Immunology and Cell Biology (1993) 7 1, 167-179
`
`Immunochetnotherapy of human colon carcinoma
`xenografts in nude mice using combinations of
`idarubicin-monoclonal antibody conjugates
`
`MARK J. SMYTH,' HAMISH McA. FOSTER,^ SARAH. M. ANDREW,'
`JIN. GHEE T E H/ KENIA KRAUER,' IAN F.C. McKENZIE' and
`GEOFFREY A. PIETERSZ'
`^Research Centre for Cancer and Transplantation, Department of Pathology, The University of
`Melbourne, ParkvUle, Victoria, Australia; ^Ludwig Institute For Cancer Research, Royal Melbourne
`Hospital Complex, Parkville, Victoria, Australia
`
`Summary Tumour cell heterogeneity is probably a principal cause of treatment failure and
`represents a formidable barrier for effective antibody-targeted chemotherapy. Idarubiciti (Ida), a more
`potent and less cardiotoxic analogue of daunomycin, has been demonstrated to specifically target and
`eradicate homogeneous, cloned, murine tumour cell populations in vitro and in vivo when coupled to
`monoclonal antibodies (MoAb); however, the antitumour activity of Ida-MoAb conjugates against
`human tumour xenografts remains to be established. In this study, the value of cotargeting conjugates
`to different human tumour-associated antigens within a solid tumour has been assessed by comparing
`the effects of combinations of Ida-anti-colon carcinoma MoAb conjugates with any one Ida-anti-
`colon carcinoma MoAb conjugate used alone. Individual Ida-MoAb conjugates have previously been
`evaluated for tbeir specific binding and cytotoxicity to one of two different buman colon carcinoma
`xenografts (Colo 205 or LIM2210) in vitro, altbough tbeir efficacy alone or in combination required
`assessment in vivo. Combinations of the most effective Ida-MoAb conjugates were demonstrated to
`enable a greater number of complete tumour regressions than tbe most efficacious Ida-MoAb
`conjugate administered alone in vivo; some combinations inhibited control tumour growth by up to
`95%. This study suggests tbat Ida-MoAb conjugates can be effective against subcutaneous buman
`tumours in nude mice, altbough it is unlikely that any single conjugate will eradicate all the tumour
`cells in a solid tumour, and the value of 'cocktails' of drug-MoAb conjugates against some xenografts
`(i.e. LIM2210) appears to be limited.
`
`Key words: idarubicin, immunochemotherapy, immunoconjugates, immunotoxins, monoclonal anti-
`body, xenografts.
`
`I
`
`Introduction
`
`Monoclonal antibody {MoAb)-recognizing
`antigens expressed selectively on buman colon
`cancers are currently being evaluated for their
`potential to target diagnostic and therapeutic
`agents.' Many tumours, however, appear to be
`pnenotypically unstable and consist of subpopu-
`lations of cells witb significant beterogeneity
`in their antigen expression.'^ However, sue-
`
`cessful therapeutic antibody-mediated targeting
`of drugs to combat colorectal carcinoma de-
`pends on tbe stability and beterogeneity of an-
`tigen expression witbin tbe tumour. As beter-
`ogeneity and quantitative differences in antigen
`expression bave been observed in primary
`tumours,^ cloned tumour celts'* and metastases,
`it is important to establish tbe antigenic reper-
`toirc of tumour cells wbeti designing effective
`antibody-targeting regimens.
`
`Correspondence; Dr Geoffrey Pietersz, The Austin Research Institute, Kronheimer Building, Austin
`Hospital. Heidelberg, Vic. 3084, Australia.
`,
`Accepted for publication 13 November 1992.
`
`IMMUNOGEN 2130, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`168
`
`M.J. Smyth etal.
`
`For small experimental tumours growing
`subcutaneously (s.c), dnig-MoAb complexes
`are particularly effective but large tumours can
`not be eradicated.** Some of the major prob-
`lems lie in the inaccessibility of some areas of
`large solid tumours and the level of antigen
`expression of the tumour cells. Several studies
`using MoAb-targeted drugs have arrested the
`growth of human tumours in athymic nude
`mice,^ however the constant trend of all tu-
`mours to produce phenotypic variants** has
`probably prevented their complete eradica-
`tion. Immunological therapy of human breast
`tumours implanted in nude mice has previ-
`ously demonstrated that 'cocktails' of MoAb
`were more effective in inhibiting the growth
`of tumours than any one MoAb alone.^ It was
`therefore of interest to use a panel of several
`Ida-MoAb conjugates^ targeted at a number
`of tumour-associated antigens on different hu-
`man colon carcinoma cells to examine the
`efficacy of immunoconjugate therapy. In pre-
`vious studies, Ida-MoAb conjugates were
`demonstrated to be effective in eradicating
`small murine tumours growing subcutane-
`ously in inbred mice but not large tumours.
`We now report on the effects of several
`Ida-MoAb conjugates on human colon cancer
`xenografts.
