Oncogene ( 1994). 9, 1829 - 1838
`
`© M acmilla n Press Ltd. 1994
`
`Antibody to HER-2/neu receptor blocks DNA repair after cisplatin in
`human breast and ovarian cancer cells
`, V.R. Chazin1, M.D. Pegram 1, S.B. HowelP & D.J . Slamon1
`R.J . Pietras1, B.M. Fendly2
`1 Dil•ision of H ematology-Oncology, University of California , L os Angeles, California 90024 ; 2Genentech, Inc., 460 Point San
`Bruno Blvd, S outh San Fran cisco, California 94080 ; and 1Cancer C enter, University of California , San Diego, California 92093 ,
`USA
`
`Approximately 30% of human breast and ovarian cancers
`have amplification and/or overexpression of HER-2/neu
`gene which encodes a cell surface growth-factor receptor.
`is assoc(cid:173)
`Overexpression of this receptor, p 185H£R·21••u,
`iated with poor outcome and may predict clinical res(cid:173)
`ponse to chemotherapy. Antibodies to HER-2/neu recep(cid:173)
`tor have a cytostatic effect in suppressing growth of cells
`with overexpression of pl8SHER·21••". To elicit a cytocidal
`effect, therapy with antireceptor antibody was used in
`combination with the DNA-damaging drug, cisplatin, and
`this combined treatment produced a synergistic decrease
`in cell growth. In addition, antibody mediated an in(cid:173)
`creased sensitivity to cisplatin in drug-resistant ovarian
`carcinoma cells containing multiple copies of HER-2/neu
`gene. To evaluate the mechanism for this synergy, un(cid:173)
`scheduled DNA synthesis was measured in cancer cells
`incorporation of (3H)thymidine and autoradio(cid:173)
`using
`graphy, and formation and repair of cisplatin-induced
`DNA adducts was also measured. Treatment with cis(cid:173)
`platin led to a marked, dose-dependent increase in un(cid:173)
`scheduled DNA synthesis which was significantly reduced
`by combined treatment with antireceptor antibody in
`HER-2/neu-overexpressing cells. Therapy with antibody
`to HER-2/neu receptor also led to a 35-40% reduction
`in repair of cisplatin-DNA adducts after cisplatin expo(cid:173)
`sure and, as a result, promoted drug-induced killing in
`target cells. This phenomenon which we term receptor(cid:173)
`enhanced chemosensitivity may provide a rationale for
`targeting and exploitation of overex(cid:173)
`more selective
`thus
`pressed growth factor receptors in cancer cells,
`leading to new strategies for clinical intervention.
`
`Introduction
`Growth factors and their receptors play pivotal roles in
`regulation of cell growth and differentiation (Carpenter
`& Cohen , 1979; Aaronson , 1991). There is now con(cid:173)
`siderable evidence that malignancy arises by a step-wise
`progression of genetic events that often include the
`unregulated expression of growth factor receptors or
`elements of their signaling pathways (Bishop, 1983;
`Aaronson , 1991; Harris et al., 1992). Among these
`receptors , the most frequently implicated in human
`cancers have been members of the epidermal growth
`factor (EGF) or c-erbB receptor family . The HER-2/
`(c-erbB-2) proto-oncogene encodes a 185 kDa
`neu
`
`C o rrespo ndence: Ri cha rd J. Pietras, Ph.D., M .D .
`in revised
`Received 15 Au gust 1993; accepted
`1994
`
`fo rm 15 M a rch
`
`tyrosine kinase, pl 85HER-2 "'", with
`transmembrane
`homology to EGF receptor (Coussens et al., 1985;
`Semba et al., 1985). This receptor has oncogenic poten(cid:173)
`tial which may be mediated through multiple genetic
`mechanisms including point mutations in the trans(cid:173)
`membrane domain (Bargmann et al., 1986), truncation
`of the extracellular domain or overexpression of the
`non-mutated proto-oncogene (DiFiore et al., 1987;
`Hudziak et al., 1987; Yarden & Ullrich , 1988; Aaron(cid:173)
`son, 1991 ). Moreover, amplification and/or overexpres(cid:173)
`in
`found
`is
`sion of the normal HER-2/neu gene
`25 - 30% of primary human breast and ovarian cancers
`l 989a) and , less frequently , in
`(Sla m on et al., 1987,
`other human cancers, including gastric (Park et al.,
`1989; Kasprzyk et al. , 1992) and endometrial (Ber(cid:173)
`chuck et al., 1991 ) adenocarcinom as. Most impor(cid:173)
`tantly, HER-2/neu amplification correlates with a poor
`prognosis in that patients whose tumors contain this
`alteration have a shorter disease-free survival as well as
`a shorter overall survival (Slamon, 1987; Slamon et al.,
`l 989a; Berchuck et al., 1991 ; Press et al., 19~ ;
`Seshadri et al., 1993).
