Oncogene (1997) 15, 547-447
`© 1997 Stockton Press
`All rights reserved 0950-9232/97 $12.00
`
`b>
`
`The effect of HER-2/neu overexpression on chemotherapeutic drug
`sensitivity in human breast and ovarian cancercells
`
`Mark D Pegram, Richard S Finn, Karo Arzoo, Malgorzata Beryt, Richard J Pietras and
`Dennis J Slamon
`
`Division vf Hematology-Oncology, University vf California at Lox Angeles, School af Medicine, Las Angeles, California 9009S,
`USA
`
`Recent studies indicate that oncogenes may be involved
`in determining the sensitivity of human cancers
`to
`chemotherapeutic agents. To define the effect of HER-
`2/neu oncogene overexpression on sensitivity to che-
`motherapeutic drugs, a full-length, human HER-2/new
`cDNA was introduced into human breast and ovarian
`cancer cells.
`/n vitro dose-response curves
`following
`exposure to 7 different classes of chemotherapeutic
`agents were compared for HER-2- and control-trans-
`fected cells, Chemosensitivity was also tested in vive for
`HER-2-
`and control-transfected human breast
`and
`ovarian cancer xenografts in athymic mice, These studies
`indicate that HER-2/neu overexpression was not suffi-
`cient
`to induce intrinsic, pleomorphic drug resistance.
`Furthermore, changes in chemosensitivity profiles result-
`ing from HER-2/neu transfection observed in vitra were
`cell
`line specific,
`Jn vivo, HER-2/neu-overexpressing
`breast and ovarian cancer xenografts were responsive to
`different classes of chemotherapeutic drugs compared to
`control-treated xenografts with no statistically significant
`differences between HER-2/neu-overexpressing and non-
`overexpressing xenografts. We found no instance in
`which HER-2/neu-oyerexpressing xenografts were ren-
`dered more sensitive to chemotherapeutic drugs in vivo.
`HER-2/neu-overexpressing xenografts consistently ex~-
`hibited more rapid regrowth than control xenografts
`following initial
`response to chemotherapy suggesting
`that a high rate of tumor cell proliferation rather than
`intrinsic drug resistance may be responsible for
`the
`adverse prognosis associated with HER-2/neuw over-
`expression in human cancers.
`
`breast
`Keywords: HER-2/new (c-erbB-2);
`ovarian cancer; drug resistance; chemotherapy
`
`cancer,
`
`Introduction
`
`The human HER-2/neu (c-erbB-2) proto-oncogene
`encodes a
`185 kD transmembrane receptor
`tyrosine
`kinase which is homologous to, but distinct from,
`the
`epidermal growth factor receptor (EGFR) as well as
`other members of the type [ receptor tyrosine kinase
`family (ie. HER-3 and HER-4). Sequence identity
`between members of
`this
`receptor
`family in
`their
`extracellular, and intracellular tyrosine kinase domains
`is 40-60% and 60-80%, respectively (Rajkumar and
`Gullick, 1994). Amplification of the HER-2/neu gene
`
`
`‘
`Correspondence: DJ Slamon
`Received 19 August 1996; revised 21 April 1997, accepted 22 April
`1997
`
`occurs in ~25—30% of human breast and ovarian
`cancers resulting in overexpression of the gene product,
`and this molecular alteration, when present,
`is an
`independent predictor of both relapse-free und overall
`survival in these diseases (Pauletti et a/., 1996; Slamon
`et ql., 1987).
`In breast cancer, overexpression of the
`HER-2/neu gene has been associated with a number of
`other adverse prognostic factors including; advanced
`pathologic stage (Seshadri ef a/., 1993), number of
`axillary lymph node metastasis (Slamon ef al., 1987),
`absence ofestrogen and progesterone receptors (Quenel
`et al., 1995: Querzoli et al., 1990; Barbareschi er al.
`1992),
`increased S-phase fraction (Borg ef al., 1991:
`Anbazhagan er af, 1991), DNA ploidy (Stal er at..
`1994: Lee et al., 1992), and high nuclear grade (Berger
`et al., 1988; Poller et a/., 1991). A role for the HER-2/
`neu alteration in metastasis has also been suggested
`given the increased occurrence of visceral metastasis
`(Kallioniema er af,
`1991) and higher
`incidence of
`micrometastatic bone marrow disease (Pantel et al.,
`1993)
`im patients with HER-2/new overexpression,
`In
`addition, expression of HER-2/neuw has prognostic
`significance in patients with gastric (Yonemura ef al.,
`1991), endometrial (Berchuck er a/., 1991; Hetzel er al,
`1992; Lukes er
`a@/., 1994; Saffari et ai., 1995), and
`salivary gland cancers (Semba ef a/., 1985; Press ef al.,
`1994). The exact role alteration of HER-2/neu receptor
`expression plays in the pathogenesis of these cancers
`remains unclear.