`
`Materials and methods
`
`Tumour cells
`
`The human colon carcinoma cell line Colo
`205 was maintained in vitro in RPMI, supple-
`mented with 10% heat-inactivated newborn
`calf serum (Flow Laboratories, Sydney, Aus-
`tralia), 2 mmol/L glutamine (Commonwealth
`Serum Laboratories [CSL], Sydney, Australia),
`100)ag/mL streptomycin (Glaxo, Melbourne,
`Australia) and 100 IU/mL penicillin (CSL).^"
`The murine
`thymoma
`ITT(l) 75NS E3
`(E3)" was maintained in Dulbecco's modified
`Eagles medium (DME) with the same addi-
`tives. For in vivo experiments, Colo 205 and
`LIM2210 (a fresh colon carcinoma recently
`excised from a patient at the Royal Melbourne
`Hospital, Australia, and established as a xe-
`nograft model in nude mice) were maintained
`by passage of s.c. tumours in Swiss nude mice.
`Mice were purchased from the Animal Re-
`
`sources Centre (Perth, Australia) and both
`female and male nude mice were grafted with
`human colon tumours at 7-8 weeks of age
`(bodyweight 16-25 g). Colo 205
`tumours
`cells were injected s.c. and LIM2210 tumour
`were implanted s.c. (approximately 30 mm^
`pieces) into the abdominal wall and were
`allowed to develop into palpable tumours
`before commencing
`treatment. Mice were
`then subjected to a series of intravenous (i.v.)
`treatments, and the size of tumours was meas-
`ured regularly with a caliper square measuring
`along the perpendicular axes of the tumours.
`The data were either recorded as mean tumour
`size (cm"^, two diameters) or mean tumour
`volume (cm^, three diameters) ± s.e.m. Indi-
`vidual mice were monitored for their tumour
`growth and response to treatment. Groups of
`8-10 mice, all of the same sex and age, were
`used in each experiment.
`
`Monoclonal antibodies
`
`The MoAb used were: (a) 250-30.6 (mouse
`IgG2b), which recognizes an antigen present
`on normal and malignant human gastrointesti-
`nal epithelium;'^ (b) 24-17.1 (mouse IgG2a)
`reactive with an antigen present on human
`colon and breast
`tumours;'"^ (c) JGT-13
`(mouse IgGl) reactive with carcinoembryonic
`antigen (CEA) on colon carcinoma but not
`with normal tissues; (d) 27.1 (mouse IgGl)
`reactive with human milk fat globule antigen
`(HMFG) on a number of colon tumours;
`(e) I-l (mouse IgGl) reactive with CEA on
`many colon carcinomas but negative on nor-
`mal tissues; and (f) anti-Ly-2.1 (IgG2a) used
`as a non-specific control antibody.
`JGT-13,
`27.1 and I-l were made by J.G. Teh and C.H.
`Thompson (Research Centre for Cancer and
`Transplantation, Melbourne, Australia) and
`satisfy classification as anti-colon carcinoma
`MoAb. The MoAb were isolated from ascites
`fluid by precipitation with 40% ammonium
`sulfate, dissolution and dialysis in phosphate-
`buffered saline (PBS) and purified on an Affi-
`gel Blue column
`(Bio-Rad Laboratories,
`Sydney, Australia) and eluted with PBS.
`MoAb were then dialysed against PBS, ali-
`quoted and stored at -70 "C and tested for
`activity by rosetting with sheep anti-mouse
`immunoglobulin.'•* It should be noted that
`
`IMMUNOGEN 2130, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Tumour therapy with IDA-MoAb conjugates
`
`169
`
`none of these MoAb were cytotoxic in vitro or
`in vim.
`
`Preparation oflda-MoAh
`
`conjugates
`
`in 0.05 mol/L
`Intact MoAb (2-3 mg/niL)
`Borate buffer
`(pH 7.8-8.0) were mixed
`with a 10-20 mol/L excess of 14-bromo-4-
`deniethoxydaunomycin (Br-Ida) dissolved in
`(N.N)-dimethylformainide (DMF) at 10 mg/
`mL. The reaction was maintained at room
`temperature for 4 h before centrifuging {400 ^
`for 5 min) to remove any precipitate. Free
`Br-lda and other unreacted starting materials
`were removed by gel filtration chromatogra-
`phy using a Sephadex G-25 column (PD-10,
`Pharmacia, Uppsala, Sweden), and the conju-
`gates were then passed through a column of
`Porapak Q to remove any adsorbed drug.'^
`The amount of Ida incorporated in the conju-
`gate (3-6 molecules of Ida/molecule MoAb)
`was determined by absorbance spectropho-
`tometry at 483 nm (64^3 = 3.4 x lO^mol/L
`per cm) and protein estimation;^*" further de-
`tails are provided elsewhere.''