`The human HER-2/neu gene is a homologue of the
`rat c-neu proto-oncogene whose activated form was
`initially identified as a dominant transforming onco(cid:173)
`gene in DNA from ethylnitrosourea-induced rat neuro(cid:173)
`glioblastomas (Shih et al., 1981 ) . Comparison of the
`transforming neu oncogene sequence with its normal
`identified a point
`rat proto-oncogene counterpart
`mutation in the transmembrane domain which confers
`increased tyrosine kinase activity to the altered p 185"'"
`gene product. This increased activity is believed to be
`responsible for cell transformation mediated by the
`mutated gene (Bargmann et al., 1986). To date, no
`analogous point mutation has been found in the HER-
`2/neu gene product in human tumors (Slamon et al. ,
`1987, l 989a ; Lemoine et al. , 1990; Lofts & Gullick,
`1992). In contrast, the alteration occurring in human
`is overexpression of a normal gene
`malignant cells
`product which is most frequently but not uniformly
`l 989a,b;
`to gene amplification (Slamon et al.,
`due
`Lemoine et al., 1990). Additionally, overexpression of a
`structurally-unaltered HER-2/neu gene leads to neop(cid:173)
`lastic transformation of both NIH3T3 cells (DiFiore et
`al., 1987; H udziak et al., 1987) and immortalized, but
`non-transformed , human breast cells (Pierce et al.,
`1991; Pietras et al., 1991 ), indicating that this altera(cid:173)
`tion may play a pathogenic role in promoting tumori(cid:173)
`genicity of non-malignant cells.
`the extracellular
`Monoclonal antibodies against
`domain of the mutated rat neu membrane receptor can
`reversibly suppress tumorigenesis by neu-transformed
`NIH3T3 cells (Drebin et al., 1988). In related studies ,
`
`1 of 10
`
`Celltrion, Inc., Exhibit 1015
`
`

`

`1830
`
`R.J . PIETRAS er al.
`
`monoclonal antibodies against portions of the extracel(cid:173)
`lular domain of the non-mutated human gene product
`can specifically inhibit the growth of human breast
`carcinoma cells overexpressing the HER-2/neu gene
`product (Hudziak et al., 1989). These antibodies may
`accomplish their growth-inhibitory effects by blocking
`a putative autocrine/ paracrine growth-stimulatory loop
`involving pl 85H FR-Vn•u receptor (Aaronson , 1991; Harris
`et al. , 1992). Studies with human heregulin , a newly(cid:173)
`identified 45 kDa
`ligand
`to pl85H ER-lNu
`receptor
`(Holmes et al., 1992), or other possible p I 85HER-2
`n•u
`ligands (Peles et al., 1992) may help to further define
`this pathway and its role in malignancy. As with the
`p J 85HER-l neu
`receptor, OVerexpreSSiOn Of epidermal
`growth factor (EGF) receptor
`is found
`in several
`human cancers and is suspected to play a role in
`tumorigenesis (Yamamoto et al., 1986). Similarly,
`monoclonal antibodies against the extracellular domain
`of EGF receptor exhibit significant antitumor activity
`among cells overexpressing this receptor (Masui et al.,
`1984). In related studies, Aboud-Pirak et al. (1988)
`identified a poorly understood but probable synergistic
`tumor cell inhibitory effect between monoclonal anti(cid:173)
`bodies to EGF receptor and the widely-used chemo(cid:173)
`therapeutic drug, cisplatin. The combined treatment
`elicited a significant reduction in both the number and
`size of tumors generated by human epidermoid carcin(cid:173)
`oma cells which overexpress the EGF receptor. Pur(cid:173)
`suant to this report, we and others investigated the
`possibility of a similar effect in human cells over(cid:173)
`expressing the pl 85HER·21n•u receptor. Preliminary data
`indicated an enhanced cytotoxicity of cisplatin
`in
`breast and ovarian cells overexpressing the HER-2/neu
`gene when grown concomitantly in the presence of
`antibody specific to an extracellular epitope of the
`pl85HER·2
`MV protein (Hancock et al., 1991; Pietras et al.,
`1991). These studies, however, do not indicate whether
`the phenomenon is true synergy nor do they provide
`data regarding the possible mechanism(s) by which it
`occurs.
