`
`
`
`ROBERT K. OLDHAM, EDITOR
`
`
`
`
`
`
`
`IMMUNGEN 2307, pg. 1
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2307, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`Journals Manager
`Susan L. Patterson
`
`Managing Editor
`Rita S. Kessel
`
`Molecular Biotherapy is an international
`journal for the application of biologicals in
`clinical or veterinary practice. Authors are
`urged to study the notes on the inside back
`cover.
`
`Publishers, Editorial, Advertisement, and
`Reprint Offices: Butterworth Publishers, 80
`Montvale Avenue, Stoneham, MA 02180, USA.
`Telephone: (617) 438-8464. Telex: 880052.
`
`Publishing Director: C.J. Rawlins
`
`Molecular Biotherapy (ISSN 0952-8172) is
`published four times a year (March, June,
`September and December) by Butterworth
`Publishers, 80 Montvale Avenue, Stoneham,
`MA 02180. US POSTMASTER: Send address
`changes to Molecular Biotherapy,
`Butterworths, 80 Montvale Avenue, Stoneham,
`MA 02180.
`
`Subscription Orders/Inquiries: North America:
`Journal Fulfillment Department, Butterworth
`Publishers, 80 Montvale Avenue, Stoneham,
`MA 02180, USA. Telephone: (617) 438-8464.
`
`Annual subscription (4 issues): $45.00
`individuals, U.S. and Canada; $50.00
`individuals, elsewhere; $120.00 institutions,
`U.S. and Canada; $145.00 institutions,
`elsewhere; single copies $13.50 individuals,
`U.S. and Canada; $15.00 individuals,
`elsewhere; $35.00 institutions, U.S. and
`Canada; $43.00 institutions, elsewhere.
`
`Prices include packing and delivery by sea
`mail. Airmail prices available on request.
`Copies of this journal sent to subscribers in
`Bangladesh, India, Pakistan, Sri Lanka, Japan,
`and Europe are air-speeded for quicker
`delivery.
`
`International mailing agents: Mercury
`Airfreight International Ltd., Unitair Centre, Gt.
`Southwest Rd., Feltham, Mddx. TW14 8NJ, UK.
`
`Back issues: Available from Butterworth
`Publishers, 80 Montvale Avenue, Stoneham,
`MA02180.
`
`Reprints (minimum order 100): Contact
`Reprint Services, Sheridan Press, Fame
`Avenue, Hanover, PA 17331, USA, for details.
`
`Copyright: Readers who require copies of
`papers published in this journal may either
`purchase reprints or obtain permission to copy
`from the Publisher at the following address:
`Butterworth Publishers, 80 Montvale Avenue,
`Stoneham, MA 02180, USA. For readers in the
`USA, permission to copy, beyond that
`permitted by sections 107 and 108 of the US
`copyright laws, is given on the condition that
`libraries and other users registered with the
`Copyright Clearance Center (CCC)
`Transactional Reporting Service pay the base
`fee of $3.00 per copy directly to CCC, 21
`Congress Street, Salem, MA 01970 (telephone:
`(508) 744-3350), quoting the following:
`0952-8172189 $3.00 + 0.
`© 1989 Butterworth Publishers, a division of
`Reed Publishing (USA) Inc.
`All Rights Reserved. No part of this publication
`may be reproduced, stored in a retrieval
`system, or be transmitted, in any form or by
`any means (electronic, mechanical,
`photocopying, recording, or otherwise) without
`the written permission of the Publisher.
`
`Mol. Biother., 1989, vol. 1, no. 4
`
`MOLECULAR
`BIOTHERAPY
`
`THE INTERNATIONAL JOURNAL FOR THE APPLICATION
`OF BIOLOGICALS IN CLINICAL OR VETERrNARY PRACTICE
`
`VOLUME l • NUMBER 4 • 1989
`
`ROBERT K. OLDHAM, EDITOR
`
`Review
`
`186 Staphylococcal protein A adsorption in neoplastic disease:
`analysis of physicochemical aspects
`R.M. Murphy, C.K. Colton, and M.L. Yarmush
`
`Papers
`
`208
`
`Immunologic monitoring studies in advanced cancer patients
`treated with recombinant human gamma interferon (IFN-'Y4A)
`P.J. Romano, A. Lipton, H.A. Harvey, M.J. Bartholomew,
`G. Giudice, F. Kloszewski, L. Witkowski, and M.R. Downing
`
`213 Action of a thymic cytokine TsiF<® in reversing the
`autoimmune disease state of the MRL/1 pr mouse
`H. C. Miller and C. Vito
`
`218
`
`lmmunogenicity and tumor protective activity of 816
`melanoma vaccines
`D. Johnston and J-C Bystryn
`
`223 Cross-reacting tumor associated transplantation antigen on
`primary 3-methylcholanthrene-induced BALB/c sarcomas
`J.H. Coggin, Jr.