`
`Drug activity
`
`In a 24 h cytotoxicity assay, 100 |iL of Colo
`205 tumour cells (2-5 x 10''/mL) were added
`to a 96-well flat-bottomed microtitre plate
`and incubated for 1 h at 37 "C."' Free Ida
`(prepared by dissolution in PBS) and Ida-
`MoAb conjugates were filtered sterile through
`a 0.22 (im Millipore filter and diluted in PBS;
`50 ^iL of Ida or Ida-MoAh were added to the
`cells in duplicate; control wells received 50 |iL
`of PBS and the cells were cultured at 37 °C in
`a 7% CO2 atmosphere for 24 h. In a 30 min
`inhibition assay. 200 |iL of tumour cells (2-
`5 X 10''/mL) were collected in sterile Eppen-
`dorf tubes, resuspended in sterile conjugate
`and mixed for 30 min at 37 °C. Cells were
`then centrifuged
`(4001^ for 5 min). resus-
`pended in growth medium, and 100 |JL ali-
`quots were seeded into a microtitre plate using
`quadruplicate wells/sample before an incuba-
`tion period of 16-24 h. After the incubation
`period in both assays, 50 (iL of medium con-
`taining 1 |iCi of ['^H]-thymidine (specific ac-
`tivity 5 Ci/mmol; Amersham International
`Ltd. Amersham, England) was added, and the
`plates were incubated for 2-4 h; cells were
`
`then harvested onto a glass fibre filter paper
`using a cell harvester, dried for 10 min at
`SO'C and the individual samples were separ-
`ated and counted on a B-scintillation counter.
`The incorporation of [" H]-thymidine was ex-
`pressed as a percentage of the inhibition of
`incorporation of controls; the standard error
`was generated by multiple determinations and
`did not exceed 5% for any given experimental
`point.
`
`Flow cytometry
`
`The reactivity of different MoAb against hu-
`man colon carcinoma was assessed by flow
`cytometry using
`tluorescein
`isothiocyanate
`(FITC)-labelled MoAb and freshly prepared
`Colo 205 or LIM2210 tumour cells. Briefly,
`80 (iL of FITC
`solution
`(1 mg/mL
`in
`acetone)/mg of MoAb were coated as a fine
`layer in a glass bijou bottle. The MoAb
`(0.5 mol/L bicarbonate buffer, pH 9.0) was
`added and incubated in the dark at room
`temperature for 1 h. Free FITC was removed
`by gel filtration
`(PD-10, Pharmacia) and
`FITC-MoAb conjugates were quantitated
`spectrophotometrically.'^ LIM2210 and Colo
`205 tumour cells were prepared from an
`excised s.c. tumour by gently breaking the
`tumour into small pieces, aspirating with a
`fine needle to a single suspension and washing
`three times in PBS with 1% bovine serum
`albumin (BSA). These cells (5 x 10^) were
`added
`to Eppendorf
`tubes and
`incubated
`with 100 ML (10 ng MoAb) FITC-MoAb
`(3-4 mol/L FlTC/mol MoAb) at 4°C for
`1 h. Unbound antibody was removed by
`three washes with PBS (9000 j? for 5-10 s)
`before being finally resuspended in 500 pL of
`fresh PBS with 1% BSA. Samples were run
`on a FACScan (Becton Dickinson, Mountain
`View, CA, USA), and data were recorded as
`the percentage of cells bound relative to a
`negative control (anti-Ly-2.1) MoAb (murine
`IgG2a antibody reactive with the murine Ly-
`2.1 alloantigen) ± the standard error (quadru-
`plicate determinations) as calculated by a
`HP310 system (Hewlett Packard Australia
`Ltd, Melbourne, Australia).
`
`Immunoperoxidase analysis
`Analysis of MoAb reactivity was also as-
`sessed by immunoperoxidase staining of tresh
`
`IMMUNOGEN 2130, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`170
`
`M.J. Smyth etal.
`
`100 n
`
`90-
`
`8 0-
`
`2- VO-
`
`ID
`o 60
`
`Eo o
`
`50-
`
` 40-
`
`to
`^ 30
`
`20-
`
`10-
`
`0-
`1.0 X Iff'
`
`4.1 X Iff*
`1.6 X Itf*
`Antibody dilution"''
`
`6.A K
`
`Fig. 1. Antibody titre (% rosette-fonning cells) vs
`antibody dilution of 24-17.1 conjugates on Colo
`205 target cells. Serial dilutions were performed
`upon a 1.0 mg/mL solution of cither 24-17.1 (A)
`or Ida-24-17.1, four molecules Ida/molecule con-
`jugate (O) or lda-24-17.1, six molecules Ida/
`molecule conjugate (•) or Ida-24-17.1, or 15
`molecules Ida/molecule conjugate ( • ).
