HNTIBDDEI
`munocunjugates,and
`
`adinpharmaceuticals
`
`IMMUNOGEN 2329, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`GENERAL INFORMATION
`
`Antibody, lmmunoconjugates, and Radinpharmncemicals will publish fundamental scientific, preclinical,
`and clinical studies which examine the use of potential antibody, immunoconjugates, or radiopharmaceui
`ticals.
`
`Antibody, lmmunaconjugales, and Radinpharmaceuticalx (ISSN: 0892-7049) is published quarterly for
`$90 per year by Mary Ann Lichen, lnc., 1651 Third Avenue, New York, NY 10128, (212) 289-23001
`
`Postmaster: Send address changes to: Antibody, Immunoconjugaies, and Rudiaphurmaceufit‘als,
`”/0 Subscription Department, Mary Ann Liebert, lnc., 1651 Third Avenue, New York, NY 10128.
`
`Antibody, Immunaconjugalex, and Radiopharmaccuiicals is owned and published by Mary Ann Liebert,
`Inc.
`
`Subscriptions should be addressed to the Publisher and are payable in advance. Rates for subscriptions
`are $90 per volume in the United States and Possessions and $123 elsewhere, Subscriptions begin with the
`first issue of the current volume.
`
`Reprints, except special orders of 100 or more, are available from the authors.
`
`Information for Manuscript Submission is given elsewhere in the publication.
`
`Business Communications should be addressed to the Publisher.
`
`Advertising Inquiries should be addressed to Mary Ann Licbert, lnc., 165] Third Avenue, New York,
`NY 10128. (212) 289-2300.
`
`Manuscripts should be directed to the Editor, Stanley E. Order, MIL, Sc.D_, F.A.C.R.. Radiation
`Oncology, Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21205. (301) 955—6980,
`Contributions to this journal are published free of charge.
`
`All authored papers and editorial news and comments, opinions, findings, conclusions, or recommenda—
`tions in Antibody, Immunaconjugates, and Rudiopharmaceuticals are those of the author(s), and do not
`necessarily reflect the views of the journal and its publisher, nor does their publication in Antibody,
`[mmunocanjugalex and Radiopharmaccuticali imply any endorsement.
`
`Copyright © 1988 by Mary Ann Liebert, Inc. Printed in the United States of America.
`
`M7 M4
`
`, M. W - New York
`
`IMMUNOGEN 2329, pg. 2
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`ANTIBODY, IMMUNOCONJUGATES, AND
`HADIOPHARMACEUTICALS
`Volume 1, Number I, 1988
`Mary Ann Licbert, Inc, Publishers
`
`Preclinical and Clinical Studies with a
`
`Variety of Immunoconjugates
`
`GEOFFREY A. PIETERSZ, MARK J. SMYTI—I, JERRY KANELLOS.
`ZITA CUNNINGHAM, NIGEL P.Ml SACKS,
`and IAN EC. MCKENZIE
`
`Research Centre for C(nicer am! Transplanmmm, Deparmzuu! ufPuI/mlugv. University of
`Melbourne. Par/(ville, Vic.. 3052, Amrruliu
`
`ABSTRACI‘
`
`The use of monoclonal antibodies to target cytotoxic drugs to solid
`tumors is an attractive concept, but has yet to make a radical impact on
`the therapy of cancer.
`A number of variables that could influence the
`efficacy of drug-antibody conjugates were investigated - (a) Drug
`‘ the
`covalent attachment of drug to monoclonal antibody was variable;
`anthracyclines such as adriamycin (AD) coupled poorly, yet analogs SUCh as
`branoidarubicin were coupled successfully; chlorambucil
`(CBL) and m91Pha13-n
`(MEL) were unique in that a greater number of drug molecules could be
`coupled to monoclonal antibody. The requirement for more cytotoxic drugS
`was clear when aminopterin (AMN) , a more cytotoxic analog of methotrexate
`(MI'X) , was coupled to monoclonal antibody and found to be more effective
`than methotrexate conjugates. Of all the drugs used,
`idarubicin (Ida) was
`the most effective in vivo (Ida > MEL > Am > Ml‘X > 03L 1’ AD)-
`03)
`Antibody -— in several tumor growth models Flab')2 drug complexes were as
`effective as IgG, but not more so.
`(0) Access to tumor — the importance Of
`tumor access was demonstrated when tumors growing subcutaneously were
`eradicated by the local injection of whole ricin-antibody COUjUQates and
`intraperitcneal (i.p.) tumors were also easily eradicated by 1-P-
`treatment.
`By contrast,
`tumors growing in the subcutaneous site are less
`susceptible to therapy, however, vasoactive agents increased the w
`efficacy of drug—antibody conjugates.
`(d)
`’I'umor - the problem of tumor
`heterogeneity was addressed by using a cocktail of two drug—antibody-
`conjugates for tumor therapy;
`the cocktails were clearly more effective
`than either conjugate used alone.