`Retrospective data from two large clinical trials in
`breast cancer suggests an association between HER-2/
`new overexpression and resistance lo chemotherapy.
`Results from the Intergroup Study 0011 (Allred ef ai..
`1992) and the International (Ludwig) Breast Cancer
`Study Group (Gustersonef a/., 1992) led investigators
`to conclude that node-negative breast cancer patients
`whose tumors contain HER-2/neu overexpression have
`a less favorable prognosis due to a lack of response to
`adjuvant
`cyclophosphamide
`(CPA), methotrexate
`(MTX). and 5-fluorouracil
`(5-FU)-based chemother-
`apy (CMF). In addition, in a study of 68 patients with
`advanced
`breast
`cancer, Wright
`and
`colleagues
`reported a shortened survival for patients with HER-
`2/mey overexpression who were treated with mitoxan-
`trone despite the fact
`that
`response rates between
`HER-2/new-overexpressing
`and
` non-overexpressing
`tumors were
`similar,
`50% vs
`58%,
`respectively
`(Wright e¢ al., 1992). A study of HER-2/new over-
`expression in epithelial ovarian cancer demonstrated
`that patients whose tumors had the alteration were
`more
`likely to fail chemotherapy with CPA and
`carboplatin (CBDCA) (Felip er al., 1995). Conversely,
`in a clinical series reviewed by Klijn ef a/. patients with
`Genentech 2139
`Genentech 2139
`Celltrion v. Genentech
`Celltrion v. Genentech
`IPR2017-01122
`IPR2017-01122
`
`

`

` “ HER-2/neu and chemotherapeutic drug sensitivity
`
`rr
`MD Pegram et al
`538
`
`metastatic breast cancer and amplification of the HER-
`2/neu
`gene
`had
`a
`superior
`response
`to CMF
`chemotherapy (75%) compared to patients without
`HER-2/new amplified tumors (45%) and the smedian
`length of progression-free survival
`from the start of
`chemotherapy was superior in patients whose tumors
`exhibited amplification (Berns ef a/., 1995; Klijn ef al.,
`1993), Recently, data from the Cancer and Leukemia
`Group-B demonstrated
`thal
`node-positive
`breast
`cancer patients with HER-2/new overexpression de-
`rived
`a
`benefit
`from doxorubicin
`(DOX)-based
`adjuvant
`chemotherapy which
`is
`dose-dependent
`indicating that HER-2/new overexpression may be
`associated with an increased response to DOX (Muss
`et al., 1994). In composite, the clinical data to date are
`somewhat contradictory and do not adequately define
`what role, if any, HER-2/mev overexpression plays in
`chemotherapy response. Moreover,
`there
`is
`little
`experimental data to address this potentially impor-
`tant question.
`In one
`study evaluating m vitro
`chemosensitivity
`in HER-2/neu-transfected MCF7
`breast carcinoma cells, no significant difference in
`response to either 5-FU or DOX was seen, while
`HER-2 overexpression was associated with a 2—4-fold
`increase in resistance to cisplatin (CDDP) (Benz er al,,
`1992).
`In another study, HER-2/neu transfection of
`MDA-MB-435 cells conferred resistance to paclitaxel
`(TAX) via an mdr-l-independent mechanism (Yu ef wl.,
`1996).
`Jn
`vitro studies of
`lung cancer
`cell
`lines
`demonstrated
`an
`association
`between HER-2/new
`expression levels and intrinsic chemoresistance to six
`different chemotherapeutic drugs (Tsai ef al., 1993),
`and transfection of HER-2/neuw cDNA into one lung
`cancer
`cell
`line
`resulted in an increase
`in drug
`resistance (Tsai er al., 1995),
`In an attempt to further define the effect of HER-2/
`neu overexpression on sensitivity to chemotherapeutic
`drugs
`in human breast and ovarian cancers, we
`introduced a full-length, human HER-2/new cDNA,
`via a retroviral expression vector,
`into four different
`breast cancer cell lines; MCF7, MDA-MB-231, MDA-
`MB-435
`and BT-20,
`and
`two
`different
`ovarian
`carcinoma cell
`lines; 2008 and Caov-3. All of the
`parental cell lines used for this study contain a single
`copy of the HER-2/new gene and express basal levels of
`the gene product while
`the matched HER-2/neu
`retroviral
`transfectants overexpress
`the gene. Dose-
`response
`curves using seven different
`classes of
`chemotherapeutic agents were constructed for
`the
`HER-2/neu-overexpressing cell
`lines as well as their
`mock-transfected parental controls. The rationale for
`this
`experimental
`approach was
`to allow direct
`comparison of genetically identical parent/daughter
`cells which differ only in that one member of the pair
`overexpresses
`the human HER-2/neuw gene. This
`approach was taken to circumvent
`the difficulty of
`comparing cell
`lines derived from separate sources
`which may inherently differ
`in characteristics other
`than HER-2/new overexpression which could impact on
`drug sensitivity. The rationale for evaluating more than
`one cell
`line representing each of these two human
`malignancies is to avoid the possibility that any given
`observation could be unique to an individual cell
`line
`rather than béing representative of a more generic
`biologic effect associated with HER-2/neu overexpres-
`sion. Finally, to avoid the possibility that any observed
`
`effects might be restricted to an in vitro setting and
`because monolayer cell culture assays may not detect
`important multicellular mechanisms of drug resistance
`(Kerbel ef a/., 1994; Kerbel, 1995), chemosensitivity
`was tested in vive for breast and ovarian cancer parent/
`daughter xenografts in an athymic mouse model.
