Antitumor Activity of Carcinoma-reactive BR96-Doxorubicin
`Conjugate against Human Carcinomas in Athymic Mice and
`Rats and Syngeneic Rat Carcinomas in Immunocompetent Rats
` 
`Hans Olov Sjögren, M. Isaksson, David Willner, et al.
`Cancer Res  
`
`1997;57:4530-4536.
`
`Updated version
` 
`
`Access the most recent version of this article at:
`http://cancerres.aacrjournals.org/content/57/20/4530
` 
`
`
`
` 
` 
` 
` 
` 
` 
`
`E-mail alerts
` 
`Reprints and
`Subscriptions
` 
`Permissions
` 
`
`Sign up to receive free email-alerts
`
` related to this article or journal.
`
`To order reprints of this article or to subscribe to the journal, contact the AACR Publications
`.
`pubs@aacr.org
`Department at
` 
`To request permission to re-use all or part of this article, contact the AACR Publications
`.
`permissions@aacr.org
`Department at
` 
`
`
`
`
`
`Downloaded from cancerres.aacrjournals.org Downloaded from cancerres.aacrjournals.org on November 14, 2014. © 1997 American Association for Canceron November 14, 2014. © 1997 American Association for Cancer
`
`
`
`
`
`Research. Research.
`
`IMMUNOGEN 2143, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`[CANCER RESEARCH 57. 4530-4536, October 15. 1997]
`
`Antitumor Activity of Carcinoma-reactive BR96-Doxorubicin Conjugate against
`Human Carcinomas
`in Athymic Mice and Rats and Syngeneic Rat
`Carcinomas
`in Immunocompetent Rats1
`
`Hans Olov Sjögren,2M. Isaksson, David Willner,
`
`Ingegerd Hellström, Karl Erik Hellström, and Pamela A. Trail
`
`Unit of Tumor Immunology. Department of Cell and Molecular Biology. Wallenberg Laboratory: University1 of Lund, 22007 Lumi, Sweden [H. O. S.. M. 1.1; Brisrol-Mvers Squibb
`Pharmaceutical
`Institute. Wallingford. Connecticut 06492 ¡D.W.¡;and Bristol-Myers
`Squibb Pharmaceutical
`Institute. Seattle. Washington 9X121 ¡I.H.. K. E. H.. P. A. T.I
`
`ABSTRACT
`
`to the
`antibody BR96 was conjugated
`The internalizing monoclonal
`anticancer drug doxorubicin (DOX) using an acid-labile hydrazone bond
`to DOX and a thioether bond to the monoclonal
`antibody. The resulting
`conjugate,
`termed BR96-DOX, binds to a tumor-associated
`Lewis ' anti
`gen that is abundantly expressed on the surface of human carcinoma cells.
`BR96-DOX binds
`to RCA, a human colon carcinoma
`cell
`line, and
`BN7005, a transplantable
`colon carcinoma
`induced in a Brown Norway
`(BN) rat by 1,2-dimethyl-hydrazine.
`BR96-DOX produces cures of estab
`lished s.c. RCA human colon carcinomas
`in athymic mice and rats.
`BR96-DOX also cured both s.c. and intrahepatic BN7005
`tumors
`in
`immunocompetent
`BN rats. Unconjugated DOX, given at its maximum
`tolerated dose, and matching doses of nonbinding
`IgG-DOX conjugate
`were not active against RCA or BN7005 carcinomas. An anticonjugate
`antibody
`response was produced in BN rats treated with BR96-DOX.
`However,
`this could be largely prevented by administering
`the immuno-
`suppressive
`drug deoxyspergualin.
`These results confirm the concept of
`antibody-directed
`therapy
`in models
`in which the targeted antigen is
`expressed
`both in normal
`tissues and tumors. The findings
`in BN7005
`further demonstrate
`efficacy of BR96-DOX therapy in a model
`in which
`the tumor is syngeneic and the host is immunocompetent.
`
`INTRODUCTION
`MAbs3 to tumor-associated
`
`antigens have been used with variable
`
`success in the treatment of cancer. With the exception of treatment of
`minimal residual disease colon carcinoma (1), unmodified MAbs have
`shown little antitumor activity. MAbs have also been used to deliver
`a variety of toxic moieties to malignant cells (2-4). Antigen-specific
`activity has been demonstrated
`in vitro and in nude mouse models
`with several such conjugates. However,
`to date, the clinical efficacy of
`MAb-mediated
`delivery has been demonstrated definitively only for
`radiolabeled MAb conjugates used in the therapy of B cell lymphoma
`(5, 6).
