Vol. 4, 697-711, March [998
`
`Clinical Cancer Research 697
`
`Tamoxifen-resistant Fibroblast Growth Factor-transfected MCF-7
`
`Cells Are Cross-Resistant in Vivo to the Antiestrogen
`ICI 182,780 and Two Aromatase Inhibitors‘
`
`Sandra W. McLeskey, Lurong Zhang,
`Dorraya El-Ashry, Bruce J. Trock,
`Cecilia A. Lopez, Samir Kharbanda,
`Christopher A. Tobias, Lori A. Lorant,
`Rachel S. Hannum, Robert B. Dickson, and
`Francis G. Kernz
`Lombardi Cancer Center [S. W. M.. L. Z.. B. J. T.. C. A. L.. S. K..
`C. A. T.. R. S. H.. D. E-A.. R. B. D.. F. G. K.]. Departments of
`Biochemistry and Molecular Biology [D. E-A.. F. G. K.]. Cell
`Biology [R. B. D.. L. Z.]. Medicine [B.}. T.]. and Pharmacology
`[S. W. M.]. and the School of Nursing [S. W. M.]. Georgetown
`University Medical Center. Washington. D. C. 20007
`
`ABSTRACT
`
`Although the antiestrogen tamoxifen has been the
`mainstay of therapy for estrogen receptor (ER)-positive
`breast cancer, successful treatment of responsive tumors is
`often followed by the acquisition of tamoxifen resistance.
`Subsequently, only 30-40% of patients have a positive re-
`sponse to second hormonal therapies. This lack of response
`might be explained by mechanisms for tamoxifen resistance
`that sensitize ER pathways to small amounts of estrogenic
`activity present in tamoxifen or that bypass ER pathways
`completely. To elucidate one possible mechanism of tamox-
`ifen resistance, we treated ovariectomized tumor-bearing
`mice injected with fibroblast growth factor (FGF)-trans-
`fected MCF-7 breast carcinoma cells with the steroidal an-
`
`tiestrogen ICI 182,780 or one of two aromatase inhibitors,
`4-OHA or letrozole. These treatments did not slow estrogen-
`independent growth or prevent metastasis of tumors pro-
`duced by FGF-transfected MCF-7 cells in ovariectomized
`nude mice. FGF-transfected cells had diminished responses
`to ICI 182,780 in vitro, suggesting that autocrine activity of
`the transfected FGF may be replacing estrogen as a mito-
`
`Received 7/3/97: revised ll/26/97: accepted 12/10/97.
`The costs of publication of this article were defrayed in pan by the
`payment of page charges. This article must therefore be hereby marked
`aziwrriselneltt
`in accordance with 18 U.S.C. Section 1734 solely to
`indicate this fact.
`‘This work was supported by NIH Grants CA50376 (to F. G. K.).
`CA092l8 (to F. G. K. and S. W. M.). CA53l85 (to F. G. K. and
`R. B. D.). CA66l54 (to S. W. M.). CA’/1465 (to D. E-A.). and Cancer
`Center Grant CA5l()08; American Cancer Society Grant IRG-193 (to
`S. W. M.): U.S. Army Medical Research and Material Command Grants
`DAMD I7-94-4l72 (to D. E-A.)
`and DAMD l7-94-J-4173 (to
`S. W. M.): and a Susan Komen Foundation Fellowship (to L. Z.).
`2To whom requests for reprints should be addressed. at Southem
`Research lnstitute. P. 0. Box 55305. 2()(X) Ninth Avenue South. Bir-
`mingham. AL 35255-5305. Phone: (205)581-2480; Fax: (2()5)58l-
`2877; E-mail: kern@sri.org.
`
`genie stimulus for tumor growth. ER levels in FGF trans-
`fectants were not down-regulated, and basal levels of tran-
`scripts for estrogen-induced genes or of ER-mediated
`transcription of estrogen response element (ERE) luciferase
`reporter constructs in the FGF expressing cells were not
`higher than parental cells, implying that altered hormonal
`responses are not due to down-regulation of ER or to FGF-
`mediated activation of ER. These studies indicate that estro-
`
`gen independence may be achieved through FGF signaling
`pathways independent of ER pathways. If so, therapies di-
`rected at the operative mechanism might produce a thera-
`peutic response or allow a response to a second course of
`antiestrogen treatment.
`
`INTRODUCTION
`
`Because conventional therapy is not usually curative in
`clinical breast cancer. development of tamoxifen resistance.
`in
`which breast tumors previously growth-inhibited by tamoxifen
`become refractory.
