`© 1994 Kluwer Academic Publishers. Printed in the Netherlands.
`
`Report
`
`High dose toremifene in advanced breast cancer resistant to or relapsed
`during tamoxifen treatment
`
`Seppo Pyrhonenf Ritva Valavaara} Jouni Vuorinen3 and Alajos Hajba3
`I Department of Radiotherapy and Oncology, Helsinki University Central Hospital; 2 Department of
`Radiotherapy and Oncology, Tarku University Central Hospital; and 3 Orion Corporation Farmos,
`R & D Pharmaceuticals
`
`Key words: advanced breast cancer, antiestrogenic therapy, tamoxifen resistance, second—line treatment,
`toremifene
`
`Summary
`
`Fifty patients with advanced breast cancer refractory to prior tamoxifen therapy were assigned to investiga-
`tional treatment with high-dose toremifene administered 120 mg orally twice a day. Treatment was generally
`well tolerated. The majority (80%) of the patients had no side effects, and among the remaining 10 patients
`reported side effects were mostly mild and/or transient. Two objective tumor responses were observed: one
`complete response (CR), duration 6.2 months, and one partial response (PR), duration 8 months. The re-
`sponse rate was thus 4% (95% CI: 0.5 to 14%). In addition 3 patients experienced a mixed response, some
`metastatic sites responding, while at other sites disease progressed; 22 patients had disease stabilization for > 2
`months. A subset analysis disclosed that a small subgroup of patients, including 7 patients in this study, who
`had achieved CR at some of the sites during preceding tamoxifen therapy, experienced a long progression-
`free time during high dose toremifene treatment. The median time to progression in this subgroup of patients
`was 9.4 months (95% CI: 3.8 to 9.4) as opposed to 2.1 months (95% CI: 2.0 to 2.8) for all the remaining 43
`patients, which is a significant decrease in disease progression (p < 0.03). Such results reveal that although this
`kind of second-line hormonal treatment with high dose toremifene cannot be recommended for all tamoxifen
`failures, there might be a subset of patients, i.e. those who achieve CR in some lesion during tamoxifen ther-
`apy, who benefit from this type of treatment.
`
`Introduction
`
`Toremifene is a chlorinated triphenylethylene de-
`rivative, chemically related to tamoxifen. At least
`five different phase III trials comparing tamoxifen
`with toremifene in postmenopausal patients with
`breast cancer are currently underway or under
`analysis [1]. In preclinical studies some differences
`have been observed between tamoxifen and tore-
`
`mifene in efficacy as well as in toxicity in favor of
`toremifene [2]. Especially in high doses, toremifene
`has been less toxic both in animal experiments [2]
`and in clinical phase I [3, 4] and II studies [5—7]. Dai—
`ly doses up to 240 mg have been well tolerated even
`during prolonged treatment [5—7]. Also some dose-
`response relationship has been observed in animal
`
`model tumors, high doses of toremifene being more
`effective than low and moderate doses [2]. Even es-
`
`Address for offprints: S. Pyrhonen, Department of Radiotherapy and Oncology, Helsinki University Central Hospital, Haartmaninkatu
`4, SF-00290 Helsinki, Finland
`
`Astrazeneca Ex. 2069 p. l
`Mylan Pharms. Inc. v. Astrazeneca AB IPR20l6-01324
`
`
`
`224
`
`S Pyrhénen et al.
`
`trogen-receptor (ER)-negative tumors like murine
`uterine sarcoma have responded to high doses of
`toremifene [2]. The subsequent phase II clinical
`studies as primary hormonal treatment in advanced
`breast cancer also reveal existence of some dose-
`
`response relationship, the highest response rate
`
`(over 60%) being achieved with a 240 mg daily dose
`[6] while a 20 mg daily dose has yielded only a 21%
`response rate in similar patients [8, 9]. All these ob-
`servations have focused interest on high dose tore-
`mifene as a second— or third-line treatment in ad-
`
`vanced breast cancer for tamoxifen-refractory pa-
`tients. Preliminary observations from the U.K. re-
`
`vealed that high-dose toremifene might be an
`effective second-line treatment for such patients
`
`[10]. Consequently this confirmatory study was ini-
`tiated to provide more knowledge of the feasibility
`of high—dose toremifene as second-line treatment
`for patients with advanced breast cancer resistant
`to tamoxifen or patients relapsed during tamoxifen
`therapy.