`Several factors contribute to the lack of clinical predictability of the
`models commonly used to evaluate MAb conjugates. Typically,
`treat
`ment starts shortly after xenografting of human tumors to congenitally
`athymic
`(nude) mice when the tumor burden is very small, and
`conclusions
`about efficacy are commonly based on a delay in tumor
`outgrowth rather than on regression or cure of established tumors. The
`tumors used are usually sensitive to the free drug administered by an
`optimized route and schedule, although activity of the drug may not be
`observed
`at "conjugate
`equivalent"
`doses. The majority of MAbs
`
`Received 2/5/97; accepted 8/7/97.
`The costs of publication of this article were defrayed in pan by the payment of page
`charges. This article must
`therefore be hereby marked advertisement
`in accordance with
`18 U.S.C. Section 1734 solely to indicate this fact.
`1This work was supported by the Bristol Myers Squibb Pharmaceutical Research
`Institute and the Swedish Medical Research Council.
`Immunology.
`2 To whom requests for reprints should be addressed, at Unit of Tumor
`Department of Cell and Molecular Biology. Wallenberg Laboratory. Box 7031. 22007
`Lund. Sweden.
`antibody; DOX. doxonibicin; Ley.
`used are: MAb. monoclonal
`3 The abbreviations
`tumor volume doubling delay; PR.
`Lewis y; MTD, maximum tolerated dose; TVDD.
`partial
`tumor regression; CR. complete tumor regression: DSG. 15-deoxyspergualin; BN.
`Brown Norway.
`
`identified to date and, therefore, conjugates of these MAbs are tumor
`selective rather than tumor specific in patients because binding to cells
`of some normal
`tissues is typically observed. However,
`the preclinical
`models
`typically used do not address
`the issue of tumor-selective
`targeting because rodents do not, for the most part, express the target
`antigen in normal
`tissues. Furthermore, patients in Phase I/II clinical
`trials usually have bulky tumors,
`into which immunoconjugates
`pen
`etrate poorly (7), and the tumors are frequently resistant
`to chemo-
`therapeutic
`drugs as a consequence
`of extensive prior
`therapy.
`In
`addition, whereas an antibody response to the conjugate is common in
`patients,
`it cannot occur in athymic rodents.
`is expressed
`that
`MAb BR96 recognizes
`a Ley-related epitope
`(>100,000 molecules/cell)
`at the surface of cells from the majority of
`human cancers of breast, colon, and lung (8, 9). The conjugate,
`termed
`BR96-DOX, was prepared using a chimeric (mouse-human)
`form of
`murine BR96 ( 10) and the DOX derivative 6-maleimidocaproyl DOX
`hydrazone so that it contained both an acid labile bond and a thioether
`linker (11, 12). Upon binding to the tumor cell surface, BR96-DOX
`internalizes
`into the acidic environment of lysosomes/endosomes
`and
`liberates DOX. The BR96-DOX conjugate was shown to produce
`cures of antigen-expressing
`tumors,
`to be at least 8-fold as potent as
`the unconjugated
`parent drug DOX, and to be active
`at a dose
`equivalent
`to 5% of its MTD in athymic mice (11).
`The antigen to which BR96 binds is not expressed in mice. It is,
`however, detectable
`in the gastrointestinal
`epithelium of humans,
`dogs, and rats. It was encouraging,
`therefore,
`that cures were obtained
`also in nude rats xenotransplanted with DOX-sensitive
`human lung
`tumors (11). Although athymic rats provide an appropriate model for
`conjugate efficacy and toxicity,
`they lack the ability to produce an
`anticonjugate
`response. Athymic animals may also mount a limited
`host
`response
`to xenotransplanted
`tumors, which may facilitate re
`gression after an initial damage
`inflicted by the immunoconjugate
`(13). Therefore, a model of a syngeneic tumor in immunocompetent
`animals
`that expresses
`the target antigen in normal
`tissues would
`reflect
`the clinical situation more closely.
`This study was designed to address
`several of these issues. The
`efficacy of BR96-DOX was investigated
`against
`several
`types of
`colon tumor models. These were s.c.
`implanted RCA human colon
`tumor xenografts
`studied in athymic mice and rats and a rat colon
`carcinoma, BN7005, which was evaluated as a s.c. metastasis model
`and an experimental
`liver metastasis model
`in syngeneic,
`immuno
`competent
`rats. In the rat models,
`the BR96 MAb was tumor selective
`rather than tumor specific. The effect of an anticonjugate
`response on
`the efficacy of BR96-DOX against BN7005 carcinomas was also
`evaluated.
`
`MATERIALS AND METHODS
`
`MAbs and Immunoconjugates
`
`(8, 9). Human IgG (Rockland
`MAb BR96 has been described previously
`Inc., Gilbertsville.
`PA) was used to produce nonbinding control
`immunocon
`jugates.