`
`Results
`
`Characterization of human breast and ovarian cancer
`cells engineered to overexpress the HER-2/neu gene
`A full-length HER-2/new cDNA was introduced via
`retroviral vector into a panel of human breast and
`ovarian carcinoma cells which are known to have a
`single copy of the HER-2/new gene and to express
`‘normal’ levels of the gene product. Breast cell lines BT-
`20 and MDA-MB-435 were established from previously
`untreated patients making them less
`likely to have
`treatment-induced chemotherapeutic drug resistance
`while the MCF7 cell
`line was established from a
`patient with prior radiation and hormonal therapy and
`the MDA-MB-231 cell line was derived from a patient
`previously treated with multidrug chemotherapy (5-FU.
`CPA, DOX, MTX, and prednisone), The ovarian
`carcinoma cell line 2008 was established from a patient
`who had not had prior chemotherapy, whereas the
`Caov-3 cell
`line was derived from a patient whose
`tumor had been exposed to prior 5-FU, DOX, and
`CPA in viva. This spectrum of cell
`lines allows for
`response data representative of a diverse group of
`human
`breast
`and
`ovarian
`cancers, HER-2/nen-
`engineered and control cells were identically infected
`using a neomycin phosphotransferase-based vector
`which either contained, or did not contain, a full-
`length HER-2/new CDNA. Retroviral
`infectants were
`selected for neomycin resistance and subjected to
`fluorescence activated cell sorting (FACS) analysis for
`detection of the p1$5"'=*" protein, Western blot analysis
`confirmed a marked increase in p185!""*? expression in
`cells engineered to overexpress the gene relative to
`mock (NEQO)-infected controls (Figure la and b). SK-
`BR-3 human breast carcinoma cells and SK-OV-3
`human ovarian carcinoma cells naturally overexpress
`the HER-2 receptor and were included in these studies
`for comparison of non-manipulated overexpressing
`
`=
`
`MDA-MB-231/HER-2
`
`MDA-MB-231/NEO
`
`Se
`
`MDA-MB-435/HER-2
`
`MDA-MB-435/NEO
`
`Caov-3/HER-2
`MCF7/HER-2
`BT-20/HER-2
`Caov-3/NEO
`2008/HER-2
`MCFT/NEO
`BT-20/NEO
`2008/NEO
`SK-OV-3
`SKBR3
`fa e@e@-- »
`3
`
`Figure 1 Western blot analysis of HER-2/new- and mock (NEQO)-
`vector infected breast (a) and ovarian (b) carcinoma cell
`lines
`demonstrating high-level expression of p18s''""* in transfected!
`cell
`lines. SK-BR-3 breast cells and SK-OV-3 ovarian cells have
`native amplification/overexpression of the HER-2/mew gene and
`ure shown as posilive controls
`
`

`

`=“
`HER-2/neu and chemotherapeutic drug sensitivity
`“-n
`MD Pegram et ai
`—eree————
`539
`
`reproducibility, all sets of in vitro assays were repeated
`at
`least
`two times. This assay yielded 4-parameter,
`sigmoidal curve fits with correlation coefficients ranging
`from 0.938 -0.999. Differences between dose-response
`curves Were assessed using 2-factor analysis of variance
`(ANOVA) of data points which fell between the [Cay
`and [Cgo, Representative data from these experiments
`
`"]SKBR3 "|MCF7/NEO _]MCF7/HER-2 _]BT-20/NEO _]BT-20/HER-2 "|MDA-MB-231/NEO
`
`t$-— + = +
`
`' a ' te
`
`|]mpa-mB-231/HER-2
`
`q
`
`HRG-B1
`
`IP: HER-2
`IB:PY —
`
`_]MDA-MB-435/NEO _]MDA-MB-435/HER-2
`
`= ie
`
`+
`
`+
`itfas.
`:
`IB:HER2 > ge 9°-Ge"-ee mt
`
`Figure 2. To demonstrate the phosphorylation state of p1gs'"="*
`in HER-2/new-transfected breast carcinoma cell
`lines,
`anti-
`phosphotyrosine immunoblots were performed following immu-
`noprecipitation with a monoclonal anti-p18s'"*** antibody both
`in the presence (>) or absence (—) of recombinant heregulin B-1
`(a). The sameblot is reprobed with anti-p185"""* (b), These data
`demonstrate constitutive tyrosine phosphorylation of p1gs!lE®?