`
`(250-30.6 and 24-17.1) and formalin-fixed
`(27.1, I-l and JGT-13) LIM2210 tumour.'^
`Purified antibodies were used at a concentra-
`tion of 1
`
`Results
`
`Previous i« vitro (blocking with free MoAb)
`and in vivo (conjugates of irrelevant specificity)
`studies have demonstrated that the antitumour
`effects of Ida-MoAb conjugates depended pri-
`marily upon the antigen-binding specificity of
`the antibody.*" The aim of these studies was to
`demonstrate that a combination of Ida-MoAb
`conjugates using different tumour-reactive an-
`tibodies could be more effective against an
`established tumour than any one
`tumour-
`reactive Ida-MoAb conjugate alone. Initially,
`therefore, a panel of Ida-MoAb conjugates
`had to be established with varying cytotoxici-
`ties and reactivities to the tumours of interest.
`The 24 h cytotoxicity assay was used to esti-
`mate the relative cytotoxicity of two different
`Ida-MoAb conjugates to Colo 205; however,
`this assay was not feasible using LIM2210
`cells as the line cannot be established in vitro.
`Consequently, the relative reactivities of dif-
`ferent MoAb with the LIM2210 tumour were
`tested using immunoperoxidase and flow cy-
`tometry to establish the various combinations
`of Ida-MoAb conjugates to be assessed in vivo.
`
`Coupling of Ida to MoAh
`Five conjugates composed of Ida and different
`MoAb were prepared at drug : antibody molar
`ratios of 3-6 : 1 following the establishment
`of suitable reaction conditions using different
`molar excesses of Br-Ida. At least 50% of
`MoAb activity'** was maintained in all the
`conjugates containing less than six molecules
`of Ida/molecule of MoAb. For example, Ida-
`24-17.1 conjugates of four and six molecules
`of Ida/molecule of MoAb retained between
`50 and 70% of the original unmodified 24-
`17.1 antibody activity, while the antibody
`activity was reduced at an Ida incorporation
`ratio of 15 (Fig. 1). Therefore, Ida-MoAb
`conjugates that were used in vitro and in vivo
`had between three and six molecules of Ida/
`molecule of MoAb.
`
`Cytotoxicity of Ida-MoAb conjugates to
`Colo 205 in vitro
`The in vitro cytotoxicity of Ida and Ida-MoAb
`conjugates were measured in the 24 h cytotox-
`icity assay, and IC50 values were determined
`on the reactive Colo 205 and non-reactive E3
`tumour cell lines (Table 1). The IC<^,) for the
`Ida-24-17.1 conjugate was 11 times greater
`and four times greater for Ida-250-30.6 than
`that of free Ida. The selective cytotoxicity of
`both of
`these
`ida-MoAb conjugates
`for
`antigen-positive Colo 205 cells was demon-
`strated by their 5-9-fold lower cytotoxicity
`on antigen-negative E3 cells. Therefore, both
`Ida-250-30.6 and Ida-24-17.1 conjugates had
`satisfactory in vitro cytotoxicity to use against
`human colon carcinoma Colo 205 xcnografts
`in nude mice. The combined cytotoxicity of
`
`IMMUNOGEN 2130, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Tumour therapy with IDA-MoAb confugates
`171
`Table 1. Cytotoxicity of idarubicin and idaruhicin-MoAb conjugates; IC50 determinations usihg a 24 h
`cytotoxicity assay
`
`Tumour
`
`Ida
`
`Ida-250-30.6
`
`Mean IC,;,,(mol/L)
`Ida-24-17.1
`
`Ida-anti-Ly-2.1
`
`Colo 205
`E3
`
`8.0 X
`1.0 X
`
`10-
`3.3x10-'-(3)
`10- '(5)
`4.3x10-^(3)
`IC^o = 50% inhibition of ['^H]-thyniidine incorporation of controls. Parentheses indicate the number of
`preparations tested.
`
`3.0x10-^(3)
`2.6x10-''(2)
`
`
`9.0 >= 10-'(3)
`
`4.5 >clO''*(l)
`
`Ida-250-30.6 and Ida-24-17.1 conjugates was
`compared with that of either conjugate alone
`using a 30 min cytotoxicity assay (Fig. 2).
`Given
`that
`the
`Ida-24-17.1
`conjugate
`(ICso = 6.0 X 10^ * moI/L) was 6-7 times less
`cytotoxic than the Ida-250-30.6 conjugate
`(IC5o = 9.2x 10-^moI/L) against the Colo
`205 cell line, the cytotoxic effect of a mix-
`ture of these two conjugates was at least
`additive. Indeed the IC^^ of an equimolar
`mixture of Ida-250-30.6 and Ida-24-17.1
`conjugates (1.5 x 10-^moI/L) was compara-
`ble to that predicted by the sum of the
`
`individual conjugate growth inhibition curves
`('
`
`MoAb binding to colon cancer cells
`Flow cytometry indicated that both 250-30.6
`and 24-17.1 MoAb were reactive with the
`Colo 205 colon carcinoma binding 99 ± 1%
`and 94 ± 4% of the tumour cells, respectively.