`
`229 Phase I trial of mitomycin C immunoconjugates cocktails in
`human malignancies
`D- Orr, R. Oldham, M. Lewis, J. Ogden, S-K Liao, K. Leung,
`S. Dupere, R. Birch, and B. A vner
`
`185
`
`IMMUNOGEN 2307, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`Phase I Trial of Mitomycin C
`Immunoconjugates Cocktails in
`Human·. Malignancies
`
`Douglas Orr, MD, *t Robert Oldham, MD, *t Marvin Lewis, MD, *t
`John Ogden, PhD,* Shuen-Kuei Liao, PhD,* Kiu Leung, PhD,*
`Sherry Dupere, PhD,* Robert Birch, PhD,§ and Belina Avner, BA*
`Williamson Medical Center,* Biological Therapy Institute, t National Biotherapy Study Group,§ and
`Biotherapeutics, Inc.,:f: Franklin, TN, USA
`
`Nineteen patients with refractory solid malignancies received individualized combinations of mitom~
`cin C conjugated murine monoclonal antibodies selected by immunohistochemical and flow cytometr~c
`screening of tumor specimens. There were no responses in this Phase I study. !~rombo_c~topema
`precluded escalation above a mitomycin C dose of60 mg per treatment cycle. Preclmzca~, clznzcal, and
`toxicity experiences with this investigational approach to the treatment of cancer are dzscussed.
`
`Keywords: Mitomycin C; immunoconjugates; antibody cocktails
`
`Introduction
`The signal achievement of hybridoma technology de(cid:173)
`scribed by Kohler and Milstein 1 has provided re(cid:173)
`searchers a virtually unlimited repertoire of molecular
`probes of high purity and unequaled specificity with
`which to characterize and purify macromolecules and
`to unscramble heretofore obscure biologic processes
`(e.g., immunologic networks). Unfortunately, with a
`few notable exceptions, 2- 4 the specificity afforded by
`murine monoclonal antibodies (MoAb) has not yet
`been successfully exploited in the treatment of cancer.
`Responses in hematologic malignancies to unconju(cid:173)
`gated murine MoAb have generally been transient and
`nonsustainable. 5•6 Attempts to utilize unmodified anti(cid:173)
`bodies recognizing tumor-associated antigens in the
`treatment of solid tumors have likewise been discour(cid:173)
`aging/ although antibody delivery to tumor deposits
`has consistently been demonstrable. 8•9 Antibodies
`have been utilized to serve as carriers of toxic agents
`to tumor cells. Examples include radionuclide immu(cid:173)
`noconjugates, t0- 13 immunotoxins, 4•14 and chemother(cid:173)
`apy-antibody conjugates. 15•16 These studies have dem(cid:173)
`onstrated
`that antibodies are,
`in general, well
`tolerated, that the risk of major allergic reaction to
`these foreign proteins on first exposure is small, that
`myelosuppression may accompany radioimmunocon(cid:173)
`jugates and chemoimmunoconjugates, that edema and
`a vascular leak syndrome may follow immunotoxin in-
`
`Address reprint requests to Dr. Douglas Orr, Biological Therapy
`Institute, P.O. Box 1676, Franklin, TN 37065-1676.
`Accepted for publication March 17, 1989.
`
`fusions, and that the emergence of human antimurine
`antibodies may limit ongoing treatments.