`We report here a proven synerg1st1c decrease in
`growth of breast and ovarian cancer cells treated with
`pl 85HER-2
`receptor antibody
`in combination with
`n•u
`cisplatin or carboplatin, drugs used
`in
`therapy of
`human neoplasms (McClay & Howell, 1990; Martin et
`al., 1992). Maintenance of the integrity of cell DNA by
`intricate repair pathways is essential to cell survival
`(Pera et al. , 1981; Kwok & Sutherland, 1989). Block(cid:173)
`ade of cisplatin-induced DNA repair by antireceptor
`antibody may underlie this effect, offering a new bio(cid:173)
`logic strategy for targeted killing of cells with HER-2/
`neu overexpression. Elucidation of a pathway for sup(cid:173)
`pression of DNA repair triggered by receptor-specific
`interactions could have broad significance in cancer
`therapy. In view of past obstacles to long-term mono(cid:173)
`clonal antibody therapies in human cancer, an alterna(cid:173)
`tive therapeutic use of antireceptor antibodies may be
`in combination with cytotoxic agents .
`
`Results
`
`Characterization of breast and ovarian cancer cells with
`H ER-2/neu gene overexpression
`To investigate the basis of the reported effects between
`cytotoxic drugs and antireceptor antibodies, we con-
`
`ducted a series of studies with human breast cancer
`cells containing marked overexpression of pl 85H ER·2
`neu
`receptor as well as with cisplatin-resistant human
`ovarian carcinoma cells which had either low or high
`expression of the receptor. SKBR3 breast adenocar(cid:173)
`cinoma cells, initially derived from a malignant pleural
`effusion, overexpress the receptor protein based on
`several-fold amplification of the gene which occurred in
`the original tumor (Kraus et al., 1987). Ovarian cancer
`cells 2008 were established from a patient with serous
`cystadenocarcinoma of the ovary, and the 2008/ Cl3*
`5.25 subline (designated Cl3 here) was obtained by in
`vitro selection for resistance to cisplatin (Andrews et
`al., 1988). Both the 2008 parental cells and the Cl3
`subline contain a single copy of the HER-2/neu gene
`and express low levels of the gene product (Pietras et
`al., 1991). To generate Cl3 cells containing high ex(cid:173)
`pression of the HER-2/neu receptor, multiple copies of
`full-length human HER-2/neu cDNA (pRVH2) were
`introduced into parental Cl 3 ovarian cells as pre(cid:173)
`viously described (Chazin et al. , 1992). Control cells
`were identically prepared utilizing a control vector
`devoid of HER-2/neu cDNA (pRVCON). Retrovirally(cid:173)
`infected clones were first selected for neomycin resis(cid:173)
`tance and then selected for pl 85HER-i n•u overexpression
`by detection of receptor protein using Western blot
`analyses as shown in Figure 1. As in 2008 ovarian
`parental Cells, minimal expression Of pl 85HER-l n•u pro(cid:173)
`tein was
`found
`in both
`the Cl3 parental and
`Cl3pRVCON 'cells. This is in contrast to marked exp(cid:173)
`ression of pl85 HER-l n•u protein found in Cl3pRVH2
`cells which were engineered to overexpress the gene
`product (Figure 1). SKBR3 cells naturally overexpress
`the receptor and were included in these studies for
`comparison
`(Figure
`l).
`Independent
`immuno(cid:173)
`histochemical analyses confirmed the relative expre(cid:173)
`ssion levels in the various cell lines and verified a
`plasma membrane distribution of the receptor (data
`not shown). Overexpression of the HER-2/neu gene
`product in Cl 3 cells was found to be associated with
`pronounced alterations
`in growth properties. The
`
`('")
`
`ex>
`0
`
`N
`
`a:
`CD
`~ 0 u
`en
`
`('")
`
`z
`0
`u
`>
`a:
`a.
`('")
`
`u
`
`N
`:I:
`>
`a:
`a.
`u
`
`('")
`
`- 200 kd
`
`- 97.4 kd
`- 69 kd
`
`- 46 kd
`
`plS5 HER-2/neu -
`
`Figure 1 Characterization of pl 85HER-i """ receptor expressed in
`human ovarian and breast carcinoma cells. Using Western blot
`analyses, SKBR3 breast cancer cells represent a positive control
`fo r p I 85H ER·2 "'" receptor expression (lane I from left) as com(cid:173)
`pared with 2008 ovarian cells (lane 2), C 13 ovarian parent cells
`(lane 3), Cl3pRVCON cells (lane 4) or Cl3pRVH2 cells engi(cid:173)
`neered lo overexpress p I 85HER-2 """ protein (lane 5). As indicated
`in the figure , p I 85HER-l neu receptor is a 185 kDa protein. Blots
`were performed as described elsewhere (Slamon er al .. 1989a.b)
`using anti-HER-2/neu receptor specific antibody
`
`2 of 10
`
`Celltrion, Inc., Exhibit 1015
`
`

`

`extent of tumor formation by HER-2/neu-overexpress(cid:173)
`ing cells in athymic mice exceeds that of corresponding
`parental control cells by 3.8-fold after 28 days (P <
`0.001; data not shown). Given these data, we had
`available cells with native and molecularly-engineered
`overexpression of H ER-2/neu gene as well as parental
`control cells containing low levels of the p 185HER·2
`receptor for further investigation.