`in SKBR3, MCF7/HER-2 and BT-20/HER-2 even in the absence
`of heregulin B-1. In mock (NEO)-transtected MCF7 and BT-20
`cells, heregulin B-]
`induced both an increase in plgs'®®?
`tyrosine phosphorylation (a) and downregulation of pigs"®®?
`expression (b). MDA-MB-231 cells exhibited neither basal nor
`heregulin-induced tyrosine phosphorylation of pl&s""®= despite
`high expression levels of the protein
`
`a
`
`o
`wu
`
`ee
`oo
`wi
`
`nN
`
`&
`=
`£
`of
`o
`oS
`3
`a
`a
`w
`™
`~
`oO
`oO
`w
`Ihe st le Se
`-—-
`+
`-
`+
`= += +
`=
`4
`
`HAG-BI
`
`IP: HER-2
`IB:PY
`
`B: HER-2 ———>
`
`b 1
`
`Figure 3. Anti-phosphotyrosine immunoblot of HER-2/new- or
`mock
`(NEO)-transfected ovarian carcinoma
`cells
`following
`jmmunoprecipitation with an anti-p185""** specific monoclonal
`antibody either in the presence (+) or absence (—) of exogenous
`recombinant heregulin B-1 (a). The same blot
`is reprobed with
`unti-pI8S"=®~ (bj). The
`data demonstrate an
`increase
`in
`pisst®*®= tyrosine phosphorylation and downregulation of
`plgs'"©"= expression on exposure to heregulin B-|
`in Caov-3/
`NEO cells. Caov-3/HER-2 cells demonstrate both basal and
`heregulin-induced tyrosine phosphorylation of pisstl®= whereas
`2008/HER-2 have neither
`increased basal or heregulin-induced
`pigsters phosphorylation despile overexpression of the protein
`
`cells (o the engineered cells. The levels of HER-2/neu
`overexpression in the engineered cells are comparable
`to, but do not exceed, the levels found in actual human
`tumors circumventing the possibility that any observed
`biologic changesare artifacts of levels of overexpression
`which do not occur
`in nature. As
`a measure of
`functional activity of p18S"&"?,
`the phosphorylation
`state of p185"*** was. assessed using immunoblotting
`techniques. Protein lysates from each of the transfected
`cell lines were subjected to immunoprecipitation with a
`pl85""** specific monoclonal antibody. These experi-
`ments were performed on cell
`lines both with and
`without prior exposure to heregulin B-1, a growth
`factor ligand cloned on the basis ofits ability to induce
`tyrosine phosphorylation of pl85"®** through the
`formation of
`HER-2/HER-3 and/or HER-2/HER-4
`heterodimeric
`¢gomplexes
`(Sliwkowski
`ef
`al.,
`1994;
`Plowman et a/., 1993). The resulting immunoprecipi-
`tates were
`then resolved
`by polyacrylamide
`gel
`electrophoresis (SDS-PAGE) and probed with an
`anti-phosphotyrosine antibody (Figures 2a and 3a).
`These results indicate that HER-2/neu cDNA transfec-
`tion results in expression of a p[85"*** protein whichis
`either constitutively tyrosine phosphorylated or can be
`phosphorylated on exposure to heregulin B-1 in each of
`the breast cell lines with the exception of MDA-MB-
`231 (Figure 2a). Similarly, ovarian Caov-3/HER-2 cells
`exhibited heregulin-induced tyrosine phosphorylation
`of pI&5"&** while 2008/HER-2 cells did not (Figure
`3a), In Figures 2b and 3b the same blots from Figures
`2a and 3a have been probed with the same anti-
`pl85"®®* antibody used for the immunoprecipitations.
`These
`results confirm overexpression of pl85'""**
`protein in the HER-2/neu-transfected cell lines, and in
`addition, demonstrate that exposure of the mock-vector
`(NEO)transfected cell lines to heregulin B-1 in most
`cases resulted in tyrosine phosphorylation as well as
`down-regulation of pI&5"*"*? expression (Figures 2b
`and 3b). The relative degree of heregulin induced
`tyrosine phosphorylation of pl85"*** correlated with
`the expression level of HER-3 in these cells. For
`example, MCF7 cells have 2.5x 10' HER-3 molecules
`per cell whereas MDA-MB-231 and 2008 cells have
`only 1.4» 10", and 1.0 10" HER-3 molecules per cell,
`respectively by quantitative ELISA (Aguilar ef al.