`represents an important
`therapeutic di-
`lemma. However, the development of tamoxifen resistance is
`not necessarily associated with progression to an ER’-negative
`phenotype. In many cases of clinical tamoxifen resistance, ER
`expression may be retained (1-4). implying that the resistance is
`due an alteration in activity of the tamoxifen/ER complex.
`Tamoxifen resistance in such a case could result from three
`
`possible mechanisms that. according to present knowledge,
`would not preclude successful
`treatment with an alternative
`hormonal therapy. First. alterations in the ER could arise. which
`might diminish or extinguish inhibitory responses to tamoxifen.
`leaving only its partial agonist effects to predominate (5-8).
`Second. tamoxifen resistance arising in the setting of an intact
`ER could be a result of altered intratumoral tamoxifen metab-
`
`olism. which might produce more estrogenic metabolites locally
`(7, 9—l 1). Third, available tamoxifen could be sequestered by an
`increase in antiestrogen binding sites not associated with ERs
`( I2). As mentioned. in each of these three instances. substitution
`of a hormonal therapy different from tamoxifen might result in
`a clinical response. Two such alternative therapies used in this
`report are steroidal estrogen antagonists. such as ICI 182.780.
`which lack the partial agonist activity of tamoxifen. and aro-
`matase inhibitors, which inhibit endogenous estrogen produc-
`tion by all tissues. depriving the ER of its ligand.
`Although the mechanisms of tamoxifen resistance de-
`
`" The abbreviations used are: ER. estrogen receptor: FGF. fibroblast
`growth factor:
`IMEM.
`improved minimal essential medium: X-gal.
`5-bromo-4-chloro-3-indoyl-B-D-galactopyranoside: FBS.
`fetal bovine
`serum: 4-OHA. 4—hydroxyandrostenedione; NK. natural killer: CCS.
`charcoal-stripped calf serum: ERE. estrogen response element: CAT.
`chloramphenicol acetyltransferasez RT. reverse transcription.
`
`MYLAN PHARMS. INC. EXHIBIT 1005 PAGE 1
`
`

`
`698 ICI 182.780 Effects on FGF-transfected MCF-7 Cells
`
`scribed above should be amenable to alternative honnonal ther-
`
`apy, early results for small numbers of tamoxifen-resistant pa-
`tients have shown that only about 30-40% of such patients have
`a positive response to subsequent ICI 182,780 or aromatase
`inhibitor therapy (13-20). These data imply alternative mecha-
`nisms for tamoxifen resistance. Constitutive production of au-
`tocrine growth factor(s) or growth factor receptors by tumor
`cells has been proposed as a mechanism for tamoxifen resist-
`ance that may or may not
`involve ER pathways. Evidence
`supporting this hypothesis is gained from the acquisition of
`estrogen-independent growth in tumor models,
`including the
`one used in this report. in which growth factors or growth factor
`receptors have been overexpressed in estrogen-dependent breast
`carcinoma cell
`lines (21-26). In addition. recent clinical data
`showing decreased efficacy of tamoxifen in treating tumors
`overexpressing c-erbB2 (27) supports a role for growth factor
`signaling in clinical tamoxifen resistance. Because some growth
`factor signaling pathways. including the ERB-B pathway. have
`been shown to interact with ER signaling pathways (25. 28-32).
`increased growth factor signaling could be one mechanism by
`which cells could become sensitive to previously ineffective
`amounts of estrogenic stimulation produced by the partial ago-
`nist activity of tamoxifen itself or its estrogenic metabolites.
`above. In cases in which such interactions have been demon-
`
`strated. the growth factor and ER pathways may act collabora-
`tively (25), making the final outcome susceptible to pharmaco-
`logical manipulations of either pathway and implying that
`second line honnonal therapies might have an effect. However.
`increased autocrine or intracrine growth factor signaling might
`also bypass the need for ER-mediated growth stimulation in
`tumor cells or affect stromal components of the tumor. such as
`endothelial or immune cells (33-36). to alter the tumor envi-
`ronment in ways conducive to tumor growth.
`In either case.
`alternative hormonal therapies might not be effective.
`Recently, cell-specific coactivators and corepressors have
`been identified for steroid hormone receptors. including the ER,
`which may influence steroid receptor-induced transcription pos-
`itively or negatively (37, 38). Thus. the activity of tamoxifen in
`inhibiting or even stimulating tumor growth might depend on
`the relative expression of various stimulatory or inhibitory co-
`factors in a particular tumor (39. 40). However. transient trans-
`fection experiments suggest that tamoxifen-resistant tumors pro-
`duced by such mechanisms should still be sensitive to pure
`antiestrogens (40).