`
`Patients and methods
`
`Patients
`
`Fifty consecutive postmenopausal women with ER-
`positive tumors (2 10 fmol/mg prot) were recruited
`into this study between June 1986 and April 1990.
`All had advanced disease: inoperable primary or
`
`metastatic breast cancer progressing during tamox-
`ifen treatment. Prior chemotherapy, except adju-
`
`vant treatment, was not allowed. Other prerequi-
`sites included performance status 2 50% (Karnof—
`
`sky), life expectancy >3 months, measurable or
`evaluable disease, no evidence of severe heart, liver
`or renal disease nor uncontrolled diabetes. The dis-
`
`ease could either be primarily resistant to tamoxi-
`fen treatment or relapsed after a response. The
`study was approved by the local ethical committees
`and verbal informed consent was obtained from all
`
`the patients. The main patient characteristics and
`localization of lesions are depicted in Table 1.
`Twelve patients had only soft tissue lesions, i.e. cu-
`taneous, subcutaneous, or lymph node metastases,
`while the majority of the patients had more ad-
`
`in prognostically less favorable
`vanced disease,
`sites. Four patients were treated previously with ad-
`juvant chemotherapy. Twenty—four patients had dis-
`ease progression during adjuvant tamoxifen treat-
`
`ment and 26 patients during the treatment of ad-
`vanced disease. Median disease free interval among
`these two groups of patients was 20.2 and 30.0
`months, respectively. This difference was not statis-
`tically significant (p = 0.3, Mann—Whitney test).
`
`Treatment
`
`Toremifene was administered as 120 mg orally
`twice a day. Toremifene treatment was started im-
`mediately (next day) after cessation of tamoxifen
`
`therapy. The treatment was discontinued on signs of
`progressive disease, or intolerable side effects, or
`
`according to the patient’s desire.
`
`Evaluation
`
`The pretreatment evaluation included physical ex-
`
`Table 1. Patient characteristics
`
`Patients entered
`Evaluable
`
`Mean age (range)
`Mean postmenopausal period (range)
`Median Karnofsky index (range)
`Receptors
`Median ER (range)
`
`Median PR (range)
`
`50
`48
`
`68.9 years (40-83)
`19.0 years (2.5—39.4)
`80% (50-100)
`
`62.0 fmol/mg prot
`(11-921)
`39.0 fmol/mg prot
`(0—1502)
`
`Localization of lesions (total: 85 sites)
`Primary
`Local relapse
`Soft tissue
`Visceral
`Skeletal
`
`Distribution of lesions in 50 patients
`Only soft tissue
`Only visceral
`Visceral + soft tissue
`
`Only skeletal
`Skeletal + primary or soft tissue or
`visceral
`
`1
`13
`20
`24
`27
`
`12
`8
`3
`
`4
`
`23
`
`AstraZeneca Ex. 2069 p. 2
`
`
`
`High dose toremifene in tamoxifemresistant breast cancer
`
`225
`
`Table 2. Overall responses
`
`Response
`
`No. of patients (%) Time—to—progression,
`in months
`
`CR
`PR
`SD 2 5 months
`SD < 5 months
`PD
`NE
`
`1 ( 2)
`1 ( 2)
`9 (18)
`13 (26)
`24 (48)
`2 ( 4)
`
`5.8
`8.2
`5.5—17.0
`2.8- 4.3
`0.9- 23
`1.1- 1.9
`
`amination, laboratory tests, chest radiography, ul-
`trasonography of the liver, and bone scan. Patients
`came for clinical assessment every 4 weeks during
`the first 16 weeks and thereafter every 8 weeks. Le-
`
`sions evaluable by physical examination, routine X-
`rays, or ultrasonography were checked during ev-
`ery Visit. Otherwise chest X—rays, ultrasonography
`of the liver, and bone scan were repeated every 6
`months. The tumor response and duration of re-
`sponse were evaluated according to UICC criteria
`{11]. The evaluation of adverse effects followed the
`World Health Organization guidelines [12].
