio
`assay~complete
`antagonist. No agonist
`activity
`Reduced cancellous bone
`in rats
`~ No effect on bone in rats
`— No effect on
`gonadotrophins
`
`Alternative mechanisms of ~ None reported
`action
`
`[Ref.]
`
`[26]
`
`[26]
`
`[54]
`
`[26]
`
`[55]
`
`[65]
`[56]
`
`Activity against tamoxifen- — Yes
`[56-
`resistant cell lines
`58]
`
`
`Table 10. Clinical results with ICI 182,780
`
`
`
`Assessment
`
`Effect
`
`[Ref.]
`
`Activity first line
`
`Activity in tamoxifen-
`resistant tumours
`
`— No studies
`— Reduces tumour proliferation
`and ER before surgery
`
`— PR 7/ 19(37%), NC 6/ 19
`(32%)
`— 19 patients treated
`~ Median duration of response
`>22/12
`— Active in vim:
`
`Side-effects
`
`— None of note
`
`Peripheral
`antagonist/agonist ratio
`
`— ? No effect on gonadotrophins
`or SHBG
`— Inhibition of endometrial
`proliferation
`— Bone—no data
`
`[1 1]
`
`[61]
`
`[62]
`
`[61]
`
`[61]
`
`[65]
`
`[65]
`
`Pharmacokinetics
`
`~ Given by monthly depot
`— Therapeutic levels present
`throughout month
`
`[61]
`
`indicate no agonist or antagonist effects of ICI 182,780 on
`gonadotrophins, or SHBG, although uterine proliferation is
`inhibited [61, 64]. There are no clinical data with respect to
`the effect of ICI 182,780 on bone.
`Thus, in preliminary studies, ICI182,780 appears to be
`highly active, with little toxicity or agonist activity and the
`preclinical data appear to be predicting the clinical effects.
`
`
`
`582
`
`A. Howell er al.
`
`STEROIDAI. INHIBITORS
`
`First generation
`
`OH
`
`4—hydroxyandrostenedione
`
`NON-STEROIDAL INHIBITORS
`
`Second generation
`0
`
`\\
`
`Plomestane
`
`O
`
`Exemestane
`
`First generation
`
`Second generation
`
`Third generation
`
`C2H5
`
`
`
`2H5C .
`
`Nx
`//NN
`
`Vorozole
`
`mI
`
`C1 CH
`
` NH2 N Rogletimide
`
`o
`
`N
`
`o
`
`o
`
`N
`
`0
`
`Aminoglutethimide
`
`/
`
`Fadrozole
`
`CN
`
`|
`
`Anastrozole
`
`Figure 3. Structural formulae of aromatase inhibitors clinically available or in clinical trial.
`
`NC
`
`CN
`
`All three compounds interact covalently with the aromatase
`enzyme during the first oxidation cycle and cause irreversible
`inhibition. This provides prolonged peripheral aromatase inhi-
`bition in spite of rapid plasma clearance of the drug (cross-
`reactivity of plomestane and its metabolites complicates sim-
`ple oestrogen measurements and has slowed its development).
`None of the second and third generation aromatase inhibi-
`tors in trial have the toxicity associated with aminoglutethim-
`ide. The major side-effects are mild gastrointestinal disturb-
`ance and hot flushes in approximately 40% of patients [72,
`73, 81-83].
`A major advantage of the new aromatase inhibitors com-
`pared with progestogens is the difference in weight gain. In a
`recently reported three-arm phase III trial of anastrozole 1 mg
`
`or 10 mg compared with megestrol acetate 160 mg, more than
`30% of patients treated with megestrol acetate had weight
`gain of 5% or more and over 10% had a weight gain of
`10% or more. Furthermore, the weight gain experienced with
`rnegestrol acetate increased over time (Figure /1) [83].
`The results of many of the clinical
`trials with the new
`aromatase inhibitors are shown in Table 13 [72, 73, 81-87].
`Except for the randomised trials, the response rates should be
`viewed with caution as the response rate in small phase II
`studies can be greatly influenced by patient selection. With
`the exception of the comparison of tamoxifen with fadrazole
`[82], all aromatase inhibitors have been tested as second— or
`third—line treatments in ER—positive or ER unknown patients.
`Three phase II trials of exemestane are currently ongoing in
`
`|nnoPharma Exhibit 10510007
`
`
`
`New Endocrine Therapies for Breast Cancer
`
`583
`
`Table 11. Suppression of serum oestradiol, oestrone and oestrone sulphate by second and third
`generation aromatase inhibitors
`
`Exemestane (FCE 24304)
`Pyridoglutethimide
`Fadrozole (CGS 16949A)
`Vorozole (R83842)
`Letrozole (CGS 20267)
`Letrozole (CGS 20267)
`Anastrozole (ZD 1033)
`
`Dose
`(mg/day)
`
`Oestradiol
`(%)*
`
`Oestrone
`(%)*
`
`25
`1600
`2-16
`5
`0.1-2.5
`0.1—5
`1
`
`28
`50
`65
`11
`211'
`<10]
`151'
`
`35
`17
`27
`45
`21']
`<10]
`151'
`
`Oestrone
`sulphate
`(%)*
`
`39
`—~
`30
`31
`~
`<10]
`81'
`
`[Ref]
`
`[69]
`[70]
`[71]
`[72]
`[73]
`[74]
`[75]
`
`*Percentage of baseline. ']'At or below detection limit of assay.
`
`74] and is now also in a phase III trial in a three-arm study
`comparing letrozole 0.5 mg and letrozole 2.5 mg against
`megestrol acetate.
`Two phase III studies of anastrozole (Arimidex) have been
`completed and the early results reported fo