`FGFs and their receptors have been shown to be present
`with high frequency in breast cancer specimens (41-50). Evi-
`dence for a possible role for FGF signaling in the estrogen-
`independent growth of breast tumors is gained from study of
`clonal and polyclonal FGF-transfected MCF-7 cell lines. which
`are capable of fonning large. progressively growing tumors in
`ovariectomized or tamoxifen-treated nude mice. Moreover, the
`
`FGF-transfected cells are metastatic, forming micrometastases
`in lymph nodes.
`lungs, and other organs (21. 22. 51). The
`estrogen-independent and tamoxifen-resistant growth of FGF-
`transfected MCF-7 cells suggests an interaction between FGF
`signaling pathways and ER—activated pathways that could occur
`at the level of the ER itself or at the end point of both pathways.
`where they impinge on growth mechanisms. If FGF-mediated
`growth pathways bypass the ER pathway to affect growth di-
`
`rectly, we would expect that growth would be unaffected by
`hormonal treatments devoid of agonist activity. We therefore
`sought to determine the sensitivity of the estrogen-independent
`tumor growth of FGF-transfected MCF-7 cells to ICI 182.780 or
`aromatase inhibitors. In contrast to what was seen with ERB-B
`
`that FGF-mediated pathways
`signaling pathways. we report
`appear to provide an alternative growth stimulatory signal that is
`not dependent on ER activation.
`
`MATERIALS AND METHODS
`Cell Lines. FGF-transfected MCF-7 cell lines have been
`
`described previously (21. 22. 51. 52). Briefly, the ML-20 clonal
`cell line is a MCF-7-derived cell line that is stably transfected
`with a [1162 expression vector. The in vitra and in viva growth
`characteristics of ML-20 cells are indistinguishable from wild-
`type MCF-7 cells (51). and >90% of the cells routinely stain
`positive for B-galactosidase expression by X-gal staining (52).
`MKL-F (FGF-4-transfected; Ref. 52) and FGF-1 clone 18 (FGF-
`1-transfected) cells (22) resulted from the stable transfection of
`the ML-20 clonal cell line with expression vectors for FGF-4
`(also known as hst-1/K-FGF) and FGF-1 (also known as acidic
`FGF or aFGF). respectively. Both cell lines continue to stably
`express B-galactosidase. allowing effects of FGF overexpres-
`sion on metastatic capability to be assessed by X-gal staining of
`organs and tissues of tumor-bearing mice. The MKL-4 cell line
`was derived by transfecting wild-type MCF-7 cells (of similar
`passage number used for the ML-20 transfection) with an ex-
`pression vector for FGF-4, which produced the clonal MKS-1
`cells (21). These cells were then retransfected with an expres-
`sion vector for lacZ. yielding MKL-4 cells (51). Cells were
`maintained in IMEM (Biofluids, Rockville. MD) supplemented
`with 5% FBS in a humidified. 37°C. 5% CO2 incubator in
`routine culture until used for tumor cell injection.
`Drugs.
`ICI 182,780 was kindly donated by Dr. Alan
`Wakeling of Zeneca Pharmaceuticals (Macclesfield, England).
`and was administered s.c. at a dose of 5 mg in 0.1 ml of vehicle
`every week. For the experiment depicted in Fig. 1, powdered
`drug was first dissolved in 100% ethanol and spiked into
`warmed peanut oil (Eastman Kodak. Rochester. NY) to give a
`final concentration of 50 mg/ml. For the experiments depicted in
`Fig. 1. B and C. 50 mg/ml preformulated drug in a vehicle of
`10% ethanol. 15% benzyl benzoate. 10% benzyl alcohol,
`brought to volume with castor oil, was supplied by B. M. Vose
`(Zeneca Pharmaceuticals). 4-OHA was donated by Angela Bro-
`die (University of Maryland. Baltimore, MD) and was admin-
`istered s.c. at a dose of 1 mg/mouse/day 6 days of the week in
`a vehicle of 0.3% hydroxypropylcelluose. Letrozole was do-
`nated by Dr. Ajay Bhatnagar (Novartis, Ltd., Basil, Switzerland)
`and was administered via gavage at a dose of 1 mg/mouse/day
`6 days of the week in a vehicle of 0.3% hydroxypropylcellulose.
`Sustained-release (60 day) pellets containing 5 mg of tamoxifen
`were obtained from Innovative Research of America (Sarasota,
`FL) and implanted s.c. in the interscapular area at the time of
`tumor cell injection.