`
`Statistical analysis
`
`An estimate for the true response rate and an exact
`
`95% confidence interval for the true response was
`calculated. The distribution of time—to—progression
`was estimated by the Kaplan—Meier method. Median
`times—to-progression with 95% confidence intervals
`(C1) are reported. Kaplan—Meier curves were com-
`
`Table 3. Patients with mixed responses, achieving partial or com-
`plete response at any site
`
`Patient
`
`Duration of
`treatment (wks)
`
`Localization
`
`Response
`
`H—15
`
`H—22
`
`T-8
`
`16
`
`8
`
`8
`
`lymph node
`skeletal
`liver
`skeletal
`soft tissue and
`skin
`
`lungs
`liver
`
`PR
`PD*
`CR
`PD*
`
`PR
`
`PD
`PD
`
`pared with a log rank—test. An estimate for the true
`hazard ratio and a 95 % CI for the true hazard ratio
`were calculated in order to assess the difference in
`
`time-to-progression between the two groups under
`consideration. A hazard ratio of 1 is indicative of
`
`identical proportions of patients with a subsequent
`event of interest such as disease progression at a giv-
`en time point. A hazard ratio of less than 1 is indica-
`tive that a smaller proportion of patients progressed
`
`in the first group of patients compared to the second
`group of patients at a given time point.
`
`Results
`
`Response
`
`A1150 patients were evaluable for toxicity. Two pa-
`tients discontinued the treatment in less than 2
`
`months and were not evaluable for response. The
`antitumor effect of toremifene in all 50 patients is
`
`presented in Table 2. There were two objective re-
`sponses; response rate 4% (95 % CI: 0.5 to 14%).
`One patient with lung and pleural metastases
`achieved a complete response (CR) of 6.2 months’
`duration. She succumbed to chronic cardiovascular
`
`disease while in complete remission. The heart dis-
`ease was diagnosed already prior to the toremifene
`treatment, so her death was not considered to be
`
`attributed to that drug. In another patient a partial
`response (PR) in multiple skin metastases was ob-
`served, duration 8.0 months. In addition to these
`
`two objective responses, 22 patients (44%) had a
`stable disease (SD) for longer than 2 months. Alto-
`gether 9 of these stabilized diseases remained stable
`
`longer than 5 months, the longest duration being 17
`months. In addition to the 2 overall objective re-
`
`sponses, another 3 patients experienced mixed re-
`
`sponses, i.e. although some of the lesions regressed,
`the disease progressed in other sites. The metastatic
`
`sites and the corresponding responses of these
`mixed—responders are presented in Table 3.
`
`Time-t0—progressi0n
`
`* increase in size of lytic lesion in x—rays.
`
`The median time—to-progression (TTP) of all the
`
`Astrazeneca Ex. 2069 p. 3
`
`
`
`(71%) with CR at any site during previous tamoxi-
`fen treatment suffered no disease progression be-
`fore 5 months (Table 4). Figure 1 demonstrates the
`differences in probability of managing without pro-
`gression among the 7 patients who had experienced
`CR in some of the evaluable lesions during previous
`
`tamoxifen treatment and all other patients. These 7
`had a median TTP of 9.4 months (95 % CI: 3.8 to 9.4)
`as opposed to 2.1 (95% CI: 2.0 to 2.8) for all the
`other 43 patients. This difference is statistically sig-
`nificant (x2 = 5.0, p = 0.03, Log rank—test). No other
`apparent factor such as age, length of postmeno-
`pausal period, receptor content, the initial disease
`free interval, or localization of the lesions was asso-
`
`ciated with this prolongation of TTP.
`
`Toxicity
`
`The treatment was generally well tolerated. Forty
`patients experienced no side effects, and among the
`other patients side effects were mainly mild or mod-
`erate (grade I or II) and sweating mostly only tran-
`sient (Table 5). In 2 cases the treatment was discon-
`tinued within 2 months without evidence of disease
`
`progression; one of these patients had thrombosis
`in the left thigh, which obviously had been there pri-
`or to this treatment due to major surgery on that leg
`
`(prophylactic bone fixation a.m. Kiintscher). The
`other patient, 74, experienced a cerebrovascular in-
`sult one month after the start of treatment. Relation
`of this event to the treatment cannot be excluded.