`Immunoconjugates
`consisting of BR96 (BR96-DOX)
`or human IgG
`(IgG-DOX)
`conjugated
`to DOX were prepared as described previously (12).
`
`4530
`
`Downloaded from
`
`on November 14, 2014. © 1997 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2143, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Assay for Rat Antibodies
`
`against BR96-DOX
`
`Rat Carcinoma Line BN7005
`
`AN'Ã(cid:143)Ã(cid:143)TUMOKACTIVITY OF BR'XvDOX IMMl'NÃ(cid:173)X'ONJl
`
`(ÃŒAII.S
`
`sera of blood samples
`antibody responses,
`rat antieonjugate
`To evaluate
`drawn from treated rats were assayed in an ELISA measuring specific binding
`of
`rat
`serum antibodies
`to chimeric BR96-DOX conjugate
`expressed
`as
`A,,n
`at a dilution of 1:100.
`
`Carcinoma Lines
`
`of the colon obtained from M. Brattain
`RCA is a human adenocarcinoma
`(Medical College of Ohio, Toledo, OH). BN7005 is a rat adenocarcinoma
`of
`the colon,
`induced by 1,2-dimethylhydrazine
`in a BN rat and cloned by
`limiting dilution in the absence of selection pressure.
`
`In Vitro Cytotoxicity Assays
`
`cytotoxicity was evaluated as described previously (14).
`Antigen-specific
`cultures of colon carcinoma
`cells were harvested using
`Briefly, monolayer
`trypsin-EDTA (Life Technologies,
`Inc., Grand Island, NY), and the cells were
`resuspended
`to I X lOVml
`in RPMI 1640 containing
`10% heat-inactivated
`FCS. The cells were added to flat-bottomed
`96-well microtiter
`plates
`(0.1
`ml/well) and incubated overnight
`at 37°Cin a humidified atmosphere of 5%
`
`of
`and serial dilutions
`from the plates,
`CO2 in air. Media were removed
`BR96-DOX,
`IgG-DOX, or DOX were added to each of the wells. Quadrupli
`cate samples were assayed. The cells were exposed to the drug or conjugates
`for 2 h at 37°Cin a humidified atmosphere
`of 5% CO2 in air. The drug or
`
`conjugate was then removed, and the cells were washed three times with RPMI
`and cultured in RPMI containing
`\()c/c heat-inactivated
`FCS. Approximately
`48 h later,
`the cells were pulsed for 2 h with 1.0 ¿iCi/wellof |'H|thymidine
`
`and trypsin
`(New England Nuclear, Boston MA). The media were removed,
`Instruments,
`was added to the wells. The cells were harvested
`(Skatron
`Sterling, VA) onto glass fiber filter mats and dried, and filter-bound |'H]thy-
`midine radioactivity was determined (ß-platescintillation counter, Pharmacia
`LKB Biotechnology,
`Piscataway, NJ). Inhibition of ['Hlthymidine
`uptake was
`
`the mean cpm for treated samples with the mean
`by comparing
`determined
`cpm of the untreated control.
`
`Experimental
`
`Animals
`
`inbred male rats, 3-5 months old, of the BN strain were
`Immunocompetent
`obtained
`from a closed colony maintained
`at
`the Wallenberg Laboratory.
`Athymic
`female mice of BALB/c background
`and athymic (Rowett)
`female
`rats were obtained from HarÃ(cid:173)anSprague Dawley. Animals
`received food and
`water ad libitum.
`
`Tumor Models
`
`at weekly or
`directions
`tumors were measured in two perpendicular
`s.c.
`biweekly intervals using calipers. Tumor size was defined as follows:
`/ X w2/2,
`where / = measurement
`of longest axis (mm) and w = measurement
`of axis
`perpendicular
`to / (mm). Data are presented as median tumor size. Antitumor
`activity is expressed in terms of TVDDs as described previously ( 15). A tumor
`growth delay equivalent
`to S3.3 TVDD was considered evidence of biological
`activity. PR reflects a decrease in tumor volume to £50% of the initial
`tumor
`volume; CR refers to a tumor that has regressed completely and is not palpable
`for a period of time equal
`to the tumor volume doubling time: and cure is
`defined as an established
`tumor
`that has regressed completely and that, after
`regression,
`is not palpable for a period of time a 10 TVDTs.
`
`Tumors were excised and minced in PBS to obtain
`s.c. BN7005 Tumors.
`a single cell suspension, which was inoculated s.c. in the thigh of the right hind
`leg.
`commonly me-
`BN7005 Tumors. Human colon carcinomas
`Intrahepatic
`tastasize
`to the liver, whereas
`those of rats do not. To evaluate whether
`a
`therapeutic
`response could be achieved against BN7(X)5 growing in the liver,
`a model was developed
`in which approximately Mr BN7005
`cells were
`implanted under
`the capsule of two liver
`lobes.