`Tumor Cell Injection. The procedure for tumor cell
`injection has been described previously (21). Briefly.
`tumor
`cells were scraped into their normal growth medium. and viable
`cells were quantified using trypan blue exclusion. The cells were
`
`MYLAN PHARMS. INC. EXHIBIT 1005 PAGE 2
`
`

`
`Clinical Cancer Research
`
`699
`
`resuspended in their normal growth medium at a density of
`66.7 X 10“ cells/ml, and 0.15 ml (containing 10 million cells)
`were used to inject ovariectomized mice (nude or I2eige/nude/
`xid) into the mammary fat pad. For the experiment involving
`MKL-4 cells and nude mice (Fig. 1A). each mouse was injected
`bilaterally into the thoracic mammary fat pads (two injections
`per mouse). There were seven mice in the vehicle group and five
`mice in each treatment group. For the experiments involving
`MKL-4 cells and beige/nude/xid mice (Fig. 2). four tumor cell
`injections were given, two on each side in the thoracic mam-
`mary fat pad and two in the inguinal mammary fat pad; treat-
`ment groups consisted of four mice. For the experiments involv-
`ing MKL-F and FGF-1, clone 18 cells (Fig. 1. B and C). each
`mouse was injected once in the right thoracic mammary fat pad.
`There were seven mice in the each vehicle group, and treatment
`groups consisted of five or six mice each. Tumors resulting from
`the injections were measured twice weekly in three dimensions
`using calipers. Tumor volume is the product of the largest
`dimension. the orthogonal measurement. and the tumor depth. as
`described previously (21). Because the FGF-l-transfected clone
`18 cell line produces tumors that in some cases are surrounded
`by a fluid-filled sac that confounds tumor measurements (22).
`these tumors were measured postmortem by weighing them.
`Determination of Metastasis. Organs were harvested
`from tumor-bearing animals. fixed briefly. and stained with
`X-gal
`reported previously (51) and viewed through a dissect-
`ing microscope (Olympus SZH). Clusters of blue-staining cells
`were identified as micrometastases. In accordance with previous
`results, no macrometastases were identified (21. 22, 51. 53).
`Growth Assays. Anchorage-dependent and anchorage-
`independent growth assays were performed as described (21).
`Briefly. for anchorage-dependent growth. cells were plated in
`24-well culture dishes at a density of 10,000 cells/well for the
`time course experiments (Fig. 4) and 20,000 or 30,000 cells/well
`for the concentration-response experiments (Fig. 5). For growth
`in FBS. following overnight attachment. treatments were added
`at the indicated concentrations. and cells were counted on the
`indicated days. For growth assays under estrogen-depleted con-
`ditions, cells were stripped of estrogens during a 24-h period the
`day following plating by changing the medium four times to
`phenol red-free IMEM supplemented with 5% CCS (21). We
`have found that this stripping procedure allows complete re-
`moval of estrogens without substantial proliferation of cells
`before treatments are added. Following the stripping procedure.
`on day 0. treatments were added. and counting of cells was done
`as above.
`
`Doubling times were determined according to the follow-
`ing equation: doubling time = t3 - t,/3.32log(N3/N,). where N3
`and N, are the number of cells at times [2 and t,. respectively. N,
`and N: are the means of quadruplicate determinations.
`Anchorage-independent assays in FBS-containing medium
`were done as described previously (21 ). For experiments using
`estrogen-depleted conditions. cells were stripped of estrogens
`over a 24-h period as described above before being plated in soft
`agar. Colonies greater than 60 jun were counted using an
`Omnicon 3600 Image Analysis system.
`ER Assays.
`[3H]Estradiol binding has been described
`previously (54. 55). Briefly. cells grown to 70% confluence
`were stripped with twice daily medium changes over 4 days
`
`with 5% CCS in phenol red-free IMEM. The prolonged strip-
`ping method allows ERs to become up-regulated to maximal
`levels. Cells were harvested. washed sequentially at 4°C with
`serum-free. phenol red-free IMEM followed by TEG (l() mM
`Tris. pH 7.4, 1 mM EDTA, 10% glycerol). and resuspended in 1
`ml of TEG plus 1 mM DTT. 0.5 M NaCl and a cocktail of
`protease inhibitors (1 mg/ml leupeptin. 77 pg/ml aprotinin.