`manuscript
`in preparation). HER-4 expression levels
`are very low,<10' molecules/cell, relative to HER-2 or
`HER-3 in this panel of cell lines, therefore heregulin-
`induced HER-2
`phosphorylation
`appears
`to
`be
`predominantly influenced by the abundance of HER-
`2/HER-3 heterodimers in these cells. Having success-
`fully engineered the breast
`and ovarian cells
`to
`overexpress p185"***, we next evaluated the effects of
`overexpression on their sensitivity to chemotherapeutic
`drugs in vitro and in vive,
`
`Effect of HER-2j/neu overexpression on sensitivity of
`human breast and ayvarian cells to chemotherapeutic
`agents in vitro
`The effects of HER-2/neu overexpression in human
`breast and ovarian carcinomacell lines on sensitivity to
`a variety of chemotherapeutic agents was determined in
`vitro. The effective dose range for each drug ([C\o—
`1Cyy) was identified using a range often different doses,
`each tested in quintuplicate, To assure accuracy and
`
`a-
`
`i S
`
`e
`
`

`

`HER-2/neu and chemotherapeutic drug sensitivity
`MO Pegram et al
`2
`
`are shawn in Tables | and 2. These data include the
`ICco+one standard deviation and the significance level
`for differences between control (NEO) and HER-2-
`engineered
`cell
`lines.
`Introduction of neomycin
`phosphotransferase gene via the NEO control vector
`and selection in neomycin resulted in no change in
`chemosensitivity in MCF7 cells
`(data not
`shown)
`indicating thal neomycin resistance does not confer
`cross-resistance to chemotherapeutic agents in vitro.
`Clinically achievable peak plasma levels of chemother-
`apeutic drugs from standard dosing schedules used in
`humans are shown for reference in Table 1,
`HER-2/new overexpression in MCF7 breast carcino-
`ma cells resulted in a 2.5-fold decrease in sensilivily to
`the platinum analog CBDCA,as well as a twofold
`decrease in 5-FU sensitivity, Conversely, a twofold
`increase in sensitivity to TAX was noted while no
`changein sensitivity to the other four drugs tested was
`found (Table 1). These resulls are similar
`to those
`reported by Benz e/ a/. who noted a 2—3-fold decrease in
`sensitivity to CDDP but no changein sensitivity to DOX
`or 5-FU in MCF7cells which overexpress HER-2/neu
`(Benz er af., 1992). In contrast, MDA-MB-231/HER-2
`cells were rendered more sensitive to four of the seven
`drugs tested (Table 1). This increase in sensitivity ranged
`from 1.4-fold for thiotepa (TSPA) to > 100-fold for
`TAX. The BT-20/HER-2 cells were also 2-4-fold more
`sensitive to TSPA and 5-FU, but
`like MCF7/HER-2
`cells, they were more resistant to platinum compounds.
`Lastly, MDA-MB-435/HER-2cells exhibited no change
`in chemosensitivity to any of the seven classes of
`Effect of HER-2)i|expression on chemosensitiviay of
`chemotherapeutic agents tested. Among the ovarian
`breast and evaria “enografis in vivo
`carcinoma cell
`lines, Caoy-3/HER-2 cells were slightly
`more sensitive to DOX andvinblastine (VBL) compared
`ta Caoy-3/NEO: however, HER-2/neu overexpression in
`2008 cells resulted in a threefold and 7.5-fold increase in
`resistance to CBDCA and TSPA,respectively (Table 2).
`
`These results indicate that HER-2/eu overexpression
`does not produce any consistent or predictable change in
`drug sensitivity profiles in vitro across the various cell
`lines tested and underscore the necessity of evaluating
`more than one cell
`line prior
`to drawing general
`conclusions on the
`effect of
`this alteration on
`chemotherapeutic response
`in human cancer
`cells,
`Moreover,
`the differences in chemosensitivity patterns
`among
`the HER-2/neu-transfected
`cell
`lines
`did
`not appear to correlate with basal or heregulin B-1-
`induced tyrosine phosphorylation of p185"**?, Despite
`the fact that chemosensitivity in HER-2/new-overexpres-
`sing cells was cell line specific, some trends did emerge
`from the data, HER-2/neu-overexpression had no major
`effect on sensitivity lo DOX in any of the six cell lines
`tested with the exception of Caov-3/HER-2 cells where
`it was associated with a
`small
`(0.5 wm to 0.3 uM)
`but
`statistically significant
`increase
`in
`sensitivity.
`Similarly, HER-2/neu overexpression had minimal
`effects on response to etoposide (VP-16) with only
`onecel] line, MDA-MB-23], exhibiting a slight increase
`in
`sensitivity
`after
`transfection with HER-2/neu.
`Increased resistance to platinum analogs was observed
`in three of the six cell
`lines with HER-2/new over-
`expression compared to their controls. Finally, when
`agents which interfere with microtubule formation (VBL
`and TAX) were studied,
`three of six HER-2/neu-
`overexpressing cell
`lines demonstrated an increase in
`sensitivity.
`
`sensitivity
`To further evaluate and expand drug.