`In untreated animals,
`tumor
`nodules were visible at laparotomy 7 days after grafting and enlarged rapidly,
`spreading
`into the peritoneal
`cavity and mesenteric
`lymph nodes yielding
`ascites and killing recipients within 3-4 weeks. Tumor growth was determined
`at laparotomy by liver inspections
`and measurements
`of perpendicular
`diam
`eters of intrahepatic
`tumors, which were detected by a distinct,
`light color
`contrasting to the normal
`liver tissue.
`
`Therapy
`
`Tumors were staged to various sizes prior to therapy. Athymic mice and rats
`received all therapy i.V., whereas BN rats received DOX i.v. and BR96-DOX
`and IgG-DOX i.p. Treatments were performed on an individual body weight
`basis: doses were presented as both mg/nr of MAb and equivalent DOX (16).
`Conjugates
`typically contained
`I mg of DOX per 35 mg of MAb. To clarify
`whether
`the anticipated generation of an antieonjugate
`response toward BR96-
`DOX might
`reduce the efficacy of BR96-DOX therapy,
`experiments were
`performed in which BN rats were treated concurrently with DSG, a drug shown
`previously to suppress
`the appearance
`of an antibody response to analogous
`conjugates
`in mice (17. 18). DSG was administered
`i.p. in 11 daily doses of 30
`mg/nr
`beginning one day after the first day of BR96-DOX therapy.
`In one
`experiment,
`a second round of BR96-DOX therapy, with or without DSG, was
`given when tumors had regrown following BR96-DOX-induced
`PR.
`
`RESULTS
`
`in
`of BR96-DOX Conjugates
`Cytotoxicity
`Antigen-specific
`Vitro. The in vitro potency and specificity of BR96-DOX was eval
`uated against RCA and BN7005 carcinoma cells following a 2-h drug
`or conjugate
`exposure. As shown in Table
`I, BR96-DOX demon
`strated
`antigen-specific
`cytotoxicity
`against
`both the RCA and
`BN7005 colon carcinoma lines. The BR96-DOX conjugate was more
`potent
`than a nonbinding
`IgG-DOX conjugate
`prepared using the
`same linker chemistry. The BR96-DOX was approximately
`10-fold
`less potent than unconjugated DOX following a 2-h exposure in vitro.
`Activity of BR96-DOX against s.c. RCA Human Colon Tumors
`in Athymic Mice and Rats. Previous studies have reported that RCA
`colon tumors in athymic mice were cured by BR96-DOX,
`although
`the xenografts were insensitive to unconjugated DOX (11). Antitumor
`activity of BR96-DOX was shown to be antigen specific because a
`nonbinding IgG-DOX conjugate was not active. In the study presented
`here,
`the activity of BR96-DOX against established s.c. RCA colon
`tumors was evaluated in both athymic mice and rats. In each case,
`unconjugated DOX was administered at its MTD; 24 and 12 mg/m2 in
`mice and rats,
`respectively. BR96-DOX was administered
`to both
`mice and rats at a dose of 420 mg/m2 BR96-DOX ( 12 mg/m2 equiv
`alent DOX).
`In athymic mice (Fig. 1), the BR96-DOX conjugate
`produced cures of established RCA xenografts, whereas unconjugated
`
`Table 1 Antigen-specific
`
`activity of BK96-DOX against RCA human and BN7005 rat
`colon carcinoma Ã(cid:141)T//.Yin vitro
`
`Human Carcinoma Line RCA
`
`
`
`IC50 </IMDOX)RCABN7005BR96-DOX2.07.0IgG-DOX20
`
`Studies in athymic mice used s.c. tumors maintained by in viva passage as
`described previously (15). For athymic rat studies, RCA cells were harvested
`in logarithmic growth using trypsin-EDTA and washed in serum-free medium,
`and 5 x IO6 cells were injected s.c.
`in the left axillary region.
`
`4531
`
`" IC50 of
`
`>30DOX0.2
`lgG-DOX:IC5(1 of BR96-DOX.
`
`0.7Specificity
`
`>4.3
`
`ratio"10
`
`Downloaded from
`
`on November 14, 2014. © 1997 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2143, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`ANTITUMOR ACTIVITY OF BR96-DOX IMMUNOCONJUGATES
`
`to
`at a dose equivalent
`administered
`the BR96-DOX conjugate,
`15,
`the MTD of unconjugated DOX, demonstrated
`a significant
`tumor
`growth delay, with 50% CRs and 50% PRs. In contrast, unconjugated
`DOX administered at the same dose was not active. Therefore, BR96-
`DOX was active and tolerated in a model of a human DOX insensitive
`
`5000
`
`1 1
`
`Days Post Implant
`
`0
`
`10
`
`20
`
`30
`
`40
`
`Days Post Treatment
`
`antitumor activity of BR96-DOX against s.c. BN7005 colon
`Fig. 2. Antigen-specific
`tumors in BN rats. Tumors were implanted s.c., and therapy was initiated when tumors
`had reached approximately
`12 mm in diameter. Results are from untreated control rats (•)
`and rats treated with 420 mg/m2 (12 mg/m2 equivalent DOX) BR96-DOX (•),IgG-DOX
`(D). or 12 mg/m2 of DOX (A).