`1
`pg/ml pepstatin A). A whole-cell extract was prepared by ho-
`mogenization with 40 strokes in a Teflon—glass Dounce homog-
`enizer followed by centrifugation at 105.000 X g for 30 min.
`Protein content of the supernatant was detennined by the
`method of Bradford (56), and protein concentrations were ad-
`justed to 2 mg/ml. Extracts were incubated with 10 nM |"H]l7B-
`estradiol with or without a 100-fold excess of unlabeled estra-
`
`diol for 16 h at 4/C. Unbound ligand was removed by absorption
`with dextran-coated charcoal followed by centrifugation. Ali-
`quots of the supernatant were counted in a Beckman liquid
`scintillation counter.
`
`Northern Blots. Cells were grown to 50% confluence in
`IMEM supplemented with 5% FBS and then stripped of estro-
`gens as described for the growth assays. above. Treatments of
`0.1% ethanol (vehicle) or 10”‘ M 17B-estradiol
`in the same
`medium were added. Cultures were harvested after 3 days of
`treatment. and RNA was extracted using RNAzol B (Tel-Test.
`Inc.) according to the manufacturer's directions. Thirty ug of
`each RNA were subjected to electrophoresis in a 1.2% formal-
`dehyde/agarose gel and transferred to nylon (Hybond-N. Am-
`ersham Corp, Arlington Heights. IL) by capillarity. “P-labeled
`antisense riboprobes for pS-2, GAPDH. and cathepsin D were
`prepared and sequentially hybridized to the membrane overnight
`at 65°C [hybridization buffer was 50% fonnamide. 50 mM
`Na2HPO_,. 0.8 M NaCl. 10 mM EDTA. 2.5 X Denhardt's solu-
`tion (IX Denhardt’s = 0.02% polyvinylpyrrolidone. 0.()2%
`BSA). 0.2% SDS. 400 p.g/ml yeast
`tRNA. and 400 ug/ml
`sonicated salmon sperm DNA with 10“ DPM/ml of the appro-
`priate probe]. The membrane was washed three times in 0.1%
`SDS/0.1 X SSC at 80°C for the PS-2 and cathepsin D probes,
`and 75°C for the GAPDH probe. Autoradiograms and Phospho-
`rlmager (Molecular Dynamics Model 44581) quantitation of
`individual hybridization signals were obtained between the se-
`quential hybridizations. For the results depicted in Fig. 7. A and
`B. Phosphorlmager values obtained for PS-2 or cathepsin were
`normalized to those obtained for GAPDH.
`
`Progesterone Receptor mRNA Determination by RT-
`PCR. The primers for human progesterone receptor that pro-
`duce a 205-bp PCR product have been described previously
`(57). The human GAPDH primers that produce a 437-bp PCR
`product are as follows: 5’-AAG GTC GGT GTG AAC GGA
`TTT G-3’ (sense) and 5’-TGG TGC AGG ATG CAT TGC
`TG-3' (antisense). RT-PCR was performed with 0.1 ug of test
`RNAs. except T47D cells, where 0.02 p.g was used. using the
`GeneAmp RNA PCR kit (PE Applied Biosystems. Foster City.
`CA) according to the manufacturer's instructions with the fol-
`lowing modifications: the RT reaction was primed with 0.0625
`p.M random hexamers in a volume of 40 p.l. with 2 ptl each of
`355-labeled UTP and 358-labeled ATP (each 3000 Ci/mmol. 10
`p.Ci/p.l. Amersham Corp.) substituted for water in the reaction.
`Then. 20 pl of each RT reaction were transferred into two tubes
`for separate GAPDH and progesterone receptor PCR reactions.
`
`MYLAN PHARMS. INC. EXHIBIT 1005 PAGE 3
`
`

`
`700 ICI l82.780 Effects on FGF-transfected MCF—7 Cells
`
`Cycle analyses using RNA from ML-20, estradiol-treated cells
`(the highest expressors of progesterone receptor) revealed that
`amplification remained logarithmic at 35 cycles for the GAPDH
`reaction and 40 cycles for the progesterone receptor reaction.
`making these assays semiquantitative. The GAPDH PCR reac-
`tion was performed using standard reagent conditions recom-
`mended by the manufacturer and cycles of 95°C for 45 s and
`50°C for 45 s for 35 cycles. For the progesterone receptor PCR
`reaction, final MgCl2 concentrations were adjusted to 1.25 mM,
`and 0.25 M acetamide was included. Cycles were of 95°C for
`45 s and 50°C for 45 s for 40 cycles. GAPDH and progesterone
`receptor reaction products were first visualized by ethidium
`bromide staining following electrophoresis in a 2% agarose gel.