`
`"""
`
`Effect of HER-2/ neu antireceptor antibodies in
`combination with chemotherapeutic drugs on growth of
`human cancer cells overexpressing H ER-2/ neu gene in
`vitro and in vivo
`Several murine monoclonal antibodies reactive to the
`extracellular domain of the HER-2/neu gene product
`have been previously characterized (Hudziak et al.,
`1989; Fendly et al., 1990; Sarup et al., 1991). One of
`these monoclonal antibodies, 405, has been shown to
`elicit a cytostatic inhibition of the growth of tumor cell
`lines exhibiting overexpression of the HER-2/neu gene
`product. Such growth-inhibitory activity may be attri(cid:173)
`butable, in part, to blockade of an autocrine or para(cid:173)
`crine growth-stimulatory loop involving the membrane
`ignaling through
`receptor or, alternatively, to direct
`the receptor signal transduction pathway. Using in
`vitro assays, we confirm that the 405 antibody pro(cid:173)
`the
`in
`motes a marked, dose-dependent reduction
`growth of SKBR3 cells (P<0.001; Figure 2a). A mar(cid:173)
`ginal reduction of growth of C 13 cells molecularly(cid:173)
`"'" is also observed
`engineered to overexpres pl 85HER·2
`with antibody treatment (P < 0.10, Figure 3a) while
`C 13 control cells with no overexpression of this protein
`are not affected by 405 (Figure 3a).
`The differential sensitivity of SKBR3 and the several
`ovarian carcinoma cells to cisplatin at concentrations
`ranging from 0.02 - 83 µM is shown in Figures 2b and
`3b, re pectively. As with drug-resistant Cl 3 parental
`cells, Cl 3pRVH2 cells maintain a cisplatin-resistant
`phenotype after introduction of the HER-2/neu expres(cid:173)
`sion vector (Figure 3b), demonstrating that overexpres(cid:173)
`the resistant
`sion of HER-2/neu has no effect on
`phenotype in these cells. A more typical dose-response
`the cisplatin-sensitive 2008 ovarian cells
`curve for
`(from which the C 13 lines were derived) is also shown
`in Figure 3b for comparison. When 405 antibody is
`used in combination with cisplatin in SKBR3 (Figure
`2b) or C 13pR VH2 (Figure 3c) cells, a further and
`significant suppression of cell proliferation occurs
`(P<0.001). The cisplatin concentration at which a
`50% reduction in cell proliferation occurs (IC 50) in the
`absence vs the presence of 405 antibody changed from
`sixfold in Cl 3pR VH2 cells to greater than 16-fold in
`SKBR3 cell groups. Treatment of cells with the com(cid:173)
`bination of cisplatin and isotype control antibody had
`no greater effect on cell proliferation than cisplatin
`alone (P> 0.40). These dose-effect relationships were
`evaluated further using the method of Chou & Talalay
`latter median-effects
`the
`in
`required
`(1984). As
`approach, cells are grown in the presence of increasing
`concentrations of antibody or cisplatin alone, and with
`fixed-molar ratio (cf.
`in a
`both agents maintained
`Figures 2 and 3). Analyses of these data show a syner(cid:173)
`gistic interaction between 405 antibody and cisplatin
`in vitro in SKBR3 breast carcinoma cells, with a Com(cid:173)
`bination Index 50 < 0.5. These effects were also found
`
`2 neu RECEPTOR ANTIBODY BLOCKS DNA REPAIR
`
`1831
`
`a
`
`100
`
`90
`
`e 80
`c
`
`70
`
`0
`u
`:::!!
`~ 60
`c
`0
`50
`·~
`~ 40
`0
`a.
`30
`Q)
`u
`
`20
`
`10
`
`100
`10
`Antibody concentration (µg m1 - 1 )
`
`1000
`
`b
`
`100
`
`90
`
`g 80
`
`c
`0
`70
`u
`:::!!
`~ 60
`c
`.2
`50
`~ 40
`~
`0
`30
`a.
`Q)
`u
`
`20
`
`10
`
`0
`0.1
`
`100
`10
`Cisplatin concentration (µM)
`
`1000
`
`Figure 2 Sensitivity of SKBR3 breast carcinoma cells to cis(cid:173)
`platin and HER-2 . neu antireceptor antibody and synergistic inter(cid:173)
`actions . (a) SKBR3 cell proliferation in vitro in the presence of
`4D5 antibody at concentrations ra nging from 0.2 to 800 µg m1- 1
`[A : 4D5). (b) Proliferation of SKBR3 cells was determined in the
`presence of cisplatin ranging from 0.02 to 83 µM [• : ci platin].