`Quantitative analysis, however, demonstrated
`that the level of 250-30.6 antigen expression
`was approximately five times higher
`than
`24-17.1 antigen expression on Colo 205 cells
`
`100 -
`
`Log [IDA concentration]
`
`Fig. 2. The inhibitory effect of lda-24-17.1, 5 mol Ida/mol conjugate (•), Ida-250-30.6, 5 mol Ida/inol
`conjugate ( •) or ati equimolat mixture of Ida-24-17.1 and Ida-250-30.6 (C) on Colo 205 cells in a 30 min
`inhibition assay. For reference the predicted inhibitory effect of the mixture has been calculated (C) and
`demotistrated presuming that the conjugate's cytotoxicities are additive.
`
`IMMUNOGEN 2130, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`172
`
`M.j. Smyth et al.
`
`Table 2. Reactivity of individual MoAb and com-
`binations on L1M2210 tumour
`
`MoAb
`
`Mean fluorescence
`
`250-30.6
`27.1
`I-l
`JGT-13
`24-17.1
`250-30.6 + 27.1
`250-30.6 +27-1+JGT-13
`JGT-13+ 1-1
`JGT-13+ 1-1+24-17.1
`
`89
`400
`27
`55
`50
`536 (489)
`565 (544)
`65
`(82)
`118(132)
`
`Values in parentheses are the sum of the mean
`fluorescence of individual MoAb on LIM2210.
`
`liminary in vivo experiment assessing the abil-
`ity of each of the MoAb to target Ida to
`LIM2210 tumours.
`
`Antitumour activities of Ida-MoAb against
`Colo 205 tumour xenografts
`tested
`Ida-250-30.6 had previously been
`against the Colo 205 xenograft and was found
`to be particularly effective at inhibiting tu-
`mour growth. However only one of the 10
`conjugate-treated mice survived tumour-free.
`In contrast the antitumour effect of Ida alone
`or in combination with MoAb was minimal
`and limited by its toxicity,
`and consequently
`Ida alone or mixed with MoAb was not
`included as a control against Colo 205 tumour
`xenografts. To assess the effect of a combina-
`tion of Ida-250-30.6 and Ida-24-17.1 conju-
`gates, two experiments were performed
`in
`which conjugates alone and in combination
`were compared against the Colo 205 xe-
`nografts in nude mice. In the first experiment,
`nude mice (10/group) were inoculated s.c.
`with 2x10'' Colo 205 cells in the abdominal
`region and were treated with intraperitoneal
`(i.p.) injections of either PBS, Ida-24-17.1,
`Ida-250-30.6 or Ida-24-17.1 combmed with
`Ida-250-30.6. Mice received a total of 120 |jg
`of Ida and 7.2 mg of MoAb over days 4
`(tumour size 0.1 cm^), 6 and 11 after tumour
`inoculation, and throughout the course of this
`experiment it was evident that Ida-MoAb
`conjugates administered alone or in combina-
`
`as determined by their mean linear fluores-
`cence (MLF) in the presence of equal concen-
`trations of either MoAb (MLF = 1 6 6 + 13
`[250-30.6] compared with MLF = 35 ± 4 [24-
`17.1]). When both antibodies were used to-
`gether there was an increase in fluorescence
`(MLF=19O±5). Thus both antibodies used
`together should convey more Ida to the tu-
`mours than either alone. Flow cytometry indi-
`cated that both the 250-30.6 and 27.1 MoAb
`were reactive with the LIM2210 colon carci-
`noma binding 85-95 ± 3% of the tumour
`cells compared with JGT-13
`(21 ±5%),
`I-l (13 ±5%) and 24-17.1 (5 ± 3 %; Fig. 3).
`When
`the antibodies were used
`together,
`additional fluorescence was again apparent
`(Table 2). Immunoperoxidase staining also
`quantitatively demonstrated that both 27.1
`and 250-30.6 MoAb were considerably more
`reactive on LIM2210
`tissue sections than
`JGT-13 (weakly positive), 24-17.1 and I-l
`(data not shown). These results led to a pre-
`
`0 200 400 600 800 1000
`m
`
`[
`
`It
`
`ull scale
`
`= o
`
` -
`
`200 400 600 eOO 1000
`
`(a)
`
`^ _
`
`•J
`;
`O "
`
`10^
`10^
`10'
`Fluoroscance 1
`
`• jMir
`
`• • ' 1 —•
`10'
`10^
`
`10^
`
`10'
`
`Fluorescence I
`
`200 400 600 80O 1000
`
`200 400 600 BOO 1000
`
`10°
`
`10'
`
`10^
`
`10^
`
`10'
`
`Fluorescence 1
`
`10'
`
`Fiuoraseence 1
`
`0
`
`200 400 60O 800 1O00
`
`0 200 400 SOO BOC tOOO
`
`Fig. 3. FACscan analysis of MoAb binding to
`LIM2210 tumour cells plotted as fluorescence in-
`tensity vs cell number, (a) aiiti-Ly-2.1 (control),
`(b) 24-17.1, (c) I-l, (d) JGT-13, {e) 250-30.6 and
`(f) 27.1.