`Products were then electrophoresed on a 4—20% acrylamide gel
`that was subjected to both autoradiography and Phosphorlmager
`quantitation as described above.
`Transient Transfection, Luciferase, and CAT Reporter
`Assays. ML-20 and clone 18 cells were plated in 6-well
`plates, allowed to attach overnight. and stripped of estrogens in
`a procedure similar to that for the growth assays (see above).
`Following stripping, cells were transfected by the calcium phos-
`phate.
`low-CO2 method (58). The luciferase plasmids pGLB-
`MERE or pGLB-MNON were obtained by inserting an approx-
`imately l.48-kb fragment containing a glucocorticoid response
`element-deleted mouse mammary tumor virus promoter with
`either a substituted double consensus ERE (MERE) or the same
`sequence with the ERE palindromes scrambled (MNON) (59)
`into the Hindlll site of pGLB (Promega, Madison. WI). Each
`dish received 2.5 ug of either pGLB-MERE or pGLB-MNON
`and l.0 p.g pCMV-CAT, which directs constitutive expression
`of CAT, cotransfected as a control for transfection efficiency.
`Following transfection. each well was washed twice with PBS
`and incubated for 48 h in medium containing vehicle (0.0l%
`ethanol). 10"’ M estradiol, IO"7 M ICI 182,780, a combination
`of E3 and ICI, 10 ng/ml FGF-l plus 10 ug/ml heparin, or a
`combination of FGF, heparin, and ICI 182,780. (Duplicate sam-
`ples of each treatment were used.) Cells were lysed and assayed
`for luciferase activity using the Luciferase Reporter Gene Assay
`(Boehringer Mannheim. Indianapolis, IN) according to the man-
`ufacturer‘s instructions. Luciferase values, expressed as relative
`light units, for each sample were corrected for background by
`subtracting the value of lysates of untransfected cells prepared
`in parallel. CAT expression was assayed using the CAT ELISA
`(Boehringer Mannheim. Indianapolis, IN) according to the man-
`ufacturer‘s instructions. Protein content of the lysates was de-
`tennined using the BCA Protein Assay Reagent (Pierce, Rock-
`ford, IL). Luciferase and CAT values, normalized for protein,
`were used to calculate mean specific relative light units/ng CAT.
`Statistical Analyses. Statistical methods used for tumor
`growth have been described previously (53. 60). For Figs. 1 and
`2. only mice surviving at
`the end of the experiment were
`included in the analysis. When no tumor developed from a
`particular injection, tumor volume was recorded as zero. The
`repeated measures ANOVA (60) was used to compare tumor
`volumes among the treatment groups using measurements taken
`over the entire time course of the experiment. In addition, final
`tumor volumes (or weights in the case of clone 18) were
`compared among treatment groups at the end of each experi-
`ment using ANOVA. For analysis of metastasis in Table I, for
`
`each transfectant, analysis of covariance was used to compare
`the effects of treatment on total metastases, total distant metas-
`tases (lung metastases plus other metastases), lymph node me-
`tastases.
`lung metastases. and other metastases. The analyses
`were all conducted with final tumor volume (or weight for the
`clone 18 cells)
`included in the model as a covariate. The
`analyses considered the effects of all treatments simultaneously,
`as well as the effects of individual
`treatment comparisons
`(which were adjusted for multiple comparisons using Dunnett’s
`method). For each transfectant, the effect of final tumor volume
`(or weight for clone 18) on the number of metastases was
`evaluated using linear regression (for each of the categories of
`metastasis described above). In Fig. 3, paired I tests were per-
`formed comparing control and transfected cells under different
`conditions of treatment. For the anchorage-dependent growth
`assays depicted in Fig. 4, we examined the effect of treatment on
`the rate of cell growth, using linear regression with an interac-
`tion between time and treatment. To compare cell growth rates
`and doubling times among the cell lines under specific treatment
`conditions, nested linear regression models were used. For Fig.
`6, ANOVA was used to determine significant differences in ER
`binding among cell lines.