`The broad clinical experience with and the well(cid:173)
`characterized response and toxicity profiles of chemo(cid:173)
`therapeutic agents make them attractive as conjugates
`to antibodies. Drug immunoconjugates have been
`shown to retain both in vitro cytotoxicity and anti(cid:173)
`body specificity. 17 However, few clinical studies have
`as yet been reported with chemotherapy immunocon(cid:173)
`jugates. We have previously studied 23 patients who
`each received individualized cocktails of component
`doxorubicin-conjugated murine MoAb; results are re(cid:173)
`ported elsewhere. 15 Five minor responses (healing of
`skin ulcers, minor shrinkage of tongue nodules,
`shrinkage oflymphadenopathy in chronic lymphocytic
`leukemia) were observed, but there were no major re(cid:173)
`sponses. This demonstration of biologic activity, al(cid:173)
`though modest, did occur in clinical situations of
`bulky tumor and previous doxorubicin resistance. Ev(cid:173)
`idence for deconjugation of doxorubicin from anti(cid:173)
`body prior to tumor cell binding and internalization
`prompted us to investigate alternative conjugation
`methodologies and screen other possible chemothera(cid:173)
`peutic candidates. Another commonly used cytotoxic
`agent, mitomycin C, has been reported to demonstrate
`both in vivo and in vitro activity following conjugation
`to an anticolon cancer murine MoAb. 18 Our preclini(cid:173)
`cal studies also suggested potentially improved conju(cid:173)
`gation characteristics of the mitomycin C-antibody
`preparation. Like the doxorubicin-antibody linkage,
`the mitomycin C-antibody conjugate is acid labile
`and, following endocytosis, subject to cleavage within
`the acidic environment of the lysosome. In this study,
`individual tumor specimens were typed against a
`
`Mol. Biother., 1989, val. 1, no. 4
`
`© 1989 Butterworth Publishers 229
`
`IMMUNOGEN 2307, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`Papers
`
`panel of previously generated murine MoAb, which
`recognize tumor-associated antigens. Two to five anti(cid:173)
`bodies exhibiting in vitro binding to tumor were se(cid:173)
`lected for each patient. A fundamental hypothesis ex(cid:173)
`amined in this trial proposes that distinct permutations
`of antibodies individually specified for each patient
`will be needed to encompass a heterogeneic tumor cell
`population. In this report, we describe tumor process(cid:173)
`ing and antibody selection, conjugation methodology,
`and our clinical experience in 19 patients treated with
`individualized mitomycin C immunoconjugates.
`
`Materials and Methods
`Patient population
`Ten females and nine males, from 28 to 74 years of age
`(median age, 59 years) and with a variety of solid
`tumors, participated in this trial (Table 1). All patients
`were screened and counseled by a board-certified
`medical oncologist following referral upon failure of
`standard treatment modalities. Performance status
`(PS) distribution was as follows: PS 0 (n = 3); PS 1 (n
`= 8); PS 2 (n = 4); and PS 3 (n = 4). Seventeen pa(cid:173)
`tients had received prior chemotherapy (three prior
`mitomycin C); 13 prior radiotherapy and nine prior
`biotherapy. One of the patients with breast cancer had
`earlier received a doxorubicin immunoconjugate. A
`patient with cancer of the jejunum had 9 months pre(cid:173)
`viously been exposed to diagnostic murine antibody
`when administered an anti-CEA radiolabeled antibody
`for staging purposes. Duration of illness ranged from 1
`to 10 years, with a median of 3.3 years.
`Patients participating in this trial were required to
`have a confirmed diagnosis of metastatic cancer not or
`no longer amenable to standard therapy. Patients ex(cid:173)
`hibited normal renal (creatinine <2 mg/dl) and hepatic
`
`Table 1. Patient characteristics
`
`Patient no.
`
`Age (yr)
`
`Sex
`
`Tumor type
`
`1
`2
`3•
`4
`5
`6
`7
`8
`9
`10
`11 6
`12
`13
`14
`15
`16
`17
`18
`19
`
`68
`74
`37
`53
`61
`48
`59
`34
`64
`42
`63
`65
`57
`57
`36
`64
`27
`64
`59
`
`M
`F
`F
`F
`F
`F
`F
`M
`M
`M
`M
`M
`F
`M
`F
`M
`F
`F
`M
`
`Prostate
`Hepatoma
`Breast
`Ovary
`Breast
`Ovary
`Uterus (leiomyosarcoma)
`Lung
`Kidney
`Schwan noma
`Jejunum
`Prostate
`Uterus (carcinosarcoma)
`Head and neck
`Breast
`Colon
`Breast
`Colon
`Penis
`
`• Had previously received doxorubicin immunoconjugate (murine).
`b Had previously received diagnostic murine radioimmunoconjugate.
`
`(bilirubin <2 mg/dl) functions, full ambulatory capa(cid:173)
`bility, a life expectancy of greater than 3 months, ade(cid:173)
`quate hematologic parameters, and evaluable or mea(cid:173)
`surable disease. A history of irreversible mitomycin
`C-related toxicity (hemolytic/uremic syndrome, pul(cid:173)
`monary toxicity, unacceptable myelosuppression) pre(cid:173)
`cluded participation.