`
`Table 2 Antigen-specific
`
`antitumor
`syngeneic
`
`activity of BR96-DOX against
`rat colon carcinoma
`
`s.c. BN7005
`
`Dose/injection
`(mg/m2)"TreatmentBR96-DOXIgG-DOXDOXMAb90175350420420DOX2.55.010.012.012.012.018.0*24.0*Tumor
`regressionsComplete01330000Partial00060000Total
`responding/
`no. ofRats0/51/53/59/130/10OK0/200/8
`
`administered every 4 days for 3 injections.
`" Treatments
`Toxicity observed as weight
`loss >15% and hind leg paralysis.
`
`Table 3 Response of established intrahepatic
`to BR96-DOX administered
`
`isografts of the BN7005 colon carcinoma
`i.p. on four occasions
`
`isogr.Treatment"UntreatedBR96-DOX
`
`Wk after tumor
`
`36/6
`
`with tumor
`
`Tumor diameter(mm)Proportion
`
`420 mg/m2
`(12 mg/m2 DOX)Proportion
`
`6/6*
`
`
`
`
`241.4e'6/623 ±
`
`6/654
`
`with tumor
`1/6
`ND^ Métastases'
`£34ifting60/6
`Tumor diameter
`(mm)1
`a First day of therapy was considered day 1; treatments were administered on days 1,
`4, 7, and 11.
`Sacrificed on day 21 because all rats in this group had large tumors in the liver, which
`spread to the peritoneal wall and mesenteric lymph nodes.
`c For each rat, the mean diameter of each tumor of two liver tumors was calculated and
`averaged. Mean ±SE for each group is shown.
`d ND. not done.
`' One rat died with multiple lung métastasesbut no liver tumors. The remaining five
`
`rats were sacrificed 5 weeks after the start of therapy and determined to be tumor free on
`necropsy.
`
`1000 -i
`
`500-
`
`250-
`
`B
`
`3500-,
`
`3000-
`
`2500
`
`</5
`
`2000-
`
`1500-
`
`1000-
`
`500-
`
`0
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`70
`
`75
`
`80
`
`Days Post Implant
`
`s.c. RCA human colon tumor
`activity of BR96-DOX against
`Fig. 1. Antitumor
`xenografts. A, athymic mice. Results are from untreated control mice (•)and mice treated
`with 420 mg/m2 (12 mg/m2 equivalent DOX) BR96-DOX (•)or 24 mg/m2 of DOX (A)
`on days 15, 19. and 23. B, athymic rats. Results are from untreated control
`rats (•)and
`rats treated with 420 mg/m2 (mg/m2 equivalent DOX) BR96-DOX (•)or 12 mg/m2 of
`DOX (A) on days 12. 16, and 20.
`
`In
`DOX administered at twice that dose was not active (<3.3 TVDD).
`the RCA colon tumor model
`in athymic mice, BR96-DOX was more
`active and more potent
`than unconjugated DOX. The athymic mouse
`model, although an appropriate model
`for demonstrating
`distal site
`antigen-specific
`activity, does not address the issue of normal
`tissue
`expression of the targeted antigen. Previous studies demonstrated that
`BR96-DOX produced cures of established DOX-sensitive human lung
`tumor xenografts
`in athymic
`rats (11). Rats,
`like humans
`and in
`contrast
`to mice, have been shown by immunohistology
`to bind BR96;
`the normal
`tissue reactivity occurs primarily on cells of the gastroin
`testinal
`tract; esophagus,
`stomach,
`intestine,
`and pancreas
`(acinar
`cells; Refs. 8 and 11). The athymic rat, therefore, provides a unique
`model to evaluate the antitumor activity of BR96-DOX against DOX-
`insensitive human colon tumors
`in a system in which the targeting
`MAb is tumor selective rather than tumor specific. As shown in Fig.
`4532
`
`Downloaded from
`
`on November 14, 2014. © 1997 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2143, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`ANTITUMOR ACTIVITY OF BR96-DOX IMMUNOCONJUGATES
`
`B
`
`appearance of rat BN7005
`Fig. 3. Macroscopic
`colon carcinoma
`growing in two liver
`lobes. A,
`livers prior to therapy: B, livers 21 days after treat
`ment with BR96-DOX; C and D,
`liver
`from un
`treated control
`rat on days 7 and day 21, respec
`tively. T. tumor; S. scar tissue.