`0n the basis of these results, phase I
`studies are in progress using MTX—anti—oolon cancer monoclonal antibodies
`given intravenously and ricin anti—breast antibodies given into the tumors.
`
`Ira-mum
`
`In contrast to the treatment of lymphcma and leukemia where a
`proportion of patients obtain a complete remission, cytotoxic agents for
`the treatment of most solid tumors are clearly less effective (1) and there
`
`79
`
`
`
`IMMUNOGEN 2329, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`is usually accompanying toxicity. Consequently there have been attempts to
`"target" cytotoxic agents to tumors. The advent of monoclonal antiodies
`(MoAbs) has been an important step in the developnent of drug targeting
`whereby the MoAbs can now be used to convey cytotoxic drugs to tumor cells
`(2—5) . Drug—MoAb conjugates, upon binding to target cells, may be
`specifically internalized and degraded to free drug, which then acts on the
`target
`(6-9) . This approach to drug targeting does however, present many
`problems both in vitro and in vivo. Firstly, it is difficult to couple
`cytotoxic drugs (mostly organic chemicals)
`to hydrophilic antibody
`molecules with retention of both drug and antibody activity; thus, despite
`retaining selective cytotoxicity for target cells, the conjugate is
`generally less cytotoxic than the free drug. As a result it is necessary
`to couple more cytotoxic drugs to MoAbs to obtain greater antitumor
`effects; our efforts have been directed towards conjugating MoAbs to more
`cytotoxic analogs of drugs. Other problems include the in vitro and i_n
`
`vivo stability of the conjugates, and the ability of immunoconjugates to
`penetrate throughout tumors. Here we present a review of our studies using
`different drug—antibody conjugates;
`including their potentiation with
`vasoactive agents and the use of cocktails of drug-antibody conjugates.
`
`WANDHETHCDS
`
`Tumor Cells
`
`E3,a clonal variant of the murine thymoma I'I‘I‘(l)75NS (10) was
`maintained in vitro in Dulbecoo' 5 Modified Eagles Medium (DME) supplemented
`with 10% heat inactivated newborn calf serum (Flow Laboratories, Sydney:
`Australia),
`21:14 glutamine (Co-rmonwealth Serum Laboratories (CSL) ,
`Melbourne, Australia),
`loopg streptanycin (Glaxo, Melbourne, Australia) and
`lOOIUml-l penicillin (CSL).
`For in vivo experiments E3 cells were
`maintained by serial passage in ascites fluid in (C57BL/6 x BALE/CH?
`(B6CE‘ ) mice. Human cell lines used were COLO 205 (11)
`«3 001011 carcmoma.
`and (IF-M (12) a ri‘-lymphocyte leukemia cell line; these were maintained in
`culture in RPMI—1640 medium (Flow Laboratories) with the same additives as
`above. Adherent cells were harvested with 0.125% trypsin (CSL) I washed
`with RPMI—1640, and either used for the in vitro assays or injected
`subcutaneously into nude mice (3—5x106) and grown in vivo as xenoglfafts 0“
`the abdominal wall of nude mice.
`
`Mice
`
`Nude mice (Swiss) were obtained from the Animal Resources Centre
`(Perth, Western Australia), and B6C15‘l mice were produced in our department.
`
`Monoclonal Antibodies and Serolggy
`
`A
`Several monoclonal antibodies were used in this study (Table 1).
`rosetting assay (18) was used to determine the antibody activity of the
`drug—antibody conjugates, as a control antibody that had undergone the same
`procedures used in the coupling methods (other than adding drug) was used.
`
`Preparation of Drug-antibody Conjugates
`
`Chlorambucil (19). N~acety1 melphalan (20, 21) and methotrexate (22)
`were coupled to MoAb as described previously using an active ester
`derivative of these drugs. Amimpterin was also coupled by a similar
`method but was dissolved in dimethyl sulphoxide rather than
`dimethylfomamide (23).
`Idarubicin was coupled to MoAb via the
`
`80
`
`IMMUNOGEN 2329, pg. 4
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`'RBLE 1
`Details of Moths used in Experimtal Studies.