`
`Antibody generation, characterization,
`and purification
`
`The broad representation of tumor-associated anti(cid:173)
`gens among common solid tumors permits the system(cid:173)
`atic screening of tumor specimens from a panel of se(cid:173)
`lected murine MoAb. 19 These antibodies have been
`generated against adenocarcinomas (breast and colon)
`and melanomas using standard hybridoma methodolo(cid:173)
`gy (Table 2). The designation BT indicates an anti(cid:173)
`body developed within the laboratories of Biothera(cid:173)
`peutics, Inc.; the designation BA indicates antibodies
`acquired from outside sources. The hybridomas se(cid:173)
`creting MoAb BA-Br-1 (B38.1) and BA-Br-5 (B72.3),
`originally developed by Horan Hand et al. at the Na(cid:173)
`tional Cancer Institute,20 were provided by American
`Type Culture Collection, Rockville, MD. Characteris(cid:173)
`tics of MoAb BA-Br-3 (BTMA8), BA-Me-4 (140.72),
`and BA-Me-5 (140.240) have previously been pub(cid:173)
`lished. 21-23 A description of antibody-binding pat(cid:173)
`terns to selected tumors in vitro has been reported
`elsewhere. 19 Liao et al. observed that such binding
`crosses organ-specific boundaries, e.g., BA-Br-1
`bound not only to all 35 breast cancers examined in
`vitro but also to all 15 colon, 11 lung, 4 prostatic and 2
`pancreatic cancers tested. Each antibody in the panel
`used in this study has been extensively tested against
`normal adult tissues, including blood components
`(polymorphonuclear neutrophils, monocytes, lympho(cid:173)
`cytes, erythrocytes, platelets), kidney, lung, heart,
`breast, and skin. 19 Some binding within liver cells, to
`renal tubules, to lung tissue, and to sweat glands was
`observed, but to date no hepatotoxicity, nephrotoxi(cid:173)
`city, pulmonary toxicity, nor cutaneous injury has
`been attributed to any of this panel of antibodies. Mu(cid:173)
`rine MoAb were expanded in ascites of BALB/c mice
`or by means of a bioreactor and then submitted for
`purification using one or more of several methodolo(cid:173)
`gies [caprylic acid, ammonium sulfate, protein-A
`sepharose CL-4b, or high performance liquid chroma(cid:173)
`tography (HPLC)].
`
`Selection ofindividualized antibody cocktails
`Frozen tissue specimens (from fresh or cryopreserved
`primary or metastatic tumors) were provided the
`Monoclonal Antibody Division for assessment of
`binding by a panel of murine monoclonal antibodies.
`Five-micron-thick sections of frozen tissue embedded
`in OCT compound (polyvinyl alcohol, benzalkonium
`chloride, polyethylene glycol, d.H20; Miles Labs,
`Elkhart, IN), then cut and placed on gelatin-coated
`slides, were evaluated by means of the avidin-biotin-
`
`230
`
`Mol. Biother., 1989, vol. 1, no. 4
`
`IMMUNOGEN 2307, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`Mitomycin C immunoconjugates: D. Orr et al.
`
`Table 2. Murine MoAb used in this study
`
`MoAb
`
`lsotype
`
`Immunogen
`
`BA-Br-1
`BA-Br-3
`BA-Br-5
`BT-Br-6
`BT-Co-1b
`
`BT-Co-2
`
`BT-Co-3
`
`BT-Co-4
`
`BT-Co-5b
`
`BT-Co-6
`
`BA-Me-4
`BA-Me-5
`BT-Me-7
`BT-Me-8
`BA-Me-11
`
`lgG1
`lgG1
`lgG1
`lgG1
`lgG3
`
`lgG3
`
`lgG3
`
`lgG3
`
`lgG3
`
`lgG3
`
`lgG1
`lgG2a
`lgG1
`lgG1
`lgG1
`
`Membrane extract of breast carcinoma tissue
`Membrane extract of breast carcinoma cell line CAMA-1
`Membrane extract of breast carcinoma tissue
`Dispersed cells from breast carcinoma tissue
`Dispersed cells from colon carcinoma grown as xenografts
`in nude mice
`Dispersed cells from colon carcinoma grown as xenografts
`in nude mice
`Dispersed cells from colon carcinoma grown as xenografts
`in nude mice
`Dispersed cells from colon carcinoma grown as xenografts
`in nude mice
`Dispersed cells from colon carcinoma grown as xenografts
`in nude mice
`Dispersed cells from colon carcinoma grown as xenografts
`in nude mice
`Melanoma cell line CaCL 78-1
`Melanoma cell line CaCL 78-1
`Melanoma cell line BUR
`Melanoma cell line BUR
`Melanoma cell line
`
`Antigen structure
`
`NDa
`>300 kD glycoprotein
`220-400 kD glycoprotein
`ND
`29 kD + 31 kD protein
`
`ND
`
`ND
`
`ND
`
`29 kD + 31 kD + protein
`
`ND
`
`95-150 kD glycoprotein(s)
`p97-like (97 kD) glycoprotein
`110 kD protein
`110 kD + 40 kD protein
`280 kD + >400 kD proteoglycan
`
`• ND = Not yet defined, although attempts to determine the molecular mass of antigen involved were made.