`
`although
`
`the targeted antigen was expressed
`
`colon carcinoma
`normal
`tissues.
`Antitumor Activity of BR96-DOX against s.c. Rat Colon Car
`cinoma BN7005. Although athymic rats provide an appropriatemodel
`for conjugate efficacy and toxicity, they may also mount a limited host
`response to xenotransplanted tumors, which may facilitate regression
`after an initial damage inflicted by the immunoconjugate (13). In addi
`tion, athymic rats lack the ability to produce an anticonjugate response
`and cannot be used to address issues of altered conjugate clearance,
`pharmacokinetics, and tumor localization when anticonjugate antibodies
`are present. Therefore, a model of a syngeneic tumor in immunocompe-
`
`in
`
`tent animals expressing the target antigen in normal tissues would reflect
`the clinical situation more closely. The BN7005 rat colon carcinoma,
`derived originally from BN rats, was used to address these issues.
`Treatment of established BN7005 s.c. tumors with BR96-DOX at a dose
`of 420 mg/m2 (12 mg/m2 equivalent DOX) administered every 4 days for
`a total of three injections resulted in antitumor activity equivalent to a6.7
`TVDD; 3 of 5 rats were cured of their tumors (Fig. 2). As shown,
`BN7005 tumors were not sensitive to unconjugated DOX because treat
`ment with the MTD of 12 mg/m2 did not result in tumor regressions or
`even a tumor growth delay. The activity of BR96-DOX against estab
`lished BN7005 tumors was antigen specific; a matching dose of 420
`4533
`
`Downloaded from
`
`on November 14, 2014. © 1997 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2143, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`ANTITUMOR ACTIVITY OF BR96-DOX IMMUNOCONJUGATES
`
`1 i V
`
`i
`
`Fig. 4. Survival of rats bearing intrahepatic BN7005 colon car
`cinomas. Results are from untreated control rats (A) and rats treated
`with 420 mg/m2 ( 12 mg/irr equivalent DOX) BR96-DOX (•)or 12
`mg/m2 of DOX (A). Arrows
`indicate when treatment doses were
`given.
`
`0 »..20
`
`Ut
`
`40
`
`60
`
`80
`
`100
`
`120
`
`140
`
`160
`
`180
`
`200
`
`Days Post Implant
`
`mg/irr of nonbinding IgG-DOX conjugate (12 mg/rrr equivalent DOX)
`was not active.
`of BN7005 to BR96-DOX was dose-
`response
`The antitumor
`dependent
`(Table 2). At doses of BR96-DOX of 90 and 175 mg/m2,
`
`administered every 4 days for a total of three injections, 0 of 5 and 1
`of 5 tumors
`regressions were observed,
`respectively. At 350-420
`mg/m2, administered
`every 4 days for a total of three injections,
`60-70% tumor
`regressions were observed (3 of 5 and 9 of 13). A
`therapeutic
`effect was not seen with control
`IgG-DOX conjugates
`administered at the same dose and schedule. BN7005 tumors were not
`sensitive to unconjugated DOX, even when it was administered
`at
`doses above the MTD. Previous experiments
`(11) have demonstrated
`that unconjugated BR96 MAb was not active against established
`tumors and that mixtures of MAb BR96 and DOX (at the MTD) were
`no more active than DOX administered alone.
`Regressions
`and Cures of Intrahepatic
`BN7005 Colon Carcino
`mas
`following Treatment
`with BR96-DOX. BN7005 cells were
`injected
`into two liver
`lobes, and treatment was initiated
`7 days
`later (four
`treatments
`given 3 days apart). Although untreated rats
`were sacrificed
`due to tumor within 3 weeks,
`six rats
`that had
`received
`420 mg/m2 BR96-DOX (12 mg/m2 DOX) were free of
`
`at 9
`(Table 3). Rats were necropsied
`tumor at 7 weeks
`detectable
`this
`time,
`five of
`the six rats were
`weeks postimplantation.
`At
`tumor
`free, and one of the rats had multiple
`lung tumors but no
`tumor elsewhere
`(Fig. 3).
`In a repeat experiment,
`the long-term survival of rats bearing intrahe
`patic BN7005 tumors was evaluated (Fig. 4). Both untreated rats and rats
`treated with the MTD (12 mg/m2) of DOX were sacrificed at 21 days
`
`postimplantation due to extensive tumor burden. At necropsy, there was
`widespread tumor growth in liver and mesenteric lymph nodes and
`formation of hemorrhagic ascites.