`Antibody
`Reference
`Purification
`Target Cell
`Procedure
`
`
`Immunoglobulin
`Subclass
`
`Ly—2.l
`
`250-30.6
`
`13
`
`l4
`
`anti-Transferrin
`Receptor
`('I'E'R) 15
`17.1
`16
`Ly-l.l
`17
`
`Protein A
`
`Affigel Blue
`
`I'I‘I‘(l)75NS E3
`thymcma
`COL0205
`(Ca colon)
`
`IgGZa
`IgG2b
`
`CEIM
`Affigel Blue
`COL0205
`Affigel Blue
`CBA thymus
`Protein A
`_——_._.___.—————-—-——————-——‘————
`
`IgGl
`19G23
`19‘323
`
`(l—2mg/ml in 0.05M
`brcmoidarubicin (Br—Ida) derivative (24). Briefly, MoAb
`Borate pH 8.0) were mixed with molar excesses (0—50) of l4—brano—4-
`demethoxydaunomycin (Br—Ida) dissolved in N,N—dimethylformamide (EMF) at
`10mg/ml. The reaction was maintained at rocm temperature for 4 hours,
`before centrifuging (400g x 5 minutes) to remove any precipitate; free
`Br-Ida and other unreacted materials were removed using a Sephadex G25
`column (PDlO) and the conjugates were then passed through a column of
`Porapak Q to remove non-specifically bound drug (25) . The amount of
`idarubicin incorporated in drug—antibody conjugates was determined by
`absorbence spectrophotanetry at 483nm (E,83 = 3.4 x 105 m-1 cm-l) , 5115
`protein estimated (26). Adriamycin was coupled to MoAb using the '
`iodoacetyl derivative (27); sulfydryl groups were exposed by treating
`antibody (lml,
`lmg/ml) with dithiothreitol (DTT)
`(75,1, 1M) for.4§m1ns;
`the mixture was desalted by gel filtration on a PDlO column equilibrated
`with deoxygenated 0.01M Tris-saline buffer pH 8.6, and the protein
`collected.
`Iodoacetyl adriamycin (0.4mg)
`in 100 l DMF was added to the
`reduced antibody (2.5ml) , and allowed to stand for 2.5—3 hrs and the
`precipitate which formed was removed by centrifugation and the supernatant
`purified by gel filtration.
`The number of residues (N) of adriamycm
`molecules bound per immunoglobulin molecule was calculated to the formula:
`N = 215,000 x A280 / (11,600 x Am) — (8,600 x Am)
`where Auso and Azuu are the absorbance of the conjugate at 480 and 280nm
`using the extinction coefficient of adriamycin at 480 and 280nm of ll, 600 m
`CHI-l and 8:600 In”l cm”1 respectively; the extinction of imxmmoglobulln at
`280nm is 215,000 m-1 cm’l.
`
`i
`
`In vitro cell inhibition assays
`
`Tm types of assays were carried out to test for residual drug activity
`0f the anugates in comparison with free drug.
`(a) 24 murassay:
`100111
`of cells (1-5 x 106 /ml) was added to a 96-well flat bottom microtltre plate
`and incubated for 2-3 hours at 37°C; sterile antibody, free drug or
`.
`conjugate was diluted in PBS and a 50141 aliquot was added to cells using
`duplicate wells per sample. Controls received 50141 of PBS and the cells
`were cultured at 37°C in 7% 00
`for 24 hrs.
`(b) 30 minute assay:
`200111 of
`09115 (1-5 x lOfi/ml) was collegted in sterile plastic tubes, resuspended 1n
`sterile antibody, free drug or conjugate and mixed for 30 minutes at 37_C.
`The cells were centrifuged (400g x 5 min) and resuspended in growth medium,
`then 100141 cells were added to microtitre wells in duplicate and 1ncubated
`for 24 hrs. After the incubation period in both assays,
`luCi of.
`[ifilrthymidine (specific activity = lSCi/nmol; Amersham Internation Ltd,
`Amersham, England) or [3H1-deoxyuridne (specific activity = lSuCi/Irmol in
`50111 medium) was added and the plates incubated for a further 3-6 hours.
`Then the cells were harvested onto glass fibre filter paper, dried and the
`samples counted for radioactivity on a 8 counter.
`Incorporation of
`
`81
`
`IMMUNOGEN 2329, pg. 5
`Phigenix v. Immunogen
`|PR2014—00676
`
`IMMUNOGEN 2329, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`radionucleotide was expressed as the percentage inhibition in incorporation
`of controls; the standard error for any given point did not exceed 5%.
`
`In vivo experiments
`
`Tumor cells
`(a) Survival:
`Several different approaches were used.
`were injected intraperitoneally and mice received various treatments by the
`same route and the survival of the mice was calculated.
`(b) Tumor growth:
`Timor cells were injected subcutaneously into the abdominal wall and
`allowed to develop into a palpable lump at which time treat-merit (i.p. or
`i.v.) was commenced. The size of the tumors was measured daily with a
`caliper square along the perpendicular axes of the tumor and the data was
`recorded as a mean tumor size (products of two diameters i standard error).
`Experimental groups of 10—20 mice, all of the same sex and age, were used
`in each experiment.
`In other studies, tumors were injected directly.