`b Based on epitope blocking and indirect immunoprecipitation experiments. BT-Co-1 and BT-Co-5 recognized different epitopes present on the same or
`similar molecules.
`
`peroxidase complex technique (Vectastain ABC kit,
`Vector Laboratories, Burlingame, CA). Following fix(cid:173)
`ation with cold acetone, tissue sections were incubat(cid:173)
`ed with normal horse serum for 15 minutes to reduce
`nonspecific binding of horse anti-mouse IgG anti(cid:173)
`serum. Murine MoAb selected from the panel were
`then incubated at a concentration of 5 !J-g/ml with tis(cid:173)
`sue sections for 1 hour, washed with phosphate-buff(cid:173)
`ered saline (PBS) at pH 7.4, and incubated with bio(cid:173)
`tinylated horse anti-mouse IgG antiserum for 30
`minutes. Reaction with methanol and hydrogen perox(cid:173)
`ide for 30 minutes blocked endogenous peroxidase in
`tissue. Sections were then incubated with a preformed
`complex of avidin-biotinylated-horseradish peroxi(cid:173)
`dase for 1 hour, oxidized by 3-amino-9-ethylcarbazol
`(AEC) and 0.002% hydrogen peroxide for 15 minutes,
`and then counterstained in Gill's hematoxylin (Fisher
`Scientific, Norcross, GA) for 20 minutes. Positive
`(anti-HLA-A,B,C monomorphic determinant MoAb
`W6/32 and anti-epithelial membrane antigen MoAb)
`and negative (nonspecific murine IgG and PBS) anti(cid:173)
`body controls were also used. One or two sections
`from the frozen tissue block were routinely stained
`with hematoxylin and eosin to permit histopathologic
`confirmation of adequately preserved tumor cells. The
`immunohistochemical preparations were scored inde(cid:173)
`pendently by two observers, using a grading system
`based on intensity [from absent (0) to very strong
`( 4 +)] and the distribution and pattern of the reddish(cid:173)
`brown stain. A continuum from homogenous mem(cid:173)
`brane and/or cytosolic to patchwork staining was ob(cid:173)
`served. A designation " ±" was given when there
`were foci of positive antibody binding to tumor within
`a predominantly negative stain. By this means, anti-
`
`bodies were chosen as potential constituents of an im(cid:173)
`munoconjugate cocktail.
`Enzymatically or mechanically disaggregated
`tumor cells (when available) were subjected to flow
`cytometric analysis to determine the percentage of
`tumor cell coverage and binding intensity by immuno(cid:173)
`histochemically selected antibodies, either singly or in
`combination (Cytofluorograph liS, Ortho Diagnostic
`System, Inc., Westwood, MA). Tumor cells were pre(cid:173)
`pared enzymatically by incubating finely minced
`tumor tissue in dispersion solution containing RPMI
`1640 medium, 5% fetal bovine serum (Hyclone), and
`Worthington collagenase II, 31-25 units/10 ml/g tumor
`tissue. The dispersion was carried out at 37°C for 70
`minutes with continuous stirring. DNase (1,000 units,
`Sigma type I) was added for the final 10 minutes to
`increase filterability. Single cells were separated from
`undigested tumor mince using a 60-~J-m mesh nylon
`filter. Flow cytometry makes use of single cell suspen(cid:173)
`sions of tumor and fluoresceinated antimurine anti(cid:173)
`body tracers to quantitate the proportion of tumor
`cells bound by each selected murine MoAb and the
`relative intensity of that binding (i.e., reflecting the
`number of antibodies bound to positive cells), as indi(cid:173)
`cated by the mean peak channel (MPC). 24 The HLE-1
`antibody, which binds to leukocytes, corrected for any
`white blood cell contamination and propidium iodide
`exclusion staining for nonviable cells. Simultaneous
`antibody exposures were examined for possible addi(cid:173)
`tivity or interference among antibody combinations
`because one antibody might sterically hinder binding
`by another antibody recognizing a different epitope on
`the same or different antigen. Such antibodies would
`be better used separately.