`
`IMMUNOGEN 2130, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Tumour therapy with IDA-MoAb confugates
`
`173
`
`tion could significantly inhibit Colo 205 tu-
`mour growth. Furthermore, several weeks
`after the completion of treatment (by day 50),
`the mice treated with a combination of Ida-
`250-30.6 and Ida-24-17.1 had a mean tumour
`size 4-5% that of PBS control-treated mice
`(i.e. a 95% decrease in tumour mass; Fig, 4).
`The mean tumour size of mice treated wtih
`Ida-24-17.1 or Ida-250-30.6 alone was 25%
`or 20% that of control treated mice, respec-
`tively. The observation that a combination of
`two Ida-MoAb conjugates was significantly
`more effective than either conjugate alone was
`somewhat explained by individual
`tumour
`growth profiles, which
`indicated that
`the
`combination of Ida-250-30.6 and Ida-24-17.1
`completely eradicated four of 10 tumours,
`whereas Ida-24-17.1 or Ida-250-30.6 alone
`only eradicated one of 10 or two of 10
`tumours, respectively (data not shown).
`A second experiment was designed to ascer-
`tain whether a combination of Ida-MoAb
`conjugates was more effective than any one
`Ida-MoAb against larger subcutaneous Colo
`205 tumours. Here, mice (10/group) were
`
`injected s.c. with 8 x 1 0 '' Colo 205 cells and
`were treated i.p. with one of PBS, Ida-24-
`17.1, Ida-250-30.6 or Ida-24-17.1 with Ida-
`250-30.6. Mice received a total of 200 ng of
`Ida and 10.0 mg of MoAb over days 17
`(tumour size 0.50 cm^), 20, 23 and 27 after
`tumour inoculation (Fig. 5). Again both Ida-
`24-17.1 and Ida-250-30.6 were less effec-
`tive at inhibiting tumour growth when given
`alone, although the mean tumour size (day
`45) of Ida-24-17.1 (69% of control)- or
`Ida-250-30.6 (54%)-treated mice was only
`slightly greater than that of Ida-24-17.1- and
`Ida-250-30.6-treated mice (40%). The com-
`bined treatment with both Ida-MoAb conju-
`gates eradicated one tumour, whereas neither
`conjugate alone was able to cause complete
`regression of the tumour mass. In addition, it
`should be noted that the antitumour effect of
`all the conjugate regimens was less dramatic in
`this second study, where treatment began with
`a tumour size five times that used in the first
`experiment (Fig. 4).
`
`Antitumour activities of Ida-MoAb against
`LIM2210 xenografis
`LIM2210 is a tumour recently obtained from a
`fresh colon cancer implant, and the effect of
`
`a
`
`t.1
`
`•- 08-
`
`E Q.G
`
`So--
`
`0 1-
`
`Time alter tumouf inoculaiion (days)
`
`t
`t 1
`I
`Time alter tumour inoculation (days)
`
`Fig. 4. Growth of the Colo 205 human colon
`carcinoma xenograft in Swiss nude mice injected
`s.c. with 2 x 1 0' cells. Groups of 10 mice were
`given i.p. treatments (arrowed) of PBS (D), Ida-24-
`17.1 (O), Ida-250-30.6 (•) or Ida-24-17.1 and
`Ida-250-30.6 (•). Error bars represent the standard
`error of the mean for tumour size (cm^). The
`non-specific conjugate (Ida-Ly-2.1) demonstrated
`an Ida effect in this model and gave the same results
`as PBS.
`
`Fig. 5. Growth of the Colo 205 human colon
`carcinoma xenoeraft in Swiss nude mice injected
`s.c. with 8x 10 cells. Groups of 10 mice were
`given i.p. treatments (arrowed) of PBS (D), Ida-24-
`17.1 (O), Ida-250-30.6 (•) or Ida-24-17.1 and
`Ida-250-30.6 (A). Error bars represent the standard
`error of the mean of tumour size (cm^). The
`non-specific cotijugate (Ida-Ly-2.1) demonstrated
`an Ida effect in this model and gave the same results
`as PBS.
`
`IMMUNOGEN 2130, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`174
`
`M. J. Smyth et al.
`
`Ida-
`Ida-24-17.1 or
`all following either
`JGT-13 conjugate treatment. It should be
`noted that in these experiments (and elswhere)
`a non-reactive
`immunoconjugate
`(Ida-anti-
`Ly-2.1) has no effect on the growth of these
`tumours.
`Based on these findings and the binding
`characteristics of these MoAb to LIM2210
`tumour cells, a larger study was undertaken to
`ascertain whether combinations of Ida-MoAb
`conjugates were more effective than any one
`conjugate alone. Mice (six or seven per group)
`were implanted s.c. with LIM2210 tumour
`(approximately 30 mm^) and were treated i.v.