`
`B iodistr ibution
`
`BSCFi mice bearing subcutaneous E3 tumors (0.5 - 1.0cmz) were used to
`compare the distribution of 125I-ant-_i—Ly—2.1 in the presence or absence of
`therapeutic levels of different vasoactive agents. Groups of 4 mice were
`sacrificed at 24 hrs after the injection of labeled anti—Ly-Z-l and the
`biodistribution of
`‘25I-anti-Ly-2.l was determined by counting the
`radioactivity in blood, heart, spleen, liver, kidneys and tumor Eran these
`mice.
`The distribution of isotope is reported as a localisation ratio5
`lztissue (cpm/g) / blood (CW9) ,. All mice received 500,000 cpm of 1‘ I
`iapprox. 20 g protein) by tail vein injection, one hour after the
`intravenous or oral administration of the vasoactive agent.
`
`Clinical Studies
`
`In a study using MTx-MoAb conjugates, patients with metastatic,
`'
`histologically confirmed, colorectal cancer were included if they had
`fulfilled the eligibility criteria of 3-month expected survival, Eastern
`Cooperative Oncology Group (EmG) performance status of 0-2, presence 0f
`measurable disease, no other therapy for at least 1 month prior to MoAb—MI‘X
`administration, and normal hepatic and renal funciton. Four patients have
`been treated, with informed consent obtained from each. The measured
`lesions were studied 1 and 3 months after treatment with the same technique
`as that used for the initial evaluation. Blood tests were done to assess
`potential hanatological, renal or hepatic toxicity, and to detect human
`anti-mouse antibody _(HAMA)
`formation (assayed by an ELISA based test).
`Serum'levels of carcmoembryonic antigen (CEA) were determined using a
`radiommunoassay for CEA (CEA-RIA, Abbott Laboratories, Chicago, 111- USA) -
`
`Administration of Antim Conjugates
`
`Three patients each received a total dose of 100mg MoAb covalently
`bound to approximately 4mg of MI'X. The second dose given to the first
`patient caused a mild allergic reaction, so the dose was not escalated
`until the lower dose was tolerated by the next patient. Dose escalation
`was awarding to the modified Fibonacci Sequence (28) up to a maximum dose
`of 5001119 MoAb : 20mg MIX/metre? of body surface area. This is well below
`the usual therapeutle dose Of MTX used in human without folinic acid rescue
`(25mg/m2) canpared With lf_300mg/m2 that can be given with folinic acid
`rescue.
`The antibody conjugate was diluted in 500mls of normal saline and
`administered over 6-8 hours whilst imder strict medical observation with
`the patient-s being carefully monitored for change in pulse, blood pressure,
`
`82
`
`IMMUNOGEN 2329, pg. 6
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`temperature and respiratory function for 24 hours. Ccmplete physical
`examinations were made after ccmpletion of the treatment and on the
`following day. All patients received a 48 hour course of systemic
`corticosteroids (to diminish hypersensitivity phenomena). Human anti—mouse
`antibodies (HAMA) were measure, before and after treatment; no patient with
`pre—existing HAMA was found. Patients were examined for signs and symptoms
`of serum sickness for 1 month after the infusion.
`Pharmacokinetics:
`To determine the pharmacokinetics of the MoAb—MI'X
`conjugates, serial blood samples were obtained (at 12, 24, 4B and 72 hours)
`after infusion. Affinity—purified sheep anti—mouse IgG (SAMG; Amersham,
`UK) was diluted to mug/soon in PBS, plated on 96 well PVC plates (Costar:
`Cambridge, MA) and incubated overnight at 4°C.
`the plates were then washed
`6 times in PBS/0.05% mean 20 (Sigma Chemical Co., St Louis, MO, USA) and
`serum samples were diluted 1:32 in PBS/Tween diluent. Diluted serum (50u1)
`was then added with 10 counts per minutes (cpm) of 125I—r’nlxb labeled
`canpetitor (in 50111 of the same diluent) to the SAMG coated wells. After
`overnight incubation at 4°C,
`the plates were washed in PBS/Tween and (31:18d
`at 37°C and the wells were counted in a ganma counter to determine the
`amount of radioactivity bound per well. Each time the assay was performed
`a standard curve was generated using dilutions of. purified unlabeled MoAbs
`of the isotype being assayed.
`The standard curve was generated by plotting
`the percentage of bound radioactivity/well
`(minus background) versus_the
`log concentration of unlabeled competitor MoAb.
`The amount of MoAb in the
`serum samples was then calculated by relating the average 0pm bound/well t0
`the concentration of unlabeled McAb producing an equivalent level of bound
`radioactivity .
`
`Reflnse Criteria
`
`Response to treatment was assessed using standard criteria as SUSSESted
`by Miller et al (29). Complete Response:
`(CR)
`is the disappearance °f all
`clinical evidence of tumor for at least 4 weeks. Partial W? (PR)
`is a reduction of at least 50% in the sum of the PIOdUCtS Of all diameters
`of measured lesions, lasting for at least 4 weeks. There must be no
`objective progression of any existing lesion and no new lemons may appear.