`SKBR3 cells were also treated with 4D5 together with cisplatin,
`with the combination given at a fixed-molar ratio of 64 : I (cis(cid:173)
`plat in :4D5). Concentrations of cisplatin ra nged from 0.02 to
`83 µM and are shown on the abscissa; corresponding conventions
`of 4D5 ranging from 0.3 nM to 1.3 µM (i .e .. 0.05 to 200 µg ml - 1)
`in order to maintain a fixed-molar ratio of 64 : I (cisplatin: 4D5)
`with cisplatin were present in this experiment but are not dis(cid:173)
`played on the absci sa [ e : cisplatin & 405]. In these studies. drug
`was added 5 min after antibody. Control experiments were con(cid:173)
`ducted with non-p185" ER-~ .... antibodies of the same cla s and
`isotype (lgG I) and /or cisplatin vehicle as appropriate
`
`when 405 was added in combination with carboplatin
`in SKBR3 cells (P < 0.001 ; data not shown). Similar
`anlayses of data for Cl 3pR VH2 cells required mathe(cid:173)
`matical extrapolation of the dose required to produce a
`median effect for treatment with antibody alone (cf.
`Figure 3a). With the proviso that the latter dose pro(cid:173)
`jection is valid, the resulting calculations predict a
`Combination Index 50 < 0.5 as found with the breast
`the relative receptor(cid:173)
`carcinoma cells . To confirm
`dependent specificity of this phenomemon, Cl3pRV(cid:173)
`CON cells which do not overexpress HER-2/neu proto(cid:173)
`oncogene were treated with identical antibody/drug
`combinations, and no apparent synergistic decrease in
`cell growth was observed (Figure 3). Similarly, the IC50
`ratio in the absence vs the presence of 405 antibody
`was 0.97 in Cl3pRVCON cells.
`
`3 of 10
`
`Celltrion, Inc., Exhibit 1015
`
`

`

`1832
`
`R.J . PIETRAS e1 al.
`
`a
`
`100
`
`95
`
`90
`
`85
`
`80
`
`e c
`
`0
`u
`~
`c
`.2
`I'!
`~ e c.
`
`Qi
`u
`
`75
`
`0.1
`10
`100
`1000 5000
`Antibody concentration (µg ml - 1
`)
`
`100 b
`
`80
`
`60
`
`40
`
`20
`
`e c
`
`0 u
`~
`c
`.2
`I'!
`~
`0 a.
`
`Qi
`u
`
`0
`O.Q1
`
`c
`
`100
`
`0.1
`10
`Cisplatin concentration (µM)
`
`100
`
`e c
`
`0 u
`~
`c
`.2
`iti
`~
`0 a.
`
`Qi
`u
`
`80
`
`60
`
`40
`
`20
`
`Cisplatin & 405
`
`0-+-~~...,..,-~~,.,.,,,~~-rm...----.......... TT"n~,
`0.01
`0.1
`10
`100
`Cisplatin concentration (µM)
`
`Figure 3 Sensitivity of ovarian carcinoma cells to cisplatin and
`HER-2/neu antireceptor antibody and synergistic interactions. (a)
`C 13 ovarian cell proliferation in vi1ro in the presence of 405
`antibody ranging from 0.4µgm1 - 1 to J.2mgm1 - 1
`; Cl3pRV(cid:173)
`[ •
`CON/405; e ; Cl3pRVH2/405]. (b) Proliferation of cisplatin(cid:173)
`resistant Cl3 cells was determined in the presence of cisplatin
`ranging from 0.04 to 83µM (• ; Cl3pRVCON/cisplatin; e ; Cl3-
`pR VH2/cisplatin]. For comparison,
`the growth
`response of
`cisplatin-sensitive 2008 cells in the presence of the same dose
`range of cisplatin are also shown (reduced points, 2008/cisplatin).
`(c) Cl3 cell groups were also treated with 405 together with
`cisplatin, with the combination given at a fixed molar ratio of
`64 : I (cisplatin :405). Concentrations of cisplatin ranged from
`0.04 to 83 µM and are shown on the abscissa. Corresponding
`concentrations of 405 ranging from 0.7 nM to 1.3 µM (i .e., 0. 1 to
`200 µg mJ - 1
`) were also present in order to maintain a fixed-molar
`ratio of 64: I (cisplatin:405), but these values are not displayed
`on the abscissa [• , Cl3pRVCON/cisplatin & 405; e, Cl3-
`pRVH2/cisplatin & 4D5]. In these experiments, drug was added
`5 min after antibody. Control experiments were conducted with
`non-HER-2/neu antibodies of the same class and isotype (lgGI)
`and/or cisplatin vehicle as appropriate
`
`To confirm and extend the in vitro studies of the
`synergistic effect between 4D5 and cisplatin, the com(cid:173)
`bination was tested for inhibition of growth of sub(cid:173)
`cutaneous tumor xenografts in athymic nude mice.