`
`Mol. Biother., 1989, val. 1, no. 4
`
`231
`
`IMMUNOGEN 2307, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`Papers
`
`Selected antibodies were also screened against pa(cid:173)
`tient blood cells prior to their final incorporation into a
`treatment cocktail. Heparinized whole blood was cen(cid:173)
`trifuged at 800 x g for 20 minutes at 4°C; platelets
`were obtained from the plasma layer and washed
`twice in PBS containing 2% fetal calf serum (FCS) by
`centrifugation at 1 ,800 x g for 20 minutes. A mixture
`of lymphocytes, monocytes, and granulocytes was re(cid:173)
`covered from the buffy coat layer and then washed in
`PBS containing 2% FCS twice by centrifugation at 400
`x g for 5 minutes. The cells were dispensed into 96-
`well Falcon microtest plates (3 x 105/well for nucleat(cid:173)
`ed cells; 5 x 105/well for platelets). The plates were
`centrifuged at 400 x g to pellet the cells. MoAb sam(cid:173)
`ples (5 J.Lg/ml) were added (100 J.Lg/well) and incubated
`for 1 hour at room temperature. Cells were pelleted,
`supernatants were removed, and cells were washed in
`PBS plus 2% FCS three times, pelleting between each
`wash. Fluoresceinated goat anti-mouse Ig antiserum
`(Coulter Immunology, Hialeah, FL) at an optimized
`concentration (I 00 J.Lg/well) was added and incubated
`for 45 minutes at 4°C. Cells were repelleted, superna(cid:173)
`tants were again removed, and cells were washed two
`times as above, pelleting between each wash. Normal
`mouse IgG (5 J.Lg/ml) and PBS were used as negative
`controls. MoAb W6/32 (anti-HLA-A,B,C monomor(cid:173)
`phic determinant) was used as a positive control.
`Platelets and buffy coats (containing lymphocytes,
`monocytes, and granulocytes) were gated from a for(cid:173)
`ward angle at 90°, and light scatter patterns were ana(cid:173)
`lyzed independently for fluorescence.
`
`Immunoconjugate preparation and
`preclinical evaluation
`The chemotherapeutic agent mitomycin C was conju(cid:173)
`gated via an N-hydroxysuccinimide · (NHS) active
`ester intermediate to amine groups (provided by ly(cid:173)
`sine, arginine, asparagine, glutamine, and NH2-termi(cid:173)
`nal amino acids) displayed by the antibody protein
`(Figure 1). The conjugation procedure, modified from
`Kato et al., 25 entailed the addition of a glutaric acid
`spacer arm to the imine group of mitomycin C, fol(cid:173)
`lowed by its conversion to an active ester by treat(cid:173)
`ment with NHS and dicyclohexylcarbodiimide (DCC).
`The resulting compound was crystallized and stored
`dessicated at - 20°C. Conjugation was accomplished
`by dissolving 86 mg of this active ester in dimethylfor(cid:173)
`mamide (DMF), which was added to 1 g of MoAb and
`borate buffer at pH 8.5 so that the DMF concentration
`was 10%. Free mitomycin C was removed by tangen(cid:173)
`tial flow ultrafiltration (Minitan) and the conjugate
`mixed with mannitol and stored in the dark at 10°C.
`High yield conjugation was confirmed by the HPLC
`(Waters HPLC with two Protein Pak 300 SW columns
`in series) demonstration of superimposable peaks of
`antibody and mitomycin C, with less than 2% free
`drug. Preservation of mitomycin C antitumor activity
`despite conjugation was documented by exposing the
`active ester to water and determining equivalence in
`cell-killing capacity between free drug and the hydro-
`
`0
`
`Figure 1. Chemical structure of mitomycin C.
`
`lyzed active ester. In vitro cytotoxicity was assayed
`by standard techniques employing tritiated thymidine
`uptake by tumor cells. To insure safety, the immuno(cid:173)
`conjugate preparation, prior to clinical use, underwent
`endotoxin screening (LAL chromagenic assay), steril(cid:173)
`ity testing (trypticase soy broth (aerobes, fastidious),
`thioglycolate broth (anaerobes, aerobes), Sabourad's
`dextrose agar (fungi)) and general safety testing (in
`guinea pigs and mice). Most clinical specimens con(cid:173)
`tained less than 1.0 endotoxin unit (EU)/ml, but
`ranged as high as 2.8 EU/ml. Immunoconjugate prepa(cid:173)
`rations were examined by HPLC for the quantity of
`free mitomycin-C, using absorbance at OD280 for anti(cid:173)
`body and OD363 for mitomycin-C.