`
`RESULTS
`
`Estrogen-independent Growth of Tumors Produced by
`FGF-transfected MCF-7 Cells Is Not Inhibited by Treat-
`ment with a Pure Antiestrogen or with Aromatase Inhibi-
`tors. We have previously shown that both FGF-l- and FGF-
`4-transfected MCF-7 cells fomi progressively growing tumors
`in ovariectomized nude mice, as well as in similar mice
`treated with tamoxifen (21, 22. 53). Although ovariectomized
`mice could be expected to have substantially lower levels of
`estrogenic compounds than reproductively intact mice, some
`estrogens are synthesized at extraovarian sites, such as adre-
`nal gland. liver, fat, or possibly the tumor itself. The trans-
`fected cells evidently still possess ERs, because they respond
`to estrogen and tamoxifen administered to the mice, as well
`as to these compounds used in tissue culture (21, 22). To test
`the hypothesis that growth of the FGF-transfected cells in
`ovariectomized or tamoxifen-treated nude mice is due to
`
`increased sensitivity to the small amounts of estrogens still
`present in ovariectomized nude mice. we tested the ability of
`a pure antiestrogen, ICI l82.780, and two aromatase inhibi-
`tors, 4-OHA and letrozole, to inhibit the estrogen-indepen-
`dent tumor growth produced by these FGF-transfected cell
`lines.
`
`In a first experiment to test the above hypothesis, FGF-4-
`transfected MKL-4 cells were injected as before, and the mice
`were treated with vehicle.
`tamoxifen, or ICI 182,780. There
`were no significant differences in tumor volume among the
`treatment groups considered over the entire time course of the
`experiment (P = 0.72) or at the final time point (Fig. IA; P =
`0.72). Treatment with ICI 182.780 did not inhibit tumor growth
`below that achieved in vehicle-treated mice (P = 0.675). Thus,
`the failure of [CI 182.780 to inhibit the estrogen-independent
`growth exhibited by this cell line supports the hypothesis that
`such growth does not result from small amounts of estrogenic
`
`MYLAN PHARMS. INC. EXHIBIT 1005 PAGE 4
`
`

`
`5
`
`A MKL-4 (FGF-4)
`
`80
`
`400
`
`
`
`83
`
`90
`
`90
`
`I00
`
`§
`
`
`
`
`
`toO0
`
`TumorVolume(mm3) §
`
`Clinical Cancer Research 701
`
`B MKL-F (FGF-4)
`
`
`
`23
`
`45
`
`63
`
`
`
`TumorWeight(mg)
`
`.t>8
`
`C FGF-l. clonel8 Hi)
`
`onOO
`
`—N88
`
`90 I00 [00 E 100
`0 iii
`ii
`
`I500
`
`
`
`
`
`
`
`MeanTumorVolume(mm3) §3
`
`27
`
`48
`
`61
`
`Days after Injection
`
`Days after Injection
`
`Treatment
`
`1 VEH R\\\\\‘ 4—OHA
`- TAM W LET
`
`IC1
`
`Fig. I Growth of FGF-transfected MCF-7 cells in ovariectomized nude mice is not inhibited by treatment with [Cl 182.780. 4-OHA. or letrozole.
`Ten million cells from the indicated cell lines were injected into the mammary fat pads of ovariectomized nude mice treated with vehicle (VEH); a
`5-mg, 60-day-release tamoxifen pellet (TAM); ICI 182.780. 5 mg s.c. every week (ICI);
`I mg of 4-OHA s.c. per day 6 days of the week (4—0HA);
`or 1 mg of letrozole per day via gavage 6 days of the week (LET). Columns. group mean; bars SE. Numbers above each column are the percentages
`of injections resulting in measurable tumors at that time point. A. volumes of tumors produced by one clonal FGF-4—transfectcd MCF-7 cell line.
`MKL-4. at the indicated number of days following tumor cell injection. B. volumes of tumors produced by a second clonal FGF-4-transfected MCF-7
`cell line. MKL-F. at the indicated number of days following tumor cell injection. C. weights of tumors produced by a clonal FGF-1-transfected MCF-7
`cell line. FGF-1. clone 18. weighed after sacrifice of the animals 28 days after tumor cell injection. (Because the FGF-1 producing MCF-7 cells may
`form fluid-filled sacs around the tumor. confounding tumor measurements before sacrifice. only postmortem weights are presented here.)
`
`growth stimulation achieved by extraovarian estrogen produc-
`tion.
`
`We wished to assess the effect of ICI 182,780 on metastasis
`as well as on tumor growth. In spite of its retention of the
`transfected IacZ expression plasmid, the MKL-4 cell line be-
`comes heterogeneous over time with respect to B-galactosidase
`expression, such that a few cells have high expression. but most
`are negative (52). We therefore used a second clonal FGF-4-
`transfected MCF-7 cell
`line, MKL-F. the B-galactosidase ex-
`pression of which is stable, for a subsequent experiment involv-
`ing FGF-4-transfected MCF-7 cells. Because FGF-1 has also
`been shown to produce estrogen—independent
`in viva growth
`when transfected into MCF-7 cells (22), we also included a
`clone of FGF-1-transfected cells designated clone 18, the B-ga-
`lactosidase expression of which is also stable. For these exper-
`iments, two aromatase inhibitors, 4-OHA (61, 62) and letrozole
`(63), were also used to inhibit extraovarian synthesis of estro-
`gens.