`Growth of ovarian CI3pRVH2 cells was monitored in
`animals treated with either 405 alone, isotype control
`antibody, cisplatin alone or a combination of antibody
`and cisplatin (Figure 4). Previous pharmacokinetic
`studies using antibody 405 have been presented and
`demonstrate that single-dose therapy with this mono(cid:173)
`clonal antibody leads to maintenance of a significant
`serum antibody concentration of several days duration
`(DeSantes et al., 1992). Treatment of tumor-bearing
`mice with a single dose of 4D5 caused a concentration(cid:173)
`dependent inhibitie>n of tumor growth at antibody
`doses ranging from 25-150mgkg- 1 (P<0.01; Figure
`4a). Growth inhibition by antibody alone is cytostatic
`since tumor growth resumes by 21-28 days after the
`antibody dose. Treatment of mice with cisplatin alone
`at doses of 6-9 mg kg- 1
`, but not at 3 mg kg- 1
`, elicited
`a similar dose-dependent decline
`in
`tumor growth
`(P < 0.01; Figure 4b ). Therapy of
`tumor-bearing
`animals with 4D5 in combination with cisplatin results
`in a significant and marked inhibition of tumor growth
`exceeding the effect of either agent given alone (P <
`0.005; Figure 4c). Analyses of these dose-effect rela(cid:173)
`tionships by the method of Chou & Talalay ( 1984)
`again show a substantial synergistic interaction between
`the 405 antibody and cisplatin in vivo, with a Com(cid:173)
`bination Index so= 0.16, further substantiating the
`superior therapeutic effect of combined therapy. Data
`presented in Figure 5 indicate that the benefit of this
`treatment is sustained over a 6 week period after one
`dose of combined therapy. Cisplatin and antireceptor
`antibody administered
`together elicit a
`logarithmic
`reduction in ovarian tumor size as compared to that
`observed with cisplatin or antibody given as single
`agents (P<0.001) .
`
`Effect of H ER-2/ neu antireceptor antibodies in
`combination with chemotherapeutic drugs on unscheduled
`DNA synthesis
`After demonstrating both in vitro and in vivo a clear
`synergistic effect of the combination of 4D5 and cis(cid:173)
`platin in HER-2/neu-overexpressing cells, experiments
`were designed to evaluate the possible mechanism(s)
`for this phenomenon. To determine if the synergistic
`increase in drug-mediated cytotoxicity occurring with
`antibody was a result of an increase in cellular accum(cid:173)
`ulation of cisplatin, we conducted independent experi(cid:173)
`ments using methods previously described (Andrews et
`al., 1988). These studies showed no significant effect of
`405 at doses up to 100 µg mJ- 1 on accumulation of
`83 µM [14qcarboplatin by SKBR3 cells over 3 to 24 h
`(data not shown), indicating that the synergistic effect
`does not appear to occur by altered cellular accumula(cid:173)
`tion of this chemotherapeutic drug.
`DNA repair is well known to play an important role
`in the recovery of cells from the toxicity of cisplatin
`(Pera et al., 1981; Scanlon & Kashani-Sabet, 1988;
`Whitaker, I 992; Zhen et al., I 992). To evaluate whether
`changes in DNA repair mechanisms might be a poten(cid:173)
`tial explanation for the synergistic interaction of anti(cid:173)
`receptor antibody and platinum-derived drugs, we
`measured unscheduled DNA synthesis (Trosko &
`
`4 of 10
`
`Celltrion, Inc., Exhibit 1015
`
`

`

`a
`
`1200
`
`2 neu RECEPTOR
`
`TIBODY BLOCKS 0 A REPAIR
`
`1833
`
`E 1000
`E
`u
`:0
`:J
`2
`
`800
`
`600
`
`400
`
`200
`
`0
`
`0
`b
`
`1200
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`Q)
`
`E
`:J
`0
`>
`0
`E
`:J
`I-
`
`Q)
`
`E
`:J
`0
`>
`0
`E
`:J
`I-
`
`E 1000
`E
`u
`:0
`:J
`2
`
`800
`
`600
`
`400
`
`200
`
`c
`
`1200
`E 1000
`E
`u
`:0
`:J
`2
`
`800
`
`600
`
`400
`
`200
`
`Q)
`
`E
`:J
`0
`>
`0
`E
`:J
`I-
`
`0
`
`0
`
`5
`
`15
`20
`10
`Treatment day
`
`25
`
`30
`
`figure 4 Growth of CJ3pRVH2 ovarian cancer cells in nude
`mice and effect of treatment with 405 antibody. ci olatin and
`antibody drug combination . (a) The growth of CI 3pRVH2
`tumors is shown in the presence of control ( • ) or 405 an ti recep(cid:173)
`). 25 mg kg - 1 ( A ). 75 mg kg - 1
`tor antibody at 5 mg kg - 1
`).