`studies associated with HER-2/ney overexpression,
`we developed
`an
`in
`vivo chemotherapeutic drug
`sensitivity assay which utilized serial measurements
`
`Table 1 Effect of HER-2/nou overexpression on sensitivity of human breast cells to chemotherapeutic agents in vitro”
`CDDP(ym)" DOX (um)
`SFU (uM)
`TAX (nM)
`TSPA (jim)
`VBL (nm)
`VP-I6 (uM)
`MCF7/NEO
`19.145.0
`0.394+0.03
`103434
`20,2439
`78.54 13.0
`0,93 40.09
`16.0410
`MCF7/HER-2
`AR4478*
`0.344007
`22.5460"
`9649.6"
`85.2496
`1.10.05
`1.0430
`MDA-MB-435/NEO
`13.0+1,3
`0.64009
`TO£0.7
`L240
`75.6242
`0440.02
`27+0.2
`MDA-MB-435/HER-2
`13.3423
`0.6+0,07
`9+ L2
`1240.05
`77.142.)
`0.34002
`3,202
`
`MDA-MB-231/NEO
`MDA-MB-231/HER-2
`
`21.6464
`20,3+4.0
`
`)3+0,03
`0.2+0,05
`
`50,0+90
`443+ 12.0
`
`\4.6+1.5
`0.08 +0,05"""
`
`23834174
`167.04 7.4"
`
`19.0425
`24,0
`
`N24+0,5
`34+0,5**
`
`P32
`0.24006
`278.3475.0
`$8412
`130.02 20,2
`0.174003
`3.6£03
`BT-20/NEO
`
`
`
`
`
`
`BT-20/HER-2 117.8+20.6"*=0.3400125.742,08"* 0.15+0,02 3204708" 42412 12.2402
`
`[Peak Plasma]
`Reference
`
`30)
`400)
`10.6
`940
`IO)
`5.6
`30
`(D'Incalei ev al.
`(Nelson ef af.,
`(Cohen et a,
`(Wierik ef al,
`(MacMillan et
`(Robert ef af.
`(Gormley et al,
`19R6) 1982) VOR)
`
`1987)
`
`al, 1978)
`1982)
`1979)
`a p<d.05; **, PoOOl ***, P< 0,001. "CBDCA substituted for CDDP for MCF7/NEO and MCF7/HER-2, Peak plasma concentration of
`CBDCA is 50 um (Harland et a/., 1984). Data shown are [Csy values for each drug. Error is reported as +one standard deviation. The peak
`plasma levels of each drug achievable in humans with standard dosing schedules are shown for reference
`
`Table 2 Effect of HER-2/neu overexpression on sensitivity of human ovarian cells to chemotherapeutic agents in wire”
`
` CDDP (um)" DOY (jim) 5-FU (uM) TAX(1M) TSPA (jIM) VAL (nM) VPI (pM)
`
`
`
`
`
`20.0413
`0.5005
`16.0438
`M8+6.2
`KU9+41.6
`Leo)
`1.2402
`19.1403
`0,340.04"
`15.3415
`21.7430
`85.6448
`0540.03"
`1.2403
`
`Caav-3/NEO
`Caov-3/HER-2
`
`0.54: 0.03
`Lo+03
`49418
`15202
`3.6405
`0,06 + 0.007
`1.3403
`2n08/NEO
`0440.05
`1640.7
`7.0+74"*
`1.6402
`§.3409
`0,0640.0)
`3.9+0.3"""
`2008/HER-2
`ar P<0.05; *%, P<0.01; ***, P=0.001. *CBDCAsubstituted for CDDP for 2008/NEO and 2008/HER-2. Data indicate [Cy values for each
`drug. Experimental error is reported as +one standard deviition
`
`

`

`HER-2/meu and chemotherapeutic drug sensitivity
`MO Pegram et al
`oa
`541
`
`=o wero
`sh Mernes
`
`=
`
`re
`
`ao
`Tine, Daye
`
`a
`
`1)
`
`ee
`SSpeee
`
`—-—y-—
`
`
`
`and MCF7/HER-2 was not statistically significant
`(P=0.12). Treatment with TAX also resulted in
`significant
`responses
`for MCF7/NEO and MCF7/
`HER-2 tumors compared to vehicle-treated controls.
`Mean TAX-treated T/C ratios at maximum response
`were 0.19+0,09 and 0.30+0,18 for MCF7/NEO and
`MCF7/HER-2 tumors,
`respectively (Figure 5d), and
`this difference was marginally significant
`(P=0.09),
`Finally,
`response
`to
`treatment with TSPA was
`significant for both MCF7/NEO and MCF7/HER-2
`tumors compared to control (Figure Se), but there was
`no significant difference between response of MCF7/
`NEO xenografts compared to MCF7/HER-2 xeno-
`grafts
`in
`response to TSPA (P=0.17). Additional
`analysis
`in a 2-fuctor ANOVA model
`failed to
`
`of subcutaneous human tumor xenografts growing in
`athymic mice. For the im vive studies, human breast
`(MCF7) and ovarian (2008) curcinoma cells were
`selected for testing because of their predictable tumor
`formation in
`athymic mice.