`
`226
`
`S Pyrhdnen et al.
`
`—— CR in some lesions
`
`Remaining Patients
`
`
`
` 0.6
`
`Probability
`
`0.4
`
`0.2
`
`1 _ _ _ _ _ _-
`
`0
`
`5
`
`10
`
`15
`
`20
`
`Time to Progression (in months)
`
`Fig. I. Time-to-progression in 7 patients with CR in some lesions
`during previous tamoxifen therapy vs. all the remaining 43 pa-
`tients. Difference is statistically significant (P = 0.03, Log rank-
`test).
`
`treated patients was 2.8 months (95 % CI: 2.0 to 3.5).
`Comparison of TTP to the outcome for previous ta-
`moxifen therapy revealed some differences. The
`patients were first divided into 5 groups: one group
`suffered progressive disease while receiving adju-
`vant tamoxifen (24 patients), and the other 4 groups
`were defined according to the response of evalua-
`ble metastases during tamoxifen therapy (26 pa-
`tients). As performed elsewhere [13], we selected 5
`months’ TTP as a cut—point of prognostic signifi-
`cance. Differences were as follows: only 3 of 24 pa-
`
`tients (13%) who experienced relapse during adju-
`vant treatment had TTP over 5 months, in contrast
`
`to 8 of 26 (31%) patients who received tamoxifen
`for metastatic disease (Table 4). Interestingly, all 3
`patients with CR for all sites and 5 of 7 patients
`
`Table 4. Prediction of prolonged time-to—progression (TTP) by response to preceding tamoxifen therapy
`
`Tamoxifen treatment and its outcome
`
`No. of patients
`
`No. (%) of patients with TTP over 5 months
`during treatment with toremifene
`
`Adjuvant treatment
`Response of evaluable disease
`PD
`SD
`PR
`CR
`
`CR in any of the lesions*
`
`24
`
`7
`12
`4
`3
`
`7
`
`3 (13)
`
`2 (29)
`2 (17)
`1 (25)
`3 (100)
`
`5 (71)
`
`* Including three patients with overall response SD and one with overall response PR.
`
`AstraZeneca Ex. 2069 p. 4
`
`
`
`High dose toremifene in tam0xifen—resistant breast cancer
`
`227
`
`Discussion
`
`In postmenopausal patients with metastatic breast
`cancer, second-line treatment with another hor-
`
`monal drug is usually feasible when the first—line
`therapy fails. Some clinical [10] and preclinical stud-
`ies [2], particularly, reveal that high dose toremi-
`fene may be active in tumors in which tamoxifen is
`ineffective. Preclinical studies have demonstrated
`
`that high concentrations of toremifene may have a
`direct receptor-independent oncolytic effect, i.e. be
`cytostatic or cytotoxic against tumors without estro-
`gen receptors [2]. Recently, new estrogen—receptor—
`independent mechanisms of antiestrogens have
`been identified. These molecules are found to be ef-
`
`fective inducers of transforming growth factor beta
`
`(TGF—beta) in fibroblastic cells [13], and TGF—beta
`excreted from stromal fibroblasts has the ability to
`inhibit growth or division of breast cancer cells [14].
`The magnitude of this TGF-beta-mediated mecha-
`nism as well as the estrogen—receptor—mediated
`
`growth-inhibition of breast cancer cells are evident-
`ly to a certain extent related to the concentration of
`an antiestrogenic drug [14]. As confirmed in pre-
`clinical [2] as well as phase I and II clinical studies
`[3~7, 10], toremifene seems to be less toxic than ta-
`moxifen in high doses although therapeutic benefit
`of high doses has not been established. Thus this
`
`type of high dose second-line treatment was consid-
`ered in principle interesting.
`In two previously published studies, experience
`with high dose toremifene as a second-line treat-
`ment after tamoxifen failure varied widely. A pre-
`
`liminary report from the U.K. [10] describes a re-
`
`Table 5. Side effects
`
`Evaluable patients
`N0 side effects
`Sweating*
`Nausea and depression*
`Dizziness*
`Thrombosis
`Fatigue
`Leucorrhea*
`
`No. of patients (%)
`
`50
`40 (80)
`6 (12)
`2 ( 4)
`2( 4)
`1 ( 2)
`1 ( 2)
`1( 2)
`
`* All these side effects were mild or moderate and sweating was
`mostly only transient.