`There must be a significant reduction in the size of an evaluable lesmn-
`Stable Disease:
`(SD)
`is an objective regression of measurable disease less
`than required to meet the criteria for partial response or less than a 25%
`decrease in a measurable lesion over 4 weeks. Progressive Disease:
`. (PD)
`is an increase in the sum of the product of the two greatest perpendicular
`diameters of any measurable lesion by 25% or more, or obv10us increase 1“
`an evaluable lesion. Appearance of new areas of malignant disease
`signifies progressive disease. niraticn of Response: Duration of response
`was measured from the achievement of maximal response to the first 5191'1 Of
`disease progression.
`
`REUL'lS
`
`Conjugation of Antinfllastic Drugs to Monoclonal Antibodies
`
`We have examined a number of different drugs in an attempt to Obtain
`the most potent drug/antibody conjugate. The conjugation procedures are
`detailed here and the activity discussed in the following sections.
`Mriamycin: Despite the side effects associated with the use of
`adriamycin, it is widely used; and although the exact nature of Its
`cytotoxicity is not clear, intercalation of DNA is important.
`To reduce
`systemic toxic effects, several groups have coupled adriamycimto
`.
`'
`monoclonal antibodies wifli varying success (30). However, adriamycm ls
`coupled to antibody with some difficulty, so a number of analogs to
`facilitate coupling were prepared (Fig.
`lA—D) . of these, only the
`
`83
`
`IMMUNOGEN 2329, pg. 7
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`(A)
`
`0
`
`OH
`
`0
`
`OH
`
`“ OH
`
`Adriamycin (Ad)
`
`o
`
`N
`
`|
`
`MPB — Ad
`
`0
`
`Succ—Ad
`
`IA—Ad
`
`OCH3 0
`
`I
`
`('3
`
`OH
`0
`
`HO
`
`””2
`
`0 l
`
`l
`Ad —— NHCCHZCHZCHZ
`
`0
`o
`II
`II
`Ad—NHCCHZCHZC— OH
`
`0 l
`
`l
`Ad—NHCCHZ—I
`
`(B)
`
`(C)
`
`(D)
`
`FICIJRE 1
`Chemical structures of Adriamycin (A) and derivatives. B, Maleimidophenlfl“
`butyryl adriamycin (MPB—Ad); C, Succinyl adriamycin (Succ—Ad) and D,
`Iodoacetyl adriamycin (IA—Ad) . Reprinted with permission fromm
`Diagnosis and 'I’nerm (27) .
`
`iodoacetyl adriamycin analog resulted in an active conjugate (Fig. 2) where
`Up to 8 molecules of adriamycin could be coupled to antibody with good
`protein recovery and antibody activity (Table 2) .
`In these studies
`adriamycin non-covalently bound to MoAb was removed using a porapak Q
`column (25).
`
`Drug
`
`'JEBLE 2
`ijugatim of Drugs and mabs.
`MoAb Activig
`Number of drug
`Protein
`MoAb
`molecules per
`Reoovery(%)
`Before
`After
`antibody molecule
`conjugation
`
`Adriamycin
`8
`70
`Ly—2.l
`1/128,000
`l/32,000
`Chlorambucil
`25
`70
`Ly—2.l
`1/50,000 1/15, 000
`Melphalan
`25
`65
`Ly-2.l
`l/75,000 1/20.000
`"
`25
`55
`Ly—2.1 (IgGl)
`F'(ab')2
`Ly—2 . 1
`80
`13
`Methotrexate
`Ly-2 . l
`97
`6
`Aminopter in
`Idarubicin
`5
`50
`Ly-2.l
`l/80,000 1/56,000
`"
`4
`60
`250-30.6
`l/33,000 1/11,000
`
`
`1/32
`
`1/32
`
`84
`
`IMMUNOGEN 2329, Pg. 8
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 8
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`ll
`Ad—NHCCH2 —I
`
`/—' Ab-SH
`
`0I
`
`I
`Ad-NHCCHa-S -Ab
`
`_
`.
`FIGJRE 2
`Coupling of adriamycin to monoclonal antibody. Reprinted With permISSiOI'l
`from Targeted Diagnosis and Therm (27) .
`
`Gibrammcil: This alkylating agent is used for the treauuent of
`various leukemias,
`lymphanas and for breast and ovarian carcinoma, although
`marrow suppression is one of the side effects. Chlorambucil
`(F19. 3A) was
`one of the first drugs used for coupling to polyclonal antibodies and
`indicated that complexes with CBL may be formed at low or high pH, 31310th
`the exact nature of the bond was unknown.
`
`CHLORAHBUCIL
`
`0
`n
`[E‘CHZCHZCHZ
`DH
`
`NHR
`I
`H—[li—CHZ
`6:!)