`( •
`( •
`or ·150 mg kg - 1 (0) given by intraperitoneal injection on day 0.
`(b) The growth ofCl3pRVH2 tumors is shown in the presence of
`control( • ) or cisplatin at 3mgkg- 1 ( e ). 6mgkg - 1 ( A ). or
`) given by i.p. injection on day 0 . (c) Finally. the
`9 mg kg - 1
`( •
`growth of CI 3pR VH2 tumors is shown on treatment with control
`(•) or 405 antibody 'c1splatin combinations at 2.5 1.5 mg kg - 1
`) respectively.
`( e ). 5.0/3.0 mg kg - 1 ( A ) and 10.0/ 6.0 mg kg - 1 ( •
`Antireceptor antibody 405 was given i.p. at day 0. and cisplatin
`was given 18 h later. Ovarian cells were injected ubcutaneously
`at 5 x 107 cells per animal. After I week. mice were randomized
`on day 0 to groups of 3 4 animals on the basis of body weight
`and tumor nodule size. Animals received either lgG I antibody
`control ( 150 mg kg - 1). 405 antibody. cisplatin or a combination
`treatment (see Materials and methods for additional details
`
`1000
`
`100
`
`E
`E
`u
`:0
`:J
`2
`Q)
`
`E
`:J
`0
`>
`0
`E
`:J
`I-
`
`0
`
`5
`
`10
`
`15
`
`25
`20
`Day
`
`30
`
`35
`
`40
`
`45
`
`figure 5 Growth of Cl3pRVH2 ovarian cancer cells in nude
`mice over 6 week and effect of a single treatment with 405
`antibody. cisplatm or a ntibody drug combination . The growth of
`Cl3pRVH2 tumor i hown in the presence of control ( e ). 405
`). cisplatin at 3 mg kg - 1
`antireceptor antibody at 5 mg kg - 1 ( •
`(0). or 405 antibody 'cisplatin combination at 5.0 3.0 mg kg - 1
`( A ). respectively. Antireceptor antibody 405 was given i.p. al
`day 0. and cisplatin wa given 18 h later as in Figure 3. Ovarian
`cells were injected subcutaneously at 5 x 107 cells per animal.
`After I week. mice were randomized on day 0 to groups of 3 4
`animals on the basis of body weight and tumor nodule size
`
`active DNA repair apparatus in the e cells (P < 0.001;
`Figure 6a). Treatment with 4D5. however, almost com(cid:173)
`pletely blocked thi cisplatin-induced increase in DNA
`this phenomenon wa
`that
`synthesis. To confirm
`pecifically due to HER-2 neu overexpression, it wa
`also tested in ci platin-resistant C 13 cells with and
`without overexpression of HER-2 neu. In these studie ,
`Cl3p RVH2 cells, but not CI3 control cells, exhibited
`repair
`suppression of DNA
`an antibody-induced
`(P < 0.0 I; Figure 6a). Carboplatin at 34 µM also pro(cid:173)
`in SKBR3 and CI3
`moted unscheduled DNA synthesi
`(P < 0.0 I). This effect was similarly blocked by
`cell
`in cells
`combined treatment with 4D5 (200 µg ml- 1
`)
`in control cell
`overexpressing HER-2/neu but not
`(data not shown). To confirm and extend these obser(cid:173)
`vations, an a lternative measure of unscheduled DNA
`synthesis was performed. Autoradiographic localization
`of silver grains due to [3H]thymidine uptake in cell
`nuclei provided independent data demonstrating that
`this phenomenon does indeed occur. Cisplatin, but not
`4D5, enhances unscheduled DNA synthesis in both
`SKB R 3 and C 13 cells (Figure 6b). Again, this drug(cid:173)
`induced effect is blocked by combined treatment with
`antireceptor antibody in cells overexpressing HER-2/
`neu but not in the C 13 control cells, confirming that
`4D5 interferes with DNA repair only in those cell
`""" receptor.
`overexpressing the p I 85HER·2
`
`Yager, 1974; Williams, 1977) induced by cisplatin in
`SKBR3 and C 13 cells (Figure 6a). As expected, treat(cid:173)
`ment of SK BR3 cells with cisplatin a lone provoked a
`significant increase in unscheduled DNA synthesis to
`2. 1-times the control level as determined by [3H]thymi(cid:173)
`dine incorporation into DNA. These data indicate an
`
`Effect of H ER-2/ neu antireceptor antibody on formation
`and repair of cisplatin-induced DNA adducts in the DNA
`of cisplatin-sensitive and -resistant human ovarian cancer
`cells
`Since measures of unscheduled DNA synthesis provide
`only an indirect as essment of actual DNA repair, we
`
`5 of 10
`
`Celltrion, Inc., Exhibit 1015
`
`

`

`1834
`
`R.J . PIETRAS et al.