`
`markable objective response rate (25%), while in a
`Swedish study no response was reported among 35
`patients [7]. In a recently published study from
`USA 5% (95 CI: 3% to 7%) response rate was de-
`tected [15]. Our observation accords very well with
`that study: two objective responses among 48 eva-
`
`luable patients, i.e. a response rate of 4%. The great
`difference between the aforementioned studies
`
`may be explained mainly by differences in patient
`characteristics. In the U.K. study the majority of the
`patients had fairly localized disease and had pri-
`
`marily responded or exhibited prolonged stabiliza-
`tion under tamoxifen treatment, while in the Nordic
`
`and USA studies the patients were less selected ta-
`moxifen failures. A large proportion of the patients
`in the latter studies had received tamoxifen as an
`
`adjuvant treatment or had never responded to pri-
`mary tamoxifen treatment. In addition a large pro-
`portion of the patients in these studies had very ex-
`tensive disease including visceral and bone involve-
`ment. Therefore the great difference in response
`
`rates with similar treatment may be expected.
`An objective response rate of 4 percent as in this
`study might be considered clinically non-signifi-
`cant, particularly if the treatment had considerable
`
`toxicity. The analysis, however, disclosed that 80%
`of the patients had no side effects, and the remain-
`ing 10 patients experienced mostly only mild or
`transient side effects. In this kind of palliative treat-
`ment, besides objective tumor regression, pro-
`longed stabilization of the disease may be meaning-
`
`ful as well, as suggested by Howell and coworkers
`[16]. They analyzed prognoses of patients with ad-
`vanced breast cancer under endocrine therapy and
`observed that patients with stabilized disease over 5
`
`months had prognoses indistinguishable from that
`of those achieving partial response. When analyz-
`ing TTP in the present study, we observed that, in
`
`addition to the two patients with objective re-
`sponses, another 9 patients were progression—free
`over 5 months and 2 of them even over 12 months. It
`
`is apparent that most of these patients also bene-
`fited from toremifene treatment.
`
`In this kind of phase II study, the number of pa-
`tients is a limiting factor for subset analysis. This
`study, however, confirms the observations that the
`outcome of previous endocrine therapy might be a
`
`AstraZeneca Ex. 2069 p. 5
`
`
`
`228
`
`S Pyrhénen er al.
`
`valuable predictor for selecting patients for second-
`line hormonal treatment [17]. All the three patients
`who achieved overall CR during preceding tamoxi-
`fen treatment had no disease progression until after
`
`5 months when changed to high dose toremifene. In
`addition there were 4 other patients who achieved
`CR with tamoxifen in some of their numerous
`
`metastatic sites. Altogether this subset of 7 patients
`with CR achieved in some location during tamoxi-
`
`fen treatment had significantly longer time-to-pro-
`gression than did the remaining 43 patients. Al-
`though the figures are small, this should provide
`some interest in this kind of analyses for similar
`
`other studies as well as subset meta-analysis from
`all the high dose toremifene studies. This manage-
`ment might disclose more clearly a subset of pa-
`tients in whom a second-line antiestrogenic treat-
`ment with high dose toremifene is beneficial, pro-
`viding these patients with another chance for a non-
`toxic hormonal treatment prior to changing to
`other endocrine therapies or more toxic cytostatic
`drugs.
`
`Acknowledgements
`
`The skilful secretarial assistance of Ms. Raija Vassi—
`
`nen is gratefully acknowledged. This investigation
`has been supported by a research grant from the
`Finnish Cancer Society.
`
`References
`
`1. Pyrhonen S: Phase III studies of toremifene in metastatic
`breast cancer. Breast Cancer Res Treat 16 (Suppl): S-41-
`S—46, 1990
`2. Kangas L, Nieminen A—L, Blanco G, Gronroos M, Kallio S,
`Karjalainen A, Perila M, Sodevall M, Toivola R: A new tri-
`phenylethylene compound, Fc-1157a. II. Antitumor effects.