`In contrast, at this time all six rats
`treated with 420 mg/m2 BR96-DOX ( 12 mg/m2 equivalent DOX) did not
`
`demonstrate any clinical signs of tumor burden, and when they were
`laparotomized at 3 weeks, three were completely free of detectable tumor
`and the other three had small tumors and only at the site of inoculation.
`The median survival time for BR96-DOX treated rats was 88 days; one
`of six rats was sacrificed at day 68 due to the presence of large liver
`tumors, and two of six rats died at day 88 of multiple lung métastases.
`Fifty % of the rats treated with BR96-DOX survived for the 26-week
`evaluation period, and necropsies performed at this time showed that
`these rats were free of detectable tumor.
`Serum Antibodies
`to BR96-DOX Interfere with Its Therapeutic
`Efficacy; Treatment
`with DSG Counteracts
`Their
`Formation.
`Because both the murine and human regions of the chimeric BR96 are
`foreign in rats, an antibody response
`against BR96-DOX was ex
`
`4534
`
`pected. Rats bearing s.c. BN7005 tumors were left as untreated
`controls,
`treated with 420 mg/m2 IgG-DOX (12 mg/m2 equivalent
`DOX) plus DSG, or treated with 420 mg/m2 BR96-DOX (12 mg/m2
`
`round of
`equivalent DOX) with or without DSG during the first
`conjugate
`therapy. All drug treatments were initiated when tumors
`were approximately
`500 mar
`in size, and treatments were adminis
`tered every 4 days for a total of three injections. As shown in the Fig.
`5, DSG did not interfere with the antitumor activity of BR96-DOX,
`which in fact was slightly better for rats that were treated with both
`BR96-DOX and DSG relative to that seen with BR96-DOX alone. As
`described above (Fig. 2 and Table 2), IgG-DOX was not active against
`established BN7005 tumors. Similarly, combined therapy with DSG
`and IgG-DOX treatment was not active against BN7005 tumors (Fig.
`5). Of the BN7005-bearing
`rats treated with BR96-DOX alone, 4 of 5
`
`Flrit
`Course
`
`Second
`Course
`
`3000-
`
`1000-
`
`Days Pose Treatment
`
`Fig. 5. Antitumor activity of BR96-DOX against s.c. BN7005 colon tumor xenografts
`with and without DSG cotreatment. Tumors were implanted s.c.. and therapy was initiated
`when tumors had reached approximately
`10 mm in diameter. Results are from untreated
`control
`rats (•)and rats treated with 12 mg/m' BR96-DOX (•).12 mg/m2 BR96-DOX
`plus 30 mg/m2 DSG (O), or 12 mg/m2 IgG-DOX plus 30 mg/m2 DSG (A). A second
`round of therapy was initiated when tumors had regrown to a size of approximately
`700
`mm1.
`
`Downloaded from
`
`on November 14, 2014. © 1997 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2143, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`ANTITUMOR ACTIVITY OF BR96-DOX IMMUNOCONJUGATES
`
`against BR96-DOX after a first course of
`Table 4 Absence of serum antibodies
`conjugate therapy correlates mlh a greater therapeutic
`effect of a second course of
`BK96-DOX conjugate
`
`Serum anti-BR96-DOX"
`
`Tumor size (mm*)
`
`Rat no.
`
`BR96-DOXalone12345BR96-DOX
`
`Day -4
`
`Day 3
`
`Day 0
`
`Day 10
`
`Day 21
`
`the Ley antigen in some normal
`syngeneic, BN rats that express
`tissues. This model was used to address
`two issues that were not
`approachable by investigating xenotransplanted
`human tumors: activ
`ity in a syngeneic
`tumor model and activity in the presence of an
`anticonjugate
`immune response. Studies on BN7005 were done with
`both s.c.
`tumors and experimentally
`induced liver métastases.At a
`(0.33)Pos
`dose equivalent
`to that used with the human colon tumor RCA in
`athymic mice and rats, 420 mg/m2, 12 mg/m2 equivalent DOX,
`(0.39)Pos
`(0.69)Pos
`BR96-DOX conjugate produced 23% CRs and 46% PRs of estab
`(0.39)Neg
`(0.20)Neg
`lished s.c.
`tumors, whereas optimized, unconjugated DOX was not
`(<0.20)5704504908504503801640I860205018606006001900281020203180700320
`active.
`Importantly,
`in the BN7005 model of experimental
`liver me
`tastasis, BR96-DOX administered
`at this dose produced 50% cures
`and 30% long-term responses, whereas unconjugated, optimized DOX
`was not active.