`with one of the following: (i) PBS; (ii) Ida;
`(iii) Ida-250-30.6; (iv) lda-250-30.6 and Ida-
`27.1 (the two most effective); (v) Ida-250-
`30.6, Ida-27.1 and Ida-JGT-13 (the three
`most effective at different target antigens); (vi)
`Ida-JGT-13 and lda-I-1 (two detecting CEA
`at the same target antigen); or (vii) Ida-JGT-
`13, Ida-I-1 and Ida-24-17.1 (the three least
`effective). Mice received a total of 115 Mg Ida
`and/or 5.8 mg of MoAb over days 32 (tumour
`volume 0.20 cm"^), 33, 35 and 39 after tumour
`implantation (Fig. 7). The LD,o of Ida has
`previously been demonstrated to be in the
`range of 0.75-1.00 mg/kg,^ and therefore it
`was not surprising that at this dose level Ida
`alone was toxic, killing five of six mice by day
`42 and the sixth mouse in this group by day
`50 (data not shown). By contrast, no Ida-
`MoAb conjugate alone or in combination with
`others was lethal, nor were there signs of
`toxicity such as weight loss, fatigue or fever.
`Despite the fact that the mean tumour volume
`of all the groups of mice were similar at the
`commencement of treatment, each individual
`group of mice contained tumours varying
`considerably in size. Consequently the large
`standard error of each group made it difficult
`to compare the mean tumour volume of
`groups of mice treated with panels of Ida-
`MoAb conjugates. However, some trends in
`the individual tumour growth profiles of mice
`were noted, which indicated that combina-
`tions of the most efficacious Ida-MoAb conju-
`gates were more effective in the treatment
`of LIM2210 tumour xenografts (Fig. 7). Indi-
`vidual tumour growth curves (Fig. 7) indi-
`cated that no treatment regimen completely
`eradicated LIM2210 tumour, although
`the
`
`immunochemotherapy in this model was of
`interest. An initial experiment was performed
`to assess the individual efficacy of a variety of
`different Ida-MoAb conjugates constructed
`with MoAb that had been shown to bind
`LIM2210 cells in vitro. Groups of Swiss nude
`mice (eight/group) were implanted s.c. with
`LIM2210
`tumour
`pieces
`approximately
`30 mm-* in volume and were treated i.v. with
`either PBS, Ida-24-17.1, Ida-JGT-13, Ida-
`27.1 or Ida-250-30.6. These mice received
`20 |ig of Ida and 1.0 mg of MoAb on days 19
`(tumour size 0.25 cm^), 20, 21, 25 and 26
`after tumour implantation (total 100 fig Ida;
`Fig. 6). All four Ida-MoAb conjugates dem-
`onstrated antitumour activity against
`the
`LIM2210 colon carcinoma xenografts. The
`most effective was the Ida-250-30.6 conjugate
`which, by day 35, had inhibited
`tumour
`growth by 75% compared with the PBS
`control (mean tumour size 0.39 cm^ com-
`pared with 1.58 cm*^). At the same time,
`Ida-27.1 conjugate inhibited control growth
`by 58% while both Ida-24-17.1 and Ida-
`JGT-13 were less effective (34%). In both
`Ida-250-30.6- and Ida-27.1-treated groups,
`one in eight tumours were completely eradi-
`cated, whereas no tumours regressed in size at
`
`1.8-
`
`1.6-
`
`" 1.0-
`
`E O . B-
`
`D c
`
` 0.6-
`ra
`IS
`
`2 0-4-
`
`0.2-
`
`25
`tt
`t
`tt
`Time after tumour inoculation (days)
`
`Fig. 6. Growth of the LIM2210 human colon
`carcinoma xenograft iti Swiss nude mice implanted
`s.c. with 30 mm pieces of tumour. Groups of eight
`mice were given i.v. treatments (arrowed) of PBS
`or Ida-Ly-2.1 (D), Ida-24-17.1 (•), Ida-250-30.6
`(O), Ida-27.1 (•) or Ida-JGT-13 (A). Error bars
`represent the standard error of the mean of tumour
`size (cm^).
`
`IMMUNOGEN 2130, pg. 8
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Tumour therapy with IDA-MoAh conjugates
`
`175
`
`0.0
`
`Time alter inocuiation (days)
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`Time after inoculation (days)
`
`Pig. 7.
`Individual growth curves of LIM2210 tumour implanted s.c. into Swiss nude mice and treated i.v.
`with: (a) PBS; (b) Ida-250-30.6; (c) Ida-250-30.6 atid Ida-27.1; (d) Ida-250-30.6, Ida-27.1 and Ida-JGT-
`13; (e) Ida-JGT-13 and Ida-I-1; and (f) Ida-JGT-13, Ida-I-1 and Ida-24-17.1.