`
`ought to obtain direct data on the formation and
`removal of cisplatin-induced lesions in total genomic
`DNA of ovarian carcinoma cells (Table I). Cisplatin(cid:173)
`en itive parental (2008) and cisplatin-resistant C 13
`ovarian
`carcinoma
`cells
`with
`overexpression
`
`a
`
`240
`
`220
`
`e
`c
`
`Q)
`
`160
`
`140
`
`0
`0
`<ft.
`"' 200
`·;;;
`..c. c 180
`>
`"' <t
`z
`0
`-0
`2
`:l
`-0 120
`Q)
`..c.
`0 "' 100
`c
`~
`
`b
`
`405
`
`Cisplatin
`
`Cisplatin/405
`
`:l
`
`"'
`Q) u
`:l
`c
`Qj
`0 ....
`Q) a.
`"' c
`·~
`Cl
`Q)
`z
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`Control
`
`405
`
`Cisplatin Cisplatin/405
`Treatment groups
`
`Figure 6 D A repair (unscheduled D A synthesis) in human
`breast and ovarian carcinoma cells after treatment with 4D5,
`cisplatin (DDP) or 4D5/cisplatin (4D5/DDP) combinations. (a)
`Unscheduled DNA synthesis was measured in Cl3pRVCO
`(black bar), Cl 3pR VH2 (white bar), and SKBR3 (hatched bar)
`cells in the presence of 4D5 (~OOµgml- 1 ), cisplatin (IOµM) or
`4D5.cisplatin (200 µg m1 - 1 and I 0 µM , respectively). (b) Measure(cid:173)
`ment of D A repair by autoradiographic localization of radio(cid:173)
`active thymidine was ~abulated in Cl3pRVCO
`(black bar),
`Cl3pRVH2 (grey bar), and SKBR3 (batched bar) cells. Counts
`of developed silver grains in the photographic emulsion overlying
`cell nuclei were compared after treatment with 4D5 (200 µg
`mJ - 1
`) , cisplatin (DDP: 10 µM), 4D5,carboplatin (200 µg mJ - 1 and
`10 µM , respectively) or control (CO ) solutions. Unscheduled
`D A synthesis was detennined as described in Materials and
`methods
`
`(Cl 3pR VH2) or normal expression (Cl 3pR VCON) of
`the HER-2/neu gene were each treated with 200 µM
`cisplatin for I h at 37°C, washed and then harvested at
`0 or 24 h after the initial cisplatin treatment. To test
`the effect of antireceptor antibody, cells were first
`exposed to 4D5 (200 µg ml - 1
`) or control solution for
`6 h prior
`to cisplatin
`treatment. After cisplatin
`exposure and cell washing, 4D5 was maintained in the
`culture medium at 200 µg mJ - 1 for the repair times
`indicated in Table I. Table I shows data for the forma(cid:173)
`tion and removal of cisplatin-DNA adducts from the
`genomic DNA of cells
`treated with and without
`antireceptor antibody in three separate experiments.
`The initial frequency of cisplatin lesions in the parental
`2008 cells is six to even times higher than in the C 13
`resistant cells as found by others (Zhen et al., 1992).
`This is consistent with the cisplatin-resistant phenotype
`of the C 13 cells.
`In
`the absence of antireceptor
`antibody, removal of cisplatin-induced DNA adducts
`at 24 h, a direct measure of DNA repair occurs at
`similar rates in the cisplatin-resistant cells (Cl 3pR VH2
`and CI 3pRVCON). As anticipated, the extent of repair
`in the resistant cells exceeds that seen in cisplatin(cid:173)
`sensitive cells (2008). Both the 2008 and C13pRVCON
`cells contain low expression of HER-2/neu, and com(cid:173)
`bined therapy with 4D5 antibody and cisplatin had no
`significant change in the rate of cisplatin-induced DNA
`adduct removal at 24 h. However, treatment of HER-
`2/neu-overexpressing CI 3pRVH2 cells with anti recep(cid:173)
`tor antibody prior to cisplatin promoted a significant
`reduction in the extent of DNA repair to 64% of that
`found in cells not treated with antibody (P < 0.05).
`Moreover, the actual rate of drug-DNA adduct repair
`found in the cisplatin-resistant C 13pR VH2 cells treated
`with the antireceptor antibody approached that found
`in the cisplatin-sensitive 2008 cells, indicating that the
`cisplatin-resistant phenotype can be reversed by this
`combined therapy.
`To assure that the effect noted above was not attri(cid:173)
`butable solely to an excessive cisplatin concentration,
`C 13pRVH2 cells were also exposed for I h to cisplatin
`at I 0 µM, a dose closer to the I Cj() for these cells.
`Otherwise,
`this experimental series was conducted
`using a
`treatment protocol as outlined above, but
`determination of platinum levels in cellular material
`was done by a sens