`immuno-
`the BR96-DOX (chimeric mouse/human)
`As expected,
`conjugate was immunogenic
`in rats. The development
`of anticonju
`gate antibodies
`resulted in reduced efficacy of BR96-DOX,
`but
`in
`duction of anticonjugate Abs was decreased
`by cotreatment with
`DSG. These data are in agreement with observations made with
`immunoconjugates
`containing Pseudomonas
`exotoxin (18). The im
`munocompetent
`rat model may overpredict
`the problem with anticon
`jugate antibodies, because a mouse-human chimeric antibody may be
`more immunogenic
`in rats than in cancer patients.
`In fact,
`the pro
`duction of clinically relevant anticonjugate
`antibodies has not been
`observed to date in the Phase I clinical
`trial of BR96-DOX (19).
`These data demonstrate that MAb-directed targeting is also feasible
`when the tumor,
`like human cancers, arises in the species in which it
`is studied and is histocompatible with its host. CRs and cures of
`established
`s.c. RCA human and BN7005 rat colon tumors were
`obtained at
`tolerated doses of BR96-DOX although the rats,
`like
`humans, express the BR96-defined
`antigen in normal
`tissues of the
`gastrointestinal
`tract (8, 11). This indicates that the presence of normal
`tissue expression of the Ley antigen does not prevent
`tumor localiza
`tion and/or
`result
`in unacceptable
`levels of toxicity of BR96-DOX,
`although no detailed investigation has been performed in cured rats on
`the possible subclinical damage of organs expressing Ley.
`It is particularly encouraging that BR96-DOX could eradicate tu
`mor nodules in the liver of immunocompetent
`rats, because this is a
`primary organ for naturally occurring métastasesof colon carcinoma
`in humans.
`antigenic
`vasculature,
`functional
`level of
`size,
`tumor
`Issues of
`and inherent
`sensitivity or resistance
`to the targeted
`heterogeneity,
`drug in the context of conjugate dose and schedule are discussed in a
`separate paper.
`
`(0.29)Pos
`(0.33)Pos
`(0.77)Pos
`(0.63)Neg
`(0.24)Neg
`(<0.20)Pos
`
`+DSG(median
`
`tumor)Pos
`" Sera of blood samples drawn 4 days prior to and 3 days after start of the second round
`of BR96-DOX trealment were assayed in an ELISA measuring specific binding of rat
`serum antibodies
`to chimeric BR96-DOX. expressed as A450/6W nm at a dilution of 1:100
`(numbers
`in parentheses).
`Pos, positive; Neg. negative.
`
`response, whereas none of the
`anticonjugate
`developed a detectable
`rats treated with BR96-DOX and DSG did (Table 4). A second round
`of BR96-DOX therapy (350 mg/m2, 10 mg/m2 equivalent DOX every
`
`4 days for a total of three injections) was initiated when tumors had
`regrown to a size of approximately 700 mm3 in size. As shown in Fig.
`
`antibod
`5, tumors of rats that did not get DSG and had anticonjugate
`ies did not respond to a second round of therapy with BR96-DOX.
`In
`contrast,
`rats that received BR96-DOX plus DSG and did not have
`anticonjugate
`antibodies
`responded to a second round of BR96-DOX
`therapy with 3 of 6 PRs.
`
`DISCUSSION
`
`the therapeutic efficacy of a MAb-drug immunocon-
`In this study,
`jugate, BR96-DOX, was evaluated in several animal models, which
`were selected to address
`some of the issues
`that are faced when
`evaluating immunoconjugates
`in the clinic. These include expression
`of the targeted antigen in normal
`tissues,
`the generation of an antibody
`response to the conjugate that may adversely affect
`its pharmacoki-
`netics, and insensitivity to the targeted drug. The activity and toxicity
`of BR96-DOX in several models was compared to that of the uncon-
`jugated parent drug DOX, using doses and schedules
`that had been
`optimized for unconjugated DOX. In contrast,
`the doses and schedules
`selected were not optimal
`for BR96-DOX in these models. Rather,
`antitumor activity can be achieved at significantly lower cumulative
`doses of BR96-DOX than shown here,
`if the administration
`schedule
`is optimized for conjugate versus free drug. The doses of BR96-DOX
`were kept constant on a mg/m2 basis (16) in the various models.
`
`in this paper were performed with colon
`All studies described
`carcinomas, a tumor type know to be relatively insensitive to DOX. A
`human colon carcinoma, RCA, evaluated in athymic mice was insen
`sitive to unconjugated DOX administered
`at
`its MTD using the
`optimal schedule for DOX. In contrast, BR96-DOX at a dose of 420
`mg/m2 (12 mg/m2 equivalent DOX) produced 89% cures and 12%
`
`by
`then increased
`the RCA model was
`of
`CRs. The stringency
`evaluating the antitumor effects of BR96-DOX against RCA colon
`tumors xenografted in athymic rats, which,
`like humans but in contrast
`to mice, express the Ley antigen in normal
`tissues, par