`
`Administration of immunoconjugates
`
`Patients participating in this clinical trial provided
`written informed· consent, following a thorough dis(cid:173)
`cussion of potential risks and toxicities of participa(cid:173)
`tion and identification of alternative therapies. This
`Phase I clinical trial was performed under a protocol
`approved by the Investigational Research Review
`Committee of the Williamson Medical Center (Frank(cid:173)
`lin, TN). Eighteen patients received treatment at the
`Williamson Medical Center; one patient was treated at
`Cottage Hospital, in Santa Barbara, CA, under com(cid:173)
`passionate Investigational New Drug (IND) authoriza(cid:173)
`tion. Patients received 300 mg allopurinol daily begin(cid:173)
`ning 12 hours prior to initiation of immunotherapy and
`continuing until completion of treatment. The immu(cid:173)
`noconjugate was prepared in 100 ml of normal saline
`and administered over 4-6 hours through a central
`venous line. Patients received a test dose of immuno(cid:173)
`conjugate, consisting of one to three drops of the
`preparation, followed by an observation period of 15
`minutes. If there were no untoward side effects, the
`infusion resumed at a rate of approximately 20 ml/h.
`For patients who developed mild pulmonary toxicity,
`infusions were slowed to 1-10 mg of antibody per
`hour. Emergency support for anaphylaxis was avail(cid:173)
`able at the bedside. Mild erythema prompted tempo(cid:173)
`rary interruption of the infusion and administration of
`100 mg Solu-Cortef intravenously. If the rash respond(cid:173)
`ed, the infusion was resumed, with an additional 100
`mg Solu-Cortef at 2 hours if indicated. Vital signs
`were monitored every 15 minutes during the first hour
`of infusion, every 30 minutes for the next 2 hours, and
`then every hour until 2 hours postinfusion. Immuno(cid:173)
`conjugates for the first cohort of patients, based upon
`the mitomycin C component, were administered on
`alternate days according to the following scheme: 5
`
`232
`
`Mol. Biother., 1989, val. 1, no. 4
`
`IMMUNOGEN 2307, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`mg, 10 mg, 10 mg, 15 mg (total of 40 mg). Total doses
`were to be escalated in 20-mg mitomycin C incre(cid:173)
`ments (i.e., to 60 mg, then 80 mg, and so on) as per(cid:173)
`mitted by toxicity experience.
`
`Pharmacokinetics and antibody targeting
`Whenever possible, accessible tumor deposits (such
`as from skin or peripheral lymph nodes) were excised
`posttreatment to permit assessment of antibody deliv(cid:173)
`ery and saturation of tumor deposits (using the avi(cid:173)
`din-biotin-peroxidase complex technique). Tissues
`were considered to be saturated if no enhancement of
`staining was noted following the in vitro application of
`the corresponding murine MoAb.
`In two patients, pharmacokinetic analyses of mu(cid:173)
`rine antibody were performed. Pharmacokinetic pro(cid:173)
`files for murine MoAb in sera were obtained by
`a double-determinant ELISA technique. Enzyme
`immunoassay plates were coated with sheep anti(cid:173)
`mouse IgG (Cappel Division of Organon Teknika
`Corp., West Chester, PA) at 10 f.Lg/ml in 0.05 M sodium
`carbonate, pH 9.6, overnight at 4°C. Plates were
`blocked with 5% bovine serum albumin in PBS for 1
`hour at room temperature and washed 5 times with
`0.2% Triton X-100/PBS. Patient sera at dilutions of
`1:10 and 1:100 in PBS and standards (murine MoAb,
`Ba-Br-1 at 1-50 pg/ml) were added to plates and incu(cid:173)
`bated at room temperature. After washing five times,
`tracer antibody (goat anti-mouse IgG-peroxidase, Or(cid:173)
`ganon Teknika) at 1:10,000 in wash solution was added
`and the plate incubated for 2 hours at room tempera(cid:173)
`ture. After washing, substrate solution [1 mg/ml a(cid:173)
`phenylenediamine and 0.04% hydrogen peroxide (30%
`solution in citric acid phosphate buffer, pH 5)] was
`added. The reaction was terminated after 15 minutesin
`the dark by the addition of 50 f.Lg/well 0.45 M sulfu-ric
`acid stop solution. Absorbance plates were mea(cid:173)
`sured at 490 nm using a Dynatech MR 600 microplate
`reader (Dynatech Laboratories, Chantilly, VA). Se(cid:173)
`rum concentrations of antibody were determined by
`linear regression analyses from a standard curve.
`
`Human antimurine antibody
`(HAMA) assessment
`Human IgG antibodies to mouse IgG were measured
`by means of the ImmunoSTRIP HAMA ELISA assay
`(Immunomedics, Inc., Warren, NJ). Samples were di(cid:173)
`luted 1:10 prior to testing; the limit of sensitivity of
`this assay is 37 ng/ml. We did not attempt to measure
`either IgM or IgE HAMA.