`
`regression of tumour mass (both > 50% and
`< 50% of initial untreated tumour size) was
`observed in a number of mice receiving either
`a combination of Ida-250-30.6 and other Ida-
`MoAb conjugates or Ida-250-30.6 conjugate
`alone (Table 2). Within 1 week of the com-
`pletion of treatment it was evident that three
`of six tumours of mice that received Ida-250-
`30.6 conjugate alone had regressed compared
`with one of six of those given combinations of
`the less effective Ida-JGT-13, Ida-24-17.1 and
`Ida-I-1 conjugates (Table 3). In addition, by
`day 46 the mean tumour volume of mice
`treated with Ida-250-30.6 alone was 0.68
`±0.23cm^^ compared with 1.00±0.33cm^
`tor
`the group receiving Ida-JGT-13 and
`Ida-I-1 conjugates (PBS mean tumour vol-
`
`ume = 1.31 ±0.26 cm'*). Interestingly, those
`groups receiving Ida-27.1 both with or with-
`out Ida-JGT-13 in addition to Ida-250-30.6
`conjugate demonstrated four of six and six of
`seven regressions, respectively (Table 3). On
`day 46 the mean tumour volume of mice
`treated with Ida-250-30.6, Ida-27.1 and Ida-
`JGT-13 was 0.33±0.12cm^ while
`those
`mice receiving a combination of Ida-250-30.6
`and Ida-27.1 had a mean tumour volume of
`0.50 ±0.23 cm"^. It was also noted that differ-
`ences between the antitumour effects of a
`combination of Ida-MoAb conjugates and
`Ida-250-30.6 conjugate became less evident
`as LIM2210 tumours began
`to regrow 2
`weeks after the completion of treatment (days
`54-67).
`
`IMMUNOGEN 2130, pg. 9
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`176
`
`M.f Smyth et al.
`
`Table 3. Antitumour activity of idarubicin-MoAb
`conjugates on LIM2210 colon carcinoma xeno-
`grafts in nude mice
`
`Complete
`regressions
`
`Partial
`regressions
`
`Minor
`response
`
`0/6
`
`0/6
`
`0/7
`
`0/6
`
`0/6
`
`0/6
`
`0/6
`
`1/6
`
`1/7
`
`1/6
`
`0/6
`
`1/6
`
`0/6
`
`3/6
`
`6/7
`
`4/6
`
`1/6
`
`1/6
`
`Treatment
`
`PBS
`
`Ida-250-30.6
`
`Ida-250-30.6
`Ida-27.1
`
`Ida-250-30.6
`Ida-27.1
`Ida-JGT-13
`
`Ida-JGT-13
`Ida-I-1
`
`Ida-JGT-13
`Ida-I-1
`Ida-24-17.1
`
`Complete regression (no evidence of tumour
`> 100 days after tumour inoculation); partial re-
`gression (> 50% reduction in original tumour vol-
`ume); minor response (< 10-50% reduction in
`original tumour volume).
`
`Discussion
`
`Tumour cell heterogeneity may have a pro-
`found effect on the use of MoAb as carriers of
`cytotoxic agents. Differences in antigen ex-
`pression have been observed between primary
`tumours and their metastases.'^"^^ The het-
`erogeneity (including not only whether a
`tumour can bind antibody but how much is
`bound) is of crucial importance for determin-
`ing the amount of drug reaching the intracel-
`lular target. Despite the improved defmition
`of tumour-associated antigens and their epi-
`topes.^^"^^ it is probable that the effective
`targeting of antineoplastic drugs to tumours
`using MoAb will require combinations of
`drug-MoAb conjugates. Tumour cell popula-
`tions resistant to any one drug-MoAb con-
`jugate may occur as a result of the preferential
`survival and growth of antigen-negative or
`antigen-low cells and the modulation or shed-
`ding of antigen following
`treatment.^^'^^
`Although panels of MoAb have been dem-
`
`than single
`onstrated to be more effective
`MoAb for both the diagnosis of cancer^*^ and
`the treatment of human tumour xenografts,'"'
`no studies to date have compared different
`mixtures of drug-MoAb conjugates and their
`antitumour activity in vivo.
`Individual Ida-MoAb conjugates have pre-
`viously been shown to be effective in the
`treatment of murine thymomas,^""^^ however
`their use has been limited to the eradication of
`small subcutaneous murine tumours and their
`less impressive effects agaiust human colon
`carcinoma xenografts in nude mice.'** Iu addi-
`tion, a combination of Ida-MoAb conjugates
`were demonstrated to be selective agents in
`the depletion of Ly-2' and L3T4* T cell
`subsets in mice for the prolongment of graft
`survival in a tumour allograft model."'" Given
`these studies and the known heterogeneity of
`some colon carcinoma cell populations," it was
`believed that combinations of Ida-MoAb con-
`jugates could be more effective in the treat-
`ment of human colon carcinoma xenografts
`than any cue Ida-MoAb alone. Indeed, using
`both established homogeneous and fresh het-
`erogeneous carcinoma of the colon xenograft
`models, the study herein has demonstrated
`that
`improved antitumour effects can be
`achieved with mixtures of the most effective
`Ida-MoAb conjugates.
`
`Ida was conjugated to MoAb reactive with
`CEA (JGT-13. 1-1).