`II)”
`
`/Cl
`CH —CH
`2
`2
`N/
`\CH —CH
`2
`2 C]
`
`/Cl
`CH -—CH
`2
`2
`N/
`\ —CH
`CH2
`2
`
`C]
`
`HELPHALAN
`R = H
`R : CH3CD N-ACETVL HELPHALAN
`FIGURE 3
`Chemical structures of chlorambucil
`(CBL) , melphalan (MEL) and N—acetyl
`melphalan (N-ACMEL) .
`
`85
`
`IMMUNOGEN 2329, pg. 9
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 9
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`An active ester derivative of CBL was coupled to monoclonal antibodies
`(Fig. 4); this is a mild procedure and 20—30 CBL molecules'could be-ooupled
`to an antibody molecule with good protein recovery and antibody act1v1ty
`(Table 2); 80% of CBL was covalently linked and 85% of the alkylating
`activity was preserved.
`
`0
`o
`f
`NHS
`I
`DRUG—C—OH haoauG—c—o—N
`CDI
`
`o
`\\
`
`/O
`
`/
`
`AB—NH2
`
`fl
`DRUG—‘C—NH—AB
`
`FIGIME 4
`(CBL) , N—acetyl melphalan (N—AcMEL) , methotrexate
`Coupling of Chlorambucil
`(MTX) and aminopterin (AMN)
`to monoclonal antiobdies using the active ester
`method.
`
`mthotrexate W) and Aminqmerin (mm: MI'X was coupled to antibody
`using an active ester (Fig. 4) as the y—glutamyl carboxyl group of MI'X
`(Fig. 5A) can be modified without effect on the dihydrofolate reductase
`binding ability. Conjugates with a drug:antibody ratio of 13 were formed
`with good yield and activity (Table 2).
`The more toxic folic acid
`antagonist, aminopterin (AMN)
`(Fig. SB) , was coupled to antibody in a
`similar way to MTX, but due to its low water solubility, the incorporation
`of AMN was less than MI‘X, with 6 molecules of aminopterin per antibody
`molecule bound (Table 2).
`Helphalan: This is a more potent alkylating agent than chlorambucil
`and was coupled to antibody using a new approach. Melphalan enters cells
`via the amino acid transport system and its multifunctional nature makes it
`
`_O
`C—NH— [iH—CUUH
`CH 2
`I
`2
`CH —I:UUH
`
`_
`
`CH2
`
`IT
`R
`
`H N
`2
`
`\{N N
`No 01N
`
`NH2
`
`R=CH3
`R:H
`FIG'JRBS
`chemical structures of methotrexate (MTX) and aminopterin (Ah/LN) .
`
`HETHUTREXATE
`AHINDPTERIN
`
`86
`
`IMMUNOGEN 2329, Pg. 10
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 10
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`difficult to couple to antibodies; the amino group of melphalan, necessary
`for cell uptake but not cytotoxicity, was blocked with an acetyl group and
`the resulting N-ACMEL (Fig. 3C) was coupled to antibody via an active ester
`(Fig. 4). The procedure removed the ability of the Melphalan to enter
`cells by active transport, however the MoAb provided the alternative route
`of cell entry. Similarly to CBL, 20-30 residues: of MEL could be coupled to
`MoAb (Table 2) .
`
`is an anthracycline analog and is 10
`Idarubicin (Fig. 6A)
`Idarubicin:
`times more cytotoxic than adriamycin. The l4—brcmo analog of idarubicin
`(Fig. 6B) was coupled to monoclonal antibodies using a reaction which could
`give rise to two types of linkage; by testing the stability to base, it was
`concluded that 50% of the drug was ester linked (Fig. 6D) and 3-5 residues
`of drug could be coupled with good recovery of antibody activity and
`protein (Table 2).
`
`0
`
`0H
`
`0
`
`R
`
`.lg. DH
`
`I]
`
`0
`
`OH
`
`\
`
`[3
`
`HD
`
`NH
`
`A IDARUBICIN
`
`R = H
`
`B BROMOIDARUBICIN
`
`R 2 Br
`
`C AHINF [INK
`
`D ESTER LINK
`
`R
`
`R
`
`NH-MoAb
`
`90-C-MOAb
`
`_
`FIGJRE 6
`Chemical structures of idarubicin (Ida) and possible linkages to antibodY-
`
`In Vitro Activi
`
`of the Dru -Antib
`
`Con 'u ates
`
`The pharmacologic activity of the free drug and MoAb bound drugS were
`tested in Vitro on cell lines by measuring the inhibition of DNA and RNA
`synthesis previously found to correlate well with cell death. The LYLE—:2
`cytotoxicity data for the various drug antibody conjugates is summarized
`(Tables 2 and 3) and there are several interesting points to note. Of all
`the drugfintibody conjugates, only Chlorambucil—antibody conjugates were
`more toxic than free drug (10 fold). Examination of the LB.