`Cancer Chemother Pharmacol 17: 109-113, 1986
`3. Kivinen S, Miienpaa J: Effect of toremifene on clinical
`chemistry, hematology and hormone levels at different dos-
`es in healthy postmenopausal volunteers: phase I study. J
`Steroid Biochem 36: 217-220, 1990
`4. Hamm JT, Tormey DC, Kohler PC, Haller D, Green M, She-
`mano 1: Phase I study of toremifene in patients with ad-
`vanced cancer. J Clin Onco19: 2036-2041,1991
`
`Horvath G, Stendahl U, Kalling M, Ferno M, Himmelmann
`A, Hajba A: Antiestrogenic treatment of advanced and re-
`current carcinoma corporis uteri — A phase II study of tore-
`mifene. Anticancer Res 10: 323-326, 1990
`. Hietanen T, Baltina D, Johansson R, Numminen S, Hakala
`T, Helle L, Valavaara R: High dose toremifene (240 mg dai-
`ly) is effective as first line hormonal treatment in advanced
`breast cancer. An ongoing phase II multicenter Finnish-Lat-
`vian cooperative study. Breast Cancer Res Treat 16 (Suppl):
`S-37-S-40, 1990
`Jonsson P-E, Malmberg M, Bergljung L, Ingvar C, Ericsson
`M, Ryden S, Nilsson I, Johansson Terje 1: Phase II study of
`high dose toremifene in advanced breast cancer progressing
`during tamoxifene treatment. Anticancer Res 11: 873-876,
`1991
`
`Pyrhonen S, Valavaara R, Heikkinen M, Rissanen P, Blanco
`G, Nordman E, Holsti LR, Hajba A: Treatment of advanced
`breast cancer with 20 mg toremifene, a phase II study. Pre-
`liminary communication. J Steroid Biochem 36: 227-228,
`1990
`
`Valavaara R, Pyrhonen S: Low-dose toremifene in the treat-
`ment of estrogen—receptor-positive advanced breast cancer
`in postmenopausal women. Curr Therap Res 46: 966-973,
`1989
`
`Ebbs SR, Roberts JV, Baum M: Alternative mechanism of
`action of ‘anti-oestrogens’ in breast cancer. Lancet II: 621,
`1987
`
`Hayward JL, Rubens RD: Assessment of response to ther-
`apy in advanced breast cancer. Br J Cancer 35: 292-298,1977
`Miller AB, Hoogstraten B, Staquet M, Winkler A: Report-
`ing results of cancer treatment. Cancer 47: 207-214, 1981
`Colletta AA, Wakefield LM, Howell FV, van Roozendaal
`KEP, Danielpour D, Ebbs SR, Sporn MB, Baum M: Anti-
`oestrogens induce the secretion of active transforming
`growth factor beta from human fetal fibroblasts. Br J Cancer
`62: 405-409, 1990
`Colletta AA, Wakefield LM, Howell FV, Danielpuor D,
`Baum M, Sporn MB: The growth inhibition of human breast
`cancer cells by a novel synthetic progestin involves the in-
`duction of transforming growth factor beta. J Clin Invest 87:
`277-283, 1991
`Vogel CL, Shemano I, Schoenfelder J, Gams RA, Green ~
`MR: Multicenter phase II efficacy trial of toremifene in ta-
`moxifen-refractory patients with advanced breast cancer. J
`Clin Oncol 11: 345-350,1993
`Howell A, Mackintosh J, Jones M, Redford J, Wagstaff J ,
`Sellwood RA: The definition of the ‘no change’ category in
`patients treated with endocrine therapy and chemotherapy
`for advanced carcinoma of the breast. Eur J Cancer Clin On-
`C0124: 1567-1572, 1988
`Henderson IC: Endocrine therapy of metastatic breast can-
`cer. In: Harris JR, Hellman S, Henderson IC, Kinne DW
`(eds) Breast Diseases, 2nd Ed. JB Lippincott Company, Phi-
`ladelphia, 1991, pp 559-603
`
`10.
`
`12.
`
`14.
`
`16.
`
`17.
`
`AstraZeneca Ex. 2069 p. 6