`In agreement with the experiment using MKL-4 cells de-
`picted in Fig.
`IA, when the FGF-4-transfected MKL-F cells
`were used, there were no differences in tumor volume among
`treatment groups over all
`time points (P = 0.382), and [C1
`182,780 did not decrease tumor growth below that obtained in
`vehicle-treated animals (Fig. 1B; P = 0.837 for the last time
`point). In addition, neither 4-OHA nor letrozole decreased tu-
`mor growth below vehicle-treated levels (P = 0.571 and 0.931
`for the last time point, respectively).
`FGF-1—transfected clone 18 cells form tumors that are
`
`accurate tumor volume measurements during the course of the
`experiment. Consequently. when these cells were used (Fig.
`1C). only terminal tumor weights were analyzed with ANOVA.
`As with the MKL-4 and MKL-F cells. ICI 182.780 did not
`
`inhibit estrogen-independent tumor growth in the clone 18 cells
`(P = 0.977). Administration of ICI 182.780 to animals injected
`with ML-20 cells, a clonal line of B-galactosidase-transfected
`wild-type MCF-7 cells (51). also produced no effect when
`compared with vehicle-treated animals [i.e., no progressively
`growing tumors were obtained in either case (data not shown)l.
`In other, separate experiments. a polyclonal population of con-
`trol vector-transfected ML-20 cells that forms progressively
`growing tumors in estrogen-supplemented mice (22) did not
`fonn tumors in either untreated or ICI 182,780-treated animals.‘
`Thus. the continued progressive in viva growth of FGF-trans-
`fected cells in ovariectomized animals treated with either a pure
`antiestrogen or aromatase inhibitors demonstrates that the estro-
`gen-independent growth of these cells in untreated ovariecto-
`mized nude mice is not due to estrogenic activity produced at
`extraovarian sites.
`Because ICI 182,780, 4-OHA, and letrozole were without
`effect in the experiments described above, we injected repro-
`ductively intact female mice for 2 weeks with these compounds
`at the same doses used in the above experiments to observe for
`activity in preventing effects of endogenous estrogens on the
`
`sometimes surrounded by a fluid-filled sac (22, 53). preventing
`
`4 Unpublished results.
`
`MYLAN PHARMS. INC. EXHIBIT 1005 PAGE 5
`
`

`
`702 ICI 182.780 Effects on FGF-transfected MCF-7 Cells
`
`Table I Metastasis of FGF-transfected MCF-7 cells is not inhibited
`by treatment with ICI 182,780 or aromatase inhibitors
`Mice were sacrificed and tumors and organs were subjected to
`X-gal staining as described previously (51). Mice bearing tumors pro-
`duced by injection of MKL-4 cells were sacrificed at 61 days; for
`MKL-F tumors, mice were sacrificed after 64 days: and for FGF—l clone
`18 tumors. mice were sacrificed after 28 days.
`Metastatic site
`
`Positive lymph nodes/
`lymph nodes
`examined
`
`Lung Other
`
`3
`4
`5
`
`Injected cells/ No. of tumor-
`treatment
`bearing mice
`MKL-4
`Vehicle
`TAM”
`ICI
`182.780
`MKL—F
`Vehicle
`TAM
`ICI
`182,780
`4-OHA
`LET
`
`6
`5
`3
`
`3
`3
`
`3/10
`5/18
`4/23
`
`0/27
`4/20
`0/14
`
`0/13
`1/ I2
`
`5/24
`3/23
`2/13
`
`5/18
`4/22
`
`3
`2
`3
`
`3
`3
`1
`
`0
`0
`
`2
`3
`3
`
`2
`3
`
`7
`2
`4
`
`I
`0
`0
`
`0
`0
`
`0
`3
`1
`
`1
`0
`
`FGF—l clone
`18
`Vehicle
`TAM
`ICI
`182.780
`4-OHA
`LET
`
`6
`6
`4
`
`5
`5
`
`" TAM. tamoxifen; LET. letrozole.
`
`endometrium. Uteri harvested from mice injected with either ICI
`182,780, 4-OHA. and letrozole weighed le