`for
`methotrexa’ce showed a 40 fold decrease in drug activity when bound to
`anti-WY: and aminopterin a 20 fold decrease in activity. However. th? .
`aminopterin conjugate was nearly as toxic as free methotrexate; emphasmlng
`the improvement in cytotoxicity with more potent analogs.
`In addition, the
`coupling of adriamycin to antibody using iodoacetyl adriamycin caused a 40
`fold decrease in cytotoxicity; however by using idarubicin, where the
`conjugate was coupled via the C-14 carbon, the resulting conjugate had
`similar cytotoxicity to that of the free drug. Melphalan, when modified to
`N-ACMEL was 25 times less active, however when this was coupled to antibody
`there was a lo fold increase in cytotoxicity.
`
`87
`
`IMMUNOGEN 2329, Pg. 11
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 11
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`mm 3
`In Vitro Cytotoxicity of Drug—Mb conjugates
`Drug
`Nlmber of drug
`Free drug
`molecules per
`IDS0 (M)l
`
` Ab molecule
`
`Drug—MoAb
`inmunoconjugate
`IDau (M)
`
`lnot done
`2.3xlO'1
`not done
`Adriamycin
`2.6x10-
`(anti-Ly—2.1)
`8.0x10-1
`53
`Iodoacetyladriamycin
`3.0:{10'1L (anti-'I'E'R)
`2. 2x10"1
`4
`Idarubicin
`1. 83:10”1 (250—30.6)
`8.21-ilO"1
`8
`Methotrexate
`8.4xlO'] (anti-Ly-2.l)
`4.2x10-1
`a
`Aminopterin
`1. 6x10—1 (anti-Ly-2.l)
`l.7xlO"1
`25-30
`Chlorambucil
`not done
`3.1xlO‘1
`not done
`Melphalan
`7.5xlO"l (anti-Ly-2.l)
`7.5xlCJ'1
`25-30
`N—acetyl melphalan
`1 Concentration at?—which 50% inhibition of the incorparation of —
`3H-—thymidine or 3H—uridine uptake relative to the control occurs.
`
`It was of interest that with certain canpomds, e.g. MTX and CBL (where
`the former was approximately 400 times more potent than CBL as free drug)
`the potency was similar in the imunoconjugate. The advantage of the
`alkylating agents is clearly the ability to bind large amounts of these to
`antibody with a resulting increase in potency. The drugs examined were
`mostly those in common usage, however our feeling is that other drugs,
`discarded because of their toxicity, would be worth investigating, as the
`toxicity is usually decreased when they are coupled to antibody to form an
`inmuncconj ugate .
`
`In Vivo Efficacy of Drug—Antibody Conjugates
`
`In the first the
`_In vivo models: Several different models were used.
`murine thymcma I‘I'I'(75)NS was grown in the 9:2 ocngenic strain either
`subcutaneously or in the peritoneum (B6.PL(75NS)) . This tumor is Ly—2.l‘
`but grows progressively, without rejection in C57BL/6 mice which are
`Ly—2.2+. Thus, the monoclonal anti-Ly—2.l antibodies are effectively
`"tumor specific" in that the antibody reacts only with the tumor and with
`no other normal tissues.
`In a second model human colon carcinoma cell
`lines were grown subcutaneously in nude mice and in a third, recently
`established model, fresh samples of colon carcinoma were growing
`subcutaneously in nude mice. All three models were used for preclinical
`studies and can also be used to examine the influence of the route of
`injection and the site of the tumor.
`.Adrialycin—Monb Conjugates: The in vivo efficacy of Ad—anti—Ly—2.l
`conjugates was tested in mice bearing I'I'I‘(l)75NS thymomas growing in the
`peritoneum. Groups of mice bearing established tumors were treated on days
`1, 3,
`5 and 7 with a total of léug of Ad in the conjugate.
`It was found
`that this increased the lifespan of 30% of mice by >200 days (Table 4);
`
`ERIE 4
`
`Efficacy of a variety of Drug—anti—Ly—ZJ conjugates on the survival of
`mice bearing the thymna I'1T(l)75bS E3.
`Treatment Schedule‘
`Days of survival
`(is survived)
`Total Dose
`Days after tumor
`PBS Drug MoAb Drug+ Drug—MoAb
`Drug
`conjugated administered( 9)
`inoculation
`MoAb
`conjugate
`Drug
`McAb
`0,1,2,3,7,l3
`6x4
`6x40
`CBL
`0,1
`2x15
`2x150
`MEL
`l,3,5,7
`4x4
`4x50
`AD
`l
`0.05
`0.22
`Ricin
`lAll treatments were administered i.p.
`
`48 3200/80% >200/80%
`35
`35
`>200/90%
`32
`>200/30%
`52
`>200/90%
`
`22
`30
`
`20
`25
`28
`55
`
`88
`
`IMMUNOGEN 2329, Pg. 12
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 12
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`when used with subcutaneously growing tumors,
`little effect
`(Fig. 7).