`Immunohistochemical
`analysis of sections from these tumors and Western
`blot analysis
`from cell
`lines derived
`from these
`xenografts confirmed that the relative expression level
`of HER-2/new was maintained during the course of the
`study (data not shown). Overexpression of HER-2/neu
`in MCF7
`breast
`carcinoma
`cells
`resulted
`in
`a
`significant change in their
`in vive growth character-
`istics (Figure 4), By day 50, MCF7/HER-2 tumors
`were
`2.7-fold
`larger
`than MCF7/NEO tumors
`(P=0.0001), At
`the onset of chemotherapy adminis-
`tration, animals were assigned to treatment groups
`such that initial tumor volumes were the same in each
`group (55+4 mm’). Because the MCF7/NEO xeno-
`grafts have a significant difference in inherent growth
`rate compared to MCF7/HER-2 xenografts, the ratio
`of chemotherapy-treated to untreated control
`tumor
`volume (T/C ratio) was calculated for each tumor. The
`maximum response to chemotherapy, defined as the
`point at which the T/C ratio was at a minimum, was
`determined for each individual tumor. The maximum
`drug responses for the MCF7/NEO xenografts were
`then directly compared to responses found in the
`MCF7/HER-2 xenografts.
`In the human breast cancer xenograft model, all
`five drugs tested resulted in significant responses for
`both MCF7/NEO and MCF7/HER-2 tumors com-
`pared to their
`respective untreated control
`tumors
`(P<0.05)
`indicating
`that HER-2/neu-transfected
`MCF7
`xenografts miatintain
`sensitivity
`to
`these
`chemotherapeutic drugs in vive (Figure 5). The mean
`time to point of maximum response was 17+5 days
`and was independent of the drug tested or tumor type
`(ic. NEO vs HER-2). Tumor regrowth following day
`21 uniformly occurred indicating a lack of prolonged
`response to the initial treatment. Of note was the fact
`that
`there was a significant difference in regrowth
`rates
`following responses
`to chemotherapy when
`comparing MCF7/NEO to MCF7/HER-2 tumors.
`The mean tumor doubling time following chemother-
`apy was
`14.6
`days
`for MCF?7/HER-2
`tumors
`compared to 23.8 days
`for MCF7/NEO tumors
`(P=0,0001). This demonstrates that HER-2/neu-over-
`expressing tumors maintain their proliferative advan-
`tage following exposure to chemotherapy in vivo. The
`T/C ratios at
`the point of maximum response are
`represented graphically by box plots
`(Figure
`5).
`Treatment with DOX resulted in significant responses
`for both MCF7/NEQ and MCF7/HER-2 tumors
`compared to their respective untreated control groups
`(Figure
`5a). The difference
`in
`the magnitude of
`maximum response to DOX between MCF7/NEO
`and MCF7/HER-2
`(tumors was
`not
`statistically
`significant
`(P=0.13). Treatment with CDDP also
`resulted in significant
`responses
`for both MCF7/
`NEO and MCF7/HER-2 tumors and again mean
`CDDP-ireated T/C ratios were
`not
`significantly
`Group means=\{CF7/NEO MCF7/HER-2 Significance Level
`
`different at
`the point of maximum response (Figure
`a
`0.43
`0.62
`P=0.13
`5b, P=0.12). Similarly, treatment with 5-FU resulted
`b
`0.30
`0.44
`P=0.12
`in significant responses compared to controls for both
`c
`0,33
`0.41
`P=0,12
`d
`0.19
`0.30
`P=0,09
`MCF7/NEO and MCF7/HER-2 tumors (Figure 5c);
`e
`0.27
`O38
`P=0,17
`but
`the difference in response between MCF7/NEO
`
`A =
`
`a
`
`- a
`
`a
`
`> i
`
`(NEO)
`Figure 4 Tumorigenicity of HER-2/new, of control
`véetor-infected human breast (MCF7) s.c. xenografts in female
`athymic mice (n=13-14/group). Ereor bars indicate standard
`error, MCF7/HER-2 xenografts (&) have a significant growth
`advanuige over MCF7/NEO (()) (P= 0.0001) im vive. Mice in this
`expermment were treuted with a vehicle control solution i.p,
`beginning on day 0 (12 days status post xenograft inoculation), at
`which time objectively measurable xenografts had formed
`
`(rolaiivebscontrol)
`Tumo!Response
`
`
`response (relalive to
`illustrating tumor
`Figure 5 Box Plots
`control) for MCF7/NEO (unshaded boxes) and MCF7/HER-2
`(shaded boxes) xenografts (v= 12-14 per group) in response lo
`treatment with: (a) DOX (5 mg/kg), (by CDDP (5 mg/kg). (¢) 5-
`FU (100 mg/kg), (4) TAX (15 my/ke x3), and (e) TSPA (5 mg/
`kg x 3). Error burs indicate 10" to 90" percentiles, boxes indicute
`35" to 75! percentiles, and notches indicate 95%. confidence
`intervils, Group mean T/C ratios and significance levels (Mann-
`Whitney Utest) for differences between MCF7/NEO and MCF7/
`HER-2 are as follows:
`
`

`

`HER-2/neu and chemotherapeutic drug sensitivity
`MD Pegram et.al
`
`demonstrate significant differences in the magnitude of
`response between MCF7/NEO and MCF7/HER-2
`xenografts to any chemotherapeutic agent tested over
`the time period during which responses were seen.