`
`Clinical monitoring and toxicity analysis
`The following pretreatment studies were required: his(cid:173)
`tory and physical examination, CBC with differential
`and platelet count, PT, aPTT, urinalysis, serum com(cid:173)
`plement, SMA-18, chest X-ray, EKG, appropriate CT
`scans and tumor markers, and an immunocompetency
`profile. Six hours after starting each infusion, a repeat
`
`Mitomycin C immunoconjugates: D. Orr et al.
`
`CBC with differential and platelet count was obtained.
`At 24 hours following infusion the following tests were
`repeated: CBC with differential, platelet count, urin(cid:173)
`alysis, serum complement, SMA-18, and an immuno(cid:173)
`competency profile. The following tests were thereaf(cid:173)
`ter done weekly for 4 weeks: CBC differential, platelet
`count, urinalysis, SMA-18. These studies were ob(cid:173)
`tained monthly for 3 months following treatment: his(cid:173)
`tory and physical, CBC with differential, platelet
`count, urinalysis, SMA-18, appropriate CT scans and
`tumor markers, and an immunocompetency profile.
`Blood samples for antibody pharmacokinetic studies
`and human antimurine antibody (HAMA) analysis
`were obtained from selected patients.
`Patients were monitored closely for possible aller(cid:173)
`gic reactions, including fever, urticaria, broncho(cid:173)
`spasm, anaphylaxis, Arthus reaction, vasculitis, or
`serum sickness. In addition to the side effects asso(cid:173)
`ciated with free mitomycin C, close attention was
`given to possible unique toxicities from antibody traf(cid:173)
`ficking to selected organs, particularly those compris(cid:173)
`ing the reticuloendothelial system (RES). Toxicities
`were graded on a scale of 1 to 4 as per standard Na(cid:173)
`tional Biotherapy Study Group (NBSG) criteria.
`Guidelines for immunoconjugate dose escalations
`were strictly delineated by protocol.
`
`Results
`In vitro and in vivo analysis
`of immunoconjugates
`Immunoreactivity. Selected murine MoAb were test(cid:173)
`ed by flow cytometry against corresponding tumor
`targets, prior to and following immunoconjugation, to
`determine retained immunoreactivity (Table 3). The
`data suggest that conjugation reduced moderately the
`percentage of tumor cells to which antibody attached
`and modestly the number of MoAb bound per tumor
`cell, as demonstrated by MPC.
`
`Molar ratios. Immunoconjugates used in this study
`exhibited mitomycin C:murine monoclonal antibody
`molar ratios (MR) from 2.5 to 15.79 at the time of their
`preparation.
`
`Purity, stability, and reproducibility. Table 4 depicts
`BA-Br-1 immunoconjugate molar ratios from sequen(cid:173)
`tial procedures, demonstrating good reproducibility.
`Table 4 also displays the percent conjugated mitomy(cid:173)
`cin c lost daily when the preparation is stored at 10°C.
`The 1-week requirement for safety testing following
`immunoconjugate formulation is accompanied by a
`15-20% loss of conjugated mitomycin C. The bulk of
`free mitomycin C is removed by ultrafiltration imme(cid:173)
`diately prior to administration.
`
`Specificity and cytotoxicity. Specific and indifferent
`tumor targets were exposed in vitro to selected im(cid:173)
`munoconjugates to evaluate relative cytotoxicity. Fig(cid:173)
`ure 2 reveals comparable cytotoxicities (as determined
`
`Mol. Biother., 1989, vol. 1, no. 4
`
`233
`
`IMMUNOGEN 2307, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`
`
`Papers
`
`Table 3. Flow cytometric evaluation of unconjugated and conjugated antibodies
`
`Antibody
`
`Concentration (fLg/ml)
`
`%Positive
`
`MPC
`
`%Positive
`
`MPC
`
`Unconjugated antibody
`
`Conjugated antibody
`
`BA-Br-1 8
`
`BA-Br-38
`
`BT-Co-1b
`
`BT-Co-6b
`
`BA-Me-4b
`
`• BT-20 target.
`bLS-174Ttarget.
`
`10
`5
`1
`10
`5
`1
`10
`5
`1
`10
`5
`1
`10
`5
`1
`
`76.9
`69.1
`33.5
`64.9
`67.4
`39.2
`53.0
`43.2
`7.0
`85.5
`76.5
`63.8
`38.9
`35.6
`17.2
`
`80
`75
`62
`74
`76
`64
`65
`60
`47
`83
`73
`66
`55
`55
`50
`
`66.4
`47.5
`26.0
`71.3
`50.4
`39.6
`23.7
`22.7
`18.4
`62.6
`54.2
`50.0
`29.9
`19.3
`2.7
`
`74
`67
`59
`77
`69
`65
`53
`52
`50
`67
`62
`59
`53
`50
`46
`
`determined an LD50 for free mitomycin C o