`
`the inmunoconjugates had
`
`3.8
`
`_._._.goingogo'm'4:inmo:oa
`
`.0 oo
`
`.04;
`
`1—4
`
`
`
`
`
`MeanTumorSize(cm?)
`
`0123456789101112
`1
`T
`T
`1
`Days after Tumor lnocuiation
`
`'
`FIGJFE 7
`Growth of the Ly—2.l+ thymcma I'I'I'(1)75NS in B6CF1 mice injected
`subcutaneously with 3 x 106 cells. Groups of 10 mice were given treatments
`i-P-
`(
`i )7 PBS ( l ), adraimycin ( o ), non conjugated mixture of
`adriamycin and anti-Ly—2.1 ( o ) and Ad—anti—Ly—2.l conjugate_( D )-_ Error
`bars represent t standard error of the mean tumor size. Reprinted w1th
`permission from Targeted Diagnosis and 'Iheragy (27) .
`
`dilorantucil—anti—Ly-ZJ conjugates: Mice with intraperitoneal tumors
`treated With 24>g Of CBL conjugate survived indefinitely (Table 4) I
`.
`althOUGh a non-covalent mixture of CBL and MoAb also increased the lifespan
`of 80% by >200 days. This is clearly a synergistic effect as similar doses
`of free drug and antibody did not have any effect when administered alone.
`In mice with subcutaneously growing tumors,
`those receiving CBL-MUN?
`ccnjugates had tumors significantly mailer than mice receiving a mixture
`of CBL and McAb or MoAb alone (Fig. 8).
`.
`Melphalan—lhab conjugates: Melphalan coupled to anti—Ly—2.l using the
`N-AcMEL derivative was tested in mice with the thymana growing in the
`peritoneal cavity (Table 4); 90% of the mice receiving the conjugate
`survived tumor free >200 days whereas all other mice died with tumor by day
`
`89
`
`IMMUNOGEN 2329, pg. 13
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 13
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`M
`
`\i—O
`
`
`
`MeantumorSIZE(cm7)
`
`
`
`
`
`HM H
`
`6
`9
`12
`0
`3
`15
`Days after tumor inoculation
`
`FIGZIRE 8
`injection of
`Growth of the thymcma I'IT(l)75NS E3 in BGCE‘1 mice given a 5.0.
`2 x 10" cells. Groups of 10 mice were given treatment i.p. denoted (
`i );
`PBS ( D ), free CBL ( I ), CBL—anti—Ly—2.l oonjguate ( o ), non-covalently
`conjugated CBL-anti—Ly—2.1 ( O ) and anti-Ly—2.l
`( A ). Error bars
`represent L standard error of the mean.
`
`The effect on solid tumors was tested in nude mice bearing COLO 205
`35.
`and using N—Acim—30.6 conjugates (60,19)
`in 10% of the mice,
`tumors were
`eradicated (Fig 9), and on day 28 the mean size of the tumors of the mice
`treated with conjugate was 50% that of mice in the control groups.
`However, while effective,
`the conjugates were limited by their cytotoxicity
`and consequently more toxic analogs were examined.
`
`In vitro experiments
`Amirnpterin and Methotrexatelkhnb conjugates:
`demonstrated AMN to be 10 times more cytotoxic than MTX and AMN-MoAb to be
`20 times more cytotoxic than MTX—McAb conjugates. Both conjugates were
`tested in vivo in nude mice bearing the COL0205 tumor,
`the total dose of
`AMN, either free or conjugated, was 35,9 and of MTX, 75,Jg.
`On day 19 the
`AMN conjugate treated tumors were 60% smaller than the MTX tumors although
`the dose was only half that of M'I'X (Fig. 10).
`The more toxic AMN analog is
`clearly more potent in vivo.
`
`Idarubicin—antibody caljugates: Colon carcinoma xenografts in nude
`mice were treated with a total of 275g of Ida given i.v. either as
`conjugate or free drug.
`It was noted that many mice receiving unconjugated
`Ida died due to tone effects (Fig ll): 80% of those receiving Ida alone,
`and 100% of those reoe1v1ng the mixture of Ida and 30.6 antibody. Of the
`group that received Ida-MoAb conjugate, 20% tumors were eradicated and the
`mean tumor size of this group was about 20% of the PBS treated mice on day
`20.
`
`90
`
`IMMUNOGEN 2329, Pg. 14
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2329, pg. 14
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`(cm?)
`MeanTumorSize
`
`O
`
`2
`
`4
`
`6 8101214161820222426283032
`ITTTTT
`1
`Days After Tumor Inoculation
`
`FIGTR‘E 9
`Tumor growth of OOLO 205 xenograft in nude mice. Groups of 10 mice bearing
`preexisti