`Unlike MCF7 cells.
`the ovarian carcinoma cells
`2008/HER-2 had only a small growth advantage over
`2008/NEO cells (Figures 6a—c, vehicle controls),
`In
`this model. both the 2008/NEO and 2008/HER-2
`xenografts were refractory to treatment with DOX
`using two different
`treatment schedules (5 mg/kg on
`day | or 3 mg/kg on days | and 14, data not shown),
`Higher doses of DOX resulted in substantial toxicity.
`Similarly, VP-16 at a dose of 25 mg/kg on days 0, 3,
`and 7 had no effect on 2008/NEO or 2008/HER-2
`tumor growth during the 21 day observation period.
`A dose of 50 mg/kg on day 0 and day 3 did result in
`a significant response compared to untreated control
`tumors by day 6 (data not shown), however there was
`no response difference between 2008/NEO and 2008)
`HER-2 tumors, and the higher dose of VP-16 resulted
`in substantial mortality beyond day 6. Treatment of
`ovarian
`2008
`tumors with CDDP resulted
`in
`significant responses by day 6 at which lime tumor
`volumes of CDDP-treated tumors were 37% of
`controls and significant differences were maintained
`during a 21 day observation period (Figure 6a). There
`was no difference, however,
`in the degree of response
`between 2008/NEO and 2008/HER-2 ovarian xeno-
`grafts,
`thus
`the
`threefold shift
`in IC,,
`suggesting
`platinum resistance in the HER-2/new-overexpressing
`cells i vitro was not observed in vivo. Treatment of
`ovarian 2008 NEO and HER-2 tumors with TAX
`resulted
`in
`a 58% reduction in
`tumor volume
`compared to control which was apparent at day 6,
`However,
`there was no difference in response when
`comparing 2008/NEO and 2008/HER-2 tumors in-
`dicating that HER-2/new overexpression in these cells
`had no impact on sensitivity to TAX in vive (Figure
`6b). Treatment of ovarian 2008 xenografts with TSPA
`also resulted in a significant
`response compared to
`untreated control tumors. For this drug, a significant
`difference between TSPA-treated 2008/NEO and 2008/
`HER-2 tumors did emerge by day 21 with TSPA-
`treated 2008/HER-2 tumors measuring 100% larger
`than TSPA-treated
`2008/NEO tumors
`(P=0.002)
`(Figure 6c). Moreover,
`this result paralleled the in
`vitro results where a 7.5-fold increase in resistance to
`TSPA was noted in 2008 cells overexpressing HER-2/
`neu. This difference, however, appeared lo be due to
`more rapid tumor regrowth for 2008/HER-2 xeno-
`grafts following response to chemotherapy rather than
`intrinsic resistance to TSPA.
`In fact. at
`the ume of
`maximal
`response to TSPA (day 10),
`there Was no
`significant difference between 2008/NEO and 2008/
`HER-2 xenografts (P=0,17), These data paralleled the
`results seen with MCF7/HER2 xenografts where rapid
`tumor
`regrowth
`occurred
`following
`response
`to
`chemotherapy in vive.
`
`Discussion
`
`The involvement of some oncogenes in the develop-
`ment of chemotherapeutic drug resistance is suggested
`by experimental data demonstrating increased expres-
`sion of c-fos, c-mive, and c-H-ray gene transcripts in
`
` 2OONNEDect cine
`Vobume,mm
`JOWNEOTERA
`
`Tumor
`
`a De-
`
`OORMERSTSrk,
`
`Figure 6 Response of human ovarian xenografts 2008/NEO and
`2008/HER-2 to lreatment with (a) CDDP (5 mg/kg),
`(bh) TAX
`(1S me/ke <3), und (ce) TSPA (5 mp/ke=3) in female athymic
`mice, Injection of a single eyele of (hese ihree drugs resulted in
`significant responses compared to a vehicle contral for both 2008)
`NEO and 2008/HER-2 xenografts; however,
`the magnitude of
`Fespotise Wis not significantly different for 2008/NEO compured
`to 2M/HER-2 xenografts. The growth rate of TSPA-treated
`2008/HER-2 xenografts (c) was significantly greater than 204)8/
`NEOxenografts (P= 0.002) following wn initial response to TSPA
`
`cells
`carcinoma
`ovarian
`human
`cisplatin-resistant
`{Scanlon er al., 1989).
`In transfection studies, c-mye
`expression was subsequently shown to result
`in